Contrasting modes of mitochondrial genome evolution in sister taxa of wood-eating marine bivalves (Teredinidae and Xylophagaidae)

Abstract

The bivalve families Teredinidae and Xylophagaidae include voracious consumers of wood in shallow and deep-water marine environments, respectively. The taxa are sister clades whose members consume wood as food with the aid of intracellular cellulolytic endosymbionts housed in their gills. This combination of adaptations is found in no other group of animals and was likely present in the common ancestor of both families. Despite these commonalities, the two families have followed dramatically different evolutionary paths with respect to anatomy, life history and distribution. Here we present 42 new mitochondrial genome sequences from Teredinidae and Xylophagaidae and show that distinct trajectories have also occurred in the evolution and organization of their mitochondrial genomes. Teredinidae display significantly greater rates of amino acid substitution but absolute conservation of protein-coding gene order, whereas Xylophagaidae display significantly less amino acid change but have undergone numerous and diverse changes in genome organization since their divergence from a common ancestor. As with many bivalves, these mitochondrial genomes encode two ribosomal RNAs, 12 protein coding genes, and 22 tRNAs; atp8 was not detected. We further show that their phylogeny, as inferred from amino acid sequences of 12 concatenated mitochondrial protein-coding genes, is largely congruent with those inferred from their nuclear genomes based on 18S and 28S ribosomal RNA sequences. Our results provide a robust phylogenetic framework to explore the tempo and mode of mitochondrial genome evolution and offer directions for future phylogenetic and taxonomic studies of wood-boring bivalves.

Keywords: Marine woodborers; deep-sea; mitochondrial gene order; shipworm; xylophagy; xylotrophy.

Fig. 1.

Wood-boring bivalves of the families Teredinidae and Xylophagaidae. (a) Xylophagaids and (b) Teredinids within their burrows in wood; (c) xylophagaid (Xylophaga dorsalis) and (d) teredinid (Bankia setacea) removed from their wooden burrows. Note the dramatic differences in morphology. cb, calcareous burrow lining; p, pallet; si, siphon; sv, shell valve. Scale bars for (A)–(D) = 1 cm.

Fig. 2.

Graphic comparison of teredinid and xylophagaid habitat preferences, reproductive strategies, anatomy, and mitochondrial genome characteristics.

Fig. 3.

Violin plots comparing genome size, GC-content, and evolutionary (amino acid substitution) rate between Teredinidae and Xylophagaidae. Statistical significance of differences between Teredinidae and Xylophagaidae with respect to each genome property was evaluated using paired t-tests with P values adjusted by Bonferroni methods (P values shown above). (a) Genome size; (b) GC content; and (c) evolutionary (amino acid substitution) rates as measured by the tip-to-root distance. Note that GC-content and amino acid substitution rate, but not genome size, are significantly different at a P value threshold of <0.01.

Fig. 4.

Phylogenetic relationships and variation in mitochondrial gene arrangement among species of Xylophagaidae and Teredinidae. Phylogenetic tree inferred by ML (single-site homogeneous model, unpartitioned, IQ-Tree 1.6.7) based on 4,135 unambiguously aligned amino acid positions selected using Gblocks from an alignment of 12 concatenated mitochondrial protein-coding genes. Bootstrap proportions, less than 100 are indicated at the nodes; where no numerical value is indicated, the bootstrap proportion = 100. Protein-coding gene arrangements associated with each species are depicted on the right. Putative origins of replication are indicated by thick vertical bars. A proposed scenario for the evolutionary history of gene rearrangement, determined by analysis of common intervals using CRex, is superimposed on the phylogenetic tree with branches color-coded according to extant and hypothetical ancestral gene orders. Symbols indicate rearrangements required to transform each hypothetical ancestral gene order to the order observed in the extant species. “/” indicates  transposition, “\” indicates reverse transposition, “<” indicates reversal, “*” indicates tandem duplication with random loss (TDRL).