Skip to main content

Full text of "Cartilage Oligomeric Matrix Protein in Idiopathic Pulmonary Fibrosis."

See other formats

OPEN 3 ACCESS Freely available online 

■0-PLOS I o-^E 

Cartilage Oligomeric Matrix Protein in Idiopathic 
Pulmonary Fibrosis 

Louis J. Vuga^*^ Jadranka Milosevic^ ^ Kusum Pandit\ Ahmi Ben-Yehudah^, Yanxia Chu\ 
Thomas Richards\ Joshua Sciurba\ Michael Myerburg\ Yingze Zhang\ Anil V. Parwani^, 
Kevin F. Gibson\ Naftali Kaminski^ 

1 Dorothy P and Richard P Simmons Center for Interstitial Lung Diseases, Division of Pulmonary, Allergy and Critical Care Medicine, University of Pittsburgh, School of 
Medicine, Pittsburgh, Pennsylvania, United States of America, 2 Department of Pathology, University of Pittsburgh, School of Medicine, Pittsburgh, Pennsylvania, United 
States of America, 3 Pittsburgh Development Center, Magee-Women's Research Institute and Foundation, University of Pittsburgh, School of Medicine, Pittsburgh, 
Pennsylvania, United States of America, 4 Pulmonary, Critical Care and Sleep Medicine, Yale School of Medicine, New Haven, Connecticut, United States of America 


Idiopathic pulmonary fibrosis (IPF) is a progressive and life threatening disease with median survival of 2.5-3 years. The IPF 
lung is characterized by abnormal lung remodeling, epithelial cell hyperplasia, myofibroblast foci formation, and 
extracellular matrix deposition. Analysis of gene expression microarray data revealed that cartilage oligomeric matrix 
protein (COMP), a non-collagenous extracellular matrix protein is among the most significantly up-regulated genes (Fold 
change 13, p-value <0.05) in IPF lungs. This finding was confirmed at the mRNA level by nCounter® expression analysis in 
additional 1 15 IPF lungs and 154 control lungs as well as at the protein level by western blot analysis. Immunohistochemical 
analysis revealed that COMP was expressed in dense fibrotic regions of IPF lungs and co-localized with vimentin and around 
pSMAD3 expressing cells. Stimulation of normal human lung fibroblasts with TGF-pi induced an increase in COMP mRNA 
and protein expression. Silencing COMP in normal human lung fibroblasts significantly inhibited cell proliferation and 
negatively impacted the effects of TGF-pi on C0L1A1 and PAIl. COMP protein concentration measured by ELISA assay was 
significantly increased in serum of IPF patients compared to controls. Analysis of serum COMP concentrations in 23 patients 
who had prospective blood draws revealed that COMP levels increased in a time dependent fashion and correlated with 
declines in force vital capacity (FVC). Taken together, our results should encourage more research into the potential use of 
COMP as a biomarker for disease activity and TGF-pi activity in patients with IPF. Hence, studies that explore modalities that 
affect COMP expression, alleviate extracellular matrix rigidity and lung restriction in IPF and interfere with the amplification 
of TGF-pi signaling should be persuaded. 

Citation: Vuga U, Milosevic J, Pandit K, Ben-Yehudah A, Chu Y, et al. (201 3) Cartilage Oligomeric Matrix Protein in Idiopathic Pulmonary Fibrosis. PLoS ONE 8(1 2): 
e83120. doi:10.1371/journal.pone.0083120 

Editor: Min Wu, University of North Dakota, United States of America 

Received June 12, 2013; Accepted October 30, 2013; Published December 20, 2013 

Copyright: © 2013 Vuga et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits 
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. 

Funding: Support for this research was provided by the Grants funded by National Institutes of Health (NIH) through the National Heart, Lung, and Blood 
Institute (NHLBI): HL108869, HL073745, R01HL095397, RC2HL101715, U01HL108642, and P50HL0894932 as well as funded by the Dorothy P. and Richard P. 
Simmons Endowed Chair for Pulmonary Research, a generous donation from the Dorothy P. and Richard P. Simmons Family. The funders had no role in study 
design, data collection and analysis, decision to publish, or preparation of the manuscript. 

Competing Interests: The authors have declared that no competing interests exist. 

* E-mail: 

9 These authors contributed equally to this work. 


Idiopathic pulmonary fibrosis is a chronic and devastating 
disease without a known etiology [1]. To date, IPF remains 
incurable with a median survival of 2.5 to 3 years [2] and it has the 
worst prognosis among interstitial lung diseases [3] . The prevailing 
hypothesis of disease pathogenesis suggests the disease begins as an 
alveolar epithelial injury with aberrant alveolar re-epithelialization 
[4] . What is believed to follow is a cascade of events including local 
changes in epithelial cell phenotypes, fibroblast-myofibroblast 
transformation, macrophage activation, epithelial cell apoptosis, 
release of a variety of cytokines, chemokines, and growth factors, 
including transforming growth factor pi (TGF-Pl). TGF-Pl is 
probably the most studied among them, because of its wide known 
roles in extracellular matrix deposition, as well as extensive effects 
on fibroblast and epithehal cell phenotypes [5-7]. While the 
relative contribution of these events is unclear, the end result is 

extensive lung remodeling, uncontrolled extracellular matrix 
deposition and formation of myofibroblast foci. 

We and others have applied genome scale transcript profiling 
techniques of human IPF lungs to better understand the disease, 
identify novel targets for therapeutic interventions as well as new 
biomarkers [8-13]. These studies have led to generation of 
expression profiles, and usually focused on one or two target 
molecules [8,10,14-16], but they still contain a wealth of 
information and should be mined for more. Recentiy, re-analyzing 
the datasets, we discovered that the cartilage oligomeric matrix 
protein (COMP), a protein never studied in the context of IPF, is 
among the top up-regulated genes in IPF lungs in published 
datasets [17]. 

Cartilage oligomeric matrix protein (COMP) is an extracellular 
matrix protein that is mainly localized to tendon, cartilage, and 
pericartUage tissues [18]. COMP has four epidermal growth factor 



December 2013 | Volume 8 | Issue 12 | e83120 


binding domains, 8 TSP-3 repeats, and a thrombospondin C- 
terminal domain, which together are responsible for binding 
interactions with other proteins and extracellular matrix compo- 
nents such as TGF-Pl [19,20]. COMP interacts with multiple 
matrix components, including coUagens type I, II, and IX, 
proteoglycans, non-coUagenous matrix proteins such as fibronectin 
and matrUins [21-23]. Most importantly COMP functions as 
matrix assembling facilitator and plays a role in the stability of the 
coUagen network. COMP binds and brings five coUagen molecules 
close to each other and promotes collagen fibril formation [24]. 
However, COMP doesn't bind to the formed collagen fibrils; 
instead it works as a catalyst to arrange the coUagen molecules for 
early and abnormal fibril formation and thus may contribute to 
matrix rigidity. 

Increases in COMP have been reported in several diseases [25- 
27]. In rheumatoid arthritis and osteoarthritis injury to chondro- 
cytes leads to increased secretion of COMP [25] and interaction of 
COMP with rheumatoid arthritis synovial fibroblasts through 
integrins has been reported [28,29] COMP secretion from skin 
fibroblasts has been reported in affected skin of keloids [30] and 
systemic sclerosis patients [31-33]. Elevations of COMP have also 
been reported in vascular atherosclerosis [34], systemic lupus 
erythematosus (SLE) [35], renal fibrosis [36], degenerating acinar 
cells of chronic pancreatitis [37], and liver cirrhosis [38]. While 
increases in COMP have not been reported in lung fibrosis, we 
noticed that COMP was increased in some of our microarray 

Based on these observations, we decided to investigate the role 
of COMP in IPF. We analyzed the expression of COMP in a 
larger set of lungs, localized its protein over-expression in IPF 
lungs and determined its regulation and effects on normal human 
lung fibroblasts and determined the relationship between elevated 
COMP serum levels and measures of disease severity in IPF. 

Materials and Methods 

Gene Expression Microarray 

The IPF and control lung tissues were obtained from University 
of Pittsburgh Health Sciences Tissue Bank (Pittsburgh, PA). The 
experimental materials, procedures, samples collection, IRB, and 
statistical analysis have been previously described by us [17,39]. 
The gene expression microarray was performed on 15 controls 
and 23 UIP samples. The data is available at Gene Expression 
Omnibus (GSE- 10667). 

nCounter® Gene Expression Analysis 

We extracted total RNA from 154 Control lungs and 1 15 IPF 
lung tissues obtained from the Lung Tissue Research Consortium 
(LTRC) for nCounter® Analysis System (Nanostring, Seattle, WA) 
validation of COMP mRNA expression. Details about the samples 
are available at the LGRC (Lung Genomics Research consortium) 
website ( 500 ng of total RNA 
were hybridized to a 3' biotinylated capture probe and a 5' 
reporter probe tagged to a fluorescent barcode. Following 
overnight hybridization at 65°C, the samples were transferred to 
the nCounter® Prep Station, excess probes were washed out, and 
the probe-RNA complexes were bound and immobihzed on 
streptavidin-coated cartridges. The cartridges were scanned in the 
nCounter® Digital Analyzer using 1 155 fields of vision. The data 
was normalized using GUSB as the housekeeping gene. 

Cell Culture and Transfection 

Early passages (1-3) of primaiy normal human lung fibroblast 
(NHLF) (Lonza Ltd, Basel Switzerland) were cultured in a 

humidified atmosphere containing 5% of COj in incubator 
(Kendro Lab, New Town, CT) at 37°C and maintained as per die 
supplier's instructions. For hypoxic conditions, cells were placed in 
hypoxic conditions with 1 % of oxygen for 24 hours. AH cells were 
grown until 70-80% confluence. Whenever indicated, cells were 
stimulated with recombinant TGF-fil (R&D, Minneapohs, MN) 
and/ or traiisfected with 50 nM siCOMP and their corresponding 
negative controls (Thermo Scientific Dharmacon, Lafayette, CO) 
using Lipofectamine 2000 (Iiivitrogen, Carlsbad, CA) according to 
the manufacturer's instructions. 

RNA Extraction and Real-time RT-PCR Analysis 

Total RNA were extracted from NHLF. AH reagents (including 
primers for COMP) and analysis software used for qRT-PCR 
experiment were obtained from ABI, (Foster City, CA) and 
performed according to the vendor recommendation as previously 
described by us [16]. 

Protein Isolation and Western Blot Analysis 

Lung tissues and NHLF were lysed, harvested following the 
manufacturers' protocol (Thermo Fisher Scientific^'^, Rockford, 
IE). The concentrations of protein were measured by using 
Pierce's Biciiichoninic acid (BCA) (Pierce, Rockford, IE). For 
Western blot analysis, equal amounts of cellular extracts (10 |J.g) 
were separated on 10% SDS-PAGE gels and transferred to PVDF- 
Plus membranes (GE Osmonics, Trevose, PA). Western blots were 
performed with antibodies against COMP (1:1,000; Lifespan 
Bioscience, Inc., Seattle, WA), |3-actin (1:10,000; Sigma - Aldrich, 
St. Louis, MO), P-SMAD3 (1:1,000; CeU Signaling Technology, 
Inc. Beverly, MA). After incubation with the respective secondary 
antibodies, specific bands were visuahzed by autoradiography 
using enhanced chemHuminescence according to the manufactur- 
er's instructions (PerkinElmer Life Sciences, Boston, MA). 
Densitometry was performed using the shareware, ImageJ 


Paraffin embedded IPF and control lungs were obtained from 
University of Pittsburgh Health Sciences Tissue Bank (Pittsburgh, 
PA). Tissue slides were deparaflinized in serials: 100%, 90%, 80%, 

Table 1. Demographic and the clinical information of the 
patients in the longitudinal study. 



IPF (N = 23) 












Non smokers 




68.1 ±8.6 





Baseline PFTs 




48.3 ±16.5 


50.7 ±11.4 

FVC: Force Vital Capacity, DLCO: Diffusing Capacity of Lung for Carbon 

CPI: Composite Physiological Index. 



December 2013 | Volume 8 | Issue 12 | e83120 



c 15 


» 13 

S" 12 


a) 11 
I 10 








n 8 

2 6 

4 H 

0 J 



(U U 

< <] 3 
z o 

O 1 


■> — 













Figure 1. COMP gene and protein levels are increased in IPF lungs. (A) Microarrays analysis revealed an increased COMP gene expression in 
23 IPF lungs compared to 15 control lungs (A) and verified in 13 IPF lungs and 13 control lungs by using qRT-PCR (B). (C) COMP mRNA levels were 
determined by nCounter® in 11 5 IPF lungs and 1 54 control lungs. (D-E) Protein levels of COMP were determined in IPF lungs (n = 6) and control lungs 
(n = 6) by using western blot and quantified by using ImageJ (p<0.05). (3- Actin was used as loading control and the western blot shown is a 
representative of three repeated experiments. 
doi:1 0.1 371 /journal.pone.00831 20.g001 

70%, 60% and 50% ethanol and rehydrated three times in PBS 
each time for 10 minutes. Slides were incubated for 45 minutes 
with 5% donkey serum in Tris-buflFered saline (TBS) pH 7.4 
containing 3% bovine serum albumin (Sigma- Aldrich) and 
incubated for 4 hours with primary antibody. After five washes 
with 0.5% BSA in TBS for 5 minutes each time, slides were 
incubated in a biotinylated donkey anti-rat secondary antibody for 
30 minutes. After 2 washes with 0.5% BSA in TBS for 10 minutes 
each, the slides and arrays were then incubated with streptavidin- 
linked alkaline-phosphatase (Jackson Immuno Research, West 
Grove, PA). Shdes were washed again and incubated for 15 
minutes in Fast Red substrate (DakoCytomation, Carpinteria, CA) 
to detect the activity of alkaline-phosphatase. Briefly, tissue 
sections and arrays were washed for five minutes in water and 
counterstained using Mayer's Hematoxylin (DakoCytomation, 
Carpinteria, CA). The images were visualized with Olympus '^'^ 
microscope, PROVIS (Olympus America Inc., Melville, NY). 

Confocal Imaging 

Frozen IPF lung slides were fixed in 2% Paraformaldehyde 
(Sigma-Aldrich, St. Louis, MO) for 20 minutes and permeabUized 
using 0.1% Triton X in PBS for 15 minutes, followed by 
rehydration in PBS, washes with 0.5% BSA in PBS and blocking 
with 5% donkey serum (Sigma) in 3% BSA in PBS for 45 minutes. 
Slides were incubated with anti-COMP (Accurate Chemical and 
scientific corporation, Westbury, NY), Vimentin (Abeam, Cam- 
bridge, MA), or pSMAD3 (Lifespan Bioscience, Inc., Seattle, WA) 

antibodies in a blocking solution overnight at 4°C. Secondary 
antibodies, nucleus staining (DAPI) and confocal imagining were 
performed as previously described by us [16,40]. 

Longitudinal Study Population 

All patients were evaluated at the University of Pittsburgh 
Medical Center, Pittsburgh, PA and studies were approved by 
the Institutional Review Board (IRB) at the University of 
Pittsburgh. The diagnosis of IPF was established on the basis of 
American Thoracic Society (ATS) and European Respiratory 
Society (ERS) Criteria [41] and surgical lung biopsy when 
clinically indicated. Clinical data were available through the 
Simmons Center Database at the University of Pittsburgh. 
Smoking status was defined as previously described [42]. All 
patients signed informed consent to participate in the study. 
Subjects enrolled in the study were followed at intervals of 3 to 
4 months according to usual care practices at the Dorothy P 
and Richard P Simmons Center for Interstitial Lung Diseases. 
Physiologic data (Pulmonary Function Tests [PFT] and oxygen 
desaturation studies) and physician assessments were performed 
at all visits. Radiographic studies (X-rays or Computed 
Tomography Scans) were performed when clinically indicated 
and blood samples were collected and pulmonary function tests 
(PFT) were examined in intervals of 3 to 4 months. The 
demographic and the clinical information of the patients in the 
longitudinal study are shown in the Table 1. 



December 2013 | Volume 8 | Issue 12 | e83120 


^ .- -i' ' "a I m 


Figure 2. COMP is localized in fibrotic regions of idiopathic pulmonary fibrosis lungs. (A-B) Localization of COMP protein (red) In tissue 
obtained from control lung: a healthy lung parenchyma without COMP (A) and perlcartllage airway region with COMP accumulation (B). (C) COMP 
protein (red) Is located In fibrotic region of IPF Lungs. (D-G) Co-locallzatlon of COMP and vimentin In tissue obtained from IPF lung. The green 
fluorescence represents COMP and the red fluorescence shows fibroblasts marker; vimentin. Nuclei were counterstalned with 4, 6- dlamldlno-2- 
phenyllndole (blue). Yellow represents co-expression of COMP (green) and vimentin (red) In IPF lung as a yellow tinge. All figures shown are 
representatives of more than 3 experiments and magnification of 40x. 
dol:1 0.1 371 /journal.pone.00831 20.g002 

Enzyme-Linked Immunosorbent Assay (ELISA) 

The IPF patients were recruited in longitudinal study as 
described in "Study Population". Participants were followed up 
for 2.5 years while their blood was drawn and PFT were 
examined. We measured COMP level in serum of IPF patients 
and controls using COMP ELISA kit (AnnaMar Medical AB, 
Goeteborg, Sweden). The experimental procedures were followed 
as recommended by the vendor and data were analyzed by 
utilizing Delta soft 1 1 1 version 2.243 (Bio-Rad, Hercules, CA). 

Statistical Analysis of Data 

AH values were presented as mean ± SD. Group comparisons 
were made using an unpaired, two-tailed Student's t-test for 
normally distributed data. A level of p<0.05 was considered 
statistically significant. Longitudinal study of COMP and all the 
PFT for 23 patients within 120 days of any blood draw were used 
for correlation analysis of COMP level to FVC %. A data set with 
one record per PFT occasion, associating each PFT with COMP 
level from the blood draw nearest it, was generated for all PFT of 
23 patients within 120 days of any blood draw. A generalized 
estimating equation was used to fit to the data, based on the 
reference of Yan et al [43] with FVC % predicted as dependent 
and COMP as independent variable, and accounting for the 
subject effect by assuming an exchangeable within-subject 
correlation structure. 


COMP Gene and Protein Expression is Higher in IPF 

COMP was one of the most significantly increased genes in our 
previously published microarray data [17] and qRT-PCR 
confirmed its up-regulation in the same tissues (Figure lA-B). 

We also verified the array result in a separate, larger cohort 
consisting of 1 1 5 IPF lung samples and 154 normal histology 
controls, by using nCounter® expression analysis and demonstrat- 
ed that COMP mRNA was significantly increased in IPF lungs 
(8.8 Fold change, P- value = <0.05 compared to normal histology 
lungs) (Figure IC). To compare COMP protein levels in IPF lungs 
to those in control lungs, we performed western blot analysis. We 
found a significant increase of COMP protein levels in IPF lungs 
(Figure ID-E). In order to localize COMP in IPF lungs, we 
performed Immunohistochemistry (IHC) analysis. In normal 
histology lungs, COMP was mainly located in the cartilaginous 
areas of the large airways (Figure 2B). The IPF lung exhibited 
expression of COMP (red in color) in areas of dense fibrosis and 
myofibroblast foci (Figure 2C). To identify the types of cell that 
secrete COMP protein in the lungs, we performed immunofluo- 
rescence stains on frozen IPF and control lungs. COMP (green) 
and Vimentin (red) were co-localized in IPF lungs (Figure 2G) 
suggesting that COMP was secreted mainly by mesenchymal cells, 
most probably fibroblasts in IPF lungs. 

COMP is a TGF-pi and Hypoxia Inducible Molecule 

After co-localizing COMP in fibroblasts of IPF lungs, we 
wanted to determine whether TGF-|3l regulates COMP expres- 
sion. Stimulation of NHLF with TGF-|3l (5 ng/ml) induced 
increase in mRNA and protein levels of COMP in a time 
dependent manner as determined by qRT-PCR, western blot and 
ELISA (Figure 3A-C). To determine whether COMP expression 
in the fibrotic lungs could be associated with TGF-pi effects, we 
co-localized phosphorylated SMAD3 (pSMAD3) and COMP. We 
found that COMP proteins were localized adjacent to cells 
containing pSMAD3 in the nuclei (Figure 3G) suggesting that this 
indeed was the case. 



December 2013 | Volume 8 | Issue 12 | e83120 



'"a\ - i -'i'- COMP 



= 4 



o 2 


(5 ng/mL) 








(5 ng/mL) 

Figure 3. TGF-pi induces COMP gene and protein expression. (A) COMP mRNA was determined by qRT-PCR in NHLF at 0, 2, and 6 hours after 
stimulation with recombinant TGF-pl (5 ng/ml). (B) Western blots show an increase in COIVIP protein and pSMADS expression after 1 2-24 hours TGF- 
(51 treatment. For all western blot experiments p-actin was used as loading control and all figures shown are representatives of more than 3 
experiments. (C) Primary fibroblasts were stimulated with TGF-pi (5 ng/mL) for 24 hours and COIVIP protein was determined using ELISA assay. (D-G) 
COMP protein (green) is adjacent to the cells containing pSMADB in nuclei. The coexpression of nuclei (blue) and pSMADB (red) is observed in tissue 
obtained from IPF lung as purple color and it is surrounded with COMP (green) in the merged figure 3G. 
doi:1 0.1 371 /journal.pone.00831 20.g003 

To determine the dose response of COMP induction by 
TGF-pi, we stimulated NHLF with 2, 5, and 10 ng/mL of 
TGF-pi. We observed increased expression of COMP with 
2 ng/mL of TGF-pi, and an additional significant induction 
when we used a concentration of 5 ng/mL TGF-Pl. Licreasing 
the TGF-pi concentration to 10 ng/mL had no additional 
effect on COMP expression (Figure 4A). In contrast, the TGF- 
Pl concentration of 10 ng/mL further increased the expression 
of PAIl, a highly TGF-pi responsive gene (Figure 4B). 

We also examined whether COMP was inducible by hypoxia. 
Exposure of NHLF cells to hypoxic conditions (1% O2) for 24 
hours induced a significant increase COMP gene expression 
(Figure 4C) without evidence of an increase in TGF-pi as can 
be observed from lack of increase in PAH (Figure 4D). This 
observation suggests that COMP may also be induced by 
extreme hypoxia conditions, potentially independent of TGF-pi. 

COMP Modulated TGF-p Signaling 

In order to determine whether COMP plays a role in 
modulating TGF-P 1 signaling, we measured mRNA levels of 
profibrotic and TGF-P responsive genes PAIl and COLIAI. 
After silencing COMP, the expression of PAH and COLlAl 
was significantly reduced (Figure 5 A-B). As it has been 
reported that TGF-P 1 induced lung fibroblast proliferation [44], 
we examined NHLF proliferation rate after treating the cells 
with siCOMP and TGF-Pl. The results showed that silencing 
COMP expression reduced TGF-P 1 induced NHLF prolifera- 
tion rate (Figure 5C) suggesting that indeed COMP had a role 
in modulating TGF-P 1 signaling as previously described [19]- 



December 2013 | Volume 8 | Issue 12 | e83120 







70 T 
0 J 




























Figure 4. COMP induction by TGF pi is dose dependent and Hypoxia induces COMP. A) COMP mRNA was determined by qRT-PCR in NHLF 
treated with 2, 5, and 10 ng/ml TGF pi. B) PAI induction by 2, 5, and 10 ng/ml TGF |31. C) Exposure of cells to extreme hypoxia (1% O2) for 24 hours 
causes induction of COiVlP mRNA (as measured by qRT-PCR), but not accompanied by a similar increase in PAI1 (D). 
doi:1 0.1 371 /journal.pone.00831 20.g004 

Increased Concentration of COMP Protein in Serum of IPF 
Patients is Associated with Decline of FVC 

We next compared concentration of COMP protein in the 
Serum of IPF patients (n = 20) and controls (n = 20). We used 
COMP ELISA Assay and determined that COMP concentrations 
were significandy increased in the serum of IPF patients when 
compared with controls (P-value = 0.004, mean = 9.977 and 
SD = 4.422 for IPF and mean = 6.475 and SD = 2.473 for controls 
(Figure 6A). To determine whether COMP changed with disease 
progression, we used samples from 23 patients who were 
prospectively followed up for 2.5 years with periodic blood draws 
and pulmonary function tests. We found that COMP protein 
levels increased in time dependent fashion and showed a 
significant correlation with the decline of force vital capacity in 
the majority of patients (Figxire 6B). 


In this study we investigated the levels and potential role of 
COMP in IPF. COMP is among the most up-regulated genes in 
IPF by microarrays, a result validated by both qRT-PCR and 
nCounter® expression system. COMP protein is localized to 
vimentin expressing cells in IPF lungs and is usually found 
adjacent to cells that have nuclear p-SMAD3. Stimulation of 

NHLF with TGF-pi induces up-regulation of COMP gene 
expression and COMP protein levels. The silencing of COMP 
reduces PAIl and COLlAl gene expression and inhibits TGF-pi 
induced fibroblast proliferation. COMP concentrations are 
increased in serum of IPF patients compared to control and the 
serum concentration of COMP in IPF patients continue to 
increase over time and correlate with disease progression as 
reflected by decline in FVC. Our observations highlight the 
potential role of COMP in IPF and support its' proposed role of as 
a modulator of TGF- (31 signaling. 

Our finding that COMP protein level is increased in IPF sera 
and IPF lungs is relevant in understanding the pathogenesis of IPF 
lungs. COMP mutations have been described as the cause of 
pseudoachondroplasia (PSACH) and multiple epiphyseal dysplasia 
(MED) [45-47]. The concentration of COMP is decreased in both 
serum and plasma of PSACH or MED patients [48,49] which 
could be explained by accumulation of mutated COMP in 
granular or lamellar in endoplasmic reticulum (ER) of the 
chondrocytes [50,51]. In contrary, COMP levels are increased 
in patients with rheumatoid arthritis (RA) and the molecule has 
been proposed as a potential biomarker for RA and osteoarthritis 
(OA) activity [25]. Similarly COMP is increased in scleroderma 
dermal fibroblasts and in serum of patients with systemic sclerosis 
[31-33] and also in cirrhotic livers [38]. While it has not been 



December 2013 | Volume 8 | Issue 12 | e83120 





- ^ 

2 O 








C 2 




Figure 5. COMP modulates TGF-p signaling. NHLF were trans- 
fected with 70 nM of siCOIVlP or scrambled (SCR), treated for 6 hours 
with TGF-pi (5 ng/mL) and RNA extracted after 24 hours. (A-B) qRT-PCR 
was used to determine mRNA levels of PAI1 and C0L1A1 in NHLF. (C) 
The effect of COMP inhibition using siRNA on TGF-pi induced NHLF 

doi:1 0.1 371 /journal.pone.00831 20.g005 

studied directly in the lung, a recent longitudinal study in patients 
with systemic scleroderma with lung involvement suggested that 
elevated COMP concentrations in serum were predictive of 
mortaUty [52]. While we did not assess mortality, we did show an 
increase in COMP proteins in the lungs and bloods of patients 
with IPF and demonstrated that the increase was associated with 
decline of FVC over time, concurrent with previous results in 
other disease and suggesting that COMP should be added to 
repertoire of proteins evaluated as potential peripheral blood 
biomarkers in IPF. 

While we could not directly demonstrate COMP induction by 
TGF-P 1 in human IPF lungs, we demonstrated that COMP 
mRNA and protein levels were increased after stimulation of 
human lung fibroblasts with TGF-P 1. This is consistent with other 
reports that demonstrated induction of COMP after TGF-P 1 
stimulation in keloid and dermal fibroblast [30,33]. Consistent 
with that, we have shown that COMP expression in the IPF lung is 
generally distributed around cells that express vimentin and that 
have abundant phosphorylated nuclear SMAD3, indicative of 
TGF-P 1 stimulation. This is of particular interest because it had 
been described that COMP direcdy binds to members of TGF-P 1 
family and enhances their signal transduction activities [19]. When 
we examined the impacts of TGF-P 1 on NHLF after the silencing 
of COMP, we found reduction in TGF-P 1 target genes as well as 
significant reduction in TGF-P 1 induced fibroblast proliferation, 
suggesting that indeed COMP does enhance TGF-P 1 signaling. 
While it is difficult to establish experimentally a similar relation in 
the human IPF lung, our results that demonstrate co-locaUzation 
of COMP to areas where there is evidence of TGF-pi activities, 
add to these observations and suggest the possibility of a positive 
feedback loop between COMP and TGF-P 1 activities in the IPF 

The fibrotic lung is characterized by the intensive accumulation 
of extracellular matrix (ECM). The pathologic ECM depositions in 
fibrotic lungs include coUagens (I, III, V, VI and VII), fibronectin, 
elastin, and cartilage related proteins [53]. Interestingly, we found 
COMP was increased in dense fibrotic areas of lung parenchyma 
in IPF. We also localized COMP in normal lung but only around 
cartilage possessing airway (Figure 2A-B). In the case of COMP 
expression in chronic diseases such as pseudoachondroplasia 
(PSACH) and multiple epiphyseal dysplasia (MED), rheumatoid 
arthritis, systemic sclerosis and pathological would healing 
(Keloid), it was reported that COMP inhibits normal coUagen 
fibrils formation and destabilize normal extracellular matrix 
formation in areas where COMP molecules are excessively higher 
in the relationship to coUagen [30,54]. It was suggested that this 
inhibition cause increased lung rigidity by the lost elasticity typical 
of normal ECM. This may be of particular interest, because of the 
evidence that abnormal ECM rigidity plays a significant role in the 
pathogenesis of fibrosis [55]. In this context, it is of interest to note 
that other cartilage related proteins such as osteopontin [10,56], 
periostin [57], and YKL-40 [58] have been described as being 
significantiy increased in IPF. 

It is critically important to acknowledge few limitations of our 
study. First, we provide evidence on COMP regulation through 
TGF-Pl and its involvement in TGF-Pl signaling cascade in 
NHLF, but we do not provide similar data in-vivo. We feel that 
the co-localization of vimentin, pSMADS and COMP in fibrotic 
foci in human IPF lungs suggests that TGF-P I induces COMP 
secretion mostly in fibrotic regions of IPF lungs and to some extent 
more informative than using a limited model of lung fibrosis. To 
that extent we did observe that in-vitro exposing NHLF to extreme 
hypoxia in-vitro did also induce COMP, but it is unclear whether 
this mechanism would be relevant in-vivo. Clinically, our findings 



December 2013 | Volume 8 | Issue 12 | e83120 


A) B) 

Control IPF " » «s ' 2 2* 

Years After First Blood Draw 

Figure 6. COMP Protein levels in serum of IPF patients are associated with decline of FVC. (A) ELISA assay was used to determine 
concentrations of COMP in serum of IPF patients (n = 20) compared to controls (n = 20) (t-test, P-value <0.05). (B) Data from patients (n = 23) showed 
the ratio of change in COMP and the decline of FVC in a longitudinal study. Red line represents average value of ratio of change in COMP level vs. 
change in FVC. The blue line represents baseline of COMP level in each participant whereas each serves as its own control. 
doi:1 0.1 371 /journal.pone.00831 20.g006 

are also limited, whUe we offer compelling evidence of the 
relationship between COMP protein levels in serum and FVC, we 
are aware that the numbers are limited and we did not capture 
patients for this study in earlier stages of the disease. Our center is 
a tertiary medical facility, which gets regional and national 
referrals, and thus the populations of the study were at later stages 
of the disease. Despite this shortcoming, we were able to show the 
continuous increase of serum COMP level in 2.5 years. 
Considering its role in regulation of matrix rigidity, the 
observation that elevated COMP concentrations in serum of IPF 
patients are associated with a decline of FVC provides support to 
our hypothesis that COMP plays a role in pathogenesis of IPF and 
should be further evaluated as a biomarker for disease activity in 

In this manuscript we demonstrate that the mRNA and protein 
expression levels of COMP, an extracellular matrix protein that 
accentuates TGF-Pl signaling and is associated with extracellular 
matrix polymerization and stiffness, are high in IPF lungs 
compared to controls. COMP serum protein concentrations are 
increased in IPF patients and correlate with the decline of FVC 
over time in individuals with IPF. We also demonstrate data that 
COMP may be induced by TGF-Pl in the IPF lung and that at 
least in-vitro serves as an enhancer of TGFpi signaling as 


1. Raghu G, Collard HR, EganJJ, Martinez FJ, Bchr J, ct al. (2011) All official 
ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis: evidence- 
based guidelines for diagnosis and management- Am J Respir Grit Gare Med 
183: 788-824. 

2. Flalierty KR, Travis WD, Colby TV, Toews GB, Kazerooni EA, et al. (2001) 
Histopadiologic variability in usual and nonspecific interstitial pneumonias. 
Am J Respir Grit Gare Med 164: 1722-1727. 

3. Lynch JP 3rd, Saggar R, Wcigt SS, Zisman DA, White ES (2006) Usual 
interstitial pneumonia. Semin Respir Grit Gare Med 27: 634—651. 

4. Sclman M, King TE, Pardo A (2001) Idiopathic pulmonary fibrosis: prevailing 
and evolving hypotheses about its pathogenesis and implications for therapy. 
Ann Intern Med 134: 136-151. 

5. Pottier N, Maurin T, Chevalier B, Puissegur MP, Lebrigand K, et al. (2009) 
Identification of keratinocyte growth factor as a target of mieroRNA-155 in lung 
fibroblasts: implication in cpitliclial-mesenchymal interactions. PLoS One 4: 

6. WUIis BC, Borok Z (2007) TGF-beta-indueed EMT: mechanisms and 
implications for fibrotie lung disease. Am J Physiol Lung Cell Mol Physiol 
293: L525-534. 

7. Pandit KV, Corcoran D, Yousef H, Yarlagadda M, Tzouvelckis A, ct al. (2010) 
Inhibition and role of let-7d in idiopathic pulmonary fibrosis. Am J Respir Grit 
Gare Med 182: 220-229. 

previously proposed. Taken together, our results should encourage 
more research into the potential use of COMP as a biomarker for 
disease activity and TGF-(3l activity in patients with IPF. Hence, 
studies that explore modalities that affect COMP expression, 
alleviate extracellular matrix rigidity and lung restriction in IPF 
and interfere with the amplification of TGF-pi signaling should be 


The authors thank members of Dr. Kamiiiski's Laboratory and IPF 
support group (The Dorothy P and Richard P Simmons Center for 
Interstitial Lung Diseases), as well as Dr. Simon Watkins, Director of 
Center of Biological Imaging, University of Pittsburgh that made these 
studies possible. Portion of these findings have been presented at American 
Thoracic Society (ATS) conference in San Diego, May 2009 and San 
Francisco, May 2012. 

Author Contributions 

Conceived and designed the experiments: LfV JM KP ABY JS NK. 
Performed the experiments: LJVJM KP MM YC JS. Analyzed the data: 
LJV JM KP TR NK ABY JS MM YC AVP. Contributed reagents/ 
materials/analysis tools: LJV YZ MM TR AVP NK. Wrote the paper: LJV 
JM NK. Patients recruitment: KFG YZ. 

8. Zuo E, Kaminski N, Eugiii E, Allard J, Yakhiiii Z, ct al. (2002) Gene expression 
analysis reveals matrilysin as a key regulator of pulmonary fibrosis in mice and 
humans. Proc Nad Acad Sci U S A 99: 6292-6297. 

9. Gosgrove GP, Brown KK, Sehicmann WP, Scrls AE, Parr JE, et al. (2004) 
Pigment epithelium-derived factor in idiopathic pulmonary fibrosis: a role in 
aberrant angiogenesis. Am J Respir Grit Gare Med 170: 242—251. 

10. Pardo A, Gibson K, Gisneros J, Richards TJ, Yang Y, et al. (2005) Up-regulation 
and profibrotic role of osteopontin in human idiopathic pulmonary fibrosis. 
PLoS Med 2: e251. 

1 1 . Sclman M, Pardo A, Barrcra L, Estrada A, Watson SR, et al. (2006) Gene 
expression profiles distinguish idiopathic pulmonary fibrosis from hypersensitiv- 
ity pneumonitis. AmJ Respir Grit Gare Med 173: 188-198. 

12. Bridges RS, Kass D, Loh K, Glackin G, Borczuk AC, et al. (2009) Gene 
expression profiling of pulmonary fibrosis identifies Twist 1 as an antiapoptotic 
molecular "rectifier" of growth factor signaling. AmJ Pathol 175: 2351—2361. 

13. Hsu E, Shi H, Jordan RM, Lyons-Wciler J, Pilewski JM, ct al. (2011) Lung 
tissues in patients with systemic sclerosis have gene expression patterns unique to 
pulmonary fibrosis and pulmonary hypertension. Artliritis Rheum 63: 783—794. 

14. Wang XM, Zhang Y, Kim HP, Zhou Z, Feghali-Bostwick GA, et al. (2006) 
CavcoIin-1: a critical regulator of lung fibrosis in idiopathic pulmonary fibrosis. 
J Exp Med 203: 289.5-2906. 



December 2013 | Volume 8 | Issue 12 | e83120 


15. Englcrt JM, Hanford LE, Kaminski N, Tobolcwski JM, l\in RJ, rt al. (2008) A 
role for the receptor for advanced glycation end products in idiopathic 
pulmonary fibrosis. Am J Pathol 172: 583-591. 

16. Vuga LJ, Bcn-Yehudah A, Kovkarova-Naumovski E, Oriss T, Gibson KF, et £il. 
(2009) WNT5A is a regulator of fibroblast proliferation and resistance to 
apoptosis. AmJ Respir Cell Mol Biol 41: 583-589. 

17. Konishi K, Gibson KF, LindcU KO, Richards TJ, Zhang Y, et al. (2009) Gene 
expression profiles of acute exacerbations of idiopathic pulmonary fibrosis. 
AmJ Respir Crit Care Med 180: 167-175. 

18. DiCesare PE, Morgelin M, Carlson CS, Pasumarti S, Paulsson M (1995) 
Cartilage oligomcric matrix protein: isolation and characterization from human 
articular cartilage. J Orthop Res 13: 422-4-28. 

19. Haudenschild DR, Hong E, Yik JH, Chromy B, Morgelin M, et al. (2011) 
Enhanced activity of transforming growth factor betal (TGF-betal) bound to 
cartilage oligomeric matrix protein. J Biol Chem 286: 43250—43258. 

20. Newton G, Weremowiez S, Morton CC, Copeland NG, Gilbert DJ, et al. (1994) 
Characterization of human and mouse cartilage oligomeric matrix protein. 
Genomics 24: 435-439. 

21. Thur J, Rosenberg K, Nitschc DP, Pihlajamaa 1\ Ala-Kokko L, ct al. (2001) 
Mutations in cartilage oligomcric matrix protein causing pseudoachondroplasia 
and multiple epiphyseal dysplasia affect binding of calcium and collagen I, II, 
and DC. J Biol Chem 276: 6083-6092. 

22. Mann HH, Ozbek S, Engel J, Paulsson M, Wagener R (2004) Interactions 
between the cartilage oligomeric matrix protein and matrilins. Implications for 
matrix assembly and the pathogenesis of chondrodysplasias. J Biol Chem 279: 

23. Holden P, Meadows RS, Chapman KL, Grant ME, Kadlcr KE, et al. (2001) 
Cartilage oligomeric matrix protein interacts with type IX collagen, and 
disruptions to these interactions identify a pathogenetic mechanism in a bone 
dysplasia family. J Biol Chem 276: 6046-6055. 

24. Halasz K, Kassner A, Morgelin M, Heinegard D (2007) COMP acts as a catalyst 
in collagen fibriliogenesis. J Biol Chem 282: 31166-31173. 

25. Chaganti RK, Kehnan A, Lui L, Yao W, Javaid MK, et al. (2008) Change in 
serum measurements of cartilage oligomeric matrix protein and association with 
the development and worsening of radiographic hip osteoarthritis. Osteoarthritis 
Cartilage 16: 566-571. 

26. Smidi RK, Zunino L, Webbon PM, Heinegard D (1997) The distribution of 
cartilage oligomeric matrix protein (COMP) in tendon and its variation with 
tendon site, age and load. Matrix Biol 16: 255-271. 

27. Hesselstrand R, Kassner A, Heinegard D, Saxne T (2008) COMP: a candidate 
molecule in the pathogenesis of systemic sclerosis with a potential as a disease 
marker. Ann Rheum Dis 67: 1242-1248. 

28. Neidhart M, Zaucke F, von Knoch R, Jungel A, Michel BA, et al. (2005) 
(ialcctin-3 is induced in rheumatoid arthritis synovial fibroblasts after adhesion 
to cartilage oligomeric matrix protein. Ann Rheum Dis 64: 419—424. 

29. Chen FH, Thomas AO, Hecht JT, Goldring MB, Lawler J (2005) Cartilage 
oligomeric matrix protein/ thrombospondin 5 supports chondrocyte attachment 
through interaction with intcgrins. J Biol Chem 280: 32655-32661. 

30. Inui S, Shono F, Nakajima T, Hosokawa K, Itami S (2011) Identification and 
characterization of cartilage oligomeric matrix protein as a novel pathogenic 
factor in keloids. AmJ Pathol 179: 1951-1960. 

31. Yamamoto M, Takahashi H, Suzuki C, Naishiro Y, Yamamoto H, et al. (2007) 
Cartilage oligomeric matrix protein in systemic sclerosis. Rheumatology 
(Oxlbrd) 46: 1858-1859. 

32. Farina G, Lemaire R, Korn JH, Widom RL (2006) Cartilage oligomeric matrix 
protein is overexpressed by scleroderma dermal fibroblasts. Matrix Biol 25: 213- 

33. Farina G, Lemaire R, Pancari P, Bayle J, Widom RL, et al. (2009) Cartilage 
oligomeric matrix protein expression in systemic sclerosis reveals heterogeneity 
of dermal fibroblast responses to transforming growth factor beta. Ann Rheum 

Dis 68: 435-441. 

34. Canfield AE, Farrington C, Dziobon MD, Boot-Handford RP, Heagerty AM, et 
al. (2002) The involvement of matrix glycoproteins in vascular calcification and 
fibrosis: an immunohistochemical study. J Pathol 196: 228—234. 

35. Wislowska M, Jablonska B (2005) Cartilage oligomeric matrix protein in serum 
in systemic lupus erythematosus and knee osteoarthritis. Preliminary commu- 
nication. Rheumatol Int 25: 373-378. 

36. Kim W, Moon SO, Lee SY, Jang KY, Cho CH, et al. (2006) COMP- 
angiopoietin- 1 ameliorates renal fibrosis in a unilateral ureteral obstruction 
model. J Am Soc Nephrol 17: 2474-2483. 

37. Liao Q, Kleclf J, Xiao Y, Di Cesare PE, Korc M, et al. (2003) COMP is 
selectively up-regulated in degenerating acinar cells in chronic pancreatitis and 
in chronic-pancreatitis-like lesions in pemcreatic cancer. ScandJ Gastroenterol 
38: 207-215. 

38. Xiao Y, Kleeflf J, Guo J, Gazdhar A, Liao Q, et al. (2004) Cartilage oligomeric 
matrix protein expression in hepatocellular carcinoma and the cirrhotic Hver. 
J Gastroenterol Hepatol 19: 296-302. 

39. Milosevic J, Pandit K, Magistcr M, Rabinovieh E, Ellwanger DC, et al. (2012) 
Profibrotic role of miR-154 in pulmonary fibrosis. AmJ Respir Cell Mol Biol 47: 

40. Gilani SR, Vuga LJ, Lindell KO, Gibson KF, Xue J, et al. (2010) CD28 down- 
regulation on circulating CD4 T-ceUs is associated with poor prognoses of 
patients with idiopathic pulmonary fibrosis. PLoS One 5: e8959. 

41. Demedts M, Costabel U (2002) ATS/ERS international multidisciplinary 
consensus classification of the idiopathic interstitial pneumonias. Eur Respir J 19: 

42. King TE, Jr., Tooze JA, Schwarz Ml, Brown KR, Cherniack RM (2001) 
Predicting survival in idiopathic pulmonary fibrosis: scoring system and survi\'al 
model. AmJ Respir Crit Care Med 164: 1171-1181. 

43. Yan J, Fine J (2004) Estimating equations for association structures. Stat Med 23: 
859-874; discussion 875-857,879-880. 

44. Khalil N, Xu YD, O'Connor R, Duronio V (2005) Proliferation of puhnonary 
interstitial fibroblasts is mediated by transforming growth factor-beta 1 -induced 
release of extracellular fibroblast growth factor-2 and phosphorylation of p38 
MAPK andJNK. J Biol Chem 280: 43000-43009. 

45. Ikegawa S, Ohashi H, Nishimura G, Kim K(.^, Sannohe A, et al. (1998) Novel 
and recurrent COMP (cartUage oligomeric matrix protein) mutations in 
pseudoachondroplasia and multiple epiphyseal dysplasia. Hum Genet 103: 

46. Deere M, Sanford T, Ferguson HL, Daniels K, Hecht JT (1998) Identification of 
twelve mutations in cartilage oligomeric matrix protein (COMP) in patients with 
pseudoachondroplasia. AmJ Med Genet 80: 510-513. 

47. Briggs MD, Hoffman SM, King LM, Olsen AS, Mohrenweiser H, et al. (1995) 
Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the 
cartilage oligomeric matrix protein gene. Nat (jcnet 10: 330—336. 

48. Chen HC, Shah SH, Li YJ, Stabler TV, Jordan JM, et al. (2008) Inverse 
association of general joint hypermobility with hand and knee osteoarthritis and 
serum cartilage ohgomeric matrix protein levels. Arthritis Rheum 58: 3854— 

49. Tufan AC, Satiroglu-Tufan NL, Jackson GC, Semerci CN, Solak S, et al. (2007) 
Serum or plasma cartilage oligomeric matrix protein concentration as a 
diagnostic marker in pseudoachondroplasia: differential diagnosis of a family. 
Eur J Hum Genet 15: 1023-1028. 

50. Stanescu R, Stanescu V, Muriel MP, Maroteaux P (1993) Multiple epiphyseal 
dysplasia, Fairbank type: morphologic and biochemical study of cartilage. 
AmJ Med Genet 45: 501-507. 

51. Hansen U, Platz N, Becker A, Bruckner P, Paulsson M, et al. (201 1) A secreted 
variant of cartilage oligomeric matrix protein cau-rying a chondrodysplasia- 
causing mutation (p.H587R) disrupts collagen fibriliogenesis. Arthritis Rheum 
63: 159-167. 

52. Hesselstrand R, Andreasson K, Wuttge DM, Bozovic G, Scheja A, et al. (2012) 
Increased serum COMP predicts mortalitv in SSc: results from a longitudinal 
study of interstitial lung disease. Rheumatolo^g\' (Oxford) 51: 915-920. 

53. Pardo A, Selman M, Kaminski N (2008) Approaching the degradome in 
idiopathic pulmonary fibrosis. Int J Biochem Cell Biol 40: 1141-1155. 

54. Agarwal P, Schulz JN, Blumbach K, Andreasson K, Heinegard D, et al. (2013) 
Enhamced deposition of CEU-tilage oligomeric matrix protein is a common feature 
in fibrotic skin pathologies. Matrix Biol 32: 325-331. 

55. Zhou Y, Huang X, Hecker L, Kurundkar D, Kurundkar A, et al. (2013) 
Inhibition of mechanosensitivc signaling in myofibroblasts ameliorates experi- 
mental pulmonary fibrosis. J Clin Invest 123: 1096-1108. 

56. KadotaJ, Mizunoe S, Mito K, Mukae H, Yoshioka S, et al. (2005) High plasma 
concentrations of osteopontin in patients with interstitial pneumonia. Respir 
Med 99: 111-117. 

57. Okamoto M, Hoshino T, Kitasato Y, Sakazaki Y, Kawayama T, et al. (2011) 
Periostin, a matrix protein, is a novel biomarker for idiopathic interstitial 
pneumonias. Eur Respir J 37: 1119-1127. 

58. Korthagen NM, van Moorsel CH, Barlo NP, Ruven HJ, Kruit A, et al. (2011) 
Serum and BALF YKL-40 levels arc predictors of survival in idiopathic 
pulmonary fibrosis. Respir Med 105: 106-113. 



December 2013 | Volume 8 | Issue 12 | e83120