The seven species are commonly sold as ornamental fish and can be found in many pet markets. Specimens were obtained from the Qiqiaoweng pet market in Nanjing, Jiangsu province, China. The specimens were identified using morphological characteristics described in FishBase (https://www.fishbase.de/). No fish were sacrificed during this study. The fish were reared at the Laboratory of Animal Molecular Evolution, Nanjing Forestry University. Total genomic DNA was extracted from each fin using a FastPure Cell/Tissue DNA Isolation Mini Kit (Vazyme, Nanjing, China), and stored at –80 °C for future use.

Seven complete mitogenomes were sequenced on an Illumina platform (Personalbio Nanjin, China) using total genomic DNA. The genomic DNA was used to generate an Illumina library with an insert size of 400 bp. The clean data were then assembled in Geneious Prime 2022 software, using Lamprologus signatus (MZ427900.1) as a template. The mitogenomes were assembled and manually revised using DNAstar v. 7.1 (Madison, WI, USA).

Conservative domains were detected using BLAST (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and MITOS WebServer (http://mitos.bioinf.uni-leipzig.de/index.py) (Bernt et al. 2013). Maps of the mitogenomes were constructed using CGView (https://cgview.ca/) (Stothard and Wishart 2005). MEGA X was used for base composition analysis, relative synonymous codon usage (RSCU) analysis, pairwise relative genetic distance (p-distance) calculation, as well as non-synonymous (Ka) and synonymous substitutions (Ks) analysis (Fay and Wu 2003; Kumar et al. 2016). Composition skew values were calculated using the following formulas: “AT-skew = (A − T) / (A + T) GC-skew = (G − C) / (G + C)” (Perna and Kocher 1995).

Phylogenetic analysis was conducted using the sequences of 13 PCGs and two rRNA genes from the complete mitogenomes of 105 species, including seven species from this study (Suppl. material 1). Channa andrao and Hyphessobrycon sweglesi were selected as outgroups, while the remaining specimens belonged to the Cichlidae family. Phylogenetic analysis was conducted using maximum likelihood (ML) and Bayesian inference (BI) methods with PhyloSuite v. 1.2.3 software package (Zhang et al. 2020; Xiang et al. 2023). All genes were aligned using MAFFT v. 7.313, and the best-fit substitution model and partitioning scheme were determined using ModelFinder. ML phylogenies were inferred using IQ-TREE with the Edge-linked partition model for 5000 ultrafast bootstraps (Minh et al. 2013; Nguyen et al. 2015). BI phylogenies were inferred using MrBayes v. 3.2.7a with a partition model (Ronquist et al. 2012). The analysis consisted of two parallel runs with 2,000,000 generations each, and the initial 25% of sampled data was discarded as burn-in. The trees were visualized and edited using iTOL v. 6 (Letunic and Bork 2021).

Seven complete mitogenomes covering two genera were obtained. L. kungweensis (16,587 bp), L. meleagris (16,582 bp), L. ornatipinnis (16,585 bp), N. brevis (16,586 bp), N. caudopunctatus (16,586 bp), and N. similis (16,580 bp) had similar lengths, while N. leleupi had the shortest length at 16,562 bp (Fig. 1) (accession numbers: OP805601.1, OP805600.1, OQ076695.1, OP930818.1, OP930816.1, OP930817.1, and OP930815.1). The seven mitogenomes possessed the typical gene composition found in most bony fish, including 13 PCGs, 22 tRNAs, two rRNAs, and a CR. Among these genes, 12 PCGs, 14 tRNA genes, and two rRNA genes, were located on the major strand (H-strand), while the remaining eight tRNA genes and a PCG were encoded on the minor strand (L-strand). The gene order of these mitogenomes was identical to that of previously published species L. signatus (MZ427900.1) and Neolamprologus brichardi (AP006014.1) (Nam et al. 2021). Seventeen intergenic regions of the same length were observed between the mitochondrial regions of species L. kungweensis, L. meleagris, L. ornatipinnis, N. brevis, with lengths ranging from 10 bp (between atp8 and atp6) to 35 bp (between trnN and trnC). However, N. caudopunctatus exhibited a 24 bp intergenic region between trnV and rrnL, and N. similis displayed a 38 bp intergenic region at the same location. The trnC and trnY of N. similis overlapped by 1 bp, whereas there was no overlap in this region in the other six species. In addition, N. leleupi had one more intergenic region (24 bp between cox1 and trnS2) than other species (Table 1).

Figure 1. 

The gene maps of the seven newly sequenced mitogenomes. Different gene types are shown in different colors.

The nucleotide composition of the seven newly sequenced Lamprologus and Neolamprologus mitogenomes were biased toward A and T (Table 2). The AT-skews exhibited inconspicuously positive values, while all GC-skews were markedly negative. The analysis revealed a clear preference for the utilization of C, along with a minor inclination towards A, across the entire genome (Table 2).

To determine the nucleotide composition of Cichlidae, the A + T content, AT-skew, G + C content, and GC-skew of 103 complete mitogenomes (including 8 subfamilies Astronotinae, Cichlasomatinae, Cichlinae, Etroplinae, Geophaginae, Pseudocrenilabrinae, Ptychochrominae, and Retroculinae of the family Cichlidae) were calculated. The H-strand in the mitogenomes of 103 cichlid species showed a similar preference for A and T nucleotides. The 103 Cichlidae mitogenomes had a comparable nucleotide composition, A + T content (52.1 ~ 58.8%) were higher than the G + C content (41.1 ~ 47.8%) (Fig. 2). The GC-skew were negative (–0.351 ~ –0.221), indicating a higher occurrence of C than G except for Andinoacara rivulatus (–0.019), Pelvicachromis pulcher (–0.005), and Etroplus canarensis (–0.002). The AT-skew were inconspicuously positive (0.002 ~ 0.076), indicating a small difference in the content of A and T in the mitogenomes. This phenomenon is also observed in other published Teleostei genomes (Liu et al. 2020; Ruan et al. 2020; Xu et al. 2021a, 2021b; Wang et al. 2023). The A nucleotide composition is commonly used to indicate gene direction and replication orientation during transcription and replication (Wei et al. 2010a, 2010b).

Figure 2. 

A + T content vs AT-skew and G + C content vs GC-skew in the 103 mitogenomes of family Cichlidae. Values are calculated on H-strands for full-length mitogenomes.

In the seven newly sequenced mitogenomes, PCG nad6 was on the L-strand, while other PCGs were on the H-strand. The average A + T content of the PCGs ranged from 53.0% (N. leleupi and N. brevis) to 54.7% (L. meleagris). Six of them had the same 13 PCGs length of 11,466 bp, while the remaining species, N. leleupi, had a slightly shorter length of 11,421bp. The reason for this difference was that the cox1 gene in N. leleupi had a mutation causing a premature stop codon compared to other species, resulting in a reduction of 45 base pairs in length (Tables 1, 2).

Most of the PCGs in the seven newly sequenced mitogenomes began with the start codon ATG, except for cox1, which started with GTG. Most PCGs terminated with the codon TAA or incomplete codon (TA− / T−−), with the exception of nad1, which ended with TAG (Table 2). The cichlid species are relatively conservative in their use of start codons, and their preferences are generally consistent with those of the seven newly sequenced species with the only exception of the occurrence of a rare start codon ATC in the cox1 and nad3. All the Cichlids share the stop codons with TAA, TAG, AGA, and incomplete codons (TA− / T−−) (Fig. 3).

Figure 3. 

Start codon and stop codon usage for the mitochondrial genome protein-coding genes of 103 cichlid species.

RSCU was calculated to identify the predominant synonymous codon (Grantham et al. 1980). The comparative analysis based on RSCU of all PCG codons showed that the codon usage patterns of these seven species were similar (Fig. 4). Genes encoding Ile and Leu2 had high frequency, while those encoding Cys, Met, and Ser1 were infrequent.

Figure 4. 

The codon distribution and RSCU of the mitogenomes of the seven newly sequenced mitogenomes.

The selection pressure was analyzed by calculating the ratio of Ka/Ks across Lamprologus and Neolamprologus for each aligned PCG (Fig. 5) (Yang and Nielsen 2002). It was found that atp8 showed the largest Ka/Ks value among the 13 PCGs, which suggested more amino acid variety in the biomolecule. This suggests that the atp8 gene might have evolved faster than other PCGs due to slight selection pressure (Hassanin et al. 2005). The faster evolution of the atp8 gene could result in greater amino acid diversity, indicating its potential as an effective marker for population classification. The Ka/Ks values for all PCGs were lower than 1, suggesting that purifying selection was likely the main driver of mitochondrial PCG evolution (Hurst et al. 2002).

Figure 5. 

Ka/Ks values for the 13 PCGs. Pale pink box plots, five species of gnus Neolamprologus; orange box plots, four species of Lamprologus; blue box plots, nine species of Lamprologus and Neolamprologus. The band inside the box represents the median; upper and lower hinges correspond to the 25^th and 75^th percentiles; circles, to outliers.

Besides the Ka/Ks analysis, an assessment of the degree of divergence in Lamprologus and Neolamprologus was conducted by analyzing the overall p-distance between nucleotides of 13 PCGs + two rRNA genes (Fig. 6). The results of p-distance indicated that the fastest-evolving gene was nad6, which was inconsistent with the results of Ka/Ks value. However, the difference in this gene might be not comparable with the selection since this force is acting in a contemporary period.

Figure 6. 

Genetic p-distances for nucleotide sequences for 13 PCGs and 2 rRNAs. Pale pink box plots, five species of Neolamprologus; orange box plots, four species of Lamprologus; blue box plots, 9 species of genera Lamprologus and Neolamprologus.

The size of the rrnS genes were between 943 bp (L. meleagris, N. brevis, and N. caudopunctatus) and 946 bp (N. similis), while the size of the rrnL genes in seven species ranged between 1,652 bp (N. similis) to 1,671 bp (N. brevis) (Table 1). The two rRNA genes located between trnF and trnL2, with trnV separating them. The A + T content of rRNAs ranged from 53.0% ~ 54.1% (Table 2).

The sizes of the tRNA genes ranged from 66 bp (trnY of N. caudopunctatus) to 74 bp (trnK). The combined length of the 22 tRNA genes varied between 1,552 bp (N. caudopunctatus) and 1,554 bp (L. kungweensis, L. ornatipinnis, and N. similis). The A + T contents of tRNA genes ranged from 54.7% to 55.8% among the seven species analyzed in this study (Table 2).

As with other fish mitogenomes, the CRs were discovered to exist between trnF and trnP in all seven species. The sizes of the CRs ranged from 884 bp (N. similis) to 891 bp (L. kungweensis). The A + T contents of PCGs, tRNAs, and rRNAs sequences were found to be similar to that of the entire mitogenomes, whereas CR sequences had a higher A + T content (62.5% ~ 63.8%) (Table 2).

To elucidate the phylogenetic inter-relationships within the family Cichlidae and genera Lamprologus and Neolamprologus, concatenated nucleotide sequences of 13 PCGs + two rRNAs from 103 cichlid species were obtained. Additionally, Channa andrao, and Hyphessobrycon sweglesi from two other families were used as outgroups. It was found that BI and ML analysis generated the same topology structure on most nodes (Fig. 7).

Figure 7. 

13 PCGs-based phylogenetic tree of 103 cichlid species and two outgroups. Numbers at nodes represent the posterior probability and bootstrap values for BI and ML analysis, respectively. “-” indicates this clade not supported by BI or ML analysis.

Specifically, the seven complete mitogenomes covered two genera in this study have good clustering in phylogenetic trees, and within the family Cichlidae, the subfamily Etroplinae and Ptychochrominae were monophyletic across analyses. They diverged with species in other subfamilies early in the evolutionary history of cichlid fishes. This result was similar to a previous molecular phylogenetic study (Astudillo-Clavijo et al. 2022). Thirty-one species from subfamily Astronotinae, Cichlasomatinae, Cichlinae, Geophaginae, and Retroculinae were clustered into one branch, indicating these five subfamilies were closely related. Moreover, 67 Pseudocrenilabrinae species also formed a monophyletic clade. Pseudocrenilabrinae tribes and their interrelationships were for the most part well supported as reported by Astudillo-Clavijo et al. (2022). Due to the addition of seven newly sequenced mitogenomes, three pairs of sisters (N. brichardi + N. leleupi, N. caudopunctatus + L. meleagris, and L. signatus + L. kungweensis) were newly identified, as shown in Fig. 7. Lamprologus meleagris did not cluster with Lamprologus species, but with species from the genus Neolamprologus. Previous studies have identified such taxonomic issues in the genera Lamprologus and Neolamprologus (Schelly et al. 2006; Sturmbauer et al. 2010). Sturmbauer et al. (2010) think a viable way might be to re-assign the genus name Lamprologus to most Neolamprologus species. Our results also support this scenario. However, the species from the genera Lamprologus and Neolamprologus used in this study were limited, making it impossible to perform a more detailed analysis. Therefore, to better understand the relationships between members of these two genera, it will be beneficial to include more species in future studies.

In conclusion, our study increased the database of mitogenome in Cichlidae, and showed that mitogenome sequences are efficient molecular markers for studying the phylogenetic relationships within Cichlidae. However, there is a lack of analyses in nuclear genes. In the future study, we will further improve these deficiencies.

The authors have declared that no competing interests exist.

No ethical statement was reported.

This research was funded by the Innovation and Entrepreneurship Training Program for College Students of China (202210298070Z), the biodiversity investigation, observation and assessment program (2019–2023) of Ministry of Ecology and Environment of China (2110404), and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).

Conceptualization: WZ, HL. Data curation: JW. Formal analysis: JT, JW. Funding acquisition: WZ, JW, HL. Methodology: HL. Project administration: HL, WZ. Resources: WZ, HL, KH. Software: JT, JW. Supervision: HL, KH. Validation: HL. Visualization: JW, JT. Writing - original draft: JW.

Jiachen Wang https://orcid.org/0000-0002-0437-7687

Jingzhe Tai https://orcid.org/0009-0001-4507-9280

Wenwen Zhang https://orcid.org/0000-0003-0142-9469

Ke He https://orcid.org/0000-0001-6446-9439

Hong Lan https://orcid.org/0009-0004-2188-3604

Hongyi Liu https://orcid.org/0000-0003-2081-5779

All of the data that support the findings of this study are available in the main text or Supplementary Information.