The combined effects of pyridostigmine and chronic stress on brain cortical and blood acetylcholinesterase, corticosterone, prolactin and alternation performance in rats.

Thousands of soldiers who served in the Gulf War have symptoms that have been collectively termed Gulf War Illness (GWI). It has been suggested that a combination of operational stress and pyridostigmine, a drug given as a pretreatment to protect soldiers against the effects of exposure to nerve agents, might have had unexpected adverse health effects causing these symptoms. Our laboratory has previously modeled operational stress in rats using a paradigm of around-the-clock intermittent signalled footshock. In the present studies, this model was used to investigate the potential synergistic effects of chronic stress and pyridostigmine on physiology and behavior. Seventy-two rats were trained to perform an alternation lever pressing task to earn their entire daily food intake. The rats were then implanted with osmotic minipumps containing vehicle, pyridostigmine (25 mg/ml pyridostigmine bromide) or physostigmine (20 mg/ml eserine hemisulfate). The pumps delivered 1 microl/h, which resulted in a cumulative dosing of approximately 1.5 mg/kg/day of pyridostigmine or 1.2 mg/kg/day of physostigmine, equimolar doses of the two drugs. The rats were then returned to their home cages where performance continued to be measured 24 h/day. After 4 days, 24 of the 72 rats were trained to escape signalled footshock (avoidance-escape group) and 24 other rats (yoked-stressed group) were each paired to a rat in the avoidance-escape group. The remaining 24 rats were not subjected to footshock (unstressed group). Shock trials were intermittently presented in the home cage 24 h/day for 3 days, while alternation performance continued to be measured. Since only 12 test cages were available, each condition was repeated to achieve a final n of six rats per group. Pyridostigmine and physostigmine each decreased blood acetylcholinesterase levels by approximately 50%. Physostigmine also decreased brain cortical acetylcholinesterase levels by approximately 50%, while pyridostigmine had no effect on cortical acetylcholinesterase activity. Alternation performance was impaired on the first day of stress and then recovered. Neither pyridostigmine nor physostigmine affected performance in the absence of stress or increased the effects of stress alone. Corticosterone was significantly increased in the yoked stress group compared to unstressed controls. These data suggest that pyridostigmine does not exacerbate the effects of stress on performance or levels of stress hormones. Furthermore, these data do not suggest that stress enables pyridostigmine to cross the blood brain barrier.

Behavioral effects Of 8-OH-DPAT in chronically stressed male and female rats.

The present study tested the hypothesis that chronic stress desensitizes serotonergic 5-HT(1A) receptors and alters behavioral changes following 5-HT(1A) agonist administration. Eating, acoustic startle response (ASR), and locomotor activity were measured in stressed and nonstressed male and female rats after 8-OH-DPAT administration. Stressed rats were paired and stressed by around-the-clock intermittent foot shock. Controllable stress (CS) rats could avoid/terminate shock for themselves and their yoked partners by pulling a ceiling chain, whereas their partners, the uncontrollable stress (UCS) rats, could not. Rats earned their entire daily ration of food by pressing a lever. In previous experiments, this paradigm was stressful, but not debilitating and rats continued to eat, groom, sleep, and avoid/escape greater than 99% of shock trials. Locomotor activity and ASR were measured in the present study after saline and 8-OH-DPAT administration (0.25 mg/kg, IP) before, 24 h, and 72 h after shock onset. 8-OH-DPAT only decreased food intake significantly in male and female rats after the first administration. Stress decreased food intake in both the CS and UCS rats, with UCS rats eating the least. However, the effects of stress and 8-OH-DPAT were not additive. 8-OH-DPAT significantly increased peak startle amplitude at 100 and 120 dB, and decreased latency to peak startle amplitude at 100 dB in male and female rats. In contrast, 8-OH-DPAT did not alter percent prepulse inhibition (%PPI) at 100 dB, but significantly decreased %PPI in males but not females at 120 dB. Stress did not have a consistent effect on ASR, but reduced %PPI in males, but not females. Neither stress nor 8-OH-DPAT significantly altered locomotor activity. Although the results do not show an increased sensitivity to 8-OH-DPAT in stressed rats, the unexpectedly weak effects of 8-OH-DPAT alone on the behavioral measures chosen limits the conclusions that can be drawn.

Chronic sustained stress increases levels of anterior pituitary prolactin mRNA.

Our laboratory is investigating the effects of chronic stress on physiological, endocrine and behavioral measures, in order to elucidate the neuronal substrates for the pathophysiological consequences of stress in humans. In these studies, we have employed a rodent model of sustained stress in which rats are exposed to around-the-clock intermittent footshock that can be avoided or escaped by rats in the controllable stress group, but not by rats in the uncontrollable stress group. Each rat in the uncontrollable stress group is paired (yoked) to a rat in the controllable stress group such that the controllable stress group rat avoids or escapes shock for both rats. A third group of rats receives no shock (controls). We have previously reported that in male rats, plasma prolactin levels were elevated after 3 days of stress in both stress groups. In the present experiments we determined whether the increases in plasma prolactin were correlated with increases in anterior pituitary prolactin mRNA. In addition, we measured hormones and mRNA at three time points and we examined these responses in female as well as male rats. Adult male and female Sprague-Dawley rats were exposed to chronic stress for 1, 3 or 14 days. In unstressed control rats, levels of anterior pituitary prolactin mRNA were fivefold higher in female as compared to male rats. However, stress increased levels of anterior pituitary prolactin mRNA over baseline in both genders. After 1 day of stress, anterior pituitary prolactin mRNA levels increased in male and female rats belonging to both stress groups, with no significant difference seen between the uncontrollable vs. controllable stress groups. After 3 days of stress, anterior pituitary prolactin mRNA levels were even more elevated, and rats in the uncontrollable stress group had higher anterior pituitary prolactin mRNA levels than those in the controllable stress group. After 14 days of stress, there were no significant differences in control and stressed groups with respect to anterior pituitary prolactin mRNA. These data suggest that chronic sustained stress increases the synthesis of anterior pituitary prolactin mRNA during the first days of stress, and that levels return to prestress values sometime between 3 and 14 days of stress.

Sustained stress disrupts the performance and acquisition of delayed alternation in rats.

The effects of sustained stress on acquisition and performance of a delayed alternation task were studied in male rats. Rats lived 24 h per day in operant cages where they earned all of their food via lever pressing. During the stress portion of each experiment, one group of rats was able to avoid or escape signaled intermittent footshock (Avoidance/Escape group), a second group (Yoked) did not have control over shock termination, a third group never received shock (Control). Shock trials were presented around-the-clock at approximately 5-min intervals and the stress portion of each study lasted 1 week. We have previously reported that rats tolerate this paradigm well and avoid/escape 99% of the shock trials. Three experiments were conducted. In Experiment 1, rats learned the delayed alternation task prior to stress onset; in Experiment 2, rats were exposed to stress and the alternation task concurrently; in Experiment 3, rats were stressed for 14 days prior to being required to perform the delayed alternation task. In the first experiment, stress decreased both food intake and the accuracy of responding during the first days of stress. In the second experiment (acquisition), stressed rats required more days to reach asymptotic performance on the alternation task. In Experiment 3, rats stressed for 14 days prior to acquisition of the delayed alternation task performed similarly to controls.

Effects of the serotonin agonists 8-OH-DPAT, buspirone, and DOI on water maze performance.

We have previously reported that the serotonin 5-HT1A agonist 8-OH-DPAT and the 5-HT2c agonist TFMPP impair performance on a water maze. In the present report we extended those studies by examining a second 5-HT1A agonist, buspirone, to see whether its effects paralleled those of 8-OH-DPAT, and by testing the effects of the 5-HT2 agonist DOI. Unlike the open pool Morris water maze, the maze used in these experiments has alleys and doorways. The maze can be easily reconfigured to present rats with both previously learned or new maze challenges. Performance is assessed by time to reach the maze exit platform and the number of wrong doorways entered (errors). At doses that did not affect performance in a previously learned maze, the 5-HT1A agonists 8-OH-DPAT (0.1 mg/kg) and buspirone (1 mg/kg) slowed acquisition of a new maze configuration as measured by both swim time to the exit platform and errors committed. A higher dose of buspirone (10 mg/kg) completely blocked acquisition of a novel maze. In contrast. DOI slowed performance as assessed by swim time on both a well-learned maze as well as acquisition of a new maze, but did not affect error rate on either task, suggesting that this 5-HT2 agonist impaired performance by depressing motor activity. These experiments demonstrate that serotonin agonists, especially the 5-HT1A subtype, can impair learning.

Sustained stress and fixed interval performance.

The effects of sustained stress on response rate and temporal patterning (quarter-life) of rats performing either a previously learned fixed-interval schedule (FI 60) or learning an FI 60 simultaneously with stress onset were determined. Rats lived 24 h/day in operant cages, where they earned all of their food via lever-pressing. During the stress portion of each experiment, one group of rats was able to avoid or escape signalled intermittent footshock (Avoidance/ Escape Group), a second group (Yoked) did not have control over shock termination, a third group never received shock (Control). Shock trials were presented around the clock at approximately 5-min intervals and the stress portion of each study lasted 1-2 weeks. We have previously reported that rats tolerate this paradigm well and avoid/escape 99% of the shock trials. In rats previously trained on the FI task, both rate of responding and quarter-life values were significantly decreased on the first day of stress for both the Avoidance/Escape and Yoked Groups. Food intakes and quarter-life values were not significantly different from the controls by stress Days 3 and 2, respectively. In the acquisition study, controls learned the F1 task by Day 4 as judged by quarter-life of responding. FI task acquisition was significantly impaired in stressed rats compared to controls, not reaching asymptotic performance until Day 9 of stress. There were no major differences between the 2 stress groups in either study. These data demonstrate that stress may impair both the rate and patterning of behavior, and suggest that this rodent paradigm may usefully model some aspects of the effects of stress in humans.

Effects of chronic stress on food acquisition, plasma hormones, and the estrous cycle of female rats.

Our laboratory has previously conducted a number of studies to determine the effects of chronic stress on the physiology and behavior of male rats. The present study was performed to extend these investigations to female rats. Female rats were chronically stressed using a behavioral paradigm of around-the-clock signalled intermittent foot shock in which some rats can pull a chain to avoid/escape shock (stress) while another group of rats is yoked to the first group (yoked-stress) and does not have control over shock termination. Control rats were never shocked but all groups lever pressed for food pellets on an FR1 schedule (one pellet per lever press). Daily vaginal samples were obtained for several weeks prior to stress onset and throughout the chronic stress period. After 14 days of stress, the experiment was terminated and morning blood samples were collected for hormonal assays. Stress transiently decreased lever pressing for food pellets and body weights, but both measures returned to prestress levels by day 14 of stress. Plasma adrenocorticotropic hormone (ACTH) concentrations were significantly elevated in the yoked-stress group compared to the other two groups, but there were no significant effects of 14 days of stress treatment on plasma corticosterone, prolactin, estradiol, or progesterone concentrations. There were no significant differences in estrous cycle length among experimental groups.

Effects of triazolam and diazepam on learning and memory as assessed using a water maze.

One of the reported adverse side effects of the frequently prescribed benzodiazepines diazepam (Valium) and triazolam (Halcion) is an impairment of anterograde memory in humans. The experiments described in this article compared the effects of triazolam and diazepam on performance in a water maze task that is sensitive to drugs that affect learning and memory. The water maze utilized is a traditional type of maze with alleyways and door choices, unlike the Morris open water maze. Time required to find an out-of-the-water platform and errors committed during the swim are used as performance measures. Rats were tested on a previously learned maze configuration and on the acquisition of new maze configurations. Neither diazepam (0.25, 1.0, or 2.0 mg/kg) nor triazolam (0.05, 0.2, or 0.3 mg/kg) injected 30 min prior to testing on the previously learned maze affected swim time or errors committed. Administration of diazepam (0.5, 1.0, or 2.0 mg/kg, IP) prior to daily training on three different new maze configurations did not affect swim time, but did increase swim errors. Triazolam administered at 0.1, 0.2, or 0.3 mg/kg markedly impaired performance as assessed by either swim time or errors. There were no differences in performance of rats previously treated with triazolam, diazepam, or vehicle in learning another new maze after drug treatment was terminated. These data demonstrate that both diazepam and triazolam affect acquisition but not recall of maze configurations and support similar conclusions reached using other types of tasks in humans and animals.

Effects of the serotonin receptor agonists 8-OH-DPAT and TFMPP on learning as assessed using a novel water maze.

We evaluated the effects of two drugs active at serotonin receptors, 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT, a 5-HT1A agonist) and N-3-trifluoromethylphenyl)piperazine hydrochloride (TFMPP, a 5-HT2C agonist) on learning using a novel water maze previously characterized in our laboratory. The water maze utilized is a traditional type of maze with alleyways and doors through which the rats learn to swim to reach a platform, unlike the open pool Morris water maze task. Performance is assessed by swim time required to reach the platform and errors committed. Following initial training on maze configuration A, rats were assigned to saline, TFMPP and 8-OH-DPAT treatment groups and tested for performance once per dose, 30 min after administration of drug (0.25, 0.5, and 1.0 mg/kg IP). Swim times were significantly increased as compared to saline for all doses for both drugs. The error rate was increased for 8-OH-DPAT at all doses, while TFMPP had no effect on error rate at any dose. Next, rats were challenged to learn new mazes following daily administration of 0.25 or 0.5 mg/kg of each drug 30 min prior to each daily swim trial. Rats given 0.25 mg/kg of 8-OH-DPAT learned new maze C more slowly than saline-treated rats, while TFMPP had no effect at this dose. At the higher dose of 0.5 mg/kg, tested on new maze B, TFMPP administration significantly increased swim times but not errors, while this dose of 8-OH-DPAT markedly increased both swim time and errors. Finally, rats from all groups were tested on maze E after drug administration was discontinued, and there were no performance differences among groups. These data suggest that serotonin1A receptors may inhibit learning.

Time cost of alternation reduced demand for food in a closed economy.

A principal purpose of the present study was to characterize rats' demand for food by increasing the required delay between the alternation of leverpresses. While living in cages that were exposed to a 12:12 light:dark cycle, rats were required to alternate leverpresses for their entire daily food ration. After a press on one lever resulted in the delivery of a food pellet, a red cuelight was illuminated for a fixed duration that equaled the imposed delay between successive leverpresses. After the imposed delay elapsed and the cuelight was turned off, a press on the alternate lever resulted in the delivery of another food pellet. Increases in the imposed delay from 0.01 s to 16 s reduced mean percent correct (alternation) and mean food intake, although both were reduced more rapidly and to a greater extent in the light than in the dark photoperiod. This interaction resulted from a reduction of meal size in the light and a compensatory increase in meal size in the dark as the imposed delay increased. Rats also defended their daily intake in the dark against increasingly larger imposed delays by spending increasingly larger amounts of time alternating and by increasing total (correct + incorrect) presses. These results suggest that photoperiod is a determinant of the motivational value of food as a reinforcer for instrumental behavior in a closed economy.

Effects of chronic stress on sleep in rats.

The present study was conducted to determine the effects of chronic stress on sleep using a rodent paradigm of around-the-clock signalled intermittent foot shock in which some rats can pull a chain to avoid/escape shock while another group of rats is yoked to the first group. We measured sleep using telemetry; four-channel EEG was collected 24 h/day in rats during 2 prestress days; days 1, 2, 3, 7, and 14 during chronic stress; and 3 poststress days. States of REM sleep, non-REM (NREM) sleep, and waking were scored for each 15-s period of the EEG recordings. During the prestress period, rats slept (REM plus NREM) 55% of available time during the light hours and 34% of the dark hours with the remainder represented by waking. On the first day of stress, total sleep and, especially REM sleep, decreased markedly. By the second day of stress, only REM sleep in the controllable stress group (but not the uncontrollable stress group) was still significantly decreased compared to prestress levels, and REM sleep returned to baseline levels by day 7 of stress. The recovery of sleep quantity was accomplished by increased sleep during the dark hours, resulting in a long-lasting disruption of normal circadian sleep patterning.

Effects of chronic stress and time of day on preference for sucrose.

Rats were allowed to lever press for two types of food pellets of equivalent caloric value and total carbohydrate content. One type of food pellet provided more of the calories as sucrose. During a 16-day prestress period, lever presses for 12 rats were recorded hourly. Following the baseline period, four rats (stressed group) were shaped to pull a ceiling chain to avoid or escape signalled foot shock presented intermittently around-the-clock. Four additional rats (yoked group) were each paired to one of the chain-pulling rats such that the rat trained to pull the ceiling chain controlled stressor termination for both rats. A third group of four rats served as the control group and received no shock. We have previously reported that rats in this model of chronic stress tolerate the paradigm well, continuing to gain weight, eat, drink water, and groom and escape more than 99% of the trials presented. During the baseline period, the sweeter pellet was preferred by most rats, but differences in preference among rats and in preference at different times of day were observed. The preference for the high-sucrose pellet was most marked in the hours preceding lights off. Overall, no changes in food preference were seen as a function of stress condition during the 14-day stress period, although one rat in the yoked group increased preference for the sweeter pellet during stress and returned to prestress food preferences when stress was terminated.

Effects of chronic stress on anterior pituitary and brain corticotropin-releasing factor receptors.

[125I]Corticotropin-releasing factor binding was measured in membrane homogenates prepared from the anterior pituitary, frontal cortex, motor cortex, somatosensory cortex, mesolimbic area (olfactory tubercle and nucleus accumbens), caudate putamen, hypothalamus, midbrain, and cerebellum from control and chronically stressed rats. The stressor consisted of 3 or 14 days of around-the-clock intermittent foot-shock (approximately one trial per 5-min frequency) that could be avoided or escaped on 90% of the trials presented by pulling a ceiling chain. Plasma corticosterone levels were almost doubled in stressed rats following 3 days of chronic stress and remained significantly elevated in rats stressed for 14 days as compared to controls. Plasma corticotropin levels were similar in controls and stressed animals in both the 3- and 14-day experiments. [125I]Corticotropin-releasing factor binding was decreased in anterior pituitary and frontal cortex following 3 days of chronic stress; binding affinity of anterior pituitary membranes was not different between control and stressed animals. [125I]Corticotropin-releasing factor binding was similar in control and 3-day-stressed animals in the other brain areas examined. After 14 days of chronic stress, hypothalamic [125I]corticotropin-releasing factor binding was decreased in stressed rats as compared to control animals but no other differences were seen. The decrease in the apparent number of anterior pituitary corticotropin-releasing factor receptors following 3 days of stress may be due, in part, to increased plasma corticosterone levels and/or increased corticotropin-releasing factor secretion during that time.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of psychoactive drugs or stress on learning, memory, and performance as assessed using a novel water maze task.

A novel water maze was used to assess the potential performance-disrupting effects of psychoactive drugs and stressors (4 mg/kg amphetamine sulfate; 1, 2, or 4 mg/kg diazepam; 30 mg/kg caffeine; 5 or 30 mg/kg atropine sulfate; 15 min of either intermittent foot-shock, forced running, or immobilization). The task utilized a traditional type of maze with walls and doorways set inside a pool. The apparatus could easily be reconfigured to present different mazes of approximately equal difficulty by opening or closing multiple doorways. Performance was measured by number of errors and time required to swim from the "start" to "finish" (a raised platform not in the rat's line of sight). After initial maze training, rats were divided into two groups. One group ran three daily trails through the same maze each day; this group was used to assess memory. The second group was challenged to swim three consecutive trials in a new maze configuration each day as a measure of learning. On any given day, rats from both groups received the same treatment. Drug or stress treatments were interspersed with vehicle or no-treatment trials days. The new maze task was more sensitive than the well-learned maze to the performance disrupting effects of amphetamine, caffeine, and diazepam, while atropine had no significant effect on performance on either maze. Foot-shock stress impaired performance on both mazes, while the other stressors had no significant effect.

Effect of stress on hypothalamic insulin in rats.

Four experiments examined the effects of stress on hypothalamic insulin and plasma hormones in rats. Two hours daily of immobilization (IM) stress for 2 and 4 days resulted in an increase in hypothalamic insulin. In contrast, 15 min of daily IM over 13 days or 48 h of continuous signalled shock avoidance did not alter hypothalamic insulin. These findings are interpreted to indicate that changes in hypothalamic insulin are part of the stress response. Possible reasons for the different effects of time and paradigm on the hypothalamic insulin responses to stress are discussed. Plasma insulin and glucose levels were not responsive to any of the stressors. Brief acute stress caused increases in the stress-responsive hormones ACTH, corticosterone, and prolactin, as expected, and these responses attenuated or disappeared with repeated or longer stress exposures.

Effects of controllable vs. uncontrollable chronic stress on stress-responsive plasma hormones.

We have previously reported effects of chronic stress on circadian rhythms of temperature, eating, and locomotor activity. These studies were conducted using an around-the-clock signalled intermittent footshock paradigm in which some rats have control over shock termination while other rats are yoked to the rats with control. Although this paradigm is stressful, as suggested by decreases in food intake and disrupted circadian rhythms, rats tolerate the paradigm well, continuing to eat, drink, gain weight, and groom. In the present studies, rats were sacrificed following 3 or 14 days of stress, and plasma was collected for hormonal assays. After 3 days of stress, plasma corticosterone and prolactin levels were elevated in both stress groups compared to controls; yoked rats had higher levels of corticosterone than rats in the group with control over shock termination, while prolactin levels in both stressed groups were similar. ACTH levels were similar in stressed and control rats. After 14 days of stress, ACTH and corticosterone levels in both stress groups were similar to control levels. Prolactin levels were elevated in the yoked experimental group compared to levels in control or controllable stress groups. These data support previous studies suggesting that control over stressors attenuates the effects of stress on physiology and demonstrate that two hormones with diverse biological effects are elevated by chronic stress.

Circadian effects of escapable and inescapable shock on the food intake and wheelrunning of rats.

The feeding and running activities of rats were used to characterize the stressful effects of footshock. A 12-hour light/dark cycle was continuously in effect. During the preshock stage, all rats lived in individual cages where they could drink water, press a lever once for each 45 mg food pellet or run in an attached running wheel. During the shock stage, rats were moved to cages where they could drink and eat, but not run. Five s of light followed by 5 s of a light + tone compound preceded five progressively more intense shock levels. Rats in group AE were trained to pull a ceiling chain to avoid/escape from footshocks that were scheduled 24 hours per day. In group YS, the frequency, intensity and pattern of shocks for a rat were yoked to a rat in group AE. Rats in group CN were not shocked. During recovery, rats were allowed to resume running. During the dark, but not the light of the first day of shock, the intake of groups AE and YS was less than the intake of group CN. By the third day of shock, the food intake of group AE recovered and was no longer different from the intake of group CN. Intake of group YS remained less than the intakes of groups AE and CN during the remainder of the shock stage. When shock was terminated, the intake of the YS group recovered, but neither running nor food intake differed among groups. These results suggest that although inescapable shock disrupts food intake more than escapable shock, this disruption is specific to the duration of the shock stage and depends on rat's circadian pattern of intake.

Effects of MK-801 on learning and memory as assessed using a novel water maze.

The effects of the NMDA receptor antagonist MK-801 [(+)-10,11-dihydro-5-methyl-5H-dibenzo [a,d]-cyclohepten-5,10 imine hydrogen maleate] on learning and memory were assessed using a water maze. The maze was a traditional type of maze with alleys and choices between various paths, but set inside a pool of water to a height of 25 cm. Different mazes could be configured by altering the arrangement of open vs. closed doors. Both the time required to reach an out-of-the-water exit platform and the errors made during the swim from start to finish were recorded. Learning was assessed during the first 10 to 20 trials in a new maze configuration, while memory was tested after the maze was well learned. Three experiments, some with several phases, were performed. These experiments compared the effects of 0.1 mg/kg of either (+)-MK-801, or (-)-MK-801 vs. saline on learning new maze configurations as well as swimming well-learned mazes. Neither of the MK-801 isomers impaired performance of a previously learned maze. (+)-MK-801 clearly slowed learning of new mazes as measured by both maze completion time and errors committed, while (-)-MK-801 had a significant but smaller effect on learning. Rats given (+)- or (-)-MK-801 (0.1 mg/kg) for 16 days while learning one maze and then challenged to learn a new maze without drug administration performed no differently on the new maze than controls, suggesting that the acute effect of MK-801 on learning is not long lasting.

Effects of soman on neuroendocrine and immune function.

We have previously reported that plasma growth hormone (GH) and prolactin levels were markedly decreased in rats two weeks following a single dose (100 micrograms/kg, SC) of soman. We have now conducted additional experiments to attempt to determine whether neuroendocrine responses to physiological or pharmacological challenge are altered in rat survivors of soman exposure, and whether immune function, which can be affected by circulating hormones, is altered in the soman-exposed rats. In the present study, basal prolactin levels were not significantly lower in the soman-treated rats although prolactin increases in response to physiological or pharmacological challenge were attenuated. Also, basal growth hormone levels in soman survivors were similar to control levels in 2 of 3 experiments in the present report. In the third experiment, growth hormone levels were lower in soman-treated animals. Endocrine abnormalities appeared to be related to the severity of soman insult as assessed by changes in body weight following exposure. Both ACTH and prolactin responses to stress were impaired in a severely affected subpopulation of soman survivors. The thymus, an important immune organ, was decreased in weight in severely affected soman survivors, but other tests of immune function did not show differences between control and soman-exposed rats.

Effects of sustained avoidance/escape on demand for food.

Rats were concurrently exposed to a sustained avoidance/escape (SAE) procedure and a procedure for assessing demand for food. In the sustained SAE procedure, signalled shock avoidance/escape trials were presented at varying intervals averaging five minutes. The shock could be avoided by pulling on a ceiling chain early in the trial, or escaped by pulling on the chain later in the trial. Demand curves for food were generated by requiring 1, 5, 10, 20, 40, 80, 160, and 320 lever presses for each food pellet on successive days. The demand curve procedure was introduced after either brief (3 or 5 days) or extended (21 or 23 days) exposure to SAE. Following brief exposure to SAE, SAE animals showed decreased food intake and less elasticity of demand relative to non-SAE controls. Following extended exposure to SAE, these effects were diminished or absent.