The role of AVR4 and AVR4E proteins in virulence and avirulence of the tomato pathogen Cladosporium fulvum: molecular aspects of disease susceptibility and resistance

Active disease resistance in plants relies on highly sensitive and specific surveillance systems that enable recognition and restriction of growth of invading pathogens. Numerous resistance ( R ) genes have been characterized from various plant species that provide typical gene-for-gene-based resistance to a variety of different pathogens that carry the matching avirulence ( Avr ) genes. Because an Avr gene product is not beneficial for the pathogen in the presence of the matching R gene, the primary function of Avr gene products is expected to be associated with virulence rather than with avirulence (Chapter 1). Indeed, evidence is accumulating that Avr genes encode effector proteins that condition establishment of a compatible interaction between a pathogen and a susceptible host, either by interacting with host-encoded virulence targets to suppress (basal) defence responses or by interfering with activities of pathogenesis-related (PR) proteins or other induced defence responses. Thus, losing an avirulence determinant, in order to overcome R gene-mediated resistance, might carry a virulence penalty for the pathogen. On the introgressed region in tomato plants carrying the Cf-4 locus (MM-Cf4 plants), five homologues of Cladosporium fulvum resistance gene Cf-9 ( Hcr9-4 s) are present. In Chapter 2, we demonstrated that homologue Hcr9-4E confers a similar level of resistance to C. fulvum as homologue Hcr9-4D , which represents the Cf-4 gene, through recognition of a novel avirulence determinant, designated AVR4E. The Avr4E gene encodes a cysteine-rich protein of 121 amino acids that is secreted into the apoplastic space of tomato as a mature protein of 101 amino acids (Chapter 2). The genomic Avr4E sequence of C. fulvum race 5 confers avirulence (after transformation) to strains of C. fulvum that are normally virulent on plants carrying Hcr9-4E , indicating that AVR4E is a genuine a race-specific avirulence determinant. Strains of C. fulvum evade Hcr9-4E- mediated resistance either by lack of Avr4E gene expression or by production of a stable AVR4E mutant protein that is elicitor-inactive and carries two amino acid substitutions, Phe 62Leu and Met 73Thr. Moreover, we demonstrated by site-directed mutagenesis that the single amino acid substitution Phe 62Leu in AVR4E is sufficient to evade Hcr9-4E- mediated resistance (Chapter 2).Strains of C. fulvum that produce race-specific elicitor AVR4 induce a hypersensitive response (HR) leading to resistance in tomato plants carrying the Cf-4 resistance gene. In Chapter 3, the mechanism of AVR4 perception by Cf-4 plants is examined by performing binding studies with 125I-AVR4 on microsomal membranes fractions (MFs) of MM-Cf0 (susceptible) and MM-Cf4 (resistant) tomato plants and non-host plant species. We identified an AVR4-high affinity-binding site (HABS) ( KD = 0.05 nM) that exhibited all the characteristics expected for ligand-receptor interactions, such as saturability, reversibility and specificity. Surprisingly, the AVR4-HABS present in MFs appeared to originate from fungi present on partially infected tomato plants rather than from the tomato plant itself. This fungus-derived, AVR4-specific HABS is heat- and proteinase K-resistant, suggesting that it might be a non- proteinaceous component. Affinity cross-linking demonstrated that AVR4 specifically binds to a component of about 75 kDa of fungal origin, a phenomenon that is possibly related to the intrinsic function of AVR4 for C. fulvum . As no AVR4-specific HABS could be detected in MFs of tomato plants that were grown under contained conditions, the mechanism of perception of AVR4 by MM-Cf4 plants appears to be different from that found for several other fungus-derived peptides and elicitors, including AVR9, for which a HABS has been identified in MFs of several plant species.The AVR4 protein of C. fulvum contains 8 cysteine residues, all of which are involved in disulfide bonds. Assignment of the disulfide bond pattern of AVR4 was partly achieved by a biological analysis using potato virus-X (PVX)-mediated expression of AVR4 disulfide bond mutants in plants carrying Cf-4 , and subsequently confirmed and completed by partial chemical reduction of AVR4, followed by mass mapping (Chapter 4). The four disulfide bonds present in AVR4 were identified as Cys 11-Cys 41, Cys 21-Cys 27, Cys 35-Cys 80and Cys 57-Cys 72. The disulfide bond pattern and the spacing of the cysteine residues were subsequently used to carry out a motif-based search, which revealed that AVR4 contains a chitin-binding domain that is also present in chitin-binding proteins of invertebrates (inv ChBD). Three disulfide bonds in AVR4 are conserved amongst members of the inv ChBD family and these bonds are required for conformational stability of AVR4 (Chapter 4). In natural strains of C. fulvum disruption of two of these three conserved disulfide bonds, i.e. between Cys 11-Cys 41and Cys 35-Cys 80, results in lack of AVR4 recognition by Cf-4 plants. However, all four disulfide bond mutants of AVR4 retained the ability to bind to chitin in vitro. Moreover, when bound to chitin, these disulfide bond mutants of AVR4 are less sensitive to proteolytic breakdown. Thus, while evasion of Cf-4- mediated resistance by C. fulvum appears to result from instability and protease sensitivity of AVR4 mutant proteins present in the apoplast of tomato, AVR4 mutants retained their putative intrinsic role in protecting the cell wall of C. fulvum against plant chitinases.AVR4 does not require binding to chitin to induce Cf-4- mediated plant defence responses. In Chapter 5, we analyzed whether domains responsible for necrosis-inducing activity (NIA) and chitin-binding ability within AVR4 can be distinguished. Therefore, we performed a peptide scan (PEPSCAN) analysis using polyclonal antibodies raised against native AVR4 to identify antigenic domains containing putatively solvent-exposed residues that might condition NIA of AVR4. This antibody-affinity analysis identified one major (Cys 41-Cys 57) and two minor antigenic domains in AVR4 (Ile 17-Asn 31and Gln 62-Asn 76). Selective alanine substitutions were performed on all, except for Cys, Pro and Gly, residues present in the major antigenic domain of AVR4, as this domain exhibits the highest affinity for AVR4 polyclonal antibodies. Moreover, to determine which residues in AVR4 are important for chitin binding, all aromatic residues and two additional residues present in conserved domains of the anticipated chitin-binding domain(s) were individually substituted by alanine. We showed that replacement of single amino acids in the major antigenic domain of AVR4 does not affect its NIA (Chapter 5). Substitution of aromatic residues Tyr 38and Trp 63by Ala, however, reduced the NIA of AVR4. However, these AVR4 mutant proteins appeared to be unstable, suggesting that Tyr 38and Trp 63contribute to conformational stability rather than to NIA of AVR4. In this study, no individual amino acid residues could be identified that are essential for binding of AVR4 to chitin, suggesting that multiple rather than single amino acid residues contribute to chitin-binding ability.It would be beneficial for C. fulvum to evade induction of plant defense responses by modifying the Avr gene products in such way that their avirulence but not their virulence functions are lost (Chapter 6). For most single AVRs, however, no pronounced role in virulence of C. fulvum could be assigned, possibly due to the fact that other proteins compensated for their intrinsic functions, suggesting that AVRs are redundant.