Phylogenetic and familial estimates of mitochondrial substitution rates: study of control region mutations in deep-rooting pedigrees

Abstract

We studied mutations in the mtDNA control region (CR) using deep-rooting French-Canadian pedigrees. In 508 maternal transmissions, we observed four substitutions (0.0079 per generation per 673 bp, 95% CI 0.0023-0.186). Combined with other familial studies, our results add up to 18 substitutions in 1,729 transmissions (0.0104), confirming earlier findings of much greater mutation rates in families than those based on phylogenetic comparisons. Only 12 of these mutations occurred at independent sites, whereas three positions mutated twice each, suggesting that pedigree studies preferentially reveal a fraction of highly mutable sites. Fitting the data through use of a nonuniform rate model predicts the presence of 40 (95% CI 27-54) such fast sites in the whole CR, characterized by the mutation rate of 274 per site per million generations (95% CI 138-410). The corresponding values for hypervariable regions I (HVI; 1,729 transmissions) and II (HVII; 1,956 transmissions), are 19 and 22 fast sites, with rates of 224 and 274, respectively. Because of the high probability of recurrent mutations, such sites are expected to be of no or little informativity for the evaluation of mutational distances at the phylogenetic time scale. The analysis of substitution density in the alignment of 973 HVI and 650 HVII unrelated European sequences reveals that the bulk of the sites mutate at relatively moderate and slow rates. Assuming a star-like phylogeny and an average time depth of 250 generations, we estimate the rates for HVI and HVII at 23 and 24 for the moderate sites and 1.3 and 1.0 for the slow sites. The fast, moderate, and slow sites, at the ratio of 1:2:13, respectively, describe the mutation-rate heterogeneity in the CR. Our results reconcile the controversial rate estimates in the phylogenetic and familial studies; the fast sites prevail in the latter, whereas the slow and moderate sites dominate the phylogenetic-rate estimations.

Figure 1

Maternal pedigrees relating 61 analyzed individuals shown at the tips of the genealogical trees. The individuals in whom substitutions were found are indicated by a number reflecting the position of the mutated site.

Figure 2

Substitution-density distribution among the sites in the alignment of 973 HVI sequences from Europe (HVI being delimited by positions 16024–1683). Bars represent the counts of the positions characterized by a given site occupancy (i.e., X=0, 1, 2, 3,…; for each position, the occupancy describes the number of sequences in the alignment that carry the less frequent allele). Note that, to accommodate all data, the scales on both axes are not kept uniform. The continuous line is the theoretical curve based on the fitted parameters describing “slow” and “moderate” sites (table 3). Asterisks (*) indicate the sites that were mutated in the familial studies from table 1; the corresponding sequence positions are shown above. The mean mutation density of 48, used to draw the theoretical curve for the “fast” sites, represented the average substitution density among these “familial” sites; we assumed 19 such sites in the whole sequence (table 2). Inverted triangles indicate the sites found to be correlated with the haplogroups; if not rapidly mutating, these sites should be IBD. Inset shows the low-occupancy sites (bars) and a theoretical curve generated according to a single-rate model (eq. [3]) using a cumulative substitution density of the sites up to X=12.

Figure 3

Substitution-density distribution among the sites in 650 European HVII sequences (HVII being delimited by positions 58–370). The mean mutation density used to draw the theoretical curve for the fast sites was 48.9, and their number was assumed to be 22 (according to table 2). The meaning of the bars, asterisks, numbers, and triangles is the same as in figure 2. Inset shows the low-occupancy sites (bars) and a theoretical curve generated according to a single-rate model (eq. [3]) using a cumulative substitution density of the sites up to X=12.