Volume 17, Issue 5 pp. 1067-1072
Full Access

Nucleus accumbens β-endorphin levels are not elevated by brain stimulation reward but do increase with extinction

Abraham Zangen

Abraham Zangen

Behavioural Neuroscience Branch, National Institute on Drug Abuse, National Institute of Health, Department of Health and Human Services, Baltimore, MD 21224, USA

Search for more papers by this author
Uri Shalev

Uri Shalev

Behavioural Neuroscience Branch, National Institute on Drug Abuse, National Institute of Health, Department of Health and Human Services, Baltimore, MD 21224, USA

Search for more papers by this author
First published: 14 March 2003
Citations: 28
: Dr A. Zangen, as above.
E-mail: [email protected]

Abstract

β-Endorphin is an endogenous opioid peptide implicated in reward processes, but the brain sites directly involved in its putative role in reward have not been identified. Here we used in vivo microdialysis in rats to study the effect of a potent reinforcer, lateral hypothalamus self-stimulation (LHSS), on the extracellular levels of β-endorphin in the nucleus accumbens (NAS). The NAS is involved in the reinforcing effects of natural and artificial rewards, has high density of opioid receptors and is innervated by arcuate nucleus β-endorphin neurons. LHSS had no effect on extracellular levels of β-endorphin in the NAS. Surprisingly, extinction of the self-stimulation behaviour induced a rapid increase in NAS β-endorphin levels. In a subsequent experiment in rats previously trained to self-administer heroin for 10 days, β-endorphin levels also were increased during a test for extinction of the heroin-reinforced behaviour. Finally, the increase in extracellular β-endorphin levels in the NAS was also observed during exposure to an aversive stimulus, intermittent footshock (20 min). These results indicate a possible role for increased levels of NAS β-endorphin in the organism's adaptive response to stress and frustration.

Introduction

Endogenous opioids have been implicated in brain reward processes, including those underlying drug reward and the development of addiction (Van Ree et al., 2000; De Vries & Shippenberg, 2002). β-Endorphin is an endogenous opiate that binds with high affinity to µ- and δ-opioid receptors, although its affinity to the κ-opioid receptor is lower (Akil et al., 1984). A role for β-endorphin in reward is suggested by demonstrations that ventricular infusions of β-endorphin have rewarding effects as measured in the conditioned place preference and the drug self-administration methods (Van Ree et al., 1979; Spanagel et al., 1991). The specific brain sites involved directly in rewarding effects of endogenous β-endorphin have not yet been identified, but there are indications for a role for the nucleus accumbens (NAS), a brain region critically involved in reward pathways of both natural and unnatural (e.g. drugs of abuse) reinforcers (Wise, 1996b). The cell bodies of β-endorphin-releasing neurons are located primarily in the arcuate nucleus (ArN) of the hypothalamus (Watson et al., 1978) and there is a strong reciprocal connection between the ArN and the NAS (Finley et al., 1981; Hu & Jin, 2000). Furthermore, relatively high concentrations of opioid receptors are found in the NAS (Mansour et al., 1995) and activation of these receptors is rewarding. Rats self-administer opioids directly into the NAS (Goeders et al., 1984) and microinjections of opioid agonists into the NAS facilitates electrical brain stimulation reward (Johnson et al., 1995). Finally, a role for NAS β-endorphin in drug reward has been suggested recently following the demonstration that acute administration of ethanol, cocaine and amphetamine, but not nicotine, increases extracellular levels of endorphins measured directly by microdialysis in the NAS (Olive et al., 2001). Moreover, cocaine self-administration also increases β-endorphin levels in the NAS (Roth-Deri et al., 2003).

Here we initially assessed whether the reported effects of psychostimulants and ethanol on extracellular levels of NAS β-endorphin would generalize to non-drug rewards. Electrical stimulation of the medial forebrain bundle (in regions such as the lateral hypothalamus) is strongly rewarding and is considered a highly productive approach in the study of the neuronal mechanisms underlying reward (Wise, 1996a). In previous studies, β-endorphin levels in hypothalamic and whole brain tissue were increased during lateral hypothalamus self-stimulation (LHSS; Stein, 1985) and systemic injections of high doses of naloxone (an opiate antagonist) and intracerebroventricular injections of antibodies to β-endorphin reduced responding for LHSS, indicating involvement of β-endorphin in self-stimulation reward (Schaefer, 1988; Carr, 1990). In this study we used a microdialysis procedure to measure the effect of LHSS on extracellular levels of β-endorphin in the NAS. Surprisingly, we found that LHSS did not change β-endorphin levels in the NAS, but extinction of lever pressing following termination of the electrical stimulus significantly increased the peptide levels. We subsequently assessed whether this effect would also be observed during extinction from heroin, a drug reinforcer. Finally, we tested the hypothesis that the observed increases in NAS β-endorphin levels are because of the frustration/stressful component of extinction training (Amsel, 1962) by measuring β-endorphin levels in the NAS during exposure to an intermittent footshock stressor.

Materials and methods

Animals

Male Long Evans rats (270–330 g, Charles River, Raleigh, NC, USA) were used. Rats were housed under a reversed 12-h light : 12-h dark cycle (lights off at 09.00 h) and tested during their dark phase. Food and water were available ad libitum. The experimental procedures followed the NIH Guide for the Care and Use of Laboratory Animals (1996).

Surgical procedures

Rats were implanted unilaterally, under anaesthesia (pentobarbital + chloral hydrate, 31 + 142 mg/kg), with a CMA guide cannula (Carnegie Medicine; Stockholm, Sweden) 2.2 mm in diameter directed at the NAS (A +1.4 mm; L +1.2 mm from bregma; and V 5.8 mm from skull; Paxinos & Watson, 1998). One group (n = 15) was also implanted with a monopolar stimulating electrodes (Plastics One, Roanoke, VA, USA) into the lateral hypothalamus (A −2.5 mm; L +1.7 mm from bregma; and V −8.5 mm from skull) with a skull screw serving as the current return. Other rats (n = 18) were implanted with intravenous catheters into the right jugular vein under anaesthesia as described previously (Shalev et al., 2000) during the intracranial cannula implantation surgery. Briefly, the rats were surgically implanted with intravenous (i.v.) Silastic catheters (Dow Corning, Midland, MI, USA) into the right jugular vein. The analgesic buprenorphine (0.01 mg/kg, s.c.) was given at the time of surgery. The catheter was secured to the vein with a silk suture and passed subcutaneously to the top of the skull where it was connected to a modified 22-gauge cannula (Plastics One). The cannula was mounted to the skull with jeweller screws and dental cement. After surgery, the catheters were flushed every 24–48 h with sterile saline (0.05 mL).

Apparatus

The experimental chambers (30 × 24 × 21 cm; MED Associates, Georgia, VT, USA) had two levers located 9 cm above the floor, but only one lever (an active, retractable lever) activated the infusion pump (Razel Sci., Stamford, CT, USA) or the stimulator (PHM/152B/2, MED Associates). Presses on the other lever (an inactive, stationary lever) were also recorded, but had no programmed result. The grid floors of the chambers were connected to electric shock generators (MED Associates).

Microdialysis

Microdialysis was performed as previously described (Zangen et al., 1999). A microdialysis probe (2 mm in length, 20 kDa cutoff value, CMA/10; Carnegie Medicine; Stockholm, Sweden) was inserted into the guide cannula to 7.8 mm from skull surface. Probes were continuously perfused with artificial cerebrospinal fluid (aCSF) containing (in mm): NaCl, 125; KCl, 2.5; NaH2PO4, 0.5; Na2HPO4, 5; MgCl2, 1; CaCl2, 1.2; and 0.05% (w/v) bovine serum albumin, pH 7.4. Rats were allowed to recover from probe implantation overnight, and in the following days the aCSF flow rate was set at 2.5 µL/min for all experiments. Dialysates were collected during 20-min intervals and immediately frozen until assayed for β-endorphin by an ELISA kit (Peninsula, Belmont, CA, USA) as described previously (Zangen et al., 1999). The concentrations of β-endorphin in the dialysates obtained here were within the linear segment of the standard curve. The intra-assay variation was 4% and the interassay variation was 13%. According to the manufacturer, the antiserum used in the assay cross-reacts 100% with rat β- and α-endorphin, 60% with γ-endorphin and 0% with met-enkephalin, adrenocorticotrophic hormone or α-melanocyte stimulating hormone.

Experiment 1: effect of LHSS on β-endorphin levels in the NAS

Training

Rats were trained for electrical self-stimulation over 5–7 days with two 20-min sessions/day. The rats were housed in the animal colony and brought daily to the experimental chambers. The training phase started after 2–3 ‘pretraining’ days, during which the parameters for brain stimulation were optimized individually for each rat as described previously (Shalev et al., 2000). During the subsequent 5–7 days, rats pressed the active lever for brain stimulation (monopolar cathodal pulses). Each lever press resulted in a 500-ms train of 0.1 ms, 100 Hz rectangular cathodal pulses at the current level that sustained reliable responding (400–800 µA); a cue light located above the lever was turned on for 1 s. Each session began 2 h after placing the rats in the training cages with the introduction of the active lever into the cage and the illumination of the white cue light above this lever for 30 s. After 20 min the active lever was retracted and 60 min later the second 20-min session began. Twelve out of the 15 rats were successfully trained for LHSS.

Microdialysis procedure

Following the last training session, the dialysis probe was inserted into the guide cannula approximately 15 h before the first test session. Baseline samples were taken during 100 min (five 20-min intervals) before the first 20-min test session. During this 15-h period, including during baseline samples collection, the rats were not exposed to reward or reward-related cues. In one group of rats (n = 6) the first test session was a regular brain stimulation reward session (similar to training) and in the second group of rats (n = 6) the first test session was an extinction session (i.e. the stimulator was turned off). During the second test session, that started 60 min after termination of the first session, test conditions were reversed. Forty minutes following the completion of the second test session, all the rats were exposed to 20-min noncontingent brain stimulation (one stimulation per second) using the training parameters for each rat, with the active lever retracted. At the end of this and subsequent experiments, rats were overdosed, an Evans Blue solution was injected through the microdialysis probe and brains were removed and soaked in 4% paraformaldehyde. The brains were then frozen, sliced into 40 µm coronal sections and examined under the microscope.

Experiment 2: effect of extinction of heroin self-administration behaviour on β-endorphin levels in the NAS

Training

Rats were trained for heroin self-administration over 10 days, for three 3-h sessions per day, separated by 3 h. The rats were housed chronically in the experimental chambers. A red houselight was turned on for the entire session. Each response on the active lever resulted in the delivery of 0.1 mg/kg heroin (diacetylmorphine HCl, NIDA, USA) and the initiation of a 20-s timeout period. During this period, lever presses were recorded, but not reinforced, and the cue light was turned on. At the end of each session, the houselight was turned off and the active lever was retracted.

Microdialysis procedure

Following the training phase, the rats were left undisturbed in their home cages for 3 days in order to allow testing under ‘heroin-free’ conditions, and avoid interference from physical withdrawal effects. The dialysis probe was inserted into the guide cannula approximately 15 h before the test. The test was conducted under conditions similar to training except that the drug syringes were removed from the pumps, i.e. extinction conditions. Baseline samples were taken for 3 h (nine 20-min intervals) before the test session and six additional samples were collected during the 2 h extinction session.

Experiment 3: effect of intermittent foot shock on β-endorphin levels in the NAS

A different group of naive rats (n = 6) were placed in the experimental chambers 3–4 days after surgery, and the dialysis probe was inserted into the guide cannula approximately 15 h before testing. Intermittent, inescapable, electric footshocks were delivered through the grid floor for 20 min (0.5 mA, 0.5 s ON; mean OFF period of 40 s). These shock parameters were based on previous reports (Shaham & Stewart, 1995). Dialysates were collected during 20-min intervals, starting with five baseline collections before exposure to the footshock, and over five additional collections following termination of the footshock session.

Statistical analysis

The dialysate levels of β-endorphin, for each rat were transformed into a percentage of an average basal level, assigning a value of 100% to the average β-endorphin level in the baseline samples collected before the test sessions. The analysis of experiment 1 was carried out first by a repeated measures anova, which included the baseline data points, the LHSS point and the post-session points, and then by a second repeated measures anova, which included the above data points, the extinction point and one additional post-session data point. The noncontingent data point was not included in these analyses. Experiment 2 was analysed using a repeated measures anova, which included four pre-session data points and the six data points collected during the session. Experiment 3 was analysed using a repeated measures anova, which included four pre-shock data points, the shock session data point and one post-session data point. anovas were followed by the Student–Newman–Keuls post hoc test.

Results

Experiment 1: effect of LHSS on β-endorphin levels in the NAS

The mean (±SEM) rate of lever presses for brain stimulation reward during the microdialysis experiment was 62.3 ± 1.9 presses/min and was similar to the rate recorded during the previous training days. During the extinction session (electrical stimulation off), a rapid drop in the rate of lever presses was recorded (Fig. 1A). The basal level of β-endorphin in the dialysates was 108 ± 21 pm. Dialysate β-endorphin levels were not altered during the 20-min LHSS session (F8,40 = 0.61, P > 0.7; n = 6; Fig. 1B). However, a significant increase in β-endorphin levels was observed during the 20-min extinction session that started 1 h after the LHSS session, which then rapidly dropped to baseline levels following the session (F10,50 = 3.57, P < 0.001; n = 6; Fig. 1B). In a different group of rats, the LHSS session started 1 h after the extinction session. Again, β-endorphin levels were not altered during the LHSS session (F8,40 = 0.60, P > 0.7; n = 6; Fig. 1C). By contrast, a significant increase in β-endorphin levels was observed during the 20-min extinction session; the level dropped rapidly to baseline levels following the session (F10,50 = 3.15, P < 0.005; n = 6; Fig. 1C). Non-contingent brain stimulation did not affect dialysate β-endorphin levels (Fig. 1B and C). No correlation was observed between the number of lever presses and the percentage of increase in dialysate β-endorphin levels during extinction (R2 = 0.009).

Details are in the caption following the image

(A) Number (mean ± SEM) of responses on the active lever during tests for the effect of LHSS and extinction of self-stimulation behaviour on β-endorphin levels in the NAS. One group of rats was exposed to LHSS first (n = 6) and the other group of rats was exposed to extinction first (n = 6). Because the response rate pattern during LHSS and extinction was similar in both groups, data from both groups were combined (n = 12). (B) Changes in extracellular levels of β-endorphin (mean ± SEM) in the NAS during LHSS, extinction (EXT) of self-stimulation behaviour (when extinction followed the LHSS session) and during noncontingent (NC) stimulation (n = 6). (C) Changes in extracellular levels of β-endorphin in the NAS during LHSS and extinction of self-stimulation behaviour, when extinction preceeded the LHSS session (n = 6). Data points expressed as a percentage of mean baseline levels of β-endorphin obtained in the five pretest samples. *P < 0.01 vs. baseline and LHSS data points (Student–Newman–Keuls post hoc test).

Experiment 2: effect of extinction of heroin self-administration behaviour on β-endorphin levels in the NAS

Seven of the rats were excluded from the analysis because of failure to acquire heroin self-administration (n = 1), catheter failure during training (n = 5) and microdialysis probe failure (n = 1). The mean ± SEM number of drug infusions, number of responses on the active lever and number of responses on the inactive lever over 2 h on the last day of training were 10.6 ± 2.1, 32.3 ± 10.0 and 0.8 ± 0.4, respectively. The mean ± SEM number of responses on the active lever and number of responses on the inactive lever during the 2-h extinction session were 62.0 ± 13.8 and 2.5 ± 0.4, respectively. Most of the lever responses (40.5 ± 9.7) were observed during the first 20 min of the extinction session (Fig. 2A). Thus, the rats demonstrated an ‘extinction burst’ that is typically observed when drug is no longer available (Yokel, 1987).

Details are in the caption following the image

(A) Number (mean ± SEM) of responses on the active lever during tests for the effect of extinction of heroin self-stimulation behaviour on β-endorphin levels in the NAS (n = 11). (B) Changes in extracellular levels of β-endorphin (mean ± SEM) in the NAS during extinction of heroin self-administration (120 min). Data points are expressed as a percentage of mean baseline levels of β-endorphin obtained in the nine pre-extinction test samples. *P < 0.05 vs. baseline (n = 11; Student–Newman–Keuls post hoc test).

The basal level of β-endorphin in the dialysates, collected from the extracellular fluid in the NAS before the extinction session, was 162 ± 46 pm. The dialysate β-endorphin level was increased significantly at the onset of the 2-h extinction session and gradually returned to the baseline level by the end of the session (F9,90 = 4.18, P < 0.0005; n = 11; Fig. 2B). No correlation was observed between the number of lever presses and the percentage of increase in dialysate β-endorphin levels during extinction (R2 = 0.017).

Experiment 3: effect of intermittent footshock on β-endorphin levels in the NAS

In a different group of naive rats, the basal level of β-endorphin in the dialysates collected from the extracellular fluid in the nucleus accumbens was 137 ± 29 pm (n = 6). During the 20-min intermittent footshock session dialysate β-endorphin levels were increased significantly (F5,25 = 9.48, P < 0.0001) and returned to baseline levels within 20 min (Fig. 3).

Details are in the caption following the image

Changes in extracellular levels of β-endorphin (mean ± SEM) in the NAS during exposure to 20 intermittent footshocks (FS, 0.5 mA). Data points expressed as a percentage of mean baseline levels of β-endorphin obtained in the five pretest samples. *P < 0.05 vs. baseline (Student–Newman–Keuls post hoc test).

Discussion

We found that extracellular β-endorphin levels in the NAS were not affected by LHSS. In contrast, rapid and transient elevations of β-endorphin levels in the NAS were observed during extinction of LHSS- or heroin-reinforced behaviour. Finally, exposure to footshock also increased NAS β-endorphin levels.

The role that opioid peptides in the NAS play in brain reward mechanisms is unclear. A previous study indicated rewarding effects of methionine enkephalin in the NAS using the intracranial self-administration technique (Goeders et al., 1984); however, the selective endogenous µ-opioid endomorphin-1 failed to be self-administered into the NAS (Zangen et al., 2002). In other studies, opiates failed to establish conditioned place preference when injected into the NAS (Bals-Kubik et al., 1993; Olmstead & Franklin, 1997). In addition, there are contradictory reports on the effect of opiates injected into this area on facilitation of brain stimulation reward (Johnson et al., 1995; Van Ree et al., 2000). More recently, Olive et al. (2001) reported an increase in extracellular levels of β-endorphin in the NAS following acute injections of ethanol, cocaine and d-amphetamine and Roth-Deri et al. (2003) reported an increase in extracellular levels of β-endorphin in the NAS following cocaine self-administration. Our present findings, demonstrating lack of effect for LHSS on NAS β-endorphin levels, indicate that the role that NAS β-endorphin plays in the brain reward mechanism might be limited to specific rewards, i.e. ethanol and psychostimulants. This suggestion is supported by data showing that acute exposure to nicotine (Olive et al., 2001) and consumption of sucrose pellets (A. Zangen, unpublished data) do not increase NAS β-endorphin.

However, one cannot ignore the possibility of an additional site of action for β-endorphin in the brain. Thus, in contrast to the ambiguous effects of opiates in the NAS, injections of opiates into the ventral tegmental area (VTA) are consistently reported to induce rewarding effects and to facilitate brain stimulation reward, presumably by disinhibition of dopaminergic cells projecting to the NAS (Jenck et al., 1987; Devine & Wise, 1994; Olmstead & Franklin, 1997). Studies on the effects of opioid antagonists or antibodies to β-endorphin on LHSS (Schaefer, 1988; Carr, 1990) did not include specific microinjections into the NAS; therefore, the reported effects might be mediated by other brain sites. Thus, an appealing alternative site for a rewarding activity of β-endorphin in the brain is the VTA.

In contrast to the lack of changes during LHSS, extinction of the LHSS behaviour dramatically and transiently increased NAS β-endorphin levels. A similar robust and transient increase in NAS β-endorphin was observed during extinction of heroin-reinforced lever pressing (in the present study) and during extinction of cocaine-reinforced lever pressing (Roth-Deri et al., 2003). Interestingly, the increases in β-endorphin levels during extinction of cocaine-reinforced lever pressing were much greater then those induced by cocaine self-administration (Roth-Deri et al., 2003). It is thought that behaviour during extinction is controlled by reward-associated cues (Catania, 1992). It is therefore possible that NAS β-endorphin is involved in the appetitive motivational processes that underlie reward seeking that is triggered by reward-associated cues. This idea is supported by reports that opioid receptor antagonists decrease the conditioned reinforcing effects of natural and other rewards (Gerrits et al., 1995; Delamater et al., 2000). In addition, it has been suggested that endorphins are released before drug self-administration, when drug is expected but not yet available (Sweep et al., 1989; Gerrits et al., 1999). Similarly, in our study, increases in NAS β-endorphin levels were observed when rats were exposed to cues associated with LHSS during extinction but not during the self-stimulation session. Although it seems likely that reward expectation would increase NAS β-endorphin levels during LHSS, we might have missed this effect because of the limited time-resolution of our method and the fact that these expectations were immediately met by reward administration. The degree, however, that reward ‘expectancy’ can account for the present data is unknown in light of the findings of Carroll et al. (1986), that in cocaine-trained rats the opioid antagonist naltrexone had no effect on lever pressing during extinction.

Finally, the increase in NAS β-endorphin reported here may reflect an adaptive response of the brain opioid system to the frustrative/stressful component of the extinction training (Amsel, 1962; Gray, 1967; Feldon & Gray, 1981). The removal of a reinforcer during an extinction test is known to activate the hypothalamic–pituitary–adrenal axis (Mason, 1983) and can also lead to other behavioural and physiological stress responses (Gray, 1987). In addition, exposure to different stressors activates the endogenous opioid system, resulting in the release of β-endorphin in the plasma and the brain (Akil et al., 1984). An obvious example for the role of the opioid system in stress response regulation is the phenomenon of stress-induced analgesia (Amit & Galina, 1986). More importantly, β-endorphin was shown to be critically involved in this phenomenon (Kelsey et al., 1986; Rubinstein et al., 1996). Further support for the above idea comes from our recent findings that stress induced by formalin injection increases extracellular β-endorphin levels in the ArN (Zangen et al., 1998), and that stress induced by intermittent footshock increases extracellular β-endorphin levels in the NAS (present study), and from the reported decreased responding to painful stimuli following intra-NAS β-endorphin injections (Tseng & Wang, 1992). On the basis of the above data we suggest that the increase in NAS β-endorphin level during extinction training is involved in the adaptive response to stress and frustration induced by reward removal. Such adaptive response is not induced (and not required) during LHSS. Moreover, this adaptive response in the NAS is not necessarily involved in the control of lever pressing during extinction training, thus explaining the lack of effect of naltrexone on extinction behaviour (Carroll et al., 1986).

In conclusion, we found that brain stimulation reward does not affect extracellular levels of β-endorphin in the NAS but that extinction of LHSS and heroin self-administration behaviour dramatically increases β-endorphin levels in the NAS. In addition, we found that exposure to footshock stressor elevates NAS β-endorphin levels. We suggest that although we cannot categorically rule out the possibility of independent processes, the elevated extracellular levels of β-endorphin in the NAS during extinction training and exposure to footshock reflect a common adaptive response to stress.

Acknowledgements

This work was supported by the NIDA/IRP programme. We thank Drs Roy Wise and Yavin Shaham for helpful comments and Dr David Highfield for help in computer programming.

Abbreviations

  • ArN
  • arcuate nucleus
  • LHSS
  • lateral hypothalamus self-stimulation
  • NAS
  • nucleus accumbens
  • VTA
  • ventral tegmental area.
    • The full text of this article hosted at iucr.org is unavailable due to technical difficulties.