NEW TOPIC: MOLECULAR EVOLUTION

THE NEUTRAL THEORY OF MOLECULAR EVOLUTION Genetic drift causes more substitutions than does natural selection. Molecular evolution is a balance between drift and mutation. Motoo Kimura (1968)

OBSERVATIONS THAT PROMPTED THE NEUTRAL THEORY Observed rates of amino acid substitutions in proteins were surprisingly high. The amount of heterozygosity in natural populations seemed too high to be explained by selection.

UNIFORM RATE OF AMINO ACID SUBSTITUTION Amino Acid Differences

FUNCTIONALLY IMPORTANT PARTS OF THE AMINO ACID SEQUENCE EVOLVE MORE SLOWLY THAN OTHER PARTS

Variation in the Rate of evolution in different types of DNA

THE FIXATION OF NEUTRAL MUTATIONS For any locus, 2Nµ mutations enter a population each generation. Each new neutral mutation has a probability of being fixed equal to its initial frequency, 1/(2N). The average number of substitutions per locus per generation is equal to the product of these, 2Nµ x 1/(2N) = µ

= µ BASIC TENANT OF THE NEUTRAL THEORY: The substitution rate = the mutation rate = µ

EQUILIBRIUM HETEROZYGOSITY Basic Neutral Theory models consider only panmictic populations, so migration and nonrandom mating are ignored. By eliminating considerations of selection as an important force in molecular evolution, neutral theory models a balance between drift and mutation. As a result, at equilibrium, F = 1/(1 + 4Nµ), and H = 4Nµ/(1 + 4Nµ)

PREDICTION: Because the strength of random genetic drift depends on Ne, neutral mutations should take longer to drift to fixation in large populations (on average, 4Ne generations). Therefore, large populations should contain many alleles at intermediate frequency, whereas small populations should exhibit little polymorphism.

POSITIVE RELATIONSHIP BETWEEN HETEROZYGOSITY AND POPULATION SIZE

PREDICTION: Low µ High µ GENES, OR GENE REGIONS, WITH A HIGHER SUBSTITUTION RATE SHOULD SHOW INCREASED POLYMORPHISM. Low µ High µ

REGIONS OF PROTEIN CODING GENES WITH HIGH SUBSTITUTION RATES HAVE HIGH LEVELS OF POLYMORPHISM

FIXATION OF NEUTRAL VS. BENEFICIAL ALLELES Drift Dominates if: Selection dominates if: 1/2N >> 2s 1/2N << 2s Probability of fixation of a new mutant allele: Drift: Selection: 1/2N 2s

The rate of mutation (per year) may increase with metabolic rate and decrease with generation length: ENDOTHERMS ECTOTHERMS

Mutations may have different effects on fitness Synonymous mutation: do not alter the amino acid sequence of the protein Often selectively neutral Non-synonymous mutation: alter the amino acid sequence of the protein More likely to be subject to selection

COROLLARY OF THE NEUTRAL THEORY: NEUTRAL (SILENT) MUTATIONS HAVE HIGHER SUBSTITUTION RATES THAN MUTATIONS UNDER SELECTION. Most of the human genome (>95%) is noncoding. Most mutations in these regions are silent. Approx. 24% of base substitutions in protein coding regions are silent (synonymous).

SYNONYMOUS VS. NONSYNONYMOUS SUBSTITUTIONS IN THE  - GLOBIN PROTEIN CODING REGION

SYNONYMOUS VS. NONSYNONYMOUS SUBSTITUTIONS

SYNONYMOUS RATE NONSYNONYMOUS RATE

Detecting selection on DNA sequences Synonymous substitutions: do not change protein Should evolve at a neutral rate Estimated by ks (number of synonymous substitutions / 1 kb) Nonsynonymous substitutions: change protein Faster evolution than synonymous sites indicates positive selection Slower evolution than synonymous sites indicates purifying selection Estimated by ka (number of nonsynonymous substitutions / 1 kb) The ratio ka/ks can be used to infer the action of selection on protein coding genes.

Signature of positive selection

Positive selection on FOXP2

ESTIMATION OF THE MUTATION RATE PER NUCLEOTIDE SITE ANCESTRAL SPECIES SPECIES A SPECIES B t years µ = mutation rate per nucleotide site per year Expected number of changes per site between species A and B at synonymous (silent) sites = 2 t µ Observed number of changes per site between species A and B = D Estimated mutation rate = µ = D / (2 t)

1 sub./site/100 MY 1 sub./site/100 BY

ESTIMATED RATES OF MUTATION PER NUCLEOTIDE SITE SYNONYMOUS SUBSTITUTION RATE (substitutions/site/billion years) n = NUCLEAR GENOME m = MITOCHONDRIAL GENOME c = CHLOROPLAST GENOME FROM: Lynch & Blanchard. 1998.Genetica 102-103:29-39.

THE MOLECULAR CLOCK Because the substitution rate = µ, the rate of molecular evolution should be a constant over time as long as the mutation rate does not change over time.

DATING THE TIME OF EVOLUTIONARY DIVERGENCE BETWEEN TWO LINEAGES ANCESTRAL SPECIES SPECIES 1 SPECIES 2 AATGGTGGCT AATCCGGGCT t d = estimated rate of nucleotide substitutions / site/ year (from a calibrated molecular clock) D = observed fraction of nucleotide sites differing between two species (0.3 in the example above) 2 x t x d = expected fraction of nucleotide differences between species Estimated time of divergence = t = D / (2d)

The molecular clock can be calibrated with dates from the fossil record 46 32 30 21 16

DATING THE TIME OF EVOLUTIONARY DIVERGENCE BETWEEN TWO LINEAGES EXAMPLE: The substitution rate per site is d = 10-8, and the nucleotide sequences from two species differ at a fraction D = 0.2 of the nucleotide sites. We estimate the time since isolation between the two lineages as: t = 0.2 / (2 x 10-8) = 107 years

Species A: T C G Species B: T A G CAVEATS TO THE MOLECULAR CLOCK Two sequences may have similar sequences BUT not because of common descent. Species A: T C G Species B: T A G Convergence – lineages may diverge and then acquire the same base by further mutation.

MUTATIONAL SATURATION Since nucleotides have only four possible states (A,G,T,C), DNA sequences will not be expected to diverge 100% even after an infinitely long period of time. Maximum Divergence: 75% Mathematical models can be used to correct for the saturation effect by making the relationship between divergence and time linear.

DIFFERENT GENES EVOLVE AT DIFFERENT RATES Variation in the selective constraints on different genes leads to differing rates of molecular evolution. Thus, a calibration of the molecular clock must be done separately for each gene.

NEUTRAL THEORY: SUMMARY The Neutral Theory provides an intentionally simplified model of nature (drift/mutation balance), which makes quantitative and testable predictions about the patterns of molecular evolution in nature. The Neutral Theory provides a “NULL MODEL” for statistical tests of selection at the molecular level. The predictions of the Neutral Theory are generally supported by the empirical data.

NEUTRAL THEORY OF PHENOTYPIC EVOLUTION For quantitative traits the rate of increase of between-population variance = 2 x U x average squared mutation effect = 2Vmt The rate of divergence of a neutral character is equal to twice the rate of polygenic mutation.

THE RATE OF DIVERGENCE OF SKELETAL MORPHOLOGY IN MAMMALS TENDS TO BE WELL BELOW THE NEUTRAL EXPECTATION Upper and lower bounds of the neutral rate (Vm / Ve) Rate of increase of between-population variance (in units of within-population variance) FROM: M. Lynch. 1990. Am. Nat. 136:727-741.