Radiolabelled soman binding to sera from Rats, Guinea Pigs and Monkeys.

Soman is a highly toxic organophosphorus chemical warfare compound that binds rapidly and irreversibility to a variety of serine active enzymes, i.e., butyryl- and acetyl-cholinesterases and carboxylesterase. The in vivo toxicity of soman has been reported to vary significantly in different animal species, such as rats and guinea pigs or non-human primates. This species variation makes it difficult to identify appropriate animal models for therapeutic drug development under the US Food and Drug Administration (FDA) Animal Rule. Since species variation in soman toxicity has been correlated with species variation in serum carboxylesterase, we undertook to determine if serum from guinea pigs, rats and non-human primates bound different levels of soman in vitro in the presence of equimolar concentrations of soman. Our results demonstrated that the amount of soman bound in the serum of rats was 4 uM, but essentially null in guinea pigs or non-human primates. The results strongly correlate with the presence or absence of carboxylesterase in the serum of animals and the difference in the toxic dose of soman in various species. Our results support prior suggestions that guinea pigs and non-human primates may be better animal models for the development of antidotes under the FDA Animal Rule.

Design, synthesis, and characterization of novel, nonquaternary reactivators of GF-inhibited human acetylcholinesterase.

The goal of this research was to identify structurally novel, non-quaternarypyridinium reactivators of GF (cyclosarin)-inhibited hAChE that possess the capacity to mediate in vitro reactivation of GF-inhibited human acetylcholinesterase (hAChE). New compounds were designed, synthesized and assessed in GF-inhibited hAChE assays. Structure activity relationships for AChE binding and reactivation of GF-inhibited hAChE were developed. Lead compounds from two different chemical series, represented by compounds 17 and 38, displayed proficient in vitro reactivation of GF-inhibited hAChE, while also possessing low inhibition of native enzyme.

A common mechanism for resistance to oxime reactivation of acetylcholinesterase inhibited by organophosphorus compounds.

Administration of oxime therapy is currently the standard approach used to reverse the acute toxicity of organophosphorus (OP) compounds, which is usually attributed to OP inhibition of acetylcholinesterase (AChE). Rate constants for reactivation of OP-inhibited AChE by even the best oximes, such as HI-6 and obidoxime, can vary >100-fold between OP-AChE conjugates that are easily reactivated and those that are difficult to reactivate. To gain a better understanding of this oxime specificity problem for future design of improved reactivators, we conducted a QSAR analysis for oxime reactivation of AChE inhibited by OP agents and their analogues. Our objective was to identify common mechanism(s) among OP-AChE conjugates of phosphates, phosphonates and phosphoramidates that result in resistance to oxime reactivation. Our evaluation of oxime reactivation of AChE inhibited by a sarin analogue, O-methyl isopropylphosphonofluoridate, or a cyclosarin analogue, O-methyl cyclohexylphosphonofluoridate, indicated that AChE inhibited by these analogues was at least 70-fold more difficult to reactivate than AChE inhibited by sarin or cyclosarin. In addition, AChE inhibited by an analogue of tabun (i.e., O-ethyl isopropylphosphonofluoridate) was nearly as resistant to reactivation as tabun-inhibited AChE. QSAR analysis of oxime reactivation of AChE inhibited by these OP compounds and others suggested that the presence of both a large substituent (i.e., ≥ the size of dimethylamine) and an alkoxy substituent in the structure of OP compounds is the common feature that results in resistance to oxime reactivation of OP-AChE conjugates whether the OP is a phosphate, phosphonate or phosphoramidate.

Comparison of oxime reactivation and aging of nerve agent-inhibited monkey and human acetylcholinesterases.

Non-human primates are valuable animal models that are used for the evaluation of nerve agent toxicity as well as antidotes and results from animal experiments are extrapolated to humans. It has been demonstrated that the efficacy of an oxime primarily depends on its ability to reactivate nerve agent-inhibited acetylcholinesterase (AChE). If the in vitro oxime reactivation of nerve agent-inhibited animal AChE is similar to that of human AChE, it is likely that the results of an in vivo animal study will reliably extrapolate to humans. Therefore, the goal of this study was to compare the aging and reactivation of human and different monkey (Rhesus, Cynomolgus, and African Green) AChEs inhibited by GF, GD, and VR. The oximes examined include the traditional oxime 2-PAM, two H-oximes HI-6 and HLo-7, and the new candidate oxime MMB4. Results indicate that oxime reactivation of all three monkey AChEs was very similar to human AChE. The maximum difference in the second-order reactivation rate constant between human and three monkey AChEs or between AChEs from different monkey species was 5-fold. Aging rate constants of GF-, GD-, and VR-inhibited monkey AChEs were very similar to human AChE except for GF-inhibited monkey AChEs, which aged 2-3 times faster than the human enzyme. The results of this study suggest that all three monkey species are suitable animal models for nerve agent antidote evaluation since monkey AChEs possess similar biochemical/pharmacological properties to human AChE.

A structure-activity analysis of the variation in oxime efficacy against nerve agents.

A structure-activity analysis was used to evaluate the variation in oxime efficacy of 2-PAM, obidoxime, HI-6 and ICD585 against nerve agents. In vivo oxime protection and in vitro oxime reactivation were used as indicators of oxime efficacy against VX, sarin, VR and cyclosarin. Analysis of in vivo oxime protection was conducted with oxime protective ratios (PR) from guinea pigs receiving oxime and atropine therapy after sc administration of nerve agent. Analysis of in vitro reactivation was conducted with second-order rate contants (k(r2)) for oxime reactivation of agent-inhibited acetylcholinesterase (AChE) from guinea pig erythrocytes. In vivo oxime PR and in vitro k(r2) decreased as the volume of the alkylmethylphosphonate moiety of nerve agents increased from VX to cyclosarin. This effect was greater with 2-PAM and obidoxime (>14-fold decrease in PR) than with HI-6 and ICD585 (<3.7-fold decrease in PR). The decrease in oxime PR and k(r2) as the volume of the agent moiety conjugated to AChE increased was consistent with a steric hindrance mechanism. Linear regression of log (PR-1) against log (k(r2)[oxime dose]) produced two offset parallel regression lines that delineated a significant difference between the coupling of oxime reactivation and oxime protection for HI-6 and ICD585 compared to 2-PAM and obidoxime. HI-6 and ICD585 appeared to be 6.8-fold more effective than 2-PAM and obidoxime at coupling oxime reactivation to oxime protection, which suggested that the isonicotinamide group that is common to both of these oximes, but absent from 2-PAM and obidoxime, is important for oxime efficacy.

An in vitro comparative study on the reactivation of nerve agent-inhibited guinea pig and human acetylcholinesterases by oximes.

The reactivation of nerve agent-inhibited acetylcholinesterase (AChE) by oxime is the most important step in the treatment of nerve agent poisoning. Since the evaluation of nerve agent antidotes cannot be conducted in humans, results from animal experiments are extrapolated to humans. Guinea pig is one of the animal models that is frequently used for conducting nerve agent antidote evaluations. Several investigations have demonstrated that the efficacy of an oxime primarily depends on its ability to reactivate nerve agent-inhibited AChE. If the in vitro oxime reactivation of nerve agent-inhibited animal AChE is similar to that of human AChE, it is likely that the results of an in vivo animal study will reliably extrapolate to humans. Therefore, the goal of this study was to compare the reactivation of guinea pig and human AChEs inhibited by six different G and V type nerve agents. Reactivation kinetic studies with five mono- and bis-pyridinium oximes showed that oxime reactivation of nerve agent-inhibited human AChE in most cases was faster than guinea pig AChE. The most significant enhancement was observed in the reactivation of human AChE inhibited by nerve agents containing bulky side chains GF, GD, and VR, by H-series oximes HLo-7, HI-6, and ICD-585. In these cases, species-related differences observed between the two AChEs, based on the second-order reactivation rate constants, were 90- to over 400-fold. On the other hand, less than 3-fold differences were observed in the rates of aging of nerve agent-inhibited guinea pig and human AChEs. These results suggest that the remarkable species-related differences observed in the reactivation of nerve agent-inhibited guinea pig and human AChEs were not due to differences in the rates of aging. These results also suggest that guinea pig may not be an appropriate animal model for the in vivo evaluation of oxime therapy.

Crystal structures of human carboxylesterase 1 in covalent complexes with the chemical warfare agents soman and tabun.

The organophosphorus nerve agents sarin, soman, tabun, and VX exert their toxic effects by inhibiting the action of human acetylcholinesterase, a member of the serine hydrolase superfamily of enzymes. The current treatments for nerve agent exposure must be administered quickly to be effective, and they often do not eliminate long-term toxic side effects associated with organophosphate poisoning. Thus, there is significant need for effective prophylactic methods to protect at-risk personnel from nerve agent exposure, and protein-based approaches have emerged as promising candidates. We present the 2.7 A resolution crystal structures of the serine hydrolase human carboxylesterase 1 (hCE1), a broad-spectrum drug metabolism enzyme, in covalent acyl-enzyme intermediate complexes with the chemical weapons soman and tabun. The structures reveal that hCE1 binds stereoselectively to these nerve agents; for example, hCE1 appears to react preferentially with the 10(4)-fold more lethal PS stereoisomer of soman relative to the PR form. In addition, structural features of the hCE1 active site indicate that the enzyme may be resistant to dead-end organophosphate aging reactions that permanently inactivate other serine hydrolases. Taken together, these data provide important structural details toward the goal of engineering hCE1 into an organophosphate hydrolase and protein-based therapeutic for nerve agent exposure.

Effect of octanol:water partition coefficients of organophosphorus compounds on biodistribution and percutaneous toxicity.

Knowledge of partition coefficient (log P) data can play a critical role in understanding the pharmacokinetic and pharmacodistributive properties of toxic organophosphorus (OP) compounds. Using a recently published gas chromatographic method, the octanol:water log P values for the compounds tabun (GA), sarin (GB), cyclosarin (GF), and O-ethyl-S-(2-diisopropylaminoethyl) methylphosphonothiolate (VX) were determined to be 0.384 +/- 0.033, 0.299 +/- 0.016, 1.038 +/- 0.055, and 0.675 +/- 0.070, respectively. Based on these data, the log P value of the fluorophosphonate fragment, common to GB, soman (GD), and GF, was determined to be -2.256 +/- 0.273. The predictive value for absorption and distribution of the determined log P values was compared to measured values. The time to onset of local fasciculations (47.3, 29.0, 8.8, 8.5, and 6.3 min, respectively) in guinea pigs exposed percutaneously to equilethal doses of GA, VX, GF, GB, or GD was used as an indicator of dermal penetration. There was a good correlation (r = 0.95) between the measured log P value and the rate of onset of local fasciculations. Assuming a direct correspondence, equilibrium tissue:blood log P may be estimated from octanol:water log P. Comparison of the estimated and directly measured tissue:blood log P revealed a correlation of 0.8 for GD in liver, muscle, and adipose tissue. Our results demonstrate the use of log P data to both predict absorption and determine the distribution of OP compounds in tissues. This facilitates further estimates of in vivo OP effects from in vitro experiments.

Acetylcholinesterase inhibition: does it explain the toxicity of organophosphorus compounds?

The hypothesis that acetylcholinesterase (AChE) inhibition is the mechanism of toxicity of organophosphorus (OP) compounds was examined by mathematically modeling the in vivo lethal effects of OP compounds and determining the amount of variation in OP toxicity that is explained by AChE inhibition. Mortality dose-response curves for several OP compounds (i.e., VX, soman, cyclosarin, sarin, tabun, diisopropylfluorophosphate and paraoxon) exhibited steep probit slopes (> 9.6) in guinea pigs. Steep probit slopes were also observed when the mortality dose-response curves for soman were examined in mice, rats, rabbits and non-human primates. The consistently steep probit slopes of the dose-response curves for highly toxic OP compounds suggested that these compounds have a single specific mechanism of toxicity regardless of the OP compound or the species in which it was tested. Regression analysis indicated that 93% of the 3,280-fold variation in the median lethal doses (i.e., LD(50)) of OP compounds in rats was explained by the variation in their in vitro rate constants for inhibition of AChE. Conversely, 91% of the 23-fold variation in the ability of the oximes pralidoxime and obidoxime to protect against the toxicity of OP compounds in guinea pigs was explained by the variation in the in vitro ability of oximes to reactivate OP-inhibited AChE. The best explanation for this variety of observations was that the primary mechanism of in vivo toxicity for highly toxic OP compounds is the inhibition of AChE, and the residual unexplained variation in OP toxicity that might be explained by other mechanisms represents < 10% of the total variation in OP toxicity.

A physiologically based pharmacokinetic (PB/PK) model for multiple exposure routes of soman in multiple species.

A physiologically based pharmacokinetic (PB/PK) model has been developed in advanced computer simulation language (ACSL) to describe blood and tissue concentration-time profiles of the C(+/-)P(-) stereoisomers of soman after inhalation, subcutaneous and intravenous exposures at low (0.8-1.0 x LD(50)), medium (2-3 x LD(50)) and high (6 x LD(50)) levels of soman challenge in three species (rat, guinea pig, marmoset). Allometric formulae were used to compute the compartment volumes, blood flow rates, tidal volume and respiratory rate based upon total animal weight. Blood/tissue partition coefficients for soman, initial carboxylesterase and acetylcholinesterase levels and the rate constants for interactions between soman and these enzymes were species-dependent and were obtained from in vitro measurements reported in the literature. The model incorporated arterial and venous blood, lung, kidney, liver, richly perfused, poorly perfused and fat tissue compartments as well as subcutaneous and nasal exposure site compartments. First-order absorption from linearly filled soman deposits into metabolizing exposure site compartments was employed to model subcutaneous and inhalation exposures. The model was validated by comparing the predicted and observed values for C(+/-)P(-)-soman in arterial blood at various times following exposure and by regression analysis. Sensitivity analysis was used to determine the effects of perturbations in the model parameters on the time-course of arterial C(-)P(-)-soman concentrations for different exposure routes. In our evaluation of 28 datasets, predicted values were generally within 95% confidence limits of the observed values, and regression coefficients comparing predicted and observed data were greater than 0.85 for 95% of the intravenous and subcutaneous datasets and 25% of the inhalation datasets. We conclude that the model predicts the soman toxicokinetics for doses >or=1 x LD(50) for intravenous and subcutaneous exposures and inhalation exposures of 8 min or less sufficiently well to allow its use in the modeling of bioscavenger protection.

Protection against soman or VX poisoning by human butyrylcholinesterase in guinea pigs and cynomolgus monkeys.

Human butyrylcholinesterase (HuBuChE), purified from outdated human plasma, is being evaluated for efficacy against nerve agents in guinea pigs and cynomolgus monkeys. Previous studies in rodents and nonhuman primates demonstrated that pretreatment of animals with enzymes that can scavenge nerve agents could provide significant protection against behavioral and lethal effects of nerve agent intoxication. In preparation for evaluation of efficacy of HuBuChE prior to initiating an investigational new drug (IND) application, the pharmacokinetics of HuBuChE were evaluated in guinea pigs and in cynomolgus monkeys. HuBuChE was injected intramuscularly (i.m.) at two doses, and blood samples were taken to follow the time-course of HuBuChE in blood for up to 168 h after administration. In guinea pigs, the two doses of HuBuChE, 19.9 and 32.5 mg/kg, produced similar times of maximal blood concentration (T(max) of 26.0 and 26.8 h, respectively) and similar elimination half-times (t(1/2) of 64.6 and 75.5 h, respectively). Enzyme levels were still 10-fold over baseline at 72 h. Based on these data, guinea pigs were administered 150 mg/kg of enzyme i.m. and challenged at T(max). Soman or VX doses were approximately 1.5, 2.0 and 2.0 x LD50 administered subcutaneously (s.c.) in sequence at 90-120 min apart. None of the animals displayed signs of organophosphorus (OP) anticholinesterase intoxication at any of the challenge levels, and all survived for the 14-day duration of the experiment. Similar experiments were carried out with cynomolgus monkeys to determine the pharmacokinetics of HuBuChE and its efficacy against soman. The complete survival of nearly all animals tested to date, coupled with the maximal blood concentration and half-life elimination profile obtained for HuBuChE after i.m. injection, provides strong support for the continued development of HuBuChE as a product to protect against nerve agents.

Two possible orientations of the HI-6 molecule in the reactivation of organophosphate-inhibited acetylcholinesterase.

The inhibition of acetylcholinesterase (AChE) by organophosphorus compounds (OPs) causes acute toxicity or death of the intoxicated individual. One group of these compounds, the OP nerve agents, pose an increasing threat in the world due to their possible use in the battlefield or terrorist acts. Antidotes containing oxime compounds to reactivate the inhibited enzyme are highly valued for treatment against OP poisoning. One of these reactivators, HI-6, was shown to be significantly more effective in treating soman toxicity than other oximes, such as 2-PAM, TMB4, and obidoxime. However, HI-6 was less effective in reactivating AChE inhibited by the OP pesticide, paraoxon. In this study, the mechanism for HI-6-induced reactivation of OP-AChE conjugates was investigated using mouse mutant AChEs inhibited with different OPs including organophosphate paraoxon, and several methylphosphonates. Results indicate that the HI-6 molecule may assume two different orientations in the reactivation of AChE inhibited by organophosphate and Sp methylphosphonates. These conclusions were further corroborated by reactivation studies using an analog of HI-6 in which the bispyridinium moieties are linked by a methylene bridge rather than an ether oxygen.

A theoretical expression for the protection associated with stoichiometric and catalytic scavengers in a single compartment model of organophosphorus poisoning.

The ability of certain organophosphorus (OP) compounds to inhibit acetylcholinesterase (AChE) has made them useful for industrial (insecticides) and military (nerve agents) purposes. We have previously published a single compartment mathematical model of the interactions between OP nerve agents and the enzymes affected by these agents. That model, which could be used to predict the LD50 of seven nerve agents in rats, has been extended to include the protective actions of stoichiometric and catalytic OP-scavenger enzymes (delivered as pretreatments) so that protective ratios attributable to the scavengers may be predicted. Prediction of expected human protection from in vitro rate constant and initial enzyme level measurements is the ultimate goal for this work. The enhanced model predicts the LD50 from rate constants of the OP agent's binding reactions with AChE, carboxylesterase (CaE) and a stoichiometric scavenger (S); a first-order OP elimination rate (including a contribution due to a catalytic scavenger); and whole body estimates of AChE, CaE and S. The ratio of the scavenger-treated LD50 estimate to the scavenger-free LD50 estimate provided a theoretical expression describing the scavenger's contributions to the protective ratio. Published in vivo protective ratios for two stoichiometric scavengers (fetal bovine serum AChE and human utyrylcholinesterase) against challenge by several OP agents in mice were compared with ratios predicted by the model. A linear regression analysis of in vivo protective ratios in mice versus the ratios predicted by the model from the in vitro measurements resulted in an R(2) value of 0.902. The catalytic scavenger portion of the theory could not be validated due to a lack of published data. We conclude that the one-compartment model can be used to make reasonable estimates of the protective ratio attributable to stoichiometric scavengers, but can make no conclusions regarding the ability of the model to predict catalytic scavenger protection ratios.