Evaluation of oxime efficacy in nerve agent poisoning: Development of a kinetic-based dynamic model

Toxicology and Applied Pharmacology 209 (2005) 193 – 202 www.elsevier.com/locate/ytaap

Evaluation of oxime efficacy in nerve agent poisoning: Development of a kinetic-based dynamic model Franz Woreka,*, Ladislaus Szinicza, Peter Eyerb, Horst Thiermanna b

a Bundeswehr Institute of Pharmacology and Toxicology, Neuherbergstrasse 11, 80937 Munich, Germany Walther-Straub-Institute of Pharmacology and Toxicology, Ludwig-Maximilians-University, Goethestrasse 33, 80336 Munich, Germany

Received 27 January 2005; accepted 9 April 2005 Available online 17 May 2005

Abstract The widespread use of organophosphorus compounds (OP) as pesticides and the repeated misuse of highly toxic OP as chemical warfare agents (nerve agents) emphasize the necessity for the development of effective medical countermeasures. Standard treatment with atropine and the established acetylcholinesterase (AChE) reactivators, obidoxime and pralidoxime, is considered to be ineffective with certain nerve agents due to low oxime effectiveness. From obvious ethical reasons only animal experiments can be used to evaluate new oximes as nerve agent antidotes. However, the extrapolation of data from animal to humans is hampered by marked species differences. Since reactivation of OP-inhibited AChE is considered to be the main mechanism of action of oximes, human erythrocyte AChE can be exploited to test the efficacy of new oximes. By combining enzyme kinetics (inhibition, reactivation, aging) with OP toxicokinetics and oxime pharmacokinetics a dynamic in vitro model was developed which allows the calculation of AChE activities at different scenarios. This model was validated with data from pesticide-poisoned patients and simulations were performed for intravenous and percutaneous nerve agent exposure and intramuscular oxime treatment using published data. The model presented may serve as a tool for defining effective oxime concentrations and for optimizing oxime treatment. In addition, this model can be useful for the development of meaningful therapeutic animal models. D 2005 Elsevier Inc. All rights reserved. Keywords: Acetylcholinesterase; Organophosphates; Oximes; Human; Model

Introduction Various highly toxic organophosphorus compounds (OP) were developed in the past for use as chemical warfare agents (Fnerve agents_; Sidell, 1997). Large stocks of nerve agents are still available and present a persistent threat to the population. The repeated use of nerve agents during military conflicts (MacIlwain, 1993) and terrorist attacks (Nagao et al., 1997) underlines the necessity to develop an effective medical treatment regimen. The toxic effects of nerve agents are mainly due to a progressive inhibition of cholinesterases by phosphylation (denotes both phosphorylation and phos-

* Corresponding author. Fax +49 89 3168 2333. E-mail address: [email protected] (F. Worek). 0041-008X/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.taap.2005.04.006

phonylation) of their active center serine leading to inactive enzyme (MacPhee-Quigley et al., 1985; Taylor et al., 1995). The inability of inhibited AChE to hydrolyze acetylcholine results in accumulation of the transmitter and subsequently in generalized over-stimulation of cholinergic receptors followed by breakdown of neuromuscular and ganglionic transmission. Standard treatment of nerve agent poisoning includes a muscarine antagonist, e.g., atropine, and an AChE reactivator (oxime). Presently, the oximes obidoxime and pralidoxime are approved as antidotes against OP poisoning but are considered to be rather ineffective against certain nerve agents, e.g., soman (pinacolylmethylphosphonofluoridate) and cyclosarin (cyclohexylmethylphosphonofluoridate) (Lundy et al., 1992; Sidell, 1992; Worek et al., 2004). To overcome these gaps numerous new oximes were

194

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

synthesized and tested for their antidotal efficacy in the past decades (Bismuth et al., 1992). The efficacy of novel nerve agent antidotes cannot be investigated in humans for ethical reasons. Therefore, testing is primarily performed with animals (Dawson, 1994). However, substantial species differences in the toxicokinetics of inhibitors (Benschop and de Jong, 2001), pharmacokinetics and dosing of antidotes (Baggot, 1994; Clement et al., 1995; Dawson, 1994), and reactivating potency of oximes (Clement and Erhardt, 1994; Worek et al., 2002) hamper the extrapolation of animal data to humans. Reactivation of inhibited AChE by removal of the phosphyl moiety from the AChE active site serine is considered to be the primary mechanism of oximes (Wong et al., 2000) and recent clinical data from OP pesticide poisoned patients provide strong evidence for the validity of this assumption (Eyer, 2003). The evaluation of oxime efficacy requires the consideration of further post-inhibitory reactions of AChE, spontaneous dealkylation through alkyl-oxygen bond scission (‘‘aging’’) (Shafferman et al., 1996), and spontaneous dephosphylation (‘‘spontaneous reactivation’’) of the phosphyl-AChE-complex (Aldridge and Reiner, 1972). Recently, the kinetic constants of inhibition, reactivation, and aging (Fig. 1) were determined for different nerve agents, pesticides, and oximes with human erythrocyte AChE in order to provide a kinetic basis for the proper assessment of oxime efficacy (Worek et al., 2004). Clinical findings indicate that kinetic data generated with human erythrocyte AChE in vitro (Worek et al., 1997, 1999) correlate well with the in vivo cholinesterase status in OP pesticide poisoned patients and could be used to optimize oxime treatment (Eyer, 2003; Thiermann et al., 1999). This

was achieved mainly by calculating effective oxime concentrations applying different theoretical models and using half-time of reactivation of inhibited AChE at steady-state AChE activity in the presence of different OP and oxime concentrations (Thiermann et al., 1999). However, these static models do not provide information about timedependent changes of AChE activity. Moreover, these models do not allow dynamic simulation of AChE activities in the presence of concomitant changes of inhibitor and oxime concentrations. To get more insight into these complexities we developed a kinetic model which enables the simulation of dynamic changes of AChE activity by including enzyme kinetic (inhibition, reactivation, aging), pharmacokinetic, and toxicokinetic parameters. This dynamic model was used for the evaluation of reactivating efficacy of oximes with nerve agent-inhibited human AChE at different scenarios.

Materials and methods Model development. The development of a dynamic model for the simulation of AChE activities at changing inhibitor and oxime concentrations requires the consideration of different reactions between AChE, inhibitor, and oxime (Fig. 1), which are characterized by respective kinetic constants and may be expressed as differential equations (Aldridge and Reiner, 1972; Green and Smith, 1958; Skrinjaric-Spoljar et al., 1973; Su et al., 1986; Worek et al., 2004):


‘‘Inhibition__

dE d½EP ¼ ¼ ki 4 ½E 4 ½OP dt dt

ð1Þ

(This model ignores that inhibition rate may approach an asymptotic value at [OP] > K D. Such a situation, however, is not expected in reality. Hence, the model may be simplified to the second-order reaction depicted above.) ‘‘Reactivation__

hereby Fig. 1. Interactions of AChE, organophosphorus compounds, and oximes. In this scheme the respective concentrations are denoted for [OP] the organophosphorus compound, [EP] the phosphylated AChE, [EPOX] the Michaelis-type phosphyl-AChE-oxime complex, [OX] the reactivator, [E] the active enzyme, [POX] the phosphylated oxime, and [EA] the dealkylated (‘‘aged’’) AChE. The different reactions can be described by kinetic constants for inhibition (k i), aging (k a), spontaneous (k s), and oxime-induced reactivation. K D is equal to the ratio [EP]  [OX] / [EPOX] and describes the dissociation constant which is inversely proportional to the affinity of the oxime to [EP], and k r denotes the rate constant for the displacement of the phosphyl residue from [EPOX], indicating the reactivity of the oxime.

kobs ¼

d½ E  ¼ kobs 4 ½EP þ EPOX dt kr 4 ½OX KD þ ½OX

‘‘Spontaneous reactivation__ ‘‘Aging__

d½EA ¼ ka 4 ½EP dt

ð2Þ

ð3Þ d½E ¼ ks 4 ½EP dt

ð4Þ ð5Þ

(This simplified model assumes that aging and spontaneous reactivation occurs with EP and not with EPOx, which can be ignored at [Ox] N K D anyhow.)

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

Table 2 Dissociation (K D) and reaction rate constants (k r) for the oxime-induced reactivation of OP-inhibited AChE

Combining Eqs. (1) – (5) gives d½E ¼
ki 4 ½OP4 ½E þ ks 4 ½EP þ kobs 4 ½EP þ EPOX dt

OP

d½EP ¼
ki 4 ½OP4 ½E
ks 4 ½EP
kobs 4 ½EP þEPOX dt
ka 4 ½EP

ð6Þ

which enables the simulation of changes of AChE activities upon simultaneously changing OP and oxime concentrations. The two differential equations were numerically solved with Maple 9.0 (Maplesoft, Waterloo, Canada). The various equations used for the calculation of AChE activities after exposure by sarin (isopropylmethylphosphonofluoridate), cyclosarin, and VX (O-ethyl S-[2-(diisopropylamino)ethyl)methylphosphonothioate) are presented in Appendix A. Model validation. The applicability of the kinetic constants obtained in vitro to calculate the expected AChE activity was compared with in vivo data from a parathionpoisoned patient (Eyer et al., 2003). A 45-year-old man (patient BR 3/97) having ingested intentionally 50 g parathion, presented to the emergency physician unconscious with cholinergic signs and seizures. He received atropine and obidoxime in the toxicologic clinic and blood samples were taken repeatedly to determine paraoxon and obidoxime concentration and erythrocyte AChE activity. With the kinetic constants of inhibition and reactivation (Tables 1 and 2) and with the measured paraoxon and oxime concentrations the expected AChE activity was calculated and compared with the determined values. Eq. (7) was applied as derived previously (Eyer, 2003): ½E ¼ ½EP þ EPOX

k r ki 4 ½OP4

1 þ KD ½OX



ð7Þ

Calculation of AChE activities at different scenarios. The simulation of AChE activities upon nerve agent exposure was performed for different scenarios, i.e., intravenous exposure by sarin and cyclosarin, compounds having a rather short biological half-life (Spruit et al., 2000), and intravenous and percutaneous exposure to VX as an example of a persistent OP (van der Schans et al., 2003). Table 1 Rate constants for the inhibition of AChE by OP (k i) and for the spontaneous dealkylation (k a) and reactivation of OP-inhibited AChE (k s) OP

k i (M
1min
1)

VXa Sarina Cyclosarina Paraoxon-ethylb

1.2 2.7 4.9 2.2

   

108 107 108 106

k a (min
1)

k s (min
1)

0.00032 0.0038 0.0016 0.00036

0.00035 0 0 0.00036

Data for human erythrocyte AChE, pH 7.4, 37 -C. a From Worek et al. (2004). b From Mast (1997).

195

VXa Sarina Cyclosarina Paraoxon-ethylb

Obidoxime

Pralidoxime

HI 6

KD (AM)

kr (min
1)

KD (AM)

kr (min
1)

KD (AM)

kr (min
1)

27.4 31.3 945.6 32.2

0.893 0.937 0.395 0.81

28.1 27.6 3159 187.3

0.215 0.25 0.182 0.17

11.5 50.1 47.2 548.4

0.242 0.677 1.3 0.2

Data for human erythrocyte AChE, pH 7.4, 37 -C. a From Worek et al. (2002). b From Worek et al. (2004).

Hereby, it was assumed that nerve agent exposure and intramuscular oxime treatment occurred simultaneously. Antidotal oxime dose should resemble 1– 3 autoinjector equivalents, i.e., 250 –750 mg obidoxime chloride (MW 359.2), 600– 1800 mg pralidoxime chloride (MW 172.6), and 500 – 1500 mg 1-[[[4-(aminocarbonyl)pyridinio]methoxy]methyl]-2-[(hydroxyimino)methyl]pyridinium dichloride monohydrate (HI 6; MW 377.2) (Clement et al., 1995; Sidell, 1997; Szinicz and Kullmann, 1990). Plasma oxime concentrations were calculated by using human pharmacokinetic data. For obvious reasons no human toxicokinetic data of nerve agents are available. Therefore, respective data from guinea pigs were used for calculating plasma nerve agent concentrations (Spruit et al., 2000; van der Schans et al., 2003). The data of intravenous sarin exposure (0.8  LD50 sarin) were extrapolated to 4  LD50 sarin, assuming a linear dose– concentration relationship. Sarin data were used for the simulation of intravenous cyclosarin exposure since no toxicokinetic data of cyclosarin are available. Toxicokinetic parameters of intravenous VX exposure of guinea pigs (1 and 2  LD50 VX; van der Schans et al., 2003) were linearly extrapolated to 3 and 5  LD50. A one-compartment open model with first-order absorption and elimination was used for calculation of VX concentrations after percutaneous exposure. Constants for inhibition, reactivation, and aging (Tables 1 and 2) were taken from the literature (Worek et al., 2004).

Results The validation of the model by using in vitro enzyme kinetic constants (Tables 1 and 2) and distinct in vivo inhibitor and oxime concentrations gave a good agreement between the patient’s and the calculated AChE activities in the case of parathion poisoning (Fig. 2). In the dynamic model, the calculated time course of AChE activity following intravenous sarin exposure and intramuscular injection of 250 mg obidoxime, 600 mg pralidoxime, or 500 mg HI 6 (Fig. 3) indicates that these oximes should be able to restore the immediately blocked (not shown) enzyme activity rapidly and sufficiently even at higher sarin doses (Fig. 4). In contrast, obidoxime and

196

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

tion, results in a progressive decrease of AChE activity (Fig. 7). Again, differences between the oximes, based on their specific reactivities, can be observed and increasing the oxime dose has only little effect.

Discussion The validity of the dynamic model was supported by the reasonable agreement between the model predictions and

Fig. 2. Model validation. Paraoxon and obidoxime concentrations were determined in plasma samples of a parathion-poisoned patient (top, data from Eyer, 2003). The pharmaco- and toxicokinetic parameters and the enzyme kinetic data (cf. Tables 1 and 2) were used for the calculation of AChE activities (bottom). Enzyme activities, determined in diluted whole blood of an OP-poisoned patient, were included for comparison. Time p.i. gives the time after ingestion of the OP in hours.

pralidoxime would be rather ineffective after intravenous cyclosarin exposure (Fig. 5), due to the low reactivity and affinity of both oximes (Table 2). In addition, the substantially higher inhibitory potency of cyclosarin, 18-fold higher compared to sarin (Table 1), thwarts the oxime effectiveness. On the other hand, HI 6 should be an effective reactivator of cyclosarin-inhibited AChE. The simulation of AChE activities after intravenous VX exposure and oxime treatment shows a rather slow increase at 1  LD50 VX, the maximum reactivation being after approximately 2 and 3 h with obidoxime, pralidoxime, and HI 6, respectively (Fig. 6). Due to the long persistence of VX (Fig. 6G, note the different time scale), resulting in long-lasting toxicologically relevant concentrations at higher VX doses, and the fast elimination of the oximes (Fig. 3), the enzyme activity decreases again. Despite of a 5fold higher molar dose of pralidoxime, compared to obidoxime, this oxime was substantially less effective. The difference between the oximes is mainly based on different reactivities (k r, Table 2). Due to the comparably high potency of the oximes (K D, Table 2) near-maximal effects are already obtained and increased oxime doses, i.e., 3 autoinjector equivalents (Figs. 6B, D, F), have only little additional effect on AChE activity. According to published data (van der Schans et al., 2003) percutaneous VX exposure is characterized by a protracted increase of VX concentration followed by a slow decrease (Fig. 7H, insert). Application of these data to the dynamic model, assuming simultaneous VX and oxime administra-

Fig. 3. Calculated plasma concentrations of obidoxime (A), pralidoxime (B), and HI 6 (C) after intramuscular injection of 1 to 3 autoinjector equivalents, i.e., 250 mg obidoxime, 600 mg pralidoxime, and 500 mg HI 6. The following values for the different parameters were used: Parameter

Obidoxime

Pralidoxime

HI 6

Dose (g) Dose (mmol) V Dss (ml/kg) k abs (min
1) k el (min
1)

0.25/0.5/0.75 0.70/1.39/2.09 173a 0.0578b 0.0083c

0.6/1.2/1.8 3.48/6.95/10.43 815a 0.103d 0.0088a

0.5/1.0/1.5 1.33/2.65/3.98 240b 0.0578b 0.0113b

V Dss was normalized for 70 kg body weight. Oxime plasma concentrations were calculated with the Bateman equation (Appendix A). The values for the different parameters were taken from: aSidell et al. (1972), bClement et al. (1995), cSidell and Groff (1970), and dJovanovic (1989). No k abs value is known for obidoxime; therefore, the respective value of HI 6 was used, assuming comparable absorption kinetics of the structurally similar bispyridinium compounds.

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

197

nM (74 h) to 43 nM (96 h) with virtually no change in obidoxime concentration during this period results in substantial decrease of measured and calculated AChE activities and indicates a sufficient sensitivity of the model. Admittedly, the model reactions as outlined in Fig. 1 ignore, for instance, the involvement of phosphyloximes (POX) that are inevitably formed during oxime-induced reactivation. Human paraoxonase (PON 1) with the prevailing Q192 phenotype (94%; Brophy et al., 2001) is able to hydrolyze effectively O,O-diethylphosphoryl obidoxime (Kiderlen et al., 2000). This enzyme, however, is apparently unable to inactivate phosphonyl oxime derivatives (Worek et al., 2000). Hence, POX accumulation may occur in the many reactivation/re-inhibition cycles, particularly with the more stable compounds deriving from 4-pyridinium oximes (Leader et al., 1999) and may alter the efficacy of the 4pyridinium oximes, which is not considered in the presented model. Due to the extreme lability of POX formed during reactivation of inhibited AChE by HI 6 no accumulation of this reaction product could be observed so

Fig. 4. Calculated human AChE activities after intravenous sarin and simultaneous intramuscular oxime injection. The changes in AChE activity were calculated by using toxicokinetic data of (
)-sarin (Spruit et al., 2000) and pharmacokinetic data of the oximes (cf. Fig. 3) for 250 mg obidoxime (A), 600 mg pralidoxime (B), and 500 mg HI 6 (C) and exposure to 0.8  LD50 and 4  LD50 of (
)-sarin (assuming a linear dose concentration relationship of sarin). The (
)-sarin concentrations (D) were calculated with the equation [sarin] = A  e
at + B  e
bt (Spruit et al., 2000) using the following data:

A (nM) B (nM) a (min
1) b (min
1)

0.8  LD50 256.24 0.642

4  LD50 1281 3.21 4.6 0.012

the patient’s AChE activities. The example shown in Fig. 2 demonstrates that changes of OP and oxime concentrations were immediately reflected by alterations in AChE activity that correspond closely with the measured values. The increase of plasma paraoxon concentration from 32

Fig. 5. Calculated human AChE activities after intravenous cyclosarin and simultaneous intramuscular oxime injection. The changes in AChE activity were calculated by using toxicokinetic data of (
)-sarin (Spruit et al., 2000) and pharmacokinetic data of the oximes (cf. Fig. 3) for 1, 2, and 3 autoinjector equivalents of obidoxime (A), pralidoxime (B), and HI 6 (C) and exposure to 0.8  LD50 cyclosarin.

198

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

Fig. 6. Calculated human AChE activities after intravenous (T)-VX and simultaneous intramuscular oxime injection. The changes in AChE activity were calculated by using toxicokinetic data of VX (van der Schans et al., 2003) and pharmacokinetic data of the oximes (cf. Fig. 3) for 250 mg (A) and 750 mg obidoxime (B), 600 mg (C) and 1800 mg pralidoxime (D), and 500 mg (E) and 1500 mg HI 6 (F) and exposure to 1, 3, and 5  LD50 VX. The (T)-VX concentrations (G, log scale) were calculated with the three-exponential equation [VX] = A  e
at + B  e
bt + C  e
ct (van der Schans et al., 2003). A linear relationship between VX dose and A, B, C was anticipated and a linear regression line forced through the origin. For the elimination rate constants, a, b, and c, the mean values as given by van der Schans et al. (2003) were used throughout. The following data were applied: Data from van der Schans et al. A (nM) B (nM) C (nM) a (min
1) b (min
1) c (min
1)

1  LD50 179.5 32.5 1.46 0.71 0.045 0.0071

Data used for simulation 2  LD50 288.0 63.6 1.80 0.67 0.033 0.0042

1  LD50 151.1 32.9 1.01

3  LD50 453.3 98.8 3.03 0.69 0.039 0.0057

5  LD50 755.5 164.7 5.06

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

199

Fig. 7. Calculated human AChE activities after percutaneous VX and simultaneous intramuscular oxime injection. The changes in AChE activity were calculated by using plasma (T)-VX concentrations as measured in guinea pigs and human pharmacokinetic data of the oximes (cf. Fig. 3) for 250 mg (A) and 750 mg obidoxime (B), 600 mg (C) and 1800 mg pralidoxime (D), and 500 mg (E) and 1500 mg HI 6 (F) and exposure with 1, 3, and 5  LD50 VX. The AChE activity in the absence of oximes is shown too (G). The VX concentrations (H) were calculated with the equation [VX] = [VX0]  e
el*t
[VX0]  e
abs*t using the published plasma VX concentrations after percutaneous exposure (van der Schans et al., 2003) and assuming a linear dose concentration relationship of VX. The following data were used:

VX0 (nM) abs (min
1) el (min
1)

1  LD50 4.784

far (Luo et al., 1999). The reactivation rate constants, k r and K D, used for the calculations with the dynamic model were determined under conditions of minimized POX effects (Worek et al., 2004). Therefore, the data presented may over-estimate the effect of obidoxime in vivo, since a

3  LD50 14.35 0.005005 0.003375

5  LD50 23.92

substantial POX formation may occur at physiological AChE concentrations leading to an impaired net reactivation of the enzyme. The fundamental prerequisite for the application of the dynamic model is the availability of enzyme kinetic as well

200

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

as toxico- and pharmacokinetic data. Recently, inhibition, reactivation, and aging kinetics of different nerve agents and analogs were investigated with human erythrocyte AChE (Worek et al., 2004) and can be used for the simulation of AChE activities at different scenarios. Human pharmacokinetics after intramuscular injection of obidoxime, pralidoxime, and the newer oxime HI 6 were determined by different investigators (Clement et al., 1995; Jovanovic, 1989; Sidell and Groff, 1970, 1971, 1972). Standard autoinjector doses, i.e., 250 mg obidoxime, 600 mg pralidoxime, and 500 mg HI 6, as well as 2- and 3-fold higher doses were used for the simulations. Due to substantial differences of the apparent volume of distribution (V Dss) between obidoxime and pralidoxime, the peak plasma oxime concentrations were similar despite of a 5-fold higher molar dose of pralidoxime. The major limitation of the application of the dynamic model to nerve agents is the lack of human toxicokinetic data and the availability of only few animal toxicokinetic data, mainly from rats and guinea pigs (Benschop and de Jong, 2001). In order to exclude additional variability due to species differences we confined to a single species and selected sarin and VX data from guinea pigs throughout for the model calculations (Spruit et al., 2000; van der Schans et al., 2003). Selecting guinea pigs appeared most appropriate since this species is notorious for its low carboxylesterase activity, like humans, thus preventing from scavenging of the highly potent nerve agents (Maxwell and Brecht, 1991). Inhalation and percutaneous absorption are considered as the main routes of entry of nerve agents (Sidell, 1992). Hereby, intravenous injection may serve as a model for inhalation exposure (Aas et al., 1985). The rapid onset of signs of poisoning requires preclinical antidotal treatment by intramuscular injection of atropine and oxime (Sidell, 1992). On basis of these premises, simulations were performed for compounds with a short biological half-life, sarin and cyclosarin, and for the persistent agent VX. According to the model calculations, the three oximes should be sufficiently effective after sarin exposure (Fig. 4), which concurs with in vivo animal data (Dawson, 1994; Schoene and Oldiges, 1973). The low potency of obidoxime and especially pralidoxime towards cyclosarin-inhibited human AChE is the major cause for the low efficacy of these oximes in the model calculation with a low cyclosarin dose (Fig. 5). Even administration of 3 autoinjector equivalents did not enable pralidoxime (1800 mg) to be significantly effective and it would require 30 autoinjector equivalents (18,000 mg) to increase AChE activity by 25% after 100 min. Due to the high reactivity and affinity of HI 6 to cyclosarininhibited human AChE, this oxime should be an effective reactivator. It was tempting to look at literature data on oxime effectiveness in cyclosarin-poisoned experimental animals. Guinea pigs intoxicated by 6  LD50 cyclosarin were

treated with pralidoxime (30 mg/kg), obidoxime (20 mg/ kg), or HI 6 (110 mg/kg). Survival rates were 0/5, 1/5, and 5/5, respectively (Lundy et al., 1992). Equimolar administration of 100 Amol/kg pralidoxime, obidoxime, or HI 6 i.m. to male rats poisoned by 1  LD50 cyclosarin resulted in reactivation of AChE in diaphragm of 3%, 16%, and 99%, respectively (Kassa and Cabal, 1999). Male rhesus monkeys pretreated with pyridostigmine and challenged with 5  LD50 cyclosarin were treated with atropine and pralidoxime (26 mg/kg) or HI 6 (38 mg/kg). 48 h after poisoning red blood AChE increased to 40% and 95%, respectively (Koplovitz et al., 1992). These data indicate that the model prediction correlates well with in vivo AChE and survival. The simulation of VX exposure indicates the necessity of repetitive oxime administration in order to maintain a sufficient level of AChE activity. The long persistence of VX and the rather rapid elimination of the oximes results in a final re-inhibition of AChE after higher intravenous VX doses (Fig. 6). This situation is even more pronounced after percutaneous VX exposure (Fig. 7). In guinea pigs a slow but long-lasting increase of plasma VX concentration was recorded (van der Schans et al., 2003), and in domestic pigs, considered to be a more relevant species for modeling human percutaneous absorption (Simon and Maibach, 2000), erythrocyte AChE activity decreased slowly after percutaneous challenge with 2  LD50 VX (Chilcott et al., 2003). These findings and the fact that the major human endogenous OP hydrolyzing enzyme, paraoxonase, fails to metabolize VX (Masson et al., 1998) suggests that a similar toxicokinetic pattern can be expected in humans after percutaneous VX exposure. The model calculations indicate that repetitive oxime injections or an oxime infusion would be more appropriate in this scenario. In conclusion, the proposed dynamic model may serve as a valuable tool for the evaluation of oximes, for defining effective oxime concentrations and for optimizing oxime treatment protocols in human OP pesticide and nerve agent poisoning. Depending on the availability of kinetic data this model may also be used for the development of refined animal models which are indispensable for the licensing of new oximes. Notwithstanding, the lack of published human toxicokinetic data of nerve agents and the possible interference of phosphyloximes, particularly in the case of obidoxime, lets us behind with some uncertainty.

Acknowledgments The authors are grateful to Prof. Dr. B. Fichtl, WaltherStraub-Institute, for help with the mathematics and valuable suggestions. Conflict of interest statements: The authors have no conflicts of interest that are directly relevant to the content of this paper.

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

Appendix A Eq. (6) were used for the simulation of dynamic changes of AChE activities by considering enzyme kinetics, toxicokinetics of inhibitor, and pharmacokinetics of oxime: d½E ¼
ki 4 ½OP4 ½E þ ks 4 ½EP þ kobs 4 ½EP þ EPOX dt

201

Formula for Maple 9.0: DðeÞðt Þ ¼
ki4ð A4expð
a4t Þ þ B4expð
b4t Þ þ C4expð
c4t ÞÞ4eðt Þ þ ks4pðt Þþ pðt Þ4kr 4 Dose=VDss 4 k abs =ðkabs
kel Þ4ðexpð
kel 4t Þ
expð
kabs 4t ÞÞÞ=ðKD þ ðDose=VDss4kabs =ðkabs
kel Þ4ðexpð
kel4t Þ
expð
kabs 4t ÞÞÞÞ; Dð pÞðt Þ ¼
ka 4pðt Þ
ks 4pðt Þþ ki 4ðA4expð
a4t Þ

d½EP ¼ ki 4 ½OP 4 ½E
ks 4 ½EP
kobs 4 ½EP þ EPOX dt
ka 4 ½EP The time course of the oxime concentrations was calculated with the following Bateman function: ½Oxime ¼


Dose kabs 4 4 e
kel 4t
e
kabs 4 t VDss kabs
kel

The time course of the OP concentrations was calculated by multi-exponential functions as shown below. The resulting differential equations were numerically resolved with Maple 9.0. Due to different procedures for the calculation of inhibitor concentrations three different templates were used for intravenous sarin (cyclosarin), intravenous VX, and percutaneous VX exposure. Intravenous sarin and cyclosarin exposure ½sarin=cyclosarin ¼ A 4 e
at þ B 4 e
bt

þ B4expð
b4t Þ þC4expð
c4t ÞÞ4eðt Þ
pðt Þ4kr4ðDose=VDss 4k abs=ðkabs
kelÞ 4ðexpð
kel 4t Þ
expð
kabs 4t ÞÞÞÞ =ðKD þðDose=VDss 4k abs =ðkabs
kel Þ4ðexpð
kel 4t Þ
expð
kabs4t ÞÞÞÞ Percutaneous VX exposure ½VX ¼ ½VX0 e
et
½VX0 e
at Formula for Maple 9.0: DðeÞðt Þ ¼
ki4ðVX0 4ðexpð
e4t ÞÞ
VX0 4ðexpð
a4t ÞÞÞ 4eðt Þþks 4pðt Þþpðt Þ4kr4ðDose=VDss 4 k abs =ðkabs
kel Þ4ðexpð
kel 4t Þ
expð
kabs 4 t ÞÞÞ =ðKD þ ðDose=VDss 4 k abs =ðkabs
kel Þ 4ðexpð
kel 4t Þ
expð
kabs 4t ÞÞÞÞ; Dð pÞðt Þ ¼
ka 4pðt Þ
ks 4pðt Þ þ ki4ðVX0 4ðexpð
e4t ÞÞ
VX04ðexpð
a4t ÞÞÞ4eðt Þ
pðt Þ4kr 4ðDose=VDss 4 k abs =ðkabs
kel Þ4ðexpð
kel 4t Þ

Formula for Maple 9.0: DðeÞðt Þ ¼
ki 4ð A4expð
a4t Þ þ B4expð
b4t ÞÞ4eðt Þ


expð
kabs 4t ÞÞÞÞ=ðKD þðDose=VDss 4k abs

þ ks 4pðt Þ þ pðt Þ4kr 4ðDose=VDss 4k abs

=ðkabs
kel Þ4ðexpð
kel 4t Þ
expð
kabs 4 t ÞÞÞÞ

=ðkabs
k el Þ4ðexpð
kel 4t Þ
expð
kabs 4t ÞÞÞ =ðKD þ ðDose=VDss 4kabs =ðkabs
kel Þ4ðexpð
kel 4t Þ
expð
kabs 4 t ÞÞÞÞ; Dð pÞðt Þ ¼
ka 4pðt Þ
ks 4pðt Þ þ ki 4ðA4expð
a4t Þ þ B4expð
b4t ÞÞ4eðt Þ
pðt Þ4kr 4ðDose=VDss 4 k abs =ðkabs
kel Þ 4ðexpð
kel 4t Þ
expð
kabs 4t ÞÞÞÞ =ð K D þ ðDose=VDss 4 k abs =ðkabs
kel Þ 4ðexpð
kel 4t Þ
expð
kabs 4t ÞÞÞÞ Intravenous VX exposure ½VX ¼ A 4 e
at þ B 4 e
bt þ C 4 e
ct

References Aas, P., Sterri, S.H., Hjermstad, H.P., Fonnum, F., 1985. A method for generating toxic vapors of soman: toxicity of soman by inhalation in rats. Toxicol. Appl. Pharmacol. 80, 437 – 445. Aldridge, W.N., Reiner, E., 1972. Enzyme Inhibitors as Substrates— Interactions of Esterases with Esters of Organophosphorus and Carbamic Acids. North-Holland Publishing Company, Amsterdam, London. Baggot, J.D., 1994. Application of interspecies scaling to the bispyridinium oxime HI-6. Am. J. Vet. Res. 55, 689 – 691. Benschop, H.P., de Jong, L.P.A., 2001. Toxicokinetics of nerve agents. In: Somani, S.M., Romano, J.A. (Eds.), Chemical Warfare Agents: Toxicity at Low Levels. CRC Press, Boca Raton, pp. 25 – 81. Bismuth, C., Inns, R.H., Marrs, T.C., 1992. Efficacy, toxicity and clinical use of oximes in anticholinesterase poisoning. In: Ballantyne, B., Marrs, T.C. (Eds.), Clinical and Experimental Toxicology of Organophosphates and Carbamates. Butterworth and Heinemann, Oxford, pp. 555 – 577. Brophy, V.H., Jampsa, R.L., Clendenning, J.B., McKinstry, L.A., Jarvik, G.P., Furlong, C.E., 2001. Effects of 5V regulatory-region polymor-

202

F. Worek et al. / Toxicology and Applied Pharmacology 209 (2005) 193 – 202

phisms on paraoxonase-gene (PON1) expression. Am. J. Hum. Genet. 68, 1428 – 1436. Chilcott, R.P., Dalton, C.H., Hill, I., Davidson, C.M., Blohm, K.L., Hamilton, M.G., 2003. Clinical manifestations of VX poisoning following percutaneous exposure in the domestic white pig. Hum. Exp. Toxicol. 22, 255 – 261. Clement, J.G., Erhardt, N., 1994. In vitro oxime-induced reactivation of various molecular forms of soman-inhibited acetylcholinesterase in striated muscle from rat, monkey and human. Arch. Toxicol. 68, 648 – 655. Clement, J.G., Bailey, D.G., Madill, H.D., Tran, L.T., Spence, J.D., 1995. The acetylcholinesterase oxime reactivator HI-6 in man: pharmacokinetics and tolerability in combination with atropine. Biopharm. Drug Dispos. 16, 415 – 425. Dawson, R.M., 1994. Review of oximes available for treatment of nerve agent poisoning. J. Appl. Toxicol. 14, 317 – 331. Eyer, P., 2003. The role of oximes in the management of organophosphorus pesticide poisoning. Toxicol. Rev. 22, 165 – 190. Eyer, F., Meischner, V., Kiderlen, D., Thiermann, H., Worek, F., Haberkorn, M., Felgenhauer, K., Zilker, T., Eyer, P., 2003. Human parathion poisoning: a toxicokinetic analysis. Toxicol. Rev. 22, 143 – 163. Green, A.L., Smith, H.J., 1958. The reactivation of cholinesterase inhibited with organophosphorus compounds: 2. Reactivation by pyridinealdoxime methiodides. Biochem. J. 68, 32 – 35. Jovanovic, D., 1989. Pharmacokinetics of pralidoxime chloride—A comparative study in healthy volunteers and in organophosphorus poisoning. Arch. Toxicol. 63, 416 – 418. Kassa, J., Cabal, J., 1999. A comparison of the efficacy of acetylcholinesterase reactivators cyclohexyl methylphosphonofluoridate (GF agent) by in vitro and in vivo methods. Pharmacol. Toxicol. 84, 41 – 45. Kiderlen, D., Worek, F., Klimmek, R., Eyer, P., 2000. The phosphoryl oxime-destroying activity of human plasma. Arch. Toxicol. 74, 27 – 32. Koplovitz, I., Gresham, V.C., Dochterman, L.W., Kaminskis, A., Stewart, J.R., 1992. Evaluation of the toxicity, pathology and treatment of cyclohexylmethylphosphonofluoridate (CMPF) poisoning in rhesus monkeys. Arch. Toxicol. 66, 622 – 628. Leader, H., Vincze, A., Manisterski, B., Rothschild, N., Dosoretz, C., Ashani, Y., 1999. Characterization of O,O-diethylphosphoryl oximes as inhibitors of cholinesterases and substrates of phosphotriesterases. Biochem. Pharmacol. 58, 503 – 515. Lundy, P.M., Hansen, A.S., Hand, B.T., Boulet, C.A., 1992. Comparison of several oximes against poisoning by soman, tabun and GF. Toxicology 72, 99 – 105. Luo, C., Saxena, A., Smith, M., Garcia, G., Radic, Z., Taylor, P., Doctor, B.P., 1999. Phosphoryl oxime inhibition of acetylcholinesterase during oxime reactivation is prevented by edrophonium. Biochemistry 38, 9937 – 9947. MacIlwain, C., 1993. Study proves Iraq used nerve gas. Nature 363, 3. MacPhee-Quigley, K., Taylor, P., Taylor, S., 1985. Primary structures of the catalytic subunits from two molecular forms of acetylcholinesterase—A comparison of NH2-terminal and active center sequences. J. Biol. Chem. 260, 12185 – 12189. Masson, P., Josse, D., Lockridge, O., Viguie, N., Taupin, C., Buhler, C., 1998. Enzymes hydrolyzing organophosphates as potential catalytic scavengers against organophosphate poisoning. J. Physiol. (Paris) 92, 357 – 362. Mast, U., 1997. Reaktivierung der Erythrozyten-Acetylcholinesterase durch Oxime. Thesis, Mu¨nchen. Maxwell, D.M., Brecht, K.M., 1991. The role of carboxylesterase in species variation of oxime protection against soman. Neurosci. Biobehav. Rev. 15, 135 – 139. Nagao, M., Takatori, T., Matsuda, Y., Nakajima, M., Iwase, H., Iwadate, K., 1997. Definitive evidence for the acute sarin poisoning diagnosis in the Tokyo subway. Toxicol. Appl. Pharmacol. 144, 198 – 203. Schoene, K., Oldiges, H., 1973. Die Wirkungen von Pyridiniumsalzen gegenu¨ber Tabun—Und Sarinvergiftungen in vivo und in vitro. Arch. Int. Pharmacodyn. Ther. 204, 110 – 123.

Shafferman, A., Ordentlich, A., Barak, D., Stein, D., Ariel, N., Velan, B., 1996. Aging of phosphylated human acetylcholinesterase: Catalytic processes mediated by aromatic and polar residues of the active centre. Biochem. J. 318, 833 – 840. Sidell, F.R., 1992. Clinical considerations in nerve agent intoxication. In: Somani, S.M. (Ed.), Chemical Warfare Agents. Academic Press, San Diego, pp. 155 – 194. Sidell, F.R., 1997. Nerve agents. In: Sidell, F.R., Takafuji, E.T., Franz, D.R. (Eds.), Medical Aspects of Chemical and Biological Warfare. Borden Institute, Washington, DC, pp. 129 – 179. Sidell, F.R., Groff, W.A., 1970. Toxogonin: blood levels and side effects after intramuscular administration in man. J. Pharm. Sci. 59, 793 – 797. Sidell, F.R., Groff, W.A., 1971. Intramuscular and intravenous administration of small doses of 2-pyridinium aldoxime methochloride to man. J. Pharm. Sci. 60, 1224 – 1228. Sidell, F.R., Groff, W.A., Kaminskis, A., 1972. Toxogonin and pralidoxime: kinetic comparison after intravenous administration to man. J. Pharm. Sci. 61, 1765 – 1769. Simon, G.A., Maibach, H.I., 2000. The pig as an experimental animal model of percutaneous permeation in man: qualitative and quantitative observations—An overview. Skin Pharmacol. Appl. Skin Physiol. 13, 229 – 234. Skrinjaric-Spoljar, M., Simeon, V., Reiner, E., 1973. Spontaneous reactivation and aging of dimethylphosphorylated acetylcholinesterase and cholinesterase. Biochim. Biophys. Acta 315, 363 – 369. Spruit, H.E.T., Langenberg, J.P., Trap, H.C., Van Der Wiel, H.J., Helmich, R.B., Van Helden, H.P.M., Benschop, H.P., 2000. Intravenous and inhalation toxicokinetics of sarin stereoisomers in atropinized guinea pigs. Toxicol. Appl. Pharmacol. 169, 249 – 254. Su, C.T., Wang, P.H., Lui, R.F., Shih, J.H., Ma, C., Lin, C.H., Liu, C.Y., Wu, M.T., 1986. Kinetic studies and structure – activity relationships of bispyridinium oximes as reactivators of acetylcholinesterase inhibited by organophosphorus compounds. Fundam. Appl. Toxicol. 6, 506 – 514. Szinicz, L., Kullmann, R., 1990. Therapie der Vergiftungen mit hochtoxischen organischen Phosphorverbindungen. Wehrmed. Monatsschr. 34, 176 – 181. Taylor, P., Radic, Z., Hosca, N.A., Camp, S., Marchot, P., Berman, H.A., 1995. Structural basis for the specificity of cholinesterase catalysis and inhibition. Toxicol. Lett. 82 – 83, 453 – 458. Thiermann, H., Szinicz, L., Eyer, F., Worek, F., Eyer, P., Felgenhauer, N., Zilker, T., 1999. Modern strategies in therapy of organophosphate poisoning. Toxicol. Lett. 107, 233 – 239. van der Schans, M.J., Lander, B.J., van der Wiel, H., Langenberg, J.P., Benschop, H.P., 2003. Toxicokinetics of the nerve agent (T)-VX in anesthetized and atropinized hairless guinea pigs and marmosets after intravenous and percutaneous administration. Toxicol. Appl. Pharmacol. 191, 48 – 62. Wong, L., Radic, Z., Bru¨ggemann, J.M., Hosea, N., Berman, H.A., Taylor, P., 2000. Mechanism of oxime reactivation of acetylcholinesterase analyzed by chirality and mutagenesis. Biochemistry 39, 5750 – 5757. Worek, F., Ba¨cker, M., Thiermann, H., Szinicz, L., Mast, U., Klimmek, R., Eyer, P., 1997. Reappraisal of indications and limitations of oxime therapy in organophosphate poisoning. Hum. Exp. Toxicol. 16, 466 – 472. Worek, F., Diepold, C., Eyer, P., 1999. Dimethylphosphoryl-inhibited human cholinesterases: inhibition, reactivation, and aging kinetics. Arch. Toxicol. 73, 7 – 14. Worek, F., Eyer, P., Kiderlen, D., Thiermann, H., Szinicz, L., 2000. Effect of human plasma on the reactivation of sarin-inhibited human erythrocyte acetylcholinesterase. Arch. Toxicol. 74, 21 – 26. Worek, F., Reiter, G., Eyer, P., Szinicz, L., 2002. Reactivation kinetics of acetylcholinesterase from different species inhibited by highly toxic organophosphates. Arch. Toxicol. 76, 523 – 529. Worek, F., Thiermann, H., Szinicz, L., Eyer, P., 2004. Kinetic analysis of interactions between human acetylcholinesterase, structurally different organophosphorus compounds and oximes. Biochem. Pharmacol. 68, 2237 – 2248.