The use of gene expression profiling to dissect the stress sensitivity of reproductive development in maize

With the advent of large-scale genomic sequencing, complete or partial genomic sequencing has been conducted on a range of species resulting in a significant increase in gene sequence information. Although the number of genes identified is increasing, our knowledge of the function and interaction of these newly discovered genes is lacking. Functional genomic tools have been developed to improve our understanding of these genes. Gene expression profiling, also known as transcript profiling or mRNA profiling, is one such tool and its use is the focus of this paper. We utilized two unique gene expression profiling technologies, the molecular dynamic (MD)/Amersham microarray and the Affymetrix GeneChip, to investigate the effect of abiotic stress on gene expression in female reproductive tissues of maize (Zea mays L.). The MD microarray, described here as a targeted genomic approach, was populated with 384 maize genes largely comprising four important metabolic pathways with rigorous gene coverage within each pathway. In contrast, the Affymetrix GeneChip, defined here as a non-targeted genomic approach, was designed to provide a survey of metabolism, spanning 27 pathways, including >300 unknown genes. We utilized these technologies in conjunction with traditional physiological approaches to sample tissues subjected to various forms of abiotic stress. From these experiments, we conclude that flowering remains a susceptible developmental period to abiotic stress in elite maize hybrids and yield losses are associated with abnormal floral development and impaired ear growth. With the targeted approach, we show that genes within the starch biosynthetic pathway are coordinately regulated under stress and the decrease in expression of these genes co-varies with the loss of starch. The non-targeted approach: (1) revealed a set of genes that were consistently affected by abiotic stress regardless of tissue type; (2) demonstrated that gene expression in some tissues is more responsive to stress than other tissues; (3) identified a group of genes that were drought responsive that would not have been predicted a priori. Collectively, these experiments provide examples of how these technologies can be used for both gene and pathway discovery. (C) 2002 Published by Elsevier Science B.V.