Meiotic recombination and linkage disequilibrium in the human genome
Evidence is accumulating that recombination events in humans are not randomly distributed, but clustered into hotspots 1-2 kb wide. The Xp/Yp pseudoautosomal region (PAR1) is a male-specific recombination hot domain. The PAR1 gene PPP2R3B shows unusually high levels of nucleotide diversity. I have used LD studies in this region to reveal a dearth of historical recombination, arguing against earlier speculation that high diversity is driven by mutagenic recombination. Chimp LD studies, the existence of recombination hotspots inside LD blocks, and the observation of meiotic drive within a hotspot favouring recombination suppressors all support the hypothesis that recombination hotspots turn over rapidly during evolution. To identify very recently arisen recombination hotspots, LD patterns between African-American and European-American populations were compared in 147 genes, revealing PON2 as the only gene with convincing evidence for a recently arisen recombination hotspot. A second approach, whereby contemporary linkage data is compared to historical LD data, was employed to identify putative young recombination hotspots. The 6 kb D6S383-D6S970 interval on chromosome 6 shows a recombination rate of 33 cM/Mb in pedigrees, yet lies within a block of strong LD. Sperm crossover analyses across this region in two men did not reveal a recombination hotspot, raising the possibility of a hotspot polymorphism at this locus. Finally, the release of data from phase II of the HapMap project allowed construction of genome-wide linkage disequilibrium unit (LDU) maps at high resolution. Sliding window analyses of these maps suggested putative recombination hotspots, directing future research into the characteristics and distribution of recombination hotspots in humans.