Identification of Candidate Genes Related to Water Stress in Grapevine by Co-expression Analysis and Comparative Genomics

Grapevine is one of the most important fruit crops in the world. Water shortage restricts where grape can commercially be grown and drought stress threatens the yield and quality in non-irrigated growing regions. Understanding of molecular responses to water stress is critical for sustainable global grape production. We combined gene coexpression network analysis and comparative genomics approach to identify candidate genes responding to water stress in grapevine. We first obtained 241 Arabidopsis genes that had the experimental evidence responding to water stress by searching the PubMed publications. Then these 241 genes were used to retrieve 1,145 coexpressed genes via querying the Arabidopsis coexpression database. In grapevine, 727 genes were obtained as orthologs to the water stress-responding genes in Arabidopsis. Next, we constructed a grapevine genome-wide gene coexpression network by using 374 available microarrays. In the coexpression network, 3,019 nodes were linked with 13,758 edges, and 196 coexpressed gene modules were identified. Of the 196 modules, 31 modules were found to be enriched with GO terms or KEGG pathways, providing systems-level functional modules that contain the coexpressed genes. Finally, we used grapevine orthologous genes to those in Arabidopsis as a query to search the water stress-responding genes in the grapevine coexpression network. We identified 447 genes in 31 modules which were considered as candidate water stress-responding genes. Of the 447 genes, 98 genes were identified as differentially expressed genes based on grapevine microarrays containing treatments of water deficit and the non-deficit conditions. These grapevine genes are high likelihood candidate genes for responding water stress in grapevine. We are in the process of validating the functions of some of these genes.