Palmitate impairs and eicosapentaenoate restores insulin secretion through regulation of SREBP-1c in pancreatic islets

Abstract

Objective: Chronic exposure to fatty acids causes beta-cell failure, often referred to as lipotoxicity. We investigated its mechanisms, focusing on contribution of SREBP-1c, a key transcription factor for lipogenesis.

Research design and methods: We studied in vitro and in vivo effects of saturated and polyunsaturated acids on insulin secretion, insulin signaling, and expression of genes involved in beta-cell functions. Pancreatic islets isolated from C57BL/6 control and SREBP-1-null mice and adenoviral gene delivery or knockdown systems of related genes were used.

Results: Incubation of C57BL/6 islets with palmitate caused inhibition of both glucose- and potassium-stimulated insulin secretion, but addition of eicosapentaenoate (EPA) restored both inhibitions. Concomitantly, palmitate activated and EPA abolished both mRNA and nuclear protein of SREBP-1c, accompanied by reciprocal changes of SREBP-1c target genes such as insulin receptor substrate-2 (IRS-2) and granuphilin. These palmitate-EPA effects on insulin secretion were abolished in SREBP-1-null islets. Suppression of IRS-2/Akt pathway could be a part of the downstream mechanism for the SREBP-1c-mediated insulin secretion defect because adenoviral constitutively active Akt compensated it. Uncoupling protein-2 (UCP-2) also plays a crucial role in the palmitate inhibition of insulin secretion, as confirmed by knockdown experiments, but SREBP-1c contribution to UCP-2 regulation was partial. The palmitate-EPA regulation of insulin secretion was similarly observed in islets from C57BL/6 mice pretreated with dietary manipulations. Furthermore, administration of EPA to diabetic KK-Ay mice ameliorated impairment of insulin secretion in their islets.

Conclusions: SREBP-1c plays a dominant role in palmitate-mediated insulin secretion defect, and EPA prevents it through SREBP-1c inhibition, implicating a therapeutic potential for treating diabetes related to lipotoxicity.

FIG. 1.

Lipotoxic effects of palmitate and protective effects of EPA on insulin secretion in murine-isolated islets. A: Low GSIS (2.8 mmol/l), high GSIS (20 mmol/l), and KSIS from murine-isolated islets incubated without (control, white bars) or with palmitate (black bars), palmitate-EPA (bold hatched bars), or EPA (regular hatched bars). B: Insulin content of islets incubated without (control) or with palmitate, palmitate-EPA, or EPA. C: Palmitate uptake in islets isolated from C57BL/6 mice. Three independent experiments were performed using four sets of islets for each repetition, and results are expressed as means ± SE. Statistical analysis was performed using one-way ANOVA followed by Dunnett's procedure. **P < 0.01 and *P < 0.05 vs. palmitate group, respectively.

FIG. 2.

Gene expression and protein profiles in murine-isolated islets treated with palmitate or palmitate-EPA. A and B: Levels of mRNA of various genes in pancreatic islets isolated from C57BL/6 mice without (control, white bars) or with palmitate (black bars), palmitate-EPA (hatched bars), or EPA (regular hatched bars) as determined by real-time PCR. mRNA quantities were calculated as a ratio to the cyclophilin level in the each cDNA sample. Data are shown as the relative expression ratio to control samples. C: Cellular TG levels of islets incubated with palmitate, palmitate-EPA, or EPA. D: Immunoblot analysis of indicated proteins in the islets. Cont., control; mSREBP-1, membrane form of SREBP-1; nSREBP-1, nuclear form of SREBP-1. α-Tubulin protein was used as a loading control. Three independent experiments were performed using four sets of islets, and results are expressed as means ± SE. Statistical analyses were performed using one-way ANOVA followed by Dunnett's procedure. **P < 0.01 vs. palmitate group.

FIG. 3.

Protection from palmitate-induced lipotoxicity in islets isolated from SREBP-1–null mice. Islets were isolated from wild-type littermates (white bars) and SREBP-1–null mice (black bars) and incubated without (control) or with palmitate, palmitate-EPA, or EPA for 48 h. GSIS and KSIS (A) and cellular TG contents (B) were measured. Three independent experiments were performed using four sets of islets, and results are expressed as means ± SE. Statistical analyses were performed using two-way ANOVA followed by Tukey's procedure. **P < 0.01 and *P < 0.05, respectively.

FIG. 4.

Gene expression and protein profiles in islets isolated from SREBP-1–null mice treated with palmitate or palmitate-EPA. Islets were isolated from SREBP-1–null mice and wild-type littermates and incubated without (control) or with palmitate, palmitate-EPA (PE), or EPA for 48 h. A: mRNA levels of the indicated genes were measured. mRNA levels were determined by real-time PCR, calculated as ratio to cyclophilin expression levels. B: Immunoblot analysis of SREBP-1 and insulin-signaling proteins. Relative expression ratios to control samples are shown. Three independent experiments were performed using four sets of islets, and results are expressed as means ± SE. Statistical analyses were performed using two-way ANOVA followed by Tukey's procedure. **P < 0.01 and *P < 0.05, respectively.

FIG. 5.

Effects of overexpression of constitutively active Akt in islets treated with palmitate or palmitate-EPA. Islets were isolated from C57BL/6 mice and incubated without (control) or with palmitate or palmitate-EPA for 48 h. Islets were infected (100 multiplicity of infection, respectively) with adenoviral-GFP (Ad-GFP) or adenoviral–constitutively active Akt (Ad-Akt-CA) for 48 h before incubation with palmitate or palmitate-EPA. GSIS and KSIS (A) and protein levels of indicated insulin-signaling molecules (B) were measured. Amounts of insulin-signaling proteins were estimated by immunoblot analysis using indicated antibodies, and α-tubulin protein was used as a loading control. Levels of mRNA of SREBP-1c and UCP-2 were determined by real-time PCR (C), calculated as ratio to cyclophilin expression levels. Relative expression ratios to control samples are shown. Three independent experiments were performed using four sets of islets, and results are expressed as means ± SE. Statistical analyses were performed using two-way ANOVA followed by Tukey's procedure. **P < 0.01 and *P < 0.05, respectively.

FIG. 6.

Effects of UCP-2 gene silencing on murine-isolated islets treated with palmitate or palmitate-EPA. Islets were isolated from C57BL/6 mice and infected (500 multiplicity of infection, respectively) with adenoviral siRNA for LacZ (Ad-LacZ RNAi) or UCP-2 (Ad-UCP-2 RNAi) and cultured without (control) or with palmitate, palmitate-EPA, or EPA for 48 h. A: The effects of UCP-2 siRNA on mRNA levels of UCP-2 in pancreatic islets isolated from C57BL/6 mice incubated with palmitate, palmitate-EPA, or EPA were determined by real-time PCR. GSIS (B) and ATP-to-ADP ratio (C) from the islets after UCP-2 gene silencing and incubation with palmitate, palmitate-EPA, or EPA were measured. Level of mRNA of SREBP-1c was determined by real-time PCR (D), calculated as ratio to cyclophilin expression levels. Relative expression ratios to control samples are shown. Three independent experiments were performed using four sets of islets, and results are expressed as means ± SE. Statistical analyses were performed using two-way ANOVA followed by Tukey's procedure. **P < 0.01.

FIG. 7.

Effect of EPA on insulin secretion in vivo. C57BL/6 mice were fed control diet (white bars), 20% Tripalmitin diet (black bars), and Tripalmitin + 5% EPA-E diet (hatched bars) for 28 days. Islets were isolated from individual animals. GSIS and KSIS (A) and mRNA levels of SREBP-1c (B) were measured. KK-Ay mice were administered vehicle (black bars) or EPA at a dose of 1 g · kg⁻¹ · day⁻¹ (hatched bars) for 28 days. Islets were isolated from pool pancreas (three to four animals). GSIS (C) and mRNA levels of SREBP-1c (D) were measured. Three independent experiments were performed using four sets of islets, and results are expressed as means ± SE. Statistical analyses between indicated groups were performed using one-way ANOVA followed by Dunnett's procedure. **P < 0.01 and *P < 0.05, respectively.

FIG. 8.

Mechanism by which palmitate induces and EPA protects impairment of insulin secretion in pancreatic islets.