Exendin-4 Protects against Hyperglycemia-Induced Cardiomyocyte Pyroptosis via the AMPK-TXNIP Pathway

Diabetic cardiomyopathy is a common cardiac condition in patients with diabetes mellitus, which results in cardiac hypertrophy and subsequent heart failure. Chronic inflammation in the diabetic heart results in loss of cardiomyocytes and subsequentially cardiac dysfunction. Accumulated evidence implicated pyroptosis as a vital contributor to the hyperglycemia-induced cardiac inflammatory response. Exendin-4, a GLP analog, promotes survival of cardiomyocytes in cardiovascular diseases, including diabetic cardiomyopathy. However, the role of Exendin-4 in cardiac pyroptosis remains to be elucidated. Our study revealed that Exendin-4 treatment protected against heart remolding and dysfunction and attenuated cardiac inflammation in high-fat diet-fed rats. The activity of caspase-1 and production of pyroptotic cytokines were significantly inhibited by Exendin-4 treatment in the diabetic heart and in high glucose-treated cardiomyocytes as well. In an effort to understand the signaling mechanisms underlying the antipyroptotic property of Exendin-4, we found that blockade of AMPK, an oxidative stress sensor, activity diminished the antipyroptotic property of Exendin-4. Phosphorylation of AMPK resulted in degeneration of TXNIP that promoted the activation of the NLRP3 inflammasome. Exendin-4 treatment decreased the protein level of TXNIP. Moreover, RNA silencing of TXNIP mimicked the antipyroptotic actions of Exendin-4. These findings promoted us to propose a new signaling pathway mediating cardioprotective effect of Exendin-4 under hyperglycemic conditions: Exendin-4 → ROS↓ → pAMPK↑ → TXNIP↓ → caspase-1↓ → IL-1β and IL-18↓ → pyroptosis↓. In general, our study identified Exendin-4 as a pyroptotic inhibitor protecting against hyperglycemia-induced cardiomyocyte pyroptosis via the AMPK-TXNIP pathway.


Introduction
Diabetes mellitus (DM) is a group of metabolic disorders characterized by hyperglycemia. Patients with diabetes suffer from cardiovascular diseases 2 to 4 times likely than individuals without diabetes [1]. Diabetic cardiomyopathy (DCM), the leading diabetic complication, is a critical cause of fatalities in chronic DM patients. DCM is defined by abnormal myocardial structure and cardiac function in the absence of coronary atherosclerosis and hypertension [2]. In DCM, chronic cardiac inflammation is characterized contributing to loss of cardiomyocytes that results in impaired systolic function [3,4]. However, the mechanism and medical treatment remain to be elucidated.
Glucagon-like peptide-1 (GLP-1) is identified with metabolic effects for glycemic control in type 2 DM [15]. Furthermore, GLP-1 is involved in cardiovascular physiology and protects cardiomyocytes against hyperglycemia-induced toxicity in DCM [16]. Exendin-4, a GLP analog, has a longer half-life than GLP-1 through evading clearance by dipeptidyl peptidase IV (DPP-4) [17]. In diabetic rodent models, Exendin-4 treatment improved cardiac function and glucose uptake [18]. Pretreatment of Exendin-4 also reduces the infarct area in the rat heart after ischemia reperfusion [19]. In these models, previous studies focused on the protective effect of Exendin-4 on cardiomyocyte apoptosis. However, the role of Exendin-4 in pyroptosis is largely unknown. The aim of this study is to investigate the potential benefits of Exendin-4 in hyperglycemia-induced cardiac remolding, inflammation, and cardiomyocyte pyroptosis and clarify the underlying molecular mechanisms.

Materials and Methods
2.1. Animals. C57BL/6J mice were obtained from the Model Animal Research Center (MARC) of Nanjing University (Nanjing, China). Animals were maintained in an SPFgrade laminar-flow housing apparatus under controlled temperature, humidity, and 12 h light/dark regimen. The experimental animal facility has been accredited by the AAALAC (Association for Assessment and Accreditation of Laboratory Animal Care International), and the IACUC (Institutional Animal Care and Use Committee) of Model Animal Research Center of Nanjing University approved all animal protocols used in this study.

2.3.
High-Fat Diet-Induced Type 2 Diabetic Mouse Model. For the high-fat diet-induced type 2 diabetic mouse study, 21 mice were randomly distributed to two groups fed a control diet (CON) (10% kcal fat, 70% kcal carbohydrate, and 23% kcal protein with a total caloric value of 3.85 kcal/gm, n = 7) or a high-fat diet (consisting of 45% kcal fat, 35% kcal carbohydrate, and 20% kcal protein with a total caloric value of 4.73 kcal/gm, n = 14). After a 16-week dietary intervention, the high-fat diet-fed group was then randomly subdivided to receive Exendin-4 (HFD+EXE) (25 nmol/kg/d, n = 7) or normal saline (HFD) by intraperitoneal injection during the light cycle. For 8-week administration of Exendin-4, the animals were sacrificed for cardiac histology and inflammation analysis.
2.5. Echocardiography. Echocardiography was performed 3 days before sacrifice. The mice in each group were anesthetized with isoflurane. Transthoracic two-dimensional M-mode echocardiography and pulsed-wave Doppler spectral tracings were obtained using the Vevo 2100 Imaging System (VisualSonics, Canada). The percentages of ejection fraction (EF%) were measured using M-mode tracings. The percentage of fractional shortening (FS) was calculated according to the following formula ½ðLVDD − LVSDÞ/LVDD × 100%.
2.6. Histology. For 24-week dietary intervention, hearts were isolated and fixed in 4% paraformaldehyde for 4 hours followed by gradual dehydration. Then, the heart tissues were embedded in paraffin and cut into 6 μm sections. Masson's trichrome staining was performed as MARC SOP, and the samples were imaged with an Olympus inverted microscope (Olympus BX51, Japan).

2.7.
Immunofluorescence and Immunohistochemistry. Immunofluorescence staining of WGA was performed as previously reported. Briefly, the hearts were fixed with 4% paraformaldehyde for 2 hours on ice and embedded in optimum cutting temperature compound (OCT, Leica, 4853, Germany). The frozen samples were cut into 10 μm sections. The sections were rinsed with PBS and incubated with Wheat Germ Agglutinin (WGA; Thermo Fisher, W11261, USA) and 4′,6-diamidino-2-phenylindole (DAPI; Jackson ImmunoResearch, USA) at 4°C for 30 min. FLICA staining was performed according to the manufacturer's instructions (ICT097, Bio-Rad, USA). The fluorescent images were captured with a ZEISS LM780 laser scanning confocal microscope (Zeiss, Germany).
Immunohistochemistry was performed to examine the level of pyroptotic proteins in the diabetic myocardium. The preparation of section samples was the same as in histological analysis. After antigen retrieval with heating, specimens were incubated with primary antibodies against cleaved caspase-1 (Cell Signaling Technology, #4199, MA, USA), NLRP3 (Novus Biologicals, NBP2-12446, USA), IL-1β (Abcam, ab2105, UK), and IL-18 (Abcam, ab71495, UK) at 4°C overnight. After incubation with secondary antibodies, the sections were stained with diaminobenzidine and imaged with an Olympus inverted microscope (Olympus BX51, Japan).

Isolation and Culture of Primary
Cardiomyocytes. Primary cardiomyocytes were isolated from 1-to 3-day-old neonatal C57BL/6 mice via collagenase digestion according to the manufacturer's protocol (Worthington, USA). Briefly, cardiac tissues were rinsed with HBSS and sequentially digested by pancreatin and collagenase type II without Ca 2+ . The isolated cells were resuspended in Ca 2+ containing L15 medium with oxygen awaking. Cardiomyocytes were filtered and purified by differential plating, and 0.1 mM 5-bromo-2-deoxyuridine (BrdU; Sigma-Aldrich, B5002, Germany) was added to the medium to prevent the proliferation of nonmyocytes. Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Hyclone, USA) were used to culture the cardiomyocytes, which were divided into four groups and treated with low glucose (5 mM glucose, CON) or high glucose (40 mM glucose, HG) with or without Exendin-4 (25 nM, HG+EXE) and CC (10 μM, HG+EXE+CC) for 24 h. All cells were incubated at 37°C in humidified air with 5% CO 2 .
2.10. Reverse Transcriptase qPCR. The animals were euthanized, and the hearts were dissected and frozen in liquid nitrogen. Total RNAs from cultured neonatal cardiomyocytes or heart tissues were extracted using 1 ml of TRIzol reagent (Invitrogen, USA) according to the manufacturer's instructions. cDNA synthesis was performed using the SuperScript III Reverse Transcriptase Kit (Thermo Fisher, 18080044, USA) according to the manufacturer's instructions. The SYBR Green PCR Master Mix Kit (Applied Biosystems, 4309155, USA) was used to quantify the relative mRNA levels of IL-1β, IL-6, IL-10, TNF-α, ICAM-1, VCAM-1, caspase-1, NLRP3, and TXNIP. Real-time PCR was performed with the 7500 FAST Real-Time PCR System (Applied Biosystems, USA) for 40 cycles, with GAPDH serving as internal controls. The following primers were used in the study: 2.12. Caspase-1 Activity Assay. Caspase-1 activity was determined using the Caspase-1 Assay Kit (fluorometric) from Abcam (ab39412, UK) according to the manufacturer's instructions. Briefly, cells were lysed and the nuclei and organelles were removed by centrifugation at 20,000g. 50 μg of total cytosolic protein was used to assess cytosolic caspase activity. Cell homogenates were incubated up to 4 hours at 37°C with corresponding caspase substrate conjugated to the chromophore p-nitroanilide. Cleavage of the substrate was quantified spectrophotometrically at 405 nm using a plate reader (Bio-Rad, USA).

Western
Blotting. The Western blot assay was conducted as previously described. Briefly, cell lysates prepared in standard cell lysis buffer were separated in 10% SDS-polyacrylamide, and proteins were transferred to PVDF membranes (Millipore, USA

Exendin-4 Ameliorates Glucose Metabolic Disorder in
High-Fat Diet-Induced Rodent Models. A high-fat diet-(HFD-) fed mouse model is robust to mimic metabolic abnormalities in human diabetic patients, including obesity, hyperglycemia, and complications of diabetes [20]. After 24-week dietary intervention, the mean body weights of CON, HFD, and HFD+EXE were measured. Compared to that of CON, the body weight of HFD was increased. Exendin-4 treatment (25 nmol/kg/d) prevented gain of weight by HFD (Figure 1(a)). Blood plasma was collected for glucose metabolic analysis. HFD feeding resulted in a significantly increased level of fasting blood glucose (FBG), while HFD +EXE mice showed identical FBG level with the CON (Figure 1(b)). Exendin-4 treatment also facilitated insulin secretion after glucose injection (Figure 1(c)). The homeostasis model assessment (HOMA-IR) level and IPGTT indicated that Exendin-4 improved glucose tolerance and insulin resistance (Figures 1(d) and 1(e)). The HFD+EXE group showed a reduced area under the curve compared to the other two groups (Figure 1(f)). These results exhibited the metabolic effects for glycemic control of Exendin-4 in a type 2 DM mouse model.

Exendin-4 Protects against Cardiac Dysfunction and
Remolding in Diabetic Mice. To determine the cardioprotective role of Exendin-4 in DCM, we studied the cardiac function and histology of the three groups of animals. Echocardiography was performed 7 days prior to sacrifice. Eject fraction (EF) and fractional shorting (FS) were measured for systolic and diastolic function assessment. HFD-fed mice displayed impaired systolic and diastolic function (Figure 2(a)). FS and EF were decreased in HFD compared to CON mice (Figures 2(b) and 2(c)). Exendin-4 treatment significantly rescued HFD-induced cardiac dysfunction (Figures 2(a)-2(c)). Masson's trichrome stain showed thickened left ventricular free walls (Figure 2(d)) and fibrotic deposition in HFD hearts (Figure 2(e)), which implicated pathological cardiac remolding. Treatment of Exendin-4 protected against HFD-induced myocardial hypertrophy and fibrosis (Figures 2(d) and 2(e)). To study the morphology of cardiomyocytes, we stained the heart section with WGA, a widely used dye labeling cell membrane (Figure 2(f)). Quantification of the transverse sectional area of cardiomyocytes indicated that Exendin-4 treatment prevented HFD-driven cardiomyocyte hypertrophy (Figure 2(g)).
Taken together, these evidences demonstrated that Exendin-4 protects against cardiac dysfunction and remolding in a type 2 diabetic rodent model.

Exendin-4 Treatment Attenuated Inflammatory
Response in the Diabetic Myocardium. Chronic inflammation involved in cardiac remolding is considered a consequence of cardiomyocyte pyroptosis. To assess the strength of cardiac inflammatory response in the three groups of animals, we performed reverse transcriptase qPCR to quantify the mRNA level of proinflammatory genes, including IL-1β, IL-6, TNF-α, TGF-β, ICAM-1, and VCAM-1. Compared to those in the CON, most proinflammatory genes were upregulated over 3 times in HFD samples. Exendin-4 treatment attenuated the HFD-driven transcriptional activity of these genes (Figure 3(a)). Next, we purified ventricular protein extraction and examined the enrichment of proinflammatory cytokines: IL-1β, IL-6, TNF-α, and TGF-β. ELISA results displayed that Exendin-4 negatively regulated the production of these proinflammatory cytokines induced by HFD (Figures 3(b)-3(e)). Next, we examined the level of the anti-inflammatory cytokine IL-10 in heart tissues. qPCR results showed that HFD feeding enhanced the transcriptional activity of IL-10, which was not altered by Exendin-4 ( Figure 3(f)). However, HFD+EXE samples showed increased protein level of IL-10, compared to the HFD ones ( Figure 3(g)). Taken together, these evidences implicated that Exendin-4 attenuated inflammatory response via regulation of both proinflammatory and anti-inflammatory cytokines.

Treatment of Exendin-4 Suppressed the Activation of
Caspase-1 in Diabetic Cardiomyocytes. Pyroptosis is a great contributor to cardiac inflammation in DCM [21]. Exendin-4 treatment significantly downregulated cardiac IL-1β level induced by HFD feeding. Meanwhile, IL-1β is a direct target of caspase-1, the pyroptotic-specific proteinase caspase [11]. Therefore, it is interesting to ask whether Exendin-4 is involved in the regulation of the pyroptotic signaling pathway. We performed immunohistochemistry to visualize the expression level of pyroptotic proteins in the heart, including cleaved caspase-1, NLRP3, IL-1β, and IL-18. Compared to the CONs, seriously increased levels of cleaved caspase-1, IL-1β and IL-18, and NLRP3 were detected in HFD samples. Exendin-4 treatment suppressed HFD-induced caspase-1 cleavage and decreased the protein levels of IL-1β and IL-18, but not NLRP3 (Figure 4(a)). Meanwhile, qPCR results showed no significant difference of caspase-1 and NLRP3 mRNA levels between HFD and HFD+EXE samples (Figures 4(b) and 4(c)). These results suggested that Exendin-4 regulated the activation of caspase-1 in the diabetic heart.
To determine the role of Exendin-4 in cardiomyocyte pyroptosis, we examined pyroptotic components in high glucosetreated cardiomyocytes. The primary cardiomyocytes were incubated in the absence (CON) or presence of 40 mM glucose (HG) in combination with Exendin-4 (25 nM) (HG+EXE). Culture medium was collected for quantification of the secretion of IL-1β and IL-18 using ELISA. Compared to the CON, HG samples exhibited dramatically  increased level of both ILs. Exendin-4 treatment attenuated the increased production of IL-1β and IL-18 induced by high glucose (Figures 4(d) and 4(e)). To examine the expression level of caspase-1, we purified protein extraction and performed Western blotting. High glucose facilitated the generation of cleaved caspase-1 (p20) and slightly increased the level of NLRP3 and ASC. Treatment of Exendin-4 inhibited the cleavage of caspase-1 but did not alter the enrichment of NLRP3 and ASC (Figure 4(f)). To determine the activity of caspase-1, firstly, we applied the fluorometric assay based on YVAD-AFC, a substrate of caspase-1 emitting 400 nm light after cleavage. Incubation of cardiomyocytes with high glucose caused significant upregulation of caspase-1 activity, which was diminished by Exendin-4 treatment (Figure 4(g)).
The role of Exendin-4 in caspase-1 activation was further identified by the FLICA assay, which labels active caspase-1 in living cells. Quantification of FLICA enrichment in 10 single cells each group confirmed the inhibitory effects of Exendin-4 on caspase-1 activation (Figure 4(h)). In general, these evidences demonstrated that Exendin-4 inhibited the activation of caspase-1 and protected cardiomyocytes against T2DM-induced pyroptosis.
To determine the antioxidative role of Exendin-4 in high glucose-treated cardiomyocytes, we incubated the cells with fluorescent probe labeling intracellular ROS. Flow cytometry exhibited that high-glucose treatment enlarged the cellular population with high enrichment of ROS. Exendin-4 restored the FACS profile of HG to that of CON ( Figure 5(a)). It indicated the antioxidative role of Exendin-4 in high glucoseinduced cardiomyocytes. ROS-driven activation of the NLRP3 inflammasome was characterized by the interaction between NLRP3 and TXNIP [22][23][24][25]. As a link between oxidative stress and cardiac inflammation in various cardiovascular diseases [26], TNXIP is a potential regulator involved in the antipyroptotic property of Exendin-4 in high glucose-induced cardiomyocytes. To verify this hypothesis, firstly, we examined the expression level of TXNIP. Incubation of cardiomyocytes in HG medium caused the upregulated mRNA level of TXNIP, compared to CON. Exendin-4 treatment did not alter the transcriptional activity of the TXNIP gene ( Figure 5(b)). However, Western blot showed that Exendin-4 treatment significantly decreased the protein level of TXNIP ( Figure 5(c)). It was suggested that TXNIP is posttranscriptionally regulated by Exendin-4 in diabetic cardiomyocytes. Furthermore, we administered cardiomyocytes with siRNA targeting TXNIP and examined the pyroptotic activity. RNA silencing of TXNIP significantly decreased the expression level of TXNIP ( Figure 5(d)) and diminished high glucose-induced caspase-1 activation and IL-1β production (Figures 5(e) and 5(f)). It mimicked the inhibitory effects of Exendin-4 on pyroptosis. These results implicated the degeneration of TXNIP as a mediator of Exendin-4 in antipyroptotic regulation.
It is crucial to identify the ROS sensors linking antioxidative and antipyroptotic properties of Exendin-4 in diabetic cardiomyocytes. Previous studies revealed that AMPK is a key regulator of energy metabolism and inflammation in DCM [4,27,28]. Activated AMPK degenerates TXNIP to manipulate the activity of the NLRP3 inflammasome [29]. Meanwhile, the activation of AMPK is driven by oxidative stress via ROS-dependent phosphorylation [30]. In high glucose-treated cardiomyocytes, a slight increase in phosphorylated AMPK (pAMPK) was observed. Interestingly, Exendin-4 treatment further facilitated the phosphorylation but did not alter the level of total AMPK (Figures 5(c) and 5(g)). To determine the role of AMPK in antipyroptotic effects of Exendin-4, we pretreated HG+EXE cardiomyocytes with CC (10 μM), a specific AMPK inhibitor, and examined the level of TXNIP and of IL-1β. CC treatment increased the protein level of TXNIP ( Figure 5(c)) and majorly blocked the inhibitory effect of Exendin-4 on IL-1β production ( Figure 5(g)). Moreover, the activity of caspase-1 was significantly increased in HG+EXE+CC, compared to HG+EXE ( Figure 5(h)). These evidences strongly supported that ROS-sensitive AMPK mediated the antipyroptotic properties of Exendin-4 in hyperglycemia-induced cardiomyocytes.

Discussion
Exendin-4 has been characterized by cardioprotective effects on cardiovascular diseases, including DCM, in human being [31]. Exendin-4 is involved in apoptosis, autophagy, and various biological processes to promote cardiomyocyte survival [32][33][34]. As a vital contributor to the development of myocardial inflammation, pyroptosis has not yet been reported to be regulated by Exendin-4 treatment. Our studies revealed the antipyroptotic property of Exendin-4 mediated by the AMPK-TXNIP pathway in hyperglycemia-induced cardiomyocytes.
In a type 2 diabetic mouse model, Exendin-4 treatment mostly rescued cardiac remolding and dysfunction. It has been determined that myocardial inflammation causes loss of cardiomyocytes that eventually results in heart failure [4]. As shown in previous studies [3,6,35,36], diabetic cardiac tissues express increased level of multiple proinflammatory cytokines promoting progression of myocardial  inflammation. Exendin-4 treatment negatively regulates the transcription of proinflammatory cytokines and decreased their protein level. In addition, our results showed that the mRNA level of anti-inflammatory IL-10 was increased in the heart of HFD-fed animals. The increased level of IL-10 has also been observed associated with inflammation in various cardiovascular diseases [37,38]. IL-10 is well characterized as an antiinflammatory factor and protective factor in myocardial infarction in diabetic animals [39][40][41][42]. Therefore, we also proposed that upregulation of IL-10 as a protective mechanism in the HFD-induced diabetic heart attenuates immune response and facilitates cell survival [43]. Exendin-4 treatment did not affect the transcriptional activity of IL-10 but further increased its protein level. It was suggested that Exendin-4 regulates IL-10 production via a posttranscriptional mechanism. Taken together, we characterized Exendin-4 as anti-inflammatory with clinical potential for DCM treatment.
In the diabetic heart, the NLRP3 inflammasome responds to hyperglycemia-induced toxicity and initiates the progression of pyroptosis [6,7,14]. In this study, the protein level of inflammasome components, NLRP3 and ASC, was not affected by Exendin-4-treated hyperglycemia cultured cardiomyocytes, while the upstream activator of inflammasome, TXNIP, was downregulated significantly [22,24]. The inflammasome is characterized as a molecular dock recruiting procaspase-1 and facilitates its autoactivation [13]. Substantial evidences provided by both in vivo and in vitro experiments strongly supported that Exendin-4 blocked the activation of caspase-1. Meanwhile, the production of pyroptotic proinflammatory cytokines IL-1β and IL-18 was also diminished by Exendin-4. Based on the above evidence, we proposed that Exendin-4 inhibited pyroptosis via blocking the activation of core pyroptotic regulators, inflammasome and caspase-1.
Oxidative stress and the accelerated ROS production induced by high glucose are known to play key roles in the progression of diabetic cardiovascular disease and cardiomyocyte pyroptosis [21]. In this study, Exendin-4 exhibited antioxidative and antipyroptotic effects in high glucosetreated cardiomyocytes. Previous studies revealed that AMPK, as a ROS sensor activated by Exendin-4, protects against cardiac dysfunction in DCM [33,44]. Our results indicated that activation of AMPK played an important role in the inhibition of caspase-1 activity by Exendin-4 treatment. In our proposal, on the hand, Exendin-4 treatment promotes the activation of AMPK to suppress the activation of the downstream NLRP3 inflammasome and caspase-1. The phosphorylation of AMPK, in response to hyperglycemiainduced oxidative stress, is upregulated as a resistant mechanism against ROS-induced cardiac inflammation. However, in pathological condition, the expression of TXNIP is significantly upregulated that the activity of AMPK is not sufficient for TXNIP degeneration in the absence of Exendin-4. On the other hand, Exendin-4 may facilitate degeneration of TXNIP by improving its sensitivity to AMPK activation. It is known that degeneration of TXNIP is an AMPK-dependent process [29]. Desensitization of TXNIP to AMPK activity may attenuate efficiency of AMPK on TXNIP degeneration. The later hypothesis requires further evidence to prove in future studies.

Conclusion
Our study revealed that Exendin-4 improved cardiac function and attenuated inflammatory response in the diabetic heart and protected cardiomyocytes against hyperglycemia-induced pyroptosis. The antipyroptotic property of Exendin-4 was mediated by the ROS-AMPK-TXNIP pathway, which regulated the activity of the inflammasome and caspase-1 in diabetic cardiomyocytes.

Data Availability
All data generated and analyzed during the current study are included in this published article.