Chronic Treatment with Naltrexone Prevents Memory Retention Deficits in Rats Poisoned with the Sarin Analog Diisopropylfluorophosphate (DFP) and Treated with Atropine and Pralidoxime

Abstract

Humans and rats poisoned with sarin develop chronic neurological disabilities that are not prevented with standardized antidotal therapy. We hypothesized that rats poisoned with the sarin analogue diisopropylfluorophosphate (DFP) and resuscitated with atropine and pralidoxime would have long-term memory deficits that were preventable with naltrexone treatment. Long Evans rats (250–275 g) were randomized to: DFP (N = 8): single subcutaneous (SC) injection of DFP (5 mg/kg). Treatment (N = 9): DFP (5 mg/kg) followed by chronic naltrexone (5 mg/kg/day × 12 weeks). Control (N = 12): single SC injection of isopropyl alcohol, (DFP vehicle) followed by chronic naltrexone (5 mg/kg/day). If toxicity developed after injection, antidotal therapy was initiated with atropine (2 mg/kg) and pralidoxime (25 mg/kg) and repeated as needed. After 12 weeks, rats underwent testing for place learning (acquisition) across 5 days of training using the Morris Water Maze. On day 6 a memory retention test was performed. Statistical analysis was performed using IBM SPSS Statistics. Rats receiving DFP rapidly developed toxicity requiring antidotal rescue. No differences in acquisition were seen between the DFP vs. DFP + naltrexone rats. During memory testing, DFP-poisoned rats spent significantly less time (29.4 ± 2.11 versus 38.5 ± 2.5 s, p < 0.05) and traveled less distance (267 ± 24.6 versus 370 ± 27.5 cm, p < 0.05) in the target quadrant compared to the treatment group. Treatment rats performed as well as control rats (p > 0.05) on the test for memory retention. Poisoning with DFP induced impaired memory retention. Deficits were not prevented by acute rescue with atropine and pralidoxime. Chronic naltrexone treatment led to preserved memory after DFP poisoning.

Introduction

Acute and chronic exposures to nerve agents produce chronic neuropsychological deficits in humans and animals. Survivors of the sarin attack in the Tokyo subway suffered chronic deficits including impaired psychomotor functions such as finger tapping, a tendency towards worse performance on the Benton visual memory retention test, and smaller than normal brain volumes in the hippocampus and other brain regions compared to unexposed controls [1, 2]. Similarly, exposure to sarin from the demolition of depositories containing sarin during the bombing phase of the 1991 Gulf War and after the ground war is a suggested cause of the neuropsychological disabilities suffered by 30 % or more of the veterans [3]. It should be noted that, despite this apparent link between sarin exposure and cognitive impairment, a second school of thought purports that these deficits may be linked to the emotional response to the trauma of exposure or warfare and are more consistent with the symptoms of post-traumatic stress disorder [see [4] for review].

Similarly, both acute and chronic exposures to sarin-like chemicals have been shown to impact neurocognitive function. Rats acutely exposed to sarin show specific signs of neurodegeneration [5] while exposure to a lethal dose of sarin analog diisopropylfluorophosphate (DFP) developed difficulty with spatial learning, even after being rescued with atropine and pralidoxime [6, 7]. Both rats and guinea pigs chronically exposed to sarin develop neurological and neurobehavioral disabilities similar to those reported with DFP exposure [8–12]. Therefore, DFP poisoning provides a good surrogate for exposure to the more volatile sarin.

Since antidotal therapy with atropine and oximes may not prevent the neuropsychological disabilities and neurodegeneration associated with nerve agent poisoning [13–15], other antidotal therapy is needed to act on different pathways. Based on what is known about the pathophysiology of organophosphate-induced neurodegeneration, targeting neuroinflammation and apoptosis may provide an effective strategy [16]. The diterpenoid 4R-cembranoid has been found to decrease the number of dead neurons and astrocyte activation in rats poisoned with DFP [17]. Naltrexone, a neuroprotective agent with anti-inflammatory properties [18, 19], was found to prevent deficits in spatial learning in rats given lethal doses of DFP [6]. The current study tested the hypothesis that deficits in memory induced by acute poisoning with a lethal dose of DFP and rescued with atropine and pralidoxime, can be effectively prevented by chronic therapy with naltrexone.

Methods

A randomized controlled trial examining the ability of naltrexone to prevent deficits in learning and memory in rats acutely poisoned with a lethal dose of DFP, and resuscitated with atropine and pralidoxime, was performed. DFP was used in the experiment because it has comparable toxicity to sarin while being less volatile. Availability of sarin is also highly restricted while DFP is readily available for experimentation. The Institutional Animal Care and Use Committee approved all procedures. Subjects were female Long-Evans rats weighing 250 to 275 g (approximately 2.5 months of age) purchased from Charles River Laboratories (Wilmington, MA). Rats were housed 2 per cage with woodchip bedding and environmental enrichments which included nesting materials and PVC pipes for burrowing. Animal rooms are ventilated to maintain consistent temperature (72 ° F) and humidity with normal 12-h light–dark cycles. Rats were allowed unrestricted access to normal rat chow (hard pellets) and water, and were acclimated to their setting for 7 days prior to use. These animals were naïve prior to this study.

DFP and isopropyl alcohol (IPOH) were purchased from Sigma Aldrich (St. Louis, MO). Figure 1 illustrates the overall experimental design. Thirty rats were randomly assigned by drawing numbers to receive either a single subcutaneous (SC) injection of DFP (5 mg/kg) dissolved in 0.25 M IPOH (0.2–0.23 ml; N = 18) or an equal concentration and volume of SC IPOH (N = 12). Sample size was based on a power analysis done using previously generated data [6]. The dose chosen was toxic in all animals in a prior experiment [6]. After injection, rats were monitored by an investigator blinded to the injection assignments for signs and symptoms of cholinesterase toxicity for up to 90 min. If toxicity developed, antidotal therapy was initiated with SC atropine (2 mg/kg in 0.40–0.45 ml) and SC pralidoxime (25 mg/kg in 0.20–0.30 ml) and repeated as needed. Doses chosen were the same used in other studies of OP intoxication [20] and also replicated those used in our prior study [6]. Twenty-four hours later, the rats that received DFP were then randomized to the DFP-naltrexone group to receive SC naltrexone (5 mg/kg/day in 0.2–0.3 ml; N = 9) for 12 weeks, or the DFP-saline group to receive an equal volume of SC saline for the same time period (N = 8). The IPOH-naltrexone group of rats received a single injection of SC isopropyl alcohol instead of DFP and naltrexone (5 mg/kg/day in 0.20–0.30 ml). Injection syringes were filled by a single un-blinded investigator and assigned a code that matched the code on each animal cage. Administration of injections was performed by technical staff based on the assigned code, and all injections were performed between 8:00–10:00 am each day. These staff members were blinded to the contents of the syringes and meanings of the codes. Animals were weighed weekly to assess general health status and to adjust doses of naltrexone as needed. No animals displayed any obvious signs of distress or altered daily behavior during the survival period (including failure to grow or aggressiveness towards cage mate).

Fig. 1.

Summary of experimental design and timeline

At the end of the 12-week treatment period, all rats underwent 6 days of testing using the place version of the Morris Water Maze. This task was first developed to assess hippocampal-based spatial navigation learning in rats [21] and is now widely adopted for rodent models of traumatic brain injury [22] and neurodegenerative disorders [23]. It has also been extensively used in assessment of the cognitive effects of specific exposures [24–27]. Study investigators involved in behavioral testing were blinded to treatments groups of each rat. On days 1 through 5, rats underwent testing for acquisition (spatial learning). On day 6, rats were tested for memory of the place in which they were previously trained to escape water from. A 1.8 m diameter open field pool was filled with water (26 °C) made opaque with Crayola® non-toxic, washable paint. Animals were required to locate a submerged platform that was 10 cm in diameter and 1.5 cm below the surface of the water.

Test of Spatial Learning

Each animal underwent 4 trials per day. A trial began with the animal being placed in the water, face-in towards pool wall at 1 of 12 starting positions, and ended once the animal found the platform. The starting location was counterbalanced for order effects across trials, while the platform location remained in 1 of 4 target quadrants across days. The target quadrant of the platform was counterbalanced across all animals. If the animal did not find the platform within 60 s, it was guided there by hand. After a 30-s period on the platform, the animal was re-placed in the pool at a different start position until all 12 starting positions were tested. The latency to find the platform from each starting position, distance traveled, and the time spent in each quadrant of the pool was recorded.

Test of Spatial Memory

A probe trial to test spatial memory was performed by removing the platform. The rat was placed in the pool from any of the previous start positions. The time spent in each quadrant location was recorded, as well as the distance traveled within each quadrant. There were 5 possible quadrant locations, identified as: target, adjacent, adjacent-counterclockwise, opposite to target, and center of pool. Rats were removed after 120 s of swimming during the probe trial.

Statistical Analysis

Time to find platform and distance traveled for each day was compared across groups and across days of testing using Repeated Measures Analysis of Variance (RMANOVA) followed by Fisher’s post-hoc test, with p < 0.05 indicating significance. Statistical analysis was performed using IBM SPSS Statistics.

Results

All rats who received DFP developed acute toxicity within five minutes, manifested by fasciculations and seizures. One rat who received DFP seized and died immediately, bringing the DFP group to n = 17. Ten of the remaining DFP-injected rats required a single dose of atropine and pralidoxime for successful rescue within 30 min of DFP injection. Five required 2 doses, and two required 3 doses of rescue medications.

On the first day of testing, mean latency times and mean distances traveled to find the platform were comparable in all groups (DFP-saline 5.99 ± 0.65 s and 86.6 ± 8.03 cm; DFP-naltrexone 4.73 ± .31 s and 80.92 ± 10.86 cm, IPOH-naltrexone 6.55 ± 0.84 s and 84.03 ± 6.13 cm; p > 0.05). On days 2 through 5, rats in the DFP-saline group and DFP-naltrexone group had comparable mean latency times that were shorter than the latency time of rats in the IPOH-naltrexone group, as shown in Fig. 2a. Similarly, on days 2 through 4, rats in the DFP-saline group and DFP-naltrexone group had comparable mean distances traveled to find the platform that were shorter than the mean distance of rats in the IPOH-naltrexone group, as shown in Figs. 2b and 3.

Fig. 2.

Learning phase of the Morris Water Maze trials on Days 1 to 5. a Time spent finding the hidden platform. b Distance traveled to find the hidden platform. On the first day of testing, mean latency times and mean distances traveled to find the platform were comparable in all groups (p > 0.05). On days 2–5, Rats in the DFP-saline group and DFP-naltrexone group had comparable mean latency times that were shorter than the latency time of rats in the naltrexone group (a) (*p < 0.05). On days 2–4, rats in the DFP-saline group and DFP-naltrexone group had comparable mean distances traveled to find the platform that were shorter than the mean distance of rats in the IPOH-naltrexone group (b) (*p < 0.05)

Fig. 3.

Retention (memory) of previous place learning in the Morris water maze on the sixth day, consisting of a single probe trial (120-s trial) without the escape platform. Regions are categorized as Target, Adjacent (Adj), Adjacent Counter-Clockwise (ADJ-CC), Opposite (Opp), and Center (of pool). The Target is the region which contained the hidden platform during acquisition (Days 1–5). a Mean distance traveled within a particular region. b Mean percentage of time spent in a given region. DFP-poisoned rats that did not receive naltrexone spent significantly less time and traveled less distance in the target quadrant compared to DFP-poisoned rats treated with naltrexone. Those rats who were not DFP-poisoned performed comparably to DFP-poisoned rats who received naltrexone in both time and distance in the target sector (*NS within DFP-saline group; **significantly different within DFP-naltrexone and IPOH-naltrexone groups)

During probe testing for memory retention, DFP-poisoned rats that did not receive naltrexone spent significantly less time (29.4 ± 2.11 vs. 38.5 ± 2.5 s, p < 0.05) and traveled less distance (267 ± 24.6 vs. 370 ± 27.5 cm, p < 0.05) in the target quadrant compared to DFP-poisoned rats treated with naltrexone. Those rats who were not DFP-poisoned performed comparably to DFP-poisoned rats who received naltrexone, spending comparable amounts of time in the target sector (39.6 ± 3.26 vs. 38.5 ± 2.5 s) and traveling comparable distances in the target sector (353 ± 21.5 vs. 370 ± 27.5 cm). In summary, retention memory was impaired in rats that were poisoned with DFP but this impairment was blocked by treatment with naltrexone.

Discussion

The results of this experiment demonstrate that, in a rodent model, 5 mg/kg/day of subcutaneous naltrexone administered daily after a lethal poisoning with the sarin analogue DFP ablated the effects of DFP on memory based on performance in a Morris Water Maze. In the acquisition phase of the trials, naltrexone did not have an effect. During acquisition, animals are required to escape from water aversion, and with repeated exposure across trials, tend to develop learning curves regardless of the type brain damage they may have incurred. However, the severity of brain damage can contribute to more extensive abolition of water escape learning. In this study, the lack of performance differences between both DFP-treated groups may reflect recovery from early motor-related impairments initially produced by DFP exposure. On the other hand, the memory impairments that were observed in DFP-treated rats without naltrexone reflect enduring effects of DFP exposure on compromising hippocampal function. The hippocampus receives acetylcholinergic input from the medial septum and studies show that the ability to remember spatial location is mediated by cholinergic activity within the hippocampus [28]. It is quite possible that the DFP-induced memory deficits observed in these rats affected hippocampal place cells to encode the spatial location of the target platform properly. This idea, however, requires further testing, including histopathological studies to determine that memory deficits can be correlated to damage, specifically in the hippocampus. Previous studies in rats [29], in Gulf War Veterans [30, 31] and in victims of the Tokyo subway sarin attack [2] have demonstrated that a single exposure to sarin results in reduced volumes of specific hippocampal areas, supporting the idea that the memory deficits in our rats is likely due to structural changes in the hippocampus.

DFP was used in this investigation due to its similarity to sarin, which has greater volatility and has been used as both a military nerve gas and in terrorist attacks. Irreversible inhibition of acetylcholinesterase at muscarinic and nicotinic acetylcholine synapses leads to parasympathetic overstimulation and muscle weakness and is the mechanism of acute lethality. The ability of the antidotes atropine and pralidoxime to enhance survival so long as pralidoxime is given before aging occurs; that is, before nerve agents bind irreversibly to cholinesterase, is debated. However, it is generally agreed that antidotes to acute cholinergic poisoning do not prevent neurological sequelae [14, 32], suggesting that other mechanisms are involved in the chronic cerebral toxicity.

Limitations of this study are that only one nerve agent and one antidote were investigated in one species. Extrapolation to other species, and in particular to humans, is unknown. The rats exposed to DFP required varying doses of pralidoxime for rescue from the acute poisoning, creating the possibility that the pralidoxime may have an effect on outcome. Examination of the group assignments show that of the 5 rats that received 2 doses of pralidoxime, 3 were assigned to the naltrexone treatment group and 2 to the saline control group. The 2 rats that received 3 doses of pralidoxime were divided evenly between the naltrexone treatment and saline control groups. A further limitation is the lack of histopathology on the brains of the animals which would support the assertion that the behavioral changes seen were due to hippocampal damage. The time duration chosen for the experiment may also affect the measured outcomes. An assessment of behavioral changes at multiple time points over a longer period will provide more detailed information about the progression of the DFP-induced deficits. Finally, limitations of the water maze as an assay for psychological function in rodents have been noted across a variety of research disciplines [33–35]. Rodents use a variety of search strategies to navigate towards a goal (i.e., escape from water) and may explain differences in swim latencies, even slower latencies in control animals (i.e., the Naltrexone-only rats in this study). About two-thirds of the variability in Morris maze results are accounted for by non-cognitive factors such as thigmotaxis and passivity [33]. Furthermore, variations in system setup and protocol (i.e., pool size, extra-maze cues, experimenter movement), species and strain differences and even stress in the animal may contribute largely to inconsistent acquisition (latency, distance traveled) results across labs, thereby adversely influencing the allocentric aspects of rodent spatial learning [34, 36]. With the latter, a more sensitive measure of spatial learning is the amount of zones crossings that animals exhibit; it is a measure that was not obtained in this study and needs future consideration. The incorporation of a probe trial to assess memory retention, however, is widely accepted as a feasible way to separate learning from performance. Indeed, the findings in this study that DFP-poisoned rats showed poorer memory retention in the probe trial do lend support to the idea that cognitive function in these animals was compromised under heavier task demands. The clinical significance of this work and related studies is that it may be possible to prevent the delayed neurological consequences of nerve agent poisoning by chronically administering agents with anti-inflammatory properties in addition to the atropine and pralidoxime given at the time of acute poisoning.