Natural loss-of-function mutation of EDR1 conferring resistance to tomato powdery mildew in Arabidopsis thaliana accession C24

Abstract

To screen for potentially novel types of resistance to tomato powdery mildew Oidium neolycopersici, a disease assay was performed on 123 Arabidopsis thaliana accessions. Forty accessions were fully resistant, and one, C24, was analysed in detail. By quantitative trait locus (QTL) analysis of an F2 population derived from C24 × Sha (susceptible accession), two QTLs associated with resistance were identified in C24. Fine mapping of QTL-1 on chromosome 1 delimited the region to an interval of 58 kb encompassing 15 candidate genes. One of these was Enhanced Disease Resistance 1 (EDR1). Evaluation of the previously obtained edr1 mutant of Arabidopsis accession Col-0, which was identified because of its resistance to powdery mildew Golovinomyces cichoracearum, showed that it also displayed resistance to O. neolycopersici. Sequencing of EDR1 in our C24 germplasm (referred to as C24-W) revealed two missing nucleotides in the second exon of EDR1 resulting in a premature stop codon. Remarkably, C24 obtained from other laboratories does not contain the EDR1 mutation. To verify the identity of C24-W, a DNA region containing a single nucleotide polymorphism (SNP) unique to C24 was sequenced showing that C24-W contains the C24-specific nucleotide. C24-W showed enhanced resistance to O. neolycopersici compared with C24 not containing the edr1 mutation. Furthermore, C24-W displayed a dwarf phenotype, which was not associated with the mutation in EDR1 and was not caused by the differential accumulation of pathogenesis-related genes. In conclusion, we identified a natural edr1 mutant in the background of C24.

Keywords: EDR1; Oidium neolycopersici; resistance; tomato.

Figure 1

Quantitative trait locus (QTL) analysis of Oidium neolycopersici (On) resistance and fine mapping of QTL‐1 on chromosome 1 in C24 × Sha. (A) Symptoms of On infection on C24, F₁ C24 × Sha and Sha. (B) Two QTLs located on chromosomes 1 and 2 were identified. LOD, logarithm of odds. (C) Disease index (DI) after On infection. Class C/C is homozygous for the C24 allele, C/S is heterozygous and S/S is homozygous for the Sha allele. For each class, the DI value is the average score of F₂ plants with the designated genotype for marker 159 linked to QTL‐1 on chromosome 1, or for marker 515 linked to QTL‐2 on chromosome 2. (D) Markers used for the fine mapping of QTL‐1 (Table S4) are indicated. The distance between markers is proportional to the physical distance. White bars represent regions homozygous for the C24 allele, and shaded bars represent heterozygous regions. The space in between white and shaded bars denotes a crossover event between two flanking markers for each recombinant. The arrows point towards the interval in which QTL‐1 resides. For each recombinant (REC), the phenotype of F₃ and/or F₄ populations is indicated. S, susceptible; SNP, single nucleotide polymorphism; R, resistant.

Figure 2

The edr1 (enhanced disease resistance 1) mutation confers resistance to tomato powdery mildew O idium neolycopersici. (A) Fungal growth on Col‐0 plant and edr1 mutant. (B) Fungal biomass quantification. Values were normalized relative to act2, and calibrated to levels on edr1 mutants. Error bars represent standard deviation of three biological replicates and, for each replicate, rosette leaves were collected. Asterisk indicates significant difference from the control according to independent‐samples t‐test: *P < 0.05. A representative of two experiments is presented.

Figure 3

Characterization of C24‐W. (A) Mutation in the EDR1 (Enhanced Disease Resistance 1) gene of C24‐W. Top panel shows that, among three GT repeats (red, underlined), one repeat was missing in C24‐W, leading to the occurrence of a premature stop codon TGA (red, underlined). The middle panel shows the resulting protein sequence when one GT repeat is missing. The bottom panel depicts the position of the two‐nucleotide (nt) deletion in EDR1 of C24‐W. (B) Size differences of C24‐U and C24‐W plants. (C) Relative level of expression of defence genes PR1 (Pathogenesis‐Related gene 1) and PR2 (Pathogenesis‐Related gene 2) in plants not inoculated with the pathogen. (D) Fungal biomass quantification in C24‐W compared with C24‐U. Values were normalized relative to act2, and calibrated to levels in C24‐U plants. Error bars represent standard deviation of three biological replicates and, for each replicate, rosette leaves were collected. Asterisks indicate significant difference from the control according to independent‐samples t‐test: *P < 0.05. A representative of two experiments is presented.

Figure 4

Macroscopic and microscopic images of F₄ plants homozygous for the C24‐W  edr1 (enhanced disease resistance 1) allele (edr1/edr1), or heterozygous (EDR1/edr1), at 8 days post‐inoculation with Oidium neolycopersici (On). Left panel: F₄ plants homozygous for the C24‐W  edr1 allele; right panel: heterozygous F₄ plants. (A) Macroscopic symptoms after On inoculation. (B) Images of histological study showing fungal growth and cell death. Compared with heterozygous F₄ plants, fungal growth was greatly restricted after haustorium formation on the resistant edr1edr1  F₄ plants, and cell death was observed in many epidermal cells intruded by fungal haustoria. HR, Hypersensitive Response.

Figure 5

Relative PR (Pathogenesis‐Related) gene expression levels in F₄ plants homozygous for the C24‐W  edr1 allele (edr1edr1), or heterozygous (EDR1edr1), at 6 days post‐inoculation with Oidium neolycopersici.

Figure 6

Suppression of putative homologues of EDR1 (Enhanced Disease Resistance 1) in tomato did not affect the susceptibility level of cultivar Moneymaker (MM). (A) Phylogenetic analysis of protein sequences of tomato EDR1 candidates compared with EDR1 from Arabidopsis and other plant species. (B) Representation of fragments from the coding sequences of the tomato EDR1 candidates used in RNAi constructs. (C) Relative expression of Solyc01g097980 (Solyc01g) and Solyc06g068980 (Solyc06g) in silenced lines. (D) Relative fungal growth in T2 plants containing the silencing construct [T2(+)] compared with MM (results of the first experiment). (E) Relative fungal growth in T2 plants containing the silencing construct [T2(+)] compared with T2 plants without the silencing construct [T2(–)] (results of the second experiment). Values were normalized relative to Elongation Factor 1α (EF), calibrated to levels in untransformed MM plants or non‐transgenic T2 plants. Error bars represent standard deviation of at least three biological replicates and, for each replicate, third and fourth leaves were pooled. Asterisks indicate significant difference from the control according to one‐way analysis of variance or t‐test: *P < 0.05.