Sulfurous Gases As Biological Messengers and Toxins: Comparative Genetics of Their Metabolism in Model Organisms

Gasotransmitters are biologically produced gaseous signalling molecules. As gases with potent biological activities, they are toxic as air pollutants, and the sulfurous compounds are used as fumigants. Most investigations focus on medical aspects of gasotransmitter biology rather than toxicity toward invertebrate pests of agriculture. In fact, the pathways for the metabolism of sulfur containing gases in lower organisms have not yet been described. To address this deficit, we use protein sequences from Homo sapiens to query Genbank for homologous proteins in Caenorhabditis elegans, Drosophila melanogaster, and Saccharomyces cerevisiae. In C. elegans, we find genes for all mammalian pathways for synthesis and catabolism of the three sulfur containing gasotransmitters, H2S, SO2 and COS. The genes for H2S synthesis have actually increased in number in C. elegans. Interestingly, D. melanogaster and Arthropoda in general, lack a gene for 3-mercaptopyruvate sulfurtransferase, an enzym for H2S synthesis under reducing conditions.


Introduction
Despite initially being thought of only as toxic gases, hydrogen sulfide (H 2 S), nitric oxide (NO), and carbon monoxide (CO) are now recognized as important endogenously produced signalling molecules known as gasotransmitters. Wang describes gasotransmitters as small gas molecules that are membrane permeable, endogenously generated, and which have functions at physiologically relevant concentrations [1]. The first two gasotransmitters to be discovered were NO [2] and CO [3]. H 2 S was the third identified gasotransmitter [1,[4][5][6][7]. Like NO and CO, H 2 S is also a toxic air pollutant [8][9][10]. Sulfur dioxide (SO 2 ) and carbonyl sulfide (COS) are gaseous toxins that only recently have been shown to be endogenously produced and to transmit biological signals [11]. In this paper, we will discuss the biology of the sulfur containing gasotransmitters and refer to their use as toxins. Our primary objective is to relate what is known in mammals to an understanding of the action of these compounds on invertebrate pests of agriculture. As such, we have augmented the paper with comparative bioinformatics of genes involved in the synthesis and catabolism of H 2 S, SO 2 and COS. This will facilitate future detailed genetic studies into the mode of action of these gasotransmitters/sulphurous fumigants The strongest evidence that SO 2 and COS are gasotransmitters comes from their effect on smooth muscle cells. Dilation of vascular smooth muscle is caused by the endothelial release of vasodilator substances referred to as endothelium derived relaxing factor (EDRF) [2]. NO is a major mediator of EDRF-induced vasodilation [12], and H 2 S has been suggested as a secondary EDRF component [13]. However, EDRF causes hyperpolarization in smooth muscle cells, but neither NO nor H 2 S cause this effect. It has been suggested that EDRF contains more than one component that causes hyperpolarization, designated endothelium derived hyperpolarizing factors (EDHFs) [14,15]. Both SO 2 and COS are produced by the porcine coronary artery (PCA), and both have short half-lives of 1-2 seconds, similar to EDHF [11,16,17]. Therefore, SO 2 and COS are potential candidates for EDHF [18].

Hydrogen Sulfide
H 2 S is a colourless, odorous, flammable, and water-soluble gas [27]. It is also highly toxic as evidenced by its use in the First World War as a chemical warfare agent [28]. H 2 S is also a significant air pollutant, particularly in sewerage treatment plants, where it can accumulate to dangerous levels [10]. The toxicity of these gases was initially presumed to be caused by the reversible inhibition of cytochrome c oxidase (COX), the terminal electron acceptor of the electron transport chain (ETC) [29].
It is now recognized that H 2 S has widespread biological roles. Thus, while H 2 S does inhibit COX at high concentrations of approximately 80 ppm similar to cyanide, at low concentrations H 2 S actually stimulates oxygen consumption [30]. H 2 S is found to efficiently compete with other electron donors. When H 2 S concentration is high in colonocytes, complex I of the ETC operates in reverse mode and accepts electrons from quinone in order to reduce NAD to NADH [31]. Inhibition of respiration caused by H 2 S is accompanied by a reversion of the ETC complex II [32].
Exposure to 150 ppm of H 2 S has been shown to induce a suspended animation like state in mice [33]. C. elegans acclimatized in 50 ppm of H 2 S results in thermotolerant and an increase in longevity [34]. Resistance to high temperatures in C. elegans often correlates with increased lifespan [35]. The lethal dose, 100% for C. elegans is 150 ppm of H 2 S. However, acclimatized C. elegans are able to survive being exposed to 500 ppm [36]. H 2 S also affects the cardiovascular [37], neural [4], digestive, respiratory, endocrine [38], and immune systems at physiological concentrations [39] (Table 1). H 2 S is endogenously produced during the metabolism of sulfur containing amino acids, in solution H 2 S dissociates to HS − and H + [40]. These biological activities have led to H 2 S being acknowledged as the third gasotransmitter following NO and CO [1,[4][5][6][7].
2.1. Synthesis of H 2 S. H 2 S is enzymatically generated via the desulfhydration of cysteine by two pyridoxal-phosphate (PLP) dependent enzymes cystathionine-β-synthase (CBS) [47] and Cystathionine-γ-lyase (CSE/CTH) [48]. As well as a PLP-independent enzyme 3-mercaptopyruvate sulfurtransferase (3MST/MPST) [49]. Both CBS and CSE are located in the cytosol [50], whereas 3MST is present in the cytosol and the mitochondria [51]. The biosynthetic pathway of H 2 S is dependent on the tissue location. CBS is the primary source in the brain [52] whereas CSE is the primary source of H 2 S in blood vessels [41]. Disruption of CSE results in an elevation of blood pressure [53].
Both CBS and CSE affect not only the levels of H 2 S, but also the metabolism of sulfur containing amino acids and the redox state of the cell via their effect on the availability of glutathione (GSH). CBS and CSE are each involved in the homocysteine-dependent transsulfuration pathway. CBS catalyzes the first step in the catabolism of homocysteine to cystathionine, whereas CSE catalyzes the synthesis of cystathionine to cysteine (Figure 1  Availability of the sulfur containing amino acid cysteine is a critical factor in the synthesis of glutathione (GSH) [56,57]. Roughly half of the intracellular GSH in the liver is derived from the transsulfuration pathway [58]. GSH and glutathione disulphide (GSSG) are the main thiol/disulphide couple involved in cellular redox maintenance (2GSH/GSSG) [59,60]. H 2 S increases γ-Glutamylcysteine, which is a precursor to GSH and causes a recovery of cysteine transport [42,43].
The heme in the CBS enzyme is redox-active and is capable of reversibly regulating the activity of the enzyme according to the redox state. Under reducing conditions cystathionine production is decreased by approximately 1.7 fold [54]. Whereas under oxidising conditions cystathionine production is increased between 1.6 and 2.1 fold [58]. Additionally, the expression of the CSE gene is also induced under oxidising conditions [61]. The redox responsiveness of these two pathways is likely important in order to maintain an appropriate intracellular glutathione pool [58,62].
A third enzyme, 3MST, participates in a two-step pathway of H 2 S synthesis. Firstly, aspartate aminotransferase (AAT/ASAT/AspAT/GOT (Glutamic oxaloacetic transaminase)) [63] deaminates cysteine in the presence of αketoglutarate to generate 3-mercaptopyruvate and glutamate [64]. Secondly, 3-mercaptopyruvate is desulfhydrated to pyruvate and H 2 S by 3MST (Figure 1(b)). However 3MST activity is decreased under oxidative conditions, unlike CBS or CSE [65]. The inhibition results from oxidation of a catalytic cysteine in the active site of 3MST to sulfenate [66]. This inhibition helps to conserve cysteine in the cell, contributing to the maintenance of cellular redox homeostasis.

Catabolism of H 2 S.
A paralogue of 3MST called Rhodanese (RHOD) is the principle enzyme involved in the detoxification of H 2 S in the mitochondria [67]. RHOD is also involved in the detoxification of cyanide [68]. H 2 S is rapidly oxidized to thiosulfate (S 2 O 3 2− ) and then converted to sulfite (SO 3 2− ) and sulfate (SO 4 2− ) [69]. Vertebrate 3MST, which has 59% homology to RHOD can also potentially detoxify cyanide and H 2 S [67,70].  Figure 2). Of these sequences, all but the C. elegans sequences have been characterized and shown to be CBS. There is also a somewhat more divergent yeast sequence (YGR012W) that defines a second orthologous cluster containing four C. elegans paralogues but no sequences from the other two organisms ( Figure 2). YGR012W is a cysteine synthase located on the mitochondrial outer membrane [71].  Figure 3). S. cerevisiae protein MET17, catalyzes the reaction between O-acetylhomoserine and sulfide, leading to the production of homocysteine [72,73]. S. cerevisiae proteins STR3 and IRC7 are cystathionine-βlyase proteins not found in H. sapiens, which are involved in the biosynthesis of methionine [74,75]. The C. elegans sequence CBL-1/C12C8.2 forms a second orthologous cluster with S. cerevisiae protein IRC7 (Figure 3).

S .c C Y S 3 S .c S T R 3
Cystathionine-γ-lyase orthologous cluster

Carbonyl Sulfide
Carbonyl sulfide was first described in 1841 [77]. It is an air pollutant that also has been used as a fumigant [78,79]. COS is also naturally present in the atmosphere, in water, soil, and plants [80]. COS is biologically generated in bacteria via the enzyme thiocyanate hydrolase, but this enzyme is not present in eukaryotes [81]. Interestingly, COS is detectable in both porcine coronary artery (PCA) and cardiac muscle and is able to induce arterial dilation [18]. As of the writing of this review the eukaryotic pathway of COS biosynthesis is not known. It has been shown, however that stimulation of PCA with acetylcholine causes an increase in synthesis of COS within the coronary artery. This suggests that muscarinic acetylcholine receptors (mAChRs) and not nicotinic acetylcholine receptors (nAChRs) are involved in regulating COS synthesis [18], because mAChRs but not nAChRs are found in the coronary artery [82]. COS is converted via α-carbonic anhydrase (α-CAH) to H 2 S and CO 2 . In eukaryotes, α-CAH is primarily responsible for pH regulation [83]. The enzyme is widely distributed in mammalian blood and tissue [84]. The toxicity of COS is mediated by H 2 S as inhibition of α-CAH activity decreased the toxicity of COS [85]. It is interesting to note that α-CAH activity can be inhibited via H 2 S [86]. After exposure to COS, the redox balance of the cell is disrupted and genes that respond to oxidative stress such as glutathione reductase and superoxide dismutase are upregulated [23]. The gene expression effect of exposure to COS is similar to that of phosphine exposure [87].

C. elegans: Genes Involved in COS Metabolism and Detoxification. The genes responsible for the synthesis of COS
has not yet been identified, but it is known that α-carbonic anhydrase (α-CAH) is responsible for the conversion of COS to carbon dioxide and hydrogen sulfide. The C. elegans αcarbonic anhydrase gene family has been studied previously. It has six family members, two of which (CAH-3 and CAH-4) have been demonstrated to encode functional α-CAH enzymes ( Table 2) [88]. Additionally C. elegans, S. cerevisiae, and D. melanogaster also contain a β class of CAH which is not found H. sapiens [89,90].

Sulfur Dioxide
Like the gases mentioned previously, SO 2 is a toxic air pollutant [91,92]. It also has the distinction of being the oldest recorded chemical fumigant, as it was used by ancient Egyptians, Greeks, and Romans [93]. It was also used as a chemical warfare agent in a conflict between the Athenians and the Peloponnesians circa 431 B.C. [94] as well as during the Roman siege of Dura-Europos in 256 C.E. [95]. Sulfite, a dissociation product of SO 2 , is used as a preservative in beverages and food [96]. SO 2 is likely to be a signalling molecule as it is produced endogenously from the metabolism of sulfur containing amino acids [97]. Additionally, SO 2 has been found to produce biological effects at physiological concentrations [18], such as vasodilation in isolated rat aortic rings [98] and a decrease in blood pressure of male rats [99,100]. For these reasons, SO 2 has been suggested to be a gasotransmitter [98]. SO 2 can also dissociate to its derivatives in solution, which may also be biologically active.   dissociates into sulfite (SO 3 2− ) and bisulfite (HSO 3 − ) in a 3 : 1 ratio in a neutral solution [101]. Furthermore, both SO 3 2− and HSO 3 − can be oxidized to SO 4 2− via sulfite oxidase (SOX) [102]. Vasodilation via SO 2 , however, has been found to be much greater than dilation induced by SO 2 derivatives [103]. It is, therefore, unlikely that activity attributed to SO 2 is actually due to the action of derivative compounds. 2 . SO 2 is generated via two different pathways, one enzymatically and one nonenzymatic. The enzymatic metabolism occurs via catabolism of cysteine [49].  Cysteine catabolism to cysteinesulfinate is dependent upon cysteine dioxygenase (CDO) [97,104]. CDO concentration is regulated by the availability of methionine and cysteine or protein [105]. Therefore, CDO is one of the enzymes that regulates free cysteine levels [106]. The key enzyme in the generation of SO 2 is AAT, which is constantly being expressed (Figure 1(c)) [107]. However, expression of AAT is increased via glucocorticoids [11,108]. AAT is expressed in cytosol and mitochondria [109]. Nonenzymatic generation of SO 2 occurs at neutrophils as a result of oxidative stress, which causes the conversion of H 2 S to sulfite [110].

Toxicity.
There is little information available on the mechanism of SO 2 toxicity. It potentially involves oxidative damage caused by free radicals formed during sulfite oxidation [111]. Exposure to SO 2 has been found to cause lipid peroxidation as well as increase the levels of enzymes that protect cells against oxidative damage caused by reactive oxygen species, including superoxide dismutase, glutathione peroxidase and catalase [112]. Exposure to SO 2 also induces chromosomal aberrations, chromatid exchanges and micronuclei formation, as shown in cultured human blood lymphocytes [113,114]. Sulfite oxidase (SOX) is involved in oxidative detoxification of sulfite [102,115]. Deficiency of SOX has previously been demonstrated to increase SO 2 , SO 3 2− , and HSO 3 − toxicity [116,117]. SOX activity has been shown to be significantly different in different mammalian tissues [118]. It is expected the oxidative damage caused by exposure to SO 2 would be tissue specific. However it is found that exposure of SO 2 caused nearly equal oxidative damage, suggesting that SO 2 toxicity is systemic [119].

C. elegans: Genes Involved in SO 2 Metabolism and
Detoxification. Phylogenetic analysis was not carried out on the other proteins involved in the synthesis or breakdown of SO 2 as no gene duplication had occurred, resulting in a simple one-to-one correspondence between nematode and human sequences. The human gene for cysteine dioxygenase (cdo-1) carries out the initial step in sulfur dioxide synthesis from cysteine. C. elegans has a single counterpart which is also called cdo-1 ( Table 3). The second step in sulfur dioxide synthesis is aspartate amino transferase, which is also used elsewhere in sulfur metabolism and is the subject of Figure 5. The final step in the synthesis of sulfur dioxide is the nonenzymatic decomposition of sulfinyl-pyruvate to sulfur dioxide and pyruvate. The oxidation of sulfur dioxide is carried out by SOX-1 in humans, the orthologous C. elegans gene is H13N06.4 (Table 3).

Cross-Talk between Gasotransmitters
H 2 S and SO 2 have been found to act synergistically with NO to enhance the vasorelaxant effect [100,120]. The vasorelaxant effect of H 2 S in rat aortic rings can be decreased by removal of the endothelium, interruption of NO synthase or blocking of Ca 2+ -dependent K + channels [121]. This suggests that NO and potentially EDHF contribute to the vasorelaxant effect of H 2 S. However, others have not observed synergistic vasorelaxation when H 2 S and NO treatments are combined but instead see a decrease in vasorelaxation [122]. This is likely due to reaction between H 2 S and NO to generate a nitrosothiol [123].
Synthesis of H 2 S is increased when exposed to NO due to the activation of the cyclic guanosine monophosphate pathway, which potentially increases the activity CSE [124]. NO also increases the expression of CSE which is involved in H 2 S synthesis [37]. It has also been found that metabolism of CO is also enhanced when exposed to H 2 S [125]. Additionally, NO and H 2 S can result in an increase in cysteine uptake [42,126]. This increase in cystine uptake can then enhance the synthesis of H 2 S [127]. In contrast, H 2 S has been found to suppress NO synthesis in rats by down regulating the aortic L-arginine-dependent NO pathway [128]. Exposure to SO 2 is found to increase NO pathway [129]. Despite the poor understanding of the cross-talk between gasotransmitters, it is clear that their functions are tightly integrated.

Storage, Release, and Transport
After the sulfur containing gases are endogenously produced, they can either be immediately released or stored for later release in response to a physiological signal. Unlike the short half-lives of COS and SO 2 , the half-life of H 2 S in mammalian plasma is about 30 minutes [130]. Sulfur from the sulfur containing gases can be stored in two forms, acid-labile sulfur or bound sulfane-sulfur. Exogenously H 2 S is absorbed and stored as bound sulfane-sulfur [131]. Sulfur is released from bound sulfane-sulfur by reduced glutathione and cysteine, or at pH higher than 8.4 [131,132]. Acid-labile sulfur is generally found as iron-sulfur enzymes in the mitochondria. Sulfur is released from acid-labile sulfur at a pH of <5.4, with little or no H 2 S released at pH ≤ 6 [131].
Bound sulfane-sulfur is generally located in the cytoplasm though there is also evidence of long distance transport [131,133]. One mechanism of transport has been determined in Lucina pectinata in which the heme group of hemoglobin I (HbI) binds and transports sulfide [97,134]. In the C1 Hb from Riftia pachyptila, the sulfide is bound to zinc ions [135]. Neither HbI nor C1 Hb are found in C. elegans, though other globular proteins exits which could potentially bind to sulfide, providing a mechanism for long distance sulfide transport. H 2 S COS and SO 2 are able to freely diffuse across the hydrophobic cellular membrane without facilitation via membrane channels [136][137][138]. Furthermore, no evidence of active membrane transport of H 2 S has been found [137]. This situation is analogous that of ammonia transport, which is also able to freely diffuse across the cellular membrane. Despite this ability, however, three distinct transport systems exist that actively transport ammonia across cell membranes [139].

Future Directions
The three sulfur containing gases H 2 S, SO 2 , and COS act as gasotransmitters in vertebrates. The primary bioassay that is used to study the three compounds is vasodilation, though metabolic arrest has also been demonstrated in the case of H 2 S and all three compounds are known to affect the redox state of cells. Though the compounds have barely been studied in invertebrates, H 2 S has been shown to mediate lifespan extension and heat tolerance in C. elegans [34] as well as desiccation tolerance in D. melanogaster [140]. Exposure of C. elegans to SO 2 induces ovoviviparity [141], which is a stress response in C. elegans [142]. The toxicity of the chemicals when administered at concentrations greater than normal endogenous levels likely reflects their role as potent neurological and physiological signalling molecules. The toxicity has been exploited commercially through the use of these compounds as fumigants.
Characterisation of the roles of these molecules in the model organism, C. elegans, will facilitate the genetic analysis of their function and toxicology with benefits to be gained in agriculture and medicine. A first step toward genetic analysis is to determine the extent to which the metabolic pathways exist in the primary eukaryotic genetic model organisms; S. cerevisiae, C. elegans, and D. melanogaster. Our analysis reveals that all of the metabolic genes are present in yeast. Studies in this organism will be of limited value in understanding how the compounds act in a multicellular animal, however. D. melanogaster and indeed all insects are missing one of the metabolic genes altogether, which will limit the studies that can be considered in this species. The model system C. elegans contains all of the mammalian genes involved in the metabolism of the sulfur containing gases. C. elegans has an additional interesting property. Most genes for synthesis of H 2 S have been duplicated in C. elegans even though single genes exist in the other organisms. This situation likely reflects the stereotypic biology of C. elegans in which specific genes act in specific cell types to a greater extent than in other organisms. This may provide a research advantage as genetic manipulations may allow gasotransmitter signalling to be disrupted more specifically than in other organisms.
C. elegans is ideal for the genetic investigation of gasotransmitter action and toxicity as the nematodes are cultured on agar medium which facilitates simultaneous exposure to dissolved chemicals and gases. C. elegans reproduces rapidly as a self-fertilizing hermaphrodite which facilitates the creation of mutant strains. There are also well-defined techniques for transformation and genetic manipulation of gene expression. Furthermore, the stereotypic development of C. elegans means the origin of each cell in the adult is known and the physiological role of each cell is reproducible. Because the organisms are transparent, the physiology of individual cells in the living organism also can often be studies microscopically using fluorescent probes.
Genetic analysis can be carried out in the "forward" direction, which refers to the traditional approach of mutagenesis of all genes in the genome followed by a screen for mutant individuals that exhibit a specific effect (e.g., resistance to H 2 S). Analysis then reveals the gene that was mutated and how the phenotype is mediated. Genetic analysis can also be carried out in the reverse direction, which refers to the molecular genetic approach of disrupting a characterised gene that ought to affect a process and then analysing the result. An example would be to individually suppress each CSE gene in C. elegans to see which of them result in phenotypes related to H 2 S synthesis. Genetic analysis can also be combinatorial, an approach that is greatly facilitated in genetic model organisms, which usually have large collections of characterised mutant strains that are distributed to researchers on request. An example of a combinatorial approach would be to determine the effect of H 2 S depletion on lifespan in strains carrying known longevity mutations. This would indicate whether the effect of H 2 S on lifespan is related to any previously described lifespan enhancing mutations.
Major issues remain to be investigated regarding the roles of sulfur containing gases in biology, particularly in invertebrates. These include the possibility of unique roles of the gases within specific subcellular compartments, in specific tissues or at specific times during development. Genetic analysis can be used to identify interactions between gaseous signals and other signalling pathways as well as the influence of redox state on the activity of the gases. Understanding the mechanisms of action can also be used to identify novel fumigants or fumigation synergists of commercial importance. Model organism genetics can also be used for the identification and genetic manipulation of physiological parameters of medical significance that are controlled by the sulfur containing gases. Such physiological states include thermotolerance, desiccation tolerance, reversible metabolic arrest, and hypoxic preconditioning. All of these research targets can be addressed effectively and meaningfully in genetic model organisms such as C. elegans.