BIOLOGICAL CONTROL OF PIERCE’S DISEASE WITH NON-PATHOGENIC STRAINS OF XYLELLA FASTIDIOSA Project Leader:

This project is to construct and test nonpathogenic strains of X. fastidiosa derived from a pathogenic Pierce’s disease strain for competitive exclusion of the pathogen in grapevines. Potential virulence genes were selected from comparative genome sequence analyses as well as DNA macroarray studies of differential gene expression. A more comprehensive analysis of differential gene expression with a DNA microarray approach is an outcome of this project and is funded under a new grant. Disruption of xanthan gum production by mutations in gumD and gumH resulted in less of a slime layer and fewer adhering cells than the wild-type strain on plastic and wood surfaces. However, when biofilm formation was quantitatively measured in polystyrene microtiter plates, both gumD and gumH mutants formed significantly more biofilm than the wild-type X. fastidiosa. In addition, the disruption of rsmA resulted in significantly more biofilm that the parent strain. Virulence assays in grapevine are still in progress, but gum mutants showed fewer symptoms than the parent strain in an alternative host assay. The analysis of potential virulence genes has also had side benefits for epidemiological work on Xylella by providing new primer sets for differentiate certain host strains of X. fastidiosa. INTRODUCTION A general approach that is being considered to manage Pierce’s disease is biological control of the bacterial pathogen. Specific biological control approaches include the use of antagonistic endophytic bacteria isolated from the xylem of grape, bacteriophages, and interference with bacterial intercellular signaling. Another approach is the possible use of a nonpathogenic strain of X. fastidiosa derived from a pathogenic Pierce’s disease strain for competitive exclusion of the pathogen in grapevines. This concept has certain advantages over other biological control approaches and considerable precedent in bacterial, fungal, and viral systems, including the biological control of xylem-inhabiting bacterial pathogens. Colonization and protection of plants with less virulent or completely nonpathogenic strains of plant pathogens has been demonstrated in a number of bacterial (Wilson and Lindow, 1993), fungal (Sneh, 1998), and viral systems (Fulton, 1986). Some studies relied on naturally occurring avirulent strains, while other researchers have developed defined nonpathogenic mutants of pathogenic strains for this purpose, following the expectation that they would have the same ecological requirements for growth and are therefore ideal competitors (Wilson and Lindow, 1993). Another advantage of this approach is the specificity of the interaction, which reduces or eliminates possible deleterious effects on non-target organisms (Cook et al., 1996). This concept has also been demonstrated for the xylem-inhabiting vascular wilt pathogen, Ralstonia (Pseudomonas) solanacearum (Frey et al., 1994). Nonpathogenic mutants of this pathogen still colonized vascular tissues and resulted in high protection rates against the pathogenic strain. Our goal is to test this concept of competitive exclusion with nonpathogenic, or reduced virulence mutants, of the xylem-inhabiting Xylella fastidiosa for biological control of Pierce’s disease. OBJECTIVES 1. Construct deletion mutations in putative virulence genes of 2. Test mutant strains for virulence in grapevines. 3. Test mutant strains for biological control of pathogenic strains in grapevines. RESULTS AND CONCLUSIONS Results Selection of candidate virulence genes We have utilized the full genome sequences of Xylella fastidiosa strains (Bhattacharyya, A., et al. 2002; Simpson et al., 2000; Van Sluys et al., 2003) to select open reading frames specifying putative pathogenicity and virulence factors. In addition, we constructed a DNA macroarray with about 100 of these genes to analyze their expression in different Xylella strains in planta and in vitro (Hernandez-Martinez et al., 2002). This work follows the hypothesis that many genes important in virulence and symptom will be differentially expressed in the bacterium grown in culture vs. during infection of plants. We have shown that these genes are expressed to varying degrees ranging from none to very high. However, since over 50% of the Xylella fastidiosa genome consists of genes with no known function, a more comprehensive approach toward the identification of virulence genes is necessary. We have recently obtained funding from the CDFA in a separate project to continue this work through the use of full genome microarrays.