A gene expression biomarker identifies factors that modulate sterol regulatory element binding protein

Abstract

Lipid levels are regulated in mammals by sterol regulatory element-binding proteins (SREBPs). To identify chemicals that might cause steatosis through SREBP, a mouse gene expression biomarker was characterized that can predict modulation of SREBP family members. Using microarray comparisons, genes were identified that were commonly regulated by constitutively-active SREBP family members and required SREBP cleavage-activating protein. Pair-wise correlations between the biomarker and microarray profiles were measured using the Running Fisher algorithm; the test achieved a balanced accuracy of 95% and 94% for prediction of SREBP activation or suppression, respectively. Our approach uncovered a number of previously unrecognized features of SREBP regulation in a mouse liver gene expression compendium of ∼2600 microarray comparisons. SREBP modulation is 1) a relatively common phenomenon, with 10% and 12% of all comparisons exhibiting activation or suppression, respectively; 2) often modulated by diets and chemicals; 3) linked to steatosis induced by perfluoroalkyl acids; and 4) in most cases (77%) leads to coordinated increases or decreases in lipogenic and cholesterogenic genes. Conditions were identified in which regulation was uncoupled. Together, these findings identify chemicals that perturb SREBP activation and demonstrate a diversity of coupling levels among lipogenic and cholesterogenic genes across experimental conditions.