BackgroundThe present study coupled expression profiling with chromatin immunoprecipitation sequencing (ChIP-seq) to examine peroxisome proliferator-activated receptor-β/δ (PPARβ/δ)-dependent regulation of gene expression in mouse keratinocytes, a cell type that expresses PPARβ/δ in high concentration.ResultsMicroarray analysis elucidated eight different types of regulation that modulated PPARβ/δ-dependent gene expression of 612 genes ranging from repression or activation without an exogenous ligand, repression or activation with an exogenous ligand, or a combination of these effects. Bioinformatic analysis of ChIP-seq data demonstrated promoter occupancy of PPARβ/δ for some of these genes, and also identified the presence of other transcription factor binding sites in close proximity to PPARβ/δ bound to chromatin. For some types of regulation, ATF4 is required for ligand-dependent induction of PPARβ/δ target genes.ConclusionsPPARβ/δ regulates constitutive expression of genes in keratinocytes, thus suggesting the presence of one or more endogenous ligands. The diversity in the types of gene regulation carried out by PPARβ/δ is consistent with dynamic binding and interactions with chromatin and indicates the presence of complex regulatory networks in cells expressing high levels of this nuclear receptor such as keratinocytes. Results from these studies are the first to demonstrate that differences in DNA binding of other transcription factors can directly influence the transcriptional activity of PPARβ/δ.

Although

Ligands for peroxisome proliferator-activated receptor-beta/delta (PPARbeta/delta) increase skeletal muscle fatty acid catabolism, improve insulin sensitivity, increase serum high-density lipoprotein cholesterol, elicit anti-inflammatory activity and induce terminal differentiation. Contradictory findings are also reported suggesting that PPARbeta/delta ligands potentiate tumorigenesis by increasing cell proliferation, by inhibiting apoptosis through phosphorylation of Akt and by increasing cyclooxygenase-2 (COX2) and vascular endothelial growth factor (VEGF) expression. The contradictory findings could be due to differences in the model system (cancer cell line versus in vivo), differences in cell culture conditions (with and without serum) or differences in ligands. The present study examined the effect of two different PPARbeta/delta ligands (GW0742 and GW501516) in human cancer cell lines (HT29, HCT116, LS-174T, HepG2 and HuH7) cultured in the presence or absence of serum and compared in vitro analysis with in vivo analysis. Neither PPARbeta/delta ligand increased cell growth or phosphorylation of Akt and no increase in the expression of VEGF or COX2 were detected in any cancer cell line in the presence or absence of serum. Similarly, liver, colon and colon polyps from mice administered these PPARbeta/delta ligands in vivo did not exhibit changes in these markers. Results from these studies demonstrate that serum withdrawal and/or differences in ligands do not underlie the disparity in responses reported in the literature. The quantitative nature of the present findings are inconsistent with the hypothesis that cancer cell lines respond differentially as compared with normal cells, and provide further evidence that PPARbeta/delta ligands do not potentiate tumorigenesis.

Cyclooxygenase (COX) 2-derived prostaglandin E(2) (PGE(2)) promotes colorectal carcinoma growth and invasion, and inhibition of COX2 by non-steroidal anti-inflammatory drugs is known to inhibit these processes. There is controversy regarding the effect of ligand activation of peroxisome proliferator-activated receptor (PPAR)-beta/delta on colon carcinogenesis, although collective evidence from independent laboratories suggest that ligand activation of PPARbeta/delta leads to the induction of terminal differentiation coupled with inhibition of cell growth in a variety of models. The present study examined the hypothesis that ligand activation of PPARbeta/delta and inhibition of COX2 attenuate colon cancer through independent mechanisms and that combining these two mechanisms will enhance this inhibition. Colon cancer was induced by administering azoxymethane to wild-type and PPARbeta/delta-null mice. Cohorts of mice were treated with GW0742 (a PPARbeta/delta ligand), nimesulide (a COX2 inhibitor) or a combination of GW0742 and nimesulide. Inhibition of COX2 by nimesulide attenuated colon cancer and ligand activation of PPARbeta/delta by GW0742 had inhibitory effects. However, the combined treatment of GW0742 and nimesulide did not cause an enhancement in the attenuation of colon cancer. Mechanistically, the effects of these compounds occurred through independent mechanisms as increased levels of differentiation markers as a result of ligand activation of PPARbeta/delta were not found with COX2 inhibition, and a reduction in PGE(2) levels resulting from COX2 inhibition was not observed in response to ligand activation of PPARbeta/delta. Results from these studies effectively dissociate COX2 inhibition and PPARbeta/delta activity during colon carcinogenesis.