Zhang et al. BMC Ecol Evo (2021) 21:19 BMC Ecology and Evolution https://doi.org/10.1186/s12862-021-01755-3

RESEARCH ARTICLE Open Access Quantifcation and evolution of mitochondrial genome rearrangement in Jifeng Zhang1,2,3,4,5* , Guopen Miao1, Shunjie Hu1, Qi Sun1, Hengwu Ding2, Zhicheng Ji6, Pen Guo7, Shoubao Yan1, Chengrun Wang1, Xianzhao Kan2* and Liuwang Nie2*

Abstract Background: Rearrangement is an important topic in the research of mitochondrial genomes ("mitog- enomes" hereafter), whose causes and mechanisms remain enigmatic. Globally examining mitogenome rearrange- ments and uncovering their characteristics can contribute to a better understanding of mitogenome evolution. Results: Here we systematically investigated mitogenome arrangements of 232 amphibians including four newly sequenced Dicroglossidae mitogenomes. The results showed that our new sequenced mitogenomes all possessed a trnM tandem duplication, which was not exclusive to Dicroglossidae. By merging the same arrangements, the mitogenomes of ~ 80% species belonged to the four major patterns, the major two of which were typical vertebrate arrangement and typical neobatrachian arrangement. Using qMGR for calculating rearrangement frequency (RF) (%), we found that the control region (CR) (RF 45.04) and trnL2 (RF 38.79) were the two most frequently rearranged components. Forty-seven point eight percentage= of amphibians= possessed rearranged mitogenomes including all neobatrachians and their distribution was signifcantly clustered in the phylogenetic trees (p < 0.001). In addition, we argued that the typical neobatrachian arrangement may have appeared in the Late Jurassic according to possible occurrence time estimation. Conclusion: It was the frst global census of amphibian mitogenome arrangements from the perspective of quantity statistics, which helped us to systematically understand the type, distribution, frequency and phylogenetic character- istics of these rearrangements. Keywords: Mitogenomics, Amphibians, qMGR, Rearrangement score, Rearrangement frequency, Phylogenetic characteristics

Background transduction and intermediary metabolism. Te content As semi-autonomous organelles, mitochondria retain of vertebrate mitogenomes is conservative, including 13 their own genomes and participate in many essential protein-coding genes (PCGs), 22 tRNA genes (trns), 2 biological processes in eukaryotic cells such as energy rRNA genes (rrns) and a control region (CR) with repli- cation and gene transcriptional regulatory signals [1, 2]. Because of the moderate evolutionary rate of the mitoge- *Correspondence: [email protected]; [email protected]; [email protected] nome, it has been an excellent molecular marker for phy- 1 School of Biological Engineering, Huainan Normal University, Huainan, logenetic studies [3–6]. Anhui 232001, People’s Republic of China Te gene order of vertebrate mitogenomes is as con- 2 College of Life Science, Anhui Normal University, Wuhu, Anhui 241000, People’s Republic of China servative as its content, such that human, mouse, clawed Full list of author information is available at the end of the article and zebrafsh all share the same gene arrangement

© The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativeco​ mmons​ .org/licen​ ses/by/4.0/​ . The Creative Commons Public Domain Dedication waiver (http://creativeco​ ​ mmons.org/publi​ cdoma​ in/zero/1.0/​ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Zhang et al. BMC Ecol Evo (2021) 21:19 Page 2 of 14

in their mitogenomes [7–10]. However, with an increas- in mitogenome rearrangements has been reported [34]. ing number of diferent species’ mitogenomes completely Tese fndings inspire us to further explore the landscape sequenced, gene rearrangements have been reported in of mitogenome rearrangements in amphibians. many diferent taxa. Generally, these rearrangements are Except for earlier studies on a few species of amphib- inferred to the results from events such as gene transpo- ians, most studies have focused only on a few lineages sition, gene inversion, gene duplication, and gene loss. rather than a more global and systematic analysis. Also, Summaries and comparisons of mitogenome rear- previous studies have paid little attention to the role of rangements have been reported [3, 11]. An initial impor- individual genes in rearrangement. Recently we proposed tant study undertaken by Boore summarized all gene a method, qMGR, for quantifying mitogenome rear- arrangements known for Metazoa, based on no more rangement and developed a web service (http://qmgr. than 100 mitogenomes available at that time [1]. Further hnnu.edu.cn/) that provides large-scale and accurate comparative studies of mitogenome arrangement have analysis of mitogenome rearrangement information [35]. subsequently been published (e.g., [12–15]). In addition, In this study, we newly determined mitogenomes for several possible rearrangement models were proposed four in the neobatrachian frog family Dicroglossi- to explain mechanisms, including tandem duplication– dae, and found that they all had a tandem duplication of random loss (TDRL) [16, 17], tandem duplication and trnM (IQMM trn cluster), which was not a feature exclu- non-random loss (TDNL) [18, 19], intramitochondrial sive to this family [36, 37]. To identify common charac- recombination [20, 21], etc. Using the CREx and TreeREx teristics of mitogenome rearrangement in amphibians, programs, based on four rearrangement models, other we then focused on the study of gene rearrangement studies have tried to fnd common intermediate steps patterns of all known amphibian mitogenomes, quanti- (common intervals) in two or more gene orders being fed the rearrangement frequency (RF) for each single investigated to elucidate the evolution of mitogenome gene and the rearrangement score (RS) for each mitog- rearrangements in diferent lineages such as insects and emome by qMGR, detected phylogenetic characteristics crabs [22–25]. of species with identical mitogenome arrangements, and In general, most retain the conserved rather estimated possible time for rearrangement patterns. Our than rearranged mitogenome components and order. Te fndings contribute to understanding characteristics and conservative arrangement is called the “typical vertebrate evolution of mitogenome rearrangement in amphibians. (mitogenome) arrangement” in vertebrates and the “typi- cal invertebrate (mitogenome) arrangement” in inverte- Results brates [1, 26]. To date, the number of vertebrate species trnM tandem duplication of amphibian mitogenomes with typical vertebrate arrangement accounts for more Lengths of the four newly sequenced mitogenomes are than half of all species for which mitogenomes have been 18,520 bp (Quasipaa robertingeri), 16,640 bp (Limnon- determined. Among the amphibians, neobatrachians are ectes fragilis), 18,154 bp (Limnonectes fujianensis (Tai- the majority of frogs, accounting for ~ 92% of the total, wan)) and 18,293 bp (Limnonectes fujianensis (Fujian)), over 6600 species (AmphibiaWeb, http://www.Amphi​ respectively, and the full range of their GC contents biawe​b.org/, accessed February, 2020). Most sequenced was relatively narrow (39.7–42.9%). Te four new neobatrachian frogs possess a derived mitogenome mitogenomes all contain a tandem duplication of trnM arrangement, “typical neobatrachian arrangement” [14, (IQMM trn cluster) that also occurs in other sequenced 27, 28]. Furthermore, there are some other types of rear- dicroglossids [14]. rangements in addition to typical neobatrachian arrange- Among the 35 amphibian mitogenomes presenting ment in amphibians. For example, the mitogenome of the evidence of duplication or loss of genes as well as CRs neobtrachian frog Limnonectes bannaensis lacks trnA, (excluding gene rearrangement with the same total num- trnN, trnC and trnE and contains a tandem duplication ber of genes), 19 species (including our four new mitog- of trnM [29], the mitogenome of the caecilian Crotapha- enomes) possess a tandem duplication of trnM (IQMM trema lamottei includes the duplications of trnF, trnP and trn cluster) (see Table 1). Tese 19 species represent 16 trnT and lack of trnK [30], and there are two CRs found species of Dicroglossidae, two species of in the mitogenomes of the neobatrachian frogs Mantella (the non-neobatrachian frog family), and one species madagascariensis [31] and Rhacophorus schlegelii [32]. of the neobatrachian Ceratobatrachidae. Tus, tandem In addition, while nad6 is located between nad5 and cob duplication of trnM is not exclusive to Dicroglossidae, genes within typical vertebrate mitogenome, it is rear- and more than one rearrangement event is required ranged to between rrnL and nad1 in the mitogenome of to explain this pattern given that these families are dis- the plethodontid salamander Aneides hardii [33]. Fur- tantly related to each other [38, 39]. Te frogs Lepto- thermore, in some amphibians intraspecifc variation lalax oshanensis (I-Q-M-V-P-M-nad2) and Mantella Zhang et al. BMC Ecol Evo (2021) 21:19 Page 3 of 14

Table 1 List of amphibian mitogenomes with two trnM genes in this study Species Taxon Accession Nos

Platymantis vitianus Anura; Ceratobatrachidae NC_027671 Euphlyctis hexadactylus Anura; Dicroglossidae NC_014584 Fejervarya cancrivora Anura; Dicroglossidae NC_012647 Fejervarya limnocharis Anura; Dicroglossidae NC_005055 Hoplobatrachus rugulosus Anura; Dicroglossidae NC_019615 Hoplobatrachus tigerinus Anura; Dicroglossidae NC_014581 Nanorana parkeri Anura; Dicroglossidae NC_026789 Nanorana pleskei Anura; Dicroglossidae NC_016119 Nanorana taihangnica Anura; Dicroglossidae NC_024272 Occidozyga martensii Anura; Dicroglossidae NC_014685 Quasipaa boulengeri Anura; Dicroglossidae NC_021937 Quasipaa robertingeri Anura; Dicroglossidae this study Quasipaa spinosa Anura; Dicroglossidae NC_013270 Quasipaa yei Anura; Dicroglossidae NC_024843 Limnonectes bannaensis Anura; Dicroglossidae AY899242 Limnonectes fragilis Anura; Dicroglossidae this study Limnonectes fujianensis (Fujian & Taiwan) Anura; Dicroglossidae this study Mantella madagascariensis* Anura; NC_007888 Leptobrachium boringii Anura; Megophryidae NC_024427 Leptolalax oshanensis* Anura; Megophryidae NC_020610 Oreolalax major Anura; Megophryidae NC_030605

* All species with two trnM are tandem duplication except Leptolalax oshanensis and Mantella madagascariensis madagascariensis (nad5-I-M-L2-P-F-rrnS-V1-rrnL-L1- as Patterns 1 ~ 4 in descending order of the frequency T-nad1-M-CR) have two separate copies of trnM rather of rearrangement types. And we identifed two seg- than a tandem duplication (Additional fle 1: Table S1). ments that were subject to highly frequent rearrange- Te diference in position suggests that they may pos- ments as Region 1 and Region 2, respectively. Between sess diferent mechanisms of occurrence [29, 31]. Te them, Region 1 covered the nad5-nad6-cob gene cluster amphibian species with gene duplication or loss (involved and CR, and Region 2 was mainly composed of IQM trn with other tRNA genes, rRNA genes, PCGs and CRs) are cluster and WANCY trn cluster. Compared with typical also shown in Additional fle 1: Table S1. vertebrate arrangement (Pattern 1), 55 neobatrachians shared the second most frequent pattern of mitogenome Diferent patterns of mitogenome rearrangement arrangement, known as typical neobatrachian arrange- in amphibians ment, which had an LTPF trn cluster in their Region 1, We subsequently investigated more rearrangement and we found that all neobatrachians belonged to Pat- types of 232 amphibian complete mitogenomes fltered tern 2. Due to the existence of novel IQMM trn cluster in by available annotation information. Among them, 121 Region 2, the six species of Dicroglossidae including the species’ mitogenomes represented typical vertebrate new mitogenomes determined were distinguished from arrangement, and the remaining 111 species possessed typical neobatrachian arrangement and formed a new 43 types of gene arrangement, when we examined the pattern, Pattern 3. Pattern 4 occurred in three species of arrangements of all mitochondrial (mt) genes involved, the genus Odorrana, whose HLTPF trn cluster replaced which were defned as global arrangement. Teir LTPF trn cluster of neobatrachians in Region 1. arrangements were summarized in Additional fle 1: Fig- PCGs play an important role in mitogenome rear- ure S1. However, ignoring the species with rare arrange- rangement in vertebrates [40, 41]. To reduce the com- ments (only one or two occurrences), they all belonged plexity of mitogenome rearrangement caused by RNA to major arrangements that occurred at least three times genes and CRs, we further investigated the local arrange- (irrespective of phylogeny) in amphibian mitogenomes, ment (not all genes involved) limiting analyses on only as shown in Fig. 1a. In the fgure, there were only four PCGs. Figure 1b shows the 3 patterns (Pattern 1 ~ 3) of patterns in these global arrangements. We labeled them PCGs arrangements (frequency ≥ 3), which only the frog Zhang et al. BMC Ecol Evo (2021) 21:19 Page 4 of 14

a

121 species (74 of Caudata, 28 of Gymnophiona, and 19 of Archaeobatrachia)

a8 cox3 nad3 R nad4L nad4 H S2L2 nad5 T F V nad1 I M W cox1 D cox2 Pattern 1 a6 G cob CR rrnS rrnL L1 nad2 K nad6 E P Q A N C Y S1 Region 1 Region 2 55 species (Nebatrachia)

nad3 R nad4 H nad5 L2 T F nad1 I M W D K Pattern 2 a8 a6 cox3 G nad4L S2 cob CR rrnS V rrnL L1 nad2 cox1 cox2 nad6 E P Q A N C Y S1

6 species (Dicroglossidae)

nad3 R nad4 H nad5 L2 T F nad1 I M M W cox1 D cox2 K Pattern 3 a8 a6 cox3 G nad4L S2 cob CR rrnS V rrnL L1 nad2 nad6 E P Q A N C Y S1

3 species (Odorrana)

nad3 R nad4L nad4 nad5 L2 T F V nad1 I M W cox1 D cox2 K Pattern 4 a8 a6 cox3 G S2 cob CR H rrnL rrnL L1 nad2 nad6 E P Q A N C Y S1

Global arrangement (all genes) b c

210 amphibians 126 species (77 of Caudata, 30 of Gymnophiona, and 19 of Archaeobatrachia) nad1 nad2 cox1 cox2 a8 a6 cox3 nad3 nad4L nad4 nad5 cob Pattern 1 nad4 H S2L2 nad5 T F nad6 cob CR rrnS Pattern 1 nad6 E P

14 species () Region 3 62 species (Nebatrachia) nad1 nad2 cox1 cox2 a8 a6 cox3 nad3 nad4L nad4 cob nad5 Pattern 2 nad4 H nad5 L2 T F nad6 S2 cob CR rrnS Pattern 2 nad6 E P

5 species (3 of Caudata and 2 of Archaeobatrachia) 3 species (Megophryidae) nad1 nad2 cox1 cox2 a8 a6 cox3 nad3 nad4L nad4 nad5 cob nad4 H S2 L2 nad5 T W F Pattern 3 Pattern 3 cob CR rrnS nad6 nad6 E P

3 species (Fejervarya) nad5 L2 Local arrangement (PCGs) nad4 H S2 cob CR T F rrnS Pattern 4 nad6 E P d 3 species (Odorrana) 214 amphibians 212 amphibians nad4 nad5 L2 T F S2 cob CR H rrnS nad2 W cox1 Pattern 5 nad6 E P nad1 I M nad2 Pattern 1 A N C Y Pattern 1 Q

4 species (Microhyloidea) 19 species (16 of Dicroglossidae, nad2 W cox1 Local arrangement (Region 1) 2 of Megophryidae, Pattern 2 N A C Y and 1 of Ceratobatrachidae) Pattern 2 nad1 I M M nad2 3 species (Megophryidae) Q nad2 cox1 Pattern 3 A N C Y

Local arrangement (Region 2) Fig. 1 The patterns of global and local mitogenome arrangement in amphibians (occurrence frequency 3 times among sampled taxa). The ≥ number of occurrences and categorization are shown above each illustration of patterns, main rearranged genes or regions are marked with color, red dashed box or underline). a Global arrangement patterns involved with all mt genes. b Local arrangement patterns only involved with PCGs. c Local arrangement patterns in Region 1. d Local arrangement patterns of IQM region and WANCY region in Region 2. Abbreviations of mt genes follows ref. [14]

Hoplobatrachus rugulosus, a species with double mito- major diferences among the 3 patterns related to the chondrial nad5 genes (only present once among sampled gene orders of nad5, nad6, and cob. Terefore, we consid- amphibians) was excluded. Among them, 91.3% of the ered that the nad5-nad6-cob segment may be “an active sampled amphibians shared the frst pattern (Pattern 1) region” of mt PCGs rearrangement, and the segment was of PCGs arrangement, while the other 19 species from defned as Region 3. Anura and Caudata belonged to the remaining two pat- Figure 1c and d show the patterns of mitogenomic terns, implying that the probability of rearrangement local arrangements that occurred at least three times involved proteins is much smaller than other genes. Te in Region 1 and Region 2, respectively. Te Region 1 of Zhang et al. BMC Ecol Evo (2021) 21:19 Page 5 of 14

Fig. 1c included two extra trn clusters: a TPW trn cluster Accurate quantifcation of rearrangement frequency at the 5′ end of CR and a TPLF trn cluster at the 3′ end of qMGR [35] is a method for large-scale and rapid quanti- CR, which was derived from 3 species of Megophryidae fcation of mitochondrial genome rearrangements. Com- and 3 species of genus Fejervarya, respectively, compared pared with the arrangement of the reference genome, it with Fig. 1a of mitogenomic global arrangements. Te can calculate the rearrangement score (RS) of each gene left part of Fig. 1d shows that a novel local arrangement, in each genome one by one based on the changes of genes the IQMM trn cluster, was derived from 19 species out on its both sides, and accumulate RS of all genes in this of 3 families (all within Anura), namely Dicroglossidae, genome to obtain the genome’s RS. In a given taxonomic Megophryidae, and Ceratobatrachidae (see also results of group, dividing the actual RS of a gene by its maximum trnM tandem duplication above). Similarly, we also found possible RS can be used to extrapolate the relative rear- other rearrangement patterns in the WANCY region of rangement frequency (RF) of the individual gene. qMGR Region 2, the pattern of ‘WNACY’ was present in 4 spe- was used to accurately calculate the RF of amphibian cies of Microhyloidae and the pattern of ‘ANCY’ was pre- mitochondrial genes. Figure 3a and b show the RF dis- sent in 3 species of Megophryidae (whose mitogenome tributions of all individual genes of amphibians and their possessed the TPW trn cluster). three consistent orders, respectively. As shown in Fig. 3a, Compared to major arrangements, 47 species with RF of most genes were less than 10 (%), the genes with rare arrangements accounted for 42.3% of all rearranged high RF (%, coloring deeper) were mainly concentrated amphibian mitogenomes including 7 pairs of species with in the nad5-CR segment and IQM-WANCY segment double rearrangements (the same rearrangement type (namely, Region 1 and Region 2 mentioned above), con- appeared only twice in all sampled species) and 33 spe- sistent with previous fndings for rearrangement hot- cies with single rearrangement (this rearrangement type spots in vertebrate mitogenomes [16, 42]. For each single is unique in all sampled species) (Fig. 2). gene, CR was assigned the highest RF(%) with a score of

Platymantis vitianus nad4 nad5 S2 cob L2 H F rrnS V rrnLL1nad1 I M M nad2 T W cox1 nad6 E P Q CR A Y C N

Typical vertebrate

40 nad4 HS2L2 nad5 cob T F rrnS V rrnLL1 nad1 I M nad2 W cox1 nad6 E P CR Q A N C Y arrangement

Single rearrangement Double rearrangements

30 •

• 20

• Pattern 4 Species RS

• average RS (=12.9) • 0 Pattern 3 • Pattern 2 (Typical • neobatrachian arrangement)

• Pattern 1 (Typical 01 vertebrate arrangement)

Tylototriton verrucosus Boulengerula taitana Rhyacotriton variegatus cob T P P CR cob T T P Fig. 2 Scatter plot of RS of species with rare arrangements (Single rearrangement and Double rearrangements). The red dotted line represents the average RS of 47 species with rare arrangements; The black dotted lines indicate RS of species belonging to four major patterns (global arrangement). The fgure also shows the species with the highest and the lowest RS and their rearrangement information, and the names of genes with two changed fanking genes are marked in red, the names of genes with changes of one-sided gene are marked in purple Zhang et al. BMC Ecol Evo (2021) 21:19 Page 6 of 14

a T b Anura CaudataGymnophiona CR CR CR CR cob CR F rrnS rrnS rrnS rrnL rrnL rrnL nad6 P a6 a6 a6 E a8 a8 a8 cox1 cox1 cox1 rrnS V cox2 cox2 cox2 051 10 0 cox3 cox3 cox3 nad5 cob cob cob nad1 nad1 nad1 nad2 nad2 nad2 nad3 nad3 nad3 13 PCGs nad4 nad4 nad4 L2 rrnL nad4L nad4L nad4L S2 nad5 nad5 nad5 232 amphibians nad6 nad6 nad6 H A A A C C C L1 D D D E E E nad4 0.10 F F F nad1 G G G 0.05 H H H

Density I I I 0 •••••••••••• ••• • •• •• • • Q K K K I L1 d4L 0 10 20 30 40 50 L1 L1 na L2 L2 L2

M 22 trns M (%) M M R RF N N N nad3 P P P Conserved segme nad2 Q Q Q G A R cox3 R R CN S1 S1 S1 Y S2 S2 S2 T T T a6 S1 V V V W W W W a8 cox2 cox1 Y Y Y n t K 050 100 01530 01530 D RF (%) RF (%) RF (%) Fig. 3 RF distributions of mt genes of all sampled amphibians (a) and species of three orders (b) using the qMGR method. In (a), the outer ring shows the mitogenome arrangement of most amphibians (namely, typical vertebrate arrangement). Linear density distribution is in its inner bottom part, and the depth of colors corresponding to the diferent values of RF are placed in its inner upper part, and the higher the value, the darker the color. The control region (CR) and trnL2 gene are the two most frequently rearranged components. We also mark three or more contiguous conserved genes with RF of zero by dotted lines; In (b), the histograms indicates that RF of each gene in three separate groups: frogs, caecilians, and salamanders

45.04 by using qMGR [35], indicating that about 45% of tried to choose it as the benchmark for calculating RF. the fanking genes of CR have been changed, that is, the Te results showed that the genes with changed RF were changes have been accumulated more than 200 times those rearranged genes relative to typical vertebrates (trnP and trnF are located at the 5′ and 3′ end of CR, arrangement (Fig. 4). Interestingly, all of their scores had respectively). trnL2 was ranked second with a score of declined, but the relative rank of their RF changes lit- 38.79, and nad5 had the highest RF(%) among all PCGs, tle. Te top 3 of them were both CR, trnL2, and nad5 in with a score of 24.14. In addition, according to the def- order, implying that they are the most active elements of nition of RF, if consecutive adjacent genes all have the mitogenome rearrangement in frogs. smallest RF value (here 0), we believed that they them- To further investigate the rearrangement degree of selves formed a rearranged conserved segment in the species with rare arrangement, based on qMGR [35], we mitochondrial genome. We identifed the two most con- also calculated their species RS, which was the cumula- served segments in amphibian mitogenomes as atp6- tive value of the rearrangement scores of all genes in the cox3-G-nad3-R-nad4L and S-D between cox1 and cox2. genome of a given species. As shown in Fig. 2, we found In Fig. 3b, there were fewer genes with a non-zero RF that the species with the highest RS is the neobtrachian (< 50%) and with relatively low scores (< 15) in Caudata frog Platymantis vitianus (species RS = 32), which had and Gymnophiona. In comparison, RF of most genes 17 genes involved in rearrangement compared with typi- (27 of the total 38) in Anura were greater than zero, and cal vertebrate arrangement. Te species with the lowest some of them were greater than 30. Te primary reason scores, consisting of the salamander Tylototriton verru- was that most species with rearranged mitogenomes cosus, the salamander Rhyacotriton variegatus and the belonged to typical neobatrachian arrangement in this caecilian Boulengerula taitana (species RS = 2), had only group (56.3%). Also, RF of anuran trnL2 and CR were one rearranged gene, which was a tandem duplcation both greater than 50, mainly due to the fact that their of trnT, trnT, and trnP, respectively (Fig. 2). Te average rearrangements were accompanied by > 50% changes in RS of all species with rare arrangements was 12.9, which adjacent genes on both sides compared to typical verte- was slightly higher than the species with typical neoba- brate arrangement. trachian arrangement (Patten 2). In addition, Fig. 2 also Considering that typical neobatrachian arrangement is shows RS of the species belonging to the other three the dominant gene order in frogs’ mitogenomes, we also global arrangement patterns. Zhang et al. BMC Ecol Evo (2021) 21:19 Page 7 of 14

Anura Benchmark: typical vertebrate arrangement RF(%) Benchmark: typical neobatrachian arrangement 1 1

0.8 2

0.6 3 5 4 67 1 8 0.4 2 3 5 4 7 0.2 9 8

0 S2 L2 nad5 cob TPCR F

Typical vertebrate nad4 H S2L2 nad5 cob T F rrnS arrangement (partial) CR nad6 E P

Typical neobatrachian arrangement (partial) H S2 nad5 cob L2 T F rrnS nad4 CR nad6 E P Fig. 4 Histogram of rearranged genes’ RF in anuran mitogenomes based on two diferent benchmarks, and comparison of two typical arrangements (partial) below it. In the fgure above, the number above strips represents rank of RF of single gene in all genes sorted in descending order. In the segments rearranged of the fgure below, and the color of the gene names represents the same meaning as in Fig. 2

Phylogenetic characteristics of mitogenome reports (e.g., [38, 44, 47, 48]). Overall support was high rearrangement in amphibians within all species. 84% of the nodes had a bootstrap value For phylogenetic characteristics of rearranged amphib- (BS) ≥ 75% (Fig. 5), and 98% had Bayesian posterior prob- ian mitogenomes, the optimal ML tree and BI tree of 232 abilities (BPPs) ≥ 0.90 (Additional fle 1: Figure. S2). Our amphibians were shown in Fig. 5 and Additional fle 1: results suggested that the superfamily and family-level Figure. S2, respectively, and they had similar basic topol- relationships of Neobatrachia were mostly consistent ogies. In them, the monophylies of the three amphibian with most previous studies (i.e., superfamily Ranoidea groups (frogs, caecilians, and salamanders) were strongly includes Ranidae, Dicroglossidae, , Mantel- supported, as found in most phylogenetic analyses [38, linae, , etc.) [38, 44, 49, 50]. Additionally, 43–46]. Within frogs, the relationship of two higher- BS and BPPs also strongly supported most families of level groups, Neobatrachia and Archaeobatrachia (non- salamanders and caecilians, for instance, Salamandridae, neobatrachian frogs) was consistent with most previous Plethodontidae, and Hynobiidae within Caudata [51], Zhang et al. BMC Ecol Evo (2021) 21:19 Page 8 of 14

s

i s

P

s

u

t

s

n

Batrachuperus_londongensis Batrachuperus_yenyuanensi sis a

e

e l n

u

g u

d

n c

o Batrachuperus_tibetanus Batrachuperus_pinchoni e a ii nge

h

h Hyno Hynobius_qu g m

y

c Hynobius_maoershanensis la i n o

n i

o y

u v

Hynobius_nigrescen o

h a ct s b l

b s f Hynobius_kimura serl

i

ius_guabangshanensis u tu _ angche Hynobius_nebulosus _ y ida a s s s r i s t _

u Batrachupe u

i H i eg Hynobius_yangi p Hynobius_chinensis _ke i

b b ltzi Hyno r Hynob ynobius_leechi u Ony a o o o s Hynobius_amj lla_ a x

Hyno e n n e v ighti Batrachuperus_mustersi i yeri i

o bius_sh i ch cu y r lpaer y rell _ m chodacty s

n d

h bius_yiwuensis h d s

g ndr o anguinus ius_formosanus o on _w bius_arisanensi t ch ubrichti ntris e us d _ _fus tatu Onyc Onychodactylus d avipunctatuss t har s rus_gorga n s tensis u u e ve Ranodon_ u rum

s

e e h o

i

s s laman a_ h lus_zhangyapingi s ongatu s ht e i

P a n o P cturus_be nigri s Pseudohynobius_jinf oteu Liua_shihi g Liua_tsinpaensi des_fl Luet d Salama nat S e ac i Pachyhyno g ensis i Pr ei M N sati ps_ wri Andrias_ja Rhyacotri on_el icroc k S Andr ty Aneides_ dromantes_brunus enotyph lus_zhaoermi n An ctylium_scu ip s ensi En esmognathussmo n._sp.i h i Hy hod a o i bi D t ida oseps_attenuaa_ a a _ s De h ecilia nops_annulatuss r e m s us Microcaecilia_unicolo_da fischeri icus s Phaeognathus_h c c lus_brasi Pl a _poelz s p * He _ * n._sp.islineat v onicus * Batrachose oglos dermatophag Gymnopis_mul idianu Batr it b rphyriticusnatu * _ _palleu i ni Mic Batrachosol eps_ o sis Sc Dermophis_mexicanus i n * B orius_cea e rocaecilia_sp l histometopum_gre iensi s Oedipina sis Caudata (84)Th ophilus Geotrypetes_ * Eury us_marg s seudotriton_rubern * P a wenxian tipl Gyri Hypogeop r rucosus nsis Grandisonia_sechPras i Gyrinophilus_ponotriton_anders ensis e cata . i Gege * Stereochiltotritonton_t_ aliangen Indotyphlus_mah lini s Ech o er g 100 yl otri ton_ver eichow a * 100 w 100 a ori 0 neophis_ra h 100 T _coop 100100 61 lot hongkong is_rostratusphin 97 *

0 i 100 1 99 Ty lototri 9 78 51 1 i 00 100 riton_ 100 96 10 3 Ty e 1 63 53 s ri 73 Potomotyphlus_kaupiel 10 _feii 00 68 Tylototriton_shanjing maswamil 0 lototriton_kn s a ensi 61 r 100 100 Chthonerpeton_indTyp ashtraensi 0 Ty rito 100 62 s 100 10 6 0 2 h 100 100 Paramesot lone 6 8 6 * armoratuicus i 0 10 100 Pachyt ct 100 Caecilia_tentaculates_natans s 0 Triturus_pygmaeu s 66 0084 100 istinctum i 1 6400 100 Triturus_m 10 100 61 86 1 67 Caecilia_gra * 87 0 Triturus_dobrog 99 100 0 10 atus 100 78 100 Triturus_carnifet x s Oscaecilia_oCaecili 10 i a_volc a Triturus_macedonicurus_cris cili 0 73 itu karelini 100 r _perstriatu Bo c a s 100 T rus_ ifi ulengeruhrocephalal ni 56 8610000 _at Boulenger _ 10 10 8 511 Tritu ra 10100 0 4 tophthalmusnd i u a_taitana 0 0 00 No alama He la_bo 100 1 ias a_bishop rpele_sq ulen 100 100 59 00 100 Lyc e u ger 1 00 alostom i 1 100 Ambystom niense Crot Chikila_ful a 10 99 101000 mbystoma_lateralalifor ap 100 0 51 A stoma_c m hatrema leri * 100 1 0070 70 grinu Scolecomor _lamott 1 00 1 Amby toma_ti ph ei 00 Ambys mexicanum us_vit 68 toma_ Uraeotyp tatus Ambys ndersoni hlus_cf. 100 65 ystoma_a Uraeo * 93100 83 Amb ii typhlus_gansi 100 100100 toma_dumeril Ichthyophis_b 90 Ambys ombayensis 100 100 93 100100 Ambystoma_texanum Ichthy 100 ophis_glutinosus Gymnophiona (33) 100 47.8% Ambystoma_unisexual 100 100 Outgroups 100 Ichthyophis_bannanicus 100 100 Ambystoma_barbouri Rhinatrema_bivittatum Heleophryne_regis Chinemys_reevesi 51 * Hyperolius_marmoratus a_iguana 100 0 93 Trichobatrachus_robustu Iguan 10100 40 1000 * s 100 9393 Microhyla_ Xenopus_laevis 48 10000 * butleri nus 100 100 56 Microhyla_ Xenopus_victoria 100 M ornata orealis 83 10 icrohyla enopus_b s 000 Microh _heymonsi X ropicali 89 1001 yla_o enopus_t hoi 00 00 Microhyl kinaven X rval 716 1 77 a_pul sis Pipa_ca eri 99 96 100 Kaloul chra 100 9 96 Kaloula_boa_pulc erlini 7 0.08 hra menochirus_boettg s Kaloula_rugire Hy 10010 alis hymenochirus_dorsali_m ata 00 829600 65 1000 Mant hrynus 1 1 Buergeria_buergerella_ fera Pseud _varieg is 0 100 0 * ma Rhinop ina al 10 Rh dagas omb _orient 10 81 * aco B 100 Rh ph carie ombina anensisalis 67 * aco or nsi B _lichu a 7 10 R phoruus_sc i s 59 * ugosa bina tor 10 10 10 0 Glandirana_tie hl m 0 _e s_dennysegeli Bo ina_fortinupti ma 0 1 0 Amol m ii deladigi oi 1 0 el cro 0 1000 A jano Bomb 0 0 moloops galgan 100 Am _wuyi vi ombina_maxis_ r Anura (1 919 969 0 p nta B u * 0 36 59 A o s_ricketti ie ombina_mi ss s 100 86 55 96 lops e n B o 00 00 87 m n sis x_majo * 100 10 97 Ran ol sis i 1001 00 _mantzorum a 0 0 o es_obstetricans ngi * 1 626 1 9 Rana_nigrops_l anensii 000 a_p Discogllyt u is 100 0 9 A * 10 10 Pe o 606 Oreolallich is 100 0 333 la loe _ s 100 0 Pelophylop n 0 00 4 n n s 0 989 c e 101 10010 y si hium_borhanens 00 80 54 Pe hylax_ i s c s ipe 101 71 1 ing 10081 0 Pelol m reolalax ap atus 7 0 o acul O 99 ph h i * 70 5 Pelo la s 0 100 x er * P p yl _epeiroticuhqipericuat Leptobratolalax_o t i 10010 h ax_c a 0 10 1 80 elop phylax_cylaxf rys_s 100 00 Pe Lep h set 0 0079 p f._punca_atrcheyi 15) 101 P 99 1 . m 0 9 e lo h _ CR0 no 101 Pel ph yl bedr s s_c 0 a Pelobates_cultrel mas 0 lophylax_cf s 00 p te 100 Pelophyl a hochsteto 100110 yla 989 3 s _ anus 0 101 0 x_ i is 0 00 o th 101 P i 1001 1 p ._b. agae ody _ iv s 00 100 c l 100 el x Atym Leiopma s 0 Babina_ hyl f._b.MPFC1 us 1 _cf._b.A Pe u c 101 0 Babi op GM178 s chusmisensis rien 100 0 a _bol 0 _ nlingensiu Hyl hyla x_ iopel ns 101 os i 0 Babina_hols s 100 ax us s * ._t.G 100 c Le Babi arana_na_adeno _c f 0 * ddei tu s Odorrana_margaretae x ._t. F a_ts _japoni n s Am okinavan Soogl c 101 0 _ . L s ti s Odorra kurtmuellG M 1 i * n M 08 Hyl te s R M157 8 annecta 0 Ran a_s P Hyla_us a o 7-23 2 Telmatobi tes_ra * Rana lops_ g FC1 Telmatobius Rana_c na_i Hyla_chinensiso gariza Rana_drayt uenthe r * Rana_h ophy u p ponicu * Ra a_ a - 9 y Rana n basp t le 1 7 r Hyla_ * i Rana i Buf a i Ra e 1 36 * O s a_ u D Hoplobatrachus * _ * Hoplobatrach s tormo a i Euphlyctis_hexadactylus r Fe n _melano F F * _j a * h ra * o y i o_ga * c la_gayi ** a_ ***s * i s ri e e n i c ka lv k e eri W l atesb c le jer j Bufo_tibetanu i j a_kunyu awae tianus idoz _ r e e atic i _a h Buf T cf._c a Bufo g dybow uanr loosae Bufo al . r r ma t

varya_cancrivora v a v us Bufo_stejneger h s i

r m

a a a

f

s o y ckeri

r r ur eian _ . n n ga_marten y y e

ntis_v s ii e a us_baeobatrachus a ens nsis

s e s

n _ _ t ensis ki a

a m c l

i i

j us_tigerinus m i e is Quasipaa_yei

Telmatobufo_australis u

u

n

f n _ Nanorana_parker l

t Calyptocep Nanorana_pleske o _ o i r

Hemisus_marmoratus s Platyma ug Breviceps_adspersus n s Quasipaa_spinos c Lechriodus_melanopyga t

e r h s m

t i

i ulos

a a i

c Quasipaa_boulenger i nomaloglos Nanorana_taihangnic L t r

e a

Quasipaa_roberting i

A s

n

u

o

Limnonectes_fujianensis s

n

m Neobatrachia (90)

i

L

Fig. 5 Phylogenetic tree of all amphibians studied inferred using ML method based on the nucleotide dataset of 13 mt PCGs. The tree divides extant amphibians into three major taxa (orders): Gymnophiona, Caudata and Anura (including Neobatrachia) and the number of sampled species are shown in brackets. We have marked the branches of the species with rearrangements and species names only involved with PCGs rearrangements in dark red, names of species with rare arrangement are marked with an asterisk, and a pale red background is given to Neobatrachia, the group with the most intensive rearrangements. The middle pie chart shows the proportion of sampled amphibian species with mt gene rearrangements

and Caeciliidae, Typhlonectidae, Ichthyophiidae within branches marked in Fig. 5). In the 111 species with rear- Gymnophiona [52]. ranged mitogenomes, only 15 species were members of Compared with other vertebrate taxa, mitogenome the 117 total species sampled of Caudata (11.9%) and rearrangement of amphibians is seemingly more frequent Gymnophiona (15.2%), while most of them were con- [27], particularly in neobatrachian frogs [14, 28]. Among centrated in Anura (83.5%), especially Neobatrachia (90 the 232 amphibian mitogenomes examined, there were species: 100%). Figure 5 also shows 20 species with rear- 111 species (47.8%) with non-typical organization (red rangements of PCGs (only the gene order of 13 PCGs in Zhang et al. BMC Ecol Evo (2021) 21:19 Page 9 of 14

the mitogenome was considered), accounting for only appeared [14]. We also marked possible occurrence time 8.6% of total species (red species name marked in Fig. 5). of other rearrangement patterns in Additional fle 1: Fig- Species with rearranged mitogenomes tended to be ure. S3. phylogenetically clustered whether studied on all genes According to the estimated time of the rearrangement (p < 1.0e–10, hypergeometric test) or PCGs (p < 0.001, patterns, as shown in the lower left corner of Additional hypergeometric test), suggesting that mitogenomic rear- fle 1: Figure. S3, we speculated that the rearrangement rangement of amphibians also possesses phylogenetic of amphibian mitogenomes might present the following characteristics, similar with insects [22] and birds [53]. trends: (1) the trn genes near CR tended to be rearranged In Fig. 5, these species with rare arrangement were to the 3′ end of CR which was consistent with the tran- marked with asterisks. We found that they included scription direction of protein-coding genes; (2) the nd5 90% of the species with rearrangements of PCGs, which gene also had a tendency to rearrange to the 3′ end of CR. might be one of the reasons why their rearrangements were “rare”. Among the 232 species examined, 28 species Discussion belong to 22 families, each of which includes only one or Mitogenome rearrangement and qMGR method two species. We also further examined whether the rarity Just like our newly determined species, rearrangement could be due to the density of taxon sampling. Te result events are often accompanied by related gene duplica- showed that it was highly dependent on density of taxon tion as well as gene loss. Most existing models of rear- sampling (p < 1.0e-04, hypergeometric test). rangement mechanisms often involve gene duplication and gene deletion (e.g. [17–19, 55, 56],). Among these Rearrangement time estimation of amphibian models, TDRL (tandem duplication–random loss) model mitogenomes has been widely accepted to explain the rearrangement Because mitogenome rearrangement of amphibians pos- mechanism of amphibian mitochondrial genomes[34], sessed phylogenetic characteristics, and the same gene Such as Fejervarya limnocharis [57], L. bannaensis[29], arrangement among diferent species was generally Hoplobatrachus tigerinus [58], archeyi [59]. unlikely to be the result of convergent evolution [11], we With TDRL model, Xia et al. [34] perfectly explained the could estimate the (latest) possible occurrence time of rearrangement events within the species. TDNL (tandem their pattern according to the divergence time of com- duplication and non-random loss) is also another simi- mon ancestor lineages with the same rearrangement pat- lar model involving gene duplication and gene loss, and tern. So we referred to the literature on divergence time it has been used to rationally explain the rearrangement of diferent taxa of amphibians [14, 39, 43, 45], and esti- of the millipedes (Narceus annularus and Tyropygus sp.) mated the possible occurrence time of various patterns [18]. In addition, DMNR (the Dimer-Mitogenome and from three rearrangement enrichment regions, namely, Non-Random Loss) and DRRL (the Double Replications Region 1, Region 2 and Region 3, as shown in Fig. 1. and Random Loss) have also been proposed to explain Compared with the older Amphioxus (Branchiostoma the course of the rearrangement in two fatfsh mitog- foridae) mitogenome arrangement [54], the pattern enomes (Crossorhombus azureus and Samariscus latus) of typical amphibian arrangement was consistent with [19, 60]. that of vertebrates, including H-S2-L2-nad5-nad6-E- Our research has found that trn genes were more prone cob-T-P-CR-F (Region 1), IQM-nad2-WANCY (Region to rearrangement than other genes, and some trn genes 2), and nad5-nad6-cob (Region 3). Diferent from their (such as trnL2, trnT, and trnP) preferred to turn to the 3′ Region 3, the pattern, nad5-cob-nad6, were identifed in end of CR. Satoh et al. [61] proposed that there was a sig- 3 caudatans and 2 archaeobatrachians. It implied that the nifcant correlation between trn gene position (that is, trn rearrangement pattern could have occurred before the gene order) and codon usage in vertebrate mitogenomes, divergence of frogs and salamanders, that is, it may be and the closer to the 3′ end of CR the trn gene was, the earlier than 350 million years ago (Ma) [39]. Four newly higher its usage was, inferring that the mt gene arrange- sequenced species of neobatrachians, and 2 archaeoba- ment of vertebrates is afected by translation constraints, trachians, all had the IQMM pattern (IQMM trn clus- which may help maintain the gene order for a long time. ter), suggesting that the pattern can appear earlier than However, Xia et al. [28] claimed that the correlation was 250 Ma, before the divergence of neobatrachians and not signifcant for them of typical neobatrachian mitog- archaeobatrachians in Anura [39, 43] (Additional fle 1: enome. Furthermore, Xia et al. [34] also believed that Figure. S3). Similarly, the pattern, H-S2-nad5-nad6- the gene distributions of rearranged frogs were related E-cob-CR-L2-T-P-F (from Region 1 of typical neoba- to their non-adaptive forces. Terefore, the mecha- trachian arrangement) may have existed in the Late nism of trn gene order is still under debate in amphibian Jurassic (earlier than 140 Ma) when neobatrachians mitogenomes. Zhang et al. BMC Ecol Evo (2021) 21:19 Page 10 of 14

Nevertheless, due to the relatively rarity of rearrange- Non‑coding regions of mitogenome rearrangement ments and their constitution of useful synapomorphies, In the mitochondrial genomes, two non-coding regions, they may provide more useful information for phyloge- CR and ­OL, play important roles in the rearrangement netic inference with increasing difculty of their own studies [16, 42]. Te duplicated CR and mitochondrial research [16, 28]. A large number of studies have been gene rearrangement have been found in many parrot spe- carried out to clarify possible units of rearrangement cies [68]. Our study showed that CR was the most active and possible evolutionary processes between two rear- element in the mitogenome rearrangement of amphib- ranged mitogenomes (e.g. [17, 22, 23, 29]). qMGR has ians. When we disregarded it in the study of PCGs more fexibility than other rearrangement analysis algo- rearrangements, the results may have missed some infor- rithms (e.g., CREx [25], TreeRex [62], amGRP [63] and mation. For example, nad6-cob-CR-nad5 of neobtrachian GRAPPA [64]), and can analyze mitogenomes (or mt frog Buergeria buergeri [69] and nad6-cob-CR-nad5-CR gene fragments) of diferent numbers and diferent rear- of neobtrachian frog R. schlegelii [32] were treated as the rangement types, even other circular genomes such as same arrangement type. chloroplasts. In the absence of prior knowledge, within Unfortunately, we found that a few species in the Gen- a certain group, qMGR [35] was able to flter out highly Bank database have incomplete or even incorrect anno- rearranged genes and genomes, which would contribute tations for ­OL (as found by the references [70]). So we to the study of the choice of rearrangement models and ignored this important element in the WANCY region, rearrangement pathways. which was considered as a hotspot for rearrangement [16, 42]. Tis made us unable to distinguish between Mitogenome rearrangement and phylogenetic analysis ACW-OL-NY of marsupial Trichosurus vulpecula [71] Many previous rearrangement studies were often accom- and A-OL-CWNY of caecilian Siphonops paulensis [16], panied by phylogenetic analysis (e.g., [14, 32, 34, 50]), and between WA-OL-NCY and WAN-OL-CY of neob- which helped to speculate on the evolution of rearrange- trachian frog Q. boulengeri (intraspecifc rearrangement ment events. In our study, we reconstructed two phylo- of mitogenome) [34]. Tese implied that the components genetic trees (BI tree and ML tree). We found that the examined in the study of gene rearrangements had a great phylogenetic relationship of most neobtrachians was con- infuence on the results. In fact, some reports have found sistent with previous studies (e.g., [38, 43, 44, 49, 65]), the that ­OL is absent in birds, crocodiles, fsh, scorpions, etc. relationship between teresomatans and all other caecil- [42, 72, 73]. However, in vertebrates with mitochondrial ians was in full agreement with previous reports (e.g. [44, ­OL, the WANCY region of amphibians possessed the 49, 52],), and our trees also supported a sister-group rela- most frequent rearrangements, which also involved gene tionship between frogs and salamanders (the Batrachia duplication, gene loss, and pseudogenes (e.g., [16, 29, 34, hypothesis), as found in most previous studies(e.g. [44, 42, 50]). 52, 66],). In addition, the relationship between Leiopel- matidae and all other frogs was compatible with previous Conclusion reports in BI tree (e.g. [50, 59, 67]). Te sister relation- In this study, we frst examined the characteristics of ships of Pipoidea and Discoglossoidea were consistently trnM tandem duplication in four newly sequenced supported (BPPs = 1, BS = 83%), however, some recent Dicroglossidae mitogenomes as well as in other amphib- studies did not support the view (e.g. [38, 44, 50]). Te ian taxa, and found that it was not an exclusive feature inconsistent results may be afected by diferent species of Dicroglossidae. We then applied qMGR for calculat- and selected molecular markers. But these subtle difer- ing RS and RF of each single mt gene, and screened out ences did not change the overall distribution of amphib- high-frequency genes and conservative genes involved ian rearrangement types in the phylogenetic trees. Based in diferent taxa of amphibian mitogenome rearrange- on the results of the phylogenetic study above, species ment. Based on phylogenetic analysis, we found that with the same pattern of mitogenome rearrangement mitogenomic rearrangement of amphibians possessed mostly belonged to closely related taxa. Just like the pre- phylogenetic implications, and were concentrated in vious reports of avian mitogenome rearrangement [53], Neobatrachia. Still, typical vertebrate arrangement was the phylogenetic characteristics of rearrangement were the most dominant type of arrangements for amphib- further interpreted that they could originate from a com- ians. In addition, qMGR can also obtain RS of species mon ancestor and were then retained during subsequent with rare arrangements, which can be used to measure lineage diversifcation. Terefore, this makes it possible the rearrangement degree of a single species. Tis was a to use the divergence time of ancestral species with com- systematic survey of the rearrangement of the amphibian mon mitogenome arrangement to estimate the occur- mitochondrial genome and its evolution. Nevertheless, rence time of rearrangement patterns. as the currently available data was less than 5% of the Zhang et al. BMC Ecol Evo (2021) 21:19 Page 11 of 14

total number of known species, our results may still lack arrangement, and those of not all genes were called local representativeness and integrity. More data and more in- arrangements. We surveyed major arrangement whose depth analysis will help us fgure out thoroughly the rear- the cumulative number with the same arrangements rangement characteristics of amphibians. was ≥ 3 times, For the type of arrangement less than or equal to two times, we called it "rare arrangement". We Methods further subdivided the high-frequency rearrangement Determination and analysis of 4 new mitogenomes regions in the global arrangements or local arrange- Samples of four dicroglossid frogs Q. robertingeri (code ments into diferent regions, which also belonged to local Zhang-YBJW031), L. fragilis (code DT-CP005), L. fuji- arrangements. To facilitate research, we also numbered anensis (Taiwan population) (code Zhang-TWDT034) the diferent arrangement types of these major arrange- and L. fujianensis (Fujian population) (code DT-FJ002) ments into multiple patterns. were collected from Sichuan, Hainan, Taiwan and Fujian province in China, respectively. Total DNA was Precise quantifcation of rearranged genes and genomes extracted from their fresh muscle tissues after the frogs qMGR can be used to accurately calculate the rear- were euthanized using 0.5% MS-222. Ten tissues of the rangement frequency of each single gene or each single frst two were stored at − 80 °C at the College of Life Sci- genome within a given taxonomic group [35]. When a ences, Anhui Normal University, China, and the latter reference arrangement (benchmark) of mitogenome was two were stored at − 20 °C at the School of Bioengineer- selected, it can accumulate changes in genes at the two ing, Huainan Normal University, China. In accordance nearest fanking positions of a gene to be tested and give with Regulation for the collection of genetic resources of the gene a score. Based on this principle, qMGR can cal- China (HJ 628-2011), we collected all laboratory animals, culate the rearrangement score (RS) of a complete mitog- and animal experiments and follow-up disposal were enome and the relative rearrangement frequency (RF) of carried out based on Regulations for the management individual genes within a certain group (e.g., neobatra- of laboratory animals in Huainan Normal University chians, anurans or amphibians) (referring to reference (2015). We designed multiple primer pairs (Additional [35] and its website for more details on the method). fle 1: word S1) based on sequence alignments of mito- We chose typical vertebrate arrangement as a bench- chondrial genes from closely related species [29, 57, 74], mark for the comprehensive analysis of mitochondrial and carried out shotgun sequencing and assembling. We genome rearrangement in amphibians, and also chose identifed genomic components by the MITOS2 (http:// typical neobatrachian arrangement for comparative mitos​.bioin​f.uni-leipz​ig.de/index​.py) [75], tRNAscan-SE analysis of RF. In the calculation process, we regarded the (http://lowel​ab.ucsc.edu/tRNAs​can-SE/) [76], and man- CR as a single gene, while ignored pseudogenes and the ual sequence comparison [2, 37]. Base composition was origin of L-strand replication (O­ L) for their incomplete determined using DNASTAR (www.dnast​ar.com). annotations.

Preparation of rearrangement data Phylogenetic analysis and rearrangement time estimation We downloaded sequences data of amphibian mitoge- We performed phylogenetic analysis using maximum nomes from NCBI: Organelle Genome Resource (http:// likelihood (ML) [77] and Bayesian inference (BI) [78] www.ncbi.nlm.nih.gov/genom​e/organ​elle/) on May 10, methods based on the combined nucleotide dataset of 2019. In total, we retrieved mitogenomes of 228 species 13 PCGs of 232 species (including 43 families, about half (in addition to the four species newly sequenced here) by of them contain only 1–2 species) as well as two out- fltering out the data without complete annotation infor- group species (turtle: Chinemys reevesi, NC_006082, and mation. We then extracted location information of genes iguana: Iguana iguana NC_002793). Multiple sequence and their nucleotide sequences for subsequent rearrange- alignment was carried out using MAFFT 7.2 [79]. Te ment analysis and phylogeny using designed Perl and R substitution saturation analyses for each codon positions scripts. of each gene were assessed by using DAMBE 7.2.25 [80]. And we detected the saturation on their third codon posi- Diferent rearrangement patterns tions. Te information of substitution saturation for all To investigate major patterns of amphibian mitogenome codon positions of 13 genes could be seen in Additional rearrangement, we sorted all genes based on their posi- fle 1: Table S2. Terefore, phylogenetic analysis was only tion information in the genome. When starting with the based on their frst two codon positions. For BI analysis, same gene, we merged and counted species with identi- the best schemes for partition and substitution models cal gene orders. We defned the arrangement involved (Additional fle 1: Table S3) were determined in Parti- with all mt genes (37 genes + CR in this study) as a global tionFinder version 2.1.1 [81] according to the Bayesian Zhang et al. BMC Ecol Evo (2021) 21:19 Page 12 of 14

Information Criterion (BIC) and greedy search algo- Authors’ contributions This study was conceived by J.Z. and the project was led by J.Z., X.K. and L.N. rithm. Ten the BI analysis was implemented in MrBayes Collection of the sample, mitogenomic data and references were done by J.Z., 7 v.3.2.7a [82, 83]. We ran four Markov chains for 5 × 10 G.M., S.Y., Z.J. and P.G. Data analysis and the experiments were performed by generations (sampling every 1000 generations) and cal- J.Z., S.H., Q.S. X.K., H.D and C.W.. The manuscript was written by J.Z., S.H., Q.S. and Z.J. All authors read and approved the fnal manuscript. culated a 50% majority-rule consensus tree and Bayesian posterior probabilities (BPPs) after discarding the initial Funding 25% trees as burn-in. All MCMC runs were repeated This study was funded by National Natural Science Foundation of China (No. 31501461, 31800256); Anhui Science and Technology Major Project (No. twice to avoid spurious results. Furthermore, the conver- 18030701189); Independent Research fund of Key Laboratory of Industrial gence of the MrBayes analyses was checked with Tracer Dust Prevention and Control & Occupational Health and Safety, Ministry of 1.7.1 [84]. Subsequently, the ML tree was inferred with Education (Anhui University of Science and Technology) (No. EK20201004), Huainan science and technology project (2017A0421); the Key Support RAxML v.8.2.12 [85] using the GTRGAMMA model, Program for Outstanding Young Talents in University of Anhui Province (No. 100 starting trees and 1000 bootstrap replicates to assess gxyqZD2016264); the Research Projects of Huainan Normal University (Nos. node support [86]. After that, the ML bootstrap conver- 2019hsjy22, 2017hsyxkc91, 2015xj49zd, 2015hssjjd05, and 2015hsyxkc22). The funding bodies had no role in the design of the study and collection, analysis, gence test was carried out with parameter "-I autoMRE". and interpretation of data and in writing the manuscript. To estimate the possible occurrence time of gene rear- rangement, we referred directly to the results of the stud- Availability of data and materials The source code of qMGR was freely available under GNU GPL at https​://githu​ ies on divergence time estimations [14, 39, 43, 45]. Based b.com/zhang​lab20​19/qMGR, and its web service was available at http://qmgr. on the divergence timetable of amphibians and related hnnu.edu.cn/. The four complete sequences of our newly identifed frogs were taxa, we estimated the latest possible occurrence time of submitted to GenBank (https​://www.ncbi.nlm.nih.gov/) with accession num- bers KY441640 (Q. robertingeri), AY899241 (L. fragilis), MF678821 (L. fujianensis the diferent rearrangement patterns. (Taiwan)) and AY974191 (L. fujianensis (Fujian)), respectively. We downloaded sequences data of amphibian mitogenomes from NCBI: Organelle Genome Supplementary Information Resource (http://www.ncbi.nlm.nih.gov/genom​e/organ​elle/). The online version contains supplementary material available at https​://doi. Ethics approval and consent to participate org/10.1186/s1286​2-021-01755​-3. Frog experiments were conducted according to the guidelines established by the Regulations for the Administration of Afairs Concerning Experimental Additional fle 1: Table S1. Amphibian species with mt gene duplication Animals (Ministry of Science and Technology, China, 2004). Our research was or loss in this study. Figure S1. Mitogenomic rearrangement patterns of approved by the Animal Ethics Committee of Huainan Normal University all amphibian species investigated. The numbers in frst column indicated (approval number: AE20151020). The collection of the samples was approved the occurrence frequencies of patterns (a number less than three is by the Academic Committee of Huainan Normal University and the cor- defned as a rare arrangement in this study). Figure S2. Phylogenetic responding local forestry management departments. The collection behavior tree of all amphibians studied using BI method based on the nucleotide complied with Regulation for the collection of genetic resources of China (HJ dataset of 13 mt PCGs. The number of species sampled for each lineage 628-2011), is shown in brackets. We marked the branches of the species with mt gene rearrangements and the names of species only involved with PCGs Consent for publication rearrangements in dark red, and a pale red background was set on the Not applicable. branch of neobatrachians, the group with the most intensive rearrange- ments. Figure S3. Possible occurrence time and trends of local rearrange- Competing interests ment patterns of mitogenomes according to divergence times of their The authors declare that they have no competing interests. ancestors. The possible trends were shown in the box at lower left. Word S1. Amplifcation scheme and primers sequences for sequencing four Author details 1 mitochondrial genomes in this study. Table S2. Determination of substitu- School of Biological Engineering, Huainan Normal University, Huainan, 2 tion saturation for each codon position of each gene. Table S3. The best Anhui 232001, People’s Republic of China. College of Life Science, Anhui 3 partitioning scheme selected by PartitionFinder for diferent data matrices. Normal University, Wuhu, Anhui 241000, People’s Republic of China. Anhui Key Laboratory of Low Temperature Co‑Fired Materials, Huainan Normal University, Huainan 232001, People’s Republic of China. 4 Key Laboratory Abbreviations of Industrial Dust Prevention and Control and Occupational Health and Safety, 5 a6, ATP6: ATP synthase subunit 6; a8, ATP8: ATP synthase subunit 8; BI: Bayes- Ministry of Education, Huainan 232001, People’s Republic of China. Anhui ian inference; BIC: Bayesian Information Criterion; BPPs: Bayesian posterior Shanhe Pharmaceutical Excipients Co., Ltd., Huainan 232001, People’s Republic 6 probabilities; cox 1–3, COX1-3: Cytochrome c oxidase subunit I–III; CR, D-loop: of China. Department of Biostatistics, Bloomberg School of Public Health, 7 Control regions; cob, Cytb: Cytochrome b; IQM(M): trnI, trnQ, trnM (trnM); LTPF: Johns Hopkins University, Baltimore, MD 21205, USA. Life Science and Food trnL (CUN), trnT, trnP, and trnF; KS test: Kolmogorov–Smirnov test; Ma: Million Engineering College, Yibin University, Yibin, Sichuan 644000, People’s Republic years ago; ML: Maximum likelihood; mt: Mitochondrial; nad1-6, 4L, ND1-6, 4L: of China. NADH dehydrogenase subunit 1–6, 4 L; NJ: Neighbor-joining method; OL: The origin of light strand replication; PCR: Polymerase chain reaction; qMGR: Received: 16 May 2020 Accepted: 28 January 2021 Quantifying mitogenome rearrangements; RF: Rearrangement frequency; RS: Rearrangement score; TDRL: Tandem duplication and random loss; WANCY: trnW, trnA, trnN, OL, trnC, and trnY; tRNA, trn: Transfer ribonucleic acid; trnL1: trnL (UUR); tCL2: trnL (CUN); trnS1: trnS (UCN); trnS2: trnS (AGY). References Acknowledgements 1. Boore JL. Animal mitochondrial genomes. Nucleic Acids Res. We thank Changzhi Pu, Wenze Yang, Simin Lin and Xuelong Jiang for their 1999;27(8):1767–80. assistance in samples collection. 2. Sumida M, Kanamori Y, Kaneda H, Kato Y, Nishioka M, Hasegawa M, Yonekawa H. Complete nucleotide sequence and gene rearrangement of Zhang et al. BMC Ecol Evo (2021) 21:19 Page 13 of 14

the mitochondrial genome of the Japanese pond frog Rana nigromacu- 25. Bernt M, Merkle D, Ramsch K, Fritzsch G, Perseke M, Bernhard D, Schlegel lata. Genes Genet Syst. 2001;76(5):311–25. M, Stadler PF, Middendorf M. CREx: inferring genomic rearrangements 3. Boore JL, Brown WM. Big trees from little genomes: mitochondrial gene based on common intervals. Bioinformatics. 2007;23(21):2957–8. order as a phylogenetic tool. Curr Opin Genet Dev. 1998;8(6):668–74. 26. Boore JL, Lavrov DV, Brown WM. Gene translocation links insects and 4. Brown WM, George M Jr, Wilson AC. Rapid evolution of animal mitochon- crustaceans. Nature. 1998;392(6677):667–8. drial DNA. Proc Natl Acad Sci USA. 1979;76(4):1967–71. 27. Machado DJ, Lyra ML, Grant T. Mitogenome assembly from genomic 5. Cummings MP, Meyer A. Magic bullets and golden rules: data sampling in multiplex libraries: comparison of strategies and novel mitogenomes for molecular phylogenetics. Zoology. 2005;108(4):329–36. fve species of frogs. Mol Ecol Resour. 2016;16(3):686–93. 6. Dellaporta SL, Xu A, Sagasser S, Jakob W, Moreno MA, Buss LW, Schi- 28. Xia Y, Zheng Y, Miura I, Wong PB, Murphy RW, Zeng X. The evolution of erwater B. Mitochondrial genome of Trichoplax adhaerens supports mitochondrial genomes in modern frogs (Neobatrachia): nonadaptive placozoa as the basal lower metazoan phylum. Proc Natl Acad Sci USA. evolution of mitochondrial genome reorganization. BMC Genomics. 2006;103(23):8751–6. 2014;15:691. 7. Anderson S, Bankier AT, Barrell BG, de Bruijn MH, Coulson AR, Drouin J, 29. Zhang JF, Nie LW, Wang Y, Hu LL. The complete mitochondrial genome Eperon IC, Nierlich DP, Roe BA, Sanger F, et al. Sequence and organization of the large-headed frog, Limnonectes bannaensis (Amphibia: Anura), of the human mitochondrial genome. Nature. 1981;290(5806):457–65. and a novel gene organization in the vertebrate mtDNA. Gene. 8. Bibb MJ, Van Etten RA, Wright CT, Walberg MW, Clayton DA. Sequence 2009;442(1–2):119–27. and gene organization of mouse mitochondrial DNA. Cell. 1981;26(2 Pt 30. San Mauro D, Gower DJ, Cotton JA, Zardoya R, Wilkinson M, Massing- 2):167–80. https​://doi.org/10.1016/0092-8674(81)90300​-7. ham T. Experimental design in phylogenetics: testing predictions from 9. Roe BA, Ma DP, Wilson RK, Wong JF. The complete nucleotide expected information. Syst Biol. 2012;61(4):661–74. sequence of the Xenopus laevis mitochondrial genome. J Biol Chem. 31. Kurabayashi A, Usuki C, Mikami N, Fujii T, Yonekawa H, Sumida M, 1985;260(17):9759–74. Hasegawa M. Complete nucleotide sequence of the mitochondrial 10. Zardoya R, Meyer A. The complete nucleotide sequence of the genome of a Malagasy poison frog Mantella madagascariensis: evolution- mitochondrial genome of the lungfsh (Protopterus dolloi) supports its ary implications on mitochondrial genomes of higher anuran groups. phylogenetic position as a close relative of land vertebrates. Genetics. Mol Phylogenet Evol. 2006;39(1):223–36. 1996;142(4):1249–63. 32. Sano N, Kurabayashi A, Fujii T, Yonekawa H, Sumida M. Complete 11. Zhong J, Li G, Liu ZQ, Li QW, Wang YQ. Gene rearrangement of mitochon- nucleotide sequence of the mitochondrial genome of Schlegel’s tree frog drial genome in the vertebrate. Yi chuan xue bao Acta genetica Sinica. Rhacophorus schlegelii (family Rhacophoridae): duplicated control regions 2005;32(3):322–30. = and gene rearrangements. Genes Genet Syst. 2005;80(3):213–24. 12. Ki JS, Dahms HU, Hwang JS, Lee JS. The complete mitogenome of the 33. Mueller RL, Boore JL. Molecular mechanisms of extensive mitochon- hydrothermal vent crab Xenograpsus testudinatus (Decapoda, Brachyura) drial gene rearrangement in plethodontid salamanders. Mol Biol Evol. and comparison with brachyuran crabs. Comp Biochem Physiol D: 2005;22(10):2104–12. Genomics Proteomics. 2009;4(4):290–9. 34. Xia Y, Zheng Y, Murphy RW, Zeng X. Intraspecifc rearrangement of mito- 13. Kilpert F, Podsiadlowski L. The complete mitochondrial genome of chondrial genome suggests the prevalence of the tandem duplication- the common sea slater, Ligia oceanica (Crustacea, Isopoda) bears a random loss (TDLR) mechanism in Quasipaa boulengeri. BMC Genomics. novel gene order and unusual control region features. BMC Genomics. 2016;17(1):965. 2006;7:241. 35. Zhang J, Kan X, Miao G, Hu S, Sun Q, Tian W. qMGR: a new approach for 14. Kurabayashi A, Sumida M. Afrobatrachian mitochondrial genomes: quantifying mitochondrial genome rearrangement. Mitochondrion. genome reorganization, gene rearrangement mechanisms, and evo- 2020;52:20. lutionary trends of duplicated and rearranged genes. BMC Genomics. 36. Chen G, Wang B, Liu J, Xie F, Jiang J. Complete mitochondrial genome 2013;14:633. of Nanorana pleskei (Amphibia: Anura: Dicroglossidae) and evolutionary 15. Mindell DP, Sorenson MD, Dimchef DE. Multiple independent characteristics. Curr Zool. 2011;57(6):785–805. origins of mitochondrial gene order in birds. Proc Natl Acad Sci USA. 37. Shan X, Xia Y, Zheng YC, Zou FD, Zeng XM. The complete mitochondrial 1998;95(18):10693–7. genome of Quasipaa boulengeri (Anura: Dicroglossidae). Mitochondrial 16. San Mauro D, Gower DJ, Zardoya R, Wilkinson M. A hotspot of gene order DNA. 2014;25(2):83–4. rearrangement by tandem duplication and random loss in the vertebrate 38. Feng YJ, Blackburn DC, Liang D, Hillis DM, Wake DB, Cannatella DC, Zhang mitochondrial genome. Mol Biol Evol. 2006;23(1):227–34. P. Phylogenomics reveals rapid, simultaneous diversifcation of three 17. Shi W, Gong L, Wang SY, Miao XG, Kong XY. Tandem duplication and major clades of Gondwanan frogs at the Cretaceous-Paleogene bound- random loss for mitogenome rearrangement in Symphurus (Teleost: ary. Proc Natl Acad Sci USA. 2017;114(29):E5864–70. Pleuronectiformes). BMC Genomics. 2015;16:355. 39. Roelants K, Gower DJ, Wilkinson M, Loader SP, Biju SD, Guillaume K, 18. Lavrov DV, Boore JL, Brown WM. Complete mtDNA sequences of two Moriau L, Bossuyt F. Global patterns of diversifcation in the history of millipedes suggest a new model for mitochondrial gene rearrangements: modern amphibians. Proc Natl Acad Sci USA. 2007;104(3):887–92. duplication and nonrandom loss. Mol Biol Evol. 2002;19(2):163–9. 40. Hickerson MJ, Cunningham CW. Dramatic mitochondrial gene rearrange- 19. Shi W, Dong XL, Wang ZM, Miao XG, Wang SY, Kong XY. Complete mitog- ments in the hermit crab Pagurus longicarpus (Crustacea, anomura). Mol enome sequences of four fatfshes (Pleuronectiformes) reveal a novel Biol Evol. 2000;17(4):639–44. gene arrangement of L-strand coding genes. BMC Evol Biol. 2013;13:173. 41. Mu X, Yang Y, Liu Y, Luo D, Xu M, Wei H, Gu D, Song H, Hu Y. The complete 20. Lunt DH, Hyman BC. Animal mitochondrial DNA recombination. Nature. mitochondrial genomes of two freshwater snails provide new protein- 1997;387(6630):247. coding gene rearrangement models and phylogenetic implications. 21. Maarouf N, Arno G, Carter ND, Syrris P, Yusuf S, Camm AJ, Poleiniki J, Parasites Vectors. 2017;10(1):11. Al-Saady NM. Quantifcation of mitochondrial sublimons in human fbril- 42. Qian L, Wang H, Yan J, Pan T, Jiang S, Rao D, Zhang B. Multiple independ- lating atria. Clin Sci. 2004;106(6):653–9. ent structural dynamic events in the evolution of snake mitochondrial 22. Babbucci M, Basso A, Scupola A, Patarnello T, Negrisolo E. Is it an ant or genomes. BMC Genomics. 2018;19(1):354. a butterfy? Convergent evolution in the mitochondrial gene order of 43. Pyron RA. Divergence time estimation using fossils as terminal taxa and Hymenoptera and Lepidoptera. Genome Biol Evol. 2014;6(12):3326–43. the origins of Lissamphibia. Syst Biol. 2011;60(4):466–81. 23. Basso A, Babbucci M, Pauletto M, Riginella E, Patarnello T, Negrisolo E. The 44. Pyron RA, Wiens JJ. A large-scale phylogeny of Amphibia including over highly rearranged mitochondrial genomes of the crabs Maja crispata and 2800 species, and a revised classifcation of extant frogs, salamanders, Maja squinado (Majidae) and gene order evolution in Brachyura. Sci Rep. and caecilians. Mol Phylogenet Evol. 2011;61(2):543–83. 2017;7(1):4096. 45. San Mauro D. A multilocus timescale for the origin of extant amphibians. 24. Bernt M, Merkle D, Middendorf M. An algorithm for inferring mitoge- Mol Phylogenet Evol. 2010;56(2):554–61. nome rearrangements in a phylogenetic tree. In: Comparative genomics, 46. Zhang P, Zhou H, Chen YQ, Liu YF, Qu LH. Mitogenomic perspec- international workshop, RECOMB-CG 2008, Paris, France, October 13–15, tives on the origin and phylogeny of living amphibians. Syst Biol. 2008 Proceedings: 2008. Pp 143–157. 2005;54(3):391–400. Zhang et al. BMC Ecol Evo (2021) 21:19 Page 14 of 14

47. Hoegg S, Vences M, Brinkmann H, Meyer A. Phylogeny and compara- mitochondrial genomes and nuclear RAG1. Mol Phylogenet Evol. tive substitution rates of frogs inferred from sequences of three nuclear 2004;33(2):413–27. genes. Mol Biol Evol. 2004;21(7):1188–200. 67. Irisarri I, Vences M, San Mauro D, Glaw F, Zardoya R. Reversal to air-driven 48. Roelants K, Bossuyt F. Archaeobatrachian paraphyly and pangaean diver- sound production revealed by a molecular phylogeny of tongueless sifcation of crown-group frogs. Syst Biol. 2005;54(1):111–26. frogs, family . BMC Evol Biol. 2011;11(1):114. 49. Frost DR, Grant T, Faivovich J, Bain RH, Haas A, Haddad CFB, De Sa RO, 68. Eberhard JR, Wright TF. Rearrangement and evolution of mitochondrial Channing A, Wilkinson M, Donnellan SC. The amphibian tree of life. Bull genomes in parrots. Mol Phylogenet Evol. 2016;94:34–46. Am Museum Nat History. 2006;297(297):1–370. 69. Sano N, Kurabayashi A, Fujii T, Yonekawa H, Sumida M. Complete nucleo- 50. Irisarri I, San Mauro D, Abascal F, Ohler A, Vences M, Zardoya R. The tide sequence and gene rearrangement of the mitochondrial genome of origin of modern frogs (Neobatrachia) was accompanied by accelera- the bell-ring frog, Buergeria buergeri (family Rhacophoridae). Genes Genet tion in mitochondrial and nuclear substitution rates. BMC Genomics. Syst. 2004;79(3):151–63. 2012;13:626. 70. Prada CF, Boore JL. Gene annotation errors are common in the mamma- 51. Vieites DR, Zhang P, Wake DB. Salamanders (Caudata). The timetree of life lian mitochondrial genomes database. BMC Genomics. 2019;20(1):73. 2009:365–368. https​://www.googl​e.com/books​?hl zh-CN&lr &id 9rt1c​ 71. Paabo S, Thomas WK, Whitfeld KM, Kumazawa Y, Wilson AC. Rearrange- 1hl49​MC&oi fnd&pg PA365​&ots PW-ES7dx​zp&sig= YoUdZ=​BYhl9=​ ments of mitochondrial transfer RNA genes in marsupials. J Mol Evol. ICq3w​yMok9=​cSEZb​eE.= = = 1991;33(5):426–30. 52. San Mauro D, Gower DJ, Müller H, Loader SP, Zardoya R, Nussbaum RA, 72. Macey JR, Larson A, Ananjeva NB, Fang Z, Papenfuss TJ. Two novel gene Wilkinson M. Life-history evolution and mitogenomic phylogeny of orders and the role of light-strand replication in rearrangement of the caecilian amphibians. Mol Phylogenet Evol. 2014;73:177–89. vertebrate mitochondrial genome. Mol Biol Evol. 1997;14(1):91–104. 53. Haring E, Kruckenhauser L, Gamauf A, Riesing MJ, Pinsker W. The com- 73. Seutin G, Lang BF, Mindell DP, Morais R. Evolution of the WANCY region in plete sequence of the mitochondrial genome of Buteo buteo (Aves, amniote mitochondrial DNA. Mol Biol Evol. 1994;11(3):329–40. Accipitridae) indicates an early split in the phylogeny of raptors. Mol Biol 74. Chen Z, Zhai X, Zhu Y, Chen X. Complete mitochondrial genome of the Evol. 2001;18(10):1892–904. Ye’s spiny-vented frog Yerana yei (Anura: Dicroglossidae). Mitochondrial 54. Boore JL, Daehler LL, Brown WM. Complete sequence, gene arrange- DNA. 2015;26(3):489–90. ment, and genetic code of mitochondrial DNA of the cephalochordate 75. Bernt M, Donath A, Juhling F, Externbrink F, Florentz C, Fritzsch G, Putz Branchiostoma foridae (Amphioxus). Mol Biol Evol. 1999;16(3):410–8. J, Middendorf M, Stadler PF. MITOS: improved de novo metazoan mito- 55. Dowton M, Campbell NJ. Intramitochondrial recombination—is chondrial genome annotation. Mol Phylogenet Evol. 2013;69(2):313–9. it why some mitochondrial genes sleep around? Trends Ecol Evol. 76. Lowe TM, Eddy SR. tRNAscan-SE: a program for improved detec- 2001;16(6):269–71. tion of transfer RNA genes in genomic sequence. Nucleic Acids Res. 56. Zhang J-Y, Zhang L-P, Yu D-N, Storey KB, Zheng R-Q. Complete mito- 1997;25(5):955–64. chondrial genomes of Nanorana taihangnica and N. yunnanensis (Anura: 77. Felsenstein J. Evolutionary trees from DNA sequences: a maximum likeli- Dicroglossidae) with novel gene arrangements and phylogenetic hood approach. J Mol Evol. 1981;17(6):368–76. relationship of Dicroglossidae. BMC Evol Biol. 2018;18(1):1–13. 78. Huelsenbeck JP, Ronquist F, Nielsen R, Bollback JP. Bayesian infer- 57. Liu ZQ, Wang YQ, Su B. The mitochondrial genome organization of the ence of phylogeny and its impact on evolutionary biology. Science. rice frog, Fejervarya limnocharis (Amphibia: Anura): a new gene order in 2001;294(5550):2310–4. the vertebrate mtDNA. Gene. 2005;346:145–51. 79. Katoh K, Standley DM. MAFFT multiple sequence alignment software 58. Alam MS, Kurabayashi A, Hayashi Y, Sano N, Khan MMR, Fujii T, Sumida M. version 7: improvements in performance and usability. Mol Biol Evol. Complete mitochondrial genomes and novel gene rearrangements in 2013;30(4):772–80. two dicroglossid frogs, Hoplobatrachus tigerinus and Euphlyctis hexadacty- 80. Xia X. DAMBE7: new and improved tools for data analysis in molecular lus, from Bangladesh. Genes Genet Syst. 2010;85(3):219–32. biology and evolution. Mol Biol Evol. 2018;35(6):1550–2. 59. Irisarri I, San Mauro D, Green DM, Zardoya R. The complete mitochondrial 81. Lanfear R, Frandsen PB, Wright AM, Senfeld T, Calcott B. PartitionFinder genome of the relict frog Leiopelma archeyi: insights into the root of the 2: new methods for selecting partitioned models of evolution for frog tree of life. Mitochondrial DNA. 2010;21(5):173–82. molecular and morphological phylogenetic analyses. Mol Biol Evol. 60. Shi W, Miao X-G, Kong X-Y. A novel model of double replications and ran- 2016;34(3):772–3. dom loss accounts for rearrangements in the Mitogenome of Samariscus 82. Ronquist F, Huelsenbeck JP. MrBayes 3: Bayesian phylogenetic inference latus (Teleostei: Pleuronectiformes). BMC Genomics. 2014;15(1):352. under mixed models. Bioinformatics. 2003;19(12):1572–4. 61. Satoh TP, Sato Y, Masuyama N, Miya M, Nishida M. Transfer RNA gene 83. Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Hohna S, arrangement and codon usage in vertebrate mitochondrial genomes: a Larget B, Liu L, Suchard MA, Huelsenbeck JP. MrBayes 3.2: efcient Bayes- new insight into gene order conservation. BMC Genomics. 2010;11:479. ian phylogenetic inference and model choice across a large model space. 62. Bernt M, Merkle D, Middendorf M. An algorithm for inferring mitoge- Syst Biol. 2012;61(3):539–42. nome rearrangements in a phylogenetic tree. In: Research in computa- 84. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior tional molecular biology. 2008; pp 143–157. summarization in Bayesian phylogenetics using Tracer 1.7. Syst Biol. 63. Bernt M, Merkle D, Middendorf M. Using median sets for inferring phylo- 2018;67(5):901–4. genetic trees. Bioinformatics. 2007;23(2):129–35. 85. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post- 64. Moret BME, Siepel A, Tang J, Liu T. Inversion medians outperform break- analysis of large phylogenies. Bioinformatics. 2014;30(9):1312–3. point medians in phylogeny reconstruction from gene-order data. In: 86. Stamatakis A. Phylogenetic models of rate heterogeneity: a high perfor- Workshop on algorithms in bioinformatics. 2002. 521–536. mance computing perspective. In: Parallel and distributed processing 65. Grant T, Frost DR, Caldwell JP, Gagliardo R, Haddad CFB, Kok PJR, Means symposium, 2006 IPDPS 2006 20th International: 2006. IEEE: 8 pp. DB, Noonan BP, Schargel WE, Wheeler WC. Phylogenetic systematics of dart-poison frogs and their relatives (amphibia: athesphatanura: dendro- batidae). Bull Am Museum Nat History. 2006;299:1–262. Publisher’s Note 66. San Mauro D, Gower DJ, Oommen OV, Wilkinson M, Zardoya R. Phy- Springer Nature remains neutral with regard to jurisdictional claims in pub- logeny of caecilian amphibians (Gymnophiona) based on complete lished maps and institutional afliations.