The sample used in this study was collected from the Cailaohe River, Fengshan, Guangxi, China (24.61°N, 106.97°E). Total genomic DNA was extracted from fin-tissue samples using a DNA isolation kit (Qiagen) with a final elution volume of 50 µl. The quality and purity of the isolated DNA were assessed prior to downstream applications. Agarose gel electrophoresis was used to analyze DNA integrity and assess the presence of contaminants. DNA purity was evaluated using a NanoDrop One spectrophotometer (Thermo Fisher Scientific, USA). Final DNA concentrations were accurately determined using a Qubit 3.0 Fluorometer (Thermo Fisher Scientific, USA).

The collected DNA sample was used for paired-end (PE) library construction using standard protocols of the NEBNext Ultra II DNA Library Prep Kit for Illumina (NEB, USA) in accordance with the manufacturer’s instructions. It was sequenced using the Illumina NovaSeq 6000 platform (Illumina, USA) with a 350-bp insert size. Adaptor and low-quality reads were filtered using fastp (Chen 2023), resulting in a total of 69.21 Mb of clean reads (150 bp). The mitogenome was de novo assembled using MitoZ (Meng et al. 2019). The assembled mitogenome was annotated using the online tool MITOS using the default parameters (Bernt et al. 2013). The protein-coding sequences were checked and confirmed using Geneious R10 (Kearse et al. 2012). Start/stop codons, codon usages, relative synonymous codon usage (RSCU), and nucleotide composition were analyzed using MEGA v. 7 (Kumar et al. 2016) and PhyloSuite (Xiang et al. 2023). Skew compositions were calculated using: AT-skew = (A − T) / (A + T) and GC-skew = (G − C) / (G + C) (Perna and Kocher 1995). TRNAscan-SE v. 2.0 (Lowe and Chan 2016) was used to predict the secondary structures and anticodons of transfer RNAs (tRNAs). The online mitochondrial visualization tool OGDRAW (Greiner et al. 2019) was used to draw a graphical map of the complete mitogenome. The newly generated complete mitogenome sequence and its annotation were submitted to GenBank using BankIt (accession number OR492308).

To determine the phylogenetic position of G. cailaoensis, 92 complete mitogenomes and 68 Rag1 sequences of Labeoninae were downloaded from GenBank, and one species of Torinae, two species of Xenocypridinae, and three species of Opsariichthyinae were used as the outgroups (Mayden et al. 2009). Three datasets were constructed for analyses: (i) protein-coding gene (PCG) sequences of the complete mitogenome (dataset 1); (ii) PCG sequences of the complete mitogenome combined with nuclear DNA (ncDNA) (Rag1) sequences (dataset 2); and (iii) ncDNA (Rag1) sequences (dataset 3). All sequences were first aligned using MAFFT v. 7.475 (Katoh and Standley 2013), then trimmed using trimAl (Salvador et al. 2009). Maximum-likelihood (ML) and Bayesian-inference (BI) approaches were used to construct phylogenetic trees based on the three datasets. The ML analysis was performed using IQ-TREE v. 2.1.4 (Minh et al. 2020) based on the best-substitution model selected by ModelFinder in the IQ-TREE package (Kalyaanamoorthy et al. 2017). Nodal support was assessed based on 1,000 bootstrap replicates (Felsenstein 1985). The BI analysis was performed using MrBayes v. 3.2.7 (Ronquist et al. 2012), with the best-fit nucleotide substitution model also determined using ModelFinder. Four chains (three hot, one cold) were run for 5 million generations, with tree sampling every 1,000 generations and the first 25% of samples discarded as burn-in. Convergence was confirmed by ascertaining that the average standard deviation of split frequencies was below 0.01. The phylogenetic trees were viewed and edited using FigTree v. 1.4.4 (Rambaut 2014).

The mitogenome of Guigarra cailaoensis was identified as a circular double-stranded DNA sequence of 16,593 base pairs (bp) in length and included 13 protein-coding genes, 22 tRNA genes, two ribosomal RNA (rRNA) genes, and a putative control region (Table 1, Fig. 1). The base composition of G. cailaoensis was A = 32.2%, G = 15.5%, T = 26.4%, and C = 26.0%, with higher AT content (58.6%) than GC content (41.4%) (Table 2).

Figure 1. 

Circular map of complete mitogenome of Guigarra cailaoensis.

The PCGs had a total length of 11,412 bp, accounting for 68.78% of the total length of the complete mitogenome. The ND5 coding DNA sequence (CDS) had the highest number of base pairs (1 824 bp), while ATPase8 had the lowest (165 bp). The base percentage composition revealed a lower G + C content (41.1%) compared to the A + T content (58.9%). All PCGs were encoded on the heavy (H) strand, except for the ND6 gene, which was encoded on the light (L) strand. All PCGs were initiated with the methionine codon ATG, except for COX1, which was initiated with GTG, consistent with previous labeonine mitochondrial DNAs (Wang et al. 2019). Two types of stop codon were identified: TAA (ATP6, COX1, COX3, ND1, ND4L, ND5, and ND6) and TAG (ATP8, ND2, and ND3). Incomplete stop codons were detected for COX2, CYTB, and ND4 (Table 1).

The RSCU results indicated that six codons, CUA (2.35%), CGA (2.35%), CCA (2.33%), GGA (2.20%), UCA (2.19%), and GUA (2.16%), were the most frequently used. Additionally, the amino acids Pro, Thr, Leu1, Arg, Ala, Ser2, Val, and Gly were encoded by four codons, while all the other amino acids were encoded by two codons (Fig. 2).

Figure 2. 

Relative synonymous codon usage (RSCU) in mitogenomes of Guigarra cailaoensis.

The 12S rRNA and 16S rRNA were 952 and 1,641 bp in length, respectively. They were located between trnF and trnL1, separated by trnV. The nucleotide composition of the rRNAs was A = 35.4%, C = 23.8%, G = 20.6%, and T = 20.2%. Thus, G. cailaoensis displayed a higher percentage of AT (55.6%) than GC (44.4%) (Table 2).

Twenty-two tRNA genes were identified in G. cailaoensis mitogenome, including two for trnL and trnS, and one for each of the other amino acids (Table 1). Of these, 21 tRNA genes exhibited the typical cloverleaf secondary structure with four domains, while the trnS1 gene lacked the D domain (D-stem and D-loop) (Fig. 3).

Figure 3. 

Secondary structures of 22 tRNA genes in Guigarra cailaoensis.

Fourteen tRNAs were encoded on the H-strand, while the remaining tRNAs were encoded on the L-strand (trnQ, trnA, trnN, trnC, trnY, trnS1, trnE, and trnP; Table 1). The length of these tRNAs varied, ranging from 66 bp (trnC) to 76 bp (trnL1 and trnK), with a total length of 1,562 bp and accounting for 9.41% of the total mitogenome. Nucleotide composition of the tRNAs was A = 29.0%, C = 20.7%, G = 23.1%, and T = 27.2%, showing a higher AT content (56.2%) than GC content (43.8%) (Table 2).

The non-coding control region in the mitogenome, identified as the D-loop, was located between the trnP and trnF genes (Fig. 1). Spanning 918 bp in length, the region accounted for 5.53% of the whole mitogenome. The region exhibited a higher AT content (67.9%) than GC content (32.1%), with a nucleotide composition of A = 33.9%, T = 34.0%, C = 18.7%, and G = 13.4% (Table 2).

The best-fit models for ML and BI analyses were identified, as shown in Table 3. Phylogenetic trees derived from dataset 1 and 2 were remarkably similar, while those from dataset 3 showed slight variations. Within each dataset, the trees generated from ML and BI analyses were consistent across all taxa, differing only slightly in their support values. Consequently, the ML trees were presented here together with the nodal support values generated by ML and BI analyses, respectively. Notably, all taxa within Labeoninae were recovered as a monophyletic clade, further subdivided into four major lineages. In dataset 1, G. cailaoensis formed a sister taxon to the lineage consisting of Paraqianlabeo lineatus and Pseudogyrinocheilus prochilus, and in dataset 2 formed a sister taxon to Pseudogyrinocheilus prochilus. However, in dataset 3, G. cailaoensis formed a sister taxon to the lineage consisting of Discogobio and Discocheilus (Figs 4–6).

Figure 4. 

Phylogenetic tree of Guigarra cailaoensis and 98 species downloaded from GenBank based on PCG sequences of complete mitogenomes (dataset 1). Nodal numbers are ML bootstrap values and BI posterior probability values, respectively. Only values above 50% are given.

Figure 5. 

Phylogenetic tree of Guigarra cailaoensis and 72 species downloaded from GenBank based on PCG sequences of complete mitogenome combined with ncDNA (Rag1) sequences (dataset 2). Nodal numbers are ML bootstrap values and BI posterior probability values, respectively. Only values above 50% are given.

Figure 6. 

Phylogenetic tree of Guigarra cailaoensis and 72 species downloaded from GenBank based on ncDNA (Rag1) sequences (dataset 3). Nodal numbers are ML bootstrap values and BI posterior probability values, respectively. Only values above 50% are given.

The authors have declared that no competing interests exist.

This study does not require ethical approval because no procedures were performed on live animals, and the tissue samples used were collected from the dead specimens.

This work was supported by the National Natural Science Foundation of China (31960103) and Yunnan Provincial Basic Research Special Project (202301AT070254).

Conceptualization: LPZ. Data curation: LPZ. Funding acquisition: LPZ. Methodology: LPZ, YMG. Project administration: LPZ. Software: LPZ, YMG. Supervision: LPZ. Visualization: LPZ. Writing – original draft: LPZ. Writing – review and editing: LPZ, YMG.

Lan-Ping Zheng  https://orcid.org/0000-0002-9855-6503

All of the data that support the findings of this study are available in the main text. The genome sequence data are openly available in GenBank of NCBI at (https://www.ncbi.nlm.nih.gov/) under the accession no. OR492308.