Lecture 2: The neutral theory Neutral theory of molecular evolution In 1968 Motoo Kimura published the neutral theory of molecular evolution. This was followed by a similar proposal by King and Jukes in 1969. This was the starting point of an ongoing debate about the causes of genetic variation within and between species. Kimura was convinced that most polymorphisms that occur in a population do not influence the fitness of an individual and consequently are not subjected to selection. He suggested neutral mutations to be the cause of genetic variation in species. Mutations are considered to be neutral even if they happen at non-degenerate sites. Another important statement of the neutral theory is that most changes in allele frequencies in a population can be attributed to genetic drift. Once a neutral mutation arises in a gamete of an individual it is subjected to many forces: the carrier of the gamete must survive to reach the reproductive age, this same gamete must be fertilized and develop into an embryo. Furthermore, the embryo has to be viable to be part of the next generation. Thereby neutral mutations create new alleles which then can rise in frequency and become widely spread in the population or they can get lost by chance. For this reason another term for ‘Neutral theory of molecular evolution’ is ‘mutation-drift hypothesis’. Although Kimura considered genetic drift to be the main force in evolution he never denied “... the role of natural selection in determining the course of adaptive evolution" (Kimura 1986). To model the impact of genetic drift on a population of finite size N we use the Wright-Fisher model. In general, models without selection are much easier to treat because we can neglect the influence of genotype dependent fitness effects. For this reason the Wright-Fisher model is one of the most elementary models in population genetics. Lecture SS Population Genetics I 1 Lecture 2: The neutral theory Assuming the Wright-Fisher model, i.e. a population of diploid organisms with 1) constant population size 2) discrete (non overlapping) generations 3) random mating 4) and an equal sex ratio, we can model the changes of allele frequencies over generations due to genetic drift. The finite population size, which is retained unchanged at N individuals over generations, consists of N1 females and N2 males (N1 = N2). We will observe one locus with two possible alleles A and a to follow the frequency of one allele over one generation (A = white and a = blue, Figure 1). In the parental generation both alleles have the same allele frequency, f A = f a = Figure 1: The Wright-Fisher model describes the random transfer of gametes from one generation to the next. 4 . 8 Due to the assumption of random mating we can combine all parental alleles in one common gene pool (Figure 1). By sampling randomly from this gene pool we can pick alleles to build up the next generation (F1) with size N. In other words, the parents form gametes of a certain genotype, which are pooled together. Afterwards these gametes are combined to form a zygote, thus forming the next generation. The allele frequencies of both alleles changed over one generation (Figure 1). The white allele increased in frequency f a = 5 while the blue allele decreased in 8 3 frequency f A = . 8 Lecture SS Population Genetics I 2 Lecture 2: The neutral theory From the analytical point of view we can make the model more convenient if we treat the N diploid individuals as 2N haploid individuals. Consequently we do not have to think about mating, the gene pool, or the fusion of gametes because every individual is the immediate genetic descendant of one parental individual. To calculate homozygosity of the daughter generation, we sample from the gene pool as follows. First we can sample two alleles which are identical by descent. This means we pick one allele, e.g. a blue one, put it back and sample exactly the same one again. The probability that this happens is 1/2N. Second we can sample two alleles, which are not identical by descent but identical by state (= DNA sequence). Together, we have for the homozygosity of the daughter generation: G' = 1 1 + 1 G . 2N 2N (1) Next we consider the heterozygosity: H = 1 G . The change of heterozygosity from one generation H to the next can be described by 1 H' = 1 H = 1 G' . 2N If the heterozygosity of the parental generation (2) H is compared to the heterozygosity of the next generation H' we can document the change of heterozygosity as ΔH , where ΔH = H' H = 1 H. 2N (3) The important impact of the population size N can be seen in equation (3). The smaller a population is, the more heterozygotes are lost from one generation to another. Lecture SS Population Genetics I 3 Lecture 2: The neutral theory Genetic drift removes genetic variation from the population. This has several consequences, namely non-directional changes in allele frequency, a permanent loss of alleles and evolution which cannot be repeated. If genetic variation is removed by genetic drift, where does the measurable genetic variation come from? Genetic drift and mutations are two opposing forces. While drift removes variation from a population, the appearance of new mutations increases variation. The frequency at which new mutations enter the population can be described by the mutation rate u . This rate displays the chance by which an organism gains a new mutation during life time and is given as probability per site per generation. It can vary between species and even between genomic regions within a single species. Here are some average values of u per generation: Eukaryotes 10-8- 10-9 Bacteria 10-8 DNA-Viruses 10-6- 10-8 RNA-Viruses 10-3- 10-5 As mentioned in the previous lecture, mutations do not necessarily change the protein structure. Such mutations are synonymous changes in the DNA sequence and thus may not be affected by selective forces. All individuals in the population have the same probability to gain a mutation according to the mutation rate u . Mutations at a particular nucleotide site occur according to the infinite-site model. The two mechanisms mutation and genetic drift determine heterozygosity H . ΔH can be approximated by ΔH 1 H + 2u 1 H . 2N (4) The right-hand side of equation (4) is the sum of both mechanisms that contribute to the change of the genetic variation of a population. Lecture SS Population Genetics I 4 Lecture 2: The neutral theory The left term 1 H 2N is the change of genetic variation due to random genetic drift. It will always have a ΔN H = negative value. The right term Δm H = 2u1 H is the change of genetic variation due to new mutations. It will always have positive ~ values. At equilibrium ΔH = 0 . Therefore the equilibrium value of heterozygosity H has to be: 4N u ~ H= . 1+ 4 N u (5) The substitution rate k One very important quantity in population genetics and molecular evolution is the number of new neutral mutations that go to fixation due to genetic drift. This so called substitution rate k can be obtained by comparing two sequences from different species. We use the ratio of nucleotide differences K of the sequences and the time of divergence T of both species to get an estimate of k as k = K . 2T All neutral mutations which occur in a population have the same probability to go to fixation of 1/2N. The substitution rate k for a population is determined by new mutations that happen in the population ( 2 N u ) times their probability to go to fixation 1/2N. Therefore, k = 2N u 1 /2N = u . As population size N cancels out, the equation given above can be reduced to k u . Lecture SS Population Genetics I 5 Lecture 2: The neutral theory This is one of the most striking results in population genetics. Against all intuition the substitution rate is independent of population size and equals the mutation rate u . Consequently it is possible to measure the mutation rate of a species by looking at the number of nucleotide differences at neutral sites compared to a closely related species. Application: The divergence between Drosophila melanogaster and D. simulans is roughly 5 %. Phylogenetic analysis shows that the divergence time T is about 2.3 106 years, which corresponds to about 2.3 107 generations. Therefore using the equation 2Tu k we find u 1.09 109 per generation. The nearly neutral theory of molecular evolution (proposed by Tomoko Ohta in 1973) posits that mutations are neutral or slightly deleterious. Within populations the slightly deleterious mutations show similar dynamics as neutral alleles; however, over long time scales differences in substitution rates may become observable. Lecture SS Population Genetics I 6
© Copyright 2024