The Complete Mitochondrial Genome of Rhacophorus dennysi (Anura: Rhacophoridae) with Novel Gene Arrangements and its Phylogenetic Implications

Vertebrate mitochondrial (mt) DNAs form closed circular molecules which have lengths varying from 15 to 21 Kb (Boore, 1999; Sano et al., 2005; Chen et al., 2011; Zhang et al., 2015). These mt genomes typically contain 13 protein-coding genes (PCGs), 2 ribosomal RNA (rRNA) genes, and 22 transfer RNA (tRNA) genes and a control region (CR) (Boore, 1999). The CR is a long non-coding region (approximately 0.5 Kb–9 Kb) (Kurabayashi et al., 2008), which includes signals for regulating and initiating mitochondrial genome replication and transcription and a short non-coding sequence referred to as the light-strand replication origin (OL) (Boore, 1999).

Mitochondrial genes organization is usually conserved in nearly all vertebrates (Boore, 1999). However, the mt gene arrangements of the neobatrachians are different, and a variety of reorganizations have occurred (Kurabayashi and Sumida, 2013; Zhang et al., 2013, 2018; ; ). For instance, there is a lack of the tRNAHis gene in the Odorrana schmackeri mitogenome (); a tandem duplication of tRNAMet gene has been found in the mtDNA of Quasipaa boulengeri (); ND5 gene between tRNASer and ND6 has been translocated to the region between the CR and the LTPF tRNA cluster in Buergeria buergeri (). The phenomena of mt gene reorganizations have generally been interpreted by the Tandem Duplication and Random Loss model ().

In the Neobatrachia, the phylogenetic relationships among Rhacophoridae, Mantellidae, Dicroglossidae and Ranidae remain controversial. Some phylogenetic studies supported the relationship of ((Rhacophoridae + Mantellidae) + (Dicroglossidae + Ranidae))

Abbreviations

PCR, polymerase chain reaction; rRNA, Ribosomal RNA; tRNA, transfer RNA; ATP6, ATPase subunit 6; ATP8, ATPase subunit 8; bp, base pairs; COI-III, cytochrome c oxidase subunit I-III; Cyt b, cytochrome b; CR, control region; H strand, heavy strand; L strand, light strand; mtDNA, mitochondrial DNA; ND1-6, and ND4L, NADH dehydrogenase subunit 1-6, and 4L; OL, L-strand replication; ML, maximum likelihood; BI, Bayesian inference.

(Zhang et al., 2009, 2018; Chen et al., 2011), but others supported a sister-taxon relationship between ((Rhacophoridae+Mantellidae) + Ranidae) and Dicroglossidae (Frost et al., 2006; Pyron and Wiens, 2011; Kakehashi et al., 2013; Kurabayashi and Sumida, 2013; Kurabayashi et al., 2010; Zhang et al., 2013; Li et al., 2014; Xia et al., 2014; Yuan et al., 2016; Chen et al., 2017).

mtDNA is an important molecular marker and has been widely used in the studies of genetics, phylogenetics and phylogeography. In the present study, we determined the complete nucleotide sequence of the mitochondrial genome of Rhacophorus dennysi. We performed phylogenetic analyses based on complete mt genomes of the newly sequenced and other reported species of Ranoidea to assess the taxonomic position of Rhacophorus dennysi, and to test the phylogenetic relationship of Rhacophoridae and Ranidae.

The R. dennysi sample was collected from Qifeng, Guniujiang, Anhui province in China. This frog sample used was stored at -40°C (Sample No. AM12026) in the Conservation Biology Laboratory, College of Life Sciences of Anhui Normal University. Total DNA was extracted from a piece of muscle tissue by proteinase K digestion, phenol-chloroform extraction, and ethanol precipitation (Sambrook et al., 2001).

To determine the complete mitochondrial genomic sequence of R. dennysi, polymerase chain reaction (PCR) was carried out with the primers for the mtDNAs of frogs described in the literatures (Kurabayashi and Sumida, 2009; Zhang et al., 2013). Furthermore, based on the complete mtDNA sequences of R. schlegelii (AB202078) and P. megacephalus (AY458598), we also designed two pairs of primers to amplify mt fragments from the Cytb gene to the ND5 gene. PCR reaction volume of 30 μl contained 21µl sterile double distilled water, 3 µl 10× reaction buffer (with Mg2+), 2.5 µl (2.5 mmol/l) dNTPs, 1 µl each primer (10µmol/l), 0.5µl Taq DNA polymerase (TaKaRa Bio Inc., Otsu, Shiga, Japan) and 1 µl template DNA. Amplification was performed using Applied Biosystems 2720 Thermal Cycler with the following conditions: initial denaturation at 94°C for 4 min, 32 cycles of denaturation at 94°C for 40 s, annealing at 52-58°C for 40 s and elongation at 72°C for 60 s, and a final extension at 72 °C for 10 min. The resulting PCR fragments were separated by electrophoresis in 1.0% agarose gels, then PCR products were purified using TIANquick Midi Purification Kit (TIANGEN Bio Inc., Beijing, China), and then directly sequenced on an automated sequencer (ABI 3730) from both strands.

Nucleotide sequences were checked and assembled using the program SeqMan (DNASTAR Inc., Madison, WI, USA). The 13 protein-coding and two rRNA genes were annotated by comparison with the known complete mtDNA sequences of Rhacophorus schlegelii (Sano et al., 2005), Polypedates megacephalus (Zhang et al., 2005) and Buergeria buergeri (Sano et al., 2004). The 22 tRNA genes were identified by their cloverleaf secondary structure and anticodon sequences using tRNA Scan-SE v.2.0.2 (http://lowelab.ucsc.edu/tRNAscan-SE; Lowe and Chan, 2016). The complete mtDNA sequence of R. dennysi was deposited in GenBank with the accession number KM035412.

In order to address the phylogenetic relationships among Rhacophoridae, 3 additional, previously published Rhacophoridae mitogenomes were included in the analysis. In addition, mitochondrial genomes from one species in the family Mantellidae, twenty-two species in Ranidae, and thirteen species in Dicroglossidae were retrieved from GenBank to further confirm the phylogenetic position of the family Rhacophoridae among Ranoidea. Additionally, three Microhylidae species were used as the outgroups based on Pyron and Wiens (2011) (Table I).

We constructed the phylogenies using the concatenated 12 mt protein-coding genes and partitioned these genes by codon position. The best fitted substitution model for each partition was estimated using Akaike Information Criterion (AIC) implemented in jModeltest v.2.1.7 (Darriba et al., 2012). The GTR+I+G model was chosen for ML and Bayesian inference (BI) analyses, which were separately performed using RaxML (Kozlov et al., 2019) with 1000 bootstrap replications and MrBayes v.3.2.7 (Ronquist et al., 2012). Besides, the following settings were applied in the BI analysis: 10 million Markov chain Monte Carlo (MCMC) generations, a sampling frequency of 1000, burn-in = 1000.

The R. dennysi mt genome was 18,052 bp in length, containing 13 protein-coding genes, 2 rRNA genes, 22 tRNA genes and a control region (Table II). The base composition of the light strand (L-strand) was 31.5.9% A, 31.0% T, 23.2% C, and 14.3% G, which is similar to other vertebrates (Zhang et al., 2015; Li et al., 2016).

Remarkably, the tree frog R. dennysi possessed a novel mitogenomic gene organization much different from other neobatrachians. In the R. dennysi mt genome, the ND5 gene between tRNASer (AGY) and ND6 was translocated to a position between the CR and tRNAThr. In this new mitogenome, four tRNA genes (tRNAThr, tRNALeu(CUN), tRNAPro and tRNAPhe) Formed a TLPF tRNA cluster, different from the neobatrachian-type arrangement (Sumida et al., 2001; Irisarri et al., 2012; Li et al., 2014).

Among the 13 protein-encoding genes in the R. dennysi mitogenome, most of these protein-coding genes started with the common initiation codon ATG except two genes (ND1 and COI) beginning with ATA, one gene (ND2) beginning with ATT. Stop codons were variable for all protein-coding genes. Seven protein genes (ND2, COII, ATP8, ND4L, ND5, ND6 and Cytb) used complete stop codon TAR, and COI ended with AGG, whereas other genes (ND1, ATP6, COIII, ND3 and ND4) ended with incomplete stop codon T.

The noncoding regions in the R. dennysi mtDNA contained the control region and some spacers. The control region was located between the Cytb and ND5 genes with the length of 2,603 bp. The length of the CR in this study is obviously longer than that of R. Dennysi (2,122 bp) in the literature of Huang et al. (2016).

The BI and ML analyses of the molecular dataset produced the identical topologies and very similar branch support (Fig. 1). In the phylogeny of Rhacophoridae, Ranidae, Dicroglossidae and Mantellidae, the monophyly of Dicroglossidae, Ranidae and Rhacophoridae are well supported.

In our tree, the 40 in group species referred in this study were divided into four major clades: Mantellidae, Rhacophoridae, Ranidae and Dicroglossidae, which strongly supported the monophyly of Dicroglossidae, Ranidae and Rhacophoridae. Rhacophoridae was a sister clade to Mantellidae with strong supports (BP = 100, PP = 1.00) and the clade of (Rhacophoridae + Mantellidae) appeared as the sister taxon to Ranidae (BP = 100, PP = 0.92), then together as a sister group of the Dicroglossidae (BP = 100, PP = 1.00). The Rhacophoridae clade included 4 species R. schlegelii, R. dennysi, P. megacephalus, B. buergeri. R. schlegelii and R. dennysi were grouped as the sister clade of P. megacephalus with high supports (BP = 100, PP = 1.00), then together as a sister taxon of B. buergeri (BP = 100, PP = 1.00).

In present study, we discovered a novel gene arrangement of R. dennysi mt genome. The ND5 gene and four tRNA genes (tRNAThr, tRNALeu(CUN), tRNAPro and tRNAPhe forming TLPF tRNA cluster) were located between the CR and the 12S rRNA gene, which differed from the neobatrachian-type arrangement (LTPF tRNA cluster) but shared similarities to those of R. schlegelii (Sano et al., 2005) and P. megacephalus (Zhang et al., 2005), two other species from the same family. Gene rearrangement in animal mtDNA is generally believed to take place through the Tandem Duplication and Random Loss (TDRL) model (San Mauro et al., 2006). According to the TDRL model, a multigene portion of the genome is duplicated, then one copy becomes nonfunctional and is subsequently deleted from the genome.

Generally, gene rearrangements have been considered as useful indicators to resolve some phylogenetic relationships (San Mauro et al., 2006; Zhang et al., 2018). For example, the duplication of tRNAMet likely appeared in all the descendants of Dicroglossidae, indicating that the duplicated tRNAMet genes can be regarded as a synapomorphic character of Dicroglossidae (Alam et al., 2010; Chen et al., 2011, 2017).

Within neobatrachians, the translocation of ND5 was discovered in two distinct lineages: Rhacophoroidea and Mantellidae, and a part of Dicroglossidae (Occidozyga, Fejervarya, Euphlyctis and Hoplobatrachus) (Sano et al., 2004, 2005; Kurabayashi et al., 2008; Alam et al., 2010; Chen et al., 2017). Thus, convergent gene rearrangements occur frequently in non-sister lineages (Kurabayashi and Sumida, 2013). We should require careful consideration when genomic features are employed for phylogenetic relationship.

Phylogeny of Ranoidea (i.e., Rhacophoridae, Mantellidae, Ranidae and Dicroglossidae) has not reached a consensus (Chen et al., 2011; Yuan et al., 2016). In our analyses, the sister-group relationship between ((Rhacophoridae + Mantellidae) + Ranidae) and Dicroglossidae has been well supported (BP = 100, PP = 1.00), consistent with the molecular studies of Kakehashi et al. (2013), Kurabayashi and Sumida (2013), Zhang et al. (2013), Xia et al. (2014), Yuan et al. (2016) and Chen et al. (2017).

In our phylogenetic trees, two species of the genus Rhacophorus (R. schlegelii and R. dennysi) were clustered together with the representative of the genus Polypedates (P. megacephalus), indicating a very close phylogenetic relationship, meanwhile, the representative of the genus Buergeria (B. buergeri) occupied the basal position in the clade of Rhacophoridae. The phylogenetic relationship among the representatives of the family Rhacophoridae revealed here is congruent with the results of most phylogenetic analyses (Yu et al., 2009; Pyron and Wiens, 2011; Zhang et al., 2013, 2018; Chen et al., 2017; Chan et al., 2018).

As a whole, this study provides some evidence for the generic classification of Rhacophoridae proposed by Pyron and Wiens (2011), and our phylogenetic analyses supported the sister-group relationship between ((Rhacophoridae + Mantellidae) + Ranidae) and Dicroglossidae.

This research was supported by the National Natural Science Foundation of China (31872253), the Provincial Nature Science Research Projects of Anhui Colleges (KJ2018A0331; KJ2018A0350), the Government-University Cooperation Projects (XDHX201718; XDHXTD201701; 2017TSJY04) and the Scientific Research Projects of the Inner Mongolian Higher Educational System (NJZY19025).

The authors have declared no conflict of interest.