The Biphasic Effect of Morphine on Odor Conditioning in Neonatal Rats CHRISTOPHER K. RANDALL PHILIPP J. KRAEMER JOHN M. DOSE TIMOTHY J. CARBARY MICHAEL T. BARD0 University of Kentucky Lexington, Kentucky
Three experiments examined the dose-dependent biphasic effect of morphine on odor conditioning in neonatal rats. In Experiment 1 , a single pairing of an odor and a low dose of morphine (0.5 mg/kg) in 5-day-old rats produced an odor preference, relative to an unpaired control group. In Experiment 2, pairing an odor with a high dose of morphine (2.0 mglkg) produced an odor aversion, relative to an unpaired control group. A third experiment compared performance of a group given odor and morphine (2.0 mg/kg) paired to that of two unpaired groups: one given morphine 24 hr prior to and the other 24 hr after odor exposure. The paired group showed an odor aversion relative to both of the unpaired groups, which did not differ. The latter finding suggests that even if morphine metabolism is incomplete after 24 hr, behavior is unaffected. These results are discussed in reference to the functional development of the opioid system in rats. 0 1992 John Wiley & Sons, Inc.
The effects of endogenous and exogenous opioids on development and learning in rats has been studied extensively (see Kehoe, 1988). In fact, in the last decade, the opioid system has emerged as a particularly fruitful area for research, some of which has focused on opiates and associative learning. Adult rats, for example, have been shown to exhibit either conditioned taste preferences or aversions when novel tastes are paired with an injection of a low or high dose of morphine, respectively (Mucha & Herz, 1985, 1986; Lett & Grant, 1989). Several studies have provided evidence that the opioid systems are behaviorally functional in neonatal rats as early as Day 5 postpartum (Coyle & Pert, 1976a; Kehoe & Blass, 1986a, 1986b, 1986c, 1986d, 1986e, 1989). Furthermore, these and other studies Reprint requests should be sent to Christopher K. Randall, Department of Psychology, University of Kentucky, Lexington, KY 40506-0044. Received for publication 21 November 1991 Revised for publication 25 March 1992 Accepted at Wiley 2 April 1992 Developmental Psychobiology 25(5):355-364 (1992) 0 1992 by John Wiley & Sons, Inc.
CCC 0012-1630/92/050355-10$04.00
356
RANDALL ET AL.
provide evidence that exogenous opioids may produce reinforcing effects in neonatal rats. It appears that the neural mechanisms that mediate the biphasic dose-response curve for morphine effects on taste conditioning are present early in development. For example, Kehoe and Blass conducted a number of experiments that directly examined the effects of different morphine doses on taste conditioning in neonatal rats. They first demonstrated that immature rats can form associations between morphine injections and novel tastes (Kehoe & Blass, 1986a). In these studies, a saccharin solution was paired with a low dose (0.5 mg/kg) of morphine on postpartum Day 5. These subjects subsequently drank significantly more saccharin than did unpaired controls when tested on Day 10. In contrast to the taste preference induced by a low dose of morphine, pups given an injection of a higher dose of morphine (2.0 mg/kg) ingested significantly less saccharin than did control subjects. There is presently little information available to indicate whether the biphasic dose-dependent effect of morphine is unique to taste conditioning or whether other types of conditioned stimuli may yield a similar pattern of results. In a second series of experiments, Kehoe and Blass (1986b, 1989) conditioned pups on Day 5 with an artificialorangeodoreitherpairedorunpairedwithintraperitonealinjections ofalow dose of morphine (0.5 mg/kg). Consistent with the taste conditioningdata associated with this low dose, subjects in the paired group spent significantly more time over the orange odor than did subjects in control groups. What is not known, however, is whether the biphasic effect of morphine that occurs with taste conditioning also occurs with odor conditioning; that is, will pairings of a high dose of morphine and odor produce a conditioned aversion to that odor? Theprimarygoalofthepresent series ofexperiments was to replicateandextend the findings by Kehoe and Blass (1986a, 1986b, 1989)outlined above. Specifically, we set out to test the generality of the biphasic dose-dependent effect of morphine in immature rats observed with taste conditioning by employing an odor conditioning procedure. Because the biphasic effect of morphine has been repeatedly shown with tastes, one could argue that this effect might be consumption-based. Demonstrating the same effect with an odor conditioned stimulus (CS) would rule out that possibility. Furthermore, there is evidence that immature ratsfunctionally equate tastes and odors when both are paired with lithium chloride (Kraemer, Kraemer, Smoller, & Spear, 1989).A demonstrationofthe biphasic effect withan odorconditioningprocedure would further support a case for the functional equivalence of odor and taste learning in neonatal rats.
Experiment 1 The goal of the first experiment was simply to replicate in our laboratory the conditioned odor preference effect that Kehoe and Blass (1986a) demonstrated using a low dose of morphine.
Method
Subjects Subjects were twenty-four 4-day-old Sprague-Dawley-derived rats from three litters. The pups were born in our colony maintained at the University of Kentucky.
BIPHASIC EFFECT O F M O R P H I N E
357
All pups were housed with their dam and littermates in standard opaque maternity cages, partially filled with pine shavings. Litters were culled to between 10-12 pups 24 hr following birth, with the date of parturition designated as Day 0. The colony was maintained on a 12 : 12 hr light/dark illumination cycle, with light onset at 0800 hr. Conditioning and testing occurred during the light portion of the cycle.
Appa ra t us Pups were conditioned in standard maternity cages divided into eight individual compartments, each measuring I 1 x 11.5 cm. Approximately 0.5 inches of pine shavings lined the bottom of each cage. Two commercially available heating pads were placed beneath the cages to maintain a floor surface ranging between 34-36°C. The test arena consisted of a black wooden box (27 x 21 x 9 cm), placed on a wire-screen floor that was raised 3 cm above the surface of the test box. Cotton strips were placed underneath the screen floor on both sides of the test box. As many as eight identical boxes could be placed simultaneously under a ceilingmounted video camera. The camera provided video input for a Videomex-V (Columbus Instruments) activity monitor. This device records the movement of a user-defined object across a video monitor screen, measuring the amount and direction of movement. For purposes of this experiment, the system was used to measure the time pups spent on each side of the test box. One cotton strip under each test box contained I ml of orange extract (Kroger brand), while the strip on the remaining side was left untreated.
Procedure Subjects were randomly divided into two groups (n = 12), designated Paired (P) and Unpaired (UP), with each litter equally represented in each group. On Day 4, pups were separated from the dam and permitted to group huddle for 3 hr prior to the experimental manipulation. At the conclusion of the 3-hr period, subjects were weighed and immediately placed into individual compartments in the heated maternity chambers. Subjects were administered an intraperitoneal (IP) injection of either a 0.5 mg/kg morphine sulfate (Group UP) or saline (Group P). The dose of morphine was based upon the salt form of the drug (obtained from National Institute on Drug Abuse, Rockville, MD). Pups were monitored in the maternity chambers for a 1-hr period prior to being returned to the home cage. On Day 5, approximately 24 hr following the treatment on Day 4, subjects were again allowed to group huddle in a heated maternity cage for 3 hr. After this period, subjects were again weighed and were immediately separated into individual compartments in a heated maternity cage. Pine chips in each individual compartment contained 1.O ml of orange extract (Kroger brand). Pups were placed in the odor for 30 min prior to being injected with 0.5 mg/kg morphine (Group P) or saline (Group UP). Pups remained isolated in the heated, odor-filled maternity cages for 1 hr following injections, after which they were returned to the home cage where they remained until they were tested on Day 10. A 5-day retention interval between conditioning and testing has been used previously in neonatal rats (Rudy & Cheatle, 1979; see also Kehoe, 1988). This interval was necessary in order to enhance the sensitivity of our behavioral measure because motoric activity
358
RANDALL ET AL. Experiment 1 .-
W
c
i:! I : 0 0
0 0
10
x 0
I_
+
P
UP
Treatment Group Fig. 1 . Mean percent time visiting the odor CS zone for paired (P) and unpaired (UP) groups in Experiment 1. (Asterisk denotes a significant difference from control group).
may be too low at a younger age to use the spatial odor preference test as a valid measure of conditioned responding. On the test day, pups were removed from the home cage and allowed to group huddle in a heated maternity cage for 3 hr. Each pup was then given a 10-min spatial odor preference test, which was divided into two 5-min intervals. For the initial 5-min interval, pups were placed on the center line separating the odor/ nonodor sides, facing one side wall, and allowed to move freely around the chamber. After 5-min, the pups were briefly removed from the chamber and then reintroduced to the test apparatus, oriented facing the side wall opposite to their initial placement, and allowed to move freely for the final 5-min interval. Total time each pup spent on the odor side of the chamber was recorded during the 10min test period.
Results and Discussion The mean percent time spent on the odor side of the test chamber by each group for the 10-min test session is presented in Figure 1. Odor preference data were analyzed with a two-way mixed factors analysis of variance (ANOVA), with conditioning as the between groups factor (pairedhnpaired) and test as the repeated measure (first and second 5-min test period). Subjects in Group P spent significantly more time over the orange odor (X = 29.6) than did subjects in Group UP (X = 13.9), as revealed by the significant main effect of conditioning, F(1,22) = 5.30, p < .03. Neither the main effect of test, nor the Conditioning x Test interaction reached significance. The results of this first experiment show that, consistent with Kehoe and Blass (1986a), pairing an odor with an injection of a low dose of morphine in 5-day-old rats is sufficient to increase the behavioral preference for that odor, relative to a control group. As Kehoe (1988) pointed out, low doses of morphine consistently produce conditioned taste preferences, whereas high doses of morphine produce conditioned taste aversions. In contrast to taste conditioning, however, this biphasic effect of morphine has not been demonstrated with odor conditioning. The
BIPHASIC EFFECT OF MORPHINE
359
question remaining, therefore, is whether or not pairing an odor with a higher dose of morphine (2.0 mg/kg) would produce an aversion to that odor. This question was addressed in a second experiment.
Experiment 2 In Experiment 1 , although a significant conditioned odor preference was found, subjects in Group P spent only 29% of the test period over the odor side of the test chamber. Rats in Group UP spent only 13.9%of the test session over the orange odor. These rather low preference levels in unpaired control animals do not permit much room for a conditioned odor aversion to appear. For that reason, we attempted to increase the baseline preference level in the unpaired group in this experiment. Unpublished data from our lab suggest that rat pups prefer banana odor over orange odor. Given this information, we hypothesized that an unpaired group exposed to a more preferred odor (banana) should spend more time over the odor zone, thereby augmenting our chances of detecting a conditioned odor aversion. The morphine dose used in this study (2.0 mg/kg) has been shown to produce conditioned taste aversions in neonatal rats (see Kehoe, 1988).
Method Subjects and Apparatus Subjects were forty-eight 4-day-old Sprague-Dawley-derived rats from 4 litters. The pups were housed and maintained exactly as in Experiment 1 and the apparatus used in this study were the same as those described previously.
Procedure Pups were randomly divided into four groups ( n = 12), designated pairedorange (P-0), unpaired-orange (UP-01, paired-banana (P-B), and unpaired-banana (UP-B), with each litter equally represented in each group. The dose level of morphine used in this study (2.0mg/kg) and the addition of the banana odor (Kroger brand) were the only deviations from the procedure described in Experiment 1.
Results The mean percent time spent over the odor zone by each group is presented in Figure 2. Odor preference data were analyzed with a 2 x 2 x 2 ANOVA; the two between factors were conditioning (pairedhnpaired) and odor (orange/ banana), and test (first and second 5-min interval) was the repeated measure. Regardless of the specific odor used, paired subjects spent significantly less time (Z = 8.7%) over the odor zone than did pups in the unpaired groups (X = 26.3%), as revealed by the main effect of conditioning, F(1,44) = 20.46, p < .001. Neither the Conditioning x Test interation nor the main effects of odor or test reached significance.
360
RANDALL ET AL. Experiment 2
I
a) C
0 N L
0 -0
0
C 0
*
P-0
UP-0
P-B
UP-B
Treatment Group
Fig. 2. Mean percent time visiting the orange (0)or banana (B) CS zone for paired (P) and unpaired (UP) groups in Experiment 2. (Asterisks denote significant difference from control group).
Discussion The results of this experiment, in combination with those of Experiment 1, demonstrate that morphine does indeed have a biphasic effect on odor conditioning in immature rats. While a low dose of morphine (0.5 mg/kg) produces a conditioned odor preference, a high dose (2.0 mg/kg) produces a conditioned odor aversion, relative to control subjects. In contrast to our expectations, however, utilizing a preferred odor (banana) in this experiment did not affect the baseline preference level. As shown in Figure 2, subjects in the unpaired groups spent roughly equivalent amounts of time over the odor zone, regardless of the specific odor used. The same is true for the two paired groups. This noticeable lack of a between-group odor difference signifies the generality of the conditioned odor aversion effect. Although between experiment comparisons should be made cautiously, one difference in the results of the two experiments should be mentioned. Specifically, the average odor preference level of unpaired subjects in Experiment 2 6 = 26.3%) appears to be somewhat higher than that of Group UP in Experiment 1 6 = 13.9%). The reason for this baseline difference is not clear. One obvious difference between studies is the difference in drug intensity. It is well known that immature rats metabolize morphine more slowly than adults (Kupferberg & Way, 1963; Yeh & Krebs, 1980). Perhaps the higher dose of morphine (2.0 mg/kg) used in Experiment 2 may have remained in the rat’s system longer than 24 hr in the unpaired control group. Behaviorally, this could result in different levels of odor preferences due to weak, long-delay conditioning. The third experiment was designed to assess this possibility.
Experiment 3 We conducted a third experiment to test the hypothesis that morphine may not be completely metabolized after 24 hr and may residually affect odor preferences in the unpaired control group. In order to test this hypothesis, the procedure was
BIPHASIC EFFECT OF MORPHINE
361
Experiment 3
* P
UP
UP-BACK
Treatment Group Fig. 3. Mean percent time visiting the odor CS zone for paired (P) and two unpaired (UP and UP-BACK) groups in Experiment 3. (Asterisk denotes significant difference from both control groups).
modified to include an additional unpaired group. The unpaired group used in the previous two experiments received a morphine injection on Day 4, 24 hr prior to odor exposure (a forward, explicitly unpaired procedure). Therefore, another group (UP-BACK) was added that received a morphine injection on Day 6, approximately 24 hr following odor exposure (a backward, explicitly unpaired procedure).
Method Subjects and Apparatus Subjects were thirty 4-day-old Sprague-Dawley-derived rats from three litters. The pups were housed and maintained exactly as in the first two experiments. Again, both the conditioning and test apparatus in this study were the same as those previously described.
Procedure Subjects were randomly divided into three groups (n = lo), with each litter represented in each group. The procedure closely followed Experiment 1, differing only in the addition of the second unpaired group (UP-BACK). Group UP-BACK received IP injections of saline on Days 4 and 5 and an IP injection of morphine on Day 6. Groups P and UP both received an additional IP injection of saline on Day 6. The dose level of morphine (2.0 mg/kg) in this experiment was that which produced a conditioned odor aversion in Experiment 2. Only artificial orange odor was used as the CS, since banana odor did not have a significant effect on baseline responding in Experiment 2. A spatial odor preference test was conducted on Day 10, exactly as described in the previous two experiments.
Results The mean percent time spent over the odor zone by each group is presented in Figure 3. The odor preference data were analyzed with a two-way ANOVA,
362
RANDALL ET AL.
with conditioning (paired, unpaired, unpaired-backward) as the between groups factor and test (first and second 5-min interval) as the repeated measure. Animals in Group P spent significantly less time (X = 3.3%) over the odor zone than did subjects in either Group UP (X = 20.3%) or Group UP-BACK (X = 18.5%), as revealed by the main effect of conditioning, F(2,27) = 4.41, p < .03. Both the Conditioning x Test interaction and the main effect of test failed to reach significance. Newman-Keuls’ post-hoc comparisons revealed no significant difference between Groups UP and UP-BACK, although both groups did differ significantly from Group P.
Discussion The replication of the odor aversion following pairings with a high dose of morphine further supports the generality of this effect. The absence of a difference between Groups UP and UP-BACK indicates that, although morphine may not be completely metabolized 24 hr after injection, residual amounts do not noticeably influence odor preferences relative to unpaired subjects. Thus, the apparent baseline shift in the unpaired group given a low dose of morphine in Experiment 1 , relative to the unpaired group given a high dose of morphine in Experiment 2, may simply reflect random variation across experiments.
General Discussion These experiments confirm that morphine does indeed have a biphasic dosedependent effect on odor conditioning, not unlike that which has been shown with taste conditioning in both immature and adult rats (Kehoe & Blass, 1986a, 1986d; Kehoe & Sakurai, 1991; Mucha & Herz, 1985, 1986). Furthermore, demonstrating this effect with an exteroceptive CS, an artificial odor, confirms that the biphasic effect of morphine is not simply a consumption-based effect. Replicating this effect with a second CS modality also advances our understanding of the associative capabilities of the immature rat. Specifically, neonatal rats as young as 5 days postpartum can readily associate an exteroceptive CS with a proprioceptive, or internal, US. The present results reaffirm Kehoe’s (1988) assertion that the neural mechanisms that produce conditioned aversions and preferences in adults are functional in rat pups as young as Day 5. Even if the maturation of the neural mechanisms necessary to produce morphine preferences and aversions is incomplete in immature rats, complete functional maturation might not be a precursor for the behavioral effects demonstrated in these and previous studies. A threshold level of functioning might be sufficient to produce conditioned preferences and aversions. The odor and taste aversions produced in rats with high doses of morphine are similar to those produced with lithium chloride (LiCI). One trial appears to be sufficient to produce both morphine and LiC1-odor aversions in infant and adult rats (Rudy, & Cheatle, 1977, 1979; Miller, Jagielo, & Spear, 1990; Takulis, 1974). Further evidence suggests that LiCl aversions are similar to morphine-produced aversions in that they are reversible with the opioid antagonist naloxone (Kehoe, 1988). In fact, Kehoe (1988) demonstrated that LiCl produces a biphasic dosedependent effect in infant rats similar to that shown with morphine. Because the
BIPHASIC EFFECT OF MORPHINE
363
similarities between LiCl and morphine conditioning are undeniable, it should not be surprising that the opioid receptors have been implicated in the biphasic dosedependent effect of LiCl. The precise neural mechanisms responsible for the biphasic dose-dependent effect of morphine on conditioned responding to odor cues in 5-day-old rats are not presently known. One interesting possibility, advocated by Bechara and van der Kooy (1989, points to the involvement of two distinct opioid receptor subtypes in mediating the rewarding and aversive properties of opiates. The authors point out that microinjections of morphine directly into specific brain areas can produce reinforcing effects by stimulating mu receptors, while systemic injections of morphine produce aversive effects by stimulating peripheral kappa receptors. Other research has corroborated both the existence of these receptor subtypes and their effects on conditioned responding (Leysen, Gommeren, & Niemegeers, 1983; Wuster, Schulz, & Herz, 1981; Mucha & Herz, 1985). Although the ontogeny of the mu receptor in the brain is well documented (Coyle & Pert, 1976b; Spain, Roth, & Coscia, 1985; Wohltmann, Roth, & Coscia, 1982), not much is presently known about the development and distribution of the kappa receptors in the periphery. If these receptor subtypes are responsible for the biphasic effect of morphine, the present results, as well as other data (see Kehoe, 1988), suggest that both of these opioid systems may be functional in rats by 5 days of age.
Notes This research was supported in part by a National Institute of Health grant (2 SO7 RR07114-21) to P. J. Kraemer and a United States Public Health Service grant (DAO5312) to M. T. Bardo. All correspondence should be addressed to Christopher K. Randall, Department of Psychology, University of Kentucky, Lexington, Kentucky, 40506-0044.
References Bechara, A., & van der Kooy, D. (1985). Opposite motivational effects of endogenous opioids in brain and periphery. Nature, 314, 533-534. Coyle, J. T., & Pert, C. B. (1976a). Ontogenetic development of opiate receptor in rodent brain. Brain Research, 118, 157-160. Coyle, J. T., & Pert, C. B. (1976b). Ontogenetic development of 3H-naloxone binding in rat brain. Neuropharmacology, 15, 555-560. Kehoe, P. (1988). Opioids, behavior, and learning in mammalian development. In E. M. Blass (Ed.), Handbook of behavioral neurobiology: Developmental psychobiology and behavioral ecology, Vol. 9, (pp. 309-346). New York: Plenum Press. Kehoe, P . , & Blass, E. M. (1986a). Behaviorally functional opioid systems in infant rats. I. Evidence for olfactory and gustatory classical conditioning. Behauioral Neurosciencc, ZOO, 359-367. Kehoe, P., & Blass, E. M. (1986b). Behaviorally functional opioid systems in infant rats. 11. Evidence for pharmacological, physiological, and psychological mediation of pain and stress. Behavioral Neuroscience, 5 , 624-630. Kehoe, P., & Blass, E. M. (1986~).Central nervous system mediation of positive and negative reinforcement in neonatal albino rats. Developmental Brain Research, 27, 69-75. Kehoe, P., & Blass, E. M. (3986d). Conditioned aversions and their memories in 5-day-old rats during suckling. Journal of Experimental Psychology: Animal Behavior Processes, 12, 40-47. Kehoe, P., & Blass, E. M. (1986e). Opioid-mediationof separation distress in 10-day-old rats: Reversal of stress with maternal stimuli. Developmental Psychobiology, 19, 385-398.
364
RANDALL ET AL.
Kehoe, P., & Blass, E. M. (1989). Conditioned opioid release in ten-day-old rats. Behavioral Neuroscience, 103, 423-428. Kehoe, P., & Sakurai, S. (1991). Preferred tastes and opioid-modulated behaviors in neonatal rats. Developmental Psychobiotogy, 24, 135-1 48. Kraemer, P. J., Kraemer, E. G., Smoller, D. E., &Spear, N. E. (1989). Enhancement of flavor aversion conditioning in weanling but not adult rats by prior conditioning to an odor. Psychobiology, 17 (l), 34-42. Kupferberg, H. J., & Way, E. L. (1963). Pharmacologic basis for the increased sensitivity of the newborn rat to morphine. Journal ofPharmacology and Experimental Therapeutics, 141, 105-1 12. Lett, B. T., & Grant, V. L. (1989). Conditioned taste preference produced by pairing a taste with a low dose of morphine or sufentanil. Psychopharmacofogy, 98, 236-239. Leysen, J. E., Gommeren, W., & Niemegeers, J. E. (1983). 3H sufentanil, a superior ligand for popiate receptors: Binding properties and regional distribution in rat brain and spinal cord. European Journal of Pharmacology, 87, 209-255. Miller, J . S., Jagielo, J. A., & Spear, N . E. (1990). Changes in the retrievability of associations to elements of the compound CS determine the expression of overshadowing. Animal Learning & Behavior, 18, 157-161. Mucha, R. F., & Herz, A. (1985). Motivational properties of kappa and mu opioid receptor agonists studied with place and taste preference conditioning. Psychopharmacology, 86, 274-280. Mucha, R. F., & Herz, A. (1986). Preference conditioning produced by opioid active and inactive isomers of levorphanol and morphine in rat. Life Sciences, 38, 241-249. Rudy, J. W., & Cheatle, M. D. (1977). Odor-aversion learning by neonatal rats. Science, 198, 845-846. Rudy, J. W., & Cheatle, M. D. (1979). Ontogeny of associative learning: Acquisition of odor aversions by neonatal rats. In N. E. Spear & B. A. Campbell (Eds.), Ontogeny of learning and memory (pp. 157-188), Hillsdale, NJ: Erlbaum. Spain, J. W., Roth, B. L., & Coscia, C. J. (1985). Differential ontogeny of multiple opioid receptors (mu, delta, and kappa). Journal of Neuroscience, 5 , 584-588. Takulis, H. (1974). Odor aversions produced over long CS-US delays. Behavioral Biology, 10,505-510. Wohltmann, M., Roth, B. L . , & Coscia, C. J. (1982). Differential postnatal development of mu and delta opiate receptors. Developmental Brain Research, 3, 679-684. Wuster, M., Schulz, R., & Herz, A. (1981). Multiple opiate receptors in peripheral tissue preparations. Biochemical Pharmacology, 30, 1883-1887. Yeh, S. Y ., & Krebs, 8.A. (1980). Development of narcotic drug metabolizing enzymes in the newborn rat. Journal of Pharmacology and Experimental Therapeutics, 213(1), 28-32.
Three experiments examined the dose-dependent biphasic effect of morphine on odor conditioning in neonatal rats. In Experiment 1 , a single pairing of an odor and a low dose of morphine (0.5 mg/kg) in 5-day-old rats produced an odor preference, relative to an unpaired control group. In Experiment 2, pairing an odor with a high dose of morphine (2.0 mglkg) produced an odor aversion, relative to an unpaired control group. A third experiment compared performance of a group given odor and morphine (2.0 mg/kg) paired to that of two unpaired groups: one given morphine 24 hr prior to and the other 24 hr after odor exposure. The paired group showed an odor aversion relative to both of the unpaired groups, which did not differ. The latter finding suggests that even if morphine metabolism is incomplete after 24 hr, behavior is unaffected. These results are discussed in reference to the functional development of the opioid system in rats. 0 1992 John Wiley & Sons, Inc.
The effects of endogenous and exogenous opioids on development and learning in rats has been studied extensively (see Kehoe, 1988). In fact, in the last decade, the opioid system has emerged as a particularly fruitful area for research, some of which has focused on opiates and associative learning. Adult rats, for example, have been shown to exhibit either conditioned taste preferences or aversions when novel tastes are paired with an injection of a low or high dose of morphine, respectively (Mucha & Herz, 1985, 1986; Lett & Grant, 1989). Several studies have provided evidence that the opioid systems are behaviorally functional in neonatal rats as early as Day 5 postpartum (Coyle & Pert, 1976a; Kehoe & Blass, 1986a, 1986b, 1986c, 1986d, 1986e, 1989). Furthermore, these and other studies Reprint requests should be sent to Christopher K. Randall, Department of Psychology, University of Kentucky, Lexington, KY 40506-0044. Received for publication 21 November 1991 Revised for publication 25 March 1992 Accepted at Wiley 2 April 1992 Developmental Psychobiology 25(5):355-364 (1992) 0 1992 by John Wiley & Sons, Inc.
CCC 0012-1630/92/050355-10$04.00
356
RANDALL ET AL.
provide evidence that exogenous opioids may produce reinforcing effects in neonatal rats. It appears that the neural mechanisms that mediate the biphasic dose-response curve for morphine effects on taste conditioning are present early in development. For example, Kehoe and Blass conducted a number of experiments that directly examined the effects of different morphine doses on taste conditioning in neonatal rats. They first demonstrated that immature rats can form associations between morphine injections and novel tastes (Kehoe & Blass, 1986a). In these studies, a saccharin solution was paired with a low dose (0.5 mg/kg) of morphine on postpartum Day 5. These subjects subsequently drank significantly more saccharin than did unpaired controls when tested on Day 10. In contrast to the taste preference induced by a low dose of morphine, pups given an injection of a higher dose of morphine (2.0 mg/kg) ingested significantly less saccharin than did control subjects. There is presently little information available to indicate whether the biphasic dose-dependent effect of morphine is unique to taste conditioning or whether other types of conditioned stimuli may yield a similar pattern of results. In a second series of experiments, Kehoe and Blass (1986b, 1989) conditioned pups on Day 5 with an artificialorangeodoreitherpairedorunpairedwithintraperitonealinjections ofalow dose of morphine (0.5 mg/kg). Consistent with the taste conditioningdata associated with this low dose, subjects in the paired group spent significantly more time over the orange odor than did subjects in control groups. What is not known, however, is whether the biphasic effect of morphine that occurs with taste conditioning also occurs with odor conditioning; that is, will pairings of a high dose of morphine and odor produce a conditioned aversion to that odor? Theprimarygoalofthepresent series ofexperiments was to replicateandextend the findings by Kehoe and Blass (1986a, 1986b, 1989)outlined above. Specifically, we set out to test the generality of the biphasic dose-dependent effect of morphine in immature rats observed with taste conditioning by employing an odor conditioning procedure. Because the biphasic effect of morphine has been repeatedly shown with tastes, one could argue that this effect might be consumption-based. Demonstrating the same effect with an odor conditioned stimulus (CS) would rule out that possibility. Furthermore, there is evidence that immature ratsfunctionally equate tastes and odors when both are paired with lithium chloride (Kraemer, Kraemer, Smoller, & Spear, 1989).A demonstrationofthe biphasic effect withan odorconditioningprocedure would further support a case for the functional equivalence of odor and taste learning in neonatal rats.
Experiment 1 The goal of the first experiment was simply to replicate in our laboratory the conditioned odor preference effect that Kehoe and Blass (1986a) demonstrated using a low dose of morphine.
Method
Subjects Subjects were twenty-four 4-day-old Sprague-Dawley-derived rats from three litters. The pups were born in our colony maintained at the University of Kentucky.
BIPHASIC EFFECT O F M O R P H I N E
357
All pups were housed with their dam and littermates in standard opaque maternity cages, partially filled with pine shavings. Litters were culled to between 10-12 pups 24 hr following birth, with the date of parturition designated as Day 0. The colony was maintained on a 12 : 12 hr light/dark illumination cycle, with light onset at 0800 hr. Conditioning and testing occurred during the light portion of the cycle.
Appa ra t us Pups were conditioned in standard maternity cages divided into eight individual compartments, each measuring I 1 x 11.5 cm. Approximately 0.5 inches of pine shavings lined the bottom of each cage. Two commercially available heating pads were placed beneath the cages to maintain a floor surface ranging between 34-36°C. The test arena consisted of a black wooden box (27 x 21 x 9 cm), placed on a wire-screen floor that was raised 3 cm above the surface of the test box. Cotton strips were placed underneath the screen floor on both sides of the test box. As many as eight identical boxes could be placed simultaneously under a ceilingmounted video camera. The camera provided video input for a Videomex-V (Columbus Instruments) activity monitor. This device records the movement of a user-defined object across a video monitor screen, measuring the amount and direction of movement. For purposes of this experiment, the system was used to measure the time pups spent on each side of the test box. One cotton strip under each test box contained I ml of orange extract (Kroger brand), while the strip on the remaining side was left untreated.
Procedure Subjects were randomly divided into two groups (n = 12), designated Paired (P) and Unpaired (UP), with each litter equally represented in each group. On Day 4, pups were separated from the dam and permitted to group huddle for 3 hr prior to the experimental manipulation. At the conclusion of the 3-hr period, subjects were weighed and immediately placed into individual compartments in the heated maternity chambers. Subjects were administered an intraperitoneal (IP) injection of either a 0.5 mg/kg morphine sulfate (Group UP) or saline (Group P). The dose of morphine was based upon the salt form of the drug (obtained from National Institute on Drug Abuse, Rockville, MD). Pups were monitored in the maternity chambers for a 1-hr period prior to being returned to the home cage. On Day 5, approximately 24 hr following the treatment on Day 4, subjects were again allowed to group huddle in a heated maternity cage for 3 hr. After this period, subjects were again weighed and were immediately separated into individual compartments in a heated maternity cage. Pine chips in each individual compartment contained 1.O ml of orange extract (Kroger brand). Pups were placed in the odor for 30 min prior to being injected with 0.5 mg/kg morphine (Group P) or saline (Group UP). Pups remained isolated in the heated, odor-filled maternity cages for 1 hr following injections, after which they were returned to the home cage where they remained until they were tested on Day 10. A 5-day retention interval between conditioning and testing has been used previously in neonatal rats (Rudy & Cheatle, 1979; see also Kehoe, 1988). This interval was necessary in order to enhance the sensitivity of our behavioral measure because motoric activity
358
RANDALL ET AL. Experiment 1 .-
W
c
i:! I : 0 0
0 0
10
x 0
I_
+
P
UP
Treatment Group Fig. 1 . Mean percent time visiting the odor CS zone for paired (P) and unpaired (UP) groups in Experiment 1. (Asterisk denotes a significant difference from control group).
may be too low at a younger age to use the spatial odor preference test as a valid measure of conditioned responding. On the test day, pups were removed from the home cage and allowed to group huddle in a heated maternity cage for 3 hr. Each pup was then given a 10-min spatial odor preference test, which was divided into two 5-min intervals. For the initial 5-min interval, pups were placed on the center line separating the odor/ nonodor sides, facing one side wall, and allowed to move freely around the chamber. After 5-min, the pups were briefly removed from the chamber and then reintroduced to the test apparatus, oriented facing the side wall opposite to their initial placement, and allowed to move freely for the final 5-min interval. Total time each pup spent on the odor side of the chamber was recorded during the 10min test period.
Results and Discussion The mean percent time spent on the odor side of the test chamber by each group for the 10-min test session is presented in Figure 1. Odor preference data were analyzed with a two-way mixed factors analysis of variance (ANOVA), with conditioning as the between groups factor (pairedhnpaired) and test as the repeated measure (first and second 5-min test period). Subjects in Group P spent significantly more time over the orange odor (X = 29.6) than did subjects in Group UP (X = 13.9), as revealed by the significant main effect of conditioning, F(1,22) = 5.30, p < .03. Neither the main effect of test, nor the Conditioning x Test interaction reached significance. The results of this first experiment show that, consistent with Kehoe and Blass (1986a), pairing an odor with an injection of a low dose of morphine in 5-day-old rats is sufficient to increase the behavioral preference for that odor, relative to a control group. As Kehoe (1988) pointed out, low doses of morphine consistently produce conditioned taste preferences, whereas high doses of morphine produce conditioned taste aversions. In contrast to taste conditioning, however, this biphasic effect of morphine has not been demonstrated with odor conditioning. The
BIPHASIC EFFECT OF MORPHINE
359
question remaining, therefore, is whether or not pairing an odor with a higher dose of morphine (2.0 mg/kg) would produce an aversion to that odor. This question was addressed in a second experiment.
Experiment 2 In Experiment 1 , although a significant conditioned odor preference was found, subjects in Group P spent only 29% of the test period over the odor side of the test chamber. Rats in Group UP spent only 13.9%of the test session over the orange odor. These rather low preference levels in unpaired control animals do not permit much room for a conditioned odor aversion to appear. For that reason, we attempted to increase the baseline preference level in the unpaired group in this experiment. Unpublished data from our lab suggest that rat pups prefer banana odor over orange odor. Given this information, we hypothesized that an unpaired group exposed to a more preferred odor (banana) should spend more time over the odor zone, thereby augmenting our chances of detecting a conditioned odor aversion. The morphine dose used in this study (2.0 mg/kg) has been shown to produce conditioned taste aversions in neonatal rats (see Kehoe, 1988).
Method Subjects and Apparatus Subjects were forty-eight 4-day-old Sprague-Dawley-derived rats from 4 litters. The pups were housed and maintained exactly as in Experiment 1 and the apparatus used in this study were the same as those described previously.
Procedure Pups were randomly divided into four groups ( n = 12), designated pairedorange (P-0), unpaired-orange (UP-01, paired-banana (P-B), and unpaired-banana (UP-B), with each litter equally represented in each group. The dose level of morphine used in this study (2.0mg/kg) and the addition of the banana odor (Kroger brand) were the only deviations from the procedure described in Experiment 1.
Results The mean percent time spent over the odor zone by each group is presented in Figure 2. Odor preference data were analyzed with a 2 x 2 x 2 ANOVA; the two between factors were conditioning (pairedhnpaired) and odor (orange/ banana), and test (first and second 5-min interval) was the repeated measure. Regardless of the specific odor used, paired subjects spent significantly less time (Z = 8.7%) over the odor zone than did pups in the unpaired groups (X = 26.3%), as revealed by the main effect of conditioning, F(1,44) = 20.46, p < .001. Neither the Conditioning x Test interation nor the main effects of odor or test reached significance.
360
RANDALL ET AL. Experiment 2
I
a) C
0 N L
0 -0
0
C 0
*
P-0
UP-0
P-B
UP-B
Treatment Group
Fig. 2. Mean percent time visiting the orange (0)or banana (B) CS zone for paired (P) and unpaired (UP) groups in Experiment 2. (Asterisks denote significant difference from control group).
Discussion The results of this experiment, in combination with those of Experiment 1, demonstrate that morphine does indeed have a biphasic effect on odor conditioning in immature rats. While a low dose of morphine (0.5 mg/kg) produces a conditioned odor preference, a high dose (2.0 mg/kg) produces a conditioned odor aversion, relative to control subjects. In contrast to our expectations, however, utilizing a preferred odor (banana) in this experiment did not affect the baseline preference level. As shown in Figure 2, subjects in the unpaired groups spent roughly equivalent amounts of time over the odor zone, regardless of the specific odor used. The same is true for the two paired groups. This noticeable lack of a between-group odor difference signifies the generality of the conditioned odor aversion effect. Although between experiment comparisons should be made cautiously, one difference in the results of the two experiments should be mentioned. Specifically, the average odor preference level of unpaired subjects in Experiment 2 6 = 26.3%) appears to be somewhat higher than that of Group UP in Experiment 1 6 = 13.9%). The reason for this baseline difference is not clear. One obvious difference between studies is the difference in drug intensity. It is well known that immature rats metabolize morphine more slowly than adults (Kupferberg & Way, 1963; Yeh & Krebs, 1980). Perhaps the higher dose of morphine (2.0 mg/kg) used in Experiment 2 may have remained in the rat’s system longer than 24 hr in the unpaired control group. Behaviorally, this could result in different levels of odor preferences due to weak, long-delay conditioning. The third experiment was designed to assess this possibility.
Experiment 3 We conducted a third experiment to test the hypothesis that morphine may not be completely metabolized after 24 hr and may residually affect odor preferences in the unpaired control group. In order to test this hypothesis, the procedure was
BIPHASIC EFFECT OF MORPHINE
361
Experiment 3
* P
UP
UP-BACK
Treatment Group Fig. 3. Mean percent time visiting the odor CS zone for paired (P) and two unpaired (UP and UP-BACK) groups in Experiment 3. (Asterisk denotes significant difference from both control groups).
modified to include an additional unpaired group. The unpaired group used in the previous two experiments received a morphine injection on Day 4, 24 hr prior to odor exposure (a forward, explicitly unpaired procedure). Therefore, another group (UP-BACK) was added that received a morphine injection on Day 6, approximately 24 hr following odor exposure (a backward, explicitly unpaired procedure).
Method Subjects and Apparatus Subjects were thirty 4-day-old Sprague-Dawley-derived rats from three litters. The pups were housed and maintained exactly as in the first two experiments. Again, both the conditioning and test apparatus in this study were the same as those previously described.
Procedure Subjects were randomly divided into three groups (n = lo), with each litter represented in each group. The procedure closely followed Experiment 1, differing only in the addition of the second unpaired group (UP-BACK). Group UP-BACK received IP injections of saline on Days 4 and 5 and an IP injection of morphine on Day 6. Groups P and UP both received an additional IP injection of saline on Day 6. The dose level of morphine (2.0 mg/kg) in this experiment was that which produced a conditioned odor aversion in Experiment 2. Only artificial orange odor was used as the CS, since banana odor did not have a significant effect on baseline responding in Experiment 2. A spatial odor preference test was conducted on Day 10, exactly as described in the previous two experiments.
Results The mean percent time spent over the odor zone by each group is presented in Figure 3. The odor preference data were analyzed with a two-way ANOVA,
362
RANDALL ET AL.
with conditioning (paired, unpaired, unpaired-backward) as the between groups factor and test (first and second 5-min interval) as the repeated measure. Animals in Group P spent significantly less time (X = 3.3%) over the odor zone than did subjects in either Group UP (X = 20.3%) or Group UP-BACK (X = 18.5%), as revealed by the main effect of conditioning, F(2,27) = 4.41, p < .03. Both the Conditioning x Test interaction and the main effect of test failed to reach significance. Newman-Keuls’ post-hoc comparisons revealed no significant difference between Groups UP and UP-BACK, although both groups did differ significantly from Group P.
Discussion The replication of the odor aversion following pairings with a high dose of morphine further supports the generality of this effect. The absence of a difference between Groups UP and UP-BACK indicates that, although morphine may not be completely metabolized 24 hr after injection, residual amounts do not noticeably influence odor preferences relative to unpaired subjects. Thus, the apparent baseline shift in the unpaired group given a low dose of morphine in Experiment 1 , relative to the unpaired group given a high dose of morphine in Experiment 2, may simply reflect random variation across experiments.
General Discussion These experiments confirm that morphine does indeed have a biphasic dosedependent effect on odor conditioning, not unlike that which has been shown with taste conditioning in both immature and adult rats (Kehoe & Blass, 1986a, 1986d; Kehoe & Sakurai, 1991; Mucha & Herz, 1985, 1986). Furthermore, demonstrating this effect with an exteroceptive CS, an artificial odor, confirms that the biphasic effect of morphine is not simply a consumption-based effect. Replicating this effect with a second CS modality also advances our understanding of the associative capabilities of the immature rat. Specifically, neonatal rats as young as 5 days postpartum can readily associate an exteroceptive CS with a proprioceptive, or internal, US. The present results reaffirm Kehoe’s (1988) assertion that the neural mechanisms that produce conditioned aversions and preferences in adults are functional in rat pups as young as Day 5. Even if the maturation of the neural mechanisms necessary to produce morphine preferences and aversions is incomplete in immature rats, complete functional maturation might not be a precursor for the behavioral effects demonstrated in these and previous studies. A threshold level of functioning might be sufficient to produce conditioned preferences and aversions. The odor and taste aversions produced in rats with high doses of morphine are similar to those produced with lithium chloride (LiCI). One trial appears to be sufficient to produce both morphine and LiC1-odor aversions in infant and adult rats (Rudy, & Cheatle, 1977, 1979; Miller, Jagielo, & Spear, 1990; Takulis, 1974). Further evidence suggests that LiCl aversions are similar to morphine-produced aversions in that they are reversible with the opioid antagonist naloxone (Kehoe, 1988). In fact, Kehoe (1988) demonstrated that LiCl produces a biphasic dosedependent effect in infant rats similar to that shown with morphine. Because the
BIPHASIC EFFECT OF MORPHINE
363
similarities between LiCl and morphine conditioning are undeniable, it should not be surprising that the opioid receptors have been implicated in the biphasic dosedependent effect of LiCl. The precise neural mechanisms responsible for the biphasic dose-dependent effect of morphine on conditioned responding to odor cues in 5-day-old rats are not presently known. One interesting possibility, advocated by Bechara and van der Kooy (1989, points to the involvement of two distinct opioid receptor subtypes in mediating the rewarding and aversive properties of opiates. The authors point out that microinjections of morphine directly into specific brain areas can produce reinforcing effects by stimulating mu receptors, while systemic injections of morphine produce aversive effects by stimulating peripheral kappa receptors. Other research has corroborated both the existence of these receptor subtypes and their effects on conditioned responding (Leysen, Gommeren, & Niemegeers, 1983; Wuster, Schulz, & Herz, 1981; Mucha & Herz, 1985). Although the ontogeny of the mu receptor in the brain is well documented (Coyle & Pert, 1976b; Spain, Roth, & Coscia, 1985; Wohltmann, Roth, & Coscia, 1982), not much is presently known about the development and distribution of the kappa receptors in the periphery. If these receptor subtypes are responsible for the biphasic effect of morphine, the present results, as well as other data (see Kehoe, 1988), suggest that both of these opioid systems may be functional in rats by 5 days of age.
Notes This research was supported in part by a National Institute of Health grant (2 SO7 RR07114-21) to P. J. Kraemer and a United States Public Health Service grant (DAO5312) to M. T. Bardo. All correspondence should be addressed to Christopher K. Randall, Department of Psychology, University of Kentucky, Lexington, Kentucky, 40506-0044.
References Bechara, A., & van der Kooy, D. (1985). Opposite motivational effects of endogenous opioids in brain and periphery. Nature, 314, 533-534. Coyle, J. T., & Pert, C. B. (1976a). Ontogenetic development of opiate receptor in rodent brain. Brain Research, 118, 157-160. Coyle, J. T., & Pert, C. B. (1976b). Ontogenetic development of 3H-naloxone binding in rat brain. Neuropharmacology, 15, 555-560. Kehoe, P. (1988). Opioids, behavior, and learning in mammalian development. In E. M. Blass (Ed.), Handbook of behavioral neurobiology: Developmental psychobiology and behavioral ecology, Vol. 9, (pp. 309-346). New York: Plenum Press. Kehoe, P . , & Blass, E. M. (1986a). Behaviorally functional opioid systems in infant rats. I. Evidence for olfactory and gustatory classical conditioning. Behauioral Neurosciencc, ZOO, 359-367. Kehoe, P., & Blass, E. M. (1986b). Behaviorally functional opioid systems in infant rats. 11. Evidence for pharmacological, physiological, and psychological mediation of pain and stress. Behavioral Neuroscience, 5 , 624-630. Kehoe, P., & Blass, E. M. (1986~).Central nervous system mediation of positive and negative reinforcement in neonatal albino rats. Developmental Brain Research, 27, 69-75. Kehoe, P., & Blass, E. M. (3986d). Conditioned aversions and their memories in 5-day-old rats during suckling. Journal of Experimental Psychology: Animal Behavior Processes, 12, 40-47. Kehoe, P., & Blass, E. M. (1986e). Opioid-mediationof separation distress in 10-day-old rats: Reversal of stress with maternal stimuli. Developmental Psychobiology, 19, 385-398.
364
RANDALL ET AL.
Kehoe, P., & Blass, E. M. (1989). Conditioned opioid release in ten-day-old rats. Behavioral Neuroscience, 103, 423-428. Kehoe, P., & Sakurai, S. (1991). Preferred tastes and opioid-modulated behaviors in neonatal rats. Developmental Psychobiotogy, 24, 135-1 48. Kraemer, P. J., Kraemer, E. G., Smoller, D. E., &Spear, N. E. (1989). Enhancement of flavor aversion conditioning in weanling but not adult rats by prior conditioning to an odor. Psychobiology, 17 (l), 34-42. Kupferberg, H. J., & Way, E. L. (1963). Pharmacologic basis for the increased sensitivity of the newborn rat to morphine. Journal ofPharmacology and Experimental Therapeutics, 141, 105-1 12. Lett, B. T., & Grant, V. L. (1989). Conditioned taste preference produced by pairing a taste with a low dose of morphine or sufentanil. Psychopharmacofogy, 98, 236-239. Leysen, J. E., Gommeren, W., & Niemegeers, J. E. (1983). 3H sufentanil, a superior ligand for popiate receptors: Binding properties and regional distribution in rat brain and spinal cord. European Journal of Pharmacology, 87, 209-255. Miller, J . S., Jagielo, J. A., & Spear, N . E. (1990). Changes in the retrievability of associations to elements of the compound CS determine the expression of overshadowing. Animal Learning & Behavior, 18, 157-161. Mucha, R. F., & Herz, A. (1985). Motivational properties of kappa and mu opioid receptor agonists studied with place and taste preference conditioning. Psychopharmacology, 86, 274-280. Mucha, R. F., & Herz, A. (1986). Preference conditioning produced by opioid active and inactive isomers of levorphanol and morphine in rat. Life Sciences, 38, 241-249. Rudy, J. W., & Cheatle, M. D. (1977). Odor-aversion learning by neonatal rats. Science, 198, 845-846. Rudy, J. W., & Cheatle, M. D. (1979). Ontogeny of associative learning: Acquisition of odor aversions by neonatal rats. In N. E. Spear & B. A. Campbell (Eds.), Ontogeny of learning and memory (pp. 157-188), Hillsdale, NJ: Erlbaum. Spain, J. W., Roth, B. L., & Coscia, C. J. (1985). Differential ontogeny of multiple opioid receptors (mu, delta, and kappa). Journal of Neuroscience, 5 , 584-588. Takulis, H. (1974). Odor aversions produced over long CS-US delays. Behavioral Biology, 10,505-510. Wohltmann, M., Roth, B. L . , & Coscia, C. J. (1982). Differential postnatal development of mu and delta opiate receptors. Developmental Brain Research, 3, 679-684. Wuster, M., Schulz, R., & Herz, A. (1981). Multiple opiate receptors in peripheral tissue preparations. Biochemical Pharmacology, 30, 1883-1887. Yeh, S. Y ., & Krebs, 8.A. (1980). Development of narcotic drug metabolizing enzymes in the newborn rat. Journal of Pharmacology and Experimental Therapeutics, 213(1), 28-32.
