- Email: [email protected]
Differential effects of benzodiazepine receptor agonists on hippocampal long-term potentiation and spatial learning in the Morris water maze
Brain Research, 626 (1993) 63-70 © 1993 Elsevier Science Publishers B.V. All rights reserved 0006-8993/93/$06.00
63
BRES 19325
Differential effects of benzodiazepine receptor agonists on hippocampal long-term potentiation and spatial learning in the Morris water maze Robert
K. McNamara
~, G r e g o r y
E. dePape
and Ronald
W. Skelton
Department of Psychology, University of Victoria, Victoria, B.C. (Canada) (Accepted 25 May 1993)
Key words: Diazepam; Chlordiazepoxide; CL 218,872; Long-term potentiation; Hippocampus; Spatial learning; Morris water maze
The amnesic effect of benzodiazepine drugs has been well documented, though the mechanisms mediating this effect are unknown. Long-term potentiation (LTP) has been proposed as a mechanism by which information is stored in the mammalian central nervous system. This experiment sought to determine if benzodiazepines impair mnemonic processes by blocking LTP. Rats implanted with a stimulating electrode in the perforant path and a recording electrode in the dentate gyrus were given high-frequency stimulation after the administration of either chlordiazepoxide (5 mg/kg), diazepam (5 mg/kg) or CL 218,872 (10 mg/kg). None of these drugs completely blocked the induction of LTP as measured by changes in the magnitude of the population spike amplitude, though CL 218,872 significantly suppressed potentiation over the duration of recording (24 h). Moreover, the potentiation observed in diazepam-treated rats returned to baseline after 24 h. Two weeks after the last recording, the same implanted rats were given their previous drug and dose and then tested for spatial learning ability in the Morris water maze. Each drug resulted in a severe impairment of spatial learning, but had no effect on cue learning. Two days later, in the absence of drugs, the same rats readily acquired a reversed platform location. Together these results suggest that CL 218,872 may impair spatial learning by suppressing LTP in the perforant path but that chlordiazepoxide and diazepam can impair spatial learning in the absence of LTP suppression in this pathway.
INTRODUCTION Soon after their discovery, benzodiazepines (BZ) were f o u n d to have amnesic properties (e.g., ref. 16). Since these early observations, n u m e r o u s reports in both h u m a n (reviewed in ref. 17) and animal (reviewed in ref. 11)literatures have confirmed that BZs p r o d u c e amnesia. M o r e specifically, BZs impair the transfer of information from short-term m e m o r y to long-term memory storage but spare retention/retrieval processes 11'17. T h e specificity with which BZs impair m e m o r y suggests that they are interfering with an e n d o g e n o u s storage mechanism. L o n g - t e r m potentiation (LTP) is an e n h a n c e m e n t of synaptic efficacy resulting from a brief episode of high-frequency stimulation and has m a n y properties indicative of an e n d o g e n o u s information storage mechanism. For example, L T P is e n d u r i n g 6, has associative properties 29 and has b e e n shown to a c c o m p a n y asso-
ciative learning 24"39m'45. Moreover, N M D A receptor blockade, which also prevents L T P induction ~2, impairs spatial learning 34 but not p e r f o r m a n c e in the Morris water maze ( M W M ) 18. Together, these results suggest that L T P does indeed play a role in the storage of certain types of information (spatial) in the m a m m a l i a n CNS. T h e r e are several reasons why it might be thought that BZs impair m n e m o n i c processes by blocking the induction of hippocampal LTP: (1) BZs resemble N M D A receptor antagonists in that they impair spatial learning but not p e r f o r m a n c e or cue learninglS.26,34; (2) d i a z e p a m inhibits depolarization-induced release of [3H]glutamate in hippocampal slices3; (3) maximal L T P is induced at the 0-frequency ( 5 - 7 Hz) 2° and BZs depress 0-frequency9,31; (4) BZs may prevent a series o f stimulations from producing maximal excitation by p r o l o n g i n g r e c u r r e n t inhibitionl.47; (5) G A B A antagonists 46 and B Z inverse-agonists 42 facilitate the
Correspondence: R.W. Skelton, University of Victoria, Department of Psychology,P.O. Box 3050, Victoria, B.C., Canada, V8W 3P5. Fax: (1) (604) 721-8929; e-mail: [email protected]. 1 Present address: Northwestern University, Department of Psychology, 2029 Sheridan Road, Evanston, IL 60208, USA.
~4 induction of L'FP in the hippocampus; (6) postsynaptic inhibition, induced either through the injection of hyperpolarizing current 2~, focal application of GABA ~s or commissural afferent stimulation 15 blocks the induction of LTP in the hippocampus; (7) LTP induction has been correlated with reductions of GABA release 5 and GABA-mediated inhibition ~9'22, although the sensitivity of GABA a receptors does not change37; and (8) area CA1 and the dentate gyrus both possess a high density of BZ receptors as. Together, these data strongly suggest that a reduction in GABA-mediated inhibition is required for LTP induction and that enhanced GABA-mediated inhibition resulting from BZs will suppress LTP induction. To date, the few studies that have assessed the effects of BZs on the induction of LTP have produced inconsistent results, with BZ agonists either blocking, having no effect or facilitating LTP induction in the hippocampus. For example, diazepam, triazolam H and lorazepam s were found to block the induction of LTP in the Schaffer collateral pathway (CA3--+ CA1) in hippocampal slices. Further, diazepam (5 and 10 r a g / kg) was found to suppress post-tetanic potentiation in the Schaffer collateral pathway 23 or in the CA1 region after high-frequency stimulation delivered to the medial septum of urethane-anaesthetized rats l°. Conversely, midazolam failed to block LTP in CAI in hippocampal slices 4 and diazepam (2-4 m g / k g ) failed to block LTP in CA1 in urethane-anaesthetized rats 44 but enhanced post-tetanic potentiation in the perforant pathway (PP):3. The effect of BZs on LTP induced in the PP ~ dentate pathway in the unanaesthetized rat has not been tested. A problem with the slice data, however, is that it is difficult to determine whether the BZ concentration found to block LTP in hippocampal slices is sufficient to impair mnemonic processes. Accordingly, the present experiment sought to examine the effects of CDP, diazepam and CL 218,872, a ligand with a selective affinity for the BZ~ receptor subtype 43, on the induction of LTP in the PP in unanaesthetized rats. To confirm that the doses tested were indeed amnestic, spatial learning was later tested in the same rats. In previous reports, PP transection 4° and the drugs and doses chosen for the present study 2'26'27'3° impaired spatial learning in the MWM, confirming both the importance of this pathway as well as the amnesic potency of these drugs. MATERIALS
AND METHODS
Animals Male l o n g - E v a n s hooded rats served as subjects. They were housed in pairs in shoebox cages and maintained on a 12/12 h light/dark cycle. Testing was conducted during the light phase of the
cycle. Food and w a t e r w e r c ctmtinuously available Ihe rals weighed approximately 350 g at the time of surgeis.
Surgeo' Under pentobarbital anesthesia (6(t m g / k g ) , chronic electrodes were implanted into the dentate gyrus ( 3.5 mm AP; 2.0 mm ML: -~3.4 m m DV from dural fl)r the recording of EEG and ew)ked potentials and into the PP (bipolar: - 7 . 9 mm AP; 4.2 mm ML; 2.9 mm DV from dural for the delivery of single pulses and highfrequency stimulation. Recording electrodes consisted of either a single strand of stainless-steel wire, 76 /xm in diameter, coated with enamel to a total diameter of 114 /zm or two single strands twisted together. Two skull screws, connected to uninsulated stainless-steel wire, served as current-return and ground-reference electrodes, Gold-plated pins {Amphenol 220-$02) were soldered to lhe ends of the electrodes, inserted into a connector (Mclntyre) and affixed to the skull using dental acrylic and two additional anchoring skull screws. Both PP and dentate electrodes were imphmted under elec trophysiological guidance. While PP and dentate gyrus electrodes were lowered, a storage oscilloscope displayed field potentials recordcd from the electrode in the dentate gyrus, evoked by single square wave pulses delivered to the PP electrode (0.1 ms, 0.2 ttz, 25{) #A). The positions and depths of the electrodes were adjusted to produce a dentate gyrus-evoked potential with a population spike of a maximum amplitude and minimum threshold, on lhe rising edge of the EPSP.
Drugs and group assignment Rats were treated with either chlordiazepoxide hydrochloride (5 m g / k g ; dissolved in 0.9% NaCI; Hoffman-La Roche; n = 6), diazepam (5 m g / k g ; suspended in saline with Tween-80, 1 d r o p / 1 0 ml; Hoffman-La Roche; n = 6), CL 218,872 (10 m g / k g ; suspended in saline with Tween-80; Cyanamid; n = 6), saline (0.9% NaCI; n = 6) or saline plus Tween-80 (1 d r o p / 1 0 ml; n = 2). All injections were administered i.p. 30 rain prior to either high-frequency stimulation or behavioral testing.
High-#equeno, stimulation One week after surgery, individual rats were placed in a grounded Faraday cage and connected to the electronics. A series of i n p u t / output ( I / O ) curves were recorded with five passes (one pass every 30 s) at each of the following intensities: 50, 100, 250, 500, 750 i*A. After two baseline I / O s were complete a drug was injected and 30 rain later a third, post-drug I / O was recorded. After the post-drug I / O , ten trains consisting of 10 pulses at 200 Hz were applied at 750 ~ A (interburst interval, 30 s). l / O s were recorded immediately, 30 min, 60 min, 24 h and 1 week after high-frequency stimulation. Evoked potentials recorded while the rat was moving were excluded from further analysis. Brain stimulation was controlled by, and evoked potentials were collected with, Brainwave hardware and the data were analyzed using Brainwave software. Population spike amplitude was determined by fitting a line tangent to the response wave form at two points, corresponding to the onset and offset of the spike and measuring the distance of a vertical line drawn from the tangent line to the trough (see Fig. 1A). An example of a population spike before and after high-frequency stimulation is illustrated in Fig. lB. For each of the sampling periods, an I / O curve was constructed for each rat by plotting the five population spike amplitudes against the five current intensities (see Fig. 2). Each curve was then reduced to a single number (the I / O area) by calculating the area between it and the X-axis as calculated by the graphing program Fig P (Version 6.0).
Behacioral testing Two weeks after high-frequency stimulation was administered, population spike amplitudes had returned to baseline levels (data not shown) and individual rats were administered their same drug and dose and trained in the M W M 33. During initial acquisition, the submerged escape platform was located in the center of the northwest quadrant. Training consisted of 20 consecutive trials followed by a probe trial and visible platform training (four trials). During each trial, the rat's swim path length (distance) and escape latency were
65 recorded with a video tracking system (Chromotrak, SD Instruments). Once the rat located the platform, it was permitted to remain on it for 15 s. If the rat did not locate the platform within 60 s, it was guided to it and allowed to remain on it for 15 s. After each trial, the rat was returned to a holding cage positioned 90 cm under a 250 W brooding lamp (for warmth) and allowed to remain there for the 45 s intertrial interval. Total run time for 20 trials ranged from 30 to 45 min. Two days following initial acquisition, undrugged rats were required to learn the location of the platform now located in the diagonally opposite quadrant. The latter reversal was conducted to determine if exposure to high-frequency stimulation and drugs produced a permanent learning impairment. Soon after maze training, rats were perfused transcardially, brains removed, sectioned and examined histologically. The position of the electrode tips were documented on corresponding coronal plates obtained from Paxinos and Watson 35.
Data analysis Differences of I / O area and swim path lengths were assessed using an analysis of variance (ANOVA) procedure with repeated measures. Post hoc comparisons were assessed using Dunnett's test.
12
A. Vehicle
10
•
8 6
Pre-Tetany Post-Tetany 136%
4 20
> E v
• "0 ::3 I~ '~
1816; 14. 12' 10"
B. CL 218,872
8" 4. 2. 0
A.
181416"
I
C.
I
l
I
I
CDP 1 3 ~ ~
12 10 468 ~ 2"
0
i
I
i
i
i
50
100
250
500
750
Intensity (gA) Fig. 2. Representative input/output curves from rats treated with (A) vehicle, (B) CL 218,872 and (C) chlordiazepoxide (CDP). All curves shown were measured 60 min after high-frequency stimulation.
-2
mv
Ims
B.
In every case the acceptable level of statistical significance was P < 0.05.
RESULTS
Fig. 1. A: a representative population spike illustrating how amplitude was measured from the trough of the wave vertically to the tangent joining the leading and trailing edges of the spike. B: representative wave forms taken before and after (60 min) highfrequency stimulation, illustrating LTP.
Electrode placements in the PP and dentate gyrus are illustrated in Fig. 3. Stimulating PP electrodes terminated in the angular bundle and dentate recording electrodes terminated in both the upper and lower blades of the dentate gyrus granule cell layer. The electrophysiological and behavioral responses of rats treated with saline and saline plus Tween-80 were very similar and their data were therefore combined. None of the drugs affected the size of the population spike prior to high-frequency stimulation. The mean postdrug population spike amplitudes (percent of baseline) were 99 + 3.5% (saline/saline + Tween), 98.5 + 3.0% (CDP), 99 + 1.3% (CL 218,872) and 96 + 2.9% (di-
66 azepam). None of these mean spike amplitudes differed significantly from baseline levels ( P > 0.05). The degree of LTP obtained in the dentate gyrus after systemically administered saline or CL 218,872 is illustrated in Fig. 4A. Rats treated with saline demonstrated a robust increase in the population spike amplitude that persisted for up to 24 h. However, rats treated with CL 218,872 showed only a small increase in population spike amplitude, which was less than that seen in saline controls. An overall A N O V A on the data indicated a significant group effect, Fl.l.~ = 4.3, P < 0.05, time effect, F6,TS= 13.7, P < 0.001 and group by time interaction, F6.7s = 2.9, P < 0.01. Nevertheless, at each of the sampling points after high-frequency stimulation, the population spike amplitudes of both
A.
G,I E
180"
"~
16o
~0
140
"5
Vehicle CL 218,872
0 •
12O
C ~ L..
188 B1
B2
P'D
5
30
24h
60
Time B. 2400 "
2000
¢,) 1600 1200 f-
ca
8o0
D
4OO i
i
I
,
i
i
i
i
[
1
2
3
4
5
6
7
8
9
Trial 50
i
10
Block
C.
40
3o
~,
lO
Vehicle
CL
218,872
Fig. 4. Effects of systemic CL 218,872 (10 mg/kg) on population spike amplitudes (expressed as I / O curve areas) (A), the distance taken to locate the submerged platform (blocks of two trials) (B) and performance during the probe trial (C). Note that CL 218,872 suppresses LTP induction and impairs spatial learning. (Arrow in A indicates point at which high-frequency stimulation applied; * P < 0.01 compared to chance in C). (B1 and B2 indicate the first and second baseline recordings; error bars = S.E.M.). Dashed line in C represents the chance level.
-7.30
/"
,i
Fig. 3. Representative placements (four rats from each treatment group) of recording electrodes in the dentate gyrus (top two coronal sections) and stimulating electrodes in the perforant path (bottom three sections). (Numbers are mm from bregma.)
vehicle and CL 218,872 groups differed significantly from baseline levels ( P < 0 . 0 5 ) showing that both groups exhibited LTP to some degree. Two weeks later, CL 218,872 also impaired spatial learning (Fig. 4B). Rats treated with the vehicle rapidly acquired the location of the platform while rats treated with CL 218,872 showed a severe impairment. An A N O V A on distance data revealed a significant group difference, F1j 8 = 30.3, P < 0.001, trial-block difference, F9,16 2 = 15.2, P < 0.001, as well as a significant group by trial block interaction, F9,162 = 2.2, P < 0.05. Performance on the probe trial confirmed the deficit, in that rats treated with CL 218,872 failed to show a bias for the correct quadrant (Fig. 4C) unlike controls who did ( P < 0.01 compared to chance levels). The mean total distance required by rats treated with CL 218,872 (185 + 25 cm) to reach the visible platform did not differ significantly from the mean total distance required by
67 rats treated with the drug vehicle (204 + 30 cm; P >
0.05). The LTP obtained in the dentate gyrus after systemically administered saline or CDP is illustrated in Fig. 5A. CDP did partially suppress LTP, though the magnitude of the potentiation was not significantly different from controls. An ANOVA on population spike amplitude data revealed a significant time difference, F6,78 = 15.2, P < 0.001, but not a significant group difference, F1,13= 1.1, n.s. or group by time interaction, F6,78 = 0.9, n.s. At all sampling phases, the population spike amplitudes of the CDP group differed significantly from baseline levels (P < 0.01). CDP did, however, impair spatial learning (Fig. 5B). Rats treated with CDP took longer swim paths than controls over the course of training. An ANOVA on distance data revealed a significant group difference, FI,16 = 70.8, P < 0.001, trial-block difference, F9,144 = 9.5, P < 0.001 and group by trial block interaction, F9,144 = 2.3, P < 0.01. The probe trial confirmed the CDP impairment
tO Vl
O Vehicle 160"
M
m
140 "
120 " E 0 ~
A,
180"
21
11111-
o
B1
. . . . . . . . . .
B2
P'D
5
30
6'0 24h
Time 2400 " A
B.
20001600 1
• 0 c
1200-
~
4000 Trial
Block
C.
40
4)
180'
4)
160'
r"
ao
A.
2o
O Vehicle • COP
~o
140" 0 '~ 4)
Vehicle
120" 100"
-
~
.....................................
4)
B'I
B'2
P'D
A
5
3'0
6'0 24h
Time 2400
E o
B.
2000' 1600
4) ¢:
1200 '
¢o
800
a
400.
0
;;;;;;;;;10 Trial
Block
C.
¢J
1 a
Dlazepam
Fig. 6. Effects of systemic diazepam (5 mg/kg) on population spike amplitudes (A), the distance taken to locate the submerged platform (B) and performance during the probe trial (C). Note that diazepam partially suppresses LTP induction but severely impairs spatial learning. (arrow in A indicates point at which high-frequency stimulation applied; * P < 0.01 compared to chance in C).
lO. 0 Vehicle
CDP
Fig. 5. Effects of systemic CDP (5 mg/kg) on population spike amplitudes (A), the distance taken to locate the submerged platform (B) and performance during the probe trial (C). Note that CDP only partially suppresses LTP induction but severely impairs spatial learning. (arrow in A indicates point at which high-frequency stimulation applied; * P < 0.01 compared to chance in C).
(Fig. 5C). Rats treated with vehicle, but not CDP, showed a significant bias for the correct quadrant (P < 0.01). The mean total distance required by rats treated with CDP (191 _+ 22 cm) to reach the visible platform did not differ from the mean total distance required by rats treated with the drug vehicle (as above; P > 0.05). The effects of diazepam on LTP and spatial learning are illustrated in Fig. 6. Diazepam partially suppressed LTP, though the magnitude of the potentiation did not differ significantly from control levels. An ANOVA on population spike data failed to reveal a significant group difference, F1A3 = 2.4, n.s. or group by time interaction, F6,78= 3.1, n.s., but there was a significant time difference, F6,78= 12.3, P < 0.001. Comparison with baseline levels revealed that the diazepam group's population spike amplitudes were significantly greater than baseline levels over the first three sampling phases (P < 0.05). However, the diazepam group's population spike amplitude did not differ significantly from base-
68
1000 E
800
¢"
400
'~ t-,t
200
A. - -o---* •
.\\
Vehicle CL 218,872
c?.
600
0 1
2
3
4
Trial
5
6
7
8
Block
B.
50
G) 0 C
40 30 20 ~o0
Vehicle
CL 218,672
CDP
Olezepam
Fig. 7. Performance of each treatment group previously given highfrequency stimulation and drugs on the distance taken to locate the submerged platform (now in southwest quadrant) (A) and performance during the probe trial (B). Note that each of the drug groups acquire the new platform location at a rate comparable to controls (* P < 0.01 compared to chance). line levels at the 24 h sampling phase and was significantly different from controls ( P < 0.05). Rats treated with diazepam showed a spatial learning deficit, taking longer swim paths to reach the escape platform (Fig. 6B). This was confirmed by an A N O V A which revealed a significant group difference, F~,~6= 40.1, P < 0.001, trial-block difference, F9,14 4 = 10.6, P < 0.001 and group by trial interaction, F9,144 = 2.3, P < 0.01. This impairment was confirmed during the probe trial (Fig. 6C) where only rats treated with vehicle showed a bias for the correct quadrant ( P < 0 . 0 1 compared to chance). The mean total distance required by rats treated with diazepam (211 +_ 31 cm) to reach the visible platform did not differ significantly from the mean total distance required by rats treated with the drug vehicle (as above; P > 0.05). During the drug-free reversal, all of the treatment groups acquired the new platform location at a rate comparable to controls (Fig. 7A). This was confirmed by an A N O V A which revealed a significant trial-block difference, F7,196 = 22.4, P < 0.001, but not a significant group difference, F2,28 = 1.6, n.s. or group by trial interaction, F2~,~96= 0.9, n.s. During the probe trial, all of the treatment groups showed a robust bias for the correct quadrant (Fig. 7B; P < 0.01), confirming that each group had acquired the platform's spatial location. DISCUSSION The present study found that acute treatment with CL 218,872, but not with CDP or diazepam, signifi-
cantly suppressed LTP at a dose that impaired spatial learning in the MWM. Rats treated with either CDP ol diazepam showed potentiation only slightly lower than that observed in rats treated with the vehicle, whereas rats treated with CL 218,872 showed only minimal potentiation. However, all rats treated with drugs showed a severe spatial learning impairment. The absence of an impairment on the visible platform task indicates that drug-treated rats were motivated to escape from the water and were capable of swimming in a goal directed and coordinated manner. Additionally, when these rats were undrugged, they were able to acquire the reversed platform location at rates comparable to controls, suggesting that the impairment observed during initial acquisition was not due to damage produced by the electrodes or by high-frequency stimulation. Thus, the drugs and doses tested in the present study were sufficient to impair spatial memory in the absence of sensorimotor disruption, replicating previous findings 2'2<~v'-~°. While it is possible that these drugs impaired acquisition by reducing core body temperature (compounding with the cold water), this account seems unlikely, since treated rats were kept under a heating lamp between trials and did not manifest behaviors indicative of chilling (shivering). Moreover, a previous report e(' found that the reduction of core body temperature produced by both diazepam and cold water were well below that required to produce amnesia alone ( - 5 ° C ) 3('. The failure of the BZs diazepam and CDP to block hippocampal LTP contrasts with previous reports (hippocampal slice, C A 1 ) s'H but is consistent with other reports in which neither midazolam (hippocampal slice, CA14) n o r diazepam (anaesthetized rats, CA144) blocked LTP-induction. This discrepancy does not appear to be related to the pathway stimulated, since BZs have failed to block LTP in both the Schaffer collateral pathway 4 and the PP (present findings), though differences in drug concentration and stimulation parameters may have contributed to the discrepancy. For example, LTP elicited by primed burst stimulation, but not LTP elicited by trains of high-frequency pulses, was blocked by serotonin ~3, suggesting that the amnesic doses of BZs used in the present study could impair LTP induced by alternative stimulation parameters. Assessment of amnesic doses of BZs on LTP induced by 0-patterned stimulation are warranted, particularly since BZs block both hippocampal 0-rhythm '~'-~[ and LTP elicited by 0-patterned stimulation in vitro ~4. Interestingly, the dose of diazepam that failed to block LTP induction in the present study is much greater than that required to prolong recurrent inhibition in the P P - d e n t a t e pathway j, suggesting that increasing
69 the inhibitory threshold is not by itself sufficient to prevent LTP induced by high-frequency stimulation. The finding that CL 218,872, a triazolopyridazine with a selective affinity for BZ 1 receptors, impairs LTP induction in the PP raises the possibility that selective activation of BZ 1 receptors has a greater suppressing effect on LTP than does activation of both BZ 1 and BZ 2 receptors by diazepam and CDP. While this may reflect a difference in relative drug concentrations, different receptor subtype distribution may also be a contributing factor. For example, although BZ 1 and BZ 2 receptors are proportionately equal in the hippocampus (50/50), BZ 1 receptors appear to be primarily postsynaptic whereas BZ 2 receptors are primarily presynaptic 32. It is therefore possible that the selective activation of BZ 1 receptors with CL 218,872 produces a greater hyperpolarizing effect on postsynaptic cells than does concurrent BZ~ and BZ 2 activation by CDP and diazepam and therefore a greater suppressing effect on LTP 21'38. The concurrent activation of presynaptic BZ 2 receptors with CDP and diazepam may reduce the release of inhibitory neurotransmitter (GABA?), preempting maximal postsynaptic hyperpolarization and maximal LTP suppression. However, future research employing multiple doses of each drug would be required to determine their relative efficacies on LTP suppression. The present study found that CL 218,872 both suppressed LTP induction and impaired spatial learning, suggesting that the BZ 1 receptor is an important modulator of synaptic plasticity. This notion is corroborated by the finding that the inverse-agonist /3-carboline, which also binds preferentially to the BZ~ receptor 7, was found to both enhance LTP 42 and facilitate spatial learning in the MWM 28. Thus, although BZ t receptors have differential regional and neuronal distributions relative to BZ 2 receptors, their selective activation is sufficient to reduce the neuronal excitability required for long-term synaptic changes. Consistent then is the recent finding that CL 218,872, like other BZ receptor agonists, retarded kindling 25, a durable form of synaptic plasticity. Transection of the PP impairs spatial learning in the MWM 4° and AP5, an NMDA receptor antagonist which blocks LTP in this pathway, also impairs spatial learning in the MWM 34, suggesting that LTP in this pathway is required for spatial learning. The data from the present study do not uniformly support this contention. Consistent with this notion is the finding that CL 218,872 suppressed LTP and impaired spatial learning. However, the differential effects of both diazepam and CDP on LTP induction and spatial learning suggest that spatial information storage is not de-
pendent on LTP in this pathway. It remains possible, however, that the LTP induced by high-frequency stimulation does not accurately mimic LTP subserving spatial learning, which might be blocked by CDP and diazepam. In sum, the present experiment found that CL 218,872, a selective BZ~ receptor agonist, suppressed, while CDP and diazepam had little effect on, LTP induction in the PP-dentate pathway. The same drug doses were also found to, in the same rats, severely impair spatial learning in the MWM. Together these findings suggest that suppression of PP-dentate LTP elicited by stimulus trains is not required to produce a spatial learning deficit in the MWM. Acknowledgments. This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada. The authors would like to thank Hoffmann-La Roche for their generous contribution of chlordiazepoxide and diazepam and American Cyanamid Co. for the CL 218,872. The authors thank M.E. Corcoran for his helpful comments on an earlier draft of this manuscript.
REFERENCES 1 Adamec, R.E., McNaughton, B., Racine, R. and Livingston, K.E., Effects of diazepam on hippocampal excitability in the rat: action in the dentate area, Epilepsia, 22 (1981) 205-215. 2 Arolfa, M.P. and Brioni, J.D., Diazepam impairs place learning in the Morris water maze, Behat,. Neural BioL, 55 (1991) 131-136. 3 Baba, A., Okumura, S., Mizou, H. and Iwata, H., Inhibition by diazepam and y-aminobutyric acid of depolarization-induced release of [14C]cysteine sulfinate and [3H]glutamate in rat hippocampal slices, J. Neurochem., 40 (1983) 280-284. 4 Birnstiel, S. and Haas, H.L., Anticonvulsants do not suppress long-term potentiation (LTP) in the rat hippocampus, Neurosci. Lett., 122 (1991) 61-63. 5 Bliss, T.V.P., Douglas, R.M., Errington, M.L and Lynch, M.A., Correlation between long-term potentiation and release of endogenous amino acids from dentate gyrus of anaesthetized rats, J. Physiol., 377 (1986) 391-408. 6 Bliss, T.V.P. and Lomo, T., Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path, J. Physiol., 232 (1973) 331-356. 7 Braestrup, C. and Nielson, M., [3H]propyl /3-carboline-3carboxylate as a selective radioligand for the BZ 1 benzodiazepine receptor subclass, J. Neurochem., 37 (1981) 333-341. 8 Brown, M.W., Riches, I.P., Cairns, N.J. and Smithson, J.E., Memory-affecting drugs and hippocampal synaptic plasticity. In H.L. Haas and G. Buzsaki (Eds.), Synaptic Plasticity in the Hippocampus, Springer, Berlin, 1988, pp. 146-149. 9 Caudarella, M., Durkin, T., Galey, D., Jeantet, Y. and Jaffard, R., The effect of diazepam on hippocampal EEG in relation to behavior, Brain Res., 435 (1987) 202-212. 10 Clement-Cormier, Y., DeFrance, J., Divakaran, P., Stanley, J., Taber, K. and Marchand, J., Inhibition of cyclic nucleotide accumulation following hippocampal potentiation: effects of diazepam, J. Neurosci. Res., 5 (1980) 531-536. 11 Cole, S.O., Effects of benzodiazepines on acquisition and performance: a critical assessment, Neurosci. Biobehac. Ret,., 10 (1986) 265-272. 12 Collingridge, G.L., Kehl, S.J. and McLennan, A., Excitatory amino-acids in synaptic transmission in the Schaffer colleteralcommissural pathway of the rat hippocampus, J. Physiol., 334 (1983) 33-46.
7(I 13 Corradetti, R., Balterini, L., Pugtiese, A.M. and Pepeu. G., Serotonin blocks the long-term potentiation induced by primed burst stimulation in the CA1 region of rat hippocampal slices, Neuroscience, 46 (1992) 511-518. 14 Del Cerro, S., Jung, M. and Lynch, G., Benzodiazepines block long-term potentiation in slices of hippocampus and piriform cortex, Neuroscience, 49 ( 1991) 1-6. 15 Douglas, R.M., Goddard, G.V. and Riives, M., Inhibitory modulation of long-term potentiation: evidence for a postsynaptic locus of control, Brain Res., 240 (1982) 259 272. 16 Feldman, S.A., A comparative study of four premedications, Anaesthesia, 18 (1963) 169. 17 Ghoneim, M.M. and Mewaldt, S.P., Benzodiazepines and human memory: a review, Anesthesiology, 72 (1990) 926-938. 18 Heale, V. and Harley, C., MK-801 and AP5 impair acquisition, but not retention, of the Morris milk maze, Pharmacol. Biochem. Behar., 36 (1990) 145-149. 19 Kanda, M., Maru, E., Ashida, fi., Tatsuno, O. and Takatani, O., Effects of LTP-inducing tetanic stimulations of the perforant path on the commissural inhibition and facilitation of dentate granule gell discharge, Brain Res., 484 (1989) 325-332. 20 Larson, J., Wong, D. and Lynch, G., Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation, Brain Res., 368 (1986) 347-350. 21 Malinow, R. and Miller, J.P., Postsynaptic hyperpolarization during conditioning reversibly blocks induction of long-term potentiation, Nature, 320 (1986) 529-530. 22 Maru, E., Ashida, H. and Tatsuno, J., Long-lasting reduction of dentate paired-pulse depression following LTP-inducing tetanic stimulations of perforant path, Brain Res., 478 (1986) 112-120. 23 Matthews, W.D. and Connor, J.D., Effects of diphenylhydantoin and diazepam on hippocampal evoked responses, Neuropharmacology, 15 (1976) 181-186. 24 Matthies, H.H., Ruethrich, T., Ott, H.K., Matthies, H.K. and Matthies, R., Low frequency perforant path stimulation as a conditioned stimulus demonstrates correlates between long-term potentiation, Physiol. Behat~., 36 (1986)811-821. 25 McNamara, R.K. and Corcoran, M.E. CL 218,872, a triazolopyridazine with a selective affinity for the w~ benzodiazepine receptor subtype, retards the development and expression of amygdaloid-kindled seizures: effects of flumazenil, Epilepsy Res., in press. 26 McNamara, R.K. and Skelton, R.W., Diazepam impairs acquistion but not performance in the Morris water maze, Pharmacol. Bioehem. Behat., 38 (1991) 651-658. 27 McNamara, R.K. and Skelton, R.W., Like diazepam, CL 218,872, a selective ligand for the benzodiazepine Wl receptor subtype, impairs place learning in the Morris water maze, Psychopharmacology, 107 (1992) 347-351. 28 McNamara, R.K. and Skelton, R.W., Benzodiazepine receptor antagonists flumazenil and CGS 8216 and inverse-agonist/3-CCM enhance spatial learning in the rat: dissociation from anxiogenic effects, Psychobiology, 21 (1993) 101-108. 29 McNaughton, B.L., Douglas, R.M. and Goddard, G.V., Synaptic enhancement in fascia dentata: cooperativity among coactive afferents, Brain Res., 157 (1978) 277-293. 30 McNaughton, N. and Morris, R.G.M., Chlordiazepoxide, an anxiolytic benzodiazepine, impairs place navigation in rats, Behalf. Brain Res., 24 (1987) 39-46. 31 McNaughton, N.. Richardson, J. and Gore, C., Reticular elicita-
tion of hippocampal slow waves: common effecls of some anxiolytic drugs, Neuroscience, 19 (1986) 899-903. 32 Medina, JH., Novas, M.L. and De Robertis, E., t|eterogeneity ol benzodiazepine receptors: experimental differences between [~H]flunitrazepam and [~H]ethyl-/J-carboline-3-carboxylate binding sites in rat brain membranes, J. Neurochem., 41 (1983) 7{)3709.
33 Morris, R.G.M., Developments of a water-maze procedure for studying spatial learning in the rat. J. Neurosci. Meth., 11 (1984) 47-60. 34 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-D-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. 35 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic Press, Orlando, FL, 1988. 36 Richardson, R., Riccio, D.C. and Morilak, D., Anterograde memory loss induced by hypothermia in rats, Behm'. Neural Biol., 37 (1983) 76-88. 37 Scharfman, H.E. and Sarvey, J.M., y-Aminobutyrate sensitivity does not change during long-term potentiation in rat hippocampal slices, Neuroscience, t5 (1985)695-702. 38 Scharfman, H.E. and Sarvey, J.M., Postsynaptic firing during repetitive stimulation is required for long-term potentiation in hippocampus, Brain Res., 331 (1985) 267-274. 39 Sharp, P.E., McNaughton, B.L. and Barnes, C.A., Explorationdependent modulation of evoked responses in fascia dentata: fundamental observations and time course, Psychobiology, 17 (1989) 257-269. 40 Skelton, R.W. and McNamara, R.K., Bilateral knife cuts to the perforant path disrupt spatial learning in the Morris water maze, Hippocampus, 2 (1992) 73-80. 41 Skelton, R.W., Scarth, A.S., Wilkie, D.M., Miller, J.J. and Phillips, A.G., Ix)ng-term increases in dentate granule cell responsivity accompany operant conditioning, J. Neurosci., 7 (1987) 3081-3087. 42 Sorensen, S., M., Zwolshen, J.M. and Humphreys, T.M., An electrophysiological comparison of MDL 26,479, a novel triazole and beta carbolines in the hippocampal slice: evidence for cognitive enhancing potential, Soc. Neurosci. Abstr., 16 (1990) 1331. 43 Squires, R.F., Benson, D.I., Braestrup, C., Coupet, J., Klepner, C.A., Myers, V. and Beer, B., Some properties of brain specific benzodiazepine receptors: new evidence for multiple receptors, Pharmacol. Biochem. Behat,., 111 (1979)825-830. 44 Stringer, J.L. and Guyenet, P.G., Elimination of long-term potentiation in the hippocampus by phencyclidine and ketamine, Brain Res., 258 (1983) 159-164. 45 Weisz, D.G., Clark, G.A. and Thompson, R.F., Increased responsivity of dentate granule cells during nictitating membrane response conditioning in rabbit, Behat,. Brain Res., 12 (1984) 145 154. 46 Wigstrom, H. and Gustafsson, B., Facilitated induction of hippocampal long-lasting potentiation during blockade of inhibition, Nature, 301 (1983) 603-604. 47 Wolf, P. and Haas, H.L., Effects of diazepines and barbiturates on hippocampal recurrent inhibition, Naunyn-Schmiedeberg's Arch. Pharmacol., 299 (1977) 211-218. 48 Young, W.S. and Kuhar, M.J., Radiohistochemical localiziation of benzodiazepine receptors in rats brain, J. Pharm. Exp. Therap., 212 (1980) 337-346.
63
BRES 19325
Differential effects of benzodiazepine receptor agonists on hippocampal long-term potentiation and spatial learning in the Morris water maze Robert
K. McNamara
~, G r e g o r y
E. dePape
and Ronald
W. Skelton
Department of Psychology, University of Victoria, Victoria, B.C. (Canada) (Accepted 25 May 1993)
Key words: Diazepam; Chlordiazepoxide; CL 218,872; Long-term potentiation; Hippocampus; Spatial learning; Morris water maze
The amnesic effect of benzodiazepine drugs has been well documented, though the mechanisms mediating this effect are unknown. Long-term potentiation (LTP) has been proposed as a mechanism by which information is stored in the mammalian central nervous system. This experiment sought to determine if benzodiazepines impair mnemonic processes by blocking LTP. Rats implanted with a stimulating electrode in the perforant path and a recording electrode in the dentate gyrus were given high-frequency stimulation after the administration of either chlordiazepoxide (5 mg/kg), diazepam (5 mg/kg) or CL 218,872 (10 mg/kg). None of these drugs completely blocked the induction of LTP as measured by changes in the magnitude of the population spike amplitude, though CL 218,872 significantly suppressed potentiation over the duration of recording (24 h). Moreover, the potentiation observed in diazepam-treated rats returned to baseline after 24 h. Two weeks after the last recording, the same implanted rats were given their previous drug and dose and then tested for spatial learning ability in the Morris water maze. Each drug resulted in a severe impairment of spatial learning, but had no effect on cue learning. Two days later, in the absence of drugs, the same rats readily acquired a reversed platform location. Together these results suggest that CL 218,872 may impair spatial learning by suppressing LTP in the perforant path but that chlordiazepoxide and diazepam can impair spatial learning in the absence of LTP suppression in this pathway.
INTRODUCTION Soon after their discovery, benzodiazepines (BZ) were f o u n d to have amnesic properties (e.g., ref. 16). Since these early observations, n u m e r o u s reports in both h u m a n (reviewed in ref. 17) and animal (reviewed in ref. 11)literatures have confirmed that BZs p r o d u c e amnesia. M o r e specifically, BZs impair the transfer of information from short-term m e m o r y to long-term memory storage but spare retention/retrieval processes 11'17. T h e specificity with which BZs impair m e m o r y suggests that they are interfering with an e n d o g e n o u s storage mechanism. L o n g - t e r m potentiation (LTP) is an e n h a n c e m e n t of synaptic efficacy resulting from a brief episode of high-frequency stimulation and has m a n y properties indicative of an e n d o g e n o u s information storage mechanism. For example, L T P is e n d u r i n g 6, has associative properties 29 and has b e e n shown to a c c o m p a n y asso-
ciative learning 24"39m'45. Moreover, N M D A receptor blockade, which also prevents L T P induction ~2, impairs spatial learning 34 but not p e r f o r m a n c e in the Morris water maze ( M W M ) 18. Together, these results suggest that L T P does indeed play a role in the storage of certain types of information (spatial) in the m a m m a l i a n CNS. T h e r e are several reasons why it might be thought that BZs impair m n e m o n i c processes by blocking the induction of hippocampal LTP: (1) BZs resemble N M D A receptor antagonists in that they impair spatial learning but not p e r f o r m a n c e or cue learninglS.26,34; (2) d i a z e p a m inhibits depolarization-induced release of [3H]glutamate in hippocampal slices3; (3) maximal L T P is induced at the 0-frequency ( 5 - 7 Hz) 2° and BZs depress 0-frequency9,31; (4) BZs may prevent a series o f stimulations from producing maximal excitation by p r o l o n g i n g r e c u r r e n t inhibitionl.47; (5) G A B A antagonists 46 and B Z inverse-agonists 42 facilitate the
Correspondence: R.W. Skelton, University of Victoria, Department of Psychology,P.O. Box 3050, Victoria, B.C., Canada, V8W 3P5. Fax: (1) (604) 721-8929; e-mail: [email protected]. 1 Present address: Northwestern University, Department of Psychology, 2029 Sheridan Road, Evanston, IL 60208, USA.
~4 induction of L'FP in the hippocampus; (6) postsynaptic inhibition, induced either through the injection of hyperpolarizing current 2~, focal application of GABA ~s or commissural afferent stimulation 15 blocks the induction of LTP in the hippocampus; (7) LTP induction has been correlated with reductions of GABA release 5 and GABA-mediated inhibition ~9'22, although the sensitivity of GABA a receptors does not change37; and (8) area CA1 and the dentate gyrus both possess a high density of BZ receptors as. Together, these data strongly suggest that a reduction in GABA-mediated inhibition is required for LTP induction and that enhanced GABA-mediated inhibition resulting from BZs will suppress LTP induction. To date, the few studies that have assessed the effects of BZs on the induction of LTP have produced inconsistent results, with BZ agonists either blocking, having no effect or facilitating LTP induction in the hippocampus. For example, diazepam, triazolam H and lorazepam s were found to block the induction of LTP in the Schaffer collateral pathway (CA3--+ CA1) in hippocampal slices. Further, diazepam (5 and 10 r a g / kg) was found to suppress post-tetanic potentiation in the Schaffer collateral pathway 23 or in the CA1 region after high-frequency stimulation delivered to the medial septum of urethane-anaesthetized rats l°. Conversely, midazolam failed to block LTP in CAI in hippocampal slices 4 and diazepam (2-4 m g / k g ) failed to block LTP in CA1 in urethane-anaesthetized rats 44 but enhanced post-tetanic potentiation in the perforant pathway (PP):3. The effect of BZs on LTP induced in the PP ~ dentate pathway in the unanaesthetized rat has not been tested. A problem with the slice data, however, is that it is difficult to determine whether the BZ concentration found to block LTP in hippocampal slices is sufficient to impair mnemonic processes. Accordingly, the present experiment sought to examine the effects of CDP, diazepam and CL 218,872, a ligand with a selective affinity for the BZ~ receptor subtype 43, on the induction of LTP in the PP in unanaesthetized rats. To confirm that the doses tested were indeed amnestic, spatial learning was later tested in the same rats. In previous reports, PP transection 4° and the drugs and doses chosen for the present study 2'26'27'3° impaired spatial learning in the MWM, confirming both the importance of this pathway as well as the amnesic potency of these drugs. MATERIALS
AND METHODS
Animals Male l o n g - E v a n s hooded rats served as subjects. They were housed in pairs in shoebox cages and maintained on a 12/12 h light/dark cycle. Testing was conducted during the light phase of the
cycle. Food and w a t e r w e r c ctmtinuously available Ihe rals weighed approximately 350 g at the time of surgeis.
Surgeo' Under pentobarbital anesthesia (6(t m g / k g ) , chronic electrodes were implanted into the dentate gyrus ( 3.5 mm AP; 2.0 mm ML: -~3.4 m m DV from dural fl)r the recording of EEG and ew)ked potentials and into the PP (bipolar: - 7 . 9 mm AP; 4.2 mm ML; 2.9 mm DV from dural for the delivery of single pulses and highfrequency stimulation. Recording electrodes consisted of either a single strand of stainless-steel wire, 76 /xm in diameter, coated with enamel to a total diameter of 114 /zm or two single strands twisted together. Two skull screws, connected to uninsulated stainless-steel wire, served as current-return and ground-reference electrodes, Gold-plated pins {Amphenol 220-$02) were soldered to lhe ends of the electrodes, inserted into a connector (Mclntyre) and affixed to the skull using dental acrylic and two additional anchoring skull screws. Both PP and dentate electrodes were imphmted under elec trophysiological guidance. While PP and dentate gyrus electrodes were lowered, a storage oscilloscope displayed field potentials recordcd from the electrode in the dentate gyrus, evoked by single square wave pulses delivered to the PP electrode (0.1 ms, 0.2 ttz, 25{) #A). The positions and depths of the electrodes were adjusted to produce a dentate gyrus-evoked potential with a population spike of a maximum amplitude and minimum threshold, on lhe rising edge of the EPSP.
Drugs and group assignment Rats were treated with either chlordiazepoxide hydrochloride (5 m g / k g ; dissolved in 0.9% NaCI; Hoffman-La Roche; n = 6), diazepam (5 m g / k g ; suspended in saline with Tween-80, 1 d r o p / 1 0 ml; Hoffman-La Roche; n = 6), CL 218,872 (10 m g / k g ; suspended in saline with Tween-80; Cyanamid; n = 6), saline (0.9% NaCI; n = 6) or saline plus Tween-80 (1 d r o p / 1 0 ml; n = 2). All injections were administered i.p. 30 rain prior to either high-frequency stimulation or behavioral testing.
High-#equeno, stimulation One week after surgery, individual rats were placed in a grounded Faraday cage and connected to the electronics. A series of i n p u t / output ( I / O ) curves were recorded with five passes (one pass every 30 s) at each of the following intensities: 50, 100, 250, 500, 750 i*A. After two baseline I / O s were complete a drug was injected and 30 rain later a third, post-drug I / O was recorded. After the post-drug I / O , ten trains consisting of 10 pulses at 200 Hz were applied at 750 ~ A (interburst interval, 30 s). l / O s were recorded immediately, 30 min, 60 min, 24 h and 1 week after high-frequency stimulation. Evoked potentials recorded while the rat was moving were excluded from further analysis. Brain stimulation was controlled by, and evoked potentials were collected with, Brainwave hardware and the data were analyzed using Brainwave software. Population spike amplitude was determined by fitting a line tangent to the response wave form at two points, corresponding to the onset and offset of the spike and measuring the distance of a vertical line drawn from the tangent line to the trough (see Fig. 1A). An example of a population spike before and after high-frequency stimulation is illustrated in Fig. lB. For each of the sampling periods, an I / O curve was constructed for each rat by plotting the five population spike amplitudes against the five current intensities (see Fig. 2). Each curve was then reduced to a single number (the I / O area) by calculating the area between it and the X-axis as calculated by the graphing program Fig P (Version 6.0).
Behacioral testing Two weeks after high-frequency stimulation was administered, population spike amplitudes had returned to baseline levels (data not shown) and individual rats were administered their same drug and dose and trained in the M W M 33. During initial acquisition, the submerged escape platform was located in the center of the northwest quadrant. Training consisted of 20 consecutive trials followed by a probe trial and visible platform training (four trials). During each trial, the rat's swim path length (distance) and escape latency were
65 recorded with a video tracking system (Chromotrak, SD Instruments). Once the rat located the platform, it was permitted to remain on it for 15 s. If the rat did not locate the platform within 60 s, it was guided to it and allowed to remain on it for 15 s. After each trial, the rat was returned to a holding cage positioned 90 cm under a 250 W brooding lamp (for warmth) and allowed to remain there for the 45 s intertrial interval. Total run time for 20 trials ranged from 30 to 45 min. Two days following initial acquisition, undrugged rats were required to learn the location of the platform now located in the diagonally opposite quadrant. The latter reversal was conducted to determine if exposure to high-frequency stimulation and drugs produced a permanent learning impairment. Soon after maze training, rats were perfused transcardially, brains removed, sectioned and examined histologically. The position of the electrode tips were documented on corresponding coronal plates obtained from Paxinos and Watson 35.
Data analysis Differences of I / O area and swim path lengths were assessed using an analysis of variance (ANOVA) procedure with repeated measures. Post hoc comparisons were assessed using Dunnett's test.
12
A. Vehicle
10
•
8 6
Pre-Tetany Post-Tetany 136%
4 20
> E v
• "0 ::3 I~ '~
1816; 14. 12' 10"
B. CL 218,872
8" 4. 2. 0
A.
181416"
I
C.
I
l
I
I
CDP 1 3 ~ ~
12 10 468 ~ 2"
0
i
I
i
i
i
50
100
250
500
750
Intensity (gA) Fig. 2. Representative input/output curves from rats treated with (A) vehicle, (B) CL 218,872 and (C) chlordiazepoxide (CDP). All curves shown were measured 60 min after high-frequency stimulation.
-2
mv
Ims
B.
In every case the acceptable level of statistical significance was P < 0.05.
RESULTS
Fig. 1. A: a representative population spike illustrating how amplitude was measured from the trough of the wave vertically to the tangent joining the leading and trailing edges of the spike. B: representative wave forms taken before and after (60 min) highfrequency stimulation, illustrating LTP.
Electrode placements in the PP and dentate gyrus are illustrated in Fig. 3. Stimulating PP electrodes terminated in the angular bundle and dentate recording electrodes terminated in both the upper and lower blades of the dentate gyrus granule cell layer. The electrophysiological and behavioral responses of rats treated with saline and saline plus Tween-80 were very similar and their data were therefore combined. None of the drugs affected the size of the population spike prior to high-frequency stimulation. The mean postdrug population spike amplitudes (percent of baseline) were 99 + 3.5% (saline/saline + Tween), 98.5 + 3.0% (CDP), 99 + 1.3% (CL 218,872) and 96 + 2.9% (di-
66 azepam). None of these mean spike amplitudes differed significantly from baseline levels ( P > 0.05). The degree of LTP obtained in the dentate gyrus after systemically administered saline or CL 218,872 is illustrated in Fig. 4A. Rats treated with saline demonstrated a robust increase in the population spike amplitude that persisted for up to 24 h. However, rats treated with CL 218,872 showed only a small increase in population spike amplitude, which was less than that seen in saline controls. An overall A N O V A on the data indicated a significant group effect, Fl.l.~ = 4.3, P < 0.05, time effect, F6,TS= 13.7, P < 0.001 and group by time interaction, F6.7s = 2.9, P < 0.01. Nevertheless, at each of the sampling points after high-frequency stimulation, the population spike amplitudes of both
A.
G,I E
180"
"~
16o
~0
140
"5
Vehicle CL 218,872
0 •
12O
C ~ L..
188 B1
B2
P'D
5
30
24h
60
Time B. 2400 "
2000
¢,) 1600 1200 f-
ca
8o0
D
4OO i
i
I
,
i
i
i
i
[
1
2
3
4
5
6
7
8
9
Trial 50
i
10
Block
C.
40
3o
~,
lO
Vehicle
CL
218,872
Fig. 4. Effects of systemic CL 218,872 (10 mg/kg) on population spike amplitudes (expressed as I / O curve areas) (A), the distance taken to locate the submerged platform (blocks of two trials) (B) and performance during the probe trial (C). Note that CL 218,872 suppresses LTP induction and impairs spatial learning. (Arrow in A indicates point at which high-frequency stimulation applied; * P < 0.01 compared to chance in C). (B1 and B2 indicate the first and second baseline recordings; error bars = S.E.M.). Dashed line in C represents the chance level.
-7.30
/"
,i
Fig. 3. Representative placements (four rats from each treatment group) of recording electrodes in the dentate gyrus (top two coronal sections) and stimulating electrodes in the perforant path (bottom three sections). (Numbers are mm from bregma.)
vehicle and CL 218,872 groups differed significantly from baseline levels ( P < 0 . 0 5 ) showing that both groups exhibited LTP to some degree. Two weeks later, CL 218,872 also impaired spatial learning (Fig. 4B). Rats treated with the vehicle rapidly acquired the location of the platform while rats treated with CL 218,872 showed a severe impairment. An A N O V A on distance data revealed a significant group difference, F1j 8 = 30.3, P < 0.001, trial-block difference, F9,16 2 = 15.2, P < 0.001, as well as a significant group by trial block interaction, F9,162 = 2.2, P < 0.05. Performance on the probe trial confirmed the deficit, in that rats treated with CL 218,872 failed to show a bias for the correct quadrant (Fig. 4C) unlike controls who did ( P < 0.01 compared to chance levels). The mean total distance required by rats treated with CL 218,872 (185 + 25 cm) to reach the visible platform did not differ significantly from the mean total distance required by
67 rats treated with the drug vehicle (204 + 30 cm; P >
0.05). The LTP obtained in the dentate gyrus after systemically administered saline or CDP is illustrated in Fig. 5A. CDP did partially suppress LTP, though the magnitude of the potentiation was not significantly different from controls. An ANOVA on population spike amplitude data revealed a significant time difference, F6,78 = 15.2, P < 0.001, but not a significant group difference, F1,13= 1.1, n.s. or group by time interaction, F6,78 = 0.9, n.s. At all sampling phases, the population spike amplitudes of the CDP group differed significantly from baseline levels (P < 0.01). CDP did, however, impair spatial learning (Fig. 5B). Rats treated with CDP took longer swim paths than controls over the course of training. An ANOVA on distance data revealed a significant group difference, FI,16 = 70.8, P < 0.001, trial-block difference, F9,144 = 9.5, P < 0.001 and group by trial block interaction, F9,144 = 2.3, P < 0.01. The probe trial confirmed the CDP impairment
tO Vl
O Vehicle 160"
M
m
140 "
120 " E 0 ~
A,
180"
21
11111-
o
B1
. . . . . . . . . .
B2
P'D
5
30
6'0 24h
Time 2400 " A
B.
20001600 1
• 0 c
1200-
~
4000 Trial
Block
C.
40
4)
180'
4)
160'
r"
ao
A.
2o
O Vehicle • COP
~o
140" 0 '~ 4)
Vehicle
120" 100"
-
~
.....................................
4)
B'I
B'2
P'D
A
5
3'0
6'0 24h
Time 2400
E o
B.
2000' 1600
4) ¢:
1200 '
¢o
800
a
400.
0
;;;;;;;;;10 Trial
Block
C.
¢J
1 a
Dlazepam
Fig. 6. Effects of systemic diazepam (5 mg/kg) on population spike amplitudes (A), the distance taken to locate the submerged platform (B) and performance during the probe trial (C). Note that diazepam partially suppresses LTP induction but severely impairs spatial learning. (arrow in A indicates point at which high-frequency stimulation applied; * P < 0.01 compared to chance in C).
lO. 0 Vehicle
CDP
Fig. 5. Effects of systemic CDP (5 mg/kg) on population spike amplitudes (A), the distance taken to locate the submerged platform (B) and performance during the probe trial (C). Note that CDP only partially suppresses LTP induction but severely impairs spatial learning. (arrow in A indicates point at which high-frequency stimulation applied; * P < 0.01 compared to chance in C).
(Fig. 5C). Rats treated with vehicle, but not CDP, showed a significant bias for the correct quadrant (P < 0.01). The mean total distance required by rats treated with CDP (191 _+ 22 cm) to reach the visible platform did not differ from the mean total distance required by rats treated with the drug vehicle (as above; P > 0.05). The effects of diazepam on LTP and spatial learning are illustrated in Fig. 6. Diazepam partially suppressed LTP, though the magnitude of the potentiation did not differ significantly from control levels. An ANOVA on population spike data failed to reveal a significant group difference, F1A3 = 2.4, n.s. or group by time interaction, F6,78= 3.1, n.s., but there was a significant time difference, F6,78= 12.3, P < 0.001. Comparison with baseline levels revealed that the diazepam group's population spike amplitudes were significantly greater than baseline levels over the first three sampling phases (P < 0.05). However, the diazepam group's population spike amplitude did not differ significantly from base-
68
1000 E
800
¢"
400
'~ t-,t
200
A. - -o---* •
.\\
Vehicle CL 218,872
c?.
600
0 1
2
3
4
Trial
5
6
7
8
Block
B.
50
G) 0 C
40 30 20 ~o0
Vehicle
CL 218,672
CDP
Olezepam
Fig. 7. Performance of each treatment group previously given highfrequency stimulation and drugs on the distance taken to locate the submerged platform (now in southwest quadrant) (A) and performance during the probe trial (B). Note that each of the drug groups acquire the new platform location at a rate comparable to controls (* P < 0.01 compared to chance). line levels at the 24 h sampling phase and was significantly different from controls ( P < 0.05). Rats treated with diazepam showed a spatial learning deficit, taking longer swim paths to reach the escape platform (Fig. 6B). This was confirmed by an A N O V A which revealed a significant group difference, F~,~6= 40.1, P < 0.001, trial-block difference, F9,14 4 = 10.6, P < 0.001 and group by trial interaction, F9,144 = 2.3, P < 0.01. This impairment was confirmed during the probe trial (Fig. 6C) where only rats treated with vehicle showed a bias for the correct quadrant ( P < 0 . 0 1 compared to chance). The mean total distance required by rats treated with diazepam (211 +_ 31 cm) to reach the visible platform did not differ significantly from the mean total distance required by rats treated with the drug vehicle (as above; P > 0.05). During the drug-free reversal, all of the treatment groups acquired the new platform location at a rate comparable to controls (Fig. 7A). This was confirmed by an A N O V A which revealed a significant trial-block difference, F7,196 = 22.4, P < 0.001, but not a significant group difference, F2,28 = 1.6, n.s. or group by trial interaction, F2~,~96= 0.9, n.s. During the probe trial, all of the treatment groups showed a robust bias for the correct quadrant (Fig. 7B; P < 0.01), confirming that each group had acquired the platform's spatial location. DISCUSSION The present study found that acute treatment with CL 218,872, but not with CDP or diazepam, signifi-
cantly suppressed LTP at a dose that impaired spatial learning in the MWM. Rats treated with either CDP ol diazepam showed potentiation only slightly lower than that observed in rats treated with the vehicle, whereas rats treated with CL 218,872 showed only minimal potentiation. However, all rats treated with drugs showed a severe spatial learning impairment. The absence of an impairment on the visible platform task indicates that drug-treated rats were motivated to escape from the water and were capable of swimming in a goal directed and coordinated manner. Additionally, when these rats were undrugged, they were able to acquire the reversed platform location at rates comparable to controls, suggesting that the impairment observed during initial acquisition was not due to damage produced by the electrodes or by high-frequency stimulation. Thus, the drugs and doses tested in the present study were sufficient to impair spatial memory in the absence of sensorimotor disruption, replicating previous findings 2'2<~v'-~°. While it is possible that these drugs impaired acquisition by reducing core body temperature (compounding with the cold water), this account seems unlikely, since treated rats were kept under a heating lamp between trials and did not manifest behaviors indicative of chilling (shivering). Moreover, a previous report e(' found that the reduction of core body temperature produced by both diazepam and cold water were well below that required to produce amnesia alone ( - 5 ° C ) 3('. The failure of the BZs diazepam and CDP to block hippocampal LTP contrasts with previous reports (hippocampal slice, C A 1 ) s'H but is consistent with other reports in which neither midazolam (hippocampal slice, CA14) n o r diazepam (anaesthetized rats, CA144) blocked LTP-induction. This discrepancy does not appear to be related to the pathway stimulated, since BZs have failed to block LTP in both the Schaffer collateral pathway 4 and the PP (present findings), though differences in drug concentration and stimulation parameters may have contributed to the discrepancy. For example, LTP elicited by primed burst stimulation, but not LTP elicited by trains of high-frequency pulses, was blocked by serotonin ~3, suggesting that the amnesic doses of BZs used in the present study could impair LTP induced by alternative stimulation parameters. Assessment of amnesic doses of BZs on LTP induced by 0-patterned stimulation are warranted, particularly since BZs block both hippocampal 0-rhythm '~'-~[ and LTP elicited by 0-patterned stimulation in vitro ~4. Interestingly, the dose of diazepam that failed to block LTP induction in the present study is much greater than that required to prolong recurrent inhibition in the P P - d e n t a t e pathway j, suggesting that increasing
69 the inhibitory threshold is not by itself sufficient to prevent LTP induced by high-frequency stimulation. The finding that CL 218,872, a triazolopyridazine with a selective affinity for BZ 1 receptors, impairs LTP induction in the PP raises the possibility that selective activation of BZ 1 receptors has a greater suppressing effect on LTP than does activation of both BZ 1 and BZ 2 receptors by diazepam and CDP. While this may reflect a difference in relative drug concentrations, different receptor subtype distribution may also be a contributing factor. For example, although BZ 1 and BZ 2 receptors are proportionately equal in the hippocampus (50/50), BZ 1 receptors appear to be primarily postsynaptic whereas BZ 2 receptors are primarily presynaptic 32. It is therefore possible that the selective activation of BZ 1 receptors with CL 218,872 produces a greater hyperpolarizing effect on postsynaptic cells than does concurrent BZ~ and BZ 2 activation by CDP and diazepam and therefore a greater suppressing effect on LTP 21'38. The concurrent activation of presynaptic BZ 2 receptors with CDP and diazepam may reduce the release of inhibitory neurotransmitter (GABA?), preempting maximal postsynaptic hyperpolarization and maximal LTP suppression. However, future research employing multiple doses of each drug would be required to determine their relative efficacies on LTP suppression. The present study found that CL 218,872 both suppressed LTP induction and impaired spatial learning, suggesting that the BZ 1 receptor is an important modulator of synaptic plasticity. This notion is corroborated by the finding that the inverse-agonist /3-carboline, which also binds preferentially to the BZ~ receptor 7, was found to both enhance LTP 42 and facilitate spatial learning in the MWM 28. Thus, although BZ t receptors have differential regional and neuronal distributions relative to BZ 2 receptors, their selective activation is sufficient to reduce the neuronal excitability required for long-term synaptic changes. Consistent then is the recent finding that CL 218,872, like other BZ receptor agonists, retarded kindling 25, a durable form of synaptic plasticity. Transection of the PP impairs spatial learning in the MWM 4° and AP5, an NMDA receptor antagonist which blocks LTP in this pathway, also impairs spatial learning in the MWM 34, suggesting that LTP in this pathway is required for spatial learning. The data from the present study do not uniformly support this contention. Consistent with this notion is the finding that CL 218,872 suppressed LTP and impaired spatial learning. However, the differential effects of both diazepam and CDP on LTP induction and spatial learning suggest that spatial information storage is not de-
pendent on LTP in this pathway. It remains possible, however, that the LTP induced by high-frequency stimulation does not accurately mimic LTP subserving spatial learning, which might be blocked by CDP and diazepam. In sum, the present experiment found that CL 218,872, a selective BZ~ receptor agonist, suppressed, while CDP and diazepam had little effect on, LTP induction in the PP-dentate pathway. The same drug doses were also found to, in the same rats, severely impair spatial learning in the MWM. Together these findings suggest that suppression of PP-dentate LTP elicited by stimulus trains is not required to produce a spatial learning deficit in the MWM. Acknowledgments. This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada. The authors would like to thank Hoffmann-La Roche for their generous contribution of chlordiazepoxide and diazepam and American Cyanamid Co. for the CL 218,872. The authors thank M.E. Corcoran for his helpful comments on an earlier draft of this manuscript.
REFERENCES 1 Adamec, R.E., McNaughton, B., Racine, R. and Livingston, K.E., Effects of diazepam on hippocampal excitability in the rat: action in the dentate area, Epilepsia, 22 (1981) 205-215. 2 Arolfa, M.P. and Brioni, J.D., Diazepam impairs place learning in the Morris water maze, Behat,. Neural BioL, 55 (1991) 131-136. 3 Baba, A., Okumura, S., Mizou, H. and Iwata, H., Inhibition by diazepam and y-aminobutyric acid of depolarization-induced release of [14C]cysteine sulfinate and [3H]glutamate in rat hippocampal slices, J. Neurochem., 40 (1983) 280-284. 4 Birnstiel, S. and Haas, H.L., Anticonvulsants do not suppress long-term potentiation (LTP) in the rat hippocampus, Neurosci. Lett., 122 (1991) 61-63. 5 Bliss, T.V.P., Douglas, R.M., Errington, M.L and Lynch, M.A., Correlation between long-term potentiation and release of endogenous amino acids from dentate gyrus of anaesthetized rats, J. Physiol., 377 (1986) 391-408. 6 Bliss, T.V.P. and Lomo, T., Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path, J. Physiol., 232 (1973) 331-356. 7 Braestrup, C. and Nielson, M., [3H]propyl /3-carboline-3carboxylate as a selective radioligand for the BZ 1 benzodiazepine receptor subclass, J. Neurochem., 37 (1981) 333-341. 8 Brown, M.W., Riches, I.P., Cairns, N.J. and Smithson, J.E., Memory-affecting drugs and hippocampal synaptic plasticity. In H.L. Haas and G. Buzsaki (Eds.), Synaptic Plasticity in the Hippocampus, Springer, Berlin, 1988, pp. 146-149. 9 Caudarella, M., Durkin, T., Galey, D., Jeantet, Y. and Jaffard, R., The effect of diazepam on hippocampal EEG in relation to behavior, Brain Res., 435 (1987) 202-212. 10 Clement-Cormier, Y., DeFrance, J., Divakaran, P., Stanley, J., Taber, K. and Marchand, J., Inhibition of cyclic nucleotide accumulation following hippocampal potentiation: effects of diazepam, J. Neurosci. Res., 5 (1980) 531-536. 11 Cole, S.O., Effects of benzodiazepines on acquisition and performance: a critical assessment, Neurosci. Biobehac. Ret,., 10 (1986) 265-272. 12 Collingridge, G.L., Kehl, S.J. and McLennan, A., Excitatory amino-acids in synaptic transmission in the Schaffer colleteralcommissural pathway of the rat hippocampus, J. Physiol., 334 (1983) 33-46.
7(I 13 Corradetti, R., Balterini, L., Pugtiese, A.M. and Pepeu. G., Serotonin blocks the long-term potentiation induced by primed burst stimulation in the CA1 region of rat hippocampal slices, Neuroscience, 46 (1992) 511-518. 14 Del Cerro, S., Jung, M. and Lynch, G., Benzodiazepines block long-term potentiation in slices of hippocampus and piriform cortex, Neuroscience, 49 ( 1991) 1-6. 15 Douglas, R.M., Goddard, G.V. and Riives, M., Inhibitory modulation of long-term potentiation: evidence for a postsynaptic locus of control, Brain Res., 240 (1982) 259 272. 16 Feldman, S.A., A comparative study of four premedications, Anaesthesia, 18 (1963) 169. 17 Ghoneim, M.M. and Mewaldt, S.P., Benzodiazepines and human memory: a review, Anesthesiology, 72 (1990) 926-938. 18 Heale, V. and Harley, C., MK-801 and AP5 impair acquisition, but not retention, of the Morris milk maze, Pharmacol. Biochem. Behar., 36 (1990) 145-149. 19 Kanda, M., Maru, E., Ashida, fi., Tatsuno, O. and Takatani, O., Effects of LTP-inducing tetanic stimulations of the perforant path on the commissural inhibition and facilitation of dentate granule gell discharge, Brain Res., 484 (1989) 325-332. 20 Larson, J., Wong, D. and Lynch, G., Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation, Brain Res., 368 (1986) 347-350. 21 Malinow, R. and Miller, J.P., Postsynaptic hyperpolarization during conditioning reversibly blocks induction of long-term potentiation, Nature, 320 (1986) 529-530. 22 Maru, E., Ashida, H. and Tatsuno, J., Long-lasting reduction of dentate paired-pulse depression following LTP-inducing tetanic stimulations of perforant path, Brain Res., 478 (1986) 112-120. 23 Matthews, W.D. and Connor, J.D., Effects of diphenylhydantoin and diazepam on hippocampal evoked responses, Neuropharmacology, 15 (1976) 181-186. 24 Matthies, H.H., Ruethrich, T., Ott, H.K., Matthies, H.K. and Matthies, R., Low frequency perforant path stimulation as a conditioned stimulus demonstrates correlates between long-term potentiation, Physiol. Behat~., 36 (1986)811-821. 25 McNamara, R.K. and Corcoran, M.E. CL 218,872, a triazolopyridazine with a selective affinity for the w~ benzodiazepine receptor subtype, retards the development and expression of amygdaloid-kindled seizures: effects of flumazenil, Epilepsy Res., in press. 26 McNamara, R.K. and Skelton, R.W., Diazepam impairs acquistion but not performance in the Morris water maze, Pharmacol. Bioehem. Behat., 38 (1991) 651-658. 27 McNamara, R.K. and Skelton, R.W., Like diazepam, CL 218,872, a selective ligand for the benzodiazepine Wl receptor subtype, impairs place learning in the Morris water maze, Psychopharmacology, 107 (1992) 347-351. 28 McNamara, R.K. and Skelton, R.W., Benzodiazepine receptor antagonists flumazenil and CGS 8216 and inverse-agonist/3-CCM enhance spatial learning in the rat: dissociation from anxiogenic effects, Psychobiology, 21 (1993) 101-108. 29 McNaughton, B.L., Douglas, R.M. and Goddard, G.V., Synaptic enhancement in fascia dentata: cooperativity among coactive afferents, Brain Res., 157 (1978) 277-293. 30 McNaughton, N. and Morris, R.G.M., Chlordiazepoxide, an anxiolytic benzodiazepine, impairs place navigation in rats, Behalf. Brain Res., 24 (1987) 39-46. 31 McNaughton, N.. Richardson, J. and Gore, C., Reticular elicita-
tion of hippocampal slow waves: common effecls of some anxiolytic drugs, Neuroscience, 19 (1986) 899-903. 32 Medina, JH., Novas, M.L. and De Robertis, E., t|eterogeneity ol benzodiazepine receptors: experimental differences between [~H]flunitrazepam and [~H]ethyl-/J-carboline-3-carboxylate binding sites in rat brain membranes, J. Neurochem., 41 (1983) 7{)3709.
33 Morris, R.G.M., Developments of a water-maze procedure for studying spatial learning in the rat. J. Neurosci. Meth., 11 (1984) 47-60. 34 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-D-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. 35 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic Press, Orlando, FL, 1988. 36 Richardson, R., Riccio, D.C. and Morilak, D., Anterograde memory loss induced by hypothermia in rats, Behm'. Neural Biol., 37 (1983) 76-88. 37 Scharfman, H.E. and Sarvey, J.M., y-Aminobutyrate sensitivity does not change during long-term potentiation in rat hippocampal slices, Neuroscience, t5 (1985)695-702. 38 Scharfman, H.E. and Sarvey, J.M., Postsynaptic firing during repetitive stimulation is required for long-term potentiation in hippocampus, Brain Res., 331 (1985) 267-274. 39 Sharp, P.E., McNaughton, B.L. and Barnes, C.A., Explorationdependent modulation of evoked responses in fascia dentata: fundamental observations and time course, Psychobiology, 17 (1989) 257-269. 40 Skelton, R.W. and McNamara, R.K., Bilateral knife cuts to the perforant path disrupt spatial learning in the Morris water maze, Hippocampus, 2 (1992) 73-80. 41 Skelton, R.W., Scarth, A.S., Wilkie, D.M., Miller, J.J. and Phillips, A.G., Ix)ng-term increases in dentate granule cell responsivity accompany operant conditioning, J. Neurosci., 7 (1987) 3081-3087. 42 Sorensen, S., M., Zwolshen, J.M. and Humphreys, T.M., An electrophysiological comparison of MDL 26,479, a novel triazole and beta carbolines in the hippocampal slice: evidence for cognitive enhancing potential, Soc. Neurosci. Abstr., 16 (1990) 1331. 43 Squires, R.F., Benson, D.I., Braestrup, C., Coupet, J., Klepner, C.A., Myers, V. and Beer, B., Some properties of brain specific benzodiazepine receptors: new evidence for multiple receptors, Pharmacol. Biochem. Behat,., 111 (1979)825-830. 44 Stringer, J.L. and Guyenet, P.G., Elimination of long-term potentiation in the hippocampus by phencyclidine and ketamine, Brain Res., 258 (1983) 159-164. 45 Weisz, D.G., Clark, G.A. and Thompson, R.F., Increased responsivity of dentate granule cells during nictitating membrane response conditioning in rabbit, Behat,. Brain Res., 12 (1984) 145 154. 46 Wigstrom, H. and Gustafsson, B., Facilitated induction of hippocampal long-lasting potentiation during blockade of inhibition, Nature, 301 (1983) 603-604. 47 Wolf, P. and Haas, H.L., Effects of diazepines and barbiturates on hippocampal recurrent inhibition, Naunyn-Schmiedeberg's Arch. Pharmacol., 299 (1977) 211-218. 48 Young, W.S. and Kuhar, M.J., Radiohistochemical localiziation of benzodiazepine receptors in rats brain, J. Pharm. Exp. Therap., 212 (1980) 337-346.