- Email: [email protected]
Spatial performance correlates with in vitro potentiation in young and aged Fischer 344 rats
Brain Research, 554 (1991) 1-9 © 1991 Elsevier Science Publishers B.V. 0006-8993/91/$03.50 ADONIS 000689939116770V
BRES 16770
Research Reports
Spatial performance correlates with in vitro potentiation in young and aged Fischer 344 rats David L. Deupree, Dennis A. Turner and Carter L. Watters Department of Neurosurgery and Research Services, Durham VAMC and Division of Neurosurgery, Department of Surgery, Duke University Medical Center, Durham, NC 27710 (U.S.A.)
(Accepted 5 February 1991) Key words: Potentiation; Aging; Hippocampus; CA1 extracellular field potential; Spatial memory
Young adult (2-4 months old) and aged (24-26 months old) Fischer 344 (F344) rats were trained for spatial behavior (locating a hidden escape platform) in a circular water maze. The aged rats showed deficits in both the acquisition and retention of the learned response. Following the behavioral training, hippocampal slices from the rats were prepared. Potentiation of CA1 extracellular, somatic field potentials was studied in vitro following either a short stimulus train (4 pulses) or a longer train (50 pulses). Slices from the aged rats showed less short-t,~,'m potentiation (124.8 _+ 4.9% baseline, mean _+ S.E.M.) at 1 min following the short train in comparison to slices from the young rats (151.8 _+ 7.5%, P < 0.05). However, following the longer train, no differences were found between the groups in the degree of either short-term (measured at 1 min after stimulation) or long-term potentiation (measured at 60 min). The amount of potentiation seen at various time points after either train correlated with the behavioral measure of retention. These results indicate that F344 rats exhibit age-related behavioral deficits, and age-related synaptic potentiation deficits in response to short stimulation trains. The correlation between the degree of potentiation (both short-term and long-term) and retention of a behavioral task adds strength to the hypothesis that potentiation mechanisms may underlie memory processes. INTRODUCTION Although not all individuals exhibit adverse aging effects, when they do occur, aging can produce deterioration of brain, behavior and cognitive functioning 13'19. The hippocampus is one of the initial brain regions adversely effected by aging. The hippocampus in particular appears to become partially deafferented through the process of aging, with deterioration and loss of afferents over time 8'17'1s. This deafferentiation and isolation is even more severe with Alzheimer's Disease 11,22, and may be partially responsible for the early cognitive decline seen in Alzheimer's patients. In rats, integrity of the hippocampus is important in the acquisition and retention of several learned behaviors 21, among these being so-called 'spatial behavior '36'37. Lesions of the entorhinal cortex or the medial septal nucleus 31'38"44, structures which provide afferent input to the hippocampus and which also exhibit spontaneous degeneration with aging and Alzheimer's Disease, lead to deficits of spatial behavior in rats. In addition, aged rats have been found to display significant deficits in spatial acquisition and retention 2'3A4-16,28,29,4°. Potentiation, or enhancement, of hippocampal syn-
apses has been proposed as a neurophysiological model for memory 4'7'34. Since aged rats often exhibit m e m o r y deficits, one might expect that synaptic potentiation in the hippocampus of aged rats would also be deficient. However, the results relating hippocampal potentiation and aging are mixed. A number of studies have found age-related deficits of population spike increases in the CA1 region following induction of long-term potentiation (LTP), and during frequency potentiation 25'26'42. However, other investigators could find no age effect in population spike amplitude of dentate granule or CA1 neurons following LTP induction a'lg'2a. Some of these same groups did find that the decay of dentate granule cell population excitatory postsynaptic potential (EPSP) slope following the induction of LTP was faster in aged rats a'3"14. In addition, Barnes and McNaughton 3 found that when using repeated stimulus trains to induce maximal LTP over a period of time, young rats reached their maximum faster than aged rats. The discrepancies noted here could be due to a variety of methodological and technical differences between studies, including varying stimulation parameters and physiological measurements. In order to strengthen the position for LTP (and
Correspondence: D. Deupree, Neurosurgery, Box 3807, Duke University Medical Center, Durham, NC 27710, U.S.A.
p o t e n t i a t i o n / e n h a n c e m e n t in general) as a memory model, it would be important to show within subjects correlations between LTP and behavioral measures of memory. Further, manipulation of one variable should influence the outcome of the other. Morris and Baker 34 review research strategies intended to strengthen the argument for potentiation as a model for memory. The strongest methodologies were argued to be correlational and manipulation strategies 34. Morris and colleagues 33 showed that chronic infusions of D-2-amino-5-phospho-
behavioral deficits with differences in potentiation parameters. This was accomplished by first training both young adult and aged Fischer 344 (F344) rats in a water maze for spatial behavior, and then inducing potentiation in the hippocampal CA1 region in vitro using either a brief stimulus train (4 pulses) or a longer train (50 pulses). The measures of spatial learning and memory were then compared with the physiological measures of potentiation, both between groups and on an individual basis across age groups.
novalerate (AP5) into the lateral ventricles produced spatial learning deficits in rats during water maze training. AP5 is a N-methyl-D-aspartate receptor antagonist which is known to block the induction of LTP in the hippocampus m'33. Further experiments 35 showed correlations between AP5 tissue content within the hippocampus during chronic intraventricular infusion, decreases in the degree of dentate LTP induction and water maze performance in rats. Induction of dentate LTP in vivo prior to behavioral training has been shown to inhibit spatial maze performance in rats 9'3° or to enhance behavioral acquisition of classical conditioning in rabbits 5. It has been argued that the opposing influence produced by pre-training LTP induction in these two circumstances may be due to a differential influence of heightened hippocampal excitability upon acquisition of the two responses 6. Regardless, these studies demonstrate correlations between LTP and behavior 33"35, and that m a n i p u l a t i o n of LTP can influence resulting behavioral acquisition5"9'3°. If one accepts the argument that aging can lead to spontaneous hippocampal lesions, then it could be argued that aging studies fall u n d e r the manipulation strategy (age-induced hippocampal lesions being a manipulation). In adding an age variable to the correlational strategy, one combines correlation and manipulation strategies. There are two studies which have used the correlation approach to investigate possible relationships between aging, various aspects of potentiation and behavior. Landfield 23 found that synaptic depression seen in the later stages of frequency potentiation could distinguish between aged rats that exhibited retention deficits (of avoidance conditioning) and aged rats that did not exhibit deficits. Barnes and McNaughton 3 found that the acquisition and decay of EPSP slope during and after LTP induction corresponded to acquisition and retention behavior of young and aged rats in a radial maze. Thus it appears that there may be correlations between behavioral measures of learning and memory in aged rats and physiological measures thought to reflect memory processes. The purpose of the present study was to extend the results of these earlier attempts to correlate age-related
MATERIALS AND METHODS Water maze procedures
A water maze was selected as the tool for behavioral assessment of the rats because, as noted above, measurements obtained from use of this type of maze can differentiate between young adult and aged rats 15'16'28'4°, as well as being sensitive to hippocampal integrity36. Subjects were young and aged male F344 rats. The aged rats (n = 20) were obtained from the National Institute on Aging, were 24-26 months old, and healthy (free of respiratory problems and no visible tumors) at the time of behavioral testing and subsequent in vitro experimentation. Young rats (n = 15) were obtained from Charles Rivers, and were 2-4 months old. Although the different age groups of rats were obtained from different suppliers, both suppliers obtained their breeding stock from the same source (National Institute of Health), and thus the rats were from identical genetic sources. Rats were trained for so-called spatial memory using a Morris water maze32. Briefly, rats were trained to locate an escape platform within a circular water maze. The maze was 1.5 meters in diameter by 0.6 meters in height. The maze was filled with water (approximately 22 °C) mixed with opaque food dye, in order to hide the location of the escape platform from the rats. The escape platform was always located just below the water surface (1-2 cm). Located within the room were several stationary visual distal cues such as the overhead light rack and rat housing racks along two of the walls. Also there were significant non-visual distal cues which also remained stationary (i.e. the sounds and smells generated by rats within the housing racks). Thus there were several distal sensory cues within the testing environment which the rats could utilize in locating the hidden platform. Each rat was given 8 training (acquisition) trials per day, over 4 consecutive days (32 total trials). Each trial consisted of gently placing the rat into the maze, facing the wail, at one of 4 entry sites. The entry sites were located in a repeatable manner by dividing the maze into 4 equal quadrants. During each day, the rat would receive two trials from each entry point, randomly determined. The escape platform was located in a constant position (placed in the center of one quadrant) for each rat during all training trials. During each trial the rat was allowed 60 s to find the platform. Once the rat had located the platform, it was allowed a brief rest period (30-60 s) on the platform before a new trial was begun. Escape latency (in s) for each trial was measured. If the rat had not located the platform with 60 s, it received a latency score of 60 s, was gently placed on the platform and allowed to rest there briefly before a new trial began. Rats were dried off with towels between every 2-3 trials. Each rat was observed carefully during each trial. If a rat exhibited any difficulty in swimming about the maze, it was removed from the maze, and discarded from the study. Four aged rats were found to exhibit swimming abnormalities and these animals were not included in the study, thus the behavioral results were obtained from 16 rats from this group. Following the completion of acquisition training, each rat received a retention trial 24 h later. The retention trial was identical to the training trials, except that the escape platform was removed
from the maze. The amount of time the rat spent swimming in the quadrant where the platform had been located during the previous training trials (training quadrant) was measured during the 60-s retention trial. Time spent in the training quadrant during the retention trial was referred to as the retention score.
Hippocampal slice preparation and recording procedures All physiological experiments were performed within two weeks of the completion of behavioral testing. Rats were sacrificed using Halothane, the hippocampus from each hemisphere was dissected out, and transverse mid-hippocampal slices 500 /zm thick were obtained using a chopping type microtome 43. Slices were allowed to incubate in a physiological bath for at least 90 min before being transferred to a gas interface type slice chamber where somatic hippocampal CA1 field potentials were recorded. The temperature of the bath within the chamber was maintained at 33 + 1 °C. Concentrations of the bath were (in mM): D-glucose 10, KCI 3, NaCI 124, NaH2PO 4 H 2 0 1.2, NaHCO 3 26, MgSO 4 7H20 2, CaCl 2 6H20 2.4. The bath and slices were exposed to a mixture of 5% CO 2 and 95% 0 2 gas. Recording electrodes were glass micropipettes (5-10 MI2) filled with 2 M NaCl. Stimulation was delivered with bipolar electrodes placed in the stratum radiatum of the CA1 field. Test stimuli were monophasic constant current pulses, 50 /~s in duration. Estimates of maximal population spike amplitudes were determined by increasing the duration of pulse, while keeping the current intensity at 5.0 mA, until there was no further increase in population spike size. Potentials were amplified by a factor of 10, filtered at 0-10 kHz, and recordings were stored on computer disks for off-line analysis. Criteria for acceptable field potentials were population spikes from the pyramidal cell layer, with maximal amplitude of at least 5 inV. Population spike amplitude was measured from the initial positive peak of the extracellular EPSP to the subsequent negative peak of the population spike (see Fig. 1). Once an acceptable field potential was obtained, 1-2 h were allowed for stabilization, during which potentials were generated once per min. Following stabilization, current intensity was adjusted to evoke a population spike 50% of maximal amplitude. The current range used to evoke half amplitude population spikes was 3.2-5.5 mA at 50 /~s. After ensuring that the potentials were stable following current adjustment, baseline potentials were recorded for 4 rain, then either a short (4 pulses at 100 Hz) or long (50 pulses at 100 Hz) stimulus train was delivered through4he stimulating electrodes. Post-train poten-
A
tials were recorded for 10 min following the 4-pulse train, and for 1 h following the 50-pulse train. Post-stimulus train population spike amplitudes were converted to a percentage of the mean population spike amplitude of the 4 baseline potentials. Population spikes were chosen for analysis because the changes in population spike amplitudes as a result of potentiating stimuli are more robust than changes seen in synaptic potentials. Therefore measurement of population spike changes was chosen to maximize the chances of observing a potentiating effect following the short stimulus train (4 pulses).
Data analysis Analysis of data (both behavioral and physiological) was performed using multivariate analysis of variance, with post-hoc analysis being performed using Tukey's test for differences between group means. Correlations were performed using Pearson correlation coefficients. Behavioral variables included average escape latencies per block of 8 training trials and retention score (time spent in the training quadrant during the retention trial). Potentiation measures included for correlation with behavioral retention scores were percent baseline values obtained at various time points following the presentation of the stimulus train. Frequency data were analyzed using a 2 x 2 chi square table.
RESULTS
Behavioral Average escape latencies for each training day (block of 8 trials) are summarized in Fig. 2. Over the 4 training days the aged rats as a group showed no reduction in escape latencies (/'3,45 -----0.17, n.s.), while the young rats did show reductions across days (F3,42 = 23.20, P < 0.01). Overall there was a significant age effect (F1,29 = 23.48, P < 0.01). Post-hoc analysis between groups found the escape latencies of the two groups were not significantly different from each other during the first two training days. However, the escape latencies were found
Young rats (n = 15) BB
C
~ e d rots (n = IS) r-~
~' X +
6O
40
B
D
(J _~
20
i1/1i1i BLOCKS OF EIGHTTRIALS
Fig. 1. Examples of CA1 field potentials obtained from hippocampal slices of aged and young rats, both before and after presentation of the 4-pulse stimulus train. A and B show potentials before (A) and 1 min after (B) the train of a slice from a young rat. Amplitudes of population spikes were measured as the absolute value (in mV) from the initial positive peak of the extracellular EPSP to the adjoining negative peak of the population spike. C and D show representative potentials of a slice from an aged rat at the same time points. The vertical bar for traces A and B is 10.0 mV and 6.0 mV for C and D. The horizontal bar is 30 ms for all traces in the figure.
Fig. 2. Summary of escape latencies (mean + S.E.M.) across blocks of training trials. The latencies were not significantly different between age groups during the first two blocks of trials. However, during the last two blocks (days 3 and 4) escape latencies of the aged rats were found significantly greater than the latencies of the young rats. These results show the average latencies of the young rats progressively declining across the 4 blocks of trials (indicating acquisition of the behavioral response), while the average latencies of the aged rats do not show such a trend across days. **P < 0.01 difference between groups.
to be significantly different on days 3 and 4 ( P < 0.01 for both comparisons). Fig. 3 summarizes retention scores for the two groups. The escape platform had been located in q u a d r a n t 2 during the training trials, thus time spent in that q u a d r a n t during the retention trial served as the retention score. The young rats were found to have significantly greater retention scores (30.5 + 3.0 s, mean + S . E . M . ) than the aged rats (17.6 + 1.5 s) (F1.29 = 14.99, P < 0.01). Also shown in Fig. 3 is time spent in each of the o t h e r quadrants during the retention trial, whereas the d o t t e d line represents time spent in any quadrant based upon chance (15 s). On average the aged rats were spending time in all quadrants around chance levels during the retention trial, while the young rats spent the majority of time within the training quadrant.
In vitro potentiation A c c e p t a b l e field potential recordings were obtained from 14 of the 15 young rats from which behavioral data were collected (93% frequency of acceptance), and of the 16 aged rats, acceptable potentials were obtained from 13 rats (81%). These acceptance frequencies were not found to be significantly different between the groups (X2(1) = 0.22, n.s.). The two age groups did not differ in current intensity (either in overall range or average intensity) n e e d e d to evoke a half maximal population spike. Slices from 10 young rats and 7 aged rats received the short (4 pulses at 100 Hz) stimulus train. Slices from 10 young rats and 6 aged rats received the longer stimulus train (50
pulses at 100 Hz). T h e r e were 6 young rats from which it was possible to obtain two viable slices, using one slice for each stimulus train experiment. T h e r e were no aged rats which provided a slice for each experiment. A few slices did not show potentiation following the 4-pulse train (operationally defined as an increase of greater than 10% above the average baseline value when m e a s u r e d at 1 min following the train). Thus, data from the slices of 1 young rat (90% potentiation frequency) and 2 aged rats (71%) were discarded for not showing sufficient potentiation following the 4-pulse train. These frequencies for potentiation occurrence between the two groups were not significantly different (X2(1) = 0.12, n.s.). All slices from both age groups showed p o t e n t i a t i o n following the 50-pulse train. Fig. 4 summarizes the effects of the 4-pulse train on population spike amplitudes from slices which showed potentiation (more than a 10% increase over baseline levels at 1 min). A l t h o u g h by definition slices from both age groups showed potentiation, it can be seen that, at least during the first few min following the train, the slices from young rats on average showed greater potentiation than the slices from aged rats. The initial increase in population spike amplitudes seen during the first few min after stimulation will be referred to here as short-term potentiation (STP). A f t e r about 5 rain, potentiation levels of both groups of slices showed a tendency to stabilize (long-term potentiation, LTP). In o r d e r to avoid unwieldy levels of the r e p e a t e d measures variable (time), statistical analysis was restricted to percent baseline values at - 1 (last baseline measure), 1,
Young Rats (n = 15) BIB Aged Rats (n = 16) [S:] Slices from young rots (n = 9) o - - e Slices from aged rots (n = 5) o - - o
30 +
160
'
-L 20
8
+
-
10
i
1
3
"i.... I" 4
1
2
}'L+
.
,?, 3
1O0
4
QUADRANTS
Fig. 3. Summary of time (mean + S.E.M.) spent in each of the 4 maze quadrants during the retention trial. During the training trials, the escape platform was located in quadrant 2. The dotted line represents chance level (15 s) for amount time spent in any particular quadrant during the trial. The young rats on average were spending time in quadrant 2 that was well above chance levels, while the aged rats were close to chance. The retention scores (time spent in quadrant 2) of the aged rats were significantly shorter than the scores of the young rats. One can see that the young rats by far spent the majority of the retention test in quadrant 2. Although the aged rats, on average, did spend more time in quadrant 2 than any of the other 3 quadrants during the trial, this was not a significant trend. **P < 0.01 difference between groups.
=o ^
1
5
10
MINUTES FOLLOWING STIMULUS TRAIN (4 pulses at 100 Hz)
Fig. 4. Summary of change in percent baseline values (mean + S.E.M.) of CA1 population spikes following a 4-pulse at 100 Hz stimulus train. Included are baseline values obtained during the 4 min prior to train presentation (arrow on time axis, min 0). Slices from rats of both age groups showed increases over baseline levels following the 4 pulse train, however, the slices from young rats showed greater increases. Between groups analysis was performed at 1, 5 and 10 min following the train. A significant difference between potentiation values was found at 1 min following the train, but not at 5 or 10 min. *P < 0.05 difference between groups.
5 and 10 min following the train. Using those time points only, it was found that overall the potentiation values of the slices from young rats were significantly greater (F1,12 = 6.12, P < 0.05). Post-hoc analysis also showed that potentiation values of slices from young rats were significantly greater than those of slices from aged rats at 1 min following the train (P < 0.05). As can be seen in Fig. 4, there was no significant difference between potentiation values at either 5 or 10 rain following the train. These results indicate that the slices from young rats showed greater initial potentiation following the 4-pulse train (STP), but that the two groups did not differ in terms of later phases of potentiation (LTP) following the 4-pulse train. At 1 min following the train the slices from the young rats showed potentiation values of 151.8 + 7.5 (mean + S.E.M.) percent baseline, while the slices from the aged rats showed potentiation values of 124.8 + 4.9 percent baseline. At 10 min after the train, the potentiation values for the slices from young rats were 131.4 + 5.7, and the values for slices from aged rats were 112.8 + 6.1. Thus slices from both age groups showed initial potentiation (STP) which decreased in magnitude over a few minutes, but were still slightly above original baseline values when measured at 10 min (LTP). Recordings from several of these slices were carried out for periods up to 1 h. In these slices, the LTP values on average did not return to original baseline values (remaining just slightly above as was the case for the 10 min values), in slices from either age group. Fig. 5 summarizes effects of the 50-pulse train on population spike amplitudes. The initial increase in population spike amplitudes (first 10 min) seen following the 50-pulse train is again referred to as STP, while the potentiation during the next 50 min (min 10-60) is referred to as LTP. The left side of Fig. 5 summarizes the data for STP (min 1-10, at 1-min intervals), while the right side summarizes the data for LTP (min 10-60, at 10-min intervals). One can see that there were no significant differences in potentiation values at any time point following the 50-pulse train between slices from young rats (n = 10 slices from 10 rats) and aged rats (n = 6 slices from 6 rats). As can be seen in Fig. 5, the potentiation values of both groups of slices, on average, appeared to decline at similar rates during the first few min following the 50-pulse train (STP), then remain relatively stable during the remaining time (LTP). At 1 rain following the train the potentiation values for slices from young rats were 204.5 + 9.4 percent baseline, and 196.5 + 3.7 for the slices from aged rats. These values were seen to decrease to 174.3 + 7.7 (slices from young rats) and 175.7 + 7.4 (slices from aged rats) after 10 min. At 60 min the potentiation values for both groups had
decreased slightly (but not significantly) from their 10 min values (slices from young rats = 165.7 + 14.6%, and slices from aged rats = 159.0 + 19.6%). Correlation o f behavior with potentiation Correlation coefficients between retention scores and in vitro potentiation measures obtained from the slices at 1, 5 and 10 min following the 4-pulse train were calculated. The correlations for combined data collapsing across the two age groups (n = 14; aged rats = 5 and young rats = 9) showed significant positive correlations between retention scores and potentiation values at 1
min: r13 = 0.63 (P < 0.05) and at 5 min: r13 ---- 0 . 5 7 ( P < 0.05), but not at 10 min: r13 = 0 . 4 0 ( n . s . ) . Fig. 6A shows the scatter plot of the individual data points for the correlation of retention scores with potentiation at 1 min following the 4-pulse train (STP). One can see the relationship between the two variables was also positive within each of the age groups, although not significant. The correlation between retention scores and STP for young rats was: r s = 0.31 (n.s.) and for the aged rats: r 4 = 0.84 (n.s.).
Correlation coefficients between retention scores and in vitro measures of potentiation following the 50-pulse train were also calculated. These data were also from an analysis that collapsed across age groups (n = 16; aged rats = 6 and young rats = 10). Fig. 6 B - D summarizes 3 correlations. Retention scores were found to show a
Slices from young rate (n = 10) e ~ e Slices from aged rats (n == 6) o ~ o E J +
200
8 E z" '
150
.J LIJ
I--
z
STP
100
LTP
hJ
o_
,
. . . .
"
. . . .
~,
....
io'
,
10
•
,
,
,
.
,
60
MINUTES FOLLOWINGSTIMULUSTRAIN (50 pulses at 100 Hz) Fig. 5. Summary of change in percent baseline values (means _+
S.E.M.) of CA1 population spikes following a 50-pulse at 100 Hz stimulus train. The left portion of the figure shows baseline values during the 4 min prior to train presentation (arrow on time axis, min 0), as well as the average increases above baseline during the first 10 min following the train (STP). The right portion of the figure shows percent baseline averages for both groups during the 10-60 min time period following the train, at 10-min intervals (LTP). There were no significant differences between the two groups at any time period analyzed.
A
C
Young rots (n - 9) •
Young rats (n = 10) • Aged rots (n = 6) O
Aged rots (n - 5) O
•
...--
z
200
g
15o
~
loo
Z _
•
O 0
~
..
.._ •:.~-'~7., • • •
•
.~
O •
5O O 100
1o
;o
~o
do
REI-tNnON SCORE ( N c )
B
;o
do
RETENTION SCORE ( ~ ¢ )
D
Young rotl (n - I 0 ) • Agod mr, (n - e) o
Young rotl (n - 10) • Aged ~ , (. - e) o
300
i
lOO o
:oo
.
~o
~o
. . . . . . . . . . •
.
0
1o
~o
do
RETENTION SCORE (sec)
do
RETENTION SCORE (sac)
Fig. 6. Scatter plots of correlations between behavioral retention scores and in vitro potentiation measures. Within each scatter plot, individual data points from young rats (solid dots) and aged (open dots) rats are displayed. Also displayed are regression lines representing correlations for the data collapsed across groups (solid line), as well as for correlations within each group (regression line for young rats is dashed line, and regression line for aged rats is dotted line). See text for details concerning the various correlation coefficients. A: the scatter plot for the correlation between retention scores and potentiation values measures at 1 rain following the 4-pulse train (STP). The overall correlation across groups was positive. In addition, one can see that individually both groups showed a positive relationship between the two variables. B: the scatter plot of the correlation between retention scores and potentiation values measured at 1 min following the 50-pulse train (STP). The correlation between the two variables across groups was positive. One can see, however, that although the young rats maintained this positive relationship, the aged rats did not, and in fact show a slightly negative correlation between the variables, C: the scatter plot of retention scores correlated with potentiation values measured at 60 min following the 50-pulse train (LTP). Here also the overall relation between the variables was positive, and the relationship between these variables within each group remained positive. D: the scatter plot of the correlation of retention scores with a measure of LTP persistence (potentiation measured at 60 min as a percent of the potentiation measured at 10 min) following the 50-pulse train. Once again, the overall relationship between the variables was positive, as was the relationship within the two groups.
significant positive c o r r e l a t i o n with p o t e n t i a t i o n values at b o t h 1 min: rl5 = 0.50 ( P < 0.05) and 60 min: r15 = 0.51 ( P < 0.05) f o l l o w i n g t h e train. T h u s , o v e r a l l the r e t e n t i o n scores f r o m b o t h g r o u p s o f rats s h o w e d positive correlations with m e a s u r e s of S T P and LTP following the 50-pulse train. Fig. 6B shows the scatter plot of individual d a t a p o i n t s for the c o r r e l a t i o n b e t w e e n r e t e n t i o n scores and p o t e n t i a t i o n at 1 m i n f o l l o w i n g the 50-pulse train (STP). O n e can see a dissociation b e t w e e n the two age g r o u p s in h o w the two variables c o r r e l a t e d . T h e correlation b e t w e e n the two variables for y o u n g rats was still
positive, a l t h o u g h not significant; r 9 = 0.52 (n.s.), while the c o r r e l a t i o n for the a g e d rats was slightly n e g a t i v e , a l t h o u g h also n o t significant; r 5 = - 0 . 0 9 (n.s.). Fig. 6C shows the scatter plot for t h e c o r r e l a t i o n b e t w e e n retention
scores
and
potentiation
values
at
60 min
f o l l o w i n g the 50-pulse train (LTP). In this p l o t b o t h age groups retain a p o s i t i v e r e l a t i o n b e t w e e n the two variables, a l t h o u g h again b o t h r e l a t i o n s h i p s w e r e n o t significant. F o r y o u n g rats; r 9 = 0.55 (n.s.), and for a g e d rats; r 5 = 0.76 (n.s.). C o r r e l a t i o n coefficients w e r e also calculated b e t w e e n r e t e n t i o n scores and p o t e n t i a t i o n
7 values at 1 min and all 10 min increments following the train; however, significant correlations were found only at 1 and 60 min. Previous studies have found that the decay of LTP in vivo was faster in aged r a t s 2'3'14 and that the rate of decay in LTP was similar to the rate of decay in behavioral retention of spatial memory in rats 3. It was recognized that the decay in LTP from the studies cited a b o v e 2'3'14 was measured in terms of changes in field EPSP slope, and that population spike LTP can be decoupled from EPSP LTP (implying the possibility of different mechanisms), thus there was no a priori reason for expecting similar results during the present study. However, a measure of population spike change following the 50pulse train was compared with retention latencies. The estimate of LTP change (persistence) chosen was the comparison of potentiation at 60 min with potentiation at 10 min (min 60 as a percentage of min 10) following the 50-pulse train. This measure thus focused on the time period summarized in the right portion of Fig. 5 (LTP), avoiding contamination from the STP phase. Across the two groups LTP persistence showed a significant positive correlation with retention scores; r15 = 0.53 (P < 0.05). Fig. 6D shows the scatter plot of individual data points for correlation of retention scores with LTP persistence (potentiation at 60 rain as a percentage of potentiation at 10 min). Here again, the individual groups retain positive relationships between the two variables. For the young rats; r 9 = 0.54 (n.s.), while for the aged rats; r 5 = 0.82 (P < 0.05). DISCUSSION Behavior
The behavioral results show severe deficits (both in acquisition and retention) by the aged rats. The behavioral results are consistent with a variety of past studies 2' 3.14-16,28,29,4o, which demonstrate deficits in behavioral measures of 'spatial' learning and memory in aged rats. The severity of the deficit exhibited by the aged rats is troublesome, in that on average, the aged rats in this study showed no signs of acquiring the spatial task (a flat learning curve). Several reports regarding aged rats have found a learning curve during spatial training 2'3'14'15'4°. However, the present behavioral results are consistent with other studies which looked at spatial performance in aged F344 rats 2s'29, which could find no evidence for behavioral acquisition across trials (a flat learning curve). It is certainly possible that the severe behavioral results seen in the present study may have been due to other factors besides deficit learning abilities. For example, visual deficits cannot be ruled out as an alternative explanation for the poor behavioral acquisition and
retention exhibited by the aged rats. Recent data show a correlation between degree of retinal atrophy and water maze performance by aged rats 41. In order to test for visual deficits for local, intra-maze cues, a group of 24-month-old F344 rats were tested in the water maze in this lab while using a visual platform. These rats did exhibit a steady decrease in escape latencies over training trials (data not shown). Unfortunately this only demonstrates that these aged rats could see, and utilize, cues within the maze itself. The possibility still exists that the aged rats might not have been able to appreciate, or to utilize the many other sensory distal (extra-maze) cues which were present, which could account for the acquisition deficit seen during this study. Thus, the behavioral results presented here as they stand cannot rule out alternative explanations for the spatial performance deficit displayed by the aged rats. Potentiation
Although the frequency of obtaining potentiation following the 4-pulse train was lower in slices from aged rats, this was not a significant difference. Therefore, it appears that the threshold for potentiation did not differ between the two age groups. The in vitro results show a difference between slices of the two age groups in the magnitude of STP, but not LTP, following the short stimulus train (4 pulses). There were also no differences between the two groups in measures of both STP and LTP following the longer train (50 pulses). Therefore, it appears that the level, or amount, of stimulation used to induce potentiation can be an important variable in trying to delineate differences in hippocampal responsiveness between aged and young adult F344 rats. These results suggest that use of minimal stimulation paradigms (brief stimulus train) may prove more valuable in this regard than the use of 'over-stimulation' paradigms (longer stimulus trains). Thus, these brief stimulation trains may highlight deficits in either presynaptic or postsynaptic mechanisms underlying potentiation, which can be overcome by the more intense stimulation. Although short-lasting forms of synaptic plasticity (augmentation, frequency potentiation, paired-pulse facilitation and STP) have generally been thought to reflect presynaptic mechanisms (and thus a presynaptic deficit if abnormal), postsynaptic components of STP have also been suggested 1'27. LTP, in addition, may also possess both presynaptic and postsynaptic components, but clearly the dendritic integration of synaptic signals may be critical for adequate LTP induction and maintenance 4' 7,27. Dendritic integration depends upon a number of factors, including the postsynaptic receptor density, electronic distance between the cluster of synapses activated by the stimulus train, the voltage achieved with
the stimulus train and the total influx of the intracellular signal responsible for the maintenance of LTP (such as Ca2÷). A functional lengthening of overall dendritic electrotonic length in aged rats has previously been described 43, which may be an important aspect of the a b n o r m a l integration leading to the deficit shown here in STP, and may also be pertinent to LTP. The observed changes in electronic length could certainly be overcome by the increased depolarization during the 50-pulse train, as c o m p a r e d to the 4-pulse train. Thus, alterations in both pre- and postsynaptic mechanisms may accompany aging, and further studies will be required to describe these in m o r e detail. Correlation o f behavior and potentiation The relative degree of STP at 1 and 5 min (increases above baseline values in population spike amplitudes) following the 4-pulse train correlated with retention scores. The relationship between these variables remained positive when looked at within each age group, which suggests that STP, as well as LTP, may prove a useful neurophysiological m o d e l of m e m o r y processes. Retention scores were also found to show a positive
relationship with estimates of STP (potentiation at 1 min) and LTP (potentiation at 60 min) following the 50-pulse train. The within groups correlations for LTP and retention scores both r e m a i n e d positive, while the within groups of STP and retention scores showed a dissociation. The relationship r e m a i n e d positive within the young group, but showed a light negative correlation within the aged group. The finding that correlations between degree of potentiation and retention were significant at only 1 and 60 min is troubling. H o w e v e r , the fact that all correlations o b t a i n e d between potentiation and retention scores following the 50-pulse train were positive lends support to the argument that the two significant correlations were not spurious.
REFERENCES 1 Anwyl, R., Mulkeen, D. and Rowan, M.J., The role of N-methyl-D-aspartate receptors in the generation of short-term potentiation in the rat hippocampus, Brain Research, 503 (1989) 148-151. 2 Barnes, C.A., Memory deficits associated with senescence: a behavioral and neurophysiological study in the rat, J. Comp. Physiol. Psychol., 93 (1979) 74-104. 3 Barnes, C.A. and McNaughton, B.L., An age comparison of the rates of acquisition and forgetting of spatial information in relation to long-term enhancement of hippocampal synapses, Behav. Neurosci., 99 (1985) 1040-1048. 4 Baudry, M., Larson, J. and Lynch, G., Long-term changes in synaptic efficacy: potential mechanisms and implications. In P.W. Landfield and S.A. Deadwyler (Eds.), Long-Term Potentiation: From Biophysics to Behavior, Alan R. Liss, New York, 1988, pp. 109-136. 5 Berger, T.W., Long-term potentiation of hippocampal synaptic
Lastly, an estimate of the rate of change (persistence) in LTP over the 10- to 60-min time p e r i o d (potentiation at 60 min as a percentage of potentiation at 10 min) was found to show a positive correlation with retention scores, which also r e m a i n e d positive when looked at within the separate age groups. This result suggests that slices showing greater LTP persistence over the 10- to 60-min period t e n d e d to come from rats with longer retention scores (efficient memory). Overall the results following the 50-pulse train suggest that rats which d e m o n s t r a t e d efficient behavioral retention of the spatial task in vivo, yielded slices which p r o d u c e d greater initial potentiation (STP), and t e n d e d to show stable LTP over the next 50 min. On the o t h e r hand, rats demonstrating inefficient behavioral retention t e n d e d to provide slices which p r o d u c e d less STP, and showed less LTP persistence. In summary, the overall trend of these correlational results supports the hypothesis that potentiation processes (both STP and LTP) can serve as physiological models for m e m o r y processes. H o w e v e r , it is not meant to imply that one variable is causative of the other, as correlative relationships cannot imply causation. In fact, it is p r o b a b l e that these variables are effected by an intermediate variable (or variables). T h e r e are several factors which could be i n t e r m e d i a t e variables capable of influencing both behavior and h i p p o c a m p a i potentiation. A m o n g these are age-related changes in N M D A r e c e p t o r kinetics t2'39 and age-related deficits in calcium homeostasis, which could alter both pre- and postsynaptic calcium currents 2°'24. Acknowledgements. The authors wish to thank Jim Davis, Larry Goldstein, Steven Stasheff and Bill Wilson for providing helpful comments during discussions of this study, and Toni Shaw for assistance in preparation of the manuscript. This research was supported through a VA Research Service Award, and Grant IIRG-88-01 from the ADRDA.
transmission affects rate of behavioral learning, Science, 224 (1984) 627-630. 6 Berger, T.W. and Sclabassi, R.J., Long-term potentiation and its relation to hippocampal pyramidal cell activity and behavioral learning during classical conditioning. In P.W. Landfield and S.A. Deadwyler (Eds.), Long-term Potentiation: From Biophysics to Behavior, Alan R. Liss, New York, 1988, pp. 467-498. 7 Bliss, T.V.P. and Lynch, M.A., Long-term potentiation of synaptic transmission in the hippocampus: properties and mechanisms. In P.W. Landfield and S.A. Deadwyler (Eds.), LongTerm Potentiation: From Biophysics to Behavior, Alan R. Liss, New York, 1988, pp. 3-72. 8 Bondareff, W., Synaptic atrophy in the senescent hippocampus, Mech. Ageing Dev., 9 (1979) 163-171. 9 Castro, C.A., Silbert, L.H., McNaughton, B.L. and Barnes, C.A., Recovery of spatial learning deficits after decay of electrically induced synaptic enhancement in the hippocampus, Nature, 342 (1989) 545-548. 10 Collingridge, G.L., Kehl, S.J. and McLennan, H., Excitatory
9 amino acids in synaptic transmission in the Schaffer collateralcommissural pathway of the rat hippocampus, J. Physiol., 334 (1983) 33-46. 11 Cotman, C.W. and Anderson, K.J., Synaptic plasticity and functional stabilization in the hippocampal formation: possible role in Alzheimer's Disease. In G.S. Waxman (Ed.), Advances in Neurology (Vol. 47): Functional Recovery in Neurological Disease, Raven Press, New York, 1988, pp. 313-335. 12 Cotman, C. and Monaghan, D., Multiple excitatory amino acid receptor regulation of intracellular Ca2÷: implications for aging and Alzheimer's Disease, Ann. N.Y. Acad. Sci., 568 (1989) 138-148. 13 Creasey, H. and Rapoport, S.I., The aging human brain, Ann. Neurol., 17 (1985) 2-10. 14 De Toledo-Morreil, L. and Morrell, E, Electrophysiological markers of aging and memory loss in rats, Ann. N. Y. Acad. Sci., 444 (1985) 296-311. 15 Gage, EH. and Bjrrklund, A., Cholinergic septal grafts into the hippocampal formation improve spatial learning and memory in aged rats by an atropine-sensitive mechanism, J. Neurosci., 6 (1986) 2837-2847. 16 Gage, EH., Dunnett, S.S. and Bjrrklund, A., Spatial learning and motor deficits in aged rats, Neurobiol. Aging, 5 (1984) 43-48. 17 Geinisman, Y. and Bondareff, W., Decreases in the number of synapses in the senescent brain: a quantitative electron microscopic analysis of the dentate gyrus molecular layer in the rat, Mech. Ageing Dev., 5 (1976) 11-23. 18 Geinisman, Y., Bondareff, W. and Dodge, J.T., Partial deafferentation of neurons in the dentate gyrus of the senescent rat, Brain Research, 134 (1977) 541-545. 19 Giaquinto, S., Aging and Nervous System, John Wiley, New York, 1988. 20 Gibson, G.E. and Peterson, C., Calcium and the aging nervous system, Neurobiol. Aging, 8 (1987) 329-343. 21 Horel, J.A., The neuroanatomy of memory, Brain, 101 (1978) 403-445. 22 Hyman, B.T., Van Hoesen, G.W., Damasio, A.R. and Barnes, C.L., Alzheimer's Disease: cell-specific pathology isolates the hippocampal formation, Science, 225 (1984) 1168-1170. 23 Landfield, P.W., Correlative studies of brain neurophysiology and behavior during aging. In D.G. Stein (Ed.), The Psychobiology of Aging: Problems and Perspectives, Elsevier, New York, 1980, pp. 227-251. 24 Landfield, P.W., Campbell, L.W., Hao, S. and Kerr, S., Aging-related increases in voltage-sensitive, inactivating currents in rat hippocampus: implications for mechanisms of brain aging and Alzheimer's Disease, Ann. N.Y. Acad. Sci., 568 (1989) 95-105. 25 Landfield, P.W. and Lynch, G., Impaired monosynaptic potentiation in vitro hippocampal slices from aged, memory-impaired rats, J. Gerontol., 32 (1977) 523-533. 26 Landfield, P.W., McGaugh, J.L. and Lynch, G., Impaired synaptic potentiation processes in the hippocampus of aged, memory-deficient rats, Brain Research, 150 (1978) 85-101. 27 Larson, J. and Lynch, G., Role of N-methyl-D-aspartate receptors in the induction of synaptic potentiation by burst stimulation patterned after the hippocampal 0-rhythm, Brain
Research, 441 "(1988) 111-118. 28 Lindner, M.D. and Schallert, T., Aging and atropine effects on spatial navigation in the Morris water task, Behav. Neurosci., 102 (1988) 621-634. 29 Luine, V. and Hearns, M., Spatial memory deficits in aged rats: contributions of the cholinergic system assessed by CHAT, Brain Research, 523 (1990) 321-324. 30 McNaughton, B.L., Barnes, C.A., Rao, G., Baldwin, J. and Rasmussen, M., Long-term enhancement of hippocampal synaptic transmission and the acquisition of spatial information, J. Neurosci., 6 (1986) 563-571. 31 Mitchell, S.J., Rawlins, J.N.P., Steward, O. and Olton, D.S., Medial septal lesions disrupt theta rhythm and cholinergic staining in medial entorhinal cortex and produce impaired radial arm maze behavior in rats, J. Neurosci., 2 (1982) 292-302. 32 Morris, R., Developments of a water-maze procedure for studying spatial learning in the rat, J. Neurosci. Methods, 11 (1984) 47-60. 33 Morris, R.G.M., Anderson, E., Lynch, G.S. and Baudry, M., Selective impairment of learning and blockade of long-term potentiation by an N-methyl-o-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. 34 Morris, R. and Baker, M., Does long-term potentiation/synaptic enhancement have anything to do with learning and memory? In L.R. Squires and N. Butters (Eds.), Neuropsychology of Memory, Guilford Press, New York, 1984, pp. 521-535. 35 Morris, R.G.M., Davis, S. and Butcher, S.P., The role of NMDA receptors in learning and memory. In J.C. Watkins and G.L. Collingridge (Eds.), The NMDA Receptor, IRL Press, Oxford, 1989, pp. 137-152. 36 Morris, R.G.M., Garrud, P., Rawlins, J.N.P. and O'Keefe, J., Place-navigation impaired in rats with hippocampal lesions, Nature, 297 (1982) 681-683. 37 O'Keefe, J. and Nadel, L.,The Hippocampus as a Cognitive Map, Clarendon-Oxford University Press, New York, 1978. 38 Olton, D.S., Walker, J.A. and Gage, EH., Hippocampal connections and spatial discrimination, Brain Research, 139 (1978) 295-308. 39 Peterson, C. and Cotman, C.W., Strain-dependent decrease in glutamate binding to the N-methyl-o-aspartic acid receptor during aging, Neurosci. Lett., 104 (1989) 309-313. 40 Rapp, ER., Rosenberg, R.A. and Gallagher, M., An evaluation of spatial information processing in aged rats, Behav. Neurosci., 101 (1987) 3-12. 41 Spencer, R.L., Ganem, J., Chapman, S., O'Steen, W.K. and McEwen, B.S., Evaluation of L-acetyl carnitine (LAC) on brain and behavioral measures in aged rats, Soc. Neurosci. Abstr., 15 (1989) 261. 42 Tielen, A.M., Mollevanger, W.J., Lopes da Silva, EH. and Hollander, C.E, Neuronal plasticity in hippocampal shces of extremely old rats. In W.H. Gispen and J. Traber (Eds.), Aging of the Brain, Elsevier, Amsterdam, 1983, pp. 73-84. 43 Turner, D.A. and Deupree, D.L., Functional elongation in CA1 pyramidal neurons in senescent F-344 rats, Neurobiol. Aging, in press. 44 Winson, J., Loss of hippocampal theta rhythm results in spatial memory deficit in the rat, Science, 201 (1978) 160-162.
BRES 16770
Research Reports
Spatial performance correlates with in vitro potentiation in young and aged Fischer 344 rats David L. Deupree, Dennis A. Turner and Carter L. Watters Department of Neurosurgery and Research Services, Durham VAMC and Division of Neurosurgery, Department of Surgery, Duke University Medical Center, Durham, NC 27710 (U.S.A.)
(Accepted 5 February 1991) Key words: Potentiation; Aging; Hippocampus; CA1 extracellular field potential; Spatial memory
Young adult (2-4 months old) and aged (24-26 months old) Fischer 344 (F344) rats were trained for spatial behavior (locating a hidden escape platform) in a circular water maze. The aged rats showed deficits in both the acquisition and retention of the learned response. Following the behavioral training, hippocampal slices from the rats were prepared. Potentiation of CA1 extracellular, somatic field potentials was studied in vitro following either a short stimulus train (4 pulses) or a longer train (50 pulses). Slices from the aged rats showed less short-t,~,'m potentiation (124.8 _+ 4.9% baseline, mean _+ S.E.M.) at 1 min following the short train in comparison to slices from the young rats (151.8 _+ 7.5%, P < 0.05). However, following the longer train, no differences were found between the groups in the degree of either short-term (measured at 1 min after stimulation) or long-term potentiation (measured at 60 min). The amount of potentiation seen at various time points after either train correlated with the behavioral measure of retention. These results indicate that F344 rats exhibit age-related behavioral deficits, and age-related synaptic potentiation deficits in response to short stimulation trains. The correlation between the degree of potentiation (both short-term and long-term) and retention of a behavioral task adds strength to the hypothesis that potentiation mechanisms may underlie memory processes. INTRODUCTION Although not all individuals exhibit adverse aging effects, when they do occur, aging can produce deterioration of brain, behavior and cognitive functioning 13'19. The hippocampus is one of the initial brain regions adversely effected by aging. The hippocampus in particular appears to become partially deafferented through the process of aging, with deterioration and loss of afferents over time 8'17'1s. This deafferentiation and isolation is even more severe with Alzheimer's Disease 11,22, and may be partially responsible for the early cognitive decline seen in Alzheimer's patients. In rats, integrity of the hippocampus is important in the acquisition and retention of several learned behaviors 21, among these being so-called 'spatial behavior '36'37. Lesions of the entorhinal cortex or the medial septal nucleus 31'38"44, structures which provide afferent input to the hippocampus and which also exhibit spontaneous degeneration with aging and Alzheimer's Disease, lead to deficits of spatial behavior in rats. In addition, aged rats have been found to display significant deficits in spatial acquisition and retention 2'3A4-16,28,29,4°. Potentiation, or enhancement, of hippocampal syn-
apses has been proposed as a neurophysiological model for memory 4'7'34. Since aged rats often exhibit m e m o r y deficits, one might expect that synaptic potentiation in the hippocampus of aged rats would also be deficient. However, the results relating hippocampal potentiation and aging are mixed. A number of studies have found age-related deficits of population spike increases in the CA1 region following induction of long-term potentiation (LTP), and during frequency potentiation 25'26'42. However, other investigators could find no age effect in population spike amplitude of dentate granule or CA1 neurons following LTP induction a'lg'2a. Some of these same groups did find that the decay of dentate granule cell population excitatory postsynaptic potential (EPSP) slope following the induction of LTP was faster in aged rats a'3"14. In addition, Barnes and McNaughton 3 found that when using repeated stimulus trains to induce maximal LTP over a period of time, young rats reached their maximum faster than aged rats. The discrepancies noted here could be due to a variety of methodological and technical differences between studies, including varying stimulation parameters and physiological measurements. In order to strengthen the position for LTP (and
Correspondence: D. Deupree, Neurosurgery, Box 3807, Duke University Medical Center, Durham, NC 27710, U.S.A.
p o t e n t i a t i o n / e n h a n c e m e n t in general) as a memory model, it would be important to show within subjects correlations between LTP and behavioral measures of memory. Further, manipulation of one variable should influence the outcome of the other. Morris and Baker 34 review research strategies intended to strengthen the argument for potentiation as a model for memory. The strongest methodologies were argued to be correlational and manipulation strategies 34. Morris and colleagues 33 showed that chronic infusions of D-2-amino-5-phospho-
behavioral deficits with differences in potentiation parameters. This was accomplished by first training both young adult and aged Fischer 344 (F344) rats in a water maze for spatial behavior, and then inducing potentiation in the hippocampal CA1 region in vitro using either a brief stimulus train (4 pulses) or a longer train (50 pulses). The measures of spatial learning and memory were then compared with the physiological measures of potentiation, both between groups and on an individual basis across age groups.
novalerate (AP5) into the lateral ventricles produced spatial learning deficits in rats during water maze training. AP5 is a N-methyl-D-aspartate receptor antagonist which is known to block the induction of LTP in the hippocampus m'33. Further experiments 35 showed correlations between AP5 tissue content within the hippocampus during chronic intraventricular infusion, decreases in the degree of dentate LTP induction and water maze performance in rats. Induction of dentate LTP in vivo prior to behavioral training has been shown to inhibit spatial maze performance in rats 9'3° or to enhance behavioral acquisition of classical conditioning in rabbits 5. It has been argued that the opposing influence produced by pre-training LTP induction in these two circumstances may be due to a differential influence of heightened hippocampal excitability upon acquisition of the two responses 6. Regardless, these studies demonstrate correlations between LTP and behavior 33"35, and that m a n i p u l a t i o n of LTP can influence resulting behavioral acquisition5"9'3°. If one accepts the argument that aging can lead to spontaneous hippocampal lesions, then it could be argued that aging studies fall u n d e r the manipulation strategy (age-induced hippocampal lesions being a manipulation). In adding an age variable to the correlational strategy, one combines correlation and manipulation strategies. There are two studies which have used the correlation approach to investigate possible relationships between aging, various aspects of potentiation and behavior. Landfield 23 found that synaptic depression seen in the later stages of frequency potentiation could distinguish between aged rats that exhibited retention deficits (of avoidance conditioning) and aged rats that did not exhibit deficits. Barnes and McNaughton 3 found that the acquisition and decay of EPSP slope during and after LTP induction corresponded to acquisition and retention behavior of young and aged rats in a radial maze. Thus it appears that there may be correlations between behavioral measures of learning and memory in aged rats and physiological measures thought to reflect memory processes. The purpose of the present study was to extend the results of these earlier attempts to correlate age-related
MATERIALS AND METHODS Water maze procedures
A water maze was selected as the tool for behavioral assessment of the rats because, as noted above, measurements obtained from use of this type of maze can differentiate between young adult and aged rats 15'16'28'4°, as well as being sensitive to hippocampal integrity36. Subjects were young and aged male F344 rats. The aged rats (n = 20) were obtained from the National Institute on Aging, were 24-26 months old, and healthy (free of respiratory problems and no visible tumors) at the time of behavioral testing and subsequent in vitro experimentation. Young rats (n = 15) were obtained from Charles Rivers, and were 2-4 months old. Although the different age groups of rats were obtained from different suppliers, both suppliers obtained their breeding stock from the same source (National Institute of Health), and thus the rats were from identical genetic sources. Rats were trained for so-called spatial memory using a Morris water maze32. Briefly, rats were trained to locate an escape platform within a circular water maze. The maze was 1.5 meters in diameter by 0.6 meters in height. The maze was filled with water (approximately 22 °C) mixed with opaque food dye, in order to hide the location of the escape platform from the rats. The escape platform was always located just below the water surface (1-2 cm). Located within the room were several stationary visual distal cues such as the overhead light rack and rat housing racks along two of the walls. Also there were significant non-visual distal cues which also remained stationary (i.e. the sounds and smells generated by rats within the housing racks). Thus there were several distal sensory cues within the testing environment which the rats could utilize in locating the hidden platform. Each rat was given 8 training (acquisition) trials per day, over 4 consecutive days (32 total trials). Each trial consisted of gently placing the rat into the maze, facing the wail, at one of 4 entry sites. The entry sites were located in a repeatable manner by dividing the maze into 4 equal quadrants. During each day, the rat would receive two trials from each entry point, randomly determined. The escape platform was located in a constant position (placed in the center of one quadrant) for each rat during all training trials. During each trial the rat was allowed 60 s to find the platform. Once the rat had located the platform, it was allowed a brief rest period (30-60 s) on the platform before a new trial was begun. Escape latency (in s) for each trial was measured. If the rat had not located the platform with 60 s, it received a latency score of 60 s, was gently placed on the platform and allowed to rest there briefly before a new trial began. Rats were dried off with towels between every 2-3 trials. Each rat was observed carefully during each trial. If a rat exhibited any difficulty in swimming about the maze, it was removed from the maze, and discarded from the study. Four aged rats were found to exhibit swimming abnormalities and these animals were not included in the study, thus the behavioral results were obtained from 16 rats from this group. Following the completion of acquisition training, each rat received a retention trial 24 h later. The retention trial was identical to the training trials, except that the escape platform was removed
from the maze. The amount of time the rat spent swimming in the quadrant where the platform had been located during the previous training trials (training quadrant) was measured during the 60-s retention trial. Time spent in the training quadrant during the retention trial was referred to as the retention score.
Hippocampal slice preparation and recording procedures All physiological experiments were performed within two weeks of the completion of behavioral testing. Rats were sacrificed using Halothane, the hippocampus from each hemisphere was dissected out, and transverse mid-hippocampal slices 500 /zm thick were obtained using a chopping type microtome 43. Slices were allowed to incubate in a physiological bath for at least 90 min before being transferred to a gas interface type slice chamber where somatic hippocampal CA1 field potentials were recorded. The temperature of the bath within the chamber was maintained at 33 + 1 °C. Concentrations of the bath were (in mM): D-glucose 10, KCI 3, NaCI 124, NaH2PO 4 H 2 0 1.2, NaHCO 3 26, MgSO 4 7H20 2, CaCl 2 6H20 2.4. The bath and slices were exposed to a mixture of 5% CO 2 and 95% 0 2 gas. Recording electrodes were glass micropipettes (5-10 MI2) filled with 2 M NaCl. Stimulation was delivered with bipolar electrodes placed in the stratum radiatum of the CA1 field. Test stimuli were monophasic constant current pulses, 50 /~s in duration. Estimates of maximal population spike amplitudes were determined by increasing the duration of pulse, while keeping the current intensity at 5.0 mA, until there was no further increase in population spike size. Potentials were amplified by a factor of 10, filtered at 0-10 kHz, and recordings were stored on computer disks for off-line analysis. Criteria for acceptable field potentials were population spikes from the pyramidal cell layer, with maximal amplitude of at least 5 inV. Population spike amplitude was measured from the initial positive peak of the extracellular EPSP to the subsequent negative peak of the population spike (see Fig. 1). Once an acceptable field potential was obtained, 1-2 h were allowed for stabilization, during which potentials were generated once per min. Following stabilization, current intensity was adjusted to evoke a population spike 50% of maximal amplitude. The current range used to evoke half amplitude population spikes was 3.2-5.5 mA at 50 /~s. After ensuring that the potentials were stable following current adjustment, baseline potentials were recorded for 4 rain, then either a short (4 pulses at 100 Hz) or long (50 pulses at 100 Hz) stimulus train was delivered through4he stimulating electrodes. Post-train poten-
A
tials were recorded for 10 min following the 4-pulse train, and for 1 h following the 50-pulse train. Post-stimulus train population spike amplitudes were converted to a percentage of the mean population spike amplitude of the 4 baseline potentials. Population spikes were chosen for analysis because the changes in population spike amplitudes as a result of potentiating stimuli are more robust than changes seen in synaptic potentials. Therefore measurement of population spike changes was chosen to maximize the chances of observing a potentiating effect following the short stimulus train (4 pulses).
Data analysis Analysis of data (both behavioral and physiological) was performed using multivariate analysis of variance, with post-hoc analysis being performed using Tukey's test for differences between group means. Correlations were performed using Pearson correlation coefficients. Behavioral variables included average escape latencies per block of 8 training trials and retention score (time spent in the training quadrant during the retention trial). Potentiation measures included for correlation with behavioral retention scores were percent baseline values obtained at various time points following the presentation of the stimulus train. Frequency data were analyzed using a 2 x 2 chi square table.
RESULTS
Behavioral Average escape latencies for each training day (block of 8 trials) are summarized in Fig. 2. Over the 4 training days the aged rats as a group showed no reduction in escape latencies (/'3,45 -----0.17, n.s.), while the young rats did show reductions across days (F3,42 = 23.20, P < 0.01). Overall there was a significant age effect (F1,29 = 23.48, P < 0.01). Post-hoc analysis between groups found the escape latencies of the two groups were not significantly different from each other during the first two training days. However, the escape latencies were found
Young rats (n = 15) BB
C
~ e d rots (n = IS) r-~
~' X +
6O
40
B
D
(J _~
20
i1/1i1i BLOCKS OF EIGHTTRIALS
Fig. 1. Examples of CA1 field potentials obtained from hippocampal slices of aged and young rats, both before and after presentation of the 4-pulse stimulus train. A and B show potentials before (A) and 1 min after (B) the train of a slice from a young rat. Amplitudes of population spikes were measured as the absolute value (in mV) from the initial positive peak of the extracellular EPSP to the adjoining negative peak of the population spike. C and D show representative potentials of a slice from an aged rat at the same time points. The vertical bar for traces A and B is 10.0 mV and 6.0 mV for C and D. The horizontal bar is 30 ms for all traces in the figure.
Fig. 2. Summary of escape latencies (mean + S.E.M.) across blocks of training trials. The latencies were not significantly different between age groups during the first two blocks of trials. However, during the last two blocks (days 3 and 4) escape latencies of the aged rats were found significantly greater than the latencies of the young rats. These results show the average latencies of the young rats progressively declining across the 4 blocks of trials (indicating acquisition of the behavioral response), while the average latencies of the aged rats do not show such a trend across days. **P < 0.01 difference between groups.
to be significantly different on days 3 and 4 ( P < 0.01 for both comparisons). Fig. 3 summarizes retention scores for the two groups. The escape platform had been located in q u a d r a n t 2 during the training trials, thus time spent in that q u a d r a n t during the retention trial served as the retention score. The young rats were found to have significantly greater retention scores (30.5 + 3.0 s, mean + S . E . M . ) than the aged rats (17.6 + 1.5 s) (F1.29 = 14.99, P < 0.01). Also shown in Fig. 3 is time spent in each of the o t h e r quadrants during the retention trial, whereas the d o t t e d line represents time spent in any quadrant based upon chance (15 s). On average the aged rats were spending time in all quadrants around chance levels during the retention trial, while the young rats spent the majority of time within the training quadrant.
In vitro potentiation A c c e p t a b l e field potential recordings were obtained from 14 of the 15 young rats from which behavioral data were collected (93% frequency of acceptance), and of the 16 aged rats, acceptable potentials were obtained from 13 rats (81%). These acceptance frequencies were not found to be significantly different between the groups (X2(1) = 0.22, n.s.). The two age groups did not differ in current intensity (either in overall range or average intensity) n e e d e d to evoke a half maximal population spike. Slices from 10 young rats and 7 aged rats received the short (4 pulses at 100 Hz) stimulus train. Slices from 10 young rats and 6 aged rats received the longer stimulus train (50
pulses at 100 Hz). T h e r e were 6 young rats from which it was possible to obtain two viable slices, using one slice for each stimulus train experiment. T h e r e were no aged rats which provided a slice for each experiment. A few slices did not show potentiation following the 4-pulse train (operationally defined as an increase of greater than 10% above the average baseline value when m e a s u r e d at 1 min following the train). Thus, data from the slices of 1 young rat (90% potentiation frequency) and 2 aged rats (71%) were discarded for not showing sufficient potentiation following the 4-pulse train. These frequencies for potentiation occurrence between the two groups were not significantly different (X2(1) = 0.12, n.s.). All slices from both age groups showed p o t e n t i a t i o n following the 50-pulse train. Fig. 4 summarizes the effects of the 4-pulse train on population spike amplitudes from slices which showed potentiation (more than a 10% increase over baseline levels at 1 min). A l t h o u g h by definition slices from both age groups showed potentiation, it can be seen that, at least during the first few min following the train, the slices from young rats on average showed greater potentiation than the slices from aged rats. The initial increase in population spike amplitudes seen during the first few min after stimulation will be referred to here as short-term potentiation (STP). A f t e r about 5 rain, potentiation levels of both groups of slices showed a tendency to stabilize (long-term potentiation, LTP). In o r d e r to avoid unwieldy levels of the r e p e a t e d measures variable (time), statistical analysis was restricted to percent baseline values at - 1 (last baseline measure), 1,
Young Rats (n = 15) BIB Aged Rats (n = 16) [S:] Slices from young rots (n = 9) o - - e Slices from aged rots (n = 5) o - - o
30 +
160
'
-L 20
8
+
-
10
i
1
3
"i.... I" 4
1
2
}'L+
.
,?, 3
1O0
4
QUADRANTS
Fig. 3. Summary of time (mean + S.E.M.) spent in each of the 4 maze quadrants during the retention trial. During the training trials, the escape platform was located in quadrant 2. The dotted line represents chance level (15 s) for amount time spent in any particular quadrant during the trial. The young rats on average were spending time in quadrant 2 that was well above chance levels, while the aged rats were close to chance. The retention scores (time spent in quadrant 2) of the aged rats were significantly shorter than the scores of the young rats. One can see that the young rats by far spent the majority of the retention test in quadrant 2. Although the aged rats, on average, did spend more time in quadrant 2 than any of the other 3 quadrants during the trial, this was not a significant trend. **P < 0.01 difference between groups.
=o ^
1
5
10
MINUTES FOLLOWING STIMULUS TRAIN (4 pulses at 100 Hz)
Fig. 4. Summary of change in percent baseline values (mean + S.E.M.) of CA1 population spikes following a 4-pulse at 100 Hz stimulus train. Included are baseline values obtained during the 4 min prior to train presentation (arrow on time axis, min 0). Slices from rats of both age groups showed increases over baseline levels following the 4 pulse train, however, the slices from young rats showed greater increases. Between groups analysis was performed at 1, 5 and 10 min following the train. A significant difference between potentiation values was found at 1 min following the train, but not at 5 or 10 min. *P < 0.05 difference between groups.
5 and 10 min following the train. Using those time points only, it was found that overall the potentiation values of the slices from young rats were significantly greater (F1,12 = 6.12, P < 0.05). Post-hoc analysis also showed that potentiation values of slices from young rats were significantly greater than those of slices from aged rats at 1 min following the train (P < 0.05). As can be seen in Fig. 4, there was no significant difference between potentiation values at either 5 or 10 rain following the train. These results indicate that the slices from young rats showed greater initial potentiation following the 4-pulse train (STP), but that the two groups did not differ in terms of later phases of potentiation (LTP) following the 4-pulse train. At 1 min following the train the slices from the young rats showed potentiation values of 151.8 + 7.5 (mean + S.E.M.) percent baseline, while the slices from the aged rats showed potentiation values of 124.8 + 4.9 percent baseline. At 10 min after the train, the potentiation values for the slices from young rats were 131.4 + 5.7, and the values for slices from aged rats were 112.8 + 6.1. Thus slices from both age groups showed initial potentiation (STP) which decreased in magnitude over a few minutes, but were still slightly above original baseline values when measured at 10 min (LTP). Recordings from several of these slices were carried out for periods up to 1 h. In these slices, the LTP values on average did not return to original baseline values (remaining just slightly above as was the case for the 10 min values), in slices from either age group. Fig. 5 summarizes effects of the 50-pulse train on population spike amplitudes. The initial increase in population spike amplitudes (first 10 min) seen following the 50-pulse train is again referred to as STP, while the potentiation during the next 50 min (min 10-60) is referred to as LTP. The left side of Fig. 5 summarizes the data for STP (min 1-10, at 1-min intervals), while the right side summarizes the data for LTP (min 10-60, at 10-min intervals). One can see that there were no significant differences in potentiation values at any time point following the 50-pulse train between slices from young rats (n = 10 slices from 10 rats) and aged rats (n = 6 slices from 6 rats). As can be seen in Fig. 5, the potentiation values of both groups of slices, on average, appeared to decline at similar rates during the first few min following the 50-pulse train (STP), then remain relatively stable during the remaining time (LTP). At 1 rain following the train the potentiation values for slices from young rats were 204.5 + 9.4 percent baseline, and 196.5 + 3.7 for the slices from aged rats. These values were seen to decrease to 174.3 + 7.7 (slices from young rats) and 175.7 + 7.4 (slices from aged rats) after 10 min. At 60 min the potentiation values for both groups had
decreased slightly (but not significantly) from their 10 min values (slices from young rats = 165.7 + 14.6%, and slices from aged rats = 159.0 + 19.6%). Correlation o f behavior with potentiation Correlation coefficients between retention scores and in vitro potentiation measures obtained from the slices at 1, 5 and 10 min following the 4-pulse train were calculated. The correlations for combined data collapsing across the two age groups (n = 14; aged rats = 5 and young rats = 9) showed significant positive correlations between retention scores and potentiation values at 1
min: r13 = 0.63 (P < 0.05) and at 5 min: r13 ---- 0 . 5 7 ( P < 0.05), but not at 10 min: r13 = 0 . 4 0 ( n . s . ) . Fig. 6A shows the scatter plot of the individual data points for the correlation of retention scores with potentiation at 1 min following the 4-pulse train (STP). One can see the relationship between the two variables was also positive within each of the age groups, although not significant. The correlation between retention scores and STP for young rats was: r s = 0.31 (n.s.) and for the aged rats: r 4 = 0.84 (n.s.).
Correlation coefficients between retention scores and in vitro measures of potentiation following the 50-pulse train were also calculated. These data were also from an analysis that collapsed across age groups (n = 16; aged rats = 6 and young rats = 10). Fig. 6 B - D summarizes 3 correlations. Retention scores were found to show a
Slices from young rate (n = 10) e ~ e Slices from aged rats (n == 6) o ~ o E J +
200
8 E z" '
150
.J LIJ
I--
z
STP
100
LTP
hJ
o_
,
. . . .
"
. . . .
~,
....
io'
,
10
•
,
,
,
.
,
60
MINUTES FOLLOWINGSTIMULUSTRAIN (50 pulses at 100 Hz) Fig. 5. Summary of change in percent baseline values (means _+
S.E.M.) of CA1 population spikes following a 50-pulse at 100 Hz stimulus train. The left portion of the figure shows baseline values during the 4 min prior to train presentation (arrow on time axis, min 0), as well as the average increases above baseline during the first 10 min following the train (STP). The right portion of the figure shows percent baseline averages for both groups during the 10-60 min time period following the train, at 10-min intervals (LTP). There were no significant differences between the two groups at any time period analyzed.
A
C
Young rots (n - 9) •
Young rats (n = 10) • Aged rots (n = 6) O
Aged rots (n - 5) O
•
...--
z
200
g
15o
~
loo
Z _
•
O 0
~
..
.._ •:.~-'~7., • • •
•
.~
O •
5O O 100
1o
;o
~o
do
REI-tNnON SCORE ( N c )
B
;o
do
RETENTION SCORE ( ~ ¢ )
D
Young rotl (n - I 0 ) • Agod mr, (n - e) o
Young rotl (n - 10) • Aged ~ , (. - e) o
300
i
lOO o
:oo
.
~o
~o
. . . . . . . . . . •
.
0
1o
~o
do
RETENTION SCORE (sec)
do
RETENTION SCORE (sac)
Fig. 6. Scatter plots of correlations between behavioral retention scores and in vitro potentiation measures. Within each scatter plot, individual data points from young rats (solid dots) and aged (open dots) rats are displayed. Also displayed are regression lines representing correlations for the data collapsed across groups (solid line), as well as for correlations within each group (regression line for young rats is dashed line, and regression line for aged rats is dotted line). See text for details concerning the various correlation coefficients. A: the scatter plot for the correlation between retention scores and potentiation values measures at 1 rain following the 4-pulse train (STP). The overall correlation across groups was positive. In addition, one can see that individually both groups showed a positive relationship between the two variables. B: the scatter plot of the correlation between retention scores and potentiation values measured at 1 min following the 50-pulse train (STP). The correlation between the two variables across groups was positive. One can see, however, that although the young rats maintained this positive relationship, the aged rats did not, and in fact show a slightly negative correlation between the variables, C: the scatter plot of retention scores correlated with potentiation values measured at 60 min following the 50-pulse train (LTP). Here also the overall relation between the variables was positive, and the relationship between these variables within each group remained positive. D: the scatter plot of the correlation of retention scores with a measure of LTP persistence (potentiation measured at 60 min as a percent of the potentiation measured at 10 min) following the 50-pulse train. Once again, the overall relationship between the variables was positive, as was the relationship within the two groups.
significant positive c o r r e l a t i o n with p o t e n t i a t i o n values at b o t h 1 min: rl5 = 0.50 ( P < 0.05) and 60 min: r15 = 0.51 ( P < 0.05) f o l l o w i n g t h e train. T h u s , o v e r a l l the r e t e n t i o n scores f r o m b o t h g r o u p s o f rats s h o w e d positive correlations with m e a s u r e s of S T P and LTP following the 50-pulse train. Fig. 6B shows the scatter plot of individual d a t a p o i n t s for the c o r r e l a t i o n b e t w e e n r e t e n t i o n scores and p o t e n t i a t i o n at 1 m i n f o l l o w i n g the 50-pulse train (STP). O n e can see a dissociation b e t w e e n the two age g r o u p s in h o w the two variables c o r r e l a t e d . T h e correlation b e t w e e n the two variables for y o u n g rats was still
positive, a l t h o u g h not significant; r 9 = 0.52 (n.s.), while the c o r r e l a t i o n for the a g e d rats was slightly n e g a t i v e , a l t h o u g h also n o t significant; r 5 = - 0 . 0 9 (n.s.). Fig. 6C shows the scatter plot for t h e c o r r e l a t i o n b e t w e e n retention
scores
and
potentiation
values
at
60 min
f o l l o w i n g the 50-pulse train (LTP). In this p l o t b o t h age groups retain a p o s i t i v e r e l a t i o n b e t w e e n the two variables, a l t h o u g h again b o t h r e l a t i o n s h i p s w e r e n o t significant. F o r y o u n g rats; r 9 = 0.55 (n.s.), and for a g e d rats; r 5 = 0.76 (n.s.). C o r r e l a t i o n coefficients w e r e also calculated b e t w e e n r e t e n t i o n scores and p o t e n t i a t i o n
7 values at 1 min and all 10 min increments following the train; however, significant correlations were found only at 1 and 60 min. Previous studies have found that the decay of LTP in vivo was faster in aged r a t s 2'3'14 and that the rate of decay in LTP was similar to the rate of decay in behavioral retention of spatial memory in rats 3. It was recognized that the decay in LTP from the studies cited a b o v e 2'3'14 was measured in terms of changes in field EPSP slope, and that population spike LTP can be decoupled from EPSP LTP (implying the possibility of different mechanisms), thus there was no a priori reason for expecting similar results during the present study. However, a measure of population spike change following the 50pulse train was compared with retention latencies. The estimate of LTP change (persistence) chosen was the comparison of potentiation at 60 min with potentiation at 10 min (min 60 as a percentage of min 10) following the 50-pulse train. This measure thus focused on the time period summarized in the right portion of Fig. 5 (LTP), avoiding contamination from the STP phase. Across the two groups LTP persistence showed a significant positive correlation with retention scores; r15 = 0.53 (P < 0.05). Fig. 6D shows the scatter plot of individual data points for correlation of retention scores with LTP persistence (potentiation at 60 rain as a percentage of potentiation at 10 min). Here again, the individual groups retain positive relationships between the two variables. For the young rats; r 9 = 0.54 (n.s.), while for the aged rats; r 5 = 0.82 (P < 0.05). DISCUSSION Behavior
The behavioral results show severe deficits (both in acquisition and retention) by the aged rats. The behavioral results are consistent with a variety of past studies 2' 3.14-16,28,29,4o, which demonstrate deficits in behavioral measures of 'spatial' learning and memory in aged rats. The severity of the deficit exhibited by the aged rats is troublesome, in that on average, the aged rats in this study showed no signs of acquiring the spatial task (a flat learning curve). Several reports regarding aged rats have found a learning curve during spatial training 2'3'14'15'4°. However, the present behavioral results are consistent with other studies which looked at spatial performance in aged F344 rats 2s'29, which could find no evidence for behavioral acquisition across trials (a flat learning curve). It is certainly possible that the severe behavioral results seen in the present study may have been due to other factors besides deficit learning abilities. For example, visual deficits cannot be ruled out as an alternative explanation for the poor behavioral acquisition and
retention exhibited by the aged rats. Recent data show a correlation between degree of retinal atrophy and water maze performance by aged rats 41. In order to test for visual deficits for local, intra-maze cues, a group of 24-month-old F344 rats were tested in the water maze in this lab while using a visual platform. These rats did exhibit a steady decrease in escape latencies over training trials (data not shown). Unfortunately this only demonstrates that these aged rats could see, and utilize, cues within the maze itself. The possibility still exists that the aged rats might not have been able to appreciate, or to utilize the many other sensory distal (extra-maze) cues which were present, which could account for the acquisition deficit seen during this study. Thus, the behavioral results presented here as they stand cannot rule out alternative explanations for the spatial performance deficit displayed by the aged rats. Potentiation
Although the frequency of obtaining potentiation following the 4-pulse train was lower in slices from aged rats, this was not a significant difference. Therefore, it appears that the threshold for potentiation did not differ between the two age groups. The in vitro results show a difference between slices of the two age groups in the magnitude of STP, but not LTP, following the short stimulus train (4 pulses). There were also no differences between the two groups in measures of both STP and LTP following the longer train (50 pulses). Therefore, it appears that the level, or amount, of stimulation used to induce potentiation can be an important variable in trying to delineate differences in hippocampal responsiveness between aged and young adult F344 rats. These results suggest that use of minimal stimulation paradigms (brief stimulus train) may prove more valuable in this regard than the use of 'over-stimulation' paradigms (longer stimulus trains). Thus, these brief stimulation trains may highlight deficits in either presynaptic or postsynaptic mechanisms underlying potentiation, which can be overcome by the more intense stimulation. Although short-lasting forms of synaptic plasticity (augmentation, frequency potentiation, paired-pulse facilitation and STP) have generally been thought to reflect presynaptic mechanisms (and thus a presynaptic deficit if abnormal), postsynaptic components of STP have also been suggested 1'27. LTP, in addition, may also possess both presynaptic and postsynaptic components, but clearly the dendritic integration of synaptic signals may be critical for adequate LTP induction and maintenance 4' 7,27. Dendritic integration depends upon a number of factors, including the postsynaptic receptor density, electronic distance between the cluster of synapses activated by the stimulus train, the voltage achieved with
the stimulus train and the total influx of the intracellular signal responsible for the maintenance of LTP (such as Ca2÷). A functional lengthening of overall dendritic electrotonic length in aged rats has previously been described 43, which may be an important aspect of the a b n o r m a l integration leading to the deficit shown here in STP, and may also be pertinent to LTP. The observed changes in electronic length could certainly be overcome by the increased depolarization during the 50-pulse train, as c o m p a r e d to the 4-pulse train. Thus, alterations in both pre- and postsynaptic mechanisms may accompany aging, and further studies will be required to describe these in m o r e detail. Correlation o f behavior and potentiation The relative degree of STP at 1 and 5 min (increases above baseline values in population spike amplitudes) following the 4-pulse train correlated with retention scores. The relationship between these variables remained positive when looked at within each age group, which suggests that STP, as well as LTP, may prove a useful neurophysiological m o d e l of m e m o r y processes. Retention scores were also found to show a positive
relationship with estimates of STP (potentiation at 1 min) and LTP (potentiation at 60 min) following the 50-pulse train. The within groups correlations for LTP and retention scores both r e m a i n e d positive, while the within groups of STP and retention scores showed a dissociation. The relationship r e m a i n e d positive within the young group, but showed a light negative correlation within the aged group. The finding that correlations between degree of potentiation and retention were significant at only 1 and 60 min is troubling. H o w e v e r , the fact that all correlations o b t a i n e d between potentiation and retention scores following the 50-pulse train were positive lends support to the argument that the two significant correlations were not spurious.
REFERENCES 1 Anwyl, R., Mulkeen, D. and Rowan, M.J., The role of N-methyl-D-aspartate receptors in the generation of short-term potentiation in the rat hippocampus, Brain Research, 503 (1989) 148-151. 2 Barnes, C.A., Memory deficits associated with senescence: a behavioral and neurophysiological study in the rat, J. Comp. Physiol. Psychol., 93 (1979) 74-104. 3 Barnes, C.A. and McNaughton, B.L., An age comparison of the rates of acquisition and forgetting of spatial information in relation to long-term enhancement of hippocampal synapses, Behav. Neurosci., 99 (1985) 1040-1048. 4 Baudry, M., Larson, J. and Lynch, G., Long-term changes in synaptic efficacy: potential mechanisms and implications. In P.W. Landfield and S.A. Deadwyler (Eds.), Long-Term Potentiation: From Biophysics to Behavior, Alan R. Liss, New York, 1988, pp. 109-136. 5 Berger, T.W., Long-term potentiation of hippocampal synaptic
Lastly, an estimate of the rate of change (persistence) in LTP over the 10- to 60-min time p e r i o d (potentiation at 60 min as a percentage of potentiation at 10 min) was found to show a positive correlation with retention scores, which also r e m a i n e d positive when looked at within the separate age groups. This result suggests that slices showing greater LTP persistence over the 10- to 60-min period t e n d e d to come from rats with longer retention scores (efficient memory). Overall the results following the 50-pulse train suggest that rats which d e m o n s t r a t e d efficient behavioral retention of the spatial task in vivo, yielded slices which p r o d u c e d greater initial potentiation (STP), and t e n d e d to show stable LTP over the next 50 min. On the o t h e r hand, rats demonstrating inefficient behavioral retention t e n d e d to provide slices which p r o d u c e d less STP, and showed less LTP persistence. In summary, the overall trend of these correlational results supports the hypothesis that potentiation processes (both STP and LTP) can serve as physiological models for m e m o r y processes. H o w e v e r , it is not meant to imply that one variable is causative of the other, as correlative relationships cannot imply causation. In fact, it is p r o b a b l e that these variables are effected by an intermediate variable (or variables). T h e r e are several factors which could be i n t e r m e d i a t e variables capable of influencing both behavior and h i p p o c a m p a i potentiation. A m o n g these are age-related changes in N M D A r e c e p t o r kinetics t2'39 and age-related deficits in calcium homeostasis, which could alter both pre- and postsynaptic calcium currents 2°'24. Acknowledgements. The authors wish to thank Jim Davis, Larry Goldstein, Steven Stasheff and Bill Wilson for providing helpful comments during discussions of this study, and Toni Shaw for assistance in preparation of the manuscript. This research was supported through a VA Research Service Award, and Grant IIRG-88-01 from the ADRDA.
transmission affects rate of behavioral learning, Science, 224 (1984) 627-630. 6 Berger, T.W. and Sclabassi, R.J., Long-term potentiation and its relation to hippocampal pyramidal cell activity and behavioral learning during classical conditioning. In P.W. Landfield and S.A. Deadwyler (Eds.), Long-term Potentiation: From Biophysics to Behavior, Alan R. Liss, New York, 1988, pp. 467-498. 7 Bliss, T.V.P. and Lynch, M.A., Long-term potentiation of synaptic transmission in the hippocampus: properties and mechanisms. In P.W. Landfield and S.A. Deadwyler (Eds.), LongTerm Potentiation: From Biophysics to Behavior, Alan R. Liss, New York, 1988, pp. 3-72. 8 Bondareff, W., Synaptic atrophy in the senescent hippocampus, Mech. Ageing Dev., 9 (1979) 163-171. 9 Castro, C.A., Silbert, L.H., McNaughton, B.L. and Barnes, C.A., Recovery of spatial learning deficits after decay of electrically induced synaptic enhancement in the hippocampus, Nature, 342 (1989) 545-548. 10 Collingridge, G.L., Kehl, S.J. and McLennan, H., Excitatory
9 amino acids in synaptic transmission in the Schaffer collateralcommissural pathway of the rat hippocampus, J. Physiol., 334 (1983) 33-46. 11 Cotman, C.W. and Anderson, K.J., Synaptic plasticity and functional stabilization in the hippocampal formation: possible role in Alzheimer's Disease. In G.S. Waxman (Ed.), Advances in Neurology (Vol. 47): Functional Recovery in Neurological Disease, Raven Press, New York, 1988, pp. 313-335. 12 Cotman, C. and Monaghan, D., Multiple excitatory amino acid receptor regulation of intracellular Ca2÷: implications for aging and Alzheimer's Disease, Ann. N.Y. Acad. Sci., 568 (1989) 138-148. 13 Creasey, H. and Rapoport, S.I., The aging human brain, Ann. Neurol., 17 (1985) 2-10. 14 De Toledo-Morreil, L. and Morrell, E, Electrophysiological markers of aging and memory loss in rats, Ann. N. Y. Acad. Sci., 444 (1985) 296-311. 15 Gage, EH. and Bjrrklund, A., Cholinergic septal grafts into the hippocampal formation improve spatial learning and memory in aged rats by an atropine-sensitive mechanism, J. Neurosci., 6 (1986) 2837-2847. 16 Gage, EH., Dunnett, S.S. and Bjrrklund, A., Spatial learning and motor deficits in aged rats, Neurobiol. Aging, 5 (1984) 43-48. 17 Geinisman, Y. and Bondareff, W., Decreases in the number of synapses in the senescent brain: a quantitative electron microscopic analysis of the dentate gyrus molecular layer in the rat, Mech. Ageing Dev., 5 (1976) 11-23. 18 Geinisman, Y., Bondareff, W. and Dodge, J.T., Partial deafferentation of neurons in the dentate gyrus of the senescent rat, Brain Research, 134 (1977) 541-545. 19 Giaquinto, S., Aging and Nervous System, John Wiley, New York, 1988. 20 Gibson, G.E. and Peterson, C., Calcium and the aging nervous system, Neurobiol. Aging, 8 (1987) 329-343. 21 Horel, J.A., The neuroanatomy of memory, Brain, 101 (1978) 403-445. 22 Hyman, B.T., Van Hoesen, G.W., Damasio, A.R. and Barnes, C.L., Alzheimer's Disease: cell-specific pathology isolates the hippocampal formation, Science, 225 (1984) 1168-1170. 23 Landfield, P.W., Correlative studies of brain neurophysiology and behavior during aging. In D.G. Stein (Ed.), The Psychobiology of Aging: Problems and Perspectives, Elsevier, New York, 1980, pp. 227-251. 24 Landfield, P.W., Campbell, L.W., Hao, S. and Kerr, S., Aging-related increases in voltage-sensitive, inactivating currents in rat hippocampus: implications for mechanisms of brain aging and Alzheimer's Disease, Ann. N.Y. Acad. Sci., 568 (1989) 95-105. 25 Landfield, P.W. and Lynch, G., Impaired monosynaptic potentiation in vitro hippocampal slices from aged, memory-impaired rats, J. Gerontol., 32 (1977) 523-533. 26 Landfield, P.W., McGaugh, J.L. and Lynch, G., Impaired synaptic potentiation processes in the hippocampus of aged, memory-deficient rats, Brain Research, 150 (1978) 85-101. 27 Larson, J. and Lynch, G., Role of N-methyl-D-aspartate receptors in the induction of synaptic potentiation by burst stimulation patterned after the hippocampal 0-rhythm, Brain
Research, 441 "(1988) 111-118. 28 Lindner, M.D. and Schallert, T., Aging and atropine effects on spatial navigation in the Morris water task, Behav. Neurosci., 102 (1988) 621-634. 29 Luine, V. and Hearns, M., Spatial memory deficits in aged rats: contributions of the cholinergic system assessed by CHAT, Brain Research, 523 (1990) 321-324. 30 McNaughton, B.L., Barnes, C.A., Rao, G., Baldwin, J. and Rasmussen, M., Long-term enhancement of hippocampal synaptic transmission and the acquisition of spatial information, J. Neurosci., 6 (1986) 563-571. 31 Mitchell, S.J., Rawlins, J.N.P., Steward, O. and Olton, D.S., Medial septal lesions disrupt theta rhythm and cholinergic staining in medial entorhinal cortex and produce impaired radial arm maze behavior in rats, J. Neurosci., 2 (1982) 292-302. 32 Morris, R., Developments of a water-maze procedure for studying spatial learning in the rat, J. Neurosci. Methods, 11 (1984) 47-60. 33 Morris, R.G.M., Anderson, E., Lynch, G.S. and Baudry, M., Selective impairment of learning and blockade of long-term potentiation by an N-methyl-o-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. 34 Morris, R. and Baker, M., Does long-term potentiation/synaptic enhancement have anything to do with learning and memory? In L.R. Squires and N. Butters (Eds.), Neuropsychology of Memory, Guilford Press, New York, 1984, pp. 521-535. 35 Morris, R.G.M., Davis, S. and Butcher, S.P., The role of NMDA receptors in learning and memory. In J.C. Watkins and G.L. Collingridge (Eds.), The NMDA Receptor, IRL Press, Oxford, 1989, pp. 137-152. 36 Morris, R.G.M., Garrud, P., Rawlins, J.N.P. and O'Keefe, J., Place-navigation impaired in rats with hippocampal lesions, Nature, 297 (1982) 681-683. 37 O'Keefe, J. and Nadel, L.,The Hippocampus as a Cognitive Map, Clarendon-Oxford University Press, New York, 1978. 38 Olton, D.S., Walker, J.A. and Gage, EH., Hippocampal connections and spatial discrimination, Brain Research, 139 (1978) 295-308. 39 Peterson, C. and Cotman, C.W., Strain-dependent decrease in glutamate binding to the N-methyl-o-aspartic acid receptor during aging, Neurosci. Lett., 104 (1989) 309-313. 40 Rapp, ER., Rosenberg, R.A. and Gallagher, M., An evaluation of spatial information processing in aged rats, Behav. Neurosci., 101 (1987) 3-12. 41 Spencer, R.L., Ganem, J., Chapman, S., O'Steen, W.K. and McEwen, B.S., Evaluation of L-acetyl carnitine (LAC) on brain and behavioral measures in aged rats, Soc. Neurosci. Abstr., 15 (1989) 261. 42 Tielen, A.M., Mollevanger, W.J., Lopes da Silva, EH. and Hollander, C.E, Neuronal plasticity in hippocampal shces of extremely old rats. In W.H. Gispen and J. Traber (Eds.), Aging of the Brain, Elsevier, Amsterdam, 1983, pp. 73-84. 43 Turner, D.A. and Deupree, D.L., Functional elongation in CA1 pyramidal neurons in senescent F-344 rats, Neurobiol. Aging, in press. 44 Winson, J., Loss of hippocampal theta rhythm results in spatial memory deficit in the rat, Science, 201 (1978) 160-162.