Introduction

Nesolagus is a rabbit genus that comprises three distinct species: the Annamite striped rabbit (Nesolagus timminsi), the Sumatran rabbit (Nesolagus netscheri), and the extinct N. sinensis1. The N. netscheri is a lagomorph species found endemically in Sumatra, Indonesia2,3. Because insufficient information is available to estimate population size, range, and density, the International Union for Conservation of Nature (IUCN) has classified this species as Data Deficient4. The N. netscheri has been a legally protected species in Indonesia since then, and it continues to be protected under the recently updated Indonesian law. It has been classified as the rarest lagomorph owing to the small number of museum specimens and the rarity of historical observations5.

The relationship between N. netscheri and N. timminsi adds complexity to the genus’s evolutionary context because geographically they are separated and both have significant threats from habitat loss and fragmentation3,6. Despite its historical recognition and occasional sightings, there remain substantial gaps in understanding the evolutionary relationships of N. netscheri within the Nesolagus genus and the broader family Leporidae, indicating a need for molecular phylogenetic studies to elucidate the evolutionary history and genetic diversity of N. netscheri. A comprehensive understanding of these aspects is crucial, as it would provide insights into the evolutionary processes that have shaped the species and inform conservation strategies.

Mitochondrial (mtDNA) genomes provide a wealth of genetic markers that are crucial for conservation efforts, particularly for species like the Sumatran striped rabbit (Nesolagus netscheri). The complete sequencing of mtDNA can uncover specific genetic markers, such as cytochrome c oxidase subunit I (COI) and cytochrome b, which is widely recognized as a universal DNA barcode for animal species identification7,8,9. This marker is particularly valuable for assessing genetic diversity and population structure, allowing conservationists to monitor the health of populations and identify distinct genetic lineages that may require targeted conservation efforts10,11,12,13,14. For instance, in the case of the Sumatran striped rabbit, identifying unique mtDNA haplotypes could inform habitat management and restoration strategies, ensuring that genetic diversity is preserved in fragmented landscapes. Additionally, mtDNA can reveal information about historical population dynamics and gene flow, critical for understanding how habitat fragmentation affects species like the Sumatran striped rabbit. By analyzing variations in mtDNA, researchers can identify genetic bottlenecks and assess the impacts of habitat loss on gene flow between populations15,16. This information is particularly relevant in the context of Sumatra’s rapidly changing landscape, where deforestation and agricultural expansion have led to isolated populations that may be at risk of inbreeding17,18,19,20.

The use of mtDNA markers can thus help prioritize conservation actions, such as creating wildlife corridors to enhance connectivity between fragmented habitats21. Moreover, the high mutation rate and maternal inheritance of mtDNA make it an ideal candidate for studying evolutionary relationships and phylogenetic analyses within Leporidae9,10,22,23. For example, the identification of specific mitochondrial markers can aid in distinguishing between closely related species or subspecies, providing insights into their evolutionary history and informing conservation strategies that consider the unique genetic makeup of each population9.

This research is the first successful characterization of the mitochondrial genome of N. netscheri, which has not been previously published or available in any database. This study aims to characterize the mitogenome of N. netscheri and reconstruct the evolutionary tree in the family Leporidae using a novel mitochondrial genome sequence. This study provides valuable information, including molecular markers, which can be utilized to address conservation needs, considering the limited extent of studies on rabbits. These findings will provide valuable information for phylogenetic, evolutionary, and population genetic studies. Additionally, findings can potentially be important to conserving the rarest rabbit, N. netscheri.

Results and discussion

Mitogenome composition and organization

The complete assembly of N. netscheri’s circular mitochondrial genome was generated by the Oxford Nanopore Technolgy (ONT) long reads. The length of this final assembly was 16,709 bp, with a coverage of 5021×. It is 3 bp smaller than the other N. timminsi mitogenome (NC_063946.1). The complete mitochondrial genome sequence of Sumatran striped rabbit Nesolagus netscheri has been deposited in GenBank under the accession number PQ047138. This genome size is within the range of available Leporidae mitochondrial genome (Table 1.). As in other mammals24, the mitogenome of N. netscheri encodes 37 classical mitochondrial genes (13 protein-coding genes [PCGs], 22 tRNAs, 2 rRNAs, and 1 D-loop) (Fig. 1). The complete mitogenome had a significant A + T bias in its nucleotide composition (30.05% A, 31.8% T, 13.3% C, and 24.4% G). In this mitogenome, the H strand transcribed twelve PCGs, two rRNAs, and thirteen tRNAs, while the L strand transcribed the remaining ten genes (Table 2). This transcription pattern is the same as that of the genus N. timminsi. N. timminsi and N. netscheri belong to the same genus, and their gene orders are similar (Supplementary File 1).

Fig. 1
figure 1

(a) Circular representation of the mitochondrial genome of Nesolagus netscheri. The colored blocks indicate different gene types, with genes encoded on the light and heavy strands in counterclockwise and clockwise orientations, respectively. The black inner circle represents the GC content of the mitogenome, while the pink and red regions indicate GC skew. The rabbit illustration was adapted from https://scentsindonesia.com; (b) Geographic distribution of N. netscheri (orange regions) within its habitat. The map was created using QGIS version 3.40.1 Bratislava (https://www.qgis.org).

Full size image
Table 1 List of Leporidae mitochondrial genomes chosen to reconstruct the phylogenetic tree.
Full size table
Table 2 Annotation of the N. netscheri mitochondrial genome.
Full size table

Protein-coding genes

Standard initiation codons (ATN) were used for the initiation of all PCGs, and conventional stop codons (TAG or TAA) were used for the termination of 9 out of 13 PCGs. The genes ND3, ND4, and COX3 all terminated with incomplete stop codons (either T or TA; Table 2), while the CYTB gene had an alternate putative stop codon (AGG). The start codons found in the PCGs are consistent with the standard vertebrate mitochondrial start codons, except for ATT. There have been comparable reports of this terminal codon among various Leporidae28,31,33. The same phenomenon has been observed with the regular and common stop codons TAG and TAA in mammals24. AGG is an uncommon stop codon; however, it has been identified in other Leporidae species28,31,33 and other vertebrates in previous study34. The COX3, NAD3, and NAD4 genes in vertebrate mitochondrial genomes, including Leporidae28,31,33, frequently exhibit reduced stop codons. These reduced stop codons have been reported to be addressed through post-transcriptional polyadenylation35.

Relative synonymous codon usage values for N. netscheri are presented in Fig. 2. The PCGs contained 3,510 codons. The most prevalent codons in these mitogenomes encode Leu, Ile, Gly, and Thr. In contrast, the least frequently observed were those that encoded Trp. (Fig. 2).

Fig. 2
figure 2

Relative synonymous codon usage (left) and codon distribution (right) in the mitochondrial protein-coding genes of N. netscheri.

Full size image

Transfer and ribosomal RNA genes

The mitochondrial genome of N. netscheri had a total of 22 transfer RNA (tRNA) genes. The length of the tRNA genes varied from 54 bp (trnG) to 72 bp (trnN). A typical ‘cloverleaf’ secondary structure was inferred in all tRNA genes, with the exception of trnS1 (see Fig. 3). This is a commonly observed characteristic in genes for tRNA in mammals36,37 and vertebrates38. The lack of an arm in trnS1 may potentially have a functional purpose in the process of structural compensation with other structures39. The 16 S RNA (rrnL) is positioned between tnrL2 and trnV and has a length of 1,580 bp, whereas the 12 S RNA (rrnS) is placed between trnV and trnF and has a length of 950 bp. The positions of the control region and rRNA genes (rrnS and rrnL) in the N. netscheri mitochondrial genome are similar to those already identified for other leporid mitogenomes28,31,33.

Fig. 3
figure 3

Secondary structure of tRNAs in the mitochondrial genome of N. netscheri.

Full size image

The Ka/Ks ratio is an important parameter for quantifying selection pressure40. To assess the effect on mitochondrial PCGs, we investigated the Ka/Ks ratios for N. netscheri and two leporid species, Lepus europaeus and Nesolagus timminsi, which were used as reference species. The 13 PCGs showed varying Ka/Ks ratios, which suggests that the genes are exposed to a range of functional constraints (Fig. 4). The data indicates that all the PCGs are undergoing evolutionary changes due to purifying selection (< 1), except for atp8. The lowest calculated value of Ka/Ks, which indicates selected pressure, was found for the nd4l gene in the Genus Nesolagus (Ka/Ks = 0.025). This suggests that nd4l faces the highest level of selective pressure and evolves at the slowest rate among the genes analysed. Purifying selection in mammalian mitochondrial DNA (mtDNA) is evidenced by studies that demonstrate a consistent pattern of selective pressure against deleterious mutations. For example, Stewart et al. highlighted that strong purifying selection operates during the transmission of mtDNA across generations, effectively maintaining the integrity of mitochondrial genomes by favoring the survival of less mutated variants41. Similarly, Tsai et al. reported that reducing the dosage of mitochondrial DNA polymerase enhances the elimination of defective mitochondrial genomes, indicating that purifying selection plays a critical role in maintaining mtDNA quality during oogenesis42. Furthermore, Ennis et al. found that the complete mitochondrial genome of Baird’s tapir exhibited patterns of purifying selection in its protein-coding genes, reinforcing the notion that mtDNA is subject to evolutionary constraints across various mammalian species43. These studies underscore the importance of purifying selection in preserving the functional integrity of mitochondrial genomes in mammals.

The atp8 gene in the genus Nesolagus exhibits a Ka/Ks ratio of 1.22, indicating positive selection and a rapid rate of evolution. This suggests its role in adaptation to ecological challenges. The atp8 gene exhibits the same pattern of high Ka/Ks ratios in various mammals, including Rattus44, Chiroptera45, caviomorph rodents46, as well as between Puma concolor47. It has been reported in other studies that positive selection also affects mitochondrial genes, particularly in animals that are adapted to harsh environments23,48,49,50. In extreme environments, positive selection in ATP synthase genes enhances metabolic efficiency and energy production, as seen in species inhabiting high-altitude or low-oxygen conditions51,52. For tropical mammals, atp8 has been identified as critical for adaptation to the metabolic demands of dense rainforest ecosystems, which are characterized by high biodiversity and fluctuating conditions53,54. Studies on bats, cetaceans, and tropical chickens highlight the evolutionary significance of atp8 in overcoming environmental stressors specific to their habitats, such as low oxygen, high energy demands, or thermal adaptation55,56. In the context of Nesolagus, which inhabits the tropical forests of Sumatra, the positive selection of atp8 may enhance its capacity to adapt to the unique challenges of its environment, including dense vegetation, resource competition, and fluctuating climatic conditions46,52,57,58,59,60. These findings collectively underscore the ecological and evolutionary importance of the atp8 gene in facilitating adaptation to harsh environments.

Fig. 4
figure 4

Selective pressure analysis of the protein-coding genes (PCGs) of N. netscheri. By employing two distinct leporid mitogenomes as references, the Ka/Ks ratios were determined for each of the 13 PCGs.

Full size image

Phylogenetic of N. netscheri

Mitochondrial DNA is a commonly employed genetic marker for understanding molecular systematic in animals61. To better comprehend the evolutionary relationships within the Leporidae family, we collected concatenated nucleotide sequences of 13 PCGs and two ribosomal RNA (rRNA) genes from 13 leporids. The topologies of the phylogenetic trees obtained from the Bayesian inference (BI) and maximum likelihood (ML) analyses were nearly identical, as illustrated in Fig. 5. Both trees generated from the two models were supported by robust statistics values.

The phylogenetic analysis confirmed the monophyly of the family Leporidae (posterior probability = 0.99; bootstrap = 100), with strong support for the Nesolagus genus (1; 100), the genus Lepus (1; 100), and the group comprising Oryctolagus cuniculus, Brachylagus idahoensis, and Sylvilagus sp. (1; 91), forming a single single and well-supported clade. In the Leporidae family, Nesolagus is positioned basally as a sister group to a clade consisting of Oryctolagus cuniculus, Brachylagus idahoensis, Sylvilagus sp., Romerolagus diazi, and Lepus spp. The topology in our study is consistent with previous research that examined cranial morphology in Leporidae62,63 and molecular data64,65. The data analysis shows a split between N. netscheri and N. timminsi during the Late Pleistocene (95%HPD, 0.43 mya ± 0.19). This study’s divergence time matches the study of Nesolagus sinensis fossils from the Early Pleistocene period1. These fossils have been reported to be more primitive and directly related to the currently existing Nesolagus. In further studies, it is necessary to include the mitogenome of additional genera such as Pentalagus, Bunolagus, Poelagus, Caprolagus, and Pronolagus into phylogenetic analysis to perform a more comprehensive investigation of Leporidae phylogeny.

Fig. 5
figure 5

The phylogenetic tree of Leporidae is based on 13 PCGs (11,396 bp), which is estimated using the Bayesian relaxed-molecular clock method. Posterior probability values from Bayesian inference and percentages of bootstrap from maximum likelihood analyses are listed on each node, respectively. Purple bars were used to show the 95% HPD for each node, with each genus represented by a different color.

Full size image

Conclusions

Mitochondrial genomes are increasingly utilized as valuable markers in phylogenetic, population genetic, and evolutionary studies. This paper reports and annotates the first complete mitochondrial genome of the elusive Sumatran striped rabbit (Nesolagus netscheri). There was 16,709 bp of circular double-stranded DNA. The mitochondrial genome’s organization and gene arrangement is similar to the one described for its sister taxon, N. timminsi. According to our Bayesian-based phylogenetic tree, the genus Nesolagus is at the basal of the Leporidae tree, and N. netscheri diverged in the Late Pleistocene (0.43 Mya). This research provides vital genetic information essential for developing effective conservation strategies for this rabbit. In the context of Sumatra’s unique and threatened forests, the mitochondrial genome data is critical for assessing genetic diversity, identifying evolutionary adaptations, and informing strategies such as inbreeding or bottleneck detection and translocation. This new information can be used as a reference mitogenome for future mitochondrial diversity studies to diagnose the species’ population status. It is a crucial point for adaptation to environmental change because of its limited habitat and distribution. Other Leporidae mitochondrial genomes are still required to provide a more thorough evolutionary history, and population genetic research and monitoring are needed to support effective conservation management—particularly in the case of Sumatran striped rabbits.

Materials and methods

Sample collection and DNA extraction

A tissue sample from the leg muscular of Nesolagus netscheri was taken from a specimen from Setiawan et al. research2. The tissue sample was collected immediately after the animal’s death to reduce DNA degradation caused by post-mortem damage. The sample was promptly frozen to ensure its preservation. Additionally, sterilized tools and gloves were used during sample handling to minimize the risk of exogenous contamination. Permission to access genetic resources for the Sumatran rabbit sample has been obtained from the Ministry of Environment and Forestry (no. SK.154/KSDAE/SET.3/KSA.2/8/2023). No live animal was involved in this research. This individual was reportedly caught in a forest within the Dempo mountain, South Sumatra province, Indonesia. The tissue sample was extracted using the Qiagen Blood and Tissue Kit. Then the DNA samples were quantified using a Qubit Fluorometer.

Mitogenome sequencing and assembly

Following Genome DNA Extraction, genome DNA was sequenced using PromethION Flow Cell (Oxford Nanopore Technologies) and Ligation Sequencing Kit V14 SQK-LSK114 to prepare the genomic DNA library. A total of 1000 ng in 48µL of gDNA sample was used to produce the library. The first step is the DNA repair and end-prep step using NEBNext FFPE DNA Repair Mix and Ultra II End-prep Enzyme Mix. In the next step, the sample is given adapter ligation using Ligation Adapter reagent, Ligation Buffer, and NEBNext Quick Ligation Module. The genomic DNA library was sequenced on a PromethIon 24 device with super accuracy (SUP) base-calling mode and running for approximately 72 h. A total number of 14,512,100 reads were generated and made available in FASTQ. All these reads were employed to conduct de novo assembly of the mitogenome of N. netscheri.

Mitogenome annotation and analysis

The Guppy base caller ONT v.3.2.4 was utilized to determine the base sequence from the raw data accurately. Only sequences with a Phred score greater than 13, indicating good quality, were chosen for the de novo mitogenome assembly using Oxford Nanopore Technology66,67. The assembly was performed utilizing the Flye program v.2.568 in the mitogenome assembly mode and functioned as baits to identify all potential mitochondrial sequences using Minimap269. A subsequent set of ONT assemblies was generated and subsequently refined over four iterations applying the same dataset using racon70 and medaka (https://github.com/nanoporetech/medaka).

The mitogenome sequences that were assembled were annotated using MITOS271. The tRNA scan-SE 1.21 software was employed to detect tRNA genes, using the genetic code specific to vertebrate mitochondria72. A visual illustration of the mitogenome of N. netscheri was drawn using the Proksee73, an online tool for visualizing mitochondrial data. The Ka_Ks calculator was utilized to quantify the ratio of nonsynonymous to synonymous substitution rates (Ka/Ks) for all 13 protein genes in three leporids, using MEGA X74. Bayesian and Maximum Likelihood phylogenetic trees were constructed using the 13 concatenated protein-coding genes (PCGs) of Leporids mitogenomes, as shown in Table 1. The best evolutionary model for analysis has been identified using jModeltest 275 to be the GTR model of sequence evolution, which includes a discrete gamma distribution and a proportion of invariable sites to account for variation in rates among sites (GTR+ Г + I). The Maximum-Likelihood (ML) phylogenetic was inferred using IQ-TREE (version 2.1) via the IQ-TREE web server with76 model selection and ultrafast bootstrap approximation (1000 iterations) based on the alignment of 13 PCGs. Beast 2.7.6 77 was used to estimate divergence times based on PCGs. The time tree was constructed using two calibrated points: the divergence of Lepus americanus and Lepus europaeus, which occurred 8.6 mya78, and the divergence of L. americanus and Lepus tolai, which was estimated based on fossil evidence to have occurred 0.78 mya79. The divergence time analysis employed an uncorrelated lognormal relaxed molecular clock and Yule process, with a total of 50 million generations. The convergence of the analysis was corroborated using Tracer80. All parameters show an ESS greater than 200, confirming sufficient sampling and reliability. The software TreeAnnotator was used to discard the initial 25% of trees and calculate the nodes’ ages along with their 95% credible intervals. The produced tree was viewed using FigTree Version 1.4.4.