Empagliflozin ameliorates diabetic cardiomyopathy probably via activating AMPK/PGC-1α and inhibiting the RhoA/ROCK pathway

Abstract

Background: Diabetic cardiomyopathy (DCM) increases the risk of hospitalization for heart failure (HF) and mortality in patients with diabetes mellitus. However, no specific therapy to delay the progression of DCM has been identified. Mitochondrial dysfunction, oxidative stress, inflammation, and calcium handling imbalance play a crucial role in the pathological processes of DCM, ultimately leading to cardiomyocyte apoptosis and cardiac dysfunctions. Empagliflozin, a novel glucose-lowering agent, has been confirmed to reduce the risk of hospitalization for HF in diabetic patients. Nevertheless, the molecular mechanisms by which this agent provides cardioprotection remain unclear.

Aim: To investigate the effects of empagliflozin on high glucose (HG)-induced oxidative stress and cardiomyocyte apoptosis and the underlying molecular mechanism.

Methods: Twelve-week-old db/db mice and primary cardiomyocytes from neonatal rats stimulated with HG (30 mmol/L) were separately employed as in vivo and in vitro models. Echocardiography was used to evaluate cardiac function. Flow cytometry and TdT-mediated dUTP-biotin nick end labeling staining were used to assess apoptosis in myocardial cells. Mitochondrial function was assessed by cellular ATP levels and changes in mitochondrial membrane potential. Furthermore, intracellular reactive oxygen species production and superoxide dismutase activity were analyzed. Real-time quantitative PCR was used to analyze Bax and Bcl-2 mRNA expression. Western blot analysis was used to measure the phosphorylation of AMP-activated protein kinase (AMPK) and myosin phosphatase target subunit 1 (MYPT1), as well as the peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α) and active caspase-3 protein levels.

Results: In the in vivo experiment, db/db mice developed DCM. However, the treatment of db/db mice with empagliflozin (10 mg/kg/d) for 8 wk substantially enhanced cardiac function and significantly reduced myocardial apoptosis, accompanied by an increase in the phosphorylation of AMPK and PGC-1α protein levels, as well as a decrease in the phosphorylation of MYPT1 in the heart. In the in vitro experiment, the findings indicate that treatment of cardiomyocytes with empagliflozin (10 μM) or fasudil (FA) (a ROCK inhibitor, 100 μM) or overexpression of PGC-1α significantly attenuated HG-induced mitochondrial injury, oxidative stress, and cardiomyocyte apoptosis. However, the above effects were partly reversed by the addition of compound C (CC). In cells exposed to HG, empagliflozin treatment increased the protein levels of p-AMPK and PGC-1α protein while decreasing phosphorylated MYPT1 levels, and these changes were mitigated by the addition of CC. Adding FA and overexpressing PGC-1α in cells exposed to HG substantially increased PGC-1α protein levels. In addition, no sodium-glucose cotransporter (SGLT)2 protein expression was detected in cardiomyocytes.

Conclusion: Empagliflozin partially achieves anti-oxidative stress and anti-apoptotic effects on cardiomyocytes under HG conditions by activating AMPK/PGC-1α and suppressing of the RhoA/ROCK pathway independent of SGLT2.

Keywords: AMPK; Apoptosis; Diabetic cardiomyopathy; Empagliflozin; Oxidative stress; ROCK.

Figure 1

Effects of empagliflozin on cardiac function in db/db mice. A: M-mode imaging revealed reduced systolic and diastolic function in db/db mice; B: Quantitation of E/A ratio, left ventricular ejection fraction, and left ventricular fractional shortening. NC: Normal control mice; DB: Db/db mice; EM/DB: Db/db mice treated with empagliflozin; LVEF: Left ventricular ejection fraction; LVFS: Left ventricular fractional shortening. ^aP < 0.05 vs NC; ^bP < 0.05 vs DB.

Figure 2

Effects of empagliflozin on mitochondrial injury, apoptosis, and AMP-activated protein kinase, peroxisome proliferator-activated receptor-γ coactivator-1α, and the RhoA/ROCK pathway in db/db mice. A: Left ventricular sections stained with hematoxylin and eosin and by TdT-mediated dUTP-biotin nick end labeling to assess cardiomyocyte apoptosis: The nuclei of normal cardiomyocytes were blue while the nuclei of apoptosis-positive cardiomyocytes were brown. Transmission electron micrographs showed the effects of empagliflozin on the ultrastructure of the myocardium in db/db mice; B: Quantitation of apoptotic cells; C and D: The phosphorylation of AMP-activated protein kinase and myosin phosphatase target subunit 1, as well as the protein expression of peroxisome proliferator-activated receptor-γ coactivator-1α were measured by Western blot in three groups. Bars indicate the mean ± SD from three independent experiments (n = 3). β-tubulin was set as a control for normalization. NC: Normal control mice; DB: Db/db mice; EM/DB: Db/db mice treated with empagliflozin; HE: Hematoxylin and eosin; TEM: Transmission electron micrographs; p-AMPK: Phosphorylated AMP-activated protein kinase; p-MYPT1: Phosphorylated myosin phosphatase target subunit 1; TUNEL: TdT-mediated dUTP-biotin nick end labeling. ^aP < 0.05 vs NC; ^bP < 0.05 vs DB.

Figure 3

Effects of empagliflozin, fasudil, and overexpression of peroxisome proliferator-activated receptor-γ coactivator-1α on high glucose-induced cardiomyocyte apoptosis in vitro. A: Effects of different concentrations of empagliflozin on viability of cardiomyocytes exposed to high glucose (HG); B and C: Cardiomyocyte apoptosis measured by flow cytometry; D: TdT-mediated dUTP-biotin nick end labeling staining of cardiomyocytes in each group. Bars indicate the mean ± SD from three independent experiments (n = 3). NG: Normal glucose; HG: High glucose; EM: Empagliflozin; CC: Compound C; FA: Fasudil; HG + Ad–PGC: Cells were transfected with peroxisome proliferator-activated receptor-γ coactivator-1α-GFP-Ad; HG + Ad-GFP: Cells transfected with GFP-Ad. ^aP < 0.05 vs NG; ^bP < 0.05 vs HG; ^cP < 0.05 vs HG + EM.

Figure 4

Effects of empagliflozin, fasudil, and overexpression of peroxisome proliferator-activated receptor-γ coactivator-1α on apoptotic genes and related proteins in vitro. A and B: Bax and Bcl-2 mRNA expression was quantified using real-time quantitative PCR in cardiomyocytes; C: Protein expression of active caspase-3 measured by Western blot. Bars indicate the mean ± SD from three independent experiments (n = 3). β-tubulin was set as a control for normalization. NG: Normal glucose; HG: High glucose; EM: Empagliflozin; CC: Compound C; FA: Fasudil; HG + Ad–PGC: Cells were transfected with peroxisome proliferator-activated receptor-γ coactivator-1α-GFP-Ad; HG + Ad-GFP: Cells transfected with GFP-Ad. ^aP < 0.05 vs NG; ^bP < 0.05 vs HG; ^cP < 0.05 vs HG + EM.

Figure 5

Effects of empagliflozin, fasudil, and overexpression of peroxisome proliferator-activated receptor-γ coactivator-1α on high glucose-induced mitochondrial injury and oxidative stress in vitro. A: Levels of intracellular high glucose; B and C: Changes in mitochondrial membrane potential; D: Cellular ATP production; E: Cellular superoxide dismutase activity. Bars indicate the mean ± SD from three independent experiments (n = 3). ROS: Reactive oxygen species; MMP: Mitochondrial membrane potential; SOD: Superoxide dismutase; NG: Normal glucose; HG: High glucose; EM: Empagliflozin; CC: Compound C; FA: Fasudil; HG + Ad–PGC: Cells were transfected with peroxisome proliferator-activated receptor-γ coactivator-1α-GFP-Ad; HG + Ad-GFP: Cells transfected with GFP-Ad. ^aP < 0.05 vs NG; ^bP < 0.05 vs HG; ^cP < 0.05 vs HG + EM.

Figure 6

Relationship of empagliflozin with AMP-activated protein kinase, peroxisome proliferator-activated receptor-γ coactivator-1α, and the RhoA/ROCK pathway in cardiomyocytes in HG conditions in vitro. A-D: The phosphorylation of AMP-activated protein kinase and myosin phosphatase target subunit 1, as well as the protein expression of peroxisome proliferator-activated receptor-γ coactivator-1α were measured by Western blot in three groups; E: Sodium-glucose cotransporter (SGLT)1 and SGLT2 protein expression in cardiomyocytes. Bars indicate the mean ± SD from three independent experiments (n = 3). β-tubulin was set as a control for normalization. NG: Normal glucose; HG: High glucose; EM: Empagliflozin; CC: Compound C; FA: Fasudil; HG + Ad–PGC: Cells were transfected with PGC-1α-GFP-Ad. ^aP < 0.05 vs NG; ^bP < 0.05 vs HG; ^cP < 0.05 vs HG + EM.