VPI-MD: a multi-omics database for Verticillium–plant interaction

Abstract

Verticillium wilt (VW), caused by Verticillium species, is one of the most severe soil-borne plant diseases globally and has been ranked as one of the world's top quarantine concerns. Over 400 plant species including crops are severely affected by VW which results in significantly reduction of yield and quality, causing a substantial economic loss of billions of dollars worldwide. Controlling key pathogens of Verticillium species is always a significant challenge and hot concern (Chen et al., 2021; Wang et al., 2020). Elucidating the mechanisms of pathogen–host interactions provides the foundation and effective reference to govern the fungus. Recent attentions on high-throughput omics technologies offer crucial basis and excellent tools for investigating the control of VW. Comparative genomics analysis revealed lineage-specific regions (LSRs) contain potential key genes for host adaptation (Zhang et al., 2019). Transcriptomic studies discovered that RVE2 could enhance cotton resistance to VW by regulating jasmonic acid signalling (Liu et al., 2023). Secretomic evidence identified the V. dahliae effector protein VP2 that could trigger hypersensitive response and cell death in tobacco, Arabidopsis and cotton (Qiu et al., 2024). Epigenomic reports explored how epigenetic conformation changes within V. dahliae affect the evolution and transcriptional regulation (Torres et al., 2024).

Despite the wealth of sequencing data for both Verticillium spp. and the host plants, however, there still exists significant limitations for high-effective utilization of these data sets as the non-direct-usable raw data formats and different standards and types of public databases. Thus, it is necessary and urgent to establish a unified platform to high-efficiently facilitate these omics data. Herein, we developed a multi-omics database for Verticillium spp. and their host–plant interactions (VPI-MD, https://www.bic.ac.cn/VPI-MD/). The database integrates data collection, organization, analysis, visualization and download functionalities (Figure 1A) and covers over 3000 multi-omics samples containing all existing data from 104 Verticillium spp. and related fungi and 14 host plants as well as 12 sets (316 samples) of our newly submitted multi-omics data for V. dahliae V592 and sea-island cotton. These data sets have a volume of totally over 6.5T and include multiple types of omics data of genomic (119), transcriptomic (1382), proteomic (36), variomic (1551), metabolomic (54) and epigenetic (271) information as well as all previously identified VW-related genes of 242 fungal and 565 host plant genes. The database offers 14 analytical tools for processing these multi-omics data (Figure 1A) and supports user-uploaded omics data for differential expression (DE) (Figure 1B,a) and GO/KEGG (Figure 1B,f) analysis. VPI-MD offers online searching of genes of desired fungus or host plants and personalized analysis via the online tools, including Gene expression profile to compare the expression profiles (Figure 1B,c), Gene fishing to identify the co-expression genes (Figure 1B,g) and Gene map viewer to visualize the genomic distributions (Figure 1B,h). It also supplies the functionalities containing Multiple sequence alignment to identify conserved domains or motifs (Figure 1B,d), Primer design for experimental verification, BLAST for cross-species comparison, GO/KEGG for enrichment analysis (Figure 1B,f), Protein–protein interaction to investigate possible interactions (Figure 1B,b) as well as Methyl view (Figure 1B,e) and IGV browser (Figure 1B,i) for visualization analysis of epigenetic modifications and genome-wide variations.

Regarding the underutilized research in Verticillium spp. and the host plants (Hultman et al., 2015), we performed an example to utilize the VPI-MD for an integrated transcriptomic and proteomic analysis of V. dahliae strain V592 incubated with different resistant sea-island cotton cultivars of resistant Xinhai21 (V592^−XH21) and susceptible Xinhai7 (V592^−XH7) (Figure 1C,a). DE and GO/KEGG analysis showed that several heat shock protein (HSP) genes were significantly down-regulated in V592^−XH21 compared to V592^−XH7 (Figure 1C,b–e). The proteomic data indicated four HSPs and a transcription factor HSF_type SFL1 were significantly down-regulated in V592^−XH21 (Figure 1C,f,g). The interaction between VdSFL1 and VdHSP90 promoters was further validated through yeast one-hybrid and dual-luciferase reporter assays (Figure 1C,h,i). The pathogenicity of VdSFL1 knockout mutant VdSFL1-KO of V. dahliae towards cotton plants was significantly reduced (Figure 1C,j). These results suggest that the resistant cultivar XH21 might reduce the pathogenicity of V592 by down-regulating VdSFL1 and VdHSP genes. This example demonstrates the convenience and potential of the VPI-MD in integrating multi-omics data to quickly generate reliable output.

LSRs in fungi often harbour crucial genes for unique biological functions (Dodds, 2010). In this study, we compiled 13 genomes predicted to contain coding genes of V. dahliae for a pan-genome analysis (Figure 1D,a). The result showed a stabilized tendency of the core gene family along with the increase of genomes (Figure 1D,b,c). High-density dispensable gene and low-density core gene regions on ctg5 and ctg6 were identified as LSR1-V592 and LSR2-V592, respectively (Figure 1D,d). Sequence alignment confirmed that LSR1-V592 is a previously identified LSR named G-LSR2 in Vd991, which is horizontally transferred from the fungus F. oxysporum f. sp. vasinfectum (Zhang et al., 2019), whereas LSR2-V592 is a novel region identified by the VPI-MD (Figure 1D,e). Analyses of density distribution and methyl viewer of 5-methylcytosine (5mC) at CpG sites across the genomes indicated that LSR2-V592 is absent in the JR2 strain (Figure 1D,d,f,g), corroborating the authenticity of our results. Single-copy gene phylogenetic tree analysis indicated a high degree of relatedness among V. dahliae strains (Figure 1D,h). While a BLASTP search of genes in this region showed that the strains with the highest homology are predominantly in the Fusarium spp., with homologues found only in a few V. dahliae strains. Further analysis displayed an identical distribution between LSR2-V592 and LSR1-V592 in V. dahliae (Figure 1D,i). These results suggest that this segment likely originated from Fusarium spp. Through a comprehensive sequence analysis, we found that the LSR2-V592 displays optimal collinear alignment related to its counterpart sequence within the F. keratoplasticum LHS11.1 strain (Figure 1D,j). These results suggest that LSR2-V592 may derive from the horizontal transfer of F. keratoplasticum LHS11.1. This example demonstrates a quick and accurate discovery of data mining and provides effective guidance for further research.

In summary, we for the first time developed the VPI-MD database for research of Verticillium spp. and their host plants. The VPI-MD integrates 14 online tools for multi-omics data mining and presentation, coupled with two examples for database application. Regular update and improvement for the database are also expected in the future. The VPI-MD will play a significant role in the prevention and treatment of VW and make an effective contribution to the improvement of agriculture and related industries.

The authors declare no conflict of interest.

H.L. and X.J. conceived and designed the study. F.W., R.T., K.C., Q.X., S.X., A.I. and Z.M. collected the data. S.S., M.L. and C.J. processed the data and built the database and website. S.S., Y.C. and X.Z. performed the experiments. F.W., X.J. and H.L. directed the project. S.S. and H.L. wrote the paper.