Psychopharmacology (t995) 117 : 23-31

© Springer-Verlag 1995

Gregory I. Elmer • Jeanne O. Pieper Steven R. Goldberg • Frank R. George

Opioid operant self-administration, analgesia, stimulation and respiratory depression in p-deficient mice

Received: 31 December 1993 / Final version: 28 March 1994

It is commonly thought that/,-receptors play an important role in the reinforcing effects of opioids. In the present study, inbred strains widely divergent in CNS opiate receptor densities were used to investigate the influence of genetic variation in receptor concentration on opioid-reinforced behavior. In particular, the CXBK/ByJ mice were used as an investigative tool because of their significantly lower number of CNS /Z opioid receptors. The behavioral pharmacology of opioids in tlhe/z-deficient CXBK/ByJ mice was compared to other commonly used inbred mouse strains, C57BL/6J and BALB/cJ, and the opiate receptor rich CXBH/ByJ mice. Operant opioid reinforced behavior, opioid-induced locomotor stimulation, analgesia and respiratory depression were investigated in all four inbred strains, To assess the acquisition and maintenance of opioid reinforced behavior, oral selfadministration of the potent benzimidazole opioid, etonitazene, was determined using an operant fixedratio schedule of reinforcement (FR 8). Acquisition of etonitazene-reinforced behavior was established in all four strains including the #-deficient CXBK/ByJ mice. However, there were significant genetic differences in the amount of drug intake during the maintenance of opioid-reinforced behavior and extinction behavior following vehicle substitution. For example, drug intake was significantly greater in the BK versus BH mice during the maintenance phase and an extinction burst was seen in the BH but not the BK mice following vehicle Abstract

Gregory I. Elmer ( ~ ) . Jeanne O. Pieper - Steven R. Goldberg Behavioral Pharmacology and Genetics Section, Preclinical Pharmacology Laboratory, National Institute on Drug Abuse, Addiction Research Center Box 5180, Baltimore, Maryland 21224, USA Frank R. George Department of Psychology and Center on Alcoholism Substance Abuse Addictions (CASAA), University of New Mexico, Albuquerque, New Mexico, 87131, USA

substitution.Thus, #-receptor density may not account for individual variability in the acquisition of opioidreinforced behavior under these conditions. Sensitivity to etonitazene-induced respiratory depression, stimulation of locomotor activity and analgesia were unrelated to drug intake during self-administration sessions across these four inbred strains. These data indicate that inherited differences in CNS g-opiate receptor concentrations do not affect acquisition of etonitazenereinforced behavior. Key words Opioid

. Genetics

• Self-administration

•

CXBK/ByJ • Reinforcement

Introduction

The #-receptor subtype has commonly been suggested to be responsible for the primary (Young et al. 1981; Bertalmio et al. 1989; Negus et al. 1993) and conditioned (Shippenberg et al. 1987; Bals-Kubic et al. 1990) reinforcing effects of opioids. In particular, information derived from biochemical and behavioral data demonstrate a significant correlation between an opioids affinity at the/,-receptor and the dose that maintains maximal rates of operant drug self-administration behavior (Woods et al. 1981). In addition to/z-agonists, c~-agonists have also been shown to serve as a reinforcer in operant self-administration and conditioned place preference studies (Tortella et al. 1980; Young et al. 1983; Shippenberg et al. 1987) and the g-antagonist, naltrindole, is effective in altering/z-agonist self-administration in rats (Negus et al. 1993). Thus, variability in an agonists' relative affinity at /~- and possibly 6-receptors may account for differences in the reinforcing property of opioids. In addition to pharmacological variability in an agonists' affinity at/~- or 6 -receptors, genetic variability across subjects has been shown to be an important

24

component in determining an individual's response to opioids. Genetic factors have been demonstrated to contribute significantly to the analgesic (Baran et al. 1976; Vaught et al, 1988), locomotor stimulant (Oliverio and Castellano 1974), respiratory depressant (Muraki and Kato 1986), and conditioned (Cunningham et al. 1992; Suzuki et al. 1993) and operant reinforcing properties of opioids (Suzuki et al. 1992). Therefore, pharmacological variability in #receptor affinity and individual genotypic variability appear to be important components in the acute and reinforcing effects of opioids. If agonist activity at #receptor subtypes is responsible for opioid self-administration behavior, then variability across individuals in opioid self-administration behavior may be a function of genetic variation in CNS #-receptor concentration. Inbred strains are valuable research tools for defining heritability and genetic covariation in response to drugs and in determining biochemical mechanisms underlying acute drug effects (Howerton et al. 1983; Miner et al. 1986; George and Goldberg 1989; Elmer et al. 1990). The CXBK/ByJ (BK) and CXBH/ByJ (BH) mice are excellent examples of genotypes useful for testing hypotheses related to the actions of opioids. The BK mice demonstrate a 33% decrease in opiate receptors compared to the average of the 7 CXB recombinant inbred strains, their progenitors and reciprocal F1 hybrids (Baran et al. 1975), a predominant decrease in the #1 receptor subtype in specific neuroanatomical regions (Moskowitz and Goodman 1985), large shifts to the right in morphine-induced analgesic dose-effect curves (Moskowitz et al. 1985), and similar rightward shifts in other g-receptor mediated behaviors (Baran et al. 1975; Moskowitz et al. 1985; Raffa et al. 1987). As an example of their utility, the use of BK mice was important in demonstrating a significant CNS ~5-component in opiate-induced analgesia (Vaught et al. 1988) a central role for CNS #-receptors in defeat-induced analgesia (Miczek et al. 1982) and the presence of genetically and pharmacologically distinct #-receptor population mediating morphine-induced analgesia (Pick et al. 1993). To complement the BK mice, BH mice have slightly greater than normal CNS naloxone binding sites and are more sensitive to several opiaterelated drug effects (Baran et al. 1975; Jacob et al. 1983; Marek et al. 1988, 1990). The C57BL/6J (C57) and BALB/cJ (BALB) mice are commonly used inbred mouse strains that are intermediate in CNS opiate receptor concentration (Baran et al. 1975; Reith et al. 198t; Belknap et al. 1989) are extremely divergent in measures of opioid preference (C57 high/BALB low: Horowitz et al. 1977) and have been used in numerous studies investigating the operant reinforcing effects of ethanol (Elmer et al. 1987a, b, 1988). In general, these mouse strains are some of the most divergent available for investigating possible causal relationships between opiate receptor concentration, opioid-related pheno-

types, the reinforcing effects of opioids and the commonalty of drug-reinforced behavior across drug class. The purpose of the present study was to use the BK, BH, BALB and C57 mice to investigate operant opioid self-administration behavior in mouse strains that vary significantly in opiate receptor population. The use of specific genotypes with known biochemical differences will assist in determining potential relationships between #-receptor concentration, sensitivity to the acute effects of opioids and opioid-reinforced behavior. Specifically, if #-receptor concentration is responsible for opioid-reinforced behavior and individual variability in susceptibility to opioid self-administration, it would be expected that self-administration behavior and drug intake in these genotypes would reflect their relative sensitivities to opiates and inherited g-receptor concentrations. In addition, analysis of the relationships between spontaneous locomotor activity, the ability of an opioid to induce analgesia, locomotor stimulation, respiratory depression and maintenance of drug-reinforced behavior may prove useful in building a data base necessary to determine the genetic covariance of these traits and in developing predictive models of opioid drug-seeking behavior.

Materials and methods Animals Adult male and female CXBK/ByJ and CXBH/ByJ mice (Addiction Research Center) and male C57BL/6J and BALB/cJ mice (Jackson Laboratory), 6 months old and weighing approximately 27-29 g at the start of their training were used. Male and female BH and BK mice bred at the Addiction Research Center consisted of first and second generation offspring from Jackson Laboratory parental strains. Male and female mice were used in order to maximize the number of these difficult to obtain animals. Animals of each gender were equally distributed across dose. All animals were experimentally naive, housed individually in a temperature-controlled room (21°C) with a 12-h light-dark cycle (0700-1900 hours lights on), and given free access to Purina Laboratory Chow and tap water prior to initiation of the experiments. Drug The potent benzimidazole opioid, etonitazene, was used in all experimental procedures. Etonitazene was used instead of morphine primarily because it is more potent, less bitter and readily absorbed when given orally, as opposed to morphine that is bitter tasting and poorly absorbed orally. Etonitazene was administered as the base for all dosing procedures. Analgesia Etonitazene-induced analgesia was measured by the hot-plate test (Franklin and Abbot 1989). The subject was placed on the 5 5 ° C hot-plate 30 min post-SC injection of 0, 1, 3, 10, 30, or 56 mg/kg etonitazene. A cut-off time of four times each respective strain's baseline latency was used to avoid tissue damage. Two dependent measures were used to assess antinociception: 1) latency to first

25 observable response to heat as described by Belknap et al. (1990) (i.e. paw shake, extended paw lift) and 2) latency to paw lick (hind or front). Data are presented as the percent of maximal analgesic response as determined by the following formula: 100 x [(latency to first response or paw-lick) - (saline baseline latency)] + [(4 x baseline latency) - (baseline latency)]. A between subjects design was used to determine dose-effect curves (n = 6-9 mice per dose per strain).

the number of spout contacts yielding liquid delivery during each stimulus presentation was held constant throughout the study (10), the reinforcement schedule can be more simply referred to as a Fixed Ratio (FR) n, where n equals the number of lever presses required to activate the liquid delivery system. The start of the session was signaled by illumination of a white house light. System control and data acquisition and storage were accomplished on Apple lie computers located in an adjacent room.

Locomotion

Procedure overview

Baseline locomotor activity and locomotor activity induced by etonitazene was monitored in an Omnitech Activity Monitor (Omnitech Electronics, Columbus Ohio). Activity in the monitor was recorded by photobeam interruptions. Subjects were confined to one-quarter of the activity monitor (22.5 cm L x 22.5 cm W) by Plexiglas walls. All activity measurements were conducted in an isolated room under red light between the hours of 12:00 p.m. and 4:00 p,m. Baseline levels of activity were determined for each strain by placing the naive subject in the activity monitor without prior treatment (n = 5-6 per strain). To measure etonitazene-induced locomotor activity, subjects were placed in the activity monitor for 10 min in order to acclimate, then injected SC with 0, 0.03, 0.10, 0.30, 1.0, 3.0, 10.0, 30.0, 100.0, 300.0, or 417.0 mg/kg etonitazene. The dose-range administered to each strain was chosen based on preliminary data. All doses were administered at an injection volume of 10 ml/kg. Data were collected in 5-rain intervals and are presented as the total number of counts ctunulated during the 40-rain session. The effect of each dose was determined in separate animals (n = 6-9 mice per dose per strain).

The following sequence of procedures was used in each strain to assess the establishment of etonitazene as a reinforcer; a) all mice were trained to an F R 8 schedule of water reinforcement, b) gradually exposed to the pharmacological effects of etonitazene using post-prandial procedures, then c) tested for drug reinforced behavior in the absence of post-prandial inducement. This approach was used to establish satisfactory baseline levels of responding for water in all genotypes and to provide sufficient exposure to the interoceptive effects of the drug. Exposure to the pharmacological effects of etonitazene was achieved via post-prandial drinking of gradually increasing etonitazene concentrations. It was equally important that all strains be exposed in such a way as to avoid development of aversion to the taste or other pharmacological aspects of etonitazene. Since these inbred strains vary greatly in sensitivity to the various effects of etonitazene, the initial etonitazene concentrations were low and increases in concentration were made in small increments to avoid possible aversion development. Tests of etonitazene reintbrcement were made in the absence of post-prandial induction at a final concentration of 1 gg/ml. All sessions throughout the study were 30 min in length and ran between 1200 and 1700 hours.

Respiration

Training Respiratory rate was determined by monitoring minute changes in pressure caused by the subjects breathing in an isolated chamber. The system consists of two sealed 1.5 1 glass containers (one experimental chamber and the other a filter chamber) and one reference chamber used to minimize the effects of ambient room pressure changes (Columbus Instruments, Columbus, Ohio). R o o m air was continuously pumped through each chamber at a rate of approximately 1 1/min. Respiratory rate was based on the relative number of minute pressure changes in the experimental chamber versus a reference chamber. Each system was enclosed in a sound attenuating chamber illuminated by red lights. Respiration rate was monitored for 60 min immediately following administration of etonitazene. Full dose-response curves were obtained for each strain using a range of etonitazene doses administered SC (0, 1, 3, 10, or 30 gg/kg) in an injection volume of 10 ml/kg. Male and female BH mice were used in these experiments. The effect of each dose was determined in separate animals (n = 6 4 mice per dose per strain). Operant etonitazene-reinforced behavior

Apparatus Eight mouse operant chambers were used and have been described in detail elsewhere (Elmer et aI. 1986, 1990). Briefly, the lever consisted of a balanced rocker arm designed to break an infrared photo beam when 0.5 g of force was applied. A solenoid delivery system, adapted from a system developed by Beardsley and Meisch (1981), delivered a small amount of liquid (approximately 2 gl) in response to a spout lick. Initially, a single lever press turned on stimulus lights above the spout that signaled availability of liquid delivery; in the presence of the stimulus lights, spout contacts resulted in liquid delivery. According to Ferster and Skinner (1957) this schedule is termed a heterogeneous chain Fixed Ratio 1 (lever press): 10 (Fixed Ratio l) (spout contact) schedule of reinforcement. Since

The procedures used to establish lever pressing and spout licking responses were similar to those used in other mouse strains as well as other spedes (Elmer et al. 1986, 1990; Meisch 1975). All mice were food restricted to 80% of free feeding weights at 6 months of age and maintained at this weight for the duration of the study by rationing daily food allotments. Daily allotments of Purina Laboratory Chow (approximately 3.5 g depending upon each mouse) were placed in the home cage 60 rain prior to the beginning of each session. During each session water was available from the drinking spout. Once drinking reliably occurred, activation of the spout was made contingent upon a single lever press. Water bottles were restored to the home cages when the response chain was reliably established under the FR 1 condition (typically 1 day). The response requirement in the first component of the reinforcement schedule (lever presses) was then increased from 1 to 2, 4, 6, and 8. Three to five sessions were run at each FR for all strains. Responding for water at the F R 8 condition was maintained for at least ten sessions to ensure that extensive training and satisfactory response patterns were successfully established in all mice in order to control for any initial effects of Iearning on subsequent responding when water was replaced by etonitazene delivery.

Post-prandial exposure to concentrations of 0.25-1.0 t~g/ml etonitazene. Sixty minutes prior to each session, the daily allotment of food was given in the home cage to each of the four strains. Once responding became stable in all strains at 0 gg/ml (tap water), a series of logarithmically increasing etonitazene solutions, in the order 0.25, 0.5, 0.7, and 1.0 gg/ml, replaced water. Five 30-rain sessions were run at each concentration. Behavior maintained by 1 #g/ml etonitazene without prior food presentation. To determine if 1 ~tglml etonitazene had come to function as a reinforcer, behavior maintained by etonitazene was tested without prior food presentation. The amount of food given before

26

Table 1 Baseline Activity 2149 3242 3987 4757

Strain BH C57 BALB BK

Stimulant EDs0* 13. t 0.3 31.0 124.6

Stimulant Efficacy# 294 171 573 535

Drug Intake* 1.9 6.4 8.5 9,7

Analgesic EDs0* 8.2 13.7 5.1 22.7

Respir. ED40* 4.4 3.9 2.1 11.1

* gg/kg; # maximum% increase

the session was decreased in the order of 1, 0.6, 0.3, 0.1, and 0 g, with the remainder of the daily allotment being given after session; at least three sessions at each food amount were run in each strain. In the absence of prior food presentation, 1 tag/rnl etonitazene was tested for five sessions followed by 15 sessions at 0 gg/ml and a 1 gg/ml etonitazene retest. All sessions were 30 min in length. Data analysis The EDs0 and ED40 values were derived from the regression analysis of the linear portion of each dose-response curve. A two-way analysis of variance (ANOVA) was performed for each doseresponse curve in the analgesia, locomotor stimulant and respiratory depression experiments. A repeated-measures analysis of variance (ANOVA) was performed to analyze genetic differences in post-prandial exposure to etonitazene and subsequent tests of reinforcement. Preliminary genetic correlations across the six measured phenotypes and/t and 6 Bmaxdata were conducted to determine the degree of genetic covariance between these traits. Genetic correlations were performed using the inbred strain mean for each phenotype. Binding data for # and 6 were taken from Reith et al., (1981); ~c data from this study were not used due to the questionable specificity of ethlyketocyclazocine in the assay conditions.

[F(Genotype) = 21.7; df = 3, 78; P < 0.0001], [F(Dose) = 113.9; df= 4, 78; P < 0.0001].The EDs0 for latency to first response and latency to paw lick were significantly correlated (P < 0.04), therefore all subsequent measures using analgesic values refer to the latter method for determining antinociception. The BK mice were the least sensitive and the BALB mice the most sensitive to the analgesic effects of etonitazene. There was no significant effects of gender in the BH or BK mice. The EDs0 for etonitazene-induced analgesia as determined by paw lick latency was 5.1, 8.2, 13.7, and 22.7 gg/kg in the BALB, BH, C57, and BK mice, respectively (Table 1). Locomotion Figure 2 (panel A) shows baseline locomotor activity for BK, BALB, C57, and BH mice in the absence of 5000 "~

Results

Analgesia

4000

Eo~ 2 0 0 0

Figure 1 shows etonitazene-induced analgesia as determined by paw lick latency in the BK, BALB, C57, and BH mice as a function of increasing etonitazene dose. There were significant effects of genotype and dose on the latency to the first response [F(Genotype) = 4.8; df = 3, 78; P < 0.004], [F(Dose) = 36.5; df = 4, 78; P < 0.0001] and latency to the paw lick response

o

.J

1000

BH

C57

BALB

BK

Strain

600 500

..........

80

1-

i ,,111

t

- i

i Jl~ll

I'"l

BALB',

I

l--T--I

.~../~..L

IIIii

,~

"

/..

P O

o

400 3O0 200

o-e

10o o 0.01

0.1

1

Etonitazene -

-

0.1

~

f

,

iiiiii

i

I

i

ii,lr

1 10 Etonitazene (pglkg)

i

i,~

i

iii

100

Fig. 1 Percent of the maximal analgesic effect in BK, BALB, C57 and BH mice as a function of increasing etonitazene dose. Each point represents the mean (+ SEM) from six to nine mice

10

100

1000

(gg/kg)

Fig. 2A Baseline locomotor activity counts without prior pretreatment in BK, BALB, C57 and BH mice. Each point represents the mean (+ SEM) from five or six mice. B Locomotor activity in BK, BALB, C57 and BH mice as a function of increasing etonitazene dose. Each point represents the mean (+ SEM) from five or six mice per group

27 any pretreatment. Genotype significantly affected the level of baseline activity [F(Genotype) = 4.6; df = 2,18; P < 0.01]. BK mice were the most active whereas the BH mice were least active under these conditions. Figure 2 (panel B) shows percent of saline control locomotor activity as a function of increasing doses of etonitazene. There were significant effects of genotype, dose and a genotype x dose interaction: [F(Genotype) = 53.6; df = 3,79; P<0.0001], [F(Dose) = 26.7; df = 5,79; P < 0.01], [F(Genotype x Dose) = 3.8; df = 9,79; P < 0.01]. In addition, there was a significant difference across genotype in the efficacy of etonitazene to stimulate locomotor activity. Within these four genotypes there was no relationship between baseline activity in the absence of any pretreatment and the potency or efficacy of etonitazeneinduced locomotor stimulation. There was no significant effects of gender in the BH or BK mice. The etonitazene dose that produced half-maximal stimulation differed significantly across strains; C57 mice were more sensitive, tbllowed by BH, BALB then BK mice. The EDs0s for etonitazene-induced locomotor activity are presented in Table 1.

significantly greater decreases in respiration rate in a linear manner (data not shown). The EL40 (40% of the maximal possible decrease, i.e. death) was chosen because it represents an 80% decrease in the linear portion of the dose response curve and represents a dramatic change in respiratory function. The ED40s were 2.1, 3.9, 11.1 and 4.4 gg/kg in the BALB, C57, BK, and BH mice, respectively (Table 1).

Respiration

Figure 4, panels A and B, illustrate lever presses and drug intake (gg/kg) per session, respectively, for C57, BALB, BK and BH mice as a function of increasing etonitazene concentration at F R 8. C57, BALB, BK and BH mice all drank etonitazene concentrations up to 1 gg/ml. There was a significant effect of concentration [F(Concentration) = 8.5, df = 3,104

Figure 3 shows the effects of etonitazene on respiration rate in the C57, BALB, BK and BH mice. There were significant effects of genotype, dose and a genotype x dose interaction on etonitazene-induced changes in respiration rate: [F(Genotype) = 4.5; df = 3,98; P < 0.00541, [F(Dose) = 56.4; df = 4,98; P < 0.0001], [F(Genotype x Dose) = 4.5; df = 12,98; P < 0.0001]. There was no significant effects of gender in the BH or BK mice. The BK mice were less sensitive to the respiratory depressant effects of etonitazene and the only strain to show respiratory stimulation at low doses. The rank order sensitivity to the respiratory depressant effects of etonitazene was BALB > C57 > BH > BK. Doses greater than 30gg/kg did not produce

Operant self-administration

Training All strains were successfully trained to stable rates of lever pressing behavior on an F R 8 schedule of water reinfbrcement. There were no significant differences in rate of responding across increasing FR's as a function of genotype. Rates of lever pressing at F R 8 were 3.7, 3.8, 4.0 and 4.2 responses/s in the BH, B A L K C57 and BK mice, respectively.

Post-prandial exposure to concentrations of 0.25-1.0 #g/kg etonitazene

2500

==2 0 0 0 ¢~ t 5 0 0 >~ 1000

50O

0

o Pr,

120

, - , ~ - - r

,

,,,

,

100

---T

o-,

,

,

I

0.4

'

"

I

0.6

"

~

-

0.8

-

'

°

~

1

~

~=~== 20

o

' s

8O o

0.2

~o~° ~i~.i.~.°.....lii ..............

100

60 40

o 0

0

20

0.2

0.4

0,6

0.8

1

Etonitazene ( l i g / m l ) 1

Etonitazene

10

100

(lig/kg)

Fig. 3 Respirationrate as a functionof increasingetonitazenedose in BK, BALB,C57 and BH mice. Each point represents the mean respiration rate taken at 5-min intervals over a 1-h period. Each point represents the mean (+_SEM) from fiveor six mice per group

Fig. 4 Lever pressing and drug intake as a function of increasing etonitazene concentrationin BK, BALB,C57 and BH mice during post-prandial exposure to etonitazene. Panels A and B illustrate lever presses and drug intake (gg/kg), respectively,at FR 8. Each point represents the conditionmean (_+SEM) of results from 8 BK, BALK C57 and BH mice over a minimum of four consecutive sessions

28 P < 0.001] and a significant genotype x concentration interaction [F(Genotype x Concentration) = 5.8, df = 12,104 P < 0.001] for lever presses. C57 mice showed a slight U-shaped curve as a function of increasing concentration whereas BH mice showed a slight inverted U-shaped curve. Within each strain, no significant decrease was seen between the number of lever presses for water solution and etonitazene (1 gg/ml). Thus, the absolute amount of etonitazene consumed (gg/kg) increased as the concentration of etonitazene was increased. There was a significant difference between strains for drug intake as a function of increasing etonitazene concentration [F(Genotype) = 3.4, df= 3,78, P < 0.04], BALB mice consumed less etonitazene at all concentrations. At 1 gg/ml, drug intake was 71.9, 83.5, 85.0 and 94.2 gg/kg for BALB, C57, BH and BK mice, respectively. At the higher etonitazene concentrations the C57 mice demonstrated maximal signs of straub tail. In all strains the temporal distribution of responding within the session was similar under post-prandial drinking conditions, greater than 80% of responding occurred within the first 15 min of the session.

250
200

P o.

150

>~ .J

100

50

Behavior maintained by 1.0 #g/ml etonitazene without prior food presentation Figure 5 shows the mean number of lever presses during the last five sessions for BALB, C57, BK and BH mice at F R 8 without prior food presentation when drug (1 gg/ml) or vehicle (water) was available (panel A). Each drug condition (drug, vehicle and drug) was run for 15 sessions in order to obtain stable rates of responding and to obtain reliable extinction in all four strains. There was a significant effect of drug on the number of lever presses per session [F(Drug) -- 24.6, df = 2,78, P < 0.001]. Drug maintained significantly greater amounts of behavior than vehicle for each strain. There was a large effect of genotype on drug intake (panel B) [F(Genotype) = 26.8, df = 3,78, P < 0.001]. Drug intake exceeded that of the vehicle for each strain. Rates of extinction during vehicle substitution differed across strain (Fig. 6). In the BH mice a significant increase in behavior during the first 3 days of vehicle substitution (up to 275% of drug responding) occurred followed by a gradual decrease to 36 % of drug responding. Lever pressing behavior remained stable for the first 2-3 days in C57 mice fbllowed by a gradual reduction to 46% of drug responding. The BALB mice showed the most rapid decline in lever press behavior following vehicle substitution but were much more variable during the course of the extinction process. In the BK mice lever press behavior dropped to 65% of drug responding the first day and decreased rapidly across sessions to 28 % of drug responding by the last 5 days of vehicle substitution. Genetic relationships across phenotypes

BH

C57

BALB

BK

12

In order to begin investigating the genetic relationships among opioid-induced and drug naive phenotypes, genetic correlations across phenotypes were performed

10 v,

<:--o~

4

.=-= 250

==

""

2O0

2

.........

'~-~

.......

....

f \

~:=

B~Ea

~-

C57

0 BH

C57

BALB

BK

•~ 100

Strain Fig. 5 Lever pressing and drug intake in BK, BALB, C57 and BH

mice respondingat FR 8. Panel A illustrateslever presses at 1 pg/ml versus vehicle in the absence of prior food presentation. Panel B illustrates drug intake (gg/kg) at 1 gg/ml in the absence of prior food presentation. Eachbarrepresents the conditionmean(+ SEM) from eight BK, C57 and BH mice and five BALB mice over the last 5 days at each condition. Each point represents the condition mean (_+SEM) of results from eight BK, C57 and BH mice and five BALB mice during the last five consecutive sessions of drug and vehicleconditions

vehicle"~ substitution

4

8

12

16

Session

Fig. 6 Extinction behavior in BK, BALB, C57 and BH mice responding at FR 8 during vehicle substitution. Lever pressing is presented as a percentage of previous drug-reinforcedresponding. Each point represents the daily sessionmean of eight BK, C57 and BH mice and five BALB mice. SEM for each data point are less than 18%

29 on the complete matrix of phenotypes obtained in the current study (drug intake, baseline activity, stimulant ED50, stimulant efficacy, analgesic ED50, respiratory depressant ED50) along with whole brain binding data ([3Hldihydromorphine (DHM) and [3H](D-Ala2,DLeu5) enkephalin) obtained from Reith et al. 1981. The only two behavioral phenotypes that a significant genetic correlation was found were drug intake and baseline activity. (r = +0.94, P < 0.05). Of the in vitro correlations, there was a significant genetic correlation between [3H]DHM and analgesic and respiratory depressant ED50s; r = -0.94, P < 0.05 and r = -0.99, P < 0.05, respectively. The statistical power of the study is restricted by the N and limits the generality of these data to the study population (C57, BALB, BH, BH) and the methods used to assess each phenotype until further replication in additional genotypes or study populations are completed.

Discussion

The current study is the first to describe genetic differences in operant oral self-administration of an opioid in mice. Four inbred strains were trained with specific procedures designed to carefully expose naive subjects to the pharmacological effect of the drug prior to tests of operant drug-reinforced behavior (Meisch 1975). In the absence of food-induced drinking conditions, etonitazene maintained high rates of responding in the g-deficient BK mice, intermediate rates in BALB and C57 mice and relatively low rates of responding in the opiate-rich BH mice. Importantly, etonitazene served as a reinforcer in all four strains; rates of lever pressing were significantly less during the vehicle condition than drug condition. Etonitazene maintained rates and patterns of responding in these inbred mice similar to those maintained by etonitazene in other species under similar conditions (Meisch and Stark 1977; Carroll and Meisch 1978; Carroll et al. 1981). Extinction rates differed significantly as a function of genotype, BH mice demonstrated a significant extinction burst not seen in the other three strains. The differences in extinction rate were not directly related to differences seen in the other opioid-related phenotypes (including binding). The differences seen in extinction rate may be a function of differential conditioning effects related to stimuli associated with drug delivery (Katz and Goldberg 1987), the particular maintenance dose used in these experiments (Woods and Schuster 1971; Grant and Johanson 1987) or could possibly represent differences in the magnitude of the reinforcer (Mackintosh 1974). Context-specific tolerance to etonitazene-induced analgesia only in the BH mice may support the first hypothesis (Elmer et al. 1993); however, further parametric studies are required.

Specific genotypes with known variations in opiate receptor concentration, BK, BALB, C57 and BH, were chosen to investigate the role of opiate receptor subtypes in the acquisition of opioid-reinforced behavior. All four strains acquired opioid-reinforced behavior despite large differences in opiate receptor concentration and the ability of this paradigm to demonstrate qualitative differences in drug-reinforced behavior (Elmer et al. 1987a,b, 1990; Suzuki et al. 1992). Importantly, etonitazene served as a reinforcer in BK mice; a mouse strain with a significant deficit in CNS g-receptor concentration and relatively insensitive to the stimulant, analgesic and respiratory depressant effect of etonitazene. Thus, #-receptor concentration may not account for genetic variability in the acquisition of opioid-reinforced behavior under these training procedures. The fact that all strains acquired opioidreinforced behavior diminishes the role that whole brain opiate binding plays in the acquisition phase using these methods. Specific regional variation in #-receptor population or in strain specific opioid interactions with other neurotransmitters may further elucidate the observed differences seen in the maintenance and extinction of opioid-reinforced behavior. The C57BL subline C57BL/ByJ and the BK mice are the only strains in the current studies for which regional binding data are available (Moscowitz and Goodman 1985). These two strains demonstrate significant differences in g-receptor concentration in areas thought to mediate many of the observed differences tbund in the stimulant, analgesic and reinforcing effects of opioids (striatal, periaquaductal gray and nucleus accumbens). Autoradiography techniques wiU enable finer resolution of region-specific opiate concentrations in opioidrelated phenotypes (see Belknap et al. 1991). The results presented in the current study as well as others (Woods et al. 1981; Ling et al. 1985; Belknap 1989), support the independent inheritance of numerous opioid-related phenotypes. Since each phenotypic component of the drug response is independently affected by genotype, individual genetic variability may put some individuals at risk for acute or chronic druginduced toxicity's during drug self-administration (Devor et al. 1988). For example, the BALB mice are significantly more sensitive to the respiratory depressant effects of etonitazene and the immunosuppressant effects of opioids (current data; Taub et al. 1991; Eisenstein et al. 1990). Drug intake, however, was greater in the BALB mice than C57 and BH mice. As a potential consequence, the majority of the BALB mice (6/8) succumbed to infectious disease following completion of these experiments. These studies suggest genotype to be an important factor in determining individuals at risk for the acute toxic effects of opioids during drug-taking behavior. Recent hypotheses suggest spontaneous locomotor activity in response to a novel environment and

30

drug-induced locomotor activity as possible phenotypes characterizing predisposing factors in the acquisition of drug-reintbrced behavior (Wise and Bozarth 1987; Piazza et al. 1989). However, all strains acquired drug-reinforced behavior in the current study despite significant differences in spontaneous locomotor activity and the potency and efficacy of etonitazene-stimulated activity. Interestingly, drug intake during the maintenance of etonitazene-reinforced behavior was significantly correlated with spontaneous locomotor activity. These data lend support to the potential relationship between spontaneous locomotor behavior and drug-intake during self-administration sessions but do not address the relative reinforcement efficacy of opioids as a function of genotype. In addition to demonstrating etonitazene-reinforced behavior in all four strains, these studies demonstrate a third drug class for which C57 mice will respond under fixed ratio schedules of reinforcement (ethanol: Elmer et al. 1987; cocaine: George et al. 1990) and possibly represent a departure from previous reports of a link between the intake of ethanol and opioids (Nichols and Hsiao 1967; Hyytiaa and Sinclair 1993). The current data indicate that for BALB mice operant oral self-administration of etonitazene was somewhat greater than for the water vehicle. However, these mice showed the lowest rates of responding during the postprandial drinking phase and showed the smallest difference between rates of responding for drug versus rates of responding for vehicle. Further studies using a broad range of conditions (i.e. schedule of reintbrcement, dose, route of administration) will be necessary to confirm genetic independence/covariance between ethanol- and opioid-reinforced behavior. The use of inbred strains with specific biochemical alterations was important for directly testing the hypothesis that #-receptors mediate the reinforcing properties of opioids. The results of these studies indicate that #-receptor populations may not account for individual variability in the acquisition of opioid-reinforced behavior. The relationship between drug intake and /~-receptor population awaits further verification via receptor autoradiography techniques. In general, the current series of studies using a behavioral genetics approach suggest independent inheritance of severN opioid-related phenotypes. This approach may lead to techniques designed to characterize individuals at risk for the acute toxic effects of opiates during drug self-administration and provide valuable background information that may help to separate the therapeutic value of opioids from unwanted side-effects such as respiratory depression and addiction liability. Acknowledgements The authors would like to thank Carla Highkin and Laurica Graham for secretarial support. This work was supported in part by NIAAA grants AA-07754 and AA-09549 to F.R.G. The animals used in this study were maintained in facilities fully accredited by the American Association for the Accreditation

of Laboratory Animal Care (AAALAC) and the studies were conducted in accordance with the Guide for Care and Use of Laboratory Animals provided by the NIH and adopted by NIDA.

References Bals-Kubic R, Shippenberg TS, Herz A (1990) Involvement of central p and 6 opioid receptors in mediating the reinforcing effects of/~-endorphin in the rat. Eur J Pharmacol 175 : 63-69 Baran A, Shuster L, Eleftheriou BE, Bailey DW (1975) Opiate receptors in mice: genetic differences. Life Sci 17 : 633-640 Beardsley PM, Meisch RA (1981) A precision drinking device for rats tested with water, etonitazene, and ethanol. PharmacoI Biochem Behav 14 : 871-876 Belknap JK, O'Toole LA (1992) Studies of genetic differences in response to opioid drugs. In: Harris RA and Crabbe JC (eds) The genetic basis of alcohol and drug actions. Plenum Press, New York, pp 225-252 Belknap JK, Noordewier B, Lame M (1989) Genetic dissociation of multiple morphine effects among C57BL/6J, DBA/2J and C3H/HeJ inbred mouse. Physiol Behav 46 : 69-74 Belknap JK, Lame M, Danielson PW (1990) Inbred strain differences in morphine-induced analgesia with the hot plate assay: a reassesment. Behav Genet 20 : 333-338 Belknap JK, Laursen SE, Sampson KE, Wilkie A (1991) Where are the # receptors that mediate opioid analgesia? An autoradiographic study in the HAR and LAR selection lines. J Addict Dis 10 : 29~t4 Bertalmio A, Woods JH (1989) Reinforcing effect of alfentanil is mediated by # opioid receptors: apparent pA2 analysis. J Pharmacol Exp Ther 25 t : 455M60 Carroll ME, Meisch RA (1978) Etonitazene as a reinforcer: Oral intake of etonitazene by rhesus monkeys. Psychopharmacology 59 : 225-229 Carroll ME, France CP, Meisch RA (1981) Intravenous self- administration of etonitazene, cocaine and phencyclidine in rats during food deprivation and satiation. J Pharmacol. Exp Ther 214 : 339-346 Crabbe JC, Phillips TJ, Kosobud A, Belknap JK (1990) Estimation of genetic correlation: Interpretation of experiments using selectively bred and inbred animals. Alcohol Clin Exp Res 14 : 141-151 Cunningham CL, Niehus DR, Malott DH, Prather LK (1992) Genetic differences in the rewarding and activating effects of morphine and ethanol. Psychopharmacology 107 : 385-393 Devor EJ, Reich T, Cloninger CR (1988) Genetics of alcoholism and related end-organ damage. Semin Liver Dis 8 : I-11 Elmer GI, Meisch RA, George FR (t987a) Mouse strain differences in operant self-administration of ethanol. Behav Genet 17 : 439451 Elmer GI, Meisch RA, George FR (1987b) Differential concentration-response curves for oral ethanol self-administration in C57BL/6J and BALB/cJ mice. Alcohol 4 : 63~8 Elmer GI, Meisch RA, Goldberg SR, George FR (1988) A fixed ratio analysis of oral ethanol reinforced behavior in inbred mouse strains. Psychopharmacology 96:431436 Elmer GI, Meisch RA, Goldberg SR, George FR (t990) Ethanol self-administrationin Long Sleep and Short Sleep mice: evidence for genetic independence of neurosensitivity and reinforcement. J Pharmacol Exp Ther 254:1054-1062 Elmer GI, Mathura CB, Goldberg SR (1993) Genetic factors in conditioned tolerance to the analgesic effects of etonitazene. Pharmacol Biochem Behav 45:251-254 Ferster CB, Skinner BF (t957) Schedules of reinforcement. Appleton-Century-Crofts, New York Franklin KBJ, Abbot FV (1989) Techniques for assessing the effects of drugs on nociceptive responses. In: Boulton AA, Baker GB,

31 Greenshaw AJ (eds) Neuromethods. Humana Press, Clifton, New Jersey, pp 145--216 Funada M, Narita M, Hayashi Y, Suzuki T, Misawa M, Nagase H (1991) Reinforcing property of morphine in the CXBK mouse: characteristics of # opioid receptor subtypes. In: Van Ree JM, Mulder AH, Wiegant VM, Van Wimersma-Greidanus TB (eds) New leads in opioid research, pp !30.-131 George FR, Elmer GI, Meisch RA, Goldberg SR (199t) Orally delivered cocaine functions as a positive reinforcer in C57BL/6J mice. Pharmacol Biochem Behav 38 : 897 903 George FR, Goldberg SR (t989) Genetic approaches to the analysis of addiction processes. Trends Pharmacol Sci 10 : 73-83 George FR, Ritz MR, Elmer GI (1991) The role of genetics in drug dependence. In: Pratt J (ed) The biological basis of drug tolerance and dependence. Academic Press, New York, pp 265-290 Goldberg SR, Gardner ML (1981) Second-order schedules: extended sequences of behavior controlled by brief environmental stimuli associated with drug self-administration. In: Thompson T, Johanson CE (eds) NIDA Research Monograph 37. US Govt PrintingOffice, Washington, D.C., pp 241~70 Grant KA, Johanson CE (1987) Diazepam self-administration and resistance to extinction. Pharmacol Biochem Behav 28 : 81-86 Horowitz GP, Whimey G, Smith JC, Stephan F K (1977) Morphine ingestion: genetic control in mice. Psychopharmacology 52:119 122 Howerton TC, O'Connor ME, Collins AC (1983) Lipid solubility is correlated with hypnotic and hypothermic responses of LongSleep (LS) and Short-Sleep (SS) mice to various depressant drugs. J Pharmacol Exp Ther 227 : 38%393 Hyytiaa P, Sinclair JD (1993) Oral etonitazene and cocaine consumption by AA, ANA and Wistar rats. Psychopharmacology 111 : 409-414 Jacob J, Michaud G, Nicola M-A, Prudhomme N (1983) Genetic differences in opioid binding and in antinociceptive activities of morphine and ethylketocyclazocine. Life Sci 33 : 645-648 Jiang Q, Mosberg HI, Porreca F (1990) Modulation of the potency and efficacy of mu-mediated antinociception by delta agonists in the mouse. J Pharmacol Exp Ther 254 : 683-689 Ling GSF, Spiegel K, Lockhart SH, Pasternak GW (1985) Separation of opioid analgesia frmn respiratory depression: evidence for different receptor mechanisms. J Pharmacol Exp Ther 232 : 149-155 Mackintosh NJ (1974) The psychology of animal learning. Academic Press, New York Marek P, Yirmiya R, Liebeskind JC (1988) Strain differences in the magnitude of swimming-induced analgesia in mice correlate with brain opiate receptor concentration. Brain Res 447:188 190 Marek P, Yirmiya R, Liebeskind JC (1990) Genetic influences on brain stimulation-produced analgesia in mice: II. Correlation with brain opiate receptor concentration. Brain Res 507:155-157 Meisch RA (1975) The function of schedule-induced polydipsia in establishing ethanol as a positive reinforcer. Pharmacol Rev 27 : 465-473 Miczek KA, Thompson ML, Shuster L (1982) Opioid-like analgesia in defeated mice. Science 215:1520-1522 Miner LL, Marks MJ, Collins AC (1986) Genetic analysis of nicotine-induced seizures and hippocampal nicotinic receptors in the mouse. J Pharmacol Exp Ther 239 : 853-860 Moskowitz AS, Goodman RR (1985) Autoradiographic analysis of #~, #2, and delta opioid binding in the central nervous system of C57BL/By and CXBK (opioid receptor-deficient) mice. Brain Res 360:108-116 Moskowitz AS, Terman GW, Carter KR, Morgan JM, Liebeskind JC (1985) Analgesic, locomotor and lethal effects of morphine in the mouse: strain comparisons. Brain Res 361 : 46-51 Muraki T, Kato R (1986) Strain differences in the effects of morphine on the rectal temperature and respiratory rate in male mice. Psychopharmacology 89 : 60-64

Negus SS, Henriksen SJ, Mattox R, Pasternack GW, Portoghese PS, Takemori AE, Weinger MB, Koob GS (1993) Effects of antagonists selective for # and ~ on the reinforceing effects of heroin in the rat. J Pharmacol Exp Ther 265 : 1245-1252 Nichols JR, Hsiao S (1967) Addiction liability of albino rats: breeding for quantitative differences in morphine drinking. Science 157:561 563 Oliverio A, Castellano C (1974) Genotype-dependent sensitivity and tolerance to morphine and heroin: dissociation between opiateinduced running and analgesia in the mouse Psychopharmacologia 39 : 13-22 Oliverio A, Castellano C (1981) Behavioral effects of opiates: a pharmacogenetic analysis. Curr Dev Psychopharmacol 6 : 45-64 Piazza PV, Deminiere J-M, Le Moal M, Simon H (1989) Factors that predict individual vulnerability to amphetamine self-administration. Science 245 : 1511-1513 Pick CG, Nejat R J, Pasternak GW (1993) Independent expression of two pharmacologically distinct supraspinal mu analgesic systems in genetically different mouse strains. J Pharmacol Exp Ther 265 : 166-17I Raffa RB, Mathiasen JR, Jacoby HI (1987) Colonic bead expulsion time in normal and #-opioid receptor deficient (CXBK) mice folloing central (ICV) administration of #- and c~-opioid agonists. Life Sci 41 : 2229-2234 Reith MEA, Sershen H, Vadasz C, Lajtha A (1981) Strain differences in opiate receptors in mouse brain. Eur J Pharmacol 74 : 377 380 Shippenburg TS, Bals-Kubik R, tterz A (1987) Motivational properties of opioids: evidence that an activiation of &receptors mediates reinforcement processes. Brain Res 436 : 234-239 Spuhler K, Hoffer B, Weiner N, Palmer MR (1983) Evidence for genetic correlation of hypnotic effects and cerebettar Purkinje neuron depression in response to ethanol in mice. Pharmacol Biochem Behav 17:1073-1076 Suzuki T, George FR, Meisch RA (1988) Differential establishment and maintenance of oral ethanol reinforced behavior in Lewis and Fischer 344 inbred rat strains. J Pharmacol Exp Ther 245 : 164 170 Suzuki T, George FR, Meisch RA (1992) Etonitazene delivered orally serves as a reinforcer for Lewis but not Fischer 344 rats. Pharmacol Biochem Behav 42 : 579-586 Taub DD, Eisentsein TK, Geller EB, Adler MW, Rogers TJ (1991) Immunomodulatory activity of # and K-selective opioid agonists. Proc Natl Acad Sci 88 : 360-364 Tortella FC, Moreton JE (1980) D-Ala2-Methionine-enkephalinamide self-administration in the morphine-dependent rat. Psychopharmacology 69 : 143-147 Vaught JL, Mathiasen JR, Raffa RB (1988) Examination of the involvement of supr~spinal and spinal # and 6 opioid receptors in analgesia using the g receptor deficient CXBK mouse. J Pharmacol Exp Ther 245:13-16 Wise RA, Bozarth MA (1987) A psychomotor stimulant theory of addiction. Psychol Rev 94 : 469-492 Woods JH, Schuster CR (1971) Opiates as reinforcing stimuli. In: Thompson T, Pickens RW (eds) Stimulus properties of drugs. Appleton-Century-Crofts, New York, pp 163--175 Woods JH, Katz JL, Young AM, Medzihradsky F, Smith CB (1981) Correlations among certain behavioral, physiological, and biochemical effects of narcotic agonists. In: Harris LS (ed) Problems of drug dependence. DHHS Publication, Washington, D.C., pp 43 57 Young AM, Swain HH, Woods JH (1981) Comparison of opioid agonists in maintaining responding and in suppressing morphine withdrawal in rhesus monkeys. Psychopharmacology 74 : 329-335 Young AM, Woods JH, tterling S, Hein DW (1983) Comparison of the reinforcing and discriminative stimulus properties of opioids and opioid peptides. In: Smith JE, Lane JE (eds) The neurobiology of opiate reward processes, pp 147-174