Molly Przeworski seminar - regulation of recombination

Post date: Dec 5, 2016 11:23:26 PM

Meiotic recombination hotspots are found in most sexual organisms, but they arise through very different mechanisms. Molly Przeworski gave an excellent seminar today with an overview of strategies for controlling where recombination occurs. Fundamentally, the existence of a recombination hotspot means that some genetic feature - be it a sequence-based feature or a chromatin feature - locally enhances the likelihood of a recombination. However, a devilish side effect of recombination is that if there are two alleles that differ in their propensity to direct local recombination, individuals heterozygous for these alleles will be slightly more likely to have progeny with the recombination-poor allele, due to a process known as biased gene conversion. In effect, this means there is some inherent tendency to remove recombination-enhancing features from the genome, barring selection to maintain these features. In many vertebrate species, the sites of recombination hotspots are specified by specific DNA motifs bound by the protein Prdm9. These motifs were indeed found to be under negative selection in primates. Molly showed that Prdm9 is found in many vertebrates (and also has been lost in many vertebrate lineages). In all lineages that retain Prdm9, Prdm9's DNA-binding domain is rapidly evolving, implying that it is universally driving its binding sites out of existence.

It occurred to me that Prdm9 could avoid biased gene conversion by directing the double-stranded break to occur a short distance away from where it binds. Gene conversion tracts are typically around 300 bp long, so if the double-stranded break were a couple nucleosomes away from Prdm9's binding site, gene conversion would usually not include the recombination-enhancing sequence. But upon further thought, it seems challenging for recombination-at-a-distance to gain a foothold if it came into existence. Imagine a Prdm9 allele gains a mutation such that it no longer removes its own binding site. This new Prdm9 allele confers no fitness advantage over the original, as it is still binding the same depleted sites. It is only if it gains a new binding site and proceeds to maintain that binding site in the genome while the other Prdm9 alleles remove theirs that it can confer a fitness advantage.