Introduction

Invertebrate mitochondrial genome (mitogenome) is typically double-stranded and closed circular molecules, approximately 15–18 kb in length1,2. Invertebrate mitogenome consists of 13 protein-coding genes (PCGs), 22 transfer RNAs (tRNAs), 2 ribosomal RNAs (rRNAs), and one non-coding control region (CR)2,3. The mitogenome has characteristics such as small size, simple structure and fast evolution, it has been extensively studied and widely used for species identification and molecular phylogeny researches4,5.

Diplopoda (millipedes) is one of the most diverse groups of arthropods, with more than 7000 species described6. Millipedes Spirobolus grahami belongs to the Spirobolidae family of the Diplopoda class7. Millipedes are an important part of modern terrestrial ecosystems and play an important role in the decomposition of organic matter6,8,9. However, few studies have documented the phylogeny, evolution, behavior, physiology, and ecology of Millipedes8,10,11. Therefore, the use of mitogenome might be expected to provide valuable data on their phylogenetic relationship.

In order to further investigate the relationship between the Diplopoda, in this study, we firstly sequenced and characterized the mitogenome of S. grahami. The structural organization, nucleotide composition, codon usage, and AT/GC-skew were analyzed. Additionally, we conducted phylogenetic analyses based on 13 PCGs available elsewhere for the purpose of investigating the phylogenetic position of S. grahami within Diplopoda, which we believe might be helpful for further evolutionary and phylogenetic studies on millipedes within the Diplopoda.

Materials and methods

Sample collection and DNA extraction

Sample used in this study collected from Guilin Seven Star Park (Guilin, China). The collected sample was morphologically characterized based on the images and morphological features on GBIF (https://www.gbif.org/) and MilliBase (https://millibase.org/), with specific reference to Keeton12. The collection of the specimen was reviewed and approved by Nanjing Forestry University. Specimen for this study was collected in accordance with Chinese laws. Sample was stored at the Zoology Laboratory of Nanjing Forestry University. Total DNA was extracted from muscular tissue using a FastPure Cell/Tissue DNA Isolation Mini Kit (Vazyme™, Nanjing, China). The remaining tissue was stored at − 20 °C in 90% ethanol to preserve the specimens.

Next-generation sequencing

Library construction and sequencing were carried out by Novogene (Nanjing, China) on the HiSeq 2500 platform (Illumina Inc., San Diego, USA) following the manufacture’s protocol for 150-bp paired-end reads. Clean reads were used to assemble the full mitogenome in Geneious Prime 2020 using Spirobolus bungii (NC056899.1) as the template, and both ends of the final assembly were manually examined for overlap to build a circular mitogenome.

Annotation and sequence analysis

The BLAST CD-search (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and MITOS Webserver (http://mitos.bioinf.uni-leipzig.de/index.py) were used to detect PCGs, tRNAs, rRNAs, and CR13,14,15. The gene map of the mitogenome was generated with the CG view Server (http://cgview.ca/)16. Nucleotide compositional differences between genes were calculated according to the following formulae: AT-skew = (A − T)/(A + T) and GC-skew = (G − C)/(G + C)17,18. Relative synonymous codon usage (RSCU) was calculated in MEGA X and image rendered by PhyloSuite v1.2.119,20. The synonymous replacement rate (dS), non-synonymous replacement rate (dN), and the ratio of non-synonymous replacement rate to synonymous replacement rate (dN/dS) were determined using MEGA X for Spirobolus species19.

Phylogenetic analysis

We constructed a concatenated set of base sequences of the 13 PCGs from 24 species to study the phylogenetic relationship in Diplopoda (Table 1). Lithobius forficatus was used as an outgroup. Phylogenetic analyses were conducted for each dataset using Bayesian inference (BI) and maximum likelihood (ML) methods. All operations were performed in PhyloSuite v1.2.120. MAFFT was used to perform multiple sequence alignment, with strategy of L-INS-i. ModelFinder was used to select the best-fit model. The best fit models of BI were GTR + F + I + G4 for COX1, COX2, COX3, ND1, ND2, ND4, ND4L, ND5, ND6, HKY + F + I + G4 for ATP6, ND3, Cytb, and HKY + F + G4 for ATP8. The best fit models of ML were GTR + F + R4 for COX1, COX2; GTR + F + R5 for ND1, ND4, ND4L, ND5, HKY + F + I + G4 for ATP6, HKY + F + G4 for ATP8, TIM2 + F + I + G4 for COX3, TPM3u + F + I + G4: Cytb, ND3, TPM3u + F + I + I + R3 for ND2, ND6. BI tree was performed with MrBayes 3.2.6 and run for 1,000,000 generations, with a burn-in of 25% trees, while ML tree was performed in the IQ-TREE21,22. The phylogenetic trees were viewed and edited using iTOL (https://itol.embl.de/)23.

Table 1 The mitogenomes used in phylogenetic analyses.
Full size table

Results and discussion

Mitochondrial genome organization

The total mitogenome of S. grahami was typical circular, double-stranded molecules, with 14,875 bp in length (Fig. 1). The mitogenome has been submitted to GenBank (Table 1). Mitogenomes of S. grahami encoded all 37 classical mitochondrial genes (13 PCGs, 22 tRNAs, and 2 rRNAs) and one CR. In this mitogenome, 15 genes (four PCGs, two rRNAs, and nine tRNAs) were transcribed from the majority strand (J strand), and the remaining 22 genes were transcribed from the minority strand (N strand) (Table 2), which is identical to S. bungii of the same genus11. The gene order of S. grahami was also consistent with that of S. bungii and Spirobolus walkeri in the same genus11. The gene order of millipede mitogenome is diverse24, but the gene order of this genus is relatively stable.

Figure 1
figure 1

Mitochondrial genome of S. grahami. Yellow blocks: CR; green blocks: rRNAs; light purple blocks: tRNAs; dark purple blocks: PCGs.

Full size image
Table 2 Annotation and organization of the complete mitogenome of S. grahami.
Full size table

Base composition analysis suggested that this mitogenome was biased toward A and T, the content ratio of A + T is 58.68% (Table 3), which is consistent with a previous study11. Besides, the PCGs, tRNAs, rRNAs, and CR were all biased in nucleotide composition (A + T > G + C), which is consistent with other invertebrate researches25,26. The AT-skew of S. grahami was negative, while the GC-skew was positive. The low GC-skew values of the analyzed mitogenome indicated the occurrence of more Cs than Gs. However, the AT-skew of tRNAs and CR were slight positive.

Table 3 Nucleotide composition and skewness of S. grahami mitogenome.
Full size table

Multiple overlaps between contiguous genes were calculated. There were five gene overlaps in this mitogenome, ranging from 3 to 7 bp. The longest overlap region of the mitogenome was found between Cytb and ND6, as well tRNA-Cys and tRNA-Trp, with 7 bp in length.

Protein-coding genes and codon usage

The total length of the PCGs was 10,988 bp, accounting for 73.87%. Four PCGs, ND1, ND4L, ND4, and ND5 were transcribed from the J-stand, and the other PCGs from the N-strand. The sizes of 13 PCGs ranged from 156 (ATP8) to 1702 bp (ND5) in the mitogenome. The start codon of all PCGs is ATN (ATG, ATT, and ATA), except COX1 starts with the CTA codon. This unusual start codon, CTA, have previously been reported in Spirobolus11. In addition, three stop codons were found in the PCGs of S. grahami, namely TAA, TAG, and T. In the mitogenome, the occurrence frequency of the stop codon T was higher than those of the other two stop codons, while the stop codon TAG occurred the least.

The RSCUs of the PCGs in the mitogenome were calculated, as shown in Fig. 2. The RSCUs analysis of S. grahami showed that codons tended to use more A or T at the third codon, which is consistent with some previous studies27,28. The dN/dS of the PCGs in the mitogenome of Spirobolus (S bungii, S. grahami, and S. walkeri) were calculated, too (Table 4). In evolutionary analysis, it is necessary to understand the rate at which dN and dS mutations occur, analyzing their ratios to detect selective pressures, if any, among PCGs. In this study, ND4L having the lowest evolutionary rate, and COX1 having the highest sequence variability. The faster evolution of COX1 of the genus Spirobolus might result in greater amino acid diversity, indicating its potential as an effective marker for classification. The dN/dS values for most PCGs were lower than 1, suggesting that purifying selection was likely the main driver of mitochondrial PCG evolution29.

Figure 2
figure 2

Relative synonymous codon usage of S. grahami, the stop codon is not included.

Full size image
Table 4 The dN/dS values among Spirobolus species.
Full size table

Transfer RNA, ribosomal RNA genes and control regions

22 tRNAs and two rRNAs were discontinuously distributed throughout the whole mitogenome. The tRNA genes of the mitogenome were 1376 bp, which account for 9.3% of the entire mitogenome. There were nine tRNAs from the J-strand and 14 transcribed from the N-strand. Among all secondary structures of the 22 tRNA genes from the S. grahami mitogenome, except for tRNA-Ser1, all had a typical cloverleaf structure (Fig. 3), as observed in other Diplopoda mitogenomes8,11,30. The 16S rRNA (1033 bp) was encoded between tRNA-Val and tRNA-Leu1, and the 12S rRNA was 757 bp long. The total size of the two rRNAs was 1790 bp, accounting for 12.03%.

Figure 3
figure 3

Secondary structure of 22 tRNA genes from the S. grahami mitogenome.

Full size image

One CR was found between the genes tRNA-Ile and 12S rRNA in the mitogenome, with 450 bp in length, accounting for 3.03%. The content of A + T is 71.78%, consistent with research that mitochondrial CR is typically characterized by high A + T content in most invertebrates25,31,32.

Phylogenetic analyses

We included 23 species of Diplopoda in the phylogenetic analyses and selected L. forficatus in Chilopoda as an outgroup to root the phylogenetic trees, using BI and ML methods. Phylogenetic trees were constructed based on sequences of 13 PCGs (Fig. 4). The topologies of the BI and ML trees were similar to each other. S. grahami is clustered together with S. bungii and S. walkeri, which belong to the same genus Spirobolus. Narceus annularus is closely related to genus Spirobolus, which is consistent with the result of previous study11. Glomeridesmus spelaeus is distantly related to the other Diplopoda species, similar to the previous study33. In addition, phylogenetic trees also support the classification of genus Tropostreptus. Previous study on millipede mitochondria have shown that genus Tropostreptus is phylogenetically more closely related to Archispirostreptus gigas and Macrolenostreptus orestes34. The results of the our study on the phylogenetic analysis of mitochondria also support this. We demonstrate that the mitogenome might be an effective tool for millipede classification. Our study shows that mitogenome sequences are effective molecular markers for studying the phylogenetic relationships and evolution within Diplopoda, but the data that covered only 22 species, meaning it’s still limited.

Figure 4
figure 4

Phylogenetic trees of 23 Diplopoda species and an outgroup (Lithobius forficatus) based on 13 PCGs using the BI (a) and ML (b) method.

Full size image

Conclusions

The mitogenome of S. grahami was determined to be 14,875 bp in length, with A + T content of 58.68%. The nucleotide composition showed that the mitogenomes of S. grahami exhibited negative AT and positive GC skews. The COX1 having the highest sequence variability. The dN/dS values for most PCGs were lower than 1, suggesting that purifying selection was likely the main driver of mitochondrial PCG evolution Both BI and ML trees support the classification of genus Spirobolus and Tropostreptus. Our results would contribute to the future resolution of phylogenetic relationships in Diplopoda.