Dapagliflozin Suppresses ER Stress and Improves Subclinical Myocardial Function in Diabetes: From Bedside to Bench

Diabetic cardiomyopathy remains a main contributor of mortality in patients diagnosed with diabetes (1,2). Most patients with diabetes present with diastolic dysfunction, despite having preserved left ventricular ejection fraction (LVEF), and have other subclinical myocardial abnormalities, which are harder to detect by conventional echocardiography (3). Speckle-tracking echocardiography (STE), an echocardiographic imaging technique, represents the motion of cardiac tissue, by using the naturally occurring speckle pattern, and sensitively detects myocardial dysfunction at an early stage (4). Previous literature suggests that LV global longitudinal strain (GLS) is a viable metric to describe contractility and relaxation in both human and animal hearts, and such measurements have been used in models of diabetes (1,2,4).

Dapagliflozin (DAPA) is a sodium–glucose cotransporter 2 inhibitor (5). It serves as a novel antihyperglycemic agent through selectively blocking renal glucose reabsorption and thereby facilitating the elimination of blood glucose to the urine (6). A recently launched clinical trial demonstrated that DAPA reduced risk of worsening heart failure or death from cardiovascular causes for patients with heart failure with reduced ejection fraction compared with those who received a placebo, regardless of type 2 diabetes status (5). Sodium-hydrogen antiporter 1 (NHE1), a ubiquitous membrane-bound enzyme, and protein kinase C β-II isoform (PKCβII), a key regulator in a variety of cell functions, have been previously reported to be upregulated in hyperglycemia-induced cardiac dysfunction (7,8). Recent evidence showed that sodium–glucose cotransporter 2 inhibitors ameliorate cardiac hypertrophy and fibrosis through inhibiting cardiac NHE1 and PKCβ activity (9,10). Xu et al. (11) reported pathways involving endoplasmic reticulum (ER) stress inhibitors that prevent diabetes-induced ER stress–mediated apoptosis and subsequent cardiac dysfunction (12,13). Nevertheless, a detailed understanding of how DAPA reverses high glucose–induced myocardial dysfunction is lacking. In this translational study, by way of speckle-tracking imaging technique, we aimed to investigate the DAPA effect and underlying molecular mechanism of action on myocardial remodeling using both human and animal models of diabetes.

Patients at the Chi-Mei Medical Center who were diagnosed with diabetes according to the definition of the World Health Organization and preparing to receive DAPA (10 mg/day) were initially included between June 2016 and March 2017. Patients were to be treated with other types of antidiabetes medications, but the regimen could not be adjusted during the follow-up period. Patients with poor image windows, atrial fibrillation, documented heart failure, coronary arterial disease, and structural heart disease, including LVEF <50%, above-moderate valvular heart disease, and documented cardiomyopathy were excluded (Supplementary Fig. 1). All participants underwent echocardiographic evaluation before and after 6 months of DAPA therapy. The study was conducted according to the recommendations of the Declaration of Helsinki on biomedical research involving human subjects and was approved by the local ethics committee (Tainan, Taiwan) (IRB: 10307-003). Written informed consent was obtained from each participant.

Male Sprague-Dawley rats, 12-weeks-old, at ∼250–300 g, were obtained from the Animal Center of National Cheng Kung University College of Medicine. The animal experiments were approved and conducted in accordance with local institutional guidelines for the care and use of laboratory animals in Chi-Mei Medical Center (no. 100052307) and conformed to the Guide for the Care and Use of Laboratory Animals. Rats were divided into the following groups: control, streptozotocin (STZ)-induced diabetes, STZ + DAPA treatment for 6 weeks, and STZ + insulin treatment for 6 weeks. The details of the study design are presented in Supplementary Fig. 2. The details of pressure volume loop and histology measurement are addressed in Supplementary Material.

All data are available from the corresponding authors upon request.

After exclusion of 14 patients, 54 patients with diabetes who were receiving DAPA therapy were enrolled in the study. The average age of the included patients was 61.8 ± 7.7 years (Supplementary Table 1). More often the patients were male and overweight (BMI 27.1 ± 3.3 kg/m²). More than 80% of them had comorbidities, including hypertension and hyperlipidemia, and received antihypertensive medication and statins, respectively. More than one-half of the patients, other than DAPA, were also on metformin, another antidiabetes drug. During follow-up period, there was no change of antidiabetic drugs. After 6 months of DAPA, there was significant improvement in levels of hemoglobin A_1c. Despite the trend of body weight reduction and blood pressure lowering, which did not reach significance, there was a significant reduction of waist circumference. Baseline echocardiography prior to the use of DAPA therapy demonstrated that despite preserved LVEF (67.6 ± 7.2%), there was impaired diastolic function (E-to-A ratio 0.8 ± 0.2, E-to-e′ ratio 11 ± 4.7) and strain (GLS −14.8 ± 5.2%) (Supplementary Table 2). Notably, after 6 months of DAPA therapy, although there was no significant change regarding LVEF, both diastolic function (E-to-A ratio 0.9 ± 0.4, E-to-e′ ratio 7.1 ± 2.3) and strain (GLS −19.2 ± 3.7) improved significantly.

Post–STZ administration, in rats, blood glucose levels increased from 100 mg/dL to ∼300 mg/dL, demonstrating successful induction of diabetes and hyperglycemia in these rats. Subsequently, those rats were assigned to the following groups: control, STZ, STZ + DAPA (6 weeks), and STZ + insulin (6 weeks). Rats that were treated with STZ and either DAPA or insulin showed marked reduction in hyperglycemia (Supplementary Fig. 3A), while the diabetic rats (STZ alone) presented with a trend of body weight loss (Supplementary Fig. 3B). The hyperglycemic state induced in the diabetic rats (STZ alone) additionally led to increased food and water intake, and this increase and the effect could be attenuated by treatment with either DAPA or insulin (Supplementary Fig. 3C and D). Regarding the hemodynamic parameters, the heart rate and systolic and diastolic blood pressure were not significantly different among the different groups (Supplementary Fig. 4). At baseline, the echocardiography-measured cardiac structure and function were normal for all groups. Supplementary Figure 5 demonstrates the changes of interventricular septal end diastolic dimension, left ventricular end diastolic internal diameter, ejection fraction, and fractional shortening. Although there was no difference in interventricular septal end diastolic dimension and left ventricular end diastolic internal diameter among the four groups, STZ rats showed a significant decline of left ventricular systolic function, including ejection fraction and fractional shortening, but the adverse functional effects were partially reversed by the treatment with DAPA.

Given that the conventional parameters of echocardiography are not sensitive enough to detect subtle myocardial dysfunction, we investigated the diastolic function and STE in the study rats (Fig. 1A–D). First, diastolic function parameters including E-to-A ratio and TDI-derived average e′ reflected a decrease in diastolic function in the post-STZ diabetes state. Treatment with DAPA significantly mitigated the STZ-induced decline. In terms of GLS, in the rats, high glucose–induced impairment of GLS was rescued by DAPA treatment. Notably, and in contrast to DAPA, insulin failed to show a similar functional benefit on diastolic function and GLS. These results, taken together, indicate a potentially protective effect of DAPA against high glucose–induced cardiac dysfunction in STZ rats.

Using pressure-volume loop relationship analysis, we further investigated the effect of DAPA on hemodynamic changes in high glucose–induced cardiac dysfunction in STZ rats. Compared with the untreated control, both end-systolic volume and end-diastolic volume were higher in STZ rats and could be rescued by DAPA treatment (Fig. 1E–I). Likewise, high glucose–induced suppression of maximal velocity of pressure rise (+dP/dt) and fall (−dP/dt) was reversed by the specific treatment of DAPA and not insulin. In rats treated with STZ, there was evident decline of exponential decay of left ventricular pressure in isovolumic relaxation (τ), despite lack of significant change in arterial elastance. As we had anticipated, DAPA specifically reversed τ-values in the aforementioned STZ treatment group. Through temporal clamping of the abdominal inferior vena cava, we found that although the end-diastolic pressure-volume relationship was not significantly different among the groups, the end-systolic pressure-volume relationship was blunted in rats treated with STZ. DAPA reversed that effect as well.

In a postmortem study, we investigated the effect of DAPA on high glucose–induced cardiac and lung injuries. High glucose increased the ratio of heart to body weight as well the wet–to–dry lung weight ratio, while DAPA significantly alleviated the cardiac hypertrophy and fluid accumulation–related lung injuries (Fig. 2A–D). Masson trichrome staining revealed a significant increase of fibrotic intensity in the hearts of rats with STZ-induced diabetes compared with the control hearts. Notably, DAPA attenuated cardiac fibrosis in STZ-treated rats with diabetes but not insulin (Fig. 2E and F).

Using MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, we further studied the effect of DAPA on high glucose–induced cardiotoxicity in vitro. Cardiomyocytes were cultured in high glucose (30 mmol/L, for 24 h) before exposure to DAPA (20 μmol/L) for 1 h. Under high glucose, cell viability declined in cardiomyocytes, while DAPA was able to rapidly attenuate the cardiomyocyte death (Fig. 3A). Given that reactive oxygen species (ROS) is a known major contributor to apoptosis and apoptotic cell death in myocytes, the accumulation of ROS was measured in the cultured cells. After exposure to high glucose, the level of ROS in cardiomyocytes was seen to be significantly elevated, while DAPA treatment reduced the observed increase in ROS generation (Fig. 3B and C). To further confirm whether observed cell death occurs primarily by apoptosis, we used flow cytometry and found that DAPA significantly reduced high glucose–induced apoptosis in cardiomyocytes (Fig. 3D and E). Additionally, TUNEL assay demonstrated a reduction in cell death suggesting a possible protective effect of DAPA on suppression of high glucose–mediated apoptosis (Fig. 3F and G). The F-actin staining results showed that DAPA decreased high glucose–triggered hypertrophic changes in cardiomyocytes (Fig. 3H).

After incubation with 30 mmol/L glucose (high glucose), cardiomyocytes showed a significant upregulation of ER stress– and apoptosis-associated proteins, including GRP78, eIF-2α, CHOP, and cleaved caspase3, while treatment with DAPA rescued the expression of these proteins (Fig. 4). The same high glucose condition triggered upregulation of ATF4 and BAX in cardiomyocytes, while DAPA treatment partially attenuated the high glucose–triggered upregulation of ATF4 (P = 0.08) and BAX (P = 0.12) in cultured H9C2. Likewise, high glucose activated expressions of NHE1 and PKCβII, while the treatment of DAPA attenuated the phenomenon. These cellular results suggest that DAPA may act as a suppressor of high glucose–induced apoptosis via mitigating the rise in levels of ER stress proteins.

Diabetic cardiomyopathy in the absence of discernible coronary arterial or valvular disease has received attention (1). While diabetes is characterized by myocyte loss and myocardial fibrosis, it also leads to decreased elasticity in the heart and resultant impaired contractility (1,3,11,14). The main focus behind the pathophysiology has been on hyperglycemia-induced cardiotoxicity (15). Along the molecular lines, intracellular accumulation of ROS has been linked to ER stress, a perturbation of ER-associated basic cellular function resulting in a variety of dysfunction including that in protein folding, lipid synthesis, calcium homeostasis, and apoptosis (11,16). Recent reports also indicated the involvement of ER stress in the development of diabetic cardiomyopathy (11). To date, a direct molecular link for the development of diabetic cardiomyopathy, has remained elusive. Although the subjects with diabetes included in this study were free from heart failure, we still observed a significant improvement of diastolic function and strain post–DAPA use, while lacking a comparison of antidiabetes drugs is a limitation. The changes that we detected may or may not be reliably picked up by conventional echocardiography, while STE may overcome the limitation. Our goal was to use a multidisciplinary approach to investigate the therapeutic potential of DAPA as a protective molecule in the setting of diabetes-induced cardiac dysfunction (Supplementary Graphic Abstract). This study offers a glimpse into the mechanistic basis for DAPA’s pleiotropic action—and one that might prove to be of translational value in treatment of diabetes, diabetic cardiomyopathy, or both.

The results taken together indicate that 1) DAPA has a cellular protective effect that may occur through suppressing ER stress, 2) DAPA may prevent high glucose–induced cardiac apoptosis and fibrosis, and 3) STE may offer a clinical benefit in detecting subtle changes in early stages of diabetes such that patients may be selected for DAPA treatment in trials.

Funding. This study was supported by National Science Council, Chi-Mei Medical Center, Ministry of Science and Technology (MOST 108-2628-B-384, 109-2326-B-384-001-MY3), and the National Health Research Institute (NHRI-EX106-10618SC).

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

Author Contributions. W.-T.C. and Y.-W.L. contributed to the conception and design of the study, acquired and analyzed the data, and drafted the manuscript. S.F., J.-T.C., N.-W.K., C.-S.H., Z.-C.C., and J.-Y.S. critically revised the manuscript. W.-T.C. and J.-Y.S. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.