Skip to main content

Full text of "Century-scale methylome stability in a recently diverged Arabidopsis thaliana lineage"

See other formats


Downloaded from http://biorxiv.org/on September 18, 2014 



bioRviv 

f Vbeta 

THE PREPRINT SERVER FOR BIOLOGY 



Century-scale methylome stability in a recently diverged 
Arabidopsis thaliana lineage 

Joerg Hagmann, Claude Becker, Jonas Muller, et al. 
bioRxiv first posted online September 16, 2014 

Access the most recent version at doi: http://dx.doi.Org/10.1 101/009225 



Creative The copyright holder for this preprint is the author/funder. It is made available 
Commons under a CC-BY-NC-ND 4.0 International license. 
License 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Century-scale methylome stability in a recently diverged 
Arabidopsis thaliana lineage 



Jorg Hagmann'*, Claude Becker'*, Jonas Muller', Oliver Stegle 23 , Rhonda C. Meyer 4 , Korbinian 
Schneeberger 5 , Joffrey Fitz', Thomas Altmann 4 , Joy Bergelson 6 , Karsten Borgwardt 27 , Detlef Weigel' 



'Department of Molecular Biology, Max Planck Institute for Developmental Biology, 72076 Tubingen, 
Germany 

2 Machine Learning and Computational Biology Research Group, Max Planck Institute for Developmental 
Biology and Max Planck Institute for Intelligent Systems, 72076 Tubingen, Germany 

3 European Molecular Biology Laboratory, European Bioinformatics Institute, Wellcome Trust Genome 
Campus, Hinxton, Cambridge CBIO ISD, United Kingdom 

4 The Leibniz Institute of Plant Genetics and Crop Plant Research, 06466 Gatersleben, Germany 

department of Plant Developmental Biology, Max Planck Institute for Plant Breeding Research, 50829 
Cologne, Germany 

department of Ecology and Evolution, University of Chicago, Chicago, IL 60637, USA 

7 Center for Bioinformatics (ZBIT), Eberhard Karls Universitat Tubingen, 72076 Tubingen, Germany 



I 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



ABSTRACT 

There has been much excitement about the possibility that exposure to specific environments can induce 
an ecological memory in the form of whole-sale, genome-wide epigenetic changes that are maintained over 
many generations. In the model plant Arabidopsis thal'iana, numerous heritable DNA methylation differences 
have been identified in greenhouse-grown isogenic lines, but it remains unknown how natural, highly 
variable environments affect the rate and spectrum of such changes. Here we present detailed methylome 
analyses in a geographically dispersed A. thaliana population that constitutes a collection of near-isogenic 
lines, diverged for at least a century from a common ancestor. We observed little DNA methylation 
divergence whole-genome wide. Nonetheless, methylome variation largely reflected genetic distance, and 
was in many aspects similar to that of lines raised in uniform conditions. Thus, even when plants are grown 
in varying and diverse natural sites, genome-wide epigenetic variation accumulates in a clock-like manner, 
and epigenetic divergence thus parallels the pattern of genome-wide DNA sequence divergence. 

INTRODUCTION 

Differences in DNA methylation between individuals can be due to genetic variation, stochastic events or 
environmental factors. Epigenetic marks such as DNA methylation are not randomly distributed across 
plant genomes, but associate with certain classes of genomic loci, especially with transposable elements 
(TEs). Changes in the DNA sequence or structure caused by, for instance, TE insertion, can induce 
secondary epigenetic effects at the concerned locus [1, 2], or, via RdDM, even at distant loci [3-5]. The high 
degree of sequence variation, including insertions/deletions (indels), copy number variants (CNVs) and 
rearrangements among natural accessions in A thaliana provides ample opportunities for linked epigenetic 
variation [6-I0]. The genomes of A thaliana accessions from around the globe are rife with differentially 
methylated regions (DMRs) [10], but it remains unclear how many of these cannot be explained by closely 
linked genetic mutations and how many are pure epimutations [II] that occur in the absence of any genetic 
differences. 

The seemingly spontaneous occurrence of heritable DNA methylation differences has been documented 
for wild-type Arabidopsis thaliana isogenic lines grown for several years in a stable greenhouse environment 
[1 2, 1 3]. Truly spontaneous switches in methylation state are most likely the consequence of incorrect 
replication or erroneous establishment of the methylation pattern during DNA replication [1 4- 1 6]. A 
potential amplifier of stochastic noise is the complex and diverse population of small RNAs that are at the 
core of RNA-directed DNA methylation (RdDM) [1 7] and that serve as epigenetic memory between 
generations. The exact composition of small RNAs at silenced loci can vary considerably between 
individuals [1 3], and stochastic inter-individual variation has been invoked to explain differences in 
remethylation, either after development-dependent or induced demethylation of the genome [1 8, 1 9]. Such 
epigenetic variants can contribute to phenotypic variation within species, and epigenetic variation in 



2 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



otherwise isogenic individuals has been shown to affect ecologically relevant phenotypes in A. thaliana [20- 
22]. 

In addition to these spontaneous epigenetic changes the environment can induce demethylation or de novo 
methylation in plants, for example after pathogen attack [23]. Recently, it has been proposed that repeated 
exposure to specific environmental conditions can lead to epigenetic differences that can also be 
transmitted across generations, constituting a form of ecological memory [24-27]. The responsiveness of 
the epigenome to external stimuli and its putative memory effect have moved it also into the focus of 
attention for epidemiological and chronic disease studies in animals [28,29]. How the rate of trans- 
generational reversion among induced epivariants with phenotypic effects compares to the strength of 
natural selection, which in turn determines whether natural selection can affect the population frequency of 
epivariants, is largely unknown [30-33]. 

To assess whether a variable and fluctuating environment is likely to have long-lasting effects in the absence 
of large-scale genetic variation, we have analyzed a lineage of recently diverged A. thaliana accessions 
collected across North America. Using a new technique for the identification of differential methylation, we 
found that in a population of thirteen accessions originating from eight different locations and diverged for 
more than one hundred generations, only 3% of the methylome had undergone a change in methylation 
state. Epimutations at the DNA methylation level did not accumulate at higher rates in the wild as they did 
in a benign greenhouse environment. Using genetic mutations as a timer, we demonstrate that 
accumulation of methylation differences was non-linear, corroborating our previous hypothesis that shifts in 
methylation states are generally only partially stable and that reversions to the initial state are frequent 
[12,34]. Many methylation variants that segregated in the natural North American lineage could also be 
detected in the greenhouse-grown population, indicating that similar forces determined spontaneous 
methylation variation, independently of environment and genetic background. Population structure could be 
inferred from differences in methylation states, and the pairwise degree of methylation polymorphism was 
linked to the degree of genetic distance. Together, these results suggest that the environment makes only a 
small contribution to trans-generationally inherited epigenetic variation on a genome-wide scale. 

RESULTS 

Characterization of the near-isogenic HPGI lineage from North America 

Previous studies of isogenic mutation accumulation (MA) lines raised in uniform greenhouse conditions 
identified many apparently spontaneously occurring pure epimutations [12,13]. To determine whether 
variable and fluctuating environments in the absence of large-scale genetic variation substantially alter the 
genome-wide DNA methylation landscape in the long term, we analyzed a lineage of recently diverged A 
thaliana accessions collected across North America. Different from the native range of the species in 
Eurasia, about half of all North American individuals appear to be identical when genotyped at 139 genome- 

3 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



wide markers [35]. We selected 1 3 individuals of this lineage, called haplogroup-l (HPGI), from locations 
in Michigan, Illinois and on Long Island, including pairs from four sites (Figure la, Table SI). Whole-genome 
sequencing of pools of eight to ten siblings from each accession identified a shared set of 670,979 single 
nucleotide polymorphisms (SNPs) and 170,998 structural variants (SVs) relative to the Col-0 reference 
genome, which were then used to build a HPG I pseudo reference genome (SOM: Genome analysis of 
HPGI individuals; Table S2; Figure SI). 

Only 1,354 SNPs and 521 SVs segregated in this population (Table S3, Figure S2 and S3), confirming that 
the 1 3 strains were indeed closely related. Segregating SNPs were noticeably more strongly biased towards 
GC^AT transitions than shared SNPs, especially in TEs, although the bias was not as extreme as in the 
greenhouse-grown MA lines (Figure lb) [36]. A phylogenetic network and STRUCTURE analysis based on 
the segregating polymorphisms reflected the geographic origin of the accessions (Figure la, c; Figure S4). 
Three of the pairs of accessions from the same site were closely related, and were responsible for many 
alleles with a frequency of 2 in the sampled population (Figure Id). If the spontaneous genetic mutation rate 
is similar to that seen in the greenhouse [36], the HPG I accessions would be 1 5 to 384 generations 
separated from each other. With a generation time of one year, their most recent common ancestor 
would have lived about two centuries ago, which is consistent with A. thaliana having been introduced to 
North America during colonization by European settlers [37]. Lastly, we observed only a weak positive 
correlation between genetic distance and phenotypic difference (Figure S5). We conclude that the HPG I 
accessions constitute a near-isogenic population that should be ideal for the study of heritable epigenetic 
variants that arise in the absence of large-scale genetic change under natural growth conditions. 

Differentially methylated positions the HPGI lineage 

To assess the long-term heritable fraction of DNA methylation polymorphisms in the HPGI lineage, we 
grew plants under controlled conditions for two generations after collection at the natural sites, before 
performing whole methylome bisulfite sequencing on two pools of 8- 1 0 individuals per accession (SOM: 
Primary analysis of methylation; Table S4). After mapping reads to the HPG I pseudo reference genome, we 
first investigated epigenetic variation at the single-cytosine level. There were 535,483 unique differentially 
methylated positions (DMPs), with an average of 147,975 DMPs between any pair of accessions 
(SD = 23,745); thus, 86% of methylated cytosines accessible to our analyses were stably methylated across 
all HPG I accessions. The vast majority of DMPs (97%) was detected in the CG context (CG-DMPs). As we 
have discussed previously [12], this can be largely attributed to the differences in methylation rates. 
Because of lower average CHG and CHH methylation compared to CG methylation at individual sites, 
statistical tests of differential methylation fail more often for CHG and CHH sites. The finding that only 
about 2% of all covered cytosines were differentially methylated strongly contrasts with a previous 
population epigenomic study [10], which despite lower sequencing depth than our experiments concluded 
that the vast majority, over 90%, of all cytosines in the A. thaliana genome is differentially methylated among 



4 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



1 40 natural, more divergent accessions [10], with a third being found at minor allele frequencies of over 
10%. 

Using the geographic outlier LISET-036 as a reference strain, we found that 61% of CG-DMPs as well as 
36% of the small number of CHG- and CHH-DMPs were present in at least two independent accessions 
(Figure S6a), many of them shared between accessions from the same site. As is typical for A thaliana [38], 
most methylated positions clustered around the centromere and localized to TEs and intergenic regions 
(Figure 2a; Figure S6b). In contrast, CG-DMPs were over-represented on chromosome arms, localizing 
predominantly to coding sequences (Figure 2a; Figure S6b), similar to what we had previously observed in 
the greenhouse-grown MA lines [1 2]. 

We asked whether DMPs had accumulated more quickly in natural environments than in the greenhouse, 
using DNA mutations in the HPG I and MA populations as a molecular clock (SOM: Estimating DMP 
accumulation rates). Our null hypothesis was that a variable and highly fluctuating natural environment 
increases the rate of heritable methylation changes. In contrast, DMPs accumulated in sub-linear fashion in 
both the HPG I and MA populations [1 2] (Figure 2b) - with similar trends for DMPs in all three contexts - 
and DMPs did not accumulate more rapidly in the HPG I than in the MA lines. The steeper initial increase 
relative to SNP differences as well as the broader distribution of MA line differences relative to HPG I 
differences were most likely the result of having compared individual plants in the MA experiment [1 2], 
rather than pools of siblings, as in the HPG I experiment (Figure S7; SOM: Estimating DMP accumulation 
rates). Furthermore, if the genetic mutation rate in the wild were higher than in the greenhouse, for 
example because of increased UV exposure, we would underestimate the epimutation rate per generation 
in the HPG I strains. 

Differentially methylated regions the HPG I lineage 

Because it is unclear what consequences variation at individual methylated cytosines has in plants, we next 
investigated differentially methylated regions (DMRs) in the HPG I population. A limitation of previous plant 
methylome studies using short read sequencing has been that these relied on DMPs or fixed sliding 
windows along each chromosome to identify DMRs, rather than beginning with what appears intuitively to 
be more appropriate, namely regions that are known to be methylated in individual strains (methylated 
regions, MRs; SOM: Methylated regions) [39]. We therefore adapted a Hidden Markov Model (HMM), 
which had been developed for segmentation of animal methylation data [40], to the more complex DNA 
methylation patterns in plants (SOM: Differentially methylated regions). We identified on average 32,529 
MRs per strain (median length 1 22 bp), with almost a quarter of the HPG I reference genome, 22.6 Mb, 
covered by an MR in at least one strain (Figure 2a, c; Table S5, Figure S8a). MRs overlapping with coding 
regions were over-represented in genes responsible for basic cellular processes (p-value « 0.00 1), in 
agreement with gene body methylation being a hallmark of constitutively expressed genes [4 1]. Only l% of 
m CHH and 2% of m CHG positions were outside MRs (Figure 2d), consistent with the dense CHH and CHG 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



methylation found in repeats and silenced TEs [38]. Compared to m CGs within MRs, m CGs outside MRs 
localized almost exclusively to genes (94%), were spaced much farther apart, and were separated by many 
more unmethylated loci (Figure 2e; Figure S8b, c). This explains why sparsely methylated genes were 
under-represented in HMM-determined MRs, even though gene body methylation accounts for a large 
fraction of m CGs. The accuracy of our MR detection method was well supported by independent methods 
(SOM: Differentially methylated regions). 

Using the unified set of MRs, we tested all pairs of accessions for differential methylation, identifying 4,821 
DMRs with an average length of 159 bp. Of the total genome space occupied by MRs, only 3% were 
contained in DMRs, indicating that the heritable methylation patterns had remained largely stable in this set 
of geographically dispersed accessions (Figure S8a, e; Figure S9; Table S6). Indeed, 91% of genie and 98% of 
the TE sequence space were devoid of DMRs. Of the DMRs, 3,199 were classified as highly differentially 
methylated (hDMRs; Table S7). Their allele frequency spectrum was similar to that of DMPs (Figure 2f). 
Most DMRs and hDMRs showed highly variable methylation in only one cytosine context, often CG 
(Figure 2g). Different from DMPs, the densities for DMRs and hDMRs were highest in centromeric and 
pericentromeric regions, and overlapped more often with TEs than with genes (Figure 2a, c). In relation to 
the full complement of MRs, however, genie regions were two-fold overrepresented in the genome 
sequence covered by DMRs, and three-fold in the genome sequence covered by hDMRs (Figure 2c). In a 
recent report of 140 divergent accessions [10], DMRs were also biased towards genie regions, but not 
quite as extreme as in the HPG I lines, likely reflecting the much greater genetic variation among TEs in this 
set of accessions [10], compared to the only recently diverged HPG I lines. 

Methylation variation and transcriptome changes 

DNA methylation in gene bodies has been proposed to exclude H2A.Z deposition and thereby stabilize 
gene expression levels [41]. We therefore asked what impact differential methylation had on transcriptional 
activity. We identified 269 differentially expressed genes across all possible pairwise comparisons (Table S8, 
S9), most of which were found in more than one comparison. When we clustered accessions by 
differentially expressed genes, closely related pairs were placed together (Figure SI I). We identified 28 
differentially expressed genes that overlapped with an hDMR either in their coding or I kb upstream 
region, but the relationship between methylation and expression was variable (Table SIO). By visual 
examination of hDMRs, we found not more than five instances of demethylation that were associated with 
increased expression; examples are shown in Figure SI 2. 

Comparison between genetic and epigenetic differentiation 

With the caveat that there are uncertainties about the genetic mutation rate in the wild, and therefore how 
the number of SNPs relates to the number of generations since the last common ancestor, there was no 
evidence for faster accumulation of DMPs in the HPG I population, nor for very different epimutation rates 

6 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



among HPG I lines (Figure 2b). Importantly, the overlap between DMPs in the two populations was much 
greater than expected by chance: the chance of a random m C site in the MA population of being a DMP in 
the HPG I population was only 7%, but it was 41% among sites that were also DMPs in the MA population. 
In other words, compared to all m C sites in the MA population, DMPs in the HPG I population were four- 
to six-fold enriched among sites that were also DMPs in the MA population, and vice versa (Figure 3a) 
(SOM: Similarity of epigenetic variation profiles in independent populations). Shared DMPs were more 
heavily biased towards the chromosome arms and towards genie sequences than population-specific DMPs 
(Figure SI 3a and SI 3b). Conversely, DMPs from one population were more likely to be unmethylated 
throughout the other population when compared to random methylated sites (Figure 3a), as one might 
expect for sites that sporadically gain methylation. 

DMPs private to the HPG I lineage appeared to be less frequent in the pericentromere compared to DMPs 
private to the MA lines (Figure SI 3a), which was also reflected in an apparently higher epimutation 
frequency in the MA lines for these regions (Figure SI 3b). We therefore investigated whether the 
annotation spectrum differed between these two classes of DMPs. Even though MA-private DMPs were 
more often found in TEs compared to HPG I -private DMPs, this bias was also observed for all cytosines 
accessible to our methylome analyses (Figure SI 3c), and can therefore be explained by a more accurate 
read mapping and better TE annotation in the Col-0 reference compared to the HPG I pseudo-reference 
genome. Indeed, except for chromosome 4, the average sequencing depth in the pericentromere was 
higher in the MA lines (Figure SI 3b). 

DMPs distinguishing MA lines that were separated by only a few generations more frequently overlapped 
with HPG I DMPs than DMPs identified between distant MA lines (Figure SI 4). We interpret this 
observation as an indication of privileged sites that are more labile and therefore more likely to have 
changed in status already after a small number of generations. 

Similar to variable single positions, or DMPs, the overlap between 2,523 DMRs in the MA lines and the 
4,821 DMRs of the HPG I accessions was highly significant (Z-score = 32.9; 100,000 permutations) (Figure 
3b). We observed similar degrees of overlap independently of DMR sequence context. Overlapping DMRs 
were, in contrast to shared DMPs, not biased towards genie regions (Figure SI 5). DMRs of the HPG I 
lineage, however, overlapped with genie sequences more often than MA-DMRs (Figure SI 5), which might 
again be explained by the different efficiencies in mapping to repetitive sequences and TEs (Figure SI 3b). 
We identified DMRs that distinguish the MA and HPG I populations using a randomly chosen MA and a 
randomly chosen HPG I line; these DMRs, which differentiate distantly related accessions, were also 
enriched in each of the two sets of within-population DMRs (MA or HPG I) (Figure 3c). Finally, we 
compared HPG I -DMRs to DMRs that had been identified with a different method among 140 natural 
accessions from the global range of the species[IO] (Figure 3d). Although only 9,994, less than one fifth, of 
the DMRs from the global accessions were covered by methylated regions in the HPG I strains, the overlap 
of DMRs was highly significant (Z-score = 19.8; 100,000 permutations). In conclusion, the high recurrence 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



of DMPs and DMRs from different datasets points to the same loci being inherently biased towards 
undergoing changes in DNA methylation independently of genetic background and growth environment. 

To quantify how many methylation differences were co-segregating with genome-wide genetic changes in 
cis and trans, we estimated heritability for each hDMR by applying a linear mixed model-based method. We 
used segregating sequence variants with complete information as genotypic data and average methylation 
rates of hDMRs with complete information as phenotypes. The median heritability of all hDMRs was 0.4 1 
(mean 0.44), which means that genetic variance across the entire genome contributed less than half of 
methylation variance (Figure 4a). Regions classified as hDMRs in the HPG I strains that were not methylated 
in the greenhouse-grown MA lines had a higher median heritability, 0.48, than HPG I hDMRs also found 
among MA DMRs (0.29), which held true for all sequence contexts (Figure 4a; Figure SI 6). hDMRs found 
only in the HPG I population, especially those in unmethylated regions of the MA lines, were thus more 
likely to be linked to whole-genome genotype than hDMRs found in both populations. For 1 9% of all 
hDMRs (21% CG-hDMRs, 1 4% CHG-hDMRs, 7% CHH-hDMRs), the whole-genome genotype explained 
more than 90% of their methylation differences (with a standard error of at most 0.I). Of these hDMRs, 
half had a heritability of greater than 0.99. That 6.7% of the sequence space of these heritable hDMRs still 
overlapped with MA DMRs (versus 9.4% for the less heritable hDMRs) was in agreement with the 
hypothesis that there are regions that vary highly in their methylation status independently of genetic 
background. 

To identify genetic variants that potentially directly cause methylation changes in their local genomic 
neighborhood, we focused on DMRs with segregating SNPs or indels located within I kb. Of I9I such 
DMRs, only three showed a systematic correlation with nearby sequence polymorphisms. We noticed, 
however, that coding regions with SVs larger than 20 bp that distinguished the MA and HPG I populations 
were more likely to be methylated in both the MA and HPG I lines than non-polymorphic coding regions 
(Figure 4b). Consequently, HPG I -specific DMPs were on average closer to SVs than DMPs shared between 
the HPG I and MA populations (Figure 4c). 

Next, we asked whether the genome-wide methylation pattern reflected genetic relatedness, i.e., 
population structure. Hierarchical clustering by methylation rates of DMPs and hDMRs grouped strains by 
sampling location (Figure 4d, e). This result was largely independent of the sequence or the annotation 
context of the DMPs and hDMRs, and not seen with N-DMPs (Figure SI 7). That MRs not classified as 
DMRs (N-DMRs) grouped the accessions similar to DMPs, albeit with less confidence (shorter branch 
lengths; Figure SI 7), suggested that our DMR calling algorithm was conservative. Methylation data thus 
paralleled similarity between accessions at the genetic level, in agreement with methylation differences 
reflecting the number of generations since the last common ancestor. 



8 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



DISCUSSION 

We have tested the hypothesis that under natural conditions, epigenetic variation accumulates over the 
short term in a manner that is very different from the clock-like behavior of genetic variation [24-27], by 
taking advantage of a unique natural experiment, a lineage that has likely diverged for at least a century 
throughout North America. Our analyses have revealed little evidence for long-term heritable genome- 
wide epigenetic differentiation that might have been induced by the variable and fluctuating environmental 
conditions experienced by the HPG I accessions since they separated from each other. While the exact 
conditions these plants have been subjected to since their separation remain unknown, the time scale and 
diversity of geographic provenance are strong indicators of the variability of the environment between the 
different sampling sites. The general framework enabled by the HPG I lineage - nearly isogenic lines grown 
for more than a century under variable and fluctuating conditions - could not have been achieved in a 
controlled greenhouse experiment. 

Studies of epiRIL populations have shown that pure epialleles can be stably transmitted across 
generations[5, 1 9], but how often this is the case for environmentally induced epigenetic changes has been 
heavily debated [33,42-44]. The recent excitement about the transmission of induced epigenetic variants 
comes from such induced variants having been proposed to be more often adaptive than random genetic 
mutations [25-27]. Contrary to the expectations discussed above, we found that epimutation rates under 
natural growth conditions at different sites did not exceed those observed in a controlled greenhouse 
environment, with polymorphisms accumulating sub-linearly in both situations, apparently because of 
frequent reversions. Note that we grew the HPG I plants under controlled conditions for two generations 
after sampling at the natural site, to reduce the range of epigenetic variation to the long-term heritable 
fraction. We cannot exclude that in field-grown HPG I individuals epigenetic variation is increased and 
carries a stronger signature of the sampling site. However, such a hypothetical fraction of epigenetic 
variation, if it existed, is not heritable, because we did not find evidence for it after two extra generations in 
the greenhouse. Additional studies comparing plants grown outdoors to their progeny grown in a stable 
and controlled environment will help to further clarify these issues. 

That DMPs between closely related MA lines are more likely to overlap with HPG I DMPs than DMPs 
between more distantly related MA lines supports the hypothesis that there are different classes of 
polymorphic sites. One of these includes 'high lability' sites that are independent of the genetic background, 
that change with a high epimutation rate, and that are therefore more likely to appear in each population. 
Another class of DMPs comprises more stable sites that gain or lose methylation more slowly and that 
therefore are less likely to be shared between different populations. 

Differences between accessions in terms of DNA methylation recapitulated their genetic relatedness, 
further corroborating our hypothesis that heritable epigenetic variants arise predominantly as a function of 
time rather than as a consequence of rapid local adaptation. Epigenetic divergence thus does not become 
uncoupled from genetic divergence when plants grow in varying environments, nor does the rate of 

9 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



epimutation increase. A minor fraction of heritable epigenetic variants may be related to habitat, which 
could be responsible for LISET-036 being epigenetically a slight outlier (Figure 4e), even though it is not any 
more genetically diverged from the most recent common ancestor of HPG I than other lines. Such local 
epigenetic footprints could also explain fluctuations in epimutation frequency between the MA and HPG I 
lineages. Subtle adaptive changes at a limited number of loci would go unnoticed in the present analysis of 
genome-wide patterns and can therefore not be excluded. However, on a genome-wide scale there was 
little indication of adaptive change: neither were LISET-036 specific DMRs in and near genes enriched for 
GO terms with an obvious connection to environmental adaptation, nor were there overlapping 
differentially expressed genes (Figure SI 8, SOM: Analysis of LISET-036 specific hDMRs). In combination 
with the general lack of correlation between differential methylation and changes in gene expression, our 
findings suggest that epigenetic changes in nature are mostly neutral, and thus comparable to genetic 
mutations. 

Because of the near-isogenic background of the HPG I accessions, we were also able to gauge how much of 
epigenetic variation is either caused by, or stably co-segregates with genetic differences. HPG I -specific 
hDMRs were more often linked to genotype variation than regions that were variably methylated in both 
the HPG I and MA populations. This suggests that heritable hDMRs can, to a certain extent, be considered 
facilitated epigenetic changes [II]. 

Even though DMRs, like DMPs, are over-represented in genes, they are mainly located in TEs and 
intergenic regions, which is different from the situation for DMPs. Altogether our data indicate that both 
DMPs and constitutively methylated sites in genes are typically separated by many unmethylated sites and 
that a large fraction is therefore not classified as being within (D)MRs. Variability of DNA methylation in 
genes thus mainly affects single, sparsely distributed cytosines. Further experiments are necessary to clarify 
the biological relevance of variation of single-site DNA methylation in genie regions. 

In summary, comparisons between MA laboratory strains and natural HPG I accessions revealed that DMPs 
overlapped much more than expected by chance, despite these populations having experienced very 
different environments that also differ greatly in their stability, and despite completely different genetic 
backgrounds. The observation that changes at many sites and loci are independent of the genetic 
background and geographic provenance suggests that spontaneous switches in methylation predominantly 
reflect intrinsic properties of the DNA methylation and gene silencing machinery. Our most important 
finding is probably that DNA methylation is highly stable across dozens, if not hundreds of generations of 
growth in natural habitats; 97% of the total methylated genome space was not contained in a DMR. This is 
in stark contrast to published data describing more than 90% of the genome as variably methylated in a set 
of 1 40 divergent natural accessions[IO]. We propose that heritable polymorphisms that arise in response 
to specific growth conditions appear to be much less frequent than those that arise spontaneously. These 
conclusions are of importance when considering epimutations as a potential evolutionary force. 



10 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



MATERIAL AND METHODS 
Plant growth and material 

Accessions [35] were collected in the field at locations indicated in Table SI. Seeds had been bulked in the 
Bergelson lab at the University of Chicago before starting the experiment. Plants were then grown at the 
Max Planck Institute in Tubingen on soil in long-day conditions (23 °C, 1 6 h light, 8 h dark) after seeds had 
been stratified at 4 °C for 6 days in short-day conditions (8 h light, 1 6 h dark). We grew one plant of each 
accession under these conditions; seeds of that parental plant were then used for all experiments. Eight 
plants of the same accession were grown per pot in a randomized setup. All accessions used in this paper 
have been added to the 1 00 1 Genomes project ( http:// 1 00 1 genomes.org ) and have been submitted to the 
stock center. 

Nucleic acid extraction 

DNA was extracted from rosette leaves pooled from eight to ten individual adult plants. Plant material was 
flash-frozen in liquid nitrogen and ground in a mortar. The ground tissue was resuspended in Nuclei 
Extraction Buffer (10 mM Tris-HCI pH 9.5, 10 mM EDTA, 1 00 mM KCI, 0.5 M sucrose, O.I mM spermine, 
0.4 mM spermidine, 0. 1% b-mercaptoethanol). After cell lysis in nuclei extraction buffer containing 10% 
Triton-X- 1 00, nuclei were pelleted by centrifugation at 2000 g for 1 20 s. Genomic DNA was extracted 
using the Qiagen Plant DNeasy kit (Qiagen GmbH, Hilden, Germany). Total RNA was extracted from 
rosette leaves pooled from eight to ten individual adult plants using the Qiagen Plant RNeasy Kit (Qiagen 
GmbH, Hilden, Germany). Residual DNA was eliminated by DNasel (Thermo Fisher Scientific, Waltham, 
MA, USA) treatment. 

Library preparation 

DNA libraries for genomic and bisulfite sequencing were generated as described previously [1 2]. Libraries 
for RNA sequencing were prepared from I ug of total RNA using the TruSeq RNA sample prep kit from 
lllumina (lllumina) according to the manufacturer's protocol. 

Sequencing 

All sequencing was performed on an lllumina GAM instrument. Genomic and bisulfite-converted libraries 
were sequenced with 2 x I0I bp paired-end reads. For bisulfite sequencing, conventional A. thaliana DNA 
genomic libraries were analyzed in control lanes. Transcriptome libraries were sequenced with I0I bp 
single end reads. Four libraries with different indexing adapters were pooled in one lane; no control lane 
was used. For image analysis and base calling, we used the lllumina OLB software version 1. 8. 

I I 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Processing of genomic reads 

The SHORE pipeline vO.9.0 [45] was used to trim and quality-filter the reads. Reads with more than 2 (or 
5) bases in the first 1 2 (or 25) positions with a base quality score of less than 4 were discarded. Reads were 
trimmed to the right-most occurrence of two adjacent bases with quality values equal to or greater than 5. 
Trimmed reads shorter than 40 bases and reads with more than 10% (of the read length) of ambiguous 
bases were discarded. 

Genetic variant identification 

Genetic variants were called in an iterative approach. In each step, SNPs and structural variants common to 
all strains were detected and incorporated into a new reference genome. The thus refined "HPGI-like" 
genomes served as the reference sequence in the subsequent iterations (Figure SI). We performed three 
iterations to call segregating variants and built two reference genomes to retrieve common polymorphisms. 
The steps performed in each iteration will be described in the following. 

Read mapping 

Reads were aligned against the Arabidopsis thaliana genome sequence version TAIR9 in iteration I and 
against updated "Haplogroup I -like" genomes in further iterations. The mapping tool GenomeMapper 
v0.4.5s [46] was used, allowing for up to 10% mismatches and 7% single-base-pair gaps along the read 
length to achieve high coverage. All alignments with the least amount of mismatches for each read were 
reported. A paired-end correction method was applied to discard repetitive reads by comparing the 
distance between reads and their partner to the average distance between all read pairs. Reads with 
abnormal distances (differing by more than two standard deviations) were removed if there was at least 
one other alignment of this read in a concordant distance to its partner. The command line arguments used 
for SHORE are listed in Supplementary File I. 

SNP and small indel calling 

Base counts on all positions were retrieved by SHORE vO.9.0 [45] and a score was assigned to each site 
and variant (SNP or small indel of up to 7% of read length) depending on different sequence and alignment- 
related features. Each feature was compared to three different empirical thresholds associated with three 
different penalties (40%, 20% and 5% reduction of the score, initial score: 40). They can be found in 
Table SI 2. 

For comparisons across lines, positions were accepted if at most one intermediate penalty on their score 
was applicable to at least one strain (score > 32). In this case, the threshold for the other strains was 
lowered, accepting at most one high and two intermediate penalties (score > 1 5). In this way, information 

12 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



from other strains was used to assess sites from the focal strain under the assumption of mostly conserved 
variation, allowing the analysis of additional sites. 

Only sites sufficiently covered (> 5x) and with accepted base calls in at least half of all strains (> 7 out of 
1 3) were processed further. Variable alleles with a frequency of 100% were classified as "common" and 
variants with a lower frequency as "segregating". 

Additional SNPs were called using the targeted de novo assembly approach described below. 
Structural variant (SV) calling 

Although a plethora of SV detection tools have been developed, their predicted SV sets show little overlap 
between each other on the same data sets. Furthermore, the false positive rate of many methods can be 
drastic [47]. Hence, rather than taking the intersection of the output from different tools, which would 
yield only a low amount of SVs, we combined different tools and applied a stringent evaluation routine to 
identify as many true SVs as possible. Since SVs common to all strains should be incorporated into a new 
reference, only methods that identify SVs on a base pair level could be used. Currently, there are four 
different SV detection strategies (based on depth of coverage, paired-end mapping, split read alignments or 
short read assembly, respectively). Only tools based on split read alignments and assemblies are capable of 
pinpointing SV breakpoints down to the exact base pair. Programs that were used include Pindel v2.4t [48], 
DELLY vO.0.9 [49], SV-M vO. I [50] and a custom local de novo assembly pipeline targeted towards 
sequencing gaps (described below). We reported deletions and insertions from all tools, and additionally 
inversions from Pindel. DELLY combines split read alignments with the identification of discordant paired- 
end mappings. Thus, our SV calling made use of three out of four currently available methodologies. 

Reads for DELLY were mapped using BWA vO.6.2 [51] against the TAIR9 Col-0 reference genome to 
produce a BAM file as DELLY' s input format. 

The arguments for the command line calls of all tools are listed in Supplementary File I. 
Targeted de novo assembly 

While using a re-sequencing strategy, there are regions without read coverage ("sequencing gaps") because 
either the underlying sequence is being deleted in the newly sequenced strain, or highly divergent to the 
reference sequence, or present in the focal strain, but not represented in the read set. To access the 
strain's sequences of the first two cases, a local de novo assembly method was developed. 

Insertion breakpoints or small deletions, however, can mostly not be detected by zero coverage due to 
reads ranging with a few base pairs into or beyond the structural variants. Therefore, we defined a "core 
read region" as the read sequence without the first and last 10 nucleotides. To be able to assemble the 

13 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



latter cases, the definition of "sequencing gaps" was extended from zero-covered regions to stretches not 
spanned by a single read's core region. 

All reads aligned to the surrounding 1 00 nucleotides of such newly defined sequencing gaps as well as the 
unmappable reads from the re-sequencing approach together with their potential mapped partners 
constituted the assembly read set. Two assembly tools were used to generate contigs, SOAPdenovo2 v2.04 
[52] and Velvet vl.2.0 [53] (command line arguments in Supplementary File I). Contigs shorter than 200 bp 
were discarded. To map the remaining contigs of each assembler against the iteration-specific reference 
genome, their first and last 1 00 bp were aligned with GenomeMapper v0.4.5s [46], accepting a maximal edit 
distance of 10. If both contig ends mapped uniquely within 5,000 bp, the thus framed region on the 
reference was aligned to the contig using a global sequence alignment algorithm after Needleman-Wunsch 
('needle' from the EMBOSS v6.3.l package). In addition, non-mapping contigs were aligned with blastn 
(from the BLAST v2.2.23 package) [54] to yield even more variants. 

All differences between contig and reference sequences were parsed (including SNPs, small indels and SVs) 
for each assembly tool. Only identical variants retrieved from both assemblers were selected. 

Generating and filtering consolidated variant set of each strain 

For each strain, all variants from the SV tools and the de novo assemblies were consolidated (Figure SI a) 
and positioned to consistent locations to be comparable using the tool Dindel vl.OI [55]. In the case of 
contradicting or overlapping variants, identical variants (having the same coordinates and length after re- 
positioning) predicted by a majority of tools were chosen and the rest discarded, or all were discarded if 
there was no majority. 

Despite sequencing errors or cross-mapping artifacts of the re-sequencing approach, genomic regions 
covered by reads are generally trusted. Chances of long-range variations in the inner 50% of a mapped 
read's sequence ("inner core region" of a read) are assumed to be low, since gaps would deteriorate the 
alignment capability towards the ends of the read. 

Therefore, we filtered out variants from the consolidated variant set spanning a genomic region already 
covered by at least one inner core region of a mapped read of the corresponding strain (Figure SI a), 
assuming homozygosity throughout the genome. This "core read criterion" had to be fulfilled at each 
genomic position spanned by the variant. 

Using branched reference to validate variants 

Variants passing the core read filter in all strains were classified as common variants and were incorporated 
into the reference sequence of the previous iteration, thus replacing the reference allele. Segregating 
variants, which could not be detected in all strains, were additionally built into the reference in separate 

14 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



"haplotype regions" (or "branches" of the reference sequence) to eventually be able to assess whether 
reads preferentially mapped to the reference or the alternative haplotype sequence (Figure SI a). Linked 
variant haplotypes of a strain (distance between consecutive variants < 1 07 bp, the maximal possible span of 
a read on the reference) as well as identical haplotype regions among strains were merged into one branch 
sequence. 

For each strain, all reads were re-mapped to this new reference sequence yielding read counts at the 
variant site on each branch (rb) and at the corresponding site on the reference haplotype sequence (r re f) 
(Figure SI a). Here, the read count of a site was defined as the number of inner core regions spanning the 
site. To increase certainty of variant calling and to rule out heterozygosity, the read count of the major 
allele was tested against a binomial distribution that assumed 95% allele frequency out of a total of rb+r re f 
observations, i.e. sole presence of either the branch or the reference haplotype (if 100% had been assumed, 
it would not yield a distribution). The null hypothesis of homozygosity was rejected after P value correction 
by Storey's method[56] for q values below 0.05. 

The same variant could be part of several different haplotypes and thus, could be included into different 
branch sequences. Reads supporting this variant would map at multiple locations in the reference. 
Therefore, we allowed all aligned rather than only unique reads to contribute to read counts and omitted 
the paired-end correction procedure. 

Final sets of common and segregating variants 

We followed a similar "population-aware" approach to prefer commonalities among strains as was used for 
the SNP calling for labeling variants as being common or segregating. Here, variable sites with accumulated 
coverage over both branch and reference sequence of < 3x were marked as "missing data". If there was at 
least one haplotype in a strain with a q value above 0.05, it was assumed to be present in the population. If 
the test on the same haplotype failed in another strain, but the absolute read count of the haplotype 
sequence exceeded the alternative haplotype read count by > 2-fold, then this haplotype was considered 
present in the corresponding strain as well. 

We classified variants where at least 7 out of 1 3 strains did not show missing data as 'common' if the 
branched haplotype was present in all strains, as 'not present' if the reference haplotype was present in all 
strains, or into 'segregating* if there was support for both haplotypes. 

To combine common variants identified by the described stepwise algorithm into potentially less 
evolutionary events, we aligned 200 bp around each variant of the last iteration's genome back to the 
TAIR9 Col-0 reference genome using a global alignment strategy ('needle' from the EMBOSS v6.3.l 
package). 



15 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



In total, we found 842,103 common and 2,017 segregating polymorphisms without removing linked loci 
compared to Col-0 after two iterations, to which the different tools contributed to different extent 
depending on the variant type (Figure Sic). 

Methylome sequencing 

Genomic and bisulfite sequencing were performed as described in ref. [12]. 
Processing and alignment of bisulfite-treated reads 

The procedure followed one described [ 1 2], except that we aligned reads against the HPG I -like as well as 
against the Col-0 reference genome sequences. Command line arguments for SHORE are listed in 
Supplementary File I. 

Determination of methylated sites 

We followed the same procedures as described [12]. Here, we restricted the set of analyzed positions to 
cytosine sites with a minimum coverage of 3 reads and sufficient quality score (Q25) in at least half of all 
strains (i.e. > 7), that is, 21 million positions in total. Out of those, we identified 3.8 million methylated 
cytosines in at least one strain by applying a false discovery rate (FDR) threshold at 5%, and between 
2,120,310 and 2,927,447 methylated sites per strain (Table S4). False methylation rates retrieved from read 
mapping against the chloroplast sequence can be found in Table S4. 

Identification of differentially methylated positions (DMPs) 

We performed the same methods as in Becker et al. to obtain DMPs[l2]. First, cytosine positions were 
tested for statistical difference between both replicates of a sample using Fisher's exact test and a 5% FDR 
threshold. Because individual samples consisted of a pool of several plants, the number of DMPs between 
replicates was negligible (between 0 and 161). After excluding them, we applied Fisher's exact test on the 
3.8 million cytosine sites methylated in at least one strain for all pairwise strain comparisons. Using the 
same P value correction scheme as in Becker et a/., we identified 535,483 DMPs across all 1 3 strains. 

Identification of methylated regions (MRs) 

To statistically detect stretches of positions consistently methylated higher as their flanking regions, we 

used a Hidden Markov Model (HMM) implementation modified from Molaro and colleagues [40]. It assumes 

that the number of methylation-supporting reads at each cytosine follows a beta binomial distribution and 

that distributions over positions within and between methylated regions will differ from each other, 

providing a way to distinguish them. To do so, the HMM uses two states for high and low methylation. The 

16 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



method of Molaro and colleagues was designed for calling MRs in human samples, where the vast majority 
of methylated cytosines are in a CG context. In plants, however, one observes considerable methylation in 
all three contexts (CG, CHG and CHH), each with a different methylation rate distribution. Hence, we 
extended the HMM by learning the parameters of three different beta binomial distributions per state, one 
for each context. Additionally, in contrast to humans, only the minority of cytosines in the CG context is 
methylated, as are cytosines in the other contexts. Hence, methylation rates were inverted to find 
hypermethylated, rather than hypomethylated regions as in the original HMM implementation. 

Apart from these changes, we followed the same computational steps as described by Molaro and 
colleagues [40]: The describing parameters of the - in our case - six distributions (determining the 
emission probabilities) and the transition probabilities between states were iteratively trained (using the 
Baum-Welch algorithm) from methylation rates of all cytosines in the corresponding context throughout 
the genome. After each iteration, all cytosines were probabilistically classified into the most likely state via 
Posterior Decoding, given the trained model. After training of the HMM, i.e. after maximally 30 iterations 
or when convergence criteria were met, consecutive stretches of high methylation state were scored, in 
our case by the sum of all contained methylation rates. Next, P values were computed by testing the scores 
against an empirical distribution of scores obtained by random permutation of all cytosines throughout the 
genome. After FDR calculation, consecutive stretches in high state with an FDR < 0.05 are defined as 
methylated regions (MRs). 

The HMM was run on all genome-wide cytosines, independent of their coverage. Methylation rates were 
obtained using accumulated read counts from the strain replicates, resulting in one segmentation of the 
genome per strain. Gaps of at least 50 bp without a covered C position within a high methylation state 
automatically led to the end of the high methylation segment. Positions with a methylation rate below 10% 
at the start or end of highly methylated regions (until the first position with a rate larger than 10%), were 
assigned to the preceding or subsequent low methylation region, respectively. 

Identification of differentially methylated regions (DMRs) 

The method to identify MRs yielded 1 3 different segmentations of the genome, one for each strain. We 
selected regions being in different or highly methylated states between strains and statistically tested them 
for differential methylation (including FDR calculation) as described below. To obtain epiallele frequencies, 
we clustered strains into groups based on their pairwise comparisons and statistically tested the groupings 
against each other. Regions that showed statistically significant methylation differences between at least two 
sets of strains were identified as DMRs. Finally, because of the sensitivity of the statistical test, we 
empirically filtered DMRs for strong signals and defined highly differentially methylated regions (hDMRs). 



17 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Selecting regions to test for differential methylation 

We defined a breakpoint set containing the start and end coordinates of all predicted methylated regions. 
Each combination of coordinates in this set defined a segment to perform the test for differential 
methylation in all pairwise comparisons of the strains, if at least one strain was in a high methylation state 
throughout this whole segment (Figure S9a). To also detect quantitative differences rather than solely 
presence/absence methylation, we also compared entirely methylated regions in more than one strain to 
each other. 

Because of the sheer number of such regions, we applied the following greedy filter criteria: Regions were 
discarded from any pairwise comparison if less than 2 strains contained at least 10 cytosines covered by at 
least 3 reads each (accumulated over strain replicates) in this region (Figure S9a (a)). Regions were 
discarded from any pairwise comparison if the reciprocal overlap of this region to at least one previously 
tested region was more than or equal to 70% (Figure S9a (b)). This was done to prevent "similar" regions 
to be tested twice. Pairwise tests of a region were not performed if both strains were in low methylation 
state throughout the whole region (Figure S9a (c)). Strains were excluded from pairwise comparisons in a 
region if the number of positions covered by at least 3 reads each was less than half of the maximum 
number of such positions of all strains in the same region (Figure S9a (d)). This prevented comparing 
regions with unbalanced coverage to each other, e.g. a strain with 10 data points against another one with 
only 2. 

These filters reduced the set of regions to test from -2.5 million to -230,000 per pairwise comparison. 
Testing regions for differential methylation between strains 

We designed a statistical test for differential methylation between two strains for a given region. The test 
assumes that the number of methylated and unmethylated read counts per position along a region follows a 
beta binomial distribution - similar to the HMM in MR calling. More precisely, there are 3 distributions for 
each sequence context and for each strain. Using gradient-based numerical maximum likelihood 
optimization, we fitted the parameters for each beta binomial distribution on the available read count data 
of the region in the respective strain. This was done a) for each of the two strains separately (while taking 
strain replicates into account), resulting in (two times three) strain-specific beta binomial distributions, and 
b) for the read counts of both strains including their replicates together, resulting in (three) common beta 
binomial distributions. In this way, we obtained each distribution's mean \i and standard deviation a. We 
selected only regions for potential DMRs, whose intervals - 2a i, [i\ + 2ai] for strain I and [^2 - 2a2, 
\ii + 2a2] for strain 2 did not overlap. 

To further corroborate statistical significance, we computed P values by calculating the ratio of the strain- 
specific and the common log likelihoods of the available read count data using the corresponding beta 

18 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



binomial distributions and by testing it against a chi-squared distribution (with 6 degrees of freedom). Let 
sample S have Ns c cytosines in context c in total and Cs c p reads at position p in context c, from which xs c p 
are methylated, then we compute: 



a max a , fe (D^f {^j)B(a + x A( , t . h H- C A( , f - s A( , t )) * fe^f (^-)^(« + t + C /Jai - )) 

\ 1 — , ^! i — 



After correction for multiple testing using Storey's method[56], an FDR threshold of 0.01 defined 
statistically different methylated regions (DMRs) between two strains. 

Additionally, this method allowed calling differential methylation in a region for each context separately by 
computing P values as described above without summing over the contexts (c = I, 2 or 3). We termed 
resulting DMRs CG-DMRs if the methylation at only CG sites within this region was statistically significantly 
different, and similarly CHG-DMRs and CHH-DMRs. 



Grouping differentially methylated strains in each region 

For 1 3 strains there are at maximum 78 pairwise comparisons per region. To summarize pairwise 
comparisons and obtain epiallele frequencies, we assigned strains into differentially methylated groups. To 
achieve such clustering, we constructed a graph for each region where strains were represented as vertices 
and connected to other strains by an edge if the region was identified as a DMR between them (Figure 
S9b). We assume that strains within a group are then similarly methylated. The task is to find the smallest 
number of groups of vertices so that no two strains within a group are connected by an edge. 

We set up a custom algorithm, which iteratively solves the "vertex coloring problem" for an increasing 
number of different colors, starting with two and quitting once all strains could be successfully assigned a 
color (Figure S9b). In each iteration, strains were processed in descendent order of their degree (i.e. 
number of edges it is connected to). Each strain was assigned to all possible colors that did not invoke a 
collision. Subsequently, the algorithm continued recursively to assign the color of the next strain. 

Each strain had 3 context-dependent means of its beta binomial distributions per region (termed strain 
means from now on). We roughly approximated each group's mean methylation values (group means) as the 
mean values of all strain means within a group. The grouping diversity describes the accumulated absolute 
differences between the strain means and their respective group means divided by the number of strains. As 
an example, consider Figure S9b. For simplicity, it only displays methylation rates for one out of three 
contexts. In the real data, the respective values were accumulated over all three contexts. The group mean 
for the blue strains in the example is (89+90+90+93+87)/5 = 89.8% and for the white strains 52%. The 
grouping diversity of the clustering shown here would be (from strains A to K): (|56-52| + |59-52| + |64-52| + 
|89-89.8| + |4I-52| + |93-89.8| + |90-89.8| + |45-52| + |47-52| + |90-89.8| + |45-52| + |87-89.8|) / I I = 2.84. 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



If there was more than one possible grouping of the strains, we chose the one with the lowest grouping 
diversity. A strain with no edges (i.e. which is not statistically differentially methylated to any other strain) 
was assigned into the group to which the accumulated absolute difference between its strain mean and the 
group mean was lowest. In the example of Figure S9b, strain L is grouped to the blue strains because its 
mean methylation value (81%) is closer to the blue group mean (90%) than to the white one (52%). 

This procedure summarized the ~22l,000 DMRs of all pairwise strain comparisons into I 1,323 DMRs 
between groups of strains. 

Testing regions for differential methylation between groups of strains 

Once grouped, the same statistical test as for differential methylation between two strains was used to test 
groups of strains. Beta binomial distributions were approximated using the read counts of all strains in a 
group as if they were replicate data. This procedure identified 10,645 groups of regions showing 
significantly different methylation. Because the method used for the selection of the regions to perform the 
differential test can result in overlapping regions, DMRs can still overlap each other. From sets of 
overlapping DMRs, the non-overlapping DMR(s) with the lowest 'grouping diversity' was (were) retained, 
resulting in 4,821 final DMRs. For the vast majority of DMRs (98%), strains were classified into two groups, 
i.e. there are only two epialleles. 

Heritability analysis of methylated regions 

For each differentially methylated region, we considered a linear mixed model to estimate the proportion 
of variance that is attributable to genetic effects (heritability) and its standard error. The approach is similar 
to variance component models used in GWAS, e.g. refs. [57,58]. Briefly, we considered the log average 
methylation rate of DMRs as phenotype and assessed the variance explained by genotype using a Kinship 
model constructed from all segregating genetic variants. We considered only DMRs and genetic 
polymorphisms that had no missing data in all accessions. 

Population structure analysis 

We identified non-synonymous SNPs using SHOREmap_annotate [59] and excluded them from population 
structure analyses. We ran STRUCTURE v.2.3.4 [60] with K=2 to K=9 with a burn-in of 50,000 and 
200,000 chains for 10 repetitions and determined the best K value using the AK method [61]. The 
phylogenetic network was generated using SplitsTree v.4. 12.3 [62]. 



20 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Mapping to genomic elements 

We used the TAIRIO annotation for genes, exons, introns and untranslated regions; transposon annotation 
was done according to [63]. Positions and regions were hierarchically assigned to annotated elements in 
the order CDS > intron > 5' UTR > 3' UTR > transposon > intergenic space. We defined as intergenic 
positions and regions those that were not annotated as either CDS, intron, UTR or transposon. 

Positions were associated to the corresponding element when they were contained within the boundaries 
of that element. (D)MRs were associated to a class of element if they overlapped with that class of element; 
a (D)MR could only be associated to one class of element. When summing up basepairs of an element class 
covered by (D)MRs, the number of basepairs of a (D)MR overlapping with that class of element were 
considered. In that case the space covered by a (D)MR could be assigned to different classes of elements, 
while each basepair of the (D)MR could be assigned to only one class. 

Overlapping region analysis 

We tested for significant overlap of DMRs using multovl version 1. 2 (Campus Science Support Facilities 
GmbH (CSF), Vienna, Austria). We reduced the genome space to the basepair space covered by MRs 
identified in at least one HPG I accession. DMRs were considered in the analysis if their start and end 
positions were contained within the MR space. DMRs that only partially overlapped with the MR space 
were trimmed to the overlapping part. Overlap between DMRs from different datasets was analyzed by 
running 100,000 permutations of both DMR sets within the MR basepair space, multovl commands are 
listed in Supplementary File I. 

Processing and alignment of RNA-seq reads 

Reads were processed in the same way as genomic reads, except that trimming was performed from both 
read ends. Filtered reads were then mapped to the TAIR9 version of the Arabidopsis thaliana 
( http://www.arabidopsis.org ) genome using Tophat version 2.0.8 with Bowtie version 2.1.0 [64,65]. 
Coverage search and microexon search were activated. The command lines for Tophat are listed in 
Supplementary File I. 

Gene expression analysis 

or quantification of gene expression we used Cufflinks version 2.0.2[66]. We ran a Reference Annotation 
Based Transcript assembly (RABT) using the TAIRIO gene annotation 

( ftp://ftp.arabidopsis.org/home/tair/Genes/TAIRIO genome release/TAIRI 0 gff3/ ) supplied with the most recent 
transposable element annotation [63] Fragment bias correction, multi-read correction and upper quartile 
normalization were enabled; transcripts of each sample were merged using Cuffmerge version 2.0.2, with 

21 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



RABT enabled. For detection of differential gene expression we ran Cuffdiff version 2.0.2 on the merged 
transcripts; FDR was set to < 0.05 and the minimum number of alignments per transcripts was 1 0. 
Fragment bias correction, multi-read correction and upper quartile normalization were enabled. The 
command lines for the Cufflinks pipeline are listed in Supplementary File I. Analysis and graphical display of 
differential gene expression data was done using the cummeRbund package version 2.0.0 under R version 
3.0.1. 

Data visualization 

When not mentioned otherwise in the corresponding paragraph, graphical displays were generated using R 
version 3.0.1 ( www. r- p ro j ect. o rg ) . Circular display of genomic information in Figure 2a was rendered using 
Circos version 0.63 [67]. 

Phenotyping 

Leaf area was determined using the automated IPK LemnaTec System and the IAP analysis pipeline [68]. 
Plants were grown in a controlled-environment growth-chamber in an alpha-lattice design with eight 
replicates and three blocks per replicate, taking into account the structural constraints of the LemnaTec 
system. Each block consisted of eight carriers, each carrying six plants of one line. Stratification for 2 days 
at 6°C was followed by cultivation at 20/l8°C, 60/75% relative humidity in a 16/8 h day/night cycle. Plants 
were watered and imaged daily until 21 days after sowing (DAS). Adjusted means were calculated using 
REML in Genstat 14 th Edition, with genotype and time of germination as fixed effects, and replicate|block as 
random effects. 

Data accessibility 

The DNA and RNA sequencing data have been deposited at the European Nucleotide Archive under 
accession number XXX and XXX. A GBrowse instance for DNA methylation and transcriptome data is 
available at (to be released upon publication). DNA methylation data and MR coordinates have also been 
uploaded to the EPIC-CoGe browser (data will be made publicly available upon publication of the 
manuscript). 

ACKNOWLEDGEMENTS 

We are grateful to C. Lanz for help with the lllumina sequencing, Q. Song and A. Smith for making the 
source code of the Hidden Markov Model implementation available, the group of V. Colot for sharing the 
Col-0 MeDIP-Seq data, and C. Klukas for assistance with the processing of the phenotyping data. We thank 
R. Schwab for critical reading of the manuscript. This work was supported by a Marie Curie FP7 fellowship 

22 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



(O.S.), grant NIH GM083068 (J.B.), FP7 Collaborative Project AENEAS (contract KBBE-2009-226477), a 
Gottfried Wilhelm Leibniz Award of the DFG, and the Max Planck Society (D.W.). 



AUTHOR CONTRIBUTIONS 

C.B., J.H., T.A., J.B., and D.W. conceived the study; C.B. and R.C.M. performed the experiments; J.H., C.B., 
J.M., O.S., K.S. analyzed the data; J.F. implemented the data visualization; K.B. provided advice on statistical 
analysis; and C.B., J.H. and D.W. wrote the paper with contributions from all authors. 



COMPETING FINANCIAL INTERESTS 

The authors declare that no competing interests exist. 



AUTHOR INFORMATION 

Correspondence and requests for materials should be addressed to D.W. (e-mail: 
weigel@weigelworld.org ). 

SUPPORTING INFORMATION 

Supplementary information is linked to the paper. 



23 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



REFERENCES 

1 . Martin A, Troadec C, Boualem A, Rajab M, Fernandez R, et al. (2009) A transposon-induced 

epigenetic change leads to sex determination in melon. Nature 46 1: I 1 35- 1 1 38. 

2. Wang X, Weigel D, Smith LM (20 1 3) Transposon variants and their effects on gene expression 

in Arabidopsis. PLoS Genet 9: el 003255. 

3. Bender J, Fink GR (1 995) Epigenetic control of an endogenous gene family is revealed by a novel 

blue fluorescent mutant of Arabidopsis. Cell 83: 725-734. 

4. Melquist S, Luff B, Bender J (1 999) Arabidopsis PAI gene arrangements, cytosine methylation and 

expression. Genetics 1 53: 40 1 -4 1 3. 

5. Silveira AB, Trontin C, Cortijo S, Barau J, Del Bern LE, et al. (20 1 3) Extensive natural epigenetic 

variation at a de novo originated gene. PLoS Genet 9: el 003437. 

6. Gan X, Stegle O, Behr J, Steffen JG, Drewe P, et al. (20 1 I) Multiple reference genomes and 

transcriptomes for Arabidopsis thaliana. Nature 477: 4 1 9-423. 

7. Cao J, Schneeberger K, Ossowski S, Gunther T, Bender S, et al. (20 1 I) Whole-genome 

sequencing of multiple Arabidopsis thaliana populations. Nat Genet 43: 956-963. 

8. Schneeberger K, Ossowski S, Ott F, Klein JD, Wang X, et al. (20 1 I) Reference-guided assembly 

of four diverse Arabidopsis thaliana genomes. Proc Natl Acad Sci USA 1 08: 1 0249- 1 0254. 

9. Long Q, Rabanal FA, Meng D, Huber CD, Farlow A, et al. (20 1 3) Massive genomic variation and 

strong selection in Arabidopsis thaliana lines from Sweden. Nat Genet 45: 884-890. 

10. Schmitz RJ, Schultz MD, Urich MA, Nery JR, Pelizzola M, et al. (20 1 3) Patterns of population 

epigenomic diversity. Nature 495: 1 93- 1 98. 
I I. Richards EJ (2006) Inherited epigenetic variation - revisiting soft inheritance. Nature reviews 
Genetics 7: 395-40 1. 

1 2. Becker C, Hagmann J, Miiller J, Koenig D, Stegle O, et al. (20 1 I) Spontaneous epigenetic 

variation in the Arabidopsis thaliana methylome. Nature 480: 245-249. 

1 3. Schmitz RJ, Schultz MD, Lewsey MG, O'Malley RC, Urich MA, et al. (20 1 I) Transgenerational 

epigenetic instability is a source of novel methylation variants. Science 334: 369-373. 

1 4. Alabert C, Groth A (20 1 2) Chromatin replication and epigenome maintenance. Nat Rev Mol 

Cell Biol 1 3: 1 53- 1 67. 

1 5. Genereux DP, Miner BE, Bergstrom CT, Laird CD (2005) A population-epigenetic model to 

infer site-specific methylation rates from double-stranded DNA methylation patterns. Proc 
Natl Acad Sci USA 1 02: 5802-5807. 

1 6. Liu Q, Gong Z (20 1 I) The coupling of epigenome replication with DNA replication. Curr Opin 

Plant Biol 1 4: 1 87- 1 94. 

24 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



1 7. LawJA, Jacobsen SE (20 1 0) Establishing, maintaining and modifying DNA methylation patterns 

in plants and animals. Nat Rev Genet I 1 : 204-220. 

1 8. Calarco JP, Borges F, Donoghue MT, Van Ex F, Jullien PE, et al. (20 1 2) Reprogramming of DNA 

methylation in pollen guides epigenetic inheritance via small RNA. Cell I5I: 1 94-205. 

1 9. Teixeira FK, Heredia F, Sarazin A, Roudier F, Boccara M, et al. (2009) A role for RNAi in the 

selective correction of DNA methylation defects. Science 323: 1 600- 1 604. 

20. Latzel V, Allan E, Bortolini Silveira A, Colot V, Fischer M, et al. (20 1 3) Epigenetic diversity 

increases the productivity and stability of plant populations. Nat Commun 4: 2875. 

21. Roux F, Colome-Tatche M, Edelist C, Wardenaar R, Guerche P, et al. (20 1 I) Genome-wide 

epigenetic perturbation jump-starts patterns of heritable variation found in nature. 
Genetics 1 88: I0I5-I0I7. 

22. Cortijo S, Wardenaar R, Colome-Tatche M, Gilly A, Etcheverry M, et al. (20 1 4) Mapping the 

epigenetic basis of complex traits. Science: ePub Feb 06, 20 1 4. 

23. Dowen RH, Pelizzola M, Schmitz RJ, Lister R, Dowen JM, et al. (20 1 2) Widespread dynamic 

DNA methylation in response to biotic stress. Proc Natl Acad Sci U S A 1 09: E2 1 83-2 1 9 1 . 

24. Jablonka E, Raz G (2009) Transgenerational epigenetic inheritance: prevalence, mechanisms, 

and implications for the study of heredity and evolution. Q Rev Biol 84: 1 3 I - 1 76. 

25. Bonduriansky R (20 1 2) Rethinking heredity, again. Trends Ecol Evol 27: 330-336. 

26. Bossdorf O, Richards CL, Pigliucci M (2008) Epigenetics for ecologists. Ecol Lett II: 1 06- 1 1 5. 

27. Danchin E, Charmantier A, Champagne FA, Mesoudi A, Pujol B, et al. (20 1 I) Beyond DNA: 

integrating inclusive inheritance into an extended theory of evolution. Nat Rev Genet 1 2: 
475-486. 

28. Bergman Y, Cedar H (20 1 3) DNA methylation dynamics in health and disease. Nat Struct Mol 

Biol 20: 274-28 1. 

29. Feil R, Fraga MF (20 1 I) Epigenetics and the environment: emerging patterns and implications. 

Nat Rev Genet 13:97-109. 

30. Geoghegan JL, Spencer HG (2012) Population-epigenetic models of selection. Theor Popul Biol 

81:232-242. 

31. Daxinger L, Whitelaw E (2012) Understanding transgenerational epigenetic inheritance via the 

gametes in mammals. Nat Rev Genet 13: 153-162. 

32. Mirouze M, Paszkowski J (201 I) Epigenetic contribution to stress adaptation in plants. Curr 

Opin Plant Biol 14: 267-274. 

33. Paszkowski J, Grossniklaus U (2011) Selected aspects of transgenerational epigenetic 

inheritance and resetting in plants. Curr Opin Plant Biol 14: 195-203. 



25 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



34. Becker C, Weigel D (20 1 2) Epigenetic variation: origin and transgenerational inheritance. Curr 

Opin Plant Biol 1 5: 562-567. 

35. Piatt A, Horton M, Huang YS, Li Y, Anastasio AE, et al. (20 1 0) The scale of population 

structure in Arabidopsis thaliana. PLoS Genet 6: el 000843. 

36. Ossowski S, Schneeberger K, Lucas-Lledo Jl, Warthmann N, Clark RM, et al. (20 1 0) The rate 

and molecular spectrum of spontaneous mutations in Arabidopsis thaliana. Science 327: 92- 
94. 

37. O'Kane SL, Al-Shehbaz IA (1 997) A synopsis of Arabidopsis (Brassicaceae). Novon 7: 323-327. 

38. Zhang X, Yazaki J, Sundaresan A, Cokus S, Chan SW, et al. (2006) Genome-wide high- 

resolution mapping and functional analysis of DNA methylation in Arabidopsis. Cell 1 26: 
I I89-I20I. 

39. Seifert M, Cortijo S, Colome-Tatche M, Johannes F, Roudier F, et al. (20 1 2) MeDIP-HMM: 

genome-wide identification of distinct DNA methylation states from high-density tiling 
arrays. Bioinformatics 28: 2930-2939. 

40. Molaro A, Hodges E, Fang F, Song Q, McCombie WR, et al. (20 1 I) Sperm methylation profiles 

reveal features of epigenetic inheritance and evolution in primates. Cell 1 46: 1 029- 1 04 1. 

41. Coleman-Derr D, Zilberman D (20 1 2) Deposition of histone variant H2A.Z within gene bodies 

regulates responsive genes. PLoS Genet 8: el 002988. 

42. Daxinger L, Whitelaw E (20 1 2) Understanding transgenerational epigenetic inheritance via the 

gametes in mammals. Nat Rev Genet 1 3: 1 53- 1 62. 

43. Manolio TA, Collins FS, Cox NJ, Goldstein DB, Hindorff LA, et al. (2009) Finding the missing 

heritability of complex diseases. Nature 46 1: 747-753. 

44. Slatkin M (2009) Epigenetic inheritance and the missing heritability problem. Genetics 1 82: 

845-850. 

45. Ossowski S, Schneeberger K, Clark RM, Lanz C, Warthmann N, et al. (2008) Sequencing of 

natural strains of Arabidopsis thaliana with short reads. Genome Res 1 8: 2024-2033. 

46. Schneeberger K, Hagmann J, Ossowski S, Warthmann N, Gesing S, et al. (2009) Simultaneous 

alignment of short reads against multiple genomes. Genome Biol 10: R98. 

47. Mills RE, Walter K, Stewart C, Handsaker RE, Chen K, et al. (20 1 I) Mapping copy number 

variation by population-scale genome sequencing. Nature 470: 59-65. 

48. Ye K, Schulz MH, Long Q, Apweiler R, Ning Z (2009) Pindel: a pattern growth approach to 

detect break points of large deletions and medium sized insertions from paired-end short 
reads. Bioinformatics 25: 2865-287 1. 

49. Rausch T, Zichner T, Schlattl A, Stiitz AM, Benes V, et al. (20 1 2) DELLY: structural variant 

discovery by integrated paired-end and split-read analysis. Bioinformatics 28: i333-i339. 

26 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



50. Grimm D, Hagmann J, Koenig D, Weigel D, Borgwardt K (20 1 3) Accurate indel prediction 

using paired-end short reads. BMC Genomics 1 4: 1 32. 

5 1. Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler 

transform. Bioinformatics 25: 1 754- 1 760. 

52. Luo R, Liu B, Xie Y, Li Z, Huang W, et al. (20 1 2) SOAPdenovo2: an empirically improved 

memory-efficient short-read de novo assembler. GigaScience I: 1 8. 

53. Zerbino DR, Birney E (2008) Velvet: algorithms for de novo short read assembly using de 

Bruijn graphs. Genome Res 1 8: 82 1 -829. 

54. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1 990) Basic local alignment search tool. 

J Mol Biol 2 1 5: 403-4 1 0. 

55. Albers CA, Lunter G, MacArthur DG, McVean G, Ouwehand WH, et al. (20 1 I) Dindel: 

accurate indel calls from short-read data. Genome Res 21: 96 1 -973. 

56. Storey JD, Tibshirani R (2003) Statistical significance for genomewide studies. Proc Natl Acad 

Sci USA 1 00: 9440-9445. 

57. Kang HM, Sul JH, Service SK, Zaitlen NA, Kong SY, et al. (20 1 0) Variance component model to 

account for sample structure in genome-wide association studies. Nat Genet 42: 348-354. 

58. Lippert C, Listgarten J, Liu Y, Kadie CM, Davidson Rl, et al. (20 1 I) FaST linear mixed models 

for genome-wide association studies. Nat Methods 8: 833-835. 

59. Schneeberger K, Ossowski S, Lanz C, Juul T, Petersen AH, et al. (2009) SHOREmap: 

simultaneous mapping and mutation identification by deep sequencing. Nat Methods 6: 550- 
55I. 

60. Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus 

genotype data. Genetics 1 55: 945-959. 

61. Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using 

the software STRUCTURE: a simulation study. Mol Ecol 1 4: 26 1 I -2620. 

62. Huson DH, Bryant D (2006) Application of phylogenetic networks in evolutionary studies. Mol 

Biol Evol 23: 254-267. 

63. Slotte T, Hazzouri KM, Agren JA, Koenig D, Maumus F, et al. (20 1 3) The Capsella rubella 

genome and the genomic consequences of rapid mating system evolution. Nat Genet 45: 
83 1 -835. 

64. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, et al. (20 1 3) TopHat2: accurate alignment 

of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol 
1 4: R36. 

65. Langmead B, Salzberg SL (20 1 2) Fast gapped-read alignment with Bowtie 2. Nat Methods 9: 

357-359. 

27 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



66. Trapnell C, Hendrickson DG, Sauvageau M, Goff L, Rinn JL, et al. (20 1 3) Differential analysis of 

gene regulation at transcript resolution with RNA-seq. Nat Biotechnol 3 1: 46-53. 

67. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, et al. (2009) Circos: an information 

aesthetic for comparative genomics. Genome Res 1 9: 1 639- 1 645. 

68. Klukas C, Pape JM, Entzian A (20 1 2) Analysis of high-throughput plant image data with the 

information system IAP. J Integr Bioinform 9: I9I. 



28 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 

FIGURE LEGENDS 




Transitions Transversions Allele frequency 



Figure I: North American HPGI accessions belong to a genetically homogeneous 
population. 

(a) Sampling locations of the 1 3 strains analyzed in this study. Pie charts indicate population structure 
inferred from segregating SNP data. Data were analyzed using STRUCTURE[60], with K=6. 
CT=Connecticut, IL=lllinois, IN=lndiana, MI=Michigan, NJ=New Jersey, NY=New York, WI=Wisconsin. (b) 
Single-nucleotide mutation spectrum. Bars represent the accession average, error bars indicate 95% 
confidence intervals, (c) Phylogenetic network of HPGI accessions based on segregating SNPs and SVs with 
SplitsTree v.4. 1 2.3 [62]. Numbers indicate bootstrap confidence values (10,000 iterations). Dashed line 
delimits close-up in Figure S4. (d) Allele frequencies of SNPs. 



29 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



























0 


1 



b 60 



Relative density 



10- 



0 




















V 




HPG1 





































































100 200 300 400 
SNPs 

CHH BCHG BCG 






3 



8 
.6 

c 
g 

o .4 

LL 

.2 
.0 



Inside Outside 
MRs MRs 

(4,404) (25,333) 



I DMRs 
I hDMRs 



UL 

0 0 J5> ^ 0^ ° 

<y & & & & 



Figure 2: Epigenetic variation in a nearly isogenic population. 

(a) Genome-wide features: average coverage in 1 00 kb windows, the remainder in 500 kb windows. 
Outside coordinates in Mb. (b) DMP number in relation to SNP number in pairwise comparisons. MA data 
are based on single individuals, HPGI data on pools of 8-I0 individuals; each data point represents an 
independent comparison of two lines. DMPs in each pairwise contrast were scaled to the number of 
methylated sites compared, (c) Annotation of cytosines in MRs and hDMRs. (hD)MR sequences were 
assigned to only one annotation in the following order: CDS > intron > UTR > transposon > intergenic. (d) 
Sequence context of methylated positions relative to MRs and DMRs. (e) Fraction of 5m CGs among all CG 
sites for each gene and transposable element, with at least 5 CGs. (f) Minor epiallele frequencies of 2,304 
hDMRs that could be split into only two groups and for which at least four strains showed statistically 
significant differential methylation. Strains not tested statistically significant for a particular hDMR were not 
considered for this plot, (g) DMRs and hDMRs according to sequence contexts in which significant 
methylation differences were found. 'C' denotes (h)DMRs in all three contexts. 



30 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Status in MA Status in HPG1 Status in MA Status in HPG1 Status in MA Status in HPG1 Status in HPG1 Status in 



Schmitz et al. 




5m Csin DMPsin 5m Cs in DMPs MRs in DMRs in MRs DMRs MRs MA DMRs MA MRs MA DMRs MA DMRs in DMRs in 

HPG1 HPG1 MA in MA HPG1 HPG1 in MA in MA / HPG1 vs. HPG1 / HPG1 vs. HPG1 Schmitz et al. HPG1 



Figure 3: Epigenetic variation in independent populations and conditions. 

(a) Comparison of 5m Cs and DMPs identified in pairwise comparisons of HPGI and MA lines' 2 . Left: sites in 
HPGI strains and their status in the MA data; right: sites in the MA strains and their status in the HPGI 
data, (b) Comparison of MRs and DMRs identified in pairwise comparisons of HPGI and MA lines. Left: 
regions in HPGI strains and their status in the MA data; right: regions in the MA strains and their status in 
the HPGI data, (c) MRs and DMRs identified in comparison between one randomly chosen MA line (30-39) 
and one randomly chosen HPGI line (MuskSP-68), and their overlap with within-population DMRs. (d) 
Comparison of HPGI DMRs with DMRs identified in 1 40 natural A. thaliana accessions [10]. Because MRs 
were not reported in ref. [10], the overlap of DMRs with non-DMRs could not be assessed. 



31 



Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



*1.0- 



> _ i 



Status in 
MA lines 



' 80 -| Conserved sequence 
Divergent sequence 

i 60 



o 20 



1 0 



Jjjj 

J ■ ■ ■ ■ 



2 Q 
o o 
CD 



2 Q 
o O 
CD 



MA 



HPG1 



Methylation rate (%) 0 




iBfflA n n n n n lift M fin 



p 4,000- 



2,000- 



Shared HPG1- 
DMPs private 
DMPs 



cocococo^^cococx>cx>c^c^r^r^cx>cx>^^ooc\]c\i'si-'si-r^h- cocococofst-cococMCMr- 



t en oxo cocDCD^i-^i-^i-^-cDCDi^t^i-i-cDCDcccc^Ln co co i- ■■- a? cold ID CD CD IT 



(J)ifi cn cn u.u.u.u.0) WC 0 CO^O-O. mDD COCO Q_ Q_ CO CO CO CO ^5 5" collllllll 

jj^zzzzmcDmffi' 2 ^ [nnimm^^^zzzz 

**** 22^222 



Figure 4: Genetic effects on epigenetic variation. 

(a) Heritability values based on genome-wide genetic differentiation for all hDMRs, hDMRs with randomly 
permuted methylation rates and subsets of hDMRs depending on their overlap with MA-MRs and MA- 
DMRs. P: Permuted (2,945 hDMRs); A: All (2,945); U: Unmethylated in MA (1,310); M: Methylated in MA 
(1,243); D: DMR in MA (392). (b) Correlation between SVs and probability of overlap with MRs. Divergent 
sequences are insertions of at least 20 bp relative to the other population. This analysis is based on 3,256 
SVs overlapping with genes, 641 with CDS and 4,020 with TEs (Table SI I), (c) Distances between common 
SVs of at least 20 bp and the closest DMP, depending on whether it is shared between the MA and HPG I 
populations. Triangles represent the mean, (d) Hierarchical clustering of HPG I strains based on 
methylation rates at 50,000 CG-DMPs. (e) Hierarchical clustering of HPG I strains based on average 
methylation rates of 2,829 hDMRs with full information across all strains. Methylation rates per region 
were calculated as the average methylation rate of each methylated cytosine in that region. 



32 



Downloaded from http://biorxiv.org/ on September 18, 2014 



Supplementary Online Material 

Century-scale methylome stability in a recently diverged 
Arabidopsis tha liana lineage 

Jorg Hagmann, Claude Becker, Jonas Muller, Oliver Stegle, Rhonda C. Meyer, Korbinian Schneeberger, Joffrey Fitz, Thomas 
Altmann, Joy Bergelson, Karsten Borgwardt, Detlef We/gef 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Genome analysis of HPGI individuals 

To answer how heritable epigenetic differences are affected by long-term exposure to fluctuating and diverse 
environmental conditions, we wished to identify a nearly isogenic lineage among natural accessions of A. thaliana that 
had diverged for at least several dozens of generations. In the natural range of the species in Eurasia, nearly isogenic 
individuals are generally found only at single sites, many of which are transient [1, 2]. An exception is North America, 
where about half of all individuals sampled appear to be identical when genotyped at 1 39 genome-wide markers [2], in 
agreement with a limited number of founders that were introduced from Europe in historic times [3]. We here refer 
to this lineage as haplogroup-l (HPG I), because it dominates the North American population. 

We selected 1 3 HPGI individuals from seven locations in the Eastern Lake Michigan area, from one location in 
Western Illinois, and one location on Long Island; the median distance between sites was 1 55 km. The set consisted of 
pairs of accessions from each of four sites, and single individuals from the other five sites (Figure la, Table SI). 

We assessed genetic divergence among the 1 3 HPG I lines by lllumina paired-end sequencing (see table below). 

Reads 







Reads 


Average 


Positions 


mapped 


Positions 


Accession 


Accession 


mapped 


coverage 


covered > 


to HPGI 


covered > 


ID 


name 


to Col-0 


(x) 


5x 


reference 


5x 


8699 


328PNA-062 


71,064,284 


48 


107,019,787 


72,348,378 


108,702,343 


470 


BRR-4 


69,155,280 


45.2 


106,981,548 


70,401,862 


108,680,008 


504 


BRR-57 


41,1 19,975 


28.9 


106,831,177 


41,814,926 


108,627,670 


1 739 


KBS-Mac-68 


65,629,252 


42.6 


106,955,923 


66,742,616 


108,677,940 


I74I 


KBS-Mac-74 


61,925,749 


39 


106,926,171 


62,923,233 


108,665,771 


742 


LISET-036 


78,397,323 


42 


107,083,267 


62,671,495 


108,712,472 


1 942 


MNF-Che-47 


42,262,867 


28.2 


106,848,829 


42,925,314 


108,634,616 


1 943 


MNF-Che-49 


43,052,687 


28.8 


106,844,153 


43,729,285 


108,622,042 


208 1 


MuskSP-68 


63,558,21 1 


42.7 


106,989,784 


42,298,899 


108,688,853 


2I06 


MSGA-I0 


41,684,061 


26 


106,809,482 


64,718,515 


108,608,920 


2I59 


Paw- 1 3 


68,356,951 


43.9 


106,948,820 


69,661,005 


108,660,130 


2370 


Yng-4 


69,221,983 


41.7 


106,965,670 


70,384,702 


108,679,280 


24I2 


Yng-53 


67,134,738 


46.5 


107,026,441 


68,362,136 


108,707,709 



We identified an initial set of common single-nucleotide polymorphisms (SNPs), small-scale indels and structural 
variants (collectively referred to as SVs henceforth) by read alignment to the Col-0 reference genome. We iteratively 
built an HPGI pseudo-reference genome through integration of the common variants into the Col-0 genome, re- 
alignment of HPGI reads to this new reference and re-calling of SNPs and SVs (Figure SI a, c), following the rationale 
of Gan and colleagues [4]. After two iterations, the number of common variants had increased by 12% and the 
number of reads that could not be mapped had decreased by a third (Figure Sib). We ultimately identified 670,979 
common SNPs and 170,998 SVs compared to the Col-0 reference (Table S2). Considering the corresponding 
nucleotides in the close relative A. lyrata as the ancestral states [5], a little bit more than half of the SNPs at positions 
alignable to the A. lyrata genome were classified as derived in the HPG I population, close to what would be expected 
for a comparison of an arbitrary set of accessions (i.e., the HPG I ancestor and Col-0) (Table S2). 

We additionally called variants in each of the 13 individuals based on the HPGI pseudo-reference genome. Compared 
to the common variants, a much smaller number, 1,354 SNPs and 521 SVs, segregated in the HPGI population (Table 
S3), confirming that the 13 strains were indeed closely related. As for common variants, segregating variants were 
fairly equally distributed across all ten chromosome arms (Figure S2). Eighteen percent of both segregating and 



S2 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



common SNPs mapped to genes, and 30% of segregating and 35% of common SNPs to non-transposable element 
intergenic regions (Figure S3; Table S3). This was similar to SNPs in other natural accessions [6], indicating that HPG I 
is representative of natural accessions of A. thaliana. On average, two HPG I accessions differed by 294 SNPs. 

Primary analysis of methylation 

We performed whole methylome bisulfite sequencing to an average depth of 1 8x per strand (Table S4) on two pools 
consisting of 8-I0 individuals per accession. The pooling was performed to reduce variation in methylation rate caused 
by stochastic fluctuations in coverage or read sampling biases. Using the HPG I pseudo reference genome instead of 
the Col-0 reference genome increased the number of cytosines sufficiently covered for statistical analysis by 5% on 
average, and the number of positions called as methylated by 7% (Table S4). Of 24 million cytosines that were covered 
by at least three independent reads, on average 2.5 million were methylated per strain (Table S4). 

For the analysis of differentially methylated positions (DMPs), we considered only cytosines that were covered by at 
least three independent reads in at least seven strains, 2 1. 1 million in total. We found mostly DMPs in the CG context 
(97%). This can be partially explained by our statistical test, which more easily identifies large differences in 
methylation, as is typical for variation at CG sites (see discussion in SOM of ref. [7]). In addition, stable silencing of 
repeats and TEs that are rich in CHG and CHH sites may lead to a similar patterns. To assess epiallele frequencies, 
we compared methylation in each of the 1 2 accessions from near Lake Michigan to the geographic outlier LISET-036 
from Long Island, which we selected as a reference strain. Sixty-one percent of CG-DMPs were recurrent in at least 
two independent Lake Michigan accessions (Figure S6a), which is almost double of what we had previously observed 
for ten equidistant greenhouse-grown MA lines [7]. This can be partly explained by the fact that we sequenced four 
pairs of strains from the same location. Forty-five percent of all CG-DMPs with a frequency of 2 were attributable to 
such pairs, while 6% would be expected if they were randomly distributed across strains. 

Estimating DMP accumulation rates 

We had previously sequenced the genomes of only five of the 1 2 MA lines for which we had reported DMPs [7] ' [8]. 
We therefore generated additional genome sequence data for all ten MA lines in generation 3 I, counting from the 
founder plant of the population, as well as from the two lines in generation 3 [7]. To increase the number of data 
points in the low range of genetic differences, we inferred the number of SNPs between siblings (which had been 
included in the previous MA methylome analyses [7]) from the mutation rates determined by Ossowski and colleagues 
on the same lines. The greater variance in DMP rates and the more rapid initial increase in DMPs in the MA lines in 
Figure 2b is presumed to be due to the methylome data having come from individual plants, instead of from pools of 
individuals as for the HPG I lines. By pooling strains, low frequency epimutations are diluted and less likely to be 
detected. This assumption is corroborated by the fact that we see only 46 DMPs between replicates of HPG I pools 
compared to about 1,300 in replicates of the MA lines. To further investigate this, we compared the number of DMPs 
after in silico pooling of individuals. We first combined data from two siblings of MA line 30-39 or 30-49 in generation 
31 with two siblings of their generation 32 offspring. We then calculated DMPs in comparison of pooled data from 
two times two individuals from two different generation 3 I MA lines. We compared the results with those from 
individual comparisons of all 16 pairwise comparisons between the four early- and four late-generation individuals. The 
number of DMPs distinguishing pools was at the lower end of the DMP distribution from the individual comparisons 
(Figure S7). Hence, it is likely that we underestimate the epimutation rate of the HPG I accessions (Figure 2b). 

S3 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Moreover, we assumed the same genetic mutation rate in the two populations. A potentially faster genetic mutation 
rate in the wild would result in a steeper slope of the HPGI curve, if plotted against number of generations. Finally, 
the initial increase of the HPGI epimutation rate is based on only few comparisons between strains from the same 
sampling site, which might not be sufficient for an accurate estimate. 

Methylated regions 

The value of an approach that defines methylated regions (MRs) before identifying differentially methylated regions 
(DMRs) has been demonstrated before with a Hidden Markov Model (HMM) method developed for the analysis of 
methylated-DNA-immunoprecipitation followed by array hybridization (MeDIP-chip) [9]. We therefore implemented a 
two-state HMM that allows MR identification based solely on within-genome variation in methylation rate. The model, 
based on one for animal DNA methylation data [10], learns methylation rate distributions for both an unmethylated 
and a methylated state for each sequence context separately (CG, CHG and CHH) while simultaneously estimating 
transition probabilities between the two states from genome-wide data. On the trained model, the most probable 
path of the HMM along the genome is then used to define regions of high and low methylation. Applied to the HPG I 
population, the HMM identified on average 32,530 (SD = 1,629) MRs in each strain. The unified MRs had a median 
length of 122 bp (mean = 649 bp), with a maximum length of 87.7 kb (Figure S8a; Table S5). 

For validation we compared data generated from Col-0 (see below) to data from methylated-DNA 
immunoprecipitation followed by sequencing (MeDIP-seq; Vincent Colot and co-workers, pers. communication). Of 
the genome space enriched in MeDIP-seq, 89% was classified as MR by our HMM approach. 

To also evaluate whether the identified MRs sufficiently capture methylated sites within gene bodies consisting almost 
exclusively of CG sites, we tested how many MRs fall into gene body methylated genes as defined previously [II]. We 
re-implemented their method and called between 4,330 and 4,626 gene body methylated (BM) genes and between 
14,998 and 15,489 unmethylated (UM) genes for the HPGI strains. These figures are similar to the 4,361 BM and 
15,753 UM genes reported in ref. [II] for Col-0. A quarter of the BM genes identified in this study and in ref. [II] 
did not overlap, which may be due to genetic differences and/or different sequencing depths and analysis pipelines. 
MRs in this work overlap with 58% of the HPG I BM genes. This compares with an overlap of MeDIP domains of Col- 
0 (Colot lab, see above) of only 42% with Col-0 BM genes. Moreover, the concepts of our approach and that in ref. 
[II] differ considerably: by modeling the density of methylated sites within a gene with a binomial distribution allowing 
only little variance, genes with slightly increased density of methylated sites compared to the global average are quickly 
classified as BM in the method of ref. [II]. Such sites can still be located far apart from each other. In contrast, MRs 
are called by our HMM-based method only when there is a locally restricted, dense region of methylated sites. BM 
genes that are covered by MRs have a higher density of methylated sites compared to BM genes without overlapping 
MRs (Figure S8d). 

Differentially methylated regions 

To identify DMRs, we performed pairwise comparisons of overlapping MRs or parts thereof that were classified either 
(i) as in high methylation state in both accessions of a pair, or (ii) as high methylation state fragments in one and low 
methylation state in the other accession. For each DMR we then assigned all accessions to groups, based on their 
significant methylation differences (Figure S9). We expected to find fewer DMRs in regions that had a high methylation 
state in both tested accessions. In agreement, only 0.4% of those tested fragments (31,531 out of 7,355,716) were 

S4 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



significantly differentially methylated. In contrast, we identified as differentially methylated 4.4% (107,988 out of 
2,450,278) of fragments where the tested accessions had been assigned to contrasting methylation states. 
Consolidation of all pairwise comparisons, resolving overlapping DMRs and differentially testing after grouping 
resulted in a final set of 4,821 DMRs with an average length of 159 bp (Figure S8a; Figure S9; Table S6). 

Our sensitive statistical test classified as differential some regions with low variance and only subtle methylation 
difference; we therefore defined as highly differentially methylated regions (hDMRs) with potentially greater biological 
relevance all DMRs that were longer than 50 bp and that showed a more-than-three-fold difference in methylation 
rate in at least one sequence context, when considering at least five cytosines of that context (Figure S9). In addition, 
the overall methylation rate of the DMR in the more highly methylated strain had to be greater than 20%. Of 3,909 
size-filtered DMRs, 3,199 (80%) were classified as hDMRs (Table S7). The grouping of hDMRs yielded similar epiallele 
frequencies as for the DMPs (54% with frequency larger than I; Figure 2f). The independent analysis of cytosines from 
different contexts allowed us to call context-specific (h)DMRs. In 71% of DMRs and 76% of hDMRs, only cytosines in 
the CG context significantly differed in their methylation rate. Only a minority, 15% of DMRs and 7% of hDMRs, 
showed highly variable methylation in more than one cytosine context (Figure 2g). 

In contrast to previously used methods for the analysis of whole-genome bisulfite sequencing data from plants, our 
HMM for the detection of MRs does not require information about methylation differences at the single-site level. By 
first identifying blocks of methylation, our approach limits the number of tested regions to the methylated space of the 
genome, thereby reducing the multiple testing problem and the requirement for arbitrary filters. Importantly, limiting 
DMR detection to HMM-identified MRs revealed that the location of DMRs in the genome follows the overall 
distribution of methylation. Most of these DMRs thus overlap with TEs and intergenic regions, which is in contrast to 
previously published DMRs relying on user-defined criteria including arbitrary sliding windows or DMPs [7,12-21]. The 
HMM analysis also revealed that non-CG methylation is almost exclusively organized in regions of contiguous DNA 
methylation. We suggest such an approach to identifying MRs be applied to bisulfite sequencing data in future studies. 

Analysis of LISET-036 specific hDMRs 

Strain LISET-036 was the most different when strains were clustered by CHG-DMPs, CHH-DMPs and DMRs. Since 
CHG and CHH-DMPs constituted only a minor fraction of all DMPs (~3%), we focused on hDMRs private to the 
HPGI strains to investigate the possible basis of LISET-036 being an outlier. While LISET-036 had the most private 
hDMRs among all accessions (Figure SI 8a), their spectrum in terms of context and overlap with genomic features did 
not deviate from that of the other strains (Figure SI 8b). 44 LISET-036 private hDMRs overlapped with genes and 30 
overlapped with the gene-adjacent regions, defined as 1,000 bp upstream or downstream of genes. The only GO term 
for which these 74 genes were enriched was "intrinsic to membrane" (p-value 0.01). However, there were no 
overlapping differentially expressed genes. Taken together, there is little evidence for a pronounced phenotypic effect 
of the LISET-036-specific epivariants. 

Differential gene expression and epigenetic variation 

We performed RNA-seq (Table S8) on rosette leaves of all 13 strains and identified 251 differentially expressed (DE) 
genes across all possible pairwise comparisons (Table S9). A majority of these genes were identified as DE in more 
than one comparison. Gene Ontology (GO) analysis identified several defense-related GO terms to be over- 
represented (p « 0.001); which may be linked to these genes evolving fast [22]. As could be expected from the small 

S5 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



numbers of DE genes, clustering of accessions based on expression of all genes revealed no structure reflecting 
genetic distance or geographical origin. When we limited the clustering to DE genes, however, accessions originating 
from the same geographical location, except Yng-4 and Yng-53, clustered together (Figure SI la). A similar observation 
could be made when counting the number of DE genes per comparison: while accessions from the same site generally 
showed no or only few DE genes, comparison of accessions from different sites revealed up to 1 49 differences. The 
two Yng strains accounted for most of the DE genes identified in pairwise comparisons (Figure SI lb). Although we 
identified some loci where changes in contiguous stretches of DNA methylation correlated with alterations in 
transcriptional activity (Table SIO), we did not observe a general relationship between these two features, suggesting 
either that transcriptional differences in the HPGI population is mostly due to DNA mutations, or due to epigenetic 
changes independent of DNA methylation. 

Similarity of epigenetic variation profiles in independent populations 

We used the data we had previously generated on the greenhouse-grown MA lines [7] to investigate whether the MA 
epigenetic variants were similar to the HPG I variants. Almost half of all positions that had been classified as DMPs in 
the MA lines and that were sufficiently covered in the HPGI accessions were also polymorphic in the HPGI 
accessions (41%; p«l x I0" 5 , Fisher's exact test). Similarly, about a third of HPGI DMPs were also MA DMPs (29%; 
p«l x I0" 5 ) (Figure 3a). 

We recalled DMRs in the MA strains [7] using the methods implemented for the HPGI population. We detected on 
average 22,868 MRs (SD = 1,282) per MA line, covering 22.3 Mb. Of 3,837 DMRs in the MA lines, 2,523 coincided 
with MRs detected in the HPGI population. DMRs in the HPGI population were 4-fold more likely to coincide with 
DMRs in the MA lines than with a random MR from this set (Figure 3b). 

We also identified DMRs between one of the MA lines (30-39) and one of the HPGI lines (MuskSP-68). These DMRs 
were also enriched in each of the two sets of within-population DMRs (MA or HPG I) (Figure 3b). 

Finally, we performed a similar analysis with DMRs from 140 natural accessions that represent the entire range of the 
species [14]. While the overlap between those 53,752 DMRs and the HPGI DMRs was greater than expected by 
chance (Z-score = 19.8; 100,000 permutations), most of the DMRs in the 140 accessions (9,994) do not even overlap 
with MRs in the HPGI or MA population (Figure 3d). While this is almost certainly due at least in part to the very 
different DMR detection methods, it could also indicate that differential methylation in the global population was 
influenced by genetic variation to a larger degree than in the HPG I accessions. 

Supplementary References 

1. Bomblies K, Yant L, Laitinen R, Kim S-T, Hollister JD, et al. (2010) Local-scale patterns of genetic 

variability, outcrossing and spatial structure in natural stands of Arabidopsis thaliana. PLoS Genet 6: 
el 000890. 

2. Piatt A, Horton M, Huang YS, Li Y, Anastasio AE, et al. (2010) The scale of population structure in 

Arabidopsis thaliana. PLoS Genet 6: el 000843. 

3. O'Kane SL, Al-Shehbaz IA (1997) A synopsis of Arabidopsis (Brassicaceae). Novon 7: 323-327. 

4. Gan X, Stegle O, Behr J, Steffen JG, Drewe P, et al. (2011) Multiple reference genomes and 

transcriptomes for Arabidopsis thaliana. Nature 477: 419-423. 



S6 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



5. Hu TT, Pattyn P, Bakker EG, Cao J, Cheng JF, et al. (20 1 I) The Arabidopsis lyrata genome sequence and 

the basis of rapid genome size change. Nat Genet 43: 476-48 1 . 

6. Cao J, Schneeberger K, Ossowski S, Gunther T, Bender S, et al. (20 1 I) Whole-genome sequencing of 

multiple Arabidopsis thaliana populations. Nat Genet 43: 956-963. 

7. Becker C, Hagmann J, Miiller J, Koenig D, Stegle O, et al. (20 1 I) Spontaneous epigenetic variation in the 

Arabidopsis thaliana methylome. Nature 480: 245-249. 

8. Ossowski S, Schneeberger K, Lucas-Lledo Jl, Warthmann N, Clark RM, et al. (20 1 0) The rate and 

molecular spectrum of spontaneous mutations in Arabidopsis thaliana. Science 327: 92-94. 

9. Seifert M, Cortijo S, Colome-Tatche M, Johannes F, Roudier F, et al. (20 1 2) MeDIP-HMM: genome-wide 

identification of distinct DNA methylation states from high-density tiling arrays. Bioinformatics 28: 
2930-2939. 

10. Molaro A, Hodges E, Fang F, Song Q, McCombie WR, et al. (20 1 I) Sperm methylation profiles reveal 

features of epigenetic inheritance and evolution in primates. Cell 1 46: 1 029- 1 04 1. 

I I. Takuno S, Gaut BS (20 1 2) Body-methylated genes in Arabidopsis thaliana are functionally important and 
evolve slowly. Mol Biol Evol 29: 2 1 9-227. 

1 2. Schmitz RJ, Schultz MD, Lewsey MG, O'Malley RC, Urich MA, et al. (20 1 I) Transgenerational epigenetic 

instability is a source of novel methylation variants. Science 334: 369-373. 

1 3. Dowen RH, Pelizzola M, Schmitz RJ, Lister R, Dowen JM, et al. (20 1 2) Widespread dynamic DNA 

methylation in response to biotic stress. Proc Natl Acad Sci USA 1 09: E2 1 83-2 1 9 1 . 

14. Schmitz RJ, Schultz MD, Urich MA, NeryJR, Pelizzola M, et al. (20I3) Patterns of population epigenomic 

diversity. Nature 495: 1 93- 1 98. 

1 5. Schmitz RJ, He Y, Valdes-Lopez O, Khan SM, Joshi T, et al. (20 1 3) Epigenome-wide inheritance of 

cytosine methylation variants in a recombinant inbred population. Genome Res. 

1 6. Qian W, Miki D, Zhang H, Liu Y, Zhang X, et al. (20 1 2) A histone acetyltransferase regulates active 

DNA demethylation in Arabidopsis. Science 336: 1 445- 1 448. 

1 7. Calarco JP, Borges F, Donoghue MT, Van Ex F, Jullien PE, et al. (20 1 2) Reprogramming of DNA 

methylation in pollen guides epigenetic inheritance via small RNA. Cell I5I: 1 94-205. 

1 8. Ausin I, Greenberg MV, Simanshu DK, Hale CJ, Vashisht AA, et al. (20 1 2) INVOLVED IN DE NOVO 2- 

containing complex involved in RNA-directed DNA methylation in Arabidopsis. Proc Natl Acad Sci 
USA 109:8374-8381. 

19. Stroud H, Greenberg MV, Feng S, Bernatavichute YV, Jacobsen SE (2013) Comprehensive analysis of 

silencing mutants reveals complex regulation of the Arabidopsis methylome. Cell 152: 352-364. 

20. Regulski M, Lu Z, Kendall J, Donoghue MT, Reinders J, et al. (2013) The maize methylome influences 

mRNA splice sites and reveals widespread paramutation-like switches guided by small RNA. 
Genome Res 23: 1 65 1 - 1 662. 

21. Stroud H, Ding B, Simon SA, Feng S, Bellizzi M, et al. (2013) Plants regenerated from tissue culture 

contain stable epigenome changes in rice. Elife 2: e00354. 

22. Jones JD, Dangl JL (2006) The plant immune system. Nature 444: 323-329. 



S7 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Supplementary figure legends 

Figure SI: Iterative re-alignment strategy and statistics. 

(a) Iterative re-alignment approach to evaluate predicted variants and to build a HPGI pseudo-reference genome. (I) 
For each strain, variants called by diverse structural variant detection tools and a local de novo assembly pipeline were 
combined into a consolidated variant set. (2) Variants with core read coverage were filtered out, the remaining 
variants were classified into common and potentially segregating. Brown triangles symbolize insertions/deletions, the 
brown X represents a SNP. (3) All common variants were incorporated into the reference genome. All segregating 
variants were introduced in branches of the reference genome, which incorporated polymorphisms linked by less than 
1 07 bp. (4) After mapping the reads against the branched reference, a binomial test was performed in each strain and 
for each variable site to call the allele, i.e., to determine whether there was statistical evidence for the presence of 
only one haplotype covering the variant's coordinates. Variants with the same non-reference allele call in all strains 
were considered as "common"; those with a reference call in at least one strain and a variant call in at least one other 
strain were classified as "segregating". (5) All common variants from the previous step were incorporated into the 
new reference sequence, and a new iteration was started over from (I), or this new genome served as the HPGI 
pseudo-reference genome after iteration 2. (b) Increase of detected variants and decrease of unsequenced genome 
space and unmappable reads by iterative read mapping. The legend on the right side denotes absolute values after 
iteration 2. The reference value (100%) derives from the mapping against the Columbia-0 genome sequence (TAIR9), 
and for common variants it is the number of variants leading to the genome of iteration I. Thus, -842,000 common 
variants led to the genome of iteration 2, -864,000 to the genome of iteration 3. (c) Composition of common 
polymorphisms by variant type (top) and by detection tool (bottom). Variants found by more than one tool 
contributed to the count for all respective tools. 

Figure S2: Distribution of genetic variants along the five chromosomes. 

Relative density of common variants in 100 kb sliding windows with a step size of 10 kb. 

Figure S3: Annotation of genetic variants. 

Polymorphisms were hierarchically assigned to CDS > intron > 5' UTR > 3' UTR > transposon > intergenic. 

Figure S4: Magnification of the central area of the phylogenetic network in Figure Ic. 

Numbers indicate bootstrap confidence values (10,000 iterations). 

Figure S5: Phenotypic analysis. 

(a) Leaf growth measured over time (Materials and Methods). Error bars represent 95% confidence intervals. On 
average, 36 plants were measured per accession, (b) Correlation of genetic distance, represented by number of SNPs 
per pairwise comparison, and difference in leaf area at 21 days after germination. 

Figure S6: DMPs. 

(a) Epiallele frequencies of DMPs for CG sites only (left), and comparison of all three sequence contexts (right), (b) 
Annotation of 5m Cs and DMPs. Cytosines were hierarchically assigned to CDS > intron > 5' UTR > 3' UTR > 
transposon > intergenic. 

S8 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure S7: Effect of sample pooling on the number of identified DMPs. 

Small filled triangles and box-and-whisker plot indicate distribution of DMPs that were called by comparing individual 
plants of generations 3 I and 32 of lines MA30-39 and MA30-49 with individual plants of MAO-4-26 and MAO-8-87, 
which represent generation 3 of two independent lines (1 6 comparisons in each group). The large unfilled triangles 
indicate the number of DMPs that were called when data from four lines of generations 3 1 and 32 were pooled in 
silico and compared against all four lines from generation 3. On average, a substantially lower number of DMPs is 
called with pooled data. 

Figure S8: Characteristics of (differentially) methylated regions. 

(a) Histogram of the length of unified MRs, DMRs and hDMRs. The red and green lines indicate mean and median 
length, respectively. For better visibility, regions larger than 2,000 bp were excluded from the representation. The 
minimum size of hDMRs was 50 bp. (b)-(d) are based on data of one HPGI strain (LISET-036). (b) Number of 
unmethylated cytosines in-between methylated CG sites within genes in dependence of whether these sequences are 
inside or outside of MRs. (c) Distances in bp between methylated CG sites within genes in dependence of whether 
these sequences are inside and outside MRs (minimal distance 2 bp), (d) Distances between methylated CG sites 
within body methylated (BM) genes identified with the method from ref. [II] (SOM: Methylated regions) and within 
genes not identified as BM (minimal distance 2 bp), (e) Length distributions of DMRs that overlap and that do not 
overlap coding regions. Triangles show mean values. 

Figure S9: Selecting subregions of HMM -detected MRs to test for differential methylation. 

(a) Example illustrating the selection procedure of regions and pairwise strain comparisons to be tested for differential 
methylation. (*) For simplicity, the illustration uses a minimum number of covered sites of two reads per region (10 
reads for the real data set). (I) We selected all possible regions where two strains presented different states of 
methylation (regl to reg5) and applied filter criteria (a), (b) and (c). (2) If a region passed filters (a), (b) and (c) (in the 
example only regl and regl), criteria (a), (c) and (d) were checked for each pairwise comparison between strains on 
that region. Note, the selection of a region in (I) must not necessarily lead to a differential test between any two 
strains (e.g., regl). Refer to the Material and Methods section for elaborate descriptions of criteria, (b) Assignment of 
strains to different groups based on differential methylation. Left: An exemplary DMR represented as a graph: strains 
are represented as nodes, edges reflect a statistically significant test between two strains. Right: Finding the minimal 
number of sets, where no edge connects nodes from two different groups is known as the colouring vertex problem. 
In this example, the solution is two sets of strains (blue and white nodes). Strains without statistically significant tests 
(e.g., strain L) are grouped into the set of strains where the difference between the strain's and the group's mean 
methylation rate is minimal. 

Figure SIO: Difference in local methylation rate between lines classified as "low methylation" 
versus lines classified as "high methylation" over a given (h)DMR. 

Histograms of the absolute mean methylation rate differences of DMRs (grey) and hDMRs (black) of different 
sequence contexts. 



S9 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure SI I: Differential gene expression. 

(a) Hierarchical clustering of HPGI accessions by expression of differentially expressed genes, (b) Differentially 
expressed genes per pairwise comparison. 

Figure SI 2: DMRs and gene expression. 

Examples of DMRs (top panel) overlapping with a protein-coding gene (AT4G09360, left), a non-coding RNA 
(AT4G04223, middle) and a transposable element (ATIG62460, right). The expression of the corresponding locus is 
represented in the bottom panel. 

Figure SI 3: Methylated cytosines independently identified as DMPs in HPGI accessions and/or 
MA lines. 

(a) Relative density of DMPs along the 5 chromosomes. For each class and each chromosome, the window with the 
maximal density was set to I. Sliding window; window size 100,000 bp; step size 10,000 bp. (b) Ratios between 
epimutation frequencies and sequencing depth along the 5 chromosomes for MA and HPG I lines. Epimutation 
frequencies were determined as the number of DMPs per cytosine with at least threefold coverage per window. 
Coverage is represented as average coverage per window across all accessions of each population. Dashed lines mark 
the ideal coverage ratio of I. Sliding window; window size 100,000; step size 10,000 bp. (c) Annotation of Cs, N- 
DMPs and DMPs. Cytosines were hierarchically assigned to CDS > intron > 5' UTR > 3' UTR > transposon > 
intergenic. 

Figure SI 4: MA DMPs shared with HPG I DMPs according to MA generational distance. 

We computed DMPs between two randomly chosen MA strains separated by specific numbers of generations and 
plotted the fraction of those DMPs shared with a randomly chosen HPGI strain. Each boxplot summarizes ten such 
random comparisons. 

Figure SI 5: Overlap of MA DMRs and HPGI DMRs per genomic feature and DMR sequence 
context. 

DMRs were hierarchically assigned to CDS > intron > 5' UTR > 3' UTR > transposon > intergenic. CG-DMRs had 
significantly different methylation in the CG context only and C-DMRs in any other (additional) context(s). 

Figure SI 6: Heritability by sequence context of hDMRs in HPGI, and their overlap with 
unmethylated, methylated and differentially methylated regions in MA lines. 

Figure SI 7: Hierarchical clustering by polymorphic and non-polymorphic positions and regions. 

Figure SI 8: Analysis of LI-SET-036 private hDMRs. 

(a) The number of strains sharing the same methylation status for DMRs found in each strain is plotted (determined 
by the strain grouping procedure; see Methods), (b) Stacked bar plots showing the distributions of sequence contexts 
(bottom) and overlapping genomic features (top three plots) for each strain's private hDMRs. 'CG only' exclusively 
considers CG-hDMRs whereas 'CHG' and 'CHH' might also include hDMRs of other contexts than CHG and CHH, 



SIO 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 

respectively. The distribution across intergenic space, TEs and genes was similar for all strains. See section "Analysis of 
LI-SET-036 specific hDMRs" above for more details. 



SI I 



SOM Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Figure SI 

a 

0) 




Assemblers 
Velvet, SOAPdenovo 




SNPs, 


\ Non-sharec 


indels, 


\ variants 


SVs 







Consolidated variant set 



non-shared 
variants 



non-shared 
variants 




(4) 



Variant 1 






variant 2 






Strain 1 


Strain 2 


Strain 3 


Strain 4 




Strain 13 

































variant s 


missing x 


unknownx 


referenced 




variant Y 



0 ref calls, 
>7 var calls 



Common variants 



/>1 ref calls, 

>1 var calls £7 calls in total) 



Segregating variants 



(5) 



New reference tt 





120- 




110 - 




100- 






g 


90 


O 




LL 


80 




70 



Iteration 1 
(Col-0) 



Iteration 2 Iteration 3 




77% 



_ 160 

o 

o 

CO 

*, 120 

c 

0 

g. 80 

o 

LL 

40 - 
0 




— 2,017 

segregating 
118% variants 
112% _ 842,000 
common 
variants 
4.6 Mbp 
0-coverage 
region 
2.1x10 6 non- 
g 7 o/ o mappable 
reads 



Insertion 
in HPG1 

Deletion 
in HPG1 



SNPs Small SVs SVs 
indels (>8bp) (>20bp) 
(<8bp) 



20 




Assembly 

SHORE 

Pindel 

SV-M 

DELLY 



SNPs Small SVs SVs 
indels (>8bp) (>20bp) 
(<8bp) 



SI2 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure S2 




0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 

Chromosome 1 (Mb) 




Chromosome 2 (Mb) 




0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 

Chromosome 3 (Mb) 




0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 

Chromosome 4 (Mb) 




^ i i i i i i i i i i i i i i i i i i i i i i i i i i i i 

> 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 

Chromosome 5 (Mb) 



SI3 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure S3 




SNPs SVs 



SI4 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure S4 




SI5 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 




SI6 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure S6 




SI7 



SOM Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Figure S7 



40,000 - 




n r 

30-39 30-49 
MA line in generation 31/32 



SI8 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 





DMRs 



— I — 

500 



1000 
Region length (bp) 




hDMRs 



1000 
Region length (bp) 



Inside Outside 
MRs MRs 

(50,742) (21 5,468) 



# 200 
O 



100 



T 



Inside Outside 
MRs MRs 

(50,742) (215,468) 



O 200- 
O 



J 150- 



T 



Inside Outside 
MRs MRs 

(16,035) (117,438) 



Q 200 



Inside Outside 
CDS CDS 

(2,008) (2,851 ) 



SI9 



SOM Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Figure S9 



Strain 1 



1) Select regions: 
4c-c-c — c — c| — 



2) Select pairwise strain comparisons: 



#covered reads: 

Strain 2 

#covered reads: 

Strain 3 

#covered reads: 

Methylated region (MR) 

Regions 
for 
differential 
testing 





i 2 0 




jC-C 




3 3 


( m 










5 6 




regl 





4c-c-c — c — cf - 

2 0 1 6 3 1 

fC-C fC C— Cf - ((d)^ 

(d)X) 




A A Strain A with 56% methylation 
56% J rate in focal region 



Statistically significant differential methylation 
between two strains in focal region (i.e. DMR) 



S20 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure SIO 




Absolute methylation rate difference (%) 



CHG-DMRs 

■ l J ^ 

o 15- 

LLh ..iih.Ll.j_J _I_1L 



0 J r 



20 40 60 80 100 

Absolute methylation rate difference (%) 



>> 20-, 

I 15" 
= 10- 
a> 5- 

o-L 



40 60 
Absolute methylation rate difference (%) 



S2I 



SOM Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Figure S 1 1 

a 



Log10(FPKM+1) 



■ 



<t ^ CC in CO (D D) i — LO CD i — cd r**- 
U_U_ m M « Q_ ^ CO CO 



BRR4 
BRR57 
KBSMac68 
KBSMac74 
LISET036 
MNFChe47 
* MNFChe49 
MSGA10 
MuskSP68 
Pawl 3 
Yng4 
Yng53 



to 

















23 






































































31 


































1 
































23 


14 


-| -| 




























































14 


36 


20 


1 4 






















































3 


24 


57 


41 


1 4 




































52 


28 


16 


11 


8 


21 




















4 


56 


40 




17 




20 




















18 


18 
















7 




28 






4 


13 


12 














8 


1 


9 


19 












16 


36 


34 






8 


12 


40 


33 


4 












28 


51 


47 










17 


51 


36 


5 
















8 






2 




43 


84 


71 






17 


35 


79 


53 


22 






1 




1 


4 


3 


2 


94 


149 


119 


4 


34 


71 


127 


80 


67 



CO 00 o 

t- CD t- 

5 CL < 

CO CO O 

°- % w 



CD LO 

" cr 
■S cr 



100 

50 

0 



Sample 1 



S22 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 

Figure SI2 



Non-methylated C, Watson strand ■ 5 m CG, Watson strand 





■ Nor 


-methylated C, Crick strand 


:: " - 















5 m CHG, Watson strand 
1 5 m CHG, Crick strand 



5 m CHH, Watson strand 
5 m CHH, Crick strand 



-flw:i,;.qi^;%^.>iiV'jTjii^ 




AT4G04223.1 



AT4G04223.2 



0.5- 



W]D]DODDDcozZ(y)C g ,3 ,3 
£ o o co r " 



10 



B+4 




0.75- 



^0.50- 



0.25- 



0.00- 



AT1G62460 




























/ 



«:D3Jcoco OT zzcne: g = = 

? ^||HOO>| u oo 

00 -v| CD 00 



(33 

r\o 



Accession 



Accession 



Accession 



S23 



SOM Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Figure SI 3 



a 

sz 

CD 
T3 

So.o: 



Shared 

Non-differential 



Private MA 
Private HPG1 



0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 

Chromosome 1 (Mb) 



■f 1-0- 



1 0.0 



■£i-o 



0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 

Chromosome 2 (Mb) 



0 1 2 3 4 5 6 7 



9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 
Chromosome 3 (Mb) 



^^^^^^^^^^^^^^^^ 

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 

Chromosome 4 (Mb) 



100 n 



0 J 



CDS 
. 5'UTR 
i Intergenic 
i Intron 
i 3'UTR 
. TE 



JZ<CLCLCLCLO-CLCL 



0123456789 



10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 
Chromosome 5 (Mb) 



CD 
Q_ 
X 

w 

b 
< 



Q Q Q Q 



< 9 
2 ± 



Q 

5 



CD 

£ 1.0 

< 

o 0.6 



CD 

3=1.2. 

2 0.8 J 
o 0.6J 



CD 

£ 0.9 
< 

o 0.5 



CD 

^0.8 
^0.7 
o 



CD 
Q_ 

10.8 
< 0.7 




Epimutation frequency MA/HPG1 
Coverage MA/HPG1 



i 1 1 1 1 1 1 r~ 

0 1 2 3 4 5 6 7 



9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 



Chromosome 1 (Mb) 



0 1 2 3 4 5 6 7 




10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
Chromosome 2 (Mb) 




0 1 2 3 4 5 6 7 



9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
Chromosome 3 (Mb) 



0 1 2 3 4 5 6 7 



10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
Chromosome 4 (Mb) 




0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 



2.0 
1.6 
1.2 
0.8 



2.0 
1.6 
1.2 
0.8 



2.0 
1.6 
1.2 
0.8 



2.0 
1.6 
1.2 
0.8 



Chromosome 5 (Mb) 



S24 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure SI4 



0.20 




Number of generations separating two MA lines 



S25 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure SI5 




S26 



SOM Hagmann et al. 



Downloaded from http://biorxiv.org/ on September 18, 2014 

Stability of the Arabidopsis thaliana methylome in nature 



Figure SI6 



Heritability of HPG1 CG-hDMRs 



1.0 
0.8 
0.6 
0.4 
0.2 
0.0 



All Unmethylated in MA Methylated in MA DMR in MA 

(2,289) (1,074) (939) (276) 



Heritability of HPG1 CHG-hDMRs 




All Unmethylated in MA Methylated in MA DMR in MA 

(472) (177) (210) (84) 



Heritability of HPG1 CHH-hDMRs 




All Unmethylated in MA Methylated in MA DMR in MA 

(215) (72) (105) (38) 



S27 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 



Figure SI7 



CG_N-DMPs 



CHG_DMPs 



CHG_N-DMPs 





Q.CO 



I- CO CO'tf t- CO m i- 

- - w < a. co< o o u 

U rn ^ rn 2 2 

u-^ 2> ^coco 2 
z ^ .2 ^ do m ^ 
^ 2 ^ ^ «= 



N-DMRs 



^'toococoo)0)^ 



1 LL LL 
■JZZ 
22 



COCO^-^-r-~-N-C0C0^-^C0C0C\JC\J 

cococcccwWi-t- co com m co co 
oo^fccir S S c c o) coo o - - 

WW mCU Q. Q. CO CO CO CO ^ ^ "J CO LL LL LL Li. 



C£>COC\J^^ncn^OC003C»r^-a>Tl-COTl-CX3h-0^1-CON-I^C\JCD 

co cd cd cc m 1- 0)1- co cd m o>i- rrcoinT-N.in^i-^i-coc: 
r-i on oit oig c<d_ a> 00c a> c grrtL(r< <-> co a> a>oo 

2 2 ^ 55 

CHH_N-DMPs 



CHH_DMPs 



wizi 

CO CO Q_ CO 




N(MWWWt-CMwV^(Mi-t-^^^^CMt-IMCMNW 



o o g g c c c 



^i-i-cocot^^^u 



^■^■cocooococooiroh 



cofcococo^i-cococNjor- 



? ?9 9 ^ g cr tr tr 

"^3325 



^COCO-^-^COCOTl-Tl-Tl-Tl- CO CD CO in CO OT-i- CO CO 1- 



^^2222 

CG-DMPs in genes 



I- I- > > t= c < < vu „ 
LULU >>ZZ0-D- 
COCO CLD. 



< < Q_ CL a 
OOC0C0£ 
CO CO CO CO 3 ^ 2> COLLLLLLLL 

rjooorjorjQ^'^.g.gzzzz 



Dcoincoin^i-i-1- 



^2222 



0 CO 2^ 0. 2> Q_ LL ^CO LLLLCOU-CO 

j DQ ^ 2 3 z ^ m ZZjZffl 

* ^ 5 2 * 22 2* 



CG-DMPs in TEs 




Methylation rate (%) 



COCOCOCO'^-Tl-COCO^l-'^-OOCMOJ 

■1- -i- m m cd coco co r~- n- t- t- cd cd 
sscdcdccoooo<<oo 

CLCL>> 2222^^ZZOOOO 

C0C0C0C0^^£LQ.LLLLLLLLC0C0 2 
CD CQ CQ CQ ^ ^ Z Z Z Z □ □ .= 

**** 2222 

CGJiDMRs 



d i- i- co coir: 



^22^^9255ccroo) a - [[ a:(r<<ooD.D. 

llii5666il££ >> -^^SgLTrxoofc^c2 



JZZ 

CO CO CO CO LL LL LL LL q. D. 
CQCQCQCQZZZZ 

****2222 



CHG_hDMRs 



CHHJiDMRs 



U^lfoU^ toiW^Mi?! 1 ! iffiSiffiSHfi 




cocoinina:o:inir 



coco-^t-co com 



°>> rxo:LULU2222^^^^0000 LULUrxoC 

CO CO CO CO CO CO 3 2> COLLLLLLLL COCO 
jjCD[D[D[I1 22 HZZZZ 

^^^^ 22^^22 



oococo^ , . 

Ofrir-^-^CDCO^-I^CDCDTl-Tl-Tl-Tl- 

Lrra:<< " " " "a " 

^ZZrDm m(n2222i 

11 53COCOCOCO 2^ ^jU-lLlLlL 

^■^DQmmm^^zzzz 
^^^^22^222 



inmrrrrcDcoi-i-inincoco co co-^- ■^cocococon-i^ti-ti-ti-ti- 
DCDCrrrrOO 5 5 co coo occ<<ooq_dloocdcdcdcd 
rr rr £ & < < S S ch I— >- >- rn 10 10 en to to orr r r 
gg CDCD zzrxL?>>LULU^^^^22^^220000 

0000 CLD. COCO Sg^COCO 2? 2? CO CO LL LL LL LL 

□ □ codq55cqlxizzzz 
^^22^^2222 



S28 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 




S29 



Downloaded from http://biorxiv.org/ on September 18, 2014 
SOM Hagmann et al. Stability of the Arabidopsis thaliana methylome in nature 

Supplementary tables 

Table SI : Haplogroup- I (HPG I) accessions used in this study. 
Table S2 : Common SNPs and SVs. 
Table S3 : Segregating SNPs and SVs. 

Table S4 : Summary statistics on methylome sequencing. 

Table S5 : Methylated regions (MRs). 

Table S6 : Differentially methylated regions (DMRs). 

Table S7 : Highly differentially methylated regions (hDMRs). 

Table S8 : Summary statistics on transcriptome sequencing. 

Table S9 : Differentially expressed (DE) genes identified in pairwise comparisons between 
HPG I accessions. All DE genes (q-value < 0.05) between any two accessions are listed. If more than one gene 
name appears in column I, the read counts could not be assigned to one gene in particular and/or a fused transcript 
was suggested by the read data. 

Table SI0 : Statistics of overlapping DE genes and hDMRs. 

Table SI I : Overlap of SVs with MRs in HPG I and Col-0 in different genomic features. 

Table SI2 : Scoring matrices for SNP calling and assessing cytosine site statistics (for bisulfite 
sequencing). 



S30