Dapagli ozin Protects Doxorubicin-induced Cardiotoxicity in Breast Cancer Patients With Diabetes via Suppressing ER Stress


 Background: Cancer patients with diabetes have an increasing risk of Dox-induced cardiotoxicity. Despite previous studies reporting benefits of dapagliflozin on the cardiovascular system, it remains unknown whether dapagliflozin has a cardioprotective effect in cancer patients with diabetes receiving Dox therapy. The purpose of this study was to investigate the potential of dapagliflozin for preventing doxorubicin (Dox)-induced cardiotoxicity.Methods: Using Taiwan National Health Insurance Database, the incidence of heart failure of cancer patients with or without diabetes was investigated. Streptozotocin (STZ)-induced diabetic rats were pretreated with oral dapagliflozin (10 mg/kg/day) for 6 weeks followed by Dox (5 mg/kg/week) for 4 weeks via intraperitoneal injection. Sequential echocardiography was applied to assess cardiac function. For in vitro analysis, cardiomyocytes cultured in high glucose (30 mM) were treated with dapagliflozin at 10 M and subsequently exposed to Dox at 1 M. Apoptosis and endoplasmic reticulum (ER) stress-related protein expression were measured by immunohistochemistry and Western blotting.Results: Among the studied patients, those with diabetes had a higher risk of major adverse cardiovascular events including the development of heart failure. In diabetic rats, dapagliflozin reduced cardiac fibrosis and significantly improved cardiac function. Dapagliflozin effectively inhibited Dox-induced apoptosis and reactive oxygen species in cardiomyocytes under high glucose. Mechanistically, we showed that dapagliflozin decreased the cardiac expression of Bax and cleaved caspase 3 but increased Bcl-2. Dapagliflozin also significantly reduced ER stress-associated proteins including GRP78, PERK, eIF-2, ATF-4, and CHOP. Conclusions: Our study revealed for the first time that dapagliflozin mitigated Dox-induced cardiomyocyte apoptosis in diabetes via inhibiting ER stress. These results indicate that dapagliflozin could be useful for preventing cardiotoxicity in diabetic cancer patients receiving Dox treatment.

At the end of the experiments, the rats were sacri ced and fresh heart tissues were immediately collected. The weight of the heart tissue was measured. For histopathological examination, the heart tissue was xed in 4% paraformaldehyde and embedded in para n (Alfa Aesar, Lancashire, UK). Sections were stained with hematoxylin-eosin (HE) and Masson's trichrome stain. The rest of the heart tissue was frozen in liquid nitrogen and stored at -80℃ for further biochemical assays.

Measurement of cell apoptosis via ow cytometry
Apoptosis of the H9C2 cardiomyocytes was measured using the annexin V/propidium iodide (PI) doublestaining method. After treatment, the cells were harvested and washed twice with ice-cold PBS. The cells were resuspended in binding buffer and then incubated with annexin V and PI working solution for 15 min in the dark at room temperature. Cellular uorescence was measured via ow cytometry (Becton Dickinson, Franklin Lakes, NJ, USA).

TUNEL staining
A terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay was conducted using an in situ cell death detection kit (BioVision, Milpitas, CA, USA) according to the manufacturer's protocol.

Western blotting
After the indicated treatments, the H9C2 cardiomyocytes were harvested and lysed with ice-cold RIPA buffer (Merck Millipore, Burlington, MA, USA). The total protein concentrations were determined using a BCA Protein Assay kit (Thermo Fisher Scienti c, Waltham, MA, USA). The method has been addressed previously [4,5]. Antibodies were listed in Supplement Materials.

Statistical analysis
The chi-squared test was used to compare differences in age and comorbidity frequencies between breast cancer patients with and without diabetes. After testing for normality, continuous variables were compared between diabetic and non-diabetic patients using the Mann-Whitney U test. The Kaplan-Meier method was used to plot MACCEs, and group differences were compared via the log-rank test. The hazard ratio (HR) of MACCEs between cancer patients with and without diabetes was estimated using the Cox proportional hazard regression model adjusted for the potential confounding factors age and comorbidities. A two-tailed P value < 0.05 was considered statistically signi cant for all of the tests. All of the analyses were conducted using SAS software version 9.4 (SAS Institute, Cary, NC, USA). Kaplan-Meier curves were plotted using STATA (version 12; Stata Corp., College Station, TX, USA).

Results
Diabetes increased the cardiovascular risks but not all cause mortality in patients with breast cancer Compared with breast cancer patients naïve to anti-diabetic medications (N37, 962), those receiving more than three months of anti-diabetic medications (N = 762) were relatively older and had more comorbidities (Table 1). The breast cancer stages at diagnosis were similar between the two groups and only small amounts of patients received cardiovascular drugs. Notably, despite no signi cant difference in all-cause mortality between breast cancer with and without diabetes, signi cantly more of the diabetic cancer patients had subsequent MACCEs, especially hospitalization for heart failure. To reduce the potential bias, age and comorbidities were adjusted for the nal study subjects. The adjusted hazard ratio showed that the diabetic cancer patients had higher risks of MACCEs (hazard ratio: 2.12; con dence interval: 1.75-2.57, P < 0.0001) and heart failure (hazard ratio: 2.17; con dence interval: 1.63-2.90, P < 0.0001) than those without diabetes (Table 2). Nevertheless, the risk of all-cause mortality was similar between the two groups. With the increases in age and clinical stages of breast cancer, the risks of MACCEs, heart failure, and all-cause mortality elevated incrementally. Even after adjusting for comorbidities, chronic kidney disease still contributed to the risks of all endpoints. Correspondingly, the Kaplan-Meier survival plots presented higher cumulative incidence rates of MACCEs and heart failure in the diabetic cancer patients than those free from diabetes. Interestingly, the probabilities of all-cause mortality were similar between the breast cancer patients with and without diabetes (Fig. 1).
Dapagli ozin improves cardiac function, survival and heart weight in Dox-treated STZ rats To investigate whether dapagli ozin exerts bene cial actions on Dox-induced cardiotoxicity under high glucose, we established a Dox-induced cardiotoxicity model in the STZ rats. After the induction of diabetes using STZ, the rats were administered dapagli ozin for six weeks followed by another four weeks of Dox treatment (shown in Supplement Figure 2). The Dox-treated STZ rats had signi cantly lower weight that the non-Dox treated rats. Pretreatment with dapagli ozin mitigated the reduction in body weight compared with the Dox-treated STZ rats (shown in Supplement Figure 3a, P < 0.001). Pretreatment with dapagli ozin decreased blood glucose but had no effects on heart rates compared with the Dox-treated STZ rats (shown in Supplement Figure 3b and c, P < 0.001). Cardiac function was measured by sequential echocardiography in the control, STZ, STZ+Dox, and STZ+Dox+Dapa groups. Although there was no difference in IVSd and LVIDd among the four groups, the STZ+Dox group presented a signi cant decline in left ventricular systolic function including EF and FS (P < 0.05) while pretreatment with dapagli ozin mitigated Dox-induced cardiotoxicity (Fig. 2a, P < 0.001). This result indicated the protective potential of dapagli ozin against Dox-induced cardiotoxicity in STZ rats.
The mortality of the rats in the STZ+Dox+Dapa group was lower than that in the STZ+Dox group ( Fig. 2b, P = 0.16). In the post-mortem study, we investigated the effect of dapagli ozin on cardiac structure and lung injury after Dox treatment. Dox-induced cardiac toxicities were observed through an increase in the ratio of heart to body weight ( Fig. 2c, P < 0.05) as well as the weight to dry (D/W) lung weight ratio in the Doxtreated STZ rats while dapagli ozin signi cantly alleviated injuries ( Fig. 2d, P < 0.05).
Dapagli ozin improves cardiac function and structure in Dox-treated STZ rats Using the P-V loop analyses, we further studied the effect of dapagli ozin on hemodynamics in Dox-induced cardiac dysfunction in the STZ rats. Fig. 3a shows representative results of the P-V loop analyses with different preloads in the control, STZ, STZ+Dox, and STZ+Dox+Dapa groups. Compared with the controls, both Ves and Ved were higher in the STZ rats, especially those treated with Dox, and recovered by pretreatment with dapagli ozin (Fig. 3b, P < 0.05). Likewise, the maximal velocity of the pressure rise (+dP/dt) and fall (-dP/dt) was suppressed in the STZ rats treated with Dox and mitigated by dapagli ozin (Fig. 3c, P < 0.05). Despite no signi cant changes in arterial elastance (Ea), the decline in the exponential decay of the left ventricular pressure in isovolumic relaxation (tau) in the rats treated with STZ+Dox was also reversed by dapagli ozin (Fig. 3d, P < 0.05). Through temporal clamping the abdominal inferior vena cava, we found that although the EDPVR was not signi cantly different among the groups, the ESPVR was blunted in the rats treated with STZ+Dox, which was reversed by dapagli ozin (Fig. 3e, P < 0.05). Our ndings implied that pretreatment with dapagli ozin mitigated the Dox-induced hemodynamic suppression in the STZ rats.
Dapagli ozin attenuates Dox-induced cardiac brosis and apoptosis through inhibiting ER stress in STZ rats Compared with the STZ rats, cardiac brosis was signi cantly increased in the STZ rats treated with Dox for 28 days (P < 0.001) but signi cantly attenuated in the rats pretreated with dapagli ozin (Fig. 4a, P < 0.001).
Using TUNEL and F-actin staining, we further evaluated the apoptotic cardiomyocytes in the four groups. Compared with the STZ group, Dox signi cantly increased the numbers of apoptotic cardiomyocytes (P < 0.001), which was alleviated in the rats pretreated with dapagli ozin (Fig. 4b, P < 0.001). The apoptosisrelated proteins, including Bax, cleaved caspase 3, and Bcl-2, were measured in the cardiac tissue by Western blotting (Fig. 4c). The results showed that Dox treatment markedly upregulated the levels of pro-apoptotic proteins, such as Bax and cleaved caspase 3 (P < 0.05), in the STZ rats, while protein expression was signi cantly suppressed in the rats pretreated with dapagli ozin compared with the control group (P < 0.001). Conversely, Bcl-2 expression, a key regulator of apoptosis, was markedly downregulated in the Doxtreated STZ rats (P < 0.05), whereas pretreatment with dapagli ozin preserved the expression of Bcl-2 (P < 0.05).
To study the effects of dapagli ozin on Dox-induced ER stress, we evaluated the cardiac expression of ER stress-associated proteins including GRP 78, p-PERK, eIF-2 , ATF4, and CHOP among the four groups.
Compared to the control group, Dox treatment signi cantly increased the expression of GRP 78, p-PERK, and eIF-2 (P < 0.01), while pretreatment with dapagli ozin suppressed the increase in p-PERK and eIF-2 expression (Fig. 4d, P < 0.01 and P < 0.05, respectively). For ER stress-induced apoptosis, CHOP is the downstream signal of the p-PERK-eIF2 -ATF4 pathway in unfolded protein response. The results revealed that Dox treatment markedly induced the expression of ATF4 and CHOP in the cardiac tissue of the STZ rats (P < 0.05). Compared to the Dox+STZ group, pretreatment with dapagli ozin signi cantly inhibited the expression of ATF4 and CHOP (P < 0.05).
Dapagli ozin attenuates Dox-induced ROS generation and apoptosis mediated by ER stress in cardiomyocytes under high glucose Using MTT assays, rst we evaluated the effects of Dox and dapagli ozin on cell viabilities in the cardiomyocytes. The cells exposed to Dox at concentrations of 1, 10, and 100 M for 24 h showed dosedependent cytotoxicity (shown in Supplement Figure 4a; P < 0.01), while those exposed to dapagli ozin at concentrations of 0.1, 1, 10, and 20 M presented no cytotoxic effects (shown in Supplement Figure 4b). To study whether dapagli ozin has bene cial actions on Dox-induced cardiotoxicity, cardiomyocytes were pretreated with 20 M of dapagli ozin before exposure to 1 M of Dox under high glucose (30 mM). Under high glucose, Dox enhanced the death of cardiomyocytes (P < 0.001) while dapagli ozin attenuated the death (Fig. 5a, P < 0.05). ROS is one of the main reasons for apoptotic death in cardiomyocytes. After exposure to Dox, the level of ROS in the cardiomyocytes increased ( Fig. 5b, P < 0.05). Pretreatment with dapagli ozin for 1 h signi cantly suppressed the Dox-induced ROS generation (P < 0.01), which might lead cardiomyocytes to undergo apoptosis under high glucose. To con rm this hypothesis, we pretreated cardiomyocytes with or without dapagli ozin followed by administration of Dox under high glucose, and the percentages of apoptosis were determined by TUNEL assays and annexin V/propidium iodide staining ( Fig.  5c and d, P < 0.001). Dapagli ozin signi cantly reduced Dox-induced apoptosis in the cardiomyocytes under high glucose (P < 0.001).
The apoptosis-associated proteins were detected by Western blotting. Dox administration signi cantly increased the expression of apoptosis-related proteins such as Bax and cleaved caspase 3 (P < 0.05 and P < 0.01, respectively). It also signi cantly decreased the expression of anti-apoptotic protein Bcl-2 in the cardiomyocytes under high glucose (Fig. 5e, P < 0.05). Furthermore, pretreatment with dapagli ozin effectively inhibited the expression of both Bax and cleaved caspase 3 (P < 0.01 and P < 0.05, respectively) but increased the expression of Bcl-2 (P < 0.05) in the cardiomyocytes after Dox exposure under high glucose. To investigate the anti-apoptosis mechanism of dapagli ozin on Dox-induced apoptosis under high glucose, we further measured the expression of ER stress. Dox treatment signi cantly triggered the upregulation of GRP78 (P < 0.001), p-PERK (P < 0.05), eIF-2 (P < 0.05), ATF4 (P < 0.05), and CHOP (P < 0.01) in the cardiomyocytes under high glucose (Fig. 5f). Pretreatment with dapagli ozin effectively inhibited ER stress-associated protein expression in the Dox-treated cardiomyocytes under high glucose. These results suggested that dapagli ozin could inhibit Dox-induced apoptosis via suppressing of ER stress under high glucose.

Discussion
In this study, we found that rst, among cancer patients receiving Dox therapy, those with diabetes had higher risks of cardiovascular events and especially heart failure, but not all-cause mortalities. This emphasized the importance of treating diabetes optimally in cancer patients preparing for Dox therapy, which is another stress on the heart. Second, by suppressing ER stress-associated proteins, dapagli ozin signi cantly reduced myocardial brosis and restored cardiac function in the Dox-treated rats. Also, in the cardiomyocytes, pretreatment with dapagli ozin effectively inhibited Dox-induced apoptosis and reactive oxygen species. Based on these ndings, dapagli ozin may mitigate chemotherapy-induced cardiotoxicity in patients with concomitant diabetes by regulating ER stress (Fig. 6).
Dox is one of the most common chemotherapeutic drugs and is typically used to treat patients with breast cancer [11]. However, Dox-induced myocardial dysfunction remains a major challenge in clinical practice. The mechanisms involved in Dox-induced cardiotoxicities have been reported including oxidative stress by ROS [12], mitochondria dysregulation and topoisomerase II b inhibition [13]. This suggests that Dox-induced cardiotoxicity is an indicator of multiplex biological processes. Some strategies have been reported to possibly prevent Dox-induced cardiotoxicity such as dexrazoxane, an iron chelator. It has been found to bind free iron and remove iron from its complex with Dox [14]. Despite being approved as a cardioprotective agent to prevent Dox-induced cardiotoxicity, dexrazoxane has side effects [15]. It may increase risks of infection, cause secondary malignant neoplasms, reduce the e cacy of Dox, and is expensive. Thus, many ongoing studies that attempt to discover new agents against Dox-induced cardiotoxicity focus on interfering with oxidative stress, in ammation, and apoptosis [5]. Nevertheless, to date an optimal regimen for preventing and managing Dox-induced cardiotoxicity remains lacking.
With the increasing prevalence of diabetes in cancer patients, it remains unclear how Dox affects the diabetic heart. Dox-induced irreversible cardiotoxicity may happen in a variety of patients, especially cancer patients with diabetes [1]. Reports have demonstrated that ROS-mediated apoptosis in cardiomyocytes is a major mechanism of Dox-induced cardiotoxicity [16]. Likewise, increased ROS has been regarded as a central mechanism of cardiac dysfunction in patients with diabetes [17] ROS generation may contribute to a double hit of stress in diabetic cancer patients receiving Dox treatment. Therefore, inhibiting the generation of ROS could be a therapeutic target to reduce Dox-induced cardiotoxicity in diabetic cancer patients.
The goal of diabetes treatment is not only to control blood glucose but also maintain myocardial function. The EMPA-REG, CANVAS, and DAPA-HF trials have shown that SGLT2 inhibitors have cardioprotective effects in reducing cardiovascular adverse events including heart failure hospitalizations and cardiovascular mortality [18]. Therefore, SGLT2 inhibitors could be a priority choice in diabetic patients who are vulnerable to cardiovascular stress such as Dox therapy. In diabetic cardiomyopathy animal models, dapagli ozin has been found to improve cardiac morphologic and function including cardiac hypertrophy, brosis, and heart failure as well as both systolic and diastolic left ventricle function [19]. From another perspective, while some studies indicated that dapagli ozin could possibly decrease Dox-induced cardiotoxicity, the effect of dapagli ozin treatment on diabetic cancer patients receiving Dox therapy remains largely unknown. In this study, we recruited breast cancer patients with diabetes and found that diabetes increased their cardiovascular risks. Further, echoing our clinical ndings, using a diabetic cardiotoxicity animal model, we illustrated that Dox-induced cardiac dysfunction, fibrosis, and activated ER stress could be significantly reserved by pretreatment with dapagli ozin.
Previous studies indicated that ER stress plays a signi cant role in mediating Dox damage in hearts [20]. Accumulating studies demonstrated that Dox-induced oxidative stress and ER stress subsequently cause myocardial cell death through the apoptosis pathway [5,21]. Collectively, these studies indicated that the heart is more susceptible to Dox-induced oxidative stress given its high mitochondrial density. Further, the heart also lacks the anti-oxidant enzyme to detoxify oxidative stress-induced ROS. Hence, a large number of free radicals accumulating in the myocardium results in the destruction of the mitochondrial membranes and ER [22]. In this study, we hypothesized that the diabetic heart would be more susceptible to Dox-induced cardiotoxicity. Our results demonstrated that high glucose exacerbated Dox-induced apoptosis in cardiomyocytes by upregulating ER stress.

Conclusion
Dapagli ozin mitigated Dox-induced cardiomyocyte apoptosis in diabetes via inhibiting ER stress. Our results suggest that dapagli ozin could be useful for preventing cardiotoxicity in diabetic cancer patients receiving Dox treatment. Ethics approval and consent to participate: This study was approved by the Institutional Review Board of Chi-Mei Medical Center (CV code: 10406-E01). Given that the data is derived from the NHIRD databank, the consent to participate is not applicable.

Consent for publication: Not applicable
Availability of data and materials: The data is available upon the reasonable request to the corresponding author.
Competing interests: no con icts of interest Author contributions: All authors were involved in the conception and design of the study and data interpretation. WC and YL drafted the paper and performed data analysis. WC and YL were involved in the data analysis and interpretation. All authors critically revised the paper and approved it for submission.  Figure 1 Kaplan-Meier survival plots of the cumulative incidence rates of (a) major adverse cardio-and cerebrovascular events (MACCEs), (b) heart failure, and the probabilities of free from (c) all-cause mortality in breast cancer patients with diabetes   Expression of apoptosis-associated protein, including Bax, Bcl-2, and caspase 3, in the rat hearts were measured by Western blotting. (d) Expression of ER stress-associated protein, including GRP78, p-PERK, eIF-2 , ATF4, and CHOP, in the rat hearts were measured by Western blotting. *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the indicated groups (N = 3-6). including GRP78, p-PERK, eIF-2 , ATF4, and CHOP, in the cardiomyocytes. *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the indicated groups (N = 3-5).

Figure 6
Summary of the double-hit stresses of diabetes and doxorubicin therapy in patients with cancer through ROS, apoptosis, and ER stress that could be prevented by treatment with dapagli ozin.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.