In mammalian societies, dominance hierarchies translate into inequalities in health, reproductive performance and survival. DNA methylation is thought to mediate the effects of social status on gene expression and phenotypic outcomes, yet, a study of social status-specific DNA methylation profiles in different age classes in a wild social mammal is missing. We tested for social status signatures in DNA methylation profiles in wild female spotted hyenas (Crocuta crocuta) in cubs and adults using non-invasively collected gut epithelium samples. In spotted hyena clans, female social status influences access to resources, foraging behaviour, health, reproductive performance and survival. We identified 149 differentially methylated regions between 42 high- and low-ranking spotted hyena females (cubs and adults). Differentially methylated genes were associated with energy conversion, immune function and glutamate receptor signalling. Our results provide evidence that socio-environmental inequalities are reflected at the molecular level in cubs and adults in a wild mammal.

In the Serengeti National Park in Tanzania, we collected gut epithelium cells from faecal samples of 42 free-ranging female hyenas (24 cubs and 18 adults) of either low (15 cubs, 9 adults) or high (9 cubs, 9 adults) social status. In spotted hyenas, social status is determined by the animals’ interactions within their social group (clan). We extracted DNA and RNA from each sample. Epithelium cells contain mainly one cell-type, reducing the effect of cell-type specific methylation. We first performed whole genome bisulfite sequencing (WGBS), but mapping yielded in up to 99% bacterial data and parasitic DNA and the analysis with MACAU did not give coherent and in trustworthy results (Supplementary Table 7). Using methylation capture targeting CpG-methylation (with MBD2, MethylMiner, Invitrogen), we enriched methylated DNA fragments of hyenas. After Illumina sequencing (HiSeqX 150 bp PE, Macrogen), we mapped the quality-filtered reads to the novel spotted hyena reference sequence⁵⁰ with the bowtie2 mapper⁷⁴. We identified rank-associated differentially methylated regions (DMRs), by splitting the genome in 300bp windows (selecting windows with 10 reads per position and a minimum of 10 different individuals) and fitting a negative-binomial model using the EdgeR package⁷⁵ with age class (cub, adult) and social status class (high-ranking, low-ranking) as fixed effects. We estimated the variance using the weighted likelihood empirical Bayes method. P-values were computed with Likelihood Ratio Tests, and False Discovery Rate (FDR) controlled with the Benjamini-Hochberg procedure⁷⁶. Windows with a FDR ≤ 0.05 were considered significantly different and adjacent significant windows were merged to create DMRs (‘rankDMRs’). We annotated the rankDMRs by using the striped hyena annotation (Hyena hyena, RefSeq- Assembly ID: GCF_003009895.1) and by adding Transcription Start Sites (TSS) and promoter regions. We identified coherent DMRs using a subsampling procedure and estimated the discriminative power of rankDMRs and mean methylation using a random-forest approach⁷⁷. Further details for all experimental, methodical and statistical and reproducibility can be found in Supplementary Material. We conducted this study under research permits from the Tanzania Commission for Science and Technology and permission from the Tanzanian National Parks Authority and Tanzanian Wildlife Research Institute. All procedures were performed in accordance with the Leibniz Institute for Zoo and Wildlife Research Ethics Committee on Animal Welfare (permit number: 2017-11-02).