- Email: [email protected]
Duration of retrograde amnesia induced by tetrodotoxin inactivation of the parabrachial nuclei is inversely related to the intensity of footshock in rat's passive avoidance response
Behavioural Brain Research, 49 (1992) 175-180 9 1992 Elsevier Science Publishers B.V. All rights reserved. 0166-4328/92/$05.00
175
BBR 01302
Duration of retrograde amnesia induced by tetrodotoxin inactivation of the parabrachial nuclei is inversely related to the intensity of footshock in rat's passive avoidance response Corrado Bucherelli and Giovanna Tassoni Diparthnento di Scienze Fisiologiche, Firenze (Italy) (Received 26 September 1991) (Revised version received 20 December 1992) (Accepted 15 April 1992)
Key words: Memory consolidation; Parabrachial nucleus; Footshock intensity; Tetrodotoxin; Functional ablation
Previous work has'demonstrated that in rats post-trial bilateral functional blockade of the parabrachial nuclei by local injection of tetrodotoxin (10 ng in 1 lal) partially disrupts retention of an overtrained passive avoidance response. The time-course of functional blockade disrupting effects on passive avoidance response has been studied by altering both the acquisition-tetrodotoxin injection interval and footshock intensity as independent variables. When weak footshocks (0.8 mA) were delivered, functional ablation of the parabrachial nuclei was still effective when induced up to 8 days after acquisition training. On the other hand, when rats were shocked with the stronger stimuli (1.2 mA), 2 days after acquisition functional ablation was no longer effective. The results are discussed in terms of unconditioned stimulus (US) intensity and engram consolidation.
INTRODUCTION
Previous work has shown that bilateral reversible blockade of the parabrachial nuclei (PBN) by tetrodotoxin (TTX) injected shortly after the acquisition session, disrupts inhibitory conditioned responses ~'~z'19. TTX has a well-known pharmacological action, namely the specific blocking of sodium channels. This type of PBN blockade has a significant amnesic effect on conditioned taste aversion (CTA) when induced up to 4 days after the taste-poison pairing ~2. Similarly the passive avoidance response (PAR) was disrupted with the difference that the amnesic effect was weaker and could be elicited by TTX applied 24 h but not 48 h after acquisition TM. These quantitative differences between CTA and PAR disruption were tentatively explained on the assumption that the PBN and contiguous reticular formation are directly involved in the representation of CTA, but that they participate less directly in PAR 19. On the other hand, overtrained CTA was unaffected by bilateral parabrachial TTX injection 12. Moreover,
the passive avoidance research showed that the amnesic effect of electroconvulsive shock (ECS) does not depend only on the acquisition-treatment delay 9'~~ but also on the duration 3"~3 and the intensity4'~6 of the unconditioned stimulus (US). Similar results were obtained with protein synthesis inhibitors ~5. In an earlier experiment PAR has been elicited in the step-through apparatus with three consecutive series of 5 footshocks (i mA, 1 s) serving as the US 19. Given the high number of footshocks, and their fairly high intensity, the partial amnesic effect of post-acquisition TTX in this paradigm could be perhaps attributed to overtraining. The aim of the present work is to study in rats after a well reinforced training the impairment of PAR by post-acquisition injection of TTX into the PBN while using the acquisition-TTX interval and footshock intensity as independent variables.
MATERIALS AND METHODS
Animals Correspondence: C. Bucherelli, Dipartimento di Scienze Fisiologiche, Viale G.B. Morgagni 63, 1-50134 Firenze, Italia.
Sixty-day-old male albino rats of the Wistar strain (average body weight 280 g) were employed. The animals were individually housed in stainless steel cagbs in
176 a room with natural light-dark cycle and constant temperature of 20 + 1 ~ They had free access to food and water throughout the experiment.
Surgely and drug application Functional ablation of the PBN was induced by bilateral injection of 10 ng TTX dissolved in 1 Ill saline into points with the stereotaxic coordinates AP 8.8; L + 1.8; V 7. The accuracy ofthe setting had been controlled in preliminary trials. These coordinates are quite close to thosegiven by Paxinos and Watson 14for Wistar rats with body weight between 270 and 320 g. TTX was injected in anesthetized animals fixed in the stereotaxic apparatus. The injection needle (diameter 0.4 mm) connected to the Hamilton syringe was fixed in the electrode holder of the stereotaxic apparatus and introduced into the target structure. The animals were anesthetized with ketamine (Ketalar, Parke Davis, 100 mg/kg i.p.) and trephine openings 1 mm in diameter were made 8.8 mm caudal from bregma and 1.8 mm lateral f r o m t h e midline. One lal of the solution was injected over 1.5 min and the needle was left in place for another 1 min before it was slowly withdrawn. Sham-operated animals received parabrachial injection of saline under the above condition. Unoperated control animals underwent only conditioning procedures.
Apparatus The step-through apparatus ~ consisted of a light chamber (30 x 21 • 15 cm) made of white opaque plastic with a transparent lid, and of a dark chamber (30 x 21 • 15 cm) with walls and ceiling made of dark opaque plastic. The floor of both chambers was made of 2-mm stainless steel rods spaced 1 cm apart; the floor of the dark chamber could be electrified. The two chambers were connected by an opaque guillotine door (6 • 8 cm). The apparatus was placed in an acoustically isolated room, kept at a constant temperature of 20+ 1 ~ Illumination inside the light chamber was 60 lux.
Procedure PAR acquisition. The subject was placed into the illuminated compartment of the apparatus and 5 s later the guillotine door was raised. The latency of entrance into the dark compartment (step-through latency) was recorded with a stop-watch when the animal had placed all 4 paws beyond the threshold. After the animal had entered the dark compartment (or after it was placed there by hand when it had stayed more than 60 s in the illuminated compartment) the guillotine door was lowered and one 3-s unavoidable scrambled electrie shock was applied. The door was opened again, and 5 s later,
if the rat was still inside the dark compartment, up to 5 avoidable shocks (1 s) were administered, at 2-s intervals. After the animal had spontaneously gone inside the light compartment or after having been placed inside it by hand, the guillotine door remained open and the animal was allowed to stay for up to 60 s in the light compartment. The whole procedure was then repeated; at the end of the second trial the rat was returned to the home cage. PAR retrieval. Two days after PAR acquisition and/ or after the experimental treatment (i.e. after complete disappearance of the TTX effect11"2~ the animal was placed into the illuminated compartment, the guillotine door was raised and rat's step-through latency was recorded up to 180 s.
Experimental groups The rats were randomly divided into 23 groups of 8-10 animals each. Fifteen groups received 0.8 mA footshocks, 8 groups 1.2 mA footshocks. The groups differed also according to (1)the diverse procedures which were employed: PBN TTX injection (T), shamoperated (S), non-operated controls (C) and (2) according to'the interval between PAR acquisition and TTX or saline administration. Thus the 0.8 mA groups were: C groups with acquisition-retrieval delay respectively of 2, 3, 4, 6, 10 days; S and T groups with acquisitiontreatment delay respectively of 0, 1, 2, 4, 8 days. The 1.2 mA groups were: S and T groups with acquisitiontreatment delay respectively of 0, 1, 2, 4 days.
Statistical analysis Statistical analysis was performed employing oneway, two-way and three-way ANOVA and N e w m a n Keuls' multiple comparisons test.
Morphology After conclusion of the experiment, the injection site was histologically verified in a random selection of about one-third of the animals. Rats were deeply anesthetized and intracardially perfused with saline, which was followed by 4% formaldehyde. The brains were cut with a freezing microtome and the injection needle tracks were identified in Nissl-stained serial sections.
RESULTS
Spontaneous preference of the dark compartment Step-through latencies in the first acquisition trial were similar in all 23 groups. The average values ranged from 13.6 + 3.2 (mean + S.E.M.) to 7.3 + 2.9. One-way ANOVA showed absence of significant between groups
177 180 -
g 120-
T
i -r
i= a.
6o.
I
'
sl O
1
2
3
1
2
i
n
4 DAYS
8
6
Ar
- Treat ment Delay Acqulsitlon-Retrieval Delay
10
Fig. 1. glean ( _+S.E.M.) step-through latencies during retrieval test in the 0.8 mA shock groups. C: unoperated groups. S: sham-operated groups. T: TTX-injected groups.
differences (F22,286 = 0.07, n.s.) and indicated uniformity of the employed groups.
PAR hz the weak footshock (0.8 mA) groups After the first footshock in the dark compartment almost all rats remained in the light compartment for 60 s and had to be placed inside the dark compartment by hand. Fig. 1 shows good PAR retrieval in both the S and C groups. This indicates that ketamine anesthesia and surgical procedures do not influence PAR acquisition and retention, irrespective of the time interval after acquisition. On the contrary, functional ablation of the PBN by TTX always induces retrograde amnesia, irrespective of the duration of the acquisitiontreatment interval. The last assertion is documented by poor PAR in all the T groups. All above statements are supported by the statistical analysis. Two-way ANOVA (3 x 5) revealed significant main effect of treatment (F2,122 = 28.02, P < 0.0001) but not of the post-acquisition time (F4,122 = 0.29, n.s.). There was no significant interaction (Fs.12z = 0.30, n.s.). Newman-Keuls' multiple comparisons showed that the T groups were significantly different from the S and C groups at each postacquisition interval (P<0.01) and that the C and S groups were never significantly different from each other.
the PBN by TTX induced amnesia when performed immediately after the acquisition test or 24 h later. The same intervention performed after either 48 or 96 h was ineffective. Step-through latency in the S groups was inversely related to the acquisition-retrieval interval. Two-way ANOVA (2 x 4) showed significant main effects of treatment (Fl,15 7 = 17.09, P < 0.0001) but not of the post-acquisition time (F3,57 = 0.64, n.s.). There was significant interaction (F3,57= 8.31, P<0.0001). Newman-Keuls' multiple comparisons showed significant differences ( P < 0.01) between the S and T groups 180 -
I v Z w I-
120
,,r 0
60 w I-
S O-
PAR hz the strong footshock (1.2 mA) groups Since weak shock intensity revealed no differences between the C and S groups, only the S and T groups were used with the 1.2 mA shock intensity which caused complete avoidance of the dark compartment after the first trial. Fig. 2 shows that the functional ablation of
0 2
i
S
i
:T
I
2
3
4 DAYS
4
Acquisition -Treatment Delay Acquisltlon- Retrieval Delay
Fig. 2. Mean (+ S.E.M.) step-through latencies during retrieval i'~st in the 1.2 mA shock groups. S: sham-operated groups. T: TTXinjected groups.
178 at the 0- and 1-day but not at the 2- and 4-day postacquisition interval. Step-through latency in the S group with an acquisition-retrieval delay of 2 days was significantly longer than in the S groups with 4- and 6-day delays ( P < 0.05, in both instances). Also the T groups with 0- and 1-day acquisition-treatment delays were significantly different from the T groups with the 2- and 4-day delays (P<0.01, in all instances).
Comparison between PAR elicited by 0.8 and 1.2 mA footshocks . In this comparison, only groups tested at both shock intensities, i.e. the S and T groups treated 15 min, I, 2, and 4 days after acquisition, are considered. The comparison of Figs. I and 2 indicates the stronger PAR in the S groups receiving 1.2 mA footshocks, but only at the shortest acquisition-retrieval intervals. The differences between the S groups receiving 1.2 mA and 0.8 mA footshock disappear at longer acquisitionretrieval intervals. On the other hand, there is a clearcut difference between the effects of functional ablation of the PBN on PAR in the 0.8 and 1.2 mA groups. Both the 0.8 and 1.2 mA S groups exhibit the same PAR 96 h after the acquisition, but, at that acquisition-treatment delay, the 0.8 and 1.2 mA T groups exhibit a clearly different PAR, strictly related to the footshock intensity. All these conclusions are supported by statistical analysis. Three-way ANOVA (2 • 2 • 4) showed significant main effects of treatment (F~.tzo=53.98, P<0.0001) and footshock intensity (F1a2o=20.29, P<0.0001) but not of delay (F3,12o= 0.63, n.s.). There were significant treatment-delay (F3,12o = 4.45, P<0.01) and treatment-intensity-delay (F3a2o = 3.96, P < 0.0 I) interactions. Newman-Keuls' multiple comparisons showed that the 0.8 mA and 1.2 S groups displayed significantly different PAR retrieval 2 and 3 days ( P < 0.05, in both instances) but not 4 and 6 days after acquisition. Significant differences in step-through behavior were found between the 0.8 and 1.2 m A T groups receiving TTX 4 days after acquisition (P
DISCUSSION The behavioural homogeneity of the employed rats was shown by similar step-through latencies in the acquisition trial in all groups. After the unavoidable shock almost always rats spontaneously left the dark compartment. This means that rats very seldom received the supplementary escapable shocks. Thus, the groups of rats were fairly homogeneous also with respect to shock exposition. The spontaneous preference of rats
for darkness and its reversal by PAR training demonstrated that the employed procedure was adequate. The present results confirm previous reports according to which post-trial general anesthesia and surgical procedures do not negatively influence the avoidance responses by interfering with some later stage of engram formation 1~'12a9. In the present study ketamine was employed instead of pentobarbital because of the shorter general anesthesia duration and of the less depressant effects on respiratory function. Also preliminary experiment had shown that in Wistar rats TTX bilateral functional blockade of PBN was well tolerated under ketamine anesthesia. It must be pointed out that partial blockade on NMDA-receptor-gated channels, induced by ketamine, does not interfere with PAR memory storage tested 2 days after surgery. The extent and duration of the blockade induced by the employed TTX dosage was described by other authors 2~ They monitored pupillary diameter in anesthetized rats at various intervals after injecting TTX, 2, 1.5, 1, 0 mm lateral from the Edinger-Westphal nucleus. The resulting mydriasis indicated that the mean blockade radius is 1 mm and that the maximum effect is obtained 30-120 min after injection and decays exponentially to control level over the subsequent 24 h. Our histological controls showed a good repetitivity of injection in the PBN site. Given the accuracy of the administration site TTX surely affected the parabrachial region. Because of the constancy of the TTX diffusion we feel that the possibility of TTX having induced a significant functional inactivation also of the adjacent reticular formation was small enough. As in previous experiments ~9 the initial motor impairment caused by the bilateral TTX injection into the PBN disappeared completely within 24 h, as shown by the neurological tests which were back to normal. Although impaired retrieval certainly accompanies early stages of post-TTX recovery, 2 days are obviously a sufficiently long interval to reinstate normal readout of all available memory traces. Our results indicate that the retrograde amnesia elicited by TTX blockade of the PBN depends on the intensity of the footshocks employed during PAR acquisition. When the animals were subjected to mild punishments the induction of retrograde amnesia by the functional ablation of the PBN was not related to the delay of the TI'X blockade. On the contrary, a definite temporal relationship emerged when more intense nociceptive stimuli were employed. This finding confirms the results of a previous experiment in which rats had been subjected to a higher number of footshocks of an intensity (1 mA) intermediate between the ones employed in the present experiment ~9.
179 Obviously the avoidance response was less evident in the groups of animals that had received the weaker footshocks. This was shown not only by the unequal PAR exhibited by 0.8 and 1.2 mA groups, but also by the different effects of the TTX blockade on the PAR. We can surmise that the stage of memory formation may be significantly shortened by the increase of the number and intensity of the aversive stimuli. On the contrary, when milder footshocks are used, engram consolidation needs more time. This can explain the disruptive effect of 3"I'X blockade lasting up to 8 days after acquisition when using the 0.8 mA US. This extremely delayed anmesic effect resembles the anticholinesterase-induced amnesia described by other authors 5. Our results fit well with the analysis of retrograde-amnesia data forming the basis of the consolidation hypothesis: consolidation time is inversely related to US intensity2. This can be mainly due to different initial engram strengths. According to other authors 6'7, who found that at longer post-acquisition intervals amnesi~i can be elicited by the same amount of puromycin injected into several sites of the brain, stronger US may activate multiple storage sites. In this case spatially circumscribed functional ablation of a definite area does not cause engram disruption. It is possible that 48 h after acquisition (1.2 mA footshocks), PAR engram could be present also in other brain sites distant f r o m t h e PBN. The comparison of the results of the 0.8 and 1.2 mA groups suggests that engram strength may be expressed not only by the intensity of PAR measured during retrieval testing, but also by the resistance of the engram to TTX disruption. In fact, the effects of functional ablation of the PBN on PAR were quite different in the 0.8 and 1.2 mA groups when induced 96 h after acquisition at a time when both the respective S groups exhibited the same PAR. These findings indicate that the response intensity is not a reliable index of resistance of the acquired response to the disrupting influences. Absence of consolidation gradient in the low shock groups could indicate a TTX effect on retrieval. However, this explanation is improbable because good retrieval was seen in the high-shock groups with the same TTX-retrieval time interval. Previous studies seemed to indicate that functional ablation of the PBN by TTX elicits stronger and longer lasting retrograde anmesia for CTA than for PAR, although both reactions are inhibitory I m2.~9. The present results show that the TTX blockade of the PBN may differentially affect responses of the same type, acquired with different US intensity. These results do not support the hypothesis that the PBN is more important for CTA than for PAR consolidation ~9.
The present study confirms the hypothesis that consolidation of passive avoidance responses depends on the intensity of the punishment employed during acquisition. This was previously demonstrated by employing ECS as the amnesic agent 3"4"8A6"17. We may conclude that the PBN play a crucial role in PAR engram consolidation, but that the time period during which this area appears to be crucial is variable and depends on the conditioning procedure employed.
ACKNOWLEDGEMENTS The authors wish to express their warmest thanks to Dr. J. Bure~ for his valuable advice and constructive criticism during the course of this work. The authors also wish to thank Prof. Carlo Ambrogi Lorenzini for his critical suggestions. They also wish to thank Mr. A. Aiazzi, Mr. S. Batacchi, Mr. S. Cammarata, Mr. M. Dolfi and Mr. A. Vannucchi for their technical assistance.
REFERENCES 1 AmbrogiLorenzini,C., Bucherelli,C. and Giachetti,A., Some factors influencingconditionedand spontaneousbehaviorof rats in the light-'dark box test, PhysioL Behav., 36 (1986) 97-101. 2 Cherkin,A., Memoryconsolidation:probit analysisofretrogradeamnesia data, Psychonom. Sci., 4 (1966) 169-170. 3 Chorover,S.L. and Schiller, P.H., Short-termretrogradeamnesia in rats, J. Comp. PhysioL PsychoL, 59 (1965) 73-78. 4 Del Prato, D.J. and Thompson,R.W., Effectofelectroconvulsive shock on passive avoidancelearningwith highand low intensity foot-shock, Psychol. Rep., 17 (1965) 209-210. 5 Deutsch,J.A., Hambnrg,M.D. and Dahl, H., Anticholinesteraseinduced amnesia and its temporal aspects, Science, 151 (1966) 221-223. 6 Flexner, L.B., Church, A.C., Flexner, J.B. and Rainbow,T.C., The effectof the beta-adrenergieantagonist,propranolol, on the cerebral spread of a memorytrace in mice,PharmacoL Biochem. Behav., 21 (1984) 633-639. 7 Flexner,L.B., Flexner,J.B., De La Haba, G. and Roberts, R.B., Loss of memoryas related to the inhibitionof protein synthesis. J. IVeurochem., 12 (1965) 535-541. 8 Geller,A., Robustelli, F. and Jarvick, M.E., A parallel studyof the amnesic effect of cycloheximideand ECS under different strengths of conditioning,Psychophannacology, 16 (1970) 281289. 9 Huges,R.A., Barrett, R.J. and Ray,O.S., Retrogradefimnesiain rats increasesas a functionof ECS-test intervaland ECS intensity, PhysioL Behav., 5 (1970) 27-30. I0 Huges,R.A., Barrett, R.J. and Ray, O.S., Training-to-testinterval as a determinant of a temporallygraded ECS-produced response decrement in rats, J. Comp. PhysioL Psychol., 71 (1970) 318-324. '~ 11 Ivanova, S.F. and Buret, J., Acquisition of conditioned taste aversion in rats is preventedby tetrodotoxinblockadeof a-small
180
12
13 14 15
16
midbrain region centered around the parabrachial nuclei, Physiol. Behav., 48 (1990) 543-549. Ivanova, S.F. and Buret, J., Conditioned taste aversion is disrupted by prolonged retrograde effects of intracerebral injection of tetrodotoxin in rats, Behav. Neurosci., 104 (1990) 948-954. McGaugh, J.L., Time-dependent processes in memory storage, Science, 153 (1966) 1351-1358. Paxinos, G. and Watson, C., The Rat Brain h2 Stereotaxic Coordinates, 2nd edn., Academic Press, Sydney, 1986. Quartermain, D. and McEwan, B.S., Temporal characteristics of amnesia induced by protein s)aathesis inhibitor: determination by shock level, Nature, 288 (1970) 677-678. Ray, O.S.,.and Bivens, L.W., Reinforcement magnitude as a determinant of performance decrement after electroconvulsive shock, Science, 160 (1968) 330-332.
17 Robustelli, F., Geller, A., Aron, C. and Jarvick, M.E., The relationship between the amnesic effect of electroconvulsive shock and strength of conditioning in a passive avoidance task, Comm. Behav. BioL, 3 (1969) 233-239. 18 Spevack, A.A., Rabedau, R.G. and Spevack, M.E., A temporal graded ECS function following one trial learning, Psychonont. ScL, 9 (1967) 153-154. 19 Tassoni, G., Bucherelli, C. and Burel, J., Post-acquisition injection of tetrodotoxin into the parabrachial nuclei elicits partial disruption of passive avoidance reaction in rats, Behav. Neural BioL, 57 (1992) 116-123. 20 Zhuravin, I.A. and Buret, J., Extent of the tetrodotoxin induced blockade examined by pupillary paralysis elicited by intracerebral injection of the drug, Exp. Brahl Res., 83 (1991) 687-690.
175
BBR 01302
Duration of retrograde amnesia induced by tetrodotoxin inactivation of the parabrachial nuclei is inversely related to the intensity of footshock in rat's passive avoidance response Corrado Bucherelli and Giovanna Tassoni Diparthnento di Scienze Fisiologiche, Firenze (Italy) (Received 26 September 1991) (Revised version received 20 December 1992) (Accepted 15 April 1992)
Key words: Memory consolidation; Parabrachial nucleus; Footshock intensity; Tetrodotoxin; Functional ablation
Previous work has'demonstrated that in rats post-trial bilateral functional blockade of the parabrachial nuclei by local injection of tetrodotoxin (10 ng in 1 lal) partially disrupts retention of an overtrained passive avoidance response. The time-course of functional blockade disrupting effects on passive avoidance response has been studied by altering both the acquisition-tetrodotoxin injection interval and footshock intensity as independent variables. When weak footshocks (0.8 mA) were delivered, functional ablation of the parabrachial nuclei was still effective when induced up to 8 days after acquisition training. On the other hand, when rats were shocked with the stronger stimuli (1.2 mA), 2 days after acquisition functional ablation was no longer effective. The results are discussed in terms of unconditioned stimulus (US) intensity and engram consolidation.
INTRODUCTION
Previous work has shown that bilateral reversible blockade of the parabrachial nuclei (PBN) by tetrodotoxin (TTX) injected shortly after the acquisition session, disrupts inhibitory conditioned responses ~'~z'19. TTX has a well-known pharmacological action, namely the specific blocking of sodium channels. This type of PBN blockade has a significant amnesic effect on conditioned taste aversion (CTA) when induced up to 4 days after the taste-poison pairing ~2. Similarly the passive avoidance response (PAR) was disrupted with the difference that the amnesic effect was weaker and could be elicited by TTX applied 24 h but not 48 h after acquisition TM. These quantitative differences between CTA and PAR disruption were tentatively explained on the assumption that the PBN and contiguous reticular formation are directly involved in the representation of CTA, but that they participate less directly in PAR 19. On the other hand, overtrained CTA was unaffected by bilateral parabrachial TTX injection 12. Moreover,
the passive avoidance research showed that the amnesic effect of electroconvulsive shock (ECS) does not depend only on the acquisition-treatment delay 9'~~ but also on the duration 3"~3 and the intensity4'~6 of the unconditioned stimulus (US). Similar results were obtained with protein synthesis inhibitors ~5. In an earlier experiment PAR has been elicited in the step-through apparatus with three consecutive series of 5 footshocks (i mA, 1 s) serving as the US 19. Given the high number of footshocks, and their fairly high intensity, the partial amnesic effect of post-acquisition TTX in this paradigm could be perhaps attributed to overtraining. The aim of the present work is to study in rats after a well reinforced training the impairment of PAR by post-acquisition injection of TTX into the PBN while using the acquisition-TTX interval and footshock intensity as independent variables.
MATERIALS AND METHODS
Animals Correspondence: C. Bucherelli, Dipartimento di Scienze Fisiologiche, Viale G.B. Morgagni 63, 1-50134 Firenze, Italia.
Sixty-day-old male albino rats of the Wistar strain (average body weight 280 g) were employed. The animals were individually housed in stainless steel cagbs in
176 a room with natural light-dark cycle and constant temperature of 20 + 1 ~ They had free access to food and water throughout the experiment.
Surgely and drug application Functional ablation of the PBN was induced by bilateral injection of 10 ng TTX dissolved in 1 Ill saline into points with the stereotaxic coordinates AP 8.8; L + 1.8; V 7. The accuracy ofthe setting had been controlled in preliminary trials. These coordinates are quite close to thosegiven by Paxinos and Watson 14for Wistar rats with body weight between 270 and 320 g. TTX was injected in anesthetized animals fixed in the stereotaxic apparatus. The injection needle (diameter 0.4 mm) connected to the Hamilton syringe was fixed in the electrode holder of the stereotaxic apparatus and introduced into the target structure. The animals were anesthetized with ketamine (Ketalar, Parke Davis, 100 mg/kg i.p.) and trephine openings 1 mm in diameter were made 8.8 mm caudal from bregma and 1.8 mm lateral f r o m t h e midline. One lal of the solution was injected over 1.5 min and the needle was left in place for another 1 min before it was slowly withdrawn. Sham-operated animals received parabrachial injection of saline under the above condition. Unoperated control animals underwent only conditioning procedures.
Apparatus The step-through apparatus ~ consisted of a light chamber (30 x 21 • 15 cm) made of white opaque plastic with a transparent lid, and of a dark chamber (30 x 21 • 15 cm) with walls and ceiling made of dark opaque plastic. The floor of both chambers was made of 2-mm stainless steel rods spaced 1 cm apart; the floor of the dark chamber could be electrified. The two chambers were connected by an opaque guillotine door (6 • 8 cm). The apparatus was placed in an acoustically isolated room, kept at a constant temperature of 20+ 1 ~ Illumination inside the light chamber was 60 lux.
Procedure PAR acquisition. The subject was placed into the illuminated compartment of the apparatus and 5 s later the guillotine door was raised. The latency of entrance into the dark compartment (step-through latency) was recorded with a stop-watch when the animal had placed all 4 paws beyond the threshold. After the animal had entered the dark compartment (or after it was placed there by hand when it had stayed more than 60 s in the illuminated compartment) the guillotine door was lowered and one 3-s unavoidable scrambled electrie shock was applied. The door was opened again, and 5 s later,
if the rat was still inside the dark compartment, up to 5 avoidable shocks (1 s) were administered, at 2-s intervals. After the animal had spontaneously gone inside the light compartment or after having been placed inside it by hand, the guillotine door remained open and the animal was allowed to stay for up to 60 s in the light compartment. The whole procedure was then repeated; at the end of the second trial the rat was returned to the home cage. PAR retrieval. Two days after PAR acquisition and/ or after the experimental treatment (i.e. after complete disappearance of the TTX effect11"2~ the animal was placed into the illuminated compartment, the guillotine door was raised and rat's step-through latency was recorded up to 180 s.
Experimental groups The rats were randomly divided into 23 groups of 8-10 animals each. Fifteen groups received 0.8 mA footshocks, 8 groups 1.2 mA footshocks. The groups differed also according to (1)the diverse procedures which were employed: PBN TTX injection (T), shamoperated (S), non-operated controls (C) and (2) according to'the interval between PAR acquisition and TTX or saline administration. Thus the 0.8 mA groups were: C groups with acquisition-retrieval delay respectively of 2, 3, 4, 6, 10 days; S and T groups with acquisitiontreatment delay respectively of 0, 1, 2, 4, 8 days. The 1.2 mA groups were: S and T groups with acquisitiontreatment delay respectively of 0, 1, 2, 4 days.
Statistical analysis Statistical analysis was performed employing oneway, two-way and three-way ANOVA and N e w m a n Keuls' multiple comparisons test.
Morphology After conclusion of the experiment, the injection site was histologically verified in a random selection of about one-third of the animals. Rats were deeply anesthetized and intracardially perfused with saline, which was followed by 4% formaldehyde. The brains were cut with a freezing microtome and the injection needle tracks were identified in Nissl-stained serial sections.
RESULTS
Spontaneous preference of the dark compartment Step-through latencies in the first acquisition trial were similar in all 23 groups. The average values ranged from 13.6 + 3.2 (mean + S.E.M.) to 7.3 + 2.9. One-way ANOVA showed absence of significant between groups
177 180 -
g 120-
T
i -r
i= a.
6o.
I
'
sl O
1
2
3
1
2
i
n
4 DAYS
8
6
Ar
- Treat ment Delay Acqulsitlon-Retrieval Delay
10
Fig. 1. glean ( _+S.E.M.) step-through latencies during retrieval test in the 0.8 mA shock groups. C: unoperated groups. S: sham-operated groups. T: TTX-injected groups.
differences (F22,286 = 0.07, n.s.) and indicated uniformity of the employed groups.
PAR hz the weak footshock (0.8 mA) groups After the first footshock in the dark compartment almost all rats remained in the light compartment for 60 s and had to be placed inside the dark compartment by hand. Fig. 1 shows good PAR retrieval in both the S and C groups. This indicates that ketamine anesthesia and surgical procedures do not influence PAR acquisition and retention, irrespective of the time interval after acquisition. On the contrary, functional ablation of the PBN by TTX always induces retrograde amnesia, irrespective of the duration of the acquisitiontreatment interval. The last assertion is documented by poor PAR in all the T groups. All above statements are supported by the statistical analysis. Two-way ANOVA (3 x 5) revealed significant main effect of treatment (F2,122 = 28.02, P < 0.0001) but not of the post-acquisition time (F4,122 = 0.29, n.s.). There was no significant interaction (Fs.12z = 0.30, n.s.). Newman-Keuls' multiple comparisons showed that the T groups were significantly different from the S and C groups at each postacquisition interval (P<0.01) and that the C and S groups were never significantly different from each other.
the PBN by TTX induced amnesia when performed immediately after the acquisition test or 24 h later. The same intervention performed after either 48 or 96 h was ineffective. Step-through latency in the S groups was inversely related to the acquisition-retrieval interval. Two-way ANOVA (2 x 4) showed significant main effects of treatment (Fl,15 7 = 17.09, P < 0.0001) but not of the post-acquisition time (F3,57 = 0.64, n.s.). There was significant interaction (F3,57= 8.31, P<0.0001). Newman-Keuls' multiple comparisons showed significant differences ( P < 0.01) between the S and T groups 180 -
I v Z w I-
120
,,r 0
60 w I-
S O-
PAR hz the strong footshock (1.2 mA) groups Since weak shock intensity revealed no differences between the C and S groups, only the S and T groups were used with the 1.2 mA shock intensity which caused complete avoidance of the dark compartment after the first trial. Fig. 2 shows that the functional ablation of
0 2
i
S
i
:T
I
2
3
4 DAYS
4
Acquisition -Treatment Delay Acquisltlon- Retrieval Delay
Fig. 2. Mean (+ S.E.M.) step-through latencies during retrieval i'~st in the 1.2 mA shock groups. S: sham-operated groups. T: TTXinjected groups.
178 at the 0- and 1-day but not at the 2- and 4-day postacquisition interval. Step-through latency in the S group with an acquisition-retrieval delay of 2 days was significantly longer than in the S groups with 4- and 6-day delays ( P < 0.05, in both instances). Also the T groups with 0- and 1-day acquisition-treatment delays were significantly different from the T groups with the 2- and 4-day delays (P<0.01, in all instances).
Comparison between PAR elicited by 0.8 and 1.2 mA footshocks . In this comparison, only groups tested at both shock intensities, i.e. the S and T groups treated 15 min, I, 2, and 4 days after acquisition, are considered. The comparison of Figs. I and 2 indicates the stronger PAR in the S groups receiving 1.2 mA footshocks, but only at the shortest acquisition-retrieval intervals. The differences between the S groups receiving 1.2 mA and 0.8 mA footshock disappear at longer acquisitionretrieval intervals. On the other hand, there is a clearcut difference between the effects of functional ablation of the PBN on PAR in the 0.8 and 1.2 mA groups. Both the 0.8 and 1.2 mA S groups exhibit the same PAR 96 h after the acquisition, but, at that acquisition-treatment delay, the 0.8 and 1.2 mA T groups exhibit a clearly different PAR, strictly related to the footshock intensity. All these conclusions are supported by statistical analysis. Three-way ANOVA (2 • 2 • 4) showed significant main effects of treatment (F~.tzo=53.98, P<0.0001) and footshock intensity (F1a2o=20.29, P<0.0001) but not of delay (F3,12o= 0.63, n.s.). There were significant treatment-delay (F3,12o = 4.45, P<0.01) and treatment-intensity-delay (F3a2o = 3.96, P < 0.0 I) interactions. Newman-Keuls' multiple comparisons showed that the 0.8 mA and 1.2 S groups displayed significantly different PAR retrieval 2 and 3 days ( P < 0.05, in both instances) but not 4 and 6 days after acquisition. Significant differences in step-through behavior were found between the 0.8 and 1.2 m A T groups receiving TTX 4 days after acquisition (P
DISCUSSION The behavioural homogeneity of the employed rats was shown by similar step-through latencies in the acquisition trial in all groups. After the unavoidable shock almost always rats spontaneously left the dark compartment. This means that rats very seldom received the supplementary escapable shocks. Thus, the groups of rats were fairly homogeneous also with respect to shock exposition. The spontaneous preference of rats
for darkness and its reversal by PAR training demonstrated that the employed procedure was adequate. The present results confirm previous reports according to which post-trial general anesthesia and surgical procedures do not negatively influence the avoidance responses by interfering with some later stage of engram formation 1~'12a9. In the present study ketamine was employed instead of pentobarbital because of the shorter general anesthesia duration and of the less depressant effects on respiratory function. Also preliminary experiment had shown that in Wistar rats TTX bilateral functional blockade of PBN was well tolerated under ketamine anesthesia. It must be pointed out that partial blockade on NMDA-receptor-gated channels, induced by ketamine, does not interfere with PAR memory storage tested 2 days after surgery. The extent and duration of the blockade induced by the employed TTX dosage was described by other authors 2~ They monitored pupillary diameter in anesthetized rats at various intervals after injecting TTX, 2, 1.5, 1, 0 mm lateral from the Edinger-Westphal nucleus. The resulting mydriasis indicated that the mean blockade radius is 1 mm and that the maximum effect is obtained 30-120 min after injection and decays exponentially to control level over the subsequent 24 h. Our histological controls showed a good repetitivity of injection in the PBN site. Given the accuracy of the administration site TTX surely affected the parabrachial region. Because of the constancy of the TTX diffusion we feel that the possibility of TTX having induced a significant functional inactivation also of the adjacent reticular formation was small enough. As in previous experiments ~9 the initial motor impairment caused by the bilateral TTX injection into the PBN disappeared completely within 24 h, as shown by the neurological tests which were back to normal. Although impaired retrieval certainly accompanies early stages of post-TTX recovery, 2 days are obviously a sufficiently long interval to reinstate normal readout of all available memory traces. Our results indicate that the retrograde amnesia elicited by TTX blockade of the PBN depends on the intensity of the footshocks employed during PAR acquisition. When the animals were subjected to mild punishments the induction of retrograde amnesia by the functional ablation of the PBN was not related to the delay of the TI'X blockade. On the contrary, a definite temporal relationship emerged when more intense nociceptive stimuli were employed. This finding confirms the results of a previous experiment in which rats had been subjected to a higher number of footshocks of an intensity (1 mA) intermediate between the ones employed in the present experiment ~9.
179 Obviously the avoidance response was less evident in the groups of animals that had received the weaker footshocks. This was shown not only by the unequal PAR exhibited by 0.8 and 1.2 mA groups, but also by the different effects of the TTX blockade on the PAR. We can surmise that the stage of memory formation may be significantly shortened by the increase of the number and intensity of the aversive stimuli. On the contrary, when milder footshocks are used, engram consolidation needs more time. This can explain the disruptive effect of 3"I'X blockade lasting up to 8 days after acquisition when using the 0.8 mA US. This extremely delayed anmesic effect resembles the anticholinesterase-induced amnesia described by other authors 5. Our results fit well with the analysis of retrograde-amnesia data forming the basis of the consolidation hypothesis: consolidation time is inversely related to US intensity2. This can be mainly due to different initial engram strengths. According to other authors 6'7, who found that at longer post-acquisition intervals amnesi~i can be elicited by the same amount of puromycin injected into several sites of the brain, stronger US may activate multiple storage sites. In this case spatially circumscribed functional ablation of a definite area does not cause engram disruption. It is possible that 48 h after acquisition (1.2 mA footshocks), PAR engram could be present also in other brain sites distant f r o m t h e PBN. The comparison of the results of the 0.8 and 1.2 mA groups suggests that engram strength may be expressed not only by the intensity of PAR measured during retrieval testing, but also by the resistance of the engram to TTX disruption. In fact, the effects of functional ablation of the PBN on PAR were quite different in the 0.8 and 1.2 mA groups when induced 96 h after acquisition at a time when both the respective S groups exhibited the same PAR. These findings indicate that the response intensity is not a reliable index of resistance of the acquired response to the disrupting influences. Absence of consolidation gradient in the low shock groups could indicate a TTX effect on retrieval. However, this explanation is improbable because good retrieval was seen in the high-shock groups with the same TTX-retrieval time interval. Previous studies seemed to indicate that functional ablation of the PBN by TTX elicits stronger and longer lasting retrograde anmesia for CTA than for PAR, although both reactions are inhibitory I m2.~9. The present results show that the TTX blockade of the PBN may differentially affect responses of the same type, acquired with different US intensity. These results do not support the hypothesis that the PBN is more important for CTA than for PAR consolidation ~9.
The present study confirms the hypothesis that consolidation of passive avoidance responses depends on the intensity of the punishment employed during acquisition. This was previously demonstrated by employing ECS as the amnesic agent 3"4"8A6"17. We may conclude that the PBN play a crucial role in PAR engram consolidation, but that the time period during which this area appears to be crucial is variable and depends on the conditioning procedure employed.
ACKNOWLEDGEMENTS The authors wish to express their warmest thanks to Dr. J. Bure~ for his valuable advice and constructive criticism during the course of this work. The authors also wish to thank Prof. Carlo Ambrogi Lorenzini for his critical suggestions. They also wish to thank Mr. A. Aiazzi, Mr. S. Batacchi, Mr. S. Cammarata, Mr. M. Dolfi and Mr. A. Vannucchi for their technical assistance.
REFERENCES 1 AmbrogiLorenzini,C., Bucherelli,C. and Giachetti,A., Some factors influencingconditionedand spontaneousbehaviorof rats in the light-'dark box test, PhysioL Behav., 36 (1986) 97-101. 2 Cherkin,A., Memoryconsolidation:probit analysisofretrogradeamnesia data, Psychonom. Sci., 4 (1966) 169-170. 3 Chorover,S.L. and Schiller, P.H., Short-termretrogradeamnesia in rats, J. Comp. PhysioL PsychoL, 59 (1965) 73-78. 4 Del Prato, D.J. and Thompson,R.W., Effectofelectroconvulsive shock on passive avoidancelearningwith highand low intensity foot-shock, Psychol. Rep., 17 (1965) 209-210. 5 Deutsch,J.A., Hambnrg,M.D. and Dahl, H., Anticholinesteraseinduced amnesia and its temporal aspects, Science, 151 (1966) 221-223. 6 Flexner, L.B., Church, A.C., Flexner, J.B. and Rainbow,T.C., The effectof the beta-adrenergieantagonist,propranolol, on the cerebral spread of a memorytrace in mice,PharmacoL Biochem. Behav., 21 (1984) 633-639. 7 Flexner,L.B., Flexner,J.B., De La Haba, G. and Roberts, R.B., Loss of memoryas related to the inhibitionof protein synthesis. J. IVeurochem., 12 (1965) 535-541. 8 Geller,A., Robustelli, F. and Jarvick, M.E., A parallel studyof the amnesic effect of cycloheximideand ECS under different strengths of conditioning,Psychophannacology, 16 (1970) 281289. 9 Huges,R.A., Barrett, R.J. and Ray,O.S., Retrogradefimnesiain rats increasesas a functionof ECS-test intervaland ECS intensity, PhysioL Behav., 5 (1970) 27-30. I0 Huges,R.A., Barrett, R.J. and Ray, O.S., Training-to-testinterval as a determinant of a temporallygraded ECS-produced response decrement in rats, J. Comp. PhysioL Psychol., 71 (1970) 318-324. '~ 11 Ivanova, S.F. and Buret, J., Acquisition of conditioned taste aversion in rats is preventedby tetrodotoxinblockadeof a-small
180
12
13 14 15
16
midbrain region centered around the parabrachial nuclei, Physiol. Behav., 48 (1990) 543-549. Ivanova, S.F. and Buret, J., Conditioned taste aversion is disrupted by prolonged retrograde effects of intracerebral injection of tetrodotoxin in rats, Behav. Neurosci., 104 (1990) 948-954. McGaugh, J.L., Time-dependent processes in memory storage, Science, 153 (1966) 1351-1358. Paxinos, G. and Watson, C., The Rat Brain h2 Stereotaxic Coordinates, 2nd edn., Academic Press, Sydney, 1986. Quartermain, D. and McEwan, B.S., Temporal characteristics of amnesia induced by protein s)aathesis inhibitor: determination by shock level, Nature, 288 (1970) 677-678. Ray, O.S.,.and Bivens, L.W., Reinforcement magnitude as a determinant of performance decrement after electroconvulsive shock, Science, 160 (1968) 330-332.
17 Robustelli, F., Geller, A., Aron, C. and Jarvick, M.E., The relationship between the amnesic effect of electroconvulsive shock and strength of conditioning in a passive avoidance task, Comm. Behav. BioL, 3 (1969) 233-239. 18 Spevack, A.A., Rabedau, R.G. and Spevack, M.E., A temporal graded ECS function following one trial learning, Psychonont. ScL, 9 (1967) 153-154. 19 Tassoni, G., Bucherelli, C. and Burel, J., Post-acquisition injection of tetrodotoxin into the parabrachial nuclei elicits partial disruption of passive avoidance reaction in rats, Behav. Neural BioL, 57 (1992) 116-123. 20 Zhuravin, I.A. and Buret, J., Extent of the tetrodotoxin induced blockade examined by pupillary paralysis elicited by intracerebral injection of the drug, Exp. Brahl Res., 83 (1991) 687-690.