Induction of the ChREBPβ Isoform Is Essential for Glucose-Stimulated β-Cell Proliferation

Abstract

Carbohydrate-responsive element-binding protein (ChREBP) is a glucose-sensing transcription factor required for glucose-stimulated proliferation of pancreatic β-cells in rodents and humans. The full-length isoform (ChREBPα) has a low glucose inhibitory domain (LID) that restrains the transactivation domain when glucose catabolism is minimal. A novel isoform of ChREBP (ChREBPβ) was recently described that lacks the LID domain and is therefore constitutively and more potently active. ChREBPβ has not been described in β-cells nor has its role in glucose-stimulated proliferation been determined. We found that ChREBPβ is highly expressed in response to glucose, particularly with prolonged culture in hyperglycemic conditions. In addition, small interfering RNAs that knocked down ChREBPβ transcripts without affecting ChREBPα expression or activity decreased glucose-stimulated expression of carbohydrate response element-containing genes and glucose-stimulated proliferation in INS-1 cells and in isolated rat islets. Quantitative chromatin immunoprecipitation, electrophoretic mobility shift assays, and luciferase reporter assays were used to demonstrate that ChREBP binds to a newly identified powerful carbohydrate response element in β-cells and hepatocytes, distinct from that in differentiated 3T3-L1 adipocytes. We conclude that ChREBPβ contributes to glucose-stimulated gene expression and proliferation in β-cells, with recruitment of ChREBPα to tissue-specific elements of the ChREBPβ isoform promoter.

Figure 1

ChREBPβ is expressed in β-cells and is responsive to glucose. A: Diagram of the primers used in RT-PCR assays. Primers were designed to be specific for exon 1b, exon 1a, or for a region that is common to both isoforms. B: Time course of ChREBPβ expression in response to 20 mmol/L glucose. 832/13 INS-1–derived rat insulinoma cells were cultured in media containing 2 or 20 mmol/L glucose for the indicated times, and RT-PCR was performed using primers specific for the three forms of ChREBP or for β-actin as a control. C: Dose response of ChREBPβ expression. 832/13 cells were cultured in the indicated concentrations of glucose for 18 h, and RT-PCR was performed as in B. D and E: 832/13 cells were cultured in the indicated concentrations of glucose for 18 h and treated with a scrambled control siRNA (siCon) or with siRNAs directed against ChREBPβ (siChREBP01 and -02). RT-PCRs were performed using primers for the indicated genes. Data are presented as the fold change relative to starting time point (B), or the lowest concentration (C), or to the low glucose scramble control (D and E) for each primer pair after normalizing to β-actin using the ΔΔCT method. All experiments were performed at least three times. Error bars represent the SEM. *P < 0.05; ns, not significant.

Figure 2

Glucose induces ChREBPβ in primary rat and human islet cells. A–G: Dispersed rat islet cells were incubated in media containing 5.5 or 15 mmol/L glucose for 1 to 4 days, and the expression of the indicated genes were determined and shown as fold change relative to each day after normalizing to β-actin using the ΔΔCT method. H: Isolated human islet cells were cultured in media containing 5.5 or 15 mmol/L glucose for 1, 2, or 4 days. Total RNA was isolated and subjected to RT-PCR using primers specific for the indicated genes. The data are expressed as a fold change from 5.5 mmol/L glucose. I: Rat islets were isolated, dispersed, and incubated with lipid-conjugated Accell siRNA for 4 days. Total RNA was isolated and subjected to RT-PCR. Data are presented as relative to the scramble control (SiCon) 15 mmol/L treatment, after normalization to β-actin using the ΔΔCT method. Error bars are the SEM (n = 3–4 for rat islets, n = 5–9 for human islets). ChREBPComm, ChREBP-common; siChREBPβ01 and -02, siRNAs directed against ChREBPβ. *P < 0.05.

Figure 3

ChREBP binds to tissue-specific ChoREs of exon 1b. A: Luciferase assays identify the exon 1b E-box as a functional ChoRE. A total of 832 cells were transfected with luciferase reporter plasmids driven by exon 1b and upstream sequences containing wild-type (WT) or mutant versions of an upstream E-box element and a previously defined downstream ChoRE, as indicated, or an empty vector control (PGL3). Cells were treated for 18 h with 2 or 20 mmol/L glucose, and luciferase activity was measured from cell lysates. Results shown are relative fold luciferase activity, normalized to protein (n = 3). Error bars are SEM. *P < 0.05 when comparing 20 vs. 2 mmol/L glucose; #P < 0.05 when compared with the WT ChREBPβ promoter construct. B: Conservation of flanking sequence of the upstream E-box of ChREBP exon 1b were aligned from data obtained from the University of California, Santa Cruz genome browser (http://genome.ucsc.edu/). Sequence deviation from human is denoted with bolding and underlining. Together, the conserved sequences are consistent with a consensus ChoRE, with two E-box–like elements separated by 5 bp. C: EMSA demonstrates ChREBP binding to the newly identified upstream ChoRE. HeLa cells were transfected with plasmids expressing Flag-tagged ChREBP and HA-tagged Mlx (the obligate heterodimer binding partner of ChREBP), and cell lysates were used for gel shift assays using the upstream ChoRE and flanking sequences as fluorescently labeled double-stranded probes with the indicated incubation parameters. The gel was visualized with a LI-COR Odyssey System and is representative of three independent experiments with essentially identical results. Abs, antibodies; ACC, Acaca ChoRE.

Figure 4

ChIP assays reveal tissue-specific recruitment of ChREBP to exon 1b ChoREs. A: INS-1–derived 832/13 cells were treated with 2 or 20 mmol/L glucose for 18 h, and cells were fixed with formaldehyde and subjected to a ChIP assay using antibodies directed against ChREBP or an IgG control. The results are presented as fold enrichment over the control IgG signal. Numerous primer pairs were chosen to scan regions of the ChREBP gene upstream of exons 1a and 1b as indicated. The results are from four to seven independent experiments. Chromatin isolated from rat islets cultured for 4 days in 15 mmol/L glucose (B) or from frozen mouse liver from ad libitum–fed mice (C) was sheared and subjected to a ChIP assay using antibodies against ChREBP or IgG as the control. Results relative to the IgG control and are from three mice and four rats. D: NIH 3T3-L1 cells were differentiated into mature adipocytes, cultured for 18 h in 20 mmol/L glucose, and processed for a quantitative ChIP assay. Results of three to four independent experiments are expressed as the signal from an antibody directed against ChREBP relative to the IgG control. Error bars are the SEM. *P < 0.05; ns, not significant.

Figure 5

Correlation of proliferation and ChREBPβ expression of pancreatic β-cells. A: The ratios of ChREBPβ to ChREBPα are displayed in relation to percent BrdU incorporation for each of the indicated model systems after expression for the indicated times in 15 or 20 mmol/L glucose was measured in absolute terms, using the ΔΔCT method relative to β-actin and adjusting for primer efficiency. Data are listed in the order of highest to lowest BrdU incorporation, corresponding to highest to lowest ChREBPβ-to-ChREBPα ratio. nd, not determined. B: Data from primary rodent and human cells from A were plotted on a double log scale and fitted with a power least squares fit and R² value, calculated in Excel. We note that inclusion of data from INS-1–derived insulinoma cells decreased the R² value to 0.84. Data from ChREBPα (C) and ChREBPβ (D) and their relation to proliferation are presented separately (n = 3–5). The error bars represent the SEM.

Figure 6

ChREBPβ depletion attenuates glucose-stimulated β-cell proliferation in INS-1–derived 832/13 cells. A: INS-1–derived 832/13 cells were treated with control siRNAs (siCon) or siRNAs targeted against ChREBPβ mRNA (Siβ-01 and -02) and 24 h later cultured for 16 h in 2 or 20 mmol/L glucose. Immunoblotting was performed using antibodies against ChREBP (ChREBPα is visualized) and β-actin (note that these siRNAs decreased ChREBPβ mRNA but had no effect on ChREBPα, Fig. 1). pos, positive control. B: After 48 h of the siRNA treatment, cells were treated for 2 h with 2 or 20 mmol/L glucose and fixed and stained with an antibody recognizing ChREBPα. C: After 48 h, cells were cultured in 2 or 20 mmol/L glucose for 16 h, and BrdU was added 30 min before fixation and staining for BrdU (red) and DAPI (blue). D: Quantification of the results in C, wherein at least 1,000 cells were counted. Results are from four independent experiments. Error bars are the SEM. siChREBPβ, siRNAs directed against ChREBPβ. *P < 0.05.

Figure 7

Depletion of ChREBPβ attenuates glucose-stimulated β-cell proliferation in isolated rat islet cells. A and B: Isolated rat islet cells were treated with lipid-conjugated (Accell) siRNAs and cultured in 5.5 or 15 mmol/L glucose for 96 h. Total RNA was collected and subjected to RT-PCR using primers specific for ChREBPβ or ChREBPα mRNAs. C: Protein extracts were subjected to immunoblotting using an antibody against ChREBPα and β-actin. D: To determine the effects of the siRNAs on ChREBP translocation, cells were treated with control or Accell siRNA against ChREBPβ cultured in 5.5 or 15 mmol/L glucose for 32 h and were fixed and stained for insulin, ChREBP, and DNA (DAPI). E: Isolated rat islet cells were treated with Accell siRNAs and cultured in 5.5 or 15 mmol/L glucose for 96 h and fixed and stained with Ki67 and insulin. F: Quantification of the results in E, wherein at least 1,000 insulin-positive cells were counted. Results are from four different rat islet isolations. siChREBPβ-01 and -02, siRNAs directed against ChREBPβ; siCon, scramble control siRNA. Error bars are the SEM. *P < 0.05.

Figure 8

A model of ChREBPβ-mediated glucose-stimulated β-cell proliferation. ChREBPα is mostly cytoplasmic in β-cells, and only a small percentage actually enters the nucleus in response to increased glucose metabolism (37,39). ChREBPβ is a target gene of ChREBPα and is constitutively nuclear and more transcriptionally potent than ChREBPα (2). Thus, glucose drives a feed-forward amplification signal that continues as long as glucose metabolism remains elevated. We propose that elevation of glucose metabolism for a period of time is required for glucose-stimulated proliferation but that hyperglycemia for too long results in ChREBPβ-contributed glucose toxicity.