Does (Population) Size Matter (in Evolution)?

In the 28 April issue of Science, Eric Bazin, Sylvain Glémin, and Nicolas Galtier report that population size matters for nuclear DNA diversity, but not for mitochondrial DNA diversity. The paper was forwarded to me by one of my favorite philosophers of biology because of the explanation Bazin et al. offer for explaining the disconnection between mtDNA diversity and population size: genetic draft. Here's the story.

A basic tenet of population genetics theory is that population size and extent of genetic diversity are intimately connected. Large populations, in fact, should be genetically more diverse than small ones. Why? Neutral genetic variation is modulated by drift and mutation. Drift removes genetic variation while mutation adds it. In small populations, drift removes variation faster than mutation can add it. But as populations get larger and larger, drift is less and less effective at removing genetic variation.

This basic tenet of population genetics was challenged by allozyme polymorphism studies of genetic diversity in the 1960s. What the data suggested was that if this tenet were correct, the population sizes of the represented species should be roughly equal (see Lewontin 1974 for discussion). But of course population sizes are not equal.

By the 1980s, biologists learned that the situation suggested by the allozyme data was not nearly so bad (see Gillespie 1991 for discussion). Yet, the puzzle of the connection between population size and genetic diversity persists. In fact, the nature of genetic variation has long been something of a mystery from the point of view of population genetics.

John Maynard Smith and John Haigh (1974) offered genetic hitchhiking, or linked selection, as a solution to the apparent population size paradox suggested in Richard Lewontin's (1974) review. The idea is that when an advantageous mutant is swept to fixation by selection, genetic variation at loci that are linked to the mutant is reduced. Genetic homogeneity in large populations could be due to such a process. Maynard Smith and Haigh's hitchhiking explanation languished in large part because it depends on populations having high linkage disequilibrium.

Allozyme polymorphism studies of population diversity have been replaced by DNA-based marker studies, especially mitochondrial DNA (mtDNA) marker studies. In their Science paper, Bazin et al. analyzed diversity data for nuclear DNA from 417 species, mitochondrial DNA from 1,683 species, as well as allozyme data for 912 species. The team tested the connection between population size and genetic diversity against several ecological and phylogenetic factors. Every comparison of nuclear DNA data was congenial with the idea that population size and levels of genetic variation are connected. But not so for the mitochondrial DNA data. mtDNA variation across species, according to Bazin et al., is independent of population size.

By the end of the 1980s, some data became available that showed that genetic variation is reduced in regions of low recombination in Drosophila (see Gillespie 2001). For this reason, among others, John Gillespie revisited the hitchhiking hypothesis (2000a, 2000b, 2001). But rather than following Maynard Smith and Haigh's deterministic model, Gillespie developed a stochastic model of a process he calls genetic draft. The essential difference between genetic draft and genetic hitchhiking is that, for draft, random variables are used for the timing of a hitchhiking event and the probability that the neutral allele is linked to a selectively advantageous mutation. Gillespie was able to disconnect population size and levels of genetic variation.

Bazin et al. suggest that the homogeneity of genetic variation across species for the mtDNA data can be explained by genetic draft. Why draft? There were only very low levels of recombination in mtDNA relative to nuclear DNA. They thus think that mtDNA is ripe for hitchhiking events of the sort Gillespie describes (including his assumptions about the rate of substitution).

Of course, low levels of recombination in mtDNA relative to nuclear DNA is not enough to establish the conclusion that genetic draft explains the independence of genetic variation from population size for mtDNA. Bazin et al. also need evidence of selective substitutions --the triggers for hitchhiking events. For that, they appeal to the neutrality index as a measurement of selection. A neutrality index that is <1 is taken to show that selective amino acid substitutions have occurred. The median values of the neutrality index for both vertebrates and invertebrates is <1. Indeed, Bazin et al. argue that 58% of amino acid substitutions are selectively advantageous in invertebrate mtDNA and 12% in vertebrate mtDNA.

Other explanations for the mtDNA genetic diversity data include variations in mitochondrial mutation rate, population bottlenecks, and background selection. Bazin et al. consider each of these alternatives, but argue for various reasons that they are less likely than the genetic draft explanation. They claim, for instance, that variation in mitochondrial mutation rate is an unlikely explanation because they think it's unlikely that that mutation rate is inversely related to population size throughout the animal taxa represented in the data. They think bottlenecks are unlikely because they would have to affect the nuclear genome as well, but that was not observed. Background selection is unlikely because there is still an expectation that genetic variation increases as population size increases, which was not observed for the mtDNA data. By elimination, then, Bazin et al. end with genetic draft as the best explanation of the data.

Note: For those of you into human evolution, the summary article by Adam Eyre-Walker preceding Bazin et al.'s paper  points out that humans appear to be an exception to their observed pattern. Eyre-Walker says, "[i]f the authors are correct, then the effective population size estimated from the mitochondrial DNA should be lower than that estimated from autosomal DNA. This is not what we see in humans; the effective population sizes estimated from autosomal DNA, Y-chromosome DNA, and mitochondrial DNA are all approximatley 10,000" (p. 538). Eyre-Walker doesn't conclude that Bazin et al.'s observations are wrong. Just that "humans have such small effective population sizes that adaptive evolution in the mitochondrial genome is very rare...." (p.538).

References

Bazin, E., Glémin, S., and Galtier, N. (2006), "Population Size Does Not Influence Mitochondrial Genetic Diversity in Animals", Science 312: 570-572.

Eyre-Walker, A. (2006), "Size Does Not Matter for Mitochondrial DNA", Science 312: 537-538.

Gillespie, J. H. (1991), The Causes of Molecular Evolution. New York: Oxford University Press.

Gillespie, J. H. (2000a), "Genetic Drift in an Infinite Population: The Pseudohitchhiking Model", Genetics 155: 909-919.

Gillespie, J. H. (2000b), "The Neutral Theory in an Infinite Population", Gene 261: 11-18.

Gillespie, J. H. (2001), "Is the Population Size of a Species Relevant to its Evolution?", Evolution 55: 2161-2169.

Gillespie, J. H. (2004), Population Genetics: A Concise Guide. Baltimore, MD: Johns Hopkins University Press.

Lewontin, R. C. (1974), The Genetic Basis of Evolutionary Change. New York: Columbia University Press.

Maynard Smith, J. & J. Haigh (1974), “The Hitch-hiking Effect of a Favourable Gene”, Genetical Research 23: 23-25.