Background

Eucalyptus, belonging to the Myrtaceae family, grows rapidly, is widely cultivated worldwide (over 20 million hm2) [1], and has a high economic value because of its wide range of uses. Introduced to China more than 120 years ago, eucalyptus provides an annual wood output of more than 40 million m3 and more than 50% of China’s pulping materials [2].

Eucalyptus urophylla is a hybrid species with rapid growth and strong stress resistance among eucalyptus trees. It is the dominant species in artificial forestation in China and occupies the largest area.

E. urophylla × E. grandis is a successful hybrid species of Eucalyptus and is the leading variety of artificial afforestation, occupying the largest area in China with rapid growth and strong stress resistance. The transfer of these excellent economic and agronomic traits to other Eucalyptus species via interspecific hybridization is important for broadening the genetic diversity of eucalyptus. However, recent research on E. urophylla × E. grandis has focused on wood processing, with few studies exploring the genomic biology of these superior qualities. To compensate for this deficiency and overcome these obstacles in basic and applied biology research, it seems essential to assemble a high-quality reference genome of E. urophylla × E. grandis.

Comparative genomics provides important information on evolutionary biology and useful information on interspecific genomic differences [3,4,5]. For example, identifying conserved genome structures, inferring common ancestry, and analyzing genomic similarities and differences are important for evolutionary genetics and the transmission of genetic information [6]. The increasing number of genomic resources allows us to use comparative genomics to gain new insights into the evolutionary variation of individual genes and gene families [7, 8] in whole genomes [9,10,11]. Eucalyptus belongs to the family Myrtaceae and includes more than 700 different species [12, 13], most of which belong to the Eucalyptus subgenus [14, 15]. Genomic research on eucalyptus is relatively underdeveloped. The reference genome of E. grandis was published only in 2014 [16], which has greatly hindered genomics and population genetics research on eucalyptus. Therefore, little research has been conducted on eucalyptus in this area.

Here, we reveal the high-quality reference genome of E. urophylla × E. grandis using a combination of Illumina sequencing, PacBio HiFi sequencing, and Hi-C sequencing platforms with a size of 545.75 Mb, containing 34,502 protein-coding genes. Repetitive elements occupied 58.56% of the genome. Comparative genomic analyses revealed that E. urophylla × E. grandis had recently undergone a WGD event and large-scale structural variation among the 34 Eucalyptus accessions. The results presented in this study provide a foundation for E. urophylla × E. grandis genomic studies seeking to affirm the genetic variation, genome evolution, and genealogical structure of Eucalyptus as well as for breeding studies for the genetic improvement of E. urophylla × E. grandis.

Results

The high-quality E. urophylla × E. grandis genome

We obtained 78.13 gigabases (Gb ~ 142.31-fold) Illumina paired-end sequences (Supplementary Table 1), which indicated that the estimated E. urophylla × E. grandis genome size was 549 Mb, with 2.16% heterozygosity, using K-mer analysis (Supplementary Table 2). We sequenced E. urophylla × E. grandis by generating 28.16 gigabases (Gb) PacBio high-fidelity (HiFi) data, yielding a preliminary genome assembly of 545.93 Mb (contig N50, 39.94 Mb) with a GC content of 39.89% (Table 1 and Supplementary Table 1). To obtain a high-quality genome of E. urophylla × E. grandis, a total of 59.32 Gb of data were obtained with 108.69-fold genome coverage by Hi-C sequencing, which was used to construct a chromosome interaction heatmap. The total number of scaffolds was 209, the longest being 68 Mb. Subsequently, the final assembly captured a 545.75 Mb genome sequence (scaffold N50, 51.62 Mb) containing 34,502 protein-coding genes with 58.56% of repeat sequences. According to the Hi-C contact maps, 98.29% of the entire genome was organized and divided into 11 chromosomes (Fig. 1a; Table 1). The longest chromosome was 64.4 Mb (Chr8) and the shortest was 38.5 Mb (Chr4) (Supplementary Table 3). The mean long terminal repeat (LTR) assembly index (LAI) score was 18.51. The genetic region assembly integrity of the highly conserved core proteins was supported by 98.4% (1588) (Table 1 and Supplementary Table 4) using the Benchmarking Universal Single-Copy Orthologs (BUSCO) analysis. This further confirms the integrity and high quality of the genome assembly.

Table 1 Summary of Eucalyptus urophylla × E. grandis genome assembly
Full size table
Fig. 1
figure 1

The global landscape of E. urophylla × E. grandis. (a) Hi-C interaction heatmap of E. urophylla × E. grandis. (b) Genomic landscapes of E. urophylla x E. grandis. (I–VII are chromosomes, different types of TE, TE heatmap (500 kb), gene density (500 kb), GC content heatmap (500 kb), different types of noncoding RNAs, and gene pairs of E. urophylla × E. grandis

Full size image

The genomic features of E. urophylla × E. grandis are shown in Fig. 1. Based on the highly contiguous E. urophylla × E. grandis genome, 34,502 protein-coding genes were identified, which was lesser than the number in E. grandis (36,349). The average gene and coding sequence (CDS) lengths were 5,148 bp and 1,218 bp, respectively. The average exon and intron lengths were 306.77 and 813.98 bp, respectively (Supplementary Fig. 1). Among the 34,502 predicted genes, 31,241 (90.55%) were functionally annotated from 10 known databases: InterPro (23,485 genes, 68.07%), GO (Gene Ontology; 16,385 genes, 47.49%), KEGG_ALL (30,405 genes, 88.13%), KEGG_KO (11,116 genes, 32.22%), Swiss-Prot (20,220 genes, 58.61%), TrEMBL (30,391 genes, 88.08%), TF (1,840 genes, 5.33%), Pfam (22,760 genes, 65.97%), NR (30,970 genes, 89.76%), and KOG (23,839 genes, 69.09%) (Supplementary Table 5). Among the predicted repetitive elements, 39.4% were long terminal repeats (LTRs), 14.00% were DNA transposons, and 5.98% were long interspersed nuclear elements (LINEs) (Supplementary Tables 7 and Supplementary Fig. 2). We identified 828 miRNAs, 9,657 rRNAs, 411 snRNAs, and 450 tRNAs (Supplementary Tables 8, Fig. 1bVI). The GC content was unevenly distributed (Fig. 1bV).

Evolutionary analysis of the E. urophylla × E. grandis genome

To investigate the E. urophylla × E. grandis genome evolution, we collected 12 other representative plant species genomes using comparative genomic analyses and identified the gene families, which revealed that there were 16,280 gene families and 3050 species-specific single-copy genes (Supplement Table 9). In addition, clustering analyses revealed that there were 27,032 genes in the families, accounting for 78.3% of the predicted genes in E. urophylla × E. grandis, which was similar to the % of predicted genes in E. grandis (Supplementary Table 9). Based on gene family analyses, we constructed a phylogenetic tree with 652 shared single-copy orthologs of the 13 species, which indicated that E. urophylla × E. grandis was most closely related to E. grandis, comprising a monophyletic group and specialized approximately 0.00086 million years ago (Mya), whereas the divergence time estimated between Eucalyptus and P. granatum was approximately 63.2 Mya (Fig. 2a).

Fig. 2
figure 2

The E. urophylla × E. grandis genome evolution. (a) Phylogenetic tree construction and divergence time estimation for E. urophylla × E. grandis and 12 representative plants. (b) The gene family expansion (green), contraction (red), and gene copy number distribution. (c) Venn diagram showing the gene family clusters in E. urophylla × E. grandis, M. domestica, A. thaliana, E. grandis, and C. papaya. (d) Ks distribution in E. grandis, P. granatum, and E. urophylla × E. grandis

Full size image

Comparative evolutionary analysis of the gene families in 13 plant species showed that 341 gene families of E. urophylla × E. grandis showed significant expansion relative to the gene families of the most recent common ancestor, whereas 767 showed significant contraction (p < 0.01, Fig. 2b). We found that E. urophylla × E. grandis had fewer gene family expansions and more gene family contractions than the other Myrtaceae species (Fig. 2b), which is consistent with the lower gene number. We identified 4958 single-copy genes in E. urophylla × E. grandis via clustering analyses, accounting for 14.37%, which was similar to that in E. grandis (Fig. 2b). Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) showed that the expanded gene families were involved in ion binding, plant–pathogen interactions, carbohydrate derivative binding, flavonoid biosynthesis, starch and sucrose metabolism, and pentose and glucuronate interconversion (Supplementary Figs. 3 and 4). However, functional analysis of the contracted gene families revealed that they were involved in the Toll-like receptor signaling pathway, NF-kappa B signaling pathway, cell recognition, protein modification process, ion binding, MAPK signaling pathway, and phenylpropanoid biosynthesis (Supplementary Figs. 5 and 6). The results of gene family clustering showed that E. urophylla × E. grandis and E. grandis shared more gene families than M. domestica, A. thaliana, and C. papaya, which was consistent with their phylogenetic relationships (Fig. 2b, c).

To estimate WGD events, E. grandis and P. granatum were selected and their synonymous nucleotide substitutions (Ks) were characterized. In addition, we found that a WGD event recently occurred in E. urophylla × E. grandis genome after its divergence from P. granatum with the in-depth comparison genomic analyses (Fig. 2d). E. grandis and P. granatum also underwent genome-wide replication following divergence. Furthermore, based on the Ka/Ks ratios, we found that 113 candidate genes were under strict positive selection in E. urophylla x E. grandis (p < 0.05). GO and KEGG enrichment analysis showed that the positive selection genes were enriched in “DNA metabolic process,” “Protein-containing complex assembly,” “Cellular response to DNA damage stimulus,” “Fanconi anemia pathway,” “DNA replication,” “Cholesterol metabolism,” “Homologous recombination,” “Cell cycle,” and “Steroid biosynthesis,” indicating that they may improve DNA damage resistance and the related metabolic pathways in adverse environments (Supplementary Figs. 7 and 8).

Comparative genomic analysis of Eucalyptus species

To understand phylogenetic relationships, 34 Eucalyptus accessions were collected (Supplementary Table 10). A phylogenetic tree was constructed using 798,492 high-quality SNPs, which revealed that the tree was divided into six branches. The phylogenetic relationship between Eucalyptus species showed that E. virginea (VIR) and E. decipiens (DEC) were clearly separated (Fig. 3). E. globulus (GLO) and E. viminalis (VIM) were the closest relatives. E. albens (ALB) and E. polyanthemos (POL) belonged to the same evolutionary branch. E. curtisi (CUR) and E. tenuipes (TEN) were closely related. E. regnans (REG) and E. pauciflora (PAU) belonged to the same evolutionary clade. To explore their evolutionary relationships, genomic synteny analyses of E. urophylla × E. grandis (EUC) and the other 30 Eucalyptus species were performed, which exhibited high levels of genomic synteny (Fig. 4). Interestingly, the comparative genome structure of EUC, E. grandis, and E. globulus (GLO) showed higher collinearity, indicating no large-scale structural variation after divergence, which was consistent with their evolutionary relationships (Fig. 3 and Supplementary Fig. 8). Interestingly, we found that Chr9 showed large structural variations among EUC, E. fibrosa (FBI), E. brandiana (BRA), E lansdowneana (LAN), E. regnans (REG), E. pumila (PUM), E. shirleyi (SHI), E. polyanthemos (POL), and E. victrix (VIC) according to the synteny analysis shown in Fig. 4. Large structural variations in chromosomes Chr2, Chr4, and Chr6 were found between EUC and E. erythrocorys (ERY) (Fig. 4). EUC and E. guilfoylei (GUI) showed large chromosomal rearrangements in Chr2 and Chr6 (Fig. 4). Chr3 and Chr8 showed large inversions between EUC and E. paniculata (PAN) (Fig. 4).

Fig. 3
figure 3

The SNP-based phylogeny of 34 Eucalyptus species

Full size image
Fig. 4
figure 4

Comparative genomes resolve genome synteny in 31 eucalyptus species

Full size image

Discussion

The high quality E. urophylla × E. grandis genome provided community resources for Eucalyptus genetic breeding research

Myrtaceae is the 8th largest flowering plant family and is mainly grown in the subtropics and tropics, with 5950 species in 132 genera [17]. Eucalyptus is the most planted hardwood species worldwide and a member of the Myrtaceae family, showing unique economic value as a global renewable energy resource [16]. However, studies on E. urophylla × E. grandis genomics and molecular levels are limited and almost nonexistent. Resolving the genomic resources of E. urophylla × E. grandis is, therefore, of great significance, because it can promote eucalyptus evolutionary studies and molecular breeding. With the development of long-read sequencing techniques, complex heterozygous genomes have been successfully assembled [18,19,20]. Here, we completed a high-quality E. urophylla × E. grandis reference genome, which is an economically valuable source of natural products and accelerates the application for molecular breeding, evolution, and genetics of Myrtaceae.

We generated a high-quality genome assembly of 545.75 Mb for E. urophylla × E. grandis, which was smaller than that of E. grandis (605 Mb) [16]. The genome scaffold N50 was 51.62 Mb, which is larger than that of other Myrtaceae family, such as Psidium guajava (443.8 Mb, N50 40.4 Mb) [21]. We predicted 34 and 502 protein-coding genes. This number was higher than that in Psidium guajava [21] and lower than that in E. grandis [16]. As an important component of genome composition, repetitive sequences play crucial roles in chromosomal rearrangement, gene regulation, and genome evolution, but also have significantly affect high-quality genome assembly. Here, we found 58.56% repeated sequences in the E. urophylla × E. grandis genome, with 39.4% LTRs, which was higher than those in E. grandis and Psidium guajava. Furthermore, BUSCO analysis revealed that 98.4% of the highly conserved core proteins supported the assembly integrity of genetic regions, indicating that the reference genome quality was superior to that of the guava genome (95.7%) [21]. Overall, the assembled E. urophylla × E. grandis genome in this study was complete and accurate, providing valuable genomic resources for subsequent studies on Eucalyptus population evolution and genetic improvement.

Phylogenetic analysis contributed to the evolutionary relationship

To explore the phylogenetic analyses, 13 other genomes of representative plant species were selected, which showed E. urophylla × E. grandis were most closely related to E. grandis, supporting the placement of E. urophylla × E. grandis and E. grandis in Myrtaceae, and P. granatum in the order Myrtales. WGD events can cause gene family expansion, chromosomal rearrangement, genome size variation, and species evolution [22]. Gene family analysis was performed for E. urophylla × E. grandis, E. grandis, Arabidopsis, M. domestica, and C. papaya, which revealed 8882 common gene families. In contrast, the E. urophylla × E. grandis genome had unique 850 gene families, which was more than those of E. grandis and less than those of M. domestica, Arabidopsis, and C. papaya. The Ks analysis revealed that E. urophylla × E. grandis shared a recent WGD event with E. grandis and P. granatum (Fig. 2d). The chromosomal regions of E. urophylla × E. grandis showed a one-to-one correspondence with E. grandis (Fig. 4), which is consistent with an evolutionary relationship (Fig. 3), possibly because they did not have large-scale structural variation after species divergence (Fig. 4).

Because genome research on eucalyptus is still relatively lacking, there is some controversy regarding the classification of eucalyptus. To understand phylogenetic relationships, 34 Eucalyptus species accessions were collected (Supplementary Table 10). A phylogenetic tree was constructed using 798,492 high-quality SNPs, which showed that the phylogenetic tree was divided into six branches (Fig. 3). The phylogenetic relationship between eucalyptus species showed that VIR and DEC were clearly separated. GLO and VIM were the closest relatives. ALB and POL belonged to the same evolutionary branch. Similarly, CUR and TEN were closely related. REG and PAU belonged to the same evolutionary clade. These results are consistent with those of previous studies [23] and contribute to our understanding of the evolutionary relationships between different Eucalyptus species at the genome level.

Comparative genomics reveals interspecific structural variation in Eucalyptus

Studying structural variations (SVs) is a challenging yet important for understanding trait differences in highly repetitive genomes as well as an important component of genetic diversity, and has important implications in evolution and breeding. However, owing to the relative lag in eucalyptus genomic research, eucalyptus research remains in its infancy. Genomic synteny analyses showed that our genome assembly of E. urophylla × E. grandis had a high level of genome synteny with other Eucalyptus species. Interestingly, the comparative genome structures of EUC, GRA, GLO, and E. viminalis (VIM) showed higher collinearity, indicating no large-scale structural variation after divergence, which was also consistent with their evolutionary relationships (Fig. 3 and Supplementary Fig. 8). Structural variation was weighed more heavily for genomic and trait effects. Interestingly, we found that chromosomes Chr2, Chr3, Chr6, Chr8, and Chr9 showed large structural variations, which may be an important reason underlying the differentiation of these eucalyptus species. This lays the foundation for our follow-up research on important characteristics such as material and stress resistance.

Conclusions

Overall, in this study, we assembled 545.75 Mb of the E. urophylla × E. grandis high-quality genome and annotated 34,502 protein-coding genes, which demonstrated a complete genome landscape. Comparative genomic analysis revealed that E. urophylla × E. grandis underwent a recent WGD event. We characterized the phylogenetic relationships of 34 eucalypts at the genome-wide level for the first time. Interspecific structural variations were identified using genomic collinearity analysis. This will accelerate the application of molecular genetic breeding of E. urophylla × E. grandis, deepen our understanding of eucalyptus biology and genetic improvement of eucalyptus, and lay the foundation for population genome research.

Materials and methods

Library construction and sequencing of E. urophylla × E. grandis

Young leaves of E. urophylla × E. grandis plants were collected and immediately frozen in liquid nitrogen. We extracted DNA from fresh young leaves for Illumina and PacBio sequencing with a QIAGEN® Genomic Kit (QIAGEN, Germany). Sequencing libraries for Illumina, PacBio HiFi, and Hi-C were constructed.

Genome assembly and correction

K-mer analysis was used to estimate genomic size, heterozygosity, and repeats by plotting the 17-mer depth distribution [24, 25]. HiFi data were assembled with the Hifiasm software using default parameters [26]. Juicer (v.1.5) [27], a 3D-DNA scaffolding pipeline [28], Juicebox (v.1.11.08) [29], and HiCUP [30] were used to correct the initial orientations for genome assembly. BUSCO analyses were used to evaluate the completeness of the reference assembly (https://busco.ezlab.org/frame_wget.html) [31]. Genome assembly continuity was evaluated using the LTR assembly index (LAI) [32], as described by Ou et al. [29].

Repetitive sequence and noncoding RNA annotation

Complementary methods were used to identify repetitive sequences as described by Shen et al. [4]. Tandem Repeats Finder (v4.09) software [33] was used to identify tandem repeats. TEs were predicted using a complementary strategy with RepeatMasker (v4.06), LTR_FINDER (v1.05) [34], RepeatProteinMasker, RepeatScout (v1.05) [35], and RepeatModeler (v1.05). Eventually, we obtained a non-redundant genome. The miRNAs, tRNA, rRNA, and snRNAs were annotated. The tRNAs were identified using tRNAscan-SE (v1.3.1) [36]. The rRNAs were predicted using BLAST [37]. The Rfam database (release 13.0) [38] was used to search for snRNAs and miRNAs using Infernal (v1.1) [39].

Prediction and annotation of E. urophylla × E. grandis

Protein-coding gene prediction was performed using three independent methods as previously described [4]. Five programs were used to conduct de novo prediction: GlimmerHMM (v3.0.4) [40], Augustus (v3.2.1) [41], Genscan (v1.0) [42], GeneID (v1.4.4) [43], and SNAP [44]. Five representative species (R. argentea, P. granatum [45], C. citriodora [46], S. oleosum, and E. grandis [47]) were used for homolog-based predictions using the GeMoMa [48] software. Transcriptome assembly prediction was performed using the Hisat [49] and Stringtie [50]. Next, we integrated the gene sets into a non-redundant with the MAKER2 [51]. Functional annotations were performed using InterProScan [52], TrEMBL [53], NCBI-NR (V2013), SwissProt [53], Pfam (http://pfam.xfam.org), GO [54], KOG, and KEGG [55] (Table S3). We obtained a reliable set of gene annotations by integrating 1549 (96.00%) complete BUSCO results.

Gene family identification and evolutionary analysis

We used OrthoMCL (v2.0.9) to obtain gene clusters from E. urophylla × E. grandis and 12 other species genomes obtained from the phytozome database (https://data.jgi.doe.gov/), including A. thaliana (TAIR10), C. papaya (ASGPBv0.4), E. grandis (v2.0), M. domestica (HFTH1), O. sativa (v7.0), P. trichocarpa (v4.1), P. persica (v2.1), P. granatum (GCF_007655135.1_ASM765513v2), S. lycopersicum (ITAG4.1), T. cacao (v2.1), V. vinifera (v2.1), and Z. jujuba (GCF_000826755.1_ZizJuj_1.1) with default parameters. Subsequently, we obtained gene families (Supplementary Table S9) and generated a shared gene family subset for five species (E. urophylla × E. grandis, A. thaliana, M. domestica, E. grandis, and C. papaya) (Fig. 2c). The divergence times was estimated to use the MCMCTree program (v4.9) [56] with the parameters “the clock = 3 and model = 0”. CAFÉ (v4.2) [57] was used to identify gene family expansion and contraction. The results of gene family expansions or contractions were obtained as described by Chen et al. [20]. WGD analysis was performed using BLASTP (E-value = 1 × 10− 10), ML estimation in CODEML (v4.9) of the PAML software [56], and Multi-tAxon Paleopolyploidy Search software [58] based on age distributions [20, 59] for the three selected Myrtales species: E. grandis, P. granatum, and E. urophylla × E. grandis. Finally, the WGD ages described by Chen et al. [20] were estimated. We first used MUMMER [60] to call SNPs, then used a script to convert the output format of the MUMMER software to VCF, and used bcftools (http://github.com/samtools/bcftools) to merge. A neighbor-joining phylogeny was constructed using VCF2Dis (https://github.com/BGI-shenzhen/VCF2Dis) based on the P-distance matrix. Finally, the matrix file was uploaded to a website (http://www.atgc-montpellier.fr/fastme/) to obtain the tree file. Comparative genomic analyses were performed using MUMmer (V4.0) [60] as described by Shen et al. [4].