Fig. S1. Validation of Chip-Seq Binding Sites by Single Gene Chip-PCR

Fig. S1. Validation of ChIP-seq binding sites by single gene ChIP-PCR

ChIP-PCR was performed on PPARγ- and RXR-enriched chromatin harvested during adipocyte differentiation at day 0 and day 6 as described in Fig. 1. Thirty out of 31 randomly selected sites displaying both PPARg and RXR binding in ChIP-seq could be validated by single gene ChIP-qPCR. Three out of three randomly selected day 6 PPARg-only sites were found to display modest occupancy of RXR as well as PPARg. Positive controls regions (Acox1, Fabp4 and Plin) and a negative control (Sox9 promoter, no gene control region (NC), myoglobin promoter and myoglobin exon 2) were included.

Fig. S2. Transactivation potential of PPARγ:RXR binding sites identified by ChIP-seq

Promoter or enhancer sequences identified as PPARγ:RXR binding sites in ChIP-seq were cloned into pGL3-basic and pGL3-promoter luciferase-reporter vectors, respectively. Constructs were transiently co-transfected with PPARã and RXRá expression vectors in NIH-3T3 cells to test for transactivation. Four of the eight (only 5 shown) PPARã:RXR associated regions supported PPARγ:RXR-mediated transactivation. For comparison, PPARã:RXR mediated transactivation of p3xPPRE-TK-Luc where three copies of the Acox PPRE in tandem were cloned in front of a minimal thymidine kinase promoter to control luciferase expression. All results are displayed as fold activity of respective vectors without transfected PPARã and RXR. Results are representative of three independent experiments performed in triplicate.

Fig. S3. Biological replicates of PPARg binding during adipocyte differentiation

PPARg ChIP-PCR was performed on three biological replicates and analyzed as described. Selected target sites adjacent to the following genes were investigated; Acox (chr11:116.060.282-116.060.362), Plin (chr7:86.879.737-86.879.817), Insig1 (chr5:28.364.246-28.364.610), Angptl4 (chr17:33.916.262-33.916.505), Alas1 (chr9:106.176.081-106.176.331), Btd (chr14:32.457.663-32.457.847) and Cidec (chr6:113.385.818-113.386.091). The data is illustrated relative to day 6 (100%) and representative of three biological experiments.

Fig. S4. Genome wide mapping of PPARg:RXR binding sites during adipocyte differentiation

(A-D) ChIP-seq was performed on PPARg and RXR enriched chromatin harvested during adipocyte differentiation (days 0-6). Target sites were detected using FindPeak peak detection (<0.001 FDR). Y-axis indicates the number of annotated tags. ChIP-seq data from days 0, 1, 2, 3, 4, 6 viewed in the UCSC browser showing PPARg and RXR binding sites for (A) Abca1, (B) Plin and Pex11a, (C) Pnpla2 and (D) Agpat2.

Fig. S6. Efficiency of PPARd and PPARg knockdown

(A) RNA was harvested from 3T3-L1 preadipocytes (day 0) exposed to the indicated shRNAi expressing lentivirus and the levels of PPARg and PPARd mRNA were determined by real-time PCR. Expression levels were normalized to expression of TFIIB. Levels in cells treated with RNAi against LacZ were set to 1. (B) Protein was harvested from 3T3-L1 preadipocytes (day 1) exposed to the indicated shRNAi expressing lentivirus. Protein levels of PPARg were determined by Western blotting using PPARg specific antibody (sc 7196). Arrows indicate the position of PPARg1 and -2, respectively.

* denotes an unspecific protein.

Fig. S7. PPARg:RXR binding and RNAPII activity

PPARg, RXR and RNAPII ChIP-seq were performed and analyzed as described. (A-F) Equalized ChIP-seq data viewed in the UCSC browser for (A) Lpin1, (B) Acsl, (C) Scd1, (D) Dgat2, (E) Taldo1, and (F) Hk2. Y-axis shows the number of annotated tags.

Fig. S8. Correlation between RNAPII occupancy, pre-mRNA and mRNA for genes during differentiation

Transcriptional activity was measured as percentage of day 0 (vertical axis) by RNAPII occupancy, and levels of pre-mRNA and mRNA at day 0, 1, 2, 3, 4 and 6 during adipogenesis (horizontal axis, in category scale). (A-E) Profiles of two genes for each of the clusters A, B, C , D and E are shown. RNAPII occupancy was obtained from ChIP-seq analyses, and levels of pre-mRNA and mRNA were measured by real-time PCR using intron-exon and exon-exon spanning primers, respectively. All pre-mRNA and mRNA levels were normalized to TFIIB mRNA and expression at day 0 was set to 100. Corresponding correlation coefficients for these and additional genes can be found in Table S3.

Fig. S9. Knock-down of PPARg inhibits induction of novel PPARg target genes early in adipogenesis

3T3-L1 cells were infected with lentivirus expressing shRNA against LacZ or PPARg as described in Fig. 4. RNA was harvested at day 0 and day 2 of adipogenesis and the levels of known and novel putative PPARg target genes involved in glycerolipid metabolism (Lipe, Fabp4, Acsl1, Agpat2, Gpat3 and Gpd1), glucose metabolism (Hk2, Taldo and Rpia) and regulation of transcription (Srebf1c, Cebpa, and PPARg), respectively, were determined by real-time PCR. Expression levels were normalized to expression of TFIIB. Expression levels at day 0 in LacZ shRNAi expressing cells were set to 1.

* denotes mRNAs significantly induced from day 0 to day 2 (P < 0,05 in Student’s t-test)

Fig. S10. Rosiglitazone induces mRNA expression of novel putative PPARg target genes

Day 6 3T3-L1-adipocytes were treated with the PPARg specific agonist rosiglitazone (BRL40653) or vehicle (DMSO) for 12 h. RNA was harvested, and the levels of known and novel putative PPARg target genes involved in glycerolipid metabolism (Fabp4, Acsl1, Agpat2, Gpat3 and Gpd1), glucose metabolism (Hk2, Taldo and Rpia) and regulation of transcription (Insig1), respectively, were determined by real time PCR. Expression levels were normalized to expression of TFIIB and mRNA levels in vehicle treated cells were set to 1. * denotes significantly induced mRNAs (P <0,05 in Student’s t-test)