Research ArticleCardiovascular Biology

Depolysulfidation of Drp1 induced by low-dose methylmercury exposure increases cardiac vulnerability to hemodynamic overload

See allHide authors and affiliations

Science Signaling  25 Jun 2019:
Vol. 12, Issue 587, eaaw1920
DOI: 10.1126/scisignal.aaw1920

Methylmercury makes hearts more fragile

Although the widespread environmental contaminant methylmercury is largely associated with neurotoxic effects, it is also associated with increased risk for cardiovascular disease. Nishimura et al. found that mice exposed to a dose of methylmercury that was too low to cause neurotoxicity were more vulnerable to heart failure in response to pressure overload. Methylmercury removed a polysulfide group from Drp1, thereby removing an inhibitory brake on this protein, which resulted in increased Drp1-mediated mitochondrial fission. Treating mice or human cardiomyocytes with a polysulfide group–releasing compound reversed fragility to mechanical overload induced by methylmercury. These results provide a molecular mechanism for the cardiotoxic effects of methylmercury and a possible strategy to avert these effects.

Abstract

Chronic exposure to methylmercury (MeHg), an environmental electrophilic pollutant, reportedly increases the risk of human cardiac events. We report that exposure to a low, non-neurotoxic dose of MeHg precipitated heart failure induced by pressure overload in mice. Exposure to MeHg at 10 ppm did not induce weight loss typical of higher doses but caused mitochondrial hyperfission in myocardium through the activation of Drp1 by its guanine nucleotide exchange factor filamin-A. Treatment of neonatal rat cardiomyocytes with cilnidipine, an inhibitor of the interaction between Drp1 and filamin-A, suppressed mitochondrial hyperfission caused by low-dose MeHg exposure. Modification of cysteine residues in proteins with polysulfides is important for redox signaling and mitochondrial homeostasis in mammalian cells. We found that MeHg targeted rat Drp1 at Cys624, a redox-sensitive residue whose SH side chain forms a bulky and nucleophilic polysulfide (Cys624-S(n)H). MeHg exposure induced the depolysulfidation of Cys624-S(n)H in Drp1, which led to filamin-dependent activation of Drp1 and mitochondrial hyperfission. Treatment with NaHS, which acts as a donor for reactive polysulfides, reversed MeHg-evoked Drp1 depolysulfidation and vulnerability to mechanical load in rodent and human cardiomyocytes and mouse hearts. These results suggest that depolysulfidation of Drp1 at Cys624-S(n)H by low-dose MeHg increases cardiac fragility to mechanical load through filamin-dependent mitochondrial hyperfission.

INTRODUCTION

Intracellular thiol-containing molecules, especially cysteine (Cys) thiol groups, have unique nucleophilic properties that allow them to easily react with oxidants and electrophiles. Reduction/oxidation (redox)–dependent posttranslational modifications of Cys in proteins play crucial roles in various physiological and pathophysiological processes (1). Cys polysulfides (Cys-SS(n)H), such as Cys persulfides (CysSSH) and Cys trisulfides (CysSSSH) whose sulfhydryl groups are covalently bound to sulfur (sulfane sulfur), are formed in proteins during translational and posttranslational processes (2, 3). Low–molecular weight persulfides are abundant and exist endogenously in organisms. Moreover, low–molecular weight persulfides react with oxidized Cys thiol of various proteins to form polysulfidation. These polysulfidation events can modulate protein functions (4). Because Cys-SS(n)H have higher nucleophilicity (pKa = 4.3) than Cys thiol (pKa = 8.29) or hydrogen sulfide (pKa = 6.8) (5), low–molecular weight polysulfides and polysulfidated proteins preferentially react with and scavenge electrophiles and oxidants, which contributes to the negative regulation of redox signaling as well as mitochondrial biogenesis and bioenergetics (2, 3).

Methylmercury (MeHg) is an electrophilic environmental neurotoxicant that is ubiquitously detected in seawater (0.2 to 1 parts per thousand) and is biologically concentrated into seafood (6), particularly large predatory fish like swordfish, which contain relatively high concentrations of MeHg [0.49 to 2.6 parts per million (ppm)] (7). Several epidemiological studies and in vivo animal studies have established the neurotoxic actions of MeHg (810). High-dose MeHg exposure can induce covalent modification of protein sulfhydryl groups to form MeHg–S-protein complex (S-mercuration). This pathological protein modification disrupts enzyme functions and, in part, explains the neurotoxicity of MeHg (1114). Furthermore, epidemiological studies have also suggested MeHg as a potent risk factor for cardiac events (1517). Mercury increases cardiac risk at a lower concentration than that associated with neurotoxicity (18), although the underlying molecular mechanism is obscure.

Structural and morphological changes of the heart (known as remodeling) are a clinical outcome of chronic heart failure (19). When the heart is exposed to hemodynamic overload, it will increase cardiomyocyte size, a phenomenon called adaptive hypertrophy. However, if mechanical stress on the heart is sustained, then the heart will gradually undergo maladaptive hypertrophy, including fibrosis and senescence, as well as decrease in left ventricular (LV) function. A growing body of evidence has suggested the involvement of abnormal mitochondrial dynamics in cardiac transition from adaptive to maladaptive hypertrophy induced by hemodynamic overload (20). Mitochondrial dynamics are regulated by the balance between fusion factors including optic atrophy 1 (Opa1), mitofusin 1 (Mfn1), and Mfn2 and the fission factor dynamin-related protein 1 (Drp1) (21). Genetic aberration of these factors results in cardiac dysfunction (2224). We have reported that hypoxic stress induced by myocardial infarction promotes the interaction of Drp1 with actin-binding protein filamin-A (FLNa), which acts as a guanine nucleotide exchange factor for Drp1. The resulting Drp1-mediated mitochondrial hyperfission triggers myocardial premature senescence and cardiac dysfunction in mice (25).

Drp1 is redox-sensitive, and its function is regulated by Cys-based redox modifications such as S-nitrosylation (26) and sulfenylation (27). We have previously found that Drp1 is polysulfidated in the basal state, and its polysulfidation negatively regulates guanosine 5′-triphosphate (GTP)–binding activity (3). This evidence suggests that Drp1 polysulfidation and activity could be affected by electrophiles. MeHg readily reacts with polysulfides and deprives sulfur atom from polysulfides, generating nonelectrophilic bismethylmercury sulfide (MeHg)2S (28). We thus speculated that MeHg increases risk of cardiac events by affecting myocardial Drp1 activity and mitochondrial dynamics. In this study, we revealed that MeHg promoted depolysulfidation of Drp1 at Cys624, leading to Drp1 activation and mitochondrial hyperfission. These findings provide a better understanding of the molecular basis of mitochondrial hyperfission by low-dose MeHg and the associated cardiac fragility in response to hemodynamic overload.

RESULTS

Exposure to subneurotoxic low-dose MeHg exacerbates cardiac dysfunction induced by pressure overload

We first examined whether a neurotoxic dose (100 ppm) of MeHg (high MeHg) caused cardiac abnormalities in mice. Exposure to high MeHg for 2 weeks resulted in decreased body weight (fig. S1A). In contrast, high MeHg exposure had little impact on heart weight (fig. S1B) and LV geometries (table S1). Fasting blood glucose levels were similar in vehicle- and MeHg-treated mice. The rate of decay of blood glucose after intraperitoneal glucose injection in MeHg-exposed mice tended to be slower than that in vehicle-treated mice, but this difference was not statistically significant (fig. S1C). The protein abundance of antiapoptotic factors [B cell lymphoma 2 (Bcl2) and glutathione peroxidase 1 (Gpx1)] and hypertrophy-related factors (RhoA and cRaf) was decreased, whereas that of the stress-responsive heme oxygenase 1 (HO-1) was increased (figs. S1D and S10). Akt, extracellular signal–regulated kinase (ERK), and endothelial nitric oxide synthase (eNOS) are key signaling molecules in cell survival and proliferation. Phosphorylation of Akt but not that of ERK and eNOS was decreased in MeHg-treated hearts (figs. S1E and S10). Thus, high MeHg exposure seems to not affect LV morphology and functions, although it induced stress responses and decreased survival signaling in mouse hearts.

We next asked whether MeHg promotes a transition from adaptive to maladaptive hypertrophy in response to hemodynamic overload in mouse heart. Exposure to a subneurotoxic dose (10 ppm) of MeHg (low MeHg) is reported to have no effect on neuronal functions in mice over the course of 1 year (29). Oral exposure of mice to low MeHg for 1 week did not decrease body weight (Fig. 1A) but tended to increase mortality after transverse aortic constriction (TAC) compared with vehicle-treated TAC mice (Fig. 1B). Mercury accumulation in heart and other tissues were similar between sham- and TAC-operated mice (Fig. 1C). Pressure overload–induced increases in heart size and weight were enhanced in mice exposed to low MeHg (Fig. 1, D to F), whereas lung weight was not changed by pressure overload and MeHg exposure (Fig. 1G). Low MeHg exposure enhanced pressure overload–induced increases in myocardial cell size (Fig. 1H) and in the expression of hypertrophy- and fibrosis-related genes (fig. S2A). Consistent with the gene expression analysis, interstitial fibrosis was increased in the hearts of MeHg-exposed TAC mice (figs. S2B and S11). Echocardiographic analyses showed that low MeHg exposure substantially exacerbated LV dysfunction induced by pressure overload at 1 or 4 weeks after TAC (table S2). Compared with high MeHg, low MeHg exposure did not increase the abundance of stress-responsive proteins (fig. S2C). These results suggest that low MeHg promotes the cardiac transition from adaptive to maladaptive hypertrophy induced by pressure overload without apparent neurotoxicity.

Fig. 1 Exposure to subneurotoxic dose of MeHg promotes maladaptive cardiac hypertrophy induced by pressure overload.

(A and B) Changes in body weight (A) and survival rate (B) in mice after TAC. Mice were orally given MeHg (10 ppm, in drinking water) for 1 week before TAC (n = 5 to 11 mice per treatment). (C) The mercury concentrations in the indicated organs from mice given drinking water (vehicle) or MeHg-containing water (MeHg) for 1 week (n = 4 to 8 mice per treatment). (D) Representative images of mouse hearts 1 week after TAC (n = 4 to 6 mice per treatment). Scale bars, 1 μm. (E to G) Heart weight (HW)/body weight (BW) (E), HW/tibia length (TL) (F), and lung weight (LW)/TL (G) in mice 1 week after TAC (n = 4 to 6 mice per treatment). (H) Myocardial cell size in mice 1 week after TAC. Left: Representative images of wheat germ agglutinin staining of LV sections. Scale bars, 200 μm. Right: Relative changes in the average cross-sectional area (CSA) of cardiomyocytes in LV myocardium (n = 3 mice per treatment). Data are shown as means ± SEM. *P < 0.05 and **P < 0.01, one-way analysis of variance (ANOVA) (E to H).

Low-dose MeHg induces mitochondrial hyperfission but not cytotoxicity in cardiomyocytes

We have previously reported that mitochondrial hyperfission is observed in the peri-infarct region in mouse hearts after myocardial infarction, which plays a key role in inducing premature myocardial senescence and cardiac dysfunction (25). Because high MeHg exposure reportedly induces mitochondrial dysfunction (30, 31), we next asked whether low MeHg exposure also affected mitochondrial morphology in mouse hearts. Using a transmission electron microscope, we found that mitochondria were significantly fragmented in the LV myocardium of mice exposed to low MeHg (Fig. 2A). Consistent with altered mitochondrial fragmentation, the GTP-binding activity of mitochondrial fission factor Drp1 was significantly increased in MeHg-treated hearts (Fig. 2B and fig. S8). TAC operation also promoted Drp1 activity. On the other hand, MeHg treatment did not affect the protein abundance of the mitochondrial fusion proteins Opa1, Mfn1, and Mfn2 in mouse hearts and neonatal rat cardiomyocytes (NRCMs) (figs. S3, A and B, and S11). Consistent with a previous report about reduced Opa1 levels in failing human hearts (32), the protein abundance of Opa1, Mfn1, and Mfn2 was reduced after TAC in MeHg-treated hearts (figs. S3A and S11). Next, we classified the mitochondrial morphology of NRCMs into three forms: tubule, intermediate, and vesicle, as previously done (25). Exposure of NRCMs to low-dose (30 nM) MeHg for 3 days increased the number of NRCMs with vesicle-type mitochondria (Fig. 2C). MeHg induces apoptosis and necrosis in various cell types (33). Mitochondrial fission occurred at a much lower concentration of MeHg compared with the concentration that induced cytotoxicity in cardiomyocytes (Fig. 2D). Apoptosis as detected by terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling (TUNEL) staining and caspase-3 activation was increased in NRCMs treated with only high-dose (500 nM) MeHg (figs. S4, A and B, and S12). Knockdown of Drp1 by small interfering RNA (siRNA) inhibited MeHg-induced mitochondrial fission (Fig. 2E). In addition, immunostaining revealed that endogenous Drp1 colocalized with and encircled mitochondrial fragments in MeHg-treated but not control NRCMs (Fig. 2F). These results suggest that low-dose MeHg preferentially induces mitochondrial fission through Drp1 activation.

Fig. 2 MeHg induces Drp1 activation and mitochondrial fission in cardiomyocytes.

(A) Representative electron micrographs of LV myocardium in mice 1 week after MeHg exposure (n = 3 mice per treatment). Scale bars, 1 μm. (B) GTP-binding activity of Drp1 in mouse hearts 4 weeks after TAC (n = 5 mice per treatment). (C) Representative images of mitochondrial morphology in NRCMs treated with MeHg (0.03 μM) for 3 days. Right: Mitochondrial morphologies of NRCMs classified into three groups (vesicle, intermediate, and tubule) (n = 3 independent experiments). Scale bars, 10 μm. DMSO, dimethyl sulfoxide. (D) Dose-dependent effect of MeHg on cytotoxicity (circles) and mitochondrial fission (triangles) in NRCMs (n = 3 independent experiments). LDH, lactate dehydrogenase. (E) Effect of Drp1 knockdown on MeHg-induced mitochondrial fission. NRCMs were transfected with negative control siRNA (siNC) or siRNA for Drp1 (n = 3 independent experiments). Right: Percentage of cells with vesicle-type mitochondria. Scale bars, 10 μm. (F) Colocalization of mitochondria and Drp1 in NRCMs treated with MeHg. Green, mitochondria; red, endogenous Drp1; blue, 4′,6-diamidino-2-phenylindole (DAPI) (n = 3 independent experiments). Scale bars, 5 μm. Data are shown as means ± SEM. *P < 0.05 and **P < 0.01, one-way ANOVA (B, C, and E). ATP, adenosine 5′-triphosphate.

MeHg increases Drp1 activity through Drp1 depolysulfidation

Drp1 activity is regulated by its phosphorylation (34). For instance, Drp1-mediated mitochondrial fission is stimulated by phosphorylation of Ser616 and inhibited by phosphorylation of Ser637. Therefore, we investigated whether MeHg regulates Drp1 phosphorylation. Basal Ser616 and Ser637 phosphorylation of Drp1 in mouse hearts was lower than that in forskolin- or calyculin A–stimulated controls, and low or high MeHg exposure did not alter the extent of phosphorylation at Ser616 and Ser637 (figs. S5A and S12). In addition, exposure of NRCMs to low-dose MeHg promoted mitochondrial fission (Fig. 2C) but did not change the extent of phosphorylation at Ser616 and Ser637 (figs. S5B and S12). These results suggest that MeHg increases Drp1 activation independently of phosphorylation at these sites.

The balance between electrophiles and nucleophiles is suggested to play a key role in maintaining redox and cardiac homeostasis. We have previously reported that Cys persulfides and polysulfides, rather than H2S/HS, function as reactive nucleophiles, and Drp1 activity is inhibited by its polysulfidation (3). We investigated whether MeHg regulates the polysulfidation-depolysulfidation cycle of Drp1. Although a sulfane sulfur–specific probe, SSP4, directly reacts with Na2S4 but not NaHS (35), treatment of NRCMs with NaHS substantially increased SSP4 fluorescence intensity (Fig. 3A), suggesting that incorporated NaHS acts as donor for reactive polysulfides. Tag-switch-tag assays revealed that endogenous Drp1 protein was polysulfidated and that exposure to low-dose MeHg reduced Drp1 polysulfide levels (Fig. 3B and fig. S8). Treatment with NaHS prevented MeHg exposure from inducing depolysulfidation and activation of Drp1 (Fig. 3, B and C, and fig. S8). Protein polysulfidation level is controlled by the balance between polysulfidation and depolysulfidation reactions. We previously reported that the reduction of intracellular reactive polysulfide decreases Drp1 polysulfidation (3). However, MeHg did not affect total intracellular reactive polysulfide levels (Fig. 3A), suggesting that MeHg does not reduce the polysulfidation of Drp1. Therefore, we investigated whether MeHg promotes the depolysulfidation of Drp1. MeHg treatment directly induced depolysulfidation of recombinant Drp1 in a dose-dependent manner (figs. S6, A and B, and S12), suggesting that the polysulfide group of Drp1 is highly reactive with electrophiles and that MeHg preferentially removes the sulfur atom from the polysulfide group of Drp1. Consistent with this notion, organic mercury reportedly induces the depolysulfidation of some proteins (3, 28). Drp1 polysulfidation was also inhibited in MeHg-treated hearts (Fig. 3D). Consistent with the enhancement of Drp1 activity in pressure-overloaded hearts (Fig. 2B), TAC also promoted Drp1 depolysulfidation (Fig. 3D and fig. S8). These results suggest that Drp1 activity is regulated by the balance between electrophiles and nucleophiles through the polysulfidation-depolysulfidation cycle.

Fig. 3 MeHg increases Drp1 activity through depolysulfidation of Drp1 proteins.

(A) Intracellular sulfane sulfur levels in NRCMs. NRCMs pretreated with MeHg (0.5 μM) for 14 hours, NaHS (100 μM) for 14 hours, or Na2S4 (10 μM) for 3 hours were incubated with SSP4 (n = 4 independent experiments). (B) Effect of MeHg on the polysulfidation (polyS) of endogenous Drp1 proteins in the presence or absence of NaHS (n = 3 independent experiments). (C) Effect of MeHg with or without NaHS on the GTP-binding activity of Drp1 (n = 3 independent experiments). (D) Effect of MeHg on Drp1 polysulfidation in mouse hearts 1 week after TAC (n = 4 mice per treatment). Data are shown as means ± SEM. *P < 0.05 and **P < 0.01, one-way ANOVA.

The balance of electrophiles and nucleophiles is important for cardiac fragility to hemodynamic overload

We investigated whether NaHS administration prevents Drp1 activation and improves cardiac function of MeHg-exposed mice. As described above, oral exposure of mice to low MeHg for 1 week promoted Drp1 activation (Fig. 4A and fig. S8) and decreased Drp1 polysulfidation (Fig. 4B and fig. S8). Intraperitoneal administration of NaHS for 3 days restored the activity and polysulfidation of Drp1 (Fig. 4, A to C). Next, we analyzed the effect of NaHS administration on the cardiac function of TAC-operated mice. Ejection fraction, fractional shortening, and LV internal diameter at end-systole were improved by NaHS (Fig. 4, D to G). Moreover, low MeHg exposure enhanced pressure overload–induced increases in TUNEL-positive cardiac apoptosis, and NaHS administration suppressed apoptosis (Fig. 4H). Basal cardiac parameters were not changed in mice with MeHg and NaHS administration (table S3). TAC increased the end systolic pressure in mice regardless of whether they were given vehicle, MeHg, or NaHS, indicating that similar levels of pressure overload occurred. These results suggest the pivotal role of the sulfur-dependent regulation of Drp1 in determining the risk of heart failure.

Fig. 4 Administration of NaHS rescues MeHg-mediated Drp1 activation and cardiac dysfunction.

(A and B) Effect of NaHS on Drp1 activation (A) and polysulfidation (B) in MeHg-exposed mouse hearts. Bands corresponding to Drp1 (arrowheads) were quantified (n = 5 mice per treatment). (C) Quantification of GTP-Drp1 (A) and polyS-Drp1 (B). (D) Representative images of echocardiographs from mice 1 week after TAC (n = 5 to 10 mice per treatment). Scale bars, 2 μm. (E to G) Effect of NaHS on ejection fraction (EF) (E), fractional shortening (FS) (F), and LV internal diameters at end-systole (LVIDs) (G) in vehicle-, MeHg-, and NaHS-exposed mice 1 week after TAC (n = 5 to 10 mice per treatment). (H) TUNEL staining to detect apoptosis in myocardium 2 weeks after TAC. Red, TUNEL-positive nuclei; blue, DAPI-stained nuclei. The population of TUNEL-positive apoptotic cells (arrowheads) was quantified (n = 5 mice per treatment). Scale bars, 50 μm. Data are shown as means ± SEM. *P < 0.05 and **P < 0.01, one-way ANOVA (C and E to H).

Polysulfidation of Drp1 at Cys624 negatively regulates Drp1 guanosine triphosphatase binding activity

Drp1 has nine Cys residues. To identify the Cys residue that confers MeHg sensitivity, we constructed a depolysulfidation mimic Drp1 mutant [substitution of Cys to Ser (CS)] and compared the basal mitochondrial fission activity in HeLa cells overexpressing each Drp1 CS mutant. All of the CS mutants promoted mitochondrial fission similar to Drp1 wild type (WT), except for the Drp1 C624S mutant that induced greater mitochondrial fission than Drp1 WT (Fig. 5A). Consistent with the effects on mitochondrial morphology, the Drp1 C624S mutant displayed higher basal GTP-binding activity, and MeHg did not enhance the activity of this mutant (Fig. 5B and fig. S9). The basal polysulfidation of Drp1 C624S was lower than that of Drp1 WT, and MeHg did not affect the polysulfidation of Drp1 C624S (Fig. 5C). To examine whether Cys624 was sufficient for functional polysulfidation, we also generated the Drp1 8CS(C624) mutant, in which all Cys residues except Cys624 were replaced by serine. Although Drp1 8CS(C624) was still polysulfidated, MeHg almost completely abolished polysulfidation (Fig. 5C and fig. S9). These results suggest that Cys624 is a MeHg-sensitive polysulfidation site. Because a polysulfidated Cys thiol group would be bulkier than a normal Cys thiol group, we tested whether the substitution of Cys by tryptophan (a large and bulky amino acid that structurally mimics polysulfidated Cys) alters Drp1 activity. Drp1 C624W displayed lower mitochondrial fission activity than Drp1 WT, and MeHg did not enhance mitochondrial fission through Drp1 C624W (Fig. 5D). These results suggest that MeHg induces mitochondrial fission through depolysulfidation of Drp1 at electrophile-sensitive Cys624.

Fig. 5 Polysulfidation of Cys624 negatively regulates Drp1 activity.

(A) Effect of Drp1 overexpression on mitochondrial fission. HeLa cells overexpressed Drp1 WT or the depolysulfidation mimic CS mutants, and mitochondrial morphologies were visualized (n = 3 independent experiments). Right: Percentage of cells with tubule-type mitochondria. Scale bars, 20 μm. GFP, green fluorescent protein. (B) Drp1 activity in response to MeHg in Drp1 (WT)– or Drp1 (C624S)–overexpressing HeLa cells (n = 3 independent experiments). (C) Effect of MeHg on polysulfidation of Drp1 WT or CS mutants. Drp1 8CS(C624); all eight Cys residues except Cys624 were mutated to serine (n = 3 independent experiments). (D) Effect of MeHg on mitochondrial fission in HeLa cells expressing Drp1 WT or the polysulfidation mimic C624W mutant (n = 3 independent experiments). Right: Percentage of cells with tubule-type mitochondria. Scale bars, 20 μm. Data are shown as means ± SEM. *P < 0.05 and **P < 0.01, one-way ANOVA.

FLNa participates in MeHg-induced mitochondrial fission

We have previously identified FLNa as a guanine nucleotide exchange factor for Drp1 and shown that the interaction between Drp1 and FLNa induces mitochondrial fission in a hypoxia-dependent manner (25). Moreover, cilnidipine inhibits Drp1-FLNa interaction and prevents hypoxia-induced mitochondrial fission (25). We investigated whether FLNa was involved in MeHg-induced mitochondrial fission. Cilnidipine treatment inhibited MeHg-induced mitochondrial fission (Fig. 6A). Moreover, knockdown of FLNa by siRNA also inhibited MeHg-induced mitochondrial fission (Fig. 6B). MeHg treatment promoted the interaction of Drp1 with FLNa, and the interaction of the polysulfidation mimic Drp1 C624W with FLNa was reduced compared to WT (Fig. 6C and fig. S9). We have previously shown that Drp1 and FLNa form punctate structures and colocalize with each other after hypoxia (25). After MeHg treatment, GFP-Drp1 was colocalized with mCherry-FLNa in punctate dots, and cilnidipine pretreatment inhibited the punctate formation and colocalization of Drp1 with FLNa (Fig. 6D). Moreover, Drp1 C624W did not form punctate structures after MeHg treatment (Fig. 6D). These results suggest that Drp1 polysulfidation inhibits the interaction with FLNa.

Fig. 6 FLNa mediates MeHg-induced Drp1 activation.

(A) Effect of cilnidipine (CIL) on MeHg-induced mitochondrial fission. NRCMs were incubated with CIL (1 μM) for 1 hour before MeHg (0.03 μM) treatment. Right: Percentage of cells with vesicle-type mitochondria (n = 3 independent experiments). Scale bars, 20 μm. (B) Effect of FLNa knockdown on MeHg-induced mitochondrial fission. NRCMs were transfected with negative control siRNA (siNC) or two different siRNAs for FLNa (siFLNa #1 and #2). Right: Percentage of cells with vesicle-type mitochondria (n = 3 independent experiments). Scale bars, 20 μm. (C) Interaction of FLNa and Drp1 WT or C624W. HeLa cells transfected with myc-FLNa and FLAG-Drp1 WT or C624W were treated with MeHg for 3 days (n = 3 independent experiments). (D) Localization of GFP-Drp1 WT or C624W with mCherry-FLNa (mChy-FLNa) in cardiac fibroblasts. Cardiac fibroblasts expressing GFP-Drp1 WT or C624W and mCherry-FLNa were incubated with CIL for 1 hour before MeHg treatment. The arrowheads show punctate where Drp1 colocalizes with FLNa (n = 3 independent experiments). Scale bars, 20 μm. Data are shown as means ± SEM. *P < 0.05 and **P < 0.01, one-way ANOVA (A and B).

MeHg-induced depolysulfidation of Drp1 underlies MeHg-induced cardiac vulnerability to mechanical stress

We asked whether Drp1 depolysulfidation and activation by MeHg confers susceptibility to myocardial cell injury induced by mechanical stress. NRCMs cultured in stretchable silicone chambers for 48 hours were subjected to a 20% one-dimensional static stretch for an additional 12 hours. Stimulation of MeHg-exposed NRCMs with static stretch reduced viability as assessed by 3-(4,5-dimethylthiazol-2-yl)-2,5-triphenyltetrazolium bromide (MTT) assays (Fig. 7A). Furthermore, LDH cytotoxicity was increased in MeHg-exposed NRCMs by stimulation with hypotonic stress, another type of mechanical stress (Fig. 7B). Apoptosis (as assessed by TUNEL staining) and caspase-3 activation were also increased by hypotonic stress (figs. S7, A and B, and S12). These effects were attenuated by mitochondrial division inhibitor 1 (Mdivi-1)–mediated inhibition of Drp1 (Fig. 7, A and B). Moreover, NaHS treatment substantially suppressed the static stretch and hypotonic stress–induced injury of MeHg-exposed NRCMs (Fig. 7, C and D). Cilnidipine treatment and FLNa knockdown also suppressed the hypotonic stress–induced injury of MeHg-exposed NRCMs (Fig. 7, E and F). Moreover, overexpression of the polysulfidation mimic Drp1 C624W prevented the MeHg-induced cardiomyocyte vulnerability to mechanical stress (Fig. 7G). These results suggest that the regulation of Drp1 activity by polysulfidation-depolysulfidation cycle at Cys624 underlies cardiac vulnerability to mechanical stress through filamin-dependent mitochondrial hyperfission (Fig. 7H).

Fig. 7 Vulnerability of MeHg-exposed rat cardiomyocytes to mechanical stress.

(A and B) Effect of Mdivi-1 on static stretch (A) or hypotonic-induced (B) cardiac injury in MeHg-exposed NRCMs. NRCMs were preincubated with Mdivi-1 (10 μM) for 30 min and then stretched statically for 12 hours (A) or cultured in hypotonic media for 3 days (B) with or without MeHg (0.05 μM). Cell viability and toxicity were measured by MTT and LDH release assay [n = 4 (A) and 3 (B) independent experiments]. (C and D) Effect of NaHS on static stretch (C) or hypotonic-induced (D) cardiac injury in MeHg-exposed NRCMs. NRCMs were preincubated with NaHS (100 μM) for 24 hours and then stretched statically for 12 hours (C) or cultured in hypotonic media for 3 days (D) with or without MeHg (0.05 μM) (n = 4 independent experiments). (E) Effect of CIL on hypotonic-induced cardiac injury. NRCMs were preincubated with CIL (1 μM) for 1 hour and then cultured in hypotonic media for 3 days with or without MeHg (0.05 μM) (n = 3 independent experiments). (F) Effect of FLNa knockdown on hypotonic-induced cardiac injury. NRCMs were transfected with negative control siRNA (siNC) or two different siRNAs for FLNa (siFLNa #1 and #2) and then cultured in hypotonic media for 3 days with or without MeHg (0.05 μM) (n = 3 independent experiments). (G) Effect of Drp1 polysulfidation on hypotonic-induced cardiac injury. H9c2 cells expressing FLAG-Drp1 WT or C624W were cultured in hypotonic media for 3 days with or without MeHg (n = 4 independent experiments). Data are shown as means ± SEM. **P < 0.01, one-way ANOVA. (H) Working model for cardiac fragility to hemodynamic overload through reactive sulfur-dependent regulation of Drp1 activity. GDP, guanosine diphosphate.

MeHg induced mitochondrial hyperfission and cardiac vulnerability to mechanical load in human cardiomyocytes

We investigated whether the electrophile-mediated regulation of Drp1 and cardiac vulnerability are conserved in human induced pluripotent stem (iPS) cardiomyocytes. Exposure of iPS cardiomyocytes to low-dose MeHg induced mitochondrial fission, which was inhibited by cilnidipine (Fig. 8A). Hypotonic mechanical stress increased LDH release and apoptosis (as assessed by TUNEL staining) of MeHg-exposed iPS cardiomyocytes, which were prevented by pretreatment with cilnidipine or NaHS (Fig. 8, B and C). Moreover, overexpression of Drp1 C624W but not WT suppressed MeHg-induced injury to iPS cardiomyocytes (Fig. 8, D and E). These results suggest that the depolysulfidation of Cys624 in Drp1 increases cardiac fragility to mechanical load through filamin-dependent mitochondrial hyperfission in human cardiomyocytes.

Fig. 8 MeHg-mediated cardiac vulnerability to mechanical stress in human cardiomyocytes.

(A) Effect of CIL on MeHg-induced mitochondrial hyperfission. Human iPS cardiomyocytes were preincubated with CIL (0.5 μM) for 1 hour and then treated with MeHg (0.05 μM) for 3 days. Left: Representative images of mitochondrial morphology (green) in human cardiomyocytes with DAPI (blue). Right: Percentage of cells with vesicle-type mitochondria (n = 3 independent experiments). Scale bars, 50 μm. (B) Effect of CIL and NaHS on hypotonic-induced cytotoxicity of human cardiomyocytes. Cells were preincubated with CIL (0.5 μM) for 1 hour or NaHS (100 μM) for 24 hours and then cultured in hypotonic media for 2 days with or without MeHg (n = 3 independent experiments). (C) Effect of CIL and NaHS on hypotonic stress–induced apoptosis of human cardiomyocytes. Left: Representative images of TUNEL-positive apoptotic cells (green) merged with DAPI (blue). The proportion of TUNEL-positive nuclei (arrowheads) was quantified (n = 3 independent experiments). Scale bars, 50 μm. (D) Effect of Drp1 polysulfidation on hypotonic stress–induced cytotoxicity. Cells were transfected with FLAG-Drp1 WT or C624W and then cultured in hypotonic media for 2 days with or without MeHg. FLAG-Drp1 WT or C624W-positive human cardiomyocytes were visualized with DsRed fluorescence (n = 3 independent experiments). (E) Effect of Drp1 polysulfidation on hypotonic stress–induced apoptosis. The proportion of TUNEL-positive nuclei in DsRed-expressing cells (arrowheads) was calculated (n = 3 independent experiments). Scale bars, 50 μm. Data are shown as means ± SEM. **P < 0.01, one-way ANOVA.

DISCUSSION

Proper mitochondrial fusion-fission dynamics are pivotal in maintaining mitochondrial quality and integrity. Disorders of mitochondrial dynamics affect diverse cellular functions including energy synthesis, metabolism, proliferation, apoptosis, calcium signaling, and redox signaling (36). In particular, cardiomyocytes are highly dependent on mitochondria for energy production. About 30% of the total cell volume of cardiomyocytes is occupied by mitochondria, and about 90% of the required energy in cardiomyocytes is produced by mitochondria (37). Therefore, mitochondria quality control is important for cardiac homeostasis, and its impairment causes the progression of many cardiovascular diseases (38). We demonstrated that a subneurotoxic dose of MeHg exposure did not directly affect cardiac morphology and functions but increased cardiac vulnerability after pressure overload (table S2). MeHg induced mitochondrial hyperfission in cardiomyocytes through Drp1 depolysulfidation (Figs. 2C and 3B), and mitochondrial hyperfission increased mechanical stress–induced fragility of cardiomyocytes (Fig. 7, A and B). Mitochondrial hyperfission alters myocyte Ca2+ handling in a reactive oxygen species–dependent manner, and abnormal Ca2+ signaling initiates stretch-induced cell death through caspase-3 and caspase-9 activation in neonatal heart cells (39).

Drp1 activity is predominantly regulated by posttranslational modifications, including phosphorylation (34), oxidation (27), S-nitrosylation (26), SUMOylation (40), and O-GlcNAcylation (41). We have previously reported that GTP-binding activity of Drp1 is inhibited by reactive persulfide-mediated polysulfidation (3). In this study, we showed that electrophile MeHg promoted the depolysulfidation of rat Drp1 at Cys624 and increased Drp1-mediated mitochondrial fission (Fig. 5, C and D). MeHg-induced Drp1 depolysulfidation increased the GTP-binding activity of Drp1 by promoting the interaction with FLNa, a guanine nucleotide exchange factor for Drp1 (Fig. 6C) (25). Cys624 of rat Drp1 (Cys644 of human Drp1) is a redox-sensitive residue. Sulfenylation or S-nitrosylation of human Drp1 at Cys644 promotes mitochondrial fission (26, 27), whereas covalent modification by the electrophile 15-deoxy-Δ12,14-prostaglandin J2 results in Drp1 inactivation (42). These results suggest that Drp1 functions are intricately regulated by redox conditions. Polysulfidation (sulfhydration)–mediated regulation of protein functions have been reported in several proteins (4). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is polysulfidated at Cys150 by H2S under physiologically relevant conditions, and polysulfidation of GAPDH increases its catalytic activity (43). However, nitric oxide induces S-nitrosylation of Cys150 of GAPDH, which inhibits its catalytic activity (44), indicating that GAPDH activity is competitively regulated by S-nitrosylation and polysulfidation at Cys150. Moreover, Cys443 in eNOS and Cys38 in p65 are also competitively modified by H2S and NO to induce polysulfidation and S-nitrosylation, respectively (45, 46). Protein polysulfidation could be coordinately regulated by other redox modifications such as S-nitrosylation and oxidation. In addition, Drp1-mediated mitochondrial fission is stimulated by phosphorylation of Ser616 and inhibited by phosphorylation of Ser637 (34). Although MeHg did not affect Drp1 phosphorylation directly (fig. S5, A and B), the possibility of cross-talk between Drp1 phosphorylation and redox modification would be important to investigate in the future.

MeHg is a strong electrophile that readily reacts with nucleophilic Cys thiols, leading to alteration of protein functions through Cys S-mercuration (47). High-dose MeHg induces S-mercuration of various proteins including Mn-superoxide dismutase (13), neuronal nitric oxide synthase (11), phosphatase and tensin homolog deleted from chromosome 10 (48), and cyclic adenosine monophosphate response element–binding protein (48), which results in disruption of cellular homeostasis and neuronal toxicity. In cardiomyocytes, high-dose MeHg also induced cytotoxicity, although mitochondrial fission occurred in a lower concentration (Fig. 2D). MeHg reportedly reacts with reactive persulfides and abstracts sulfur, generating nonelectrophilic (MeHg)2S (28). Under our assay conditions, MeHg treatment did not reduce total sulfane sulfur levels in cardiomyocytes (Fig. 3A), whereas Drp1 polysulfidation was reduced by MeHg (Fig. 3B). We also confirmed that MeHg directly induced Drp1 depolysulfidation using recombinant rat Drp1 proteins (fig. S6, A and B). These results suggest that low-dose MeHg selectively reacts with polysulfidated Drp1 and induces Drp1-mediated mitochondrial fission through depolysulfidation. Moreover, NaHS pretreatment inhibited depolysulfidation of Drp1 by MeHg. Increased reactive polysulfides by NaHS may induce repolysulfidation of Drp1. Another possibility is that NaHS directly reacts with MeHg, generating (MeHg)2S (Fig. 7H) (28).

In this study, we showed that depolysulfidation of Drp1 Cys624 by MeHg increases cardiomyocyte vulnerability to mechanical stress in human cardiomyocytes as well as rodent cardiomyocytes (Figs. 7G and 8D), and exposure to subneurotoxic low-dose MeHg exacerbates cardiac dysfunction induced by pressure overload in mice in vivo (table S2). Although epidemiological studies have suggested MeHg as a potent risk factor for cardiac diseases such as myocardial infarction in human (1517), its underlying molecular mechanism and the involvement of Drp1 have been obscure. Changes in Drp1 expression and phosphorylation have been reported in human pathological samples including those from patients with Alzheimer’s disease (49), and the relationship between Drp1 polysulfidation and human diseases has not been analyzed. Future studies focusing on the role of Drp1 polysulfidation in human diseases are necessary to understand the clinical importance of this redox-dependent posttranslational modification.

In conclusion, we demonstrated that subneurotoxic MeHg increases cardiac fragility after pressure overload in mouse hearts. The electrophilic depolysulfidation of Drp1 underlies MeHg-induced Drp1 activation and mitochondrial hyperfission in cardiomyocytes. NaHS protected polysulfidation of Drp1 after MeHg treatment and suppressed MeHg-induced cardiac vulnerability to mechanical load. H2S has been reported to affect multiple biological processes to exert cardioprotective actions in diverse pathological states (50). In particular, H2S-induced polysulfidation on various enzymes, receptors, transcription factors, and ion channels is prevalent in cardiovascular systems (4). Therefore, polysulfidated proteins may be potential targets for therapeutic intervention in cardiovascular systems.

MATERIALS AND METHODS

Reagents and antibodies

MeHg (II) chloride, thiocarbohydrazide solution, durcupan resin, and GTP-agarose were from Sigma-Aldrich. Anti-Drp1 (H-300), anti-Mfn1 (H-65), and anti-GAPDH (FL-335) were from Santa Cruz Biotechnology. Anti–cleaved caspase-3 (no. 9664), anti–phospho-Drp1 (Ser616) (no. 4494), anti–phospho-Drp1 (Ser637) (no. 4867), anti-Akt (C67E7) (no. 4691), anti–phospho-Akt (Ser473) (no. 4060), anti–p44/42 mitogen-activated protein kinase (MAPK) (Erk1/2) (no. 4695), anti–phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) (no. 4370), anti–phospho-eNOS (Ser1177) (no. 9571), and horseradish peroxidase–conjugated anti-mouse and anti-rabbit immunoglobulin G were from Cell Signaling. Anti-Mfn2 (ab56889) and anti-Drp1 (ab56788) antibodies were from Abcam. Anti-eNOS mouse antibody (no. 610296) and anti-Opa1 (no. 612606) were from BD Biosciences. Anti-ATP synthase β monoclonal antibody (3D5AB1) was from Thermo Fisher Scientific.

Animals and MeHg exposure

All protocols using mice and rats were reviewed and approved by the ethics committees at the National Institute for Physiological Sciences or the Animal Care and Use Committee at Kyushu University and were performed according to Institutional Guidelines Concerning the Care and Handling of Experimental Animals. Five-week-old male C57BL/6J mice were purchased from Japan SLC Inc. (Shizuoka, Japan). All mice were kept in plastic cages in a climate-controlled animal room with a 12-hour light/dark cycle. MeHg was dissolved into drinking water and freely available to the mice from 1 week before TAC surgery to the end point of the in vivo experiments. A mini-osmotic pump (ALZET) filled with vehicle [phosphate-buffered saline (PBS)] or NaHS (50 μmol/kg per day) was implanted intraperitoneally into the mice 3 days before TAC surgery.

TAC surgery, transthoracic echocardiography, and cardiac catheterization

Pressure overload was induced by surgical TAC operation. Briefly, male mice (6 weeks old) were anesthetized by a mixture of domitor (Nippon Zenyaku Kogyo, Japan), midazolam (Sandoz, Japan), and butorphanol (Meiji Seika Pharma, Japan). Under artificial respiration, the transverse aorta between the brachiocephalic artery and the left carotid artery was constricted by the width of a 27-gauge needle using a 5-0 silk braid. Sham treatment was performed similarly but without constriction. Echocardiography was performed using Nemio XG echocardiography (Toshiba, Japan) with a 14-MHz transducer every week. LV functions were measured using a micromanometer catheter (Millar 1.4F, SPR 671; Millar Instruments).

Transmission electron microscopy

Mouse LV tissues were prefixed for 3 hours on ice using 2% paraformaldehyde solution containing 0.15 M sodium cacodylate and 2 mM CaCl2 (pH 7.4) and cut out into 1- to 2-mm cubes. After washing with 0.15 M cacodylate solution, block tissues were immersed in solution containing 2% osmium tetroxide, 1.5% potassium ferrocyanide, 0.15 M sodium cacodylate, and 2 mM CaCl2 (pH 7.4) for 1.5 hours at room temperature. After washing with distilled water, tissue cubes were immersed in thiocarbohydrazide (0.01 mg/ml) solution for 40 min and then postfixed using 2% osmium for 1 hour. En bloc staining was performed by immersing tissues in a solution of 1% uranium acetate overnight at 4°C and then immersing them in an aqueous solution of lead aspartic acid for 60 min with oven frying. After dehydration with graded series of ethanol and acetone, specimens were embedded with durcupan resin. The surface (70-nm thickness) of resin-embedded tissue was exposed using a diamond knife on an Ultracut UC7 (Leica Microsystems, Austria). The surface of embedded tissue was imaged with a Veleta charge-coupled device camera (Olympus, Germany) equipped on JEOL1010 (JEOL, Tokyo, Japan).

Morphological analysis

Mouse hearts were removed, washed in PBS, and fixed with 10% neutral-buffered formalin (Nacalai Tesque). For the assessment of collagen I and III deposition, the heart was stained with Picrosirius Red Stain Kit (Polysciences Inc.), following the manufacturer’s instructions.

Cell cultures and transfection

Neonatal rat cardiac myocytes and fibroblasts were isolated from Sprague-Dawley rat pups (1 to 2 days) as described previously (25). The iPS cell–derived cardiomyocyte product, iCell Cardiomyocytes2, was purchased from Cellular Dynamics and maintained according to the manufacturer’s instructions. HeLa cells cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum and antibiotics [penicillin (100 U/ml) and streptomycin (100 μg/ml)] were seeded at 5 × 105 cells/cm2 on a culture dish. Plasmid DNA was transfected using ViaFect (Promega) for iCell Cardiomyocytes2 and HeLa cells or a Neon transfection system (Invitrogen) for cardiac fibroblasts. For siRNA knockdown, cells were transfected with each siRNA (20 nM) using Lipofectamine RNAiMAX reagent (Invitrogen) for 72 hours. Stealth siRNAs for rat FLNa (#1, RSS308835; #2, RSS308836) and rat Drp1 (RSS300120) were from Invitrogen.

GTP-agarose pulldown assay

The GTP-agarose pulldown assay was performed as described (25). Briefly, mouse hearts were homogenized using Physcotron (Microtec, Japan) in ice-cold GTP-binding buffer [50 mM Hepes (pH 7.4), 1% Triton X-100, 10% glycerol, 150 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA, and 1% protease inhibitor cocktail]. NRCMs were washed with ice-cold PBS, collected in GTP-binding buffer, and homogenized by sonication for 5 s (Qsonica, WakenBtech, Japan). The lysate was centrifuged (16,000g for 10 min at 4°C), and an aliquot of the supernatant (100 μg of protein) was incubated with 50 μl of GTP-agarose beads (equilibrated in GTP-binding buffer) for 45 min at room temperature. The beads were centrifuged (1000g for 1 min at 4°C) and washed twice with GTP-binding buffer. The GTP-bound proteins were eluted by 2× Laemmli buffer with 2-mercaptoethanol and subjected to SDS–polyacrylamide gel electrophoresis (SDS-PAGE).

Measurement of protein S-polythiolation

Measurement of persulfide/polysulfide levels in cells was performed using a polysulfide-specific fluorescent probe SSP4 [3′,6′-di(O-thiosalicyl)fluorescein]. Fluorescence intensity at the excitation wavelength of 482 nm and emission wavelength of 515 nm was measured using SpectraMax i3 plate reader (Molecular Devices). Protein polysulfidation was detected by tag-switch-tag assay, as described previously (2). For lysate samples, cell lysates (100 μg of protein) were prepared using radioimmunoprecipitation assay buffer and incubated with methylsulfonyl benzothiazole (MSBT; 2 mM) at 37°C for 1 hour to label free thiols. Lysates were then reacted with CN-biotin (3 mM) at 37°C for 1 hour, and methanol/chloroform precipitation was performed to remove free probes. The biotinylated (persulfide/polysulfide) proteins were pulled down using streptavidin-agarose beads. The beads were washed thrice with PBS, and bound proteins were eluted by 2× Laemmli buffer with 2-mercaptoethanol and subjected to SDS-PAGE. For recombinant proteins, recombinant Drp1 (10 μM) were reacted with MeHg (10 or 50 μM) at 37°C for 1 hour; then, MeHg was removed by Sephadex G-25 desalting column. Drp1 was incubated with MSBT (1 mM) at 37°C for 1 hour and then reacted with CN-biotin (2.5 mM) at 37°C for 1 hour. Free probes were removed by desalting columns. Proteins were subjected to SDS-PAGE.

Mitochondrial morphology

NRCMs or iCell Cardiomyocytes2 were plated on Matrigel-coated glass coverslips or fibronectin-coated CELLview 10-well slide (Advanced TC treated, Greiner), respectively. Cells were treated with MeHg for 3 days. Mitochondria were stained using ATP synthase β antibody and detected by BZ-X700 microscope (KEYENCE). At least 10 images were randomly taken for each sample. The average mitochondrial fragment length per cell was quantified using ImageJ, and the mitochondrial morphology was classified into three groups (vesicle, intermediate, and tubule) according to the average mitochondrial fragment length, as described previously (25).

Cell toxicity and viability

NRCMs were seeded at 1.5 × 105 cells/ml on Matrigel-coated 96-well plate or glass bottom dish and treated with MeHg for 3 days. Cell toxicity was measured using Cytotoxicity LDH Assay Kit-WST (Dojindo). Apoptosis was measured by DeadEnd Fluorometric TUNEL System (Promega) for culture cells, TUNEL Assay Kit (Red) (Abcam) for tissue, or immunoblotting with cleaved caspase-3. To analyze mechanical vulnerability, NRCMs plated onto laminin-coated silicone rubber culture dishes (Menicon, Japan) were statically stretched by 20% using a stretching machine. Cell viability was assessed by MTT assay. After 12 hours of static stretch, NRCMs were incubated with MTT (0.5 mg/ml) in culture medium for 3 hours. The supernatant was aspirated, and formazan crystals were dissolved in DMSO. The formazan concentration was determined by optical density at 570 nm using a SpectraMax i3 plate reader. For hypotonic stimulation, NRCMs were incubated in normal or 50% hypotonic media (50% DMEM and 50% water) with or without 50 nM MeHg for 3 days. H9c2 cells infected with retrovirus encoding FLAG-Drp1 WT or C624W were seeded on 96-well plate. iCell Cardiomyocytes2 plated onto fibronectin-coated CELLview 10-well slides were transfected with both FLAG-Drp1 WT or C624W and DsRed. After 8 hours, cells were incubated in 30% hypotonic media with or without 50 nM MeHg for 2 days. Cell toxicity and apoptosis were measured by LDH release assay and TUNEL assay, respectively.

Statistical analysis

The results are shown as means ± SEM. All experiments were repeated at least three times. Statistical comparisons were made with two-tailed Student’s t test (for two groups) or one-way ANOVA followed by a Tukey comparison procedure (for three and more groups). Values of P <0.05 were considered to be statistically significant.

SUPPLEMENTARY MATERIALS

stke.sciencemag.org/cgi/content/full/12/587/eaaw1920/DC1

Fig. S1. Effect of MeHg (100 ppm) on body weight and protein amounts.

Fig. S2. Subneurotoxic dose of MeHg promotes induction of maladaptive gene expression and remodeling after pressure overload.

Fig. S3. Changes in the abundance of mitochondrial fusion- and fission-related proteins.

Fig. S4. Effect of MeHg exposure on cardiomyocyte apoptosis.

Fig. S5. MeHg exposure minimally affects Drp1 phosphorylation in the heart.

Fig. S6. Depolysulfidation of recombinant Drp1 by MeHg.

Fig. S7. Hypotonic stress–induced apoptosis of MeHg-exposed NRCMs.

Fig. S8. Uncropped Western blots for Figs. 2 to 4.

Fig. S9. Uncropped Western blots for Figs. 5 and 6.

Fig. S10. Uncropped Western blots for fig. S1.

Fig. S11. Uncropped Western blots for figs. S2 and S3.

Fig. S12. Uncropped Western blots for figs. S4 to S7.

Table S1. Echocardiographic parameters of mice with 100-ppm MeHg exposure at 1 or 2 weeks.

Table S2. Echocardiographic parameters of TAC-operated mice with MeHg exposure.

Table S3. Millar catheter analysis of TAC-operated mice given MeHg and NaHS.

REFERENCES AND NOTES

Acknowledgments: We thank S. Oda, N. Nagai, and H. Ishihara (NIPS, Okazaki, Japan) for technical support and the Spectrography and Bioimaging Facility, NIBB Core Research Facilities. We also appreciate the technical assistance from The Research Support Center, Research Center for Human Disease Modeling, Kyushu University Graduate School of Medical Sciences. Funding: This work was supported by grants from JSPS KAKENHI [to M.N. (16KT0013 and 19H03383), A.N. (17K15464 and 19K07085), T. Tanaka (19K16363), Y.K. (25220103 and 18H05293), and T.A. (18H05277)]. This work was also supported by JSPS KAKENHI grant no. JP16H06280 and Platform Project for Supporting Drug Discovery and Life Science Research [Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS)] from AMED (JP19am0101091), a Japan Agency for Medical Research and Development grant (JP18mk0104117) to Y. Kanda, and a grant from The Cell Science Research Foundation (to M.N.). Author contributions: M.N. and A.N. designed the research and wrote the paper. A.N., K.S., T. Tanaka, T. Toyama, and K.N. performed experiments. Y.S., T.N.-T., D.Y., Y. Kanda, Y. Kumagai, and T.A. contributed new reagents/analytic tools. A.N., T. Tanaka, T. Toyama, K.N., D.Y., Y. Kanda, Y. Kumagai, and T.A. analyzed and interpreted data. M.N. edited the paper. Competing interests: A patent (JPWO2016080516A1) has been filed in the International Patent System for part of this work. M.N. is named as an inventor on this patent. The other authors have declared that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials. A material transfer agreement is required for pcDNA3-myc-FLNa WT plasmid by The President and Fellows of Harvard College.
View Abstract

Stay Connected to Science Signaling

Navigate This Article