Endogenous Hydrogen Sulfide Persulfidates Caspase-3 at Cysteine 163 to Inhibit Doxorubicin-Induced Cardiomyocyte Apoptosis

Doxorubicin (DOX) is an efficient antitumor anthracycline drug, but its cardiotoxicity adversely affects the prognosis of the patients. In this study, we explored whether endogenous gasotransmitter hydrogen sulfide (H2S) could protect against DOX-induced cardiomyocyte apoptosis and its mechanisms. The results indicated that DOX significantly downregulated endogenous H2S production and endogenous synthetase cystathionine γ-lyase (CSE) expression and obviously stimulated the apoptosis in H9C2 cells. The supplement of H2S donor sodium hydrosulfide (NaHS) or overexpression of CSE inhibited DOX-induced H9C2 cell apoptosis. DOX enhanced the activities of caspase family members in cardiomyocytes, while NaHS attenuated DOX-enhanced caspase-3, caspase-2, and caspase-9 activities by 223.1%, 73.94%, and 52.29%, respectively. Therefore, taking caspase-3 as a main target, we demonstrated that NaHS or CSE overexpression alleviated the cleavage of caspase-3, suppressed caspase-3 activity, and inhibited the cleavage of poly ADP-ribose polymerase (PARP). Mechanistically, we found that H2S persulfidated caspase-3 in H9C2 cells and human recombinant caspase-3 protein, while the thiol-reducing agent dithiothreitol (DTT) abolished H2S-induced persulfidation of caspase-3 and thereby prevented the antiapoptotic effect of H2S on caspase-3 in H9C2 cells. The mutation of caspase-3 C148S and C170S failed to block caspase-3 persulfidation by H2S in H9C2 cells. However, caspase-3 C163S mutation successfully abolished the effect of H2S on caspase-3 persulfidation and the corresponding protection of H9C2 cells. Collectively, these findings indicate that endogenous H2S persulfidates caspase-3 at cysteine 163, inhibiting its activity and cardiomyocyte apoptosis. Sufficient endogenous H2S might be necessary for the protection against myocardial cell apoptosis induced by DOX. The results of the study might open new avenues with respect to the therapy of DOX-stimulated cardiomyopathy.


Introduction
Doxorubicin (DOX) is commonly used for the chemotherapy of a variety of solid tumors and hematological tumors. However, its cardiotoxicity is an important issue in the field of oncocardiology. With the increase in the cumulative dose of DOX, the prevalence of congestive heart failure also increases [1]. Moreover, the mortality rate of patients treated with DOX-induced heart failure was relatively increased [2]. Since cardiomyocyte apoptosis is an important pathological basis of DOX-induced cardiomyopathy [3], it is of great significance to clarify the underlying mechanism for improving cardiac function and life quality in patients treated with DOX. Previous data have shown that the inhibition of DNA/RNA/protein synthesis in cardiomyocytes [4,5], oxidative stress injury [6], lipid peroxidation [7], calcium homeostasis disorders, endoplasmic reticulum stress [8], mitochondrial dysfunction [9], increased inflammatory mediators [10], and abnormal autophagy regulation [11] are all important mechanisms for DOX-induced apoptosis in myocardial cells. However, the molecular mechanisms underlying apoptosis in cardiomyocytes exposed to DOX have not yet been fully elucidated and urgently require further in-depth exploration.
Cysteine aspartic acid-specific proteases (caspases) exist in the cytoplasm, and their active sites all contain cysteines which can specifically cleave peptide bonds on target aspartic acid residues and play a significant role in apoptosis. A previous study has shown that DOX upregulates the mRNA levels of caspase-6 in rat cardiomyocytes [12]. Caspase-6, an effector caspase, depends on caspase-3 activation [13,14] and cleaves the lamina required for cell morphology, accelerating chromatin condensation and apoptotic body formation [14]. Caspase-6 can also activate caspase-8 [13]. In the cytoplasm, activated caspase-8 cleaves BH3-interacting domain death agonist (Bid) to form the active molecule truncated Bid (tBid), which facilitates cytochrome c release from mitochondria [15]. DOX-stimulated oxidative stress decreases the Bcl-2/Bax ratio [12], resulting in changes in the structure of the mitochondrial membrane to facilitate cytochrome c translocation to the cytosol. Cytochrome c triggers the cleavage of caspase-9, which further increases the activity of the effector molecule caspase-3 [16,17]. DOX can also upregulate the expression of death receptors, such as TNF receptor 1 [18], Fas [19], DR4, and DR5 [20], which combine with the corresponding free ligand to form a specific complex that initiates caspase-8 to directly activate caspase-3 [21]. Caspase-4 and caspase-12 are highly homologous, and both exist in the endoplasmic reticulum as precursors [22]. DOX-stimulated endoplasmic reticulum stress activates caspase-4 and caspase-12 and promotes the release of the two caspases from the endoplasmic reticulum to the cytoplasm. And then, the cytoplasmic activated caspase-4 and caspase-12 initiate the caspase cascade including caspase-9 and caspase-3, ultimately leading to apoptosis [22]. The above studies suggest that caspases are key molecules in the mechanisms of DOX-stimulated cardiomyocyte apoptosis. Previous studies have shown that the phosphorylation of caspase-9 at Thr-125 and Ser-196 inhibits the activation of caspase-9 [23,24], while phosphorylation at Tyr-251 enhances the activity of caspase-9 and accelerates apoptosis [25]. The glutathionylation of caspase-3 at Cys-220 and Cys-163 suppresses its activity [26]. These findings suggest that the posttranslational modification of caspases is important for the regulation of their activities.
Hydrogen sulfide (H 2 S) is a newly identified gasotransmitter [27][28][29][30]. It is primarily generated endogenously by cystathionine γ-lyase (CSE) catalysis in the cardiovascular system. H 2 S plays a major regulatory role in maintaining cardiovascular function and structure [31,32]. Previous studies have shown that persulfidation targeting sulfhydryl groups at key cysteine residues of proteins is an important molecular mechanism by which H 2 S regulates protein function [33,34]. H 2 S persulfidates CSE to enhance its activity and decrease homocysteine levels to reduce atherosclerosis [35]. H 2 S inhibits IKKβ activity through persulfidating Cys-179 on IKKβ, inhibiting the inflammatory response of pulmonary artery endothelial cells [33]. Previous research by our group demonstrated that H 2 S donors significantly improved cardiac function and attenuated cardiac pathological micro-and ultrastructural changes in DOX-treated rats, including inflammatory cell infiltration and mitochondrial edema [36]. However, it is not clear whether and how H 2 S regulates caspase activity to inhibit the apoptosis of DOXinduced cardiomyocytes.
Therefore, in the present study, we examined the impact of endogenous H 2 S on caspases and the underlying molecular mechanism to reveal new insights into the development of DOX-stimulated cardiomyocyte apoptosis.

Materials and Methods
2.1. Cell Culture and Experimental Treatments. The rat embryonic cardiomyocyte H9C2 cell line was procured from Wuhan Procell Company (Wuhan, China). Cells were cultured in DMEM supplemented with 10% FBS (Thermo, Gibco, United States) and 1% penicillin/streptomycin (37°C, 5% CO 2 ). When H9C2 cells reached semiconfluence, they were starved in the basal DMEM for 8 h. The H9C2 cell model was prepared by exposure to DOX (0.3 μmol/L) for 24 h [37], and H 2 S donor sodium hydrosulfide (NaHS, 100 μmol/L) or CSE overexpression was used to provide sufficient H 2 S levels.

Cell Viability Measured
Using the Cell Counting Kit-8 (CCK-8) Assay. Cell viability was analyzed using an enhanced CCK-8 kit (Beyotime, Shanghai, China) [38]. Cells were plated at a density of 5 × 10 3 /well in microplates. H9C2 cells were exposed to DOX (0, 0.1, 0.2, 0.3, 0.4, and 0.5 μmol/ L) after starvation in basal DMEM for 8 h. Twenty-four hours after treatment, the basic medium was changed to medium containing 10% CCK-8 and cultured for 2 h at 37°C in the dark. Cell-free wells with reagent were used as blank controls, and absorbance values were measured at 450 nm. Cell viability was calculated using the following equation: ðOD stimulated − OD blankÞ/ðOD unstimulated − OD blankÞ × 100%.

Apoptosis Assessment Using a TdT-Mediated dUTP
Nick-End Labeling (TUNEL) Kit. A TUNEL kit (Roche, USA) was applied to assess the apoptosis in H9C2 cells [39]. The cells were gently rinsed with PBS after discarding the supernatant and fixed in 4% paraformaldehyde for 15 min. Afterward, they were permeabilized for 30 min following incubation with the TUNEL reagent for 1 h at 37°C protected from light. Finally, nuclei were labeled using DAPI dye. Fluorescence images were captured using a confocal laser scanning microscope (Olympus, Japan). Green fluorescence represents the TUNEL-labeled signal marking apoptotic cells. The apoptotic index was described as the percentage of the number of TUNEL-labeled nuclei (green) to the total DAPI-labeled nuclei (blue).

Caspase Activity Detected Using the Caspase Family
Colorimetric Assay Kit in H9C2 Cells. The activities of caspases were assessed using a series of caspase activation assay kits (Applygen, Beijing, China) according to the manufacturer's instructions [38]. Caspase activity kits in the present study included caspase-1 colorimetric assay kit (C1111), 2 Oxidative Medicine and Cellular Longevity caspase-2 colorimetric assay kit (C1112), caspase-3 colorimetric assay kit (C1113), caspase-4 colorimetric assay kit (C1114), caspase-5 colorimetric assay kit (C1115), caspase-6 colorimetric assay kit (C1116), caspase-8 colorimetric assay kit (C1118), and caspase-9 colorimetric assay kit (C1119). The active caspases recognize specific peptides and cleave the substrate peptides at specific aspartic acid residues. Different caspase activity kits contain synthetic corresponding peptides in which the aspartic acid residue is conjugated with p-nitroanilide (pNA). The active caspase cleaves the peptide-pNA compound at the aspartic acid residue and releases free pNA which has a maximum absorbance at 405 nm. One unit of caspase enzymatic activity is defined as the hydrolysis of substrate peptide-pNA at 37°C for 1 h to release 1 nmol of free pNA. In brief, the cells were rinsed gently with precooled PBS and incubated in lysis solution on ice at 4°C for 15 min. The total protein was harvested into centrifuge tubes and chilled on ice for 10 min, followed by a centrifugation at 12000 g and 4°C. The sample and the substrate were added to a 96-well plate in sequence and then incubated for 2 h at 37°C protected from light. The pNA content was detected by measuring the absorbance OD value at 405 nm. Caspase activity was calculated according to the pNA standard curve and the OD value of the sample. The protein concentration was detected by the Bradford method and used to normalize the caspase activity.

H 2 S Detection Using a Fluorescent Probe in H9C2 Cells.
The H 2 S-specific fluorescent probe SF7-AM (Cayman, Ann Arbor, USA) was used to determine the H 2 S content in cells [40]. Cells were gently rinsed with PBS after discarding the medium. Subsequently, the cells were incubated with 2.5 μmol/L H 2 S fluorescent probe in the incubator for 30 min protected from light. The cells were incubated in 4% paraformaldehyde for 20 min at 25°C. Green fluorescence was captured using a confocal laser scanning microscope, indicating H 2 S content in the cells. DAPI dye was used for nuclear staining.

Caspase-3 Persulfidation Detection Using Biotin Switch
Analysis (BSA). The BSA method was used to determine the persulfidation of caspase-3 in H9C2 cells [34]. DOXtreated H9C2 cells were divided into DOX, DOX+NaHS, and DOX+NaHS+dithiothreitol (DTT) groups. DOXuntreated H9C2 cells were divided into control, NaHS, and NaHS+DTT groups. For the H9C2 cells, the treatment conditions included the incubation with 100 μmol/L NaHS for 2 h, 200 μmol/L DTT for 15 min, and 0.3 μmol/L DOX for 24 h. After the experiment, the cells were gently rinsed with PBS, and then, 120 μL of the cell lysis was added and incubated for 20 min at 4°C. The cells were collected into an EP tube and then centrifuged at 13000 g at 4°C to obtain the supernatant. Ten microliters of the supernatant were mixed with 2x SDS loading buffer and boiled to detect the total caspase-3 using denaturing SDS-PAGE electrophoresis. The remaining supernatant was used to analyze the caspase-3 persulfidation. First, the supernatant was incubated with 2.5% SDS and 20 mM S-methyl methanethiosulfonate (MMTP) for 20 min at 50°C. And then precooled acetone was added and kept at -20°C for 2 h to precipitate the protein. The precipitated protein pellets were collected after the centrifugation for 10 min at 4°C. Subsequently, the precipitants were dissolved in 200 μL of the cell lysis buffer containing 3 μL of EZ-linked iodoacetyl-PEG2 biotin (10 mg/mL) and then incubated on a shaker at 4°C for 12 h. To enrich the persulfidated protein, 6 μL of UltraLink Immobilized NeutrAvidin was used to combine the biotinylated proteins and incubated for 4 h at 4°C. The precipitation was collected by centrifuging at 5000 g at 4°C for 10 min. After rinsing 6 times with PBS containing 0.001% SDS, the samples were mixed with 1x β-mercaptoethanol-free SDS loading buffer, boiled, centrifuged at 5000 g to collect the persulfidated proteins, and loaded onto 10% SDS-PAGE to analyze caspase-3 persulfidation. Recombinant human caspase-3 protein (Origene, Rockville, USA) was divided into control, NaHS, and NaHS +DTT groups. In the NaHS group, 0.15 μg of caspase-3 3 Oxidative Medicine and Cellular Longevity protein was incubated with 100 μmol/L NaHS for 30 min at 37°C, while in the NaHS+DTT group, the protein was treated with NaHS and 1 μmol/L DTT for the last 15 min. The protocol for analyzing total and persulfidated caspase-3 protein was the same as that in the cell experiments.

Protein Expression Detected by Western Blot Analysis.
The expression of proteins in H9C2 cells was analyzed by western blotting analysis as described in a previous study [41]. Briefly, H9C2 cells were collected and rinsed gently with precooled PBS. Each well of the 6-well plate was lysed using 60 μL of cell lysis buffer on ice for 30 min. The supernatant was obtained after the centrifugation of 12000 g at 4°C, and 1 μL of the supernatant was used for the quantification of the protein concentration using the BCA method. The remaining protein was denatured in 2x loading buffer for 10 min. The denatured proteins (15-30 μg) were separated using 10%-15% SDS-PAGE electrophoresis and transferred onto nitrocellulose (NC) membranes. Nonspecific binding was blocked using 5% skimmed milk at 25°C for 1 h. The NC bands were incubated at 4°C for 12 h with the primary antibodies diluted in PBST: caspase-3 (1 : 1000, AC030, Beyotime, China), cleaved caspase-3 (1 : 500, AC030, Beyotime), PARP (1 : 1000, AP102, Beyotime), CSE (1 : 1000, 12217-1-AP, Proteintech, China), β-tubulin (1 : 4000, TA-10, ZSGB, China), and β-actin (1 : 4000, TA-09, ZSGB, China). After washing with PBST, the bands were immersed with the corresponding horseradish peroxidaselabeled secondary antibodies at a working concentration of 1 : 5000 at 25°C for 1 h. Finally, the NC bands were incubated with ECL chemiluminescence reagents. A FluorChem M chemiluminescence imaging analysis system was used to take pictures, and ImageJ software (version 1.4.3) was used for gray value analysis.
2.11. Statistical Analysis. SPSS 17.0 (IBM, USA) and Graph-Pad (GraphPad, CA, USA) were applied for the statistical analysis. All data are expressed as the mean ± SEMs. A comparison among multiple groups was performed using ANOVA followed by Tukey's test if the data were normally distributed or followed by a Dunnett T3 procedure if the data were not normally distributed. p < 0:05 was regarded as statistically significant.

H 2 S Donor or CSE Overexpression Prevented DOX-Stimulated Apoptosis in Cardiomyocytes.
Cells were treated with DOX (0.1-0.5 μmol/L) for 24 h, and the viability of H9C2 cells was determined using the CCK-8 assay. The data indicated that cell viability treated with 0.1, 0.2, 0.3, 0.4, and 0.5 μmol/L DOX was decreased by 22.77%, 33.24%, 47.79%, 49.8%, and 52.28%, respectively, compared to the controls (p < 0:01, Figure 1(a)). By using 0.3 μmol/L DOX as an intervention dose in the following experiment, we showed that compared to the controls, the proportion of TUNELlabeled cells treated with DOX was increased by 85.94%, while CSE protein level decreased by 26.18% (p < 0:01, Figures 1(b) and 1(c)). A fluorescence probe assay was used to quantify H 2 S content in cells. A marked decrease in H 2 S content in DOX-stimulated H9C2 cells was observed, as evidenced by a significant decrease in H 2 S-specific green fluorescence compared to the controls (Figure 1(d)). These findings suggest that DOX downregulates the H 2 S/CSE pathway in cells and induces H9C2 cell apoptosis.
To explore the impact of endogenous H 2 S reduction on apoptosis in DOX-stimulated cardiomyocytes, treatment with the H 2 S donor or overexpression of the H 2 S-producing enzyme CSE was performed to increase the H 2 S content in DOX-treated H9C2 cells. The results of H 2 S donor NaHS (100 μmol/L) treatment revealed that the H 2 S fluorescence intensity of the cells was increased, while the proportion of the TUNEL-labeled cells decreased by 65.97% (p < 0:01, Figures 1(b) and 1(d)) in the DOX+NaHS group compared to the DOX group. The expression of CSE and H 2 S content in CSE-overexpressing cells was significantly increased (Figures 1(e) and 1(f)) compared to that in the EV group. In the cells transfected with EV, the treatment of DOX decreased the expression of CSE and H 2 S content (Figures 1(e) and 1(f)), but increased the proportion of apoptotic cells (p < 0:01, Figure 1(g)). However, DOX treatment did not affect CSE protein expression, H 2 S content, or the proportion of TUNEL-positive cells (Figures 1(e)-1(g)) in CSE-overexpressed cells. These results suggest that sufficient H 2 S content in H9C2 cells prevents the cardiomyocyte from DOX-stimulated apoptosis.

H 2 S Inhibited Caspase-3 Activity and Cardiomyocyte
Apoptosis by Persulfidating Caspase-3. To explore the molecular mechanism by which H 2 S inhibited caspase-3 activity and cardiomyocyte apoptosis, H9C2 cells were divided into DOX, DOX+NaHS, and DOX+NaHS+DTT groups. Compared with the DOX+NaHS group, caspase-3 activity was enhanced by 77.05%, the ratio of cleaved caspase-3/caspase-3 increased by 48.74%, the ratio of cleaved PARP/ PARP increased by 45.61%, and the proportion of TUNELpositive cells increased by 55.22% (p all <0.01) (Figures 3  (a)-3(c)) in the cells of the DOX+NaHS+DTT group. These findings suggested that the sulfhydryl reducing agent DTT abolished the suppressive effect of H 2 S on DOX-stimulated caspase-3 activity and cardiomyocyte apoptosis in H9C2 cells. Therefore, H 2 S might inhibit caspase-3 activity by targeting the sulfhydryl group.
To demonstrate how H 2 S affected the sulfhydryl group of caspase-3, BSA method was used to detect the persulfidation of caspase-3 in the DOX-treated H9C2 cells. Figure 3(d) shows that the persulfidation of caspase-3 in the cells of the DOX+NaHS group was increased by 99.7% compared with the DOX group (p < 0:01), while the H 2 S-induced caspase-3 persulfidation was weakened in the cells of the DOX +NaHS+DTT group, accompanied by the reactivated caspase-3 (p < 0:01). Furthermore, the treatment of NaHS also induced a significant persulfidation of caspase-3 in the DOX-untreated H9C2 cells and purified human caspase-3 protein, which was abolished by the intervention with DTT (p all <0.01) (Figures 3(e) and 3(f)). The abovementioned results suggested that H 2 S might suppress caspase-3 activity by persulfidating the sulfhydryl group of caspase-3 protein to inhibit the cardiomyocyte apoptosis.

Cys-163 Was the Key Site for H 2 S to Persulfidate
Caspase-3. It was reported that Cys-163 was the active catalytic site of caspase-3 [42]. Here, we investigated the exact  9 Oxidative Medicine and Cellular Longevity site on which H 2 S persulfidated caspase-3, targeting Cys-163. In the experiment, its two neighboring cysteine residues were used as the control residues. The caspase-3 wild-type plasmid (caspase-3-WT-His) and three caspase-3 mutant plasmids, including caspase-3-C163S-His, caspase-3-C148S-His, and caspase-3-C170S-His plasmids, were transfected into caspase-3 knockdown H9C2 cells, respectively. Each type of cells was, respectively, divided into control, NaHS, and NaHS+DTT groups. The results indicated that the persulfidation of caspase-3 in the NaHS group was greater than that of the control H9C2 cells transfected with caspase-3-WT-His, caspase-3-C148S-His, or caspase-3-C170S-His plasmid, which was abolished by the DTT (p all <0.01). However, there was no difference in the ratio of persulfidated caspase-3/caspase-3 with or without NaHS treatment in the cells transfected with the caspase-3-C163S-His plasmid (Figure 4(a)). These results suggested that Cys-163 might be the target residue on which H 2 S persulfidated caspase-3.
Furthermore, to examine the significance of caspase-3 persulfidation in the H 2 S-inhibited caspase-3 activation and apoptosis of H9C2 cells treated by DOX, we compared the inhibitory effect of H 2 S on caspase-3 activity and apoptosis between H9C2 cells transfected with caspase-3-WT-His and caspase-3-C163S-His plasmids. The results revealed that in the H9C2 cells transfected with caspase-3-WT-His plasmid, H 2 S donor inhibited DOX-stimulated activation of caspase-3 and the increase in the proportion of TUNELpositive cells (p < 0:01) (Figures 4(b) and 4(c)). However, in the cells transfected with the caspase-3-C163S-His plasmid, neither caspase-3 activity nor cell apoptosis was affected by the treatment of H 2 S donors (Figures 4(d) and 4(e)).

12
Oxidative Medicine and Cellular Longevity These above findings suggested that H 2 S persulfidated caspase-3 at Cys-163 to inhibit the DOX-induced caspase-3 activation and accordingly cardiomyocyte apoptosis.

Discussion
In the present study, we revealed that DOX induced cardiomyocyte apoptosis by downregulating the CSE/H 2 S pathway. Furthermore, sufficient endogenous H 2 S persulfidated caspase-3 to inhibit its activity, alleviating DOX-stimulated apoptosis of cardiomyocytes. The Cys-163 site of caspase-3 acted as the target residue for H 2 S to persulfidate caspase-3 protein ( Figure 5).
DOX, an efficient anthracycline antitumor agent, can cause cardiomyopathy, cardiac damage, arrhythmia, and heart failure. Cardiomyocyte apoptosis is an important pathological basis of DOX-induced cardiomyopathy [3,6,9]. Previous studies have shown that DOX directly damages DNA, generates reactive oxygen species through protein/ DNA interactions, and leads to cell death [43]. DOX also inhibits topoisomerase II, resulting in double-strand DNA breakage and hindering cell replication and transcription [44]. However, the molecular mechanisms underlying DOX-stimulated cardiomyocyte apoptosis have not yet been entirely clear and need to be further explored.
Endogenous H 2 S is a novel gasotransmitter identified after nitric oxide (NO) and carbon monoxide (CO) [27-30, 45, 46]. It is mainly generated endogenously by CSE catalyzed in the cardiovascular system [47]. H 2 S is widely involved in the pathophysiological processes in the body, especially in the cardiovascular system. In preliminary animal experiments, H 2 S improved cardiac function in DOXinduced cardiomyopathy in rats and reduced myocardial damage. H 2 S donors have been demonstrated to significantly reduce cardiomyocyte apoptosis [36]. In this study, we explored the regulatory effect of H 2 S on the DOX- (e) Biotin switch analysis was used to detect the persulfidation of caspase-3 protein in DOX-untreated H9C2 cells and β-actin as an internal control (n = 9). (f) Biotin switch analysis was used to detect the persulfidation of human recombinant caspase-3 protein (n = 9). The data were shown as mean ± SEMs, * * p < 0:01. DOX: doxorubicin; DOX+NaHS: doxorubicin+sodium hydrosulfide; DOX+NaHS+DTT: doxorubicin +sodium hydrosulfide+dithiothreitol; Ctrl: control; NaHS: sodium hydrosulfide; NaHS+DTT: sodium hydrosulfide+dithiothreitol. 14 Oxidative Medicine and Cellular Longevity stimulated apoptosis of cardiomyocytes and its molecular mechanisms. It has been reported that the steady-state plasma concentration of DOX is 22.6-334 ng/mL (0.03-0.5 μmol/L) in patients treated with DOX [48]. Therefore, we used the CCK-8 method to assess the effect of different concentrations (0-0.5 μmol/L) of DOX on the viability of H9C2 cells, and the data indicated that 0.3 μmol/L DOX administration reduced H9C2 cell viability by 50%. The TUNEL assay indicated that the apoptotic rate of H9C2 cells exposed to 0.3 μmol/L DOX was as high as 85.94%, thus confirming the successful establishment of the apoptosis model induced by 0.3 μmol/L DOX for 24 h.

Oxidative Medicine and Cellular Longevity
The results demonstrated that DOX induced apoptosis of H9C2 cells and significantly downregulated the endogenous H 2 S production and the expression of endogenous synthase CSE. Previous studies showed that reduced H 2 S content could cause cardiomyocyte apoptosis during ische-mia/reperfusion injury [49,50]. Treatment with H 2 S donors to restore H 2 S content in heart tissue significantly inhibited isoprenaline-induced cardiomyocyte apoptosis [51]. In addition, endogenous H 2 S was involved in the development of diabetic cardiomyopathy [52,53]. These findings suggest that endogenous H 2 S has a strong myocardial protective effect and inhibits cardiomyocyte apoptosis in the pathogenesis of cardiac diseases. In the present study, H 2 S donor treatment and CSE overexpression were used to increase the content of H 2 S, which inhibited the apoptosis of DOXstimulated cardiomyocytes, suggesting that endogenous H 2 S/CSE downregulation might be an important mechanism for DOX-induced cardiomyocyte apoptosis.
The caspase family is synthesized in the form of inactive proenzymes, and the activated caspase molecules play significant roles in the apoptotic pathway in cardiomyocytes [54][55][56], gradually disintegrating cell components through the cleavage of specific substrates that can reconstruct the cytoskeleton and degrade nuclei. Furthermore, apoptotic cells exhibit shrunken morphology, condensed chromatin, and finally internuclear chromosomal DNA fragmentation [57]. Previous studies have shown that members of the caspase family are widely involved in the process of DOXinduced apoptosis in cardiomyocytes [12,55,56,58,59].
In this study, we found that DOX activated almost all members of the caspase family in cardiomyocytes, and H 2 S donor NaHS intervention inhibited caspase-2, caspase-3, and caspase-9 activities, exerting the strongest inhibitory effect on caspase-3. Among the caspase family members, caspase-3 is a critical regulator of apoptosis and serves as the hub of endogenous and exogenous apoptotic pathways, playing an important role at the final stage of apoptosis [60]. Therefore, caspase-3 has attracted widespread attention since its discovery in 1994 [61]. Caspase-3 precursor is a single-chain inactive peptide with a molecular weight of 32 kDa. Caspase-3 can be activated by its upstream regulators including caspase-8 or caspase-9 to produce active tetramer structures composed of two large (17 kD) and small (12 kD) subunits [62]. Activated caspase-3 further cleaves its downstream PARP into two fragments. Then, cleaved PARP fails to inhibit the activity of Ca 2+ /Mg 2+ -dependent nucleic acid endonuclease, leading to DNA fragmentation and programmed apoptosis [63]. In the present study, we found that either H 2 S donors or overexpression of CSE blocked the effects of DOX on caspase-3 activation and PARP cleavage. The results suggest that caspase-3 might be a target of H 2 S to prevent cardiomyocyte apoptosis. However, the mechanism by which H 2 S inhibits caspase-3 activity remains unclear.
Caspase-3 activity depends on a conserved QACRG pentapeptide sequence, in which a reducing sulfhydryl group at the cysteine residue is acquired to maintain the activity of caspase-3 [64]. As a strong reducing agent, DTT can block the oxidation process of sulfhydryl group and keep it in a reduced state (R-SH) [65]. In this study, we found that DTT blocked the suppressive effect of H 2 S on caspase-3 activation and apoptosis in DOX-treated cardiomyocytes, suggesting that the suppressive effect of H 2 S on the activation of caspase-3 and cell apoptosis might be related to the sulfhydryl group on the caspase-3. It was reported that H 2 S regulates target protein function by persulfidation, a posttranslational modification occurring at a specific sulfhydryl group, which converted a reduced sulfhydryl group (R-SH) to a persulfidated sulfhydryl group (R-SSH) and then affected the structure and function of the protein [34,66,67]. Therefore, we investigated the persulfidation of caspase-3 in NaHS-treated H9C2 cells and purified human caspase-3 protein. As expected, both the cellular and cell-free experimental findings suggested that H 2 S persulfidated caspase-3, which could be reversed by DTT treatment.
Furthermore, it is important to demonstrate the exact site at which H 2 S persulfidates caspase-3 to elucidate the cardioprotective mechanism responsible for H 2 S. Previous studies reported that Cys-163 is the active site catalyzing protease hydrolysis [62]. The glutathionylation of caspase-3 inhibits the activation of caspase-3 [26]. NO also exerts antiapoptotic effects through the nitrosylation of caspase-3 at Cys-163 [68][69][70][71][72]. Therefore, in this study, Cys-163 was selected as the candidate residue of H 2 S-induced persulfidation of caspase-3, while Cys-148 and Cys-170, the two neighbor cysteine residues, were used as control residues. Interestingly, we found that when Cys-163 in caspase-3 was replaced with Ser, the persulfidation of caspase by H 2 S no longer occurred. On the contrary, neither the mutation at Cys-148 nor Cys-170 impacted the persulfidation of caspase-3 by H 2 S. Moreover, H 2 S did not affect caspase-3 activity and cardiomyocyte apoptosis in DOX-treated H9C2 cells transfected with the mutated caspase-3-C163S-His plasmid. These data demonstrated that H 2 S inhibited the apoptosis of DOXstimulated cardiomyocytes by persulfidating caspase-3 at Cys-163 to suppress caspase-3 activation.
Regarding the mechanisms by which H 2 S-induced persulfidation of caspase-3 at Cys-163 inhibited the caspase-3   16 Oxidative Medicine and Cellular Longevity activity, we speculated the following possibilities. (1) The cleavage of substrate peptide bonds catalyzed by the caspase-3 depends on the attack of the nucleophile thiol group of catalytic Cys-163 on the aspartic acid at P1 position [73]. Based on the above characteristics of activated caspase-3, caspase-3 inhibitors were developed to inhibit caspase-3 activity by forming a covalent bond with the thiol group of Cys-163 [74]. Therefore, we speculated that the persulfidated sulfhydryl group of Cys-163 induced by H 2 S might change the nucleophilic ability of Cys-163 and then inhibit the caspase-3 activity. (2) Acting as an effector caspase, caspase-3 cannot be autoactivated. However, one activated caspase-3 could cleave another procaspase-3 to form an active caspase-3 [75] and then promote caspase-3 activation cascade. Therefore, the block of procaspase-3 cleavage might also be involved in the mechanism responsible for the inhibitory effect of caspase-3 persulfidation on the caspase-3 activity. In addition, the detection of caspase-3 persulfidation and activity in the heart from DOX-treated animals might also strengthen the causality relationship between the persulfidation of caspase-3 and the inhibition of caspase-3 activity, which merits further investigation.

Conclusion and Perspective
In summary, the present study revealed a novel mechanism by which H 2 S inhibited DOX-induced apoptosis of cardiomyocytes. We first showed that H 2 S inhibited the activity of caspase-3 by persulfidating Cys-163 residue on caspase-3, suppressing apoptosis in DOX-treated cardiomyocytes. Therefore, the results of this study clearly revealed the protective effect of endogenous H 2 S on the cardiotoxicity induced by DOX and its mechanisms, providing a new idea and target for the potential therapeutic intervention against DOX-stimulated cardiomyopathy. The effect of H 2 S on the antineoplastic efficacy of DOX should be paid great attention when it is considered as a potential treatment strategy in the anticancer therapy. Chegaev et al. developed new DOX derivatives and designed a series of H 2 S-donating DOXs which continuously released H 2 S in DMEM culture and human serum. Furthermore, H 2 S-donating DOXs exhibited a stronger cytotoxicity than DOX alone in both DOX-sensitive and DOX-resistant osteosarcoma cells via impairing membrane P glycoprotein transporter to block DOX efflux and enhance DOX effectiveness [76]. Wang et al. found that compared with DOX only, the implantation of DOX-ZnS@SiO 2 fibrous mesh significantly shrank the tumor size and made a more seriously histological damage in the Huh7 mouse tumor model by releasing H 2 S to enhance the proapoptotic effect of DOX [77]. The abovementioned findings suggested that cotreatment with H 2 S effectively enhanced chemotherapeutic efficacy of DOX in vitro and in vivo, which further supported the potential clinical application of H 2 S in the antineoplastic treatment.
Another important issue that should be concerned is about the PARP. PARP, acting as a DNA damage sensor and signal transducer, plays a key role in the DNA repair, and therefore, PARP inhibitor has been developed to block the DNA repair in the cancer cells treated with chemotherapy or radiotherapy [78]. Pettitt et al. found that siRNAmediated PAPR1 deficiency abolished the cytotoxicity of PARP1 inhibitors in both human CAL51 cells and DLD1 cells, suggesting that the normal expression of the intact PARP1 is essential for anticancer efficacy of PARP1 inhibitor [79]. In our present study, the integrity of PARP1 was preserved by H 2 S treatment since H 2 S suppressed the activated caspase-3-mediated cleavage of PARP1. Therefore, we speculated that H 2 S might cooperate with PARP inhibitor to enhance its anticancer effect in the cancer cells although the above speculation needs to be confirmed in the future.

Data Availability
The data used to support the findings of the study are available from the corresponding authors upon request.

Conflicts of Interest
The authors declare that there is no conflict of interest.

Authors' Contributions
Xiaoyun Ye conducted the experiments, analyzed data, and participated in manuscript writing. Yingying Li, Boyang Lv, Bingquan Qiu, Shangyue Zhang, and Hanlin Peng performed the experiments and analyzed data. Wei Kong, Yaqian Huang, Junbao Du, and Chaoshu Tang designed the research, interpreted the results of the experiment, and edited the manuscript. Hongfang Jin designed the experiments, provided the overall guidance, and contributed to manuscript writing.