Skip to main content

Full text of "Archives Of Biochemistry Vol-xxiii-xxiv (1949)"

See other formats


Indian Agricultural 
Research 1 Institute, New Delhi. 


L A. R. L 6. 

MGIPC-S4-10 ARs—21-6-49—1,000. 





ARCHIVES 

OF 

BIOCHEMISTRY 


EDITORS 


E. S. G. Barr6n 

A. Neuberger 

G. 0. Burr 

F. F. Nord 

M. L. Crossley 

H. Theorell 

M. A. Lauffer 

K. V. Thimann 

K. Linderstr0m-Lang 

C. H. Werkman 

C. M. McCay 

R. J. Williams 

VOLUME 




1949 

ACADEMIC PRESS INC. PUBLISHERS 
NEW YORK, N. Y. 



Copyright 1949, by Academic Press Inc. 
Made In United States of America 



CONTENTS OF VOLUME 23 

No. 1, August, 1949 

S. Lassen, E. K. Bacon and II. J. Dunn. The Digestibility of 

Polymerized Oils . 1 

1). Siminovitch and D. R. Briggs. The Chemistry of the Living 
Bark of the Black Locust Tree in Relation to Frost Hardiness. 

I. Seasonal Variations in Protein Content. 8 

D. R. Briggs and D. Siminovitch. The Chemistry of the Living 

Bark of the Black Locust Tree in Relation to Frost Hardiness. 

II. Seasonal Variations in the Electrophoresis Patterns of the 

Water-Soluble Proteins of the Bark. 18 

M. Frank Mallette and Charles R. Dawson. On the Nature 

of Highly Purified Mushroom Tyrosinase Preparations. 29 

D. C. Carpenter and William C. Smith. Separation of a Crys¬ 
talline Globulin from Tomato Juice and Determination of Its 

Isoelectric Point . 45 

Robert C. Ottke. The Occurrence of Ergosterol in Neurospora 

crassa . 49 

Chou Hao Li, Miriam E. Simpson and Herbert M. Evans. The 
Influence of Growth and Adrenocorticotropic Hormones on 

the Fat Content of the Liver. 51 

A. C. Paladini, R. Caputto, L. F. Leloir, R. E. Trucco and 
C. E. Cardini. The Enzymatic Synthesis of Glucose-1,6- 

Diphosphate . 55 

Harry Rudney. Studies on the Mechanism of the Inhibition of 

Glucolysis by Glyceraldehyde. 67 

Gilberto G. Villela, Mario Vianna Dias and Laura T. Que- 

iroga. Distribution of Thiamine in the Brain. 81 

Dorothy S. Genghof. The Sulfur Amino Acid Requirement of 

Tetrahymend geleii . 85 

IIalvor N. Christensen and Jean A. Streicher. Concentration 
of Amino Acids by the Excised Diaphragm Suspended in 
Artificial Media. I. Maintenance and Inhibition of the Con¬ 
centrating Activity . 96 

IIalvor N. Christensen, Mary K. Cushing and Jean A. 
Streicher. Concentration of Amino Acids by the Excised 
Diaphragm Suspended in Artificial Media. II. Inhibition of 
the Concentration of Glycine by Amino Acids and Related 

Substances . 106 

Klas-Bertil Augustinsson. Substrate Concentration and Speci¬ 
ficity of Choline Ester-Splitting Enzymes. Ill 

















Louis E. Wise. The Polysaccharide from lies mannane . 127 

Kurt I. Altman, George W. Casarett, T. R. Noonan and K. 
Salomon. The Distribution in Rat Tissues of the Methylene 

Carbon Atom of Glycine Labeled with C u . 131 

Frederick C. Bauer, Jr. and Edwin F. Hirsch. Esterified Fatty 

Acid Levels of Normal Human Sera. 137 

Daniel I. Arnon and F. R. Whatley. Factors Influencing Oxy¬ 
gen Production by Illuminated Chloroplast Fragments. 141 

Letters to the Editors : 

Samuel Natelson, Julius K. Lugovoy and Joseph B. 
Pincus. A New Fluorometric Method of Epinephrine 157 


Hempstead Castle and Flora Kubsch. The Production of 
Ilsnic, Didymic, and Rhodocladonic Acids by the Fungal 


Component of the Lichen Cladonia cristatella . 158 

D. J. D. IIockenhull, K. Ramachandran and T. K. Walker. 

The Biosynthesis of the Penicillins. 160 

Joseph F. Nyc, Francis A. Haskins and Hersctiel K. Mit¬ 
chell. A Metabolic Relationship between the Aromatic 

Amino Acids . 161 

Dominick Papa. X-Ray Diagnostic Agents—2-(3,5-Diiodo- 

4-Hydroxybenzyl)-Benzoic Acid . 163 

Arthur M. Hartman, Leslie P. Dryden and Charles A. 
Cary. A Role of Vitamin B 12 in the Normal Mammal .. 165 
Book Reviews . 169 


No. 2, September, 1949 


Eric Conn, Birgit Vennesland and L. M. Kraemer. Distribu¬ 
tion of a Triphosphopyridine Nucleotide-Specific Enzyme 
Catalyzing the Reversible Oxidative Decarboxylation of Malic 

Acid in Higher Plants. 179 

Victor Schocken. The Genesis of Auxin during the Decompo¬ 
sition of Proteins. 198 

Bernard S. Gould, Henry M. Goldman and John T. Clarke, Jr. 
Antiscorbutic Substances. 3-Methyl-L-Ascorbic Acid and 1- 

Methylheteroascorbic Acid . 205 

Albert A. Dietz. Chemical Composition of Normal Bone Marrow 211 
Albert A. Dietz and Bernhard Steinberg. Chemical Composi¬ 
tion of Irradiated Bone Marrow. 222 

Virgil L. Koenig. Ultraceutrifugal Studies on Some Porcine 

Plasma Protein Fractions . 229 

W. R. Ruegamer. Occurrence of an Unidentified Rat Growth 

Factor in Cottonseed Meal. 236 

Harry J. Deuel, Jr., Samuel M. Greenberg, Evelyn Straub, 
Tomoko Fukui, A. Chatterjee and L. Zechmeister. Ste- 


















reochemical Configuration and Provitamin A Activity. VII. 

Neocryptoxanthin TJ . 239 

L. Zechmeister, J. II. Pinckard, Samuel, M. Greenberg, Evelyn 
Straub, Tomoko Pukui and IIarry ,T. Deuel, Jr. Stereo¬ 
chemical Configuration and Provitamin A Activity. VIII. 
Pro-Carotene (a Poly-cis Compound) and Its All-trans 

Isomer in the Rat. 242 

Otto Meyerhof and Jean 11. Wilson. Comparative Study of 
the Glycolysis and ATP-ase Activity in Tissue Homogenates 246 
P. G. Stansly and N. H. Ananenko. Candidulin : An Antibiotic 

from Aspergillus candidus . 256 

Max A. Lauffer and Martha Wheatley. Destruction of Influ¬ 
enza A Virus Infectivity by Formaldehyde. 262 

Ines Mandl and Carl Neuberg. An Unknown Effect of Amino 
Acids. II. Interaction of Nitrogenous Polycarboxylic Acids 
(N-Substituted Amino Acids) and Insoluble Metal Sulfides 

and Mercaptides . 271 

Roger J. Williams, L. Joe Perry and Ernest Beersteciier, Jr. 
Biochemical Individuality. III. Genetotrophic Factors in 

the Etiology of Alcoholism . 275 

Marie A. Fischer and Max A. Lauffer. The Reaction of To¬ 
bacco Mosaic Virus with Formaldehyde. I. Electrophoretic 

Studies. 291 

James Franck. An Interpretation of the Contradictory Results 
in Measurements of the Photosynthetic Quantum Yields and 

Related Phenomena . 297 

E. Geiger, E. B. Hagerty and H. D. Gatchell. Transformation 
of Tryptophan to Nicotinic Acid Investigated with Delayed 

Supplementation of Tryptophan . 315 

Letters to the Editors : 

Bernard Siiacter. Increase in Yeast Respiration in Pres¬ 
ence of Several Steroids and Diethylstilbestrol. 321 

J. J. Oleson and J. C. Van Meter. Inhibition of Ergo- 
stanyl Acetate by 7-Dehydrocholesteryl Bromide. 323 


T. J. Cunha, J. E. Burnside, D. M. Busciiman, R. S. Glass¬ 
cock, A. M. Pearson and A. L. Siiealy. Effect of Vita¬ 
min B IS , Animal Protein Factor and Soil for Pig Growth 324 
Esmond E. Snell and Harry P. Broquist. On the Prob¬ 
able Identity of Several Unidentified Growth Factors .. 326 
J. E. Burnside, T. J. Cunha, A. M. Pearson, R. S. Glass¬ 
cock and A. L. Shealy. Effect of APF Supplement' on 


Pigs Fed Different Protein Supplements. 328 

Otto Warburg, Dean Burk, Victor Schocken, Mitchell 
Korzenovsky and Sterling B. Hendricks. Does Light 
Inhibit the Respiration of Green Cells?. 330 
















W. G. Frankenburg and A. M. Gottscho. Myosmine in 


Cigar Tobacco . 333 

Book Reviews . 336 


No. 3, October, 194!) 


Grant N. Smith and Cecilia S. Worrei,. Studies on the Action 
of Chloramphenicol (Chloromycetin) on Enzymatic Systems. 

1. Effect of Chloramphenicol on the Activity of Proteolytic 

Enzymes . 341 

Simon Black. A Microanalytical Method for the Volatile Fatty 

f Acids .’. 347 

William C. Day, Michael J. Pklozar, Jr. and Sidney Gottlieb. 

The Biological Degradation of Lignin. I. Utilization of 
Lignin by Fungi . 360 


Margaret E. Greig and William C. Holland. Studies on the 
Permeability of Erythrocytes. I. The Relationship between 
Cholinesterase Activity and Permeability of Dog Erythrocyte 370 
Aurin M. Chase. Studies on Cell Enzyme Systems. 11. Evi¬ 
dence for Enzyme-Substrate Complex Formation in the Re¬ 


action of Cypridina Luciferin and Lueiferase. 385 

Georg Cronheim, Mary E. Baird and Warren N. Dannenburg. 

Studies with Penicillinase in the Presence of Sulfonamides 394 
Saul H. Rubin and Jacob Sciieiner. Antibiotin Effect of 

Ilomologs of Biotiu and Biotin Sulfone. 400 

Anthony J. Glazko, Loretta M. Wole and Wesley A. Dill. 
Biochemical Studies on Chloramphenicol (Chloromycetin). 

I. Colorimetric Methods for the Determination of Chlo¬ 
ramphenicol and Related Nitro Compounds. 411 

S. C. Pan, A. A. Andreasen and Paul Kolaciiov. Factors In¬ 
fluencing Fat Synthesis by llhoilotorula gracilis . 419 

J. A. Muntz, E. S. Guzman Barron and C. L. Prosser. Studies 
on the Mechanism of Action of Ionizing Radiations. III. 

The Plasma Protein of Dogs after X-Ray Irradiation. An 
Electrophoretic Study . 434 


Virginia Marie Smith. On the Mechanism of Enzyme Action. 
XXXIX. A Comparative Study of the Metabolism of Carbo¬ 
hydrates, in the Presence of Inorganic and Organic Phos¬ 
phates by Mcrulius lacrymans and Marasmius chordalis .... 446 
Josedh V. Fiore and F. F. Nord. Lipase Determinations with 


the Aid of Polyvinyl Alcohol. 473 

F. F. Nord, J. V. Fiore, G. Kreitman and S. Weiss. On the 
Mechanism of Enzyme Action. XL.' The Interaction of Sola- 
nione, Riboflavin, and Nicotinic Acid in the Carbohydrate 
Fat Conversion by Certain Fusaria . 480 















Letters to the Editors : 

Adrian M. Srb. Lack of Effect of (larbamyl-L-Glutamic 
Acid on the Growth of Certain Arginineless Mutants of 


Neurospora . 495 

Hans Burstrom. n-Diamylacetic Acid and Nitrate Assimi¬ 
lation . 497 

Carl Neuberg and Inks Mandl. Functions of ATP and 

Other Phosphoric Acid Derivatives. 499 

Donald J. Hanaiian, N. B. Everett and C. D. Davis. Fate 
of S 3r, -Na Estrone Sulfate in Pregnant and Non-Pregnant 

Rats . .....501 

II. Lehr and J. Berger. The Isolation of a Crystalline 

Actinomycin-like Antibiotic . 503 

B. David Polis, Edith Pons, Maxinf. Kerrigan and Lillian 
Jedeikin. Effect of Insulin and Adenosinetriphosphatase 
on a Reaction Coupling Oxidation with Phosphorylation 505 


G. Poimak, S. J. Folley and T. H. French. Synthesis of 
the Short-Chain Fatty Acids of Milk Fat from Acetate 508 
T. J. Cuniia, II. II. Hopper, J. E. Burnside, A. M. Pearson, 

R. S. Glasscock and A. L. Siiealy. Effect of Vitamin 
B t , and APF Supplement on Methionine Needs of the 


Pig . 510 

Ciioh Hao Li, C. Kalman and II. M. Evans. The Effect 
of Adrenocorticotropic and Growth Hormones on the 
Glucose Uptake and Glycogen Synthesis by the Isolated 

Diaphragm with and without Insulin. 512 

Author Index. 515 

Subject Index . 520 

Index op Book Reviews . 525 














The Digestibility of Polymerized Oils 

S. Lassen, E. K. Bacon and H. J. Dunn 

From Van Camp Laboratories , Terminal Island , California 
Received August 9, 1948 

Introduction 

Edible polymerized oils have, for many years, been used in small 
amounts in the food industry. In view of the recent increase in the use 
of these oils in foods (1, 2, 3), it was thought of interest to study their 
chemical and biological properties. 

By suitable polymerization, unsaturated glycerides usually lose their 
characteristic odor and taste. As polymerization progresses, the 
viscosity of the oil increases, and the amount of unsaturation is de¬ 
creased as measured by the decrease in iodine value. The emulsifying 
qualities of the oil also change characteristically as the molecular 
weight is increased. According to Bradley (4), the main structural 
changes that unsaturated glycerides undergo upon polymerization 
seem to involve their double bonds. 

More recently, Brocklesby (5) has suggested that, during the polymerization of 
unsaturated glycerides, the most likely course is for two double bonds to react to 
form an unstable 4-carbon ring which breaks open and rearranges as follows: 

R-CH-CH-R' R CH—CH R' RCH 2 CII*R' 

+ - * I ! - > I 

R-CH = CH-R' R CH-CH R' R-C-CH-R' 

This type of polymerization, according to Brocklesby, might take place between two 
unsaturated fatty acids both attached to the same glyceride molecule (intramolecu¬ 
lar) or between fatty acids of different glyceride molecules (intermolecular). Our 
results suggest that this reaction is more likely to occur between fatty acids of differ¬ 
ent glycerides than within the same glyceride. Furthermore, Brocklesby's theory for 
intramolecular ploymerization involves the elimination of two double bonds between 
adjacent fatty acid molecules within the same triglyceride, thereby forming a 4- 
carbon ring. Since this ring, as visualized by Brocklesby, is unstable, a shift in the 
position of one fatty acid chain with respect to the other takes place, resulting in the 
reformation of one of the two double bonds involved in the ring formation. The molec¬ 
ular structure of triglycerides makes such a change unlikely, as this undoubtedly 
would tend to rupture the glyceride molecule. 

1 



2 


8. LASSEN, E. K. BACON AND H. J. DUNN 


A considerable amount of research has been done on the polymerization of the 
methyl esters of unsaturated fatty acids (7, 8) and glycerides (4,5, 6). The literature 
appears to contain few references to the biological effect of thermal treatment of 
edible oils (9). 

To investigate this problem, a typical unsaturated oil was selected, 
namely, sardine oil (California pilchard, Sardinops caerulea). This oil 
is used, to a large extent, in animal nutrition as a carrier of vitamin A 
and D oils and is a natural constituent of fish meal, which has an 
extensive use in animal nutrition. 

Expebimental 

A sample of refined dcstearinated sardine oil was subjected to poly¬ 
merization by heating it in a glass vessel at 250°C. in a nitrogen atmos¬ 
phere. By taking samples of the oil at regular intervals, and checking 
its iodine value and refractive index, the degree of polymerization was 
followed. Three samples of polymerized oil were produced having 
iodine values (Hanus) of approximately 160,140, and 120, respectively. 
These samples showed only a moderate increase in free fatty acid over 
the original oil. They were, however, all subsequently alkali-refined, 
and this brought the free fatty acid content of all of them down below 
0.5%. The samples were stored under nitrogen in a refrigerator until 
ready for use. The sardine oil, the samples of polymerized sardine oil 
produced therefrom, and a sample of commercially available edible 
polymerized herring oil, were subjected to a detailed analysis as shown 
in Table I. 

To determine the effect polymerization has upon the biological value 
of sardine oil, a series of animal experiments were undertaken. It was 
decided to use rats as the experimental animal and to determine the 
coefficient of digestibility of the polymerized oils by the fat balance 
method. The details of the experiments were as follows: 

Mature albino rats of the Sprague-Dawley strain weighing approximately 275 g. 
were divided into groups of 10 each, balanced according to sex and weight, and 
placed in individual metabolism cages so adapted that the urine and feces could be 
collected separately. The experiment was divided into 3 feeding periods, a prelimin¬ 
ary depletion period of 4 days, followed by a test period of 7 days, and a final 3 day 
after period. During the depletion period all of the rats received a fat free basal diet 
ad libitum. During the 7 day test period the rats received the basal diet supplemented 
with 5% of the various oils under study. During this test period an accurate account 
was kept of the food intake, the various groups being restricted to 15 g. of diet/day/ 
rat. During the after period the rats were again given the fat free basal diet ad libitum. 



DIGESTIBILITY OF POLYMERIZED OILS 


3 


TABLE I 

Effect of Polymerization on Physical Constants of Samples Tested 


Sample ^ 

Iodine ( 

Per cent 1 
unsaponi- 1 
Sable 
fraction | 

Refractive 

index 

Per cent 
free fatty 
acid 

Molecular weight® 

no. 1 
Hanua 

27°C. 

Ni> 

Whole 

oil 

Fatty 

acids 6 

Control. Sardine oil 
Polymerized sardine oil 

177.7 

1.61 

1.4800 

0.19 

828 

286 

first fraction 

Polymerized sardine oil 

155.5 

2.09 

1.4830 

0.15 

942 

306 

second fraction 
Polymerized sardine oil 

138.1 

2.45 

1.4843 

0.30 

1062 

319 

, third fraction 
Polymerized herring oil 

124.1 

2.39 

1.4858 

0.50 

1151 

336 

commercial sample 

106.4 

1.61 

1.4755 

0.20 

962 

301 


a Cryoscopic method, using benzene as a solvent. 

b The oils were saponified, the unsaponifiable fractions were removed, and the 
molecular weight determinations were run on the free fatty acids resulting from 
acidulation of the soaps. A correction was made for dimerization. 

The composition of the basal ration, and the rations fed during the test period are 
shown in Table II. It was decided to feed the oils at a level of 5% in the ration; this, 
it was thought, would compare more closely to the fat content of the average rat 
ration than the higher levels used by other workers in fat digestibility studies ( 10 , 
H, 12). 

During the test period and after period, the feces were collected daily and stored as 
composite group samples in a measured amount of aqueous 0.5% formaldehyde solu- 

TABLE II 
Composition of Diets 

Per cent 


Ingredient 

Fat-free 
basal diet 

Groups 2 
through 6 

Vitamin test casein G.B.I. 

30.0 

30.0 

Cystine, C. P. 

0.3 

0.3 

U.S.P. salt mixture No. 2 

4.5 

4.5 

Rice bran concentrate (NOPCO) 0 

8.0 

8.0 

Ruffex, Fisher 

5.0 

5.0 

Fat 

— 

5.0 

Sucrose 

52.2 

47.2 

2-Methyl-l ,4-naphthoquinone 

b 

b 


* Added with rice bran conccntrate/100 g. of diet: Riboflavin 1.0 mg., Calcium 
pantothenate 2.5 mg., Chlorine chloride 50.0 mg. 
h Added to all diets at a level of 5 7 /g. of diet. 



4 S. LASSEN, E. K. BACON AND H. J’. DUNN 

tion. The animals appeared to be in good health and no deficiency symptoms were 
noticed. The inclusion of Ruffex in the diet eliminated the greasy consistency of the 
feces observed in an earlier experiment, thus minimizing any mechanical losses during 
their collection. Samples of urine from the various groups of animals were tested for 
pH, and the ketone bodies were determined using the method of Greenberg et al (13). 
The results of these urine tests showed no significant trends. At the end of the experi¬ 
ment, the total amount of fat consumed by the individual groups was noted, and a 
fat determination on aliquots of the feces of the various groups was carried out using 
the Saxon method (14). 

To obtain some information as to the nature of the unabsorbed fat, the balance of 
the fecal samples were used for a determination of the average molecular weights of 
the fatty acids. The fat contained in the feces consists to some extent of fatty acids, 
partially free and partially in the form of soaps, besides fatty acids in ester form. 
To obtain a sample of fatty acid for molecular weight determinations which would 
include fat in all the above-mentioned forms, but exclude sterols and any oxidized 
fats, the following analytical procedure was used. 


TABLE III 

Analytical Results of Feeding Experimeid 


Type of fat fed 

Group 

No. of 
rats 

Total 

fat 

con¬ 

sumed 

Total 
fat re¬ 
covered 
in feces 

Unab¬ 
sorbed 
fat re¬ 
covered 
in feces* 

Coeffici¬ 
ent of 
digesti¬ 
bility* 

Molec¬ 
ular 
weight 
of fecal 
fatty 
acids' 

Molec¬ 
ular 
weight 
of fatty 
acids 
from 
unab¬ 
sorbed 
oil d 




0 . 

0 . 

0 . 




None (Fat-free basal) 

1 

10 


3.34 

— 


257 


Control sardine oil 

2 

10 

49.67 

4.19 

0.85 

98.3 

280 

374 

Polymerized herring 









oil 

3 

10 

49.75 

5.87 

2.53 

94.9 

399 

587 

Polymerized sardine 









oil, first fraction 

4 

10 

49.68 

5.33 

1.99 


345 

492 

Polymerized sardine 









oil, second fraction 

5 

10 

49.73 

8.56 

5.22 

89.5 

413 

513 

Polymerized sardine 









oil, third fraction 

6 

10 

49.25 

10.85 

7.51 

84.8 

565 

702 


0 The correction for metabolic fat was made by subtracting 3.34 g., the metabolic 
fat recovered in Group 1, from each of the other values. 

6 Coefficient of digestibility as used here is defined as that fraction of total ingested 
fat which is retained. 

6 Average molecular weight of total fecal fatty acids determined by the cryoscopic 
method using benzene as solvent and correcting for dimerization. 

d Calculated from the molecular weights of total fecal fatty acids by correcting for 
the influence of metabolic fat. 







DIGESTIBILITY OF POLYMERIZED OILS 


5 


The samples of feces were saponified, and sterols and other unsaponifiable matter 
removed from the saponificate by a thorough extraction with low-boiling petroleum 
ether. After acidulation the fatty acids were removed from the aqueous solution by 
repeated ethyl ether extractions. Following evaporation of the ethyl ether, the fatty 
acids were finally isolated by re-extracting the residue with low-boiling petroleum 
ether. Molecular weight determinations were then carried out on the fatty acid 
samples using the same method as applied to the original polymerized oils. 



Fig. 1. Relation between degree of polymerization in per cent and the coefficient 
of digestibility of sardine oil. The degree of polymerization is defined as: 

( Iodine number of Iodine number of\ 

unpolymerized oil — polymerized oil / X 100. 

Iodine number of unpolymerized oil 


The results of the fat balance study, as well as the molecular weight 
determinations, are shown in Table III. Table III and Fig. 1 show, as 
expected, that the digestibility of polymerized oils decreases as the 
degree of polymerization is increased. Table III also shows that the 
molecular weights of the unabsorbed fat (as determined on the com¬ 
posite fatty acids of fecal fat) increase as fats of increasing degree of 
polymerization are fed. 



6 


S. LASSEN, E. K. BACON AND H. I. DUNN 


Discussion and Results 

The determination of molecular weights on the polymerized oils and 
their fatty acids does not reveal with any large degree of accuracy to 
what extent sardine oil may be polymerized intramolecularly or inter- 
molecularly. The evidence does, however, speak in favor of a pre¬ 
dominantly intermolecular polymerization under the conditions of our 
experiment. This conclusion is based on a comparison of the increase 
noted in molecular weight of the polymerized sardine oils with the 
increase in the molecular weight of their corresponding separated fatty 
acids. If it were purely intermolecular polymerization, the molecular 
weight of an unsaturated glyceride would double in case of complete 
dimeric polymerization, whereas the molecular weight of its fatty acids 
would only increase by about 20%. If, on the other hand, it were ex¬ 
clusively intramolecular polymerization, the molecular weight of such 
a glyceride would not increase at all, while that of its separated fatty 
acids would increase by about 50%. 

The values in Table I show that the increase in molecular weight of 
sardine oil, due to polymerization, amounted to about 39%, whereas 
the increase in molecular weight of the corresponding separated fatty 
acids amounted to only 18%. These percentages favor intermolecular 
polymerization. 

The healthy appearance of the animals at the end of these experi¬ 
ments, and the fact that no significant trends were discovered in the 
analysis of the urine from these rats with respect to pH or ketone 
bodies, would seem to indicate that the ingestion of a polymerized oil 
does not produce any visible metabolic disturbances in the rat. The 
increase noted in molecular weight of the fatty acids from the fecal 
fat to nearly double its value, although the degree of polymerization of 
the corresponding oil has only increased to about 30%, suggests that 
the polymerized portion of the oil is not absorbed to any large extent. 
Such an assumption would explain the lower digestibility that we have 
found for the polymerized oils. 

The sample of polymerized herring oil, representing a commercially 
available oil which is being used in increasingly large amounts in the 
food industry, exhibited the same characteristics as the polymerized 
sardine oils. In addition to the lower digestibility found for polymerized 
sardine oil, preliminary experiments have indicated that feeding of 
these oils to rats on a vitamin K-free ration may produce vitamin 



DIGESTIBILITY OF POLYMERIZED OILS 


7 


K deficiency symptoms in 7—10 days. These findings are being further 
investigated and will be reported on separately. 

Acknowledgment 

Our appreciation is expressed to Frank M. Campbell for his valuable assistance 
with the animal experiments. 


Summary 

The data presented indicate that during the polymerization of 
sardine oil, intermolecular dimers are largely formed. The coefficient of 
digestibility of the polymerized oils was compared with that of the un¬ 
polymerized oil and found to decrease with increasing degree of poly¬ 
merization. 

A determination of the molecular weights of the total fecal fatty 
acids (free as well as combined) indicated that the polymerized portion 
of the oils is not absorbed to any large extent, if at all. 

References 

1. Jakobsen, F., Nergaard, R., and Mathiesen, E., Tids. Hermetikind. 27, 225 

(1941). 

2. Jakobsen, F., and Nergaard, R., Tids. Kjemi , Bergvesen Met. 3, 68 (1943). 

3. Anonymous, Food Manufacture XXIII, No. 9, 423-425 (September, 1948). 

4. Bradley, T. F., Ind. Eng. Chem. 29, 440 (1937). 

5. Brocklesby, if. N., Fisheries Research Board of Canada , Bull. 59, 107 (1941). 

6. Bradley, T. F., Ind. Eng. Chem. 30, 689 (1938). 

7. Bradley, T. F., and Johnston, W. B., ibid. 32, 802 (1940). 

8. Bradley, T. F., and Johnston, W. B., ibid. 33, 86 (1941). 

9. Roy, A., Ann. Biochem. and Exptl. Med. {India) 4, 17 (1944). 

10. Mattil, K. F., Higgins, J. W., and Robinson, H. E., Science 104, 255 (1946). 

11. Augur, V., Rollman, H. S., and Deuel, H. J., Jr., J. Nutrition 33, 177 (1947). 

12. Deuel, H. J., Jr., Cheng, A. L. S., and Morehouse, M. G., ibid. 35, 295 (1948). 

13. Greenberg, L. A., and Lester, D., J. Biol. Chem. 154, 177 (1944). 

14. Hawk, P. B., Oser, B. L., and Summerson, W. II., Practical Physiological 

Chemistry, p. 409. Blakiston, Philadelphia, Penna., 1947. 



The Chemistry of the Living Bark of the Black Locust Tree 
in Relation to Frost Hardiness. I. Seasonal 
Variations in Protein Content 12 

D. Siminovitch 3 and D. R. Briggs 

From the Division of Agricultural Biochemistry, University of Minnesota , 
University Farm , St. Pauly Minnesota 
Received September 13, 1948 

Introduction 

The development of frost hardiness in plants is a phenomenon which 
has occupied the interest and attention of botanists and plant physi¬ 
ologists for years. Considerable advances have been made toward 
arriving at an understanding of the physiological phases of this phe¬ 
nomenon (1, 2), but the problem of the biochemical mechanism of the 
development of hardiness remains obscure. The application of the 
newer biochemical techniques, developed from the study of animal 
tissues, to the study of the problems of plant physiology, has led to 
considerable advances in the understanding of the processes of photo¬ 
synthesis, metabolism, and salt accumulation in plants. With most 
plants of the north temperate zone, particularly the perennials, the 
development of frost hardiness is no less a part of their physiology than 
is photosynthesis or salt accumulation, but little has been done toward 
studying the problem from a biochemical point of view. 

1 Paper No. 2427. Scientific Journal Series, Minnesota Agricultural Experiment 
Station. 

* The studies presented in this paper were initiated in 1942-43, while D. Siminovitch 
held a Royal Society of Canada Fellowship for study under the late Prof. R. A. 
Gortner at the University of Minnesota. The contents of this paper are condensed in 
part from a thesis presented by D. Siminovitch to the faculty of the Graduate School 
of the University of Minnesota in partial fulfilment of the requirements for the degree 
of Doctor of Philosophy, July, 1948, and in part from a continuation of the studies 
made under a grant from the Graduate School, University of Minnesota. 

* Herman Frasch Foundation Fellow, Division of Agricultural Biochemistry, 
University of Minnesota. 


S 



BLACK LOCUST BARK. I 


9 


The trees and shrubs of the forests in the north temperate zone exhibit remarkable 
adaptation to low temperatures since they survive winter temperatures of —20° to 
—50°C. or lower. That the mechanism of this adaptation is not a permanent feature 
of the tree or shrub throughout the year but is elaborated annually and alternately in 
response to the seasonal periodicity of climate is evident from the fact that these trees 
and shrubs will be killed even by — 5°C. when frozen artificially in midsummer during 
their period of most active growth. 

Chemical investigation of the nature of this frost resistance has been deterred be 
cause only recently has it been established that it is largely a property of the proto¬ 
plasm of the living cell. Through the concerted efforts of a group working under G. W. 
Scarth, who had been occupied with the study of cell physiology over a period of many 
years, the problem has been considerably clarified. As a result of this work (3, 4, 5, 6, 
7), it has become possible to establish directions for a biochemical approach to the 
problem. It has been shown that certain physical changes, microscopically observable 
and measurable, occur in the protoplasm and protoplasmic membrane of living cells 
when plants develop frost hardiness. These changes consist of an increase in the 
permeability of the protoplasmic membrane to water and other polar substances, an 
increase in non-solvent space in the cell and an increased capacity of the proto¬ 
plasmic membrane to retain its plasticity after dehydration. The degree to which these 
living cells could withstand freezing is positively correlated with the extent of these 
physical changes, and a quick test of hardiness based on these protoplasmic changes 
has proven to be a valuable guide to studies on the real hardiness of plants. The 
intimate association of proteins and lipides with the cell, either as components of 
enzyme systems or of the structural organization within the cell, and the high surface 
activity of proteins and lipides which would indicate their participation in the struct¬ 
ure of the surface of the cell suggest that changes in these constituents may be involved 
in the development of cold resistance. 

This, the first in a series of studies dealing generally with the prob¬ 
lem of the biochemistry of frost hardiness of tree cells, will report the 
results of an analysis of the seasonal variations which occur in the 
proteins of living bark cells when these cells undergo changes in their 
resistance to frost injury. Some observations on the variations which 
occur in the carbohydrates are included. Because some experiments 
which were conducted on the bark of stumps of trees from which the 
crowns had been removed in late winter, yielded information as to the 
relative importance of protein and carbohydrate in relation to frost 
hardiness, analytical results on this stump material have also been 
included in this report. 


Experimental 

Choice of Material for Study 

The living bark of trees serves as excellent experimental material for an investiga¬ 
tion of the biochemical changes associated with the development of frost hardiness. 



10 


D. SIMINOVITCH AND D. E. BRIGGS 


In this cortex,’ the physical changes observable microscopically in the protoplasm 
during hardening and dehardening are greatest, occurring seasonally in autumn and 
spring, respectively. Because the living bark of the tree in winter is exposed to the 
greatest extremes of low temperature, it would be expected that the amplitude of 
change in hardiness, and, therefore, of physiological and chemical change, will be 
greatest in the cells of bark and should consequently be easiest to detect there. 

The selection of a suitable species of potentially hardy tree was contingent upon 
the success with which extraction of proteins from the bark tissue could be made. 
Apple trees were first tried because of the availability of a large number of varieties 
possessing recognized different degrees of hardiness. However, very little water- 
soluble protein could be extracted from apple bark, as has been previously reported 
by Thomas (8) and others, or from the barks of the other species of trese which were 
examined. 

The authors were directed to the black locust, Robinia pseudo-acacia L., by a 
publication of Jones, Gersdorff and Moeller (9), in which is described in detail the 
extraction and isolation of soluble proteins in considerable quantity from the bark 
tissue of this tree. These investigators used drastic extraction method#. The tissue 
was dried, ground thoroughly in a ball mill (both procedures being known to denature 
proteins), and then extracted with water or salt solutions. For the present studies, it 
was desirable that less drastic methods be used for maceration of the living bark and 
that the tissue used be in the fresh condition so that proteins could be extracted in 
the undenatured state. 

Determination of Fro8t Hardiness 

The availability of a reliable measure of hardiness was essential to the chemical 
investigation of the problem. The degree of hardiness of the bark cells must be known 
at any time that a chemical analysis is made if any relation between the hardiness and 
the chemistry of the bark is to be determined. The hardiness tests employed were 
based on studies of the properties of hardy and unhardy living cells (2), which showed 
that the degree of resistance of hardy cells to intercellular freezing can be estimated 
from the extent to which they can withstand dehydration, whether this is produced 
by plasmolysis in salt or sugar solutions or desiccation of the exposed cells in atmos¬ 
pheres of low relative humidities (10). Normally, when plant tissues are frozen, ice 
crystals grow in the intercellular spaces at the expense of water within the cells (5). 
The degree to which the cells arc dehydrated during freezing is determined by the 
temperature to which they are exposed. Plasmolysis of the cells followed by deplas- 
molysis, or desiccation and subsequent rehydration of the cells, in effect reproduce the 
conditions observed when the tissues are frozen and thawed. The lowering of the 
freezing point, as calculated from the osmotic pressure of the solution which the 
cells can withstand without injury, is roughly indicative of the freezing temperature 
which they will survive and, therefore, of their hardiness. 

The hardiness test on a bark sample was made as soon as possible after collection 
of the material. Thin, tangential sections were .taken from the living bark and placed 
in a graded series of balanced salt solutions (NaClrCaCh: :9:1) of increasing con¬ 
centration (1 Af-5 Af). Ten sections were plasmolyzed in each solution for 10 min. 
and then transferred to water. After an interval of 5 min., they were tested for survival 
by staining with neutral red solution (a few drops of a 1% solution in 25 cc. water). 



BLACK LOCUST BARK. I 


11 


If the cells absorbed and retained neutral red they were considered alive. An estimate 
was then made of the percentage of survival of the cells in each tissue section. The 
hardiness was recorded as the percentage of cells that remained alive after exposure 
to the various solutions. Since each tissue section contained over 100 cells, the estimate 
of survival after exposure to each of the solutions was based on at least 1000 cells. 

It is to be noted that the seasonal course of frost hardening observed in the locust 
with this hardiness test corresponded very well with the seasonal course of frost 
hardening observed in apple trees by Hildreth as determined by actual freezing tests 
( 11 ). 

Analytical Procedures 

1 . Preparation of Material 

The collection and treatment of material for these experiments were guided by the 
results of exploratory work done during 1942-43. Periodically through the fall, 
winter, spring, and summer of 1946-47, black locust trees 9-13 years old, from a 
single neighboring group, were cut near the base of the trunks and log sections 3 ft. in 
length and 3-4 in. in diam., and complete with outer and inner bark attached, were 
collected. The bark was promptly removed, the outer dead corky bark being peeled 
off with a knife and discarded, exposing the inner white living bark which was re¬ 
moved by holding the log horizontally against a rotating wire-brush wheel. Sufficient 
pressure was used to permit the wheel to brush lightly against the bark so as to avoid 
grinding the wood just underneath the bark. In this way the bark was shredded and 
removed uniformly over the length of the log. Largo samples of shreds were collected 
for extraction of protein. Small samples were collected in moisture dishes for the 
determination of moisture and nitrogen fractionation. To obtain a representative 
sample of the bark, the small samples were usually taken in complete narrow 1 in. 
rings encircling the log at 1 ft. intervals. Sampling in this way was necessary, because 
the bark shreds were not readily mixed and the removal of a small representative 
sample from a large quantity of shreds was not possible. 

2 . Nitrogen Determinations on Small Fresh Bark Samples 

Samples of fresh bark (equivalent to 4-5 g. dry matter) were thoroughly extracted 
with 150 cc. distilled water and filtered. The residue was then re-extracted with 
another 50 cc. water, filtered, and washed. The total filtrate and washings were 
combined, trichloroacetic acid added to form a 10% solution, and the precipitated 
proteins filtered off and washed with 10% trichloroacetic acid. Nitrogen determina¬ 
tions were made by the macro Kjeldahl method to secure values for the water insolu¬ 
ble nitrogen (insoluble protein in residue), the water soluble protein (trichloroacetic 
insoluble), and the water soluble non-protein nitrogen (soluble in 10% trichloroacetic 
acid). 


8. Determinations of Soluble Protein in Large Fresh Bark Samples 

Large samples of shredded bark (200-300 g.) were extracted with distilled water 
(10 g. water/g. bark) and filtered in the cold at 4°C. to yield clear, straw-yellow 
extracts. The extracts were saturated with (NHO 2 SO 4 , the heavy flocculum of 
protein collected by filtration and dialyzed against distilled water overnight at 4°C. 



12 


D. SIMINOVITCH AND D. E. BRIGGS 


The sacs were then transferred to a buffer solution of total ionic strength of 0.20 
consisting of a mixture of 0.19 ionic strength NaCl and 0.01 ionic strength phosphate 
at pH 7.2, and equilibrated against changes of this solution until free of sulfate. 

The nitrogen content in these extracts was determined in an aliquot by a micro 
Kjeldahl method. The total nitrogen extracted was calculated from the volume of 
extract obtained in each case. 

4. Reducing and Non-reducing Sugar Determinations 

Determinations of reducing and non-reducing sugars were made on small air-dried 
samples according to the method described in “Cereal Laboratory Methods” pub¬ 
lished by the Am. Assoc, of Cereal Chemists (1947, p. 32). 

5 . Moisture Determinations 

Moisture content of the bark tissue was determined by drying in a vacuum oven 
at 100°C. for 4 hr. 


REStJLTS 

The analytical findings are presented graphically in Figs. 1, 2, and 
3. In each of these figures the results of the hardiness tests are shown as 



stump! 


k ■ 


* 


5 2 
Q 

i— 

4 i 

$ 

35 

>- 

h 

QC 
< 


J JL 


Fig. 1. Showing the annual variations in the hardiness and in the water-insoluble 
protein nitrogen, the water-soluble protein nitrogen, and the water-soluble non-protein 
nitrogen contents (per cent dry weight of bark tissue) of the living bark of normal 
trees of the black locust. Data obtained on bark of stumps of trees of the same species 
during the months of June and July are also shown. (Data taken from small samples.) 
(X) ** non-soluble protein nitrogen. (O) ** soluble protein nitrogen. (#) * soluble 
non-protein nitrogen. 


BLACK LOCUST BARK. I 


13 



JLASONDJAFMAMJJL JJL 


Fig. 2 . Showing the changes in the hardiness and the water-soluble protein 
nitrogen (extracted from large samples of bark) in the living bark of the normal trees 
of the black locust during a year and in the bark of stumps of trees of the same species 
during the months of June and July. 



Fig. 3. Showing the variations which occur in the hardiness and in the soluble 
total, reducing, and non-reducing sugars in the living bark of normal trees of the 
black locust during a year and in the bark of stumps of trees of the same species 
during the months of June and July. (Data taken from small samples.) 



14 


D. SIMINOVITCH AND D. E. BRIGGS 


a crosshatched area representing degrees of survival of cells between 
zero and complete survival after being subjected to the dehydrating 
action of the salt solutions of the concentrations indicated. Fig. 1 shows, 
in addition, the analytical values, in per cent of dry weight of tissue, 
for insoluble (protein) nitrogen, soluble protein nitrogen and soluble 
non-protein nitrogen obtained on the small samples of bark. Fig. 2 
presents the data, calculated again as percentage nitrogen on the dry 
weight basis, for the soluble protein obtained by actual extraction from 
large samples of bark. Fig. 3 gives the analyses for sugars, total, re¬ 
ducing, and non-reducing, present in the bark at the time of collection 
of the logs from the field. 

It must be recognized that some degree of variation will always occur 
between individual trees collected at the same time of year, that small 
variations in composition may occur between different logs from the 
same tree, and that, in the sampling technique employed, it has not 
been possible to obtain samples which are entirely free of outer bark 
and underlying woody material. Consequently, the extent of uncon¬ 
trolled variation between the samples is rather large. Due to a limited 
number of trees available for these experiments, elimination of these 
uncontrolled variables by increasing the number of samples could not 
be attained. It is to be emphasized, therefore, that only variations 
which are relatively large between samples and which show a definite 
trend with respect to the measured hardiness and with the season of 
the year can be considered significant. 

It may be noted also that the amount of nitrogen extracted as 
soluble protein from the large samples of bark is, in all cases, somewhat 
smaller than that extracted as soluble protein from small samples. 
This discrepancy undoubtedly arises from the circumstance that, while 
the procedure used in extracting all samples in each class was consistent, 
the degree of extraction of the large samples was not as exhaustive as 
that employed on the small samples. 

Discussion 

It is apparent from these analyses that, of the nitrogen fractions of 
the living bark cells of the locust tree, only that corresponding to the 
water-soluble proteins shows a direct and consistent correlation with 
the frost hardiness of these cells. The water-insoluble (protein) fraction 
appears, within a certain degree of variation, which includes that which 
normally exists between individual trees and which includes variations 



• BLACK LOCUST BABK. I 


15 


in sampling accuracy, to remain constant through the year and is not 
related to hardiness variation. The water-soluble non-protein nitrogen, 
while showing a large degree of variation with time of year (a larger 
variation than is observed between samples from different trees which 
are collected at the same time) shows no correlation with the state of 
hardiness of the bark cells. 

From the data presented in Fig. 3, it would appear that the soluble 
sugar content of bark cells also increases in late summer and autumn 
in a manner which is roughly proportioned to the development of 
hardiness. In the spring however, it is seen that the soluble carbohy¬ 
drate content decreases precipitously with the initiation of sap flow 
(late April) and starts to increase again (in June) and these changes 
bear no close correlation with the dehardcning of the cells. This is in 
marked contrast to the close correlation between soluble protein con¬ 
tent of the cells and their state of hardiness observed through the same 
time period. 

This absence of correlation of hardiness with soluble carbohydrate 
content of the cells and the contrasting close correlation with their 
soluble protein content is emphasized in some analyses made upon the 
bark tissue of a number of stumps of trees from which the crowns had 
been removed in late winter (in March). Analyses made on these 
tissues during June and July (as shown at the right in each of the 
figures) indicated, in every case, that the degree of dehardening of the 
cells was markedly retarded in comparison with bark tissue of normal 
trees, and that the content of soluble proteins, which would normally 
have reached a minimum in July, was still above the normal minimum 
and present in amounts which were equal to those present in normal 
trees at such times as these showed comparable hardiness. The carbo¬ 
hydrate content of the bark cells of the stumps were found to be as 
low as, or lower than, that of normal trees at any time during the year, 
indicating again that no direct relationship exists between degree of 
hardiness of the cells and their sugar contents. 

On the basis of these analytical studies, made on the bark cells of 
some 40 locust trees collected at intervals throughout the period of 
one year, it is to be concluded that seasonal variations in the frost 
hardiness of this tissue are closely proportional to the simultaneous 
variations in the soluble protein content of the cells. No such correla¬ 
tion exists with respect to other nitrogen fractions nor with respect to 
soluble carbohydrate content of the cells. 



16 


D. SIMINOVITCH AND D. R. BRIGGS 


It is still too early in this investigation to state definitely that the 
observed correlation between soluble protein content of the bark cells 
and their ability to resist frost injury is a causal relationship. That such 
a relationship actually exists, however, is strongly supported by the 
physiological studies mentioned above. It would appear reasonable to 
assume that the permeability of a cell’s membrane to water or the 
extent to which the cell could be dehydrated or distorted without dis¬ 
ruption of its vital organization might be dependent upon its hydro¬ 
philic protein content. 

Although the correlation here demonstrated is one between protein 
concentration in the cells and resistance to the injurious effects of 
dehydration as would be produced by plasmolysis in the salt solutions 
used, or to the normal type of extracellular freezing, as measured by 
the hardiness test used, it is also possible that the increase in protein 
serves another purpose namely, protection of the cell against intra¬ 
cellular freezing in the event of rapid drop in temperature. It is con¬ 
ceivable in the latter case then that the higher content of soluble protein 
in winter protects the cells merely by inhibiting ice formation within 
the protoplasm by promoting undercooling so that, as cooling below 
the freezing point occurs, ice crystals are formed first in the intracellu¬ 
lar spaces, the undercooled water within the cells diffuses to these 
extracellular crystals and no ice crystals are allowed to form within the 
cells. It is not yet possible to picture specifically the mechanisms in¬ 
volved, nor is it possible to state positively that the observed correla¬ 
tion is causal. 

Summary 

The changes which occur seasonally in the water-insoluble protein 
nitrogen, the water-soluble protein nitrogen, the water-soluble non¬ 
protein nitrogen, and the reducing and non-reducing sugars of the 
living bark tissue of the black locust are studied in relation to the 
seasonal variations in its frost hardiness. 

Of the nitrogen fractions studied only the water-soluble proteins 
are found to increase in concentration in the bark in the fall along with 
the development of frost hardiness and to decrease in concentration in 
spring with the disappearance of hardiness. 

Reducing and non-reducing sugars decrease in spring but more 
rapidly than the rate at which hardiness is lost from the tissues. 

When the crown is removed from a winter tree, dehardening in the 



BLACK LOCUST BARK. I 


17 


bark of the stump is retarded and the seasonal decrease in protein, 
which is observed in normal trees, is also delayed in proportion to the 
retention of hardiness in the stump. At the same time, the reducing and 
non-reducing sugars in the stump decrease in concentration as in 
normal trees and to a greater extent. 

The correlation observed between the changes in water-soluble 
protein and hardiness suggest that this constituent of the bark bears 
some causal relationship to the mechanism of development of frost 
hardiness. It is also indicated that the sugars are not primary factors 
concerned in the mechanism. 

References 

1. Levitt, J., Frost Killing and Hardiness of Plants. A Critical Review. 211 pp. 

Burgess Pub. Co., Minneapolis, 1941. 

2. Scarth, G. W., New Phytologist 43, 1 (1944). 

3. Levitt, J., and Scarth, G. W., Can. J. Research 14C, 267 (1936). 

4. Scarth, G. W., and Levitt, J., Plant Physiol. 12, 50 (1937). 

5. Siminovitch, D., and Scarth, G. W., Can. J. Research 16C, 467 (1938). 

6. Levitt, J., and Siminovitch, D., ibid. 18C, 550 (1940). 

7. Siminovitch, D., and Levitt, J., ibid. 19C, 9 (1941). 

8. Thomas, W., Plant Physiol. 2, 109 (1927). 

9. Jones, D. B., Gersdorff, C. E. F., and Moeller, O., J. Biol. Chem. 64, 656 

(1925). 

10. Scarth, G. W., Plant Physiol. 16, 171 (1941). 

11. Hildreth, A. C., Univ. Minnesota Agr. Sta. Bull. 42, 1 (1926). 



The Chemistry of the Living Bark of the Black Locust Tree 
in Relation to Frost Hardiness. H. Seasonal Variations 
in the Electrophoresis Patterns of the Water- 
Soluble Proteins of the Bark 12 

D. R. Briggs and D. Siminovitch 3 

From, the Division of Agricultural Biochemistry, University of Minnesota, 
University Farm, St. Paul, Minnesota 
Received September 13,1948 

Introduction 

In Part I of this series (1), it was shown that the concentration of 
the total water-soluble proteins in the living bark of the black locust 
tree at any time during the year was in direct correlation with the degree 
of frost hardiness of the cells of the bark at that time. No attempt to 
characterize or fractionate the proteins so obtained was described in 
that paper. It was considered desirable to learn, however, whether the 
water-soluble cell protein involved was a single protein entity or a 
mixture. Also, if it were a mixture, it was of interest to learn whether 
or not the quantitative variation with hardiness could be identified 
with any single component or group of components in the mixture. 

The Tiselius electrophoretic technique has proved to be a useful 
tool in the characterization and analysis of animal proteins such as 
those obtained from the blood of animals (2). While some electro¬ 
phoretic studies have been reported on the proteins of seeds, the cyto¬ 
plasmic proteins of the vegetative parts of plants present rather an 
open field for investigation (3). Wildman and Bonner (4) have made a 

1 Paper No. 2428. Scientific Journal Series, Minn. Agr. Expt. Sta. 

1 Contents of this paper are condensed in part from a thesis presented by D. Sim¬ 
inovitch to the faculty of the Graduate School of the University of Minnesota in 
partial fulfilment of the requirements for the degree of Doctor of Philosophy, July, 
1948, and in part from a continuation of the studies made under a grant from the 
Graduate School of the University of Minnesota. 

* Herman Frasch Foundation Fellow, Division of Agricultural Biochemistry, 
University of Minnesota. 


18 



BLACK LOCUST BABK. II 


19 


study of the cytoplasmic proteins of spinach leaves with the aid of 
electrophoresis. Aside from the interest which an electrophoretic study 
of the cytoplasmic proteins of bark cells would have in itself, this 
technique appeared particularly well suited for the kind of seasonal 
analysis of the proteins required in this frost hardiness study. This 
paper will describe the results obtained from an investigation of the 
seasonal changes in the electrophoretic patterns of the water-soluble 
cellular proteins of the black locust bark in relation to changes in 
frost hardiness. 


Methods 

The proteins were extracted from fresh shredded bark tissue with distilled water 
and precipitated by saturating the solution with (NH^jSCh. The precipitate was 
filtered off and dialyzed until free of sulfate against phosphate buffer of ionic strength 
0.20 (consisting of a mixture of 0.10 ionic strength NaCl and 0.01 ionic strength 
phosphate buffer) of pH 7.2 and having a specific conductivity of 0.0115. After 
determination of the protein concentration in the dialyzed extracts by micro Kjeldahl 
analysis, they were all adjusted with buffer to approximately 1% protein concentra¬ 
tion and equilibrated against fresh buffer overnight previous to electrophoretic 
analysis. The proteins were examined in the conventional Tiselius electrophoresis 
apparatus, utilizing the Longsworth scanning device. Scanned photographs were 
taken after 2 hr. runs using 50 milliampcres current with a voltage drop of 6.0 volts/ 
cm. through the cell. 

The same protein extracts were used for electrophoretic analysis as were used for 
determination of total extractable protein reported in the preceding paper (1) so that 
the values for total water-soluble protein nitrogen and for frost hardiness given in 
that paper apply in the present experiments. 

Results 

Variations in Electrophoresis Patterns between 
Individuals of the Species 

When the location of the locust trees used in seasonal studies of 
frost hardiness was first selected, it was assumed that individual trees 
at the same stage of hardiness would not differ from each other greatly 
in the protein composition of their bark cells and only one or two trees 
were felled at each time. The results of the quantitative analyses 
described in the preceding paper supported this assumption. 

Electrophoretic analysis soon demonstrated, however, that, to make 
seasonal comparisons of the protein patterns, it was necessary to restrict 
the comparisons to trees which were growing in closely associated 
groups or clones. Electrophoretic analysis showed the following rela- 



20 


D. R. BRIGGS AND D. 8IMINOVITCH 


Ascamm aescarntw 

A "'A 

A"'h 


Fig. 1 . The differences in Tiselius electrophoresis pattern, in winter, between 
the bark proteins of trees of different groups or clones of the black locust. 

tionships to hold: (1) that the water-soluble protein from bark cells is 
a mixture of several electrophoretically distinguishable components 
and is not just a single protein entity; (2) that the ascending and de¬ 
scending patterns, while approximately symmetrical are not completely 
so, indicating some interaction between components in the mixture; 
(3) that proteins extracted from the bark tissues of different parts of 
the same tree give identical patterns, i.e., identical ratios of the various 
constituent proteins in the mixture; (4) that proteins extracted from 
different trees cut on the same day from a group in a closely adjacent 
locality give closely similar patterns; but (5) the patterns obtained 
from the proteins of trees cut on the same day (and yielding roughly 
the same total ’amount of protein extractable) but from groups in 
localities as little as 100 yards apart can vary to a considerable degree 
from each other with respect to the relative amounts of electrophoreti¬ 
cally distinguishable protein components present in the mixture. Each 
group of trees employed in this study seemed to be a clone because of 
underground root connections between adjacent trees. It is known 
that the locust can propagate vegetatively in this manner, and pre¬ 
sumably each group or clone arose from a single seed. 

The difference in the electrophoresis patterns for the proteins from 
individual trees collected in the winter from two clones is illustrated in 
Fig. 1. The characteristic pattern of the bark proteins of trees from 
each clone was almost exactly reproducible throughout the winter, 
which was the period of least seasonal fluctuation in the quantity of 
protein present in the bark cells. 



BLACK LOCUST BAKK. II 


21 


In spite of these differences in the electrophoresis patterns as observed 
between proteins obtained from individuals of separate groups or 
clones, it was found that the seasonal variations in the total quantity 
of soluble proteins, as described in the preceding paper, were the same 
in all individuals of the black locust independent of their location. 
For the study of the complete cycle of seasonal changes in electro¬ 
phoresis patterns occurring through the year, however, trees from 
a single clone had to be selected. The proteins from different trees of 
such a clone, consisting of trees 9—13 years of age, were determined to 
be so nearly identical, both with respect to total protein concentration 
and pattern on any one day, that any variations observed in the 
Tiselius patterns can be considered to constitute definite seasonal vari¬ 
ations in the ratios of component soluble proteins of the bark cells 
of trees belonging to that clone. 

Seasonal Variations in Electrophoresis Patterns 

The Tiselius electrophoresis patterns obtained with the water-soluble 
proteins from bark samples collected periodically during the hardening 
period from July to December, 1946, are shown in Fig. 2. It is remark¬ 
able that there is little or no change in the electrophoresis patterns and, 


JUL 22,1946 

AUG 3, 1946 

AUG 30,1946 

SEP 22, 1946 

NOV 3,1946 

DEC 3,1946 

Fig. 2. Tiselius electrophoresis patterns of the soluble proteins from the bark 
of normal trees during the autumn hardening changes of 1946. Descending patterns 
only. 




22 


D. R. BRIGGS ANDjD.JSIMINOVITCH 


therefore, in the relative proportions and number of soluble protein 
components in the tissue during the period after early September in 
which the greatest change in degree of hardiness is found to occur, and 
in which the greatest change in the total amount of soluble protein in 
the tissues occurs, as described in the preceding paper. There is 
observed a marked change, however, during July and early August 
preceding or in the initial stages of the development of hardiness. 

The series of electrophoretic analyses were continued into late winter, 
spring and early summer of 1947. As expected, the changes in the 
various patterns and protein fractions observed during dehardening 
were very much the same as in fall but in the reverse directions (Fig. 3). 


jkL 

jL. 

jLl 

A 

A 

A 

A 

A. 


JUL 15,1947 

JUL I, 1947 

JUL I, 1947 
STUMP 

JUN 17, 1947 
STUMP 

JUN 13,1947 
MAY 15,1947 
APR 3,1947 

FEB 6,1947 


Fig. 3. Tiselius electrophoresis patterns (descending) of the soluble proteins from 
the bark of normal trees and stumps during the spring dehardening period of 1947. 


The persistence of the same type of pattern throughout the whole 
winter and even in spring, after considerable dehardening and protein 
mobilization from the cells has occurred, is clearly in evidence. Again 
the most marked changes in seasonal patterns are produced during the 
midsummer after dehardening. The relative ratios of the various 
fractions of protein existing in July, 1946, are resumed to a fair degree 
of approximation in July, 1947. 



BLACK LOCUST BARK. II 


23 


The total area in each Tiselius pattern is proportional to the total 
protein concentration in the sample being electrophoretically analyzed. 
The proportions of the various protein fractions present can be esti¬ 
mated from the percentage of the total area which appears under each 
peak. When the percentage of each fraction is multiplied by the total 
amount of protein obtained from 100 g. dry weight of tissue, the actual 
amount of each of the various fractions in percentage of the dry weight 
of the tissue can be calculated. 

For the purposes of reporting the relative amounts of the fractions 
which are present in the cell proteins, the three peaks which appear as 
consistently symmetrical in the patterns are designated as A, B, and C 
(see Fig. 2 or Fig. 3) in the order of their increasing mobilities. Material 
which migrates slower than A and is of variable symmetry from pattern 
to pattern is designated as A' while that fraction (again of variable 
symmetry) which moves faster than C is designated C'. Thus, the 
areas of the patterns are divided into 5 fractions, 3 of which correspond 
to peaks of a generally high degree of symmetry. Patterns for the de¬ 
scending boundary were analyzed in each case. As mentioned before, 
the area relationship of the fractions in the ascending and the descend¬ 
ing patterns were not completely identical. Also the limits chosen as 
bounding the area under each peak were those drawn to the base line 
from the low point between peaks. This is an arbitrary and inaccurate 
procedure but no other, more accurate, procedure was possible in this 
case. For > these reasons, it is not considered that too much emphasis 
can be placed upon the ratios obtained except to that extent wherein 
broad trends are indicated. 

In Fig. 4 are plotted the amounts of total water-soluble protein as 
found in the bark tissue at various times throughout the year, together 
with the amounts of each of the fractions obtained from the Tiselius 
patterns as described above. The actual percentages of the various 
fractions in the total protein are found to vary in a broadly consistent 
manner between the summer and winter extremes. Average percent¬ 
ages of the fractions A', A, B, C, and C' in the July total soluble protein 
are found to be 7, 24, 25, 26, and 17, respectively, and for the period 
from early September to late May, to be 6, 29, 37, 20, and 8, respect¬ 
ively. It is evident that, with the development of frost hardiness, the 
total amount of each fraction increases (see Fig. 4) but that fractions 
A and B increase in greater proportions than do the others. This is 
especially true of fraction B. In this connection it is of significance to 



24 


D. R. BRIGGS AND D. SIMINOVITCH 



Fia. 4. Variations in the total water-soluble proteins of bark cells and in the 
various fractions of the soluble proteins as they occur throughout the year in normal 
trees and in stumps, as estimated from electrophoresis patterns. 

point out that the total protein content, the distribution among the 
various fractions (Fig. 4) and the general contour of the patterns (Fig. 
3) for the protein extracted from stump tissue (See Part I for descrip¬ 
tion of this tissue) collected in July are more nearly like those of early 
spring normal tissue than they are like normal July tissue. This is in 
agreement with the finding of a retardation in loss of hardiness of this 
stump tissue and with the idea that hardiness is causally related to 
the total protein content of the tissue. 

Discussion 

Results of the analytical estimations of amounts of water-soluble 
proteins present in the bark cells of the black locust tree at various 



BLACK LOCUST BARK. II 


25 


times of year, together with the results of electrophoretic analyses of 
these water-extractable proteins obtained from individual trees belong¬ 
ing to one closely associated group, or clone, throughout the same time 
period, yield evidence for the following sequence of events as occurring 
in the bark cells curing one year’s cycle in relation to the physiological 
states of the cells at the various periods of the year. While the actual 
studies reported here extended from July, 1946, to August, 1947, it is 
more convenient to discuss the results as covering the time period from 
the deep winter of one year through the spring, summer, and fall into 
the winter of a following year. 

If we start our cycle in January, it is found that, at this period, the 
bark cells exhibit a maximum of frost hardiness, contain a maximum 
percentage of their dry weight in the form of water-extractable pro¬ 
tein which, by electrophoretic analysis, yields a characteristic pattern 
in which the percentage amounts of the 5 fractions A', A, B, C, and C' 
are approximately 6, 29, 37, 20, and 8, respectively. These conditions 
are maintained without detectable change through February and 
March, although dormancy is broken sometime after the month of 
January. In April, sap begins to rise in the tree. By early May the 
buds have begun to swell, followed by leaf emergence during May, and 
growth of leaves to full size by the last of June. During the period from 
late April to the middle of June there occurs, in the bark cells, a rapid 
drop in water-soluble protein content with an accompanying and 
proportional drop in frost hardiness of these cells. During the early 
part of this drop in total water-soluble protein content, it is found that 
very little change in the percentage composition of this protein mixture 
occurs (in terms of the arbitrarily chosen fractions), all components 
being mobilized from the cells in about the same proportions. During 
the latter part of May and the early part of June, however, in the 
period of most active leaf growth, the total protein content drops 
rapidly, and it becomes apparent that a shift in percentage of the 
various components of that protein which remains in the bark cells is 
taking place. Also, at this period is observed the beginning of new 
growth of the bark. Cambial growth becomes increasingly rapid. By 
early July, leaf growth has ceased, cambial growth has reached a 
maximum rate and begins to fall off in rate. At this time, the content of 
water-soluble proteins in the bark cells is at a minimum, frost hardi¬ 
ness is at a minimum, and the percentage amounts of the 5 fractions 
detectable by electrophoretic analysis have changed to the extreme 



26 


D. R. BRIGGS AND D. SIMINOVITCH 


difference from those of winter tissue. At this time the percentages of 
A', A, B, C, and C' fractions are approximately 7, 24, 25, 26, and 17, 
respectively. 

In late July, cambial division is terminated and the new annual ring 
of wood has been virtually completed for the year. Final lignification 
of the wood then proceeds during August and the bark cells begin to 
prepare themselves for winter. By early August their protein content 
is already increasing rapidly, the maximum being approached at about 
the time the leaves are lost in the middle of October and attained in 
November. Maximum frost hardiness is reached at this time. The 
electrophoresis pattern has become nearly characteristic of winter 
proteins by the latter part of August, and this pattern is then main¬ 
tained throughout the winter and the cycle is complete. 

These electrophoresis observations were made on the proteins of the 
bark of trees belonging to what appears to be a single clone. A study of 
the complete cycle of patterns seasonally within other clones has not 
yet been made. Some electrophoretic analyses of the proteins from 
winter bark of trees belonging to a second clone have led to the unex¬ 
pected conclusion that a variation exists in the ratios of the component 
proteins of bark cells within the species, black locust. The patterns of 
the two clones were compared in winter because that seems to be the 
period of least seasonal change. It is probable that the differences in 
the respective patterns of the two clones can be accounted for by 
differences in the relative proportions of the various protein fractions 
that make up the total soluble cytoplasmic proteins rather than by 
differences in the nature of the protein fractions, although no attempt 
has been made to separate the fractions by solubility procedures. 
These differences in protein patterns in winter between groups, or 
clones, of trees are consistent and are not reflected in any apparent 
pathological symptoms in one or other of the groups or in any mor¬ 
phological difference. This is interesting because very little variation 
has been observed in the Tiselius patterns of normal human blood 
proteins, and even small deviations from the normal reflect pathological 
conditions (5). 

No taxonomical study has been attempted to determine whether 
different varieties of the species were involved in the study. Also, it 
has not been established that the differences observed between groups 
of trees are Hot environmental. However, if environment were re¬ 
sponsible, it is difficult to see how the characteristic pattern is so closely 



' BLACK LOCUST BARK. II 


27 


maintained from tree to tree within a group or clone and reproduced 
from year to year, without some small variation in response to some 
local environmental condition. Underground root connections were 
clearly in evidence between some of the trees in each group which led 
to the belief that each group is a clone. Apparently, the characteristic 
protein patterns of each clone are due to genetic differences and are 
transmitted vegetatively. This would warrant further investigation 
in which the Tiselius patterns of the proteins of individual plants of 
the same age and grown in the same environment, but derived from 
different seeds, would be compared. 

Summary 

A seasonal study of the water-soluble cytoplasmic proteins of the 
bark of the black locust tree by electrophoretic analysis is made in 
relation to its frost hardiness. 

The extractable proteins consist of at least 5 major clectrophoreti- 
cally distinguishable components. 

During the periods of the greatest changes in hardiness and in con¬ 
centration of the total protein, in fall and spring, no significant devi¬ 
ation occurs in the electrophoresis patterns from that which is char¬ 
acteristic of the winter protein. Marked changes occur in the pattern 
in midsummer, however. These changes are produced by relatively 
large shifts in the concentrations of 2 or 3 of the components relative 
to the other components of the total protein and not in the number of 
electrophoretically detectable components present. 

When the crown of a tree is removed, the normal development of the 
electrophoresis pattern characteristic of summer tissue proteins are 
somewhat retarded in the proteins of the stump. This retardation 
coincides with the inhibition of dehardening and of protein hydrolysis 
in the stump tissue. 

On the same day the electrophoresis patterns of the proteins taken 
from trees growing in close proximity, or within a “clone,” are essenti¬ 
ally identical, as are the patterns of proteins taken at various levels 
on the same tree. 

The electrophoresis patterns of proteins taken on the same day 
from different trees in more widely separated groups or “clones” are not 
identical. These differences seem to be accounted for by differences in 
relative concentrations of the various components of the total protein. 



28 


D. R. BRIGGS AND D. SIMINOVITCH 


References 

1. Siminovitch, D., and Briggs, D. R., Arch. Biochem. 23, 8 (1949). 

2. Edsall, J. T., Advances in Protein Chem. Ill, 383 (1947). 

3. Vickery, H. B., Physiol. Revs. 25, 347 (1945). 

4. Wildman, S. G., and Bonner, J., Arch. Biochem. 14, 381 (1947). 

5. Abramson, H. A., Moyer, L. S., and Gorin, M. H., Electrophoresis of Proteins 

and the Chemistry of Cell Surfaces, pp. 183-193. Reinhold Pub. Corp. New 
York, 1942. 



On the Nature of Highly Purified Mushroom 
Tyrosinase Preparations 

M. Frank Mallette 1 and Charles R. Dawson 

From the Department of Chemistry, Columbia University, New York, N. Y. 

Received November 24, 1948 

Introduction 

In previous communications from this laboratory (1,2,3), two types 
of highly purified mushroom tyrosinase have been described. Whereas 
both types of tyrosinase possess the ability to catalyze the aerobic 
oxidation of both monohydric and o-dihydric phenols, the ratio of 
these two activities is markedly different in the two types of prepara¬ 
tions. Tyrosinase preparations whose ratio of activities (catecholase to 
cresolase) 2 is relatively high have been called high catecholase prep¬ 
arations, and those in which the ratio of the catecholase to cresolase 
activity is relatively low have been called high cresolase preparations. 

Much effort has been directed in the past toward ascertaining 
whether or not the two activities of tyrosinase are those of one enzyme 
or a mixture of two enzymes, but the results thus far have not led to a 
satisfactory conclusion. The problem is a fundamental one because of 
the probable role of tyrosinase in plant respiration and other natural 
processes. 

Attempts have been made previously in this laboratory to obtain 
further information concerning the physical nature of tyrosinase by 
means of electrophoresis. Preliminary experiments were carried out 
using crude mushroom tyrosinase preparations, for it was known that 
the ratio of activities could most easily be changed by chemical means 
when crude preparations were employed. The results of these prelim¬ 
inary experiments led to the conclusion that no clean cut separation of 

1 Present address: Department of Biochemistry, Johns Hopkins University, Balti¬ 
more, Md. 

* The activities of the enzyme tyrosinase in catalyzing the aerobic oxidation of the 
monohydric phenol, p-cresol, and the o-dihydric phenol, catechol, are referred to as 
cresolase and catecholase activities, respectively. 

29 



30 


M. FRANK MALLETTE AND CHARLES it. DAWSON 


the two activities could be brought about in crude preparations by 
means of electrophoresis. It is the purpose of this communication to 
report on the results obtained using highly purified enzyme prepara¬ 
tions,* and from the results to propose a model that appears to recon¬ 
cile the conflicting views concerning the nature of the enzyme tyrosi¬ 
nase. 

Enzyme Preparations 

Five highly purified tyrosinase preparations made from the common 
mushroom Psalliota campestris, were used in this investigation. They 
were prepared by methods developed over a number of years in this 
laboratory, involving repeated fractional salt precipitations and alum¬ 
ina adsorptions. Details of the preparative procedures are published 
elsewhere (4,5). The copper content and activity data on the 5 prepara¬ 
tions used arc summarized in Table I. It can be seen from the activity 
ratio data in the table that 4 of the preparations were of the high cate- 
cholase type. It was originally planned to examine electrophoretically 


TABLE I 

The Enzymatic Properties and Copper Contents of 5 Highly Purified Tyrosinase 
Preparations from the Common Mushroom, Psalliota compestris 


Prepa¬ 

ration® 

Copper 6 

Activity 6 unita/ml. 

Unita/7 Cu 

Units/mg. dry weight 

Activity 

ratio 


Cat. 

Cre. 

Cat. 

Cre. 

Cat. 

Cre. 

Cat./Cre. 

I 

■per cent 

0.206 

21,500 

450 

2130 

48 

4400 

95 

48 

IT 

0.204 

49,000 

1140 

1980 

47 

4050 

95 

43 

III 

0.103 

11,,500 

570 

2100 

104 

2150 

107 

20 

IV 

0.098 

18,,500 

1100 

2320 

137 

2270 

135 

17 

V 

0.028 

1,880 

1180 

856 

536 

237 

149 

1.6 


0 For simplicity of discussion in this paper, the enzyme preparations have been 
assigned numbers different from those used in the original data and published else¬ 
where (5). The number key is as follows: 


I = C211-228F2II, II = C211-228F2I, III = C175BI, IV = C172-5AB, V = C189. 

* For methods used in measuring the activities and copper content see experimnetal 
section. 

* Certain of the results of the electrophoretic investigation of crude and purified 
mushroom tyrosinase preparations have been briefly described earlier in a review 
article on tyrosinase (3) and referred to as unpublished work. 




MUSHROOM TYROSINASE 


31 


more than one purified preparation of each type. However, due to 
preparational difficulties it was possible to include only one purified 
high cresolase preparation. 

Before considering the results of the electrophoresis studies, it will 
be helpful to study in some detail the data given in Table I. It will be 
observed that the preparations listed in Table I are arranged in the 
order of decreasing copper content. Also, it is worthy of note that this 
order of arrangement corresponds to an increase in the units of creso¬ 
lase activity /y of Cu as seen in Col. 6 of the table. The constancy of the 
catecholase activity /7 of Cu for the high catecholase preparations 
I to IV (Col. 5) confirms the dependence of the catecholase activity on 
the copper content as previously reported by Keilin and Mann ((5) and 
Ludwig and Nelson ( 1 ). 

Using units of activity/mg. dry weight of protein as a measure of the 
purity of the enzyme, it can be seen from Col. 7 of Table I that the 
“catecholase purity” decreases with a decrease in copper content. On 
the other hand the “cresolase purity” increases with decrease in copper 
content (see Col. 8 ). This relationship between purity and copper con- 



% COPPER 

Fig. 1 . The relationship between the activity ratios and copper contents of highly 
purified tyrosinase preparations. The copper data on the two preparations taken from 
Roth (7) were obtained by the manometric method of Warburg and Krebs (8). For 
other details see section on methods. 



32 M. FRANK MALLETTE AND CHARLES R. DAWSON 

tent suggests the presence of a copper-free protein possessing cresolase 
activity as a contaminant in preparations III, IV, and particularly V. 4 
However, such an explanation of the data is not the only one possible; 
indeed, as will be pointed out subsequently, there are reasons for 
favoring another interpretation. 

Before leaving Table I, one additional comparison of data should be 
made, namely, that the activity ratio and the copper content change 
in the same direction. Furthermore, when plotted (see Fig. 1), a linear 
relationship between these two quantities is suggested. The significance 
of this observation will be considered later. 

Electrophoresis 

The 5 purified tyrosinase preparations described in Table I were electrophoreti- 
cally analyzed in a small volume electrophoresis apparatus of the Tiselius type 
equipped to analyze the migrating boundaries optically 6 (for experimental details, see 
legends of Figs, and Tables). The studies were made with 3 purposes in mind. In the 
first place, a criterion of the homogeniety of these preparations was desired. Secondly, 

TABLE II 

Showing How the Electrophoretic Mobility of the Main Component in Tyrosinase 
Preparations of High Electrophoretic Homogeneity is Related to the 
Activity Ratio of Such Preparations? 


Preparation 

Ratio 

Cat. 

Cre. 

mX 10 & (Main 
component) 

Homogeneity 

pH 

Buffer (primaiy- 
secondary 
phosphate) 




per cent 



I 

48 

6.38 

100 

7.58 

0.05 M 

II 

43 

6.32 

100 

7.67 

0.05 M 

III 

20 

5.3 

95 

7.71 

0.10 M 

IV 

17 

4.5 

93 

7.58 

0.15 M 

V 

1.6 

3.0 

70-80 

7.71 

0.10 M 


tt Unfortunately, only preparations III and V were run at the same ionic strength 
(0.28), and the mobility values for preparations I, II, and IV should probably be 
corrected to this same ionic strength for a better comparison. Such a correction (see 
Abramson, Moyer and Gorin, Electrophoresis of Proteins, Rcinhold, 1942, p. 164, Eq. 
8) yields mobility values of 5.21, 5.16, and 5.05 X 10~ 6 for preparations I, II, and IV, 
respectively. The corrected values aro plotted as a broken line in Fig. 2. 

4 On the basis of copper content and “catecholase purity” it would appear that the 
high catecholase preparations III and IV were only about half as pure as the high 
catecholase preparations I and II. 

• All electrophoresis and ultracentrifuge experiments were performed by Dr. D. H. 
Moore, College of physicians and Surgeons, Columbia University. 






mushroom tyrosinase 


33 


if tyrosinase should be a mixture of two enzyme proteins, a method of separation was 
needed whereby the monophcnolase activity alone could be recovered. And finally if 
tyrosinase should be a single protein enzyme, a knowledge of the surface properties 
as indicated by the mobility.) of these purified preparations might yield Ze b- 
formation regarding the nature of the enzyme molecule. 



Fig SJ. The relationship between the activity ratios and electrophoretic mobilities 
° i llg r,v P unfied tyrosinase preparations. Measurements were taken with the small- 
scale Tisehus apparatus at pH values from 7.58 to 7.71 at 1°C. See footnote to Table 
11 ror significance of broken line curve. 


Tab e II contains electrophoretic mobility and homogenity data at about the same 
prt value for the 5 purified preparations presented in Table I. The relationship be¬ 
tween the mobility of the main component and the activity ratio is shown graphically 
m . 2 j* 50 footnote. Table II). The homogeneity data in Col. 4 of the table, 

estimated from scanning diagrams of the moving boundaries (such as are produced in 
!g. 3) show that, for preparations III, IV, and V, significant amounts of an “im¬ 
purity can be clectrophoretically resolved. In all 3 cases this material migrated more 
rapidly than the main component. Preparations I and II, however, were completely 
homogeneous within the limits of the electrophoretic method. 

In cases of preparations III, IV, and V, all of which contained more 
than one component, the electrophoresis experiments were prolonged to 
permit the partial separation of the components. By means of a motor- 

• The electrophoretic mobility of a protein is largely dependent on its surface prop¬ 
erties. See H. A. Abramson, L. S. Moyer and M. H. Gorin, Electrophoresis of Pro¬ 
teins. Rcinhold, New York. 




34 


M. FRANK MALLETTE AND CHARLES R. DAWSON 



Fig. 3. Scanning diagrams made during the electrophoresis of 5 different highly 
purified tyrosinase preparations (I, II, III, IV, and V) in the small-scale Tiselius 
apparatus. All diagrams except IVa show both the ascending (left) and descending 
(right) boundaries taken by the Longs worth scanning method (9). Diagram IVa 
(ascending boundary only) was made on preparation IV using the Philpot-Svensson 
(10,11) scanning method. Each diagram shows the homogeneity of the protein bound¬ 
aries after migration in an electric field for a given time interval. The distance of 
migration can be estimated for each component by comparing the position of each 
peak with the boundary position before the electrophoresis was started. 0 In diagrams 
I, II, III, and IV the initial position of the boundary is recorded as a slight break in 
the white line just above each scanning diagram. This white line is a photograph of the 
slit taken before the electrophoresis was begun. 6 All of the experiments were carried 
out at 1 rfc 0.01°C. using primary-secondary phosphate buffers. The constant for the 
conductivity cells, used to measure the resistance of the buffers and buffered protein 
solutions, was 1.089 reciprocal cm. Other data pertinent to each experiment are listed 
in the table below. 


Diagram 

I 

II 

III 

IV 

IVa 1 

V 

Time, hr. 

1 

1 

1.5 

3 

23 

3 

pH 

7.58 

7.67 

7.71 

7.58 

7.55 

7.71 

Buffer, cone. M 

0.05 

0.05 

0.10 

0.15 

0.15 

0.10 

Resistance of buffer and protein 

291 

293 

163 

116 

116 

161 

Current milliamps. 

18 

18 . 

26 

30 

50 

28 

Voltage between electrodes 

170 

180 

120 

120 

100 

120 

Cross-section area of cell, cm. 2 

0.761 

0.761 

0.800 

0.761 

— 

0.761 

Magnification factor 

0.560 

0.645 

0.560 

0.645 

— 

0.645 


° Except for diagram IVa and V. 

6 The scale of distance of migration along the abscissa of diagram IVa is different 
from that of IV. 








MUSHROOM TYROSINASE 


35 


driven hypodermic syringe, samples were then taken of the main com¬ 
ponents, the fast components, and the unseparated middle portions 
corresponding to the original enzyme preparations. In every case the 
enzymatic properties of the main components were the same as the 
properties of the respective original preparations. As was also expected, 
the unseparated middle portions were the same as the originals. 



Fig. 4. View A represents sedimentation diagrams for preparation I in an air- 
driven ultracentrifuge taken by the Longsworth scanning method. The experiment 
was conducted using 0.05 M primary-secondary phosphate buffer pH 7.58 at 32°C. 
The time interval between the upper and lower diagrams was 10 min. with the cell 
making 790 r.p.s. about an axis 57.06 mm. from the bottom of the cell. Reference lines 
appear as very faint streaks at the edges of the diagrams. Two other diagrams (not 
shown) were made at earlier times. Ten to 20% of a lightweight material is shown at 
the right of the main component. 

View B is a diffusion diagram taken with preparation I in a small-scale Tiselius 
electrophoresis apparatus using the Longsworth scanning method. The experiment was 
carried out in 0.05 M primary-secondary phosphate buffer pH 7.58 at 1 ± 0.01 °C. 
These views of the diffusing boundaries in both legs of the U-tube were taken 23 hr. 
8 min. after beginning the experiment. Other diagrams were taken up to a diffusion 
time of 48 hr. The magnification factor of the lens system was 0.645. 

The samples of the fast components all contained only traces of 
enzyme as evidenced by activity measurements. Furthermore, the 
properties of the activities and the activity ratios were the same as for 
the pure main components. It seems likely, therefore, that the pure 
fast components actually possessed no tyrosinase activity, but that the 
samples all showed a very slight activity as a result of mixing during the 
sampling process. From these studies it may be concluded that no 
electrophoretic separation of enzyme activities took place in any of the 
5 highly purified preparations investigated, and the electrophoretic 
| homogeneity of the high catecholase preparations was very high under 
the conditions employed. 





36 


M. FRANK MALLETTE AND CHARLES R. DAWSON 


Mobility data are recorded in Table III as taken with preparations III and IV at 
various conditions of pH. The experiments leading to these results were performed in 
the hope of ascertaining the isoelectric point of tyrosinase. Although no definite con¬ 
clusion can be drawn from the data of Table III, it is certain for preparations III and 
IV that the isoelectric point is below pH 5, probably in the vicicity of pH 3 or lower. 
Of importance in such studies is the fact that tyrosinase starts to lose its Cu and 
enzyme activity in solutions buffered below pH 5; and, therefore, electrophoresis in¬ 
vestigations are restricted to pH values near 5 or above. There is some reason to be¬ 
lieve that the isoelectric point may vary with the activity ratio of the enzyme. 

TABLE III 


Showing the Dependence on pH of the Electrophoretic Mobility of the Main Component 
of Two Highly Purified Mushroom Tyrosinase Preparations 


Preparation 

Ratio 

Cat. 

Cre. 

mXIO 5 (Main 
component) 

Homogeneity 

pH 

Buffer 




per cent 



III 

20 

5.3 

95 

7.71 

0.1 M primary-secon- 






dary phosphate 



4.8 

97 

5.86 

Mcllvaine 

IV 

17 

7.9 

90 

8.9 

S0rensen borate-HCl 



4.5 

93 

7.58 

0.15 M primary-secon¬ 






dary phosphate 



2.5 

95-100 

4.97 

Mcllvaine 


Kubowitz (12) reported that purified potato tyrosinase was isoelectric at pH 5.4. 
Experience in this laboratory has shown that the potato enzyme resembles the mush¬ 
room high cresolase preparation more closely than it does the high catecholase enzyme. 
It seems possible, therefor, that a high cresolase preparation might have a higher 
isoelectric point than suggested for the high catecholase preparations of Table III. 7 

Since it is known that two substances may have fairly similar electro¬ 
phoretic mobilities and yet differ greatly in other respects, an additional 
criterion of the purity of these tyrosinase preparations was desired. 
The ultracentrifuge provides one such criterion, since it is capable of 
resolving particles which differ in their sedimentation constants. Un¬ 
fortunately, however, only preparation I was available in sufficient 
quantity for both electrophoretic and ultracentrifugal analysis. 

The data obtained in the ultracentrifuge 5 using preparation I yielded a sedimen¬ 
tation constant for the main component of 6.4 Svedberg units at 20°C. A measurement 
of the diffusion constant at 1°C. gave a value of 6.1 X 10“ 7 when recalculated to 20°C. 

7 It is unfortunate that not enough of the high cresolase preparation V was available 
for a determination of its isoelectric point. 





MUSHROOM TYROSINASE 


37 


Using these values and assuming a partial specific volume of 0.75 cm. 3 /g. [an average 
for many proteins (13)], the molecular weight of the tyrosinase molecule in prepara¬ 
tion I calculated to be 102,000 with an uncertainty of about ±10%. 

An examination of the plate reproduced in Fig. 4 [a photograph of the sedimenting 
boundary taken by the Longsworth scanning method (14)] reveals that 80-90% of 
preparation I was present as a single homogeneous component. The impurity, con¬ 
sisting of 10-20% of the total, appears to be heterogeneous and is lighter than the 
main component. Ultracentrifugation was continued in this experiment until the main 
component was precipitated and could be collected as a solid. The cell contents were 
then removed and assayed for enzyme activity. The relatively unsedimented lighter 
portion was found to contain less than 1% of the enzyme activity with the activity 
ratio the same as for the original. Resuspension of the precipitated main component 
yielded a solution whose enzymatic properties were the same as for.those of the 
original preparation. Hence, although the preparation was not homogeneous in the 
ultracentrifuge, it may be concluded that the two activities can probably not be 
separated by means of the ultracentrifuge. 

A calculation based on the diffusion constant and the electrophoretic mobility 
indicated that the net charge on the enzyme molecules of preparation I (assuming 
spherical symmetry) was 17.8 electron units at pH 7.78. A negative sign for the net 
charge may be deduced from the fact that the enzyme migrated toward the anode. 
The assumption of spherical symmetry may not be justified for the tyrosinase mole¬ 
cule. As a matter of fact, the sedimentation and diffusion constants are in agreement 
with some deviation from the spherical form. This is revealed by a calculation of the 
molecular frictional factor, which in this case is 1.26, corresponding to an axial ratio 
of about 5.5, assuming no hydration of the molecule. On the other hand, proteins are 
commonly hydrated and the apparent asymmetry may be due in considerable part to 
such hydration. Adair and Adair (15) suggest that an asymmetry of the magnitude 
calculated above cannot be exclusively due to hydration and some must be inherent 
in the molecule. 


Discussion 

As indicated in the foregoing, preparation I was 80-90% homoge¬ 
neous in the ultracentrifuge and had a molecular weight of about 
100,000. Assuming that the heterogeneous impurity contained no 
copper, then the copper content of the molecule can be revised from 
0.206% upward to about 0.25%, a value which corresponds to 4 Cu 
atoms per enzyme molecule. To justify the assumption that the impur¬ 
ity contained no copper, it should be recalled that the catecholase 
activity of this preparation bore about the same relationship to copper 
as was observed for the other high catecholase preparations (see Table 
I). Furthermore, the parallelism between the catecholase activity and 
copper content during many stages of purification indicates that the 
impurities are copper-free and that all non-enzyme copper has been 
removed. 



38 


M. FRANK MALLETTE AND CHARLES R. DAWSON 


It will be recalled that a consideration of the copper content data and 
“catecholase purity” data shown in Table I lead to the suggestion that 
preparations III and IV might be only about half as pure as prepara¬ 
tions I and II by reason of contamination with copper-free protein. To 
explain the increase in “cresolase purity” with decrease in copper con¬ 
tent from Preparation I to V, it was only necessary to ascribe cresolase 
activity to the copper-free protein. However, as shown in Table II, 
preparations III and IV were found to be over 90% homogeneous' using 
electrophoresis as a criterion, rather than about 50% pure as suggested 
by the reasoning above. Preparation V was found to be 70-80% homo¬ 
geneous rather than less than 15% homogeneous as suggested on the 
basis of copper content. Furthermore, no evidence for the existence of 
a cresolase protein was found on enzymatic examination of the rela¬ 
tively small amounts of other protein components separated from the 
main components of each of these preparations during electrophoresis. 

As illustrated by the situation encountered with preparation I, 
where 10-20% of lightweight material was found in the ultracentri¬ 
fuge, although this preparation appeared homogeneous by electro¬ 
phoresis, it might be argued that a relatively large amount of copper- 
free non-enzyme material might be present in preparations III, IV, and 
V and not be clectrophoretically distinguishable. Such a possibility 
seems rather remote, however, since the data in Table III reveal that a 
considerable shift in the pH of electrophoresis did not greatly alter the 
observed homogeneities of preparations III and IV. It would be very 
unusual for two proteins to have the same mobility-pll characteristics 
over a pH range from 5 to 9. Moreover, the assumption involving an 
electrophoretically indistinguishable protein component does not satis¬ 
factorily explain the relationship observed (see Fig. 2) between the 
electrophoretic mobility of the main component and the activity ratio; 
a fact which can be due only to a change in the surface characteristics 
of the tyrosinase molecule. It would seem, therefore, that the physical, 
analytical and activity data on these 5 purified preparations cannot be 
accounted for on the basis of the presence of more than one protein 
component. 

It is possible, however, to account for these and other experimental 
tyrosinase data on another basis; a basis involving quite a new con¬ 
cept of the tyrosinase molecule. Before presenting this concept or model, 
however, it will be helpful to list certain important facts that have now 



MUSHROOM TYROSINASE 


39 


been established in regard to the nature and mode of action of highly 
purified tyrosinase preparations. These facts may be summarized as 
follows: 

1 . Crude mushroom tyrosinase preparations possess the ability to 
catalyze the aerobic oxidation of both monohydric and o-dihydric 
phenols. The two activities, which appear to involve quite different 
reactions, are referred to as cresolase and catecholase activities, res¬ 
pectively (3). 

2 . The ratio of catecholase to cresolase activity may be readily in¬ 
creased during purification (1,6,12,16) and the activity ratio of partially 
purified preparations may be changed by mild physical treatments such 
as warming (16,17), treatment with charcoal (18), and by adsorption 
on kaolin or alumina (16). 

3. In purified preparations possessing a high catecholase to cresolase 
activity ratio (high catecholase preparations) only the catecholase 
activity is proportional to the copper content (1,6,12), and such prep¬ 
arations have properties differing from those of the enzyme as it 
occurs in the mushroom juice. In purified high cresolase preparations 
both activities are proportional to the copper content (2,19) and the 
properties of such preparations more closely resemble the properties of 
the enzyme in the plant juice. 

4. A change in the activity ratio is always accompanied by a change 
in the purity level of both activities (cresolase and catecholase activi¬ 
ties/mg. dry weight) and in highly purified preparations the activity 
ratio and the copper content are related (see Fig. 1). 

5. A tyrosinase preparation having only monophenolase activity 
has never been reported (3) and the simultaneous oxidation of o-dihy- 
dric phenol is necessary Tor the enzymatic oxidation of monohydric 
phenol (20,21). 

6 . Both activities are inhibited to the same extent by the copper 
inhibitors, cyanide and diethyldithiocarbamate (22). 

7. The electrophoretic mobility varies with the activity ratio (see 
Table II and Fig. 2.) and neither electrophoresis nor ultracentrifu¬ 
gation separate the two activities. 

8 . Several purified preparations differing widely in copper content 
and activity ratio have been found to possess a high degree of homo¬ 
geneity (see Tables I and III). 



40 


M. PRANK MALLETTE AND CHARLES R. DAWSON 


A New Model for the Enzyme 

The several hypotheses involved in the construction of an enzyme 
model that reconciles all of the above tyrosinase data, stem from the 
observation made in this study that several highly purified tyrosinase 
preparations having different copper contents and different activity 
ratios were each judged to possess a high degree of homogeneity and 
activities that could not be separated. These conditions lead to the 
fundamental hypothesis of the model to be proposed, i.e., the mole¬ 
cular sizes of the enzyme molecule are different in tyrosinase prepara¬ 
tions having different activity ratios, but the number of copper atoms 
per enzyme molecule is essentially constant. 

To account for the relationships between the copper contents, molec¬ 
ular weights, and activity ratios of various types of tyrosinase prep¬ 
arations, another important hypothesis of the model states that a 
fragmentation or loss of portions of the natural tyrosinase molecule 
occurs as the result of chemical or physical treatment during the prep¬ 
aration or purification of the enzyme solution. As a third and final 
postulate, the model requires that the fragments lost during the 
preparative procedures contain none of the copper originally present in 
the natural molecule, but such fragments must be in some way impor¬ 
tant elements in the cresolase activity. It also seems likely that, during 
the early stages of the fragmentation process, the catecholase activity 
is influenced. Furthermore, a loss of fragments of the molecule would 
be expected to alter the net charge and therefore the electrophoretic 
mobility of the enzyme due to removal, alteration or formation of new 
charged groups. As previously shown (Fig. 2) a relationship between 
the mobility and the activity ratio was observed. 

Suggesting that the size of the tyrosinase molecule varies through¬ 
out the preparations listed in Table I amounts to proposing a higher 
molecular weight for high cresolase than for high catecholase prep¬ 
arations. Unfortunately, it was only possible in this study to determine 
the molecular weight of the high catecholase preparation I (molecular 
weight 100,000). However, Tenenbaum (17) determined the molecular 
weight of the enzyme in a mushroom tyrosinase preparation by the 
diffusion method of Northrop and Anson (23) and reported a value of 
about 200,000. The preparation used by Tenenbaum apparently had a 
considerably lower activity ratio than preparation I, since the described 
enzymatic properties resemble those of preparations III and IV. No 



MUSHROOM TYROSINASE 


41 


data on the copper contents were included in Tenenbaum’s report. How¬ 
ever, on the basis of the model presented above, a preparation having a 
lower activity ratio than preparation I would be expected to have a 
higher molecular weight . 8 

If it be assumed, as in the model, that the natural tyrosinase mole¬ 
cule is a single protein complex possessing 2 characteristic activities 
and containing 4 atoms of copper, and the size of the molecule de¬ 
creases during the process of obtaining a high catecholase type prep¬ 
aration, and the cresolase activity is associated with the fragments 
lost in this process (as well as on the copper), then the facts previously 
enumerated as applying to the nature of tyrosinase, may be explained 
as follows: (The points arc numbered to correspond with the previous 
summarization.) 

1 . The fundamental definition of the enzyme model as stated above 
is based on a single protein entity possessing two different activities. 

2 . A change in the activity ratio on treatment of the molecule in 
various ways would arise from the partial loss of portions of the mole¬ 
cule responsible for the cresolase activity as well as from the “unmask¬ 
ing” of the “catecholase centers” as the large high cresolase type 
molecules decrease in size. 

3. In a series of high catecholase preparations the catecholase activ¬ 
ity would be proportional to copper while the cresolase activity would 
not be proportional owing to the loss, in varying amounts, of the groups 
involved in the catalysis of p-cresol oxidation. In high cresolase prep¬ 
arations both activities should be more or less proportional to copper 
with a lowered value for the catecholase activity /7 of Cu. A decrease 
in the catecholase activity per unit of Cu might be expected for large 
molecules where some “catecholase active centers are masked” by the 
large size of the molecule. 

4. The catecholase activity/rag. of dry weight (catecholase purity) 

8 The authors are fully aware of the fact that the proposed model for the enzyme 
tyrosinase needs further experimental evidence that the molecular size of the enzyme 
actually does vary with the activity ratio. Such experiments will be performed as 
soon as possible. Due to the great difficulty of preparing sufficient amounts of reason¬ 
ably homogeneous tyrosinase specimens of relatively low catecholase/cresolase ratio 
it is likely that some time may elapse before the obvious ultracentrifuge and diffusion 
experiments can be performed. In the meanwhile, it is hoped that the proposed model 
will stimulate research on this important enzyme and lead to a better understanding, 
of its action. 



42 


M. FRANK MALLETTE AND CHARLES R. DAWSON 


would increase with increasing activity ratio due to a decrease in molec¬ 
ular weight without loss of catecholase activity. The “cresolase purity” 
might increase, decrease, or remain constant, depending on the relative 
rates of change in molecular size and cresolase activity per molecule. 
According to Table I there is actually a slight decrease with increasing 
ratio. As protions of the molecule or complex are lost, the percentage 
copper content would increase, always providing that the fragments 
lost contain relatively little copper. 

5. By definition, the model precludes the possibility of a separate 
cresolase enzyme. The cresolase activity is associated with parts of the 
molecule which when separated as fragments possess no enzymatic 
(cresolase) activity. The close association of the two activities in the 
model is in line with the observation that the simultaneous oxidation 
of o-dihydric phenol is necessary for the enzymatic oxidation of mono- 
hydric phenol. 

6. Since, in the model, both activities are dependent on the same Cu 
atoms, substances which combine with Cu would necessarily inhibit 
both activities to the same extent. 

7. Electrophoretic properties of the molecule might be expected to 
change as fragments are removed from the molecule, since it is probable 
that both the diffusion constant and the net charge would change. 

8. According to the model, highly purified tyrosinase preparations 
(of either type) should possess the properties of a single homogeneous 
protein entity. 


Experimental 

The experimental details associated with the electrophoresis studies are given in the 
legends of the several tables and figures. 

Three different methods were employed to determine the copper content of the 
several enzyme preparations. Preparations I, II, and V were assayed by the polaro- 
graphic method of Ames and Dawson (24). Preparations III, IV, and V were assayed 
by the manometric method of Warburg and Krebs (8), based on the copper catalysis 
of the oxidation of cysteine. Preparation IV was also analyzed spcctrographically.’ 

It will be noted that the copper of preparation V was determined in two ways. The 
cysteine method yielded 0.0296% and the polarographic method 0.0255%. The prep¬ 
aration of the copper-free water and other precautions found necessary have been 
previously described (24). 

• Determination made by Dr. A. F. Daggett of the University of New Hampshire 
who reported 0.095% copper compared to 0.098% found by the Warburg-Krebs 
method. No metals other than copper were found except for an estimated 0.5% of cal¬ 
cium. 




MUSHROOM TYROSINASE 


43 


Catecholase activity was determined by the chronometric method (25) and cre- 
solase activity was assayed manometrically (26). Protein concentrations were esti¬ 
mated as dry weights of undialyzable solids (5). 

Acknowledgments 

The authors gratefully acknowledge the helpfulness of preliminary experiments on 
the electrophoresis of crude tyrosinase preparations carried out in this laboratory 
with Dr. Stanley R. Ames, now of Distillation Products, Inc., Rochester, New York. 
The authors are also indebted to Mr. Stanley Lewis of these laboratorise for aid in the 
preparation of the enzyme specimens, to Dr. Albert F. Daggett of the University of 
New Hampshire for the spectrographic analysis, and to Dr. Daniel H. Moore of the 
College of Physicians and Surgeons, Columbia University for the electrophoretic and 
ultracentrifuge analyses. The preliminary phases of this investigation were made 
possible by a grant from the American Philosophical Society. 

Summary 

1. Five highly purified tyrosinase preparations obtained from the 
common mushroom, Psalliota campestris, and possessing different 
amounts of copper, different ratios of catecholase to cresolase activities, 
and different levels of purity as indicated by the activities per mg. of 
dry weight, have been eleetrophoretically analyzed in an apparatus of 
the Tiselius type. One preparation was also examined in the ultra¬ 
centrifuge. 

2. No separation of the two activities of tyrosinase was effected by 
the electrophoresis and ultracentrifuge procedures. Purified prepara¬ 
tions of markedly different copper content and activity ratio were 
found to possess equally high orders of electrophoretic homogeneity. 

3. On correlating the activity, copper and homogeneity data, it has 
been found possible to explain all of the known properties of mush¬ 
room tyrosinase in terms of a single copper-protein entity. The princi¬ 
pal feature of this explanation is that different types of tyrosinase 
(different activity ratios, etc.) arise as the result of fragmentation of the 
protein molecule during the preparative procedures. 

4. A model of the tyrosinase molecule is suggested, which accounts 
for a decrease in the ability of the enzyme to catalyze the oxidation of 
p-cresol and an increase in the copper content as the size of the mole¬ 
cule is decreased during treatment in various ways. The model also 
accounts for the proportionality between the catecholase activity and 
copper content after preliminary stages of purification. Although addi¬ 
tional experimentation is necessary to prove definitely the validity of 
the model, all tyrosinase data now available are in agreement. 



44 


M. FRANK MALLETTE AND CHARLES R. DAWSON 


5. The ultracentrifuge results indicate that purified mushroom tyrosi¬ 
nase of the high catecholase type is a copper protein of molecular 
weight about 100,000 and containing 4 atoms of copper/molecule 
(0.25% Cu). 


References 

1. Ludwig, B. J., and Nelson, J. M., J. Am. Chem. Soc. 61, 2601 (1939). 

2. Parkinson, G. G., and Nelson, J. M., ibid. 62, 1693 (1940). 

3. Nelson, J. M., and Dawson, C. R., Advances in Enzymol. 4, 99 (1944). 

4. Mallette, M. F., Dissertation, Columbia University, 1945. 

5. Lewis, S., Mallette, M. F., Ames, S. R., Nelson, J. M., and Dawson, C. R., 

Arch. Biochem . 16, 283 (1948). 

6. Keilin, D., and Mann, T., Proc. Roy. Soc., London 125B, 187 (1938). 

7. Roth, L. J., Dissertation, Columbia University, 1944. 

8. Warburg, O., and Krebs, II. A., Biochem. Z. 190, 143 (1927). 

9. Longsworth, L. G., J. Am. Chem. Soc. 61, 529 (1939). 

10. Philpot, J. S. L., Nature 141, 283 (1938). 

11. Svenson, H., KoUoid-Z. 87, 181, (1939); 90, 141 (1940). 

12. Kubowitz, F., Biochem. Z. 292, 221 (1937); 299, 32 (1938). 

13. Svedburg, T., and Pedersen, K. O., The Ultracentrifuge, Oxford, 1940. 

14. Moore, D. H., Rev Sci. Instruments 14, 293 (1943). 

15. Adair, G. S., and Adair, M. E., Proc. Roy. Soc., London 120B, 422 (1936). 

16. Adams, M. H., and Nelson, J. M., J. Am. Chem. Soc. 60, 2472 (1938). 

17. Tenenbaum, L. E., Dissertation, Columbia University, 1940. 

18. Onslow, M. W., and Robinson, M. E., Biochem. J. 22, 1327 (1928). 

19. Dalton, H. R., and Nelson, J. M., J. Am. Chem. Soc. 61, 2946 (1939). 

20. Bordner, C. A., and Nelson, J. M., ibid. 61, 1507 (1934). 

21. Behm, R. C., and Nelson, J. M., ibid. 66, 711 (1944). 

22. Gregg, D. C., and Nelson, J. M., ibid. 62, 2500 (1940). 

23. Northrop, J. H., anq Anson, M. L., J. Gen. Physiol. 12, 543 (1928-29). 

24. Ames, S. R., and Dawson, C. R., Ind. Eng. Chem., Anal. Ed. 17, 249 (1945). 

25. Miller, W. H., Mallette, M. F., Roth, L. J., and Dawson, C. R., J. Am. 

Chem. Soc. 66, 514 (1944). 

26. Mallette, M. F., and Dawson, C. R., ibid. 69, 466 (1947). 



Separation of a Crystalline Globulin from Tomato Juice 
and Determination of Its Isoelectric Point 1 

D. C. Carpenter and William C. Smith 

From the Division of Food Science and Technology , New York State 
Experiment Station , Geneva , New York 
Received December 14, 1048 

Introduction 

In connection with the separation of a new food accessory related to 
thiamine metabolism (4), we have had occasion to isolate a crystalline 
globulin from tomato juice serum (Lycopersicon esculentum ) and deter¬ 
mine its isoelectric point. It was necessary to know whether the 
globulin behaved as a cation or anion, so one could know whether it 
would remain in solution or come out on the cation exchange resin 
employed to separate the new food accessory. 

Many of the plant globulins (1) have an isoelectric point around 
pH 5.2-5.5. However, we found that asclepain from milkweed had an 
isoelectric point of pH 3.11 (2). The globulin from tomato has an 
isoelectric point of pH 3.43 as reported herein. 

Experimental 

Preparation of Tomato Globulin 

Fresh juice from John Baer tomatoes was run slowly through a Sharpies super¬ 
centrifuge to remove the red cellular debris and finally filtered crystal clear through 
asbestos and paper pulp. The filtered juice was light yellow in color and had a pH of 
4.2 and contained 4.5% total solids, the bulk of the latter being sugars. The juice was 
concentrated to 35-40% solids by freezing out ice in 5-6 successive freezings, sepa¬ 
rating ice from the mother liquor each time in a perforated bowl centrifuge and re¬ 
freezing the mother liquor. The concentrated syrup was subjected to dialysis against 
water. The dialyzate contained the new food factor as well as sugar and other dialyz- 
able substances. The globulin precipitated out inside the dialyzer in amorphous con¬ 
dition and brownish in color. This was transferred to cylinders and sedimented in a 

1 Approved by the Director of the New York State Experiment Station for publica¬ 
tion as Journal Paper No. 768. 


45 



46 


D. C. CARPENTER AND WILLIAM C. SMITH 


centrifuge. The globulin was dissolved in warm (50°C.) M NaCl solution, stirring 
occasionally for 2 hr., and the liquid separated from undissolved matter in a centri¬ 
fuge. The somewhat turbid centrifugate was filtered crystal clear through paper pulp 
and asbestos, diluted with 5 volumes of water at 50°C. and again filtered through 
paper pulp and asbestos. Toluene was added and the globulin solution was placed in 
the cold room at 0°C. for 48 hr. to crystallize. The mother liquor was decanted off and 
the colorless crystalline globulin separated in the centrifuge. The globulin was redis¬ 
solved in warm M salt Solution, placed inside a dialyzer tube and attached to a slowly 
driven stirrer which rotated dialyzer tube and contents in a beaker of cone. (NH 4 ) 2 - 
S0 4 solution. As the concentration of (NH 4 )2S0 4 increased within the dialyzer tube, 
solid (NH 4 )2S0 4 was added to the outer solution. Small, needle-like crystals of the 
globulin began to separate, and after 3-4 days the operation was concluded and the 
crystals separated centrifugally from the mother liquor. The crystals were redis¬ 
solved in warm M salt solution, filtered and recrystallization carried out by 5-fold 
dilution with water as above. The crystals were finally suspended in distilled water 
and electrodialyzed between parchment membranes to remove the last trace of salt. 
This suspension was preserved with toluene in the cold room until used in the electro¬ 
phoresis work. From 40 gal. of tomato juice some 5-6 g. of globulin were separated. 

Electrophoresis Experiments 

Preliminary electrophoresis experiments indicated that the iso¬ 
electric point of the globulin lay between pH 3 and 4. Accordingly, a 
series of N/100 phthalate buffer solutions were prepared covering the 
pH range 2 to 6, and 1 ml. of an aqueous suspension of micro crystals 
of the globulin added to 100 ml. of each buffer solution, protected with 
toluene, and kept in the cold room until the electrophoresis experiments 
‘were completed. 

We employed the method of Northrop and Kunitz (5), measuring 
under the microscope the migration velocity of the suspended crystal 
fragments and reversing the applied potential repeatedly. Measure¬ 
ments were made at both upper and lower stationary levels and the 
mean of a dozen or more closely agreeing observations for each pH 
were averaged. 

The Helmholtz-Lamb equation V = f ND/iirh was used in calculat¬ 
ing the electrokinetic potential. The values of D and h (dielectric con¬ 
stant and viscosity, respectively) have been assumed to be those for 
water at 25°C., namely, 81 and 0.009, respectively. All quantities in 
the above equation are expressed in e.g.s. electrostatic units. 

The data are shown graphically in Fig. 1, from which it is clear that 
the isoelectric point of the tomato globulin is pH 3.43. Confirmation of 
this value for the isoelectric point was had by combining all of the 



SEPARATION OF A CRYSTALLINE GLOBULIN 


47 



Fig. 1. Migration velocity and f-potential of tomato globulin. 


various globulin solutions left over from the electrophoresis work and 
bringing the pH to 3.43, whereupon the protein separated out cleanly 
in a few minutes. 


Discussion 

Cohn and coworkers (3) carried out some work on the electrophoresis 
of the red cellular debris of whole tomato juice and found zero migration 
around pH 4.69. It was of interest to repeat Cohn’s work with the red 
cellular matter of juice sedimented in the Sharpies centrifuge, after 
washing this several times with water and separating it by centrifuging. 
A small amount was suspended in phthalate buffer solutions of various 
pHs and the migration velocity, shown graphically as a broken line 
in Fig. 1, was obtained. Zero migration occurred at pH 2.40. It is 
evident that any protein adsorbed on the red cellular debris is not the 
globulin. 

It follows that the globulin occurs in tomato juice (pH 4.2) as anion 
and can be removed by an anion exchange resin (IR-4) without going 
through the tedious dialysis operation. The new food accessory is left 
in solution and is removed subsequently by cation exchangers. 



48 


D. C. CARPENTER AND WILLIAM C. SMITH 


Acknowledgments 

We are glad to acknowledge a grant from the W T illiams-Waterman Fund in partial 
support of the work, and also for supplies of exchange resin from the Resinous 
Products and Chemical Company. 


Summary 

The separation of a new crystalline globulin from tomatoes is de¬ 
scribed and its isoelectric point found to be pH 3.43. 

References 

1. Carpenter, D. C., and Lovelace, F. E., J. Am. Ckem. Soc. 65, 3738 (1033). 

2. Carpenter, D. C., and Lovelace, F. E., ibid. 66, 2364 (1943). 

3. Cohn, E. J., Cross, J., and Johnson, 0. C., J. Gen. Physiol. 2, 145 (1920). 

4. Metcalf, D., Hucker G. J., and Carpenter, D. C., J. Bad. 51, 381 (1946). 

5. Northrop, J. H., and Kunitz, M., J. Gen. Physiol. 7, 729 (1925). 



The Occurrence of Ergosterol in Neurospora crassa 
Robert C. Ottke 

From the Department of Botany and Microbiology , 

Yale University , New Haven , Connecticut 
Received January 17, 1949 

Introduction 

Ergosterol, C 28 H 44 O, has been found to occur in molds thus far in¬ 
vestigated, and hence may be considered the principal sterol of the 
lower fungi. In some instances (1), ergosterol has not actually been 
isolated, but its presence has been inferred because irradiation of the 
mycelium with ultraviolet light produced antirachitic activity. 

Large quantities of N. crassa mycelia (strain 4540A) were accumulated in connec¬ 
tion with the isolation and identification of a natural precursor of nicotinic acid (2). 
The mold was grown under carefully controlled conditions on Fries minimal medium 
(3) (2% sucrose) in 20-1. pyrex bottles at 30°C. for 4 days. The cultures were constant¬ 
ly aerated through a sterile filter. The mycelia were filtered off and dried in a forced- 
draft oven at 65°C. for 24 hr. and then ground in a Wiley mill. Yield of mycelia was 
approximately 80 g. (dry weight) from 20 1. of medium. 

The identification of ergosterol is based upon the excellent agreement of the ultra¬ 
violet absorption spectrum of the sterol isolated from Neurospora with the spectrum of 
authentic ergosterol determined on the same instrument, as well as good agreement of 
the physical constants of the sterol, steryl acetate, and steryl benzoate with those of 
the corresponding derivatives prepared from authentic ergosterol. The sterol gives 
the typical color reactions (Liebermann-Burchard and Tortelli-Jaff6) of ergosterol. 

Experimental 

All melting points arc corrected. All optical rotations were taken at room tempera¬ 
ture in a 1.000 dm. tube, the sample being dissolved in 2.056 cc. of chloroform. 

Isolation of Ergosterol 

2610 g. of dry, finely ground mycelia were extracted with acetone in a Soxhlet 
apparatus for 110 hr. After removal of the acetone by distillation, the brown, oily 
residue weighed approximately 90 g. This was saponified with 400 g. of KOH in 21. of 
70% ethanol; allowed to stand at room temperature for 72 hr., diluted to 4 1. with 

49 



50 


ROBERT C. OTTKE 


water and extracted 6 times with 300 cc. portions of ether. The ether extracts were 
combined, washed 3 times with water, and dried for 24 hr. over anhydrous NasSCh. 
The ether was distilled off in a N* atmosphere, and the crude sterol fraction, which 
weighed 4.8 g., was digested with methanol, crystallized, and dried. An orange color 
persisted, so the sterol was dissolved in ethanol and decolorized by the addition of a 
small amount of Norit. The recovered sterol weighed 3.22 g., m.p. 157-160°C. After 
2 recrystallizations from ethyl acetate and 3 from ether in a Skau tube, small white 
plates were obtained, m.p. 160°C., [<x]d — 132°. Ergosteryl acetate and ergosteryl 
benzoate were prepared by the usual methods. Acetate: m.p. 172°C., [c*]d — 89°. 
Benzoate: m.p. 164°C., [<x]d — 64°. 

The ultraviolet absorption spectrum of the sterol (0.0599 mg./cc. in 
absolute ethanol) was determined on a Beckmann ultraviolet spectro¬ 
photometer. Peaks were found at 272 m/i, 282 m/*, and 294 m/*; this is in 
excellent agreement with the spectrum of authentic ergosterol deter¬ 
mined on the same instrument. The spectra were found to be practically 
identical. 

In subsequent determinations, the ergosterol concentration in 
Neurospora crassa was found, by quantitative digitonin precipitation, 
to be approximately 0.13% of the total dry weight of the mycelia. 

Acknowledgments 

The author wishes to express his appreciation to Prof. Werner Bergmann of the 
Department of Chemistry, Yale University, for his invaluable assistance and advice, 
and also to Dr. David Bonner of the Department of Botany and Microbiology, Yale 
University, for Neurospora mycelia. 

These investigations were supported, in part, by a grant from the Williams-Water- 
man Fund for the Combat of Dietary Diseases. 

Summary 

1. The principal sterol of Neurospora crassa has been isolated. 

2 . This sterol has been identified as ergosterol by means of the 
physical constants of the sterol, the steryl acetate, and the steryl ben¬ 
zoate; and also by means of the ultraviolet absorption spectrum. 

3. The ergosterol concentration in N. crassa, grown in the manner 
described, has been found to be approximately 0.13% of the dry 
mycelial weight. 

References 

1. Preuss, L. M., Peterson, W. H., SteenbociC, H., and Fred, E. B., J. Biol. 

Chem. 90 , 369 (1931). 

2. Bonner, D., Proc. Nall. Acad. Sci. U. S. 34, 5 (1948). 

3. Beadle, G. W., and Tatum, E. L., Am. J. Botany 32, 678 (1945). 



The Influence of Growth and Adrenocorticotropic 
Hormones on the Fat Content of the Liver 1 

Choh Hao Li, Miriam £. Simpson and Herbert M. Evans 

From the Institute of Experimental Biology , University of Califomia } 

Berkeley , California 
Received January 17, 1949 

Introduction 

The increase in liver fat following injections of certain anterior 
pituitary extracts into fasting rats has been known for some time (1,2). 
It was not clear, however, which of the known factors in pituitary ex¬ 
tracts is responsible for this effect. In fact, a separate ketogenic factor 
in the anterior pituitary has been suggested. It is the purpose of this 
paper to report the influence of acute dosage with pure growth and 
adrenocorticotropic hormones on the liver fat content in fasting normal 
and hypophysectomized rats. 

Experimental 

Experiments in Normal Rats 

Male rats of 40 days of age were fasted 24 hr.; on the second day of fasting, a total 
dose of 5.0 mg. of growth or adrenocorticotropic (ACTII) hormone, divided into 3 
injections at 2-hr. intervals was administered intraperitoneally and the animals killed 
2 hr. after the last injections. The livers were removed, weighed and placed into a 
dry-ice bath. The frozen tissues were dried in vacuo. The difference in weight before 
and after drying was taken as the water content. The dried powder was then extracted 
continuously by petroleum-ether in a Soxhlet apparatus. The ether extract was evap¬ 
orated and the residue was dried to constant weight which was taken as the amount 
of total fat. The protein content in the fat-free dried tissue was estimated by multi¬ 
plying Kjeldahl nitrogen by 6.25. The growth and adrenocorticotropic hormones were 
prepared by methods previously described (3,4). 

It should be mentioned that the animals were fed ad libitum before fasting. The 

1 Aided by grants from the American Cancer Society (through the National Re¬ 
search Council, Committee on Growth); the U. S. Public Health Service RG-409, and 
the Research Board of the University of California, Berkeley, California. 


51 


52 


C. H. LI, M. E. SIMPSON AND H. M. EVANS 


diet contained 19.3% protein, 56.7% carbohydrate, and 6.5% fat and was derived 
from 68.5% ground whole wheat, 5% casein, 10% alfalfa leaf meal, 5% fish oil, and 
1.5% NaCl. 

The results are summarized in Table I. It will be seen that the livers 
of rats receiving either the growth or adrenocorticotropic hormone have 
a higher fat content than do the livers of the controls. Adrenocortico¬ 
tropic hormone appears to be somewhat more effective in this respect. 
The livers of the control animals contained 2.59% fat, whereas the 
amount of fat in the ACTH-treated livers was approixmately double 
this value (5.39%). On the other hand the liver fat of the growth-hor¬ 
mone injected rats was elavated from 2.59 to 4.27%. Statistical analy¬ 
sis of these differences showed them to be highly significant. 


TABLE I 

Effect of Growth and Adrenocorticotropic (ACTH) Hormones on the 
Liver Composition of Fasted Male Rats 


Hormone 

No. of 

Body weight 


Liver 

Composition: G./100 g. liver 

injected® 

rats 

at autopsy 

G. 

G./100 g. 
b. wt. 

Water 

Fat 

Protein 

Growth 

8 

0 . 

154.1 ±9.1 b 

6.37 

4.12 ±0.15 

70.20 ±0.54 

4.27 ±0.39 

21.2 ±0.31 

ACTH 

12 

129.0 ±7.3 

5.82 

(>0.7)‘ 
4.51 ±0.15 

(>0.3) 
69.65 ±0.46 

(0.001) 

5.39 ±0.35 

(>0.05) 
19.97 ±0.74 

Control 

20 

133.5 ±5.1 

7.67 

(0.1) 

4.13 ±0.15 

(<0.1) 
70.97 ±0.47 

(<0.001) 
2.59 ±0.21 

(<0.05) 
21.66 ±0.33 


“ 40 days old male rats were fasted 24 hr.; on the second day of fasting, 3 injections 
with a total dose of 5.0 mg. growth hormone or ACTH were administered intraperi- 
toneally to the animals, in 2-hr. interval. Animals were killed 2 hr. after the last 
injection. 

1 Mean ± standard error. 

e Fisher’s p value. 

It may be noted that the liver weight was not changed appreciably 
by the injections. The water content remained unchanged in spite of 
the increment of the liver fat. There appears to be a slight decrease in 
liver protein following the administration of adrenocorticotropic 
hormone. 


Experiments in Hypophysectomized Rats 

Similar experiments were carried out oh male rats hypophysecto¬ 
mized at 40 days of age. Fasting began 6 days after operation. The 
results are shown in Table II. It will be observed that the animals 




LIVER FAT 


53 


TABLE II 


Effect of Growth and Adrenocorticotropic {ACT 11) Hormones on the 
Liver Composition of Fasted Hypophysectomized Male Rats 


Hormone 

No. of 

Body weight 
at autopsy 


Liver 

Composition: G./100 g. liver 

injected® 

rats 

G. 

G./100 g. 
b. wt. 

Water 

Fat 

Protein 

Growth 

10 

121.0*4.8* 

4.13 

3.43 ±0.09 

70.93 *0.40 

3.38*0.25 

19.07*0.25 

ACTH 

8 

105.9*3.1 

3.69 

(<0.02) e 
3.49 *0.11 

(<0.01) 
69.02 *0.27 

(<0.001) 
3.09*0.25 

(<0.001) 
21.03*0.15 

Control 

16 

114.63 *3.8 

3.69 

(<0.05) 
3.22 *0.06 

(<0.02) 
69.82 ±0.17 

(0.001) 

2.19*0.10 

(<0.01) 
22.00*0.17 


■ Hypophysectomized male rats (operated at 40 days of age and 6 days postopera¬ 
tive) were fasted 24 hr.; on the second day of fasting, 3 injections with a total dose of 
5.0 mg. growth or adrenocorticotropic hormone were administeree intraperitoneally 
to the animals in 2-hr. interval. Animals were killed 2 hr. after the last injections. 

* Mean ± standard error. 

c Fisher’s p value. 

receiving either growth or adrenocorticotropic hormone continued to 
show a definite increase in liver fat. The increase is about the same in 
both groups. While the liver weight per 100 g. body weight appears less 
in the hypophysectomized rats than that for unoperated animals there 
is no significant difference in the composition of protein and water. 
However, the fat content seems to be higher in the normal liver, a 
value of 2.59% vs. 2.19% for the hypophysectomized rats. 

Discussion 

The earlier experiments of Best and Campbell (1,2) established the 
fact that certain anterior pituitary extracts produce a rapid increase in 
size and an intense fatty infiltration of the liver in fasting rats. These 
observations have been confirmed subsequently by other investigators 
(5,6). It was further shown by MacKay and Barnes (5), and Fry (6), 
that the fatty liver normally produced by the administration of anter¬ 
ior pituitary extract does not occur following adrenalectomy. Thus, it 
may be said that the production of fatty liver is mediated by the adre¬ 
nal cortex. The data herein reported using adrenocorticotropic hormone 
agree with this conclusion. In fact, Baker et al., (7) have already shown 
that fatty infiltration occurs in adult rats treated with adrenocortico¬ 
tropic hormone, as demonstrated by histochemical technique. The fact 
that adrenocorticotropic hormone directly or indirectly affects fat 



54 


C. H. LI, M. E. SIMPSON AND H. M. EVANS 


metabolism has also been demonstrated by Bennett et al. (8). These 
investigators found that adrenocorticotropic hormone exhibits keto- 
genic activity in fasting rats and that the ketogenic effect of the hor¬ 
mone disappears in adrenalectomized animals. 

The fact that growth hormone might cause an increase in liver fat 
could possibly have been expected. Earlier investigators working with 
only partially purified growth preparation (9,10) found growth-pro¬ 
moting and ketogenic activities were associated. The recent data of 
Bennett et al. (8) have clearly shown that pure growth hormone pos¬ 
sesses ketogenic activity. 

In unfasted normal or hypophysectomized rats treated with growth 
hormone for 10 days or more, we have found in each case a decrease of 
the fat content in the liver (11). The present experiment with fasting 
and acute dosage gives an opposite result. These data may be looked 
upon as indicating that the growth hormone causes at first a mobili¬ 
zation of depot fat to the liver, the fat being then either utilized for 
protein synthesis or oxidized as an energy source under further injec¬ 
tions of the hormone. 


Summary 

The composition of the liver of fasting normal or hypophysecto¬ 
mized rats acutely treated with adrenocorticotropic or growth hormone 
has been analyzed. It has been shown that both growth and adrenocor¬ 
ticotropic hormone cause increase in liver fat with less significant 
changes in the water and protein content. 

References 

1. Best, C. H., and Campbell, J., J. Physiol. 86, 190 (1936). 

2. Best, C. H., and Campbell, J., ibid. 92, 91 (1938). 

3. Li, C. H., Evans, H. M., and Simpson, M. E., J. Biol. Chem. 169, 353 (1945). 

4. Li, C. H., Evans, H. M., and Simpson, M. E., ibid. 149, 413 (1943). 

5. MacKay, E. M., and Barnes, R. H., Am. J. Physiol. 118, 525 (1937). 

6. Fry, E. G., Endocrinology 21, 283 (1937). 

7. Baker, B. L., Ingle, D. J., Li, C. H., and Evans, H. M., Am. J. Anal. 82, 75 

(1948). 

8. Bennett, L. L., Kreiss, R. E., Li, C. H., and Evans, H. M., Am. J. Physiol. 

162, 210 (1948). 

9. Greaves, J. D., Freiberg, J. K, and Jones, H. E .;J. Biol. Chem. 133, 243 (1940). 

10. Shipley, R. A., and Seymour, W. B., Endocrinology 31, 634 (1942). 

11. Unpublished data of this laboratory. 



The Enzymatic Synthesis of Glucose-1,6-Diphosphate 

A. C. Paladini, R. Caputto, L. F. Leloir, R. E. Trucco and C. E. Cardini 

Institute de Investigaciones Bioquimicas, Fundacidn Campomar 
Jutiin Alvarez 1719, Buenos Aires, Argentina 
Received February 7, 1949 

Introduction 

Evidence on the identity of the coenzyme of phosphoglucomutase 
(1) with glucose-1,6-diphosphate has been reported previously (2). 
This substance has been found to be synthesized by yeast under the 
same conditions which lead to an accumulation of fructose-1,6-diphos¬ 
phate, and this was the starting point of this investigation in which the 
enzymatic synthesis of the coenzyme has been studied by incubating 
adenosinetriphosphate (ATP) with different sugar derivatives in the 
presence of muscle and yeast enzymes. 

The coenzyme can be estimated fairly accurately by a method based 
on the acceleration of the phosphoglucomutase reaction: glucose-1- 
phosphate —* glucose-6-phosphate. This is a very sensitive method by 
which amounts of the order of 10~ 10 moles can be measured, and it 
permits detection of the synthesis even with crude enzymes, in which 
case it is a very slow process in comparison with other competing re¬ 
actions which transform the sugar derivatives. 

The reaction mixture which gave by far the best yield of coenzyme 
was that of glucose-l-phosphate with ATP and the enzyme, and al¬ 
though the identity of the coenzyme was not known with certainty, at 
that time, the reaction was formulated provisionally as follows: 

glucose-l-phosphate+ATP—>glucose-l,6-diphosphate-j-ADP. (I) 

With crude enzymes the yield was only about 0.5% of the glucose-l- 
phosphate added and this low yield was mainly due to the presence of 
phosphoglucomutase which disposed of most of the substrate. 

The muscle enzyme could be purified sufficiently, so that, as much 
as 80% of the glucose-l-phosphate was transformed into coenzyme and 
it has been possible to test the validity of the previous equation. 

65 



56 


PALADINI, CAPUTTO, LELOIR, TRUCCO AND CARDINI 


According to this equation, one acid-labile phosphate group should 
become stable since the reactants contain 3 acid-labile groups and the 
reaction products only two. Moreover, the amount of glucose diphos¬ 
phate should be equivalent to half of the acid-labile phosphate of ATP 
and one acid equivalent should be liberated per molecule of glucose 
diphosphate. These 3 conditions were found to agree with the experi¬ 
mental data. 

The above mentioned reaction is similar to the hexokinase and phos- 
phohexokinase reactions and it seems justified to call the enzyme: 
glucose-l-phosphate kinase. It has a pH optimum at 6.8, and it is 
activated by magnesium and manganese ions. Glucose and glucose-l- 
phosphate are phosphorylated in position 6 by hexokinase and glucose- 
l-phosphate kinase, respectively. The possibility that they were one 
and the same enzyme was investigated, and it was found that the two 
effects could be separated. 


Methods 

Substrates and Analytical Methods 

Methods of estimation were as follows. Phosphoglucomutase and cophosphogluco- 
mutase as described previously (2). Hexokinase by an adaptation of the method des¬ 
cribed by Trucco et al. (3) for galactokinase. Phosphate: Fiske and SubbaRow (4). 
Acid-labile phosphate after 7 min. hydrolysis in 1 A H2SO4 at 100°C. Protein: Kunitz 
and MacDonald (5). Glucose-6-phosphate was prepared from glucose-l-phosphate 
according to Colo wick and Sutherland (6). Glucose-l-phosphate was obtained by 
chemical (7) or enzymatic (8) synthesis. Fructose-6-phosphate plus fructose-l-phos- 
phate according to Macleod and Robison (9). 

Estimation of the Enzyme 

Unless otherwise stated, the reaction mixture for the determination of the enzyme 
was as follows: 0.5 »M of glucose-l-phosphate, 0.37 i*M of ATP, 1 nM of MgS04,* 
total volume, 0.1 ml. All solutions adjusted to pH 7.0. The reaction was started by 
adding variable amounts of enzyme. After 10 min. at 37°C., the reaction was inter- 
Oipted by immersing the tubes in a boiling water bath. With the crude extracts it was 
often necessary to centrifuge and take an aliquot of the supernatant. Glucose diphos¬ 
phate was then estimated by addition of glucose-l-phosphate and yeast phospho¬ 
glucomutase as described previously (2) using 70% pure glucose-1,6-diphosphate as 
standard. Blanks to measure any preformed coenzyme and reducing substances 
formed during the first reaction were run at the same time. These consisted of samples 
in which the glucose-l-phosphate kinase reaction was stopped at 0 time, and samples 
which were incubated but to which no phosphoglucomutase was added. 



ENZYMATIC SYNTHESIS 


57 


Criteria of Purification 

The ratio of activities: glucose- 1 -phosphate kinase/phosphoglucomutase was 
generally used as a criterion of purification. As mentioned in the introduction, the 
low yield in glucose diphosphate with crude extracts was due to the competition of 
other enzymes for the same substrates. With the yeast extracts, the activity of phos- 
phoglucomutase explained almost quantitatively the lack of recovery, and, for this 
reason, the above-mentioned ratio was used to measure the purification. 

Another criterion was the maximum yield of coenzyme relative to the glucose- 1 - 
phosphate added. It was measured in the same mixture as that used for the estimation 
of activity but with an excess of ATP (0.7 yM). Aliquots were withdrawn from the 
mixture every 30 min., and the glucose diphosphate formed was measured until a 
maximum was reached. 

Experimental 

Specificity of Glucose-1-Phosphate 

Several sugars and hexose phosphates were incubated with Lebedew’s juice and an 
excess of ATP. Glucose, fructose and their monophosphates increased the formation 
of coenzyme. Glucose- 1 -phosphate seemed to be the J)est substrate, but the results 
were not conclusive. However, when the maceration juice was fractionated with 
(NH 4 ) 2 SO 4 and the fraction which precipitates at 45% saturation tested, it became 
clear that glucose-1-phosphate gave the largest yield of coenzyme. When muscle en¬ 
zyme was purified no detectable amount of coenzyme was synthesized with any sub¬ 
strate except glucose- 1 -phosphate (Table I). 

TABLE I 

Coemyme Formation from Different Substrates 

1 yM of substrate, 0.5 yM of ATP, 2.0 yM MgSCL, and 0.005 ml. of yeast enzyme, 
or 0.01 ml. of muscle enzyme, were incubated for 10 min. at 37°C. Total volume, 0.5 ml. 
Results in moles 10~ 10 . 

Substrate Cophosphoglucomutase formed 



Yeast enzyme 

Muscle enzyme 

None 

0.0 

0.0 

Glucose-l-phosphate (no ATP) 

0.05 

0.0 

Glucose-l-phosphate 

4.00 

>8.0 

Glucose 

1.83 

— 

Fructose 

1.03 

— 

Glucose-6-phosphate 

0.38 

0.0 

Glucose-6-phosphate -f glucose- 

4.08 

>8.0 

l-phosphate 

Fructose-l-phosphate and 

0.22 

— 

fructose-6-phosphate 

Fructose-l-phosphate and 

3.03 

— 

fructose-6-phosphate+glu- 
cose-l-phosphate 

Glucose-6-phosphate+fructose- 

— 

0.0 


6-phosphate 



58 


PALADINI, CAPUTTO, LELOIR, TRUCCO AND CARDINI 


Purification of Yeast Enzyme 

A maceration juice of dried brewers’ yeast (Cerveceria palermo) was prepared accord¬ 
ing to Neuberg and Lustig (10). To 400 ml. of this juice 1 N acetic acid was added 
until pH 4.9 was reached. The small isoelectric precipitate was centrifuged at 6000 
r.p.m. and the supernatant discarded. The precipitate was dissolved in 20 ml. of 
water, and the pH raised to 6.8. The solution was then precipitated by adding 0.9 
volumes of saturated (NHOiSOi and the precipitate dissolved in 1-2 ml. of water. By 
this procedure the relation phosphoglucokinase/phosphoglucomutase was increased 
about 40 times, and the recovery of glucose-1-phosphate as coenzyme increased from 
less than 0.6% in the maceration juice to 12-17% in the purified solution (Table II). 
Unfortunately, the loss of enzyme was so great (about 95%) that further attempts of 
purification of the yeast enzyme were discontinued. 

TABLE II 


Purification of Yeasts 9 Phosphoglucokinase 
Results in moles X 10~ 9 of reaction product. 



i 

Coenzyme formed 

Glucose-C-phos- 
phate formed 

Maximum. Per cent 
glucose-1-phos¬ 
phate recovered 
as coenzyme 

Lebedew juice 

1.2 

150.0 

M).5 

Isoelectric precipitation 

2.2 

40.0 

— 

Ammonium sulphate precipitate 

3.0 

10.0 

12.0 


Purification of Rabbit Muscle Enzyme 

Rabbits were killed by a blow and bled. The muscle was passed through a cooled 
mincer and suspended with 2 volumes of cold water. After 10 min. extraction it was 
strained through muslin and extracted again with 1 volume of water. A third extrac¬ 
tion was made with 1 volume of salt solution containing 0.5 M KC1 and 0.1 M phos¬ 
phate buffer pH 7.4. This procedure was carried out with 14 rabbits. In 4 of them the 
first water extract was the richest in phosphoglucokinase (highest value 1.2 X 10 _s 
pM of glucose diphosphate synthesized by 0.01 ml. in 10 min.); and in 5 cases the 
saline extract was the richest (highest value 2.2 X 10“ 3 pM of glucose diphosphate 
synthesized by 0.01 ml. in 10 min.); finally, in 3 oases, no appreciable activity was 
obtained in any of the 3 extracts. In every case the saline extract was chosen for 
further purification, because even in those instances in which the activities were 
poorer than in the aqueous extracts, the relation kinase/mutase was always highest. 

In the saline extract the phosphoglucokinase behaves as if it were 
loosely adsorbed on myosin. When this solution was dialyzed as a thin 
layer against distilled water, protein started to precipitate after about 
1 hr., and finished in about 4 hr. This precipitate, which was mainly 







ENZYMATIC SYNTHESIS 


59 


myosin, carried down the phosphoglucokinase activity. On the con¬ 
trary, when myosin was precipitated with (NH 4)2804 at 33% satura¬ 
tion, practically no activity followed the precipitate. Fifty per cent 
saturated (NH 4 ) 2 S 0 4 is necessary to precipitate the activity (Table 
III). 

TABLE III 

Precipitation 0 / the Muscle Enzyme by Dialysis and Ammonium Sulphate Fractionation 

Results in moles X 10~ 10 . 


Treatment 

Glucose diphosphate formed 

Protein/mg./ml. 

Supernatant 

Precipitate 

Supernatant 

Precipitate® 

Dialysis 





2 hr. 

1.2 

2.0 

— 

— 

4 hr. 

0 

3.8 

1 

0.73 

Ammonium sulphate 





treatment 





33% saturation 

— 

0.7 

— 

1.1 

50% saturation 

— 

>8.0 

— 

0.9 


• Dissolved in 0.05 M phosphate buffer pH 7.4. 


Precipitation with (NH 4 ) 2 S0 4 was the best way of purifying the 
saline extracts. To 110 ml. of this extract, 80 ml. of saturated (NH 4 ) 2 S 0 4 
solution, brought to pH 7.5 with NH 4 0H, was added. The precipitate 
was separated by filtration through fluted paper in the ice-box. One 
hundred fifty ml. of clear filtrate were obtained, to which 25 ml. of 
(NH 4 ) 2 S0 4 (pH: 7.5) were added. The precipitate was now collected 
by filtration and dissolved in water to make 8 ml. of solution. 

As shown in Table IV, this procedure improves the relation kinase/ 
mutase about 40 times, and the relation kinase/protein 4-fold. The 
liberation of inorganic phosphate effected by the crude enzyme was not 
detectable with the purified solution. 

Relation between Hexokinase and Glucose-1-Phosphate 
Kinase Activities 

The possibility that the hexokinase and the glucose-l-phosphate 
kinase reactions were catalyzed by the same enzyme was investigated 
by measuring both activities in different extracts. 





60 PALADINI, CAPUTTO, LELOIR, TBUCCO AND CARDINI 

TABLE IV 

Activities of the Muscle Enzymes during Purification 
Tests as described in text, referred to 0.01 ml. enzyme. Results in moles X 10 - ' of 
reaction product. 



Coensyme 

formed 

Glucose-6- 

phosphate 

formed 

Inorganic 

phosphate 

liberated 

Protein 
mg. /ml. 

Aqueous extract (rejected) 

0.24 

12.0 

— 

— 

Crude saline extract 

0.6 

8.0 

2.0 

8.0 

Fraction insoluble in 0.33 
saturated ammonium sul¬ 
phate 

0.35 

2.0 

3.0 


Fraction precipitated in 0.5 
saturated ammonium sul¬ 
phate 

3.6 

1.2 

0 

10.5 


These two activities were measured in: (a) Crude Lebedew juice; 
(b) A solution of hexokinase obtained from brewers’ yeast by the 
method of Berger et al. (11) and partially purified by acetone precipi¬ 
tation; and (c) The partially purified preparations of glucose-l-phos- 
phate kinase. 

The results are shown in Table V. The purified hexokinase had the 
lowest glucose-l-phosphate kinase activity. On the other hand, the 
ratio hexokinase/glucose-l-phosphate kinase was about 30 times 
lower in the purified glucose-l-phosphate kinase than in Lebedew 
juice. 

The two activities “were thus found to change independently from 
the point at which it was concluded that different enzymes were in¬ 
volved in these reactions. 


TABLE V 

Hexokinase and Glucose-l-Phosphate Kinase Activities 
in Different Enzyme Preparations 

Values referred to 0.01 ml. enzyme. Results in moles X 10“* of reaction product. 


Preparation 

Hexokinase 

Glucose-l-phos¬ 
phate kinase 

Lebedew juice 

10200 

0.7 

Purified glucose-l-phosphate 
kinase from yeast 

1240 

2.4 

Partially purified hexokinase 

8000 

0.08 

Purified glucosc-l-phosphate 
kinase from muscle 

0 

3.7 



ENZYMATIC SYNTHESIS 


61 


Yield of Coenzyme in Relation to the Amount 
of Adenosinetriphosphate 

Known amounts of ATP were incubated at 37°C. with glucose-1- 
phosphate, magnesium and the enzyme. Samples were withdrawn at 
various times, and the reaction stopped by heating. The amount of 
coenzyme was then estimated at suitable dilutions as usual. The 
enzyme used was a purified preparation from muscle in which the 
(NHOiSO* fractionation had been repeated twice and which was 
practically free from adenosinetriphosphate and phosphatase activity. 



Fig. 1 . Yield of coenzyme in relation to the amount of ATP. Curve 1. Incubation 
of 0.8 nM glucose-l-phosphate + 1.5 nM magnesium + 0.4 nM ATP + 0.01 ml. of 
enzyme. At point A, 0.2 nM ATP added to an aliquot. Curve 2. Same as 1 but with 
0.8 of ATP. 

The results of one of several experiments appear in Fig. 1, where 
Curve 1 shows that the maximum amount of coenzyme formed is 
equal to the amount of ATP, in agreement with Eq. 1. That the reaction 
stopped due to the lack of ATP was proved by adding more ATP to an 
aliquot, and finding that more coenzyme was synthesized. 




62 


PALADINI, CAPUTT0, LELOIR, TRUCCO AND CARD INI 


Curve 2 in Fig. 1 shows the same type of experiment, but with double 
the amount of ATP. Here the end of the reaction was not attained, and 
the reaction seemed to take place more slowly. This inhibition by higher 
concentration of ATP has been repeatedly observed. 

Yield of Coenzyme in Relation to the Decrease in 
Acid-Labile Phosphate 

According to Eq. 1, one mole of acid-labile phosphate should become 
acid-stable per mole of coenzyme formed. Experiments were carried out 
in which the acid-labile phosphate and coenzyme concentration were 
estimated during the course of the reaction. 

As shown in Fig. 2, the results agree with the proposed equation. 

Yield of Coenzyme in Relation to Acid Production 

Colowick and Kalckar (12) showed that, in the hexokinase reaction, 
one acid equivalent is liberated for every mole of phosphate trans- 



Fig. 2. Yield of coenzyme in relation to the increase in acid-stable phosphate. 
Incubation of 0.86 pM glucose-l-phosphate-1-0.56 pM ATP+1.5 pM MgSO«+0.01 
ml. of purified muscle enzyme. Reaction stopped by heating. 




ENZYMATIC SYNTHESIS 


63 



ferred from ATP to glucose. In the similar equation proposed for the 
synthesis of cophosphoglucomutase one acid equivalent should be 
produced for every mole of coenzyme formed. 

The acid production was measured manometrically in a Warburg 
apparatus. In the main compartment were added: 10 tiM of glucose-1- 
phosphate; 5 uM of MgS04j 3.7 pM ATP; 0.2 ml. of 0.5 M bicarbonate, 
and 0.1 ml. toluene. In the side bulb 0.2 ml. of purified muscle glucose- 
1 -phosphate kinase. Total volume, 2.5 ml. Gas phase 95% N + 5% 
COj. A control without glucose-l-phosphate was run at the same time. 
After 35 min. the reaction was finished as judged by the pressure, 
which remained constant for 15 min. The C0 2 evolved was 77 #*1 
corresponding to 3.5 microequivalents of acid. The enzymatic deter¬ 
mination gave 3.8 pM of coenzyme. 

Optimum pH Curve 

Maleate “buffer” according to Smits (13) was used after it was 
ascertained that it does not interfere with the activity of the enzyme. 
The experiment of Fig. 3 was carried out with the muscle enzyme, and 




64 


PALADINI, CAPUTTO, LELOIR, TRUCCO AND CARDINI 


shows that the optimum is at pH 6.8. With the yeast enzyme a closely 
similar curve was obtained. 

Activation by Magnesium and Manganese 

Fig. 4 shows that both Mg ++ and Mn ++ activate glucose-l-phos- 
phate kinase. With magnesium the enzyme is fully activated at about 
1.5 X 10 -3 mole/1. Manganese is slightly less effective. 



concentration in moles x 10“^ per liter 

Fig. 4. 

Since the variable concentration of ions might affect the activity of 
phosphoglucomutase during the estimation of glucose diphosphate, 
in these experiments, the amount of salt was made the same in all the 
tubes, including the coenzyme standard, before adding the phospho¬ 
glucomutase. 

Discussion 

The evidence presented seems sufficient to show that in yeast and 
animal tissues glucose diphosphate is formed by the interaction of 
adenosinetriphosphate with glucose-1-phosphate. This raises the 






ENZYMATIC SYNTHESIS 


65 


question as to whether this reaction can be a quantitatively important 
pathway in carbohydrate dissimilation, or if it is only limited to the 
formation of coenzyme for phosphoglucomutase. The evidence at hand 
favors the second alternative. In crude extracts of yeast, the reaction 
catalyzed by hexokinase is several thousand times faster than that of 
glucose-l-phosphate kinase, and in either muscle or yeast the relation 
phosphoglucomutase/glucosc-l-phosphate kinase is about 100, even 
under conditions devised to favor the activity of glucose-l-phosphate 
kinase. 

The possibility remains that the glucose diphosphate is destroyed 
almost as fast as it is formed, but preliminary experiments indicate that 
this is not the case with muscle. However, the question is not definitely 
settled since extracts of liver, kidney, brain, and Escherichia coli 
destroy glucose diphosphate. 

Another question which arises is whether the reaction of glucose-l- 
phosphate and ATP is the only mechanism by which glucose diphos¬ 
phate is formed. Experiments indicate that E. coli can synthesize 
glucose diphosphate by the transference of phosphate between two 
molecules of glucosc-1-phosphate without the intervention of ATP. 

Summary 

The enzymatic syntheses of glucose diphosphate have been studied 
with extracts of yeast and rabbit muscle. The best yields were obtained 
on incubation of the enzyme with adenosinetriphosphate and glucose- 
l-phosphate. 

Using partially purified extracts for the estimation of acid-labile 
phosphate, the total yield of glucose diphosphate and the acid forma¬ 
tion agree with the equation : 

ATP -f- glucose-l-phosphatc —> glucose diphosphate + ADP. 

Magnesium and manganese ions accelerate the reaction. The opti¬ 
mum pH is 6.8. 


References 

1. Caputto, R., Leloir, L. F., Trucco, R. E., Cardini, C. E., and Paladini, A., 

Arch. Biochem. 18 , 201 (1948). 

2. Leloir, L. F., Trucco, R. E., Cardini, C. E., Paladini, A., and Caputto, R., 

ibid. 19 , 339 (1948) ; Cardini, C. E., Paladini, A. C., Caputto, R., Leloir, 
L. F., and Trucco, R. E., ibid. 22, 87 (1949). 



66 


PALADINI, CAPUTTO, LELOIR, TRUCCO AND CARDINI 


3. Trucco, R. E., Caputto, R., Leloir, L. F., and Mittelman, N., ibid. 18, 137 

(1948). 

4. Fiske, C. H., and SubbaRow, Y., J. Biol. Chern. 66, 375 (1925). 

5. Kunitz, M., and MacDonald, M. J., J. Gen. Physiol. 29, 393 (1946). 

6. Colowick, S. P., and Sutherland, E. W., J. Biol. Chern. 144, 423 (1942). 

7. Cori, C. F., Colowick, S. P., and Cort, G. T., ibid. 121* 465 (1937). 

8. Sumner, J. B., and Somers, G. F., Arch. Biochem. 4, 11 (1944). 

9. Macleod, M., and Robison, R., Biochem. J. 27, 286 (1933). 

10. Neuberg, C., and Liistig, H., Arch. Biochem. 1, 191 (1942). 

11. Berger, L., Slein, M. W., Colowick, S. P., and Cori, C. F., J. Gen. Physiol. 29, 

379 (1946). 

12. Colowick, S. P., and Kalckar, II. M., J. Biol. Chem. 148, 117 (1943). 

13. Smits, G., Biochem. Biophys. Ada 1, 280 (1947). 



Studies on the Mechanism of the Inhibition of 
Glucolysis by Glyceraldehyde 1 

Harry Rudney 2 , 

From the Banting and Best Department of Medical Research, 

University of Toronto , Toronto, Canada 
Received February 11, 1949 

Introduction 

Twenty years ago, Mendel (1) discovered the inhibitory effect of 
glyceraldehyde on glucolysis. He found that 10~ 3 M d/-glyderaldehyde 
inhibited the anaerobic glucolysis of tumor cells almost completely 
without affecting their respiration or the respiration of normal cells. 
Later, Mendel, Bauch and Strelitz (2) showed that the inhibitory effect 
of glyceraldehyde is completely abolished by 10~ 3 M pyruvic acid and 
partially reversed by smaller amounts of this compound. These findings 
have been confirmed and extended to other tissues (9). 

Studies by Adler (3) and Sullmann (4,5) suggested that glyceraldehyde interfered 
with the initial phosphorylation of glucose, a conclusion reached on the basis of the 
absence of any inhibitory effect of glyceraldehyde on the glycolysis of hexose mono- 
and diphosphates. 

Neither Holmes (6) nor Lehmann (7) could obtain an inhibition by glyceraldehyde 
of muscle glycolysis in extracts to which yeast hexokinase was added. Adler et at . (3) 
could obtain no inhibition of yeast hexokinase activity with the above compound. 
Stickland (8), on the other hand, found that the glucolysis of muscle extracts fortified 
with yeast hexokinase was inhibited by 3 X 10~ 3 M glyceraldehyde, and that the 
inhibition could be reversed by gradually increasing the concentration of hexokinase. 
Thus, the amount of hexokinase present seemed to be the critical factor in determin¬ 
ing the extent of the inhibition, and the aforementioned negative results were ex¬ 
plained on this basis. Stickland also found that pyruvic acid (4 X 10“ 4 M) could reverse 
the effect of glyceraldehyde. These results suggested that the inhibition of hexokinase 
activity was the means whereby glucolysis was inhibited by glyceraldehyde. 

Dorfman (9) has adequately reviewed and discussed the early literature on the 

1 This research was supported by a grant from the Ontario Cancer Treatment and 
Research Foundation. 

2 Present address: Department of Biochemistry, Western Reserve University, 
Cleveland, Ohio. 


67 


68 


HARRY RUDNEY 


inhibitory action of glyceraldehydc on glycolysis, and points out that several questions 
regarding the mechanism of the effect remain unanswered: (1) How does pyruvic acid 
reverse the inhibition? (2) If glyceraldehyde inhibits glucolysis, why are the same 
concentrations not effective in inhibiting respiration? This last point has been raised 
by Barker et al. (10) in the case of the lack of inhibition of respiration by iodoacctic 
acid in concentrations which inhibit glycolsyis. 

In view of these questions, it was decided to reexamine the problem 
of the mechanism of the inhibition of glucolysis by gfyceraldehyde, and 
the action of pyruvate thereon, by studying the effect of these sub¬ 
stances on the hexokinase of brain, muscle, tumor, and yeast. 

Experimental 

M ethods 

The method of Colowick et al. (11), whereby glucose disappearance is determined as 
a measure of hexokinase activity, was used. It is based on Nelson's colorimetric micro 
method (12) involving copper reduction and the use of Ba(OH) 2 and ZnS0 4 as protein 
precipitants, and has the desirable feature of eliminating all phosphorylated com¬ 
pounds during the protein precipitation. 

In the experiments to be reported below, the procedure described by Colowick et al. 
was followed closely, and the concentrations of the components of the test system 
were exactly as given by them, unless otherwise indicated. The final volume of the 
incubation mixture was 2.35 mi. and the incubation period was 20 min. at 30°C. 
Glyceraldehyde and pyruvate were always placed in the main compartment of the 
Warburg vessel. After deproteinization, glucose was determined according to the 
method of Somogyi (13), using Nelson's arsenomolybdate reagent. Measurements of 
optical density were made in a Colemen Junior Spectrophotometer at 550 m/*. 

The differences in reduction obtained before and after the experimental period were 
taken as reflecting the change in concentration of glucose. Terrors due to the disap¬ 
pearance of small amounts of glyceraldehyde during the course of the experiment were 
corrected by heating an aliquot of the deproteinized filtrate for 4 min. at 80°C., and 
then immediately cooling in ice water. Under these conditions most of the reduction 
due to glyceraldehyde had already taken place, while that due to glucose was just 
beginning. From the difference in optical density readings obtained between controls 
and experimental reaction mixtures at the above time and temperature, it was possible 
to determine the amounts of glyceraldehyde disappearing with an error of 10%. 
The difference in density readings obtained from the reaction mixtures treated as 
above was then subtracted from the difference in densities found after complete re¬ 
duction by glyceraldehyde and glucose had taken place in another aliquot of the same 
filtrate. Using this method of correcting for the disappearance of glyceraldehyde, it 
was found possible to determine with an accuracy of ± 5-8% the amount of glucose 
disappearing during an interval when small amounts of glyceraldehyde also disap¬ 
peared. This method was considered sufficiently accurate for the purpose of deter¬ 
mining whether a marked inhibition of hexokinase activity existed. In the experi¬ 
mental results tabulated below, differences in inhibition of less than 15% are not 



GLUCOLYSIS INHIBITION 


69 


considered significant. Experiments were usually arranged so that the differences in 
optical density readings in the absence of glyceraldehyde after heating for 10 min. at 
100°C. were very large as compared with those obtained after heating for 4 min. at 
80°C. All values listed in the tables below are corrected for glyceraldehyde disappear- 
ance. 


Materials 

The glyceraldehyde used was the dl -form prepared by Schering-Kahlbaum as a 
white powder. The latter was washed by refluxing with acetone for 30 min, filtered, 
and then dried. (M.l\ 138°C.) Solutions used for enzymatic experiments were heated 
for f) min. at 85°C. to ensure only the monomeric form being present. Needham and 
Lehmann (7) have shown that the dimeric form does not inhibit glucolysis. Pyruvic 
acid was added in the form of the sodium salt prepared according to Robertson (14). 
Adenosine triphosphate was prepared by the method of Kerr (15). Samples from com¬ 
mercial sources were also used. 3 

The hexokinase extracts from beef brain acetone powders and from rat skeletal 
muscle were prepared as described by Colowick et al. (11), the muscle extract being 
obtained by filtration through cheese cloth. Tumor hexokinase was prepared in two 
ways, as a homogenate and as an aqueous extract from an acetone powder of rat 
sarcoma 39. The former was made by homogenizing 1 g. of tumor tissue freed from 
necrotic areas with 1.5 ml. of distilled water at 5°C., while the latter preparation was 
made in the same manner as the brain acetone powder extract. The hexokinase from 
yeast was prepared according to the method of Meyerhof (16). 

Results 

The Effect of Glyceraldehyde and Pyruvate on the Hexokinase 
Activity of Beef Brain Extracts 

Experiments were performed to determine whether glyceraldehyde 
exerted an inhibitory effect on the activity of hexokinase from beef 
brain extract and the effect of pyruvic acid thereon. Table I shows the 
results of these experiments. From this table it will be seen that 
glyceraldehyde in concentrations varying from 2 X 10~ 3 il/ to 4.5X10' 3 
M exerts an inhibitory effect on hexokinase activity ranging from 44 
to 100%. These results support the assumption of earlier workers that 
glyceraldehyde acts by inhibiting the phosphorylation of glucose. 

The effect of pyruvate on the inhibitory action of glyceraldehyde on 
beef brain hexokinase is somewhat uncertain. In one case the inhibition 
was partially reversed (Expt. 5, Table I), while in the majority of cases 
no significant effect could be noted. Control experiments show that 
glyceraldehyde did not react with pyruvate in the presence of the 

8 Armour and Company, Chicago, Ill., and Nutritional Biochemicals, Cleveland. 
Ohio. 



70 


HARRY RUDNEY 


TABLE I 

The Effect of Glyceraldehyde and Pyruvate on the Activity of the 
Hezokinase from Beef Brain Cortex 





Normal y 

Per cent inhibition 

Experi¬ 
ment no. 

Molar cone, of 
glyceraldehyde 

Molar cone, of 
pyruvate 

of glucose 
disappear¬ 
ing 

With glyc¬ 
eraldehyde 

With pyru¬ 
vate +glyc¬ 
eraldehyde 

With 

pyruvate 

1 

4.5 X10"’ 


450 

100 



2 

4.5 X10 -3 


505 

86 



3 

2.2 XlO' 3 


465 

44 



4 


1X10" 4 

220 



0 

5 

3.5 XlO" 3 

1X10~ 3 

530 

52 

24 


6 

4.5X10- 3 

5 X10“ 4 

400 

82 

66 


7 

4.0 XlO" 3 

2 X10" 3 

375 

64 

52 


8 

2.2 X10" 3 

5 XlO 4 

500 

54 

56 


a 9 

4.5 XlO' 3 

2X10“ 3 

410 

53 

48 


«10 

4.5X10- 3 

2X10- 3 

580 

72 

50 


fc ll 

4.5 X10" 3 


630 

14 



b 12 

4.5 X10" 3 


485 

17 



13 

4.5 X10" 3 

4X10- 3 

325 

100 

100 


14 

4.5XKF 3 

8XIO- 3 

325 

100 ! 

100 



0.6 ml. of brain extract were added to the main compartment of the Warburg cup 
containing 1.0 ml. of 0.02 M MgCl 2 ,0.06 M NaHC0 3 , with water, glyceraldehyde, and 
pyruvate to a final volume of 1.9 ml. The sidearm contained 0.1 ml. 1% glucose, 0.15 
ml. 0.9 M NaF, and 0.05 M NaHC0 3 , and 0.2 ml. of 0.05 M adenosine triphosphate. 
Final volume 2.35 ml. Incubation period: 20 min. at 30°C. 

a Extract incubated for 20 min. at 20°C. with glyceraldehyde before reaction ini¬ 
tiated. 

b d-Glyceraldehyde used in these experiments. In all other cases the racemic form 
was employed. 


extract, nor did pyruvate display any reducing activity on the copper 
reagent. The lack of reaction of pyruvate with glyceraldehyde was 
found to be true for all other extracts from different sources mentioned 
forthwith. 

Tests with slices of beef brain cortex showed that the amounts of 
glyceraldehyde necessary to inhibit glucolysis of this tissue are much 
lower than those normally required to cause the same degree of inhi¬ 
bition of the hexokinase in such preparations. For example, 10“ 3 M 
glyceraldehyde, which has no effect on the activity of the extract, 
inhibits the anaerobic glucolysis of slices 80-90%, irrespective of the 





GLUC0LY8IS INHIBITION 


71 


temperature at which the estimation was carried out (37.5°C. or 30°C.). 
Meyerhof and Randall (17) obtained similar results in their studies on 
the inhibitory effect of adrenochrome on glycolysis. There may be 
several reasons for this phenomenon, which will be discussed further 
on in this paper. No effect of pyruvate could be noted on the inhibition 
of glucolysis in beef brain slices effected by glyceraldehyde. This fact 
harmonizes with the findings obtained with extracts. 

Needham (18) and Mendel (19) found that the l isomer of glycer- 
aldehydc was responsible for the inhibition of gucolysis. To ascertain 
which enantiomer was exerting the inhibitory effect in the investiga¬ 
tion reported here experiments were carried out with d-glyceraldehyde. 4 
Expts. 11 and 12, Table I, show that d-glyceraldehyde has a negligible 
effect on the activity of the hexokinasc. It can be concluded, therefore, 
that the l form only is responsible for the inhibition observed with a 
racemic mixture. It was noted, furthermore, that, during the course 
of the experiments, small amounts of d-glyecraldehyde disappeared, 
which were comparable to those amounts disappearing in experiments 
with (//-glyceraldehyde. It is, therefore, very probable that the glycer¬ 
aldehyde being metabolized in the experiments where the racemic form 
was employed was of the A form. 

The Effect of Glyceraldehyde on Tumor Hexokinasc Activity 

The effects of glyceraldehyde and pyruvate were determined on 
homogenates and on aqueous extracts of acetone powders of rat sar¬ 
coma 39. The results are shown in Table II. From this table it will be 
seen that 4.5 X 10~ 3 glyceraldehyde is needed to bring about a 50-60% 
inhibition of the purified hexokinase extracts, while the homogenates 
show only a 30% inhibition when the same concentration is used. The 
greater sensitivity of the extracts could be due to the lack of protective 
substances present in the homogenates. In this connection, it is inter¬ 
esting to note that Boyland (21) found that, with glycolizing tumor 
extracts, a 90% inhibition could be achieved only with 2.2 X 10“ J M 
glyceraldehyde. Pyruvate does not significantly affect the inhibition 
by glyceraldehyde in either type of preparation. The presence or ab¬ 
sence of fluoride does not seem to influence the lack of effect of pyruvate 
(Expts. 9 and 10, Table II). 

* The author is indebted to Dr. E. Baer for his kindness in supplying a sample of 
•d-glyceraldehyde used in this investigation. 



72 


HARRY RUDNEY 


TABLE II 

The Effect of Glyceraldehyde and Pyruvate on the Activity of 
Hexokinase from Rat Tumor 


Experi¬ 
ment no. 

Molar cone, of 
glyceraldehyde 

Molar cone, of 
pyruvate 

Normal y 
of glucose 
disappear¬ 
ing 

Pci 

With glyc¬ 
eraldehyde 

■ cent mhibiti 

With pyru¬ 
vate +glyc- 
eraldohyde 

on 

With 

pyruvate 

1 

1.0X10-* 

2.0 X10' 3 

480 

26 

9 

0 

2 

1.5X 10“ 3 

2 . 0 X 10- 3 

370 

31 

32 

0 

3 

3.0X10- 3 

4.0 X10' 3 

515 

12 

12 

0 

4 

4.5 XlO' 3 

4.0X 10~ 3 

470 

25 

16 

0 

5 

4.5X10* 3 

4.0 X10" 3 

400 

26 

26 


6 

4.5 X 10' 3 

4.0X10' 3 

500 

58 

58 


7 

4.5X10” 3 

4.0X10~ 3 

470 

55 

58 

0 

8 

4.5 X 10~ 3 

4.0 X10' 3 

690 

61 

71 


9 

4.5 XlO -3 

4.5 X10' 3 

890 

45 

45 


10 

4.5 XlO -3 

4.5 X10' 3 

860 

48 

48 

0 


In Expts. 1 to 5, 0.4 ml. of homogenate was used. In all other experiments, 0.4 ml. 
of an aqueous extract of acetone powder was used. In Expts. 0 and 10, no fluoride was 
present. Reaction time: 20 min. at 30°C. in all cases. Other conditions as described in 
footnote to Table I. 


When the action of glyceraldehyde and pyruvate was tested on the 
anaerobic glucolysis of tumor tissue slices, profound differences from 
the behavior of cell-free extracts were noted. The sensitivity of slices 
to small concentrations of glyceraldehyde was much greater than that 
observed with homogenates. Moreover, pyruvate definitely reversed 
the inhibition, the extent of the reversal varying according to the con¬ 
centration of glyceraldehyde used. Typical experiments are outlined in 
Table III. It was also found that, when the disappearance of glucoes 
was used as the criterion for the extent of glucolysis, similar results 
were obtained (Table IV). It is enteresting to note that in the second 
experiment of Table IV, 10 -4 M pyruvate was sufficient to abolish a 
59% inhibition by 10~ 3 M glyceraldehyde. In later experiments with 
tissue slices, where glucose disappearance was measured, it was also 
found that about 50% of the glyceraldehyde present had disappeared 
at the end of the experimental period, both in the absence and in the 
presence of pyruvic acid. It was not possible to asceratin with any 



GLUCOLYSIS INHIBITION 


73 


great accuracy the amounts disappearing and it was difficult to deter¬ 
mine whether only d-glyceraldehyde disappeared or the l form also. 
If more of the l form disappears in the presence of pyruvic acid, it 
would readily explain how the reversal of the inhibition is affected. 


TABLE III 


The Effect of Glyceraldehyde and Pyruvate on the 

Glucolysis 


of Slices of Rat Sarcoma 39 



Additions to normal medium 

Q^- 

Inhibition 
per cent 

(1) None 

30.0 

— 

10“ 3 M pyruvate 

34.5 

— 

1.5 X 10“ 3 M glyceraldehyde 

5.0 

84.0 

1.5 X10 -3 M glyceraldehyde-fl0~ 3 M pyruvate 

29.5 

0.0 

10~ 3 M glyceraldehyde 

7.0 

77.0 

10“ 3 M glyceraldehyde-flO” 3 M pyruvate 

33.5 

0.0 

5X10 -4 Af glyceraldehyde 

27.0 

10.0 

5X10~ 4 M glyceraldehyde -fl0~ 3 Af pyruvate 

35.5 

0.0 

(2) None 

20.0 

— 

4.0X10“ 3 M pyruvate 

22.5 

— 

4.5 X 10~ 3 M glyceraldehyde 

4.0 

80.0 

4.5Xl0~ 3 Af glyceraldehyde+■ 4 X 10“ 3 M pyruvate 

11.0 

45.0 


Glucolysis was measured at 37.5°C. in section (1) and at 30°C. in section (2). The 
medium consisted of Ringer's bicarbonate solution containing 200 mg.-% glucose. The 
gas phase was 05% N 2 -5% C0 2 . 


^C0 2 


•pi. C0 2 /hr./mg. dry weight, calculated on the basis of pressures obtained 


during the first 30 min. period after temperature equilibration. 


TABLE IV 


The Effect of Glyceraldehyde ( GA) and Pyruvate (PY ) on the Disappearance 
of Glucose Effected by Slices of Rat Sarcoma 39 



y of glucose disappearing/ing. of tissue 
(dry weight) in 2.5 hr. 

I 

II 

III 

Normal 

■MS 

■■ 

96 

Normal+10" 8 M Py 

HIS 


98 

Normal+10“ 8 M Py+10 -8 Af GA 

! 90 

110 

83 

Normal-fl0“ 4 M Py-fl0~ 3 M GA 


90 

45 

Normal-f 10 -3 M GA 

WM 

56 

52 


The medium used was the samo as that used for the experiments with tissue extracts 
except that no fluoride or adenosine triphosphate were present. 








74 


HARRY RUDNEY 


The Effect of Glyceraldehyde on Yeast Hexokinase 

The measurements of yeast hexokinase activity are the most accurate 
of all those reported in this paper, since it was found that no glycer¬ 
aldehyde disappeared during the course of the reaction. The prepara¬ 
tion obtained according to the procedure of Meyerhof had a dry weight 
of 55 mg./ml. 0.15 ml. of the 10-fold diluted preparation, corresponding 
to 0.83 mg. of protein, was used in the experiments. A similar prepara¬ 
tion of yeast hexokinase, used by Stickland (8), had a dry weight of 
18 mg./ml. and, since he added 0.2 ml. of extract (corresponding to 


TABLE V 

The Effect of Glyceraldehyde and Pyruvate on the Activity of 
the Hexokinase from Yeast 




Molar cone, of 
pyruvate 

Normal y of 
glucose disap¬ 
pearing 

Per cent inhibition 

Experiment 

no. 

Molar cone, of 
glyceraldehyde 

With glyc¬ 
eraldehyde 

With pyru¬ 
vate +gl> cer- 
aldehyde 

1 

1.0 X10 3 


550 

5 


2 

1.0 X10-* 


475 

20 


3 

3.0 X10 3 

4.0X10 3 

495 

13 

9 

4 

3.0X10 3 

3.0X10 3 

610 

56 

63 

5 

3.0X10 3 

3.0X10 3 

655 

30 

26 

6 

6.0 X10" 3 

6 .0X10 3 

480 

48 

48 

7 

6.0 X10- 3 

6.0 XIO' 3 

460 

59 

0 

8 

6.0 X10~ 3 

6 .0X10-’ 

380 

59 

59 

9 

6.0 X10~ 3 


515 

4 


10 

1 .2X10-* 


680 

50 



0.15 ml. of the 10-fold diluted extract was used as the source of enzyme. No fluoride 
was present in any of the above experiments. Other conditions as described in footnote 
to Table I. 


3.6 mg. of protein), the amount of yeast hexokinase used in the follow¬ 
ing experiments was much less than that used by Stickland. Despite 
this fact, the inhibition of yeast hexokinase requires extremely large 
amounts of glyceraldehyde (Table V). Concentrations which would 
inhibit muscle hexokinase 80-100% have a comparatively mild effect 
on the yeast hexokinase. Under the conditions of Stickland’s experi¬ 
ments, where yeast hexokinase was added to a muscle extract, 3 X 10~ 3 
M glyceraldehyde almost completely inhibited glucolysis. Pyruvic acid 



GLUCOLY8I8 INHIBITION 


75 


was found to be completely without effect on the inhibitory action of 
glyceraldehyde, except for one instance (Expt. 7, Table V). In no case 
could this experiment be repeated, however. The foregoing results with 
pyruvate appear to be directly at variance with those of Stickland, who 
found that 4 X 10 -4 M pyruvate could reverse completely the inhibition 
of glucolysis caused by 3 X 10 -3 M glyceraldehyde. However, it will 
be recalled that Stickland was measuring glucolysis presumably in the 
presence of an intact series of glycolytic enzymes from muscle. Thus, the 
action of glyceraldehyde on glycolizing cell-free tissue extracts may 
present features which are similar to those observed in the present study 
with tumor slices, where it has been shown that the glyceraldehyde 
inhibition of the glucolysis of tumor slices can be reversed by small 
amounts of pyruvic acid. 


The Effect of Glyceraldehyde on Muscle Hexokinase Activity 

The effect of various amounts of glyceraldehyde and pyruvic acid on 
the hexokinase activity of rat skeletal muscle extracts was examined 
and the results are shown in Table VI. From these results it can be 
seen the 2 X 10~ 3 M glyceraldehyde inhibits muscle hexokinase activ¬ 
ity 80-100%. Thus, under the conditions adopted in these experiments 

TABLE VI 


The Effect of Glyceraldehyde and Pyruvate on the Activity of the Hexokinase 
from Rat Skeletal Muscle 



Molar cone, of 
glyceraldehyde 

Molar cone, of 
pyruvate 

Normal y 

Per cent inhibition 

Experi¬ 
ment no. 

of glucose 
disappear¬ 
ing 

With glyc¬ 
eraldehyde 

With pyru¬ 
vate -f glyc- 
eraldehyde 

With 

P> ruvate 

1 

5.0 XlO" 3 


300 

76 



2 

4.5X10-’ 

2 .OXIO - 3 

350 

100 

49 

0 


4.5X10-’ 

4.0X10" 3 

350 

100 

50 


3 

2.2X10 ’ 

2.0X10" 3 

550 

100 

88 



1.1X10-’ 

2.0X10-3 

550 

63 

72 


4 

2.2X10-’ 

2.0X10” 3 

455 

87 

76 


5 

4.5X10-’ 

2.0X10-3 

410 

86 

92 


6 

4.5X10-’ 

2.0X10-3 

240 

58 

41 

0 


0.8 ml. of rat skeletal muscle extract was used in these experiments. Other condi¬ 
tions as described in footnote to Table I. 



76 


HARRY RUDNEY 


the hexokinase in muscle seems to be the most sensitive of all the 
enzymes obtained from the various sources mentioned above. 

In general, pyruvate does not appear to have any significant effect, 
although in Expt. 2 (Table VI) it decreased the inhibition from 100% 
to 49%. It is interesting to note that doubling the concentration of 
pyruvate had no further effect. It was also found that pyruvate alone 
had little influence on hexokinase activity, nor could any evidence be 
uncovered to indicate that there was an increased disappearance of 
glyceraldehyde in the presence of pyruvate. Further studies showed 
that the breakdown of hexoscdiphosphate to lactic acid in the same 
muscle extracts used above was unaffected by 3 X 10~ 3 M glyceralde¬ 
hyde. 


TABLE VII 

The Effect of Increasing Concentrations of Adenosine Triphosphate (ATP) 
on the Inhibition of M uscle Hexokinase by Glyceraldehyde 


Experi- 

3.0 micromoles ATP 

G.O micromoles ATP 

10 micromoles ATP 

20 micromoles ATP 

ment no. 

Normal 

Inhibition 

Normal 

Inhibition 

Normal 

Inhibition 

Normal 

Inhibition 

1 


per rent 


per cent 

6.50 

per cent 

83 

595 

per cent 
80 

2 

300 

67 

420 

62 

510 

66 



3 

147 

100 

320 

65 

320 

82 



4 

200 

78 

255 

100 

— 

— 

385 

88 

5 

107 

57 

265 

82 





6 

115 

100 

215 

90 

— 

— 

434 

77 


Conditions similar to those listed under Table VI, Concentration of glyceralde¬ 
hyde = 1.2 X 10~ 3 M . Total volume in vessels = 2.35 ml. 


Other investigators (20,17) have found that increases in the concen¬ 
tration of adenosine triphosphate could decrease the effect of various 
inhibitors on the hexokinase reaction. This point was tested with muscle 
extracts and the results are shown in Table VII. From this table it can 
be seen that an increase in adenosine triphosphate concentration of 
from 3 to 20 micromoles/2.35 ml. does not relieve an existing inhibi¬ 
tion by glyceraldehyde. A decrease of an inhibition of 100% to 77% 
occurred in only one experiment, while in the others the variations can 
be accounted for as being within the limits of the experimental error. 



GLUCOLYSIS INHIBITION 


77 


The Relationship between the Concentration of Enzyme and the 
Inhibitory Action of Glyceraldehyde on Hexokinase Activity 

Stickland (8) found that the inhibition of glucolysis of muscle ex¬ 
tracts fortified with yeast hexokinase could be reversed if the hexo¬ 
kinase concentration was slightly increased. Thus, in one instance an 
increase in enzyme concentration of 14% reduced an inhibition of 90% 
to one of 50%. Stickland stressed these experiments as indicating that 
the concentration of hexokinase was the limiting factor in determining 
the extent of the action of glyceraldehyde on this enzyme. The same 
author was also unable to obtain a direct relationship between the 
enzyme concentration and the measure of enzyme activity as deter¬ 
mined by the rate of glucolysis. In the present investigation experi¬ 
ments were performed on the hexokinase reaction in rat skeletal 
muscle, yeast and tumor extracts, to test these points. 

It was observed that, under the conditions adopted in these experi- 



Fig. 1. Solid lines indicate normal, while dotted lines indicate normal + glycer- 
aldehydc. X — Rat skeletal muscle extract. Glyceraldehyde concentration^.OX 10 -3 
M . 1 enzyme unit == 0.8 ml. of extract. 0 = Aqueous extract of sarcoma 39 acetone 
powder. Glyceraldehyde concentration = 4.5 X 10~ 8 M. 1 enzyme unit = 0.4 ml. of 
extract. A = Yeast hexokinase extract diluted ten times. Glyceraldehyde concentra¬ 
tion = 9.0 X 10~ 3 M. 1 enzyme unit = 0.1 ml. of diluted extract. 



78 


HARRY RUDNEY 


ments, a direct linear relationship was obtained between the amount 
of glucose phosphorylated and the concentration of the hexokinase 
extract (Fig. 1). Furthermore, an increase in the enzyme concentration 
did not reverse the inhibition out of all proportion to the amount added. 
A possible explanation of the differences between the results of the 
present investigation and those obtained by Stickland might lie in the 
different experimental conditions adopted. 

Discussion 

The foregoing experiments demonstrate that /-glyceraldehyde in¬ 
hibits the activity of mammalian tissues. The fact that concentrations 
of glyceraldehyde smaller than those required to inhibit the hexokinase 
in extracts of brain and tumor can completely inhibit the glucolysis of 
these tissues can be explained in several ways. 

1. It is possible that other enzymes in the glycolytic cycle beyond the 
hexokinase stage may be even more sensitive to glyceraldehyde than the 
hexokinase, e.g., fructokinase (22). On the other hand, it has been 
shown by various authors (3,21) and verified in this study with muscle 
extract, that the breakdown of hexose diphosphate to lactic acid is not 
inhibited by glyceraldehyde. However, such experiments do not offer 
conclusive proof that all enzymes beyond the aldolase stage in the 
glycolytic cycle are not affected by glyceraldehyde for the following 
reasons: 

a. It has not been shown in the above-mentioned experiments that 
the breakdown of hexose diphosphate to lactic acid was operating at a 
rate comparable to that of glycolysis. Thus, under conditions where 
the rate of breakdown of hexose diphosphate to lactic acid is less than 
that observed in glucolysis, there might be a partial inhibition by glyc¬ 
eraldehyde of one of the enzymes involved, yet this inhibition will not 

be observable in the process hexose diphosphate -» lactic acid, 

because the reaction catalyzed by the inhibited enzyme is not the 
rate-determining step. 5 

b. It may be that glyceraldehyde acts by uncoupling the oxidative 
phosphorylation step in glucolysis. If such were the case, the break¬ 
down of hexose diphosphate to lactic acid would remain unaffected, 
yet there would be observed an inhibition of glucose breakdown, since 

* The author wishes to thank Dr. M. F. Utter of the Deaprtment of Biochemistry, 
Western Reserve University, for bringing this point to the author’s attention. 



GLUCOLYSIS INHIBITION 


79 


the regeneration of adenosine triphosphate would be decreased by the 
uncoupling, thus leading to a decreased phosphorylation of glucose. 6 

2. The inhibition may be of the type noted by Racker and Krimsky 
(23), wherein a partial inhibition of an energy-yielding reaction in the 
glycolytic cycle led to a greatly increased inhibition of glycolysis as 
measured in terms of lactic acid production. 

3. The proteins of the extract may exert a protective effect on the 
hexokinase. 

The mechanism whereby pyruvate reverses the inhibition of glyco¬ 
lysis by glyceraldehyde still remains obscure. The foregoing experiments 
show that pyruvate does not act directly on the hexokinase in tissue 
extracts. It is possible that the enzymes mediating the pyruvate effect 
are destroyed in tissue extracts, or that pyruvic acid can bring about 
the removal of /-glyceraldehyde in quantities sufficient to lift the inhi¬ 
bition. The method at our disposal could not decide this point. In¬ 
creased amounts of pyruvate appearing under aerobic conditions in 
tissue slices might also explain why concentrations of glyceraldehyde 
which inhibit glycolysis do not inhibit respiration. 

Acknowledgment 

The author wishes to express his thanks to Dr. Bruno Mendel for his advice and 
encouragement during the course of this work. 


Summary 

To determine the mechanism whereby glucolysis is inhibited by dl- 
glyceraldchyde, the effect of this substance was tested on the activity 
of the enzyme hexokinase obtained from various sources, such as rat 
skeletal muscle, rat sarcoma 39, beef brain and yeast, with the follow¬ 
ing results: 

1. The hexokinase from the above-mentioned sources is inhibited by 
(//-glyceraldehyde and the inhibition is caused by the /-isomer. 

2. The sensitivity of the enzyme to the inhibitor varies with the 
source of the enzyme; thus, muscle hexokinase activity is inhibited 
completely by 2 X 10 -3 M (//-glyceraldehyde, while that from yeast is 
only partially inhibited by 10~ 2 M. 

3. Aqueous extracts of acetone powders of rat sarcoma 39 are more 
sensitive to the inhibitor than are homogenates. 



80 


HARRY RUDNEY 


4. Increasing the concentration, either of the hexokinase or of 
adenosine triphosphate, does not affect the extent of the inhibition. 

5. Under the experimental conditions adopted, pyruvate does not 
reverse the inhibitory effect of glyceraldehyde on hexokinase activity 
from various sources. On the other hand, pyruvate does abolish the 
inhibition of anaerobic glucolysis by glyceraldehyde in slices of rat 
sarcoma 39. 

6. The amounts of glyceraldehyde required to inhibit glucolysis in 
slices are smaller than those required to inhibit the hexokinase reaction 
in extracts. 


References 

1. Mendel, B., Klin. Wochschr. 8, 169 (1929). 

2. Mendel, B., Bauch, M., and Strelitz, F., ibid. 10, 118 (1931). 

3. Adler, E., Calvet, F., and Gunther, G., Z. physiol. Chern. 249, 40 (1937). 

4. Sullmann, H., Biochem. Z. 296, 325 (1938). 

5. SCillmann, IT., Enzymylogia 6, 372 (1939). 

6. Holmes, E. G., Ann. Rev. Biochem. 3, 381 (1934). 

7. Lehmann, H., and Needham, J., Enzymologia 6, 95: (1938). 

8. Stickland, L. IL, Biochem. J. 36, 859 (1941). 

9. Dorfman, A., Physiol. Revs. 23, 124 (1943). 

10. Barker, S. B., Shorr, E., and Mallam, M., J. Biol. Chem. 129, 33 (1939). 

11. Colowick, S. I\, Cori, G. T., and Slein, M. W., ibid. 168, 583 (1947). 

12. Nelson, M., ibid. 163, 375 (1944). 

13. Somocyi, M., ibid. 160, 61 (1945). 

14. Robertson, W. B., Science 96, 93 (1942). 

15. Kerr, S. E., J. Biol. Chem. 139, 121 (1941). 

16. Meyerhof, ()., Biochem. Z. 138, 176 (1927). 

17. Meyerhof, O., and Randall, L. O., Arch. Biochem. 17, 171 (1948). 

18. Needham, J., and Lehmann, II., Biochem. J. 31, 1913 (1937). 

19. Mendel, B., Strelitz, F., and Mundell, D., Science 88, 149 (1938). 

20. Greio, M., Arch. Biochem. 17, 129 (1948). 

21. Boyland, E., and Boyland, M. E., Biochem. J. 32, 321 (1938). 

22. Cori, G. T., and Slein, M. W., Federation Proc. 6, 246 (1947). 

23. Racker, E., and Krimsky, I., J. Biol. Chem. 173, 519 (1947). 



Distribution of Thiamine in the Brain 


Gilberto G. Villela, Mario Vianna Dias and Laura T. Queiroga 

From the Divisions of Biochemistry and Physiology, 

Instituto Oswaldo Cruz, Rio de Janeiro, Brazil 
Received February 25, 1949 


Introduction 

Peters and associates have shown that cocarboxylase (thiamine 
pyrophosphate) is an essential factor in pyruvate oxidation in the 
brain calling attention to the importance of vitamin Bx in nerve function 
(1). More recently, the findings of Minz and of von Muralt have re¬ 
vealed a possible new role played by thiamine in the mechanism of 
nerve impulse transmission (2,3). Vianna Dias has shown that thia¬ 
mine, when administered locally to the brain, produces a very marked 
excitatory effect (4). This property is specific for the entire thiamine 
molecule and gives evidence of its possible action on the central ner¬ 
vous system. Moreover, it is known that thiamine inhibits cholines¬ 
terase and reinforces many of the effects of acetylcholine (5,6). 

As acetylcholine and cholinesterase are distributed differently in the various parts 
of the brain (7,8), it would be desirable to know whether thiamine has a similar 
distribution. The published values refer only to the thiamine content of the whole 
brain; analytic data for the thiamine content of the white and gray matter are still 
lacking. Lcong, determining the thiamine of rat brain by the bradycardia method 
found 2.0 7 /g. of fresh tissue (9). Ochoa and Peters give values for the rat and the 
pigeon which varied from 2.7 to 3.6 7 , expressed as eocarboxylase ( 10 ). Williams and 
coworkers reported 4.1-4.8 7 for the rat, and 1.4-1 .8 7 /g. for the human brain (11). 
Muralt, working with the thiocrome and Phijeomyces tests obtained a value of 2.6 7 
for the gray matter ( 12 ). 

The present paper deals with the results obtained for the white and 
gray matter (cortex) and for the nucleus caudatus of normal dogs. 
Some experimental data on animals injected with large doses of 
thiamine are also included. The thiamine of the nervous tissues of dogs 
under varied conditions is being studied and will be reported upon later. 


81 



82 


G. G. VILLELA, M. V. DIAS AND L. T. QUEIROGA 


Material and Methods 

Twenty-three adults dogs weighing 8-16 kg. were used. The animals were killed by 
intravenous injection of chloroform (about 2 . ml) and the brain removed as soon as 
possible and cleaned from contaminating blood. We observed that chloroform causes 
the brain vessels to empty and therefore avoids undesirable accumulation of blood. 
The hemispheres were cut in thin slices and the gray matter was separated from the 
white mass using a sharp knife and a razor blade. Nucleus caudatus was also removed 
and weighed separately. For the analysis, 250 mg. to 1 g. of tissue was weighed on an 
analytical balance and ground in an agate mortar until a fine homogeneous mass was 
obtained. Small volumes (1 ml. each) of 0.1 N H 2 SO 4 were added and the mass well 
mixed. The resulting suspension was carefully transferred to a Pyrex test tube and 
made up to 5 ml. with acid solution. The tube was left in a boiling water bath for 15 
min. After cooling, 2 ml. of a 10 % suspension of takadiastase (Parke, Davis & Co.) 
was added and the solution adjusted to pH 4.5. The solution was then incubated 24 hr. 
at 37°C. After hydrolysis, the solution was filtered and clarified with Celite (Hyfio- 
supercel) and the volume made up to 10 ml. with distilled water. The clear filtrate was 
used for the thiamine determinations. 

Sarett and Cheldelin’s microbiological method using Lactobacillus fermentum 36 was 
employed for all determinations (13). In some cases, the slant cultures were fortified 
with high doses of thiamine as recommended by Cheldelin et al. (14). Kach assay was 
duplicated and accompanied by an assay with a standard thiamine solution. The 
thiochrome method of Hennessy and Cerecedo, slightly modified, was performed on 
only 8 samples and the results were in good agreement with the microbiological 
determinations. The fluorescence of the thiochrome was measured in a Pfalz and Bauer 
fluorometer and the adsorption run in a Decalso column (15). 

Thiamine hydrochloride 1 dissolved in 0.85% saline was injected subcutaneously 
daily into 6 normal dogs and the animals killed with chloroform. The animals were 
sacrificed one day after the last injection. Four dogs were injected daily with 1 mg. of 
thiamine/kg. body weight, while 2 others received, respectively, 8.5 and 10.0 mg./day. 

Results 

In Table I are shown the values obtained in 17 normal dogs for the 
thiamine content of the gray matter (cortex), white matter and nucleus 
caudatus. The average values for the white matter revealed lower 
amounts of this vitamin than for the cortex and nucleus caudatus. 
The same distribution was maintained in the brain of the dogs saturated 
with thiamine (Table II). Probably the variations encountered in the 
normal dogs are probably due in part to the varied origin of these 
animals, since they were of different breeds and environmental condi¬ 
tions which may have affected their nutritional state. 

thiamine hydrochloride was kindly supplied by Hoffman La Roche, Rio de 
Janeiro, to whom we are grateful. 



THIAMINE IN BRAIN 


83 


The storage of thiamine in the brain seems not greatly affected when 
large doses of thiamine are administered. Leong reported that the rat 
fed a diet rich in vitamin Bi attained a maximum storage with about 
30 I. U. of thiamine per day (9). Ochoa and Peters similarly observed 
that the thiamine content of the total brain reaches a level not far 
beyond the normal in animals receiving extra doses of this vitamin 
(10). In our experiments, where large amounts of thiamine were ad¬ 
ministered, the values increased to only slightly over the normal aver¬ 
age, suggesting that the brain is nearly saturated with thiamine. Thus, 
dogs 4 and 5, which received 8.5 and 10.0 mg. of thiamine per day, 
presented a thiamine content comparable with those injected with 
lower doses and only slightly higher than some of the normal dogs. 

TABLE I 


Thiamine Distribution in the Brain of Normal Dogs 
Total thiamine in 7 /g. of tissue 


*No. of dogs 

Gray matter 

White matter 

Nucleus caudatus 

1 

2.8 

1.1 

2.7 

2 

1.7 

0.7 

2.1 

3 

1.9 

1.4 

2.3 

4 

2.4 

1.6 

3.4 

5 

1.3 

0.8 

1.3 

6 

2.7 

1.2 

2.3 

7 

1.7 

1.3 

1.7 

8 

r 1.3 

1.0 

2.2 

9 

1.4 

2.0 

1.5 

10 

1.4 

1.3 

1.6 

11 

2.1 

0.8 

2.4 

12 

2.1 

1.3 

1.5 

13 

1.6 

0.7 

1.4 

14 

2.0 

1.5 

1.7 

15 

1.3 

1.0 

2.1 

16 

1.3 

1.1 

1.9 

17 

1.7 

1.2 

1.8 

Mean and S.D: 

1.8zb0.48 

1.2±0.37 

2.0 ±0.53 

M 

Stand, error: 

0.12 

0.09 

0.13 


Jsfx)* o 

<r = \ ——v Stan, error = 7 =. 
\ n—1 V n 



84 


G. G. VILLELA, M. V. DIAS AND L. T. QUEIROGA 


TABLE II 


Thiamine Content of the Brain of Injected Dogs 



Thiamine inj. 


Thiamine in y/g. of tissue 

Dogs 

mg./kg. body 
wt./day 

Time in days 

Cortex 

White matter 

Nucleus 

caudatus 

1 

1 

21 

2.6 

1.7 

2.9 

2 

1 

10 

3.2 

1.2 

4.0 

3 

1 

13 

3.0 

3.0 

3.1 

4 

1 

13 

2.7 

2.1 

4.0 

5 

8.5 

3 

2.7 

2.5 

3.6 

6 

10.0 

8 

2.3 

2.1 

2.6 

Mean and S. D. (<r) 

Stan, error 

2.7±0.32 

0.14 

2.1 ±0.62 
0.28 

3.4±0.62 

0.28 


Summary 

Thiamine was determined in the brain of 17 normal dogs and showed 
a varied distribution. The values averaged 1.8 ± 0.48 y for the cortex, 
1.2 dz 0.37 7 for the white matter, and 2.0 ± 0.53 7 /g. of tissue for the 
nucleus caudatus. The injection of large doses of thiamine hydrochlo¬ 
ride subcutaneously provoked a limited increase of this vitamin in the 
brain with a distribution similar to that presented by the normal 
uninjected animals. 


References 

1. Peters, R. A., Lancet 230, 1161 (1936). 

2. Minz, B., Compt. rend. soc. biol. 127, 125 (1938). 

3. von Muralt, A., Nature 154, 767 (1944). 

4. Vianna Dias, M., Science 105, 211 (1947); Abstracts, XVIII Intern. Physiol. 

Cong. 1947, 338. 

5. Click, D., and Antopol, W., J. Pharmacol. Exptl. Therap . 66, 389 (1939). 

6. Agid, R., Beauvallet, M., and Minz, B., Compt. rend. soc. biol 126, 982 (1937). 

7. Macintosh, F. C., J. Physiol. 99, 436 (1941). 

8. Nachmansohn, D., Bull. soc. chim. biol. 21, 761 (1939). 

9. Leong, P. C„ Biochem. J . 31, 375 (1937). 

10. Ochoa, S., and Peters, R. A., ibid . 32, 1501 (1938). 

11. Williams, R. J., Univ . Texas Publ. 4237, 43 (1942). 

12. Muralt, G., and von Muralt, A., Vitamins and Hormones 5, 93 (1947). 

13. Sarett, II. P., and Cheldelin, V. H., J. Biol Chem. 155, 153 (1944). 

14. Cheldelin, V. II., Bennett, E. J., and Kornberg, H. A., ibid . 166 779 (1946). 

15. IIennessy, D. J., and Cerecedo, L. R., J. Am. Chem. Soc . 61, 179 (1939). 





The Sulfur Amino Acid Requirement of Tetrahymena geleii 1 

Dorothy S. Genghof 

From the Department of Biochemistry , Cornell University 
Medical CoUege ) New York City 
Received March 3, 1949 

Introduction 

Recently, a paper by Kidder and Dewey (1) showed that, on a 
methionine-low medium, Tetrahymena geleii W, a ciliated protozoon, 
gave a growth response to the addition of homocystine to the medium 
in the presence of liver extract. 2 They explained their results on the 
basis of the methylation of homocysteine to form methionine through 
the mediation of a factor in the liver which they believed could be de¬ 
scribed as a co-transmethylase. They also reported that homocystine 
plus choline did not permit growth of Tetrahymena in the absence of the 
liver factor. Furthermore, they stated that exogenous choline did not 
influence the methylation of homocystine. To account for a source of 
methyl groups, they assumed the synthesis of the methyl group from 
some unknown precursor. The small amount of methionine in their 
liver extract was reported to be insufficient to explain their results. 

In view of the fact that dimethylthetin was a methyl donor for the rat, it occurred 
to us that perhaps dimethylthetin might possibly serve as a methyl donor for Tetrahy¬ 
mena . Experiments were undertaken to test this possibility. Dimethylthetin, however, 
was unable to act as a methyl donor when added to increasing amounts of homo¬ 
cystine in the presence of the liver extract. Furthermore, choline, betaine and di- 
methylpropiothetin were equally ineffective although the quantities added were 
sufficient to combine with all of the homocystine present. 

We then turned to the examination of the liver factor. Kidder and Dewey found 
that, in the presence of the liver extract (L.E.L.) and graded amounts of homocystine, 
the organism showed a growth effect greater than could be accounted for on the basis 
of the methionine content of the medium and the added liver extract. Since they could 
not account for this additional growth on the basis of the methionine present, they 

1 The author wishes to thank the Lederle Laboratories Division, American Cyana- 
mid Company, for a research grant which has aided greatly in this work. 

2 Liver Extract Lilly No. 343, referred to as L.E.L. 


85 



86 


DOROTHY S. GENGHOF 


assumed that additional methionine must have been formed from the homocystine 
present. 

When we attempted to reproduce this growth effect, we obtained only as much 
growth as could be accounted for on the basis of the methionine present in the partic¬ 
ular samples of Liver Extract Lilly No. 343 we had obtained. However, our method of 
assay for methionine was different, as will be explained later. We suspected that one 
might be dealing with a simple question of requirement of sulfur amino acids in the' 
presence of a minimal required amount of methionine. Homocystine might thus be 
supplying the remaining requirement for organic sulfur. If this were true, cystine 
might be capable of replacing homocystine in t he presence of this minimal amount of 
methionine. And indeed, this proved to be the case, for cystine plus liver extract, or an 
amount of methionine equivalent to that present in the liver extract, gave just as 
good growth as was obtained with homocystine plus liver extract. 

Obviously, one was not dealing with a methylation problem involving 
a co-transmethylase, but rather with the significance of total amounts 
of sulfur amino acids to the growth of Tetrahymena . The situation would 
seem to be quite comparable to that in the rat (2) and in the chick (3), 
where cystine gives an increased growth effect when added to a limited 
amount of methionine in the diet. 

The most probable explanation for the differences between our 
results and theirs lies in the methionine assay data. According to Kidder 
and Dewey, the amount of L.E.L. they added contained 0.4 y of 
L-methionine as assayed by a method using Leuconostoc mesenteroides , 
whereas our value for the same weight of L.E.L. was 1.4 y of l- 
methionine. In our assay method Streptococcus faecalis was used as the 
test organism but similar values were also obtained when Tetrahymena 
itself was used as the assay organism. Furthermore, when a much 
earlier preparation of L.E.L. was tested, methionine values of the 
same order of magnitude were obtained. If one assumes that the methi¬ 
onine value for the liver extract used in the experiment by Kidder and 
Dewey was the same as we found, the excess growth they obtained 
could be explained almost completely on the basis of the methionine 
content of the liver extract and medium. 

If our explanation of the phenomenon be correct, the elaborate 
theories of transmethylation proposed by Kidder and Dewey for 
Tetrahymena and for higher forms of animal life derive no support from 
their experimental results on Tetrahymena} 

3 Prior to submission to the journal, this paper was sent to Dr. Kidder for his criti¬ 
cism. He has informed us that they have re-evaluated the methionine content of thbir 
liver fraction using Tetrahymena as the assay organism. He stated that, using th r 
purified medium and supplying the Factor II by the addition of protogen [Stokstad 



SULFUR AMWO ACID REQUIREMENT OF T. OELEII 


87 


Materials and Methods 

Organism and Inoculum 

Tetrahymena geleii W , 4 the ciliated protozoon used for these experiments, was 
maintained on a liquid medium composed of 1% dextrose, 1% peptone (Bacto), and 
1% yeast extract. The stock culture was transferred at monthly intervals and was 
incubated and stored at 25°C. 

The cultures for inoculum were grown for 2 or 3 days in an unslanted position. For 
inoculation of the experimental tubes 1 drop of a 1-10 dilution in saline was added to 
each tube and the tubes were incubated at 25°C. for 6 days in a slanted position 
according to the method of Kidder and Dewey. Readings were made with a Klett- 
Summerson colorimeter using a No. 66 filter (640«-700 m/u). 

Medium 

The medium was prepared according to the formula of Kidder and Dewey (1) and 
the pH was adjusted to 7.0 before the medium was autoclaved. It consisted of amino 
acids, with the exception of the sulfur-containing amino acids, 1 % glucose, vitamins 
including 1 7 /ml. of choline, hydrolyzed yeast nucleic acid and Cerophyl 6 at a 1-5 
dilution. In some experiments, the choline was omitted and the Cerophyl was used at 
a 1-10 dilution. The hydrolyzed yeast nucleic acid was prepared as follows: 10 g. of 
yeast nucleic acid were dissolved in 50 ml. of water and 5 ml. of 25% NH 4 OH and 
autoclaved in a sealed tube for 1 hr. at 118°C. according to the method of Levene (5); 
after removal of the ammonia under reduced pressure, the pil of the solution was 
adjusted to 7.0 and the solution was filtered and diluted to 500 ml. Whenever neces¬ 
sary, the medium or its ingredients were stored with chloroform and toluene. 

The medium, prepared in double strength lots, and addenda were placed in a tube 
20 X 150 mm. to make a total volume of 6 ml. The tubes were plugged with cotton 
and autoclaved for 6 min. at 15 lbs. pressure. 

Three samples of liver extract were used; two were from Lilly and Company 6 and 
one was from Lederle Laboratories 7 . The Lilly samples, Liver Extract No. 343, were 

et al.y Arch. Biochem., 20 , 75 (1949)], their assay results for methionine in the liver 
fraction which they used turned out to be 15 7 /mg. or 1.8 7 /ml. on the 120 7 of L.E.L. 
used. Recovery experiments averaged 94%. When cysteine was added to the assay 
medium, they obtained 10.8 r/mg., or 1.3 7 /ml. on the 120 7 of L.E.L. used. A 
recovery value of 106% was obtained in the latter case. Dr. Kidder has authorized us 
to state that they are in full accord with the results expressed in this paper and to 
include the above data in a footnote to this paper. 

4 This organism was obtained through the kindness of Dr. E. L. R. Stokstad of 
Lederle laboratories, Pearl River, N. Y. 

6 The Cerophyl was obtained through the courtesy of Cerophyl Laboratories, Inc., 
Kansas City, Mo. 

6 The recent Lilly sample was obtained through the courtesy of Dr. E. D. Campbell, 
Eli Lilly and Co., Indianapolis, Ind. 

1 ./The Lederle Liver Extract was furnished by Lederle Laboratories, Pearl River, 

W V 

u. Y. 



88 


DOROTHY S. GENGHOF 


of two Lot Numbers, 209S—861400 and 436627. The Lcderle preparation was Solu¬ 
tion Liver Extract—3.3 U. S. P. injectable units per ml. prepared for intramuscular 
injection. The pH of the Lilly samples was adjusted to 7.0 before they were added to 
the medium. All liver extracts were added in a volume of 0.1 ml. to 6 ml. of medium. 

L-Methionine Assay 

A microbiological method was used for the determination of the L-methionine con¬ 
tent of the liver samples and Cerophyl preparations. The earlier assays were performed 
by the method of Stokes et at. (5) and the later ones by the method of Henderson and 
Snell ( 6 ). The same organism was employed for both assays, namely, Streptococcus 
faecalis, ATCC No. 8043. The Lilly extracts were assayed before and after an 18 hr. 
hydrolysis in 6 N HC1, but the Lederle Liver Extract and the Cerophyl were assayed 
as used. 


Experimental Results 

The first evidence to be presented is an experiment designed to 
duplicate the findings obtained by Kidder and Dewey. The results are 
summarized by the curves in Fig. 1 . The constant amount of homo¬ 
cystine used in the preparation of Control Curve 3 was the amount 
chosen by Kidder and Dewey as giving maximal stimulation under the 
conditions of their experiment. 

All values plotted in Fig. 1 were obtained after 3 serial transfers through the same 
medium. This technique involved an original inoculum as described above plus two 
more transfers through the same concentration of test substances, e.g. } control tube 
to control tube, 1 7 of methionine to 1 7 of methionine, etc: Readings were made after 
each 6 day growth period. The value of the serial transfer technique is questionable 
for this experiment. The only difference obtained after serial transfer was a drop 
of 10-15 colorimeter units in the readings of tubes containing small amounts of methio¬ 
nine, as was the case with the values of Control Curve 1. However, the technique 
was applied in order to parallel the work of Kidder and Dewey as closely as possible. 

The growth response to increasing amounts of homocystine in the 
presence of constant amounts of the liver extract, Curves A and B, 
is paralleled by that of Control Curve 2 where the constant amount of 
liver extract is replaced by an amount of methionine equal to the 
amount present in 120 7 of hydrolyzed L.E.L. Lot No. 436627. 
On the basis of methionine assay values of the hydrolyzed material, 
120 7 of L.E.L. Lot No. 436627 added 1.4 7 of methionine to the 
medium and 120 7 of Lot No. 209S added 1.2 7 of methionine. Control 
Curve 2 shows that the L.E.L. is causing no greater amount of growth 
than that which would be expected on the basis of its methionine 
content. 



SULFUR AMENO ACID REQUIREMENT OF T. GELEII 


89 


There is one point on Control Curve 3 that is of particular interest 
to the experiment. The growth represented by this point (marked by a 
dotted line) is to be compared with that represented by the 40 y points 


rt-METHIONINE PER ML. 



* DL- HOMOCYSTINE PER ML. 

Fig. 1. Utilization of DLrhomocystine by Tetrahymena in the presence of Liver 
Extract Lilly (L.E.L.) or L-methionine. 

These values were obtained after 3 serial transfers. Control Curve 1 : graded amounts 
of homocystine; Control Curve 2: graded amounts of homocystine + 2.2 y of methio¬ 
nine (0.8 7 from the Cerophyl of the medium and 1.4 7 of added methionine); Curve 
A: graded amounts of homocystine + 120 7 /ml. of L.E.L. No. 343, Lot 436627 
(methionine content = 2.2 7 : 0.8 7 from the Cerophyl + 1.4 7 from L.E L.); Curve 
B: graded amounts of homocystine + 120 7 /ml. of L.E.L. No. 343, Lot 209S- 
861400 (methionine content - 2 .O 7 :0.8 7 from the Cerophyl -f 1.2 7 from L.E.L.); 
Control Curve 3: graded amounts of methionine + 40 7 /ml. of homocystine. The 
growth values are plotted at the level of methionine represented by the amount of 
methionine added per ml. + 0.8 7 of methionine from the Cerophyl. The point on Con¬ 
trol Curve 3 shown by the dotted line indicates growth at a level of 2.2 7 of methio¬ 
nine and 40 7 of homocystine. This amount of growth corresponds to the amount of 
growth on Control Curve 2 and Curve A at the point of 40 7 of homocystine. 



90 


DOROTHY S. GENGHOF 


on the homocystine curves—Curve A and Control Curve 2—because 
all 3 of these points represent the growth resulting from the addition 
of 2.2 7 of methionine and 40 y of homocystine to the medium. Since 

ft L- METHIONINE PER ML. 



ft Di - HOMOCYSTINE PER ML. 

Fig. 2. Utilization of DL-homocystine by Tetrahymena in the presence of Lederle 
Liver Extract. 

The values were obtained without serial transfer. Curve A: graded amounts of 
homocystine + 0.0007 ml. of Lederle Liver Extract (methionine content = 1.8 y: 
0.4 7 from the Cerophyl + 1.4 y from the Lederle Liver Extract); Control Curve 1: 
graded amounts of homocystine; Control Curve 3: graded amounts of methionine + 
40 7 /ml. of homocystine. The point on Control Curve 3 shown by the dotted line 
indicates growth at a level of 1.8 7 of methionine and 40 7 of homocystine. This 
amount of growth corresponds to the amount of growth on Curve A at the point of 
40 7 of homocystine. Control Curve 4: graded amounts of methionine; Control Curve 
5: graded amounts of methionine -f 0.0007 ml. of lederle Liver Extract. In Control 
Curves 3,4, and 5, the growth values are plotted at the level of methionine represented 
by the amount of methionine added per ml. + 0.4 7 of methionine from the Cerophyl. 
In Control Curve 5, an additional 1.4 7 of methionine has been added, due to the 
amount present in the liver extract. 




SULFUR AlftNO ACID REQUIREMENT OF T. GELEII 


91 


these points on Curves A and B are about opposite the corresponding 
points on Control Curve 3, there is obviously no extra growth over and 
above that which can be attributed to the methionine and homocystine 
present. This is in sharp contrast to the results reported by Kidder and 
Dewey. They obtained a great deal more growth in the presence of 40 y 
of homocystine plus 120 y of L.B.L. than that observed in the presence 
of the same amount of homocystine plus an amount of methionine 
corresponding to the amount in 120 y of L.E.L. 

The previous experiment has shown that two different samples of L.E.L. gave a 
utilization of homocystine that could be accounted for by the amount of methionine 
present in the liver. At an early stage in this work, before L.E.L. was available, a 
sample of Lederle Liver Extract solution was tried. An amount of Iiederle Liver Ex¬ 
tract equivalent to 120 y of L.E.L. was calculated to be 0.0007 ml., on the basis of 
data given in the paper of Bennett and Toennies (7). In this case the medium was 
different in two respects from the previous experiment. The Cerophyl was used at a 
dilution of 1-10 and the choline was omitted. Still another difference from the condi¬ 
tions of the first experiment was that the readings for these tubes were made after 6 
days of growth without carrying the experiment through serial transfers. 

The results plotted in Fig. 2 with a different liver product, that of 
Lederle Laboratories, are essentially the same as those found in the 
serial transfer experiment with L.E.L. When the height of Curve A 
at 40 y of homocystine is compared to the point on Control Curve 3 at 
1.8 7 of L-methionine, the levels are found to be about the same. There 
is no extra growth over and above that which can be accounted for by 
the amount of L-methionine in the Lederle Liver Extract. The fact that 
Control Curves 4 and 5 are practically coincidental when plotted shows 
that the liver is adding no unknown growth factor. 

Although these results were obtained with a different liver extract, 
after a single inoculation, without serial transfer, and in the absence of 
added choline, they are similar to those obtained with L.E.L. 

Since we had found that methionine could substitute for the liver 
extract in the presence of increasing amounts of homocystine, it seemed 
of interest to attempt to substitute cystine for the homocystine. The 
data for this experiment have been summarized in the 4 curves of Fig. 3 
which are similar to those for the homocystine experiment (Fig. 1 ). 
Only one liver extract, L.E.L. Lot No. 436627, was used. A slightly 
greater amount of L-methionine, 1.5 7 /ml., replaced the L.E.L. in 
obtaining the values plotted in Control Curve 2 , although 1.4 7 /ml. of 
L-methionine is probably the value closer to the equivalent amount of 



92 


DOROTHY S. GENQHOF 


methionine. The medium was the same as that for the homocystine 
experiment and this experiment was likewise carried through 3 serial 
transfers. 


i L- METHIONINE PER ML. 



Fig. 3. Utilization of L-cystine by Tetrahymena in the presence of Liver Extract 
Lilly (L.E.L.) or Lrinethionine. 

The values were obtained after 3 serial transfers. Control Curve 1 : graded amounts 
of homocystine; Control Curve 2: graded amounts of homocystine + 2.3 y of methio¬ 
nine (0.8 y from the Cerophyl of the medium and 1.5 7 of added methionine); Curve 
A: graded amounts of homocystine + 120 7 /ml. of L.E.L. No. 343, Lot 436627 
(methionine content = 2.2 7 *. 0.8 7 from the Cerophyl 4* 1.4 7 from L.E.L.); 
Control Curve 3: graded amounts of homocystine + 20 7 /ml. of L-cystine. The growth 
values are plotted at the level of methionine represented by the amount of methionine 
added per ml. 4 0.8 7 of methionine from the CerophyL The point on Control 
Curve 3 shown by the dotted line indicates growth at a level of 2.3 7 of methionine and 
40 7 of homocystine. This amount of growth may be compared to the amount of 
growth on Control Curve 2 and Curve A at the point of 40 7 of homocystine. 





SULFUR AMINO ACID REQUIREMENT OF T. GELEII 


93 



Fig. 4. Effect of increasing levels of Liver Extract Lilly (L.E.L.) or Lr-methionine 
upon the utilization of DL-homocystine by Tetrahymena. 

All values have been plotted after subtraction of the control tube reading. ExpL I: 
Control Curve: graded amounts of homocystinc; Curve A: graded amounts of homo¬ 
cystine + 100 y/ml. of L.E.L. No. 343, Lot 436627 (methionine content = 1.2 y ); 
Curve B: graded amounts of homocystine + 400 y/ml. of L.E.L. (methionine con¬ 
tent = 4.8 y); Curve C: graded amounts of homocystine + 1000 y/ml. of L.E.L. 
(methionine content = 12 y). Expt. II: Control Curve: graded amounts of homo¬ 
cystine; Curve A': graded amounts of homocystine + 1 y/ml. of methionine; Curve 
B': graded amounts of homocystine + 4 y/ml. of methionine; Curve C': graded 
amounts of homocystine + 10 y/ml. of methionine; Curve D': graded amounts of 
homocystine + 20 y/ml. of methionine. 

The results show that L-cystine can replace DL-homocystine in the 
presence of either L.E.L. or an equivalent amount of methionine. The 
Control Curves 2 and 3 of Fig. 3 bear the same relationship to Curve A 
as did the Control Curves 2 and 3 of Fig. 1 to Curves A and B. 

Further evidence that the liver extract or an equivalent amount^of 




94 


DOROTHY S. GENQHOF 


methionine stimulate the utilization of graded amounts of homocystine 
in a similar manner is presented in Fig. 4. The amounts of methionine 
or liver extract in the medium were maintained at given levels and the 
homocystine was added in graded amounts. The final medium con¬ 
tained Cerophyl at a dilution of 1-10 but no added choline. In plotting 
the curves of Fig. 4, the values from the control tubes were used as the 
zero point in order to compare more readily the results at the various 
levels of liver extract (Fig. 4, Expt. I) or methionine (Fig. 4, Expt. II). 

There is a relative increase in the utilization of the homocystine with 
amounts of methionine up to 4 7 /ml. and with amounts of liver up to 
400 7 /ml. equivalent to 4.8 7 /ml. of methionine. At 10 7 /ml. of 
methionine the organism can no longer use the homocystine as effi¬ 
ciently as it uses it at 4 7 /ml., and at the 20 7 /ml. level of methionine 
there is no increase in growth, with amounts of homocystine up to 50 
7 /ml., over that in the control tube containing methionine alone. How¬ 
ever, the level of 1000 7 /ml. of liver, which contains only 12 7 /ml. of 
methionine, gives a curve very similar to that given by 20 7 /ml. of 
methionine. This is due in part, we believe, to the cystine content of the 
L.E.L. One thousand 7 of L.E.L. would contain 5 7 of cystine, on the 
basis of an L.E.L. analysis (7). This cystine would add to the stimu¬ 
lating effect of the methionine present and so cause Curve C (Fig. 4, 
Expt. I) to approach the curve of the 20 7 /ml. level of methionine 
(Curve D', Fig. 4, Expt. II). 

The results of this experiment show that with 1 7 /ml. of methionine in the medium 
there is a limited use of increasing amounts of homocystine. With about 4 7 /ml. of 
methionine, there is a maximal use of homocystine, and with increasing amounts of 
methionine, up to 20 7 /ml., there is a diminishing use of homocystine. 

Similar results were obtained when the growth on increasing amounts of methionine, 
up to 30 7 /ml., was compared with the growth on increasing amounts of methionine 
plus 40 7 /ml. of homocystine. At a level of 15-20 7 /ml. of methionine, the amount of 
growth was the same, with or without added homocystine. 

When a limited amount of methionine is present, increased growth 
of Tetrahymena is observed in the presence of added homocystine. 
However, with a maximal amount of methionine present, addition of 
homocystine does not result in any additional growth of Tetrahymena . 
Comparable results are obtained when cystine replaces homocystine. 

Acknowledgment 

The author wishes to express appreciation for the invaluable advice and counse 
given by Dr. Vincent du Vigneaud throughout the course of this work. The capable 



SULFUR AMINO ACID REQUIREMENT OF T. GELEII 


95 


assistance of Miss Mary R. Lloyd and Mrs. Susan M. Wing in the microbiological 
work, and of Mrs. Josephine T. Marshall in the preparation of the figures is gratefully 
acknowledged. 

Summary 

We have been unable to confirm the findings of Kidder and Dewey 
upon which they base their contention that a co-transmethylase is 
partly responsible for the utilization of homocystine by Tetrahymena 
geleii W in the presence of Liver Extract Lilly No. 343. In our experi¬ 
ments, the increased growth of Tetrahymena in the presence of graded 
amounts of homocystine and a constant amount of liver extract seemed 
to be due entirely to the methionine content of the liver extract and 
medium. Furthermore, it was found that cystine could replace homo¬ 
cystine in the nutrition of Tetrahymena under similar conditions. 

The response of Tetrahymena to increasing amounts of homocystine 
or cystine in the presence of limited amounts of methionine appears to 
depend upon an overall deficiency of sulfur amino acids. In the pres¬ 
ence of an amount of methionine up to 15 7 /ml., addition of cystine or 
homocystine resulted in an increased amount of growth. In the 
presence of maximal amounts of methionine, addition of cystine or 
homocystine did not increase the growth of Tetrahymena. 

References 

1. Kidder, G. W., and Dewey, V. C., Proc. Natl Acad. Sci. U. S. 34, 81 (1948). 

2. du Vigneaud, V., Brown, G. B., and Chandler, J. P., J. Biol. Chem. 143, 59 

(1942). 

3. Grau, C. R., and Almquist, II. J., J. Nutrition 26, 631 (1943). 

4. Levene, P. A., J. Biol. Chem. 33, 425 (1918). 

5. Stokes, J. L., Gunness, M., Dwyer, I. M., and Caswell, M. C., ibid. 160, 35 

(1945). 

6. Henderson, L. M., and Snell, E. E., ibid. 172, 15 (1948). 

7. Bennett, M. A., and Toennies, C., ibid. 163, 235 (1946). 



Concentration of Amino Acids by the Excised Diaphragm 
Suspended in Artificial Media. I. Maintenance 
and Inhibition of the Concentrating Activity 1 

Halvor N. Christensen and Jean A. Streicher 

From the Division of Laboratories , Children’s Medical Center , Department of 
Biological Chemistry , Harvard Medical School , and the Department of 
Biochemistry and Nutrition, Tufts College Medical School 
Boston , Massachusetts 
Received March 10, 1949 

Introduction 

This communication reports a study in vitro of the activity of cells in 
maintaining internal concentrations of amino acids much higher than 
those found in the fluids which bathe them (1,2). The ability of the 
excised diaphragm of the rat and of the guinea pig to concentrate gly¬ 
cine has been maintained during 1-3 hours of immersion in saline- 
bicarbonate solutions containing sodium pyruvate. The influences of 
various factors, and of various substances added to these media, upon 
the assimilative activity have been studied. Some of the conclusions 
reached were confirmed when the distribution of total a-amino acid 
nitrogen instead of glycine was studied. 

The concentrative assimilation of glutamic acid by brain cortex 
slices has been described by Stern (3) in a preliminary communication. 
The disappearance of glutamic acid from the suspending fluid was 
accounted for partially by the accumulation of glutamic acid within the 
slices and partially by chemical change of the glutamic acid. The 
assimilation occurred in the presence of glucose or certain other sub¬ 
strates, and was inhibited by malonate, azide, luminal, and iodoacetate, 
and by the absence of oxygen. 

Changes in the amount of free amino acids in the system were small 
when a hemidiaphragm carefully excised from a rat was shaken in saline 
solutions under the conditions described here. The diaphragm of the rat 

1 Assisted by a grant-in-aid from the Abbott Laboratories, Inc., North Chicago, Ill. 

96 



AMINO ACID CONCENTRATION. I 


97 


contained about 5 times as high a glycine concentration as the plasma, 
on a fresh tissue basis, corresponding to a ratio of about 7.5 for the con¬ 
centration in the cellular water to that in the extra-cellular water. No 
assumption is made that the apparently free amino acids of the tissues 
are dispersed in the cell water. 

Experimental 

Sprague-Dawley rats weighing 180-210 g. were fasted 20 hr. then anesthetized 
lightly with ether, and blood was drawn from the heart into tubes containing heparin. 
The hemidiaphragms were removed quickly with minimal trauma, weighed on a 
precision torsion balance, and placed at once in 2-2.2 ml. of the freshly prepared 
medium in a 25 ml. Erlenmeyer flask at room temperature. This was set to shaking in 
a water bath at 37°C. In some cases, the second half-diaphragm was extracted at once 
for analysis after weighing (no equilibration in artificial media). The agitation was by 
horizontal to and fro motion 6.8 cm. long, at 60-78 cycles/min., with a gas mixture 
containing 95% 02-5% C0 2 passing through the flask. At the end of the experimental 
period the hemidiaphragm was blotted on filter paper and extracted carefully with 
saturated aqueous picric acid (2.5 ml. for every 100 mg. of tissue) by grinding in a mor¬ 
tar. Three aliquots of 1 ml. of filtrate could be obtained, permitting duplicate glycine 
determinations as well as a blank analysis. When a-amino nitrogen was to be deter¬ 
mined, the tissue was extracted repeatedly by grinding with aqueous picric acid, 
filtering into the special tube for this determination (4), to a total volume of about 
5 ml. This filtrate was buffered at pH 6.5 with phosphate and heated 90 min. to 
eliminate the a-amino group of glutamine and half the a-amino nitrogen of glutathione 
(5). Plasma and suspending media were deproteinized by adding 5 volumes of picric 
acid. Whenever a-amino nitrogen was to be determined, these filtrates also were 
heated 90 min. at 100°C. at pH 6.5. 

Glycine was determined by the method of Alexander, Landwehr and Seligman (6), 
using, however, twice as much ninhydrin as originally recommended. This analytical 
procedure is discussed in another report (7). The small specimens available necessi¬ 
tated a halving of the volumes used in the glycine analyses of diaphragms. The total 
volume of distillate collected in the procedure (6) was adjusted to 5 ml. rather than 
10 ml. The volume of water added midway in the distillation was usually 1.9 ml. or 
more. These distillations were completed in about 9 min. Entirely satisfactory results 
were obtained with this modification. A rough neutralization of the residual picric 
acid in the filtrates was necessary (about 1 drop 0.75 N NaOH/ml. of filtrates of 
diaphragm and saline media; 1 drop/2 ml. of plasma filtrates and 1:11 tissue filtrates). 

Results 

When the hemidiaphragm was shaken in a glycine-free salt solution, 
from half to two-thirds of the diaphragm glycine diffused into the 
medium in 1 hr. Only when enough glycine was placed in the medium 
to prevent or diminish this glycine transfer were we able to compare 



98 


HALVOR N. CHRISTENSEN AND JEAN A. STREICHER 


the activity of various substances in promoting the retention or loss 
of glycine by the tissue. A fluid glycine concentration approximately 
equal to that of the plasma prevented the fall of the glycine level of the 
diaphragm during 1-3 hr. of equilibration in an appropriate medium 
(see below). If the fluid glycine level was higher than the plasma con¬ 
centration, glycine was transferred from the fluid into the diaphragm; 
if lower, glycine was transferred from the diaphragm to the fluid 
(Table I). 

TABLE I 

An Experiment Illustrating the Responsiveness of the Diaphragm Concentration 
to the Level in the Suspending Fluid 

Two paired portions of diaphragm were shaken for 2 hr. in media alike except in 
glycine concentration. A third control portion was analyzed at once upon excision. 



Fluid glycine N 

Diaphragm 
glycine N 

Distribution ratio 


mg.-% 

mg.-% 


In the intact rat 

0.48“ 

2.40 

5.0 

Equilibrated portion 1, after 2 hr. 

0.60 

2.91 

4.9 

Equilibrated portion 2, after 2 hr. 

0.38 

1.95 

5.1 


“Plasma concentration X 1.05. 


When the glycine of the medium was made twice as high as that of 
the plasma, the glycine concentration of the tissue was nearly doubled 
in 2 hr. of equilibration (Table IV). These results gave some justifica¬ 
tion for comparing the distribution of glycine between the diaphragm 
and the plasma on one hand, and between the diaphragm and an 
artificial suspending fluid on the other. These comparisons have been 
made in terms of the distribution ratio , that is, the ratio of the dia¬ 
phragm concentration (fresh weight basis except as indicated) to 
extracellular concentration. 

The total amount of “free” glycine nitrogen in the system, hemi- 
diaphragm-suspending fluid, was increased on an average by only 0.8 y 
or 6% during 1 or 2 hr. of incubation in glycine-containing media 
(27 Expts.). The small amount of extra glycine may have been formed 
by either cellular or extracellular proteolysis; experiments with tissue 
slices have suggested that the cut boundaries may have been a factor. 

One could be surer of how strongly a substance affected glycine assimilation using 
a second technique. Two paired diaphragms tvere shaken simultaneously in two sets 
of media, alike except in one ingredient. The final glycine concentrations of the two 



AMINO ACID CONCENTRATION. I 


99 


suspending fluids rarely differed by as much as 10%; therefore, similar conclusions 
were reached whether we compared the final concentrations of the 2 hemidiaphragms 
or compared the 2 distribution ratios (as recorded here). The standard deviation for 
pairs of hemidiaphragms shaken in two samples of the same medium was about 5% 
(6 trials). 

Illustrative experiments leading to the selection of a medium (medium 
No. 1) containing 103 mil/ NaCl, 25 mM NaIIC0 3 , 2.5 ml CaCU, and 
20 mil/ sodium pyruvate (and glycine) are recorded in Tables II and 
III. During 60 min. of equilibration in this medium, the distribution 
ratio for glycine increased up to 30% over the original value in the rat. 
If the medium was replaced hourly, the distribution ratio did not fall 

TABLE II 

Changes of the Distribution of Glycine when the Diaphragm 
was Suspended in Various Solutions 

One hemidiaphragm was analyzed immediately and the ratio of its glycine concen¬ 
tration to that of the plasma (X 1.05) calculated. The other hemidiaphragm was 
analyzed after 60 min. in the indicated medium, and its glycine referred to that of the 
suspending medium. 


Animal 

no. 

Medium 

Plasma 

glycine 

N 

Fluid 

glycine 

Dia¬ 
phragm 
in rtra 

Glycine 
m vitro 

Per cent 
change of 
distribu¬ 
tion ratio 



(XI.05) 

mg.-7c 

mg.- ( i 

mg. -Vo 


34 

Rat plasma 

0.42 

0.52 

1.90 

1.78 

-29 

35 

Krebs-Ringer-bicarbonate solution 

0.55 

0.53 

2.4 

1.8 

-22 


(KRB) 






33 

Same, but with 80 meq. K/l. 

0.38 

0.71 

2.1 

1.50 

-62 

39 

NaCl 129 meq/1. NaHC0 3 , 25 meq/1. 

0.59 

0.62 

3.0 

1.7 

-45 

45 

KRB+KON, 3 m M. 

0.17 

0.66 

2.03 

1.4 

-51 

44 

KRB+sodium pyruvate, 20 mM. 

0.34 

0.37 

1.50 

1.80 

+ 4 

51 

KRB+sodium citrate, 15 mM. 

0.46 

0.46 

2.13 

1.56 

-26 

54 

KRB-f pyruvate, less KC1 and 

0.65 

0.42 

2.30 

1.91 

+31 


KH 2 PO 4 






62 

Medium 1°, 20 meq. Ca ++ /1« 

0.36 

0.41 

1.70 

: 1 . 6 O 

-20 

90 

Medium 1, 180 min. equilibration, 

0.63 

0.43 

2.04 

1.59 

+ 14 


renewed hourly 






63 

Medium 1. 0.75% glucose replacing 

0.41 

0.44 

1.83 

1.40 

-29 


pyruvate 






58 

Medium 1+DNP, 10' 4 M 

0.24 

0.52 

1.28 

1.11 

-60 

64 

Medium 1+DNP, 10'* M. 

0.40 

0.46 

2.13 

1.67 

-32 

56 

Rat plasma+pyruvate, 20 mM. 

0.36 

0.32 

! 1.33 

1.39 

+ 16 


° NaCl 103, NaHCOi 25, CaCl 2 5, and sodium pyruvate 20 meq./l. 



100 


HALVOR N. CHRISTENSEN AND JEAN A. STREICHER 


TABLE III 

Changes in the Distribution Ratios of Glycine ([Diaphragm-]/[Suspending Fluid]) 
Resulting from Modification of the Suspending Fluid 
One hemidiaphragm was equilibrated 60 minutes in the usual control medium, the 
other simultaneously in this medium modified as indicated. Where electrolytes were 
added the sodium chloride content of the medium was correspondingly decreased. 


Animal no. 

Modification 

Por cent change 
in distribution 
ratio 

80 

Succinate replacing pyruvate 

-13 

89 

Succinate replacing pyruvate 

-14 

85 

a-Ketoglutarate replacing pyruvate 

+ 

87 

rt-Ketoglutarate replacing pyruvate 

- 3 

145 

Fumarate replacing pyruvate 

-17 

146 

+ 20 m¥ fumarate 

-14 

127 

Additional pyruvate, total 40 mA//I. 

-15 

148 

+ ATP, 10 m M a 

-23 

93 

Higher pH, (HCOr) = 35 ml 

+ 11 

96 

Higher pH, (HCOa") = 50 m M 

- 8 

159 

+ 10 m M KC1 

0 

128 

+ 1.2 mAf NaHsPO« 

+ 12 

133 

+ 1.2 ml MgS0 4 

+ 6 

135 

+ both NaH 2 P0 4 and MgS0 4 

+ 8 

78 

+ 3 m M malonate 

+ 10 

79 

+ 20 mAf malonate 

-13 

92 

+ 33 m M malonate 

+ 6 

119 

+ 55 m M malonate 

-11 

88 

4- sodium arsenite 10 m M 

-10 

91 

4- sodium arsenite 25 mM 

-56 

102 

4- dinitrophenol 2 X 10 -8 M 

-22 

97 

+ dinitrophenol 1 X 10“ 5 Af 

-16 

94 

4- dinitrophenol 5 X 10 -8 M 

-10 

137 

4- iodoacetate, 1 mA/ 

-10 

139 

4- aminopterin, 0.8 mg./ml. 

+ 4 


° Neither experimental nor control medium contained pyruvate. 

appreciably below the original value during 3 hr. This medium was 
equally effective for the diaphragm of the weanling guinea pig. Re¬ 
placement of pyruvate by citrate (15 mM/1.) diminished glycine re¬ 
tention, probably because of the binding of calcium ions. ATP (10 mil/) 
was inhibitory (Table III), perhaps for the same reason. a-Ketoglu- 
tarate replaced pyruvate; succinate replaced pyruvate but less effec¬ 
tively. Replacement of half the sodium of the medium by potassium 
led to about as rapid loss of glycine from the diaphragm as under any 
circumstances observed. 



AMINO ACID CONCENTRATION. I 


101 


When rat plasma was used as the suspending medium, the concen¬ 
tration of a-amino acids (measured manometrically by ninhydrin) in 
the diaphragm was maintained less successfully than for glycine, the 
distribution ratio falling, as an average, 52%. The concentration of 
glycine (Table II), and of the amino acids collectively, was improved 
by the addition of 20 mM/1. of sodium pyruvate to the plasma. The 
average fall in distribution ration of a-amino nitrogen in 1 hr. from the 
value in the intact animal was 35% in the presence of pyruvate. 

Effects of Certain Poisons 

In a N 2 -CO 2 atmosphere, or in a medium containing 3 m M cyanide 
(Table II), glycine loss from the diaphragm was rapid, being faster in 
the absence of oxygen. Arsenite at 25 m M concentration had a similar 
effect (Table III). A surprising resistance to malonate was encountered. 
After tests at lower levels, all of the NaCl of the medium was replaced 
by sodium malonate (55 m M). The distribution ratio after 1 hr. was 
still only 11 % lower than with the control medium. 

2,4-Dinitrophenol at 10 -4 M led to a very rapid loss of glycine (Tables 
II and III). A strong inhibition was still evident at 10 ~ 5 M, with no 
effect apparent at 10 -6 M. Adding dinitrophcnol to plasma containing 
20 mM/ 1 . of pyruvate inhibited the retention of a-amino nitrogen by 
the suspended diaphragm, the distribution ratios falling by 52% as an 
average. 

Male rats showed higher plasma glycine concentrations than females, 
44 males averaging 0.48 mg.-% of glycine nitrogen, standard deviation 
0.08, 46 females 0.35 mg.-%, standard deviation 0.06. 

Diaphragm of the Scorbutic Guinea Pig 

We had hoped to be able to study in vitro the depression of the level 
of free glycine in the muscle in the scorbutic guinea pig (8). The dia¬ 
phragm, however, did not share this decrease with skeletal muscle. 
Diaphragms of either normal or scorbutic guinea pigs (Table IV) were 
highly responsive to elevations of the extracellular glycine, whether in 
the intact animal or in the artificial medium. When the glycine concen¬ 
tration of the suspending fluid was double that of the plasma, the 
glycine content of the diaphragm was almost doubled in 2 hours. 



102 


HALVOR N. CHRISTENSEN AND JEAN A. STREICHER 


TABLE IV 

Responsiveness of Diaphragm of Scorbutic Guinea Pigs 
to Elevations of Extracellular Glycine 

In the first 5 experiments ono hemidiaphragm was suspended for 2 hr. in the arti¬ 
ficial medium. In the other 2 cases, 20 mAf of glycine/kg. body weight was fed, in 3 
doses an hour apart (2). The distribution ratios given in this table are the calculated 
ratios cellular concentration/extracellular concentration (6). Concentrations are in 
mg.-%. Animals 109 and 111 were normal animals. 


Expt. 

Pretest glycine N 

Glycine N, after exposure to high glycine 

Fluid 

Diaphragm 

Dist. ratio 

Fluid 

Diaphragm 

Dist. ratio 

109 

0.86 

2.86 

5.1 

1.77 

4.03 

3.4 

111 

1.10 

2.66 

3.6 

0.96 

2.41 

3.7 

112 

0.98 

2.66 

4.1 

1.80 

4.68 

3.9 

113 

0.41 

1.45 

5.5 

1.20 

2.80 

3.5 

114 

0.78 

1.92 

3.7 

1.16 

2.72 

3.2 

115 




11.1 

14.0 

1.8 

116 




10.0 

11.7 

1.6 


Temperature Effects 

Pairs of rat hemidiaphragms were shaken in an initially glycine-frec 
medium No. 1, one of the hemidiaphragms at 37°C., the other at 1°C. 
The “cold” diaphragm lost half as much glycine nitrogen in an hour as 
the “warm” one (e. g., 0.65 mg.-% compared with 1.30 mg.-%, as 
determined by the glycine found in the fluids). A coefficient of approxi¬ 
mately 2 for this 36° temperature difference is compatible with the view 
that the leakage of glycine from the tissue into a glycine-low fluid was a 
simple diffusion process. Subsequently the “warm” diaphragm was 
restored to a medium of glycine concentration similar to that of the 
plasma. Reassimilation against the gradient occurred in 2 hr. to 
approach the original distribution ratio ( e. g., 4.2, compared with the 
ratio of 4.7 in the intact rat). 

The temperature coefficient of the uptake of glycine was studied by 
placing one hemidiaphragm in the usual medium containing about 1 
mg.-% of glycine nitrogen, which is about twice the plasma concentra¬ 
tion, and equilibrating at 1°, 20°, 30° or 38°C. The other hemidia¬ 
phragm and the plasma were extracted for analysis at once after sacri¬ 
fice of the rat. The net transfer of glycine nitrogen to the diaphragm 
(i. e., mg.-% after equilibration minus mg.-% in the intact animal) is 



AMINO ACID CONCENTRATION. I 


103 



Fig. 1. The change in the concentration of glycine nitrogen of the diaphragm, in 
2 hr. in a medium containing about 1 mg.-% of glycine nitrogen, plotted against the 
temperature. The glycine nitrogen concentrations found, in mg. per 100 g. water, for 
the plasma and for the artificial medium at the end of the experiment were, respec¬ 
tively: for the experiment at 1°, 0.50 and 0.90; at 20°, 0.48 and 1.00; at 30°, 0.53 and 
0.97; at 38°, 0.53 and 1.06. 

shown in Fig. 1. Between 30° and 20°C. a reversal occurred in the 
direction of glycine transfer. While these experiments do not give us a 
temperature coefficient for the inward migration of glycine, they do 
indicate that the concentration process depends upon temperature- 
sensitive reactions. 

Discussion 

These experiments appear to show that the high glycine content of 
the diaphragm is maintained by an active concentrating process. The 
diaphragm glycine level was responsive to the fluid concentration, 
falling when placed in a medium of one glycine concentration and rising 
in a medium of slightly higher concentration. Cdycine could be elimi¬ 
nated from and then restored to the diaphragm by manipulating the 
level of glycine in the fluid. 

The stimulating effect of pyruvate and a-ketoglutarate, and the in¬ 
hibiting effect of several factors (anoxia, low temperature, presence of 
cyanide and particularly of 2,4-dinitrophenol (9)) bear upon the 





104 


HALVOR N. CHRISTENSEN AND JEAN A. STRETCHER 


question of the energy source for the concentrative assimilation of 
amino acids. Our observations on the distribution of a-amino nitrogen 
indicate that our conclusions are not limited to glycine but may apply 
for other amino acids. 

Gale (10) has found that gram-positive bacteria also require exer- 
gonic metabolism to concentrate certain amino acids from the suspend¬ 
ing fluid. Strains of Strep, faecalis do not establish equilibrium with a 
new environment either by gaining or by losing cellular glutamic acid, 
unless glucose or another appropriate energy source is present. Cells of 
S. aureus, in contrast (and more like the rat diaphragm) show a slow 
leakage of glutamic acid in the absence of glucose, which is checked 
rather than enhanced by the addition of glucose. Brain slices likewise 
appear to require energy-yielding metabolic reactions in order to 
assimilate glutamic acid (3). 

Human erythrocytes, with their low metabolic activity, contain 
a-amino acids (collectively by ninhydrin, correcting for glutathione) 
only 1.2-1.5 times as concentrated in the cell water as in the plasma 
water (11); glycine was concentrated about 1.7 times, alanine little, if 
at all (12). Similarly, the rate of entrance of these amino acids into 
erythrocytes relative to other tissues is extremely slow (12). 

Summary 

1 . The distribution of glycine and of a-amino acid nitrogen between 
excised diaphragm and artificial suspending medium has been investi¬ 
gated. 

2 . In a NaCl-NaHCCh medium containing Ca ++ ions and sodium 
pyruvate, the distribution of glycine between diaphragm and fluid 
was maintained at distribution ratios as high as in the intact animal 
for 1-3 hr. If the medium was richer in glycine than the plasma, gly¬ 
cine entered the diaphragm against the concentration gradient; at 
lower fluid concentrations glycine moved from the tissue to the medium. 

3. The concentrating activity was strongly inhibited by the following 
agents and conditions: anoxia, cyanide, 2,4-dinitrophenol at very low 
concentrations, potassium ion, and arsenite. a-Ketoglutarate and 
succinate were active in replacing pyruvate. Malonate had compara¬ 
tively little effect. These observations bear upon the energy source for 
the concentration of amino acids. 

4. The stimulating effect of pyruvate and the inhibiting effect of 



AMINO ACID CONCENTRATION. I 


105 


dinitrophenol were confirmed for the a-amino acids collectively as 
measured manometrically by ninhydrin. 

5. Male rats were observed to have about one-third higher plasma 
glycine levels than females. 

6 . The diaphragm of the guinea pig, whether scorbutic or normal, 
was similarly responsive as to its glycine content to the glycine level 
of the suspending medium. 

7. The temperature coefficient of the outward migration of glycine 
from a diaphragm placed in a glycine-free medium was compatible with 
the process of simple diffusion. The direction in which glycine moved in 
a glycine-rich medium was reversed from inward to outward by 
lowering the temperature, indicating that the steepness of the gradient 
was dependent upon temperature-sensitive reactions. 

References 

1. Van Slyke, D. D., and Meyer, G. M., J. Biol Chem. 16, 197 (1913-1914). 

2. Christensen, II. N., Streiciier, J. A., and Elbinger, R. L., ibid. 172, 515 

(1948). 

3. Stern, J. R., Biochem. J. 42, LVII (1948). 

4. Hamilton, P. B., and Van Slyke, D. D., J. Biol Chem. 150, 231 (1943). 

5. Hamilton, P. B., ibid. 158, 375 (1945). 

6. Alexander, B., Landwehr, G., and Selioman, A. M., ibid. 160, 51 (1945). 

7. Christensen, H. N., Cushing, M. K., and Streicher, J. A., Arch. Biochem. 

23, 106 (1949). 

8. Christensen, H. N., and Lynch, E. L., J. Biol Chem. 172, 107 (1948). 

9. Loomis, W. F., and Lipmann, F., ibid. 173, 807 (1948). 

10. Gale, E. F., Bull. Johns Hopkins Hosp. 83, 119 (1948). 

11. Christensen, H. N., and Lynch, E. L., J. Biol. Chem. 163, 741 (1946). 

12. Christensen, H. N., Cooper, P. F., Jr., Johnson, R. D., and Lynch, E. L., 

ibid. 168, 191 (1947). 



Concentration of Amino Acids by the Excised Diaphragm 
Suspended in Artificial Media. II. Inhibition of the 
Concentration of Glycine by Amino Acids and 
Related Substances 

Halvor N. Christensen, Mary K. Cushing and Jean A. Streicher 

From, the Division of Laboratories , Children’s M edical Center , Boston , Department 
of Biological Chemistry , Harvard Medical School , and the Department of 
Biochemistry and Nutrition , Tufts College Medical School , Boston 
Received March 10, 1949 

Introduction 

Whenever the amino acid concentration of the plasma was elevated 
by the feeding of any of a number of amino acids to the guinea pig, the 
free glycine content of the liver or muscle was decreased, or that of the 
plasma increased; or, as we interpreted the observation, the extent to 
which these tissues concentrated glycine was diminished (1). The mag¬ 
nitude of this effect was roughly proportional to the elevation of the 
plasma amino nitrogen, no matter which of several amino acids was 
fed. Conversely, high plasma glycine levels diminished the “concentra¬ 
tion” of non-glycine amino acids. This inhibition of the concentrating 
of one amino acid by another was interpreted as a competition based 
upon structural similarities. This type of inhibition has now been ob¬ 
served and studied using the isolated rat diaphragm, suspended in an 
artificial medium (2). 


Experimental 

Paired hemidiaphragms were shaken for 3 hr. at 37°C. in separate flasks containing 
2-2.2 ml. of the artificial medium (NaCl 103 m M } NaHC0 3 25 mM, CaCl 2 2.5 ml, 
sodium pyruvate 20 m M } in an atmosphere of oxygen containing 5% C0 2 ). To the 
medium used for one hemidiaphragm the amino acid was added at a concentration of 
21.4 mM . The media were replaced hourly. In the cases where the amino acid had a 
net charge at pH 7.4 ( e.g. } sodium glutamate, arginine monohydrochloride) the 
NaCl of the experimental medium was correspondingly diminished by 21 mM/1. 
Glycine was incorporated into all media at a concentration estimated to be similar to 

106 



AMINO ACID CONCENTRATION. II 


107 


that of the plasma, i.e. } about 0.48 mg.-% of glycine nitrogen for male rats, about 
0.35 for females. During the first and second hours of incubation there was a barely 
measurable augmentation of the glycine concentration of the 2 successive portions of 
media supporting the experimental hemidiaphragm, relative to the media for the 
control hemidiaphragm due to loss of glycine from the tissue. The 2 final solutions 
serving for the third hour were analytically identical in their final glycine concentra¬ 
tions. Therefore, it was necessary to compare only the final glycine concentrations of 
the 2 hemidiaphragms. The glycine concentrations of the plasma and the 2 final media 
were determined to confirm the similarity of the glycine concentration of these three. 
Other experimental and analytical details have been described previously (2). 

Effect of Excesses of Various Amino Acids upon glycine Determination 

Three of the ammo acids added in excess (20-60 moles/mole of glycine) interfered 
so strongly with the glycine analyses as to preclude their study, unless they were to be 
removed before glycine analyses. Histidine and tryptophan prevented good recoveries 
of the formaldehyde resulting from the reaction of glycine and ninhydrin. Alexander, 
Landwehr and Scligman (3) suggested that such effects were due to combination of 
formaldehyde with amino acids. Phenylalanine in excess (30 moles/mole of glycine) 
interfered by giving a volatile product, presumably phenylacetaldehyde, which halved 
color formation and, in larger excesses, led to turbidity in the color reaction of for¬ 
maldehyde with chromotropic acid. These interferences depended upon the presence 
of large excesses of the amino acids. j8-Phenylglycine did not interfere. 

Several other amino acids (e.g. } aspartic and glutamic acids, proline, ornithine) 
diminished the recoveries of glycine by 10 or 15% when present in excesses of 60 
moles/mole of glycine. In the proportions actually found present in diaphragms at the 
end of an equilibration ( e . g ., 7:1, 5:1), the interference was not measurable. By 
doubling the ninhydrin taken (using 2% rather than 1% ninhydrin solution), the 
interference was decreased. Other amino acids studied ( e.g ., leucine, valine, serine) 
were without effect upon the glycine analyst's at excesses of 60 moles/mole of glycine. 
Recently, Krueger (4) has reported interferences by a number of amino acids with the 
method of Alexander et al. for the determination of glycine. These losses are, for the 
most part, far larger than we have encountered. The only suggestion that we can make 
to account for these differences is the empirical and highly critical nature of the con¬ 
ditions required for the distillation. In agreement with our results, Kreuger found that 
increasing the excess of ninhydrin diminished the interference. 

Because glycine is one of the more abundant of free amino acids in the diaphragm 
(about one mole out of every 8, excluding glutamine) interference by such amino acids 
as histidine and tryptophan normally present in diaphragm should not be serious, 
although the possibility of such losses exist s. The levels of these free amino acids in the 
diaphragm are not known. Added glycine could, however, be recovered satisfactorily 
from diaphragm filtrates. 


Results 

The consequences of including amino acids and other substances 
structurally related to glycine in the medium are illustrated in Table I. 



108 H. N. CHRISTENSEN, M. K. CUSHING AND J. A. STREICHER 


TABLE I 

Changes in the Olycine Concentration of Diaphragm Resulting from the Presence 
of Amino Adds and Related Substances in the Suspending Fluid 
One hemidiaphragm of a rat was equilibrated in the control medium, the other 
simultaneously in a medium identical except for the presence of the amino acid at 
21.4 mAT concentration. After 3 hr., the media being renewed hourly, the tissue 
samples and final fluid media were analyzed. The values given are the percentages by 
which the glycine contents of the experimental hemidiaphragms were less than the 
glycine contents of the control samples. Since the glycine contents of the two media 
were practically identical at the end of each experiment, the values given correspond 
also with the decrease of the distribution ratios. 


Substance 

Decrease in 
glycine con¬ 
centration 

t 

Substance 

Decrease in 
glycine con¬ 
centration 


per rent 


per cent 

DL-Serine 

-37 

D-Leucine vs. L-leucine 

0 

L-Valine® 

-30 

dl-A spartic vs. L-aspartic 

0 

L-Prolinc 

-35 

d i/-a- Ami no-a-met hy lbu ty r i c 
acid 

-20 

L-Leucinc 

-25 

a-Aminoisobutyric acid 

-24 

L-Aspartic acid 

-35 

Betaine 

-15 

l- Asparagine 

-33 

Glycocyamine 

-14 

LrGlutamic acid 

-21, -16 

Creatine 

© 

1 

00 

L-Glutamine 

-18 

/3-Alanine 

-3 

L-Ornithine 

-20 

DL-/3-Aminobutyric acid 

0 

L-Arginine 

-28 

p-Aminobenzoic acid 

-4 

10 Amino acids 6 

-29 

N-Formyl-L-valine d 

-1 

7 Amino acids c vs. l - 
leucine 

-17 

Niacinamide 

-2 


a From Dr. R. C. Corley, Purdue University. 

6 irAspartic, L-glutamic, L-asparaginc, Lrglutamine, L-arginine, L-ornithine, ir 
valine, Lrlcucine, DL-serine and Lrproline, each 2.1 m M . 

c DL-Aspartic, L-asparagine, L-arginine, dl- valine, L-leucinc, DL-serine and Lrpro¬ 
line, each 3.0 mM. The control medium contained L-leucine, 21.4 mAf. 
d Gift of Dr. J. W. Hinman, Research Laboratories, The Upjohn Company. 

All a-amino acids studied were inhibitory. Seven a-amino acids di¬ 
minished the glycine retention of diaphragm by 25-37%. The effect of 
glutamine and glutamic acid appeared to be smaller, 16-21%. When 
L-leucine and D-leucine were added to the suspending media for each of 
a pair of hemidiaphragms, the final glycine contents of the two hemi¬ 
diaphragms were identical. The same was true for L-aspartic acid and 
DL-aspartic acid. Three amino acids in which the amino group was not 



AitlNO ACID CONCENTRATION. II 


109 


in the a position (/3-alanine, DL-/3-aminobutyric acid, p-aminobenzoic 
acid) had no significant effect upon the glycine distribution. Two a- 
methyl-a-amino acids also showed inhibiting effects. Introduction of 
an N-formyl group into valine abolished its inhibitory action. Betaine 
and glycocyamine were somewhat inhibitory to glycine concentration, 
but under the present conditions creatine did not appear so. Inhibition 
occurred whether the a-amino acid concentration was brought to 21 
mM using a single amino acid or a mixture of 7 or 10 amino acids. 

Discussion 

The plausibility of attributing the inhibition of the type observed to 
metabolic products of the amino acid rather than to the amino acid 
itself is much less with the isolated diaphragm. This is particularly true 
in the case of the two a-methyl-a-amino acids, which are degraded little 
if at all by the dog (5). The present technique did not permit us to 
examine for the criterion of the competitive nature of the inhibition, 
since we could vary the glycine concentration only narrowly without 
stimulating excessive transfers of glycine. Several inferences as to the 
structural requirements for competition may be made: 

1. The size and charge of the side-chain did not appear to be impor¬ 
tant in determining the degree of inhibition, as if the differentiation 
between glycine and other amino acids were simply on the basis of the 
absence or presence of a side-chain. 

2 . Competition did not occur when the amino group was not in the 
a position. Similar results were obtained upon feeding amino acids to 
guinea pigs. 

3. The presence of N-alkyl groups did not necessarily eliminate 
competition (e.g., proline, betaine, glycocyamine). 

4. N-Acylation, with loss of the positive charge, eliminated inhibition 
in the case of formyl valine. 

5. d- or L-amino acids appeared to be equally effective. This is per¬ 
haps not surprising in view of the absence of optical asymmetry in 
glycine. 

6 . A high a-amino acid concentration, rather than an extreme con¬ 
centration of a single amino acid, appeared responsible for the inhibi¬ 
tion of glycine retention. 

The probable relation of the loss of amino acids from the interior of 



110 H. N. CHRISTENSEN, M. K. CUSHING AND J. A. STREICHER 

the cell, as a result of the process reported here, to the nutritional 
disadvantage of diets containing excessive quantitites of one or several 
amino acids has already been discussed (1). Certain amino acid mixtures 
developed recently can be infused intravenously with extreme rapidity 
into patients, producing very high plasma amino acid concentrations, 
but no nausea. Until the nutritional consequences of such high amino 
acid levels (which do not represent all amino acids) are known, it is 
suggested that the convenience of rapid infusion may be overempha¬ 
sized. 

Summary 

Each of a group of a-amino acids inhibited, at a 21 m M concentration 
the concentrating of glycine by the excised diaphragm suspended in an 
artificial medium. The effect was obtained only when the amino group 
was in the a position. An N-formyl group abolished the effect of valine. 
N-alkyl groups did not necessarily eliminate the inhibition. No optical 
specificity was observed. These effects are related to similar inhibitions 
observed in the intact animal. 

References 

1. Christensen, H. N., Streicher, J. A., and Elbincer, R. L., J. Biol. Chern. 172, 

515 (1948). 

2. Christensen, H. N., and Streicher, J. A., Arch. Biochem. 23, 96 (1949). 

3. Alexander, B., Landwehr, G., and Seligman, A. M., J. Biol. Chern. 160, 51 

(1945). 

4. Krueger, R., Helv . Chirn. Acta. 32, 239 (1949). 

5. Leighty, J. A., and Corley, R. C., J. Biol. Chern. 120, 331 (1937). 



Substrate Concentration and Specificity of 
Choline Ester-Splitting Enzymes 1 

Klas-Bertil Augustinsson 2 

From the Department of Neurology, College of Physicians and Surgeons , 
Columbia University, New York 
Received March 16, 1949 

Introduction 

For the assumption that the release and removal of acetylcholine are events essen¬ 
tial for conduction as proposed by Nachmansohn (1), it is of interest to know how far 
the esterase present in conductive tissue is specific for acetylcholine and distinguish¬ 
able from otTier esterases. Stedman, Stedman and Easson (2) and Simonart (3) found 
that propionylcholine and butyrylcholine are split by serum esterase at a higher rate 
than acetylcholine. Their findings, confirmed by Glick (4), do not support the assump¬ 
tion that the serum esterase is an enzyme specific for acetylcholine. Nachmansohn 
and Rothcnberg (5) demonstrated that, in striking contrast to serum esterase and 
many other esterases, the esterase in a great variety of conductive tissue, nerve, and 
muscle, hydrolyzes propionylcholine at a lower rate than acetylcholine or, in a few 
cases, at the same rate, whereas butyrylcholine is split at a very low rate or not at all. 
Only red blood cell esterase showed the same pattern of hydrolysis rates when different 
substrates were used. Nachmansohn and Rothcnberg came to the conclusion that the 
esterases in conductive tissue have a high, though not absolute, specificity for acetyl¬ 
choline, since no other ester tested is split at a higher rate. 

A second difference by which the choline ester-splitting enzymes may be distin¬ 
guished, is the effect of substrate concentration. As was shown by Alles and Hawes (6), 
there is a definite optimum concentration of acetylcholine for red blood cell esterase. 
Excess of the ester strongly inhibits the enzyme. This finding has been confirmed and 
extended to the esterases of brain (7) and many other conductive tissues (5,19). How¬ 
ever, the relationship between activity and substrate concentration may differ for a 
given enzyme from substrate to substrate. The importance of a careful consideration 
of the substrate concentration in following the enzymic hydrolysis of various esters 
has been recently demonstrated by Augustinsson (8,9). The erythrocyte esterase 
splits, e. g., acetyl-/3-methylcholine more slowly than acetylcholine at low substrate 
concentration but more rapidly when the substrate concentrations are high. This 

1 The work has been carried out under a grant from the U. S. Public Health Service. 

2 Fellow of the American-Scandinavian Foundation, with additional grants from tho 
Swedish National Medical Research Council and the U. S. Public Health Service. 
Present address: Biochemical Institute, University of Stockholm. 


Ill 



112 


KLAS-BERTIL AUGUSTINSSOK 


shows that, although both esters inhibit the enzyme in high concentration, the opti¬ 
mum substrate concentration is not the same for the two esters. 

Bodansky (10) reported that the brain and erythrocyte cholinesterases 
split triacetin, human brain esterase more rapidly than acetylcholine. 
To clear up this problem and other conflicting results by various inves¬ 
tigators regarding the specificity of choline ester-splitting enzymes, the 
following study has been performed. 

Methods 

The esterase activity was measured by the Warburg manometric method. In modi¬ 
fication of the usual technique, 1.60 ml. of the substrate solution was placed in the 
main compartment of the vessel and 0.40 ml. of the enzyme solution in the side bulb 
(8). Substrates and enzyme preparations were dissolved in a bicarbonate-buffer 
solution containing 0.15 M NaCl, 0.04 M MgCl 2 and 0.025 M NaHC0 3 (5). The 
hydrolysis was carried out in a gas mixture of 95% N 2 -5% C0 2 . After attaining 
temperature equilibrium, the first manometer was red and the contents of the other 
flasks were mixed at zero time. At 1 min. intervals, the contents of the other flasks 
were mixed. Each manometer was red at 6 min. intervals, 1 min. between each manom¬ 
eter reading. Readings were made continuously for 36 min. Measurements were 
made at 23-24°C. 

The output of C0 2 expressed in /d. was plotted against time. The interpolated 30 
min. value, minus the amount of C0 2 evolved during the same time period by non- 
enzymic (spontaneous) hydrolysis, evaluated in each case, has been used as unit (& 3 o) 
in expressing the esterase activity. 

The tissue was homogenized according to the technique described by Potter and 
Elvehjem (11). Aliquot parts of the homogenized tissue or of the supernatant fluid 
obtained after centrifugation were then used as enzyme preparations. 

Fresh solutions of the substrates (in bicarbonate-buffer solution) were made for each 
experiment. The esterase activity was measured at 6 various substrate concentrations. 
The final percentage concentrations of the substrates after mixing with the enzyme 
solutions were: 2.00, 0.60, 0.20, 0.060, 0.020, and 0.006. 

The spontaneous hydrolysis of the substrates at various concentrations was ob¬ 
served for each substrate. The enzymic activity, expressed by 6 3 o, was plotted against 
pS , the negative logarithm of molar substrate concentration. 

Selection of M alerial 

For the study of enzyme kinetics in general the ideal condition would be the use of 
pure enzymes. Such preparations, however, are not readily available. For the present 
study, two highly purified preparations were available in which the enzyme may be 
considered as nearly pure: esterase from human serum and from electric tissue of 
Eledrophorua electricus . 3 The first was used as an example of a choline ester-splitting 

9 I am greatly obliged to Dr. A. Goldstein, Harvard Medical School, for the prepara¬ 
tion of the scrum esterase (Dr. E. Cohn’s fraction IV-6-3), and to Dr. M. A. Rothen- 
berg, of this laboratory, for the electric tissue esterase. 


CHOLINE ESTER-SPLITTING ENZYMES 


113 


enzyme, which, on the basis of previous experiments, does not split acetylcholine at 
the highest rate; the second as an example of an esterase of conductive tissue assumed 
to have the highest affinity to acetylcholine. 

As pure enzyme preparations from other tissues were not available, those conduc¬ 
tive tissues were selected which contain a high concentration of the enzyme, so that 
relatively small amounts of tissue could be used. In this way, the probability of inter¬ 
ference by other enzymes or proteins becomes smaller. Examples of 4 types of conduc¬ 
tive tissue were chosen: nerve and muscle tissue of vertebrate and invertebrate animals. 
One of the highest concentrations of the choline ester-splitting enzyme in mam¬ 
malian nerve tissue has been found in the nucleus caudatus, and in the invertebrate 
nerve tissue in the head ganglion of squid. Both were found to be relatively specific 
for acetylcholine (5). In addition to the figures published previously, new data con¬ 
cerning the activity and relative specificity of esterases in conductive tissue are 
summarized in Table I. 4 On the basis of these data, Lebistes muscle was chosen as 

TABLE I 

Rate of Hydrolysis of Acetylcholine ( ACh ), Propionylcholine (PrCh ) and 
Butyrylcholine ( BuCh ), and Triacetin (7’A) by Various Conductive 
Tissue at Optimum Acetylcholine Concentration 


Tissue 

Animal 

ACh split 
mg./g./hr. 

PrCh 

BuCh 

TA 

In per cent of ACh hydrolysis 

Brain 

Pigeon 

240 

80 

0 

5 


Sparrow 

300-350 


5 



Humming bird 

300 


0 



Turtle 

80-100 

70 

0 

4 


Frog 

60 

50 

0 

15 

Muscle 

Lebistes 

300-400 

75 

12 

10 


Lizard 

40 


3 



Nereis 

60-65 



20 


Lumbricus 

140-160 

70 

15 

10 

. 


vertebrate and the body wall of Lumbricus as invertebrate muscle. 

Erthrocyte esterase has been tested since it is known to be similar to nerve and 
muscle esterase (5,8). The blood of Helix aspersa has also been studied, since it has 
been found that the cell-free blood of Helix pomatia is similar to the nerve and ery¬ 
throcyte esterase, although different in a few respects (8,9). The dart sac of Helix 
pomatia has a very high esterase activity (8,9). It appeared desirable to obtain addi¬ 
tional information regarding some peculiar properties of this enzyme. 

4 These determinations were carried out by Mrs. Claire Marshall and Mrs. Emily 
Feld-Hedal in Dr. Nachmansohn’s Laboratory and the unpublished data were kindly 
put at my disposal. 



114 


KLAS-BERTIL AUGUSTINSSON 


Snake venom from Colubridae has a very high cholinesterase activity. Zeller (12) 
assumes that the snake venom cholinesterase is different from his “s” type and “e” 
type cholinesterases and classifies it as a new “e” type (colubercholinesterase). 

Results 

1. Purified Esterase Prepared from Human Blood Serum 

The results obtained with the highly purified esterase from human 
blood plasma are presented in Fig. 1. The dissociation curve for acetyl¬ 
choline is the same as that obtained with the partly purified preparation 
from horse serum described previously; the pK t is 2.5 (8). Propionyl- 



Fig. 1. Activity-p$ curves for the enzymic hydrolysis of various esters by a puri¬ 
fied esterase preparation from human blood serum (Fraction IV-6-3 according to 

Cohn (18)). 0.8 mg. enzyme per vessel. •-•, Acetylcholine chloride (ACh); 

•-•, Propionylcholine chloride (PrCh) .#, Butyrylcholine 

chloride (BuCh); O-O, d^Acetyl-0-methylcholine chloride; O.O, 

Benzoylcholine chloride; X-X, Triacetin (TA); X.X, Methyl 

butyrate; □-Q Nu 2017; □.□, Nu 2416. 









CHOLINE ester-splitting enzymes 


115 


choline is split at a higher rate than acetylcholine and butyrylcholine 
at a still higher rate, confirming earlier findings. Both these esters give 
the same value of pK, as acetylcholine. The hydrolysis of benzoyl- 
choline is depressed by excess of substrate. The optimum substrate 
concentration is at a pS of 2, which is again in agreement with the 
previous observation. Methyl butyrate, tributyrin and triacetin are 
split very slowly. The affinities of the noncholine esters are definitely 
lower than that of the choline esters. 

Two other compounds similar in structure to prostigmine were 
tested: 3-acetoxyphenyl trimethylammonium methylsulfate (Nu 2017) 
and 2-acetoxybenzyl trimethylammonium bromide (Nu 241(5): 5 


0—CON(CH,)j 

A 


V 


-N(CH,), 

Br~ 


0—COCH 3 


()—COCH, 


Ln(CII 3 ) 3 
—so 4 ch 3 


A, 

V 


■CH,N(CH a )j 

Br~ 


Prostigmine bromide 


Nu 2017 


Nu 2416 


Nu 2017 has a relatively high affinity for this esterase ( pK , 3.7). 
Nu 2416, on the other hand, is split at a very low rate. The differences 
in pK„ for the hydrolysis of acetylcholine and Nu 2017 may explain 
why acetylcholine is split at a lower rate than Nu 2017 at low substrate 
concentration, at a higher rate when the concentrations are higher. 
The affinity of Nu 2017 for the enzyme is higher, but the hydrolysis 
rate of the enzyme-substrate complex is lower than for acetylcholine. 
The low value of K a in the case of Nu 2017 is consistent with the high 
affinity of prostigmine for the enzyme. 

The spontaneous hydrolysis of Nu 2017 and Nu 2416 is more rapid 
than that of acetylcholine. At 23°C., the amounts of C® 2 in gl. evolved 
during 30 min. from 2.0 ml. of 0.025 M bicarbonate solution are the 
following: 


Molar 

cone. 

Acetylcholine 

chloride 

Nu 2017 

Nu 2416 

ixio-« 

8 

18 

14 

3X10"* 

5 

8.5 

4 

ixio-» 

3 

5 

1 

3XlO-> 

2 

4 

0 

1X10~‘ 

1 

3 

0 


5 1 wish to express my thanks to Dr. J. A. Aeschlimann of Hoffmann-LaRoche, Inc., 
Nutley, New Jersey, for supplying these compounds. 



116 


KLA8-BERTIL AUGUSTINSSOft 


2. Purified, Esterase Prepared from Electric Tissue 
of Eledrophorus electricus 

The results obtained with the crude preparation and the highly 
purified solution from the electric tissue of Eledrophorus electricus are 
presented in Fig. 2. The activity substrate concentration relationship 
has been studied for the hydrolysis of various esters. In previous experi¬ 
ments, the substrate concentration was varied only for the acetyl¬ 
choline hydrolysis, whereas the other substrates were tested only at 
concentrations close to the optimum of acetylcholine hydrolysis. 



Fig. 2. Activity-/)* 1 ? curves for the enzymic hydrolysis of various esters by the 
electric organ of Eledrophorus electricus. Left: crude extract; right: purified prep¬ 
aration (1 mg. of protein splitting 20 g. of ACh/hr.). Symbols as in Fig. 1. 

Propionylcholine is split at about the same rate as acetylcholine, and 
butyrylcholine not at all. The patterns are the same for the crude and 
highly purified preparation and confirm the previous finding (5) that 
the properties of the esterase of homogenized tissue are unchanged 
after purification and that acetylcholine at high concentration depresses 
the enzyme activity. The optimum acetylcholine concentration is 
3 X 10“ 8 M. The curves are symmetrically bell-shaped, consistent with 
Haldane’s hypothesis that the formation of an ES 2 complex incapable of 
breaking down at high substrate concentration may be valid for the 
hydrolysis of acetylcholine (13). The optimum substrate concentrations 
are the same for acetyl- and propionylcholine. The contrast to the 



CHOLINE ESTER-SPLITTING ENZYMES 


117 


serum cholinesterase is striking. Acetyl-j3-methylcholine is split at a 
much lower rate. The optimum substrate concentration (pS op t 1.6) is 
higher than in the case of acetylcholine and propionylcholine, and the 
same as for the erythrocyte and brain cholinesterase (8). 

The enzyme splits triacetin, but the activity substrate concentration relationships 
for the hydrolysis of acetylcholine and triacetin are entirely different. At high sub¬ 
strate concentration triacetin is split at a higher rate than acetylcholine; at low sub¬ 
strate concentrations, triacetin is split at a much lower rate. The affinity of the enzyme 
for acetylcholine is much greater than for triacetin. 

The effect of the esterase of electric tissue on the synthetic substrates also differs 
from that produced by the serum esterase; Nu 2416 is split slowly and Nu 2017 not at 
all. 

8. Mammalian Nerve Tissue (Nucleus Caudatus of Ox) 

In Fig. 3 are shown the results with an esterase obtained from mammalian brain 
tissue (nucleus caudatus of ox). The substrates used were acetylcholine, butyrylcho- 
line, and triacetin. The data confirm the previous finding that the enzyme is inhibited 
by high acetylcholine concentrations. Bulyrylcholine is split at. a very low rate. The 
pattern of the activity-pfi curves with the 3 substrates used is strikingly similar to 
that of the highly purified electric tissue esterase. 



Fig. 3. Activity-ptf curves as in Fig. 1 obtained with 
an extract from nucleus caudatus (ox). 

4. Invertebrate Nerve Tissue (Head Ganglion of Squid) 

The esterase of the head ganglion of the squid splits acetylcholine at a higher rate 
than the other substrates tested, confirming the previous statement (5). Propionyl¬ 
choline is split at a lower rate, and butyrylcholine still more slowly. The activity- 
substrate concentration relationships are the same as those obtained with the electric 




118 


KLAS-BERTIL AUGU8TINSS0N 


tissue (Fig. 4). As in the case of electric tissue esterase, the optimum substrate concen¬ 
tration for the hydrolysis of acetyl-0-methylcholine (pS opt 1.6) is higher than that for 
acetylcholine (pS QP 1 2.5). Therefore, acetyl-0-methylcholine may be split at the same, 
or even a higher, rate than acetylcholine when the concentrations of the two sub- 



Fig. 4. Activity-p£ curves as in Fig. 1 obtained with 
an extract from squid ganglion. 

strates arc high, close to 0.1 M. In low concentrations, on the other hand, close to 
0.001 A/, acetylcholine is split much more rapidly than acetyl-0-methylcholinc. These 
differences apparently explain the findings of Richards and Cutkomp (14) and Tobias 
and coworkers (15), that insect nerve tissue was more active on acetyl-0-methyl- 
choline than on acetylcholine, since the molar concentrations of the substrates used 
by these authors were high (0.1 M). Benzoylcholine is not split at all by the squid 
ganglion esterase. 

The effect of substrate concentration on the hydrolysis of triacetin is similar to that 
found in the electric tissue esterase. The esterase of squid ganglion splits Nu 2017 very 
slowly. 



Fig. 5. Activity-p*5 curves as in Fig. 1 obtained with 
an extract from Lebistes tail muscle. 




CHOLINE ESTER-SPLITTING ENZYMES 119 

5 . Vertebrate Muscle (Lebistes reticulatus) 

The results obtained with extracts prepared from homogenized tail muscle of 
Lebistes reticulatus are shown in Fig. 5. The optimum concentration of acetylcholine is 
the same as that obtained with the esterases of nerve tissue. Butyrylcholine is split at a 
very low rate; the optimum concentration for butyrylcholine is slightly lower than for 
acetylcholine. The rate of hydrolysis of triacetin differs in the same way as with nerve 
esterase. 


6. Invertebrate Muscle (Lumbricus lerrestris ) 

The act.ivity-p& curves for the hydrolysis of different substrates by washed, homog¬ 
enized Lumbricus muscle (Fig. 6) resemble those with vertebrate muscle, but. the 



Fig. 6 . Activity-p£ curves as in Fig. 1 obtained with 
an extract from Lumbricus muscle. 

optimum substrate concentration for choline esters is higher (p$ op t 1.7). Only benzoyl- 
choline, which is split at a low rate at all concentrations, has an optimum at a lower 
concentration. 


7. Erythrocytes from Human Blood 

Human red blood cells were washed 3 times with 0.9% NaCl solution and then 
hemolyzed with twice the volume of distilled water. The hemolyzatc was diluted with 
10 volumes of bicarbonate Ringer’s solution. This solution was used in the experiments 
and the results are shown in Fig, 7. The optimum substrate concentration for acetyl¬ 
choline is 3 X 10-* M. Propionylcholine is split at a lower rate with approximately the 
same optimum. Butyrylcholine is split at a very low rate or not at all. At the optimum 




120 


KLAS-BERTIL AUGUSTINSSON 


acetylcholine concentration, acetylcholine is split 10 times faster than triacetin. In 
0.1 M concentration, triacetin is split at a rate which is 2.7 times higher than for 
acetylcholine. 



Fin. 7. Activity-p<S curves as in Fig. 1 obtained with 
hemolyzate of human erythrocytes. 

The pattern for this esterase shows the similarity to the nerve esterases previously 
stated (7,5,8). Both enzymes seem to be localized in the surface membrane and certain 
similarities have been reported concerning the properties of nerve and erythrocyte 
membranes. 



Fig. 8. Activity-p& curves as in Fig. 1 obtained 
with the blood from Helix aspersa. 

8. Blood from Helix aspersa 

The results obtained with the blood of Helix aspersa are shown in Fig. 8. The enzyme 
displays its optimum activity at pS oP t 2.7 when acetylcholine is employed as sub- 




CHOLINE ESTER-SPLITTING ENZYMES 121 

strate. In the hydrolysis of propionylcholine, the optimum substrate concentration 
is higher, about the same as has been previously obtained in the hydrolysis of acetyl- 
0-mcthylcholine (pS ovt 1.5) (8). This difference in pS op t explains why propionyl¬ 
choline is split at a higher rate than acetylcholine when the substrate concentrations 
are high (0.1 M ). Butyrylcholine is split at a much lower rate and the optimum is the 
same as for propionylcholine. Triacetin gives the usual dissociation curve. The pattern 
thus demonstrated for the Helix blood is not quite the same as that, found for the 
nerve-muscle-erythrocyte esterases. Differences in other respects have been reported 
( 8 ). 

9. Dart Sac from Helix aspersa 

It has been pointed out that, the dart sac enzyme displays its maximum activity at a 
higher acetylcholine concentration than is observed with the erythrocyte and brain 
esterases (9). Fig. 9 confirms this result with the dart sac from Helix aspersa. Prop¬ 
ionylcholine is split at a low rate and butyrylcholine at a still lower rate. The opti¬ 
mum substrate concentration for all 3 choline esters is 3 X 10 2 M , which is 10 times 
higher than for the nerve and muscle esterases. 


7 

/ 



Fig. 9. Activity-p$ curves as in Fig. 1 obtained 
with the dart sac from Helix aspersa. 

Triacetin is split at an extremely high rate in contrast to the findings obtained with 
all esterases hitherto described. It was assumed that the same enzyme splits acetyl¬ 
choline and tributyrin (8). However, the very high activity toward triacetin suggests 
the presence of a second esterase distinct from the acetylcholine-hydrolyzing enzyme. 
To test this assumption, the rate of hydrolysis was measured in a mixture of acetyl¬ 
choline and either triacetin or butyrylcholine. The amount of CO 2 evolved was then 
compared with the amounts when the substrates were hydrolyzed separately. Table 
II shows the results. Butyrylcholine inhibits the hydrolysis of acetylcholine, which 




122 


KLAS-BERTIL AUGUSTINSSON 


TABLE II 

Rate of Enzymic Hydrolysis of Acetylcholine (ACh), Butyrylcholine ( BuCh ) and 
Triacetin (TA), and of Mixtures of these Substrates by the Dart Sac 
Rate is expressed as ** 1 . CO 2 evolved in 30 min. ( 630 ). Substrate concentration in 
each case as follows: ACh: 1.1 X 10”* M ; BuCh: 9.5 X. 10 ~ 3 M ; TA: 9.2 X 10 ” 3 M 


Substrate 

frio 

ACh 

80 

BuCh 

11 

TA 

67 

ACh -f- BuCh 

67.5 

ACh+TA 

131 


indicates that these 2 substrates may compete for the same enzyme. In a mixture of 
acetylcholine and triacetin, an additive effect is obtained. This apparently indicates 
that the 2 substrates are split by 2 esterases, acting independently of each other. It is 
possible that both enzymes split acetylcholine, and this may explain the higher opti¬ 
mum acetylcholine concentration. 

It was observed that the dart sac of Helix aspersa consists of 2 parts: an outer and 
an inner tissue. It is possible that one esterase is in the outer, the other esterase in the 
inner shell, which may be muscular element. Analogous observations have not been 
reported for Helix pomatia. 

10. Snake Venom from Naja naja 

In the present experiments, a dried and crystallized preparation of the venom was 
used. The activity of this preparation was of the same magnitude as for the electric 



Fig. 10. Activity-ptf curves as in Fig. 1 obtained with the dried and crystallized 
venom from Naja naja. (ACh 1: 0.12 mg. of fresh venom per vessel; ACh II: 0.4 mg 
of 11 year old venom per vessel.) 



CHOLINE ESTER-SPLITTING ENZYMES 


123 


organ, on the basis of dry weight. Acetylcholine in high concentration inhibits the 
activity and the optimum substrate concentration was 3 X 10~ 8 AT, the same as was 
found for the nervc-muscle-erythrocyte esterases (Fig. 10—ACh I). Propionylcholine 
is split at a lower rate than acetylcholine, but the optimum substrate concentration is 
slightly higher and about the same as for acetyl-/3-methylcholine; the latter is split at 
a lower rate. Butyrylcholine and benzoylcholine are practically unaffected. Triacetin 
gives the usual curve. 

/Discussion 

Two essential results emerge- from the experiments reported. 

1 . Importance of Substrate Concentration . Esterases of the type found 
in serum show the familiar dissociation curve when the activity is 
plotted against the log of acetylcholine concentration. This relationship 
is strikingly different from that found with esterases of conductive 
tissue and erythrocytes, and of a few special cases like the Helix blood 
and the snake venom. In addition, the various choline esters frequently 
have different substrate optima. For example, whereas acetylcholine 
does not inhibit the serum esterase in relatively high concentration 
(0.1 Af), benzoylcholine has a substrate optimum at about pS = 2. 
The activity of the second type of esterase has an optimum for acetyl- 
0-methylcholine which differs markedly from that for acetylcholine, 
propionylcholine, and butyrylcholine. Therefore, there is no fixed ratio 
between the rates of hydrolysis of these 3 esters and that of acetyl-]3- 
methylcholine. Moreover, even for these 3 esters, optimum substrate 
concentration may not always be the same, as is demonstrated by the 
results with Helix blood esterase. 

Considering the hydrolysis of non-choline esters by these esterases, 
the importance of the activity substrate concentration relationship 
becomes still more obvious. For instance, triacetin is split by the ester¬ 
ases of the type found in conductive tissue at a low rate at the optimum 
acetylcholine concentration, but at high substrate concentration, the 
rate of hydrolysis of triacetin may be even higher than that of acetyl¬ 
choline. The affinity of such esterases is in all cases much higher for 
acetylcholine than for triacetin. Testing the hydrolysis rate at that 
high concentration only gives a distorted picture. One example of this 
is Bodansky’s finding that human brain esterase splits triacetin faster 
than acetylcholine. Another example is the statement of Mendel, 
Mundell and Rudney (16) that, in a mixture of esterases, a quantita¬ 
tive distinction may be made between 2 types by the use of acetyl-j8- 
methylcholine and benzoylcholine. The data presented in this paper 
are incompatible with this view. 



124 


KLAS-BERTIL AUGUSTINSSON 


It is, at present, impossible to evaluate the physiological significance 
of the relationship between enzyme activity and substrate concentra¬ 
tion. The concentration at which acetylcholine may appear during 
activity at the site of action is unknown. But is it known that the 
enzyme is present in an excess of about 10 times, leaving a considerable 
margin of safety (20). A similar situation may be assumed for substrate 
concentration, namely, that it usually remains below the optimum; 
maximum velocity due to optimum concentration may occur only in 
special conditions, such as pathological disturbances. 

2. Specificity of Acetylcholine Hydrolyzing Enzyme. The second im¬ 
portant outcome of the data presented is a confirmation and extension 
of the observations of Nachmansohn and Rothcnberg on the specificity 
of the esterase present in conductive tissues and erythrocytes. The test 
of the hydrolysis rates of propionylcholinc and butyrylcholine appears 
indeed essential for distinguishing this type of esterase from that pres¬ 
ent in serum and pancreas. Without the use of these two choline esters, 
all those enzymes were defined as cholinesterases, which, at optimum 
conditions, hydrolyze acetylcholine at a higher rate than any other 
ester (8). Thus, the type of esterases present in conductive tissue— 
and erythrocytes -did not appear to have sufficiently strong substrate 
specificity to justify a sharp distinction between this type and other 
esterases. However, the present study confirms the view of Nachman¬ 
sohn and Rothenberg that the esterases present in conductive tissue 
may be distinguished from other choline ester-splitting enzymes by the 
use of propionylcholine and butyrylcholine. 

The patterns of the aetivity-p<S curves in nerve and muscle tissue of 
vertebrate and invertebrate and in the red blood cells are indeed 
strikingly similar in all essential features. The affinity of acetylcholine 
to the enzyme is high in contrast to non-choline esters. The bell-shaped 
curve obtained if the activity is determined as a function of substrate 
concentration is another distinction and confirms the previous findings 
of various authors of the inhibitory effect of high acetylcholine concen¬ 
tration on this special enzyme. 

All these features support the assumption of the existence of a type 
of enzyme relatively specific for acetylcholine and distinctly different 
from other choline ester-splitting enzymes. Therefore, and in view of 
the rather confusing terminology used by various investigators, 
Augustinsson and Nachmansohn have proposed for this type of enzyme 
the term acetylcholine-esterase (ACh-esterase) (17). In addition to the 



CHOLINE ester-splitting enzymes 


125 


esterase present in all conductive tissues and erythrocytes, there are 2 
special cases in which the esterases appear to have similar features: 
Helix blood and snake venom. The snake venom contains a great 
variety of enzymes. The hemolytic effect of the venom is well known. 
It has been recently demonstrated that lysolecithin, which is assumed 
to be a hemolytic factor, releases the esterase within the red blood cells 
(8). It is not impossible that red blood cells may be one source, if not 
the main source, of the effect exerted by the venom. No suggestion can 
be made as to the presence of the esterase in Helix blood. The enzyme 
has a few features which distinguish it from the ACh-esterase, but more 
information is necessary before a conclusion is permissible. 

The properties of the esterase preparation obtained from the dart 
sac of Helix , which offered some difficulty as to classification (8), have 
found* a satisfactory explanation from the evidence that the prepara¬ 
tion contains two different types of esterases. 

Acknowledgments 

I wish to express my thanks to Dr. David Naehmansohn for his stimulating interest 
and his valuable suggestions. I want to thank Mrs. M. August insson for her unfailing 
assistance in the experimental work. 


Summary 

Whereas choline ester-splitting enzymes of the type present in serum 
give the usual dissociation curve, the esterases from all conductive 
tissue tested and from erythrocytes show a rather sharp optimum 
concentration of acetylcholine. The optimum substrate concentration is 
for some choline esters almost the same, but different for others. In 
contrast, triacetin is split at a low rate in low, and at a high rate in 
high concentrations. 

The experiments illustrate the importance of substrate concentration 
when the enzyme activity is measured toward different substrates. 
They support the assumption that the type of esterase present in con¬ 
ductive tissue and erythrocytes, and possibly in a few special cases, has 
well-defined properties distinctly different from other choline ester¬ 
splitting enzymes. 

References 

1. Nachmansohn, D., Bull. Johns Hopkins Hosp. 83, 463 (1948). 

2. Stkdman, E., Stedman, E., and Easson, L. H., Biochem. J. 26, 2056 (1932). 

3. Simonart, A., Rev. beige sci. med. 5, 73 (1933). 



126 


KLAS-BERTIL AUGUSTINSSON 


4. Click, D., J. Biol Chem . 125, 729 (1938); 130, 527 (1939); 137, 357 (1941). 

5 . Nachmansohn, D., and Rothenberg, M. A., ibid. 158, 653 (1945). 

6 . Alles, S. A., and Hawes, R. C., ibid. 133, 375 (1940). 

7. Zeller, E. A., and Bissegger, A., Helv. Chim. Acta 26, 1619 (1943). 

8 . Augustinsson, K.-B., Acta physiol Scand. 15, Suppl. 52 (1948); Nature 162, 194 

(1948). 

9. Augustinsson, K.-B., Nature 157, 587 (1946); Biochem. J. 40, 343 (1946). 

10. Bodansky, O., Ann . N. Y. Acad. Sci . 47, 521 (1946). 

11 . Potter, V. R., and Elvehjem, C. A., J. Biol Chem. 114, 495 (1936). 

12 . Zeller, E. A., Advances in Enzymol 8, 459 (1948). 

13. Haldane, J. B. S., Enzymes, IiOndon, 1930. 

14. Richards, A. G., Jr., and Cutkomp, L. K., J. Cellular Comp. Phy&iol. 26, 57 

(1945). 

15. Tobias, J. M., Kollros, J. J., and Savit, J., ibid. 28, 159 (1945). 

16. Mendel, B., Mundell, D. B , and Rudney, H., Biochem. J. 37, 473 (1943). 

17. Augustinsson, K.-B., and Nachmansohn, D., Science in press. 

18. Cohn, E. J., J. Hematol. 3, 471 (1948). 

19. Bullock, T. H., Grundfest, H., Nachmansohn, D., and Rothenberg, M. A., 

J. Neurophysiol. 10, 11 (1947). 

20. Nachmansohn, D., and Feld, E. A., J. Biol. Chem. 171, 715 (1947). 



The Polysaccharide from lies mannane 

Louis E. Wise 

From The Institute of Paper Chemistry , Appleton , Wisconsin 
Received March 18, 1949 

Introduction 

Various species of Amorphophallus from Java are known to contain 
appreciable amounts of polysaccharidic material. The polysaccharides 
may be isolated by water extraction of the dried roots. In 1939, a num¬ 
ber of such roots (representing various species) were examined by de 
Groot, van Hulssen and Koolhaas (1), and all were found to contain 
mannose-yielding carbohydrates in appreciable amounts. Among them 
was the so-called lies mannane (from A. oncophyllus ), the polysac¬ 
charide of which gave rise to over 50% mannose on hydrolysis. The 
aqueous solutions of this polysaccharide were extremely viscous, and 
the carbohydrate could be precipitated from such solutions by means 
of alcohol. Experiments at The Institute of Paper Chemistry have 
shown that the lies mannane polysaccharide is an excellent beater 
additive when used in paper manufacture. 

Analytical data, given in detail in the experimental part, show that 
the carbohydrate consists mainly of glucose and mannose units, with 
subordinate amounts of pentosans and uronic acid units. Tests for 
fructose, galactose, and galacturonic acid groups were negative. Thus, 
the polysaccharide of lies mannane may be a mannoglucan (but the 
possibility of a mixture of mannan and glucan is not excluded). 

Experimental 

An orienting experiment was carried out with a small samp le of ground powder of 
lies mannane (also termed lies ties) obtained through the courtesy of the Trade 
Commissioner of the Netherlands Indies. The powder was sifted into boiling water and 
the mixture stirred for 10-15 min., cooled, and centrifuged. The nearly colorless liquid 
was decanted, and the aqueous solution precipitated by admixture with several 
volumes of 95% ethanol. The resulting precipitate was obtained both in a gelatinous 
fibrous form and in a flocculent form. On extensive centrifuging of the aqueous-alco- 

127 



128 


LOUIS E. WISE 


holic suspension, the “fibrous” form (which predominated) remained in suspension, 
whereas the flocculent type formed a compact deposit on the bottom of the centrifuge 
bottle. The fibrous polysaccharide was best removed by filtration on mercerized 
broadcloth and could then be dehydrated by suspending successively in alcohol, 
acetone, and ether, using the broadcloth each time as a filtering medium and squeezing 
out the solvent. The smaller floes of polysaccharide could be treated with solvents 
directly in the centrifuge. In either case, the product was dried at room temperature. 

The mixture of fibrous and flocculent material was used for the preliminary hydrol¬ 
yses. Later, larger samples of purified (fibrous) iles mucilage were isolated by a similar 
procedure, the principal variant being omission of the acetone trituration— i. e., only 
ethanol and ether were used in the dehydration. In one experiment, 37 g. of air-dried 
powdered Iles mannane (34.8 g. oven dried) were treated with 4.5 1. of water at 90- 
95°C. and gave about 14.5 g. of (total) polysaccharide. This must be taken as a 
minimal figure, inasmuch as complete extraction is very difficult. 

The mucilage in aqueous solution evidently gives the same type of borax-gel test 
as that given by the mannogalactans— e. g. y locust bean gum (2). Physically, it also 
resembles the latter. However, it is chemically very different. The cold suspension of 
iles mucilage in water gives a deep blue coloration with iodine solution. This is not 
given by the mannogalactans. 

Hydrolysis of the air-dried mucilage with 1% H 2 S0 4 at the boiling point of the mix¬ 
ture showed (from a study of the hydrolysis-time curve) that the hydrolysis was vir¬ 
tually complete in 12.5 hr. Thus, 150 mg. of air-dried mucilage (139.7 mg. oven dry) 
yielded (by the Munson-Walker method) 141 mg. of reducing sugars (calculated as 
glucose) after 12.5 hr. During the hydrolysis, a small amount of dark residue also 
formed, but this was not studied further. 

The filtered neutralized hydrolyzate contained mannose (identified as the phenyl- 
hydrazone, m.p. 194.5-195.5°C.) and the filtrate from a quantitative mannose deter¬ 
mination, on heating, yielded a voluminous precipitate qualitatively identified as 
* phenylglueosazone, m.p. 207-208°C. 

The presence of mannose and glucose in the hydrolyzate of Iles mannane was con¬ 
firmed by the following procedure. The hydrolyzate was neutralized with BaCOs. 
The filtrate was evaporated to a small volume, treated with alcohol to precipitate 
barium salts, and filtered. This second filtrate was evaporated to a thin sirup, a drop 
of which was then subjected to the paper partition chromatographic separation devised 
by Partridge (5). Typical dark brown glucose and mannose “spots” were obtained 
(identical in their relative positions with those obtained from known pure sugar 
samples). In this procedure, the uronic acids were presumably removed as barium 
salts, and pentosans were present in such small quantities that they failed to register 
on the paper strip. When hydrolyzed for very brief periods with HC1, the iles mucilage 
failed to respond to the Seliwanoff test for n-fructose, whereas, under identical condi¬ 
tions, inulin gave a characteristic, deep red pigment, soluble in amyl alcohol. Evi¬ 
dently Jruetosans are absent from the iles gum. 

Another 500 mg. sample of the iles mucilage was hydrolyzed with 
2% HN0 3 for several hours, and then carried through the mucic acid 
determination for galactose. At the end of 2 days in the refrigerator, the 
HN0 3 solution showed a faint cloudiness but no weighable precipitate 



POLYSACCHARIDE FROM ILES MANNANE 


129 


was obtained. This indicated the absence of galactose and galacturonic 
acid . The absence of galactose was further confirmed by quantitative 
differential fermentations (6) of the neutralized iles hydrolyzate. When 
the hydrolyzate corresponding to 100 mg. of air-dried mucilage was 
fermented by organism N.R.R.L. No. 379, the Munson-Walker reduc¬ 
ing value was 36.2 mg. of Cu 2 0. When fermented with organism 
N.R.R.L. No. 966 in a parallel experiment, the final reducing value was 
35.8 mg. of Cu 2 0. These values are identical (within the experimental 
error), and clearly indicate the absence of galactose. As indicated above, 
the mannose in the H 2 SC >4 hydrolyzate was determined quantitatively 
as the phenylhydrazone. Uronic anhydride was determined by Brown¬ 
ing’s procedure (3), and pentosans were estimated by Krdber’s method 
(4). 

The total Cu 2 0 Munson-Walker reducing value obtained from 100 
mg. of the polysaccharide after hydrolysis corresponded to 101 mg. of 
sugar calculated as glucose. The Cu 2 0 reducing values corresponding to 
the predetermined mannose and uronic anhydride values could be 
read from the Munson-Walker tables of Wise and McOammon (7) and, 
for each of these, a corresponding “glucose” value could be calculated. 
The sum of these calculated values, subtracted from the original 101 
mg., gave a measure of the actual glucose in the hydrolyzate. Admitted¬ 
ly, such a calculation is highly proximate, but it serves to show that a 
substantial portion of the hydrolyzate is D-glucose. 

A proximate analysis of the mucilage (oven dry basis) follows: 

Per cent 


Mannan 41 

Glucan (calculated from Munson-Walker reducing values) 48.6 

Uronic anhydride 3.6 

Pentosans 1.76 

Ash 0.53 


The above values for glucan and pentosans are given with reserva¬ 
tions, but it is apparent that the mucilage contains largely mannose 
and glucose groups. How these exist in the mucilage is problematical. 
The glucose may emanate from a true mannoglucan or, possibly, from 
a mixture of mannan and glucan. 

Acknowledgments 

Thanks are due to Miss Ruth C. Rittenhouse, who made the chromatographic 
separations. 



130 


LOUIS E. WISE 


References 

1. Groot, J. E. de, Hulssen, C. J. van, and Koolhaas, D. R., Chem. Weekblad 36, 

69 (1939). 

2. Rowland, B. W., Paper Trade J. 121, No. 25, 38 (Dec. 20, 1945). 

3. Browning, B. L., Tappi 32, 119 (1949). 

4. Krober, E., and Rimbach, C., Z. angew. Chem. 16, 508 (1902). 

5. Partridge, S. M., Biochem. J. 42, 238, 251 (1948). 

6. Wise, L. E., and Appling, J. W., Ind. Eng. Chem ., Anal. Ed. 16, 28 (1944). 

7. Wise, L. E., and McCammon, D. C., J. Assoc. OJJic. Agr. Chemists 28, 167 (1945). 



The Distribution in Rat Tissues of the Methylene 
Carbon Atom of Glycine Labeled with C 141 

Kurt I. Altman, George W. Casarett, T. R. Noonan and K. Salomon 

From the Department of Radiation Biology, University of Rochester 
School of Medicine and Dentistry, Rochester 7, N. Y. 

Received January 10, 194!); revised May 20, 1949 

Introduction 

In the course of studies concerning the role of glycine as a precursor 
of hemin (1), data on the distribution of the labeled a-carbon atom of 
glycine have been gathered. These data are presented in this paper. 
The metabolism of glycine labeled with N 16 in the amino group, or 
with C 13 and C 14 in the carboxyl group, has been studied previously by 
several groups of investigators (2,3,4,5,6). Since, on the basis of these 
and other studies (7), it seems reasonable to suspect differences in 
metabolic behavior between, on the one hand, the amino nitrogen and 
the carbon skeleton of glycine, and, on the other hand, between the 
methylene and the carboxyl carbon atom, our findings on a variety of 
tissues, and on urine, feces, and expired C0 2 , may be of interest. No 
attempt has been made to account for all of the C l4 -activity adminis¬ 
tered. 

Methods 

Glycine was administered to young adult male rats (ranging from 200 to 275 g. 
body weight) by one of three routes: (1) intravenously (4 rats), (2) intraperitoneally 
(1 rat), and (3) by stomach tube (2 rats). All rats received 1 yc. of the C'Mabeled 
amino acid, with the exception of rats No. 13 and No. 21 (<•/. Table I) which received 
2 yc. The rats were allowed free access to a diet of purina fox chow and water until 
the time of glycine administration, after which all food was withdrawn. All but one of 
the animals were sacrificed and processed 24 hr. after the administration of glycine, 
as described previously (1). Rat No. 55 was killed 5 hr. after the injection of glycine. 

In all experiments glycine containing C 14 in its a-carbon atom and having a C' 4 - 

1 This paper is based on work performed under contract with the United States 
Atomic Energy Commission at The University of Rochester Atomic Energy Project, 
Rochester, New York. 


131 




132 


ALTMAN, CASARETT, NOONAN AND SALOMON 


activity of 4.0 X 10 6 disintegrations/min. (= 1.83 /ic.)/mg. was used . 2 All isotope 
measurements were made by the method of Bale and Masters, as briefly described 
elsewhere (1). After administration of glycine all animals were placed in metabolism 
cages to permit separate collection of urine and feces. The total amount of urine 
excreted after glycine injection was diluted to 10 ml. with water and then aliquots 
were taken for C 14 -analysis, and for the preparation of dixanthyl urea according to 
Fosse ( 8 ). The tissues and the gastrointestinal tract contents were dried by lyoph- 
ilizing, whereas fecal samples were dried in air and then over P 2 O 5 in vacuo. In one 
experiment, expired CO 2 was collected in NaOII traps and aliquots were then taken 
for C 14 -analysis, CO 2 being liberated by acid addition. 


Results 


Results pertaining to tissue concentrations of C 14 are presented in 
Table 1, where C ,4 -activity is expressed in terms of “corrected” C 14 - 
activity 3 in order to permit comparison of specific activities. The total 
amount of C 14 incorporated into the tissues is given as percentage of the 
total C 14 -activity administered, based on the total dry weight of the 
tissue as determined experimentally. 

As might be expected from the magnitude of the contribution of 
muscle to the total body weight, the general mass of dissectablc skele¬ 
tal muscle contains the highest percentage of the total administered 
activity, although the C 14 -concentration 4 of this tissue is low. The 
C 14 -concentration of diaphragm is higher than that of other skeletal 
muscle (except in the case of rat No. 55). This observation is probably 
attributable to the higher physiological activity of diaphragm as com¬ 
pared to other skeletal muscle. Similarly, the heart, another muscle of 
high physiological activity, also shows a higher C ,4 -concentration than 
the large mass of skeletal muscle analyzed. The highest (^-concentra¬ 
tions were found in the gastrointestinal tract, liver, kidney, and lungs. 
Brain was found to have the lowest (^-concentration of all tissues 
examined. Since preliminary fractionations of brain (rats No. 55, A32, 
and 28) have indicated that only approximately 30% of the total 
C 14 -activity of the brain resides in the protein fraction (freed from 


2 We are indebted to Dr. B. M. Tolbert of the Radiation Laboratory of the Univer¬ 
sity of California for making available this preparation of pure glyeine synthesized by 
Dr. R. Ostwald. 


• “Corrected” ^activity: 10 - diHintegrat io ns/min./g. . dry weigh ty tjague 

disintegrations/min./kg. body weight 
1 That is, disintegrations/min./g. dry weight of tissue. 



Stomach tube j Stomach tube Intraperitoneal Intravenous Intravenous Intravenous Intravenous 


METHYLENE CAItBON ATOM 


CM 

»H 

e 

8 

CM 

b» 

co 

CM 

x 

lO 

8 8 8 

X 

o 

*-< CM 
X X 





d 

o 


d 

O 

o 

rH 

oo 

co 

CM 

o 

rH O 






rH 

CM 














CM 

CM 


CO 

iO 

h- 

iO 

oo 

3 

X 


rH 10 

iO 




O 

b» 



CO 

O 

CO 

CM 

05 

»o 

CM i0 

CO 




rH 

rH 


CM 

CM 

»0 

00 


CO 

o 

rH 

o oo 

05 













rH 


rH 






o 

o 

CM 

CM 

»o 

00 

CO 

b* 





10 

X 

iQ 

© 



rf< 

CM 



co 

c 





CM 

CM 

OO 

o 

CO 

N 

O 

o 

d 

rH 

iO 

05 





CM 

CM 

id 



CM 














00 

CO 


CO 

CM 


co 

© 

CM 




o 




CO 

o 


rH 

05 

00 

rH 

o 

iO 




X 




d 

CM 


CM 

CM 

«0 

cd 

10 

00 




** 




T*< 

00 

iO 

CO 

00 

N 

rH 

CM 

1^ 

CD 




CO 


** 

1-H 

00 

Tf 

CO 


CM 

oo 

CM 

CM 

X 




X 

CO 

iO 

o 

00 

o 

o 

o 

o 

o 

l> 


o 




o 

o 

o 



CO 















CM 



■** 

• 0 

rH 

CO 

© 

05 



8 





iO 


CO 

"t 

o 

i—i 

-f 

© 

CM 







CM 


CM 

00 

-t 

d 

05 

05 

X 



cd 




05 

CO 

iO 

© 

CO 

O0 

iO 

l> 

CM 

CM 

CM 



co 

rH 

cO 

o 

00 

00 

00 

CM 


00 

00 

CM 

X 





05 

X 

o 

cd 

o 

O 

o 

o 

rH 

id 

CO 

o 

o 



rH 

O 

b- 


** 

co 














co 

cO 


CO 


CM 


-f 


X 

cO 


o 




05 

00 


CD 

-f 

05 

rH 

00 

X 

CO 

O 


X 




o 

CM 


CM 

CO 

*0 

b- 

r- 


IO 

od 






CM 

00 

© 

CM 

»o 

CM 

CO 



X 

‘0 




rH 

Q 


'•t 

O 

co 



00 

CM 

iO 






05 

25 

o 

r-i 

>o 

o 

o 

rH 

fH 

cd 

00 

o 

o 




O 

o 



CM 














05 

»0 


iO 

oo 


lO 


X 

b- 

o 


Q 




05 

05 


04 

00 

i—* 

CO 

o 

X 

co 



25 




o 

fH 


Tt? 

CM 

id 

b- 

cd 

05 

00 



id 





fH 


CM 

** 

t> 

00 

00 







X 

iO 

rH 

05 


CM 

rH 

CM 

iO 

rH 









o 



o 

o 

o 

O 

CM 







CM 

cd 

o 



o 

CO 


05 

rH 










o 


00 

co 


10 










o 

rH 


fH 



CM 

CM 









iO 

05 


o 

o 

CO 

10 

iO 

© 





CM 

X 

10 

o 

00 



CM 

CM 

00 

»o 

X 





X 

iO 

05 

o 

00 


rH 

o 

o 

rH 

rH 

cd 





o 

cd 

05 








rH 









CO 

00 


00 



05 

CM 

co 









*o 


05 


o so 


d 








o 

rH 


CO 

CM 

CO 

05 

CM 
















rH 






































h 



























-C 





£ 











H> 





-H 





a 






CJ 





0 





T3 






c3 



u 


o 





H> 






a 



aS 


G 




-o 

fl 

rS 

0 

C J 






13 

a 



0 

o 

3 

*0 

S 

13 

.S 




0 

o 

13 

o 






1 


S 

s 

'■P 

8 



Brain 

Muscle— 
Muscle— 

Testes 

Heart 

Spleen 

Kidney 

Liver 

Gastroint 

Pancreas 

Diaphrag 

Duodenal 
Lungs 
Bone mai 
Gastroint 

Feces 

Urine 


• a -m o 
G £ . 

^ £ ** S 

»o 


w • — 

£ w +-> +* 

n "I? C w 

s 

13 5 2 

5 j^'p 

w a -4 g 

TJ S S-g 

§ 3 ia 


£ tJ *> • 2 v.o > ^ 

| g3£i | Ill's 
O^.s E f 

Oqo^ £T3 § g sS v 

£ 2tj- 3 d-c^ 


§ 8 .S J i 2 

IrfMPfi 

IPStiillN 

| iH^fcS^i 

*“ l CM 0 w 0 TJ » » $ 

-a SSI 




134 


ALTMAN, CASARETT, NOONAN AND SALOMON 


nucleic acid)®, it is at present difficult to state whether the slow rate of 
turnover of brain protein is the cause of its low C ,4 -concentration. 
In two experiments (rats No. 28 and 55) the liver was fractionated and 
approximately 85-90% of the total activity was found to be present 
in the protein fraction (from which nucleic acids had been removed). 

The elimination of the a-carbon atom of glycine as C0 2 within a 
period of 5 hr. after intravenous injection (rat No. 55) accounts for 
11.8% of the total activity administered. It appears that the methylene 
carbon atom of glycine differs in this respect from the carboxyl carbon 
atom, since the rate of elimination of the carboxyl carbon atom labeled 
with C 14 (in mice) is very rapid, approximating that of the carboxyl 
group of acetate (6). The data presented below also indicate that the 
largest percentage of the C l4 -activity expired as C0 2 is eliminated in the 
first hour. The C0 2 collections were made within the 5-hr. experi¬ 
mental period and analyzed for C 14 with the following results: 

Total C u -activity administered 
per cent 

First hour 4.91 

Second and third hour 5.71 

Fourth and fifth hour 1.20 

The urinary and fecal excretion of C 14 -activity are also shown in 
Table I. As can be seen, the amount of C 14 -activity excreted through 
these pathways is relatively small, ranging up to 10%. In two cases 
(rats No. 21 and 13) urea was isolated from 1 ml. of diluted urine (as 
dixanthyl urea) and was found to have C' 4 -activities of 1.19 and 0.99 
(10 4 disintegrations/min./mA/ urea), i.e., 1.7% and 2.3% of the total 
urinary activity or 0.15% and 0.17% of the total C 14 -activity adminis¬ 
tered. 

Comments 

As can be seen from Table I, animals which were given labeled 
glycine by intravenous and intraperitoneal routes show reasonably 
good agreement (allowing for a 10% experimental error) with respect 
to tissue isotope concentration, while animals given glycine by stomach 
tube exhibited wider variations. Since the animals fed by stomach tube 
were not fasted prior to the administration of the amino acid, such 

* These fractionations of brain also revealed the presence of approximately 60-65% 
of the total C l4 -activity of the brain in the lipide fraction. The lipides of the gastro¬ 
intestinal tract were also found to contain significant (Inactivity in several cases. 



METHYLENE CARBON ATOM 


135 


variations may be due to differences in the rate of amino acid absorp¬ 
tion. It is also of interest to note that there appears to be relatively 
little difference between the 24-hr. rats and the 5-hr. animal. The only 
marked difference appears to be in the case of muscle, whose C 14 - 
concentration is lower in the 5-hr. rat, and in the case of the pancreas 
which shows a considerably higher C 14 -concentration in this animal. 

The data indicate that C 14 -incorporation into the gastrointestinal 
tract in animals which were given labeled glycine by intravenous and 
intraperitoneal routes was slightly greater than into the liver, whereas 
incorporation into the kidney was of approximately the same order as 
liver in this group of animals. Skeletal muscle, in general, seems to 
contain about 1/3 of the activity of liver, whereas the ratio of dia¬ 
phragm or heart to liver C 14 -activity is somewhat higher. Because of the 
differences in route of administration and labeling, the relationships 
between the C 14 -activity of the liver, internal organs and muscle differ 
from those found by other workers (3). 

Sprinson (10) reported data on the N 15 -excretion in man fed N 15 - 
labeled glycine showing that 12% of total N 15 -activity was excreted 
within 5 hr., and 25-30% within 24 hr. These values are somewhat 
higher than the values reported here. Greenberg and Winnick's (5) 
value of 3.7% total activity excreted in urine is not comparable to our 
values since the collection period is not identical. 

Although quantitative data presently available are limited, it seems 
likely that a large portion of the C u -activity of the tissues is actually 
a measure of protein metabolism, except in the case of brain and gastro¬ 
intestinal tract, which has been discussed before. The differences in 
the anabolic and catabolic activity, particularly as concerns proteins, 
are reflected in the wide variations in C 14 -act.ivity in the individual 
tissues. 

Acknowledgments 

The authors gratefully acknowledge the technical assistance of Miss Mildred K. 
Taylor and Miss Ann Marie Noonan in carrying out the C 14 determinations. 

Summary 

1. Data on the distribution in a variety of rat tissues of the methylene 
carbon atom of glycine labeled with C 14 have been presented, in terms 
of corrected C 14 -activity and percent of total administered C 14 -activity. 

2. Urinary and fecal excretion of C 14 subsequent to administration 



136 


ALTMAN, CASARETT, NOONAN AND SALOMON 


of labeled glycine, as well as C 14 -activity of the gastrointestinal con¬ 
tents have been reported. In two cases, data on C 14 -activity of urea 
have been presented. 

3. The C l4 -activity of C0 2 expired over a period of 5 hr. was meas¬ 
ured and found to constitute ca. 12% of the total C 14 -activity adminis¬ 
tered. 


References 

1. Altman, K. I., Casarett, G. W., Masters, R. E., Noonan, T. R., and Salomon, 

K., J. Biol. Chem. 176, 319 (1948). 

2. Ratner, S., Rjttenberg, D., Keston, A. S., and Sciioenheimer, R., ibid. 134, 

665 (1940). 

3. Shemin, D., and Rittenberg, D., ibid. 153, 401 (1944). 

4. Olsen, N. S., Hemingway, A., and Nier, A. O., ibid. 148, 611 (1943). 

5. Greenberg, I). M., and Winnick, T., ibid. 173, 199 (1948). 

6. Heidelberger, C., Personal communication. 

7. Grinstein, M., Kamen, M. D., and Moore, C. V., ibid. 174, 767 (1948). 

8. Fosse, O., Ann. 6, 13 (1916). 

9. Skelton, H., Arch. Internal Med. 40, 140 (1927). 

10. Sprinson, D. B. in The Use of Isotopes in Biology and Medicine, p. 183. 
Wisconsin, 1948. 



Esterified Fatty Acid Levels of Normal Human Sera 

Frederick C. Bauer, Jr. 1 and Edwin F. Hirsch 
With the technical assistance of 
Lynn Carbonaro 

From the Henry Baird Favill Laboratory, St. Luke’s Hospital, Chicago, Illinois, 
and the Department of Pathology, The University of Chicago 
Received March 14, 1949 

Introduction 

Recently the authors described a simple colorimetric procedure for 
determining total esterified serum fatty acids (1). The method is based 
upon the conversion of the fatty acid esters into the corresponding 
hydroxamic acids by hydroxylamine hydrochloride, and their subse¬ 
quent conversion to colored ferric salts. The present report records the 
esterified fatty acid content of fasting sera from 102 apparently normal 
persons, and compares these with values published by various authors 
(Table II). 

Materials and Methods 

Laboratory technicians, nurses and other hospital employees reported to the 
laboratory without breakfast at about 8:00 A.M. Blood was drawn and the serum 
esterified fatty acid content was determined in triplicate or duplicate according to the 
method described (1). Most of the subjects were young women. 

Results 

The results of the analyses are contained in Table I. The maximum 
is 12.6, the minimum 7.0, and the mean is 9.2 mcq. of esterified fatty 
acids/1, of serum. The standard deviation is ± 1.29. 

Comments 

Since free fatty acids in sera are not measured by this procedure, the 
results may not be exactly comparable to values obtained in procedures 

1 Seymour Coman Fellow in Pathology, The University of Chicago. 

137 



138 


FREDERICK C. BAUER, JR. AND EDWIN F. HIRSCH 


measuring total fatty acids. However, the free fatty acid contents of 
human sera have been reported to be small (2), calculated at 3% or 
less of the total fatty acid content (3). 

Plasma or serum total fatty acids have been determined by proce¬ 
dures employing various principles. Bloor (4) and Boyd (5) determined 
total fatty acids and cholesterol on alcohol-ether extracts of plasma by 
saponification, extraction of the acidified residue with petroleum ether, 
and oxidation of an aliquot of the solution with H 2 S 04 -K 2 Cr 2 0 7 
reagent. Cholesterol was determined separately in another aliquot, and 
its oxidation value calculated and subtracted from the oxidation value 
of the mixture of total lipide, giving the value for total fatty acid. 


TABLE I 

Serum Esterified Fatty Acid Content of 102 Normal Fasting 
Individuals Grouped According to Levels 


Ksterified serum fatty 
acid levels meq./l. 

7.0-7.9 

8.0-8.9 

9.0-9.9 

10.0-10.9 

11.0-11.9 

12.0-12.9 


No. of persons 
18 
27 
31 
13 
7 
6 


Stoddard and Drury (G), Man and Gildea (7), and Thannhauser (12), 
extracted the serum or plasma with alcohol and ether, then saponified 
the extract. The soaps were suspended or dissolved in water, and HC1 
was added to cause separation of the free fatty acids. The latter were 
recovered by filtration through a Gooch filter. Finally, the fatty acids 
were redissolved in hot alcohol and titrated with standard alkali. 
Wilson and Hansen (8) saponified alcohol-ether extracts of sera, 
evaporated these to dryness, added water, acidified, and extracted 
with petroleum ether. The fatty acids were removed from the petrol¬ 
eum ether with alcoholic KOH solution and water, and were recovered 
by evaporating this to dryness, dissolving in water, precipitating with 
acid and extracting with petroleum ether. The fatty acids were then 
weighed and titrated. 

Table II gives the serum fatty acid values obtained by the various 
authors. 

When the minimum values of total serum fatty acid levels in Table 
II are compared, all are at approximately the same level (6.9-8.6 



ESTERIFIED FA'i’TY ACID LEVELS OF NORMAL HUMAN SERA 139 


meq./l.). The range of fatty acid levels in many of the different series 
is large, with the exception of our series, wherein the range is remark¬ 
ably small. The reason for the small range observed by this method of 
analysis, in contrast to the broad range in other series, is not clear. It 
seems probable that the specificity of the colorimetric reaction involved 
may be partly responsible, for only fatty acids present in ester form 
react. Other organic compounds, such as urea and amino acids, do not 
produce & color under the conditions of analysis. 


TABLE II 

Normal Serum or Plasma Fatty Acid Levels Obtained by Various Authors 


Authors 

No. of sera 
analyzed 

Meq. of fatty aeids/1. 

Maxi muni 

Minimum 

Mean 

S. D. 

Peters and Man (9) 

355 

36.9 

7.3 

12.3 

±3.37 

Boyd (5) h 

8 

16.1 

8.6 

12.5 

±1.77“ 

Bullen and Bloor (10)'’ 

12 

19.9 

8.3 

11.8 

±3.44“ 

Wilson and Hansen (8) d 

9 

18.2 

7.5 

13.0 

±2.92“ 

Thannhauser (12) r 

— 

16.2 

7.2 

— 

— 

Stoddard and Drury (6) f 

10 

12.0 

6.9 

10.6 

— 


“ S. D. was calculated from the authors’ data after converting their values to 
meq./l., where necessary. The following formula was used: 


Standard deviation = J Sum of squares of deviations from mean 
\ Total number of observations 


6 Meq. were calcualted from fatty acid weights (mg.-%) by the following formula: 
— = meq. fatty acid/1. (11). 


'In the method of Stoddard and Drury (6) the average fatty arid molecular 
weight, is assumed to be 277.2 when fatty acid weight is calculated from their titration 
values. Therefore, the following formula is used to convert data obtained by their 
method back to meq./l.: 

—= meq. fatty aeid/1. 
d Data from scrum of individuals with illnesses is excluded. 


Other factors, in at least one study (10), may be time when blood was 
drawn for analysis, i.e., about 3 hr. after a meal; whereas in our series, 
blood was drawn after an all-night fast (about 12 hr.) and after the 
subjects had come for work to the hospital. The effect of slight muscu¬ 
lar activity on serum lipides is not completely understood (11). 




140 


FREDERICK C. BAUER, JR. AND EDWIN F. HIRSCH 


Summary 

Serum esterified fatty acids were determined on 102 healthy fasting 
individuals. The maximum was 12.6, the minimum 7.0 and the mean 
9.2 meq./l. A brief comparison is made between the fatty acid levels 
found in this study and similar data obtained by other authors. 

References 

1. Bauer, F. C., Jr., and Hirsch, K. F., Arch. Biochem. 20, 242 (1949). 

2. Bloor, W. R., Biochemistry of the Fatty Acids, pp. 116 and 117. Reinhold Pub. 

Corp., N. Y., 1943. 

3. Davis, B. D., Arch. Biochem. 15, 351 (1947). 

4. Bloor, W. R., J. Biol. Chem. 77, 53 (1928). 

5. Boyd, E. M., ibid. 101, 323 (1933). 

6. Stoddard, J. L., and Drury, P. E., ibid. 84, 741 (1929). 

7. Man, E. B., and Gildea, E. F., ibid. 99, 43 (1932 33). 

8. Wilson, W. R., and Hansen, A. E., ibid. 112, 457 (1936). 

9. Peters, J. P., and Man, E. B., J. Clin. Invest. 22, 707 (1943). 

10. Bullen, S. S., and Bloor, W. R., J. Allergy 8, 155 (1936-37). 

11. Peters, J. P., and Van Slyke, D. D., Quantitative Clinical Chemistry, Inter¬ 

pretations, Vol. 1, pp. 468, 481,495: The Williams and Wilkins Co., Baltimore, 
1946. 

12. Thannhauser, S. J., New Engl. J. Med. 237, 515, 546 (1947). 



Factors Influencing Oxygen Production by 
Illuminated Chloroplast Fragments 

Daniel I. Amon and F. R. Whatley 

From the Division of Plant Nutrition, University of California, Berkeley, Calif. 

Received April 6, 1949 

Introduction 

One of the serious obstacles in the experimental study of the mech¬ 
anism of photosynthesis has been the impossibility of separating the 
process from the activities of intact green cells. In recent years, how¬ 
ever, an important advance was made by R. Hill (1,2), who demon¬ 
strated that the long-known capacity of isolated chloroplasts to evolve 
oxygen can be greatly enhanced by the use of suitable oxidants. 

Although the entire process of photosynthesis was not reconstructed 
in the chloroplasts, in the sense that in no case was C0 2 reduction linked 
with oxygen evolution (3), yet the work of Ilill made it possible to in¬ 
vestigate outside the living cell the reaction most characteristic of 
photosynthesis in green plants: the photolysis of water resulting in the 
evolution of gaseous oxygen. The recent investigations in this field 
have been reviewed by Holt and French (4), who marshaled the evi¬ 
dence in favor of the identity of the oxygen-liberating mechanism in 
isolated chloroplasts with that in the intact green cells. 

Despite the substantial measure of agreement among the investiga¬ 
tors of isolated chloroplasts, there remained several areas of conflicting 
observations, some of them of considerable theoretical importance. 
The present communication deals with a part of our investigation of 
the reactions of isolated chloroplasts, in which a special attempt was 
made to examine the points of discordance. 

Methods 

Chloroplast fragments, rather than whole plastids, were used in all experiments. 
The preference for fragments was in accord with the general objective of using the 
smallest subcellular, chlorophyll-bearing aggregate still capable of evolving oxygen 
upon illumination. The chloroplast fragments were prepared from large leaves of chard 

141 



142 


DANIEL I. ARNON AND F. R. WHATLEY 


(spinach beet), Beta vulgaris , grown in a greenhouse in a nutrient solution of the 
composition previously used in sand culture (5). The procedure for preparing chloro- 
plast fragments was the same as previously described (5), except that operations were 
carried out in the cold and the leaves were shredded with a stainless steel knife, prior 
to grinding for 2 min. in a Waring Blendor in the presence of M /15 potassium phos¬ 
phate buffer of pH 6.5. 

Chlorophyll was determined (5) for each batch of chloroplast fragments and an 
aliquot selected to give approximately 0.5 mg. of chlorophyll (a -f b) in each manom¬ 
eter vessel. With quinone and ferricyanide, the quantity of oxidant used was also 
standardized to yield 84 mm.* of 0 2 for each vessel. Standard reagents without further 
attempt of purification were used. Solutions of three oxidants were prepared, just prior 
to use. Forty mg. of quinone were dissolved in 10 ml. 0.01 N H2SO4. The use of acid as 
a solvent was based on the recommendation of Warburg and Luttgens (6). With the 
quinone at our disposal, water solutions were equally satisfactory if used without 
delay. No difference was observed between samples of quinone from two different 
sources: Eastman-Kodak and Hopkins and Williams (England). Unless otherwise 
indicated, 0.2 ml. of the acid quinone solution containing 7.5 X 10~ 6 moles was added 
to each vessel. The ferricyanide was prepared by dissolving 2.0 g. of K 3 Fe(CN) 8 in 
100 ml. il//15 phosphate buffer, pH 6.5. Twenty-five hundredths ml. of the solution, 
corresponding to 15 X 10 -6 moles, was added to each vessel. Of the phenol indophenol 
(approximately 41% pure as determined by ascorbic acid titration) 30 mg. were 
dissolved in 5 ml. phosphate buffer, pH 6.5, and 0.5 ml., corresponding to approxi¬ 
mately 5.1 X 10~ 8 moles, were added to each vessel. 

Oxygen evolution was measured manometrically at 15°C. in a refrigerated constant 
temperature bath of a design similar to that developed by S. Aronoff, equipped with a 
plate glass strip at the bottom. Illumination was provided by a battery of 150 watt 
Mazda projector flood lamps placed under the glass strip. The intensity of illumina¬ 
tion at flask level was approximately 28,000 lux. At this relatively high level of illu¬ 
mination the rate of oxygen evolution was essentially independent of light intensity. 
All readings were made in an atmosphere of nitrogen. No attempt was made to purify 
the nitrogen gas from traces of oxygen. The total volume of the reaction mixture was 
3 ml., and all measurements were made in conical vessels of approximately 15 ml. 
capacity. 

Yields and Rates 

The overall equation of photosynthesis, 

(i) C0 2 + 2 H s 5 -► |CH s O | + H 2 0 + 0 2) 

indicates that the release of each oxygen atom from the splitting of 
water is accompanied by the release of two hydrogen atoms (7). If the 
oxygen-liberating mechanism in photolysis is the same as in photo¬ 
synthesis, 1 an identical relation between hydrogen and oxygen must 

1 In discussing the evolution of oxygen from the splitting of water by illuminated 
chloroplasts, it is convenient to adopt a short-hand expression analogous in usage to 
such terms as photosynthesis, photooxidation and photoreduction. The term photol¬ 
ysis has been adopted for this purpose. 



CHLOROPLAST OXYGEN EVOLUTION 


143 


obtain. It becomes important, therefore, to consider whether stoichi¬ 
ometric yields of oxygen are obtained in photolysis reactions with 
several oxidants in accordance with the following equations for quin- 
one, ferricyanide and phenol indophenol respectively: 


0 OH 



(iii) 4 K,Fe (CN), +4K+ + 2II 2 0-> 4 K,Fe (CN), + 4H + + 0 2) 

(iv) 2 V-OH + 2H 2 0-> 

2 HO—V~NH -OOH + O, 

Warburg and Liittgens (6), who discovered the quinone reaction, 
reported yields between 80 and 90% of theoretical. On the other hand, 
Aronoff (8) was unable to obtain with quinone oxygen yields higher 
than 35% of those expected from Eq. (ii). Aronoff (8) attributed his 
low yields to competing reactions resulting from the deterioration of 
quinone. Warburg and Liittgens (6) explained their slight discrepancy 
from theoretical yields by the utilization of oxygen by chloroplast 
fragments, concomitant with the much greater photolytic oxygen 
evolution. 

Warburg and Liittgens (6) obtained their high yields only upon the 
addition to the chloroplast fragments (“granula”) of an accessory 
factor contained in the supernatant fluid remaining after centrifuging 
off the suspension containing the chlorophyll. They identified this 
accessory factor as chloride and designated this anion as the “co¬ 
enzyme” of photolysis. 

With quinone as the oxidant, we obtained yields of oxygen at 15°C. 
(Table I) in accord with theory (Eq. ii). The marked effect of chloride 
reported by Warburg and Liittgens (6) was confirmed in this and 
numerous other experiments. The absence of this ion may account 
for the low yields obtained by Aronoff (8). Striking as is the effect of 
chloride on oxygen evolution, other evidence at hand does not support 
the conclusion of Warburg and Liittgens (6) that chloride is a coenzyme 
for oxygen evolution. A full discussion of this subject is reserved for 
another paper. For the purpose of the current investigation, the addi¬ 
tion of chloride to the reaction mixture was adopted as a standard 



144 


DANIEL I. ARNON AND F. R. WHATLEY 


TABLE I 

Oxygen Evolution by Chloroplast Fragments 


Oxidant 

Yield (theory) 
per cent 

Hate (QfJ 1 )* 

No KC1 

0.01 M KCl 

No KCl 

0.01 M KCl 

Quinone 

36 

102 

290 

1030 

Ferricyanide 

19 

102 

200 

800 

Phenol intlophenol 

90 

107 

400 

730 


° QoJ 1 = mm. 3 of oxygon /hr. /mg. of chlorophyll, computed from data obtained for 
the 6-min. period from 1 min. to 7 min. after turning on the light. 


procedure to insure maximum yields and rates as points of reference 
necessary for evaluating the influence of various factors on photolysis. 

As already noted by Warburg and Liittgens (6), the activating effect of chloride 
was not confined to quinone, but was also apparent when ferricyanide was used as an 
oxidant (Table I). In agreement with these authors, and with Holt and French (11), 
we have also found that K 3 Fe(CN)« alone can act as the oxidant in photolysis without 
the addition of the other reagents in the original reaction mixture as used by Hill and 
Scarisbrick ( ( .)) or in the modified formula used by Holt and French (10). In our 
experiments, supplementing the ferricyanide with K 2 C 2 O 4 , or with ferric iron in the 
presence of oxalate, produced no increase in the yield or rate of oxygen evolution. The 
illumination of reaction mixtures containing appreciable amounts of ferric oxalate was 
found under our conditions to yield large amounts of C0 2 . Thus, in the absence of 
KOH, the addition of ferric oxalate may give apparent- high yields owing to the in¬ 
clusion of evolved C0 2 with the oxygen measured. 

The suitability of phenol indophenol as an oxidant for quantitative mesaurement of 
oxygen evolution by chloroplasts was demonstrated by Holt and French (11). Our 
own results with this dye are given in Table I. As with the other two oxidants, 
stoichiometric yields of oxygen were obtained with this dye as well. There was a 
marked chloride activation on the rate of oxygen evolution, but it was less in magni¬ 
tude than with the other oxidants. It is possible that the impurities in the dye included 
a small amount of chloride, not sufficient, however, to bring about maximum rates. 
Unlike the other oxidants, phenol indophenol gave high yields even in the absence of 
added chloride. It is believed that this is attributable to the relatively high QJ^ 1 
(Table I) obtained without chloride. A discussion of the relation between rates and 
yields in a later section of this paper has a further bearing on this point. 


Effect of the Cytoplasmic Fluid 

The high-speed centrifugation used for separating the chloroplast 
fragments yielded a pale yellowish-green supernatant, which contained 
only a minute amount of chlorophyll. It was designated, as previously 



CHLOROPLAST OXYGEN EVOLUTION 


145 


(5), the cytoplasmic fluid. For example, in a typical preparation the 
suspension of chloroplast fragments contained 1.9 mg., as compared 
with 0.01 mg. of chlorophyll/ml. in the cytoplasmic fluid. As far as 
preparatory technique is concerned, our chloroplast fragments corre¬ 
sponded to the “granula” of Warburg and Liittgens (6) and the “grana” 
of Aronoff (8), wtiereas our cytoplasmic fluid corresponded to the 
“solution” of Aronoff (8). 

Warburg and Liittgens (0) found no evolution of oxygen upon illu¬ 
mination of the cytoplasmic fluid alone. On the other hand, Aronoff (8) 
reported a high rate of oxygen evolution in the illuminated cytoplasmic 
fluid (“solution”), about 10 times as great as that of the chloroplast 
fragments. The chlorophyll concentration in AronofPs cytoplasmic 
fluid was not given, but it was assumed that it was low as in our prep¬ 
arations. A question of fundamental interest was thus raised: is there 
some constituent of the cytoplasmic fluid, other than the previously 
noted chloride, capable of enhancing the rate of photolysis by chloro¬ 
plast fragments? 

To test this possibility, the following experiments were carried out. 
A reaction mixture was prepared with quinone as the oxidant, contain¬ 
ing a small amount of chloroplast fragments computed to give the 
same concentrations of chlorophyll as in the cytoplasmic fluid. The 
results of one experiment, summarized in Table II, disclosed that, 
when minute amounts of chlorophyll were used, the cytoplasmic 
constituents failed to increase the rate of oxygen evolution. The 
actual oxygen evolution in the vessel containing the small amount 
of chlorophyll was very much lower than that in the vessel with the 
high (0.5 mg.) chlorophyll content: 4-0 vs. 45 mm. 3 of oxygen (Table 
II). Upon inserting the respective figures into the formula, Qo * 1 = 


mm. 3 () 2 evolved m 0 min. _ . . . , , . , r . 

. - . .—--X 10, higher rates were obtained for the 

mg. chlorophyll 


TABLE II 


Oxygen Evolution from Cytoplasmic Fluid and from Chloroplast Fragments , 
Containing Varying Amounts of Chlorophyll 



Chlorophyll 

concentration 

mm.* of Oj 
evolved in 6 mins. 

< 

Cytoplasmic fluid 

0.032 

4 

1250 

Chloroplast fragments 

0.032 

6 

1870 

Chloroplast fragments 

0.49 

i 

45 

920 




146 


DANIEL I. ARNON AND F. R. WHATLEY 


two reaction mixtures containing the low chlorophyll concentration 
(Table II). It is doubtful, however, whether the differences are sig¬ 
nificant. An error of 1 or 2 mm.* of oxygen would be multiplied many 
times and have a marked effect on the Qo, 1 for 0.032 mg. of chlorophyll, 
but it would only slightly change the computed rate for 0.5 mg. 
chlorophyll. It is for this reason also that we ascribe no special signifi¬ 
cance to the difference in the rates between the two reaction mixtures 
containing the low concentrations of chlorophyll. 

Stability of the Photolytic System 
In Intact Leaves 

Warburg and Liittgens (6) found no loss in activity of chloroplast 
fragments after storing sugar beet leaves for several days at 5°C. Under 
our conditions, a comparison (with quinone as the oxidant) of the 
activity of chloroplast fragments of freshly picked spinach beet leaves 
with that of leaves kept for 4 days in the dark at 2°C. in a pliofilm bag 
to insure turgidity showed a loss on storage of about 25%: Q^, 1 of 1010 
and 740, respectively. 

The leaves used in the present investigation were generally harvested 
in the morning but, on a number of occasions, were used after dark 
storage overnight without any apparent loss of photolytic activity of 
the chloroplast fragments. Hill and Scarisbrick (13), working with 
Setellaria media, stressed the importance of time of day for collecting 
leaves: highest activity of chloroplasts was obtained in leaves picked in 
the morning, the activity falling nearly to zero in leaves picked during a 
sunny afternoon. Kumm and French (12) observed an increase in 
activity following illumination of leaves previously stored in the dark, 
just prior to the separation of chloroplasts. Under our conditions, no 
such dependence of activity on light or time of day was observed. In 
one experiment, leaves were collected in the greenhouse at about mid¬ 
night and kept in the dark at 2°C. until the chloroplast fragments were 
separated the following afternoon. All manipulations, prior to the turn¬ 
ing on of light for the photolysis reaction, were carried out in dim light. 
The Q^ 1 obtained with quinone as the oxidant was 1010, comparable to 
that of active preparations obtained from leaves harvested at other 
times of the day. 



CHLOROPLAST OXYGEN EVOLUTION 


147 


Chloroplasts and Chloroplast Fragments 

The experiments of Hill and Scarisbrick (13) and Kumm and French 
(12) with intact chloroplasts indicated a rapid loss of photolytic 
activity with storage, with half-lives of the order of 2 hr. French, Anson 
and Holt (14) found, however, much greater stability in chloroplasts 
stored at 0°C.: between 25 and 50% of the original activity was lost in 
12 hr. and, in some preparations, even smaller losses occurred in that 
period. 

There is agreement among different investigators that the photolytic 
mechanism in the chloroplast fragments is relatively more stable. 
Warburg and Liittgens (6) found that chloroplast fragments when 
stored in M /20 phosphate buffer at pH 6.21, containing 0.5% KOI, 
lost only 5-15% of their activity in 24 hr. The half-life of a similar prep¬ 
aration stored at 2-3°C., presumably without KC1, by Aronoff (8) was 
11 hr. Our own preparations of chloroplast fragments stored in phos¬ 
phate buffer at 2°C. were also relatively stable, the loss of activity in 
5 hr. amounting to about 14% (Qo 2 l of 1190 vs. 1020). 

A striking rate of deterioration of photolytic activity was observed, 
however, in chloroplast fragments prepared by another procedure. 
Instead of grinding in a Waring Blendor in phosphate buffer, batches 
of leaves were macerated in the cold in a “Vitajuicer” 2 . This instru¬ 
ment permitted a trituration of leaves without the use of liquid, 
accompanied by rapid separation of the green leaf juice undiluted by 
an extraneous solvent. The dark green fluid was then handled in the 
same manner as the previously described filtered Blendor mash (5): 
i.e.j centrifuged at low speed for 1 min., the residue discarded and 
the low-speed centrifugate (l.s.c.) centrifuged for 20 min. at high speed 
to separate the chloroplast fragments from the cytoplasmic fluid. 

The stability of the photolytic system in the chloroplast fragments 
divested of the cytoplasmic fluid was as in the Waring Blendor prepa¬ 
ration, but a rapid deterioration was noted in the activity of the l.s.c., 
that is, when the chloroplast fragments were stored (at 2°C.) in com¬ 
bination with the cytoplasmic fluid: 


Hours after 
grinding leaves 

0.75 

2.00 


4.3 

7.75 

Qo , 1 

1600 

1200 


740 

490 


distributed by Enterprise Development Corp., 231 W. Olive Ave., Burbank, 
Calif. 



148 


DANIEL I. ARNON AND F. R. WHATLEY 


The very high Qo, 1 = 1600, obtained within 45 min. of grinding the 
leaves, and the rapid falling off of activity thereafter, gives a measure 
of the rate of inactivation occasioned by the presence of the cyto¬ 
plasmic fluid. It also suggests that an unavoidably large degree of 
deterioration takes place during the relatively long period of high¬ 
speed centrifugation when the cytoplasmic fluid and the chloroplast 
fragments are combined. 

Inactivation by Heating 

Isolated chloroplast fragments when held for 5 min. at 55°C. failed 
completely to evolve oxygen even with the addition of KC1 (using qui- 
none as the oxidant). The sensitivity of the photolytic system to heat 
was previously shown by Warburg and Luttgens (6), who found com¬ 
plete inactivation by heating for 10 min. at 50°C. Holt and French (10) 
have also shown that maintaining chloroplasts at 35°C. for 15 min. 
resulted in the loss of 80% of the photolytic activity (with ferricyanidc 
as oxidant), while 5 min. at 35°C. caused a loss of 37%. 

The thermolability of the photolytic system was found to be in dis¬ 
tinct contrast with the thermostability of the polyphenoloxidase known 
to occur in the chloroplast fragments (5). The polyphenoloxidase 
activity remained virtually unchanged after heating the material at 
75°C. for 5 min. The oxidase activity was lost after heating at 100°C. for 
•3 min. 

Sensitivity to Light 

Warburg and Luttgens (6) reported that illumination of suspension 
of chloroplast fragments in the absence of an oxidant rapidly destroyed 
the photolytic activity when subsequently the oxidant was added. 
Since it was not clear whether this inactivation was peculiar to the 
quinone which these authors used, nor whether the pre-illumination 
included the oxidant as well as the chloroplasts, we undertook to re¬ 
examine this point with the use of 3 oxidants: quinone, ferricyanidc 
and phenol indophenol. 

Chloroplast fragments were placed in the main vessel and the respective oxidants 
in the side-arm. The main vessel and the side-arm were appropriately darkened 
with tinfoil or exposed to light prior to mixing in' accordance with the outline 
given in Table III. The “dark exposure” of either the chlorophyll fragments or the 
oxidant served as a check on the thermal deterioration during the 20 min. of exposure 
to light—an important point in view of the thermolability of the chloroplast material. 



CHLOROPLAST OXYGEN EVOLUTION 


149 


The pretreatment, either in light or dark, was given while the vessels were under¬ 
going shaking in a bath kept at 15°C. In the control treatments aliquots of the same 
suspension of chloroplast fragments were mixed with the respective oxidant at zero 
time. 

The results presented in Table III indicate that the thermolability of 
the shaken chloroplast fragments during a 20 min. period can account 
for serious losses in activity. Moreover, the three oxidants varied in 
their respective photo- and thermosensitivity. Quinone was very un¬ 
stable in light and an exposure of 20 min. rendered it unsuitable as an 
oxidant for the chloroplast fragments. An indication of the instability 
of phenol indophenol was also obtained. The ferricyanide was found to 
be stable either in light or in dark and it was thus most suited for the 

TABLE III 


Effect of Preillumination of Chloroplast Fragments ( chl.f '.) and OxidarUs 
on Oxygen Evolution 

Pretreatment was given in vessels undergoing shaking in a bath kopt at 15°C. 


Experimental conditions 


Qc* 


Quinone 

FeCy 

Phenol 

indophenol 

1. Chl.f. and oxidant mixed at 0 time 

1080 

800 

700 

2. Chl.f. preilluminated, oxidant kept 
in dark. Mixed at 20 min. 

530 

560 

300 

3. Chl.f. kept in dark, oxidant preil¬ 
luminated. Mixed at 20 min. 

140 

600 

260 

4. Chl.f. and oxidant kept in dark. 
Mixed at 20 min. 

610 

560 

300 


evaluation of changes in the chloroplast preparation. When due allow¬ 
ances are made for the thermal inactivation of the chloroplast fragments 
and the peculiarities of the oxidants, the data in Table III offer no 
basis for a conclusion of light inactivation in the absence of oxidant. It 
will be recalled that these experiments were carried out in an atmos¬ 
phere of nitrogen. Data obtained by one of us (Arnon) in another in¬ 
vestigation showed that, in air, the illumination of chloroplast frag¬ 
ments in the absence of an oxidant resulted in a marked photooxidation. 








150 


DANIEL I. ARNON AND F. R. WHATLEY 


Effect of Temperature on Rates and Yields 

The thermal instability of the photolytic system in chloroplasts is 
further illustrated by the data in Table IV. The highest rates of oxygen 
evolution were obtained at the lower temperatures. This is in accord 
with Holt and French (10) who observed highest activities at about 
10°C. It seems probable that the decreased activity at higher tempera¬ 
tures is a reflection of an accelerated destruction of enzymes or some 
unstable intermediates essential to the light reaction. 


TABLE IV 

Effect of Temperature on Rale of Oxygen Evolution with Different Oxidants 



acU 

10°C. 

15°C. 

20°C. 

25°C. 

Quinone 

700 

920 

820 

660 

Ferricyanide 

620 

700 

740 

460 

Phenol indophenol 

520 

730 

720 

620 


An interesting relation between temperatures and total yields of 
oxygen is illustrated in Table V. As previously noted (Table I) the 
omission of KCl was associated with a low rate of oxygen evolution. 
The corollary of a lower rate is an increase in the time required for the 
oxygen evolution to come to completion, i.e., to reach the amount 
required by the stoichiometric relations (100% yield). But this delay 
occasioned by a slower rate is conducive to the deterioration of the 
unstable photolytic system, especially at the higher temperatures. 

TABLE V 

Effect of Temperature on the Yield of Oxygen 


Percentage yields 



No KCl 

KCl Added 


10°C. 

15°C. 

20°C. 

25°C. 

m 

15°C. 

20°C. 

25°C. 

Quinone 


51 

38 

■a 

101 

101 

107 

94 

Ferricyanide 

43 

32 

27 

u 

100 

88 

88 

76 



















CHLOROPLAST OXYGEN EVOLUTION 


151 


Thus, as shown in Table V, the per cent yield falls off precipitously with 
higher temperature in the “minus KC1” series, whereas only a mild 
decline was observed in the “plus KC1” preparations. Even within the 
“plus KC1” group, the lower rates of the ferricyanide series (Table IV) 
are also reflected in a greater decline in yields. The attainment of 
a theoretical yield of oxygen seems to be dependent on the photo- 
lytic mechanism winning the “race” against time with the concomi¬ 
tant inactivation of the system. The higher the rate, the better are 
the chances for winning the race. If this interpretation is correct, it 
holds important implications for studying effects of inhibitors, pH, etc., 
on the photolytic process. Conclusions about the effects of these factors 
will have validity only if the control measurements indicate high rates 
and yields. 

Effect of pH 

Divergent results were reported on the influence of hydrogen ion 
concentration on oxygen evolution. Hill and Scarisbrick (13) found an 
optimum at pH 8, with the activity falling to half at pH 6.5 and 
vanishing at pH 5.6. Warburg and Liittgens (6) imply that pH 6.5 was 
their optimum, the oxygen evolution becoming greatly diminished or 
stopping altogether at pH 7.4. Holt and French (10) found that the 
optimum pH varied according to the temperature and method used for 
measuring the activity of chloroplasts. With ferricyanide at 15°C. the 
optimum for the evolution of oxygen was pH 7, but with the hydrogen 
ion measurement technique different optimal values were obtained at 
different temperatures: pH 7 at 10°C. and pH 7.6 at 3°C. 



Fig. 1. Effect of pH on oxygen evolution by chloroplast fragments. 



152 


DANIEL I. ARNON AND P. R. WHATLEY 


The influence of pH from 5.2 to 7.7 under our conditions (0.5 mg. of 
chlorophyll per vessel, 15°C., quinone) is shown in Fig. 1. All measure¬ 
ments were made in M/15 phosphate buffers adjusted to the respective 
pH values. The final pH at the end of the manometric measurements 
was determined with a glass electrode. The optimum pH was found to 
be approximately 6.9. The activity of the system fell off rather sharply 
on the alkaline side as compared with a more gradual decline on the 
acid side. It is possible that here again direct observations are compli¬ 
cated by the interaction of two factors: increased activity and accel¬ 
erated deterioration, both occuring at a slightly alkaline pH. This 
interpretation, if correct, would account for the apparently more 
alkaline pH optimum observed at low temperatures by Holt and French 
( 10 ). 

Effect of Inhibitors and Narcotics 
Hydroxylamine 

This inhibitor is of special significance to the oxygen evolution by 
chloroplasts and chloroplast fragments. The inhibitory action by hy¬ 
droxylamine on photosynthesis in intact green cells is attributed to 
its being a specific inhibitor for the oxygen evolution reaction of the 
process (15). Hydroxylamine would thus be expected to be the inhi¬ 
bitor par excellence if the photolytic system in chloroplasts were indeed 
the same as the oxygen-releasing mechanism in the intact green cell. 
Yet, neither Hill (2), using the oxyhaemoglobin method for measuring 
oxygen output by chloroplasts, nor Aronoff (8), with the quinone 
method and chloroplast fragments, found any hydroxylamine inhibi¬ 
tion. On the other hand French, Anson and Holt (14) found, with 
ferricyanide as the oxidant, a 72% inhibition with 10 -4 M hydroxyl¬ 
amine, and Macdowall (16) using the phenol indophenol method ob¬ 
tained a 50% inhibition with 3.1 X 10~ 4 M hydroxylamine. 

Our own results with hydroxylamine are shown in Fig. 2. A marked 
inhibition of oxygen evolution was obtained, the 50% value corre¬ 
sponding to a concentration of approximately 2 X 10~ 4 M. It is pos¬ 
sible that the absence of inhibition by hydroxylamine observed by 
Hill (2) and Aronoff (8) was due, in the first instance, to a reaction of 
the inhibitor with the haemoglobin used and, in the second, to the low 
activity of the chloroplast preparations to which no chloride was added. 



CHLOROPLAST OXYGEN EVOLUTION 


153 


Sodium Azide 

As with hydroxylamine, the evidence that this inhibitor of photo¬ 
synthesis (15) also inhibits oxygen evolution by chloroplast prepara¬ 
tions, was contradictory. Hill (2) and Aronoff (8), who measured 
oxygen evolution by the oxyhaemoglobin and quinone methods, respec¬ 
tively, found no azide inhibition. French, Anson and Holt (14), how¬ 
ever, using the ferricyanide test, obtained a complete inhibition of 
oxygen evolution with 10~ 3 M sodium azide, and Macdowall (16) 
found with the phenol indophenol technique that a 50% inhibition 
was given by 8 X 10~ 2 M azide. 



I0' 4 M 

No Hydroxylamine 


4'10' 4 M 


6<I0‘ 4 M »-o-o 
-- 


Fio. 2. Hydroxylamine inhibition of oxygen evolution by chloroplast fragments. 


The inhibition of oxygen evolution by azide under our conditions is 
shown in Fig. 3. A 50% inhibition was obtained with a sodium azide 
concentration of 8 X 10 -4 M. It is possible that the suggested explana¬ 
tion of the negative results of Hill (2) and Aronoff (8) with hydroxyl¬ 
amine is also applicable to their negative results with azide. 


o-Phenanthroline 

Warburg and Liittgens (6) found that the strong metal-binding 
agent, o-phenanthroline, is a powerful inhibitor of the oxygen evolution 



154 


DANIEL I. ARNON AND F. R. WHATLEY 


by chloroplast preparations. Because of its great effectiveness, a 50% 
inhibition being given by less than 4.2 X 10~ 5 M, these authors con¬ 
cluded that o-phenanthroline is an inhibitor rather than a narcotic. 
The inhibitory effect of o-phenanthroline was confirmed by Aronoff (8), 
and by Macdowall (16), who found a 50% inhibition with 2.7 X 10~* M. 

The striking inhibition of oxygen evolution by phenanthroline was 
fully confirmed in our experiments, using quinone as the oxidant: a 
50% inhibition was obtained with 9 X 10~ 8 M o-phenanthroline (Fig. 
4). The bearing that o-phenanthroline inhibition may have on the 
participation of a metal in the photolytic mechanism of chloroplasts 
will be discussed elsewhere. 


Phenylurelhane 

Warburg (17) has shown that phenylurethane acts as a narcotic, 
inhibiting the light reaction of photosynthesis. A concentration of 
5 X 10 -4 M reduced the photosynthesis of Chlorella cells by 50%. 
Interestingly enough, Warburg and Liittgens (6) found that phenyl¬ 
urethane produces a similar effect on oxygen evolution by chloroplast 
fragments: a 50% inhibition was obtained by a concentration of 
6.1 X 10~ 4 M, Concordant results were also obtained by Hill and 
Scarisbrick (13). 

Aronoff (8) and Macdowell (16), however, while confirming the 
narcotic action of phenylurethane, found the effective concentrations 
to be much higher than those for photosynthesis and photolysis 
reported above. The former obtained a 50% inhibition with 1.2 X 10 -2 
M, and the latter with 2 X 10~ 3 M. In our experiments the effective¬ 
ness of phenylurethane was similar to that found by Warburg and 
Liittgens (6) and Hill and Scarisbrick (13). A 50% inhibition was 
given by 2 X 10 -4 M phenylurethane. 

Summary 

1. Illuminated chloroplast fragments yielded stoichiometric quanti¬ 
ties of oxygen with 3 oxidants: quinone, ferricyanide and phenol 
indophenol. 

2. Based on chlorophyll content, the rates of oxygen evolution by 
chloroplast fragments and by the cytoplasmic fluid were comparable. 

3. The stability of the oxygen-evolving system under various con¬ 
ditions. was investigated. 



CHLOROPLAST OXYGEN EVOLUTION 


155 



Fia. 3. Sodium azide inhibition of oxygen evolution by chloroplast fragments. 



Fig. 4. o-Phenanthroline inhibition of oxygen evolution by chloroplast fragments. 



156 


DANIEL I. ARNON AND F. R. WHATLEY 


4. Support for the identity of the oxygen-liberating mechanism in 
isolated chloroplast fragments with that in the intact green cells was 
found in the inhibitory effects of hydroxylamine and phenylurethane. 
Sodium azide also inhibited the photochemical oxygen liberation by 
chloroplast fragments. The powerful inhibition by o-phenanthroline 
was confirmed. 


References 

1. Hill, R., Nature 139, 881 (1937). 

2. Hill, R., Proc. Roy. Soc. London 127B, 192 (1939). 

3. Brown, A. H., and Franck, J., Arch. Biochem. 16, 55 (1948). 

4. Holt, A. S., and French, C. S., Chap. 14 in “Photosynthesis in Plants,” J. 

Franck and W. E. Loomis, Editors. Iowa State College Press, Ames, Iowa, 
1949. 

5. Arnon, D. I., Plant Physiol. 24, 1 (1949). 

6. Warburg, 0., and LOttgens, W., Biochimia 11, 303 (1946); also published as 

Chap. 20 in Warburg, O., “Schwcrmetalle als Wirkungsgruppe von Fermenten. ,, 
W. Saenger, Berlin, 1946. 

7. Van Niel, C. B., Advances in Enzymol. 1, 263 (1941). 

8. Aronoff, S., Plant Physiol. 21, 393 (1946). 

9. Hill, R., and Scarisbrick, R., Nature 146, 61 (1940). 

10. Holt, A. S. and French, C. S., Arch. Biochem. 9, 25 (1946). 

11. Holt, A. S., and French, C. S., ibid. 19, 368 (1948). 

12. Kumm, J., and French, C. S., Am. J. Botany 32, 291 (1945). 

13. Hill, R., and Scarisbrick, R., Proc. Roy. Soc. London 129B, 238 (1940). 

14. French, C. S., Anson, M. L., and Holt, A. S., private communication. 

15. Rabinowitch, E. I., Photosynthesis. Vol. I, Interscience, New York, 1945. 

16. Macdowall, F. D. H., Plant Physiol. In press. 

17. Warburg, O., Biochem. Z. 100, 230 (1919). 



Letters to the Editors 


A New Fluorometric Method for the 
Determination of Epinephrine 

When a solution of epinephrine is made strongly alkaline and allowed 
to oxidize in air, fluorescence may be observed when the exciting wave¬ 
length is of the order of 365 mu (1). This fluorescence is not specific for 
epinephrine in that we have observed that various substances such as 
protein hydrolyzates, aged tryptophan, and tyrosine will fluoresce 
under similar conditions. Further, this fluorescence is transient. The 
fluorescent material is not completely extractable with organic solvents 
and the fluorescence is, therefore, affected by salt concentration. 

We have been able to convert epinephrine to a substance which will 
fluoresce when excited by a longer wavelength ( ca . 435 mu), which 
incident light will not excite the interfering substances listed above. 
This was accomplished by treating the epinephrine solution with am¬ 
monia containing an organic primary amine such as ethylenediamine, 
butylamine, propylenediamine, benzylamine, aniline or o-phenylene- 
diamine. A fluorescent product results which is extractable with ali¬ 
phatic alcohols such as butyl and amyl alcohols. This fluorescence, in 
contrast to that obtained when epinephrine is treated with alkali, is 
quite stable and will maintain constant fluorescence for several days. 
The fluorescence is excited by light of 435 mu, the secondary filter, 
having a maximum transmission at 570 m/u, having negligible trans¬ 
mission below 500 m^i. The fluorescence of this material resembles the 
fluorescence of riboflavin. The reaction probably involves an oxidation 
of the catechol structure to the quinone and subsequent condensation 
with the amine. Catechol itself gives a similar reaction. 

For the amines containing an aromatic nucleus, such as benzylamine, 
an excitation wavelength shorter than 435 m p is needed. We therefore 
limited our study to the aliphatic amines, particularly ethylenediamine 
to achieve specificity. Thus, by using a longer wavelength for excitation, 
and extracting with solvents, we avoid interference from many sub¬ 
stances and avoid salt effects. 


157 



158 


LETTERS TO THE EDITOBS 


Distinction from riboflavin itself is made by the ease with which 
epinephrine may be completely destroyed by oxidizing agents (KMnOi) 
to which riboflavin is resistant. 

In a typical procedure, to 0.05 y of epinephrine in 1 ml. of 1 N HC1 is 
added 0.5 ml of a 2:1 mixture of G N NH 4 OH and redistilled ethylene- 
diamine. The mixture is heated on a water bath for 1 hr. at boiling tem¬ 
perature and 5 ml. of redistilled n-amyl alcohol is added, while the mix 
is still warm. The mixture is vigorously shaken in a test tube with 
ground glass stopper in which the reaction was carried out. The solu¬ 
tion is allowed to cool to room temperature and centrifuged at 2500 
r.p.m. The fluorescence of the alcohol layer is measured in a Farrand or 
Pfaltz and Bauer fluorometer against a fluorescein standard, subtracting 
a blank due to scattered light, which was run through the same proce¬ 
dure but containing no epinephrine. 

The study of the nature of this reaction and its application to the 
assay of epinephrine in normal and pathological sera is being investi¬ 
gated and will be reported in a more detailed report later. 

References 

1 . Block, W., Hdv. Physiol. Pharmacol. Ada 6 , 122 (1948). 

Department of Biochemistry, 
and Department of Pediatrics, 

Jewish Hospital of Brooklyn, 

Brooklyn, N. Y. 

Received April 29,1949 

The Production of Usnic, Didymic, and Rhodocladonic 
Acids by the Fungal Component of the 
Lichen Cladonia cristatella 

The occurrence of certain organic substances which are peculiar to 
lichens, has been demonstrated by Zopf (1) and others. The diagnostic 
value of these substances, which in great part are organic acids, has 
become a matter of primary importance in the recognition of lichen 
species as demonstrated by Asahina (2), Evans (3), and others. 
Lichens and lichen substances, moreover, have recently attracted 
attention because of their antibiotic properties. In 1944, Burkholder 
et al. (4) demonstrated antibacterial activity of extracts from 27 lichen 
species against Staphylococcus aureus and Bacillus subtilis and from 4 


Samuel Natelson 
Julius K. Luoovoy 
Joseph B. Pincus 



LETTERS TO THE EDITORS 


159 


against Alcaligenes fecalis. In 1945, in a report of further investigations 
in this field, Burkholder and Evans (5) showed that the primary anti¬ 
biotic effect against B. subtilis was to be attributed to usnic acid, a 
specific lichen substance. In the same year, Asano et al. (6) reported 
the antibacterial effect of lichesteric acid and its derivatives against 
staphylococci. Stoll et al. (7), in 1947, reported that usnic acid had a 
high potency of activity against the tuberle bacillus, and Shibata et al. 
(8), in 1948, demonstrated antibacterial action of the same acid against 
the avian type of this bacillus as well as against Staph, aureus. 

The fungus and the alga, which constitute a lichen species, have been 
separated, in certain cases, under aseptic conditions and growth inde¬ 
pendently in synthetic media, as reported by Tobler (9) and others. 
Up to the present time, however, very little is known concerning the 
physiology of either of these organisms when cultured separately. 
Furthermore, the capacity for the synthesis of the specific substance or 
substances characteristic of any lichen species has not been established 
as a function peculiar to the lichen as a whole or to either of its com¬ 
ponents. In the case of the lichen Cladonia cristatella Tuck., however, 
cultures of the fungal and algal components isolated and grown in 
synthetic media in this laboratory, when subjected to the microchem¬ 
ical tests developed by Asahina, show that the fungal component is 
autonomous in the capacity to synthesize usnic and didymic acids. 
Further evidence in support of this has been supplied by Dr. D. M. 
Bonner, who demonstrated by paper-partition chromatography the 
presence of a substance in these cultures that agrees with usnic acid 
in both Rf value (in phenol saturated with water) and in color reaction 
with FeCl 3 . The sample of usnic acid used as a basis for comparison in 
this demonstration was obtained from Dr. Y. Asahina of the Pharma¬ 
ceutical Inst., Tokyo Imperial Univ. Rhodocladonic acid, which is 
normally present in the apothecium of this species, also developed in 
some of the cultures of the fungal component as indicated by the red 
pigmentation of portions of the mycelium. A detailed account of this 
study will be published in the future. 

References 

1. Zopp, W., Die Flechtenstoffe, 449 pp. G. Fischer, Jena, 1907. 

2. Asahina, Y., Botan. Mag. [ Tokyo ] 51, 579 (1937). 

3. Evans, A. W., Torrey Botan. (Stub Bull. 70, 139 (1943). 

4. Burkholder, P. R., etal.'P’roc. Nail. Acad. Sci. U. S. 90, 250 (1944). 



160 


LETTERS TO THE EDITORS 


5. Burkholder, P. R., and Evans, A. W., Toney Botan. Club Butt, 72, 157 (1945). 

6. Asano, M., et cd ., Igakusoran 1, 41 (1945) (in Japanese). 

7. Stoll, A., et al , Experientia 3, 11, 115 (1947). 

8. Shibata, S., et al., Jap. Med. J . 1, 151 (1948). 

9. Tobler, F., Biologie der Flechten, 265 pp. G. Bomtraeger, Berlin, 1925. 

Osborn Botanical Laboratory, Hempstead Castle 

Yale University , Flora Kubsch 

New Haven, Conn. 

Received May 9,1949 


The Biosynthesis of the Penicillins 


The following considerations have led the authors to suggest that the 
penicillin-type molecule (I) (1), is formed in Penicillium notatum from 
the amino acids, L-cysteine and D-/3-hydroxyvaline, and a carboxylic 
acid (II). 


Me 
Me—i— 


-CH—COOII 


i N 
^CH \o 

iIhco-r 


Penicillin 

I 


Me 


Me 

-A-- 


i: 


-CH—COOH 0-hydroxyvaline 


H NH, (b) 
t 

SH CO-OH cysteine 

All 2 —CH 
(a) ! 

nh 2 

HO • CO • R carboxylic acid 

Suggested as precursors of penicillin 
II 


The formation of a sulphide linkage between the mercapto acid, 
cysteine, and the hydroxyacid, /3-hydroxyvaline, could be compared 
with the synthesis of cystathionine from cysteine and homoserine 
effected by Neurospora (2). 

The assumption may, therefore, be made that the two amino acids, 
either free, or as part of larger peptide or protein molecules, may form 
the sulphide bridge at f and the peptide link at *, with further ring 
closure involving loss of one H atom from (a) and one from (b) to give 
the dicyclic structure (the order of these processes being, as yet, 
unspecified). Experiments carried out by one of us (3) indicate that the 
acylation process at t takes place after the synthesis of the dicyclic 



LETTERS TO THE EDITORS 


161 


portion, which latter appears to be associated with the cell-substance 
of the mold. 

The formation of dimethylpyruvic acid from glucose by a related 
mold, Aspergillus niger (4), appears to indicate that the production of 
substances related to valine may readily take place in similar molds. 

One of us (5) has observed that the production of dimethylpyruvic 
acid is stimulated by the presence of acetate in the mold medium. 
Acetate has also been found to increase the production of penicillins 
by P . notatum (6). These phenomena may be linked, in that a precursor 
of these products might be jS-hydroxydimethylpyruvic acid. 

Further work on the biosynthesis of dimethylpyruvic acid is pro¬ 
ceeding in this laboratory. 


References 

1. Committee on Medical Research, 0. S. R. D. (Washington), and M. R. C. (Lon¬ 

don), Science 102, 627 (1945). 

2. Horowitz, N. H., J. Biol. Chem. 171, 255 (1947). 

3. Hockenhull, D. J. D., unpublished work. 

4. Gida, T., J. Shanghai Sd. Inst. [IV] 1, 201 (1935). 

5. Ramachandran, K., Ph.D. Thesis, University of Manchester ,1948. 

6. Calam, C. T., and Hockenhull, D. J. D., J. Gen. Microbiol 3, No. 1, 19 (1949). 

Faculty of Technology , D. J. D. Hockenhull 

The University, K. Ramachandran 

Manchester , England. T. K. Walker 

Received May 14, 1949 


A Metabolic Relationship between the 
Aromatic Amino Acids 

The present work involves a further investigation of a Neurospora 
mutant, C-86, previously mentioned by Lein, Mitchell and Houlahan 
(1) as one that can utilize anthranilic acid, indole, tryptophan, kynure- 
nine, 3-hydroxyanthranilic acid, and nicotinic acid as supplements for 
growth. Mutant C-86, when crossed to a “wild-type” strain, wasr found 
to differ from this wild type strain, with respect to tryptophan biosyn¬ 
thesis, by a mutation at a single locus. 

A number of compounds were tested for growth-promoting proper¬ 
ties for this mutant. These included: 3,4-dihydroxyphenylalanine, 
anthranil, benzoic acid, aniline, p-aminobenzoic acid, formylanthranilic 
acid, isatoic acid, cfs-cinnamic acid, frans-cinnamic acid, phloroglucinol, 



162 


LETTERS TO THE EDITORS 


phenylacetic acid, p-aminophenylacetic acid, /9-phenylethyl alcohol, 
phenyl-DL-a-alanine, /3-phenylethylamine, salicylic acid, coumarin, 
coumaric acid, 2-carboxyindole, 3-carboxyindole, cinnamaldehyde, 
phenylalanine, and tyrosine. Of these compounds, phenylalanine, tyro¬ 
sine, and trans- cinnamic acid were active in promoting the growth of 
C-86. The relative growths of this mutant on supplements of trypto¬ 
phan, indole, anthranilic acid, phenylalanine, tyrosine and frans-cinna- 
mic acid are given in Table I. 


TABLE I 

The Relative Growths of Neurospora Mutant C-86 
in the Presence of Various Supplements 

(The mold weights are for 20 ml. cultures grown at pH 4.6 and 25°C.) 


Dry wt. of mold—mg.—3 days growth 


IaM 

Tryptophan 

Indole 

Anthranilic 

acid 

Phenyl¬ 

alanine 

Tyrosine 

Trans -cinnamic acid 

0.1 

n 

7 


0 

0 


0 

0.2 


18 


2 

1 


0 

0.4 

20 

29 


5 

3 


0 

0.8 


36 


12 

4 


2 

1.4 


41 


18 

8 


5 

2.0 

36 

36 

42 

21 

14 

1 

6 

2.0 

-j- 

35 

30 

40 

27 

20 

0 

7 


a Growth of mutant C-86 on supplements of frana-cinnamic acid at pH 5.6 and 25°C. 


Neurospora mutant, 40008, which utilizes anthranilic acid, indole or 
tryptophan for growth, cannot use either phenylalanine, tyrosine, or 
frans-cinnamic acid. Apparently, strain C-86 has a genetic block which 
occurs at a point earlier in a reaction series involving tryptophan than 
does the block in strain 40008. This would imply that phenylalanine, 
tyrosine and frans-cinnamic acid are involved in the biosynthesis of 
tryptophan prior to the formation of indole or anthranilic acid in 
Neurospora. Another Neurospora strain, E-5212, utilizes phenylalanine 
for growth but none of the other substances found to promote the 
growth of strain C-86. 

The evidence presented suggests the possibility of a common pre¬ 
cursor to the aromatic amino acids. 










LETTERS TO THE EDITORS 


163 


Acknowledgments 

This work was supported by grants from the Williams-Waterman Fund for the 
Combat of Dietary Diseases, the Rockefeller Foundation and the Office of Naval 
Research, United States Navy Department, N6 onr 244, Task Order No. 5. 

References 

1. Lein, J., Mitchell, H. K., and Houlahan, M. B., Proc. Natl. Acad. Sci. U. S. 

34, 435 (1948). 

Kerckhoff Laboratories of Biology , 

California Institute of Technology , 

Pasadena 4, California 

Received May 18, 1949 

X-Ray Diagnostic Agents—^2-(3,5-Diiodo- 
4-Hydroxybenzyl)-Benzoic Acid 

In a recent publication, Jones et al. (1) have reported that, in the 
course of testing 2-(3,5-diiodo-4-hydroxybenzyl)-benzoic acid (I) as a 
cholecystographic agent in dogs, the extrahepatic biliary ducts were 
visualized along with the gall bladder. Since the inception of cholecys¬ 
tography in 1924, several iodinated aromatic acids (2) have been used 
or studied clinically as cholecystographic agents, and, without exception 
these compounds were found to be too toxic or exhibited untoward 
reactions such as diarrhea, nausea, and vomiting. In addition, poor 
absorption of several of these diagnostic agents resulted in interfering 
shadows in the cholecystograms. 

As part of a comprehensive program on X-ray diagnostic agents, we 
prepared and tested I some time ago in view of its structural similarity 
to a-phenyl-j8-(3,5-diiodo-4-hydroxyphenyl)-propionic acid (II) (3). 
It is to be noted that II, and all of the other recently described iodin¬ 
ated compounds (4), suggested as cholecystographic agents are deriva- 



I II 


Joseph F. Nyc 
Francis A. Haskins 
Herschel K. Mitchell 



164 


LETTERS TO THE EDITORS 


tives of fatty acids. It has been assumed that the latter type of com¬ 
pounds have been satisfactory clinically as cholecystographic agents 
by virtue of the lipophilic character of these compounds and their 
sodium salts and their consequent property of being excreted into the 
bile and concentrated in the gall bladder. In view of the fundamental 
difference in the nature of the carboxyl group in compounds of typo I 
and II, and the clinical history of type I compounds, it is of importance 
from the standpoint of the correlation of structure and cholecys¬ 
tographic property to determine the clinical efficacy of compound I. 

In view of the pharmacological and clinical interest in I, we wish to 
record herein our synthesis of I, since Jones ef al. (1) describe only the 
pharmacological data for this compound. The requisite intermediate, 
2-(p-hydroxybenzyl)-benzoic acid (III) was obtained in 80% yield 
by the Raney reduction of 2-(p-hydroxybenzoyl)-benzoic acid. Iodi- 
nation of III with potassium triiodide gave I in 83% yield. 

Experimental 

2-(p-Hydroxybenzyl)-Benzoic Acid (III) 

To a solution of 20 g. of 2-(p-hydroxy benzoyl)-benzoic acid (5) in 350 cc. of 10% 
NaOH, there was added, with stirring, 30 g. of Raney’s nickel-aluminum alloy (6) in 
the course of 1.5-2 hr. The reaction mixture, after heating for an additional hour, was 
filtered and the residual nickel catalyst washed with hot 2% NaOH. The combined 
filtrate and washing was acidified to Congo red paper with HC1, and, after cooling the 
acidified solution, the crude benzylbenzoic acid was filtered; yield 18 g., m.p. 130- 
136°C. Recrystallization from water gave a white product which melted at 148- 
149°C.; literature m.p. 145-146°<3. (7). 

2-(3,5-Diiodo~4-Hydroxybenzyl-Benzoic Acid (I) 

To a solution of 22.8 g. (0.1 mole) of III in 800 cc. of 0.5 N NaOH solution, there 
was added dropwise, with stirring, a solution of 50.8 g. of iodine and 50.8 g. of KI 
dissolved in 250 cc. of H 2 0. The iodinated mixture was stirred for 1 hr. at room tem¬ 
perature, filtered through Supercel, and the filtrate acidified with S0 2 until acid to 
litmus paper. The white crystalline solid was filtered, washed with water and dried; 
yield 40 g. (83%), m.p. 209-211°C. Recrystallization from ethyl acetate-petroleum 
ether or chloroform-petroleum ether gave I as a fine white crystalline material melting 
at 212.5-213.5°C.; literature 208-209°C. (1). 

Anal. Calcd. for CmHioOsI 2 : I, 52.90; Neut. equiv. 240. 

Found: I, 52.82; Neut. equiv. 239. 



LETTERS TO THE EDITORS 


165 


References 

1. Jones et al ., Radiology 51 , 225 (1948). 

2. Graham and Cole, J. Am. Med. Assoc. 82 , 613 (1924); Orator and Walchshofer 

Devi. Z. Chir. 205 , 86 (1947); Dohrn and Diedrich, U. S. Pat. 2,116,104, May 
3,1938; U. S. Pat. 2,220,086, Nov. 5, 1940; Junkmann, Klin. Wochschr. 20 , [5] 
125 (1941); see also 3a. 

3. The compound, Priodax, has been extensively used clinically as a gall bladder 

contrast agent; (a) Einsel and Einsel, Am. J. Digestive Diseases 10 , 206 
(1943); (b) Vaughan and Eichwald, Radiology 43, 578 (1948); (c) Dannen- 
berg, Am. J. Roentgenol. 51 , 328 (1944); (d) Dohrn and Diedrich, U. S. Pat. 
2,345,384, March 28, 1944. 

4. Natelson, Kramer and Tekel, U. S. Pat. 2,400,433, May 14, 1946; Epstein, 

Natelson and Kramer, Am. J. Roentgenol. 56 , 201 (1946); Schwenk and 
and Papa, U. S. Pat. 2,436,270, Feb. 17, 1948; IIoppe and Archer, Federation 
Proc. 8, 303 (1949). 

5. Hubacher, J. Am. Chem. Soc. 68, 718 (1946). 

6. Papa, Schwenk and Whitman, J. Org. Chem. 7, 587 (1942). 

7. Bistrzychi and Yssel de Schepper, Ber . 31 , 2792 (1898). 

Chemical Research Division , Domenick Papa 

Schering Corporation , 

Bloomfield , N. J. 

Received May 25 , 1940 

A Role of Vitamin Bi 2 in the Normal Mammal 1 

In continuance of previously reported work on an unidentified 
nutrient, X, for rats (1,2,3), we have now compared (Table I) crystal¬ 
line vitamin B 12 with APA 15 unit liver extract fed to weanling young as 
supplements to X-deficient basal rations (BR) adequate in respect to 
all chemically identified nutrients and containing 25%, 45%, and 65% 
of protein. The test young were reared by mothers that were on an 
X-deficient ration during the nursing period. In accordance with our 
previous findings, the 14 day growths on BR cotaining an X-deficient 
casein (casein C) decreased from 36 g. to 22 g. and 4 g., respectively, 
with these increases in percentage of protein. It is clear that supple¬ 
mentation with Bn increased the rate of growth on all the levels of 
protein, and that maximally effective doses of Bi 2 produced practically 
the same effect as doses of liver extract which we have found to be 
maximally effective for male rats (i. e., 0.05 cc. or 0.10 cc. of the brand of 
liver extract used). These doses of these supplements practically 

1 This project was supported by an allotment from Bankhead-Jones Special Re¬ 
search Funds. 



166 


LETTERS TO THE EDITORS 


TABLE I 

Effect of Vitamin on Growth 


Protein in ration* 

: 

Sots of 

litter- 

mate 

male 

rats 

Duration 
of assay 

Negative 

control 

Vitamin Bu (cryst.) 

15 Unit parenteral 
liver extract 

Principal kind 

Level 

Av. gain 
in weight 

Daily 

dose 

Av. gain 
in weight 

Daily 

dose 

Av. gain 
in weight 


per cent 


days 

0 . 

7 

< 7 . 

cc. 

0. 

Casein C 

25 

10 

28 

64 

0.01 

86‘ 

0.0025 

120 






0.05 

130 



Casein C 

25 

3 

14 

35 

2.5 

78 

0.05 

81 






5.0 

78 

0.10 

85 

Casein C 

25 

5 

14 

36 

2.8 

80 

0.10 

81 

Casein C 

45 

4 

14 

22 

2.8 

74 

0.10 

72 

Casein C 

45 

2 

14 

21 

2.8 


0.10 

70 






5.6 

70 



Casein C 

65 

2 

14 

4 

5.0 

53 

0.10 

55 

Casein C 

10 

5 

28 

28 

0.5 

59 



Egg albumin 

10 

5 

28 

39 

0.5 

75 



Cottonseed 

10 

5 

28 

41 

0.5 

84 



Cottonseed 

25 

2 

10 

7 C 

0.5 

42 



Soybean 

25 

2 

10 

5 C 

0.5 

60 




° The basal rations (BR) consisted of casein C (extracted with 10 lots of hot alcohol, 
6 hr. with each lot) or soybean oil meal to furnish the principal protein as indicated, 
dried yeast 10 (o protein 5), cottonseed oil 9.85, Navitol 0.15, salts 4.5, and (A) 
lactose 15 plus added vitamins (thiamine, riboflavin, pyridoxine, Ca pantothenate, 
niacin, inositol, PABA, biotin, pteroylglutamic acid, ascorbic acid, a-tocopherol 
acetate, and menadione) and dextrin to 100 or (B) dextrin to 100 without lactose or 
or added vitamins. No yeast was fed in the deglanded defatted cottonseed flour, heat- 
coagulated egg albumin or 10% casein basal rations; the vitamins were supplied by 
the mixture given above. 

b t for difference from negative controls «* 5.4**. ** Adjacent to a t value denotes 
a probability at the 1% level or beyond; * a probability of at least the 5% level. 

«Weight gains of Birfed rats for the week before feeding this vitamin. 

doubled, tripled, and increased 13-fold the rates of growth, respectively, 
on the 25%, 45%, and 65% protein rations. The growth rates on 
similar rations containing 10% of various proteins were likewise in¬ 
creased by supplementation with Bn, as had also occurred with our 
10% protein ration containing casein C when supplemented with liver 
extract (1). Similar rations containing oil meals- as sources of protein 
were evidently deficient in Bu, the growth rates of rats that survived on 
such rations containing 25% of protein increasing several-fold upon 














LETTERS TO THE EDITORS 


167 


supplementation with this vitamin. The increased growth rates on all 
of the B 12 or liver extract-supplemented rations involved increases in 
food consumption. 

The data in Table II show that the kidneys of similarly prepared 
rats maintained on an unsupplemented Bu-deficient 25% protein ra¬ 
tion became hypertrophied; no definite pathology was detected histo¬ 
logically in these kidneys. 


TABLE II 


Effect of Vitamin Bn Deficiency on Kidney Weights 
The basal ration contained 25% of protein (20% casein C, 5% yeast protein). 
Approximately maximally effective dose of 15 unit APA liver extract was the source 
of Bn activity. 


Period on 
ration 

PairB of scx- 
littcr-mate 
rats 

Av, body wt. when 
sacrificed 

Average weight of kidneys 

No Bn 

Bn 

No Bn 

Bn 

t for diff. 

days 


9 • 

g. 

9 • 

9 . 





Males 




28-36 

6 

142 

228 

2.43 

2.55 

0.8 

54-70 

26 

215 

322 

2.97 

2.79 

2.3* 

201-307 

9 

317 

411 

4.64 

,3.27 

3.9** 


Females 


28-36 

6 

■1 

150 

1.58 

1.68 

0.8 

56-67 

10 

n 

214 

2.11 

1.79 

5.0** 


(Males on stock colony diet: 4, approximately 270 days on diet, av. body weight 
366 g., av. kidney weight, 2.79 g.; 2, approximately 2 years on diet, av. body weight 
423 g., av. kidney weight 3.28 g.) 


It is clear that, with all the levels of protein tested, a deficiency of 
Bn has a very deleterious effect on growth; we have reported that; with 
high levels of protein, such a deficiency may even be fatal (2), and we 
have found with such diets that, over a large range of growth levels, 
the B 12 required to attain a given level of growth is increased. We 
believe that the above results indicate that vitamin B 12 plays a funda¬ 
mental role affecting the capacity of the normal mammal to utilize 
protein. 








168 


LETTERS TO THE EDITORS 


References 

1. Report of the Chief of the Bureau of Dairy Industry , Agricultural Research Adminis¬ 

tration , U . S. Dept . Agr., p. 17, 1944. 

2. Cary, C. A., Hartman, A. M., Dryden, L. P., and Likely, G. D., Federation 

Proc. 5 , 128 (1946). 

3. Hartman, A. M., ibid . 6, 137 (1946). 

Bureau of Dairy Industry , Arthur M. Hartman 

Agricultural Research Administration , Leslie P. Dryden 

U . 5. Department of Agriculture , Charles A. Cary 

BeUsville , Maryland 
Received June 8 , L9^9 



Book Reviews 


Advances in Food Research. Vol. I. Edited by E. M. Mrak and George F. 
Stewart. Academic Press, New York, N. Y., 1948. 459 pp. Price $7.50. 

The rapid improvement and expansion of food research in universities and in the 
food industry has quite naturally been accompanied by a rapid improvement and 
expansion of the literature on food research. A welcome addition to this literature is 
the new annual series of reviews edited by Dr. Emil Mrak of the University of Cali¬ 
fornia and Dr. George F. Stewart of Iowa State College. Food research deals with the 
complex problems of delivering to the consumer the most nutritious and desirable 
foods at the lowest possible prices. These problems of the ancient art of food pro¬ 
duction and preservation are today being attacked by the common modern methods 
of biochemistry, physiology, and technology. The technological or engineering ap¬ 
proach is largely neglected in this first volume of Advances in Food Research —a 
neglect which no doubt will be remedied in future volumes—and the first volume 
contains surprisingly little material that would seem out of place on the pages of 
Archives of Biochemistry. In the space of a short review, the individual articles of 
Advances in Food Research can be described only briefly. 

The nutritive value of the original unprocessed foodstuff is influenced greatly by 
growing conditions, as is shown by the review of Somers and Beeson on “The In¬ 
fluence of Climate and Fertilizer Practices upon the Vitamin and Mineral Content 
of Vegetables.” 

Once food is grown, it must be preserved by some technique, such as cold storage, 
addition of chemical preservatives, canning or drying. All but two of the remaining 
reviews are concerned directly or indirectly with these various methods of preserva¬ 
tion. Two reviews deal with fresh meat and its cold storage. Bate-Smith discusses 
“The Physiology and Chemistry of Rigor Mortis, with Special Reference to the 
Aging of Beef.” This review emphasizes the relations of the practical problems of the 
storage and aging of beef to the present knowledge of muscle chemistry. Lowe dis¬ 
cusses “Factors Affecting the Palatability of Poultry, with Emphasis on Histological 
Postmortem Changes.” The work of Lowe is notable for the effective use of taste 
panels and of histological techniques in addition to the usual chemical methods, in 
the study of the quality of poultry meat and of the changes in quality on storage. 

In his review on “Microbial Inhibition by Food Preservatives,” Wyss discusses in 
terms of modern biochemistry the mechanisms of the actions of many chemical food 
preservatives. The review by Clifcorn, on “Factors Influencing the Vitamin Content 
of Canned Foods,” makes clear the sincere and organized effort now being made to 
learn the effects of canning under different conditions on the vitamins of canned foods 
and to learn how to preserve the vitamins to the greatest extent possible. During the 
war, dehydrated foods, which could be shipped in compact form and stored without 
refrigeration, were of great military importance. The work on two dehydrated foods, 

169 



170 


BOOK REVIEWS 


which were of particular military importance, is reviewed in this volume. Lightbody 
and Fevold discuss dried eggs and Ross discusses dehydrated potatoes. Both the 
initial quality and keeping quality of these products were very much improved by 
the war-time investigations. Dried eggs were finally produced for the Army which, 
even after long storage, could be made into scrambled eggs and omelets of first class 
quality. 

A main cause of deterioration of the war-time dehydrated foods during preparation 
and storage was non-enzymatic browning, now recognized as one of the major reac¬ 
tions for good and for bad with which the food industry has to deal. Stadtman’s 
review discusses the investigations of browning in fruit products, investigations which 
have been of great value for the understanding of non-enzymatic browning in general. 
The three reviews on dehydrated foods which have just been mentioned are based on 
the intensive work supported by the Quartermaster Corps, with which the Editors of 
Advances in Food Research were,’ and still are, intimately connected. 

The quality of a foodstuff depends not only on its chemical composition but also 
on its physical character. The physical character of a large group of fruit products 
is achieved by the use of the jelling agent, pectin. “High-Polymer Pectins and Their 
Deesterification” are discussed by Baker. The whole subject matter of this review is 
typical modern polymer chemistry and enzyme chemistry. 

Finally, we come to the acceptability of the food which lias been produced and 
preserved by one method or another. The review by Lepkovsky discusses “The 
Physiological Basis of Voluntary Food Intake.” The question of appetite is presented 
by Lepkovsky, not from the common psychological point of view, but from the point 
of view of the chemical composition and nutritive value of the food eaten and of the 
physiological state of the animal doing the eating. 

All told, a good beginning in the creation of a review literature in the field of food 
research, although it must be said that some of the reviews are not too well written. 
But good writers of reviews are rare. In particular, it seems that authors do not find 
it easy to write reviews which are of detailed interest to the expert, and at the same 
time, permit the non-expert, without great effort, to get a rapid grasp of the field 
being reviewed. 

M. L. Anson, Continental Foods, Inc., Hoboken, N. J. 

Biochemical Society Symposia. Committee of Publication for the Biochemical 
Society: J. H. Bushill, H. A. Krebs, E. J. King and R. T. Williams. No. I, The 
relation of optical form to biological activity in the amino acid series. A symposium 
held at University College Hospital Medical School London, on February 15, 1947. 
Organized and edited by R. T. Williams. Cambridge. At the University Press. 1948. 

This publication is devoted to the properties, especially biological ones, of the 
optical forms of amino acids. It deals essentially with L-forms and D-forms of a-amino 
acids. A. C. Chibnall discusses their nomenclature, which already has been established 
for a number of years, in the introduction. In the first place, the directional rotation 
in water of the individual amino acid was the criterion for the terms l- and D-form, 
respectively. After it had been established that all pertinent building blocks of al¬ 
bumin, according to their configuration, belong to the L-series, this property was no 
longer used; the L-form rather indicates that it belongs to the L-configuration in 



BOOK REVIEWS 


171 


general. This conclusion facilitates not only the classification of the a-amino acids to 
the natural or synthetic amino acid series; moreover, it is necessary for the recognition 
of polypeptides containing amino acids. It has been widely accepted that the various 
directional rotations of the individual amino acids be expressed within parentheses. 
Thus, l (+)-alanine. The necessity for further agreement in the classification arose 
with the discovery of those related to carbohydrates. 

H. A. Krebs describes the status of our knowledge of l- and D-amino acid dehydrases 
(oxydases). The fundamentals of our knowledge we owe to Neubauer, and primarily 
to Knoop. It is well known that Emil Fischer’s hypothesis, according to which only 
L^a-amino acids occur in albumin, in either the organism or the cell, has been dis¬ 
proved. The D-a-amino acids are present especially in microorganisms. It also has been 
established that polypeptides, which contain amino acids, are hydrolyzed by poly¬ 
peptidases from vegetable and animal tissues. Great astonishment was caused by the 
finding that D-amino acids of animal organisms are broken down in addition to 
L-amino acids. The mere fact that, after feeding of DLramino acids—depending in 
extent on their composition—not 100% of the administered D-form appeared in the 
urine, as was to be expected, indicated that the synthetic form of amino acid was 
being metabolized. An attempt to isolate a D-amino acid oxidase was successful. The 
corresponding L-enzyme system was established beyond doubt considerably later. 
Both kinds of enzymes differ substantially in their properties, but not in the mecha¬ 
nism of their action. O. Warburg and Negelein were able to crystallize the D-form as 
the Ba salt. Even before, the structural form of the enzyme had been recognized 
(flavinadeninedinucleotid proteid). It is of greatest interest that the activity of 
maximally purified L-enzyme, extracted by Green, falls far behind that of fresh tissues. 
The Lr-amino acid oxidase attacks the greater part of the L-a-amino acids. The hydroxy- 
amino acids serine and threonine, the dicarboxylic acids asparagine and glutamine, 
and the diamino acids lysine and ornithine, do not break down. 

Krebs reports on the occurrence of a-amino acid oxidases, their specificity, their 
properties, etc. There are also divers other findings which require further research. 
This, for instance, holds true for the statement that the enzyme obtained from the 
tissues of various vertebrates does not appear to be identical. 

E. A. Zeller obtained especially effective amino acid oxidases from the secretions of 
venom glands of various poisonous snakes. Krebs also mentions the fact that enzymes 
of a special sort are effective for many amino acids: reduction of histidine, arginine, 
cysteine, phenylalanine, and tyrosine. 

An excellent critical survey of the problem of the presence of D-glutamine acids in 
products of the hydrolysis of proteins of malignant tumors by means of strong HC1 is 
given by G. R. Tristram. New methods for the demonstration of amino acids have 
been developed: chromatographic methods, methods of isotope dilution, electro¬ 
dialysis, utilization of microorganisms, particularly for the differentiation of the l- 
and D-forms of amino acids. 

Mention is also made of the possibility of the formation of steroisomers according to 
Walden’s Umkehrung. It might be added here that Walden himself thought that 
Waldenases might be present in the cells. There are no indications at this stage of 
investigation about the presence of D-glutamine acids in the albumin of malignant 
tumors for Kfigl’s concept of the unique position of tumor proteins in regard to their 



172 


BOOK REVIEWS 


structure. Of course, there may be peculiarities in the sequence of the building blocks 
of proteins. It is surprising that no one points out the findings of Kogl and his co¬ 
worker, von Erxleben, in which they list individually in tabular form the extablished 
amino acids. Besides glutamic acid, other amino acids, notably hydroxyglutamic 
acid, are cited. Their discoverer, Dakin, himself doubted their existence. Several 
scientists reported previously that the yield of this amino acid is minute or even zero. 
K6gl subjected relatively little amounts of protein to hydrolysis. 

A. Neuberger investigates exhaustively the metabolism of D-amino acids in the 
organisms of mammals. The tests of feeding of dl-, l-, and D-a-amino acids have been 
mentioned previously. Numerous attempts of this kind also have been carried out on 
human beings. The occurrence of D-pyrrolidone carboxylic acid in the urine of rats, 
who were fed D-glutamine acid, is of great interest. It is pointed out that carnivores 
are better equipped with amino acid ureoxidase than herbivores. A report on the 
relation of amino acids which are foreign to the animal organism follows (Neubauer, 
Knoop): formation of a-keto acids. Entirely new is du Vigneaud and Irish’s proof of 
the transformation of an amino acid into an acetyl derivative of the L-form. The 
accepted view is that of the dehydration of D-amino acids with formation of the 
corresponding a-keto acid and asymmetrical reamination accompanied by simul¬ 
taneous acetylation. Additional investigations with labeled isotopes confirmed this 
reported finding also with other amino acids. 

There follows a survey of the influence of certain D-amino acids on growth. We 
have reports on experiments with rats, mice, and chickens, the results being assembled 
n tabular form. In every case a check is made on whether transfer of a D-form to the 
corresponding L-form can be considered. The experiments with cystine and methio¬ 
nine to determine the effects of growth are of great interest. Derivatives and stereo¬ 
chemical forms were used for comparison (du Vigneaud). The relation of L-serine and 
L-homocysteine, derived from L-cystathionine, merits special consideration. A report 
on the excretion of special derivates of D-cysteine follows. Transformations which were 
established are determined in the combination and properties of mercapto acid, 
which is formed after feeding of aromatic-aliphatic halogen compositions. 

The author deals especially with the relation of aromatic amino acids in the animal 
organism (tryptophan, phenylalanine, and tyrosine). The transformation of the first 
to p-hydrqxyphenylalanine, and the manner of reduction 2,5-dihydroxyphenylalanine 
has been known for a long time. The breaking down of dopa is also mentioned. 

In his contribution on amino acids and microbiological chemistry H. N. Rydon 
gives an excellent survey of the relation of the above mentioned amino acids in the 
metabolism of microorganisms, a field in which substantial progress has been achieved. 

T. S. Work describes the d- and L-amino acids as antibiotics. These chemical sub¬ 
stances, formed by microorganisms, are capable of inhibiting the growth of other 
species. 

Work also discusses the problem of nutrition of bacterial cells in connection with 
the permeability of the cellular border. One always finds results of fundamental im¬ 
portance in the course of these discussions. At present, it is Neurospora, the study of 
which has led to many a new finding on enzymatic cell processes. 

The study of biological properties of various amino acids and the products of their 
transformation—particularly the special composition of polypeptides—raises the 



BOOK REVIEWS 


173 


question, on which F. Bergel reports, whether d- and L-araino acids and their deriva¬ 
tives show special pharmacological effects. 

Other problems dealt with include amino acids and calcium metabolism—calcifi¬ 
cation (H. Lelimann); relation of l- and D-form of amino acids to the growth of 
anaerobic molds (G. M. Hills). M. V. Tracey's note on the possible significance of 
amino acids concludes the work. He wonders whether, for instance, the configuration 
of amino acids might influence the enzymatic reduction of polypeptides—-for example, 
in a protective sense. Scientists were already dealing with this question many years 
ago. 

Altogether the Symposium affords a good survey of our present knowledge in the 
field of amino acids of the l- and D-series, also with reference to their relation in the 
organism. 

Emil Abderhalden, Zurich, Switzerland 

Topics in Physical Chemistry. A Supplementary Text for Students of Medicine. 

By W. Mansfield Clark, DeLamar Professor of Physiological Chemistry, The 
School of Medicine, The Johns Hopkins University. Williams & Wilkins Co., Baltimore 
Maryland, 1948. xv + 738 pp. Price $10.00. 

Many teachers in schools of medicine realize that explanations of physiological 
phenomena, and of important laboratory methods, are often to be found in the prin¬ 
ciples of physical chemistry. Yet the study of physical chemistry is not required of 
students preparing to enter medical schools, and there is little time for this subject in 
courses in medical biochemistry. As a result of this situation, Dr. Clark has prepared 
this book, not as a text for a formal oourse, but “to be drawn upon as the student of 
elementary biochemistry and the maturing student of medicine may find occasion." 

This book is considerably longer than previous textbooks of physical chemistry for 
medical students, but the treatment is, in general, elementary. For the benefit of 
students without experience in quantitative chemistry, Dr. Clark has included chap¬ 
ters on the use of the analytical balance and on the exact measurement of volumes. 
Near the end of the book he gives a good deal of space to the optical instruments used 
in biochemical laboratories. The major portion of the text consists of a presentation of 
the principles of classical physical chemistry, with particular emphasis on their bio¬ 
chemical applications. Pertinent references to the literature are given directly in the 
text rather than in footnotes or in a bibliography. 

The outstanding feature of the book is the skillful way in which the author has 
introduced applications to show the practical value of the principles. In the chapter on 
the gas laws, he describes the manometric apparatus used in the study of tissue metab¬ 
olism. In the chapter on electrolytic conductance, he includes a method for esti¬ 
mating the “total base" of scrum from its conductance. A large part of the chapter on 
mass-action equilibria is devoted to the reversible reaction between hemoglobin and 
oxygen. Other subjects of biochemical interest, not usually treated in books on physi¬ 
cal chemistry, are discussed in the chapters on phenomena associated with membranes, 
protein solutions, and blood-electrolytes. Perhaps the most difficult chapters are those 
on free energy and oxidation-reduction potentials. The chapter on equilibria in proton 
exchanges is unusually good, with its account of titration curves, buffer action, and 
acid-base indicators. 



174 


BOOK REVIEWS 


This book, like Dr. Clark’s previous writings, bears witness to his sound scholarship 
and his practical turn of mind. The reviewer is glad to recommend it to teachers and 
research workers in biochemistry as well as to students of medicine. 

David I. Hitchcock, New Haven, Conn. 

Vitaminas, Metodos de Dosificacion. By Gilbebto G. Villela, Chief of the 
Chemical and Pharmacological Division of the Oswald Cruz Institute in Rio de 
Janeiro, and lecturer in Biochemistry at the Medical Faculty of the University of 
Brazil. With an introduction by Dr. Venancio Deulofeu, Professor of the University of 
Buenos Aires. Translated from Portuguese into Spanish. El Ateneo, Buenos Aires, 
1948. xvi + 442 pp. (bound in paper). 

Latin-America and the entire Spanish speaking world is to be congratulated on 
Gilberto G. Villela’s “Methods of Vitamin Determinations.” Strange as it may seem, 
there is no counterpart of this book in the English language. The subject of vitamin 
analysis has been treated quite differently in “Estimation of the Vitamins,” edited by 
W. J. Dann and G. H. Satterfield (Biological Symposia, 12, 1947) and “Methods of 
Vitamin Assay,” by the Association of Vitamin Chemists (Interscience, 1947). While 
the latter gives exact procedural details on some selected methods for a few of the 
vitamins, the former is a collection of essays on procedures presented in different ways 
by a number of different authors. Villela’s book lists quite evenly all of the important, 
and some of the less established, procedures, usually in a form sufficiently detailed to 
allow a man trained in the field of analysis to carry out the determinations without 
much difficulty. 

Villela discusses each vitamin in a separate chapter. The fat soluble vitamins A, D, 
E, and K are followed by vitamin C and the B complex, thiamine, riboflavin, nico¬ 
tinic acid, pyridoxine, pantothenic acid, biotin, p-aminobenzoie acid, choline, inositol, 
and folic acid. There is no systematic discussion on provitamin D analysis and there 
is ng discussion on vitamin P or the analysis thereof. 

The arrangement of each vitamin chapter may best be illustrated by describing one 
of the chapters more fully: Nicotinic acid and nicotinamide cover 33 pp. in this 
volume. This section begins witli a list of the names given to this vitamin, the empiri¬ 
cal and structural formulae (the ring nitrogen has been left out of the formula for 
nicotinic acid), followed by an historical “introduction” and a presentation of the 
“physicochemical properties and characterization.” Paragraphs on “toxicity” and 
“distribution” in nature are followed by “requirements” where the data recommended 
by the Food and Nutrition Committee of the National Research Council of the U. S. 
are given. “Metabolism,” “coenzymes I and II,” and a discussion of the presence of 
the nicotinic acid molecule in urine and in blood, is followed by a short paragraph on 
pellagra and nicotinic acid deficiency symptoms. All this takes up one-third of the 
total space devoted to this vitamin. Coming to the methods of nicotinic acid deter¬ 
mination, Villela discusses first the bioassay on dogs. Very appropriately he recom¬ 
mends the procedure used by Waisman and Elvehjem. 1 W. J. Dann in the “Estima¬ 
tion of the Vitamins” does not refer to this method, but to & later paper by Schaefer, 

1 Waisman, H. A., and Elvehjem, C. A., The Vitamin Content of Meat. The 
Burgess Publishing Co., Minneapolis, Minn., p. 121, 1941. 




BOOK REVIEWS 


175 


McKibbin and Elvehiem* which, in the eyes of the investigators, describes perhaps a 
more reliable method, although this procedure has not been subjected to sufficiently 
extensive tests to judge the precision. Bioassay for nicotinic acid activity has recently 
been found indispensable in spite of all the other methods available. The problem of 
fortifying rice and grits with the various B vitamins called for a water-insoluble form 
of nicotinic acid. The physiological activity of any such compound can be tested only 
on dogs. Therefore, Villela did well to include the bioassay on dogs. 

A discussion of the well-known chemical assay procedures, the cyanogen bromide 
method and the 2,4-dinitro-chlorobenzene method follow. There is also a paragraph 
on the fluorometric determination of nicotinamide in the presence of nicotinic acid, 
and there are discussions on the modifications of the basic techniques for the deter¬ 
mination of nicotinic acid and derivatives in urine and in blood. Finally there is a 
description of microbiological assay procedures. The LactobaciUus arabinosus test is 
described for routine assays and the test using Proteus vulgaris for nicotinamide in 
blood. The chapter ends with a discussion of the methods available for the determin¬ 
ation of nicotinic acid deficiency in man. There are 66 references to original literature 
in this section, plus numerous other references by author’s name only. 

The chapters on the other vitamins are arranged in the same systematic fashion. 
The selection of the specific modification for each method discussed, and the presen¬ 
tation of the general method, is perhaps not always quite as critical as one might 
expect for a treatise on analyses. One might also wish to find more information on the 
limitations of the various assays and on the sensitivities to be expected from each 
procedure. 

Generally speaking, this reviewer, who does not read Spanish very fluently, is under 
the impression that the various chapters have been prepared with about equal care. 
The chapter on vitamin D is perhaps the weakest because it contains so many state¬ 
ments which are not absolutely accurate: the provitamin D in human skin has not 
been identified (p. 101 ), nor has the involvement of the sebaceous gland been proven 
(p. 101). It was not at the 1931, but at the 1934, Conference that the relation of the 
biological activity of the irradiated ergosterol solution to crystalline vitamin Dj was 
defined, etc . Particularly puzzling to this reviewer is the table on p. 100 listing the 
relative biological activities of the vitamins D 2 , Ds, D 4 , and Ds for chicks and rats. 
Vitamin D* is given a value of 1,000 for both species, D 3 shows 25,000 for chicks and 
1,000 for rats, D 4 (which is not otherwise mentioned or identified in the text) rates 
2,000 for chicks and 100 for rats, etc . 

This book will surely find a wide and well deserved distribution in the Spanish 
speaking countries. To the English reading public, however, this book is not a “must” 
in spite of the fact that a few original contributions are included, such as the fluoro¬ 
metric determination of folic acid solutions. But to brush up on scientific Spanish 
Villela’s book will prove interesting reading. 

H. R. Rosenberg, New Brunswick, N. J. 

* Schaefer, A. E., McKibbin, J. M., and Elvehjem, C. A., J. Biol. Chem . 144, 679 
(1942). 




176 


BOOK REVIEWS 


Recent Progress in Hormone Research. The Proceedings of the Laurentian Hormone 
Conference, Vol. III. Academic Press, Inc., New York, N. Y., 1948. viii + 379 pp. 
Price $7.80. 

The third volume of the proceedings of the Laurentian Hormone Conference is a 
good demonstration of the success of this meeting. Twelve papers were presented by 
authorities in their respective problems and the ample discussion enhances the value 
of the published reports. The problems considered were grouped in 5 main subjects. 

The first paper is on the biochemistry of pituitary growth hormone by Choh Hao 
Li and H. M. Evans. The first demonstration of growth interruption by hypophy- 
sectomy (Caselli, 1900) was followed by the demonstration (Evans and Long, 1921) 
that the injection of anterior pituitary can accelerate the growth rate of the rat and 
produces gigantism. The persistent researches of this laboratory were climaxed when 
the growth hormone was isolated in pure form by Li et al. (1944-1945). In their paper 
the authors describe the physical properties, the chemical composition and the biolog¬ 
ical action of the growth hormone: (1) increase of growth; (2) lowering of blood 
amino acids; (3) nitrogen retention; (4) elevation of alkaline phosphatase level in the 
plasma; (5) increase in weight of liver and thymus. 

In the second section on “Steroid Hormones,” 3 papers were presented. The first 
is by K. Miescher, on the “Relation of Activity to Constitution of Sexogcns,” with 
special reference to doisynolic acid. This substance was obtained by opening the keto 
ring of the estrogens and has a high activity on oral administration. The second paper, 
of S. Licberman and K. Dobriner, is an authoritative contribution to the important 
problem of identification of steroids excreted in urine in health and disease. The third 
paper, by H. L. Mason, is a description of the identification and titration of many 
interesting urinary steroid compounds excreted in cases of adrenal disease, adrenal 
cortical tumors and cortical hyperplasia, observed by Kepler and Sprague at the 
Mayo Clinic. 

The histochemical and histophysical methods in hormone research were discussed 
in 2 papers. E. W. Dempsey emphasizes the difficulties of the problem in his review of 
the procedures now available for characterizing the chemical organization of the cells 
of the endocrine glands. In a short and clear paper, C. P. Leblond gives a summary of 
his work, using the radioiodine that has made it possible to trace the behavior of iodine 
from the time it enters the body through its transformation into thyroid hormone and 
the excretion of this hormone in the feces. 

The physiology and function of the testis is presented in 4 papers and was discussed 
from many angles and by many specialists in the experimental and clinical fields. 
C. W. Hooker contributed his own research and a discussion of a part of the complex 
and debated problem of the biology of the interstitial cells of the testis; there is no 
mention of the classic and fundamental historical papers (Ancel, Borun, etc.). W. 0. 
Nelson and C. G. Heller emphasize the possibility of diagnosing and classifying the 
various types of hypogonadism in male patients on the basis of physical findings, 
testicular biopsy and assays of urinary gonadotrophins. 

The paper on the testis-pituitary relationship in man by C. G. Heller and W. 0. 
Nelson was supplemented by lengthy discussions and factual contributions of many 
specialists in the field. The paper of J. B. Hamilton on the role of testicular secretions 
as indicated by the effects of castration in man and by studies on pathological condi- 



BOOK REVIEWS 


177 


tions and the short lifespan associated with maleness, is a contribution full of excellent 
information. 

The fifth section discussed was “Hormones and Hypertension.” The paper by A. C. 
Corcoran on renal pressor system and experimental and clinical hypertension is a 
masterly presentation and a good resume of the American literature, and especially 
of the personal experience and opinions of the author. The paper of Hans Selye is an 
excellent exposition of his experimental work and his theoretical interpretations of 
the diseases of adaptation described by him. 

B. A. Houssay, Buenos Aires. 




Distribution of a Triphosphopyridine Nucleotide-Specific 
Enzyme Catalyzing the Reversible Oxidative 
Decarboxylation of Malic Acid in 
Higher Plants 1 

Eric Conn, Birgit Vennesland and L. M. Kraemer 2 

From the Department of Biochemistry, University of Chicago, Chicago, Illinois 
Received March 14, 1949 

The widespread distribution of the enzyme oxaloacetic carboxylase 
in the tissues of higher plants has recently been demonstrated (1). The 
reaction catalyzed may be written: 

oxaloacetate ?=* pyruvate + CO 2 . (1) 

Previous work of Ochoa, Mehler and Kornberg (2), on a similar enzyme 
in pigeon liver, has indicated the existence of a close relationship be¬ 
tween the enzymic decarboxylation of oxaloacetic acid (OAA) and the 
oxidation of malic acid by triphosphopyridine nucleotide (TPN). The 
purified “malic enzyme” of pigeon liver catalyzes an overall reaction 
which may be written: 

Z-malate + TPN 0X <=* pyruvate + CO 2 + TPN re d. (2) 

Evidence has previously been published for the presence, in an OAA 
carboxylase preparation from parsley root, of a TPN-specific enzyme 
which catalyzes the reversible oxidative decarboxylation of malic acid 
(3). The experiments described in the present paper constitute an 
examination of the question whether this enzyme can always be dem¬ 
onstrated in OAA carboxylase preparations from higher plants. For 
sake of brevity, this activity will also be referred to as a “malic en¬ 
zyme,” although further work is necessary to establish whether the 

1 This work was supported in part by grants from the John and Mary R. Markle 
Foundation and from the Dr. Wallace C. and Clara A. Abbott Memorial Fund of the 
University of Chicago. 

* Department of Chemistry, University of Chicago. 

179 



180 ERIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMEH 


mechanism of the reaction in plants is identical with that in pigeon 
liver. During this investigation it became apparent that the wide¬ 
spread occurrence of enzymes which inactivate TPN seriously inter¬ 
fered with the demonstration of the malic enzyme. When the assay 
procedure was designed to avoid this difficulty, it was possible to show 
the presence of the malic enzyme in all of the plant carboxylase prepara¬ 
tions tested. 

Materials and Methods 

The protein preparations employed in this work all contained oxaloacetic carboxy¬ 
lase. A summary of the methods of preparation, and of the carboxylase activity of the 
various crude enzymes used, is given in Table I. 

TABLE I 


Description of Plant Preparations 


Sample 

Carboxylase activity 

K /y. jrrotein 

Dry weight. 
mg./ml. 

Wheat germ, Type A 

35 

19 

Wheat germ, Type B 

214 

69 

Wheat germ, Type C 

326 

113 

Beet 

32 

14 

Carrot 

87 

4 

Parsnip 

22 

27 

Parsley root 

75 

10 

Spinach 

4 

20 

Peas 

11 

35 


The Type A wheat germ, beet, carrot, and spinach preparations were the fractions 
obtained by precipitation with (NID 2 SO 4 in the concentration range.of 250-500 g./l. 
of extract or press juice (1). Thc r parsley root and parsnip preparations were dialyzed 
solutions of the lyophilized preparations previously described (3). 

Type B wheat germ: Water extract (4 ml./g. wheat germ) was fractionated with 
solid (NHOsjSCLat pH 6.5, retaining the precipitates, successively, between (NHOaSO* 
concentration limits of: 

t. 185 and 365 g./l. 

2. 100 and 240 g./l. 

Type C Wheat Germ: Procedure was that followed with Type B, the concentration 
limits being: 

1. 190 and 365 g./l. 

2. 240 and 340 g./l. 

3. 290 and 340 g./l. 

The pea preparation, kindly furnished by M. B. Mathews of the Dept, of Chemistry 
was a sample enriched in DPN-formic dehydrogenase. It was precipitated bv 
(XHOaSO^ within the concentration limits of 220 g./l. and 330 g./l., from an extract 
of dried green peas (equal weights of water-soaked peas and 0.1 M NasHPCL). 



ENZYME DISTRIBUTION 


181 


In all cases where (NH.O2SO4 precipitates were used, care was taken to remove the 
salt completely by dissolving the precipitate in 0.025 M phosphate buffer, pH 7.3, 
and dialyzing against water or buffer. Any residue remaining was then centrifuged and 
discarded. 

The carboxylase activity is expressed in term of K/g. dry weight of protein, where 
K is the monomolecular reaction rate constant in reciprocal minutes. Calculations of 
specific carboxylase activity have been described elsewhere (3). 

Yellow enzyme was prepared from brewer’s yeast 3 according to the method of 
Warburg and Christian (4). The partially purified material used in these studies was 
that- obtained by precipitation with methanol and drying over II 2 SO 4 at 0°C. The 
yellow powder was dissolved in water and centrifuged; the clear solution was then 
tised in the test system in amounts known to be in excess. The absorption coefficient 
of this material at 465 m/x was 9.2 cm. 2 5 /g* This figure cannot be used to calculate 
purity from Theorell’s (5) value for the absorption coefficient, since impurities which 
absorb at 465 m m are known to be present 

TPN was prepared from beef liver by a modification of the procedure of Warburg, 
Christian and Oriese (6). The purity of the material was determined by enzymatic 
1 eduction with Zwischenfermerit and glucose-6-phosphate (6) and bv reduction with 
NaBHi (7). Using the value for 0 of 1.3 X 10 7 em. 2 /mole (8), 4 the TPN purity was 
found to be 15%. The purity of diphosphopyridine nucleotide (DPN) samples obtained 
from Schwarz Laboratories was determined by reduction with NaBH 4 and ranged 
between 50 and 60%. The DPN samples did not contain any TPN since the material 
was inactive in the specific manomctric assay system for TPN described by Warburg 
et al. (6). The TPN did not contain appreciable DPN since the same amount of 
absorption at 340 m/x was obtained whether the sample was reduced by NaBH 4 (which 
reduces both DPN and TPN or whether is was reduced specifically by Zwischenfer - 
ment and glueosc-6-phosphate. When tested with the DPN-specific formic dehydro¬ 
genase of dried green peas (9), the maximum DPN content of the TPN preparation 
was determined to be less than 1%. 6 Yeast adenylic acid and adenosine were purchased 
from Schwarz Laboratories and muscle adenylic acid from the Sigma Chemical 
Company. /-Malic acid was purchased from Eastman Chemicals and recrystallized 
twice from ethyl acetate-petroleum other before use. Glycylglyeine was purchased 
from Pfanstiehl Chemicals, and catalase was the commercial product, Catalase- 
Sarett, sold by the Vita-Zyme Laboratories. Glucose-6-phosphate was synthesized 
from the monoacetonc of glucose according to Lovene and Raymond (10). All solutions 
were adjusted to pH 7.4 before use. 

The manomctric experiments were carried out in t he Warburg apparatus at 37°C. 
Experiments measureng O 2 uptake were carried out in an atmosphere of 100%, 0 2 . 
C0 2 evolution was measured according to the direct method described by Dixon (11). 

3 We wish to thank the Ambrosia Brewing Company of Chicago for a supply of 
yeast. 

0 = 73 ^ (cm.‘/mole). 

5 Unpublished experiments of Mr. M. B. Mathews, Dept, of Chemistry, Univ. of 

Chicago. 




182 ERIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMER 


The spectrophotometric measurements were carried out on a Model DU Beckman 
spectrophotometer. 

Pyruvate analyses were made by the hydrazone method of Friedemann and Haugen 
(12), using both the direct and specific extraction procedures. Pyruvate was also 
analyzed by decarboxylation with yeast pyruvic carboxylase (13). Malate was 
determined according to Speck (14). Since high blanks were obtained in this procedure 
it was necessary to use an internal standard prepared by adding known amounts of 
malate to a reaction mixture which had been incubated without malate. Small 
amounts of TPN (1-15 y) were determined manometrically by the Zmschenferme.nl 
method of Warburg el al. (6). 


Results 

Direct Spectrophotometric Tests for TPN-Malic Enzyme 

The spectrophotometric test employed by Ochoa et al. (2) could 
sometimes be applied very successfully to demonstrate the occurrence 
of Reaction 2 in various plant preparations. The test is dependent on 
the measurement of light absorption at 340 my, due to the reduced 
TPN, and has the advantage that the reversibility of the reaction can 
readily be shown. A typical experiment of this sort, with a wheat germ 
preparation, is presented in Fig. 1. The results are essentially similar 



Fia. 1. Reversal of TPN- 
nialic enzyme from wheat germ. 
Reduction and oxidation of TPN 
determined spectrophotomctri- 
cally at 340 m^t. Corex cells, 
d = 1.0 cm. Temperature, 25°C. 
Measurements made against a 
blank which received all addi¬ 
tions except TPN. Optical den¬ 
sity corrected for changes on 
dilution. 1.1 ml. 0.1 M glycyl- 
glycine buffer pH 7.4: 240 y 
TPN, and water made to a 
volume of 3.0 ml. At 0 time, 0.02 
ml. 0.5 M J-malate added. At 1, 
0.16 mg. wheat germ, Type C 
added. At 2, 0.1 ml. 0.01 M 
MnCU added. At 3, 0.1 ml. 0.5 
M pyruvate added. At 4, 0.1 ml. 
1.0 M NaHCOi saturated with 
CO s .added. At 5, 0.1 ml. 0.5 
M /-malate added. 




ENZYME DISTRIBUTION 


183 


to those obtained with parsley root (3) and show the dependence of the 
reaction on the presence of Mn ++ (Co ++ may be substituted for Mn' H ). 
Reoxidation of TPN r( .d by addition of pyruvate and bicarbonate is 
shown at times 3 and 4. The reduction obtained by addition of malate 
at time 5 further demonstrates the ease with which the equilibrium of 
the reaction can be shifted. 



MINUTES 

Fig. 2. Destruction of TPN by carrot enzyme. Reduction of TPN determined 
spectrophotometrically at 340 m^. Corex cells, d = 1.0 cm. Temperature, 25°C. 
Measurements made against a blank which received all additions except TPN. 1.0 ml. 
0.1 M glycylgvlcine buffer, pH 7.4; 2 mg. carrot enzyme; 0.1 ml. 0.01 M MnCl 2 ; 
0.1 nil. 0.1 M /-malate; and water made to a volume of 2.9 ml. At 0 time, 120 7 TPN 
added. 

Application of the spcctrophotometric test to various OAA carboxy¬ 
lase preparations did not, however, give uniformly positive results for 
the presence of the malic enzyme. In some cases no reduction of TPN 
was observed, in others only a small and very slow reduction was 
obtained, and in still others the reduction of the coenzyme did not 




184 ERIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMER 

appear to reach an equilibrium point but was followed by an apparent 
reoxidation, although no known oxidizing agent such as pyruvate or 
bicarbonate had been added. An extreme example of this latter sort, 
obtained with a carrot preparation, is shown in Fig. 2. The possibility 
of reoxidation of reduced TPN by pyruvate in the presence of lactic 
dehydrogenase could be eliminated since the preparation did not con¬ 
tain the latter enzyme. Analysis for TPN remaining at the end of the 
experiment showed that all the TPN had disappeared. 

These and similar results with other plants suggested that TPN- 
inactivating factors were present in variable amounts in the plant 
preparations. One would fail to observe the presence of the malic 
enzyme if these factors destroyed the TPN before commencement of 
observations. Even in cases where the spectrophotometric test was used 
successfully (such as that shown in Fig. 1), it. was possible to show that 
some inactivation of TPN occurred. Thus, the plateau reached pre¬ 
ceding arrow 3 in Fig. 1 is only apparent. If sufficient time was allowed 
to lapse before addition of any oxidizing agent, a decrease in light 
absorption occurred. 

Manometric Test System for TPN-Malic Enzyme 

The malic enzyme could also be detected by a method which involved a substitu¬ 
tion of malate and malic enzyme for glucose-6-phosphate and Zwischrnfermcnt in 
Warburg’s old yellow enzyme system. In this procedure, TPN is reduced by malate 
in the presence of malic enzyme, and the reduced TPN is then oxidized by yellow 
eftzyme which, in turn, is oxidized by molecular oxygen. The rate of O 2 consumption 
may, therefore, serve as a measure of the malic enzyme present provided the other 
components of the sytem are present in excess. A similar system has been used by Adler 
d al. (15) to study isocitric dehydrogenase. 

A detailed study of this reaction sequence was made with the various wheat germ 
preparations which were available. The dependence of 0* consumption on the pres¬ 
ence of all components is shown in Fig. 3. Potassium cyanide (0.0025 M) was em¬ 
ployed in this experiment, as it was in Warburg's Zwischenfcrment system, to inhibit 
catalase and thus prevent the breakdown of H 2 O 2 formed in the reaction between 
yellow enzyme and 0 2 . The O 2 consumption of the complete system containing malate 
wheat germ, TPN, Mn + +, and yellow enzyme is shown in Curve 1. Curve 2 shows that, 
when KCN was omitted, the 0 2 consumption was about one-half that observed in its 
presence. This would be expected if the peroxide formed were decomposed to H 2 0 and 
O* by the catalase known to be present in the Type C wheat germ sample used. 
Substitution of an excess of DPN for TPN resulted in a negligible O 2 uptake, as 
shown in Curve 3. A negligible 0 2 uptake also resulted upon omission of yellow enzyme 
or TPN (Curves 4 and 5), and upon omission of Mn ++ , malate, or wheat germ, or heat 
inactivation of the wheat germ enzyme (Curves 6-9). 



ENZYME DISTRIBUTION 


185 


Glycylglycine buffer was used in these experiments in preference to phosphate 
buffer since the latter precipitates Mn ++ below its effective concentration and, there¬ 
fore, causes some inhibition. Since a 10-fold increase in Mn ++ concentration gave no 
increase in the oxidation rate, the concentration was assumed to be near optimum 
value. The malate concentration (0.0125 M) used in the experiment shown in Fig. 3 



MINUTES 

Fig. 3. Demonstration of TPN-malic enzyme in wheat germ. Complete system 
(Curve 1) contained 0.0125 M J-malatc; 0.0025 M KCN (neutralized to pH 7.4); 
0.05 M glycylglycine buffer, pH 7.4; 0.00025 M MnCl>; 11.3 mg. wheat germ, Type 
C; 1.5 X 10" 6 M TPN; 20 mg. yellow-enzyme; and water to a total volume of 4.0 ml. 
Curve 2, KCN omitted. Curve 3, 5 X 10~ 6 M DPN substituted for TPN. Curves 4 
and 5, yellow enzyme or TPN omitted. Curves 6, 7, 8 and 9, wheat germ enzyme heat- 
inactivated or MnCl 2 , malate, or wheat germ omitted. Reaction started by adding 
yellow enzyme together with DPN when indicated from side arm after temperature 
equilibration. Gas, 0 2 ; temperature, 37°C. KOII in center well. 

was less than the optimum value. In most of the other experiments larger concentra¬ 
tions (0.062 and 0.083 M malate) near the optimum value were employed. An increase 
in the yellow enzyme concentration did not increase the rate of O 2 uptake. 

Evidence that the reaction proceeds according to Eq. (3) 

Z-malate + J 0 2 -► pyruvate + C0 2 + HjO (3) 



186 EHIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMER 


was obtained by measuring malate and 0 2 utilization and pyruvate and CO 2 formation 
in the presence of added catalase and in the absence of cyanide.® It was hoped that 
this would ensure the decomposition of the peroxide formed and thus avoid any 
oxidation by the peroxide itself. However, as will be described later, this objective 
was not always achieved. Since some knowledge of the applicability of the mano- 
metric test system to malic enzyme preparations of varying degrees of purity was 
desired, 3 preparations of wheat germ were employed. Experiments were performed 
with both an excess and a limiting amount of malate and representative results are 
given in Table II. When an accurate measure of malate utilization was desired, small 
amounts of malate were used. 

TABLE II 


Balance Experiments with Wheat Germ 


Type of preparation 

Malate added 
initially 

Malate found 
at end of 
experiment 

Os used 

Pviuvate 

formed 

CO 2 formed 

10 mg. Type A 

3.5 mg. Type B 

11 mg. Type C 

nwra moles 

10 

250 

250 

micro mol ex 

5.4 

micromoles 

6.0 

5.4 

3.5 

micromoles 

5.3 

5.7 

6.0 

micromoles 

7.0 

7.4 

7.3 


All vessels contained 0.067 M glycylglycine buffer, pH 7.4, 0.033 M MnCl 2 ; 20 mg. 
yellow enzyme, 1.2 X 10~ 5 M TPN; 0.03 mg. catalase, water to 3.0 ml., and the 
amounts indicated of malate and wheat germ protein. Reaction was started by tipping 
in yellow enzyme and TPN from side arm after temperature equilibration. Gas, O*; 
temperature 37°C. Reaction followed until 0 2 uptake ceased, usually a period of 
150-180 minutes. At this time, 2.5 ml. of the reaction mixture were removed and added 
to 2.5 ml. of cold 10% trichloroacetic acid. The resulting precipitate was centrifuged 
dbwn and malate and pyruvate analyses by the hydrazone method were conducted on 
properly diluted aliquots of the supernatant. Pyruvate analyses by yeast carboxylase 
were done on aliquots of the untreated reaction mixture. 

According to Eq. 3, the moles of malate removed should be equal to 
the moles of pyruvate and of CO 2 formed and twice as great as the 
moles of O 2 consumed. The data in Table II show that these expecta¬ 
tions were always fulfilled for malate disappearance, pyruvate formation 
and C0 2 formation, but for 0 2 consumption the expected values were 
obtained only with Type C preparation. In the case of Types A and B 
preparations, the 0 2 consumption was twice the theoretical value. 
This is the amount which is expected if the peroxide were not decom- 

• Cyanide can affect the reaction in 3 ways: by inhibiting catalase; by inhibiting 
peroxidase; and by binding OAA as the cyanohydrin. Cyanide effects of the type 
shown in Fig. 3 are not obtained with all preparations. A complete rationalization of 
all the effects observed has not yet been achieved. 





ENZYME DISTRIBUTION 


187 


posed to give II 2 0 and 0 2 . The preparations were found to contain an 
active peroxidase as tested by the method of Avery and Morgan (16). 
Furthermore, Keilin and Hartree (17) and Chance (18) have demon¬ 
strated that catalase may exhibit peroxidase activity under conditions 
where peroxide is generated slowly. It seems probable, therefore, that 
a catalyzed oxidation of an unidentified component of the system by 
peroxide occurred with preparations A and B, but not with, preparation 
C. The possibility of its being glycylglycine may be discarded, as the 
same results were obtained with several other buffers, namely, veronal, 
tris 7 and ammonia. Although pyruvate is reported to react with H 2 0 2 in 
bacterial preparations, even in the presence of catalase (19), no such 
reaction occurs in the plant materials under the conditions of these 
experiments, since pyruvate can be readily determined, by 3 different, 
procedures, to be present in the amount expected. 

Effects of Nucleotides other than TPN on the Manometric Test System 

The rate of 0 2 consumption observed in the manometric test for 
malic enzyme always showed a decline with time. The rapidity of this 
decline varied with different preparations tested. In all cases where the 
experiment was continued for a sufficient period of time the reaction 
eventually stopped. This effect was shown to be due to the destruction 
of TPN added, rather than to an inactivation of the malic enzyme. 

The effects of ATP and 1)PN on the synthesis of TPN in pigeon liver as reported by 
Mehler li at. (20), as well as the work of Mann and Quastel (21), and of Kornberg and 
Lindberg (22), on the ability of nicotinamide to prevent DPN destruction suggested 
that these compounds might las effective in preventing the destruction of TPN. In 
addition, Ccithaml and Vennesland (23) had shown that ATP together with TPN 
increased the incorporation of labeled CO* into isocitrate by parsley root protein and 
suggested that ATP had a sparing action on TPN. 

Examination of these various possibilities revealed that ATP, DPN, 
and muscle adenylic acid could all serve effectively to prevent the 
destruction of TPN in the wheat germ preparations. Results of a 
typical experiment are shown in Fig. 4. This particular experiment was 
designed to demonstrate the effects of DPN and ATP, and also to pro¬ 
vide evidence that the differences in the rates are associated with the 
disappearance or maintenance of TPN. The first portion of Curve 4 in 
this figure is identical with Curve 1, Fig. 3. At 90 min., when the re- 

7 Tris (hydroxy me thy l)-aminomethanc, obtainable from the Commercial Solvents 
Corporation, 17 E. 42nd Street, New York, N. Y. 



188 ERIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMEU 


action had stopped, an analysis for TPN showed that none was pres¬ 
ent. The rate of the reaction was immediately restored almost to its 
initial level when more TPN was added to the reaction mixture, but 
again fell to a low value after 100 min. These findings, therefore, are in 
agreement with the previous observations that there is a TPN-inacti- 
vating factor in wheat germ. For the sake of brevity this destructive 
factor will be referred to as a TPN-ase but the name should not be 
understood to imply any mechanism of its action. Curve 6 of Fig. 4 



Fig. 4. Demonstration of TPN-malic enzyme, in wheat germ. Complete system 
(Curve 4) contained 0.0625 M /-malate; 0.0025 M KCN (neutralized to pH 7.4); 
0.0625 M glycylglycine buffer, pH 7.4; 0.0025 M MnCh; 20 mg. yellow enzyme*, 
1 X 10~ 5 M TPN; 11.3 mg. wheat germ, Type C; and water to a total volume of 4.0 
ml. At 90 min. on Curve 4, another 30 y TPN added from a second side arm. Curve 1, 
lieat-inactivated wheat germ enzyme employed or malate omitted. Curve 2, 1.5 X 
lO" 4 M DPN substituted for TPN. Curve 3, 0.0015 M ATP substituted for TPN. 
Curve 5,1.5 X 10~ 4 M DPN added to the complete system. Curve 6,0.0015 M ATP 
added to the complete system. Reaction started by tipping in yellow enzyme and 
TPN together with ATP or DPN, when indicated, from side arm after temperature 
equilibration. Gas, 0 3 ; temperature, 37°C. KOH in center well. 



ENZYME DISTRIBUTION 


189 


demonstrates that 6 micromoles of ATP greatly prolonged the period 
of rapid 0 2 uptake, and at 90 min. an analysis for TPN showed that 
90% of the TPN initially added could be recovered. It is clear, there¬ 
fore, that ATP addition had prevented the destruction of TPN in this 
case. A similar effect of DPN is shown in Curve 5. The small differences 
between Curves 5 and 6 should not be interpreted as a measure of the 
relative effectiveness of the two compounds, since equimolar concen¬ 
trations were not employed. Blanks obtained by omitting malate, or by 
substituting DPN or ATP for TPN are given as Curves 1, 2 and 3, 
respectively. These data were obtained with Type C wheat germ prep¬ 
aration. Qualitatively identical results are obtained with the Type A 
wheat germ. 

Muscle adenylic acid (adenosine-5-phosphoric acid) gave an effect 
identical with that observed with an equivalent concentration of ATP. 
Yeast adenylic acid (adenosine-3-phosphoric acid) is about one-half 
as effective as muscle adenylic acid or ATP. Nicotinamide, adenine, 
adenosine, guanine, ribose, orthophosphate and pyrophosphate did not 
exhibit any sparing action on TPN destruction in the wheat germ 
samples. 

The manometric test system for malic enzyme has also been applied to a dialyzed 
extract of pigeon liver acetone powder (24). Using 0.5 ml of the extract, 350 /xl. of O* 
were utilized in 180 min. in the presence of 0.08 M malate, and only 15 /xl- were used 
when malate was omitted. ATP and DPN addition increased the O 2 uptake somewhat, 
although not as much as in the case of wheat germ, the values being 520 /xl. and 440 /xl., 
respectively, when 0.002 M ATP or 2 X 10 -4 M DPN are added. These results are in 
agreement with the observation that pigeon liver extracts destroy TPN only very 
slowly and, therefore, large stimulation effects of ATP or DPN should not be observed. 
In one respect, the pigeon liver preparation differs markedly from the plant prepara¬ 
tions. 0.002 M ATP without TPN causes an 0 2 utilization of 200 /xl. in 180 min., 
corresponding to 57% of the value obtained when 1.2 X 10” 6 M TPN is added without 
ATP. Such effects of ATP without added TPN were never obtained in the case of the 
plant extracts. 


Quantitative Assay for TPN Malic-Enzyme 

Attempts were made to determine a range of enzyme concentration 
in which the rate of 0 2 consumption would be directly proportional to 
the quantity of enzyme added. As would be expected from the fore¬ 
going results, no adequate proportionality was observed unless the 
destruction of TPN was prevented. Fig. 5. shows the lack of propor¬ 
tionality obtained with 22 7 TPN per vessel, but no other added 
nucleotide. This experiment illustrates particularly the fact that there 



190 ERIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMER 

is a range of protein concentration in which an increase in amount of 
protein actually causes a decrease in the quantity of 0 2 consumed over 
a given period of time. The disadvantage of using such a procedure 
for quantitative assay is obvious. When, however, the destruction of 
TPN is prevented by addition of adequate amounts of ATP, linear 



MINUTES 

Fig. 5. Oxygen uptake with different amounts of wheat germ. All vessels contained 
0.0625 M Z-malate; 0.05 M glyeylglycine buffer, pH 7.4; 0.0025 M MnCl 2 ; 0.0025 M 
KCN (neutralized to pH 7.4); 20 mg. yellow enzyme; 7 X 10"® M TPN; water to a 
volume of 4.0 ml., and the indicated amounts of wheat germ enzyme, Type 0. 
Reaction started by tipping in yellow enzyme and TPN from the side arm after 
temperature equilibrat ion. Gas, 0 2 ; temperature 37°C. KOH in center well. 

proportionality was observed over a wide range of protein concentration 
as shown in Fig. 6. The different curves do not pass through the origin 
because there is a small and reproducible consumption of 0 2 when no 
wheat germ protein is present. The blank in this experiment is due to 
the yellow enzyme since omission of the latter gives no measureable 0 2 







ENZYME DISTRIBUTION 


191 


uptake whatever. Some crude plant preparations also gave a small O 2 
consumption without added yellow enzyme. Appropriate corrections for 
these blanks can readily be made for each preparation used. 

The assay system described suffers from the disadvantage that the 
behavior of the peroxide formed varies in the different plant prepara- 



Fig. 6. Oxygen uptake as a function of wheat germ concentration. All vessels con¬ 
tained 0.083 M Z-malate; 0.067 M glycylglycine buffer, pH 7.4; 0.003 M MnCl*; 
0.0015 M ATP; 1.7 X 10“ 6 M TPN; 20 mg. yellow enzyme; water to a volume of 4.0 
ml., and the amounts indicated of wheat germ enzyme, Type B, expressed as protein 
nitrogen. Curves are 0 2 uptake for periods of 10, 20, 30, 40, and 50 min. after tipping 
in yellow enzyme, TPN, and ATP from the side arm after temperature equilibration. 
Gas. 0 2 ; temperature, 37°C. KOH in center well. 

tions. It is, therefore, necessary in each case to examine the balance 
between malate disappearance or pyruvate formation and O 2 consump¬ 
tion in order to interpret correctly the quantitative results with respect 
to malic enzyme content. 




192 ERIC 1 CONN, BIRGIT VENNESLAND AND h. M. KHAEMER 


Distribution of TPN Malic-Enzyme 

The manometric test method for TPN malic-enzyme was applied to 
6 plant preparations other than wheat germ. In each case, tests were 
run with and without ATP and DPN in order to ascertain whether 
these substances always had the same effect. Planks omitting malate 
always showed negligible 0 2 uptake. Some preparations gave results 
qualitatively similar to those observed with wheat germ. Among these 
were the samples prepared from beets, carrots, spinach, and peas. In 
each case, approximately the same number of carboxylase units were 
employed per test as was the case with wheat germ. 



Fig. 7. Demonstration of TPN-malic enzyme in parsley root. All vessels contained 
0.05 M glycylglycine buffer, pH 7.4; 0.0025 M MnCh; 20 mg. yellow enzyme; 
1 X 10“‘ M TPN; 12 mg. parsley root; 0.0025 M KCN (neutralized to pH 7.4); 
and water to total volume of 4.0 ml. Additions as indicated were: 0.0625 M /-malate; 
0.0013 M ATP; and 2 X 10“ 4 M DPN. Reaction started by tipping in TPN and 
yellow enzyme together with ATP or DPN when indicated from side arm after temper¬ 
ature equilibration. Gas, O 2 ; temperature, 37°C. KOH in center well. 








ENZYME DISTRIBUTION 


103 



MINUTES 

Fui. 8. Demonstration of TPN-malic enzyme with cytochrome reductase. Re¬ 
duction of cytochrome c determined speetrophotoinetrieally at 550 m/». Corex cell, 
if = i.o cm. Temperature, 25°C. Measurements were made against a blank which 
received all additions except wheat germ enzyme. Optical density corrected for 
changes on dilutions. 0.2 ml. 0.5 M glyeylglycine buffer, pH 7.4; 0.15 n.\l cytochrome 
c; 0.112 mg. cytochrome c reductase; 6.9 mg. wheat germ, Type B; 0.3 ml. 0.5 M l- 
malate; 0.1 mi. 0.01 M MnCl 2 ; and water made to a volume of 3.3 ml. At 0 time, 1.4 
7 TPN added. At 1,321 7 DPN added. At 2,1.4 7 TPN added. At 3,1.4 7 TPN added. 
At 40 min. (not shown on graph) addition of 14 7 TPX caused a change in log h/l of 
0.394. 

Parsley root and parsnip preparations tested likewise gave similar 
results, except that virtually no 0 2 consumption was observed with 
malate and TPN unless either ATP or DPN was added. A typical ex¬ 
periment with parsley root is shown in Fig. 7. It may be noted that the 
enzyme employed in this epxeriment was one which had previously 
given negative results when tested for the malic enzyme by the direct 
spectrophotometric procedure. It is obvious that the malic activity is 
actually present, but that the preparation apparently contains a high 




194 ERIC CONN, BIRGIT VENNESLAND AND L. M. KRAEMER 

concentration of TPN-ase. Results of this type illustrate most clearly 
the questionable nature of negative results for the malic enzyme. Even 
in the presence of ATP, the use of an excess of plant protein may give 
virtually negative results in the manometric test system. For example, 
when the enzyme concentration was increased 5-fold in the experiment 
of Fig. 7, no significant uptake of oxygen was observed. In these 
experiments no attempt was made to correlate quantitatively the malic 
and carboxylase activities in the various samples, nor for that matter 
has there been any effort to compare TPN-ase activities of the different 
preparations. 

Coupling of TPN Malic-Enzyme with Cytochorme c Reductase 

TPN malic enzyme may also be detected spectrophotometrically by 
following reduction of cytochrome c at 550 m/* in the presence of 
cytochrome c reductase (25,26). This method is extremely sensitive and 
requires only very small quantities of TPN. However, because of this 
latter fact, it is particularly susceptible to the TPN-destroying enzymes 
in the plant preparations. Where such factors are not present, reduction 
is readily observed. Where the TPN-ases are present, it is possible to 
overcome their effect by the addition of other nucleotides such as DPN. 
Experiments of this type were run with wheat germ enzyme and with 
parsnip. The results with the wheat germ enzyme are shown in Fig. 8. 
When 1.4 y TPN were added at zero time, no reduction of cytochrome 
c occurred. At time 1, 321 y DPN were added, with no change in light 
absorption. However, when "an additional 1.4 y TPN were added at 
time 2, reduction was observed. After 7 min. an apparent reoxidation of 
cytochrome c 8 seemed to occur. At time 3 another 1.4 y TPN were added 
and reduction again occurred. Muscle adenylic acid gave results similar 
to DPN. 

All these results can be interpreted in a manner similar to that given 
in the section on manometric experiments, i.e., the enzyme prepara¬ 
tions contain a TPN-ase the action of which can be prevented by DPN, 
ATP, and adenylic acid. 

* The nature of this reoxidation lias not been examined. It may bo too rapid to lie 
accounted for entirely as an autoxidation by air. This point requires additional 
investigation. 



ENZYME DISTRIBUTION 


195 


Discussion 

The work reported in this paper was designed to determine whether 
a TPN-specific enzyme catalyzing the reversible oxidative decarboxy¬ 
lation of malic acid is present in a variety of OAA carboxylase prep¬ 
arations available from various plant sources. A need was felt for a 
method which could be widely applied to a group of preparations and, 
if possible, could be used for a quantitative assay. The manometric 
procedure described for detection of the malic activity can be applied 
in this fashion provided care is taken to avoid the destruction of TPN 
by addition of muscle adenylic acid, ATP or DPN. Even in the pres¬ 
ence of added nucleotides, an excess of enzyme must be avoided. 
Care must also be taken, in a quantitative application of the method, 
to determine the ratio of malate oxidation to 0 2 utilization in order to 
interpret correctly the 0 2 uptake in terms of malic enzyme activity. 

The presence of TPN malic enzyme in the 7 plants studied shows that 
the distribution of the enzyme is extensive. No plant /8-carboxylase 
preparation tested to date has been found which does not contain the 
malic-enzyme, provided precautions are taken to prevent TPN de¬ 
struction. The successful demonstration of the activity therefore indi¬ 
cates a close relationship between the dehydrogenation and the de- 
carboxytion such as Ochoa, Mehler and Kornberg (2) have found to be 
the case in pigeon liver. However, insufficient data are at hand to 
warrant the final conclusion that the two activities are associated with 
one protein. 

The widespread distribution of the malic enzyme indicates that a 
large number of cell-free plant preparations are capable of causing the 
synthesis of malate from pyruvate and C0 2 , provided a source of re¬ 
duced TPN is available. Whether radiant energy may effect a reduction 
of TPN through the mediation .of chlorophyll is a question which re¬ 
quires further experimentation. The dependence of some “dark 
reactions” on TPN has been demonstrated. It is not impossible that 
such earlier results as those of Frenkel (27) may be explained in these 
terms. Frenkel found that dark fixation of C0 2 in cell sap of Nitella 
was dependent on intact cell structure. If the TPN were destroyed on 
disintegration of the cells, such a result might be expected, even though 
the proteins capable of catalyzing the reductive carboxylation of 
pyruvate to malate were present. 

Although the TPN-destroying factors encountered in this work were 



196 ERIC CONN, BIRGIT VENNE8LAND AND L. M. KKAEMKK 

largely obstacles to the demonstration of the malic enzyme, their wide¬ 
spread distribution is a fact of interest in itself, and suggests that 
great care must be exercised in interpreting stimulatory effects of such 
substances as adenylic acid on an oxidation system dependent on the 
pyridine nucleotides. The experiments made were not designed to 
determine the mode of action of the TPN-ase or the method whereby 
adenylic acid, ATP or DPN exert their effects. The disappearance of 
the 340 m/i band of reduced TPN might indicate a splitting of the 
nicotinamide-ribose bond. However, nicotinamide was not effective in 
preventing destruction, as it was found to do in the case of DPX- 
nucleosidase described by Mann and Quastel (21) and by Kornbcrg 
and Lindberg (22). On the other hand, the fact that substances having 
in common the structure of adenylic acid can prevent the destruction 
of TPN suggests that inactivation may involve hydrolysis near the 
adenylic acid portion of the molecule. The protective action of ATP 
and DPN may also depend on hydrolysis of these compounds to ade¬ 
nylic acid, which may be the only active agent. Kornberg (28) has de¬ 
scribed a pyrophosphatase which splits ATP, DPN, TPN, and flavine- 
adenine nucleotide. He has observed a protective action of DPN on the 
other 3 compounds and this suggests a similarity to the effects reported 
in this paper. However, the effect of adenylic acid was not reported by 
Kornberg. 

The possibility cannot be excluded that more than one factor are 
involved in the destruction of TPN by the crude plant preparations 
employed. Additional data are required before any conclusions can be 
drawn as to the identity or lack of identity of this destructive factor 
with those enzymes, which have been described by other authors. 

Acknowledgments 

We arc grateful to Dr. E. A. Evans, Jr., for his helpful advice and generous per¬ 
mission to use his facilities. We are also indebted to Dr. T. R. Ilogness for permission 
to use facilities made possible by a grant from the Rockefeller Foundation. 

Summary 

The widespread distribution of factors which destroy TPN has been 
found to constitute the main difficulty in demonstrating the presence 1 
of a TPN-specific malic enzyme in plant tissues. The action of the 
destroying factors can be prevented by ATP, DPN, and adenylic acid. 
These facts have been applied to devise a generally applicable mano- 



ENZYMIC DISTRIBUTION 


197 


metric assay procedure for the detection and the quantitative esti¬ 
mation of the malic enzyme. 

The assay procedure has been used to study the distribution of the 
enzyme. Seven different plant sources (wheat germ, beets, spinach, 
carrots, parsley root, parsnip, and peas) were all found to contain the 
malic enzyme. All of these preparations also contained oxalaeetie 
carboxylase. 

The results suggest a possible close association between the malic 
activity and the carboxylase, and indicate, further, that soluble 
enzymes which can cause the reductive carboxylation of pyruvate to 
malfttc are widely distributed in the tissues (roots, leaves and seeds) of 
higher plants. 

References 

1. Vennesland, B., J. Biol. Chem. 178, 591 (1949). 

2. Ochoa, 8., Mehler, A. H., and Kornberg, A., ibid. 174, 979 (1948). 

3. Vennesland, B., Gollub, M. C., and Speck, ,J. F., ibid. 178, 301 (1949). 

4. Warburg, O., and Christian, W., Biochem. Z. 266, 377 (1933). 

5. Theorell, II., ibid. 278, 263 (1935). 

6. Warburg, 0., Christian, W., and Griese, A., ibid. 282, 157 (1935). 

7. Mathews, M. B., J. Biol. Chem. 176, 229 (1948). 

8. Warburg, O., in Nord, F. F., and W ei den hag e n , R., Ergeb. Enzumforach . 7, 

210 (1938). 

9. Adler, E., and Sreenivasaya, M., Z. physiol. Chem. 249, 24 (1937). 

10. Levene, 1\ A., and Raymond, A. L.,./. Biol. Chem. 92, 757 (1931). 

11. Dixon, M., Manometric Methods, 2nd Edit., 62. Cambridge, 1943. 

12. Friedemann, T. E., and IIaugen, G. E., J . Biol. Chem. 147, 115 (1943). 

13. Westerkamp, H., Biochem. Z. 263, 239 (1933). 

14. Speck, J. F., manuscript. 

15. Adler, E., von Euler, II., Gunther, G., and Plass, M., Biochem. J. 33, 1028 

(1939). 

16. Avery, 0. T., and Morgan, H. J., J. Exptl. Med. 39, 275 (1924). 

17. Keilin, D., and Hartree, E. F., Biochem. J. 39, 293 (1945). 

18. Chance, B., Acta Chem. Bcand. 1, 236 (1947). Arch. Biochem. 21, 416 (1949). 

19. Sevag, M. G., Ann. Chem. 607, 92 (1933). 

20. Mehler, A. Ii., Kornberg, A., Grisolia, S., and Ochoa, S., J . Biol. Chem. 174, 

961 (1948). 

21. Mann, P. J. G., and Quastel, J. H., Biochem. J. 36, 502 (1941). 

22. Kornberg, A., and Lindberg, 0., J . Biol. Chem. 176, 665 (1948). 

23. Ceithaml, J., and Vennesland, B., ibid. 178, 133 (1949). 

24. Evans, E. A., Jr., Vennesland, B., and Slotin, L., ibid. 147, 771 (1943 >. 

25. Haas, E., Horecker, B. L., and IIogness, T. R., ibid. 136, 747 (1940). 

26. Haas, E., Harrer, C. J., and Hogness, T. R., ibid . 142, 835 (1942). 

27. Frenkel, A. W., Plant Physiol 16, 654 (1941). 

28. Kornberg, A., J . Biol. Chem. 174, 1051 (1948). 



The Genesis of Auxin during the 
Decomposition of Proteins 

Victor Schocken 1 

From the WiUiarn G. Kerchojf Laboratories of the Biological Sciences, 
California Institute of Technology, Pasadena 4, Calif. 

Received March 21, 1949 

Introduction 

The terra “auxin protein” has been applied to a protein isolated from 
spinach leaves, which upon alkaline hydrolysis yielded about 0.02-0.09 
y of auxin, calculated as indoleacetic acid/mg. of the protein (1, 8). As 
shown below, the auxin appears indeed to be indoleacetic acid but 
amounts to far less than one molecule per molecule of protein (mol. 
wt. ca. 2 X 10 6 ). In no case was the indoleacetic acid obtained without 
hydrolyzing the protein. On the contrary, in order to produce the 
auxin, the protein was subjected to treatments of many hours in 0.1 N 
NaOH at 100°C., a procedure which has been shown to give comparable 
quantities of indoleacetic acid from a variety of proteins not in any way 
related to plants and not having any enzymatic properties. Despite the 
small quantities of auxin obtained, the drastic treatment of the protein 
required to obtain them, and the non-specificity of the reaction, 
Wildman and Bonner (8) suggest that “a reasonable working hypoth¬ 
esis may be that the auxin of the protein is related in some way to the 
specific enzymatic activity of the material,” and that “probably the 
bound auxin of Fraction I [the auxin protein] should not be regarded 
as a precursor or storage form of auxin, but rather as the biochemically 
active form of the growth substance.” 

The following experiments were carried out to test this hypothesis 
and, in general, to provide a background for the appraisal of any 
possible physiological or chemical significance attributable to the 
appearance of a minute quantity of indoleacetic acid in the hydrolyzate 
of a protein. 

1 Present address: National Cancer Institute, Bethesda 14, Md. 

198 



AUXIN 'GENESIS 


199 


To determine the specificity of the reaction giving rise to the indole- 
acetic acid, a group of proteins from widely varying sources was sub¬ 
jected to the treatment found optimum for obtaining indoleacetic acid 
from the spinach “auxin protein.” This investigation reveals that the 
property of yielding small quantities of indoleacetic acid upon alkaline 
hydrolysis is by no means specific to any one enzyme, or to any one 
protein, nor is it specific to proteins of botanical origin, but, on the 
contrary, seems to be a widespread property of tryptophan-containing 
proteins of whatever origin or function. 

Materials and Methods 

The quantities of auxin occurring in the hydrolyzates of the proteins were deter¬ 
mined by means of t he standard Avena test as described by Went and Thimann (7). 
Each test was standardized by including a sample consisting of a solution of indole¬ 
acetic acid of known concentration. The curvatures produced by the samples being 
tested were compared with the curvatures produced by the known concentrations of 
indoleacetic acid, and the activities of t he samples are expressed in mg.-equivalents of 
indoleacetic acid. The assumption that the Avena curvatures produced by the extracts 
of the protein hydrolyzates are due to indoleacetic acid rather than some other auxin 
is based on the observations that the active compound is an acid, that it is relatively 
stable to boiling 0.1 N NaOH, and particularly, that it gives a positive test in the 
modified Salkowski reaction described by Tang and Bonner (6) as specific to indole¬ 
acetic acid. 

Efforts were made to obtain indoleacetic acid or any other auxin from the proteins 
particularly the spinach Fraction I, by treatments which should leave the protein 
intact. Among the methods tried were prolonged dialysis against distilled water and 
various buffers, precipitation at the isoelectric point, washes with acids, bases, and 
organic solvents. None of these was successful, but rather confirmed the conclusion 
that hydrolysis of the protein is a necessary condition for obtaining any auxin from it. 
The results here reported are those obtained by hydrolyzing the proteins in dilute 
NaOH solutions, a procedure also used by Wildman and Bonner (8). 

To assay the auxin in such a hydrolyzate by the Avena method it Ls brought to pH 3 
by the addition of HC1, and is extracted with ether. The extract is concentrated by 
boiling off the ether until a volume of about 1 ml. remains. This 1 ml. is then quanti¬ 
tatively transferred into a vial containing 0.8 ml. of hot- agar, whereupon the ether 
evaporates and leaves the dissolved material in the agar. The agar, which now con¬ 
tains any auxin which may have been present in the protein hydrolyzate, is serially 
diluted so that at least one concentration will be of the proper magnitude for the 
assay. Each dilution is then poured into a mold and cut into 12 equal-sized blocks 
which are used in the Avena test as described above. 

The proteins investigated were the following: 

Spinach Cytoplasmic Protein (Fradion /). This is the “auxin protein” of Wildman 
and Bonner (8), the preparation of which they have described. The samples used were 
generously supplied by Dr. Wildman or prepared under his supervision. 



200 


VICTOR SCHOCKKN 


Ovalbumin. A thrice-recrystallized sample prepared by the method of Keckwick and 
Cannon (4) and supplied through the courtesy of Dr. George Feigen. 

fi-Ladoglobulin . A twice-recrystallized sample prepared by the method of Palmer 
(5) and obtained through the courtesy of Dr. John Cushing. 

Chymotryp8in. A crystalline preparation obtained from the Plaut Research Labora¬ 
tory, Lehn and Fink Products Corp., Bloomfield, N. J. 

Fibrin . “Difco, Desiccated” obtained from Difco Laboratories, Inc., Detroit, Mich. 

Serum Albumin. A crystalline sample obtained from Professor Norman Horowitz. 

Experimental 

The optimum conditions for obtaining auxin from proteins upon alkaline hydrolysis 
were investigated, using the spinach Fraction I. In all of these experiments the reac¬ 
tion mixtures were 5 ml. in volume, and the normality of base specified refers to the 
final concentration in this volume. The reactions were carried out in 30 ml. test tubes 
equipped with air condensers, and heated in a boiling water bath. From Fig. 1, 



Fig. 1 . Auxin yields, calculated as indoleacctic acid equivalents, obtained from 
spinach Fraction I, the “auxin-protein,” upon refluxing at 100°C. with NaOH of 
various concentrations. 

showing the yields obtained with various concentrations of NaOH, it is evident that 
0.1 AT is the most favorable concentration of alkali for obtaining auxin from the 
spinach protein, when the hydrolysis is carried out at 100°C. 

These conditions were then applied to other proteins. A comparison of the rates of 
appearance and maximum yields of auxin in the hydrolyzates of various proteins is 
shown in Fig. 2. Naturally, not all proteins could be expected to react in the same way 
to a given set of conditions. The experiment shows, however, that the reaction giving 
rise to the auxin is by no means unique to the spinach Fraction I but is displayed by 
many other proteins when they are similarly treated. Positive Avena tests were also 
obtained when the protein of Avena coleoptiles, trypsin, and zein (a sample of un- 




AUXIN GENESIS 


201 



_L_.1_l_L_ 

19 90 43 SO 

T1MC Of HUTINQ IN HOOKS 


Fig. 2 . Auxin yields, calculated as indoleacetic acid equivalents, obtained from 
several proteins upon refluxing at 100°C. with 0.1 N NaOH. 

known purity) were treated, although the investigation of these proteins was not, 
extensive enough to determine their yields with certainty. Gelatin, which contains no 
tryptophan, and pepsin yielded no auxin. 

Since a plausible explanation for the appearance of trace amounts of auxin in the 
hydrolysates of the proteins studied seemed to be that they arise through the oxida¬ 
tive deamination of tryptophan to indoleacetic acid, the abundance of this amino 
acid in the proteins was compared with their yields of auxin. For this purpose the 
tryptophan content of the spinach Fraction I was determined by the Adamkiewicz- 
Hopkins method (2), while values for the other proteins were taken from the literature. 
In Table I are listed the tryptophan contents and auxin yields of the various proteins. 



Fio. 3. Auxin yields; calculated as indoleacetic acid equivalents, obtained upon 
refluxing Fraction I at 100°C. in 0.1 N NaOH, with and without added tryptophan 
(1 mg.). 





202 


VICTOR SCHOCKEN 


TABI.K I 

Tryptophan Content atul A uxin Vields of Various Proteins 


Maximum IAA equivalents 
obtained on heating in 0.1 S 


Piotein 

Tryptophan 
per cent 

NaOII at 100°C. 

Chymotrypsin 

5.5 

0.2 X 10 ~ 2 7 /mg. protein 

Fraction I 

4.1 

8.5 

Fibrin 

3.0 

5.8 

0 -Iiaetoglobulin 

1.9 

3.2 

Ovalbumin 

1.3 

5.4 

Serum albumin 

0.2 

0.1 

Gelatin 

0.0 

0 (< 0 . 0002 ) 


Further to test the hypothesis that, it is tryptophan which gives rise to the auxin in 
the protein hydrolyzates, tryptophan was subjected to the conditions of the protein 
hydrolysis. The results of these experiments, presented in Fig. 3, show that trypto¬ 
phan is indeed converted to some extent to indoleacetie acid under the conditions used 
to obtain auxin from the proteins. 


Discussion 

Inasmuch as the capacity to yield small amounts of auxin proves to 
he commonplace among tryptophan-containing proteins,„ and in fact 
appears to be roughly proportional to the tryptophan contents of the 
proteins, it is interesting to consider whether this auxin could be indole- 
acetic acid originating from the tryptophan. Significant in this consid¬ 
eration is the fact that, under the conditions of the hydrolysis used to 
obtain indoleacetie acid from the proteins, tryptophan is released into 
a warm alkaline reaction medium. It is known that, under these condi¬ 
tions, tryptophan undergoes oxidative deamination to indoleacetie 
acid (3). It is not surprising, therefore, that tryptophan treated along 
with the protein in 0.1 N NaOH at 100°C. contributes to the yield of 
auxin acid. The results of such an experiment as presented in Fig. 3 are, 
however, difficult to interpret quantitatively for the following reason. 
I’lie free tryptophan is exposed to the warm alkali from the beginning 
of the experiment. Whatever indoleacetie acid is produced appears 
early and is then exposed to the destructive action of the reaction 
medium. (Note how the yields fall off with time in Fig. 2.) The tryp¬ 
tophan from the protein, on the other hand, is only.slowly released into 
the medium and, therefore, whatever indoleacetie'acid is formed from 
it appears later. The indoleacetie acid from the added tryptophan is 



AUXIN GENESIS 


203 


exposed to the destructive action of the reaction medium for a longer 
period of time than the indoleacctic acid from the peptide tryptophan. 
The yields cannot, therefore, be compared directly, and, moreover, if 
suitably small quantities of tryptophan are added, and long periods of 
digestion used, then the effect of the added tryptophan can be lost 
entirely. 

That the auxin found in the hydrolyzates of the proteins investigated 
arises through the oxidative deamination of tryptophan is offered as 
an hypothesis consistent with the observations. The fact remains, how¬ 
ever, that, whatever the source of the auxin may be, the same treat¬ 
ment produces eomaprable quantities in the hydrolyzates of such wide¬ 
ly varying proteins as chymotrypsin, fibrin, /3-lactoglobulin, ovalbumin 
and serum albumin, in addition to the spinach Fraction I. This fact 
makes it appear unwarranted to attribute any physiological signifi¬ 
cance to the auxin obtainable from the spinach protein which cannot 
also be assigned to the auxin obtainable from the other proteins. This 
type of so-called “bound auxin” would, therefore, appear to be associ¬ 
ated neither with any special enzymatic activity of the protein, nor 
specifically with plant growth or plants. 

Acknowledgments 

Thanks arc due Miss Marl ha Kent for her assistance in many of the experiments 
reported, and to Professor James Bonner and Dr. Sam Wildman for their cooperation. 


Summary 

1. The treatment found to give optimum yields of auxin from spin¬ 
ach Fraction I, the “auxin protein,” was refluxing at 100°C. in 0.1 N 
NaOII. 

2. Chymotrypsin, fibrin, /?-lactoglobulin, ovalbumin, scrum albumin 
and other proteins all yielded an auxin believed to be indolcacetic acid 
when subjected to this treatment. 

3. The yields of auxin from the protein hydrolyzates were roughly 
proportional to the tryptophan contents of the proteins. 

4. When tryptophan was added to the reaction medium, the yield of 
indoleacetic acid was increased. 

5. It is suggested that the auxin from the proteins arises through 
oxidative deamination of tyrptophan to indoleacetic acid in the hot 
alkaline medium in which the proteins are hydrolyzed. 



204 


VICTOR SCHOCKEN 


References 

1. Bonner, J., and Wildman, S., 6th Growth Symp ., 51 (1947). 

2. Block, R. J., and Bolling. D., The Determination of the Amino Acids, Burgess 

Pub. Co., Minneapolis, Minn., 1940. 

3. Gordon, S., and Wildman, S., J . Biol. Chem. 147, 389 (1943). 

4. Keckwick, R. A., and Canon, R. E., Biochem. J. 30, 232 (1936). 

5. Palmer, A. H., J. Biol. Chem. 104, 359 (1934). 

6. Tang, Y., and Bonner, J., Arch. Biochem. 13, 11 (1947). 

7. Went, F. W., and Thimann, K. V., Phytohormones, Macmillan, N. Y., 1937. 

8. Wildman, S., and Bonner, J., Arch. Biochem. 14, 381 (1947). 



Antiscorbutic Substances. 3-Methyl-L-Ascorbic Acid 
and 1-Methylheteroascorbic Acid 

Bernard S. Gould, Henry M. Goldman and John T. Clarke, Jr. 

* From the Department of Biology, Massachusetts Institute of Technology, anil the 
Department of Dental Research, Beth Israel Hospital, Boston 
Received March 25, 1940 

Introduction 

It is generally accepted that, for an analog of ascorbic acid to have 
antiscorbutic activity, it is essential that, among other things, all the 
hydroxyl groups in the molecule must be free. 

However, 3-methyl-L-ascorbic acid appears to show, on the basis of 
relatively non-specific preventive and curative methods, with growth 
only as the criterion (1,2) as much as 1/10 of the activity of L-ascorbic 
acid. The responses to the elevated doses of the compound are, how¬ 
ever, not typical of those observed with ascorbic acid, and it was sug¬ 
gested that preliminary in vivo conversion to ascorbic acid is involved. 

Previous investigations (3,4,5) have established that a critical ascorbic acid intake 
(0.225 mg./day) is necessary for normal osteoblastic function in the guinea pig and the 
“alkaline” phosphatase level of the blood has been shown to be the most sensitive and 
earliest index of this function. The restoration of normal serum phosphatase levels 
upon the administration of a critical dose of ascorbic acid to scorbutic animals is 
strikingly prompt (Figs. 1 and 2) and can be used as a measure of antiscorbut ic activity 
(3). Responses that occur only after prolonged administration would be suggestive of 
the necessity for preliminary conversion of the compound to ascorbic acid. 

Using the phosphatase bioassay, studies were made to determine (1) 
the exact antiscorbutic potency of 3-mcthyl-L-ascorbic acid, and (2) 
whether preliminary conversion to L-ascorbic acid appears to be 
involved; and using the classical histological technique, to determine 
unequivocally whether the compound shows (1) true antiscorbutic 
properties, and (2) whether it may act competitively with L-ascorbic 
acid. 


205 



206 


B. S. GOIJLD, II. M. GOLDMAN AND J. T. CLARKE, JR. 



Fig. 1 Fig. 2 

Fig. 1 . Serum phosphatase responses of animals maintained on a scorbutogenic 
diet for 24 days, then fed various supplements for 5 days. 

Curve 1, animals fed 0.225 mg. ascorbic acid; Curve 2, 6.75 mg. 3-met hyl-L-ascorbic 
acid; Curve 3, 11.25 mg. 3-methyl-L-ascorbic acid; Curve 4, 1 mg. 1-methylhetero- 
ascorbic acid; and Curve 5, fed 0.225 mg. ascorbic acid throughout the preparative and 
test periods. 

Fig. 2. Typical serum phosphatase responses of animals fed ascorbic acid, 1- 
mcthylheteroascorbic acid and 3-methyl-L-a8corbic acid. 

Curve 1, animals fed a scorbutogenic diet for 25 days then supplemented by 0.225 
mg. ascorbic acid for 10 days; Curve 2, fed the diet supplemented by 0.25 mg. ascorbic 
acid throughout; Curve 3, fed the diet supplemented by 11.25 mg. 3-methyl-L-ascorbic 
acid for 25 days; Curve 4, fed the diet supplemented by 1 mg. 1-methylheteroascorbic 
acid; Curve 5, fed the diet supplemented by 5.6 mg. 3-methyl-L-ascorbic acid for 23 
days followed by 11.25 mg. (intraperitoncally) for 13 days; and Curve 6, supplemented 
by 2.25 mg. 3-mcthyl-i/-ascorbic acid for 23 days followed by 11.25 mg. (intraperi- 
toneally) for 13 days. 

Experimental Methods 

lhe methods for the maintenance and treatment of the animals 1 as well as (he 
estimation of phosphatase are those described in previous studies (3,4,5,7). Test- 
animals invariably showed evidence of scurvy and characteristically low serum 
phosphatase; levels in 18-25 days. Control animals responded characteristically to the 

1 We are extremely grateful to Mr. Leonard Stutman for considerable assistance 
in the care and preparation of the animals. 




ANTISCORBUTICS 


207 


critical level of 0.225 mg. ascorbic acid daily after 5 days administration and, at levels 
below this, there was no phosphatase response. 


Histological Methods 

The midsagittal section of the head of each animal was cut in transverse sections at 
predetermined points at different levels of the incisor teeth. The decalcified (5% 
UNO*) sections were cut, mounted and stained with hematoxylin and eosin. Some 
sections were also stained by the Wilder’s silver-stain method. 

Synthesis of 3-Mcthyl-h-Ascorbic Acid and of 

* 1 -Mcthylhctcroacsorbic A cid. 

In addition to one sample of 3-mcthyl-L-aseorbic acid kindly supplied by Dr. (\ S. 
Vestling and another by Dr. M. C. Rebstock, to whom we express our sincere thanks, 
both 3-methyl-L-ascorbic acid and 1-methylhcteroaseorbic acid were prepared by a 
slight modification of the method used by Vestling and Rebstock (2). The preparation 
showed no depression of the melting point when admixed with the sample obtained 
from Dr. Vestling (m.p. 124-125°C.) and contained no measurable amount of ascorbic 
acid, as indicated by a completely blank titration with 2,6-dichlorophenol indophenol 
reagent. 

The 1-methylhcteroascorbic acid obtained was recrystallizcd frpm hot nitromethane 
and yielded a crop of orange blocks which melted at 159-161°C. (with decomposition). 
Haworth’s preparation (6) melted at 162°C. 

Results 

3-Mcthyl-L”Ascorbic Acid 

From the results shown in Fig. 2 it is evident that 3-methyl-L- 
aseorbic acid fed to the extent of 30 times the critical dose of L-ascorbic 
acid showed no protective action. On the other hand, when 50 times the 
critical dose was fed, most of the animals were protected, as shown by 
histologica} examination as well as by the phosphatase level of the 
serum, indicating that the compound is l/50th as active as L-ascorbic 
acid. However, the results of curative tests show a striking difference 
from the response anticipated if 3-methyl-L-ascorbic acid were merely 
l/50th as active as L-ascorbic acid. The histological and phosphatase 
responses obtained upon the administration of various analogs of 
ascorbic acid (5), while quantitatively different, are essentially the 
same when the quantities administered are adjusted to be equivalent to 
the desired amount of ascorbic acid, in that the effects are evident in as 
short a time as 24 hr. and invariably, in less than 5 days (Fig. 2). The 
results obtained here indicate a markedly delayed response, both in 



208 B. S. GOULD, H. M. GOLDMAN AND J. T. CLARKE, JK. 

serum phosphatase elevation and in histological tooth structure repair, 
when 3-methyl-L-ascorbic acid is administered. Even when 50 times 
the critical dose of L-ascorbic acid is administered intraperitoneally, 
there was no significant response after 5 or even 8 days, but, if the 
compound is administered over a prolonged period of 12-13 days, the 
response is apparent (Fig. 2). It would appear, therefore, that prelimin¬ 
ary conversion of the methylated compound to L-ascorbic acid is 
essential before activity is manifested. 

1 -Melliyllietcroascorbic . I cid 

Using doses of 1 mg. per day (about 4 times the critical dose of 
L-ascorbic acid), preventive and curat ive experiments indicate that the 

TAIJLK I 

» 

Histological Examination of Some Typical Animals Fed Ascorbic Acid , 
d-Methyl-L-A scorbic A cid, l-Methylhcteroascorbic A cid , 
or Mixtures of the Compounds 


Animal 

Treatment* 

PliOHplmtase response 

IfiMtoloRicnl response'’ 

24 

j 

0.5 mg. A. A. for 23 d. 

Phosphatase, 0.5 units 

] 

No indication of scor¬ 
butic lesions 

10 

2.25 mg. MAA for 23 d. 

Phosphatase fell to 3.S 
units 

Very scorbutic 

it 

2.25 mg. MAA for 23 d., 
then 11.25 mg. for 13 d. 

Phosphatase at 23 d. 
fell to 1.7 units, then 
rose to final value of 
16.5 

Scorbutic, followed by 
very mild repair, then 
very rapid repair 

14 

5.6 mg. MAA for 23 d. 

Phosphatase fell to 2.6 
units 

Scorbutic, complete ces¬ 
sation of dentine forma¬ 
tion 

15 

Same as No. 14, then 
given 11.25 mg. for 12 
days 

■ Phosphatase fell to 4.2 
units at 23 d., then 
rose after 12 d. to 11.5 
units 

Scorbutic, followed by 
very mild repair, then 
very rapid repair 

18 

11.25 mg. MAA for 26 d. 

Phosphatase, 12.5 units 

Subacute scorbutic con¬ 
dition still laying down 
irregular dentine 







ANTISCORBUTICS 


209 


TABLE I ( Continued ) 


Vniuuil 

Treatment 0 

PhoHphatane resixmse 

Histological reHponHe'' 

19 

11.25 mg. MAA for 26 d. 

Phosphatase, 9.5 units 

Mild scorbutic condition 

20 

11.25 mg. MAA for 23 d. 

Phosphatase, 5.5 units 

Mild scorbutic condition 

21 

11.25 mg. MAA for 23 d. 

Phosphatase, 6.1 units 

Mild scorbutic condition 
no repair 

33 

Scorbutogenic diet for 

23 d., then 11.25 mg. 
MAA for 5 d. 

Phosphatase at 23 d., 
5.8 units; at 28 d., 6.4 
units 

Very scorbutic, no signi¬ 
ficant repair 

30 

Scorbutogenic diet for 

23 d., then 1 mg. IIA A 
for 5 d. 

Phosphatase fell to 2.2 
units at 23 d., then rose 
l to 7.8 units at 28 d. 

Markedly scorbutic fol¬ 
lowed by good repair 

22 

11.25 mg. MAA plus 0.5 
mg A A for 25 days 

Phosphatase, 8.5 units 

i 

i 

| 

No indication of scor¬ 
butic condition. Same 
as animal 24 

9 

Scorbutogenic diet for 

18 d., the 0.25 mg. A A 
for 5 days 

Phosphatase fell to 3.6 
units at 18 d., then 
rose to 10.1 units at 

23 days 

Very scorbutic, then 
mild repair 


" A A = L-ascorbic acid. 

MAA = 3-Mot hyl-L-ascorbic acid. 

HAA = 1-Mot hylheteroascorbic acid. 
h A dotailod histological description will be reported separately. 

response is like that of L-ascorbic acid and unlike that of 3-methyl-L- 
ascorbic acid, in that there is ho delay in the manifestation of its 
antiscorbutic activity (Fig. 2), suggesting rapid demethylation to active 
L-ascorbic, acid. 

Competitive Action between 3-Methylascorbic Acid and l- Ascorbic acid 

Experiments were carried out in which animals were fed 11.25 mg. of 
3-methyl-L-ascorbic acid and 0.5 mg. of L-ascorbic acid. There was no 
indication of an antagonistic effect between the two compounds 
(Table I) under these conditions. 




210 B. S. GOULD, H. M. GOLDMAN AND J. T. CLARKE, JR. 


Summary and Conclusions 

1. Bioassays employing (1) the serum phosphatase response and (2) 
the histological tooth structure examination indicate that 3-methyl- 
L-ascorbic acid is only l/50th as active antiscorbutically as L-ascorbic 
acid. 1-Methylheteraoscorbic acid shows strong antiscorbutic activity. 

2. Both the blood phosphatase and histological responses to the 
administration of large doses of 3-methyl-L-ascorbic acid are markedly 
delayed in contrast to the prompt responses observed with L-ascorbic 
acid; suggesting that preliminary conversion of 3-methyl-L-ascorbic 
acid to L-ascorbic acid probably occurs. 

3. There appears to be no biochemical antagonism between 3- 
methyl-L-ascorbic acid and L-ascorbic acid when mixtures of the two 
are fed, the former in relatively large and the latter in relatively small 
amounts. 


References 

1. Vestling, C. S. f and Rebstock, M. C., Federation Proc. 4, 108 (1945). 

2 . Vestling, C. S., and Rebstock, M. C., J. Biol. Chem. 164, 631 (1946). 

3. Gould, B. S., and Schwachman, H., ibid. 161, 439 (1943). 

4. Shwachman, H., and Gould, B. S., J. Nutrition 23, 271 (1942). 

5. Gould, B. S., and Shwachman, II., Am. J. Physiol. 136, 485 (1942). 

6 . Haworth, W. N., Hirst, E. L., Smith, F., and Wilson, W. J., J. Chem. Soc. 1937, 

829. 

7. Gould, B. S., Arch. Biochem. 19, 1 (1948). 



Chemical Composition of Normal Bone Marrow 

Albert A. Dietz 

From the Toledo Hospital Institute of Medical Research , Toledo 6, Ohio 
Received March 29, 1949 

Introduction 

In a previous paper, correlations were shown between the water, 
lipide, residue, total nitrogen, and lipide nitrogen components of 
normal rabbit marrow (1). In the present paper, similar relationships 
are shown for the marrow of other animals. In addition, a study was 
made of the sulfur, nitrogen, and phosphorus distribution in marrow. A 
number of isolated results may be found in the literature. Singher and 
Marinelli (2) stated that rat marrow had a high sulfur content, but 
they gave no values. A value of 25-30 mg.-% of acid-soluble phosphorus 
was recorded by Lutwak-Mann (3) for rabbit marrow. The nitrogen 
components of normal and anemic rat marrow, were determined by 
McCoy and Schultze (4). An attempt is made, in this paper, to obtain 
a broader picture of the composition of bone marrow, so that deviations 
from normal can be related to alterations in hemopoietic function. 

Experimental 

Bone marrow was obtained from rabbits, rats, and guinea pigs of all ages, from a 
4 month old cat, a 2 month old beef calf, a frog (Ram pipiens ), hogs, bulls, and dogs. 
The cat and dogs were killed by nembutal anesthesia, and the other animals by 
exsanguination following a blow at the base of the skull. With the exception of the 
hog, and a few samples of beef marrow obtained at a slaughter house, all marrow 
samples were removed immediately after death. The method of analysis for the gross 
components, in marrow taken from different bones, was the same as previously 
described (1), with the exception that the fat was extracted once with petroleum ether 
as well as 3-4 times with alcohol-ether (3:1). For rabbit marrow the petroleum ether 
was not necessary, but for some of the other samples, especially beef, the bulk of the 
fat did not dissolve readily in the alcohol-ether mixture. For the smaller animals, 
the marrow from 2 or more bones had to be grouped to obtain adequate samples. The 
data were subjected to correlation analysis, as previously described (1). The symbols 
W, L, R f N } and Nl are again used to represent percentage concentrations of the 


211 



212 


ALBERT A. DIETZ 


TABLE I 

Composition of Femur Marrow 


Animal 

Arc months 

Wat or 

Lipide 

Residue 

Nitrogen 

Lipide nitro¬ 
gen per cent 
of lipide 

Guinea pig 

0.8 

per cent 

70.4 

per cent 
. 12.1 

per rent 

17.5 

per cent 

2.54 

0.87 

Guinea pig 

24. 

70.8 

10.4 

18.8 

2.73 

1.38 

Hat 

1.8 

73.3 

5.0 

21.1 

3.06 

2.13 

Hat 

26. 

07.7 

14.0 

18.3 

2.75 

0.81 

Cat 

! 4 - 

; oo.i 

38.2 j 

! 11.7 

1.84 

0.33 

Beef 

! 1.5 

| 41.3 ! 

1 47.7 1 

11.0 i 

i 1.51 

0.18 

Beef 

>50. 

0.2 

88.2 | 

2.0 ! 

! 0.23 

0.03 

Hog 

>12. 

8.1 

! oo.i ! 

1.8 

0.20 1 

0.03 

Rabbit 

1.2 

73.0 | 

12.5 

13.0 

2.00 

0.57 

Rabbit 

24. 

30.0 1 

55.7 

8.3 1 

1.21 

0.00 

Dog 

>24. 

20.8 

72.1 

7.1 

0.87 

0.10 

Frog 

18. 

05.4 

28.4 

0.2 

0.87 

0.11 


water, lipide, residue (lipide-free solids), and nitrogen in t he whole bone marrow, and 
lipide nitrogen as per cent of the lipido, respectively. 

In addition to the gross analyses, protein-free filtrates of certain marrows were 
prepared. The marrow was ground in a mortar to give a homogeneous mixture and an 
aliquot of 2-2.5 g. was weighed into a 25 ml. volumetric flask. Ten ml. of water were* 
added and the mixture allowed to stand at room temperature for 1 hr. to hemolyze 
the cells. At this time, 12.5 ml. of 20% trichloroacetic acid were added and made up to 
volume with water. The protein-free filtrate was obtained and analyzed for non-pro¬ 
tein nitrogen by the Kjeldahl method, inorganic phosphate by the method of Tisdall 
(5) as modified by Kaucher et at. (6), and the distribution of sulfur. The method used 



Fig. 1 . Composition of normal bone marrow plotted on triangular coordinates. 
The values given in the insert are for the equation, W « ah +' bR, with the standard 
error of estimate of the equation. The line represents the composition of marrow of 
rabbits 1-24 months of age (1). 








BONK MAKKOW 


213 


for the distribution of sulfur was essentially that of Marenzi at al. (7), removing the 
phosphate with zirconium oxychloride at a pH of 9.0 ± 0.2 and isolating the benzi¬ 
dine sulfate. The sulfate, however, was titrated by the method of Cope ( 8 ). The accu¬ 
racy of this method was the same as found by Power and Wakefield (9). The total 
sulfur and phosphorus contents of the marrow were determined on another aliquot 
of 0 .1-0.4 g. A micro modification of the HN0 3 and HC10 4 oxidation of Evans and St. 
John (10) was used, and sulfate and phosphate determined as described above. About 
95% of the sulfur in 5 mg. of methionine was recovered following oxidation wilh 3.5 
ml. HNOa and 1 ml. HCIO4. A third aliquot was analyzed for the major components 
as previously described ( 1 ). 



Fig. 2 . Correlation of lipide and water in normal bone marrow. The values in the 
insert are for the equation, L = a + (b ± 07 ,) W with a standard error of estimate, 
± <r, and the correlation coefficient, r. The; shaded area represents the composition of 
normal rabbit marrow, ± a ( 1 ). 




214 


ALBERT A. DIETZ 


Results 

The plotting of the values obtained for the major bone marrow con¬ 
stituents showed a correlation in all animals. Some differences were 
noted from the values previously recorded for rabbit marrow (1). The 
chief difference was in the limits of the compositions. The marrow of the 
small animals, guinea pigs and rats, was distributed at the active end of 
the graphs, and the variations with age were less marked. Analyses 
representing samples of marrow from the femurs of the different animals 



10 

% RESIDUE 


Fig. 3. Correlation of water and residue in normal bone marrow. The values in the 
insert are for the equation W = a + (b ± <rt)R with a standard error of estimate, 
± < r , and the correlation coefficient, r. The shaded area represents the composition of 
normal rabbit marrow, ± a (1). 


are shown in Table I. With the larger animals, the marrow from the long 
bones showed^a considerable decrease in activity with age, and all 
samples showed a distribution towards the inactive end of the scale. 
The differences are further shown in Figs. 1-4, where, for comparison, 
the line in Fig. 1 and the shaded areas in Figs. 2-4 represent the compo¬ 
sition/)! normal rabbit marrow (1). The guinea pig marrow shows a tend¬ 
ency to deviate from the line given for the rabbit marrow, but this 




BONE MAKROW 


215 


deviation is hardly significant. The deviation of the rat marrow is more 
significant. In both cases, less water is associated with a unit weight of 
residue. In the values given in the legend of Fig. 1, a and b represent the 
g. II 2 0 associated with each gram of lipide and residue, respectively. 
The guinea pig, rat, and beef marrows have a significantly smaller 



% LIPIDE 

Fig. 4. Correlation of residue and lipide in normal bone marrow. The values in the 
insert are for the equation, R = a + (ft ± n)L, with a standard error of estimate, 
± <r, and the correlation coefficient, r. The shaded area represents the composition of 
normal rabbit marrow, ± a (1). 

amount of water associated with the residue than in the case of the 
rabbit marrow. In no case was water associated with the lipide as the 
values for a in Fig. 1 are not significantly different from zero. The frog 
marrow differed from all the others. An insufficient number of samples 
were available to permit correlation analysis, but the points are so 



216 


ALBERT A. DIETZ 


located as to indicate a dilution with water. The points in Fig. 2 indi¬ 
cate that the lipide and water components of the frog marrow vary in¬ 
versely with one another, but the other analyses tend to show a rela¬ 
tively constant residue and nitrogen content. 

Correlation of Nitrogen Components 

The residue showed a positive correlation with the nitrogen content 
of the marrow, and, in all cases, the regression lines fell within one 
standard deviation of the line previously given for normal rabbit 
marrow (1), and so are not included in this paper. The lines differed 
from one another in range of concentrations only. 

The regressions of the lipide with the log of the lipide nitrogen 
showed a negative correlation in all cases (Fig. 5). Thus, there is rela¬ 
tively less nitrogen associated with the lipide in the inactive marrows. 
The hog and guinea pig marrows showed a slightly better correlation 
when the log of the lipide concentration was correlated Avith the log of 
the lipide nitrogen, but not sufficiently better to include another graph. 
As shown by the values in the legend of Fig. 5 and the placement of the 
points, the rat and guinea pig marroAVs show significant differences in 
both the slopes and limits of the regression line from rabbit marrow. 

The results of the analyses for various sulfur and phosphorus com¬ 
ponents and for non-protein nitrogen could not be correlated readily 
with the activity of the marrow as expressed by its water, lipide, and 
residue content. Within a given animal there was usually an increase 
in inorganic sulfate and phosphate, total sulfur and phosphorus, and 
non-protein nitrogen with an increased activity of the marrow. Be¬ 
tween the various rabbits there was, however, a larger variation. 
Representative analyses are given in Table II. The lipide concentration 
may be obtained by subtracting the percentages of water and residue 
from 100. 

Using the values found in this table, and calculating the results on a 
lipide-frec or lipide-free-solid basis, somewhat more constant values 
can be obtained for certain of the constituents, but marked differences 
are still found between the various samples. On the lipide-free basis, 
the total nitrogen, sulfur, and phosphorus for 9 samples of active 
marrow from the humerus, femur (and proximal tibia) are: 2.90 
(2.45-3.22), 0.28 (0.21-0.41), and 0.37 (0.20-0.52)%, respectively. 
On the lipide-free-solid basis the values are: 14.5 (12.8-15.7), 1.40 
(1.28-1.85), and 1.83 (1.18-2.48)%, respectively. 



BONE MARROW 


217 


A calculation of distribution of sulfur shows that 73.6 (68.3-83.0)% 
is present in the proteins. In the non-protein fraction the average 
distribution of sulfur is 50.6% inorganic sulfate, 6.5%, ethereal sulfate, 
and 42.9% non-sulfate sulfur. A calculation of the various nitrogen, 
sulfur, and phosphorus ratios on the equivalent weight basis gives: 
P/S = 2.05 (1.38-2.97), N/P = 12.4 (9.2- 18.2) and N/S = 24.3 
(16.5-30.0). The ratio of inorganic-phosphate phosphorus to inorganic- 
sulfate sulfur on the same basis was more variable, 1.66-6.17. Additional 



Fig. 5. Correlation of lipide with the log df its nitrogen content in normal bone 
marrow. The values in the insert are for the equation, L = a + (6 ± 07 ,) log 100 Ni 
with a standard error of estimate, ± <r, and the corelation coefficient, r. The shaded 
area represents the composition of normal rabbit marrow, dt <r (l). 




TABLE II 


218 


ALBERT A. DIETZ 


05 


/, 


a 

8 

£ 


c 

3 

o 

ft 

fi 

o 


a C kQ 

I'S^ 

£ 5P 

a 




£ 

3« 

2 s 




ft. 

H by 
c 


<j 03 

g 

a®® •. 

*7 m 

o 5 


3 *; 

jo l r 

W £ 


a. 

fci -- h 


w C 


jHcc 

5>lo 

3 ti) 
o S 
H 


IS 

<h 


St 

£ a. 


S 

« 


c3 

j3 


”3 

O 

Is 

a 

■H 

o 


T5 

CS 

cS 


3 

a 

03 

«*-< 

of 

I 

3 

w 


cd 

256 


CO 

CM 

© 

© 

SO 



05 

cm 

1 

fH 

CM 

8$5 

tH 

eo 

88 

CO 

SS 

£ 

78 

88 

s 

8 

96 

rH 

00 

96 

8S 

87 1 

S£ 













*-H 

00 

CO 

f-H 

o 

C5 

CM 

f-H 

88 

00 

00 

SO 

SO 

SO 

o 

tN* 


05 

f-H 

l>- 

IH 

r*H 

r-J 

*—i 

CM 

rH 

, " H 

f-H 

CM 

i-h 

r-H 

o 

CM 

sO 

o 


CD 

OO 

05 

q 

CD 

00 

cm 

8 8 

05 

f—i 

CD 

fH 


f-H 


d 

SO 

co 

CD 


SO 

<p f 

CO 

»o 

-t* 

8 

00 



co 



N 



oo 

o 



CM 

CM 


CD 

£ 

»o 


f—H 




CO 

CM 


CM 

fH 

f-H 

CM 

CO 

05 


CM 


SO 

00 

»o 

q 

o 

»o 

cm 

05 

od 

d 

CM 

sd 



-H 

od 

od 



f-H 

T-H 

CM 

r— < 



CM 

CM 


eo 

CM 

CO 

oo 

CD 

00 

oo 

o 

00 


00 


© 

o 


cd 

CM 

-t* 

cd 


o 

CM 












CO 


r>- 

CD 

oq 

CD 

CD 

»o 

CM 

00 

00 

00 

cd 

CM 

oo 

CM 

*6 

CM 

CM 

s 

CM 

CM 

CM S 

CM 

CM 

s 

00 

"* 

8 

00 

0© 

CD 

8 


CD 

CD 

3 

CD 

N 

s 



CM 


fH 



f-H 

i-H 



1 

CM 

CO 


o 

'*r 


o 

OO 

CO 

q 

! 00 

d 

CM 

i 

od 

fH 

i-H 

od 

H< 

od 


i_ 

i—H 



rH 


i-H 

rH 


t-H 


i 

i 

O 

CO 

o 

-T 

CD 

05 



»D 

oo 

co CO 

5? 

cd 

C5 

05 

•* 

s 

CM 

SO 

d 

so 

s $ 


rt?iM»tiOtONW050 £ 

03 



TABLE II —Continued 


BONE MARROW 


21 '.) 



o 

g 

2 

Ci 



00 

00 


1 

1 

CO 

o 

o 


i-H 

rH 



o 

*o ! 

q 


00 

f—4 

<N 

csi 

o 

CO 

00 

cd 

oo 

f—1 

*C 

t> 

*o 

o 


1 

1 


X 

j 

00 


i 

_i 

! 

1 

i 


| 00 CO j 

I >o x x t 

OJ 1 -H 10 ' 




X »0 Oz »0 

l» 00 

** 55 -f t*. 

X ~ 

i 

-« OJ 

38 

45 

44 

85 

o 

^ S -t 

0* h- 

00 ** »o i-0 

q q 

dodo 

1—H 

*o x o 

o — 

d -f x 

rf ci 

-f- (N ^ 

zo 

O M O 

* ! 

-t 

CO 



c- 

»o —• 

»d 

d x , 


i 

i 


c: 

o 

j 

^ o 

-t* O O iO 

q q 

d d »d o 

d oo 



i 

1 O <N CO 


i-O -* 

CO 



X <N »0 

t- o 

«-i 00 CO 00 

00 CQ 


*—< 

1 ' 

q e- x ^ 

x *r 

d r»J »o »6 

od *-<* 

1—4 t—« 

co io 



>0 H N 


i 

rtj C5 

<3 o 


° Humerus and femur only. 

b Marrow from a pocket of active marrow in the femur of an older bull. 
c Marrow from humerus, femur and tibia. 
d Rabbit no. 7 not included in average. 








220 


ALBERT A. DIETZ 


relationships can readily be calculated from the values given in Table 
II. 

A sample of beef and cat marrow is included in Table II for comparison. The most 
marked difference is in the low values obtained for the inorganic sulfate content. Of 
several dog tissues analyzed, Denis and Leche (11) found the liver to have the largest 
concentration of total sulfate, 20-21 mg./lOO g. The values given for rabbit bone 
marrow are of the same order of magnitude, and would Ik- somewhat higher if both 
results were calculated on a lipide-free basis. 

Discussion 

The composition of many tissues can lie expressed by a point, and 
variations from the average taken as deviations from nonmil. Where 
three or more variable major components are present, the normal 
variation is no longer expressible by an average, unless one of the 
variables is neglected. This is frequently done by expressing the compo¬ 
sition on a fat-free basis. A more accurate picture, however, is obtained 
by calculating the correlation between the components and expressing 
the composition by a line. This method was applied by Fenn ami llaego 
in a study of the composition of liver (12). It is of special importance in 
the study of the chemistry of bone marrow because of the great varia¬ 
tion in the content of the three major components, water, lipide, and 
protein. Not only is the variation pronounced in different animals, but 
is very marked within a given animal, depending upon the bone from 
which it is removed. In all cases studied, the major components of the 
marrow showed a definite correlation to one another. Deviations from 
normal take the form of a shift in the limits of the line with the same 
coefficients, a change in the coefficients, or both. All types of changes 
have been found in pathologic marrow, and will be reported in other 
papers. 

The inability to obtain a correlation between the lesser components 
and the activity of the marrow may, in part, be due to the unavail¬ 
ability of a sufficient number of samples from the same animal. Within 
a given animal the nitrogen, sulfur, and phosphorus usually increased 
with increased activity of the marrow. Marrows of similar water, lipide, 
and residue concentrations, taken from different animals, showed 
variable amounts of the lesser components. The relationship was such 
that high values of these components were associated with active 
marrow, but somewhat lower values did not necessarily indicate a less 
active marrow. This is similar to the relationship found by McCoy and 



BONE MARROW 


221 


Schultze (4) for the hemoglobin concentration of animals recovering 
from anemia. The variability of the values may well indicate differences 
in the manufacture, storage, and discharge of the cellular elements. 

Summary 

There is a direct linear correlation between the water and residue 
(lipide-free solids) and an inverse linear correlation between those two 
components and the lipide content of marrow from guinea pigs, rats, 
cats, beef, hogs, and rabbits. The rat, guinea pig, and beef marrow had 
a relatively smaller amount of water associated with the residue than 
did that from rabbits. In no case was a significant amount of water 
associated with the lipide. 

Samples of frog marrow have a relatively higher concentration of 
water than the marrow from the other animals studied. 

High concentrations of total sulfur, phosphorus, non-protein nitrogen, 
and inorganic sulfate and phosphate are found in the active marrow 
samples of rabbits, but somewhat lower values for these constituents 
do not necessarily indicate a less active marrow. The variations found 
may be due to differences in the rate of manufacture and liberation of 
the cellular elements. 


References 

1. Dietz, A. A., J. Biol. Chem. 165, 505 (1946). 

2. Singiier, II. O., and Marinelli, L., Science 101, 414 (1945). 

3. Lutwak-Mann, C., Biochem. J . 41, xxx (1947). 

4. McCoy, R. H., and Schultze, R. II., Biol. Chem. 156, 479 (1944). 

5. Tisdall, F. F., ibid. 50, 329 (1922). 

6. Kaucher, M., Button, J., and Williams, II. II., J. Lab. Clin. Med. 27, 349 

(1942). 

7. Marenzi, A. D., Banfi, L. S.,-and Banfi, R. F., Anales farm, y bioquhn. 

{Buenos Aires ) 15, 113 (1944). 

8. Cope, C. L., Biochem. J. 25, 1183 (1931). 

9. Powers, M. H., and Wakefield, E. G., J. Biol. Chem. 123, 665 (1938). 

10. Evans, R. J., and St. John, J. L., Ind. Eng. Chem., Anal. Ed. 16, 630 (1944). 

11. Denis, W., and Leche, S., J. Biol. Chem. 55, 565 (1925). 

12. Fenn, W. O., and Haegk, L. F., ibid. 136, 87 (1940). 



Chemical Composition of Irradiated Bone Marrow 

Albert A. Dietz and Bernhard Steinberg 

From the Toledo Hospital Institute of Medical Research , Toledo 6, Ohio 
Received March 29, 1949 

Introduction 

The literature on the cellular changes of bone marrow resulting from 
irradiation was recently reviewed by Warren and Dunlap (1). There 
is, however, little information on the chemical changes following 
irradiation of marrow. In this paper, the composition of marrow waft 
determined at various intervals following the administration of a uni¬ 
form dose of X-rays over the legs on the right side, and compared with 
unirradiated marrow removed from similar bones of the left legs. 

Experimental 

Rabbits over 4 months old were anesthetized with nembutal and the desired area 
irradiated with 3000 r of unfiltered X-ray delivered in a field of 380 cm. 2 , at the rate 
of 160 r/min. The animals were killed at 1-32 day intervals following the irradiation. 
In most cases the right foreleg was irradiated with the rest of the body covered by 
1.5 mm. of lead. In these animals the left foreleg was used as the control to measure 
the,changes taking place in the irradiated limb, and the marrow of the hind legs was 
used to determine the effect of sampling. The marrow samples were analyzed for the 
major constituents as previously described (2). For the more complete analyses, the 
foreleg and the hind leg on the right side were irradiated separately and an aliquot of 
the combined humerus and femur from this side compared with a similar aliquot from 
the left side. Inactive marrow samples from the radius, ulna, and distal tibia weie 
also compared, but, because of the relatively small changes in their composition, most 
of these results are not included. Aliquot samples were analyzed for the major compo¬ 
nents (2) and for the nitrogen, sulfur, and phosphorus distribution (3). The terms 
X-rayed, control, and normal marrow are used in connection with the composition of 
the irradiated and non-irradiated marrow from the same animal and the composition 
of normal rabbit marrow as previously reported (2,3), respectively. Representative 
samples of marrow were used for histologic examination. 

Results 

The changes in the gross composition of rabbit bone marrow following 
irradiation are summarized in Tabic I, and representative analyses 


222 



IRRADIATED BONE MARROW 


223 


including the nitrogen, sulfur, and phosphorus components are given 
in Table II. At one day following irradiation, when the marrow showed 
a slight decrease in mature cells, no significant changes were noted in 
the gross composition, but the inorganic sulfate had decreased about 
25%. With longer periods after radiation a marked increase in the li- 
pide and decrease in the water, residue, and total and lipide nitrogen 
was found. Of the lesser components, the most marked changes were 
decreases in inorganic sulfate and non-protein nitrogen with slightly 
smaller decreases in the non-protein organic sulfur. The results on the 
•inorganic phosphate determinations were somewhat irregular, but the 
analysis of a larger group of data than is given in Table II indicated 


TABLE I 

Changes from Control in the Percentage Composition of Active Bone 
Marrow Following 3000 r of X-Ray 








Difference 




Difference in per cent total marrow 

in ner cent 
of lipide 


Days after 
X-ray 

No. of 
samples 





Histologic 

changes 



Water 

Lipide 

Residue 

Total 

nitrogen 

Lipide 

nitrogen 

1 

3 

1.8 

-2.7 

0.9 

0.21 

0.023 

Decrease in 
mature forms 

2 

5 

— 1.5 

5.6 

-4.1 

-0.62 

-0.067 

Marked hypo- 

3 

2 

-4.4 

9.3 

-4.9 

-0.97 

-0.247 

plasia and nu- 

4 

5 

-1.1 

4.6 

-3.6 

-0.63 

-0.068 

clear degener- 

8 

12 

-6.2 

9.6 

-3.5 

-0.58 

-0.065 

ation 

S. D.« 


±4.03 

±5.34 

±1.47 

±0.229 

±0.042 


t b 


4.30 

3.16 

8.55 

9.1 

3.64 


P b 


0.001 

0.005 

0.001 

0.001 

0.001 


16 

4 

-10.0 

12.4 

-2.5 


-0.066 

Decrease in 

24 

4 

-6.2 

8.8 

-2.6 


-0.059 

cellularity to 

32 

7 

-7.8 

11.3 

-3.5 

ESI 

-0.062 

no change 

Non-ir radi¬ 
ated hind legs 

15 

-0.57 

0.62 

-0.04 


-0.016 


S. D.‘ 


±2.7 

±3.23 

±0.56 

±0.088 

±0.045 



° Standard deviation 


IXx’ 


- JtXx 


b t and P —see Ref. 4. Compared with hind legs. 











224 


ALBERT A. DIETZ AND BERNHARD STEINBERG 


TABLE II 

Composition of Irradiated ami Non-lrrwliated Humerus and Femur 
Rabbit Marrow , 8000 r 


Du\ * 




1 

Xon-protoin milfur 

1 ‘hos- 

Nitrogen 

afttM 

X-ra\ 

Mftriow 


ItfNidue 

InorR. 

S0 4 

mg.- r ; s 

Ktlioi 

SO 4 

tnp.-'i S 

! Organic 

S 

s 

mg.-*,' 1 * 

Total 

N 

Lipido 
ntg - r '( \ 

Non-pro- 

teiu 

mg .-7c \ 

0 

O' 

per cent 

35.5 

per cent 

10.5 

16.3 

2.0 

14.4 

51.0 

per rent 

1.52 

80 

126 


V 

35.8 

9.9 

15.8 

2.6 

15.8 

49.8 

1.51 

90 

148 

1 

X « 

43.1 

10.4 

12.3 

0.5 

11.8 

32.6 

1.47 

73 

113 


c 

41.5 


16.1 

2.2 

14.1 

36.0 



141 

2 

X 

38.5 

8.4 

9.4 

0.5 

16.4 

35.3 

1.33 

99 

131 


c 

39.5 

11.3 

19.0 

2.8 

26.0 

66.4 

1.54 

108 

251 

4 

X 

40.7 

0.0 

2.9 

0.5 

4 8 

103.5 

1.18 

90 

92 


c 

46.0 

13.2 

20.0 

1.6 


83.0 

1.89 

I__ 

117 

251 

8 

X 

39.7 j 

<U i 

4.1 

0.9 

10.6 

25.0 


37 

64 


! o ! 

j j 

40.1 

8.1 j 

9.9 

! 1.2 

17.6 

45.8 


66 

94 

16 

x 1 

23.6 

0.7 

9.0 

1.4 

16.1 

56.5 

1.02 

100 

128 


v. 

27.9 ! 

_ 

! s.o 

15.8 

, 1.0 

24.2 

43.1 

1.23 

118 

204 

32 

X 

43.5 

1 

9.8 

0.4 

18.7 

63.0 



143 

• 

V 

51.4 

13.3 | 

19.1 

2.3 

20.1 

56.2 

1.96 

87 

206 

i h 

X 

15.1 

"IT 1 

2.3 w 

1 


79.6 

0.73 

84 

35 


c 

1 

16.7 

7.9 | 

6.f 

i 

i 

50.5 

1.08 

162 

99 


“ “C” and “X” represent Control and X-niyed marrows, respectively. 
b Radius and ulna. 


that the inorganic phosphate frequently showed a decline during the 
first 8 days and thereafter an increase over the control marrow. The 
smaller change in phosphate is also shown in the ratio of inorganic- 
phosphate phosphorus to inorganic-sulfate sulfur calculated from Table 
II on an equivalent, weight basis. In many of the cases, this ratio for 
the irradiate marrow is more than double that of the control marrow, 
for example, in the samples 32 days after irradiation, the vlaues are 
10.03 and 4.57, respectively. The distribution of the non-protein sulfur 





IRRADIATED BONE MARROW 


225 


components, showed a smaller percentage inorganic sulfate in the 
irradiated marrow in most cases. 

The changes noted above were found up to 32 days following irradi¬ 
ation. They persisted even though, in some cases, little or no/lifference 
could be found in the histologic picture. The regeneration was not uni¬ 
form in the different animals, which is consistent with the variation in 
response of individuals to X-ray, as found by others (1). 

The increase in lipide and decrease in percentage composition of the 
other constituents is consistent with a decrease in eellularity of the 
marrow. The greatest number of values were obtained for the 8>day 
Regenerated marrow. The significance P of the changes was calculated 
for these samples (4) (Table I). The likelihood of any one average 
difference occuring by chance is of the order of 1 in 1000. 



Flo. 1. Composition of irradiated and noil-irradiated bone inarrow from humerus 
8 days after irradiation, plotted on triangular coordinates. The broken line represents 
normal marrow. The irradiated and non-irradiated marrow from similar sites'in the 
same animal an* shown by the arrowed and blank ends, respectively, of the shorter 
lines. The multiple correlation equation for the X-rayed marrow is: 

IF - (0.15 ± (UK»/, + «).81 ± 0.72)/?, 

and for the control, 

IF = (0.020 ± 0A2)L + (1.13 =fc 0.52)/?. 

Standard errors of estimate of the equations are 5.00 and 7.83, respectively. 


Further analysis of the water, lipide, and residue composition of the 
irradiated marrow indicates changes other than a simple reduction in 
activity. In Fig. 1, where the composition of the 8-day regenerated 
marrow is plotted on triangular coordinates, and compared with the 
corresponding control marrow, some of the changes tend to parallel 
the line for the normal marrow. This change is consistent with a simple 
. decrease in activity. The majority of the lines, however, deviate from 
this, in that there is relatively less residue iu the irradiated marrow. 
This is further shown by the coefficients of the residue in the multiple 
correlation equations, Fig. 1. The coefficient of /£, g. H 2 0/g. residue, 



226 


ALBERT A. DIETZ AND BERNHARD STEINBERG 


for the non-irradiated control marrow is 4.43 db 0.52, and for the 
irradiated 6.81 ± 0.72. In neither case was a significant amount of 
water associated with the lipide. 

In Figs. 2-4, the simple correlations of the major components are 
graphed. The most pronounced change is in the shift of the activity of 



Fig. 2. Correlation of lipide and water 
in irradiated and non-irradiated bone 
marrow 8 days after irradiation. The line 
A-I, if connected, is the regression line of 
normal rabbit marrow; the short dashes 
being one standard error of estimate from 
the solid lines. A and I representing the 
active and inactive ends of the marrow 
scale, respectively. The standard errors of 
estimate are shown by the sides of Ihe 
rectangles, and the ranges in composition 
by the lengths. 



Fig. 3. Correlation of water and residue 
in irradiated and non-irradiated bone 
marrow 8 days after irradiation. Cf. Fig. 2. 


the irradiated marrow toward the inactive end of the curves. The loss 
of residue is relatively greater than the loss of water, Fig. 3. The slopes, 
of the curves and the position of the limits show the deviations from 
normal to be greatest for the marrow samples normally most active. 
Marked decreases in inorganic sulfate and non-protein nitrogen were 








IRRADIATED BONE MARROW 


227 



Fia. 4. Correlation of residue and 
lipide in irradiated and non-irradiated 
bone marrow 8 days after irradiation. 
Cf. Fig. 2. 


found in the inactive marrow, of the radius and ulna (Table 11), even 
though the changes in the major components were small. 

The equations for the regressions of total nitrogen on residue and for the log of 
lipide nitrogen on lipide did not deviate significantly from normal as the calculated 
lines for the irradiated marrow fell within one standard error of estimate of the normal 
(2). There is, however, a shift of the irradiated samples toward the inactive end of the 
curves. The magnitude of this shift can be seen in Table I. 

The analysis of the protein-freo filtrates of the serum of these animals showed no 
abnormal changes in the nitrogen, phosphorus, or sulfur content. In view of previous 
work (I), measurable changes in the blood would not be expected to result from the 
irradiation of a small area. 


Discussion 

Many of the chemical changes due to irradiation have been deduced 
from observations on blood or on urine (1). The analyses recorded in 
this paper show the magnitude of the effect of irradiation on the bone 
marrow itself. The increased amount of water associated with a unit 
weight of residue in the irradiated marrow is consistent with observed 
edema in and around the cells of the irradiated area (1). An increased 
urinary excretion of non-protein nitrogen is usually explained as a 
result of the destruction of protein (1), but the marked decrease in the 
non-protein nitrogen of the irradiated marrow would indicate that at 
least a portion of it is derived from the non-protein nitrogen of the 
tissue. The relatively great changes in the inorganic sulfate and non¬ 
protein organic sulfur would indicate that the various sulfur compounds 
play an important role in the hemopoietic function of bone marrow. The 
decrease in inorganic phosphate, followed by an increase after about 8 



228 


ALBERT A. DIETZ AND BERNHARD STEINBERG 


days, is in line with the observation that, “Alterations in the total 
phosphorus level is by no means a constant finding after irradiation, 
and when present is usually of brief duration.” (1). 

Summary 

Marrow, following 3000 r of X-ray, was analyzed at intervals up to 
32 days after irradiation and compared with non-irradiated marrow of 
the same animals. The irradiated marrow showed an increase in lipide 
content and decrease in water, residue (lipide-free solids), non-protein 
sulfur fractions, and total, lipide, and non-protein nitrogen. The changes 
in composition were greatest in marrow normally most active. A 
relatively greater amount of water was associated with the residue of 
the irradiated marrow. 


References 

1. Warren, S., and Dunlap, C. E., Arch. Path. 34, 562 (1942). 

2. Dietz, A. A., J. Biol. Chein. 165, 505 (1946). 

3. Dietz, A. A., accompanying paper. Arch. Biochem. 23, 211 (1949). 

4. Croxton, F. E., and Cowden, D. J., Applied General Statistics, p. 330. New York. 

1945. 



Ultracentrifugal Studies on Some Porcine 
Plasma Protein Fractions 


Virgil L. Koenig 1 

From the Institute of Physical Chemistry , Upsala University , Sweden 
Received April 11, 1040 


Introduction 

In view of the availability of some protein fractions from porcine 
plasma, a study of the ultracentrifugal properties of some of the 
protein fractions was undertaken. 

The freshly collected citrated porcine plasma was fractionated in the 
Armour Laboratories according to procedures patterned after those 
developed by Cohn at al. ( 1 , 2 ). Only those fractions were studied which 
had been examined eleetrophoretically. The following fractions were 
studied: Fraction II P -2 (a fairly eleetrophoretically homogeneous 
preparation of 7 -globulin consisting of one resolvable component), 
^Fraction III-l P -2 (a mixture which consisted of 57% of a protein with 
the mobility of 7 -globulin and 43% of a protein with the mobility of 
0 -globulin), and Fraction V P-5 (a mixture which consisted of 84% of a 
protein with the mobility of albumin, 8 % of a protein with the mobility 
of a-globulin, (5% of a protein with the mobility of 0 -globulin, and 2 %, 
of a protein with the mobility of 7 -globulin). The electrophoretic 
analyses wore furnished by the Armour Laboratories. 

Experimental 

All of the? experiments wen* carried out on the Svedberg oil turbine velocity centri¬ 
fuge. A speed of 59,000 r.p.m. was used. The Lamm Scale method for observing the 
sedimenting boundaries was used throughout the investigation. All the preparations 
were dissolved in 0.2 M NaCl. The preparations were stored in the iyophilized form 
under refrigeration. Sedimentation runs were made at 10 different protein concentra¬ 
tions, ranging from 0.15% to 5.00% based on the dry weight of the preparation; thus 
the variation of the sedimentation constant with protein concentration could be 
observed. The treatment of the ultracentrifugal data was the same as Pedersen’s (3). 

1 Fellow of the American-Scandinavian Foundation, 1946-1947, on leave from the 
Armour Laboratories, Chicago. Now at P. O. Box 1663, Los Alamos, New Mexico. 

229 




230 


VIRGIL L. KOENIG 


Results 

Fraction II ( y-globulin ). Fig. 1 shows the sedimentation diagram for 
Fraction II in a 0.5% solution after running for 80 min. The ordinate, 
Z, represents the scale line displacement and is proportional to the 
concentration gradient, while the abscissa, x, is the scale reading for the 
test run and is proportional to the distance from the axis of rotation, 
meniscus, etc. There is evidence for a small amount of heavier material 
as well as a very small amount of lighter material. The preparation is, 
however, predominately one component. The curve is symmetrical. 



Fig. 1. Sedimentation diagram for Fraction II. 

The equation for the line of regression of S} 0 on concentration, c, as 
calculated by the method of least squares and the correlation coeffi¬ 
cient, are given in Table I. Concentrations in terms of An, the refrac¬ 
tive index increment, are calculated from the diagram by the method of 
Pedersen (4). The equation for the line of regression of St o on An, 
obtained by the method of least squares, and the correlation coeffi¬ 
cient are given in Table I. 

Fraction III-l. The sedimentation diagram for a 0.5% solution after 
sedimenting for 50 min. is shown in Fig. 2. The preparation consists 



PLASMA PROTEIN FRACTIONS 


231 


TABLE 1 


Protein 

fraction 

Line of reKreHnion for 

Sm on concentration 

Correla¬ 
tion coeffi¬ 
cient r 

Line of refcresnion for .S’*o on In 

Correla¬ 
tion coeffi¬ 
cient r 

II 

N 20 = 7.28-0.39c 

-0.949 

•S'so = 7.25 — 0.002") An 

-0.951 

I IT-1 

1st comp. 

, 8 2n = 40.46-37r 

-0.895 

*820 — 44.72 — 0.83 \u 

-0.628 

2nd comp. 

N 20 = 22.64-1.68c 

-0.932 

.8*0=23.19—0.22 An 

-0.925 

3rd comp. 

N 20 = 7.28-0.18c 

-0.973 

1 - 

.S' 2 o= 7.28 -0.0020 An 

-0.967 

V 

Albumin 

<8>o = 4.68—0.18c 

-0.983 

. 820 = 4 , 67 - 0,0016 An 

-0.982 

Heavy com¬ 
ponent 

,8,o= 7.33-0.46c 

-0.992 

. 8 * 0 =7.50-0.019 An 

-0.967 



Fig. 2. Sedimentation diagram for Fraction 111-1. 







232 


VIRGIL L. KOENIG 


essentially of one major component with at least 2 faster sedimenting 
components and a slower sedimenting component in small concentra¬ 
tions. The equations of the regression lines of o on c and on An are 
given in Table I, together with the respective correlation coefficients. 
The concentration values in the equations of Table I represent total 



Fia. 3. Sedimentation diagram for Fraction V. 

concentration based on the preparation as weighed out in dry form, and 
do not represent the exact concentration of each component. The 
absolute concentration would of necessity be the total concentration of 
the protein multiplied by the percentage composition of each compo¬ 
nent in the mixture. 


PLASMA PROTEIN FRACTIONS 


233 


Fraction V. The sedimentation diagram for a 0.5% solution of Frac¬ 
tion V after running 100 min. is given in Fig. 3. The preparation pre¬ 
dominates in albumin, but has with it about 20% of a heavier sedi¬ 
menting protein. The equations for the regression lines of Sw on c and 
on An arc given in Table I, together with the respective correlation 
coefficients. The concentration values in the equations of Table I 
represent total concentration of the preparation as weighed out in dry 
form, and do not represent the concentration of each component. As 
in the case of Fraction III-l, the absolute concentration would of 

TABLE II 

Summary of Data 



Fraction II 



Fraction III-l 



Fraction V 

(’oncen- 
t ration in 
yei rent 

.S M 

An 

1st 

component 

2nd 

component 

3rd 

coin |M»nont 

Albumin 

Heavy 

Component 




Sn 

An 

! 

<S'to 

An 

»S;o J 

An 

.S,,i 

An J 

SfH 

An 

0.15 

7.48 

20 

41.26 


22.21 

3 

7.25 ! 

11 




; 

0.25 

6.06 

37 



24.87 


7.31 

26 

4.70 

26 

7.37 

0 

0.50 

0.92 

80 

40.02 


21.63 

8 

7.11 

17 ' 

4.53 ; 

54 



0.75 

6.92 

113 

38.65 

0 

20.56 

0 

7.13 

72 

4.55 

02 

0.02 

25 

1.00 

6.80 

147 

35.20 

: 

10.77 

10 

7.17 

87 

4.51 

108 

6.88 

34 

1.50 

6.86 

219 

30.04 

16 

20.16 

17 

7.07 

120 

4.42 

156 

6.67 

14 

2.00 

6.57 

285 

30.81 

12 

18.24 

24 

6.84 

188 

4.25 

215 

6.20 

71 

3.00 

6.32 

420 

36.30 

14 

17.84 

20 

6.82 

220 

4.23 

300 

5.03 

08 

1.00 

5.33 

610 

22.53 


15.71 


6.40 

324 

3.89 

420 

5.43 

115 

5.00 

5.46 

812 

22.73 


14.00 

30 

6.46 

461 

3.70 

578 

5.17 

103 


necessity be the total concentration of the preparation multiplied by 
the percentage composition of each of the two components in the 
preparation. 

Table II gives a summary of the An data. 

Discussion 

There is good agreement between the Sjo values for Fraction II 
extrapolated to zero concentration and zero An; namely 7.28 S and 
7.25 S, respectively. The slopes of the two regression lines have the 
expected difference, inasmuch as a constant relating index of refraction 





234 


VIRGIL L. KOENIG 


to concentration would be the ratio of the two slopes. The correlation 
coefficients 0.949 and 0.951 indicate well the straight-line relationship 
between Sio and concentration or An. 

Fraction III-l presents a mixture of proteins with a definite prepon¬ 
derance of one component. The diagram of Fig. 2 shows the main com¬ 
ponent and one heavier component. The heaviest component has 
already sedimented to the bottom of the cell and is not visible in the 
diagram. This component exhibits rather high concentration depend¬ 
ence, as judged by the slope of the regression line. This dependence is 
probably due in part to the viscosity of the solution because of the 
presence of the other proteins, and in part may indicate a long fila¬ 
mentous molecule. This heavy component may indicate a denatured 
portion of the fraction. The <S' 2 o values extrapolated to zero concentra¬ 
tion and zero An agree rather well in view of the fact that there was a 
limited amount of An data. The limited amount of An data is reflected 
in the lower correlation coefficient, 0.628. The r value for concentration, 
0.895, is more substantial. 

The second component, probably the 20-component of Pedersen (3), 
exhibits rather high concentration dependence as judged by the slope 
of the regression line. Here, a long filamentous molecule is suggested, 
as well as the other factors discussed under the previous component. 
There is good agreement between the S 20 values extrapolated to zero 
concentration and zero An. The correlation coefficients, 0.932 and 
0.925, indicate the straight-line relationship. 

The third component or the main component includes the greatest, 
part of the protein. There is rather low concentration dependence. The 
<S 2 o value extrapolated to zero concentration agrees well with the value 
extrapolated to zero An. The <S 2 o values are the same as those for Frac¬ 
tion II. From the electrophoretic analysis, one would conclude that this 
component is made up of both /3- and 7 -globulins, both of which have 
about the same sedimentation constant. The r values are good and 
support the straight-line relationship. 

The diagram in Fig. 2 indicates the presence of a lighter component 
and is better resolved as the sedimentation progresses, but not well 
enough resolved to be quantitatively considered. 

Fraction V represents albumin contaminated with a heavier com¬ 
ponent. It has so far not been possible to obtain a pure albumin. The 
preparation described here represents the best that has been obtained 
in any quantity. The <S 2 0 values extrapolated to zero concentration and 



PLASMA PROTEIN FRACTIONS 


235 


zero An agree well for the albumin, the lighter, and the main component. 
The correlation coefficients, 0.983 and 0.982, show the straight-line 
relationship. 

The heavy component shows fairly good agreement between the 
S 20 values extrapolated to zero concentration and zero An. The values 
are close to those for the /3- and y- globulins. The r values, 0.992 and 
0.967, respectively, support the straight-line relationship. 

Acknowledgments 

* The author is indebted to Professor The Svedberg for his kindness in making 
available the facilities of the Institute for this work, and to The Armour Laboratories, 
Chicago, for making the plasma proteins available, as well as for other help. The 
author is grateful to Laborator Kai 0. Pedersen for his advice and help. Many 
thanks are extended to Mr. Evald Heilman, who, together with his staff, carried out 
most of the calculations on the experimental data. To the technicians who operate the 
ultracentrifuge, thanks and appreciation are extended for their cooperation. 

Summary 

An ultracentrifugal study on a series of protein fractions from porcine 
plasma has been made. By extrapolating to zero concentration the 
straightline formed by plotting sedimentation against concentration, 
the sedimentation constants at infinite dilution have been determined. 
The following values were obtained: Fraction II, 7 -globulin, 7.25 
7.28 S; Fraction III-l, a mixture of 0 - and 7 -globulins, 40.46-44.72 S, 
22.64 23.19 <S, and 7.28 S; Fraction V, an impure albumin, 4.67-4.68 S 
and 7.33-7.50 S. The properties and composition of these fractions 
were discussed. 


References 

1. Cohn, E. J., Lijetscher, J. A., Jr., Oncley, J. L., Armstrong, S. H., Jr., and 

Davis, B. D., J. Am. Chem. Soc. 62, 3396 (1940). 

2. Cohn, E. J., Strong, L. E,, Hughes, W. L., Jr,, Mulford, D. J., Ashworth, J, 

N., Melin, M., and Taylor, II. L., ibid. 68, 459 (1946). 

3. Pedersen, K. O., Ultracentrifugal Studies on Serum and Serum Fractions. Alm- 

qvist & Wiksells Boktryckeri A. B., Uppsala, Sweden, 1945. 

1. Svedberg, T., and Pedersen, K. O., The IJltracentrifuge. Oxford, 1940. 



Occurrence of an Unidentified Rat Growth Factor 
in Cottonseed Meal 

W. R. Ruegamer 

From Swift & Co., Research Laboratories, Chicago, IUinois 
Received April 14, 1949 

Introduction 

In a recent paper, Register et al. (1) described an improved rat assay 
procedure for a growth factor present in liver extract. The corn-soy- 
bean ration used in these tests, even though well fortified with vitamins 
and minerals, was found to be inadequate for normal rat growth. In¬ 
cluding either desiccated thyroid or iodinated casein at an appropriate 
level depressed the growth rates of the animals still further. Afactor(s) 
present in commercial pernicious anemia liver extract was effective in 
restoring the rat growth to a more normal rate. Subsequent studies by 
Register et al. (2) have demonstrated that crystalline vitamin Bi 2 will 
give a quantitative response in this assay. 

Evidence presented in this report indicates that cottonseed meal 
contains the rat growth factor as contrasted to soybean meal which does 
not. 'This observation is in accordance with that of Zucker and Zucker 
(3) who have reported the existence of a lactation factor in cottonseed 
meal. 

Experimental 

Four test diets were prepared and fed to 4 groups of male weanling rats, 8 rats per 
group. Gp. 1 received the basic corn-soybean ration, 1 Gp. 2 received the corn-soybean 
diet plus 0.25% desiccated thyroid, 2 Gp. 3 received the same diet as Gp. 1, except for 
the substitution of hydraulic cottonseed meal for soybean meal, and Gp. 4 received 
the same diet as Gp. 3 plus 0.25% desiccated thyroid. At the end of the fourth week 
(Fig. 1) the desiccated thyroid content of the diets fed Gp. 2 and 4 was increased to 
0.4%. Starting the sixth week, Gp. 2 was subdivided into two groups of 4 rats each. 

1 This ration is the same as that of Register etal. (1) except for the inclusion of 1% 
A and D oil (30001. U. of A, 4001. U. of D/g.) at the expense qf the com and soybean 
meal. 

1 Parke-Davis desiccated thyroid containing 0.3% iodine. This preparation is 50% 
stronger than U. S. P. 


236 



RAT GROWTH FACTOR 


237 



Fig. 1. Growth rates of rats receiving soybean and cottonseed meal diets 
with and without added desiccated thyroid. 

tip. I, Corn-soybean basal diet; Gp. 2, Corn-soybean basal diet plus 0.25% desic¬ 
cated thyroid; Gp. 3, Corn-cottonseed basal diet; Gp. 4, Corn-cottonseed basal diet 
plus 0.25% desiccated thyroid. 

A. Desiccated thyroid content increased to 0.4% in diets fed gps. 2 and 4. 

B. Gp. 2 subdivided and half (broken line) given daily injections of liver extract. 

One subgroup remained on the desiccated thyroid diet as a negative control, whereas 
the other subgroup received additional supplements of 1 U. S. P. unit of liver extreet 3 
l>cr rat injected daily. 

From Fig. 1, it can be s<*en that for the first 4 weeks of the experiment, 0.25% 
desiccated thyroid produced a marked decrease in the growth rates of rats receiving 
the corn-soybean diet (Gps. 1 and 2). Those rats receiving the cottonseed meal diet 
with 0.25% desiccated thyroid added (Gp. 4) grew as well as their controls on the 


3 Armour 15 IT. S. P. units/cc. 



238 


W. H. HUEGAMEK 


unsupplemented cottonseed meal ration (Gp. 3). Elevating the level of desiccated 
thyroid to 0.4% produced a marked drop in the growth rates of rats receiving the 
corn-soybean diet and, to a much lesser extent, those animals receiving the 
corn-cottonseed meal diet. However, those rats receiving the cottonseed meal diet 
appeared to recover in part from the increased amount of desiccated thyroid and began 
to grow at a rate comparable to that of their controls (Gp. 3). The growth rates of rats 
receiving the corn-soybean diet plus 0.4% desiccated thyroid remained depressed. 
The subgroup which received liver extract as a supplement to the desiccated thyroid- 
corn-soybean diet demonstrated a marked weight gain, indicating that an actual 
deficiency had existed which could be overcome by supplying the missing factor. 

Conclusions 

Therefore, it may be concluded that hydraulic cottonseed meal con¬ 
tains activity for the rat growth factor. It is not certain whether this 
activity is due to the presence of vitamin B 12 or to the existence of some 
other antithyroid substance in cottonseed meal. Nevertheless, con¬ 
siderable heat is produced in the preparation of hydraulic meal, and it 
would, therefore, appear that the factor is relatively heat stable. Ex¬ 
periments are now in progress to determine the effect of processing 
upon the content of this factor in cottonseed meal, and to ascertain the 
amount present in several composite samples. 

References 

1. Register, U. D., Ruegamek, W. R., and Elvehjem, C. A., J. Biol. Chew. 177, 

129 (1949). 

2. Register, U. D., Lewis, U. J., Thompson, IT. T., and Elvehjem, C. A., Proc. 

. Soc. Exptl. Biol. Med. 70, 167 (1949). 

3. Zucker, T. F., and Zitcker, L. N., Arch. Biochem. 19, 323 (1948). 



Stereochemical Configuration and Provitamin A Activity. 
VII. Neocryptoxanthin U 

Harry J. Deuel, Jr., Samuel M. Greenberg, Evelyn Straub, 
Tomoko Fukui, A. Chatterjee 1 and L. Zechmeister 

From the Oates and Crellin Laboratories of Chemistry , California Institute of 

Technology , Pasadena , 2 and the Department of Biochemistry and Nutrition , 
the University of Southern California , Los Angeles* 

Received April 26, 1949 

Introduction 

The cis-trans isomerization of cryptoxanthin, C 40 H 56 OII, has been 
studied by several authors (2,6,7,8,10), and recently the formation and 
differentiation of the neocryptoxanthins U, A, and B was observed on 
the Tswett column ( 9 ). The naturally-occurring all -trans isomer was 
found to show 56% of the biopotency of that of /3-carotene (4), while 
the corresponding figure for neocryptoxanthin A (5) amounts to only 
42%. Since we have now been able to obtain neocryptoxanthin U in 
analytically pure, crystalline state ( 2 ), some comparative assays of its 
provitamin A effect in the rat are reported below. 

Experimental 

The tests were made in the same manner as earlier (2,3,4), except that only male 
rats were used. The supplements were prepared at the start of the test; each test 
substance was stored in a number of separate bottles under CO* in the deep freeze. A 
new sample was used twice weekly. The desired dose was present in 0.1 ml. of 
Wesson oil which contained an added amount of 0.5% a-tocopherol (0.5 mg. per dose). 
No appreciable destruction of 0-carotene, cryptoxanthin, or neocryptoxanthin U 
occurred after storing for one week in an ordinary refrigerator. In the deep freeze, the 
extinction values were identical with those of the original samples even after six weeks. 

Results 

Table I gives the data on the 107 rats used while the data from which 
the potency is calculated are given in Fig. 1 . 

1 N6e Mookerjee; Fellow of the Indian Government (Deputationist). 

2 Contribution No. 1288. 

8 Paper number 212 of the Department of Biochemistry and Nutrition, University 
of Southern California. 


239 



TABLE 


240 


DEUEL ET AL 


isf§ 

jf It 
3 & 4 


© o © »o 


© © © © 
05 Q ^ 
W ^ w 


rt< «0 © CO 


, OO o Ci 02 C »C t-h s 

as CM ^ CM CO CO »Q t-h CM 

w I 

© lO lO 02 H © 0005 

.CM 0^*^pr*0c;0505^ 

ts CM ^C^COC^^-COQO^^hCO 


lO 02 O *f 
00 GO CM iO © 


5 ot S 3 2 


© t* © CO 

^ 22 ?j 


C 4 ^ N iC r- O 0 

: ^ © H H 


(N - H 00 
"t O rf 0 


*-« NOON 

: '** »o cm cm cm 


•s S 

£<5 £ 

O be = 

2s.s 

CO 

CO 

Tj< 

05 

q 

i> 

oo 

CO 

1 g" 

4 - *-> M 

* srs 

»8 

s 

^H 

o 

tH 

o 

S 

8 

*"H 

o 

8 

« | 

< » * 

i—» 


*—( 

rH 

1—H 

r-H 

1 ” H 

r-H 

,s3» 

$ .g 

wqmi 

CM 


CM 

i—t 

CM 

i-H 

CM 

r— - 

CM 

CM 


O N S Ii 3 

H o H cm' 


v o d 
§ 23 e- 


weight at end of the depletion {X'riod and at the start of the assay period. 




NEOCRYPTOXANTHIN l f 


241 



Fig. 1. Relationship of gain in weigh! 
to log of daily dosage of 0 -carotene, all- 
/rarw-cryptoxanthin and neocryptoxan- 
t hin U. Points A, C, and E represent the 
projection of the growth of rats receiving 
2.5, 3.5, and 4.5 7 , respectively, daily of 
neocryptoxanthin U on the 0 -carotene 
curve while points B and D are the pro¬ 
jections of the growth of rats receiving 
0.75 and 1.25 7 , respectively, daily of all- 
trans-cryptoxanthin on the 0 -carotene 
curve. 


* The average provitamin A potency of alUraas-cryptoxanthin is 60% 
of that of 0-carotene, which is in agreement with the value reported 
earlier (4) of 56%. The biopotcncy of neocryptoxanthin U averages 
27% of that of 0-carotene. This figure corresponds with a value of 45% 
of that of the all -tram form of this carotenol. The decrease of the 
activity lies roughly in the same range as that of neo-0-carotene V 
which has a potency of 38% of that of the all-fra/w?-0-carotene (3). 

Summary 

Neocryptoxanthin U, probably a mono -cis isomer, shows a pro¬ 
vitamin A activity in the rat of 27% of that of all-/ra^-0-carotene, or 
45% of that of all-fra n s-cryptoxanthin. 

References 

1 . Baumgarten, W., Bauerxfeind, J. C., and Bowff, C. S., hid. Eng. Chem. 36 , 

344 (1944). 

2. Chatterjee, A., and Zechmeister, L., in press. 

# 3. Deuel, H. J., Jr., Johnston, C., Sumner, E., PolgAr, A., and Zechmeister, L., 
Arch. Biochem. 5, 107 (1944). 

4. Deuel, H. J., Jr., Meserve, E. R., Johnston, C. H., Polgar, A., and Zech¬ 

meister, L., ibid. 7, 447 (1945). 

5. Deuel, H. J., Jr., Meserve, K. R., Sandoval, A., and Zechmeister, L., ibid. 

10 , 491 (1946). 

6 . Fraps, G. S., and Kemmerer, A. R., Ind. Eng. Chem. y Anal. Ed. 13 , 806 (1941). 

7. White, J. W., Brunson, A. M., and Zscheile, F. P., ibid. 14 , 798 (1942). 

8 . White, J. W., Zscheile, F. P., and Brunson, A. M„ J . Am. Chem. Soc. 64, 2603 

(1942). 

9. Zechmeister, L., and Lemmon, R. M., ibid. 66, 317 (1944). 

10 . Zechmeister, L., and Tuzson, P., Biochem. J . 32, 1305 (1938); Ber. 72, 1340 

(1939). 



Stereochemical Configuration and Provitamin A Activity. 
Vm. Pro-y-Carotene (a Poly-ci's Compound) and 
Its AXL-trans Isomer in the Rat 

L. Zechmeister, J. H. Pinckard, Samuel M. Greenberg, Evelyn 
Straub, Tomoko Fukui and Harry J. Deuel, Jr. 

From the Oates and Crettin Laboratories of Chemistry, California Institute of Technology, 
Pasadena, California , l and the Department of Biochemistry and Nutrition, 
University of Southern California, Los Angeles , 5 California 
Received April 26, 1949 

Introduction 

As is well known, the introduction of one or two cis double bonds 
into a naturally-occurring, all-frans carotenoid brings about a sharp 
reduction in the provitamin A potency (1,3,5,6,7). The only instance 
where an increase in provitamin A activity has been noted is in the 
case of a naturally-occurring poly-m compound termed pro-y-carotene 
as compared with all-frans-y-carotene samples obtained from Mimulus 
longiflorus, Grant, or Pyracantha angustifolia, Schneid (4,2). 

However, since the respective all-£rans-y-carotene preparations as 
obtained from different sources show definite divergencies; e.g., in 
their melting points (8), it cannot be excluded that these are caused by 
minor variations which are of structural rather than stereochemical 
nature. 

In order to decide in an unambiguous manner whether or not the 
provitamin A potency of a poly-cts compound may be equivalent to 
that of the corresponding all-frans form, we have now prepared an 
analytically pure, chromatographically homogeneous, crystalline 
sample of all-<rans-y-carotene by iodine catalysis of a fresh solution of 
pure crystals of the naturally-occurring pro-y-carotene (ex Pyracantha 
angustifolia) (4,2). The comparative bioassays of these two provitamins 
which must be structurally identical and are only sterically different, 

1 Contribution No. 1289. 

1 Paper No. 214 of the Department of Biochemistry and Nutrition, University of 
Southern California. 


242 



PROVITAMIN A ACTIVITY. VIII 


243 


are described below. Some experiments with a sample of all-<ran$-y- 
carotene obtained from commercial tomato paste were also included. 

The tests were carried out in the same manner as earlier (1-7), except that only 
male rats were used* The samples of supplement were prepared in Wesson oil con¬ 
taining 0.5% a-tocopherol at the start of the experiment and stored under C0 2 in a 
number of bottles in deep freeze. A new sample was used twice weekly. There was no 
appreciable destruction of the carotenoids during the tests, as determined spectro¬ 
scopically. 

Experimental 

The al 1 -tram-y-c&ro tone was prepared from pro-7-carotene (ex Pyracantha) as 
follows. 

One hundred fifty mg. of crystals in 80 ml. of petroleum ether (b.p. 60-70°C.) in 
a 100 ml. Pyrex volumetric flask were combined with 4.5 mg. of iodine in the same 
solvent. The flask was irradiated with a 400 watt, water-cooled, incandescent bulb 
from a distance of 10 cm. for 5 min. Then the solution was developed with petroleum 
ether containing 6% acetone on a calcium hydroxide + Cclite (3:1) column (30 X 8 
cm.) and the main upper zone of all-2ra7is-7-carotene was cut out and eluted with 
acetone + ethanol. (The paler neo zones were catalyzed again as described.) The 
combined solutions of all-Jrans-7-carotene were rechromatographed on alumina + Ce- 
lite (4:1; benzene + petroleum ether 3:2). The main pigment zone was cut out, 
eluted, transferred with water into petroleum ether, washed ana dried. The filtered 
solution was then evaporated in vacuo to dryness and the residue was dissolved in the 
minimum amount of benzene. The all-Jrans-7-carotene was crystallized out by adding 
dropwise absolute methanol (cooling). After two recrystallizations, the yield was 40 
mg., m.p. 135-7°C. (corr.). = 15.7 X 10 at 461 m/i, in hexane; and 14.2 X 10 4 
at 473 m/u in Wesson oil. 

Anal.: Calcd. for C4oH 66 : C, 89.48; H, 10.52; found: C, 89.79; H, 10.38. 

In a mixed chromatogram test there was no separation of all-frans-7-caroteno ex 
pro-7-carotene and an authentic sample isolated from tomato paste or commercial 
carotene preparations. 

Results 

Two series of tests were carried out, but the results of only the second 
series are recorded in Table I. 

The average potency of pro- 7 -carotene was found to be 41% of that 
of all-/ran$-/3-carotene. This result agrees well with the 44% value 
recorded earlier (4). 

The provitamin A activity of the all-frans-y-carotene prepared by 
iodine catalysis from the pro-y-carotene was identical (42%) with that 
of this poly-czs isomer. A similar high result was also obtained with the 
all-Jrans-y-carotene prepared from tomato paste where the two values 
give an excellent agreement at a figure of 47%. On the other hand, it is 



244 


ZECHMEISTER ET AL. 


apparent that the response of the sample of'all-frans-Y-carotene used 
here is markedly superior to that of the product obtained from Mimulus 
or from Pyracantha where potencies of only 28 and 26% of the 0-caro¬ 
tene values were found earlier. 


TABLE I 

Summary Table Giving Body Weights and Total Gain in Weight for Male Rats 
Receiving 0-Carotene, AU-trans-y-Carotene (ex Pro-y-Carotene and Tomato 
Paste), and Pro-y-Carotene in Wesson Oil Containing 0.5% 
a-Tocopherol, and for Negative Controls Receiving only 
Wesson OH Containing 0.5% a-Tocopherol 


Supplement 

Dose 
per day 

Number 
of rate 

Average 

starting 

weight 0 

Average 

total 

gain 

Average 

final 

weight 

Number 

died 

Calculated 
potency (13- 
carotene « 
100) 


y 


U‘ 

0 • 

u- 


r 

/3-Oarotene 

0.5 

11 

103.3 

17.5 

121.1 

( 10 ) 

1 



1.0 

13 

106.7 

46.5 

( 11 ) 

153.5 

( 11 ) 

2 


AM-trcm 8-7-carotene 
(ex pro-7-carotene) 

1.2 

13 

, 105.8 

18.5 

124.4 

( 12 ) 

1 

43 


1.8 

13 

105.2 

33.3 

137.7 

( 12 ) 

1 

41 


2.4 

13 

105.7 

47.5 

153.2 

0 

42 

Pro-7-carotene 

1.2 

12 

105.8 

18.5 

124.3 

0 

43 


1.8 

14 

105.1 

30.7 

135.8 

0 

38 


2.4 

; 

13 

105.7 

46.5 

152.2 

0 

42 

All-frans-7-carotene 

i 

1.2 

13 

107.5 

22.5 

129.9 

0 

47 

(ex tomato paste) 

1.8 

13 

107.2 

39.2 

147.9 

( 12 ) 

1 

47 

Negative controls 

0.0 

12 

100.6 

- 11.0 

( 2 ) 

82.0 

( 2 ) 

10 

— 


• Body weight at the end of the depletion period and at the start of the assay period. 


We cannot offer any clear explanation for these divergencies at the 
present time. However, considering the consistency of the new results 
which were obtained on the same strains of animals, we believe we have 
proved that there exist such cis-trans isomeric differences which do not 





PROVITAMIN A ACTIVITY. VIII 


245 


affect the biopotencies in the rat. The potencies in the several groups 
(starting from lowest dosage) were as follows: all-frans- 7 -carotene 
(ex pro- 7 ) 42.7, 40.6, and 41.8; pro- 7 -carotene, 42.7, 38.0, and 41.7; 
and alWrans-y-carotene (ex tomato), 47.0 and 40.6. Moreover, 83% of 
of the group of negative controls lost weight or had died before the 
eleventh day and only two survived and these lost an average of 11.0 g. 
This would indicate that the dietary regime was vitamin A-free and 
that the rats used were sufficiently depleted of vitamin A at the start 
of the test. At least in this animal, the naturally-occurring poly-cis 
provitamin A, pro- 7 -carotene is as potent as its poly-frans form, termed 
7 -carotene. 

Summary 

The earlier reports which showed that pro- 7 -carotene has a biolo¬ 
gical potency in rats equal to 44% of that of /3-carotene have been con¬ 
firmed, since the new results are calculated at 41%. All-Jrans-y-carotene 
obtained by iodine catalysis of naturally-occurring pro-y-carotene was 
found to have a provitamin A activity of 42%. It has been demon¬ 
strated that, in this case, a poly-m provitamin A and its poly-/rans 
stereoisomer are biologically equivalent in the rat. 

References 

1. Deuel, II, J., Jr., Greenberg, S. M., Straub, E., Fukui, T., and Zechmeister, 

L., Arch. Biochem. 23, 230 (1949). 

2. Deuel, H. J., Jr., Hendrick, C., Straub, E., Sandoval, A., Pinckard, J. H., 

and Zechmeister, L., ibid . 14, 97 (1947). 

3. Deuel, H. J., Jr., Johnston, C., Meserve, E. R., PolgAr, A., and Zechmeister, 

L., ibid. 7, 247 (1945). 

4. Deuel, H. J., Jr., Johnston, C., Sumner, E., PolgAr, A., Schroeder, W. A., 

and Zechmeister, L., ibid. 5 , 365 (1944). 

5. Deuel, H. J., Jr., Johnston, C., Sumner, E., PolgAr, A., and Zechmeister, L., 

ibid. 6, 157 (1945). 

6. Deuel, II. J., Jr., Johnston, C., Sumner, E., PolgAr, A., and Zechmeister, L., 

ibid. 5, 107 (1944). 

7. Deuel, H. J., Jr., Meserve, E. R., Sandoval, A., and Zechmeister, L., ibid. 10, 

491 (1946). 

8. Zechmeister, L., and Schroeder, W. A., ibid. 1, 231 (1942). 



Comparative Study of the Glycolysis and ATP-ase 
Activity in Tissue Homogenates 1 

Otto Meyerhof and Jean R. Wilson 

From the Department of Physiological Chemistry, School of Medicine, 
University of Pennsylvania, Philadelphia, Penna. 

Received May 2, 1949 

In some foregoing papers (1,2) the balanced activities of the enzymes 
of the glycolytic cycle were studied. It was, for instance, shown that thp 
high rate of glycolysis, obtained in centrifuged extracts of brain in the 
absence of added phosphate donors, was due to a removal of the excess 
of ATP-ase 2 with the particulate matter. Only under these conditions 
are the activities of hexokinase and ATP-ase, the enzymes mainly 
responsible for phosphorylation and dephosphorylation of the sugar 
intermediates, kept in step. In the uncentrifuged homogenate, on the 
other hand, the ATP-ase is so far in excess that the glycolysis rapidly 
goes down to very low values, owing to the irreversible dephosphory¬ 
lation of ATP. This lack of balance can be overcome either by adding 
lafge amounts of phosphate donors (HDP, phosphocreatine) or by 
repeated additions of ATP itself. 

Recently, it has been shown that the situation in tumor homogenates 
and extracts is somewhat different (3); on the one hand, most of the 
ATP-ase remains in solution when the particulate matter is centrifuged 
out. On the other hand, the dissolved ATP-ase of tumor is strongly in¬ 
hibited by some higher members of the narcotic series, while the other 
glycolytic enzymes remain unaffected. This peculiarity can again be 
used to produce a very high and steady glycolysis in homogenates and 

1 This work was aided by grants from the American Cancer Society, recommended 
by the Committee on Growth; the David, Josephine and Winfield Baird Foundation; 
the Division of Research Grants and Fellowships of the National Institutes of Health, 
United States Public Health Service; and the Rockefeller Foundation. 

’Abbreviations: ATP = adenosinetriphosphatc; ATP-ase *= enzyme splitting the 
first labile group of ATP; pyro P, labile P of ATP; HDP = hexosediphosphatc; 
DPN = diphosphopyridine nucleotide = cozymase. 


246 



TISSUE HOMOGENATES ACTIVITIES 


247 


extracts of tumors in the absence of P donors. With octyl or decyl 
alcohol or toluene, Ql& values (mm, 1 CO 2 driven out by lactic acid/hr. 
and mg. dry weight of tissues at 38°C.) of 50 to 70 are obtained nearly 
constant for 60-80 min.* 

The question arose whether similar effects can be produced in brain 
homogenate and other tissue homogenates. It is shown in the following, 
that some of the narcotic substances give at least a partial inhibition 
of the adsorbed ATP-ase of the brain homogenate and induce in this 
way a temporary high rate of glycolysis which usually falls off after 
40-50 min. 

On the other hand, the homogenate of chicken embryo contains only 
a small excess of ATP-ase over hexokinase. Consequently, glucose 
alone is glycolyzed quite well, with Ql 0 values ranging from 10 to 25. 
The ATP-ase is not inhibited but is mostly activated by the narcotic 
substances in question; it is, however, inhibited by sodium azide. 
Sodium azide, therefore, can serve to stabilize glycolysis for at least 
80 min. Some additional experiments with tumor homogenates and 
ATP inhibitors are reported. 

Methods and Procedures 

The methods were the same as in the preceding papers. Lactic acid was determined 
manomctrieally by the Warburg technique, using bicarbonate solution, with 95% 
X 2 - 5 % COj in the gas space (1,2,3). Brain tissue was used only as homogenate. 
Chicken embryos, between 5 and 11 days, were obtained from West Chester Farm.* 
The water content of the embryos was between 92 and 94%. Before homogenization 
the embryos were frozen at — 12°C. because this allows a better extraction of the 
glycolyzing enzymes (see 5). 

The same narcotics were used as in the preceding papers. Moreover, digitonin and 
some other lipide solvents were tested also. 

Results 

1. Brain Homogenate 

Dissolved brain ATP-ase is not only not inhibited but is actually 
activated by octyl alcohol during the first minutes of incubation (3). 

' In a recent paper of LePage (4), on the glycolysis of tumor homogenates, most of 
the experiments were done in the presence of fluoride. Therefore, only the oxidation- 
reduction step and the phosphorylation of glucose occured. Moreover, so much HDP 
was added, that 75% of the maximum amount of lactic acid was produced even in the 
absence of glucose (see l.c. p. 1014, Table III). In our experiments, only 15 7 P of 
HDP = 11 mm.* COj are added to prime the reactions. 

4 George F. Shaw, West Chester, Pa. 



248 


OTTO MEYERHOF AND JEAN R. WILSON 


With adsorbed ATP-ase no effect was found in experiments of short 
duration. If the incubation is prolonged, and if sugar is present, which in 
itself diminishes the activity of the ATP-ase by transphosphorylation, 
a slight inhibition occurs, somewhat more with decyl alcohol then octyl 
alcohol. Simultaneously, the transphosphorylation to glucose is in¬ 
creased (Table I). 


TABLE I 


Inhibition of ATP-ase of Brain Homogenate by Higher Alcohols 
(0.3 cc. homogenate in 1 cc. total volume, temperature 38°C.) 


No. 

Time® 

Glucose 

Addition 

ATP-ase inhibitor 

InorRanio P 
split off 

Pyro P 
present 

B 

ATP-ase 

inhibition 


min. 



7 


7 

per cent f 

518 

30+0 

- 



136 




30+6 



58.7 


-4.5 



30+6 

- 

Decyl alcohol 

58.7 



0 


30+6 

+ 


43.7 


-20.4 



30+6 

+ 

Decyl alcohol 

32.4 


-27.4 

26 

590 

0 

- 



133.6 




6 

— 


64.4 


-2 



6 

— 

Octyl alcohol 

70.1 



0 


6 

+ 


40.1 


-28 



6 

+ 

Octyl alcohol 

33.8 


-37 

15 


6 • 

+ 

Decyl alcohol 

31.6 


-37 

21 

590a 

30+6 

— 

! 

55.1 


-6 



30+6 

— 

Octyl alcohol 

55.1 



0 


30+6 

+ 


42.6 


-20 



30+6 

+ 

Octyl alcohol 

33.1 

j 

-27 

22 


30+6 

+ 

Decyl alcohol j 

30.4 


-25 

29 


* 30 + means incubation of the enzyme with the inhibitor for 30 min. at 38°C. 
before adding ATP. 


In this way a high rate of glycolysis can be obtained (see Figs. 1 and 
2 for the effect of octyl alcohol, decyl alcohol and toluene). Because the 
inhibition of ATP-ase is only about 30%, the effect is not maintained for 
longer than 40 min. However, by adjusting the amount of homogenate, 
one can find conditions where the rate is constant or rises during 60- 
80 min. (see Fig. 1, Curve G + D (2/3)). The Ql« values 40-50 cor¬ 
respond to those formerly found in centrifuged extracts under favorable 












TISSUE HOMOGENATES ACTIVITIES 


249 


Fig. 1. Glycolysis of complete 
brain homogenate. In main com¬ 
partment 0.3 cc. homogenate (10 
mg. dry weight). 0.1 cc. M/10 
phosphate, 0.1 cc. 1.3% NallCOs, 
HDP with 15 7 P. Total volume, 
1 cc. Side arm: 0.1 cc. ATP with 
60 7 pyro P, 0.15 cc. cozymase 
= 0.6 mg. DPN, 0.05 cc. NallCOs 
1.3%, tipped in at 0 min. 

B, blank without glucose; (1, 
4 mg. glucose; G + T, glucose 
+ toluene; G + D, glucose + 
decyl alcohol. Dotted line, G*f D 
(2/3), glucose -f decyl alcohol, but 
only 0.2 cc. of homogenate. 



conditions. Smaller increases of glycolysis are obtained with saturated 
lauryl alcohol, 0.1% digitonin, and 0.2% sodium azide. Some such 
experiments are reproduced in Table II. 


2 . Homogenate of Chicken Embryo 

In the former work on glycolysis of embryonic extracts (5), stress was 
laid on the oxidation-reduction step and on the transformation of HDP 
into lactic acid. After recent studies had revealed several conditions 
necessary for the metabolism of free sugar in the absence of larger 
amounts of phosphate donors, we applied this knowledge to the homog¬ 
enate of chicken embryo. 

The ATP-ase of this homogenate is almost completely soluble: 
75% of the total ATP-ase of the homogenate remains in the extract 



Fig. 2. Glycolysis of brain ho¬ 
mogenate. Mixture like that of 
Fig. 1. B, blank without glucose; 
G, glucose alone; G + O, glucose 
+ octyl aclohol; G+D, glucose 
4-decyl alcohol. 





250 


OTTO MEYERHOF AND JEAN R. WILSON 


TABLE II 


Ql, Values with Glucose, of Brain Homogenate with Various ATP-ase 
Inhibitors Calculated from Ifi min. 88°C. 

(Blank value without glucose subtracted.) 


No. 

Inhibitors used 

Cone. 

QLu 

516 

_ 


6.2 


Octyl alcohol 

Satd. 

36.8 


Decyl alcohol 

Satd. 

54.7 


Lauryl alcohol 

Satd. 

10.5 

535 

— 


11.7 


Decyl alcohol 

Satd. 

42.2 


Digitonin 

0.1% 

18.6 


Sodium azide 

0.3% 

16.2 

536 

— 


10.5 


Decyl alcohol 

Satd. 

45.2 


Toluene 

Satd. 

37.2 


Digitonin 

0.12% 

16.5 


Sodium azide 

0.2% 

19.0 


after centrifugation. This dissolved ATP-ase is activated by octyl 
alcohol (30-40% for incubation time of 8 min.), quite similarly to the 
dissolved ATP-ase of brain. It is weakly inhibited by digitonin (0.1%) 
but more so by 2 X 10~ 2 M sodium azide (35-55%). The Qp values of 
ATP-ase are 30-45 in the homogenate. If there were an excess of ATP- 
ase over hexokinase, only azide could be expected to raise or stabilize 
the glycolytic rate of free -sugar. This was actually the case. While 
octyl alcohol inhibited the glycolysis increasingly, probably by activa¬ 
tion of the ATP-ase, azide could keep it in a steady state when it 
would otherwise slowly decline. In some experiments the decline during 
the first 20 min. was very strong and the azide effect pronounced. In 
such an experiment the following mm. 3 C 02 were developed for the time 
of 20-50 min after the start of the experiment and after subtracting the 
blank value: 


Glucose alone 

mmJCOt 

25 

Glucose with sat. octyl alcohol 

5 

Glucose with 2 X 10“ 3 M digitonin (0.25%) 

56 

Glucose with 2 X 10~ 2 M azide 

115 



TISSUE HOMOGENATES ACTIVITIES 


251 


In most of the experiments, however, especially if 1 X 10~ 2 M 
pyruvate was present, glycolysis remained nearly constant more than 
an hour without azide, with Qta between 10 and 25. If the activity was 
relatively low, the addition of crystallized yeast hexokinase 5 gave a 



Fir*. 3. Glycolysis of homogenate of chicken embryo (587). 7 day old embryo. 
Main compartment 0.5 cc. homogenate (12 mg. d. w.), 0.1 cc. M/ 10 phosphate, 0.1 
ec. Af /10 pyruvate, 0.1 cc. 1.5% KHCOs, 0.05 cc. M /10 glutathione, HDP with 
15 7 P filled to 1.35 cc. with 1.15% KC1. Side arm, 0.15 cc. ATP with 90 y pyro P* 
0.1 cc. DPN (0.4 mg.), 0.05 cc. 1.5% KIICOs, tipped in at 0 min. 

B, blank without glucose. Dotted line, G glucose alone (0.25%); G + A •—• = 0.25% 
glucose + 2.5 X 10 ' 2 M sodium azide, 1.25% F + A ▼—▼ = 1.25% fructose 
+ 2.5 X 10~* M sodium azide (lower values of the same curve); 0.25% F + A 
= 0.25% fructose + 2.5 X 10 “ 2 M sodium azide; Hex.: glucose 0.25% with addition 
of 10 7 yeast hexokinase. 

considerable increase of activity. In Fig. 3, 10 y of dry hexokinase 
powder was used which consists partly of protein and partly of 
(NH 4 ) 2 S 04 . This shows that, at least in these cases, the hexokinase is 
the enzyme controlling the rate. However, in other experiments, which 

6 We t hank Dr. M. Kunitz for a sample of twice crystallized yeast hexokinase. 




252 


OTTO MEYERHOF AND JEAN R. WILSON 


are summarized in Table III, the Ql» value of 25 could be obtained with 
glucose alone. In this respect it should be remembered that, according 
to Needham et al. (6), the Ql« of embryos over 5 days is only about 8 
and, with older embryos still lower. Even the lowest rate obtained in 
the homogenate is still higher than the rate of anaerobic glycolysis of 
living tissue. It can, moreover, be seen from Table III, that 0.25% 

TABLE III 


Qr.a Values in Homogenate and Extract of Chicken Embryo 
(For complete mixture see text) 


No. 

Age 

embi vos 
in days 

D.w. of 
homog. 

Prepara¬ 
tion used 

Sugar 

added 

Sugar 

ATP-ase inhibitor 

Deviations from 
standard mixtuie 

Qu- 

mmu> 

blank 



mu. 



j per 
rent 




583 

7 

8.5 

Homog. 

Glucose 

0.25 


No pyruvate 

3 





Glucose 

0.25 

Sodium azide 

No pyruvate 

9 





Glucose 

1.25 

Sodium aizde 

No pyruvate 

7 





Fructose 

0.25 

Sodium azide 

No pyruvate 

3 

586 

12 

18 

Homog. 

Glucose 

0.25 



13.2 





Glucose 

0.25 


No pyruvate 

11.1 





Glucose 

0.25 

Octyl alcohol 


11.1 








Yeast hexokinase 

34.0" 

588 

10 

10 

Homog. 

Glucose 

0.25 



25.7 





Glucose 

0.25 

Sodium azide 


22 





Glucose 

0.25 


30 7 pyro-P in¬ 

13.3 


* 



Glucose 

0.25 

Sodium azide 

stead 90 

30 7 pyro-P in¬ 

16.1 








stead 90 






Fructose 

0.25 

Sodium azide 

30 7 pyro-P in¬ 

9.2 








stead 90 


588 a 

10 

10 

Extract 

Glucose 

0.25 



14.5 





Glucose 

0.25 

Sodium azide 


14.5 



1 

l_ 

Fructose 

0.25 

Sodium azide 


8.9 


a For 60 min. expt. 
b For 30 min. expt. 


fructose in the presence of azide (F + A) gives a much lower rate than 
0.25% glucose + azide (G + A). 1.25% fructose, however, gives about 
the same rate as 0.25% glucose and the points for both are drawn on 
one curve. Galactose, which is not shown in the figure, gives the same 
as the blank. 




TISSUE HOMOGENATES ACTIVITIES 


253 


In Table III, the fortified mixture for glycolysis consisted of the following constit¬ 
uents, with 0.3-0.5 cc. homogenate: 0.1 cc. M/10 phosphate + 0.1 cc. Af/10 pyru¬ 
vate + 0.1 cc. KHCOs 1.5%, HDP with 15 y P, 0.05 cc. Af/10 glutathione, 0.2 cc. 
2% sugar, filled to 1.6 cc. with 1.15% KC1 and the other additions as stated. The 
homogenate was made with an isotonic mixture consisting of 67 parts KC1 (1.15%), 
3 parts MgCl 2 and 30 parts NaHCOj + NHJICOs. In this mixture 2% nicotinamide 
was dissolved. Cozymase, 0.4 mg. DPN and ATP with 90 y pyro P were tipped in 
from the side arm at 0 time. 

S . Additional Observations on Tumor Homogenate 

. After the foregoing papers (3) on the glycolysis of malignant tumor 
homogenates had been published, the analogy between the inhibition 
of the adsorbed ATP-ase of dried yeast and of the dissolved ATP-ase of 
malignant tumor had become more and more evident. Therefore, we 
tested with tumor homogenate those narcotics and lipide solvents 
which were able to produce the Harden-Young effect in rapidly dried 

TABLE IV 


Glycolysis of Homogenate of Rat Sarcoma 
(8 mg. d.w. of homogenate) 


No. 

I 

(JIUCOS** 

added 

ATP-ase inhibitor 

Cone, of 
inhibitor 

«u* 

minus 

blank 


mg. 





509 

— 

— 


2.7 



3 

— 


6.1 

3.4 


3 

Octyl alcohol 

Satd. 

28.4 

25.7 


3 

Benzene 

Satd. 

27 

24.3 


3 

Decyl alcohol 

Satd. 

42 

39.2 


3 

Digitonin 

0.1% 

37.8 

35.1 


3 

Na-desoxy-cholato 

0.05% 

32.5 

29.8 

513 j 

— 

— 


4.1 



3 ! 



5.7 

1.6 

3 

Decyl alcohol 

Satd. 

52.6 

48.5 

! s 

Lauryl alcohol 

Satd. 

13.9 

9.8 

i 

3 

Digitonin 

0.03% 

60.0 

55.9 

520 

— 

-- 

! 

3.8 



3 

— 


4.5 

0.7 

; 

■ 

3 

Octyl alcohol 

Satd. 

37.0 

33.2 


3 

Decyl alcohol 

Satd. 

49.0 

45.2 


3 

Digitonin 

0.08% 

41.5 

37.7 


3 

Digitonin 

0.04% 

39.0 

35.2 


Expt. 509, first 10 min., Expt. 513 and 520, first 5 min. not counted. 




254 


OTTO MEYERHOF AND JEAN R. WILSON 


yeast (7). The following solutions were all able to inhibit dissolved 
tumor ATP-ase: satd. benzene, satd. chloroform, desoxycholate 
(0.1-0.2%), and digitonin (0.05-0.25%). Tumor ATP-ase is as sensi¬ 
tive to traces of heavy metals, as had formerly been found for myosine 
ATP-ase (8). The ATP for these experiments, therefore, must be put 
through Amberlite resin to remove the traces of mercury resulting from 
the preparation of ATP. With satd. benzene and chloroform, and with 
the concentration of digitonin and desoxycholate indicated above, 
inhibitions of 40-50% are obtained. The same substances are effective 
in various degrees in eliciting a strong glycolysis by bringing hexokinase 
and ATP into step. The most suitable of these substances is digitonin, 
partly because it has no vapor pressure which may interfere with the 
exact manometric measurement. The Ql„ values of 50-60 obtained with 
digitonin are at least as high as those found formerly with octyl and 
decyl alcohols (Table IV). 


Discussion 

From the foregoing experiments in conjunction with those which 
have been previously published, it seems to follow that the balanced 
activities of hexokinase and ATP-ase are responsible for maintaining 
the steady state of glycolysis in homogenates of animal tissues. In con¬ 
trast to yeast, the hexokinase is the more sensitive enzyme and is rela¬ 
tively deficient, while ATP-ase is mostly in excess. Checking the 
latter enzyme brings both enzymes into step. In brain homogenate the 
easiest way to accomplish this is to centrifuge the homogenate, be¬ 
cause 9/10 of the ATP-ase is adsorbed on the particles. However, the 
principle which is effective injtumor, the inhibition of the ATP-ase by 
higher members of the narcotic series, can be applied in some degree to 
the adsorbed ATP-ase of brain. 

In homogenates of chicken embryos this method gives no result 
because the ATP-ase is not inhibited by these narcotics. The inhibition 
by sodium azide can be used for such a purpose. Moreover, the content 
of ATP-ase is much smaller than in brain and tumor, giving a Qp of 
about 30. Therefore, often a quite stable glycolysis is obtained from 
glucose alone, especially if pyruvate is added; this can be further sta¬ 
bilized by addition of sodium azide. 

The inhibition of the dissolved tumor ATP-ase by higher members of 
the narcotic series and lipide solvents, shows a striking parallelism to 
the effect of the same substances on the adsorbed but not on the dis- 



TISSUE HOMOGENATES ACTIVITIES 


255 


solved yeast ATP-ase. In this way we found digitonin to be an espe¬ 
cially suitable inhibitor of tumor ATP-ase. In a concentration of 
10-3 f ^ evokes a steady glycolysis of tumor homogenate with a Ql» 
of about 50. 


Summary 

The adsorbed ATP-ase of brain homogenate is inhibited 25-30% by 
(lecyl alcohol in the presence of glucose and somewhat less by octyl al¬ 
cohol. This inhibition allows a high glycolysis with a Ql» of about 50 
fo^ roughly 40 min. 

In chicken embryo homogenate the ATP-ase is not inhibited by 
these narcotics, but is inhibited about 50% by 0.02 M sodium azide. 
Because the ATP-ase of the chicken homogenate is not very much in 
excess over hexokinase, glycolysis with glucose alone, especially in the 
presence of pyruvate, is relatively stable. Addition of sodium azide 
stabilizes it still more. 

The ATP-ase of tumor homogenate is inhibited not only by higher 
alcohols, but also by other lipide solvents, like digitonin. Digitonin, in 
a concentration of 2.5 X 10“ 4 M, elicits a steady glycolysis of QLa = 50, 
similarly to decyl and octyl alcohols. 

References 

1. Meyerhof, O., and Geliazkowa, N., Arch. Biochem . 12, 405 (1947). 

2. Meyerhof, O., and Wilson, Jean R., ibid. 14, 71 (1947); 17,153 (1946). 

3. Meyerhof, O., and Wilson, Jean R., ibid. 21, 1, 22 (1949). 

4. LePage, G. A., J . Biol. Chem. 176, 1009 (1948). 

5. Meyerhof, O., and Perdigon, E., Enzymologia 8, 353 (1940). 

6. Needham, J., and Nowinski, W. W., Biochem. J . 31, 1165 (1937). 

7. Meyerhof, O., J. Biol . Chem. in press 1949. 

8. Poms, B. D., and Meyerhof, O., ibid . 169, 389 (1947). 



Candidulin: An Antibiotic from Aspergillus candidus 

P. G. Stansly and N. H. Ananenko 

From the Chemotherapy Division, Stamford Research laboratories , American Cyanamid 
Company , Stamford , Connecticut 
Received May 16, 1949 


Introduction 

An antibiotic substance having marked activity against acid-fast 
bacteria has been recovered from the fermentation liquor of an organ¬ 
ism 1 identified as a member of the Aspergillus candidus group of fungi 2 
(1). The properties of the active principle distinguish it from anti¬ 
biotic substances hitherto described, and the name “Candidulin” is 
therefore proposed for it. 

Experimental 

Production 

The mold was tested for elaborat ion of the antibiotic by the agar streak method on 
a variety of substrates, using Mycobacterium ranae as the indicator organism. The 
best activity, as determined by the zone of inhibition of M. ranae , was obtained in a 
mediunr containing glucose, glycerol, KH2PO4, NaNOs, and asparagine. Media con¬ 
taining such complex materials as corn steep, yeast extract, and soybean flour were 
completely unproductive. 

Active fermentation liquor was first produced in a neutral medium consisting of 
0.5% glucose, 1.5% glycerol, 0.2% asparagine, 0.1% NaNOj, 0.05% KH2PO4, and 
2% suspended agar. It was later found that, if asparagine was omitted from the broth, 
agar was no longer required for the elaboration of the antibiotic. In either case, 
metabolism solutions assaying 64 dilution units per ml. (tested in Kirchner’s medium 
against M. ranae ) was regularly obtained under stationary conditions after 7 days of 
incubation at 25°C. in 500 ml. Erlenmeyer flasks, 100 ml. per flask. 

1 A colony of the organism appeared on a pour plate of the vomitus of “RuppePs M 
vulture. This material was examined at the suggestion of Dr. David Rutstein, Har¬ 
vard University, and supplied through the courtesy of the Bronx Zoo, New York, N. Y 
The mold was not typical of the flora of the vomitus, since only one colony appeared on 
a single plate of a dozen or more poured. The predominant organisms present were 
yeasts. 

2 Dr. Kenneth B. Raper of the Northern Regional Research Laboratories has kindly 
examined the mold and concurs with this identification. 


256 




CANDIDULIN 


257 


Stability 

At pH 2, 5, and 7, the activity of metabolism liquor was unimpaired after 1 hr. at 
room temperature or 10 min. at 100°C. At pH 9 it was stable for at least 1 hr. at room 
temperature, but reduced after 10 mb. at 100°C. At pH 11, the activity was reduced 
after 5 min. and destroyed after 1 hr. at room temperature of 10 min. at 100°C. 

Extraction 

One volume of ether completely removed the activity from metabolism liquor at 
pH 2, 7, or 10.7, suggesting, therefore, a neutral molecule. Two or 3 extractions with 
1/10 volumes of chloroform sufficed to remove all the activity from metabolism solu¬ 
tions at their prevailing pH (7.5-7.7), 

Purification and Crystallization 

Chloroform extracts so obtained were washed with one 1/10 volume of water, dried 
with Drierite and distilled in vacuo , leavbg an oily, viscous residue contabbg the 
active principle. 

A degree of purification of this oil could be achieved by chromatographic methods. 
For example, a chloroform solution percolated through a column of MgSO* allowed 
free passage of the active prbciple while retaining certain of the colored imputities. 
On the other hand, when percolated through a column of activated alumina, the anti¬ 
biotic was firmly adsorbed, whereas certain of the impurities passed through freely. 
Other impurities could be removed by development with a solution of 10% ethanol in 
chloroform. Under these conditions, slow movement of the active principle took place 
so that, upon the addition of 7-9 column volumes (column = 9" X 3/4"), the anti¬ 
biotic appeared in the eluate. Nevertheless, chromatography was of limited prepara¬ 
tive value because there was no simple way of following the movement of the active 
principle, and because good separation from various other impurities was not readily 
achieved. It was, however, useful b working up very crude materials, such as mother 
liquors, from which further crystalline material could not otherwise be obtained. 

Routine preparation of crystalline candidulin was accomplished by 
digestion of the chloroform-free, dry residue with several portions of 
boiling n-hexane, from which the active principle separated out, on 
standing and cooling, in crystalline form. The crude, yellowish crystals 
were repeatedly Crystallized from n-hexane until white and free of 
viscous impurities. The purer the antibiotic became, the more readily 
it separated out from hexane as a voluminous mass of long, white, 
glistening, needle-like particles. The yield of purified material was 
about 5 mg./l. of medium. 

Physical and Chemical Properties 

Repeatedly recrystallized candidulin melted at 88-89°C. (hot stage, 
uncorr.) and was optically active, [a]* 4 = + 15° ± 2° (1% in 



258 


P. 0. STANSLY AND N. H. ANANENKO 


chloroform). Qualitative tests indicated nitrogen to be present and 


halogen and sulfur absent. 




Analysis: 

c 

H 

N (Dumas) 

Found 

63.47 

7.3 

6.6 


63.53 

6.9 


Calculated for CuHi 

*NOs, M. W. 209.13 




63.1 

7.2 

6.7 


The molecular weight, determined by the osmometric method (2) 
in water, was 232, indicating that the true molecular weight corres¬ 
ponds to the empirical formula given above. 

The following crystallographic properties characterized the sub¬ 
stance : 

Crystal habit: Extremely thin elongated plates with longitudinal 
striations. 

Refractive indices: N x and N y = 1.525 ± 0.003. 

Optic sign: Positive. 

Optic axial angles: 2E 113°, 2V 67° 

The material contained no titratable acid or basic groups. After 
standing at pH 10.9 and back-titrating, a fairly well-defined inflection 
appeared, corresponding to a pK of 9.4. The resulting solution was 
devoid of antibiotic activity and gave no ninhydrin or FeCl 3 tests. 

The approximate solubility (mg./ml.) of candidulin at room temper¬ 
ature was determined as follows: methanol, ethanol, acetone, ether and 
chloroform, > 6; benzene, 5; carbon tetrachloride, 4; n-hexane, 0.075; 
n-hexane at, b.p., about 1; water, 5 (slowly); 1 N HC1 and 5% NajCOs, 
insoluble or slowly soluble as in water; dilute NaOH, very rapidly 
soluble, presumably due to hydrolysis. 

Candidulin showed no ultraviolet absorption. Examination of its 
infrared spectrum, using a Nujol mull, suggested the following struc¬ 
tures to be present: methyl, internal unsaturation, carbonyl, amine and 
possibly hydroxyl. No absorption corresponding to the phenyl radical 
was present. 

The FeCl 3 , fuchsin aldehyde and ninhydrin tests were negative. No 
solid products were obtained by treatment with 2,4-dinitrophenyl- 
hydrazine, 3,5-dinitrobenzoylchloride and other acylating reagents, 
xanthydrol, and HgO. 



CANDIDULIN 


259 


Treatment with bromine in CCI 4 afforded a biologically inactive, 
crystalline, halogen-containing product in low yield. This was obtained 
as follows: 34 mg. of candidulin was dissolved in a minimal quantity of 
CCU and treated, dropwise, at room temperature' with a solution of 
4 ml. of bromine in 100 ml. of CC1 4 until the color of bromine persisted. 
The solvent was distilled in vacuo and the residue crystallized from 
ethanol-water or methyl ethyl ketone, yielding 2.G mg. of a white 
crystalline product, m.p. 143-147°C. (decompn.). The following refrac¬ 
tive indices of the columnar crystals were obtained: 

Ni parallel length = 1.065 ± 0.003, 

N 2 perpendicular length = 1.610 ± 0.003. 

TABLE I 

Antimicrobial Spectrum of Candidulin 

Inhil». oonr * 

Orpcaiimm 7 /ml. 


In Trypti case-Soy Broth 


S. aureus (152) 

320 

S. aureus (II) 

320 

Sarcina lutea 

320 

Strep, hemolyticus (( ’203) 

160 

B. polymyxa (A2) 

160 

Ii. sut)tili8 (398) 

10 

As. aeruginosa 

640 

E. coli (Maclxiod) 

160 

B. abortus (19) 

20 

K. pneumoniae (BE) 

10 

S. griseus 

320 

A . fumigatus 

80 

P. puberulum 

40 

Mycobacterium (607). 

1.25 

M. ranae 

2.5 

M. smegmatis 

2.5 

In Kirchner’s Broth h 


M. ranae 

0.4 

M. tuberculosis (H37) 

0.1 

/V. tuberculosis (D4) 

0.06 


* Observations made after an incubation period of 24 hr. at 37°C. with these ex¬ 
ceptions: S. griseus , A. fumigaius and P, puberulum, 72 hr.; the saprophytic mycobac¬ 
teria, 48 hr.; and the tubercle bacilli, 7 days. 
b Containing 0.05% Tween 80. 



260 


P. G. STANSLY AND N. H. ANANENKO 


Biological Properites 
Antimicrobial Spectrum 

Representative organisms were tested for their response to candidulin 
by serial two-fold titration in broth. The results are shown in Table I. 
It is apparent that the Mycobacteria were particularly sensitive to the 
antibiotic. 


Lethal Toxicity for Mice 

A limited number of mice (1 or 2 per dose) were injected subcutane¬ 
ously with an aqueous 0.5% solution of candidulin in doses of from 25 to 
300 mg./kg. The results suggested that the LD 60 was in the neighbor¬ 
hood of 250 mg./kg. 


Therapeutic Test 

Groups of 5 mice infected intravenously with M. tuberculosis, 
strain D4, were treated subcutaneously for a maximum of 12 days with 
daily doses of 25, 50, 100, and 200 mg./kg. beginning on the day of 
infection. The treated mice survived no longer than the control mice, 
and of those examined post mortem, all showed characteristic lesions of 
tuberculosis. It was apparent, therefore, that candidulin exerted no 
effect on the course of this experimental infection. 

Miscellaneous 

A limited investigation was undertaken to determine the possible 
fate of candidulin in the mouse when administered in a single dose 
orally (200 mg./kg.), subcutaneously, intraperitoneally or intra¬ 
venously (each at 100 mg./kg.). In no case was it possible to demon¬ 
strate antibiotic activity in the blood or urine when tested at intervals 
5 min. to 6 hr. after administration. The stomach contents of the 
mouse which received the oral dose showed appreciable activity 6 hr. 
after administration. Some activity was also present after 6 hr. at the 
subcutaneous site of injection of the mouse receiving this treatment. 
Preliminary observations indicated that incubation of candidulin with 
mouse liver, kidney, plasma, and washed erythrocytes did not affect 
activity very much, whereas whole blood had a marked neutralizing 
effect. 



CANDIDULIN 


261 


Acknowledgment 

Grateful acknowledgment is made to the following, all of these Laboratories: Dr. 
P. H. Bell and Miss K. S. Howard for the molecular weight determination and titra¬ 
tion; Mr. R. J. Francel for the infrared examination; Dr. A. F. Kirkpatrick for the 
crystallography; Dr. J. A. Kuck and staff for the microanalysis; and Dr. 11. J. White 
and staff for the chemotherapeutic test. 

Summary 

A crystalline substance having marked antibiotic activity against 
acid-fast bacteria was recovered from the fermentation liquor of a 
strain of Aspergillus candidus. The production, isolation, physical, 
chemical and biological properties of the active prinicple, for which the 
name “Candidulin” is proposed, are described. The material failed to 
influence the course of experimental mouse tuberculosis. 


References 

1. Thom, C., and Raper, K. 13., A Manual of* the Aspergilli. Williams and Wilkins, 

11)45. 

2. Bell, I 3 . If., and Howard, K. S., to be published. 



Destruction of Influenza A Virus Infectivity 
by Formaldehyde 1,2 

Max A. Lauffer and Martha Wheatley 

From the Department of Physics, University of Pittsburgh, Pittsburgh, Pennsylvania 

Received April 25, 1949 

Introduction 

The destruction of the infectivity of influenza virus is of considerable 
practical importance because influenza vaccines are prepared by inacti¬ 
vating virus. Studies on the kinetics of the inactivation of influenza 
virus by heat (1) and by the action of urea (2) have been reported 
previously. The present communication deals with the inactivation of 
influenza in the presence of formaldehyde. 

Materials and Methods 

PR8 influenza A virus, the history of which was described previously (2), was used 
in the present investigation. Preparations A through J represent the 8th to 11th and 
14th to 19th chicken embryo passages in Pittsburgh, respectively. Each virus prep- 
araiton consisted of allantoic fluid from infected chicken embryos. Potassium phos¬ 
phate buffers of ionic strength 0.2 at pH values of 4.7, 5.6, 5.8, 7.0 and 8.0, and sodium 
acetate-acetic acid buffers of ionic strength 0.2 at pH values of 5.0, 5.2 and 5.4, were 
used in these investigations. The formaldehyde was of reagent grade. 

The general method of study was to dilute 0.1 ml. of vims preparation to 10 ml. with 
buffer and formaldehyde. In most of the studies, the final formaldehyde concentration 
was 1 X 10~ 4 g./ml., but the range was from 0 to 2 X 10~ 4 g./ml. The final virus con¬ 
centrations were 1% of those in the original allantoic fluid preparations. The solution 
containing virus and formaldehyde was then heated in a water bath held at some 
particular temperature, and samples were removed periodically for analysis. Fifty 
per cent chicken embryo infectivity end-points were determined in the manner des¬ 
cribed previously (1,2). It was found that the inactivation in the presence of formalde¬ 
hyde was probably a first order reaction with respect to virus concentration. Accord¬ 
ingly, first order reaction velocity constants and their standard errors were computed 
in the manner described previously (2). 

1 Aided by a grant from the Division of Research Grants and Fellowships of the 
National Institute of Health, U. S. Public Health Service. 

1 Contribution No. 4-p-49, of the Department of Physics of the University of Pitts¬ 
burgh. 


262 




DESTRUCTION OP INFLUENZA A VIRUS 


263 


Experimental Results 

In Table I are presented specific reaction velocity constants and their 
standard errors for the destruction of infectivity .of PR8 influenza A 
virus in the presence of formaldehyde. Ten different virus preparations 
were used. The effects of three variables, temperature, formaldehyde 
concentration and pH, were investigated. The temperature range 
studied was 25-40°C.; the formaldehyde concentration range 1 was 
0-2 X 10 -4 g./ml.; and the pH interval was 4.7-8.0. 

The reproducibility of rate constants for destruction of infectivity in 
formaldehyde at 10 -4 g./ml. for different virus preparations can be 
ascertained by examining certain of the values presented in Table I. 
The standard error for the difference between two values is the square 
root of the sum of the squares of the standard errors of two values. In 
only one case, that involving preparations F and I at 25°C., pH 5.6, did 
the difference between rate constants exceed twice the standard error of 
difference. On the basis of these results, it is reasonable to conclude that 
rate constants are reproducible, within experimental error, from prep¬ 
aration to preparation. 



Fig. 1. Natural logs of reaction velocity constants for the destruction of PR8 
influenza A virus infectivity plotted as ordinates against reciprocals of absolute 
temperature. Top graph—pH8; middle graph—pH 7; bottom graph—pH 5.6. 
Formaldehyde concentration was 1 part/10,000. 



264 


MAX A. LAUFFEH AND MARTHA WHEATLEY 


TABLE I 

The Destruction of Influenza Virus Infedivity in the Presence of Formaldehyde 


pH 

Formaldehyde 
concentration 
g./ml. X10 4 

Temperature 

°C. 

k min 1 

SE k 

Preparation 

8.0 

0.5 

25 

0.25 

0.02 

D 

8.0 

1.0 

25 

0.44 

0.09 

D 

8.0 

1.0 

30 

2.21° 

0.39 

c 

8.0 

1.0 

30 

0.71 

0.09 

D 

8.0 

1.0 

35 

1.77 

0.39 

C 

8.0 

1.0 

35 

1.70 

0.14 

D 

8.0 

1.0 

40 

1.72 

0.20 

D 

8.0 

2.0 

25 

0.74 

0.07 

D 

7.0 

0.0 

25 

0 

— 

J 

7.0 

0.5 

30 

0.34 

0.05 

A 

7.0 

1 

25 

0.24 

0.04 

E 

7.0 

1 

25 

0.17 

0.04 

J 

7.0 

1 

30 

0.51 

0.05 

A 

7.0 

1 

30 

0.56 

0.06 

E 

7.0 

i 

35 

0.04 

0.09 

A 

7.0 

1 

35 

0.82 

0.08 

E 

7.0 

1 

40 

1.09 

0.12 

E 

7.0 

2 

30 

1.15 

0.25 

A 

5.8 

1 

25 

0.08 

0.02 

G 

5.0 

0.0 

25 

0.11 

0.03 

I 

5.6 

0.5 

25 

0.06 

0.01 

D 

5.6 

1 

25 

0.10 

0.01 

D 

5.6 

1 

25 

0.06 

0.02 

F 

* 5.6 

1 

25 

0.13 

0.01 

I 

5.6 

1 

30 

0.16 

0.05 

C 

5.6 

1 

w 35 

0.32 

0.05 

C 

5.6 

2 

25 

0.17 

0.02 

D 

5.4 

1 

25 

0.06 

0.02 

G 

5.2 

1 

25 

0.20 

0.04 

G 

5.0 

0 

25 

0.41 

0.09 

H 

5.0 

1 

25 

0.49 

0.10 

H 

5.0 

1 

25 

0.43 

0.06 

G 

4.7 

0 

25 

0.65 

0.13 

J 

4.7 

1 

25 

0.65 

0.11 

J 

Unbuffered 

1 

25 

0.13 

0.02 

B 

Unbuffered 

l 

30 

0.46 

0.07 

B 

Unbuffered 

1 

35 

0.78 

0.14 

B 

Unbuffered 

1 

40 

2.76 

.0.46 

B 


° This value was not used in any comput ations because it is obviously inconsistent 
with the pattern established by remaining data. 


DESTRUCTION OF INFLUENZA A VIRUS 


265 


Some of the data of Table I are grouped in Fig. 1 in a manner to 
show the way in which rate constants vary with temperature. Natural 
logarithms of rate constants are plotted against reciprocals of absolute 
temperature, according to the Arrhenius equation, for studies carried 
out at pH 5.6, 7.0 and 8.0. It can be seen that the data fall reasonably 
well upon straight lines. Energies of activation, and entropies of activa¬ 
tion, can be computed in the manner described previously (1,2.) The 
energies of activation were 23,000, 24,000 and 19,000 cal./mole, and 
the entropies of activation were 11,18 and 0 cal./mole deg., respectively 
for the reactions at pH 5.6, 7.0 and 8.0. The differences between these 
energies of activation and entropies of activation at the different pH 
values are of doubtful significance, for each figure is subject to consider¬ 
able experimental error. The important thing is that, for the destruction 



lo8 e (cxl0 4 ) 

Pig 2. Natural logs of reaction velocity constants for the destruction of PR8 
influenza A virus infectivity plotted as ordinates against natural logs of formaldehyde 
concentration expressed in terms of'parts/l 0,000. Upper graph—pH 7, 30°C ; lower 
graph—pH 8, 2f)°C 

of PR8 influenza A virus infectivity in formaldehyde solutions at a 
concentration of 10 -4 g./ml., the energy of activation is about 20,000 
cal./mole and the entropy of activation is approximately 10 cal./mole 
deg. at pH values between 5.6 and 8.0. 

In Fig. 2, natural logs of specific reaction rates are plotted against 
natural logs of formaldehyde concentrations for studies carried out at 
30°C., pH 7.0, and at 25°C., pH 8.0. In both cases, the data fell on 
straight lines with slopes of approximately 0.8. This means that the 




266 MAX A. LAUFFER AND MARTHA* WHEATLEY 

reaction velocity constants are proportional to the formaldehyde con¬ 
centration raised to approximately the 0.8 power. 

The effect of variation of pH upon the specific reaction velocity con¬ 
stants is shown in Figs. 3 and 4. In Fig. 3, logs of specific reaction rates 



PH 

Fig. 3. Logs of specific reaction velocity constants plotted as ordinates against 
pH. Upper graph—35°C.; middle graph—30°C.; lower graph —25°C. Formaldehyde 
concentration was 1 part /10,000. 



Fig. 4. Reaction velocity constants plotted as ordinates against pH for range from 
4.7 to 8.0. Broken line and open circles represent rates in absence of formaldehyde. 
Dotted line and closed circles represent rates in presence of 1 part formaldehyde/ 
10,000. Solid line represents sum of the other two. 


DESTRUCTION OF INFLUENZA A VIRUS 


267 


are plotted against pH, in the range 5.6-8.0, for formaldehyde concen¬ 
trations of 10 -4 g./ml. at 25, 30 and 35°C. The slopes of the three 
graphs are all approximately 0.3. This would indicate that the specific 
reaction rate for the destruction of influenza A virus infectivity in 
formaldehyde in media alkaline to the isoelectric point is proportional 
to the reciprocal of the hydrogen ion activity raised to the 0.3 power. 
In Fig. 4, reaction velocity constants at 25°C. are plotted against pH 
for studies carried out in the absence of formaldehyde and in 10 -4 g. for¬ 
maldehyde/ml. The pH range covered was from 4.7 to 8.0. It can be 
observed that the rate of destruction of infectivity in the presence of 
formaldehyde has a minimum value under the conditions of this experi¬ 
ment at a pH value of about 5.6, and also that the rate of destruction of 
infectivity in media more acid than pH 5.6 is the same in the presence 
of formaldehyde as in the absence of formaldehyde. 

Discussion 

The kinetics of the reaction have been investigated. According to 
most current concepts of reaction kinetics, the way in which an agent 
can accelerate a reaction is to combine with the reactant to form an 
intermediate product. There is chemical evidence (3,4) that formalde¬ 
hyde is capable of reaction with amine, amide, indole and guanidine 
residues. If the enhanced reaction rate observed in the presence of 
formaldehyde is the result of the formation of an intermediate com¬ 
pound with formaldehyde, the data of the present study indicate some¬ 
thing of the possible nature of that reaction. 

It is a necessary deduction from the above postulate, namely, that 
formaldehyde speeds up the destruction of virus activity by the forma¬ 
tion of an intermediate compound which then loses infectivity, that 
the rate of loss of infectivity should be proportional to the fraction of 
the virus particles which exist as virus-formaldehyde complex at any 
one instant. According to the mass action law, this fraction should be 
equal to the product of the equilibrium constant for complex formation, 
and the formaldehyde concentration raised to the nth power, where n 
is the number of formaldehyde molecules which react with one virus 
particle to form the complex. It follows that, if these postulates apply, 
the rate of the reaction should be proportional to the formaldehyde 
concentration raised to the nth power. The data of Fig. 2 show that the 
rate of destruction is proportional to the formaldehyde concentration 
raised to the 0.8 power. If it is assumed that this figure does not differ 



268 


MAX A. LAUFFER AND MARTHA WHEATLEY 


significantly from 1, then these data tend to show that the reaction 
between virus and formaldehyde involves only a single formaldehyde 
molecule for each reactive group on a virus particle. Thus, one might 
say that a particular group, or, perhaps, one of several particular 
groups, on a virus particle can react with one molecule of formaldehyde 
to form a complex which loses infectivity more rapidly than normal 
virus. 

The chemical groups which might conceivably react with formalde¬ 
hyde exist in nonionized form in alkaline solutions but are capable of 
capturing protons to form positively charged groups in acidic solutions. 
These groups seem to react in the nonionized form (4) with formalde¬ 
hyde. If this is so, one should expect the equilibrium between virus- 
formaldehyde complex and unaltered virus to depend upon the pH of 
the medium; the higher the pH, the higher the fraction in the complex 
state and, consequently, the higher the rate of the desturction of in¬ 
fluenza virus activity. The data of Fig. 3 show that this prediction 
holds over the pH range 5.6-8.0. However, the data of Fig. 4 show that 
the reaction has a minimum rate at approximately pH 5.6. The iso¬ 
electric point of PR8 influenza A virus has been shown to be at pH 5.3 
(5). It is tempting to dismiss the approximate coincidence between the 
isoelectric point and the pH value of minimum destruction rate by the 
generalization that many biologically active materials are most stable 
at their isoelectric points. However, it was shown (1) that, in the ab¬ 
sence, of formaldehyde, the destruction of activity proceeds at a mini¬ 
mum rate at about pH 8 and at faster rates in media with lower pH 
values. This knowledge suggested the possibility that two reactions are 
involved in the destruction of infectivity in the presence of formalde¬ 
hyde. The first of them involves reaction with formaldehyde to form a 
complex which then loses infectivity. The second does not involve 
reaction with formaldehyde, but is simply the thermal inactivation of 
the virus. Accordingly, it can be assumed that, at pH 8 in the presence 
of formaldehyde, the first reaction proceeds at an appreciable rate at 
25°C. and the second reaction proceeds at negligible rate. As pH is 
lowered, the rate of the first reaction decreases and the rate of the 
second reaction increases. The results of the present study can thus be 
accounted for by assuming that, at pH 5.6, or thereabouts, the rate of 
the second reaction becomes greater than that of the first, and that, at 
pH values below this, the variation of rate with pH depends upon the 
variation of thermal inactivation with pH. If this assumption is correct, 



DESTRUCTION OF INFLUENZA A VIRUS 


269 


then the following should be observed: (a) the rate of reaction at pH 
values substantially below 5.0 should be the same in the absence of 
formaldehyde and in the presence of formaldehyde; (b) at pH values 
substantially above 5.6, the reaction rate should be considerably 
greater in the presence of formaldehyde than in the absence of formal¬ 
dehyde. Data shown in Fig. 4 confirm this prediction, for the rates of 
inactivation at pH 4.7 and pH 5.0 are the same in the presence of 1 
part/10,000 of formaldehyde and in the absence of formaldehyde, but 
at pH 7 the reaction in the presence of formaldehyde proceeds at a 
measurable rate while that in the absence of formaldehyde proceeds 
too slowly to be measured. The possiblity also exists that, at some pH 
values much higher than 8, the recation which accounts for the thermal 
inactivation at high pH values (l) might proceed faster than that 
involving formaldehyde. 


Summary 

1. The inactivation of PR8 influenza A virus in the presence of 
formaldehyde was investigated. 

2. The reaction is of the first order with respect to virus. 

3. The energy of activation is approximately 20,000 cal./mole and 
(lie entropy of activation is approximately 10 cal./mole deg. 

4. At pH values between 7.0 and 8.0, the rate of inactivation is 
proportional to the formaldehyde concentration raised to approxi¬ 
mately the 0.8 power, but at pH values below 5.6, the rate is indepen¬ 
dent of formaldehyde concentration. 

5. Between pH 5.6 and 8.0, the rate of inactivation decreases as pH 
decreases. Below pH 5.6, the rate of reaction increases as pH decreases. 
The reaction below pH 5.6 was shown to be the same as the normal heat 
inactivation of the virus. 

6. The data are consistent with the interpretation that, in the pH 
range studied, two parallel reactions contribute to the loss of infcctivity. 
One involves the reaction of one molecule of formaldehyde with the 
nonionized form of an ionizable group on a virus particle to form a com¬ 
plex which then loses activity readily. This reaction predominates at 
pH values above 5.6. The other reaction is the normal thermal inacti¬ 
vation process which does not involve reaction with formaldehyde. 
This process predominates at pH values below pH 5.6. 



270 


MAX A. LAUFFER AND MARTHA WHEATLEY 


REFERENCEwS 

1. Lauffer, M. A., Carnelly, H. L., and MacConald, E., Arch. Biochem. 16, 321 

(1948). 

2. Lauffer, M. A., Wheatley, M., and Robinson, G., in press. 

3. Olcott, H. S., and Fraenkel-Conrat, H., Chem. Revs. 41, 151 (1947). 

4. French, D., and Edsall, J. T., Advances in Protein Chem. 2, 277 (1945). 

5. Miller, G. L., Lauffer, M. A., and Stanley, W. M., J. Exptl. Med. 80, 549 

(1944). 



An Unknown Effect of Amino Acids. II; Interaction of 
Nitrogenous Polycarboxylic Acids (N-Substituted 
Amino Acids) and Insoluble Metal 
Sulfides and Mercaptides 

Ines Mandl and Carl Neuberg 

From the Institute of Polymer Research, Polytechnic Institute of Brooklyn, 
and the Research Laboratories, Interchemical Cory., N. Y. 

Received June 6,1940 

Introduction 

Amino acids and related substances in neutral or weakly alkaline 
solution have been shown (1) to prevent precipitation of the insoluble 
sulfides of many metals, and also to redissolve freshly precipitated 
sulfides. The same was found true for mercaptides. The importance of 
this phenomenon is due to the fact that it applies to substances widely 
distributed in nature: protein degradation products, biometals, and 
thiol compounds occurring during metabolism. 

A possible explanation of the phenomenon may involve formation of 
a complex between metal and amino acid of sufficient stability to pre¬ 
vent precipitation with H 2 S or to effect a double decomposition of the 
precipitated metal sulfide on interaction with excess amino acid in 
weakly alkaline solution. 

Some understanding of these reactions might be gained by a study of 
the behavior of certain nitrogenous polycarboxylic acids such as eth- 
ylenediaminetetraacetic acid (COOHCH 2 ) 2 :N-CHrCIIrN:(CH 2 - 
COOH) 2 and trimethylaminetricarboxylic acid (COOH-CH 2 ) 2 :N- 
CHj-COOH. These substances have not been found in nature—at 
least as yet—but their great tendency to complex formation was 
established by Ender, Brintziger, Pfeiffer, Schwarzenbach, and others 
(2). They may be regarded as iV-substituted amino acids or glycine 
derivatives. 

The results tabulated below show that solutions of both compounds 
at a pH of approximately 8 prevent the precipitation of the sulfides of 
Zn, Co, Ni, Mn, Fe ++ , Fe +++ , VO ++ . In this respect they resemble 


271 



272 


INES MANDL AND CARL ^EUBERG 


natural amino acids. They also prevent sulfide precipitation of lead, 
the biological significance of which has recently been pointed out (3). 
On the other hand, they do not prevent precipitation of CuS under 
conditions in which ordinary amino acids were found effective and 
their ability to redissolve CoS is negligibly small. The sulfides of Ag, 
Cd, Hg, Tl, and Bi are precipitated in the presence of both reagents. 
In spite of the great tendency to complex formation of these nitrog¬ 
enous polycarboxylic acids, and although Ag, Cu, Cd, Hg, and Bi 1 
are known to form complexes, these criteria do not allow a prediction 
of the behavior of these systems toward II 2 S or mercaptans. Quite 
generally, a tendency to complex formation per se will not inhibit pre¬ 
cipitation as the sulfide. A well-known example applied in inorganic 
analysis is the resistance of many metal complexes toward alkali 
hydroxides. Some metals, which cannot be precipitated as hydroxides 
in the presence of glycerin, tartrates or citrates, can still be precipitated 
as sulfides, e. g., Fe, Cu, Bi, Co, Ni. 

In this connection, reference may be made to the possible feasibility 
of using nitrogenous polycarboxylic acids in the same way as poly¬ 
phosphates (5) and nucleinates (f>) when the effectiveness of the bio¬ 
elements is to be assured in enzyme systems, in metalloproteins or 
otherwise, and deionization must be retained to guard against inhibi¬ 
tion by thiol compounds. 


Experimental 


The exact procedure and all necessary precautions as well as the 
advantages of Na 2 S as precipitating agent have been fully described 
in a previous publication (1). 


The trimethylamine tricarboxylic acid (Eastman-Kodak) was purified by solution 
iu alkali and reprccipitation with HC1, or by re crystallization. A molar solution of the 
disodium salt in water, pH approximately 7, was used. In the case of the ethylene- 
diaminetetraacctic acid, which can be purified in the same way M solutions of the 
trisodium salt, pH approximately 8.5, have been applied. 

In all experiments il//100 solutions of the metal salts were precipitated with 0.5 ec. 

CHjN 


AT/10 NajS or .V/10 solutions of the Na salt of ethylene isothiourea, 



CH,-N 

H 


1 This has been specifically established in the case of Bi triglycolamate (4). 



AMINO ACIDS 


273 


as a convenient mercaptan-like substance, and redissolved by means of 2 cc. of the 
nitrogenous polycarboxylate solutions. Control experiments were made in all cases. 
They showed that 1 cc. of A//100 metal salt solution was required to obtain a definite 
ethylene isothiourea precipitate and also for CuS and ZnS. One-tenth cc. proved 
sufficient for the precipitation of all other metal sulfides. One cc. of M /10 Zn salt 
was used for the Zn ethylene-iso thiourea precipitation. 

Freshly precipitated CoS could not be redissolved completely, but precipitation 
could easily be prevented when Na 2 S was added to the Co salt in the presence of 
either polycarboxylate. 

The ethylenethiourea was dissolved in an equimolecular quantity of NaOH and a 
few drops methanol added to facilitate solution. 

*Xanthogenates, which were investigated in some cases, were similarly redissolved. 

In all cases, identically the same results were obtained with the 
disodium trimethylaminctricarboxylate and the trisodium ethylene- 
tetraacetate solutions. 

Unless the contrary is specifically indicated, all solutions tabulated 
below remained clear for days. 



NajS 

Na ethylene- 
iaothiourea 

K ethyl xunthogenatc 

Zn 

Clear on warming 

Clear 

Clear 

Ni 

Clear 

Clear 

Clear 

Co 

Prevents only 

Clear 

Prevents only 

Mn 

Clear 

Clear 

a 

Fe++ 

Clear on warming or standing 

Clear 

a 

Fe +++ 

Clear on warming and standing 

— i 

Clear 

Pb 

Clear on warming for short time only 

Clear 

Clear 

Cu 

No effect 

Clear 

Clear on warming 

T1 

No effect 

Clear 


Ag 

No effect 

Clear 


Cd 

No effect 

Clear 


Hg 

No effect 

Clear 


Hi 

No effect 

Clear 


VO ++ 

Clear 6 

Clear 


IJ0 2 ++ 

Clear on standing 

Clear 



° Lack of a definite precipitate prevented making satisfactory tests with these 
substances. 

6 The grayish brown precipitate formed on addition of Na 2 S to vanadyl sulfate, 
dissolved without difficulty in both reagents. Whether a true sulfide is involved must 
be left, however, in abeyance, since a precipitate is also formed when NaOH is added 
to the blue vanadyl compound. This precipitate too dissolves in both reagents, 
though with a different color. 


274 


INES MANDL AND CARL NEUBERG 


Acknowledgment 

The authors wish to thank Rohm & Haas Co., Philadelphia, Bersworth Chemical 
Co., Farmingham, Mass., and Alrose Chemical Co., Providence, R. I., for supplying 
them with chemicals used in this investigation. 

Summary 

iV-substituted poly carboxylic amino acids are shown to prevent 
precipitation of sulfides, mercaptides and xanthogcnates, and to 
dissolve freshly precipitated thiol compounds with an effectiveness 
equal to but not greater than that of ordinary amino acids. 

References 

1. Neuberg, C., and Mandl, I., Arch. Biochem. 19, 149 (1948). 

2. Ender, W., Fette u. Seifen 45, 144 (1938); Pfeiffer, P., et at ., Ber. 75, 1 (1942); 

76, 847 (1943); Brintziger, II., et al. } Z. anorg. allgem. Chem. 249, 113, 299 
(1942); 251, 285 (1943); Schwarzenbacii, O., et al. } Helv. Chirn. Acta 26, 418, 
452 (1943); 31, 331, 459, 1029 (1948); Beck, G., ibid. 29, 357 (1946). 

3. Speck, J. F., J. Biol. Chem. 178, 315 (1949); Stickland, L. H., Biochem. J. 44, 190 

(1949); Blackburn, C. R. B., J. Biol. Chem. 178, 855 (1949). 

4. Lehman, R. A., and Sproull, R. C., J. Am. Pharm. Assoc. 31, 190 (1942). 

5. Frankenthal, L., Roberts, I. S., and Neuberg, C., Exptl. Med. and Burg. 1, 386 

(1943). 

6. Neuberg, C., and Roberts, I. S., Arch. Biochem. 20, 185 (1949). 



Biochemical Individuality. III. Genetotrophic 
Factors in the Etiology of Alcoholism 1 

Roger J. Williams, L. Joe Berry and Ernest Beerstecher, Jr. 

From the Biochemical Institute ami the Chemistry Department of The University 
of Texas and The Clayton Foundation for Research, Austin, Texas 
Received May 9, 1949 


Introduction 

Recent developments in the field of biochemical genetics (1) make it 
inescapable that each individual possesses a distinctive hereditary 
make-up and a metabolism which is distinctive as to its details because 
of differences in enzymic patterns (2,3). Partial genetic blocks (4,5) 
somewhere along the metabolic assembly line are probably common¬ 
place in the inheritance of individuals and these may give rise to aug¬ 
mented requirements for specific minerals, vitamins, amino acids, or 
other nutritional factors. On the basis of these genetic variations, it 
becomes possible for one individual to suffer from nutritional disease 
when his diet is wholly adequate from the standpoint of many other 
individuals. 

If attention is centered on the classification of those diseases in which 
no infective agent appears to be involved, nutritional diseases such as 
pellagra, rickets, scurvy and beri beri, and diseases of genetic origin such 
as hemophilia, Huntington’s chorea, etc., constitute highly important 
groups. So far as we know, however, no one has given attention to the 
existence of genetotrophic (geiieto = genetic; trophic = nutritional) 
diseases—those in which the genetic background and nutritional factors 
jointly enter into the etiology. 

We wish to present the hypothesis that numerous diseases of obscure 
etiology are genetotrophic in origin. With respect to alcoholism, we have 
strong evidence which is to be presented in this paper. Other diseases 
such as allergies, mental diseases, cardiovascular diseases, arthritis, 
multiple sclerosis, drug addiction and cancer, have not been studied 

1 This research was supported in part by grants from the Research Corporation and 
the Research Council on Problems of Alcohol, New York. 

275 



276 R. J. WILLIAMS, L. J. BERRY AND E. BEERSTECHER, JR. 

from this standpoint, but we believe that the facts warrant such a 
study and that careful consideration should be given to the possibility, 
and even probability in some cases, that genetotrophic factors are 
operative. 

That these factors enter in an important way into the etiology of a 
particular disease does not exclude the entrance also of psychogenic or 
other influences. Even diseases which are primarily due to infective 
agents may in turn be greatly affected by genetotrophic factors. An 
individual may, for example, possess a metabolic pattern which is 
conducive to a specific nutritional deficiency; this deficiency in turn 
may constitute an invitation to infection even though the available 
evidence on this point is not impressive (6). 

Alcoholism, or compulsive drinking considered as a disease, continues 
to be one for which neither a practical means of prevention nor a satis¬ 
factory treatment has been developed (7). One of us, on the basis of new 
clarification and insights gained in the field of biochemical genetics and 
on a careful survey of the pertinent literature, has outlined reasons for 
suspecting that the extreme appetite for alcohol possessed by compul¬ 
sive drinkers has a physiological basis closely linked with the inherited 
metabolic patterns of the individuals afflicted (8). Other specific 
appetites, e.g., for salt, for calcium, for phosphate, for sugar as in cer¬ 
tain diabetics, for fat, for protein, for B vitamins, all have a physiolog¬ 
ical basis, and there is ample reason for supposing that the appetite for 
alcohol has also. This hypothesis does not exclude the influence of 
psychological and social forces in alcoholism, but emphasizes that, in 
the study of the physiological aspects of acloholism, attention must be 
directed to individual metabolic patterns as distinguished from the 
hypothetical pattern common to all individuals. 

Studies with experimental animals here reported offer confirmation 
of this hypothesis and make possible a theory of alcoholism which is in 
line with all the previously known facts as well as the new findings 
presented. In addition, these studies strongly suggest the possibility 
of developing treatments, both prophylactic and therapeutic, for the 
management of this disease. Such treatments will, in the nature of the 
case, act in a constructive manner and improve the general health of 
the patient. 

Experimental 

Approximately 100 white rats and 30 mice have been placed in individual cages and 
their alcohol consumption observed. For this purpose young weanling animals have 



BIOCHEMICAL INDIVIDUALITY. Ill 


277 


generally been used. Two drinking bottles have been provided, one containing water 
and the other 10% alcohol, and the positions of the drinking bottles have been inter¬ 
changed daily. The consumption of alcohol, water and food was followed daily for 
many of the animals. However, when they were not under treatment, as much as 3 
days sometimes elapsed between readings. The food consumption records have not 
been analyzed or utilized, partly because of their unreliability due to scattering. 

In addition to carrying out the tests with animals on stock diets, two experimental 
diets have been used. Diet A contained: 


Commercial casein 
Sucrose 

Vegetable oil (Wesson Oil) 


Dried brewers' yeast (autoclaved) 
Salt mixture No. 2 


20 parts 
60 parts 

15 parts (fortified to give 60 units 
Vitamin A and 10 units 
Vitamin D per cc.) 

10 parts 
5 parts 


The yeast was autoclaved at pH 9.0 for 90 min. at 125°C. 

Diet B had the same composition except for the addition, in mg./kg. of diet, of: 


Thiamine 

1 

Pyridoxine 

1.5 

Riboflavin 

7.5 

Choline 

1000 

Calcium pantothenate 

30 

Inositol 

30 

p-Aminobenzoic acid 

30 

Folic acid (synthetic L. casei factor) 

1 

Nicotinic acid 

10 

Biotin 

0.1 


In the first experiments, animals were used from our original stock of Wistar rats 
which had been grown in our laboratory over a period of years. These were designated 
“Strain O.” Later, similar rats were obtained from a neighboring laboratory and 
designated “Strain H,” Three strains of mice were used, including (1) a dba strain 
which was closely inbred using brother-sister matings, (2) a C 3 H strain for which, 
because of breeding difficulties, brother-sister mating had been abandoned, and (3) 
a strain of white mice purchased on‘the market. 


Results and Discussion 

Inspection of the alcohol consumption records shows clearly that 
each animal tends to exhibit a distinctive pattern of response. These 
patterns, however, fall into a few general groups, and show a parallel to 
the diverse drinking patterns of human individuals. One group of rats 
on a stock diet, exemplified by Fig. 1, consumed relatively large 
amounts of alcohol rather consistently, often beginning the first day. 
Others (Fig. 2) avoided the consumption of alcohol at first but after a 



278 B. J. WILLIAMS, L. J. BERRY AND E. BEERSTECHER, JB. 


time drank more and more until after a few weeks a large part of their 
fluid intake was 10% alcohol. Still others (Figs. 3 and 4) avoided the 
consumption of substantial amounts of alcohol over a period of many 
weeks. A closer examination of the consumption curves shows them to 
be highly individual in character. In some animals the consumption was 
relatively steady, in others it tended to fluctuate widely. In Fig. 3 it 
will be noted that the consumption was consistently higher than that 



FIG. 3 


FIG. 4 

DAYS ON STOCK DIET 


Fios. 1-4. Representative 
individual responses of rats as 
shown by ad lib. selection of 
water or 10% ethyl alcohol. 


shown in Fig. 4 but continued for a long period at a very moderate 
level. Further evidence as to the distinctive character of the individual 
responses will be presented later. 

Effect of Diet in Experiments with Groups of Animals 

When groups of animals were placed upon diets A and B instead of 
stock diets, the average responses were found to be dependent upon the 
type of diet used, in confirmation of the earlier work of Mardones, 
Segovia and Onfray (9), and Brady and Westerfeld (10). These workers 
have found that groups of rats on restricted diets consume larger 
amounts of alcohol than when the restricted diets are supplemented 
with B vitamins and liver and yeast extracts. Mardones, Segovia and 



BIOCHKMICAL INDIVIDUALITY. Ill 


279 


Onfray have postulated the existence of an unknown principle, “ Factor 
N,” which controls appetite for alcohol. Brady and Westerfeld also 
observed the efficacy of a liver extract preparation in decreasing alcohol 
consumption under certain conditions, but after the alcoholic consump¬ 
tion became relatively large they were unable to observe more than a 
temporary diminution of consumption caused by its addition to the 
diet. 

In Tables I and II are included summaries of the data on alcohol 
consumption with respect to two strains of rats on diet A and diet B. 


TABLE I 

Alcohol Consumption of Rats arul Mice 


Type animal studiod 

Number of 
animals 

Diet 0 

Av. alcohol 
consumption 
25-40th day 
as cm. 3 /100 g. 
animal/day 

Range of 
alcohol con¬ 
sumption 

Mean of 
deviation 
from mean 

Percentage 
of deviation 
from mean 

Strain “0” rats 

10 

A 

.93 

.48-1.39 

.21 

23 

Strain “0” rats 

10 

B 

.17 

.07-0.29 

.05 

30 

Strain “11” rats 

41 

A 

.73 

.18-1.83 

.25 

34 

Strain “H” rats 

38 

B 

.27 

.07-0.63 

.11 

41 

Dba Strain mice 

10 

A 

.29'' 

.24-0.38 

.036 

12 

C 3 II Strain mice 

9 

A 

.46 6 

.07-1.38 

.42 

91 

White mice 

14 

A 

.50" 

.23-1.87 

.27 

54 


" The rats were on the diet and had access to alcohol 25 days prior to the period 
covered in this table, and the mice were on the regime for 5 days. 

6 The period for the mice was from the 6th to the 14th day. 

All animals were started on diets within one week from weaning. 


TABLE II 


Comparison of Two Strains with Respect to Their Maximum Alcohol Consumption 


Strain 

Number of 
rats 

Diet 

Average maxi¬ 
mum alcohol 
consumption 
during ueriod 
on diet 

Range 

Av. no. dav s 
to reach 
maximum 

Range of days 
(indiv. rats) 

“O” 

26 

A 

.94±.24 a 

.58-1.54 

53 

35-81 

“H” 

39 

A 

1.33±.35“ 

.72-2.17 

52 

35-81 

“O" 

26 

B 

.35±.14“ 

.08-0.75 

81 

62-120 

“H” 

42 | 

B 

,47±.23“ 

.03-1.00 

61 

53-87 


“ These values are the mean deviations from the mean, and not standard deviations. 



280 R. J. WILLIAMS, L. J. BERRY AND E. HEER8TECHKR, JR, 


It will be noted that diet A induces far greater alcohol consumption 
than diet B and that the maximum consumption (Table II) is reached 
earlier on diet A than on diet B. 

While a detailed analysis of the total water consumption records has 
not been made, it appears evident that ( 1 ) the total fluid consumption 
of the rats figured on a weight basis decreased with increase in weight, 
and ( 2 ) when the total fluid consumption of groups of rats of the same 
weight exhibiting high and low alcohol consumption are compared, 
there is no striking difference. 

Genetic Basis for Alcohol Response Patterns 

It has previously been shown that ordinary inbred strains of labora¬ 
tory animals are by no means homogeneous with respect to their, 
nutritional requirements, and that substrains of such animals with 
different levels of requirement for specific vitamins can be bred ( 11 , 12 ). 
From the data presented in Table 1 several pertinent observations may 
be made: ( 1 ) While “Strain 0” and“Strain II” rats both consume much 
more alcohol on diet A than on diet B, the contrast between the two 
diets is significantly greater in the case of “Strain 0” rats. ( 2 ) On diet 
A “strain O” rats consumed 27% more alcohol than did“strain H” rats. 
(3) On diet B ,however, the situation was reversed; the “strain H” rats 
consumed (30% more alcohol. (4) Variability within the group is rela¬ 
tively high in the case of the t wo strains of rats and in the C 3 H mice and 
the‘white mice, whereas the variability is relatively low among the 
dba mice. 

. These observations find a ready explanation on the basis of genetic 
differences and no other explanation has suggested itself. In the case of 
the dba mice which arc being used extensively in our laboratories in 
cancer investigations, brother-sister mat ings have been strictly adhered 
to and this appears to account for the low variability in this group. 
The C 3 H mice ordinarily are handled in the same manner but breeding 
difficulties have been encountered and brother-sister matings had to be 
abandoned until the difficulties could be eliminated. The two strains 
of rats and the white mice are of the sort commonly used for nutritional 
investigations and no special precautions have been taken in connection 
with their breeding. 

It should be noted that the extreme variability exhibited by the 
CjH mice is due in part to the fact that, for these animals, diet A is 



HIOCHKMICAL INDIVIDUALITY. Ill 


281 


very deficient. The experiment involving these mice could not have 
been extended much longer because of inadequacy of the diet. Strain 
differences in nutritional requirements of this sort have doubtless a 
genetic origin. 

The existence of wide variations in individual responses (Tables I and 
II), which, however, are minimized in the dba mice, can hardly be 
explained except on other than a genetic basis, and the high strain 
variability with respect to treatment to be discussed later, offers 
further strong evidence as to the importance of genetic factors in 
determining the responses to alcohol. 

Control of Alcohol Appetite in Animate 

Perhaps the most striking evidence of genetic variability was ob¬ 
tained in connection with the treatments which were instituted to 
overcome the alcohol appetite of the rats. Animals of “Strain 0” and 
of “strain H,” for example, responded in a significantly different 
manner when they were administered 10 B vitamin by mouth and the 
anti-pernicious anemia vitamin by injection. In every one (100%) of the 
24 rats of “strain O,” the alcohol consumption was decreased to a low 
level by this treatment. In 70% of these rats no difficulty was encount¬ 
ered in keeping the alcohol consumption at a low level. In the other 
30%, there was a tendency to revert to a higher level as in the work of 
Brady and Westerfcld, which tendency, however, in our experiments 
could usually be overcome by repeated vitamin administration. 

Among the 45 rats of “strain H” treated in the same manner, only 3, 
or about 7%, showed a diminution of alcohol consumption to a low 
level. In view of the relatively large number of animals used and the 
consistency of the results, there seems no question as to the significance 
of the observed differences between the two strains. 

A large number of rats have been treated in the course of our investi¬ 
gation. Typical results can best be presented in the form of a series of 
individual consumption curves on which are indicated in code the 
treatments that were instituted. These results show that, for individual 
rats on the indicated diets, different nutritional factors enter into the 
creation of an appetite for alcohol. For example, one rat of “strain O” 
(Fig. 5) was dramatically and permanently cured of its desire for 
alcohol by the oral administration of 10 known B vitamins alone. At 
the other extreme is Rat No. 1, Strain H (Fig. 13), the alcohol consiimp- 



282 R. J. WILLIAMS, L. J. BERRY AND E. BEERSTECHER, JR. 

tion of which remained at a high level, even after 3 other vitamins had 
been administered, viz., vitamins A, E, and the anti-pernicious anemia 
vitamin. When some linseed oil was added to the diet, however, the 
alcohol consumption promptly fell to nearly zero. Various intermediate 
variations are depicted in Figs. 6 to 12. Fig. 6 represents the consump¬ 
tion curve of a rat which was very slow in developing an appetite which, 
however, was dramatically abolished by the administration of 10 known 



Figs. 5-8. Effect of diet 
upon individual alcohol con¬ 
sumption levels. 


days 



BIOCHEMICAL INDIVIDUALITY. Ill 


283 



DAYS 


Figs. 0-12. Effect of diet 
upon individual alcohol con¬ 
sumption levels. 


B vitamins plus the anti-pernicious anemia vitamin. Fig. 7 has to do 
with a rat which drank steadily over a long period at a low level. Its 
appetite too was abolished by a similar vitamin treatment. Another 
rat (Fig. 8) remained at a high level of intake for nearly 7 weeks, when 
the administration of 11 vitamins promptly brought the consumption 
down to nearly zero. Rat No. 70 (Fig. 9) is one for which the anti- 
pernicious anemia vitamin appears to be unusually important. This is 



284 R. J. WILLIAMS, L. J. BERRY AND E. BEERSTECHER, JR. 

somewhat true of Rat No. 69 (Fig. 10) which showed an initial re¬ 
sponse to the anti-pernicious anemia vitamin alone, showed an increase 
when the B vitamins were first administered, but subsequently a drop 
to approximately zero when a second treatment with the anti-perni¬ 
cious anemia vitamin was given. Fig. 11 depicts the case of a rat of 
“strain H” which had its appetite greatly decreased by the adminis¬ 
tration of 11 vitamins. This rat had access to alcohol for 55 days 
previous to the treatment, but its consumption record for the earlier 
period was not taken. 



Fig. 13. Effect of diet upon 
individual alcohol consump¬ 
tion levels. 


In Fig. 12 and 13 it will be noted that additional vitamin A adminis¬ 
tration appeared to be effective in bringing about a material diminution 
of alcohol consumption in “strain H” rats. A diminution was observed 
in 2 days in every one, 100%, of the 31 rats furnished additional vita¬ 
min A along with the other 12 vitamin supplements already furnished. 
In another series of experiments the addition of linseed oil to the diet 
caused a marked diminution of the alcohol consumption in 4 out of 6 
rats. Two such cases are depicted in Figs. 12 and 13. 

Our experimentation in the control of alcohol appetite in mice has 
not been extensive. The experiments which have been performed indi¬ 
cate that the problem with mice is not essentially different than with 
rats. Only 6 mice have been treated, and these for only 4 days at the 
date of this writing; of these, 5 diminished their intake, one remained 
unchanged. The average diminution for the 6 was 54%. One had dimin¬ 
ished its consumption to zero. 



BIOCHEMICAL INDIVIDUALITY. Ill 


285 


It is clear from our experience that extended studies must be made 
before it will be possible to abolish with complete success the alcohol 
appetite of very individual laboratory animal regardless of its peculiar 
inheritance. We have not yet investigated in the general field of amino 
acids or minerals; it may be that, in some animals, a partial genetic 
block causes a high requirement for one of these and that such defici¬ 
ency induces in some unknown manner an abnormal appetite for 
alcohol. Our experience convinces us, however, that it will be possible 
by an extension of the means we have already employed to abolish the 
appetite for alcohol in 100% of laboratory animals. 

Habituation Experiments 

It has Ixien tacitly assumed, in connection with a number of studies (13), that 
animals can ho habituated to alcohol simply by forcing them lo drink it over a period 
of time. According to the genetotrophic concept this should not be the case for all the 
animals if the diet of the animals is a reasonably satisfactory one. 

While our experiments on this subject have not been carried very far, certainly they 
indicate strongly that habituation, particularly in the sense of addiction, is not 
brought about bv the simple procedure that has had wide use. 


TABLE HI 

Effect of Forced Alcohol Consumption (“Habituation”) 
on Voluntary Intake of Alcohol 


Ciroup 

No. of IlltN 

Diet 

l'(*. Alcohol/ 

100 g. rat/ 
day 

Range 

Mean of 
deviation 
from mean 

Percentage of 
deviation 
from mean 

“Habituated” 

9 

B 

.09 

.06-.13 

±.01 

11 

“Habituated” 

8 

Stock 

.06 

.05-.09 

±.01 

17 

Control 

18 

B 

.15 

.07-. 25 

±.05 

33 

('ontrol 

8 

Stock 

.32 

.07-.78 

±.18 

! 51 


A group of 9 weanling rats was placed for 29 days on diet B and furnished as the 
only available fluid 10% alcohol. A similar group of rats was placed on the same 
regimen, except t hat a stock diet was used instead of diet B. Two control groups (one 
on each diet) were placed upon the same regimen except that they were given a choice 
from the start between 10% alcohol and water. 

At the end of the 29-day period the 4 groups all had a choice between water and 
10% alcohol as in the earlier experiments described. The results presented in Table III 
show that animals which are forced to take 10% alcohol for a period of time are by no 
means led to take large amounts by choice. There is the strong suggestion that, when 
rats on essentially adequate diets are forced to drink alcohol, they actually develop a 
dislike for it and drink less than if they had been given a choice from the beginning. 
In any event, this procedure, which has often been used in animal studies on alco¬ 
holism, is clearly not promising as a means of producing compulsive drinkers. 



286 R. J. WILLIAMS, L. J. BERRY AND E. BEERSTECHER, JR. 


Theory of Alcoholism 

The theory suggested and supported by these experimental findings 
may be outlined as follows: Appetite for alcohol is a physiological 
perversion based upon incompletely satisfied nutritional needs. We 
know relatively little as to the mechanism whereby physiological 
changes in the body induce special appetites, but that special appetites 
are induced by the internal environment is undeniable. Alcoholic 
craving, according to our theory, develops as an overpowering drive in 
certain individuals as a result of their unusually high requirement for 
one or more specific nutritional entities, such as the B vitamins, and 
because this high need is not fully satisfied by the common foods which 
are consumed. The use of large quantities of refined foods would, ac¬ 
cording to this theory, contribute to alcoholism; and the consumption 
of alcohol itself, in addition to developing a taste, would promote the 
appetite and start a vicious cycle because it crowds out of the diet 
wholesome foods which normally contribute to the various needs. 

There may or may not exist a few specific lacks (known and un¬ 
known) which are generally associated with the compulsive drinkers 
appetite. Our experiments with rats indicate that widely different 
deficiencies may contribute directly or indirectly in producing an 
appetite for alcohol. Among the controlling factors are various of the 
known B vitamins, the anti-pernicious anemia vitamin (vitamin B 12 ), 
vitamin A, and unsaturated fat acids (linseed oil). Our knowledge of 
precisely how the anti-pernicious anemia vitamin is obtained and 
assimilated from natural diets is limited. It was administered to the 
animals by intraperitoneal injection in order to insure its absorption. 
The failure to assimilate this or some other vitamin might lie at the 
basis of certain cases of alcoholism. 

Our experimental studies strongly suggest that the abundant satis¬ 
faction of every nutritional need will abolish alcoholic appetite. It 
should be emphasized that in providing for every nutritional need the 
chain is as strong as its weakest link, and satisfying every need except 
one should be expected to be completely ineffective. Furthermore, due 
partly to differences in assimilation, bacterial action in the intestines, 
etc., as well as distinctive features in the pattern of intermediate metab¬ 
olism, the requirement for a given vitamin or other nutritional entity 
by a particular individual may actually be far above what is considered 
normal. This will make it more difficult than might be supposed to 
insure to him an adequate supply of every potential nutritional factor. 



BIOCHEMICAL INDIVIDUALITY. Ill 


287 


It is true that the treatment of alcoholism has often included the 
administration of supposedly generous amounts of B vitamins. A 
sufficient cause for the failure of vitamin therapy as a treatment for 
alcoholism in the past is the omission of the anti-perniciuos anemia 
vitamin as well as sufficient amounts of certain other vitamins which in 
individual cases may be necessary links in the chain. Commercial 
vitamin preparations are often inadequate and poorly balanced (14), 
and the idea back of whatever therapy that has been used has been to 
remedy deficiencies arising because of excessive alcoholic consumption. 
No one, so far as we known, has ever seriously considered a nutritional 
deficiency to be a cause of alcoholism, and no attempts have been made, 
therefore, to correct an underlying deficiency. Requirements in such 
cases may be very high or hard to satisfy, and haphazard administra¬ 
tion of “ vitamin pills” and self-medication by uninformed laymen can 
be expected to be completely ineffective, or even harmful or dangerous. 

The effects of very high alcoholic consumption have not as yet been 
studied in our investigation. On a body-weight basis, the maximum 
consumption observed under our conditions is equivalent to 3.7 qt. of 
90 proof liquor per day for an average man. In proportion to the total 
food consumption of a rat, however, the amounts are much smaller. It 
may be that continued nutritional deprivation associated with high 
alcoholic consumption does irreparable damage to the individual and 
that, for this reason, advanced stages of alcoholism will be incurable by 
satisfying all the nutritional needs. We may hope that is not so, and 
the presumptive evidence is against this idea. 

The theory which we have set forth, if valid, emphasizes the impor¬ 
tance of good nutrition and the desirability of fulfilling the nutritional 
needs of everyone—not only those who have average requirements. 
In calculating nutritional needs, a far larger factor of safety than is 
generally used may well be in order. It seems, on the basis of our 
experiments, very likely irideed that consistent good nutrition from 
childhood on will strongly militate against alcoholism. Regardless of 
the efficacy of any treatment which has its basis in the theory pro¬ 
pounded, the prophylactic effects of good nutrition are strongly indi¬ 
cated. Good nutrition may, however, mean in some cases nutrition 
designed for the needs of a particular individual with specific require¬ 
ments that are unusually high. 

An interesting manner in which the proposed theory fits the facts has 
to do with the relationships between age and alcoholism. If an indivi- 



288 K. J. WILLIAMS, L. J. BERRY AND E. BEERSTECHER, JR. 


dual becomes an alcoholic at an early age, before 28 (15), up to now 
there has been practically no hope for him, In general, those who 
become alcoholic later in life are much more susceptible to treatment. 
According to our theory, those individuals who become alcoholics 
early are those who have the highest requirement for some specific 
nutritional factor; the resulting deficiency obviously shows up earlier 
and the possibility of its being met or even half-way met is remote. 
Those who become alcoholic in later years do not have such marked 
deficiencies and the urge can perhaps be broken by psychological means. 

Study of alcoholism from the point of view which we have developed 
is in its infancy, and much further research will be required to clarify 
the problem and to test various ramifications of the theory. One line of 
research must center on the problem of precisely what nutritional 
deficiencies may be responsible for inducing an appetite for alcohol. 

Another research field of large proportions has to do with developing 
means of assessing individual metabolic patterns, so that something 
other than a “shotgun” therapy can be used. We believe that it will 
eventually become possible through the efforts of many investigators 
in many laboratories to determine, in the case of any specific alcoholic, 
the biochemical roots of his particular difficulty, and, on this basis, to 
formulate an adequate treatment. For adequacy in this regard, it is 
essential that every nitritional need be met simultaneously. 

It should be noted also that, even if one has relieved the appetite of 
an.alcoholic, no lasting benefit would result were he allowed to go his 
way, become deficient again and relapse into the condition of alcohol 
craving again. How much danger will exist in this direction, only 
experience can show. It is obvious, however, that a one-time alcoholic 
should continue to have his nutritional needs satisfied continuously 
throughout life. 

Finally, another vast field of research which is opened up by the 
theory we have proposed is the possible relation of genetotrophic 
factors to other diseases of obscure etiology, which have previously 
been mentioned. 


Acknowledgments 

We gratefully acknowledge our indebtedness to our colleagues and assistants, to 
Helen Kirby, who has carried forward the investigation of individual metabolic traits, 
and to Janet Keed, Nellie KefTer, William Brown and Gene Rich, whose technical 
assistance has been invaluable. 



BIOCHEMICAL INDIVIDUALITY. Ill 


289 


Summary 

Evidence is presented, based upon animal experimentation, that 
alcoholism is a genetotrophic disease, viz., a disease in which genetic 
factors and nutritional deficiency together are etiological agents. It 
has been found possible, by the administration of nutritional factors, 
to control the appetite for alcohol in laboratory animals. It is suggested 
that genetotrophic factors may be important in may diseases of obscure 
etiology. 

Code for Alcohol Consumption Curves 
IV 

Bir -5 units of 20 unit anti-pernicious anemia liver preparation given intraperitoneally 
each day for f> consecutive days. 

4 V 

ing./day for 5 davs 


Thiamine.0.3 

Riboflavin.0.5 

Calcium panto! henato.3.0 

Folic acid (Folvit-c) .... 0.1 


5V 


Same as IV + 4 V, except that the B 12 

was partially purified. 


!0V 

mg. /day for 5 da> s 

Same as 4V with the addition of: 



Pyridoxine. 


.0.15 

Choline . 


.100.0 

I nositol. 


. 3.0 

p-Aminobenzoic acid.... 


. 3.0 

Nicotinic acid. . . 


. 1.0 

Biotin. . 

11V 

. 0.01 

Same as IV + 10V. 

13V 

units/day for 5 days 

Same as 11V with the addition of: 



Vitamin A. 


.500 

Vitamin 1). 

L 

.100 

Same as 13V with the addition of: 


Linseed oil. 


. 5% by weight to total diet. 

Note: In the alcohol consumption curves the experimentally determined values are 
plotted. These are uncorrected for spillage and evaporation. Hence, a low value may 


actually represent zero consumption. 


























290 R. J. WILLIAMS, L. J. BERRY AND E. BEBRSTECHER. JR. 


References 

1. Beadle, G. W., Chem. Revs. 37, 1 (1945). 

2. Williams, R. J., Chem. Eng. News 25, 1112 (1947). 

3. Williams, R. J., The Human Frontier. Harcourt, Brace and Co., New York, 1949. 

4. Mitchell, II. K., and Houlahan, M. B., Am. J. Botany 33, 31 (1946). 

5. Mitchell, II. K., and Houlahan, M. B., Federation Proc. 6, 506 (1947). 

6. Schneider, H. A., Vitamins and Hormones 4, 35 (1946). 

7. Editorial, J. Am. Med. Assoc. 135, 576 (1947). 

8. Williams, R. J., Quart. J. Studies Ale. 7, 567 (1947). 

9. Mardones, J., Segovia, N., and Onfray, E., Arch. Biochem. 9, 401 (1946). 

10. Brady, R. A., and Westerfeld, W. W., Quart. J. Studies Ale. 7, 499 (1947). 

11. Engel, R. W., Proc. Soc. Exptl. Biol. Med. 52, 281 (1943). 

12. Light, R. F., and Oracas, L. J., Science 87, 90 (1938). 

13. Newman, H. W., Acute Alcoholic Intoxication, pp. 88-95. Stanford University 

Press, 1941. 

14. Williams, R. J., J. Am. Med. Assoc . 119, 1 (1942). 

15. Fleming, R., Lecture 25, Quart. J. Studies Ale. (1945). 



The Reaction of Tobacco Mosaic Virus with 
Formaldehyde. I. Electrophoretic Studies 1 

Marie A. Fischer 2 and Max A. Lauffer 

From the Departments of Chemistry, Physiological Chemistry and Physics, 3 
University of Pittsburgh, Pittsburgh, Pennsylvania 
Received May 6, 194!) 

Introduction 

In 1938, Ross and Stanley (1) reported that reaction between 2% 
tobacco mosaic virus (TMV) and 2% formaldehyde, buffered with 
il f /10 phosphate to pH 7.0, causes the infectiousness of the virus to be 
destroyed and causes a decrease in the amount of color developed with 
ninhydrin and with Folin’s phenol reagent at .pH 7.7. Furthermore, 
these workers found that, if virus which had been inactivated was 
subsequently dialyzed for 3 days against a dilute pH 3.0 buffer, the 
lesion-count on N. glutinom was increased. The increased lesion-count 
was accompanied by a reversal of the chemical changes, for the amount 
of color developed by both ninhydrin and the phenol reagent was found 
to be increased by the dialysis. Kassanis and Kleczkowski (2) failed to 
observe the reversibility of the reaction and were unable to correlate 
the chemical changes effected by formaldehyde treatment with the loss 
of infectiousness. More recently, Fraenkel-Conrat et al. (3) reported 
that formaldehyde treatment at pH 7.0 decreases the availability of the 
tryptophan and tyrosine residues to the Folin reagent, but that for¬ 
maldehyde does not actually combine with these groups. Therefore, it 
seemed worthwhile to investigate the electrophoretic changes resulting 
from the treatment of TMV with formaldehyde, and to repeat those 
experiments of lloss and Stanley which indicate that the infectivity of 

1 Aided in part by a grant from the National Foundation for Infantile Paralysis, Inc. 

* Some sections were abstracted from a thesis submitted by Marie A. Fischer to the 
Department of Chemistry in partial fulfillment of the requirements for the Ph.D. 
degree. 

* Contribution no. 721 of the Department of Chemistry and 5-p-49 of the Depart¬ 
ment of Physics, University of Pittsburgh. 

291 



292 


MARIE A. FISCHER AND MAX A. LAUFFER 


formaldehyde-inactivated virus can be increased by dilute pH 3.0 
dialysis. 

Materials and Methods 

The juice expressed from infected Turkish tobacco plants was passed through a 
Sharpies Supercentrifuge. Ninety per cent of the infectious material was collected on 
the cylinder and, after suspension in M /10 phosphate, pH 7.0, the material was sub¬ 
jected to a low and high speed centrifugation cycle (4). The cycle was repeated 4 times. 
The final supernatant fluid was colorless and, on t he basis of activity measurements, a 
100-fold concentration of virus was effected. The protein content of the purified 
solution was determined by Kjeldahl analyses for nitrogen, and infectivity was meas¬ 
ured by the half-leaf method of inoculation (5), using N. glutinosa as the test plant. 
So that approximately equal numbers of local-lesion would be caused by treated and 
control samples, several dilutions of each sample were usually made and compared by 
the half-leaf method and a Latin-square pattern for inoculation. 

Formaldehyde-treated TMV was prepared by allowing solutions containing 2% 
virus and 2% formaldehyde to react at room temperature (1). In the pH 7.0 experi¬ 
ments the reaction was carried out in il//10 phosphate, and in the pH 4.0 experiments 
in 0.077 M Na 2 P0 4 -0.063 M citric acid. To stop the reaction, the formaldehyde was 
removed by dialysis for fi hours against cold, distilled water in a rocking dialysis 
apparatus. 

For the prolonged pH 3.0 dialyses, the procedure of Ross and Stanley (l) was 
followed as closely as possible. The formolized virus was dialyzed against flowing 
0.001 M phosphate-citrate-HCl buffer at pH 3.0 in a rocking dialysis apparatus. The 
temperature was kept between 5 and 10°C. The samples were subsequently dialyzed 
to pH 7.0 with J//10 phosphate, analyzed for nitrogen, and inoculated at suitable 
dilutions against a control sample of virus. In all eases the inactive virus and the, 
aliquot dialyzed at pH 3 were inoculated at the same protein concentrations. 

The electrophoretic analyses were carried out in the apparatus described by Tiselius 
(6) as irtodified by Longsworth (7). All of the preparations were diluted to a protein 
concentration of 0.5% and were dialyzed for 3 days against two 1 1. and a third 2 1. 
portions of pH 7,00, ionic strength 0.&, KH2PO4-K2HPO4 buffer. The samples migrated 
in an electric field of 4.5 volts/cm., with continouus compensation, usually for 24 hr. 
Mobilities were determined for several preparations by measuring migration distances 
after 2 hr. of electrophoresis without comi>ensation under otherwise similar experi¬ 
mental conditions. In each case the current was reversed, the boundaries were re¬ 
turned to their initial positions, and the mobilities were calculated from the averages 
of the migration distances of both boundaries. 

Experimental Results and Discussion 

The Reaction at pH 7.0 

It has been shown (1,8) that formaldehyde-inactivation of tobacco 
mosaic virus, in a neutral buffer, and at room temperature, follows the 
course of a first-order reaction. Table I shows that, under the same 



TOBACCO MOSAIC VIRUS AND FORMALDEHYDE 2W 


TABLE I 

The Effect of Formaldehyde Treatment on Tobacco Mosaic Virus n 


Time of HCHO troatemnt" 

\ptivity remaining 

Anodic mobility 

{ hr .) 

per cent 

cm * / volt-sec X/0' 1 

0 

100 

7.20 

6 

12.1 

7.31 

fi 

10.8 

7.37 

12 

1.63 

7.43 

24 

0.25 

7.48 

24 

0.078 

7.58 


a Each analysis is fora separate experiment carried out at room temperature for the 
time shown. 


conditions of treatment, an increase in the anodic mobility of tin* vims 
occurs and the increase in mobility is dependent upon the lime of con¬ 
tact between formaldehyde and the virus. Since a first-order loss of 
activity is frequently interpreted as implying an “all-or-none” mech¬ 
anism, it seemed desirable to determine whether the electrophoretic 
mobility of all the virus particles had been changed, or whether some 
had undergone a change in mobility and others had not. To study this 
problem, samples which had been inactivated to varying degrees were 
subjected to electrophoresis at pll 7.0 for approximately 24 hr. Fig. 1 
shows typical electrophoretic patterns for (A) untreated, (B) 72% 
inactivated, and (C) 99% inactivated virus after prolonged electro¬ 
phoresis. All of the samples inactivated at pH 7.0 showed essentially the* 


llii ilk 

ZOhri. Shn 5hf». 20hr». 24hr* 3hr* Shn 2«hri 

*»«• D*»e Ate Otic. 

(A) <B> 



Fig. 1 . Electrophoretic patterns of 
(A) untreated, (B) 72% inactivated, (C) 
09% inactivated, and (D) a mixture of 
t wo parts 99% inactive and one part un¬ 
treated virus. The virus was inactivated 
at pH 7.0 and the electrophoresis carried 
out at pH 7.0. 


(0) 



294 


MARIE A. FISCHER AND MAX A. LAUFFER 


same degree of homogeneity as did normal virus. The electrophoretic 
homogeneity of the preparations was verified by studying known mix¬ 
tures of active and inactivated virus. In one representative experiment 
shown in Fig. 1 (D), a mixture of two parts inactive and one part nor¬ 
mal virus was easily separated into its two components. The larger 
fraction had the faster rate of migration, an observation consistent with 
the known higher mobility of formaldehyde-treated virus. It is, there¬ 
fore, apparent that an approximately equal number of reactive groups 
on each virus particle reacts with formaldehyde at pH 7.0, and thus the 
reaction products retain essentially the original electrophoretic homo¬ 
geneity, even though they arc inhomogeneous with respect to infecti- 
vity. As is reported elsewhere (8), a theory devised to account for the 
kinetics of the reactions between formaldehyde and TMV leads to 
deductions consistent with this observation. 


The Reaction of pH J+.O 

Formaldehyde inactivation at pH 4.0 seems to be a first order re¬ 
action with a reaction velocity constant of 0.45 reciprocal hours at 
room temperature. In this respect, the reaction at pH 4.0 is similar to 
that at pH 7.0 (1,8). However, the chemical changes appear to be differ¬ 
ent. At pH 4.0, insoluble material appears during the course of the 
reaction and, after about 12 hr., gel formation is evident. These changes 
were not observed in a control sample of virus held at pH 4 or in virus 
inactivated at pH 7.0. It has already been shown that formaldehyde 



Ate. D«sc. 


(A) 



Ate. Dttc. 



Ate. Oetc* 


(B) 



Fig. 2. Electrophoretic patterns of 
TMV treated with formaldehyde at 
pH 4.0. Electrophoresis was carried 
out at pH 7.0 for 24 hr. (A) repre¬ 
sents control TMV, (B) after 2 hr. 
JICHO treatment, (C) after 6 hr. 
IICHO treatment, and (D) after 24 
hr. HCHO treatment. 


<C) 


( 0 ) 



TOBACCO MOSAIC VIKUS AND FOItMACDKHYDK 


29 .-) 


inactivation at pH 7.0 results in a single component by the criterion of 
electrophoresis at pH 7.0. Therefore, for comparison, virus was treated 
with formaldehyde at pH 4.0 and subsequently electrophoretically 
analyzed at pH 7.0. Before electrophoresis, each sample was centri¬ 
fuged at 3500 r.p.m. for 15 min. to remove any large particles of in¬ 
soluble material. Fig. 2 shows the electrophoresis patterns obtained for 
several samples of virus treated in this manner. The control sample, 
which was subjected to the same pH change but was not treated with 
formaldehyde, appeared to be essentially homogeneous. The samples 


TABLE II 

The Effect of Prolonged pH 3.0 Dialysis on the Activity 
of Formaldehyde-Inactivated TMV 


(1) 

Cone, of sample 

(2) . 

Activity remaining 
before pH 3 dialysis 

(3) 

Activity remaining 
after pH 3 dialysis 

(4) 

Ratio of (3)/(2) 

0 -Ice. 

per'cent 

per cent 


10~ 4 

0.0 

17.0 

1.0 

10- ' 

! 7.6 

7.0 

1.0 

10-* 

6.5 

0.0 

1.4 

10 1 

17.0 

16.0 

0.0 

10~ 4 

13 

13 

1.0 

10* 4 

13 

16 

1.2 # 

5X10- 3 

0.078 

0.25 

3.3 

10 3 

0.03 

1.6 

1.7 

10-3 

1.23 

1.26 

1.0 

10-3 

1.05 

1.10 

1.1 

10-3 I 

! 0.006 

0.017 

3.1 

10“ 2 

0.010 

0.084 

4.4 

10-= 

0.060 i 

1 1 

0.120 

1.0 


treated with formaldehyde at pH 4.0 at room temperature gave in¬ 
homogeneous patterns which were anomolous because the inhomoge¬ 
neities were not symmetrical with respect to the ascending and des¬ 
cending boundaries. 

The Effect of Prolonged pH S.O Dialysis on Formaldehyde 
Inactivated TMV 

The data of Table II support the observation of Ross and Stanley (1) 
that prolonged pH 3.0 dialysis of formaldehyde-treated virus causes an 
increase in the activity of the virus preparation. It is unlikely that 
errors in the testing method can explain the difference in lesion-count 



296 


MARIE A. FISCHER AND MAX A. LAUFFER 


between inactivated virus and the aliquot which was subsequently 
dialyzed at pH 3.0. Errors in the method should give rise to random 
variation in the relative activities of the samples. In that case, the 
mean of the ratios of the relative activities of the sample before and 
after pH 3 dialysis should be one. Actually, the mean for all of the 
experiments was 1.84, and the standard deviation of the mean was 
zb 0.31. Since the deviation from one is 2.7 times the standard deviation 
of the mean, it is improbable that the increased lesion-count following 
the pH 3 dialysis is the result of random error. 

It was not possible to detect electrophoretic mobility differences 
between inactive virus and the same sample subsequently dialyzed for 
3 days at pH 3.0. Furthermore, mixtures of the two did not seqaratc 
into fractions during 24 hr. of electrophoresis at pH 7.0. If the pro¬ 
longed pH 3.0 dialj’sis affects groups responsible for the net charge on 
the particle, the mobility change is so small that it is undetectible 
under the conditions of the experiment. 

Summary 

1. Reaction between tobacco mosaic virus and formaldehyde at 
pH 7.0 resulted in an electrophoretically homogeneous preparation, 
regardless of the time of formaldehyde treatment. 

2. When formaldehyde treatment was carried out at pH 4.0, the 
preparations appeared inhomogeneous when the electrophoresis was 
carried out at pH 7.0. 

3. Dialysis of inactivated virus at pH 3.0 for 3 days caused an in¬ 
crease in the lesion-count when the infectiousness was tested by inocu¬ 
lation on N, glatinosa. 

4. Under the conditions of these experiments, it was not possible to 
detect a change in the electrophoretic mobility of inactive virus after it 
had been dialyzed at pH 3.0 for 3 days. 

References 

1. Ross, A. F., anj> Stanley, \\. M., J. Gen . Physiol. 22, 165 (1938). 

2. Kassams, B., and Kleczkowski, A., Biochem. J. 38, 20 (1944). 

3. Fraenkel-Conkat, H., Brandon, B. A., and Olcott, H. S., J. Biol. Chem. 168, 

99 (1945). 

4. Stanley, YV. M., J. Am. Chem. Soc. 64, 1801 (1942). 

5. Loring, H. S., J. Biol. Chem. 121, 637 (1937). 

6. Tiseiaus, A., Tram. Faraday Soc.. 33, 524 (1937). 

7. Longsworth, L. G., J. Am. Chem. Soc. 61, 529 (1939). 

8. Fischer, M. A., and Tmitffer, M. A., J. Am. Chem. Soc . In press. 



An Interpretation of the Contradictory Results in 
Measurements of the Photosynthetic Quantum 
Yields and Related Phenomena 

James Franck 

F row the I )e part went ofChemistry (FeU Fund), University of Chicago, Chicago , Illinois 

Received April 22, 1940 

1. Contra dictions in the Values Measured for the 
Photosynthetic Quantum Yield 

Warburg and Negelein (1) were the first to measure the quantum 
yield of photosynthesis. They found that 4 light quanta are sufficient 
to reduce one CO 2 molecule. Thermodynamic considerations made it 
difficult to understand this high yield, but their measurements remained 
uncontested for many years. Doubts arose concerning them when new 
observations, made in Farrington Daniels’ laboratory in Madison (2), 
consistently gave much lower quantum yields with an upper limit of 
about 1/10 instead of 1/4. By copying the original observation method 
exactly, Rieke (3) was able to get the same quantum yield (1/4) as 
Warburg and Negelein. However, when the algae ( Chlorella) which 
served as a plant material for all of these measurements were kept in 
an alkaline buffer solution, the yield had an upper limit of ~ 1/8. An 
explanation of these contradictory results was presented by Emerson 
and Lewis (4), who found that, in their own careful measurements, 
made under conditions supposed to be identical with Warburg and 
Negelein’s ,an outburst of CO 2 at the beginning of illumination and a 
corresponding uptake of that gas during the following dark period 
changed the photosynthetic quotient to such an extent that apparent 
quantum yields even larger than 1/4 could be calculated. However, the 
rate of photosynthetic oxygen production gave a yield of ~ 1/10 
whether measured by use of a C0 2 buffer or by Warburg’s well known 
two-vessel method. This explanation of the differences was generally 
accepted in this country, especially as a number of observers using a 
diversity of methods all found ~ 1/10 as the upper limit of the 


297 



298 


JAMES FRANCK 


quantum yield (5). Recently, Warburg (6) has rejected Emerson’s 
criticism. His new measurements give the same quantum yields as his 
old ones and a photosynthetic quotient near unity. However, the quo¬ 
tient and the quantum yield were measured under different conditions, 
and it is thus still an open question to what extent the C0 2 outburst 
(Emerson effect) may have falsified Warburg’s measurements. Be¬ 
cause Warburg and Emerson used comparable experimental methods, 
emphasis will be placed on the contradictions between their results. 

The present writer, who was priviliged to have oral discussions with 
Warburg and with Emerson, finds it hard to accept the point of view 
that only Warburg’s method under special conditions will permit the 
algae to reduce C0 2 with a quantum yield of 1/4 when all other observa¬ 
tions systematically give ~ 1/10 as the highest value. On the other 
hand, it is not certain that the occurrence of the Emerson effect is the 
main cause of the difference between Enerson’s and Warburg’s new 
results, because this effect, while undoubtedly present in both observa¬ 
tions, seems to be smaller in Warburg’s new experiments than in 
Emerson’s. The main part of the C0 2 outburst during irradiation 
occurs during the first minutes of illumination, while the corresponding 
uptake of C0 2 during dark periods shows a slower decay. If we com¬ 
pare only those measurements of Warburg with Emerson’s in which 
the precaution is taken of not using these first minutes of pressure 
changes in each light and dark period for the calculations, we still find 
differences of about 100%; i.e., apparent quantum yields between 
1/4 and 1/6 according to Warburg and 1/10 according to Emerson 
and Lewis. 

These facts raise doubts as' to whether the Emerson effect suffices 
for the explanation of Warburg’s results. Experiments are now in 
progress in Warburg’s and Emerson’s groups in which the main im- 
phasis is laid upon the question of whether, under the special conditions 
of Warburg’s experiments, a gas exchange with a photosynthetic 
quotient of ~ 1 is possible giving the result of a quantum yield of 
~ 1/4. 1 However, not only this problem is at stake, because there are 
further differences, discussed later, between the results of Warburg’s 
measurements and those of Emerson and Lewis (and the other ob¬ 
servers who measured the quantum yield with different methods). 
They seem to indicate that Warburg’s results differ hot only quantita- 

1 Compare the note added in proof at the end of the paper. 



PHOTOSYNTHETIC QUANTUM YIELDS 


299 


tively but also qualitatively from those of the others. A similar situa¬ 
tion exists for another basic problem of photosynthesis—the identi¬ 
fication of the chemical nature of photosynthetic intermediates. Two 
groups engaged in studying these compounds with the help of the 
radioactive isotope, C 14 , have obtained entirely contradictory results. 

II. Proposal to Interpret the Differences as Caused 
ry Variations in the Permeability of Chloroplast 
Membranes to Respiration Intermediates 

The present writer is inclined to seek a way out of these dilemmas by 
the following assumption: Normally the process of respiration in the 
cells has another chemical pathway than, and is locally separated from, 
the process of photosynthesis in the chloroplasts. However, under 
special conditions, the chloroplast membranes become permeable to 
intermediates of respiration. In that case, the processes of dark and 
light metabolism become intertwined and interfere with each other. 
Correspondingly great changes of both processes occur during illumin¬ 
ation. They may be of brief duration if the photosynthesis rate is much 
higher than that of repiration, or permanent when the photosynthetis 
gas exchange is smaller, or not much higher, than that of respiration. 
The following chapters contain an analysis of the inconsistencies in the 
measurements of the quantum yield based on this idea. 

We introduce the assumption that Warburg’s high quantum yield 
may be connected with the reduction of respiratory intermediates 
rather than with the reduction of CO*. That is possible because War¬ 
burg’s measurements are carried out under conditions where the 
photosynthetic rates are smaller than or, at best, comparable to, the 
respiration rates. That such a process may occur and influence the 
measurements of the quantum yield of photosynthesis is by no means 
a new idea. It was proposed a long time ago by different authors but 
such statements did not receive much attention, nor were attempts 
made to give a detailed explanation of why the effect occurs only under 
special conditions, or how it is connected with other observations in the 
field of photosynthesis. It may, therefore, be useful to examine this 
possibility more closely. We will see that the assumption is not only 
able to reconcile Warburg’s quantum measurements with those of 
others, but also agrees with results of an entirely different nature. 



300 


JAMBS FRANCK 


III. Evidence Supporting the Hypothesis that the Gas 
Exchange Measured by Warburg is Due to Photochemioai, 
Reduction of Respiration Intermediates 

/. Quantum Yield and Photosynthetic Quotient of This Process 

The first questions which arise are the following: Will reduction of 
intermediate respiration products proceed with a quantum yield of 
~ 1/4? Will the gas exchange connected with this process have the 
same quotient as the photosynthesis of C0 2 ? To answer these questions, 
we do not need to enter into a detailed discussion concerning the exact 
chemical nature of the respiration products, because we know that in 
the overall process of respiratory oxidation of carbohydrates one 
oxygen molecule is consumed for each C0 2 molecule formed. We, 
therefore, may speak only of carbohydrates, of half-oxidized respiration 
products and of C0 2 . The term carbohydrates includes all substances 
of the general formula (CH 2 0) n . Half-oxidized products are substances 
containing a group which can be restored to the oxidation state of a 
carbohydrate by the addition of two hydrogen atoms, and C0 2 stands 
for free C0 2 or for substances in which a carboxylated group may 
be reduced photochemically. These latter groups need 4 hydrogen 
atoms for reduction to the carbohydrate state. Obviously, half as 
many quanta per molecule as are used for the reduction of C0 2 
should then suffice to reduce the half-oxidized products and one-half 
molecule of oxygen will be evolved for each of these molecules 
reduced to carbohydrate. The photochemical removal of half-oxidized 
molecules from the course of respiration means that oxidation by 
respiration proceeds only half way. Thus, every half-oxidized mole¬ 
cule reduced photochemically will prevent the consumption of one- 
half molecule of oxygen by respiration. The evolution of C0 2 will be 
diminished by a whole molecule. In other words, the process would add 
one oxygen molecule and take away one C0 2 molecule from the 
balance sheet of respiration for every 4 quanta absorbed by the 
photosynthetic apparatus. If, far below the compensation point, the 
photochemical gas exchange should consist solely in the reversal of 
respiration, a quantum yield of ~ 1/4 with a normal photosynthetic 
quotient would result. 2 

* Correspondingly an apparent quantum yield of J should be observed if respira¬ 
tion products reducible by one hydrogen atom are photochemically diverted from 
the course of respiration. 



PHOTOSYNTHETIC QUANTUM YIELDS 301 

With rising light intensity, the contribution of the reduction of 
respiratory intermediates to the overall photosynthetic gas exchange 
should become smaller, while the contribution made by photosynthesis 
of C0 2 will rise. Correspondingly, the quantum yield should gradually 
fall with rising light intensity to a limiting value equal to the quantum 
yield of CO 2 assimilation, a value half as great as for the reduction of 
the respiratory intermediates. The shape of the curve, quantum yield 
vx. light intensity, cannot be predicted without further hypothesis. Let 
us introduce the assumption that, in Warburg’s algae, the membranes 
‘separating the chloroplasts from the rest of the cell are so easily per¬ 
meable to these respiratory intermediates that at higher light intensi¬ 
ties they all diffuse into the photosynthctic apparatus instead of being 
further oxidized by respiration. The occurrence of a quantum yield of 
1/4 at very low light intensities may be taken as an indication that 
reduction of intermediates occurs preferentially to that of C0 2 . Thus, 
we may estimate that the light intensity necessary to divert all respira¬ 
tory intermediates from the course of respiration would probably lie 
about 3 times higher than the one necessary to reach the compensation 
point. The apparent quantum yield measured with light of that inten¬ 
sity would then be ~ 1 /7. While the extreme case of total diversion of 
intermediates from the course of respiration may not be entirely real¬ 
ized in Warburg’s measurements, it seems at least to give an approxi¬ 
mate description of the situation. 

Warburg uses a great surplus of algae in order to guarantee that at all 
times, in spite of the shaking of the vessel, all of the incident light is 
absorbed. As a result, only a small fraction of the algae is exposed at 
each moment to the full intensity of the incident light; in fact, the bulk 
of them receive no light. The stirring quickly exchanges the layer of 
exposed algae with others emerging out of the darkness. In effect, this 
situation constitutes a kind of flash illumination of the algae in which 
the duration of the flash is several times shorter than that of the dark 
period. Even though, during the brief illumination period, the respira¬ 
tory intermediates may be more quickly removed by photochemical 
reduction than they can be replaced by diffusion into the chloroplasts, 
this can be compensated for by diffusion occurring during the dark 
periods. In Warburg’s experiments we may, therefore, conclude that a 
relatively high percentage of all respiratory intermediates will be 
caught by the light and, consequently, the contribution of this reduc¬ 
tion process should be a considerable percentage of the total photo- 



302 


JAMES FRANCK 


synthetic gas exchange, even in the region of light intensities surpassing 
somewhat the ones necessary to reach the compensation point. 

The following table gives quantum yields measured by Warburg as a 
function of light intensity. They are calculated from the steady rates of 
gas exchange in the light and in the dark, discarding the first 5 min. of 


each period ((>). 

Intensities of incident light 

Micromole quantu'inin. Quantum fields 

0.158 (below compensat ion) 1/3.96 

0.33 (below compensation) 1 /4.5 

0.75 (below compensation) 1 /5.03 

1.42 (above compensation) 1 /5.56 


The highest intensity used causes a gas exchange only 1.1 times that of 
respiration. The yields show an unmistakable tendency to become 
smaller with growing light intensity. Warburg explains the lowering of 
the quantum yield with increasing light intensity by the assumption 
that the rate of photosynthesis in the part of the algae exposed to the 
full intensity of the incident light may start to approach light satura¬ 
tion even before the total respiration of the bulk of the algae is compen¬ 
sated. Rabinowitch has pointed out that such an assumption implies 
light saturation values much lower than usual. Moreover, it is in con¬ 
tradiction to Emerson and Lewis’ results who, using Chloretta in equal 
concentration, observed a linear relation between light intensity and 
photosynthetic rate up to values even higher than Warburg’s maximum 
intensity. These measurements were made in an alkaline buffer solu¬ 
tion which Warburg regards as a sufficient reason for the low quantum 
yield of ~ 1/10. However, these authors found no change in the quan¬ 
tum yield of the oxygen evolution if the alkaline buffer was replaced by 
an acid medium. Furthermore, if Warburg’s arguments that alkalinity 
is always harmful to photosyntheiss is correct, we would expect that it 
would lower not only quantum yields in the region of intensities where 
light is the limiting factor, but also the saturation intensities and satur¬ 
ation values. Therefore, the rate curve should start to bend over to 
saturation at lower intensities in the case of Emerson’s observations 
than in Warburg’s. This also is contrary to the experimental evidence. 

2. Changes of Permeability of Chloroplast Membranes by Variation 
of External and Internal Conditions 

An explanation without contradictions becomes possible if we 
accept the hypothesis that the high quantum yields are connecter! 



PHOTOSYNTHETIC QUANTUM YIELDS 


303 


with the reduction of respiratory intermediates, and that the dif¬ 
ference between the results of Warburg and Emerson lies in a differ¬ 
ence in the permeability of the chloroplast membranes to the res¬ 
piratory intermediates. On this basis, the lowering of the quantum 
yield with increasing light intensity measured by Warburg indicates 
that the percentage contribution of the reduction of respiratory prod¬ 
ucts to the overall gas exchange starts to fall far below the compensa¬ 
tion point. The quantum yield of 1/5.56, measured at the highest in¬ 
tensity, indicates that at that point it is still reponsible for ~ 60% of 
Ihe photosynthetic gas exchange. In Warburg’s algae the chloroplast 
membranes must have been very permeable; in Emerson’s the mem¬ 
branes seem to be impermeable. It is a well known fact that great 
differences in permeability of membranes can be caused by somewhat 
different treatment of the algae (7). For instance, permeability to 
alkali rises with the age of the algae and the cell density, and is made 
especially great by prolonged anaerobicity (12). If the membranes 
become permeable, they do so for a variety of substances. Thus, an 
increase in the membrane permeability of the chloroplasts to alkali 
may occur under the same conditions as an increase in the premeability 
to respiratory products. 

Support of the Main Hypothesis Taken from Measurements of 
Photosynthetic Quantum Yield Made by Other Observers 

Certain observations connected with the measurements of quantum 
yields by Rieke (5) and Kok (5) are evidence of this correlation. Both 
authors observe, in acid solutions, curvature in the plot of photosyn¬ 
thetic rate against light intensity below and in the neighborhood of the 
compensation point, indicating a higher quantum yield at these 
intensities, and both find that the quantum yield is somewhat smaller 
when an alkaline buffer is used. Both authors used Warburg’s manom¬ 
eter method for their quantum yield determinations. However, the 
method differs from the one used by Warburg, Emerson, and by Rieke 
himself in his earlier paper, in that the measurements are made with 
only 20-40% absorption rather than total light absorption (this method 
requires, in addition to the measurement of the intensity of the incident 
light, another one of the amount of light which passes through the algal 
suspension without being absorbed. Because of the light scattering, the 
last mentioned measurement is made with the help of an integrating 
sphere). Therefore, during the illumination period, all the algae are 



304 


JAMES FRANCK 


continuously irradiated and, in a first approximation, by light of a con¬ 
stant intensity. Thus, even if the chloroplast membranes are permeable, 
a high contribution of the reduction of respiratory intermediates to the 
overall photosynthetic gas exchange cannot be expected in the region 
above the compensation point. 

Most of Rieke’s quantum measurements were made in alkaline buffer. 
Quantum yields between 1/11 and 1/12 with no dependence upon light 
intensity were observed under these conditions. However, in connection 
with rate measurements of anaerobic photoreduction in Scenedcsmus 
(where hydrogen uptake replaces the oxygen evolution of photosyn¬ 
thesis (8)), he compared photosynthetic quantum yields and rates in 
slightly acid solution with those in alkaline buffer. The use of the acid 
medium gives a slightly bent rate curve and a quantum yield ~ 20 
25% higher than that given in the alkaline buffer. A few measurements 
of this kind made with Chlorclla showed a similar decrease of the quan¬ 
tum yield due to alkalinity. The lowering of the quantum yield by 
alkali in these algae is, without doubt, the result of cellular damage 
and is not caused by removal of the respiratory intermediates from 
participation in photosynthesis. The latter influence could only 
have been responsible for ~ 5% of the decrease, judging by the small 
deviation of the rate curve from linearity. It. is, therefore, justifiable to 
correct, as Rieke does, the quantum yield measured in alkaline buffer 
to values between 1/9 and 1/10. 

Kok’s paper on quantum yields deserves special mention, because he 
not only observes the alkali damage and a rate curve in acid medium 
with a bend at low light intensities, but he also indicates clearly that 
the bend may be responsible for Warburg’s high quantum yield of 1/4. 1 
His plant material was Chlorella suspended in culture solution. Since he 
uses quite low algal concentrations, the pressure changes are very 
small at low light intensities. He, therefore, refrains from measuring the 
shape of the curve in this region. However, on the basis of extrapolation 
of the curve measured at higher light intensities, he concludes that this 
part of the curve must be bent. The directly observed portion of the 

1 However, it may be emphasized that we cannot agree at all with Kok’s interpre¬ 
tation of the steeper slope of the rate curve at low light intensities. He introduces the 
assumption that normal respiration is stopped by irradiation and replaced by a photo¬ 
chemical process giving two energy-rich phosphorylated bonds per quantum. No 
attempt is made to explain why or how respiration is coupled with the hypothetical 
photochemical reaction. 



PHOTOSYNTHETIC QUANTUM YIELDS 


305 


curve is strictly linear and is measured bcwteon a light intensity giving 
a photosynthetic rate approximately three times the rate of respiration 
and one giving a rate nine times the rate of respiration. If the curve, 
corrected for respiration, is extrapolated to zero intensity, it does not 
hit the zero point of gas exchange but rather a point about halfway 
between zero and the point at which respiration is compensated. Kok 
concludes that the slope of the curve at the very lowest light intensities 
must be twice the slope of the linear portion observed at higher inten¬ 
sities. Using only the latter part of the curve, he claims that the quan¬ 
tum yield of photosynthesis measured by him has the value 1 /7.5 and, 
because that is greater than 1/8, he rejects theories based on the theo¬ 
retical yield 1/8. (The experimental limit, of course?, should be some¬ 
what lower than the theoretical one.) We can discard this part of his 
conclusions because Rieke found an error in the photometry of Kok 
which makes the actual quantum yield more than \(Y/ ( lower than the 
one calculated by this author. The corrected values are — 1 9 and an*, 
therefore, in agreement with the yield as measured by most observers, 
and half as high as Warburg’s. In a few measurements made in an 
alkaline buffer he found that the rate curve had a slope 2(YY ( less steep 
than the one observed in the culture solution. 

Connection Between the Hypothesis ami Warburg's Observations of 
Cyanide Influence on Photosynthetic Rates in CMorelia 

There arc further observations which support the assumption that 
photochemical reduction of respiratory intermediates can occur in 
plants. The first, and most important, observation of this type was 
made by Warburg (9) in the year 1920 and interpreted by him by the 
very assumption we make use of in the present paper. He found that 
respiration in CMorelia is much less sensitive to cyanide than is photo¬ 
synthesis. CConcentrations of this poison, which slightly increase 
respiration, are able to suppress photosynthesis entirely. However, 
the gas exchange during illumination does not fall below the compensa¬ 
tion point, even if much greater cyanide concentrations are used, which, 
as measurements during dark periods show, increase respiration con¬ 
siderably. Warburg’s contention is the following: CO 2 has to undergo a 
chemical dark reaction to transform it into a substance which can be 
reduced photosynthetically. Cyanide in moderate concentrations 
poisons this preparatory reaction and, therefore, makes C0 2 photo¬ 
synthesis impossible. Intermediates of respiration do not need such 



306 


JAMES FRANCK 


preparatory dark reactions to become photochemically reducible. If 
they are caught by the photochemical reaction before being evolved as 
CO 2 , the gas exchange of respiration is compensated and no pressure 
changes are observed in the light. 

Van der Paamv (10) was the first one to become suspicious as to 
whether there was not a connection between Warburg’s results con¬ 
cerning the influence of cyanide on algae and his determination of the 
quantum yield, but apparently Warburg did not regard it as likely. 
He probably saw clearly that a great difference in conditions exists 
between the two types of experiments. The reduction of respiratory 
intermediates may be the only photochemical reduction process occur¬ 
ring under conditions where photosynthetic CO 2 reduction is prevented, 
but that does not mean that the former process may be able to compete 
with the latter if the preparatory CO 2 fixation is able to proceed without 
hindrance. After all, photosynthesis can proceed approximately 20 
times as fast as respiration and thus, at low light intensities, the concen¬ 
tration of the fixed C0 2 will be great compared with the concentration 
of respiratory intermediates. The effective concentration of the latter 
must be even smaller if one takes into account the fact that most of the 
respiration (perhaps all of it) takes place in parts of the cell outside of 
the chloroplasts, and that the diffusion of the respiratory intermediates 
into the chloroplasts may be slow. 

Other observations are described in the literature which fit into the 
same pattern. Algae or leaves, irradiated under conditions where CO 2 is 
removed as efficiently as it can possibly be done, show that the gas 
exchange observed in the dark is fully compensated for in the light (11). 
The experiments indicate that, in this case, not only the intermediate 
products of respiration are photochemically reducible but also inter¬ 
mediate products of photoxidation. 

The intermediates of respiration have a chance to be photochemically 
reduced only if they penetrate into the chloroplasts. According to the 
hypothesis put forward in the present essay, the premeability of the 
chloroplast membranes to these substances is not a normal property but 
rather one easily imposed on the membranes (especially in (Morelia) 
by suitable pretreatment of the plant material. This conclusion lends 
itself to experimental tests. For instance, it would be of interest to find 
out whether very young and diffuse cultures of Chlorella, not subjected 
to dense packing by strong centrifugation, will behave like Warburg’s 
if irradiated in the presence of cyanide, or whether, in this case, cyanide 



PHOTOSYNTHETIC QUANTUM YIELDS 


307 


may be able to lower the gas exchange below the compensation point 
during exposure to light, because of the impermeability of the chloro- 
plast membranes to the respiratory intermediates. To avoid damage by 
photoxidation, moderate light intensities should be used which, in the 
presence of cyanide, are sufficient to reach light saturation. 

IV. Connection between the Main Hypothesis and 
Measueements of the Chlorophyll 
Fluorescence in Plants 

/. Respiration Intermediates are Preferentially 
Adsorbed ^at the Chlorophyll 

Our explanation is only possible if a strong preference exists for the 
reduction of respiratory intermediates. Fixperimental evidence indicates 
that the photosynthetic apparatus has indeed a special affinity for 
these substances. After Kautsky and coworkers (12) first observed that 
the chlorophyll fluorescence becomes several times stronger during the 
induction period than during the steady state of photosynthesis, much 
attention was paid to the general relation between fluorescence inten¬ 
sity and photosynthetic rates (12). Although no unanimity exists in 
respect to the interpretation, there is at least full agreement about the 
facts. The most important ones are the following: The fluorescence 
yield of the chlorophyll is always very low in plants. However, there 
exists a general relation between fluorescence intensity and the rate of 
photosynthetic oxygen production (of course, it must be recognized that 
quick changes in gas exchange cannot be recorded undistorted because 
of the slowness of the adjustment of gas equilibria between water and 
the atmosphere, whereas fluorescence intensities can be measured 
practically free of inertia of the recordihg instrument). If the fluores¬ 
cence intensity becomes several times greater than normal, photo¬ 
synthetic oxygen production becomes many times smaller. Measure¬ 
ments during the induction period clearly show the relationship. The 
fluorescence intensity shows a steep rise in the first seconds of irradia¬ 
tion followed by a slow transition to normal intensities in about one 
one minute. Correspondingly, the rate of oxygen production falls rapidly 
from the initial value to a deep minimum and rises in about a minute 
to its steady state. Sometimes with higher plants, and often with algae 
(especially Chlorella), the fluorescence intensity-time curve looks 
different. The decay after the first outburst is somewhat more rapid, 



308 


JAMES FRANCK 


and a .second broader wave of fluorescence intensity rise and fall 
develops. A corresponding time course is observed for the rate of 
oxygen production (13). The induction phenomena are more pro¬ 
nounced in old and dense algae cultures, and in leaves, than in sus¬ 
pensions of young algae. In some cases, very young and diffuse cultures 
of algae show no anomalies of the fluorescence (12) and no induction 
losses of the rates (5) at the beginning of the irradiation periods. How¬ 
ever, anaerobic treatment of young algae causes induction phenomena 
to occur. If the chloroplast membranes are permeable due to age, 
density, prolonged anaerobicity, or other conditions, alkali can pene¬ 
trate into the chloroplasts and the anomalies of the fluorescence 
intensity disappear entirely. It is, however, sufficient that alkali pene¬ 
trates merely the outer cell membranes. This enhances the degree of 
ionization of the plant acids and thereby lowers their ability to per¬ 
meate the chloroplast membrane. There is also a strong resemblance 
between the fluorescence curve and the one of C0 2 evolution vs. time 
as observed by Emerson during the first minutes of irradiation. The 
first high fluorescence maximum corresponds to the strong C0 2 out¬ 
burst and the second lower fluorescence maximum corresponds to a 
second weaker outburst of C0 2 . 

Detailed explanations of these and connected phenomena are given 
elsewhere by the author and coworkers (12). A brief summary of the 
conclusions to which the theory leads follows: The cause of the induc¬ 
tion phenomena is the inactivation of the oxygen-liberating enzyme by 
a metabolic excretion product of the algae which is water-soluble and 
which penetrates from the suspension into the chloroplasts. This 
poison is destroyed by oxidation, the photosynthetically produced 
peroxides being efficient oxidizing agents. The poison concentration in 
the cells is, therefore, greater during dark periods than during illumina¬ 
tion, and consequently, at the start of an illumination period, the 
activity of the oxygen-liberating enzyme is low. Therefore, the photo¬ 
peroxides are not removed quickly and will attack easily oxidizablc 
substances such as carbohydrates. The first oxidation products of the 
carbohydrates are acids. These acids possess surface-active properties 
and displace C0 2 complexes and photosynthetic intermediates from 
the surface of the chlorophyll complexes. Just as surface-active narcot¬ 
ics, they increase the fluorescence intensity when they are adsorbed 
at the chlorophyll. The rapid rise in fluorescence intensity at the very 
start of the irradiation period is the result of the production of these 



PHOTOSYNTHETIC QUANTUM YIELDS 


309 


acids. The decay of the fluorescence to a steady state corresponds to a 
decreasing production of these acids because the oxygen-liberating 
enzyme becomes reactivated by removal of the poison by oxidation. 
The absence of the fluorescence anomalies in the presence of alkali 
indicates that the ionized acids are not strongly adsorbed at the 
chlorophyll or that they are unable to penetrate the chloroplast 
membranes. 

. 2. Connection between Fluorescence Anomalies and 

the Emerson Effect, etc. 

The time course of the Emerson effect coincides with that of the 
fluorescence intensity because the acids, which are abundant when the 
fluorescence yield is high, cause the liberation of CO 2 from the sub¬ 
stance in which it is fixed preparatory to photosynthetic reduction. 

In Warburg’s algae the acid intermediates of respiration are already 
present at the start of the irradiation. The concentration of the prepara¬ 
tory fixation product of CO-» should, therefore, be smaller and the rise 
of acidity connected with the induction phenomena will then give a 
smaller Emerson effect than the one occurring in Emerson’s algae. If 
the surface-active acids can be made in the chloroplasts by an oxidation 
reaction between carbohydrates and photosynthetically produced 
peroxides, they should also appear under conditions where photoxida- 
tion occurs. That is indeed the case. When CO* is the limiting factor 
responsible for saturation, the fluorescence yield rises in the presence 
of sufficient oxygen to cause photoxidation. However, when only 
enough oxygen to sustain respiration is admitted, practically no rise of 
the fluorescence yield occurs because no acids are made by photoxida¬ 
tion. When oxygen is entirely .absent, fermentation occurs and the 
acids produced by this process are able to penetrate to the photosyn¬ 
thetic apparatus because anaerobicity increases the chloroplast mem¬ 
brane permeability. As a result, the fluorescence yield is high at the 
very beginning of the irradiation. A rise in fluorescence intensity is also 
observed if, by prolonged irradiation, the carbohydrate concentration 
becomes high or if sugar is introduced into the leaves from outside. A 
high concentration of carbohydrates favors the reaction between 
peroxides and carbohydrates with the formation of surface-active 
t,cids, but the main reason for the rise in fluorescence intensity in this 
case seems to be the poisoning of the oxygen-liberating enzyme by a 



310 


JAMES FRANCK 


metabolic product and the consequent increase in the peroxide con¬ 
centration. 

3. Preferential Reduction of Respiration Intermediates 

All of the observations mentioned clearly indicate that partially 
oxidized carbohydrates containing one or more acid groups are strongly 
adsorbed at the chlorophyll complexes and arc able to displace other 
substances from contact with the chlorophyll even if the latter are 
present in much higher concentration. This property of the acids is 
used by the plant to control photochemical activity and prevent photoxi- 
dation or excessive production of photosynthates. Thus, without ex¬ 
ception, we observe that a rise of fluorescence intensity is connected 
with a decline in the photosynthetic rates. This seems to be in direct 
contradiction to the hypothesis on which our interpretation of War¬ 
burg’s quantum yield measurement is based, namely, that such plant 
acids are preferentially reduced photochcmically and, under certain 
conditions, are responsible for an increase in photochemical activity 
rather than for a decrease. However, the contradiction may be only an 
apparent one. The conclusion that 8 quanta are necessary to reduce one 
COj molecule implies that two quanta are needed to split each water 
molecule. According to the kinetic theory of Franck and Herzfeld (14), 
this can be achieved if chlorophyll complexes activated by absorption 
of a light quantum serve as intermediate hydrogen donors and get back 
from water their hydrogen with the help of a second light absorption 
act. This means that the chlorophyll complexes have to react alternately 
with the substances to be reduced and with water. If the substances to 
be reduced are as strongly adsorbed as the intermediates of carbo¬ 
hydrate oxidation, this alternation will work efficiently only so long as 
their concentration is small. If these intermediates are produced by 
photoxidation or by reaction between carbohydrates and photosyn¬ 
thetic peroxides, their concentration will be large and they will cover 
both the hydrogenated and the dehydrogenated chlorophyll complexes 
and prevent the rehydrogenation of the latter. As a result, all photo¬ 
chemical activity will stop. According to Clendenning and Gorham (15), 
the rate of quinone reduction by illuminated chloroplasts first rises and 
then quickly declines as the quinone concentration is increased, even 
under conditions where light intensity is the limiting factor. The 
interpretation may be a similar one; a substance may react photochemi- 
cally at low concentrations and act as a narcotic at higher ones. 



PHOTOSYNTHETIC QUANTUM YIELDS 


311 


The complete cessation of all photochemical activity when enough 
partially oxidized carbohydrate is adsorbed on the chlorophyll com¬ 
plexes is not the only possible explanation of the changes in fluores¬ 
cence intensity and the regulation of the photosynthetic output of 
oxygen. The same result can be achieved if the partially oxidized mole¬ 
cules are reduced photochemically and then immediately reformed by 
oxidation. The same limited amounts of material can thus be used over 
and over again, and the photosynthetic apparatus will be transferred 
from a state in which it does work to a state resembling that of an 
idling engine. Suitable experiments may make possible a decision 
between the two explanations, but the writer has thus far been unable 
to propose the right kind of experiments. 

V. The Reason fok the Conflicting Results in tiie 
Determination of Piiotosynthetic Intermediates 
May Re the Same as the One Responsible for 
the Contradictions in Quantum Yield Measurements 

Finally, a few remarks may be added supplementing those made in 
the introduction about the similarity between the contradictions en¬ 
countered in the attempts to identify intermediates of photosynthesis 
and the ones discussed above in the interpretation of the measurements 
of photosynthetic quantum yield. Calvin and coworkers (16) came to 
the conclusion that substances known to be intermediates of respiration 
act also as intermediates of photosynthesis. Gaffron, Fager and Brown 
(16), on the other hand, find entirely different properties in the sub¬ 
stance or substances which, according to their results, must be an 
intermediate of photosynthesis. Fager, who has not fully identified 
the chemical nature of this substance, presents chemical evidence 
excluding all the substances mentioned by Calvin and coworkers. The 
Chicago group stated at the meeting of the Am. Assoc. Advancement 
Sci., in December, 1947, that the observations made in California seem 
to be related to respiration (and fermentation) of the plants. The 
present writer is in agreement with that statement but believes that 
all interactions between respiration and photosynthesis must also be 
taken into account. Such influences will occur if the chloroplast mem¬ 
branes of the plants used in California become permeable to respiration 
intermediates. Indeed the pH of 4.5-5 used in the California experi¬ 
ments is favorable for membrane permeation of plant acids, while 
pH 7.8 used in Chicago will prevent that process. 



312 


JAMES FRANCK 


Summary 

Warburg’s observation, that the quantum yield of photosynthesis is 
~ 1 /4, deviating from that of others by a factor of two, can be ex¬ 
plained by the assumption that, under the particular conditions of his 
experiments, a photochemical reduction of intermediate respiration 
products rather than C0 2 reduction is observed. 

The condition necessary for the replacement of normal photosyn¬ 
thesis by this other process seems to be an abnormal permeability of 
the chloroplast membranes. The consequences of this hypothesis have 
been compared with experimental evidence. 

Indications are mentioned for the assumption that the contradictions 
in the results gained so far on intermediates of photosynthesis may be 
caused by the same variability of the permeability of membranes which 
is supposed to be responsible for Warburg’s results on quantum yields. ' 

Note Added in Proof 

Since this paper was written, 0. Warburg, D. Burk, V. Schocken and S. Hendricks 
have carried out new experiments on photosynthetic quantum yield and have reported 
their findings at a meeting of the Society of General Physiologists, at Woods Hole, on 
June 22, 1949. 

The evaluation of these data must wait until they are published in extenso. However 
two sets of experiments have been presented which seem to make it necessary to 
change one of the main assumptions used in the present paper. They are: 

(a) Quantum yield measurements with the Warburg method gave again a yield of 
1/4 provided the suspension was quite acid (pH ~ 4—5); higher pH values gave 
smaller yields approaching the yield of ~ 1/10. The measurements have been 
extended well into the region above the compensation point. Since the narrow 
beam of red monochromatic light (from the monochromator) was not intense enough 
to overcompensate several times the respiration of all the algae (only about 1/10 to 
1/20 of the algae in the vessel were exposed simultaneously to the light), the beam was 
superimposed on a general irradiation of the vessel with white light of unknown abso¬ 
lute intensity. The white light, though of considerably weaker intensity than the red 
beam, illuminated a larger portion of the algae, and was adjusted to give gas exchanges 
about 3 times those of respiration. The quantum efficiency of the red light remained 
unaltered by irradiation with the white light. 

(b) When the CO 2 concentration in the vessel was kept exceedingly low, illumina¬ 
tion with the red light beam did not measurably change the oxygen consumption of 
normal respiration. 

It is difficult (though not impossible) to reconcile the hypothesis that reduction of 
half-oxidized respiration products is responsible for Warburg's high quantum yield, 
with the fact (a) that it remains high even when the total photosynthetic activity 
becomes several times higher than respiration. Moreover, the fact mentioned under 



PHOTOSYNTHETIC QUANTUM YIELDS 


313 


heading (b) is in direct contradiction to that hypothesis, which therefore fails to give 
a satisfactory explanation of Warburg's results. However, the present author is still 
convinced that the above discussion contains a good deal of material useful for the 
reconciliation of the differences in the results of quantum yields and of the chemical 
nature of intermediates of photosynthesis. He believes that there are two different 
photosynthetic processes, one with the quantum yield of 1/4, the other with 8 quanta 
and that the former is exceptional, taking place only when the chloroplast membranes 
become permeable, thus permitting mutual interactions between photosynthesis and 
respiration. High acidity, which discharges acids, is obviously one of the conditions 
necessary for the occurrence of that process. 

If, as the new experiments indicate, the replacement of CO % reduction by that of 
half-oxidized respiration products is not responsible for the higher quantum yield 
process observed by Warburg, it might be that a part of the energy of respiration can 
be used for photosynthetic processes. It is often assumed that energy-rich phosphate 
bonds are made by the light and that their energy is used to reduce CO 2 in dark reac¬ 
tions similar to the synthesis occurring as a by-product of dark respiration. We have 
many reasons to reject this idea. However, it might be possible that the energy stored 
in the phosphate bonds produced by respiration might be transferred to phosphate; 
bonds of the CO 2 complex and of intermediate products of photosynthesis. In that 
way, the energy of 12 K-cal. would be available in the molecules to be reduced before 
each photochemical reaction and, with that additional energy, photosynthesis may 
proceed with 4 quanta. However, this photosynthesis could, even if all other conditions 
are favorable to it, only proceed to a maximum rate of 1.5 times that of respiration. 
Any photosynthesis in algae beyond 1.5 times respiration would need 8 quanta. 

If we want to explain by that hypothesis some of Warburg’s new results mentioned 
under (a) wo must introduce the assumption that, in Warburg's special arrangements, 
t he algae exposed to the strong but short-lasting illumination of the red beam are in a 
better position to make use of the energy of the phosphate bonds for photosynthesis 
t han those algae exposed for a longer time to the weaker intensity of the white light. 
It is possible to give kinetic reasons for such behavior, but this would be premature as 
long as insufficient experimental evidence is available in support of the main hypo¬ 
thesis. 

References 

1. Warburg, 0., and Neoelein, E., Z. phys. Chem . 102, 235 (1922); 106, 191 (1923). 

2. Manning, W. M., Stauffer, J. F., Duggar, B. M., and Daniels, F., J. Am. 

Chem. Soc. 60, 266 (1938); Manning, W. M., Juday, C., Nolt, M., ibid. 60, 
274 (1938); Petering, H. S., Duggar, B. M., and Daniels, F., ibid . 61, 3525 
(1939); Mayer, J. L., de Witt, T. W., and Smith, E., ibid. 61, 3529 (1939); 
Dutton, H. J., and Manning, W. M., Am. J. Botany 28, 516 (1941). 

3. Hi eke, F. F., J. Chem. Phys. 7, 238 (1938). 

4. Emerson, R., and Lewis, C. M., Carnegie Inst. Wash. Yearbook 39,154 (1940); 

Am. Jour. Botany 28, 789 (1941); Emerson, R., and Nishimura, M. S., 
"Photosynthesis in Plants," Ch. 10, p. 219. Monograph Am. Soc. Plant Physiol., 
Iowa State College Press, 1949. 

5. Moore, W. E., and Duggar, B. M., "Photosynthesis in plants," op. Cit.y Ch.16, 

p. 239; Rieke, F. F., ibid. Ch. 12, p. 251; Arnold, W., ibid. Ch. 13, p. 273; 



314 


JAMES FRANCK 


Kok, B., “Critical Considerations of the Quantum Yield of ChloreUa Photo¬ 
synthesis,” Proefschrift Universiteit, Utrecht-Amsterdam, 1948. 

6. Warburg, 0., Am. J.'Botany , 35, 194 (1948). 

7. Gaffron, H., Cold Spring Harbor Symp. Quant. Biol. 7, 377 (1939). 

8. Gaffron, H., Am. J . Botany 27, 273 (1940). 

9. Warburg, 0., Biochem. 103, 188 (1920). 

10. Van per Paauw, Rec . trav. botan. neerland 29, 497 (1932). 

11. Gaffron, H., Biochem. Z . 292, 241 (1937); Franck, J., anjd French, C. S., J. 

Gen. Physiol. 25, 309 (1941). 

12. Kautsky, H., and Eberlein, E., Biochem. Z . 302, 137 (1939); Kautsky, H., 

and Hirsch. A., Z . anorg. aUgem. Chem . 222, 126 (1935); Kautsky, H., and 
Zedutz, W., Naturwissenschaften 29, 101 (1941); McAlister, E. D., and 
Myers, J., Smithsonian Inst. Misc. Collections 99, No. 6 (1940); Wassink, E. 
C., and Katz, E., Enzymologia 6, 145 (1939) ; Wassink, E. C., Vermeulen, D., 
Reman S. H., and Katz, E., ibid. 5, 100 (1938) ; Katz, E., “Photosynthesis in 
Plants,” op. cit., Ch. 15, p. 287; Franck, J., French, C. S., and Puck, T. T., J. 
Phys. Chem. 45, 1268 (1941); Shiau, Y. G., and Franck, J., Arch. Biochem. 14, 
253 (1947); Franck, J., “Photosynthesis in Plants,” op. cit. Ch. 16, p. 293. 

13. Franck, J., Pringsheim, P., and Lad, D. T., Arch. Biochem. 7, 103 (1945). 

14. Franck, J., and Herzfeld, K., J . Phys. Chem. 45, 978 (1941). 

15. Clendenning, K. A., and Gordon, P. R., Can. J. Research in press. 

16. Benson, A. A., Calvin, M., Science 105, 648 (1947); 107, 476 (1948); 109, 140 

(1949); Benson, A. M., Calvin, M., Haas, V. A., Aronoff, S., Hall, A. G., 
Bassham, J. A., and Weigel, J. G., “Photosynthesis in Plants,” op. cit. Ch. 19, 
p. 381 ; Brown, A. H., Fager, E. W., and Gaffron, H., Arch. Biochem. 19,407 
(1948); “Photosynthesis in Plants,” op. cit . Ch. 20, p. 403; Fager, E. W., 
ibid. Ch. 21, p. 423. 



Transformation of Tryptophan to Nicotinic Acid Investigated 
with Delayed Supplementation of Tryptophan 1 

E. Geiger, E. B. Hagerty and H. D. Gatchell 2 

From the Von Camp Laboratories, Terminal Island, and the Department of Physiology, 
School of Medicine, Univ. of Southern California, Los Angeles, Calif. 

Received March 7, 1040 

Introduction 

Since the experiments of Krehl, Tepley and Elvchjem (1) on the 
cmative effect of tryptophan in experimental niacin deficiency, several 
authors have reported that the excretion of nicotinic acid is increased 
more after the administration of extra tryptophan in the form of the 
pure amino acid than in the form of extra casein (2). This finding 
suggests that tryptophan may also he utilized in a way which is in¬ 
dependent of protein synthesis. 

It has been shown in earlier experiments (3,4) that tryptophan, fed 
as a delayed supplement to rats on tryptophan-free amino acid mix¬ 
tures, neither promotes growth nor prevents cataract formation. It 
was, therefore, assumed that the same technique might be used to 
determine whether or not the participation of tryptophan in niacin 
synthesis occurs independently of protein synthesis. 

It was further intended to investigate whether vitamin B 6 is specifi¬ 
cally involved in the transformation of tryptophan (5), or whether 
the absence of this essential nutrient interferes in a non-specific way 
with the nicotinic acid production. Since there are no indications that 
the oil-soluble vitamins A and D present in cod-liver oil are directly 
involved in the amino acid metabolism, we investigated the effect on 
nicotinic, acid formation of a diet containing all the necessary growth- 
factors with the exception of cod-liver oil. 

‘This work has been supported in part by a research grant from the National 
Vitamin Foundation. 

* One part of the experiments was submitted by H. D. Gatehell in partial fulfillment 
of the requirements for the degree of Master of Science in the Department of Physiol- 
ogy, School of Medicine, University of Southern California. 

315 



31G 


E. GEIGEH, E. B. HAGEHTY AND H. D. GATCHELL 


Methods 

Infantile male Sprague-Dawlcy rats of 40-45 g. body weight were placed in indivi¬ 
dual cages provided with wire mesh bottoms. The body weight of the animals and t he 
food consumption were determined daily. 

The rats had access to their diets from 4 P.M. until 7 A.M. the next day and the 
supplement was administered by dropper at 12 M. The tryptophan solution was 
prepared by dissolving 75 mg. of i.-tryptophan in 25 ml. of hot 2% gelatin solution. 


TABLE I 

Composition of Diets 


: 

, 

Component 

Diet mmilxM 

I 

II 

III 

IV 

V 

VI 

1. Basal diet" 

86 g. 

86 g. 

86 g. 

86 g. 

86 g. 

86 g. 

2. Vitamin-free 
casein 

i 

!> g- 

!> g- 

g- 

10 g. 

10 g. 

10 g. 

3. Tryptophan-free 
Casein acid hy- 
drolyzate 



— 

3g. 

3 g. 

3 g. 

4. Gelatin 

6g. 

6 g. 

6g. 

- 

.... 


5. L-Tryptophan 

— 

: 

150 mg. 

— 

— 

100 mg. 

6. Nicotinic acid 


20 mg. 

— 

— 

— 

-- 

7. Cod liver oil 

2Xwk. 

2 drops 

2Xwk. 

2 drops 

2Xwk. 

2 drops 

2Xwk. 

2 drops 

— 

— 


a Basal diet: 820 g. sucrose, 30 g. corn oil, 40 g. U. S. P. salt mix, 4.0 mg. thiamin, 
6.0 mg. riboflavin, 5.0 mg. pyrkloxine, 40 mg. calcium pant-hothenate, 1 g. choline, 
2.0 mg. vitamin K, 200 mg. inositol, 0.2 mg. biotin* 1.0 mg. folic acid, 1.6 g. cystine. 


This solution could be cooled to about 30°C. without precipitation of the amino acid. 
The niacin solution contained 10 mg. niacin in 2.5 ml. 2% aqueous gelatin solution. 
The animals not receiving supplement were given at the same time equivalent amounts 
of a 2% gelatin solution. For the composition of the diets see Table I. 

The nicotinic acid excreted in the urine was determined by the modified colori¬ 
metric method of the Research Corporation (6). 



TRYPTOPHAN-NICOTINIC ACID TRANSFORMATION 


317 


Experimental 

A. In the first group of experiments the rats were placed on a deficient diet No. I. (7). 

After a depletion period of 0 days, during which the rats did not gain weight, the 
animals were divided into 5 groups, each group containing 4 rats. The condensed 
results are presented in Table II. 

The rats of Gp. 1, continued on Diet I, did not gain more than the animals of Krehl 
el al. (7) on a similar deficient diet. In Gp. 2, the addition of nicotinic acid improved 
the growth considerably. In Gp. 3, a diet, supplemented with tryptophan, resulted in 
a similar improvement of growth. In Gp. 4, the rats received the deficient Diet I during 
the night and at 12 M., as a delayed supplement, 0.5 ml. of the niacin solution. In 
Gp. 5 the rata were kept on deficient Diet I for the night and received as a delayed 
supplement at 12 M, 0.5 ml. of the tryptophan solution. 


TABLE II 

Average Growth of Rats with Delayed Supplementation of Tryptophan for 21 Days 


(lump no.® 

Diet (see 
Table I) 

Supplement 

Avei age 
group gain 
I>ei day 

Average 
individual 
gam per day 

Average 
group dietary 
intake per 
day 

Ratio of 
dietary in¬ 
take over 
weight gain 

1 

I 

_ 

y. 

1.2 

Q. 

1.7 1.2 

0 • 

0 




' 


0.9 1.2 

<> 

2 

II 

- 

2.2 

l 

1.9 2.4 
2.3 2.4 

7 

3.2 

3 

III 

— 

2.4 

3.3 2.0 

2.3 2.0 

8 

3.3 

4 

I 

Niacin 

2.0 

3.2 2.4 

Q 

3.2 





2.5 2.5 

O 

5 

I 

Tryptophan 

2.5 

2.0 2.0 

2.0 1.8 

7 

3.5 


a Each group contained 4 rats. 


Comparing the growth of the Gps. 2 and 3 with that of 4 and 5 shows that nicotinic 
arid and tryptophan fed as a delayed supplement promotes growth as weU as when fed 
simultaneously with the deficient diet 

B. In another group of experiments we investigated whether the delayed supple¬ 
mentation of tryptophan also increased the nicotinic acid excretion. In these experi¬ 
ments, the rata were depleted 14 days on an imbalanced diet, No. IV. After this 
depletion period, Gp. 6 received the same imbalanced diet daily from 4 P. M. until 



318 E. GEIGER, E. B. HAGERTY AND H. D. GATCHELL 

7 A. M. and no supplement. Gp. 7 received an imbalanced diet at night and at 11 A. M. 
a supplement of 20 mg. tryptophan. 

Table III shows that, in both groups, nicotinic acid excretion decreased during the 
depletion period and continued to be low in Gp. 6 . In Gp. 7, which received tryptophan 
as a delayed supplement, however , with the resumption of growth there was a considerable 
increase in nicotinic acid excretion. 

C. In a third group of experiments we investigated rats which did not receive cod 
liver oil in their diet. These rats wore placed on an imbalanced Diet V for 7 days. 
After this period, the rats were divided into 2 groups. Gp. 8 received, from 4 P. M. to 
7 A. M., Diet VI containing tryptophan. The rats of Gp. 9 received the imbalanced 
diet from 4 P. M. until 7 A. M., and at 11 A. M. tryptophan as a delayed supplement. 

During the depletion period the average nicotinic acid excretion in both groups 
dropped from 147 y and 16 l 7 /day to 31? and 187 /day for Gps. 8 and 9, respectively. 
Growth also declined to about 1 g. per rat per day. 

During the next 11 days the growth and the nicotinic acid excretion increased 
slightly in both groups. At the end of this period the symptoms of vitamin A defi¬ 
ciency, such as ophthalmy, became apparent and, at the same time, growth an^l 
nicotinic acid excretion decreased in both groups to a very low value. At this time the 
animals did not* show any signs of vitamin D depletion. 

On the 19th day, cod liver oil was administered and, with the disappearance of 
symptoms of vitamin A deficiency, rapid growth started, paralleled by a considerable 
increase in nicotinic acid excretion. 

These experiments show' that omission of cod liver oil from the diet interferes not only 
with growth bid also with urinary nicotinic acid excretion, and that , after the addition 

TABLE III 


Nicotinic Acid Excretion of Hats on Delayed Supplementation of Tryptophan 


Rat group 
no. a ' 

Duration of 
period 

Diet (see Table I) 

, 

Daily av. gain 
per rat 

Daily av. food 
intake per rat 

Av. nicotinic 
acid excretion 
in urine 


day* 







A. Depletion period 



6 

14 

IV 

0.3 

3.9 

48 

7 

14 

IV 

0.4 

3.5 

65 

B. Test period 

6 

6 

IV 

-0.14 

! 

3.5 

55 

7 

6 

IV and delayed 
supplementation 
of tryptophan 

1.8 

5.5 

161 


* Each group contained four rats. 





TRYPTOPHAN-NICOTINIC ACID TRANSFORMATION 319 


TABLE IV 

Nicotinic Acid Excretion of Vitamin A and D Depleted Hats 


Rat group 
no.* 

Duration of 
period 

Diet (see 
Table I) 

Supplementation with Daily av. 
tryptophan gain per rat 

Daily av. 
food intake 
per rat 

Av. nicotinic 
acid excretion 
m urine 


clayft 


<J- 

A. Depletion period 

u. 

y 

8 

; 7 

V 

! io 

6.5 

(14 ir 31 

9 

7 

V 

1 

j 1.0 

6.3 

(161)" 18 

_ 


1 

! _ ___ J—-. 

_ __ __ __ 



B. Tryptophan supplementation period 


8 

9 

19 

19 

VI 

V 

Mixed in the diet 
Delayed supplement 

0.8r> 

0.7 

! 6.8 
j 6.2 

40 

36 


C. 

Cod liver oil and tryptophan supplementation 


8 

13 

VI 

Tryptophan mixed 

2.5 

9 

106 




in diet 


i 


9 

13 

V 

Delayed supple men- 

2.8 

8 

148 




tation 




1 Value 

on first day of depletion |K*riod. 





k Each group contained four rats. 

of cod liver oil, growth and nicotinic acid excretion are increased nearly equally, inde¬ 
pendently of whether the tryptophan is mixed in the diet or fed as a delayed supplement. 

Discussion 

These experiments indicate that tryptophan given as a delayed 
supplement to niacin-deficient rats increases growth, and augments the 
urinary excretion of nicotinic acid. It was shown earlier that feeding of 
delayed supplements of tryptophan to animals kept on tryptophan- 
free diets, does not promote growth (3) and does not prevent cataract, 
formation (4). Therefore, the corrective effect of tryptophan in niacin 
deficiency seems to be exerted independently of protein synthesis by 
direct transformation of this amino acid into nicotinic acid. 

It should be pointed out that the animals in the present experiments 
grew, even though tryptophan was given as a delayed supplement. This 
seems to be in contradiction to earlier experiments, in which a totally 



320 


E. GEIGER, E. B. HAGEllTY AND H. D. GATCIIELU 


tryptophan-free diet was supplemented. In the present experiments, 
however, the deficient diet per se contains enough tryptophan, in the 
form of casein, to promote growth, and only the harmful effect of the 
added imbalanced amino acid mixture had to be corrected by supple¬ 
mentation of tryptophan. 

These experiments show, secondly, that omission of cod liver oil 
from the diet interferes with the growth-promoting effect of tryptophan. 
There was a possibility that in such animals where growth, i.e., 
apposition of newly formed body protein, ceases, a larger amount of 
tryptophan may be transformed to niacin. The experiments, however, 
show that deficiency of vitamins, normally supplied with cod liver oil, 
interferes with the transformation of tryptophan to niacin in the same 
way as does vitamin B« deficiency. 

Since the completion of these experiments, a paper by Junqueira and 
Schweigert (5) appeared showing that, besides vitamin B« deficiency, 
a deficiency of Bj, B 2 , pantothenic acid, or folic acid decreases the 
transformation of tryptophan to nicotinic acid. Our results, and the 
results of the above authors, suggest that Vitamin B« may not specifi¬ 
cally participate in the transformation of tryptophan to nicotinic 
acid. It seems more probable that any deficiency in essential food 
elements may interfere with growth promotion and with the trans¬ 
formation of tryptophan to nicotinic acid. 

Summary 

1. Tryptophan given as a delayed supplement to niacin-deficient 
rats increases growth and augments urinary excretion of nicotinic 
acid, as well as when fed simultaneously with the deficient diet. 

2. Omission of cod liver oil from the diet interferes with the growth- 
promoting effect of supplementary tryptophan and inhibits the trans¬ 
formation of this amino acid to niacin. 

References 

1. Krehi., W. A., Tepi.y, L. J., Sarma, P. S., and Elvehjem, C. A., Science 101, 489 

(1945). 

2. Sabett, H. P., and Goldsmith, G. A., J . Biol. Chem. 177, 463 (1949). 

3. Geigeb, E., J. Nutrition 34, 97 (1947). 

4. Schaeffer, A. J., and Geiger, E., Ptoc. Soc. Exptl. Biol. Med. 66, 309 (1947). 

5. Junqueira, P. B., and Schweioert, B. S., J. Biol. Chem. 162, 403 (1946). 

6. Melnick, D., Cereal Chemist 19, 553 (1942). 

7. Krekl, W. A., Sarma, P. S., and Ei.veiijem, C. A., J. Biol. Chem. 162,403 (1946). 



Letters to the Editors 


Increase in Yeast Respiration in Presence of Several 
Steroids and Diethylstilbestrol 

While studying the effects of steroids on enzyme systems of yeast, it 
was observed that several steroids and diethylstilbestrol produced a 
marked increase in the endogenous respiration of yeast. When respira¬ 
tion was measured in the presence of glucose, on the other hand, a di¬ 
minished respiration was observed following addition of several of the 
compounds. 

It is seen from the tables that the inhibitory effect of diethylstil¬ 
bestrol, desoxycorticosterone, and testosterone on respiration with 
added glucose is in inverse order to their effect on acceleration of the 
endogenous respiration of yeast. Inhibition of respiration by rat brain 


TABLE I 

Accelerative Effects on Endogenous Respiration 11 


Substance added 6 


Oxygen u 

iptake, n\. 


30 min, 

00 min. 

90 min. 

120 min. 

Water (control) 

14 

24 

34 

41 

Diethylstilbestrol 

114 

274 

345 

384 

Cholesterol 

62 

87 

103 

114 

Desoxycorticosterone 

34 

63 

87 

106 

Testosterone 

18 

35 ' 

52 

68 

Progesterone 

14 

26 

38 

48 • 


° Four experiments in duplicate on yeast from two sources were carried out. The 
table shows typical average results from one experiment. 

6 Warburg vessels contained 7.7 mg. (dry weight) of washed bakers' yeast and phos¬ 
phate buffer (0.066 Af), pH 6.8. Equilibration was carried out for 2 hr. before tipping 
in the substances from side arms, in order to reduce endogenous metabolism to a low, 
relatively constant level. Two mg. of the steroids and of stilbestrol were added as finely 
ground suspensions in water at zero time. Gas phase, air; temperature, 37°C.; 10% 
KOI I in the central well. 


321 




322 


LETTERS TO THE EDITORS 


homogenates in the presence of several steroids lias recently been re¬ 
ported (1), while Verzar (2) has demonstrated that desoxycorticos- 
terone accelerates glycogenolysis and inhibits glycogen synthesis in rat 
diaphragm. 

It would appear, therefore, that the steroids and diethylstilbestrol 
diminish yeast respiration with added glucose by inhibiting glucose 
utilization. When no exogenous substrate is added, yeast respiration is 
probably due primarily to utilization of endogenous glycogen, and the 
rate of respiration is probably limited by the rate of glycogenolysis. 
Since several steroids and diethylstilbestrol markedly increased the 
endogenous respiration of yeast, it is suggested that these substances 
produce their effect by accelerating glycogenolysis. 


TABLE 11 

Inhibitory Effects on Respiration with Added Glucose a 


Substance added* 

1 Oxygen uptake, /il. 

Per cent 
inhibition 

Per cent inhi¬ 
bition rat brail 
homogenates 
(Data of 
Gordan (1)) 

30 min. 

60 min. 

90 min. 

Water (control) 

50 

101 

148 

— 

- 

DiethyLstilbestrol 

27 

38 

44 

70 

91 

Cholesterol 

54 

105 

154 

1 

14 

Desoxycorticosterone 

33 

66 

98 

34 1 

87 

Testosterone 

1 41 

84 

131 

11 1 

24 

Progesterone 

! 47 

j 98 ! 

i j 

146 

j 1 ! 

1 

28 


a Two experiments in duplicate were carried out. The table shows typical average 
results from one experiment. 

6 Warburg vessels contained 0.77 mg. (dry weight) of washed yeast, phosphate 
buffer (0.066 M) pH 6.8, and 0.2% glucose. Respiration was measured for 1 hr. before 
the substances (2 mg.) were tipped in from side arms. Data are uptakes following 
tipping in of side arm contents. Conditions otherwise as in Table I. 

References 

1. Gordan, G. S., and Elliott, H. W., Endocrinology 41, 517 (1947). 

2 . VerzAr, F., and Wenner, V., Riochem. J. 42, 35 (1948). 

Laboratory of Experimental Oncology , Bernard Shactkh 

National Cancer Institule } 

United States Public Health Service , 

and the University of California Medical School , 

San Francisco , California 
Received May 16 , 1949 



LETTERS TO THE EDITORS 


323 


Inhibition of Ergostanyl Acetate by 
7-Dehydrocholesteryl Bromide 

The activity of ergostanyl acetate in curing the stiffness syndrome in 
guinea pigs has been reported from this laboratory (1) and by Petering 
and coworkers (2). 

We wish to report here the results of some tests made on various 
combinations of sterols in a search for synergisms and antagonisms in 
this class of compounds, as measured by the guinea pig assay for the 
anti-stiffness factor. 

Weanling pigs of the Hartley strain were placed on the diet previously 
described (1), and allowed to deplete for approximately a week, when 
most of the animals developed a 3 + stiffness. A curative test of 5-7 
days was then employed. 


TABLE I 


Antagonism of Ergostanyl Acetate by 7-Dehydrocholesteryl Bromide 


Supplement 

Test level 

No. of 
assays 

Total no. 
animals 

Animals 

showing 

curative 

response 

1 None 


5 

38 

per cent 

2 

2 Ergostanyl acetate 

10 7 

5 

38 

63 

3 Ergostanyl acetate 

200 y 

1 

8 

69 

4 As 2+7-dehydrocholesteryl bromide 

6 mg. 

3 

22 

0 

5 As 2-f-7-dehydrocholesteryl bromide 

3 mg. 

2 

15 

0 

6 As 2+7-dehydrocholesteryl bromide 

1 mg. 

1 

7 

36 

7 As 2+7-dehydrocholesteryl bromide 

0.6 mg. 

1 

8 

31 

8 As 2+7-dehydrocholesteryl bromide 

6 mg. 

1 

7 

64 

9 As 3+7-dehydrocholesteryl bromide 

3 mg. 

1 

7 

64 

10 As 2+7-dehydrocholesterol 

6 mg. 

1 

8 

50 


All of the supplements were dissolved in butyl succinate and injected 
intramuscularly each day. The 7-dehydrocholesteryl bromide was 
prepared by the method of Bernstein et al. (3) and the ergostanyl ace¬ 
tate by standard procedures (4). 

The results of 5 separate experiments are combined in Table I. The 
expected responses were obtained from 10y daily of ergostanyl acetate. 
Of a number of compounds tested, 7-dehydrocholesteryl bromide was 
the most active as an inhibitor. Levels of 3 mg. or more per day of this 








324 


LETTERS TO THE EDITORS 


compound completely inhibited the response of IO 7 of ergostanyl 
acetate. This inhibition faded out at levels below 3 mg. per day. The 
antagonism was completely reversed by 200 7 daily of ergostanyl 
acetate, indicating ready reversibility. 

Animals receiving this inhibitor became more severely deficient by 
the end of the test period than those in the negative control group. 
However, this effect with other diets has not been tested. Thus 7- 
dehydrocholesteryl bromide appears to be a reversible antagonist of the 
anti-stiffness factor. The existence of a naturally occuring antagonist 
has been reported by van Wagtendonk and Wulzen (5). 

References 

1 . Oleson, J. J., Van Donk, E. C., Bernstein, S., Dorfm\n, L., and SubbaRow, Y. 

J. Biol. Chem. 171, 1 (1947). 

2. Petering, H. G., Stubberfield, L., and Delor, R. A., Arch. Biochem . 18, 487 

(1948). 

3. Bernstein, S., Sax, K. J., and SubbaRow, Y., J. Org. Chem. 13, 837 (1948). 

4. We are indebted to Dr. S. Bernstein and Mr. K. J. Sax for the compounds used. 

5. van Wagtendonk, W. J., and Wulzen, R., J. Biol. Chem. 164, 597 (1946) 

Lederle Laboratories Division , J. J. Oleson 

American Cyanamid Co., J. C. Van Meter 

Pearl River , N. Y. 

Received May 20, 

Effect of Vitamin B i2 , Animal Protein Factor and 
Soil for Pig Growth 

Vitamin Bu, 1 an animal protein factor supplement (Lederle), 2 and soil 3 were added 
to a control ration of ground yellow corn 57, peanut meal 41.5, bone meal 0.5, lime¬ 
stone 0.5, salt 0.5%, plus vitamins A and D at levels of 5000 I. U. and 700 I. U„ 
respectively, per pig daily. The other vitamins were added at the following levels/100 
lbs. of feed: thiamine 1 g., riboflavin 230 mg., niacin 2.33 g., pantothenic acid 1 g., 
pyridoxine 375 mg., choline 19.4 g., and folic acid 22.7 mg. The control ration con¬ 
tained all the known vitamins which the pig has been shown to need (1). The first 
trial lasted 6 weeks and the second trial for 5 weeks. 

1 Vitamin B 12 concentrate (charcoal) from Dr. D. F. Green, Merck <fc Company, 
Rahway, New Jersey. (Contained 2 mg. of B 12 activity /lb.) 

2 Animal protein factor supplement (fermentation product) (N195 and N199B) 
from Dr. T. H. Jukes, Lederle Laboratories, Pearl River, New York. 

3 Top 3 inches of soil, strained of vegetable matter and dried at room temperature. 



LETTERS TO THE EDITORS 


325 


The addition of 5% soil to the basal ration was beneficial, indicating 
that soil was supplying some unknown factor(s). This is in agreement 
with previous work (2), when 5% soil added to a purified ration con¬ 
taining all known vitamins needed by the pig also supplied an unknown 
factor (s). 

In Exp. 1 the addition of the animal protein factor supplement to the 
basal ration increased the rate of gain 26%. In Expt. 2, where much 
smaller pigs were used, the animal protein factor supplement resulted 
in approximately 2.5 times the gain of that obtained on the basal ration. 

’In Expts. 1 and 2, the addition of vitamin B t2 concentrate to the 
basal ration was of no apparent benefit. In the third week of the second 
experiment, the level of Bi 2 concentrate was doubled. The pigs de¬ 
creased in rate of gain when the level of B i2 was increased, thus show¬ 
ing no benefit for increased levels of Bn. 


TABLE I 


I'Jxpt. 

no. 

No. pigs 

Aw starting 
weight 

Ration fed 

\v flail v 
gain 

Hh a\. 



lbs. 


Ur 


1 

4 

33.3 

Basal 

1.14 

13.7 


4 

32.5 

Basal + 11% animal protein factor 

1.44 

15.2 

' 



supplement 




4 

33.0 

Basal-fO.2% vitamin Bn concentrate 

1.21 

13.3 

1 

4 

33.3 

Basal+5% soil 

1.34 

12.1 

2 

f> 

18.3 

Basal 

0.20 



(i 

18.1 

Basal-1-2.2% animal protein factor 






supplement 

0.73 



5 

18.5 

Baaal-f- 0.4% vitamin Bn concentrate 

0.25 



Catron and Culbertson (3) found that B t2 was of benefit when added 
to a corn soybean oil meal, alfalfa, mineral ration to which vitamins A 
and I) were added. Why the results obtained in t his experiment are 
different, it is difficult to state. It is possible that, since we used peanut 
meal instead of soybean oil meal, it may make a difference. In addition, 
in these experiments, alfalfa meal was not included, whereas all B- 
eomplex vitamins which the pig has been shown to need (1) were added 
to the control ration, thus, possibly, accounting for the different results. 
It is possible that the B 12 charcoal concentrate used did not have any 
activity, although this in not likely to be the case. It is also possible 
that vitamin Bi 2 is only one of the factors in the animal protein factor 




326 


LETTERS TO THE EDITORS 


supplement and that the other factor(s) must be present before vitamin 
B 12 will be of benefit. 

The addition of the APF supplement seemed to stimulate hemoglo¬ 
bin formation; however, data are needed with more pigs before this can 
definitely be stated. 

These data show that the addition of the animal protein factor 
supplement was of considerable benefit, especially with lighter hogs, 
when added to the basal ration, whereas the vitamin B i2 concentrate 
was of no apparent help. These data show that, under the conditions of 
this experiment, the animal protein factor supplement of Lederle 
Laboratories and the vitamin Bi 2 charcoal concentrate were different 
in their response, and that the animal protein factor supplement 
supplies an unknown factor or factors for the young pig fed a corn- 
peanut meal ration. Discussion is given as to possible reasons why the 
B 12 charcoal concentrate was of no benefit. 


References 


1. Cunha, T. J., Colby, R. W., Bustad, L. K., and Bone, J. F., J. Nutrition 36,215 

(1948). 

2. Cunha, T. J., Colby, R. W., IIodgskiss, H. W., Huang, T. C., and Ensmingf.r, 

M. E., J. Animal Sri. 7, 523 (1948). 

3. Catron, D., and Culbertson, C. C., Iowa Farm Sri., 3, 3 (1949). 


Department of Animal Industry, 
Agricultural Experiment Station, 
University of Florida, 

Gainesville, Fla. 

Rerieved May 24,1949 


T. J. Cunha 
J. E. Burnside 
D. M. Buschman 
R. S. Glasscock 
A. M. Pearson 
A. L. Shealy 


On the Probable Identity of Several 
Unidentified Growth Factors 1 


In 1946, Guirard, Snell and Williams (1) described an assay method 
for a naturally-occurring, water-soluble substance which duplicated the 
growth-promoting effect of acetate for Lactobacillus casei. Preparations 
concentrated approximately 40-fold from yeast extract were 400 times 
as active, on the weight basis, as acetate in promoting growth. On 

1 Supported in part by grants from the Schenley Research Foundation and from 
Parke, Davis and Co. We are indebted to Dr. I. C. Gunsalus and to Drs. T. H. Jukes 
and E. L. R. Stokstad for samples of the concentrates referred to in the text. 



LETTERS TO THE EDITORS 


327 


these grounds it appears logical to ascribe a catalytic role in the pro¬ 
duction of acetate to the active factor in such concentrates. 

These considerations, and certain similarities in stability and dis¬ 
tribution, led us to test concentrates (a) of an unidentified factor 
described by O’Kane and Gunsalus (2) that is required by resting cells 
of Streptococcus faecalis for oxidation of pyruvic acid, and (b) of 
“protogen,” an essential growth-factor for Telrahymena geleii de¬ 
scribed by Stokstad el al. (3). Highly purified preparations of each of 
these substances wore extremely active in promoting growth of Laclo- 


TABLE I 

Comparative Activities of Acetate, Pyruvate Oxidation Factor, and Protogen in 
Promoting Growth of Lactobacillus casei 


Sodium acetate 

Yea.st extract 

Pyruvate oxidation 
factor* 

Protogen e 

7/10 ml. 

Turbidity® 

7 MO ml. 

Turbidity® 

my /10 ml. 

Turbiditj* 

iny/10 ml. 

Turbidity" 

0 

94 

0 

94 

0 

94 

0 

94 

100 

70 

30 

04 

1 

88 

0.3 

39 

300 

02 

100 

53 

3 

70 

1.0 

07 

500 

50 

300 

47 

10 

06 

3.0 

00 

1,000 

50 

1,000 

42 

30 

59 

10.0 

45 

3,000 

40 

10,000 

32 

100 

49 

30.0 

43 

10,000 

37 


1 

300 

40 

i_ 



a Per cent of incident light transmitted, uninoculated medium = 100. Incubated 
30 hr. The medium and procedure were those described by Guirard, Snell and Williams 
0 ). 

6 From Dr. I. G. Gunsalus; purified approximately 3000 times over yeast extract. 
r From Dr. T. Ii. Jukes; purified approximately 10,000 times over a standard liver 
preparation. 

bacillus casei in the absence of acetate (Table I) under previously 
described conditions (1). Concentrates of the pyruvate oxidation 
factor with relative approximate purities of 1, 70, and 3000 units/ mg. 
bv enzymatic assay (2) had relative activities of approximately 1, 90, 
and 3300 for L. casei. A preparation of “protogen” purified 10,000 times 
over a standard liver preparation by Tetrahymcna assay, proved ex¬ 
tremely active for L. casei ; 3 my of this preparation sufficed to dupli¬ 
cate the growth-promoting activity of 400 y of sodium'acetate. The 
amounts of the latter concentrate required by L. casei for growth are 





328 


LETTERS TO THE EDITORS 


quantitatively similar to the amounts of pyridoxal or folic acid re¬ 
quired for growth of this organism. 

From these results it appears highly probable that protogen, the 
pyruvate oxidation factor, and the acetate-replacing factor for L. casei 
are identical. It has been indicated (3) that more than a single form of 
this factor occurs naturally. Unpublished data (4) show that this 
factor is also one of the substances which promotes rapid growth of 
Streptococcus faecalis from small inocula in acetate-free media (e.g., 
5,6). In confirmation of previous data (1,7), and in marked contrast to 
their effect on L. casei, concentrates of pyruvate oxidation factor and of 
protogen were relatively ineffective in replacing acetate for Lacto¬ 
bacillus arabinosus. 


References 

1. Gitrard, B. M., Snell, E. K., and Williams, R. J., Arch. U we hem. 9, 381 (1940). 

2. O'Kane, D. J., and (!t nsali s, J. C., J . Bart. 56, 499 (1948). 

3. Stokstad, E. L. R., Hoffman, C. E., Regan, M. A., Fordham, I)., and Ji.kes, 

T. H., Arch. Biochrm. 20, 75 (1949). 

4. McNutt, W. S., and Snell, K. E., unpublished data. 

5. Cooperman, J. M., Ruegamer, W. R., Snell, E. E., and Elvehjem, C. A., •/. 

Biol. Chem. 163, 769 (1946). 

6. Colio, L. G., and Babb, V., ibid. 174, 405 (1948). 

7. Guirard, B. M., Snell, E. E., and Williams. R. J., Arch. Biochrm. 9, 361 (1946). 

Department of Biochemistry, Esmond E. Sneli, 

University of Wisconsin, Harry P. Broquist 

Maclison, H7s. 

Received May 2o. 194!) 


Effect of APF Supplement on Pigs Fed 
Different Protein Supplements 

Cunha et al. (1) showed that an APF supplement (Lederle Laboratories) was of 
considerable benefit and supplied an unknown factor(s) when added to a corn-peanut 
meal ration (same ration as fed in Lot 1), whereas a B 12 charcoal concentrate (Merck 
& Co.) was of no benefit when added to the same ration. In this experiment, peanut 
meal, soybean oil meal and fish meal made up 41.5, 44.5, and 30%, respectively, of the 
ration. All rations contained the same amount of crude protein which was 22.78 
(dr 0.05%). The remainder of the ration in each lot consisted of corn, bone meal 0.5, 
limestone 0.5 and salt 0.5%. Pigs in all lots received vitamins A and D at levels of 
5000 I. U. and 700 I. U., respectively, daily. The other vitamins were added in all 
rations at the following levels/100 lbs. of feed: thiamine 1 g., riboflavin 230 mg., 
niacin 2.33 g., pantothenic acid 1 g., pyridoxine 375 mg., choline 19.4 g., and folic acid 



LETTERS TO THE EDITORS 


329 


22.7 mg. Thus, all the known vitamins which the pig has been shown to need (2) were 
added to all the rations fed. The trial lasted for 37 days. Five pigs were used in each 
lot. 

The addition of the APF supplement to the corn-peanut, meal ration 
resulted in 2.13 times the rate of gain obtained on the control ration. 
This is in agreement with previous work by Cunha el al. (1), where the 
addition of the APF supplement, at the same level, to the same corn- 
peanut meal ration (Lot 1) increased the rate of gain approximately 
2.5 times. 


TABLE 1 


Lot no. 

Av. .starting 
weight 

Protoin supplement fed 

Av. daily gain 


lb 


lb. 

1 

27.1 

Peanut meal 

0.62 

2 

27.1 

Peanut meal + APF supplement" 

1.40 

3 

26.4 

Soybean oil meal 

1.01 

4 

27.0 

Soybean oil meal + APF supplement" 

1.31 

f> 

27.9 

Fish meal 6 

1.29 

6 

27.8 

Fish meal + APF supplement" 

1.45 


“ Animal Protein Factor supplement (fermentation product) obtained from Dr. 
T. II. Jukes, Lederle Laboratories, Pearl River, New York. 

* Fish meal, 60.2% protein, obtained from Ralston Purina Co., St. Louis, Missouri, 
and which satisfied their standards for being of very high quality. 

The addition of the APF supplement to the corn-soybean oil meal 
ration increased the rate of gain approximately 30%. These data show 
that the APF supplement is much more beneficial when added to a corn 
ration containing peanut meal than when it contains soybean oil meal. 

The addition of the APF supplement to the corn-fish meal ration 
was beneficial to a small extent. This may mean that fish meal, or 
particularly the fish meal used, is not quite high enough in the factor 
or factors which the APF supplement supplies. 

Of much interest is the finding that the addition of the APF supple¬ 
ment to a corn-peanut, meal ration and to a corn-soybean oil meal ration 
resulted in gains being obtained similar to those on a corn-fish meal 
ration (Lot 5). This shows that the APF supplement supplied an 
unknown factor or factors which caused peanut meal and soybean oil 
meal to come up to the fish meal used in feeding value for'the pig. 
When the APF supplement was added to the peanut meal ration 




330 


LETTERS TO THE EDITORS 


(Lot 2) the results obtained were about as good as when it was added 
to fish meal ration (Lot 6). 

These data indicate that the APF supplement supplies an unknown 
factor or factors for the pig, and that it increased the feeding value of 
peanut meal and soybean oil meal so that these plant protein supple¬ 
ments were similar in feeding value to the fish meal used. 


References 


1. Cunha, T. J., Burnside, J. E., Buschman, D. M., Glasscock, R. S., Pearson, 

A. M., and Shealy, A. L., unpublished data (1949). 

2. Cunha, T, J., Colby, R. W., Bustad, L. K., and Bone, J. F., J. Nutrition 36, 215 

(1948). 


Department of Animal Industry , 
Agricultural Experiment Station , 
University of Florida, 

Gainesville, Fla. 

Received June 20, 1940 


J. K. Burnside 
T. J. Cunha 
A. M. Pearson 
R. S. Glasscock 
A. L. Siiealy 


Does Light Inhibit the Respiration of Green Cells? 

It is well known that the oxygen consumption of green cells may be de¬ 
creased by illumination. The light intensity at which the oxygen exchange 
of a given culture becomes zero may be termed “compensating.” Above 
this light intensity positive evolution of oxygen gas is observed. The 
mechanism of compensation by light has until now remained uncertain, 
but the simplest explanation is obviously the production of oxygen gas by 
photosynthesis; that is, since light clearly produces oxygen gas above 
compensation it is reasonable to suppose that it also does so below 
compensation. However, the idea, old as the science of photosynthesis, 
still persists that light inhibits respiration per se, either anticatalyti- 
cally (as by inactivating respiratory enzymes), or by reducing inter¬ 
mediates of respiration. If this idea were true, most computations of 
photosynthetic efficiency would be invalidated because they have been 
carried out below the compensation point. 

A decision on this much-discussed problem has been obtained from 
the following type of experimentation, performed many times. In the 
main compartment of a rectangular vessel, with two side-arms con¬ 
taining alkali, was placed a suspension of Chlorella pyrenoidosa cells in 
acid culture medium (pH 4.8), with air as gas phase. The vessel, 



LETTERS TO THE EDITORS 331 

attached to a manometer, was rapidly shaken and alternately darkened 
and illuminated from below with a beam of completely absorbed red 
light of the same intensity (M).25 microeinsteins/min./3 cm. 2 area) as 
employed in simultaneous quantum efficiency determinations on 
aliquot suspensions by the 2-vessel method with 5% C0 2 in air as gas 
phase. In the vessel with low C0 2 pressure, negligible light action was 
observed, the oxygen consumption in the dark and in the light being 
practically identical, whereas, with aliquots of the same suspension 
under otherwise identical conditions except for adequate C0 2 pressure, 
high efficiencies of 3 to 5 quanta absorbed per molecule of O 2 produced 
were observed both below and well above the compensation point, with 
no change in dark respiration at the widely different C0 2 pressures 
involved (see example). 

This absence of light action on respiration at low C0 2 pressures may 
appear to contradict the experience of other investigators who, since 
the use of manometry in photosynthesis, have observed, in vessels con¬ 
taining alkali in side-arms or middle compartments, that the respira¬ 
tion of Chlorella could be compensated by light. The explanation for 
this apparent discrepancy is that such experiments were carried out 
with too high light intensities that were not controlled quantita¬ 
tively by means of simultaneous efficiency determinations. Light 
and alkali compete for the small amount of C0 2 formed in respiration, 
so that, for every CO 2 pressure, however low, a light intensity exists 
that will compensate respiration, as we have confirmed when we have 
used light of sufficiently high intensity. 

Our experiments show conclusively that red light does not inhibit 
the respiration per se when light intensities are employed that yield 
high photosynthetic efficiencies. When light compensates respiration it 
does so by the independent process of photosynthesis, the gas exchange 
of which happens to be the opposite of that of respiration. This result 
has been confirmed in a different way by determination and comparison 
of photosynthetic efficiency below and above the compensation point. 
Under the conditions of our experiments, reported in detail elsewhere 1 , 
the same high quantum efficiencies of 3-5 quanta per O 2 are obtained 
up to intensities at least five times the compensating intensity. That is, 
one molecule of oxygen developed above the compensation point, or 

1 In press, Science, and Biochim. Biophys, Acta (Meyerhof Festschrift, October, 
1949); and report delivered at meeting of Society of General Physiologists, Woods 
Hole, Mass., June 22, 1949. 



332 


LETTERS TO THE EDITORS 


one less molecule of oxygen consumed below the compensation point, 
as the result of light action, represents the same gain in chemical 
energy. All theories of light action should be in harmony with this now- 
established thermodynamic fact. 


Example 

Each of three vessels contained 230 mm . 3 aliquots of Chlorella pyrenoiclosa cells 
suspended in 7 cc. of culture medium (5 g. MgS 04 - 7 H 2 0 , 2.5 g. KNO3, 2.5 g. KH4PO4, 
2 g. NaCl, and 5 mg. FeS 04 ‘ 7 H 2 0 in 1 1. of filtered, unsterilized well water at pH 
4.5-5). Temperature, 20 °C. Horizontal shaking at the rate of 150 cycles/min. at 2 
cm. amplitude. Total intensity of red light beam (630-660 mu), 0.254 microeinsteins/ 
min., equivalent actinometrically to 5.7 mm . 3 Oa/min. 

No. I. 0.2 cc. N NaOH in each side-arm, gas phase air. 

(Vessel volume 18.87 cc., liquid volume 7.40 cc., ko* 1.09) 

20' dark —40 mm. xo 2 = +43.5 mm . 3 

20' red light —39 mm. 

20' red light action —1 mm. xo 2 = +1.09 mm . 3 


Xo. II. Gas phase 5% C0 2 in air, respiration not compensated by white light. 



Vessel 3 

Vessel 5 



Total volume 

* 17.99 

13.91 



Liquid volume 

7.00 

7.00 



k(h 

1.046 

0.665 



kcoi 

1.634 

1.253 



20'‘dark 

—13 mm. 

— 27.5 mm. xo* = 

= —40.6 mm. 3 


20' red light 

— 2 mm. 

— 7.0 mm. 



20' red light action 

+11 mm. 

+20.5 mm./ X ° 2 

= +23.8 mm. 3 ' 
= — 19 mm. 3 . 

r o 2 





= - 0.8 


hv 

0' 2 

20 X 5.7 

23.8 0 



No. III. Gas phase 5% C() 2 in air, respiration overcompensated several-fold by white 

light. 

Vessel 3 

Vessel 0 



20' white light 

+39 mm. 

+57.5 mm. 



20' white + red light 

+48 mm. 

+76 mm . 



20' red light action 

+ 9 mm. 

+ 18.5 mml 10 ' 
Uoo, 

= —26 mm. 3 ) 
= -26 mm. 3 J 

co 2 
= of 


= - 1.0 



LETTERS TO THE EDITORS 




It will L)o observed that the oxygen consumption in the dark in Nos. I and II is 
essentially the same (about 42 mm. 3 /20'). This means that the CO* pressures 
used did not influence respiration, and that the 00 2 pressure required to obtain maxi¬ 
mum respiration is below that required for effective photosynthesis, where C0 2 is 
required as substrate and not merely catalytically. 


National Cancer Institute, 
National Institutes of Health , 
Bcthesda, Md., 

Plant Industry Station, 
BcUsvillc, Md., and 
Marine Biological Laboratory, 
Woods Hole, Mass. 

Received August S, iO/ft 


Otto Warburo 
Dean Burk 
Victor Schocken 
Mitchell Korzenovsky 
Sterling B. Hendricks 


Myosmine in Cigar Tobacco 

In a previous publication (I), an analytical procedure was described, 
w hich permits the separation of the alkaloids present in tobacco leaves 
from the alkaloid transformation products (1,2) which accumulate 
during the fermentation of the leaves. Fraction A of this analytical 
procedure, obtained by extraction of powdered tobacco in presence of 
MgO with a suitable hydrocarbon solvent, contains exclusively the 
alkaloids proper, free of any degradation products. 

Detailed analyses of Fractions A of the same and different samples 
obtained from unfermented tobacco leaves have proved the presence of 
considerable amounts of nicotine and smaller amounts (7-10% of the 
nicotine) of nornicotine. 1 Invariably, the sum of these two alkaloids is 
slightly lower than the total alkaloids determined in separate sub¬ 
fractions. This indicates that other alkaloids are present in amounts of 
roughly 0.5-2% of the nicotine. 1 Various qualitative observations made 
it appear likely that these additional alkaloids are dehydrogenation 
products of nicotine and of nornicotine containing double bonds in their 
pyrrolidine rings. Since dehydrogenation products may represent inter¬ 
mediates between the alkaloids proper and their oxidation products 
which have been found in fermented leaves, attempts were made to 
identify these compounds. 

According to Haines, Eisner and Woodward (3), myosmine (2-(3- 
pyridyl)-A 2 -pyrroline) (4), in aqueous solutions, yields the correct 

‘These percentages, based on nicotine, are valid for unferniented tobacco only. 
The variable decrease of nicotine on fermentation prevents our expressing a similar 
ratio for fermented leaves. 



334 


LETTERS TO THE EDITORS 


value for the one primary amino group of its hydrolyzed form (with 
open pyrroline ring) by the van Slyke procedure. Model experiments 
proved 2 that this reaction can be employed with good results for the 
quantitative determination of amounts as low as 200 y of myosmine 
(containing 20 y of amino nitrogen), in the presence of very much 
larger amounts of nicotine, nornicotine and other alkaloids. Tested by 
this method, the following values were obtained for the myosmine con¬ 
tent of various tobacco samples. 

TABLE I 


Alkaloid Contents of Various Tobacco Samples Expressed as Alkaloid Nitrogen* 
in Per Cent of Dry Weights of the Samples 


No. 

, 

Type of tobacco sample 

Nicotine 

nitrogen 

Nornicotine 

nitrogen 

Myosmine nitrogen by 
van Slyke method 

Total alkaloid 
nitrogen other 
than nicotine * 
and nornicotine* 

1 

Pa. Seedleaf #5 
unfermented 

0.701o 

0.046ft 

0.011,±.004, [14]* 

0.007-0.022 

2 

Pa. Seedleaf #5 
fermented 

0.310 6 

! 

0.030 2 

0.012,±.006o [10]* 

0.006-0.020 

3 

Pa. Seedleaf # 12 
unfermented 

0.712 

0.050 

0.005,±.000, [2]* 

- ■ 

4 

Pa. Seedleaf #12 
fermented 

0.246 

0.024 

0.007,±.000, [2]* 

— 

5 

Pa. Seedleaf #24 
unfermented 

0.796 

0.047 

0.015„±.003o [3]* 



° The values in terms of alkaloid nitrogen rather than of alkaloids permit an 
immediate nitrogen balance. TKe alkaloid values can be obtained by multiplying the 
values listed by the factor 5.8 for nicotine, 5.3 for nornicotine, and 5.2 for myosmine. 

6 The figures in brackets indicate the numbers of independent analyses made for 
each sample. 

c Being a small difference between two large analytical values, this kind of deter¬ 
mination is subject to a wide margin of error. 

Some further samples of cigar leaf tobacco types, other than Penn¬ 
sylvania Scedleaf, yielded similar values for myosmine nitrogen. Con¬ 
trary to nicotine and nornicotine, the myosmine does not seem to 
decrease during fermentation. This may indicate its function as an inter¬ 
mediate. In spite of the fairly good reproducibility of these analytical 

* Many of the pure alkaloids used for these model experiments were kindly sup¬ 
plied by Drs. Woodward, Eisner and Haines of the Eastern Regional Research 
Laboratory, Philadelphia 18, Pa. 



LETTERS TO THE EDITORS 


335 


results, a fully satisfactory identification of myosmine still has to be 
achieved.’ Thus far, this compound was only identified in tobacco 
smoke (5) and as a product of the high temperature catalytic decompo¬ 
sition of nicotine (6). 

Ultraviolet absorption spectra measured for various samples of 
Fraction A show deviation from the spectra (7) calculated for their 
nicotine and nornicotine contents. The shift of the spectra caused by 
t hese deviations may be interpreted as being caused by the presence of 
small amounts of myosmine and possibly further unsaturated alkaloids, 
but the presence of additional ultraviolet absorbing substances prevents 
a reliable quantitative evaluation. 

Hydrogenation of the substances contained in Fraction A has yielded, 
in some tentative experiments, small increases of distillable nicotine. 
This effect would indicate the presence of N-methylmyosmine (8) 
which, on hydrogenation, yields nicotine. Myosmine, if hydrogenated, 
yields nornicotine which does not distil under the conditions employed 
in our analysis. Metanicotine would yield a similar hydrogenation 
effect as would myosmine. The order of magnitude of these additional 
compounds appears to be similar to that of myosmine. 

References 

1. Fhankenburg, W. G., Science 107, 427 (1948). 

2. Frankenburg, W. G., and Gottscho, A. M., Arch. Biochem. 21, 247 (1949). 

3. Haines, P. G., Eisner, A., and Woodard, C. F., J. Am. Chem. Soc. 67,1258 (1945). 

4. Statu, E., Wencsch, A., and Zajic, E., Ber. 69, 393 (1936); SpXtii, E., and 

Mamoli, L., ibid. 69, 757 (1936). 

5. Wencsch, A., and Schoki.ler, R., Fnch. Mitllg. Oesterreich. Tnhakregie 2, 15 

(1933). 

6 . Woodward, G. F., Eisner, A., and Haines, P. G., J. Am. Chem. Soc. 66, 911 

(1944). 

7. Swain, M. L., Eisner, A., Woodward, C. F., and Brice, B. A., ibid. 71, 1341 

(1940). 

8. Spath, E., Wibaut, J. P„ and Kesztler, P., Ber. 71, 100 (1938). 

Research Laboratory, W. G. Frankenburo 

General Cigar Company, Inc., A. M. Gottscho 

60S N. Charlotte St., 

Ijincaster, Pa. 

Received August 9,1949 


3 Such an identification appears desirable because other unsaturated alkaloids re¬ 
lated to myosmine may also respond to the van Slyke reaction. 




Book Reviews 


Nature of Life: A Study on Muscle. By A. Szent-Gyoruyi, Department of Bio¬ 
chemistry, The University, Budapest. Academic Press Inc., New York, N. Y. 1948. 
91 pp. and 7 plates. Price $3.00. 

Prof. Szent-Gyorgyi has that rare gift, which we remember in Gowland Hopkins, of 
being able to make a singular contribution to any field of biochemistry in which ho 
choses to work. It is not merely that he makes a noteworthy contribution: many have 
that gift. But few have the talent to produce views and facts which demand a com¬ 
plete reorganization of a field. In his latest book Szent-Gyorgvi demands a complete* 
reorganization of our views on the biophysics and biochemistry of muscular contraction. 
It is the subtitle and not the title which describes the contents. On the nature of life he 
has little to sav except that since “life is characterised by self-reproduction, one rabbit 
could not be alive at all -only two rabbits are one rabbit.” One day perhaps the author 
will write us another book which will be two rabbits. But, for the moment, a book 
which is one rabbit will do very well. 

Muscle differs from most other celljtypes in that it has been the object of much de¬ 
tailed study at many different levels of organization. Emphasis is laid upon the dis¬ 
tinctive attributes of these' different levels, and upon the desirability of being alert to 
the emergence of properties at the higher levels of organization which would not 
readily be predicted from knowledge of the lower levels. 

Falling into this category is Szent-Gyorgyi's suggestion that in a protein molecule 
there may be a tendency towards non-localization of electrons, which may occupy 
levels characteristic of the molecule as a whole. To the extent to which this is true 
there will be resonance energy, which is likely to be lost by changes in the configura¬ 
tion of the molecule. It is possible that in organized systems of proteins, electrons may 
even occupy levels which are characteristic of the organization. It may be that the 
characteristic properties of muscle are bound up with the properties of such systems. 

No attempt is made to account for all the cytological characteristics of muscle 
fibrillae. The feature which is presented is of a structure based on two proteins, actin 
and myosin. Actin forms long threads composed of strings of globular molecules: myo¬ 
sin molecules, which are rigid rods, adsorb upon the actin threads to give actomyosin. 
A gel of actomyosin, when brought into contact with adenosine triphosphate (ATP), 
contracts vigorously and reversibly. The active system contains ATPase, and when 
the ATP is destroyed the contraction may reverse. The association of contractility 
with ATP is one of the main factors which leads Szent-Gyorgyi to postulate that the 
contractile unit in muscle is actomyosin. The myosin unit is itself a protein complex, 
for it appears to consist of a protein rod, which might perhaps be called structural 
myosin, onto which is adsorbed a variety of globular proteins, including ATPase, 
ADPase, ATP deaminase, etc. Without at least some of these adsorbed proteins, 
structural myosin is unable to give an active (contractile) complex with actin. 

336 



BOOK UKVI1SWS 


337 


The picture which is presented is, on the biochemical level, uncommonly plausible. 
But cell physiologists will not be as contented with the picture as biochemists. The 
explanation of rigor which is suggested does not carry conviction, and there is one 
basic assumption in the whole story for which there is hardly a shred of evidence: this 
is that the system actomyosin-ATP is, in fact, involved in muscle contraction in the 
same way as it is with the contraction of an actomyosin gel in a test-tube. For example 
there is no evidence 4 that ATP is, in fact, involved in the actual contraction and relax¬ 
ation of muscle, rather than wit h recovery. That t his is so makes a reasonable working 
hypothesis: but Szent-Gybrgyi, in common with many other biochemists, is inclined 
to treat it as a fact. As A. V. Hill has recently emphasised, this will not do. 

But, as was said at the l>eginning of this review, this is a really valuable book, one 
which merits much thought. We can do with more such books. 

,1. Danielli, London 

Textbook of Endocrinology. By Hants Selye, Professor and Director of the Inst it ut 
de M&lecine et de Chirurgie experimentales, Universit6 de Montreal, Montreal, 
Canada. Acta Endocrinologica University de Montreal, Montreal, Canada, xxxii -f 
914 pp. Price $12.80. 

Selye has achieved his main goal of providing a textbook of endocrinology “primar¬ 
ily for the medical student and physician.” Others, even specialists, should find this 
book useful as a ready source of reference on the various aspects of endocrinology . 
The cost of the book, however, may prohibit its purchase by medical students, 
particularly since only a very short period of time is allotted to the teaching of endo¬ 
crinology by most medical schools. On the other hand the general usefulness of the 
book for both experimental and clinical information should prompt its purchase by 
many. 

The amount of space devoted to each subject seems to be judiciously chosen. The 
topics are well organized and clearly presented. Much of the material has been sub¬ 
mitted to colleagues and associates for criticism and revision. 

In the Introduction, Selye discusses the tasks undertaken by his institute “in an 
attempt to delimit and systematize the field of endocrinology” and the role of this 
book in this general program. 

The Introduction is followed by a section entitled General Endocrinology. In 37 
pages, the author presents a brief and concise description of the general nature and 
action of the endocrine organs, the classical and recognized methods of study, a brief 
history of endocrinology prior to the twentieth century, a list of journals and mono¬ 
graphs containing endocrine literature and a compilation of the commercially avail¬ 
able endocrine preparations with the names of the respective manufacturers. 

Chapter I (43 pages) is devoted to a discussion of the occurrence and role of steroids, 
I heir chemistry, chemical structure (well illustrated by spatial models), nomenclature, 
chemical structure as related to biological activity, biogenesis and metabolism of the 
steroid hormones, including a valuable tabulation of steroids isolated from natural 
sources. 

In the next nine chapters the endocrine organs are treated in the following sequence: 
Adrenals (109 pages). Unfortunately, Selye has chosen to treat the cortex and medulla 
simultaneously. Thus, continuity and clarity have suffered. The same criticism may 



338 


BOOK REVIEWS 


be made of the Pituitary (121 pages), the various lobes of which should have been 
treated separately; Ovary (155 pages); Pancreas (58 pages); Parathyroid (54 pages); 
Pineal (6 pages) Testes (76 pages); Thymus (10 pages); and Thyroid (86 pages). The 
general plan of treatment is the same for each chapter. The various subjects are con¬ 
sidered in a logical sequence: Historic Introduction; Normal Morphology (anatomy, 
histology, comparative morphology, embryology, and theories concerning the histo- 
physiology of the organ); Chemistry of the Organ; General Pharmacology of the 
Hormones (standardization, pharmacology, mode of administration, and chief indi¬ 
cations); Experimental Physiology (explantation, transplantation, technic of extir¬ 
pation, effects of extirpation, and treatment with hormones); Metabolism of the 
Hormones including content in the body fluid and tissues in health and disease; 
Stimuli Influencing Structure and Activity of the Organ; Diseases of the Organ; Hypo- 
and Hyper-Activity and Tumors. Approximately 50% of the space is allotted to clini¬ 
cal material. 

The various gastrointestinal hormones, other postulated hormones, and the hor¬ 
mones of invertebrates and plant s are presented in a brief but adequate manner in the 
19 pages of Chapter 10. 

In the eleventh and last chapter interrelationships among various hormones arfc 
discussed with excellent schematic diagrams. Most of the space is allotted to the 
female reproductive system, pregnancy and lactation (39 pages), and the adaptation 
syndrome (30 pages). The last three pages contains a General Survey of Hormonal 
Correlations with an excellent but somewhat overburdened diagrammatic sketch. 
This portion is worthy of more consideration. 

The book is profusely illustrated with not only clinical abnormalities but also 
photographs and graphs to illustrate many physiological and pathological effects of 
the hormones. Diagrammatic schemes are used frequently to clarify or correlate 
specific effects. The author has avoided the cataloging of a large number of references 
but instead has chosen to provide at the end of each chapter a number of key refer¬ 
ences in the form of reviews, monographs and occasionally an original article. An 
extensive index is included. 

C. D. Kochakian, Rochester, New York 

Physiology of Man in tne Desert. By E. f. Adolph and Associates. Interscience 
Publishers, Inc., New York, N. Y., 1947. xiii + 357 pp. Price $6.50. 

The book is a real contribution in a field of physiology in need of information, a field 
in which commonly accepted beliefs are at such variance with the facts that they are a 
hazard to all but experienced desert travelers. “Contrary to the popular legend of the 
wanderer tortured by killing thirst , when the unfortunate victim of water shortage is 
first incapacitated, he is neither delirious nor in agony from ‘mouth thirst’; he is simply 
incapable of even mild physical effort.” 

The book also laid to rest the popular beliefs that in deserts men (1) drink more 
water than they need, (2) can be trained to do with less water, and (3) that men should 
drink no water at work or on the march. So firmly established has been the notion that 
men drink more water than they need that a plan to ration desert troops to 2 quarts of 
water daily was abandoned only when the commanding general was induced to try it. 
The decision was that shortage of water was not consistent with field operations. 



BOOK REVIEWS 


339 


Man may acclimatize to heat, but he cannot adapt himself to less water. On the 
contrary, the real danger is that, as man dehydrates, he does not voluntarily drink 
enough water to replace that lost, and men have succumbed to dehydration exhaustion 
with plenty of water in their canteens. 

The writers have given a detailed picture of the signs and symptoms of dehydration. 
“Rarely is lack of body water recognized as the major cause of reduced efficiency in 
the desert; the blame is often placed on the heat, or diet, or the difficulty of the task 
set—impaired morale is one of the earliest signs of dehydration—when fully hydrated, 
man cheerfully does tasks which he finds distasteful if he is moderately dehydrated 
and consequently in low spirits.” 

The book evaluates the 3 stresses which the desert dweller must contend with* 
namely, work, heat, and dehydration. The fully hydrated man can turn out 77% of 
the work he is capable of at moderate temperatures, at 110°F., but only 37% of the 
work when dehydrated by the loss of water equal to 4.0% of his body weight. 

There are many fine physiological contributions in the book involving cardiac 
output and changes in the composition of many tissues with progressive dehydration. 
Of special interest are two physiological processes, heat regulation and urine produc¬ 
tion, for which body water is needed. Water is available for sweat formation (heat 
regulation) and urine production when the body is severely dehydrated. The body 
temperature and body composition must remain fairly constant, regardless of 
other possible physiological derangements caused by the loss of water. While most of 
the information was obtained on human subjects, animals were also used in their 
investigations. 

Thirst received considerable study and interesting species differences came to light. 
Thirst served as an accurate indicator of the water needs of the dog and he replaced 
the water he lost by drinking an almost equivalent amount. In man, thirst seemed to 
be inhibited when dehydrated more than 2.0%, so that, w r hen most needed, thirst fails 
him as an indicator of his water requirements. Before man voluntarily makes good his 
water losses, food and leisure seem to be essential. 

Maps are drawn which show 7 average amounts of water required for the maintenance 
of one man in the hottest month in the desert areas of the United States, Northern 
Africa, Asia, and Australia. There are also other useful maps and tables. One table 
shows the mileage per gallon of water that the desert traveler is capable of under 
various conditions. 

It is unfortunate that a book containing so much valuable information should be so 
poorly put together. Actually it is not a book but a collection of about 20 independent 
contributions. The result is that the book is badly repetitious. It could have been 
integrated into a book of half its present size with great improvement. Many chapters 
are excellent, some are fair and at least one, Chapter 7, is almost entirely excess 
baggage. 

The book as a whole supplies much badly needed information and is a most wel¬ 
come addition to our knowledge on the physiology of man in the Desert. 

Samuel Lepkovsky, Berkeley, Calif. 



340 


BOOK REVIEWS 


Vitamins and Hormones t Vol. VI. Edited by Robert S. Harris, Massachusetts 
Institute of Technology, and Kenneth V. Thiman, Harvard University, Cambridge, 
Massachusetts. Academic Press, Inc., New York, N. Y., 1948. xi + 435 pp. Price 
$7.80. 

Vol. VI of Vitamins and Hormones contains 8 reviews on current topics and a useful 
cumulative index for Vol. I-V. The large number of reviews which have recently 
appeared in this general field make it necessary for some of the reviewers to confine 
their efforts to specific periods of time or limited phases of the subject matter. 

The review of the chemistry and biological action of pteroylglutamic acid and re¬ 
lated compounds by Brian L. Hutchings and John H. Mowat aids one in unraveling 
this complicated subject. It is of somewhat greater usefulness in clarifying the chem¬ 
istry than the biological action of these compounds. 

The review of Vitamin K by Henrik Dam serves as a valuable addendum to previ¬ 
ous works on this subject. B. S. Schweigert has gathered together valuable practical 
information for those interested in the use of the cotton rat and hamster in nutritional 
experimentation. 

The difficulties of definition of vitamins as pharmacological agents are pointedout 
by Hans Molitor and Gladys A. Emerson. The authors succeed in bringing together a 
large number of relatively unrelated observations in an interesting manner. The com¬ 
mon practice of using relatively large doses of vitamins makes a knowledge of the 
pharmacological properties of these compounds of some importance. 

H. M. Sinclair discusses in some detail the problems of assessing nutritional status 
of the human in the absence of manifest clinical signs and symptoms. It is obvious 
from his discussion that much remains to be done before simple criteria for such 
studies are developed. Van Lanen and Turner have reviewed in some detail the 
occurrence and synthesis of vitamins in microorganisms. The subject is reviewed 
primarily from a descriptive rather than analytical point of view. 

The review on B vitamins as plant hormones by James and Harriet Bonner is an 
•excellent summary of work in this field. The correlation of activities of various com¬ 
pounds in plants, microorganisms and mammals is of great inetrest from the point; 
of view of comparative biochemistry. 

The extensive review by Edward C. Kendall of the role of the adrenal cortex on the 
metabolism of water and electrolytes is perhaps the most useful of the group. The 
great increase in interest in adrenal physiology in recent months should make this 
summary particularly valuable. 


Albert Dorfman, Chicago, Ill. 



Studies on the Action of Chloramphenicol (Chloromycetin 1 ) 
on Enzymatic Systems. I. Effect of Chloramphenicol 
on the Activity of Proteolytic Enzymes 

Grant N. Smith and Cecilia S. Worrel 

From the Research Laboratories of Parke, Davis and Co., Detroit 92, Michigan 
Received May 18,1949 

Introduction 

With the isolation and identification of the new antibiotic chloram¬ 
phenicol (6,10,25) from culture filtrates of a Streptomyces, it seemed 
desirable to investigate the effects of this drug on isolated and intact 
enzymatic systems to determine, if possible, the mode of action of this 
new compound. 

The establishment of the structure of chloramphenicol by Rebstock 
et al. (9, 17,23) as d( — )threo-l-p-nitrophenyl-2-dichIoracetamido-l,3 
propanediol suggested the possibility that the compound might be 
acting as an antimetabolite, replacing the normal peptides which were 
being utilized by the bacteria which are sensitive to this antibiotic. In 
this case, the compound would probably act as an enzymatic inhibitor 
for the proteolytic enzymes. It was therefore decided to test the in¬ 
hibitory action of chloramphenicol on proteolytic systems. The present 
paper deals with some of the observations that were made on this 
subject. 

Since the proteolytic enzymes which are known to occur in bacteria 
differ from the corresponding enzymes of animals and higher plants, as 
has been pointed out by various investigators (12,22,26), it seemed 
desirable to study the effects of chloramphenicol on those enzymes 
which can be obtained in crystalline form from animals and higher 
plants, as well as on the enzymes isolated from bacteria which are 
sensitive to the antibiotic. The former group of enzymes would give a 
more direct insight into the action of the antibiotic, since a considerable 
amount of data is available as to the chemical nature of the peptide 
molecules which are hydrolyzed by these enzymes. The data obtained 

1 Parke, Davis & Co. trade mark. 


341 



342 


GRANT N. SMITH AND CECILIA S. WORREL 


from these studies would also indicate whether the differences in the 
response exhibited by animal cells and bacterial cells to chloramphenicol 
can be explained on the basis of the effects of the antibiotic on their 
respective proteolytic enzyme systems. 

Experimental 

The effects of chloramphenicol on isolated proteolytic enzymes from plants and 
animals were investigated by using crystalline trypsin (20), chymotrypsin (18), 
pepsin (21), and papain (5). The changes in enzymatic activity were followed by using 
methods recommended by Anson et al. (2,3,4). Casein was used as the substrate in the 
case of trypsin, chymotrypsin, and papain, while hemoglobin was used as the sub¬ 
strate for pepsin. 

The effects of chloramphenicol on each of these enzymes were determined by esti¬ 
mating the decrease or increase in the activity of a standard enzyme solution after the 
treatment with the antibiotic. One ml. of the enzyme solution was mixed with 1 ml. 
of the chloramphenicol solution containing from 3.2 to 161.5 y of the antibiotic/ml. 
The solution was allowed to stand for 30 min. at 37°C. to allow sufficient time for the 
antibiotic and enzyme to react before determining the activity. 

The typical effects of chloramphenicol on crystalline trypsin, chymotrypsin, 
pepsin and papain are shown in Table I. The enzymatic activity has been expressed 
in these cases in terms of the mg. of tyrosine liberated from the substrates by enzy¬ 
matic digestion. The results were identical with all concentrations of chloramphenicol 
tested. 

The proteolytic enzymes produced by bacteria are liberated into the culture medium 
and are generally isolated from the medium during the commercial production of 

TABLE I 

The Effect of Chloramphenicol on Crystalline Proteolytic Enzymes 


Enzyme 

Concentration of 
chloramphenicol 

Tyrosine liberated 


7 /ml. 

mo. 

Trypsin 

0 

0.41 

Trypsin 

6.4 

0.41 

Trypsin 

12.8 

0.41 

Trypsin 

161.5 

0.42 

Chymotrypsin 

0 

0.57 

Chymotrypsin 

6.4 

0.58 

Chymotrypsin 

12.8 

0.58 

Chymotrypsin 

161.5 

0.57 

Pepsin 

0 

0.39 

Pepsin 

6.4 

0.39 

Pepsin 

12.8 

0.39 

Pepsin 

161.5 

0.39 

Papain 

0 

0.20 

Papain 

6.4 

0.21 

Papain 

12.8 

0.21 

Papain 

161.5 

0.20 



CHLORAMPHENICOL (CHLOROMYCETIN) STUDIES 


343 


bacterial proteases (26). At the same time, there are proteolytic enzymes which re¬ 
main within the bacterial cell and carry on the normal metabolic processes (8). From 
the data that are now available, it is impossible to ascertain whether the proteolytic 
enzymes liberated into the culture medium are identical with-those enzymes remaining 
within the cell. Therefore, the effects of chloramphenicol were studied on proteolytic 
enzymes isolated from the culture medium and also the enzymes obtained from the 
bacterial cells by autolysis or drying and defatting with acetone. 

The test organisms, B. mycoides , B. subtilis, and E. eoli, were grown in beef infusion 
broth for 6 days at 37°C. until a luxurious growth was obtained. The bacterial cells 
were then removed from the liquid phase by passing the culture solution through a 
Seitz filter. The proteolytic enzymes which have been liberated into the medium 
passed through the filter and were collected in the filtrate. The effect of chlorampheni¬ 
col on these extracellular proteases was then tested by incubating 10 ml. of the filtrate 
with 1 ml. of chloramphenicol solution and 100 ml. of 3% gelatin solution for 72 hr. at 
37°C. under sterile conditions. The extent of hydrolysis of the gelatin was then deter¬ 
mined by using the formalin titration procedure of Sorensen. 


TABLE II 

The Effect of Chloramphenicol on Proteases Produced by B. subtilis 


Enzyme 

Concentration of 
chloramphenicol 

7 /ml. 

0.1 N NaOlI 

ml. 

Extracellular protease 

100.0 

12.10 

Extracellular protease 

25.0 

12.07 

Extracellular protease 

3.1 

12.08 

Extracellular protease 

0.4 

12.08 

Extracellular protease 

0.0 

12.10 

Cell autolyzates 

100.0 

3.70 

Cell autolyzates 

25.0 

3.65 

Cell authlyzates 

3.1 

3.65 

Cell autolyzates 

0.4 

3.60 

Cell autolyzates 

0.0 

3.65 

Acetonized cells 

100.0 

6.20 

Acetonized cells 

25.0 

6.20 

Acetonized cells 

3.1 

6.20 

Acetonized cells 

0.4 

6.25 

Acetonized cells 

0.0 

6.25 


The data obtained from a typical experiment using filtrates of B. 
subtilis cultures is given in Table II. The extent of hydrolysis of the 
gelatin is expressed in terms of ml. of 0.1 N NaOH solution. Similar 
results were obtained with concentrated enzyme solutions prepared 
from the filtrates by methyl alcohol fractionation. 

A solution of bacterial proteases was prepared from the bacterial 
cells by liberating the proteases from the cells by autolysis. The cells 
collected from the cultures were washed 4 times with physiological 



344 


GRANT N. SMITH AND CECILIA S. WOBREL 


saline solution and suspended in distilled water containing toluene. 
The suspension was shaken for 1 hr. and then allowed to stand at room 
temperature for 5-7 days. The suspension was centrifuged and the 
clear supernatant used as the enzyme solution. 

The effect of chloramphenicol on these proteases was tested by using 
a mixture of 1 ml. of the autolyzate, 1 ml. of the chloramphenicol solu¬ 
tion and 50 ml. of the gelatin solution according to the procedure pre¬ 
viously described. The mixture was incubated 16 hr. and the extent of 
hydrolysis then determined by the formalin titration procedure. An 
example of the results obtained are presented in Table II. In this case 
the enzyme preparation was prepared from B. subtilis cells. 

Acetonized cell preparations of the test organisms were prepared by 
suspending the washed cells in glass-distilled acetone for 6 hr. at 3°C. 
The acetone was changed every 2 hr. and replaced with another portion 
of cold acetone. The cells were then suspended in a mixture of 50% 
acetone and 50% ethyl ether for 1 hr., and finally in ethyl ether for 1 hr. 
The cells were then dried at room temperature in a hood until all traces 
of acetone and ether had been removed. The material was then ground in 
a glass mortar using a small quantity of powdered glass. The material 
was suspended in distilled water and used as the enzyme solution. The 
effect of chloramphenicol on this preparation using the procedure out¬ 
lined above is shown in Table II. 

Discussion 

•In the experiments which have been reported above, concentrations 
of chloramphenicol have been employed which would approximate 
bacteriostatic and subbacteriostatic concentrations of the drug if used 
with organisms possessing the same concentration of enzyme used in 
these studies. 

In the initial experiments with chloramphenicol, the plate method 
suggested by Gorini et al. (13) was used to determine the inhibitory 
action of the antibiotic on the proteases liberated by the test organisms 
into the culture medium. Stone’s agar was used in preparing the assay 
plates. In these experiments, the same results were obtained as reported 
by Gorini et al. (13) when they tested penicillin. The zones of inhibition 
appeared very turbid indicating that the gelatin had not been hydro¬ 
lyzed by the bacteria present in these zones and, therefore, chloram¬ 
phenicol had decreased the proteolytic activity in these areas. The 
enhanced growth ring appeared as a transparent halo around the zone 



CHLORAMPHENICOL (CHLOROMYCETIN) STUDIES 


345 


of inhibition. Since the gelatin has been completely hydrolyzed in this 
area, this would seem to be the zone of most intense proteolytic activity. 
The background area appeared slightly turbid suggesting partial hydrol¬ 
ysis of the proteins present. These results reported above would at 
first seem to indicate that chloramphenicol inhibits the proteolytic 
activity of the enzyme liberated into the medium by bacteria. If, 
however, the number of bacteria present in each zone are taken into 
consideration, it would appear more likely that the differences observed 
were due to the difference in the number of bacteria present in each 
region. Since there are no indications in this method that cessation of 
growth was due to inhibition of proteolytic activity, and since the 
quantity of proteases produced by the small number of bacteria in the 
zone of inhibition are not sufficient to be detected by this method, it is 
very doubtful that the method shows that chloramphenicol is inhibiting 
bacterial proteases. 

From the results obtained from the studies on the effect of chloram¬ 
phenicol on proteolytic enzymes, it is apparant that although chloram¬ 
phenicol is related to peptides which are split by proteolytic enzymes 
(7) the amide group of this compound is not able to block the reaction 
between these enzymes and their normal substrates. In fact the 
chloramphenicol itself appears to be attacked by these proteolytic 
enzymes (24). It has been found that papain and trypsin will hydrolyze 
the amide linkage of chloramphenicol to a slight extent while bacterial 
proteolytic enzymes are very effective in these regards. 

From the investigation of Herriott et al. (14,15,16) it has been found 
that the phenolic hydroxyl groups and benzenoid portion of the 
tyrosine molecules present in pepsin are important for enzymatic 
activity. It appears, therefore, that chloramphenicol is not blocking 
these groups or modifying the structure of the benzenoid portion of the 
molecule by transferring chlorine atoms to it. 

The results obtained with papain also indicated that the antibiotic 
is not influencing the relationship between the disulfide and sulfhydryl 
groups of the enzyme molecule which are important for its action 
(11,19). 

Summary 

1. The inhibitory effects of the new antibiotic chloramphenicol have 
been tested on a series of proteolytic enzymes to determine whether the 
effects of this compound on gram negative bacteria can be explained in 



346 


GRANT N. SMITH AND CECILIA S. WORREL 


terms of effects of the drug on proteolytic enzymes present in the 
bacteria. 

2. Data have been presented to indicate that chloramphenicol does 
not inhibit or activate bacterial proteases, crystalline trypsin, chymo- 
trypsin, pepsin, or papain when used in subbacteriostatic and bacterio¬ 
static concentrations. 

3. The results indicated that chloramphenicol does not block the 
phenolic hydroxyl groups, amine groups, carboxyl groups, or the — SH 
groups in proteolytic enzymes tested, and which are necessary for 
enzymatic activity. 


References 

1. Anson, M. L.,./. Gen. Physiol. 20, 561 (1937). 

2. Anson, M. L., ibid. 22, 79 (1938). 

3. Anson, M. L., and Mirsky, A. E., ibid. 16, 59 (1932). 

4. Anson, M. 1.., and Mirsky, A. E., ibid. 17, 151 (1933). 

5. Halls, A. K., and Lineweaver, H., J. Biol. Chem. 130, 669 (1939). 

6. Bartz, Q. R., ibid. 172, 445 (1948). 

7. Bergmann, M., Advances in Enzymol. 2, 49 (1942). 

8. Console, A. 1)., and Rahn, 0., J. Bart. 36, 47 (1938). 

9. Controllis, J., Rebstock, M. 0., and Crooks, TI. M., Jr., J. Am. Chem. Soc. 

in press. 

10. Ehrlich, J., Bartz, Q. R., Smith, R. M., Joslyn, D. A., and Burkholder, P. R., 

Science 106, 417 (1947). 

11. Fruton, J. G., and Bergmann, M., J. Biol. Chem. 133, 153 (1940). 

12. Gale, E. F., The Chemical Activities of Bacteria, pp. 23, 44, 68, 103. University 

Tutorial Press, Ltd., London, 1947. 

13: Gorini, L., and Torriani, A., Biochem. Biophys. Acta 2, 227 (1948). 

14. IIerriott, R. \L, J. Gen. Physiol. 19, 283 (1935). 

15. Herriott, R. M., ibid. 20r335 (1937). 

16. Herriott, R. M., and Northrop, J. H., ibid. 18, 35 (1934). 

17. Long, L. M., and Troutman, H. D., J. Am. Chem. Soc. in press. 

18. Kunjtz, M., and Northrop, J. II., J. Gen. Physiol. 18, 433 (1935). 

19. Maschmann, E., Biochem. J. 279, 225 (1935). 

20. Kunitz, M., and Northrop, J. II., J. Gen. Physiol. 19, 991 (1936). 

21. Northrop, J. II., and Kunitz, M., ibid. 30, 177 (1946). 

22. Porter, J. R., Bacterial Chemistry and Physiology, pp. 518-529. Wiley, New 

York, N. Y., 1946. 

23. Rebstock, M. C., Crooks, H. M., Jr., Controulis, J., and Bartz, Q. R., J. 

Am. Chem. Soc. in press. 

24. Smith, G. N., and Worrel, C. S., Science in press. 

25. Smith, R. M., Joslyn, D. A., Gruhzit, O. M., McLean, 1. W., Jr., Penner, M. 

A., and Ehrlich, J., J. Bad. 55, 425 (1948). 

26. Wallerstein, L., Ind. Eng. Chem. 31, 1218 (1939). 



A Microanalytical Method for the Volatile Fatty Acids 12 

Simon Black 


From the Chemical Division , Department of Medicine , 

University of Chicago , Chicago , Illinois 
Received May 2, 1949 

Introduction 

The usual methods for estimating the volatile fatty acids involve 
separation of these substances from the sample by steam distillation and 
titration with standard alkali (1). In the method to be described, dis¬ 
tillation is replaced by a microdiffusion procedure with several resultant 
advantages. Large numbers of samples can be analyzed simultaneously 
with simple apparatus and a minimum of manipulation. The acids are 
collected in a very small volume in which an accurate microtitration 
can be made. 


Apparatus 

Alicrodiffusion Unit 

The microdiffusion vessel is a 50 ml. glass-stoppered Erlenmeyer flask (Fig. 1). The 
steel center cup is spun from 0.01 in. stainless steel (No. 302). 3 Its inner surface is well 
polished. It is A in. deep and has an outside diameter of \ in. The radius of curvature 
between the wall and floor of t he cup is ^ in. These cups are cleaned by boiling 10-15 
min. in dilute (1-3%) acetic acid, rinsing copiously in water, and drying in an oven. 
The steel cup is held in one of glass made by cutting the lower 15 mm. portion from a 
shell vial (15 mm. (). D.). 

Microburette 

All liquid quantities of 0.1 ml. or less are delivered with a microburette constructed 
from a tuberculin syringe and a micrometer caliper (Figs. 2 and 3). This instrument 

1 This work was aided in part by a grant to Professor E. S. Guzman Barron by the 
American Cancer Society on recommendation of the Committee on Growth of the 
National Research Council. 

2 A portion of this work was described before the meeting of the American Society 
of Biological Chemists at Atlantic City, N. J., March 16, 1948. 

3 These cups were obtained from Mr. Lee A. Farmer of the M and L Instrument 
Co., 1307 E. 71 Place, Chicago 19, Ill. 


347 



348 


SIMON BLACK 


MICRODIFFUSION UNIT 



embodies principles used by Dean and Fetcher (2). To make the burette, a J in. 
section of the needle end of the syringe barrel is slowly warmed (2-3 min.) to red heat, 
then drawn out about 4 in. with the aid of a glass rod. A right angle bend is made near 
the origin of the drawn portion and the syringe placed in an upright position in a 
muffle furnace previously brought to 600°C. After 20-30 min. the current is shut off 
and the furnace allowed to cool slowly to room temperature. At a point about 3 in. 
from the bend, the drawn glass is heated in a small flame over a length of about i in. 
and drawn out until it is sealed. It is then cut at a point just above the seal. If the 
orifice is too small the tip is ground on an abrasive stone until it is adequate. With 


MICROTITRATION AND PIPETTING APPARATUS 


r “ T 



Fig. 2. 





VOLATILE FATTY ACID DETERMINATION 


349 


an opening of proper size it is possible to fill the syringe with 0.5 ml. of water in 
10-15 sec. by pulling the plunger gently. 

To calibrate the burette the diameter of the plunger is measured at its widest point 
and the cross-section area calculated. This value in mm. 1 is equal to the number of 
mm. 3 delivered when the plunger is advanced 1 mm. The delivery error with 1 ml. 
syringes is about 0.1 mm. 3 

The most inexpensive micrometer is adequate for constructing this burette. A small 
portion of the jaw of the caliper, including the anvil, is cut away and the syringe 
attached with the aid of a clamp made from 3 wooden blocks, as illustrated in Figs. 
2 and 3. 

Titration Assembly 

The microburette is held in a ringstand clamp as shown in Fig. 3. The shaft of this 
clamp is horizontal; it is perpendicular to the syringe barrel and is held by a second 
clamp in such a way that is can rotate when the tip of the syringe is raised or lowered. 



Fig. 3. 

The titration platform consists of two small boards nailed together. The upper one 
is about iV in. thick and has a hole slightly larger than | in. in diam. drilled through its 
center. This serves as a pit for the steelcup and prevents it from spinning or migrating 
about the platform during a titration. When glass cups are used, this platform is 
unnecessary. An inverted spot plate, which provides a white background, is more 
suitable. 

An Alnico magnet (40 mm. long, 46 mm. wide) is attached to the shaft of a small 
variable speed motor (not shown in Figs.). This motor and magnet are mounted on a 
separate ringstand so that vibrations are not communicated to the rest of the appa¬ 
ratus. The position of the magnet is shown in Fig. 2. 

The interior of the steel cup must be brightly lighted during a titration. The lamp 
(Fig. 3) must be turned off between titrations to prevent warming the solution in the 
burette. 

The magnetic stirrer is made by imbedding a small piece of iron wire (“stove-pipe” 
wire) in glass. 





350 


SIMON BLACK 


Operation of Burette 

Cups into which reagents are to be measured are held in a Petri dish on the titration 
platform. To deliver a specified volume of solution the burette tip is brought to rest 
on the cup floor and the micrometer shaft advanced the number of mm. required to 
give the desired volume. The tip is then immediately raised to avoid the possibility of 
additional fluid flowing out due to capillarity. 4 During a titration the individual cup 
is held on the platform as shown in Figs. 2 and 3. With the light and stirring motor 
switched on and a magnetic stirrer in the cup, the burette tip is placed just below the 
surface of the liquid. Solution is then run in from the burette until the end point is 
observed against the bright steel surface. The syringe is then immediately raised. The 
volume of solution consumed in the titration is calculated by multiplying the cross 
section area of the syringe plunger by the number of mm. it advances, e.g., the initial 
minus the final micrometer reading. 


Reagents 

For Basic Procedure 

0.03 N KiCOy LtiHiO. Dissolve 0.124 g. of K 2 C0 3 *1.5 H 2 0 (C. P. Baker’s) plus 
0.25 g. of KNOs in distilled water, and dilute to 50 ml. 

80% H 2 SO a . Add 400 ml. of cone. H 2 S0 4 (C. P. or Reagent) to 100 ml. of 0.0005 M 
acetic acid. This solution must be well mixed. When steel center cups are not used, 
100 ml. of distilled water may be substituted for the 0.0005 M acetic acid. 

0.1 M Monosodium Maleate. Dilute 5 ml. of 1.0 N NaOH and 1.5 ml. of isoamyl 
alcohol plus 0.58 g. of maleic acid (C. P. Pfanstiehl) with distilled water to 50 ml. 

0.1 N Standard NaOH. 

Indicator Solution. One ml. of phenolphthalein indicator, 0.1% in 95% alcohol, is 
diluted to 25 ml. with C0 2 -free (boiled) distilled water. This solution should be made 
up on the day of use. 

Ag 2 SOi (C. P. or Reagent). 

For Ether Extraction Procedure 

5 N H 2 SOa. 

Alkaline Na 2 S0i. Twenty g. of anhydrous Na 2 S0 4 (C. P. or Reagent) plus 10 ml. 
of 1 AT NaOH are dissolved and diluted to 100 ml. 

Ethyl Ether. This reagent should be of good quality, substantially free of peroxides. 

For Oxidative Procedure 

Alkaline Lactate. Dissolve 0.4 ml. of Reagent lactic acid in 100 ml. of 1.0 N NaOH. 

Saturated KMnO 4 . Dissolve 7 g. of KMn0 4 (Reagent) in hot distilled water. Allow 
to cool to room temperature and use supernatant solution. 

0.1 M SnSOi. Dissolve 2.15 g. SnS0 4 (C. P.) in distilled water and dilute to 100 ml. 
This substance forms a fine precipitate immediately on dissolving. It settles very 
slowly, however, and can be pipetted and used as a suspension. 

4 A rubber band properly attached to the plunger eliminates this danger. 




VOLATILE FATTY ACID DETERMINATION 


351 


Experimental 

Basic Procedure 

The principle of the method is to separate the volatile acid by diffusion from the 
acidified sample in the main chamber of the microdiffusion unit to the alkaline solution 
in the center cup, then to estimate it in the cup by an appropriate microtitration. The 
procedure described below is suitable for acetic, propionic, and butyric acids. The 
details are exactly the same for the higher acids, such as caproic, except that the steel 
cup is omitted and the K 2 C0 3 * 1.5H 2 0 solution placed directly in a glass cup (12 mm. 
O. D.). 

’ Acids higher than butyric induce, during the diffusion process, a creeping of the 
alkaline solution over the surface and out of the steel cup, causing serious errors in 
some instances. Without the steel cup, errors occur which are due to interaction of the 
glass with the alkali, but these are less objectionable, rarely exceeding 0.07 micro¬ 
equivalent of acid. The glass cup must be discarded after one usage, for the rate of its 
reaction with K 2 C0 j- 1.5H 2 0 increases on repeated use. Cups of alkali-resistant glass 
(Corning No. 728) were, surprisingly, less satisfactory than those of the lime glass of 
which shell vials are made. 

There is a tendency for small amounts of the alkali to creep out of the steel center 
cup in the blank run as well as in the presence of certain higher acids. Error from this 
source has been reduced to a negligible value by inclusion of KNO s in the alkaline 
solution, by inclusion of a trace of acetic acid in the H 2 S0 4 added to the main chamber, 
and by rapid, rather than slow, heating of the unit in the oven. 

The efficiency of the alkaline solution as a trapping agent is strongly affected by the 
solubility of the alkali used. KOH is rapidly converted to the carbonate when added 
to the center cup. The carbonate, K 2 COa, is considerably less soluble than the hydrate, 
K 2 C0 3 -1.5 H 2 0. The latter is the only agent which has permitted good recoveries of 
volatile acid over an extended period. 

When the diffusion of volatile acid into the center cup is complete, a solution of 
monosodium maleate is used to neutralize the excess alkali in the cup and liberate 
C0 2 . Isoamyl alcohol reduces the surface tension of the solution, allowing it to spread 
readily into a thin layer over the interior of the cup. While monosodium maleate is a 
sufficiently strong acid to liberate C0 2 and to be titrated to the phenolphthalein 
endpoint, it is too weak to liberate a substantial fraction of free fatty acids from their 
salts. The latter are, therefore, not lost by evaporation during the escape of C0 2 . 

In the steam distillation of fatty acids (1), salts such as MgS0 4 are used to increase 
the distillation rate. Salts were not found adequate, however, to promote rapid 
diffusion of these acids from the main chamber into the center cup of the apparatus 
described here. H 2 S0 4 , in relatively high concentration, was more satisfactory. 

The details of the basic procedure are as follows: One ml. of an un¬ 
known solution, containing not more than 2 microequivalents of 
volatile acid, is pipetted into a glass-stoppered Erlenmeyer flask. One ml. 
of 80% H 2 SO 4 is added and the flask shaken gently to ensure mixing. 
The glass-held steel center cup, containing 100 mm. 3 of K 2 CXVI. 5 H 2 O 



352 


SIMON BLACK 


solution, is taken from a Petri dish and placed centrally on the floor 
of the flask with the aid of forceps (Fig. 1). The glass cup stands directly 
in the acidified sample. The glass stopper is put loosely in place (do not 
twist or “lock”) and the flask set in a 100-105°C. oven. It should be set 
directly on the oven shelf or grating, not in a tray. After 14-16 hr. the 
center cup is removed and placed on a paper towel which absorbs the 
liquid from the under surface. The steel cup is then removed from the 
glass one by grasping from the inside with sharp-pointed forceps having 
the prongs bent outward, and placed in a Petri dish. Forty mm.* of 
monosodium maleate solution are added to the cup and the dish tilted 
gently, if necessary, to ensure distribution of this reagent over the cup 
bottom. After 10-20 min., during which C0 2 escapes, about 0.3 ml. of 
indicator solution and a magnetic stirrer are added. The cup is placed 
on the titration platform and the titration performed as described 
above with 0.1 N NaOH. 

TABLE I 


Recovery of Individual Volatile Acids from Standard Solutions 

Procedure described in text. Glass center cups were used for valeric, caprylic and 
caproic acids. 


Microequiva- 
lents added 

Microequivalents found 

Acetic 

Propionic 

Butyric 

Valeric 

Caproic 

Caprylic 

. 0.10 

0.09 

0.10 

0.09 

0.13 

0.07 

0.15 

1.00 

0.98 

1.00 

1.03 

1.03 

1.00 

0.87 

2.00 

2.03 

2.05 

1.97 

1.97 

1.92 

1.65 


The amount of volatile acid is calculated by subtracting a blank 
titration value from that obtained with the unknown. The theoretical 
blank is equal to the difference in the number of equivalents of maleate 
and KjCOj • 1.5H 2 0 plus the trace amount of acetic acid present in 1 ml. 
of the 80% H 2 SC> 4 . 

Table I shows the recovery of 6 volatile acids from standard solutions 
by the basic procedure just described, and Table II shows the recovery 
of 3 acids from copper-lime filtrates of yeast. 

If chloride is present in the sample, it will yield HC1 to the center 
cup. This is prevented by addition of about 10 mg. of solid Ag 2 S0 4 
prior to the H^SOi solution. The flask should be tilted and rotated 
several times to ensure distribution of the silver throughout the solution. 





VOLATILE FATTY ACID DETERMINATION 


353 


The basic procedure has been used with no significant interference 
from the following acids: acetoacetic, citric, isocitric, cw-aconitic, a- 
ketoglutaric, succinic, fumaric, and malic. Of 10 amino acids tested, 
none gave interference except tryptophan, which decomposed in the 
strong acid and yielded more than 0.5 microequivalent of volatile acid 
per micromole of the compound added. Lactic, pyruvic and oxaloacetic 
acids yielded 0.2-0.5 microequivalent of volatile acid per micromole, 


TABLE II 

Recovery of Volatile Acids from Copper-Lime Filtrates of Yeast 

20 mg. dry wt. of washed bakers' yeast/ml. was suspended in & solution containing 
0.02 M NaH 2 P 04 and 0.13 M NaCl. One ml. of this suspension was added to a centri¬ 
fuge tube containing 7 ml. H 2 0 and one of the volatile acids (neutralized) in sufficient 
amount to give the indicated final ratio of acid to yeast. One ml. of 15% CuS0 4 -5H 2 0 
was added and mixed, then one ml. of a 10% Ca(OH) 2 suspension. After shaking and 
standing 15 min. the tubes were centrifuged. One ml. aliquots of the filtered superna¬ 
tant solutions were analyzed by the basic procedure. About 10 mg. of Ag 2 S0 4 were 
added to each microdiffusion unit prior to addition of the H 2 S0 4 . Blank: 0.12 micro¬ 
equivalent of volatile acid/2 mg. yeast. 


Microequivalcnts added 
/2 mg. of yea&t 

Microequivalents found/2 mg. of yeast 

Acetic 

Butyric 

Caproic 

0.20 

0.20 

0.20 

0.19 

0.60 

0.62 

0.60 

0.54 

1.20 

1.18 

1.27 

1.27 


and /J-hydroxybutyric acid 0.5-0.6 microequivalent. Glucose and 
carbohydrate-containing substances such as adenosine and its phos¬ 
phates interfere seriously, to about the same extent per mole as 0- 
hydroxybutyric acid. Solid HgS0 4 , added to the flask, completely pre¬ 
vents diffusion of pyruvic acid. 

Ether Extraction 

The volatile acids may be separated from the majority of interfering substances by 
extracting them from aqueous solution with ethyl ether. Table III shows the fraction 
of each of 6 acids extracted when 15 ml. of ether was used to extract 3.2 ml. of aqueous 
solution, as well as when the water and ether volumes were equal. It was found possible 
to quantitatively re-extract the acids from ether into a small volume of NaOH 
solution only when a large amount of Na 2 S 0 4 was added. 



354 


SIMON BLACK 


The procedure is as follows: Three ml. of a protein-free filtrate is 
placed in a 30 ml. glass-stoppered centrifuge tube . 6 Two-tenths ml. of 
5 N H 2 SO 4 is added, 15 ml. of ether run in, and the tube shaken vigor¬ 
ously for 2 min. Ten ml. of the supernatant ether is transferred with a 
pipette* to a second stoppered tube containing 1.2 ml. of the alkaline 
Na?S 04 solution. This tube is shaken for 1 min. One ml. of the aqueous 
layer is then transferred with a fine-tipped pipette to a 50 ml. Erlen- 
meyer flask and the analysis carried out as described above by the 
microdiffusion procedure. 

TABLE III 

Extraction of Volatile Acids from Water with Ethyl Ether 
I. 15 ml. of ether used to extract 3.2 ml. of aqueous solution." II. 10 ml. of ether 
used to extract 10 ml. of aqueous solution. 


Acid extracted 

Per cent of acid found 

in ether lay 


I 

II 

Acetic 

71 

32 

Propionic 

87 

60 

Butyric 

95 

83 

Valeric 

98 

93 

Caproic 

100 

98 

Caprylic 

100 

100 


0 The small amount of calcium in copper-lime filtrates causes additional retention 
in the aqueous layer of about 0.1 micromole of volatile acid; this amount is independ¬ 
ent of the total volatile acid in the sample. Addition of 0.1 ml. of 10% Xa 2 WO, 
prior to extraction causes precipitation of the calcium (which need not be removed) 
and elimination of this source of error. 

To calculate the amount of volatile acid present in the original 3 ml. 
of filtrate, the quantity found is divided by the factor 0.55 multiplied 
by whatever fraction of the acid in question Table III indicates should 
be extracted by this procedure. Table IV shows the recovery of 3 
volatile acids from 3 different animal tissues by the procedure just 
described. 

Carried through this procedure, lactic and pyruvic acids yielded 
0 . 1 - 0.2 microequivalent of volatile acid, calculated as acetic acid, per 
micromole of either of these substances present in the filtrates. /S- 

* Maizel-Gerson reaction vessel. Obtained from the Wilkens-Anderson Co., Ill N. 
Canal St., Chicago, Ill. 

• The tip of this pipette is bent 90° with respect to the pipette axis to avoid drawing 
up aqueous solution. One pipette is used for all transfers; a stream of air is drawn 
through it with a water aspirator after each transfer to clear it of ether and volatile 
acid. 



VOLATILE FATTY ACID DETERMINATION 


355 


Hydroxybutyric acid exhibits the same distribution ratio between water 
and ether as acetic acid, hence interferes to about the same extent as 
when filtrates are analyzed directly. If it is desired to use HgS0 4 or the 
KMn0 4 procedure (below), following the ether extraction, the alkaline 
NajSOi extract must first be evaporated to dryness in the flask, then 
1 ml. of water added. 


TABLE IV 

Recovery of Individual Volatile Acids from Animal Tissues 
by Ether Extraction Procedure 

The fresh tissue samples were homogenized in sufficient water to give a suspension 
with 0.15 g. tissue/ml. Two ml. of the suspension was added to a centrifuge tube 
containing 6 ml. ofwater and an amount of the volatile acid (neutralized) to give the 
indicated ratio of acid to tissue. The copper-lime precipitation was then carried out as 
indicated in Table II. Blank volatile acid values, calculated as acetic acid, were as 
follows in microequivalents/30 mg. fresh tissue: liver 0.11, brain 0.03, and muscle 0.07. 


Microoquiv- 
alents added 
30 mg. of 
f issue 

Mieroequivalents found/30 mg. of tissue 

Rubbit h\ei 

Hat brain 

Rabbit muscle 


Aeetic 

But> lie 

Cupime 

Acetic 

But\ no 

Caprnic 

Acetic 

Butyric 

Oaproic 

0.20 

0.20 

0.20 

0.20 

0.21 

0.22 

0.18 

0.21 

0.25 

0.18 

0.60 

0.56 

0.58 

0.57 

0.68 

0.60 

0.60 

0.63 

0.61 

0.58 

1.20 

1.14 

1.18 

1.06 

1.21 

1.23 

1.05 

1.15 

1.16 

1.06 


Analyses of Mixtures 

Osburn, Wood and Werkman (3) have described procedures for analyses of mixtures 
of volatile acids in which the individual acids are distinguished on the basis of the 
differences in their distribution coefficients between water and ethyl ether. Analogous 
procedures have been used successfully in conjunction with the microdiffusion method. 

Oxidative Procedure for Removal of Pyruvic and Lactic Acids 
in Acetic Acid Determinations 

It is known that pyruvic (4) and lactic acids (5) can be oxidized in alkaline solution 
to C0 2 and H 2 0. We have attempted to remove these substances from solutions of 
acetic acid by such a procedure. 

It was found that alkaline KMn0 4 , on evaporation in a 100-105 o C. 
oven, completely removed 5 micromoles of lactic acid, using volatile 
acid recovery as the index of its presence. One micromole, however, was 





356 


SIMON BLACK 


only partially destroyed. Pyruvic acid was removed completely by the 
procedure only when lactic acid was also present. 

We believe, though it has not been directly demonstrated, that the Mn<0<, which 
is deposited relatively slowly on the flask bottom during lactate oxidation, catalyzes 
the further oxidation of organic materials in the solution. We have found, in contrast, 
that MnO or MnOj catalyze the formation of acetate from either of these two com¬ 
pounds. Fortunately, manganese ions are removed from filtrates by copper-lime 
precipitation and thus do not complicate the procedure. 

The procedure is of limited usefulness because it will convert certain 
substances to volatile acids which are not so originally, such as some of 
the amino acids. The majority of the latter substances tested yielded 
0.2-0.3 microequivalent of volatile acid per micromole. Some, however, 
such as glycine, alanine, tyrosine, and aspartic acid are completely 
oxidized. 

Acetoacetic acid, which does not yield volatile acid when carried 
through the basic procedure because it is decarboxylated by the strong 
H2SO4, is oxidized almost quantitatively to acetic acid by this proce¬ 
dure. If it is desired to use the oxidative procedure when acetoacetate 
is present, the latter may first be completely removed by making the 
filtrate 0.1 N with respect to H 2 SO 4 and heating in a boiling water bath 
for 10 min. Sufficient additional alkali must then be added in the oxida¬ 
tive procedure to neutralize the H2SO4. 

The oxidative procedure reduces the high interference of / 2 -hydroxy- 
butyric acid to about 0.2 microequivalent of volatile acid per micromole. 
Some of the following acids are oxidized completely and some not at all, 
but none yields volatile .acid when carried through the procedure: 
citric, isocitric, m-aconitic, a-ketoglutaric, succinic, fumaric, malic and 
oxaloacetic. Glucose may also be removed in this way, but not inter¬ 
ference by adenosine and related substances. 

The higher volatile acids are partially oxidized. On the basis of 
volatile acid recovered, butryric acid is about 50% destroyed and 
caproic acid 60%. 

Some unidentified neutral but volatile product of this procedure 
causes occasional errors due to alkali creeping out of the steel center 
cup. Such errors occur only when the amount of volatile acid is very 
small and can be avoided by making the following two changes in the 
reagents used in the microdiffusion unit: 1 . Substitute 0.005 M for 
0.0005 M acetic acid in making up the 80% H2SO4 solution; 2 . Substi- 



VOLATILE PATTY ACID DETERMINATION 


357 


tute 0.04 N KjCO*'1.51120-0.5% KNO 3 for the 0.03 N solution speci¬ 
fied 

The details of the procedure follow: One ml. of protein-free filtrate 
is placed in a 50 ml. glass-stoppered Erlenmeyer flask. One-tenth ml. 
of alkaline lactate and 0.1 ml. of saturated KMnO< are added in succes¬ 
sion. The flask is then tilted gently back and forth 12-15 times to 
ensure thorough mixing of the solutions (important). Care should be 
taken to avoid spattering solution high on the flask wall. The open 
vessel is then placed in the 100-105°C. oven. When the contents have 
evaporated to dryness, 1 ml. of 0.1 M SnS0 4 is added plus about 10 mg. 
of solid Ag 2 S0 4 . After gentle shaking 1.0 ml. of 80% H2SO4 is added and 
the analysis carried out by the basic procedure. 

Recovery of acetic acid by this procedure from 3 animal tissues is 
shown in Table V. 


TABLE V 

Recovery of Acetic Acid from 3 Rabbit Tissues Using Oxidative Procedure 
Filtrates were prepared as indicated in Table IV. One ml. aliquots were used for 
each determination. Volatile acid blank values, expressed as microequivalents/30 mg. 
of tissue, were as follows: liver 0.03, brain 0.05, and muscle 0.20. 


Microequivalents added 
/30 mg. of tissue 

Microequivalents found/30 mg. of tissue 

Liver 

Brain 

Skeletal muscle 

0.20 

0.25 

0.17 

0.23 

0.60 

0.60 

0.50 

0.57 

1.20 

1.30 

1.20 

1.17 


Colorimetric Method for Acetic Acid 

The method of Hutchens and Kass ( 6 ) may be applied directly to the 
acid collected in the microdiffusion center cup, and numerous sources 
of interference with color development thus avoided. Ba(NO s ) 2 
equivalent to the K 2 CO 3 *1.5^0 must be added prior to color develop¬ 
ment to neutralize the sample and precipitate carbonate. 


Deproteinization 

Lehninger and Smith, in the development of a specific method for 
caprylic acid, have determined the conditions required for good re- 









358 


SIMON BLACK 


covery of fatty acids in protein-free tissue filtrates (7). The details of 
their procedure as we have used it are indicated In Tables II and IV. 

Other Uses of the Apparatus 

It is of interest that the ungreased glass stopper of the Erlenmeyer 
flask allows no significant vapor leakage in a prolonged experiment. 
Little water is lost during the microdiffusion process despite a water 
vapor pressure in the flask of about 0.5 atmosphere. We have used the 
apparatus described here for the determination of ethanol ( 10-100 y) 
and of ammonia (1-10 7 ) by published microdiffusion methods (8,9). 
100 % recoveries were readily obtained. 

Acknowledgments 

The author is grateful to Professor E. S. Guzman Barron for supporting this work, 
and to Mrs. Marion L. Hearon for valuable assistance in some of the experiments. 

Summary 

1. A microdiffusion method for determining volatile fatty acids is 
described. The acids are caused to diffuse in a closed system from an 
acidified solution to a small cup containing standard alkali. They are 
subsequently estimated in the cup by a suitable microtitration. The 
range for the experimental conditions outlined is 0 . 2 - 2.0 microequiva¬ 
lents. The average error in determining acetic, propionic and butyric 
acids is less than 0.03 microequivalent. 

2. The method is essentially non-specific. However, individual acids 
may be distinguished in some instances by their characteristic distri¬ 
bution ratios between water and ethyl ether. 

3. A general procedure is described for separating the acids from 
such interfering materials as lactic and pyruvic acids and others by 
extraction with ethyl ether. 

4. A procedure for the removal of lactic and pyruvic acids by alka¬ 
line permanganate oxidation is described, suitable for use in acetic 
acid determinations under certain circumstances. 

5. A simple microburette and microdiffusion unit are described 
which are of general usefulness in microdiffusion analyses. 

References 

1. Fbiedeman, T. E., J. Biol. Chem. 123, 161 (1938). 

2. Dean, R. B., Fetcher, E. S., Jr., Science 96, 237 (1942). 



VOLATILE FATTY ACID DETERMINATION 


359 


3. Osburn, 0. L., Wood, H. G., and Werkman, C. H., Ind. Eng. Chem ., Anal. Ed. 

8, 270 (1936). 

4. Meyer, K., J. Biol. Chem. 103, 39 (1933). 

5. Conant, J. B., and Tongberg, C. 0., ibid. 88, 701 (1930). 

6. Hutchens, J. 0., and Kass, B. M., ibid. 177, 571 (1949). 

7. Lehninger, A. L., and Smith, S. W., ibid. 173, 773 (1948). 

8. Winnick, T., Ind. Eng. Chem., Anal. Ed. 14, 523 (1942). 

9. Conway, E. J, Microdiffusion Analysis and Volumetric Error, Rev. Ed., C. 

Lockwood, London, 1947. 



The Biological Degradation of Lignin . 1 
I. Utilization of Lignin by Fungi 

William C. Day, Michael J. Pelczar, Jr. and Sidney Gottlieb 

From the Departments of Bacteriology and Chemistry , 

University of Maryland , College Park , Maryland 
Received May 13, 1949 

Introduction 

9 

Ample evidence exists to indicate that various species of micro¬ 
organisms are capable of degrading lignin in situ in plant materials 
(1,2,3). Numerous claims have been made in the literature purporting 
to demonstrate the ability of microorganisms to utilize lignin as a sole 
carbon source, or purporting to demonstrate the existence of enzymes 
acting specifically on the lignin molecule (4,5,6). The obvious deterrent 
to a rational approach to the study of lignin enzymes and of the utili¬ 
zation of lignin by fungi has been the lack of a method for isolating 
lignin by a procedure which minimizes structural alterations in the 
lignin molecule during the isolation process. Brauns (7), in 1939, 
proposed a method for isolating part of the lignin from wood by a 
process in which no drastic agents (no heat, acid, or alkali) are used. 
Although the “native lignin” isolated by this process represents only a 
small part of the lignin in wood, Brauns presents much evidence (7) to 
indicate the identity, or at least the close similarity, of this material 
with lignin as it exists in wood. The availability of such a procedure 
seemed to us to warrant an experimental reexamination of the ability of 
microorganisms to utilize lignin and of the possible existence of lignin¬ 
degrading enzymes. 

This paper presents the results of studies on the utilization of native 
lignin by various fungi; the paper which follows will report the investi¬ 
gations made on a lignin-oxidizing enzyme. 

1 This work was done under contract No. N7 onr 397-4 between the University of 
Maryland and the Office of Naval Research. The project was initiated at the sug¬ 
gestion of the Prevention of Deterioration Center, National Research Council. 


360 



BIOLOGICAL DEGRADATION OF LIGNIN. I 


361 


Experimental 

Methods and Materials 

Cultures. Thirty-five cultures of wood-destroying (white-rot) fungi were employed 
in this study as listed in Table I. Stock cultures of these organisms were maintained 
on potato dextrose agar, subcultured at monthly intervals, and stored at 10°C. during 
the interim. 

Preparation of Native IAgnin. Native lignin was prepared from freshly cut 12 year 
old red spruce wood ( Picea rubra) by the method described by Brauns (8). 


TABLE I 

Wood-Rotting (White Rot) Fungi Screened for Lignin- 
Decomposing Activity 


Culture 

number 

Genus and species designation 

Culture 

number 

Genus and species designation 

1 

Polyporus oregonense 

18* 

Fomes pini 

2 

Fomes fomentarius 

19 

Fomes igniarius var . laevigatus 

3* 

Poria subacida 

20 

Fomes pini 

4* 

Poria subacida 

21 

Echinodontium tinctoriurn 

5 

Pholiota adioposa 

22 

Polyporus anceps 

6 

Corticium galactinum 

23 

Polyporus taugae 

7 

Polyporus valvatus 

24 

Poria weirii 

8 

Ustulina vulgaris 

25 

Fomes annosus 

9 

Fomes geotropus 

26 

Peniophora gigantea 

10 

Stereum sulcatum 

27* 

Fomes annosus 

11 

Lentinus tigrinus 

28* 

Fomes annosus 

12 

Polyporus borealis 

29* 

Fomes pini 

13 

Polyporus abietinus 

30* 

Fomes igniarius 

14 

Poria subacida 

31* 

Ganoderma applanatum 

15 

Collybia vdutipes 

32* 

Polyporus circinatus 

16 

Fomes applanatus 

33* 

Fomes igniarius 

17 

Daedelea unicolor 

34* 

Polyporus circinatus 



35* 

Ganoderma applanatum 


* Cultures received from Dr. Mildred K. Nobles, Ottawa, Canada. 

Remaining cultures were obtained from Dr. Ross W. Davidson, Beltsville, Mary¬ 
land. 


Preparation of Lignin for Incorporation into Media. A weighed amount of dry lignin 
was added to a measured amount of distilled water (2% suspension) and homogenized 
in a Waring Blendor. This very fine suspension of lignin was then dispensed into 126 
ml. flasks and autoclaved for 10 min. at 10 lbs. pressure. This served as the stock 
lignin for incorporation into media. 

The sterilized aqueous suspension of lignin was found to contain a thin layer of 
dark material as a result of a thermoplastic change during autoclaving. This material 







362 


W. C. DAY, M. J. PELCZAR, JR. AND S. GOTTLIEB 


was removed from the flask with aseptic precautions; it amounted to approximately 
half of the amount of lignin added. The lignin remaining in suspension, on chemical 
analysis was found to have a methoxyl content of 14.8%, the same as the unheated 
lignin. The dark material likewise was found to have the same methoxyl content. 

For incorporation into media, equal parts of the sterilized lignin and double strength 
liquid or solid sterile media were combined. This resulted in a final lignin concentra¬ 
tion of approximately 0.5% in the media. The stock lignin suspensions ranged in pH 
from 5.4 to 5.6, and therefore did not alter the pH of the media, which was 5.5. 

Some of the final experiments reported here were performed with lignin which was 
sterilized by exposure to freshly distilled ether. The lignin, after a 3-day contact with 
ether, was recovered by evaporation of the ether and then, with aseptic precautions, a 
stock suspension (in water) was prepared, using a Waring Blendor. 

The technique employed for incorporation of lignin into solid media consisted of 
adding 20 ml. of sterilized agar basal medium to a sterile Petri dish or a sterile 125 ml. 
Erlenmeyer flask. After this layer of medium solidified, it was overlaid with a 5 ml. 
mixture of the same basal medium containing lignin. This prevented settling out of 
the lignin, which provided for the mycelial inoculum to be in contact with lignin. • 

Preparation of Culture Inoculum. In the preparation of mycelial inoculum for test 
media, it was found necessary to use inocula with minimum carry over of nutrients. 
Inocula obtained from stock cultures grown on potato dextrose-agar proved to be 
unsatisfactory. Therefore, a mineral salt medium containing washed agar and inor¬ 
ganic salts was prepared and inoculated from potato dextrose-agar stock cultures. 
Mycelia produced on this medium were used as inocula for test media, taking pre¬ 
cautions to avoid contact with the original portion of inoculum. A third type of 
inoculum was prepared by inoculating malt extract broth with the original stock-cul¬ 
ture inocula contained in 125 ml. Erlenmeyer flasks. These flasks were aerated in a 
shaking apparatus at incubation temperatures fluctuating between 25 and 30°C. for 
one week. The mycelial pellets which had formed in these flasks were washed 6 times 
in sterile distilled water, after which they were suspended in buffered sterile water. 
These pellets were then used as inoculum. 

The following precautions were observed during inoculation of media: Use of 
uniform amounts of inoculum,"'floating inoculum on surface of still liquid cultures, 
and inoculation of central portion of all solid media. 

Incubation of Cultures. Throughout the course of this work cultures were incubated 
at a temperature range of 25-30°C. for periods of 1-4 w eeks as indicated in the tables 
of results. Agar media were dispensed in shallow layers in 125 ml. flasks as well as in 
Petri dishes. Liquid media were dispensed in 20 ml. amounts in 125 ml. flasks and 
incubated by both still and shake culture techniques. The shaking apparatus was 
adjusted to a stroke of 1J" with 100 three in. excursions/min. 

Preparation of Basal Media. At the beginning of this study numerous synthetic and 
natural media were prepared in an attempt to arrive at an “all or none” growth re¬ 
sponse to a carbon source. This was complicated by the fact that growth was usually 
very poor with an inorganic nitrogen source and when organic nitrogen was added, the 
compound would also serve to a limited degree as a carbon source. The composition 
of media included in this report are given in the various tables where growth response 
to lignin is recorded. 

All media were adjusted to pH 5.5 and then sterilized at 15 lbs. pressure (121°C.) 



BIOLOGICAL DEGRADATION OF LIGNIN. I 


363 


for 15 min. Glucose and thiamine were sterilized separately by filtering through a 
sintered glass filter, after which they were added aseptically to the already sterilized 
medium according to the concentration desired. The mineral supplement employed in 
all media was Hoagland’s A-Z mixture (9). 

Ten ml. of a 1% sterilized aqueous suspension of lignin were aseptically added to 
equal amounts of double strength concentrations of pre-sterilized basal media, when 
a lignin-containing medium was desired. 

Measurement of Growth Response 

a. Solid Media . Measurement of the diameter of giant mycelial colonies were 
recorded in mm. No attempt was made to measure the comparative thickness of the 
mycelial mats. 

b. Liquid Media. Measurement of mycelial growth in liquid media was made in 
2 ways. (1) Comparative visual observations recorded as a series of +\s or — \s if no 
growth was evident; (2) weighing of the mycelial mat, by use of the following tech¬ 
nique: Liquid culture media left in the flasks after incubation was carefully poured 
off, leaving only the mycelia in the culture flasks. To this flask was then added 25 ml. 
of dioxane, after which the flask was placed on a shaking apparatus and shaken for 8 
hr. to remove any unutilized, adsorbed lignin which may have adhered to the myce¬ 
lium during incubation. After shaking, the mycelial-dioxane mixture was filtered 
through an asbestos pad contained in a tared procelain Gooch crucible, washed again 
with 25 ml. of dioxane, and then dried at 55°C. until constant weight was obtained. 
This procedure was also applied to the mycelia growing in the control flasks which 
contained no lignin. 

Results 

Screening Experiments on Liquid and Solid Media 

During initial attempts to screen the 35 white rot fungi cultures for 
utilization of lignin, it was found that a clear-cut ‘‘all or none” growth 
response could not be obtained when lignin was the limiting carbon 
source in the medium. Some of the difficulties encountered were as 
follows: 

On agar media, most of the cultures grew to some extent on the basal 
medium without added carbohydrate (Table II). This may have been 
due to carry over of some nutrients with the inoculum, size of inoculum, 
which was variable, etc. In liquid media, an inorganic source of nitrogen, 
such as NH4NO3 or (NH 4 ) 2 S 0 4 , did not prove satisfactory. Supple¬ 
menting such inorganic nitrogen in the medium with amino acids or 
casein hydrolyzate resulted in media upon which the fungi would grow 
to some degree without added carbohydrate. As a consequence, numer¬ 
ous solid and liquid media were prepared and all 35 cultures were 
tested. Each medium was prepared in duplicate sets, one with and 
one without lignin. 



364 


W. C. DAY, M. J. PELCZAR, JR. AND S. GOTTLIEB 


TABLE II 

Results of Screening Experiments to Demonstrate Lianin Utilization 
by White Rot Fungi on Solid and Liquid Media 


Culture number 

Growth on solid media after 

2 weeks incubation 

Growth on liquid media after 

4 weeks incubation 

Basal* medium 

Basal medium plus 
0.5% lignin 

Basal 6 medium 

Basal medium plus 
0.5% lignin 

1 

mm. c 

20 

mm. r 

10 


_d 

2 

45 

90 

+ 

2+ 

3 

35 

60 

+ 

2 + 

4 

15 

30 

+ 

24- 

5 

25 

60 

— 

— 

6 

30 

55 

± 

4- 

7 

10 

- 

2 + 

24- 

8 

10 

65 

+ 

4- 

9 

90 

90 

+ 

2 + 

10 

15 

— 

— 

— 

11 

90 

90 

+ 

2 + 

12 

35 

20 

- 

— 

13 

15 

90 

± 

+ 

14 

25 

90 

dt 

+ 

15 

45 

90 

4 - 

2 + 

16 

25 

45 

dtz 

4- 

17 

25 

90 


24- 

18 

30 

5 

- 

- 

19 

25 

10 

4- 

4- 

20 

5 

30 

+ 

+ 

21 

10 

- 

- 

— 

22 

45 

60 

+ 

24- 

23 

55 

10 

+ 

4- 

24 

15 

25 

± 

+ 

25 

30 

90 

rk 

+ 

26 

60 

<>0 

± 

24- 

27 

25 

5 

- 

— 

28 

45 

10 

- 

- 

29 

50 

10 

— 

— 

30 

40 

15 

— 

— 

31 

35 

15 

— 

— 

32 

12 

- 

— 

— 

33 

30 

75 

+ 

24- 

34 

5 

— 

— 

— 

35 

5 

35 

+ 

24- 


fl Basal medium: NH 4 N0 3 0.5% Agar 2.00% 

KH 2 P0 4 0.15% Thiamine (HC1) 0.001% 

MgS0 4 -7H 2 0 0.05% Mineral supplement 1ml. 

CaCO, 0.02% Water (distilled) 1 1. 

b Same as above, but without agar incorporated. 
c Diameter of giant colony in mm. 

d Reading of — indicates no growth; d=, doubtful growth response; -f, small 
mycelial mat; 2-f, half the liquia surface covered with mycelial mat. 




BIOLOGICAL DEGRADATION OF LIGNIN. I 


365 


Examples of the type of response obtained in solid and liquid media 
containing inorganic nitrogen, with and without lignin, are presented 
in Table II. The cultures which seemed to produce greater amounts of 
mycelia on the solid medium containing lignin were as follows: No. 3, 
Poria subacida; No. 4, Poria subacida; No. 8, Ustulina vulgaris; No. 13, 
Polyporus abietinus; No. 14, Poria subacida; No. 15, Collybia velutipes; 
No. 17, Daedalea unicolor; No. 25, Fomes annosus; No. 26 Peniophora 
gigantea; No. 33, Fomes igniarius; and No. 35 Ganoderma applanatum. 

Several of the cultures gave indications of being inhibited by the 
presence of lignin, in which instance they produced no evidence of 
growth on the basal medium containing lignin. These cultures were 
as follows: No. 7, Polyporus valvatus; No. 10, Stereum sulcatum; No. 21, 
Echinodontium tinctorium; No. 32, Polyporus circinaius; and No. 34, 
Polyporus circinatus. Growth in the liquid medium described in Table 
II was irregular, and, in those cases where growth did occur, it was not 
very abundant. It was, however, significant to note that cultures No. 
6, Corlicium galadinum, No. 13, Polyporus abietinus, No. 14, Poria 
subacida, No. 16, Fomes applanatus, No. 17, Daedalea unicolor, No. 24, 
Poria weirii, and No. 25, Fomes annosus, presented evidence of growth 
in the medium containing lignin, but gave a doubtful growth 
response in the medium minus lignin. 

Although the results presented in Table II were suggestive of growth 
response to lignin by several of the fungi employed, the results were not 
interpreted as definitive. Repeated experiments following the scheme 
presented in this table gave variable results so that a consistent, 
unequivocal response to lignin could not be attained. 

Adaptation Technique for Lignin Utilization 

As a consequence of the results obtained in the screening experiments 
described above, an attempt was made to stimulate utilization of lignin 
by a process of adaptation. For this purpose, a basal medium of the 
composition indicated in Table III which contained 0.1% asparagine as 
a supplementary source of nitrogen was employed. Asparagine in con¬ 
junction with (NH 4 )N0 3 was found to provide a more suitable source 
of nitrogen than (NH 4 )NOj alone. 

For these experiments, the following cultures, which previously had exhibited 
stimulation of growth in the presence of lignin, were employed: No. 9, Fomes geotropus; 
No. 11, Lentinus tigrinus; No. 13, Polyporus abietinus; No. 14, Poria subacida; No. 17, 
Daedalea unicolor; and No. 20, Peniophora gigantea. Each of these cultures was 



366 


W. C. DAY, M. J. PELCZAR, JR. AND S. GOTTLIEB 


TABLE III 

Adaptation of White Rot Fungi to Lignin Utilization 


Amount of mycelial growth in mg. of several white rot fungi serially subcultured 
after 3 week incubation periods in a basal medium with decreasing 
glucose concentration and constant lignin concentration. 

Still Cultures 


Basal medium'* plus glucose Basal medium plus glucose and 0.5% lignin 


Culture Per cent glucose and serial subculture 6 Per cent glucose and serial subculture 

number 



0 . 1 % 

.05% 

•025% 

. 01 % 

0 % 

0 . 1 % 

.05% 

.025% 

01 % 

0 % 


1st 

2ml 

3rd 

ith 

5th 

1st 

I 

2nd 

3rd 

4 th 

5th 

9 

15 r 

17 

15 

10 

4 

18 

20 

15 

l] 

3 

11 

20 

22 

18 

14 

5 

25 

24 

21 

16 

6 

13 

10 

8 

5 

4 

1 

20 

22 

25 

26 

2 \\ 

14 

14 

11 

0 

6 

4 

37 

30 

35 

31 

33 

17 

30 

21 

15 

11 

7 

50 

40 

35 

28 

15 

26 

20 

15 

10 

8 

3 

22 

17 

14 

12 

4 


Shake Cultures 


0 

12 

15 

14 

12 

6 

22 

18 

16 

13 

6 

11 

24 

20 

18 

17 

9 

30 

26 

24 

10 

11 

13 

13 

10 

7 

5 

2 

23 

25 

20 

30 

31 

14 

28 

8 

5 

2 

3 

22 

21 

23 

25 

28 

17 

43 

32 

21 

13 

10 

52 

41 

I 25 

18 

13 

* 26 

15 

13 

11 

10 

4 

25 

10 

17 

14 

(> 


u Basal medium: 


NH 4 NO 3 . . . . •.0.5% 

Lr(-f-)-Asparagine. •.0.1% 

KH 2 P0 4 .0.15% 

MgS0 4 ’7H 2 0 ■.0.05% 

CaCOj. 0.02% 

Thiamine (HC1)-.0.001% 

Mineral supplement .. •.1 ml. 

Water (distilled) to. . •.-11. 


b Cultures incubated for 3 weeks, whereupon a small piece of mycelium was trans¬ 
ferred to a similar medium containing less glucose. This process involved 5 serial 
cultures, the last medium containing no glucose. 
e Figures represent weight of mycelia in mg. 











BIOLOGICAL DEGRADATION OF LIGNIN. I 


367 


inoculated into 2 flasks of the basal medium, one of which contained glucose (0.1 %) 
and the other both lignin (0.5%) and glucose (0.1%) as indicated in Table III. 
Duplicate sets were inoculated, one set incubated in still culture and the other by 
shake culture. After an incubation period of 3 weeks, the above cultures were sub¬ 
cultured into new flasks containing the same media except for a decrease in the glu¬ 
cose content as is indicated in Table III. This process of subculturing into media 
in which the glucose concentration had been diminished to half of its preceding con¬ 
centration was repeated at 3-week intervals until the glucose concentration had been 
reduced to 0, after which time lignin remained as the limiting source of carbon in one 
of the media. 

The pH of the media and dry weight of the mycelium was determined for each 
culture flask after each 3-week incubation period by the technique described under 
‘^Materials and Methods.” The mycelial weight determinations are recorded in 
Table III. The pH values are not recorded, as no significant trend was noted. 

From the results presented in Table III, it can be seen that cultures 
No. 13, Polyporus abietinus, and No. 14, Poria subacida , showed evi¬ 
dence of increased ability to utilize lignin during the course of 5 serial 
subcultures in the presence of lignin. Growth in the medium with glu¬ 
cose alone decreased with each decrease in glucose concentration. How¬ 
ever, in the medium containing a constant amount of lignin (0.5%). 
growth actually showed some increase, even though the glucose content 
was decreased and finally omitted entirely.,/ 

Cultures No. 9, Pomes geotropus, No. 11, Lentinus tigrinus , and No. 
26, Peniophora gigantea , showed no evidence of adaptation to lignin 
utilization under the conditions of the experiment. After the fourth 
serial subculture, growth was practically negligible in both the medium 
with 0.01% glucose and that containing 0.01% glucose and 0.5% lignin. 
Culture No. 17, Daedalea unicolor , showed little if any increase in 
ability to utilize lignin. 

It was also evident from the results presented in Table III that the 
total mycelial yield was higher for all cultures in the media containing 
glucose (at the higher concentrations) plus lignin than in the corre¬ 
sponding medium containing glucose alone.. 

During the course of the adaptation experiments, subcultures wen 
made from the liquid media onto 4 solid media of the same basic compo¬ 
sition. This was done to obtain additional evidence for the degree oi 
growth response to lignin after the culture had been in contact with 
lignin for various periods of time. It was found that cultures No. 13 
Polyporus abietinus , and No. 14, Poria subacida , produced a large 
mycelial mat on the lignin containing agar while the other 4 cultures 
grew very sparsely on the same medium. 



368 


W. C. DAY, M. J. PELCZAR, JR. AND S. GOTTLIEB 


Comparisons of the response of cultures No. 13, Polyporus abietinus. 
and No. 14, Poria subadda (adapted and unadapted) to lignin were 
made by inoculating each of these cultures onto a solid medium which 
contained lignin as the limiting source of carbon. It was clearly demon¬ 
strated that the cultures which had been grown in the presence of lignin 
(inoculum taken from 3rd serial subculture—Table III) had developed 
a considerable ability for utilization of lignin, as evidenced by the large 
mat of mycelium which developed on the lignin medium. The media 
inoculated with the unadapted culture (inoculum taken from growth 
on potato dextrose-agar slant) produced very sparse growth. 

Response to Chemically Sterilized Lignin and Autoclaved Lignin 

It has been stated that the lignin for incorporation into the medium 
was sterilized in an aqueous suspension by autoclaving. However, jt 
was possible to sterilize the lignin by exposure to ether as described 
under “Materials and Methods.” Growth response to the chemically 
sterilized lignin was the same as was observed with autoclaved lignin. 

Specificity of Growth Response to Lignin 

To eliminate the possibility that the growth response observed and 
interpreted as lignin utilization was not due to impurities, the lignin 
was subjected to further purification, and growth response determined 
on the resulting product. The original lignin was repurified by dissolving 
in-dioxane, reprecipitated in ether, and washed in ether, benzene, and 
petroleum ether. This process was repeated 10 times. The final product 
was designated repurified native lignin. The original and repurified 
lignin were then separately incorporated into solid media as the sole 
source of carbon and inoculated with cultures No. 13, Polyporus 
abietinus, and No. 14, Poria subadda. The inoculum was taken from the 
3rd serial subculture flasks containing lignin (Table III). It was ob¬ 
served that growth response did not differ on either type of lignin. It 
was, therefore, concluded that the stimulation of growth was due 
specifically to the lignin. 

Summary 

Thirty-five cultures of white rot fungi were screened for their ability 
to utilize native lignin. Under the experimental conditions described, 
many of these cultures exhibited growth response when lignin was the 



BIOLOGICAL DEGRADATION OF LIGNIN. 1 


369 


limiting source of carbon in the medium. However, the results from 
these screening experiments could not be clearly evaluated because of 
variations in growth on duplicate determinations and other irregulari¬ 
ties in growth response. 

As a consequence, 6 of the cultures were selected and an attempt 
made to adapt them to lignin utilization by growing the cultures for a 
prolonged period in a lignin-glucose medium, gradually diminishing the 
glucose content until it reached zero. By this approach, it was possible 
to obtain two cultures, No. 13, Polyporus abietinus , and No. 14, Poria 
subacida , which consistently produced an appreciable amount of growth 
on a solid or liquid medium in which lignin was the limiting source of 
carbon. Both of these media were of chemically defined composition 
except for the agar in the solid medium. 

The availability of such a technic for rapid growth of fungi on a 
synthetic medium containing lignin provides for a more rigorous and 
rational approach to the mechanism of biological degradation of lig¬ 
nin than has been hitherto possible. 

References 

1. Waksman, S. A., and Nisson, W., Science 74, 271 (1931). 

2. Lindeberg, G., Symbolae Botan. Upsalienses 8 , 5 (1944). 

3. Bose, S. R‘, and Sarkar, S. N., Proc. Roy. Soc. London 123B, 193 (1937). 

4. Pringsheim, H., and Fuchs, W., Ber. 66, 2095 (1923). 

5. Waksman, S. A., and Hutchings, I. J., Soil Set. 42, 119 (1936). 

6. Garren, K. H., Phytopathology 28, 875 (1938). 

7. Brauns, F. E., J. Am. Chem. Soc. 61, 2120 (1939). 

8. Brauns, F. E., J. Org. Chem. 10, 211 (1945). t 

9. Robbins, W. J., and Kavanagh, F., Am. J. Botany 25, 229 (1938). 



Studies on the Permeability of Erthrocytes. I. The 
Relationship between Cholinesterase Activity and 
Permeability of Dog Erthrocytes 1 

Margaret E. Greig and William C. Holland 2 

From the Department of Pharmacology, Vanderbilt University 
School of Medicine, Nashville, Tennessee 
Received March 14, 1949 

Introduction 

Evidence that the internal ionic composition of the erythrocyte is 
regulated by metabolic processes has been supplied by a number of 
workers (9,11,20,25,26,45). Wilbrandt observed that the addition of 
NaF or sodium iodoacetate is followed by a large loss of potassium 
from the mammalian red cell. This effect he attributed to the inhibitory 
action of these drugs on glycolysis. Le Fevre provided evidence that, in 
the case of hemolysis of erythrocytes by glycerol or glucose,»some phos¬ 
phorylation mechanism is involved which is sensitive to sulfhydryl 
agents. Apart from these suggestions, no specific metabolic processes 
seem to have been considered. 

• In some experiments on the matabolic effects of methadon, 3,4 we 
observed that, following the intravenous administration of this drug, 
dogs occasionally showed a hemoglobinuria and frequently an increase 
in red cell fragility (35). In experiments in vitro methadon also caused 
hemolysis of dog erythrocytes. We had previously found that methadon 
inhibited the glycolysis of glucose by brain but had no effect on the 
glycolysis of glycogen or of phosphorylated hexoses (18). The permea¬ 
bility changes in erythrocytes produced by methadon might be ex¬ 
plained by the inhibitory action of this drug on glycolysis as postulated 

‘Funds for carrying on this work were kindly supplied by the Mallinckrodt 
Chemical Works. 

* Research Fellow, U. S. Public Health Service. 

* Kindly supplied by Mallinckrodt Chemical Works. 

4 Methadon or amidone, 2-dimcthylnmino-4,4-diphenylheptanone-5-hydrochloride. 


370 



STUDIES ON PERMEABILITY OP ERYTHROCYTES. I 


371 


by Wilbrandt for the action of NaF and sodium iodoacetate. However, 
as we used washed cells, it is unlikely that there would be appreciable 
glucose in the medium. Methadon also inhibited cholinesterase (13,14, 
19) as do NaF and sodium iodoacetate (30). Red cells are a rich source 
of cholinesterase (2), which is situated in the cell membrane or stroma 
(7,33). It seemed possible, on the basis of certain considerations to be 
described later, that the acetylcholine-cholinesterase system instead of, 
or in addition to, the enzymes involved in glycolysis might be con¬ 
cerned with permeability of the red cell. To test this hypothesis further, 
the effect of physostigmine, a specific inhibitor of cholinesterase, on 
dog erythrocytes was investigated. 

Methods 

Cholinesterase activity was determined both manometrieally, using the Warburg 
technique, and by titration to constant pH with iV/200 NaOH, the pH being deter¬ 
mined by use of the Coleman pH meter. Both determinations were carried out at 37°C. 
in the media described in the tables. 

Freshly drawn blood was defibrinated, centrifuged, and the red cells washed with 
isotonic buffer or saline. In some experiments 0.5 ec. of a 50% suspension, in others, 
0.2 cc. of packed red cells were used. 

The buffers used were the following: Ringer-Krebs bicarbonate buffer (24); 
bicarbonate saline consisting of 100 parts isotonic NaCl or IvCl and 21 parts isotonic 
NaHCOj or KIICO 3 ; phosphate saline buffer consisting of 100 parts isotonic NaCl 
or KC1 and 21 parts isotonic phosphate (mixtures of dibasic and monobasic phos¬ 
phates in the proportions specified by Sprensen (38) to give the desired pH). 

The concentration of acetylcholine used in the experiments was 10~ 2 M. The con¬ 
centration of methadon was 8.7 X 10 -4 M , and that of physostigmine varied between 
7.2 X 10~* M and 3.6 X 10 “ 8 M y as indicated in the tables and graphs. It was assumed 
that, at pH 7 and higher, acetylcholine bromide and methadon hydrochloride were 
completely ionized, and on the basis of these assumptions stock solutions which w'ere 
isotonic with plasma w'ere made. Physostigmine was used in the form of the free base 
and the stock solution w f as also made isotonic with plasma. 

In some experiments the fragility of red cells was determined by Sanford’s method 
(37), in which the red cell suspension was added to solutions of varying concentrations 
of NaCl. In other experiments, the degree of hemolysis w r as determined by removing 
samples from the experimental flask at intervals throughout the experiment, centri¬ 
fuging, and determining the hemoglobin in the supernatant fluid by use of a photo¬ 
electric colorimeter. 

Results 

In the investigation described, the blood of 7 different dogs was used. 
The results of only a few experiments, which are representative of a 
much larger number, are presented. Isotonic solutions were used 
throughout. 



372 


MARGARET E. GREIG AND WILLIAM 0. HOLLAND 


In preliminary experiments, the effects of methadon and of physostigmine on red 
cell cholinesterase activity were determined manometrically in a sodium, potassium, 
or Ringer-Krebs’ bicarbonate buffer at pH 7.4 with 95% Nr-5% C0 2 as the gas 
phase. At the end of the experimental period, fragility tests by Sanford's method 
showed that methadon always increased fragility while physostigmine had little, or a 
variable, effect (Table I). It was observed, however, that, if the experimental flasks 
were allowed to stand at room temperature for 2-3 hr., complete hemolysis of the 
physostigmine-treated cells frequently occurred, while hemolysis was considerably less 
in the methadon-treated suspension. 


TABLE I 

Effect of Methadon and of Physostigmine on Dog Erythrocyte Cholinesterase 

Each Warburg vessel contained 0.5 cc. of a 50% suspension of erythrocytes and 
acetylcholine M/100 in a final volume of 2 cc. The media were: 1. Ringer-Krebs 
bicarbonate buffer; 2. 100 parts isotonic KC1, 21 parts isotonic KHC0 3 . 






111m. 3 CO2 evolved 

» 

Per cent NaCl in which hemo¬ 
lysis began and was complete 

Dura¬ 

tion 

Me¬ 

dium 

Drug 

Cone, of 
drug 

Con¬ 

trol 

With 

drug 

Inhibi¬ 

tion 

Control 

With drug 





Begun 

Com¬ 

plete 

Begun 

Com¬ 

plete 

win. 

95 

1 

Methadon 

M 

8.7 X 10-" 

238 

197 

per cent 

17 

.54 

.40 

.66 

.52 

95 

2 

Methadon 

8.7 X 10~< 

267 

238 

11 

.66 

.40 

.66 

.58 

60 

1 

Methadon 

8.7 X 10-* 

159 

104 

35 

.44 

.34 

.66 

.44 

50 

1 

Physostigmine 

7 X 10-* 

229 

77 

66 

.52 

.34 

.52 

.34 

60 

1 

Physostigmine 

7 X 10-‘ 

257 

28 

89 

.52 

*46 

.54 

.38 

60 

1 

Physostigmine 

7 X 10-« 

257 

130 

49 

.52 

.46 

.50 

.46 

60 

1 

Physostigmine 

.7 X 10-» 

257 

210 1 

19 

! 

.52 

.46 

.52 

.44 


These seemingly contradictory results might be attributed (1) to a discrepancy 
between the results of the Sanford test for fragility and the degree of hemolysis 
actually occurring in the experimental medium, or (2) to the increased pH caused by 
the escape of CO* from the medium, or (3) to physostigmine having a delayed effect 
but causing rapid hemolysis after the onset. To test these possibilities, large scale 
experiments were carried out in Erlenmeyer flasks, and 1 cc. aliquots were removed at 
intervals, diluted with isotonic solutions, centrifuged, and the free hemoglobin deter¬ 
mined in the supernatant fluid colorimetrically. It was found that fragility, as deter¬ 
mined by the Sanford method, did not necessarily parallel the hemolysis produced in 
the experimental flask. When KC1 was substituted for NaCl in the Sanford test, 
different results were obtained. It proved more satisfactory to determine the hemo¬ 
globin directly after centrifugation of a sample of the experimental suspension. 



STUDIES ON PERMEABILITY OF ERYTHROCYTES. I 


FIGURE I 

NaCI 



FIGURE 2 





374 


MARGARET E. GREIG AND WILLIAM C. HOLLAND 


FIGURE 3 



TIME IN MINUTES 


FIGURE 4 


10% NflCl + 90% KCI 



TIME IN MINUTES 




PER CENT HEMOLYSIS 


STUDIES ON PERMEABILITY OF ERYTHROCYTES. I 


FIGURE 5' 



TIME IN MINUTES 

FIGURE 6 


5 X N0HCO3 + 95% KHCOj 



TIME IN MINUTES 










STUDIES ON PERMEABILITY OF ERYTHROCYTES. I 


377 


The factors which influenced the changes in permeability produced 
by physostigmine and methadon to the greatest extent were found to 
be the ionic composition of the medium, the pH, and the presence or 
absence of acetylcholine. 

Effect of Cations and of Acetylcholine 

Na + , K + -Methadon. In NaCl, and in mixtures of isotonic solutions of 
NaCl and KC1 in proportions varying between 90% and 10% NaCl, 
methadon increased the hemolysis of dog erythrocytes (Fig. 1). The 

FIGURE t 

95% K PHOSPHATE SALINE + 5% NoCI 
p H 7.8 AND 8 



Figs. 1-9. One cc. packed erythrocytes in saline or buffer, either alone or with 
acetylcholine or the cholinesterase inhibitor, or with both of these drugs, were incu¬ 
bated in a water bath at 37°C. Final volume 10 cc. Final concentration of acetylcho¬ 
line 0.01 M. Concentration of cholinesterase inhibitors is indicated on the graphs. 
Symbols on the graphs denote isotonic solutions, e.g., 90% KCl-10% NaCl signifies 
isotonic solutions of KC1 and NaCl in the proportion 9:1. 


378 


MARGARET E. GREIG AND WILLIAM C. HOLLAND 


TABLE II 

Effect of No and of K Ions on the Inhibition by Physostigmine of Cholinesterase 

Buffers consisting of either 100 parts isotonic NaCl and 21 parts isotonic NaHCOi or 
KC1 and KHCOj in the same proportions, or mixtures of these two buffers were used. 
The concentration of acetylcholine was 0.01 M. 


Expt. 

Duration 1 

Cone, of 

Cone, of 

mm. 3 CO 2 evolved 

Effect 

Na 

K 

physostigmine 

Control 

With drug 

164 

min. 

90 

per cent 

95 

per cent 

5 

M 

7X10-' 

223 

86 

per rent 
— 61 



95 

5 

7X10' 7 

223 

223 

0 



95 

5 

7X10- 8 

223 

223 

0 



5 

95 

7X10- 8 

256 

79 

-76 



5 

95 

7X10' 7 

256 

151 

-41 



5 

95 

X 

o 

256 

223 

-13 

162 

50 

100 

_ 

7X10- 6 

178 

45 

-75 



95 

5 

7Xl0-« 

151 

39 

-74 



— 

100 

7X10-8 

151 

0 

-100 

176 

50 

100 

_ 

3.6X10* 

129 

24 

-82 





3.6 X10- 7 


66 

-49 



50 

50 

3.6X10-8 

100 

35 

— 65 





3.6 X10- 7 


82 

-17 

178 

60 

100 

— 

3.6X10 8 

115 

149 

+29.6 



— 

100 

3.6X10 8 

124 

100 

-20 


i_ _ J 

— 

100 

3.6X10- 9 

i 

124 

121 

0 


composition of the medium in which maximum hemolysis occurred 
varied somewhat in different experiments, but the time at which the 
maximum effect of methadon was produced was fairly constant. In 
3 experiments, in which bloods of 3 different dogs were used, the times 
at which maximum hemolysis occurred were 165, 165 and 180 min. 
from the start of the experiment (Fig. 2). In KC1 solutions, and occa¬ 
sionally in solutions containing 90% KC1 and 10% NaCl, methadon 
produced an increase in resistance of dog erythrocytes (Figs. 3,4). 

N& + , K + -Physostigmine. In NaCl solution, physostigmine caused 
only slight increases in hemolysis, and these effects occurred late in the 
experiment (around 150-200 min.). As the concentration of K was 
increased, the permeability of the physostigmine-treated suspension 





STUDIES ON PERMEABILITY OF ERYTHROCYTES. I 370 

deviated more and more from the control, and in KC1 as well as in 5% 
or 10% NaCl in KC1, the physostigmine-treated erythrocytes were 
more resistant than were those treated with methadon (Fig. 4). 

Acetylcholine. In KC1, in bicarbonate buffer, or in phosphate-saline 
buffer, the addition of acetylcholine increased the resistance of red 
cells and at the same time increased the differences in the degrees of 
hemolysis between the control and the drug-treated cells (Figs. 5,6,7,8). 
Under these conditions, physostigmine caused increased hemolysis of 
erythrocytes, which is in contrast to the effect produced by this drug 
in'KCI without acetylcholine, where the cells were made more resistant 
(Fig. 8). In bicarbonate buffer, physostigmine in concentrations 
between 7.3 X 10~ 4 M and 7.3 X 10 -7 M produced a maximum rela¬ 
tive hemolytic rate in about 90 min. 

Effect of pH 

While methadon produced changes in resistance of red cells over a 
wide range of Na:K concentrations, the physostigmine effect, in experi¬ 
ments of short duration, occurred most markedly in a medium in which 
the proportion of K was high compared with that of Na. Likewise, 
while methadon was active over a wide range of pH the activity of 

TABLE III 

Effect of pH on Physostigmine Inhibition of Cholinesterase 

2 cc. packed erythrocytes, 2 cc. 10% acetylcholine and 22 cc. isotonic NaCl or KC1 
were incubated at 37°C., and titrated with 0.02 N NaOH. Physostigmine 9 X 10“ 8 M. 


Medium 

Duration of 
expt. 

pH 

1 

Control 

cc. NaOH 

With physo- 
stignune 

Spontaneous 
hydrolysis of 
acetylcholine 

Inhibition 


imn . 



# 


per cent 

NaCl 

20 

7.4 

1.86 

1.20 

.41 

46 



8.0 

3.10 

1.23 

.45 

72 

NaCl 

20 

7.4 

1.55 

1.0 

.41 

48 



8.0 

4.10 

1.7 

.45 

66 

. 

KOI 


7.2 


WEM 

.18 

25 


i 


1.13 

■fl 

.23 

70 



380 


MARGARET E. QREIG AND WILLIAM C. HOLLAND 


physostigmine increased quite markedly as the pH was increased 
(Fig. 9). 

At pH values of 7.8-8 physostigmine caused an increase in hemolysis 
in contrast to its effect in KC1, in which it produced an increased 
resistance (c/. Figs. 9,4). Thus, an increased resistance produced by 
physostigmine may be changed to an increased hemolysis either by 
increasing the pH or by the addition of acetylcholine. 

The increased effect of physostigmine in media of high pH and of 
high K content parallels the increased activity of cholinesterase under 
these conditions (Fig. 4) (Tables II, III). Also these conditions appear 
to be optimum for the inhibitory action of physostigmine (Tables II, 
III). 

In low concentrations, the inhibitory effect of physostigmine on 
cholinesterase and its effect on permeability wear off earlier than when 
higher concentrations are used. This may be due to destruction of the 
drug. 

Discussion 

The experiments described show that, under certain conditions, 
both methadon and physostigmine will produce changes in permeability 
of dog erythrocytes. As already stated, methadon inhibits the glyco¬ 
lysis of glucose in addition to cholinesterase activity, but, under the 
conditions of our experiments, in which washed cells were used, it is 
unlikely that there would be significant quantities of free glucose in the 
•medium. Any carbohydrate inside the cell would probably be glycogen 
or phosphorylated hexose, the metabolism of which we found to be 
unaffected by methadon. Effects of methadon on cell permeability 
under these conditions would not likely be connected with its action 
on carbohydrate metabolism. NaF and sodium iodoacetate, which 
Wilbrandt (44,45) showed could produce changes in permeability of 
erythrocytes, are both glycolytic inhibitors and inhibitors of cholines¬ 
terase. While glycolysis is sensitive to lower concentrations of NaF and 
sodium iodoacetate than is cholinesterase, the concentrations of these 
drugs required to produce changes in permeability, as found by Wil¬ 
brandt, were of the same order of magnitude as those which produce 
significant inhibitions of cholinesterase activity. Phosphoglyceral- 
dehyde dehydrogenase activity is inhibited by iodoacetate in concen¬ 
trations of Af/3000 (1). However, Wilbrandt found that changes in 
permeability of red cells occurred only when considerably higher con- 



STUDIES ON PERMEABILITY OP ERYTHROCYTES. I 


381 


centrations (M/500-M/100) were used. According to Nachmansohn 
and Lederer (30), cholinesterase is inhibited by iodoacetate in con¬ 
centrations of AT/100. Wilbrandt found that NaF in concentrations of 
around M/200-M/145 inhibited glycolysis but had little effect on 
permeability, while concentrations of M/145-M/72 produced changes 
in permeability. Massart and Dufait (29) used concentrations of M/50 
and M/100 NaF to inhibit cholinesterase 60 and 30%, respectively. 
It would thus seem possible that cholinesterase activity as well as 
glycolytic activity might be inhibited under Wilbrandt’s experimental 
conditions. It is also possible that specific inhibitors of glycolysis could 
affect permeability by virtue of their inhibiting the formation of 
adenosine triphosphate and of pyruvate, both of which may be used in 
the synthesis of acetylcholine (27,28,32). Physostigmine is considered 
a specific inhibitor of cholinesterase. Deutsch and Raper (12) found 
that, under certain conditions, physostigmine with acetylcholine in¬ 
creased the carbohydrate metabolism of slices of the cat’s submaxillary 
gland in vitro but had no inhibitory action. If similar changes in per¬ 
meability produced by both methadon and physostigmine are due to a 
single metabolic disturbance, it would seem that, of the known effects, 
the action on cholinesterase would provide the best explanation. 

Both methadon and physostigmine produced increased fragility and 
increased resistance of dog erythrocytes depending on the medium. In 
KC1 both produced increases in resistance. In NaCl methadon caused 
increased hemolysis while physostigmine had little effect on experiments 
of short duration. Whether increased fragility or increased resistance 
results, may also depend on the composition of the erythrocytes. The 
exact interpretation of these results, however, awaits experiments in 
which the action of physostigmine and methadon on the prolytic 
changes are determined. 

Methadon was active in causing hemolysis over a wider range of 
Na-K concentrations and of pH than was physostigmine, but in a 
medium of a high K concentration and a pH of about 8, physostigmine 
was a powerful hemolytic agent. The hemolytic activity of physostig¬ 
mine was increased by acetylcholine. It should be noted that the 
optimum conditions for hemolysis by physostigmine paralleled the 
optimum conditions for cholinesterase activity and, under these 
conditions, physostigmine also exerted its maximum inhibitory action 
(Tables II, III). The optimum conditions for the action of methadon 
on cholinesterase are being investigated and will be reported later. 



382 


MARGARET E. GREIG AND WILLIAM C. HOLLAND 


It might be of interest to note that certain other compounds which, although they 
undoubtedly have other actions, are known to be inhibitors of cholinesterase activity 
also influence the permeability of erythrocytes. One of these drugs, namely morphine, 
which is a hemolytic agent in vitro (5,23,34), an inhibitor of cholinesterase in vitro 
(4,46,47), and which has been reported to cause an anemia in vivo , has been investi¬ 
gated for its effect on glycolysis by brain in this laboratory (18). It was found that, in 
concentrations up to 5.6 X 10~ 3 M, it either had no effect or accelerated glycolysis. 
This being the case, its hemolytic action cannot be explained by its inhibiting the 
anaerobic metabolism of glucose. Lead, which produces a hemolytic anemia (6,22), 
causes marked changes of permeability of red cells in vitro (21) and is also a potent 
inhibitor of cholinesterase (16). Tetanus toxin, which produces a secondary (15) ane¬ 
mia in rabbits, is also an inhibitor of cholinesterase (17). 

In pernicious anemia, red cells have been reported to show an abnormal permea¬ 
bility (3). The erythrocyte cholinesterase is also low in this condition (36). 

Davis (10) reported the production of hyperchromic anemia in dogs following the 
administration of acetylcholine with physostigmine. 

There is some evidence that the acetylcholine-cholinesterase system 
may be concerned with the permeability of tissues other than red cells. 
Swan and Hart (40) showed that physostigmine, carbaminoyl choline 
and mecholyl increased the permeability of the blood-aqueous humor 
barrier to dyes, and Stocker (39) observed that miotics increased the 
permeability of the blood-aqueous barrier when instilled into the eye. 
Cumings (8) reported changes in serum and muscle potassium after 
the administration of prostigmine, an inhibitor of cholinesterase, to 
myasthenic patients. Thompson and Tice (41) observed that prostig¬ 
mine produced a decrease in serum potassium in dogs and cats and an 
elevation of muscle potassium in rats. 

That the acetylcholine-cholinesterase system is concerned with 
changes in permeability in the nerve cell may be postulated from the 
work of von Muralt (42) and of Nachmansohn and collaborators 
(30,31,32), who have provided considerable evidence that the passage 
of an impulse along a nerve is dependent on the rapid formation or 
liberation and removal of acetylcholine within the cell. This being the 
case, the change in permeability which accompanies the change in 
potential could also be explained by the formation and removal of 
acetylcholine. 

Welch (43), in a discussion of the function of acetylcholine and cho¬ 
linesterase in nerve transmission, has suggested that acetylcholine may 
be a coenzyme which is concerned with the activity of an enzyme in the 
cell membrane, and that the role of this enzyme is to alter excitability 
of the cell through changes in membrane polarity. Such an idea might 
also be applicable to red cell permeability. 



STUDIES ON PERMEABILITY OF ERYTHROCYTES. I 


383 


Investigations of the effect of cholinesterase inhibitors on the prolytic 
exchange of erythrocytes are being undertaken. 

Summary 

1. Evidence is presented that changes in the permeability of the 
erythrocyte may be effected by inhibition of the activity of the acetyl- 
choline-cholinesterase system which is situated in the cell membrane. 

2. Both methadon and physostigmine, which are inhibitors of 
cholinesterase, produce changes in permeability of dog erythrocytes. 

3. The effect of these drugs on permeability is influenced by the 
sodium and potassium content of the medium, the pH, and the pres¬ 
ence of acetylcholine. 


References 

1. Abler, E., v. Euler, II., and Gunther, G., Skarui. Arch. Physiol. 80, 1 (1938). 

2. Augustinsson, Iv. B., Acta Physiol. Scand. 15, Suppl. 52 (1948). 

3. Bang, O., and Orskov, S. L., J. Clin. Invest. 16, 279 (1937). 

4. Bernheim, F., and Bernheim, M. L. C., J. Pharmacol. Exptl. Therap. 57, 427 

(1936). 

5. Bonanno, G., Gazz. ospedali e din. 28, 907 ; Biochern. Zentr. 8, 267 (1909) ; Chem. 

Abstracts 3, 2317 (1909). 

6. Bouchard, M., Compt. rend. soc. biol. 5, 358 (1873) (quoted from Aub, J. C\, 

Fairhall, L. T., Minot, A. S., and Reznikoff, P., Lead poisoning, Williams 
and Wilkins, 1926). 

7. Brauer, R. W., and Root, M. A., Federation Proc. 4, 1131 (1945). 

8. Cumings, J. N., J. Neurol. Psychiat. 3, 115 (1941). 

9. Danowski, T. S., J. Biol. Chem. 139, 693 (1941). 

10. Davis, J. E., Am. J. Physiol. 147, 404 (1946). 

11. Davson, H., and Danielli, J. F., Biochem. J. 32, 991 (1938). 

12. Deutsch, W., and Raper, H. S., J. Physiol. 87, 275 (1936); 92, 439 (1938). 

13. Eadie, G. S., and Fitzgerald, D. B., Federation Proc. 7, 1 (1948). 

14. Eadie, G. S., Bernheim, F., and Fitzgerald, D. B., J . Pharmacol. Exptl. 

Therap. 94, 19 (1948). 

15. Farkas, H., and Kligler, I. J., Proc. Soc. Exptl. Biol. Med. 48, 717 (1941). 

16. Frommel, E., Herschberg, A. D., and Piquet, J., Helv. Physiol. Pharmacol. 

Acta 2, 169 (1944). 

17. Genuit, H. J., and Labenz, K., Arch. Exptl. Path. Pharmakol. 198, 369 (1941). 

18. Greig, M. E., Arch. Biochem. 17, 129 (1948). 

19. Greig, M. E., and Howell, R. S., Proc. Soc. Exptl. Biol. Med. 68 , 352 (1948). 

20. Harris, J. E., J. Biol. Chem. 141,579 (1941). 

21. Hknriques, V., and Orskov, S. L., Skand. Arch. Physiol. 74, 78 (1936). 

22. Heubel, E., quoted from Aub, J. C., Fairhall, L. T., Minot, A. S., and Rez¬ 

nikoff, P., Lead Poisoning, 140. Williams and Wilkins Co., 1926. 

23. Koeppe, H., Arch . ges. Physiol. (. PflUger y s ) 99, 33 (1903). 



384 


MARGARET E. GREIG AND WILLIAM C. HOLLAND 


24. Krebs, A. A., and Henseleit, K., Z . physiol. Chem. 210, 33 (1932). 

25. Le Fevre, P. G., J. Gen. Physiol. 31, 505 (1948). 

26. Le Fevre, P. G., Biol. BuU. 93, 224 (1947). 

27. Lipton, M. A., and Barr6n, E. S. G., J. Biol. Chem. 166, 367 (1946). 

28. Mann, P., Tenennbaum, J. G., and Quastel, J. H., Biochem. J. 33, 823 (1939). 

29. Massart, L., and Dufait, R., BuU. soc. chim. bid. 21,1039 (1939). 

30. Nachmansohn, D., and Lederer, E., ibid. 21, 797 (1939). 

31. Nachmansohn, D., Vitamins and Hormones 3, 337 (1945). 

32. Nachmansohn, D., Currents in Biochemical Research. Interscience Pub., Inc., 

N. Y., 1946. 

33. Pal£us, S., Arch. Biochem. 12, 153 (1947). 

34. Rhode, H., Biochem. Z . 131, 560 (1922). 

35. Ross, P., and Greio, M. E., unpublished experiments, 1948. 

36. Sabine, J. C., J. Clin . Invest ., 19, 833 (1940). 

37. Sanford, A. H., Clinical Diagnosis by Laboratory Methods. W. B. Saunders and 

Co., Philadelphia, 1948. 

38. Sorensen, Hydrogen Ion. A. T. S. Britton. D. Van Nostrand, N. Y., 1929. 

39. Stocker, F. W., Arch. Ophthalmol. 37, 583 (1947). 

40. Swan, K. C., and Hart, W. M., Am. J. Ophthalmol 23, 1311 (1940). 

41. Thompson, V., and Tice, A., J. Pharmacol. Exptl. Therap. 73, 455 (1941). 

42. Von Muralt, A., Vitamins and Hormones 5, 93 (1947). 

43. Welch, J. W., BuU. Johns Hopkins Hosp. 83, 568 (1948). 

44. Wilbrandt, W., Arch. ges. Physiol. (Pfluger’s) 243, 519 (1939-40). 

45. Wilbrandt, W., Trans. Faraday Soc. 33, 956 (1937). 

46. Wright, C. I., and Sabine, J. C., J. Pharmacol Exptl. Therap. 78, 375 (1943). 

47. Wright, C. I., and Sabine, J. C., ibid. 93, 230 (1948). 



Studies on Cell Enzyme Systems. II. Evidence for 
Enzyme-Substrate Complex Formation in the 
Reaction of Cypridina Luciferin and Luciferase 1 

Aurin M. Chase 

From the Physiological Laboratories, Princeton University, New Jersey, and the 
Marine Biological Laboratory, Woods Hole, Massachusetts 
Received February 23, 1949 

Introduction 

Most studies on enzyme kinetics and equilibria have been carried out 
with hydrolytic enzymes by measureing the amount of substrate 
disappearing or reaction products which appear during given intervals 
of time. Oxidative enzyme reactions have received less attention, 
despite their importance as energy-controlling processes, largely be¬ 
cause the process frequently occurs in a complex series of steps. 

The enzyme-catalyzed oxidative bioluminescent reactions, which are 
probably in some forms related to cellular respiratory processes, are 
unique in that the intensity of the emitted light is a direct, instantaneous 
measure of the velocity of the underlying reaction. Moreover, lumines¬ 
cence measurements are easily made and can, if necessary, be auto¬ 
matically recorded. The ostracod crustacean, Cypridina hilgendorfii, 
is one of a few luminous organisms from which the enzyme and sub¬ 
strate of the reaction can be readily extracted and separated (10). If 
the animals are dried when collected and kept dry, the luciferin and 
luciferase remain stable apparently indefinitely. Furthermore, they are 
stored in such quantity in this animal that they can be subjected to 
purification procedures. It is thus possible to study the reaction in 
vitro, using relatively pure compounds and controlled conditions. This 
reduces the effects of unknown complicating factors. 

The theory of an intermediate enzyme-substrate complex, derived 
and tested by Michaelis and Menten (13) for the hydrolysis of sucrose 

1 This work was supported by an institutional grant for fundamental research from 
the New Jersey Section of the American Cancer Society to the Biology Department of 
Princeton University. 


385 



386 


AURIN M. CHASE 


by invertase, has often been applied in studies on enzyme-catalyzed 
reactions. It yields a numerical value of the equilibrium constant for the 
dissociation of the postulated enzyme-substrate complex. Haldane (9) 
presents a compilation of the so-called Michaelis constants for a large 
number of reactions. His table shows, in general, constants of the order 
of magnitude of 10 -2 for hydrolytic reactions and 10~ 6 for oxidative 
reactions, although exceptions to this general rule occur. 

An analysis of the reaction of Cypridina luciferin and luciferase by 

Moelwyn-Hughes (14) did not give satisfactory agreement with the 

Michaelis-Menten theory of intermediate enzyme-substrate complex 

formation. Because the data used were from studies involving crude 

extracts of Cypridina luciferin and luciferase and for other reasons, 2 

new measurements have been made with the more highly purified 

luciferin and luciferase now available. These new data show the effect 

* 

upon the velocity of the luminescent reaction of varying the luciferin 
concentration over a range of 1:200. They have been analyzed in terms 
of the Michaelis-Menten theory and yield a value for the Michaelis 
constant which is not inconsistent with other values for enzyme-cata¬ 
lyzed systems. 

Apparatus and Methods 

Measurements of the luminescent reaction were made with a modification of the 
apparatus described by Anderson (2), which determines total emitted light. In this 
way very dim light production can be accurately measured, since the instantaneous 
intensity is not recorded but, rather, the amount of light produced during the interval 
frpm the start of the reaction to any particular time. When total light i^ plotted against 
time from initiation of the reaction, the slope of the first part of the resulting curve is 
directly proportional to the initial velocity. 

The luciferase used was an extract of 5 g. of dried, powdered Cypridina organisms 
in 100 ml. of distilled water. This was filtered and subjected to prolonged dialysis, 
first against running tap water, and finally against several changes of distilled water. 

*The data used by Moelwyn-Hughes were those of Harvey and Snell (11), who 
studied the effect of total luciferin concentration (luciferin plus oxyluciferin) upon the 
velocity constant of the reaction, using the crude aqueous extracts of luciferin and 
luciferase which were at that time available. For reasons unknown at present the veloc¬ 
ity constant obtained under such conditions varies with substrate concentration as 
well as with the concentration of enzyme. It decreases as the substrate concentration is 
increased, and approaches a minimum value. This is not true for the reaction of the 
more highly purified Cypridina luciferin and luciferase now available, for which it has 
been shown (8) that the velocity constant is practically independent of luciferin 
concentration and varies only with the concentration of enzyme, as would be expected. 
However, Moelwyn-Hughes’ use of velocity constants, rather than velocities, would 
appear to be unjustified on purely theoretical grounds. 


ENZYME-SUBSTRATE COMPLEX FORMATION 


387 


This treatment removed much dialyzable material and also caused the precipitation of 
a considerable quantity of inactive protein. The final product from this dialysis, 
suitably diluted, served as the enzyme stock solution. 

The luciferin, extracted also from dried, powdered Cypridina organisms, was carried 
through two cycles of purification by the method of Anderson (3). The degree of purity 
of the product from this procedure, while not actually known, is very high compared 
with that of luciferin in a crude, aqueous extract of the organisms. 3 

Since the hydrogen ion concentration and that of other ions greatly affect this 
luminescent reaction (2,4,6) the reaction mixture used in these experiments always 
consisted of equal volumes of A//15 KH 2 PO 4 and M /15 Na 2 HPO* and was 0.01 M for 
NaCl. 

The luciferin obtained from Anderson’s purification procedure is in hydrogen-sat¬ 
urated w-butyl alcohol solution, stored under an atmosphere of hydrogen. Preliminary 
experiments showed that this alcohol has an inhibitory effect on the activity of the 
enzyme, luciferase. It was, therefore, necessary to use some other solvent for the 
luciferin since the substrate concentrations were to bo varied in the experiments by 
taking different volumes of luciferin stock solution. 0.1 A HC1 was chosen as the sol¬ 
vent. Luciferin stock solution for a day’s experiments was prepared as follows. Two 
ml. of the butanol-luciferin solution were transferred from the storage vessel to a small 
vial. By means of a vacuum desiccator, a liquid nitrogen freezing trap and a vacuum 
pump, the alcohol was removed from the solution, leaving a solid residue. This was 
dissolved in 5 ml. of 0.1 A HC1 and the resulting solution put into a small test tube 
imtnersed in an ice water bath. This last precaution is necessary because luciferin 
undergoes spontaneous oxidation in the presence of dissolved oxygen, losing the prop¬ 
erty of giving light with luciferase, and this oxidation is greatly retarded at low 
temperatures. 

For an individual luminescence measurement the desired volume of luciferin-HCl 
stock solution (0.02-2.00 ml.) was placed in the bottom of the reaction vessel. Next, 
enough 0.1 A 1IC1 was added to make the total volume of acid present (including that 
added with the luciferin) 2.00 ml. This was done to keep the pH and chloride ion con¬ 
tent constant in all experiments. Ten ml. of the reaction mixture w r ere next added and 
the reaction vessel was placed in position in the light-measuring apparatus. Another 
10 ml. volume of reaction mixture containing the luciferase was then run in from a 
fast-flowing pipette, starting the luminescent reaction. Total emitted light was re¬ 
corded at intervals for as long a period as desired. 

Experimental Results 

Exploratory experiments with ordinary dilutions of the luciferase 
stock solution showed that a linear relationship apparently existed 
between luciferin concentration and velocity of the luminescent reac¬ 
tion, with no indication of a maximum velocity. Evidently a high con- 

3 In terms of quantity of light produced per unit dry weight of material, the lucif¬ 
erin used in the present work was about 1600 times as pure as that in the dry powder 
obtained by grinding the whole Cypridina organisms. 



388 


AURIN M. CHASE 


centration of substrate was required to saturate the enzyme. Rather 
than use higher concentrations of luciferin, which was prohibitive 
because of the limited supply of Cypridina available, the concentration 
of enzyme was greatly decreased. With a 1:600 dilution of the stock 
luciferase solution a typical velocity vs. substrate concentration rela¬ 
tionship was obtained. Luminescence was necessarily very dim at the 



TIME IN MINUTES 

Fig. 1 . Total light emitted during the first 2-3.5 min. after adding a constant, 
small amount of luciferase to various volumes of luciferin solution in a constant 
volume of reaction mixture. The straight lines represent the data during the early 
part of the reaction and their slopes are proportional to the initial velocity. The plotted 
values are from Table I. 

lowest luciferin concentrations, but it could still be accurately measured 
with the light-integrating apparatus. Typical data are plotted in Fig. 1 
and all numerical values are given in Table I. To avoid confusion in the 
figure, not all of the data have been plotted. Because of the very low 
luciferase concentration that was used, it was not practical to measure 
these luminescence reactions over their entire course but the initial 



ENZYME-SUBSTRATE COMPLEX FORMATION 


389 


velocities can be determined very precisely from the slopes of the 
straight lines which fit the data in the early part of the reaction. 4 

Inspection of Fig. 1 shows that the slopes of the lines (proportional 
to initial velocity) increase asymptotically toward a maximum value as 
the luciferin concentration is increased. The relationship between initial 


TABLE I 

Luminescence, in Arbitrary Unite (Millivolts), Obtained from a Constant Amount 
of Luciferase and Various Amounts of Luciferin Solution 

These are the data as directly measured. They are plotted in Fig. 1, with the ex¬ 
ception of those luminescence experiments involving 0.90, 1.80 and 2.00 ml. of luci¬ 
ferin solution. 


Total luminescence in millivolts for indicated volumes (in ml.) of luciferin solution; 
luciferase concentration constant 


Time 



0.02 



0.50 

0.75 

0.00 

i.OO 

1.50 

1.80 

2.00 

mm. 

0.25 

■ 

12 

18 

24 

32 

30 

37 

48 

35 

37 



20 


52 

58 

62 

70 

74 

70 

73 

0.75 

B 

28 


74 

90 

88 

98 

108 

100 

108 

1.00 

12 

38 

74 

102 

119 

116 

129 

143 

133 

140 

1.25 

— 

48 

88 

125 

150 

144 

157 

172 

164 

168 

1.50 

16 

57 

105 

148 

168 

171 

183 

200 

188 

199 

1.75 

— 

65 

121 

165 

194 

198 

214 

235 

218 

227 

2.00 

22 

74 

135 

186 

219 

220 

240 

260 

248 

259 

2.25 

— 

82 

152 

208 

245 

248 

266 

287 

275 

290 

2.50 

24 

90 

164 

232 

266 

266 

290 

314 

300 

316 

2.75 

— 

98 

182 

252 

292 

291 

312 

340 

328 

344 

3.00 

30 

107 

195 

268 

314 

314 

336 

367 

352 

368 

3.25 

— 

114 

208 

287 

335 

337 

358 

392 

376 

394 

3.50 

35 

120 

223 

310 

355 

358 

380 

415 

400 

420 


velocity of the luminescent reaction and luciferin concentration is, 
therefore, essentially the same as that encountered in general for 
enzyme-catalyzed processes. 

4 It will be observed that the straight lines of Fig. 1 do not pass exactly through the 
origin but slightly above it. This is probably because at the moment of mixing the 
luciferin and luciferase an initial flash of light occurs of complicated nature. This 
initial burst was first observed in the experiments of Amberson (1). It is small com¬ 
pared with the total light produced in the present experiments. 









390 


AtlRXN Mi CHASE 


Discussion 

Since the present data relating velocity of the Cypridina luminescent 
reaction to luciferin concentration appeared to conform to the Michael- 
is-Menten theory, an analysis was made in the following way. 

The usual form of the Michaelis-Menten equation is: 

V" 08 ) 

9 K, + (S) ’ 

where v = measured velocity of the reaction, V' = maximum velocity, 
(S) = substrate concentration and K, = the dissociation constant of 
the enzyme-substrate complex (i.e., the so-called Michaelis constant). 



Fig. 2. The data of Fig. 1 plotted in terms of the equation, (S)/v = ((S)/V') + 
t K,/V '), a linear modification of the Michaelis-Menten equation. Substrate concen¬ 
tration is expressed as molarity. The data are well fitted by a straight line, indicating 
that they obey the equation. The slope of this line is equal to the reciprocal of the 
maximum velocity, V', for the particular luciferase concentration used. The intercept 
is equal to the Michaelis constant divided by the maximum velocity. K„ calculated 
from the intercept, is 5.95 X 10~ 7 . 

The value of the maximum velocity must be known in order to cal¬ 
culate the Michaelis constant for the system but it was not possible, 
in the present case, to obtain this datum experimentally. Fortunately, 
this can be done graphically if the data are otherwise adequate. Line- 
weaver and Burk (12) have shown that if the reciprocal of the classical 




ENZYME-SUBSTRATE COMPLEX FORMATION 


391 


Michaelis-Menten equation be multiplied through by (S), the following 
linear equation is obtained: 

(S) = (S) , 

v V' V' 

(all terms having the same meanings as before). If the equation is 
satisfied, a straight line should fit the data when (S)/v is plotted 
against (S). The slope of this line is equal to the reciprocal of the maxi¬ 
mum velocity, and the intercept gives the value of K a /V'. Fig. 2 shows 
the data plotted in this form. The slope can be determined with con¬ 
siderable accuracy and yields a value for the maximum velocity of 143 
millivolts/min., for the particular luciferase concentration used. Sub¬ 
strate concentration is expressed in terms of molarity 6 in Fig. 2, and K s 
has a value 6 of 5.95 X 10~ 7 . 

The results of another series of luminescence experiments, in which a 
higher luciferase concentration had been used, gave a value for K„ of 
b.03 X 10~ 7 , not essentially different from that of the series with the 
lower enzyme concentration. Lack of space prohibits inclusion of the 
data from this second series of experiments but they are of the same 
quality as those shown in Figs. 1 and 2. 

It is of course recognized that K H may not represent the true dis¬ 
sociation constant of the luciferin-luciferase complex since, as Briggs 

6 The dry weight of luciferin used and its molecular weight must be known in order 
to express substrate concentration in terms of molarity. The molecular weight of 
Cypridina luciferin is not exactly known at present, although a determination of its 
combining weight was made by Chase (7), using a reaction between luciferin and 
ferricyanide. While an exact figure could not be obtained, his data indicated a com¬ 
bining weight somewhere between about 250 and 600. On the assumption that the 
combining weight as measured is equal to the molecular weight, and for purposes of 
calculating K g in the present case, 500 was taken as the molecular weight of luciferin. 
By weighing the dry residue from a 20 ml. volume of n-butanol-luciferin solution it 
was found that each ml. contained 6 X 10“ 8 g. of solid. Since, in the present experiments 
luciferin stock solutions were made by redissolving the residue from 2.0 ml. of butanol- 
luciferin solution in 5.0 ml. of 0.1 N HC1, each ml. of the HCl-luciferin stock solution 
would contain 2.4X10” 8 g. of solid. For purposes of calculation it was assumed that 
this solid is pure luciferin. These assumptions obviously affect the calculation of a 
precise value for K However, they do not particularly affect the order of magnitude 
of the constant. 

8 If, in the present case, substrate concentrations are expressed as percentage con¬ 
centration rather than in molarity, a value of 2.6 X 10~ 8 is obtained for K t . Some of 
the constants in the literature have been calculated in this way for reactions where the 
molecular weight of the substrate is quite unknown. 



392 


AURIN M. CHASE 


and Haldane (5) pointed out, the theory of Michaelis and Menten 
assumes the velocities of formation and dissociation of the enzyme- 
substrate complex to be infinitely great compared to the velocity with 
which the end-products of the reaction are split off from the inter¬ 
mediate complex. The theoretical treatment, therefore, ignores these 
first two velocity constants and considers the third only. This assump¬ 
tion may not be justified in all cases. However, taking the value of K, 
for the luminescent reaction at its face value, the luciferin-luciferase 
intermediate complex would appear to be similar in nature to those of 
oxidative enzyme systems in general, since such systems exhibit rela¬ 
tively low K, values as compared with hydrolytic enzyme systems. 
There seems to be little doubt that the Michaelis-Menten theory applies 
in the case of the luminescent reaction of partially purified Cypridina 
luciferin and luciferase under the existing experimental conditions.. 

The very good description of these data by the Michaelis-Menten 
equation also indicates that one molecule of luciferin combines with one 
luciferase molecule in the formation of the enzyme-substrate complex. 
This conclusion was further substantiated by plotting the data in terms 
of an equation representing the reversible formation of an enzyme- 
substrate complex involving 2 molecules of luciferin for each molecule of 
luciferase instead of the 1:1 ratio. The equation has the form: 

1 = K. J_ 1 

v v , '(S) i + r' 

If it describes the data, a straight line should be obtained when 1/v is 
plotted against 1/(S) 2 . The data, plotted in this way, could not be fitted 
by a straight line. It can, therefore, be concluded with some certainty 
that the intermediate enzyme-substrate complex for this reaction in¬ 
volves not more than one molecule of luciferin for each molecule of 
luciferase. 

AcKNOWLEDGM ENT 

It is a pleasure to acknowledge the capable technical assistance of Miss Betsy 
Brigham, who performed most of the experiments upon which this paper is based. 

Summary 

Enzyme-catalyzed oxidative bioluminescent reactions have the 
unique property that the intensity of the luminescence affords a direct, 
instantaneous measure of the velocity of the underlying process. 



ENZYME-SUBSTRATE COMPLEX FORMATION 


393 


The luminescent reaction of Cypridina luciferin and luciferase, both 
partially purified, was measured under conditions where, with constant 
enzyme concentration, the substrate concentration was varied over a 
200-fold range. The velocity of the reaction increases with increase of 
luciferin concentration and approaches a maximum value asymptoti¬ 
cally. 

The Michaelis constant, K„ representing the dissociation of an 
enzyme-substrate complex, was calculated and found to be about 
6 X 10~ 7 . This order of magnitude resembles those found for many 
oxidative enzyme systems. 

The analysis of the data indicated that not more than one molecule 
of luciferin combines with one molecule of luciferase in forming the 
enzyme-substrate complex. 


References 

1. Amberson, W. R., J. Gen. Physiol. 4, 517, 535 (1922). 

2. Anderson, R. S., J. Cellular Comp. Physiol. 3, 45 (1933). 

3. Anderson, R. S., J. Gen. Physiol. 19, 301 (1935). 

4. Anderson, R. S., J. Am. Chem. Soc. 69, 2115 (1937). 

5. Briggs, G. E., and Haldane, J. B. S., Biochem. J. 19, 338 (1925). 

6. Chase, A. M., J. Cellular Comp. Physiol. 31, 175 (1948). 

7. Chase, A. M., ibid. 33, 113 (1949). 

8. Chase, A. M., and Harvey, E. N., ibid. 19, 242 (1942). 

9. Haldane, J. B. S., Enzymes, pp. 35-37. Longmans, Green & Co., New York, 1930. 

10. Harvey, E. N., Am. J. Physiol. 42, 318 (1917). 

11. Harvey, E. N., and Snell, P. A., J. Gen. Physiol. 14, 529 (1931). 

12. Lineweaver, H., and Burk, D., J. Am. Chem. Soc. 66, 658 (1934). 

13. Michaelis, L., and Menten, M. L., Biochem . Z . 49, 333 (1913). 

14. Moelwyn-Hughes, E. A., Ergeb. Enzymforsch. 6, 23 (1937). 



Studies with Penicillinase in the Presence of Sulfonamides 1 

Georg Cronheim, Mary E. Baird and Warren N. Dannenburg 

From the Research Department of the S. E. MassengiU. Co., Bristol, Tennessee 
Received October 25, 1948 

Introduction 

Recent studies indicated that penicillin and sulfonamides in combi¬ 
nation may act synergistically in vitro and in vivo (2,4,8). To explain 
this synergistic action, it has been suggested (7) that penicillin reduces 
the number of bacteria to limits within which the sulfonamides become 
completely inhibitory. 

An investigation of penicillin blood levels in normal adults after 
varying doses of penicillin given orally together with sulfonamides and 
alkalizing buffer salts, showed that the values were fairly high and quite 
uniform (1). While the buffer salts (sodium citrate and sodium lactate) 
accounted to a large extent for the good utilization of the ingested 
penicillin the question arose whether the sulfonamide provided addi¬ 
tional protection, particularly against the destructive action of peni- 
•cillinase. 

Winnek (II) combined penicillin with phthalylsulfathiazole in rectal 
suppositories to prevent the growth of penicillinase-producing bacteria 
in the large bowel. From the penicillin blood levels obtained, it was 
concluded that the drug had been protected effectively. 

While it had been reported (5) that sulfadiazine sodium was without 
any effect on penicillinase, our experiments indicated a definite inter¬ 
action between sulfonamides and penicillinase. 

Experimental 

A weighed amount of penicillinase was dissolved in 25 ml. of distilled water, the 
sulfonamide was suspended, and the mixture heated to 37°C. Ten ml. of a penicillin 
solution pre-heated to 37°C. was added, and a sample taken immediately for peni¬ 
cillin assay. The mixture was kept at 37°C. and one of more samples withdrawn at 

1 Presented before the Division of Biological Chemistry of the American Chemical 
Society, Washington, September, 1948. 


394 



PENICILLINASE STUDIES 


395 


specified intervals. All samples were diluted with water to a theoretical penicillin 
content of 1 unit/ml. and assayed by the agar cup method using Staph, aureus (12). 
Control experiments without sulfonamides were always made simultaneously. 

The activity of the penicillinase (Takamine Laboratories) used in most of this 
study was determined for varying enzyme concentrations and two penicillin levels 
and for different times of exposure. Complete inactivation in 30 min. at 37°C. of 
10,000 or 50,000 units of penicillin required about 57 and 300 mg., respectively, of the 
enzyme. When the time of exposure is increased, smaller amounts of penicillinase will 
effect complete destruction of penicillin. 

TABLE I 

Inactivation of Penicillin by Varying Amounts of Penicillinase without and 
with Addition of a Mixture of 500 mg. Each of Microcrystalline 
Sulfadiazine and Microcrystalline Sulfathiazole 


Penicillinase 

10,000 Units penicillin 

Penicillinase 

50,000 Units penicillin 

After 30 min. incubation 8 

After 30 min. incubation 8 

Without sulfon¬ 
amides 

Sulfa mixture 
added 

Without sulfon¬ 
amides 

Sulfa mixture 
added 

mg 

zone of inhibit ion in mm. 

mu- 

zone of inhibition in mm. 

0 

7.5 

7.4 

0 

7.5 

7.7 

25 

7.2 

7.7 

100 

7.5 

7.8 

50 

4.0 

7.6 

150 

7.5 

7.6 

60 

0 

7.3 

200 

3.8 

6.3 

80 


7.2 

300 

0 

5.6 

100 


3.8 

400 


0 

150 


0 





® The zone of inhibition before incubation is the same in all experiments, averaging 
7.6 mm. 


After the activity of the penicillinase had been established, the effect of sulfona¬ 
mides was studied. In the first experiments, 500 mg. each of microcrystalline sulfa¬ 
diazine and microcrystalline sulfathiazole were suspended in the reaction mixture. 
With an exposure of 30 min., and with both 10,000 and 50,000 units of penicillin, the 
sulfonamides provided a marked protection against penicillinase (Table I). The same 
protection was obtained when the amount of enzyme was kept constant and the 
exposure was varied (Table II). The figures also indicate that it makes no difference 
whether sulfadiazine or sulfathiazole, or a mixture of these two drugs, is added. 
Sulfonamides alone do not give any zone of inhibition in the high dilution in which the 
solutions are used in the cup test. 

Since the zone of inhibition before incubation is the same in all experiments and 
equals that found before and after incubation without any penicillinase added, it can 
be concluded that the small amount of enzyme left after dilution for the cup test has 
no measurable effect on the penicillin determination. 



396 G. CRONHEIM, M. E. BAIRD AND W. N. DANNENBURG 


TABLE II 


Inactivation of Penicillin by Penicillinase in the Presence of Sulfonamides 
after Varying Times of Exposure at 87°C. a 


Time of incubation 

10,000 Units penicillin+25 mg. penicillinase 

Without sulfon- 
amides 

Sulfa mixture 
added* 

Sulfathiazole 

added 6 

Sulfadiazine 

added** 

min. 


zone of inhibition in mm. 


0 

7.2 

7.3 

7.4 

7.2 

30 

6.2 

7.2 

6.5 

6.8 

60 

4.1 

6.8 

5.8 

6.3 

90 

1.9 

6.3 

5.5 

5.9 

120 

0 

5.6 

5.0 

5.4 


a The zone of inhibition before incubation is the same in all experiments, averaging 
7.6 mm. 

6 500 mg. each of microcrystalline sulfadiazine and sulfathiazolc. 
c 1 g. of microcrystalline sulfathiazole. 
d 1 g. of microcrystalline sulfadiazine. 

The experimental data seemed at first to present a case of true enyzme inhibition by 
sulfonamides. The first indication that the results had to be explained by a different 
mechanism, however, was seen in the experiments with single sulfonamides (Table II). 
The protection by either sulfathiazole or sulfadiazine alone, or by a mixture of these 
two drugs, was practically the same, even though sulfathiazole is about 8 times more 
soluble in water at 37°C. than sulfadiazine. Furthermore, saturated solutions of one 
or mixtures of several sulfonamides did not provide any protection. However, all 

TABLE III 

Protection of Penicillin ( 10,000 Units) against Penicillinase by Varying 
Amounts of Microcrystalline Sulfathiazole 
Time of exposure: 30 min. at 37°C. 


Penicillinase 

Without sulfon¬ 
amide 

! 

Microerystalline sulfathiazole added 

i 

1 g. 

2 g. 

3g. 

mg. 


zone of inhibition in mm. 


25 

7.8 

7.4 



50 

4.5 

7.2 

6.8 

6.0 

100 

0 

5.2 

6.3 

5.0 

150 


0 

. 4.2 

3.0 

200 



0 

1.6 

250 




0 









PENICILLINASE STUDIES 


397 


observed facts were in agreement with the assumption that the penicillinase was 
adsorbed by the suspended sulfonamide particles and thus removed from the solution. 

To prove this assumption, the amount of microcrystalline sulfathiazole was changed 
and it was found (Table III) that larger quantities will provide full protection against 
increasing concentrations of penicillinase. Within the range of the experiments a 
direct relationship exists between the amounts of sulfathiazole and inactivated peni¬ 
cillinase. Since 1 g. of sulfathiazole is already far in excess of its solubility at 37°C. in 
the volume used, an increase to 2 or 3 g. will not change this situation. Thus, the 
protection against higher concentrations of penicillinase points to an adsorption of the 
enzyme by sulfathiazole particles. 

A comparison of microcrystalline sulfathiazole (particle size about 5 m) with regular 
sulfathiazole (50 to 100 y) showed (Table IV) that the protective action of regular 
sulfathiazole is markedly less than that of the microcrystalline form. Since the only 
difference is in the particle size of sulfathiazole, the conclusion seems justified that the 
penicillinase is removed from the solution by physical adsorption. 

TABLE IV 

Protection of Penicillin against Penicillinase by Microcrystalline 
and Regular Sulfathiazole 

10,000 units of penicillin and 2 g. of sulfathiazole. 


Penicillinase 

Microcrystalline sulfathiazole 

Regular sulfathiazole 

Before incubation 

After 30 min. 
incubation 

Before incubation 

1 

After 30 min. 
incubation 

mg. 

zone of inhibition in mm. 

zone of inhibition in mm. 

50 

7.5 

6.3 

7.6 

5.8 

60 

7.7 

6.3 

7.5 

5.0 

70 

7.6 

5.8 

7.5 

5.3 

80 

7.3 

6.3 

7.4 

3.9 

90 

7.1 

5.0 

7.3 

0 

100 

6.9 

4.1 

6.8 

0 


The time required for the adsorption of penicillinase by the sulfonamides is fairly 
short. When a solution of the enzyme (50 mg.) is shaken continuously with 2 g. of the 
microcrystalline sulfa drug in the absence of penicillin, then centrifuged, and the 
supernatant tested for enzyme activity, the adsorption is found to be complete within 
5 min. This experiment may also serve as a final proof for the adsorption theory, 
since a solution of penicillin alone is not affected, even after prolonged contact of 
several days with microcrystalline sulfonamides. 

The foregoing experiments had all been made with a rather crude penicillinase 
preparation, and it seemed desirable to test the behavior of a purified form of this 
enzyme, such as penicillinase A-Schenle> .* Our data indicate that, on a weight basis, 
this preparation is approximately 500 times more potent. 

* We are indebted to Dr. C. E. Duchess, Schenley Laboratories, Inc., for a supply of 
Penicillinase A. 





398 G. CRONHEIM, M. E. BAIRD AND W. N. DANNENBURG 


In the experiments, the enzyme solution (20 ml.) was shaken for 5 min. with the 
specified amount of microcrystalline sulfadiazine, then centrifuged, and 10 ml. of the 
clear supernatant used for the test as previously described. Controls without the 
sulfonamide were treated in an identical manner. 

TABLE V 

Protection of Penicillin (50,000 Units) against Purified Penicillinase 
by Microcrystalline Sulfadiazine 
Time of exposure: 60 min. at 37°C. 


Penicillinase 

Without sulfadiazine 

After treatment with microcrystalline sulfadiazine 

1 K. 

2 k. 

mg. 


zone of inhibition in mm. 


None 

7.9 

7.9 

8.0 

0.09 

5.7 

7.7 

7.7 ’ 

0.18 

4.4 

6.5 

7.1 

0.27 

2.1 

5.6 

6.3 

0.36 

0 

4.2 

6.0 

0.54 


0 

4.3 


The results (Table V) show that there is no difference between the crude and the 
purified enzyme preparation as far as the adsorption of the penicillinase is concerned. 
It should be noted that the amount of sulfonamide required for complete adsorption 
is of the same order if the enzyme preparations are compared on the basis of their 
activity. 

« Preliminary experiments with microcrystal line sulfaguanidine indicate that the 
adsorptive power of this compound for penicillinase is much lower than that of either 
sulfadiazine or sulfathiazole_ 

Discussion 

The experimental observations are of interest for two reasons. First, 
penicillin can be protected by sulfonamide suspensions against the 
destructive action of penicillinase. Whether or not such a protection is 
of practical importance in the oral administration of penicillin requires 
further investigation. According to Stewart and May (10), the amount 
of penicillinase found in the upper part of the human digestive tract is 
negligible. However, their conclusion is based only on indirect proof in 
a small number of experiments. In mice and rabbits, the penicillin 
blood levels may be prolonged when the drug'is given orally together 
with substances which in vitro have an inhibiting effect on penicillinase 
( 9 ). 



PENICILLINASE STUDIES 


399 


The second factor concerns the observation that suspensions of sul¬ 
fonamides may remove, at least temporarily, an enzyme from its 
solution by physical adsorption. Such adsorption is observed frequently 
and has been utilized in the purification of penicillinase (6). The same 
principle is also applied in therapeutics by using insoluble substances 
such as kaolin in the treatment of certain intestinal disorders (3). 
However the present observation is apparently the first example that a 
drug which is given orally for its systemic effect may also act, at least 
temporarily, by adsorbing enzymes or other high molecular compounds. 
It should be mentioned that sulfonamides are also frequently applied 
in the form of dusting powder or ointments to wounds and body sur¬ 
faces. In these instances the possibility of interfering with normal bio¬ 
chemical processes is even greater, since the time of contact is much 
longer than in the case of oral administration. 

Summary 

The presence of sulfonamides (sulfadiazine and/or sulfathiazole) 
results in a marked protection of penicillin against penicillinase. This 
protection is due to an adsorption of the enzyme on the sulfonamide 
particles. The degree of protection depends upon the amount of sus¬ 
pended sulfonamides and their particle size. 

References 

1. Cbonheim, G., and Baird, M. E., The Journal-Lancet 69, 56 (1949). 

2. Dowung, H. F., Hussey, II. H., Hirsh, II. L., and Wilhelm, F., Ann. Intern. 

Med. 25, 950 (1946). 

3. Goodman, L., and Gilman, A., The Pharmacological Basis of Therapeutics, 

p. 769. Macmillan, New York, 1941. 

4. Gottlieb, B., and Forsyth, C. C., J. Am. Med. Assoc. 135, 740 (1947). 

5. Henry, R. J., and HousewriGht, R. D., J. Biol. Chem. 167, 559 (1947). 

6. Housewright, R. D., and Henry, R. J., ibid. 167, 553 (1947). 

7. Klein, M., and Kalter, S. S., J. Bad. 51, 95 (1946). 

8. Kolmer, J. A., Am. J. Med. Sci. 215, 136 (1948). 

9. Reid, R. D., Felton, L. C., and Pittroff, M. A., Proc. Soc. Exptl. Biol. Med. 63, 

438 (1946). 

10. Stewart, H. C., and May, J. R., Lancet 1947, 857 (Dec.). 

11. Winner, P. S., Abstracts, Meeting Am. Chem. Soc., Div. of Med. Chem., New 

York, September, 1947. 

12. Food and Drug Administration. Tests and Methods of Assay for Antibiotic Drugs* 



Antibiotin Effect of Homologs of Biotin and Biotin Sulfone 1,2 

Saul H. Rubin and Jacob Scheiner 

Nutrition Laboratories , Hoffmann-La Roche , Inc., Nuiley, N. J. 

Received February 15, 1949 

Introduction 

Hofmann, Chen, Bridgwater and Axelrod (1) have described the 
biological activity of a number of homologs of O-heterobiotin (oxy- 
biotin), including the bis-nor, nor, homo and bis-homo compounds. 
None of these were active antibiotins for either S. cerevisiae or L. 
arabinosus. On the other hand, the nor, the bis-homo and the homo 
compounds showed anti-O-heterobiotin activity for S. cerevisiae, while 
for L. arabinosus only the last compound was active. Since Goldberg 
et al . (2) have synthesized comparable homologs in the biotin series, it 
became possible to extend the studies to homologs of biotin and biotin 
sulfone and to compare, in a general fashion, the activities of the 
homologs of biotin and O-heterobiotin. 

* Experimental 

Methods 

The antibiotin activity of the homologs was investigated with Saccharomyces 
cerevisiae 189 , Lactobacillus casei e and Lactobacillus arabinosus 17-5. The procedures 
used with S. cerevisiae and L. casei have been reported elsewhere (2,3,4), with the 
present difference that the yeast assays were carried out in flasks. L. arabinosus was 
grown in the medium of Wright and Skeggs (5) and, in most of the experiments, incu¬ 
bated at 30°C. for periods of 20 or 67 hr. However, as it was found that greater growth 
was obtained in a shorter period at 37.5°C., some of the later experiments on the 
reversal of inhibition by biotin and the effect of the homologs in the presence of 
Tween 80 were read after 18 hr. at the higher temperature. The growth was measured 
turbidimetrically. 

The antibiotin activity expressed as the Molar Inhibition Ratio (M.I.R.) (6) was 

1 Publication No. 174. 

* Presented before the Division of Biological Chemistry, 114th Meeting of the 
American Chemical Society, Washington, D. C., September 3, 1948. 

400 




ANTIBIOTIN EFFECT 


401 


determined in the following ranges: L. caaei 0.1-0.2 my, S. cerevisiae 0.1-0.2 my, and 
L. arabinosus 0.6-1.0 my; anti-O-heterobiotin activity was determined in the ranges: 

L. caaei 0.6-1.0 my, L. arabinoaua 1.0-2.0 my, and S. cerevisiae 0.5-1.0 my. 

The common denominator in these various ranges for the estimation of M. I. R. 
has been the choice of comparable linear portions of the curves of response (c/. (4) for 
typical assay curves). The reproducibility of these determinations of M. I. R. can 
vary with many factors, including the slope of the curve, time of incubation, concen¬ 
tration of inoculum, etc., with the consequence that we have not always been able to 
reproduce values within the limits of ± 20% claimed by Dittmer and du Vigneaud 
(7). However, the M. I. R. reported here and elsewhere for various antibiotins cover 
so wide a range, that variations up to 100% are not of great consequence. 

Results 

Aclivily for S. cerevisiae (Table I) 

Esther decreasing or increasing the biotin side chain by one methyl- 
cue group gives compounds (nor- and homobiotin, respectively) which 
are potent antagonists of biotin and O-heterobiotin. Further increase 
in the side chain length (bis-homo- and tris-homobiotin) progressively 
diminishes the antibiotin activity. 

On the other hand, among the homologs of biotin sulfone, the nor- 
eompound is by far the weakest antimetabolite while the remaining 
homologs show approximately the same activity. These sulfone homo¬ 
logs are, on the whole, less effective in antagonizing the growth of 
yeast than the corresponding biotin homologs. 

When the M.I.R. is determined by the depression of growth from 
0.2 to 0.1 my of biotin, the homologs of biotin and biotin sulfone, with 
the exception of nor- and homobiotin, are 11-30 times more effective 
against O-heterobiotin than against biotin. On the other hand, the 
similarity in antibiotin and anti-O-heterobiotin activity of nor- and 
homobiotin is surprising, both with respect to the general pattern of 
the present study and the work of others (1,4,7,11). However, when the 

M. I.R. is calculated from total inhibition (Table IV), norbiotin and 
homobiotin are many times more effective against O-heterobiotin than 
against biotin. 

Activity for the Lactobacilli (Table I) 

Homobiotin and its sulfone are the most potent inhibitors, the 
M.I.R. against biotin being 130 and 460, respectively, for L. casei, and 
4000 and 1200, respectively, for L. arabinosus. Tris-homobiotin sulfone 
(M.I.R. 6200) is also a potent antibiotin for the latter organism. 



d Anti-O-Heterobiotin Effect of Homologs of Biotin and Bi 


402 


SAUL H. RUBIN AND JACOB 8CHEINER 



a n is the number of methylene units in the side chain; n = 4 in biotin and biotin sulfone. 

6 Determined for S . cerevisiae after 16 hr. at 30°C., for L. casei after 3 days at 37.5°C., and for L. arabinosus after 20 hr. at 
30°C. 



ANTIBIOTIN EFFECT 


403 


d-Homobiotin was found to be twice as active as dZ-homobiotin for 
L. arabinosus, demonstrating that only one of the enantiomorphs is 
active. 

For L. casei, the biotin homologs are 10-67 times more effective in 
inhibiting growth when O-heterobiotin is the growth factor. The ratio 
for the sulfone analogs is somewhat more constant, varying from 13 to 
26. Among the biotin homologs, the homo- and tris-homo-compounds 
are very potent, while among the homologs of biotin sulfone, only the 
homo-compound shows a marked anti-O-heterobiotin effect. The molar 

REVERSAL OF THE ANTIBIOTIN EFFECT OF HOMO¬ 
BIOTIN AND HOMOBIOTIN SULFONE BY BIOTIN 



Fig. 1 . The incubation period for the test organism, L. arabinosus, was 18 hr. at 
37.5°C. The effect of each inhibitor was measured in the presence of 1.0 my of biotin 
per tube while the effect of biotin in reversing the antibiotin activity was measured in 
the presence of 100 y of inhibitor. 

inhibition ratios of homobiotin and homobiotin sulfone against O-hetero¬ 
biotin (8 and 18, respectively) are the strongest antivitamin effects yet 
recorded in the biotin literature. 

For L. arabinosus the homologs are from 6 to 67 times more effective 
in antagonizing growth when O-heterobiotin is the growth factor. 
Homobiotin (M.I.R. 100), homobiotin sulfone (M.I.R. 100), and tris- 
homobiotin sulfone (M.I.R.850) are the most potent anti-O-hetero- 
biotins. When the incubation period is extended from 20 to 67 hr., 



404 


SAUL H. BUBIN AND JACOB SCHEINER 


there is a marked decrease in antibiotin effect (compare the L. arabino¬ 
sus data in Tables I and III). 

The biotin homologs are, in general, more effective than the biotin 
sulfone homologs against L. casei, while the reverse is true for L. 
arabinosus. 

REVERSAL OF THE ANTi—0—HETEROBIOTIN EFFECT 
OF HOMOBIOTIN AND HOMOBIOTIN 
SULFONE BY O-HETEROBIOTIN 



Fiq. 2. The incubation period for the test organism, L. arabinosus, was 18 hr. at 
37.5°C. The effect of each inhibitor was determined in the presence of 2.0 my of 
O-heterobiotin per tube, and the effect of O-heterobiotin in reversing the anti-O-hetero- 
biotin activity of homobiotin and homobiotin sulfone was determined in the presence 
of l00 y of inhibitor. 


Reversal of Inhibition by Biotin and O-Heterobiotin 

The effect of increasing amounts of biotin in nullifying the anti¬ 
biotin effect of homobiotin, tris-homobiotin, homobiotin sulfone and 
tris-homobiotin sulfone, was determined with L. arabinosus (18 hr. at 
37.5 C.). For the biotin homologs, about 5.0 and 0.5 my of biotin were 
required to initiate growth in the presence of 100 y of homobiotin and 
tris-homobiotin, respectively, while for the biotin sulfone homologs 
about 15 and 1.0 my of biotin initiated growth in the presence of 100 y 
of homobiotin sulfone and tris-homobiotin, respectively. The reversal 


ANTIBIOTIN EFFECT 


405 


of inhibition by biotin, as well as the antibiotin effect of homobiotin and 
homobiotin sulfone, are illustrated in Fig. 1. 

O-heterobiotin also reverses the effect of these antagonists, as is 
illustrated in Fig. 2. Similarly, for S. cerevisiae, the antibiotin effect 
of the homologs is reversed by biotin and O-heterobiotin. 

TABLE II 

Estimates of Molar Inhibition Ratio for L. arabinosus from Partial and Total Inhibition “ 


Molar inhibition ratio* 


Compound 

Antibiotin 

Anti-O-heterobiotin 


Partial 

Total' 

Total 

Partial 

Total' 

Total 


inhibition 

inhibition 

Partial 

inhibition 

inhibition 

Partial 

nL-Homobiotin 

12,000 

20,000 

2 

150 

500 

3 

DL-Homobiotin sulfone 

5,500 

17,000 

3 

120 

240 

1 

2 


“Total inhibition ratios represent the amount of inhibitor required to prevent 
growth in the presence of either 1.0 nvy of biotin or 2.0 1117 of O-heterobiotin. 
b Determined after 18 hr. at 37.5°C. 
e Calculated from "inhibition” curve of Fig. 1. 

4 Calculated from “inhibition” curve of Fig. 2 . 

Comparison of Estimates of M.I.R. from Partial and 
Total Inhibition (Table II) 

A comparison of the M.I.R. calculated in the usual fashion for L. 
arabinosus and the M.I.R. calculated at total inhibition of 1.0 my of 
biotin or 2.0 my of O-heterobiotin, shows that the partial inhibition 
values average approximately one-third of the total inhibition results, 
i.e., values obtained from partial inhibition indicate that the com¬ 
pounds are 3 times more effective than values obtained from total 
inhibition. This is to be expected, since calculations of this nature from 
a sigmoidal curve would vary with the changing slopes of the curve. 
Our original choice of a method which measures partial inhibition is 
predicated largely upon use of a major part of the curve which is 
essentially linear (c/. “Methods”). 

As Hofmann et al. (1) used a comparable method for determining 
M.I.R. of O-heterobiotin homologs at total inhibition, this factor of 3 
provides a basis for comparing their results with the present data, with 
the following results. 






406 


SAUL H. RUBIN AND JACOB SCHEINER 


Comparison with O-Heterobiotin Homologs 

In Table III the activity of the biotin homologs and the corre¬ 
sponding O-heterobiotin homologs for L. arabinosus is compared (incu¬ 
bation period: 3 days at 30 C.). As the activity of the O-heterobiotin 
homologs (nor-, homo- and bis-homooxybiotin) was reported by Hof¬ 
mann et al. (1) on the basis of total inhibition, their results have been 
divided by the factor of 3 discussed in the preceding section. Among the 
O-heterobiotin homologs, the most active, homooxybiotin, is effective 
only against O-heterobiotin. On the other hand, the corresponding 
biotin homolog, homobiotin, not only is many times more effective 
against O-heterobiotin, but also antagonizes biotin. Bis-homobiotin is 
also an effective antagonist of O-heterobiotin. 

TABLE III 

Comparative Antibiotin and Anti-O-Heterobiotin Effect of Homologs of Biotin 
and O-Heterobiotin for L . arabinosus 


Molar inhibition ratio 


Biotin homolog 

Present paper 0 

Hofmann et al. (l) h 


Antibiotin 

Anti-O-hetero- 

biotin 

O-Heterobiotin 

homolog 

Antibiotin 

Anti-O-hetero- 

biotin 

Nor- 

Homo- 

Bis-homo- 

>200,000 

29,000 

>200,000 

>100,000 

940 

55,000 

Nor- 

Homo- 

Bis-homo- 

>167,000 

>167,000 

>167,000 

>167,000 

75,000 

>167,000 


• Determined after 67 hr. at 30°C. 

h These results have been divided by 3 for comparative purposes (c/. Table II and 
text). 


For S. cerevisiae (Table IV), homooxybiotin shows a weak anti¬ 
biotin activity, while the nor- and bis-homo- homologs are inactive. 
On the other hand, the corresponding biotin homologs are effective 
biotin antagonists. Against O-heterobiotin, both series of homologs are 
active, but again the biotin homologs are more effective antagonists. 

As mentioned previously, the results in Table IV differ from those 
presented in Table I, particularly with respect to norbiotin and homo¬ 
biotin. The data in Table IV were obtained by determining the amount 
of inhibitor required to antagonize completely 0.2 m-y of biotin or 1.0 
my of O-heterobiotin, this method being comparable to the type of 













ANTIBIOTIN EFFECT 


407 


TABLE IV 

Comparative Antibiotin and Anti-O-Heterobiotin Effect of Homologs of Biotin 
and O-Heterobiotin for S. cerevisiae 


Molar inhibition ratio 


Biotin liomolog 

Present paper® 

Hofmann et al. (I) and Axelrod et al. (13) 


Antibiotin 

Anti-O-hetero¬ 

biotin 

O-Heterobiotin 
homolog 

Antibiotin 

Anti-O-hetero- 

biotin 

Nor- 

25,000 

1,000 

Nor- 

>500,000 

143,000 

Ilomo 

25,000 

1,000 

Homo- 

445,000 

7,400 

Bis-homo- 

50,000 

1,800 

Bis-homo- 

>500,000 

30,000 


a The molar inhibition ratios represent the amount of inhibitor required to prevent 
growth in the presence of 0.20 1117 of biotin or 1.0 m 7 of O-heterobiotin. 


measurement used by Hofmann et al. (1). A major change in antibiotin 
activity is found only in norbiotin and homobiotin, while bis-homo- 
biotin, tris-homobiotin and the homologs of biotin sulfone show little 
change in antibiotin effect. The anti-O-heterobiotin activity of the 
homologs is not markedly affected by the change in method of deter¬ 
mination of M.I.R. 


Effect of Homologs in the Presence of Tween 80 

Williams, Broquist and Snell (8) demonstrated that Tween 80 
supports near-maximal growth of several microorganisms on a semi¬ 
synthetic medium in the absence of biotin. Potter and Elvehjem (9) 
later showed that a combination of aspartic acid and oleic acid would 
replace biotin almost completely for growth of L. arabinosus. These 
observations suggested that one function of biotin is concerned with 
the synthesis of oleic acid (8,9). If oleic acid is the product of an enzyme 
system catalyzed by biotin, then compounds antagonistic to biotin 
should have no effect on oleic acid. In Fig. 3 this is shown to be the 
case when Tween 80 replaces biotin. As the same result is obtained with 
d-homobiotin, tris-homobiotin, homobiotin sulfone, and tris-homo¬ 
biotin sulfone, only the non-effect of d-homobiotin is shown. It was 
found that 750 y of Tween 80 per tube produce in 18 hr. at 37.5 C. 
about the same growth response of L. arabinosus as 1 my of biotin. The 
addition of 400 y of d-homobiotin has no effect on growth with oleic 



408 


SAUL H. RUBIN AND JACOB SCHEINER 


NON-EFFECT OF d-HOMOBIOTIN 
ON TWEEN 80 



0-0 MCG. (X 1 O’*) d-HOMOBIOTIN CONTRA TWEEN 80 
•—•MCG. •• BIOTIN 


Fig. 3. The incubation period for the teat organism, 

L. arabinosus, was 18 hr. at 37.5°C. 

acid, while 5y of inhibitor are sufficient to prevent growth with biotin. 
This specific effect of the homologs is similar to that of Y-(3,4-ureylene- 
cyclohexyl)-butyric acid (10). The failure of these inhibitors to prevent 
growth with oleic acid provides further evidence for the hypothesis that 
one function of biotin concerns the synthesis of oleic acid. 

Discussion 

The high activity of norbiotin in relation to homobiotin for S. 
cerevisiae can be contrasted with the results obtained by Dittmer and 
du Yigneaud (7) with imidazolidone aliphatic acids, and by Hofmann 
et al. (1) with homologs of oxybiotin. The former found that shortening 
the aliphatic side chain of imidazolidonecaproic acid by one methylene 
group reduced the antibiotin effect to a much greater extent than when 
the side chain was lengthened by one or two methylene groups. The 
latter found that nor-oxybiotin was a less effective O-heterobiotin 
antagonist than either homooxybiotin or bis-homooxybiotin. On the 
other hand, the progressive decrease in activity shown in Table I as the 
side fhftiTi is lengthened, is in general agreement with the work of 
Dittmer and du Vigneaud (7). The markedly decreased antibiotin 



ANTIBIOTIN EFFECT 


409 


activity of norbiotin and homobiotin when the M.I.R. is calculated 
from total inhibition is surprising in view of the slight change in anti¬ 
biotin activity of bis-homobiotin, tris-homobiotin and the homologs 
of biotin sulfone, as well as the slight changes in anti-O-heterobiotin 
activity found for all of the compounds. These results indicate that, 
although nor-biotin and homobiotin are effective biotin antagonists for 
the partial antagonism of growth, they are much less effective in antago¬ 
nizing the growth of the yeast in the presence of the small amounts of 
biotin required to initiate growth. The observation that for both L. 
casei and L. arabinosus the antibiotin and anti-O-heterobiotin effect of 
the homologs higher than biotin is greater than the corresponding 
effect of the homolog lower than biotin, is in agreement with the results 
of Hofmann el al. (1) and Dittmer and du Vigneaud (7). However, in 
contrast to the results with S. cerevisiae, there is no progressive decrease 
in antagonism as the side chain is lengthened, as was reported by the 
latter group for imidazolidone aliphatic acids. 

The greater antimetabolite effect of the homologs against O-hetero- 
biotin than biotin is in agreement with previous observations on the 
comparative effect of desthiobiotin (4), of various homologs of 0- 
heterobiotin (1), of biotin sulfone (11) and of desthiobiotin analogs 
(12). The antibiotin effect of d/-homobiotin (M.I.R.-4000) is much 
greater than that previously reported (M.I.R. 42,000) (2). However, 
the result reported initially was a range finder and, therefore, was 
indicative only of order of magnitude. 

Summary 

Norbiotin, homobiotin, bis-homobiotin, tris-homobiotin and the 
corresponding homologs of biotin sulfone are potent antibiotins for 
L. casei and, with the exception of the nor homologs and bis-homo¬ 
biotin, for L. arabinosus. For S. cerevisiae all homologs but norbiotin 
sulfone are potent biotin antagonists. 

When O-heterobiotin is the growth factor, the homologs are more 
effective against the lactobacilli and the yeast. The biotin homologs are 
generally more active against S. cerevisiae and L. casei than the biotin 
sulfone homologs, while the latter are more effective against L. arabino¬ 
sus. The inhibition has been shown to be competitive in nature for 
L. arabinosus and S. cerevisiae. 

The antibiotin activity of norbiotin and homobiotin against S. 



410 


SAUL H. RUBIN AND JACOB SCHEINER 


cerevisiae is much greater when determined from partial than from 
total inhibition. This difference was not found for bis- and tris-homo- 
biotin or for any of the sulfones tested. In the case of O-heterobiotin, 
the method of determination did not affect the M.I.R. of any of the 
compounds. 

When oleic acid is the growth factor, homobiotin, tris-homobiotin, 
homobiotin sulfone, and tris-homobiotin sulfone have no effect on L. 
arabinosus. 


References 

1. Hofmann, K., Chen, C., Bridgwater, A., and Axelrod, A. E., J. Am. Chem. 

Soc ., 69, 191 (1947). 

2. Goldberg, M. W., Sternbach, L. H., Kaiser, S., IIeineman, S. D., Scheiner, 

J., and Rttbin, S. H., Arch. Biochem. 14,480 (1947). 

3. Rubin, S. H., Drekter, L., and Moyer, E. II., Proc. Soc. Exptl. Biol. Metf. 68, 

352 (1945). 

4. Rubin, S. H., Flower, D., Rosen, F., and Drekter, L., Arch. Biochem. 8, 79 

(1945). 

5. Wright, L. D., and Skeggs, H. R., Proc. Soc. Exptl. Biol. Med. 66, 95 (1944). 

6 . Dittmer, K., and du Vigneaud, V., Science 100, 129 (1944). 

7. Dittmer, K, and du Vigneaud, V., J. Biol. Chem. 169, 63 (1947). 

8 . Williams, W. L., Broquist, H. P., and Snell, E. E., ibid. 170, 619 (1947). 

9. Potter, R. L., and Elvehjem, C. A., ibid. 172, 531 (1948). 

10. Broquist, II. P., and Snell, E. E., ibid. 173, 435 (1948). 

11. Axelrod, A. E., de Woody, J., and Hofmann, K., ibid. 163, 771 (1946). 

12. Duschinsky, R., and Rubin, S. H., J. Am. Chem. Soc. 70, 2546 (1948). 

13. Axelrod, A. E., Purvis, S. E., and Hofmann, K, J. Biol. Chem. 176, 695 (1948). 



Biochemical Studies on Chloramphenicol (Chloromycetin ). 1 
I. Colorimetric Methods for the Determination of 
Chloramphenicol and Related Nitro Compounds 2 

Anthony J. Glazko, Loretta M. Wolf and Wesley A. Dill 

From the Research Laboratories of Parke, Davis and Company, Detroit, Michigan 
Received May 11,1949 

Introduction 

Chloramphenicol (Chloromycetin 1 ) is a new antibiotic which has 
proven to be highly effective against certain Rickettsiae and gram 
negative bacteria (1). It is an aromatic nitro compound which was 
isolated and partly characterized by Bartz (2). The complete structure 
has been established by Rebstock, Crooks, Controulis and Bartz (3), 
and it has been synthesized by several methods (4,5). Chloramphenicol is 
D( — )</irco-2-dichloracetamido-l-p-nitrophenyl-l ,3-propanediol, having 
the following configuration (3): 

II HN—CO—CIIC1-. 

0,N V-di- k— CH 2 OH 

d>H H 

The analytical methods described in this paper involve the quanti¬ 
tative reduction of the aromatic nitro group to form a primary amine, 
which is then determined by the standard Bratton-Marshall diazo 
procedure (6). Aryl amines, which may be present before reduction, are 
accounted for by running a parallel determination in which the reduc¬ 
tion step is omitted. 

The chemical method is not specific for the active antibiotic, since 
inactive degradation products of chloramphenicol which still retain the 
nitro group are also included in the determination. Reliable estimates 

1 Parke, Davis and Co. trade mark for chloramphenicol. 

’Reported in part at the Second National Symposium on Recent Advances in 
Antibiotic Research, Washington, D. C., April 11,1949. 


411 



412 


A. J. GLAZKO, L. M. WOLF AND W. A. DILL 


of active chloramphenicol are best made by microbiological assay 
procedures (1,7), and the differences between the chemical and micro¬ 
biological assay figures thus represent inactive degradation products, 
including the glucuronide and the free base of chloramphenicol (8). 
The colorimetric method can be made more specific for the active anti¬ 
biotic by using the solvent extraction procedure described in this paper. 

A number of reducing agents were investigated for quantitative 
reduction of the nitro group in chloramphenicol, including titanous 
chloride, zinc, tin, Raney nickel, and electrolytic reduction. Of these, 
the first two agents were found to be most satisfactory. The titanous 
chloride reduction method was used extensively for biochemical in¬ 
vestigations, and is preferred to the zinc method because no heating 
period is required for reduction and less interference is encountered 
from normal constituents of tissue. The zinc reduction method is in¬ 
cluded here because of its simplicity, but it is satisfactory only in the 
absence of interfering substances. 

Experimental 

Reagents 

1. Zinc Dust. Reagent grade. 

2. Titanous Chloride. An aqueous solution is made as described by English (9) and 
stored under hydrogen. Sixty ml. of stock 15% TiClj is added to 300 ml. of concen¬ 
trated HC1, heated to boiling and cooled under Hj or CO 2 . It is then poured into 3.61. 
of previously boiled and cooled distilled water in the storage bottle, air is displaced 
with H, and the contents thoroughly mixed. A convenient storage apparatus and 
burette is described by Knecht and Hibbert ( 10 ). The final solution should be a clear 
amethyst color, with no evident turbidity or precipitate. 

3. Sodium Hydroxide. 1 N solution. 

4 . Hydrochloric Acid. 0.25 N and 0.5 N solutions. 

5. Sodium Nitrite. 0.1% aqueous solution, prepared fresh daily. 

6. Ammonium Sulfamate. 0.5% aqueous solution. 

7. Bratton-MarshaU Coupling Reagent. 0.1% aqueous solution of N-( 1-naphthyl) 
ethylenediamine dihydrochloride. This should be stored in a brown bottle and kept in 
the refrigerator when not in use. 


Procedure 

Samples are diluted to contain no more than 120 7 of chloramphenicol/ml. Where 
appreciable amounts of protein are present, the samples are first deproteinized with 
3% trichloroacetic acid for the titanium reduction method, or with Zn(OH)t if the 
zinc reduction procedure is to be used. 

Reduction of the nitro group of chloramphenicol with titanous chloride proceeds 
rapidly and smoothly at room temperature. Excess titanium salts are precipitated as 



CHLORAMPHENICOL STUDIES 


413 


the hydrated oxides in alkaline solution and removed by centrifugation to prevent 
interference with subsequent diazotization. An aliquot of the solution is then acidified 
and color is developed as described in the following section. A parallel determination 
is also run for aryl amines which may be present before reduction to correct for the 
color produced by these compounds. 

One ml. of sample containing up to 120 y of chloramphenicol is pipetted in each of 
two 15 X 125 mm. test tubes (A and B). To tube A is added 2 ml. of water, and 1 ml. 
of the titanous reagent is run in directly from the burette which is an integral part of 
the storage apparatus. The contents of the tube are immediately mixed by shaking. 
One ml. of 1 N NaOH is then added to precipitate the titanuim salts. The final pH 
should be approximately 9.0 to insure complete precipitation. With acid filtrates, the 
strength of the NaOH solution should be increased so that 1 ml. produces a final pH of 
9.0. After formation of the precipitate, tube A is centrifuged at 1500 R.P.M. for a few 
min. to pack down the precipitate. 

Four ml. of water is added to tube B, which serves as the aryl amine control. Finally, 
1 ml. portions of the supernatant solutions from tubes A and B are transferred by 
pipette to two optical cuvettes (A and B), each containing 4 ml. of 0.25 N HC1. Color 
development is carried out as described in the next section. Where low concentrations 
of chloramphenicol are encountered, larger volumes of sample may be taken for 
reduction and for color development. 

In the zinc reduction procedure, samples are diluted to contain less than 120 y of 
chloramphenicol. One ml. of sample is pipetted into each of two 15 X 125 mm. test 
tubes (A and B) marked at the 5 ml. level. Four ml. of 0.5 N HC1 is added to each 
tube, and approximately 50 mg. of zinc dust is added to tube A only. Both tubes are 
then heated in flowing steam or in a boiling water bath for exactly 1 hr. After cooling 
rapidly to room temperature, volumes are readjusted to the 5 ml. mark by the addition 
of water, ignoring the slight volume correction due to the presence of zinc (<0.01 ml.). 
Both samples are mixed thoroughly, the zinc is allowed to settle, and 1 ml. portions 
of the clear supernatant solution are transferred by pipette to 18 X 150 mm. optical 
cuvettes containing 4 ml. of 0.25 N HC1. These are then handled as described in the 
next section. 


Development of Color 

Following reduction by either the titanium or zinc procedures, 
samples are handled in the same manner thereafter for development of 
color. Prior to the diazotization procedure, if any appreciable amount 
of extraneous color is present in the cuvettes the optical densities should 
be read at 555 m m in a Coleman Junior Spectrophotometer against a 
reagent blank. These readings are designated Ai and Bi for tubes A 
and B. Where the solutions are colorless, or nearly so, the Ai and Bi 
readings may be omitted. 

To each cuvette is added 0.5 ml. of the NaN0 2 reagent, and the contents are mixed 
and allowed to stand for exactly 5 min. Then 0.5 ml. of the sulfamate reagent is added 
to decompose excess HNO*. After 3 min. standing, 0.5 ml. of the Bratton-Marshall 



414 


A. J. GLAZKO, L. M. WOLF AND W. A. DlLL 


reagent is added, and the contents of the tubes thoroughly shaken. All tubes are then 
placed in a water bath at 38°C. and left for exactly 1 hr. At the end of this time the 
tubes are removed, wiped dry with a clean towel, and optical densities are read against 
a reagent blank at 555 m/i (readings A 2 and B 2 ). 

Calculation of Results 

The optical density due to reducible nitro compounds is calculated from the colorim¬ 
eter readings in the following manner: 

Corrected Optical Density = (A 2 — O.77A0 — (B 2 — 0.77Bi). 

This equation corrects for any interfering color in the original sample (readings 
Ai and Bi), and for the presence of aryl amines in the original sample (reading B 2 ). 
The factor 0.77 represents a correction for decrease in the optical density of the origi¬ 
nal solution due to dilution with the diazo reagents. 

Corrected optical densities are converted to concentrations of chloramphenicol by 
reference to a standard curve prepared in the same manner as the unknowns, using 
standard solutions containing known concentrations of the drug. A direct relation 
between optical density and chloramphenicol concentration is found to exist. Final 
results are expressed in terms of “chloramphenicol-equivalents,” since the inactive 
metabolic products containing the nitro group are not differentiated from chloram¬ 
phenicol by this procedure. Analyses are run in duplicate on each sample, and at least 
one standard is included with each set of unknowns to check on the uniformity of test 
conditions. 

In replicate analyses carried out with the zinc and titanous chloride procedures, 
variations of dt 4% from the mean were observed. After the addition of known 
amounts of chloramphenicol to plasma, recoveries of 92 to 98% of corresponding 
aqueous solutions are generally obtained. Recoveries of 90% were obtained with 
, w r hole blood in dilutions of 1:40. Tissue homogenates gave recoveries of 90 to 98% in 
dilutions of 1:50, while lesser dilutions gave unsatisfactory results. 

Conditions for Color Development 

Optimum conditions for diazotization were established by studying 
maximum color formation with different concentrations of reagent, 
time of reaction and acidity. The final color was found to have maxi¬ 
mum light absorption at 555 m/z, which is somewhat higher than found 
with simpler aryl amines (535 m/z-545 m/z). The same peak light ab¬ 
sorption was found with zinc- and titanium-reduced samples, and the 
metabolic products of chloramphenicol showed similar absorption 
spectra. 

The slow rate of coupling with the Bratton-Marshall reagent under 
the conditions of this procedure is evident by the slow increase in 
optical density on standing. Data presented in Table I show that the 



CHLORAMPHENICOL STUDIES 


415 


coupling rate follows the pattern of a monomolecular reaction, and is 
accelerated at higher temperatures. Although complete color develop¬ 
ment is not attained in 1 hr. at 38°C., reproducible colorimetric readings 
can be made if standard solutions are also read in 1 hr. The coupling 
rate of the zinc-reduced samples appears to be somewhat greater than 
with the titanium-reduced samples, possibly due to splitting of the 
side chain on acid hydrolysis with the formation of a different reduction 
product. 


TABLE I 

Rate of Coupling of Reduced Chloramphenicol 
Solutions containing 20 y of titanium-reduced chloramphenicol were diazotized, 
sulfamate and coupling reagent added, and optical densities measured at 555 m#x after 
different periods of standing at 25°C. and 38°C. Readings wen 4 made against a reagent 
blank with a Coleman Junior Spectrophotometer. The rate constants are calculated 
from the equation Kt = log (100/per cent remaining uncoupled ), where the greatest 
observed optical density is taken as 100% coupling, and t represents the time of 
standing in minutes. 




25°C. 

1 


38°C. 


Tune 

Optical 

density 

Per cent 
cou pli*d 

K 

Optical 

density 

Per cent 
coupled 

K 

min. 

10 

0.102 

17 


0.169 

29 


Kl 





47 

0.014 

..E9 

■ ! 1 1 ■ 



0.344 

59 


60 

0.354 

58 

0.006 

0.475 

82 

0.013 

120 

0.498 

82 

0.006 

0.561 

97 

0.013 

180 

0.573 

95 

0.007 

0.581 

(100) 

— 

245 

0.606 

(100) 

— 

0.581 

(100) 

— 


The rate of coupling is markedly influenced by pH, with the greatest 
rate observed at pH 4.5. However, at this pH a large part of the coupled 
compound exists as a yellow-colored base and the final solution must 
be further acidified to obtain maximum absorption at 555 m n. This may 
be used as the basis for a rapid analytical procedure in which coupling 
is carried out at pH 4.5 for 20 min. at room temperature, and the 
solution is then acidified prior to reading the color intensity. The 
coupled compound appears to be fairly stable in acid and weakly 
alkaline solutions, but the diazonium salt prior to coupling is stable 
only in strongly acidic solutions. For this reason we prefer to conduct 




416 


A. J. GLAZKO, L. M. WOLF AND W. A. DILL 


the coupling reaction in strong acid, instead of using the rapid procedure 
at pH 4.5. 


Establishment of Conditions for Reduction 

Conditions for the reduction of chloramphenicol were established by varying time 
of heating, acid concentration and amount of reducing agent. The use of 0.25 N HC1 
and approximately 50 mg. zinc dust was found to give practically complete reduction 
in 1 hr. at 100°C. With titanous chloride, reduction was found to occur almost imme¬ 
diately at room temperature in acid, alkaline and neutral solutions. Six equivalents 
of titanous chloride are required for the reduction of one equivalent nitro group (10). 
However, a large excess of reagent is needed for complete reduction under the condi¬ 
tions of this test because no attempt is made to avoid atmospheric oxygen, and other 
reducible substances may be present. 

The removal of titanium salts is accomplished by precipitation of the insoluble 
hydrated oxides in the presence of alkali. 8 This step is essential, because the titanous 
reagent would otherwise interfere with diazotization, reducing the HNOj and diazon- 
ium salts. Precipitation appears to be practically complete at pH 9, but a large excess 
of alkali should be avoided. 4 Aliquots of the clear supernatant solution should be re¬ 
moved for analysis immediately after centrifuging, because the titanous oxide is 
sufficiently reactive to decompose water slowly with the evolution of hydrogen gas 
(10), causing the precipitate to rise to the surface. 

Solvent Extraction Procedure for Chloramphenicol 

The colorimetric procedure described in this paper does not distinguish between 
active chloramphenicol and its inactive degradation products which retain the aro¬ 
matic nitro group. The urinary excretion of these inactive metabolic products is often 
10-fold greater than that of active antibiotic (11). Consequently, an extraction proce¬ 
dure was developed to increase the specificity of the colorimetric method for chloram¬ 
phenicol. 

The following method was used for analysis of urine samples collected from dogs 
which had been given chloramphenicol by mouth. One ml. of urine was pipetted into 
glass-stoppered tubes containing 6 ml. of ethyl acetate and 2 ml. of 0.2 M phosphate 
buffer at pH 6.0.® The mixtures were shaken in a mechanical shaker for 10 min. The 
tubes were then centrifuged and the aqueous layers removed by aspiration and dis¬ 
carded. The organic layer was washed twice by shaking with equal volumes of pH 6.0 
buffer previously saturated with ethyl acetate. After separation of the two phases, 5 

•The precipitate is colored a deep blue-black when excess titanous reagent is 
present. With high concentrations of nitro compounds, the titanous reagent is oxi¬ 
dized to the titanic form, and a white precipitate appears on the addition of alkali. 

4 NatCOi can be used in place of NaOH for precipitation of the titanium reagent, 
but foaming occurs after reacidification due to production of CO*. 

4 This ratio of solvent to aqueous phase was found to give minimum volume change 
due to the mutual solubility of the two phases. We have also used isopropyl acetate, 
first suggested to us by Dr. E. K. Marshall, Jr., with equally good results. 




CHLORAMPHENICOL STUDIES 


417 


ml. of the ethyl acetate layer was transferred by pipette to a 20 ml. beaker and evapo¬ 
rated to dryness on a steam bath. The residue was dissolved in 3 ml. of water* and 
analyzed by the titanium reduction procedure. In addition, microbiological assays (7) 
were run on the original urine samples , 7 and colorimetric analyses were also made 
for aryl amines and total nitro compounds. The results presented in Table II show 
fair agreement between the microbiological and chemical extraction procedures. Data 
on the distribution of the purified metabolic products of chloramphenicol between 
ethyl acetate and aqueous buffers at various pH levels will be presented elsewhere. 

TABLE II 

Analysis of Dog Urine for Chloramphenicol 
Urine samples collected from dogs which had been given chloramphenicol by mouth 
were analyzed for aryl amines and total aromatic nitro compounds by colorimetric 
methods, and for unchanged chloramphenicol by the microbiological and solvent 
extraction procedures. All results are expressed as 7 of chloramphenicol-equiva- 
lents/ml. of urine. 


Sample no. 

Aryl amine* 

Total nitro 
compounds 

Chloram phenicol 

Extraction 

Microbiological 

46 

600 

190 

5 

4 

33 

340 

420 

7 

5 

25 

440 

860 

37 

28 

19 

670 

2030 

130 

120 

40 

720 

3700 

130 

120 

10 

570 

3300 

220 

220 

17 

460 

5220 

240 

230 

21 

270 

1820 

240 

230 

23 

280 

5980 

440 

380 

22 

280 

6350 

600 

610 

43 

420 

3760 

560 

610 


Interfering Substances 

A survey of normal tissue constituents showed that adenine, adeno¬ 
sine triphosphate and nucleic acids, after heating with zinc in acid 
solution, formed derivatives which would diazotize and couple with 
the Bratton-Marshall reagent, but these compounds did not react in 

* With blood and tissue specimens the solution may be turbid at this point due to the 
presence of lipide substances, which can be eliminated by extraction with an equal 
volume of petroleum ether. 

7 The microbiological assays were conducted by Dr. J. L. Schwab and Mrs. Mar¬ 
garet Galbraith, to whom we express our appreciation. 




418 


A. J. QLAZKO, L. M. WOLF AND W. A. DILL 


this manner after treatment with titanous chloride. Both the zinc and 
titanium procedures were found to yield colored derivatives with folic 
acid by reduction of the —NH—CH 2 — linkage to form a primary 
amine. The zinc method described in this paper has been applied to 
the determination of adenine, and the titanium method has been found 
suitable for the determination of folic acid (12). 

Acknowledgments 

We are especially indebted to Dr. A. C. Bratton, Jr., for his constant help and 
advice during the course of this work; to Dr. Quentin R. Barts and Dr. John Ehrlich 
for a sample of crystalline chloramphenicol; and to Dr. Harry Crooks for preliminary 
data on the structure of chloramphenicol during the early stages of the work. 

Summary 

w 

A colorimetric procedure is described for the determination of 
chloramphenicol in biological materials, based on the reduction of the 
aromatic nitro group to a primary amine with titanous chloride or 
metallic zinc, followed by diazotization and coupling with the Bratton- 
Marshall reagent. This determination includes certain inactive meta¬ 
bolic derivatives of chloramphenicol which are shown to be present in 
high concentration in urine, as well as active chloramphenicol. The 
method can be made specific for chloramphenicol by preliminary ex¬ 
traction with organic solvents. 

References 

1 . Smith, R. M., Joslyn, D. A., Gruhzit, 0. M., McLean, T. W., Penner, M. A., 

and Ehrlicii, J., J. Bad. 55, 425 (1948). 

2. Bartz, Q. R., J. Biol Chan. 172, 445 (1948). 

3. Rebstock, M. G\, Crooks, II. M., Controulis, J., and Bartz, Q. R., J. Am. 

Chem. Soc. 71, 2458 (1949). 

4. Controulis, J., Rebstock, M. C., and Crooks, H. M., ibid. 71, 2463 (1949). 

5. Long, L. M., and Troutman, H. A., ibid. 71, 2469, 2473 (1949). 

6 . Bratton, A. C., and Marshall, E. K., J. Biol. Chem. 128, 537 (1939). 

7. Joslyn, D. A., and Galbraith, M., J. Bad. 54, 26 (1947). 

8. Dill, W. A., and Glazko, A. J., Federation Proc. 8, 34 (1949). 

9. English, F. L., J. Ind. Eng. Chem. 12, 994 (1920). 

10. Knecht, E., and Hibbert, E., Now Reduction Methods in Volumetric Analysis, 

2nd Ed., Longmans, Green and Co., Manchester, 1925. 

11. Glazko, A. J., Wolf, L. M., Dill, W. A., and Bratton, A. C., J. Pharmacol. 

Exptl. Therap. in press (1949). 

12. Glazko, A. J., and Wolf, L. M., Arch. Biochem. 21, 241 (1949). 



Factors Influencing Fat Synthesis by Rhodotorula gracilis 

S. C. Pan, A. A. Andreasen and Paul Kolachov 

Joseph E. Seagram & Sons, Inc., Louisville, Kentucky 
Received May 31, 1949 

Introduction 

The production of fat by microorganisms has long been a subject of 
interest to research and to industry. Attempts to produce fat on a 
commercial scale by The growth of microorganisms were made in 
Germany during both World Wars. Endomyces vernalis was employed 
during World War I and Oospora lactis during World War II (1,6). It is 
doubtful whether they were successful either time because the enor¬ 
mous surface areas required by both of these organisms undoubtedly 
presented overwhelming technical difficulties when commercial scale 
cultivation was attempted. 

However, recent investigations have shown that there are some yeasts and mold s, 
capable of synthesizing significant amounts of fat when grown i n submerged cultur es ^ 
Fusaria (3, 10,21) , Mucor and molds related to it (4), were among the first to be noted. 
Certain Torula yeasts, e.g., T orulopsis lipofera (9 ) and R hodotor ul a gracilis (5), and 
tfie ye ast N edaromyces reuka ufii (1 3) have also been found capable of synthesizing 
significant amounts offat under jeonditions of submerged culture. The soil yeast isolated 
by Starkey (19) is_un ique in tha t it can grow and synthesize fat in media almost devoid 
of nitrogen. Since submerged cultureTecIuiiques are more applicable to industrial 
operations it was deemed advisable to investigate this method of producing fat in 
some detail in order to evaluate its economic potentialities. 

This paper deals mainly with the synthesis of fat by Rhodotorula 
gracilis grown in submerged cultures on a molasses medium* however, 
some experiments were made with glucose for purposes of comparison. 
Rh. gracilis has been reported to be one of the most efficient of the fat 
synthesizing organisms which can be grown in submerged cultures. By 
theoretical reasoning, Rippel (14) concluded that the fat coefficient— 
or grams of fat synthesized from 100 grams of carbohydrate utilized— 
could hardly exceed 15. However, Enebo, Anderson and Lundin (5) 
have shown that Rh. gracilis can attain a fat coefficient value as high as 


419 



420 


S. C. PAN, A. A, ANDBEASEN AND P. KOLACHOV 


16-18 when inverted commercial sucrose is used as a source of carbo¬ 
hydrate. 

Efficient aeration and proper nitrogen supply are known to be salient 
factorsJn microbial fat synthesis^Smedley-MacLean et al. (16,17) have 
made thorough studies of these factors with brewers’ yeast, and Heide 
(8) and Raaf (12) have reported similar studies with Endomyces vemalis 
The reviews by Smedley-MacLean (18), Bernhauer (2) and Klein- 
zeller (9a) give detailed discussions of the factors involved in fat syn¬ 
thesis. However, data for the submer ged cultivation of fat producing 
organisms on a molasse s medium^were lacking^ Since molasses offer s one 
_of the most economical growth media, these variables wer e considered 
worthy of investigation— ~ 

Experimental Procedure 

Organisms 

The culture of Rh. gracilis used in these experiments was obtained from Dr. H. 
Lundin of the Royal Technical University, Stockholm, Sweden. Other fat synthe¬ 
sizing organisms were also evaluated, including Endomyces vemalis , Torulopsis pair 
cherrima, Oospora laclis t Torulop sis lipojera* Candida reukaufii , and Starkey’? soil 
yeast 1 (19). However, RhoSotdhda gracilis was found to be far superior to any of these 
organisms, both in fat content and fat coefficient. The fat content of Starkey’s soil 
yeast was found to be comparable to Rh. gracilis but the fat coefficient and rate of 
sugar utilization were much lower. Therefore, Rh. gracilis was employed in all the 
experiments reported in this paper. 

Culture Methods 

Formulas for the 3 media that were used in these studies are given in Table l._ 
Yeast cultures were transferred once a weekon agar slants c ontainin g glucosg 
(Mediumll) and molas se ? (Medium III) . After 48-72 hr. incu bation at 28°C ., these 
slants wer e used as inocula for the fat synthesis experi ment s. The surface growth from 
an agar slant was suspended in sterile water immediately Bef ore inoculation. The 
inoculated medium contai ned appro xima tely 5million cells/ml. This type of inoculum 
was found to be just as satisfactory as that Obtained from aerated or shaken cultures. 

Cultivation Procedure 

In cultivation utilizing glucose as the sugar source, the glucose was sterilized 
separately and added to the clarified, sterilized basal medium. All glucose media were 

1 These cultures were obtained from Professor E. McCoy of the University of 
Wisconsin, Madison, Wisconsin, Dr. L. J. Wickerham of Northern Regional Research 
Laboratory, Peoria, Illinois, and Dr. R. L. Starkey of Rutgers University, New Bruns¬ 
wick, New Jersey. 







RHODOTORULA GRACILIS FAT SYNTHESIS 


421 


TABLE I 

Composition of Media? 


Medium* I 


Medium II 


Medium III 



o./l. 


a./L 


O'/l- 

(NHiJjSO, 

1.0 

KHjPO* 

1.0 

(NH,) 2 SO, 

1.0 

KHjPO* 

1.0 

Yeast extract 

5.0 

KIIjPO, 

1.0 



(Difco) 




MgS0«-7H 2 0 

1.0 



Carbohydrate* 

20.0 

NaCl 

0.5 

Carbohydrate 11 

20.0 

pH 

4.0--4.8 

CaCl 2 -6H 2 0 

0.5 

pH 

6.0 



FeCl,-6H 2 0 

0.005 





Yeast extract 

1.5 





(Difco) or malt 






^ extract (Difco) c 






Carbohydrate* 

Varying 





pIT (with H 1 SO 4 ) 

4^PIX 






a All salts and glucose were of C. P. grade. All media were sterilized at 15 p.s.i. for 
30 min. 

b This medium was essentially the same as that employed by Enebo et al. (5). 
e Yeast extract and malt extract were found interchangeable. The yeast extract 
contained 8 . 8 % nitrogen. ~ 

* Carbohydrate refers to either glucose or sugars contained in molasses . Blackstrap^, 
molasses obtained fr om U . S. Sugar Corp., Clewist on, Fla., was used throughout the 
present experimentation, 'fhis molasses had_a sp. gr. o f 1.44 (25°C. ), a total sugar, 
content of 46.5%, and a nitrogen content of 1.0%. 


buffered with sterile CaC0 3 (0.5 g./l.) to check the drop in pH caused by the 
formation of H 2 S0 4 from (NH^SOi. In cultivation utilizing molasses, the mediujp 
was prepared as fo llows : sufficient water was added to $00 g. of molasses ta make a 
tofaTvolume of 500jnl. The diluted molasses was centrifuged at 2500 r. p.m. fo rJ5 
min.Jbo remove the suspended insoluble matter. Aliquots of the centrifugate*^were 
added to solutions oTthe nutrient salts andThe resultingjnedium was sterilized at,. 
15 p.s.i. for 30 min. Separate sterilization of the dilute molasses and nutrients gave 
the same results as when all the ingredients were sterilized together. Unless stated 
otherwise, the media contained 4% sugar (either glucose pr to tal s u gar a s glucose 
after hydrolysis ). Aft gr in oculation, one drop of Foami cide A 3 /100 ml. of medium wga 
added to s upp ress foaming^ 

Aeration experiments were made in 2 1., cone-shaped flasks, each containing 300 ml. 
of medium; shake cu ltures were grown in 500 ni l. Erlen meyer flasks, each containing 

* The centrifugate still contained traces of insoluble solids. They were determined 

in uninoculated blank tests. Corrections were then applied to the dry weight of yeast, 
usually less than 2 %. « 

* Wyandotte Chemicals Corp. J 













422 


S. C. PAN, A. A. ANDREASEN AND P. KOLACHOV 


80 ml. of me dium. Humidified sterile air w as passed into the media th rou gh AloxjLt e 
gas diffuse^. The rate^Tlaeralion'lvas measured by a wet-test gas meter. Unless 
stated otherwise, 0.5-0.8 volume of air/volume of medium/min. was used. In shaking 
experiments, a sh aker with a stroke length of 4 qhl wiLS .operatedJOO cycles per 
min. All flas ks we re incubated at the o ptimum te mpera ture of the yeast, 27- 29°C. 

Uampfes were taken periodically dunng^ultivation and analyzed for sugar. When 
all available sugar was utilized, the resulting medium was'analyzed For dry weight 
of yeast and for total fat content. The yeast population was determined by means of a 
Neubauer counting chamber; the culture was checked for contamination at the same 
tiin£. 


Analytical Procedures 

Dry Yeast 

A 15 ml. sample was diluted to 40 ml. and centrifuged at 3500 r.p.m. in a weighed 
40 ml. Pyrex centrifuge tube. The cells were washed once with Af/15 KH2PO4 (7). 
After drying for 18-24 hr. at 60°C. under a vacuum of 27-28 in. Hg, the tubes were 
weighed again. Drying to constant weight at 110°C. resulted in less than 2% addi¬ 
tional loss in weight. 


Reducing Sugar 

Sugars were determined by the micro method of Shaffer and Somogyi (15). The 
m olasses med ium was analyzed for total sugar as glucose after hydrolyzing in 0.35_ 
N HC1 for 5 min. fi Ta boiling water bathT ~~ ~~ 


Fat 

A method for fat determination suggested by Gray (7) and based on one by Smed- 
Iqy-MacLean (16) was followed. The dried yeast cells were hydrolyzed in 1 N HC1 for 
2 hr. in a boiling water bathj filte red, d ried and extracted for 8 hr. with anhydr^ms 
ether in a Gold fisch e xtractor . It was later found that fat analyses coul d be ob tained 
satisfac torily by hydrolyzing the yeast celTslfirecdly in the - medium, which had been. 
Acidified to 1 AT with H(J1. 


Experimental Results 

Preliminary Experiments in Glucose Media 

Fat synthesis was first studied using a glucose medium (Medium I) 
in order to have some means of evaluating the results obtained with 
molasses media. The results agreed in general with those reported by 
Enebo et al (5). Rh. gracilis was found capable of completely utilizing 
4 g. of glucose/100 ml. of medium in 4 days, producing over 30% of dry 
yeast (based on glucose utilized) with a fat content of 60% (based on 
dry yeast). The unusually high fat coefficient of 16 to 18 was also con- 




RHODOTORULA GRACILIS FAT SYNTHESIS 


423 


firmed. Data from a typical experiment in a glucose medium is pre¬ 
sented in Expt. 1, Table II. 

Analysis of fat content and estimations of cell population during 
cultivation confirmed the earlier view that the process can be divided 
into two phases, namely, a phase of protein formation, i.e., cell multi¬ 
plication, followed by a phase of fat formation (6,11). During the first 
24 hr. there was a rapid increase in cell population (from 5 millions/ml. 
to 600 millions/ml.) when practically no fat was synthesized. In the 
following 3 days the fat accumulated steadily while the cell population 
jemained practically constant. 


TABLE II 

Fat Synthesis in Molasses vs. Glucose Media " 


Expt. 
no. 1 

. 

Sugar 

substrate 

Medium 

Sugar utili¬ 
zation 
(hawed on 
initial 
sugar) 

Yjcld of dry 
yeast 
(based on 
sugar 
utilized) 
(d) 

Fat content 
of dry 
yeast 

( B ) 

Fat coefficient 
(.4)X(B)/100 

1 

Glucose 

Medium P 

per rent 

97.7 

f cr cent 

31.2 

per cent 

59.0. 

18.4 

2 

Glucose 

Medium P with 

61.5 

23.1 

56.7 

13.1 



Y.E. C left out 





3 

Molasses 

Medium P 

87.0* 

40.6 

28.3 

11.5 

4 

Molasses 

Medium P with 

Y.E/ left out 

86 .8 d 

39.2 

31.3 

12.3 

5 

Molasses 

(NH 4 ) 2 S0 4 1 g./l. 
KH*P0 4 1 g./l. 

88 .2 d 

37.8 

35.9 

13.6 

6 

Molasses 

No nutrient added 

56.0 

37.4 

47.6 

17.7 


° Cultivation was carried out with aeration by sparging for 4-6 days. Initial sugar 
concentration was 4% in all cases. 

6 Medium I contained yeast extract, (NHd^SO*, KH 2 P0 4 , MgS0 4 -7H*0, NaCl, 
CaCl r 6H*0, and FeCl*-6H 2 0. 

c Y.E. denotes yeast extract. 

d Maximum sugar utilization was 90%, of initial sugar. These figures represent 
complete utilization. 

The effect of the concentrations of nitrogen nutrients, such as (NH 4 )*S 04 and yeast 
or malt extract, upon fat synthesis also agreed with many previous reports (5,8,12,19).^ 
Media rich in nitrogen support an abundant growth of yeast but the fat content is 
always low. However, as pointed out by Enebo et al. (5), there is an optimum concen¬ 
tration of nitrogen which must be furnished if the organisms are to synthesize the 
maximum amount of fat. Experiments with glucose showed that the optimum con¬ 
centration of nitrogen can be furnished by 1 g. of (NH 4 )*S 04 plus 1.5 g. of yeast ex¬ 
tract/1. of medium. 



424 


S. C. PAN, A. A. ANDREASEN AND P. KOLACHOV 


The report by Enebo et al. also emphasized the importance of neutralization during 
cultivation. The HiSO« liberated from (NH«) > SO < lowers the p H o f the medium 
.markedly,, resulting in an inactiv ation of the* yeas t. In the present investigation, the 
addition of a small amount of CaCOi was found convenient for neutralizing the acid 
formed. . ~~ 

Comparison of Fat Formation in Molasses and Glucose Media 

Experiments with molasses media show that molasses apparently 
contains appreciable amounts of the nutrients that are required by_the 
yeast. Total cell population and yield of dry yeast were higher in 
molasses media than in glucose. A population of over one Trillion 
cells/ml. was attained in molasses media as against 500-600 millions/ml. 
in glucose. Table II presents a comparison of growth and fat synthesis 
in these two kinds of media. This table also shows that omission of 
yeast extract from molasses media had no effect on growth and.fat 
synthesis but there was a marked reduction in the yields of yeast and 
fat when it was omitted from glucose medium. With (NH^SO* and 
KHjPO« as the only nutrients added to a molasses medium (Expt. 5) 
the yeast grew as well and synthesized fat even better than when 
grown in a molasses medium containing all the added nutrients of 
Medium I (Expt. 3). Molasses, without any added nutrients, supported 
fairly good growth although the sugar utilization was incomplete 
(Expt. 6). With the exception of the medium containing only molasses 
(Expt. 6), the nutrients present in the molasses media produced greater 
yields of yeast but the fat content was considerably less than from 
glucose media. From these results it may be concluded that it is possible 
to obtain complete sugar utilization from molasses media without 
reducing the yield of fat if the proper concentrations of (NH^jSOi and 
KH 2 PO 4 are present. 

The pH of a mo lasses-water m ixtu re containing 4% su gar was 4.6r 
. 4.7; therefQrfi*.no initial pH ad justme nt was ne cessary . During culti¬ 
vation, the pH of the medium gradually rose to 5.6-6.2 and conse¬ 
quently no neutralization was required. This rise in pH was probably 
due to the utilization by the yeast of organic acid salts present in the 
molasses media. 

Unlike glucose, utilization of sugar in a molasses medium always 
stopped when 88-90% of the initial sugar was consumed. Apparently 
the residual 10-12% consisted of non-fermentable sugar and other 
reducing substances which analyzed as sugars. 




RHODOTORULA GRACILIS FAT SYNTHESIS 


425 


Effect of KHtPOi and (NHJiSOi upon Fat Synthesis from Molasses 

To find the optimum concentrations of (NH^SCh and KH s PO« for 
fat synthesis in molasses media, experiments were performed with 
varying concentrations of these two compounds, added separately and 
in combination. Cultures supplemented in this manner were tested, 
using both aerated and shake flask cultures. 

As shown in Fig. 1 , KH 2 PO 4 alone improved the sugar utilization to 
a limited extent (from 50.2% to 62.0%) but had no effect on the fat 
coefficient. Fig. 2 shows that (NH 4 ) 2 S 04 was very effective in improving 
‘sugar utilization; the yield of yeast increased slightly at higher 
(NH 4 ) 2 S 04 concentrations and no appreciable reduction was observed 
in the yield of fat with a concentration of (NH 4 ) 2 S 04 as high as 2 g./l. 
of medium. However, 3 g./l. caused a marked decline in both fat con¬ 
tent and fat coefficient, and sugar utilization was incomplete. Presum¬ 
ably, the well nourished yeast lost some of its fat-synthesizing power 
and growth stopped before the sugar and (NH 4 ) 2 S 04 were completely 
utilized because other nutrients or growth factors had been exhausted. 


Sugar 

Utnijation 



Fia. 1. Effect of KHjPC >4 on fat synthesis in molasses medium. Cultivations were 
made in flasks with sparged air and in shake flasks, with KH2PO4 as the only nutrient 
added to the medium. Initial sugar concentration, 4 g./lOO ml. of medium. Cultivation 
was stopped after 4&-60 hr., since longer periods of time did not increase the amount 
of sugar utilized. 



426 


S. C. PAN, A. A. ANDREASEN AND P. KOLACHOV 


Sugar uttiitatioA 
Y>«ld .f Yn»4 



Fig. 2 . Effect of (NH 4 ) 2 S0 4 on fat synthesis in molasses medium. Cultivations 
were made in flasks sparged with air and in shake flasks, with (NH 4 ) 2 S0 4 as the only 
nutrient added to the medium. Cultivation was stopped after 48-60 hr. Initial sugar 
concentration, 4 g./lOO ml. of medium. 


Table III shows the effect of (NPL^SCb and KH 2 PO 4 added in 
combination to the medium. It was found that 0.5 g. of (NH^SCVl. 
was adequate to effect complete sugar utilization when 0.5 g. of KH 2 PO 4 
was also added (Expt. 4). As little as 0.25 g. of (NIE^SCVl. would 
bring about complete sugar utilization in the presence of 1 g. of 
KH 2 PO 4 /I. (Expt. 8 ). Apparently, the presence of KH 2 PO 4 enhanced 
the effect of (NH 4 )tSO«. On the other hand, 1.5 g. of (NI^HSCb/l. 
was high enough to cause a decline in fat yield if an equal amount of 
KH 2 PO 4 was added (Expt. 13). It was found that when higher con¬ 
centrations of these two compounds were used, the yield of fat was 
further reduced (Expts. 6 and 12). It is interesting to note that with 
1.0 g. each of (NH 4 )2S04 and KH 2 PO 4 /I., the fat coefficient showed 
remarkably wide variations. Results of two extreme cases are shown 
in Expts. 10 and 11. Such variation is to be expected, since this concen¬ 
tration of the two salts approaches the upper limit for efficient fat 
synthesis. It is possible that phosphate enables the yeast to utilize 
more of the nitrogen contained in molasses since the addition of phos¬ 
phate markedly enhanced the effect of (NH 4 ) 2 S 04 . 



RHODOTORULA GRACILIS FAT SYNTHESIS 


427 


Effect of Aeration upon Fat Synthesis 

Varying the rate of aeration from 0.1 to 1 volume of air/min./volume of medium 
had no effect on the yields of yeast or fat. Aer ation by shaking proved to be just as 
eff ective as by sparging , hut stationary cultures produced very inferior results (see 
Table IV). The effect of aeration on the rate of sugar utilization will be discussed 
later. 

TABLE III 


Effect of (NH^iSOa and KH 2PO4 on Fat Synthesis in Molasses Media a 


Exp!,. 

No. 

KILPO 4 

added 

(NIL) 2 SO< 

added 

1 Sugar utili- 
! zation (hawed 
on initial 
sugar) 

Yield of dry 
yeast (based 
on sugar 
utilized) 

(A) 

Fat eontent of 
dry yeast 

m 

Fat coefficient 
(A)X(£)/100 


u./L 

0-/1- 

per rent 

per rent 

per cent 


1 




37.8 

46.9 


2 



68.0 fc 

38.2 


■ I 

3 

0.5 


75.0'' 

30.0 



4 

0.5 

0.5 

88.5 

37.6 

48.3 

18.3 

f> 

0.5 

1.0 

88.5 

36.8 

46.7 

17.3 

6 

0.5 

2.5 

89.0 

41.0 

20.9 

8.6 

7 

1.0 

0.0 

72.0'' 

36.5 

49.7 

18.2 

8 

1.0 

0.25 

88.0 

37.6 

49.2 

18.5 

^ 1 

1.0 

0.5 

88.4 

37.0 

48.7 

18.0 

10' 

1.0 

1.0 

88.6 

36.2 

46.8 

17.0 

11' 

1.0 

1.0 

88.9 

37.4 

32.6 

12.2 

12 

1.0 

2.0 

89.0 

40.5 

24.2 

9.8 

13 

1.5 

1.5 

88.9 

38.8 

28.3 

11.0 


n Cultivation was carried out in both flasks with sparged air and shake flasks with 
(NHi) 2 S 0 4 and KH2PO4 as the only added nutrients. Initial sugar concentration was 
4%. Cultivation was stopped at the end of 48-60 hr., beyond which practically no 
more sugar was consumed. Results are averages of at least two experiments. 
b These figures indicate incomplete sugar utilization. 

c Data of two extreme cases are listed here to show the wide variation in results. 

Fat Synthesis at Higher Sugar Concentrations 

All of the above experiments were made with media containing 4% 
sugar. However, it was considered advisable to determine whether the 
same high fat yields could be maintained using higher levels of sugar. 
It was found that Rh . gracilis could completely utilize 6 or 8 g. of 
sugar/100 ml., in either molasses or glucose media, and yields of yeast 
and fat, based on sugar utilized, were just as high as with lower con¬ 
centrations (Table V). 




428 


S. C. PAN, A. A. ANDREASEN AND P. KOLACHOV 


TABLE IV 

Effect of Aeration upon Fat Synthesis 11 


Aeration* ratio 

Cultivation 

period 

Sugar utilisa¬ 
tion (baaed on 
initial sugar) 

Yield of dry 
yeast (based 
on sugar 
utilized) 

(A) 

Fat content 
of dry yeaat 

(B) 

Fat coefficient 
(A)X(*)/100 

1:1 

hr. 

40 

per cent 

89.7 

per cent 

36.5 

per cent 

49.6 

18.1 

1/3:1 

60 

88.5 

36.6 

48.8 

17.9 

1/8:1 

60 

88.8 

37.2 

48.2 

17.8 

0.1:1 

72 

88.7 

36.2 

48.6 

17.6 

Stationary culture* 

12 days 



30.2 

12.2* 

Shaken culture^ 

J*L 

88.7 

37.4 

48.8 

C 18.3 

' J 







a Cultivation was carried out i n molasses medium with 1 g. (NHO1SO4/I. as addej , 
nutrient. Initial sugar concentration was 4% in each case. 

* Aeration ratio denotes the ratio, volume of air per min.: volume of medium. 

* 100 ml. of the medium were used in a 500 ml. Erlenmeyer flask. 

d Since the sugar utilization was very incomplete, experimental error in the yields 
must be high. 


TABLE V 

Fat Synthesis at Higher Sugar Concentrations in Molasses Media a 


Initial 

sugar 

(NH0«SO« 

added 

Cultiva¬ 
tion period 

Sugar utili¬ 
zation (based 
on initial 
sugar) 

Yield of dry 
yeast (based 
on sugar 
utilized) 

(A) 

Fat content of 
dry yeast 

(B) 

Fat coefficient 
(A)X(R)/100 

0./1OO ml. 

9-/1 . 

hr. 

per cent 

per cent 



4.40 

1 

60 

89.Q 

34.3 

B 

■■RTiiaHI 

4.40 

2 

60 

89.0 

38.6 

B 

16.9 

6.65 

1 

85* 

60.3 d 

— 

■HI 

— 

6.65 

2 

85* 

89.2 

34.8 

46.1 


8.80 

1 

108* 

88.8 

35.9 

50.4 

18.2 

8.80 

4 

108* 

89.3 

35.9 

41.6 

14.8* 


a Cultivation was carried out in shake flasks. 

6 There was an initial lag perio3^T24^r. [see Fig. 5). 
c IwTwas an Init ialTag periocQP S hr. (see Fig. 5). 
d Sugar utilization stopped before completion, probably due to contamination. 
' Low fat coefficient was obviously due to high nitrogen content. 



















RHODOTORULA GRACILIS FAT SYNTHESIS 


429 


Factors Influencing the Rate of Sugar Utilization 

In nearly all experiments, the diminution in sugar content of the 
medium was determined byjperiodic analyses. Some of the results are 
shown in Figs. 3, 4, and 5, where sugar concentrations during cultiva¬ 
tion are plotted against time. 


Contact ration 
e/iOOml 



Fig. 3. Effect of nutrients on rate of sugar utilization in molasses medium. 
Cultivations were made in shake flasks. 


Sugar 

Concentration 

g/>ooml 



Fio. 4. Rate of sugar utilization at different initial sugar concentrations. Culti¬ 
vations in glucose media were made i n flasks spa rged with air; shake flasks were used 
for cultivations made in molasses media. 



430 


S. C. PAN, A. A. ANDREASEN AND P. KOLACHOV 


Si«9ar 

Concentration 
9/100 ml. 



Fig. 5 . Effect of a eration on rate of sugar utiliza tion in molasses medium . Culti¬ 
vations were made in molasses media containing^ 1 g./l. of (NH 4)2804 as added 
nutrient. Aeration is expressed as volume of air/volume of medium/min. 

Fig. 3 shows that the rates of sugar utilization during the first 24-30 
hr., in media containing 4% sugar, are the same in molasses media with 
or without the added nutrients—(NFL^SOi and KH2PO4. Although 
sugar utilization in molasses media with insufficient nitrogen (0.5 g. of 
(NH^jSOi/l. or lower) failed to proceed to completion, the rates in 
all media which allowed complete sugar utilization were identical— 
independent of the amount of (NH^gSOt and KH2PO4 added. The 
substances contained in molasses, rather than the added nutrients, 
evidently are the factors which govern the rate of sugar utilization. 

At sugar concentrations higher than 4%, growth was not initiated 
until after a long lag phase, the length of which depended upon the 
initial sugar concentration (Fig. 4). No such phenomenon was observed 
in glucose media (Fig. 4). However, as soon as growth was initiated, 
the sugar was completely utilized within 48-GO hr. regardless of the 
initial sugar concentration. That is, the higher the concentration of 
molasses in the medium, the greater the rate of sugar utilization. Fig. 4 
also shows that the rates of sugar utilization in molasses media were 
considerably higher than in glucose media at corresponding initial sugar 
concentrations. This acceleration of sugar utilization seems to conform 
with the observation that molasses media produced a greater number of 
cells and a higher dry weight of yeast than did glucose. 

Aeration had a limited effect on the rate of sugar utilization. Curves 



RHODOTORULA GRACILIS FAT SYNTHESIS 


431 


in Fig. 5 show that aeration ratios (volume of air/volume of medium/ 
min.) of 1/3:1 and 1/8:1 produced identical curves for sugar utilization. 
The rate of sugar utilization was higher at an aeration ratio of 1:1, and 
lower at 0.1:1. Shake flask aeration was apparently equivalent to an 
air-sparging ratio of 1/3:1. Sugar utilization was extremely slow and 
incomplete in stationary cultures (Table IV). From a practical point of 
view, aeration ratios higher than 1/3:1 have no particular advantage. 


Yeast Consumption oj Fat Formed 

It has been pointed out by many investigators that some micro¬ 
organisms are capable of util izing the f at stored in their cells as a source 
gf_ energy (12,20) ; Rh. gracilis was found to be one of these . Table VI 
shows the d iminution of fat which resulted from prolonged cultivation 
after all available sugar was.exhausted. In both glucose and molasses 
media, marked reduction in the fat content occurred under such a 
treatment, amounting in 14 days, to 22.2% and 40%, respectively. A 
comparison of the loss in dry weight of yeast with that of fat indicates 
that fat is consumed in preference to other cell constituents. Therefore, 
to obtain the maximum recovery of fa^the yeast must be harvested as 
soon as sugar is completely utilized. 


TABLE VI 

Consumption of Fat on Prolonged Incubation a 


Carbohydrate 

substrate 

Incubation 

period 

Dry weight of > east 

Fat 

Amount 

found 

I .OSS 

Amount 

found 

As content of 
diy yeast 

Loss 


days 

0./100 ml. 

per rent 

y./100 ml. 

per cent 

per cent 

(ilucose 6 

4 r 

1.23 


0.810 c 

65.0 



7 

1.18 

4.1 

0.763 

64.5 

5.8 


14 

1.04 

15.4 

0.630 

60.6 

22.2 

Molasses 6 

3 C 

1.32 c 


0.625 c 

47.8 



7 

1.25 

5.3 

0.559 

44.8 

10.3 


14 

1.06 

19.2 

0.375 

35.4 

40.0 


n Cultivation was carried o ut with aeration by sparging. initial sugar was 4% . 
b Medium I was used in experiments with glucose. 1.0 g. of (NHOaSCVl. was used 
as added nutrient in the molasses medium. 
c These values represent the results immediately after sugar was all consumed. 



432 


8. C. PAN, A. A. ANDHEASEN AND P. KOLACHOV 


Discussion and Conclusion 

Experimental data on growth and fat synthesis by Rh. gracilis in 
glucose media agree in general with many earlier reports on different 
fat-producing microorganisms. The importance of an optimum nitrogen 
concentration of the medium for maximum fat synthesis, emphasized 
by Enebo, Anderson and Lundin, has also been demonstrated in the 
present investigation. Under optimum conditions, the unusually high 
fat coefficient, 16-18, which was first reported by Enebo et al., was 
confirmed. 

Growth of Rh. gracilis and its fat synthesis in molasses media present 
many interesting features. The higher values for cell population, yield 
of dry yeast, and rate of sugar utilization for fat synthesis which were 
obtained in a molasses medium suggest that molasses conta ins sub ¬ 
stances which are nutritive or stimulatory to growth and increase th e 
rate of fat sy nthesis of the yeast. TheTact that the rate of sugar utili¬ 
zation is dependent upon The molasses concentration, rather than on 
added nutrients, e.g., (NH^SOi and KH2PO4, further supports this 
assumption. The question remains, however, whether the higher rate 
of sugar utilization is a result of higher cell population, a stimulatory 
effect, or both, since no attempt was made to correlate these factors. 

The presence of a Jag phase in a molasses medium containin g more 
lhan 4% sugar indicates the presence of substances which retard the 
initiation of growth. This retarding effect becomes more marked as the 
sugar concentration is increased. However, as soon as growth starts, 
this effect disappears and the stimulatory substances become functional. 

Due to the presence of these nutritive substances in molasses, the 
influence of adding (NH^SC^ and KH 2 P0 4 becomes more critical 
than in a glucose medium. An insufficient supply of these two com¬ 
pounds in a molasses medium results in incomplete sugar utilization, 
while an excess causes a marked decline in fat yield. It appears* that 
(NH 4 ) 2 S 0 4 has a direct effect, while KH 2 PO 4 enhances the action of the 
former. A concentration of 1.0 g. of (NH 4 ) 2 S 0 4 / 1 . alone, or 0.5 g. each 
of (NH 4 ) 2 S 0 4 and KH 2 P0 4 /1. proved to be satisfactory for complete 
sugar utilization and resulted in a fat coefficient which was as high as 
that obtained with a glucose medium, 16-18. 

The increase o f_pH to 6 du ring gro wth in molasse s me dia is of practica l 
^advantage sinc e it simplifies the process by eliminating pH adjustment 
during cultivation." —-— -—— 



RHODOTORULA GRACILIS FAT SYNTHESIS 


433 


Acknowledgments 

The authors wish to express their appreciation to Dr. M. C. Brockmann for his 
suggestions in the course of this investigation and to Mr. L. W. Nicholson for his 
assistance in the preparation of the manuscript. 

Summary 

The unusually high fat coefficient of 16-18, and the importance of 
an optimum nitrogen concentration, which have been reported pre¬ 
viously for Rhodotorula gracilis in a glucose medium, were confirmed. 
It was also shown that Rh. gracilis could be grown successfully in a 
molasses medium to produce fat with the same efficiency as in a glucose 
medium, provided proper amounts of nutrients—(NH 4 ) 2 S0 4 alone or 
with KH 2 P0 4 — are added to the medium to insure complete sugar 
utilization. 

It was found that certain substances present in molasses are capable 
of accelerating the sugar utilization and, therefore, the rate of sugar 
utilization is a function of the molasses concentration. This effect may 
be due to the increase in cell population under these conditions. 

References 

1. Balls, A. K., Production of Fat by Oidiurn ladis , O. T. S. Kept. PB 1320 (1945). 

2. Bernhauer, K., Ergeb . Enzymforsch . 9, 297 (1943). 

3. Damm, H., Chem. Ztg. 67, 47 (1943). 

4. Damm, H., U. 8 . Pal . 2,346,011, April, 1944. 

5. Enebo, L., Anderson, L. G., and Lundin, H., Arch. Bioche m. 11 , (1946 ). 

6 . Fink, H., HaehnTH., and Hoerburger, W., Chem. Ztg . 61, 689, 723, 744 (1937). 

7. Gray, W. D., J. Bad. 66 , 53 (1948). 

8 . Heide, SLj _Arch. Mikrobiol. 10, 135 (1939). 

9. Kleinzeller, A^lBiochem. J. 38, 480 (1944). 

9a. Kleinzeller, A., Advances in Enzyrnol. 8 , 299 (1948). 

10. Nord, F, F., Dammann, E., and Hofstetter, H., Biochem. Z. 286, 254 (1936). 

11 . Prescott, S. C., and Dunn,. C. G., Industrial Microbiology. McGraw-Hill, 

New York, 1940. 

' 12. Raaf, H., Arch . 131 (1941). 

13. Rippel, A., Noiuni^erwcAa/^en 31, 248 (1943). 

14. Rippel, A., Arch. Mikrobiol. 11, 271 (1940). 

15. Shaffer, P. A., and Somogyi, M., J. Biol. Chem. 100, 695 (1935). 

16. Smedley-McLean, I., and Hoffert, D., Biochem. J. 16, 370 (1922). 

17. Smedley-MacLean, I., and Hoffert, D., ibid. 17, 720 (1923); 18,1273 (1924); 

20, 343 (1924). 

18. Smedley-MacLean, I., Ergeb. Emymjorsch. 6, 285 (1936). 

^f9. Starkey, R. L. t J. Bad. 61 t 33 (1946)— 

IST White, A. G. C., and Werkman, C. H., Arch. Biochem. 17, 475 (1948). 

21 . Nord, F. F., Fiore, J. V., and Weiss, S., ibid. 17, 345 (1948). 



Studies on the Mechanism of Action of Ionizing 
Radiations. III . 1 The Plasma Protein of Dogs 
after X-Ray Irradiation. An 
Electrophoretic Study 

J. A. Muntz, E. S. Guzman Barron and C. L. Prosser 

From, the Argonne National Laboratory and the Chemical Division, 
Department of Medicine, The University of Chicago, Illinois 
Received February 14, 1949 

Introduction 

There have been several conflicting reports of altered concentrations 
in serum proteins following the administration of X-rays. In a series of 
human patients, Herzfeld and Schinz (1) reported a diminution of total 
protein and a decreased A/G ratio following therapeutic irradiation 
with X-rays. Wichels and Behrens (2) found similar changes in certain 
cases only, while Breitlander and Lasch (3) could find no significant 
variation in the blood proteins following irradiation. In an experimental 
study on dogs, Davy (4) found changes in the blood proteins following 
the administration of 500 roentgen units (r). There was an immediate 
fall in albumin, which returned to normal after 48 hr. Variations in the 
blood globulins were not consistent. 

Results obtained by the earlier workers are difficult to interpret 
because the X-ray dosage was not given. Then, too, these and later 
workers were handicapped by the use of the inadequate salt fractiona¬ 
tion methods for the determination of the serum proteins. With the 
method of electrophoresis as developed by Tiselius (5), the analytical 
difficulties have been largely overcome. It has been demonstrated that 
sera from any given animal species have a characteristic pattern (6,7). 
The dog, in particular, has been extensively studied (8). 

This paper reports the effect of X-rays on the plasma protein pattern 
of dogs. The experiments were performed during the years 1944-45. 

1 Papers I and II of these studies have appeared in the Journal of General Physi¬ 
ology. 


434 



STUDIES ON ACTION OF IONIZING RADIATIONS 


435 


Experimental 

Mongrel dogs were employed that had been treated with “Lederle" antidistemper 
vaccine. They were fed a stock diet consisting of Friskies, supplemented with ground 
meat twice weekly. X-rays, filtered with 0.5 mm. Cu and 1 mm. A1 were administered 
in one total body dose from a 200 kv peak machine. 

Control blood samples were drawn before the X-ray treatment and periodically 
thereafter until the animal died. Blood was collected in oxalated tubes and, after 
centrifugation, the plasma was withdrawn for analysis. 

Total protein determinations were performed by a micro Kjeldahl digest ion followed 
by ncsslerization or distillation and titration. Non-protein nitrogen was determined 
on a filtrate obtained by precipitating the protein with trichloroacetic acid. 

Electrophoretic analyses of the sera were carried out by the moving boundary 
method of Tiselius (5) in an apparatus equipped with the Schlieren scanning device 
described by Longsworth (9). Three to four cc. of serum or plasma were dialyzed for 
24 hr. against 2 liters of veronal buffer (0.075 M , pH = 8.6). 

Electrophoresis was carried out for 2 hr. in a long (11 cm.) analytical cell at 1.2°C., 
using a current of 23 milliamps and an E.M.F. of 310 v. 

In one experiment (dog 128), samples of the terminal plasma obtained shortly before 
death were extracted with lipide solvents. One 10 cc. plasma sample was extracted by 
the modified Blix procedure (10) described by Zcldis et at. (11). A considerable amount 
of the dry protein failed to dissolve in the normal saline solution and electrophoretic 
analysis of the soluble portion showed that the greatest reduction in relative concen¬ 
tration had occurred in the complex fibrinogen fraction of the original plasma. 

Another 12 cc. sample of the same plasma was lyophilized prior to extraction with 
lipide solvents. The dry, fluffy powder was extracted with two 80 cc. portions of 
absolute ethyl alcohol at — 70°C. for 20 min. each, followed by one portion of alcohol- 
ether (7:3), and finally with two 80 cc. portions of absolute ether, all at — 70°C. for 
20 min. After removing the last traces of ether with a vacuum pump at room tempera¬ 
ture, the resulting dry powder was readily soluble in normal saline solution. 

Results 

A typical normal pattern of the rising boundaries in dog plasma is 
shown in Fig. 1 . In most dogs 8 distinct components may be distin¬ 
guished. Following the nomenclature of Zeldis and Ailing ( 8 ), they have 
been designated as a r , a 2 -, <* 3 - and arglobulins; 0r, / 82 -globulins; 
fibrinogen (<£) and 7 -globulin. The fibrinogen in dog plasma does not 
separate out as a distinct component; it migrates with the /3 2 -globulin 
and serves to increase this peak above that seen in serum. 

Even these 8 components are not electrophoretically homogenous. As 
shown in Fig. 1. the a 2 -globulin, as well as the j3i-globulin, has a split 
peak, indicative of inhomogeneity. In dogs dying of X-ray injury, the 
pattern becomes less complex, and many of the small peaks seen in the 
normal pattern appear to merge. Sometimes there are only 2 a-globulins 



436 J. A. MUNTZ, E. S. GUZMAN BARRON AND C. L. PROSSER 

in the pathological sera. These seem to have the mobility of the a 2 - and 
arglobulin of normal plasma. It will be shown that the concentration of 
arglobulin is significantly increased in injured dogs. The a-globulins 
have been divided into 2 fractions for purposes of calculation: ou- and 
arglobulin are estimated together as well as a r and a<-. In many cases 
Arglobulin is poorly resolved and tends to merge with the 02 , <t> compo¬ 
nent; consequently all 3 components were estimated together. 


FIGURE I 



RISING BOUNDARIES 

Plasma Protein Changes Following a Single X-Ray Treatment 

Protein changes in the sera of 5 X-ray treated dogs are shown in 
Table I. Dogs 44, 111 and 130 received a lethal dose of X-rays. They 
maintained their food intake for 7-8 days following X-ray treatment. 
Then they began to lose their appetite and stopped eating 3-4 days 
before death. Despite the maintained food intake, the plasma albumin 






















438 J. A. MUNTZ, E. 8. GUZMAN BARRON AND C. h. PROSSER 


FIGURE 2 (Port 3) 


TERMINAL PLASMA 





| TERMINAL PLASMA 1 

I LIPID EXTRACTED B I 

a i (iii 

--1 i-► 

RttlNA BOUNDARIES OESCCNOINB BOUNDARIES 


Despite this albumin diminution, the total protein concentration was 
increased somewhat above normal. This must have been due entirely 
to the increased a r and jS-globulins, since the plasma volume in this 
terminal period was also increased slightly above normal. 

When normal dogs were starved for 3 or 4 days, no pronounced 
changes were observed in the plasma protein pattern. Hence, the 
alterations seen in X-rayed dogs are not due to the diminished food 
intake during the terminal period. 

Dogs treated with a sublethal dose of X-rays did not show any 
marked changes in their plasma protein pattei'ns (dog 36), or they had 
some changes 14 days after X-ray which returned to normal as recovery 
occurred (dog 24). On the other hand, a dog that died of distemper had, 






STUDIES ON ACTION OF IONIZING RADIATIONS 


439 


TABLE I 


Electrophoretic Analyses of Serum and Plasma Proteins on X-rayed Dogs 




Days be¬ 
fore or 
after 
X-ray 





Globulins 


Dog no.® and 
X-ray dose 

Serum or 
plasma 

N.P.N. 

mg.-% 

Total 

protein 

Albumin 

ai+ai 

ai +04 

0i +0i +$ 

7 






Per cent of total protein 

44 

Plasma 

-10 

36 

6.08 

44 

11.7 

14.3 

18.9 


300 r 

Serum 

-2 

34 

5.84 

48 

11.4 

13.2 



% 

Serum 

+7 

31 

5.36 

48.7 

10.6 

13.1 

18.1 

mm 


Serum 

+ 14 

67 

6 .11. 

28 

11.8 

28.1 


BB 


Serum 

+ 15 

47 

6.50 

23.8 

10.8 

31.7 


m 

111 

Plasma 

-7 

27 

4,35 

43.5 

13.6 

12.6 

23.5 

6.9 

350 r 


-3 

31 

5.58 

42 

10.9 

13.8 

21.5 

11.1 


Plasma 

+7 

39 

5.13 

38.6 

16.6 

11.7 

22 



Plasma 

+ 12 

34 

5.85 

25.5 

17.1 

23.2 

27.5 



Plasma 

+ 13 

97 

6.43 

16.3 

11.2 

22.6 

44.7 


130 

Plasma 

-7 

28 

5.64 

40.2 

8.2 

10.8 

35.8 

5.1 

350 r 

Plasma 

-3 

32 

6.20 

41.5 

5.0 

8.9 

37.6 

6.8 


Plasma 

+7 

39 

6.35 

29.8 

10.9 

10.4 


6.8 


Plasma 

+ 12 

50 

6.78 

19.2 

13.9 

23.4 

35.5 

8.1 


Plasma 

+ 13 

97 

6.43 

16.3 

11.2 

22.6 

44.7 

5.3 

24 

Serum 

-8 

— 

6.13 

42.4 

16.3 

6.5 

32.6 

3.3 

250 r 

Serum 

-4 

49 

5.94 

44.2 

8.6 

13.1 

29.6 

6.2 


Plasma 

-4 

49 

6.38 

44.2 

8.3 

9.7 

29.6 

6.3 


Plasma 

+ 8 

30 

5.18 

39.0 

13.5 

13.5 

27.4 

6.5 


Serum 

+ 14 

41 

5.05 

41.0 

12.3 

12.9 

26.5 

6.9 


Plasma 

4-14 

42 

5.37 

34.8 

13.4 

13.4 

29.8 

8.4 


Plasma 

4“ 42 

— 

5.02 

39.9 

9.0 

9.6 

33.7 

8.0 

155 

Plasma 

0 

27 

5.86 

45.1 

14.0 

14.0 

17.6 

9.0 

400 r 

Plasma 

+6 

25 

5.62 

37.2 

13.5 

16.0 

24.9 

8.3 


Plasma 

+8 

27 

6.20 

37.9 

7.9 

22.4 

25.7 

6.1 

36 

Serum 

-8 

24 

5.82 

51.8 

13.2 

11.2 

16.8 

7.0 

200 r 

Serum 

-4 

29 

5.94 

56.6 

9.3 

11.0 

17.3 

5.9 


Plasma 

4-14 

— 

5.22 

47.0 

12.3 

13.0 

28.0 

8.2 

504 

(no X-ray) 

Plasma 

None 

19 

4.19 

27.7 

9.3 

30.8 

28.6 

3.6 


a Administered as a total body dose. 

— Indicates number of days before X-ray. 
4- Indicates number of days after X-ray. 











































440 J. A. MUNTZ, E. 8. GUZMAN BARRON AND C. h. PROSSER 

FIGURE 3 



Monad ptotmo (bifort X-Aoy) 12 <toy» of lor X-Roy 


Dog No. 504 (DUkmptr) 

•» 



Ttrminol plotmo (shortly bofort dooffi) 


in the terminal period, a plasma protein pattern that was very similar 
to that seen in dogs dying of X-ray injury (Fig. 3). 

Effect of Injecting Bovine Albumin on the Abnormal 
Plasma Protein Pattern 

Because of the reciprocal relationship between plasma albumin and 
arglobulin, it seemed desirable to see what effect replacement of the 
lost albumin would have on the plasma protein pattern. For these 
experiments crystallized bovine plasma albumin was used. Dogs were 
given a single lethal dose of X-rays (total body irradiation). When 
the animals had progressed to the terminal period, as evidenced by the 
onset of fever, the bovine albumin was injected. An amount of albumin 
equal to one-half of the total normal plasma albumin was dissolved in 
25 cc. of 0.86% NaCl, and the entire solution was injected intravenously. 
Samples of blood were withdraxn 5 min., 1 hr., 6 hr., and 24 hr. after 



STUDIES ON ACTION OF IONIZING RADIATIONS 


441 


TABLE II 


Electrophoretic Analyses of Plasma Proteins on X-ltayed Dogs. Effect of Bovine 
Serum Albumin Injected in the Terminal Period 








Globulins 


Dog no. 
and X-ray 
dose 

Day and hour of sampling 

N.P.N. 

mg.-% 

Total 

protein 

Albu¬ 

min 

oi +aj 

an +au 

01 +02 +4> 

7 






Per cent of total protein 

123 

1-23-46 

25.4 

6.02 

44.3 

10.6 

10.6 

25.9 

8.5 

350 r 

1-29-46 

25.1 

5.68 

48.1 

10.7 

11.7 

22.7 

6.9 

(X-rayed 

2-6-46 (Control) 

29.6 

6.30 

24.0 

14.3 

20.3 

34.5 

7.0 

on 

2-6-46 (1 hr.) 

30.7 

6.59 

35.4 

9.7 

20.0 

28.7 

6.2 

1-29-46) 

2-6-46 (7 hr.) 

26.8 

5.86 

33.3 

9.4 

20.6 

30.0 

6.7 


2-7-46 

24.8 

6.26 

30.6 

10.1 

21.2 

31.9 

6.2 


2-8-46 

32.0 

5.88 

24.3 

10.0 

23.1 

36.4 

6.1 


2-9-46 

51.8 

5.54 

23.5 

11.6 

29.4 

28.9 

6.8 

138 

3-5-46 

23.6 

5.65 

43.5 

11.3 

12.2 

25.3 

7.4 

400 r 

3-16-46 (Control) 

29.8 

5.77 

20.6 

11.6 

26.7 

35.9 

5.2 

(X-rayed 

3-16-46(5 min.) 

28.2 

6.38 

33.2 

9.1 

24.1 

28.5 

5.0 

on 

3-5-46) 

3-16-46 (1 hr.) 

28.2 

6.06 

32.2 

9.4 

24.1 

28.7 

5.6 

138 

4-9-46 

30.0 

5.80 

44.5 

6.2 

12.4 

28.3 

6.9 

350 r 

4-16-46 

26.8 

5.63 

45.3 

7.8 

12.6 

25.8 

8.5 

(X-rayed 

4-23-46 (Control) 

31.8 

6.68 

28.4 

14.8 

17.2 

32.2 

7.5 

on 

4-23-46 (2 min.) 

32.6 

7.23 

37.1 

11.9 

16.3 

27.0 

7.7 

4-9-46) 

4-23-46“ (l hr.) 

36.4 

6.70 

43.3 

13.3 

18.2 

20.0 

5.4 


4-23-46 (6 hr.) 

35.2 

6.53 

36.1 

13.2 

17.9 

27.9 

4.7 


4-24-46 

31.0 

7.13 

28.6 

11.1 

* 25.4 

29.3 

5.6 


4-24-46 (Terminal) 
(Terminal) Lipide 

37.6 

6.73 

29.1 

10.1 

27.3 

28.8 

4.7 


extracted (a) 
(Terminal) Lipide 

16.2 

5.19 

39.7 

6.4 

24.2 

24.1 

5.8 


extracted (b) 

32.6 

5.78 

29.7 

5.5 

31.1 

28.3 

5.2 


5-13-46 (Control) 

31 

5.90 

40.5 

12 

12.7 

28.5 

6.1 

162 

5 min. 

29.8 

6.20 

53.1 

10.3 

10 

21.3 

5.3 

(no 

1 hr. 

31.2 

5.82 

48.8 

11 

11.5 

23.2 

5.3 

X-ray) 

6 hr. 

29.0 

5.72 

51.8 

12.4 

11.2 

21.0 

3.5 


Dog 123—5.12 g. albumin in 25 cc. saline injected 2-6-46. 

Dog 138—5.22 g. albumin in 25 cc. saline injected 3-16-46. 

Dog 128—4.3 g. albumin in 25 cc. saline injected 4-23-46. 

Dog 162—6.27 g. albumin in 25 cc. saline injected 5-13-46. 

° Some clotting occurred in this sample, thereby lowering the fibrinogen fraction and 
raising the albunin out of proportion to the other components. 


442 J. A. MUNTZ, E. S. GUZMAN BARRON AND C. L. PROSSER 


the injection and analyzed electrophoretically. A similar experiment 
was performed on a normal dog. 

As shown in Table II, the immediate effect of the albumin injection 
was to elevate the dog’s albumin fraction and to decrease the concen¬ 
tration of the 0 complex-fibrinogen fraction. The total amount of 
albumin injected was almost completely recoverable when analyses 
were made within 5 min. after injection. This was true of the normal 
dog, as well as of X-rayed animals. The per cent of injected albumin 
found at various time intervals after injection is given in Table III. 
Calculation of extra plasma albumin is based on the assumption that 
the volume of saline injected with the bovine albumin remained in the 
circulating blood following the injection. 

TABLE III 

Recovery of Injected Bovine Plasma Albumin at Various Times after Injection 


Per cent of injected albumin found in the plasma 4 


treatment 

5 min. after inj. 

1 hr. after inj. 

0 hr. after inj. 

24 hr. after inj. 

123 
(350 r) 


81 

47 

36 

128 
(350 r) 

91 


58 

26 

138 
(400 r) 

98 

82 



162 

(no X-ray) 

94 

53 

64 



• Calculation 

4 Plasma volume (X-rayed dogs) =* 6.2% of body weight. 

Plasma volume (normal dog) * 5.7% of body weight 

Per cent of injected albumin = (after (befon ? biection) X 100. 

Where: A =* Albumin concentration, g. per 100 cc., from electrophoretic pattern. 

cc 

Vi * Plasma volume, 

V% » Plasma volume, 
a * g. of bovine albumin injected. 

4 Determined on a series of 6 normal dogs and 6 X-rayed dogs with Evans’ blue dye. 




STUDIES ON ACTION OP IONIZING RADIATIONS 


443 


In two of the three X-rayed dogs, the (3 complex-fibrinogen compo¬ 
nent decreased after the albumin injection. This occurred also in the 
normal dog. On the other hand, the concentration of the elevated a r 
globulin was unaffected by the albumin injection. The normal dog, as 
well as the X-rayed dogs, lost approximately 50% of the injected 
albumin within 6 hr. This was not lost through the kidney. Dog No. 
128 was catheterized during the 6 hr. period following the injection of 
albumin; the urine obtained in this period contained some protein, but 
a total of only 55 mg. was excreted. 

Extraction of Lipides from the Terminal Plasma 

Zeldis et al. (11) have reported that the elevation of total globulin 
observed in hypoproteinemic, hyperlipemic dogs is largely due to the 
iipemia these animals developed. However, their electrophoretic 
patterns show that the arglobulin is not reduced by lipide extraction. 
The marked reduction occurs in the a r + a 2 -globulin. 

It was of interest to see whether the greatly increased a r + a 4 - 
globulin component in the terminal plasma of X-ray injured dogs could 
be reduced by lipide extraction. A 10 cc. sample of plasma from dog 
No. 128 was extracted, using the procedure described by Zeldis et al., 
except that all the extractions were carried out at the temperature of an 
acetone-solid C0 2 mixture, approximately — 70°C. The dry powder 
obtained by this procedure was not completely soluble in N NaCl. A 
considerable amount of a gelatinous precipitate remained insoluble. 
From the electrophoretic pattern, Fig. 2. terminal plasma (lipide- 
extracted a), it can be seen that the insoluble material was derived 
chiefly from the 0 complex-fibrinogen fraction. The a r + arglobulin 
fraction was hardly affected by this treatment. 

As described in the experimental part of this paper, another 12 cc. 
sample of the same plasma was extracted, using another procedure. 
As shown in Fig. 2, terminal plasma (lipide-extracted b), there is no 
abnormal diminution of the 0 complex-fibrinogen fraction in this ex¬ 
tracted plasma. Table II shows that the ar + arglobulin was not 
diminished by lipide extraction, but that a r + ay globulin was de¬ 
creased to one-half its original concentration. 

These results indicate that the elevated a r + arglobulin concentra¬ 
tion in the terminal plasma of X-ray injured dogs is an increase in true 
protein and is not due to an elevation in the lipide content. 



444 J. A. MUNTZ, E. 8. GUZMAN BARRON AND C. L. PROSSER 

Discussion 

There was no characteristic alteration in the electrophoretic pattern 
of a dog’s plasma proteins immediately after treatment with a lethal 
total body dose of X-rays. After such a treatment dogs survived for 
12-18 days. They maintained their food intake for about one week 
after X-ray and no consistently abnormal change in the plasma elec¬ 
trophoretic pattern was demonstrable other than a slight decrease in 
albumin.