The Gasotransmitter Hydrogen Sulfide Induces Nrf2-Target Genes by Inactivating the Keap1 Ubiquitin Ligase Substrate Adaptor Through Formation of a Disulfide Bond Between Cys-226 and Cys-613

Abstract

Aims: The signaling molecule hydrogen sulfide (H₂S) protects cells against oxidative stress and activates NF-E2 p45-related factor 2 (Nrf2), a transcription factor that regulates antioxidant genes. We sought to establish whether H₂S requires Nrf2 to protect against oxidative stress, and whether activation of Nrf2 by H₂S involves antagonism of Kelch-like ECH-associated protein-1 (Keap1), a redox-sensitive ubiquitin ligase substrate adaptor that represses Nrf2 under normal homeostatic conditions. Results: H₂S stabilizes Nrf2 protein and induces Nrf2-target genes via an antioxidant-/electrophile-response element. In mouse embryonic fibroblasts, the ability of H₂S to protect against cell death caused by the redox-cycling agent menadione is dependent on Nrf2. Moreover, Nrf2 regulates murine genes involved in the production of H₂S (Cystathionine-beta-synthase [Cbs] and Cystathionine-gamma-lyase [Cse]) and the degradation of H₂S (Sulfide:quinone reductase-like [yeast] [Sqrdl]). We found that H₂S stabilizes Nrf2 through inhibition of Keap1, an event that requires covalent modification of amino acids C226 and C613 in the substrate adaptor. Upregulation of Nrf2 by H₂S partially involves the production of H₂O₂, which inhibits Keap1 by stimulating the formation of an intramolecular disulfide bond between C226 and C613. The Keap1 C226 and C613 residues are also S-sulfhydrated by H₂S, and this may entail reduction of the C226–C613 disulfide bridge formed by H₂O₂. Innovation: Upregulation of Nrf2 by H₂S and H₂O₂ involves inactivation of Keap1 through modification of C226 and C613. Conclusion: Inhibition of Keap1 by H₂S leads to Nrf2-mediated induction of cytoprotective genes. Nrf2 controls Cbs, Cse, and Sqrdl, suggesting that a feedback loop exists between Nrf2 and H₂S. Antioxid. Redox Signal. 19, 465–481.

Introduction

T he term oxidative stress is commonly used to describe the situation in which there is an intracellular excess of reactive oxygen species (ROS) relative to the antioxidant capacity of the cell (20). In such circumstances, the imbalance between ROS and antioxidants can lead to aberrant signal transduction and cell death. Oxidative stress arises either from increased production of ROS or by depletion of antioxidants; in the latter case, for example, exposure of cells to electrophiles can lead to a marked decrease in the levels of the antioxidant-reduced glutathione (GSH). To protect against the deleterious effects of oxidative stress, cells are equipped with a number of mechanisms that limit ROS accumulation. Chief among these is the antioxidant response, which constitutes a stress-inducible genetic program that is directed by transcription factor NF-E2 p45-related factor 2 (Nrf2) (2, 13, 61).

The Nrf2 transcription factor is a member of the cap-n-collar (CNC) basic-region leucine-zipper (bZIP) family of proteins (24). It mediates induction of over a hundred genes encoding proteins that are involved in various cytoprotective functions, including synthesis of GSH, maintenance of protein thiols, inactivation of electrophiles, drug efflux, generation of NADPH, and activity of the proteasome (32, 42, 56); each of the genes regulated by Nrf2 contains an antioxidant-response element (ARE, i.e., 5′-^A/_GTGA^C/_GNNNGC^A/_G-3′)^† in its regulatory region (7, 17, 53, 58, 75). Induction of ARE-driven genes helps mitigate injury to cells caused by oxidants and electrophiles (25).

Innovation

This study provides evidence that hydrogen sulfide (H₂S) activates the Kelch-like ECH-associated protein-1 (Keap1)–NF-E2 p45-related factor-2 (Nrf2)–antioxidant-response element (ARE) cytoprotective pathway. The ubiquitin ligase substrate adaptor Keap1 serves as a versatile stress sensor. When Keap1 is exposed to H₂S, an intramolecular disulfide bond is formed between its C226 and C613 residues, which prevent it from repressing Nrf2- and ARE-driven gene expression. Inhibition of Keap1 by H₂S involves production of H₂O₂, which also entails oxidation of C226 and C613 in the adaptor protein. Thus, activation of Nrf2 by H₂S partly involves H₂O₂ signaling. Lastly, Nrf2 regulates genes involved in H₂S synthesis and catabolism, suggesting that an inter-relationship exists between Nrf2 and H₂S.

The activity of Nrf2 is tightly controlled. It is primarily regulated at the level of protein stability (43, 52) through an interaction with the ubiquitin ligase adaptor Kelch-like ECH-associated protein-1 (Keap1), which allows the Keap1-Cul3-Rbx1 E3 ubiquitin ligase complex (i.e., Keap1^CRL) to ubiquitylate Nrf2 under nonstressed conditions (35, 68). Keap1 is essential for maintenance of redox homeostasis, because its substrate adaptor activity ensures that Nrf2 has a short half-life, and the expression of ARE-driven antioxidant genes is restricted under normal homeostatic conditions. However, the activity of Keap1 is redox sensitive. Inactivation of Keap1 by agents that produce ROS or are soft electrophiles increases Nrf2 stability, and this in turn results in enhanced transactivation of ARE-driven gene expression (8, 39, 43, 52, 74). The increase in stability of Nrf2 observed upon stress appears to occur as a consequence of thiol-mediated inactivation of Keap1^CRL, which allows the CNC-bZIP factor to evade ubiquitylation.

Keap1 is a dimeric protein that contains five domains: the N-terminal region (residues 1–56)^‡, the Broad complex, Tramtrack, and Bric-à-Brac (BTB) domain (residues 57–177), the Intervening region (IVR, residues 178–326), the Kelch-repeat domain (Kelch, residues 327–598), and the C-terminal region (CTR, residues 599–624) (54). The BTB domain is responsible for dimerization of Keap1 subunits and recruitment of Cul3, while the Kelch and CTR domains form a six-bladed β-propeller to which substrates bind. Mammalian Keap1 proteins typically contain 25–27 cysteines, many of which are susceptible to adduction by thiol-reactive stressors (12, 27, 41). It is becoming clear that Keap1 contains several distinct thiol-based sensors that exhibit selectivity for different types of stressor. First, for example, Keap1 can be inactivated through covalent modification of C151 within its BTB domain by a diverse range of compounds, including tert-butylhydroquinone (tBHQ), diethylmaleate, Ebselen, nitric oxide (NO), and sulforaphane (SFN) (14, 36, 46, 73). Secondly, Keap1 can be inactivated through adduction of C273 and/or C288 within its IVR by chemicals that contain an α,β-unsaturated carbonyl moiety, such as dexamethasone mesylate, acrolein, 4-hydroxynonenal, and 15-deoxy-Δ^12,14-prostaglandin J₂ (12, 14, 36, 46). In addition to inactivation of Keap1 by electrophiles through the covalent modification of C151, C273, and C288, the transition metals Zn²⁺, Cd²⁺, As³⁺, and Se⁴⁺ can inactivate the substrate adaptor through noncovalent binding by coordination chemistry to both C226 and C613 (46). It therefore seems likely that although the C226 and C613 residues of Keap1 are far removed from each other in the primary structure, they are situated close to each other in the native protein. Weight is given to this view by the finding that the IVR domain of Keap1 partially envelops the six-bladed β-propeller formed by the Kelch and CTR domains in the folded protein (54). The notion that C226 and C613 in Keap1 are in juxtaposition in the native protein is also supported by the finding that ex vivo treatment with H₂O₂ results in formation of an intramolecular disulfide bridge between C226 and C613 (15), but the functional consequence of this oxidative modification of Keap1 is unknown.

Although xenobiotic-inducing agents are commonly used as a means to activate Nrf2 via modification of Keap1 thiols (41, 62, 64), the ability of endogenous signaling molecules to inhibit Keap1 has attracted little attention. Endogenous molecules that modify Keap1 and stimulate Nrf2 activity include 4-hydroxynonenal, 15-deoxy-Δ^12,14-prostaglandin J₂, nitrofatty acids, and 8-nitro-cGMP (18, 28, 30, 38), as well as NO and H₂O₂ (3, 10, 15, 46, 71). An example of a relatively poorly characterized second messenger that may regulate the Keap1-Nrf2 pathway is the gaseous signaling molecule hydrogen sulfide (H₂S), which is enzymatically generated in a pyridoxal-5′-phosphate-dependent manner by cystathionine β-synthase (CBS) and cystathionine γ-lyase (CSE) (6, 19, 40) and in a pyridoxal-5′-phosphate-independent manner by 3-mercaptopyruvate sulfur transferase (3MST) (60). Specifically, the H₂S gasotransmitter exerts antioxidant properties (21, 33, 34, 67). Evidence has been provided that it stimulates nuclear accumulation of Nrf2, and that preconditioning with H₂S confers protection against ischemia (4), but the mechanism underpinning such activity has not been elucidated. It is plausible that H₂S modifies and inactivates Keap1, causing Nrf2 accumulation and induction of ARE-driven genes, because the gasotransmitter has been reported to S-sulfhydrate Cys residues in proteins (51), though this has yet to be tested.

Results

The H₂S donor NaHS activates Nrf2

H₂S is unstable and volatile (70), and it is not practicable to expose cells to a sustained concentration of the gasotransmitter. Therefore, to determine its effect on Nrf2, mouse embryonic fibroblast (MEF) cells were treated with a single 400 μM dose of the H₂S donor NaHS before they were lysed 0–120 min later; the isothiocyanate electrophile SFN was used as a positive control. Figure 1A shows that NaHS stimulated a rapid increase in the steady-state level of Nrf2 protein. Using 60 min after administration of NaHS as an end-point, Figure 1B shows that the gasotransmitter elicited a dose-dependent increase in Nrf2 protein. To explore whether H₂S can delay destruction of Nrf2 protein, fibroblasts were treated with cyclohexamide (CHX) to block protein translation for various lengths of time after treatment with NaHS; in these experiments, fibroblasts were treated with a single dose of NaHS, and 60 min later, they were treated with CHX for various time intervals before Nrf2 protein was measured by Western blotting. As shown in Figure 1C and D, treatment of MEF cells with NaHS stimulated a delay in the destruction of Nrf2 protein when compared with vehicle-treated fibroblasts.

FIG. 1.

The H₂S donor NaHS activates the Nrf2 pathway. (A) MEF cells were treated with a single bolus of NaHS to give a final concentration of 400 μM, and lysed in RIPA buffer at various times thereafter. The amount of Nrf2 protein in the lysates was measured by Western blotting (43), and Gapdh was used as a loading control. (B) MEF cells were treated with a single 0–400 μM dose of NaHS, and 60 min later, they were lysed in RIPA buffer. Thereafter, Nrf2 protein in the lysates was measured by Western blotting, with Gapdh serving as a loading control. (C) The effect of H₂S on Nrf2 protein stability was assessed by pretreating MEFs with 400 μM NaHS; 60 min before, de novo protein translation was blocked using cyclohexamide (CHX, final dose 10 μM) for various lengths of time (0, 30, 60, and 120 min). Lysates were prepared, and Western blotting for Nrf2 was carried out as described above. (D) The graph depicts the natural logarithm of the relative abundance of Nrf2 measured by densitometry of the immunoblots as a function of CHX (mean of three experiments). (E) MEF cells were treated with a single 0–400 μM dose of NaHS, and 18 h later, Nqo1 mRNA levels were measured using TaqMan RT-PCR as described previously (53). The levels of Nqo1 mRNA were normalized against Actin; fibroblasts were treated with 3 μM SFN for 18 h as a positive control. Significant increases in Nqo1 mRNA, relative to that of vehicle-treated MEFs, with p-values of 0.05–0.01 or <0.001 are indicated above the histogram bars by double (**) or triple (***) asterisk signs, respectively. (F) Fibroblasts were treated with a single 0–400 μM dose of NaHS, and 24 h later, Nqo1 protein levels were measured by immunoblotting (31); treatment with 3 μM SFN for 24 h served as a positive control. H₂S, hydrogen sulfide; MEF, mouse embryonic fibroblast; Nrf2, NF-E2 p45-related factor 2; Nqo1, NAD(P)H:quinone oxidoreductase-1; RIPA, radioimmunoprecipitation assay; RT-PCR, real-time polymerase chain reaction; SFN, sulforaphane.

To establish if the increase in Nrf2 protein results in induction of its target genes, fibroblasts were treated with increasing concentrations of a single dose of NaHS before they were lysed 18 h later and the levels of mRNA for NAD(P)H:quinone oxidoreductase-1 (Nqo1) measured by TaqMan^™ real-time polymerase chain reaction (RT-PCR). Figure 1E shows that the Nqo1 mRNA levels increased in a dose-dependent manner when wild-type MEF cells were challenged with increasing concentrations of NaHS. Consistent with the TaqMan results, treatment of the fibroblasts with NaHS increased the level of Nqo1 protein, as measured 24 h after administration of the H₂S donor (Fig. 1F). Collectively, these data indicate that H₂S stabilizes Nrf2 protein and stimulates transactivation of the Nrf2 target gene Nqo1.

H₂S protects cells from oxidative stress in an Nrf2-dependent manner

To test whether H₂S protects cells from oxidative stress, fibroblasts were pretreated with NaHS, and 24 h later, they were exposed to the redox-cycling compound menadione for a further 24 h. Using the MTT assay, pretreatment of wild-type MEF cells with NaHS increased the lethal dose 50 (LD₅₀) of menadione from 16 to 23 μM, indicating that H₂S can protect cells from oxidative stress (Fig. 2A). The role of Nrf2 in mediating protection by H₂S against oxidative stress was assessed by pretreating Nrf2^−/− MEF cells with NaHS before they were exposed to menadione. This showed that the mutant fibroblasts were substantially more sensitive to menadione than their wild-type counterparts, with an LD₅₀ of ∼3.8 μM, and that pretreatment with NaHS did not influence their sensitivity to the redox-cycling agent (Fig. 2B). Therefore, the cytoprotection conferred by H₂S against menadione is dependent on Nrf2.

FIG. 2.

H₂S protects cells from oxidative stress in an Nrf2-dependent manner. (A) Wild-type fibroblasts were either pretreated with a single bolus of NaHS to give a final concentration of 400 μM (■, labeled right hand side as NaHS) or untreated (O, labeled right hand side as Control) 24 h before being exposed to increasing concentrations of menadione for a further 24 h. Thereafter, the viability of the fibroblasts was assessed by the MTT cytotoxicity assay. Significant increases in survival of fibroblasts that had been treated with NaHS, compared with vehicle-treated fibroblasts, with p-values of<0.001 are indicated by triple-asterisk (***) signs beside the data points. (B) The ability of Nrf2^−/− fibroblasts to withstand menadione was examined as above for the wild-type MEFs. (C) Nrf2^+/+ and Nrf2^−/− fibroblasts were either treated with a single 400 μM dose of NaHS (labeled NaHS) or untreated (labeled Control). Twenty-four hours later, the cells were exposed to 23 μM menadione for 8 h, before they were incubated with H₂DCF-DA to allow the level of ROS to be estimated. The significance in the differences in ROS levels is indicated as follows: p≤0.05 (*), p≤0.01 (**), and not significant (ns). H₂DCF-DA, 2′,7′-dichlorodihydrofluorescein diacetate; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; ROS, reactive oxygen species.

We next examined whether the ability of H₂S to protect against menadione entails suppression of ROS levels. To this end, we pretreated Nrf2^+/+ and Nrf2^−/− fibroblasts with NaHS before exposing them 24 h later to 23 μM menadione for a further 24 h; the concentration of menadione was chosen as it corresponds to the LD₅₀ of the compound in wild-type fibroblasts that had been pretreated with NaHS. As is evident in Figure 2C, pretreatment of wild-type fibroblasts with NaHS before exposure to menadione for 24 h diminished the amount of ROS generated by the redox-cycling agent to levels that were not statistically different from those under basal conditions (i.e., cells that had been treated with neither NaHS nor menadione). To examine if Nrf2 was involved in the H₂S-stimulated decrease in ROS, similar experiments were performed in Nrf2-null fibroblasts. Although pretreatment of the mutant fibroblasts with NaHS decreased modestly the amount of ROS created by menadione, these were substantially higher than the basal levels in Nrf2-null MEFs. By contrast, pretreatment of wild-type fibroblasts with NaHS decreased the amount of ROS created by menadione to levels that were statistically indistinguishable from those observed under basal conditions. Thus, H₂S stimulated a reduction in ROS levels and protected cells from oxidative stress in an Nrf2-dependent manner.

Nrf2 regulates the expression of Cbs and Cse that are involved in H₂S synthesis

To examine whether Nrf2 contributes to the regulation of murine Cbs, wild-type and Nrf2-null MEFs were grown under normal conditions or were treated with a single dose of 400 μM NaHS 18 h before the Cbs mRNA and 24 h before Cbs protein levels were examined. No significant difference was observed in the basal amount of Cbs mRNA in untreated Nrf2^+/+ MEFs and Nrf2^−/− MEFs as measured by TaqMan RT-PCR (Fig. 3A). By contrast, treatment of Nrf2^+/+ MEFs with NaHS produced a 2.2-fold increase in the Cbs mRNA levels relative to that in untreated wild-type fibroblasts, but this increase was not apparent in the Nrf2-null MEFs. The level of Cbs protein was also increased in Nrf2^+/+ fibroblasts treated with NaHS, but this was not observed in the mutant cells (Fig. 3B). Thus, H₂S regulates Cbs expression in an Nrf2-dependent manner.

FIG. 3.

Nrf2 regulates the expression of the H₂S biosynthetic enzymes Cbs and Cse. (A) Nrf2^+/+ and Nrf2^−/− fibroblasts were treated with a 400 μM bolus of NaHS, and 18 h later, the level of Cbs mRNA was analyzed using TaqMan; actin was used as a control. The significance of the increase in mRNA levels in NaHS-treated cells is indicated as follows: p=0.05–0.01 (**); not significant (ns). (B) Nrf2^+/+ and Nrf2^−/− fibroblasts were treated with 400 μM NaHS. Twenty-four hours later, the cells were lysed in RIPA buffer, and the level of Cbs protein was measured by immunoblotting. Gapdh served as a loading control. The relative amount of Cbs protein in each lane, determined by densitometry, is indicated below the blot. (C) The ability of NaHS to increase Cse mRNA in Nrf2^+/+ and Nrf2^−/− MEFs was examined by RT-PCR with a Cse-specific TaqMan primer and probe set using the same conditions described above. The significance of the increase in mRNA levels in NaHS-treated cells is indicated as follows: p≤0.001 (***); not significant (ns). (D) The ability of NaHS to increase Cse protein in the wild-type and mutant fibroblasts was examined by immunoblotting 24 h after treatment with 400 μM NaHS. Gapdh served as a loading control. The relative amount of Cse protein in each lane, determined by densitometry, is indicated below the blot. (E) The ability of a single dose of 400 μM NaHS to increase mRNA for 3Mst was examined by TaqMan RT-PCR in a similar way as described above for Cbs and Cse. 3Mst, 3-mercaptopyruvate sulfur transferase.

Similar experiments, to those described above, were performed to examine whether Nrf2 regulates Cse. As shown in Figure 3C, the Cse mRNA levels did not differ between untreated Nrf2^+/+ and Nrf2^−/− MEF cells. However, NaHS increased Cse mRNA 3.2-fold in Nrf2^+/+ fibroblasts, but not in their Nrf2^−/− counterparts. Western blotting similarly showed that NaHS increased Cse protein modestly in Nrf2^+/+ MEFs, but not in Nrf2^−/− MEFs (Fig. 3D). These results indicate that induction of Cse by H₂S requires Nrf2.

We also measured the levels of mRNA for murine 3Mst in Nrf2^+/+ and Nrf2^−/− MEFs by TaqMan RT-PCR. This revealed no difference in 3Mst mRNA levels in fibroblasts from wild-type and mutant mice. Moreover, treatment with NaHS did not increase the expression of 3Mst in either wild-type or mutant fibroblasts.

Nrf2 regulates the expression of Sqrdl that contributes to the removal of H₂S

We next examined whether Nrf2 controls the expression of the gene Sqrdl, which encodes the enzyme Sulfide:quinone oxidoreductase (Sqr), which catalyzes the oxidation of H₂S. No significant difference in Sqr mRNA was apparent in untreated Nrf2^+/+ and Nrf2^−/− MEF cells (Fig. 4A). However, treatment of the wild-type fibroblasts with NaHS produced a 2.0-fold increase in Sqr mRNA when compared with untreated fibroblasts, whereas no difference was observed in Nrf2^−/− fibroblasts. The level of Sqr protein was higher in wild-type MEFs grown under basal conditions than in their Nrf2 knockout counterparts (Fig. 4B), though this was not evident at an mRNA level. Moreover, treatment with NaHS produced an increase in Sqr protein in Nrf2^+/+ MEFs, but not in Nrf2^−/− MEFs. These data indicate that H₂S regulates Sqrdl expression in an Nrf2-dependent manner.

FIG. 4.

Nrf2 regulates the expression of the H₂S-oxidizing enzyme Sqr. (A) This panel shows the Sqr mRNA levels in Nrf2^+/+ and Nrf2^−/− MEF cells, determined using TaqMan RT-PCR, 18 h after they had been treated with either vehicle control or 400 μM NaHS. The results were normalized to the levels of actin mRNA. The significance of the increase in mRNA levels in NaHS-treated cells is indicated as follows: p≤0.05 (*); not significant (ns). (B) The level of Sqr protein in Nrf2^+/+ and Nrf2^−/− fibroblasts 24 h after they had been treated with either vehicle control or 400 μM NaHS is shown, and the relative amount of Sqr protein that was estimated by densitometry is indicated below the blot. An immunoblot for Gapdh was used as a loading control. Also, a Western blot for cytochrome c oxidase (Cox-IV) is shown to demonstrate that the abundance of mitochondria in lysates from Nrf2^+/+ and Nrf2^−/− fibroblasts is comparable. Sqr, Sulfide:quinone oxidoreductase.

H₂S triggers the formation of disulfide bonds in the dimeric Keap1 protein

We hypothesized that H₂S increases the abundance of Nrf2 protein, because it inactivates Keap1 by a mechanism that entails modification of thiols in the BTB-Kelch substrate adaptor. To investigate whether this notion is correct, we used nonreducing sodium dodecyl sulfate–polyacrylamide-gel electrophoresis (SDS-PAGE) to identify oxidized forms of Keap1. Simultaneously, we used conventional SDS-PAGE to analyze the abundance of Nrf2 in the same samples. In our experiments, we cotransfected COS1 cells with expression plasmids encoding untagged Keap1 and V5-tagged Nrf2 for 24 h, after which they were treated with a single dose of 400 μM NaHS and left for a further 0–360 min before examination. At appropriate time intervals after treatment with NaHS, cell extracts were prepared in Redox Lysis Buffer (defined in the Materials and Methods section) to allow subsequent resolution of oxidized and reduced protein by nonreducing SDS-PAGE, followed by identification of the various Keap1 forms through Western blotting. On all occasions, the amount of reduced Nrf2 protein in the same samples was measured by immunoblotting after they had first been treated with 2-mercaptoethanol, to give a final concentration of 6% by volume, before SDS-PAGE.

Nonreducing SDS-PAGE revealed that treatment of cells with NaHS stimulated the formation of several electrophoretically distinct Keap1 protein bands (Fig. 5A). We have labeled these bands Inter-1, Inter-2, Intra-1 and Reduced, based on the fact that a similar Keap1 electrophoretic pattern has been reported previously by Toledano and his colleagues when they used nonreducing SDS-PAGE to analyze ectopic Keap1 in HeLa cells that had been treated with H₂O₂ (15): our Inter-1, Inter-2, Intra-1, and Reduced Keap1 electrophoretic bands appear, respectively, equivalent to the Oxidized Intermolecular (OxIR) 1, OxIR2, Oxidized Intramolecular (OxIM), and Red bands described by Toledano and colleagues. Specifically, the Inter-1/OxIR1 and Inter-2/OxIR2 electrophoretic bands migrate more slowly during SDS-PAGE than the reduced Keap1 polypeptide, and represent a protein containing an intermolecular disulfide bridge between C151 of two different subunits: the C151 residue is located in the BTB dimerization domain of Keap1, and its involvement in formation of the Inter-1 and Inter-2 bands is concluded from the fact that neither band is observed when the Keap1^C151S mutant is subjected to nonreducing SDS-PAGE after treatment with H₂O₂ [see Ref. (15) and Fig. 5B below]. By contrast, the Intra-1/OxIM electrophoretic band migrates more quickly during SDS-PAGE than the reduced Keap1 polypeptide, and represents a protein containing an intramolecular disulfide bond between the C226 and C613 residues within a Keap1 subunit, as deduced from the fact that the band is not observed when either Keap1^C226S or Keap1^C613S mutants are examined by nonreducing SDS-PAGE [see Ref. (15) and Fig. 6A below].

FIG. 5.

H₂S triggers the formation of disulfide bonds in Keap1. (A) COS1 cells were transfected with pcDNA3.1 expression vectors for V5-tagged Nrf2 and untagged Keap1 along with a β-gal expression plasmid. After 18-h recovery from transfection, COS1 cells were exposed to a single bolus of 400 μM NaHS for the indicated times (0–360 min) before they were harvested in Redox Lysis Buffer, which contained 40 mM N-ethylmaleimide to block free thiol groups in proteins. To examine the disulfide bonds in Keap1, cell lysates prepared in Redox Lysis Buffer were subjected to nonreducing SDS-PAGE, followed by immunoblotting using an antibody against mouse Keap1 (blot shown at the top of panel). The positions of the Inter-1, Inter-2, and Intra-1 Keap1 bands relative to the reduced polypeptide are shown by arrows on the right hand side. The loading of samples was normalized using β-gal activity. Portions of the lysates from COS1 cells transfected with expression constructs for V5-tagged Nrf2 and untagged Keap1 that had been prepared in Redox Lysis Buffer were reduced by addition of 2-mercaptoethanol, to a final concentration of 6% by volume, and the abundance of ectopic Nrf2 determined using an antibody against V5. The immunoblot for Nrf2 is shown immediately below that for Keap1. (B) COS1 cells were transfected with the V5-tagged Nrf2 expression plasmid either alone or in combination with expression constructs for either untagged wild-type Keap1 (WT) or the Keap1^C151S mutant (C151S). After 18-h recovery from transfection, the cells were treated with a single dose of 400 μM NaHS, and 30 min later, they were lysed in Redox Lysis Buffer. As above, the extent of disulfide bond formation in Keap1 was examined using nonreducing SDS-PAGE followed by immunoblotting, and the level of Nrf2 was examined by Western blotting using an anti-V5 antibody to probe proteins that had been reduced before SDS-PAGE. All cells were cotransfected with an expression plasmid for β-gal, and protein levels were normalized to β-gal activity. (C) RL34 cells were transfected with either the wild-type P _−1016/Nqo1-Luc reporter plasmid or the mutant P _−1016/Nqo1-Luc(mut) plasmid that lacks an ARE sequence (53), each with pRL-TK Renilla, along with either untagged wild-type Keap1 or Keap1^C151S expression constructs (46). After recovery from transfection, the RL34 cells were treated with 400 μM NaHS, and 6 h later, luciferase reporter activity was measured; the results calculated were normalized against Renilla activity in all cases. As a positive control, cells were treated with 30 μM tBHQ for 8 h. Luciferase results that were significantly higher in NaHS-treated samples than vehicle-treated samples with p-values of 0.01–0.001 are indicated with double-asterisk (**) signs, whereas those indicated as not significant (ns) did not differ. ARE, antioxidant-response element; Keap1, Kelch-like ECH-associated protein-1; SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; tBHQ, tert-butylhydroquinone.

FIG. 6.

H₂S activates Nrf2 by inhibiting Keap1 through its C226 and C613 residues. (A) COS1 cells were transfected with an expression construct encoding V5-tagged Nrf2 along with those for untagged wild-type Keap1 (WT), Keap1^C226S (C226S), Keap1^C613S (C613S), or Keap1^C226S,C613S (C226S,C613S) proteins. After 18-h recovery from transfection, the COS1 cells were treated for 30 min with a single dose of 400 μM NaHS. Lysates were processed as described in Figure 5A. All cell transfection mixtures contained a β-gal plasmid, and the samples applied to the SDS-PAGE gel were normalized to their β-gal activity. (B) RL34 cells were transfected with either wild-type P _−1016/Nqo1-Luc or mutant P _−1016/Nqo1-Luc(mut) and pRL-TK Renilla reporter plasmids (53) along with expression constructs for wild-type Keap1, Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613Sproteins (46). After recovery, the transfected cells were treated with 400 μM NaHS, and 6 h later, they were harvested and lysed in a non-denaturing passive lysis buffer; treatment with 30 μM tBHQ for 8 h was employed as a positive control. Luciferase activity was normalized to Renilla activity in each case. Luciferase results from NaHS-treated RL34 cells expressing mutant Keap1 that were lower than that of similarly treated RL34 cells expressing wild-type Keap1 with p-values≤0.001 are indicated with triple-dollar ($$$) signs.

In our experiments, the Inter-1 and Inter-2 Keap1 bands were not present under basal conditions, but appeared within 10 min of treatment with NaHS (Fig. 5A). The Inter-1 and Inter-2 bands were however relatively short-lived and were no longer apparent 240 min after administration of a single dose of NaHS. Furthermore, we found that the Inter-1 Keap1 band was relatively more abundant than the Inter-2 band, which was also the case with the OxIR1 and OxIR2 bands in the experiments described by Fourquet et al. (15). By contrast with the Inter-1 and Inter-2 Keap1 bands, the Intra-1 band was present under basal conditions and was most abundant after treatment with NaHS for between 60 and 120 min. After treatment with NaHS, the increase in the abundance of ectopic Nrf2 protein mirrored most closely the increase in intensity of the Intra-1 band. It was noted that the appearance of Inter-1 and Inter-2 seemed to precede the increase in Nrf2 protein. Moreover, Inter-1 and Inter-2 disappeared 60 min before Nrf2 protein returned to basal levels. These observations suggest that a link exists between the formation of disulfide bonds in Keap1 and inhibition of its substrate adaptor activity toward Nrf2.

Keap1 contains a sensor that responds to H₂S and is distinct from that for NO

The gasotransmitter NO has previously been shown to inactivate Keap1 by S-nitrosylation of its C151 residue (46). To determine if the C151 residue in Keap1 represents a general gasotransmitter sensor that also recognizes H₂S, we examined whether the Keap1^C151S mutant was resistant to inactivation by H₂S. Thus, COS1 cells were transfected with Nrf2-V5 alone or jointly with an expression construct for either wild-type Keap1 or Keap1^C151S. After recovery from transfection, the COS1 cells were exposed to a single dose of NaHS for 30 min, harvested in Redox Lysis Buffer, and analyzed as described above. The level of ectopic Nrf2 protein was low in samples from untreated COS1 cells, but it was more abundant in samples from cells that had been challenged with NaHS (Fig. 5B). The increase in the amount of Nrf2 upon NaHS treatment was observed in both samples that expressed ectopic wild-type Keap1 and those that expressed Keap1^C151S. These results demonstrate that Keap1^C151S is inhibited by H₂S and therefore indicate that C151 is not required by the BTB-Kelch protein to sense the gasotransmitter. Figure 5B also shows that mutation of C151 in Keap1 prevents formation of the Inter-1 and Inter-2 disulfide bands, which agrees with the findings of Fourquet et al. (15). As nonreducing SDS-PAGE resolves two electrophoretically distinct forms of Keap1 that appear to contain an intermolecular disulfide bond, it is possible that C151 can form a disulfide bridge with another cysteine in the other subunit besides C151. Candidate residues are likely to reside in the BTB domain, and include C77 or C171, but this possibility has not been explored.

To determine whether failure of Keap1 to form Inter-1 and Inter-2 bands influences its ability to repress Nrf2, ARE luciferase reporter assays were performed in RL34 cells. This also allowed us to monitor the ability of Keap1^C151S to repress Nrf2 after treatment with H₂S; the hydroquinone electrophile tBHQ was used as a positive control. Figure 5C shows that NaHS can induce ARE-driven luciferase activity to a similar extent in RL34 cells expressing wild-type Keap1 and the Keap1^C151S mutant protein. Thus, formation of an intermolecular C151–C151 disulfide bond in Keap1 is not required for H₂S to inactivate the substrate adaptor protein. The fact that Keap1^C151S is resistant to inhibition by NO (46), but can be inactivated by H₂S, indicates that Keap1 senses these two gasotransmitters by different mechanisms.

H₂S upregulates Nrf2 by inhibiting Keap1 through its C226 and C613 residues

To access whether inactivation of Keap1 by H₂S is synonymous with formation of the Intra-1 electrophoretic band, COS1 cells were transfected with an expression vector for Nrf2-V5 along with ones encoding wild-type Keap1, Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S proteins. Under basal conditions, the level of Nrf2 protein in the transfected COS1 cells was low due to Keap1-mediated ubiquitylation (Fig. 6A), but treatment with NaHS resulted in a significant increase in Nrf2 in some, but not all, cases. Importantly, the accumulation of Nrf2 upon NaHS treatment only occurred in cells that expressed ectopic wild-type Keap1 and was not observed in cells expressing the Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S mutant proteins. Therefore, H₂S did not inactivate Keap1 point mutants that lacked C226 and/or C613 residues (i.e., had been replaced with Ser). Furthermore, mutation of either C226 or C613 in Keap1 prevented the formation of the Intra-1 electrophoretic band. These results indicate that inactivation of Keap1 by H₂S requires formation of the C226–C613 disulfide bond.

To examine whether both C226 and C613 in Keap1 are required for H₂S to alleviate repression of Nrf2 by the substrate adaptor, a series of ARE-driven luciferase reporter assays were carried out. Figure 6B shows that the ability of NaHS to induce luciferase activity is severely blunted in cells expressing ectopic Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S proteins. Taken together, the above data are consistent with the hypothesis that the oxidized form of Keap1 that contains the C226–C613 disulfide bridge is inactive and cannot direct ubiquitylation of Nrf2.

The environment of both C226 and C613 contributes to the sensitivity of Keap1 to H₂S

A molecular model of Keap1 around C226 predicted that its sulfhydryl group is orientated in close proximity to S224 and H225 (Fig. 7A). It is reasonable to suppose that S224 and H225 may diminish the pK _a of the C226 thiol group (46, 59), thereby increasing its reactivity. Furthermore, S224, H225, and C226 are conserved in Keap1 proteins from mammals and fish (Fig. 7B). To determine if S224 and H225 in Keap1 contribute to the reactivity of C226, COS1 cells were transfected with an expression construct for Keap1^S224A,H225N. Western blotting showed that mutation of S224 and H225 in Keap1 resulted in the loss of the Intra-1 electrophoretic band, suggesting that these amino acids contribute to the ability of the substrate adaptor to sense H₂S (Fig. 7C).

FIG. 7.

Residues in Keap1 adjacent to C226 and C613 contribute to the H₂S sensor. (A) The structure of Keap1 around C226 was modeled using Swiss-Model software with 3i3nA.pdb as a template. The protein fold of amino acids S224, H225, and C226 within Keap1 is presented, as is the distance from the S-atom of C226. (B) Residues 222–230 of mouse (m) Keap1 are shown aligned with its orthologs in human (h), rat (r), pig (p), and the two in zebrafish (z). The reactive C226 is underlined, whereas S224 and H225 are highlighted in red and blue, respectively. (C) COS1 cells were transfected with expression constructs for either untagged wild-type Keap1 (WT) or Keap1^S224A,H225N (S224A,H225N) protein, and upon recovery, they were treated with 400 μM NaHS for 30 min before lysates were prepared in Redox Lysis Buffer, and samples analyzed by nonreducing SDS-PAGE. All cells were cotransfected with an expression plasmid for β-gal, and the amount of protein for each sample to be analyzed was normalized to β-gal activity. (D) The protein fold of amino acids P612 and C613 within Keap1, which was predicted using Swiss-Model software, is shown. (E) Residues 610–617 of mouse Keap1 are shown aligned with the same orthologs in (B) above. The reactive C613 is underlined, while P612 is highlighted in red. Also shown is the active site of human TRX1. (F) COS1 cells were transfected with expression vectors for either untagged wild-type Keap1 (WT) or Keap1^P612L (P612L) and treated with 400 μM NaHS for 30 min before lysates were prepared in Redox Lysis Buffer, and oxidation of Keap1 was examined by nonreducing SDS-PAGE. All cell transfection mixtures contained a β-gal plasmid, and the amount of protein for each sample was normalized to β-gal activity. TRX, thioredoxin.

A model for Keap1 around C613 indicated that its sulfhydryl group is not orientated near any hydrogen-bonding or basic amino acids, but is located near the pyrrolidine ring of P612 and may alter its position and/or reactivity (Fig. 7D). Moreover, P612 and C613 are conserved in Keap1 across species (Fig. 7E). To test whether P612 in Keap1 is important for the reactivity of C613 and its ability to form a disulfide bond with C226, COS1 cells were transfected with either expression constructs for the wild-type protein or the Keap1^P612L mutant. After overnight recovery, the cells were treated with NaHS and lysates prepared in Redox Lysis Buffer. Immunoblotting showed that mutation of P612 resulted in the loss of the Intra-1 electrophoretic band under basal conditions, and it was still not observed even after treatment with NaHS. These results suggest that P612 influences the ability of C613 to form an intramolecular disulfide bond with C226 (Fig. 7F).

Inhibition of Keap1 by H₂S involves H₂O₂

Recent evidence suggests that H₂S may undergo auto-oxidation to form H₂O₂ (26). To test the notion that H₂O₂ contributes to the actions of H₂S, we examined whether forced expression of catalase might diminish activation of Nrf2 by H₂S. Thus, we transfected COS1 cells with an expression vector for Nrf2-V5 alone, or with vectors for Nrf2-V5 and wild-type Keap1, either with or without a construct encoding catalase. After recovery from transfection, cells were treated with NaHS for 30 min before Nrf2 levels were estimated. It is clear from Figure 8A that the increase in Nrf2 protein caused by NaHS is decreased, but not abolished, by the presence of ectopic catalase. These data suggest that the effect of H₂S on the Keap1-Nrf2 pathway is mediated in part by H₂O₂.

FIG. 8.

Inhibition of Keap1 by H₂S involves the participation of H₂O₂. (A) COS1 cells that expressed Nrf2-V5 alone, or expressed Nrf2-V5 along with either wild-type Keap1 (WT) alone or with both wild-type Keap1 and catalase (WT + Catalase), were treated with 400 μM NaHS for 30 min as indicated, and lysates were prepared in RIPA buffer. The amount of ectopic Nrf2 protein was analyzed by Western blotting utilizing an anti-V5 antibody. (B) COS1 cells that expressed Nrf2-V5 alone or coexpressed Nrf2-V5 with either wild-type Keap1 (WT) or Keap1^C151S (C151S) were treated with 400 μM H₂O₂ as indicated. Nrf2 levels were determined by immunoblotting using an anti-V5 antibody. (C) COS1 cells were transfected with an expression plasmid for Nrf2-V5 alone or with plasmids for Nrf2-V5 and wild-type Keap1, Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S protein before they were treated with 400 μM H₂O₂ as indicated. Nrf2 protein was detected by Western blotting with an anti-V5 antibody. All cells were cotransfected with a β-gal plasmid, and the amount of protein that was subjected to immunoblotting was normalized to β-gal activity.

Keap1 senses NO via C151 (15, 46), and Fourquet et al. inferred that the BTB-Kelch protein also senses H₂O₂ via C151 (15). To evaluate this possibility, we tested whether the Keap1^C151S mutant protein is resistant to inactivation by H₂O₂. COS1 cells were transfected with expression plasmids for Nrf2-V5 alone, or in combination with plasmids for either wild-type Keap1 or Keap1^C151S, before they were treated with H₂O₂ for 30 min, and Nrf2 levels measured. Figure 8B shows that treatment of the transfected cells with H₂O₂ caused stabilization of Nrf2 to the same extent in both wild-type Keap1-expressing cells and Keap1^C151S-expressing cells. Thus, the BTB-Kelch protein does not sense H₂O₂ via its C151 residue. To determine if Keap1 can be inhibited by H₂O₂ through triggering the C226/C613 sensor that responds to H₂S, the same strategy was used as described above, but in this case, the COS1 cells were transfected with expression plasmids encoding Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S proteins rather than Keap1^C151S. As is clear from Figure 8C, mutation of the Keap1 H₂S sensor cysteines to serines prevented H₂O₂ from stimulating an increase in Nrf2 protein. These results indicate that C226 and C613 in Keap1 sense H₂O₂ as well as H₂S.

Keap1 is S-sulfhydrated by H₂S

We used a Biotin-switch technique (BST) method (51) to determine if Keap1 is S-sulfhydrated by H₂S. COS1 cells were transfected with an expression vector for wild-type Keap1, and after recovery from transfection, they were treated with NaHS for 30 min before being harvested in radioimmunoprecipitation assay (RIPA) buffer. The cell lysates were subjected to the BST method, and after elution of biotin-labeled proteins from strepavidin beads, the amount of Keap1 protein pulled down was determined by Western blotting. For unknown reasons, the BST method caused Keap1 to migrate during SDS-PAGE as two bands, with the mobility of the slower band corresponding to a protein of 69 kDa (this is equivalent to the molecular weight of Keap1). Figure 9A shows that treatment of cells with NaHS increased the amount of Keap1 protein in the pull-down samples, indicating that the substrate adaptor is S-sulfhydrated. Interestingly, Keap1 was also found to be present at low levels in pull-down samples from lysates of untreated cells, indicating that it is subject to S-sulfhydration under basal conditions.

FIG. 9.

Keap1 is S -sulfhydrated by H₂S. (A) COS1 cells were transfected with an expression construct for untagged wild-type Keap1 protein. After 18-h recovery from transfection, cells were treated with 400 μM NaHS for 30 min after which they were harvested in RIPA buffer. Cell lysates were analyzed by the BST method as described in the Materials and Methods section. The extent of Keap1 S-sulfhydration was determined by examining the material associated with the Streptavidin beads in the pull-down samples using SDS-PAGE followed by immunoblotting with the Keap1 antibody, shown in the upper blot (labeled Pull down: Biotin), with the total Keap1 shown in the lower blot (labeled Input). (B) COS1 cells were transfected with an untagged wild-type Keap1 expression plasmid. After 18-h recovery from transfection, cells were treated with 400 μM NaHS for 0, 15, 30, 60, and 120 min, after which they were harvested in RIPA buffer. Cell lysates were analyzed by the BST. Western blot analysis of Biotin pull-down samples using the Keap1 antibody indicates that S-sulfhydration of Keap1 occurred in a time-dependent manner. The relative amount of S-sulfhydrated Keap1 protein (in the upper 69-kDa band of the upper blot) that was estimated by densitometry is shown below the blot of the Pull down material. (C) COS1 cells were transfected with an expression construct for untagged wild-type Keap1 (WT) or Keap1^C151S (C151S). After 18-h recovery from transfection, cells were treated with 400 μM NaHS for 30 min after which they were harvested in RIPA buffer. Cell lysates were analyzed by the BST. Western blot analysis was performed to determine the amount of Keap1 that had been pulled down with the Streptavidin beads. (D) COS1 cells were transfected with expression constructs encoding untagged wild-type Keap1 (WT), Keap1^C226S (C226S), Keap1^C613S (C613S), or Keap1^C226S,C613S (C226S,C613S) proteins. After 18-h recovery from transfection, the COS1 cells were treated with a single dose of 400 μM NaHS before lysates were prepared 30 min later. Modified protein in the lysates was identified by the BST method. The extent of Keap1 S-sulfhydration was revealed by the immunoblotting material associated with the Streptavidin beads using antibody against mouse Keap1. BST, Biotin-switch technique.

To further examine S-sulfhydration of Keap1, COS1 cells expressing wild-type Keap1 protein were treated with a single dose of 400 μM NaHS, before they were harvested 15, 30, 60, or 120 min later in the RIPA buffer, and the lysates subjected to BST. As shown in Figure 9B, the amount of Keap1 protein in the pull-down samples at the 15- and 30-min time points was greater than that at 0 min, suggesting that S-sulfhydration of the adaptor protein increases acutely upon exposure to H₂S. However, 60 min after treatment with a single dose of NaHS, the amount of Keap1 protein in the pull-down samples decreased to the level observed at the 0-min time point. This finding indicates that S-sulfhydration of Keap1 by H₂S is a reversible process and suggests that the modified protein is subject to desulfhydration.

To determine whether the NO sensor in Keap1 acts as a site of S-sulfhydration, COS1 cells expressing either wild-type Keap1 protein or Keap1^C151S protein were treated with NaHS, and 30 min later, they were harvested in RIPA buffer and the lysates subjected to BST. As shown in Figure 9C, the amount of Keap1 protein in the pull-down samples did not differ between wild-type Keap1-expressing cells and Keap1^C151S-expressing cells treated with NaHS, suggesting that C151 is not a major site of S-sulfhydration.

We next examined whether C226 and C613 in Keap1 are S-sulfhydrated, because they comprise the H₂S sensor. COS1 cells were transfected with expression vectors for wild-type Keap1, Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S, and after treatment with NaHS were examined using the BST method. Figure 9D shows that the amount of Keap1 protein in pull-down samples from wild-type Keap1-expressing cells increased upon treatment with NaHS when compared with untreated cells. Importantly, the accumulation of Keap1 protein upon NaHS treatment only occurred in cells that expressed ectopic wild-type Keap1 and was not observed in cells expressing the Keap1^C226S, Keap1^C613S, or Keap1^C226S,C613S mutant proteins. These findings indicate that C226 and C613 are the major sites of S-sulfhydration in Keap1, but that such modification can only occur when both cysteines are present.

Discussion

The results reported in the present article have demonstrated that H₂S can induce the classic Nrf2-target gene Nqo1, and that this requires a functional ARE sequence in the Nqo1 gene promoter. We also found that stimulation of ARE-driven gene expression by H₂S is associated with stabilization of the Nrf2 protein. Moreover, evidence is provided that cytoprotection afforded by H₂S against the redox-cycling agent menadione requires activation of Nrf2.

Regulation by Nrf2 of enzymes involved in the metabolism of H₂S and the production of cysteine

It is evident from our in vitro study of MEFs that Nrf2 regulates the expression of Cbs. Consistent with this view, in silico examination of the upstream region of the gene has revealed that it contains the sequence 5′-GTGATCTAGCA-3′ that is likely to function as an ARE to which Nrf2 can bind as a heterodimer with a small musculoaponeurotic fibrosarcoma (Maf) protein. Similarly, we have provided evidence that Nrf2 regulates Cse, and its upstream region contains the sequence 5′-ATGAGGCAGCT-3′ that is likely to serve as an ARE. Our finding that mouse Cse is regulated by Nrf2 is consistent with the recent observations of Hassan et al. (22), who reported that rat CSE is induced by platelet-derived growth factor-BB in an Nrf2-dependent manner. We found no evidence that Nrf2 regulates the levels of mRNA for 3Mst in mouse fibroblasts. However, Nrf2 does regulate human thioredoxin (TRX1) [for a review, see (24)], and as TRX1 is a cofactor for 3MST (48), it is possible that Nrf2 regulates 3MST indirectly.

Our TaqMan and immunoblotting experiments indicate that Cbs and Cse ought to be included as members of the ARE gene battery. In addition to catalyzing the production of H₂S, Cbs and Cse probably contribute to cytoprotection by augmenting levels of cysteine and thus aid the synthesis of GSH. In particular, CBS and CSE catalyze the reverse trans-sulfuration pathway that is responsible for production of cysteine from homocysteine (29, 34, 40). The availability of cysteine seems to be a rate-limiting factor for GSH synthesis, and in accordance with this, Cbs^−/− and Cse^−/− mice have diminished GSH levels and enhanced intracellular levels of ROS (11, 69). We have presented a model that depicts the interplay between Nrf2 and H₂S (Fig. 10A); this includes a role for Nrf2 in mediating the cytoprotective effects of H₂S and in controlling the expression of Cbs and Cse, which in turn stimulate production of both cysteine and the H₂S gasotransmitter.

FIG. 10.

Overview of the effects of H₂S and H₂O₂ on the Keap1–Nrf2 axis. (A) The endogenous molecule H₂S inhibits the substrate adaptor function of Keap1 through modification of C226 and C613, thus alleviating Nrf2 repression by Keap1^CRL. Furthermore, H₂S may also utilize H₂O₂ to inactivate Keap1 (dotted line). Once stimulated, Nrf2 regulates the expression of ARE-driven genes that include Cbs, Cse, and Sqrdl, which encode enzymes involved in H₂S metabolism. While the Cse and Cbs enzymes maintain redox homeostasis through generation of GSH, they are also involved in H₂S production. In turn, H₂S is rapidly removed through the sulfide oxidation pathway that relies on Sqr, encoded by Sqrdl. Therefore, H₂S-activated Nrf2 facilitates a feedback loop that promotes antioxidant defenses while limiting the level of the potentially toxic gas. The cartoon also shows that activation of Nrf2 by H₂S results in induction of Nqo1 that confers protection against redox-cycling quinones. (B) As an alternative to the direct S-sulfhydration of Keap1 by H₂S, the gasotransmitter may give rise to H₂O₂, by auto-oxidation (step 1). In this case, H₂O₂ will oxidize C226 and C613 in Keap1 to yield an intramolecular C226–C613 disulfide bridge (step 2). Subsequently, H₂S may reduce the C226–C613 disulfide in Keap1 to yield a free thiol group and an S-sulfhydrated moiety (step 3). GSH, reduced glutathione.

While low levels of H₂S confer cytoprotection against oxidative stress, it can cause toxicity when present in high concentrations (66). Excessive levels of H₂S can be countered by activation of the sulfide oxidation pathway, which involves the enzyme Sqr (1, 49). As demonstrated in this study, Nrf2 regulates the expression of Sqrdl, which encodes Sqr. In silico analysis of the flanking region of Sqrdl showed that it contains the sequence 5′-ATGACATAGCC-3′, that is likely to function as a binding site for an Nrf2–small Maf heterodimer. Collectively, our data suggest that Nrf2 is a principal regulator of H₂S metabolism, as it regulates genes involved in both its synthesis and its elimination.

Modification of Keap1 by H₂S

An aim of this project was to identify the mechanism by which H₂S increases Nrf2 activity. We have found that the gasotransmitter antagonizes Keap1 activity through modification of reactive cysteines in the substrate adaptor. While our data provide evidence that inactivation of Keap1 by H₂S requires the presence of both C226 and C613, we have also identified the amino acids surrounding these two residues that contribute to the sensitivity of Keap1 to inhibition by H₂S, presumably through lowering the thiol pK _a of C226 and C613 (55, 59). While the reactivity of C226 is dependent upon adjacent basic and hydrogen-bonding amino acids (i.e., S224 and H225), it appears that the reactivity of C613 is dependent upon its adjacent P612 residue. Although, proline is not usually thought to affect redox signaling, it is noteworthy that an active-site proline contributes to the activity of TRX1 (57).

Whether H₂S is capable of reacting directly with the reduced thiol groups of C226 and C613 in Keap1 is not known. This uncertainty arises because H₂S is a weak reducing agent and exhibits little reactivity toward free thiols (9). Given the fact that H₂S readily reacts with oxidized thiols (16), it is possible that the gasotransmitter reacts with C226 and C613, to form S-sulfhydrated Keap1, principally after they have been oxidized to form a disulfide bridge. Consistent with this hypothesis, our experiments involving forced expression of catalase suggest that H₂O₂ makes a significant contribution to activation of Nrf2 by H₂S. In this context, it is notable that H₂O₂ may be generated directly from H₂S by auto-oxidation (26), or indirectly by inhibition of cytochrome c, causing increased mitochondrial production of ROS (66). We therefore speculate that S-sulfhydration of Keap1 may occur by a two-step process in which the C226 and C613 residues are first oxidized by H₂O₂ to form a disulfide bridge before they are subsequently reduced by H₂S (Fig. 10B). Thus, it may be necessary for H₂S to stimulate formation of a C226–C613 disulfide bridge in Keap1 before it is able to S-sulfhydrate the BTB-Kelch protein.

H₂S can S-sulfhydrate Keap1 at C226 and C613, but it is a reversible process. Removal of the S-sulfhydrated group from the thiol indicates a level of regulation, and it will therefore be important to identify the proteins responsible for desulfhydration. Recent work by Krishnan et al. (37) suggests that TRX1 can desulfhydrate PTP1B in an in vitro setting, but the physiological significance of this remains to be determined. Of particular interest, our data indicate that mutation of either C226 or C613 prevents S-sulfhydration of Keap1, which suggests (i) that the BST is not sensitive enough to detect low levels of S-sulfhydrated protein, or (ii) that S-sulfhydration of one cysteine promotes modification of the second cysteine through direct transfer of the persulfide bond (trans-sulfhydration). In this case, both cysteines in Keap1 are required for the modification to occur, but only one cysteine is ultimately S-sulfhydrated. It remains to be determined whether the putative trans-sulfhydration process, which is analogous to transnitrosylation where the nitro group attached to one cysteine is transferred to a second cysteine, occurs in Keap1.

Mechanism by which Keap1 is inhibited through the C226/C613 sensor

We have established previously that Keap1 is inactivated by the noncovalent binding of Zn²⁺, Cd²⁺, As³⁺, and Se⁴⁺ to its C226 and C613 residues through a process that we envisage involves coordination chemistry (46). The present study has demonstrated that Keap1 is inactivated by H₂O₂ through formation of an intramolecular disulfide bridge between C226 and C613. It could be argued that in both cases (i.e., the binding of transition metals and thiol oxidation), the loss of Keap1 activity is associated with restriction in the freedom of movement of the IVR and CTR domains. If this hypothesis is correct, it follows that the mobility of residues in regions of Keap1 that contain C226 and C613 is essential for its substrate adaptor activity. It is possible that flexibility in this area influences the relative positions of the two β-propeller protein-docking sites in the Keap1 dimer, and this in turn may block binding of Nrf2 through both its DLG and ETGE motifs (45). Formation of the C226–C613 disulfide bridge in Keap1 might either prevent it from mediating the ubiquitylation of Nrf2 by Cul3-Rbx1, or prevent the transfer of ubiquitylated Nrf2 to a receptor involved in transport to the proteasome.

If the prediction that loss of mobility in the IVR and CTR is sufficient to trigger the C226/C613 sensor, it seems unlikely that S-sulfhydration of C226 and C613 by H₂S would inhibit Keap1. Indeed, if H₂S cleaves the C226–C613 bond in Keap1 formed by H₂O₂, then H₂S might restore substrate adaptor activity, because the reaction would free the IVR and CTR domains from restriction in their mobility caused by the disulfide bridge. In a related vein, electrophiles that form adducts with C226 or C613 might prevent the sensor being triggered by H₂O₂, because they block formation of the disulfide bridge.

Basal Nrf2 activity and endogenous signaling through the H₂O₂ sensor

Fourquet et al. (15) were the first group of workers to use nonreducing SDS-PAGE to show that Keap1 forms a C151–C151 intermolecular disulfide bond and a C226–C613 intramolecular disulfide bond upon treatment with H₂O₂. These workers assumed that NO and H₂O₂ trigger the same sensor in Keap1, and on the basis of the finding that NO inactivates Keap1 through C151 in the BTB-Kelch protein, they proposed that H₂O₂ also triggers the C151 sensor; in fact, they did not test whether H₂O₂ inactivates Keap1 via C151. Herein, we have demonstrated that inactivation of Keap1 by H₂O₂ occurs independently of C151. Moreover, our study of Keap1 mutants indicates that the BTB-Kelch protein is only sensitive to H₂O₂ inactivation if it possesses both C226 and C613. Our results are therefore consistent with the hypothesis that H₂O₂ inactivates Keap1 by formation of the C226–C613 intramolecular disulfide bond, not through formation of the C151–C151 intermolecular disulfide bond.

Nonreducing SDS-PAGE has revealed that the C226–C613 intramolecular disulfide bond in Keap1 is formed in the absence of exogenous stressors. It therefore appears that the C226/C613 sensor is triggered under basal conditions and may thus be responsible for the low Nrf2 activity that occurs under normal homeostatic conditions. This observation suggests that Keap1 may be subject to a modest level of inhibition by the generation of H₂O₂ that occurs during normal mitochondrial metabolic processes (65). It is notable that in previous cotransfection experiments described by Toledano and his colleagues, knockdown of endogenous thioredoxin reductase (TXNRD1) resulted in a significant increase in the portion of ectopic Keap1 that forms a C226–C613 disulfide bridge, and that this was associated with a marked increase in the relative amount of ectopic Nrf2 (15). These results suggest that inactivation of Keap1 through formation of a C226–C613 disulfide bridge may be rapidly reactivated by reduction of cystine by TRX1, which is in turn reduced by TXNRD1. As TRX1 and TXNRD1 are both members of the ARE gene battery (24), it is evident that an Nrf2-mediated feedback loop may exist between Keap1 and the thioredoxin pathway.

Concluding comments

Transcription factor Nrf2 is a master regulator of cellular redox homeostasis and mediates preconditioning, stimulated by H₂S, to confer protection against ischemia. While Nrf2 is of benefit to normal cells, it is frequently upregulated in nonsmall-cell lung cancer, and to a lesser extent in tumors of the head and neck, liver, ovary, and stomach (5, 72). Tumors in which Nrf2 is constitutively active have a poor prognosis, because the CNC-bZIP factor increases resistance to chemotherapeutic drugs and radiotherapy, increases angiogenesis, and allows higher rates of cell proliferation [reviewed in Refs. (23, 63)]. Upregulation of Nrf2 can arise as a consequence of somatic mutations or epigenetic changes that diminish Keap1 activity. Alternatively, Nrf2 may evade repression by Keap1 as a consequence of somatic mutations in the CNC-bZIP factor, or by overexpression of the Nrf2 gene. It is also possible that mutations in enzymes involved in the synthesis or metabolism of endogenous signaling molecules such as H₂S, H₂O₂, or NO that antagonize Keap1 might contribute to the permanent activation of Nrf2 in tumors. This possibility warrants investigation.

Materials and Methods

Chemicals

These were all readily available commercially, and in all cases, the highest grade of purity was used.

Cell culture

The COS1 and RL34 cells were grown in the Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS), 2 mM l-glutamine, 50 units/ml penicillin, and 50 μg/ml streptomycin, and MEF cells were grown in the Iscove's modified Dulbecco's medium (IMDM) supplemented with 10% (v/v) FBS, 50 units/ml penicillin, 50 μg/ml streptomycin, 6 μg epidermal growth factor (EGF), and 6 ml of a solution containing insulin, transferrin, selenium (ITS); DMEM and IMDM were from Invitrogen. Cells were transiently transfected with Lipofectamine 2000 according to the manufacturer's instructions.

Plasmids

Expression constructs encoding V5-tagged mouse Nrf2 (mNrf2-V5) and untagged mouse Keap1, both in pcDNA3.1, have been described previously (43, 44). Also, the expression constructs encoding mouse Keap1^C226S, Keap1^C613S, and Keap1^C226S,C613S have been described previously (46). The β-galactosidase expression plasmid pCMVβgal, used to correct for transfection efficiency, was from Clontech. The expression construct for mouse Keap1^S224A,H225N was generated using a Multi-site-directed mutagenesis kit (Stratagene) with the following primer: forward, 5′-GAGGAGTTCTTCAACCTGGCGAACTGCCAGCTGGCCACG-3′ (with mutated nucleotides underlined). The Keap1^P612L expression construct was created using a single-site mutagenesis kit (Stratagene) and involved the following primers: forward, 5′-GCCGTCACCATGGAACTCTGTCGGAAGCAAATTGATC-3′ and reverse, 5′-GATCAATTTGCTTCCGACAGAGTTCCATGGTGACGGC-3′. The expression vector for human catalase was provided by Dr. Nicholas K. Tonks (47).

Analysis of mRNA levels by real-time quantitative PCR

Total RNA was isolated from cell lines by standard methods, and cDNA was prepared using the Omniscript Reverse Transcription Kit (Qiagen) according to the manufacturer's instructions. The cDNA was diluted 1:7.5 in ddH₂O and used as a template for real-time quantitative PCR. ABI PRISM 7700 Sequence Detector (Applied Biosystems) was used for analyses; cycling conditions were 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 95°C for 15 s, and finally, 60°C for 1 min. All samples were analyzed in triplicate, and reactions containing no template were also carried out as a negative control. Actin served as an internal reference. Fold induction values were normalized to controls using the comparative threshold cycle values. Most probe and primer sets were purchased from Applied Biosystems (ABI) and were supplied in a single-tube format with the exception of the Nqo1 probe/primer set (forward, 5′-GCAGGATTTGCCTACACAATATGC-3′ and reverse, 5′-AGTGGTGATAGAAAGCAAGGTCTTC-3′) described elsewhere (47). The gene expression primer and probes for Cse, Cse, 3Mst, Sqrdl, and actin were Mm00461247_m1, Mm00460654_m1, Mm00460389, Mm00502443_m1, and Mm00607939_s1, respectively.

Luciferase reporter gene assay

ARE-driven luciferase activity was measured using P _−1016/Nqo1-Luc, as described previously (53); a mutant form of this plasmid in which point mutations have been introduced into its ARE [i.e., P _−1016/Nqo1-Luc(mut)] served as a negative control. The pRL-TK Renilla reporter plasmid was used to correct for transfection efficiency.

Nonreducing SDS-PAGE

To analyze redox-modified Keap1 protein, cells were transfected with expression constructs encoding either the wild-type or the mutant forms of the substrate adaptor. After 18-h recovery, cells were treated with NaHS, or vehicle control, before they were lysed, and the protein was extracted in Redox Lysis Buffer (100 mM Tris, 150 mM NaCl, 0.2% [w/v] deoxycholic acid, 5% [v/v] NP40, 200 μM sodium fluoride, 250 μM EDTA, 100 μM phenylmethylsulfonyl fluoride, and 40 mM N-ethylmaleimide). Proteins from cell lysates were diluted with two volumes of redox sample buffer (200 mM Tris, pH 6.8, 8% [w/v] SDS, 30% [v/v] glycerol, and 0.05% [w/v] bromophenol blue) before being subjected to Tris–glycine SDS-PAGE in 6% (w/v) polyacrylamide gels. For the analysis of reduced Nrf2 that was done in parallel with nonreduced Keap1, the samples prepared in Redox Lysis Buffer were allowed to react with 6% (v/v) 2-mercaptoethanol for 10 min at 95°C before electrophoresis.

Immunoblotting

For most immunoblotting experiments, cells were lysed in RIPA buffer (50 mM Tris, 150 mM NaCl, 1% [v/v] NP40, 0.5% [w/v] deoxycholic acid, and 0.1% [w/v] SDS) and allowed to react with 6% (v/v) 2-mercaptoethanol for 10 min at 95°C. Nonreduced proteins (in the Redox Lysis Buffer) and reduced proteins (in the RIPA buffer) that had been resolved by SDS-PAGE were transferred electrophoretically onto polyvinylidene fluoride membranes and probed with antibodies using standard methods. Antibodies against rat NQO1, mouse Nrf2, and mouse Keap1 have been described previously (31, 44, 45). Antibodies against CBS were provided by Dr. Ruma Banerjee (University of Michigan, Ann Arbor, USA) and have been described (29). The other antibodies were obtained commercially as follows: anti-GAPDH (Sigma), anti-CSE (Abnova), and anti-SQR and anti-Cox-IV (Abcam). ImageJ software was used to determine the densitometry of immunblots relative to the loading control. The density of each protein band was expressed as a fold change of the vehicle-treated control.

Detection of ROS

MEF cells, grown until they were 80%–90% confluent, were washed with PBS and incubated with 10 μM 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCF-DA) in Hanks Buffered Salt Solution (HBSS; Gibco) for 30 min at 37°C in the dark. After incubation, the fibroblasts were washed 3×5 ml HBSS before 1 ml of HBSS was added to each plate. Fluorescence was measured by excitation at 485 nm and emission at 538 nm. Cells were lysed, and fluorescence was normalized to protein levels.

Cell viability

Cell viability was assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (25, 50). Cells were challenged with xenobiotics in triplicate wells for 24 h, after which 20 μl MTT (5 mg/ml in PBS) was added to the culture medium to achieve a final concentration of 1 mg/ml. Cells were incubated at 37°C for 60 min before being lysed in the MTT lysis buffer and incubated at room temperature for 8 h. The formation of formazan was measured by absorbance at 570 nm.

Determination of the half-life of Nrf2

MEF cells were treated with NaHS for 1 h before treatment with CHX (final concentration 40 μg/ml) for between 0 and 120 min. Cell lysates were prepared, and the amount of Nrf2 protein was determined by Western blot. Densitometry analysis was carried out to calculate band intensities, and the relative amount of Nrf2 was plotted on a semi-log graph as described elsewhere (43).

Biotin-switch technique

To determine the extent of protein S-sulfhydration, the BST method was used (51). Duplicate dishes of COS1 cells that had been transfected with expression vectors for Keap1 were allowed to recover for 18 h before they were treated with NaHS for various periods of time. Thereafter, they were subjected to a modified BTS method. Briefly, cells were lysed in the RIPA buffer (50 mM Tris-Cl, pH 7.4, 150 mM NaCl, 1% [v/v] Nonidet P-40, 0.5% [w/v] deoxycholic acid, 0.1% [w/v] SDS) supplemented with 10 mM S-methyl methanethiosulfonate (MMTS) to alkylate free thiols. After centrifugation of lysates, the supernatant was added to 1.6 ml of HEN buffer (100 mM Hepes, 1 mM EDTA, and 0.1 mM neocuproine [pH 8.0]), supplemented with 2.5% (w/v) SDS and 20 μM MMTS, and incubated at 20°C for 18 min to complete the blocking of free thiols. The MMTS was removed from protein samples by acetone precipitation, and proteins were resuspended in the HEN buffer supplemented with 1% SDS. Thereafter, proteins were incubated with 4 mM Biotin–HPDP (Thermo scientific) for 3 h at room temperature to label S-sulfhydrated thiols. Biotinylated, and thus S-sulfhydrated, proteins were precipitated by streptavidin–agarose beads and eluted in an SDS-PAGE sample buffer.

Bioinformatics and statistical analyses

Swiss-Model software was used for molecular modeling of mKeap1 with 3i3nA.pdb as a template. Throughout this study, GraphPad Prism 4 software was used for statistical analysis. The luciferase data were analyzed using Student's t-test, whereas TaqMan data were analyzed using one-way analysis of variance, followed by Newman–Keuls multiple comparison test. Significant increases are expressed as p≤0.001 (***); p≤0.01 (**); p≤0.05 (*); and not significant. Significant decreases are similarly expressed using the dollar ($) symbol.

Footnotes

Acknowledgments

We thank Dr. Ruma Banerjee for providing antibodies against CBS, and Dr. Nick Tonks for providing an expression vector for catalase. We are most grateful to Dr. Simon Fourquet and Dr. Michel Toledano for helpful comments during the initial stages of this project. We are indebted to Dr. Mike McMahon and Dr. Albena Dinkova-Kostova for their expert help and critical suggestions throughout this project. This work was funded in part by a Biotechnology and Biological Sciences Research Council Ph.D. studentship (BB/E528995/1) under the BBSRC Industrial CASE Partnership scheme and by Cancer Research-UK (C4909/A13786).

Author Disclosure Statement

The authors declare no conflict of interest.

Abbreviations Used

†

The ARE has also been called an electrophile-response element (EpRE).

‡

The numbering of amino acids in Keap1 is based on the human protein.

References

Ackermann

, Kubitza

, Maier

, Brawanski

, Hauska

, Piña

. The vertebrate homolog of sulfide-quinone reductase is expressed in mitochondria of neuronal tissues. Neuroscience, 199:1–12. 2011.

Brigelius-Flohé

, Flohé

. Basic principles and emerging concepts in the redox control of transcription factors. Antioxid Redox Signal, 15:2335–2381. 2011.

Buckley

, Li

, Whorton

. Keap1 modification and nuclear accumulation in response to S-nitrosocysteine. Free Radic Biol Med, 44:692–698. 2008.

Calvert

, Jha

, Gundewar

, Elrod

, Ramachandran

, Pattillo

, Kevil

, Lefer

. Hydrogen sulphide mediates cardioprotection through Nrf2 signaling. Circ Res, 105:365–374. 2009.

Cancer Genome Atlas Research Network. Hammerman

, Hayes

, Wilkerson

, Schultz

, Bose

, Chu

, Collisson

, Cope

, Creighton

, Getz

, Herman

, Johnson

, Kucherlapati

, Ladanyi

, Maher

, Robertson

, Sander

, Shen

, Sinha

, Sivachenko

, Thomas

, Travis

, Tsao

, Weinstein

, Wigle

, Baylin

, Govindan

, Meyerson

. Comprehensive genomic characterization of squamous cell lung cancers. Nature, 489:519–525. 2012.

Chen

, Jhee

, Kruger

. Production of the neuromodulator H₂S by cystathionine beta-synthase via the condensation of cysteine and homocysteine. J Biol Chem, 279:52082–52086. 2004.

Chorley

, Campbell

, Wang

, Karaca

, Sambandan

, Bangura

, Xue

, Pi

, Kleeberger

, Bell

. Identification of novel NRF2-regulated genes by ChIP-Seq: influence on retinoid X receptor alpha. Nucleic Acids Res, 40:7416–7429. 2012.

Chowdhury

, Mo

, Gao

, Kazi

, Fisher

, Feinstein

. Oxidant stress stimulates expression of the human peroxiredoxin 6 gene by a transcriptional mechanism involving an antioxidant response element. Free Radic Biol Med, 46:146–153. 2009.

Das

, Huie

, Neta

, Padmaja

. Reduction potential of the sulfhydryl radical: pulse radiolysis and laser flash photolysis studies of the formation and reactions of •SH and HSSH•⁻ in aqueous solutions. J Phys Chem A, 103:5221–5226. 1999.

10.

Dhakshinamoorthy

, Porter

. Nitric oxide-induced transcriptional up-regulation of protective genes by Nrf2 via the antioxidant response element counteracts apoptosis of neuroblastoma cells. J Biol Chem, 279:20096–20107. 2004.

11.

Dickhout

, Carlisle

, Jerome

, Mohammed-Ali

, Jiang

, Yang

, Mani

, Garg

, Banerjee

, Kaufman

, Maclean

, Wang

, Austin

. Integrated stress response modulates cellular redox state via induction of cystathionine γ-lyase: cross-talk between integrated stress response and thiol metabolism. J Biol Chem, 287:7603–7614. 2012.

12.

Dinkova-Kostova

, Holtzclaw

, Cole

, Itoh

, Wakabayashi

, Katoh

, Yamamoto

, Talalay

. Direct evidence that sulfhydryl groups of Keap1 are the sensors regulating induction of phase 2 enzymes that protect against carcinogens and oxidants. Proc Natl Acad Sci U S A, 99:11908–11913. 2002.

13.

Dinkova-Kostova

, Talalay

. Direct and indirect antioxidant properties of inducers of cytoprotective proteins. Mol Nutr Food Res, 52,Suppl 1:S128–S138. 2008.

14.

Eggler

, Liu

, Pezzuto

, van Breemen

, Mesecar

. Modifying specific cysteines of the electrophile-sensing human Keap1 protein is insufficient to disrupt binding to the Nrf2 domain Neh2. Proc Natl Acad Sci U S A, 102:10070–10075. 2005.

15.

Fourquet

, Guerois

, Biard

, Toledano

. Activation of NRF2 by nitrosative agents and H₂O₂ involves KEAP1 disulfide formation. J Biol Chem, 285:8463–8471. 2010.

16.

Francoleon

, Carrington

, Fukuto

. The reaction of H₂S with oxidized thiols: generation of persulfides and implications to H₂S biology. Arch Biochem Biophys, 516:146–153. 2011.

17.

Friling

, Bensimon

, Tichauer

, Daniel

. Xenobiotic-inducible expression of murine glutathione S-transferase Ya subunit gene is controlled by an electrophile-responsive element. Proc Natl Acad Sci U S A, 87:6258–6262. 1990.

18.

Fujii

, Sawa

, Ihara

, Tong

, Ida

, Okamoto

, Ahtesham

, Ishima

, Motohashi

, Yamamoto

, Akaike

. The critical role of nitric oxide signaling, via protein S-guanylation and nitrated cyclic GMP, in the antioxidant adaptive response. J Biol Chem, 285:23970–23984. 2010.

19.

Gadalla

, Snyder

. Hydrogen sulfide as a gasotransmitter. J Neurochem, 113:14–26. 2010.

20.

Halliwell

. Biochemistry of oxidative stress. Biochem Soc Trans, 35:1147–1150. 2007.

21.

Han

, Dong

, Dimitropoulou

, Su

. Hydrogen sulfide ameliorates tobacco smoke-induced oxidative stress and emphysema in mice. Antioxid Redox Signal, 15:2121–2134. 2011.

22.

Hassan

, Boosen

, Schaefer

, Kozlowska

, Eisel

, von Knethen

, Beck

, Hemeida

, El-Moselhy

, Hamada

, Beck

, Pfeilschifter

. Platelet-derived growth factor-BB induces cystathionine γ-lyase expression in rat mesangial cells via a redox-dependent mechanism. Br J Pharmacol, 166:2231–2242. 2012.

23.

Hayes

, McMahon

. NRF2 and KEAP1 mutations: permanent activation of an adaptive response in cancer. Trends Biochem Sci, 34:176–188. 2009.

24.

Hayes

, McMahon

, Chowdhry

, Dinkova-Kostova

. Cancer chemoprevention mechanisms mediated through the Keap1-Nrf2 pathway. Antioxid Redox Signal, 13:1713–1748. 2010.

25.

Higgins

, Kelleher

, Eggleston

, Itoh

, Yamamoto

, Hayes

. Transcription factor Nrf2 mediates an adaptive response to sulforaphane that protects fibroblasts in vitro against the cytotoxic effects of electrophiles, peroxides and redox-cycling agents. Toxicol Appl Pharmacol, 237:267–280. 2009.

26.

Hoffman

, Rajapakse

, Shen

, Gates

. Generation of DNA-damaging reactive oxygen species via the autoxidation of hydrogen sulfide under physiologically relevant conditions: chemistry relevant to both the genotoxic and cell signaling properties of H₂S. Chem Res Toxicol, 25:1609–1615. 2012.

27.

Hong

, Sekhar

, Freeman

, Liebler

. Specific patterns of electrophile adduction trigger Keap1 ubiquitination and Nrf2 activation. J Biol Chem, 280:31768–31775. 2005.

28.

Ishii

, Itoh

, Ruiz

, Leake

, Unoki

, Yamamoto

, Mann

. Role of Nrf2 in the regulation of CD36 and stress protein expression in murine macrophages: activation by oxidatively modified LDL and 4-hydroxynonenal. Circ Res, 94:609–616. 2004.

29.

Kabil

, Vitvitsky

, Xie

, Banerjee

. The quantitative significance of the transsulfuration enzymes for H₂S production in murine tissues. Antioxid Redox Signal, 15:363–372. 2011.

30.

Kansanen

, Bonacci

, Schopfer

, Kuosmanen

, Tong

, Leinonen

, Woodcock

, Yamamoto

, Carlberg

, Ylä-Herttuala

, Freeman

, Levonen

. Electrophilic nitro-fatty acids activate NRF2 by a KEAP1 cysteine 151-independent mechanism. J Biol Chem, 286:14019–14027. 2011.

31.

Kelly

, Ellis

, Manson

, Chanas

, Moffat

, McLeod

, Judah

, Neal

, Hayes

. Chemoprevention of aflatoxin B₁ hepatocarcinogenesis by coumarin, a natural benzopyrone that is a potent inducer of aflatoxin B₁-aldehyde reductase, the glutathione S-transferase A5 and P1 subunits, and NAD(P)H:quinone oxidoreductase in rat liver. Cancer Res, 60:957–969. 2000.

32.

Kensler

, Wakabayashi

, Biswal

. Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway. Annu Rev Pharmacol Toxicol, 47:89–116. 2007.

33.

Kimura

, Dargusch

, Schubert

, Kimura

. Hydrogen sulfide protects HT22 neuronal cells from oxidative stress. Antioxid Redox Signal, 8:661–670. 2006.

34.

Kimura

, Shibuya

, Kimura

. Hydrogen sulfide is a signaling molecule and a cytoprotectant. Antioxid Redox Signal, 17:45–57. 2012.

35.

Kobayashi

, Kang

, Okawa

, Ohtsuji

, Zenke

, Chiba

, Igarashi

, Yamamoto

. Oxidative stress sensor Keap1 functions as an adaptor for Cul3-based E3 ligase to regulate proteasomal degradation of Nrf2. Mol Cell Biol, 24:7130–7139. 2004.

36.

Kobayashi

, Li

, Iwamoto

, Nakajima-Takagi

, Kaneko

, Nakayama

, Eguchi

, Wada

, Kumagai

, Yamamoto

. The antioxidant defense system Keap1-Nrf2 comprises a multiple sensing mechanism for responding to a wide range of chemical compounds. Mol Cell Biol, 29:493–502. 2009.

37.

Krishnan

, Fu

, Pappin

, Tonks

. H₂S-Induced sulfhydration of the phosphatase PTP1B and its role in the endoplasmic reticulum stress response. Sci Signal, 4:ra86. 2011.

38.

Levonen

, Landar

, Ramachandran

, Ceaser

, Dickinson

, Zanoni

, Morrow

, Darley-Usmar

. Cellular mechanisms of redox cell signalling: role of cysteine modification in controlling antioxidant defences in response to electrophilic lipid oxidation products. Biochem J, 378:373–382. 2004.

39.

, Johnson

, Calkins

, Wright

, Svendsen

, Johnson

. Stabilization of Nrf2 by tBHQ confers protection against oxidative stress-induced cell death in human neural stem cells. Toxicol Sci, 83:313–328. 2005.

40.

, Rose

, Moore

. Hydrogen sulfide and cell signaling. Annu Rev Pharmacol Toxicol, 51:169–187. 2011.

41.

Luo

, Eggler

, Liu

, Mesecar

, van Breemen

. Sites of alkylation of human Keap1 by natural chemoprevention agents. J Am Soc Mass Spectrom, 18:2226–2232. 2007.

42.

MacLeod

, McMahon

, Plummer

, Higgins

, Penning

, Igarashi

, Hayes

. Characterization of the cancer chemopreventive NRF2-dependent gene battery in human keratinocytes: demonstration that the KEAP1-NRF2 pathway, and not the BACH1-NRF2 pathway, controls cytoprotection against electrophiles as well as redox-cycling compounds. Carcinogenesis, 30:1571–1580. 2009.

43.

McMahon

, Itoh

, Yamamoto

, Hayes

. Keap1-dependent proteasomal degradation of transcription factor Nrf2 contributes to the negative regulation of antioxidant response element-driven gene expression. J Biol Chem, 278:21592–21600. 2003.

44.

McMahon

, Thomas

, Itoh

, Yamamoto

, Hayes

. Redox-regulated turnover of Nrf2 is determined by at least two separate protein domains, the redox-sensitive Neh2 degron and the redox-insensitive Neh6 degron. J Biol Chem, 279:31556–31567. 2004.

45.

McMahon

, Thomas

, Itoh

, Yamamoto

, Hayes

. Dimerization of substrate adaptors can facilitate cullin-mediated ubiquitylation of proteins by a “tethering” mechanism: a two-site interaction model for the Nrf2-Keap1 complex. J Biol Chem, 281:24756–24768. 2006.

46.

McMahon

, Lamont

, Beattie

, Hayes

. Keap1 perceives stress via three sensors for the endogenous signaling molecules nitric oxide, zinc, and alkenals. Proc Natl Acad Sci U S A, 107:18838–18843. 2010.

47.

Meng

, Buckley

, Galic

, Tiganis

, Tonks

. Regulation of insulin signaling through reversible oxidation of the protein-tyrosine phosphatases TC45 and PTP1B. J Biol Chem, 279:37716–37725. 2004.

48.

Mikami

, Shibuya

, Kimura

, Nagahara

, Ogasawara

, Kimura

. Thioredoxin and dihydrolipoic acid are required for 3-mercaptopyruvate sulfurtransferase to produce hydrogen sulphide. Biochem J, 439:479–485. 2011.

49.

Mimoun

, Andriamihaja

, Chaumontet

, Atanasiu

, Benamouzig

, Blouin

, Tomé

, Bouillaud

, Blachier

. Detoxification of H₂S by differentiated colonic epithelial cells: implication of the sulfide oxidizing unit and of the cell respiratory capacity. Antioxid Redox Signal, 17:1–10. 2012.

50.

Mosmann

. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods, 65:55–63. 1983.

51.

Mustafa

, Gadalla

, Sen

, Kim

, Mu

, Gazi

, Barrow

, Yang

, Wang

, Snyder

. H₂S signals through protein S-sulfhydration. Sci Signal, 2:ra72. 2009.

52.

Nguyen

, Sherratt

, Huang

, Yang

, Pickett

. Increased protein stability as a mechanism that enhances Nrf2-mediated transcriptional activation of the antioxidant response element. Degradation of Nrf2 by the 26 S proteasome. J Biol Chem, 278:4536–4541. 2003.

53.

Nioi

, McMahon

, Itoh

, Yamamoto

, Hayes

. Identification of a novel Nrf2-regulated antioxidant response element (ARE) in the mouse NAD(P)H:quinone oxidoreductase 1 gene: reassessment of the ARE consensus sequence. Biochem J, 374:337–348. 2003.

54.

Ogura

, Tong

, Mio

, Maruyama

, Kurokawa

, Sato

, Yamamoto

. Keap1 is a forked-stem dimer structure with two large spheres enclosing the intervening, double glycine repeat, and C-terminal domains. Proc Natl Acad Sci U S A, 107:2842–2847. 2010.

55.

Paulsen

, Carroll

. Orchestrating redox signalling networks through regulatory cysteine switches. ACS Chem Biol, 5:47–62. 2010.

56.

Pickering

, Linder

, Zhang

, Forman

, Davies

. Nrf2-dependent induction of proteasome and Pa28αβ regulator are required for adaptation to oxidative stress. J Biol Chem, 287:10021–10031. 2012.

57.

Roos

, Garcia-Pino

, Van Belle

, Brosens

, Wahni

, Vandenbussche

, Wyns

, Loris

, Messens

. The conserved active site proline determines the reducing power of Staphylococcus aureus thioredoxin. J Mol Biol, 368:800–811. 2007.

58.

Rushmore

, Morton

, Pickett

. The antioxidant responsive element. Activation by oxidative stress and identification of the DNA consensus sequence required for functional activity. J Biol Chem, 266:11632–11639. 1991.

59.

Salsbury

Jr. , Knutson

, Poole

, Fetrow

. Functional site profiling and electrostatic analysis of cysteines modifiable to cysteine sulfenic acid. Protein Sci, 17:299–312. 2008.

60.

Shibuya

, Tanaka

, Yoshida

, Ogasawara

, Togawa

, Ishii

, Kimura

. 3-Mercaptopyruvate sulfurtransferase produces hydrogen sulfide and bound sulfane sulfur in the brain. Antioxid Redox Signal, 11:703–714. 2009.

61.

Siow

, Mann

. Dietary isoflavones and vascular protection: activation of cellular antioxidant defenses by SERMs or hormesis? Mol Aspects Med, 31:468–477. 2010.

62.

Sporn

, Liby

. NRF2 and cancer: the good, the bad and the importance of context. Nat Rev Cancer, 12:564–571. 2012.

63.

Taguchi

, Motohashi

, Yamamoto

. Molecular mechanisms of the Keap1–Nrf2 pathway in stress response and cancer evolution. Genes Cells, 16:123–140. 2011.

64.

Tanigawa

, Fujii

, Hou

. Action of Nrf2 and Keap1 in ARE-mediated NQO1 expression by quercetin. Free Radic Biol Med, 42:1690–1703. 2007.

65.

Tormos

, Anso

, Hamanaka

, Eisenbart

, Joseph

, Kalyanaraman

, Chandel

. Mitochondrial complex III ROS regulate adipocyte differentiation. Cell Metab, 14:537–544. 2011.

66.

Truong

, Eghbal

, Hindmarsh

, Roth

, O'Brien

. Molecular mechanisms of hydrogen sulfide toxicity. Drug Metab Rev, 38:733–744. 2006.

67.

Tyagi

, Moshal

, Sen

, Vacek

, Kumar

, Hughes

Jr. , Kundu

, Tyagi

. H₂S protects against methionine-induced oxidative stress in brain endothelial cells. Antioxid Redox Signal, 11:25–33. 2009.

68.

Villeneuve

, Lau

, Zhang

. Regulation of the Nrf2-Keap1 antioxidant response by the ubiquitin proteasome system: an insight into cullin-ring ubiquitin ligases. Antioxid Redox Signal, 13:1699–1712. 2010.

69.

Vitvitsky

, Dayal

, Stabler

, Zhou

, Wang

, Lentz

, Banerjee

. Perturbations in homocysteine-linked redox homeostasis in a murine model for hyperhomocysteinemia. Am J Physiol Regul Integr Comp Physiol, 287:R39–R46. 2004.

70.

Vitvitsky

, Kabil

, Banerjee

. High turnover rates for hydrogen sulfide allow for rapid regulation of its tissue concentrations. Antioxid Redox Signal, 17:22–31. 2012.

71.

Wilson

, Gemin

, Espiritu

, Singh

. ets-1 is transcriptionally up-regulated by H₂O₂ via an antioxidant response element. FASEB J, 19:2085–2087. 2005.

72.

Yoo

, Kim

, An

, Lee

. Somatic mutations of the KEAP1 gene in common solid cancers. Histopathology, 60:943–952. 2012.

73.

Zhang

, Hannink

. Distinct cysteine residues in Keap1 are required for Keap1-dependent ubiquitination of Nrf2 and for stabilization of Nrf2 by chemopreventive agents and oxidative stress. Mol Cell Biol, 23:8137–8151. 2003.

74.

Zhang

, Lo

, Cross

, Templeton

, Hannink

. Keap1 is a redox-regulated substrate adaptor protein for a Cul3-dependent ubiquitin ligase complex. Mol Cell Biol, 24:10941–51093. 2004.

75.

Zhang

, Forman

. Reexamination of the electrophile response element sequences and context reveals a lack of consensus in gene function. Biochim Biophys Acta, 1799:496–501. 2010.

The Gasotransmitter Hydrogen Sulfide Induces Nrf2-Target Genes by Inactivating the Keap1 Ubiquitin Ligase Substrate Adaptor Through Formation of a Disulfide Bond Between Cys-226 and Cys-613

Abstract

Introduction

Innovation

Results

The H2S donor NaHS activates Nrf2

H2S protects cells from oxidative stress in an Nrf2-dependent manner

Nrf2 regulates the expression of Cbs and Cse that are involved in H2S synthesis

Nrf2 regulates the expression of Sqrdl that contributes to the removal of H2S

H2S triggers the formation of disulfide bonds in the dimeric Keap1 protein

Keap1 contains a sensor that responds to H2S and is distinct from that for NO

H2S upregulates Nrf2 by inhibiting Keap1 through its C226 and C613 residues

The environment of both C226 and C613 contributes to the sensitivity of Keap1 to H2S

Inhibition of Keap1 by H2S involves H2O2

Keap1 is S-sulfhydrated by H2S

Discussion

Regulation by Nrf2 of enzymes involved in the metabolism of H2S and the production of cysteine

Modification of Keap1 by H2S

Mechanism by which Keap1 is inhibited through the C226/C613 sensor

Basal Nrf2 activity and endogenous signaling through the H2O2 sensor

Concluding comments

Materials and Methods

Chemicals

Cell culture

Plasmids

Analysis of mRNA levels by real-time quantitative PCR

Luciferase reporter gene assay

Nonreducing SDS-PAGE

Immunoblotting

Detection of ROS

Cell viability

Determination of the half-life of Nrf2

Biotin-switch technique

Bioinformatics and statistical analyses

Footnotes

Acknowledgments

Author Disclosure Statement

Abbreviations Used

†

‡

References

The H₂S donor NaHS activates Nrf2

H₂S protects cells from oxidative stress in an Nrf2-dependent manner

Nrf2 regulates the expression of Cbs and Cse that are involved in H₂S synthesis

Nrf2 regulates the expression of Sqrdl that contributes to the removal of H₂S

H₂S triggers the formation of disulfide bonds in the dimeric Keap1 protein

Keap1 contains a sensor that responds to H₂S and is distinct from that for NO

H₂S upregulates Nrf2 by inhibiting Keap1 through its C226 and C613 residues

The environment of both C226 and C613 contributes to the sensitivity of Keap1 to H₂S

Inhibition of Keap1 by H₂S involves H₂O₂

Keap1 is S-sulfhydrated by H₂S

Regulation by Nrf2 of enzymes involved in the metabolism of H₂S and the production of cysteine

Modification of Keap1 by H₂S

Basal Nrf2 activity and endogenous signaling through the H₂O₂ sensor