Cyclovirobuxine D protects against diabetic cardiomyopathy by activating Nrf2-mediated antioxidant responses

Diabetic cardiomyopathy (DCM) is the principal cause of death in people with diabetes. However, there is currently no effective strategy to prevent the development of DCM. Although cyclovirobuxine D (CVB-D) has been widely used to treat multiple cardiovascular diseases, the possible beneficial effects of CVB-D on DCM remained unknown. The present aim was to explore the potential effects and underlying mechanisms of CVB-D on DCM. We explored the effects of CVB-D in DCM by using high fat high sucrose diet and streptozotocin-induced rat DCM model. Cardiac function and survival in rats with DCM were improved via the amelioration of oxidative damage after CVB-D treatment. Our data also demonstrated that pre-treatment with CVB-D exerted a remarkable cytoprotective effect against high glucose -or H2O2 -induced neonatal rat cardiomyocyte damage via the suppression of reactive oxygen species accumulation and restoration of mitochondrial membrane potential; this effect was associated with promotion of Nrf2 nuclear translocation and its downstream antioxidative stress signals (NQO-1, Prdx1). Overall, the present data has provided the first evidence that CVB-D has potential therapeutic in DCM, mainly by activation of the Nrf2 signalling pathway to suppress oxidative stress. Our findings also have positive implications on the novel promising clinical applications of CVB-D.


CVB-D improves cardiac function in DCM rats.
The type 2 diabetes rat model was reproduced by the injection of small doses of STZ and feeding high fat high sucrose(HFS)-fodder. As shown in Fig. 1c, the rats showed insulin resistance after HFS feeding for 12 weeks. The levels of FBG were remarkably increased in the model group compared with the control group (Fig. 1d). Cardiac function parameters were measured by electrocardiograph and plasma BNP levels, and our data showed that the ejection fraction (EF) (Fig. 1e,f) and LVFS (Fig. S1a) of diabetic rats showed a downward trend, while BNP levels gradually increased (Fig. 1g). Administrating with CVB-D (2 mg/kg/day) for 12 weeks could ameliorate these pathological changes. The LDH1 and CK-MB activity as biomarkers in serum were evaluated cardiac injury, which was obviously increased in DCM, and was significantly inhibited treated with CVB-D ( Fig. 1e-i). CVB-D significantly improved the survival rate of DCM rats (Fig. 1j). Compared with the control, the LV mass was increased in DCM, CVB-D has no significant effect on this change ( Supplementary Fig. S1b) similar to LVPWd and LVPWs (Fig. S1c,d). CVB-D did not alleviate the FBG levels ( Fig. 1d), which indicated CVB-D ameliorated cardiac function independent on blood glucose control. Collectively, these data suggested that the diabetic group had developed DCM, whereas CVB-D treatment alleviated cardiac dysfunction.

CVB-D ameliorates oxidative stress in DCM heart.
It is well known that excess oxidative stress is key pathophysiological process in cell death. The HE staining showed that DCM exhibited focal necrosis and infiltration of chronic inflammatory cells (black arrow), CVB-D significantly attenuated pathological changes (Fig. 2a). Mitochondria, as the main origin and attacking target of ROS, play a key role in energy metabolism. TEM revealed significant swelling and deformed mitochondria in DCM (green arrow), and CVB-D ameliorated the morphological abnormalities in the mitochondria (Fig. 2b).The mitoQ, a mitochondria-targeted antioxidant agent, could inhibit the mitochondria pathological damage in the DCM. 8-OHdG (red arrow, Fig. 2c,d), the biomarker of DNA oxidative damage, was significantly increase in DCM and attenuated by CVB-D. The SOD activity was significant decrease, on the contrary, increase MDA contents in DCM. Treated with CVB-D could alleviate (Fig. 2e,f).
Nrf2 plays a vital role in antioxidative responses through the up-regulation of multiple antioxidant components, such as NQO-1 and Prdx1. Western blotting results indicated that the protein expression of Nrf2, NQO-1, and Prdx1 were significantly decreased in DCM, however, treated with CVB-D increased Nrf2, NQO-1, and Prdx1 protein expression (Fig. 2g,h). These results were consistent with the immunofluorescence staining results (Fig. 2I,j). Taken together, CVB-D could ameliorate diabetes-induced oxidative stress which maybe involve in activating the Nrf2-related antioxidant pathways in DCM.

CVB-D attenuates HG-induced oxidative damage in PNRCMs.
To investigate the potential mechanism of CVB-D-mediated antioxidant stress in vitro, the PNRCMs were exposed to 40 mM glucose (high-glucose,HG) medium which induced oxidative stress damage. PNRCMs were isolated according to a previously published protocol 26 and the cardiomyocytes were purified by using a differential attachment methods and further identified by cardiac troponin T(cTnT) immunostaining (Fig. 3b). MTT results suggest that 25 mM glucose DMEM is beneficial to the survival of PNRCMs ( Supplementary Fig. S2). Preincubated with CVB-D (0.5 µM) could ameliorated the cardiomyocytes viability induced by HG (Fig. 3c, Supplementary Fig. S3). The osmotic pressure control, 40 mM mannitol, did not affect the cell viability. CVB-D (0.2 and 0.5 µM) inhibited ROS generation (Fig. 3d,e) and partially reversed the decreasing mitochondrial membrane potential (Fig. 3h,i). CVB-D (0.5 µM) inhibited the cardiomyocytes hypertrophy (black arrow) induced by HG by Giemsa staining (Fig. 3f,g). Exposure the PNRCMs to high glucose, the protein expression of Nrf2, and downstream protein NQO-1 and Prdx1 were significantly down-regulation by western blotting. Preincubation with CVB-D, the protein level was restored in certain extent, suggesting that CVB-D activated Nrf2-related antioxidant signalling (Fig. 3j,k). Above all, CVB-D could ameliorate HG-induced oxidative stress via the activation of the Nrf2-related antioxidant pathway in PNRCMs.

CVB-D attenuates HG-induced oxidative stress in PNRCMs via Nrf2 regulation.
To investigate the direct role of Nrf2 in CVB-D-mediated cardiac protection against HG, PNRCMs were pre-incubated with a pharmacological Nrf2 inhibitor (ML385) or activator (Bardoxolone) in with or without CVB-D 27,28 . The cell viability and the protein expression of Nrf2, NQO-1, and Prdx1 were determined. The protection effect of www.nature.com/scientificreports www.nature.com/scientificreports/ HG + Bardoxolone + CVB-D groups (Fig. 4d-f), suggesting that CVB-D protected against PNRCMs via the activation of Nrf2.
To verify the direct role of Nrf2 in mediating CVB-D at the genetic level, Nrf2 shRNA and overexpression plasmid adenoviruses were transfected into the PNRCMs. The results indicated that silencing of the Nrf2 gene abolished CVB-D-induced activation of Nrf2 and its downstream target proteins (Fig. 4g-i). The overexpression   www.nature.com/scientificreports www.nature.com/scientificreports/ CVB-D promotes Nrf2 nuclear translocation. The expression and activity of Nrf2 are regulated by its cullin3-based ubiquitin E3 ligase protein Keap1. Upon exposure to various stress conditions, Nrf2 is uncoupled from Keap1 and subsequently translocated to the nucleus, where it activates its downstream signals. To end this, we explore that CVB-D can promote the nuclear translocation of Nrf2 with immunofluorescence assay for Nrf2 and western blotting analysis of cytosolic and nuclear extracts were performed. The immunofluorescence assay showed that Nrf2 proteins predominantly accumulated around the nucleus after exposure to HG and treatment with CVB-D was closely related with the promotion on the nuclear translocation of Nrf2 (Fig. 5a). The results were further verified by the western blotting ( Fig. 5b-e).
As the ubiquitinated ligase of Nrf2 protein, the expression of keap1 is closely related to the nuclear translocation of Nrf2. As shown in Fig. 5f,g, unlike in the control group, the protein expression of Keap1 was increased after exposure to 40 mM glucose and was attenuated by the addition of CVB-D, suggesting that the nuclear translocation of Nrf2 was at least partly dependent on the Keap1 protein expression. Interestingly, the in vivo data is consistent with the results of in vitro experiments ( Supplementary Fig. S4).
To further clarify the interaction between CVB-D and the Nrf2-Keap1 complex, we performed a docking study using AutoDock Vina 1.1.2, and calculated their binding free energy by using the MM/GBSA method. In Fig. 5h-j, CVB-D was located in the Nrf2-Keap1 binding pocket, in which CVB-D forms one double hydrogens bonding with residue GLY-367 and VAL-606.MMGBSA showed that the binding free energy between Nrf2 and keap1 was −66.4 kcal/mol, and the main contribution was from electrostatic and vdW interactions. In addition, polar solvation was unfavourable for this binding. However, when CVB-D was present, the binding free energy www.nature.com/scientificreports www.nature.com/scientificreports/ between Nrf2 and Keap1 was changed to −56.7 kcal/mol, almost lower 10 cal/mol than in the absence of CVB-D (Supplementary Material, Table 1). This observation suggested that CVB-D could suppress the binding between Nrf2 and Keap1, which is an important factor for Nrf2 nuclear translocation. These results demonstrated the potential action of CVB-D as an enhancer of Nrf2 nuclear translocation, dependent on the down-regulation of the protein expression of keap1, and the reduced binding free energy of the Nrf2-Keap1 complex.

CVB-D attenuates H 2 o 2 -induced oxidative stress damage in PNRCMs.
To further explore the mechanism of CVB-D action, PNRCMs was exposed to H 2 O 2 , a common method to analyse oxidative stress 29,30 . The MTT assay results indicated that CVB-D significantly attenuated H 2 O 2 (100 µM, 24 h) induced toxicity to cardiomyocytes (Fig. 6a), and that CVB-D also inhibited ROS generation (Fig. 6b,c) similar to the HG model. Unlike cells exposed to HG, hydrogen peroxide causes vacuolar degeneration of PNRCMs (Fig. 6d, black arrow). It was found that the trends in Nrf2, NQO-1, and Prdx1 were consistent with the trends of HG model (Fig. 6e,f). These results were further confirming the antioxidant effect of CVB-D. (e,f) Protein expression of Nrf2, NQO-1, and Prdx-1 was assayed by western blotting, full-length blots are presented in Supplementary Fig. S12. The data are presented as the mean ± SEM (n = 3 in each group),*p < 0.05 versus the control group; **p < 0.05 versus the model group.

Discussion
DCM is a special form of heart disease that was first described in 1972 and remains the principal cause of death in patients with diabetes 31 . As the most common complication of diabetes, despite the application of conventional treatment strategies, such as hypoglycaemic, lipid-lowering, and anti-hypertensive agents, the incidence of DCM in patients with diabetic is still higher 4,32 . Therefore, a novel effective therapeutic agent is urgently required. CVB-D has been demonstrated multiple pharmacological activities and widely applied in cardiovascular diseases [22][23][24][25] . However, the possible beneficial effects and potential mechanism of CVB-D in T 2 DM-associated cardiomyopathy remains unknown. Our results has provided the first demonstration that CVB-D could effectively ameliorates cardiac dysfunction and increases survival in rats with DCM, and was associated with the activation of Nrf2 and its associated signalling targets.
DCM is usually characterized by cardiac structural and functional disorders. Some risk factors, such as oxidative stress and inflammation, promote cardiomyocyte death, cardiac tissue interstitial fibrosis, and cardiac stiffness, leading to diastolic and systolic dysfunction, and, eventually heart failure 2,33,34 . However, the exact pathological process of DCM is still unclear and there are no clear diagnostic indicators 35 . At present study, we evaluated the EF of rats by non-invasive electrocardiograph and, simultaneously, plasma BNP levels to confirm the occurrence of heart failure. After 12 weeks of elevated blood glucose, the ejection fraction (EF = 65.4% ± 1.8%) of diabetic and the BNP level (153.2 ± 15 pg/mL) are significantly decreased and increased compared to control group, indicating that the heart failure model was successfully established 36 . Meanwhile, the results of TEM revealed significant mitochondrial disruption in DCM; MDA and a biomarker of DNA damage (8-OHdG) were significantly increased, and SOD activity was decreased in DCM hearts. All findings were consistent with those of previous studies 37,38 suggesting that oxidative stress is one of the key important pathophysiological factor in DCM. The CVB-D can ameliorate oxidative stress damage in DCM, which suggests potential beneficial effect of CVB-D on DCM via ameliorating oxidative stress effect.
As a major transcription factor of cellular antioxidant response, accumulated evidence has suggested that activation of Nrf2 may have potential therapeutic in multiple cardiovascular diseases, such as heart failure, diabetes, diabetic nephropathy, and DCM [39][40][41][42] . At present, diabetes caused a significant down-regulation of Nrf2 and downstream molecules such as NQO-1 and Prdx-1, and CVB-D alleviated the expression of these proteins both in vivo and in vitro. To further explore whether the protective efficacy of CVB-D was related to Nrf2, the inhibitor/activator or sh-RNA/overexpression plasmids for Nrf2 were highlighted on HG challenged to PNRCMs and the results indicated that CVB-D protected PNRCMs via the activation of Nrf2. As well known, Nrf2 activators promote the expression of Nrf2 by disrupting the interaction between Keap1 and Nrf2 43,44 . Our results confirmed that CVB-D increased Nrf2 expression in the nucleus and reduced the protein expression of Keap1 both in vivo and in vitro, and molecular docking revealed that the binding free energy between Nrf2 and Keap1 was reduced by approximately 15% in the presence of CVB-D. Therefore, we concluded the CVB-D-mediated increase in Nrf2 expression via the inhibition of Keap1 expression and the decrease in binding free energy of the Nrf2-Keap1 complex, which represented a novel method that CVB-D promotes the uncoupling mechanism of Nrf2 and Keap1.
Hyperglycaemia is one of the causes of oxidative stress. To further clarify the antioxidant capacity of CVB-D, a classical oxidative stress model must be tested. It is well known that H 2 O 2 is a potent inducer of oxidative stress and an ideal model for studying the antioxidant effects of CVB-D 45,46 . Unlike cardiac hypertrophy induced by hyperglycaemia, H 2 O 2 leads to vacuolar degeneration of cardiomyocytes, and produces higher levels of ROS. Surprisingly, CVB-D demonstrated favourable anti-oxidant stress effect in both HG and H 2 O 2 models.
In summary, the present data is the first time that CVB-D could ameliorate the oxidative stress in DCM both in vivo and in vitro. CVB-D can inhibit the expression of keap1 and reduce the free binding energy of the Nrf2-Keap1 complex, promote Nrf2 translocating into the nucleus, activate its downstream signalling targets (NQO-1, and prdx1), and exert attenuating oxidative stress effects (Fig. 7). Our findings also have positive implications on the novel promising clinical applications of CVB-D.
Methods chemicals and reagents. CVB-D (purity >98%; the chemical structure is shown in Fig. 1a  www.nature.com/scientificreports www.nature.com/scientificreports/ with the NIH Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996). A total of 120, health female Sprague-Dawley (SD) rats (8-week-old, body weight 180-200 g) were obtained from the experimental animal center of Guizhou Medical University. The rats were divided into two groups as following: the control group (n = 24, normal diet) and the HFS group (n = 96), HFS group was fed daily with HFS to induce insulin resistance as previously described 48 , Control group fed with a standard chow diet containing 16% protein, 4% fat and 60% carbohydrate. All rats were maintained and fed ad libitum under a controlled 12/12 hour light/dark cycle and temperature (22 ± 2 °C) for 24 weeks. After 12 weeks of feeding, the serum samples of rats from each group were taken from the tail vein. Fasting blood glucose (FBG) and fasting insulin (INS) levels were analysed to evaluate insulin resistance. Insulin resistance (HOMA-IR) was calculated as follows: HOMA-IR = (FBG*INS)/22.5. Then, Model rats that have developed insulin resistance were treated with streptozotocin by a single intraperitoneal injection of 30 mg/kg in 10 mg/ml sodium citrate buffer (pH 4.5), Control group was injected with the same volume of sodium citrate buffer.After STZ injection for 3 days, diabetic model was considered successful as the FBG level was more than 11.1 mM. Then, the diabetic rats were randomly divided into four groups as following: DCM; DCM + CVB-D.H (2 mg/kg/day, ig); DCM + CVB-D.L (1 mg/kg/day, ig); DCM + mitoQ (reference drug, 5 mg/kg twice weekly, ip), the control group and DCM group were administered saline for the next 12 weeks (Fig. 1b), and plasma was taken from the rat tail vein to measure the FBG and BNP levels every month. The creatine kinase isoenzyme MB (CK-MB) and Lactic dehydrogenase 1 (LDH1) were detected in serum by commercial kits in accordance with the manual instruction at the end of experiment.
Cardiac function evaluation. Cardiac ultrasonography and serum BNP levels were assessed monthly to confirm the successful establishment of DCM 49 . The rats were anaesthetised with inhalation of sevoflurane. The left ventricular ejection fraction (EF), left ventricular fractional shortening (LVFS), left ventricular mass(LVmass), left ventricular posterior wall thickness diastole (LVPWd) and left ventricular posterior wall thickness systole (LVPWs) were determined and analysed M-mode echocardiography with a 30 MHz linear transducer using a VINNO echocardiography system (SUZHOU, CHINA) which was evaluated the cardiac function. The DCM model was successfully reproduced by the low EF and high BNP levels compared to control group.
Morphological analysis of cardiac tissue. The cardiac tissue samples were fixed in 4% buffered paraformaldehyde, cut to a thickness of 5 μm, stained with H&E, and imaged by using a Nikon microscope (Nikon, Japan). To observe the ultrastructure of mitochondria in the cardiomyocytes, the left ventricular sample was cut into 1 mm blocks, fixed with 4% paraformaldehyde solution containing 2.5% glutaraldehyde, and then imaged by using an electron microscope (JEM-1001, JEOL Ltd., Tokyo, Japan).
Measurement of SOD activity and MDA content. Total proteins were extracted from cardiac tissues lysed in lysis buffer. Protein concentration was meansured by Bicinchoninic acid method. The SOD activity and MDA contents were measured by commercial kits in accordance with the manufacturer's method.
IHC/IF. The 8-OHdG, Nrf2, NQO-1, Prdx1 were evaluated by immunohistochemical (IHC) or Immunofluorescence (IF) staining of heart tissue. The positive areas were detected by using ImageJ (U.S. National Institutes of Health, Bethesda, MD). cell culture. The primary neonatal rat cardiomyocyte culture. The primary neonatal rat cardiomyocytes (PNRCMs) were separated and purified from the ventricles of new-born 1-2-day-old SD rats, as previously described 26 . PNRCMs were cultured in 25 mM glucose DMEM (containing 10% FBS) in an incubator (5% CO 2 , 37 °C) for 72 h. The cells were identified as cardiomyocytes by immunofluorescence staining of cardiac troponin (cTnT). . PNRCMs were pre-treated with CVB-D or mitoQ for 2 h, and then cultured with 40 mM glucose for an additional 48 h (Fig. 3a). Equal concentration of mannitol were evaluated the effect of osmotic pressure on cell viability.
To investigate the ameliorating oxidative stress effect of CVB-D, H 2 O 2 -induced cytotoxicity in PNRCMs was established, and the groups were design as following: (1)  The role of Nrf2 was investigated in the protection of CVB-D on cardiomyocytes challenged to HG, The cardiomyocytes were incubated with or without agonists/inhibitors of Nrf2. Adenoviruses expressing short hairpin (sh) RNA or Nrf2-overexpressing plasmids were further explore the molecular mechanism of CVB-D.
The measurement of cell viability. Cell viability was analysed using MTT assay. The PNRCMs were seeded into 96-well plates at 5000 cells/well and cultured for 72 h. At the designing experimental time point, the MTT solution (5 mg/mL, 20 µL/well) was added to the well and then incubated for 4 h. The absorbance at 490 nm was measured with Varioskan Flash (Thermo fisher Scientific, Finland).
Giemsa staining. After indicated incubation, Giemsa staining and cell morphology was examined under an inversion microscope (Leica DMi8, Germany). The cell area of each group was measured by the software of the Image J Software, each group randomly selected 50 cells and calculated the average value.
Measurement of ROS. The production of ROS was measured by detecting the fluorescent intensity of oxidant-sensitive probe DCFH-DA, which was a stable nonfluorescent molecule that passively diffuses into cells, where the acetate could be cleaved by intracellular esterase to produce a polar diol that was well retained within in the cells. After the cell treatment with the designing protocol, which was loaded with DCFH-DA (10 μM) as per the manufacturer's instruction for 30 min without light. The fluorescent intensity was recorded by excitation at 488 nm and emission at 525 nm using a Varioskan Flash (Thermo fisher Scientific, Finland). Images were observed under a fluorescence microscope (Leica DMi8, Germany). Immunofluorescence analysis. Immunofluorescence staining of PNRCMs with anti-Nrf2 antibody (1:100 dilutions) was performed to observe the nuclear translocations of Nrf2. Images of cells were captured by using a confocal microscope (ZEISS, LSM710).

Molecular docking analysis.
The 3D-structure of the Nrf2-Keap1 complex (PDB ID: 2flu) was download from the PDB database. AutoDockTool 1.1.2 was used to determine the atom types and calculate the atomic charges of the receptor protein, and a.pdbqt file was saved for the docking analysis. The 2D structure of the ligand (CVB-D) was drawn by ChemDraw, and saved as a.cdx file. Then, the MM2 force field in Chem3D was used to optimise the 3D structure of the ligand. AutoDockTool was used to determine the atom types and calculate the partial charges of the ligand, and a.pdbqt file was saved for the docking analysis. Docking was simulated by using AutoDock Vina 1.1.2.
The Molecular Mechanics Generalized Born Surface Area (MM/GBSA) module was used to analyse the change in binding energy of the Nrf2-Keap1 complex before and after CVB-D insertion. Pre-treatment of Nrf2-Keap1 crystals involved the removal of crystal water, addition of protein force fields, addition of antagonist ions, addition of solvation cartridges, and preferences. Subsequently, a 50 ns molecular dynamics simulation was performed on the complex and the binding free energy of the Nrf2-keap1 complex with the CVB-D-Nrf2-Keap1-ligand ternary complex was calculated. Statistical analysis. Data from a minimum of at least three independent experiments were calculated and expressed as the mean ± SEM. Statistical analysis of the data was evaluated using one-way ANOVA with Tukey post hoc test by using Prism 6.0 (GraphPad Software, San Diego,CA), Kaplan-Meier survival curves and Mantel-Cox log rank test were used to analysis the survival rates for rats, A p value < 0.05 was considered statistically significant.

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.