Evidence for N-methyl-d -aspartate receptor mediation of cocaine induced corticosterone release and cocaine conditioned stimulant effects

ELSEVIER

Behavioural Brain Research 68 (1995) 219-228

BEHAVIOURAL BRAIN RESEARCH

Research report

Evidence for N-methyl-D-aspartate receptor mediation of cocaine induced corticosterone release and cocaine conditioned stimulant effects Ernest N. Damianopoulos 1, Robert J. Carey* SUNY Health Science Center and VA Medical Center, Research Service 151, VA Medical Center, 800 Irving Avenue, Syracuse, NY 13210, USA

Received 8 July 1994; revised 17 November 1994; accepted 18 November 1994

Abstract

The role of the N-methyl-D-aspartate (NMDA) receptors in cocaine conditioning and sensitization of locomotor activity was studied in four groups of Sprague-Dawley rats. A sub-motoric dose of the NMDA antagonist MK-801 (0.1 mg/kg, i.p.) was employed using a novel dual-compartment Pavlovian drug conditioning paradigm. The animals were placed sequentially in two different test environments in which locomotor activity was monitored. In the first compartment, the animals always received a non-drug test for 20 min. Upon completion of this test, the animals received either saline, cocaine (10 mg/kg i.p.), MK-801 or MK-801 plus cocaine depending on group assignment and were then placed immediately into the second compartment and again tested for 20 min. A total of six non-drug and six drug tests were conducted every other day over a 12-day period. Across all drug/saline treatment and post-treatment tests for conditioning, there were no statistical differences in locomotor activity among the saline and drug treatment groups in the non-drug test environment. In the drug/saline associated environment, however, cocaine had a reliable stimulant effect on locomotion when administered alone or in combination with MK-801. Following a 1-day and again after 21-days of withdrawal, all animals were administered a non-drug test for conditioning in which no injections were administered. On both tests, all groups had equivalent activity levels in the non-drug environment. In the drug/saline environment, only the cocaine group of the three drug treatment groups exhibited conditioned hyperlocomotion. Importantly, MK-801 blocked conditioned hyperlocomotion in the combined cocaine + MK-801 group. MK-801 did not alter serum or brain cocaine concentration or the cocaine effects on dopamine metabolism in limbic brain tissue. The co-administration of MK-801 with cocaine, however, blocked the corticosterone release effect of cocaine. Thus, the NMDA receptor site appears critical for cocaine induced conditioning and for corticosterone release. Key words: Cocaine; Pavlovian conditioning; NMDA; MK-801; Locomotion; Corticosterone

1. Introduction

An extensive animal behavior literature indicates that the administration of psychostimulant dopaminergic drugs, such as amphetamine, apomorphine and cocaine, induces locomotor stimulation which, with repeated treatment, can become exaggerated and expressed as a behavioral sensitization response [6,11,14,15,18,19,28,29,30]. Numerous reports suggest a pharmacological attenuation and/or a blockade of behavioral sensitization by antagonism of the N-methyl-D-aspartate ( N M D A ) receptor sites that mediate excitatory amino acid (EAA) neurotransmission [9-12,20,25,26]. O f the various EAA receptor subtypes (e.g., kainate, AMPA), it is the N M D A receptor site that appears to be linked to long term changes in neurons * Corresponding author. Fax: (1)(315)476-5348. To be addressed for reprint requests. 0166-4328/95/$9.50 © 1995 Elsevier Science B.V. All rights reserved (~(~rlrvilgb,

A~'~gOA~ON17~_R

as the activation of the N M D A receptors leads to a calcium entry into the postsynaptic neuron and in sequence to a cascade of biochemical events including G-protein activity and c-fos transcription [1,17,31,32]. These N M D A initiated postsynaptic biochemical events are thought to underlie CNS plasticity at the cellular level as expressed in long term potentiation effects (LTP) [16,33] and may also be implicated in the development of behavioral plasticity expressed as behavioral sensitization and conditioned drug effects. A conventional interpretation of context-specific behavioral sensitization resulting from repeated psychostimulant drug treatment is that the development of behavioral sensitization is a result of an algebraic summation of the drug effect with the conditioned drug effect [22,24]. Given the potential role of the N M D A receptor in conditioning processes, it would appear likely that the N M D A antagonist blocking of behavioral drug sensitization is by virtue

220

E.N. Damianopoulos, R.,I. Carey / Behavioural Brain Research 68 (1995) 219-228

of its efficacy in preventing the development of Pavlovian conditioning of drug responses through a possible interference with exteroceptive stimulus processing [4]. To date, however, there is no direct evidence for the summation hypothesis. Indeed, before one can investigate Pavlovian conditioning of drug responses, especially complex locomotor effects induced by psychostimulant drugs, one must resolve a critical problem that is endemic to such studies [5]. The problem is that the behavioral responses induced or modulated by psychostimulant drugs (e.g., locomotion, rearing, sniffing, etc.) all have a spontaneous above zero baseline. As a consequence, several behavioral process mechanisms, in addition to the drug treatment, can impact substantially on such baselines; e.g., level of arousal and the exploratory/habituated locomotor behavior in response to the novelty/familiarity dimension of the test environment. For the study of psychostimulant drug conditioning, the degree of habituation that develops to the test environment through repeated testing can impact substantially upon the magnitude of the target locomotor behavior. In this regard, there are two aspects of the drug treatment effect on the habituation process which need to be differentiated. In the first case, the drug can simply block habituation. In this circumstance, a direct drug antihabituation effect could be detected by a blockade of within-session response decline in the acquisition phase. As a consequence of this anti-habituation effect, differential response levels would be observed in post-treatment nondrug tests and mislabeled as conditioned drug effects. In the second case, the drug treatment could permit habituation to develop but then the habituation could be nontransferrable to the non-drug state. This would be a state dependent habituation effect. Again, as a consequence, differential response levels would be observed in posttreatment non-drug tests and mislabeled as conditioned drug effects. These considerations are of a substantial importance in any study of conditioned psychostimulant drug effects expressed as changes in locomotor activation detectable by post-treatment tests for conditioning at the end of repeated exposure to test environment stimulation. Inasmuch as control groups in such drug conditioning paradigms become habituated to the test environment due to repeated non-drug exposure, the drug treated group could exhibit an enhanced locomotor response in post-treatment tests for conditioning either as a result of conditioning or as a result of a drug impaired or drug state dependent habituation. Recently, we developed a dual compartmentdual withdrawal interval conditioning test paradigm to resolve this problem which is endemic to drug conditioning experiments. In the new drug conditioning protocol, the animal is first placed into a non-drug compartment. After a fixed time period, the animal is then removed and drug/saline/vehicle

treatments are administered prior to placement into a second compartment of equal area but with a different set of stimulus cues. Locomotion behavior is measured for the same duration in both environments. The first environment tracks the behavioral baseline while the second environment quantitates the drug response. Added to this treatment protocol are two post acquisition tests for conditioning at 1-day and 21-day withdrawal intervals. In these non-drug tests for conditioning, the non-drug environment identifies non-specific behavioral effects whereas the drug environment identifies effects selective to the drug associated environment in order to demonstrate stimulus specificity of any observed drug treatment effects. The results of the two tests for conditioning, separated by 21 days, in addition, distinguish antihabituation and drug state dependent habituation effects from those of conditioning. For example a day 1 positive test result might be due to a conditioned drug effect or due to a drug induced antihabituation effect or due to a drug state dependent habituation effect whenever statistically higher levels of the target locomotor response are observed in the drug vs. saline/unpaired treatment group comparisons. Besides these possible false positive results, a false negative result may also occur on the day 1 test for conditioning. In this case, no differences between drug vs. saline/unpaired treatment groups may occur as a result of drug withdrawal and such an withdrawal induced locomotor suppression effect may mask a conditioned drug effect. However, both withdrawal and habituation effects are labile and diminish as the withdrawal-non-testing interval increases whereas conditioned drug response effects remain stable [5]. Thus, the drug-behavioral process interaction effects which can lead to false positive and false negative results can be expected to drop out on the day 21 test for conditioning. As a result, an observation of a higher level of behavioral responding in the paired treatment group on day 1 in the drug compartment which is not present on day 21 would represent either an antihabituation effect or a drug state dependent habituation effect. Critically, for an inference of a conditioned drug response, the behavioral stimulant effect must be selectively observed in the drug associated environment on the second non-drug test on day 21. Using this newly established protocol, the effects of the N M D A antagonist MK-801 were studied on the development of conditioning and/or sensitization of cocaine induced hyperlocomotion. 2. Materials and methods 2.1. Subjects

Naive adult male Sprague-Dawley 6-month-old rats, weighing 450-550 g, were used. The animals were housed

E.N. Damianopoulos, R.J. Carey/ Behavioural Brain Research 68 (1995) 219-228

in individual cages with a continuous access to food and water. Cages were in a climate (22 °C and 55~o humidity) and light controlled room with 12 h light/dark cycles. Beginning with the arrival into the laboratory from the animal supplier (Taconic Farms, Germantown, NY), the animals were weighed, handled and then placed into 25 × 17 × 17 cm individual wire cages dally for 7 days. Subsequently, saline i.p. was administered once per day for 3 days to habituate the animals to the drug injection protocol. All treatments occurred during the light cycle and the experimental protocols were approved by animal care committees of both SUNY Health Science Center and VAMC at Syracuse 2.2. Apparatus and behavioral measurement

Locomotor behavior was video-recorded by a videoimage analyzing system (Colburn Instruments, Lehigh Valley, PA) in two separate compartments. One was a white 60 cm square chamber with 40 cm side panels while the other was a white round chamber of 70 cm diameter with a 40 cm high wall enclosure. The floor of the square chamber was lined by a light brown polysorb pan liner while the floor of the round chamber was lined by a white polysorb pan liner. The test chambers were in a darkened room illuminated by two overhead 12 V projection lamps (with a red filter) placed above and next to the two videocameras 50 cm above the center of each test chamber. Before placement into a test chamber, the animal's head area was blackened by a marker to facilitate monitoring of the animal's movements and to serve as the object focus. Spontaneous activity as well as field location of the animal was video-recorded and also analyzed by the videoimage analyzing system. In the system employed, on-line analog camera images of the freely-moving animal in each chamber are digitized and converted into total distance traveled (cm). The scores are calculated and printed at each 5 min bin of the 20 min session as well as the total scores at the end of the 20 min session. Two printers, one for each camera and microprocessor were outside the test room so that their operation would be inaudible to the animals in the test chambers. To directly observe specific behavioral responses, the animal's behavior in each test chamber was also videotaped during the first and last drug treatment trials and during the post-treatment non-drug tests for conditioning. Subsequently, this behavior was assessed and provided an independent check on the locomotion distance scores (m). Locomotion distance was directly measured by tracing the animal's movement on the video-screen using a Dietzgen rotary distance measurement calculator (Dietzgen, Syracuse, NY).

221

2.3. Drugs

Cocaine hydrochloride (MaUinckrodt Specialty Chemical, St. Louis, MO) and dizocilpine (MK-801) (RBI, Natick, MA) were dissolved into sterile distilled H20 in separate preparations and brought to appropriate concentration, 10 mg/ml for cocaine or 0.10 mg/ml for MK-801. All injections were i.p. 2.4. Design and procedure

Following matching of the animals on bodyweight and distance traversed in 3 days of 20 min pre-treatment assessments in each chamber, six animals each were assigned to four treatment groups: (a) saline; (b) cocaine (10 mg/kg); (c) MK-801 (0.10 mg/kg); and (d) combined MK-801 plus cocaine. In all treatment groups, the animal initially was placed into the non-drug (square) chamber without injection for a 20 min assessment. Immediately upon completion, the animal was removed and injected with saline, cocaine, MK-801 or MK-801 plus cocaine, depending upon group assignment and immediately placed into the second (round) chamber for another 20 min assessment. The white noise was turned-on during each box placement and turned-off upon removal. The saline/drug treatments were administered 6 times, every other day, over a 12-day period. Following a 1-day and then again after a 21-day period of withdrawal and non-testing, the animals were tested for conditioning of the drug induced responses to the exteroceptive cues of the drug associated chamber. To demonstrate context specificity of potential conditioned drug effects, the test protocol was carried out in both test chambers. The test protocol followed in the post-treatment test for conditioning was identical to the acquisition treatment protocol except that there were no injections. Each test was followed on the next day by a reacquisition trial in which the treatment was the same as that of the initial acquisition treatment phase and depended on group assignment. The reacquisition drug treatment was intended to off-set any loss in drug response due to extinction effects of the non-drug test trial on the previous day. On the final reacquisition trial, the animals were decapitonized and biochemical tissue and serum assays performed. 2.5. Biochemical assay procedures

The brain was rapidly removed and dissected on a chilled glass plate. A bilateral neostriatum sample dorsal to the nucleus accumbens and a bilateral limbic tissue sample which included the nucleus accumbens, olfactory tubercle on the overlying pyriform cortex were obtained from each animal. Immediately after dissection, the

222

E.N. Damianopoulos, R.J. Carey / Behavioural Brain Research 68 (1995) 219-228

samples of brain tissue were weighed, placed in tubes containing 0.5 ml of 0.1 M perchloric acid and 4.5 #1 of 10 #g/ml dihydroxybenzylamine (DHBA) as an internal standard and then homogenized and centrifuged. The resulting supernatant was filtered through 0.2 #m pore filters (Bioanalytical Systems) and the extracts were stored at - 7 0 ° C until the H P L C - E C analysis which was completed within 24-72 h. The tissue samples were analyzed for concentration levels of dopamine, DA (3-hydroxytyramine), and the dopamine metabolite, D O P A C (3,4dihydroxyphenylacetic acid). Trunk blood at decapitation was also collected. Half of each sample was diluted 1:1 with 0.4 M perchloric acid for the catecholamine and corticosterone assay. The other half of the sample was placed in a tube containing 200 #1 of 0.5 ~ sodium fluoride for the cocaine assay. To prepare serum samples for cocaine and corticosterone measurement, a solid phase extraction was used. For cocaine, the extraction column was a Narc 2, 3 ml (125 rag) column (J.T. Baker, Phillipsburg, NJ). 0.5-1.0 ml of serum was used depending upon availability. 100~o Acetonitrile was added to the serum (3:1 acetonitrile to serum) and centrifuged for 5 rain at 2500 r.p.m. The superuatant was decanted and to it a 0.1 M sodium phosphate buffer, pH 6.1 (2:5 buffer to serum) was added. The pH was between 4 and 6. A cation exchange column (125 mg and 3 ml) was conditioned with methanol and with 0.1M sodium phosphate buffer. Before the column could run dry, the prepared sample was passed through the column. The column was then washed with 3 ml H P L C grade water, 3 ml 0.1 M HC1 and 10 ml methanol. The cocaine was eluted with 1 ml methylene chloride/isopropanol/ammonium hydroxide (77:19:4). The sample was then evaporated under a stream of nitrogen and reconstituted in 0.2 ml of buffer. Both serum and brain samples were filtered through 0.2 #m pore filters and injected into the H P L C column for measurement of cocaine. For cocaine, the mobile phase was 0.02 M KH2PO4, pH 3.0 (76 ~o) and acetonitrile 24~o. A 100 mm × 4.6 mm Alltech Adsorbosphere catecholamine 3 # column was used. The flow rate was 0.5 ml/ minute with a UV setting of 235 nm. Another solid phase extraction procedure was followed in preparing the second plasma sample for corticosterone analysis. The extraction column was a C18 3 ml (500 mg) column. Under vacuum, the column was conditioned with 2 × 3 methanol followed by 2 × 3 ml H P L C grade H20. Before the column could dry, 0.5-2.0 ml of plasma (depending upon availability of samples) was passed through the column and immediately followed with a 2 ml H P L C grade H20/acetonitrile wash (80:20). Next, the column was air-dried for 3 min. Finally, the sample was eluted with 2 × 0.5 ml methanol. The samples were injected into a BAS phase II C18 reverse phase column (4.6 × 250 mm,

5/~m) with a mobile phase of 60~o MeOH, 40~o H20 run at a flow rate of 1.0 ml/min. A BAS variable wavelength UV detector was used with the setting at 254 nm.

3. Results 3.1. Acquisition treatment phase: non-drug test environment

The effects of the four drug/saline treatments (saline, MK-801, cocaine and MK-801 + cocaine) on locomotion during each 20 min test session for the 6 treatment days of acquisition were analyzed by a 2-way ANOVA with repeated measures. The statistical design in the acquisition treatment phase was 4 × 6 design. The first component of the design was the between-S variable of the four drug/ saline treatment groups, while the second component was the within-S variable of the 6 days of drug/saline treatment. The F-test results in the non-drug chamber indicated no group effect (F3/2o = 2.27, P > 0.05), nor day of treatment effect (F5/loo = 1.61, P > 0 . 0 5 ) or an interaction effect of group × day of treatment (F15/10o = 0.27, P > 0.05). Thus, the four groups were equivalent in terms of locomotion activity and habituation exhibited across days of treatment prior to the drug/saline injection and placement into the drug associated test chamber. To assess the development of habituation, within-session interval effect analyses were performed using a 2-way ANOVA for the four treatment groups × four 5-rain intervals of the 20 min session. These analyses were carried out for each day separately; i.e., for days 1-6. As in the group × day ANOVAs, again there were no significant group effects but highly significant interval effects were observed ranging from F3/6o = 37.09, P < 0 . 0 0 1 - F 3 / 6 0 = 93.35, P < 0.001). On each day, locomotion level declined from the 0-5 min to the 15-20 rain of the 20 rain session. The interaction of group × interval yielded NS-F-tests on all six analyses. 3.2. Acquisition treatment phase." drug~saline test environment

In contrast to the non-drug test chamber, the results of the same F-test analyses for locomotion in the drug associated chamber indicated a highly significant group effect (F3/2o = 69.90, P<0.001) but no day of treatment effect (F5/1oo = 0.98, P > 0.05) or an interaction effect of group × day of treatment (F15/~o0=0.54, P>0.05). Sequential 1-way ANOVAs were performed to identify pair-wise group differences. The results indicated that the cocaine treatment group and the MK-801+cocaine treatment group showed statistically significant higher levels of locomotion from the saline group during each of the 6 treatment days (P<0.05) and that MK-801 and saline were

E.N. Damianopoulos, R.J. Carey / Behavioural Brain Research 68 (1995) 219-228

statistically equivalent (P>0.05). These results demonstrate that cocaine induced a locomotor stimulant effect as shown by the increased locomotion levels of both cocaine treated groups. In addition, the MK-801 treatment was submotoric as it did not attenuate the cocaine stimulant effect in the MK-801 + cocaine treatment group and, since also it did not differ from the saline control treatment group in overall levels of locomotion throughout the 6 day acquisition treatment phase. To assess the development of habituation and the drug impact on habituation, again within-session interval effect analyses were performed using 2-way ANOVAs for the four treatment groups x four 5-min intervals of the 20 min session. These analyses were also carried out separately for each day; i.e., for days 1-6. As might be expected, there were significant group effects as already reported for the group x day of treatment analysis but also highly significant interval effects ranging from F3/6o = 22.95, P < 0.001-F3/60 = 91.10, P < 0.001). On each day, locomotion level declined from the 0-5 min to the 15-20 min interval of the 20 min session. The interaction

25

3.3. Post-acquisition tests for drug conditioning." non-drug test en vironment The results of the two post-acquisition non-drug tests are shown in Fig. 1. The locomotion results were first analyzed by an overall 2-way ANOVA for repeated measures for the two non-drug tests for conditioning in the non-drug test chamber and then also in the drug test chamber. The F-test results for the non-drug test chamber indicated no statistically significant group mean differences (F3/2o = 2.29, P > 0.05) and no repeated test effects in the comparison of the first (day 1)vs. the second (day 21)tests for conditioning (F1/2o=0.07, P>0.05. Thus, the four treatment groups were equivalent in the non-drug test chamber during both tests immediately prior to the test for

2s

r I i

15

DAY 21: Non Drug Box

T

2O

T

T 0

of group x interval yielded NS-F-tests on all six analyses. Thus, both within and across sessions, within-session habituation was unaffected by the MK-801, cocaine or MK801 + cocaine drug treatments.

DAY 1: Non Drug Box

20

223

~'

15

-k Y

i ] i

0

E 0

10-

o

0 0 ,,-I

10

O

i

5

5-

0

Saline

Cocaine

I

MK-801 Cocaine +

°

0 Saline

Cocaine

MK-$01

2s

DAY 1: Drug Box

i

2s

MK-801

Cocaine + MK-801

DAY 21: Drug Box .IF.X=

20 ,"

.o

20

15

,-

.o

T

.-x- p<0.01

15

0

_J

,.,J

5

5

0

0

Saline

Cocaine MK-801 Cocaine + MK-801

Saline

Cocaine

MK-801 Cocaine + MK-801

Fig. 1. Means and S.E.M.-values of locomotion distance on the post-treatment day 1 and day 21 tests for conditioning. Upper panels show the results in the non-drug box. Lower panels show the results in the drug box. The asterisk indicates a significant group mean difference ( P < 0.01) compared to all other groups.

224

E.N. Damianopoulos. R.J. Carey / Behavioural Brain Research 68 (1995) 219-228

conditioning in the drug/saline test chamber. This result indicates that there were no residual carry-over drug effects and, therefore, no increased level of arousal, in the MK801, cocaine and MK-801 + cocaine treatment groups (top panels of Fig. 1). 3.4. Post-acquisition tests for drug conditioning: drug/saline test environment

A 2-way ANOVA of locomotion in the drug/saline test chamber during the two tests for conditioning revealed an overall group effect (/73/20= 14.32, P<0.001), a test day effect (F1/20 = 10.16, P < 0 . 0 1 and also group ×test day interaction effect (F3/20 = 3.43, P<0.05). A subsequent 1-way ANOVA revealed that the group effect did not reach statistical significance (/73/20= 1.61, P > 0.05 on the day 1 test for conditioning (bottom left panel of Fig. 1); but, following the 21-day withdrawal, a significant group effect was evident (F3/20 = 7.57, P<0.01). A sequential test for independent group effects for the day 21 test results indicated that only the cocaine group had a higher level of locomotion compared to all other groups ( P < 0.01) while the other three groups had equivalent locomotion levels (P>0.05) (bottom right panel of Fig. 1). These results taken together indicate that cocaine treatment induced a conditioned locomotor stimulant effect which was context specific as it did not occur in the non-drug test chamber. The day 21 positive result for conditioning could not be attributed to an increased general arousal in the drug treated group as the groups were equivalent in the nondrug box and could not be attributed to blocked habituation or drug state dependent habituation as the positive test results occurred after 21 days of withdrawal when differential habituation effects are at minimal level (see [ 5 ]). Importantly, these conditioning test results, revealed that the MK-801 treatment which in acquisition was shown to be nonmotoric in terms of its effect on the cocaine induced stimulant effect on locomotion, nonetheless, blocked the development of the cocaine conditioned drug effect on locomotion. This anticonditioning effect was shown by the virtual equivalence of the MK-801 + cocaine group with the saline and MK-801 groups in both tests for conditioning. The absence of an MK-801 effect on the first conditioning test also indicates that the MK-801 or the MK-801 +cocaine treatment did not have a drug state dependent antihabituation effect. 3.5. Reacquisition drug/saline tests following each test for conditioning

The locomotion results are presented in Fig. 2 for the two drug/saline tests which were administered 1 day following each test for conditioning. As can be seen from

Fig. 2, the results for the non-drug test chamber (top panels) for the first and second drug/saline tests show a virtual group identity in the non-drug test chamber. A 2-way ANOVA of the results on the first postconditioning drug test indicated no group effect (F3/2o=0.61, P>0.05), a highly significant within-session effect from the first to the fourth 5-min interval (F3/60)=58.70, P<0.001 but no group x interval effect (F9/6o = 0.25, P > 0.05). The same F-test analyses on the second postconditioning drug/saline test indicated again no group effect (F3/20 = 0.98, P > 0.05), an interval effect (F3/60 = 97.37, P < 0.001) and no group × interval interaction (F9/6o = 2.12, P > 0.05). These results indicate group equivalence in locomotion and in withinsession habituation effects in the non-drug box as locomotion in the treatment groups declined from the first to the fourth 5-min interval. In contrast to the group equivalence shown in the nondrug test chamber, the locomotion results in the drug/ saline chamber in the first post-conditioning drug test, as shown in the bottom left panel of Fig. 2, revealed a group effect (F3/20 = 12.51, P<0.01); also, a highly significant interval effect (F3/20 = 64.22, P<0.01); but no interaction of group × interval (F9/60 = 1.29, P>0.05). The results of the second post-conditioning drug test (bottom right panel of Fig. 2) which occurred 22 days later showed a significant group effect (F3/2o = 9.11, P < 0.01), an interval effect (F3/6o = 9.11, P < 0 . 0 1 ) and again, no significant group × interval interaction (/79/6o= 0.06, P > 0.05). The presence of an within-session interval effect coupled with an absence of an interaction of interval with group effects suggests that all groups were equivalent in within-session habituation and that none of the drug treatment effects interfered with habituation within each session. These F-test analyses were sequentially followed by 1-way ANOVAs to determine the specific group effects. The analyses revealed that only the cocaine ( P < 0.01) and the MK-801 +cocaine differed statistically from the saline control group treatment results (P<0.01). To determine whether there were changes in drug efficacy following the 21 day withdrawal, the results of the two postconditioning drug tests were analyzed by a 3-way ANOVA for group x test day x interval effects. The results indicated a group effect (F3/50 = 19.32, P < 0.01) and an interval effect (F3/2o0 = 97.86, P<0.001) but no group x interval interaction (F9/2oo = 0.47, P > 0.05) or 3-way group x interval x test day interaction (F9/2oo=0.67 P>0.05). No other simple effects or interactions were significant. Altogether, the postconditioning drug/saline test results demonstrate that the efficacy of cocaine whether administered alone or in combination with MK-801 remained the same in the two tests and it maintained its stimulant effect on locomotor behavior over the 21-day withdrawal.

E.N. Damianopoulos, R.J. Carey / Behavioural Brain Research 68 (I995) 219-228

225

NON DRUG BOX Reaquisition 1

Reaquisition 2 16

15

o • [] []

141210cJ

Saline Cocaine MK-801 Cocaine/MK-801

14-

U 1,..

10

~ ._~

8-

6

"0 ¢..

c

6

6

0

4 2 0 0-5

6-10

11-15

16-20 rnln

0-5

6-10

11-16

16-20 mitt

DRUG BOX Reaquistion 1

16

14-

14

_= =)

f-

~

12 10-

®

~

8p.

1210-

C

"o (11

Reaquisition 2

16

6-

"0

~

6-

6-

Q 4-

~

42-

2" 0

0-5

6-10

11-15

16-20 min

0-5

6-10

11-15

16-20 rain

Fig. 2. Means and S.E.M.-values of locomotion distance during the drug tests on the first (left panels) and second (right panels) reacquisition trials 1 day after each test for conditioning (i.e., after the 1-day and 21-day non-drug tests for conditioning). Upper panels show the results in the non-drug box. Lower panels show the results in the drug box. See text for statistical analyses.

3.6. Biochemical analyses Tissue concentrations in/~g/g of wet tissue of dopamine (DA) and of the dopamine metabolite (DOPAC) in limbic and striatal tissue samples were determined for the four treatment groups. Inasmuch as the major effect of cocaine on central dopamine is by the attenuation of dopamine reuptake, dopamine turnover ratios were determined. The results of this analysis are shown in Fig. 3. As can be seen from Fig. 3, the results for the two cocaine treated groups both showed significantly lowered limbic, but not striatal, DA turnover ratios (P<0.01). Measurements of plasma cocaine and of plasma corticosterone concentrations in/~g/ml for the four treatment groups are shown in Fig. 4. Since the saline and MK-801 groups had no detectable cocaine, only the data for serum cocaine in the two cocaine treated groups were compared statistically. A comparison of the cocaine results by a t-test analysis, as shown in the top panel of Fig. 4, revealed no significant difference between the two cocaine treated

groups (t= 0.65, P>0.05). In addition, MK-801 did not interfere with cocaine transport to the brain as the cocaine concentration in the striatal and limbic tissue samples for the cocaine and for the cocaine + MK801 groups did not

0.14 0.12 o (tl

0.10

IT,<

0.08

0

0.06

0 t~

0.04 0.02 0

Saline

Cocaine

MK-801 Coc/MK-801

Fig. 3. Means and S.E.M.-values ofdopamine turnover ratios (DOPAC/ DA) in the limbic and striatum for the saline, cocaine, MK-801 and MK-801 +cocaine treatment groups. Asterisk indicates a significant group mean difference (P<0.01) compared to all other groups.

E.N. Damianopoulos, R.J. Carey / Behavioural Brain Research 68 (1995) 219-228

226 0.30 -

0.25

_T__ 0.20

.= 0.15

o 0 0.10 -

E

?, K

0.05 -

0 Saline

Cocaine

MK-801

Cocaine/ MK-801

0,25

-

,,ltE

0.20

5_

-

0.15 -

T

o ~0

0.10 -

o O

0.05 -

Saline

Cocaine

MK-801

Cocaine/ MK-801

Fig, 4. Means and S.E.M.-values of plasma cocaine and plasma corticosterone concentrations in #g/ml. Asterisk indicates a significant group mean difference (P
differ statistically (t = 1.23, P > 0.05 and t = 0.78, P > 0.05) for striatal and limbic tissue samples, respectively. The cocaine effect on plasma corticosterone levels for the two cocaine treated groups and for the saline and vehicle treated groups was analyzed by a 1-way ANOVA to test for group differences. The F-test showed a significant group effect (F3/2o=8.87, P<0.01). As shown in the bottom panel of Fig. 4, however, the cocaine treatment group induced a significantly elevated ( P < 0.01) plasma corticosterone concentration compared to all other groups. Critically, the MK-801 + cocaine treatment group did not differ from the saline or MK-801 groups.

4. Discussion The major observations of the present study are the findings of an MK-801 blockade of the cocaine conditioned locomotor stimulant effect and of the cocaine induced corticosterone release. As revealed by both the acquisition results as well as by the post-conditioning drug/ saline tests, the MK-801 dose level employed 0.1 mg/kg had no locomotor stimulant effects either by itself or in combination with cocaine. Thus, the MK-801 blockade of the conditioned cocaine stimulant effect could not be at-

tributed to an attenuation of the drug induced cocaine locomotor stimulant effect. In addition, MK-801 did not affect habituation during the drug treatment phase or induce a state-dependent habituation effect in that the MK801 treatment groups did not differ from the saline group on the day 1 non-drug test. Another salient observation is that while the cocaine alone group revealed a conditioned drug effect on the post-treatment day 21 test for conditioning, this effect was not observed on the day 1 test. The mean locomotion activity difference of the cocaine group compared to all the other treatment groups was in the right direction but did not reach statistical significance. Inasmuch as day 1 data alone are inconclusive with respect to conditioning because of the contribution of several factors including possible withdrawal induced response suppression effects (false negative results) and blocked habituation and state dependent habituation (false positive resuits), it is the day 21 test results which are the most relevant to the issue of conditioning. Importantly, in connection with theoretical interpretations of behavioral sensitization, the day 21 positive results for a cocaine conditioned drug effect occurred in the absence of a cocaine sensitization effect during acquisition. The present results impact on interpretations of behavioral sensitization of psychostimulant drug effects. According to one generally accepted conceptualization [24], behavioral sensitization is context specific and may be accounted for by an algebraic summation of the unconditioned and conditioned drug effects. The development of conditioned cocaine locomotor activation in the present study, therefore, should have added to the cocaine unconditioned drug effect leading to a progressive increase in the locomotor stimulant effect across trials selectively in the cocaine (alone) treatment group. Neither in this study nor in previous studies [5,7] did behavioral sensitization occur in that locomotor activity did not increase across days of acquisition treatment. Furthermore, the cocaine and cocaine+MK-801 groups had equivalent locomotion stimulant levels across all drug tests even though the MK801 + cocaine group did not develop a conditioned stimulant drug response. Thus, in the present study, no behavioral sensitization to cocaine was observed (perhaps, due to the intermittent drug regimen); but yet, a conditioned cocaine stimulant effect did occur. Since, in previous studies, we and others have observed behavioral drug sensitization without drug conditioning [6,18,19], it would appear therefore, that behavioral sensitization and conditioned psychostimulant drug effects have now been demonstrated to be doubly dissociated (see also [29]). Alongside its proposed role in cellular mechanisms underlying CNS plasticity [ 16,32,33 ], investigators delineating dopaminergic corticostriatal feedback loops involving a glutamatergic component have attempted to analyze the

E.N. Damianopoulos, R.J. Carey / Behavioural Brain Research 68 (1995) 219-228

N M D A receptor role in behavioral activation processes [9] and in attentional sensorimotor gating functions [3,9]. For example a nonmotoric dose of the non-competitive N M D A antagonist MK-801 (0.1 mg/kg) when combined with a nonmotoric dose of the D1 receptor agonist SKF38393 (1.5 mg/kg) leads to substantial motoric activation effects as expressed in contralateral rotation in the unilateral 6-OHDA lesion animal model [9]. When applied systemically or directly into the nucleus accumbens, MK-801 retarded habituation to startle reflex stimulus occurs [3], implicating a sensory blocking mechanism of sensory input at the thalamic sensorimotor convergence site. Projections from the sensory cortex into discrete neostriatal zones which then feed back through the globus pallidus and substantia nigra pars reticulata dopaminergicGABAergic outflow sites and from the thalamus into the motor cortex suggest that interference with glutamatergic neurotransmission at the medium spiny neurons of the striatum [9] may play a role in sensorimotor integration processes such as those involved in sensitization and conditioning. The present results, however, with the nonmotoric dose of MK-801 did not suggest either a blocking of between- or within-session habituation to the test environment exteroceptive cues. In this sense, the MK-801 treatment did not appear to interfere with sensory processing of exteroceptive stimulation. Thus, the blocking effect by MK-801 on the conditioned cocaine locomotor stimulant effect cannot be attributed to an information processing attenuation vis-~t-vis the exteroceptive drug/ saline compartment stimuli to which the cocaine stimulant effects might have been associated. Instead, the present results point to a linkage between a cocaine induced release of corticosterone and a cocaine conditioned effect. In the present study, the well-established neurochemical effect of cocaine, decreased DA turnover, was induced by cocaine and it was the same for the cocaine and the combined MK-801 plus cocaine groups. The decrease in the D O P A C / D A ratios is consistent with a decrease in intraneuronal dopamine metabolism induced by increased autoreceptor inhibition as a consequence of cocaine inhibition of the dopamine reuptake transporter. The observation of this effect selectively in the limbic brain sample is consistent with reports of limbic system mediation of cocaine locomotor effects [ 13]. As might be expected, only the two groups which received cocaine had detectable levels of cocaine in serum and brain; and, importantly, the levels were equivalent. In contrast to this result, however, the two groups which received cocaine did not have an equivalent elevation in plasma corticosterone. In fact, the MK-801 plus cocaine group had a plasma corticosterone level which was indistinguishable from that of the saline control group. Thus, in addition to blocking the conditioned cocaine behavioral response, MK-801 also blocked

227

the cocaine induced corticosterone response. Given that MK-801 did not impact upon the behavioral response induced by cocaine or its neurochemical action at a target site brain structure (limbic area) thought to have a critical role in the mediation of the cocaine locomotor stimulant effect [ 13], it would appear that corticosterone release by cocaine may be involved in the cocaine conditioning process. The 10 mg/kg cocaine dose used in the present study has also been shown to induce a conditioned place preference indicative of a positive hedonic effect [21,27]. Possibly, the reward effects of cocaine are also impacted by cocaine release of corticosterone. Importantly, however, the route of administration is also a relevant variable pertinent to cocaine reward mechanisms [21]. Since corticosterone can act as a neurosteroid [8] and can influence a variety of brain systems (e.g., medial prefrontal cortex and hippocampus), it may be through this hormonal feedback that certain selected central events initiated by cocaine (e.g., medial prefrontal cortex serotonin response) are augmented and strengthened and may be critical for the initiation and establishment of the conditioned cocaine behavioral response to the stimuli of the drug associated environment. Significantly, corticosterone has recently been shown to have positive reinforcement effects [23]. In a previous report [2], we observed an enhancement in serotonin metabolism in the medial prefrontal cortex in animals exhibiting a cocaine conditioned response. Since the medial prefrontal cortex is modulated by corticosterone, it may be that this hormonal release by cocaine provides the critical feedback necessary for conditioning. Additionally, the absence of both conditioning and the corticosterone response in the cocaine + MK-801 group of the present study indicates that locomotor stimulation is not sufficient either for conditioning or for corticosterone release. While the N M D A receptor has been implicated in many neurobehaviorai processes, the present observation of a potential critical role for the N M D A receptor in neuroendocrine responses provides a new framework in which the neurobehaviorai effects of N M D A manipulations need to be evaluated.

Acknowledgements This research was supported by NIDA RO1DA05366-09 and VA Merit Review Grant.

Grant

References [1] Beretta, S., Robertson, H.A. and Graybiel, A.M., Dopamine and glutamate agonists stimulate neuron-specific expression of Fos-like protein in the striatum, J. NeurophysioL, 68 (1992) 767-777.

228

E.N. Damianopoulos, R.J. Carey / Behavioural Brain Research 68 (1995) 219-228

[2] Carey, R.J. and Damianopoulos, E.N., Conditioned cocaineinduced hyperactivity: an association with increased medial prefrontal cortex serotonin, Behav. Brain Res., 62 (1994) 177-185. [3] Carlsson, A., Thirty years of dopamine research. In H. Narabayshi, T. Nagatsu, N. Yanagisawa and Y. Mizuno (Eds.), Advances in Neurology, Vol. 60, Raven Press, New York 1993, 1-10. [4] Dai, H. and Carey, R.J., The NMDA antagonist MK-801 can impair attention to exteroceptive stimuli, Behav. Brain Res., 62 (1994) 149-156. [5] Damianopoulos, E.N. and Carey, R.J., A new method to assess pavlovian conditioning of psychostimulant drug effects, J. Neurosci. Methods, 53 (1994) 7-17. [6] Damianopoulos, E.N. and Carey, R.J., Apomorphine sensitization effects: evidence for envronmentally contingent behavioral reorganization processes, Pharmacol. Biochem. Behav., 45 (1993) 655663. [7] Damianopoulos, E.N. and Carey, R.J., Conditioning, habituation and reorgzanization factors in chronic cocaine effects, Behav. Brain Res., 49 (1992) 149-157. [8] De Kloet, E.R., Adrenal steroids as modulators of nerve cell function, J. Steroid Biochem., 20 (1984) 175-181. [9] Di Chiara, G. and Morelli, M., Dopamine-acetylcholine-glutamate interactions in striatum: a working hypothesis. In H. Narabayshi, T. Nagatsu, N. Yanagisawa and Y. Mizuno (Eds.), Advances in Neurology, Vol. 60, Raven Press, New York, 1993, pp. 102-106 [10] Druhan, J.P., Jakob, A. and Stewart, J., The development of behavioral sensitization to apomorphine is blocked by MK-801, Eur. J. PharmacoL, 243 (1993) 73-77. [11] Heidbreder, C., Goldberg, S.R. and Shippenburg, T.S., Inhibition of cocaine-induced sensitization by the delta-opioid receptor antagonist naltrindole, Eur. J. Pharmacol., 243 (1993) 123-127. [12] Karler, R., Calder, L.D., Chaudhry, I.A. and Turkanis, S.K., Blockade of 'reverse tolerance' to cocaine and amphetamine by MK-801, Life Sci., 45 (1989) 599-606. [13] Koob, G.F., Drugs of abuse: anatomy, pharmacology and function of reward pathways, Trends Pharmacol. Sci., 13 (1992) 177-184. [14] Koob, G.F., Stinus, L. and LeMoal, M., Hyperactivity and hypoactivity produced by lesions to the mesolimibic dopamine system, Behav. Brain Res., 3 (1981) 35-41. [15] LeMoal, M. Galey, D. and Cardo, B., Behavioral effects of local injections of 6-hydroxydopamine in the medial ventral tegmentum in the rat. Possible role of mesolimbic dopaminergic system, Brain Res., 88 (1975) 190-194. [16] Lynch, G. and Baudry, M., Reevaluating the constraints on hypotheses regarding LTP expression, Hippocampus, 1 (1991) 9-14. [17] Massieu, L., Rocamora, N., Paiacios, J.M. and Boddeke, H.W., Administration of quinolinic acid in the rat hippocampus induces expression of c-fos and NGFI-A, Mol. Brain Res., 16 (1992) 8896. [18] Mattingly, B.A., Gotsick, J.E. and Matin, C., Locomotor activity and stereotypy in rats following repeated apomorphine treatments

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32] [33]

at 1-, 3- or 7-day intervals, PharmacoL Biochem. Behav., 31 (1989) 871-875. Mattingly, B.A., Gotsick, J.E. and Salamanca, L., Latent sensitization to apomorphine following repeated low doses, Behav. Neurosci., 102 (1988) 553-558. Morelli, M. and DiChiara, G., MK-801 potentiates dopaminergic D 1 but reduces D2 responses in the 6-hydroxydopamine model of Parkinson's disease, Eur. J. Pharmacol., 182 (1990) 611-612. Nomikos, G.G. and Spyraki C., Cocaine-induced place conditioning: importance of route of administration and other procedural variables, Psychopharmacology, 94 (1988) 119-125, Pert, A., Post, R. and Weiss, S.R., Conditioning as a critical determinant of sensitization induced by psychomotor stimulants. In L. Erinoff (Ed.), Neurobiology of Learning and Memory NIDA Research Monograph, Vol. 97, 1990, pp. 208-240. Piazza, P.V., Deroche, V., Rouge-Pont, F, Deminiere, J.M., Maccari, S., LeMoal, M. and Simon, H., Individual differences in the sensitivity to corticosterone's reinforcing effects and in corticosterone-induced dopamine release may be a biological basis for sensation-seeking, Soc. Neurosci., 18 (1992) 1076. Post, R.M., Weiss, S.R., Pert, A. and Fontana, D., Conditioned components of cocaine sensitization, Clin. Neuropharmacol., 15 (Suppl. 1, part A) (1992) 650A-651A. Pudiak, C.M and Bozarth, M.A., L-NAME and MK-801 attenuate sensitization to the locomotor-stimulating effect of cocaine, Life Sci., 53 (1993) 1517-1524. Snyder-Keller, A.M., Striatai c-fos induction by drugs and stress in neonatally dopamine-depleted rats given nigral transplants: importance of NMDA activation and relevance to sensitzation phenomena, Exp. Neurol., 113 (1991) 155-165. Spyraki, C., Nomikos, G.G. and Varonos, D.D., Intravenous cocaine-induced place preference: attenuation by haloperidol, Behav. Brain Res., 26 (1987) 57-62. Stewart, J. and Eikelboom, R., Conditioned drug effects. In L.L. Iversen, S.D. Iversen and S.H. Snyder (Eds.), Handbook of Psychopharmacology, Plenum Press, New York, 1987, pp. 1-57. Stewart, J. and Vezina, P., Conditioning and behavioral sensitization. In P.W. Kalivas and C.D. Barnes (Eds.), Sensitization in the Nervous System, Telford, Caldwell, NJ 1988, pp. 207-224. Swerdlow, N.R., Vaccarino, F.J., Amalri, M. and Koob, G.F., The neural substrates for the motor-activating properties of psychostimulants: a review of recent findings, Pharmacol. Biochem. Behav., 25 (1986) 233-248. Torres, G. and Rivier, C., Cocaine-induced expression of striatal c-fos in the rat is inhibited by NMDA receptor antagonists, Brain Res. Bull., 30 (1993) 173-176. Watkins, J.C. and Collingridge, G.L., The NMDA Receptor, Oxford University Press, London, 1989. Zalutsky, R.A. and Nicoll, R.A., Comparison of two forms of long-term potentiation in single hippocampal neurons, Science, 248 (1990) 1619-1624.