Spatial learning deficits induced by chronic prenatal ethanol exposure can be overcome by non-spatial pre-training

Neurotoxicology and Teratology 28 (2006) 333 – 341 www.elsevier.com/locate/neutera

Spatial learning deficits induced by chronic prenatal ethanol exposure can be overcome by non-spatial pre-training Umar Iqbal a , Shivani Rikhy b , Hans C. Dringenberg a,b,c , James F. Brien a,c , James N. Reynolds a,c,⁎ a

Department of Pharmacology and Toxicology, Queen's University, Kingston, ON, Canada K7L 3N6 b Department of Psychology, Queen's University, Kingston, ON, Canada K7L 3N6 c The Centre for Neuroscience Studies, Queen's University, Kingston, ON, Canada K7L 3N6

Received 20 September 2005; received in revised form 20 December 2005; accepted 27 January 2006 Available online 10 March 2006

Abstract This study tested the hypothesis that behavioural intervention, in the form of non-spatial pre-training, mitigates the deficits in spatial learning tasks induced in guinea pig offspring by chronic prenatal ethanol exposure (CPEE). Timed, pregnant guinea pigs were treated with ethanol (4 g/kg maternal body weight/day), isocaloric-sucrose/pair-feeding, or water throughout gestation. Offspring received non-spatial pre-training, in which animals were exposed to the procedural requirements of the water maze in the absence of distal spatial cues, and then were tested in both stationary-platform and moving-platform tasks with spatial cues. Saliva cortisol was quantified in non-trained and pre-trained animals before and after exposure to the water maze. Results: CPEE offspring exhibited performance deficits in the stationary-platform task, and non-spatial pre-training improved performance of CPEE offspring to control levels. In contrast, non-spatial pre-training had no effect on the impaired performance of CPEE offspring in the movingplatform task. Non-trained CPEE offspring had elevated saliva cortisol concentration after water-maze exposure compared to control offspring. Moreover, pre-trained control animals exhibited a sensitization of the cortisol response after repeated exposure to the water maze, and this was not evident in pre-trained CPEE offspring. Conclusions: These data demonstrate that CPEE produced deficits in spatial learning and memory processes that were partially overcome by nonspatial pre-training; however, more difficult tasks continued to reveal cognitive deficits. For repeated exposure to the water maze, CPEE offspring achieved a level of performance that was not different from control offspring, suggesting that it is the initial rate of acquisition of new learning, rather than the overall ability to learn, that is most adversely affected by CPEE. © 2006 Elsevier Inc. All rights reserved. Keywords: Fetal alcohol syndrome; Guinea pig; Cortisol; Stress; Water maze; Behavioural intervention; Pre-training

1. Introduction Alcohol consumption during pregnancy can lead to a number of physical, behavioural, and psychological problems in affected children, and may result in the fetal alcohol syndrome (FAS). The FAS is defined by three principal characteristics in infants: damage to the brain, facial malformations, and growth ⁎ Corresponding author. Department of Pharmacology and Toxicology, Queen's University, Kingston, Ontario, Canada K7L 3N6. Tel.: +1 613 533 6946; fax: +1 613 533 6412. E-mail address: [email protected] (J.N. Reynolds). 0892-0362/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ntt.2006.01.011

delays [22]. Of these three principal features, it is the brain injury that is the most debilitating and persistent throughout life. The brain injury of FAS involves decreased brain size, and damage to specific regions including the hippocampus, a brain structure involved in learning, memory, and regulation of behaviour [2]. In the guinea pig, a reliable model of ethanol teratogenicity [25], chronic prenatal ethanol exposure (CPEE) induces hippocampal growth restriction and selective neuronal cell loss [11]. Behaviourally, CPEE offspring demonstrate performance deficits in spatial learning tasks such as the Morris water maze [18,30], a task sensitive to hippocampal injury [25]. Similarly, in the rat, deficits in spatial acquisition in the water

334

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341

maze have been found after chronic prenatal ethanol exposure [3,41], and after binge-like exposure to ethanol during the neonatal brain growth spurt [24,21]. Environmental intervention during postnatal life can accelerate development and facilitate recovery from brain injuries [31]. It is well established that environmental enrichment can provide improvement of performance deficits in the Morris water maze for CPEE rats compared with control rats [13]. For example, group housing of CPEE rat offspring with a variety of “toys” included in the home cage, together with daily handling, results in significant improvement in offspring performance in spatial learning paradigms such as the water maze [14]. Similarly, male CPEE rat offspring raised in home cages containing a running wheel demonstrate an improvement in spatial memory [7], suggesting that voluntary exercise also can improve the performance of CPEE offspring in learning tasks. Another approach that does not involve improving the home cage environment is behavioural intervention, or non-spatial pre-training. Non-spatial pre-training has been used as an intervention to overcome drug-induced water-maze performance deficits induced by NMDA antagonists [16], acute ethanol exposure [5] and brain lesions [17]. To the best of our knowledge, non-spatial pre-training has not previously been tested in CPEE-induced learning deficits. For non-spatial pretraining, animals swim in the water maze under conditions in which distal visuo-spatial cues are absent, and the hidden platform is moved to a new random location after every trial [28]. Thus, the animal is afforded the opportunity to acquire the general behavioural requirements of the water maze (swimming away from the walls, finding and climbing on to a hidden platform for escape) in the absence of information about the spatial environment. It is important to understand that performance deficits in the water maze can be a result of many factors, including impaired motor or sensory ability, swim stress associated with the task, impaired acquisition of the appropriate behavioural strategy, and/or impaired spatial navigation and acquisition. Optimally, one should determine the effects of prenatal treatment in a number of water-maze tasks to better understand and interpret the nature of performance deficits and cognitive impairment. The objective of the present study was to test the hypothesis that a behavioural intervention, involving non-spatial pretraining of the animal in the environment to be encountered during testing, overcomes CPEE-induced deficits in water-maze performance in the guinea pig. More specifically, we were interested in determining whether the behavioural intervention would provide a global benefit in multiple water-maze paradigms. Moreover, we recently demonstrated that CPEE induces very high maternal cortisol concentration during gestation, and alters glucocorticoid receptor function in the hippocampus of the young adult guinea pig [19]. Ethanol consumption by the pregnant female rat increases glucocorticoid activity by increasing the set point of HPA axis function, thereby resulting in increased basal and stress-induced plasma corticosterone concentrations [37]. Moreover, this effect of CPEE is associated with altered adrenocortical development in young postnatal offspring [38], and HPA axis hyper-respon-

siveness to stressors in adult offspring [26,35,39]. We, therefore, also determined the effect of CPEE on the stress response of guinea pig offspring for first exposure in the water maze, and whether non-spatial pre-training caused adaptation in this stress response. 2. Methods 2.1. Experimental animals The experimental protocol was approved by the Queen's University Animal Care Committee, and was conducted in accordance with the guidelines of the Canadian Council on Animal Care. Female, nulliparous Dunkin–Hartley-strain guinea pigs (Charles River Canada Inc.), approximately 600 g body weight, were bred using an established procedure [9]. Gestational day 0 was defined as the last day of full vaginalmembrane opening, and term is about gestational day 68. Animals were housed individually in plastic cages with a 12-h light/dark cycle at an ambient temperature of 23 °C. At postnatal day 17, animals were weaned from their mothers and separated by sex. 2.2. Animal treatment regimens Pregnant animals were separated into one of three experimental groups and received via oral intubation: a) 4g ethanol (30% v/v)/kg maternal body weight/day with ad libitum access to food and water; b) isocaloric-sucrose (42% w/v) with pairfeeding to an ethanol treated animal and ad libitum access to water; c) isovolumetric water with ad libitum access to food and water. Treatment was administered in two equally divided doses, 2h apart every day throughout gestation (gestational day 2 to gestational day 67). A blood sample was taken on GD 57 via an ear blood vessel 1 h after the second divided dose. Blood ethanol concentration (BEC) was determined using an established gas–liquid chromatographic method [34]. 2.3. Behavioural testing The Morris water-maze test of spatial learning was adapted for guinea pigs [8], based on an established method [27]. The maze consisted of a pool (1.8-m diameter) filled with water that was made opaque by adding 500ml of non-toxic white paint. Each trial in the water maze was recorded using a video camera for off-line analysis. All water-maze testing was conducted in the afternoon between 1300 and 1500h. Offspring from each litter of each treatment group were randomly selected and separated into non-trained and pre-trained groups. 2.4. Non-spatial pre-training Offspring from each individual litter were randomly assigned to the pre-trained and non-trained groups. Offspring assigned to the pre-trained groups underwent non-spatial pretraining over five consecutive days, in which each guinea pig was given two trials each day with a 3-h delay between trials.

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341 Table 1 Pregnancy outcome data

335

60s in a probe trial, in which the platform was removed from the pool. Treatment Ethanol (9)

Sucrose (9)

Water (9)

Maternal death Spontaneous abortion Perinatal death Length of gestation (days) Litter size Offspring birth weight Male littermates (%) Female littermates (%)

0 0 4 68 ± 0.5 3.7 ± 0.9 74.4 ± 3.2⁎ 41 59

0 0 4 66 ± 1.3 3.8 ± 1.2 84.2 ± 3.7 53 47

1 0 2 68 ± 0.9 4.0 ± 0.9 89.6 ± 3.9 60 40

The number of pregnant guinea pigs/litters is reported in parentheses. The data for maternal death, spontaneous abortion and perinatal death are reported as the number of occurrences. The data for the other pregnancy outcome variables are reported as group means ± SEM. ⁎p b 0.05 for ethanol versus both sucrose and water treatment groups.

An opaque curtain that completely surrounded the pool was used to eliminate a view of distal spatial cues in the environment, and lighting was dimmed in the experimental room. The animal was placed in the pool at one of four quadrant locations at the edge of the pool determined randomly for each trial, and was allowed 90s to locate and climb onto a hidden platform, where it remained for 30s. The platform was moved to a new location prior to each trial. Time latency to find the hidden platform was recorded for each trial. If the guinea pig did not locate the platform within 90s, it was guided to the platform and allowed to stand on it for 30s before being removed from the maze. The person recording swim times remained outside of the curtains and viewed the darkened video feed to record latency times. Guinea pig offspring that did not receive pretraining (non-trained) were transported to the room each test day, and remained in covered cages in the testing room while their counterparts underwent pre-training. Each animal cage was covered and the room was darkened, such that the animal was unable to see spatial cues while in the test room before and after the pre-training trials, which began on postnatal day 40 for each animal and ended on postnatal day 44. Pre-trained and non-trained animals were then moved into testing in the stationary-platform task, as described below.

2.6. Spatial learning: moving-platform task Each animal was given an eight-day rest period between the stationary-platform and moving-platform tasks. As was the case for the stationary-platform experiment, spatial cues were not blocked during this phase of testing. Each guinea pig received two consecutive trials each day for six days. The animal was placed into the water maze, facing the wall, at one of the four cardinal compass points around the edge of the pool. Each animal was allowed 45 s to find the hidden platform, which was moved to a new quadrant at the beginning of each day of testing (i.e., moving-platform task) and remained in the same position for the two trials conducted on each experimental day. If the animal failed to locate the platform within 45s, it was manually led to the platform and allowed to remain on it for 15 s before initiating the next trial. The inter-trial interval was 3 min. Moving-platform testing began on postnatal day 62 and ended on postnatal day 67 for each animal. 2.7. Water-maze data collection and analysis A video camera suspended directly above the centre of the pool was used to record each trial in the stationary-platform and

A 75

Escape Latency (s)

Pregnancy outcome variable

50

25

1

2

3

4

5

Day

Escape Latency (s)

For this phase of the experiment, each animal could utilize spatial cues in the test environment during task acquisition. Each guinea pig received four trials per day with an inter-trial interval of 120s. This was repeated for seven consecutive days, during which the animal was placed in a pool at one of the four cardinal compass points around the edge of the pool and allowed a maximum of 45s to swim to the hidden escape platform. The location of the hidden platform remained constant for all trials and all days of testing. The time taken to locate the hidden platform was recorded for each trial. If an animal failed to locate the platform within 45s, it was manually placed on the platform for 15s before commencement of the next trial. Stationary-platform testing began on postnatal day 45 for each animal and ended on postnatal day 51. On the eighth day, each animal was tested for

ethanol sucrose water

0

B 2.5. Spatial learning: stationary-platform task

*

50 40

*

30 20 10 0 Ethanol

Sucrose

Water

Treatment Fig. 1. Non-spatial pre-training of offspring from pregnant guinea pigs that received chronic oral administration of 4g ethanol/kg maternal body weight/day, isocaloric-sucrose/pair-feeding, or water treatment. Data are presented as mean ± S.E.M of the escape latency on each of the 5days of training (A), and collapsed across all test days (B) (n = 8 offspring from 8 different litters for each treatment group). ⁎Ethanol group statistically different from the two control groups (p b 0.05).

336

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341

A

B

Escape Latency (s)

*

30

Ethanol (non-trained) Sucrose (non-trained) Water (non-trained)

#

20 10

Escape Latency (s)

40

40

Ethanol (pre-trained) Sucrose (pre-trained) Water (pre-trained)

30 20 10 0

0 1

2

3

4

5

6

1

7

2

C

4

5

6

7

D 20

**

Average escape latency over all testing days (s)

Average escape latency over all testing days (s)

3

Day

Day

10

0 Ethanol

Sucrose Treatment

Water

20

10

0 Ethanol

Sucrose Treatment

Water

Fig. 2. Stationary-hidden-platform task acquisition of non-trained and pre-trained offspring from pregnant guinea pigs that received chronic oral administration of 4g ethanol/kg maternal body weight/day, isocaloric-sucrose/pair-feeding, or water treatment. Data are presented as mean ± S.E.M of the escape latency on each of the 7days of testing, and collapsed across all test days, for the non-trained (A, C) and pre-trained (B, D) offspring (n = 8 offspring from 8 different litters for each treatment group). ⁎Ethanol group statistically different from the sucrose group (p b 0.05) on day 2 of testing. #Ethanol group statistically different from the water group (p b 0.05) on day 3 of testing. ⁎⁎Ethanol group statistically different from both control groups (p b 0.05).

moving-platform testing phases. It was not possible to record performance in the non-spatial pre-training sessions due to the low lighting in which this testing took place. For these pretraining sessions, swim times were recorded manually. For the stationary-platform and moving-platform trials, data were analyzed using the VideoMot® computer program (TSE Systems Inc). The program provided data on the time latency to find the platform, swim distance, and swim speed during each trial. 2.8. Saliva cortisol concentration On the first day of testing in the stationary-platform task, offspring from each of the treatment groups (both pre-trained and non-trained) were studied for saliva cortisol concentration after exposure to the water maze. These measures were all performed in the afternoon (the nadir of the circadian rhythm for cortisol in the guinea pig). For saliva collection, the cheek pouches of the guinea pig were quickly rinsed with 10ml of water such that no swallowing of water occurred. The animal was returned to a clean cage without food or water for 15 min. Subsequently, a saliva sample (approximately 25 μl) was collected using a round-tip, stainless-steel, oral-gavage needle attached to a syringe. Saliva was collected from each animal approximately 30min after the first exposure to the pool. Each saliva sample was analyzed for cortisol concentration using an enzyme-linked immunoassay (Salimetrics LLC, State College,

PA), as described by the manufacturer. Baseline cortisol concentration was determined by collecting a saliva samples in the afternoon on non-test days. 2.9. Statistical analysis For water-maze experiments and saliva cortisol concentration, data were analyzed using a mixed design-repeated measures analysis of variance (ANOVA) (SPSS v.7.0), followed by the Bonferroni post hoc test for multiple comparisons (Prism 4.0). For the mixed design ANOVA, we compared prenatal treatment and/or pre-training as the between-subject variables, and test day as the within-subject variable. Pregnancy outcome data were analyzed by one-way ANOVA coupled with Newman–Keuls post hoc test for multiple comparisons (Prism 4.0). 3. Results The chronic maternal ethanol regimen employed produced a peak maternal blood ethanol concentration of 346 ± 40 mg/dl on gestational day 57 at 1 h after the second divided dose. The ethanol regimen had no effect on the length of gestation, offspring birth weight, average number of littermates per litter, or distribution of male and female offspring compared with the offspring of the isocaloric-sucrose/pair-fed and water control

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341

A Escape latency (s)

35

Ethanol Sucrose Water

30 25 20

10 5 1

2

3

4

5

6

Day

B 35 Escape latency (s)

1.0

*

#

#

Ethanol Sucrose Water

0.5

0.0 Non-trained

Pre-trained

Fig. 4. Saliva cortisol concentration after acute swim stress in the water maze (day 1 of spatial-hidden-platform task) of non-trained and pre-trained offspring from pregnant guinea pigs that received chronic oral administration of 4g ethanol/kg maternal body weight/day, isocaloric-sucrose/pair-feeding, or water treatment. Data are presented as mean ± S.E.M of the saliva cortisol concentration 30min after completing the last trial on day 1 (n = 8 offspring from 8 different litters for each treatment group). ⁎Non-trained ethanol group statistically different from the two control groups (p b 0.05). #Pre-trained control groups (sucrose and water) statistically different from the non-trained control groups (sucrose and water) (p b 0.05).

3.1. Non-spatial pre-training acquisition

15

0

Ethanol Sucrose Water

30 25 20

The data for water-maze performance during non-spatial pretraining is shown in Fig. 1. A mixed design-repeated measures ANOVA revealed statistically significant main effects for day (F(4, 13) = 16.158, p b 0.0001) and treatment (F(2, 16) = 4.989, p b 0.05), and an interaction between day and treatment (F(8, 28) = 4.069, p b 0.05). Bonferroni post hoc analysis revealed that CPEE offspring had longer swim times on day 3 compared to the isocaloric-sucrose/pair-fed (t = 2.872, p b 0.05) and water (t = 3.351, p b 0.05) control groups (Fig. 1A, B).

15

3.2. Spatial learning: stationary-platform task

10 5 0 1

2

3

4

5

6

Day

C Average escape latency collasped over testing days (s)

1.5

Saliva [Cortisol] (µg/dl)

groups (Table 1). However, offspring birth weight was lower in the ethanol group compared with the sucrose (t = 2.505, p b 0.05) and water (t = 3.651, p b 0.01) treatment groups (Table 1). There was no difference in pregnancy outcome variables (Table 1) or in water-maze performance between the offspring of the isocaloric-sucrose/pair-fed and water control groups, which is consistent with previous reports using the guinea pig model ([18,30]). In addition, during water-maze testing, both body weight and swim speed did not differ significantly between treatment groups (data not shown). All water-maze data are presented as swim time (time latency to find the escape platform), where distance traveled to find the escape platform mirrored the time-latency data.

337

20

*

10

0 Ethanol

Sucrose

Water

Treatment

Fig. 3. Moving-hidden-platform task acquisition of non-trained and pre-trained offspring from pregnant guinea pigs that received chronic oral administration of 4g ethanol/kg maternal body weight/day, isocaloric-sucrose/pair-feeding, or water treatment. Data are presented as mean ± S.E.M of the escape latency on each of the 6 days of testing for the non-trained (A) and pre-trained (B) offspring (n = 8 offspring from 8 different litters for each treatment group), and collapsed across all test days (C) for all offspring (n = 16 offspring from 8 different litters for each treatment group). ⁎Ethanol group statistically different from the two control groups (p b 0.05).

Fig. 2 shows the times to find and mount the hidden platform across the 7days of testing for non-trained (Fig. 2A) and pretrained (Fig. 2B) guinea pig offspring in the three treatment groups. The mixed design-repeated measures ANOVA revealed significant main effects of pre-training (F(1, 9) = 14.644, p b 0.05), day (F(6, 4) = 15.235, p b 0.01), and prenatal treatment (F(2, 9) = 12.837, p b 0.05), and a significant day-by-treatmentby-pre-training interaction (F(6, 4) = 4.913, p b 0.05). Post hoc testing showed that the day-by-treatment-by-pre-training interaction was due to a differential improvement in performance in the CPEE offspring resulting from pre-training. Bonferroni post hoc analysis demonstrated that non-trained CPEE offspring had longer swim times on days 2 (t = 2.802, p b 0.05) and 3 (t = 2.803, p b 0.05) compared to the isocaloric-sucrose/pairfed and water controls, respectively, whereas the swim times for offspring in the two control groups were not different from each other on any test day (Fig. 2A, C). In contrast, pre-trained CPEE offspring were not different from pre-trained control offspring on any test day (Fig. 2B, D). Similarly, Bonferroni post hoc analysis revealed statistically significant decreases in swim times between pre-trained CPEE offspring and non-trained CPEE offspring on days 1 (t = 3.715, p b 0.01), 2 (t = 3.805, p b 0.01), and 3 (t = 2.812, p b 0.05). For the isocaloric-sucrose/

338

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341

pair-fed group, pre-training improved performance in the water maze only on the first test day (t = 5.192, p b 0.001).

t = 2.920, p b 0.05 for water). In contrast, the saliva cortisol concentration after exposure to the water maze was not different between non-trained and pre-trained CPEE offspring (Fig. 4).

3.3. Probe trial 4. Discussion There was no effect of prenatal treatment on any measure during the 60-s probe trial. The mixed-design repeated measures ANOVA revealed that animals in all three treatment groups exhibited a spatial bias for the quadrant in which the escape platform had previously been located (F(3, 168) = 20.12, p b 0.01). These values were 32%, 33% and 34% time spent in the platform quadrant for the ethanol, sucrose and water treatment groups, respectively. There were no differences in the percent time spent in the non-platform quadrants for any treatment group (data not shown). 3.4. Spatial learning: moving-platform task Fig. 3 shows the times to find and mount the hidden platform across the 6 days of testing for non-trained (Fig. 3A) and pretrained (Fig. 3B) guinea pig offspring in the three treatment groups. The mixed design-repeated measures ANOVA revealed statistically significant main effects of prenatal treatment (F(2, 20) = 3.342, p b 0.05) and day (F(5, 16) = 3.424, p b 0.05) on performance in the moving-platform version of the watermaze task. Offspring in all three treatment groups exhibited an improvement in performance over the 6 days of testing (Fig. 3A, B). However, there was no effect of pre-training on performance in the moving-platform version of the task (F(1, 20) = 2.008, p = 0.172); therefore, the data were collapsed across test days to determine the effect of prenatal treatment. Bonferroni post hoc analysis revealed that CPEE offspring had significantly higher escape latency compared with both the isocaloric-sucrose/pairfed (t = 2.531, p b 0.05) and water (t = 2.661, p b 0.05) control groups (Fig. 3C). 3.5. Saliva cortisol concentration Baseline samples obtained in the late afternoon (the nadir for saliva cortisol concentration in the guinea pig) on days when animals were not exposed to the water maze revealed no difference in saliva cortisol concentration between ethanol (0.229 ± 0.05μg/dl) and the isocaloric-sucrose/pair-fed (0.245 ± 0.06 μg/dl) and water (0.255 ± 0.06 μg/dl) control groups. However, non-trained CPEE offspring had higher saliva cortisol concentration than both non-trained isocaloric-sucrose/pair-fed (t = 3.023, p b 0.01) and water (t = 2.494, p b 0.05) control offspring after the first exposure to the water maze (Fig. 4). For saliva samples obtained after exposure to swim stress in the water maze, the mixed design ANOVA demonstrated a main effect of pre-training (F(1, 36) = 17.376, p b 0.001) and an interaction between prenatal treatment and pre-training (F(1, 36) = 3.956, p b 0.05). Bonferonni post hoc tests revealed that the interaction between treatment and pre-training was caused by the saliva cortisol concentration in non-trained control offspring being lower than that of pre-trained control offspring (t = 3.333, p b 0.01 for isocaloric-sucrose/pair-fed and

Enrichment experiments, conducted with mice, rats and primates, typically involve increased social interaction among offspring, more variety and variability in sensory experience, more opportunity for varied locomotor activity and exploration, and sometimes even specific learning trials or training [12]. In the current study, we investigated whether behavioural intervention, in the form of non-spatial pre-training, can mitigate deficits in water-maze performance in CPEE offspring in multiple water-maze paradigms. We demonstrate that in the stationary-hidden-platform task of the Morris water maze, CPEE offspring exhibited a slower rate of acquisition compared with control offspring. However, with repeated exposure to the task, all animals reached an equivalent level of performance, with similar spatial bias for the platform quadrant during the probe trial 24 h later. This spatial bias for the pool quadrant in which the hidden platform had previously been located indicates that all animals learned the platform location (spatial-learning), as opposed to simply developing a search strategy to find the platform. This finding is supported by previous studies conducted in the rat, demonstrating spatialacquisition deficits caused by prenatal ethanol exposure [3,41] as well as by binge-like exposure to ethanol during the neonatal brain growth spurt [21,24]. The water-maze paradigm is complex; it is a task that involves stress, visuo-spatial integration, and motor abilities [4]. Alterations in any of these parameters could contribute to poor performance, and may confound any interpretation of deficits in spatial learning. Non-spatial pre-training, during which the animal was exposed to the behavioural requirements of the task, improved water-maze performance of CPEE offspring to control levels in the stationary-platform task. Non-spatial pre-training afforded improvement in the performance of control offspring only on the first day of spatial learning, but significantly improved the performance of CPEE offspring over the first three days of testing. In order to determine whether non-spatial pre-training produced a global benefit in multiple water-maze paradigms, we utilized the more difficult daily moving-hidden-platform task. For this movingplatform task, the animal needs to learn a new platform location during the first trial on each experimental day and, in the second trial, remember this new platform location. This moving-platform task paradigm has been shown to be a very sensitive indicator of spatial-learning deficits in rat models of CPEE [32]. Interestingly, CPEE offspring exhibited longer escape latencies in the moving-platform version of the water maze, and, in contrast to the stationary-platform task, this deficit was not affected by non-spatial pre-training. These data suggest that non-spatial pre-training provides a functional benefit in relatively simple spatial-learning tasks, but that more difficult tasks will continue to reveal behavioural deficits. It will be of interest to determine if more aggressive interventions

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341

have an impact on CPEE-induced performance deficits in more complex tasks. For example, one might combine environmental enrichment during rearing with the behavioural intervention of pre-training, or add pharmacological interventions such as neurotrophins and/or cognition-enhancing drugs. Combination therapies may be expected to have additive or synergistic effects to improve cognitive abilities in individuals adversely affected by CPEE. To perform the water-maze task successfully, the animal must demonstrate two essential behaviours. First, the animal must learn to swim away from the wall of the pool, as the escape platform is always located in the centre of one of the quadrants. Second, the animal must learn that the platform is the only escape from the pool and that it must climb onto and remain on the platform. These general behavioural strategies must be learned in order for the animal to acquire spatial information about the platform location. Previous studies conducted in the rat indicate that exposure to all aspects of the water-maze task are required to achieve the full effect of non-spatial pre-training, and that individual components of pre-training, such as handling or swimming/exercise, are not sufficient to achieve the complete effect [6,15]. During the non-spatial pre-training phase of the experiment, CPEE offspring demonstrated delayed acquisition of behavioural strategies necessary to perform the water-maze task successfully (Fig. 1). It seems reasonable, therefore, to suggest that this decreased ability to acquire the appropriate behavioural strategies may have contributed to the early impairment in performance in the stationary-platform task for non-trained CPEE offspring. The benefit derived from nonspatial pre-training, therefore, may be due to the additional opportunity to acquire the most appropriate behavioural strategy to perform the water-maze task. These data also emphasize the importance of testing the experimental model in a number of paradigms, as water-maze learning is multi-faceted. In order to understand the mechanism of the CPEE-induced impaired acquisition of behavioural strategies and spatial mapping, saliva cortisol concentration, a marker of activation of the HPA axis, was determined in both non-trained and pretrained offspring after acute swim stress. A recent study conducted in the rat found that differences in the stress response to forced swim are associated with differences in water-maze performance during the initial acquisition trials [1]. Thus, we wanted to test whether there was any association between CPEE-induced learning impairment and the stress response to exposure to the water maze. To do this, we used saliva cortisol concentration as a marker of HPA axis activity after swimming in the water maze. The quantification of saliva cortisol has many advantages over plasma cortisol, as it is a non-invasive procedure and provides direct determination of the free (unbound) cortisol concentration. It has been established previously that there is a linear relationship between free plasma cortisol concentration and saliva cortisol concentration in the guinea pig [10] and the human [23]. Interestingly, in nontrained CPEE offspring, saliva cortisol concentration was increased 30 min after the first exposure to the water maze, compared to non-trained control offspring. Although we collected saliva samples at only one time point in this study,

339

we recognize that repeated saliva sampling to determine the time course of cortisol concentration induced by swim stress would be more informative, and this approach will be pursued in future studies. Altered responsiveness of the HPA axis also has been demonstrated in rat offspring after CPEE and in response to a variety of stressors [26,35,36,38,39]. In the rat, hyperresponsiveness to restraint stress in prenatal ethanol-exposed offspring is not associated with the rise in corticosterone concentration after stress, but rather with a slower recovery of the plasma corticosterone concentration to baseline [37,39]. We recently demonstrated that CPEE induces very high maternal saliva cortisol concentration during gestation, and causes alterations in glucocorticoid receptor function in the hippocampus of the young adult guinea pig [19]. Moreover, there is evidence for dysregulation of the HPA axis in human infants who were exposed to ethanol prenatally [20]. Specifically, 13month-old infants with a history of prenatal exposure to ethanol exhibit hyper-responsiveness of the HPA axis to activation by stress [20]. Furthermore, it has been demonstrated in the nonhuman primate that CPEE induces hyper-responsiveness of the HPA axis to activation by social-separation stress [33]. Thus, hyper-responsiveness of the HPA axis to activation by stress appears to be common to CPEE across several mammalian species and types of environmental manipulations. It has been suggested that non-spatial pre-training may offer protection from water-maze performance deficits in animals by way of a reduction in the stress induced by exposure to the water maze [1]. In the rat, non-spatial pre-training reduces the stress response induced by exposure to the water maze, as measured by the plasma corticosterone concentration, and this coincides with more efficient acquisition of the task [1]. In control guinea pig offspring, however, non-spatial pre-training led to an increase in saliva cortisol concentration compared to non-trained animals. These data suggest that, in the guinea pig, repeated exposure to forced swimming in the water maze during non-spatial pretraining results in a sensitization of the stress response, which is in contrast to the habituation in the stress response reported for the rat [1]. There is tremendous inter-individual variability in the response to stress, and the direction of change in HPA axis activation after repeated episodes of stress is influenced by both genetic factors and the intensity of the stimulus used to induce stress. In human subjects exposed to repeated episodes of mild psychosocial stress, approximately half of the subjects demonstrate the expected habituation of HPA axis activation, whereas 16% of subjects demonstrate sensitization of HPA axis activation in response to repeated exposure to the same stressor [40]. In the rat, repeated exposure to low-intensity electrical shock results in habituation of the stress response [29]. However, HPA axis activation in rats given higher intensity shock does not habituate, and instead demonstrates an increased response indicative of sensitization [29]. These data suggest that, for the guinea pig, exposure to the water maze may be a much more stressful experience than it is for the rat, leading to sensitization of the stress response after repeated swimming episodes. Moreover, it appears that, in control guinea pig offspring, improved water-maze performance over repeated trials is not

340

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341

associated with habituation of the stress response, suggesting that dissociation exists between the behavioural adaptation (improved performance in the water maze) and the neuroendocrine response to the stress of the test situation. In contrast to the results obtained in control guinea pig offspring, CPEE offspring that had undergone non-spatial pretraining did not exhibit an increase in the saliva cortisol response 30min after swim stress in the water maze compared with non-trained CPEE offspring. Future experiments will be aimed at elucidating the time course of the saliva cortisol response in CPEE guinea pig offspring after swim stress in the water maze. This will enable us to determine if this effect represents a deficit in sensitization of the HPA axis to repeated episodes of stress. Alternatively, it may be the case that CPEE guinea pig offspring are already at or near the ceiling for stressinduced activation of the HPA axis. In summary, we have provided evidence that non-spatial pretraining mitigates CPEE-induced deficits in the stationaryplatform spatial-learning task. Thus, at least some of the cognitive deficits produced by CPEE can be reversed by behavioural intervention. In addition, impairments in watermaze performance in CPEE guinea pig offspring may not be due entirely to deficits in spatial mapping, and may also involve delayed acquisition of the appropriate behavioural strategies. Furthermore, CPEE offspring demonstrate an altered saliva cortisol response to swim stress, suggesting long-lasting alterations in HPA axis function. Finally, a key consequence of CPEE may be impairment in the initial rate of acquisition of new learning, rather than in the overall ability to learn new information. Acknowledgements Supported by the Canadian Institutes of Health Research, Grant No. MOP-15150. Umar Iqbal was the recipient of a Wilson Fellowship from the School of Graduate Studies and Research, Queen's University. References [1] J. Beiko, R. Lander, E. Hampson, F. Boon, D.P. Cain, Contribution of sex differences in the acute stress response to sex differences in water maze performance in the rat, Behav. Brain Res. 151 (2004) 239–253. [2] R.F. Berman, J.H. Hannigan, Effects of prenatal alcohol exposure on the hippocampus: spatial behavior, electrophysiology, and neuroanatomy, Hippocampus 10 (2000) 94–110. [3] B.A. Blanchard, E.P. Riley, J.H. Hannigan, Deficits on a spatial navigation task following prenatal exposure to ethanol, Neurotoxicol. Teratol. 9 (1986) 253–258. [4] D.P. Cain, D. Saucier, The neuroscience of spatial navigation: focus on behavior yields advances, Rev. Neurosci. 7 (1996) 215–231. [5] D.P. Cain, C. Finlayson, F. Boon, J. Beiko, Ethanol impairs behavioural strategy use in naïve rats but does not prevent spatial learning in the water maze in pretrained rats, Psychopharmacology (Berl) 164 (2002) 1–9. [6] C. Caldji, C.H. Vanderwolf, The effects of different types of pretraining on the rat's retention performance in a swim-to-platform task following administration of scopolamine, Behav. Brain Res. 80 (1996) 217–220. [7] B.R. Christie, S.E. Swann, C.J. Fox, D. Froc, S.E. Lieblich, C. Redila, A. Webber, Voluntary exercise rescues deficits in spatial memory and long-

[8]

[9] [10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22] [23]

[24]

[25]

[26]

[27] [28]

[29]

term potentiation in prenatal ethanol-exposed male rats, Eur. J. Neurosci. 21 (2005) 1719–1726. H.C. Dringenberg, D.P. Richardson, J.F. Brien, J.N. Reynolds, Spatial learning in the guinea pig: cued vs. non-cued learning, sex differences, and comparison with rats, Behav. Brain Res. 124 (2001) 97–101. H. Elvidge, Production of dated pregnant guinea pigs without post partum matings, J. Inst. Anim. Technol. 23 (1972) 111–117. M. Fenske, The use of salivary cortisol measurements for the non-invasive assessment of adrenal cortical function in guinea pigs, Exp. Clin. Endocrinol. Diabetes 105 (1997) 163–168. M.A. Gibson, N.S. Butters, J.N. Reynolds, J.F. Brien, Effects of chronic prenatal ethanol exposure on locomotor activity, and hippocampal weight, neurons, and nitric oxide synthase activity of the young postnatal guinea pig, Neurotoxicol. Teratol. 22 (2000) 183–192. W.T. Greenough, Enduring brain effects of differential experience and training, in: M.R. Rosenzweig, E.L. Bennett (Eds.), Neural Mechanisms of Learning and Memory, MIT Press, Cambridge, MA, 1976, pp. 255–278. J.H. Hannigan, R.F. Berman, Amelioration of fetal-alcohol related neurodevelopmental disorders in rats: exploring pharmacological and environmental treatments, Neurotoxicol. Teratol. 22 (2000) 103–111. J.H. Hannigan, R.F. Berman, C.S. Zajac, Environmental enrichment and the behavioural effects of prenatal exposure to alcohol in rats, Neurotoxicol. Teratol. 15 (1993) 261–266. T.E. Hoh, D.P. Cain, Fractionating the nonspatial pretraining effect in the water maze task, Behav. Neurosci. 111 (1997) 1285–1291. T. Hoh, F. Boon, J. Beiko, S. Weiss, D.P. Cain, Complex behavioral strategy and reversal learning in the water maze without NMDA receptordependent long-term potentiation, J. Neurosci. 19 (1999) 1–5. T.E. Hoh, B. Kolb, A. Eppel, C.H. Vanderwolf, D.P. Cain, Role of the neocortex in the water maze task in the rat: a detailed behavioral and Golgi–Cox analysis, Behav. Brain Res. 138 (2003) 81–94. U. Iqbal, H.C. Dringenberg, J.F. Brien, J.N. Reynolds, Chronic prenatal ethanol exposure alters hippocampal GABAA receptors and impairs spatial learning in the guinea pig, Behav. Brain Res. 150 (2004) 117–125. U. Iqbal, J.F. Brien, S. Banjanin, M.H. Andrews, S.G. Matthews, J.N. Reynolds, Chronic prenatal ethanol exposure alters glucocorticoid signaling in the hippocampus of the postnatal guinea pig, J. Neuroendocrinol. 17 (2005) 600–608. S.W. Jacobson, J.T. Bihun, L.M. Chiodo, Effects of prenatal alcohol and cocaine exposure on infant cortisol levels, Dev. Psychopathol. 11 (1999) 195–208. T.B. Johnson, C.R. Goodlett, Selective and enduring deficits in spatial learning after limited neonatal binge alcohol exposure in male rats, Alcohol. Clin. Exp. Res. 26 (2002) 83–93. K.L. Jones, D.W. Smith, Recognition of the fetal alcohol syndrome in early infancy, Lancet 2 (1973) 999–1001. J.P. Kahn, D.R. Rubinow, C.L. Davis, M. Kling, R.M. Post, Salivary cortisol: a practical method for evaluation of adrenal function, Biol. Psychiatry 23 (1988) 335–349. S.J. Kelly, C.R. Goodlett, S.A. Hulsether, J.R. West, Impaired spatial navigation in adult female by not adult male rats exposed to alcohol during the brain growth spurt, Behav. Brain Res. 27 (1988) 247–257. K.A. Kimura, J.N. Reynolds, J.F. Brien, Ethanol neurobehavioural teratogenesis and the role of the hippocampal glutamate-N-methyl-Daspartate receptor-nitric oxide synthase system, Neurotoxicol. Teratol. 22 (2000) 607–616. S.Y. Lee, T. Imaki, W. Vale, C.L. Rivier, Effect of prenatal exposure to ethanol on the activity of the hypothalamic-pituitary-adrenal axis of the offspring: importance of the time of exposure to ethanol and possible modulating mechanisms, Mol. Cell. Neurosci. 1 (1990) 168–177. R.G.M. Morris, Spatial localization does not require the presence of local cues, Learn. Mem. 12 (1981) 239–260. R.G.M. Morris, Synaptic plasticity and learning: selective impairment in rats and blockade of long-term potentiation in vivo by the N-methyl-Daspartate receptor antagonist, AP5, Nature 316 (1989) 774–776. D.L. Pitman, J.E. Ottenweller, B.H. Natelsen, Effect of stressor intensity on habituation and sensitization of glucocorticoid responses in rats, Behav. Neurosci. 104 (1990) 28–36.

U. Iqbal et al. / Neurotoxicology and Teratology 28 (2006) 333–341 [30] D.P. Richardson, M.L. Byrnes, J.F. Brien, J.N. Reynolds, H.C. Dringenberg, Impaired acquisition in the water maze and hippocampal long-term potentiation after chronic prenatal ethanol exposure in the guinea pig, Eur. J. Neurosci. 16 (2002) 1593–1598. [31] M.R. Rosenweig, E.L. Bennett, Psychobiology of plasticity: effects of training and experience on brain and behaviour, Behav. Brain Res. 78 (1996) 57–65. [32] D.D. Savage, M. Becher, A.J. de la Torre, R.J. Sutherland, Dose-dependent effects of prenatal ethanol exposure on synaptic plasticity and learning in mature offspring, Alcohol. Clin. Exp. Res. 26 (2002) 1752–1758. [33] M.L. Schneider, C.F. Moore, G.W. Kraemer, Moderate level alcohol during pregnancy, prenatal stress, or both and limbic-hypothalamicpituitary-adrenocortical axis response to stress in rhesus monkeys, Child Dev. 75 (2004) 96–109. [34] N.A.E. Steenaart, D.W. Clarke, J.F. Brien, Gas–liquid chromatographic analysis of ethanol and acetaldehyde in blood with minimal artifactual acetaldehyde formation, J. Pharmacol. Methods 14 (1985) 199–212. [35] J. Weinberg, Differential effects of prenatal ethanol on males and females, Alcohol. Clin. Exp. Res. 12 (1988) 647–652.

341

[36] J. Weinberg, Prenatal ethanol exposure alters adrenocortical response to predictable and unpredictable stress, Alcohol 9 (1992) 427–432. [37] J. Weinberg, S. Bezio, Alcohol-induced changes in pituitary-adrenal activity during pregnancy, Alcohol. Clin. Exp. Res. 11 (1987) 274–280. [38] J. Weinberg, P.V. Gallo, Prenatal ethanol exposure: pituitary-adrenal activity in pregnant dams and offspring, Neurobehav. Toxicol. Teratol. 4 (1982) 515–520. [39] J. Weinberg, C.K. Kim, W. Yu, Early handling can attenuate adverse effects of fetal ethanol exposure, Alcohol 12 (1995) 317–327. [40] S. Wust, I.S. Federenko, E.F.C. van Rossum, J.W. Koper, D.H. Hellhammer, Habituation of cortisol responses to repeated psychosocial stress-further characterization and impact of genetic factors, Psychoneuroendocrinology 30 (2005) 199–211. [41] B. Zimmerberg, H.L. Sukel, J.D. Stekler, Spatial learning of adult rats with fetal alcohol exposure: deficits are sex-dependent, Behav. Brain Res. 42 (1991) 49–56.