CpG islands, but not their methylation level, are key in the regulation of meiotic recombination in chicken (Gallus gallus)

Meiotic recombination plays a fundamental role in many sexually reproducing species. Recombination shuffles the genetic material during the first meiotic cell division resulting in new combinations of alleles within each chromosome. In many organisms, the rate of recombination is not uniform across the genome but consists of so called hotspots where the recombination rate is remarkably higher than the genome average. In mammals, the regulation and location of recombination hotspots is regulated by a gene called PRDM9. Many nonmammalian animals, like birds, lack this gene and the precise mechanism for recombination rate regulation is still unknown. Previous findings in passerine birds have observed an association between recombination rate and a genomic feature known as CpG islands (CGIs). CGIs are often located in promoter regions of genes and depending on their methylation status constitute accessible chromatin regions. It has therefore been suggested that the proteins involved in the regulation of recombination have better access to less condense chromatin regions. In this study, I tested if the association between recombination rate and CGIs found in passerine birds is also true in chicken. I also tested if methylation levels of CGIs play a role in recombination rate regulation in chicken. To this end, I used previously published data for CGI locations and a methylation map in chicken, and unpublished data of recombination rate estimates. I found that the association between recombination rate and CGIs observed in passerine birds extends to chicken, suggesting that this is an ancestral trait in birds. I did not, however, find a negative association between methylation levels and recombination rate as hypothesised based on a relationship between methylation level and chromatin accessibility. This suggests that DNA methylation level at CGIs is not a strong determinant of recombination in chicken, although there may be some workflow artefacts or unknown factors remaining in my analysis obscuring the relationship between these two variables