HapMap methylation-associated SNPs, markers of germline DNA methylation, positively correlate with regional levels of human meiotic recombination

Abstract

Inter-individual and regional variability in recombination rate cannot be fully explained by the DNA sequence itself. Epigenetic mechanisms might be one additional factor affecting recombination. A biochemical approach to studying human germline methylation is difficult. We used density of the 434,198 non-redundant methylation-associated SNP (mSNP) in the derived allele HapMap dataset as a surrogate marker for germline DNA methylation. We validated our methodology by demonstrating that the mSNP density confirmed the known patterns of genomic methylation including hypermutability of methylated cytosine and hypomethylation of CpG islands. Using this approach, we found a genome-wide positive correlation between germline methylation and regional recombination rate (500 kb windows: r=0.622, p<10^-15). This remained significant with multiple correlation correcting for sequence features known to affect recombination, such as GC content and CpG dinucleotides (500 kb windows: r=0.172, p<10^-15). Using the ENCODE dataset for increased resolution, we found a positive correlation between germline DNA methylation and recombination rate (50 kb windows: r=0.301, p=0.002). This correlation was further strengthened when corrected for sequence features affecting recombination (50 kb windows: r=0.445, p<0.0001). In the Human Epigenome Project dataset there was increased DNA methylation in regions within recombination hotspots in male germ cells (0.632 vs. 0.557, p = 0.007). Our observed relationship between germline DNA methylation and recombination could be explained in two ways that are not mutually exclusive. DNA methylation could indicate preferred sites for recombination. Alternatively, methylation following recombination could inhibit further recombination perhaps being part of the enigmatic molecular pathway mediating crossover interference.