Insights into the structural features and phylogenetic implications of the complete mitochondrial genome of Fasin rainbow fish (Melanotaenia fasinensis)

Abstract

The Fasin rainbow fish, scientifically named Melanotaenia fasinensis, is highly prized by aquarium enthusiasts for its vibrant colors and adaptability to artificial aquatic environments. This species is endemic to the karst landscape of the Bird’s Head region in Papua, Indonesia, and belongs to the family Melanotaeniidae. Discovered relatively recently in 2010, this species was designated as endangered by the International Union for Conservation of Nature (IUCN) in 2021. However, there is currently insufficient data regarding its phylogenetic position. To address this gap, our study employed next-generation sequencing (NGS) to analyze the entire mitochondrial genome of M. fasinensis. The mitochondrial genome comprises 13 protein-coding genes, 22 transfer RNA genes, and two ribosomal RNA genes, with a total length of 16,731 base pairs. The base composition of the mitogenome revealed percentages of 27.76% adenine (A), 27.34% thymine (T), 16.15% guanine (G), and 28.75% cytosine (C) residues. Our phylogenetic analysis based on sequence data indicated that all species of the Melanotaeniidae family clustered together on the same branch. Furthermore, the intergeneric and interspecific taxonomic positions were explicit and clear. Phylogenetically, Melanotaeniidae were more closely related to the family Isonidae than to the family Atherinomorus. The phylogenetic position of M. fasinensis was relatively basal within the genus Melanotaenia. This study provides valuable molecular insights for further exploration of the phylogeography and evolutionary history of M. fasinensis and other members of the genus Melanotaenia.

Introduction

The freshwater ornamental fish industry in Indonesia has significant growth potential, with freshwater species comprising the majority of ornamental fish exports, contributing substantially to the nation’s foreign exchange earnings. Rainbowfish, particularly Melanotaenia species, are highly prized for their vibrant coloration and adaptability to captive environments, making them economically valuable. The Melanotaenia fasinensis, endemic to the Fasin River in West Papua, is one such species, recognized for its bright colors and popularity among aquarists [1, 2].

Previous phylogenetic studies of the Melanotaeniidae family, particularly within the genus Melanotaenia, have been based on partial mitochondrial gene sequences [1, 3]. While these studies provided initial insights into species relationships, the limited scope of the gene fragments used did not offer sufficient resolution to fully determine the phylogenetic relationships among species. This has left the taxonomic relationships within Melanotaenia inadequately resolved. In light of these limitations, recent studies have called for the use of more comprehensive datasets, such as complete mitochondrial genomes, to improve the accuracy of phylogenetic inference and address existing taxonomic challenges [4]. In this study, we present the complete mitochondrial genome of M. fasinensis, which provides the necessary data for a more robust phylogenetic analysis and contributes to resolving these long-standing taxonomic issues. In addition, understanding the mitochondrial genome of M. fasinensis provides valuable insights into the species’ evolutionary history and genetic diversity, which are essential for conservation efforts, especially since the species is endemic to a limited geographic range. Moreover, knowledge of the mitochondrial genome can support sustainable aquaculture practices and help manage the genetic health of both wild and cultivated populations, ensuring long-term viability and preservation of this economically and ecologically important species.

The mitochondrial genome is widely regarded as an excellent molecular marker for conducting phylogenetic analyses among fish species [5, 6]. Most teleost mitogenomes are circular and contain 37 genes, including 13 protein-coding genes, 2 ribosomal RNA genes, 22 transfer RNA genes, and two noncoding regions: the control region and the origin of the light strand [7, 8]. A typical teleost mitochondrial genome is characterized by its compact size, fast evolutionary rate, conserved gene content, high copy number, absence of intermolecular genetic recombination, and maternal inheritance [9,10,11]. Comprehensive analysis and comparison of the complete mitochondrial genome and its unique characteristics allow for extensive use in species identification and the study of evolutionary relationships, phylogeographic, and population genetics in many teleost [12].

To date, approximately 87 taxonomically accepted species within the genus Melanotaenia have been identified [13]. However, many phylogenetic studies on this genus are only based on partial gene fragments. In contrast, complete mitochondrial DNA sequences provide a more comprehensive and higher-resolution dataset for phylogenetic analysis, offering better insights into evolutionary relationships. Currently, eight complete mitochondrial genomes of Melanotaenia species are available in GenBank, serving as valuable genetic resources that support the present study. These complete sequences enable a more robust phylogenetic assessment of M. fasinensis and its relationship with other species within the genus.

M. fasinensis was recently been classified as endangered by the International Union for the Conservation of Nature (IUCN) [14]. The population of rainbow ornamental fish is facing a significant threat due to the extensive freshwater fishing practices in rivers, which involve the use of pesticides, encroachment, and the conversion of forests into plantation areas. In addition, due to its restricted habitat in the karst formations, significant economic value as an ornamental fish, and insufficient genetic research conducted thus far, efforts should be made to study this species, e.g., to evaluate its phylogenetic placement of this species within melanotaeniid fishes. We utilized next-generation sequencing (NGS) technology to sequence and characterize the complete mitochondrial genome of the Fasin rainbow fish. We compared it with other Melanotaenia mitogenomes by examining genome organization, codon usage patterns, secondary structures of tRNA genes, and strand asymmetry. In addition, we reconstructed phylogenetic trees based on mitogenome sequences to confirm the taxonomic position of M. fasinensis and its relationships within the Melanotaeniidae family. The primary objective of this investigation was to obtain the mitochondrial genome of M. fasinensis using high-throughput sequencing technology and to evaluate the phylogenetic relationships of M. fasinensis with other species in the genus Melanotaenia. This study provides the first genomic insight into this species. The results of the study will also make a significant contribution to the genomic references for the family Melanotaeniidae.

Materials and methods

Samples and DNA extraction

Live samples of M. fasinensis (Fig. 1) were collected from the Fasin River near Ween Village, District Sawiat, South Sorong, West Papua Province, Indonesia. The specimens were caught by a cast net on Thursday, March 2, 2023. The geographic coordinates of the sampling site were 1°13.856’S, 131°58.186’E (Fig. 2). Samples were preserved in a solution of 95% ethanol and stored at -80⁰C. Local permits for specimen collection were obtained from Akademi Perikanan Sorong (APSOR) in West Papua. The collection followed the guidelines of the International Union for Conservation of Nature (IUCN) Red List Data’s Guidelines for Proper Utilization (version 4.0) [14]. The genomic DNA was extracted from the fin clip samples with the DNAeasy Blood and Tissue Kits (Qiagen, Germany). The quality of the DNA was observed using a 0.8% agarose gel. The DNA concentration was measured using the Qubit™ dsDNA HS Assay Kit and the QubitVR 2.0 fluorometer (Life Technologies, United States).

Fig. 1

Photograph of a male (top) and female (bottom). Melanotaenia fasinensis. Photo: Kadarusman

Fig. 2

Location of M. fasinensis sampling (red pins) in Fasin River, near Ween Village, District Sawiat, South Sorong, West Papua Province, Indonesia

Mitogenome sequencing and assembly

The sequencing process was carried out at the Genomic Laboratory, National Research and Innovation Agency, Cibinong, Indonesia utilizing the ELSA-BRIN Point mechanism. High-quality genomic DNA was used to construct a DNA library. Long reads were generated through PromethION sequencing (Nanopore, Oxford, UK). The sequencing library was prepared with the ligation sequencing kit (SQK-LSK110). Each library was run in an FLO-PRO002 flow cell for 48 h, following the manufacturer’s instructions. Default software parameters were used unless specified otherwise. The run was base called live using default settings in MinKNOW Core version 5.4.3, Bream version 7.4.8, and Guppy version 6.4.6 (Oxford Nanopore Technologies). Reads shorter than 4000 bp and those with quality scores below 7 were excluded from downstream analysis. A total of 50.2 Gb of sequence data was generated from PromethION sequencing (Nanopore, Oxford, UK).

Mitogenome annotation, visualization, and comparative analysis

Raw nanopore reads were assembled by using Flye v2.9.2 [15]. The mitochondrial genome was subsequently extracted from the assembled reads by using GetOrganelle v1.7.7.0 (2). To reduce the error rate of the assembled mitochondrial genome, we then polished the mitochondrial genome for two rounds by using NextPolish [16]. A BAM file was generated by mapping the reads to the final assembly with Minimap2 v2.24 [17], and a coverage mitogenome plot was created using bam2plot [18]. The read coverage depth map is shown in Supplementary Figure S1. Mitochondrial genome annotation and visualization were carried out in MitoFish (http://mitofish.aori.u-tokyo.ac.jp/) v4.05 [19, 20], and the start and stop codons of protein-coding genes were verified using Geneious Prime 2020.1 [21]. The PCGs and rRNA genes were manually verified using CLC Genomics Workbench 9.0 by aligning the new complete mitochondrial genome with the closely related Melanotaenia mitogenome [22]. The 22 tRNAs were identified and their secondary structures reconfirmed using the tRNAScan-SE server version 2.0 [23]. To verify the correct reading frame, nucleotide sequences of the mitogenomes were translated using MEGA XI [24]. Using the vertebrate mitochondrial genetic code as a basis, we identified CSBs by comparing the recognition sites within the control region of M. fasinensis to those of other Melanotaenia species. A circular map of the M. fasinensis mitogenome was constructed with the online server MitoAnnotator v4.05 [19, 20, 25]. The Mfold web server was utilized to predict the secondary structure of the putative origin of light strand replication [26]. We calculated the nucleotide composition and the relative synonymous codon usage (RSCU) using MEGA XI [24]. The nucleotide composition bias was calculated using AT skew = (A − T)/(A + T) and GC skew = (G − C)/(G + C). Ultimately, the fully annotated mitogenome was submitted to the NCBI database.

Phylogenetic analyses

We conducted phylogenetic analyses using all available mitochondrial genome sequences of Melanotaenia genus deposited in the GenBank database (Table S1). These sequences represent approximately 9% of the 87 taxonomically recorded species within the genus in FishBase. For outgroup comparison, we selected the families Atherinidae and Isonidae because they are the closest available relatives to Melanotaenia within the order Atheriniformes, as determined by sequence similarity in the NCBI database. This makes them appropriate for providing an evolutionary context to root the phylogenetic tree.

Mitogenome sequences were aligned using the Clustal W program with default settings in MEGA11 [24]. The phylogenetic relationships were inferred using IQ-TREE software version 1.6.12 [27]. The maximum likelihood (ML) method was employed, with the best-fit substitution model selected using ModelFinder [28]. The ModelFinder analysis identified the GTR + F + R3 model as the most suitable based on the Bayesian Information Criterion (BIC). To assess the robustness of the inferred phylogeny, we conducted ultrafast bootstrap (UFBoot) analysis with 10,000 replicates [29]. This analysis provides branch support values, indicating the confidence in the inferred relationships. The resulting phylogenetic tree was visualized using FigTree v1.4.2 software [30].

Results

Genome structure, organization and composition

The total length of the mitochondrial genome of M. fasinensis was found to be 16,731 bp as reported in GenBank (accession number: OR879116). There are 13 genes that code for proteins, 22 genes that code for tRNA, 2 genes that code for rRNA, and a noncoding regulatory region called the D-loop in the mitochondrial genome (Fig. 3 and Table 1). The M. fasinensis mitogenome sequence is larger than the mitogenome of other Melanotaenia species, ranging from 16,487 bp in M. lacustris to 16,731 in M. fasinensis (Table 2). Of these, 28 genes (12 protein-coding genes, 14 tRNAs, and 2 rRNAs) were located on the heavy strand (H strand), while the remaining genes (nd6 and eight tRNAs) are situated on the light strand (L strand) (Table 1). The comprehensive mitochondrial genome of M. fasinensis exhibited a base composition of 27.76% for adenine (A), 27.34% for thymine (T), 16.15% for guanine (G), and 28.75% for cytosine (C), with an A + T content of 55.10%. Furthermore, the A + T content of M. fasinensis was similar to that of other Melanotaenia species mitochondrial genomes (Table 2).

Overall, the complete mitogenome displayed a preference for A + T content, comprising 54.56% of the total mitogenome (Table 2). The highest A + T content was observed in the control region (CR) (61.92%), followed by transfer RNA sequences (tRNAs) (55.75%), protein-coding genes (PCGs) (54.4%), and ribosomal RNA sequences (rRNAs) (54.42%). Similarly, in the comparison of mitogenomes among 7 Melanotaenia species, including M. fasinensis, they showed relatively similar nucleotide compositions to each other (Table 2).

According to the sequenced mitogenome of M. fasinensis, the AT and GC skews were 0.008 and − 0.310, respectively. The AT skew differed across various other representative Melanotaenia species: M. praecox showed a value of 0.021, while M. australis exhibited 0.028. Likewise, the GC skew displayed variation, with M. australis, M. splendida, and M. fluviatilis registering − 0.310, and M. boesmani showing − 0.299, as depicted in Table 2. The predominantly weakly positive AT skew observed in most genes suggests a higher abundance of Adenine than Thymine throughout the complete mitogenome of Melanotaenia species. Additionally, in terms of absolute values, the GC skew consistently remained lower than the AT skew across all observed mitogenomes of Melanotaenia species (Table 2).

Fig. 3

The circular mitochondrial genome map of Melanotaenia fasinensis in this study. Genes oriented in the reverse direction are depicted in the outermost concentric ring, while those in the forward orientation are situated in the second outermost ring. The innermost rings of the image illustrate the percentage of GC content for every 5 base pairs of the mitogenome, with longer lines indicating higher GC percentages

Table 1 Gene annotation of the complete mitogenomes of M. fasinensis and their characteristic features

Table 2 Nucleotide composition and skewness of the mitogenome in 8 Melanotaenia species and 2 outgroup species

Overlapping and intergenic spacer regions

The complete mitogenome of M. fasinensis contained six overlapping sequences totaling 26 bp in length (Table 1). These sequences exhibited varying lengths, spanning from 1 bp to 10 bp, with the most extensive overlap observed on the H strand between atp8 and atp6 (10 bp), as well as between nd4l and nd4 (7 bp).

In addition, there were five short noncoding intergenic spacers across the sequenced mitogenome, for a total length of 93 bp, and five overlapping sequences were detected, for a total length of 29 bp (Table 1). The lengthiest spacer, spanning 68 base pairs, was located between trnY and the cox1 gene on the L strand. The comparative analysis revealed variation in the number of overlapping sequences ranging from 4 (M. praecox) to 7 (M. fluviatilis), with a length variation of length variation ranging from 9 bp to 30 bp in other Melanotaenia species (Table S2). Additionally, the longest intergenic spacer (69 bp) was observed between trnG and nd3 of M. boesemani, while 11 bp were found between trnY and cox1 of M. australis, M. lacustris, M. parkinsoni, and M. splendida. Conversely, 9 bp separated trnY and cox1 of M. praecox, and only 4 bp were present between tRNA-Asp and cox2 of M. fluviatilis, indicating greater diversity in both location and intergenic nucleotide lengths compared to the overlaps (Table S2).

Protein-coding genes

The collective length of protein-coding genes (PCGs) within the entire mitochondrial genome accounted was 10,757 bp, representing 64.29% of the total length. The 13 protein-coding genes within the mitochondria have a common initiation codon, ATG. The nucleotide sequence ‘TAA’, which serves as a complete stop codon, was observed in the genes nd2, cox1, nd4l, nd4, nd5, nd6, and cytb. Additionally, an incomplete stop codon represented by ‘T––’ was identified in the genes cox2, cox3, and atp6. Among the protein-coding genes examined, it was observed that the nd5 gene had the greatest length, measuring 1,839 base pairs (bp), while the atp8 gene had the smallest length, measuring 168 bp (Table 1).

The collective A + T content of the 13 PCGs in M. fasinensis was calculated to be 54.31%, with values ranging from 51.85% for nd4l to 57.16% for cox2 (Table 3). Furthermore, to assess the extent of base bias across all PCGs, base skews were determined. The AT skew and GC skew values for all 13 PCGs of M. fasinensis are depicted in Fig. 4. Notably, twelve of the thirteen PCGs exhibited significant negative GC skewness, with a deviation observed in the nd6 region, consistent with patterns observed in most teleost fishes. Like in other Melanotaenia mitogenomes, the majority of these PCGs started with the ATG codon, except for cox1, which commenced with the GTG.

Regarding stop codons, eight protein-coding genes (PCGs) were terminated by the standard codons (TAA and TAG), whereas the remaining genes terminated with incomplete codons (TA and T) (Table 1 and Table S3). The occurrence of these truncated termination codons is potentially widespread in Melanotaenia mitogenomes and is thought to be offset by posttranscriptional polyadenylation mechanisms [4, 22, 31,32,33,34].

Table 3 Composition of protein-coding genes in M. fasinensis

Fig. 4

Graphical representation of AT and GC skew for all the 13 protein-coding genes in the M. fasinensis mitogenome

The relative synonymous codon usage (RSCU) values of the PCGs in M. fasinensis (Fig. 4) were summarized along with those of other Melanotaenia species (Fig. S2). M. fasinensis encoded a total of 3611 amino acids in its PCGs. Analysis of codon usage patterns in these genes indicated that codons for leucine (11.58%), alanin (9.50%), threonine (7.70%), isoleucine (7.31%), and serine (6.62%) were the most frequently employed codons, while cycteine was less prevalent (0.72%). The distribution of amino acids and their relative frequencies in M. fasinensis corresponded with those observed in the mitogenomes of other Melanotaenia species. Significantly, the third position of the predominant codons (CUA-Leu1, AUU-Ile, CUU-leu2, CUC-leu2, and GCC-Ala) exhibited a preference for A and T, a pattern in line with earlier findings in Melanotaenia fish species (Fig. 5, Fig. S2).

Fig. 5

The relative synonymous codon usage (RSCU) of the mitochondrial protein-coding genes of M. fasinensis

Transfer and ribosomal RNA genes

M. fasinensis possessed a full set of 22 tRNAs, totaling 1,549 base pairs. The lengths of these tRNAs varied among individuals, ranging from 68 bp for trnF to 73 bp for trnQ, trnD, and trnK (Table 1). Among these tRNAs, 15 genes were located on the H-strand, while the remainder were located on the L-strand (Table 1). The anticodons found in all tRNAs within the M. fasinensis mitogenome were consistent with those typically observed in most teleost species. Generally, there was a direct match between codons and anticodons. However, serine was represented by two different anticodons (UGA and GCU), and leucine was represented by UAG and UAA in M. fasinensis.

The tRNA secondary structures were depicted in Fig. 6. Apart from tRNASer 2, most tRNAs were predicted to adopt typical cloverleaf secondary structures. Numerous noncomplementary base pairs were identified throughout the stems of these tRNAs. The tRNASer2 found in M. fasinensis did not possess an identifiable DHU stem or loop, which is a characteristic observed in other fish mitogenomes. The tRNASer2 identified in M. fasinensis lacked a recognizable DHU stem and loop or these abnormal tRNAs to function similarly to normal tRNAs. As in other fishes, M. fasinensis also had two rRNAs. Like other fishes. The 16 S rRNA gene spanned 1,671 base pairs and was located between trnV and trnL. Conversely, the 12 S rRNA gene included 945 base pairs and resided between trnF and trnV (Table 1). Both ribosomal RNAs were positioned on the H-strand and were separated by trnV.

Fig. 6

Putative secondary structures of 22 tRNA genes in the mitochondrial genome of M. fasinensis

Non coding regions

Like in many vertebrates, the control region of M. fasinensis was observed within a cluster of five tRNA genes (trnW, trnA, trnN, trnC, and trnY) between trnN and trnC (Table 1). The putative origin of replication for the light strand was 33 base pairs long. The region could fold into a stable stem loop secondary structure, with a stem formed by 8 paired nucleotides and a loop of 13 nucleotides.

In the Melanotaenia genus, M. fasinensis possesses two D-loop or control regions (CRs): one located between nd4 and trnH, with a length of 710 base pairs, and another positioned between trnP and trnF, spanning approximately 880 base pairs (Fig. 3). In addition, at the 5’ end of the CR, a domain with termination-associated sequences was observed, featuring three TACAT motifs and two corresponding palindrome sequences, ATGTA. Additionally, a conserved sequence block (CSB-2) was detected at the end of the CR (Table S4). However, unlike most other fishes, M. fasinensis has a central conserved sequence block (CSB) domain, and the CSB-1 and CSB-3 could not be recognized. The A + T content, AT skew and GC skew in CR were 60.75%, -0.01 and − 0.25, respectively, which were also nearly consistent with the findings of previous reports on other Melanotaenia studied here (Table 2).

Phylogeny

The mitogenome arrangement of M. fasinensis closely resembles that of various species within the Melanotaenia genus, such as M. lacustris, M. boesemani, M. australis, M. splendida, M. parkinsoni, M. fluviatilis, and M. praecox, as well as species from Iriatherina genera. Similarities were observed, ranging from 88.66 to 83.64%. Phylogenetic trees were constructed using maximum likelihood, incorporating 11 species of Melanotaeniidae family, as well as the families Atherinomorus and Isonidae as the outgroup. This analysis revealed the existence of two primary clades, labeled as clades I and II (Fig. 7, Table S1). The results indicated a high level of agreement among the ML trees in term of their overall structure and exhibited strong support values. The phylogenetic tree analysis indicated that all species of the Melanotaeniidae species clustered together on the same branch. Furthermore, the intergeneric and interspecific taxonomic positions were explicit and clear. Phylogenetically, Melanotaeniidae were more closely related to the family Isonidae than to the Atherinomorus family. The phylogenetic position of M. fasinensis was relatively at the basal position in the genus Melanotaenia. The results of this study could help to characterize the evolutionary correlation between M. fasinensis and other rainbowfish species residing in the region of West Papua. The incorporation of additional taxa from the genus Melanotaenia in the phylogenetic analyses in the future may help to better understand the phylogenetic relationships and evolutionary history among species in this genus.

Fig. 7

Phylogenetic tree of M. fasinensis and 8 other Melanotaenia species. Two species of the families Atherinidae and Isonidae were designated to be outgroups. The numbers at each node indicate the bootstrap support values obtained by 10,000 replications

Discussion

In the present study, we found that the M. fasinensis mitogenome sequence is larger than the mitogenome of other Melanotaenia species. Within this group, 28 genes, comprising 12 protein-PCGs, two rRNAs, and 14 tRNAs, were situated on the heavy strand (H strand), while the remaining genes, including nd6 and 8 tRNAs, were positioned on the light strand (L strand). Like in other vertebrates, the majority of the genes were encoded on the H strand, with the exception of nd6 and eight tRNA genes. Additionally, all the genes exhibited similar lengths to those found in other bony fishes [35]. In addition, the gene arrangements in both the L and H strands were canonically identical and consistent with those of other Melanotaenia species [22, 34]. Genes located on the L-strand exhibited a notable preference for thymine (T) at the codon wobble position, while adenine or cytosine ending codons were disproportionately represented in genes on the H-strand [36].

The A + T content of M. fasinensis closely resembled that of mitochondrial genomes in other species within the Melanotaenia genus. The A + T content observed in the L-strand closely resembled that of other teleost mitochondrial genomes [37, 38]. This suggested that the mitochondrial genome of M. fasinensis exhibited the typical arrangement found in both teleosts and vertebrates. Genes encoded on the L-strand exhibited a notable preference for thymine at the wobble codon, while adenine or cytosine ending codons were more prevalent in the genes on the H-strand. This strand-specific bias is believed to stem from asymmetrical directional mutation pressure [36].

The predominance of adenine over thymine in most genes of the complete mitogenome of Melanotaenia species is indicated by a weakly positive AT skew. Additionally, the absolute value of the GC skew consistently remained lower than that of the AT skew across all the surveyed mitogenomes of Melanotaenia species. This discrepancy may arise from strand asymmetry, suggesting a strand compositional bias and a possible violation of Chargaff’s second parity rule. This could be the result of strand asymmetry, also known as strand compositional bias, which arises from the violation of Chargaff’s second parity rule [5, 39].

Overlapping of nucleotides between adjacent genes is a common characteristics found in teleost mitogenomes and aids in compacting the mitogenomes [37, 40]. The overlapping regions between nd4l/nd4 and atp8/atp6 were identified as common features found in Melanotaenia species [4, 22, 31,32,33,34] and other teleost [41,42,43,44]. In addition, in this study, there was greater diversity in both location and intergenic nucleotide length than in the overlapping regions. The same phenomenon has been observed in other teleost organisms [5, 12, 41, 45, 46].

In the majority of Melanotaenia species, a consistent trait is evident: atp8 ranks as the shortest while nd5 stands as the longest among the PCGs [4, 22, 31,32,33,34]. This pattern mirrors the common occurrence observed in various other teleost fish species [47,48,49,50]. Like those in other Melanotaenia mitogenomes, the majority of protein-coding genes (PCGs) in this study were found to be associated with the ATG codon, with the exception of cox1, which was initiated by GTG. Across fish mitogenomes, conventional start codons are predominantly utilized for these genes, while alternate start codons are rare within the Melanotaenia genus, a trend also observed in teleost and other eukaryotic organisms [36, 51, 52]. Concerning stop codons, eight protein-coding genes (PCGs) concluded with standard codons (TAA and TAG), while the rest contained truncated codons (TA and T). Incomplete termination codons are likely widespread in Melanotaenia mitogenomes and are potentially compensated for by posttranscriptional polyadenylation mechanisms [4, 22, 31,32,33,34].

The presence of anticodons in all tRNAs in the M. fasinensis mitogenome was in line with that commonly observed in the majority of teleost species. Overall, there was a direct correspondence between codons and anticodons. Nevertheless, serine was encoded by two distinct anticodons (UGA and GCU), and leucine was encoded by UAG and UAA in M. fasinensis. It is common for teleost mitogenomes to have multiple tRNAs recognizing different anticodons [5, 49].

Several nonmatching base pairs were detected within the stems of these tRNAs. These mismatches observed in tRNA sequences appeared to be common in teleost mt-tRNA genes [53, 54]. It is probable that the mismatches present in the stems of these tRNAs underwent modifications via posttranscriptional editing mechanisms [55, 56]. The tRNASer2 found in M. fasinensis did not possess an identifiable DHU stem or loop, which is a characteristic observed in other fish mitogenomes [6, 57]. The tRNASer2 found in M. fasinensis did not feature a recognizable DHU stem or loop, a feature observed in other fish mitogenomes [6, 57], or these atypical tRNAs operate in a manner akin to that of regular tRNAs. These proteins may necessitate coevolved interacting factors or posttranscriptional RNA editing [5, 36].

As in other fishes, M. fasinensis also had two rRNAs. The 16 S large ribosomal gene was located between trnV and trnL. Conversely, the 12 S small ribosomal gene resided between trnF and trnV. Both ribosomal RNAs were positioned on the H-strand and were separated by trnV, which aligns with a typical pattern found in most vertebrates [11, 58, 59].

The CR, also known as the control region, is the largest noncoding segment within vertebrate mtDNA and commonly hosts vital elements essential for both replication and transcription functions [10, 60, 61]. Unlike its counterparts within the Melanotaenia genus, M. fasinensis is unique in having two control regions (CRs). The CR is the longest noncoding segment within vertebrate mitochondrial DNA and typically contains crucial elements essential for replication and transcription processes [12, 37, 62]. CR1, spanning a length of 710 base pairs, was identified between nd4 and trnH (Fig. 3). However, typical features associated with control regions, such as conserved sequence blocks (CSBs), as identified by [63], or termination-associated sequences (TASs) as identified by [64], are lacking. The results of this study align with the earlier documented findings concerning bony fish [65]. Furthermore, CR2, positioned between trnP and trnF, spans approximately 880 base pairs. At the 5’ end of the CR, a region containing termination-associated sequences was noted, comprising three TACAT motifs and two corresponding palindrome sequences, ATGTA. Additionally, a conserved sequence block (CSB-2) was identified at the termination of the CR. This block plays a role in positioning RNA polymerase during transcription and priming replication [58]. However, unlike most other fishes, M. fasinensis has a central conserved sequence block (CSB) domain, and CSB-1 and CSB-3 could not be recognized. This study also aligns with previous reports on other teleost mitogenomes, noting that only certain portions of CSB in fish were detected [58, 59]. Moreover, the overall arrangement of the CR resembled what has been documented for other Melanotaenia species [6, 21, 32]. The A + T content, AT skew, and GC skew observed in the control region (CR) were similar to the results reported in previous studies on other Melanotaenia species examined in this study [4, 22, 31, 35].

The mitochondrial genome arrangement of M. fasinensis closely resembled that of all species within the Melanotaenia species [4, 22, 31, 33, 34], as well as the Iriatherina genera [35, 66]. Using whole mitogenomic data, 11 species of Melanotaenia were analyzed using maximum likelihood analyses, revealing the presence of two major clades. Moreover, from a phylogenetic perspective, the Melanotaeniidae family shares a closer evolutionary relationship to the Isonidae family than with the Atherinomorus family. Our phylogenetic analysis shows M. fasinensis in a basal position within Melanotaenia, indicating that it may represent one of the more ancestral lineages within the genus. This is consistent with previous studies that suggest West Papua is a center of origin and diversification for rainbowfishes [3, 67], including Melanotaenia. The observed genetic similarities between M. fasinensis and other species within Melanotaenia, and Iriatherina, further indicate a shared evolutionary history.

This study demonstrates the enhanced phylogenetic inference achieved by using whole mitogenome data. By incorporating information from multiple genes and non-coding regions, we provide a more robust and detailed evolutionary framework compared to gene fragment-based approaches. While previous studies utilizing the cytb gene have been successful in identifying broader phylogenetic relationships, they may lack the resolution to elucidate finer-scale patterns or recent divergence events. The congruence observed between our whole mitogenome analyses and cytb gene-based studies indicates that both methods can yield similar topologies [67]. However, whole mitogenome data offer a more comprehensive and well-supported view of evolutionary relationships. The placement of genera like Iriatherina near Melanotaenia in our phylogenetic trees highlights the advantages of using complete mitogenomic data. Such relationships may have been less clear when relying solely on partial gene sequences. This underscores the importance of incorporating whole mitogenome data for a more accurate and informative phylogenetic reconstruction.

Conclusion

This study presents the complete mitogenome of the rainbow fish (Melanotaenia fasinensis), which spans 16,731 base pairs. The arrangement and gene composition of the mitogenomes of M. fasinensis closely resemble those of other Melanotaenia species, with the notable observation of two distinct control regions in M. fasinensis. This observed variation could lead to the use of an important tool for the identification of Melanotaenia at the genus level. Phylogenetic analyses using the mitogenome confirmed that Melanotaeniidae showed closer relation to the Isonidae family rather than to the Atherinomorus family. Within the Melanotaenia genus, M. fasinensis appeared relatively basal in its phylogenetic position. The comprehensive mitogenome data presented here will serve as a valuable resource for future investigations into the molecular taxonomy, genetic diversity, and population structure of M. fasinensis. Additionally, this study provides new perspectives that could enhance conservation strategies for this species.