Histone code modifications repress glucose transporter 4 expression in the intrauterine growth-restricted offspring

Abstract

We examined transcriptional and epigenetic mechanism(s) behind diminished skeletal muscle GLUT4 mRNA in intrauterine growth-restricted (IUGR) female rat offspring. An increase in MEF2D (inhibitor) with a decline in MEF2A (activator) and MyoD (co-activator) binding to the glut4 promoter in IUGR versus control was observed. The functional role of MEF2/MyoD-binding sites and neighboring three CpG clusters in glut4 gene transcription was confirmed in C2C12 muscle cells. No differential methylation of these three and other CpG clusters in the glut4 promoter occurred. DNA methyltransferase 1 (DNMT1) in postnatal, DNMT3a, and DNMT3b in adult was differentially recruited with increased MeCP2 (methyl CpG-binding protein) concentrations to bind the IUGR glut4 gene. Covalent modifications of the histone (H) code consisted of H3.K14 de-acetylation by recruitment of histone deacetylase (HDAC) 1 and enhanced association of HDAC4 enzymes. This set the stage for Suv39H1 methylase-mediated di-methylation of H3.K9 and increased recruitment of heterochromatin protein 1alpha, which partially inactivates postnatal and adult IUGR glut4 gene transcription. Further increased interactions in the adult IUGR between DNMT3a/DNMT3b and HDAC1 and MEF2D and HDAC1/HDAC4 and decreased association between MyoD and MEF2A existed. We conclude that epigenetic mechanisms consisting of histone code modifications repress skeletal muscle glut4 transcription in the postnatal period and persist in the adult female IUGR offspring.

Figures

FIGURE 1.

In vitro studies. A, scheme of the rat glut4 upstream region. The rat glut4 promoter region (1 kb) contains the three CpG islands, MyoD-I, MyoD-II, and the MEF2 DNA binding consensus elements 5′ to the transcriptional (+1) and translational (ATG) start sites. In addition, NF-1, Sp-1, PU.1, and WT-1 consensus sequences are shown. The sequences deleted in mutant (Δ) constructs are depicted in parentheses. B, transient transfection of glut4-luciferase DNA constructs in C2C12 murine skeletal muscle cell line. Luciferase reporter activity of wild type (WT) and mutated glut4-luciferase DNA constructs is depicted as a ratio to that of the β-galactosidase gene. Differences between the wild type and various mutated constructs were established by analysis of variance (F = 11.6; p = 0.0001), and inter-group differences in comparison with the WT DNA construct were determined by the post-hoc Tukey's HSD test (* and **). Mutant CpG-II and MyoD-II activate (Ac) and MEF2 de-activates (de-Ac) luciferase gene transcription. C, DNA methylation. glut4-luciferase DNA construct was either pretreated or untreated with the SssI methylase enzyme. Top panel, demonstrates the methylated and unmethylated glut4 DNA that was restricted with methylation-sensitive HpaII and BstU1 enzymes, and the digested DNA products were separated by gel electrophoresis along with the unrestricted DNA. Although the unmethylated glut4 DNA was restricted yielding smaller products, the SssI methylase-treated methylated DNA was resistant to restriction enzyme digestion yielding larger DNA products. M = DNA markers. Lower panel demonstrates the luciferase activity of SssI methylase-treated and untreated glut4-luciferase (0–4 μg) transiently transfected in C2C12 cells. Differences between DNA concentration-matched methylated and unmethylated DNA-reporter activity was assessed by the Student's t test (*). D, histone de-acetylation. Transiently transfected C2C12 cells containing wild type (WT) and mutated glut4-luciferase DNA constructs were pretreated with trichostatin A (0.6 μm), a generic HDAC inhibitor. The mutant constructs were compared with the WT construct by analysis of variance (F = 96.1; p = 0.0001), and inter-group differences were established by the post-hoc Tukey's HSD test (*, **). E, in vitro nutrient restriction. The transiently transfected cells containing the wild type glut4-luciferase DNA construct were incubated in either a control or a nutrient-restricted (glucose and amino acid deprived) medium. Luciferase activity is depicted as a ratio to the β-galactosidase gene (internal control) per protein concentration. Student's t test demonstrated inter-group differences (*).

FIGURE 2.

Adult studies. A and B, skeletal muscle glut4 mRNA. Representative Northern blots demonstrating glut4 mRNA (2.8 kb) (top panel) and the internal control β-actin (1.8 kb) mRNA (bottom panel) in skeletal muscle of 450-day-old control (CON) and IUGR female (A) and male (B) rats. The quantification by PhosphorImager analysis of glut4 mRNA in arbitrary units is depicted as a ratio to β-actin and represented as a percent of control (CON). Inter-group differences were established by the Student's t test (*). NS, not significant. C and D, DNA methyltransferases and glut4 5′-upstream region. C, DNMT3a. Panels I and II, representative 2% agarose gels demonstrate the input PCR glut4 and GAPDH control without an antibody (left panels), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panels), and ChIP assay demonstrating the 384-bp PCR glut4 DNA amplification product, which contains the MEF2- and MyoD-I-binding sites and CpG-I and CpG-II regions (panel I) or the 428-bp PCR glut4 DNA amplification product containing MEF2, MyoD-II, and CpG-III regions (panel II), and the 230-bp PCR GAPDH DNA amplification product (served as an internal control) obtained from DNMT3a nuclear immunoprecipitates (IP)(right panels). M = DNA size markers, C = control, and I = IUGR. Arrowheads show the glut4 and GAPDH DNA bands. Panels III and IV, quantification of the amplified 384-bp (panel III) and 428-bp (panel IV) glut4 DNA product as a ratio to that of GAPDH, corrected for the input control and expressed as a percent of CON. Difference between CON and IUGR was assessed by the Student's t test (*). Panel V, representative Western blot demonstrating co-immunoprecipitation (co-IP) experiments where DNMT3a was detected in association with input nuclear protein (Input) obtained from CON and IUGR, anti-HDAC1 IgG chromatin IPs (HDAC1 IP) with increased association of DNMT3a observed in IUGR versus CON, and no DNMT3a detected in nonspecific IgG nuclear chromatin IP (IgG) of CON and IUGR. D, DNMT3b. Panels I and II, representative 2% agarose gels demonstrate the input PCR glut4 and GAPDH control without an antibody (left panels), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panels), and ChIP assay demonstrating the 384-bp PCR glut4 DNA (panel I) or the 428-bp PCR glut4 DNA (panel II) and the 230-bp PCR GAPDH DNA amplification products obtained from DNMT3b nuclear chromatin IPs (right panels). M = DNA size markers, C = control, and I = IUGR. Panels III and IV, quantification of the amplified 384-bp (panel III) and 428-bp (panel IV) glut4 DNA product as a ratio to that of GAPDH corrected for the input control and expressed as a percent of CON. Difference between CON and IUGR was assessed by the Student's t test (*). NS = not significant. Panel V, representative Western blot demonstrating co-IP experiments where DNMT3b was detected in association with input nuclear protein (Input) obtained from CON and IUGR, anti-HDAC1 IgG nuclear chromatin IPs (HDAC1 IP) with slightly increased association of DNMT3b observed in IUGR versus CON, and no DNMT3b detected in IgG nuclear chromatin IPs (IgG) of CON and IUGR.

FIGURE 3.

Adult studies, MEF2 nuclear protein and glut4 DNA. A, top panel, representative Western blot demonstrating total nuclear MEF2 protein concentrations in CON and IUGR skeletal muscle with the nuclear marker Lamin A protein serving as an internal loading control. Bottom panel, quantification of MEF2 protein concentrations as a ratio to Lamin A protein is depicted as a percent of CON. Inter-group difference was assessed by Student's t test (*). B, top panel, representative 2% agarose gels demonstrate the input PCR glut4 and GAPDH control without an antibody, in the presence of nonspecific (–) and anti-polymerase (+) IgGs and ChIP assay demonstrating the 384-bp PCR glut4 and the 230-bp PCR GAPDH DNA amplification products obtained from total MEF2 nuclear chromatin IPs. M = DNA size markers, C = control, and I = IUGR. Bottom panel, quantification of the ChIP glut4 amplification product represented as a ratio to that of GAPDH, corrected for the input control, and shown as a percent of CON. Inter-group difference is assessed by Student's t test (*). C, representative polyacrylamide gel demonstrating two gel-shifted bands (arrows) in the presence of a 32P-end-labeled DNA probe (free probe) that spans the glut4 gene containing the wild type (WT) MEF2-binding site that are not seen in the presence of MUT MEF2-binding site in C2C12 cell (positive control), CON and IUGR skeletal muscle nuclear extracts. D, representative polyacrylamide gel demonstrating competition for the two gel-shifted bands (arrows) in the presence of a 32P-end-labeled DNA probe that spans the glut4 gene containing the wild type MEF2-binding site and skeletal muscle nuclear extracts from CON and IUGR groups with increasing concentrations (10× to 100×) of unlabeled probe. E, representative polyacrylamide gel demonstrating supershifted bands (double arrow) in the presence of a 32P-end-labeled glut4 DNA probe containing the wild type MEF2-binding site, skeletal muscle nuclear extracts from control (C) and IUGR (I) groups, and anti-MEF2C, anti-MEF2D, anti-MEF2A IgGs and in the absence of IgG (–ve). F, quantification of the MEF2D and MEF2A supershift bands in control (CON) and IUGR groups. Difference between the two groups was assessed by Student's t test (*).

FIGURE 4.

Adult studies, MyoD nuclear protein and glut4 DNA. A, representative 2% agarose gels demonstrate the input PCR glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panel), and ChIP assay demonstrating the 384-bp PCR glut4 DNA and the 230-bp PCR GAPDH DNA amplification products obtained from total MyoD nuclear chromatin IPs (right panel). M = DNA size markers, C = control, and I = IUGR. Arrowheads demonstrate the glut4 and GAPDH DNA bands. B, representative 2% agarose gels demonstrate the input PCR glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) or anti-polymerase II (+) IgGs (middle panel), and ChIP assay demonstrating the 428-bp PCR glut4 DNA and the 230-bp PCR GAPDH DNA amplification products obtained from total MyoD nuclear chromatin IPs (right panel). M = DNA size markers, C = control, and I = IUGR. Arrowheads show the glut4 and GAPDH DNA bands. C, quantification of the ChIP glut4 amplification product represented as a ratio to that of GAPDH, corrected for the input control and shown as a percent of CON. Inter-group difference between MyoD-I and MyoD-II compared with their respective CON was assessed by Student's t test (*). D, representative polyacrylamide gel demonstrating a single gel-shifted band (arrows) in the presence of a 32P-end-labeled DNA probe (free probe) that spans the glut4 gene containing the wild type (WT) MyoD-I-binding site, which is not seen in the presence of MUT MyoD-I-binding site and nuclear extracts obtained from C2C12 cells (positive control), CON, and IUGR skeletal muscle. This gel shift band is competed effectively by excess (10×, 100×) unlabeled DNA probe in the CON skeletal muscle nuclear extract with the WT-labeled DNA as the probe. E, representative polyacrylamide gel demonstrating a single gel-shifted band (arrows) in the presence of a 32P-end-labeled DNA probe (free probe) that spans the glut4 gene containing the wild type (WT) MyoD-II-binding site, which is more intense in the presence of MUT MyoD-II-binding site and nuclear extracts obtained from C2C12 cells (positive control), CON, and IUGR skeletal muscle. This gel shift band is only partially competed by excess (10×, 100×) unlabeled DNA probe in the CON skeletal muscle nuclear extract with the WT-labeled DNA as the probe. F, quantification of the MyoD-I gel shift band in CON and IUGR groups shown as a percent of CON. Difference between the two groups was assessed by Student's t test (*). G and H, MEF2 and MyoD protein-protein interaction. G, representative Western blot demonstrated MyoD protein band (arrow) in CON and IUGR input chromatin in the absence (input) or presence of nuclear MEF2A IP or in the presence of a nonspecific IgG (IgG) alone. H, representative Western blot demonstrated MEF2A protein band (arrow) in CON and IUGR input chromatin in the absence (input) or presence of nuclear MyoD-IP or in the presence of a nonspecific IgG (IgG) alone.

FIGURE 5.

Adult studies, acetylated histones associated with the glut4 gene. A, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (2nd panel), and ChIP assay demonstrating the PCR amplification products of the 384-bp glut4 DNA or 230-bp GAPDH DNA (internal control) from CON (C) and IUGR (I) skeletal muscle chromatin in the presence of either anti-acetyl-histone 3 IgG (3rd panel) or the anti-acetyl-histone 3.lysine 14 amino acid IgG (right panel). B, quantification of the PCR glut4 amplification product in the acetyl H3 chromatin IP (panel I) and the acetyl-H3.K14 chromatin IP (panel II) as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*). Histone deacetylase enzyme interaction with the glut4 gene is shown. C, top panel, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panel), and ChIP assay demonstrating the PCR amplification products of the 428-bp glut4 DNA or 230-bp GAPDH DNA from CON (C) and IUGR (I) skeletal muscle chromatin in the presence of anti-HDAC1 IgG. Bottom panel, quantification of the PCR glut4 amplification product in the HDAC1 chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed in arbitrary units because the product was undetectable in CON. Inter-group difference was established by the Student's t test (*) after arbitrarily assigning a value of 0.1 to CON. D, top panel, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panel) and ChIP assay demonstrating the PCR amplification products of the 384-bp glut4 DNA or 230-bp GAPDH DNA from CON (C) and IUGR (I) skeletal muscle chromatin in the presence of anti-HDAC4 IgG. Bottom panel, quantification of the PCR glut4 amplification product in the HDAC4 chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*). MEF2D and HDAC interaction is shown. E, representative Western blot demonstrates MEF2D protein band (arrow) in CON and IUGR input nuclear protein in the absence (input) or presence of HDAC1 IP (HDAC1 IP) or in the presence of a nonspecific IgG (IgG) alone. F, representative Western blot demonstrates MEF2D protein band (arrow) in CON and IUGR input nuclear protein in the absence (input) or presence of nuclear HDAC4 IP (HDAC4 IP), or in the presence of a nonspecific IgG (IgG) alone.

FIGURE 6.

Adult studies, dimethylated histone 3 lysine residues associated with the glut4 gene. A, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (2nd panel), and ChIP assay demonstrating the PCR amplification product of the 384-bp glut4 DNA or 230-bp GAPDH DNA from CON (C) and IUGR (I) skeletal muscle chromatin in the presence of either anti-dimethyl H3.K9 IgG (3rd panel) or anti-dimethyl H3.K4 IgG (right panel). B, quantification of the PCR glut4 amplification product in the dimethyl H3.K9 chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*). C, quantification of the PCR glut4 amplification product in the dimethyl H3.K4 chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test, NS = not significant. Histone 3, lysine 9 methylase associated with the glut4 gene. D, top panel, representative Western blot demonstrating total nuclear Suv39H1 (H3.K9 dimethylase) protein concentrations in CON and IUGR skeletal muscle with the nuclear marker Lamin A protein serving as an internal loading control. Bottom panel, quantification of Suv39H1 protein concentrations as a ratio to Lamin A protein is depicted as a percent of CON. Inter-group difference was assessed by Student's t test (*). E, top panel, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panel), and ChIP assay demonstrating the PCR amplification product of the 384-bp glut4 DNA or 230-bp GAPDH DNA (internal control) from CON (C) and IUGR (I) skeletal muscle chromatin in the presence of anti-Suv39H1 methylase IgG. Bottom panel, quantification of the PCR glut4 amplification product in the Suv39H1 methylase chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*). Heterochromatin protein 1α (HP1α) association with the glut4 gene. F, top panel, representative Western blot demonstrating total nuclear HP1α concentrations in CON and IUGR skeletal muscle with the nuclear marker Lamin A protein serving as an internal loading control. Bottom panel, quantification of HP1α concentrations as a ratio to Lamin A protein is depicted as a percent of CON. Inter-group difference was assessed by Student's t test (*). G, top panel, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 and GAPDH control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panel), and ChIP assay demonstrating the PCR amplification product of the 384-bp glut4 DNA or 230-bp GAPDH DNA (internal control) from CON (C) and IUGR (I) skeletal muscle chromatin in the presence of anti-HP1α IgG. Bottom panel, quantification of the PCR glut4 amplification product in the HP1α chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*).

FIGURE 7.

Postnatal studies, DNMT1, MyoD, and MEF2 binding to the glut4 gene. A–C, top panel, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 (G4) and GAPDH (GP) control without an antibody (left panel), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (middle panel), and ChIP assay demonstrating the PCR amplification product of the 384-bp glut4 DNA or 230-bp GAPDH DNA from 2-day CON (C) and IUGR (I) skeletal muscle chromatin in the presence of either anti-DNMT1 (A), anti-MyoD (B), or anti-total MEF2 (C) IgG. Bottom panels, quantification of the PCR glut4 amplification product in either DNMT1 (A), MyoD (B), or total MEF2 (C) chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*).

FIGURE 8.

Postnatal studies, histone modifications, de-acetylation and di-methylation. A–G, top panels, representative 2% agarose gels demonstrate the input chromatin PCR-amplified glut4 (G4) and GAPDH (GP) control without an antibody (not shown), in the presence of nonspecific (–) and anti-polymerase II (+) IgGs (left panels) and ChIP assay (right panels) demonstrating the PCR amplification product of the 384-bp glut4 DNA or 230-bp GAPDH DNA (internal control) from 2-day CON (C) and IUGR (I) skeletal muscle chromatin in the presence of either anti-acetyl H3 (A), anti-acetyl H3.K14 (B), anti-HDAC1 (C), anti-HDAC4 (D), anti-H3-dimethyl (DiMe) K9 (E), anti-SUV39H1 methylase (F), or anti-HP1α IgG (G). Bottom panels, quantification of the PCR glut4 amplification product in the acetyl H3 (A), acetyl H3.K14 (B), HDAC1 (C) HDAC4 (D), H3-dimethyl K9 (E), SUV39H1 methylase (F), or HP1α (G) chromatin IP as a ratio to the GAPDH DNA product corrected for the input control and expressed as a percent of CON. Inter-group difference was established by the Student's t test (*).

FIGURE 9.

Schematic representation of the projected combinatorial progression of epigenetic changes leading to a decrease in SM glut4 transcription and expression in the female adult IUGR offspring. Rpr = repressor, Act = activator.