Brain Research. 9 Springer-Verlag 1990. Intrahippocampal cholinergic grafts in aged rats compensate impairments in a radial maze and in a place learning task.
Exp Brain Res (1990) 82:641-650
Experimental BrainResearch 9 Springer-Verlag 1990
Intrahippocampal cholinergic grafts in aged rats compensate impairments in a radial maze and in a place learning task F. Schenk, B. Contant, and P. Werffeli Institute of Physiology, School of Medicine, University of Lausanne, CH 1005 Lausanne, Switzerland Received March 9, 1990 / Accepted June 30, 1990
Summary. Age-related cognitive impairments were studied in rats kept in semi-enriched conditions during their whole life, and tested during ontogeny and adult life in various classical spatial tasks. In addition, the effect of intrahippocampal grafts of fetal septal-diagonal band tissue, rich in cholinergic neurons, was studied in some of these subjects. The rats received bilateral cell suspensions when aged 23-24 months. Starting 4 weeks after grafting, they were trained during 5 weeks in an 8-arm maze made of connected plexiglass tunnels. No age-related impairment was detected during the first eight trials, when the maze shape was that of a classical radial maze in which the rats had already been trained when young. The older rats were impaired when the task was made more difficult by rendering two arms parallel to each other. They developed an important neglect of one of the parallel tunnels resulting in a high amount of errors before completion of the task. In addition, the old rats developed a systematic response pattern of visits to adjacent arms in a sequence, which was not observed in the younger subjects. None of these behaviours were observed in the old rats with a septal transplant. Sixteen weeks after grafting, another experiment was conducted in a homing hole board task. Rats were allowed to escape from a large circular arena through one hole out of many, and to reach home via a flexible tube under the table. The escape hole was at a fixed position according to distant room cues, and olfactory cues were made irrelevant by rotating the table between the trials. An additional cue was placed on the escape position. No age-related difference in escape was observed during training. During a probe trial with no hole connected and no proximal cue present, the old untreated rats were less clearly focussed on the training sector than were either the younger or the grafted old subjects. Taken together, these experiments indicate that enriched housing conditions and spatial training during adult life do not protect against all age-related deterioration in spatial ability. However, it might be that the Offprint requests to." F. Schenk (address see above)
considerable improvement observed in the grafted subjects results from an interaction between the graft treatment and the housing conditions.
Key words: Neural transplantation - Acetylcholine Hippocampus - Aging - Spatial memory - Housing condition - Rat
Introduction It is now widely assumed that spatial memory tasks are very sensitive to age-related cognitive impairments in laboratory rats. Senescent subjects show acquisition deficits in place learning tasks (Barnes 1979; Barnes and McNaughton 1985; Gage et al. 1984; Gage and Bj6rklund 1986; Rapp et al. 1987) and in elimination tasks such as the Olton 8-arm maze (Geinisman et al. 1986; Wallace et al. 1980). These deficits are not restricted to the processing of information from distant cues, since old rats are also impaired in learning a sequence of right-left turns in a 14 unit T-maze in which there is little access to distant landmarks, if any (Ingram et al. 1981 ; Spangler et al. 1989). In brief, old rats are less accurate in remembering a place, they often rely upon response strategies (Barnes et al. 1980) and they maintain inappropriate alternation sequences in a 14-unit T-maze (Spangler et al. 1989). Spatial memory reduction is also a representative feature of human age-related disorders, and this has contributed to a wide use of spatial tasks for studying brain mechanisms of senescent memory decline. It has been demonstrated that the poor performance of old rats in spatial tasks was concomitant to fibre degeneration (Greene and Naranjo 1987), to change in synaptic density (Geinisman et al. 1986) and plasticity in the hippocampus (Barnes 1979; Barnes and McNaughton 1985). Moreover, as compared to younger subjects, groups of old rats trained in a place learning task show both a
642 slower acquisition and a less marked concomitant reduction in high affinity choline uptake in the hippocampus (Gallagher and Pelleymounter 1988). In all the cited studies, the extent of individual behavioral impairment was variable, and it was related to a morphological or functional change. In addition to emphasizing the link between spatial learning and hippocampal function, these data raise the question of which factors might prevent age-related changes in brain function. Dietary restriction is associated with prolonged life span in captive rats (Forbes 1980). This treatment also prevents some of the deficits in maze performance shown by old rats (Goodrick 1984; Stewart et al. 1989) and mice (Ingram et al. 1987). However, Beatty et al. (1987) and more recently Bond et al. (1989) have reported that in rats, dietary restriction does not prevent age-related deficits in the radial maze, whereas prolonged training during adult life appears to preserve a good performance in working memory up to 26 months of age (Beatty et al. 1985; Beatty 1988). As admitted by the authors, the preserved skills might be task and environment specific, since these rats were trained on a single radial maze during their whole life. Nevertheless, this finding suggests that the standard "old rat" might not be a purely specific model of age-related deficits. In fact, these animals usually spend their entire life in what can best be described as a prolonged captivity in an impoverished environment, and it is not clear how this may interact with normal aging processes. First, environmental conditions are known to affect brain development and function, e.g. cortical thickness (Diamond 1986) and hippocampal plasticity (Sharp et al. 1987; Petit and Markus 1987). Secondly, since the turnover of hippocampal acetylcholine is affected by solving spatial problems (Gallagher and Pelleymounter 1988), one can hypothesize that this type of activation is severely reduced during prolonged captivity in a stable environment. These experiments have focussed on short lasting reversible changes in brain mechanisms, but it is likely that living in an enriched environment or performing spatial tasks might also affect brain function on a long term basis and therefore play a long lasting role in maintaining optimal brain function. It thus appears important to find out whether noticeable cognitive impairments are also evident in more "normal" old rats. In other words, this asks the question of whether maintaining active cognitive functions might prevent, completely or partially, a decline of cognitive abilities during senescence. The following experiments were designed to dctect impairments in aged rats which have been kept in semienriched conditions during their whole life, and tested during ontogeny and adult life in various spatial tasks, i.e. a homing hole board task when weanlings (Schenk 1989), a place navigation and a radial maze task when adults. Moreover, suspensions of embryonic cholinergic neurones were implanted into the hippocampus of half of these subjects when aged 23-24 months. This treatment is known to compensate age-related deficits in spatial navigation (Gage et al. 1984; Gage and Bj6rklund 1986) and it also corrects some of the effects of the transection of the septohippocampal system (Nilsson et
al. 1987). Nevertheless, cholinergic transplants are also reported as having negative effects. Dalrymple-Alford et al. (1987) have shown that intrahippocampal grafts of cholinergic cells increase the deficits observed in a radial maze following extensive fimbria fornix lesions. Obviously, these negative effects are due to the considerable development of the grafted neurones in the fields of the intact host hippocampus, because of inappropriate injection sites. But the same authors (Kelche et al. 1988) have also reported that enriched housing conditions might interact with a graft treatment to promote a slight recovery in fimbria-lbrnix lesions, suggesting that the beneficial effect of the graft might be enhanced in our old subjects. We have used two different behavioural paradigms, a radial maze and a place learning task, in order to facilitate the detection of possible deterioration induced by the transplant. Two successive test phases were designed for each task, in order to assess the performance of the old rats at a low and high difficulty level. Twentyfive month old rats were first rctraincd on the 8-arm radial maze, then trained in a modified maze in which two adjacent arms had been rendered parallel. This change in maze structure is known to increase task difficulty in young adults by inducing neglect and confusion of the two parallel arms (Schenk et al. 1990). The same rats were trained 2 months later on a simple cued task on the homing hole board. Contextual learning was then assessed in a special trial in which the training position was neither baited nor cued. In the place navigation task, this type of test is known to detect impairments due to incomplete maturation (Schenk 1985), and to be sensitive to age-related deficits (Rapp et al. 1987).
Materials and methods Subjects Old rats. 30 male hooded PVG rats born in our laboratory were maintained by groups of 4 siblings in large glass cages (40 x 40 x 50 cm) provided with a wooden nest box and various wooden objects. All the subjects had been tested during 7 days on a homing hole board task in a cue controlled environment, when aged 3-5 weeks. They had been trained on a radial maze (18 trials in 24 days) and on a simple place navigation task (38 trials in 8 days) when adults (about 6 and 10-12 months of age). Then they remained untested from the age of 12 months. All the subjects were maintained on a restricted food diet adjusted to keep their body weight at about 90% of the normal value during training in the radial maze. This diet was then permanent from the age of 16 months. Youn9 rats. 21 control rats were maintained in the same conditions as the old rats (actually, our standard cages). They had been tested on a homing hole board during ontogeny, and in a radial maze when about 3 months old.
Experimental design
The whole procedure was conducted with two different groups of old rats, 8 months apart. A total of 16 rats aged about 24 months received bilateral septal suspension transplants, 6 rats received similar sham injections and 8 rats were left as unoperated controls.
643 These rats were from 11 different litters with a maximum of 4 siblings. Typically, half of the subjects in each litter were assigned randomly to either the experimental or the control condition. The results of rats dying spontaneously within 4 weeks following the end of an experiment were discarded from the final analysis. Ten of the grafted rats and 10 of the aged control rats survived throughout the two experiments. Two different groups of younger control rats aged 15 months (n= 11) and 6 months (n= 10) served as control groups for the experiments in the radial maze and on the homing board respectively.
Transplantation The septal suspension was prepared according to Bj6rklund et al. (1983a). The developing septal region was dissected from E14~15 donor rat fetuses following decapitation of the pregnant females. The tissue was collected in sterile culture medium at 4 ~ C, incubated in buffered trypsin at 37 ~ C for 20 rain, washed 5 times with fresh 0.6% glucose saline, and finally brought to a volume of 10-15 ~tl per dissected embryo. Mechanical dissociation was performed by repeated pipetting through a Pasteur pipette with fire rounded opening. The viability of the cells (usually above 90%) was controlled in an acridine orange+ ethidium bromide solution as described by Bj6rklund et al. (1983). Cells were counted and the final volume of the suspension was adjusted if necessary so that about 50 000 cells could be injected in 2 gl aliquots. This was certainly a lower limit since small tissue pieces remained undissociated by this procedure. Three 2 gl aliquots were injected stereotaxically into the hippocampus on each side. Rats were anesthetized with a mixture of Ketalar (10 mg/kg) and Rompun (5 mg/kg) and placed in a Kucera stereotaxic apparatus (Kucera 1970) with lambda and bregma horizontal. The following anteroposterior and lateral coordinates were taken from lambda, and the ventral position was measured from the dura: (1) A = +4.2 L a t = :t:3.5 V = 3 . 0 (2) A = +2.7 L a t = • V = 3.7 (3) A = + 2.7 L a t = • 4.8 V = 5.7. These coordinates had been adjusted to deposit the graft on the surface of the hippocampus, in order to allow correct reinnervation while preventing the development of the grafted cells in the hippocampal layers.
Behavioral testin9 The old rats were trained in a radial maze during 5 weeks (4~9 weeks posttransplantation), thus at the age of 24.5 to 25.5. They were tested on a homing board task during one week when aged 27
RADIAL MAZE
months (about 16 weeks posttransplantation). In the radial maze, their performance was compared with that of younger controls then aged 15-16 months. In the second task, 6 month old rats served as younger controls.
Radial maze. Eight transparent plexiglass tunnels (12 x 12 x 60 cm) were arranged on a table covered with a plastic sheet to form an enclosed radial maze. Each arm was baited with one drop of diluted condensed milk deposited on a small glass plate. For a particular training condition, two nearby tunnels were placed side by side and thereby rendered parallel (2-parallel radial maze, see Fig. 1). These two arms were separated from the adjacent arm on either side by a 67.5-degree angle. The remaining 6 arms were maintained at a 45-degree angle from each other. For a given subject, the same pair of arms was always parallel, but different combinations were used for different rats. During each trial, the rats were introduced into a large vertical cylinder placed in the central choice area (diameter 30 cm) and released after 5-10 s into the maze by lifting the cylinder. The central choice area was then covered by a plexiglass plate. The rats were allowed to visit all the tunnels or to spend 10 rain in the maze. After completion of the maze, they were left in it for one extra minute. An observer recorded the time of entry into each tunnel. Rats were given a total of 20 trials at the rate of 4 trials/week. The maze had a classical radial shape for the first 8 trials, was arranged into a 2-parallel arm maze for the following 8 trials, and was restored to its original shape for the last 4 trials.
Homing board. A large circular hole board (diameter 160 cm) was enclosed by a 40 cm high wall. It was provided with 15 possible escape holes, grouped in 5 triplets (Fig. 1). One hole was connected to a flexible tube while the remaining 14 holes were closed by tightly fitted plugs of foam rubber. All the holes were covered by a disk of white plastic of the same material as the sheet covering the board. Only the external hole positions were used for escape. The empty home cage was placed under the table and connected to the flexible tube. A drop of condensed milk was deposited on a small glass plate in the home cage in order to sustain a high motivation to reach home throughout training. The room was rich in distal cues (e.g. door, curtains, posters, filing cabinet) but had no window. Four 60-watt light bulbs and four infrared spots were used as ceiling lights, and two 60-watt light bulbs provided indirect illuminations of the distant landmarks. A pretrainin9 procedure was conducted in the animal colony to ensure that most subjects would enter the connected hole as of the first training session. One day before the first training session, the cover of the home cage was replaced by a plywood board with a
cue
HOMING HOLE BOARD
latency t r a i n i n g trials 1 - 8 , 1 7 - 2 0
trials 9-16
Fig. 1. Schematic representation of the two experimental set ups with indications of the experimental designs. On the homing board, the stars (*) indicate the starting positions, and the dark dot (e) signals the position of the proximal cue (striped cylinder) in the
t r a i n i n g trials (1-18)
place
time in sectors p r o b e trial (19)
centre of the training triplet (only the peripheral positions were used for escape). For the probe trial, the hole sectors are marked as tinted areas, the darkest one being the training sector. The time spent in each of these sectors is computed from the recorded track
644 connected hole and a short flexible tube similar to those of the experimental arena; rats were placed on this cover and allowed to reach home through the hole. A large transparent cover was placed above the board so that the animals could be left alone long enough to enter spontaneously. For each trainin9 trial, the subject was released by hand in one of 5 possible starting locations, facing the centre of the arena (see Fig. 1). For the first three trials, the rat was allowed a maximum of 5, 3 and 3 min respectively, to explore the enclosure and enter the tube leading to the home cage. During subsequent training, it was allowed a maximum of 72 s to find the training hole. Rats not finding the training hole in the allotted time were gently placed on it. After having entered the home cage, the test subject was left in it for an intertrial interval of about 3 minutes. Meanwhile, the fitting and cover of the hole through which the rat had entered were replaced by clean ones and the table was rotated. A different hole was connected to the flexible tube, but at the same spatial position. Thus the rats always escaped in the same position in space relative to distant room landmarks, whereas the olfactory cues left around the hole were made irrelevant by intertrial rotations of the table. A striped cylinder (height 14 cm, diameter 5 cm) placed in the middle of the escape hole triplet served as a proximal landmark. The rats were thus not required to use extra-arena information to locate the escape hole, although distant room landmarks remained relevant. All the holes were only partially covered by the plastic discs during the first training trial. The holes were progressively masked during the following trials so as to be completely covered on the third training session. By this time, all the rats had learnt to remove the plastic lid in order to uncover a hole. A total of eighteen trials were conducted in four days (respectively 3, 5, 5 and 5 trials). On day 4, the 19th trial was a 2-min probe trial with no hole connected and no proximal landmark present. Each trial was video recorded through an infrared TV camera placed above the enclosure, relaying the output to a video recorder and a monitor. The paths taken to reach the escape hole were analysed by an XY video tracker (Kukam S.A.) coupled to a PDP 1103 computer. In this experiment, the latency to reach the escape hole was analysed throughout training. The probe trials were analysed in greater detail. The time spent in various sectors of the table was analysed as indicated in Fig. 1.
Morphology
Between 4 and 16 weeks after completion of testing, the old rats were perfused via the ascending aorta with 200 ml of phosphate buffered formaldehyde added with 1.5 % picric acid. The brains were postfixed in the same solution for 3 h, then placed in phosphate buffer at 4 ~ C. They were successively transferred to three solutions of 10, 20 and 30% sucrose over three days before sectioning. Horizontal or coronal (50% of the brains) 60 gm thick sections were cut on a freezing microtome. Every third section was stained for acetylcholinesterase (Targo et al. 1986) or cresyl violet alternatively.
Results
Radial maze
Preliminary statistical analyses did n o t show any difference between s h a m and untreated old rats and therefore their results were c o m b i n e d for further analyses. D a t a o b t a i n e d f r o m one grafted rat in the radial maze were discarded f r o m the statistical analyses because o f an error in procedure during training. It was indeed easy to recognize the old rats by their general posture, r u n n i n g style and increased time spent sniffing the ground. There were no m a r k e d differences
5"
all radial
2-parallel
all radial
4
3'
9
young
--o--
old
........ ~ ......
graft
1t 0
,
i
i
i
3 4 5 Blocks of 4 trials Fig. 2. Mean number of errors per trial (• SEM) to complete the 8-arm radial maze (8 trials), the 2-parallel maze (8 trials) and the radial maze (4 trials) 0
1
2
between transplanted and control old rats. W h e n observed, they were subtle, and mainly related to the m a n n e r in which the rats chose the a r m to be entered. Transplanted rats seemed to be slightly m o r e steady a n d less hesitant, but this did n o t result in significant differences in running speed. The m e a n latency to 8 visits (errors included) was longer in b o t h g r o u p s o f old rats than in the y o u n g e r control subjects (mean 4- SE o f the time in 20 trials, control old rats: 99 4- 6 s; grafted rats: 93 4- 8 s; y o u n g rats 64 4- 4 s). A o n e - w a y A N O V A confirmed that the age effect was significant (F 2,26 = 8.16, p < 0.002; b o t h c o m p a r i s o n s between old and y o u n g e r rats were significant, Fisher P L S D , p < 0.001). Figure 2 shows that the error rate varied significantly t h r o u g h o u t training (two-way A N O V A , block effect F 4,108 = 5.89, p < 0.001). There was an overall decrease during the first training phase in the s t a n d a r d radial m a z e (block effect, F 1,27=6.53, p < 0 . 0 2 ) , but no treatment effect. As expected, modifying the structure o f the maze induced a significant increase in error rate between the second and the third block o f four trials ( F 1,27 = 6.33, p < 0.02). A n interaction between g r o u p and block effects revealed that this increase was n o t equal in all the groups (F 2,27 = 3.94, p < 0.05). Indeed, there was a significant difference between groups during training in the 2-paralM maze (F 2 , 2 7 = 3 . 4 1 , p < 0 . 0 5 ) , w i t h o u t significant block effect or interaction. I n fact, the control old rats showed a significantly higher error rate (Mean 4- SE, old: 3.1 4-0.5) t h a n the y o u n g e r rats (1.84-0.4) in the twoparallel radial maze (Fisher P L S D significant, p < 0.05). The error rate by the old grafted rats (1.9 + 0.3) was close to that o f the y o u n g e r rats t h o u g h n o t significantly lower than t h a t o f the c o n t r o l old subjects (Student t = 2.007, 17 df, n.s.). Finally, there was an overall r e d u c t i o n in error rate w h e n the rats were retested in the classical radial maze ( A N O V A o f the last two blocks o f f o u r trials F 1,27 = 19.45, p < 0.0001). Figure 2 indicates clearly that all the groups reached an equally g o o d performance. As reported elsewhere (Schenk et al. 1990), the low foraging efficiency in the 2-parallel m a z e was related to a neglect o f one o f the parallel arms. In fact m o s t errors were d o n e while all the radial arms h a d already been visited and only one a r m o f the parallel pair had to be
645
found. Parallel errors were thus defined as re-entries into the first visited parallel arm while the second remained unvisited (confusion), or re-entries into any radial arm when only the parallel was missing (neglect). These specific errors accounted for the poor performance of the control old rats. These rats showed a higher mean frequency of parallel errors than did the two other groups (F 2,27=4.41, p
