Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 May 1;28(9):1018-28.
doi: 10.1101/gad.237628.114.

Anti-diabetic rosiglitazone remodels the adipocyte transcriptome by redistributing transcription to PPARγ-driven enhancers

Sonia E Step  1 , Hee-Woong Lim, Jill M Marinis, Andreas Prokesch, David J Steger, Seo-Hee You, Kyoung-Jae Won, Mitchell A Lazar
Affiliations

Anti-diabetic rosiglitazone remodels the adipocyte transcriptome by redistributing transcription to PPARγ-driven enhancers

Sonia E Step et al. Genes Dev. .

Abstract

Rosiglitazone (rosi) is a powerful insulin sensitizer, but serious toxicities have curtailed its widespread clinical use. Rosi functions as a high-affinity ligand for peroxisome proliferator-activated receptor γ (PPARγ), the adipocyte-predominant nuclear receptor (NR). The classic model, involving binding of ligand to the NR on DNA, explains positive regulation of gene expression, but ligand-dependent repression is not well understood. We addressed this issue by studying the direct effects of rosi on gene transcription using global run-on sequencing (GRO-seq). Rosi-induced changes in gene body transcription were pronounced after 10 min and correlated with steady-state mRNA levels as well as with transcription at nearby enhancers (enhancer RNAs [eRNAs]). Up-regulated eRNAs occurred almost exclusively at PPARγ-binding sites, to which rosi treatment recruited coactivators, including MED1, p300, and CBP. In contrast, transcriptional repression by rosi involved a loss of coactivators from eRNA sites devoid of PPARγ and enriched for other transcription factors, including AP-1 factors and C/EBPs. Thus, rosi activates and represses transcription by fundamentally different mechanisms that could inform the future development of anti-diabetic drugs.

Keywords: PPARγ; adipocyte; diabetes; enhancer RNA; rosiglitazone; transcription.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
Rosi rapidly increases and represses gene transcription in adipocytes. GRO-seq was performed on mature 3T3-L1 adipocytes treated for 0, 10 min, 30 min, 1 h, or 3 h with rosi. (A) Increased transcription at the Fabp4 locus with rosi treatment. (B) The Rgs2 gene shows repressed transcription upon rosi treatment. (C) The heat map shows 1951 genes that displayed a significant change in transcription (false discovery rate [FDR] <0.05) due to rosi treatment in at least one time point.
Figure 2.
Figure 2.
Adipocyte gene body transcription levels correlate with steady-state mRNA levels. Correlation between microarray mRNA levels and GRO-seq transcription levels is plotted at each time point. The Pearson correlation coefficient is given for each pair of time points.
Figure 3.
Figure 3.
Adipocyte eRNA transcription is stimulated by rosi and correlates with gene body transcription. (A) Genome-wide average signal of intergenic bidirectional transcripts in untreated adipocytes from the plus and minus strands. (B) Two bidirectional eRNAs are transcribed at enhancers upstream of the Fabp4 TSS and up-regulated by rosi treatment. eRNA centers are indicated by arrows. (C) Heat map showing all rosi-regulated eRNAs found in an unbiased manner (N = 2251). (D) Correlation between rosi-regulated eRNAs and the regulation of the nearby gene for all pairs of matching time points (N = 462). For each gene, eRNAs within 100 kb of the TSS were included in the analysis. (E) Correlation between gene and eRNA rosi regulation as measured by RT-qPCR for two example genes, Fabp4 and Pdk4. N = 3. Error bars indicate SEM.
Figure 4.
Figure 4.
Rosi-up-regulated eRNAs are dependent on genomic binding of PPARγ. (A) Top hit from Homer de novo motif search at up-regulated eRNAs. The closest known motif, the DR1 motif, is shown for reference. The enrichment for this motif was 45% in the target sites and 13.5% in the background sites. (B) Percentage of up-regulated, down-regulated, and unregulated eRNA sites that overlap with a called PPARγ peak from ChIP-seq. There were 14,604 called PPARγ peaks in our data set. Among those, 5819 sites were extragenic, and 3642 sites had eRNAs. (C) Total PPARγ tag count in reads per million (RPM) within 1 kb of up-regulated, down-regulated, and unregulated eRNA sites. (*) P = 7.7 × 10−95 versus unregulated; (**) P = 1.8 × 10−20 versus unregulated. (D) RT-qPCR of eRNAs 24 h after siRNA knockdown of PPARγ. N = 3. Error bars indicate SEM. (E) Distance from TSS to the closest PPARγ sites for regulated genes. P < 10−10 for up-regulated versus unregulated sites by χ2 test. (F) Number of PPARγ-binding sites within 100 kb of the TSS for regulated genes. P < 10−10 for up-regulated versus unregulated sites by χ2 test.
Figure 5.
Figure 5.
Enrichment of C/EBP and AP-1 factors at down-regulated eRNAs. (A) Top two hits from Homer de novo motif search at down-regulated eRNAs. The closest known motif for each is shown for reference. (B) Total C/EBPα and FOSL2 tag counts in reads per million (RPM) within 1 kb of the center of regulated eRNAs from ChIP-seq. (*) P = 4.6 × 10−8 versus unregulated; (**) P = 8.9 × 10−46 versus unregulated. (C) Percentage of regulated eRNAs that overlap with called C/EBPα or FOSL2 peaks.
Figure 6.
Figure 6.
Redistribution of MED1 genomic occupancy upon rosi treatment. (A) Number of eRNAs up-regulated or down-regulated at each time point. (B) Work flow for identification of MED1-binding sites with regulated eRNAs. (C) Scatter plot comparing MED1-binding strength in reads per million (RPM) with and without 1 h of rosi treatment, with sites containing an up-regulated eRNA highlighted in red. (D) Scatter plot comparing MED1-binding strength with and without 1 h of rosi treatment, with sites containing a down-regulated eRNA highlighted in blue. (E) MED1 ChIP-qPCR at sites with up-regulated or down-regulated eRNAs. N = 5. Error bars indicate SEM. (*) P < 0.05; (**) P < 0.01; (***) P < 0.005 by paired t-test. (F) Box plot showing average change in MED1 genome-wide occupancy upon rosi treatment at sites of regulated eRNAs. (**) P < 10−15.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Ahmadian M, Suh JM, Hah N, Liddle C, Atkins AR, Downes M, Evans RM 2013. PPARγ signaling and metabolism: the good, the bad and the future. Nat Med 19: 557–566 - PMC - PubMed
    1. Bar-Joseph Z, Gifford DK, Jaakkola TS 2001. Fast optimal leaf ordering for hierarchical clustering. Bioinformatics 17: S22–S29 - PubMed
    1. Bugge A, Grontved L, Aagaard MM, Borup R, Mandrup S 2009. The PPARγ2 A/B-domain plays a gene-specific role in transactivation and cofactor recruitment. Methods Enzymol 23: 794–808 - PMC - PubMed
    1. Chao L, Marcus-Samuels B, Mason MM, Moitra J, Vinson C, Arioglu E, Gavrilova O, Reitman ML 2000. Adipose tissue is required for the antidiabetic, but not for the hypolipidemic, effect of thiazolidinediones. J Clin Invest 106: 1221–1228 - PMC - PubMed
    1. Chawla A, Lazar MA 1994. Peroxisome proliferator and retinoid signaling pathways co-regulate preadipocyte phenotype and survival. Proc Natl Acad Sci 91: 1786–1790 - PMC - PubMed

Publication types

-