Renal protection induced by physical exercise may be mediated by the irisin/AMPK axis in diabetic nephropathy

Abstract

In patients with diabetes, it has been suggested that physical exercise may reduce albuminuria and the progression of renal disease. However, the molecular mechanism by which physical exercise protects the kidney in diabetes remains poorly understood. The aim of the present study was to determine the contribution of muscle irisin secretion induced by aerobic physical exercise with the subsequent activation of AMPK for kidney protection under diabetic conditions. Aerobic physical exercise in rats protected the kidney in streptozotocin-induced diabetes. It reduced albuminuria, glomerular hypertrophy, and glomerular expression of collagen IV and fibronectin, as well as markers of kidney inflammation, when compared to sedentary diabetic rats. These effects were associated with elevation in muscle FNDC5/irisin and activity of AMPK in the diabetic kidney. However, the beneficial effects of exercise were lost when the diabetic rats were treated with CycloRGDyK, that in the bone it has been described as an irisin receptor blocker. In cultured human tubular (HK-2) cells, treatment with recombinant irisin counteracted the effect of high glucose in a dose-dependent manner. Irisin, per se, also activated AMPK in HK-2 cells. It is concluded that in diabetes, the renal protective effect of exercise may be mediated by the irisin/AMPK pathway.

Figure 1

Eight weeks of aerobic physical exercise in diabetic rats prevented markers of diabetic nephropathy. (A) Kidney weight. (B) Graphical quantitation of the glomerular area. (C) Representative photomicrography of renal tuft stained with HE (magnification: × 630). (D, E, G, H) Graphical quantitation of immunohistochemistry for type IV collagen (D), fibronectin (E), TNF-α (G), and macrophage infiltration (F4/80, H). (F, I) Photomicrography of glomerular immunohistochemistry for type IV collagen and fibronectin (F) and TNF-α and F4/80 (I) (magnification: × 630). (J, K) Western blot analysis of acetyl(Lys³¹⁰)NF-κB(p65) in the renal cortex (K), followed by quantitation of acetyl(Lys³¹⁰)NF-κB(p65)/vinculin ratio (J). Blots are representative of three independent experiments. Results are means ± SE. CT, nondiabetic; DM, sedentary diabetic; DM + Exe, exercised diabetic. HE, hematoxylin eosin; col IV, collagen IV; FN, fibronectin; F4/80, marker of macrophage infiltration. ^#p < 0.05 vs. CT; *p < 0.05 vs. DM. Scale bars = 20 μm.

Figure 2

Eight weeks of aerobic physical exercise in diabetic rats was associated with renal and muscle AMPK activation and an elevation in the expression of FNDC5-irisn and PGC1-α in the muscle. (A–C) Western blot analysis of pAMPK^(Thr172), AMPKα, pACC^(Ser79), ACC, and vinculin in the renal cortex followed by quantitation of pAMPK^(Thr172)/vinculin by AMPKα/vinculin ratio (A) and pACC^(Ser79)/vinculin by ACC/vinculin ratio (B). (D–H) Western blot analysis of FNDC5-irisn, GADPH, PGC-1α, vinculin, pAMPK^(Thr172), AMPKα, pACC^(Ser79), and ACC in the muscle (H), followed by quantitation of FNDC5-irisn/GADPH (D), PGC-1α/vinculin (E), pAMPK^(Thr172)/vinculin by AMPKα/vinculin (F), and pACC^(Ser79)/vinculin by ACC/vinculin ratio (G). (I) Muscle AMP/ATP ratio. (J–L) Correlations of muscle irisin vs. UACR (J) as well as muscle irisin vs. glomerular fibronectin (K) and renal acetyl NF-κB (L). Protein expression levels were normalized to vinculin or GAPDH. Blots are representative of three independent experiments. Results are means ± SE. CT, nondiabetic; DM, sedentary diabetic; DM + Exe, exercised diabetic. UACR, urinary albumin-to-creatine ratio. ACC, acetyl-CoA carboxylase. ^#p < 0.05 vs. CT; *p < 0.05 vs. DM.

Figure 3

Treatment with an irisin blocker abolished the beneficial effects of aerobic physical exercise in diabetic rats. (A) Albuminuria. (E) Photomicrography of glomerular immunohistochemistry for type IV collagen and fibronectin (magnification: × 630). (B, C) Graphical quantitation of immunohistochemistry for type IV collagen and fibronectin. (D) Quantitation of serum irisin levels. Results are means ± SE. DM, sedentary diabetic; DM + Cyclo, sedentary diabetic treated intraperitoneally with 1 mg/kg of αV integrin receptor inhibitor (CycloRGDyK); DM + Exe, exercised diabetic rats; DM + Exe + Cyclo, exercised diabetic treated intraperitoneally with 1 mg/kg of CycloRGDyK. ^#p < 0.05 vs. DM, *p < 0.05 vs. DM + Cyclo, and ^&p < 0.05 vs. DM + Exe + Cylo. Scale bars = 20 μm.

Figure 4

Irisin treatment reduces extracellular matrix accumulation and activates AMPK in HK-2 cells exposed to high glucose. (D, E) Western blot analysis of type IV collagen, fibronectin, and vinculin (D) and pAMPK^(Thr172), AMPKα (E) in HK-2 cells, followed by quantitation of collagen IV/vinculin (A), fibronectin/vinculin (B), and pAMPK^(Thr172)/vinculin by AMPKα/vinculin ratio (C). Protein expression levels were normalized to vinculin. Blots are representative of three independent experiments. Results are means ± SE. NG, normal glucose (5.6 mmol/L); HG, high glucose (30 mmol/L); HG + 5 ng/mL, HG plus 5 ng/mL of recombinant irisin; HG + 15 ng/mL irisin, HG plus 15 ng/mL of recombinant irisin; HG + 30 ng/mL irisin, HG plus 30 ng/mL of recombinant irisin. ^#p < 0.05 vs NG, *p < 0.05 vs HG.

Figure 5

Sera from diabetic exercised individuals and not from sedentary diabetics prevented extracellular matrix accumulation in HK-2 cells exposed to high glucose. (A–C) Western blot analysis of type IV collagen, fibronectin, and vinculin (C) in HK-2 cells, followed by quantitation of collagen IV/vinculin (A), and fibronectin/vinculin (B). Protein expression levels were normalized to vinculin. Blots are representative of three independent experiments. Results are means ± SE. NG + CT, normal glucose (5.6 mmol/L) plus 4% of serum from non-diabetic patients; HG + CT, high glucose (30 mmol/L glucose) plus 4% of serum from non-diabetic patients; HG + DM, high glucose (30 mmol/L glucose) plus 4% of serum from sedentary diabetic patients; HG + DM + Exe, high glucose (30 mmol/L glucose) plus 4% of serum from diabetic patients submitted to exercise training. ^#p < 0.05 vs NG + CT, *p < 0.05 vs HG + CT, and ^&p < 0.05 vs HG + DM.

Figure 6

Schematic representation of the mechanism by which irisin/AMPK may mediate the role of physical exercise in inducing renal protection in diabetes. Interrupted arrows linking muscle AMPK, PGC1 and FNDC-5 indicate that the pathway requires further investigation.