Sulfane Sulfur in Toxicology: A Novel Defense System Against Electrophilic Stress

Abstract

Electrophiles can undergo covalent modification of cellular proteins associated with its dysfunction, thereby exerting toxicity. Small nucleophilic molecules such as glutathione protect cells from electrophilic insult by binding covalently to electrophiles to form adducts that are excreted into the extracellular space. Recent studies indicate that sulfane sulfur, which is defined as a sulfur atom with 6 valence electrons and no charge, plays an essential role in protection against electrophile toxicity because sulfane sulfur can be highly nucleophilic compared to the corresponding thiol group. Advances in the development of assays to detect sulfane sulfur have revealed that sulfane sulfur-containing molecules such as persulfide/polysulfide species are ubiquitous in cells and tissues. Also, there is growing evidence that the binding of sulfane sulfur to electrophiles forms sulfur adducts as detoxified metabolites. Although the biosynthesis pathways of sulfane sulfur are known, its regulatory function in toxicology is still unclear. This review outlines the current knowledge of the synthesis, chemical properties, detection methods, interactions with electrophiles, and toxicological significance of sulfane sulfur, as well as suggesting directions for future research.

sulfane sulfur, nucleophile, electrophile, defense mechanism

Electrophiles are electron-deficient species and are represented by chemicals such as heavy metals, environmental pollutants, toxic metabolites of drugs, and endogenous signaling mediators. Examples of environmental electrophiles include methylmercury (MeHg) in fish (Hanna et al., 2015), cadmium (Cd) in rice (Li et al., 2017), and 1,2-naphthoquinone (1,2-NQ) and 1,4-naphthoquinone (1,4-NQ) in the atmosphere (Cho et al., 2004). Electrophiles form covalent bonds with nucleophiles such as thiols and amines (LoPachin and Gavin, 2016). In biological systems, thiol-containing small molecules such as l-cysteine (CysSH) and glutathione (GSH) are thought to be primary scavengers of electrophiles, forming electrophile-nucleophile adducts. GSH conjugation of electrophiles is well-established as a detoxification pathway to excrete such adducts into the extracellular space (Ketterer et al., 1983). Lipoic acid and its reduced form dihydrolipoic acid also endogenously function as a defense mechanism against electrophilic stresses (Rochette et al., 2013). Nucleophilic macro-biomolecules such as side chains on protein amino acids (eg, Cys, His, Arg, and Lys) and aromatic nitrogen sites on DNA bases are targeted by electrophiles to form stable adducts (LoPachin and Gavin, 2016). Although these electrophile-nucleophile interactions are involved in the dysfunction of macro-biomolecules and subsequent cellular and organ injury, sensor proteins with highly reactive thiols serve as a defense system at lower electrophile concentrations (Kumagai and Abiko, 2017). Electrophilic modification of sensor proteins such as Keap1 and heat shock protein 90 can activate transcription factors such as Nrf2 and heat shock factor 1, respectively, thereby up-regulating a battery of cytoprotective genes (Kumagai and Abiko, 2017). This adaptive cellular response to electrophiles is termed electrophilic signaling (Nishida et al., 2016; Rudolph and Freeman, 2009). Collectively, several defense systems are involved in protection against electrophilic stresses via interaction with thiol groups. In this context, the deprotonated form of thiol (thiolate anion) is the nucleophile that forms covalent bonds with electrophiles (LoPachin and Gavin, 2016). Thus, the pKa value is important to predict the reactivity of thiols.

CHEMICAL PROPERTIES OF SULFANE SULFUR

Sulfane sulfur is sulfur with 6 valence electrons and no charge, which binds covalently to other sulfur atoms (Toohey, 2011). This molecule always exists in another sulfur-attached form and not in stand-alone form, and can be present as elemental sulfur (S₈), persulfides (RSSH), polysulfides (RSSSH or RSSSR), and thiosulfate (S₂O₃) (Toohey, 2011). One unique property of sulfane sulfur is that it reversibly transfers to other sulfur atoms (trans-sulfidation) (Figure 1) (Toohey and Cooper, 2014). Thus sulfane sulfur easily transfers to other thiols to form the corresponding persulfide/polysulfide species and the length of the polysulfur can be increased. Some cellular proteins, such as cystathionine γ-lyase (CSE) (Yamanishi and Tuboi, 1981), 3-mercaptopyruvate sulfurtransferase (3MST) (Kabil and Banerjee, 2014), and rhodanese (Ploegman et al., 1978), are also known to carry sulfane sulfur, which can be released as H₂S/HS⁻ (Toohey, 1989). The released H₂S can be enzymatically oxidized to sulfane sulfur by sulfide: quinone oxidoreductase with an acceptor of sulfane sulfur such as GSH (Libiad et al., 2014).

Figure 1.

Synthesis of sulfane sulfur in biological systems. The major source of sulfane sulfur is cysteine persulfide, which is produced by the mitochondrial protein cysteinyl-tRNA synthetase (CARS2) with cysteine as a substrate or cystathionine β-synthase (CBS)/cystathionine γ-lyase (CSE) with cystine as a substrate. Sulfane sulfur can be reversibly transferred to other thiols such as glutathione (GSH) or protein-SH to form the corresponding persulfide/polysulfide species, respectively. This protein sulfidation can occur through both cotranslational and post-translational pathways. Abbreviations: 3MST, 3-mercaptopyruvate sulfurtransferase; CAT, cysteine aminotransferase.

Importantly, persulfides are more nucleophilic than the corresponding thiols because of the presence of unshared pairs of electrons on the atom adjacent to the nucleophilic atom, referred to as “α-effect” (Edwards and Pearson, 1962). The pKa value of an RSSH species is typically 1–2 pKa units lower than that of the corresponding RSH (Cuevasanta et al., 2015; Everett and Wardman, 1995). This indicates that a reaction with the persulfide anion (RSS⁻) is predominant over a reaction with the thiolate anion (RS⁻) toward electrophiles. It should be noted that increasing the length of the polysulfur chain should increase the inherent nucleophilicity of the sulfur atoms. Antioxidative properties of persulfide have also been reported (Iciek and Wlodek, 2001). Thiols and persulfides differ in that thiols can be only nucleophilic, whereas persulfides can be either nucleophilic or electrophilic, depending on the protonation state. A persulfide anion (RSS⁻) is highly nucleophilic, whereas the protonated RSSH is actually electrophilic, like a disulfide (Fukuto et al., 2018).

BIOLOGICAL SYNTHESIS OF SULFANE SULFUR

Sulfane sulfur is biologically synthesized through a cysteine desulfuration process (Figure 1). 3MST can catalyze the production of sulfane sulfur and attach it to the thiol group of the 3MST protein itself using 3-mercaptopyruvate as a substrate (Kabil and Banerjee, 2014). We herein define such proteins as sulfane sulfur-binding proteins (SSBPs). Dihydrolipoic acid was reported to be an acceptor of sulfane sulfur for 3MST (Mikami et al., 2011; Yadav et al., 2013). Another pathway of cysteine desulfuration is associated with its spontaneous oxidation to cystine, which can be a substrate for CSE (Cavallini et al., 1960) and cystathionine β-synthase (CBS) (Ida et al., 2014) to yield sulfane sulfur-containing cysteine persulfide (CysSSH). The produced sulfane sulfur in CysSSH can reversibly transfer to other thiols such as GSH or protein-SH to form glutathione persulfide (GSSH) or protein-SSH, respectively (trans-sulfidation). In the case of GSSH, the concentration was reported to be approximately 150 µM (brain) and 50 µM (liver and heart) in mice (Ida et al., 2014). Although the original function of cysteinyl-tRNA synthetase (CARS) is to catalyze the addition of cysteine to its cognate tRNA in de novo protein synthesis (Jacquin-Becker et al., 2002), our collaboration study with Akaike et al. found that CARS2, which is localized in mitochondria, catalyzes the production of cysteine persulfide from cysteine as a substrate (Akaike et al., 2017), indicating that CARS2 is a moonlighting protein and that protein sulfidation occurs via both post-translation and cotranslation pathways.

Related to the generation of sulfane sulfur, evidence has been shown for the existence of SSBPs in cells. It is classically recognized that rhodanese, an abundant protein in most mammalian mitochondria, has various functions including cyanide detoxification and transfer of sulfane sulfur (Cipollone et al., 2007). Rhodanese binds sulfane sulfur at Cys47 and catalyzes the transsulfuration of cyanide to thiocyanate (Ploegman et al., 1978), which is nontoxic and excreted in the urine. Serum albumin can also bind sulfane sulfur and catalyze its transfer to cyanide (Bonomi et al., 1977; Jarabak and Westley, 1989). The protective function of polysulfide-added albumin against reactive oxygen species, nitric oxide, and ultraviolet radiation was recently reported (Ikeda et al., 2018). Apart from generating sulfane sulfur, CSE can bind sulfane sulfur within itself. Such a sulfane sulfur is thought to be carried as a trisulfide between 2 cysteine residues and can be transferred to other thiols (Yamanishi and Tuboi, 1981). Furthermore, a regulator protein tyrosine phosphatase 1B (PTP1B) can be reversibly modified by sulfane sulfur, modulating its enzymatic activity (Krishnan et al., 2011). We previously identified GSH S-transferase P1 (GSTP1) in rat liver as an SSBP involved in MeHg inactivation (Abiko et al., 2015b). Dynamin-related protein 1 (Drp1), a major mediator of mitochondrial fission, was found to be polysulfidated and likely activated by depolysulfidation (Akaike et al., 2017). In addition, sulfane sulfur can bind to other proteins, such as cytochromes (Catignani and Neal, 1975), ferredoxin (Petering et al., 1971), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), alcohol dehydrogenase 5 (ADH5), and ethylmalonic encephalopathy 1 (ETHE1) (Akaike et al., 2017), but the exact function of such modification is unknown. We therefore speculate that numerous proteins undergo extensive S-polysulfidation, and the sulfane sulfur moiety appears to play a role in redox regulation and adaptive response to oxidative and electrophilic stresses.

METHODS OF SULFANE SULFUR DETECTION

To understand the biological function of sulfane sulfur, it is essential to establish detection methods sensitive enough to determine its levels in biological samples. As sulfane sulfur reacts with cyanide to form thiocyanate, further incubation of thiocyanate with Fe³⁺ can yield the red-colored complex Fe(SCN)₆³⁻, which is detectable by spectrophotometry (Wood, 1987) (Figure 2a). This method and its modifications, however, are now only rarely used because of the limited sensitivity and specificity.

Figure 2.

Analytical techniques to detect sulfane sulfur. There are at least 3 major methods of detecting sulfane sulfur: Cyanolysis uses the reaction of the cyanide anion and sulfane sulfur, followed by Fe³⁺ addition to measure the absorbance (a); various fluorescent probes can detect sulfane sulfur (b); and alkylating agents can be used to quantify adducts by LC-MS/MS with stable isotope-labeled standards (c).

A search for a new method of sulfane sulfur detection is ongoing. Ogasawara et al. found that sulfane sulfur in serum can be released as sulfide by dithiothreitol reduction, and that the released sulfide can be converted into a fluorescent derivative of thionin by reaction with p-phenylenediamine and ferric ion, and detected by high-performance liquid chromatography (Ogasawara et al., 1993). The released sulfide, however, is only a fraction of the sulfane sulfur because thiosulfate and thiosulfonates could not be detected by this method, even though they contain sulfane sulfur.

Chen et al. developed sulfane sulfur probe (SSP) series of fluorescent probes for sulfane sulfur detection (Chen et al., 2013) (Figure 2b). After that, several similar fluorescent probes such as DSP series (Liu et al., 2014), QS_n (Zeng et al., 2015), and AP (Gao et al., 2015) for the detection of hydrogen polysulfide were also developed. These probes are highly sensitive and specific for sulfane sulfur, and do not react with thiol groups or hydrogen sulfide. An advantage of fluorescent probes is that they can be used in real-time in biological samples. Zeng et al. successfully visualized both exogenous and endogenous H₂S/H₂S_n in living cells and zebrafish embryos (Zeng et al., 2015). Takano et al. developed a reversible on/off fluorescent probe (SSip-1 DA) for sulfane sulfur that can repeatedly visualize intracellular concentration changes in living cells (Takano et al., 2017) (Figure 2b). A drawback of these fluorescent probes is their inability to distinguish the origin molecules such as cysteine persulfide/polysulfide, GSH persulfide/polysulfide, or protein-bound sulfane sulfur.

Alkylating agents are also tools for detecting sulfane sulfur (Figure 2c). Monobromobimane (MBB) has long been used to quantify thiols by forming fluorescent derivatives containing 1 bimane per thiol group (Fahey et al., 1981; Newton et al., 1981). MBB was reported to detect H₂S/HS⁻ by forming bis-S bimane adducts (Wintner et al., 2010). An LC-MS/MS method was later developed to detect not only H₂S/HS⁻ but also cysteine persulfide/polysulfide and GSH persulfide/polysulfide using MBB to form the corresponding adducts, together with stable radioisotope-labeled standards (Ida et al., 2014). Although methods using alkylating agents involve the destruction of biological samples, they can simultaneously quantify each sulfane sulfur-containing species and CysSH, GSH, and H₂S/HS⁻ with high sensitivity and specificity. An improved method was recently developed that uses β-(4-hydroxyphenyl)-ethyl iodoacetamide instead of MBB (Akaike et al., 2017); the former is more favorable to detecting sulfane sulfur-containing species without overestimating cysteine/GSH levels and underestimating the sulfane sulfur levels, as it is less reactive with thiols.

Methods for the detection of protein-bound sulfane sulfur have also been developed. Mustafa et al. used a modified biotin-switch assay to detect protein-bound sulfane sulfur (Mustafa et al., 2009), but this assay may not have been specific for sulfane sulfur (Pan and Carroll, 2013). Ida et al. developed a tag-switch-tag assay with 2 probes (methylsulfonyl benzothiazole and CN-biotin) and confirmed the widespread presence of SSBPs in cells (Ida et al., 2014). Doka et al. developed a protein persulfide detection protocol using a biotin-tagged alkylating agent and streptavidin-coupled beads (Doka et al., 2016). Our group developed a (MeHg)₂S assay using MeHg as a sulfane sulfur-trapping agent, and identified GSTP1 as an SSBP (Abiko et al., 2015b). Akaike et al. further developed a biotin-PEG-conjugated maleimide labeling gel-shift assay to identify polysulfidated proteins (Akaike et al., 2017).

INTERACTION OF SULFANE SULFUR WITH ELECTROPHILES

As mentioned earlier, sulfane sulfur-containing molecules such as CysSSH and GSSH exhibit higher acidity and nucleophilicity than their corresponding thiols. This suggests that electrophiles would be captured by reactive persulfide/polysulfide species, resulting in formation of their sulfur adducts. Nishida et al. reported 3 possible fates for electrophiles reacting with the sulfane sulfur-related molecule HS⁻ (Nishida et al., 2012): Group I electrophiles (endogenous 8-nitro-cGMP and 15-deoxy-prostaglandin J2) form a relatively stable SH derivative that can undergo oxidative degradation of SH. Group II electrophiles (such as endogenous nitrated fatty acids and exogenous 1,2-NQ, 1,4-NQ, 2-tert-butyl-1,4-benzoquinone, and MBB) form electrophile-S-electrophile adducts by sulfhydration. Group III electrophiles (endogenous 4-hydroxynonenal and acrolein) undergo degradation to unknown products.

Figure 3 summarizes the formation of sulfur adducts when electrophiles were incubated with sulfane sulfur in our laboratory. We first reported that incubation of MeHg with HS⁻ produced (MeHg)₂S as a detoxified metabolite in 2011 (Yoshida et al., 2011). (MeHg)₂S is also formed by the reaction of MeHg with GSSH (GSSH-generating system), gluthathione trisulfide (GSSSG), a model polysulfide Na₂S₄, or protein-bound sulfane sulfur (eg, GSTP1) (Abiko et al., 2015b), indicating that MeHg can interact with the sulfane sulfur atom within persulfide/polysulfide molecules. In the case of Cd, cadmium sulfide (CdS) and cadmium thiosulfate (CdS₂O₃) were identified as sulfur adducts following incubation with Na₂S₄ (Akiyama et al., 2017). CdS is a stable adduct exhibiting almost no toxicity, whereas CdS₂O₃, whose formation is reversible, was shown to exert toxic effects (Akiyama et al., 2017). We also found that 1,4-NQ is converted to 1,4-NQ-SH and 1,4-NQ-S-1,4-NQ by CysSSH (CysSSH-generating system) (Abiko et al., 2017a) or 1,4-NQ-S-1,4-NQ-OH by Na₂S₄ (Abiko et al., 2017b). Unlike 1,4-NQ, the authentic 1,4-NQ-S-1,4-NQ-OH adduct exhibited no cytotoxicity or covalent binding capability, and did not activate electrophilic signaling in cells (Abiko et al., 2017b), indicating that it is a detoxified form of 1,4-NQ. Reaction of N-acetyl-p-benzoquinone imine (NAPQI), an electrophilic metabolite of acetaminophen, with CysSSH or GSSH formed NAPQIH₂-SSSCys and NAPQIH₂-SSCys or NAPQIH₂-SSG, respectively (Abiko et al., 2015a). Consistent with this observation, we also identified NAPQIH₂-SSSCys and NAPQIH₂-SSCys adducts as novel metabolites of acetaminophen in the urine of mice treated in vivo (Abiko et al., 2015a).

Figure 3.

Sulfur adduct formation from incubation of electrophiles with sulfane sulfur. The reaction of sulfane sulfur-containing molecules with electrophiles such as methylmercury (MeHg), cadmium (Cd), 1,4-naphthoquinone (1,4-NQ), and N-acetyl-p-benzoquinone imine (NAPQI) yields various sulfur adducts.

PROTECTIVE FUNCTION OF SULFANE SULFUR AGAINST ELECTROPHILIC STRESSES

Because the identified sulfur adducts of electrophiles such as (MeHg)₂S, CdS, and 1,4-NQ-S-1,4-NQ-OH exhibit little toxicity and covalent binding ability, sulfane sulfur appears to serve as a protective mechanism against electrophilic stresses. Supporting this notion, there is growing evidence that knockdown or knockout of sulfane sulfur-producing enzymes results in increased electrophile-mediated toxicity. For instance, CBS knockdown was shown to augment MeHg-induced cytotoxicity in SH-SY5Y cells, whereas CBS overexpression protected cells (Yoshida et al., 2011). In bovine aortic endothelial cells, CSE knockdown potentiated Cd-induced cytotoxicity, but CSE overexpression conferred protection (Shinkai et al., 2017). In vivo experiments have shown that CSE-knockout mice were sensitive to acetaminophen-induced hepatotoxicity (Hagiya et al., 2015), Cd-induced hepatotoxicity (Akiyama et al., 2017). Note that CBS and CSE manipulation cannot exclude sulfane sulfur-independent influence because they contribute to the production of not only sulfane sulfur but also cysteine.

An exogenous sulfane sulfur donor could be used as a chemopreventive agent against electrophilic stress; Na₂S₄ treatment has been shown to suppress toxicity caused by exposure to Cd (Akiyama et al., 2017), MeHg (Ihara et al., 2017), and 1,4-NQ (Abiko et al., 2017b). Furthermore, cystine polysulfide (CysSSSCys) has cytoprotective effect on N-ethylmaleimide toxicity (Bianco et al., 2019). Apart from reacting with electrophiles directly, sulfane sulfur donors can increase the level of endogenous persulfide/polysulfide species (eg, CysSSH and GSSH) by interaction with CysSH/CysSSCys or GSH/GSSG (Bianco et al., 2019).

CONCLUSIONS AND FUTURE DIRECTIONS

This review surveyed the protective function of sulfane sulfur against electrophilic stress in toxicology and its mechanisms of action. Advances in assay development have elucidated the synthesis pathway and widespread presence of sulfane sulfur in cells. Sulfane sulfur-containing molecules can capture electrophiles to form sulfur adducts, thereby protecting cells from electrophilic insult. As it appears likely that sulfane sulfur in small molecules and proteins plays a critical role in redox homeostasis (Iciek et al., 2015), the marked decrease in endogenous sulfane sulfur species caused by exposure to high concentrations of electrophiles could potentially contribute to disruption of the cellular redox status.

It has been generally recognized that electrophiles are trapped by GSH in the absence and presence of GST, yielding their GSH conjugates that are, in turn, excreted into the extracellular space as a canonical detoxification pathway (Figure 4). In this context, sulfane sulfur is likely involved in the detoxification of electrophiles by least 2 pathways. One possibility is that GSH persulfide dominantly reacts with electrophiles rather than GSH because of its higher nucleophilicity to form GSS-conjugate as shown in Figure 3. Another is that a single sulfane sulfur atom is used in the formation of sulfur adducts (eg, (MeHg)₂S). We speculate that such a noncanonical pathway mediated by sulfane sulfur serves as a primary defense system against electrophiles (Figure 4).

Figure 4.

Canonical and noncanonical detoxification pathways of electrophiles. Glutathione (GSH) adduct formation and subsequent excretion into the extracellular space is recognized as the major detoxification pathway of electrophiles. Trapping of electrophiles by sulfane sulfur to form sulfur adducts is putatively another predominant detoxification pathway.

The biological function of sulfane sulfur has received much attention in redox biology. We think that 3 important toxicological issues should be considered in future studies: (1) Resolving the biochemical fate of sulfur adducts such as (MeHg)₂S; (2) elucidating the protective function of intracellular SSBPs against electrophilic stresses; and (3) exploring the possibilities of practical use of sulfane sulfur-containing compounds in chemoprevention and chemotherapy. Understanding the function of sulfane sulfur still requires mechanistic investigation, as does the interaction between sulfane sulfur and electrophiles.

ACKNOWLEDGMENTS

We thank Dean Meyer, PhD for editing a draft of this manuscript.

DECLARATION OF CONFLICTING INTERESTS

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

FUNDING

This work was supported by a grant-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (No. 18H05293 to Y.K.).