Chromosome-level genome assembly of Corizus hyoscyami (Linnaeus, 1758) (Heteroptera: Rhopalidae)

Abstract

Corizus hyoscyami (Heteroptera: Rhopalidae) is a widely distributed and highly polyphagous insect, making it a relevant species for understanding the evolution of herbivory and host adaptation in true bugs. However, the genome of Corizus hyoscyami has not yet been sequenced, limiting investigations into its biological characteristics. In this study, a high-quality chromosome-level genome assembly was generated by integrating MGI DNBSEQ-T7 short-read, PacBio Sequel II long-read, and Hi-C scaffolding technologies. The assembly was assessed to be 97.6% complete by BUSCO. The final genome was assembled into a size of 1.47 Gb, with a contig N50 of 1.61 Mb and a scaffold N50 of 198.34 Mb. The GC content was determined to be 34%, and 94.90% of the sequences were anchored onto 6 chromosome-level scaffolds, representing the major chromosomal elements of the genome. Repetitive sequences accounted for 69.79% of the genome, and 18,540 protein-coding genes were annotated. This genome assembly is expected to facilitate research on adaptive evolution in Heteroptera.

Background & Summary

Corizus hyoscyami (Linnaeus, 1758) (Heteroptera: Rhopalidae), commonly known as the cinnamon bug or the black and red squash bug, is a widely distributed scentless plant bug. As a mesophilous species, it predominantly inhabits open areas with moderate humidity across the Palearctic region. Its distribution spans from Western Europe and North Africa, extending eastwards across the Middle East, Central Asia, and Siberia to Mongolia and Northern China¹. Within these regions, it frequently occurs in meadows, forest clearings, and grasslands². C. hyoscyami exhibits high polyphagy with a broad host range. During spring, both adults and nymphs primarily consume floral tissues of willow and young shoots of deciduous and coniferous trees including birch and pine. As the season progresses, feeding shifts to herbaceous plants such as sow-thistle, chamomile, and Euphorbia. Major host species include Hyoscyamus niger, Tabacum, Ononis spinosa, and Erodium. Additionally, it is recognized as a significant pest damaging fruits, vegetables, maize, and leguminous crops^3–5. The species is bivoltine, completing two generations per year, and overwinters in the adult stage^6,7. Adults remain active throughout the year, with new generations emerging between August and September. Owing to its ecological adaptability and broad dietary range, C. hyoscyami often becomes a widespread and dominant species in suitable habitats⁸.

Despite its ecological and agricultural importance, molecular research on C. hyoscyami remains limited. Although chromosome-level genomes have been successfully deciphered for an increasing number of heteropteran species, such as the predatory stink bugs Eocanthecona furcellata⁹ and Arma custos¹⁰, the Chinese water scorpion Ranatra chinensis¹¹, the jiuxiang bug Aspongopus chinensis¹², and the assassin bug Rhynocoris fuscipes¹³, genomic information for the vast majority of insect species remains unknown. Most existing studies have focused on morphological descriptions and field observations, leaving significant knowledge gaps in understanding its genomic architecture, evolutionary history, and molecular mechanisms underlying host adaptation and ecological resilience. As a polyphagous insect exhibiting seasonal host-switching behavior, C. hyoscyami serves as an ideal model for investigating plant-insect coevolution, adaptive radiation, and the genetic basis of dietary breadth. However, the absence of a high-quality reference genome has impeded in-depth exploration of these scientific questions.

To address this gap, we employed an integrated sequencing strategy combining PacBio long-read sequencing, MGI DNBSEQ short-read sequencing, and Hi-C-based chromatin conformation capture to successfully assemble a chromosome-level genome for C. hyoscyami. This genomic resource will not only provide a fundamental basis for research on the evolution of herbivory and pest management strategies but also significantly advance our understanding of insect adaptive responses to environmental variations.

Methods

Sample preparation and genome sequencing

Adult specimens of C. hyoscyami were collected from Zepu County, Kashgar, Xinjiang, China (38.2504°N, 77.3175°E). Specimens used for sequencing were identified based on the morphological characters described in Göllner-Scheiding (1980)¹⁴ and Moulet (1995)¹⁵. To provide molecular evidence for species identification and establish a definitive molecular marker for this reference genome, genomic DNA was extracted from a single adult specimen, from which a 621 bp fragment of the cytochrome c oxidase subunit I (COI) gene was amplified and Sanger-sequenced. To obtain sufficient high-quality DNA, genomic DNA (gDNA) was extracted from a pool of three female adult specimens using the cetyltrimethylammonium bromide (CTAB) method and purified with the Grandomics Genomic Kit (Grandomics, Wuhan, China). The extracted DNA was subsequently utilized for both Pacific Biosciences (PacBio) high-fidelity (HiFi) sequencing and genome survey sequencing. For HiFi sequencing, a SMRTbell library was prepared with the SMRTbell prep kit 3.0 (PacBio, Menlo Park, CA, USA). The PacBio Revio platform generated 46.64 Gb of HiFi data through circular consensus sequencing (CCS), which were subsequently used for de novo genome assembly (Table 1).

Table 1.

Sequencing data and assembly strategy for the Corizus hyoscyami genome.

	PacBio	Survey	Hi-C	RNA-seq
Clean data (Gb)	46.64	49.22	161.94	8.04
Coverage (X)	31.64	66.78	109.86	5.46
GC content (%)	33.56	35.35	37.27	38.75

For genome survey analysis, the same gDNA batch was processed into a short-read library using the MGIEasy Universal DNA Library Prep Kit V1.0 (MGI, Shenzhen, China). Sequencing on the DNBSEQ-T7RS platform (MGI, Shenzhen, China) in paired-end 150-base pair (bp) mode generated 51.39 Gb of raw data, with 49.22 Gb of clean data retained after Fastp version 0.23.4¹⁶ processing (Table 1). The mitochondrial genome of C. hyoscyami was obtained through de novo assembly of clean short-reads from the genome survey data using IDBA-UD version 1.1.3¹⁷ with a k-mer range of 41 to 141 bp. To evaluate assembly accuracy, clean reads were mapped back onto the obtained mitogenomic contig using Geneious Prime 2023.2.1¹⁸. The resulting assembly was subsequently annotated using the MITOZ pipeline version 3.6¹⁹ under the invertebrate mitochondrial genetic code, which identified a complete set of 37 standard mitochondrial genes. Finally, the full-length COI gene sequence was extracted and compared with the Sanger-sequenced fragment obtained from the single specimen, exhibiting 98.9% identity.

For high-throughput chromatin conformation capture (Hi-C) and ribonucleic acid (RNA) sequencing, a single adult specimen was utilized to ensure genomic consistency. The head of this individual was used for Hi-C library preparation, while the remaining body was used for total RNA extraction. Chromosome-level scaffolding was achieved through Hi-C library preparation by digesting nuclei with the DpnII restriction enzyme. The Hi-C protocol involved biotin labeling, blunt-end ligation, and DNA fragmentation to 300–600 bp fragments. The final libraries were sequenced on the DNBSEQ-T7RS platform (MGI, Shenzhen, China), generating 161.94 Gb of Hi-C reads (Table 1).

For genome annotation, transcriptome sequencing (RNA-seq) was performed. Total RNA was extracted from the body of the aforementioned individual using TRIzol reagent (Tiangen Biotech, Beijing, China). A non-strand-specific RNA-seq library was prepared using the MGIEasy RNA Library Prep Kit V3.1 (MGI, Shenzhen, China) and sequenced on the DNBSEQ-T7RS platform (MGI, Shenzhen, China) in paired-end 150-bp mode, generating approximately 8.04 Gb of high-quality short-read data (Table 1).

Genome size estimation

Genomic characteristics including genome size, duplication, and heterozygosity were estimated through K-mer analysis using 49.22 Gb clean MGI DNBSEQ short reads (Table 1). The k-mer frequency distribution was analyzed with JELLYFISH version 2.2.10²⁰ (k = 17), followed by genome profiling in GenomeScope version 2.0²¹ under default parameters. This analysis revealed a genome size of 1.56 Gb with 0.884% heterozygosity (Table 2).

Table 2.

Features of the Corizus hyoscyami genome assembly.

Genomic features	Statistics
Assembly level	Chromosome -level
Heterozygosity (%)	0.884
Survey	1,563,998,627
Genome size (bp)	1,475,691,838
Contig N50 (bp)	1,609,326
Scaffold N50 (bp)	198,335,333
Number of pseudochromosomes	6
BUSCO (%)	97.6
GC content (%)	34
Protein-coding genes	18,540
Repetitive elements (%)	69.86

Note: The 6 pseudochromosomes represent the five major autosomes and the X chromosome, consistent with the major elements of the known C. hyoscyami karyotype (n = 7, 5 A + m + X).

Genome assembly and Hi-C scaffolding

The C. hyoscyami genome was assembled using PacBio long-read sequencing data with WTDBG2 version 2.5²². Purge_Dups version 1.2.5²³ was used to remove haplotypic duplications, producing a contig-level genome assembly of 1.47 Gb with 34% GC content and a contig N50 of 1.60 Mb (Table 2). For chromosome-level scaffolding, high-quality Hi-C reads were aligned to the contig assembly using the end-to-end alignment algorithm implemented in BWA-MEM version 0.7.18²⁴. Scaffolds were assembled into pseudochromosomes using Juicer version 1.6²⁵ and 3D-DNA version 180419²⁶ with parameter -r 2. The chromosome-level assembly was manually reviewed and adjusted using Juicebox (https://github.com/aidenlab/juicebox/wiki/Download), resulting in 6 well-anchored pseudochromosomes representing the major putative chromosomal elements (the five autosomes and the X chromosome), with a scaffold N50 of 198.34 Mb (Fig. 1). The comprehensive genome architecture, including the distribution of gene density and repeat content across these pseudochromosomes, is summarized in a circular representation (Fig. 2). Although previous cytogenetic studies of C. hyoscyami reported a karyotype of n = 7 (5 A + m + X)^27,28, the m-chromosome was likely not clustered into a separate pseudochromosome due to its significantly small size and highly repetitive nature, which often results in insufficient Hi-C contact signals. Consequently, the sequences belonging to the m-chromosome are presumed to be distributed among the remaining unanchored scaffolds. Further analysis revealed that these unplaced scaffolds (comprising approximately 5.10% of the total assembly) were characterized by an exceptionally high repeat content (90.51%) and a significantly higher proportion of tandem repeats (9.49%) compared to the anchored chromosomes (2.00%). This high repetitive density likely caused substantial mapping ambiguity during Hi-C scaffolding, preventing their confident positioning. To ensure genomic transparency and data integrity, all assembled sequences, including these unplaced scaffolds, have been included in the public data release via Figshare²⁹. The completeness of the final assembly was assessed using BUSCO version 5.7.1³⁰ with the insecta_odb10 database, showing 97.6% completeness, confirming the high quality of the genome assembly (Table 2). In addition, the obtained genome assembly does not include telomeric sequences.

Fig. 1.

Genome-wide contact matrix of Corizus hyoscyami generated using Hi-C data.

Fig. 2.

Circular representation of the Corizus hyoscyami chromosome-level genome. Tracks represent the distribution of (a) pseudochromosomes karyotypes, (b) gene density, (c) GC content, (d) DNA transposon density, (e) LTR retrotransposon content, (f) transposable element density, and (g) tandem repeat content. Densities were calculated in 200 kb windows.

Repeat sequences annotation

Repetitive sequences were identified through a combined approach utilizing both de novo prediction and homology-based searches. A custom repeat library was generated using RepeatModeler version 2.0.5³¹ and subsequently integrated with RepBase (20170127 edition)³². Genome-wide repeat annotation was performed using RepeatMasker version 4.1.7³³ and RepeatProteinMask version 4.1.7³³ against this composite library. Specialized analyses were conducted for specific repeat classes: LTR retrotransposons were detected using LTR_FINDER version 1.07³⁴ with verification by LTR_retriever version 2.9³⁵, while tandem repeats were annotated with Tandem Repeats Finder version 4.09.1³⁶ (parameters: 2 7 7 80 10 50 500). The genome assembly contained 1.03 Gb (69.79%) of repetitive sequences, which fell into one major category (Table 3). Retrotransposons represented the largest proportion (49.82%), consisting of LTR retrotransposon (43.80%) and non-LTR Retrotransposon (6.02%). Among retrotransposons, we identified long terminal repeat (LTR) elements, short interspersed nuclear elements (SINEs), and long interspersed nuclear elements (LINEs). The LTR elements were further categorized into Copia, Gypsy, and other unclassified LTR families (Table 3).

Table 3.

The statistics of repeat contents in the genomes of Corizus hyoscyami.

Repeat type	Length (bp)	Content (%)
Retrotransposon	735,263,299	49.82
LTR-Retrotransposon	646,417,197	43.80
LTR/Copia	2,135,014	0.14
LTR/Gypsy	612,172,658	41.48
LTR-other	32,109,525	2.18
Non-LTR Retrotransposon	88,846,102	6.02
SINE	193,014	0.01
LINE	88,653,088	6.01
DNA Transposon	39,696,110	2.69
Terminal inverted repeat	31,250,556	2.12
EnSpm/CACTA	1,879,872	0.13
Harbinger	449,919	0.03
MuDR	2,844,844	0.19
Tcl/Mariner	22,810,676	1.55
hAT	3,271,510	0.22
Non-terminal inverted repeat	8,402,272	0.57
Helitron	2,008,516	0.14
DNA-other	6,298,008	0.43
Tandem repeat	24,965,997	1.69
Unclassified	199,611,958	13.53
Total content	1,029,816,544	69.79

Protein-coding gene prediction and functional annotation

Protein-coding genes (PCGs) were predicted using a combination of ab initio, transcriptomic, and homology-based approaches. RNA-seq reads were mapped to the genome assembly using HISAT version 2.2.1³⁷, with subsequent transcript assembly performed by StringTie version 2.4.0³⁸. For homology-based prediction, protein sequences from multiple insect species were aligned to the genome using Exonerate version 2.4.0³⁹. Subsequently, predictions were performed using protein sequences from ten reference species: Drosophila melanogaster (GCF_000001215.4)⁴⁰, Apis mellifera (GCA_003254395.2)⁴¹, Bombyx mori (GCA_030269925.2)⁴², Tribolium castaneum (GCA_031307605.1)⁴³, Gerris buenoi (GCA_001010745.2)⁴⁴, Acyrthosiphon pisum (GCA_005508785.2)⁴⁵, Cimex lectularius (GCA_049903775.1)⁴⁶, Rhodnius prolixus (GCA_049639745.1)⁴⁷, Apolygus lucorum (GCA_009739505.2)⁴⁸, and Orius laevigatus (GCA_018703685.1)⁴⁹. Ab initio prediction was conducted with AUGUSTUS version 3.5.0⁵⁰, incorporating RNA-seq evidence converted to hints files using bam2hints. The final gene set was generated through evidence integration with MAKER version 3.01.03⁵¹. A total of 18,540 protein-coding genes were predicted in the C. hyoscyami genome, with average lengths of 11,303.10 bp for genes, 1,284.53 bp for coding sequences (CDS), and 428.18 amino acids for proteins (Table 4). Gene function annotation was performed using an integrated analysis pipeline with eggNOG-mapper version 2.0.1⁵², InterProScan version 5.0⁵³, BLAST version 2.16.0⁵⁴, and HMMER version 3.4⁵⁵. These tools were used to analyze sequences against multiple databases including GO, COG, KEGG, nr, Swiss-Prot, and Pfam. A comprehensive functional annotation pipeline successfully annotated 17,051 genes (91.97%) with putative functions (Table 5).

Table 4.

The statistics of gene count, mRNA count, CDS count, exon count and intron count in the genomes of Corizus hyoscyami.

Parameter	Value
gene count	18,540
gene max	286,885
gene min	153
gene Median/AVG	5,693.5/11,303.10
mRNA count	20,971
mRNA max	21,490
mRNA min	153
mRNA Median/AVG	1,068/1,497.67
mRNA count Median/AVG	1/1.13
CDS count	20,971
CDS max	20,676
CDS min	150
CDS Median/AVG	942/1,284.53
protein Median/AVG	314/428.18
exon Median/AVG	174/329.66
exon count Median/AVG	3/4.54
utr3 Median/AVG	220/687.44
utr5 Median/AVG	97/137.16
intron Median/AVG	1,879/2761.31
intron count Median/AVG	2/3.54

Table 5.

The statistics of annotation in the genomes of Corizus hyoscyami.

Database	Annotated number	Percent (%)
NR	16,711	90.13
Swiss-Prot	13,113	70.73
GO	5,225	28.18
KEGG	8,614	46.46
COG	14,204	76.61
Eggnog	14,204	76.61
Pfam	12,882	69.48
Total	17,051	91.97

Data Records

The raw sequence data generated in this study have been deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) under BioProject PRJNA1328843. Specifically, the RNA-seq data (SRR35410016⁵⁶), genomic MGI DNBSEQ short-read data (SRR35410017⁵⁷), Hi-C sequencing data (SRR35410018⁵⁸), and PacBio HiFi sequencing data (SRR35410019⁵⁹) are publicly accessible.

The chromosome-level genome assembly and functional annotation (GFF) of Corizus hyoscyami have been uploaded to GenBank in NCBI under accession number GCA_059059285.1⁶⁰. These records, along with the mitochondrial genome and COI gene sequences (including both the 621 bp Sanger-verified fragment and the assembly-derived full-length sequence) are also publicly available in the Figshare database (10.6084/m9.figshare.32593206)²⁹. All data records are publicly accessible and resolve to live records.

Technical Validation

The genome assembly was 1.47 Gb in size with a scaffold N50 of 198.34 Mb. The assembly size showed consistency with estimates from K-mer analysis. For accuracy assessment, we mapped MGI DNBSEQ short reads to the C. hyoscyami genome using BWA-MEM version 0.7.18²⁴, with 98.85% of reads successfully mapping. Genome completeness was evaluated using BUSCO version 5.7.1³⁰ with the insecta_odb10 database, showing 97.6% completeness. From 1,334 orthologs analyzed, 1,316 were complete single-copy genes.

Author contributions

J.S. and Y.X. conceived the project. Y.Z., Z.C. and Z.H. collected insect samples and performed photography. Y.X. and L.M. extracted genomic DNA and performed data analysis. Y.X. wrote the manuscript. R.J., Y.S., G.L. and H.G. discussed the results and revised the manuscript. All authors have read and approved the final version.

Data availability

The raw sequencing data are deposited in the NCBI Sequence Read Archive under BioProject PRJNA1328843. The genome assembly and annotations are available at GenBank under accession number GCA_059059285.1⁶⁰. All other associated data files, including the mitochondrial genome and COI sequences, are available via Figshare at 10.6084/m9.figshare.32593206²⁹.

Code availability

All analyses were performed using established bioinformatics tools following the developers’ standard protocols. Software versions and parameters are fully documented in the Methods section. No custom code was generated for this study.

Competing interests

The authors declare no competing interests.