INTRODUCTION

Drug addiction is a chronic disorder characterized by compulsive drug taking at the expense of other behaviors (Jaffe, 1990). Intense craving for the drug is a fundamental feature of addiction and is thought to trigger drug use and relapse. Environmental cues associated with drug taking behavior are potent stimulators of opiate craving in humans (Childress et al, 1986; Sell et al, 1999). A better understanding of brain mechanisms involved in producing drug craving would lead to better treatments for drug addiction.

Human brain imaging studies have shown that drug-associated cues strongly activate the anterior cingulate cortex (Cg), the amygdala and other basal forebrain regions (Childress et al, 1999; Sell et al, 1999). Environments associated with drug taking can strongly motivate drug-seeking behavior in human addicts (Childress et al, 1986; Wallace, 1989) and the conditioned place preference (CPP) paradigm is an animal model of this type of cue-elicited conditioning (Bardo et al, 1995; Tzschentke, 1998). In the CPP paradigm, drug preference is measured, in a drug-free state, by the amount of time an animal spends in an environment that has been previously associated with drug administration. A recent study by Schroeder et al (2000) reported that exposure to a morphine-paired environment significantly increased Fos expression in the prefrontal and cingulate cortices, nucleus accumbens (NAc) and the preoptic area of nondependent rats, findings that are consistent with the human imaging results.

Recently, we reported that chronically morphine-pretreated animals abstinent for 2 weeks before CPP testing exhibited significantly greater preference for morphine-paired environments than placebo-pretreated animals (Harris and Aston-Jones, 2001). The present study sought to determine whether this increased preference after chronic morphine exposure is long-lasting, and which brain regions are involved in this enhanced preference. For the latter, we used Fos protein expression as a marker of neuronal stimulation (Sharp et al, 1993; Herrera and Robertson, 1996).

We focused our analysis of Fos induction on brain areas previously associated with conditioned or affective properties of opiates. The brain areas analyzed for this study were the anterior Cg, nucleus accumbens, bed nucleus of the stria terminalis (BNST), and amygdala. The Cg is an area consistently activated by drug cues in human cocaine and heroin addicts (Childress et al, 1999; Sell et al, 1999), and has been shown to be important for learning to associate environmental stimuli with cocaine administration in animals (Weissenborn et al, 1997). Although the basolateral amygdala is not necessary for the acquisition of cue-induced heroin seeking (Alderson et al, 2000), this area is important for cue-induced reinstatement of extinguished heroin-seeking behavior (Fuchs and See, 2002). Other studies found that the BNST is important for the reinforcing effects of opiates (Walker et al, 2000), while both the BNST and central nucleus of the amygdala (ACE) play roles in opiate withdrawal aversion (Delfs et al, 2000; Gracy et al, 2001). Neurons in the NAc appear to be critically involved in both the reinforcing properties of opiates (Dworkin et al, 1988; Zito et al, 1985) and opiate withdrawal responses (Harris and Aston-Jones, 1994). We also counted Fos-positive neurons in the periventricular nucleus of the thalamus (PVT) as a control region because it is interconnected with the NAc, ACE, ABL and BNST (Moga et al, 1995; Otake et al, 1995), but has not previously been linked to opiate abuse.

MATERIALS AND METHODS

Subjects

Male Sprague–Dawley rats (200–250 g) from Harlan (Indianapolis, IN) were used in both experiments. Rats were group housed in accordance with NIH guidelines on a 12-h light/dark cycle with food and water available ad libitum. All animal procedures were approved by the Institutional Animal Care and Use Committees of the Philadelphia Veterans Administration Medical Center and the University of Pennsylvania. A total of 26 animals were used for the experiment reported here, with individual group numbers of six or seven animals.

Drugs

Morphine pellets and morphine sulfate powder were provided by the National Institute on Drug Abuse. Morphine sulfate was dissolved in sterile saline, and was administered via intraperitonal injection. All vehicle injections consisted of sterile saline.

Chronic Drug Treatment

Morphine pretreatment was carried out in 13 rats by subcutaneously implanting two 75-mg morphine tablets under halothane anesthesia. Placebo-pretreated rats (n=13) were implanted at the same time with placebo pellets. In this report, chronically morphine-pretreated animals are denoted by the abbreviation ‘M’, and all placebo-pretreated animals are denoted by the abbreviation ‘P’. Previous studies have shown that morphine pellets are a reliable way to induce physical dependence (Gold et al, 1994; Yoburn et al, 1985). The signs of physical dependence wane by day 14 as the morphine pellets dissolve (Gold et al, 1994). Here, deprivation withdrawal, a model of abstinence, was induced by removing the pellets after 14 days.

Conditioned Place Preference Procedure

Training and testing occurred in a Plexiglas apparatus consisting of two distinct compartments (37×39×30 cm3) separated by a tunnel (31×11×19 cm3). One compartment had a grid floor with black walls, and the second compartment had a mesh floor with black and white stripes on the walls. Each compartment was equipped with photocells to automatically record time in each compartment (MED Associates, East Fairfield, VT). Rats were conditioned with acute injections of morphine in one of these environments beginning 4 weeks after pellet removal, with testing completed 5 weeks after the withdrawal from morphine. We choose to examine behavior and Fos expression after this period of abstinence, because by this time all morphine has been cleared from the body and all apparent withdrawal signs have dissipated. Also, it is a time period in which addicts trying to maintain abstinence are vulnerable to relapse. It has been estimated that 80% of addicted people will relapse by 1 month after treatment (Hunt et al, 1971; Vaillant, 1992). On the first day of the procedure (the preconditioning day), rats were allowed to freely explore all of the apparatus for 15 min, and the amount of time spent in each compartment was recorded. None of the animals had an initial bias for either compartment and rats were randomly assigned to one compartment for morphine conditioning in a balanced design. Conditioning began 5 days later. On each of the next 3 days, all to-be conditioned rats were injected with either saline or morphine (10 mg/kg i.p.) in the morning and afternoon. Rats were confined to one side of the box after each injection by means of an opaque Plexiglas divider for 30 min. The morphine and saline treatments were alternated in morning and afternoon sessions for every conditioned animal, so that rats given morphine in the morning were given saline in the opposite compartment in the afternoon, and on subsequent days received saline in the morning, and morphine in the afternoon. The morning and afternoon injections were at least 4 h apart. Five days after conditioning, a preference test was conducted, in which animals were given free access to the apparatus for 15 min. The amount of time spent in each compartment and the amount of activity were recorded. As a conditioning control, additional groups of P (n=7) and M animals (n=7) were given injections of saline in both compartments and morphine injections (10 mg/kg) in their home cages (1600 h).

Fos Protein Immunoreactivity

Two hours after the preference test animals were deeply anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and transcardially perfused with ice-cold 0.9% saline followed by ice-cold 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS), pH 7.4. The brains were removed and stored overnight in 4% paraformaldehyde. They were then transferred to a 20% sucrose solution and stored at 4°C for 5 days. Coronal sections (40 μm) were cut using a freezing microtome. Sections from P and M groups were processed together to equalize staining between groups. Sections were placed in a solution of 0.01 M PBS with 0.3% Triton-X added (PBS-Tx, pH 7.4) containing 2% normal donkey serum for 3 h. Sections were incubated overnight at room temperature in this same solution with the addition of primary antibody (rabbit antiserum against Fos-related antigens at 1:20,000, Oncogene Sciences, Cambridge, MA). Sections were rinsed three times in PBS-Tx and then incubated for 2 h with the secondary antibody (biotinylated donkey anti-rabbit 1:1000, Jackson Immunoresearch Laboratories, West Grove, PA). After three rinses in PBS-Tx, sections were transferred to an avidin–biotin complex (1:1000, Jackson Immunoresearch Laboratories) for 1.5 h. Sections were again rinsed two times with PBS-Tx and once with 0.05 M Tris buffer. Fos-related antigen-positive (denoted here as simply Fos) cells were visualized by placing the tissue in 3,3′-diaminobenzidine (DAB, 0.02%, Sigma, St Louis, MO) with 0.0002% H2O2 and 0.6% nickel ammonium sulfate in 0.01 M Tris buffer for 3.5 min. This reaction was arrested by immediate transfer into 0.05 M Tris buffer. The sections were mounted on gelatin-coated slides, stained with neutral red to identify specific structures, dehydrated through graded alcohols, cleared in xylene, and coverslipped with Permount.

Quantification of Fos-positive cells was done using Openlab image processing software (Improvision, Ltd.; Coventry England) on a Macintosh computer that was linked to a microscope and digital camera. Color images of the areas of interest were taken and saved to a disk. The numbers of Fos-positive nuclei in regions of interest were counted with a point counter tool on the saved image. This tool simultaneously marked and counted each cell so that no cells could be counted twice and the total of the number of cells counted was available. One section at each level was randomly selected from each animal. The levels chosen corresponded to the following distances from bregma (Paxinos and Watson, 1998); anterior Cg (Cg, +2.20 mm), accumbens shell and core (Ac-S, Ac-C, +1.70 mm), ventral and dorsal lateral bed nucleus of the stria terminalis (BNST-VL, BNST-DL, −0.26 mm), amygdala central and basolateral nuclei (ACE, ABL −2.80 mm) and the periventricular nucleus of the thalamus (PVT −2.80 mm; Figure 1). Both the right and left sides were counted and averaged into a single score.

Figure 1
figure 1

Schematic representations of regions analyzed for Fos-positive protein expression. Gray shaded regions indicate the areas where cells were counted. Numbers in the lower left-hand corner of each brain section represent the distance from bregma. Cg1 refers to anterior Cg area 1. Drawings were adapted from Paxinos and Watson (1998).

Data Analysis

Place conditioning data were analyzed by calculating the time spent in the morphine-paired chamber minus the time spent in the saline-paired chamber to determine chamber preference. The resulting difference score was compared between groups. In addition, a within-group measurement of conditioned place preference was assessed by comparing the difference in time spent in the morphine- and saline-paired side preconditioning vs postconditioning. Behavioral and Fos data were analyzed using a one-way or two-way analysis of variance. Where necessary, post hoc analysis was performed with a Newman–Keuls test.

RESULTS

Place Conditioning

As expected, our paradigm produced a significant preference for the morphine-paired environment in all conditioned rats (p<0.01; F(1,10)=73.14 for P rats, and F(1,10)=101.65 for M animals). Notably however, the preference for the morphine environment was greatly increased by morphine pretreatment 5 weeks prior to CPP testing. A two-way ANOVA (drug treatment×conditioning) revealed a significant main effect for drug treatment (F(1,22)=16.63, p<0.006), conditioning (F(1,22)=184.06, p<0.006), and an interaction between drug treatment and conditioning (F(1,22)=23.51, p<0.006). Follow-up tests revealed that all conditioned groups were significantly different from nonconditioned groups (p<0.01), and that conditioned M rats showed enhanced preference for the morphine-paired environment relative to conditioned P rats (p<0.01, Figure 2). The nonconditioned P and M groups given saline injections paired with both compartments, plus morphine in the home cage, did not show significant place preferences (p=0.12; F(1,24)=2.4 for P and M animals combined).

Figure 2
figure 2

Preference scores for the morphine-paired environment expressed as the mean time in seconds spent in the morphine-paired side minus the mean time in seconds spent in the saline-paired side on the test day. *Significantly different (p<0.01) from non-conditioned groups; +significantly different (p<0.01) from conditioned P groups.

Fos Measurements

Figure 3 and 4 show the results for Fos immunoreactivity in the various brain regions. The results of two-way ANOVAS (similar to those performed for preference data above) on each brain area are presented in Table 1. All brain areas, except the PVT, showed a significant main effect for conditioning. Follow-up tests indicated that the conditioned P and M animals showed significantly more Fos than nonconditioned animals (p<0.01). Furthermore, the conditioned M group showed significantly higher Fos levels than either the conditioned or nonconditioned P groups in the Cg, Ac-C, BNST-VL, ACE and ABL (p<0.01 for each area, Figure 3 and Figure 4). The conditioned P group exhibited significantly greater Fos than the nonconditioned groups in the CG, Ac-S, Ac-C BNST-VL, ACE, ABL, and significantly greater Fos levels than all groups (including the conditioned M group) in the BNST-DL (p<0.01). No significant changes were seen in the PVT of any group.

Figure 3
figure 3

Photomicrographs of frontal sections showing Fos expression from representative sections from conditioned M, conditioned P and nonconditioned M groups. Sections compare the Cg, the nucleus accumbens (Ac) (C=core, Sh=shell), the dorsal lateral (BST-DL) and ventral lateral (BST-VL) BNST, and the central (CE) and basolateral (BL) amygdala. Sections are counter-stained with neutral red. Dashed lines in the M nonconditioned pictures delineate the approximate area of each structure. For all sections dorsal is up, for the Cg, accumbens, and amygdala sections medial is right and for the BNST sections medial is left.

Figure 4
figure 4

Number of Fos-positive nuclei (±SEM) counted in the Cg, Ac-S, Ac-C, BST-DL, BST-VL, ACE, ABL, PVT. *Significantly different (p<0.01) from nonconditioned groups; -significantly different (p<0.05) from conditioned M groups.

Table 1 Two-way ANOVA Results

Correlations Between Preference Scores and Fos Expression

To determine the degree of association between the preference for the morphine environment and Fos induction, we calculated the Pearsons product-moment correlation coefficient (r) between the number of cells expressing Fos and the degree of preference expressed for each brain area in each group. Significant positive correlations (r values >0.72, df=4) were found in the Cg and ABL for the conditioned P group (p<0.05 each; Table 2). In the conditioned M group, significant positive correlations between preference and Fos staining were found in the Cg, ABL and BNST-VL (r values >0.81, p<0.02 each). Nonconditioned animals had no significant correlations.

Table 2 Correlation results

DISCUSSION

This study found that M animals show significantly greater preferences for morphine-associated environments when tested 5 weeks after withdrawal compared to P animals. Conditioned M rats also showed significantly greater Fos expression in a number of limbic and cortical areas relative to conditioned P rats. Conditioned P rats showed significantly greater Fos expression in most of the same brain areas as the conditioned M group when compared to nonconditioned rats. Significant positive correlations were found between the levels of behavioral preference in conditioned groups and Fos staining in the Cg and ABL. Fos in the BNST-VL was also significantly correlated with preference only in the conditioned M groups.

The increased preference in the conditioned M animals was found at 5 weeks postwithdrawal and replicates our previous findings for animals tested at 2 weeks after withdrawal (Harris and Aston-Jones, 2001). It is noteworthy that the degree of preference was similar in these two studies, indicating that enhanced preference had not decreased by 5 weeks postwithdrawal and that it is a long-lasting effect. Stress-induced reinstatement of heroin self-administration also follows a similar time course, becoming greater at 6 and 12 days postwithdrawal (Shavlev et al, 2001).

Others have also reported that preference scores increase following prolonged exposure to morphine (Lett, 1989; Contarino et al, 1997). However, those studies did not examine preference following prolonged drug abstinence. In addition, unlike our study in which a continuous-release morphine pellet was used, the prior reports used a daily injection regimen that is known to produce substantial locomotor sensitization (Vanderschuren et al, 1997). It is interesting that both modes of morphine administration produced similar results for enhanced preference following prior morphine administration. This indicates that withdrawal from opiates may be an important factor in generating enhanced preference to opiates, whether it occurs repeatedly as in the case of daily injections or just once following a continuous exposure to opiates.

Our results indicate that prior morphine exposure produces a prolonged increase in preference for morphine-associated stimuli, and furthermore, that prior withdrawal may increase the incentive value of opiates (see Hutcheson et al, 2001). Other experiments using the place preference paradigm have shown that morphine-pretreated animals develop a place preference at lower doses than nonmorphine-pretreated animals, suggesting sensitization to the rewarding properties (Shippenberg et al, 1996,1998). The enhanced preference seen in the M groups in our study may be because of sensitization to morphine's rewarding effects which is reflected by changes in neuronal excitability and subsequent Fos production. At 5 weeks postwithdrawal, place preference testing resulted in significantly greater Fos protein induction in the Cg, Ac-C, BNST-VL, ACE and ABL relative to conditioned P rats or nonconditioned M rats. In the Ac-S region, levels of Fos protein in conditioned M rats were significantly elevated relative to nonconditioned animals, but not in relation to conditioned P animals. Thus, conditioning, but not morphine exposure, may be an important factor in stimulation of these neurons.

Conditioned P animals showed significantly greater preference for the morphine-paired environment and significantly greater expression of Fos protein in the Cg, accumbens, BNST, and amygdala areas relative to nonconditioned rats. The conditioned P animals showed an elevation of Fos in the BNST-DL that was not found in the conditioned M animals. This suggests that the chronic morphine treatment may have affected activity in this region. In support of this idea, Walker et al (2000) reported that naloxone microinjections into this region of the BNST blocked heroin self-administration in dependent, but not nondependent, animals, indicating that responses in this area may be altered by the chronic morphine pretreatment.

Our results indicate that the increased preference and elevated Fos protein in the conditioned groups were the result of conditioning and not simply because of the morphine treatments. The nonconditioned animals given nonpaired exposures to both morphine and the conditioning chambers did not show a preference for either chamber or increased Fos expression on the test day. Furthermore, the elevations in Fos found in the conditioned groups cannot be explained by differences in locomotor activity because all groups showed similar activity counts on the test day.

Our results are consistent with earlier reports showing that drug-related cues elicit increased Fos expression in the brain (Brown et al, 1992; Crawford et al, 1995; Franklin and Druhan, 2000; Neisewander et al, 2000; Schroeder et al, 2000). Those studies found conditioned Fos induction in many of the same cortical and limbic brain areas as in the present study. These data suggest a common neuroanatomical pathway for the ability of cues associated with different addictive drugs to activate the brain. However, no previous study has examined the impact of prior dependence on subsequent conditioned Fos expression.

The changes we found in conditioned Fos protein expression in the anterior cingulate, nucleus accumbens, amygdala, and BNST suggest that these regions may play a role in the expression of morphine-conditioned behavior. Although most of these areas (Cg, accumbens, and amygdala) are not necessary for the acquisition of morphine-rewarded responding (Olmstead and Franklin, 1997a,1997b; Alderson et al, 2000), the finding that Fos expression in these regions was particularly increased in M animals indicates that these areas may be involved in the long-lasting effects of addiction on behavior, especially the expression of preference for drug-associated cues.

Conditioned M animals spent more time (an average of 2.5 min) in the morphine-paired environment than the conditioned P group. This difference may have contributed to the higher Fos levels seen in the M group. However, this explanation for the differences in Fos levels between the M and P groups is unlikely. In a recent study by Schroeder et al (2000), animals were confined to a previously morphine-paired environment prior to Fos measurements. Increased Fos expression was found in the prefrontal and anterior cingulate cortices, nucleus accumbens, and preoptic area. Unlike our study, however, they did not find conditioned Fos activation in the amygdala or BNST, and the Fos induction in the anterior cingulate was at a more posterior level than in our study. This indicates that the activation of these areas by our paradigm may be because of the fact that CPP elicits an operant approach response, and not because of the amount of time spent in the CS+ environment.

The amount of preference strongly correlated with Fos levels in the ABL and Cg. These two areas showed significant correlations in all conditioned groups. The BNST-VL was the one area where a significant positive correlation between Fos and behavior was found only in the M animals, indicating that this area may be particularly involved in the increased preference seen in the M animals.

Recent imaging studies in humans have paralleled the preclinical findings in animal studies. Cocaine- (Childress et al, 1999; Maas et al, 1998; Grant et al, 1996) or morphine-associated cues (Sell et al, 1999) that induced high levels of drug craving in drug-experienced subjects also only elicited brain activation in the amygdala and anterior Cg of these subjects. The involvement of the amygdala and Cg in response to conditioned drug cues in our study and others is consistent with the wealth of evidence supporting their role in both Pavlovian conditioning (Hatfield et al, 1996; Bussey et al, 1997) and cue-elicited drug-seeking behavior (Whitelaw et al, 1996; Meil and See, 1997; Weissenborn et al, 1997; Everitt et al, 2000). Although the amygdala may not be involved in the acquisition of morphine-conditioned responding (Alderson et al, 2000), it is involved in the reinstatement of extinguished opiate seeking (Fuchs and See, 2002). This latter finding is particularly relevant to our study.

Chronic morphine alters second messenger systems (Collier, 1980; Nestler et al, 1993; Rasmussen et al, 1990; Terwilliger et al, 1991) and expression of intermediate early genes like c-fos (Curran et al, 1996; Nye and Nestler, 1996; Erdtmann-Vourliotis et al, 1999; Frankel et al, 1999). Changes in gene expression could be involved in alterations within the CNS that lead to drug addiction. For example, Fos induction in the accumbens has been found to be necessary for the acquisition but not expression of morphine-induced CPP (Tolliver et al, 2000). Thus, Fos itself may not play a large role in the immediate expression of preference, but may instead be involved in associated plasticity, for example maintenance or extinction of the conditioning.

Genes like c-fos not only serve as markers of activation but also act as transcription factors that regulate the expression of other genes. Determining brain areas where gene expression is altered following drug dependence could reveal mechanisms for long-term changes in CNS function associated with addiction. Such insights would be valuable for designing new treatments for opiate addiction.