Hydrogen sulfide anion regulates redox signaling via electrophile sulfhydration

Abstract

An emerging aspect of redox signaling is the pathway mediated by electrophilic byproducts, such as nitrated cyclic nucleotide (for example, 8-nitroguanosine 3',5'-cyclic monophosphate (8-nitro-cGMP)) and nitro or keto derivatives of unsaturated fatty acids, generated via reactions of inflammation-related enzymes, reactive oxygen species, nitric oxide and secondary products. Here we report that enzymatically generated hydrogen sulfide anion (HS(-)) regulates the metabolism and signaling actions of various electrophiles. HS(-) reacts with electrophiles, best represented by 8-nitro-cGMP, via direct sulfhydration and modulates cellular redox signaling. The relevance of this reaction is reinforced by the significant 8-nitro-cGMP formation in mouse cardiac tissue after myocardial infarction that is modulated by alterations in HS(-) biosynthesis. Cardiac HS(-), in turn, suppresses electrophile-mediated H-Ras activation and cardiac cell senescence, contributing to the beneficial effects of HS(-) on myocardial infarction-associated heart failure. Thus, this study reveals HS(-)-induced electrophile sulfhydration as a unique mechanism for regulating electrophile-mediated redox signaling.

Figure 1. Regulation of 8-nitro-cGMP–induced protein S-guanylation by HS⁻-producing enzymes via sulfhydration of 8-nitro-cGMP by HS⁻

(a,b) Enhancement of protein S-guanylation by knockdown of CBS (a) and CSE (b) in A549 cells treated or untreated with 8-nitro-cGMP (200 μM). In a and b, quantitative data (via densitometric analysis) of S-guanylation western blots are shown at left, and NO₂ ⁻ release from 8-nitro-cGMP with A549 cells in culture with and without knockdown of CBS or CSE is shown at right. Data represent mean ± s.e.m. (n = 4). *P < 0.05, **P < 0.01 versus control. Original western blot images are shown in Supplementary Figure 1b. (c) Possible reaction mechanisms in HS⁻-dependent sulfhydration of 8-nitro-cGMP. Transition metals (M) and cysteine thiolate anion (Cys-S⁻) may participate in sulfhydration by stabilizing the HS⁻ anion. (d) Mass spectrum of the peak identified as 8-SH-cGMP formed in the reaction of 8-nitro-cGMP with NaHS. (e) Effects of cysteine and metals on HS⁻-mediated sulfhydration of 8-nitro-cGMP. The reaction of 8-nitro-cGMP (100 μM) with NaHS (1 mM) was carried out in 100 mM sodium phosphate buffer (pH 7.4) containing 100 μM DTPA in the absence or presence of additives including cysteine (100 μM), metals (150 μM FeSO ₄, FeCl₃, MnCl₂, CuSO ₄ or ZnCl₂) and metal-containing compounds or proteins (10 μM hemin, Cu,Zn-SO D, Mn-SO D, catalase or HRP) at 37 °C for 5 h. Data represent mean ± s.d. (n = 3).

Figure 2. Electrophile sulfhydration by HS⁻

(a) LC/MS analysis for the OANO₂ and HS⁻ reaction. OANO₂ (100 μM) was reacted with different concentrations of NaHS in 200 mM potassium phosphate buffer at 25 °C for 1 h. (b) LC/MS for identification of 15d-PGJ₂-SO ₃H as a product of the reaction of 15d-PGJ₂ (1 μM) and HS⁻ in 100 mM phosphate buffer at 37 °C for 2 h. (c) Time-dependent effects of HS⁻ on stability of various electrophiles. Left, single-ion recordings of 15d-PGJ₂, 15d-PGJ₂-SH and 15d-PGJ₂-SO₃H in the same reaction as in b with 15d-PGJ₂ (10 μM) and NaHS (1 mM). The MS spectrum of each PGJ₂ derivative is shown in Supplementary Figure 4c. Right, HNE and acrolein determined by RP-HPLC analyses. HNE (10 μM) or acrolein (10 μM) was incubated with NaHS (200 μM for HNE; 100 μM for acrolein) in 100 mM phosphate buffer at 37 °C. AU, arbitrary unit. (d) Effects of HS⁻ on protein electrophile adductions in A549 cells. Cells were pretreated with NaHS or were untreated for 12 h, followed by treatment with 1,2-NQ (10 μM) for 1 h. Total cellular proteins were subjected to western blotting (WB). Cells were transfected with control siRNA or CBS siRNA and were then exposed to 1,2-NQ (10 μM) for 1 h or to acrolein (3 μM or 10 μM) for 30 min. Western blots for 1,2-NQ and acrolein protein adducts formed in A549 cells with CBS knockdown or untreated cells (middle and right). Uncropped blots are shown in Supplementary Figure 17. MW, molecular weight.

Figure 3. Cellular formation of 8-SH-cGMP and its metabolic fate

(a) Metabolic fate of electrophiles after reaction with HS⁻. Details of the classification are shown in Supplementary Figure 6. (b) Formation of 8-SH-cGMP and its suppression by CBS knockdown in 8-nitro-cGMP–treated A549 cells. 8-SH-cGMP formation was determined via LC-ESI-MS/MS (8-[³⁴SH]-cGMP was used as an internal standard). Data represent mean ± s.d. (n = 3). **P < 0.01. (c) Oxidative desulfhydration of 8-SH-cGMP by H₂O₂ to form cGMP. Left, the reaction of 8-SH-cGMP (10 μM) with H₂O₂ (1 mM) led to the disappearance of 8-SH-cGMP (elution time, 6 min) and appearance of a new peak (elution time, 4 min). Right, LC/MS identified cGMP as a major product in the reaction of 8-SH-cGMP with H₂O₂. AU, arbitrary unit. (d) Oxidative desulfhydration of 8-SH-cGMP by H₂O₂ and RNOS. 8-SH-cGMP (10 μM) was reacted with various concentrations of H₂O₂ (left) and with H₂O₂ and RNOS (1 mM each) (right). NO/O₂: NO under aerobic conditions, producing nitrogen oxides; NO/O₂ ⁻: simultaneous generation of NO and O₂ ⁻ by 3-morpholinosydnonimine (SIN-1) to form ONOO⁻ in situ; ONOO⁻: synthetic ONOO⁻. Data represent mean ± s.d. (n = 3). (e) Stability of 8-SH-cGMP formed from 8-nitro-cGMP in A549 cells enhanced by antioxidant enzymes. A549 cells were treated with 8-nitro-cGMP (200 μM) in the absence and presence of antioxidant PEG-SO D (100 U ml⁻¹) and PEG-catalase (200 U ml⁻¹) for 6 h. LC-ESI-MS/MS was used to determine 8-SH-cGMP and cGMP formation. Data represent mean ± s.d. (n = 3). *P < 0.05.

Figure 4. Cellular HS⁻ formation and its physiological relevance to electrophile metabolism

(a) Quantification by LC-ESI-MS/MS coupled with a monobromobimane-based assay of HS⁻ formed by A549 cells treated with CBS siRNA or untreated cells. Data represent mean ± s.d. (n = 3). **P < 0.01. CBS knockdown was confirmed by western blot (Supplementary Fig. 7b). (b) Quantification of low-molecular-weight thiols in A549 cells with and without CBS knockdown. Data represent mean ± s.d. (n = 3). **P < 0.01. GS-bimane, glutathione-bimane. (c) Comparison of HS⁻ production in cultured cells. Data represent mean ± s.d. (n = 3). (d) Western blots for expression of CBS and CSE in various cultured cells and rat cardiac cells. Mouse cardiac tissues were analyzed 4 weeks after sham operation or myocardial infarction or 6 weeks after sham operation or TAC (top). Cultured cells (A549, HepG2, C6, mouse hepatocytes) and rat cardiac fibroblasts and myocytes were analyzed for expression of CBS and CSE. Uncropped blots are shown in Supplementary Figure 18.

Figure 5. Electrophilic H-Ras activation regulated by HS⁻ in cells and in vivo in cardiac tissues

(a) Immunohistochemical determination of 8-nitro-cGMP formation in mouse hearts after myocardial infarction (MI) after NaHS treatment or in untreated (vehicle) hearts (original images are in Supplementary Fig. 8d). Data represent mean ± s.e.m. **P < 0.01. (b) Quantitative data for western blots of H-Ras activation and S-guanylation in mouse hearts (original blots are in Supplementary Fig. 8f). Data represent mean ± s.e.m. *P < 0.05, **P < 0.01. (c) Protective effects of HS⁻ on chronic heart failure caused by myocardial infarction. Cardiac functions of mice after myocardial infarction or sham operation and treated with NaHS or untreated (treated with vehicle). LV IDd, left ventricular end-diastolic internal diameter; FS, fractional shortening; dP/dt_max, maximal rate of pressure development; EDP, end-diastolic pressure. Data represent mean ± s.e.m. **P < 0.01. (d) Quantitative data for western blots measuring amount of phosphorylation (p-) of ERK, p38 MAPK, p53 and Rb and expression of ERK, p38 MAPK, p53 and Rb in mouse hearts after myocardial infarction (original blots are in Supplementary Fig. 8f). Because western blotting detected no expression of Rb and p53 in sham-operated mouse hearts, the amounts of phosphorylation of p53 and Rb were compared after normalization with GAPDH as an internal control. Data represent mean ± s.e.m. *P < 0.05, **P < 0.01.

Figure 6. HS⁻ regulation of H-Ras electrophile sensing and signaling

(a) Suppression of LPS-induced iNOS expression, S-guanylation and activation of H-Ras by iNOS knockdown in rat cardiac fibroblasts in culture. (b) Identification of the S-guanylation site in H-Ras. Top, quantitative data for western blotting of S-guanylation in recombinant H-Ras (wild-type (WT) and C118S and C184S mutants) treated with 8-nitro-cGMP (10 μM) and its suppression by NaHS (100 μM). Bottom, MS identification of S-guanylation at Cys184 in recombinant H-Ras treated with 8-nitro-cGMP (10 μM). (c) Induction of rat cardiomyocyte senescence by ATP (100 μM) treatment (left) and in cardiac fibroblasts stimulated with LPS and ATP (right) and their suppression by NaHS (100 μM). Cells were stimulated with LPS (1 μg ml⁻¹) or ATP (100 μM) for 3 h, followed by incubation in 0.5% (v/v) serum-containing medium for 4 d. In some experiments, cells were pretreated with 100 μM NaHS for 1 h before stimulation with LPS or ATP. (d) Senescence induction of rat cardiac fibroblasts expressing WT or H-Ras^C184S by 8-nitro-cGMP (10 μM) or LPS (1 μg ml⁻¹). Scale bars, 20 μm. (e) 8-Nitro-cGMP–induced (10 μM) senescence of rat cardiac fibroblasts transfected with control vector or vectors expressing wild-type H-Ras or H-Ras^C118S or H-Ras^C184S mutants. In c and d, cell senescence was determined by SA–β-gal staining with quantitative analysis. Data represent mean ± s.e.m. *P < 0.05, **P < 0.01. Original images for a–c and e are shown in Supplementary Figure 11.

Figure 7. 8-Nitro-cGMP–induced H-Ras activation in cardiac cells and their suppression by HS⁻

(a) 8-Nitro-cGMP–induced H-Ras activation with its concomitant S-guanylation in rat cardiomyocytes and their suppression by HS⁻. Rat cardiomyocytes were pretreated with NaHS (100 μM) for 24 h, followed by incubation with 8-nitro-cGMP (10 μM) for 4 d. Data represent mean ± s.e.m. **P < 0.01. (b) 8-Nitro-cGMP–induced activation of H-Ras via S-guanylation at Cys184 in membrane preparation of rat cardiac fibroblasts. Cardiac membranes expressing GFP-fused WT H-Ras or H-Ras^C184S mutants were treated with 8-nitro-cGMP (10 μM) for 1 h. Data represent mean ± s.e.m. *P < 0.05, **P < 0.01. (c) Time courses of phosphorylation (p-) of ERK, p38 MAPK, Akt, p53 and Rb induced by 8-nitro-cGMP (10 μM) in rat cardiomyocytes. Cells were untreated or treated with NaHS (100 μM) 24 h before 8-nitro-cGMP treatment. Data represent mean ± s.e.m. Original western blots for a–c are shown in Supplementary Figure 14.

Figure 8. Dissociation of H-Ras from rafts and its activation induced by H-Ras Cys184 S-guanylation

(a) 8-Nitro-cGMP–induced colocalization of H-Ras with the Ras-binding domain (RBD) at the plasma membrane of rat cardiac fibroblasts. Cells expressing GFP-fused H-Ras and DsRed-fused cRaf-RBD proteins were pretreated for 3 h with cycloheximide (50 μg ml⁻¹) and then incubated for 3 h in cycloheximide with or without NaHS (100 μM). Then, cells were untreated or treated with 8-nitro-cGMP (10 μM) for 3 h. Merge images of GFP-fused H-Ras and DsRed-fused cRaf-RBD are shown (individual images of GFP-fused H-Ras, DsRed-fused cRaf-RBD, differential interference contrast (DIC) images are in Supplementary Fig. 16a,b). Scale bars, 50 μm. (b) Quantitative data obtained from morphometric analysis for colocalization of H-Ras and cRaf-RBD at the plasma membrane shown in a. Data represent mean ± s.e.m. **P < 0.01. (c) 8-Nitro-cGMP–induced H-Ras dissociation from rafts with concomitant S-guanylation and the implication of this dissociation in H-Ras activation. Western blotting of H-Ras, flotillin 1 and S-guanylated H-Ras for the raft fraction and active H-Ras, S-guanylated H-Ras and H-Ras for the nonraft fraction in the presence or absence of 8-nitro-cGMP (uncropped blots are in Supplementary Fig. 16c). Data represent mean ± s.e.m. *P < 0.05, **P < 0.01. (d) Schematic model showing activation of H-Ras induced by 8-nitro-cGMP. Gray lipid indicates palmitoylation (at Cys181), and black lipid indicates isoprenylation (at Cys186). The yellow balls are cysteine residues. Magenta ribbons represent α-helices, and yellow ribbons represent β-sheets. T indicates that HS⁻/H₂S inhibits the reaction mediated by the electrophile 8-nitro-cGMP. The three-dimensional structures of H-Ras and Raf-RBD were obtained from the Protein Data Bank under accession codes 3L8Z (right) and 3KUD (left).