Evaluation of common gene expression patterns in the rat nervous system.

Abstract

In the postgenomic era, integrating data obtained from array technologies (e.g., oligonucleotide microarrays) with published information on eukaryotic genomes is beginning to yield biomarkers and therapeutic targets that are key for the diagnosis and treatment of disease. Nevertheless, identifying and validating these drug targets has not been a trivial task. Although a plethora of bioinformatics tools and databases are available, major bottlenecks for this approach reside in the interpretation of vast amounts of data, its integration into biologically representative models, and ultimately the identification of pathophysiologically and therapeutically useful information. In the field of neuroscience, accomplishing these goals has been particularly challenging because of the complex nature of nerve tissue, the relatively small adaptive nature of induced-gene expression changes, as well as the polygenic etiology of most neuropsychiatric diseases. This report combines published data sets from multiple transcript profiling studies that used GeneChip microarrays to illustrate a postanalysis approach for the interpretation of data from neuroscience microarray studies. By defining common gene expression patterns triggered by diverse events (administration of psychoactive drugs and trauma) in different nerve tissues (telencephalic brain areas and spinal cord), we broaden the conclusions derived from each of the original studies. In addition, the evaluation of the identified overlapping gene lists provides a foundation for generating hypotheses relating alterations in specific sets of genes to common physiological processes. Our approach demonstrates the significance of interpreting transcript profiling data within the context of common pathways and mechanisms rather than specific to a given tissue or stimulus. We also highlight the use of gene expression patterns in predictive biology (e.g., in toxicogenomics) as well as the utility of combining data derived from multiple microarray studies that examine diverse biological events for a broader interpretation of data from a particular microarray study.