Hydrogen sulfide-mediated regulation of cell death signaling ameliorates adverse cardiac remodeling and diabetic cardiomyopathy

Abstract

The death of cardiomyocytes is a precursor for the cascade of hypertrophic and fibrotic remodeling that leads to cardiomyopathy. In diabetes mellitus (DM), the metabolic environment of hyperglycemia, hyperlipidemia, and oxidative stress causes cardiomyocyte cell death, leading to diabetic cardiomyopathy (DMCM), an independent cause of heart failure. Understanding the roles of the cell death signaling pathways involved in the development of cardiomyopathies is crucial to the discovery of novel targeted therapeutics and biomarkers for DMCM. Recent evidence suggests that hydrogen sulfide (H₂S), an endogenous gaseous molecule, has cardioprotective effects against cell death. However, very little is known about signaling by which H₂S and its downstream targets regulate myocardial cell death in the DM heart. This review focuses on H₂S in the signaling of apoptotic, autophagic, necroptotic, and pyroptotic cell death in DMCM and other cardiomyopathies, abnormalities in H₂S synthesis in DM, and potential H₂S-based therapeutic strategies to mitigate myocardial cell death to ameliorate DMCM.

INTRODUCTION

Myocardial cell death is a critical element in the pathogenesis and progression of several etiologies of cardiomyopathies, one of which is present specifically in individuals with diabetes mellitus (DM) (11, 106). The diabetic environment of hormones, cytokines, hyperglycemia, fatty acids, and adrenergic tone leads to a structural and functional cardiac muscle disorder known as diabetic cardiomyopathy (DMCM). DMCM develops independently of atherosclerosis, valvular disease, hypertension, and other common cardiovascular disease contributors (8, 10) and as such presents itself as an independent mechanism by which heart failure may occur. The cell death initiated in DMCM contributes to the loss of contractile units, arrhythmias, compensatory hypertrophy, and fibrosis (143), thus increasing the risk of DM-induced heart failure. It is well demonstrated that major cell death pathways like apoptosis, autophagy, necrosis, and pyroptosis all play a role in the development and progression of DMCM and other cardiovascular diseases (8).

The accumulation of reactive oxygen species (ROS) and reactive nitrogen species (RNS) have been shown to play a critical signaling role in the myocardial cell death process (153). More recently, studies have revealed a growing role of reactive sulfide species in cell death (69). Hydrogen sulfide (H₂S) is an endogenous antioxidant gaseous signaling molecule that interacts with proteins present in body tissue and the circulation (112). Plasma H₂S levels are decreased in T2DM patients and high-fat diet-treated mice (4, 54, 58). In addition, preclinical studies have shown that exogenous H₂S treatment ameliorates DMCM, suggesting that restoring H₂S levels may be a potential novel treatment option for DMCM (4). Whereas other gaseous signaling molecules such as nitric oxide (NO) have well-established therapeutic roles, comparatively little is known about signaling by H₂S, which has the potential to act as a therapeutic gaseous molecule as well. Here, we review the role of H₂S in cell death mechanisms in cardiomyopathies, with a focus on DMCM. We elaborate on the pathways mediating these effects and discuss the therapeutic and cardioprotective potential of H₂S.

H₂S PRODUCTION AND SIGNALING

H₂S production is highly regulated, and it is produced on demand due to its potential toxicity at higher levels. All three enzymatic pathways that produce H₂S are centered on homocysteine and cysteine metabolism. These pathways are cystathionine β-synthase (CBS), cystathionine γ-lyase (CSE), and the coupling of cysteine aminotransferase (CAT) and 3-mercapto-pyruvate sulfur transferase (3-MST) (40, 69, 191). The expression of H₂S-generating enzymes is subcellular and tissue specific. On a cellular level, CSE occurs strictly in the cytoplasm, whereas CAT is located in the mitochondria. Regarding tissue specificity, CSE is most abundant in the cardiovascular system (83), whereas CBS is predominant in the nervous system and liver and expressed in the heart.

A growing body of research is uncovering a role of H₂S in cellular signaling similar to the other gasotransmitters. One of the mechanisms by which H₂S signals cellular processes is by the sulfhydration of cysteine residues of target proteins (112). This posttranslational modification is reversible by the thioredoxin system, further lending credence to its role in signaling (6, 71). Whereas NO acts on proteins through nitrosylation, a more substantial portion of proteins is sulfhydrated by H₂S, thus increasing enzyme activity. Notably, cases of inhibition of protein activity with sulfhydration also exist (105, 112). H₂S also interacts with proteins to produce reactive persulfide that in turn causes structural changes in proteins (69). This modulates the function of K⁺ and Ca²⁺ ion channels, attenuates apoptosis and oxidative stress, and protects metabolic and mitochondrial functions (112). Finally, evidence suggests that H₂S signaling regulates NO and CO production (112, 116). The roles of H₂S in these signaling pathways in relation to cell death in cardiomyopathies, specifically DMCM, are elaborated on further in this review.

H₂S SIGNALING IN CARDIOMYOPATHIES

H₂S has been shown to have significant effects on the cardiovascular system. H₂S reduces hypertension (2, 3, 136, 162) and improves glucose uptake and metabolism in cardiomyocytes (9, 89). Furthermore, multiple studies have shown that free H₂S levels are reduced in patients with congestive heart failure (35, 117). Inhibition of H₂S synthesis by knockout of CSE increases leukocyte adherence in vessels, suggesting a role for H₂S in atherosclerosis development (184). Regarding DMCM specifically, plasma H₂S levels are decreased in T2DM patients, DM rats, and high-fat, diet-fed mice (9, 64, 66). H₂S donor administration prevents cardiomyocyte hypertrophy and increases ejection fraction in DM rats and mice (142, 194). H₂S supplementation mitigates hyperglycemia-induced cardiomyocyte dysfunction, as demonstrated in multiple studies (72, 168). Furthermore, a protective effect of H₂S has been shown in other cardiomyopathy models, including doxorubicin-induced cardiomyopathy, smoking-induced cardiomyopathy, and ischemia-reperfusion injury (I/R)-induced cardiomyopathy (12, 81, 90, 193). Together, these findings suggest that H₂S supplementation may act as a potential therapy for DMCM and other cardiomyopathies

H₂S IN CARDIOVASCULAR CELL DEATH MECHANISMS

Cell death mechanisms are a hallmark of the progression of all cardiomyopathies, including DMCM. Myocyte, endothelial, and fibroblast cell death can occur by necrosis, deregulated autophagy, or apoptosis. Each of these cell death pathways has been shown in failing human DM hearts (11, 29). In type 1 DM (T1DM) rats and STZ-induced T1DM mice, myocardial apoptosis was observed starting on day 3 after DM induction, suggesting that it occurs early in the progression of the disease (27). Recent reports have also shown a role of pyroptosis in cardiomyopathies (92). H₂S signaling has been implicated in these processes within the heart, and thus it may help regulate cell death in cardiomyopathy (summarized in Table 1).

Table 1.

H₂S regulation of cell death in multiple organ systems

Type of Cell Death (Organ/Disease)	Effect of H2S	Source (Ref. Nos.)
Apoptosis
Heart/ischemia-reperfusion	Anti-apoptotic, ↑Bcl-2, Bcl-XL, ↓miR-1	47, 61, 75
Heart/hyperglycemia	Anti-apoptotic, ↓p-JNK, p-cJun, NF-κB, ROS, ↑ PI3K/Akt, Nrf2	4, 72, 158
Heart/smoke induced	Anti-apoptotic	193
Neurons	Anti-apoptotic, ↑miR-485–5p	18
Brain/traumatic injury, Parkinson’s	Anti-apoptotic, ↑Bcl-2, ↑p62	48, 185
Gastric epithelium/ischemia-reperfusion	Anti-apoptotic, Keap1 S-sulfhydration, ↑ Nrf2 activation, ↓ NF-κB activation	38
Liver/ischemia-reperfusion	Anti-apoptotic, ↓ROS	50, 57
Aorta	↑ Apoptosis, ↑ MAPK	173
Autophagy
Brain/traumatic injury	Anti-autophagic, ↓LC3-II, ↑p62	185
Aorta/diabetes	Anti-autophagic, ↑ Nrf2 activation, ↓ROS	89
Heart/diabetes	Anti-autophagic, ↑PI3K/Akt1	166
Heart/ischemia-reperfusion	Anti-autophagic, ↑ mTOR	165
Liver/NAFLD	↑ Autophagy, AMPK activation, ↑ TG clearance	141
Pyroptosis
Bone derived macrophages	Antipyroptotic, ↓caspase 1 activity, ↓ chemokines, ↓ ROS	14
Retina	Antipyroptotic, ↓ER stress	34
Aortic endothelium	Antipyroptotic, ↓caspase 1 expression	164
Necroptosis
Heart/infarction	Antinecrotic	33
Heart/hyperglycemia	Antinecroptotic	87

AMPK, AMP-activated protein kinase; ER, endoplasmic reticulum; H₂S, hydrogen sulfide; Keap1, Kelch-like ECH-associated protein 1; MAPK, mitogen-activated protein kinase; mTOR, mammalian target of rapamycin; NAFLD, nonalcoholic fatty liver disease; Nrf2, nuclear factor-E2-related factor; PI3K, phosphatidylinositol 3-kinase; ROS, reactive oxygen species; TG, triglyceride. ↑, Increase; ↓, decrease.

APOPTOSIS

Apoptosis is a regulated, noninflammatory cell death pathway that utilizes initiator and effector caspases to cleave cellular and nuclear components. It results in the formation of apoptotic bodies and maintenance of intact plasma membrane. Mitochondrial membrane proteins such as Bcl-2, which inhibits apoptosis, and Bax, which activates apoptosis, regulate it. Perturbations in both the intracellular and extracellular environment can induce apoptosis (30). Intrinsic apoptosis is initiated by several factors, including DNA damage, endoplasmic reticulum (ER) stress, and intracellular ROS (114). These stressors lead to irreversible permeabilization of the outer membrane of mitochondria, which leads to the release of mitochondrial proteins such as cytochrome c, which activates caspase-9 and caspase-3 (82). Alterations in the extracellular environment such as hyperglycemia and inflammation activate death receptors such as the TNF receptor that ultimately activate caspase-8, which in turn activates caspase-3 (30).

Apoptosis has been demonstrated extensively in hypertrophic cardiomyopathy, ischemic cardiomyopathy, myocarditis, and dilated cardiomyopathy (36, 73, 128, 161, 183). Cardiomyocyte stretching (hypertrophy) activates many pathways that in turn initiate apoptosis (43, 62, 130). High glucose concentrations induce apoptosis in cardiomyocytes, and apoptosis has been confirmed in pathological specimens of patients with DMCM (11, 181). In addition, high concentrations of ROS, circulating and local inflammatory cytokines, ER stress, and activation of the renin angiotensin system (RAS) are known activators of caspases that mediate apoptosis, and these all are present in DMCM (10). Interestingly, nitrosylation of caspases by NO have been shown to activate apoptosis in DMCM, suggesting that sulfhydration by H₂S also plays an important regulatory role (120).

In fact, in a rat cigarette smoking-induced cardiomyopathy model, H₂S supplementation decreased apoptosis, as measured by increased Bcl-2 expression and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining. Pig and rat models of ischemia-reperfusion injury showed that H₂S treatment reduced apoptosis (109). A study by Shi et al. (138) demonstrated reduced apoptosis in cells treated with H₂S during I/R and was mediated by inhibiting JNK kinase, which is required for cytochrome c release into the cytosol. The role of H₂S on apoptosis may be organ dependent or context dependent, as studies show that H₂S prevents apoptosis in leukocytes but induces apoptosis in aorta smooth muscle cells (124, 175).

AUTOPHAGY

Autophagy is a cellular homeostatic mechanism for clearing damaged or unnecessary organelles and macronutrients by lysosomal degradation (167). Autophagy is a critical cellular process, playing both normal and pathological roles in the cardiovascular system. First described in the failing heart, it has been shown to be involved all cardiomyopathies (76). An optimal window of autophagy maintains homeostasis in response to cardiovascular events while excessive or deficient autophagy becomes maladaptive, leading to increased cell death. In ischemia-reperfusion injury, cardiomyocyte autophagy is triggered by increased afterload with both adaptive and maladaptive features (76, 79, 127). Patient biopsies from dilated cardiomyopathy showed high levels of autophagy when afterload was high (68). Increased autophagy is also observed in doxorubicin-induced cardiomyopathy (94, 95).

Autophagy is adapted to meet myocardial metabolic demand by releasing cellular components and employs a signaling network involving AMPK, which induces autophagy, and insulin pathways for regulation, making it an essential process in DM (20, 23). The nutrient-sensing mTOR complex 1 (mTORC1) integrates metabolic signals from AMPK and insulin, and when activated, mTOR inhibits the initiation of autophagy. Several studies demonstrate AMPK activation ameliorates DM-related consequences in hyperglycemic cardiomyocytes. However, in DMCM, a consensus on the role of autophagy has not been reached (20). In T1DM, some studies show increased autophagy in cardiomyocytes, whereas others show decreased autophagy (60, 169). Similarly, in T2DM mice, some studies show reduced autophagic markers, whereas others show increased autophagy (44, 60, 104). These differences may be due to differences in the model of DM used, the time of measurement, the method of measurement of autophagic flux, and glucose concentrations.

Oxidative and nitrosative mechanisms have been shown to regulate autophagy in cardiomyopathies, and recent evidence shows the importance of H₂S signaling in this process (153). In smoking-induced cardiomyopathy, H₂S administration decreased autophagy (193). A downregulation in autophagy markers with H₂S was also seen in an ischemia-reperfusion model (165). A H₂S donor has been shown to protect against cell death in high-glucose-treated cardiomyocytes via activation of AMPK. However, this study did not test whether the autophagic flux was also altered (158). Nitric oxide signaling alters autophagy in cardiomyopathy through a combination of cytokine expression, mitochondrial oxidative stress signaling, protein nitration, MMP activation, epigenetic alterations, and activation of transcription factors (153). It is likely that H₂S employs similar pathways to alter autophagy. It is important to note that the effects of H₂S on autophagy appear to be dependent on the pathological condition and the model system, as H₂S activates AMPK in DMCM models (supposedly increasing autophagic flux) but activates mTOR to inhibit autophagy in I/R injury (165).

NECROSIS, NECROPTOSIS, AND MITOCHONDRIAL PERMEABILITY TRANSITION PORE-DRIVEN NECROSIS

Necrosis is a highly inflammatory form of cell death following severe injury, whereas necroptosis is a programmed form of necrosis that utilizes TNF receptor signaling and the caspase-independent RIPK1/RIPK3 necrosome (26, 151). Myocardial cell death by necrosis occurs after a severe injury such as ischemia. Necrosis has been observed in the autopsy of T2DM patient hearts and T1DM rat and mouse hearts (11, 29). An increase in necroptosis has been shown in the heart after ischemic insults (186). The evidence showing an effect of H₂S on necrosis or necroptosis in cardiomyopathies is limited. A study by Geng et al. (33) demonstrated a decrease in necrosis markers after injection of NaHS. This decrease was measured by creatine phosphokinase and lactate dehydrogenase leakage into plasma and histology in a rat infarction model. Two more recent studies with cardiomyocytes treated with high glucose demonstrate an inhibition of necrosis and necroptosis by H₂S (86, 87). Taken together, these studies suggest an anti-necroptotic effect by H₂S, although the mechanism by which this occurs remains unclear.

Mitochondrial permeability transition (MPT) pore-driven necrosis is another regulated cell death pathway that ends with necrosis. It is initiated specifically by extremely high ROS and Ca²⁺ overload in the cytosol (30). These stressors lead to opening of the nonspecific MPT pore in the mitochondrial inner membrane, dissipate inner membrane potential, and rupture both mitochondrial membranes through osmotic swelling (30). Whereas loss of the outer membrane activates cytochrome c and caspase-3 similarly to apoptosis, the lack of ATP prevents formation of apoptotic bodies, and cell death occurs by necrosis instead (24). Hyperglycemia in DMCM induces MPT-driven necrosis through advanced glycosylation end products that elevate cytosolic ROS production (25). DMCM rodents also show impaired calcium uptake by mitochondria and elevated cytosolic calcium, which induces MPT pore opening (28). Studies show that H₂S protects against MPT-driven necrosis in the heart and brain (91, 182), although its role in MPT-driven necrosis in DMCM is unclear. In I/R-induced cardiomyopathy, however, whereas H₂S had no effect on MPT pore opening when applied directly to isolated mitochondria, it prevented MPT pore opening in the disease model (179).

PYROPTOSIS

Pyroptosis is a unique form of cell death in which the canonical pathway is specifically dependent on caspase-1 and the noncanonical pathway is reliant on caspase-11 (7). This cell death mechanism has several major characteristics that set it aside from apoptosis and necrosis. Originally discovered in monocytes and macrophages, pyroptosis has undergone a paradigm shift from being involved in immune response to taking part in pathological inflammation that contributes to diseases that are affiliated with inflammation and cell death (99). The rapid swelling and lysis of the cell membrane that result in the release of the inflammatory cytokines interleukin-1 β (IL-1β) and IL-18 characterize pyroptosis. This inflammatory lysis is in stark contrast to the organized membrane blebbing and chromosome packaging that occurs during apoptosis (26). Pyroptosis and its downstream inflammatory markers have been linked to pathogen defense, toxic shock, neuropathy, and nephropathy (85, 137, 176). However, we will focus on the role of pyroptosis in cardiovascular disease and DMCM.

The NOD-like receptor protein 3 (NLRP3) inflammasome is charged with activating caspase-1 in the canonical pyroptosis pathway. Blocking the inflammasome has been shown to mitigate cardiac remodeling post-infarction (96). NLRP3 also localizes to the mitochondria and contributes to increased reactive oxygen species (ROS) production in cardiomyocytes, which is linked to cardiomyopathy (190). IL-18 has been shown to induce cardiac hypertrophy through increasing the expression of atrial natriuretic peptide (ANP) (97). It is elevated in the serum and smooth muscle of infarction mouse models, and neutralizing IL-18 before ischemia significantly reduced infarct size (146). Elevated serum levels of IL-1β, a second inflammatory cytokine downstream of the pyroptosis pathway, has also been reported post-MI and in patients with cardiomyopathy (37). Although many inflammatory cytokines are elevated in cardiomyopathies, activation of IL-18 and IL-1β is unique to pyroptosis.

There is limited knowledge on the role of pyroptosis in DMCM. Pyroptosis may be triggered directly and indirectly in the DM heart. Knockout of NLRP3 that initiates pyroptosis improves glucose sensitivity in high-fat diet-fed mice (150). NLRP3-dependent pyroptosis is observed in the DM rat heart, and it contributes to DMCM (92, 93, 150). Palmate, a fatty acid upregulated in DM, also triggers the formation of the inflammasome (21). In both humans and mice, hyperglycemia and lipotoxicity upregulate cardiac NLRP3, caspase-1, and IL-1β (136, 150). Importantly, rosuvastatin treatment alleviated DMCM by inhibiting the NLRP3 inflammasome in T2DM rats (94). Targeting ELAVL1, an RNA binding protein, with miRNA-9 inhibits hyperglycemia-induced pyroptosis in human ventricular cardiomyocytes (56). In a clinical trial of 551 T2DM patients, canakinumab, an IL-1β inhibitor, resulted in a small decrease in Hb A_1c; however, its effect on cardiac function is unexplored (45).

Several hypotheses have been proposed for how DM induces NLRP3 inflammasome activation and its association with H₂S. One hypothesis is that DM increases thioredoxin-interacting protein expression and insulin resistance, causing cellular oxidative/nitrosative stress, leading to NLRP3 inflammasome activation and subsequent cardiomyocyte pyroptosis (92, 139). Because of similarities between nitric oxide and H₂S, a role for sulfide signaling is suggested in pyroptosis. Another hypothesis is that, reduced CBS and CSE in the DM heart decreases H₂S generation and increases homocysteine accumulation, causing hyperhomocysteinemia (133). Hyperhomocysteinemia activates matrix metalloproteinase-9 (MMP9) in the DM heart. Notably, MMP9 ablation in the DM heart decreases pyroptosis markers in the heart (102, 171). This link suggests a vital role of CBS, CSE, H₂S signaling, and MMP9 in the regulation of pyroptosis in cardiomyopathies, especially in DMCM.

A limited number of studies have shown a role of H₂S signaling in pyroptosis in other organs and in pathological conditions. In mouse bone-derived macrophages stimulated with uric acid crystals, H₂S donors inhibited caspase-1 activity and IL-1β secretion usually seen after stimulation by the crystals (14). Treatment with GYY4137, a H₂S donor, mitigates endoplasmic reticulum (ER) stress and pyroptosis and improves visual function in the CBS^+/− mice (34). CBS^−/− mice, a model of severe hyperhomocysteinemia, induces caspase-1 activation along with other pyroptosis markers determined by flow cytometry gating in isolated lung endothelial cells and in aortic endothelium (164). To date, only one study has reported a role for H₂S in pyroptosis in the heart. Toldo et al. (147) demonstrated that an H₂S donor suppressed several markers of pyroptosis activity (caspase-1 activity, caspase-1 recruitment via the ASC adapter protein) in an I/R cardiomyopathy model. The role of pyroptosis in cardiomyopathies and the mechanistic role of sulfide signaling in pyroptosis require further study.

Of note, whereas most studies show that H₂S mediates suppression of aberrant cell death, some studies show deleterious effects of H₂S, such as a study by Yang et al. (173) that demonstrated increased apoptosis after the addition of H₂S gas to human aorta smooth muscle cells in culture. The study by Yang et al. (173) used a relatively high concentration of 500 µM using pure gas instead of a H₂S donor. However, variability in effects of H₂S is likely due to varying concentrations and active forms of H₂S used. Measured physiological H₂S concentrations range from nanomolar to micromolar concentrations, which is likely due to the variety of methods used to measure H₂S (40, 134).

POTENTIAL MECHANISMS OF H₂S-MEDIATED PROTECTION AGAINST CELL DEATH

Antioxidant Mechanisms

Oxidative damage from reactive oxygen species (ROS) is an established trigger for cell death in cardiomyopathies. Increased ROS production and decreased antioxidant capacity are observed in patients and animal models of DMCM, doxorubicin-induced cardiomyopathy, and congestive heart failure (22, 63, 160). Higher H₂O₂ induces apoptosis, whereas lower myocardial oxidative stress by a reduction in NADPH oxidase activity reduces apoptosis and blunts hypertrophy (129). Although there is a lack of data connecting DMCM, pyroptosis, and ROS, a transverse aortic constriction (TAC)-induced hypertrophy model showed that ROS levels were increased along with increased NLRP3 and IL-1B (157). This increase in pyroptosis and ROS was also observed in an I/R injury model (154). Taken together, there is a clear connection with ROS and apoptosis in DMCM as well as ROS and pyroptosis in other models of cardiomyopathy.

H₂S appears to signal changes in ROS production and antioxidant pathways, potentially connecting H₂S levels with the activation of apoptosis and pyroptosis by way of ROS signaling. The incubation of HeLa cells with H₂O₂, ${\text{O}}_2^{–}$, and ONOO⁻ resulted in the release of H₂S, suggesting a homeostatic feedback mechanism (189). Incubation of cardiomyocytes with H₂S led to a dose-dependent decrease in ROS and inhibition of caspase-3-dependent apoptosis in high-glucose-treated cardiomyocytes (72). Moreover, H₂S treatment reversed decreased activity of superoxide dismutase and decreased autophagy in aortic endothelial cells from T2DM mice (89).

Several hypotheses have been proposed for the mechanism by which H₂S reduces ROS. H₂S may be a direct reducing agent, as it has been shown in endothelial cells to reduce hydrogen peroxide and oxidized lipoproteins levels (131). H₂S may also impart antioxidant actions by an indirect or feedback mechanism (65). Many studies now demonstrate that H₂S drives the downstream regulation of canonical transcription factors that regulate oxidative damage and antioxidant genes, including FOXO3, NF-κB, and Nrf2 (116).

Noncoding RNA

Cell death mechanisms are some of the central pathways regulated by miRNA. miRNAs are ∼22 nucleotide, noncoding RNA molecules that negatively regulate target gene by mRNA degradation or translational repression (101). Several miRNAs have been implicated in cell death during cardiomyopathies (101). miR-195 overexpression induces cardiac hypertrophy and promotes ROS generation and apoptosis and suppresses Bcl-2 in cardiomyocytes (152, 195). miR-1 and miR-144 increase apoptosis in high-glucose-treated cardiomyocytes (180, 181). Increased miR-30d was shown to increase pyroptosis in diabetic rats (84).

Cross-talk between miRNA and sulfide signaling may mediate the beneficial cardiac effects of H₂S (40). Rats treated with H₂S donor before I/R injury showed downregulated miR-1 and decreased apoptosis compared with the untreated I/R group in the heart (61). Several studies also show that H₂S increases expression of miR-133a, which ameliorates cardiomyocyte hypertrophy and reduces apoptosis (64, 88, 123). miR-21 has anti-apoptotic effects in cardiomyocytes, and H₂S donor induces cardiomyocyte expression of miR-21 both in vitro and in vivo (19, 147). In addition, miRNAs also regulate H₂S production by regulating CSE, which could act as a feedback mechanism (174). Nevertheless, the list of miRNA known to be affected by H₂S are small compared with the list of miRNAs known to regulate cell death, leaving ample room for further investigation.

Cytokine Signaling

Cytokines are small proteins that provide both proinflammatory and anti-inflammatory signals to the cell (119). Increased cytokine production in serum and heart muscle is a common feature of cardiovascular diseases involving cell death (119). The proinflammatory cytokine TNFα has been shown to induce both apoptosis and necrosis. Other proinflammatory cytokines such as IL-6 and IL-1 induce TNFα production. In a doxorubicin-induced cardiomyopathy cell model, treatment with NaHS, a H₂S donor, reduced the expressions of IL-1β, IL-6, and TNFα, which was mediated by the p38 MAPK/NF-κB pathway (39, 168). Numerous studies have shown reduced cytokine production after H₂S administration in models of sepsis and lung injury (83, 125, 126). The inflammatory and fibrotic remodeling in cardiomyopathies is also mediated by inflammatory cells such as macrophages and Treg cells in the heart, which act upon cardiomyocytes (103).

Interestingly, TNFα stimulation in the liver and macrophages leads to increased H₂S production, suggesting the existence of compensatory signaling pathways (132). Alternately, two studies have shown that a decrease in cytokine production (IL-6 and TNFα) after H₂S treatment is associated with increased histone acetylation, suggesting a mechanism that involves epigenetic regulation (125, 126). This effect was shown in macrophages challenged with LPS, and it may occur in cardiac macrophages in cardiomyopathies.

In cardiomyopathy following I/R injury, CD11b⁺Gr-1⁺ neutrophils are recruited to the myocardium to express IL-1β and TNFα. H₂S donor treatment of mice reduced the presence of these myeloid cells, which was correlated with promoted Bcl-2 anti-apoptotic signaling, reduced cytokine release, and preserved cardiac function (187).

VEGF is a proangiogenic cytokine that promotes survival and growth of endothelial cells (78). In DMCM, reduced VEGF leads to rarefaction of microvessles in the heart and causes a redox imbalance resulting in ischemic cell death (13, 41). H₂S supplementation increases cardiac VEGF in myocardial infarcted mice and improves proliferation and migration of endothelial cells (111). It suggests a cardioprotective effect of H₂S on endothelial cells in cardiomyopathy via VEGF cytokine signaling (122).

TGFβ is a critical cytokine in the cardiac remodeling process during cardiomyopathies, and it induces the differentiation of fibroblasts into myofibroblasts (5). Myofibroblasts in turn promote maladaptive collagen deposition and secrete cytokines like TNFα, which induces cell death. The cardioprotective effects of H₂S on cell death appear to act via cell types other than cardiomyocytes, where it reduces TGFβ expression and inhibits its signaling cascade in fibroblasts (98, 188). This limits the differentiation and proliferation of fibroblasts into myofibroblasts and prevents collagen deposition in the heart.

Renin-Angiotensin System

The role of the renin-angiotensin system (RAS) in cardiovascular diseases is well established and far-reaching. The RAS regulates oxidative stress, hypertrophy, and neuronal control of cardiovascular function (149). Studies also show that RAS partly regulates cardiomyocyte cell death. Stretching myocytes releases angiotensin II (ANG II), increases p53 binding to the ANG II promotor as well as the AT₁ receptor, and results in a four to sevenfold increase in apoptosis (77). Increased apoptosis was also observed in cardiomyocytes with an adenoviral increase in p53 alone (113). Blockade of the AT₁ receptor with losartan abolished stretch-induced apoptosis (77).

Laggner et al. (74) hypothesized that H₂S interacts with angiotensin-converting enzyme (ACE), a zinc metalloproteinase, based on studies showing that H₂S reacts with sulfide groups on metalloproteins and acts as a potent vasodilator. Their study showed a dose-dependent decrease in ACE activity in human endothelial cells after treatment with H₂S. This decrease in ACE activity was without a decrease in ACE mRNA level, and it was counteracted by the addition of Zn²⁺. Another study showed that supplementation of H₂S in DM rats reversed RAS activation and reduced ROS production (170). These studies suggest that H₂S could alter RAS signaling, which may in turn reduce oxidative stress to mitigate cell death. A direct role of H₂S-mediated RAS signaling in cardiomyopathy cell death, however, has not been reported.

Regulation of Epigenetic Modifications

H₂S also signals through epigenetic changes. As mentioned above, H₂S induces chromatin remodeling to reduce cytokine production (125). Several studies have established the role of NAD-dependent deacetylase sirtuin 1 (SIRT1) in apoptosis of cardiomyocytes. Resveratrol activation of SIRT1 in cultured cardiomyocytes after hypoxia inhibits cell death detected by TUNEL staining (15). A study by Wu et al. (163) also demonstrated a similar SIRT1-dependent protection from apoptosis in cardiomyocytes after hydrogen peroxide oxidative damage. This study also showed that treatment by H₂S donor initiated this pathway by increasing the expression of SIRT1. One possible mechanism of this effect of H₂S on SIRT1 is the role of miRNA. miR-34a has been shown to target SIRT1, which leads to an increase in acetylated p53 and ultimately an increase in apoptosis (172). H₂S has been shown to regulate miR-34a in hepatocytes; however, this mechanism has not been shown in cardiomyocytes yet (50).

After myocardial infarction, FOXP3 transcription factor-positive CD4⁺ Treg cells mediate the resolution of inflammation, as opposed to maladaptive remodeling and fibrosis (159). H₂S has been shown to be crucial for the differentiation of these Treg cells (177). Furthermore, the authors found that H₂S drove the demethylation of Foxp3 DNA, which promotes FOXP3 expression by promoting the expression of the Ten eleven translocation (Tet) family of enzymes that reverse DNA methylation (177).

Regulation of Other Gaseous Signaling Molecules by H₂S

H₂S is one of several biologically crucial gaseous signaling molecules. Like H₂S, NO has been shown to have cardioprotective effects (59). Adding physiological levels of NO to cardiomyocytes or overexpressing endothelial NO synthase (eNOS) during I/R injury reduced oxidative stress and cell death (52). A similar study of cultured cardiomyocytes treated with doxorubicin demonstrated a reduction in apoptosis, especially after pretreatment with NO (95). In this study, caspase-3 activity decreased with NO in addition to S-nitrosylation of caspase-3.

Several studies show that H₂S regulates NO expression and activity, which suggests that H₂S may protect against cell death via NO. In a clinical trial of a H₂S prodrug in patients with heart failure, increased H₂S levels were correlated with an increase in NO (117). These effects appear to exist in cardiac cells other than cardiomyocytes, as H₂S can also increase NO production in endothelial cells (118). NO regulation by H₂S may be mediated by eNOS. H₂S treatment increases eNOS expression and activity (1, 70). Mice lacking H₂S show reduced NO and eNOS levels (66, 67).

The presence of cross-talk between H₂S and NO is further supported by studies showing that H₂S and NO can sulfhydrate or nitrosylate, respectively, the same cysteine residues of proteins (162). For example, nitrosylation of glyceraldehyde-3-phosphate dehydrogenase at Cys₁₅₀ abolishes its activity, whereas sulfhydration at this residue by H₂S increases it’s activity (42) Altaany et al. demonstrated that H₂S sulfydrated eNOS and inhibited its nitrosylation at the same cysteine residue (1). This competitive cysteine modification appears to have effects on cell death since NF-κB can be nitrosylated at Cys₃₈ to inhibit its activity and it can be sulfhydrated at the same residue to increase transcription of anti-apoptotic genes (132). These examples suggest the two gaseous molecules signal in concert and regulate each other, but the role of this interaction in cell death pathways and DMCM needs further investigation.

The signaling effects of H₂S discussed in the preceding sections, which mediate cardioprotection of cell death in cardiomyopathies, are summarized in Fig. 1.

Fig. 1.

Hydrogen sulfide (H₂S) cell death signaling by effects on oxidative stress, noncoding RNA, cytokines, and epigenetics. Proposed and experimentally demonstrated mechanisms that have been shown to be increased (↑) or decreased (↓) by H₂S on cell death pathways. ACE, angiotensin-converting enzyme; AMPK, AMP-activated protein kinase; ANG II, angiotensin II; Cys-SNO, S-nitroso-cysteine; Cys-SSH, cysteine persulfide; eNOS, endothelial nitric oxide synthase; mTORC1, mammalian target of rapamycin complex 1; NO, nitric oxide; RAS, renin angiotensin system; ROS, reactive oxygen species; SIRT1, sirtuin 1.

Regulation of Transcription Factors

Many of the antioxidant, cytokine, and cell death mechanisms involving H₂S are mediated by the effect of H₂S on transcription factors. The roles of FOXO3, NF-κB, and Nrf2 in cell death mechanisms are highlighted here and are depicted in Fig. 2 with potential pathways.

Fig. 2.

Role of transcription factors in hydrogen sulfide (H₂S) signaling. Proposed and experimentally demonstrated effects of H₂S on the transcription factors NF-κB (top), nuclear factor-E2-related factor (Nrf2; middle), and FOXO3 (bottom). Pathways that are empirically not proven but could potentially be important in cardiomyocytes are depicted with a question mark (?). H₂S can sulfhydrate the p65 subunit of NF-κB to promote its anti-apoptotic gene regulation. Sulfhydration of Kelch-like ECH-associated protein 1 (Keap1) releases its inhibition of Nrf2, allowing translocation of Nrf2 to the nucleus to increase expression of antioxidant genes. miRNA in extracellular vesicles released from fibroblasts regulate Nrf2 mRNA in cardiomyocytes, and that may interact with H₂S. H₂S may promote degradation of FOXO3 to limit proapoptotic effects of FOXO3. Reactive oxygen species (ROS) also induce antioxidant gene transcription via FOXO3, which may be promoted by H₂S. H, hydrogen; S, sulfur.

FOXO3 belongs to the O subclass of the forkhead family of transcription factors along with FOXO1, FOXO4, and FOXO6. FOXO3 is inhibited under normal conditions and translocated out of the nucleus on phosphorylation by Akt. During cell stress, it functions as a trigger for apoptosis through upregulation of proapoptotic genes such as Bim and PUMA or downregulation of anti-apoptotic proteins such as FLIP in cardiomyocytes (3). FOXO3a can also protect against oxidative stress by upregulating antioxidants, depending on the context of its induction (155).

H₂S has been shown to modulate the apoptotic or protective actions of FOXO3a. In doxorubicin-induced cardiomyopathy, NaHS treatment prevented the decrease in FOXO3 phosphorylation seen after doxorubicin treatment to maintain degradation of FOXO3 and to limit apoptosis (90). Moreover, FOXO3 has been implicated in the epigenetic effects of H₂S. For example, SIRT1 deacetylation of FOXO3 increases the protective anti-apoptotic activities of FOXO3 (9). Thus, the anti-apoptotic effects of H₂S in cardiomyocytes that were dependent on SIRT1 shown by Wu et al. (163) could be mediated by FOXO3. The cross-talk between H₂S targets and FOXO3 is thus an important pathway for H₂S-mediated cardioprotection.

The nuclear factor-E2-related factor (Nrf2) and Kelch-like ECH-associated protein 1 (Keap1) axis is a key transcriptional regulator of oxidative stress. Under oxidative stress conditions, Keap1 is degraded, and Nrf2 translocates to the nucleus to induce expression of antioxidant response elements. As highlighted above, oxidative stress plays a crucial role in cell death in cardiomyopathies, and numerous publications show that change in Nrf2 expression or function alters the progression of DMCM and other cardiomyopathies (17, 80, 157). For example, a study by Qin et al. (121) demonstrated that Nrf2-knockout increased myocardial necrosis. Nrf2-activating compounds inhibit NLRP3 inflammasome activation, and Nrf2 silencing abrogated the decrease in pyroptosis in vascular endothelial cells after antioxidant treatment (32, 49).

New evidence links H₂S signaling to Nrf2-mediated cell death. Two studies have shown that incubation of mouse fibroblasts with a H₂S donor directly sulfhydrated Keap1, which enhanced Nrf2 translocation and increased expression of antioxidant genes (46, 175). This protected against age-related cellular senescence and induced cytoprotective genes. Also, this H₂S-mediated stabilization of Nrf2 subsequently controlled the expression of H₂S-producing and -degrading enzymes. Although the effect of H₂S sulfhydration of Keap1 on apoptosis or pyroptosis is yet to be explored, it is assumed that cardioprotective effects of H₂S in cardiomyopathy may be mediated by this pathway.

The effects of H₂S on Nrf2 may be more complex than just a direct sulfhydration effect. A recent study demonstrated that miRNA-28a, miR-27a, and miR-34a is secreted in exosomes from fibroblasts and absorbed by cardiomyocytes to decrease Nrf2 translation (145). Interestingly, in hepatic I/R injury, rats injected with NaHS attenuated miR-34a and upregulated Nrf2 to prevent I/R injury (50). These findings suggest a mechanism where H₂S either directly regulates Nrf2 or indirectly regulates it through miRNA to reduce oxidative stress and cell death. Investigating these mechanisms in cardiomyopathies could reveal a promising mechanism of H₂S signaling in the heart.

NF-κB is another vital transcription factor activated in response to cell stress. Typically, phosphorylation of the inhibitor-κB unmasks the p65 subunit of NF-κB, which then translocates to the nucleus to promote transcription of inflammatory and apoptotic genes (135, 148). H₂S, on the other hand, has been shown to activate protective anti-apoptotic effects on NF-κB (132). In this pathway, following TNFα stimulation, CSE expression is increased to generate H₂S, which sulfhydrates p65 (132). Sulfhydrated p65 translocates to the nucleus in liver macrophage cells and increases the transcription of anti-apoptotic gene targets TRAF, cIAP-2, Bcl-XL, and XIAP (132). Ablation of CSE and reduction of H₂S abolished these beneficial effects. NF-κB-mediated proapoptotic effects have been shown in models of DMCM and doxorubicin-induced cardiomyopathy (81, 135, 148). The role of p65 sulfhydration by H₂S in cardiomyocytes is unclear. However, a similar anti-apoptotic activation upon sulfhydration as seen in the liver may be the signaling pathway in DMCM.

THERAPEUTIC POTENTIAL OF H₂S IN CVD

Because of the cardioprotective benefits of H₂S seen in cardiomyocytes in culture and animal models, a novel long-acting prodrug of H₂S (SG1002) has been developed. SG1002 has been tested in preclinical models of cardiovascular disease and in clinical trials (2, 117). In an animal model of systolic heart failure, SG1002 protects against adverse remodeling and heart failure via upregulation of endothelial nitric oxide synthase (117). It also prevented QT interval prolongation in mice with ischemic cardiomyopathy, suggesting a decrease in cell death (2). In phase I clinical trials, SG1002 reduced BNP and increased NO in heart failure patients (115, 117). Other novel methodologies of H₂S delivery are under development, such as organ-targeted delivery by nanoparticles, microbubbles, and ultrasound (16, 156). H₂S encapsulated in microbubbles and released in the myocardium with ultrasound after I/R injury in rats reduced apoptosis and oxidative stress (16). The molecular and preclinical work highlighted in this review suggests that H₂S-based therapies could prevent cardiomyocyte cell death; however, it remains to be seen whether these therapies can mitigate the progression of DMCM in patients. In addition, all studies highlighted in this review used prevention strategies to demonstrate that H₂S protected against cell death and subsequent remodeling in cardiomyopathy. It has yet to be seen whether H₂S can reverse pathological remodeling.

FUTURE DIRECTIONS

With multiple cell death mechanisms shown to be present in DMCM, the relative contribution of each mechanism is important to address (Fig. 3). It is unclear whether apoptosis, necrosis, pyroptosis, and autophagy work in concert simultaneously, if one is predominant over another, or if one pathway affects the other. The predominant cell death mechanism may be dependent on the type of stress and thus the etiology of the cardiomyopathy or on the time course of disease progression (26). Another possibility is that one cell death process may occur before another in cardiomyocytes, raising the opportunity of discovering early biomarkers for cardiomyopathy based on the earlier cell death process.

Fig. 3.

Future directions for the role of hydrogen sulfide (H₂S) in protection against cell death signaling and diabetic cardiomyopathy. Both type 1 (T1DM) and type 2 diabetes mellitus (T2DM) lead to cardiomyopathy and cell death independent of other cardiovascular effects on the heart. H₂S may mediate signaling that prevents cell death pathways; however, many of these effects have not been proven in cardiomyocytes. H₂S and its signaling effects may also show beneficial therapeutic and preventative effects in diabetic cardiomyopathy (DMCM).

In general, suppression of one cell death pathway leads to compensation by another cell death pathway (153). For instance, inactivation of Beclin 1 protein, which is necessary for autophagy, increases apoptotic bodies in many organisms (144). Beclin 1 has a complex functional and structural interaction with anti-apoptotic protein Bcl-2, and apoptotic signals lead to caspase-mediated cleavage of Beclin 1, thereby inhibiting autophagy and vice versa (108, 192). However, there is also evidence suggesting that targeting one specific pathway may be sufficient to prevent the development of cardiomyopathy. For example, inhibiting NLRP3 gene expression in T2DM rats ameliorates DMCM (93), and inhibition of autophagy prevents hyperglycemia-induced apoptosis in cardiomyocytes (68).

H₂S may also suppress or regulate multiple cell death pathways simultaneously. H₂S has been shown to alter expression of both Bcl-2 and Beclin 1 (185). Similarly, Toldo et al. (147) demonstrated that H₂S administration during I/R cardiomyopathy suppressed pyroptosis, apoptosis, and necrosis. However, it is unclear how H₂S could regulate multiple cell death pathways together. One possibility is that H₂S could regulate both apoptosis and pyroptosis simultaneously via cross-talk with autophagy. Autophagy has been shown to inhibit pyroptosis, as activation of autophagy can clear damaged mitochondria that activates pyroptosis (108) It will be essential to understand the temporal and spatial relationship of these cell death pathways in DMCM, the cross-talk between autophagy and cell death, and their relationship to an upstream signaling molecule such as H₂S. If H₂S can regulate both apoptosis, necroptosis, and pyroptosis simultaneously in diseases such as DMCM, future studies will be required to compare the quantitatively effects of H₂S on each pathway to determine whether there is a relative gradient of cell death protection.

Although clinical trials based on H₂S have already begun, many unanswered questions remain, particularly relating to its signaling mechanisms. As highlighted in this review, many gaps in understanding exist. It is unclear how H₂S interacts transcription factors such as Nrf2 and NF-κB and how these pathways work in concert, especially in cardiomyocytes and cardiomyopathies. Bioinformatics such as next-generation sequencing are likely valuable unbiased approaches to study the numerous pathways that H₂S affects. Because H₂S is difficult to administer in optimum doses in vivo, understanding its specific signaling targets may improve existing therapies or uncover new therapies for cardiomyopathy using the H₂S cardioprotection pathway.

There are also unexplored avenues regarding the role of noncoding RNAs in H₂S signaling. Although studies have shown that H₂S regulates miRNA expression and vice versa, the underlying mechanisms of this cross-talk are unknown. Long noncoding RNA (lncRNAs) are a class of regulatory RNAs that are >200 nucleotide in length. They can regulate miRNA expression, hypertrophy, and cell death (67). For example, lncRNA H19 has been found to be both increased and decreased in different models of DM mice hearts, and overexpression of H19 in diabetic rats decreased oxidative stress and apoptosis (59, 79). Another lncRNA, MALAT1, increases cardiomyocyte apoptosis in DM rats and is overexpressed in dilated cardiomyopathy (27, 164). lncRNA MEG3 is associated with inflammasome formation and pyroptosis, albeit in aortic endothelial cells (168). Similarly, lncRNA Kcnq1ot1 promotes pyroptosis in DCM by binding to miR-214-3p (154). To our knowledge, there is no study testing whether H₂S affects lncRNA expression. However, these recent studies showing that lncRNA regulates cardiomyocyte cell death provide a rationale for determining whether H₂S could modulate miRNA expression via lncRNA.

Questions also remain as to how H₂S levels are reduced in disease states such as DM. One hypothesis is based on homocysteine, a precursor of H₂S. Studies in some mouse models of T1DM, and T1DM and T2DM patients with renal dysfunction show elevated levels of plasma homocysteine, which is strongly associated with cardiovascular disease, including DMCM (53, 100, 107, 133). The homocysteine metabolic enzymes CBS, cystathionine γ-lyase (CSE), and methyl tetrahydrofolate reductase (MTHFR) are all downregulated in a mouse model of T1DM, leading to the buildup of homocysteine (100). Elevated homocysteine also competes with cysteine for binding to CSE, thus decreasing the production of H₂S through substrate inhibition (140). Hyperhomocysteinemia also decreases expression of CSE, leading to decreased H₂S generation (33, 178). Finally, homocysteine appears to directly precipitate proteins post-translationally by homocysteinylation (55), and it is possible that DM-induced hyperhomocysteinemia may precipitate CSE, resulting in decreased H₂S production. This homocysteine/H₂S hypothesis and/or whether DM exerts other effects on H₂S metabolism requires further investigation.

H₂S is also difficult to measure in vivo due to its volatility and different reactive forms. Measurement of H₂S in different matrices using different methods produces highly variable results (51, 134). A consensus method for H₂S measurement is necessary to reduce the variabilities in the different H₂S-based experiments. Understanding the mechanisms of H₂S may uncover diagnostic markers of cardiomyopathies and biomarkers of H₂S-based therapies. The current gold standard for diagnosis of DMCM uses noninvasive two-dimensional and three-dimensional echocardiography and magnetic resonance imaging. However, this is problematic when early progression is asymptomatic. For example, in DMCM, diastolic dysfunction is the first clinical sign of the disease, but the heterogeneity of disease progression makes identification of diastolic dysfunction difficult, necessitating the discovery of molecular biomarkers (110). Based on the cell death mechanisms of DMCM, many molecules have been studied for potential biomarkers in DMCM when this disease is still reversible. For example, cardiotrophin-1 (CT-1) promotes fibrosis and inhibits apoptosis. Although CT-1 is elevated in plasma of patients with DMCM, it is also highly expressed in other tissues and in other cardiomyopathies (31, 110). Many of the H₂S signaling pathways and its diverse effects on cell death highlighted in this review are yet to be explored, and they may reveal potential biomarkers and therapeutic targets of cardiomyopathies, including DMCM.

GRANTS

This work is supported in part by the National Heart, Lung, and Blood Institute Grants HL-113281 and HL-116205 (to P. K. Mishra).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.K. and P.K.M. conceived and designed review; S.K. prepared figures; S.K., T.N.K., and P.K.M. drafted manuscript; S.K., T.N.K., and P.K.M. edited and revised manuscript; S.K., T.N.K., and P.K.M. approved final version of manuscript.