A haplotype-resolved and chromosome-scale genome assembly of Hainan muntjac (Muntiacus nigripes)

Abstract

The Hainan muntjac (Muntiacus nigripes) is a wild animal endemic to Hainan, China. Its species distribution and the diversity of muntjac karyotypes have attracted much attention. Although genomic resources have increased in recent years, relevant genome assembly data of Hainan muntjac are still lacking. Meanwhile, molecular evidence for the taxonomic units of this species remains lacking. In this study, we successfully assembled the Hainan muntjac haplotype genome at the chromosome level using Pacbio long read and long sequencing technologies and Hi-C data. The final assembly size was 2.66 Gb, with allele and scaffold N50 values of 29.27 and 700.27 Mb, respectively, and we scaffolded the genome sequence onto three chromosomes. The genome contains a total of 21,451 genes and 10,056 gene families. Phylogenetic analysis using single-copy gene families revealed that Hainan muntjac is most closely related to red muntjac, with a divergence time of 8–11 Ma. This new genomic resource of Hainan muntjac will be crucial for future comparative genomic analyses and genetic evolutionary studies.

Background & Summary

The Hainan muntjac (Muntiacus nigripes), is a small deer species belonging to the genus Muntiacus^1,2. This genus is widely distributed across regions such as China, India, Thailand, and Malaysia. The Hainan muntjac is an endemic species of China’s Hainan Province and has garnered attention for its distinctive presence in a geographically unique location³. The Muntiacus species is frequently cited as an excellent model for studying vertebrate evolution due to the unique nature of its chromosome. The range of chromosome numbers varies from 2n = 6 in female Indian muntjac to 2n = 46 in Chinese muntjac(Muntiacus reevesi)^4,5.

The chromosome evolution of the Muntiacus genus has been extensively researched. Yang et al.⁶ and Frönicke et al.⁷ have provided essential evidence for the homology and occurrence of chromosomal fission-fusion events in Muntiacus species using molecular techniques such as chromosome painting, digital imaging and heterologous fluorescence in situ hybridization (FISH). Mudd et al.⁸ have substantiated the karyotypic differences and chromosomal evolution in musk deer species through genomic data. In addition, the Red Muntjac wide distribution indicates significant genetic diversity and differentiation among subspecies^9,10. At the same time, Hainan muntjac, as an endemic species in Hainan, the time of divergence from its close relatives is still unclear.

High-quality chromosome-level genomes are essential for studying chromosome evolution, species evolution, and population genetic diversity¹¹. A chromosomal-level genome of high quality can improve the detection of chromosomal evolutionary events. It can improve the detection of chromosomal evolutionary events and provide more accurate calculations of population genetic parameters, such as genetic diversity, gene flow, phylogenetic relationships, and genetic load within the genome. Therefore, we have generated the first high-quality chromosome-level haplotype genome of Hainan muntjac by combining Pacbio long-read data, DNBSEQ short-read data, and HiC sequencing data. However, the absence of a Y chromosome in the assembled genome may restrict its utility in investigating male-specific genetics and sex determination. Nevertheless, the female Hainan muntjac genome retains its intrinsic value for examining chromosomal evolution, particularly in light of its distinctive chromosomal characteristics, such as fusions. Overall, the genome enhances the genomic resources of genus Muntiacus, offering crucial support for future research on karyotypic evolution, ecological studies, and species conservation.

Methods

Samples collection and ethics statements

Blood samples were collected from a female Hainan muntjac in Hainan Province, China, for high-molecular-weight DNA extraction. The Hainan muntjac was rescued and anesthetized with an anticoagulant tube before the samples were immediately transferred to liquid nitrogen and stored in a −80 °C refrigerator. The study design, experiments, and sample collection were approved by the Review Board of Hainan University (HNUAUCC-2024-00214).

DNA/RNA extraction, libraries preparation and sequencing

The high molecular weight genomic DNA for PacBio HiFi sequencing was isolated using 3.5 ml blood sample and the cetyltrimethylammonium bromide (CATB) method, and purified with the Blood & Cell Culture DNA Midi Kit (QIAGEN). DNA quality was assessed using a 1% agarose gel and a NanoDrop One UV-Vis spectrophotometer (Thermo Fisher Scientific, USA), while DNA quantity was measured using a Qubit 4.0 (Invitrogen, USA). The PacBio HiFi library was constructed at the Genome Center of NextOmics Bioscience Co. (Wuhan, China). The genome center of Ltd. in Wuhan, China followed the standard protocol of PacBio (Pacific Biosciences, USA) using SMRTbell 15 kb preparation solution to prepare the libraries. The libraries were then sequenced on a PacBio Sequel II sequencer using Sequencing Primer V2 and Sequel II Binding Kit 2.0. Total RNA was extracted from a 2 ml blood sample using the TRNzol Universal Kit (TIANGEN). RNA concentration and purity were assessed using NanoDrop One (Thermo Scientific, USA) and Qubit 4 Fluorometer-1 (Thermo Scientific, USA), while RNA integrity was evaluated with the Agilent 2100 Bioanalyzer. For short-threaded RNA sequencing, libraries were prepared using the TruSeq mRNA Library Kit (Illumina, USA) and sequenced on an Illumina HiSeq X Ten Sequencer (Illumina, USA) or the MGIEasy RNA Library Preparation Kit (MGI, China). Hi-C libraries were prepared for Hainan muntjac using the dnpII restriction endonuclease. The libraries were constructed on the DNBSEQ platform following the manufacturer’s instructions (MGIEasy Universal DNA Library Preparation Kit, BGI) and sequenced on a DNBSEQ-T1 sequencer (MGI, China).

Genome assembly and assessment

The study sequenced and assembled the genome of Hainan muntjac using a combination of Pacbio (HIFI) technology, HiC chromosome conformation sequencing, and short read length sequencing. The heterozygous genome was assembled using hifiasm software¹² with the raw HIFI data (Table S1), and then combined with the HIC data to decompose the hybrid genome into chromosome-level haplotypes. The genome was improved using GapCloser¹³ (v. 1.12) to intercombine short read data and gap-fill the genomes, resulting in improved assembly.

The completeness of the final genome assembly was assessed using BUSCO v5¹⁴ and compared to the mammalian database (mammal_odb10). The Burrows-Wheeler aligner was used to map reads to the genome assemblies, and coverage, depth, and alignment were calculated to assess the completeness and uniformity of the assembly.

A chromosome-level and haplotype-resolved genome for the Hainan muntjac was generated by combining ~23x PacBio HiFi long-read data, DNBSEQ short-read data, and Hi-C data. The Hainan muntjac genome, referred to as Haimjac, was assembled with a size of 2.66 Gb and a scaffold N50 of 700.27 Mb (Fig. 1, Table 1). Approximately 99.34% of the sequence bridges, equivalent to over 2.40 Gb, were mapped to three chromosome-scale pseudomolecule (Figure S1, Table S2). Our assembly results and the specific karyotype composition of this species are very similar to that of the closely related Muntiacus muntjak species⁸.

Fig. 1.

Genomic landscapes of the Hainan muntjac. A. sequencing depth; B. GC contents; C. gene density; D. Schematic diagram of chromosomes; The photo of a female Hainan muntjac taken with a normal color camera that was infrared controlled in the field.

Table 1.

Statistics of the sequencing data, assembly, and annotation results of the Hainan muntjac genome.

	Category	Haimuj
Sequencing data	PacBio (Gb)	63.24
	WGS (Gb)	195.26
	Hi-C (Gb)	224.39
	RNA-seq (Gb)	22.25
Assembly	Assembled genome size (Gb)	2.66
	Contig N50 (Mb)	29.27
	Scaffold N50 (Mb)	700.27
	Longest scaffold (Mb)	1,132
	GC content (%)	42.19
Annotation	Repeat sequences (%)	36.20
	Number of protein-coding genes	21,451
	Number of functionally annotated genes	21,139

Through the utilization of (Benchmarking Universal Single-Copy Orthologs) to evaluate genome quality, we achieved a score of 95.9%, demonstrating a high level of genomic integrity and quality (Table S3).

Genome annotation

Firstly, we annotated repetitive elements using a combination of de novo and homologous methods. To identify novel repetitive sequences, we used RepeatModeler2¹⁵ (v2.0.1) and LTR finder¹⁶ (v1.0.6), and merged them with known elements in the RepBase database. We then performed a conserved BLASTN search on the RepBase library using RepeatMasker¹⁷ (v4.0.5) to classify transposable elements. The RepeatProteinMask program from RepeatMasker¹⁷ (v4.0.5) was utilized to identify homologous repeat proteins. Furthermore, the Finder software¹⁸ (v4.07) was employed to annotate tandem repeats. Additionally, blastall¹⁹ (v2.2.26), tRNAscan-SE²⁰ (v2.0.9), and INFERNAL v1.1.1²¹ were used to characterize ribosomal RNA (rRNA), transfer RNA (tRNA), and microRNA (miRNA), respectively, with the Rfam database.

The results based on the above methodology show that genome contains 963.24 Mb (36.7%) of repetitive elements, with LINE being the most prevalent at 35.24% (Table S4, S5). The LTR is the next largest, occupying 11.03% of the genome. In the Hainan muntjac genome, we predicted 250 rRNA, 386 miRNA, 339 tRNA, and 352 snRNA (Table S6).

For protein-coding gene annotation, we masked all repetitive elements in the genome and combined transcripts, homology evidence, and de novo evidence to construct the final gene set. To annotate genes based on homology, we used blastall¹⁹ (v2.2). To conduct comparisons, set the E-value truncation to 1e-5 and include the following species: Cervus canadensis, Bos taurus, Ovis aries, Homo sapiens, Cervus elaphus. For de novo gene annotation, we used Augustus²² (v3.0.3), GlimmerHMM²³ (v3.0.1), and SNAP²⁴ (v11/29/2013) for prediction. Long-read and short-read RNA sequencing data were employed for transcript-based gene prediction. Transcripts were identified using the IsoSeq (v3) pipeline. Short-read mapping was performed using HISAT2²⁵ (v2.1.0). The final gene set was generated by the MAKER pipeline²⁶ (v3.01.03) through the combination of genes predicted through RNA-seq, homology, and novel approaches.

Using de novo prediction, homology-based protein alignment, and RNA-seq mapping methods, we identified 21,451 gene models in the genome (Figure S2). These gene regions span over 732.68 Mb, accounting for 27.53% of the Haimjac genome (Table S7). The final integrated gene set was subjected to BUSCO analysis, and the gene set integrity of the Hainan muntjac genome was high with a score of 91.5% (Table S8). Notably, 98.55% of the genes in the Hainan muntjac genome received functional annotations (Table S9 and Figure S3).

The genome was functionally annotated by conducting BLAST searches against the SwissProt, TrEMBL, and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases, using E-value cutoffs of 1e-5. To predict motifs, domains, and gene ontology (GO) terms, InterProScan²⁷ (v5.52–86.0) was utilized.

Phylogeny construction and gene family expansion

The protein sequences of single-copy gene families were compared using the MUSCLE algorithm²⁸ to perform sequence alignment. RAxML²⁹ was used to construct phylogenetic trees, and the corresponding dendrogram files were generated. To enhance clarity, we visualized the dendrograms using MEGA software. To estimate the divergence time of species, we utilized the software packages PAML 4³⁰ and MCMCTREE³¹. This step aimed to provide a more precise estimate of the evolutionary timeline of the species. To cluster gene families, we utilized the BLASTP tool to compare protein sequences among all species, estimating similarities and differences between genes. Setting the E-value threshold at 1e-7 ensured the fidelity of our outcomes. Subsequently, gene clustering was executed via H-cluster software, enhancing our comprehension of intra-family correlations. Employing the TreeFam methodology, we elucidated homologous and paralogous associations across species, thereby unveiling the evolutionary trajectory of single-copy gene families within the phylogenetic framework. The analysis of gene family clustering in Hainan muntjac identified 10,056 gene families, which included 4,538 single-copy orthologs, 12822 multiple-copy orthologs 173 unique paralogs and 3364 other orthologs (Fig. 2a, Table S10). Additionally, the analysis of gene family expansion and contraction (CAFE) revealed 336 expanded and 1156 contracted gene families (P < 0.05; Fig. 2b).

Fig. 2.

(a) Comparison of the number of homologous genes. (b) The phylogenetic relationship of 9 species and the estimated divergence time. Numbers on the branch of the phylogenetic tree represent the number of significantly expanded (green) and contracted (red) gene families.

Data Records

The Hainan muntjac genome assembly has been deposited in the NCBI BioProject database under accession number PRJNA1090610. The genomic RNA sequencing data are available in the NCBI Sequence Read Archive (SRA) under accession number SRR28810459³². The Hi-C sequencing data are deposited under SRA accession numbers SRR28810461³³ and SRR28810462³⁴, while the genomic Pacbio sequencing data are available under accession number SRR28810463³⁵. The assembled genome has been deposited in GenBank under accession number GCA_039877825.1³⁶. The annotation of the genome, including information on repetitive sequences, gene structure and functional predictions, is available in the Figshare database³⁷.

Technical Validation

Genomic integrity, fragmentation, and potential loss rates were measured using BUSCO V5. Among 9226 prospective conserved core genes in the mammalian database, 95.9% and 1% were identified as complete BUSCOs and fragment BUSCOs, respectively, indicating that the assembled genome had high integrity and validity and could be used for further analysis.

Author contributions

Y.C., Y.L. and H.L. wrote the manuscript. Y.L., J.L. and M.C. collected the samples. Y.C., M.C., X.L. and Z.X. performed the RNA/DNA isolation. Y.C., Y.L., Z.X. performed the data analysis. X.L. and H.L. reviewed the manuscript. H.L. provided the supervision of this project All authors read and approved the final manuscript.

Code availability

No specific scripts were used in this work. All codes and pipelines for data processing were executed following the manuals and protocols of the respective bioinformatics software. The specific software versions are detailed in the Methods section.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41597-024-04167-2.