Chapter Mutation

5.1 The source of all genetic variation

• Types of mutations and rates of mutation.

• How can a low-probability event like mutation account for genetic variation?

• The spectrum of fitness for mutations.

The previous four chapters have discussed in detail genotype frequencies under random and non-random mating, the relationship between genetic drift and the effective population size, as well as population subdivision and gene flow. These and all other processes in populations act to shape or change existing genetic variation. But where does genetic variation come from in the first place? The Hardy-Weinberg expectation shows clearly that particulate inheritance itself does not alter genotype or allele frequencies and so it is not a source of genetic variation. Any form of non-random mating alters only genotype frequencies and leaves allele frequencies constant. Genetic drift serves to erode genetic variation as sampling error leads to allele frequency change and eventually to fixation and loss. Gene flow just serves to partition genetic variation among subpopulations, thereby altering patterns of population structure. The process of mutation, the permanent incorporation of random errors in DNA that results in differences between ancestral and descendant copies of DNA sequences, is the ultimate source of all genetic variation. This chapter will cover the process of mutation starting out with a description of the patterns and rates of mutation. The following sections will present classical population genetic models for the fate of a new mutation, the impact of mutation on allele frequencies in a population, and the predicted balance between removal of genetic variation by genetic drift and its replacement by mutation. This chapter will also cover several models of the way new alleles are introduced by mutation commonly employed in population genetics, illustrated with applications that highlight the consequences of these models. The final section of the chapter will show how the process of mutation can be incorporated into genealogical branching models.

Mutation is a broad term that encompasses a wide variety of events that lead to alterations in DNA sequences. Point mutations lead to the replacement of a single base pair by another nucleotide. Point mutations to chemically similar nucleotides (purine to purine (A o G) or pyrimidine to pyrimi-dine (C o T)) are called transitions, while point mutations to chemically dissimilar nucleotides (purine to pyrimidine or pyrimidine to purine) are called transversions. Base substitutions that occur within coding genes may or may not alter the protein produced by that gene. Synonymous or silent mutations result in the same translation of a DNA sequence into a protein due to the redundant nature of the genetic code, while nonsynonymous or missense mutations result in a codon that does change the resulting amino acid sequence.

Mutation can take the form of insertion or deletion of DNA sequences, often referred to as indels. Indels within coding regions result in frameshift mutations if the change in sequence length is not an even multiple of three, altering the translation of a DNA sequence and possibly creating premature stop codons. Indels may range in size from a single base pair to segments of chromosomes containing many thousands of base pairs. Arrays of multiple copies of homologous genes called multigene families are formed by duplication events. Some copies of such duplicated genes may lose functions due to the accumulation of mutations, becoming pseudogenes. Gene conversion may result in the homogenization of the sequences of multiple loci within multigene families. Gene conversion occurs because of inappropriate mismatch repair that takes place during meiosis. Sections of two homologous chromosomes may anneal when they are single stranded during DNA replication. If these regions differ slightly in sequence, the annealed stretch will contain single nucleotide mismatches. These mismatches will then be repaired to proper Watson-Crick base pairing by enzymes normally involved in proofreading during DNA replication. The process of annealing between two sister chromosomes tends to happen frequently when the same gene has been repeated many times, because the gene copies have very similar sequences and the chromosomes can anneal anywhere along the length of the gene array. The result is that all gene copies within multigene regions tend to converge on one random version of DNA sequence without recombination taking place.

Mutation may also take the form of rearrangements where a chromosomal region forms a loop structure that results in a segment breaking and being repaired in reversed orientation, called an inversion. Translocations are mutations where segments of chromosome break free from one chromosome and are incorporated by repair mechanisms into a non-homologous chromosome. Transposable elements, segments of DNA that are capable of moving and replicating themselves within a genome, are frequent causes of translocation mutations. Lateral or horizontal gene transfer, the movement and incorporation of DNA segments between different individuals and even different species, is another possible avenue of mutation that occurs relatively frequently in prokaryotes. For more detail on the molecular mechanisms that underlie these different types of mutations consult a text such as Lewin (2003).

The probability that a locus or base pair will experience a mutation is a critical parameter in population genetics since the rate of mutation describes how rapidly novel genetic variation is added to populations. Although it seems counterintuitive, mutation rates are actually quite difficult to estimate with precision in many types of organism (see Drake et al. 1998; Fu & Huai 2003). Consider the case of mutation rates at a single locus that has a well-understood effect on the phenotype of an organism, like coat color in mice. The data available to estimate mutation rates are numbers of progeny that have a different coat color than expected based on the known coat-color genotypes of the parents. It is simple to divide the number of progeny with unexpected coat colors by the total number of progeny examined. However, that calculation estimates the frequency of detectable changes to coat color due to some molecular change at the coat color locus. That is an estimate of the frequency of all types of mutation anywhere at the locus rather than an estimate of the mutation rate. Such an estimate of mutation frequency could also be biased since only mutational changes that caused an obvious change in coat color are included. Not all mutations will be reflected in coat colors, like changes to the third position nucleotide of a codon that are silent, or synonymous, and do not change the resulting amino acid sequence of a gene. Additionally, mutations may vary in their effect on coat color with some mutations having little or no easily observable effect on the phenotype. Therefore, the frequency of observable changes to the phenotype is not equivalent to the mutation rate.

An estimate of the mutation rate requires more information. One critical detail is the number of replications a locus or genome experiences, because mutational changes usually occur during the replication process. Different cell types and different species experience different numbers of cell replications during growth and reproduction. For example, in mammals mutations are more frequent in male gametes than in female gametes because there are many more cell divisions before the production of a sperm than there are before production of an egg. However, the underlying mutation rate could be identical for male and female gametes with the difference due only to the different number of genome replications that occur. Another set of considerations is the size of a locus or genome available to mutate. In the hypothetical mouse coat color example, the number of base pairs at the coat color locus is a critical piece of information. The rate of mutation per base pair estimated from the frequency of coat color changes is very different if the locus has 900 or 90 base pairs.

The distinction between mutation frequency and mutation rate highlights the fact that mutation rates in population genetics are expressed in a variety of terms depending on experimental methods and the life cycle of an organism. The target of mutation can be an entire genome, a locus, or a single base pair, while the rate can be expressed in time units per DNA replication or per sexual generation. Comparisons of mutation rates only make sense when the target size and time period are expressed in identical units. Generally, for population genetic predictions involving sexual eukaryotes, mutation rates per sexual generation are the relevant units. Predictions for prokaryotes such as Escherichia coli or yeast would more naturally use mutation rates per cell division.

The most general rule of mutation is that it is a rare event with a low probability of occurrence. In a classic experiment involving literally millions of

Table 5.1 Per-locus mutation rates measured for five loci that influence coat-color phenotypes in inbred lines of mice (Schlager & Dickie 1971). Dominant mutations were counted by examining the coat color of F1 progeny from brother-sister matings. Recessive mutations required examining the coat color of F1 progeny from crosses between an inbred line homozygous for a recessive allele and a homozygous wild-type dominant allele. The effort to obtain these estimates was truly incredible, involving around 7 million mice observed over the course of 6 years.

Locus Gametes tested Mutations observed Mutation rate per locus x 10-6(95% CI)

Mutations from dominant to recessive alleles



0 0

Post a comment