Comparative Mitochondrial Features Across Characiformes (Teleostei: Ostariophysi) and Mitogenomic Architecture of Nematobrycon lacortei

Abstract

Characiformes, a highly diverse group, are of great economic and ecological significance; however, their complex morphological convergence and adaptive radiation pose challenges to taxonomy, and phylogenetic controversies remain. In this study, the complete mitochondrial genome (mitogenome) of Nematobrycon lacortei was sequenced and a comparative analysis within 51 Characiformes mitogenomes was conducted. The mitogenome of N. lacortei was a 17,585 bp circular DNA molecule, with standard organization, and showed an A/T bias with an A + T content of 62.29%. The nucleotide composition analysis of Characiformes mitogenomes revealed consistency, and the A + T content exceeded the G + C content throughout. Analyses of protein-coding genes (PCGs) revealed that purifying the selection dominates their evolution, with ND1 and ND6 having the lowest Ka/Ks values, while ND3 has the highest. Phylogenetic analyses were conducted on 51 mitogenomes from 15 families of Characiformes. The BI and ML trees supported the classification of most families but showed that Curimatopsis evelynae in Curimatidae did not cluster with its family members; in addition, relationships within Characidae were unstable. These findings contribute to the research on the evolution of Characiformes and highlight the need for more mitogenome data from diverse species in order to further clarify their phylogenetic relationships and evolutionary history.

1. Introduction

Characiformes, a highly diverse group within Ostariophysi, are characterized by remarkable species richness and have ecological implications. The order Characiformes comprises over 2000 fish species, widely distributed in the freshwater ecosystems of South America and Africa [1]. Such fish species exhibit a wide range of body sizes, ranging from miniature and diminutive species to midsized and giant species (varying from 20 mm to over 20 cm) [2]. The species within this order are of crucial importance both economically and ecologically [3]. Certain species, such as Prochilodus mariae, play a pivotal role in the internal energy flow and material cycling within the ecosystems of river systems, and they also act as ecosystem engineers [4]. Their prominent morphological convergence and adaptive radiation have long been hot topics in evolutionary biology, but they also pose significant challenges to taxonomy [3,5]. Despite the continuous development of molecular technology, phylogenetic controversies remain in the order Characiformes between molecular phylogenies and morphology-based classifications [5]. This has greatly restricted our in-depth understanding of evolution within this group.

Mitochondrial genomes (mitogenomes) possess unique value in evolutionary research. Mitochondrial DNA, characterized by its rapid evolutionary rate, compact size, and simple structure, has been widely applied as a useful tool in taxonomy, population genetics, and phylogenetic analysis [6,7]. Typically, animal mitogenomes contain 37 highly conserved genes and adhere to a strict maternal inheritance pattern [8,9]. Mitogenomes also contain 13 protein-coding genes (PCGs), 22 transfer RNA genes (tRNAs), 2 ribosomal RNA genes (rRNAs), and 1 non-coding control region (CR) [10]. In addition, features such as gene order and tRNA secondary structure in the mitogenome contain rich phylogenetic information, providing powerful indicators for clarifying complex evolutionary relationships [11]. The patterns of nucleotide substitution, biases in codon usage, and alterations in gene structures all record key events in the evolutionary history of species [12,13]. Through the analysis of mitochondrial gene sequences, genetic differences between species can be accurately identified, thereby inferring their relationships.

To address the deficiencies in mitogenomic data availability and phylogenetic resolution within Characiformes, we carried out a series of works. First, partial mitochondrial sequences provide only limited information, lacking detailed data [14]. Therefore, the complete mitogenome of Nematobrycon lacortei, a species with scarce genomic research in Characidae, was sequenced to supplement the mitogenome database of Characiformes and the structural characteristics of N. lacortei were deeply analyzed. Second, a large-scale integrative analysis of mitogenomes in Characiformes was conducted, covering the complete mitogenomes of 51 species from 15 families (

Table 1. Information on the mitogenomes analyzed in this study.

Order	Family	Species	Size (bp)	Accession No.
Characiformes	Alestiidae	Phenacogrammus interruptus	16,652	AB054129.1
	Anostomidae	Abramites hypselonotus	16,685	MW541938.1
		Megaleporinus elongatus	16,774	KU980144.1
		Megaleporinus obtusidens	16,682	KY825191.1
	Bryconidae	Brycon henni	16,885	KP027535.1
		Brycon nattereri	16,837	MT428073.1
		Brycon orbignyanus	16,802	KY825192.1
		Salminus brasiliensis	17,721	KM245047.1
	Characidae	Astyanax aeneus	16,769	BK013055.1
		Astyanax mexicanus	16,768	BK013062.1
		Hemigrammus armstrongi	16,789	MW742324.1
		Hemigrammus bleheri	17,021	LC074360.1
		Hemigrammus erythrozonus	16,710	MT484070.1
		Hyphessobrycon amapaensis	17,824	MW742322.1
		Hyphessobrycon elachys	17,224	MW315747.1
		Hyphessobrycon flammeus	16,008	MW315748.1
		Hyphessobrycon herbertaxelrodi	17,417	MT769327.1
		Hyphessobrycon pulchripinnis	17,618	MW331227.1
		Moenkhausia costae	15,811	MW366831.1
		Moenkhausia sanctaefilomenae	18,437	MW407181.1
		Nematobrycon lacortei	17,585	PP760379
		Nematobrycon palmeri	17,340	MN861079.1
		Paracheirodon axelrodi	17,100	AB898197.1
		Paracheirodon innesi	16,962	KT783482.1
	Chilodontidae	Chilodus punctatus	16,869	AP011984.1
	Citharinidae	Citharinus congicus	16,453	AP011985.1
	Curimatidae	Curimata mivartii	16,705	KP025764.1
		Curimatopsis evelynae	16,779	AP011988.1
		Ichthyoelephas longirostris	16,840	KP025763.1
		Prochilodus magdalenae	16,692	PQ510211.1
		Prochilodus argenteus	16,697	KR014816.1
		Prochilodus costatus	16,699	KR014817.1
		Prochilodus lineatus	16,699	KM245045.1
	Distichodontidae	Distichodus sexfasciatus	16,555	AB070242.1
	Erythrinidae	Hoplias intermedius	16,629	KU523584.1
		Hoplias malabaricus	16,638	AP011992.1
	Gasteropelecidae	Carnegiella strigata	17,852	AP011983.1
	Hemiodontidae	Hemiodopsis gracilis	16,731	AP011990.1
	Hepsetidae	Hepsetus odoe	16,803	AP011991.1
	Lebiasinidae	Lebiasina multimaculata	16,899	AP006766.1
		Nannostomus beckfordi	16,690	OR857846.1
		Nannostomus marilynae	16,667	OR857847.1
		Nannostomus marginatus	16,661	OR857848.1
		Nannostomus unifasciatus	16,681	OR857849.1
	Parodontidae	Apareiodon affinis	16,679	AP011998.1
	Serrasalmidae	Colossoma macropomum	16,703	KP188830.1
		Metynnis hypsauchen	16,737	MH358334.1
		Myloplus rubripinnis	16,662	MH358336.1
		Piaractus brachypomus	16,722	KJ993871.2
		Piaractus mesopotamicus	16,722	KM245046.1
		Pygocentrus nattereri	16,706	AP012000.1
Cypriniformes	Cyprinidae	Cyprinus carpio	16,592	OL693871.1

). Through the comprehensive application of nucleotide analysis, codon analysis, and implementing phylogenetic analysis for Bayesian inference (BI) and maximum likelihood (ML), we aimed to comprehensively and deeply reveal the comparative evolutionary patterns of Characiformes and provide new insights and strong evidence for phylogenetic controversies.

2. Materials and Methods

2.1. Sample Collection and DNA Extraction

For this study, a specimen of N. lacortei (native to South America) was procured from the Nanjing Pet Market located in Nanjing, China. Species identification followed the morphological characteristics described in FishBase (https://www.fishbase.se/home.htm, accessed on 1 April 2025), and molecular identification was performed according to 12S rRNA from GenBank (LC036727.1). This study was approved by the ethics committee of Nanjing Forestry University (No. 2022001). No animals were killed in this study. Genomic DNA was isolated from the tail fin sample using a FastPure Cell/Tissue DNA Isolation Mini Kit (Vazyme, Nanjing, China) while strictly adhering to the manufacturer’s provided protocol. The high-quality DNA was stored at −20 °C for use in future experimental procedures.

2.2. Next-Generation Sequencing

Library preparation and sequencing were conducted by Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China). Sequencing was performed using the NovaSeq X Plus platform (Illumina, San Diego, CA, USA), yielding 150 bp paired-end reads. This approach was adopted to ensure comprehensive and accurate data acquisition. The depth of the sequencing was 3×. To ensure the production of high-quality data, stringent quality control measures were put in place to remove low-quality sequences. Using Fastp for quality assessment and data filtering of sequencing data, the main criteria include removing adapter contamination at the 3′ end, performing quality filtering using the sliding window method with a window size of 9 bp and a step size of 1 bp, and removing sequences containing more than 5 Ns. After the filtering process, the resulting clean reads were used for assembling the complete mitogenome. The assembly was performed using the Geneious Prime 2024 software, with Nematobrycon palmeri (MN861079.1) as the reference template [15]. The assembly was carried out under the medium sensitivity/speed setting, which effectively balanced the requirements for accuracy and efficiency. Consensus sequences were generated with a base call threshold of 50%, which ultimately led to the acquisition of the complete mitogenome.

2.3. Annotation and Sequence Analysis

Conservative domains within the mitogenome were identified through two distinct computational approaches: BLAST CD-Search and the MITOS server [16,17]. A gene map was generated via the CG View server [18]. The details of the mitogenomes analyzed in this study are presented in Table 1. Nucleotide bias, a crucial metric in understanding the compositional characteristics of the mitogenome, was acquired using the formulas “AT-skew = (A − T)/(A + T)” and “GC-skew = (G − C)/(G + C)” [19]. Furthermore, analyses of relative synonymous codon usage (RSCU), as well as non-synonymous (Ka) and synonymous substitutions (Ks), were carried out using MEGA 12 software [20]. The visualization of RSCU data was performed using PhyloSuite v1.2.3 [21], and tRNAscan-SE 2.0 was utilized for the prediction of tRNA secondary structures [22].

2.4. Phylogenetic Analysis

A total of 52 mitogenomes were selected from GenBank for phylogenetic analysis, including 16 families (Table 1); only verified complete mitogenomes were selected. Among these mitogenomes, Cyprinus carpio (OL693871.1) was used as the outgroup. All procedures were completed using the PhyloSuite v1.2.3 software package [21]. Sequences of 13 PCGs were extracted and were then used for the phylogenetic analyses. MAFFT v7.313 was used for multiple gene alignment [23]. Sequence pruning in the form of triplet codons was performed using Gblocks [21]. The constructed PCGs generated concatenate sequences in PhyloSuite. With Bayesian information criterion (BIC) as the standard criterion, partition analysis was performed for MrBayes and IQ-TREE using ModelFinder v2.2.0 [24]. The best-fit model was GTR + F + I + G4 for BI and ML. The BI tree was reconstructed using MrBayes v3.2.6 with four Markov chains [25]. Two independent runs of 1,000,000 generations were conducted, with sampling carried out every 1000 generations. The first 25% of tree samples were deleted to reduce the number of simulation errors. IQ-TREE was used to reconstruct the ML tree with 5000 bootstraps [26]. Phylogenetic trees were visualized and edited using iTOL (https://itol.embl.de, accessed on 1 April 2025) [27].

3. Results and Discussion

3.1. Mitochondrial Genome Organization

The complete mitogenome sequence of N. lacortei was a classical circular DNA molecule, with a size of 17,585 bp (Figure 1). The mitogenome is consistent with those of other fish species in the order Characiformes [28,29]. It contains 13 PCGs, 22 tRNAs, 2 rRNAs, and 1 CR, and the gene order is also identical (

Table 2. General features of the mitogenome of Nematobrycon lacortei.

Gene	Position		Size (bp)	Orientation	Codon		Intergenic Nucleotides (bp)
	From	To			Start	Stop
tRNA-Phe	1	68	68	+			0
12S rRNA	69	1017	949	+			0
tRNA-Val	1018	1089	72	+			0
16S rRNA	1110	2761	1652	+			20
tRNA-Leu	2764	2838	75	+			2
ND1	2839	3810	972	+	ATG	TAA	0
tRNA-Ile	3819	3890	72	+			8
tRNA-Gln	3889	3959	71	−			−2
tRNA-Met	3972	4041	70	+			12
ND2	4043	5101	1059	+	ATG	TAA	1
tRNA-Trp	5126	5196	71	+			24
tRNA-Ala	5160	5235	76	−			−37
tRNA-Asn	5238	5310	73	−			2
tRNA-Cys	5341	5407	67	−			30
tRNA-Tyr	5407	5476	70	−			−1
COX1	5480	7039	1560	+	ATG	AGG	3
tRNA-Ser	7027	7098	72	−			−13
tRNA-Asp	7103	7174	72	+			4
COX2	7190	7880	691	+	ATG	T	15
tRNA-Lys	7881	7953	73	+			0
ATP8	7955	8122	168	+	ATG	T	1
ATP6	8113	8794	682	+	ATG	T	−10
COX3	8975	9578	604	+	ATT	T	180
tRNA-Gly	9579	9651	73	+			0
ND3	9652	10,002	351	+	ATG	TAA	0
tRNA-Arg	10,001	10,070	70	+			−2
ND4L	10,071	10,367	297	+	ATG	TAA	0
ND4	10,361	11,741	1381	+	ATG	T	−7
tRNA-His	11,742	11,810	69	+			0
tRNA-Ser	11,811	11,878	68	+			0
tRNA-Leu	11,880	11,952	73	+			1
ND5	11,953	13,791	1839	+	ATG	TAA	0
ND6	13,788	14,303	516	−	ATG	TAA	−4
tRNA-Glu	14,304	14,371	68	−			0
Cytb	14,377	15,513	1137	+	ATG	TAA	5
tRNA-Thr	15,519	15,591	73	+			5
tRNA-Pro	15,590	15,659	70	−			−2
CR	15,660	17,585	1926	/			0

).

Nucleotide composition analysis indicated that the mitogenome of N. lacortei was biased toward A/T. Its PCGs, tRNAs, rRNAs, and CR also showed a preference for A/T (

Table 3. Composition and skewness of Nematobrycon lacortei mitogenome.

Region	Size (bp)	A (%)	T (%)	G (%)	C (%)	A + T (%)	G + C (%)	AT Skew	GC Skew
Total genome	17,585	31.02	31.27	14.75	22.96	62.29	37.71	−0.004	−0.218
PCGs	11,257	29.51	32.84	13.54	24.11	62.35	37.65	−0.053	−0.281
tRNAs	1566	31.80	26.56	18.14	23.50	58.36	41.64	0.090	−0.129
rRNAs	2601	34.68	23.03	19.30	22.99	57.71	42.29	0.202	−0.087
CR	1926	34.84	37.33	13.14	14.69	72.17	27.83	−0.035	−0.056

). This phenomenon was consistent with the reported mitogenomes of fish species [30,31]. The skewness of this mitogenome showed a negative GC skew, revealing a bias toward the usage of C rather than G in the genome.

To determine the nucleotide composition on the order Characiformes, the AT skew and A + T content were computed for 51 mitogenomes, encompassing 15 families: Alestiidae, Anostomidae, Bryconidae, Characidae, Chilodontidae, Citharinidae, Curimatidae, Distichodontidae, Erythrinidae, Gasteropelecidae, Hemiodontidae, Hepsetidae, Lebiasinidae, Parodontidae, and Serrasalmidae (Figure 2). The nucleotide composition analysis of the 51 Characiformes mitogenomes revealed consistency. Across the total genome, as well as in PCGs, tRNAs, and rRNAs, the A + T content exceeded the G + C content. The AT skews predominantly exhibited positive values, indicating that A generally occurred more frequently than T. Significantly, the ratio of AT skew to AT content was nearly consistent among the total genome, PCGs, tRNAs, and rRNAs.

Multiple overlaps between adjacent genes were identified. Nine gene overlaps were detected in N. lacortei. The most extensive overlap was found between tRNA-Trp and tRNA-Ala.

3.2. Protein-Coding Genes and Codon Usage

The cumulative size of the PCGs measured 11,257 bp, constituting 64.01% of the mitogenome. With the exception of ND6, all PCGs were situated on the major strand, as detailed in Table 2. Among the 13 PCGs within the mitogenome, ATP8 was the smallest, with 168 bp, whereas ND5 was the largest, stretching to 1839 bp.

The Ka/Ks ratios of Nematobrycon PCGs was calculated (Figure 3a). The Ka/Ks ratios for all PCGs were determined to be less than 1. This finding implies that purifying selection likely plays a dominant role in molding the evolutionary patterns of PCGs. In essence, under most circumstances, selection acts to remove deleterious mutations, thereby maintaining the stability of the protein structure and function [32]. Among the PCGs, ND1 and ND6 exhibited the lowest Ka/Ks values. This suggests that these genes have been subjected to intense selective pressure, leading to a slow rate of evolution [33]. On the contrary, ND3 displayed the highest average Ka/Ks value, which indicates a relatively high rate of non-synonymous substitutions within its PCG sequence. Furthermore, the RSCU analysis revealed a preference for A and T nucleotides at the third codon position (excluding Met) (Figure 3b). This observation is consistent with the noted bias towards the use of A + T nucleotides, as evidenced by the nucleotide composition of the PCGs.

Within the mitogenome, the majority of PCGs started with the ATG codon (Table 2). An exception was COX3, which utilized the alternative initiation codon ATT, a phenomenon also detected in the mitogenomes of other animals [34]. The termination codons of these PCGs were quite varied, consisting of TAA, AGG, and T. Significantly, the incidence of the TAA was persistently greater than that of the other two termination codons, with the AGG termination codon being the least frequently occurring.

Thereafter, an investigation into the utilization of initiation and termination codons across the 51 mitogenomes was conducted (Figure 4). In the studied Characiformes species, a distinct conservative trend emerged in the selection of initiation codons. As shown in Figure 4a, there was a pronounced preference for the ATG codon, which is consistent with the results of previous studies [28,29,30,31,34]. Nevertheless, an exception was observed in COX1, which predominantly began with GTG. When it comes to termination codons, all the investigated Characiformes species made use of TAA, TAG, AGG, and T (Figure 4b). Notably, ND1, ATP8, ATP6, COX3, ND4L, and ND5 mainly terminated with TAA. Conversely, COX2, ND3, ND4, and Cytb predominantly employed T as their termination codon.

3.3. Transfer RNA, Ribosomal RNA Genes, and Control Region

Twenty-two tRNAs were dispersed across the mitogenome. The cumulative size of the tRNA regions amounted to 1566 bp, constituting 8.91% of the total size of the mitogenome. Among these tRNAs, eight were transcribed from the minor strand. The secondary structures of these tRNAs are illustrated in Figure 5a. In addition to the typical A-U and C-G, several mismatched base pairs were identified within the various stem regions of the tRNAs. Specifically, A-C, C-U, U-G, and A-G mismatches were detected among these tRNAs.

Within the mitogenome, both rRNAs, namely, the 12S rRNA and 16S rRNA, were transcribed from the major strand (Table 2).

The CR was located between the tRNA-Pro and tRNA-Phe (Table 2). The CR accounts for 10.95% of the mitogenome. Notably, the A + T content of the CR consistently exceeded that of the PCGs and RNAs, suggesting a higher prevalence of A/T (Table 3). The structural organization of the CR in the mitogenome of N. lacortei, as depicted in Figure 5b, featured a significant number of repeating units.

3.4. Phylogenetic Analyses

The primary objective of phylogenetic analysis was to elucidate the phylogenetic relationships among taxa within the order Characiformes. Both BI and ML trees exhibited a comparable topological structure, and most of the clades were supported by statistical values (Figure 6).

The families Citharinidae and Distichodontidae diverged from other families early in the evolutionary history of Characiformes. The phylogenetic trees provided substantial support for the established classification of each family within the order Characiformes. However, an exception was noted for Curimatopsis evelynae, belonging to the family Curimatidae, which do not cluster with other Curimatidae species. This phylogenetic incongruence may arise from limitations of mitochondrial markers in resolving recently diverged lineages due to incomplete lineage sorting. Future studies combining nuclear loci and morphological reappraisals, as proposed by Melo et al. [35], are crucial to clarify its systematic position. In contrast, all other species are successfully grouped with species within the same family. Furthermore, the phylogenetic trees also highlighted the incongruent relationships within the family Characidae, a finding that aligns with previous research [5]. Certain species within the family Characidae were found to cluster with species from different genera, such as Hemigrammus armstrongi (Figure 6). The clustering of species in the Hemigrammus, Hyphessobrycon, and Moenkhausia genera is confusing; this might also be due to the limitations of mitochondrial markers. The results of the phylogenetic analysis in this study are similar to those of Javonillo et al. in relation to phylogeny within the order Characiformes [36]. Moreover, a substantial number of Characiformes species exist whose complete mitogenomes remain unpublished. Currently, our understanding of the structure of Characiformes mitogenomes is still rather limited and far from being comprehensive. Consequently, there is an imperative need to acquire mitogenome data from a greater diversity of species within the order Characiformes.

In conclusion, this study presents the first complete mitogenome of N. lacortei and a comparative analysis across 51 Characiformes species. Our findings confirm a conserved mitogenomic architecture and a strong A/T bias within the order Characiformes, while purifying the selection dominates PCG evolution, particularly in ND1 and ND6. Phylogenetic analyses supported most familial relationships but revealed inconsistencies in Curimatidae and Characidae, highlighting persistent incongruences between molecular and morphological classifications. In addition, since only the complete mitochondrial genome of one species was sequenced in this study, for the highly diverse Characiformes, the current mitochondrial data are still insufficient. This indicates that it is necessary to continue to supplement the mitochondrial database of Characiformes, ultimately advancing the resolution of their phylogeny and adaptive diversification patterns.