Inter-Individual Variation in DNA Methylation Patterns across Two Tissues and Leukocytes in Mature Brahman Cattle

Abstract

Quantifying the natural inter-individual variation in DNA methylation patterns is important for identifying its contribution to phenotypic variation, but also for understanding how the environment affects variability, and for incorporation into statistical analyses. The inter-individual variation in DNA methylation patterns in female cattle and the effect that a prenatal stressor has on such variability have yet to be quantified. Thus, the objective of this study was to utilize methylation data from mature Brahman females to quantify the inter-individual variation in DNA methylation. Pregnant Brahman cows were transported for 2 h durations at days 60 ± 5; 80 ± 5; 100 ± 5; 120 ± 5; and 140 ± 5 of gestation. A non-transport group was maintained as a control. Leukocytes, amygdala, and anterior pituitary glands were harvested from eight cows born from the non-transport group (Control) and six from the transport group (PNS) at 5 years of age. The DNA harvested from the anterior pituitary contained the greatest variability in DNA methylation of cytosine-phosphate-guanine (mCpG) sites from both the PNS and Control groups, and the amygdala had the least. Numerous variable mCpG sites were associated with retrotransposable elements and highly repetitive regions of the genome. Some of the genomic features that had high variation in DNA methylation are involved in immune responses, signaling, responses to stimuli, and metabolic processes. The small overlap of highly variable CpG sites and features between tissues and leukocytes supports the role of variable DNA methylation in regulating tissue-specific gene expression. Many of the CpG sites that exhibited high variability in DNA methylation were common between the PNS and Control groups within a tissue, but there was little overlap in genomic features with high variability. The interaction between the prenatal environment and the genome could be responsible for the differences in location of the variable DNA methylation.

Keywords: Brahman; methylation; prenatal stress; variability.

Figure 1

Pearson correlation coefficient estimates of samples (animals) using the beta values of cytosine-phosphate-guanine sites across the genome in the (A) Control amygdala, (B) Prenatally Stressed (PNS) amygdala, (C) Control anterior pituitary (D) PNS anterior pituitary (E) Control leukocytes, and (F) PNS leukocytes.

Figure 2

Histograms of the inter-individual beta value ranges for the cytosine-phosphate-guanine sites across the genome in the (A) Control amygdala, (B) Prenatally Stressed (PNS) amygdala, (C) Control anterior pituitary (D) PNS anterior pituitary (E) Control leukocytes, and (F) PNS leukocytes. The dashed lines represent the mean inter-individual beta value ranges.

Figure 3

Box plots of standard deviations (SD) for the cytosine-phosphate-guanine sites (genome wide) analyzed in each tissue and leukocytes from the Prenatally Stressed (PNS) and Control groups.

Figure 4

Overlap of genome wide cytosine-phosphate-guanine sites with a beta value standard deviation $\ge$0.1 across the tissues and leukocytes in the (A) Control and (B) Prenatally Stressed group and cytosine-phosphate-guanine sites with a p $\le$ 0.001 for the chi-square test for the variance in the (C) Control and (D) Prenatally Stressed group.

Figure 5

Number of sites with a beta value standard deviation $\ge$ 0.1 associated with different types of repetitive elements, long interspersed nuclear elements (LINE), short interspersed nuclear elements (SINE), long terminal repeat (LTR), satellite regions, ribosomal RNA (rRNA) and simple repeats throughout the genome.

Figure 6

Overlap of gene bodies with high DNA methylation variation across the tissues and leukocytes in the Control (A) and Prenatally Stressed (B) group.