Characterization of the complete mitochondrial genome sequences of three Merulinidae corals and novel insights into the phylogenetics

Introduction

Merulinidae (Verrill, 1865) is a clade of corals that belongs to the order Scleractinia and currently comprises 149 species across 24 genera (Huang et al., 2014a; see also WoRMS Editorial Board, 2019). These species are mainly distributed in the Indo-Pacific and Caribbean regions, but are absent in the eastern Pacific. Initially, Verrill (1865) posited that the family Merulinidae (type genus Merulina) fell within the suborder Fungacea. Merulinidae, however, was not recognized as a valid family by subsequent authors (Quenstedt, 1881; Quelch, 1886; Vaughan, 1918; Hoffmeister, 1925; Faustino, 1927; Matthai, 1928; Yabe, Sugiyama & Eguchi, 1936), until Vaughan & Wells (1943) revived it. Some modifications were proposed successively (Umbgrove, 1940; Chevalier, 1975; Veron, 1985). Nested within the ‘Bigmessidae’ (Budd, 2009), Merulinidae was polyphyletic and its species belonged to multiple divergent subclades (Fukami et al., 2004; Fukami et al., 2008; Huang et al., 2009; Huang et al., 2011; Benzoni et al., 2011; Budd & Stolarski, 2011; Arrigoni et al., 2012; Budd et al., 2012). In recent years, based on the molecular phylogenies by Fukami et al. (2008) and Huang et al. (2011), Merulinidae was redefined and expanded to include all members of ‘Bigmessidae’; Faviidae Gregory, 1900 was demoted to a subfamily (Faviinae) of the new family Faviidae Milne Edwards & Haime, 1957. Meanwhile, Pectiniidae and Trachyphylliidae were regarded as junior synonyms of the family Merulinidae (Huang et al., 2011; Budd et al., 2012).

Members of Merulinidae—namely, Dipsastraea de Blainville, 1830, Favites Link, 1807, and Hydnophora Fischer von Waldheim, 1807—have been closely associated in the past. Dipsastraea had never been applied since it was established. Until the recent revision, Dipsastraea was redefined to include the Indo-Pacific lineages of Favia and all species of Barabattoia (now a junior synonym of Dipsastraea) (Budd et al., 2012; Huang et al., 2014a). Favites is a controversial genus because Favites pentagona (Esper, 1795) is more closely related to species of other genera, indicating that Favites is polyphyletic (Huang et al., 2011; Huang et al., 2014a; Huang et al., 2014b; Huang et al., 2016; Arrigoni et al., 2012). Huang et al. (2011); Huang et al. (2014a)) discovered that F. pentagona clustered more closely with species of Caulastraea, Oulophyllia and Pectinia, than with other Favites species based on three nuclear and two mitochondrial loci (28S rDNA, histone H3, ITS rDNA, mt COI and mt IGR). Huang et al. (2014b) and Arrigoni et al. (2012) showed similar results based on histone H3 and COI/ITS genes. F. pentagona’s morphology, however, is consistent with other Favites species. Hydnophora is a distinct genus, and molecular data have supported its monophyly (Huang et al., 2011; Huang, 2012). All three genera are part of the family Merulinidae, but the current interpretations of phylogenetic relationships between Dipsastraea, Favites, and Hydnophora conflict. For instance, Huang et al. (2014a) concluded that Hydnophora and Favites (with the exception of F. pentagona) were more closely related to each other than to Dipsastraea species. Huang et al. (2014b), however, found that Hydnophora, F. pentagona, and subclade B (including Coelastrea, Dipsastraea and Trachyphyllia geoffroyi) were closely related; other Favites species formed a separate subclade F. A maximum likelihood genus-level phylogeny (576 species) of Scleractinia based on 12 DNA markers suggested that Hydnophora was more closely related to Dipsastraea than to Favites (Kitahara et al., 2016). More work is needed to clarify the evolutionary relationships between Dipsastraea, Favites, and Hydnophora.

In recent decades, mitochondrial DNA (mtDNA) has frequently been used in phylogeny and molecular evolution studies (Boore, 1999; Curole & Kocher, 1999). Mitochondrion is an important eukaryotic organelle, the mitochondrial genome has become highly economized. It typically includes 13 oxidative phosphorylation (OXPHOS) related genes, two rRNAs that encode the two subunits of mitochondrial ribosomes, and an array of tRNAs used for translation within the organelle. Previous studies revealed that the evolutionary rate of Scleractinia’s mitochondrial genome is 10–20 times slower than that of other metazoan taxa, and 5 times slower than that of its nuclear genome (Van Oppen et al., 2000; Chen et al., 2009). Mitochondrial genome rearrangements occur relatively infrequently within Scleractinia. The mitochondrial genome, therefore, played a significant role in scleractinian studies on phylogeny reconstruction (Fukami & Knowlton, 2005; Arrigoni et al., 2016; Capel et al., 2016; Terraneo et al., 2016a; Terraneo et al., 2016b). It could help us explore Scleractinia’s evolutionary process and clarify the evolutionary relationship between Scleractinia and other Hexacorallia members, such as Actiniaria, Antipatharia, and Corallimorpharia (Fukami & Knowlton, 2005; Kitahara et al., 2014; Lin et al., 2014).

In this study, we sequenced the complete mitochondrial genomes of three corals, Dipsastraea rotumana, Favites pentagona, and Hydnophora exesa, using a next-generation sequencing (NGS) strategy. Moreover, we reexamined the phylogeny of Scleractinia based on 86 species across 15 families using the mitochondrial genomes obtained in this study and those available in GenBank. We aimed to (1) describe and compare the mitogenomes of these corals and (2) provide new perspectives on the phylogenetic relationships within the family Merulinidae. The information obtained in this study may facilitate future phylogenetic and molecular evolutionary studies of Scleractinia.

Materials & Methods

Results & Discussion

Genome organization and composition

The complete mitochondrial genomes of D. rotumana (GenBank accession no. MH119077), F. pentagona (KY247139), and H. exesa (MH086217) were 16,466 bp, 18,006 bp, and 17,790 bp in length, respectively. They carried the typical composition, including 13 PCGs, two transfer RNA genes (tRNA^Met, tRNA^Trp) and two ribosomal RNA genes (Fig. 1, Table 1). Length differences were primarily the result of variation in intergenic nucleotides. As found in other Scleractinia species, all PCGs, tRNA, and rRNA genes were encoded on the H-strand. The mitochondrial genome of these three corals were identical to most published scleractinian mitogenomes (Van Oppen et al., 2000; Chen et al., 2008; Wang et al., 2013).

Table 1:

Summary of gene/element feature of Dipsastraea rotumana, Favites pentagona, and Hydnophora exesa.

Gene/ Element	Strand	Size (bp)	GC-Percent (%)	Amino Acids (aa)	InferredInitiation Codon	InferredTermination Codon	Anticodon	Intergenic Nucleotide* (bp)
tRNAMet	H	72	43.06–44.44				UAC (Y)	674–1448
16S rRNA	H	1698–1951	31.08–31.27					0–193
ND5 5′	H	711	32.35–32.84	236	ATG			0–49
ND1	H	948	34.81–35.17	315	ATG	TAG		109–110
Cyt b	H	1140	33.33–33.77	379	ATG	TAA		2
ND2	H	1287	31.88–32.42	428	ATA	TAA		25
ND6	H	561	30.84–31.55	186	ATG	TAA		1
ATP6	H	678	32.45–32.68	225	ATG	TAA		−1
ND4	H	1440	32.24–33.82	479	ATG	TAG		−1
12S rRNA	H	910–911	34.58–35.05					108–138
COIII	H	780	36.79–37.18	259	GTG	TAA		67–125
COII	H	708	33.63–34.6	235	ATG	TAA		626–713
ND4L	H	300	28.67–29.96	99	ATG	TAA		–19
ND3	H	342	31.29–31.76	113	ATG	TAA		2
ND5 3′	H	1104	30.34–30.53	367		TAG		56
tRNATrp	H	70–71	39.44–40.00				AUC (I)	−2
ATP8	H	198	19.70–20.71	65	ATG	TAA		3
COI	H	1590	35.43–35.94	529	ATG	TAA		−1–724

DOI: 10.7717/peerj.8455/table-1

*As for the adjacent genes, positive number represented spaced, negative number indicated overlap.

Mitochondrial gene codon usage

The amino acids Leu, Ser, and Arg, which were encoded by six different codons, appeared more frequently than other amino acids (Fig. 2). For all 13 mitochondrial PCGs of D. rotumana, F. pentagona, and H. exesa, the nonsynonymous/synonymous mutation ratio (Ka/Ks) varied from 0 to 0.5043 (Fig. 3). Nonsynonymous substitutions are generally more harmful than synonymous substitutions. In mitogenomes, some genes may play more important roles than other genes. Therefore, it is reasonable to assume that in order to maintain their function some genes have undergone stronger selective constraints to eliminate deleterious mutations than others. The analysis of the Ka/Ks ratios indicated that the mitochondrial PCGs evolved under strong purifying selection, which signified natural selection against deleterious mutations with negative selection coefficients (Yang & Bielawski, 2000). The Ka/Ks ratios varied for different genes, implying that different genes accumulated different amounts of deleterious mutations. The results showed that ATP8 genes in 13 PCGs presented the highest Ka/Ks values, indicating that the ATP8 gene was under minor selection pressures (Lynch, Koskella & Schaack, 2006).

Phylogenetic analyses

In this study, the Bayesian analysis was constructed based on the 13 concatenated PCG sequences of 86 Scleractinia species (Table S1). The Bayesian phylogenetic tree indicated that all of the interspecific nodes were robust, with strong posterior probabilities (Fig. 4). Species in the same family were grouped together, with the exception of the Polycyathus sp. (NC_015642), which was distinct from other Caryophylliidae species. The family Merulinidae was conventionally clustered into a robust branch and affined with the family Lobophylliidae. Unexpectedly, based on the mitochondrial genome phylogeny, the topological structure of the relationship between H. exesa, F. pentagona, D. rotumana, and F. abdita, was not consistent with previous phylogenies (Huang et al., 2011; Huang et al., 2014a; Huang et al., 2014b; Huang et al., 2016; Katahara et al., 2016). For example, Huang et al. (2014a) found that F. pentagona and D. rotumana shared a close evolutionary relationship, whereas H. exesa clustered more closely with other Favites species according to three nuclear and two mitochondrial markers. Almost all previous molecular studies rendered the genus Favites as polyphyletic. As described above, we also found that, the genus Favites was polyphyletic, with F. pentagona clustered closely with H. exesa, and D. rotumana clustered closely with F. abdita. These results indicate novel interspecific relationships.

Conclusions

In the present study, the complete mitochondrial genome sequences of three Merulinidae corals (D. rotumana, F. pentagona, and H. exesa) were determined for the first time using next-generation sequencing. The gene arrangement and composition of the three corals were identical to each other and consistent with other Scleractinia mitogenomes. The Bayesian inference results showed that the family Merulinidae was clustered into the robust branch, with H. exesa clustered closely with F. pentagona and D. rotumana clustered closely with F. abdita. This study provides reliable mitogenome data and novel insight into the phylogenetic relationships of the genera in the family Merulinidae. Our study may also facilitates future phylogenetic and evolutionary studies of Scleractinia.

Supplemental Information