Extracellular proteins of the tomato pathogen Cladosporium fulvum; role in pathogenicity and avirulence

The interaction between the biotrophic fungus Cladosporium fulvum (Cooke) and its only host, tomato ( Lycopersicon esculentum Miller) complies with the gene-for-gene model. Early genetic studies have shown that many different avirulence ( Avr ) and resistance ( R ) gene pairs exist between strains of the fungus and genotypes of the plant, respectively. Previous studies in our laboratory, as summarized in Chapter 1, led to the isolation and characterization of several proteins of fungal and plant origin that occur in the apoplast during infection. The search for proteinaceous fungal elicitors that could act as avirulence factors on resistant tomato genotypes, led to the isolation of two proteins, AVR4 and AVR9. The development of molecular tools to manipulate C. fulvum provided means to confirm that the two encoding genes, Avr 4 and Avr 9, are the avirulence genes that match the tomato resistance genes Cf-4 and Cf-9 , respectively. The genes encoding two additional fungal e xtra c ellular p roteins (ECPs), Ecp 1 and Ecp 2, have also been cloned. Based on their specific induction in planta , both genes were proposed to be putative pathogenicity genes.

The object of the research presented in this thesis was to obtain a better understanding of the function of the ECPs in pathogenicity and virulence of C. fulvum on tomato and of their potential role as specific elicitors on particular tomato genotypes, which would possibly fit into additional Avr / R gene-for-gene pairs. Accordingly, most of the work presented in this thesis involves studies on the biological function of the ECP1 and ECP2 proteins and additionally characterized ECPs of C. fulvum .

In Chapter 2 a review on our current knowledge on fungal avirulence genes is presented. Although only a few fungal Avr genes have been cloned so far, they encode very different proteins and are differently regulated. Their involvement in pathogenicity or virulence, the type of responses that their products trigger on plants carrying the matching resistance genes, and their potential application in molecular resistance breeding are discussed.

Pathogenicity assays with transgenic strains of C. fulvum in which the Ecp 1 and/or Ecp 2 genes were deleted are presented in Chapter 3. Both genes are virulence factors, as their encoded products, ECP1 and ECP2, are required for full virulence. Nevertheless, when absent, the fungus is still able to cause disease on tomato. The studies also indicate that both genes might function as suppressors of plant defense responses. It is envisaged that pathogenicity of C. fulvum on tomato results from synergism between multiple virulence factors. That is, while each single virulence factor is partly involved in causing disease on tomato, the presence of a complete set of virulence factors ensures maximal parasitic ability.

Both ECP1 and ECP2 proteins are secreted, have a low molecular weight and are cysteine-rich proteins. They are secreted early and abundantly in the intercellular space, in a similar manner as observed for the two race-specific elicitors, AVR4 and AVR9. In Chapter 4 a search for plant genotypes that could recognize these proteins as specific elicitors of HR is described. One genotype of tomato, which had been preselected for resistance to C. fulvum , showed specific HR upon exposure to ECP2. Significantly, resistance of this genotype was shown to rely solely on ECP2 recognition. The resistance character is conferred by a single dominant gene, designated Cf-ECP2 . Analogous to the Avr 4/ Cf-4 and Avr 9/ Cf-9 gene pairs, Ecp 2/ Cf-ECP2 represents a genuine gene-for-gene pair within the C. fulvum -tomato interaction.

Since the Cf-ECP2 gene operates through recognition of a crucial virulence factor, it may, therefore, provide durable protection against C. fulvum in tomato crops. Thus, the same protein may have a role in virulence on all tomato plants and a role in avirulence on only particular tomato genotypes.

In Chapter 5, we subsequently initiated an extensive screen for plants among breeding lines and wild accessions of Lycopersicon species that could respond specifically with HR to any of the five interaction-specific fungal ECPs that we had isolated. Indeed, for each of the five ECPs that were tested, we found at least one individual that responded with HR. This observation indicates that recognition specificities are scattered among populations of plants, suggesting that recognition of proteins of C. fulvum is a generic property of the Lycopersicon genus. To investigate whether this is also true for a non-host species of C. fulvum , wild accessions of Nicotiana spp. were screened for similar specific HR-associated recognition of proteins of C. fulvum . We identified a few N. paniculata plants that exhibited a specific HR towards ECP2. This suggests that generic recognition specificities of proteinaceous elicitors of a fungal pathogen exist in solanaceous plants outside its host range. Consequently, the plant system for recognition of foreign proteins seems to generate new specificities at random, rather than specificities against classes of proteins produced by the pathogens of a given host plant.

Recently, the cloning of four tomato Cf genes, among which are the Cf-4 and the Cf-9 genes that match the avirulence genes Avr 4 and Avr 9 respectively, was reported. They all belong to the same gene family, which is predicted to encode membrane-anchored leucine-rich-repeat (LRR) proteins. As such, they might participate in the perception of the different fungal avirulence factors. The Cf genes are members of clusters containing several homologues.

In Chapter 6 we tested whether some of the Cf-9 homologues, referred to as Hcr9-9 genes ( H omologue of C ladosporium fulvumR esistance gene Cf- 9 present at the Cf- 9 locus), could confer additional resistance against C. fulvum . Inoculation of fungal strains in which the Avr 9 gene was deleted onto Cf9 tomato genotypes, showed that additional partial resistance(s) are present on the L.pimpinellifolium genomic segment introgressed into L. esculentum . These additional resistance(s) are independent of the Avr 9/ Cf-9 gene pair and are most probably due to one or more of the Hcr9-9 genes. However, we could not identify any specific proteinaceous elicitor activity in apoplastic fluids of compatible C. fulvum /tomato interactions that could match the identified additional resistance(s). The suggestion that Hcr9-9 genes are responsible for the additional resistance(s) has been proven in an independent study in another research group. Therefore, the Lycopersicon genus is not only rich in recognition specificities for secreted C. fulvum proteins (AVRs and ECPs), also a single plant genotype appears to contain multiple recognition capacities towards a single fungal strain.

The data obtained in this research are discussed in Chapter 7. They indicate that HR-associated recognition of extracellular, low-molecular weight, cysteine-rich proteins of C. fulvum , some of which are involved in pathogenicity, is a general phenomenon in the Lycopersicon genus. The finding that some of these proteins, that act as avirulence factors, are also important virulence factors of the fungus might help to develop a strategy for identification of putative durable resistance genes, as proposed for the Cf-ECP2 gene. The demonstration of partial resistance(s) conferred by the Cf-9 homologues, without detection of a HR-phenotype, could suggest that recognition capacities of these genes are not limited to proteins, or that gene-for-gene responses other than HR occur in the C. fulvum -tomato interaction. In addition, our data indicate that HR-recognition of C. fulvum proteins is not confined to its host genus, Lycopersicon , but also occurs in the non-host genus Nicotiana .

In conclusion, it appears that the diversity of the HR-associated recognition of proteins by a variety of Cf genes in a population of tomato relatives, generates plants with a genuine surveillance system against invading pathogens. The importance of the recognized proteins in pathogenicity or virulence of the pathogen is likely to be positively correlated with the durability of the matching resistance gene. The wide occurrence of gene-for-gene systems in crop plants is most probably a result of human plant breeding activities, involving transfer and dissection of gene clusters that originate from wild species.