Mitochondrial genome of Trichoplax adhaerens supports placozoa as the basal lower metazoan phylum

Abstract

Mitochondrial genomes of multicellular animals are typically 15- to 24-kb circular molecules that encode a nearly identical set of 12-14 proteins for oxidative phosphorylation and 24-25 structural RNAs (16S rRNA, 12S rRNA, and tRNAs). These genomes lack significant intragenic spacers and are generally without introns. Here, we report the complete mitochondrial genome sequence of the placozoan Trichoplax adhaerens, a metazoan with the simplest known body plan of any animal, possessing no organs, no basal membrane, and only four different somatic cell types. Our analysis shows that the Trichoplax mitochondrion contains the largest known metazoan mtDNA genome at 43,079 bp, more than twice the size of the typical metazoan mtDNA. The mitochondrion's size is due to numerous intragenic spacers, several introns and ORFs of unknown function, and protein-coding regions that are generally larger than those found in other animals. Not only does the Trichoplax mtDNA have characteristics of the mitochondrial genomes of known metazoan outgroups, such as chytrid fungi and choanoflagellates, but, more importantly, it shares derived features unique to the Metazoa. Phylogenetic analyses of mitochondrial proteins provide strong support for the placement of the phylum Placozoa at the root of the Metazoa.

Fig. 1.

Scale drawing of the mitochondrial genome of T. adhaerens. The complete sequence of the 43,079-bp mitochondrial genome from a Red Sea isolate of T. adhaerens (59) was determined and annotated by identifying ORFs with the National Center for Biotechnology Information’s orf finder using genetic code 4. Known mitochondrial proteins (blue rectangles) were identified by blast and by alignment to corresponding proteins found in poriferans (NC_006894, NC_006990, and NC_006991), cnidarians (NC_000933 and NC_003522), and the choanoflagellate Monosiga (NC_004309) to infer the start of translation. Genes transcribed in the clockwise direction are shown on the outer circumference; genes transcribed in the counterclockwise direction are shown on the inner circumference. Large (rnLa and rnLb) and small (rnS) ribosomal genes are represented as gray rectangles. The tRNAs (black lines) were identified by using trnascan-se (62) and are annotated by their International Union of Pure and Applied Chemistry (IUPAC) amino acid codes. ORFs encoding unknown proteins >100 aa in length are identified by their amino acid coding capacity (green rectangles). Introns in the cox1 and nad5 genes are shown as red lines connecting exons (blue rectangles). A 103-bp imperfect direct repeat is shown as black triangles. Note that the carboxy-terminal region of cox1 (exons 5–7) is inverted with respect to cox1 exons 1–4 because of the presence of a large 16-kb inversion encompassing the region from trnP to trnV. This inversion has been confirmed experimentally in the Red Sea isolate but does not exist in another placozoan taxon (A. Signorovitch, L. Buss, and S.L.D., unpublished data).

Fig. 2.

Phylogenetic analysis of concatenated mitochondrial proteins. The data set consisted of a total of 2,730 amino acid positions concatenated from 12 mitochondrial protein sequences (atp6 cob, cox1–3, nad1–4, and 4L, 5, and 6). Partitioned Bayesian analysis was performed with mrbayes for 500,000 generations by using four chains and the mtREV amino acid substitution model. Substitution rate variation among sites was modeled by a discrete approximation of the γ-distribution with a proportion of invariable sites (I + Γ). ML analysis performed with paml using the mtREV amino acid substitution model and star decomposition tree search gave an identical tree topology. Posterior probability (Upper) and bootstrap (Lower) values are shown for each node. In these analyses, the output trees were rooted by using the chytrid fungus Monoblepharella.