Quantitative investigation of hepatic genomic response to hormonal and pathophysiological stimuli by multivariate analysis of two-dimensional mRNA activity profiles.

We have applied techniques of multivariate analysis to the characterization and comparison of the effects of various pathophysiological and hormonal stimuli on the expression of the rat hepatic genome at a pretranslational level. In vitro translated products were resolved by two-dimensional gel electrophoresis. We analyzed 10 pathophysiological states brought about by variation in thyroidal status, starvation, administration of high carbohydrate diet, and the production of experimentally induced diabetes mellitus. Each state differed significantly from every other state in the two-dimensional electrophoretic profiles. The set consisting of the minimal number of products necessary for maintaining the distinctive patterns was identified. The analysis also defined those clusters of products that behaved in a coordinate fashion in response to the various stimuli. Lastly, the similarity and dissimilarity of hepatic mRNA activity profiles to each other could be geometrically represented in three-dimensional space. Our finding that the hepatic mRNA activity profile could distinguish reliably between closely related hormonal and pathophysiological stimuli indicates the specificity of hepatic genomic expression. A systematic analysis of such profiles may be useful as an overall index of the biologic response at the hepatocellular level.