The role of metformin and resveratrol in the prevention of hypoxia-inducible factor 1α accumulation and fibrosis in hypoxic adipose tissue

Abstract

Background and purpose: Hypoxic activation of hypoxia-inducible factor 1α (HIF-1α) and fibrosis in adipose tissue contribute to adipose dysfunction. This study was designed to investigate the effects of metformin and resveratrol on the regulation of HIF-1α and fibrosis in hypoxic adipose tissue.

Experimental approach: Mice were fed a high-fat diet to induce hypoxia and fibrosis in adipose tissue; adipose tissue incubated in vitro in 1% O2 showed a similar change. The effects of metformin and resveratrol on hypoxia, HIF-1α accumulation, endoplasmic reticulum stress and gene expressions of extracellular matrix components and pro-inflammatory cytokines were examined.

Key results: Oral administration of metformin or resveratrol prevented hypoxia and reduced HIF-1α accumulation with dephosphorylation of inositol-requiring enzyme 1α and eukaryotic initiation factor 2α, indicative of suppression of hypoxic HIF-1α activation and endoplasmic reticulum stress. Metformin and resveratrol down-regulated gene expressions of Col3α, Col6α, elastin and lysyl oxidase and thereby reduced collagen deposition in adipose tissue. The increased gene expressions of TNF-α, IL-6, monocyte chemoattractant protein 1 and F4/80 were also down-regulated by metformin and resveratrol. Metformin and resveratrol had similar effects in adipose tissue exposed to 1% O2 . Metformin reduced ATP production and prevented the reduction in oxygen tension in 3T3-L1 cells, suggesting that it prevented hypoxia by limiting oxygen consumption, whereas resveratrol reduced HIF-1α accumulation by promoting its proteasomal degradation via the regulation of AMPK/SIRT1.

Conclusion and implications: Hypoxia and fibrosis are early causes of adipose dysfunction in obesity. Both metformin and resveratrol effectively inhibited HIF-1α activation-induced fibrosis and inflammation in adipose tissue, although by different mechanisms.

Figure 1

Oral administration of metformin (Met) and resveratrol (Res) prevented hypoxic HIF‐1α induction in adipose tissue. Mice were fed a HFD for 7 days and simultaneously administered Met, Res or TUDCA by oral gavage. (A) Hypoxia staining in epididymal adipose tissue was viewed by using the hypoxia probe pimonidazole (200× magnification); (B) mRNA expression of HIF‐1α in epididymal adipose tissue was measured by Q‐PCR; (C) HIF‐1α protein expression in epididymal adipose tissues was determined by western blot. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus HFD feeding‐only treatment; ^# P < 0.05 versus the indicated treatment. NCD, normalchow diet.

Figure 2

Metformin (Met) and resveratrol (Res) suppressed ER stress in adipose tissue. Mice were fed a HFD for 7 days and simultaneously administered Met, Res or TUDCA by oral gavage. The phosphorylations of IRE‐1α (A) and eIF‐2α (B) in epididymal adipose tissues were assessed by western blot. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus HFD feeding‐only treatment; ^# P < 0.05 versus the indicated treatment.

Figure 3

Metformin (Met) and resveratrol (Res) inhibited adipose fibrosis and inflammation in HFD‐fed mice. Mice were fed a HFD for 7 days and simultaneously administered Met, Res or TUDCA by oral gavage. (A) mRNA expressions of Col3α, Col6α, elastin and LOX in epididymal adipose tissue were determined by Q‐PCR; (B) Masson's trichrome staining in adipose tissue (left). Bar, 50 μM. Adipocyte size indicated by average diameter (right); (C) mRNA expressions of IL‐6, TNF‐α, monocyte chemoattractant protein 1 (MCP‐1) and F4/80 in epididymal adipose tissue were determined by Q‐PCR. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus HFD feeding‐only treatment; ^# P < 0.05 versus the indicated treatment. NCD, normalchow diet.

Figure 4

Metformin (Met) and resveratrol (Res) inhibited palmitate (PA)‐induced HIF‐1α and LOX expression in adipose tissue. Adipose tissues from normal mice were pretreated with Met, Res, TUDCA or PX‐478 at given concentrations and then incubated with PA for 24 h. (A) mRNA expressions of HIF‐1α and LOX in adipose tissue were measured by Q‐PCR, and (B) HIF‐1α protein expression was assessed by western blot. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus PA‐only treatment; ^# P < 0.05 versus the indicated treatment.

Figure 5

Metformin (Met) and resveratrol (Res) inhibited HIF‐1α expression and ER stress in adipose tissue exposed to hypoxia. (A–E) Epididymal adipose tissue from normal mice was pretreated with the indicated agents at given concentrations and then cultured under hypoxic conditions (1% O₂) for 24 h. The HIF‐1α protein expression and the phosphorylation of IRE‐1α and eIF‐2α were detected by western blot. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus hypoxia‐only treatment; ^# P < 0.05 versus the indicated treatment.

Figure 6

Metformin (Met) and resveratrol (Res) inhibited fibrosis and inflammation in hypoxic adipose tissue. Epididymal adipose tissue from normal mice was pretreated with Met or Res at the given concentrations and then cultured under hypoxic conditions (1% O₂) for 24 h. (A) mRNA expressions of Col3α, Col6α, elastin and LOX were determined by Q‐PCR; (B) western blot was used to detect the phosphorylation of p65 protein; (C) mRNA expressions of TNF‐α and IL‐6 were measured by Q‐PCR. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus hypoxia‐only treatment; ^# P < 0.05 versus the indicated treatment.

Figure 7

Effects of metformin (Met) and resveratrol (Res) on AMPK. (A) Epididymal adipose tissue from normal mice was pretreated with Met and Res and then exposed to 1% O₂ for 24 h; (B) mice were simultaneously fed a HFD for 7 days and administered Met or Res by oral gavage. Western blot was used to detect the phosphorylation of AMPK. (C, D) 3T3‐L1 cells were transfected with AMPKA1/2 (AMPKα1/2)‐specific siRNA to silence AMPK and then incubated with Met and Res under 1% O₂ for 24 h. HIF‐1α protein expression was determined by western blot (C), and mRNA expressions of HIF‐1α and LOX were measured by Q‐PCR (D). Data are expressed as mean ± SD of five mice per group or n = 5 in quintuplicate for the cell line experiment. *P < 0.05 versus hypoxia or HFD feeding‐only treatment; ^# P < 0.05 versus the indicated treatment. NCD, normalchow diet.

Figure 8

Effects of metformin (Met) and resveratrol (Res) on ATP production and hypoxia in adipocytes. (A) 3T3‐L1 cells were incubated with Met or Res under normoxia for 4 h. ATP was assayed by an ATP Assay Kit; (B, C) 3T3‐L1 cells were exposed to hypoxia (1% O₂) (B) or palmitate (PA) (C) in the presence of Met, Res or AICAR, and hypoxia staining was viewed with the hypoxia probe pimonidazole at a 200× magnification; (D) OCR was measured in 3T3‐L1 cells under basal conditions and after exposure to PA in the presence of Met and Res using an XFe96 Extracellular Flux Analyzer. Data are expressed as mean ± SD (n = 5 in quintuplicate) for cell line experiment. ^# P < 0.05 versus the indicated treatment. FCCP, Carbonyl cyanide 4‐(trifluoromethoxy)phenylhydrazone.

Figure 9

Effect of resveratrol (Res) on the expression of SIRT1. Epididymal adipose tissue from normal mice was pretreated with the indicated agents and then cultured under hypoxic conditions (1% O₂) for 24 h. Protein expression of SIRT1 and HIF‐1α abundance was determined by western blot. Data are expressed as mean ± SD of five mice per group. *P < 0.05 versus hypoxia‐only treatment; ^# P < 0.05 versus the indicated treatment. Met, metformin; NA, niacinamide.