Gypenosides improve diabetic cardiomyopathy by inhibiting ROS‐mediated NLRP3 inflammasome activation

Abstract NLRP3 inflammasome activation plays an important role in diabetic cardiomyopathy (DCM), which may relate to excessive production of reactive oxygen species (ROS). Gypenosides (Gps), the major ingredients of Gynostemma pentaphylla (Thunb.) Makino, have exerted the properties of anti‐hyperglycaemia and anti‐inflammation, but whether Gps improve myocardial damage and the mechanism remains unclear. Here, we found that high glucose (HG) induced myocardial damage by activating the NLRP3 inflammasome and then promoting IL‐1β and IL‐18 secretion in H9C2 cells and NRVMs. Meanwhile, HG elevated the production of ROS, which was vital to NLRP3 inflammasome activation. Moreover, the ROS activated the NLRP3 inflammasome mainly by cytochrome c influx into the cytoplasm and binding to NLRP3. Inhibition of ROS and cytochrome c dramatically down‐regulated NLRP3 inflammasome activation and improved the cardiomyocyte damage induced by HG, which was also detected in cells treated by Gps. Furthermore, Gps also reduced the levels of the C‐reactive proteins (CRPs), IL‐1β and IL‐18, inhibited NLRP3 inflammasome activation and consequently improved myocardial damage in vivo. These findings provide a mechanism that ROS induced by HG activates the NLRP3 inflammasome by cytochrome c binding to NLRP3 and that Gps may be potential and effective drugs for DCM via the inhibition of ROS‐mediated NLRP3 inflammasome activation.

NLRP3 is best known for sensing the widest array of stimuli, such as nigericin, ATP, ROS, MSU and cholesterol. [10][11][12] When the host receives a signal stimulus, NLRP3 itself can undergo oligomerization, increasing pro-caspase-1 through the adapter protein ASC, which activates caspase-1 and then leads to the cleavage of the interleukin-1b precursor (pro-interleukin-1b) and the interleukin-18 precursor (prointerleukin-18), causing corresponding inflammation factor IL-1b and IL-18 generation, thus inducing inflammation and necrosis (pyroptosis) and participating in the inflammatory response of cell and tissue. 7,13,14 Although the inflammasome plays an important role in the innate immune response and anti-infection immune responses, inflammasome overactivation was shown to be involved in some chronic inflammatory diseases, such as Alzheimer's, ulcerative colitis, atherosclerosis and type 2 diabetes and its complications. [15][16][17][18][19] Hence, inhibiting inflammasome activation may be a promising strategy for DCM. In rats with type 2 diabetes, the NLRP3 inflammasome is up-regulated and promotes the development of DCM. [20][21][22] Inhibiting the activation of the NLRP3 inflammasome can suppress the release of proinflammatory cytokines and attenuate the development of DCM. 23,24 In addition, NLRP3 gene silencing alleviates cardiac inflammation and exerts a protective effect on DCM. 25 However, the molecular mechanisms by which high glucose (HG) activates the NLRP3 inflammasome in cardiomyocytes remain to be further explored.
In this study, we utilized HG-stimulated H9C2 and high-fat-diet/ STZ-fed rats to study the mechanisms of NLRP3 inflammasome activation. Our study identifies that HG induces mitochondria damage and results in the overproduction of reactive oxygen species (ROS) and the release of cytochrome c, which lead to NLRP3 inflammasome activation and cell death. We further find that ROS activates the NLRP3 inflammasome, which is dependent on cytochrome c in H9C2 treated with HG. Besides these, HG-induced NLRP3 inflammasome activation is inhibited by Gps, which reduces IL-1b release and improves DCM.
Consequently, Gps may be a potential therapeutic strategy for DCM.

| Cell culture and treatment
H9C2 cells were purchased from the Library of Typical Culture of the Chinese Academy of Sciences (Shanghai, China). The cells were cultured in a DMEM medium supplemented with D-glucose (5.5 mmol/L), 10% FBS, penicillin (100 U/mL) and streptomycin (100 mg/mL). Moreover, neonatal rat ventricular myocytes (NRVMs) were obtained and cultured as previously described. 31 In HG-cultured groups, the medium contained 25 or 35 mmol/L of D-glucose.
Z-YYAD-FMK, NAC, and cyclosporin A were obtained from Abcam Trading (Shanghai) Company Ltd. Gps were purchased from Yuanye Biology Company (Shanghai) and dissolved in PBS for in vitro and in vivo experiments.

| Rats
SD rats (7 weeks of age) were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd (Beijing, China). All animal studies and experiments were permitted by the Animal Experimental Ethics Committee of Henan University. Male rats were bred in a specific pathogen-free facility. After being fed a basal diet for 1 week, rats were randomly divided into chow diet (n = 6) and highfat diet (n = 30) groups. Four weeks later, 35 mg/kg of streptozotocin (STZ, Solarbio, China) was administrated intraperitoneally. One week later, rats fasted for 10 hours and then blood glucose levels were measured by a glucometer (Roche, Germany). The rats whose blood glucose exceeded 12 mmol/L had diabetes and were used for the following study. Then, the diabetic rats were respectively treated with Z-YVAD-FMK (caspase-1 inhibitor, intravenous, 60 mg/kg, n = 6), Gps (intragastric, 200 mg/kg, n = 6), or vehicle (PBS, n = 6) once a day for 8 weeks. Once a week, blood glucose and body weight were detected and recorded. Finally, all rats were killed under anaesthesia. Blood was collected from the carotid artery and centrifuged for 10 minutes at 1800 g at 4°C to obtain the serum. The hearts were rapidly frozen in liquid nitrogen to extract the related genes and proteins or embedded in 4% paraformaldehyde for pathological analysis after lavage.

| Western blot, reverse-transcription and realtime PCR
Western blot (WB) and real-time PCR (RT-PCR) were carried out as previously described. 34 Antibodies for NLRP3 and ASC were purchased from Santa Cruz Technology (Santa Cruz, CA). Antibodies for caspase-1, IL-1b, cytochrome c, VDAC, b-actin, GAPDH and the secondary HRP-conjugated antibody were obtained from Cell Signaling Technology (CST). Primers for ASC, NLRP3, caspase-1 and IL-1b genes were obtained from Sangon Biotech Company (Shanghai). The primer sequences are presented in Table S1.

| Immunofluorescence
Immunofluorescence was used to detect the expression of NLRP3 inflammasome markers as previously described. 35 The first antibodies contained anti-NLRP3, ASC, caspase-1 and IL-1b antibodies. An Alexa Fluor 488 goat anti-rabbit IgG (H+L) antibody and an Alexa Fluor 594 goat antimouse IgG (H+L) antibody were used as secondary antibodies. Fluorescence was detected with an inverse fluorescent microscope (Nikon, Japan).

| ELISA
Supernatants of the cell culture or serum were used to measure IL-1b, IL-18 and C-reactive protein (CRP) (Thermo Fisher Scientific) according to the manufacturer's instructions.

| Histological examination
After the rats were killed, their hearts were perfused with PBS and bisected at the mid-ventricular level. Then, the hearts were fixed in 4% paraformaldehyde and embedded in paraffin. Four micrometre sections were stained with haematoxylin and eosin (H&E), and images were randomly captured by microscope.

| Co-immunoprecipitation
Co-immunoprecipitation (CoIP) was carried out as previously described. 36 1 mg of cell lysate was treated with protein A/G agarose beads for 1 hour. After being centrifuged for 10 minutes, the supernatants were incubated with an anti-NLRP3 antibody overnight at 4°C. Then, 100 lL protein A/G agarose beads were added to the supernatants and mixed for 2 hours. The beads were centrifuged and washed five times with a lysis buffer. The samples were subjected to immunoblotting with the anti-cytochrome c antibody and corresponding secondary antibodies.

| Statistical analysis
All data were shown as mean AE SD. Samples were analysed using a one-way ANOVA on SPSS 16.0 software. P < .05 was considered a statistically significant difference.

| HG-induced cell injury in cardiomyocytes
To investigate the effect of HG on H9C2 cells, the cells were treated with different concentrations of glucose for 6, 12, 24, 36 and 48 hours. The supernatants were measured using commercial LDH kits, while the residual cells were used for an MTS assay. Although 25 and 35 mmol/L of glucose could induce H9C2 to release more LDH than 5.5 mmol/L of glucose, there are no statistical differences within a short period of time (6 and 12 hours) ( Figure 1A). After the H9C2 cells were treated at 24, 36 and 48 hours, we found that HG dramatically resulted in cell damage in a dose-dependent manner ( Figure 1A). The MTS assay also showed that HG reduced myocardial viability in the same manner ( Figure 1B). In addition to the H9C2 cells, NRVMs treated with HG simultaneously showed cell damage ( Figure S1A,B). Meanwhile, HG signally induced H9C2 cell apoptosis, and this effect was dose-dependent ( Figure 1C-E). Cell apoptosis in response to HG was confirmed using NRVMs ( Figure S1C,D).
To examine whether osmotic pressure affected cardiomyocyte viability, the H9C2 cells were treated with equal concentrations of mannitol. We found that the cell damage and activity were not significantly changed, which suggested that the cell injury depended on glucose itself, not osmotic pressure ( Figure 1A,B).

| High glucose-induced cardiocyte injury mainly depended on NLRP3 inflammasome activation
To determine the role of NLRP3 inflammasome activation on HG- To determine whether HG-induced NLRP3 inflammasome activation is responsible for cardiocyte injury, the effects of HG on NLRP3 inflammasome activation and cardiocyte injury were assessed after NLRP3 was silenced in H9C2 cells. As shown in Figure 3A, the level of ZHANG ET AL.

| 4439
NLRP3 protein was obviously reduced in H9C2 cells treated with siRNA, even under a HG condition. Meanwhile, NLRP3 deficiency blocked HG-induced NLRP3 inflammasome activation, caspase-1 p20 production and mature IL-1b secretion ( Figure 3B). Additionally, an ELISA assay for IL-1b in supernatants showed that IL-1b diminished in the absence of NLRP3 ( Figure 3C). Most of all, the cell injury induced by HG improved when NLRP3 was silenced ( Figure 3D,E). Correspondingly, the effects of HG on NLRP3 inflammasome activation and cardiocyte injury were confirmed using the caspase inhibitor Z-YVAD-FMK ( Figure 3F-I). Thus, these data indicated that HG-induced NLRP3 inflammasome activation was responsible for H9C2 cell injury.

| High glucose-induced ROS release accounted for NLRP3 inflammasome activation
To ascertain the mechanisms of NLRP3 inflammasome activation, we examined ROS release, a common stimulator associated with inflammasome activation. In agreement with previous reports, stimulation was done with HG-induced cytochrome c influx into cytoplasm, which meant the mitochondrion was damaged ( Figure 4A). Correspondingly, HG promoted ROS generation, and this increase was dose-dependent ( Figure 4B).
Next, we examined the impact of ROS on NLRP3 inflammasome activation. As shown in Figure 4C, stimulation with NAC reduced the generation of ROS in H9C2, which was also done in a dose-dependent manner. Additionally, the level of NLRP3 was lower in groups treated with NAC than in the control group ( Figure 4D). The expressions of mIL-1b and cleaved caspase-1 (P20) were also downregulated in NAC-treated groups than in the DMSO-treated group ( Figure 4D). The cell damage and viability of H9C2 were obviously improved with the pretreatment of NAC compared with DMSO (Figure 4E,F). Thus, these data indicate that ROS release accounted for NLRP3 inflammasome activation induced by HG.

NLRP3 inflammasome activation
Although emerging data showed that ROS can induce NLRP3 inflammasome activation, the exact mechanisms were still uncertain. We next found that NAC not only inhibited ROS release, but also improved cytochrome c influx into cytoplasm ( Figure 5A). To identify whether cytochrome c regulated NLRP3 inflammasome activation, cyclosporine A was used to stimulate H9C2 cells along with HG. Cyclosporine A, as an inhibitor of cytochrome c, could effectively inhibit cytochrome c influx into cytoplasm, which was also proved by our data ( Figure 5B). Furthermore, cyclosporine A could inhibit the expression of NLRP3, which was increased by HG ( Figure 5C). The protein levels of mIL-1b and P20 were also reduced by cyclosporine A (Figure 5D). The cell damage and viability of H9C2 were obviously improved with the pretreatment of cyclosporine A compared with DMSO ( Figure 5D,E). To detect the mechanisms by which cytochrome c activated NLRP3 inflammasome, CoIP was used. The data showed that cytochrome c could bind to NLRP3, even in the presence of cyclosporine A ( Figure 5F). Collectively, these data indicated that cytochrome c could bind to NLRP3, which mediated the ROS-induced NLRP3 inflammasome activation.

| Gps improved HG-induced cardiomyocyte damage
To investigate the effect of Gps on HG-induced cardiomyocyte injury, H9C2 cells were treated with Gps for 48 hours along with 35 mmol/L glucose. We found that Gps noticeably inhibited LDH release and improved cell viability in a dose-dependent manner (Figure 6A,B). Alternatively, cardiocyte injury could be confirmed by cell apoptosis via FCM. In agreement with Figure 6A,B, HG-induced cell apoptosis was dose-dependently inhibited by Gps ( Figure 6C). The mean percent of apoptosis cells was 17.18% in the HG-treated group, compared to 5.46%, 5.88% and 5.41% in several concentrations of the Gps-treated groups ( Figure 6D).

| Gps suppressed NLRP3 inflammasome activation by inhibiting ROS release
To determine whether Gps's improvement of cardiocyte injury relied on regulating NLRP3 inflammasome activation, the effects of Gps on NLRP3 inflammasome activation were assessed. Gps dramatically inhibited the secretion of IL-1b and cleaved caspase-1 (p20), which were induced by 35 mmol/L of glucose ( Figure 7A). Meanwhile, the production of NLRP3 was also dose-dependently inhibited by Gps ( Figure 7A). Corresponding to the immunoblotting results, the levels of IL-1b and IL-18 proteins in medium supernatants of H9C2 cells were dramatically reduced by Gps via ELISA ( Figure 7B,C). However, in H9C2 cells treated with 5.5 mmol/L of glucose, Gps did not influence the production of NLRP3 and ASC ( Figure S4A-C). The secretion of IL-1b and p20 had also not changed ( Figure S4A,D and E).
Meanwhile, the ELISA results showed that Gps made no difference in the secretion of IL-1b and IL-18 ( Figure S4F,G).
To determine the mechanisms by which Gps inhibited NLRP3 inflammasome activation, we next examined ROS release. As shown in Figures 4B and 7D, stimulation with HG-induced ROS release in  Figure 7F). Collectively, the above data indicate that Gps inhibited ROS release and cytochrome c influx into cytoplasm, and then regulated NLRP3 inflammasome activation.  Figure 8A,B). Compared to normal mice, mice fed with HFD and then administered with STZ obviously had myocardial injury, whereas the caspase-1 inhibitor (Z-YVAD-FMK) and Gps significantly alleviated the induction and injury by HFD ( Figure 8C). As an important marker of activation, the level of CRP was markedly higher in mice treated with HFD and STZ than in normal mice. Z-YVAD-FMK and Gps significantly reduced CRP production ( Figure 8D). These data suggest that Gps attenuated disease development by reducing blood glucose and inhibiting inflammatory activation in vivo.

| In vivo, gypenosides prevented NLRP3 inflammasome activation and attenuated HG-induced myocardial injury
To evaluate the role of Gps on NLRP3 inflammasome activation in vivo, the total heart protein in these groups was used to detect NLPR3 inflammasome activation. The levels of NLRP3 and ASC were significantly higher in mice treated with HFD and STZ than in normal mice ( Figure 8E). Moreover, IL-1b secretion and P20 production were also dramatically higher ( Figure 8E). Z-YVAD-FMK and Gps significantly reduced NLRP3, ASC and P20 production and IL-1b secretion ( Figure 8E). Consistent with the WB results, Gps and Z-YVAD-FMK inhibited the production of mature IL-1b and IL-18 in the serum of mice ( Figure 8F,G). Meanwhile, the mRNA levels of NLRP3, ASC, caspase-1 and IL-1b were notably reduced by Gps and Z-YVAD-FMK ( Figure 8H-K). In a word, Gps prevented NLRP3 inflammasome activation in vivo and attenuated HG-induced myocardial injury.

| DISCUSSION
In this study, we aimed to explore the mechanism of HG-induced myocardial damage and the effect of Gps in HG-induced myocardial damage. As shown in Figure S5, our work suggests that HG could induce remarkable apoptosis in H9C2 cells, which relied on NLRP3 inflammasome activation and the release of IL-1b. Moreover, NLRP3 inflammasome activation induced by HG was dependent on mitochondrial cytochrome c influx into cytoplasm, which was regulated by ROS. In addition, we observed that Gps improved HG-induced myocardial damage by inhibiting ROS-dependent NLRP3 inflammasome activation in H9C2 cells and in mice suffering from DCM.
Diabetic cardiomyopathy has become a worldwide health problem, and its primary feature is progressive damage and cardiomyocyte death. 37 The occurrence of DCM is covert, and the mechanisms remain unclear. Recently, many studies have shown that chronic inflammation and NLRP3 inflammasome activation caused by  14,18,24 The levels of CCL2, IL-1b and other proinflammatory cytokines are dramatically reduced in NLRP3-deficient mice. 38 Moreover, inhibiting NLRP3 activation or NLRP3 gene silencing exerts a protective role on DCM. 25,39 Interestingly, some researchers declare that NLRP3 inflammasome activation and subsequent IL-1b secretion in the heart mainly exist in fibroblasts, endothelial cells and leucocytes, while cardiomyocytes cannot secrete IL-1b. However, there is considerable evidence that NLRP3 inflammasome activation and IL-1b secretion induced by HG exist in cardiomyocytes and cause myocardial damage, including in our results. 14,39,40 In this study, we also found that HG significantly induced NLRP3 inflammasome formation and Although these data show a close relationship between HGinduced myocardial damage and NLRP3 inflammasome activation, the mechanism by which HG activates NLRP3 inflammasome is still not clear. Recent studies suggest that mitochondrial ROS is required for NLRP3 inflammasome activation. [41][42][43] Moreover, many studies have indicated that mitochondrial damage and oxidative stress are important risk factors for DCM and that mitochondrial ROS is a pivotal link between oxidative stress and NLRP3 inflammasome activation. 44,45 Recently, a study has shown that monocytes and/or MDMs from T2D or DCM patients stimulated with DAMP release an elevated level of ROS. 18 Meanwhile, ROS inhibitors dramatically inhibit the secretion of mature IL-1b and IL-18. Consistent with these data, we found that H9C2 cells stimulated by HG produced elevated levels of ROS, which induced NLRP3 inflammasome activation and subsequent myocardial damage. NAC, a kind of ROS inhibitor, dramatically inhibited NLRP3 inflammasome activation and decreased the secretion of mature IL-1b.
The NLRP3 inflammasome activation contains two steps. 7,46 In the first step, which is also called the priming step, PAMP recognizes pathogen response receptors (PRRs) and then activates NF-jB, which leads to increased levels of NLRP3, IL-1b and IL-18 genes and proteins. In the second step, many stimuli (uric acid, cardiolipin, dsDNA and so on) induce the formation of the NLRP3 inflammasome and subsequently activate it to produce mature IL-1b and IL-18. Initially, ROS were supposed to be one key stimulator participating in the priming step of NLRP3 inflammasome activation, because they were required for NK-jB activation. 47 Indeed, we found that the inhibition of ROS by NAC could reduce the production of NLRP3 protein and mature IL-1b, while the levels of NLRP3, caspase-1 and IL-1b did not obviously change (data not shown). This means that ROS may be crucial to the second step of NLRP3 inflammasome activation. However, the mechanism by which ROS activate the NLRP3 inflammasome is still unclear. Some studies find that ROS can dissociate the thioredoxin-interacting protein (TXNIP) from the TXNIP-TXN complex, and free TXNIP induces NLRP3 inflammasome formation and activation. 46,48 Inhibiting TXNIP can effectively downregulate NLRP3 inflammasome activation and IL-1b secretion. 48 Moreover, one study has recently shown that ROS can directly induce the formation of the caspase-1-ASC complex and activate the NLRP3 inflammasome. 45 In the current study, we found that ROS could induce cytochrome c influx into cytoplasm, which directly bound to NLRP3 and led to NLRP3 inflammasome activation.
Cyclosporin A (a cytochrome c inhibitor) could significantly inhibit NLRP3 inflammasome activation and mature IL-1b secretion, while the ROS level was not influenced. Even so, the mechanisms by which ROS regulate cytochrome c influx into cytoplasm and cytochrome c activates NLRP3 inflammasome deserve further research.
Gps, the major component of Gynostemma pentaphylla (Thunb.) Makino, have shown the properties of hypoglycaemic action and anti-inflammation. In the current study, we found that Gps can inhibit NLRP3 inflammasome activation and IL-1b and IL-18 secretion, which might result from their ability to reduce the level of mitochondrial ROS. In vivo, Gps could reduce the level of blood glucose, inhibit the expression of NLRP3 protein and IL-1b production and improve myocardial damage. Meanwhile, the generic levels were also reduced in mice administrated with Gps, which suggested that Gps might regulate the priming step of the NLRP3 inflammasome.
In summary, our data demonstrated that Gps inhibited HGinduced myocardial damage through the ROS-induced NLRP3 inflammasome activation in vitro and in vivo. Meanwhile, the ROS-activated NLRP3 inflammasome might result from cytochrome c binding to NLRP3. Therefore, Gps may be a potential therapeutic solution for inhibiting NLRP3 inflammasome activation and NLRP3-associated syndromes.

CONFLI CT OF INTEREST
The authors declare no conflict of interests.