Background & Summary

Coregonus ussuriensis Berg belong to the order Salmoniformes, family Salmonidae, subfamily Coregoninae, and genus Coregonus. They inhabit cold-water basins above 45° N latitude, such as Siberia and Sakhalin in Russia and Heilongjiang in China, which have typical migration characteristics. The body of C. ussuriensis is long, flat, and fusiform, with a shorter caudal stalk, shorter head, and a larger mouth in the terminal position. The eyes are larger and closer to the rostral end. The fish scale is round, large and easy to fall off. The back of the body is bluish-grey, and the side of the body is silver-white (Fig. 1)1,2. As one of the rare fishes, the meat of C. ussuriensis is delicate, has high nutritional and economic value, and is popular among consumers3. However, owing to the deterioration of the living environment, overfishing, and other factors, the C. ussuriensis resources have shown a significant decline and are included in the Red Book of Endangered Animals (Fish) of China4. For sustainability of C. ussuriensis genetic resources, research on its breeding and reproduction has been gradually undertaken.

Fig. 1
figure 1

Picture of two-years-old female Coregonus ussuriensis Berg.

Full size image

Whole-genome sequencing of specific species is essential for solving practical problems in biological research and aquaculture. Genome sequencing technology is fast maturing with developments in science and technology5. The ray-finned fish have strong reproductive ability and adaptability; they constitute the most varied, abundant, and widely distributed vertebrate species6. Fish have gradually become ideal models for vertebrate genome analysis and gene function identification. More than 200 aquatic animal genomes have been sequenced and are available in the public database of the National Center for Biotechnology Information (NCBI)7. Accurate analysis of the fish genome can reveal essential regulatory genes for the desired economic traits, providing vital data for improving production and breeding efficiency8,9.

Recent research on C. ussuriensis has mainly focused on muscle nutritional quality, gonadal and embryonic development, reproductive performance, and pathogen infection10,11,12,13,14,15,16. However, reports on the genome of C. ussuriensis have been lacking, which seriously hampers research on genetic selection at the molecular level. To overcome this gap in knowledge, in this study, we assembled and constructed a reference chromosome-level genome of C. ussuriensis using PacBio sequencing and Hi-C assembly technology. The genome assembly had a total length of 2.51 Gb, with a contig N50 of 4.27 Mb and a scaffold N50 of 62.20 Mb. This reference genome provides essential data for developing molecular markers for economic traits and should help conserve and utilise the germplasm resources of C. ussuriensis.

Methods

Ethics statement

All experiments were approved by the Animal Husbandry Department of the Heilongjiang Animal Care and Use Committee. All fish involved in this research were bred following the guidelines of the Animal Husbandry Department of Heilongjiang, China.

Sample collection and DNA extraction

Samples of C. ussuriensis were collected from Bohai Cold Water Fish Experimental Station of Heilongjiang Fisheries Research Institute (129° 04′ 64.7753′′ E; 44° 14′ 5.983′′ N). The liver tissue of the C. ussuriensis shown in Fig. 1 was collected and stored in liquid nitrogen for DNA extraction, genome library construction, and high-throughput sequencing. The CTAB-based extraction method was used to extract DNA from the liver tissue17,18. The DNA concentration was 254.0 ng/μL, and the OD260/280 value was 1.83. The quality and quantity of the extracted DNA were examined using a NanoDrop 2000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA), Qubit dsDNA HS Assay Kit on a Qubit 3.0 Fluorometer (Life Technologies, Carlsbad, CA, USA) and electrophoresis on a 0.8% agarose gel, respectively.

SMRTbell library construction and PacBio sequencing

The SMRTbell library was constructed using the SMRTbell Express Template Prep kit 2.0 (Pacific Biosciences). Briefly, 5 μg of the genomic DNA mentioned above was carried into the first enzymatic reaction to remove single-stranded overhangs, followed by treatment with repair enzymes to repair any damage that may be present on the DNA backbone. After DNA damage repair, the ends of the double-stranded fragments were polished and subsequently tailed with an A-overhang. Ligation with T-overhang SMRTbell adapters was performed at 20 °C for 60 minutes. Following ligation, the SMRTbell library was purified with 1X AMPure PB beads. The size distribution and concentration of the library were assessed using the FEMTO Pulse automated pulsed-field capillary electrophoresis instrument (Agilent Technologies, Wilmington, DE) and the Qubit 3.0 Fluorometer (Life Technologies, Carlsbad, CA, USA). Following library characterisation, 3 μg was subjected to a size selection step using the BluePippin system (Sage Science, Beverly, MA) to remove SMRTbells ≤25 kb. After size selection, the library was purified with 1 X AMPure PB beads. The FEMTO Pulse and the Qubit dsDNA HS reagents Assay kit assessed library size and quantity. Sequencing primer and Sequel II DNA Polymerase were annealed and bound to the final SMRTbell library, respectively. The library was loaded at an on-plate concentration of 35 pM using diffusion loading. SMRT sequencing was performed using a single 8 M SMRT Cell on the Sequel II System with Sequel II Sequencing Kit19,20.

Hi-C library construction and sequencing

Four steps were performed for the In situ Hi-C library construction. The first was formaldehyde cross-linking; 1 g of the same C. ussuriensis shown in Fig. 1 was cross-linked for 10 min with 1% fresh formaldehyde and quenched with 0.2 M final concentration glycine for 5 min. The second was the cell lysis; the cross-linked cells were subsequently lysed in lysis buffer (10 mM Tris-HCl (pH 8.0), 10 mM NaCl, 0.2% NP40, and complete protease inhibitors (Roche)). The extracted nuclei were re-suspended with 150 μl 0.1% SDS and incubated at 65 °C for 10 min, then SDS molecules were quenched by adding 120 μl water and 30 μl 10% Triton X-100, and incubated at 37 °C for 15 min. The DNA in the nuclei was digested by adding 30 μl 10x NEB buffer 2.1 (50 mM NaCl, 10 mM Tris-HCl, 10 mM MgCl2, 100 μg/ml BSA, pH 7.9) and 150U of MboI, and incubated at 37 °C overnight. The third step was the digestion and biotin labelling. After the MboI enzyme was inactivated at 65 °C for 20 min, the cohesive ends were filled in by adding 1 μl of 10 mM dTTP, 1 μl of 10 mM dATP, 1 μl of 10 mM dGTP, 2 μl of 5 mM biotin-14-dCTP, 14 μl water and 4 μl (40 U) Klenow, and incubated at 37 °C for 2 h. The fourth step was the ligation and DNA purification, 663 μl water, 120 μl 10 x blunt-end ligation buffer (300 mM Tris-HCl, 100 mM MgCl2, 100 mM DTT, 1 mM ATP, pH 7.8), 100 μl 10% Triton X-100 and 20 U T4 DNA ligase were added to start proximity ligation. The ligation reaction was placed at 16 °C for 4 h. After ligation, the cross-linking was reversed by 200 μg/mL proteinase K (Thermo) at 65 °C overnight. According to manufacturers’ instructions, DNA purification was achieved through the QIAamp DNA Mini Kit (Qiagen). Sequencing was performed after the library quality was verified using a BGI MGISEQ-2000 platform (PE150) sequencer.

RNA extraction and transcriptome sequencing

In total, five fish were taken for RNA extraction and transcriptome sequencing; the heart, liver, spleen, intestine, kidney, and muscle tissues of each fish were mixed to extract RNA for sequencing. Total RNA was extracted using the Trizol (Invitrogen, CA, USA), RNA purity and integrity was monitored by NanoDrop 2000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and a Bioanalyzer 2100 system (Agilent Technologies, CA, USA). RNA contamination was assessed by 1.5% agarose gel. RNA concentrations ranged from 587.0 to 2475.8 ng/μL, and the OD260/280 values ranged from 1.98 to 2.06. The integrity of RNA detection showed that the RNA integrity number ranged from 8.7 to 10, and 28 S/18 S values ranged from 0.9 to 1.8. Transcriptome sequencing of qualified RNA was performed on an Illumina NovaSeq. 6000 platform and the results were used for gene prediction.

Genome survey analysis

Before genome assembly, SOAPnuke v2.1.021 was used to control the quality of the DNA sequencing library. Reads containing joints and low-quality reads were removed to obtain 196.84 Gb of clean reads. Based on the effective sequence information, the K-mer analysis was performed using the GCE v1.0.2 software22 to estimate the genome size, heterozygosity rate, repeat sequences, and other information. The K value was set at 17 to ensure sufficient generation of K-mer species to cover the entire genome. The results of K-mer analysis showed that the estimated genome size was 2560.3 Mb, heterozygosity rate was 0.66%, proportion of repeated sequences was 73.32%, and GC content was approximately 42.75%.

Genome assembly

The Trimmomatic software was used to trim the original data to reduce the number of adapter sequences and low-quality fragments23. Long-read data of 282.97 Gb (clean reads) were assembled using the assembly software, NextDenovo24. The genome sequence was assembled after error correction and removal of redundancy. Finally, the 3D-DNA software was used to cluster and construct an interaction matrix. The Juicebox software25 was used to build the chromosome interaction map, and JuiceBox was used for visual error correction. The assembled genome was 2,627.19 Mb in size, and contained 3,109 contigs and 4.27 Mb contig N50 (Table 1). Hi-C data analysis was performed to assemble further the contigs obtained from the initial assembly to the scaffold level. Finally, 2.51 Gb of the genome sequence was obtained, the scaffold N50 was 62.20 Mb, and 95.45% of the original assembly sequence was attached to 40 pairs of chromosomes (Fig. 2a), it was noticed that there was the smallest scaffolds in chr40, only 1.12 Mb in size, which was consistent with the small super-scaffold of Coregonus sp. Balchen26. The microchromosomes were difficult to find histologically, so the chromosomal karyotype were 39 pairs (Fig. 2b). Specific information regarding the chromosomes is shown in Table 2. A chromosomal circle diagram was drawn based on the 40 constructed chromosomes (Fig. 2c) using the CIRCOS27. Our input data comprised annotated gene and ncRNA gff files, alongside the genome sequence and its masked version post-repetition sequence filtration. Circos generated a comprehensive circular plot, delineated into five concentric layers. The outermost layer represents the chromosomes, followed inwardly by gene density, repeat sequence density, and ncRNA regions—subdivided into rRNA, snRNA, and miRNA zones for detailed visualization. Due to the excessive quantity of tRNA annotations, their representation was omitted in this iteration as we are currently refining the tRNA annotation results. The innermost layer illustrates the GC content.

Table 1 Summary of genome sequencing, assembly, and annotation results.
Full size table
Fig. 2
figure 2

Characteristics of the Coregonus ussuriensis genome. (a) Hi-C intrachromosomal contact map of the C. ussuriensis genome assembly; the frequency of interactions was calculated using a window size of 500 kb. The color bar shows the contact density from low (white) to high (red). (b) Cytological karyotype map of C. ussuriensis. (c) Circos plot of the C. ussuriensis genome assembly. The tracks from outside to inside are 40 chromosome-level scaffolds; gene density; repeat density; region of ncRNAs (the first little ring is rRNA distribution, the second small ring is snRNA distribution, the third small ring is miRNA distribution); and GC content.

Full size image
Table 2 Summary information for each chromosome of Coregonus ussuriensis.
Full size table

Repeat sequence annotation

The repeat sequences of the genome were predicted using homologous prediction based on the RepBase library (http://www.girinst.org/repbase) in conjunction with de novo prediction. RepeatMasker (open-4.09)28 and RepeatProteinMask (open-4.09) were used to search for tandem repeats in the RepBase (release 21.01)29. RepeatModeler (v open 1.0.11)30 and LCR-Finder (v 1.0.5) software31 were used to create the de novo (de novo sequencing) repeat sequence database. Finally, we used the RepeatMasker (open 4.0.9) and TRF software to annotate the repeat sequences using TE and de novo libraries. After removing the overlapping non-redundant parts, the combined annotation results were used for statistical analysis. We also identified the length of DNA transposons was 672.59 Mb, the long interspersed repeated sequences (LINE) was 353.29 Mb, the short interspersed repeated sequences (SINE) was 17.57 Mb, the long terminal repeat (LTR) was 177.46 Mb. All transposable element (TE) sequences accounted for 58.63% of the whole genome sequence (Table 3). The repeat sequence density statistics are shown in Fig. 2c.

Table 3 Classification statistics of repeated sequences in the genome of Coregonus ussuriensis.
Full size table

Gene function prediction and annotation

We combined the de novo prediction, homologous annotation, and RNA-Seq methods to predict and annotate protein-coding genes in the genome of C. ussuriensis. The Augustus (v3.3), GlimmerHMM (v3.0.4), and Genscan software were used for ab initio gene prediction. For homologous annotation, we selected five closely related species, Oncorhynchus mykiss (GCF_013265735.2)32, Salmo trutta (GCF_901001165.1)33, Oncorhynchus tshawytscha (GCF_002872995-1)34, Salmo salar (GCF_000233375-1)35, and Coregonus sp. Balchen (GCA_902810595-1)26 for comparison with the genome of C. ussuriensis using the TblastN software (with an e-value cutoff of 1e-5)36. The Exonerate software (v2.2.0; parameters: -model protein2 genome -percent 20-minintron 10, -maxintron 50000)37 was used for precise homologous genome sequence splicing of the matched proteins. For RNA-seq prediction and annotation, the Trinity38 software was used to assemble RNA-seq data from six tissues, namely the heart, liver, spleen, intestine, stomach, and muscle, and the PASA software was used to predict the gene structure. A total of 43,320 protein-coding genes were predicted and annotated, with an average length of 19,815.01 bp, average coding sequence length of 1,524.46 bp for each gene, and average exon number of 8.68 (Table 4 and Fig. 3). It is worth noting that C. ussuriensis had shorter gene lengths, shorter CDSs and less exons, but the number of genes was the largest, probably because of gene redundancy and even functional differentiation after duplication39. SwissProt, TrEMBL, KEGG, InterPro, GO, NR, and other protein databases were used to annotate protein functions of the gene prediction sets. Among them, 37,987 genes were annotated to the InterPro, 29,061 genes to the GO, 42,922 genes to the KEGG_ALL, 26,747 genes to the KEGG_KO, 39,392 genes to the Swissprot, 42,973 to TrEMBL, and 43,048 to the NR. A total of 43,066 protein-coding genes were annotated in the genome of C. ussuriensis, accounting for 99.41% of the predicted genes. The annotation results for each database are presented in Table 5.

Table 4 Statistical results of the predicted genes in the genome of Coregonus ussuriensis.
Full size table
Fig. 3
figure 3

Statistical map of gene sets for gene structure prediction. (a) Exon number. (b) Exon length. (c) Intron number. (d) Intron length. (e) Gene length. (f) GC-content of genes. (g) Coding sequence (CDS) length. (h) GC-content of CDS.

Full size image
Table 5 Statistical results of gene function annotation.
Full size table

Annotation of non-coding RNA

Noncoding RNAs include tRNAs, rRNAs, miRNAs, and snRNAs. tRNAs were annotated based on their structural characteristics with tRNAscan-SE (v1.3.1)40 using the default parameters. Owing to their high conservation, the rRNAs of related species are usually selected as the reference sequences. BLASTN (v2.6.0) was used to find the rRNA sequences in the genome. The covariance model in Rfam (v14.1) was used to predict miRNA and snRNA sequences in the genome using the INFERNAL (v1.0) software41. A total of 770 miRNAs, 150,183 tRNAs, 976 rRNAs, and 1,828 snRNAs were annotated in the C. ussuriensis genome (Table 6 and Fig. 2c).

Table 6 Statistical results of noncoding RNA annotation.
Full size table

Genome collinearity analysis

The genome sequences of O. mykiss, S. salar, and Coregonus sp. Balchen were selected as references for comparison with the genome sequence of C. ussuriensis. The comparison and sequencing software used was Mummer (v4.0.0 beta2)42. The JCVI software43 was used to sequence the results and build a genomic collinear map (Fig. 4). The collinearity analysis revealed that C. ussuriensis and Coregonus sp. Balchen had the highest genomic homology with the other species. The super-scaffolds of C. ussuriensis were numbered to match with C. sp Balchen26 for consistency. The linear section indicates the large homologous fragments of the genome sequence within a species or between two species as a consequence of copy or species differentiation. The functions and sequences of genes in the homologous fragments are conserved. The MCScan software (http://chibba.agtec.uga.edu/duplication/mcscan/) was used to search for linear sections of the genome between these species. The plot figures of chromosome-by-chromosome comparison to C.sp.Balchen, O. mykiss and S. salar were shown as Figs S1-S3.

Fig. 4
figure 4

Collinearity analysis of reference genomes of Coregonus ussuriensis and other homologous species (Oncorhynchus mykiss vs. Coregonus sp. Balchen vs. Coregonus ussuriensis vs. Salmo salar).

Full size image

The Ks value (mutation rate of the homologous site) of the gene pairs in the collinear segment was calculated. The Ks value can reflect the relative species differentiation and whole-genome replication events in the evolution of a species. The peak Ks position in the species comparison reflects the whole-genome replication event of the species44. According to the Ks value distribution map, the horizontal coordinate represents the Ks value and the vertical coordinate represents the number of gene pairs. By comparing the two peaks, it was possible to distinguish between the earlier and later stages of genome-wide replication and relative species differentiation (Fig. 5). As is evident from the figure, a whole genome replication (WGD) event occurred near Ks values of 0.1246, 0.2951, 0.1305, and 0.1246 for C. ussuriensis, O. mykiss, S. salar, and Coregonus sp. Balchen, respectively. Species differentiation occurred at a Ks value of 0.1650 for Coregonus sp. Balchen and O. mykiss, whereas it occurred at a Ks value of 0.1485 for Coregonus sp. Balchen. Species differentiation between C. ussuriensis and S. salar occurred at a Ks value of 0.1645. All Ks peaks were located between 0.12 and 0.3, demonstrating that they underwent whole-genome duplication, giving rise to tetraploid genomes with salmon-specific 4 R whole-genome duplications identical to those of S. salar, O. mykiss and O. kisutch45. The duplicated count in Busco (Table 1) and the off-target HiC hits (Fig. 2a) were also indicative of residually tetraploid regions, which providing further evidence that a burst of WGD occurred.

Fig. 5
figure 5

The Ks value (mutation rate of the homologous site) distribution diagram. Cus: Coregonus ussuriensiss; Omy: Oncorhynchus mykiss; Cba: Coregonus sp. Balchen; Ssa: Salmo salar.

Full size image

Data Records

The BGI-SEQ, Pacbio, and Hi-C sequencing data used for genome assembly were deposited in the NCBI Sequence Read Archive database with accession numbers SRR2524845846, SRR2534350547, and SRR2549405448 under the BioProject accession number of PRJNA1000111.

The whole genome sequence project has been deposited at GenBank under the accession JAVHNI00000000049.

The transcriptomic sequencing data were stored under accession numbers SRR17306694-SRR1730671150.

The attachment results of BGI, Pacbio and Hi-C sequencing, gene annotation and evolutionary analysis were deposited in the Figshare51 database.

Technical Validation

Genome assembly assessment

To validate the accuracy and completeness of the genome assembly, Minimap228 (v2.5, default parameters) was first used to compare the three-generation sequencing data with the assembled genome of C. ussuriensis. The comparison rate, extent of genome coverage, and depth distribution of reads were calculated, and 94.55% of the reads were mapped to contigs, with an average sequencing depth of 84.86× and an average genome coverage of 99.91%. Bases with sequencing depths greater than 10× accounted for 99.14% and those with sequencing depths greater than 20× accounted for 98.51% of the total number of bases. We assessed the integrity of the genome based on single-copy homologous genes from the OrthoDB set using the BUSCO software (v3.0.2) (https://gitlab.com/ezlab/busco), and the vertebrata_odb9 gene sets were chosen for BUSCO assessment. A total of 2,457 genes were compared using BUSCO (95.00%, complete), of which 1,306 were single-copy genes (50.50%, complete and single-copy), 1,151 were duplicated genes (44.50%, complete and duplicated), 31 were partially duplicated genes (1.20%, fragmented), and 98 were not duplicated (3.80%, missing). The genome assembly and annotation are summarised in Table 1. These results indicated that the genome assembly of C. ussuriensis was complete and of high quality.

Karyotype analysis of C. ussuriensis

To verify the correctness of the C. ussuriensis genome assembly using Hi-C data, we fixed and stained the chromosomes and confirmed the chromosome number. For karyotyping, C. ussuriensis (n = 20) was raised at 18 °C for a week before sampling. The body weight of the experimental fish was 30 ± 0.8 g. Our preliminary experiments obtained the best chromosome fixation with phytohemagglutinin (PHA) and colchicine injected under the left pectoral fin. Sample preparation: Ten micrograms of PHA per gram of fish was injected; colchicine (3 mg/g of fish) was injected 24 h after the PHA injection. The samples were collected 4 h after colchicine injection.

The gill arches were cut and placed in water. The kidney tissue (whole kidney) was then collected immediately, washed two or three times with normal saline (85% NaCl solution), cut into pieces, and placed in a 10 mL beaker (approximately 8 mL of normal saline was added to the beaker). Cell low-osmosis: The filtrate obtained after filtration through 100 mesh gauze was placed in a 10 mL tube and centrifuged at 1200 rpm for 8 min. The supernatant was discarded, the pellet was gently dislodged, and the cells were incubated for 50 min in 6 mL of a hypotonic solution (0.075 mol/L KCl solution). Chromosome fixation: The above samples were mixed with 500 μL of a fixing liquid (methanol: glacial acetic acid = 3:1), centrifuged at 1200 rpm for 8 min, and the supernatant was discarded. After gently dislodging the pellet, 6 mL of the fixing solution was added, and the chromosome fixation was allowed to proceed for 20 minutes; this procedure was repeated three times. Drop slides and staining: The sample liquid (3¬–8 drops) was dropped onto a slide from a height and spread on the slide by gently blowing the liquid. The slide was passed over an alcohol lamp until the liquid almost dried, after which it was air dried with the side having the cells placed upright. Finally, the slide was placed with cells facing down on a staining plate for 30 min, rinsed with tap water, dried in air, and observed under a light microscope. By analysing with the Argus software, the microchromosomes were not visible under ordinary microscopes, so the karyotype analysis showed 39 pairs of chromosomes.