- Email: [email protected]
Plasmid DNA damage caused by stibine and trimethylstibine
Toxicology and Applied Pharmacology 194 (2004) 41 – 48 www.elsevier.com/locate/ytaap
Plasmid DNA damage caused by stibine and trimethylstibine
$
Paul Andrewes, Kirk T. Kitchin, * and Kathleen Wallace Environmental Carcinogenesis Division, Office of Research and Development, National Heath and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC 27709, USA Received 10 July 2003; accepted 25 August 2003
Abstract Antimony is classified as ‘‘possibly carcinogenic to humans’’ and there is also sufficient evidence for antimony carcinogenicity in experimental animals. Stibine is a volatile inorganic antimony compound to which humans can be exposed in occupational settings (e.g., lead-acid battery charging). Because it is highly toxic, stibine is considered a significant health risk; however, its genotoxicity has received little attention. For the work reported here, stibine was generated by sodium borohydride reduction of potassium antimony tartrate. Trimethylstibine is a volatile organometallic antimony compound found commonly in landfill and sewage fermentation gases at concentrations ranging between 0.1 and 100 Ag/m3. Trimethylstibine is generally considered to pose little environmental or health risk. In the work reported here, trimethylstibine was generated by reduction of trimethylantimony dichloride using either sodium borohydride or the thiol compounds, dithioerythritol (DTE), L-cysteine, and glutathione. Here we report the evaluation of the in vitro genotoxicities of five antimony compounds—potassium antimony tartrate, stibine, potassium hexahydroxyantimonate, trimethylantimony dichloride, and trimethylstibine—using a plasmid DNA-nicking assay. Of these five antimony compounds, only stibine and trimethylstibine were genotoxic (significant nicking to pBR 322 plasmid DNA). We found stibine and trimethylstibine to be about equipotent with trimethylarsine using this plasmid DNA-nicking assay. Reaction of trimethylantimony dichloride with either glutathione or L-cysteine to produce DNA-damaging trimethylstibine was observed with a trimethylantimony dichloride concentration as low as 50 AM and Lcysteine or glutathione concentrations as low as 500 and 200 AM, respectively, for a 24 h incubation. D 2003 Elsevier Inc. All rights reserved. Keywords: Antimony; L-Cysteine; DNA nicking; Genotoxicity; Glutathione; Reactive oxygen species; Stibine; Trimethylstibine; Trimethylantimony dichloride; Trimethylarsine
Introduction The International Agency for Research on Cancer (IARC, 1989) has classified antimony (Sb), as antimony trioxide, as ‘‘possibly carcinogenic to humans’’. There is also sufficient evidence for the carcinogenicity of antimony trioxide in experimental animals (IARC, 1989). Antimony is in column 15 of the periodic table below arsenic, a wellknown human carcinogen (IARC, 2002), and antimony shares many chemical properties with arsenic. For instance, the biomethylation of antimony by microorganisms is well-
$ This manuscript has been reviewed by the National Health and Environmental Effects Research Laboratory at the U.S. Environmental Protection Agency and approved for publication. The views expressed in this paper are those of the authors and do not necessarily represent the views or policy of the U.S. Environmental Protection Agency. * Corresponding author: Fax: +1-919-685-3276. E-mail address: [email protected] (K.T. Kitchin).
0041-008X/$ - see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.taap.2003.08.012
established (Bentley and Chasteen, 2002) and is thought to proceed via a mechanism similar to the Challenger mechanism for arsenic biomethylation (Andrewes and Cullen, 2003; Andrewes et al., 2000; Challenger, 1945). The modes of carcinogenic action for arsenic (IARC, 2002; NRC, 1999) and antimony (Gebel, 1997; Gebel et al., 1997) are not established. However, methylated arsenic(III) species have been shown to be genotoxic (Ahmad et al., 2002; Andrewes et al., 2003; Mass et al., 2001; Nesnow et al., 2002) and potent clastogens (Kligerman et al., 2003), but not point mutagens. Human exposure to antimony is not as pervasive as it is for arsenic. For example, there is no instance of a population exposed to antimony that parallels that in Bangladesh, where at least 25 million people are exposed to high (>50 ppb) arsenic concentrations in drinking water (IARC, 2002; Mandal and Suzuki, 2002; NRC, 1999). However, antimony often occurs as a co-contaminant with arsenic and it may be a confounding variable in the carcinogenesis of arsenic
42
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
(Gebel, 2000). Furthermore, antimony is a component of lead-acid batteries and some of this antimony is volatilized as gaseous stibine (SbH3) at each charging of the battery. Exposure to antimony in the production of starter batteries can reach levels as high as 0.04 mg/m3 (Kentner et al., 1995) (compare with threshold limit value: 0.5 mg/m3). The health consequences of long-term exposure to stibine are unknown. Exposure to the corresponding arsenic compound, arsine (AsH3), causes oxidative damage to red blood cells (Blair et al., 1990). The end product of antimony biomethylation, trimethylstibine (Me3Sb), is commonly detected, along with trimethylarsine (Me3As), in biological fermentation gases from landfills and sewage treatment plants (Feldmann and Hirner, 1995; Feldmann et al., 1998). However, the health and environmental effects of these gases in the atmosphere surrounding the landfill are unknown. These organometallic gases pose an environmental problem for landfill-gas-fired power stations that must dispose of the toxic oxides produced by their combustion (Glindemann et al., 1996). As noted above, the toxicity of antimony and antimony compounds, such as stibine and trimethylstibine, in humans has not been investigated to the same extent as arsenic (Gebel, 1997; Gebel et al., 1997). In a recent review, Gebel (1997) concluded that there are no data available on the mechanistic genotoxicity of antimony. In the few studies that have been done with antimony, it was found to be clastogenic but not a gene mutagen (Gebel, 1997; Gebel et al., 1997); similar results exist for arsenic (Kligerman et al., 2003). We recently reported that the dithiol compound dithioerythritol (DTE), a good model for biological thiol compounds, reduced a trimethylarsenic(V) compound, trimethylarsine oxide (TMAO), to DNA-damaging trimethylarsine (Andrewes et al., 2003). Trimethylarsine oxide can be reduced by a variety of biological thiol compounds, including lipoic acid, glutathione, and cysteine, to form trimethylarsine (Cullen et al., 1984). Thus, we suggested that trimethylarsine oxide might react with any number of thiol compounds in cellular environments to form genotoxic trimethylarsine in vivo. There is some evidence that supports this hypothesis (Nishikawa et al., 2002). Because humans biomethylate inorganic antimony to trimethylantimony(V) species (Krachler and Emons, 2001), apparently, trimethylantimony(V) species formed in cells might be reduced to trimethylstibine. In the present study, we have examined the in vitro genotoxicity of five antimony compounds [three antimony(III) compounds—potassium antimony tartrate, stibine, and trimethylstibine—and two antimony(V) compounds— potassium hexahydroxyantimonate, and trimethylantimony dichloride]. Stibine was generated by sodium borohydride reduction of potassium antimony tartrate. The use of sodium borohydride to generate hydride compounds from metal and metalloid compounds is well-known in analyt-
ical and organometallic chemistry and is generally very efficient (Howard, 1997). Trimethylstibine was generated by reduction of trimethylantimony dichloride using either sodium borohydride, dithioerythritol, or the more physiologically relevant compounds, L-cysteine and glutathione. The ease of trimethylstibine formation by reduction of trimethylantimony dichloride by thiols and the high genotoxicity of trimethylstibine (comparable with trimethylarsine) suggest that trimethylstibine formation in vivo could be a possible mechanism of antimony carcinogenesis. In this paper, we also discuss the relevance of trimethylarsine and trimethylstibine genotoxicity with respect to possible health and environmental hazards posed by these compounds occurring in landfill and sewage fermentation gases.
Methods Materials. Caution: Antimony compounds are toxic and potentially carcinogenic. Trimethylstibine is a reactive volatile liquid. These compounds should be handled using appropriate safety measures. Trimethylantimony(V) dichloride, potassium antimony(III) tartrate, potassium hexahydroxyantimonate(V), dithioerythritol (DTE), L-cysteine (hydrochloride, monohydrate), and glutathione (reduced form, anhydrous) were purchased from Sigma (St. Louis, MO). Trimethylarsine oxide (TMAO) was synthesized using a method described previously (Nelson, 1993) in the laboratory of Dr. W.R. Cullen (University of British Columbia, Vancouver, Canada). Solutions of these compounds were made by dissolving an appropriate amount in TE buffer (10 mM Tris –HCl, 1 mM EDTA, pH 8.0) and adjusting the pH to 8 (as determined by pH indicator strips). Sodium borohydride, glutathione, cysteine, and DTE solutions were prepared fresh each time an experiment was performed by dissolving the appropriate amount of solid in water and, if necessary, adjusting the pH to 8. Supercoiled (SC) naked (essentially protein-free) plasmid DNA (pBR 322) was purchased from Roche (Indianapolis, IN). Agarose gel electrophoresis. The amounts of damaged (nicked) and intact DNA were determined by using agarose gel electrophoresis methods described elsewhere (Ahmad et al., 2002; Nesnow et al., 2002). Briefly, pBR 322 plasmid DNA, suspended in 45 Al of TE buffer (10 mM Tris –HCl, 1 mM EDTA, pH 8.0), was mixed with 5 Al of gel-loading solution (0.05% bromophenol blue, 50% v/v sucrose). The mixtures were loaded in 1% agarose gels containing 0.5 Ag/ml ethidium bromide (eight lanes per gel). Supercoiled (SC) fast migrating DNA was separated from linear (L) DNA (double-strand break) with moderate electrophoretic mobility and open circular (OC) slow migrating DNA (single-strand break) by electrophoresis for 2– 3 h at 50 V (constant voltage). Ethidium
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
bromide-stained DNA bands were visualized under UV light. Nicking of DNA exposed to gaseous antimony compounds. Cellulose nitrate filter paper squares (Whatman, 0.2-Am pore size, f5 5 mm) were rinsed with TE buffer and briefly allowed to dry. A pushpin was pushed through the center of an Eppendorf tube (500 Al) cap and through a filter paper square. The filter paper was gently pushed into the roof of the cap. Immediately before generating the volatile antimony compound, the target DNA solution (1 Al of 250 ng/Al) was pipetted onto the filter paper square. Antimony compounds, either potassium antimony tartrate or trimethylantimony dichloride, were pipetted into Eppendorf tubes with TE buffer to give a total liquid volume of 45 Al with final antimony concentrations ranging from 5 to 5000 AM. Next, each Eppendorf tube was held almost horizontal so that sodium borohydride (5 Al of 5 mM) could be pipetted onto the Eppendorf side wall without mixing with the other reagents. The Eppendorf was capped with the filter paper containing cap, then held vertically and tapped gently so the sodium borohydride drop would mix with the other reagents and generate either the stibine (from potassium antimony tartrate) or trimethylstibine (from trimethylantimony dichloride). After 30 min at room temperature, the filter paper was removed and placed in another Eppendorf tube to which TE buffer (50 Al) was added. The tube was thoroughly agitated for 1 min to release DNA from the filter paper. The amount of DNA damage was determined by using agarose gel electrophoresis as described above. Controls were also run where the antimony compounds were replaced with distilled water. Nicking of DNA exposed to antimony compounds in solution. In an experiment comparing the DNA-damaging activities of five different antimony compounds, solutions were prepared containing DNA (250 ng), an antimony compound (10 mM or 1 mM) and TE buffer to give a total volume of 50 Al. Stibine or trimethylstibine were generated in situ by mixing potassium antimony tartrate (10 mM) or trimethylantimony dichloride (1 mM), respectively, with sodium borohydride (10 mM). Sodium borohydride was always the last reagent added. For comparison, one sample was also run where DNA was exposed to trimethylarsine, generated in situ by reacting TMAO (1 mM) with sodium borohydride (10 mM). The control lanes consisted of DNA and sodium borohydride (10 mM). The compounds were incubated with the DNA for 1 h at 37 jC, and then the amount of DNA-nicking was determined by using agarose gel electrophoresis as described above. In one experiment, solutions were prepared containing DNA (250 ng) and trimethylantimony dichloride (5 –5000 AM). Trimethylstibine was generated by adding 5 Al of sodium borohydride (5 mM). After 30 min, the amount of
43
DNA damage was determined by using agarose gel electrophoresis as described above. In the same experiment, DNA suspended on filter paper squares was exposed to the gaseous antimony compounds that were volatilized from the reaction mixtures into the headspace of the vials, as described above. In other experiments, solutions were made containing 1 Al of DNA (250 ng/Al), TE buffer, trimethylantimony dichloride (20 – 5000 AM), and a thiol-reducing agent (20 –5000 AM), either DTE, L-cysteine, or glutathione, to give a total volume of 45 Al. The solutions were made in 500-Al Eppendorf tubes. The reducing agent was always the last reagent added. Controls were also run consisting of either DNA alone or DNA and reducing agent (5 mM) or DNA and trimethylantimony dichloride (5 mM). The solutions were gently agitated and then incubated for 24 h at 37 jC, and the amount of DNA damage was determined by using agarose gel electrophoresis as described above.
Results The results reported here are representative of duplicate experiments that were reproducible in all cases. No evidence from control assays suggests that nuclease contamination caused DNA damage. Supercoiled plasmid DNA (pBR 322) was exposed to five different antimony compounds and, for comparison, trimethylarsine. DNA damage was determined by using agarose gel electrophoresis (Fig. 1). We used
Fig. 1. Agarose gel electrophoresis of pBR 322 plasmid DNA after a 1-h incubation (37 jC, pH 8.0) with five different antimony compounds and, for comparison, trimethylarsine. Stibine, trimethylstibine, and trimethylarsine were made by reacting sodium borohydride (10 mM) with potassium antimony tartrate, trimethylantimony dichloride, and trimethylarsine oxide, respectively. The two control lanes contained 10 mM sodium borohydride. SC, supercoiled form of plasmid DNA; OC, open circular form of plasmid DNA (after a single-strand break).
44
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
Fig. 2. Agarose gel electrophoresis of pBR 322 plasmid DNA that was blotted onto cellulose nitrate filter paper and suspended for 30 min above reaction mixtures producing stibine. The reaction mixtures contained sodium borohydride (50 mM) and potassium antimony tartrate (100 – 5000 AM). SC, supercoiled form of plasmid DNA; OC, open circular form of plasmid DNA (after a single-strand break).
potassium antimony tartrate as a source of antimony(III) because it is much more water soluble than antimony trioxide or antimony trichloride, which hydrolyses to antimony trioxide. Potassium antimony tartrate and antimony trioxide in solution are probably in equilibrium with antimonite [Sb(OH)3] (Gebel et al., 1997). We also examined
the genotoxicity of antimony pentachloride (which hydrolyses to antimony pentoxide). Although this compound was not very water soluble, a saturated solution of antimony pentachloride did not damage DNA (results not shown). The two antimony compounds that were genotoxic were trimethylstibine and stibine (Fig. 1). Trimethylstibine and stibine were generated in situ by the reaction of sodium borohydride with trimethylantimony dichloride and potassium antimony tartrate, respectively. There was no evidence that sodium borohydride or any other component of the reaction mixture except stibine or trimethylstibine damaged DNA. Sodium borohydride and byproducts of the sodium borohydride reaction do not damage DNA in this assay (Andrewes et al., 2003). If it is assumed that trimethylstibine and trimethylarsine are generated with equal efficiency from their precursor compounds, then apparently, trimethylstibine and trimethylarsine were about equipotent at damaging plasmid DNA. In one experiment, we placed DNA in reaction mixtures generating stibine and also suspended DNA above the reaction mixtures. Here, stibine did minimal damage to the DNA in solution (results not shown) but substantial damage to the DNA on the filter paper (Fig. 2). Stibine is a volatile compound (boiling point: 17 jC) that will quickly volatilize into the headspace of the reaction vessel. Although we could not determine the concentration of stibine in the gas phase to which the DNA on the filter paper was exposed, we can estimate the minimum stibine concentration needed to damage DNA assuming the borohydride reaction is 100% efficient and all generated stibine remains in the gas phase. We estimated the minimum gas-phase stibine concentration required to do DNA damage to be 6000 mg/m3. The exposure time in this experiment was
Fig. 3. Agarose gel electrophoresis of pBR 322 plasmid DNA that was blotted onto cellulose nitrate filter paper and suspended for 30 min above reaction mixtures producing trimethylstibine (a) and agarose gel electrophoresis of pBR 322 plasmid DNA that was dissolved in those reaction mixtures (b). The reaction mixtures contained sodium borohydride (5 mM) and trimethylantimony dichloride (5 – 5000 AM). SC, supercoiled form of plasmid DNA; L, linear form of plasmid DNA (after a double-strand break); OC, open circular form of plasmid DNA (after a single-strand break).
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
30 min. In previous work with trimethylarsine generation, we showed that the DNA on filter paper squares suspended in vial headspace gases was exposed only to gaseous compounds but not to compounds in the aqueous phase that are transferred to the filter paper as an aerosol (Andrewes et al., 2003). Trimethylstibine is a volatile compound (boiling point: 81 jC) with low water solubility. Trimethylstibine is less volatile than stibine (boiling point: 17 jC). Thus, when we placed DNA in a reaction mixture generating trimethylstibine and also suspended DNA above the reaction mixture, DNA damage was seen for both DNA in solution and DNA suspended in the vial headspace (Fig. 3). For the DNA placed in the aqueous reaction mixture, damage was clearly
45
observed relative to the controls with trimethylantimony dichloride concentrations of 200 AM and higher (Fig. 3b). DNA damage for the DNA on filter paper became clearly discernable, relative to the controls, with a liquid-phase trimethylantimony dichloride concentration of f5000 AM (Fig. 3a). The different response for exposures from the gas phase versus the liquid phase is mostly a consequence of dilution, due to going from a volume of f50 Al (liquid volume) to f500 Al (vial headspace volume). The DNA damage occurred in 30 min or less. We do not know the concentration of gaseous trimethylstibine to which the DNA on the filter paper was exposed, when it was suspended above reaction mixtures. But assuming 100% efficiency of the borohydride reaction and that all trimethylstibine
Fig. 4. Agarose gel electrophoresis of pBR 322 plasmid DNA after a 24-h incubation at 37 jC with trimethylantimony dichloride (20 – 5000 AM) and either 5 mM glutathione (a) or 5 mM L-cysteine (b) in TE buffer (pH 8.0). In comparison, under the same conditions, the trimethylantimony dichloride concentration was held constant (5 mM) and either the glutathione (c) or L-cysteine (d) concentration was titrated (20 – 5000 AM). SC, supercoiled form of plasmid DNA; OC, open circular form of plasmid DNA (after a single-strand break).
46
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
remains in the gas phase, we estimate that the minimum gas-phase trimethylstibine concentration required to cause DNA damage, in 30 min or less, was 10 000 mg/m3. If the above assumptions are not correct, then the potency of trimethylstibine will be underestimated. In a preliminary experiment, supercoiled DNA was incubated with trimethylantimony dichloride and dithioerythritol (DTE) for 2 h at 37 jC. DNA damage occurred for mixtures of DTE (5 mM) and trimethylantimony dichloride concentrations greater than or equal to 200 AM (results not shown). DTE is a dithiol compound that can reduce arsenic(V) and antimony(V) compounds to arsenic(III) and antimony(III) compounds, and in a previous study, we found that DTE reduced trimethylarsine oxide to DNA-damaging trimethylarsine (Andrewes et al., 2003). Analogously, DTE reduces trimethylantimony dichloride to DNA-damaging trimethylstibine. Exposure to DTE or trimethylantimony dichloride alone did not cause DNA damage (results not shown). Further experiments were performed using the physiologically relevant thiol-reducing agents, glutathione and Lcysteine (Fig. 4). Supercoiled DNA was incubated for 24 h at 37 jC in TE buffer (pH 8.0) with various concentrations of L-cysteine, glutathione, and trimethylantimony dichloride. In one series (Figs. 4a, b) of experiments, the concentration of reducing agent was held constant and the trimethylantimony dichloride was titrated. In both cases, the lowest amount of trimethylantimony dichloride required to cause DNA damage in the presence of excess glutathione (Fig. 4a) and excess L-cysteine (Fig. 4b) was approximately 50 AM. In a second series of experiments (Figs. 4c, d), the concentration of trimethylantimony dichloride was kept constant and the concentration of reducing agent was titrated. The lowest amount of glutathione required to cause DNA damage, by reducing trimethylantimony dichloride, was approximately 500 AM (Fig. 4c), and for L-cysteine, the amount required was approximately 200 AM (Fig. 4d). Both glutathione and L-cysteine are found in the body in millimolar concentrations. Again, the DNA-damaging action of trimethylantimony dichloride in the presence of Lcysteine or glutathione was most likely due to the reduction of trimethylantimony dichloride to trimethylstibine that damages DNA.
Discussion Recently, we demonstrated the arsenic compound, trimethylarsine (Me3As), to be genotoxic (Andrewes et al., 2003). In this paper, we have examined the genotoxicity of the corresponding antimony compound, trimethylstibine (Me3Sb), as well as four other antimony compounds: potassium antimony tartrate, stibine, potassium hexahydroxyantimonate, and trimethylantimony dichloride. Of these five antimony compounds, only stibine and trimethylstibine were genotoxic.
Notably, trimethylstibine and trimethylarsine (Andrewes et al., 2003) are approximately equipotent genotoxins in the plasmid DNA-nicking assay (of the eleven arsenicals that we have previously evaluated, Andrewes et al., 2003, trimethylarsine was one of the most potent genotoxic arsenicals). Under equivalent conditions, the minimum concentration of compound in solution required to cause DNA damage for both trimethylstibine and trimethylarsine was 200 AM. It is assumed that trimethylstibine and trimethylarsine are generated with equal efficiency from their precursor compounds by reaction with sodium borohydride. Sodium borohydride is a commonly used reagent in analytical chemistry that is known to react rapidly and in high yield with many metal and metalloid compounds to form hydrides (Howard, 1997). The contrast between the genotoxicity of stibine and the lack of genotoxicity of arsine (Andrewes et al., 2003) is striking. The most likely means for arsines or stibines to cause oxidative damage to DNA are by the production of reactive oxygen species (ROS). There is a growing body of evidence suggesting that oxidative stress plays a central role in both arsenic and antimony toxicity (Gebel et al., 1997; Kitchin and Ahmad, 2003). This ROS-dependant mechanism has been studied extensively for dimethylarsine (Yamanaka et al., 1990), and arsine causes oxidative damage to red blood cells (Blair et al., 1990). If DNA damage occurs by the generation of ROS, then the capacity of arsines and stibines to damage DNA may reflect their oxidative stability. For arsines, this hypothesis is correct. Trimethylarsine and dimethylarsine, which are more susceptible to oxidation (Haas and Feldmann, 2000), were far more damaging to DNA (Andrewes et al., 2003) than were monomethylarsine and arsine, which are more resistant to oxidation (Haas and Feldmann, 2000). Our work on antimony compounds illustrates the same trend. Stibines are more susceptible to oxidation than the corresponding arsines (Haas and Feldmann, 2000). Hence, stibine caused significant damage to DNA whereas arsine did not (Andrewes et al., 2003). In addition, 1 mM trimethylstibine was slightly more damaging to DNA than was 1 mM trimethylarsine (Fig. 1). In the case of arsenic, trimethylarsine was much more DNA damaging than was arsine (Andrewes et al., 2003). Thus, we expected trimethylstibine to be much more damaging to DNA than stibine; however, this was not the case. Stibine and trimethylstibine are formed with approximately equal efficiency from their precursor compounds (Andrewes, 2000). In conclusion, these results for antimony are consistent with our earlier results for arsenic, and the correlation between oxidative stability of arsines and stibines with DNA damage possibly suggests that group 15 compounds may damage DNA by the generation of ROS. We estimated the minimum gas-phase trimethylstibine concentration required to damage DNA (in 30 min or less) to be 10 000 mg/m3. This is many orders of magnitude greater than the typical trace quantities of trimethylstibine
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
(10 – 100 Ag/m3) found in fermentation gases (Feldmann and Hirner, 1995; Glindemann et al., 1996). The permissible exposure level (OSHA PEL) for trimethylstibine is 500 Ag/m3. Thus, this suggests that trimethylstibine in fermentation gases poses a minimal health hazard. Similarly, for stibine, we estimated that the minimum gas-phase concentration required to cause DNA damage was 6000 mg/m3, which is four orders of magnitude greater than the OSHA PEL. Although concentrations of arsines and stibines in the environment are typically quite low, high concentrations have been reported under certain circumstances. For instance, a very high concentration of trimethylarsine (1000 mg/m3) was reported for gas extracted from arseniccontaminated soil (Turpeinen et al., 2002), and some interesting cases of occupational exposure to arsine have been reported (Risk and Fuortes, 1991; Romeo et al., 1997). Arsines and stibines (including Me3As, AsH3, Me3Sb, and SbH3) are used in the semiconductor industry with operating procedures that should minimize exposure. Given the in vitro genotoxicity of these compounds demonstrated in our studies, it would be prudent to examine the genotoxicity and mutagenicity of arsines and stibines using more sensitive and physiologically relevant assays. Antimony(V) and arsenic(V) compounds are readily reduced to the corresponding antimony(III) and arsenic(III) compounds by thiols, including DTE, glutathione, and Lcysteine (Cullen et al., 1984; Ferreira et al., 2003). Thus, we have previously shown that trimethylarsine oxide is reduced to trimethylarsine in the presence of DTE, which then causes DNA damage (Andrewes et al., 2003). In this work, we have shown that DNA damage occurs when trimethylantimony dichloride is incubated with DNA and either DTE, glutathione, or L-cysteine. Damage occurs because these thiol compounds reduced the trimethylantimony dichloride to trimethylstibine, which damages DNA. Analogously, when trimethylarsine oxide was incubated with glutathione or L-cysteine and DNA, nicking of DNA occurred due to the production of trimethylarsine, which damages DNA (Andrewes and Kitchin, unpublished results). Human exposure to trimethylstibine, generated during metabolism of inorganic antimony, should be further considered. If trimethylantimony(V) species (e.g., trimethylantimony dichloride) are formed in vivo, then as demonstrated above, these are likely to be reduced by cellular thiols to trimethylstibine, which damages DNA. Mammalian and human metabolism of antimony has not been investigated in much detail (Andrewes and Cullen, 2003), but antimony biomethylation is well-known in microorganisms (Bentley and Chasteen, 2002). Trimethylantimony dichloride was detected in urine of humans occupationally exposed to antimony (Krachler and Emons, 2001), and Kresimon et al. (2001) detected trace levels ( < 0.1 nM) of monomethylantimony, dimethylantimony, and trimethylantimony spe-
47
cies in urine of humans not occupationally exposed to antimony. Thus, one mode of action for antimony carcinogenesis could be metabolism to trimethylantimony(V) species and reduction to trivalent trimethylstibine that could cause DNA damage and mutation. However, trimethylantimony dichloride was not mutagenic in Salmonella TA104 (DeMarini and Funasaka, unpublished results). In summary, we identified stibine and trimethylstibine as two genotoxic forms of antimony that damaged DNA probably by the generation of ROS. No other forms of antimony were genotoxic. Our conclusions were based on assuming that the generation of stibine and trimethylstibine using sodium borohydride is 100% efficient, if this is not the case then we might have underestimated the potency of these antimony compounds. The effect of arsines and stibines on the environment and on human health warrants further consideration.
Acknowledgments This work was partially supported (P.A.) by a cooperative agreement between the National Research Council (Washington, DC) and the U.S. Environmental Protection Agency. We thank Bill Cullen, David DeMarini, Julian Preston, Jeff Ross, Miroslav Styblo, and David Thomas for their assistance.
References Ahmad, S., Kitchin, K.T., Cullen, W.R., 2002. Plasmid DNA damage caused by methylated arsenicals, ascorbic acid and human liver ferritin. Toxicol. Lett. 133, 47 – 57. Andrewes, P., 2000. The biomethylation of arsenic and antimony by Scopulariopsis brevicualis. PhD Thesis. University of British Columbia, Vancouver, B.C. Andrewes, P., Cullen, W.R., 2003. Organoantimony compounds in the environment. In: Craig, P.J. (Ed.), Organometallic Compounds in the Environment. Wiley, West Sussex, pp. 277 – 304. Andrewes, P., Cullen, W.R., Polishchuk, E., 2000. Antimony biomethylation by Scopulariopsis brevicualis: characterization of intermediates and the methyl donor. Chemosphere 41, 1717 – 1725. Andrewes, P., Kitchin, K.T., Wallace, K., 2003. Dimethylarsine and trimethylarsine are potent genotoxins in vitro. Chem. Res. Toxicol. 16, 994 – 1003. Bentley, R., Chasteen, T.G., 2002. Microbial methylation of metalloids: arsenic, antimony, and bismuth. Microbiol. Mol. Biol. Rev. 66, 250 – 265. Blair, P.C., Thompson, M.B., Bechtold, M., Wilson, R.E., Moorman, M.P., Fowler, B.A., 1990. Evidence for oxidative damage to red blood cells in mice induced by arsine gas. Toxicology 63, 25 – 34. Challenger, F., 1945. Biological methylation. Chem. Rev. 36, 315 – 361. Cullen, W.R., McBride, B.C., Reglinski, J., 1984. The reduction of trimethylarsine oxide to trimethylarsine by thiols: a mechanistic model for the biological reduction of arsenicals. J. Inorg. Biochem. 21, 45 – 60. Feldmann, J., Hirner, A.V., 1995. Occurrence of volatile metal and metalloid species in landfill and sewage gases. Int. J. Environ. Anal. Chem. 60, 339 – 359.
48
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
Feldmann, J., Koch, I., Cullen, W.R., 1998. Complementary use of capillary gas chromatography mass spectrometry (ion trap) and gas chromatography inductively coupled plasma mass spectrometry for the speciation of volatile antimony, tin and bismuth compounds in landfill and fermentation gases. Analyst 123, 815 – 820. Ferreira, C.D., Martins, P.S., Demicheli, C., Brochu, C., Ouellette, M., Frezerd, F., 2003. Thiol-induced reduction of antimony(V) into antimony(III): a comparative study with trypanothione, cysteinyl – glycine, cysteine and glutathione. BioMetals 16, 441 – 446. Gebel, T., 1997. Arsenic and antimony: comparative approach on mechanistic toxicology. Chem. Biol. Interact. 107, 131 – 144. Gebel, T., 2000. Confounding variables in the environmental toxicology of arsenic. Toxicology 144, 155 – 162. Gebel, T., Christensen, S., Dunkelberg, H., 1997. Comparative and environmental genotoxicity of antimony and arsenic. Anticancer Res. 17, 2603 – 2607. Glindemann, D., Morgenstern, P., Wennrich, R., Stottmeister, U., Bergmann, A., 1996. Toxic oxide deposits from the combustion of landfill gas and biogas. Environ. Sci. Poll. Res. 3, 75 – 77. Haas, K., Feldmann, J., 2000. Sampling of trace volatile metal(loid) compounds in ambient air using polymer bags: a convenient method. Anal. Chem. 72, 4205 – 4211. Howard, A.G., 1997. Boro(hydride) techniques in trace element speciation. J. Anal. At. Spectrom. 12, 267 – 272. International Agency for Research on Cancer [IARC], 1989. Monographs on the evaluation of carcinogenic risk to humans: some organic solvents, resin monomers and related compounds, pigments and occupational exposures in paint manufacture and painting, vol. 47. Lyon, France. International Agency for Research on Cancer [IARC], 2002. Monographs on the evaluation of carcinogenic risks to humans: some drinkingwater disinfectants and contaminants, including arsenic, vol. 84. Lyon, France. Kentner, M., Leinemann, M., Schaller, K.H., Welte, D., Lehnert, G., 1995. External and internal antimony exposure in starter battery production. Int. Arch. Occup. Environ. Health 67, 123 – 199. Kitchin, K.T., Ahmad, S., 2003. Oxidative stress as a possible mode of action for arsenic. Toxicol. Lett. 137, 3 – 13. Kligerman, A.D., Doerr, C.L., Tennant, A.H., Harrington-Brock, K., Allen, J.W., Winkfield, E., Poorman-Allen, P., Kundu, B., Funusaka, K., Roop, B.C., Mass, M.J., DeMarini, D.M., 2003. Methylated tri-
valent arsenicals as candidate ultimate genotoxic forms of arsenic: induction of chromosomal mutations but not gene mutations. Environ. Mol. Mutagen. 42, 192 – 205. Krachler, K., Emons, H., 2001. Urinary antimony speciation by HPLCICP-MS. J. Anal. At. Spectrom. 16, 20 – 25. Kresimon, J., Gruter, U.M., Hirner, A.V., 2001. HG/LT-GC/ICP-MS coupling for identification of metal(loid) species in human urine after fish consumption. Fresenius J. Anal. Chem. 371, 586 – 590. Mandal, B.K., Suzuki, K.T., 2002. Arsenic round the world: a review. Talanta 58, 201 – 235. Mass, M.J., Tennant, A.H., Roop, B.C., Cullen, W.R., Styblo, M., Thomas, D.J., Kligerman, A.D., 2001. Methylated trivalent arsenic species are genotoxic. Chem. Res. Toxicol. 14, 355 – 361. National Research Council, 1999. Arsenic in Drinking Water. National Academy Press, Washington, DC. Nelson, J.C., 1993. PhD Thesis. The biotransformations and biomobility of monomethylarsonate and dimethylarsinate in the marine environment. University of British Columbia, Vancouver, B.C. Nesnow, S., Roop, B.C., Lambert, G., Kadiiska, M., Mason, R.P., Cullen, W.R., Mass, M.J., 2002. DNA damage induced by methylated trivalent arsenicals is mediated by reactive oxygen species. Chem. Res. Toxicol. 15, 1627 – 1634. Nishikawa, T., Wanibuchi, H., Ogawa, M., Kinoshita, A., Morimura, K., Hiroi, T., Funae, Y., Kishida, H., Nakae, D., Fukushima, S., 2002. Promoting effects of monomethylarsonic acid, dimethylarsinic acid and trimethylarsine oxide on induction of rat liver preneoplastic glutathione S-transferase placental form positive foci: a possible reactive oxygen species mechanism. Int. J. Cancer 100, 136 – 139. Risk, M., Fuortes, L., 1991. Chronic arsenicalism suspected from arsine exposure—A case report and literature review. Vet. Hum. Toxicol. 33, 590 – 595. Romeo, L., Apostoli, P., Kovacic, M., Martini, S., Brugnone, F., 1997. Acute arsine intoxication as a consequence of metal burnishing operations. Am. J. Ind. Med. 32, 211 – 216. Turpeinen, R., Pantsar-Kallio, M., Kairesalo, T., 2002. Role of microbes in controlling the speciation of arsenic and production of arsines in contaminated soils. Sci. Total Environ. 285, 133 – 145. Yamanaka, K., Hoshino, M., Okamoto, M., Sawamura, R., Hasegawa, A., Okada, S., 1990. Induction of DNA damage by dimethylarsine, a metabolite of inorganic arsenics, is for the major part likely due to its peroxyl radical. Biochem. Biophys. Res. Commun. 168, 58 – 64.
Plasmid DNA damage caused by stibine and trimethylstibine
$
Paul Andrewes, Kirk T. Kitchin, * and Kathleen Wallace Environmental Carcinogenesis Division, Office of Research and Development, National Heath and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC 27709, USA Received 10 July 2003; accepted 25 August 2003
Abstract Antimony is classified as ‘‘possibly carcinogenic to humans’’ and there is also sufficient evidence for antimony carcinogenicity in experimental animals. Stibine is a volatile inorganic antimony compound to which humans can be exposed in occupational settings (e.g., lead-acid battery charging). Because it is highly toxic, stibine is considered a significant health risk; however, its genotoxicity has received little attention. For the work reported here, stibine was generated by sodium borohydride reduction of potassium antimony tartrate. Trimethylstibine is a volatile organometallic antimony compound found commonly in landfill and sewage fermentation gases at concentrations ranging between 0.1 and 100 Ag/m3. Trimethylstibine is generally considered to pose little environmental or health risk. In the work reported here, trimethylstibine was generated by reduction of trimethylantimony dichloride using either sodium borohydride or the thiol compounds, dithioerythritol (DTE), L-cysteine, and glutathione. Here we report the evaluation of the in vitro genotoxicities of five antimony compounds—potassium antimony tartrate, stibine, potassium hexahydroxyantimonate, trimethylantimony dichloride, and trimethylstibine—using a plasmid DNA-nicking assay. Of these five antimony compounds, only stibine and trimethylstibine were genotoxic (significant nicking to pBR 322 plasmid DNA). We found stibine and trimethylstibine to be about equipotent with trimethylarsine using this plasmid DNA-nicking assay. Reaction of trimethylantimony dichloride with either glutathione or L-cysteine to produce DNA-damaging trimethylstibine was observed with a trimethylantimony dichloride concentration as low as 50 AM and Lcysteine or glutathione concentrations as low as 500 and 200 AM, respectively, for a 24 h incubation. D 2003 Elsevier Inc. All rights reserved. Keywords: Antimony; L-Cysteine; DNA nicking; Genotoxicity; Glutathione; Reactive oxygen species; Stibine; Trimethylstibine; Trimethylantimony dichloride; Trimethylarsine
Introduction The International Agency for Research on Cancer (IARC, 1989) has classified antimony (Sb), as antimony trioxide, as ‘‘possibly carcinogenic to humans’’. There is also sufficient evidence for the carcinogenicity of antimony trioxide in experimental animals (IARC, 1989). Antimony is in column 15 of the periodic table below arsenic, a wellknown human carcinogen (IARC, 2002), and antimony shares many chemical properties with arsenic. For instance, the biomethylation of antimony by microorganisms is well-
$ This manuscript has been reviewed by the National Health and Environmental Effects Research Laboratory at the U.S. Environmental Protection Agency and approved for publication. The views expressed in this paper are those of the authors and do not necessarily represent the views or policy of the U.S. Environmental Protection Agency. * Corresponding author: Fax: +1-919-685-3276. E-mail address: [email protected] (K.T. Kitchin).
0041-008X/$ - see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.taap.2003.08.012
established (Bentley and Chasteen, 2002) and is thought to proceed via a mechanism similar to the Challenger mechanism for arsenic biomethylation (Andrewes and Cullen, 2003; Andrewes et al., 2000; Challenger, 1945). The modes of carcinogenic action for arsenic (IARC, 2002; NRC, 1999) and antimony (Gebel, 1997; Gebel et al., 1997) are not established. However, methylated arsenic(III) species have been shown to be genotoxic (Ahmad et al., 2002; Andrewes et al., 2003; Mass et al., 2001; Nesnow et al., 2002) and potent clastogens (Kligerman et al., 2003), but not point mutagens. Human exposure to antimony is not as pervasive as it is for arsenic. For example, there is no instance of a population exposed to antimony that parallels that in Bangladesh, where at least 25 million people are exposed to high (>50 ppb) arsenic concentrations in drinking water (IARC, 2002; Mandal and Suzuki, 2002; NRC, 1999). However, antimony often occurs as a co-contaminant with arsenic and it may be a confounding variable in the carcinogenesis of arsenic
42
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
(Gebel, 2000). Furthermore, antimony is a component of lead-acid batteries and some of this antimony is volatilized as gaseous stibine (SbH3) at each charging of the battery. Exposure to antimony in the production of starter batteries can reach levels as high as 0.04 mg/m3 (Kentner et al., 1995) (compare with threshold limit value: 0.5 mg/m3). The health consequences of long-term exposure to stibine are unknown. Exposure to the corresponding arsenic compound, arsine (AsH3), causes oxidative damage to red blood cells (Blair et al., 1990). The end product of antimony biomethylation, trimethylstibine (Me3Sb), is commonly detected, along with trimethylarsine (Me3As), in biological fermentation gases from landfills and sewage treatment plants (Feldmann and Hirner, 1995; Feldmann et al., 1998). However, the health and environmental effects of these gases in the atmosphere surrounding the landfill are unknown. These organometallic gases pose an environmental problem for landfill-gas-fired power stations that must dispose of the toxic oxides produced by their combustion (Glindemann et al., 1996). As noted above, the toxicity of antimony and antimony compounds, such as stibine and trimethylstibine, in humans has not been investigated to the same extent as arsenic (Gebel, 1997; Gebel et al., 1997). In a recent review, Gebel (1997) concluded that there are no data available on the mechanistic genotoxicity of antimony. In the few studies that have been done with antimony, it was found to be clastogenic but not a gene mutagen (Gebel, 1997; Gebel et al., 1997); similar results exist for arsenic (Kligerman et al., 2003). We recently reported that the dithiol compound dithioerythritol (DTE), a good model for biological thiol compounds, reduced a trimethylarsenic(V) compound, trimethylarsine oxide (TMAO), to DNA-damaging trimethylarsine (Andrewes et al., 2003). Trimethylarsine oxide can be reduced by a variety of biological thiol compounds, including lipoic acid, glutathione, and cysteine, to form trimethylarsine (Cullen et al., 1984). Thus, we suggested that trimethylarsine oxide might react with any number of thiol compounds in cellular environments to form genotoxic trimethylarsine in vivo. There is some evidence that supports this hypothesis (Nishikawa et al., 2002). Because humans biomethylate inorganic antimony to trimethylantimony(V) species (Krachler and Emons, 2001), apparently, trimethylantimony(V) species formed in cells might be reduced to trimethylstibine. In the present study, we have examined the in vitro genotoxicity of five antimony compounds [three antimony(III) compounds—potassium antimony tartrate, stibine, and trimethylstibine—and two antimony(V) compounds— potassium hexahydroxyantimonate, and trimethylantimony dichloride]. Stibine was generated by sodium borohydride reduction of potassium antimony tartrate. The use of sodium borohydride to generate hydride compounds from metal and metalloid compounds is well-known in analyt-
ical and organometallic chemistry and is generally very efficient (Howard, 1997). Trimethylstibine was generated by reduction of trimethylantimony dichloride using either sodium borohydride, dithioerythritol, or the more physiologically relevant compounds, L-cysteine and glutathione. The ease of trimethylstibine formation by reduction of trimethylantimony dichloride by thiols and the high genotoxicity of trimethylstibine (comparable with trimethylarsine) suggest that trimethylstibine formation in vivo could be a possible mechanism of antimony carcinogenesis. In this paper, we also discuss the relevance of trimethylarsine and trimethylstibine genotoxicity with respect to possible health and environmental hazards posed by these compounds occurring in landfill and sewage fermentation gases.
Methods Materials. Caution: Antimony compounds are toxic and potentially carcinogenic. Trimethylstibine is a reactive volatile liquid. These compounds should be handled using appropriate safety measures. Trimethylantimony(V) dichloride, potassium antimony(III) tartrate, potassium hexahydroxyantimonate(V), dithioerythritol (DTE), L-cysteine (hydrochloride, monohydrate), and glutathione (reduced form, anhydrous) were purchased from Sigma (St. Louis, MO). Trimethylarsine oxide (TMAO) was synthesized using a method described previously (Nelson, 1993) in the laboratory of Dr. W.R. Cullen (University of British Columbia, Vancouver, Canada). Solutions of these compounds were made by dissolving an appropriate amount in TE buffer (10 mM Tris –HCl, 1 mM EDTA, pH 8.0) and adjusting the pH to 8 (as determined by pH indicator strips). Sodium borohydride, glutathione, cysteine, and DTE solutions were prepared fresh each time an experiment was performed by dissolving the appropriate amount of solid in water and, if necessary, adjusting the pH to 8. Supercoiled (SC) naked (essentially protein-free) plasmid DNA (pBR 322) was purchased from Roche (Indianapolis, IN). Agarose gel electrophoresis. The amounts of damaged (nicked) and intact DNA were determined by using agarose gel electrophoresis methods described elsewhere (Ahmad et al., 2002; Nesnow et al., 2002). Briefly, pBR 322 plasmid DNA, suspended in 45 Al of TE buffer (10 mM Tris –HCl, 1 mM EDTA, pH 8.0), was mixed with 5 Al of gel-loading solution (0.05% bromophenol blue, 50% v/v sucrose). The mixtures were loaded in 1% agarose gels containing 0.5 Ag/ml ethidium bromide (eight lanes per gel). Supercoiled (SC) fast migrating DNA was separated from linear (L) DNA (double-strand break) with moderate electrophoretic mobility and open circular (OC) slow migrating DNA (single-strand break) by electrophoresis for 2– 3 h at 50 V (constant voltage). Ethidium
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
bromide-stained DNA bands were visualized under UV light. Nicking of DNA exposed to gaseous antimony compounds. Cellulose nitrate filter paper squares (Whatman, 0.2-Am pore size, f5 5 mm) were rinsed with TE buffer and briefly allowed to dry. A pushpin was pushed through the center of an Eppendorf tube (500 Al) cap and through a filter paper square. The filter paper was gently pushed into the roof of the cap. Immediately before generating the volatile antimony compound, the target DNA solution (1 Al of 250 ng/Al) was pipetted onto the filter paper square. Antimony compounds, either potassium antimony tartrate or trimethylantimony dichloride, were pipetted into Eppendorf tubes with TE buffer to give a total liquid volume of 45 Al with final antimony concentrations ranging from 5 to 5000 AM. Next, each Eppendorf tube was held almost horizontal so that sodium borohydride (5 Al of 5 mM) could be pipetted onto the Eppendorf side wall without mixing with the other reagents. The Eppendorf was capped with the filter paper containing cap, then held vertically and tapped gently so the sodium borohydride drop would mix with the other reagents and generate either the stibine (from potassium antimony tartrate) or trimethylstibine (from trimethylantimony dichloride). After 30 min at room temperature, the filter paper was removed and placed in another Eppendorf tube to which TE buffer (50 Al) was added. The tube was thoroughly agitated for 1 min to release DNA from the filter paper. The amount of DNA damage was determined by using agarose gel electrophoresis as described above. Controls were also run where the antimony compounds were replaced with distilled water. Nicking of DNA exposed to antimony compounds in solution. In an experiment comparing the DNA-damaging activities of five different antimony compounds, solutions were prepared containing DNA (250 ng), an antimony compound (10 mM or 1 mM) and TE buffer to give a total volume of 50 Al. Stibine or trimethylstibine were generated in situ by mixing potassium antimony tartrate (10 mM) or trimethylantimony dichloride (1 mM), respectively, with sodium borohydride (10 mM). Sodium borohydride was always the last reagent added. For comparison, one sample was also run where DNA was exposed to trimethylarsine, generated in situ by reacting TMAO (1 mM) with sodium borohydride (10 mM). The control lanes consisted of DNA and sodium borohydride (10 mM). The compounds were incubated with the DNA for 1 h at 37 jC, and then the amount of DNA-nicking was determined by using agarose gel electrophoresis as described above. In one experiment, solutions were prepared containing DNA (250 ng) and trimethylantimony dichloride (5 –5000 AM). Trimethylstibine was generated by adding 5 Al of sodium borohydride (5 mM). After 30 min, the amount of
43
DNA damage was determined by using agarose gel electrophoresis as described above. In the same experiment, DNA suspended on filter paper squares was exposed to the gaseous antimony compounds that were volatilized from the reaction mixtures into the headspace of the vials, as described above. In other experiments, solutions were made containing 1 Al of DNA (250 ng/Al), TE buffer, trimethylantimony dichloride (20 – 5000 AM), and a thiol-reducing agent (20 –5000 AM), either DTE, L-cysteine, or glutathione, to give a total volume of 45 Al. The solutions were made in 500-Al Eppendorf tubes. The reducing agent was always the last reagent added. Controls were also run consisting of either DNA alone or DNA and reducing agent (5 mM) or DNA and trimethylantimony dichloride (5 mM). The solutions were gently agitated and then incubated for 24 h at 37 jC, and the amount of DNA damage was determined by using agarose gel electrophoresis as described above.
Results The results reported here are representative of duplicate experiments that were reproducible in all cases. No evidence from control assays suggests that nuclease contamination caused DNA damage. Supercoiled plasmid DNA (pBR 322) was exposed to five different antimony compounds and, for comparison, trimethylarsine. DNA damage was determined by using agarose gel electrophoresis (Fig. 1). We used
Fig. 1. Agarose gel electrophoresis of pBR 322 plasmid DNA after a 1-h incubation (37 jC, pH 8.0) with five different antimony compounds and, for comparison, trimethylarsine. Stibine, trimethylstibine, and trimethylarsine were made by reacting sodium borohydride (10 mM) with potassium antimony tartrate, trimethylantimony dichloride, and trimethylarsine oxide, respectively. The two control lanes contained 10 mM sodium borohydride. SC, supercoiled form of plasmid DNA; OC, open circular form of plasmid DNA (after a single-strand break).
44
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
Fig. 2. Agarose gel electrophoresis of pBR 322 plasmid DNA that was blotted onto cellulose nitrate filter paper and suspended for 30 min above reaction mixtures producing stibine. The reaction mixtures contained sodium borohydride (50 mM) and potassium antimony tartrate (100 – 5000 AM). SC, supercoiled form of plasmid DNA; OC, open circular form of plasmid DNA (after a single-strand break).
potassium antimony tartrate as a source of antimony(III) because it is much more water soluble than antimony trioxide or antimony trichloride, which hydrolyses to antimony trioxide. Potassium antimony tartrate and antimony trioxide in solution are probably in equilibrium with antimonite [Sb(OH)3] (Gebel et al., 1997). We also examined
the genotoxicity of antimony pentachloride (which hydrolyses to antimony pentoxide). Although this compound was not very water soluble, a saturated solution of antimony pentachloride did not damage DNA (results not shown). The two antimony compounds that were genotoxic were trimethylstibine and stibine (Fig. 1). Trimethylstibine and stibine were generated in situ by the reaction of sodium borohydride with trimethylantimony dichloride and potassium antimony tartrate, respectively. There was no evidence that sodium borohydride or any other component of the reaction mixture except stibine or trimethylstibine damaged DNA. Sodium borohydride and byproducts of the sodium borohydride reaction do not damage DNA in this assay (Andrewes et al., 2003). If it is assumed that trimethylstibine and trimethylarsine are generated with equal efficiency from their precursor compounds, then apparently, trimethylstibine and trimethylarsine were about equipotent at damaging plasmid DNA. In one experiment, we placed DNA in reaction mixtures generating stibine and also suspended DNA above the reaction mixtures. Here, stibine did minimal damage to the DNA in solution (results not shown) but substantial damage to the DNA on the filter paper (Fig. 2). Stibine is a volatile compound (boiling point: 17 jC) that will quickly volatilize into the headspace of the reaction vessel. Although we could not determine the concentration of stibine in the gas phase to which the DNA on the filter paper was exposed, we can estimate the minimum stibine concentration needed to damage DNA assuming the borohydride reaction is 100% efficient and all generated stibine remains in the gas phase. We estimated the minimum gas-phase stibine concentration required to do DNA damage to be 6000 mg/m3. The exposure time in this experiment was
Fig. 3. Agarose gel electrophoresis of pBR 322 plasmid DNA that was blotted onto cellulose nitrate filter paper and suspended for 30 min above reaction mixtures producing trimethylstibine (a) and agarose gel electrophoresis of pBR 322 plasmid DNA that was dissolved in those reaction mixtures (b). The reaction mixtures contained sodium borohydride (5 mM) and trimethylantimony dichloride (5 – 5000 AM). SC, supercoiled form of plasmid DNA; L, linear form of plasmid DNA (after a double-strand break); OC, open circular form of plasmid DNA (after a single-strand break).
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
30 min. In previous work with trimethylarsine generation, we showed that the DNA on filter paper squares suspended in vial headspace gases was exposed only to gaseous compounds but not to compounds in the aqueous phase that are transferred to the filter paper as an aerosol (Andrewes et al., 2003). Trimethylstibine is a volatile compound (boiling point: 81 jC) with low water solubility. Trimethylstibine is less volatile than stibine (boiling point: 17 jC). Thus, when we placed DNA in a reaction mixture generating trimethylstibine and also suspended DNA above the reaction mixture, DNA damage was seen for both DNA in solution and DNA suspended in the vial headspace (Fig. 3). For the DNA placed in the aqueous reaction mixture, damage was clearly
45
observed relative to the controls with trimethylantimony dichloride concentrations of 200 AM and higher (Fig. 3b). DNA damage for the DNA on filter paper became clearly discernable, relative to the controls, with a liquid-phase trimethylantimony dichloride concentration of f5000 AM (Fig. 3a). The different response for exposures from the gas phase versus the liquid phase is mostly a consequence of dilution, due to going from a volume of f50 Al (liquid volume) to f500 Al (vial headspace volume). The DNA damage occurred in 30 min or less. We do not know the concentration of gaseous trimethylstibine to which the DNA on the filter paper was exposed, when it was suspended above reaction mixtures. But assuming 100% efficiency of the borohydride reaction and that all trimethylstibine
Fig. 4. Agarose gel electrophoresis of pBR 322 plasmid DNA after a 24-h incubation at 37 jC with trimethylantimony dichloride (20 – 5000 AM) and either 5 mM glutathione (a) or 5 mM L-cysteine (b) in TE buffer (pH 8.0). In comparison, under the same conditions, the trimethylantimony dichloride concentration was held constant (5 mM) and either the glutathione (c) or L-cysteine (d) concentration was titrated (20 – 5000 AM). SC, supercoiled form of plasmid DNA; OC, open circular form of plasmid DNA (after a single-strand break).
46
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
remains in the gas phase, we estimate that the minimum gas-phase trimethylstibine concentration required to cause DNA damage, in 30 min or less, was 10 000 mg/m3. If the above assumptions are not correct, then the potency of trimethylstibine will be underestimated. In a preliminary experiment, supercoiled DNA was incubated with trimethylantimony dichloride and dithioerythritol (DTE) for 2 h at 37 jC. DNA damage occurred for mixtures of DTE (5 mM) and trimethylantimony dichloride concentrations greater than or equal to 200 AM (results not shown). DTE is a dithiol compound that can reduce arsenic(V) and antimony(V) compounds to arsenic(III) and antimony(III) compounds, and in a previous study, we found that DTE reduced trimethylarsine oxide to DNA-damaging trimethylarsine (Andrewes et al., 2003). Analogously, DTE reduces trimethylantimony dichloride to DNA-damaging trimethylstibine. Exposure to DTE or trimethylantimony dichloride alone did not cause DNA damage (results not shown). Further experiments were performed using the physiologically relevant thiol-reducing agents, glutathione and Lcysteine (Fig. 4). Supercoiled DNA was incubated for 24 h at 37 jC in TE buffer (pH 8.0) with various concentrations of L-cysteine, glutathione, and trimethylantimony dichloride. In one series (Figs. 4a, b) of experiments, the concentration of reducing agent was held constant and the trimethylantimony dichloride was titrated. In both cases, the lowest amount of trimethylantimony dichloride required to cause DNA damage in the presence of excess glutathione (Fig. 4a) and excess L-cysteine (Fig. 4b) was approximately 50 AM. In a second series of experiments (Figs. 4c, d), the concentration of trimethylantimony dichloride was kept constant and the concentration of reducing agent was titrated. The lowest amount of glutathione required to cause DNA damage, by reducing trimethylantimony dichloride, was approximately 500 AM (Fig. 4c), and for L-cysteine, the amount required was approximately 200 AM (Fig. 4d). Both glutathione and L-cysteine are found in the body in millimolar concentrations. Again, the DNA-damaging action of trimethylantimony dichloride in the presence of Lcysteine or glutathione was most likely due to the reduction of trimethylantimony dichloride to trimethylstibine that damages DNA.
Discussion Recently, we demonstrated the arsenic compound, trimethylarsine (Me3As), to be genotoxic (Andrewes et al., 2003). In this paper, we have examined the genotoxicity of the corresponding antimony compound, trimethylstibine (Me3Sb), as well as four other antimony compounds: potassium antimony tartrate, stibine, potassium hexahydroxyantimonate, and trimethylantimony dichloride. Of these five antimony compounds, only stibine and trimethylstibine were genotoxic.
Notably, trimethylstibine and trimethylarsine (Andrewes et al., 2003) are approximately equipotent genotoxins in the plasmid DNA-nicking assay (of the eleven arsenicals that we have previously evaluated, Andrewes et al., 2003, trimethylarsine was one of the most potent genotoxic arsenicals). Under equivalent conditions, the minimum concentration of compound in solution required to cause DNA damage for both trimethylstibine and trimethylarsine was 200 AM. It is assumed that trimethylstibine and trimethylarsine are generated with equal efficiency from their precursor compounds by reaction with sodium borohydride. Sodium borohydride is a commonly used reagent in analytical chemistry that is known to react rapidly and in high yield with many metal and metalloid compounds to form hydrides (Howard, 1997). The contrast between the genotoxicity of stibine and the lack of genotoxicity of arsine (Andrewes et al., 2003) is striking. The most likely means for arsines or stibines to cause oxidative damage to DNA are by the production of reactive oxygen species (ROS). There is a growing body of evidence suggesting that oxidative stress plays a central role in both arsenic and antimony toxicity (Gebel et al., 1997; Kitchin and Ahmad, 2003). This ROS-dependant mechanism has been studied extensively for dimethylarsine (Yamanaka et al., 1990), and arsine causes oxidative damage to red blood cells (Blair et al., 1990). If DNA damage occurs by the generation of ROS, then the capacity of arsines and stibines to damage DNA may reflect their oxidative stability. For arsines, this hypothesis is correct. Trimethylarsine and dimethylarsine, which are more susceptible to oxidation (Haas and Feldmann, 2000), were far more damaging to DNA (Andrewes et al., 2003) than were monomethylarsine and arsine, which are more resistant to oxidation (Haas and Feldmann, 2000). Our work on antimony compounds illustrates the same trend. Stibines are more susceptible to oxidation than the corresponding arsines (Haas and Feldmann, 2000). Hence, stibine caused significant damage to DNA whereas arsine did not (Andrewes et al., 2003). In addition, 1 mM trimethylstibine was slightly more damaging to DNA than was 1 mM trimethylarsine (Fig. 1). In the case of arsenic, trimethylarsine was much more DNA damaging than was arsine (Andrewes et al., 2003). Thus, we expected trimethylstibine to be much more damaging to DNA than stibine; however, this was not the case. Stibine and trimethylstibine are formed with approximately equal efficiency from their precursor compounds (Andrewes, 2000). In conclusion, these results for antimony are consistent with our earlier results for arsenic, and the correlation between oxidative stability of arsines and stibines with DNA damage possibly suggests that group 15 compounds may damage DNA by the generation of ROS. We estimated the minimum gas-phase trimethylstibine concentration required to damage DNA (in 30 min or less) to be 10 000 mg/m3. This is many orders of magnitude greater than the typical trace quantities of trimethylstibine
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
(10 – 100 Ag/m3) found in fermentation gases (Feldmann and Hirner, 1995; Glindemann et al., 1996). The permissible exposure level (OSHA PEL) for trimethylstibine is 500 Ag/m3. Thus, this suggests that trimethylstibine in fermentation gases poses a minimal health hazard. Similarly, for stibine, we estimated that the minimum gas-phase concentration required to cause DNA damage was 6000 mg/m3, which is four orders of magnitude greater than the OSHA PEL. Although concentrations of arsines and stibines in the environment are typically quite low, high concentrations have been reported under certain circumstances. For instance, a very high concentration of trimethylarsine (1000 mg/m3) was reported for gas extracted from arseniccontaminated soil (Turpeinen et al., 2002), and some interesting cases of occupational exposure to arsine have been reported (Risk and Fuortes, 1991; Romeo et al., 1997). Arsines and stibines (including Me3As, AsH3, Me3Sb, and SbH3) are used in the semiconductor industry with operating procedures that should minimize exposure. Given the in vitro genotoxicity of these compounds demonstrated in our studies, it would be prudent to examine the genotoxicity and mutagenicity of arsines and stibines using more sensitive and physiologically relevant assays. Antimony(V) and arsenic(V) compounds are readily reduced to the corresponding antimony(III) and arsenic(III) compounds by thiols, including DTE, glutathione, and Lcysteine (Cullen et al., 1984; Ferreira et al., 2003). Thus, we have previously shown that trimethylarsine oxide is reduced to trimethylarsine in the presence of DTE, which then causes DNA damage (Andrewes et al., 2003). In this work, we have shown that DNA damage occurs when trimethylantimony dichloride is incubated with DNA and either DTE, glutathione, or L-cysteine. Damage occurs because these thiol compounds reduced the trimethylantimony dichloride to trimethylstibine, which damages DNA. Analogously, when trimethylarsine oxide was incubated with glutathione or L-cysteine and DNA, nicking of DNA occurred due to the production of trimethylarsine, which damages DNA (Andrewes and Kitchin, unpublished results). Human exposure to trimethylstibine, generated during metabolism of inorganic antimony, should be further considered. If trimethylantimony(V) species (e.g., trimethylantimony dichloride) are formed in vivo, then as demonstrated above, these are likely to be reduced by cellular thiols to trimethylstibine, which damages DNA. Mammalian and human metabolism of antimony has not been investigated in much detail (Andrewes and Cullen, 2003), but antimony biomethylation is well-known in microorganisms (Bentley and Chasteen, 2002). Trimethylantimony dichloride was detected in urine of humans occupationally exposed to antimony (Krachler and Emons, 2001), and Kresimon et al. (2001) detected trace levels ( < 0.1 nM) of monomethylantimony, dimethylantimony, and trimethylantimony spe-
47
cies in urine of humans not occupationally exposed to antimony. Thus, one mode of action for antimony carcinogenesis could be metabolism to trimethylantimony(V) species and reduction to trivalent trimethylstibine that could cause DNA damage and mutation. However, trimethylantimony dichloride was not mutagenic in Salmonella TA104 (DeMarini and Funasaka, unpublished results). In summary, we identified stibine and trimethylstibine as two genotoxic forms of antimony that damaged DNA probably by the generation of ROS. No other forms of antimony were genotoxic. Our conclusions were based on assuming that the generation of stibine and trimethylstibine using sodium borohydride is 100% efficient, if this is not the case then we might have underestimated the potency of these antimony compounds. The effect of arsines and stibines on the environment and on human health warrants further consideration.
Acknowledgments This work was partially supported (P.A.) by a cooperative agreement between the National Research Council (Washington, DC) and the U.S. Environmental Protection Agency. We thank Bill Cullen, David DeMarini, Julian Preston, Jeff Ross, Miroslav Styblo, and David Thomas for their assistance.
References Ahmad, S., Kitchin, K.T., Cullen, W.R., 2002. Plasmid DNA damage caused by methylated arsenicals, ascorbic acid and human liver ferritin. Toxicol. Lett. 133, 47 – 57. Andrewes, P., 2000. The biomethylation of arsenic and antimony by Scopulariopsis brevicualis. PhD Thesis. University of British Columbia, Vancouver, B.C. Andrewes, P., Cullen, W.R., 2003. Organoantimony compounds in the environment. In: Craig, P.J. (Ed.), Organometallic Compounds in the Environment. Wiley, West Sussex, pp. 277 – 304. Andrewes, P., Cullen, W.R., Polishchuk, E., 2000. Antimony biomethylation by Scopulariopsis brevicualis: characterization of intermediates and the methyl donor. Chemosphere 41, 1717 – 1725. Andrewes, P., Kitchin, K.T., Wallace, K., 2003. Dimethylarsine and trimethylarsine are potent genotoxins in vitro. Chem. Res. Toxicol. 16, 994 – 1003. Bentley, R., Chasteen, T.G., 2002. Microbial methylation of metalloids: arsenic, antimony, and bismuth. Microbiol. Mol. Biol. Rev. 66, 250 – 265. Blair, P.C., Thompson, M.B., Bechtold, M., Wilson, R.E., Moorman, M.P., Fowler, B.A., 1990. Evidence for oxidative damage to red blood cells in mice induced by arsine gas. Toxicology 63, 25 – 34. Challenger, F., 1945. Biological methylation. Chem. Rev. 36, 315 – 361. Cullen, W.R., McBride, B.C., Reglinski, J., 1984. The reduction of trimethylarsine oxide to trimethylarsine by thiols: a mechanistic model for the biological reduction of arsenicals. J. Inorg. Biochem. 21, 45 – 60. Feldmann, J., Hirner, A.V., 1995. Occurrence of volatile metal and metalloid species in landfill and sewage gases. Int. J. Environ. Anal. Chem. 60, 339 – 359.
48
P. Andrewes et al. / Toxicology and Applied Pharmacology 194 (2004) 41–48
Feldmann, J., Koch, I., Cullen, W.R., 1998. Complementary use of capillary gas chromatography mass spectrometry (ion trap) and gas chromatography inductively coupled plasma mass spectrometry for the speciation of volatile antimony, tin and bismuth compounds in landfill and fermentation gases. Analyst 123, 815 – 820. Ferreira, C.D., Martins, P.S., Demicheli, C., Brochu, C., Ouellette, M., Frezerd, F., 2003. Thiol-induced reduction of antimony(V) into antimony(III): a comparative study with trypanothione, cysteinyl – glycine, cysteine and glutathione. BioMetals 16, 441 – 446. Gebel, T., 1997. Arsenic and antimony: comparative approach on mechanistic toxicology. Chem. Biol. Interact. 107, 131 – 144. Gebel, T., 2000. Confounding variables in the environmental toxicology of arsenic. Toxicology 144, 155 – 162. Gebel, T., Christensen, S., Dunkelberg, H., 1997. Comparative and environmental genotoxicity of antimony and arsenic. Anticancer Res. 17, 2603 – 2607. Glindemann, D., Morgenstern, P., Wennrich, R., Stottmeister, U., Bergmann, A., 1996. Toxic oxide deposits from the combustion of landfill gas and biogas. Environ. Sci. Poll. Res. 3, 75 – 77. Haas, K., Feldmann, J., 2000. Sampling of trace volatile metal(loid) compounds in ambient air using polymer bags: a convenient method. Anal. Chem. 72, 4205 – 4211. Howard, A.G., 1997. Boro(hydride) techniques in trace element speciation. J. Anal. At. Spectrom. 12, 267 – 272. International Agency for Research on Cancer [IARC], 1989. Monographs on the evaluation of carcinogenic risk to humans: some organic solvents, resin monomers and related compounds, pigments and occupational exposures in paint manufacture and painting, vol. 47. Lyon, France. International Agency for Research on Cancer [IARC], 2002. Monographs on the evaluation of carcinogenic risks to humans: some drinkingwater disinfectants and contaminants, including arsenic, vol. 84. Lyon, France. Kentner, M., Leinemann, M., Schaller, K.H., Welte, D., Lehnert, G., 1995. External and internal antimony exposure in starter battery production. Int. Arch. Occup. Environ. Health 67, 123 – 199. Kitchin, K.T., Ahmad, S., 2003. Oxidative stress as a possible mode of action for arsenic. Toxicol. Lett. 137, 3 – 13. Kligerman, A.D., Doerr, C.L., Tennant, A.H., Harrington-Brock, K., Allen, J.W., Winkfield, E., Poorman-Allen, P., Kundu, B., Funusaka, K., Roop, B.C., Mass, M.J., DeMarini, D.M., 2003. Methylated tri-
valent arsenicals as candidate ultimate genotoxic forms of arsenic: induction of chromosomal mutations but not gene mutations. Environ. Mol. Mutagen. 42, 192 – 205. Krachler, K., Emons, H., 2001. Urinary antimony speciation by HPLCICP-MS. J. Anal. At. Spectrom. 16, 20 – 25. Kresimon, J., Gruter, U.M., Hirner, A.V., 2001. HG/LT-GC/ICP-MS coupling for identification of metal(loid) species in human urine after fish consumption. Fresenius J. Anal. Chem. 371, 586 – 590. Mandal, B.K., Suzuki, K.T., 2002. Arsenic round the world: a review. Talanta 58, 201 – 235. Mass, M.J., Tennant, A.H., Roop, B.C., Cullen, W.R., Styblo, M., Thomas, D.J., Kligerman, A.D., 2001. Methylated trivalent arsenic species are genotoxic. Chem. Res. Toxicol. 14, 355 – 361. National Research Council, 1999. Arsenic in Drinking Water. National Academy Press, Washington, DC. Nelson, J.C., 1993. PhD Thesis. The biotransformations and biomobility of monomethylarsonate and dimethylarsinate in the marine environment. University of British Columbia, Vancouver, B.C. Nesnow, S., Roop, B.C., Lambert, G., Kadiiska, M., Mason, R.P., Cullen, W.R., Mass, M.J., 2002. DNA damage induced by methylated trivalent arsenicals is mediated by reactive oxygen species. Chem. Res. Toxicol. 15, 1627 – 1634. Nishikawa, T., Wanibuchi, H., Ogawa, M., Kinoshita, A., Morimura, K., Hiroi, T., Funae, Y., Kishida, H., Nakae, D., Fukushima, S., 2002. Promoting effects of monomethylarsonic acid, dimethylarsinic acid and trimethylarsine oxide on induction of rat liver preneoplastic glutathione S-transferase placental form positive foci: a possible reactive oxygen species mechanism. Int. J. Cancer 100, 136 – 139. Risk, M., Fuortes, L., 1991. Chronic arsenicalism suspected from arsine exposure—A case report and literature review. Vet. Hum. Toxicol. 33, 590 – 595. Romeo, L., Apostoli, P., Kovacic, M., Martini, S., Brugnone, F., 1997. Acute arsine intoxication as a consequence of metal burnishing operations. Am. J. Ind. Med. 32, 211 – 216. Turpeinen, R., Pantsar-Kallio, M., Kairesalo, T., 2002. Role of microbes in controlling the speciation of arsenic and production of arsines in contaminated soils. Sci. Total Environ. 285, 133 – 145. Yamanaka, K., Hoshino, M., Okamoto, M., Sawamura, R., Hasegawa, A., Okada, S., 1990. Induction of DNA damage by dimethylarsine, a metabolite of inorganic arsenics, is for the major part likely due to its peroxyl radical. Biochem. Biophys. Res. Commun. 168, 58 – 64.