Psychopharmacology (1992) 107:221-228 Psychopharmacology ~ Springer-Verlag 1992

Ontogeny of cocaine hyperactivity and conditioned place preference in mice Giovanni Laviola, Giacomo Dell'Omo, Enrico Alleva, and Giorgio Bignami

Section of Behavioural Pathophysiology, Laboratorio di Fisiopatologia di Organo e di Sistema, Istituto Superiore di Sanitfi, Viale Regina Elena, 299, 1-00161 Roma, Italy

Received July 30, 1991 / Final version October 30, 1991

Abstract. Conditioned place preference (CPP) pro- cedures using jointly visual and tactile cues (white compartment with a wide-mesh metal floor versus black compartment with a narrow-mesh floor) were employed to assess the ontogenetic pattern of cocaine reinforcing properties in outbred CD 1 mice. A classical 11-day-long schedule, in which the drug experience occurred in the initially less-preferred compartment ("biased" procedure, Spyraki 1988), served to study cocaine (0, 1, 5, or 25 mg/kg IP repeated four times at 48 h intervals) during the early postweaning stage (21-32 days). The result was a fully-fledged CPP at all cocaine doses. A subsequent experiment used a shortened (4-day) "un- biased" CPP schedule (animals assigned at random to drug experience in one or the other compartment); this enabled an assessment of the ontogenetic pattern of the drug action (single treatment, same dose range) in pups of both sexes at three different developmental ages (14-17, 21-24, or 28-31 days). At the 25 mg/kg dose, CPP developed in animals of all ages, while the 5 mg/kg dose was effective only in 21-24 day pups and the 1 mg/kg dose was ineffective. No significant sex differences were found, but the use of the unbiased procedure enabled a demonstration of an interaction between treatment, age, and type of CS. At the preweaning stage, CPP was due mainly to an increased preference for the black/narrow- mesh compartment, while at the early postweaning stage it consisted mainly of an increased preference for the white/wide-mesh compartment; at the late postweaning stage the cue and the treatment factor did not interact. At all ages tested cocaine induced a dose-dependent in- crease in locomotor activity which was much more mark- ed at both post-weaning stages than before weaning.

Key words: Behavioural development - Ontogeny of locomotor activity - Conditioned place preference - Co- caine - Mouse

Off))rint requests to: G. Bignami

Literature data on the rewarding (positive reinforcing) properties of cocaine as assessed by Conditioned Place Preference (CPP) procedures, are mainly limited to adult male subjects (see reviews by Spyraki 1988; Cart et al. 1989; Hoffman 1989; for recent data on interactions between treatment and test variables see Nomikos and Spyraki 1988; Lawley and Kantak 1990). In fact, an ontogenetic analysis of such properties must overcome some important methodological difficulties. For exam- ple, the more commonly used CPP procedure (van der Kooy 1987; Carr et al. 1989; Swerdlow et al. 1989) requires a training period [association of repeated drug administration (unconditioned stimulus, UCS) and ex- ternal stimuli (CS)] which lasts between 6 and 8 days, while the entire schedule needs several additional days to be completed. This may cause a confounding between treatment and organismic variables during periods of fast developmental change, particularly in the case of a drug such as cocaine, whose effects can strongly depend on previous exposure (see e. g. the sensitization to the activ- ity enhancing effects over repeated treatments; Roy et al. 1978; Pan et al. 1991). On the other hand, Bardo and colleagues (1986) have shown in adult rats that a single IP cocaine administration can suffice to produce CPP, which may help circumventing some of the difficulties just mentioned.

In general, developmental work on altricial rodent CPP has been limited so far to the early postnatal phase; due to sensory and motor immaturity of the subjects, only olfactory, gustatory, and/or tactile CS have been used (Camp and Rudy 1988; Sullivan and Hall 1988; Lariviere et al. 1990; Spear 1990; Weller and Blass 1990). As concerns c0caine, the drug appears to be able to produce important effects, such as hyperactivity, from the early postnatal stage (Spear and Brick 1979). More- over, an early development of the positive reinforcing properties of the drug is suggested by the finding in rats of a CPP on postnatal day 18 after a drug experience in the presence of an odor cue at the age of 2 days (Smith and Holman t987). At this point, it seemed advisable to perform a more systematic study of cocaine CPP in

222

immature animals, using cues known to be effective in adults.

In a first experiment (expt 1), weanling mice pretested to assess their initial (unconditioned) place preference underwent the cocaine experience on the non-preferred side ("biased" procedure), according to a classical CPP schedule using combined visual and tactile stimuli as CS and multiple treatments. To better characterize cocaine reinforcing properties at successive pre- and postweaning stages, we then used (expt 2) a much shorter (4-day, single treatment) CPP "unbiased" schedule (animals as- signed at random to drug experience in one or the other compartment). The focus was on three developmental ages; namely, (i) preweaning, with tests on days 14-17 (eye opening starts around day 12-13 in the strain used; Calamandrei and Alleva 1989), (ii) early postweaning, with tests on days 21-24, and (iii) late postweaning (days 28-31). Experiment 2 was also aimed at comparing males and females, since only a few reports have taken into account sex differences (Goudie et al. 1978; Lett 1989), and at providing adequate data on activity both during the cocaine conditioning session and in the drug-free state after conditioning.

Materials and methods

Animals, breeding, and rearing conditions

Male and nulliparous female mice of a Swiss-derived outbred CD 1 strain (25-28 g) were purchased from Charles River Italia (22050 Calco, Italy). Upon arrival at the laboratory the animals were housed in an air conditioned room (temperature 21 :t: 1 ° C, relative humidity 60:1:10 %) with a reversed light-dark cycle (lights on from 9:30 p.m. to 9:30 a.m.). Pellet food (Enriched Standard Diet pur- chased from Piccioni, 25100 Brescia, Italy) and tap water were con- tinuously available. Males and females were housed separately in groups of 8-10 in 42 x 27 x t5 cm Plexiglas boxes with sawdust as bedding and a metal top. After 2-3 weeks, breeding pairs were formed and housed in 33 x 13 x 14 cm Plexiglas boxes. The females were inspected daily at noon for delivery (postnatal day 0). The stud was removed 10 days before the expected date of delivery. A total of ten litters (culled at birth to six male pups) were used for experi- ment 1, while a total of 36 litters (culled at birth to four male and four female pups) were used for experiment 2. At weaning (day 20) all pups were separated from their mothers and maintained not more than six animals per cage (experiment I) or in same-sex groups of four (experiment 2).

Apparatus

Tests were done in a laboratory room isolated from the animal colony and maintained under standard humidity and temperature conditions. The place conditioning apparatus consisted of a rectan- gular wood chamber divided in three different compartments separated by removable walls. The two end-compartments mea- sured 1 6 x 1 5 x 3 0 cm high, while the central one measured 8 x 15 × 30 cm high. Two cues, one visual and one tactile, were associated with each of the end-compartments, called respectively White (white painted walls and wide-mesh metal floor; wire of 0.7 mm diameter, mesh width 6 mm), and Black (black painted walls and narrow-mesh floor; wire of 0.2 mm diameter, mesh width 1 mm). The middle compartment (Middle) was a neutral chamber with gray walls and smooth floor, and served as starting point. The two partitions separating Middle from White and Black could be

replaced by similar partitions with a 4 × 7 cm opening which allow- ed free access to all compartments. Suspended from the ceiling above the apparatus, at a distance of 160 cm, were two 60 W white bulbs and a video camera used to record the animals' behaviour.

Drug treatment

Cocaine hydrochloride (Merck, 233 Darmstadt) in physiological saline vehicle (O.9% NaCI) was injected IP at 10 ml/kg body weight. The range of doses (see Procedures) and the route of drug ad- ministration were chosen on the basis of literature data (Spear and Brick 1979; Nomikos and Spyraki 1988) and of pilot experiments.

Experiment 1: extended (multiple treatment) schedule, "biased" procedure

This CPP procedure required 11 days, including three phases in which animals were trained between 11 a.m. and 6 p.m., each at the same time of the day in successive exposures.

Phase 1 (days 1-2, habituation). Animals were placed singly in the Middle compartment with free access to all compartments, for 30 min on the first and 10 min on the second day. On the latter, the time spent in each of the two end-compartments was recorded to provide a measure of initial (unconditioned) preference to be used for the side assignment of animals (see phase 2), after excluding extreme cases [see Scotes and Siegel (1986) for selection criteria].

Phase 2 (days 3-10, conditioning period). Six balanced groups of eight animals each were formed by random assignment of mice with opposite initial place preferences. Three groups received cocaine hydrochloride treatments (1, 5 or 25 mg/kg given IP) on alternate days immediately before confinement for 30 rain in the initially non-preferred (paired) end-compartment (half of the animals on even days and the other half on odd days); saline was given on the remaining days after confinement to the initially preferred (un- paired) compartment. Three control groups were used: the first (0/0) received saline in both the "paired" and "unpaired" compart- ments; the second (0/no inj) and the third (5/no inj) received respec- tively saline or cocaine (5 mg/kg) in the "paired" compartment, but no injection in the other one.

Phase 3 (day 11, CPP test). CPP assessment was conducted in the drug-free state 24 h after the last exposure of the previous phase. Uninjected mice were placed in the Middle compartment and allow- ed free access to both end-compartments for 10 rain. Time spent in each end-compartment (defined as both forepaws in the compart- ment) was then scored by an observer blind to the assignment of animals to the different groups.

The experimental design was also balanced for time of testing. Due to the use of six groups and a "biased" procedure, such a design did not allow the splitting of individual litters between the various treatment-side combinations. Therefore, a caveat must be that the factorial ANOVAs considering treatment and apparatus side (see Results) could not discount possible litter effects (Denenberg 1984; Sachs and Thomas t985; Chiarotti et al. 1987), which were dealt with in experiment 2.

Experiment 2: shortened (single treatment) schedule, "unbiased" procedure

This experiment (see also Introduction) was designed so as to allow (i) the study of cocaine CPP at specified developmental stages before and after weaning, (ii) the comparison of mice of both sexes, (iii) the collection of data on activity in both the conditioning and the test session, and (iv) the discounting of possible litter effects (split

litter procedure and appropriate mixed-model ANOVA designs). The whole procedure took 4 days, consisting of three phases in which animals were trained or tested between 3 and 8 p.m. (except fbr the ftrst day), each at the same time of the day.

Phase 1 (day 1, habituation). Without prior treatment, animals were allowed to explore the apparatus for 30 rain between 9 and 11 a.m. and for 10 rain between 4 and 7 p.m.

Phase 2 (days' 2 and 3, conditioning). According to a split-litter design, the four males and four females of each of 12 litters within each age were randomly assigned to one of four treatment groups (0, 1, 5 or 25 mg/kg of cocaine). The male and the female pups from the same litter receiving the same drug treatment were assigned to opposite end-compartments, while balancing for side within each treatment-sex combination was achieved across litters. On the first day of the conditioning phase, animals were injected and confined immediately after in the appropriate end-compartment for 30 rain. Animals were videotaped, and locomotor activity was subsequently scored by dividing the video screen in squares (corresponding to 8 x 7.5 cm areas in the apparatus) and counting the crossings of square limits with both forepaws. Twenty-four hours later, all subjects received a saline injection before being confined in the opposite end-compartment.

Phase 3 (day 4, CPP test). CPP assessment was conducted 24 h after the second exposure in the previous phase, and side preference was recorded in the drug-free state, as in expt 1. The following parameters were considered: Time spent in the White, Middle or Black compartments (defined as both forepaws in the compart- ment); time without ambulation in each of the three compartments; total locomotor activity, scored by measuring the number of pas- sages from the Middle to the end-compartments.

This design allowed the performance of various types of ANOVAs. A preliminary series combined all fixed grouping vari- ables (age, sex, side, and treatment) but still had to leave out the litter random variable, since only two of the four sex-side combina- tions in each treatment were represented in each litter. The absence of significant effects or interactions of the sex variable allowed subsequently the performance of additional ANOVAs considering the litter variable (litters within ages as well as sides and treatments within litters). When appropriate, simplified mixed-model ANOVAs were also performed for each age. Finally, the ANOVAs on activity data in the conditioning session also considered two repeated measures variables within subjects; that is, the initial and the final 5-min periods of the session and the five 1-min periods (hereinafter called blocks) within each of the former.

30-

(3 0 g

20

g $

o) E

0 / i -M- 0 /0 1/0 25 /0 O/no inj 5/no inl 5/O

cocaine dose (mg/kg)

Fig. l. Experiment 1 (multiple treatment schedule, "biased" procedure). Conditioning of a place preference in postweaning mouse pups. Mean difference (SEM) between preference scores in the paired end-compartment before and after conditioning. Preference scores were derived by dividing the time spent in the paired compartment by the total time spent in both end-compart- ments (N=8). *P<0.01 in comparisons with the 0/0 group

223

Results

Experiment 1 : extended (multiple treatment) schedule, "biased" procedure

The da ta on init ial (uncond i t ioned) place preference do not need to be repor ted in detail, since an appa ren t slight preference for the b l ack /na r row mesh c o m p a r t m e n t (on the average 56% of the total t ime spent in the two end- compar tmen t s ) failed to reach statistical significance.

40

30

20

10"

14-17 days

0 I ~ t I, l l

4O

r- 3O

~} 2O 1 O 10

0 O

21-24 days

=

40-

30 ¸

20 ¸

10 ¸

28-31 days

0 | ,, | I [ l L l [ l, l

1 2 3 4 5 26 27 28 29 30

m i n u t e s

Fig. 2. Experiment 2 (singte treatment schedule, "unbiased" procedure). Activity (crossings of square limits) of cocaine and control mice of different ages in the single conditioning session on day 2. The vertical bars to the right indicate pooled SEMs (x 2) derived fi'om the appropriate error mean square in the ANOVA. Male and female data were pooled in the absence of a main effect or interactions of the sex variable, resulting in an N of 24 in each age-treatment combination. Data are limited to the initial and final 5-min periods of the 30-min session. [Cocaine dose (mg/kg): (-D) 0 ; ~.-) 1; ( - . ~ 5; ~e ) 251

224

As shown in Fig. 1, all cocaine doses produced a prominent CPP. A 6 (treatment)x 2 (side) ANOVA yielded a significant interaction of these two variables [F(5,24)=4.94; P<0.01; for between-groups com- parisons by Tukey's post hocs see Fig. 1]. There were no significant differences between the three control groups (0/0, 0/no inj, 5/no inj).

Experiment 2." shortened (single treatment) schedule, "unbiased" procedure

The data concerning the initial exposure were similar to the corresponding data in expt 1, apparently suggesting a slight preference for the black/narrow-mesh compart- ment at all ages (56-60%) which was not statistically significant.

Age and cocaine effects on locomotor activity (condition~ in9 session). The activity scores of cocaine and control mouse pups are given in Fig. 2. Overall activity increased with age, with most of the change occurring between the preweaning and the early postweaning stage; this was confirmed by a 3 (age)x 4 (treatment)x 2 (period)x 5 (block) ANOVA [age F(2,33)= 18.60, P < 0.01]. Baseline within-session trends also varied. Before weaning, a low and uniform initial level was followed by an activity increase at the end of the session; by contrast, an adult- like profile prevailed after weaning, which consisted of a high initial activity level followed by a substantial de- crease indicating the occurrence of within-session habituation [in the same ANOVA, age xperiod F(2,33)= 21.83, P<0.01]

300.

g ~,~ 200.

¢• 100' o E

14-17 21-24 28-31

cocaine dose [ ] o

O . . . . 14-17 21-24 28-31

days

Fig. 3. Experiment 2 (same animals as in Fig. 2 and Table 1). Mean time spent in the Middle compartment (upper graph) and mean percent of time without ambulation ("inactivity") in the same com- partment (lower graph) during the 10-rain CPP test after condition- ing (drug-free state)

Cocaine effects were mainly in the direction of an activity enhancement. Sensitivity to drug increased with age, with most of the change again occurring between the preweaning and the early postweaning stage [age x treat- ment x period x block F(24,396) = 3.13, P < 0.01 ; see also below]. A fine-grain (age by age) analysis showed in the younger animals a significant difference between the ini- tial and the final period of the activity session [in the 4 (treatment) x 2 (period) x 5 (block) ANOVA, period

oA

(D

E

c

E

300 14-17 da'

100

0

IS 30

! 0

0

21-24 days

1 0 0

0

300

200

100

0

28-31 days

0 ! 5 25

cocaine dose (mg/kg)

>,

>

o

c~

o~

14-17 days

30 21-24 days

10 ~ ~

0 I . . . . .

28-31 days

2O

0 I 0 1 5 25

cocaine dose (mg/kg)

Fig. 4. Experiment 2 (same animals as in Table 1 and Figs. 2-3). Conditioning of a place preference as a function of age. Mean time (graphs to the left) and mean percent of time without ambulation ("inactivity", graphs to the right) in the paired ([]) and unpaired ([]) compartments during the 10-rain CPP test (drug-free state), *P < 0.05, **P<0.01

F(1,11)=5.72; P<0.05], and a main effect of cocaine [F(3,33)=3.51; P<0.05], without interaction between the two variables. Post-hoes showed that in the first period both 5 and 25 mg/kg of cocaine increased locomo- tion, while in the second period only 5 mg/kg produced such an effect (P < 0.05 or less).

At both postweaning ages, cocaine hyperactivity was generally more marked at the start than at the end of the session [in an ANOVA design like that indicated in the previous paragraph, treatment x period x block Fs(12,132)=8.17 (21-24 days); 12.89 (28-31 days); Ps<0.01]. At both ages, the lowest (1 mg/kg) dose produced some statistically significant effects which, however, were quite small and limited to very short initial and final portions of the test session; that is, a reduction of activity at the start of the session (Ps < 0.01 in the first 1-min block), and in older animals an activity increase at the end of the session (last two blocks, Ps<0.01). The intermediate dose (5 mg/kg) increased activity during the initial 5-rain period in both the 21-24 and the 28-31 day groups (Ps < 0.0t), and this effect was maintained in the last period only in the latter animals (P<0.01). The highest dose (25 mg/kg) produced a much greater hyperactivity at both ages and throughout the session (Ps < 0.01 in all post-hoes on periods and blocks).

Conditionin# effects on activity and place preference (CPP test in the dru9 free-state). The mean number of passages from the Middle to both end-compartments during the CPP test (Table l) was lower at the prewean- ing than at the postweaning ages [in the 3 (age) x 4 (treat- ment) ANOVA, age F(2,33)=30.42, P<0.01], while previous treatment did not affect this measure. As shown in the upper part of Fig. 3, age also affected time spent in the Middle compartment during the CPP test [F(2,33)=14.09; P<0.001],. again in the absence of proactive drug effects. Specifically, mice of the 21-24 day groups spent more time in the said compartment than younger and older animals (Ps < 0.05). The lower part of the same figure shows a similar nonmonotonic trend of time without ambulation [age F(2,276)= 5.06, P < 0.01], which was also affected by cocaine [F(3,276)=3.23, P < 0.05], apparently in an age-dependent fashion. In the absence of a significant interaction of the said variables, post hoes (see the rationale in Wilcox 1987, Chapter 9, pp 187-189) showed that at the lowerage more time was spent without ambulation in the 1 mg/kg than in the other three groups (P<0.05), while no significant dif- ferences were found at subsequent ages.

Table 1. Experiment 2 (same animals as in Fig. 2). Activity (mean number of passages from the Middle to both end-compart- ments ± SEM) during the 10-min CPP test after conditioning (drug- free state)

Age (days) Previous treatment (mg/kg of cocaine)

0 1 5 25

1~17 23.7±1.3 24.2±1.9 5 . 4 ± 1 . 5 27.3±1.5 21-24 39.5±2.3 37.5~2.8 36.2±1.8 36.0±2.1 2~31 39.5±2.5 ~ . 2 ± 2 . 4 41.2±1.7 ~ .2±5 .1

225

As concerns CPP scores (left part of Fig. 4) previous cocaine treatment increased the time spent in the paired compartment [in the 3 (age)x 4 (treatment)x 2 (con- ditioning side) ANOVA, treatment x side F(3,99) = 7.69; P<0.01], without effect or interactions of the age vari- able. The single 1 mg/kg cocaine treatment was obviously ineffective at all ages. Post hocs showed a significant effect of the 5 mg/kg dose only at the early post-wean- ing age (21-24 days, P<0.05), while the highest dose (25 mg/kg) was effective at all ages (Ps < 0.05 or less).

Time without ambulation in the end-compartments (right part of Fig. 4) was progressively reduced with increasing age [in an ANOVA design like that indicated in the previous paragraph, age F(2,33)= 35.03, P < 0.01; Ps<0.05 for post hoes comparing the 14-17 day mice with the 21-24 and the 28-31 day mice]. In the absence of a significal~t main effect or interaction of the treatment variable, appropriate post-hoes (see rationale above) provided evidence that animals treated previously with

150

1 4 - 1 7 days

50

-50

~ " -150 ~

0

o 21-24 days

- - - 150

5O

O

C 5O

150-

28-31 days

°1 E ~s0

-50

-150 0 1 5 25

Cocaine dose (mg/kg)

Fig. 5. Experiment 2 (same animals as in Table 1 and Figs. 2--4). Age effects on conditioning of a place preference to different cues. Mean difference in time spent in paired and unpaired compartments in the 10-min CPP test after cocaine experience in either the black/narrow- mesh ("Black") or the white/wide-mesh ("White") compartment. (~) White; (m) black

226

25 mg/kg cocaine, moved less in the paired than in the unpaired compartment (Ps < 0.05 or less).

Additional analyses were performed to ascertain whether or not the development of CPP was affected by the cues to which the animals were exposed in the drug state (white/wide mesh versus black/narrow mesh com- pound; remark that these analyses had to use difference scores for each age-dose-pairing side combination, to prevent an inordinate proliferation of measures). Besides providing a further confirmation of the CPP produced by cocaine, the data showed significant cue effects and in- teractions. Spedfically (Fig. 5), the CPP produced by the highest dose of cocaine was due to an increase in the preference for the black/narrow mesh compartment at the preweaning stage, and viceversa, to an increase in the preference for the white/wide mesh compartment at the early postweaning stage (Ps<0.01 in all comparisons between the 25 mg/kg Black or White groups, respective- ly, and other Black or White groups of the same age). At the later postweaning stage, the CPP was due to an increase in preference roughly equivalent with both types of compound cues.

Discussion

The main results of the work reported above can be briefly summarized as follows:

1. cocaine is able to produce a dose-dependent loco- motor hyperactivity in developing mice, with a marked increase in the effect between postnatal day 15 and the early postweaning stage (day 22) and a maintenance of a fully-fledged drug hyperactivity at a later stage (day 29), which is partly in contrast with rat data (see below);

2. immature mice have a fully-ftedged capability to develop cocaine CPP in a classical (l 1-day-long) "biased" procedure starting immediately after weaning, the effects of the lowest dose tested (1 mg/kg given four times) being indistinguishable from those of higher doses (5 and 25 mg/kg);

3. CPP can also be obtained at an earlier age and with a single cocaine treatment, requiring a minimum of 5 mg/kg at the early postweaning stage and a higher dose (25 mg/kg) both a week earlier and a week later (for CPP in adult rats exposed to cocaine only once see Bardo et al. 1986).

An initial remark must concern the fact that the as- sessment of the initial (unconditioned) preferences (on postnatal day 22) in the first experiment showed only a slight and non-significant preference for the black/ narrow mesh compartment. This is apparently at vari- ance with the well-known fact that small rodents, such as rats and mice, generally show a marked preference for dark or dimly illuminated places (see e.g. the review by Hill 1978; see, however, Bignami et al. 1985 for a discus- sion of the data which suggest that such a preference may be less clear-cut in mice than in rats). The present results show the need for a more systematic analysis of the role of both species and age factors in determining uncon-

ditioned preferences. In fact, differences in preference profiles might influence the results of CPP tests either per se or by interacting with organismic and test variables such as age, number of exposures, and use of either a "biased" or an "unbiased" procedure (see also Introduc- tion).

As concerns the marked differences between expts 1 and 2 in the threshold dose for the production of CPP, the present data cannot establish whether the failure to produce CPP with single treatment by the lower cocaine dose(s) is due to an insufficient exposure to a weak reinforcer. In fact, sensitization with repeated cocaine treatment is a well-known phenomenon (Lett 1989; Pan et al. 1991); therefore, the CPP obtained after weaning with four doses of 1 mg/kg may be due either to the repetition of CS-US pairings, to sensitization making later treatments more positively reinforcing than the ini- tial ones, or to both. In any event, the present results confirm and extend the data showing the early appear- ance of the cocaine capacity to produce CPP (Smith and Holman 1987; see Introduction), although further ex- periments are needed to assess whether or not visual cues can serve as CS in the absence of tactile cues so soon as the visual function is established.

The experiments using a single cocaine exposure and an "unbiased" procedure failed to show sex differences, but suggested a joint role of the cue and the age factor. Specifically, the cue and the treatment factor did not interact in older animals. By contrast, at the preweaning stage cocaine CPP consisted mainly of an increased preference for the black/narrow mesh compartment, while at the early postweaning stage it consisted mainly of an increased preference for the white/wide mesh com- partment. In the absence of significant place preferences in the initial exposure in the no-drug state, such an interaction is difficult to interprete. It deserves, however, to be taken into account in future studies aimed at a more thorough understanding of age-related differences in CPP mechanisms.

Another interesting aspect of the cocaine CPP de- velopmental profile, which will require a more thorough analysis, is the apparently lower effectiveness of the drug in 28-31 day mice (treated on day 29) than in 21-24 day mice (treated on day 22). Cocaine effects on activity are essentially the same at both ages, which points to a partial ontogenetic dissociation between the effects of the drug on activity and its reinforcing properties. In this context, it appears difficult to account for the initial reduction of activity by the lower cocaine dose at these ages; this statistically significant effect, however, was quite small and its duration was minimal.

Additional dissociation phenomena excluding a role of activity changes in the production of CPP emerge from the analysis of the proactive effects of cocaine on activity in the drug-free state (test session). In fact, these were limited to an increase of the time without ambulation in both the paired and the middle compartment; in the latter case, the effect was clear-cut and statistically signifi- cant only in preweanlings receiving the 1 mg/kg dose, which did not produce CPP. More importantly, overall activity was unaffected by previous treatment.

227

This part of the results profile also deserves to be discussed from a different viewpoint, inasmuch as it ex- tends the known range of nonmonotonic developmental changes in behaviour, including drug responses; that is, changes that can be exploited to investigate successive interactions between regulatory mechanisms which ma- ture at different paces. (A well-known example is the marked increase after eye opening, of rat and mouse activity and sensitivity to monoaminergic stimulants, followed in subsequent weeks by a decrease occurring in some combinations of species and test, but not in others; see data and discussion in Alleva and Bignami 1985; Laviola et al. 1988). This applies in particular to the time spent in the smaller middle compartment of the CPP apparatus and to the percentage of this time without ambulation. In fact, these were significantly greater shortly after weaning than in younger or in older mice, while a small dose of cocaine significantly reduced activ- ity in the said compartment in the pre-weaning, but not in the late post-weaning stage.

On the basis of ethological considerations, such a profile could be tentatively ascribed to the greater ten- dency to return frequently to the familiar nest area, after exploration of the surroundings, around the time of weaning than later on; in fact, the smaller middle com- partment is less dissimilar from the said area than the larger end-compartments. Two-week-old pups are some- where between incompetence and full competence for this type of response, which could account for their reduced tendency to stay and to suppress ambulation in the middle compartment. Moreover, the "paradoxical" (activity reducing) effect of a small cocaine dose, occur- ring at a time of low sensitivity to the more typical (activity enhancing) effect of the drug, could be ascribed to the facilitating effect of mild CNS stimulation on a response not yet completely developed.

With respect to the mechanisms of cocaine CPP production at successive ontogenetic stages, no firm in- ferences can be drawn in the absence of data on other drug effects and interactions, particularly as concerns dopaminergic agonists and antagonists. It should be noted, however, that at least some of the age differences in cocaine CPP appear to be compatible with current hypotheses on ontogenetic changes in monoaminergic mechanisms. For example, Spear and Brake (1983) have suggested that some interactions between age and treat- ment, such as the different responses of 21- and 28-day- old rats to catecholaminergic agonists, could be ex- plained by a late maturation of dopamine autoreceptors, which appear to be part of a negative feedback system (Nowycky and Roth 1978), relative to the maturation of dopamine postsynaptic receptors. The present mouse data on CPP are apparently in agreement with this hypothesis, but the same does not apply to cocaine effects on mouse activity. In this regard, it should be noted that the available behavioural and physiological-biochemical data are still far from being able to explain a striking discrepancy between rat and mouse activity data; that is, the fact that rats show a parallel attenuation of am- phetamine and cocaine hyperactivity between 21 and 28 days of age (Spear and Brake 1983), while developing

mice show a marked reduction of amphetamine effects after a peak sensitivity around the end of the second postnatal week (e.g., Alleva and Bignami 1985), and an opposite trend in the case of cocaine (present data).

In this context, it should also be mentioned that the nigro-striatal mechanism (caudate-putamen) which is implicated in the production of stereotypy (Gazzara et al. 1986), apparently achieves functional maturation later than the mesolimbic mechanism (nucleus accumbens sep- ti), which is held responsible for both the activity en- hancement produced by cocaine and the reinforcing ef- fects of the drug (Gazzara et al. 1986; Leccese and Lyssen 1987). Therefore, after the functional activation of the former mechanism, a single intermediate (5 mg/kg) co- caine dose might exceed the threshold for stereotypy without constituting a strong reward in the case of a single exposure; in other words, an initial aversive com- ponent of the drug action might conflict with the positive reinforcing properties of a limited experience with it.

In conclusion, the demonstration of cocaine CPP in immature mice of different ages and by two widely differ- ing procedures seems to provide a useful tool for de- velopmental studies of reinforcement mechanisms at the behavioural and physiological-biochemical levels. More- over, cocaine CPP in immature animals adds a specific response end-point to those currently available for be- havioural-teratological assessments, particularly those concerning (i) a variety of habit-forming agents whose positive reinforcing properties may rely on a common mesolimbic dopaminergic mechanism, and (ii) anti- dopaminergic neuroleptics widely employed in the clinic.

Acknowledgements. This research was supported as part of the Sub-project on Behavioural Pathophysiology (Project on Non- infectious Pathology) of the Istituto Superiore di Sanit/l and by grants No. 89.04677.CT04 and No. 90.04056.CTO4 from the Con- siglio Nazionale delle Ricerche. We acknowledge Dr. Flavia Chiar- otti for expert statistical advice.

References

Alleva E, Bignami G (1985) Development of mouse activity, stim- ulus reactivity, habituation and response to amphetamine and scopolamine. Physiol Behav 34:519-523

Bardo MT, Neisewander JL, Miller JS (1986) Repeated testing attenuates conditioned place preference with cocaine. Psycho- pharmacology 89: 239-243

Bignami G, Alleva E, Amorico L, De Acetis L, Giardini V (1985) Bidirectional avoidance by mice as a function of CS, US, and apparatus variables. Anim Learn Behav 13:439-450

Calamandrei G, Atleva E (1989) Epidermal growth factor has both growth-promoting and growth-inhibiting effects on physical and neurobehavioral development of neonatal mice. Brain Res 477: 1-6

Camp LL, Rudy JW (1988) Changes in the categorization of ap- petitive and aversive events during postnatal development of the rat. Dev Psychobiol 21 : 25-42

Carr GD, Fibiger HC, Phillips AG (1989) Conditioned place preference as a measure of drug reward. In: Liebman JM, Cooper SJ (eds) The neuropharmacological basis of reward. Clarendon Press, Oxford, pp 264-319

Chiarotti F, Alleva E, Bignami G (1987) Problems of test choice and data analysis in behavioral teratology: the case of prenatal benzodiazepines. Neurobehav Toxicol Teratol 9:179-186

228

Denenberg VH (1984) Some statistical and experimental considera- tions in the use of analysis-of-variance procedure. Am J Physiol 246 :R403-R408

Gazzara RA, Fisher RS, Howard SG (1986) The ontogeny of amphetamine-induced dopamine release in the caudate- putamen of the rat. Dev Brain Res 28:213-220

Goudie AJ, Dickins DW, Thornton EW (1978) Cocaine-induced conditioned taste aversion in rats. Pharmacol Biochem Behav 8: 757-761

Hill WF (1978) Effects of mere exposure on preferences in non- human mammals. Psychol Bull 85:1177-1198

Hoffman DC (1989) The use of place conditioning in studying the neuropharmacology of drug reinforcement. Brain Res Bull 23: 373-387

Lariviere NA, Chen W J, Spear NE (1990) The influence of olfactory context on Pavlovian conditioning and its expression in pre- weanling (16-day-old) and adult rats. Anita Learn Behav 18:179-190

Laviola G, Renna G, Bignami G, Cuomo V (1988) Ontogenetic and pharmacological dissociation of various components of loco- motor activity and habituation in the rat. Int J Dev Neurosci 24: 431-438

Lawley SI, Kantak KM (1990) Postconditioning effects of mag- nesium on cocaine conditioned place preference in mice. Pharmacol Biochem Behav 36:531-538

Leccese AP, Lyssen WH (i987) Lesions of dopamine neurons in the medial prefrontal cortex: effects on self-administration of am- phetamine and dopamine synthesis in the brain of the rat. Neuropharmacology 26: 1303-1308

Lett BT (1989) Repetead exposures intensify rather than diminish the rewarding effects of amphetamine, morphine, and cocaine. Psychopharmacology 98: 357-362

Nomikos GG, Spyraki C (1988) Cocaine-induced place condition- ing: importance of route of administration and other procedural variables. Psychopharmacology 94: 119-t 25

Nowycky M, Roth R (1978) Dopaminergic neurons: role of pre- synapfic receptors in the regulation of transmitter biosynthesis. Prog Neuropsychopharmacol 8: 139-158

Pan H-T, Menacherry S, Justice JB (1991) Differences in the pharmacokinetics of cocaine in naive and cocaine-experienced rats. J Neurochem 56:1299-1306

Roy SN, Bhattacharyya AK, Pradhan S, Pradhan SN (1978) Be- havioural and neurochemical effects of repeated administration of cocaine in rats. Neurophai'macology 17:559-564

Sachs BD, Thomas DA (1985) Differential effects of perinatal an- drogen treatment on sexually dimorphic characteristics in rats. Physiol Behav 34:735-742

Scoles MT, Siegel S (1986) A potential role of saline trials in mot~hine-induced place-preference conditioning. Pharmacol Biochem Behav 25:1169-1173

Smith CA, Holman EW (1987) Rewarding and aversive effects of stimulant drugs in infant rats. Pharmacol Biochem Behav 26:211-215

Spear LP (1990) Neurobehavioral assessment during the early post- natal period. NeurotoxicoI Teratol 12:489-495

Spear LP, Brake SC (1983) Periadolescence: age-dependent behav- ior and psychopharmacological responsivity in rats. Dev Psychobiol 16:83-109

Spear LP, Brick J (1979) Cocaine-induced behavior in the develop- ing rat. Behav Neural Biol 26:401-415

Spyraki C (I988) Drug reward studied by the use of place condition- ing in rats. In: Lader M (ed) The psychopharmacology of addiction. British Association for Psychopharmacology Mono- graph, No 10. Oxford University Press, Oxford, pp 97-114

Sullivan RM, Hall WG (1988) Reinforcers in infancy: classical conditioning using stroking or intra-oral infusion of milk as UCS. Dev Psychobiol 21 : 215-223

Swerdlow NR, Gilbert D, Koob GF (1989) Conditioned drug effects on spatial preference: critical evaluation. In: Boulton AA, Baker GB, Greenshaw AJ (eds) Neuromethods 13: Psycho- pharmacology. Humana Press, Clifton, New Jersey, pp 399-446

van der Kooy D (1987) Place conditioning: a simple and effective method for assessing the motivational properties of drugs. In: Bozarth MA (ed) Methods of assessing the reinforcing proper- ties of abused drugs. Springer, Berlin Heidelberg New York, pp 229-240

Weller A, Blass EM (1990) Cholecystokinin conditioning in rats: ontogenetic determinants. Behav Neurosci t04:199-206

Wilcox RR (i 987) New statistical procedures for the social sciences. Modern solutions to basic problems. Erlbaum, Hillsdale, New Jersey