Over half of primate genomes are annotated as transposable elements (TEs) (Jurka, 2000; de Koning et al., 2011). Primate TE sequences reflect their evolutionary history; some TEs are conserved over multiple phylogenetic clades and orders, whereas others are restricted to particular lineages. Considering the degree of sequence similarity between the human and chimpanzee genomes, it is not surprising that many of the same TEs are present in both species (Chimpanzee Sequencing and Analysis Consortium, 2005; Ramsay et al., 2017). In general, TE activity, especially from endogenous retroviruses (ERVs), has declined in the hominid lineage relative to other mammalian lineages (Lander et al., 2001). However, a fraction of evolutionarily recent TEs are active in the human genome, including HERV-K ERVs (Tönjes et al., 1996; Medstrand and Mager, 1998; Fuchs et al., 2013; Klawitter et al., 2016; Wildschutte et al., 2016), members of the Alu (Batzer and Deininger, 1991; Batzer et al., 1991), L1 (Kazazian et al., 1988, Brouha et al., 2003), and SVA (Wang et al., 2005; Ostertag et al., 2003) families. Most studies on mammalian TEs have focused on humans and mice, with a handful that describe the ERV and L1 TE families in chimpanzees (Yohn et al., 2005; Mun et al., 2014; Marchetto et al., 2013).

Because of their transposition competence, TEs are considered a potential source of regulatory innovation (Kazazian, 2004; Cordaux and Batzer, 2009). Indeed, TEs can serve as primate-specific regulatory sequence (Jacques et al., 2013; Trizzino et al., 2017), act as enhancer elements (Lynch et al., 2011; Chuong et al., 2013; Xie et al., 2013), carry transcription factor binding sites (Wang et al., 2007; Bourque et al., 2008; Kunarso et al., 2010; Schmidt et al., 2012), and contribute themselves to the transcriptome (Faulkner et al., 2009; Kelley and Rinn, 2012; Kapusta et al., 2013). The influence of TEs on transcriptional programs has been studied in many broad contexts from the immune system (Chuong et al., 2016) to pregnancy (Lynch et al., 2015), to pluripotency (Macfarlan et al., 2012). Yet, ultimately, the regulatory impact of TEs has not been extensively studied at a genome-wide scale in a comparative framework.

Earlier work using a Down syndrome mouse model carrying a copy of human chromosome 21 showed that in general, genes on human chromosome 21 are similarly regulated regardless of whether the chromosome is in a human or a mouse cellular environment (Wilson et al., 2008). However, human-specific TEs become aberrantly activated (marked by H3K4me3) on human chromosome 21 in the mouse cellular environment (Ward et al., 2013). It was therefore speculated that host repressive factors needed to properly silence these human TEs may be absent in the mouse nuclear environment, which is ~60 million years diverged from human. This suggested a relatively rapid evolution of TE silencing mechanisms, in line with the evolutionary arms race hypothesis.

Indeed, due to the potential deleterious effects of new TE insertions, especially in the context of developmental regulatory programs, TE activity is thought to be tightly controlled by host repressive machinery. In particular, it has been shown that in embryonic stem cells (ESCs), KRAB-containing Zinc Finger genes (KRAB-ZNFs) can recognize TEs in a sequence-specific manner, recruit the co-factor TRIM28 (also known as KAP1), and histone methyltransferase SETDB1 (also known as ESET), to effect silencing through the deposition of the repressive histone three lysine nine trimethylation (H3K9me3) modification (Matsui et al., 2010; Rowe et al., 2010; Najafabadi et al., 2015; Turelli et al., 2014; Friedli et al., 2014; Wolf and Goff, 2009; Schultz et al., 2002). After the establishment of TE silencing by H3K9me3, TEs are subjected to de novo DNA methylation, which persists in somatic and germ cells (Rowe and Trono, 2011; Quenneville et al., 2012; Rowe et al., 2013; Smith et al., 2014; Karimi et al., 2011; Walter et al., 2016; Bourc'his and Bestor, 2004; Barau et al., 2016). Some TEs can evade detection by host KRAB-ZNFs by acquiring mutations. It was thus hypothesized that the host repressor mechanisms need to evolve rapidly in order to maintain genome integrity (Kapopoulou et al., 2016; Huntley et al., 2006; Thomas and Schneider, 2011).

TEs are silenced by KRAB-ZNFs particularly in early embryonic development (Chuong et al., 2017), their most likely niche for retrotransposition (Richardson et al., 2017); however, it is difficult to obtain biological material from humans and other primates to determine how this process evolves. Induced pluripotent stem cells (iPSCs) provide a model of the cells of the embryonic inner cell mass to sidestep this challenge. This cell type is of further interest as we know of examples of primate-specific TEs that are not silenced, but actively transcribed and functionally relevant in this cell type in human (notably HERVH elements) (Klawitter et al., 2016; Santoni et al., 2012; Fort et al., 2014; Lu et al., 2014; Wang et al., 2014; Grow et al., 2015; Wissing et al., 2012). This might imply specificity of TE regulatory mechanisms, and hints at the fact that evolutionarily recent TE sequence could play a role in differentiating gene regulatory networks between species.

The establishment of a panel of non-retroviral reprogrammed iPSCs from multiple humans and chimpanzees has allowed us to comprehensively characterize the extent of TE silencing in both species (Gallego Romero et al., 2015; Banovich et al., 20162018; Burrows et al., 2016). In order to gain insight into how TE regulation, specifically TE silencing, evolves between humans and chimpanzees, we globally profiled the predominant output of KRAB-ZNF-targeted silencing mechanisms in pluripotent cells: the histone modification H3K9me3. Our goals were to identify inter-species differences in TE silencing patterns, characterize gene regulatory divergence that can potentially be explained by differences in TE silencing between species, and develop hypotheses regarding the mechanisms that underlie such inter-species differences in TE regulation. Our observations led us to conclude that, at least in our comparative iPSC system, differences in TE silencing do not drive substantial gene regulatory divergence between these two closely related species.

To perform a comparative study of TE silencing, we characterized genome-wide distributions of the repressive histone modification H3K9me3 in iPSCs from 7 chimpanzees and 10 humans (Figure 1A, Table 1 in Supplementary file 1). The iPSC lines we used were deeply characterized in this study and previously (see Materials and methods; and [Gallego Romero et al., 2015; Banovich et al., 20162018; Burrows et al., 2016]). Briefly, all cell lines were confirmed to have a normal karyotype (Figure 1—figure supplement 1A), have the ability to form embryoid bodies that consist of cells from the three germ layers (Figure 1—figure supplement 1B), display global gene expression signatures consistent with pluripotent cells (Figure 1—figure supplement 2), and express key pluripotency factors detectable by immunocytochemistry (Figure 1—figure supplement 3). All samples were processed in species-balanced batches at each experimental step (Table 2 in Supplementary file 1).

Figure 1 with 14 supplements see all

Download asset Open asset

We characterized the distribution of H3K9me3 in the 17 iPSC lines by performing chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq). In order to increase genome-mapping confidence, paired-end sequencing libraries were prepared, and only high-quality, properly-paired reads retained (see Materials and methods and Figure 1—figure supplement 4 and Table 3 in Supplementary file 1). As a first measure of quality of our data, consistent with previous reports, we observed an enrichment of H3K9me3 ChIP-seq reads at ZNF genes in both species (Figure 1—figure supplement 5)(Roadmap Epigenomics Consortium et al., 2015).

We identified regions of broad H3K9me3 ChIP enrichment in each individual independently (at an FDR of 10%; see Materials and methods and Figure 1—figure supplement 6). A read sub-sampling analysis indicates that we sequenced our samples to a depth that allows us to detect most enriched regions (Figure 1—figure supplement 7). In order to compare H3K9me3 enrichment across species, we focused exclusively on reciprocally best match orthologous regions in human and chimpanzee (see Materials and methods). We excluded from our data peaks that mapped outside those orthologous genomic regions (approximately 10% of the ChIP-seq data were excluded; Figure 1—figure supplement 8 and Table 3 in Supplementary file 1). We further filtered the data based on overall genome mappability, to exclude regions in which sequencing reads can be mapped to one species at a substantially higher probability than the other (using a cutoff of 0.8 in both species; see Materials and methods). We defined orthologous H3K9me3 ChIP-seq regions as those where a ChIP-seq peak, contained within orthologous human-chimpanzee genomic regions, was identified in at least one individual, regardless of species (see Materials and methods). We refer to this combined set as 160,113 orthologous ChIP-seq regions (rather than ChIP-seq peaks), because these regions are defined across all individuals, while the ChIP-seq peak may have been identified in only a subset of individuals.

The definition of ChIP-seq regions allows a quantitative comparison of H3K9me3 enrichment across species. We therefore calculated the number of sequencing reads that were mapped to each orthologous ChIP-seq region in each individual. By using this approach, we sidestep the difficult challenge of accounting for incomplete power to detect ‘significant’ ChIP-seq peaks in each individual or species independently. To avoid considering regions where we clearly have insufficient power to compare read counts across species, we included only the 150,390 regions with at least one read count in more than half of the individuals, regardless of species.

We considered overall properties of the read count data in the ChIP-seq orthologous regions. As expected, data from different individuals cluster by species (with the exception of one human individual, which therefore was removed from subsequent analysis; Figure 1—figure supplements 9–10), and inter-species variation in read counts is somewhat greater than intra-species variation (Spearman’s correlation in human vs. human = 0.90, chimpanzee vs. chimpanzee = 0.89, and human vs. chimpanzee = 0.80, Figure 1—figure supplement 11). We then used the framework of a linear model with a fixed effect for species to identify orthologous ChIP-seq regions with differences in H3K9me3 enrichment between humans and chimpanzees (Figure 1B, Supplementary file 2, and Materials and methods). At FDR < 0.01, we classified 16,288 (11%) regions as differentially enriched between the two species (Figure 1C–D, Figure 1—figure supplement 12 and see examples in Figure 1E). The proportion of differentially enriched regions is robust with respect to our data filtering choices (Figure 1—figure supplement 13 and Table 4 in Supplementary file 1), and the choice of approach to analyze the data (Table 5 in Supplementary file 1).

Orthologous TEs tend to be silenced more often than species-specific TEs

To compare TE silencing by H3K9me3 across species, we identified a comprehensive set of 4,036,865 autosomal orthologous TEs in the human and chimpanzee genomes (80.2% of all annotated human TEs, see Materials and methods and Figure 2A, Figure 2—figure supplement 1). We considered the ChIP-seq data in the context of the orthologous TEs, and observed that 12% of orthologous TEs (436,049 instances) overlap an orthologous H3K9me3 region with at least 50% of their length (denoted ‘overlapping’, Figure 2B and Supplementary file 3; we note that our subsequent conclusions are robust with respect to the specific cutoff used to classify a TE as overlapping with H3K9me3 regions). The TEs that do not overlap with an H3K9me3 region (‘non-overlapping’; 3,600,816 instances), and are putatively not silenced by this repressive mechanism, are an equally interesting category of elements and are therefore included in subsequent analyses. Consistent with previous reports (Turelli et al., 2014; Karimi et al., 2011), we found enrichment of LTRs (Pearson’s Chi-squared test; p<0.001), and SVA elements (p<0.001), among TEs that overlap H3K9me3 regions (Figure 2C).

Figure 2 with 1 supplement see all

Download asset Open asset

Given the proposed evolutionary arms race between TEs and the host repressive machinery (Jacobs et al., 2014), we sought to investigate the within-species patterns of silencing at both orthologous and non-orthologous TEs. We further categorized non-orthologous TEs into those that are only annotated in that species (defined as ‘species-specific TEs’; for example, SVA-E and SVA-F, which are only annotated in human; Table 6 in Supplementary file 1). Generally, we found that the same TE classes that most frequently overlap with H3K9me3 regions in the orthologous TE set, also most frequently overlap with H3K9me3 in the non-orthologous TE set (Figure 3). However, within each species, the SVA and LINE TE classes show a reduction in the proportion of TEs overlapping H3K9me3 regions in the non-orthologous, and species-specific categories (p (SVA) <0.001; p (LINE) <0.001); Figure 3—figure supplement 1). The magnitude of the effect differs between these two TE classes: For example, 85% of orthologous SVA elements overlap H3K9me3, but only 41% of non-orthologous SVA elements overlap the regions marked by H3K9me3. In contrast, 17% and 14% orthologous and non-orthologous LINE elements overlap H3K9me3 regions, respectively. In this case, the difference in effect size could reflect heterogeneity in silencing amongst the diverse elements within the large LINE class. We also found a reduction of TE and H3K9me3 overlap in human LTRs (p<0.001), but an increase in H3K9me3 overlap with LTRs that are chimpanzee-specific (p<0.001; Figure 3). In general, these observations are consistent with the notion that more recent TEs are generally less likely to be silenced, in line with the evolutionary arms race theory (Kapopoulou et al., 2016; Huntley et al., 2006; Thomas and Schneider, 2011).

Figure 3 with 1 supplement see all

Download asset Open asset

The majority of orthologous TEs are similarly silenced in human and chimpanzee

To determine the extent to which orthologous TEs are similarly silenced in both species, we performed a comparative quantitative analysis of the ChIP-seq read counts in the orthologous TEs that overlap H3K9me3 regions. As mentioned previously, our comparative analysis is not dependent on the identification of ‘significant’ ChIP-seq peaks in both species. Instead, we tested for evidence of differences in H3K9me3 read counts between humans and chimpanzees in all ChIP-seq regions. When we could not reject the null hypothesis (at FDR < 0.01) that similar numbers of H3K9me3 ChIP-seq reads are mapped to orthologous TEs, we referred to the silencing status of these TEs as ‘shared’. We thus classified TEs into the following categories (Supplementary file 3) : (i) TEs that do not overlap with an orthologous H3K9me3 region (non-overlapping); (ii) TEs that are associated with similar enrichment of H3K9me3 in both species (shared); (iii) TEs that are associated with significantly higher enrichment of H3K9me3 in humans or (iv) in chimpanzees.

We observed that the proportion of TEs that overlap with H3K9me3 regions differs substantially across different TE classes (for example, 60% of SVA elements overlap H3K9me3 regions compared to only 10% of DNA transposons). Yet, the proportion of TEs for which we have evidence for inter-species differences in silencing is consistent, regardless of TE class (Figure 4A). Indeed, while ultimately most TEs (88%) do not overlap H3K9me3 regions, when TEs do overlap H3K9me3 regions, they are generally (82% of TEs) silenced to the same extent in the two species. Our analysis includes hundreds of thousands of TEs that overlap H3K9me3 regions; thus even just 18% of TEs that overlap H3K9me3 and are silenced differently across species still comprise of tens of thousands of instances (53,201 TEs enriched for H3K9me3 in human, and 28,110 TEs in chimpanzee).

Figure 4 with 3 supplements see all

Download asset Open asset

Thousands of instances of differential silencing notwithstanding, the overall emerging pattern is that most TEs are silenced similarly between species. Yet, it is challenging to draw conclusions based on the inability to reject a null hypothesis of no differences between the species. We therefore used an alternative approach, based on an assumption of a unimodal distribution of effects (ashr) (Stephens, 2017) to obtain a more sensitive estimate of the proportion of inter-species differences at different effect sizes. This unimodal assumption in an Empirical Bayes framework produces smaller (more conservative) estimates of the proportion of tests under the null through adaptive shrinkage. This test is therefore more sensitive to smaller differences in effect size across species. Using this approach, we again find a high degree of conservation in TE silencing, with practically all inter-species differences in silencing associated with small effect sizes (smaller than two fold; Figure 1—figure supplement 14).

Given the diversity of TEs within a class, we proceeded by considering a higher resolution classification of TEs to the 11 most predominant TE families representing each of the 5 TE classes, and then to TE types within specific families (Figure 2). There is some degree of heterogeneity in inter-species silencing differences (‘effect size’) between different TE families (75–84% depending on the family; Figure 4—figure supplement 1). We particularly focused on the LTR and LINE classes, given the previously reported association of these TEs with TRIM28-mediated silencing mechanisms (Turelli et al., 2014; Castro-Diaz et al., 2014). Human LTRs are comprised of HERV elements which can be alternatively classified into Class I (HERV-F,H,I,E,R,W families), Class II (HERV-K), and Class III (HERV-L), based on the original retrovirus that integrated into the genome (van der Kuyl, 2012). Using a subset of these HERV types curated by Turelli et al., where the annotation is likely to be correct, we observed considerable heterogeneity in the frequency with which different HERV types overlap H3K9me3 regions (Figure 4—figure supplement 2A). Notably, less than 20% of 2,079 orthologous HERVH elements, which are generally known to be expressed in pluripotent cells, overlap with H3K9me3 regions (compared to 80% of 3,484 THE1B elements, which are not known to be expressed in pluripotent cells, for example). This heterogeneity is also reflected in the proportion of TEs that are similarly silenced in the two species (55–95% sharing, Figure 4—figure supplement 2B). However, the total number of elements within the more divergent TE types is small (fewer than 100 TE instances per category with less than 75% sharing). In contrast, when turning our attention to a subset of mammalian-specific L1 LINE elements (Castro-Diaz et al., 2014), we found less variability in overlap with H3K9me3 across types (5–25% are overlapping; Figure 4—figure supplement 3A). However, of those L1 elements that overlap H3K9me3, the majority (82% of 21,710 elements) are similarly silenced in both species (Figure 4—figure supplement 3B). These results suggest that there is some degree of sequence specificity in the silencing mechanisms across TE types.

Silenced TEs tend to have similar functional properties across species

We next sought to identify properties that might distinguish TEs based on our silencing classifications (shared, species-specific, or not silenced). We first considered the length of a TE, which may relate to its transposition competency and hence need for silencing. On average, TEs silenced in at least one species are longer than those that do not overlap orthologous H3K9me3 regions (Wilcoxon-rank sum test; p<0.001; Figure 5A). To confirm that longer TEs do not overlap H3K9me3 more often by chance alone, we artificially increased the length of the non-overlapping TEs to match the median length of TEs marked by H3K9me3 in both species. The artificially longer TEs do not overlap with H3K9me3 regions more often, suggesting that TE elements are silenced in a non-random manner, which cannot be simply explained by length in itself. We found the same pattern when we analyzed TEs by class, which in itself is associated with different lengths (p<0.001; Figure 5—figure supplement 1A). Interestingly, across TE families there is a strong correlation between the proportion of TEs that overlap H3K9me3 regions and the median length of the TEs in the family (Pearson’s correlation = 0.93; Figure 5—figure supplement 1B–C). However, within families the median length of TEs that do not overlap H3K9me3 is highly similar to that of TEs that do (Pearson correlation for median TE lengths across families = 0.99; Figure 5—figure supplement 1D), again suggesting that this is not a feature of length itself. When we stratified TEs by the magnitude of inter-species silencing divergence (absolute effect size), we found that shorter TEs are associated with higher silencing divergence (p<0.001; Figure 5B). Put together, these observations suggest that longer TEs may be silenced more often because they are more likely to have unwanted regulatory potential, and are therefore more likely to contain sequences that recruit silencing machinery. This is consistent with the finding that more KRAB-ZNF proteins (sequence-specific mediators of TE silencing) bind longer TEs (Imbeault et al., 2017).

Figure 5 with 6 supplements see all

Download asset Open asset

To gain insight into the relationship between TE silencing and TE transposition and mobility, we asked whether TE copy number is correlated with silencing. Across TE families, we found only a weak correlation between TE copy number and the proportion that overlap with H3K9me3 regions (Pearson’s correlation = −0.47; permutation p=0.15; Figure 5—figure supplement 2). Moreover, in contrast to the general trend, TEs in the two smallest families (780 SVAs and 7,983 ERVKs) overlap H3K9me3 regions most often (Figure 5—figure supplement 2).

Given that TEs can acquire mutations and transpose over time, we compared the sequence divergence between each individual TE instance and the corresponding consensus TE sequence (see Materials and methods). We found that TEs that are silenced similarly in both species are evolutionarily less diverged from the consensus sequence than TEs that do not overlap H3K9me3 regions (p<0.001; Figure 5C). This pattern is observed within the LTR, LINE, SINE and DNA TE classes (p<0.001; Figure 5—figure supplement 3A), but this pattern does not hold within all TE families, as might be expected considering the diversity of TE types and families within a class (Figure 5—figure supplement 3B). Across all TEs, we did not find a difference in sequence divergence between TEs that are silenced in just one or both species (Figure 5C) but stratification of TEs by the magnitude of the inter-species silencing effect reveals that when inter-species differences in silencing are larger, the sequences of the TEs are more diverged from the consensus sequence (p<0.001; Figure 5D).

We next considered proximity of TEs to genes by determining the distance between each TE and the closest orthologous annotated Transcription Start Site (TSS). TEs that are silenced in both species are significantly further away from the TSS than TEs that are preferentially silenced in one species, or those that do not overlap with H3K9me3 regions (p<0.001; Figure 5E). Again, this pattern is observed within the LTR, LINE, SINE and DNA classes, but not within all TE families (p<0.001; Figure 5—figure supplement 4). TEs that show greater divergence in silencing across species are also more likely to be closer to the TSS than TEs with lower divergence (p<0.001; Figure 5F). These results suggest that, across families, there are some differences in the functional properties of silenced TEs.

In order to determine whether silenced TEs are also associated with other genomic features, we overlaid orthologous TEs with chromatin state annotations defined in human ESCs (Ernst and Kellis, 2010). The majority of TEs, regardless of silencing status, are associated with chromatin regions annotated as ‘heterochromatin and low signal’. We found that TEs that are preferentially silenced in human overlap annotated regulatory regions less frequently compared with TEs that are preferentially silenced in chimpanzee (p<0.001; Figure 5G and Figure 5—figure supplement 5). We also considered published ChIP-seq and DNase I hypersensitivity (DHS) data in H1 hESCs. We found that a relatively small fraction of TEs overlap regions of open chromatin or ChIP-seq peaks: Less than 20% of TEs overlap active histone marks and less than 1% of TEs overlap DHS. We found that TEs that are preferentially silenced in human overlap regions of active gene regulation less often than TEs preferentially silenced in chimpanzee. The active chromatin annotations we considered include marking by H3K27ac, DHS, enhancer elements (marked by H3K4me1, p300) and regions marked by RNA polymerase II occupancy (for all comparisons p<0.01; Figure 5—figure supplement 6). Notably, there is no significant difference between TEs preferentially silenced in either species with respect to overlap with the active promoter-associated mark, H3K4me3.

TE silencing is not a major driver of inter-species gene expression divergence

To determine the potential impact of TE silencing on the regulation of neighboring genes, we characterized gene expression levels in the same iPSC lines from both species (using RNA-sequencing; see Materials and methods, Figure 6A, Figure 6—figure supplement 1A and Tables 2-3 in Supplementary file 1). After restricting the data to a set of human-chimpanzee orthologous exons (Blekhman et al., 2010) the RNA-seq data clusters primarily by species, and then by population (i.e. Caucasian or Yoruba/cell type of origin; Figure 6—figure supplement 1B). Indeed, after filtering lowly expressed genes, species emerges as the primary driver of variation in the RNA-seq data, as expected (Figure 6—figure supplement 1C).

Figure 6 with 10 supplements see all

Download asset Open asset

To characterize the level of TE silencing associated with each orthologous gene, we defined a cis-regulatory region 10 kb upstream of the TSS of each gene that includes at least one annotated TE (Materials and methods and Figure 6A). As expected, we observed that gene expression levels decrease when TEs in the cis-regulatory window overlap with H3K9me3 regions (p<0.01; Figure 6B). This observation is consistent with the general reported role of the H3K9me3 histone modification in gene regulation (Bilodeau et al., 2009; Mozzetta et al., 2015). The effect of TE silencing on gene expression is recapitulated when extending the window size to 20 kb, while TEs silenced within 1 kb upstream of the TSS have a stronger effect on gene expression (Figure 6—figure supplement 2). We then asked whether inter-species differences in TE silencing are associated with expression divergence of nearby genes.

To obtain a quantitative measure of overall TE silencing in each species, we summed H3K9me3 ChIP-seq read counts from orthologous H3K9me3 regions overlapping TEs within 10 kb of annotated genes. We used the same linear model framework described above to estimate inter-species difference in H3K9me3 counts, within a 10 kb window upstream of the TSS. Of the windows that contain at least one orthologous H3K9me3 region overlapping a TE, 79% have similar levels of H3K9me3 enrichment in human and chimpanzee (3,523), with 11% of regions enriched in human (499), and 10% of regions enriched in chimpanzee (463), at FDR of 1% (Figure 6C). We then considered inter-species differences in expression levels for genes associated with similar or species-specific H3K9me3 enrichment at TEs.

We found no significant difference in the expression divergence effect size between genes associated with species-specific or shared TE silencing, using 10, 20 or 40 kb windows upstream of the TSS (Figure 6C and Figure 6—figure supplement 3). We did find a weak, negative association between gene expression and silencing divergence when we considered silenced TEs that are located within 1 kb upstream of the TSS (only 165 genes in that set; Figure 6—figure supplements 3 and 4; Table 7 in Supplementary file 1). It is tempting to conclude that TEs (or silenced TEs) near the TSS can have a marked contribution to expression divergence; however, H3K9me3 enrichment in regions that overlap a 1 kb region around the TSS is similarly associated with expression divergence, irrespective of the presence of a TE (Figure 6—figure supplement 4). Our observations are robust with respect to a wide range of FDR cutoffs used to classify inter-species differences in silencing, suggesting that incomplete power is not a likely explanation for the overall lack of association between expression divergence and TE silencing (Table 8 in Supplementary file 1).

To further investigate the overall lack of association between silencing and gene expression divergence, we performed an analogous analysis, this time using gene expression divergence as the anchor. We found no association between silencing and inter-species expression divergence: genes that are classified as differentially expressed between humans and chimpanzees are not more likely to be associated with inter-species differences in H3K9me3 at TEs in their putative regulatory regions (Figure 6D and Figure 6—figure supplement 5A). Again, we could only find a relationship between silencing and expression divergence when considering H3K9me3 overlapping TEs within 1 kb upstream of the TSS. The observation of no association of inter-species expression divergence and TE silencing is robust with respect to a wide range of statistical cutoffs used to classify either differential expression or differential silencing (Figure 6—figure supplement 5B and Table 9 in Supplementary file 1). Indeed, when we considered the entire distribution, the correlation between inter-species effect sizes for gene expression divergence and H3K9me3 enrichment is weak (Pearson’s correlation = −0.0005 in a 10 kb window; the correlation in divergence increases somewhat to −0.14 when considering a 1 kb window upstream of the TSS; Figure 6—figure supplement 3). Moreover, we observed nearly as many genes whose expression divergence is consistent with the observed patterns of TE silencing divergence (namely, TE silencing associated with lower expression level) as those that show the opposite, inconsistent pattern (Figure 6E).

We looked at the relationship between expression divergence and TE silencing more closely by considering orthologous TEs that are most likely to regulate neighboring gene expression. Specifically, we focused on the 532 genes closest to TEs that overlap either enhancer (marked by H3K4me1) or promoter (marked by H3K4me3) regions in human, and yet are silenced in chimpanzee (Figure 5—figure supplement 6). The results are consistent with our observations for the entire set. We found a relationship between expression divergence and silencing of TEs that overlap the promoter-associated mark but not of TEs that overlap enhancers (typically further from the TSS; Figure 6F and Figure 6—figure supplement 6).

We next reasoned that perhaps TEs that are actively expressed in this cell type, in at least one species, are more likely to affect gene expression divergence. Indeed, TEs can be transcribed and contribute to lncRNA and neighboring gene expression (Ramsay et al., 2017; Karimi et al., 2011). Using our RNA-seq data, we identified 11 poly-adenylated TE types that are expressed in chimpanzee, and 7 that are expressed in human (Materials and methods; Table 10A in Supplementary file 1; note that this approach does not allow us to capture non-polyadenylated transcripts such as eRNAs). In both species, we identified the expression of HERVH elements, which are known to be expressed in human pluripotent cells. Of the nine TE types that are annotated in both species and expressed in at least one species, 1,025 instances are contained within our orthologous TE set (Table 10B in Supplementary file 1 and 123 instances overlap an H3K9me3 region. Only 14 of these TE instances are enriched for H3K9me3 in human, and 22 in chimpanzee (FDR < 0.01). These TEs are proximal (within 10 kb) of only 19 orthologous genes. Thus, due to the small numbers we are unable to properly test the association between TE silencing/expression and gene expression divergence. In any case, even if there is such an association in these cases, the small number of instances suggests that gene regulation by an actively expressed poly-adenylated TE in one species is not a major mechanism of gene expression divergence.

Further attempting to find a relationship between TE silencing and expression divergence, we hypothesized that perhaps imperfect annotation of orthologous TEs may affect our observations. A clear association between silencing of non-orthologous TEs and gene expression divergence would provide some measure of support for this possible explanation. We thus performed a similar analysis focused only on the non-orthologous TEs. In this analysis, we considered non-orthologous TEs that are located 10 kb upstream of annotated orthologous genes. We classified the non-orthologous TEs as ‘silenced’ or ‘not silenced’ based on all H3K9me3 regions identified in that species (we did not require H3K9me3 regions to be orthologous in this analysis). We found, once again, that there is no association between TE silencing and expression divergence (Figure 6—figure supplement 7). A significant effect is only consistently observed when considering a 1 kb window immediately upstream of the TSS (a set that includes 289 genes; Figure 6—figure supplement 7). In other words, TE silencing by H3K9me3, regardless of orthology, has a modest effect on gene expression divergence in human and chimpanzee iPSC lines.

We next focused our attention on known mediators of TE silencing, the KRAB-ZNF genes, in an attempt to identify species-specific regulatory effects of TE silencing. KRAB-ZNF genes, in addition to directing the repression of TEs, are themselves marked by H3K9me3 (Roadmap Epigenomics Consortium et al., 2015; O'Geen et al., 2007). We intersected a curated list of 256 KRAB-ZNF genes (Kapopoulou et al., 2016), filtered for orthology, with our list of orthologous H3K9me3 regions. We found that 104 (41%) of these genes overlap with orthologous H3K9me3 regions (Figure 6—figure supplement 8A). The proportion of KRAB-ZNFs with similar H3K9me3 enrichment in both species (78% of overlapping genes or 81 instances) is comparable to that observed for TEs (Figure 6—figure supplement 8B). 17 KRAB-ZNFs are enriched for H3K9me3 in human and 6 are enriched for H3K9me3 in chimpanzee. We found that the 256 KRAB-ZNF genes are no more likely to be differentially expressed between human and chimpanzee than all 17,354 genes (52% of KRAB-ZNFs are classified as differentially expressed and 49% of all genes; FDR of 1%; Figure 6—figure supplement 9A and Table 11 in Supplementary file 1). There are 14 KRAB-ZNF genes whose TSS overlaps an H3K9me3 region that is enriched in human (eight) or chimpanzee (six). The association between silencing and gene expression divergence is consistent with the pattern we observed for all genes whose TSS overlaps an H3K9me3 region (Figure 6—figure supplement 4) but due to the small numbers in each category, we cannot reject the null hypothesis of no association (Figure 6—figure supplement 9B).

Finally, as H3K9me3 is not the only repressive histone modification in the genome, it is possible that we do not observe a strong association between silencing and expression divergence because there is redundancy in histone-based TE silencing mechanisms. To address this possibility, we considered genome-wide data from the Polycomb-repressed associated histone modification H3K27me3 in human H1 hESCs. We found that TEs preferentially silenced by H3K9me3 in chimpanzee are not associated more often with increased levels of H3K27me3 compared to TEs preferentially silenced by H3K9me3 in human (Figure 6—figure supplement 10). This suggests that H3K27me3 and H3K9me3 are not redundant in silencing TEs across species.

The potential for TEs to participate in the regulation of gene expression has received increasing attention as genomic technologies to study these elements advance (Davidson and Britten, 1979; Jordan et al., 2003; Feschotte, 2008; Göke and Ng, 2016). However, the role of TEs can be paradoxical. They can serve as regulatory sequence when they are bound by transcription factors, or localize in regions of open chromatin. Conversely, TEs can be actively silenced to prevent unwanted activity in a cell-type dependent manner. One cell type where TE activity can be particularly deleterious to genome integrity is embryonic stem cells. Indeed, it was found that TE activity is generally restricted through repressive histone modifications in mouse embryonic stem cells (Schlesinger and Goff, 2015; Robbez-Masson and Rowe, 2015). TE activity and regulation has been studied in many contexts, but there are no published comparative studies of TE silencing in primates.

In order to gain insight into how TE silencing mediated by the TRIM28/SETDB1 pathway evolves in pluripotent cells from the hominid lineage, we profiled the distribution of the repressive histone modification H3K9me3 in human and chimpanzee iPSCs. The rationale behind our approach is the expectation that TEs that are marked by H3K9me3 are silent and are therefore not mobile, nor able to act as regulatory elements in this cell type. In contrast, TEs that are unmarked and not silenced, could be expressed, or capable of transposition, or participation in gene regulation. Alternatively, these elements may have acquired sufficient mutations to lose the ability to transpose, or influence gene regulatory programs, and are therefore not silenced because they do not ‘threaten’ genome integrity. We were particularly interested in finding differences in silencing of the same TEs between species, as this could imply that these TEs act as regulatory elements in only one species, and could contribute to phenotypic differences between species.

We found that there are differences in the TE families that are preferentially silenced in both species. This suggests specificity of silencing and targeting mechanisms, which are shared in humans and chimpanzees. In our study, up to 60% of SVA, 50% of ERV, and 15% of L1 families overlap with H3K9me3 in both species. It is likely that TE silencing in our human and chimpanzee iPSC lines is mediated by the TRIM28 co-repressor as it is known to be involved in silencing ERVs in mouse ESCs (Rowe et al., 2010), and SVAs, HERVs and L1 elements in human ESCs (Turelli et al., 2014; Castro-Diaz et al., 2014). This specificity is likely achieved through sequence-specific KRAB-ZNFs that recruit TRIM28. Indeed, specific KRAB-ZNFs recognize specific TE families (Jacobs et al., 2014; Imbeault et al., 2017; Schmitges et al., 2016).

While TEs silenced by H3K9me3 in both species tend to be less diverged from the corresponding TE consensus sequence than those that are not silenced, TEs that have large inter-species differences in silencing (more than twofold) tend to resemble TEs that are not silenced. This could imply that TEs whose silencing differs across species are less likely to require silencing, or that they are more likely to have enhancer function in just one species. Indeed, amongst TE-derived enhancers defined by CAGE in human, there is a depletion of younger, primate-specific TEs (Simonti et al., 2017). Assuming an evolutionary arms race model, one expects that TEs and their KRAB-ZNF repressors would co-evolve; indeed there is a correlation between the age of KRAB-ZNF genes and their target TEs (Najafabadi et al., 2015), and related KRAB-ZNFs bind related TEs (Schmitges et al., 2016). However, the recent observation that many KRAB-ZNF gene – TE pairs arose at the same time during evolutionary history, and that this relationship is maintained to this day despite the fact that these TEs are no longer able to transpose, challenges the idea of a broad evolutionary arms race (Imbeault et al., 2017). In our study, we found silencing by H3K9me3 in TEs that are no longer transposition competent. It could be that these TEs are not silenced to prevent their mobility, but rather to prevent their potential regulatory activity.

TEs that are silenced in both species are located further away from the TSS than TEs that are silenced preferentially in one species. This difference between the ‘shared’ and ‘species-specific’ silenced TEs indicates that we do not have an overwhelming proportion of false negatively classified TEs in the ‘shared’ group. This observation also raises the possibility that TEs that are silenced only in one species are more likely to contribute to gene expression differences between species. However, we did not find this to be the case. Within species, genes with silenced TEs that overlap promoters have lower expression levels, on average, than genes with proximal TEs that are not silenced. Across species, inter-species differences in TE silencing are only associated with divergence in gene expression when the TEs are located in promoter regions, within 1 kb of the TSS. Further from genes, we focused on TEs that overlap enhancer regions, as Imbeault et al., have shown that KRAB-ZNF gene-bound TEs only affect neighboring gene expression if the TE overlaps marks of enhancers (Imbeault et al., 2017). In our data, however, we did not find such an effect.

Our observation of overall lack of association between TE silencing and expression divergence should be considered in the following context: First, while it is intuitive to relate a promoter to a gene, it is more challenging to relate a distant enhancer to the gene it regulates. Second, it has recently been shown that while many TEs contribute to primate-specific cis-regulatory elements (CREs), many of these TEs act as transcriptional repressors and hence silencing of these elements may in fact lead to increased expression of neighboring genes (Trizzino et al., 2017). Third, we did not investigate the functional consequences of sequence changes to orthologous TEs, so our results do not account for the fact that mutations in the sequence of TEs in one species may inactivate them, eliminating the need for histone-based silencing. Fourth, we have restricted our analysis to orthologous genes; it is possible that regulatory novelty co-occurs with gene novelty, though admittedly this is rare when the species considered are human and chimpanzee. Fifth, we were unable to interrogate silencing of recently active TEs, due to our stringent H3K9me3 ChIP-seq read mapping strategy, which excludes reads that do not map uniquely.

Overall, we identified tens of thousands of inter-species differences in silencing at individual TE instances; however, considering that there are roughly four million human-chimpanzee orthologous TE sequences in the genome, these differences account for a minority of elements (2%). Of the TEs that are silenced, the observed level of conservation of TE silencing in human and chimpanzee is in line with what is observed for all H3K9me3 regions in the genome irrespective of the presence of a TE. Conservation of active regulatory regions, determined by H3K27ac, in human and chimpanzee is also thought to be similar (Prescott et al., 2015). It may be that even divergence at a small fraction of regulatory elements between human and chimpanzee – reflecting tens of thousands of instances – is sufficient to explain much of the observed divergence in gene expression levels. Yet, although we performed multiple analyses to uncover such a relationship, we concluded that, at most, inter-species differences in TE silencing have a modest effect on gene expression divergence in iPSCs.

We acknowledge that proving the absence of a species-specific effect is challenging. In general, it is difficult to draw strong conclusions based on the inability to reject a null hypothesis. Nevertheless, our observations are robust with respect to a wide range of statistical thresholds used to identify TE silencing or gene expression differences. Our observations should also be interpreted in the context of our initial assumption about what can be learned from characterizing TE silencing by H3K9me3 in iPSCs. First, while H3K9me3-mediated TE silencing is a well-established repressive mechanism in pluripotent cells, additional mechanisms of TE silencing exist; including RNA interference, CpG methylation and other repressive histone modifications such as H3K27me3 (Slotkin and Martienssen, 2007; Huda et al., 2010; Day et al., 2010; Leeb et al., 2010), and thus our findings may not extend to these other silencing pathways. For example, Marchetto et al., showed that there is differential regulation of L1 elements by piRNA-associated PIWIL2 and APOBEC3B in human and chimpanzee iPSCs (Marchetto et al., 2013). Pluripotent stem cells can be considered to be an usual cell type in many respects, including TE regulation. In these cells, TE activity is known to be regulated by a histone-based silencing mechanism following the global resetting of DNA methylation. While there are examples of this TE repression pathway in differentiated cell types (Ecco et al., 2016; Brattås et al., 2017), it is largely believed that DNA methylation replaces histone-mediated repression following the DNA de-methylation phase in embryogenesis. Therefore, alternative TE silencing mechanisms predominant in other cell types, such as DNA methylation, may be more dynamic across species. Similarly, one should also consider that the ability of TEs to act as enhancer elements has been shown to be cell-type specific (Xie et al., 2013). Our results may therefore not extend to other somatic cell types, which are not under the same level of regulatory constraint as iPSCs.

Our results complement several previous studies, which explored the role of transposable elements in primate gene regulation. The primate-specific HERVH LTR element is a human pluripotency marker (Wang et al., 2014) and has been shown to have similar levels of expression in chimpanzee iPSCs (Ramsay et al., 2017). Indeed, we find HERVH to be expressed in human and chimpanzee iPSCs, and to be depleted for H3K9me3, suggesting that this element is essential for the pluripotency network in both humans and chimpanzees. Many open chromatin regions in human are derived from primate-specific TEs (Jacques et al., 2013), and newly evolved CREs are enriched in young SVA and LTR elements (Trizzino et al., 2017). Trizzino et al. found that 2% of all LTRs are active (marked by H3K27ac) in liver but a quarter of genes with an LTR insertion affect gene expression divergence between multiple primate species, and that less than 1% of CREs are differentially active between human and chimpanzee. This suggests that while many TEs, specifically young TEs, are important for influencing inter-species differences in expression, the majority of TEs in the genome do not act as regulatory sequences.

When considering the broader context for the contribution of mammalian TEs to mediating inter-species differences in gene regulation, previous elegant work has dissected the role of specific TEs in inter-species differences between mice and rats (Chuong et al., 2013; Sundaram et al., 2017). TE activity is substantial in the rodent lineage compared to hominids (Lander et al., 2001), which suggests that TEs can more quickly rewire regulatory networks in these species. Consistent with the reduction in TE activity in primates compared to other mammals, ChIP-seq assays for the highly conserved transcriptional insulator CTCF, have shown reduced TE-mediated propagation of CTCF binding sites in primates compared to other mammalian lineages (Schmidt et al., 2012; Schwalie et al., 2013).

In summary, to date, there have been few genome-wide studies demonstrating the contribution of TEs to gene regulation in primates. TEs have been shown to comprise most primate-specific regulatory sequence (Jacques et al., 2013), to contribute to primate-conserved non-coding transcripts (Ramsay et al., 2017), and lineage-specific gene regulation (Trizzino et al., 2017). More broadly, while gene expression differences between primates have been shown to often be associated with inter-species differences in histone modifications (Cain et al., 2011; Zhou et al., 2014), the chromatin landscape is generally highly conserved in primates. For example, as many as 84% of genomic regions enriched for the active histone modification, H3K27ac, are similar between humans and chimpanzees in iPSC-derived neural crest cells (Prescott et al., 2015). Our results provide further support for the notion that chromatin state is generally highly conserved between closely related primate species. We have not found evidence that TEs, whether orthologous or not, have substantially rewired pluripotent stem cell gene regulatory networks between humans and chimpanzees.

1. 1. Ward MC

   (2017) Epigenomic conservation of transposable element silencing

   Publicly available at the NCBI Gene Expression Omnibus (accession no: GSE96712).

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE96712

1. 1. Rosenbloom

   (2013) ENCODE data in the UCSC Genome Browser

   Publicly available at the UCSC Genome Browser.

   http://hgdownload.cse.ucsc.edu/goldenPath/hg19/database

1. 1. Anders S
   2. Pyl PT
   3. Huber W

   (2015) HTSeq--a Python framework to work with high-throughput sequencing data

   Bioinformatics 31:166–169.

   https://doi.org/10.1093/bioinformatics/btu638

   • PubMed
   • Google Scholar
2. 1. Banovich NE
   2. Li YI
   3. Raj A
   4. Ward MC
   5. Greenside P
   6. Calderon D
   7. Tung PY
   8. Burnett JE
   9. Myrthil M
   10. Thomas SM
   11. Burrows CK
   12. Romero IG
   13. Pavlovic BJ
   14. Kundaje A
   15. Pritchard JK
   16. Gilad Y

   (2018) Impact of regulatory variation across human iPSCs and differentiated cells

   Genome Research 28:122–131.

   https://doi.org/10.1101/gr.224436.117

   • Google Scholar
3. 1. Barau J
   2. Teissandier A
   3. Zamudio N
   4. Roy S
   5. Nalesso V
   6. Hérault Y
   7. Guillou F
   8. Bourc'his D

   (2016) The DNA methyltransferase DNMT3C protects male germ cells from transposon activity

   Science 354:909–912.

   https://doi.org/10.1126/science.aah5143

   • PubMed
   • Google Scholar
4. 1. Batzer MA
   2. Deininger PL

   (1991) A human-specific subfamily of Alu sequences

   Genomics 9:481–487.

   https://doi.org/10.1016/0888-7543(91)90414-A

   • PubMed
   • Google Scholar
5. 1. Batzer MA
   2. Gudi VA
   3. Mena JC
   4. Foltz DW
   5. Herrera RJ
   6. Deininger PL

   (1991) Amplification dynamics of human-specific (HS) Alu family members

   Nucleic Acids Research 19:3619–3623.

   https://doi.org/10.1093/nar/19.13.3619

   • PubMed
   • Google Scholar
6. 1. Bilodeau S
   2. Kagey MH
   3. Frampton GM
   4. Rahl PB
   5. Young RA

   (2009) SetDB1 contributes to repression of genes encoding developmental regulators and maintenance of ES cell state

   Genes & Development 23:2484–2489.

   https://doi.org/10.1101/gad.1837309

   • Google Scholar
7. 1. Blekhman R
   2. Marioni JC
   3. Zumbo P
   4. Stephens M
   5. Gilad Y

   (2010) Sex-specific and lineage-specific alternative splicing in primates

   Genome Research 20:180–189.

   https://doi.org/10.1101/gr.099226.109

   • PubMed
   • Google Scholar
8. 1. Bourc'his D
   2. Bestor TH

   (2004) Meiotic catastrophe and retrotransposon reactivation in male germ cells lacking Dnmt3L

   Nature 431:96–99.

   https://doi.org/10.1038/nature02886

   • PubMed
   • Google Scholar
9. 1. Bourque G
   2. Leong B
   3. Vega VB
   4. Chen X
   5. Lee YL
   6. Srinivasan KG
   7. Chew J-L
   8. Ruan Y
   9. Wei C-L
   10. Ng HH
   11. Liu ET

   (2008) Evolution of the mammalian transcription factor binding repertoire via transposable elements

   Genome Research 18:1752–1762.

   https://doi.org/10.1101/gr.080663.108

   • Google Scholar
10. 1. Brattås PL
    2. Jönsson ME
    3. Fasching L
    4. Nelander Wahlestedt J
    5. Shahsavani M
    6. Falk R
    7. Falk A
    8. Jern P
    9. Parmar M
    10. Jakobsson J

    (2017) TRIM28 controls a gene regulatory network based on endogenous retroviruses in human neural progenitor cells

    Cell Reports 18:1–11.

    https://doi.org/10.1016/j.celrep.2016.12.010

    • PubMed
    • Google Scholar
11. 1. Brouha B
    2. Schustak J
    3. Badge RM
    4. Lutz-Prigge S
    5. Farley AH
    6. Moran JV
    7. Kazazian HH

    (2003) Hot L1s account for the bulk of retrotransposition in the human population

    PNAS 100:5280–5285.

    https://doi.org/10.1073/pnas.0831042100

    • Google Scholar
12. 1. Burrows CK
    2. Banovich NE
    3. Pavlovic BJ
    4. Patterson K
    5. Gallego Romero I
    6. Pritchard JK
    7. Gilad Y

    (2016) Genetic variation, not cell type of origin, underlies the majority of identifiable regulatory differences in iPSCs

    PLoS Genetics 12:e1005793.

    https://doi.org/10.1371/journal.pgen.1005793

    • PubMed
    • Google Scholar
13. 1. Cain CE
    2. Blekhman R
    3. Marioni JC
    4. Gilad Y

    (2011) Gene expression differences among primates are associated with changes in a histone epigenetic modification

    Genetics 187:1225–1234.

    https://doi.org/10.1534/genetics.110.126177

    • PubMed
    • Google Scholar
14. 1. Castro-Diaz N
    2. Ecco G
    3. Coluccio A
    4. Kapopoulou A
    5. Yazdanpanah B
    6. Friedli M
    7. Duc J
    8. Jang SM
    9. Turelli P
    10. Trono D

    (2014) Evolutionally dynamic L1 regulation in embryonic stem cells

    Genes & Development 28:1397–1409.

    https://doi.org/10.1101/gad.241661.114

    • PubMed
    • Google Scholar
15. 1. Chimpanzee Sequencing and Analysis Consortium

    (2005) Initial sequence of the chimpanzee genome and comparison with the human genome

    Nature 437:69–87.

    https://doi.org/10.1038/nature04072

    • PubMed
    • Google Scholar
16. 1. Chuong EB
    2. Elde NC
    3. Feschotte C

    (2016) Regulatory evolution of innate immunity through co-option of endogenous retroviruses

    Science 351:1083–1087.

    https://doi.org/10.1126/science.aad5497

    • PubMed
    • Google Scholar
17. 1. Chuong EB
    2. Elde NC
    3. Feschotte C

    (2017) Regulatory activities of transposable elements: from conflicts to benefits

    Nature Reviews Genetics 18:71–86.

    https://doi.org/10.1038/nrg.2016.139

    • PubMed
    • Google Scholar
18. 1. Chuong EB
    2. Rumi MA
    3. Soares MJ
    4. Baker JC

    (2013) Endogenous retroviruses function as species-specific enhancer elements in the placenta

    Nature Genetics 45:325–329.

    https://doi.org/10.1038/ng.2553

    • PubMed
    • Google Scholar
19. 1. Cordaux R
    2. Batzer MA

    (2009) The impact of retrotransposons on human genome evolution

    Nature Reviews Genetics 10:691–703.

    https://doi.org/10.1038/nrg2640

    • PubMed
    • Google Scholar
20. 1. Davidson EH
    2. Britten RJ

    (1979) Regulation of gene expression: possible role of repetitive sequences

    Science 204:1052–1059.

    https://doi.org/10.1126/science.451548

    • PubMed
    • Google Scholar
21. 1. Day DS
    2. Luquette LJ
    3. Park PJ
    4. Kharchenko PV

    (2010) Estimating enrichment of repetitive elements from high-throughput sequence data

    Genome Biology 11:R69.

    https://doi.org/10.1186/gb-2010-11-6-r69

    • Google Scholar
22. 1. de Koning AP
    2. Gu W
    3. Castoe TA
    4. Batzer MA
    5. Pollock DD

    (2011) Repetitive elements may comprise over two-thirds of the human genome

    PLoS Genetics 7:e1002384.

    https://doi.org/10.1371/journal.pgen.1002384

    • PubMed
    • Google Scholar
23. 1. Derrien T
    2. Estellé J
    3. Marco Sola S
    4. Knowles DG
    5. Raineri E
    6. Guigó R
    7. Ribeca P

    (2012) Fast computation and applications of genome mappability

    PLoS One 7:e30377.

    https://doi.org/10.1371/journal.pone.0030377

    • PubMed
    • Google Scholar
24. 1. Ecco G
    2. Cassano M
    3. Kauzlaric A
    4. Duc J
    5. Coluccio A
    6. Offner S
    7. Imbeault M
    8. Rowe HM
    9. Turelli P
    10. Trono D

    (2016) Transposable elements and their KRAB-ZFP controllers regulate gene expression in adult tissues

    Developmental Cell 36:611–623.

    https://doi.org/10.1016/j.devcel.2016.02.024

    • Google Scholar
25. 1. ENCODE Project Consortium

    (2012) An integrated encyclopedia of DNA elements in the human genome

    Nature 489:57–74.

    https://doi.org/10.1038/nature11247

    • PubMed
    • Google Scholar
26. 1. Ernst J
    2. Kellis M

    (2010) Discovery and characterization of chromatin states for systematic annotation of the human genome

    Nature Biotechnology 28:817–825.

    https://doi.org/10.1038/nbt.1662

    • PubMed
    • Google Scholar
27. 1. Faulkner GJ
    2. Kimura Y
    3. Daub CO
    4. Wani S
    5. Plessy C
    6. Irvine KM
    7. Schroder K
    8. Cloonan N
    9. Steptoe AL
    10. Lassmann T
    11. Waki K
    12. Hornig N
    13. Arakawa T
    14. Takahashi H
    15. Kawai J
    16. Forrest ARR
    17. Suzuki H
    18. Hayashizaki Y
    19. Hume DA
    20. Orlando V
    21. Grimmond SM
    22. Carninci P

    (2009) The regulated retrotransposon transcriptome of mammalian cells

    Nature Genetics 41:563–571.

    https://doi.org/10.1038/ng.368

    • Google Scholar
28. 1. Feschotte C

    (2008) Transposable elements and the evolution of regulatory networks

    Nature Reviews Genetics 9:397–405.

    https://doi.org/10.1038/nrg2337

    • Google Scholar
29. 1. Fort A
    2. Hashimoto K
    3. Yamada D
    4. Salimullah M
    5. Keya CA
    6. Saxena A
    7. Bonetti A
    8. Voineagu I
    9. Bertin N
    10. Kratz A
    11. Noro Y
    12. Wong C-H
    13. de Hoon M
    14. Andersson R
    15. Sandelin A
    16. Suzuki H
    17. Wei C-L
    18. Koseki H
    19. Hasegawa Y
    20. Forrest ARR
    21. Carninci P

    (2014) Deep transcriptome profiling of mammalian stem cells supports a regulatory role for retrotransposons in pluripotency maintenance

    Nature Genetics 46:558–566.

    https://doi.org/10.1038/ng.2965

    • Google Scholar
30. 1. Friedli M
    2. Turelli P
    3. Kapopoulou A
    4. Rauwel B
    5. Castro-Díaz N
    6. Rowe HM
    7. Ecco G
    8. Unzu C
    9. Planet E
    10. Lombardo A
    11. Mangeat B
    12. Wildhaber BE
    13. Naldini L
    14. Trono D

    (2014) Loss of transcriptional control over endogenous retroelements during reprogramming to pluripotency

    Genome Research 24:1251–1259.

    https://doi.org/10.1101/gr.172809.114

    • PubMed
    • Google Scholar
31. 1. Fuchs NV
    2. Loewer S
    3. Daley GQ
    4. Izsvák Z
    5. Löwer J
    6. Löwer R

    (2013) Human endogenous retrovirus K (HML-2) RNA and protein expression is a marker for human embryonic and induced pluripotent stem cells

    Retrovirology 10:115.

    https://doi.org/10.1186/1742-4690-10-115

    • Google Scholar
32. 1. Gallego Romero I
    2. Pavlovic BJ
    3. Hernando-Herraez I
    4. Zhou X
    5. Ward MC
    6. Banovich NE
    7. Kagan CL
    8. Burnett JE
    9. Huang CH
    10. Mitrano A
    11. Chavarria CI
    12. Friedrich Ben-Nun I
    13. Li Y
    14. Sabatini K
    15. Leonardo TR
    16. Parast M
    17. Marques-Bonet T
    18. Laurent LC
    19. Loring JF
    20. Gilad Y

    (2015) A panel of induced pluripotent stem cells from chimpanzees: a resource for comparative functional genomics

    eLife 4:e07103.

    https://doi.org/10.7554/eLife.07103

    • PubMed
    • Google Scholar
33. 1. Göke J
    2. Ng HH

    (2016) CTRL+INSERT: retrotransposons and their contribution to regulation and innovation of the transcriptome

    EMBO Reports 17:1131–1144.

    https://doi.org/10.15252/embr.201642743

    • Google Scholar
34. 1. Grow EJ
    2. Flynn RA
    3. Chavez SL
    4. Bayless NL
    5. Wossidlo M
    6. Wesche DJ
    7. Martin L
    8. Ware CB
    9. Blish CA
    10. Chang HY
    11. Reijo Pera RA
    12. Wysocka J

    (2015) Intrinsic retroviral reactivation in human preimplantation embryos and pluripotent cells

    Nature 522:221–225.

    https://doi.org/10.1038/nature14308

    • Google Scholar
35. 1. Huda A
    2. Mariño-Ramírez L
    3. Jordan IK

    (2010) Epigenetic histone modifications of human transposable elements: genome defense versus exaptation

    Mobile DNA 1:2.

    https://doi.org/10.1186/1759-8753-1-2

    • Google Scholar
36. 1. Huntley S
    2. Baggott DM
    3. Hamilton AT
    4. Tran-Gyamfi M
    5. Yang S
    6. Kim J
    7. Gordon L
    8. Branscomb E
    9. Stubbs L

    (2006) A comprehensive catalog of human KRAB-associated zinc finger genes: insights into the evolutionary history of a large family of transcriptional repressors

    Genome Research 16:669–677.

    https://doi.org/10.1101/gr.4842106

    • PubMed
    • Google Scholar
37. 1. Imbeault M
    2. Helleboid P-Y
    3. Trono D

    (2017) KRAB zinc-finger proteins contribute to the evolution of gene regulatory networks

    Nature 543:550–554.

    https://doi.org/10.1038/nature21683

    • Google Scholar
38. 1. Lander ES
    2. Linton LM
    3. Birren B
    4. Nusbaum C
    5. Zody MC
    6. Baldwin J
    7. Devon K
    8. Dewar K
    9. Doyle M
    10. FitzHugh W
    11. Funke R
    12. Gage D
    13. Harris K
    14. Heaford A
    15. Howland J
    16. Kann L
    17. Lehoczky J
    18. LeVine R
    19. McEwan P
    20. McKernan K
    21. Meldrim J
    22. Mesirov JP
    23. Miranda C
    24. Morris W
    25. Naylor J
    26. Raymond C
    27. Rosetti M
    28. Santos R
    29. Sheridan A
    30. Sougnez C
    31. Stange-Thomann Y
    32. Stojanovic N
    33. Subramanian A
    34. Wyman D
    35. Rogers J
    36. Sulston J
    37. Ainscough R
    38. Beck S
    39. Bentley D
    40. Burton J
    41. Clee C
    42. Carter N
    43. Coulson A
    44. Deadman R
    45. Deloukas P
    46. Dunham A
    47. Dunham I
    48. Durbin R
    49. French L
    50. Grafham D
    51. Gregory S
    52. Hubbard T
    53. Humphray S
    54. Hunt A
    55. Jones M
    56. Lloyd C
    57. McMurray A
    58. Matthews L
    59. Mercer S
    60. Milne S
    61. Mullikin JC
    62. Mungall A
    63. Plumb R
    64. Ross M
    65. Shownkeen R
    66. Sims S
    67. Waterston RH
    68. Wilson RK
    69. Hillier LW
    70. McPherson JD
    71. Marra MA
    72. Mardis ER
    73. Fulton LA
    74. Chinwalla AT
    75. Pepin KH
    76. Gish WR
    77. Chissoe SL
    78. Wendl MC
    79. Delehaunty KD
    80. Miner TL
    81. Delehaunty A
    82. Kramer JB
    83. Cook LL
    84. Fulton RS
    85. Johnson DL
    86. Minx PJ
    87. Clifton SW
    88. Hawkins T
    89. Branscomb E
    90. Predki P
    91. Richardson P
    92. Wenning S
    93. Slezak T
    94. Doggett N
    95. Cheng JF
    96. Olsen A
    97. Lucas S
    98. Elkin C
    99. Uberbacher E
    100. Frazier M
    101. Gibbs RA
    102. Muzny DM
    103. Scherer SE
    104. Bouck JB
    105. Sodergren EJ
    106. Worley KC
    107. Rives CM
    108. Gorrell JH
    109. Metzker ML
    110. Naylor SL
    111. Kucherlapati RS
    112. Nelson DL
    113. Weinstock GM
    114. Sakaki Y
    115. Fujiyama A
    116. Hattori M
    117. Yada T
    118. Toyoda A
    119. Itoh T
    120. Kawagoe C
    121. Watanabe H
    122. Totoki Y
    123. Taylor T
    124. Weissenbach J
    125. Heilig R
    126. Saurin W
    127. Artiguenave F
    128. Brottier P
    129. Bruls T
    130. Pelletier E
    131. Robert C
    132. Wincker P
    133. Smith DR
    134. Doucette-Stamm L
    135. Rubenfield M
    136. Weinstock K
    137. Lee HM
    138. Dubois J
    139. Rosenthal A
    140. Platzer M
    141. Nyakatura G
    142. Taudien S
    143. Rump A
    144. Yang H
    145. Yu J
    146. Wang J
    147. Huang G
    148. Gu J
    149. Hood L
    150. Rowen L
    151. Madan A
    152. Qin S
    153. Davis RW
    154. Federspiel NA
    155. Abola AP
    156. Proctor MJ
    157. Myers RM
    158. Schmutz J
    159. Dickson M
    160. Grimwood J
    161. Cox DR
    162. Olson MV
    163. Kaul R
    164. Raymond C
    165. Shimizu N
    166. Kawasaki K
    167. Minoshima S
    168. Evans GA
    169. Athanasiou M
    170. Schultz R
    171. Roe BA
    172. Chen F
    173. Pan H
    174. Ramser J
    175. Lehrach H
    176. Reinhardt R
    177. McCombie WR
    178. de la Bastide M
    179. Dedhia N
    180. Blöcker H
    181. Hornischer K
    182. Nordsiek G
    183. Agarwala R
    184. Aravind L
    185. Bailey JA
    186. Bateman A
    187. Batzoglou S
    188. Birney E
    189. Bork P
    190. Brown DG
    191. Burge CB
    192. Cerutti L
    193. Chen HC
    194. Church D
    195. Clamp M
    196. Copley RR
    197. Doerks T
    198. Eddy SR
    199. Eichler EE
    200. Furey TS
    201. Galagan J
    202. Gilbert JG
    203. Harmon C
    204. Hayashizaki Y
    205. Haussler D
    206. Hermjakob H
    207. Hokamp K
    208. Jang W
    209. Johnson LS
    210. Jones TA
    211. Kasif S
    212. Kaspryzk A
    213. Kennedy S
    214. Kent WJ
    215. Kitts P
    216. Koonin EV
    217. Korf I
    218. Kulp D
    219. Lancet D
    220. Lowe TM
    221. McLysaght A
    222. Mikkelsen T
    223. Moran JV
    224. Mulder N
    225. Pollara VJ
    226. Ponting CP
    227. Schuler G
    228. Schultz J
    229. Slater G
    230. Smit AF
    231. Stupka E
    232. Szustakowki J
    233. Thierry-Mieg D
    234. Thierry-Mieg J
    235. Wagner L
    236. Wallis J
    237. Wheeler R
    238. Williams A
    239. Wolf YI
    240. Wolfe KH
    241. Yang SP
    242. Yeh RF
    243. Collins F
    244. Guyer MS
    245. Peterson J
    246. Felsenfeld A
    247. Wetterstrand KA
    248. Patrinos A
    249. Morgan MJ
    250. de Jong P
    251. Catanese JJ
    252. Osoegawa K
    253. Shizuya H
    254. Choi S
    255. Chen YJ
    256. Szustakowki J
    257. International Human Genome Sequencing Consortium

    (2001) Initial sequencing and analysis of the human genome

    Nature 409:860–921.

    https://doi.org/10.1038/35057062

    • PubMed
    • Google Scholar
39. 1. Jacobs FMJ
    2. Greenberg D
    3. Nguyen N
    4. Haeussler M
    5. Ewing AD
    6. Katzman S
    7. Paten B
    8. Salama SR
    9. Haussler D

    (2014) An evolutionary arms race between KRAB zinc-finger genes ZNF91/93 and SVA/L1 retrotransposons

    Nature 516:242–245.

    https://doi.org/10.1038/nature13760

    • Google Scholar
40. 1. Jacques Pierre-Étienne
    2. Jeyakani J
    3. Bourque G

    (2013) The majority of primate-specific regulatory sequences are derived from transposable elements

    PLoS Genetics 9:e1003504.

    https://doi.org/10.1371/journal.pgen.1003504

    • Google Scholar
41. 1. Jordan IK
    2. Rogozin IB
    3. Glazko GV
    4. Koonin EV

    (2003) Origin of a substantial fraction of human regulatory sequences from transposable elements

    Trends in Genetics 19:68–72.

    https://doi.org/10.1016/S0168-9525(02)00006-9

    • Google Scholar
42. 1. Jurka J

    (2000) Repbase Update: a database and an electronic journal of repetitive elements

    Trends in Genetics 16:418–420.

    https://doi.org/10.1016/S0168-9525(00)02093-X

    • Google Scholar
43. 1. Kapopoulou A
    2. Mathew L
    3. Wong A
    4. Trono D
    5. Jensen JD

    (2016) The evolution of gene expression and binding specificity of the largest transcription factor family in primates

    Evolution 70:167–180.

    https://doi.org/10.1111/evo.12819

    • Google Scholar
44. 1. Kapusta A
    2. Kronenberg Z
    3. Lynch VJ
    4. Zhuo X
    5. Ramsay L
    6. Bourque G
    7. Yandell M
    8. Feschotte C

    (2013) Transposable elements are major contributors to the origin, diversification, and regulation of vertebrate long noncoding RNAs

    PLoS Genetics 9:e1003470.

    https://doi.org/10.1371/journal.pgen.1003470

    • PubMed
    • Google Scholar
45. 1. Karimi MM
    2. Goyal P
    3. Maksakova IA
    4. Bilenky M
    5. Leung D
    6. Tang JX
    7. Shinkai Y
    8. Mager DL
    9. Jones S
    10. Hirst M
    11. Lorincz MC

    (2011) DNA methylation and SETDB1/H3K9me3 regulate predominantly distinct sets of genes, retroelements, and chimeric transcripts in mESCs

    Cell Stem Cell 8:676–687.

    https://doi.org/10.1016/j.stem.2011.04.004

    • PubMed
    • Google Scholar
46. 1. Karolchik D
    2. Hinrichs AS
    3. Furey TS
    4. Roskin KM
    5. Sugnet CW
    6. Haussler D
    7. Kent WJ

    (2004) The UCSC table browser data retrieval tool

    Nucleic Acids Research 32:493D–496.

    https://doi.org/10.1093/nar/gkh103

    • PubMed
    • Google Scholar
47. 1. Kazazian HH
    2. Wong C
    3. Youssoufian H
    4. Scott AF
    5. Phillips DG
    6. Antonarakis SE

    (1988) Haemophilia A resulting from de novo insertion of L1 sequences represents a novel mechanism for mutation in man

    Nature 332:164–166.

    https://doi.org/10.1038/332164a0

    • Google Scholar
48. 1. Kazazian HH

    (2004) Mobile elements: drivers of genome evolution

    Science 303:1626–1632.

    https://doi.org/10.1126/science.1089670

    • PubMed
    • Google Scholar
49. 1. Kelley D
    2. Rinn J

    (2012) Transposable elements reveal a stem cell-specific class of long noncoding RNAs

    Genome Biology 13:R107.

    https://doi.org/10.1186/gb-2012-13-11-r107

    • Google Scholar
50. 1. Kim D
    2. Pertea G
    3. Trapnell C
    4. Pimentel H
    5. Kelley R
    6. Salzberg SL

    (2013) TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions

    Genome Biology 14:R36.

    https://doi.org/10.1186/gb-2013-14-4-r36

    • PubMed
    • Google Scholar
51. 1. Klawitter S
    2. Fuchs NV
    3. Upton KR
    4. Muñoz-Lopez M
    5. Shukla R
    6. Wang J
    7. Garcia-Cañadas M
    8. Lopez-Ruiz C
    9. Gerhardt DJ
    10. Sebe A
    11. Grabundzija I
    12. Merkert S
    13. Gerdes P
    14. Pulgarin JA
    15. Bock A
    16. Held U
    17. Witthuhn A
    18. Haase A
    19. Sarkadi B
    20. Löwer J
    21. Wolvetang EJ
    22. Martin U
    23. Ivics Z
    24. Izsvák Z
    25. Garcia-Perez JL
    26. Faulkner GJ
    27. Schumann GG

    (2016) Reprogramming triggers endogenous L1 and Alu retrotransposition in human induced pluripotent stem cells

    Nature Communications 7:10286.

    https://doi.org/10.1038/ncomms10286

    • PubMed
    • Google Scholar
52. 1. Kunarso G
    2. Chia NY
    3. Jeyakani J
    4. Hwang C
    5. Lu X
    6. Chan YS
    7. Ng HH
    8. Bourque G

    (2010) Transposable elements have rewired the core regulatory network of human embryonic stem cells

    Nature Genetics 42:631–634.

    https://doi.org/10.1038/ng.600

    • PubMed
    • Google Scholar
53. 1. Law CW
    2. Chen Y
    3. Shi W
    4. Smyth GK

    (2014) voom: Precision weights unlock linear model analysis tools for RNA-seq read counts

    Genome Biology 15:R29.

    https://doi.org/10.1186/gb-2014-15-2-r29

    • PubMed
    • Google Scholar
54. 1. Leeb M
    2. Pasini D
    3. Novatchkova M
    4. Jaritz M
    5. Helin K
    6. Wutz A

    (2010) Polycomb complexes act redundantly to repress genomic repeats and genes

    Genes & Development 24:265–276.

    https://doi.org/10.1101/gad.544410

    • Google Scholar
55. 1. Li H
    2. Durbin R

    (2009) Fast and accurate short read alignment with Burrows-Wheeler transform

    Bioinformatics 25:1754–1760.

    https://doi.org/10.1093/bioinformatics/btp324

    • Google Scholar
56. 1. Li H
    2. Handsaker B
    3. Wysoker A
    4. Fennell T
    5. Ruan J
    6. Homer N
    7. Marth G
    8. Abecasis G
    9. Durbin R
    10. 1000 Genome Project Data Processing Subgroup

    (2009) The sequence alignment/map format and SAMtools

    Bioinformatics 25:2078–2079.

    https://doi.org/10.1093/bioinformatics/btp352

    • PubMed
    • Google Scholar
57. 1. Liao Y
    2. Smyth GK
    3. Shi W

    (2014) featureCounts: an efficient general purpose program for assigning sequence reads to genomic features

    Bioinformatics 30:923–930.

    https://doi.org/10.1093/bioinformatics/btt656

    • PubMed
    • Google Scholar
58. 1. Love MI
    2. Huber W
    3. Anders S

    (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2

    Genome Biology 15:550.

    https://doi.org/10.1186/s13059-014-0550-8

    • PubMed
    • Google Scholar
59. 1. Lu X
    2. Sachs F
    3. Ramsay L
    4. Jacques Pierre-Étienne
    5. Göke J
    6. Bourque G
    7. Ng H-H

    (2014) The retrovirus HERVH is a long noncoding RNA required for human embryonic stem cell identity

    Nature Structural & Molecular Biology 21:423–425.

    https://doi.org/10.1038/nsmb.2799

    • Google Scholar
60. 1. Lynch VJ
    2. Leclerc RD
    3. May G
    4. Wagner GP

    (2011) Transposon-mediated rewiring of gene regulatory networks contributed to the evolution of pregnancy in mammals

    Nature Genetics 43:1154–1159.

    https://doi.org/10.1038/ng.917

    • PubMed
    • Google Scholar
61. 1. Lynch VJ
    2. Nnamani MC
    3. Kapusta A
    4. Brayer K
    5. Plaza SL
    6. Mazur EC
    7. Emera D
    8. Sheikh SZ
    9. Grützner F
    10. Bauersachs S
    11. Graf A
    12. Young SL
    13. Lieb JD
    14. DeMayo FJ
    15. Feschotte C
    16. Wagner GP

    (2015) Ancient transposable elements transformed the uterine regulatory landscape and transcriptome during the evolution of mammalian pregnancy

    Cell Reports 10:551–561.

    https://doi.org/10.1016/j.celrep.2014.12.052

    • Google Scholar
62. 1. Macfarlan TS
    2. Gifford WD
    3. Driscoll S
    4. Lettieri K
    5. Rowe HM
    6. Bonanomi D
    7. Firth A
    8. Singer O
    9. Trono D
    10. Pfaff SL

    (2012) Embryonic stem cell potency fluctuates with endogenous retrovirus activity

    Nature 487:57–63.

    https://doi.org/10.1038/nature11244

    • Google Scholar
63. 1. Marchetto MCN
    2. Narvaiza I
    3. Denli AM
    4. Benner C
    5. Lazzarini TA
    6. Nathanson JL
    7. Paquola ACM
    8. Desai KN
    9. Herai RH
    10. Weitzman MD
    11. Yeo GW
    12. Muotri AR
    13. Gage FH

    (2013) Differential L1 regulation in pluripotent stem cells of humans and apes

    Nature 503:525–529.

    https://doi.org/10.1038/nature12686

    • Google Scholar
64. 1. Matsui T
    2. Leung D
    3. Miyashita H
    4. Maksakova IA
    5. Miyachi H
    6. Kimura H
    7. Tachibana M
    8. Lorincz MC
    9. Shinkai Y

    (2010) Proviral silencing in embryonic stem cells requires the histone methyltransferase ESET

    Nature 464:927–931.

    https://doi.org/10.1038/nature08858

    • Google Scholar
65. 1. Medstrand P
    2. Mager DL

    (1998)

    Human-specific integrations of the HERV-K endogenous retrovirus family

    J Virol 72:9782–9787.

    • Google Scholar
66. 1. Mozzetta C
    2. Boyarchuk E
    3. Pontis J
    4. Ait-Si-Ali S

    (2015) Sound of silence: the properties and functions of repressive Lys methyltransferases

    Nature Reviews Molecular Cell Biology 16:499–513.

    https://doi.org/10.1038/nrm4029

    • Google Scholar
67. 1. Mun S
    2. Lee J
    3. Kim Y-J
    4. Kim H-S
    5. Han K

    (2014) Chimpanzee-specific endogenous retrovirus generates genomic variations in the chimpanzee genome

    PLoS ONE 9:e101195.

    https://doi.org/10.1371/journal.pone.0101195

    • Google Scholar
68. 1. Müller FJ
    2. Schuldt BM
    3. Williams R
    4. Mason D
    5. Altun G
    6. Papapetrou EP
    7. Danner S
    8. Goldmann JE
    9. Herbst A
    10. Schmidt NO
    11. Aldenhoff JB
    12. Laurent LC
    13. Loring JF

    (2011) A bioinformatic assay for pluripotency in human cells

    Nature Methods 8:315–317.

    https://doi.org/10.1038/nmeth.1580

    • PubMed
    • Google Scholar
69. 1. Najafabadi HS
    2. Mnaimneh S
    3. Schmitges FW
    4. Garton M
    5. Lam KN
    6. Yang A
    7. Albu M
    8. Weirauch MT
    9. Radovani E
    10. Kim PM
    11. Greenblatt J
    12. Frey BJ
    13. Hughes TR

    (2015) C2H2 zinc finger proteins greatly expand the human regulatory lexicon

    Nature Biotechnology 33:555–562.

    https://doi.org/10.1038/nbt.3128

    • Google Scholar
70. 1. O'Geen H
    2. Squazzo SL
    3. Iyengar S
    4. Blahnik K
    5. Rinn JL
    6. Chang HY
    7. Green R
    8. Farnham PJ

    (2007) Genome-wide analysis of KAP1 binding suggests autoregulation of KRAB-ZNFs

    PLoS Genetics 3:e89.

    https://doi.org/10.1371/journal.pgen.0030089

    • PubMed
    • Google Scholar
71. 1. Ostertag EM
    2. Goodier JL
    3. Zhang Y
    4. Kazazian HH

    (2003) SVA elements are nonautonomous retrotransposons that cause disease in humans

    The American Journal of Human Genetics 73:1444–1451.

    https://doi.org/10.1086/380207

    • Google Scholar
72. 1. Prescott SL
    2. Srinivasan R
    3. Marchetto MC
    4. Grishina I
    5. Narvaiza I
    6. Selleri L
    7. Gage FH
    8. Swigut T
    9. Wysocka J

    (2015) Enhancer divergence and cis-regulatory evolution in the human and chimp neural crest

    Cell 163:68–83.

    https://doi.org/10.1016/j.cell.2015.08.036

    • PubMed
    • Google Scholar
73. 1. Quenneville S
    2. Turelli P
    3. Bojkowska K
    4. Raclot C
    5. Offner S
    6. Kapopoulou A
    7. Trono D

    (2012) The KRAB-ZFP/KAP1 system contributes to the early embryonic establishment of site-specific DNA methylation patterns maintained during development

    Cell Reports 2:766–773.

    https://doi.org/10.1016/j.celrep.2012.08.043

    • PubMed
    • Google Scholar
74. 1. Ramsay L
    2. Marchetto MC
    3. Caron M
    4. Chen S-H
    5. Busche S
    6. Kwan T
    7. Pastinen T
    8. Gage FH
    9. Bourque G

    (2017) Conserved expression of transposon-derived non-coding transcripts in primate stem cells

    BMC Genomics 18:214.

    https://doi.org/10.1186/s12864-017-3568-y

    • Google Scholar
75. 1. Richardson SR
    2. Gerdes P
    3. Gerhardt DJ
    4. Sanchez-Luque FJ
    5. Bodea G-O
    6. Muñoz-Lopez M
    7. Jesuadian JS
    8. Kempen M-JHC
    9. Carreira PE
    10. Jeddeloh JA
    11. Garcia-Perez JL
    12. Kazazian HH
    13. Ewing AD
    14. Faulkner GJ

    (2017) Heritable L1 retrotransposition in the mouse primordial germline and early embryo

    Genome Research 27:1395–1405.

    https://doi.org/10.1101/gr.219022.116

    • Google Scholar
76. 1. Roadmap Epigenomics Consortium
    2. Kundaje A
    3. Meuleman W
    4. Ernst J
    5. Bilenky M
    6. Yen A
    7. Heravi-Moussavi A
    8. Kheradpour P
    9. Zhang Z
    10. Wang J
    11. Ziller MJ
    12. Amin V
    13. Whitaker JW
    14. Schultz MD
    15. Ward LD
    16. Sarkar A
    17. Quon G
    18. Sandstrom RS
    19. Eaton ML
    20. Wu YC
    21. Pfenning AR
    22. Wang X
    23. Claussnitzer M
    24. Liu Y
    25. Coarfa C
    26. Harris RA
    27. Shoresh N
    28. Epstein CB
    29. Gjoneska E
    30. Leung D
    31. Xie W
    32. Hawkins RD
    33. Lister R
    34. Hong C
    35. Gascard P
    36. Mungall AJ
    37. Moore R
    38. Chuah E
    39. Tam A
    40. Canfield TK
    41. Hansen RS
    42. Kaul R
    43. Sabo PJ
    44. Bansal MS
    45. Carles A
    46. Dixon JR
    47. Farh KH
    48. Feizi S
    49. Karlic R
    50. Kim AR
    51. Kulkarni A
    52. Li D
    53. Lowdon R
    54. Elliott G
    55. Mercer TR
    56. Neph SJ
    57. Onuchic V
    58. Polak P
    59. Rajagopal N
    60. Ray P
    61. Sallari RC
    62. Siebenthall KT
    63. Sinnott-Armstrong NA
    64. Stevens M
    65. Thurman RE
    66. Wu J
    67. Zhang B
    68. Zhou X
    69. Beaudet AE
    70. Boyer LA
    71. De Jager PL
    72. Farnham PJ
    73. Fisher SJ
    74. Haussler D
    75. Jones SJ
    76. Li W
    77. Marra MA
    78. McManus MT
    79. Sunyaev S
    80. Thomson JA
    81. Tlsty TD
    82. Tsai LH
    83. Wang W
    84. Waterland RA
    85. Zhang MQ
    86. Chadwick LH
    87. Bernstein BE
    88. Costello JF
    89. Ecker JR
    90. Hirst M
    91. Meissner A
    92. Milosavljevic A
    93. Ren B
    94. Stamatoyannopoulos JA
    95. Wang T
    96. Kellis M

    (2015) Integrative analysis of 111 reference human epigenomes

    Nature 518:317–330.

    https://doi.org/10.1038/nature14248

    • PubMed
    • Google Scholar
77. 1. Robbez-Masson L
    2. Rowe HM

    (2015) Retrotransposons shape species-specific embryonic stem cell gene expression

    Retrovirology 12:45.

    https://doi.org/10.1186/s12977-015-0173-5

    • Google Scholar
78. 1. Rosenbloom KR
    2. Sloan CA
    3. Malladi VS
    4. Dreszer TR
    5. Learned K
    6. Kirkup VM
    7. Wong MC
    8. Maddren M
    9. Fang R
    10. Heitner SG
    11. Lee BT
    12. Barber GP
    13. Harte RA
    14. Diekhans M
    15. Long JC
    16. Wilder SP
    17. Zweig AS
    18. Karolchik D
    19. Kuhn RM
    20. Haussler D
    21. Kent WJ

    (2013) ENCODE data in the UCSC Genome Browser: year 5 update

    Nucleic Acids Research 41:D56–D63.

    https://doi.org/10.1093/nar/gks1172

    • PubMed
    • Google Scholar
79. 1. Rowe HM
    2. Friedli M
    3. Offner S
    4. Verp S
    5. Mesnard D
    6. Marquis J
    7. Aktas T
    8. Trono D

    (2013) De novo DNA methylation of endogenous retroviruses is shaped by KRAB-ZFPs/KAP1 and ESET

    Development 140:519–529.

    https://doi.org/10.1242/dev.087585

    • PubMed
    • Google Scholar
80. 1. Rowe HM
    2. Jakobsson J
    3. Mesnard D
    4. Rougemont J
    5. Reynard S
    6. Aktas T
    7. Maillard PV
    8. Layard-Liesching H
    9. Verp S
    10. Marquis J
    11. Spitz F
    12. Constam DB
    13. Trono D

    (2010) KAP1 controls endogenous retroviruses in embryonic stem cells

    Nature 463:237–240.

    https://doi.org/10.1038/nature08674

    • Google Scholar
81. 1. Rowe HM
    2. Trono D

    (2011) Dynamic control of endogenous retroviruses during development

    Virology 411:273–287.

    https://doi.org/10.1016/j.virol.2010.12.007

    • PubMed
    • Google Scholar
82. 1. Santoni FA
    2. Guerra J
    3. Luban J

    (2012) HERV-H RNA is abundant in human embryonic stem cells and a precise marker for pluripotency

    Retrovirology 9:111.

    https://doi.org/10.1186/1742-4690-9-111

    • PubMed
    • Google Scholar
83. 1. Schlesinger S
    2. Goff SP

    (2015) Retroviral transcriptional regulation and embryonic stem cells: war and peace

    Molecular and Cellular Biology 35:770–777.

    https://doi.org/10.1128/MCB.01293-14

    • PubMed
    • Google Scholar
84. 1. Schmidt D
    2. Schwalie PC
    3. Wilson MD
    4. Ballester B
    5. Gonçalves A
    6. Kutter C
    7. Brown GD
    8. Marshall A
    9. Flicek P
    10. Odom DT

    (2012) Waves of retrotransposon expansion remodel genome organization and CTCF binding in multiple mammalian lineages

    Cell 148:335–348.

    https://doi.org/10.1016/j.cell.2011.11.058

    • PubMed
    • Google Scholar
85. 1. Schmidt D
    2. Wilson MD
    3. Spyrou C
    4. Brown GD
    5. Hadfield J
    6. Odom DT

    (2009) ChIP-seq: Using high-throughput sequencing to discover protein–DNA interactions

    Methods 48:240–248.

    https://doi.org/10.1016/j.ymeth.2009.03.001

    • Google Scholar
86. 1. Schmitges FW
    2. Radovani E
    3. Najafabadi HS
    4. Barazandeh M
    5. Campitelli LF
    6. Yin Y
    7. Jolma A
    8. Zhong G
    9. Guo H
    10. Kanagalingam T
    11. Dai WF
    12. Taipale J
    13. Emili A
    14. Greenblatt JF
    15. Hughes TR

    (2016) Multiparameter functional diversity of human C2H2 zinc finger proteins

    Genome Research 26:1742–1752.

    https://doi.org/10.1101/gr.209643.116

    • Google Scholar
87. 1. Schultz DC
    2. Ayyanathan K
    3. Negorev D
    4. Maul GG
    5. Rauscher FJ

    (2002) SETDB1: a novel KAP-1-associated histone H3, lysine 9-specific methyltransferase that contributes to HP1-mediated silencing of euchromatic genes by KRAB zinc-finger proteins

    Genes & Development 16:919–932.

    https://doi.org/10.1101/gad.973302

    • PubMed
    • Google Scholar
88. 1. Schwalie PC
    2. Ward MC
    3. Cain CE
    4. Faure AJ
    5. Gilad Y
    6. Odom DT
    7. Flicek P

    (2013) Co-binding by YY1 identifies the transcriptionally active, highly conserved set of CTCF-bound regions in primate genomes

    Genome Biology 14:R148.

    https://doi.org/10.1186/gb-2013-14-12-r148

    • Google Scholar
89. 1. Simonti CN
    2. Pavlicev M
    3. Capra JA

    (2017) Transposable element exaptation into regulatory regions is rare, influenced by evolutionary age, and subject to pleiotropic constraints

    Molecular Biology and Evolution 34:2856–2869.

    https://doi.org/10.1093/molbev/msx219

    • PubMed
    • Google Scholar
90. 1. Slotkin RK
    2. Martienssen R

    (2007) Transposable elements and the epigenetic regulation of the genome

    Nature Reviews Genetics 8:272–285.

    https://doi.org/10.1038/nrg2072

    • PubMed
    • Google Scholar
91. 1. Smith ZD
    2. Chan MM
    3. Humm KC
    4. Karnik R
    5. Mekhoubad S
    6. Regev A
    7. Eggan K
    8. Meissner A

    (2014) DNA methylation dynamics of the human preimplantation embryo

    Nature 511:611–615.

    https://doi.org/10.1038/nature13581

    • PubMed
    • Google Scholar
92. 1. Smyth GK

    (2004) Linear models and empirical bayes methods for assessing differential expression in microarray experiments

    Statistical Applications in Genetics and Molecular Biology 3:1–25.

    https://doi.org/10.2202/1544-6115.1027

    • PubMed
    • Google Scholar
93. 1. Speir ML
    2. Zweig AS
    3. Rosenbloom KR
    4. Raney BJ
    5. Paten B
    6. Nejad P
    7. Lee BT
    8. Learned K
    9. Karolchik D
    10. Hinrichs AS
    11. Heitner S
    12. Harte RA
    13. Haeussler M
    14. Guruvadoo L
    15. Fujita PA
    16. Eisenhart C
    17. Diekhans M
    18. Clawson H
    19. Casper J
    20. Barber GP
    21. Haussler D
    22. Kuhn RM
    23. Kent WJ

    (2016) The UCSC genome browser database: 2016 update

    Nucleic Acids Research 44:D717–D725.

    https://doi.org/10.1093/nar/gkv1275

    • PubMed
    • Google Scholar
94. 1. Stephens M

    (2017)

    False discovery rates: a new deal

    Biostatistics 18:275–294.

    • Google Scholar
95. 1. Sundaram V
    2. Choudhary MN
    3. Pehrsson E
    4. Xing X
    5. Fiore C
    6. Pandey M
    7. Maricque B
    8. Udawatta M
    9. Ngo D
    10. Chen Y
    11. Paguntalan A
    12. Ray T
    13. Hughes A
    14. Cohen BA
    15. Wang T

    (2017) Functional cis-regulatory modules encoded by mouse-specific endogenous retrovirus

    Nature Communications 8:14550.

    https://doi.org/10.1038/ncomms14550

    • PubMed
    • Google Scholar
96. 1. Thomas JH
    2. Schneider S

    (2011) Coevolution of retroelements and tandem zinc finger genes

    Genome Research 21:1800–1812.

    https://doi.org/10.1101/gr.121749.111

    • PubMed
    • Google Scholar
97. 1. Tönjes RR
    2. Löwer R
    3. Boller K
    4. Denner J
    5. Hasenmaier B
    6. Kirsch H
    7. König H
    8. Korbmacher C
    9. Limbach C
    10. Lugert R
    11. Phelps RC
    12. Scherer J
    13. Thelen K
    14. Löwer J
    15. Kurth R

    (1996) HERV-K: the biologically most active human endogenous retrovirus family

    Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology 13 Suppl 1:S261–S267.

    https://doi.org/10.1097/00042560-199600001-00039

    • PubMed
    • Google Scholar
98. 1. Trizzino M
    2. Park Y
    3. Holsbach-Beltrame M
    4. Aracena K
    5. Mika K
    6. Caliskan M
    7. Perry GH
    8. Lynch VJ
    9. Brown CD

    (2017) Transposable elements are the primary source of novelty in primate gene regulation

    Genome Research 27:1623–1633.

    https://doi.org/10.1101/gr.218149.116

    • PubMed
    • Google Scholar
99. 1. Turelli P
    2. Castro-Diaz N
    3. Marzetta F
    4. Kapopoulou A
    5. Raclot C
    6. Duc J
    7. Tieng V
    8. Quenneville S
    9. Trono D

    (2014) Interplay of TRIM28 and DNA methylation in controlling human endogenous retroelements

    Genome Research 24:1260–1270.

    https://doi.org/10.1101/gr.172833.114

    • PubMed
    • Google Scholar
100. 1. van der Kuyl AC

     (2012) HIV infection and HERV expression: a review

     Retrovirology 9:6.

     https://doi.org/10.1186/1742-4690-9-6

     • PubMed
     • Google Scholar
101. 1. Walter M
     2. Teissandier A
     3. Pérez-Palacios R
     4. Bourc'his D

     (2016) An epigenetic switch ensures transposon repression upon dynamic loss of DNA methylation in embryonic stem cells

     eLife 5:e11418.

     https://doi.org/10.7554/eLife.11418

     • PubMed
     • Google Scholar
102. 1. Wang H
     2. Xing J
     3. Grover D
     4. Hedges DJ
     5. Han K
     6. Walker JA
     7. Batzer MA

     (2005) SVA elements: a hominid-specific retroposon family

     Journal of Molecular Biology 354:994–1007.

     https://doi.org/10.1016/j.jmb.2005.09.085

     • PubMed
     • Google Scholar
103. 1. Wang J
     2. Xie G
     3. Singh M
     4. Ghanbarian AT
     5. Raskó T
     6. Szvetnik A
     7. Cai H
     8. Besser D
     9. Prigione A
     10. Fuchs NV
     11. Schumann GG
     12. Chen W
     13. Lorincz MC
     14. Ivics Z
     15. Hurst LD
     16. Izsvák Z

     (2014) Primate-specific endogenous retrovirus-driven transcription defines naive-like stem cells

     Nature 516:405–409.

     https://doi.org/10.1038/nature13804

     • PubMed
     • Google Scholar
104. 1. Wang T
     2. Zeng J
     3. Lowe CB
     4. Sellers RG
     5. Salama SR
     6. Yang M
     7. Burgess SM
     8. Brachmann RK
     9. Haussler D

     (2007) Species-specific endogenous retroviruses shape the transcriptional network of the human tumor suppressor protein p53

     PNAS 104:18613–18618.

     https://doi.org/10.1073/pnas.0703637104

     • PubMed
     • Google Scholar
105. 1. Ward MC
     2. Wilson MD
     3. Barbosa-Morais NL
     4. Schmidt D
     5. Stark R
     6. Pan Q
     7. Schwalie PC
     8. Menon S
     9. Lukk M
     10. Watt S
     11. Thybert D
     12. Kutter C
     13. Kirschner K
     14. Flicek P
     15. Blencowe BJ
     16. Odom DT

     (2013) Latent regulatory potential of human-specific repetitive elements

     Molecular Cell 49:262–272.

     https://doi.org/10.1016/j.molcel.2012.11.013

     • PubMed
     • Google Scholar
106. 1. Wildschutte JH
     2. Williams ZH
     3. Montesion M
     4. Subramanian RP
     5. Kidd JM
     6. Coffin JM

     (2016) Discovery of unfixed endogenous retrovirus insertions in diverse human populations

     PNAS 113:E2326–E2334.

     https://doi.org/10.1073/pnas.1602336113

     • PubMed
     • Google Scholar
107. 1. Wilson MD
     2. Barbosa-Morais NL
     3. Schmidt D
     4. Conboy CM
     5. Vanes L
     6. Tybulewicz VL
     7. Fisher EM
     8. Tavaré S
     9. Odom DT

     (2008) Species-specific transcription in mice carrying human chromosome 21

     Science 322:434–438.

     https://doi.org/10.1126/science.1160930

     • PubMed
     • Google Scholar
108. 1. Wissing S
     2. Muñoz-Lopez M
     3. Macia A
     4. Yang Z
     5. Montano M
     6. Collins W
     7. Garcia-Perez JL
     8. Moran JV
     9. Greene WC

     (2012) Reprogramming somatic cells into iPS cells activates LINE-1 retroelement mobility

     Human Molecular Genetics 21:208–218.

     https://doi.org/10.1093/hmg/ddr455

     • PubMed
     • Google Scholar
109. 1. Wolf D
     2. Goff SP

     (2009) Embryonic stem cells use ZFP809 to silence retroviral DNAs

     Nature 458:1201–1204.

     https://doi.org/10.1038/nature07844

     • PubMed
     • Google Scholar
110. 1. Xie M
     2. Hong C
     3. Zhang B
     4. Lowdon RF
     5. Xing X
     6. Li D
     7. Zhou X
     8. Lee HJ
     9. Maire CL
     10. Ligon KL
     11. Gascard P
     12. Sigaroudinia M
     13. Tlsty TD
     14. Kadlecek T
     15. Weiss A
     16. O'Geen H
     17. Farnham PJ
     18. Madden PA
     19. Mungall AJ
     20. Tam A
     21. Kamoh B
     22. Cho S
     23. Moore R
     24. Hirst M
     25. Marra MA
     26. Costello JF
     27. Wang T

     (2013) DNA hypomethylation within specific transposable element families associates with tissue-specific enhancer landscape

     Nature Genetics 45:836–841.

     https://doi.org/10.1038/ng.2649

     • PubMed
     • Google Scholar
111. 1. Yohn CT
     2. Jiang Z
     3. McGrath SD
     4. Hayden KE
     5. Khaitovich P
     6. Johnson ME
     7. Eichler MY
     8. McPherson JD
     9. Zhao S
     10. Pääbo S
     11. Eichler EE

     (2005) Lineage-specific expansions of retroviral insertions within the genomes of African great apes but not humans and orangutans

     PLoS Biology 3:e110.

     https://doi.org/10.1371/journal.pbio.0030110

     • PubMed
     • Google Scholar
112. 1. Zhang Y
     2. Liu T
     3. Meyer CA
     4. Eeckhoute J
     5. Johnson DS
     6. Bernstein BE
     7. Nusbaum C
     8. Myers RM
     9. Brown M
     10. Li W
     11. Liu XS

     (2008) Model-based analysis of ChIP-Seq (MACS)

     Genome Biology 9:R137.

     https://doi.org/10.1186/gb-2008-9-9-r137

     • PubMed
     • Google Scholar
113. 1. Zhou X
     2. Cain CE
     3. Myrthil M
     4. Lewellen N
     5. Michelini K
     6. Davenport ER
     7. Stephens M
     8. Pritchard JK
     9. Gilad Y

     (2014) Epigenetic modifications are associated with inter-species gene expression variation in primates

     Genome Biology 15:547.

     https://doi.org/10.1186/s13059-014-0547-3

     • PubMed
     • Google Scholar

Author details

1. Michelle C Ward

   1. Department of Human Genetics, University of Chicago, Chicago, United States
   2. Department of Medicine, University of Chicago, Chicago, United States

   Contribution

   Conceptualization, Formal analysis, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   mcward@uchicago.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1485-320X

2. Siming Zhao

   Department of Human Genetics, University of Chicago, Chicago, United States

   Contribution

   Formal analysis

   Competing interests

   No competing interests declared

3. Kaixuan Luo

   Department of Human Genetics, University of Chicago, Chicago, United States

   Contribution

   Formal analysis, Writing—review and editing

   Competing interests

   No competing interests declared

4. Bryan J Pavlovic

   Department of Human Genetics, University of Chicago, Chicago, United States

   Contribution

   Resources

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7751-5315

5. Mohammad M Karimi

   MRC London Institute of Medical Sciences, Imperial College, London, United Kingdom

   Contribution

   Formal analysis, Writing—review and editing

   Competing interests

   No competing interests declared

6. Matthew Stephens

   1. Department of Human Genetics, University of Chicago, Chicago, United States
   2. Department of Statistics, University of Chicago, Chicago, United States

   Contribution

   Formal analysis, Supervision, Writing—review and editing

   Competing interests

   No competing interests declared

7. Yoav Gilad

   1. Department of Human Genetics, University of Chicago, Chicago, United States
   2. Department of Medicine, University of Chicago, Chicago, United States

   Contribution

   Conceptualization, Formal analysis, Supervision, Funding acquisition, Writing—original draft, Writing—review and editing

   For correspondence

   gilad@uchicago.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8284-8926

• 5,176

  views

• 398

  downloads

• 24

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.