Effector Polymorphisms of the Sunflower Downy Mildew Pathogen Plasmopara halstedii and Their Use to Identify Pathotypes from Field Isolates

Abstract

The obligate biotroph oomycete Plasmopara halstedii causes downy mildew on sunflower crop, Helianthus annuus. The breakdown of several Pl resistance genes used in sunflower hybrids over the last 25 years came along with the appearance of new Pl. halstedii isolates showing modified virulence profiles. In oomycetes, two classes of effector proteins, key players of pathogen virulence, are translocated into the host: RXLR and CRN effectors. We identified 54 putative CRN or RXLR effector genes from transcriptomic data and analyzed their genetic diversity in seven Pl. halstedii pathotypes representative of the species variability. Pl. halstedii effector genes were on average more polymorphic at both the nucleic and protein levels than random non-effector genes, suggesting a potential adaptive dynamics of pathogen virulence over the last 25 years. Twenty-two KASP (Competitive Allele Specific PCR) markers designed on polymorphic effector genes were genotyped on 35 isolates belonging to 14 Pl. halstedii pathotypes. Polymorphism analysis based on eight KASP markers aims at proposing a determination key suitable to classify the eight multi-isolate pathotypes into six groups. This is the first report of a molecular marker set able to discriminate Pl. halstedii pathotypes based on the polymorphism of pathogenicity effectors. Compared to phenotypic tests handling living spores used until now to discriminate Pl. halstedii pathotypes, this set of molecular markers constitutes a first step in faster pathotype diagnosis of Pl. halstedii isolates. Hence, emerging sunflower downy mildew isolates could be more rapidly characterized and thus, assessment of plant resistance breakdown under field conditions should be improved.

Fig 1. Workflow of the study.

In grey, the procedure used to select candidate effector genes. Transcriptomic data from sunflower leaves infected by Pl. halstedii or isolated zoospores were analyzed by PSI-tBLASTn using the annotated sequences that were available in NCBI in March 2010 as models for RXLR and CRN effectors. In white, the procedure used to build the Pl. halstedii pathotype determination key. Identification of effector polymorphism was done by comparisons between genomic sequences obtained in 7 representative pathogen pathotypes (100, 300, 304, 334, 700, 703 and 710). Among the 22 KBioscience Competitive Allele Specific PCR (KASP) markers, eight were used in a determination key to discriminate Pl. halstedii pathotypes.

Fig 2. Polymorphism analysis of Pl. halstedii effector and non-effector genes.

(A) Mean nucleotide diversity (π) calculated on non-effector genes (black bar) and effector genes (grey bar). Π values were calculated on DnaSP v5 software. Error bars represent two SEM. (B, C) Comparisons of polymorphism distributions (represented by SNP sparseness, i.e. minimum average distance between two polymorphisms) in Pl. halstedii effector (grey) and non-effector genes (black). (B) Frequency of genes in each class of index. (C) Frequency of predicted peptides in each class of SNP sparseness, among genes with nucleotide polymorphism.

Fig 3. Genotypes of 14 reference isolates (_ref) and 21 geographical isolates of Pl. halstedii for 22 KASP markers based on effector gene SNPs.

The DNA base involved in polymorphism is indicated by a specific colour. Heterozygous DNA bases are separated with a slash. (–) sign corresponds to the absence of the indel (versus T in PhRXLR58_1). *Genomic sequences of effectors available for these pathotypes (S1_File). Not determined results: nd.

Fig 4. Identification key for Pl. halstedii pathotypes using KASP markers designed on effector gene SNPs.

First level (Key 1) separated the 14 Pl. halstedii pathotypes in 5 groups with 3 markers. Second level (Key 2) used five other markers to distinguish 12 subgroups of pathotypes, and especially 6 subgroups of multi-isolate pathotypes (Red boxes). 100 and 304, 314 and 714 pathotypes could not be distinguished. ¹Pl. halstedii pathotypes with only one isolate available.

Fig 5. Putative associations between resistance (R) or susceptibility (S) profiles of sunflower lines (D4, D7 and D8) and genetic profiles of effector markers (PhCRN33 and PhRXLR15) in Pl. halstedii pathotypes from sequenced data (A) and from KASP marker results (B).

(A) PhCRN33 and PhRXLR15 type B sequenced alleles from 7 pathotypes (100, 304, 334, 700, 703, 710 and 730) were respectively associated with resistance profiles of D4, D7, and D8 sunflower lines. (B) Respective associations of PhCRN33_1 and PhRXLR15_1-PhRXLR15_3 observed polymorphisms in 35 Pl. halstedii isolates with resistance (R) profiles of D4, D7, and D8 sunflower lines. The amino acids of the corresponding translated sequences are indicated for each SNP position.