Hitching a Ride on a Drosophila Genome

Not all regions of the genome are equal in the eyes of evolution. For example, natural selection is more effective on genes in regions of higher recombination. We have known this for a while. The connection between recombination rate and natural selection was nicely refined when it was shown that DNA polymorphism is lower in regions of low recombination and higher in regions of high recombination (see Begun and Aquadro). This could be due to higher mutation rates in regions of high recombination (and vice versa), greater reach of selective sweeps in regions of low recombination (the hitchhiking hypothesis), or background selection removing more polymorphism in regions of low recombination (reviewed here). It was recently shown that the mutagenic effects of recombination can explain the link between recombination rate and DNA sequence polymorphism in the human genome.

The original demonstration of a correlation between recombination rate and polymorphism used data from Drosophila, and it seems highly unlikely that the relationship is due to mutagenic effects. That leaves hitchhiking and background selection as plausible explanations for the correlation in Drosophila, which can be distinguished by comparing different measures of polymorphism. If there are too many low frequency single nucleotide polymorphisms (SNPs), then there is good evidence for adaptive evolution (see here). It's been said that if we had enough data, we would be able have definitive support for either the hitchhiking or background selection hypotheses. We now have those data, thanks to Josh Shapiro, Chung-I Wu and colleagues.

Shapiro et al's study allowed them to test various hypotheses regarding the role of natural selection in the evolution of DNA sequences. Shown here is plot of the relationship between DNA sequence polymorphism and recombination rate (top graph) and Tajima's D statistic and recombination rate (bottom graph). In the top graph we see further evidence that polymorphism and recombination rate are positively correlated. But that's not all that surprising.

The bottom graph, however, is pretty interesting. Tajima's D can be used to summarize the frequency of SNPs in a gene. A negative D statistic indicates an excess of low frequency variants, which is a hallmark of natural selection acting on a linked site. The first thing to notice in the bottom graph is the positive correlation between Tajima's D and recombination rate. That suggests there are more signatures of adaptive evolution in regions of low recombination, supporting the hitchhiking hypothesis as an explanation for decreased polymorphism

But that correlation, while significant, is pretty sloppy. One prediction of the background selection hypothesis is that selection against deleterious mutations is quite uniform across the entire genome. Regions with low recombination rates will show larger signatures of background selection (in the form of decreased polymorphism), but they won't have a pattern of polymorphism consistent with adaptive evolution. In the graph of Tajima's D versus recombination rate, however, we see a lot of variation in D statistics across the genome. Shapiro et al argue that the extreme variation is not consistent with background selection. They conclude that the correlation between DNA sequence polymorphism and recombination rate is the result of hitchhiking for two reasons:

1. There is a positive correlation between Tajima's D and recombination rate.

2. There are many negative D statistics across the genome, in both regions of high and low recombination. But there is a lot of variation in Tajima's D as you go from regions of low recombination to regions of high recombination. That suggests many different instances of positive selection shaping the pattern of polymorphism. The scattered nature supports adaptive evolution because background selection should be more uniform -- all genes should be eliminating deleterious mutations, whereas only some genes will obtain an advantageous mutation.

To recap: Begun and Aquadro described a correlation between DNA sequence polymorphism and recombination rate and attributed it to hitchhiking. It was then pointed out that background selection could also lead to a decrease in polymorphism in regions of low recombination. Now, we have new data that favor the hitchhiking hypothesis. If these results hold true, adaptive evolution plays a significant role in shaping the DNA sequence polymorphism in Drosophila genomes.

Begun DJ and Aquadro CF. 1992. Levels of naturally occurring DNA polymorphism correlate with recombination rates in D. melanogaster. Nature 356: 519-520. doi: 10.1038/356519a0

Charlesworth B, Morgan MT, and Charlesworth D. 1993 The effect of deleterious mutations on neutral molecular variation. Genetics 134: 1289-1303. link

Hill WG, Robertson A. 1966. The effect of linkage on limits to artificial selection. Genet Res 8: 269-294.

Shapiro JA, Huang W, Zhang C, Hubisz MJ, Lu J, Turissini DA, Fang S, Wang H-Y, Hudson RR, Nielsen R, Chen Z, and Wu C-I. 2007. Adaptive genic evolution in the Drosophila genomes. PNAS 104: 2271-2276. doi: 10.1073/pnas.0610385104

Spencer CCA, Deloukas P, Hunt S, Mullikin J, Myers S, et al. 2006. The Influence of Recombination on Human Genetic Diversity. PLoS Genet 2: e148. doi: 10.1371/journal.pgen.0020148

Tajima F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585-595. link