MICROARRAY ANALYSIS OF GLOBAL GENE EXPRESSION OF VITIS VINIFERA IN RESPONSE TO XYLELLA FASTIDIOSA INFECTION Project Leader:

In previous years we have analyzed gene expression profiles of Pierce’s Disease (PD) resistant and susceptible genotypes of Vitis arizonica hybrids in response to infection by Xylella. fastidiosa (Xf), the bacterium causing PD. Here we report gene expression of the PD susceptible European grapevine, V. vinifera in response to Xf infection. RNA was extracted from healthy and infected leaf tissues at 4, 8 and 10 weeks post inoculation. Isolated RNA was converted into cDNA and used for microarray-based global gene expression analysis. Data analysis results indicated that there were a total of 2,385 differentially expressed transcripts from early (4 week), middle (8 week) and late (10 week) stages of disease development. Of these 2,385 transcripts, 1,050 transcripts were up-regulated (2 to 100 fold) and 1,335 transcripts were down-regulated (0.5 to 0.02 fold) across the three stages. Comparative analysis of the differentially regulated transcripts has identified common and distinctive features of the host response from V. arizonica hybrids and V. vinifera to Xf infection that are important for understanding of PD resistant and susceptible mechanisms. An online relational database is now publicly available that has Vitis transcriptome data along with other relevant information and bioinformatics tools. INTRODUCTION The impact of PD on the California grape industry has been significant since the introduction and establishment of a more effective vector, Homalodisca coagulata, the glassy-winged sharpshooter (Almeida and Purcell 2003). Development of resistance in grape is stymied by the relatively limited amount of genetic and molecular information regarding genotype specific resistance to PD infection (Davis et al. 1978). From genotypic screening and genetic mapping studies, it was concluded that a dominant allele controls PD resistance and recently, Krivanek et al. (2006) identified a major quantitative trait locus that controls PD resistance and denoted it as ‘Pierce’s disease resistance 1’ (PdR1). The above studies confirm that the genetic basis of PD resistance in grapes varies from tolerance to resistance and suggest that host responses to the pathogen are genotype dependent. Our recent studies further confirmed that PD response differs between resistance and susceptible genotypes at molecular and physiological levels (Lin et al., 2007; Fritschi et al, 2007). Further, in the PD resistant genotypes, differential responses between stem and leaf tissues were also noted (Krivanek and Walker, 2005). The results from these studies prompted study of genome-wide molecular basis of this host / pathogen interaction. Plants respond to pathogen attack through a variety of signaling pathways consisting of a large number of regulatory as well as effector genes. Microarrays facilitate automated analysis of transcriptional profiling data to enable an understanding of such gene function and interactions. The goal of this study was to identify and characterize the molecular events in the grape/Xf interaction using genome wide transcriptome profiling between resistant and susceptible genotypes and among the different tissue types. OBJECTIVES 1. Perform a microarray gene expression analysis. 2. Develop a grape transcriptional relational database. RESULTS AND DISCUSSIONS Objective 1 Microarray gene expression analysis. Experimental set-up: Total RNA from leaf tissues of V. vinifera from 4, 8 and 12 weeks post-infection with Xf was hybridized to nine slides in a two-color experiment using the monochromatic dyes Cy5 and Cy3. For each time point, there were three slides (biological replicates) including a dye flip. Data analysis: For each gene there were 54 data points per each stage (18 per slide x 3 biological replications) of disease development. Data representing raw spot intensities generated by the GenePix software were first normalized with RMA algorithm (Robust multichip average) and data was further subjected to quantile normalization. SAM software was used to identify statistically significant genes expression changes using a cut-off value of two-fold differential expression and a qvalue of 0.5. Clustering of the significantly differentially expressed genes was carried out using TMEV software.