|Home | About | Journals | Submit | Contact Us | Français|
A series of prodrugs of 3-mercaptopyruvate (3-MP), the substrate for the enzyme 3-mercaptopyruvate/cyanide sulfurtransferase (3-MPST) that converts cyanide to the nontoxic thiocyanate, which are highly effective cyanide antidotes, have been developed. These prodrugs of 3-MP are unique in being not only orally bioavailable, but may be administered up to an hour prior to cyanide as a prophylactic agent and are both rapid- or slow-acting when given parenterally.
Antidotes for the therapeutic management of cyanide poisoning, especially in the U.S., have relied mainly on the enzyme rhodanese (thiosulfate/cyanide sulfurtransferase, EC 18.104.22.168) for detoxification. This enzyme extracts the sulfane sulfur from its substrate, thiosulfate, to form an activated-sulfane complex, which is susceptible to nucleophilic attack by cyanide with the formation of the less-toxic thiocyanate, that is excreted in the urine.1 Rhodanase is concentrated in the liver and kidneys where it is found in the mitochondrial matrix, a site of low accessibility for ionized, inorganic species, such as thiosulfate.2 This compartmentation of rhodanase in mammalian tissues leaves major targets of cyanide lethality, namely, the heart3 and central nervous system4–6 unprotected. (Rhodanase is also found in red blood cells, but its relative function has not been clarified.3)
Clinically, the vasodilator sodium nitrite is used in combination with sodium thiosulfate, both administered intravenously in tandem.1,6,7 Sodium nitrite oxidizes hemoglobin to methemoglobin, which has a high affinity for cyanide, thereby allowing the sequestration of cyanide as cyanomethemoglobin and preventing it from binding to cytochrome oxidase.1 However, this combination is far from a simple antidote. Methemoglobin formation by nitrite is slow and must be accelerated by intranasal administration of a more rapid acting and volatile amyl nitrite, although this may not be the only mechanism of action by this compound.8 In addition, nitrite must be carefully monitored to avoid toxicity and is contraindicated when the formation of carboxyhemoglobin by carbon monoxide inhalation (as in victims of fires) further compromises the oxygenation of hemoglobin. Recently, a cyanide-trapping agent used in France, namely, hydroxocobalamin,9 has been approved by the FDA for use in the U.S. Hydroxocobalamin (MW 1355), which is administered intravenously, sequesters one mole of cyanide (MW 27) per mole of antidote to form cyanocobalamin (vitamin B-12).10
The need for a new antidote capable of being administered in a timely manner to the mass-exposed cyanide victims in the event of a chemical disaster or terrorist action is acute, and the present lack of preparedness for such a disaster is alarming.11 Accordingly, the accelerated development of an alternative cyanide antidote is of high priority. The exploitation of the ubiquitous enzyme present in both the cytoplasm and mitochondria of cells, namely, 3-MPST,a whose natural substrate is the cysteine catabolite, 3-MP, has been suggested by Porter and Baskin12 and by Nagahara et al.13 However, 3-MP is chemically unstable, and attempts at intravenous administration to counteract the toxicity of cyanide were unsuccessful due to this instability,14 hence, our prodrug approach.
The three structurally diverse prototype prodrugs of 3-MP that we developed are depicted in Figure 1. Compounds 1 and 2 are sulfhydryl-protected on the mercaptopyruvate, whereas for compound 3, only the cysteine sulfur is protected. However, the thiomethyl group of 3 is relatively unreactive; for example, we were unable to acetylate this group using standard reagents. Other derivatives related to this series will be reported in a later paper. Compounds 1a and 1b require sequential cleavage of the enol acetate and the thiol acetate functional groups (or vice versa) by esterase action in vivo to liberate 3-MP, whereas the prototype compounds 2 and 3 dissociate nonenzymatically to liberate 3-MP or its ethyl ester. In addition, the prototype compound 3 was designed to provide not only 3-MP (or its ethyl ester) in vivo, but also to release an equivalent of L-cysteine, the immediate biochemical precursor of and the endogenous reservoir for 3-MP. 2-Substituted thiazolidine-4-carboxylic acids similar to 3 are known to undergo nonenzymatic, hydrolytic opening of the thiazolidine ring in vivo to provide L-cysteine, the rate-limiting amino acid for the biosynthesis of glutathione.15,16 L-Cysteine can also be oxidized in vivo to L-cystine, which can be further converted metabolically to L-thiocystine, a known sulfane sulfur-generating cyanide antagonist.17 Compounds 1 and 2 (Figure 1) were readily prepared from commercially available starting materials, while the prototype 3 was synthesized from compound 2b by condensation of its dissociated, monomeric form (of ethyl-3-MP) in situ with L-cysteine (Supporting Information). The configurations of compounds 1a, 2a, and 2b were verified by X-ray crystallography, while the structure of 1b (an oil) was assigned by analogy to 1a. Of interest is the 6-fold alternating axis (center) of symmetry18 displayed by compound 2 (Supporting Information).
These prototype prodrugs of 3-MP were evaluated for their antidotal efficacy against a toxic, but nonlethal, cyanide dose in mice, using a test paradigm that we developed to allow for 24 h survival of the animals, a requirement imposed by our Institutional Animal Care and Use Committee (IACUC). Thus, instead of attempting the duplication of a historical cyanide LD50 dose and measuring the shift in this LD50 by preadministration of the putative cyanide antidote, we measured the righting reflex recovery times of animals given cyanide alone and cyanide-plus-antidote, using minimal (but sufficient) numbers (n) of mice commensurate with a statistical power that allowed for detailed analyses of the data.19
Table 1 documents our observation that compounds 1a, 1b, 2a, 2b, and 3, when administered orally by gavage 30 min prior to the dose of cyanide, were fully protective against cyanide toxicity according to our recovery time criteria, with an average Protective Index (P.I., the ratio of the righting time of the cyanide-only controls divided by the righting times of the cyanide-plus-antidote-treated animals) of 3.78, with the P.I. for compounds 2a and 2b averaging 4.3. Because compounds 1b and 3 were specifically designed as double prodrugs of 3-MP, that is, requiring two activating steps for the release of 3-MP in vivo, they were evaluated for their oral prophylactic efficacy against cyanide by giving these compounds to mice a full one hour before the administration of the cyanide. As can be seen (Table 2), the results were still very significant, with P.I.s of 2.0 and 2.1 (p < 0.0001). Prophylactic agents against cyanide, heretofore nonexistent, are highly desirable for firemen and rescue workers responding to industrial and residential fires where the presence of cyanide in the smoke is life-threatening.
The comparative antidotal efficacies of several of our prototype 3-MP prodrugs against a standard (human) dose of the currently available cyanide antidotes in the U.S., namely, nitrite/thiosulfate (N/T) combinations and hydroxocobalamin (H), are shown in Table 3. The antidotes were administered intraperitoneally (i.p.; N/T and H cannot be given orally) 5 min before cyanide, and the average righting times of the mice were analyzed statistically. Except for compound 3, which was expected to be slow-acting by design, compounds 1a, 1b, and 2b were found to be rapid-acting and superior to either the N/T combination or hydroxocobalamin based on their relative protective indices (Table 3).
When administered i.p. 5 min postcyanide, compound 2b with a P.I. of 3.9 was found to be highly effective in countering the toxic effect of cyanide, with compound 1a being somewhat less (Table 4). This is contrasted to the minimal protection, offered by H and the N/T combination.
In summary, by providing cells directly with prodrug forms of 3-MP, the substrate for 3-MPST, thus effectively bypassing the necessity for its enzymatic generation from L-cysteine, an amino acid whose tissue concentrations are known to be tightly controlled at low levels,20 a unique series of highly effective cyanide antidotes has been discovered. Although it has been reported that, except for 3-MP itself and its oxidized disulfide form, there are no other substrates for 3-MPST, the fact that all of the compounds that initially liberate the ethyl ester of 3-MP showed good and sometimes better potency than those that liberated 3-MP in protecting mice against cyanide, suggest that ethyl 3-mercaptopyruvate must also be a substrate for this enzyme as well. Further proof requires kinetic analysis with purified 3-MPST. These 3-MP prodrugs were shown to be not only protective by the intraperitoneal route in mice, but also when given orally, even 30–60 min prior to cyanide. Thus, they are amenable for use as prophylactic agents by first responders in domestic fires, as well as by military personnel in preparation for an impending cyanide threat. Also, because they are synthesized from readily available starting materials, they can be produced in quantity for wide distribution to public health agencies for stockpiling to protect the populace in the event of a major cyanide disaster resulting from an industrial accident or terrorist activity.
We thank Jacquie Briggs and Mary Mullett for technical assistance, Jonathan Cohen for recording the NMR spectra, and Victor G. Young, Jr. and Benjamin E. Kucera for the x-ray data. This study was supported by a Seed Grant from the Center for Drug Design, Academic Health Center, University of Minnesota and by the National Institutes of Health CounterACT Program through the National Institute of Neurological Disorders and Stroke (Award #UO1NS058087-01).
aAbbreviations: 3-MP, 3-mercaptopyruvate; 3-MPST, 3-mercaptopyruvate sulfurtransferase (EC 22.214.171.124); N/T, sodium nitrite-sodium thiosulfate combination; H, hydroxocobalamine.
Supporting Information Available: Experimental details for the preparation of 1a, 1b, 2a, 2b, and 3. The X-ray crystal structures of 1a, 2a, and 2b are also provided. This material is available free of charge via the Internet at http://pubs.acs.org.