Table 1.

ChIP studies comparing binding events between species, strains or individuals. This table is an extension of the data compiled by Dowell [39]. We added studies that compared histone modifications, the tissue or cell type/line being compared and several notes. Evolutionary distances between organisms were obtained from Hedges et al. [40] except where noted. Distance is given in relation to the first species. D. mel, D. melanogaster; D. sim, D. simulans.

organisms compared	tissue/cell line or type	transcription factor/histone modification	conservation level of binding/histone modification region or putatively bound genes	reference conservation level of binding/histone modification region or putatively bound genesa	notes	reference
D. melanogaster, D. yakuba (6.5 Myr)	whole embryo (collected within 1 h preceding gastrulation)	bcd (bicoid) hb (hunchback) Kr (Kruppel) gt (giant) kni (knirps) cad (caudal)	98–99% 85–97% 97–99% 98–99% 97–100% 98–99%		— binding intensities presented differences for all six TFs — stronger binding events are less conserved — suggested an epigenetic cause for correlated binding intensity differences	[26]
D. melanogaster, D. annanassae (44.2 Myr), D. erecta (12.8 Myr), D. pseudoobscura (33.6 Myr), D. yakuba (6.5 Myr)	whole embryo (approx. 2–4 h after egg laying)	twi (twist) sna (snail)	approximately 80% (D. mel versus D. sim and yakuba) 57–65% (D. mel versus all other) 34% shared across all 6 species approximately 91% (D. mel versus D. sim)	98% D. mel twi biological replicate 88% D. mel sna biological replicate	— clustering using the differences among binding maps recapitulated the known phylogenetic tree — observed differences in binding intensities — conservation of binding is higher for genes with functions related to the known role of twi — approximately 50% of binding events assigned to genes downregulated in twi knockouts are conserved, whereas conservation is lower than average for binding events near genes unresponsive to twi knockout.	[41]
S. cerevisiae, S. mikatae (5–20 Myr [42]), S. bayanus (14.3 Myr)	all species kept under low-nitrogen condition (pseudo-hyphal growth)	Ste12 Tec1	21% 20% (genes putatively bound in three species)	97% 95% (genes shared between biological replicates)	— observed differences in binding intensity — different functions for conserved and species-specific putatively bound genes	[14]
two yeast strains: S. cerevisiae strains S96 and HS959	grown in presence and absence of mating pheromone α-factor	Ste12	approximately 70% of the genes between S96 and HS959	>0.96 mean Pearson correlation between biological replicates, <0.89 between strains	— approximately 78% of the variable binding regions exhibit Mendelian segregation — several cases of transgression were noted — 35% of the motifs in cis-variable binding traits were affected by variation versus 1% of the non-variable	[25]
S. cerevisiae, K. lactis (>100 Myr [43]), C. albicans (489.8 Myr)	—	Mcm1	7–42% (genes), 13–18% (genes in three species)	duplicates were pooled	— genes bound in all three species are enriched for cell cycle and mating type. Species-specific targets have new functions	[15]
six human individuals (one trio from CEU and one from YRI)	lymphoblastoid cell lines	CTCF	82% (all six individuals), approximately 99% within populations		— 11% of the sites are allele-specific	[24]
10 human individuals	lymphoblastoid cell lines	NFKB Pol II	92.5% 75% NFKB: 94% TSSb, 92% non-TSS Pol II: 75% TSSb, 72% non-TSS	Pol II 68% between humans and chimpanzee	— Pol II binding events: 79% Mendelian, 5% transgression. NKFB binding events: 68% Mendelian, 4% transgression — 35% of the NFKB and 26% Pol II divergent binding events coincided with genetic variations — Spearman correlation between gene expression and NFKB: 0.475 and Pol II: 0.461 — stronger binding events are more often conserved — distinct functions associated with common versus species-specific events	[23]
human, mouse (mutant mouse contains a human chromosome 21) (92.3 Myr)	liver	HNF1A HNF4A HNF6 H3K4me3	18% 34% 14% 91% (TSS), 32% (non-TSS)	— mouse chromosome 16 in mutant versus wild-type mouse: 95–100% — human chromosome 21 in mouse versus in human: 86% HNF1A 82% HNF4A 74% HNF6 86% H3K4me3 (TSS) 78% H4K4me3 (non-TSS)	— used a Down syndrome mouse model to control for experimental and species-specific differences	[16]
human, mouse (92.3 Myr)	liver	FOXA2 HNF1A HNF4A HNF6	11–45% (genes) 18–20% (genes) 31–59% (genes) 13–27% (genes)	HNF6: 66% of the genes bound in human liver were bound in HepG2		[17]
human, mouse (92.3 Myr)	ESCs	OCT4 NANOG CTCF	2%, 3.8%c 1.9%, 5.3%c 16.7%, 49.6%c		— 11/137 genes downregulated in OCT4 knockouts in mouse and human had a nearby OCT4–NANOG binding event — 62/137 genes were cases of turnover — 55/137 of these genes had no binding events — 82-fold enrichment for overlap of OCT4 sites with LTR9B repeats — binding events overlapping repeats: 20.9% for OCT4, 14.6% for NANOG and 11.1% for CTCF in human and 7.2%, 17.1% and 28.3% in mouse — two LTR9B (repeat) binding regions were validated as enhancers	[18]
human, mouse (92.3 Myr)	ESCs	OCT4 NANOG	9.1% (genes) 13% (genes) — 14% of OCT4 peaks <10 kb of the TSS — 21% of NANOG peaks <10 kb of the TSS		— data compared were generated by different studies and platforms (human data from [44])	[19]
human, mouse (92.3 Myr)	human liver, HepG2, islets, acinar; mouse liver, spleen, kidney, brain, testes, pancreatic islets, pancreatic acinar and Min6	E2F4	20% (genes)	genes overlapping among mouse tissues >65–85% of the genes putatively bound; among human tissues: 70–84%	— did not find a relationship between genes downregulated upon E2F4 knockout — approximately 50 genes with nearby E2F4 binding conserved, GO enrichment for cell cycle, proliferation and DNA repair functions. DNA packaging enriched in mouse-only binding events	[20]
human, mouse (92.3 Myr)	human adipose stromal cells versus 3T3-L1 mouse adipocytes	CTCF PPARG H3K4me3, H3K4me1, H3K4me2, H3K27ac	approximately 53% (L1) 21% (L1) approximately 15–30% (>2kb from TSS)		— species-specific histone marks were associated with species-specific expression — histone conservation was higher near the TSS and increased with enrichment of the peaks — conservation near TSS decreases with enrichment for PPARG and CTCF — cases of turnover observed	[21]
human, mouse (92.3 Myr), chicken (296 Myr)	human CD4+, HeLa and Jurkat cells; mouse ESCs and embryonic fibroblasts; chicken 5- and 10-day-old red blood cells	CTCF	7% of the chicken sites (or 2% of the mouse sites) found in both red blood cell are conserved in all three species	conservation within species (% of the largest set): mouse tissues: 27% human cells: 36% chicken: 16%	— only sites that were present in all cells of each species and across species were considered	[45]
human, chimpanzee (6.3 Myr)	lymphoblastoid cell lines	H3K4me3	69.5±0.3% (human–chimpanzee) 63.9±0.4% (human–rhesus) 63.2±0.3% (chimpanzee–rhesus)	overlap within the same species: 77.7±0.4%	— estimated that at most 7% of gene expression differences between human, chimpanzee and 3% between chimpanzee and rhesus correlated with differences in H3K4me3 — conservation was higher for peaks <1kb of the TSS	[46]
human, mouse (92.3 Myr)	various	H3K4me1 H3K4me3 H3K27ac	approximately 20–40% approximately 80% TSSb, >approximately 10–40% non-TSS approximately 80% TSSb, >approximately 20–40% non-TSS		— data compared were generated by different studies and platforms — conserved fraction was higher near TSS — conservation increased when regions were present in multiple cell types — fraction of conserved regions increased with number of transcription factors bound	[22]
human, mouse (92.3 Myr), dog (94.2 Myr), opossum (162.6 Myr), chicken (296 Myr)	liver	CEBPA	2% human–chicken 7% human–opossum 13% human–mouse 0.3% (all five species)	71% human–human 66% mouse–mouse 61% dog–dog 57% opossum–opossum 56% chicken–chicken	— binding events conserved in at least two species are more likely to occur near genes that are differentially expressed upon knockout of CEPBA/HNF4A — no correlation between binding intensity and conservation — observed many cases of turnover events — 20–40% of the motifs located in species-specific binding events were intact. A ‘larger fraction’ of the motifs were found to be disrupted	[47]
human, mouse (92.3 Myr), dog (94.2 Myr)	liver	HNF4A	12% human–mouse 9% human–dog	72% human–human 77% mouse–mouse 63% dog–dog		[47]
human, chimpanzee (6.3 Myr)	lymphoblastoid cell lines	Pol II	68%		— species-specific events were associated with different functions than conserved events — higher conservation for stronger Pol II binding events	[23]

^aData in this column can be used to assess the significance of the divergence shown in the previous column. When available, the data reports the conservation level in a ‘reference’ group, such as between biological replicates or between different tissues of the same species, or different species when the previous column refers to a comparison within the same species. For example, a 20% conservation level between species is not surprising if replicates shown 20% conservation only.

^b<1kb of the TSS.

^cTop 10% binding events.