Proteome size as the major factor determining mutation rates

In an interesting piece of work, Sung et al. (1) proposed that population size is a major factor determining mutation rate. However, as reported previously (2, 3), mutation rates are expected to negatively scale with proteome size (the total number of codons in a genome; P) with an exponent of −1, assuming that the coding DNA in a genome exerts a constraint on genetic fidelity. The influence of proteome size was originally termed the “proteomic constraint,” and was proposed to affect a variety of aspects of genetic fidelity in addition to mutation rates (3). The following relationship was derived:

where μ is mutation rate, and C is the “constraint factor,” which refers to the strength of selection on a genome. The latter may be influenced by population size, but other factors could include recombination rate, variability of the environment, and degree of mutational robustness. The relative contribution of these different factors to the overall strength of selection remains to be determined.

Our conclusions contradict those of Sung et al. (1). In particular, we found that 94% of variation in mutation rates in DNA genomes is attributable to proteome size (3); thus, this appears to be the major factor in determining mutation rates. In contrast, the authors imply that 84% of variation in mutation rates is attributable to population size (figure 1C in ref. 1). However, the number of germ-line divisions in the eukaryotic taxa were not included in the analysis; when included, a weaker relationship of r² = 0.41 is observed (Fig. 1). Thus, the impact of population size and drift in determining mutation rates appears less than claimed. Although the authors also claimed a relationship between proteome size and mutation rates, no analysis was presented to support this assertion. When their data are used to plot proteome size against mutation rates instead of genome size, the second (partially shown) positive correlation among eukaryotic taxa (figure 1A in ref. 1) is abolished (Fig. 2). This difference in plots is because of the larger proportion of noncoding DNA present in eukaryotic genomes. In Fig. 2, both prokaryotes and eukaryotes conform overall to a negative power law relationship with exponent ∼−1 (−0.96), as predicted by Eq. 1. The strength of the correlation (r² = 0.71) again implies that differences in the constraint factor are secondary to proteome size in determining mutation rates. The reanalysis further contradicts the authors’ claim that enhanced drift in eukaryotes is a factor in determining mutation rates.

Fig. 1.

Effective population size versus mutations/proteome size/cell divison, which includes a correction for the number of germ-line divisions in eukaryotes (obtained from ref. 4).

Fig. 2.

Proteome size versus mutation rate, including a correction for germ-line divisions in eukaryotes. Eukaryotes are in red.

Finally, the authors finished with the prediction that larger genomes should have more complex DNA repair. This theory has been tested (5), with the observation that free-living bacteria with smaller proteome sizes are less likely to possess DNA repair genes. Given the apparent importance of proteome size in determining mutation rates, this is also likely to be the primary factor determining the complexity of DNA repair in a genome, and not population size.