Adding by subtracting: Impairing tomato susceptibility genes to obtain resistance to Verticillium wilt

In the field of plant-microbe interactions the concept of plant disease susceptibility (S) genes is relatively new as research has largely focused on plant resistance in the past. S genes are host factors that are needed by the pathogen to establish disease and the use of S genes in breeding requires their impairment in order to establish resistance. The use of impaired S genes provides a parallel strategy to breed for resistance, especially in those cases in which dominant resistance is not available or quickly overcome. One example of a pathogen for which only one dominant R gene is described so far is the notorious soil-borne vascular wilt fungus Verticillium dahliae which affects a wide range of crops, including tomato. For V. dahliae, only a few S genes have been described in literature so far. Generally, many S genes have been identified in the model plant Arabidopsis. However, in order to be able to use impaired S genes in resistance breeding, it is essential to identify and functionally characterize S genes in crop species as well as in model species. In this thesis, strategies to identify plant S genes are reviewed (Chapter 1) and two reverse genetics strategies were pursued to identify S genes for V. dahliae in tomato.

Firstly, in order to be able to screen for loss of susceptibility against V. dahliae in tomato, a phenotyping assay was set up in Chapter 2. Several resistant and susceptible genotypes were tested to find the most reliable and reproducible phenotype caused by V. dahliae. To this end, several plant growth-related parameters were evaluated and plant canopy area was found to provide the highest discriminative power between mock and V. dahliae-inoculated plants. The relative difference in canopy area between mock and V. dahliae-inoculated plants was used to asses V. dahliae susceptibility in subsequent experiments. To determine whether the discriminative power based on canopy area measurements could be improved, the inoculation procedure was further evaluated. Neither an increased inoculum concentration, nor root trimming at the time of inoculation, nor nutrient application to the soil after inoculation significantly improved the discriminative power.

The first strategy to identify S genes for V. dahliae in tomato builds on the observation that the expression of many S genes is induced upon pathogen challenge. Therefore, available expression data were mined for candidate genes that were specifically induced in a compatible V. dahliae – tomato interaction. In Chapter 3, the most highly induced genes were selected for transient silencing using virus-induced gene silencing (VIGS). Subsequently, VIGS-treated plants were challenged with V. dahliae to screen for reduced susceptibility. Two candidates could be implicated in Verticillium wilt disease as potential S gene. The first candidate, Solyc06g067950, encodes an acyl-protein thioesterase 2 (APT 2) which catalyses the deacylation of proteins required for protein interactions with membranes. As knowledge on APTs is limited in plants, a link to plant immunity is not yet established. The second candidate, Solyc03g093140, encodes a glycerol-3-phosphate (G3P) transporter and G3P is known to function as signalling molecule in systemic acquired resistance. In order to study these candidates further, CRISPR-Cas9-mediated genome editing was used in Chapter 4 to generate knock-outs. Overall, targeted deletions in both candidates did not affect plant growth in early stages of development when compared with control plants. Surprisingly, however, none of the mutant lines showed loss of susceptibility upon challenge with V. dahliae. Collectively, these findings do not confirm the role of the G3P transporter in susceptibility to V. dahliae in tomato as three independent mutant lines showed no loss of susceptibility to V. dahliae. However, the role of APT 2 as S gene for V. dahliae requires further study because the generated mutant line only affected the rear part of the gene.

The second employed reverse genetics strategy is based on the fact that many S genes have a role in plant susceptibility which is conserved in different plant species. As extensive knowledge on S genes is available in Arabidopsis, a literature search was conducted in Chapter 5 to select candidates in this model species. For three previously identified S genes in Arabidopsis their tomato homologues were identified, and their role in V. dahliae susceptibility was determined in tomato using VIGS followed by disease phenotyping. Targeting of the tomato orthologue of Pyruvate Decarboxylase 1 (PDC1) and of WRKY27 did not result in reduced susceptibility to V. dahliae. However, transient silencing of the tomato orthologue of Walls Are Thin 1 (WAT1) indicated an involvement in susceptibility to V. dahliae. WAT1 encodes a tonoplast-localized auxin transporter and was previously found to be involved in V. dahliae susceptibility in both Arabidopsis and cotton. To further study the role of WAT1 in susceptibility to V. dahliae in tomato, CRISPR-Cas9- mediated knock-outs as well as WAT1-silenced lines using RNA interference (RNAi) were generated in Chapter 6. Silencing of WAT1 did not confirm its role in V. dahliae susceptibility in the RNAi lines. This can be attributed to the relatively high levels of residual WAT1 expression in these RNAi lines which likely compromised the silencing efficacy too much to monitor effects on V. dahliae infection. By means of CRISPR-Cas9, one WAT1 mutant line was obtained, which carried a biallelic mutation. Both mutant alleles affected the middle part of the gene and the deletions were predicted to affect the number of transmembrane domains of WAT1. Plants which were hetero- or homozygous for these deletions in WAT1 displayed severe growth defects in early plant development, such as severe discoloration of the leaves and strongly reduced overall growth. However, the targeted deletions in WAT1 also enhanced resistance to V. dahliae, as reduced stunting as well as reduced fungal biomass was monitored when compared with control plants. Furthermore, in line with previous findings that WAT1 is involved in susceptibility to multiple vascular pathogens, WAT1 CRISPR lines also showed loss of susceptibility to V. albo-atrum and Fusarium oxysporum f. sp. lycopersici in tomato.

In the Chapter 7, a general discussion on the identification of S genes in crops using different approaches is provided.