Genomic Investigation of the Strawberry Pathogen Phytophthora fragariae Indicates Pathogenicity Is Associated With Transcriptional Variation in Three Key Races

Abstract

The oomycete Phytophthora fragariae is a highly destructive pathogen of cultivated strawberry (Fragaria × ananassa), causing the root rotting disease, "red core". The host-pathogen interaction has a well described gene-for-gene resistance relationship, but to date neither candidate avirulence nor resistance genes have been identified. We sequenced a set of American, Canadian, and United Kingdom isolates of known race type, along with three representatives of the closely related pathogen of the raspberry (Rubus idaeus), P. rubi, and found a clear population structure, with a high degree of nucleotide divergence seen between some race types and abundant private variation associated with race types 4 and 5. In contrast, between isolates defined as United Kingdom races 1, 2, and 3 (UK1-2-3) there was no evidence of gene loss or gain; or the presence of insertions/deletions (INDELs) or Single Nucleotide Polymorphisms (SNPs) within or in proximity to putative pathogenicity genes could be found associated with race variation. Transcriptomic analysis of representative UK1-2-3 isolates revealed abundant expression variation in key effector family genes associated with pathogen race; however, further long read sequencing did not reveal any long range polymorphisms to be associated with avirulence to race UK2 or UK3 resistance, suggesting either control in trans or other stable forms of epigenetic modification modulating gene expression. This work reveals the combined power of population resequencing to uncover race structure in pathosystems and in planta transcriptomic analysis to identify candidate avirulence genes. This work has implications for the identification of putative avirulence genes in the absence of associated expression data and points toward the need for detailed molecular characterisation of mechanisms of effector regulation and silencing in oomycete plant pathogens.

Keywords: RNA-Seq; host–microbe interactions; nanopore sequencing; oomycete; population resequencing; race structure; red core.

FIGURE 1

Observed Phytophthora fragariae symptoms in the cultivated strawberry (Fragaria × ananassa). (A,B,D,E) Roots of Fragaria × ananassa harvested 6 weeks after inoculation with Phytophthora fragariae mycelial slurry. (A) Successful infection of the P. fragariae isolate NOV-27 (race CA2) on a susceptible “Redgauntlet” plant (Rpf2 only). (B) Unsuccessful infection of the P. fragariae isolate A4 (US4/UK2) on a resistant “Redgauntlet” plant (Rpf2 only). (C) Example of “red core” symptoms observed in “Hapil” roots infected with BC-16, 3 weeks post-inoculation. (D) Unsuccessful infection of the P. fragariae isolate NOV-27 (race CA2) on a resistant “Allstar” plant (Rpf1, Rpf2 and Rpf3). (E) Successful infection of the P. fragariae isolate A4 (race US4/UK2) on a susceptible “Hapil” plant (no Rpf genes). (F) Example of BC-16 oospores observed in “Hapil” roots, 3 weeks post-inoculation, confirming infection.

FIGURE 2

Analysis of unique and expanded orthogroups for Phytophthora fragariae isolates of the UK1, UK2, and UK3 races did not lead to the identification of candidate avirulence genes. Orthology groups were identified by OrthoFinder (Emms and Kelly, 2015) and Venn diagrams were plotted using the VennDiagram R package (Chen and Boutros, 2011) in R Core Team (2016). (A) Analysis focused on the P. fragariae isolates of race UK1: BC-1 and NOV-5 compared to isolates of races UK2 and UK3. (B) Analysis focused on the P. fragariae isolates of race UK2: A4 and BC-16 compared to isolates of races UK1 and UK3. (C) Analysis focused on the P. fragariae isolates of race UK3: NOV-5, NOV-27 and NOV-71 compared to isolates of races UK1 and UK2.

FIGURE 3

Phylogenetic and population analysis of high quality, biallelic SNP sites split the isolates into three populations with SCRP245, potentially forming an ancestral or hybrid isolate between the UK1-2-3 population and the population represented by BC-23 and ONT-3. (A) Distruct plot of fastSTRUCTURE (Raj et al., 2014) results carried out on all sequenced isolates of Phytophthora fragariae. Each colour represents a different population. 1,469 variant sites were retained for this analysis after filtering. (B) Neighbour joining tree based on 545,365 high quality, biallelic SNP sites, node labels represent the number of bootstrap replicates supporting the node. Variant sites were identified by aligning Illumina reads of all the sequenced isolates to the reference assembly of the BC-16 isolate of P. fragariae with Bowtie 2 (Langmead and Salzberg, 2012) and analysis with the Genome Analysis Toolkit (GATK) haplotypecaller (McKenna et al., 2010). Sites were filtered with VCFtools (Danecek et al., 2011) and VCFlib (Garrison, 2012) to leave only high quality, biallelic SNP sites.

FIGURE 4

Analysis of expression data showed grouping of biological replicates, a separation of the BC-16 timepoints, changes during the infection process by the BC-16 isolate and confirmation of the differential expression of a candidate avirulence gene. (A) Principal component analysis of the differentially expressed transcripts for all analysed RNA-Seq experiments. RNA-Seq reads were aligned to the assembly of the BC-16 isolate of Phytophthora fragariae using STAR version 2.5.3a (Dobin et al., 2013). Predicted transcripts were then quantified with featureCounts version 1.5.2 (Liao et al., 2014) and differential expression was identified with the DESeq2 version 1.10.1 R package (Love et al., 2014). Following this, an rlog transformation of the expression data was plotted as a principal component analysis with R Core Team (2016). (B) Venn diagram of all differentially expressed transcripts during the BC-16 infection timecourse experiment. (C) Venn diagram of all differentially expressed predicted RxLR effectors during the BC-16 infection timecourse experiment. (D) Venn diagram of all differentially expressed predicted Crinkler effectors during the BC-16 infection timecourse experiment. (E) Venn diagram of all differentially expressed predicted apoplastic effectors during the BC-16 infection timecourse experiment. Venn diagrams were plotted using the VennDiagram R package (Chen and Boutros, 2011) in R Core Team (2016). (F) Quantitative reverse transcription PCR of a strong candidate for the avirulence gene possessed by BC-16 and A4, but not BC-1 and NOV-9 (PF003_g27513.t1). Plots created by the ggplot2 R package (Wickham, 2016) in R version 3.4.3 (R Core Team, 2017).

FIGURE 5

Differential expression of putative PfAvr2 and PfAvr3 is not due to sequence variation in Phytophthora fragariae BC-16 (UK2) and NOV-9 (UK3) genomes. Regions surrounding candidate avirulence genes, PfAvr2 and PfAvr3, from BC-16 and NOV-9 aligned with MAFFT in Geneious R10 (Katoh et al., 2002; Katoh and Standley, 2013). Black bars indicate contiguous sequences and gaps are represented by dashes. Identity is shown for all sequences in the alignment, green denotes residues at that position are the same across all sequences, yellow denotes less than complete identity and red denotes very low identity for the given position. (A) Putative PfAvr2, showing a 30 bp insertion in the NOV-9 sequence upstream of PF003_g27513 (shown in green) and a T to G SNP in NOV-9 downstream of the gene of interest. (B) Putative PfAvr3, showing an extra G insertion in BC-16 upstream of PF003_g27386 (an orthologue of PF009_g26267; shown in green) and an extra G insertion in BC-16 3,761 bp downstream.