Science Wiki

Neutral Theory of Molecular Evolution is the foundational mathematical and conceptual framework in molecular evolutionary biology which holds that at the molecular level of DNA nucleotides and protein amino acid sequences, the overwhelming majority of evolutionary substitutions fixed between diverging species—as well as the bulk of standing genetic polymorphism within populations—are caused not by positive Darwinian natural selection, but by the stochastic fixation of selectively neutral (or nearly neutral) mutations through random genetic drift. Formulated by the Japanese population geneticist Motoo Kimura in 1968 and independently supported by Jack Lester King and Thomas H. Jukes in 1969 ("Non-Darwinian Evolution"), the neutral theory does not deny that adaptive phenotypic evolution is driven by natural selection; rather, it demonstrates that most mutations that survive purifying selection do not significantly alter organismal fitness and therefore serves as the essential null model for modern comparative genomics and population genetics.

Prior to the late 1960s, Neo-Darwinian orthodoxy assumed that almost all amino acid replacements between species were driven by positive directional selection and that genetic polymorphism was maintained primarily by balancing selection. However, when Emile Zuckerkandl and Linus Pauling compared hemoglobin and cytochrome *c* amino acid sequences across vertebrates in 1962–1965, they discovered that substitutions accumulated at a remarkably uniform rate over geological time (the molecular clock), and Richard Lewontin and Jack Hubby (1966) uncovered far higher allozyme heterozygosity in natural populations than J.B.S. Haldane's "cost of natural selection" (substitutional load) could tolerate.

Kimura resolved both paradoxes through a remarkably elegant population-genetic derivation linking the mutation rate (μ₀) in individuals to the evolutionary substitution rate (k) across an entire species population of diploid census size *N* (and effective population size *N*_e):

1. Number of New Neutral Mutations per Generation: In a diploid population of size *N*, there are 2N homologous copies of every autosomal gene locus. If neutral mutations arise at a rate of μ₀ per gene copy per generation, the total number of new neutral mutations entering the entire population each generation is:

1. Probability of Fixation (*P*_fix) of a Single Neutral Mutation: A newly arisen mutation starts as a single copy among *2N* total alleles in the gene pool, giving it an initial allele frequency of *p*₀ = 1 / (2N). Under pure random genetic drift with zero selective advantage or disadvantage (*s* = 0), every allele copy currently in the population has an identical probability of eventually becoming the sole ancestor of all alleles in a distant future generation. Therefore, the probability that a single new neutral mutation eventually drifts to 100% fixation (*p* = 1) equals its initial frequency:

1. Rate of Evolutionary Substitution (*k*): Multiplying the number of new neutral mutations arising per generation by the probability that each individual mutation reaches fixation yields the long-term rate of molecular substitution (*k*) per generation:

This fundamental equation (k = μ₀) is the cornerstone of molecular evolution: for strictly neutral alleles, the rate of evolutionary substitution at the population level equals the neutral mutation rate in a single gamete, completely independent of population size (*N*). In a gigantic population (*N* = 10⁸), vastly more neutral mutations arise each generation (2Nμ₀ is huge), but each individual mutation has a correspondingly tiny chance of drifting to fixation [1/(2N) is tiny]; the two *2N* factors cancel out exactly. Furthermore, Kimura showed that the average time required for a neutral mutation that eventually fixes to travel from appearance (*p* = 1/2N) to fixation (*p* = 1) is 4N_e generations.

A frequent misconception is that the neutral theory predicts all genes or proteins should evolve at the exact same rate. In reality, Kimura defined the neutral mutation rate μ₀ as a fraction (*f*₀) of the total spontaneous mutation rate (μ_T):

where f₀ (ranging from 0 to 1) is the fraction of all newly arising mutations at a locus that are selectively neutral, while the remaining fraction (1 − f₀) consists of deleterious mutations that disrupt protein folding or active-site catalysis and are rapidly purged from the population by purifying (negative) natural selection (since advantageous mutations represent a tiny fraction *f*₊ ≪ 1).

This relationship directly explains both why pseudogenes and synonymous third-codon positions evolve at the fastest rate in the genome (*f*₀ ≈ 1, so *k* ≈ μ_T) and why functionally constrained proteins like histone H4 evolve nearly a thousand times more slowly (*f*₀ ≈ 0.002). It also anchors the molecular clock used in phylogenetics to date speciation divergences across the tree of life.

While Kimura's original 1968 model partitioned mutations into strictly neutral (|s| = 0) and strongly deleterious (|s| ≫ 1/2N_e), two empirical anomalies soon emerged:

1. The Generation-Time Paradox: Spontaneous mutations arise primarily during germline DNA replication errors, which scale with generations rather than absolute calendar years. Short-lived rodents (mice, generation time ~0.3 years) undergo ~50 times more germline cell divisions per year than long-lived primates or elephants (generation time ~20–25 years), yet protein substitution rates per calendar year are much more uniform across mammals than their generation times would predict.
2. Polymorphism Ceiling: Large-population invertebrates (*Drosophila*, marine mollusks) do not exhibit the orders-of-magnitude higher neutral genetic diversity (*θ* = 4N_eμ₀) over small-population vertebrates predicted by strict neutrality.

In 1973, Kimura's colleague Tomoko Ohta resolved these paradoxes by formulating the Nearly Neutral Theory of Molecular Evolution. Ohta demonstrated that the evolutionary fate of a mutation is governed not by its selection coefficient (*s*) alone, but by the dimensionless product |2N_es| (or |4N_es| in diploids), which measures the relative strength of natural selection versus random genetic drift:

• Effectively Neutral Regime (|s| < 1 / [2N_e], or |2N_es| < 1): Even if a mutation is slightly deleterious (*s* < 0, e.g., slightly destabilizing a protein fold or using a suboptimal tRNA codon), natural selection is too weak to overcome stochastic sampling noise, and the allele drifts and fixes as if it were completely neutral.
• Purifying Selection Regime (|s| ≫ 1 / [2N_e], or |2N_es| ≫ 1): Natural selection deterministically purges deleterious mutations and fixes advantageous ones.

Crucially, across biology, body size and generation time are inversely correlated with effective population size (*N*_e): small-bodied mice have short generation times (high μ per year) but huge *N*_e (~5 × 10⁵), meaning purifying selection is hyper-efficient and purges almost all slightly deleterious mutations (small *f*₀). Conversely, large-bodied primates and whales have long generation times (low μ per year) but small *N*_e (~10⁴), widening the "nearly neutral window" (1/2N_e is large) so that more slightly deleterious mutations drift to fixation (large *f*₀). For non-synonymous protein mutations, the opposing effects of generation time and *N*_e partially cancel out, producing an approximately year-constant protein molecular clock!

Because the neutral theory provides a precise mathematical expectation for how sequences evolve in the absence of adaptive selection, evolutionary geneticists test for positive or purifying selection by measuring deviations from the neutral null hypothesis:

The Nonsynonymous-to-Synonymous Rate Ratio (ω = d_N / d_S)

Within a protein-coding gene, nucleotide substitutions are classified into two categories based on the degeneracy of the genetic code:

• d_S (or K_s): The number of synonymous (silent) substitutions per synonymous site, which do not change the encoded amino acid and are assumed to be largely neutral (*d*_S ≈ μ₀).
• d_N (or K_a): The number of nonsynonymous (missense) substitutions per nonsynonymous site, which replace one amino acid with another.

The ratio ω = d_N / d_S quantifies selective pressure on the protein:

• ω = d_N / d_S = 1.0 (Strict Neutrality): Nonsynonymous mutations fix at the exact same rate as silent mutations (seen in pseudogenes).
• ω = d_N / d_S < 1.0 (Purifying Selection): Most amino acid changes are harmful and removed by selection. Across >98% of human–chimpanzee orthologous genes, average d_N / d_S ≈ 0.20 to 0.25, confirming Kimura's thesis that purifying selection and neutral drift dominate genome-wide molecular evolution.
• ω = d_N / d_S > 1.0 (Positive Darwinian Selection): Amino acid replacements fix significantly faster than neutral expectations, a hallmark of evolutionary arms races seen in Major Histocompatibility Complex (MHC) antigen-binding grooves, viral surface glycoproteins (e.g., influenza hemagglutinin and SARS-CoV-2 Spike), and snake venom toxins.

Standing Polymorphism (θ = 4N_eμ) and the McDonald–Kreitman Test

Within a single interbreeding population at mutation–drift equilibrium, Kimura proved that expected heterozygosity (*H*)—the probability that two randomly sampled homologous chromosomes differ at a neutral site—is governed by the population mutation parameter θ = 4N_eμ₀:

In 1991, John McDonald and Martin Kreitman combined intra-species polymorphism with inter-species divergence in a 2 × 2 contingency table (the McDonald–Kreitman or MK test). Under strict neutrality, the ratio of nonsynonymous to synonymous fixed differences between species (*D*_N / *D*_S) must equal the ratio of nonsynonymous to synonymous polymorphisms within a species (*P*_N / *P*_S). An excess of fixed nonsynonymous divergence (*D*_N / *D*_S > *P*_N / *P*_S) reveals the exact fraction of amino acid substitutions (α = 1 − [D_SP_N / D_NP_S]) driven to fixation by positive natural selection.

Does the neutral theory mean that Darwin's theory of natural selection is wrong?

No. Kimura explicitly stated that adaptive morphological, physiological, and behavioral evolution (such as the vertebrate eye, wings, or enzyme active sites) is driven by Darwinian natural selection. What the neutral theory shows is that at the underlying level of DNA and protein sequences, the vast majority of mutations are either deleterious (and rapidly eliminated by purifying selection) or selectively neutral (and fixed randomly by genetic drift).

Why is the neutral substitution rate *k* equal to the mutation rate μ₀ regardless of population size?

In a diploid population of size *N*, the number of new neutral mutations arising each generation is proportional to population size (2Nμ₀), while the probability that any single neutral mutation randomly drifts to 100% fixation is inversely proportional to population size (1/[2N]). When multiplied together to calculate the overall substitution rate (k = 2Nμ₀ × 1/[2N]), the population size *2N* cancels out, leaving k = μ₀.

How does the neutral theory support the molecular clock for dating evolutionary divergences?

If the neutral mutation rate μ₀ per year remains relatively constant along diverging lineages, then neutral substitutions will accumulate linearly with elapsed time *t* since two species split from their last common ancestor (Divergence D = 2μ₀t). By calibrating *D* at synonymous sites or constrained proteins against radiometrically dated fossils, paleogeneticists can estimate when lineages without a fossil record diverged.