Empagliflozin Ameliorates Diabetic Cardiomyopathy via Attenuating Oxidative Stress and Improving Mitochondrial Function

Diabetic cardiomyopathy (DCM) is considered to be a critical contributor to the development of heart failure. Empagliflozin (EMPA), a sodium-glucose cotransporter 2 inhibitor, has been shown to prevent cardiovascular events and reduce the incidence of heart failure in randomized clinical trials. However, the mechanism of how EMPA prevents DCM is poorly understood. To study the potential mechanisms involved in the therapeutic effects of EMPA, we assessed the protective effects of EMPA on myocardial injury in type 2 diabetic db/db mice and H9C2 cardiomyocytes. 9–10-week-old male db/db mice were treated with EMPA (10 mg/kg) via oral gavage daily for 20 weeks. Afterward, cardiac function of treated mice was evaluated by echocardiography, and pathological changes in heart tissues were determined by histopathological examination and western blot assay. EMPA markedly reduced blood glucose levels, improved insulin tolerance, and enhanced insulin sensitivity of db/db mice. In addition, EMPA significantly prevented cardiac dysfunction, inhibited cardiac hypertrophy and fibrosis, and reduced glycogen deposition in heart tissues. Furthermore, EMPA improved diabetes-induced oxidative stress and mitochondrial dysfunction in both heart tissues of db/db mice and palmitate exposed H9C2 cells. EMPA significantly increased the expression of nuclear factor erythroid 2-related factor 2 (Nrf2) and its downstream genetic targets in cardiac tissue of type 2 diabetic db/db mice and H9C2 cells. EMPA also downregulated the expression of mitochondrial fission-related proteins and upregulated the expression of mitochondrial fusion-related proteins. Collectively, these findings indicate that EMPA may prevent DCM via attenuating oxidative stress and improving mitochondrial function in heart tissue.


Introduction
According to the International Diabetes Federation, diabetes is one of the fastest-growing chronic diseases worldwide and is projected to affect 693 million adults by 2045 [1,2]. One of the major chronic complications associated with diabetes is diabetic cardiomyopathy (DCM) which affects approximately 70% of diabetic individuals. The clinical manifesta-tions of DCM include left ventricular (LV) hypertrophy, myocardial fibrosis, and consequently impaired LV systolic and diastolic function, which eventually cause coronary artery disease, cardiac insufficiency, heart failure (HF), and sudden death [3].
Multiple mechanisms play direct and/or indirect causal roles in the development of DCM, including insulin resistance, glucotoxicity, lipotoxicity, impaired calcium handling, oxidative stress, mitochondrial dysfunction, and inflammation [3,4]. Oxidative stress is considered to be one of the most pivotal factors involved in both the early and late stages of DCM [5]. In type 2 diabetes mellitus (T2DM), hyperglycemia, hyperlipidemia, hyperinsulinemia, and lipotoxicity result in the excess generation of reactive oxygen species (ROS) or reactive nitrogen species (RNS) and insufficient antioxidant capacity in the heart, thereby causing the pathogenesis of oxidative stress [6,7]. Oxidative stress is detrimental to cardiac function and promotes the development of DCM by inducing cardiac cell death, cardiac hypertrophy, extracellular matrix (ECM) remodeling, and progressive cardiac fibrosis [8,9]. Mitochondria are the powerhouses of cardiomyocytes and the main site for ROS production in cardiomyocytes [10]. Under physiological conditions, the excessive production of mitochondria-derived ROS can cause mitochondrial dysfunction, impair adenosine triphosphate (ATP) production in cardiomyocytes, and ultimately induce cell death and cause cardiac dysfunction [11]. In addition, oxidative stress regulates the morphology and function of mitochondria. Maintaining normal mitochondrial function is a prerequisite for providing energy to cells, but an imbalance between mitochondrial fusion and fission may cause excessive ROS production and result in dysfunctional mitochondria [12,13]. Therefore, alleviating oxidative stress and improving mitochondrial function may be a potential strategy for preventing the development of DCM.
Common therapeutic options for T2DM focus on insulin, metformin, and sulfonylureas. However, there is no evidence these conventional hypoglycemic drugs reduce the incidence of cardiovascular disease in people with T2DM, and some hypoglycemic drugs may even increase the risk of HF [14][15][16]. Empagliflozin (EMPA) is a novel hypoglycemic agent that was approved for clinical use by the Food and Drug Administration on August 1, 2014 [17]. Unlike conventional hypoglycemic agents, EMPA lowers blood glucose levels by specifically inhibiting the activity of sodiumglucose cotransporter-2 (SGLT-2) in the S1 segment of the proximal renal tubule [18]. By doing so, treatment with EMPA results in excess blood glucose being transferred into the urine, thus lowering blood glucose levels in the body. The EMPA-REG OUTCOME trial demonstrated that treatment with EMPA reduces the risk of cardiovascular-related death by 38% in patients with T2DM with or without cardiovascular disease, and especially reduces the risk of allcause mortality, cardiovascular mortality, and hospitalization (or readmission) for HF, and also decreases the probability of myocardial infarction or stroke [19]. The previous findings suggest that EMPA obviously improved cardiac function and ameliorated cardiac hypertrophy and fibrosis in different T2DM models [20][21][22]. Furthermore, a prospective observational study demonstrated that the protective effects of EMPA on LV dysfunction were more effective in the early stage of DCM than in advanced DCM, which means the early intervention of EMPA may be preferable for DCM [23]. These findings indicate promising potential for the application of EMPA in the treatment and/or reversal of DCM. However, the mechanisms of cardioprotective effects of EMPA are still not clear.
Here, we aimed to determine the potential protection of EMPA against myocardial injury in type 2 diabetic db/db mice and investigate the underlining mechanisms. In addition, we studied the effects of EMPA treatment on diabetic-like cardiac damage in vitro on the palmitate-(PA-) exposed H9C2 cells. These findings demonstrated that EMPA significantly improved cardiac function and inhibited cardiac hypertrophy and fibrosis via attenuating oxidative stress and improving mitochondrial function. Furthermore, EMPA attenuates oxidative stress and improves mitochondrial function mainly through activating the nuclear factor erythroid 2-related factor 2 (Nrf2) signaling pathway.

Materials and Methods
2.1. Animal Procedure and Drug Treatment. 9-10 week-old diabetic db/db male mice (BKS.Cg-Dock7m+/+Leprdb) and  [24]. To study the effects of EMPA on the elevated levels of NEFAs that occur in obesity and T2DM, H9C2 cells were exposed to PA to mimic the cellular metabolic environment of obesity and type 2 diabetes. PA stock solution (10 mM) was prepared by dissolving sodium palmitate 3 Oxidative Medicine and Cellular Longevity (Sigma-Aldrich, St. Louis, MO, USA) in 50% ethanol solution mixed with 2% fatty acid-free bovine serum albumin (BSA) solution, as previously described [25]. H9C2 cells were exposed to 200 μM PA under a range of concentrations of EMPA (Selleck, Houston, TX).

Oral Glucose Tolerance Test and Intraperitoneal
Injection Insulin Tolerance Test. Mice were fasted overnight and then treated with either oral glucose (1 g glucose/kg body weight) for oral glucose tolerance test (OGTT) procedures, or intraperitoneal insulin (1 U insulin/kg body weight) for intraperitoneal injection insulin tolerance test (IPITT) procedures. Blood glucose level of each mouse was measured at the stated time points using a glucometer and blood sampled from the tail vein.

Histological Staining and Immunohistochemical
Analysis. Mouse heart tissues were fixed overnight in 4% paraformaldehyde, then dehydrated, embedded in paraffin, and sectioned at 5 μm thickness. To evaluate pathological changes, paraffin sections of heart tissues were stained with hematoxylin and eosin (H&E; Keygen Biotech, China). Sirius-red staining was used to detect heart tissue fibrosis.

Transmission Electron Microscope (TEM) Observation.
Tissue specimens were fixed in 2.5% glutaraldehyde solution at 4°C and repaired with 1% osmium tetroxide. Tissue specimens were dehydrated and then embedded in epoxy resin. Slides mounted with 50 nm sections were incubated with lead citrate or uranium acetate and observed under a Hitachi H600 electron microscope (Hitachi, Japan). 2.13. Mitochondrial Membrane Potential. Mitochondrial membrane potential was assessed using a tetramethylrhodamine ethyl ester (TMRE) probe (Thermo Fisher). H9C2 cells were seeded in 96-well plate one day before treatment, and then, cells were exposed to 200 μM PA in the presence of different doses (100 nM, 250 nM, and 500 nM) of EMPA for 24 h. Then, the cell culture medium was removed and replaced with serum-free medium containing 100 nM TMRE incubating for 20 min at 37°C. After washing three times with PBS, the fluorescence intensity was examined at 549 (excitation) and 576 nm (emission) wavelengths with a microplate reader (VICTOR® Nivo™, perkinelmer, MA).
2.14. Statistical Analysis. All data are presented as mean ± standard deviation (SD). Group differences were determined by the One-way analysis of variance using Graphpad Prism version 8.0 software. P values of <0.05 were considered statistically significant.

Effects of EMPA on Metabolic Parameters of db/db Mice.
Diabetes-related biochemical assessments for the three experimental groups of mice are shown in Table 1. Compared with WT mice, db/db mice treated with vehicle had higher average body weights and heart weight/tibial length ratios. Compared to db/db mice in vehicle, db/db mice treated with EMPA had slightly lower average body weights and heart/tibial length ratios, but these values were not 11 Oxidative Medicine and Cellular Longevity significantly different compared with db/db mice (P > 0:05). Both the fasting and nonfasting blood glucose levels of db/db mice treated with either vehicle or EMPA were obviously higher than those in the WT group. After treatment for 20 weeks, nonfasting blood glucose in db/db mice treated with EMPA was lower than in db/db mice treated with vehicle. In addition, the plasm TC, LDL-C, and TG were significantly higher in db/db groups than in WT mice (P < 0:05). However, db/db mice receiving EMPA treatment had significantly lower TC but not TG levels than db/db mice receiving vehicle. In addition, db/db mice receiving EMPA treatment had significantly higher HDL-C levels than db/ db or WT mice receiving vehicle. The effects of EMPA on glucose tolerance and insulin resistance in db/db mice were evaluated by OGTT and IPITT assay, respectively. During OGTTs, blood glucose levels in db/db mice were markedly higher than those of WT mice at all time points, with greater areas under the curve (AUC). EMPA treatment significantly improved the glucose tolerance in db/db mice, as shown by the better glucose disposal measures and lower AUC in db/ db mice treated with EMPA group (Figures 1(a) and 1(b)). During IPITTs, db/db mice showed greater insulin resistance, manifested by a reduced hypoglycemic effect of insulin and increased AUC compared to WT mice (P < 0:05). However, in EMPA-treated db/db mice, the AUC during IPITT was obviously lower than in db/db mice treated with vehicle, suggesting that EMPA improves insulin tolerance and enhances insulin sensitivity in db/db mice (Figures 1  (c) and 1(d)).

EMPA Prevents Cardiac Dysfunction in db/db Mice.
Echocardiographic assays were performed to determine whether EMPA treatment improved cardiac function of db/db mice ( Figure 2 and Table 2). The results demonstrated that cardiac function in db/db mice was impaired, reflected by a lower HR, LVEF, and LVFS and higher LVIDd and LVIDs than WT mice. In contrast, db/db mice treated with EMPA had a significantly better cardiac function, with higher LVEF and LVFS scores, than that in db/db mice in the vehicle group.

EMPA Inhibited Diabetes-Induced Pathological Changes in Myocardial
Tissue. H&E histological staining showed that whereas the myocardial fibers of WT mice are neatly arranged, the myocardial fibers of db/db mice are disordered. However, db/db mice treated with EMPA showed improvements in the arrangement of their myocardial tissue fibers compared to db/db mice in the vehicle group (Figure 3(a)).
We investigated the role of EMPA treatment on glycogen deposition in the heart tissue of db/db mice by PAS staining. As shown in Figure 3(b), db/db mice treated with vehicle had a greater level of heart tissue glycogen deposition (indicated in magenta) than WT mice, EMPA treatment effectively inhibited glycogen deposition.
FITC-conjugated WGA staining was used to determine myocyte CSA. As shown in Figure 3(c), myocyte CSA in db/db mice treated with vehicle was markedly larger than in WT mice. While EMPA treatment significantly prevents the increase of myocyte CSA of db/db mice, compared to db/db mice in the vehicle group (P < 0:05).
Sirius-red staining was conducted to evaluate collagen accumulation and the degree of fibrosis of cardiac tissues. This demonstrated that a large amount of collagen was deposited in heart tissues in db/db mice treated with vehicle, but the extent of collagen deposition was markedly lower in db/db mice treated with EMPA (P < 0:05, Figure 3(d)). In addition, the expression of the profibrotic mediator, TGF-β1, was higher in the heart tissue of db/db group mice treated with vehicle than that of WT mice. However, this increase was no longer present in db/db mice treated with EMPA (Figure 3(e)).

EMPA Ameliorates Oxidative Stress in db/db Mouse
Heart Tissues and PA-Treated H9C2 Cells. Excessive oxidative stress in response to hyperglycemia and hyperlipemia is an inducer of DCM in mice [27]. The fluorescent probes DHE and Mito-SOX Red were used to measure the extent of accumulation of intracellular ROS or Mito-ROS in heart tissues, respectively. As shown in Figures 4(a) and 4(b), compared with WT mice, both intracellular and mitochondrial ROS levels were obviously increased in the heart tissues of db/db mice treated with vehicle. However, db/db mice treated with EMPA showed significantly reduced levels of ROS. Coinciding with these findings, the expressions of lipid peroxide 4-hydroxynonenalal and nitrosation production 3nitrotyrosine, as well as oxidative stress-related markers, were considerably higher in db/db mice treated with vehicle compared with WT mice, but these changes were markedly attenuated by EMPA treatment (Figures 4(c) and 4(d)).
We further evaluated the effects of EMPA on oxidative stress by studying intracellular ROS and Mito-ROS levels in PA-induced H9C2 cells with or without EMPA treatment. The results showed that PA increases both intracellular and Mito-ROS levels, while EMPA treatment significantly attenuates these effects ( Figure 5).

EMPA Improves Oxidative Stress by Activating the Nrf2
Signaling Axis. Nrf2 is crucial for oxidative stress regulation [28]. Thus, we investigated whether EMPA attenuates oxidative stress via activating the Nrf2 pathway. As shown in Figure 6(a), immunohistochemical analyses demonstrated the expression levels of Nrf2 and SOD2 were markedly decreased in heart tissues of db/db mice treated with vehicle compared with WT mice, EMAPA treatment prevented the decrease of Nrf2 and SOD2 expression. Western blot results also showed that the expression of Nrf2 and the target genes (NQO-1and SOD2) are decreased in the heart tissues of db/ db mice treated with vehicle and H9C2 cells treated with PA. However, in both db/db mice and H9C2 cells treated with PA, EMPA treatment resulted in increased expression of Nrf2 and upregulated expression of NQO-1and SOD2 (Figures 6(b) and 6(c)). Furthermore, we evaluated whether EMPA promoted Nrf2 translocating from the cytosol into the nucleus via immunofluorescence staining assay. The results showed that EMPA treatment not only increase Nrf2 expression but also promoted Nrf2 translocating into the nucleus of H9C2 cells (Figure 6(d)).

EMPA Improves Mitochondrial Morphology and
Function in the Heart of db/db Mice and PA-Treated H9C2 Cells. TEM was used to assess the effects of EMPA on mitochondrial morphology in heart tissues. Figure 7(a) demonstrates the mean size of mitochondria of the heart tissues in db/db mice treated with vehicle is lower than that in WT mice, indicating that diabetes induces mitochondrial fragmentation. In contrast, EMPA treatment increases the size of mitochondria and inhibits mitochondrial fragmentation. Mitochondrial morphology in H9C2 cells was evaluated using MitoTracker® Red CMXRos probes. As shown in Figures 7(b) and 7(d), the mitochondria of H9C2 cells become smaller and shorter in length after exposure to PA, compared with the BSA group. However, when cells are treated with different concentrations of EMPA under PA condition, the mitochondria become filamentous and larger. In addition, EMPA treatment significantly inhibits the decline of the mitochondrial membrane potential levels induced by PA (Figures 7(c) and 7(e)).

EMPA Regulates the Mitochondrial Dynamics-Related
Proteins in Myocardial Tissues and H9C2 Cells. The above results suggested that EMPA is able to inhibit the mitochondrial fission induced by diabetes, meaning that EMPA influences mitochondrial dynamics. Mitochondrial dynamics are mainly mediated by the activity of the dynamin superfamily of large GTPases, DRP1, OPA1 and MFN1, and MFN2 [29]. Therefore, we examined the expression of factors involved in the regulation of mitochondrial dynamics in heart tissues and H9C2 cells. In heart tissues, the expressions of mitochondrial fusion-related proteins (MFN1 and OPA1) were greatly lower in db/db mice in the vehicle group than in WT mice, while EMPA treatment resulted in greater expression of OPA1 and MFN1 (Figures 8(a)-8(c)), but not MFN2 (data not shown). The expression of DRP1 was markedly higher in db/db mice in vehicle than in WT mice, but EMPA treatment prevented this increase (Figure 8(a)). Consistent with our results in heart tissues, PA treatment downregulated the expression of OPA1 and MFN1 and upregulated the expression DRP1 in H9C2 cells, but EMPA treatment upregulated the expressions of OPA1 and MFN1 and inhibited the increase of DRP1 expression induced by PA (Figure 8(b)).

Discussion
EMPA is a novel hypoglycemic agent in diabetic patients, showing the benefits for protection against the development of DCM. However, the exact functions and mechanisms by which EMPA protects against DCM are not been studied. Here, we evaluated the effects of EMPA on DCM in db/db mice and PA-treated H9C2 cells. Our results demonstrated that EMPA protected against diabetes-induced myocardial dysfunction, inhibited myocardial fibrosis and hypertrophy, and attenuated glycogen deposition in the hearts. Furthermore, EMPA treatment attenuated oxidative stress and improved mitochondrial function in heart tissues of db/db mice and H9C2 cells treated with PA. Finally, our results demonstrated that EMPA treatment reverses Nrf2 signaling in diabetic heart and H9C2 cells to attenuate oxidative stress and improve mitochondrial function through increasing the expression of Nrf2 and its downstream genes (NQO-1 and SOD2) (Figure 8(d)).
DCM is a common complication of diabetes. Myocardial fibrosis and hypertrophy are important pathological features of DCM, which reduce cardiac compliance, aggravate ventricular stiffness, and eventually lead to cardiac insufficiency [30]. The present study found that EMPA prevented DCM as reflected by the reduced myocardial collagen fiber area and glycogen deposition and inhibited myocardial hypertrophy. EMPA, as an SGLT-2 inhibitor, is a new class of drugs for the treatment of T2DM via increasing glucose excretion in the urine. In addition to its excellent glycemic effects, SGLT-2 inhibitors also have multiple protective effects on cardiac function via attenuating cardiac fibrosis, inflammation, and oxidative stress [31][32][33]. Nevertheless, SGLT-2 is predominantly expressed in the kidney, but not in other tissues such as the heart, which implies that the protective effects of EMPA on cardiac function are independent of organ-specific SGLT-2 expression [34][35][36]. Future studies need to be performed to elucidate whether the cardiacprotective effects of EMPA described here are dependent on the expression of the SGLT-2 receptor or not via experiments in SGLT-2 knockout or knockdown models.
Our findings reveal the potential mechanisms involved in EMPA's protective effects against DCM. Oxidative stress is caused by an imbalance between the production of oxygen/nitrogen free radicals and endogenous physiological antioxidant mechanisms. Oxidative damage caused by free radicals is considered an important factor that contributes to the development of T2DM and its complications. Growing evidence indicates excessive ROS levels in the heart of diabetes increase the risk of arrhythmia and mitochondrial dysfunction, which leads to myocardial cell necrosis and apoptosis [37][38][39]. In addition, ROS can also modify protein excitation-contraction coupling, activate hypertrophy signaling kinases, impair endothelial function and vascular homeostasis, and gradually exacerbate the damage associated with angiogenesis and new blood vessels information, eventually, resulting in cardiac dysfunction [1,26]. Thus, alleviating oxidative stress is a potential strategy for preventing DCM. As expected, ROS and lipid peroxides were dramatically accumulated in the heart tissues of db/db mice and in PA-treated H9C2 cells, while EMPA treatment significantly attenuated the oxidative stress via inhibiting the accumulation of ROS and lipid peroxides. We also detected the dose-response of EMPA on H9C2 cells and found a medium dose (250 nM) of EMPA is more effective than a high or low dose, implying that the effects of EMPA are not dose-dependent. One reason for this may be that the high concentration of EMPA used here caused cytotoxicity. Thus, optimal doses of EMPA will need to be evaluated for clinical use.
Nrf2 is a basic leucine zipper stress-responsive transcription factor that plays a critical role in ameliorating oxidative stress by activating the expression of antioxidant proteins [40][41][42][43]. To investigate whether EMPA attenuates oxidative stress via activating the Nrf2 signaling pathway, we studied 13 Oxidative Medicine and Cellular Longevity the expression of Nrf2 and its downstream antioxidant genes (NQO1 and SOD2) in the heart tissues and H9C2 cells. The results demonstrated that the decreased expression of Nrf2 and its downstream antioxidant genes in diabetes was reversed by the treatment with EMPA, indicating that the antioxidant effects of EMPA are largely through activating Nrf2 signaling. However, how EMPA activates Nrf2 requires further elucidation.
Mitochondria are the primary organelles of ROS production and are also the primary targets of the adverse effects of ROS. Exposure to high glucose conditions enhances mitochondrial ROS production and induces oxidative stress, which in turn impairs mitochondrial function [44]. Emerging evidence has confirmed that mitochondrial dysfunction is a critical contributor to the development of DCM [45,46]. Mitochondrial function is reflected in mitochondrial structure and morphology. Mitochondria in T2DM patients are smaller than that in healthy controls; moreover, hyperglycemia can result in mitochondrial fragmentation in various cell types [13,47,48]. In our study, we found the size of mitochondria in the heart tissues of db/db mice is much smaller than that of WT mice. In addition, PA indues mitochondrial fragmentation in H9C2 cells, whereas EMPA treatment almost normalized the size of mitochondria in db/db mice and remarkably prevented this diabetesinduced mitochondrial fragmentation in cardiomyocytes in PA treated H9C2 cells. Mitochondrial morphology is dynamically regulated by mitochondrial fusion and fission, and several dynamin-related GTPases are involved in this process. In our experiments, EMPA significantly promoted the expressions of mitochondrial fusion-related proteins (MFN1 and OPA1) and also inhibited the expression of DRP1. These findings indicate that EMPA may ameliorate mitochondrial function by inhibiting the mitochondrial fission in the heart of diabetes. However, how EMPA regulates the expression of mitochondrial-related proteins is needed to be clarified in future.

Conclusion
In summary, this study indicated that DCM is associated with oxidative stress and mitochondrial dysfunction, and that EMPA protects against DCM through effectively inhibiting oxidative stress by activating Nrf2 and its downstream antioxidant genes and improving mitochondrial function via inhibiting mitochondrial fragmentation. Our present study provides a promising therapeutic approach for clinical translation to manage the progressive DCM in individuals with diabetes.

Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest
No potential conflicts of interest relevant to this article were reported.

Authors' Contributions
J.W., X.H., H.L., and Y.C. were responsible for the methodology; J.W., X.H., J.L., and P.L. were responsible for the data curation; Y.R., J.H., and E.X. were responsible for the formal analysis; J.W., L.L., Z.T., and X.D. were responsible for writing the original draft; X.D. was responsible for the project administration; Z.T. and X.D. were responsible for the funding acquisition. All authors read and approved the final manuscript. Jinwu Wang, Xinyuan Huang, Hanjie Liu, and Yuhang Chen contributed equally to this work.