Stimulation of Acetyl-CoA Carboxylase Gene Expression by Glucose Requires Insulin Release and Sterol Regulatory Element Binding Protein 1c in Pancreatic MIN6 β-Cells

Culture media and antibiotics were obtained from Life Technologies (Paisley, U.K.). Other reagents were from Sigma (Poole, Dorset, U.K.) or BDH (Poole).

MIN6 cells (22) were used between passages 20 and 30 and grown in Dulbecco’s modified Eagle’s medium containing 15% (vol/vol) heat-inactivated FCS, 25 mmol/l glucose, 2 mmol/l glutamine, 50 μmol/l β-mercaptoethanol, 100 IU/ml penicillin, and 100 mg/ml streptomycin in a humidified atmosphere at 37°C with 5% CO₂, unless specified otherwise. MIN6 cells were maintained and infected with viruses at 25 mmol/l glucose. Cells were transferred to medium containing 3 mmol/l glucose 16 h before exploring the subsequent effects of glucose.

Adenoviruses expressing constitutively active SREBP1c (amino acids 1–403, wild-type sequence) or dominant-negative (1–403, Y³²⁰A) SREBP1c were constructed and amplified as described (18,23). Virus particles were purified on CsCl gradients before infection (24) at a multiplicity of infection of 30 viral particles per cell. Infection efficiency was judged using enhanced green fluorescent protein, also present in the adenoviral genome; >95% individual MIN6 cells were infected after 24–48 h. In all experiments, adenovirus expressing enhanced green fluorescent protein cDNA only (null) was used as control.

MIN6 cells (1.5 × 10⁵) were perifused (0.5 ml/min) at 37°C in Krebs-Ringer bicarbonate HEPES buffer (1 ml) comprising 125 mmol/l NaCl, 3.5 mmol/l KCl, 1.5 mmol/l CaCl₂, 0.5 mmol/l MgSO₄, 0.5 mmol/l KH₂ PO₄, 2.5 mmol/l NaHCO₃, and 10 mmol/l HEPES-Na⁺, pH 7.4, and containing initially 3 mmol/l glucose and equilibrated with 95:5 O₂:CO₂. Fractions (0.5 ml) were collected and stored at −20°C. Total cellular insulin was extracted in acidified ethanol (25). Radioimmunoassay was performed by competition with ¹²⁵I-labeled rat insulin (Linco Research, St. Charles, MO) (24).

Staining of neutral lipid in MIN6 cells was performed as described (26). Lipid-specific staining was visualized with a Leica TCS NT confocal microscope (excitation at 564 nm).

Total RNA was isolated by cell lysis in TRI Reagent (Sigma). First-strand cDNA synthesis was performed as described (27). RT reactions (1 h) were performed at 42°C using 3.5 μg RNA (determined from A₂₆₀ assuming 0.025 A · μg⁻¹ · ml⁻¹), oligo (15)-dT primers (1.25 μg), and RT (500 units, Moloney-murine leukemia virus; Promega) in a total volume of 50 μl. PCRs were also performed in a total volume of 50 μl, comprising 5 μl cDNA product, 0.2 mmol/l nucleotide triphosphate, 20 pmol of each primer, and 0.875 units Taq polymerase (Expand High Fidelity; Roche Diagnostics, Lewes, U.K.). Endogenous SREBP1c cDNA was amplified using primers corresponding to the murine exon 1c sequence (28): forward, 5′ATCGGCGCGGAAGCT GTCGGGGTAGCGTC; reverse, 5′ ACTGTCTTGGTTGTTGATGAGCTGGAGCAT (predicted product size, 116 bp). For SREBP 1a, the following primers were used: forward, 5′-TAGTCCGAAGCCGGGTGGGCGCCGGCGCCAT; reverse, 5′-GATGTCGTTCAAAACCGCTGTGTGTCCAGTTC (predicted product size, 106 bp). To amplify exogenously expressed SREBP1c, primers corresponding to the rat sequence (20) were used: forward, 5′-GGAGCCATGGATTGCACATT; reverse, 5′-AGGAAGGCTTCCAGAGAGGA, corresponding to amino acids 40–59 and 211–230 of mature SREBP1c. Oligonucleotide primers for β-actin mRNA were used as described (29). Primers for the amplification of ACCI were as follows: forward, 5′-AACTATGAGGTGGATCGGAG; reverse, 5′-TCCATACCGCATTACCATGC, corresponding to nucleotides 4021–4040 and 4481–4500, respectively, of human ACCI cDNA (accession no. U19822) (30). PCR products were separated on 2% agarose gels containing ethidium bromide (0.5 μg/ml) and quantitated by digital imaging (ImageQuant; Molecular Dynamics).

MIN6 cells (1.2 × 10⁶) were extracted into radio-immunoprecipitation assay buffer comprising PBS supplemented with 1.0% (vol/vol) Nonidet P40, 0.5% Na⁺ deoxycholate, and 0.1% (wt/vol) sodium dodecylsulfate. Protein concentration was determined with a commercially available kit using BSA as standard (BCA, category no. 23225; Pierce, Rockford, IL). To obtain nuclear extracts, cells were lysed by swelling in hypotonic buffer (10 nmol/l HEPES, pH 7.9, 10 mmol/l KCl, and 0.1 mmol/l Na⁺ EDTA) as described (31). Proteins were separated by SDS-PAGE (7.5% wt/vol) and transferred to immobilon membranes using a discontinuous buffer system as described (32). Blots were probed with monoclonal antibody to SREBP1 (1:1,000 dilution) and revealed with horseradish peroxidase-conjugated anti-mouse IgG (1:10,000) using an enhanced chemiluminescence detection system (Roche Diagnostics).

To generate the SREBP1c promoter.luciferase vector (pSREBP.Luc_FF), genomic DNA was isolated from MIN6 cells by lysis in TRI reagent. The immediate 5′ flanking region of the SREBP1c gene, lying between exons 1 and 2 (33), was amplified by PCR using the primer pair: forward, 5′-TTTTAGATCT⁻⁵⁷⁴GGATCCAGAACTGGATCATCAG; reverse, 5′-TTTTAAGCTT⁺²⁵CTAGGGCGTGCAGACGCTAC (BglII and HindIII sites are underlined; superscript indicates nucleotide positions with respect to transcriptional start site). PCR was performed in a total volume of 50 μl, comprising 0.3 μg DNA, 0.2 mmol/l nucleotide-triphosphate, 20 pmol of each primer, and 0.875 units Taq polymerase (Expand High Fidelity). After purification on a 2% (wt/vol) agarose gel, the 636-bp product was subcloned upstream of humanized firefly luciferase cDNA in plasmid pGL3 basic (Promega). Correct amplification and fragment insertion were confirmed by automated sequencing. Plasmid pACCIpII-1082.Luc_FF contained the immediate flanking region (−1082 to 62) of the ACCI(pII) promoter in plasmid pGL3, as reported by Luo and Kim (34). Plasmid pACCIpII-285.Luc_FF contained the fragment −285 to 62 upstream of firefly luciferase and corresponded to deletion mutant pPII-CAT4 (35). pACCIpII-217.Luc_FF was generated by subcloning the 291 bp AvaI-SalI fragment from pII-1082-Luc_FF into the SmaI site of pGL3-basic. pCMV.Luc_FF was as described earlier (36). The expression plasmid for Renilla luciferase (pRL.CMV) was purchased from Promega. cDNAs encoding the NH₂-terminal fragment of SREBP1c (1–403), or the mutated (Y³²⁰A) form, in plasmid pcDNA3 (Invitrogen) were obtained from Dr. B. Spiegelman (Harvard Medical School, Boston, MA).

PCR mutagenesis of plasmid pACCIpII-285.Luc_FF was carried out using the QuikChange Site-Directed Mutagenesis Kit (Stratagene), according to the manufacturer’s protocol. The following oligonucleotide primers (with mutated nucleotides in bold) were used: distal sterol response element (SRE) mutant: forward, 5′-GGTACCGAGCTCGGAAACTCGGCATCGCCTCACG-3′; reverse, 5′-CGTGAGGCGATGCCGAGTTTCCGAGCTCGGTACC-3′; proximal SRE mutant: forward, 5′-CGGCATCGCCGACCGTCGCCCCGCCCCG-3′; reverse, 5′-CGGGGCGGGGCGACGGTCGGCGATGCCG-3′.

Intranuclear micro-injection of plasmids was performed using an Eppendorf 5121/5246 micromanipulator/microinjector (7,37) at plasmid concentrations of 0.1 mg/ml (Luc_FF-based vectors) or 0.05 mg/ml (pCMV.RL, pcDNA3.SREBP1c [1–403]/Y³²⁰A) in injection buffer (2 mmol/l Tris-HCl, 0.2 mmol/l Na⁺ EDTA, pH 8.0) (38). SREBP1 antibody was dialyzed into injection buffer and co-injected into the nuclear compartment at 0.1 mg/ml as described (39). Individual experiments involved the injection of 100–200 separate cells per condition, with an efficiency of 5–20% productive injection. Cells were imaged 6 h after micro-injection and cultured under the conditions given in the figure legends. Photon-counting imaging was performed in single living cells using an Olympus IX-70 inverted microscope (×10 air objective, 0.4 numerical aperture) and an intensified charge-coupled device camera (Photek; Lewes, East Sussex, U.K.) (7,40).

Assays were performed on nuclear extracts (see above) according to the study by Daniel and Kim (35).

Data are expressed as means ± SE. Comparison of groups was performed with a one-tailed Student’s t test.

ACCI mRNA was strongly (approximately fourfold) induced by incubation of MIN6 cells for 6 h at 30 vs. 3 mmol/l glucose (Fig. 1A) or at 3 mmol/l glucose in the presence of insulin (>1 nmol/l). Glucose- and insulin-induced changes in ACCI mRNA level largely reflected a stimulation of mRNA synthesis, and 30 mmol/l glucose caused a >30-fold activation of an ∼1-kb fragment of the ACCI (pII) promoter assayed in single cells after microinjection and imaging of a firefly luciferase reporter construct (plasmid ACCI pII-1082.Luc_FF; Fig. 1B).

We previously reported that incubation of MIN6 cells under the conditions used here increases insulin levels, measured in the medium at the end of the 6-h incubation, from ∼7 to 14 nmol/l (12). To determine whether the effects of high [glucose] on ACCI mRNA levels and promoter activity could largely be explained by activated insulin release, we suppressed the latter with the l-type Ca²⁺ channel blocker, verapamil (41), or with an opener of ATP-sensitive K⁺ channels, diazoxide. Either agent completely inhibited ACCI (pII) promoter activation by 30 mmol/l glucose (Fig. 1B). To exclude the possibility that these drugs acted via a decrease in intracellular free [Ca²⁺] rather than a blockade of insulin release per se, we determined the effect of the α₂-adrenoreceptor agonist, clonidine, which inhibits glucose-induced insulin secretion but not increases in [Ca²⁺] (12,42). Clonidine completely reversed the stimulation of ACCI promoter activity by 30 mmol/l glucose, and this effect was rescued by 20 nmol/l insulin (Fig. 1C).

We next explored the possibility that a factor other than insulin, present within secretory vesicles and cosecreted with the hormone (e.g., ATP), may be responsible for triggering ACCI promoter activation. Arguing against this possibility, the effects of 30 mmol/l glucose were completely suppressed by blockade of signaling by insulin, but not ATP receptors, with the phosphatidylinositol 3′ kinase inhibitor, LY294002 (not shown), or through coexpression of the phosphoinositide 3′ phosphatase phosphatase and tensin homologue (43) (Fig. 1D).

To determine which cis-acting element(s) was involved in its regulation, we next performed 5′ deletion analysis of the ACCI (pII) promoter. Truncation of the promoter as far as nucleotide −285 and consequent removal of proximal SREs (an E-box and an Sp1 binding site) had no effect on ACCI promoter activity at 30 mmol/l glucose but led to a small increase in basal promoter activity (Fig. 2). By contrast, further removal of nucleotides −285 to −217 and thus two SRE half-sites (44) led to a marked (>85%) reduction in glucose-stimulated promoter activity (Fig. 2).

To explore the potential role of SREBP1, acting at the SRE half-sites, we next co-microinjected a monoclonal antibody specific for SREBP1 (Fig. 2, right panel). This led to a complete inhibition of the stimulation by 30 mmol/l glucose of the full-length (−1,082 kb) and −285-tuncated ACCI promoters (Fig. 2, top). By contrast, anti-SREBP1 antibodies had no further inhibitory effect on the construct in which the two SRE half-sites had been removed (truncation at −217 bp; Fig. 2).

To further dissect the role of the two SRE half-sites, we mutated each of these individually (44). Mutation of either site led to a strong glucose-independent induction of promoter activity (Fig. 2). Suggesting that SREBP1c may bind to either half-site as a monomer (44,45), anti-SREBP1 antibodies caused substantial inhibition of the promoter mutated at either SRE half-site (Fig. 2).

Our next experiments explored the impact of SREBP1 inhibition on the activity of the endogenous ACCI gene. Suggesting that SREBP1c is probably the predominant SREBP1 isoform in MIN6 β-cells, SREBP1c mRNA was clearly quantifiable after 20 PCR cycles, whereas SREBP1a mRNA was undetectable after 20 PCR cycles and only weakly detectable after 25 cycles (data not shown). Correspondingly, infection with an adenovirus expressing a dominant-negative-acting form of SREBP1c (SREBP1c DN) (residues 1–403, Y³²⁰A), which is unable to bind DNA (46), completely suppressed the glucose-induced increase in ACCI mRNA (Fig. 3A). This inhibition could be ascribed to the blockade of ACCI transcription because microinjection of a plasmid encoding SREBP1c DN totally abrogated the glucose-induced increase in ACCI (pII) promoter activity (Fig. 3B).

To test whether an increase in SREBP1c amount or activity in response to glucose may be responsible for the stimulation of the ACCI gene, MIN6 cells were infected with an adenovirus bearing the constitutively active NH₂-terminal fragment of SREBP1c (SREBP1c CA) (SREBP1c 1–403) (19). Infection with this virus caused a marked increase in ACCI mRNA levels in cells incubated for 6 h at either 3 or 30 mmol/l glucose (Fig. 4A). Similarly, microinjection of single cells with a plasmid construct bearing SREBP1c CA caused a marked increase in ACCI (pII) promoter activity (Fig. 4B) at either glucose concentration.

We next determined whether the effects of SREBP1c CA on ACCI gene expression might be due to enhanced insulin secretion. As evidence against this possibility, SREBP1 CA expression caused a near-complete abolition of glucose-stimulated insulin release (Fig. 4C, right panel) and was associated with a marked accumulation of intracellular lipid (Fig. 4C, left panel).

Elevated glucose concentrations (30 vs. 3 mmol/l) increased the levels of SREBP1c mRNA (Fig. 5A), total and mature SREBP1 protein (Fig. 5B), and SREBP1c transcription (Fig. 5C). However, in contrast to the regulation of the ACCI gene (Fig. 1), exogenously added insulin had no effect on any of the above parameters (Fig. 5). Moreover, neither high [glucose] nor insulin (20 nmol/l) enhanced SREBP1c processing (Fig. 5B). Finally, the effects of glucose on the SREBP1c promoter were unaltered by blockade of insulin secretion with verapamil or diazoxide (Fig. 5C), and neither exogenously added insulin nor the insulin secretagogue KCl affected SREBP1c promoter activity (Fig. 5C).

To further explore the role of the induction of SREBP1c or other transcription factors in the stimulation of the ACCI gene by glucose, we tested the impact of inhibiting protein synthesis globally. Although incubation of cells for 6 h with cycloheximide led to a stimulation of ACCI mRNA levels at 3 mmol/l glucose, a significant further increase in ACCI [mRNA] in response to 30 mmol/l glucose was still observed (ratios of ACCI: β-actin mRNA in the absence of cycloheximide 0.59 ± 0.10 and 0.96 ± 0.14 at 3 and 30 mmol/l glucose, respectively; corresponding values with 50 μg/ml cycloheximide were 1.08 ± 0.02 and 1.68 ± 0.19, n = 3 experiments in each case, P < 0.05 for the effect of 30 mmol/l glucose).

The above data suggest that glucose acts to stimulate ACCI gene expression through a mechanism that requires preexisting SREBP1c protein but not the activation of SREBP1 gene expression or SREBP1c processing. Therefore, to determine whether glucose or insulin enhances the DNA binding activity of preexisting SREBP1c, we used electrophoretic mobility shift assay. However, binding to an oligonucleotide corresponding to the two SRE half-sites delineated as the likely site of SREBP1c binding by promoter deletion analysis (Fig. 2) was too weak to be detected in extracts from cells incubated at either 3 or 30 mmol/l glucose (not shown).

Similar to preproinsulin (13,38, 47) and l-PK (12) gene expression, we show here that, at relatively early time points (≤6 h), ACCI gene expression in β-cells is probably stimulated by high glucose concentrations through a mechanism principally involving stimulated insulin release (see Fig. 1 and results). Thus, addition of the hormone was as efficient as high [glucose] in increasing ACCI mRNA levels (Fig. 1A). However, suggesting that intracellular metabolites of glucose such as glucose-6-phosphate (3) may also play some role in transcriptional stimulation, added insulin (20 nmol/l) was less effective than high glucose in stimulating ACCI promoter activity (Fig. 1B). Furthermore, 30 mmol/l glucose still activated the ACCI promoter after expression of active SREBP1c (Fig. 4A and B)—conditions under which glucose-stimulated insulin release was substantially, although not completely, inhibited (Fig. 4C). However, under these conditions, where the ACCI promoter was “ super-activated” by active SREBP1c, it is conceivable that even the modest stimulation of insulin release by glucose (Fig. 4C) may contribute to further transcriptional activation of the ACCI gene. Moreover, the failure of elevated glucose concentrations to affect ACCI transcription when insulin release was completely inhibited with verapamil or diazoxide (Fig. 1B) suggests that the impact of intracellular metabolites of glucose may be limited to a potentiation of the effects of released insulin. These findings thus contrast results previously reported in INS1 cells (3), where non-glucose nutrients and other secretagogues (KCl and forskolin) had no apparent effect on ACCI mRNA levels. One possible explanation of this apparent discrepancy is that glucose regulates ACCI gene expression by distinct mechanisms in INS1 and MIN6 cells. However, we have noted that incubation with KCl has deleterious effects on MIN6 cell morphology and does not provoke the accumulation of ACCI mRNA at the 24-h time point used by Brun et al. (3). Moreover, the effects of nutrient secretagogues on the release of insulin were not examined by these authors. Thus, it would seem difficult to exclude a role for secreted insulin in these earlier studies, especially because the effects of blocking insulin release (while preserving glucose metabolism) were not explored.

Consistent with a mechanism involving the action of secreted insulin, the effects of glucose on ACCI gene expression reported here required the presence of nuclear SREBP1c (Figs. 2 and 3) (18). However, induction of SREBP1c gene expression by glucose should not be considered responsible for the effects of glucose on the ACCI gene at early time points because ACCI mRNA was still induced by glucose when the synthesis of new proteins was suppressed (see results). Moreover, high glucose concentrations stimulated SREBP1c promoter activity independently of insulin release (Fig. 5C), whereas the latter apparently activated ACCI transcription (Fig. 1). In this respect, the regulation of the SREBP1c gene in β-cells is different from that in the liver, where insulin alone, but not glucose, induces SREBP1c expression (19).

Our data suggest instead that an insulin-stimulated modification of the NH₂-terminal fragment of SREBP1c, and thus the transactivation or DNA binding capacity of this factor, may be responsible for the activation of the ACCI gene in response to high glucose. Although the nature of this modification is unknown, it is of note that SREBP1c can be phosphorylated in vitro by mitogen-activated protein kinases (48,49), a family of enzymes strongly activated by glucose in β-cell lines (50,51). Further studies of the phosphorylation state of SREBP1c in intact cells will be required to resolve these issues. The mechanisms whereby glucose may regulate the ACCI and other genes in the β-cell are summarized in Fig. 6.

Lopez et al. (52) have previously shown that purified recombinant SREBP1c is capable of binding oligonucleotides corresponding to the two half-site SREs in the region of −285 to −247 with respect to the ACCI transcriptional start site. We show here that this region is required for the effect of glucose on the ACCI promoter in MIN6 cells (Fig. 3) and that mutation of either half-site leads to an increase in ACCI promoter activity. However, we were unable to detect the presence of any activity capable of binding these oligonucleotides in nuclear extracts from MIN6 cells by electrophoretic mobility shift assay (see results). This result was unsurprising because DNA binding activity of SREBP1c is difficult to demonstrate, even in nuclear extracts where SREBP1c is abundant, such as liver (F.F., P.F., unpublished observations). Nevertheless, our preliminary data suggest that the binding activity of adenovirally overexpressed SREBP1c to a consensus oligonucleotide derived from the fatty acid synthase (FAS) promoter (53) may be stimulated by elevated glucose concentrations (F.D., G.A.R., unpublished observations; data not shown).

An important finding made during these studies is that overexpression of SREBP1c leads to a dramatic deposition of intracellular lipid and suppression of insulin secretion (Fig. 4C). Thus, by regulating ACC and FAS gene expression, glucose-induced increases in SREBP1c may play an important role in β -cell glucolipotoxicity (54) in some forms of type 2 diabetes.

This work was supported by project grants from the Wellcome Trust, Diabetes United Kingdom (formerly the British Diabetes Association), the Medical Research Council (U.K.), the Biotechnology and Biological Sciences Research Council, and the European Union (CT 97.3011 to P.F., F.F., and G.A.R.). F.D. was supported by fellowships from Association de Langue Française pour l’Étude du Diabète et des Maladies Métaboliques, Institut Français de la Nutrition (France), and the European Union (Marie Curie).

We thank Dr. Isabelle Leclerc for helpful discussion.