Identifying the Evolutionary Forces Testing the Neutral Theory Maintaining polymorphism Choosing between the adaptive selection theory and the neutral theory is not simple, for they The Adaptive Selection Theory count for much of the data on gene polymorphism in nat As evidence began to accumulate in the 1970s that natural ural populations. A few well-characterized instances where populations exhibit a great deal of genetic polymorphism selection acts on enzyme alleles do not settle the more gen question arose: What evolutionary force is maintaining the ing large-scale patterns of polymorphism sheds light on the polymorphism? As we have seen, there are in principle five difficulty of choosing between the two theories processes that act on allele frequencies: mutation, migra Population size: According to the neutral theory, tion, nonrandom mating, genetic drift, and selection. Be- polymorphism as measured by H should be proportional cause migration and nonrandom mating are not major in- tion size m fluences in most natural populations, attention focused tion rate among neutral alleles u is constant. Thus, H the other three forces The first suggestion, advanced by R. C Lewontin(one should be much greater for insects than humans, as there are far more individuals in an insect population of the discovers of enzyme polymorphism) and many oth- than in a human one. When DNA sequence variation is ers, was that selection was the force acting to maintain the examined, the fruit fly Drosophila melanogaster indeed polymorphism. Natural environments are often quite het exhibits sixfold higher levels of variation, as the theory erogeneous, so selection might reasonably be expected to predicts; but when enzyme polymorphisms are exam pull gene frequencies in different directions within differ ined. levels of variation in fruit flies and humans are ent microhabitats, generating a condition in which many similar. If the level of dNa variation correctly mirrors alleles persist. This proposal is called the adaptive selec- the predictions of the neutral theory, then something tion theory (selection? )is increasing variation at the enzyme level in humans. These sorts of patterns argue for rejection of The Neutral theor the neutral theory A second possibility, championed by the great Japanese The nearly neutral model: One way to rescue the geneticist Moto Kimura, was that a balance between mu neutral theory from these sorts of difficulties is to retreat tation and genetic drift is responsible for maintaining from the assumption of strict neutrality, modifying the polymorphism. Kimura used elegant mathematics to theory to assume that many of the variants are slightly demonstrate that. even in the absence of selection, nat deleterious rather than strictly neutral to selection. With ural populations could be expected to contain consider this adjustment, it is possible to explain many of the able polymorphism if mutation rates(generating the vari population-size-dependent large-scale patterns. How ation)were high enough and population sizes(promoting ever, little evidence exists that the wealth of enzyme genetic drift) were small enough. In this proposal, selec- polymorphism in natural populations is in fact slightly ion is not acting, differences between alleles being"neu tral to selection. The proposal is thus called the neutral As increasing amounts of DNA sequence data become theory available, a detailed picture of variation at the DNA level is imura's theory, while complex, can be stated simply emerging. It seems clear that most nucleotide substitutions H=1/(4N+1) that change amino acids are disadvantageous and are elimi- nated by selection. But what about the many protein alleles H, the mean heterozygosity, is the likelihood th that are seen in natural populations? Are they nearly neu- randomly selected member of the population will be het- tral or advantageous? No simple answer is yet available, al erozygous at a randomly selected locus. In a population though the question is being actively investigated. Levels of without selection, this value is influenced by two vari- lymorphism at enzyme-encoding genes may depend ables, the effective population size(N) and the mutation both the action of selection on the gene(the adaptive selec. rate(a. tion theory)and on the population dynamics of the specie The peculiar difficulty of the neutral theory is that the(the nearly neutral theory), with the relative contribution vel of polymorphism, as measured by H, is determined varying from one gene to the next by the product of a very large number, Ne, and a very Adaptive selection clearly maintains some enzyme poly small number, u, both very difficult to measure with pre- morphisms in natural populations. Genetic drift seems to cision. As a result, the theory can account for almost any play a major role in producing the variation we see at the value of H, making it very difficult to prove or disprove. DNA level. For most enzyme-level polymorphism, investi- As you might expect, a great deal of controversy has gators cannot yet choose between the selection theory and resulted he nearly neutral theory. 430 Part vI EvolutionIdentifying the Evolutionary Forces Maintaining Polymorphism The Adaptive Selection Theory As evidence began to accumulate in the 1970s that natural populations exhibit a great deal of genetic polymorphism (that is, many alleles of a gene exist in the population), the question arose: What evolutionary force is maintaining the polymorphism? As we have seen, there are in principle five processes that act on allele frequencies: mutation, migra￾tion, nonrandom mating, genetic drift, and selection. Be￾cause migration and nonrandom mating are not major in￾fluences in most natural populations, attention focused on the other three forces. The first suggestion, advanced by R. C. Lewontin (one of the discovers of enzyme polymorphism) and many oth￾ers, was that selection was the force acting to maintain the polymorphism. Natural environments are often quite het￾erogeneous, so selection might reasonably be expected to pull gene frequencies in different directions within differ￾ent microhabitats, generating a condition in which many alleles persist. This proposal is called the adaptive selec￾tion theory. The Neutral Theory A second possibility, championed by the great Japanese geneticist Moto Kimura, was that a balance between mu￾tation and genetic drift is responsible for maintaining polymorphism. Kimura used elegant mathematics to demonstrate that, even in the absence of selection, nat￾ural populations could be expected to contain consider￾able polymorphism if mutation rates (generating the vari￾ation) were high enough and population sizes (promoting genetic drift) were small enough. In this proposal, selec￾tion is not acting, differences between alleles being “neu￾tral to selection.” The proposal is thus called the neutral theory. Kimura’s theory, while complex, can be stated simply: H¯¯ = 1/(4Neµ +1) H¯¯, the mean heterozygosity, is the likelihood that a randomly selected member of the population will be het￾erozygous at a randomly selected locus. In a population without selection, this value is influenced by two vari￾ables, the effective population size (Ne) and the mutation rate ( µ). The peculiar difficulty of the neutral theory is that the level of polymorphism, as measured by H¯¯ , is determined by the product of a very large number, Ne, and a very small number, µ, both very difficult to measure with pre￾cision. As a result, the theory can account for almost any value of H¯¯ , making it very difficult to prove or disprove. As you might expect, a great deal of controversy has resulted. Testing the Neutral Theory Choosing between the adaptive selection theory and the neutral theory is not simple, for they both appear to ac￾count for much of the data on gene polymorphism in nat￾ural populations. A few well-characterized instances where selection acts on enzyme alleles do not settle the more gen￾eral issue. An attempt to test the neutral theory by examin￾ing large-scale patterns of polymorphism sheds light on the difficulty of choosing between the two theories: Population size: According to the neutral theory, polymorphism as measured by H¯¯ should be proportional to the effective population size Ne, assuming the muta￾tion rate among neutral alleles µ is constant. Thus, H¯¯ should be much greater for insects than humans, as there are far more individuals in an insect population than in a human one. When DNA sequence variation is examined, the fruit fly Drosophila melanogaster indeed exhibits sixfold higher levels of variation, as the theory predicts; but when enzyme polymorphisms are exam￾ined, levels of variation in fruit flies and humans are similar. If the level of DNA variation correctly mirrors the predictions of the neutral theory, then something (selection?) is increasing variation at the enzyme level in humans. These sorts of patterns argue for rejection of the neutral theory. The nearly neutral model: One way to rescue the neutral theory from these sorts of difficulties is to retreat from the assumption of strict neutrality, modifying the theory to assume that many of the variants are slightly deleterious rather than strictly neutral to selection. With this adjustment, it is possible to explain many of the population-size-dependent large-scale patterns. How￾ever, little evidence exists that the wealth of enzyme polymorphism in natural populations is in fact slightly deleterious. As increasing amounts of DNA sequence data become available, a detailed picture of variation at the DNA level is emerging. It seems clear that most nucleotide substitutions that change amino acids are disadvantageous and are elimi￾nated by selection. But what about the many protein alleles that are seen in natural populations? Are they nearly neu￾tral or advantageous? No simple answer is yet available, al￾though the question is being actively investigated. Levels of polymorphism at enzyme-encoding genes may depend on both the action of selection on the gene (the adaptive selec￾tion theory) and on the population dynamics of the species (the nearly neutral theory), with the relative contribution varying from one gene to the next. Adaptive selection clearly maintains some enzyme poly￾morphisms in natural populations. Genetic drift seems to play a major role in producing the variation we see at the DNA level. For most enzyme-level polymorphism, investi￾gators cannot yet choose between the selection theory and the nearly neutral theory. 430 Part VI Evolution