Chromosome-level genome assembly of stem borer Batocera rufomaculata using PacBio HiFi and Hi-C sequencing

Abstract

Batocera rufomaculata (Cerambycidae: Lamiinae), a prominent representative of longhorned beetles, is a globally significant stem-boring pest, infesting over 50 species of deciduous trees. Despite its substantial ecological and economic impact, the genomic basis underlying its host adaptation remain poorly understood. Here, we present a chromosome-level genome assembly of B. rufomaculata, constructed using a combination of Illumina, PacBio HiFi, and Hi-C sequencing data. The genome spans 338.08 Mb, with a scaffold N50 of 37.00 Mb, and is organized into 10 pseudo-chromosomes, including a chromosome X validated by genome collinearity and sequencing depth analyses. Repetitive elements constitute 27.89% of the genome, totaling 94.29 Mb. Out of 17,887 predicted genes, 12,729 were functionally annotated with at least one supporting evidence. The high-quality genome assembly and annotation were confirmed by multiple metrics, including genome size, reads mapping rate (>99.5%), BUSCO completeness (>97.1%), and collinearity analyses. This comprehensive genomic resource provides a foundation for investigating the ecological adaptation of B. rufomaculata and offers valuable insights into the genetic mechanisms that could inform pest management strategies.

Background & Summary

Longhorned beetles (Cerambycidae) are among the most ecologically significant and morphologically charismatic insects. They are distinguished by their prominent long antennae (1–2 times the length of their body) and are one of the most species-rich families, with approximately 35,000 extant species globally distributed^1–3. These beetles exhibit remarkable ecological versatility, serving roles as decomposers, pollinators, vectors, and pests, while also displaying a wide phytophagous host range^1–3. Batocera rufomaculata (Cerambycidae: Lamiinae) is a quintessential representative of this family. Known commonly as the mango stem borer, fig borer, or tropical fig borer due to its destructive tendencies⁴, this species poses a significant threat as a stem-boring pest in both non-wood product and timber forests. It infests over 50 deciduous tree species, including economically important crops such as Mangifera indica (mango), durian, Coffee, Morus spp., Moringa spp.^5,6. Its distribution spans across Asia and Europe, with reported occurrences in countries such as India, Nepal, Pakistan, Thailand, Malaysia, China, Israel, Turkey, and France^5,7–9. Despite its ecological and economic importance, the genomic basis underlying its host adaptation remain poorly understood.

Throughout its life cycle, B. rufomaculata poses a persistent threat to its host plants (Fig. 1), yet effective management strategies remain limited. A single female can lay up to 200 eggs onto the bark¹⁰. Once the larvae hatch, they burrow deep into the stems or shoots, making them difficult to control¹⁰. The larval stage lasts approximately one year, after which the insects pupate and emerge as adults with a lifespan of 60–100 days. Adults feed on green growth tips and twig bark, further damaging the host plants^10,11. Current control methods primarily rely on physical or mechanical removal of the pests, which is both costly and inefficient¹⁰. However, advancements in molecular genetics and genomics offer promising avenues for pest management¹². Techniques such as RNA interference (RNAi) and CRISPR/Cas9 have demonstrated significant potential in controlling invasive insect species^13,14. Additionally, the decreasing costs of genomic sequencing technologies provide opportunities to unravel the molecular basis of pest evolution and outbreaks. Despite these advancements, no comprehensive genome resource is currently available for B. rufomaculata, hindering research into its biology and the development of targeted control measures.

Fig. 1.

Overview of Batocera rufomaculata (Degeer, 1775). Photos and permission by Yongke Zhang and Tianqi Bai. (a) Frass holes and larval excrement around the infected trunk of Mangifera indica. (b) Emergence holes created by adult on the trunk of Mangifera indica. (c) Longitudinal section of larval galleries within the trunk of Mangifera indica. (d) Egg. (e) Pupa. (f) Adult.

In this study, we present a high-quality chromosome-level genome assembly of B. rufomaculata, integrating PacBio high-fidelity (HiFi) reads, Illumina short-reads, and high-throughput chromosome conformation capture (Hi-C) data. The genome spans 10 pseudo-chromosomes (Fig. 2a,b), with chr10 identified as the X chromosome based on collinearity analysis and sequencing depth (Fig. 2c,d). The assembly and annotation were rigorously validated through metrics such as genome size comparison, read mapping rates, Benchmarking Universal Single-Copy Orthologs (BUSCO) analysis, and synteny assessment (Tables 1–7). This high-confidence genomic resource provides a foundation for understanding the adaptive and demographic processes of B. rufomaculata and paves the way for the development of novel, targeted pest management strategies.

Fig. 2.

Assembly of the chromosome-level genome of Batocera rufomaculata. (a) The heatmap displays all interactions among 10 chromosomes. The color of the Hi-C interaction linkages represents the frequency, which ranges from yellow (low) to red (high). (b) Blocks on the outmost circle represent all 10 chromosomes. Peak plots from inner to outer circles represent gene density, repeat density, GC ratio, and the coverage of Hi-C reads, HiFi-reads, and Illumina short-reads of each chromosome, respectively (sliding window size = 50 Kb). (c) Mean sequencing depth of Illumina short-reads and PacBio HiFi-reads for each chromosome of B. rufomaculata genome. (d) Chromosome-level synteny analysis of the B. rufomaculata (Bru) and another two Cerambycidae beetles, Leptura quadrifasciata (Lqu) and Rutpela maculata (Rma).

Table 1.

Summary of sequence reads.

Sequencing strategy	Platform	Usage	Clean data (Gbp)	Mapping rate (%)
Short-reads	Illumina	Genome survey	89	99.55
HiFi-reads	PacBio Revio	Genome assembly	21	99.98
Hi-C reads	Illumina	Hi-C assembly	61	99.78
RNA-seq	Illumina	Annotation	6	—

Table 7.

Evaluation of the completeness of assembly and predicted protein-coding genes (PCGs) of Batocera rufomaculata by BUSCO with endopterygota_odb10.

BUSCO (%)	Complete	Single-copy	Duplicated	Fragmented	Missing
HiFi assembly	99.1	97.9	1.2	0.3	0.6
Hi-C assembly	99.3	98.9	0.4	0.1	0.6
PCGs	97.2	96.8	0.4	0.5	2.3

Methods

Sample information

Samples of B. rufomaculata were collected in May 2023 from a mango plantation in Lujiang Township, Longyang District, Baoshan City, Yunnan Province (24°48’ N, 98°40’ E, 705 m above sea level) for this study. The collected samples were then loaded into 50 mL centrifuge tubes, quick-frozen with liquid nitrogen, and stored in an ultra-low-temperature refrigerator at −80 °C until use. Voucher specimens were deposited in the Tropical and Subtropical Cash Crop Research Institute, Yunnan Academy of Agricultural Sciences.

Illumina, PacBio HiFi, Hi-C, and RNA sequencing

For dissecting the reference genome of B. rufomaculata, genomic DNA was extracted from a male individual and prepared by the CTAB method followed by purification with a Grandomics Genomic kit for regular sequencing, according to the standard operating procedure provided by the manufacturer. DNA quality was monitored on 1% agarose gels, detected using NanoDrop™ One UV-Vis spectrophotometer (Thermo Fisher Scientific, USA), and measured by Qubit® 4.0 Fluorometer (Invitrogen, USA). All sequencing was performed at NextOmics (Wuhan, China) (Table 1).

The short-reads libraries (350-bp insert size) were generated using Truseq Nano DNA HT Sample Preparation Kit (Illumina USA) following the manufacturer’s recommendations. The DNA fragments were then end-polished, A-tailed, and ligated with the full-length adapter for Illumina sequencing, followed by PCR amplification. 150-bp paired-end reads were generated on the Illumina NovaSeq 6000 platform. The quality per base was evaluated with FastQC¹⁵. Both low-quality reads ( > 90% bases with quality < Q30) and adaptor sequences were removed. A total of 89 Gb clean data was obtained.

The HiFi-reads libraries were prepared using SMRTbell® prep kit 3.0 kit following the manufacturer’s recommendations. After DNA quality control, DNA shearing and cleanup, DNA repair and A-tailing reactions, adapter ligation and cleanup, nuclease treatment, and size selection, the cleanup of libraries transferred to the PacBio Revio equipment according to the operating manual provided by PacBio. The min passes = 3 and min RQ = 0.99 default parameters in CCS software (https://github.com/PacificBiosciences/ccs) were utilized to generate 21 Gb high-precision HiFi reads with quality over Q20.

The Hi-C libraries for Illumina sequencing were prepared using the NEBNext Ultra II DNA Library Prep Kit for Illumina (NEB) according to the manufacturer’s instructions. During the process, the collected 10⁶ cells were cross-linked, and the restriction enzyme DpnII was used to lyse and digest the isolated cells. The final library was sequenced on the Illumina Novaseq 6000 platform. After quality control, 61 Gb clean data were generated for Hi-C assembly.

Total RNA was also extracted from the same individual for extracting genomic DNA. The RNA libraries were generated using TruSeq RNA Library Preparation Kit (Illumina, USA) following the manufacturer’s recommendations. After cluster generation, the library preparations were sequenced on an Illumina Novaseq 6000 platform. About 6 Gb paired-end reads (150-bp) were generated and used to annotate the protein-coding genes of the genome of B. rufomaculata.

Chromosome-level genome assembly

The genome was assembled using hifiasm v0.20.0-r639¹⁶ with the parameter (‘-l 3 -n 4’). The haploids and contig overlaps were removed using purge_dups (https://github.com/dfguan/purge_dups). The non-redundant genome was polished using the HiFi reads and Illumina short-reads by NextPolish2 v0.2.1¹⁷. Next, we pre-processed the Hi-C data with Juicer v1.6¹⁸ and 3D-DNA v180114¹⁹. Manually visualized and corrected the pseudo-chromosomes with JuiceBox v2.15 (https://github.com/aidenlab/Juicebox). Finally, we ran 3D-DNA again to obtain the final chromosome-level genome assembly. The final genome size was 338.08 Mb, with a scaffold N50 of 37.03 Mb (Table 2), and 99.77% of the sequence anchored into 10 pseudo-chromosomes.

Table 2.

Statistics for the Batocera rufomaculata genome assembly.

Genome features	HiFi assembly	Hi-C assembly
N90	2,171,673	21,804,110
N80	7,483,781	29,473,481
N70	13,653,612	35,220,785
N60	15,465,244	36,715,325
N50	16,611,624	37,032,474
N40	16,826,451	37,032,474
N30	18,649,591	40,211,048
N20	20,971,516	41,303,899
N10	35,242,381	51,781,017
Max length	35,242,381	51,781,017
Total length	346,017,884	338,076,939
Total number	158	13
Average length	2,189,986	26,005,918

Repeat annotation

A combination of de novo and homology-based approaches was applied to predict the repeated sequences in the B. rufomaculata genome. RepeatModeler (http://www.repeatmasker.org/RepeatModeler/) was used to identify and classify novel TEs. The homology-based approach was used to detect known transposable elements (TEs) at DNA and protein levels. RepeatMasker was used to identify the DNA level TEs with the Repbase TE library 21.04²⁰. RepeatProteinMask²¹ in RepeatMasker was used to search the protein level TEs with parameters “-noLowSimple -pvalue 0.0001”. In addition, Tandem Repeat Finder²² was used to discover the tandem repeats with parameters “2 7 7 80 10 50 2000 –d –h”. In total, we identified 94.29 Mb of repetitive elements, accounting for 27.89% of the genome (Table 3). Among these elements, 24.93% were identified as TEs, including the most abundant DNA transposons, followed by LTR, LINE, and SINE retrotransposons (Table 4).

Table 3.

Statistics of repeats in the assembled genomes.

Type	Repeat size	% of genome
Trf	18,938,174	5.60
Repeatmasker	15,141,052	4.48
Proteinmask	6,972,149	2.06
De novo	73,445,725	21.72
Total	94,285,011	27.89

Table 4.

Transposable elements (TEs) content of the assembled genome.

Type (%)	Repbase TEs	TE protiens	De novo	Combined TEs
DNA	2.97	0.36	5.74	7.53
LTR	1.35	1.70	2.51	3.45
LINE	0.58	0.00	0.67	1.14
SINE	0.01	0.00	0.26	0.27
Other	0.00	0.00	0.00	0.00
Unknown	0.00	0.00	12.70	12.70
Total	4.48	2.06	21.67	24.93

Gene structure prediction

The protein-coding genes (PCGs) of the repeat-masked genome were predicted based on de novo prediction, homology prediction, and RNA-seq prediction. For de novo prediction, AUGUSTUS²³ was used to predict coding genes. For homology prediction, protein data sets of other four long-horned beetles, Anoplophora glabripennis (GCF_000390285.2), Monochamus alternatus (GCA_035320865.1), Leptura quadrifasciata (GCA_963675865.1), and Rutpela maculata (GCA_936432065.2), were aligned against the assembled genome using TBLASTN with an E-value cut-off of 1e-5. Solar software (The Beijing Genomics Institute development) was used to conjoin the BLAST hits and GeneWise²⁴ was applied to predict gene structures based on each BLAST hit. For RNA-seq prediction, we aligned RNA-seq using HISAT2 v2.2.1²⁵ and assembled unigenes using StringTie v2.1.6²⁶. TransDecoder (https://github.com/TransDecoder/TransDecoder) was used to predict the gene models. Through integration with EvidenceModeler v1.1.1²⁷, we predicted a total of 17,887 PCGs in the genome.

Gene function prediction

The function of predicted genes was annotated using BLASTP (E-value < 1e-05) against databases including SwissProt, TrEMBL, and NCBI non-redundant proteins database (NR). The structural domains and motifs of all genes were scanned against SMART, ProDom, Pfam, PRINTS, PROSITE, and PANTHER databases using InterProScan v5.25²⁸. The Gene Ontology (GO) terms were extracted based on the corresponding InterPro entry. The metabolic pathways in which the genes might be involved were assigned by BLAST (E-value < 1e-05) against the Kyoto Encyclopedia of Genes and Genomes (KEGG) protein database²⁹. In total, 12,729 predicted genes were supported by at least one functional clue (Table 5).

Table 5.

Statistics of functional annotation in Batocera rufomaculata genome.

Database	Number	Percentage (%)
InterPro	10,332	57.76
GO	7,367	41.19
KEGG	8,193	45.80
Swissprot	9,598	53.66
TrEMBL	7,636	42.69
NR	12,709	71.05
Annotated	12,729	71.16
Unanotated	5,158	28.84

Data Records

All sequencing data (Short-reads, HiFi-reads, Hi-C reads, and RNA-seq) used for genome assembly and annotation, were submitted to NCBI under the BioProject PRJNA1209233. Illumina sequencing data for genome survey were deposited in the Sequence Read Archive at NCBI under accession number SRR32458641³⁰. Hi-C sequencing data were deposited in the Sequence Read Archive at NCBI under accession number SRR32456401³¹. PacBio HiFi sequencing data were deposited in the Sequence Read Archive at NCBI under accession number SRR32456009³². RNA-seq data were deposited in the Sequence Read Archive at NCBI under accession numbers SRR32458638³³. The final chromosome assembly was deposited in GenBank under accession number JBKOFD000000000³⁴. The final gene annotation results of B. rufomaculata have been deposited in Figshare repository under a DOI number of 10.6084/m9.figshare.28466675.v1³⁵.

Technical Validation

Genome assembly and annotation assessment

The assembled size is slightly larger than the size of 325.90 Mb estimated by analyzing 17-mer frequency³⁶ utilizing Illumina short-reads (Table 6). We segmented the assembled chromosomes into 50 kb-sized fragments and plotted a heatmap using juicerbox 2matrix (https://github.com/yukaiquan/biotools/tree/main/Hic/juicerbox2matrix). The heatmap shows that on all 10 chromosomes, the interaction strength at diagonal positions is significantly higher than at non-diagonal regions, indicating good quality of chromosome assembly (Fig. 2a). BUSCO v5.2.2³⁷ was used to evaluate the completeness and redundancy of gene regions of the assembly and the protein-coding genes based on the lineage dataset of endopterygota_odb10. The complete BUSCOs of draft genomes, final genomes, and PCGs have reached over 97.00% (Table 7).

Table 6.

The genomic characteristics were estimated by analyzing 17-mer frequency utilizing Illumina short-reads.

Kmer_num (Gb)	Used_base (Gb)	Used_read (Mb)	Peak	Genome size (Mb)	Repeat (%)	Heterozygosity (%)
78.77	88.18	587.86	237	325.90	29.3	1.21

The Illumina short-reads, HiFi-reads, and Hi-C reads were mapped onto the assembled genome using BWA v0.7.12³⁸ with default settings. All mapping rates counted using the Samtools v1.2.1³⁹, have exceeded 99.50% (Table 1). The sequencing depths of chr10 had half the sequencing depth of the other chromosomes (Fig. 2c), suggesting chr10 may be the sex chromosome X.

Genome synteny

The chromosome X has been identified in the genome of L. quadrifasciata⁴⁰ and R. maculata⁴¹ of the same family as B. rufomaculata. Collinearity analysis was performed by Python MCScanX pipeline⁴² in TBtools-II⁴³. These three species showed good chromosomal collinearities and chr10 of B. rufomaculata has a good collinearity with chromosome X of L. quadrifasciata and R. maculata (Fig. 2d), confirming that chr10 should be sex chromosome X.

Author contributions

L.X.Y. conceived and supervised the study. B.T.Q. and D.Z.W. collected samples. W.W.T. and W.Y.J. performed sequencing. H.J.W. performed genome assembly and annotation analysis. H.J.W. and B.T.Q. wrote the draft manuscript. H.J.W., L.X.Y, B.T.Q. and Z.H.R. improved and revised the manuscript. All authors improved, revised, and approved the manuscript.

Code availability

All software and pipelines were executed following the manuals and protocols provided by the published bioinformatic tools. The version and parameters of the software have been described in the Methods section.

Competing interests

The authors declare no competing interests.