Analysis and annotation of genome-wide DNA methylation patterns in two nonhuman primate species using the Infinium Human Methylation 450K and EPIC BeadChips

Abstract

The Infinium Human Methylation450 and Methylation EPIC BeadChips are useful tools for the study of the methylation state of hundreds of thousands of CpG across the human genome at affordable cost. However, in a wide range of experimental settings in particular for studies in infectious or brain-related diseases, human samples cannot be easily obtained. Hence, due to their close developmental, immunological and neurological proximity with humans, nonhuman primates are used in many research fields of human diseases and for preclinical research. Few studies have used DNA methylation microarrays in simian models. Microarrays designed for the analysis of DNA methylation patterns in the human genome could be useful given the genomic proximity between human and nonhuman primates. However, there is currently information lacking about the specificity and usability of each probe for many nonhuman primate species, including rhesus macaques (Macaca mulatta), originating from Asia, and African green monkeys originating from West-Africa (Chlorocebus sabaeus).Rhesus macaques and African green monkeys are among the major nonhuman primate models utilized in biomedical research. Here, we provide a precise evaluation and reannotation of the probes of the two microarrays for the analysis of genome-wide DNA methylation patterns in these two Cercopithecidae species. We demonstrate that up to 162,000 of the 450K and 255,000 probes of the EPIC BeadChip can be reliably used in Macaca mulatta or Chlorocebus sabaeus. The annotation files are provided in a format compatible with a variety of preprocessing, normalization and analytical pipelines designed for data analysis from 450K/EPIC arrays, facilitating high-throughput DNA methylation analyses in Macaca mulatta and Chlorocebus sabaeus. They provide the opportunity to the research community to focus their analysis only on those probes identified as reliable. The described analytical workflow leaves the choice to the user to balance coverage versus specificity and can also be applied to other Cercopithecidae species.

DNA methylation changes in metabolic and immune-regulatory pathways in blood and lymph node CD4 + T cells in response to SIV infections

“…One potential limitation of this study is that human microarrays were used to measure DNA methylation on non-human primate genomes. However, alignment and stringent quality control of the included probes to the AGM and MAC genomes revealed that still 162,053 and 158,970 probes could be confidently used covering 16,295 and 16,611 genes, respectively [38]. Of those, 157,164 and 154,108 probes were detected with high confidence in the current study.…”

“…Using the Infinium HumanMethylation450K Bead-Chip (Illumina) by restricting the analysis to a subset of probes that target CpGs at a unique location on the MAC and AGM genomes, we were able to study methylation at 157,164 and 154,108 CpG sites for AGM and MAC, respectively [38]. As expected, genome-wide methylation pattern showed a typical bimodal distribution (Additional file 1: Figure 4A) and a strong correlation was observed when all probes were compared between MAC and AGM (Additional file 1: Figure 4B).…”

“…Since the 450K BeadChip is designed for the analysis of human samples, probes compatible with the Macaca mulatta and Chlorocebus sabaeus genomes were identified, yielding > 160,000 of the 450 K probes targeting functional CpGs in the respective genomes with a maximum of a single mismatch underlying the probe sequence with the exception of the last 5 bases from the 3′ end. This resulted in the ability to study 162,053 probes for AGM and 158,970 probes for MAC, covering 16,259 and 16,611 genes, respectively [38]. The recently developed annotations files [38] were used for each species, and data were filtered and normalized (AGM and macaques separately) based on the CpG island categories of the speciesspecific annotation files with the SQN Pipeline [82].…”

“…This resulted in the ability to study 162,053 probes for AGM and 158,970 probes for MAC, covering 16,259 and 16,611 genes, respectively [38]. The recently developed annotations files [38] were used for each species, and data were filtered and normalized (AGM and macaques separately) based on the CpG island categories of the speciesspecific annotation files with the SQN Pipeline [82]. All probes with a detection p value above 0.01 were removed.…”

Innate fear responses are reflected in the blood epigenome of rhesus macaques

“…A broad body of evidence suggests that environmental and experiential factors can trigger cascades of epigenetic modification that in turn mediate downstream molecular and behavioral phenotypes. In this study, we focused on DNA methylation in rhesus macaque and human blood samples using the flexibility of Illumina DNA methylation arrays as a translational platform to investigate DNA methylation in both human and non-human primate species [42][43][44][45] . In fact, we were able to identify Moreover, a DMR in PM20D1 similar to our results was recently identified in human blood samples in a longitudinal study of post-traumatic stress disorder 48 , and CAMK2A is one of the most well-understood genes mediating synaptic plasticity underlying fear learning in a variety of paradigms 49 .…”