Enhanced efficacy of anticonvulsants when combined with levetiracetam in soman-exposed rats

NeuroToxicology 32 (2011) 923–930

Contents lists available at ScienceDirect

NeuroToxicology

Enhanced efficacy of anticonvulsants when combined with levetiracetam in soman-exposed rats Trond Myhrer a,*, Siri Enger a, Morten Jonassen a,b, Pa˚l Aas a a b

Norwegian Defence Research Establishment, Protection Division, P.O. Box 25, NO-2027 Kjeller, Norway Norwegian Defence Medical Services, NO-2058 Sessvollmoen, Norway

A R T I C L E I N F O

A B S T R A C T

Article history: Received 24 February 2011 Accepted 25 April 2011 Available online 6 May 2011

Results from studies based on microinfusions into seizure controlling brain sites (area tempestas, medial septum, perirhinal cortex, posterior piriform cortex) have shown that procyclidine, muscimol, caramiphen, and NBQX, but not ketamine, exert anticonvulsant effects against soman-induced seizures. The purpose of the present study was to examine whether levetiracetam (Keppra1) may enhance the anticonvulsant potency of the above drugs to become optimally effective when used systemically. Levetiracetam has a unique profile in preclinical models of epilepsy and has been shown to increase the potency of other antiepileptic drugs. The rats were pretreated with pyridostigmine (0.1 mg/kg) to enhance survival and received anticonvulsants 20 min after onset of seizures evoked by soman (1.15  LD50). The results showed that no single drug was able to terminate seizure activity. However, when levetiracetam (LEV; 50 mg/kg) was combined with either procyclidine (PCD; 10 mg/kg) or caramiphen (CMP; 10 mg/kg) complete cessation of seizures was achieved, but the nicotinic antagonist mecamylamine was needed to induce full motor rest in some rats. In a subsequent experiment, rats were pretreated with HI-6 (125 mg/kg) to enhance survival and treatment started 40 min following seizure onset of a soman dose of 1.6  LD50. LEV (50 mg/kg) combined with either PCD (20 mg/kg) or CMP (20 mg/kg) terminated seizure activity, but the survival rate was considerably higher for LEV + PCD than LEV + CMP. Both therapies could also save the lives of rats that were about to die 5–10 min after seizure onset. Thus, the combination of LEV and PCD or CMP may make up a model of a future autoinjector being effective regardless of the time of application. ß 2011 Elsevier Inc. All rights reserved.

Keywords: Soman intoxication Seizures Treatment Levetiracetam Procyclidine Muscimol Caramiphen NBQX Ketamine

1. Introduction Organophosphorus nerve agents are lethal chemical warfare means that may be encountered during military combats, terrorist use, or during chemical disarmament. Nerve agents act by irreversibly inhibiting acetylcholinesterase, the enzyme that hydrolyzes acetylcholine. Accumulation of acetylcholine results in excessive stimulation of muscarinic and nicotinic receptors. The signs of poisoning are seen as increased salivation, respiratory distress, tremor, seizures/convulsions, coma, and death (Taylor, 2001). Increased cholinergic activity in the brain is probably related to the initial phase of seizures, whereas sustained seizures are probably associated with increased glutamatergic activity leading to neuronal damage (McDonough and Shih, 1997). Medical management of nerve agent poisoning is based on pretreatment with a carbamate cholinesterase inhibitor (pyridostigmine) to shield a fraction of the cholinesterase from irreversible

* Corresponding author. Tel.: +47 63 80 78 52; fax: +47 63 80 75 09. E-mail address: trond.myhrer@ffi.no (T. Myhrer). 0161-813X/$ – see front matter ß 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.neuro.2011.04.008

inhibition by the nerve agent. Treatment after exposure to nerve agent is based on a cholinergic antagonist (atropine sulfate) along with an oxime (obidoxime, 2-PAM, HI-6) to reactivate any unaged inhibited enzyme (Aas, 2003). Such treatment regimen can increase the survival rate significantly, but it does not effectively reduce nerve agent-induced seizure activity resulting in brain injury. For this reason, efforts in search for effective countermeasures have aimed at drugs exerting cholinergic and glutamatergic antagonism along with GABAergic agonism (McDonough and Shih, 1997). However, determination of critical receptor subtypes would provide clues for the designing of more specific anticonvulsive therapeutic strategies as it has been made in epilepsy research. Within the rat brain, there are control mechanisms capable of attenuating all aspects of convulsive activity. Several target areas have been identified, and it is assumed that the ability of a systemically administered drug to confer seizure protection depends on the drug’s relative impact on the defined action sites (Gale, 1988). A series of recent lesion and microinfusion studies have been devoted to the identification of control sites for soman-induced seizure activity and specification of critical receptors for pharma-

924

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

cological management. It has been shown that lesion of the area tempestas, medial septum, perirhinal cortex, or posterior piriform cortex produces anticonvulsant effects (prevention of convulsions or delayed onset of convulsions) in rats exposed to soman, whereas damage to nucleus accumbens, nucleus basalis magnocellularis, amygdala, hippocampus, or entorhinal cortex does not cause anticonvulsant impact (Myhrer et al., 2007, 2008a). These results are in compliance with findings that seizures can be generated in area tempestas, medial septum, perirhinal cortex, and posterior piriform cortex by means of nerve agents, chemoconvulsants, or kindling (Myhrer, 2010). Results from microinfusion studies show that anticonvulsant efficacy is obtained by GABAA modulators (muscimol, ethanol, propofol) or cholinergic antagonists (M1–M5) (atropine, scopolamine, caramiphen, procyclidine) in area tempestas (Myhrer et al., 2006a, 2008b), cholinergic antagonists (M1– M5) (atropine, scopolamine, procyclidine) in medial septum (Myhrer et al., 2009), combined glutamatergic (NMDA) and cholinergic antagonist (M1–M4) (procyclidine), AMPA antagonist (NBQX), or modulators of metabotropic glutamate receptors (mGluR5, mGluR2/3) (MPEP, DCG-IV) in the perirhinal cortex (Myhrer et al., 2010a,b), and cholinergic antagonist (M1–M5) (scopolamine) or GABAA agonist (muscimol) in the posterior piriform cortex (Myhrer et al., 2010a). Calculation of impact factors for the most potent drugs (percentage of positive effects in the seizure controlling sites) showed that scopolamine and procyclidine were ranking highest (75) followed by muscimol (50), NBQX (33), and caramiphen (33), whereas the impact factor of ketamine was 0 (Myhrer, 2010). When the percentage of nonconvulsing rats from both lesion and microinfusion studies is used as guidance for selecting the most influential seizure controlling brain sites, the area tempestas and perirhinal cortex stand out as the most prominent ones. Procyclidine exerts anticonvulsant efficacy in both areas (high impact), whereas scopolamine, NBQX, caramiphen, and muscimol yield anticonvulsant effects in 1 of the areas only (low impact) (Myhrer, 2010). The high anticonvulsant potency of scopolamine will have limited use as a post-exposure agent, because the comparatively narrow cholinergic window of the 3-phase model (about 5 min after seizure onset) makes the efficacy of anticholinergics gradually weaker with elapse of time since seizure onset (McDonough and Shih, 1997). On the other hand, procyclidine (impact factor 75) has proved very useful as a post-exposure means when combined with either muscimol, ethanol, or propofol. However, these GABAergic modulators can depress respiratory function, and about 17% of the rats died within 24 h with each treatment regimen (Myhrer et al., 2006b). To avoid adverse effects on the brainstem, enhancement of procyclidine’s anticonvulsant impact on the seizure controlling sites of the forebrain would make up a novel and interesting approach. Levetiracetam with a unique profile in preclinical models of epilepsy has been shown to increase the potency of other antiepileptic drugs up to 19-fold (Kaminski et al., 2009). The purpose of the present study was to make a comparative assessment of anticonvulsant effects of levetiracetam, procyclidine, muscimol, caramiphen, NBQX, and ketamine or each drug in combination with levetiracetam 20 min after onset of somanevoked seizures (experiment 1). In experiment 2, levetiracetam was combined with either procyclidine or caramiphen and the treatment started 40 min following seizure onset. In attempt to prolong survival, the rats were pretreated with pyridostigmine in experiment 1 and with HI-6 in experiment 2. A single dose of each anticonvulsant previously shown to cause optimal efficacy was applied. One exception was the use of 2 doses with procyclidine and caramiphen in combination with levetiracetam in experiment 1 to examine impact of nicotinic antagonism on motor dysfunctions after seizure termination. Additionally, mecamylamine was

given as adjunct in both experiments. It was also of interest to investigate whether the most potent combination(s) with levetiracetam may prevent onset of seizure activity (experiment 2). 2. Materials and methods 2.1. Animals Male Wistar rats from a commercial supplier (Taconic Breeding Laboratories, Denmark) weighing 300–330 g served as subjects. The experiments were approved by the National Animal Research Authority. The animals were housed individually and had free access to commercial rat pellets and water. The rats were handled individually 3 days preoperatively and 3 days postoperatively, being allowed to explore a table top (80 cm  60 cm) for 3 min a day. The climatized vivarium (21 8C) was illuminated from 0700 to 1900 h. 2.2. Surgery The rats were anesthetized ip with diazepam (4.5 mg/kg) and fentanyl fluanisone (2 mg/kg). Of 2 stainless screws, one was lowered 1 mm into the parietal cortex, and the contralateral one served as ground. The screws were fixed with dental cement (Durelon; ESPE, Seefeldt, Germany). The electrodes were connected with the polygraph (Grass Model 79E) with alligator clips and leads. The use of a swivel allowed the rats to move freely. The rats were given a recovery period of 7 days. 2.3. Drugs The drug doses chosen were derived from previous studies of anticonvulsant effects against soman-evoked seizures in rats; procyclidine hydrochloride 10 mg/kg, caramiphen edisylate 10 mg/kg, muscimol hydrobromide 20 mg/kg, NBQX 40 mg/kg, ketamine hydrochloride 50 mg/kg (Lallement et al., 1994; Myhrer et al., 2006b; Raveh et al., 2003; Shih et al., 1999) or antiepileptic effects in preclinical epilepsy models; levetiracetam 50 mg/kg (Kaminski et al., 2009). HI-6 dimethanesulphonate (125 mg/kg) was used to reduce death, since this oxime has been demonstrated to cause a survival rate of 60% 24 h after a soman dose of 1.6  LD50 (Shih et al., 1991). The carbamate pyridostigmine bromide (0.026 mg/kg) has also been used to prolong survival among rats (Scremin et al., 1997). Results from our pilot experiment showed that when the soman dose is 1.15  LD50, the survival rate 20 min following seizure onset is 50%. This percentage is increased to 75% when pyridostigmine (0.1 mg/kg) is administered 20 min before soman exposure. Mecamylamine (10 mg/kg) (Shih et al., 1991) was given to rats that displayed nicotinic motor dysfunctions after termination of seizures. The soman doses were either 1.15  LD50 (92 mg/kg) resulting in convulsions and death in 85% of the rats or 1.6  LD50 (128 mg/kg) resulting in convulsions and death in all rats of our strain (Sterri et al., 1980). When using a prophylactic regimen, the soman dose was 1.3  LD50 (100 mg/kg) to ensure that all rats would convulse without appropriate pretreatment. Soman was given sc. All drugs were purchased from Sigma (St Louis, MO, USA), except HI-6 that was a gift from Defence Research and Development (Canada). The drugs were dissolved in 0.9% saline, and control rats in experiment 2 received 0.3 ml of saline (0.9%)  2 as treatment. The drugs were administered ip. 2.4. Experimental design For overview, see Fig. 1. Experiment 1. (A) Pyridostigmine given 20 min before soman (1.15  LD50). Testing anticonvulsant effect of each drug (levetiracetam, procyclidine, caramiphen, muscimol,

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

925

in order to see both simultaneously. A digital microscope camera (AxioCam, Zeiss, Jena, Germany) was used to make photomicrographs. This technique allows processing of the photographs so that elements of particular interest can be made clearer by adjusting contrasts. 2.6. Evaluation of neuropathology A grading system of 0–4 previously described (McDonough et al., 1995), was used to determine severity of neuronal damage in the hippocampal CA1 region, the basolateral amygdala, and the piriform cortex, based on the approximate percentage of tissue involvement: 0, no lesion; 1, minimal, 1–10%; 2, mild, 11–25%; 3, moderate, 26–45%; 4, severe > 45%. Each animal was given a total neuropathology score expressed as a mean of the 3 individual brain areas chosen. The criterion used to characterize the pathology was neuronal degeneration. Fig. 1. Schematic overview of the experimental designs. All drugs were initially examined alone in experiment 1. Then each drug was combined with levetiracetam (not shown). All drugs (except levetiracetam) were derived from the results of microinfusion studies (see Section 1).

NBQX, ketamine) 20 min after onset of seizures. (B) Pyridostigmine given 20 min before soman (1.15  LD50). Testing anticonvulsant effects of the above drugs (2 doses of procyclidine and caramiphen) combined with levetiracetam 20 min after onset of seizures. When needed, mecamylamine given as adjunct. Experiment 2. (A) HI-6 given 20 min before soman (1.6  LD50). Testing anticonvulsant effects of levetiracetam combined with procyclidine or caramiphen (1 dose) 40 min after onset of seizures. When needed, mecamylamine given as adjunct. (B) Levetiracetam combined with procyclidine or caramiphen as prophylactic regimens 20 min before soman (1.3  LD50). 2.5. Histology The rats were anesthetized as described for surgery, perfused intracardially with 10% formalin, and the brains post-fixed in 10% formalin for at least 24 h. The brains were dehydrated and embedded in paraffin (Schmued et al., 1997). The sections were cut 5 mm thick and dried in an incubator (37 8C) for 12 h before they were stained with hematoxylin and eosin (HE) or Fluoro-Jade B (Schmued et al., 2000). Since Fluoro-Jade staining requires perfusion of the brain, only live rats could be used for this purpose. Rats that recently died or rats about to die were decapited, and the brain sections were stained with HE. Because Fluoro-Jade has been considered to be the compound most suitable for the detection of neuronal degeneration (Schmued et al., 1997), this fluorescent staining technique was used to supplement the more conventional HE staining technique. A degenerating neuron presumably expresses a strong basic molecule, since it has an affinity for the strongly acidic Fluoro-Jade (Schmued et al., 1997). The Fluoro-Jade method had previously been described in detail (Schmued et al., 1997, 2000). In order to make a distinct contrast between degenerated neurons and intact ones the sections were co-stained with 40 ,6-diamidino-2-phenylindole dihydrochloride (DAPI) resulting in a blue fluorescence of cellular nuclei. Such nuclear staining is seen in all viable cells. A 0.01% stock solution of DAPI (10 mg/100 ml distilled water) was prepared and 2 ml of this stock solution was added to 98 ml of the Fluoro-Jade B staining solution. Blue counterstained normal cell nuclei can be visualized when excited by ultraviolet (330–380 nm) light (Schmued et al., 2000). Fluoro-Jade B staining is seen with blue excitation filter, whereas DAPI is visualized by filter set 49 with excitation 365 nm and emission 445/50. One picture was superimposed on the other

2.7. EEG Seizure activity was defined as terminated when epileptiform waves had ceased (absence of continuous high amplitude rhythmic spike or sharp wave activity). EEG recording was made while the animals were situated in their home cages (50 cm  30 cm  15 cm). Measures were made 24 h prior to drug treatment, immediately after, and 24 or 48 h after treatment. 2.8. Observation of animals Each rat was observed for overt behavioral changes and signs of intoxication. Dampness of the lips and nose was interpreted as indication of hyper-rhinorrhea and -salivation. Unconsciousness was determined by loss of both righting and corneal reflexes. The rats were observed for convulsions and visible signs of intoxication continuously for the first 2–3 h and then 10 min at 24 and 48 h after soman injection. Rats that displayed aphagia/adipsia following termination of convulsions were given 5 ml of saline (ip) twice a day for 1–3 days until they started to eat. 2.9. Statistics Overall analyses were carried out by using parametric or nonparametric one-way analysis of variance (ANOVA). Group comparisons were made with Newman–Keuls or Dunn’s post hoc test, two-tailed t test, or with two-sided Fisher’s exact test. Use of the grading system of neuropathology resulted in nonparametric data. Computations were made with the Prism statistical software program (GraphPad Software, Sand Diego, CA, USA). 3. Results 3.1. Seizures and drugs 3.1.1. Experiment 1 The mean latency to seizure onset was between 7 and 9 min in all groups (Table 1). The seizures/convulsions were allowed to last for 20 min before treatment started. As seen from Table 1, no single drug was effective in terminating the seizure activity. The rats either died spontaneously or they were decapitated within 30 min after drug injection. The proportion of animals that died spontaneously versus those that were euthanized was similar among the groups. Based on previous observations rats challenged with 1.15  LD50 die within 2 h. Table 2 shows the results when the drugs were given in combination with levetiracetam. Anticonvulsant efficacy was only obtained when levetiracetam was combined with either procyclidine or caramiphen (both doses of the latter

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

926

Table 1 Anticonvulsant treatment of soman-induced (1.15  LD50) seizures 20 min after onset with single drugs in experiment 1. The rats were pretreated with pyridostigmine (0.1 mg/kg). Mean latencies in min (SEM). Drug

Dose (mg/kg)

N

Latency to seizure onset

Antiseizure response ratio

Levetiracetam Procyclidine Caramiphen Muscimol NBQX Ketamine

50 10 10 20 40 50

6 6 6 6 6 6

7.7  2.1 9.7  1.8 6.9  1.6 7.4  1.3 7.2  1.5 9.2  2.2

0/6 0/6 0/6 0/6 0/6 0/6

Table 2 Anticonvulsant treatment of soman-induced (1.15  LD50) seizures 20 min after onset with combination of drugs in experiment 1. The rats were pretreated with pyridostigmine (0.1 mg/kg). Mean latencies in min (SEM). Drug

Dose (mg/kg)

Levetiracetam Plus Procyclidine Levetiracetam Plus Procyclidine Levetiracetam Plus Caramiphen Levetiracetam Plus Caramiphen Levetiracetam Plus Muscimol Levetiracetam Plus NBQX Levetiracetam Plus Ketamine

50

*

N

Latency to seizure onset

Antiseizure response ratio

Latency to seizure termination

Lethality response ratio (48 h)

6

9.5  1.7

6/6

9.2  2.2

2/6

6

9.0  1.6

6/6

11.5  1.9

0/6*

6

8.7  1.5

6/6

12.0  1.7

3/6

6

10.2  2.2

6/6

11.5  1.2

3/6

6

8.3  1.4

0/6

–

6/6

6

9.2  1.7

0/6

–

6/6

6

9.8  2.1

0/6

–

6/6

10 50 20 50 10 50 20 50 20 50 40 50 50

Significantly different from the levetiracetam plus ketamine rats (all died spontaneously) with Fisher’s exact test; P = 0.0022.

drugs). ANOVA did not reveal significant overall effect for latency to seizure termination (F(3,20) = 0.5032, P = 0.6844). Epileptiform cortical EEG was usually seen simultaneously with convulsions or prior to convulsions. When convulsions ceased, the epileptiform activity vanished, and apparently normal cortical activity was observed (data not shown). Normal EEG was also seen 24 or 48 h following soman exposure and drug treatment. The lethality rate for the group with levetiracetam combined with the highest dose of procyclidine (20 mg/kg) was significantly lower than for the levetiracetam + ketamine group in which all rats died within 30 min after treatment (Fisher’s exact test, P = 0.0022). The combination of levetiracetam and muscimol or NBQX produced evident anticonvulsant effects, since the body of the rats temporarily became less tense, and the epileptiform activity subsided. However, the convulsions and seizure activity recurred, and the rats died spontaneously or were decapitated. In the group that received levetiracetam and ketamine, no anticonvulsant effect was observed, and death followed within 3 min after injections in 4 rats. The remaining 2 rats in the group died within 30 min. In rats that were about to die from the soman intoxication before they reached the criterion of 20 min for treatment, we administered the combination of levetiracetam and procyclidine or caramiphen (both doses). The latter combinations were given to 5 rats 5–10 min following seizure onset, and they all recovered completely without additional treatment. When seizure activity was terminated by either combination therapies with 10 mg/kg (Table 2), the rats regained consciousness after 3–5 min, and some rats exhibited nicotinic motor dysfunctions (intentional tremor/ jerks) when attempting to walk. In rats with these symptoms (50%), administration of the nicotinic antagonist mecamylamine

(10 mg/kg) resulted in complete motor rest. Since both procyclidine and caramiphen exert nicotinic antagonism, attempt was made to counteract the nicotinic dysfunction by doubling the doses of these drugs (Table 2). However, the use of 20 mg/kg did not induce motor rest, and mecamylamine had to be given as adjunct. On the other hand, the increased dose of procyclidine resulted in survival of all rats in this group (Table 2). Rats treated with levetiracetam and procyclidine or levetiracetam and caramiphen displayed similar loss of body weight on the first and second day after soman intoxication (6.4% and 3.7%, respectively). 3.1.2. Experiment 2 The mean latency to seizure onset was between 5 and 7 min (Table 3). The seizures/convulsions were allowed to last for 40 min before treatment started. Both combination therapies terminated seizures in all rats. The use of t test did not reveal a significant difference in latency to seizure termination between the groups (t (df 10) = 1.597, P = 0.1414). The survival rate was reliably higher for the levtiracetam plus procyclidine group than the levetiracetam plus caramiphen group (Fisher’s exact test, P = 0.0022). All rats were given mecamylamine to stop nicotinic motor dysfunctions. Rats that were about to die 5–10 min after seizure onset (N = 4) survived when treated with either combination therapy. The rats treated with levetiracetam and procyclidine displayed loss of body weight on the first and second day following soman exposure (10.6% and 16.2%, respectively). The group that received levetiracetam and caramiphen had a weight loss of 10.2% on the first day after intoxication and were not alive on the second day.

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

927

Table 3 Anticonvulsant treatment of soman-induced (1.6  LD50) seizures 40 min after onset with combination of drugs in experiment 2. The rats were pretreated with HI-6 (125 mg/ kg). Mean latencies in min (SEM). Drug

Dose (mg/kg)

Levetiracetam Plus Procyclidine Levetiracetam Plus Caramiphen

50

*

N

Latency to seizure onset

Antiseizure response ratio

Latency to seizure termination

Lethality response ratio (48 h)

6

7.7  1.1

6/6

11.8  1.8

0/6*

6

5.6  0.4

6/6

17.8  3.3

6/6

20 50 20

Significantly different from the levetiracetam plus caramiphen group with Fisher’s exact test; P = 0.0022.

Table 4 Protective anticonvulsant treatment given 20 min before soman intoxication (1.3  LD50) in experiment 2. Group

Dose (mg/kg)

N

Antiseizure response ratio

Lethality response ratio (24 h)

Control Levetiracetam Plus Procyclidine Levetiracetam Plus Caramiphen

– 50

5

0/5

5/5

*

5

4/5

5

2/5

*

2/5

20 50 5/5

20

Significantly different from the saline-treated control group with Fisher’s exact test; P = 0.0476.

Table 5 Median (range) neuropathology scores based on HE staining for rats treated with anticonvulsants (both doses of procyclidine and caramiphen) 20 min after soman-induced seizures in experiment 1. N

Group

Levetiracetam Plus Procyclidine Levetiracetam Plus Caramiphen

Neuropathology score Piriform cortex

Hippocampal CA1

Amygdala

10

1.5 (0–2)

0.0 (0–0)

0.5 (0–2)

6

0.5 (0–1)

0.5 (0–1)

1.0 (0–3)

Prophylactic treatment produced a significantly higher antiseizure response rate for the levetiracetam and procyclidine group than for the control group (Fisher’s exact test, P = 0.0476), whereas this was not the case for the levetiracetam and caramiphen group (Table 4). However, the rats pretreated with the combination therapies showed great problems in moving even 24 h after soman exposure. For this reason, the 3 surviving rats by 24 h were anesthetized and perfused. 3.2. Histology 3.2.1. Experiment 1 In rats that convulsed for about 30 min, only minor morphological changes were observed (Table 5). To include all brains the Table 6 Median (range) neuropathology scores based on HE staining for rats treated with anticonvulsants 40 min after soman-induced seizures in experiment 2. Group

Levetiracetam Plus Procyclidine Levetiracetam Plus Caramiphen *

N

Neuropathology score Piriform cortex

Hippocampal CA1

Amygdala

6

4.0 (4–4)

1.5* (0–4)

3.5 (3–4)

2

4.0 (4–4)

0.0 (0–0)

3.5 (3–4)

Significantly different from the piriform cortex; P < 0.01.

neuropathology evaluations were based on HE stained sections, because some brains could not be perfused. ANOVA revealed no significant overall effect in neuropathology across brain regions in none of the groups (P > 0.05). Most brains appeared undamaged in the index areas used (Fig. 2A). 3.2.2. Experiment 2 Rats that convulsed for about 55 min had marked neuropathology in the piriform cortex and amygdala, but not the hippocampal CA1 (Table 6). For the levetiracetam and procyclidine group ANOVA revealed a reliable treatment effect across brain areas (H(3) = 10.22, P = 0.0060). Multiple comparisons showed that CA1 had a significantly lower neuropathology score than the piriform cortex (P < 0.01), but not the amygdala (P > 0.05). In the levetiracetam plus caramiphen group, 2 rats that were about to die by 24 h were anesthetized and perfused. Some rats included in Table 6 displayed intact CA1 area (Fig. 2B). In 3 rats that received levetiracetam and procyclidine as prophylactic treatment, no neuropathology was seen in the perirhinal cortex, CA1, or amygdala in sections stained with DAPI/Fluoro-Jade B (survival time 24 h after soman exposure). 4. Discussion The results from experiment 1 showed that no single drug (levetiracetam, procyclidine, caramiphen, muscimol, NBQX, ketamine) was able to stop soman-induced seizures when administered 20 min after onset in rats pretreated with pyridostigmine.

928

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

Fig. 2. Sections stained with DAPI/Fluoro-Jade B (48 h following soman exposure) show undamaged tissue from a rat that had been treated with levetiracetam and procyclidine and convulsed for 30 min in experiment 1 (A). Neuropathology in the piriform cortex and amygdala, but not the hippocampal CA1 is seen in a rat that had been treated with levetiracetam and procyclidine and convulsed for 55 min in experiment 2 (B). The magnification was 100.

However, when levetiracetam was combined with either procyclidine or caramiphen the seizure activity was terminated. When regaining consciousness after seizure termination, some rats displayed nicotinic motor dysfunctions that were effectively treated with mecamylamine. In experiment 2, the combination of levetiracetam and procyclidine/caramiphen terminated seizure activity when administered 40 min following onset in rats pretreated with HI-6. However, levetiracetam and procyclidine resulted in a considerable higher rate of survival than levetiracetam and caramiphen. Prophylactic use of the combination therapies 20 min before exposure to soman prevented onset of seizures more effectively when levetiracetam was combined with procyclidne than caramiphen. In both experiments, it was observed that the combination therapies could save the lives of rats that were about to die 5–10 min after seizure onset. Anticonvulsant properties of GABAA modulators revealed by prophylactic microinfusions into seizure controlling brain sites (area tempestas, substantia nigra) have been shown to apply to systemic administration following seizure onset (Myhrer et al., 2006a,b). However, the anticonvulsant impact of drugs given well after onset of seizures is markedly reduced (Carpentier et al., 2001; Lallement et al., 1999). For this reason, combinations of anticonvulsant drugs are usually needed. Most of the presently used drugs emerged as the highest ranking anticonvulsants in several microinfusion studies (cf., Section 1). Of these drugs, only procyclidine and caramiphen ensured evident anticonvulsant efficacy in combination with levetiracetam. Procyclidine and caramiphen belong to a group of antiparkinson agents known to exert both cholinergic and glutamatergic antagonism in rodents (Gao et al., 1998; McDonough and Shih, 1995; Raveh et al., 2002). These unique properties reduce the need for multiple drug therapy in organophosphate intoxications. Procyclidine inhibits the phencyclidine site at the NMDA receptor very potently (Reynolds and Miller, 1988) in a concentration-dependent manner (Myhrer et al., 2004), whereas caramiphen appears to bind to the Zn2+ site at the NMDA receptor (Raveh et al., 1999) and additionally has blocking effect on the AMPA receptor (Raveh et al., 2002). Procyclidine and caramiphen differ in their capabilities to

antagonize a lethal dose of NMDA in mice. Procyclidine at a dose of 30 mg/kg protects 80% of NMDA-injected mice, whereas caramiphen doses of 10–80 mg/kg protect only 11–40% of the mice (Raveh et al., 1999). This difference in NMDA antagonism may explain why rats treated with levetiracetam and procyclidine have markedly higher survival rate by 48 h than rats that received levetiracetam and caramiphen (Tables 2 and 3). Both procyclidine and caramiphen have been demonstrated to be potent prophylactics when combined with physostigmine (Myhrer et al., 2004; Raveh et al., 2002). When given in combination with levetiracetam, procyclidine appears to have more powerful prophylactic effect than caramiphen (Table 4). However, the rats pretreated with the combination of levetiracetam and procyclidine or caramiphen were incapacitated even 24 h after intoxication, presumably because of excessive cholinergic activity. A crucial issue for the recovery process is that a fraction of acetylcholinesterase must be reversibly inhibited by a carbamate before exposure. During decarbamylation the enzyme subsequently becomes active again as a function of time (Harris et al., 1984). Although full protection against neuronal damage was achieved with levetiracetam and procyclidine, the present combination therapies do not seem to be well suited for prophylactic use, but they can probably work in emergency situations. Although different drugs for pretreatment and different doses of soman for initiating seizures were used in experiments 1 and 2, the degree of neuropathology and loss of body weight appears to be associated with the different length of convulsing time (about 30 min versus 55 min) in the 2 experiments. It has, however, been reported that rate of weight loss does not predict degree of neuropathology 24 h after soman exposure, but high level of EEG delta activity by 24 h predicts severity of brain lesions (McDonough et al., 1998). The combination of levetiracetam and ketamine may have lethal effect in rats exposed to soman. In 4 of 6 rats, death followed within 3 min after injections, and 2 rats died within 30 min. Ketamine given as adjunct therapy to rats exposed to a convulsant dose of soman turns out to be lethal in high doses (McDonough

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

et al., 2010). These findings appear intriguing, because ketamine is supposed to exert no or little respiratory depression (Annetta et al., 2005). Levetiracetam (Keppra1) is an antiepileptic drug that strongly enhances the anticonvulsant effects of compounds affecting either glutamatergic or GABAergic neurotransmission (Kaminski et al., 2009). The distinctive binding site of levetiracetam appears to be the synaptic vesicle protein 2A (SV2A) (Lynch et al., 2004). Although the exact mechanisms are not well known, levetiracetam probably reduces release of glutamate by which the effects of glutamatergic antagonists are highly increased (Kaminski et al., 2009). Levetiracetam does not have affinity to SV2A for GABAA receptors, but it seems to affect postsynaptic mechanisms resulting in enhanced GABAergic neurotransmission (Kaminski et al., 2009). Effects of levetiracetam on cholinergic synapses do not seem to have received much attention in experimental epilepsy. It has, however, been reported that levetiracetam may both reduce release of acetylcholine and reduce postsynaptic responsiveness in cholinergic synapses (Oliveira et al., 2005). In our laboratory, we have not observed increased cholinergic antagonism of procyclidine when it was combined with levetiracetam in microinfusions into the area tempestas of rats exposed to soman (Myhrer, Enger, and Jonassen, unpublished data). It has been reported that drugs affecting GABAA or AMPA receptors can have their potencies increased by up to 16-fold and 19-fold, respectively when combined with levetiracetam (Kaminski et al., 2009). Even if levetiracetam potentiated the anticonvulsant efficacy of muscimol and NBQX, it was not measurable in the present study. Drugs with a single mechanism of action seem to have relatively weak anticonvulsant efficacy when used well after onset of seizures, unless a potent pretreatment is administered (McDonough and Shih, 1997). Termination of soman-induced seizures by trihexyphenidyl (same group of antiparkinson drugs as procyclidine and caramiphen) 20 or 40 min after onset resembles the anticonvulsant effects provided by the combination of scopolamine and MK-801 at the same times. The latter drugs are without effects when administered separately. The anticonvulsant activity of trihexyphenidyl against somanevoked seizures at the longer seizure durations may be ascribed to mixed anticholinergic and NMDA antagonistic properties of this drug (McDonough and Shih, 1993). Likewise, the combination of atropine and ketamine has anticonvulsant impact when administered 30–120 min after soman intoxication (Dorandeu et al., 2005). NMDA antagonists can produce lethal interactions on the respiration function in soman poisoned subjects in absence of anticholinergic drugs (McDonough and Shih, 1993). The multifunction of procyclidine and caramiphen may explain why these drugs proved efficacious in combination with levetiracetam, even if the cholinergic antagonism of procyclidine and caramiphen might not have been potentiated by levetiracetam. Without any pretreatment, the survival time following a convulsant dose of soman will be markedly limited. Pyridostigmine does not readily pass the blood–brain barrier because of its quaternary amine structure. Not even under stress does the blood– brain barrier seem to become more permeable for pyridostigmine in rats (Amourette et al., 2009). Pyridostigmine may have central effects by way of peripheral pathways like cholinergic transmission at autonomic ganglia, cholinergic terminals in adrenal medulla, and vagal cholinergic afferents (Amourette et al., 2009). In order to have a reasonable number of rats to survive for 40 min after seizure onset, HI-6 was chosen as pretreatment in experiment 2. However, in addition to act as a reactivator of inhibited acetylcholinesterase, HI-6 appears to have several pharmacological effects. HI-6 has been shown to cause recovery of neuronal transmission in the respiratory center possibly by GABAergic mechanisms (Van Helden et al., 1996). Furthermore, HI-

929

6 produces a marked reduction in the evoked release of acetylcholine following challenge with soman in hippocampal slices (Øydvin et al., 2005). Thus, the anticonvulsant efficacy obtained by the combination of levetiracetam and procyclidine/ caramiphen in rats pretreated with pyridostigmine may reflect a genuine effect of the anticonvulsants unsupported by the pretreatment. When HI-6 is used as pretreatment, it cannot be ruled out that the anticonvulsant efficacy observed might have been confounded by auxiliary effects of HI-6. In rats pretreated with HI-6, soman-elicited seizures can be terminated 40 min after onset by a single drug (McDonough and Shih, 1993; Shih et al., 1999), whereas a combination of drugs is required for termination of seizures 30 min following onset in soman exposed rats that did not receive pretreatment (Myhrer et al., 2003, 2006b). It has been a matter of debate whether the content of quaternary nitrogen atoms allows HI-6 to penetrate the blood–brain barrier. However, some studies conclude that at least a certain concentration of HI-6 enters the central nervous system after systemic administration of the oxime (Cassel and Fosbraey, 1996; Ligtenstein and Kossen, 1983). Mecamylamine is a nicotinic receptor antagonist that readily crosses the blood–brain barrier (Shytle et al., 2002). The prominent nicotinic signs of nerve agent intoxication such as intentional tremor and jerks when attempting to walk following seizure termination (McDonough and Shih, 1993) were effectively treated with mecamylamine. This drug has additionally antihypertensive effect (Shytle et al., 2002) and might therefore be beneficial for rats with elevated blood pressure caused by soman-induced seizures (Goldman et al., 1993). Furthermore, mecamylamine has been shown to exert potent NMDA antagonism in mice injected with a lethal dose of NMDA (McDonough and Shih, 1995). There is an urgent need for potent anticonvulsants suitable for an autoinjector being effective at any time of application by the military personnel themselves or civilian first responders. The combination of levetiracetam and procyclidine may make up a model of such autoinjector, since this combination turns out to have promising anticonvulsant impact without exerting hypnotic effects when administered 5–10, 20, or 40 min after seizure onset. However, the anticonvulsant efficacy is more moderate when this combination is used as pretreatment. In conclusion, levetiracetam potentiated the anticonvulsant properties of procyclidine and caramiphen sufficiently to terminate soman-evoked seizure activity. In combination with levetiracetam procyclidine ensured a markedly higher survival rate than caramiphen. The potent anticonvulsant capacity of procyclidine observed in the present study is in compliance with the high impact of this drug in seizure controlling sites in the forebrain as shown in microinfusion studies of rats (Table 4 in Myhrer, 2010).

Conflict of interest statement The authors declare that there are no conflicts of interest.

References Aas P. Future considerations for the medical management of nerve agent intoxication. Prehosp Disast Med 2003;18:208–16. Amourette C, Lamproglou I, Barbier L, Fauquette W, Zoppe A, Viret R, Diserbo M. Gulf War illness: effects of repeated stress and pyridostigmine treatment in blood– brain barrier permeability and cholinesterase activity in rat brain. Behav Brain Res 2009;203:207–14. Annetta MG, Iemma D, Garisto C, Tafani C, Proietti R. Ketamine: new indications for an old drug. Curr Drug Targets 2005;6:789–94. Carpentier P, Foquin A, Kamenka J-M, Rondouin G, Lerner-Natoli M, de Groot DMG, Lallement G. Effects of thienylphencyclidine (TCP) on seizure activity and brain damage produced by soman in guinea-pigs: ECoG correlates of neurotoxicity. Neurotoxicology 2001;22:13–28.

930

T. Myhrer et al. / NeuroToxicology 32 (2011) 923–930

Cassel GE, Fosbraey P. Measurement of the oxime HI-6 after peripheral administration in tandem with neurotransmitter levels in striatal dialysates: effects of soman intoxication. J Pharmacol Toxicol Methods 1996;35:159–66. Dorandeu F, Carpentier P, Baubichon D, Four E, Bernabe´ D, Burckhart M-F, Lallement G. Efficacy of ketamine–atropine combination in the delayed treatment of somaninduced status epilepticus. Brain Res 2005;1051:164–75. Gale K. Progression and generalization of seizure discharge: anatomical and neurochemical substrates. Epilepsia 1988;29(Suppl. 2):S15–34. Gao ZG, Liu BY, Cui WY, Li LJ, Fan QH, Liu CG. Anti-nicotinic properties of anticholinergic antiparkinson drugs. J Pharm Pharmacol 1998;50:1299–305. Goldman H, Berman RF, Hazlett J, Murphy S. Cerebrovascular responses to soman: time and dose dependent effects. Neurotoxicology 1993;14:469–84. Harris LW, McDonough JH Jr, Stitcher DL, Lennox WJ. Protection against both lethal and behavioral effects of soman. Drug Chem Toxicol 1984;7:605–24. Kaminski RM, Matagne A, Patsalos PN, Klitgaard H. Benefit of combination therapy in epilepsy: a review of preclinical evidence with levetiracetam. Epilepsia 2009;50:387–97. Lallement G, Pernot-Marino I, Foquin-Tarricone A, Baubichon D, Piras A, Blanchet G, Carpentier P. Antiepileptic effects of NBQX against soman-induced seizures. Neuroreport 1994;5:425–8. Lallement G, Baubichon D, Clarenc¸on D, Gallonier M, Peoc’h M, Carpentier P. Review of the value of gacyclidine (GK-11) as adjuvant medication to conventional treatments of organophosphate poisoning: primate experiments mimicking various scenarios of military of terrorist attack by soman. Neurotoxicology 1999;20:675–84. Ligtenstein DA, Kossen SP. Kinetic profile in blood and brain of the cholinesterase reactivity oxime HI-6 after intravenous administration to the rat. Toxicol Appl Pharamacol 1983;71:177–83. Lynch BA, Lambeng N, Nocka K, Kensel-Hammes P, Bajjalieh SM, Matagne A, Fuks B. The synaptic vesicle protein SV2A is the biding site for the antiepileptic drug levetiracetam. Proc Natl Acad Sci U S A 2004;101:9861–6. McDonough JH Jr, Shih T-M. Pharmacological modulation of soman-induced seizures. Neurosci Biobehav Rev 1993;17:203–15. McDonough JH Jr, Shih T-M. A study of the N-methyl-D-aspartate antagonistic properties of anticholinergic drugs. Pharmacol Biochem Behav 1995;51:249–53. McDonough JH Jr, Shih T-M. Neuropharmacological mechanisms of nerve agentinduced seizure and neuropathology. Neurosci Biobehav Rev 1997;21:559–79. McDonough JH Jr, Dochterman W, Smith CD, Shih T-M. Protection against nerve agentinduced neuropathology, but not cardiac pathology, is associated with the anticonvulsant action of drug treatment. Neurotoxicology 1995;15:123–32. McDonough JH Jr, Clark T, Slone TW Jr, Zoeffel D, Brown K, Kim S, Smith CD. Neural lesions in the rat and their relationship to EEG delta activity following seizures induced by the nerve agent soman. Neurotoxicology 1998;19:381–92. McDonough J, Van Shura K, Lyman M, Eisner C, Mazza A, Kan R, Shih T-M. Adjuncts for the delayed treatment of nerve agent-induced status epilepticus seizures. In: Proceedings of Bioscience Review, U S Army Medical Research. Hunt Valley, MD, May; 2010. Myhrer T. Identification of neuronal target areas for nerve agents and specification of receptors for pharmacological treatment. Neurotoxicology 2010;31:629–38. Myhrer T, Skymoen LR, Aas P. Pharmacological agents, hippocampal EEG, and anticonvulsant effects on soman-induced seizures in rats. Neurotoxicology 2003;24:357–67. Myhrer T, Nguyen NHT, Andersen JM, Aas P. Protection against soman-induced seizures: relationship among doses of prophylactics, soman, and adjuncts. Toxicol Appl Pharmacol 2004;19:327–36. Myhrer T, Nguyen NHT, Enger S, Aas P. Anticonvulsant effects of GABAergic modulators microinfused into area tempestas or substantia nigra in rats exposed to soman. Arch Toxicol 2006a;80:502–7. Myhrer T, Enger S, Aas P. Pharmacological therapies against soman-induced seizures in rats 30 min following onset and anticonvulsant impact. Eur J Pharmacol 2006b;548:83–9.

Myhrer T, Enger S, Aas P. Anticonvulsant effects of damage to structures involved in seizure induction in rats exposed to soman. Neurotoxicology 2007;28: 819–28. Myhrer T, Enger S, Aas P. Anticonvulsant impact of lesions in the ventrolateral forebrain of rats challenged with soman. Brain Res 2008a;1226:241–7. Myhrer T, Enger S, Aas P. Anticonvulsant efficacy of drugs with cholinergic and/or glutamatergic antagonism microinfused into area tempestas of rats exposed to soman. Neurochem Res 2008b;33:348–54. Myhrer T, Enger E, Aas P. Anticonvulsant efficacy of pharmacological agents microinfused into medial septum of rats exposed to soman. J Med CBR Defense 2009;7:1–14. Myhrer T, Enger S, Aas P. Roles of perirhinal and posterior piriform cortices in control and generation of seizures: a microinfusion study in rats exposed to soman. Neurotoxicology 2010a;31:147–53. Myhrer T, Enger S, Aas P. Modulators of metabotropic glutamate receptors microinfused into perirhinal cortex: anticonvulsant effects in rats challenged with soman. Eur J Pharmacol 2010b;636:82–7. Oliveira AA, Nogueira CRA, Nascimento VS, Aguiar LMV, Freitas RM, Sousa FCF, Viana GSB, Fonteles MMF. Evaluation of levetiracetam effects on pilocarpine-induced seizures: cholinergic muscarinic system involvement. Neurosci Lett 2005;385:184–8. Øydvin OK, Tansø R, Aas P. Pre-junctional effects of oximes on [3H]-acetylcholine release in rat hippocampal slices during soman intoxication. Eur J Pharmacol 2005;516:227–34. Raveh L, Chapman S, Cohen G, Alkalay D, Gilat E, Rabinovitz I, Weissman BA. The involvement of NMDA receptor complex in the protective effect of anticholinergic drugs against soman poisoning. Neurotoxicology 1999;20:551–60. Raveh L, Weissman BA, Cohen G, Alkalay D, Rabinovitz I, Sonego H, Brandeis R. Caramiphen and scopolamine prevent soman-induced brain damage and cognitive dysfunction. Neurotoxicology 2002;23:7–17. Raveh L, Brandeis R, Gilat E, Cohen G, Alkalay D, Rabinovitz I, Sonego H, Weissman BA. Anticholinergic and antiglutamatergic agents protect against soman-induced brain damage and cognitive dysfunction. Toxicol Sci 2003;75:108–16. Reynolds IJ, Miller RJ. [3H]MK-801 binding to the N-methyl-D-aspartate receptor reveals drug interaction with the zinc and magnesium binding sites. J Pharmacol Exp Ther 1988;24:1025–31. Schmued LC, Albertson C, Slikker W Jr. Fluoro-Jade: a novel fluorochrome for the sensitive and reliable histochemical localization of neuronal degeneration. Brain Res 1997;751:37–46. Schmued LC, Hopkins KJ, Fluoro-Jade B. A high affinity fluorescent marker for the localization of neuronal degeneration. Brain Res 2000;874:123–30. Scremin OU, Shih T-M, Li MG. Mapping of cortical metabolic activation in somaninduced convulsions in rats. Brain Res Bull 1997;43:425–34. Shih T-M, Koviak TA, Capacio BR. Anticonvulsants for poisoning by the organophosphorous compound soman: pharmacological mechanisms. Neurosci Biobehav Rev 1991;15:349–62. Shih T-M, McDonough JH Jr, Koplovitz I. Anticonvulsants for soman-induced seizure activity. J Biomed Sci 1999;6:86–96. Shytle RD, Penny E, Silver AA, Goldman J, Sanberg PR. Mecamylamine (Inversine1): an old antihypertensive with new research directions. J Hum Hypertens 2002;16:453–7. Sterri SH, Lyngaas S, Fonnum F. Toxicity of soman after repetitive injection of sublethal doses in rat. Acta Pharmacol Toxicol 1980;46:1–7. Taylor P. Anticholinesterase agents. In: Hardman JG, Limbird LE, Gilman AG, editors. The pharmacological basis of therapeutics 10th ed. New York: McGraw-Hill Companies Inc.; 2001, pp. 175–91. Van Helden HPM, Busker RW, Melchers BPC, Bruijnzeel PLB. Pharmacological effects of oximes: how relevant are they. Arch Toxicol 1996;70:779–86.