- Email: [email protected]
AF64A (ethylcholine aziridinium ion), a cholinergic neurotoxin, selectively impairs working memory in a multiple component T-maze task
Brain Research, 414 (1987) 15-21 Elsevier
15
BRE 12640
AF64A (ethylcholine aziridinium ion), a cholinergic neurotoxin, selectively impairs working memory in a multiple component T-maze task James J. C h r o b a k 1, Israel H a n i n 2 and T h o m a s J. Walsh 1 1Department of Psychology, Rutgers University, New Brunswick, NJ08903 (U.S.A.) and 2Departmentof Pharmacology and Experimental Therapeutics, Loyola UniversitySchool of Medicine, Maywood, IL 60153 (U.S.A.) (Accepted 4 November 1986)
Key words: AF64A; Cholinergic neurotoxin; Learning and memory; Working memory; Acetylcholine; Hippocampus
The present study examined the nature of the cognitive deficits associated with a selective decrease of cholinergic activity in the hippocampus. Male Fischer rats were trained to perform a multiple component T-maze task which simultaneously assessed their ability to perform on the basis of trial-specific information (working memory) and trial-independent information (reference memory). Following 125 acquisition trials rats were bilaterally injected with AF64A (3 nmol/side) or artificial CSF into the lateral ventricles and allowed 14 days to recover before behavioral testing resumed. The controls rapidly returned to their preoperative level of performance on both components of the maze task. AF64A-treated animals were transiently impaired on the reference memory task. Their performance rapidly improved and they were performing at preoperative levels within 4 days of testing. In contrast, these animals exhibited a marked and long-lasting impairment in their performance of the working memory component. After behavioral testing was completed, neurochemical analysis revealed that AF64A produced a significant decrease in choline acetyltransferase (CHAT) activity in the hippocampus (43%) 42 days following surgery. This dosing regimen produced no alterations of striatal or cortical ChAT activity. These data suggest that alterations of hippocampal cholinergic activity severely impair an animal's ability to perform working memory tasks. INTRODUCTION Central cholinergic systems appear to be a critical neurobiological substrate for learning and m e m o r y processes. Pharmacological studies in both humans and laboratory animals have shown that cholinergic agonists improve, while antagonists impair, cognition 1'6. These studies provided evidence that memory processes are dependent upon the integrity of cholinergic systems. Analysis of behavioral changes induced by damage to the cholinergic nuclei in the basal forebrain 7'n'12'16'27 and the observation that these same nuclei degenerate in patients with various forms of dementia further support the involvement of forebrain cholinergic systems in cognitive functionS'361 In senile dementia of the Alzheimer type (SDAT) the decreases in cholinergic parameters have been shown to correlate with the severity of the
cognitive impairment 29'37. While other neurotransmitters or peptides might be altered in subpopulations of S D A T patients or late in the course of the disease, these alterations have not been shown to be related to the prevailing cognitive impairments 2°. A further understanding of the behavioral properties of cholinergic systems should provide insight into the neurobiology of learning and m e m o r y and could also provide information about the pathophysiology of S D A T and potential approaches to its treatment. Ethylcholine aziridinium ion (AF64A) is a neurotoxic analog of choline which disrupts the high affinity transport of choline ( H A C h T ) , the rate limiting step in the synthesis of acetylcholine (ACh) l°,2t. Injection of A F 6 4 A into the lateral ventricles produces a persistent reduction in presynaptic cholinergic
Correspondence: T.J. Walsh, Rutgers University, Department of Psychology, Busch Campus, New Brunswick, NJ 08903, U.S.A. 0006-8993/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)
16 markers including HAChT, choline acetyltransferase (CHAT) activity, ACh concentrations, and K+-in duced release of ACh from hippocampai slices Is' 21-23,32,34 This loss of cholinergic markers is not accompanied by alterations in the regional content of norepinephrine, dopamine, serotonin, their metabolites, or choline 4'34. We recently reported that injection of AF64A into the lateral ventricles or the dorsal hippocampus impairs learning and memory processes 34'3-~. For example, administration of relatively large doses of AF64A into the lateral ventricles reduced ACh content 45-60% in the hippocampus (7.5 or 15 nmol/ side) and frontal cortex (15 nmol/side). This neurochemical change was accompanied by a significant impairment in the retention of a passive avoidance response and the acquisition of an 8-arm radial maze task 34'35. Doses of AF64A lower than 7.5 nmol/side appear to produce a selective loss of hippocampal cholinergic markers 4"34. The studies presented here attempt to further characterize the nature of the cognitive impairments associated with this selective decrease of cholinergic activity in the hippocampus. We were particularly interested in whether loss of hippocampal cholinergic tone would selectively impair memory for trial-specific information (working memory) but not for trial-independent information (reference memory). Working memory involves flexible stimulus-response associations whose value depends on recently experienced events, whereas reference memory involves fixed stimulus-response associations whose values are not altered over consecutive trials 28. This dichotomy in memory systems is similar to the pattern of impaired and preserved memory processes observed in a variety of human, primate and rodent amnesic syndromes which involve hippocampal damage 26'2s'31. To further examine the potential dissociation of cognitive impairments produced by AF64A we developed a T-maze task which can simultaneously assess the animals ability to perform, based upon trialindependent and trial-specific information within the same trial. In this task the rat is placed into one of two start/goal locations on either arm of the maze. It is trained to first run down the stem of the maze, and then to return to the alternate start/goal location on each trial. The different aspects of this task can be characterized as (a) a trial-independent (reference
memory) component which involves an invariant response of proceeding down the stem of the maze regardless of the start location, and (b) a trial-specific (working memory) component of returning to the alternate start/goal location. Successful performance of the latter requires the animal to use some representation or memory of the initial start/goal location for any trial. MATERIALS AND METHODS
Subjects Male Fischer 344 rats (Charles River Breeders, Wilmington, DE) approximately 90 days old and weighing between 250 and 300 g were used in the following experiment(s). The animals were individually housed and maintained in a colony room with a 12 h light/dark cycle. Water was freely available at all times. Five days prior to acquisition and reacquisition trials animals were reduced to 85% of their free-feeding body weight by limiting their daily ration of food. This level was maintained throughout testing except for 7 days immediately prior and subsequent to surgery. Over the course of the experiment the rats were allowed to gain 5 g of weight each week.
Apparatus A white plexiglass T-maze with white vinyl flooring was utilized in this experiment. The stem of the maze was 71 cm long and each arm extended 36 cm from the stem. Arms were 14 cm high and 14 cm wide. A removable plexiglass barrier could be positioned at the end of each arm providing a 15 cm start compartment. The maze was elevated approximately 46 cm above the floor in the center of a small room which contained a rich array of visual stimuli.
Preparation of ethylcholine mustard aziridinium ion (AF64A) AF64A was prepared from 10 mM acetylethylcholine mustard hydrochloride by dissolving the starting material in distilled water, adjusting the pH to 11.5-11.7 with concentrated NaOH, and then vigorously stirring the solution for 20 min 8. The solution was next diluted with artificial CSF, and finally was titrated to a pH of 7.3-7.4 using HC1. Solutions of AF64A were always prepared immediately prior to use and kept on ice at all times.
17
Intracerebroventricular injection of AF64A Six days following the preoperative behavioral tests animals were anesthetized with sodium pentobarbital (45 mg/kg) and positioned in a stereotaxic instrument. Blunt ear bars were used to minimize damage to the auditory meatus. A sagittal incision was made in the scalp, and two holes were drilled through the skull for placement of the injection cannula into the lateral ventricles. Rats were bilaterally infused with isotonic CSF or 3 nmol (6 nmol/brain) of AF64A. Solutions were slowly infused (0.125/A/min) into each site using a 1 ~1 Hamilton syringe positioned in a microinjector unit mounted on the stereotaxic frame. A total volume of 3.0/tl was delivered per side. To promote the local diffusion of the perfusate the injection cannula was left in place for a period of 3 min following injection.
T-maze acquisition Animals were given 5 days of habituation trials in which they were free to explore the maze for a 5-rain period. During these trials food pellets (94 mg chocolate-flavored) were scattered throughout the maze. Following the habituation period each rat was given 5 trials a day for 5 weeks, for a total of 125 trials. There was a variable inter-trial interval of 30-90 s. Each trial began by placing the rat into either start/goal location for a 10-s period. The guillotine door was then T-MAZE TASK
I CORRECT I
ERROR- - - - - - ~
-
ol START RIGHT
LEFT
• TRAVERSE STEM
- RETU RN TO ALTERNATE S/G BOX (REFERENCE ERROR--) ( WORKINGERROR - - - )
TRIAL
FT
.I Io
START LEFT
=TRAVERSE STEM
- RETURN TO ALTERNATE S/G BOX
RANDOMLY ALTERNATED TRIALS Fig. l. A schematic representation of the T-maze task is presented. The rat is initially placed into the right (right trial) or left (left trial) start compartment. The reference memory component of the task consists of turning into and transversing the stem of the maze. The working memory component consists of running to the alternate start/goal box on any trial.
manually opened and the animal was allowed 60 s to complete the trial. The animal was rewarded for turning into and traversing the stem (see Fig. 1) and for subsequently returning to the alternate start/goal location. If the animal crossed directly to the alternate start/goal location without transversing the stem it was promptly removed and the trial ended. If the animal returned to the original start location it was removed and the trial ended. At the end of each trial the animal was returned to a cage located in the maze room. The choice of start location varied quasi-randomly, with the provision that no more than 3 trials during a given day start from the same location. Following the 125 acquisition trials animals were performing with 85-95% accuracy on both the working and reference memory components of the task. Animals were then randomly assigned to either the AF64A or control group. They were allowed 6 days to recover body weight and were injected with AF64A or CSF according to the surgical protocol described above. The number of working and reference memory errors were recorded each day for the session of 5 trials. The data were grouped by 3-day blocks and analyzed according to the percentage correct per block.
Neurochemical analysis Rats were sacrificed 42 days following surgery, after behavioral testing was completed. Their brains were rapidly removed and dissected on ice into the following regions: hippocampus, frontal cortex and striatum. ChAT was measured according to the radiometric assay of McCaman and Hunt 24 as modified by Fonnum 9. Protein content of tissue was determined according to Lowry et al. t9. Since previous studies have shown that this dose of AF64A does not alter indices of catechol- or indoleamine function 3.4 we only examined cholinergic parameters in the present study. RESULTS
Animals injected with CSF rapidly recovered from surgery and appeared normal within 2 days. Rats in the AF64A group displayed a transient syndrome consisting of diarrhea, piloerection, aphagia, adipsia, and hyperreactivity. Following this initial postoperative period, animals rapidly recovered and their
18 TABLE I
REFERENCE MEMORY ACQUISITION
Effects of AF64A (3 nmol/side, i.c.v.) on choline acetyltransferase activity
REACQUISITION
-__-- _---_
,oo
r,~.,._._;U~._~___ ~
,oo
Values are nmol A C h formed/mg protein/h. a= 80
CSF (n = 9) A F 6 4 A (n = 12)
Hippocampus
Frontal cortex
Striatum
123 + 3 71 + 7*
70 _+ 3 84 + 6
197 + 13 204 + 15
(2
80
70
Z C~ 60 bJ o- 50
* P < 0.01 vs CSF, Student's t-test.
/'
7O t CSF (N;9) IO--'OAF(N 12) I
~/
60
~
50
i
i
i
i
i
15
30
45
60
75
ii
i
i
0 105 120
l
l
I
I
15
50
45
60
.J
75
TRIALS
appearance and overt behavior was indistinguishable from that of controls. We observed no mortality with the 3 nmol dose of AF64A. ChAT activity was significantly reduced (43%) in the hippocampus of AF64A-treated animals (see Table I). No significant alterations of cortical or striatal ChAT activity were observed following AF64A although ChAT activity in the frontal cortex was 20% higher in this group. Prior to surgery both groups were performing with comparable accuracy on the reference and working memory components of the task. Both groups of rats were performing with greater than 90% accuracy on the reference memory task after 75 training trials. However, this level of performance on the working memory component was not achieved until the last 15 trials. Analysis of preoperative performance using a twoway repeated A N O V A revealed no significant group effects (P > 0.1) or group by trial interactions (P > 0.1) for either working or reference memory. However, there was a significant trial effect for both components of the task (F7,13 3 -- 46.3-107.9, P < 0.01). These analyses indicate that the performance of both groups improved at the same rate over time and they were performing with comparable accuracy prior to surgery (see Figs. 2 and 3). The performance of the CSF-injected group following surgery was not significantly different from their preoperative levels of performance on either the reference or working memory task. In contrast, the AF64A group exhibited a short-term reversible impairment of reference memory and a severe and persistent impairment of working memory. The behavior of AF64A-treated animals during testing suggested no alteration of motivation or motoric ability which might have interfered with task
Fig. 2. The effects of A F 6 4 A on the acquisition and the reacquisition of the reference m e m o r y c o m p o n e n t of the T-maze task described in Fig. 1. This figure presents the percentage of correct responses summarized over each 3-day block of sessions, 5 trials per day. Data are presented as m e a n percent correct + S.E.M. Following the acquisition trials rats were injected with A F 6 4 A (3 nmol/side, i.c.v.), allowed 14 days to recover from surgery, and were then re-exposed to the behavioral task (reacquisition), * P < 0.01 vs CSF controls, Spjotvall-Stoline t-test.
performance. These animals readily navigated the Tmaze, consumed the pellets, and completed each trial in the same amount of time as the CSF-injected controls. Furthermore, we have observed that this dose of AF64A does not alter locomotor activity4. Reacquisition of the reference memory task was significantly slower in the AF64A group. A two-way
WORKING M E M O R Y ACQUISITION
REACOUISlTION
I00
I00
90
90
8O
8O
cc 70 nO t.~ 60 pZ
70
I-
~ Q.
i
60' /
5o 4O 30
/
5O 4O
/
~ i 15
i,/
~° l 30
i 45
i 60
~ 75
i 90
i i 105 120
15 30
45
60
75
TRIALS
Fig. 3. The effects of A F 6 4 A on the acquisition and reacquisition of the working m e m o r y c o m p o n e n t of the T-maze task described in Fig. 1. This figure presents the percentage of correct responses summarized over each 3-day block of sessions, 5 trials per day. Data are presented as mean percent correct _+ S.E.M. * P < 0.01 vs CSF controls, Spjotvall-Stoline t-test.
19 repeated ANOVA revealed significant treatment (El,19 = 12.2, P < 0.01) and trials effects (F4.76 = 13.2, P < 0.01), as well as a significant interaction (F4,76 = 6.8, P '~ 0.01). Post-hoc comparisons utilizing the Spjotvall-Stoline t-test 14 revealed that the performance of the AF64A group was significantly different from that of controls only during the first block of reacquisition trials (P < 0.01). Animals treated with AF64A were performing as well as the controls within 4 days of postoperative testing. A two-way repeated ANOVA demonstrated that the AF64A group was significantly impaired in their performance of the working memory component (F1,19 = 13.9, P < 0.01). The analysis showed a significant trials effect (F4,76 = 13.4, P < 0.01) but no treatment by trials interaction (F4,76 = 1.7, P > 0.10). The performance of the control group on the working memory task steadily improved during reacquisition trials (see Fig. 3), while the AF64A-treated animals remained impaired despite a small improvement observed following the first block of trials. The performance of AF64A-treated rats on the working memory component did not improve beyond the first 3 trials block even when they were optimally performing the reference memory part of the task. DISCUSSION
Bilateral injection of a low dose of AF64A into the cerebroventricles produced a significant decrease in hippocampal, but not cortical or striatal, ChAT activity. This biochemical deficit was associated with a significant impairment in the animal's ability to perform the working memory component of the T-maze task. Performance of the reference memory component was initially impaired following AF64A; however, these animals recovered their ability to perform this task within 4 days of postoperative testing. The AF64A group did develop a pattern of responding which was inconsistent with the cognitive demands of the task. They consistently chose one particular goal location regardless of their initial start location. This pattern of responding gradually developed and persisted throughout the postoperative testing period. For many animals the preferred location varied from day to day but typically remained constant within any daily session. Similar positional preferences have also been reported for animals with
damage to the hippocampus 15 and septum 3°. The findings of this study are consistent with recent reports that ibotenic acid-induced lesions of the medial septum impair working, but not reference, memory processes 11'12'16. Since the medial septum consists of neurons which are known to contain G A B A 17, substance p33 and perhaps other neuropeptides 38, as well as ACh 2'25, it is difficult to ascribe these behavioral deficits to the disruption of a particular neurotransmitter system. The data presented here, however, strongly suggest that working memory processes are dependent upon an intact cholinergic innervation of the hippocampus. In contrast to the consistency of the data supporting this hypothesis, the effects of nucleus basalis damage on working and reference memory processes are inconsistent. Some studies report that ibotenic acid lesions of the nucleus basalis disrupt working, but not reference, memory processes 11'12'16, while other reports have found quite the opposite behavioral dissociations 27. More work is clearly needed to delineate the behavioral properties of different forebrain cholinergic systems. Evidence of behavioral recovery has been reported in some 1,11, but not all studies 27, following ibotenic acid lesions of the medial septum or nucleus basalis. We have observed no indication that AF64A-treated animals recover their ability to perform on the basis of trial-specific or working memory 3'4. The present data indicate that these animals may 'adopt alternate strategies' of performing the task which maximize their efficiency in the absence of a working memory system. Such a dependence upon alternate strategies might actually prevent recovery of normal behavior. Previously, we have shown that rats pretrained to perform a radial-arm maze task with a 2-h delay imposed between the fourth and fifth arm choices are significantly impaired following injection of AF64A (ref. 3). AF64A-treated rats are unable to perform this task at even extremely short delay intervals (30 s). Although working memory is impaired in these animals, they are able to learn a spatial discrimination (reference memory) task in a T-maze as rapidly as controls4. Thus, a variety of reports support the general hypothesis that insult to hippocampal cholinergic processes impairs the animal's ability to perform on the basis of working memory. However, their ability to perform tasks which depend solely on
20 reference m e m o r y regardless of whether they are learned prior to or following cholinergic insult, remains intact. It must be noted, however, that in the present study rats were performing the reference task with greater than 90% accuracy for the last 50 trials prior to surgery while a similar degree of proficiency was obtained for the working m e m o r y task only during the last block of 15 trials. Studies are now in progress to d e t e r m i n e whether these quantitative differences in the rate of acquisition render the different components of the task more susceptible to disruption by A F 6 4 A . A basic assumption of this study is that A F 6 4 A only decreased cholinergic parameters. While we did not assess the integrity of other transmitter systems in the present study recent work supports the longterm cholinotoxicity of our A F 6 4 A regimen (3 nmol/ side, i.c.v.). This dose of A F 6 4 A significantly decreased C h A T activity in the hippocampus but did not alter the activity of this enzyme in the frontal or parietal cortex, striatum, cingulate gyrus or the amygdala. F u r t h e r m o r e , the regional concentrations of d o p a m i n e , serotonin, norepinephrine, and their metabolites were not affected by this dose of A F 6 4 A (ref. 4). Therefore, the dose of A F 6 4 A used in the present study does produce a persistent and selective
REFERENCES 1 Bartus, R.T., Flicker, C., Dean, R.L., Pontecorvo, M., Figueiredo, J.C. and Fisher, S.K., Selective memory loss following nucleus basalis lesions: long term behavioral recovery despite persistent cholinergic deficiencies. Pharmacol. Biochem. Behav., 23 (1985) 125-135. 2 Butcher, L.L. and Woolf, N.J., Histochemical distribution of acetylcholinesterase in the central nervous system: clues to the localization of cholinergic neurons. In A. Bj6rklund, T. H6kfelt and M.J. Kuhar (Eds.), Handbook of Chemical Neuroanatomy, Vol. 3, Elsevier/North Holland, Amsterdam, 1984, pp. 1-50. 3 Chrobak, J.J., Hanin, I. and Walsh, T.J., AF64A impairs working memory in radial-arm and T-maze tasks, Soc. Neurosci. Abstr., 11 (1985) 535. 4 Chrobak, J.J., Hanin, I., Sehmechel, D.E. and Walsh, T.J., AF64A-induced working memory impairments: behavioral, neurochemieal, and histological correlates, Soc. Neurosei. Abstr., 12 (1986) 716. 5 Coyle, J.T., Price, D.L. and DeLong, M.R., Alzheimer's disease: a disorder of cortical cholinergic innervation, Science, 219 (1983) 1184-1190. 6 Drachman, D.A. and Sahakian, B.J., Memory, aging, and pharmacosystems. In D.G. Stein (Ed.), The Psychobiology
decrease in presynaptic cholinergic markers in the hippocampus. H o w e v e r , a recent report indicates that both noradrenergic and serotonergic systems might be transiently altered at short intervals following A F 6 4 A (ref. 13). The potential relationship of these changes to the behavioral p h e n o m e n o n rep o r t e d here are under investigation. The data presented here d e m o n s t r a t e that intraventricular injection of A F 6 4 A produces a selective decrease in C h A T activity in the h i p p o c a m p u s which is associated with an impairment of working memory. A F 6 4 A appears to be a useful tool to further investigate the contribution of cholinergic systems to cognitive function. F u r t h e r m o r e , these efforts should help to identify effective strategies for reversing the memory impairments associated with pathologies of the cholinergic system.
ACKNOWLEDGEMENTS This work was s u p p o r t e d in part by Biomedical Research Support Grant (PHS R R 07058) to T.J.W. and M H 34893 to I.H. A portion of this research will be submitted as part of a doctoral dissertation at the University of North Carolina at Chapel Hill by J.J.C.
of Aging: Problems and Perspectives, Elsevier/North Hol-
land, Amsterdam, 1980, pp. 348-368. 7 DuBois, B., Mayo, W., Agid, Y., Le Moal, M. and Simon, H., Profound disturbances of spontaneous and learned behaviors following lesions of the nucleus basalis magnocellularis in the rat, Brain Research, 338 (1985) 249-258. 8 Fisher, A., Mantione, C.R., Abraham, D.J. and Hanin, I., Long-term central cholinergic hypofunction induced in mice by ethylcholine aziridinium ion (AF64A) in vivo, J. Pharmacol. Exp. Ther., 222 (1982) 140-145. 9 Fonnum, F., Radiochemical microassays for the determination of choline acetyltransferase and acetylcholinesterase activities, Biochem. J.. 115 (1975) 465-472. 10 Freeman, J.J. and Jenden, D.J., The source of choline for acetylcholine synthesis in brain, Life Sci., 19 (1976) 949-962. ll Hepler, D.J,, Olton, D.S., Wenk, G.L. and Coyle, J.T., Lesions in nucleus basalis magnocellularis and medial septal area of rats produce qualitatively similar memory impairments, J. Neurosci., 5 (1985) 866-873. 12 Hepler, D.J., Wenk, G.L., Cribbs, B.L., Olton, D.S. and Coyle, J.T., Memory impairments following basal forebrain lesions, Brain Research, 346 (1985) 8-14. 13 Hortnagl, H., Potter, P.E. and Hanin, I., Activation of serotonergic systems in response to decreased cholinergic
21 function following ethylcholine mustard aziridinium ion (AF64A) treatment, Fed. Proc. Fed. Am. Soc. Exp. Biol., 45 (1986) 923. 14 Kirk, R E . , Experimental Design: Procedures for the Behavioral Sciences, 2rid Edn., Brooks/Cole, Montery, CA. 1982. 15 Kimble, D.P., Choice behavior in rats with hippocampai lesions. In R.L. Isaacson and K.H. Pribram (Eds.), The Hippocarnpus, Vol. 2, Plenum, New York, 1975, pp. 309-326. 16 Knowlton, B.J., Wenk, G.L., Olton, D.S. and Coyle, J.T., Basal forebrain lesions produce a dissociation of trial-dependent and trial-independent memory processes, Brain Research, 345 (1985) 315-321. 17 Kohler, C., Chan-Palay, V. and Wu, J.-Y., Septal neurons containing glutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain, Anat. Embryol., 169 (1984) 41-44. 18 Leventer, S., McKeag, D., Clancy, M., Wulfert, E. and Hanin, I., Intracerebroventricular administration of ethylcholine mustard aziridinium ion (AF64A) reduces release of acetylcholine from rat hippocampal slices, Neuropharmacology, 24 (1985) 453-459. 19 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951)265-275. 20 Mann, J.J., Stanley, M., Neophytidcs, A., DeLeon, M.J., Ferris, S.H. and Gershon, S., Central amine metabolism in Alzheimer's disease: in vivo relationship to cognitive deficit, Neurobiol. Aging, 2 (1981) 57-60. 21 Mantione, C.R., Fisher, A. and Hanin, I., The AF64Atreated mouse: possible model for central cholinergic hypofunction, Science, 213 (1981) 579-580. 22 Mantione, C.R., Zigmond, M.J., Fisher, A. and Hanin, I., Selective presynaptic cholinergic neurotoxicity following intrahippocampal AF64A injection in rats, J. Neurochem., 41 (1983) 251-255. 23 Mantione, C.R., Fisher, A. and Hanin, I., Possible mechanisms involved in the presynaptic cholinotoxicity due to ethylcholine aziridinium (AF64A) in vivo, Life Sci., 35 (1984) 33-41. 24 McCaman, R.E. and Hunt, J.M., Microdetermination of choline acetylase in nervous tissue, J. Neurochem., 12 (1965) 253-259. 25 Mesulam, M.-M., Mufson, E.J., Wainer, B.H. and Levey, A.I., Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Chl-Ch6), Neuroscience, 10 (1983) 1185-1201. 26 Mishkin, M., A memory system in the monkey, Philos. Trans. R. Soc. London Ser. B, 298 (1982) 85-95.
27 Murray, C.L. and Fibiger, H.C., Learning and memory deficits after lesions of the nucleus basalis magnocellularis: reversal by physostigmine, Neuroscience, 19 (1985) 1025-1032. 28 Olton, D.S. and Papas, B.C., Spatial memory and hippocampal function, Neuropsychologia, 17 (1979) 669-682. 29 Perry, E.K., Tomlinson, B.E., Blessed, G., Bergmann, P.H. and Perry, R.H., Correlation of cholinergic abnormalities with senile plaques and mental test scores in senile dementia, Br. Med. J., 2 (1978) 1457-1459. 30 Stanton, M.E., Thomas, G.J. and Brito, G.N.O., Posterodorsal septal lesions impair performance on both shift and stay working memory tasks, Behav. Neurosci., 98 (1984) 405-415. 31 Squire, L.R., The hippocampus and the neuropsychology of memory. In W. Seifert (Ed.), Neurobiology of the Hippocarnpus, Academic, New York, 1983, pp. 491-511. 32 Vickroy, T.W., Watson, M., Leventer, S.M., Roeske, W.R., Hanin, I. and Yamamura, H.I., Regional differences in ethylcholine mustard aziridinium ion (AF64A)-induced deficits in presynaptic cholinergic markers for the rat central nervous system, J. Pharmacol. Exp. Ther., 235 (1986) 577-582. 33 Vincent, S.R. and McGeer, E.G., A substance P projection to the hippocampus, Brain Research, 215 (1981) 349-351. 34 Walsh, T,J., Tilson, H.A., DeHaven, D.L., Mailman, R.B., Fisher, A. and Hanin, I., AF64A, a cholinergic neurotoxin, selectively depletes acetylcholine in hippocampus and cortex and produces long-term passive avoidance and radial-arm maze deficits in the rat, Brain Research, 321 (1984) 91-102. 35 Walsh, T.J. and Hanin, I., A review of the behavioral effects of AF64A, a cholinergic neurotoxin. In A. Fisher, C. Lachman and I. Hanin (Eds.), Alzheimer's and Parkinson's Disease: Therapeutic and Basic Research Strategies, Plenum, New York, 1986, pp. 461-468. 36 Whitehouse, P.J., Price, D.L., Struble, R.G., Clark, A.W., Coyle, J.T. and DeLong, M.R., Alzheimer's disease and senile dementia: loss of neurons in the basal forebrain, Science, 215 (1982) 1237-1239. 37 Wilcock, G.K., Esiri, M.M., Bowen, D.M. and Smith, C.C.T., Alzheimer's disease: correlation of cortical choline acetyltransferase activity with severity of dementia and histological abnormalities, J. Neurol. Sci., 57 (1982)407-417. 38 Woodhams, P.L., Roberts, G.W., Polak, J.M. and Crow, T.J., Distribution of neuropeptides in the limbic system of the rat: the bed nucleus of the stria terminalis, septum and preoptic area, Neuroscience, 8 (1983) 677-703.
15
BRE 12640
AF64A (ethylcholine aziridinium ion), a cholinergic neurotoxin, selectively impairs working memory in a multiple component T-maze task James J. C h r o b a k 1, Israel H a n i n 2 and T h o m a s J. Walsh 1 1Department of Psychology, Rutgers University, New Brunswick, NJ08903 (U.S.A.) and 2Departmentof Pharmacology and Experimental Therapeutics, Loyola UniversitySchool of Medicine, Maywood, IL 60153 (U.S.A.) (Accepted 4 November 1986)
Key words: AF64A; Cholinergic neurotoxin; Learning and memory; Working memory; Acetylcholine; Hippocampus
The present study examined the nature of the cognitive deficits associated with a selective decrease of cholinergic activity in the hippocampus. Male Fischer rats were trained to perform a multiple component T-maze task which simultaneously assessed their ability to perform on the basis of trial-specific information (working memory) and trial-independent information (reference memory). Following 125 acquisition trials rats were bilaterally injected with AF64A (3 nmol/side) or artificial CSF into the lateral ventricles and allowed 14 days to recover before behavioral testing resumed. The controls rapidly returned to their preoperative level of performance on both components of the maze task. AF64A-treated animals were transiently impaired on the reference memory task. Their performance rapidly improved and they were performing at preoperative levels within 4 days of testing. In contrast, these animals exhibited a marked and long-lasting impairment in their performance of the working memory component. After behavioral testing was completed, neurochemical analysis revealed that AF64A produced a significant decrease in choline acetyltransferase (CHAT) activity in the hippocampus (43%) 42 days following surgery. This dosing regimen produced no alterations of striatal or cortical ChAT activity. These data suggest that alterations of hippocampal cholinergic activity severely impair an animal's ability to perform working memory tasks. INTRODUCTION Central cholinergic systems appear to be a critical neurobiological substrate for learning and m e m o r y processes. Pharmacological studies in both humans and laboratory animals have shown that cholinergic agonists improve, while antagonists impair, cognition 1'6. These studies provided evidence that memory processes are dependent upon the integrity of cholinergic systems. Analysis of behavioral changes induced by damage to the cholinergic nuclei in the basal forebrain 7'n'12'16'27 and the observation that these same nuclei degenerate in patients with various forms of dementia further support the involvement of forebrain cholinergic systems in cognitive functionS'361 In senile dementia of the Alzheimer type (SDAT) the decreases in cholinergic parameters have been shown to correlate with the severity of the
cognitive impairment 29'37. While other neurotransmitters or peptides might be altered in subpopulations of S D A T patients or late in the course of the disease, these alterations have not been shown to be related to the prevailing cognitive impairments 2°. A further understanding of the behavioral properties of cholinergic systems should provide insight into the neurobiology of learning and m e m o r y and could also provide information about the pathophysiology of S D A T and potential approaches to its treatment. Ethylcholine aziridinium ion (AF64A) is a neurotoxic analog of choline which disrupts the high affinity transport of choline ( H A C h T ) , the rate limiting step in the synthesis of acetylcholine (ACh) l°,2t. Injection of A F 6 4 A into the lateral ventricles produces a persistent reduction in presynaptic cholinergic
Correspondence: T.J. Walsh, Rutgers University, Department of Psychology, Busch Campus, New Brunswick, NJ 08903, U.S.A. 0006-8993/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)
16 markers including HAChT, choline acetyltransferase (CHAT) activity, ACh concentrations, and K+-in duced release of ACh from hippocampai slices Is' 21-23,32,34 This loss of cholinergic markers is not accompanied by alterations in the regional content of norepinephrine, dopamine, serotonin, their metabolites, or choline 4'34. We recently reported that injection of AF64A into the lateral ventricles or the dorsal hippocampus impairs learning and memory processes 34'3-~. For example, administration of relatively large doses of AF64A into the lateral ventricles reduced ACh content 45-60% in the hippocampus (7.5 or 15 nmol/ side) and frontal cortex (15 nmol/side). This neurochemical change was accompanied by a significant impairment in the retention of a passive avoidance response and the acquisition of an 8-arm radial maze task 34'35. Doses of AF64A lower than 7.5 nmol/side appear to produce a selective loss of hippocampal cholinergic markers 4"34. The studies presented here attempt to further characterize the nature of the cognitive impairments associated with this selective decrease of cholinergic activity in the hippocampus. We were particularly interested in whether loss of hippocampal cholinergic tone would selectively impair memory for trial-specific information (working memory) but not for trial-independent information (reference memory). Working memory involves flexible stimulus-response associations whose value depends on recently experienced events, whereas reference memory involves fixed stimulus-response associations whose values are not altered over consecutive trials 28. This dichotomy in memory systems is similar to the pattern of impaired and preserved memory processes observed in a variety of human, primate and rodent amnesic syndromes which involve hippocampal damage 26'2s'31. To further examine the potential dissociation of cognitive impairments produced by AF64A we developed a T-maze task which can simultaneously assess the animals ability to perform, based upon trialindependent and trial-specific information within the same trial. In this task the rat is placed into one of two start/goal locations on either arm of the maze. It is trained to first run down the stem of the maze, and then to return to the alternate start/goal location on each trial. The different aspects of this task can be characterized as (a) a trial-independent (reference
memory) component which involves an invariant response of proceeding down the stem of the maze regardless of the start location, and (b) a trial-specific (working memory) component of returning to the alternate start/goal location. Successful performance of the latter requires the animal to use some representation or memory of the initial start/goal location for any trial. MATERIALS AND METHODS
Subjects Male Fischer 344 rats (Charles River Breeders, Wilmington, DE) approximately 90 days old and weighing between 250 and 300 g were used in the following experiment(s). The animals were individually housed and maintained in a colony room with a 12 h light/dark cycle. Water was freely available at all times. Five days prior to acquisition and reacquisition trials animals were reduced to 85% of their free-feeding body weight by limiting their daily ration of food. This level was maintained throughout testing except for 7 days immediately prior and subsequent to surgery. Over the course of the experiment the rats were allowed to gain 5 g of weight each week.
Apparatus A white plexiglass T-maze with white vinyl flooring was utilized in this experiment. The stem of the maze was 71 cm long and each arm extended 36 cm from the stem. Arms were 14 cm high and 14 cm wide. A removable plexiglass barrier could be positioned at the end of each arm providing a 15 cm start compartment. The maze was elevated approximately 46 cm above the floor in the center of a small room which contained a rich array of visual stimuli.
Preparation of ethylcholine mustard aziridinium ion (AF64A) AF64A was prepared from 10 mM acetylethylcholine mustard hydrochloride by dissolving the starting material in distilled water, adjusting the pH to 11.5-11.7 with concentrated NaOH, and then vigorously stirring the solution for 20 min 8. The solution was next diluted with artificial CSF, and finally was titrated to a pH of 7.3-7.4 using HC1. Solutions of AF64A were always prepared immediately prior to use and kept on ice at all times.
17
Intracerebroventricular injection of AF64A Six days following the preoperative behavioral tests animals were anesthetized with sodium pentobarbital (45 mg/kg) and positioned in a stereotaxic instrument. Blunt ear bars were used to minimize damage to the auditory meatus. A sagittal incision was made in the scalp, and two holes were drilled through the skull for placement of the injection cannula into the lateral ventricles. Rats were bilaterally infused with isotonic CSF or 3 nmol (6 nmol/brain) of AF64A. Solutions were slowly infused (0.125/A/min) into each site using a 1 ~1 Hamilton syringe positioned in a microinjector unit mounted on the stereotaxic frame. A total volume of 3.0/tl was delivered per side. To promote the local diffusion of the perfusate the injection cannula was left in place for a period of 3 min following injection.
T-maze acquisition Animals were given 5 days of habituation trials in which they were free to explore the maze for a 5-rain period. During these trials food pellets (94 mg chocolate-flavored) were scattered throughout the maze. Following the habituation period each rat was given 5 trials a day for 5 weeks, for a total of 125 trials. There was a variable inter-trial interval of 30-90 s. Each trial began by placing the rat into either start/goal location for a 10-s period. The guillotine door was then T-MAZE TASK
I CORRECT I
ERROR- - - - - - ~
-
ol START RIGHT
LEFT
• TRAVERSE STEM
- RETU RN TO ALTERNATE S/G BOX (REFERENCE ERROR--) ( WORKINGERROR - - - )
TRIAL
FT
.I Io
START LEFT
=TRAVERSE STEM
- RETURN TO ALTERNATE S/G BOX
RANDOMLY ALTERNATED TRIALS Fig. l. A schematic representation of the T-maze task is presented. The rat is initially placed into the right (right trial) or left (left trial) start compartment. The reference memory component of the task consists of turning into and transversing the stem of the maze. The working memory component consists of running to the alternate start/goal box on any trial.
manually opened and the animal was allowed 60 s to complete the trial. The animal was rewarded for turning into and traversing the stem (see Fig. 1) and for subsequently returning to the alternate start/goal location. If the animal crossed directly to the alternate start/goal location without transversing the stem it was promptly removed and the trial ended. If the animal returned to the original start location it was removed and the trial ended. At the end of each trial the animal was returned to a cage located in the maze room. The choice of start location varied quasi-randomly, with the provision that no more than 3 trials during a given day start from the same location. Following the 125 acquisition trials animals were performing with 85-95% accuracy on both the working and reference memory components of the task. Animals were then randomly assigned to either the AF64A or control group. They were allowed 6 days to recover body weight and were injected with AF64A or CSF according to the surgical protocol described above. The number of working and reference memory errors were recorded each day for the session of 5 trials. The data were grouped by 3-day blocks and analyzed according to the percentage correct per block.
Neurochemical analysis Rats were sacrificed 42 days following surgery, after behavioral testing was completed. Their brains were rapidly removed and dissected on ice into the following regions: hippocampus, frontal cortex and striatum. ChAT was measured according to the radiometric assay of McCaman and Hunt 24 as modified by Fonnum 9. Protein content of tissue was determined according to Lowry et al. t9. Since previous studies have shown that this dose of AF64A does not alter indices of catechol- or indoleamine function 3.4 we only examined cholinergic parameters in the present study. RESULTS
Animals injected with CSF rapidly recovered from surgery and appeared normal within 2 days. Rats in the AF64A group displayed a transient syndrome consisting of diarrhea, piloerection, aphagia, adipsia, and hyperreactivity. Following this initial postoperative period, animals rapidly recovered and their
18 TABLE I
REFERENCE MEMORY ACQUISITION
Effects of AF64A (3 nmol/side, i.c.v.) on choline acetyltransferase activity
REACQUISITION
-__-- _---_
,oo
r,~.,._._;U~._~___ ~
,oo
Values are nmol A C h formed/mg protein/h. a= 80
CSF (n = 9) A F 6 4 A (n = 12)
Hippocampus
Frontal cortex
Striatum
123 + 3 71 + 7*
70 _+ 3 84 + 6
197 + 13 204 + 15
(2
80
70
Z C~ 60 bJ o- 50
* P < 0.01 vs CSF, Student's t-test.
/'
7O t CSF (N;9) IO--'OAF(N 12) I
~/
60
~
50
i
i
i
i
i
15
30
45
60
75
ii
i
i
0 105 120
l
l
I
I
15
50
45
60
.J
75
TRIALS
appearance and overt behavior was indistinguishable from that of controls. We observed no mortality with the 3 nmol dose of AF64A. ChAT activity was significantly reduced (43%) in the hippocampus of AF64A-treated animals (see Table I). No significant alterations of cortical or striatal ChAT activity were observed following AF64A although ChAT activity in the frontal cortex was 20% higher in this group. Prior to surgery both groups were performing with comparable accuracy on the reference and working memory components of the task. Both groups of rats were performing with greater than 90% accuracy on the reference memory task after 75 training trials. However, this level of performance on the working memory component was not achieved until the last 15 trials. Analysis of preoperative performance using a twoway repeated A N O V A revealed no significant group effects (P > 0.1) or group by trial interactions (P > 0.1) for either working or reference memory. However, there was a significant trial effect for both components of the task (F7,13 3 -- 46.3-107.9, P < 0.01). These analyses indicate that the performance of both groups improved at the same rate over time and they were performing with comparable accuracy prior to surgery (see Figs. 2 and 3). The performance of the CSF-injected group following surgery was not significantly different from their preoperative levels of performance on either the reference or working memory task. In contrast, the AF64A group exhibited a short-term reversible impairment of reference memory and a severe and persistent impairment of working memory. The behavior of AF64A-treated animals during testing suggested no alteration of motivation or motoric ability which might have interfered with task
Fig. 2. The effects of A F 6 4 A on the acquisition and the reacquisition of the reference m e m o r y c o m p o n e n t of the T-maze task described in Fig. 1. This figure presents the percentage of correct responses summarized over each 3-day block of sessions, 5 trials per day. Data are presented as m e a n percent correct + S.E.M. Following the acquisition trials rats were injected with A F 6 4 A (3 nmol/side, i.c.v.), allowed 14 days to recover from surgery, and were then re-exposed to the behavioral task (reacquisition), * P < 0.01 vs CSF controls, Spjotvall-Stoline t-test.
performance. These animals readily navigated the Tmaze, consumed the pellets, and completed each trial in the same amount of time as the CSF-injected controls. Furthermore, we have observed that this dose of AF64A does not alter locomotor activity4. Reacquisition of the reference memory task was significantly slower in the AF64A group. A two-way
WORKING M E M O R Y ACQUISITION
REACOUISlTION
I00
I00
90
90
8O
8O
cc 70 nO t.~ 60 pZ
70
I-
~ Q.
i
60' /
5o 4O 30
/
5O 4O
/
~ i 15
i,/
~° l 30
i 45
i 60
~ 75
i 90
i i 105 120
15 30
45
60
75
TRIALS
Fig. 3. The effects of A F 6 4 A on the acquisition and reacquisition of the working m e m o r y c o m p o n e n t of the T-maze task described in Fig. 1. This figure presents the percentage of correct responses summarized over each 3-day block of sessions, 5 trials per day. Data are presented as mean percent correct _+ S.E.M. * P < 0.01 vs CSF controls, Spjotvall-Stoline t-test.
19 repeated ANOVA revealed significant treatment (El,19 = 12.2, P < 0.01) and trials effects (F4.76 = 13.2, P < 0.01), as well as a significant interaction (F4,76 = 6.8, P '~ 0.01). Post-hoc comparisons utilizing the Spjotvall-Stoline t-test 14 revealed that the performance of the AF64A group was significantly different from that of controls only during the first block of reacquisition trials (P < 0.01). Animals treated with AF64A were performing as well as the controls within 4 days of postoperative testing. A two-way repeated ANOVA demonstrated that the AF64A group was significantly impaired in their performance of the working memory component (F1,19 = 13.9, P < 0.01). The analysis showed a significant trials effect (F4,76 = 13.4, P < 0.01) but no treatment by trials interaction (F4,76 = 1.7, P > 0.10). The performance of the control group on the working memory task steadily improved during reacquisition trials (see Fig. 3), while the AF64A-treated animals remained impaired despite a small improvement observed following the first block of trials. The performance of AF64A-treated rats on the working memory component did not improve beyond the first 3 trials block even when they were optimally performing the reference memory part of the task. DISCUSSION
Bilateral injection of a low dose of AF64A into the cerebroventricles produced a significant decrease in hippocampal, but not cortical or striatal, ChAT activity. This biochemical deficit was associated with a significant impairment in the animal's ability to perform the working memory component of the T-maze task. Performance of the reference memory component was initially impaired following AF64A; however, these animals recovered their ability to perform this task within 4 days of postoperative testing. The AF64A group did develop a pattern of responding which was inconsistent with the cognitive demands of the task. They consistently chose one particular goal location regardless of their initial start location. This pattern of responding gradually developed and persisted throughout the postoperative testing period. For many animals the preferred location varied from day to day but typically remained constant within any daily session. Similar positional preferences have also been reported for animals with
damage to the hippocampus 15 and septum 3°. The findings of this study are consistent with recent reports that ibotenic acid-induced lesions of the medial septum impair working, but not reference, memory processes 11'12'16. Since the medial septum consists of neurons which are known to contain G A B A 17, substance p33 and perhaps other neuropeptides 38, as well as ACh 2'25, it is difficult to ascribe these behavioral deficits to the disruption of a particular neurotransmitter system. The data presented here, however, strongly suggest that working memory processes are dependent upon an intact cholinergic innervation of the hippocampus. In contrast to the consistency of the data supporting this hypothesis, the effects of nucleus basalis damage on working and reference memory processes are inconsistent. Some studies report that ibotenic acid lesions of the nucleus basalis disrupt working, but not reference, memory processes 11'12'16, while other reports have found quite the opposite behavioral dissociations 27. More work is clearly needed to delineate the behavioral properties of different forebrain cholinergic systems. Evidence of behavioral recovery has been reported in some 1,11, but not all studies 27, following ibotenic acid lesions of the medial septum or nucleus basalis. We have observed no indication that AF64A-treated animals recover their ability to perform on the basis of trial-specific or working memory 3'4. The present data indicate that these animals may 'adopt alternate strategies' of performing the task which maximize their efficiency in the absence of a working memory system. Such a dependence upon alternate strategies might actually prevent recovery of normal behavior. Previously, we have shown that rats pretrained to perform a radial-arm maze task with a 2-h delay imposed between the fourth and fifth arm choices are significantly impaired following injection of AF64A (ref. 3). AF64A-treated rats are unable to perform this task at even extremely short delay intervals (30 s). Although working memory is impaired in these animals, they are able to learn a spatial discrimination (reference memory) task in a T-maze as rapidly as controls4. Thus, a variety of reports support the general hypothesis that insult to hippocampal cholinergic processes impairs the animal's ability to perform on the basis of working memory. However, their ability to perform tasks which depend solely on
20 reference m e m o r y regardless of whether they are learned prior to or following cholinergic insult, remains intact. It must be noted, however, that in the present study rats were performing the reference task with greater than 90% accuracy for the last 50 trials prior to surgery while a similar degree of proficiency was obtained for the working m e m o r y task only during the last block of 15 trials. Studies are now in progress to d e t e r m i n e whether these quantitative differences in the rate of acquisition render the different components of the task more susceptible to disruption by A F 6 4 A . A basic assumption of this study is that A F 6 4 A only decreased cholinergic parameters. While we did not assess the integrity of other transmitter systems in the present study recent work supports the longterm cholinotoxicity of our A F 6 4 A regimen (3 nmol/ side, i.c.v.). This dose of A F 6 4 A significantly decreased C h A T activity in the hippocampus but did not alter the activity of this enzyme in the frontal or parietal cortex, striatum, cingulate gyrus or the amygdala. F u r t h e r m o r e , the regional concentrations of d o p a m i n e , serotonin, norepinephrine, and their metabolites were not affected by this dose of A F 6 4 A (ref. 4). Therefore, the dose of A F 6 4 A used in the present study does produce a persistent and selective
REFERENCES 1 Bartus, R.T., Flicker, C., Dean, R.L., Pontecorvo, M., Figueiredo, J.C. and Fisher, S.K., Selective memory loss following nucleus basalis lesions: long term behavioral recovery despite persistent cholinergic deficiencies. Pharmacol. Biochem. Behav., 23 (1985) 125-135. 2 Butcher, L.L. and Woolf, N.J., Histochemical distribution of acetylcholinesterase in the central nervous system: clues to the localization of cholinergic neurons. In A. Bj6rklund, T. H6kfelt and M.J. Kuhar (Eds.), Handbook of Chemical Neuroanatomy, Vol. 3, Elsevier/North Holland, Amsterdam, 1984, pp. 1-50. 3 Chrobak, J.J., Hanin, I. and Walsh, T.J., AF64A impairs working memory in radial-arm and T-maze tasks, Soc. Neurosci. Abstr., 11 (1985) 535. 4 Chrobak, J.J., Hanin, I., Sehmechel, D.E. and Walsh, T.J., AF64A-induced working memory impairments: behavioral, neurochemieal, and histological correlates, Soc. Neurosei. Abstr., 12 (1986) 716. 5 Coyle, J.T., Price, D.L. and DeLong, M.R., Alzheimer's disease: a disorder of cortical cholinergic innervation, Science, 219 (1983) 1184-1190. 6 Drachman, D.A. and Sahakian, B.J., Memory, aging, and pharmacosystems. In D.G. Stein (Ed.), The Psychobiology
decrease in presynaptic cholinergic markers in the hippocampus. H o w e v e r , a recent report indicates that both noradrenergic and serotonergic systems might be transiently altered at short intervals following A F 6 4 A (ref. 13). The potential relationship of these changes to the behavioral p h e n o m e n o n rep o r t e d here are under investigation. The data presented here d e m o n s t r a t e that intraventricular injection of A F 6 4 A produces a selective decrease in C h A T activity in the h i p p o c a m p u s which is associated with an impairment of working memory. A F 6 4 A appears to be a useful tool to further investigate the contribution of cholinergic systems to cognitive function. F u r t h e r m o r e , these efforts should help to identify effective strategies for reversing the memory impairments associated with pathologies of the cholinergic system.
ACKNOWLEDGEMENTS This work was s u p p o r t e d in part by Biomedical Research Support Grant (PHS R R 07058) to T.J.W. and M H 34893 to I.H. A portion of this research will be submitted as part of a doctoral dissertation at the University of North Carolina at Chapel Hill by J.J.C.
of Aging: Problems and Perspectives, Elsevier/North Hol-
land, Amsterdam, 1980, pp. 348-368. 7 DuBois, B., Mayo, W., Agid, Y., Le Moal, M. and Simon, H., Profound disturbances of spontaneous and learned behaviors following lesions of the nucleus basalis magnocellularis in the rat, Brain Research, 338 (1985) 249-258. 8 Fisher, A., Mantione, C.R., Abraham, D.J. and Hanin, I., Long-term central cholinergic hypofunction induced in mice by ethylcholine aziridinium ion (AF64A) in vivo, J. Pharmacol. Exp. Ther., 222 (1982) 140-145. 9 Fonnum, F., Radiochemical microassays for the determination of choline acetyltransferase and acetylcholinesterase activities, Biochem. J.. 115 (1975) 465-472. 10 Freeman, J.J. and Jenden, D.J., The source of choline for acetylcholine synthesis in brain, Life Sci., 19 (1976) 949-962. ll Hepler, D.J,, Olton, D.S., Wenk, G.L. and Coyle, J.T., Lesions in nucleus basalis magnocellularis and medial septal area of rats produce qualitatively similar memory impairments, J. Neurosci., 5 (1985) 866-873. 12 Hepler, D.J., Wenk, G.L., Cribbs, B.L., Olton, D.S. and Coyle, J.T., Memory impairments following basal forebrain lesions, Brain Research, 346 (1985) 8-14. 13 Hortnagl, H., Potter, P.E. and Hanin, I., Activation of serotonergic systems in response to decreased cholinergic
21 function following ethylcholine mustard aziridinium ion (AF64A) treatment, Fed. Proc. Fed. Am. Soc. Exp. Biol., 45 (1986) 923. 14 Kirk, R E . , Experimental Design: Procedures for the Behavioral Sciences, 2rid Edn., Brooks/Cole, Montery, CA. 1982. 15 Kimble, D.P., Choice behavior in rats with hippocampai lesions. In R.L. Isaacson and K.H. Pribram (Eds.), The Hippocarnpus, Vol. 2, Plenum, New York, 1975, pp. 309-326. 16 Knowlton, B.J., Wenk, G.L., Olton, D.S. and Coyle, J.T., Basal forebrain lesions produce a dissociation of trial-dependent and trial-independent memory processes, Brain Research, 345 (1985) 315-321. 17 Kohler, C., Chan-Palay, V. and Wu, J.-Y., Septal neurons containing glutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain, Anat. Embryol., 169 (1984) 41-44. 18 Leventer, S., McKeag, D., Clancy, M., Wulfert, E. and Hanin, I., Intracerebroventricular administration of ethylcholine mustard aziridinium ion (AF64A) reduces release of acetylcholine from rat hippocampal slices, Neuropharmacology, 24 (1985) 453-459. 19 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951)265-275. 20 Mann, J.J., Stanley, M., Neophytidcs, A., DeLeon, M.J., Ferris, S.H. and Gershon, S., Central amine metabolism in Alzheimer's disease: in vivo relationship to cognitive deficit, Neurobiol. Aging, 2 (1981) 57-60. 21 Mantione, C.R., Fisher, A. and Hanin, I., The AF64Atreated mouse: possible model for central cholinergic hypofunction, Science, 213 (1981) 579-580. 22 Mantione, C.R., Zigmond, M.J., Fisher, A. and Hanin, I., Selective presynaptic cholinergic neurotoxicity following intrahippocampal AF64A injection in rats, J. Neurochem., 41 (1983) 251-255. 23 Mantione, C.R., Fisher, A. and Hanin, I., Possible mechanisms involved in the presynaptic cholinotoxicity due to ethylcholine aziridinium (AF64A) in vivo, Life Sci., 35 (1984) 33-41. 24 McCaman, R.E. and Hunt, J.M., Microdetermination of choline acetylase in nervous tissue, J. Neurochem., 12 (1965) 253-259. 25 Mesulam, M.-M., Mufson, E.J., Wainer, B.H. and Levey, A.I., Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Chl-Ch6), Neuroscience, 10 (1983) 1185-1201. 26 Mishkin, M., A memory system in the monkey, Philos. Trans. R. Soc. London Ser. B, 298 (1982) 85-95.
27 Murray, C.L. and Fibiger, H.C., Learning and memory deficits after lesions of the nucleus basalis magnocellularis: reversal by physostigmine, Neuroscience, 19 (1985) 1025-1032. 28 Olton, D.S. and Papas, B.C., Spatial memory and hippocampal function, Neuropsychologia, 17 (1979) 669-682. 29 Perry, E.K., Tomlinson, B.E., Blessed, G., Bergmann, P.H. and Perry, R.H., Correlation of cholinergic abnormalities with senile plaques and mental test scores in senile dementia, Br. Med. J., 2 (1978) 1457-1459. 30 Stanton, M.E., Thomas, G.J. and Brito, G.N.O., Posterodorsal septal lesions impair performance on both shift and stay working memory tasks, Behav. Neurosci., 98 (1984) 405-415. 31 Squire, L.R., The hippocampus and the neuropsychology of memory. In W. Seifert (Ed.), Neurobiology of the Hippocarnpus, Academic, New York, 1983, pp. 491-511. 32 Vickroy, T.W., Watson, M., Leventer, S.M., Roeske, W.R., Hanin, I. and Yamamura, H.I., Regional differences in ethylcholine mustard aziridinium ion (AF64A)-induced deficits in presynaptic cholinergic markers for the rat central nervous system, J. Pharmacol. Exp. Ther., 235 (1986) 577-582. 33 Vincent, S.R. and McGeer, E.G., A substance P projection to the hippocampus, Brain Research, 215 (1981) 349-351. 34 Walsh, T,J., Tilson, H.A., DeHaven, D.L., Mailman, R.B., Fisher, A. and Hanin, I., AF64A, a cholinergic neurotoxin, selectively depletes acetylcholine in hippocampus and cortex and produces long-term passive avoidance and radial-arm maze deficits in the rat, Brain Research, 321 (1984) 91-102. 35 Walsh, T.J. and Hanin, I., A review of the behavioral effects of AF64A, a cholinergic neurotoxin. In A. Fisher, C. Lachman and I. Hanin (Eds.), Alzheimer's and Parkinson's Disease: Therapeutic and Basic Research Strategies, Plenum, New York, 1986, pp. 461-468. 36 Whitehouse, P.J., Price, D.L., Struble, R.G., Clark, A.W., Coyle, J.T. and DeLong, M.R., Alzheimer's disease and senile dementia: loss of neurons in the basal forebrain, Science, 215 (1982) 1237-1239. 37 Wilcock, G.K., Esiri, M.M., Bowen, D.M. and Smith, C.C.T., Alzheimer's disease: correlation of cortical choline acetyltransferase activity with severity of dementia and histological abnormalities, J. Neurol. Sci., 57 (1982)407-417. 38 Woodhams, P.L., Roberts, G.W., Polak, J.M. and Crow, T.J., Distribution of neuropeptides in the limbic system of the rat: the bed nucleus of the stria terminalis, septum and preoptic area, Neuroscience, 8 (1983) 677-703.