PPARg/RXR transcriptional circuitry governing adipocyte biology

C67

Perixosome Proliferator-Activated Receptor gamma (PPARg) transcription factor plays an indispensable role in adipogenesis, and has been linked to obesity and type 2 diabetes. In addition, it plays a significant role in inflammation, artheroschlerosis and cancer. We have employed the well characterized mouse fibroblastic cell line, 3T3-L1, to map the binding sites of PPARg, as well as it’s heterodimeric parther, RXR during adipocyte differentiation. Using chromatin immunoprecipitation paired-end ditags method, we have identified high-confidence binding sites for these two factors. After sequencing both libraries, we performed saturation analysis using Hill-function function and found that each library was approximately 66% saturated. Through Monte-Carlo analysis, we identified 2953 PPARg PET clusters and 5142 RXR PET clusters as significant, denoting the sets of PPARg and RXR binding regions, respectively. A subset of both binding sites was validated by realtime PCR. For both libraries the saturation using the Hill-equation was approx 66%. (as expected). We observed a good correlation between binding of PPARg and RXR within an interval of 500bp around a PPARg site. We found that binding of both PPARg and RXR was significantly enriched in the close proximity (2000bp) around known transcription start sites but no such enrichment was observed around the gene 3’ end.
 The co-occurrence of both binding sites proved to be an additional measure of confidence, since PPARg sites alone were validated to a much lesser degree. Comparative location analysis indicated that simultaneous binding of PPARg and RXR is essential for transcription of their target genes.
 In addition, expression profile analysis coupled with RNA interference for PPARg and PPARg/RXR binding site data, we have identified direct target genes that are dependent on PPARg transcriptional activity for their expression. The microarray data was analyzed using 3-way ANOVA to eliminate batch effects. We identified 1846 genes that showed PPARg-dependent differential expression during the course of adipocyte differentiation (FDR 5%). The identified genes significantly over-represented biological functions connected to lipid and fatty acid metabolism. Combining the ChIP-PET data and the expression profiles we identified a number of genes as novel direct targets of PPARg. Exploiting this information, we integrated RNA interference to characterize PPARg direct target genes essential for adipogenesis, insulin sensitization, and white/brown fat determination.
 In addition, ongoing analysis using of our binding site data around putative PPARg/RXR response elements shows a significant enrichment of a number of candidate transcription factor binding sites, at frequencies much greater than by chance. Further analysis using ChIP and RNA interference for these transcription factors will state the importance of these co-factors in adipocyte biology.
 Taken together, this genome-wide analysis of PPARg and RXR binding sites suggests a complex but definable rules governing gene regulation, adipogenesis and insulin sensitization.