Figure 5.

Epigenome-wide association study of pair bonding status from prairie voles (Microtus ochrogaster). The association of DNAm and pair bonding was examined by a multivariate regression model for each tissue, adjusting for chronological age as a covariate, and stratified by sex. All voles younger than 0.3 years were sex-naive and were therefore restricted from these analyses. Adult sex-naive females and males (> 0.3 years old) were considered as the reference for comparison with their pair-bonded counterparts. (A) Manhattan plots of the EWAS of pair bonding status. Significant CpGs (nominal two-sided p < 0.005, red dotted line) are colored in red (methylation increase in pair-bonded animals compared to sex-naive animals) or blue (methylation decrease in pair-bonded animals compared to sex-naive animals). The top 15 CpGs are labeled by the neighboring genes. The x-axis reports genome coordinates based on the alignment of mammalian array probes to the MicOch1.0.100 genome assembly. (B) Location of top CpGs in each tissue relative to the closest transcriptional start site. The odds ratio of the proportion changes than the background is reported in for each bar. Fisher exact p values: *p < 0.05, **p  < 0.01, ***p < 0.001, ****p  < 0.0001. The number of selected CpGs: female blood, 465 CpGs; male blood, 569 CpGs; female brain, 119; male brain, 62 CpGs; female ear, 481 CpGs; male ear, 137 CpGs; female liver, 289 CpGs; male liver, 59 CpGs. (C) Upset plot representing the overlap of EWAS of pair bonding in different tissues and sexes. Neighboring genes of the overlapping CpGs were labeled in the figure. (D) Chromatin state enrichment of the top CpGs related to pair bonding. The nominal two-sided enrichment p-values were calculated with Fisher's hypergeometric test. The universal chromatin states⁶⁰ are based on the stackHMM states in human Hg19 genome. The PRC2 binding is defined by the available ENCODE ChipSeq results of EED, EZH2, and SUZ12 transcriptional factors.