- Email: [email protected]
Interaction between the cholinergic system and CRH in the modulation of spatial discrimination learning in mice
Brain Research 906 (2001) 46–59 www.elsevier.com / locate / bres
Research report
Interaction between the cholinergic system and CRH in the modulation of spatial discrimination learning in mice Thomas Steckler*, Florian Holsboer Max Planck Institute of Psychiatry, Kraepelinstr. 2 – 10, D-80804 Munich, Germany Accepted 17 April 2001
Abstract Both cholinergic and CRH systems have been linked to cognitive processes such as learning and memory, and neuroanatomical as well as neurochemical evidence suggests important interactions between these two systems. Moreover, recent reports of pro-mnestic effects of CRH open the possibility that CRH could have beneficial effects in animals with cholinergic dysfunction. In a first experiment, spatial discrimination of C57BL / 6 mice treated with various doses of scopolamine (0.5–2.0 mg / kg IP) was tested in a two-choice water maze task. Scopolamine, but not methylscopolamine, impaired accuracy and decreased responsivity. In contrast, similar doses of the nicotinic antagonist mecamylamine had no effect on choice accuracy but altered responsivity, as indicated by increased errors of omission and a reduction in swim speed during early experimental stages. ICV CRH (0.5–1.0 mg) also failed to significantly affect accuracy, but a strong tendency was observed to impair percentage correct responses. Measures of responsivity, such as errors of omission, choice latency and distance traveled, and of thigmotaxis were not significantly affected by CRH. However, initial swim speed was reduced by the peptide. Combined treatment with scopolamine (0.5 mg / kg IP) and CRH (0.5 mg ICV) had only mild, and primarily independent, effects, but overall suggested that concomitant blockade of muscarinic receptors and activation of the CRH system would rather act synergistically to disrupt spatial discrimination learning. Synergistic effects were also observed when animals receiving a combination of mecamylamine (2.0 mg / kg IP) and CRH (0.5 mg ICV) were tested, both in terms of responsivity and thigmotaxis, and there was limited evidence that part of these effects were potentiating. Thus, the cholinergic and CRH systems interact in the modulation of learning, but CRH, contrary to prediction, worsens the impairment caused by cholinergic blockade. 2001 Elsevier Science B.V. All rights reserved. Theme: Neural basis of behavior Topic: Neuropeptides and behavior Keywords: Acetylcholine; Corticotropin-releasing hormone; CRH; Interaction; Learning; Memory; Spatial discrimination
1. Introduction Neurochemical and anatomical evidence suggests that there are important reciprocal interactions between the cholinergic and the corticotropin-releasing hormone (CRH) systems: first, frontoparietal cortical CRH release has been reported to be inhibited by acetylcholine (ACh) [69], while hypothalamic CRH activity is enhanced following ACh administration [5,47,69]. Lesions of the nucleus basalis, the main basal forebrain cholinergic nucleus projecting to neocortical areas, on the other hand, increase parietal *Corresponding author. Present address: Janssen Research Foundation, Turnhoutseweg 30, B-2340 Beerse, Belgium. Tel.: 132-14-60-7373; fax: 132-1460-22-6121. E-mail address: [email protected] (T. Steckler).
CRH-like immunoreactivity some months post-lesioning [44]. Both nicotinic and muscarinic processes appear to be involved: nicotine has been reported to activate Fos expression in CRH-positive neurons in the bed nucleus of the stria terminalis, the central nucleus of the amygdala, the dorsal raphe and Barrington’s nucleus [42], and to stimulate CRH mRNA expression in immortalized amygdala cells [36]. At the level of the hypothalamic– pituitary–adrenal (HPA) axis, nicotine has also been reported to activate CRH neurons within the paraventricular hypothalamic nucleus (PVN) [42] and to stimulate CRH secretion from rat hypothalami in vitro [7], but the main action of nicotine to stimulate ACTH release may be indirect via activation of noradrenaline release at the level of the PVN [26,41]. Chronic muscarinic blockade, on the other hand, leads to a significant increase in CRH receptors
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02555-0
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
in rat frontoparietal cortex, but not in the hippocampus [19]. Thus, a reduction in cholinergic muscarinic activity may lead to an enhanced activity of the CRH system, at least at cortical level, which contrasts the stimulation of CRH activity by activation of nicotinic receptors at subcortical level. Furthermore, acute peripheral muscarinic blockade has been reported to reduce ICV CRH-induced tachycardia through parasympathetic mechanisms [49], pointing towards important interactions at the level of the autonomic nervous system as well. Second, one of the CRH receptor subtypes, the CRH 1 receptor, is highly abundant in mouse and rat cholinergic basal forebrain and brainstem nuclei [6,9,10] and a high degree of co-localization of choline–acetyltransferase (ChAT) and CRH receptor-like immunoreactivity has been observed in both the murine basal forebrain and brainstem, with the notable exception of the nucleus basalis magnocellularis [61]. In rats, intracerebroventricular (ICV) CRH produces moderate to strong stimulation of Fos expression within the basal forebrain and brainstem nuclei, i.e., within the medial septum and diagonal band of Broca, the laterodorsal and the pedunculopontine tegmental nuclei, while only relatively weak Fos expression is induced within the substantia innominata [6]. Furthermore, ICV but not peripheral injections of CRH increase hippocampal ACh release [16,17], CRH alters hippocampal theta rhythm, possibly through interaction with cholinergic systems [37], and it may be suggested that CRH induces these effects directly at the level of the cholinergic basal forebrain via CRH 1 receptor activation. Various stressors have also been reported to increase hippocampal ACh release [1,17,22,27,33]. These stressinduced changes in hippocampal cholinergic activity reach a maximum 20–30 min after stress exposure and have been suggested to be independent of the hypothalamic– pituitary–adrenal (HPA) axis [33] (but see Ref. [27]). Obviously, one potential candidate mediating this response could be centrally released CRH. Third, CRH may interact with cholinergic mechanisms by stimulating common targets, as it has been shown that activation of the M1 muscarinic receptor subtype and of CRH receptors synergistically increase cAMP production in membranes from rat frontal cortex [48]. Since septohippocampal ACh release is activated following sensory stimulation [1,8,22,34], it may be argued that the central interaction between CRH and the cholinergic septohippocampal system serves to maintain appropriate processing of environmental information, in particular under stressful conditions. This in turn would suggest that the interaction between the two systems is of relevance for performance in a range of cognitive tasks, including spatial learning and memory paradigms. Indeed, both the cholinergic and the CRH systems have been, independently from each other, implicated in the modulation of learning and memory [3,11,13,18,28,29,32, 39,40,45,54,56–58,64,66,67]. Some of these studies sug-
47
gest that CRH may have pro-mnemonic effects [3,10, 29,32,39,54], and it is conceivable that CRH enhances cognitive function through interaction with the cholinergic system. However, the exact nature of this interaction between the two systems in the modulation of these processes remains unclear at the behavioral level. This question was addressed in the present study, testing the effects of combined cholinergic blockade and CRH stimulation in mice performing on a water maze spatial discrimination paradigm. In particular, we tested the hypothesis that ICV administration of CRH would attenuate the deleterious effects of cholinergic blockade caused by administration of either the muscarinic antagonist scopolamine or the nicotinic antagonist mecamylamine.
2. Material and methods
2.1. Animals Male 3-month-old C57BL / 6CrlBR mice (Charles River, Sulzfeld, Germany) were housed individually and maintained on a 12:12 h light / dark cycle (lights on at 07:00). Experiments were conducted during the light phase of the cycle. Food and water were available ad libitum. All animal procedures were approved by the Ethical Committee on Animal Care and Use of the Government of Bavaria, Germany.
2.2. Surgery Mice were anaesthetized with a mixture of ketamine (10%, 0.5 ml / kg, IP) and rompun (2.0 ml / kg, IP) and placed in a stereotaxic frame (TSE, Bad Homburg, Germany). A stainless steel tube (27 gauge) was implanted into the right lateral ventricle at the following coordinates according to the atlas of Franklin and Paxinos [25]: AP, 0.6 mm posterior to bregma, L, 1.8 mm lateral to the midline, V, 2.2 mm below the exposed dura mater. The tube was fixed to the skull with two stainless steel screws, placed in the vicinity of the tube, and dental cement applied to cover the screws and tube. A stainless steel stylet, with a length identical to the tube, was inserted to keep the cannula patent. Mice were allowed 1 week recovery before testing. On completion of the experiments, animals were anaesthetized with isofluran and correct placement of the cannula was confirmed by ICV injection of 5 ml of cresyl violet solution. Only animals with correct placement of the cannulae were included for further analysis.
2.3. Drugs Scopolamine hydrobromide, scopolamine methylbromide and mecamylamine hydrochloride (all Sigma Chemicals, Deisenhofen, Germany) were dissolved in saline and administered intraperitoneally (IP). CRH (Ba-
48
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
chem Biochemica, Heidelberg, Germany) was administered over a period of 2 min via an injection cannula which protruded 1.0 mm below the guide tube, and which was connected to a 10 ml Hamilton syringe.
2.4. Apparatus A circular swimming pool (80 cm in diameter, 30 cm high, white plastic) was filled to a depth of 20 cm with water (21618C; rendered opaque by the addition of a non-toxic dye). At the outside of the maze eight start boxes (10310326 cm, also filled with water) and fitted with sliding doors were fixed, from which the animal could swim into the center of the maze when the door was raised. Two identically looking circular platforms with white surface and dark gray rim were used (each 10 cm in diameter, protruding approximately 0.5 cm above the water surface). One platform was stable and provided support (‘correct platform’), the other was floating and sank when a mouse tried to climb on it (‘incorrect platform’). Performance was recorded by a video tracking system (TSE).
group) were tested after treatment with saline, scopolamine hydrobromide (0.5, 1.0 or 2.0 mg / kg IP) or scopolamine methylbromide (2.0 mg / kg), 20 min prior to each session. In a second experiment, four groups of mice (n58 per group) received treatment with saline or mecamylamine hydrochloride (0.5, 1.0 or 2.0 mg / kg IP), 20 min prior to each session. In the third experiment, three groups of cannulated animals received ICV infusions of saline (n59), 0.5 mg or 1.0 mg CRH (n57 and 6, respectively), 30 min before training. In Experiment 4, four groups of cannulated animals were first treated ICV with 0.5 mg CRH or vehicle (30 min prior to training), followed by an IP injection with scopolamine hydrobromide (0.5 mg / kg) or saline (20 min prior to training), resulting in the following treatment combinations: group saline / saline (n510), group saline / scopolamine (n59), group CRH / saline (n58), and group CRH / scopolamine (n510). In a final experiment, the interaction between CRH 0.5 mg ICV, 30 min prior to test, and mecamylamine (2.0 mg / kg), 20 min prior to test, was studied in four groups: group saline / saline (n59), group saline / mecamylamine (n58), group CRH / saline (n58), and group CRH / mecamylamine (n57).
2.5. Training procedure Mice were trained over five sessions (one session per day, 10 trials per session) to choose between the two platforms (stable vs. unstable) [2,68]. On each trial, the stable platform (correct platform) remained in the same position (approx. 10 cm distance to the wall), while the unstable platform (incorrect platform) changed position from trial to trial in a pseudorandom manner (five possible positions). It was ensured that the spatial relationship between the platforms did not consistently reward turns into one direction, and that the distance between the start position and each of the two platforms was equal over the ten trials. A trial started by placing a mouse in one of six possible start boxes (with the door closed) in pseudorandom sequence. Then the door was raised and the animals had access to the open field of the maze. Tracking started automatically when the animal swum over a start area located at the rim of the open field, directly in front of the start box, i.e., data recording started as soon as the animal had left the start box. All boxes except the start positions in front of and opposite to the correct platform were used. A trial terminated when a mouse climbed onto one of the two platforms or after 30 s. If a mouse climbed onto the stable platform within 30 s, it was allowed to stay there for another 10 s before it was returned to the holding cage. If an animal made an incorrect choice (climbing the incorrect platform) or after 30 s had lapsed, it was gently placed on the correct platform and allowed to stay there for 10 s before it was returned to the holding cage. Animals were trained in squads of four. ITI’s ranged from 2 to 4 min and each session lasted approximately 30 min. In Experiment 1, five groups of animals (n510 per
2.6. Behavioral measures A choice was made if an animal touched a platform with its forepaws or its snout. The occasional incident of brushing past the floating platform was not considered a choice. Percentage correct choices, errors of omission (i.e., the number of trials the animal failed to make a choice within 30 s), average choice latency and choice distance, swim speed, and measures of thigmotactic behavior, i.e. the relative distance swum and the relative time spent in the margin, were calculated. If an error of omission was made, no latency or distance traveled were scored. In contrast, indices of thigmotaxis were calculated for all trials as thigmotaxis may be one factor leading to errors of omission.
2.7. Data analysis Data were transformed as appropriate (arcsine after division by 100, all percentage measures; logarithmic, latency, distance and speed measures; square-root after addition of 0.5, errors of omission) and analyzed by repeated measures ANOVA, including one-factor (treatment) independent measures and a two-factor mixed measures analysis, with session as the repeated measure (treatment3session). Data from the last two experiments were analyzed using a three-factor analysis (treatment3 treatment; treatment3treatment3session), which allowed to dissociate main treatment effects. If ANOVA revealed a significant difference, this was followed by post-hoc testing using the Tukey–Kramer procedure.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
3. Results
3.1. Experiment 1: Water maze spatial discrimination after muscarinic blockade 3.1.1. Accuracy There was a main effect of dose on percentage correct responses (F(4.45)57.04, P,0.001; Fig. 1A). Post-hoc testing failed to reveal a difference between methylscopolamine- and saline-treated animals, while all scopolamine-treated groups differed significantly from saline, but not from each other. Furthermore, ANOVA showed a dose3session interaction for this measure (F(16.180)52.22, P50.012): groups started to differ from the second session onwards, with only the two highest doses of scopolamine being significantly different from saline during this session, while all three groups treated with scopolamine differed from saline during session 3. Furthermore, all three groups treated with scopolamine performed at lower accuracy level than methylscopolamine during the last training session. 3.1.2. Responsivity Overall, choice latency was increased by all three doses of scopolamine relative to saline treatment (F(4.45)52.62, P50.047; mean choice latency over sessions: saline: 3.6160.27 s, scopolamine 0.5 mg / kg: 4.7760.34 s, 1.0 mg / kg: 4.4760.17 s, 2.0 mg / kg: 4.8860.42 s, methylscopolamine: 4.5860.50 s). Moreover, there was a dose3session interaction for this measure (F(16.180)5 1.94, P50.028), and further post-hoc testing indicated that the group treated with the highest dose of scopolamine had a significantly longer choice latency than saline-treated animals during the initial training session (mean choice latency during first session: saline: 6.7161.09 s, scopolamine 0.5 mg / kg: 10.1861.46 s, 1.0 mg / kg: 9.0360.84 s, 2.0 mg / kg: 11.6261.33 s, methylscopolamine: 8.6461.05 s). Likewise, scopolamine increased choice distance at every dose tested relative to saline (Fig. 1B). In addition, choice distance was also increased in the 0.5 mg and 2.0 mg scopolamine-treated groups relative to the performance of animals treated with methylscopolamine (F(4.45)56.87, P,0.001), while the dose3session interaction did not reach significance for this parameter (P.0.05). Scopolamine induced a dose-dependent increase in errors of omission (F(4.45)52.65, P50.046), and the highest dose (2.0 mg / kg) caused significantly more omissions than saline or methylscopolamine. There was a dose3session interaction (F(16.180)54.02, P50.004), and further analysis indicated that this high-dose effect was restricted to the first training session (Fig. 1C), while groups rapidly approached zero errors of omission from session two onwards and did not differ during subsequent sessions. Overall, swim speed did not differ between groups
49
(P.0.05; mean swim speed over sessions: saline: 20.6760.51 cm / s, scopolamine 0.5 mg / kg: 20.5860.50 cm / s, 1.0 mg / kg: 20.8760.42 cm / s, 2.0 mg / kg: 20.6460.62 cm / s, methylscopolamine: 21.3760.46 cm / s).
3.1.3. Thigmotaxis The relative time spent in the margin of the maze was decreased by all three doses of scopolamine when compared to saline (F(4.45)55.85, P50.001; Fig. 1D). However, analysis of the relative distance travelled failed to show an effect of treatment (P.0.05; mean relative distance traveled in the margin: saline: 37.2560.83%, scopolamine 0.5 mg / kg: 35.2360.92%, 1.0 mg / kg: 34.1560.93%, 2.0 mg / kg: 36.1261.53%, methylscopolamine: 37.6761.03%). 3.2. Experiment 2: Water maze spatial discrimination after nicotinic blockade 3.2.1. Accuracy Mecamylamine at the doses tested had no effects on accurate responding (both P.0.05; Fig. 2A). 3.2.2. Responsivity Likewise, choice latency and choice distance remained unaffected by the drug (all P.0.05; Fig. 2B). However, mecamylamine increased errors of omission, with all doses tested being significantly different from saline-treated animals (F(3.28)54.40, P50.012), and there was a treatment3session interaction (F(12.112)56.84, P, 0.001). Further analysis indicated that this effect of mecamylamine was restricted to the first training session (Fig. 2C), while groups approached zero errors of omission during subsequent sessions (data not shown). There was also a treatment3session interaction in swim speed (F(12.112)52.32, P50.033), with animals treated with 1.0 and 2.0 mg / kg mecamylamine swimming slower than vehicle-treated mice during the first session, but not thereafter (Fig. 2D). 3.2.3. Thigmotaxis Neither measure of thigmotaxis revealed any difference between groups, although there was a tendency for salinetreated animals to spend less time in the margin of the maze than drug-treated animals during session one (F(12.112)51.95, P50.055; saline: 37.4165.68%, mecamylamine 0.5 mg / kg: 44.7364.36%, 1.0 mg / kg: 55.9165.91%, 2.0 mg / kg: 47.3363.20%). 3.3. Experiment 3: Water maze spatial discrimination after CRH stimulation 3.3.1. Accuracy Overall, there was a strong tendency for a dose-dependent effect on the percentage correct responses measure, which just failed to reach significance (F(2.19)53.50,
50
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
Fig. 1. Scopolamine disrupts acquisition of a water maze spatial discrimination. Percentage correct responses were reduced (A) and the distance traveled until a choice was made was increased. Errors of omission were increased by scopolamine during the first session only (C). Thigmotaxis in terms of the relative time spent in the margin was reduced (D). Data are presented as means with error bars denoting S.E.M.; meth, methylscopolamine.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
51
Fig. 2. Mecamylamine at a dose up to 2 mg / kg had no effect on water maze spatial discrimination accuracy in terms of percentage correct responses (A) or on the distance travelled until a choice was made (B), but increased errors of omission (C) and decreased swim speed (D) during the first training session. Data are presented as means with error bars denoting S.E.M.
52
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
P50.051). Visual inspection of the data suggested that the 1.0 mg dose of CRH tended to disrupt accuracy, while the 0.5 mg dose had no effect (Fig. 3A).
3.3.2. Responsivity Choice latency and choice distance remained unaffected by ICV administration of the peptide (mean choice latency over sessions: saline: 3.6860.03 s, CRH 0.5 mg: 3.6060.09 s, 1.0 mg: 3.7460.07 s; all P.0.05; Fig. 3B). However, analysis revealed an effect of CRH on swim speed (F(2.19)54.66, P50.023), and further post-hoc testing indicated that animals treated with 1.0 mg CRH swum slower than the other two groups. Furthermore, a dose3session interaction was observed for this measure (F(8.76)52.87, P50.021), and post-hoc testing indicated that mice receiving the high dose of CRH (1.0 mg) differed from saline-treated animals during the first but not subsequent sessions (Fig. 3C). CRH did not significantly affect errors of omission at the two doses tested (P.0.05; Fig. 3D). 3.3.3. Thigmotaxis Groups did not differ in the relative time spent or the relative distance traveled in the margin (all P.0.05, data not shown). 3.4. Experiment 4: Effects of CRH on spatial discrimination in scopolamine-treated mice 3.4.1. Accuracy Visual inspection of the data suggested that animals treated with scopolamine plus CRH, but not mice treated with either of the two drugs, differed significantly from saline-treated animals (Fig. 4A). However, three-way analysis showed that this effect was primarily caused by scopolamine: muscarinic blockade induced a significant reduction in percentage correct choices (F(1.33)55.75, P50.022), and a treatment3session interaction was seen (F(4.132)53.31, P50.018), with significant differences to IP saline treatment during the last two training sessions. In contrast, neither CRH nor CRH plus scopolamine caused any additional significant effect (all P.0.05). Thus, statistical analysis suggested that this was primarily a parallel shift, i.e., that CRH caused a small decrease in performance and that the effect of scopolamine was additive. 3.4.2. Responsivity Three-way ANOVA also indicated a main effect of CRH on choice latency and choice distance (F(1.33)510.11, P50.003 and F(1.33)510.96, P50.002, respectively), while scopolamine or combined treatment had no additional significant effect on these measures (all P.0.05; Fig. 4B; mean latency over all training sessions: saline / saline: 4.9860.49 s, saline / scopolamine: 5.7260.34 s, CRH / saline: 7.0360.49 s, CRH / scopolamine: 8.3361.11 s).
ANOVA failed to reveal significant interactions with session and the choice latency measure (all P.0.05), but an interaction between CRH-treatment and session was observed for distance traveled (treatment 13session: F(4.132)55.47, P50.001; treatment 23session and treatment 13treatment 23session: both P.0.05). Further analysis showed that the distance traveled was longer following CRH administration during all but the first session (Fig. 4B). Three-way analysis indicated that CRH on its own led to the increase in errors of omission (F(1.33)513.75, P5 0.001), while scopolamine or combined treatment had no significant effects (both P.0.05; Fig. 4C). This suggests that errors of omission were primarily caused by administration of CRH. Statistical analysis revealed an effect of treatment on swim speed, and this reduction was primarily caused by CRH (F(1.33)530.54, P,0.001), while the CRH3session interaction just failed significance (F(4.132)52.76, P5 0.057; Fig. 4D). No effects of treatment with scopolamine or with scopolamine plus CRH on swim speed were observed (all P.0.05), suggesting that alterations in swim speed were caused by CRH and that the drugs did not interact in the modulation of this parameter.
3.4.3. Thigmotaxis Analysis of the relative time spent in the margin failed to reveal a significant treatment effect (P.0.05; mean relative time over sessions: saline / saline: 37.061.43%, saline / scopolamine: 41.5563.54%, CRH / saline: 38.386 1.66%, CRH / scopolamine: 47.4963.24%). However, analysis of the relative distance traveled in the margin of the maze reached significance, and this effect was primarily caused by scopolamine treatment, which increased the relative distance traveled (F(1.33)56.60, P50.015; treatment 23session: F(4.132)50.23), while CRH on its own or in combination with scopolamine failed to cause additional significant effects (all P.0.05; Fig. 4E). 3.5. Experiment 5: Effects of CRH on spatial discrimination in mecamylamine-treated mice 3.5.1. Accuracy Three-way analysis showed an overall effect of mecamylamine on accurate performance (F(1.28)55.74, P50.024), and the treatment3session interaction also reached significance (F(4.112)52.80, P50.032; Fig. 5A). Post-hoc analysis revealed that mecamylamine impaired performance during the last training session only. Neither CRH nor combined CRH plus mecamylamine treatment induced further changes in accuracy (all P.0.05). Thus, mecamylamine at the 2 mg / kg dose induced a mild impairment in accurate responding, while CRH had no additional effects in Experiment 5. However, floor effects could have confounded these data.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
53
Fig. 3. There was a tendency for impaired accuracy following 1 mg but not 0.5 mg CRH (A). Choice distance (B) and errors of omission (first session; D) were not significantly affected. Swim speed (C) was reduced during initial sessions only. Data are presented as means with error bars denoting S.E.M.
54
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
Fig. 4. Interaction between scopolamine and CRH, effects on water maze spatial discrimination. Percentage correct responses (A), distance traveled until a choice was made (B), errors of omission (C), swim speed (D) and thigmotaxis (E). Data are presented as means with error bars denoting S.E.M.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
55
Fig. 5. Interaction between mecamylamine and CRH, effects on water maze spatial discrimination. Percentage correct responses (A), distance traveled until a choice was made (B), errors of omission (C), swim speed (D) and thigmotaxis (E). Data are presented as means with error bars denoting S.E.M.
56
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
3.5.2. Responsivity ANOVA indicated effects of treatment on both choice latency and choice distance, and both parameters were affected by CRH (F(1.28)520.22, P,0.001 and F(1.28)5 8.69, P50.006, respectively) as well as by nicotinic blockade (F(1.28)55.81, P50.023 and F(1.28)56.43, P50.017, respectively; Fig. 5B; mean latency over all training sessions: saline / saline: 5.3660.53 s, saline / mecamylamine: 5.7960.56 s, CRH / saline: 6.9660.56 s, CRH / mecamylamine: 8.5860.60 s). However, combined treatment had no additional effects (both P.0.05). Furthermore, visual inspection of the data indicated that combined treatment with CRH and mecamylamine caused an increase in errors of omission, but statistical analysis indicated that this effect was primarily due to CRH (CRH3session: F(4.112)55.89, P50.003; mecamylamine3session and treatment 13treatment 23session: both P.0.05; Fig. 5C). As before, CRH slowed swim speed (F(1.28)530.77, P,0.001), while no effects of treatment with mecamylamine or interactions were seen (all P.0.05; Fig. 5D). 3.5.3. Thigmotaxis Three-way analysis confirmed a main effect of mecamylamine (F(1.28)510.09, P50.004), and a tendency for combined treatment to cause additional impairment was seen (F(1.28)53.94; P50.057) when the relative time spent in the margin was analyzed, while CRH on its own had no effect (P.0.05; mean relative time over sessions: saline / saline: 42.1361.68%, saline / mecamylamine: 44.2461.79%, CRH / saline: 37.556 1.79%, CRH / mecamylamine: 46.9261.91%). Likewise, analysis of the relative distance traveled indicated that drug effects on the relative distance traveled was in part caused by mecamylamine, which led to an increase in the relative distance traveled (F(1.28)517.48, P,0.001). CRH on its own had no effect (both P.0.05), but the combination of CRH with mecamylamine caused additional effects (treatment 13treatment 2: F(1.28)55.54; P50.026), suggesting potentiating mechanisms (Fig. 5E).
4. Discussion This study compared the interaction between the muscarinic antagonist scopolamine and the nicotinic antagonist mecamylamine with CRH on water maze spatial discrimination learning. Main findings were a dose-dependent performance deficit caused by scopolamine, as indicated by reduced accuracy (percentage correct responses), decreased responsivity (increased choice latency, choice distance, and errors of omission) and a decrease in the relative time spent, but not the relative distance traveled, in the margin of the maze. Overall swim speed was not
affected, suggesting relatively spared motor function. Methylscopolamine at a dose of 2.0 mg / kg did not significantly differ from saline in any of the measures, suggesting that the scopolamine effects seen in this experiment were primarily centrally mediated. In contrast, mecamylamine had no effect on accuracy in Experiment 2, while three-way analysis of variance revealed a mild impairment after nicotinic blockade in Experiment 5. Comparable to muscarinic blockade, nicotinic blockade resulted in an increase in errors of omission during initial training. Thigmotactic behavior remained unaffected by mecamylamine in the second experiment. Likewise, ICV CRH had only mild effects on accuracy, with a tendency for impairment. Measures of responsivity, such as errors of omission, choice latency and distance, and of thigmotaxis were not significantly affected by CRH. However, swim speed was reduced by the peptide. Contrary to the hypothesis that CRH would attenuate some of the detrimental effects of scopolamine, we observed rather mild synergistic effects of the two drugs on accurate responding. However, for none of the measures did we observe convincing potentiating effects between the two substances. Likewise, combined treatment of CRH with mecamylamine led to some additional effects, but not improvement, both in terms of responsivity and thigmotactic behavior. Only for thigmotactic behaviour did we observe clear evidence for interaction in terms of potentiation between mecamylamine and CRH. Scopolamine is a drug well known to affect spatial performance in a range of maze tasks (e.g. [4,45,50,53,56,70]), presumably by action on both mnemonic and non-mnemonic (i.e., motivational, motor, attentional) mechanisms [60,65,68]. In similar lines, in the present study we observed a decrease in accuracy in combination with reduced responsivity. Thigmotaxis was not significantly affected in terms of the relative distance measure, but the relative time spent in the margin was shortened by scopolamine. It has been argued that the distance measure may be the more reliable index for thigmotactic behavior, while the time measure may be more easily confounded by swim speed [2]. This in turn would suggest that the effects of scopolamine on thigmotaxis in the first experiment were mild at best. Likewise, mecamylamine has been reported to impair mouse and rat maze performance in a range of studies. However, these studies tend to use higher doses [40,43,45], while lower doses in the range used in the present study also failed to affect accurate performance [12,40,45]. Moreover, non-mnemonic mechanisms are altered by mecamylamine [65,72], and the present study also showed effects on errors of omission and swim speed at doses well below those known to affect maze learning. The literature on the effects of CRH on learning and memory is even more equivocal: some studies report performance deficits following ICV CRH administration [3] or CRH overexpression in mice [28], while others find
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
‘cognitive enhancement’ [11,32,39]. Part of this discrepancy may be due to the exact site of action [3,29,54,64] and the CRH receptor subtype activated by the peptide. For example, it has been reported that CRH facilitates fear conditioning via activation of the CRH 1 receptor subtype but impairs performance in this task via activation of the CRH 2 receptor subtype [54]. Part of the discrepancy may also be due to the fact that the majority of studies did not allow a clear dissociation between mnemonic and non-mnemonic effects of the peptide. In the present study, the effects of CRH on accuracy were rather mild at the doses tested but, if at all, CRH at the doses tested tended to impair percentage correct responses. The most clear-cut effect was seen in a reduction in swim speed during initial training sessions. This is consistent with the hypoactivity often seen in novel environments following CRH administration [23], which has been suggested to reflect altered anxiety-related behavior rather than altered motor function [64]. Clearly, alterations in anxiety-related behavior induced by CRH could have confounded the results of the present and previous studies, and have been shown to at least in part contribute to the performance deficits seen in some learning paradigms [28]. This may be of particular relevance with respect to the interaction studies as both scopolamine and mecamylamine have been shown to also increase anxiety-like behavior in rats, both after systemic administration and following intra-hippocampal injection [24,62,63]. In general, the effects of combined treatment with scopolamine and CRH were, if at all, synergistic, but only very mild additive effects were observed and no potentiation of drug effects was seen. Combined nicotinic blockade and CRH activation also produced only limited synergistic effects, with evidence for potentiation in thigmotaxis. Given that CRH increases hippocampal cholinergic activity, these rather mild and, if at all, synergistic effects seem unexpected. One possible explanation could be that we missed the optimal dose range of CRH and overstimulated the cholinergic system, thus leading to overarousal and hence performance deficits. However, given the rather clear cut independency of effects and the relative paucity of clear interaction effects, this explanation seems unlikely. Alternatively, mice receiving a combination of CRH and cholinergic blockade by scopolamine or mecamylamine might have been overresponsive to the stimuli related to the test environment, as the cholinergic system may be of relevance for increasing the signal-to-noise ratio for salient events, thereby reducing the effects of distracting extraneous stimuli [55]. Consequently, scopolamine or mecamylamine may render an animal hyperresponsive to its environment as it cannot monitor and amend its behavior in an appropriate manner when exposed to fearful stimuli [5,63], and CRH may have worsen this state. In fact, it has been suggested that the exaggerated stress response in the absence of functional cholinergic activity might be due to overestimation of the magnitude of threat
57
as a consequence of hypersensitivity to environmental stimuli normally filtered out [63]. Thus, in addition to an already compromised stimulus processing and an overactive HPA axis, additional treatment with CRH would be expected to further augment the scopolamine- (and possibly also mecamylamine)-induced HPA hyperactivity, which might explain the mild synergistic effects of combined treatment. Synergistic effects could also be expected as cholinergic blockade at hippocampal level leads to increased CRH activity at the level of the paraventricular hypothalamic nucleus (PVN) through disruption of the descending hippocampal pathway, thereby enhancing the hypothalamic–pituitary–adrenal axis response [63]. However, it should be noted that ACh stimulates CRH activity at HPA level when injected directly into the PVN and that this effect is blocked by the muscarinic antagonist atropine [47]. Therefore, it is difficult to disentangle the net effect of systemic scopolamine on CRH release from the PVN at this stage. Yet another explanation could be that CRH impaired performance by acting at the CRH 2 receptor subtype [54], i.e., that it is the combination of muscarinic blockade and CRH 2 receptor stimulation which is required to cause the additive effects, while the potentially beneficial effects of CRH 1 receptor stimulation on cholinergic activity might be insufficient to overcome the muscarinic block, especially as scopolamine on its own already increases hippocampal ACh release [30,59,71]. Moreover, it has been noted that CRH also has an indirect inhibitory input on the cholinergic basal forebrain: ICV administration of CRH decreases sodium-dependent high affinity choline uptake (HACU) at the level of the rat frontal cortex and hippocampus. This effect can be blocked by naltrexone, suggesting that CRH exerts its inhibitory action through interaction with the opioidergic system [38]. If this mechanism would predominate, it could offer yet another explanation for the present results. These findings may be of relevance for psychiatric disorders such as depression and dementia: an unrestrained CRH hyperdrive has been proposed to play a major role in the development and course of depression [31] and cholinergic hyperactivity has been found in this disorder [35]. Conversely, in Alzheimer’s disease, a decrease in cholinergic function is well documented [14,15,46,51,52], which coincides with a decrease in cortical CRH and a reciprocal increase in cortical CRH receptors [15,20,21]. Stimulation of parts of the CRH system has been proposed as a potential treatment for Alzheimer’s disease [3]. The present results fail to support this concept and would suggest that this may even worsen the cognitive deficit seen in these patients. In summary, the present data lend only limited support for the concept of important interactions between CRH and ACh in the modulation of learning and memory. Moreover, the behavioral consequences of cholinergic dysfunction seem to be enhanced and not antagonized by CRH.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
58
Acknowledgements We thank A. Mederer for excellent technical support.
References [1] E. Acquas, C. Wilson, H.C. Fibiger, Conditioned and unconditioned stimuli increase frontal cortical and hippocampal acetylcholine release: effects of novelty, habituation, and fear, J. Neurosci. 16 (1996) 3089–3096. [2] M. Arns, M. Sauvage, T. Steckler, Excitotoxic lesions disrupt allocentric spatial learning in mice: effects of strain and task demands, Behav. Brain Res. 106 (1999) 151–164. [3] D.P. Behan, S.C. Heinrichs, J.C. Troncoso, X.J. Liu, C.H. Kawas, N. Ling, E.B. De Souza, Displacement of corticotropin releasing factor from its binding protein as a possible treatment for Alzheimer’s disease, Nature 378 (1995) 284–287. [4] J. Berger-Sweeney, A. Arnold, D. Gabeau, J. Mills, Sex differences in learning and memory in mice: effects of sequence of testing and cholinergic blockade, Behav. Neurosci. 109 (1995) 859–873. [5] S. Bhatnagar, B. Costall, J.W. Smythe, Hippocampal cholinergic blockade enhances hypothalamic–pituitary–adrenal responses to stress, Brain Res. 766 (1997) 244–248. [6] J.C. Bittencourt, P.E. Sawchenko, Do centrally administered neuropeptides access cognate receptors?: An analysis in the central corticotropin-releasing factor system, J. Neurosci. 20 (2000) 1142– 1156. [7] A.E. Calogero, W.T. Gallucci, R. Bernardini, C. Saoutis, P.W. Gold, G.P. Chrousos, Effect of cholinergic agonists and antagonists on rat hypothalamic corticotropin-releasing hormone secretion in vitro, Neuroendocrinology 47 (1988) 303–308. [8] I. Ceccarelli, F. Casamenti, C. Massafra, G. Pepeu, C. Scali, A.M. Aloisi, Effects of novelty and pain on behavior and hippocampal extracellular ACh levels in male and female rats, Brain Res. 815 (1999) 169–176. [9] D.T. Chalmers, T. W Lovenberg, E.B. De Souza, Localization of novel corticotropin-releasing factor receptor (CRF 2 ) mRNA expression to specific subcortical nuclei in rat brain: comparison with CRF 1 receptor mRNA expression, J. Neurosci. 15 (1995) 6340– 6350. [10] J.C. Chen, K.L. Brunson, M.B. Muller, W. Cariaga, T. Z Baram, Immunocytochemical distribution of corticotropin-releasing hormone receptor type-1 (CRF1)-like immunoreactivity in the mouse brain: light microscopic analysis using an antibody directed against the C-terminus, J. Comp. Neurol. 420 (2000) 305–323. [11] M.F. Chen, T.H. Chiu, E.H. Lee, Noradrenergic mediation of the memory-enhancing effect of corticotropin-releasing factor in the locus coeruleus of rats, Psychoneuroendocrinology 17 (1992) 113– 124. [12] Y. Chen, E. Shohami, S. Constantini, M. Weinstock, Rivastigmine, a brain-selective acetylcholinesterase inhibitor, ameliorates cognitive and motor deficits induced by closed-head injury in the mouse, J. Neurotrauma 15 (1998) 231–237. [13] R. Cozzolino, D. Guaralsi, A. Giuliani, O. Ghirardi, M.T. Ramacci, L. Angelucci, Effects of concomitant nicotinic and muscarinic blockade on spatial memory disturbance in rats are purely additive. Evidence from the Morris water task, Physiol. Behav. 56 (1994) 111–114. [14] P. Davies, A.J.F. Maloney, Selective loss of cholinergic neurons in Alzheimer’s disease, Lancet 25 (1976) 1403. [15] K.L. Davis, R.C. Mohs, D.B. Martin, D.P. Purohit, D.P. Perl, M. Lantz, G. Austin, V. Haroutunian, Neuropeptide abnormalities in patients with early Alzheimer’s disease, Arch. Gen. Psychiatry 56 (1999) 981–987.
[16] J.C. Day, M. Koehl, M. Le Moal, S. Maccari, Corticotropinreleasing factor administered centrally, but not peripherally, stimulates hippocampal acetylcholine release, J. Neurochem. 71 (1998) 622–629. [17] J.C. Day, M. Koehl, V. Deroche, M. Le Moal, S. Maccari, Prenatal stress enhances stress- and corticotropin-releasing factor-induced stimulation of hippocampal acetylcholine release, J. Neurosci. 18 (1998) 1886–1892. [18] M.W. Decker, M.J. Majchrzak, Effects of systemic and intracerebroventricular administration of mecamylamine, a nicotinic cholinergic antagonist, on spatial memory in rats, Psychopharmacology 107 (1992) 530–534. [19] E.B. De Souza, G. Battaglia, Increased corticotropin-releasing factor receptors in rat cerebral cortex following chronic atropine treatment, Brain Res. 397 (1986) 401–404. [20] E.B. De Souza, P.J. Whitehouse, M.J. Kuhar, D.L. Price, W.W. Vale, Reciprocal changes in corticotropin-releasing factor immunoreactivity and CRF receptors in cerebral cortex of Alzheimer’s disease, Nature 319 (1986) 593–595. [21] E.B. De Souza, P.J. Whitehouse, M.J. Kuhar, D.L. Price, W.W. Vale, Abnormalities in corticotropin-releasing hormone (CRH) in Alzheimer’s disease and other human disorders, Ann. N.Y. Acad. Sci. USA 512 (1987) 237–247. [22] J.D. Dudar, I.Q. Whishaw, J.C. Szerb, The role of the septal nuclei in the release of acetylcholine from the hippocampus of freely moving rats during sensory stimulation and running, Neuropharmacology 18 (1979) 673–678. [23] A.J. Dunn, C.W. Berridge, Physiological and behavioral responses to corticotropin-releasing factor administration: is CRF a mediator of anxiety or stress responses? Brain Res. Rev. 15 (1990) 71–100. [24] S.E. File, P.J. Kenny, S. Cheeta, The role of the dorsal hippocampal serotonergic and cholinergic systems in the modulation of anxiety, Pharmacol. Biochem. Behav. 66 (2000) 65–72. [25] K.B.J. Franklin, G. Paxinos, The Mouse Brain in Stereotaxis Coordinates, Academic Press, San Diego, 1997. [26] Y. Fu, S.G. Matta, J.D. Valentine, B.M. Sharp, Adrenocorticotropin response and nicotine-induced norepinephrine secretion in the rat paraventricular nucleus are mediated through brainstem receptors, Endocrinology 138 (1997) 1935–1943. [27] G.M. Gilad, B.D. Mahon, Y. Finkelstein, B. Koffler, V.H. Gilad, Stress induced activation of the hippocampal cholinergic system and the pituitary adrenocortical axis, Brain Res. 347 (1985) 404–408. [28] S.C. Heinrichs, M.P. Stenzel-Poore, L.H. Gold, E. Battenberg, F.E. Bloom, G. Koob, W.W. Vale, E. Merlo Pich, Learning impairment in transgenic mice with central over expression of corticotropin-releasing factor, Neuroscience 74 (1996) 303–311. [29] S.C. Heinrichs, E.A. Vale, J. Lapsansky, D.P. Behan, L.V. McClure, N. Ling, E.B. De Souza, G. Schulteis, Enhancement of performance in multiple learning tasks by corticotropin-releasing factor-binding protein ligand inhibitors, Peptides 18 (1997) 711–716. [30] M. Hiramatsu, H. Murasawa, T. Nabeshima, T. Kameyama, Effects of U-50,488H on scopolamine-, mecamylamine- and dizocilpineinduced learning and memory impairment in rats, J. Pharmacol. Exp. Ther. 284 (1998) 858–867. [31] F. Holsboer, The rationale for corticotropin-releasing hormone receptor (CRH-R) antagonists to treat depression and anxiety, J. Psychiat. Res. 33 (1999) 181–214. [32] H.C. Hung, C.K. Chou, T.H. Chiu, E.H. Lee, CRF increases protein phosphorylation and enhances retention performance in rats, Neuroreport 2 (1992) 181–184. [33] A. Imperato, S. Puglisi-Allegra, P. Casolini, L. Angelucci, Changes in brain dopmaine and acetylcholine release during and following stress are independent of the pituitary–adrenocortical axis, Brain Res. 538 (1991) 111–117. [34] F.M. Inglis, H.C. Fibiger, Increases in hippocampal and frontal cortical acetylcholine release associated with presentation of sensory stimuli, Neuroscience 66 (1995) 81–86.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59 [35] D.S. Janowsky, D.H. Overstreet, J.I. Nurnberger Jr., Is cholinergic sensitivity a genetic marker for the affective disorders? Am. J. Med. Genet. 54 (1994) 335–344. [36] J.W. Kasckow, A. Regmi, S. Sheriff, J. Mulchahey, T.D. Geracioti Jr., Regulation of corticotropin-releasing factor messenger RNA by nicotine in an immortalized amygdalar cell line, Life Sci. 65 (1999) 2709–2714. [37] R. Kortekaas, B. Costall, J.W. Smythe, Changes in hippocampal theta following intrahippocampal corticotropin-releasing hormone (CRH) infusions in the rat, Brain Res. Bull. 48 (1999) 603–607. [38] H. Lai, M.A. Carino, Effects of noise on high-affinity choline uptake in the frontal cortex and hippocampus of the rat are blocked by intracerebroventricular injection of corticotropin-releasing factor antagonist, Brain Res. 527 (1990) 354–358. [39] E.H. Lee, C.P. lee, H.I. Wang, W.R. Lin, Hippocampal CRF, NE, and NMDA system interactions in memory processing in the rat, Synapse 14 (1993) 144–153. [40] E.D. Levin, M. Castonguay, G.D. Ellison, Effects of nicotinic receptor blocker mecamylamine on radial-arm maze performance in rats, Behav. Neural Biol. 48 (1987) 206–212. [41] S.G. Matta, J. Singh, B.M. Sharp, Catecholamines mediate nicotineinduced adrenocroticotropin secretion via alpha-adrenergic receptors, Endocrinology 127 (1990) 1646–1655. [42] S.G. Matta, J.D. Valentine, B.M. Sharp, Nicotine activation of CRH neurons in extrahypothalamic regions of the rat brain, Endocrine 7 (1997) 245–253. [43] T. Maurice, T.P. Su, D.W. Parish, T. Nabeshima, A. Privat, PRE084, a sigma selective PCP derivative, attenuates MK-801-induced impairment of learning in mice, Pharmacol. Biochem. Behav. 49 (1994) 859–869. [44] W.J. Millard, G.W. Arendash, A.J. Dunn, E.M. Meyer, Effects of nucleus basalis lesions on cerebral cortical concentrations of corticotropin-releasing hormone (CRH)-like immunoreactivity in the rat, Neurosci. Lett. 113 (1990) 233–239. [45] P.M. Moran, Differential effects of scopolamine and mecamylamine on working and reference memory in the rat, Pharmacol. Biochem. Behav. 45 (1993) 533–538. [46] P.A. Newhouse, Alzheimer’s disease and the cholinergic system: an introduction to clinical pharmacological research, in: L.L. Heston (Ed.), Progress in Alzheimer’s Disease and Similar Conditions, American Psychiatric Press, Washington, 1997, pp. 213–231. [47] N. Ohmori, K. Itoi, F. Tozawa, Y. Sakai, K. Sakai, N. Horiba, H. Demura, T. Suda, Effects of acetylcholine on corticotropin-releasing factor gene expression in the hypothalamic paraventricular nucleus of conscious rats, Endocrinology 136 (1995) 4858–4863. [48] P. Onali, M.C. Olianas, Identification and characterization of muscarinic receptors potentiating the stimulation of adenylyl cyclase activity by corticotropin releasing hormone in membranes of rat frontal cortex, J. Pharmacol. Exp. Ther. 286 (1998) 753–759. [49] J.M. Overton, G. Davis-Gorman, L.A. Fisher, Central nervous system cardiovascular actions of CRF in sinuaortic-denervated rats, Am. J. Physiol. 258 (1990) R596–R601. [50] D.B. Peele, S.P. Baron, Effects of selection delays on radial maze performance: acquisition and effects of scopolamine, Pharmacol. Biochem. Behav. 29 (1988) 143–150. [51] E.K. Perry, The cholinergic hypothesis – 10 years on, Br. Med. Bull. 42 (1986) 63–69. [52] E.K. Perry, B.E. Tomlinson, G. Blessed, K. Bergman, P.H. Gibson, R.H. Perry, Correlation of cholinergic abnormalities with senile plaques and mental test scores in senile dementia, Br. Med. J. 2 (1978) 1457–1459. [53] J.J. Pilcher, G.R. Sessions, S.A. McBride, Scopolamine impairs spatial working memory in the radial maze: an analysis by error type and arm choice, Pharmacol. Biochem. Behav. 58 (1997) 449–459. ¨ [54] J. Radulovic, A. Ruhmann, T. Liepold, J. Spiess, Modulation of learning and anxiety by corticotropin-releasing factor (CRF) and
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64] [65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
59
stress: differential roles of CRF receptors 1 and 2, J. Neurosci. 19 (1999) 5016–5025. R.T. Richardson, M.R. DeLong, A reappraisal of the functions of the nucleus basalis of Meynert, Trends Neurosci. 11 (1988) 264– 267. ¨ M. Aaltonen, P. Riekkinen, Effects of P. Riekkinen Jr., J. Sirvio, concurrent manipulations of nicotinic and muscarinic receptors on spatial and passive avoidance learning, Pharmacol. Biochem. Behav. 37 (1990) 405–410. ¨ P. Riekkinen, Pharmacological conseP. Riekkinen Jr., J. Sirvio, quences of cholinergic plus serotonergic manipulations, Brain Res. 527 (1991) 23–26. T.W. Robbins, G. McAlonan, J.L. Muir, B.J. Everitt, Cognitive enhancers in theory and practice: studies of the cholinergic hypothesis of cognitive deficits in Alzheimer’s disease, Behav. Brain Res. 83 (1997) 15–23. J.K. Robinson, A. Zocchi, A. Pert, J.N. Crawley, Galanin microinjected into the medial septum inhibits scopolamine-induced acetylcholine overflow in the rat ventral hippocampus, Brain Res. 709 (1996) 81–87. A. Sahgal, A.B. Keith, S. Lloyd, J.M. Kervin, E.K. Perry, J.A. Edwardson, Memory following cholinergic (NBM) and noradrenergic (DNAB) lesions made singly or in combination: potentiation of disruption by scopolamine, Pharmacol. Biochem. Behav. 37 (1990) 597–605. M. Sauvage, T. Steckler, Detection of corticotropin-releasing hormone receptor 1 immunoreactivity in cholinergic, dopaminergic and noradrenergic neurons of the basal forebrain and brainstem nucleipotential implication for arousal and attention, Neuroscience (2001) in press. J.W. Smythe, S. Bhatnagar, D. Murphy, C. Timothy, B. Costall, The effects of intrahippocampal scopolamine infusions on anxiety in rats as measured by the black–white box test, Brain Res. Bull. 45 (1998) 89–93. J.W. Smythe, D. Murphy, S. Bhatnagar, C. Timothy, B. Costall, Muscarinic antagonists are anxiogenic in rats tested in the black– white box, Pharmacol. Biochem. Behav. 54 (1996) 57–63. T. Steckler, F. Holsboer, Corticotropin-releasing hormone receptor subtypes and emotion, Biol. Psychiatry 46 (1999) 1480–1508. T. Steckler, A.B. Keith, R.G. Wiley, A. Sahgal, Cholinergic lesions by 192 IgG-saporin and short-term recognition memory: role of the septohippocampal projection, Neuroscience 66 (1995) 101–114. T. Steckler, A. Sahgal, The role of serotonergic–cholinergic interactions in the mediation of cognitive behaviour, Behav. Brain Res. 67 (1995) 165–199. T. Steckler, A. Sahgal, J.P. Aggleton, W.H.I.M. Drinkenburg, Recognition memory in rats. III. Neurochemical substrates, Prog. Neurobiol. 54 (1998) 333–348. T. Steckler, C. Weis, M. Sauvage, A. Mederer, F. Holsboer, Disrupted allocentric but preserved egocentric spatial learning in transgenic mice with impaired glucocorticoid receptor function, Behav. Brain Res. 100 (1999) 77–89. Y. Tizabi, A.E. Calogero, Effects of various neurotransmitters and neuropeptides on the release of corticotropin-releasing hormone from the rat cortex in vitro, Synapse 10 (1992) 341–348. A. Urani, A. Privat, T. Maurice, The modulation of neurosteroids of the scopolamine-induced learning impairment in mice involves an interaction with sigma1 (sigma1) receptors, Brain Res. 799 (1998) 64–77. M.G. Vannucchi, C. Scali, S.R. Kopf, G. Pepeu, F. Casamenti, Selective muscarinic antagonists differentially affect in vivo acetylcholine release and memory performances of young and aged rats, Neuroscience 79 (1997) 837–846. D.V. Widzowski, E. Cregan, P. Bialobok, Effects of nicotinic agonists and antagonists on spatial working memory in normal adult and aged rats, Drug Dev. Res. 31 (1994) 24–31.
Research report
Interaction between the cholinergic system and CRH in the modulation of spatial discrimination learning in mice Thomas Steckler*, Florian Holsboer Max Planck Institute of Psychiatry, Kraepelinstr. 2 – 10, D-80804 Munich, Germany Accepted 17 April 2001
Abstract Both cholinergic and CRH systems have been linked to cognitive processes such as learning and memory, and neuroanatomical as well as neurochemical evidence suggests important interactions between these two systems. Moreover, recent reports of pro-mnestic effects of CRH open the possibility that CRH could have beneficial effects in animals with cholinergic dysfunction. In a first experiment, spatial discrimination of C57BL / 6 mice treated with various doses of scopolamine (0.5–2.0 mg / kg IP) was tested in a two-choice water maze task. Scopolamine, but not methylscopolamine, impaired accuracy and decreased responsivity. In contrast, similar doses of the nicotinic antagonist mecamylamine had no effect on choice accuracy but altered responsivity, as indicated by increased errors of omission and a reduction in swim speed during early experimental stages. ICV CRH (0.5–1.0 mg) also failed to significantly affect accuracy, but a strong tendency was observed to impair percentage correct responses. Measures of responsivity, such as errors of omission, choice latency and distance traveled, and of thigmotaxis were not significantly affected by CRH. However, initial swim speed was reduced by the peptide. Combined treatment with scopolamine (0.5 mg / kg IP) and CRH (0.5 mg ICV) had only mild, and primarily independent, effects, but overall suggested that concomitant blockade of muscarinic receptors and activation of the CRH system would rather act synergistically to disrupt spatial discrimination learning. Synergistic effects were also observed when animals receiving a combination of mecamylamine (2.0 mg / kg IP) and CRH (0.5 mg ICV) were tested, both in terms of responsivity and thigmotaxis, and there was limited evidence that part of these effects were potentiating. Thus, the cholinergic and CRH systems interact in the modulation of learning, but CRH, contrary to prediction, worsens the impairment caused by cholinergic blockade. 2001 Elsevier Science B.V. All rights reserved. Theme: Neural basis of behavior Topic: Neuropeptides and behavior Keywords: Acetylcholine; Corticotropin-releasing hormone; CRH; Interaction; Learning; Memory; Spatial discrimination
1. Introduction Neurochemical and anatomical evidence suggests that there are important reciprocal interactions between the cholinergic and the corticotropin-releasing hormone (CRH) systems: first, frontoparietal cortical CRH release has been reported to be inhibited by acetylcholine (ACh) [69], while hypothalamic CRH activity is enhanced following ACh administration [5,47,69]. Lesions of the nucleus basalis, the main basal forebrain cholinergic nucleus projecting to neocortical areas, on the other hand, increase parietal *Corresponding author. Present address: Janssen Research Foundation, Turnhoutseweg 30, B-2340 Beerse, Belgium. Tel.: 132-14-60-7373; fax: 132-1460-22-6121. E-mail address: [email protected] (T. Steckler).
CRH-like immunoreactivity some months post-lesioning [44]. Both nicotinic and muscarinic processes appear to be involved: nicotine has been reported to activate Fos expression in CRH-positive neurons in the bed nucleus of the stria terminalis, the central nucleus of the amygdala, the dorsal raphe and Barrington’s nucleus [42], and to stimulate CRH mRNA expression in immortalized amygdala cells [36]. At the level of the hypothalamic– pituitary–adrenal (HPA) axis, nicotine has also been reported to activate CRH neurons within the paraventricular hypothalamic nucleus (PVN) [42] and to stimulate CRH secretion from rat hypothalami in vitro [7], but the main action of nicotine to stimulate ACTH release may be indirect via activation of noradrenaline release at the level of the PVN [26,41]. Chronic muscarinic blockade, on the other hand, leads to a significant increase in CRH receptors
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02555-0
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
in rat frontoparietal cortex, but not in the hippocampus [19]. Thus, a reduction in cholinergic muscarinic activity may lead to an enhanced activity of the CRH system, at least at cortical level, which contrasts the stimulation of CRH activity by activation of nicotinic receptors at subcortical level. Furthermore, acute peripheral muscarinic blockade has been reported to reduce ICV CRH-induced tachycardia through parasympathetic mechanisms [49], pointing towards important interactions at the level of the autonomic nervous system as well. Second, one of the CRH receptor subtypes, the CRH 1 receptor, is highly abundant in mouse and rat cholinergic basal forebrain and brainstem nuclei [6,9,10] and a high degree of co-localization of choline–acetyltransferase (ChAT) and CRH receptor-like immunoreactivity has been observed in both the murine basal forebrain and brainstem, with the notable exception of the nucleus basalis magnocellularis [61]. In rats, intracerebroventricular (ICV) CRH produces moderate to strong stimulation of Fos expression within the basal forebrain and brainstem nuclei, i.e., within the medial septum and diagonal band of Broca, the laterodorsal and the pedunculopontine tegmental nuclei, while only relatively weak Fos expression is induced within the substantia innominata [6]. Furthermore, ICV but not peripheral injections of CRH increase hippocampal ACh release [16,17], CRH alters hippocampal theta rhythm, possibly through interaction with cholinergic systems [37], and it may be suggested that CRH induces these effects directly at the level of the cholinergic basal forebrain via CRH 1 receptor activation. Various stressors have also been reported to increase hippocampal ACh release [1,17,22,27,33]. These stressinduced changes in hippocampal cholinergic activity reach a maximum 20–30 min after stress exposure and have been suggested to be independent of the hypothalamic– pituitary–adrenal (HPA) axis [33] (but see Ref. [27]). Obviously, one potential candidate mediating this response could be centrally released CRH. Third, CRH may interact with cholinergic mechanisms by stimulating common targets, as it has been shown that activation of the M1 muscarinic receptor subtype and of CRH receptors synergistically increase cAMP production in membranes from rat frontal cortex [48]. Since septohippocampal ACh release is activated following sensory stimulation [1,8,22,34], it may be argued that the central interaction between CRH and the cholinergic septohippocampal system serves to maintain appropriate processing of environmental information, in particular under stressful conditions. This in turn would suggest that the interaction between the two systems is of relevance for performance in a range of cognitive tasks, including spatial learning and memory paradigms. Indeed, both the cholinergic and the CRH systems have been, independently from each other, implicated in the modulation of learning and memory [3,11,13,18,28,29,32, 39,40,45,54,56–58,64,66,67]. Some of these studies sug-
47
gest that CRH may have pro-mnemonic effects [3,10, 29,32,39,54], and it is conceivable that CRH enhances cognitive function through interaction with the cholinergic system. However, the exact nature of this interaction between the two systems in the modulation of these processes remains unclear at the behavioral level. This question was addressed in the present study, testing the effects of combined cholinergic blockade and CRH stimulation in mice performing on a water maze spatial discrimination paradigm. In particular, we tested the hypothesis that ICV administration of CRH would attenuate the deleterious effects of cholinergic blockade caused by administration of either the muscarinic antagonist scopolamine or the nicotinic antagonist mecamylamine.
2. Material and methods
2.1. Animals Male 3-month-old C57BL / 6CrlBR mice (Charles River, Sulzfeld, Germany) were housed individually and maintained on a 12:12 h light / dark cycle (lights on at 07:00). Experiments were conducted during the light phase of the cycle. Food and water were available ad libitum. All animal procedures were approved by the Ethical Committee on Animal Care and Use of the Government of Bavaria, Germany.
2.2. Surgery Mice were anaesthetized with a mixture of ketamine (10%, 0.5 ml / kg, IP) and rompun (2.0 ml / kg, IP) and placed in a stereotaxic frame (TSE, Bad Homburg, Germany). A stainless steel tube (27 gauge) was implanted into the right lateral ventricle at the following coordinates according to the atlas of Franklin and Paxinos [25]: AP, 0.6 mm posterior to bregma, L, 1.8 mm lateral to the midline, V, 2.2 mm below the exposed dura mater. The tube was fixed to the skull with two stainless steel screws, placed in the vicinity of the tube, and dental cement applied to cover the screws and tube. A stainless steel stylet, with a length identical to the tube, was inserted to keep the cannula patent. Mice were allowed 1 week recovery before testing. On completion of the experiments, animals were anaesthetized with isofluran and correct placement of the cannula was confirmed by ICV injection of 5 ml of cresyl violet solution. Only animals with correct placement of the cannulae were included for further analysis.
2.3. Drugs Scopolamine hydrobromide, scopolamine methylbromide and mecamylamine hydrochloride (all Sigma Chemicals, Deisenhofen, Germany) were dissolved in saline and administered intraperitoneally (IP). CRH (Ba-
48
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
chem Biochemica, Heidelberg, Germany) was administered over a period of 2 min via an injection cannula which protruded 1.0 mm below the guide tube, and which was connected to a 10 ml Hamilton syringe.
2.4. Apparatus A circular swimming pool (80 cm in diameter, 30 cm high, white plastic) was filled to a depth of 20 cm with water (21618C; rendered opaque by the addition of a non-toxic dye). At the outside of the maze eight start boxes (10310326 cm, also filled with water) and fitted with sliding doors were fixed, from which the animal could swim into the center of the maze when the door was raised. Two identically looking circular platforms with white surface and dark gray rim were used (each 10 cm in diameter, protruding approximately 0.5 cm above the water surface). One platform was stable and provided support (‘correct platform’), the other was floating and sank when a mouse tried to climb on it (‘incorrect platform’). Performance was recorded by a video tracking system (TSE).
group) were tested after treatment with saline, scopolamine hydrobromide (0.5, 1.0 or 2.0 mg / kg IP) or scopolamine methylbromide (2.0 mg / kg), 20 min prior to each session. In a second experiment, four groups of mice (n58 per group) received treatment with saline or mecamylamine hydrochloride (0.5, 1.0 or 2.0 mg / kg IP), 20 min prior to each session. In the third experiment, three groups of cannulated animals received ICV infusions of saline (n59), 0.5 mg or 1.0 mg CRH (n57 and 6, respectively), 30 min before training. In Experiment 4, four groups of cannulated animals were first treated ICV with 0.5 mg CRH or vehicle (30 min prior to training), followed by an IP injection with scopolamine hydrobromide (0.5 mg / kg) or saline (20 min prior to training), resulting in the following treatment combinations: group saline / saline (n510), group saline / scopolamine (n59), group CRH / saline (n58), and group CRH / scopolamine (n510). In a final experiment, the interaction between CRH 0.5 mg ICV, 30 min prior to test, and mecamylamine (2.0 mg / kg), 20 min prior to test, was studied in four groups: group saline / saline (n59), group saline / mecamylamine (n58), group CRH / saline (n58), and group CRH / mecamylamine (n57).
2.5. Training procedure Mice were trained over five sessions (one session per day, 10 trials per session) to choose between the two platforms (stable vs. unstable) [2,68]. On each trial, the stable platform (correct platform) remained in the same position (approx. 10 cm distance to the wall), while the unstable platform (incorrect platform) changed position from trial to trial in a pseudorandom manner (five possible positions). It was ensured that the spatial relationship between the platforms did not consistently reward turns into one direction, and that the distance between the start position and each of the two platforms was equal over the ten trials. A trial started by placing a mouse in one of six possible start boxes (with the door closed) in pseudorandom sequence. Then the door was raised and the animals had access to the open field of the maze. Tracking started automatically when the animal swum over a start area located at the rim of the open field, directly in front of the start box, i.e., data recording started as soon as the animal had left the start box. All boxes except the start positions in front of and opposite to the correct platform were used. A trial terminated when a mouse climbed onto one of the two platforms or after 30 s. If a mouse climbed onto the stable platform within 30 s, it was allowed to stay there for another 10 s before it was returned to the holding cage. If an animal made an incorrect choice (climbing the incorrect platform) or after 30 s had lapsed, it was gently placed on the correct platform and allowed to stay there for 10 s before it was returned to the holding cage. Animals were trained in squads of four. ITI’s ranged from 2 to 4 min and each session lasted approximately 30 min. In Experiment 1, five groups of animals (n510 per
2.6. Behavioral measures A choice was made if an animal touched a platform with its forepaws or its snout. The occasional incident of brushing past the floating platform was not considered a choice. Percentage correct choices, errors of omission (i.e., the number of trials the animal failed to make a choice within 30 s), average choice latency and choice distance, swim speed, and measures of thigmotactic behavior, i.e. the relative distance swum and the relative time spent in the margin, were calculated. If an error of omission was made, no latency or distance traveled were scored. In contrast, indices of thigmotaxis were calculated for all trials as thigmotaxis may be one factor leading to errors of omission.
2.7. Data analysis Data were transformed as appropriate (arcsine after division by 100, all percentage measures; logarithmic, latency, distance and speed measures; square-root after addition of 0.5, errors of omission) and analyzed by repeated measures ANOVA, including one-factor (treatment) independent measures and a two-factor mixed measures analysis, with session as the repeated measure (treatment3session). Data from the last two experiments were analyzed using a three-factor analysis (treatment3 treatment; treatment3treatment3session), which allowed to dissociate main treatment effects. If ANOVA revealed a significant difference, this was followed by post-hoc testing using the Tukey–Kramer procedure.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
3. Results
3.1. Experiment 1: Water maze spatial discrimination after muscarinic blockade 3.1.1. Accuracy There was a main effect of dose on percentage correct responses (F(4.45)57.04, P,0.001; Fig. 1A). Post-hoc testing failed to reveal a difference between methylscopolamine- and saline-treated animals, while all scopolamine-treated groups differed significantly from saline, but not from each other. Furthermore, ANOVA showed a dose3session interaction for this measure (F(16.180)52.22, P50.012): groups started to differ from the second session onwards, with only the two highest doses of scopolamine being significantly different from saline during this session, while all three groups treated with scopolamine differed from saline during session 3. Furthermore, all three groups treated with scopolamine performed at lower accuracy level than methylscopolamine during the last training session. 3.1.2. Responsivity Overall, choice latency was increased by all three doses of scopolamine relative to saline treatment (F(4.45)52.62, P50.047; mean choice latency over sessions: saline: 3.6160.27 s, scopolamine 0.5 mg / kg: 4.7760.34 s, 1.0 mg / kg: 4.4760.17 s, 2.0 mg / kg: 4.8860.42 s, methylscopolamine: 4.5860.50 s). Moreover, there was a dose3session interaction for this measure (F(16.180)5 1.94, P50.028), and further post-hoc testing indicated that the group treated with the highest dose of scopolamine had a significantly longer choice latency than saline-treated animals during the initial training session (mean choice latency during first session: saline: 6.7161.09 s, scopolamine 0.5 mg / kg: 10.1861.46 s, 1.0 mg / kg: 9.0360.84 s, 2.0 mg / kg: 11.6261.33 s, methylscopolamine: 8.6461.05 s). Likewise, scopolamine increased choice distance at every dose tested relative to saline (Fig. 1B). In addition, choice distance was also increased in the 0.5 mg and 2.0 mg scopolamine-treated groups relative to the performance of animals treated with methylscopolamine (F(4.45)56.87, P,0.001), while the dose3session interaction did not reach significance for this parameter (P.0.05). Scopolamine induced a dose-dependent increase in errors of omission (F(4.45)52.65, P50.046), and the highest dose (2.0 mg / kg) caused significantly more omissions than saline or methylscopolamine. There was a dose3session interaction (F(16.180)54.02, P50.004), and further analysis indicated that this high-dose effect was restricted to the first training session (Fig. 1C), while groups rapidly approached zero errors of omission from session two onwards and did not differ during subsequent sessions. Overall, swim speed did not differ between groups
49
(P.0.05; mean swim speed over sessions: saline: 20.6760.51 cm / s, scopolamine 0.5 mg / kg: 20.5860.50 cm / s, 1.0 mg / kg: 20.8760.42 cm / s, 2.0 mg / kg: 20.6460.62 cm / s, methylscopolamine: 21.3760.46 cm / s).
3.1.3. Thigmotaxis The relative time spent in the margin of the maze was decreased by all three doses of scopolamine when compared to saline (F(4.45)55.85, P50.001; Fig. 1D). However, analysis of the relative distance travelled failed to show an effect of treatment (P.0.05; mean relative distance traveled in the margin: saline: 37.2560.83%, scopolamine 0.5 mg / kg: 35.2360.92%, 1.0 mg / kg: 34.1560.93%, 2.0 mg / kg: 36.1261.53%, methylscopolamine: 37.6761.03%). 3.2. Experiment 2: Water maze spatial discrimination after nicotinic blockade 3.2.1. Accuracy Mecamylamine at the doses tested had no effects on accurate responding (both P.0.05; Fig. 2A). 3.2.2. Responsivity Likewise, choice latency and choice distance remained unaffected by the drug (all P.0.05; Fig. 2B). However, mecamylamine increased errors of omission, with all doses tested being significantly different from saline-treated animals (F(3.28)54.40, P50.012), and there was a treatment3session interaction (F(12.112)56.84, P, 0.001). Further analysis indicated that this effect of mecamylamine was restricted to the first training session (Fig. 2C), while groups approached zero errors of omission during subsequent sessions (data not shown). There was also a treatment3session interaction in swim speed (F(12.112)52.32, P50.033), with animals treated with 1.0 and 2.0 mg / kg mecamylamine swimming slower than vehicle-treated mice during the first session, but not thereafter (Fig. 2D). 3.2.3. Thigmotaxis Neither measure of thigmotaxis revealed any difference between groups, although there was a tendency for salinetreated animals to spend less time in the margin of the maze than drug-treated animals during session one (F(12.112)51.95, P50.055; saline: 37.4165.68%, mecamylamine 0.5 mg / kg: 44.7364.36%, 1.0 mg / kg: 55.9165.91%, 2.0 mg / kg: 47.3363.20%). 3.3. Experiment 3: Water maze spatial discrimination after CRH stimulation 3.3.1. Accuracy Overall, there was a strong tendency for a dose-dependent effect on the percentage correct responses measure, which just failed to reach significance (F(2.19)53.50,
50
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
Fig. 1. Scopolamine disrupts acquisition of a water maze spatial discrimination. Percentage correct responses were reduced (A) and the distance traveled until a choice was made was increased. Errors of omission were increased by scopolamine during the first session only (C). Thigmotaxis in terms of the relative time spent in the margin was reduced (D). Data are presented as means with error bars denoting S.E.M.; meth, methylscopolamine.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
51
Fig. 2. Mecamylamine at a dose up to 2 mg / kg had no effect on water maze spatial discrimination accuracy in terms of percentage correct responses (A) or on the distance travelled until a choice was made (B), but increased errors of omission (C) and decreased swim speed (D) during the first training session. Data are presented as means with error bars denoting S.E.M.
52
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
P50.051). Visual inspection of the data suggested that the 1.0 mg dose of CRH tended to disrupt accuracy, while the 0.5 mg dose had no effect (Fig. 3A).
3.3.2. Responsivity Choice latency and choice distance remained unaffected by ICV administration of the peptide (mean choice latency over sessions: saline: 3.6860.03 s, CRH 0.5 mg: 3.6060.09 s, 1.0 mg: 3.7460.07 s; all P.0.05; Fig. 3B). However, analysis revealed an effect of CRH on swim speed (F(2.19)54.66, P50.023), and further post-hoc testing indicated that animals treated with 1.0 mg CRH swum slower than the other two groups. Furthermore, a dose3session interaction was observed for this measure (F(8.76)52.87, P50.021), and post-hoc testing indicated that mice receiving the high dose of CRH (1.0 mg) differed from saline-treated animals during the first but not subsequent sessions (Fig. 3C). CRH did not significantly affect errors of omission at the two doses tested (P.0.05; Fig. 3D). 3.3.3. Thigmotaxis Groups did not differ in the relative time spent or the relative distance traveled in the margin (all P.0.05, data not shown). 3.4. Experiment 4: Effects of CRH on spatial discrimination in scopolamine-treated mice 3.4.1. Accuracy Visual inspection of the data suggested that animals treated with scopolamine plus CRH, but not mice treated with either of the two drugs, differed significantly from saline-treated animals (Fig. 4A). However, three-way analysis showed that this effect was primarily caused by scopolamine: muscarinic blockade induced a significant reduction in percentage correct choices (F(1.33)55.75, P50.022), and a treatment3session interaction was seen (F(4.132)53.31, P50.018), with significant differences to IP saline treatment during the last two training sessions. In contrast, neither CRH nor CRH plus scopolamine caused any additional significant effect (all P.0.05). Thus, statistical analysis suggested that this was primarily a parallel shift, i.e., that CRH caused a small decrease in performance and that the effect of scopolamine was additive. 3.4.2. Responsivity Three-way ANOVA also indicated a main effect of CRH on choice latency and choice distance (F(1.33)510.11, P50.003 and F(1.33)510.96, P50.002, respectively), while scopolamine or combined treatment had no additional significant effect on these measures (all P.0.05; Fig. 4B; mean latency over all training sessions: saline / saline: 4.9860.49 s, saline / scopolamine: 5.7260.34 s, CRH / saline: 7.0360.49 s, CRH / scopolamine: 8.3361.11 s).
ANOVA failed to reveal significant interactions with session and the choice latency measure (all P.0.05), but an interaction between CRH-treatment and session was observed for distance traveled (treatment 13session: F(4.132)55.47, P50.001; treatment 23session and treatment 13treatment 23session: both P.0.05). Further analysis showed that the distance traveled was longer following CRH administration during all but the first session (Fig. 4B). Three-way analysis indicated that CRH on its own led to the increase in errors of omission (F(1.33)513.75, P5 0.001), while scopolamine or combined treatment had no significant effects (both P.0.05; Fig. 4C). This suggests that errors of omission were primarily caused by administration of CRH. Statistical analysis revealed an effect of treatment on swim speed, and this reduction was primarily caused by CRH (F(1.33)530.54, P,0.001), while the CRH3session interaction just failed significance (F(4.132)52.76, P5 0.057; Fig. 4D). No effects of treatment with scopolamine or with scopolamine plus CRH on swim speed were observed (all P.0.05), suggesting that alterations in swim speed were caused by CRH and that the drugs did not interact in the modulation of this parameter.
3.4.3. Thigmotaxis Analysis of the relative time spent in the margin failed to reveal a significant treatment effect (P.0.05; mean relative time over sessions: saline / saline: 37.061.43%, saline / scopolamine: 41.5563.54%, CRH / saline: 38.386 1.66%, CRH / scopolamine: 47.4963.24%). However, analysis of the relative distance traveled in the margin of the maze reached significance, and this effect was primarily caused by scopolamine treatment, which increased the relative distance traveled (F(1.33)56.60, P50.015; treatment 23session: F(4.132)50.23), while CRH on its own or in combination with scopolamine failed to cause additional significant effects (all P.0.05; Fig. 4E). 3.5. Experiment 5: Effects of CRH on spatial discrimination in mecamylamine-treated mice 3.5.1. Accuracy Three-way analysis showed an overall effect of mecamylamine on accurate performance (F(1.28)55.74, P50.024), and the treatment3session interaction also reached significance (F(4.112)52.80, P50.032; Fig. 5A). Post-hoc analysis revealed that mecamylamine impaired performance during the last training session only. Neither CRH nor combined CRH plus mecamylamine treatment induced further changes in accuracy (all P.0.05). Thus, mecamylamine at the 2 mg / kg dose induced a mild impairment in accurate responding, while CRH had no additional effects in Experiment 5. However, floor effects could have confounded these data.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
53
Fig. 3. There was a tendency for impaired accuracy following 1 mg but not 0.5 mg CRH (A). Choice distance (B) and errors of omission (first session; D) were not significantly affected. Swim speed (C) was reduced during initial sessions only. Data are presented as means with error bars denoting S.E.M.
54
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
Fig. 4. Interaction between scopolamine and CRH, effects on water maze spatial discrimination. Percentage correct responses (A), distance traveled until a choice was made (B), errors of omission (C), swim speed (D) and thigmotaxis (E). Data are presented as means with error bars denoting S.E.M.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
55
Fig. 5. Interaction between mecamylamine and CRH, effects on water maze spatial discrimination. Percentage correct responses (A), distance traveled until a choice was made (B), errors of omission (C), swim speed (D) and thigmotaxis (E). Data are presented as means with error bars denoting S.E.M.
56
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
3.5.2. Responsivity ANOVA indicated effects of treatment on both choice latency and choice distance, and both parameters were affected by CRH (F(1.28)520.22, P,0.001 and F(1.28)5 8.69, P50.006, respectively) as well as by nicotinic blockade (F(1.28)55.81, P50.023 and F(1.28)56.43, P50.017, respectively; Fig. 5B; mean latency over all training sessions: saline / saline: 5.3660.53 s, saline / mecamylamine: 5.7960.56 s, CRH / saline: 6.9660.56 s, CRH / mecamylamine: 8.5860.60 s). However, combined treatment had no additional effects (both P.0.05). Furthermore, visual inspection of the data indicated that combined treatment with CRH and mecamylamine caused an increase in errors of omission, but statistical analysis indicated that this effect was primarily due to CRH (CRH3session: F(4.112)55.89, P50.003; mecamylamine3session and treatment 13treatment 23session: both P.0.05; Fig. 5C). As before, CRH slowed swim speed (F(1.28)530.77, P,0.001), while no effects of treatment with mecamylamine or interactions were seen (all P.0.05; Fig. 5D). 3.5.3. Thigmotaxis Three-way analysis confirmed a main effect of mecamylamine (F(1.28)510.09, P50.004), and a tendency for combined treatment to cause additional impairment was seen (F(1.28)53.94; P50.057) when the relative time spent in the margin was analyzed, while CRH on its own had no effect (P.0.05; mean relative time over sessions: saline / saline: 42.1361.68%, saline / mecamylamine: 44.2461.79%, CRH / saline: 37.556 1.79%, CRH / mecamylamine: 46.9261.91%). Likewise, analysis of the relative distance traveled indicated that drug effects on the relative distance traveled was in part caused by mecamylamine, which led to an increase in the relative distance traveled (F(1.28)517.48, P,0.001). CRH on its own had no effect (both P.0.05), but the combination of CRH with mecamylamine caused additional effects (treatment 13treatment 2: F(1.28)55.54; P50.026), suggesting potentiating mechanisms (Fig. 5E).
4. Discussion This study compared the interaction between the muscarinic antagonist scopolamine and the nicotinic antagonist mecamylamine with CRH on water maze spatial discrimination learning. Main findings were a dose-dependent performance deficit caused by scopolamine, as indicated by reduced accuracy (percentage correct responses), decreased responsivity (increased choice latency, choice distance, and errors of omission) and a decrease in the relative time spent, but not the relative distance traveled, in the margin of the maze. Overall swim speed was not
affected, suggesting relatively spared motor function. Methylscopolamine at a dose of 2.0 mg / kg did not significantly differ from saline in any of the measures, suggesting that the scopolamine effects seen in this experiment were primarily centrally mediated. In contrast, mecamylamine had no effect on accuracy in Experiment 2, while three-way analysis of variance revealed a mild impairment after nicotinic blockade in Experiment 5. Comparable to muscarinic blockade, nicotinic blockade resulted in an increase in errors of omission during initial training. Thigmotactic behavior remained unaffected by mecamylamine in the second experiment. Likewise, ICV CRH had only mild effects on accuracy, with a tendency for impairment. Measures of responsivity, such as errors of omission, choice latency and distance, and of thigmotaxis were not significantly affected by CRH. However, swim speed was reduced by the peptide. Contrary to the hypothesis that CRH would attenuate some of the detrimental effects of scopolamine, we observed rather mild synergistic effects of the two drugs on accurate responding. However, for none of the measures did we observe convincing potentiating effects between the two substances. Likewise, combined treatment of CRH with mecamylamine led to some additional effects, but not improvement, both in terms of responsivity and thigmotactic behavior. Only for thigmotactic behaviour did we observe clear evidence for interaction in terms of potentiation between mecamylamine and CRH. Scopolamine is a drug well known to affect spatial performance in a range of maze tasks (e.g. [4,45,50,53,56,70]), presumably by action on both mnemonic and non-mnemonic (i.e., motivational, motor, attentional) mechanisms [60,65,68]. In similar lines, in the present study we observed a decrease in accuracy in combination with reduced responsivity. Thigmotaxis was not significantly affected in terms of the relative distance measure, but the relative time spent in the margin was shortened by scopolamine. It has been argued that the distance measure may be the more reliable index for thigmotactic behavior, while the time measure may be more easily confounded by swim speed [2]. This in turn would suggest that the effects of scopolamine on thigmotaxis in the first experiment were mild at best. Likewise, mecamylamine has been reported to impair mouse and rat maze performance in a range of studies. However, these studies tend to use higher doses [40,43,45], while lower doses in the range used in the present study also failed to affect accurate performance [12,40,45]. Moreover, non-mnemonic mechanisms are altered by mecamylamine [65,72], and the present study also showed effects on errors of omission and swim speed at doses well below those known to affect maze learning. The literature on the effects of CRH on learning and memory is even more equivocal: some studies report performance deficits following ICV CRH administration [3] or CRH overexpression in mice [28], while others find
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
‘cognitive enhancement’ [11,32,39]. Part of this discrepancy may be due to the exact site of action [3,29,54,64] and the CRH receptor subtype activated by the peptide. For example, it has been reported that CRH facilitates fear conditioning via activation of the CRH 1 receptor subtype but impairs performance in this task via activation of the CRH 2 receptor subtype [54]. Part of the discrepancy may also be due to the fact that the majority of studies did not allow a clear dissociation between mnemonic and non-mnemonic effects of the peptide. In the present study, the effects of CRH on accuracy were rather mild at the doses tested but, if at all, CRH at the doses tested tended to impair percentage correct responses. The most clear-cut effect was seen in a reduction in swim speed during initial training sessions. This is consistent with the hypoactivity often seen in novel environments following CRH administration [23], which has been suggested to reflect altered anxiety-related behavior rather than altered motor function [64]. Clearly, alterations in anxiety-related behavior induced by CRH could have confounded the results of the present and previous studies, and have been shown to at least in part contribute to the performance deficits seen in some learning paradigms [28]. This may be of particular relevance with respect to the interaction studies as both scopolamine and mecamylamine have been shown to also increase anxiety-like behavior in rats, both after systemic administration and following intra-hippocampal injection [24,62,63]. In general, the effects of combined treatment with scopolamine and CRH were, if at all, synergistic, but only very mild additive effects were observed and no potentiation of drug effects was seen. Combined nicotinic blockade and CRH activation also produced only limited synergistic effects, with evidence for potentiation in thigmotaxis. Given that CRH increases hippocampal cholinergic activity, these rather mild and, if at all, synergistic effects seem unexpected. One possible explanation could be that we missed the optimal dose range of CRH and overstimulated the cholinergic system, thus leading to overarousal and hence performance deficits. However, given the rather clear cut independency of effects and the relative paucity of clear interaction effects, this explanation seems unlikely. Alternatively, mice receiving a combination of CRH and cholinergic blockade by scopolamine or mecamylamine might have been overresponsive to the stimuli related to the test environment, as the cholinergic system may be of relevance for increasing the signal-to-noise ratio for salient events, thereby reducing the effects of distracting extraneous stimuli [55]. Consequently, scopolamine or mecamylamine may render an animal hyperresponsive to its environment as it cannot monitor and amend its behavior in an appropriate manner when exposed to fearful stimuli [5,63], and CRH may have worsen this state. In fact, it has been suggested that the exaggerated stress response in the absence of functional cholinergic activity might be due to overestimation of the magnitude of threat
57
as a consequence of hypersensitivity to environmental stimuli normally filtered out [63]. Thus, in addition to an already compromised stimulus processing and an overactive HPA axis, additional treatment with CRH would be expected to further augment the scopolamine- (and possibly also mecamylamine)-induced HPA hyperactivity, which might explain the mild synergistic effects of combined treatment. Synergistic effects could also be expected as cholinergic blockade at hippocampal level leads to increased CRH activity at the level of the paraventricular hypothalamic nucleus (PVN) through disruption of the descending hippocampal pathway, thereby enhancing the hypothalamic–pituitary–adrenal axis response [63]. However, it should be noted that ACh stimulates CRH activity at HPA level when injected directly into the PVN and that this effect is blocked by the muscarinic antagonist atropine [47]. Therefore, it is difficult to disentangle the net effect of systemic scopolamine on CRH release from the PVN at this stage. Yet another explanation could be that CRH impaired performance by acting at the CRH 2 receptor subtype [54], i.e., that it is the combination of muscarinic blockade and CRH 2 receptor stimulation which is required to cause the additive effects, while the potentially beneficial effects of CRH 1 receptor stimulation on cholinergic activity might be insufficient to overcome the muscarinic block, especially as scopolamine on its own already increases hippocampal ACh release [30,59,71]. Moreover, it has been noted that CRH also has an indirect inhibitory input on the cholinergic basal forebrain: ICV administration of CRH decreases sodium-dependent high affinity choline uptake (HACU) at the level of the rat frontal cortex and hippocampus. This effect can be blocked by naltrexone, suggesting that CRH exerts its inhibitory action through interaction with the opioidergic system [38]. If this mechanism would predominate, it could offer yet another explanation for the present results. These findings may be of relevance for psychiatric disorders such as depression and dementia: an unrestrained CRH hyperdrive has been proposed to play a major role in the development and course of depression [31] and cholinergic hyperactivity has been found in this disorder [35]. Conversely, in Alzheimer’s disease, a decrease in cholinergic function is well documented [14,15,46,51,52], which coincides with a decrease in cortical CRH and a reciprocal increase in cortical CRH receptors [15,20,21]. Stimulation of parts of the CRH system has been proposed as a potential treatment for Alzheimer’s disease [3]. The present results fail to support this concept and would suggest that this may even worsen the cognitive deficit seen in these patients. In summary, the present data lend only limited support for the concept of important interactions between CRH and ACh in the modulation of learning and memory. Moreover, the behavioral consequences of cholinergic dysfunction seem to be enhanced and not antagonized by CRH.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59
58
Acknowledgements We thank A. Mederer for excellent technical support.
References [1] E. Acquas, C. Wilson, H.C. Fibiger, Conditioned and unconditioned stimuli increase frontal cortical and hippocampal acetylcholine release: effects of novelty, habituation, and fear, J. Neurosci. 16 (1996) 3089–3096. [2] M. Arns, M. Sauvage, T. Steckler, Excitotoxic lesions disrupt allocentric spatial learning in mice: effects of strain and task demands, Behav. Brain Res. 106 (1999) 151–164. [3] D.P. Behan, S.C. Heinrichs, J.C. Troncoso, X.J. Liu, C.H. Kawas, N. Ling, E.B. De Souza, Displacement of corticotropin releasing factor from its binding protein as a possible treatment for Alzheimer’s disease, Nature 378 (1995) 284–287. [4] J. Berger-Sweeney, A. Arnold, D. Gabeau, J. Mills, Sex differences in learning and memory in mice: effects of sequence of testing and cholinergic blockade, Behav. Neurosci. 109 (1995) 859–873. [5] S. Bhatnagar, B. Costall, J.W. Smythe, Hippocampal cholinergic blockade enhances hypothalamic–pituitary–adrenal responses to stress, Brain Res. 766 (1997) 244–248. [6] J.C. Bittencourt, P.E. Sawchenko, Do centrally administered neuropeptides access cognate receptors?: An analysis in the central corticotropin-releasing factor system, J. Neurosci. 20 (2000) 1142– 1156. [7] A.E. Calogero, W.T. Gallucci, R. Bernardini, C. Saoutis, P.W. Gold, G.P. Chrousos, Effect of cholinergic agonists and antagonists on rat hypothalamic corticotropin-releasing hormone secretion in vitro, Neuroendocrinology 47 (1988) 303–308. [8] I. Ceccarelli, F. Casamenti, C. Massafra, G. Pepeu, C. Scali, A.M. Aloisi, Effects of novelty and pain on behavior and hippocampal extracellular ACh levels in male and female rats, Brain Res. 815 (1999) 169–176. [9] D.T. Chalmers, T. W Lovenberg, E.B. De Souza, Localization of novel corticotropin-releasing factor receptor (CRF 2 ) mRNA expression to specific subcortical nuclei in rat brain: comparison with CRF 1 receptor mRNA expression, J. Neurosci. 15 (1995) 6340– 6350. [10] J.C. Chen, K.L. Brunson, M.B. Muller, W. Cariaga, T. Z Baram, Immunocytochemical distribution of corticotropin-releasing hormone receptor type-1 (CRF1)-like immunoreactivity in the mouse brain: light microscopic analysis using an antibody directed against the C-terminus, J. Comp. Neurol. 420 (2000) 305–323. [11] M.F. Chen, T.H. Chiu, E.H. Lee, Noradrenergic mediation of the memory-enhancing effect of corticotropin-releasing factor in the locus coeruleus of rats, Psychoneuroendocrinology 17 (1992) 113– 124. [12] Y. Chen, E. Shohami, S. Constantini, M. Weinstock, Rivastigmine, a brain-selective acetylcholinesterase inhibitor, ameliorates cognitive and motor deficits induced by closed-head injury in the mouse, J. Neurotrauma 15 (1998) 231–237. [13] R. Cozzolino, D. Guaralsi, A. Giuliani, O. Ghirardi, M.T. Ramacci, L. Angelucci, Effects of concomitant nicotinic and muscarinic blockade on spatial memory disturbance in rats are purely additive. Evidence from the Morris water task, Physiol. Behav. 56 (1994) 111–114. [14] P. Davies, A.J.F. Maloney, Selective loss of cholinergic neurons in Alzheimer’s disease, Lancet 25 (1976) 1403. [15] K.L. Davis, R.C. Mohs, D.B. Martin, D.P. Purohit, D.P. Perl, M. Lantz, G. Austin, V. Haroutunian, Neuropeptide abnormalities in patients with early Alzheimer’s disease, Arch. Gen. Psychiatry 56 (1999) 981–987.
[16] J.C. Day, M. Koehl, M. Le Moal, S. Maccari, Corticotropinreleasing factor administered centrally, but not peripherally, stimulates hippocampal acetylcholine release, J. Neurochem. 71 (1998) 622–629. [17] J.C. Day, M. Koehl, V. Deroche, M. Le Moal, S. Maccari, Prenatal stress enhances stress- and corticotropin-releasing factor-induced stimulation of hippocampal acetylcholine release, J. Neurosci. 18 (1998) 1886–1892. [18] M.W. Decker, M.J. Majchrzak, Effects of systemic and intracerebroventricular administration of mecamylamine, a nicotinic cholinergic antagonist, on spatial memory in rats, Psychopharmacology 107 (1992) 530–534. [19] E.B. De Souza, G. Battaglia, Increased corticotropin-releasing factor receptors in rat cerebral cortex following chronic atropine treatment, Brain Res. 397 (1986) 401–404. [20] E.B. De Souza, P.J. Whitehouse, M.J. Kuhar, D.L. Price, W.W. Vale, Reciprocal changes in corticotropin-releasing factor immunoreactivity and CRF receptors in cerebral cortex of Alzheimer’s disease, Nature 319 (1986) 593–595. [21] E.B. De Souza, P.J. Whitehouse, M.J. Kuhar, D.L. Price, W.W. Vale, Abnormalities in corticotropin-releasing hormone (CRH) in Alzheimer’s disease and other human disorders, Ann. N.Y. Acad. Sci. USA 512 (1987) 237–247. [22] J.D. Dudar, I.Q. Whishaw, J.C. Szerb, The role of the septal nuclei in the release of acetylcholine from the hippocampus of freely moving rats during sensory stimulation and running, Neuropharmacology 18 (1979) 673–678. [23] A.J. Dunn, C.W. Berridge, Physiological and behavioral responses to corticotropin-releasing factor administration: is CRF a mediator of anxiety or stress responses? Brain Res. Rev. 15 (1990) 71–100. [24] S.E. File, P.J. Kenny, S. Cheeta, The role of the dorsal hippocampal serotonergic and cholinergic systems in the modulation of anxiety, Pharmacol. Biochem. Behav. 66 (2000) 65–72. [25] K.B.J. Franklin, G. Paxinos, The Mouse Brain in Stereotaxis Coordinates, Academic Press, San Diego, 1997. [26] Y. Fu, S.G. Matta, J.D. Valentine, B.M. Sharp, Adrenocorticotropin response and nicotine-induced norepinephrine secretion in the rat paraventricular nucleus are mediated through brainstem receptors, Endocrinology 138 (1997) 1935–1943. [27] G.M. Gilad, B.D. Mahon, Y. Finkelstein, B. Koffler, V.H. Gilad, Stress induced activation of the hippocampal cholinergic system and the pituitary adrenocortical axis, Brain Res. 347 (1985) 404–408. [28] S.C. Heinrichs, M.P. Stenzel-Poore, L.H. Gold, E. Battenberg, F.E. Bloom, G. Koob, W.W. Vale, E. Merlo Pich, Learning impairment in transgenic mice with central over expression of corticotropin-releasing factor, Neuroscience 74 (1996) 303–311. [29] S.C. Heinrichs, E.A. Vale, J. Lapsansky, D.P. Behan, L.V. McClure, N. Ling, E.B. De Souza, G. Schulteis, Enhancement of performance in multiple learning tasks by corticotropin-releasing factor-binding protein ligand inhibitors, Peptides 18 (1997) 711–716. [30] M. Hiramatsu, H. Murasawa, T. Nabeshima, T. Kameyama, Effects of U-50,488H on scopolamine-, mecamylamine- and dizocilpineinduced learning and memory impairment in rats, J. Pharmacol. Exp. Ther. 284 (1998) 858–867. [31] F. Holsboer, The rationale for corticotropin-releasing hormone receptor (CRH-R) antagonists to treat depression and anxiety, J. Psychiat. Res. 33 (1999) 181–214. [32] H.C. Hung, C.K. Chou, T.H. Chiu, E.H. Lee, CRF increases protein phosphorylation and enhances retention performance in rats, Neuroreport 2 (1992) 181–184. [33] A. Imperato, S. Puglisi-Allegra, P. Casolini, L. Angelucci, Changes in brain dopmaine and acetylcholine release during and following stress are independent of the pituitary–adrenocortical axis, Brain Res. 538 (1991) 111–117. [34] F.M. Inglis, H.C. Fibiger, Increases in hippocampal and frontal cortical acetylcholine release associated with presentation of sensory stimuli, Neuroscience 66 (1995) 81–86.
T. Steckler, F. Holsboer / Brain Research 906 (2001) 46 – 59 [35] D.S. Janowsky, D.H. Overstreet, J.I. Nurnberger Jr., Is cholinergic sensitivity a genetic marker for the affective disorders? Am. J. Med. Genet. 54 (1994) 335–344. [36] J.W. Kasckow, A. Regmi, S. Sheriff, J. Mulchahey, T.D. Geracioti Jr., Regulation of corticotropin-releasing factor messenger RNA by nicotine in an immortalized amygdalar cell line, Life Sci. 65 (1999) 2709–2714. [37] R. Kortekaas, B. Costall, J.W. Smythe, Changes in hippocampal theta following intrahippocampal corticotropin-releasing hormone (CRH) infusions in the rat, Brain Res. Bull. 48 (1999) 603–607. [38] H. Lai, M.A. Carino, Effects of noise on high-affinity choline uptake in the frontal cortex and hippocampus of the rat are blocked by intracerebroventricular injection of corticotropin-releasing factor antagonist, Brain Res. 527 (1990) 354–358. [39] E.H. Lee, C.P. lee, H.I. Wang, W.R. Lin, Hippocampal CRF, NE, and NMDA system interactions in memory processing in the rat, Synapse 14 (1993) 144–153. [40] E.D. Levin, M. Castonguay, G.D. Ellison, Effects of nicotinic receptor blocker mecamylamine on radial-arm maze performance in rats, Behav. Neural Biol. 48 (1987) 206–212. [41] S.G. Matta, J. Singh, B.M. Sharp, Catecholamines mediate nicotineinduced adrenocroticotropin secretion via alpha-adrenergic receptors, Endocrinology 127 (1990) 1646–1655. [42] S.G. Matta, J.D. Valentine, B.M. Sharp, Nicotine activation of CRH neurons in extrahypothalamic regions of the rat brain, Endocrine 7 (1997) 245–253. [43] T. Maurice, T.P. Su, D.W. Parish, T. Nabeshima, A. Privat, PRE084, a sigma selective PCP derivative, attenuates MK-801-induced impairment of learning in mice, Pharmacol. Biochem. Behav. 49 (1994) 859–869. [44] W.J. Millard, G.W. Arendash, A.J. Dunn, E.M. Meyer, Effects of nucleus basalis lesions on cerebral cortical concentrations of corticotropin-releasing hormone (CRH)-like immunoreactivity in the rat, Neurosci. Lett. 113 (1990) 233–239. [45] P.M. Moran, Differential effects of scopolamine and mecamylamine on working and reference memory in the rat, Pharmacol. Biochem. Behav. 45 (1993) 533–538. [46] P.A. Newhouse, Alzheimer’s disease and the cholinergic system: an introduction to clinical pharmacological research, in: L.L. Heston (Ed.), Progress in Alzheimer’s Disease and Similar Conditions, American Psychiatric Press, Washington, 1997, pp. 213–231. [47] N. Ohmori, K. Itoi, F. Tozawa, Y. Sakai, K. Sakai, N. Horiba, H. Demura, T. Suda, Effects of acetylcholine on corticotropin-releasing factor gene expression in the hypothalamic paraventricular nucleus of conscious rats, Endocrinology 136 (1995) 4858–4863. [48] P. Onali, M.C. Olianas, Identification and characterization of muscarinic receptors potentiating the stimulation of adenylyl cyclase activity by corticotropin releasing hormone in membranes of rat frontal cortex, J. Pharmacol. Exp. Ther. 286 (1998) 753–759. [49] J.M. Overton, G. Davis-Gorman, L.A. Fisher, Central nervous system cardiovascular actions of CRF in sinuaortic-denervated rats, Am. J. Physiol. 258 (1990) R596–R601. [50] D.B. Peele, S.P. Baron, Effects of selection delays on radial maze performance: acquisition and effects of scopolamine, Pharmacol. Biochem. Behav. 29 (1988) 143–150. [51] E.K. Perry, The cholinergic hypothesis – 10 years on, Br. Med. Bull. 42 (1986) 63–69. [52] E.K. Perry, B.E. Tomlinson, G. Blessed, K. Bergman, P.H. Gibson, R.H. Perry, Correlation of cholinergic abnormalities with senile plaques and mental test scores in senile dementia, Br. Med. J. 2 (1978) 1457–1459. [53] J.J. Pilcher, G.R. Sessions, S.A. McBride, Scopolamine impairs spatial working memory in the radial maze: an analysis by error type and arm choice, Pharmacol. Biochem. Behav. 58 (1997) 449–459. ¨ [54] J. Radulovic, A. Ruhmann, T. Liepold, J. Spiess, Modulation of learning and anxiety by corticotropin-releasing factor (CRF) and
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64] [65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
59
stress: differential roles of CRF receptors 1 and 2, J. Neurosci. 19 (1999) 5016–5025. R.T. Richardson, M.R. DeLong, A reappraisal of the functions of the nucleus basalis of Meynert, Trends Neurosci. 11 (1988) 264– 267. ¨ M. Aaltonen, P. Riekkinen, Effects of P. Riekkinen Jr., J. Sirvio, concurrent manipulations of nicotinic and muscarinic receptors on spatial and passive avoidance learning, Pharmacol. Biochem. Behav. 37 (1990) 405–410. ¨ P. Riekkinen, Pharmacological conseP. Riekkinen Jr., J. Sirvio, quences of cholinergic plus serotonergic manipulations, Brain Res. 527 (1991) 23–26. T.W. Robbins, G. McAlonan, J.L. Muir, B.J. Everitt, Cognitive enhancers in theory and practice: studies of the cholinergic hypothesis of cognitive deficits in Alzheimer’s disease, Behav. Brain Res. 83 (1997) 15–23. J.K. Robinson, A. Zocchi, A. Pert, J.N. Crawley, Galanin microinjected into the medial septum inhibits scopolamine-induced acetylcholine overflow in the rat ventral hippocampus, Brain Res. 709 (1996) 81–87. A. Sahgal, A.B. Keith, S. Lloyd, J.M. Kervin, E.K. Perry, J.A. Edwardson, Memory following cholinergic (NBM) and noradrenergic (DNAB) lesions made singly or in combination: potentiation of disruption by scopolamine, Pharmacol. Biochem. Behav. 37 (1990) 597–605. M. Sauvage, T. Steckler, Detection of corticotropin-releasing hormone receptor 1 immunoreactivity in cholinergic, dopaminergic and noradrenergic neurons of the basal forebrain and brainstem nucleipotential implication for arousal and attention, Neuroscience (2001) in press. J.W. Smythe, S. Bhatnagar, D. Murphy, C. Timothy, B. Costall, The effects of intrahippocampal scopolamine infusions on anxiety in rats as measured by the black–white box test, Brain Res. Bull. 45 (1998) 89–93. J.W. Smythe, D. Murphy, S. Bhatnagar, C. Timothy, B. Costall, Muscarinic antagonists are anxiogenic in rats tested in the black– white box, Pharmacol. Biochem. Behav. 54 (1996) 57–63. T. Steckler, F. Holsboer, Corticotropin-releasing hormone receptor subtypes and emotion, Biol. Psychiatry 46 (1999) 1480–1508. T. Steckler, A.B. Keith, R.G. Wiley, A. Sahgal, Cholinergic lesions by 192 IgG-saporin and short-term recognition memory: role of the septohippocampal projection, Neuroscience 66 (1995) 101–114. T. Steckler, A. Sahgal, The role of serotonergic–cholinergic interactions in the mediation of cognitive behaviour, Behav. Brain Res. 67 (1995) 165–199. T. Steckler, A. Sahgal, J.P. Aggleton, W.H.I.M. Drinkenburg, Recognition memory in rats. III. Neurochemical substrates, Prog. Neurobiol. 54 (1998) 333–348. T. Steckler, C. Weis, M. Sauvage, A. Mederer, F. Holsboer, Disrupted allocentric but preserved egocentric spatial learning in transgenic mice with impaired glucocorticoid receptor function, Behav. Brain Res. 100 (1999) 77–89. Y. Tizabi, A.E. Calogero, Effects of various neurotransmitters and neuropeptides on the release of corticotropin-releasing hormone from the rat cortex in vitro, Synapse 10 (1992) 341–348. A. Urani, A. Privat, T. Maurice, The modulation of neurosteroids of the scopolamine-induced learning impairment in mice involves an interaction with sigma1 (sigma1) receptors, Brain Res. 799 (1998) 64–77. M.G. Vannucchi, C. Scali, S.R. Kopf, G. Pepeu, F. Casamenti, Selective muscarinic antagonists differentially affect in vivo acetylcholine release and memory performances of young and aged rats, Neuroscience 79 (1997) 837–846. D.V. Widzowski, E. Cregan, P. Bialobok, Effects of nicotinic agonists and antagonists on spatial working memory in normal adult and aged rats, Drug Dev. Res. 31 (1994) 24–31.