Class II histone deacetylases downregulate GLUT4 transcription in response to increased cAMP signaling in cultured adipocytes and fasting mice

Abstract

Insulin-mediated glucose uptake is highly sensitive to the levels of the facilitative glucose transporter protein, GLUT4. Repression of GLUT4 expression is correlated with insulin resistance in adipose tissue. We have shown that differentiation-dependent GLUT4 transcription was under control of class II histone deacetylases (HDACs). We hypothesized that HDACs may regulate gene expression in adipocytes as a result of adrenergic activation. To test this hypothesis, we activated cAMP signaling in 3T3-L1 adipocytes and in mice after an overnight fast. Chromatin immunoprecipitation experiments showed the association of HDAC4/5 with the GLUT4 promoter in vivo and in vitro in response to elevated cAMP. Knockdown of HDACs by small interfering RNA in cultured adipocytes prevented the cAMP-dependent decrease in GLUT4 transcription. HDAC4/5 recruitment to the GLUT4 promoter was dependent on the GLUT4 liver X receptor (LXR) binding site. Treatment of cells with an LXR agonist prevented the cAMP-dependent decrease in GLUT4 transcription. A loss of function mutation in the LXR response element was required for cAMP-dependent downregulation of GLUT4 expression in vitro, in fasted mice, and in mice subjected to diet-induced obesity. This suggests that activation of LXR signaling can prevent loss of GLUT4 expression in diabetes and obesity.

FIG. 1.

Class II HDACs are functionally redundant, but HDAC5 is the primary GLUT4 regulator in 3T3-L1 preadipocytes. Preadipocytes were transiently transfected with scrambled siRNA or HDAC4/5/9-specific siRNA, as indicated by +. The cells were used for the following experiments: A: After 3 days of siRNA treatment, cells were transiently transfected again with pcDNA3 (empty vector), MEF2A and GEF (transcription factors), or human HDAC4/human HDAC5, as indicated. Data are from at least three independent experiments and are expressed as mean and SEM and analyzed by one-way ANCOVA. *P < 0.05 (statistically significant) over scrambled siRNA, empty vector control. #P < 0.05 (statistically significant) over all other conditions. B: After 3 days of siRNA treatment, cells transiently transfected again with luciferase reporter plasmids -895-hGLUT4-luc (for promoter activity) and pRLTK (for transfection efficiency), with and without plasmids encoding GEF, MEF2A, wild-type HDAC4, and wild-type HDAC5. Mean data and SEM from three independent experiments done in duplicate are shown; data were analyzed by a one-way ANCOVA. #P < 0.05 (statistically significant) over scrambled siRNA, reporters alone. *P < 0.05 (statistically significant) over all other conditions. C: After 3 days of siRNA treatment, ChIP was performed using anti-HDAC4, anti-HDAC5, and nonimmune IgG antibodies. q-PCR was used to analyze the results. A ratio of expression to input was taken and normalized to control nonimmune IgG. Mean and SEM from three independent experiments is shown; data were analyzed by a one-way ANCOVA. #P < 0.003 (statistically significant) increase over background nonimmune IgG.

FIG. 2.

Forskolin increased intracellular cAMP levels, induced HDAC4 nuclear localization, and downregulated the GLUT4 promoter. A: SDS-PAGE was used to analyze 120 μg of nuclear extracts of 3T3-L1 adipocytes 6 days after differentiation treated with or without 25 μmol/L forskolin, which were immunoblotted (IB) for endogenous CREB and phospho-CREB, as indicated. Densitometry was analyzed by Li-Cor Imagining software and quantification represents SEM of three independent experiments. #P < 0.05 (statistically significant) over untreated control extracts. B: SDS-PAGE was used to analyze 120 μg of nuclear extracts from 3T3-L1 adipocytes 6 days after differentiation, treated with or without 25 μmol/L forskolin, which were immunoblotted for endogenous HDAC4 or HDAC5, as indicated. Densitometry was analyzed by infrared spectroscopy and quantification represents mean and SEM of three independent experiments. #P < 0.05 (statistically significant) over untreated control extracts. C: 3T3-L1 adipocytes 6 days after differentiation were treated or untreated with 25 μmol/L forskolin for the indicated time points, and ChIP was performed using the indicated anti-HDAC4, anti-HDAC5, and anti-nonimmune IgG antibodies. Results were analyzed by q-PCR and determined by an expression-to-input ratio and normalized to nonimmune IgG controls. Data are the mean and SEM from three independent experiments analyzed by one-way ANCOVA (P < 0.005 is statistically significant). D: 3T3-L1 adipocytes 6 days after differentiation were treated or untreated with 25 μmol/L forskolin for 3 h and isolated RNA was subjected to q-RT-PCR to determine relative GLUT4 mRNA. #P < 0.003 (statistically significant). E: 3T3-L1 adipocytes 6 days after differentiation were transiently transfected with -895-hGLUT4-luc (for promoter activity) and pRLTK (for transfection efficiency), with and without plasmids encoding GEF and MEF2A (transcription factors). Cells were treated or untreated with 25 μmol/L forskolin throughout the 18-h transfection. Data are the mean and SEM from three independent experiments analyzed by a one-way ANCOVA. #P < 0.05 (statistically significant) compared with reporters alone. *P < 0.05 significant compared with all other conditions.

FIG. 3.

Isoproterenol treatment increased intracellular cAMP levels, induced HDAC4 nuclear localization and down-regulated the GLUT4 promoter. A: A total of 120 μg of nuclear extracts of 3T3-L1 adipocytes 6 days after differentiation treated with or without 25 μmol/L isoproterenol were analyzed by SDS-PAGE and immunoblotted for endogenous CREB and phospho-CREB, as indicated. Densitometry analysis was quantified using Li-Cor Imagining software. Date represent mean and SEM of three independent experiments. #Statistical significance vs. untreated control extracts. B: A total of 120 μg of nuclear extracts of 3T3-L1 adipocytes 6 days after differentiation treated with or without 25 μmol/L isoproterenol were analyzed by SDS-PAGE and immunoblotted for endogenous HDAC4 or HDAC5 as indicated. Densitometry analysis was quantified using Li-Cor Imagining software. Data represent mean and SEM of three independent experiments. #Statistical significance vs. untreated control extracts. C: 3T3-L1 adipocytes 6 days after differentiation were treated or untreated with 25 μmol/L isoproterenol (Iso) for 3 h, and ChIP was performed using the indicated anti-HDAC4, anti-HDAC5, and anti-nonimmune IgG antibodies. Results were analyzed by q-PCR and determined by an expression-to-input ratio and normalized to nonimmune IgG controls. Data represent mean and SEM from three independent experiments analyzed by one-way ANCOVA. #,*P < 0.05 (statistically significant) for controls compared with all other conditions. D: 3T3-L1 adipocytes 6 days after differentiation were treated or untreated with 25 μmol/L isoproterenol for 3 h, and isolated RNA was subjected to q-RT-PCR to determine relative GLUT4 mRNA. #P < 0.003. E: 3T3-L1 adipocytes 6 days after differentiation were transiently transfected with luciferase reporter plasmids -895-hG4-luc (for promoter activity) and pRLTK (for transfection efficiency), with and without plasmids encoding GEF and MEF2A (transcription factors). Cells were treated or untreated with 25 μmol/L isoproterenol, as indicated. Data represent mean and SEM of three independent experiments analyzed by a one-way ANCOVA. #P < 0.05 vs. reporters alone. *P < 0.05 vs. all other conditions.

FIG. 4.

siRNA-mediated knockdown of class II HDACs abolish the isoproterenol effect on GLUT4 inhibition. 3T3-L1 preadipocytes were transiently transfected with scrambled siRNA or HDAC4/5/9-specific siRNA and incubated for 3 days, then transfected a second time with luciferase reporter constructs -895-hGLUT4-luc (for promoter activity) and pRLTK (for transfection efficiency), with and without plasmids encoding GEF and MEF2A, and treated or untreated with forskolin or isoproterenol (+ indicates inclusion in transfection). Data represent mean and SEM from three independent experiments analyzed by a one-way ANCOVA. #Statistical significance (P < 0.05) vs. the respective scrambled siRNA condition. *P > 0.003 vs. other conditions.

FIG. 5.

Fasting mice exhibit elevated cAMP and increased enrichment of HDAC4 on the GLUT4 promoter. A: C57BL/6 mice were allowed free access to food or fasted overnight and white adipose tissue (WAT) was collected. A whole-cell lysate was made from the WAT and analyzed by SDS-PAGE and immunoblotted (IB) for endogenous CREB and p-CREB, as indicated. Densitometry was analyzed by Li-Cor Imagining software, and quantification represents the mean and SEM of three independent experiments. #Statistical significance vs. extracts from fed controls. B: WAT from fed or fasting mice was collected, and q-RT-PCR of isolated RNA was used to determine relative GLUT4 mRNA. #P < 0.003. C: C57BL/6 mice were allowed free access to food or fasted overnight and WAT was collected. ChIP analysis on the WAT was performed using the indicated anti-HDAC4, anti-HDAC5, and anti-nonimmune IgG antibodies. Results were analyzed by qPCR and determined by an expression-to-input ratio and normalized to nonimmune IgG controls. Mean and SEM from three independent experiments is shown and was analyzed by one-way ANCOVA. #P < 0.001 vs. nonimmune IgG control.

FIG. 6.

The LXR agonist TO901317 rescues the inhibitory isoproterenol effect by dislodging HDAC4 from the GLUT4 promoter. A: 3T3-L1 adipocytes 6 days after differentiation were transiently transfected with luciferase reporter plasmids -895-hGLUT4-luc or ΔLXR-hGLUT4-luc and pRLTK (for transfection efficiency), with and without plasmids encoding GEF and MEF2A (transcription factors). Cells were treated or untreated with 25 μmol/L isoproterenol (Iso) and/or 0.1 μmol/L TO901317 (TO), as indicated. Mean and SEM from three independent experiments is shown; data were analyzed by a one-way ANCOVA. #P < 0.05 (statistically significant) vs. with reporters alone. *Significant vs. all other conditions. B: 3T3-L1 adipocytes 6 days after differentiation were treated or untreated with 25 μmol/L isoproterenol and/or 0.1 μmol/L TO901317 for 3 h, and ChIP was performed using the indicated anti-HDAC4, anti-HDAC5, anti-LXR, and anti-nonimmune IgG antibodies. Results were analyzed by q-PCR and determined by an expression-to-input ratio and normalized to nonimmune IgG controls (Con). Mean and SEM from three independent experiments is shown and was analyzed by one-way ANCOVA. #P < 0.001 compared with nonimmune IgG control; *Significant compared with all other conditions. C: Nuclear extracts (60 μg) from 3T3-L1 adipocytes 6 days after differentiation treated without or with 25 μmol/L isoproterenol, 0.1 μmol/L TO901317, or 25 μmol/L isoproterenol and 0.1 μmol/L TO901317, were analyzed by SDS-PAGE and immunoblotted (IB) for endogenous HDAC4 or HDAC5, as indicated. Densitometry was analyzed by infrared spectroscopy and quantification represents mean and SEM of three independent experiments. *P < 0.05 vs. untreated control extracts.

FIG. 7.

GLUT4 LXRE is required for downregulation of adipose GLUT4 mRNA during fasting and diet-induced obesity. C57BL/6 mice (2- to 3-month-old) mice were fed a 10% fat diet (control) or 60% high-fat diet (HFD) for 8 weeks. Fasted animals were animals fed the 10% fat diet that were deprived of food. A: Schematic drawing depicts the constructs used to generate transgenic animals. B: Total mRNA from perigonadal adipose tissue was isolated and used to determine mRNA levels of CAT, GLUT4, and actin. CAT and GLUT4 mRNA levels were normalized to actin, and the normalized mRNA levels were expressed as percent of the control mRNA for each experimental group. Data were analyzed using a two-tailed Student t test. *P < 0.05 for difference between experimental and control groups.