- Email: [email protected]
Age-related structural changes in the rat hippocampus: correlation with working memory deficiency
Brain Research, 512 (1990) 113-120 Elsevier
113
BRES 15297
Age-related structural changes in the rat hippocampus: correlation with working memory deficiency Tamar Kadar 1, Michael Silbermann e, Rachel Brandeis I and Aharon Levy 1 1Department of Pharmacology, Israel Institute for Biological Research, Ness-Ziona (Israel) and 2Department of Morphological Sciences, Rappaport Institute for Medical Research, Faculty of Medicine, Technion, Haifa (Israel) (Accepted 15 August 1989)
Key words: Rat; Hippocampus; Muscarinic receptor; Cholinergic system; Acetylcholinesterase; Lipofuscin; Working memory
Age-related histopathological changes in the hippocampal formation were correlated with cognitive performance, evaluated in rats at the 8-arm radial maze. Experiments were conducted using young (3 months), mature (12 months), middle-aged (17 months) and aged (24 months) Wistar rats. Significant memory impairments were already observed at the age of 12 months in all the measured parameters (correct choices, percent errors and total time). No further decline was observed between 12 and 17 months of age, while at 24 months additional decline was monitored mainly in the percent errors parameter. Morphometric analysis revealed a decrease in the area of cells within the hippocampus and the number of cells in the CA3 subfield. This pattern of morphological changes with age corresponded well with the cognitive impairments, with high correlation especially to lesions at the CA3 subfield. It had also been confirmed in this study that lipofuscin appeared to be a good histochemical marker for CNS cell degeneration. It is concluded that 12-month-old Wistar rats may serve as the animal model of choice for the study of specific age-related behavioral deficits and that the hippocampal CA3 region might play a major role in the age-dependent cognitive decline. INTRODUCTION The role of the central cholinergic system in cognitive processes has been well established, particularly in association with the functional decline accompanying normal as well as pathological aging. Several attempts have been reported whereby behavioral deficits have been correlated with morphological changes in the brains of animals, and these referred to experimental models which used destruction of specific brain regions by mechanical procedures 14, intracerebral injection of neurotoxins 6A7 or via non-invasive approaches such as cerebral ischemia 4. Most aging animals exhibit a decline in their m e m o r y functions and, therefore, may serve as an alternative suitable model for morphobehavioral correlation studies. Brizze and Ordy 2 suggested that the significant age-related changes noted in short-term memory passive-avoidance tests in rats could be related to cell loss and/or increased lipofuscin accumulation in regions such as the hippocampal CA1 zone and visual cortex area. On the other hand, Freund 7 indicated that there was no correlation between age-related impairment of avoidance learning and the increased deposition of lipofuscin in brain homogenates. It has been well established that part of the aging animals continue to perform like young ones. Thus,
senescent animals can be divided into two subpopulations and at least one study analyzed these two groups separately, trying to correlate between morphological brain changes and behavioral decrements. Geinisman et al. 8 have indeed found that the decrease in the number of perforated synapses in the dentate gyrus correlated with the degree of m e m o r y impairment in aged rats. As cognitive functions were shown to depend on the structural integrity of the hippocampal formation 15, we elected to focus on this region as a representative of the cholinergic system for morphobehavioral evaluations. This study utilized a behavioral paradigm, in which spatial working memory was measured, since this cognitive function has been known to be most sensitive to cholinergic manipulations. Hence, this study monitored hippocampal morphological changes in rats of different ages, whose Working memory performance had been previously measured. Thus, we attempted to correlate between the changes in behavior and structure in the aging brain of rats on an individual basis. MATERIALS AND METHODS
Animals Male Wistar rats (obtained from Charles River, U.K.) were used in this study. The animals were divided into the following age
Correspondence: T. Kadar, Department of Pharmacology, Israel Institute for Biological Research, 70450 Ness-Ziona, Israel. 0006-8993/90/$03.50 (~ 1990 Elsevier Science Publishers B.V. (Biomedical Division)
114 groups: 3, 12, 17 and 24 months. Animals were housed two in a cage, were maintained on a 12 h light-dark cycle and provided with rat purina chow and water ad libitum. The animals were initially tested for their working memory performance using an 8-arm radial maze and were then sacrificed for either histological or histochemical examinations.
Behavioral procedure The 8-arm radial maze, which was used as the behavioral paradigm in this study, enabled the monitoring of subtle changes in working memory performance• This task involved the successive selection of arms, radiating from the center of the maze in order to obtain a food reward (45 mg precision pellets from Bioserv Inc.). The maze was made of white painted wood and the arms (75 cm long and 9 cm wide), extended from an octagonal central arena (30 cm wide). Prior to the training program, the rats were put on a restricted diet until their body weight declined to approximately 80% of their initial weight. Following three days of shaping, daily training sessions were begun. At the start of a trial, the rat was placed in the center of the maze and was allowed to run throughout the maze until collecting all eight pellets or until 15 min had elapsed (whichever came first)• Rats were trained in the maze for 23 days (4-5 days a week), during which time all responses were recorded. Each age group was trained with a control group of young animals•
included 4 animals at the least. The average surface area ol a single cell in each region was calculated as the mean value of 30 cells (10 cells per animal, 3 animals per group). The number of lipofuscin-positive cells was determined in PHG-stained sections in regions CA1, CA3 and DG. Cells were counted in both hemispheres in at least two serial sections with a × 16 microscope objective, using a frame of 81,0(10 #m -~. The same setting of the CIAS was used through all the measurements.
Data analysis Two-way analysis of variance (ANOVA) with repeated measures (unequal 'n'), was performed across age groups for the behavioral data of the study and for lipofuscin accumulation. One way ANOVA was carried out for the morphometric results. Whenever statistical significance was observed, further analysis was carried out (Scheffe contrasts analysis). The relation between the behavioral data and the morphometric findings was determined using the Pearson's correlation-coefficient.
RESULTS
Behavior Three behavioral parameters
w e r e m o n i t o r e d in t h e
Histology and histomorphometry
8 - a r m r a d i a l m a z e : (a) t h e n u m b e r o f c o r r e c t c h o i c e s o u t
Brains were rapidly removed from the skull, fixed in 4% neutral buffered paraformaldehyde and processed routinely for paraffin embedding. Coronal sections, 5 # m thick, cut at the hippocampal level, were stained with hematoxylin and eosin (HE), and with the PHG staining ~2. The latter is a modification of the periodic acid-Schiff reaction and was used for the localization of lipofuscin. All animals were primarily coded, and the histological evaluation was performed with no previous knowledge of the behavioral scores. Morphometric analysis was performed using a computerized image analysis system (CIAS) connected to a colored video camera (The Olympus Cue-3, Galai, Migdal Haemek, Israel). All measurements were carried out in sections which corresponded to Fig. 21 in Paxinos' and Watson's rat brain atlas a6, at the hippocampus subfields CA1, CA3 and dentate gyrus (DG) and in the medial habenula. The latter was selected as a highly innervated cholinergic region, whose role in cognitive processes is yet to be discovered• The morphometric CIAS analysis included: number of cells per defined area, single cells' surface area and the number of lipofuscin positive cells. Measurements were performed in a frame of 32,000 #me with a ×25 objective magnification, using both hemispheres. Three serial sections were used per each animal and each age group
o f t h e first 8 e n t r i e s ; ( b ) t h e p e r c e n t o f t o t a l e r r o r s in o n e s e s s i o n ; (c) t h e t o t a l t i m e f o r t h e c o m p l e t i o n o f a s e s s i o n . T h e daily s c o r e s w e r e a n a l y s e d in w e e k l y b l o c k s . Figs. 1 a n d 2 i l l u s t r a t e d a t a a c q u i r e d in t h e m a z e t a s k , at t h e v a r i o u s a g e g r o u p s • It b e c a m e a p p a r e n t t h a t y o u n g r a t s ( 3 - m o n t h - o l d ) , p e r f o r m e d in t h e m a z e at a m u c h f a s t e r r a t e t h a n o l d e r rats. A l t h o u g h
the performance
o f all
g r o u p s i m p r o v e d f o l l o w i n g five w e e k s o f t r a i n i n g , t h e ultimate degree of performance
a t t a i n e d b y a g e d rats
( 2 4 - m o n t h - o l d ) , w a s less t h a n t h a t o f y o u n g rats. Twoway A N O V A
test f o r r e p e a t e d
measures
(across age
groups and weeks of training) revealed highly significant age effects (F3,45 = 7.69, P < 0.001 f o r c o r r e c t c h o i c e s a n d E~.45 = 57.72, P < 0.001 f o r t o t a l e r r o r s ) . O f g r e a t i n t e r e s t was t h e f i n d i n g t h a t a l r e a d y m a t u r e a n i m a l s (12 m o n t h s ) d i f f e r e d s i g n i f i c a n t l y f r o m t h e y o u n g a n i m a l s as
80 O 3 II1~.
O 3 I1~.
g
• t21m.
70
•
liB.
60
•
24m~.
• ~z mm. •
IN!
17raft.
• z4 lion.
50
~W
| t
2 3 4 TRAINING SESSIONS (NEEKLY RLOCKS)
5
Fig. 1. Mean (+S.E.M.) correct choices during the first 8 trials in the 8-arm radial maze at various age groups.
t
2 3 4 TRAINING SESSIONS (WEEKLY8LOCK£)
5
Fig. 2. Mean (+_S.E.M.) percent o f total errors in the 8-arm radial
maze at various age groups.
115 indicated by Scheffe's contrast analysis (P < 0.001). No difference was found between the performance rate of the 12- versus the 17-month-old animals, yet both differed significantly from the 24-month-old rats (P < 0.01). The 'correct choices' parameter increased significantly with training (for all age groups), however, the tendency of decrease with respect to the 'percent total errors' was not statistically significant. Fig. 3 illustrates the total time that was needed to complete a session. It could be noted that the pattern of the latter followed those of the other two behavioral parameters. A high degree of significance was recorded using A N O V A for the aged groups effect (F3,45 = 13.10, P < 0.001). A significant training effect was observed (P < 0,001) and an interaction was found between age and weeks of training (P < 0.005), according to which the young animals completed the task much faster than the other aged animals throughout the training weeks. An additional important feature which characterized the performance of mature but not young animals in the maze, was the large variation between animals as expressed by the relatively high values of standard errors.
O~.
la
•
12 ~n.
16 -,-:,
14
=:
12 tO 8 B 4 2 0
1
2 3 4 TRAINING SESSIONS (WEEKLYBLOCKS)
5
Fig. 3. Mean (+S.E.M.) total time for completing a session in the 8-arm radial maze at various age groups. This phenomenon which was found in all the aged groups, suggested a non-homogenous population in these groups.
Histology Distinct morphological changes were noted in various
Fig. 4. Sections through the hippocampal CA3 (A,B) and CA1 (C,D) subfields of 12- (A,B) and 17- (C,D) month-old Wistar rats. A and C represent intact morphology, whereas B and D represent examples of the degenerative changes such as those noted in half of the animals in these age groups. Note the shrinkage of cells and the intense staining for lipofuscin in B,D. PHG staining. A,B - - original magnification x400. C , D - original magnification xl60. Arrows (D) indicate the presence of tangle-like structures in CA1 pyramidal cells.
116 regions in the brain but especially in the hippocampus already by the age of 12 months. At this time interval degenerative changes were observed in about 50% of the subjects (n = 6) in regions CA3, CA1 and in the DG, but not always in both hemispheres. Similar changes were observed in the 17-month-old group. These included shrinkage of pyramidal cells at the CA3 and CA1 subfields, and nucleic heterochromasia. In sections stained with PHG, 'dark' cells represented cellular
accumulation of lipofuscin which was resistant to diastase pretreatment (Fig. 4). Such reactivity was not observed in specimens of the young animals. Pyknotic cells in the CA1 region had a tangle-like appearance (Fig. 4D). However, intact pyramidal cells were also encountered in damaged regions. The ratio between the two cell populations (intact vs. damaged) varied among areas, hemispheres and animals. A pattern of hemispheric asymmetry was found in reference to the morphological changes.
Fig. 5. Sections through the hippocampal CA3 (A) and CAt (B) subfields of 24-month-old Wistar rats. Note degenerative pyramidal cells in both regions. HE staining. ×160.
117 TABLE I
Age-related morphometric analysis of single-cell surface area in cholinergic regions of the Wistar rat brain The mean surface area of single cells was calculated from 120 measurements (20 cells in each animal, and 6 animals per age group).
Age (months)
Surface area (Itrn2 + S. E. M. )
3 12 17 24
CA3 subfield
CA1 subfield
Dentate gyrus
Medial habenula
271.2 205.4 190.13 150.7
162.3 121.05 121.9 126.7
97.0 68.7 80.05 71.7
85.83 63.7 72.16 72.9
+ 10.94 + 23,07** + 16.60 ___12.23"
+ + + +
17.35 9.30 8.28 4.07
+ + + +
2.94 2.45** 7.2 3.96
+ + + +
6.52 1.93"* 3.66 3.73
*P < 0.01; **P < 0.001.
In some of the animals, necrotic cells were observed at the inner layer of the DG. As already indicated above, in half of the animals (aged 12 and 17 months), the hippocampus appeared structurally intact and did not differ in its morphology from that noted in the young group. At the age of 24 months, the number of necrotic cells increased in all areas of the hippocampus, accompanied by atrophy and vacuolization (Fig. 5), a feature that was evident in 85% of the animals. The tangle-like appearance of pyramidal cells at the CA1 area became pronounced along with the presence of cells' residues (Fig. 5B). It should be however, stressed that even in the 24-month-old group, some of the animals revealed normal hippocampal morphology.
Morphometry Tables I and II represent the morphometric data related to cells (surface area and number) in regions CA1, CA3 and D G of the hippocampus as well as the surface area of single cells in the medial habenula. A significant age effect was found according to a one-way ANOVA. Specifically, a remarkable decrease was found from the 3 to the 12 month groups in surface area (Shifee contrasts; P < 0.001) for CA3, DG and medial habenula
and in number of CA3 cells (P < 0.001). In some regions an additional decrease was noted with increasing age. This phenomenon was particularly pronounced at the CA3 region, in which a significant decline was also noted between 17 and 24 months (P < 0.01). Statistical analysis concerning the CA1 subfield revealed a significant decrease in the total area occupied by the cells between 3 and 12 months (P < 0.025). The number of cells was almost constant with age, except for the CA3 subfield (Table II). Quantitative analysis of lipofuscin accumulation in the hippocampus showed an increase especially between 3 and 12 months of age (Fig. 6). The number of lipofuscin positive cells in CA3 subfield continued to rise until 24 months. As already mentioned above, a large variation was found between animals in each age group, except for the young ones. Noticeable accumulation of the pigment was observed only in approximately half of the animals, thus leading to an additional division of each age group into two sub-populations: 'impaired' versus 'non-impaired'. Animals showing at least 10 lipofuscin-positive hippocampal cells in a frame of 81,000~m 2 (the measured area) were considered as 'impaired'. Table III represents
LIPOFUSCIN ACCIlHULA1|OH
20 -2.
x.
W
T A B L E II
Or01 w
Age-related morphometric analysis of the number of cells in hippocampal subfields in the Wistar rat brain
x
N u m b e r of cells was counted in a frame of 32,000/~m 2 in at least 3 serial sections.
N
Age (months)
3 12 17 24 * * P < 0.001.
mcl3
!o
No. of cells (+ S.E.M.) CA3 subfield
CA1 subfield
Dentate gyrus
41 27.5 33.2 27.12
27.5 23.5 24.07 23.5
41.6 37.0 42.33 37.27
+ + + +
3.75 1.91"* 2.03 2.28
___1.52 + 1.08 + 0.7 + 0.88
_+ 5.07 + 2.02 + 4.55 + 2.85
0
5
10
15
20
25
A6E (Mt~T~I Fig. 6. Age related analysis of the n u m b e r of lipofuscin-positive cells in the hippocampus of Wistar rats.
118 T A B L E III
T A B L E IV
Quantitative analysis of lipofuscin-positive cells in the hippocampus of aging Wistar rats
Correlation between behavioral scores' (percent total errors) and hippocampal morphometric parameters in aged Wistar rats (Pearson's test)
Animals were divided into two sub-groups in each age group: 'impaired' and 'non-impaired'. A hippocampus that contained at least 10 lipofuscin-positive cells (per area examined) was considered as 'impaired'. Data represent the m e a n value of both hemispheres of 4 animals (in each subgroup).
Age (months)
No. of lipofuscin-positive cells _+S. E. M. CA3
CA1
DG
12 12"
1.18+0.22 12.75 _+ 7.18
2.57_+ 1.24 12.5 _+ 3.36
0.87_+0.21 26.33 _+ 7.32
17 17"
3.33 _+ 0.08 10.66+2.44
2.16 _+ 0.37 11.16_+6.45
1.5 _+ 1.06 13.16_+6.22
24 24*
6.06 + 1.42 22.58 + 4.5
2.87 + 0.71 11.08 + 6.15
3.62 + 1.0 8.58 + 4.6
P < 0.001
P < 0.01
P < 0.001
* Two-way A N O V A revealed significant differences between the two subgroups throughout in all areas examined.
the relevant data which were found to be significant throughout.
Morphobehavioral correlation The decline in the cognitive performance, which was noted already at the age of 12 months, was accompanied by morphological changes in the hippocampaus. The division of each age group into two sub-groups (according to their histology), indicated the existence of two different patterns of performance; a feature that was eminent mainly in the 12 and 17 months age groups, in which each of these two subgroups comprised half of the population (Fig.7).
N = 22 for all age groups; N = 9 for the 3- and 12-month-old age groups.
Morphometric parameters
r
Significance
CA3 - - s i n g l e cell area All age groups 3- and 12-month-old age groups
-0.544 -0.842
P < 0.01 P < 0.01
C A 1 - - area occupied by cells All age groups 3- and 12-month-old age groups
-0.362 -0.705
N.S. P < 0.05
Lipofuscin-positive cells 3- and 12-month-old age groups
0.669
P < 0.01
Correlative analysis of the morphometric data and behavioral findings as related to the working memory deficiency (expressed by total errors) revealed a high degree of relationship. The correlation data are summarized in Table IV. As can be seen, a significant correlation was found especially between cellular area in CA3 subfield and the percent of total errors r = -0.544; P < 0.01 for all age groups (Fig. 8). Higher correlation was found regarding these parameters for the 3- and 12month-old groups (r = -0.842; P < 0.01) for the individual scores and r = -0.98; P < 0.02 for the mean group scores). For lipofuscin, a correlation was found between the number of positive cells and working memory deficiency (total errors) only at the age groups of 3 and 12 months (r = 0.669; P < 0.01).
12 ~ T ~
N O ~ V I O B A L C(X~LATION Q Ml'lll hllI~Ompus
50
•
e--o.544
llllnrld
5O
70
a~
|
40
60
30 2O
~
5o
~
4o
~-
ao 2O
tO
tO t
2 3 4 TRAINII~ 9ESSIONS (WEEKLYBLOCKS)
5
Fig. 7. Age-related changes in working m e m o r y deficiency (expressed by total errors + S . E . M . ) as recorded in 12-month-old rats which were divided into two subgroups according to morphological criteria: the a m o u n t of lipofuscin in hippocampal cells.
50
100
t50 200 250 300 350 CA3 -SINGLE CELL AREA (tlIL)
400
450
Fig. 8. Linear correlation between the area of a single cell in CA3 region and working m e m o r y deficiency (total errors) in Wistar rats in various age groups.
119 DISCUSSION Age-related performance decrements were observed already by 12 months of age, an age group which is usually omitted from aging studies. This behavioral change was highly significant when compared to young animals. The 12-month-old group comprised of a heterogeneous population as was deduced from the large values of standard deviation in all the measured behavioral parameters. In another study, using a different behavioral paradigm (passive avoidance), a 30-40% decline was reported in 12-month-old mice, a decline which correlated well with the decrease in acetylcholine synthesis 9. The findings of the present study appear consistent with a previous study which showed an age-related impairment in both working memory and reference memory tasks 1. Most of the animals in all the age groups improved their performance following training, a fact that was statistically significant with regard to correct choices and the time needed for the completion of a session. The parameter of total errors, however, did not improve with training, because of a learning deficiency in the aged groups. The sharp decline in performance, as observed in rats at 12 months of age, was accompanied by changes in brain morphology among which the hippocampus was found to be the most affected region. Fifty percent of all animals in this age group revealed an apparent normal hippocampal structure, a feature that was associated with an unimpaired performance. An analysis of the middleaged group as two subgroups according to their morphobehavioral scores, revealed the existence of two different populations. One of these subpopulations appeared relatively unaffected, in which the animals performed as well as young animals. A similar division of aged rats to 'impaired' vs 'non-impaired' subgroups was suggested by De Toledo-Morrell et al. 5 and Landfield et al. 13. In both studies the hippocampal synapses of 'impaired' aged animals were shown to exhibit pronounced electrophysiological and morphological aberrations found neither in young nor in 'non-impaired' aged rats. The present morphological study supports previous data indicating the sensitivity of the hippocampus to age-related degeneration. Not all the hippocampal subfields showed a similar pattern of age-related degeneration. Despite the fact that parameters, such as number and area of cells, decreased by the age of 12 months, the REFERENCES 1 Barnes, C.A., Nadel, L. and Honig, W.K., Spatial memory deficit in senescent rats, Can J. Psychol., 34 (1980) 29-39. 2 Brizzee, K.R. and Ordy, J.M., Age pigments, cell loss and hippocampal function, Mech. Aging Dev., 9 (1979) 143-162.
best correlation between structural changes and working memory performance was found with regard to surface area of pyramidal cells in the CA3 region. This observation suggests a possible relationship between a degeneration of cholinergic components, especially at CA3 subfields, and memory deficiency. Similar observations in reference to CA3 cells was reported for mice 11. Handelmann and Olton TM have arrived to the same conclusion following kainic acid lesion of CA3, thus emphasizing the importance of CA3 pyramidal cells for hippocampal circuitry and as a junction point between cortical and subcortical structures. Another marker for age-related changes in nerve cells is lipofuscin accumulation. This 'aging pigment' is believed to represent an oxidative degeneration process and therefore, might indicate neuronal dysfunction. In rats, the accumulation of this pigment increased with age, starting at the age of 12 months. It should be emphasized that alike other morphometric parameters, the appearance and deposition of the pigment varied among animals with high correlation to the level of performance. At the old age group (24 months of age), the morphobehavioral correlations was found to be much lower, presumably due to the fact that many other non-specific age-related processes were taking place. Cognitive decrements at this age might have stemmed from additional non-central physiological injuries. It is therefore proposed that mature and middle-aged rats are the group of choice for morphobehavioral correlation studies. In our opinion, at this age group, behavioral decrements stem mostly from cognitive deficiency, while with older rats impaired physical capacity may affect behavioral results. Light microscopy studies, such as the one hereby reported, yield data on a large number of cells, while ultrastructural EM analysis may provide further information on subcellular elements involved. In conclusion, this study provides further support to the notion that the hippocampus plays a key role in the neurobiology of dementia and that the degree of cognitive impairments is highly correlated with degenerative structural changes in this region of the brain, in particular with changes in the CA3 pyramidal cells. Acknowledgements. This study was supported in part by a grant from the United States-Israel Binational Science Foundation. The authors wish to thank Ms. R. Sahar for her skilled technical assistance.
3 Coleman, P.O. and Hood, D.G., Review. Neuron numbers and dendritic extent in normal aging and Alzheimer's disease, Neurobiol. Aging, 8 (1987) 521-545. 4 Davis, H.P., Tribuna, J., Pulsinelli, W.A. and Volpe, B.T., Reference and working memory of rats following hippocampal damage induced by transient forebrain ischemia, Physiol. Be-
120 hay., 37 (1986) 387-392. 5 De-Toledo-Morrell, L., Geinisman, Y. and Morrell, E, Review: Age-dependent alterations in hippocampal synaptic plasticity: relation to memory disorders, Neurobiol. Aging, 9 (1988) 581-590. 6 Fisher, A. and Hanin, I., Mini review: choline analogues as potential tools in developing selective animal models of central cholinergic hypofunction, Life Sci., 27 (1980) 1615-1634. 7 Freund, G., The effects of chronic alcohol and vitamin E consumption on aging pigments and learning performance in mice, Life Sci., 24 (1979) 145-152. 8 Geinisman, Y., Toledo-Morrell, L. and Morrell, F., Aged rats need a preserved complement of perforated axospinous synapses per hippocampal neuron to maintain good spatial memory, Brain Research, 398 (1986) 266-275. 9 Gibson, G.E. and Jenden, D.J., Brain acetylcholine synthesis declines with senescence, Science, 213 (198t) 674-676. 10 Jucker, M., Oettinger, R. and Battig, K., Age-related changes in working and reference memory performance and locomotor activity in the Wistar rat, Behav. Neural Biol., 50 (1986) 24-36. 11 Kadar, T., Silbermann, M. and Levy, A., Age-related changes in cholinergic components within the central nervous system (hippocampus, NBM and habenula) of CWI female mice, Mech.
Aging Dev.. 47 (1989) pp. 133-144. 12 Kuttin, E.S. and Beemer, A.M., PHG-staln for histology, frozen sections and cytology. In E. Levy (Ed.), .ldvances in Pathology. Vol. 1, 1982, pp. 115-116. 13 Landfield, P.W., Review: hippocampal ncurobiological mechanisms of age-related memory dysfunction, Neurobiol. Aging, 9 (1988) 571-579. 14 Olton, D., Walker, .I.A. and Gage, F.H., Flippocampal connections and spatial discrimination, Brain Research, 139 (1978) 295-308. 15 Olton, D.S., Memory functions and the hippocampus. In W. Seifert (Ed.), Neurobiology of the Hippocampus, Academic Press, London, 1983, pp. 335-373. 16 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic (2)ordinates, Academic Press, 1982. 17 Sinden, J.D., Rawlins, J.N.P., Gray, J.A. and Jarrard, L.E., Selective cytotoxic lesions of the hippocampal formation and DRL performance in rats, Behav. Neurosci., 100 (1986) 320329. 18 Handelmann, G.E. and Olton, D.S., Spatial memory following damage to hippocampal CA3 pyramidal cells with kainic acid: irr,pairment and recovery with preoperative training, Brain Research, 217 (1981) 41-58.
113
BRES 15297
Age-related structural changes in the rat hippocampus: correlation with working memory deficiency Tamar Kadar 1, Michael Silbermann e, Rachel Brandeis I and Aharon Levy 1 1Department of Pharmacology, Israel Institute for Biological Research, Ness-Ziona (Israel) and 2Department of Morphological Sciences, Rappaport Institute for Medical Research, Faculty of Medicine, Technion, Haifa (Israel) (Accepted 15 August 1989)
Key words: Rat; Hippocampus; Muscarinic receptor; Cholinergic system; Acetylcholinesterase; Lipofuscin; Working memory
Age-related histopathological changes in the hippocampal formation were correlated with cognitive performance, evaluated in rats at the 8-arm radial maze. Experiments were conducted using young (3 months), mature (12 months), middle-aged (17 months) and aged (24 months) Wistar rats. Significant memory impairments were already observed at the age of 12 months in all the measured parameters (correct choices, percent errors and total time). No further decline was observed between 12 and 17 months of age, while at 24 months additional decline was monitored mainly in the percent errors parameter. Morphometric analysis revealed a decrease in the area of cells within the hippocampus and the number of cells in the CA3 subfield. This pattern of morphological changes with age corresponded well with the cognitive impairments, with high correlation especially to lesions at the CA3 subfield. It had also been confirmed in this study that lipofuscin appeared to be a good histochemical marker for CNS cell degeneration. It is concluded that 12-month-old Wistar rats may serve as the animal model of choice for the study of specific age-related behavioral deficits and that the hippocampal CA3 region might play a major role in the age-dependent cognitive decline. INTRODUCTION The role of the central cholinergic system in cognitive processes has been well established, particularly in association with the functional decline accompanying normal as well as pathological aging. Several attempts have been reported whereby behavioral deficits have been correlated with morphological changes in the brains of animals, and these referred to experimental models which used destruction of specific brain regions by mechanical procedures 14, intracerebral injection of neurotoxins 6A7 or via non-invasive approaches such as cerebral ischemia 4. Most aging animals exhibit a decline in their m e m o r y functions and, therefore, may serve as an alternative suitable model for morphobehavioral correlation studies. Brizze and Ordy 2 suggested that the significant age-related changes noted in short-term memory passive-avoidance tests in rats could be related to cell loss and/or increased lipofuscin accumulation in regions such as the hippocampal CA1 zone and visual cortex area. On the other hand, Freund 7 indicated that there was no correlation between age-related impairment of avoidance learning and the increased deposition of lipofuscin in brain homogenates. It has been well established that part of the aging animals continue to perform like young ones. Thus,
senescent animals can be divided into two subpopulations and at least one study analyzed these two groups separately, trying to correlate between morphological brain changes and behavioral decrements. Geinisman et al. 8 have indeed found that the decrease in the number of perforated synapses in the dentate gyrus correlated with the degree of m e m o r y impairment in aged rats. As cognitive functions were shown to depend on the structural integrity of the hippocampal formation 15, we elected to focus on this region as a representative of the cholinergic system for morphobehavioral evaluations. This study utilized a behavioral paradigm, in which spatial working memory was measured, since this cognitive function has been known to be most sensitive to cholinergic manipulations. Hence, this study monitored hippocampal morphological changes in rats of different ages, whose Working memory performance had been previously measured. Thus, we attempted to correlate between the changes in behavior and structure in the aging brain of rats on an individual basis. MATERIALS AND METHODS
Animals Male Wistar rats (obtained from Charles River, U.K.) were used in this study. The animals were divided into the following age
Correspondence: T. Kadar, Department of Pharmacology, Israel Institute for Biological Research, 70450 Ness-Ziona, Israel. 0006-8993/90/$03.50 (~ 1990 Elsevier Science Publishers B.V. (Biomedical Division)
114 groups: 3, 12, 17 and 24 months. Animals were housed two in a cage, were maintained on a 12 h light-dark cycle and provided with rat purina chow and water ad libitum. The animals were initially tested for their working memory performance using an 8-arm radial maze and were then sacrificed for either histological or histochemical examinations.
Behavioral procedure The 8-arm radial maze, which was used as the behavioral paradigm in this study, enabled the monitoring of subtle changes in working memory performance• This task involved the successive selection of arms, radiating from the center of the maze in order to obtain a food reward (45 mg precision pellets from Bioserv Inc.). The maze was made of white painted wood and the arms (75 cm long and 9 cm wide), extended from an octagonal central arena (30 cm wide). Prior to the training program, the rats were put on a restricted diet until their body weight declined to approximately 80% of their initial weight. Following three days of shaping, daily training sessions were begun. At the start of a trial, the rat was placed in the center of the maze and was allowed to run throughout the maze until collecting all eight pellets or until 15 min had elapsed (whichever came first)• Rats were trained in the maze for 23 days (4-5 days a week), during which time all responses were recorded. Each age group was trained with a control group of young animals•
included 4 animals at the least. The average surface area ol a single cell in each region was calculated as the mean value of 30 cells (10 cells per animal, 3 animals per group). The number of lipofuscin-positive cells was determined in PHG-stained sections in regions CA1, CA3 and DG. Cells were counted in both hemispheres in at least two serial sections with a × 16 microscope objective, using a frame of 81,0(10 #m -~. The same setting of the CIAS was used through all the measurements.
Data analysis Two-way analysis of variance (ANOVA) with repeated measures (unequal 'n'), was performed across age groups for the behavioral data of the study and for lipofuscin accumulation. One way ANOVA was carried out for the morphometric results. Whenever statistical significance was observed, further analysis was carried out (Scheffe contrasts analysis). The relation between the behavioral data and the morphometric findings was determined using the Pearson's correlation-coefficient.
RESULTS
Behavior Three behavioral parameters
w e r e m o n i t o r e d in t h e
Histology and histomorphometry
8 - a r m r a d i a l m a z e : (a) t h e n u m b e r o f c o r r e c t c h o i c e s o u t
Brains were rapidly removed from the skull, fixed in 4% neutral buffered paraformaldehyde and processed routinely for paraffin embedding. Coronal sections, 5 # m thick, cut at the hippocampal level, were stained with hematoxylin and eosin (HE), and with the PHG staining ~2. The latter is a modification of the periodic acid-Schiff reaction and was used for the localization of lipofuscin. All animals were primarily coded, and the histological evaluation was performed with no previous knowledge of the behavioral scores. Morphometric analysis was performed using a computerized image analysis system (CIAS) connected to a colored video camera (The Olympus Cue-3, Galai, Migdal Haemek, Israel). All measurements were carried out in sections which corresponded to Fig. 21 in Paxinos' and Watson's rat brain atlas a6, at the hippocampus subfields CA1, CA3 and dentate gyrus (DG) and in the medial habenula. The latter was selected as a highly innervated cholinergic region, whose role in cognitive processes is yet to be discovered• The morphometric CIAS analysis included: number of cells per defined area, single cells' surface area and the number of lipofuscin positive cells. Measurements were performed in a frame of 32,000 #me with a ×25 objective magnification, using both hemispheres. Three serial sections were used per each animal and each age group
o f t h e first 8 e n t r i e s ; ( b ) t h e p e r c e n t o f t o t a l e r r o r s in o n e s e s s i o n ; (c) t h e t o t a l t i m e f o r t h e c o m p l e t i o n o f a s e s s i o n . T h e daily s c o r e s w e r e a n a l y s e d in w e e k l y b l o c k s . Figs. 1 a n d 2 i l l u s t r a t e d a t a a c q u i r e d in t h e m a z e t a s k , at t h e v a r i o u s a g e g r o u p s • It b e c a m e a p p a r e n t t h a t y o u n g r a t s ( 3 - m o n t h - o l d ) , p e r f o r m e d in t h e m a z e at a m u c h f a s t e r r a t e t h a n o l d e r rats. A l t h o u g h
the performance
o f all
g r o u p s i m p r o v e d f o l l o w i n g five w e e k s o f t r a i n i n g , t h e ultimate degree of performance
a t t a i n e d b y a g e d rats
( 2 4 - m o n t h - o l d ) , w a s less t h a n t h a t o f y o u n g rats. Twoway A N O V A
test f o r r e p e a t e d
measures
(across age
groups and weeks of training) revealed highly significant age effects (F3,45 = 7.69, P < 0.001 f o r c o r r e c t c h o i c e s a n d E~.45 = 57.72, P < 0.001 f o r t o t a l e r r o r s ) . O f g r e a t i n t e r e s t was t h e f i n d i n g t h a t a l r e a d y m a t u r e a n i m a l s (12 m o n t h s ) d i f f e r e d s i g n i f i c a n t l y f r o m t h e y o u n g a n i m a l s as
80 O 3 II1~.
O 3 I1~.
g
• t21m.
70
•
liB.
60
•
24m~.
• ~z mm. •
IN!
17raft.
• z4 lion.
50
~W
| t
2 3 4 TRAINING SESSIONS (NEEKLY RLOCKS)
5
Fig. 1. Mean (+S.E.M.) correct choices during the first 8 trials in the 8-arm radial maze at various age groups.
t
2 3 4 TRAINING SESSIONS (WEEKLY8LOCK£)
5
Fig. 2. Mean (+_S.E.M.) percent o f total errors in the 8-arm radial
maze at various age groups.
115 indicated by Scheffe's contrast analysis (P < 0.001). No difference was found between the performance rate of the 12- versus the 17-month-old animals, yet both differed significantly from the 24-month-old rats (P < 0.01). The 'correct choices' parameter increased significantly with training (for all age groups), however, the tendency of decrease with respect to the 'percent total errors' was not statistically significant. Fig. 3 illustrates the total time that was needed to complete a session. It could be noted that the pattern of the latter followed those of the other two behavioral parameters. A high degree of significance was recorded using A N O V A for the aged groups effect (F3,45 = 13.10, P < 0.001). A significant training effect was observed (P < 0,001) and an interaction was found between age and weeks of training (P < 0.005), according to which the young animals completed the task much faster than the other aged animals throughout the training weeks. An additional important feature which characterized the performance of mature but not young animals in the maze, was the large variation between animals as expressed by the relatively high values of standard errors.
O~.
la
•
12 ~n.
16 -,-:,
14
=:
12 tO 8 B 4 2 0
1
2 3 4 TRAINING SESSIONS (WEEKLYBLOCKS)
5
Fig. 3. Mean (+S.E.M.) total time for completing a session in the 8-arm radial maze at various age groups. This phenomenon which was found in all the aged groups, suggested a non-homogenous population in these groups.
Histology Distinct morphological changes were noted in various
Fig. 4. Sections through the hippocampal CA3 (A,B) and CA1 (C,D) subfields of 12- (A,B) and 17- (C,D) month-old Wistar rats. A and C represent intact morphology, whereas B and D represent examples of the degenerative changes such as those noted in half of the animals in these age groups. Note the shrinkage of cells and the intense staining for lipofuscin in B,D. PHG staining. A,B - - original magnification x400. C , D - original magnification xl60. Arrows (D) indicate the presence of tangle-like structures in CA1 pyramidal cells.
116 regions in the brain but especially in the hippocampus already by the age of 12 months. At this time interval degenerative changes were observed in about 50% of the subjects (n = 6) in regions CA3, CA1 and in the DG, but not always in both hemispheres. Similar changes were observed in the 17-month-old group. These included shrinkage of pyramidal cells at the CA3 and CA1 subfields, and nucleic heterochromasia. In sections stained with PHG, 'dark' cells represented cellular
accumulation of lipofuscin which was resistant to diastase pretreatment (Fig. 4). Such reactivity was not observed in specimens of the young animals. Pyknotic cells in the CA1 region had a tangle-like appearance (Fig. 4D). However, intact pyramidal cells were also encountered in damaged regions. The ratio between the two cell populations (intact vs. damaged) varied among areas, hemispheres and animals. A pattern of hemispheric asymmetry was found in reference to the morphological changes.
Fig. 5. Sections through the hippocampal CA3 (A) and CAt (B) subfields of 24-month-old Wistar rats. Note degenerative pyramidal cells in both regions. HE staining. ×160.
117 TABLE I
Age-related morphometric analysis of single-cell surface area in cholinergic regions of the Wistar rat brain The mean surface area of single cells was calculated from 120 measurements (20 cells in each animal, and 6 animals per age group).
Age (months)
Surface area (Itrn2 + S. E. M. )
3 12 17 24
CA3 subfield
CA1 subfield
Dentate gyrus
Medial habenula
271.2 205.4 190.13 150.7
162.3 121.05 121.9 126.7
97.0 68.7 80.05 71.7
85.83 63.7 72.16 72.9
+ 10.94 + 23,07** + 16.60 ___12.23"
+ + + +
17.35 9.30 8.28 4.07
+ + + +
2.94 2.45** 7.2 3.96
+ + + +
6.52 1.93"* 3.66 3.73
*P < 0.01; **P < 0.001.
In some of the animals, necrotic cells were observed at the inner layer of the DG. As already indicated above, in half of the animals (aged 12 and 17 months), the hippocampus appeared structurally intact and did not differ in its morphology from that noted in the young group. At the age of 24 months, the number of necrotic cells increased in all areas of the hippocampus, accompanied by atrophy and vacuolization (Fig. 5), a feature that was evident in 85% of the animals. The tangle-like appearance of pyramidal cells at the CA1 area became pronounced along with the presence of cells' residues (Fig. 5B). It should be however, stressed that even in the 24-month-old group, some of the animals revealed normal hippocampal morphology.
Morphometry Tables I and II represent the morphometric data related to cells (surface area and number) in regions CA1, CA3 and D G of the hippocampus as well as the surface area of single cells in the medial habenula. A significant age effect was found according to a one-way ANOVA. Specifically, a remarkable decrease was found from the 3 to the 12 month groups in surface area (Shifee contrasts; P < 0.001) for CA3, DG and medial habenula
and in number of CA3 cells (P < 0.001). In some regions an additional decrease was noted with increasing age. This phenomenon was particularly pronounced at the CA3 region, in which a significant decline was also noted between 17 and 24 months (P < 0.01). Statistical analysis concerning the CA1 subfield revealed a significant decrease in the total area occupied by the cells between 3 and 12 months (P < 0.025). The number of cells was almost constant with age, except for the CA3 subfield (Table II). Quantitative analysis of lipofuscin accumulation in the hippocampus showed an increase especially between 3 and 12 months of age (Fig. 6). The number of lipofuscin positive cells in CA3 subfield continued to rise until 24 months. As already mentioned above, a large variation was found between animals in each age group, except for the young ones. Noticeable accumulation of the pigment was observed only in approximately half of the animals, thus leading to an additional division of each age group into two sub-populations: 'impaired' versus 'non-impaired'. Animals showing at least 10 lipofuscin-positive hippocampal cells in a frame of 81,000~m 2 (the measured area) were considered as 'impaired'. Table III represents
LIPOFUSCIN ACCIlHULA1|OH
20 -2.
x.
W
T A B L E II
Or01 w
Age-related morphometric analysis of the number of cells in hippocampal subfields in the Wistar rat brain
x
N u m b e r of cells was counted in a frame of 32,000/~m 2 in at least 3 serial sections.
N
Age (months)
3 12 17 24 * * P < 0.001.
mcl3
!o
No. of cells (+ S.E.M.) CA3 subfield
CA1 subfield
Dentate gyrus
41 27.5 33.2 27.12
27.5 23.5 24.07 23.5
41.6 37.0 42.33 37.27
+ + + +
3.75 1.91"* 2.03 2.28
___1.52 + 1.08 + 0.7 + 0.88
_+ 5.07 + 2.02 + 4.55 + 2.85
0
5
10
15
20
25
A6E (Mt~T~I Fig. 6. Age related analysis of the n u m b e r of lipofuscin-positive cells in the hippocampus of Wistar rats.
118 T A B L E III
T A B L E IV
Quantitative analysis of lipofuscin-positive cells in the hippocampus of aging Wistar rats
Correlation between behavioral scores' (percent total errors) and hippocampal morphometric parameters in aged Wistar rats (Pearson's test)
Animals were divided into two sub-groups in each age group: 'impaired' and 'non-impaired'. A hippocampus that contained at least 10 lipofuscin-positive cells (per area examined) was considered as 'impaired'. Data represent the m e a n value of both hemispheres of 4 animals (in each subgroup).
Age (months)
No. of lipofuscin-positive cells _+S. E. M. CA3
CA1
DG
12 12"
1.18+0.22 12.75 _+ 7.18
2.57_+ 1.24 12.5 _+ 3.36
0.87_+0.21 26.33 _+ 7.32
17 17"
3.33 _+ 0.08 10.66+2.44
2.16 _+ 0.37 11.16_+6.45
1.5 _+ 1.06 13.16_+6.22
24 24*
6.06 + 1.42 22.58 + 4.5
2.87 + 0.71 11.08 + 6.15
3.62 + 1.0 8.58 + 4.6
P < 0.001
P < 0.01
P < 0.001
* Two-way A N O V A revealed significant differences between the two subgroups throughout in all areas examined.
the relevant data which were found to be significant throughout.
Morphobehavioral correlation The decline in the cognitive performance, which was noted already at the age of 12 months, was accompanied by morphological changes in the hippocampaus. The division of each age group into two sub-groups (according to their histology), indicated the existence of two different patterns of performance; a feature that was eminent mainly in the 12 and 17 months age groups, in which each of these two subgroups comprised half of the population (Fig.7).
N = 22 for all age groups; N = 9 for the 3- and 12-month-old age groups.
Morphometric parameters
r
Significance
CA3 - - s i n g l e cell area All age groups 3- and 12-month-old age groups
-0.544 -0.842
P < 0.01 P < 0.01
C A 1 - - area occupied by cells All age groups 3- and 12-month-old age groups
-0.362 -0.705
N.S. P < 0.05
Lipofuscin-positive cells 3- and 12-month-old age groups
0.669
P < 0.01
Correlative analysis of the morphometric data and behavioral findings as related to the working memory deficiency (expressed by total errors) revealed a high degree of relationship. The correlation data are summarized in Table IV. As can be seen, a significant correlation was found especially between cellular area in CA3 subfield and the percent of total errors r = -0.544; P < 0.01 for all age groups (Fig. 8). Higher correlation was found regarding these parameters for the 3- and 12month-old groups (r = -0.842; P < 0.01) for the individual scores and r = -0.98; P < 0.02 for the mean group scores). For lipofuscin, a correlation was found between the number of positive cells and working memory deficiency (total errors) only at the age groups of 3 and 12 months (r = 0.669; P < 0.01).
12 ~ T ~
N O ~ V I O B A L C(X~LATION Q Ml'lll hllI~Ompus
50
•
e--o.544
llllnrld
5O
70
a~
|
40
60
30 2O
~
5o
~
4o
~-
ao 2O
tO
tO t
2 3 4 TRAINII~ 9ESSIONS (WEEKLYBLOCKS)
5
Fig. 7. Age-related changes in working m e m o r y deficiency (expressed by total errors + S . E . M . ) as recorded in 12-month-old rats which were divided into two subgroups according to morphological criteria: the a m o u n t of lipofuscin in hippocampal cells.
50
100
t50 200 250 300 350 CA3 -SINGLE CELL AREA (tlIL)
400
450
Fig. 8. Linear correlation between the area of a single cell in CA3 region and working m e m o r y deficiency (total errors) in Wistar rats in various age groups.
119 DISCUSSION Age-related performance decrements were observed already by 12 months of age, an age group which is usually omitted from aging studies. This behavioral change was highly significant when compared to young animals. The 12-month-old group comprised of a heterogeneous population as was deduced from the large values of standard deviation in all the measured behavioral parameters. In another study, using a different behavioral paradigm (passive avoidance), a 30-40% decline was reported in 12-month-old mice, a decline which correlated well with the decrease in acetylcholine synthesis 9. The findings of the present study appear consistent with a previous study which showed an age-related impairment in both working memory and reference memory tasks 1. Most of the animals in all the age groups improved their performance following training, a fact that was statistically significant with regard to correct choices and the time needed for the completion of a session. The parameter of total errors, however, did not improve with training, because of a learning deficiency in the aged groups. The sharp decline in performance, as observed in rats at 12 months of age, was accompanied by changes in brain morphology among which the hippocampus was found to be the most affected region. Fifty percent of all animals in this age group revealed an apparent normal hippocampal structure, a feature that was associated with an unimpaired performance. An analysis of the middleaged group as two subgroups according to their morphobehavioral scores, revealed the existence of two different populations. One of these subpopulations appeared relatively unaffected, in which the animals performed as well as young animals. A similar division of aged rats to 'impaired' vs 'non-impaired' subgroups was suggested by De Toledo-Morrell et al. 5 and Landfield et al. 13. In both studies the hippocampal synapses of 'impaired' aged animals were shown to exhibit pronounced electrophysiological and morphological aberrations found neither in young nor in 'non-impaired' aged rats. The present morphological study supports previous data indicating the sensitivity of the hippocampus to age-related degeneration. Not all the hippocampal subfields showed a similar pattern of age-related degeneration. Despite the fact that parameters, such as number and area of cells, decreased by the age of 12 months, the REFERENCES 1 Barnes, C.A., Nadel, L. and Honig, W.K., Spatial memory deficit in senescent rats, Can J. Psychol., 34 (1980) 29-39. 2 Brizzee, K.R. and Ordy, J.M., Age pigments, cell loss and hippocampal function, Mech. Aging Dev., 9 (1979) 143-162.
best correlation between structural changes and working memory performance was found with regard to surface area of pyramidal cells in the CA3 region. This observation suggests a possible relationship between a degeneration of cholinergic components, especially at CA3 subfields, and memory deficiency. Similar observations in reference to CA3 cells was reported for mice 11. Handelmann and Olton TM have arrived to the same conclusion following kainic acid lesion of CA3, thus emphasizing the importance of CA3 pyramidal cells for hippocampal circuitry and as a junction point between cortical and subcortical structures. Another marker for age-related changes in nerve cells is lipofuscin accumulation. This 'aging pigment' is believed to represent an oxidative degeneration process and therefore, might indicate neuronal dysfunction. In rats, the accumulation of this pigment increased with age, starting at the age of 12 months. It should be emphasized that alike other morphometric parameters, the appearance and deposition of the pigment varied among animals with high correlation to the level of performance. At the old age group (24 months of age), the morphobehavioral correlations was found to be much lower, presumably due to the fact that many other non-specific age-related processes were taking place. Cognitive decrements at this age might have stemmed from additional non-central physiological injuries. It is therefore proposed that mature and middle-aged rats are the group of choice for morphobehavioral correlation studies. In our opinion, at this age group, behavioral decrements stem mostly from cognitive deficiency, while with older rats impaired physical capacity may affect behavioral results. Light microscopy studies, such as the one hereby reported, yield data on a large number of cells, while ultrastructural EM analysis may provide further information on subcellular elements involved. In conclusion, this study provides further support to the notion that the hippocampus plays a key role in the neurobiology of dementia and that the degree of cognitive impairments is highly correlated with degenerative structural changes in this region of the brain, in particular with changes in the CA3 pyramidal cells. Acknowledgements. This study was supported in part by a grant from the United States-Israel Binational Science Foundation. The authors wish to thank Ms. R. Sahar for her skilled technical assistance.
3 Coleman, P.O. and Hood, D.G., Review. Neuron numbers and dendritic extent in normal aging and Alzheimer's disease, Neurobiol. Aging, 8 (1987) 521-545. 4 Davis, H.P., Tribuna, J., Pulsinelli, W.A. and Volpe, B.T., Reference and working memory of rats following hippocampal damage induced by transient forebrain ischemia, Physiol. Be-
120 hay., 37 (1986) 387-392. 5 De-Toledo-Morrell, L., Geinisman, Y. and Morrell, E, Review: Age-dependent alterations in hippocampal synaptic plasticity: relation to memory disorders, Neurobiol. Aging, 9 (1988) 581-590. 6 Fisher, A. and Hanin, I., Mini review: choline analogues as potential tools in developing selective animal models of central cholinergic hypofunction, Life Sci., 27 (1980) 1615-1634. 7 Freund, G., The effects of chronic alcohol and vitamin E consumption on aging pigments and learning performance in mice, Life Sci., 24 (1979) 145-152. 8 Geinisman, Y., Toledo-Morrell, L. and Morrell, F., Aged rats need a preserved complement of perforated axospinous synapses per hippocampal neuron to maintain good spatial memory, Brain Research, 398 (1986) 266-275. 9 Gibson, G.E. and Jenden, D.J., Brain acetylcholine synthesis declines with senescence, Science, 213 (198t) 674-676. 10 Jucker, M., Oettinger, R. and Battig, K., Age-related changes in working and reference memory performance and locomotor activity in the Wistar rat, Behav. Neural Biol., 50 (1986) 24-36. 11 Kadar, T., Silbermann, M. and Levy, A., Age-related changes in cholinergic components within the central nervous system (hippocampus, NBM and habenula) of CWI female mice, Mech.
Aging Dev.. 47 (1989) pp. 133-144. 12 Kuttin, E.S. and Beemer, A.M., PHG-staln for histology, frozen sections and cytology. In E. Levy (Ed.), .ldvances in Pathology. Vol. 1, 1982, pp. 115-116. 13 Landfield, P.W., Review: hippocampal ncurobiological mechanisms of age-related memory dysfunction, Neurobiol. Aging, 9 (1988) 571-579. 14 Olton, D., Walker, .I.A. and Gage, F.H., Flippocampal connections and spatial discrimination, Brain Research, 139 (1978) 295-308. 15 Olton, D.S., Memory functions and the hippocampus. In W. Seifert (Ed.), Neurobiology of the Hippocampus, Academic Press, London, 1983, pp. 335-373. 16 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic (2)ordinates, Academic Press, 1982. 17 Sinden, J.D., Rawlins, J.N.P., Gray, J.A. and Jarrard, L.E., Selective cytotoxic lesions of the hippocampal formation and DRL performance in rats, Behav. Neurosci., 100 (1986) 320329. 18 Handelmann, G.E. and Olton, D.S., Spatial memory following damage to hippocampal CA3 pyramidal cells with kainic acid: irr,pairment and recovery with preoperative training, Brain Research, 217 (1981) 41-58.