Alan S. Keys

Self-administration of cocaine increases the release of acetylcholine to a greater extent than response-independent cocaine in the nucleus accumbens of rats.

Abstract  Rationale: The neurochemical effects of psychostimulant exposure may depend on how these drugs are encountered. A useful method for examining this issue is to compare neurotransmitter release following response-dependent, or self-administered, drug exposure and response-independent exposure. Objectives: This experiment examined the effect of active and passive cocaine administration on acetylcholine (ACh) efflux in the shell region of the nucleus accumbens (NAc) in rats. Methods: One group of rats (CSA: cocaine self-administration) was trained to lever-press for intravenous infusions of cocaine (0.42 mg/kg per infusion) on a fixed-ratio-1 schedule of reinforcement. Cocaine infusions were accompanied by the onset of a stimulus light that signaled a 20-s time-out period. Control rats received intravenous cocaine (cocaine non-contingent: CNC) or saline (SAL) in a manner that was not contingent upon their behavior. Drug infusions in these groups were determined by the lever-press behavior of the animals in the CSA group, i.e. they were yoked to rats in the self-administration group such that CNC animals received equal amounts of cocaine as CSA rats. Animals received cocaine or saline in 3-h sessions for 13 consecutive days before testing. On day 14, extracellular ACh was measured in 15-min intervals before, during and after a 3-h session of cocaine exposure using unilateral microdialysis probes located in the NAc shell coupled with HPLC. Results: ACh efflux was significantly increased above baseline in both groups of rats that received cocaine but CSA rats had significantly higher ACh levels during the self-administration period compared to their yoked counterparts. In addition, ACh efflux remained elevated longer in CSA animals relative to CNC rats following cessation of cocaine exposure. Conclusions: These results demonstrate that ACh interneurons in the NAc shell are responsive to cocaine exposure. In addition, these findings suggest that the manner in which the drug is administered (i.e. either by active self-administration or passive exposure) may be relevant to the magnitude of the neural response.

Reduced glutamate immunolabeling in the nucleus accumbens following extended withdrawal from self-administered cocaine

Alterations in the density of GABA and glutamate immunolabeling within nerve terminals in the shell region of the nucleus accumbens were assessed in rats withdrawn from intravenous cocaine exposure. Four groups of rats were used: one group self‐administered cocaine (0.42 mg/ kg/ infusion) in daily 3‐h sessions for approximately 2 weeks, two additional groups received either saline or cocaine in a noncontingent fashion, and a fourth comprised a drug‐naive, age‐matched control group. Immunogold electron microscopy was used to quantify presynaptic terminal GABA and glutamate density within the vesicular and mitochondrial pools approximately 18 days following the last drug or saline exposure in the treatment groups. A significant 27.7% decrease in vesicular glutamate density within asymmetrical nerve terminals was observed in animals that self‐administered cocaine as compared to controls. This group also showed an 18.6% decrease in vesicular nerve terminal glutamate immunolabeling as compared to animals that were administered a similar total dose of cocaine in a response‐independent fashion. No significant changes in the density of nerve terminal GABA vesicular immunolabeling were observed in any groups. For both transmitters, no differences were detected in the density of immunolabeling within the presynaptic mitochondrial (i.e., metabolic) pool. These results demonstrate that glutamate density is suppressed in the shell region of the nucleus accumbens following withdrawal from 2 weeks of cocaine exposure. The findings also suggest that the motivational aspects that accompany self‐administration may participate in this reduction. Synapse 30:393–401, 1998.

D1 and D2 dopamine receptor mediation of amphetamine-induced acetylcholine release in nucleus accumbens.

To assess the interaction of dopamine and acetylcholine systems in the rat nucleus accumbens in response to direct D-amphetamine administration, in vivo microdialysis measures of acetylcholine were used during reverse dialysis of amphetamine alone and in combination with D1 and D2 receptor antagonists SCH 23390 and sulpiride, respectively. During a 15-min exposure to amphetamine (50 microM) in the nucleus accumbens, acetylcholine increased to 33% above pre-infusion levels, became maximal at 15 min post-infusion (+41%) and gradually returned to baseline levels by 60 min post-amphetamine. Conversely, amphetamine (1 mM) administration caused a biphasic change in acetylcholine release with a trend toward a decrease (-14%) during exposure followed by a significant increase (+36%) at 30 min post-amphetamine that returned to baseline levels by 60 min after infusion. The increases observed during amphetamine (50 microM) exposure and during recovery from amphetamine (1 mM) were both blocked by co-administration with the D1 antagonist, SCH 23390 (10 microM), but not with the D2 antagonist, sulpiride (10 microM). Co-infusion of sulpiride eliminated the trend toward reduced acetylcholine release observed during 1 mM amphetamine whereas co-administration of SCH 23390 potentiated this decrease. A possible tonic D1 facilitation of nucleus accumbens acetylcholine release was indicated by the consistent reductions in acetylcholine release observed during infusion of SCH 23390. These results suggest that amphetamine administration in the nucleus accumbens induces a bidirectional change in acetylcholine release that is dependent on dose and opposing effects of nucleus accumbens D1 and D2 activation. In general, relatively low doses of amphetamine administered into the nucleus accumbens caused an increase in acetylcholine release that was dependent on dopamine D1 receptors whereas higher doses of amphetamine resulted in a D2-mediated decrease.

Involvement of acetylcholine in the nucleus accumbens in cocaine reinforcement.

The nucleus accumbens (NAc) in the mesolimbic system has been identified as a key component of the neurobiological circuit that mediates the reinforcing properties of many drugs of abuse. The NAc consists primarily of GABAergic neurons but also contains a smaller population of acetylcholine (ACh)-containing interneurons.1,2 These large aspiny interneurons may have an important integrative function, and their involvement in drug reward has not been systematically investigated. Our first studies using microdialysis found that cocaine self-administration increased ACh levels in the NAc to a greater extent than response-independent (i.e., yoked) cocaine.3 These findings suggested that the context of drug experience may influence the neurochemical response to cocaine. In the two experiments described here, we have examined the interaction between cocaine self-administration and the activity of cholinergic interneurons in the NAc using microdialysis for acetylcholine and intracerebral microinjections. In our first experiment, we sought to determine if the response of NAc ACh neurons to the glutamate agonist AMPA was different in animals that self-administered cocaine versus those that received passive cocaine exposure. In the second experiment, we measured cocaine self-administration on a progressive ratio schedule of reinforcement in rats that received intraaccumbens injections of cholinergic agonists.

Continuous cocaine induces persisting alterations in dopamine overflow in caudate following perfusion with a D1 agonist.

The present study examined whether exposure to 5 days of continuous cocaine in rats would produce any persisting alterations of the decrease in striatal dopamine (DA) overflow produced by local infusion of a D1 receptor agonist. Using a microdialysis probe in striatum, changes in DA, DA metabolites, and GABA were assessed 14 to 21 days following a 5-day continuous cocaine treatment. There were no differences in baseline levels of DA and its metabolites. SKF 38393 (10−6) infusion into the striatum decreased striatal DA levels in the controls and this effect was attenuated in cocaine-pretreated rats. This result, together with other observations, supports the hypothesis of a persistently altered D1-mediated negative feedback produced by previous exposure to continuous cocaine.

Persisting changes in brain glucose uptake following neurotoxic doses of phencyclidine which mirror the acute effects of the drug

Phencyclidine (PCP) can induce a model psychosis which has a number of similarities to dementias and schizophrenia. In some cases the psychosis persists for prolonged periods after drug discontinuation.N-Methyl-d-aspartate (NMDA) antagonists such as PCP induce increases in glucose metabolism in a variety of brain structures but most notably in limbic regions such as retrosplenial, piriform, and entorhinal cortex, hippocampus, and olfactory tubercle. When given continuously for several days, these NMDA antagonists induced neural degeneration in these same critical limbic areas. In the present study regional 2-fluorodeoxyglucose (FDG) uptake was measured in rats at both 24 h and 10 days after neurotoxic, 5-day “binge” PCP administration. At 24 h after minipump removal there were persisting and large increases in glucose uptake in many brain regions, with maximal changes in the same limbic structures in which neurotoxicity has been observed. Surprisingly, many of these regions still showed elevated glucose metabolism after 10 days of recovery. These findings suggest an anatomical and neurochemical substrate for the persisting psychosis which can occur following PCP.

Oral movement patterns induced in rats by local infusions into striatum depend upon the regimen of prior neuroleptic exposure

Rats were pretreated for 11 months with vehicle or with chronic haloperidol (HAL), administered either continuously (in the drinking water) or intermittently (via weekly injections). During this time the animals were habituated to an enclosed tube and periodically monitored by a computerized video device which measured their oral movements. The rats were then withdrawn from chronic HAL and bilateral cannulae were implanted in the ventrolateral striatum (VLS) and substantia nigra (SN). One week later oral movements were observed in an open cage and then measured by the computerized video device following bilateral infusions into VLS of the muscarinic agonist pilocarpine or the dopamine D1 agonist SKF38393, or following infusions of the GABA antagonist bicuculline into SN. Agonist infusions into VLS had different effects depending upon the prior regimen of chronic HAL. Infusions of pilocarpine into VLS led to an exaggeration of the distinctive oral movement form which follows continuous HAL but an attenuation of the different oral syndrome in the intermittent chronic HAL animals. Infusions of SKF38393 into VLS had similar, but considerably smaller effects. Infusions of bicuculline into SN did not induce either effect. These results indicate differences exist in either striatum or its output circuitry in the neurochemical mechanisms which mediate the different oral movement forms induced by different chronic neuroleptic regimens.

Computer-Assisted Video Assessment of Dyskinetic Movements

Although tardive dyskinesia (TD) is a severe side effect of neuroleptics, animal models of TD have not proven generally fruitful in developing therapeutic counterstrategies or a coherent theory of why chronic neuroleptics induce this and other disorders. This is partially because relatively few primates have been maintained on the continuous proper drug regimen of neuroleptics for sufficient periods to produce TD, and also because rodent models based on the simple observational technique of counting mouth movements have yielded highly controversial results.