Journal of Molecular Neuroscience 129 Volume 30, 2006

*Author to whom all correspondence and reprint requests should be addressed. E-mail: donald.maxwell@us.army.mil

Toxicodynamic Modeling of Highly Toxic Organophosphorus Compounds

D. M. Maxwell,* K. M. Brecht, F.-C. T. Chang, I. Koplovitz, T.-M. Shih, and R. E. Sweeney

United States Army Medical Research Institute of Chemical Defense, Research Division,Aberdeen Proving Ground, MD 21010-5400

IntroductionAlthough the in vitro effect of organophosphorus

(OP) compounds on acetylcholine-esterase (AChE)has been studied extensively, the hypothesis that OPinhibition of AChE is the primary mechanism ofacute in vivo OP toxicity has been controversial. Forexample, a recent review (Pope and Liu, 2004)suggested that OP compounds have direct toxiceffects on other enzymes, ACh receptors, and recep-tor/channel complexes that are independent ofAChE inhibition. The purpose of this report is toexamine the hypothesis that AChE inhibition is themechanism of acute toxicity of OP compounds bymathematically modeling the in vivo lethal effectsof highly toxic OP compounds and determining theamount of variation in OP toxicity that is explainedby AChE inhibition.

Materials and MethodsMedian lethal doses (LD50) were determined from

the 24-h mortality of animals receiving OP com-pounds by subcutaneous (sc) administration. LD50values were calculated by probit analysis (Finney,1971) of mortality fractions with at least 6 doses ofOP compounds and 10 animals per dose.

The efficacy of oximes was determined from the24-h mortality of guinea pigs receiving variable dosesof oxime with a fixed dose (50 µmol/kg) of atropineby intramuscular injection 1 min after sc adminis-

Journal of Molecular NeuroscienceCopyright © 2006 Humana Press Inc.All rights of any nature whatsoever are reserved.ISSN0895-8696/06/30:129–132/$30.00JMN (Online)ISSN 1559-1166DOI 10.1385/JMN/30:1-2:129

ORIGINAL ARTICLE

tration of the OP compound. An oxime’s efficacywas expressed as a protective ratio (PR), which isthe ratio of OP LD50 in oxime/atropine-treated ani-mals divided by OP LD50 in untreated animals.

In vitro rate constants for OP inhibition of AChE(kAChE) and carboxylesterase (kCaE) were expressed asbimolecular rate constants, and OP hydrolysis rates(kHydro) were expressed as first-order rate constants.Rate constants for kAChE, kCaE, and kHydro in rats weretaken from Sweeney and Maxwell (1999). In vitrobimolecular rate constants for oxime reactivation (kR)of OP-inhibited AChE were taken from Worek andcoworkers (2002).

Mathematical models of relationships between invitro and in vivo variables were developed by regres-sion analysis and graphed with SigmaPlot 2000. Thefraction of total variation explained by each mathe-matical model was determined from its regressioncorrelation coefficient (r), where r2= (explained vari-ation)/(total variation).

Results and DiscussionIn as much as the OP inhibition of AChE is an

extremely rapid and specific reaction in vitro (DeJongand Benschop, 1988), it would be expected that thein vivo dose response of an AChE inhibitor wouldreflect this specificity across a variety of animalmodels. The mortality dose-response curves forsoman, a highly toxic OP compound, in a variety ofmammalian species are shown in Fig. 1. Although

57_Maxwell 10/11/06 4:28 PM Page 129

130 Maxwell et al.

Journal of Molecular Neuroscience Volume 30, 2006

the LD50 values varied 14-fold between species, theprobit slopes of the dose-response curves were sim-ilar, monophasic, and very steep (i.e., slope 15),which suggests that soman has a single specificmechanism of acute toxicity (Faustman and Omenn,2001).

The mortality dose-response curves for a varietyof other highly toxic OP compounds (i.e., VX,cyclosarin, and sarin) that had sc LD50 values in guineapigs that varied from 9 g/kg for VX to 42 g/kg forsarin also exhibited similar steep monophasic probitslopes ranging from 14.8 to 15.3 (data not shown).This observation suggests that highly toxic OPcompounds other than soman also have a singlespecific mechanism of acute toxicity.

Evaluation of AChE inhibition as the in vivo mech-anism of OP toxicity was initially performed usingdata from Sweeney and Maxwell (1999) to examinethe relationship between OP LD50 values in rats andthe bimolecular rate constants for OP inhibition ofAChE (kAChE). Regression analysis showed that LD50

(kAChE), with r2 = 0.93, which indicated that 93% ofthe variation in the acute toxicity of OP compoundswas explained by the variation in their ability toinhibit AChE.

A more comprehensive mathematical model ofOPtoxicity was subsequently derived(Sweeney andMaxwell, 1999) that included kAChE and the rates ofreaction of OP compounds with kCaE and OP kHydro,the most important OP-metabolizing enzymes. Agraphic depiction of this mathematical model, whichexplained 99% of the variation of OP LD50 values inrats, is shown in Fig. 2. This toxicity dose-responsesurface for OP compounds describes the relation-ship between in vivo LD50 and in vitro values for kCaEand kHydro that have been normalized by division

with either kAChE, the initial whole body AChE level(AChEInit), or their product. This complex dose-response surface is composed of an AChE floor forOP compounds that reacts only with AChE, a hydro-lysis plane for OP compounds that reacts primarilywith AChE and a first-order hydrolytic enzyme, anda CaE shelf for OP compounds that reacts primarilywith AChE and a stoichiometric metabolic enzyme,such as CaE. The elevation of the CaE shelf abovethe AChE floor was determined by the ratio of theinitial whole body levels of CaE (CaEInit) and AChE.This analysis illustrates the importance of includingOP metabolism for a complete understanding of thein vivo mechanism of OP toxicity.

If inhibition of AChE is the mechanism of toxic-ity of OP compounds, then the ability to reactivateOP-inhibited AChE should correlate with protectionagainst the toxicity of OP compounds. A regressionanalysis of the relationship between in vivo PRs oftoxogonin and pralidoxime at different oxime dosesand their in vitro bimolecular rate constants for oximereactivation (kR) of OP-inhibited AChE is shown inFig. 3. Efficacy of oximes against the toxicity of VX,sarin, and cyclosarin in guinea pigs was expressedas PR-1, because a PR of 1 denotes an absence ofoxime protection, and kR was multiplied by theadministered dose of oxime ([Oxime]) to normalizethe effect of different oxime doses. Regressionanalysis showed that PR-1 (kR[Oxime])0.56 withr2= 0.91, which indicated that the ability to reactivateOP-inhibited AChE explained 91% of the variationin the efficacy of oximes against highly toxic OPcompounds.

Fig. 1. Comparison of mortality dose-response curves forsoman in different species. Data for marmoset and rhesusmonkey were taken from Dirnhuber and coworkers (1979).

Fig. 2. A mathematical model of dose-response surfacefor OP toxicity in rats.

57_Maxwell 10/11/06 4:28 PM Page 130

Toxicodynamic Modeling of OP Compounds 131

Journal of Molecular Neuroscience Volume 30, 2006

ConclusionsProbit analysis of mortality dose-response curves

for a number of OPcompounds in a variety of speciessuggested that highly toxic OP compounds have asingle specific mechanism of toxicity. Regressionanalysis indicated that 93% of the variation in thetoxicity of OPcompounds was explained by the vari-ation in their in vitro rate constants for inhibition ofAChE, and the residual unexplained variation in OPtoxicity was explained by a more comprehensivemathematical model that included the reaction ofOP compounds with OP-metabolizing enzymes aswell as AChE. Conversely, a high correlation was

observed between the ability of oximes to reactivateOP-inhibited AChE and their in vivo protectionagainst OP compounds. The conclusion that consis-tently provides the best explanation for this varietyof observations is that the mechanism of toxicity ofhighly toxic OPcompounds is the inhibition of AChE.

ReferencesDawson R. M. (1994) Review of oximes available for treat-

ment of nerve agent poisoning. J. Appl. Toxicol. 14,317–331.

DeJong L. P. A. and Benschop H. P. (1988) Stereoselectivityof Pesticides: Biological and Chemical Problems, Ariens, E. J., Van Rensen, J. J. S., and Welling, W., eds., Elsevier,Amsterdam, pp. 109–149.

Dirnhuber P., French M. C., Green D. M., Leadbeater L.,and Stratton J. A. (1979) The protection of primatesagainst soman poisoning by pretreatment with pyri-dostigmine. J. Pharm. Pharmacol. 31, 295–299.

Faustman E. M. and Omenn G. S. (2001) Casarett and Doull’sToxicology: The Basic Science of Poisons, 6th ed., Klassen,C. D., ed., McGraw-Hill, New York, pp. 83–104.

Finney D. J. (1971) Probit Analysis, 3rd ed., Cambridge Uni-versity Press, Cambridge, UK, pp. 50–124.

Pope C. and Liu J. (2004) Pharmacological Perspectives ofToxic Chemicals and Their Antidotes, Flora, S. J. S.,Romano, J. A., Baskin, S. I., and Sekhar, K., eds., NarosaPublishing, New Delhi, pp. 215–239.

Sweeney R. E. and Maxwell D. M. (1999) A theoreticalmodel of the competition between hydrolase and car-boxylesterase in protection against organophosphoruspoisoning. Math. Biosci. 160, 175–190.

Worek F., Reiter G., Eyer P., and Szinicz L. (2002) Reacti-vation kinetics of acetylcholinesterase from differentspecies inhibited by highly toxic organophosphates.Arch. Toxicol. 76, 523–529.

Fig. 3. Regression analysis of in vitro oxime reactivationrate constants and in vivo oxime protection in guinea pigs.Data for some PR values were taken from Dawson (1994).

57_Maxwell 10/11/06 4:28 PM Page 131