Whole-Genome Resequencing and Pan-Transcriptome Reconstruction Highlight the Impact of Genomic Structural Variation on Secondary Metabolite Gene Clusters in the Grapevine Esca Pathogen Phaeoacremonium minimum

Abstract

The Ascomycete fungus Phaeoacremonium minimum is one of the primary causal agents of Esca, a widespread and damaging grapevine trunk disease. Variation in virulence among Pm. minimum isolates has been reported, but the underlying genetic basis of the phenotypic variability remains unknown. The goal of this study was to characterize intraspecific genetic diversity and explore its potential impact on virulence functions associated with secondary metabolism, cellular transport, and cell wall decomposition. We generated a chromosome-scale genome assembly, using single molecule real-time sequencing, and resequenced the genomes and transcriptomes of multiple isolates to identify sequence and structural polymorphisms. Numerous insertion and deletion events were found for a total of about 1 Mbp in each isolate. Structural variation in this extremely gene dense genome frequently caused presence/absence polymorphisms of multiple adjacent genes, mostly belonging to biosynthetic clusters associated with secondary metabolism. Because of the observed intraspecific diversity in gene content due to structural variation we concluded that a transcriptome reference developed from a single isolate is insufficient to represent the virulence factor repertoire of the species. We therefore compiled a pan-transcriptome reference of Pm. minimum comprising a non-redundant set of 15,245 protein-coding sequences. Using naturally infected field samples expressing Esca symptoms, we demonstrated that mapping of meta-transcriptomics data on a multi-species reference that included the Pm. minimum pan-transcriptome allows the profiling of an expanded set of virulence factors, including variable genes associated with secondary metabolism and cellular transport.

Keywords: Esca; comparative genomics; intraspecific genetic diversity; pan-transcriptome; pathogenomics; secondary metabolism; structural variation.

FIGURE 1

Esca symptoms in grapevine and phylogenetic relation of the five Phaeoacremonium minimum isolates used in this study. (A) Typical foliar symptoms of Esca in a red grape cultivar, (B) berry spotting (measles) and (C) black streaking (arrows) caused by wood colonization of Esca pathogens. (D) Dendrogram illustrating that Pm1119 and Pm448 clustered with Pm1118 and Pm449, respectively, reflecting their geographic origins. The dendrogram was constructed with MEGA7 (Kumar et al., 2016) using the Neighbor-Joining method (Saitou and Nei, 1987) and was based on a total of 22,242,282 positions. Bootstrap confidence values (100 replicates) are shown next to the branches (Felsenstein, 1985). The evolutionary distances were computed using the Maximum Composite Likelihood method (Tamura et al., 2004) and are in the units of the number of base substitutions per site. The tree was visualized with FigTree v.1.4.3 (http://tree.bio.ed.ac.uk/software/figtree). Insets show images of in vitro cultures of the five isolates.

FIGURE 2

Circular representation of the genome of Pm1119. Different tracks denote: (a) percentage of GC content; (b) repeat density; (c–g) mapping coverage of short Illumina reads: (c) Pm1119, (d) UCR-PA7, (e) Pm1118, (f) Pm448, (g) Pm449; (h) SNP density; (i) omega values; (j) CDS density; (k) BGCs (known classes in red, putative classes in black); (l) Number of genes entirely deleted in UCR-PA7, Pm1118, Pm448, or Pm449; (m) Number of expressed protein-coding genes inserted in UCR-PA7, Pm1118, or Pm449. Figure was prepared using the OmicCircos Bioconductor package (Hu et al., 2014).

FIGURE 3

(A) Venn diagrams showing the overlap between results of the three methods used to call structural variants (SVs) sites. (B) Size distribution of the deleted genomic regions in the four isolates compared to Pm1119. (C) Example of a deletion event occurring in UCR-PA7, Pm448, and Pm449 relative to Pm1118 and Pm1119, which affected the composition of a BGC associated with polyketide synthesis. Arrows represent genes coding for core biosynthetic genes (red), additional biosynthetic genes (pink), P450s (blue), cellular transporters (yellow), and FAD-binding proteins (orange). Gray arrows correspond to genes predicted to be part of the biosynthetic gene cluster (BGC), but with other annotations.

FIGURE 4

Size and amount of detected indel events encompassing protein-coding genes. The bar plot shows the number of genes and the size of the gene clusters in structural variant sites as well as the proportion of genes associated with BGCs. Venn diagrams show the overlap between genes in structural variant sites detected in the different Pm. minimum isolates.

FIGURE 5

Examples of indels involving co-expressed gene clusters. In (A), the whole co-expressed gene cluster BGC_187 identified in UCR-PA7 is deleted in Pm1119. In (B), the deletion of adjacent co-expressed genes did not alter the co-expression pattern of the genes flanking the structural variant site. Asterisks identify genes significantly differentially expressed in vitro between rotating and stationary cultures. Arrows represent genes coding for core biosynthetic genes (red), additional biosynthetic genes (pink), P450s (blue), cellular transporters (yellow), and transcription factors (green). Gray arrows correspond to genes predicted to be part of the BGC, but with other annotations.

FIGURE 6

Mapping results of metatranscriptomics data on a multi-species reference including different Pm. minimum transcriptome references. (A) Stacked bars show the counts of metatransciptomics reads aligned to the transcriptomes of each trunk pathogen included in the multi-species reference using for Pm. minimum, from left to right, the transcriptomes of UCR-PA7 (Blanco-Ulate et al., 2013), Pm1119, or the pan-transcriptome. (B) Percentage of the core and variable transcripts detected in each sample when using Pm. minimum pan-transcriptome in the multi-species reference transcriptome. Pie charts represent the composition of functional categories among the variable transcripts detected in each sample. RNAseq data from Pm1119 in vitro cultures were included to determine the level of non-specific mapping on variable genes. BGCs, biosynthetic gene clusters; CWDEs, cell wall-degrading enzymes; P450s, cytochromes P450; Pm., Phaeoacremonium; Pha., Phaeomoniella; Dia., Diaporthe; E., Eutypa; Dip., Diplodia; F., Fomitiporia; B. Botryosphaeria; I., Ilyonectria. App. healthy, apparently healthy plant without trunk-disease symptoms.