The combination of and imidazenil is an effective strategy to prevent diisopropyl fluorophosphate toxicity in mice

Fabio Pibiri*, Alan P. Kozikowski†, Graziano Pinna*, James Auta*, Bashkim Kadriu*, Erminio Costa*‡, and Alessandro Guidotti*

*Psychiatric Institute, Department of Psychiatry, College of Medicine, and †Department of Medicinal Chemistry and Pharmacognosy, University of Illinois, Chicago, IL 60612

Contributed by Erminio Costa, July 24, 2008 (sent for review May 15, 2008) Diisopropyl fluorophosphate (DFP) causes neurotoxicity related to tiary (PHY) can cross the BBB. It thus an irreversible inhibition of (AChE). Manage- protects brain AChEs from the irreversible binding of OP (10). ment of this intoxication includes: (i) pretreatment with reversible Thus, PHY weakens OP intoxication, thereby remaining the blockers of AChE, (ii) blockade of muscarinic receptors with atro- most effective treatment so far developed (8). However, PHY pine, and (iii) facilitation of GABAA receptor signal transduction by has a short half-life, and to reach a blood concentration sufficient benzodiazepines. The major disadvantage associated with this to counteract OP toxicity, it requires doses that induce unwanted treatment combination is that it must to be repeated frequently centrally mediated psychological or behavioral side effects (11– and, in some cases, protractedly. Also, the use of diazepam (DZP) 13). These side effects can be partly eliminated by a low dose of and congeners includes unwanted side effects, including sedation, scopolamine (SCO) (13). However, the administration of a amnesia, cardiorespiratory depression, and anticonvulsive toler- muscarinic blocker may lead to unwanted effects in persons ance. To avoid these treatment complications but safely protect under warm to hot weather conditions or in persons wearing against DFP-induced seizures and other CNS toxicity, we adopted protections against chemical agents. SCO administration blocking the strategy of administering mice with (i) small doses of huperzine muscarinic receptors of secretal glands may inhibit A (HUP), a reversible and long-lasting (half-life Ϸ5 h) inhibitor of sweating and thereby enhance the risk of heat injury (14, 15). AChE, and (ii) imidazenil (IMI), a potent positive allosteric modu- An additional limitation on the use of AChE inhibitors is the ␣ lator of GABA action selective for 5-containing GABAA receptors. incomplete protection from OP-induced seizures (5). Thus, this Coadministration of HUP (50 ␮g/kg s.c., 15 min before DFP) with IMI pretreatment must be followed rapidly by an administration of: (i) (2 mg/kg s.c., 30 min before DFP) prevents DFP-induced convulsions antimuscarinic agents, usually (ATR), which prevents and the associated neuronal damage and mortality, allowing (Ach) from binding to muscarininc receptors [ATR complete recovery within 18–24 h. In HUP-pretreated mice, the cannot be given prophylactically because of its short half-life and at Ϸ ED50 of IMI to block DFP-induced mortality is 10 times lower than the doses required to overcome OP-induced hypercholinergic tox- that of DZP and is devoid of sedation. Our data show that a icity, it induces disorientation, confusion, and hallucinations (16– combination of HUP with IMI is a prophylactic, potent, and safe 19)]; (ii) oxime, which reactivates AChEs blocked by OP before the therapeutic strategy to overcome DFP toxicity. ‘‘aging’’ of the binding to the (20); and (iii) anticonvulsants, usually DZP, which allosterically enhance GABA-mediated inhi- GABA receptor ͉ neurotoxicity A bition at various GABAA receptors. Unfortunately, DZP and congeners at the required anticonvulsant doses act at GABAA rganophosphate (OP) neurotoxins are among the most receptors expressing ␣1 subunits and cause sedation, amnesia, and Olethal chemical poisons ever developed and in the present cardiorespiratory depression (6). When given for protracted peri- NEUROSCIENCE political climate, should be considered possible threats to both ods, these benzodiazepines cause not only tolerance to their anti- civilians and military personnel in terrorist attacks (1, 2). convulsant effects but also dependence (21). Hence, there is an Most of the insecticides used worldwide are OP, and intoxi- urgent need to develop more effective and safe therapeutic strat- cation with these compounds represents a major public health egies against OP poisoning. concern in nonindustrialized countries (3, 4). OP neurotoxins Newly tested reversible AChE inhibitors include huperzine A target cholinergic neurotransmission by irreversibly inhibiting (HUP), an alkaloid that crosses the BBB, selectively inhibits acetylcholinesterase (AChE) and thereby inducing life-threatening AChE, and fails to bind to (BuChE) (5). disorders, including bronchial spasms and gland hypersecretion, HUP, isolated from Huperzia serrata, is currently approved in muscle fasciculation, cardiovascular impairment, tremors, convul- China for the treatment of Alzheimer’s disease and other sions, coma, and in the worst case, may even include death (5, 6). memory impairments (22) and has passed phase II clinical trials Subjects who survive severe OP-induced seizures will likely develop in the U.S. (http://clinicaltrials.gov/ct2/show/NCT00083590). At irreversible brain damage (7). doses of 100–500 ␮g/kg, HUP increases the survival of labora- The current prophylactic strategy to treat OP poisoning tory animals exposed to lethal doses of OP (5, 10, 23, 24). recommended by military organizations relies on pyridostigmine However, at these high doses, HUP inhibits blood and brain (PYR) bromide tablets taken over several days (5). PYR is a AChEs by 50–70%, a value that limits its use in healthy subjects reversible AChE blocker, which prevents the irreversible binding (5). Moreover, at a dose of 500 ␮g/kg, HUP produces untoward of OP to AChEs. In the event of OP poisoning, this pretreatment side effects such as chewing, salivation, tremors, and behavioral must be supported by a triple therapy based on atropine, oxime, and a benzodiazepine compound, usually diazepam (DZP). Several limitations are associated with this conventional prophy- Author contributions: F.P., A.P.K., G.P., J.A., B.K., E.C., and A.G. designed research, per- lactic treatment for subjects at risk for OP exposure (5). PYR formed research, analyzed data, and wrote the paper. prevents OP-induced hypercholinergic toxicity and lethality but The authors declare no conflict of interest. because of its low blood–brain barrier (BBB) penetration, fails to ‡To whom correspondence should be addressed. E-mail: [email protected]. protect against OP-induced convulsions (8, 9). However, the ter- © 2008 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807172105 PNAS ͉ September 16, 2008 ͉ vol. 105 ͉ no. 37 ͉ 14169–14174 Downloaded by guest on September 30, 2021 HUP (ED =34±3.0 µg/kg) 150 125 50 LD50 = 2.7 ± 0.2 µg/kg HUP+IMI (ED50=16±3.0 µg/kg) 100 100 75

50 50

lethality (%) 25

0 0 Protection of DFP-induced

% of DFP-induced lethality 0.1 0.3 1 3 10 30 1 10 100 Huperzine [µg/kg s.c.] DFP [µg/kg s.c.] Fig. 2. Protective action of HUP and the combination of HUP with IMI against Fig. 1. Dose–response of DFP-induced lethality in mice. Results were obtained DFP-induced lethality. Mice were treated with DFP (6 ␮g/kg s.c., Ϸ2ϫ LD50)15 by using 6–10 mice per dose of DFP. Lethality was established 48 h after DFP min after HUP (0.3–100 ␮g/kg s.c.) and 30 min after IMI (2 mg/kg s.c.). All values intoxication. are the average of at least 6 animals per group. Lethality was established 48 h after DFP intoxication. deficits (5, 10, 24). Because prophylactic treatment with HUP fails to prevent the development of seizures in OP-intoxicated LD50 of 2.7 Ϯ 0.2 ␮g/kg s.c. (Fig. 1). DFP, at a dose of 6 ␮g/kg (Ϸ2ϫ animals, HUP treatment should be associated with a nonsedative LD50), elicits a cholinergic hyperactivity in 2–3 min. After 5–6 min, benzodiazepine [i.e., imidazenil (IMI)]. severe convulsions appear (range 4–5 of Racine scale), which evolve IMI is a positive selective allosteric modulator of GABA action into a status epilepticus in 7–8 min. Mice generally die within 10 min ␣ at 5-containing GABAA receptors, which, unlike DZP, is inactive (Table 1). Based on these characteristics, this DFP dose was ␣ at 1-containing GABAA receptors (25, 26). Thus, IMI elicits a selected to study the dose-dependent protective efficacy of pre- potent anticonvulsant and anxiolytic action at doses that are treatment with HUP alone or in combination with IMI. virtually devoid of sedative and amnesic effects (6, 25–31). Of note, Fig. 2 shows that HUP given 15 min before DFP (6 ␮g/kg s.c.) IMI also potently antagonizes diisopropyl fluorophosphate (DFP)- dose-dependently attenuates DFP-induced severe muscarinic ef- induced seizures and mortality in rodents (6, 32). This drug, at doses fects and protects mice against DFP lethality (ED50,34Ϯ 3.0 up to 60-fold greater than those producing anxiolytic and anticon- ␮g/kg). However, Table 1 shows that HUP doses (50 ␮g/kg s.c. or vulsant actions, antagonizes the sedative and amnesic actions of higher) that effectively protect against DFP-induced lethality fail to DZP or alprazolam without causing tolerance (26–30, 33). Fur- prevent DFP-induced recurrent seizures (range 4–5 of Racine thermore, in non-human primates, IMI is virtually devoid of scale). Seizures generally last for 8–12 h after a DFP challenge. tolerance and dependence liability (29, 30). The IMI clearance rate in vivo is slower (in rats the t1͞ is 180 min, and in non-human 2 IMI Potentiates HUP Efficacy Against DFP-Induced Mortality. Fig. 3 primates its t1͞ is Ͼ8 h) than that of DZP (26). 2 compares the potency of IMI with that of DZP in protecting mice For a prophylactic treatment that is devoid of side effects and ␮ effective against DFP-induced toxicity, we used a combined pretreated with 25 g/kg s.c. HUP from DFP-induced death. treatment of HUP and IMI. We show that pretreatment with a This dose of HUP alone protects only 20% of mice from low dose of HUP (50 ␮g/kg s.c., 15 min before DFP) in DFP-induced death (Table 1). Although IMI or DZP alone fails combination with IMI (2 mg/kg s.c., 30 min before DFP) is a safe to protect against DFP-induced lethality (Fig. 3), a synergistic prophylactic treatment against OP toxicity. This treatment not protective interaction occurs between these benzodiazepines and only prevents death but also protects against the seizures and HUP. IMI is Ϸ10-fold more potent than DZP in protecting neurodegeneration that follow OP intoxication. Our data also HUP-pretreated mice from DFP-induced death (ED50 IMI, 0.08 suggest that a combination of HUP and IMI could safely be used mg/kg; ED50 DZP, 0.83 mg/kg; see Fig. 3). Moreover, in IMI- as a treatment after OP intoxication. treated mice (2 mg/kg s.c. 30 min before DFP), the dose– response curve of HUP protection against DFP-induced mor- Results tality shifts 2-fold toward the left (ED50 16 Ϯ 3 ␮g/kg; Fig. 2). At Doses of HUP That Protect Mice Against DFP-Induced Mortality Fail to a dose of 50 ␮g/kg s.c. HUP fails per se to alter locomotion or Abolish DFP-Induced Seizures. DFP in doses of 0.3–12 ␮g/kg s.c. memory retention (Fig. 4), whereas a dose of 100 ␮g/kg s.c. elicits a clear dose-dependent hypercholinergic toxicity with a strongly impairs motility and memory (Fig. 4). IMI (2 mg/kg s.c.)

Table 1. Pretreatment combination with huperzine and imidazenil protects mice against DFP-induced seizures and lethality Seizures

Grade 0–3, Grade 4–5, Status epilepticus, Seizures Mortality, Death, Drug treatment min min min end, h % min

Vehicle ϩ DFP 4.2 Ϯ 0.3 7.4 Ϯ 0.6 8.4 Ϯ 0.3 100 9.3 Ϯ 0.9 HUP (100 ␮g/kg) ϩ DFP 9.7 Ϯ 1.2 14.3 Ϯ 1.3 No 9 Ϯ 20 HUP (50 ␮g/kg) ϩ DFP 7.8 Ϯ 0.2 12.3 Ϯ 1.8 No 10 Ϯ 20 HUP (25 ␮g/kg) ϩ DFP 6.0 Ϯ 0.9 8.5 Ϯ 0.6 24 Ϯ 3.6 Ͼ12 80 27 Ϯ 5.6 HUP (50 ␮g/kg) ϩ IMI ϩ DFP 17.3 Ϯ 2.3 No No 4 Ϯ 20 HUP (25 ␮g/kg) ϩ IMI ϩ DFP 12.4 Ϯ 0.9 No No 6 Ϯ 20

Huperzine (HUP) was injected s.c. 15 min before, and imidazenil (IMI, 2 mg/kg s.c.) was injected 30 min before the challenge with 2ϫ LD50 of DFP (6 ␮g/kg s.c.). Mean Ϯ SEM, n ϭ 10–30 mice per group.

14170 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807172105 Pibiri et al. Downloaded by guest on September 30, 2021 120 IMI or DZP in absence of Veh DFP 100% mortality HUP 100 -30 -15 0 15 30 min 80 HUP DFP 100% mortality // // 60 -60 -30 -15 0 1 min 3 hours IMI 5 40 DZP HUP DFP 25% mortality // 20 -30 -15 0 1 min 14 hrs 5 % of DFP-induced lethality % 0 // // HUP DFP 0% mortality 0 0 0.06 0.18 0.56 1.8 5.6 // IMI or DZP [mg/kg s.c.] -30 -15 0 1 min 24 hrs (-) + 5 + + + ++ HUP + IMI DFP 100% mortality HUP [25 µg/kg s.c.] // // -60 -30 -15 0 1 min 6 hours Fig. 3. IMI is Ϸ10 times more potent than DZP in protecting against DFP- 5 induced mortality (ED50 IMI, 0.08 mg/kg; ED50 DZP, 0.83 mg/kg). Mice were HUP + IMI DFP 0% mortality pretreated with HUP (25 ␮g/kg s.c., 15 min before DFP) and with various doses of // DZP (p)orIMI(ࡗ) s.c. 30 min before DFP (6 ␮g/kg s.c.). Groups of mice were -30 -15 0 15 min 24 hrs pretreated with DZP or IMI alone (Œ), with HUP alone (■) or with vehicle alone (ᮀ)15 min before the challenge with DFP. Each point is the average of 5 different mice. HUP + IMI DFP 0% mortality // -30 -15 0 1 min 24 hrs either alone or in combination with HUP (50 ␮g/kg s.c.) fails to 5 affect locomotion and mnemonic functions (Fig. 4). Of note, Fig. 5. Time course of the protective action of HUP and the combination of unlike DZP, IMI in combination with HUP at a dose that HUP with IMI against DFP-induced lethality. Mice were treated with DFP (6 reduces DFP-induced lethality fails to induce sedation, amnesia, ␮g/kg s.c., Ϸ2ϫ LD50) at various times after HUP (50 ␮g/kg s.c.) alone or in and muscle relaxation (6, 29). combination with IMI (2 mg/kg s.c.). All values are the average of at least 6 Prophylactic treatment against OP exposure could be given animals per group. Lethality was established 24 h after DFP intoxication. before intoxication. Hence, we studied whether IMI prolongs the duration of HUP protection against DFP-induced lethality. As shown in Fig. 5, if given 15 min before DFP exposure, a dose of induced mortality, but also against DFP-induced seizures (Table 50 ␮g/kg HUP is fully protective against DFP-induced lethality. 1). Seizure onset is delayed from 8 to 10 min in mice receiving However, its potency is reduced to 75% if given 30 min before only HUP, to 15–20 min when HUP is given with IMI, and in DFP and entirely loses its efficacy if given 1 h before DFP these mice, seizures never reached level 4–5 on the Racine scale intoxication. Nonetheless, IMI [as expected by its long half-life (Table 1). The window for recurrent seizures was considerably in rodents (26)] potentiates the protective action of HUP. Also, reduced from 8–12 h to 3–4 h in mice receiving HUP in IMI prolongs the efficacy of HUP at 1 h pretreatment, delaying combination with IMI (Table 1). occurrence of death from 3 to 6 h (Fig. 5). Clear signs of DFP-induced neurotoxicity (TUNEL-positive nuclei) in the cortex and hippocampus of mice pretreated with Combination of IMI and HUP Protects Mice Against DFP-Induced HUP (50 ␮g/kg s.c., 15 min before DFP) appear 48 h after a DFP Seizures, Neurotoxicity, and Cognitive Impairment. A combination challenge (Fig. 6). In contrast, these brain areas do not show

of IMI (2 mg/kg s.c. 30 min before DFP) and HUP (25 or 50 NEUROSCIENCE ␮g/kg s.c., 15 min before DFP) not only protects against DFP- Vehicle HUP (50 µg/kg) HUP+IMI (2 mg/kg) A B C VH HUP 100 HUP 50 IMI IMI+HUP 50 Cingulate Cortex 6000 A 160 B A B C 120 4000 CA1 80 * 2000 ** A B C 40 Dentate Gyrus Freezing Time (sec/5 min) Freezing Time

Locomotion (counts/15 min) 0 0

Fig. 4. Locomotor activity (A) and contextual fear conditioning (B) in mice Fig. 6. DFP-induced TUNEL gray-positive reaction is not present in neurons treated with HUP (100 or 50 ␮g/kg s.c.), IMI (2 mg/kg s.c.), and a combination of the cingulate cortex (Top), CA1 (Middle), and dentate gyrus (Bottom)ofthe of the two drugs. All drugs were injected 30 min before each test. Each group hippocampus in mice pretreated 48 h before with a combination of HUP (50 is the average of 5 different mice. (A) **, P Ͻ 0.01 when vehicle-treated group ␮g/kg s.c., 15 min before DFP) and IMI (2 mg/kg s.c., 30 min before DFP). (A) is compared with drug-treated groups (ANOVA followed by Newman–Keuls Vehicle-treated mice. (B) HUP-pretreated mice intoxicated with DFP showing multiple comparison test). (B) *, P Ͻ 0.01 when vehicle-treated group is TUNEL-positive neurons in all of the three index areas. (C) HUP ϩ IMI- compared with drug-treated groups (one-way ANOVA followed by pretreated mice intoxicated with DFP showing almost complete absence of Newman–Keuls multiple comparison test). VH, vehicle; HUP 50, HUP 50 ␮g/kg TUNEL-positive neurons. Photomicrographs are representative of results ob- s.c.; HUP 100, HUP 100 ␮g/kg s.c.; IMI, IMI 2 mg/kg s.c. tained from groups of 5 mice. (Scale bars: 40 ␮m.)

Pibiri et al. PNAS ͉ September 16, 2008 ͉ vol. 105 ͉ no. 37 ͉ 14171 Downloaded by guest on September 30, 2021 ABVH By reducing DFP-induced cholinomimetic toxicity, HUP pre- HUP 50 treatment in doses of 100 ␮g/kg s.c. or higher effectively prevents 5000 200 IMI + HUP 50 OP-induced lethality (Table 1). However, at the dose of 100 ␮g/kg s.c., HUP causes a significant decrease in locomotor 4000 150 activity and reduces mnemonic function (Fig. 4). In rodents, 3000 100 * higher doses of HUP (200 ␮g/kg or higher) induce dangerous 2000 unwanted side effects, including profuse salivation, bronchial 1000 50 hypersecretion, arterial , tremors, and convulsions (5, 10, 24). 0 (sec/5 min) Freezing Time 0 Locomotion (counts/15 min) Of note, HUP alone, even if administered prophylactically in Fig. 7. Locomotor activity (A) and contextual fear conditioning (B) in mice doses that protect mice from DFP-induced death, provides only treated with vehicle, HUP (50 ␮g/kg s.c., 15 min before DFP) ϩ DFP (6 ␮g/kg marginal protection against DFP-induced seizures (Table 1). s.c.), and HUP (50 ␮g/kg s.c., 15 min before DFP) ϩ IMI (2 mg/kg s.c., 30 min Moreover, HUP fails to protect DFP-induced cortical and before DFP). The experiments were carried out 1 week after DFP intoxication. hippocampal neurotoxicity (Fig. 6), likely caused by the pro- Ͻ Each group is the average of 5 different mice. *, P 0.01 when vehicle-treated tracted occurrence and duration of seizures (Table 1). This group is compared with drug-treated groups (one-way 〈⌵⌷VA followed by Newman–Keuls multiple comparison test). VH, vehicle; HUP 50, HUP 50 ␮g/kg evidence underscores the relevance of optimizing prophylactic s.c.; IMI, IMI 2 mg/kg s.c. action against OP toxicity using HUP by adding an anticonvul- sant to this drug. To abolish DFP-induced seizures and neurotoxicity, we inves- signs of nuclear neuronal damage in HUP- and IMI- (30 min tigated whether the combination of HUP with DZP or IMI before DFP) treated mice (Fig. 6), suggesting that these drugs improves the protective action of HUP. Pretreatment adminis- together offer powerful neuroprotection. tration of these benzodiazepines combined with HUP provides One week after DFP intoxication, the combination of HUP full protection against DFP-induced lethality (Fig. 3) and exerts and IMI prevents cognitive performance deficits (Fig. 7B) that a potent anticonvulsant action (Table 1). Of note, IMI is 10 times develop in mice treated with HUP alone (Fig. 7B). Of note, these more potent than DZP in counteracting DFP-induced lethality, neuroprotective doses of IMI and HUP fail to produce behav- and unlike DZP, it fails to produce sedation, amnesia, or ioral dysfunctions (Fig. 4). respiratory depression and fails to induce tolerance and depen- dence after repeated treatments (6–29). Moreover, the half-life HUP and IMI in Posttreatment Prevent DFP-Induced Toxicity. Because of IMI (Ϸ3–4 h in rodents) is longer lasting than that of it is difficult to predict when a person will be exposed to toxic DZP (26). levels of nerve agents (for example, in case of terrorist attacks), After the coadministration of HUP and IMI, in our histolog- we studied whether posttreatment with HUP and IMI could also ical studies of toxicity, we failed to detect the signs of neuro- effectively counteract the acute toxicity of OPs. In posttreat- toxicity that were found in the brain of mice pretreated with ment, it is imperative to add ATR to reduce the acute muscarinic HUP alone (Fig. 6). Moreover, HUP in combination with IMI syndrome that follows a few minutes after an administration of also fails to affect motility and memory functions (Fig. 4). Thus, DFP (Table 2). Posttreatment with ATR (10 mg/kg i.p.) and HUP and IMI provide an effective combination therapy that is HUP (50 ␮g/kg s.c.) given soon after intoxication with DFP devoid of side effects and can efficaciously be used as a medi- protects all affected mice from DFP-induced mortality (Table 2). cation for life-threatening exposure to OP. The half-life of HUP Furthermore, when mice are treated immediately after DFP and IMI in humans is Ϸ5 and Ϸ12 h, respectively (26, 34). injection with HUP, ATR, and IMI (2 mg/kg s.c.), these mice are Moreover, even after repeated treatment, HUP and IMI do not also protected from the seizures that follow DFP intoxication develop tolerance toward their protective effects (5, 26). If (Table 2). applied to military personnel or civilians at risk for nerve toxin exposure, the prophylactic treatment against OP toxicity re- Discussion ported in the present work is particularly significant and should The present work reports on the remarkable efficacy and low guarantee potent and efficacious protection for the various side effect liability of prophylactic treatment with IMI in com- symptoms of OP toxicity. At the same time, this treatment fails bination with HUP against DFP-induced neurotoxicity and to modify cognitive performance and alertness (Fig. 4). It is lethality in mice. noteworthy that the prophylactic efficacy of this treatment

Table 2. Posttreatment combination with huperzine, atropine, and imidazenil protects mice against DFP-induced seizures and lethality Seizures

Grade 0–3, Grade 4–5, Status epilepticus, Seizures Mortality, Drug treatment min min min end, h % Death

Vehicle ϩ DFP 4.2 Ϯ 0.3 7.4 Ϯ 0.6 8.4 Ϯ 0.3 100 9.3 Ϯ 0.9 min DFP ϩ HUP (100 ␮g/kg) 4.7 Ϯ 0.8 8.4 Ϯ 0.8 9.1 Ϯ 0.1 100 9.9 Ϯ 0.2 min DFP ϩ HUP (100 ␮g/kg) ϩ IMI 5.3 Ϯ 0.8 8.7 Ϯ 0.1 9.5 Ϯ 0.9 100 10 Ϯ 0.9 min DFP ϩ ATR 6.1 Ϯ 0.9 9.7 Ϯ 0.9 11 Ϯ 0.8 100 4 h DFP ϩ ATR ϩ HUP (25 ␮g/kg) 6.1 Ϯ 0.5 9.1 Ϯ 0.4 No 100 10 h DFP ϩ ATR ϩ HUP (50 ␮g/kg) 6.8 Ϯ 0.5 8.9 Ϯ 0.4 No Ͼ12 0 DFP ϩ ATR ϩ HUP (100 ␮g/kg) 7.8 Ϯ 0.9 9.9 Ϯ 0.3 No Ͼ12 0 DFP ϩ ATR ϩ HUP (50) ϩ IMI 6.9 Ϯ 0.3 no No Ͼ12 0 DFP ϩ ATR ϩ HUP (100) ϩ IMI 8.6 Ϯ 0.5 no No Ͼ12 0

Huperzine (HUP), imidazenil (IMI, 2 mg/kg s.c.), and atropine (ATR, 10 mg/kg i.p.) were injected 1 min after the challenge with 2ϫ LD50 of DFP (6 ␮g/kg s.c.). Mean Ϯ SEM, n ϭ 10–30 mice per group.

14172 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807172105 Pibiri et al. Downloaded by guest on September 30, 2021 combination is not achieved by administering HUP alone or in dissolved in sterile saline, and IMI and DZP were dissolved in vegetable oil. The combination with DZP or other commercially available benzo- drugs were injected s.c. between the shoulder blades of the animals. All diazepines whose pharmacological profile includes sedative ef- injections (100 ␮l/10 g) were performed by using disposable tuberculin sy- fects among other unwanted features. ringes with 26-gauge needles. The precise mechanism that accounts for the efficacy of treatment with HUP and IMI against DFP intoxication is still to Fear Conditioning. Fear conditioning was performed by using a computerized fear-conditioning system as reported in detail in ref. 39. Mice were placed into the be fully elucidated. HUP is a specific reversible blocker of the training chamber (10 ϫ 10ϫ 12 cm, clear plastic) and allowed to explore for 3 min. AChEs and does not interact with the BuChE and carboxyles- After this time, they received an electric foot shock (2 s, 0.5 mA) delivered through terase (CarbE) present in the periphery, so these two proteins the floor grid three times every 2 min. After the last shock, mice were allowed to can act as endogenous scavengers and detoxify DFP in the explore the context for an additional 1 min before removal from the training bloodstream before it reaches critical concentration levels in the chamber and placed back into the home cage. Freezing behavior was measured brain (35). Additionally, this drug has been shown to dose- 24 h after training (for 5 min) in which the percentage of time spent in the context dependently inhibit the binding of two antagonists at the N- in which they were shocked was assessed. Freezing involved the absence of all methyl-D-aspartate (NMDA) glutamate receptors, thus showing movement except for respiratory-related movements while the mouse was in a a possible antagonistic action toward this receptor subtype (36). stereotypical crouching posture (40). It is well established that this receptor subtype triggers neuronal damage caused by seizures (36). Measurement of Locomotor Activity in a Novel Cage. A computerized AccuScan 12 animal activity monitoring system (Columbus Instruments) assisted by IMI acts as a nonsedative and potent antiepileptic agent VERSAMAX software (AccuScan Instruments) monitored locomotor activity in endowed with a long duration of action (37), and unlike DZP and mice (41). Each activity cage consisted of a Perspex box (20 ϫ 20 ϫ 20 cm) other benzodiazepines, it protects against OP agent-induced surrounded by horizontal and vertical infrared sensor beams. The locomotor seizures without producing sedation, tolerance, or dependence activity of these mice was recorded between 1:00 and 3:00 p.m. in the facility liability (6). Because of these characteristics, IMI is an excellent room where they had been housed. candidate as the drug of choice to counteract OP-induced seizures. IMI acts specifically at ␣5-containing GABAA recep- Measurement of Seizure Activity. The Racine scale was used (42) with the tors (25), which are the receptor subtypes involved in its anti- introduction of minor modifications to this seizure model. Behaviors were represented by the following numbers: 0, behavioral arrest (motionless), hair convulsant action (38). IMI is inactive at ␣1-containing GABAA receptors, which accounts for the lack of sedative and amnesic raising, excitement, and rapid breathing; 1, mouth movement of lips and actions and most likely for its failure to develop tolerance and tongue, vibrissae movements and salivation; 2, head clonus and eye clonus; 3, dependence during protracted treatment (6, 26–31). More im- forelimb clonus, ‘‘wet dog shakes’’; 4, clonic rearing; 5, clonic rearing with loss of postural control and uncontrollable jumping. portantly, because it is devoid of the unwanted side effects of classical benzodiazepines (37), IMI can be safely used as a Neuronal Damage Evaluation. Forty-eight hours after the acute DFP challenge, prophylactic treatment. mice were anesthetized with Nembutal (50 mg/kg i.p.) and perfused with 4% Relevant to accidental or unforeseen exposure to OP, the paraformaldehyde in PBS (pH 7.4). Brains were removed and postfixed in 4% present work also demonstrates that HUP and IMI in combi- paraformaldehyde for at least 72 h in PBS (pH 7.4) at 4°C and then transferred nation with ATR (administered as a single dose immediately to 30% sucrose in PBS (pH 7.4) at 4°C for cryoprotection. Fixed brains were then after DFP exposure) fully counteracts DFP-induced neurotox- embedded in frozen section medium, and 20-␮m thick sections were cut in a icity, including lethality (Table 2). cryostat at 4°C and mounted on glass slides for TUNEL assays. In conclusion, this article demonstrates the efficacy of a Tissues were permeabilized with 0.1% Triton X-100 in freshly prepared combination treatment with IMI and HUP that can safely be 0.1% citrate for 2 min on ice, rinsed briefly, and incubated in protein- used against the toxicity of OP compounds. The therapeutic ase K (15 mg/ml) for 10 min at room temperature and washed three times with PBS buffer for a total of 30 min. Endogenous peroxidase was quenched with strategy of using this combination of IMI and HUP offers an 0.3% (vol/vol) hydrogen peroxide in methanol for 30 min, and the tissue was excellent prophylactic tool against OP exposure because of its rinsed with PBS for 10 min. After equilibration in PBS buffer, sections were long-lasting action and the lack of unwanted side effects when incubated with terminal deoxynucleotide transferase (TdT)-containing NEUROSCIENCE given to healthy subjects. The prophylactic use of this treatment digoxigenin-dUTP at 37°C for1hinahumidified chamber. The TdT reaction is particularly advisable for those workers who are at risk for was terminated by three washes in PBS buffer for 10 min. Antidigoxigenin being accidentally exposed to OP substances, including farmers peroxidase (converter-POD) was added to the tissue sections, and the sections or military personnel during war actions (1, 2). Of course, this were incubated for an additional 30 min at 37°C in a humidified chamber. Two treatment would become extremely useful for civilians in the controls per assay were performed: incubating sections with DNase I served as case of terrorist attacks. positive control, and omission of the terminal transferase from the reaction mixture served as the negative control. Staining was developed with diami- Materials and Methods nobenzidine–nickel ammonium sulfate, and light microscopic examination of the sections was performed at 40ϫ magnification. Animals and Drugs. Adult male Swiss–Webster mice (Harlan Breeders), 25–30 g of body weight, maintained under a 12-h dark/light cycle and food and ACKNOWLEDGMENTS. We thank Drs. Maria Luisa Barbaccia (University of water ad libitum were used for these studies. Mice were housed in groups of Rome, Tor Vercata, Italy) and Norton H. Neff (Ohio State University College of five per cage (24 ϫ 17 ϫ 12 cm). The Internal Review Board at the University Medicine) for constructive criticisms and suggestions in the preparation of the of Illinois at Chicago approved all animal experiment protocols for Animal article. This work was supported by National Institute of Mental Health/ Welfare. DFP and diazepam were purchased from Sigma–Aldrich. HUP was National Institutes of Health Grants MH5680 (to A.G.) and MH062090 (to E.C.) kindly offered by A. P. Kozikowski from the Department of Medicinal Chem- and a Regione Autonoma della Sardegna, Italy, ‘‘Master and Back’’ postdoc- istry and Pharmacology (University of Illinois, Chicago, IL). HUP and DFP were toral fellowship (to F.P.).

1. Coupland R, Leins KR (2005) Science and prohibited weapons. Science 308: 6. Auta J, Costa E, Davis J, Guidotti A (2004) Imidazenil: A potent and safe protective 1841. agent against diisopropyl fluorophosphate toxicity. Neuropharmacology 46:397–403. 2. Romano JA, King JM (2001) Psychological casualties resulting from chemical and 7. Lemercier G, Carpentier P, Sentenac-Roumanou H, Morelis P (1983) Histological and biological weapons. Mil Med 166:21–22. histochemical changes in the central nervous system of the rat poisoned by an irre- 3. Karalliedde L, Senanayake N (1989) Organophosphorus insecticide poisoning. Br J versible anticholinesterase organophosphorus compound. Acta Neuropathol 61:123– Anaesth 63:736–750. 129. 4. Buckley NA, Roberts D, Eddleston M (2004) Overcoming apathy in research on organo- 8. Miller SA, Blick DW, Kerenyi SZ, Murphy MR (1993) Efficacy of physostigmine as a phosphate poisoning. BMJ 329:1231–1233. pretreatment for organophosphate poisoning. Pharmacol Biochem Behav 44:343–347. 5. Lallement G et al. (2002) Review of the value of huperzine as pretreatment of 9. Kerenyi SZ, Murphy MR, Hartgraves SL (1990) Toxic interactions between repeated organophosphate poisoning. Neurotoxicology 23:1–5. and chronic pyridostigmine in rodents. Pharmacol Biochem Behav 37:267–271.

Pibiri et al. PNAS ͉ September 16, 2008 ͉ vol. 105 ͉ no. 37 ͉ 14173 Downloaded by guest on September 30, 2021 10. Grunwald J, Raveh L, Doctor BP, Ashani Y (1994) Huperzine A as a pretreatment 27. Auta J, Giusti P, Guidotti A, Costa E (1994) Imidazenil, a partial positive allosteric candidate drug against toxicity. Life Sci 54:991–997. modulator of GABAA receptors, exhibits low tolerance and dependence liabilities in 11. Coelho F, Birks J (2001) Physostigmine for Alzheimer’s disease. Cochrane Database Syst the rat. J Pharmacol Exp Ther 270:1262–1269. Rev 2:CD001499. 28. Auta J, et al. (1995) Comparison of the effects of full and partial allosteric modulators 12. Philippens IH, Wolthuis OL, Busker RW, Langenberg JP, Melchers BP (1996) Side effects of GABAA receptors on complex behavioral processes in monkeys. Behav Pharmacol of physostigmine as a pretreatment in guinea pigs. Pharmacol Biochem Behav 55:99– 6:323–332. 105. 29. Auta J, Guidotti A, Costa E (2000) Imidazenil prevention of alprazolam-induced acqui- 13. Philippens IH, et al. (1998) Subchronic physostigmine pretreatment in guinea pigs: sition deficit in patas monkeys is devoid of tolerance. Proc Natl Acad Sci USA 97:2314– Effective against soman and without side effects. Pharmacol Biochem Behav 59:1061– 2319. 1067. 30. Costa E, Auta J, Guidotti A (2001) Tolerance and dependence on ligands of the 14. Meshulam Y, Davidovici R, Wengier A, Levy A (1995) Prophylactic transdermal treat- benzodiazepine sites expressed by GABA receptors. Handbook of Experimental ment with physostigmine and scopolamine against soman intoxication in guinea pigs. A Pharmacology, Mohler H, ed (Springer, Berlin), Vol 150, pp 227–250. J Appl Toxicol 15:263–266. 31. Impagnatiello F, et al. (1996) Modifications of ␥-aminobutyric acid A receptor subunit 15. Philippens IH, Melchers BP, Olivier B, Bruijnzeel PL (2000) Scopolamine augments the efficacy of physostigmine against soman poisoning in guinea pigs. Pharmacol Biochem expression in rat neocortex during tolerance to diazepam. Mol Pharmacol 49:822–831. Behav 65:175–182. 32. Rump S, et al. (2000) Effects of imidazenil, a new benzodiazepine receptor partial 16. Anderson DR, et al. (1994) Efficacy comparison of scopolamine and diazepam against agonist, in the treatment of convulsions in organophosphate intoxications. Neurotox soman-induced debilitation in guinea pigs. Fundam Appl Toxicol 22:588–593. Res 2:17–22. 17. Janowsky DS (2002) Central to treat nerve agent poisoning. Lancet 33. Thompson DM, Auta J, Guidotti A, Costa E (1995) Imidazenil, a new anxiolytic and 359:265–266. anticonvulsant drug, attenuates a benzodiazepine-induced cognition deficit in mon- 18. Lallement G et al. (2001) Subchronic administration of various pretreatments of nerve keys. J Pharmacol Exp Ther 273:1307–1312. agent poisoning. I. Protection of blood and central cholinesterases, innocuousness 34. Qian BC, et al. (1995) of tablet huperzine A in six volunteers. Acta toward blood–brain barrier permeability. Drug Chem Toxicol 24:151–164. Pharmacol Sin 16:396–398. 19. Lallement G, Foquin A, Dorandeu F, Baubichon D, Carpentier P (2001) Subchronic 35. Broomfield CA, et al. (1991) Protection by butyrylcholinesterase against organophos- administration of various pretreatments of nerve agent poisoning. II. Compared phorus poisoning in nonhuman primates. J Pharmacol Exp Ther 259:633–638. efficacy against soman toxicity. Drug Chem Toxicol 24:165–180. 36. Gordon RK et al. (2001) The NMDA receptor ion channel: A site for binding of 20. Dawson RM (1994) Review of oximes available for treatment of nerve agent poisoning. huperzine A. J Appl Toxicol 21:S47–S51. J Appl Toxicol 14:317–331. 37. Costa E, Guidotti A (1996) Benzodiazepines on trial: A research strategy for their 21. Steppuhn KG, Turski L (1993) Diazepam dependence prevented by glutamate antag- rehabilitation. Trends Pharmacol Sci 17:192–200. onists. Proc Natl Acad Sci USA 90:6889–6893. 38. Rudolph U, et al. (1999) Benzodiazepine actions mediated by specific ␥-aminobutyric 22. Kozikowski AP, et al. (1991) Delineating the pharmacophoric elements of huperzine A: acid (A) receptor subtypes. Nature 401:796–800. Importance of the unsaturated three-carbon bridge to its AChE inhibitory activity. 39. Pibiri F, Nelson M, Guidotti A, Costa E, Pinna G (2008) Decreased corticolimbic J Med Chem 34:3399–3402. allopregnanolone expression during social isolation enhances contextual fear: A 23. Ashani Y, Peggins JO, III, Doctor BP (1992) Mechanism of inhibition of cholinesterases model relevant for posttraumatic stress disorder. Proc Natl Acad Sci USA 105:5567– by huperzine A. Biochem Biophys Res Commun 184:719–726. 5572. 24. Tonduli LS, Testylier G, Masqueliez C, Lallement G, Monmaur P (2001) Effects of huperzine used as pretreatment against soman-induced seizures. Neurotoxicology 40. Blanchard RJ, Blanchard DC (1969) Passive and active reactions to fear-eliciting stimuli. 22:29–37. J Comp Physiol Psychol 68:129–135. 25. Guidotti A, et al. (2005) GABAergic dysfunction in schizophrenia: New treatment 41. Pinna G, Galici R, Schneider HH, Stephens DN, Turski L (1997) Alprazolam dependence strategies on the horizon. Psychopharmacology (Bere) 180:191–205. prevented by substituting with the ␤-carboline abecarnil. Proc Natl Acad Sci USA 26. Giusti P, et al. (1993) Imidazenil: A new partial positive allosteric modulator of 94:2719–2723. ␥-aminobutyric acid (GABA) action at GABAA receptors. J Pharmacol Exp Ther 42. Racine RJ (1972) Modification of seizure activity by electrical stimulation. II. Motor 266:1018–1028. seizure. Electroencephal Clin Neurophysiol 32:281–294.

14174 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807172105 Pibiri et al. Downloaded by guest on September 30, 2021