Exceptional Enlargement of the Mitochondrial Genome Results from Distinct Causes in Different Rain Frogs (Anura: Brevicipitidae: Breviceps)

Abstract

The mitochondrial (mt) genome of the bushveld rain frog (Breviceps adspersus, Brevicipitidae, Afrobatrachia) is the largest (28.8 kbp) among the vertebrates investigated to date. The major cause of genome size enlargement in this species is the duplication of multiple genomic regions. To investigate the evolutionary lineage, timing, and process of mt genome enlargement, we sequenced the complete mtDNAs of two congeneric rain frogs, B. mossambicus and B. poweri. The mt genomic organization, gene content, and gene arrangements of these two rain frogs are very similar to each other but differ from those of B. adspersus. The B. mossambicus mt genome (22.5 kbp) does not differ significantly from that of most other afrobatrachians. In contrast, the B. poweri mtDNA (28.1 kbp) is considerably larger: currently the second largest among vertebrates, after B. adspersus. The main causes of genome enlargement differ among Breviceps species. Unusual elongation (12.5 kbp) of the control region (CR), a single major noncoding region of the vertebrate mt genome, is responsible for the extremely large mt genome in B. poweri. Based on the current Breviceps phylogeny and estimated divergence age, it can be concluded that the genome enlargements occurred independently in each species lineage within relatively short periods. Furthermore, a high nucleotide substitution rate and relaxation of selective pressures, which are considered to be involved in changes in genome size, were also detected in afrobatrachian lineages. Our results suggest that these factors were not direct causes but may have indirectly affected mt genome enlargements in Breviceps.

Figure 1

Mitochondrial genome organization of afrobatrachians and other anurans. The mitochondrial (mt) genome organization of Breviceps mossambicus and B. poweri determined in this study is compared with that of other afrobatrachians, neobatrachians, and vertebrates reported in previous studies (^aKurabayashi and Sumida [23] and ^bZhang et al. [16]). Genes, pseudogenes, control regions (CRs), and light-strand replication origins (O_L) are shown in boxes. The heavy- and light-strand encoded genes are denoted above and below each gene box, respectively. The boxes do not reflect the actual sizes of the genes and CRs. The single-letter amino acid codes designate the corresponding transfer RNA genes (trns). L1, L2, S1, and S2 indicate trns for Leu (UUR), Leu (CUN), Ser (UCN), and Ser (AGY), respectively. “ψ” shows a pseudogene. Other gene abbreviations are as follows. 12S and 16S: 12S and 16S ribosomal RNAs; CO1–3: cytochrome c oxidase subunits 1–3; Cytb: cytochrome apoenzyme b; ND1–6 and 4L: NADH dehydrogenase subunits 1–6 and 4L. The genes, pseudogenes, O_L, and CRs with duplications and/or rearrangements in afrobatrachians are colored. “Copy” with number shows the duplicated regions within a species. Closed arrows between species indicate the rearranged genes and the presumed evolutionary direction of the translocations. The photos of afrobatrachian species are also provided (excluding Callulina kreffti).

Figure 2

Time tree of anurans. A phylogenetic tree reflecting the divergence ages estimated using a Bayesian relaxed dating method with the 15,093 bp nucleotide data. The tree topology of amphibians is the same as that of the resultant ML and BI trees. Bold branches indicate the lineages leading to the extant anurans. Horizontal blue bars on each node indicate 95% credibility intervals of estimated divergence age. The bootstrap probability (BP) of ML and Bayesian postprobabilities (BPP) are also shown on the right side of each node (BP value/^∗, ^∗∗ > 95 and 99 BPPs), and the calibration points used in the dating analysis are indicated on the corresponding nodes (A to G). The scale of the horizontal axis is in million years.

Figure 3

Changes in the dN/dS ratio (ω) among ranoid lineages. The estimated ω values of neobatrachian branches are shown (based on model 4 in Table 2). The tree topology is the same as those of the ML and BI trees reconstructed in this study. The constant ω (0.053) of nonranoid neobatrachian lineages was regarded as the background value. The estimated ω is shown on each ranoid branch. A high ω indicates the relaxation of purifying pressure. The branches for which ω values are lower (<0.053) and 1.5 times higher (>0.08) than the background are shown in blue and red colors, respectively. The frog taxa having mtDNAs exceeding >20 kbp and 28 kbp are also highlighted by orange and red colors, respectively.