PPARβ/δ recruits NCOR and regulates transcription reinitiation of ANGPTL4

Figure 1.. PPARβ/δ inverse agonists suppress transcription initiation of ANGPTL4 and recruitment of TFIIB to a scaffold complex. Caki-1 cells were treated for 30 min as indicated and subjected to ChIP-qPCR analysis. (A) Scanning ChIP-qPCR of the ANGPTL4 locus. An antibody against the CTD of the large subunit of RNAPII was used. (B) Binding of TFIIA, TFIIB and RPB1 at the ANGPTL4 and TSC22D3 TSSs. (C) TFIIB binding at the ANGPTL4 TSS in the presence of TGFβ, PT-S264, or both. The right panel shows scanning ChIP around the ANGPTL4 TSS from the same samples. Commercially available antibodies are denominated by the terms in parentheses. Means and standard deviations of three technical replicates from representative ChIP assays are plotted.

Figure 2.. PPARβ/δ inverse agonists prevent TGFβ-stimulated recruitment of Mediator subunits to the ANGPTL4 promoter. Caki-1 cells were treated for 30 min as indicated and subjected to scanning ChIP-qPCR analysis with primer pairs that amplify regions close to the TSS of ANGPTL4. Antibodies against MED1(A), MED26 (B) and MED13L (D) were used. Panels (C and F) show ChIP-qPCR data from these samples using a negative control region primer pair. In (E) RNAPII binding is shown as an additional control. Commercially available antibodies are denominated by the terms in parentheses. Means and standard deviations of three technical replicates from representative ChIP assays are plotted.

Figure 3.. NCOR and SMRT complex subunits are recruited to PPARβ/δ in the presence of PT-S264. MDA-MB231-luc2 cells were treated as indicated for 30 min, and ChIP-qPCR was performed with antibodies against subunits of HDAC3-containing complexes using primer pairs encompassing the ANGPTL4 and PDK4 PPREs (+3500 or −12 200 bp relative to the TSS of the gene, respectively). (A) Binding of NCOR in the presence of 300 nM PT-S264, 1 ng/ml TGFβ, or both. Two different antibodies against NCOR were used. (B) Binding of PPARβ/δ, RXR, HDAC3, TBLR1, NCOR, and SMRT in the presence of solvent (DMSO), 300 nM PT-S264, or 1 μM L165,041. Both panels: representative ChIP data are plotted. Commercially available antibodies are denominated by the terms in parentheses. Values denote means and standard deviations from three technical replicates.

Figure 4.. Analysis of PPAR isoform binding to chromatin in PPARβ/δ knockout clones. ChIP-qPCR analysis of PPARα, PPARβ/δ, PPARγ, and RXR binding at the ANGPTL4 and PDK4 loci in the four MDA-MB231-luc2 PPARβ/δ knockout clones (2B2, 2B3, 2A3, and 2A6). Parental cells (WT) were processed in parallel. A representative experiment is shown. The error bars denote standard deviations from three technical replicates.

Figure 5.. A retroviral reconstitution screen identifies PPARβ/δ mutants deficient in basal repression and response to PT-S264. MDA-MB231-luc2 2B3 PPARD KO cells stably expressing the murine ecotropic receptor were transduced with retroviruses derived from the empty vector or carrying PPARD cDNAs. The WT cDNA and a panel of 80 mutants (N between 1 and 8 per construct) were used. ANGPTL4 expression was measured by RT-qPCR after treatment for 6 h. Values were normalized to the RPL27 transcript. (A) Effects on basal repression and PT-S264-dependent repression. ANGPTL4 expression was plotted on the x-axis as ΔΔ Ct(basal WT–basal mutant) versus ΔΔ Ct(PT-S264 WT–PTS264 mutant) on the y-axis. Open triangles denote mutants with modified basal repression (arbitrary cutoff of two PCR cycles relative to WT). Mean Ct values relative to the Ct value obtained from cells expressing the WT PPARD cDNA (N= 20, mean Ct 28.9) were used. (B) The relative expression of ANGPTL4 was measured in cells transduced with the empty vector (N≥ 3), a retrovirus carrying the C91A-E92A DNA binding mutant PPARD cDNA (N≥ 4), the WT (N ≥ 8), the T427I (N ≥ 6), the K421A-K422A (N ≥ 5) or the R314A mutant (N ≥ 3) cDNAs and…

Figure 6.. Corepressor recruitment by PPARβ/δ mutants deficient in basal repression. 2B3 PPARD KO cells transduced with the indicated constructs were treated with solvent, 1 μM L165,041 or 300 nM PT-S264 for 30 min and subjected to ChIP-qPCR analysis using antibodies against NCOR or SMRT. The PPAR binding sites of ANGPTL4 (+3500 bp relative to the TSS, (A), PDK4 (−12 200 bp, (B), PLIN2 (−34 300 bp, (C), and a negative control region (D) were amplified as indicated. Non-cognate IPs (IgG) were processed in parallel from the solvent-treated cells. Means and standard deviations of three technical replicates from a representative ChIP assay are plotted.

Figure 7.. Regulation of RNA polymerase II binding by PPARβ/δ mutants deficient in basal repression. 2B3 PPARD KO cells transduced with the indicated constructs were treated with solvent or 300 nM PT-S264 for 30 min and subjected to ChIP-qPCR analysis using an antibody against RNA polymerase II. The TSS regions of ANGPTL4 (A), PDK4 (B), PLIN2 (C), ABCA1 (D), a negative control region (E), and the TSC22D3 TSS (F) as an additional negative control were amplified as indicated. Means and standard deviations of three technical replicates from a representative ChIP assay are plotted.

Figure 8.. HDAC inhibitor sensitivity of PT-S264-mediated repression. Parental MDA-MB231-luc2 and 2B3 PPARD KO cells (A) or 2B3 PPARD KO cells transduced with the indicated constructs (B) were treated with solvent, 300 nM PT-S264, 500 nM TSA, and 250 nM apicidin as indicated for 6 h, and RT-qPCR with primers against the ANGPTL4, PDK4, PLIN2, and TSC22D3 transcripts was performed. Means and standard deviations of (A) n= 3 independent experiments or (B) n = 3–4 independent experiments are plotted. Mean normalized Ct values in the control condition are indicated. Significance levels were calculated via paired t-tests. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; n.s., not significant.

Figure 9.. Ligand-dependent regulation of ANGPTL4 transcription initiation by PPARβ/δ and NCOR/SMRT. Data from this study show that the PPARβ/δ inverse agonist PT-S264 interferes with activator-stimulated recruitment of MED1, MED13L, MED26, TFIIB, and RNAPII. Mutations of PPARβ/δ which abrogate NCOR and SMRT binding in the basal state allow for increased binding of RNAPII, and NCOR and SMRT binding is restored upon addition of PT-S264. The latter observations were made at other PPAR target genes as well. Taken together with our previous observation that synthetic PPAR agonists together with other activating stimuli synergistically induce ANGPTL4 transcription (18), we propose the model that basal and ligand-dependent repressor recruitment limit the ability of activators to induce transcription reinitiation (and possibly initiation) via NCOR, SMRT, or both. The corepressors use both deacetylase-dependent and deacetylase-independent mechanisms, the latter of which are not elucidated as of now.