The repressor element 1-silencing transcription factor regulates heart-specific gene expression using multiple chromatin-modifying complexes

Abstract

Cardiac hypertrophy is associated with a dramatic change in the gene expression profile of cardiac myocytes. Many genes important during development of the fetal heart but repressed in the adult tissue are reexpressed, resulting in gross physiological changes that lead to arrhythmias, cardiac failure, and sudden death. One transcription factor thought to be important in repressing the expression of fetal genes in the adult heart is the transcriptional repressor REST (repressor element 1-silencing transcription factor). Although REST has been shown to repress several fetal cardiac genes and inhibition of REST function is sufficient to induce cardiac hypertrophy, the molecular mechanisms employed in this repression are not known. Here we show that continued REST expression prevents increases in the levels of the BNP (Nppb) and ANP (Nppa) genes, encoding brain and atrial natriuretic peptides, in adult rat ventricular myocytes in response to endothelin-1 and that inhibition of REST results in increased expression of these genes in H9c2 cells. Increased expression of Nppb and Nppa correlates with increased histone H4 acetylation and histone H3 lysine 4 methylation of promoter-proximal regions of these genes. Furthermore, using deletions of individual REST repression domains, we show that the combined activities of two domains of REST are required to efficiently repress transcription of the Nppb gene; however, a single repression domain is sufficient to repress the Nppa gene. These data provide some of the first insights into the molecular mechanism that may be important for the changes in gene expression profile seen in cardiac hypertrophy.

FIG. 1.

Ectopic REST expression prevents increased Nppb and Nppa mRNA levels in adult rat ventricular myocytes in response to ET-1. Data are changes in cyclophilin (Cyc), Nppb, and Nppa mRNA levels in cells infected with a control adenovirus (Ad) or REST-expressing adenovirus treated with ET-1 relative to untreated cells (means ± standard errors of the means; n = 4). *, P < 0.05.

FIG. 2.

H9c2 cells express REST. (a) PCR from H9c2 RNA either treated with reverse transcriptase (lane 1) or left untreated (lane 2). M, 1-kb Plus DNA ladder (Life Technologies). (b to d) Immunohistochemistry of H9c2 cells with anti-REST serum (b and c) or preimmune serum (d). DAPI-stained nuclei appear blue, and REST protein detected by FITC staining appears green. The DAPI and FITC images for anti-REST serum are shown separately (panels b and c, respectively), whereas the two images for preimmune serum have been overlaid (d). (e) Gel mobility shift assay of the Nppb (left panel) and Nppa (right panel) gene RE1 sites with 20 μg of nuclear extracts from H9c2 cells in the presence of no competitor (None), an RE1 site from the Chrm4 gene (Chrm4 RE1), an unrelated binding site (Sp1), or RE1 sites from the Nppa and Nppb genes (Nppa RE1 and Nppb RE1, respectively). Specific DNA-protein complexes are indicated by the filled arrow. REST protein was identified in the complex using an anti-REST antibody (α-REST), indicated by the open arrow, and an anti-Sp1 antibody (α-Sp1) was used to control for specificity.

FIG. 3.

REST represses Nppb and Nppa transcription in H9c2 cells. (a) Quantification of Nppb RE1, Nppa RE1, and control sequences precipitated by anti-REST (solid bars) and anti-Myc (hatched bars) antibodies from control adenovirus (Ad)- and DN-REST-infected H9c2 cells (means ± standard errors of the means; n = 4). Dashed line represents level of background precipitation. (b) Quantitative RT-PCR of cyclophilin (Cyc), Nppb, and Nppa mRNA levels in H9c2 cells infected with a control adenovirus (Ad) (solid bars) or DN-REST (hatched bars) relative to those in uninfected control cells (means ± standard error of the means; n = 3). *, P < 0.05.

FIG. 4.

Derepression of Nppb and Nppa in H9c2 cells is associated with increased histone acetylation. (a) Diagram indicating the locations of Nppb RE1 and Nppa RE1 sites in relation to the Nppb and Nppa genes on rat chromosome 5. Sequences amplified by PCR for ChIP experiments are represented by three double-headed arrows. The transcription initiation sites are indicated by arrows. Open boxes, noncoding sequences; hatched boxes, coding sequences. (b) Quantification of Nppb RE1, Nppa promoter (Nppa Pro), Nppa RE1, and control sequences precipitated by an acetylated histone H4 antiserum in uninfected H9c2 cells (solid bars), H9c2 cells infected with a control adenovirus (Ad) (hatched bars), or H9c2 cells infected with DN-REST (shaded bars) (means ± standard errors of the means; n = 5). *, P < 0.05. (c) Quantitative RT-PCR of cyclophilin (Cyc), Nppb, and Nppa mRNA levels in H9c2 cells infected with a control adenovirus (Ad) (solid bars) or with an adenovirus containing DN-REST (hatched bars) in the presence of TSA, expressed relative to levels in vehicle-treated control cells (means ± standard errors of the means; n = 4). *, P < 0.05.

FIG. 5.

Derepression of Nppb and Nppa in H9c2 cells is associated with increased dimethyl H3K4 levels. (a) Quantification of Nppb RE1, Nppa promoter (Nppa Pro), Nppa RE1, and control sequences precipitated by an anti-dimethyl H3K4 antibody in uninfected H9c2 cells (solid bars), H9c2 cells infected with a control adenovirus (Ad) (hatched bars), or H9c2 cells infected with an adenovirus containing DN-REST (shaded bars) (means ± standard errors of the means). *, P < 0.05. (b) Quantitative RT-PCR of cyclophilin (Cyc), Nppb, and Nppa mRNA levels in DN-REST-expressing H9c2 cells in the presence (solid bars) or absence (hatched bars) of MTA expressed relative to those in vehicle-treated control cells (means ± standard errors of the means; n = 5). *, P < 0.05.

FIG. 6.

Both REST repression domains are required for Nppb but not Nppa repression in H9c2 cells. (a) Gel mobility shift assay of the Nppb gene RE1 site with 6 μg of nuclear extracts from C-REST-, N-REST-, and DN-REST-infected H9c2 cells in the presence of no competitor, an RE1 site from the Chrm4 gene (RE1), or an unrelated binding site (Sp1). Specific DNA-protein complexes are indicated by a open (C-REST and N-REST) or filled (DN-REST) arrow. (b) Quantitative RT-PCR of cyclophilin (Cyc), Nppb, and Nppa mRNA levels in H9c2 cells infected with a control adenovirus (Ad) (solid bars), C-REST (hatched bars), or N-REST (shaded bars) relative to those in uninfected control cells (means ± standard errors of the means; n = 3).

FIG. 7.

Chromatin changes associated with C-REST and N-REST in H9c2 cells. (a and b) Quantification of Nppb RE1, Nppa promoter (Nppa Pro), Nppa RE1, and control sequences precipitated by anti-acetylated H4 (a) or anti-dimethyl H3K4 (b) in H9c2 cells infected with a control adenovirus (Ad), C-REST, or N-REST (means ± standard errors of the means; n = 3).

FIG. 8.

Derepression of Nppb in H9c2 cells is not due to squelching. (a) Quantitative RT-PCR of ectopically expressed REST, C-REST, or N-REST introduced using 2.5 × 10¹¹ or 5 × 10¹¹ PFU/ml of adenovirus. (b) Quantitative RT-PCR of Nppb mRNA in the presence of ectopically expressed REST, C-REST, or N-REST introduced using 2.5 × 10¹¹ or 5 × 10¹¹ PFU/ml of adenovirus.

FIG. 9.

Chromatin changes in adult rat ventricular myocytes in response to ET-1. (a and b) Quantification of Nppb RE1, Nppa promoter (Nppa Pro), Nppa RE1, and control sequences precipitated by anti-acetylated H4 (a) or anti-dimethyl H3K4 (b) in control (solid bars) and ET-1 treated (hatched bars) rat ventricular myocytes (means ± standard error of the means; n = 4).