Cocaine serves as a peripheral interoceptive conditioned stimulus for central glutamate and dopamine release

Abstract

Intravenous injections of cocaine HCl are habit-forming because, among their many actions, they elevate extracellular dopamine levels in the terminal fields of the mesocorticolimbic dopamine system. This action, thought to be very important for cocaine's strong addiction liability, is believed to have very short latency and is assumed to reflect rapid brain entry and pharmacokinetics of the drug. However, while intravenous cocaine HCl has almost immediate effects on behavior and extracellular dopamine levels, recent evidence suggests that its central pharmacological effects are not evident until 10 or more seconds after IV injection. Thus the immediate effects of a given intravenous cocaine injection on extracellular dopamine concentration and behavior appear to occur before there is sufficient time for cocaine to act centrally as a dopamine uptake inhibitor. To explore the contribution of peripheral effects of cocaine to the early activation of the dopamine system, we used brain microdialysis to measure the effects of cocaine methiodide (MI)--a cocaine analogue that does not cross the blood brain barrier--on glutamate (excitatory) input to the dopamine cells. IP injections of cocaine MI were ineffective in cocaine-naïve animals but stimulated ventral tegmental glutamate release in rats previously trained to lever-press for cocaine HCl. This peripherally triggered glutamate input was sufficient to reinstate cocaine-seeking in previously trained animals that had undergone extinction of the habit. These findings offer an explanation for short-latency behavioral responses and immediate dopamine elevations seen following cocaine injections in cocaine-experienced but not cocaine-naïve animals.

Figure 1. Mean lever-pressing (top panels) and ventral tegmental glutamate (middle panels) and dopamine (lower panels) fluctuations during reinstatement of lever-pressing by cocaine HCl (left panels) and cocaine MI (right panels).

IP injections of cocaine HCl (10 mg/kg) or cocaine MI (13 mg/kg) were given at Time = 0. Abbreviations: aCSC (artificial cerebrospinal fluid), Kyn (kynurenic acid, glutamate antagonist), cocaine HCl (cocaine hydrochloride), cocaine MI (cocaine methiodide). The notation “Cocaine HCl-aCSF” indicates the condition where the IP injection was cocaine HCl and where the dialysate included aCSF.

Figure 2. Mean glutamate top panels) and dopamine (bottom panels) fluctuations during yoked injections of cocaine HCl (left panels) or cocaine MI (right panels) in cocaine-naïve animals or animals experienced with cocaine HCl.

The dotted line in B represents the effects in cocaine-experienced animals of the peripheral-plus-central actions of cocaine HCl (circles in B) minus the peripheral-only action of cocaine MI (circles in D). The difference between this line and the line reflecting the effects of cocaine HCl in cocaine-naïve animals (triangles in B) is the consequence of sensitization to cocaine in the cocaine-trained animals .

Figure 3. Levels of free and methylated cocaine (M Cocaine) in blood and brain following IP injections of cocaine MI (13 mg/kg) (top panel) or cocaine HCl (10 mg/kg) (bottom panel).

Cocaine HCl dissociates at physiological pH and free cocaine enters the brain readily. The methyl group does not dissociate readily from cocaine MI at physiological pH, and the methylated quaternary salt does not cross the blood brain barrier.

Figure 4. Illustrative chromatograms from blood (red line) and brain (blue line) dialysates taken from an animal given IP cocaine MI (13 mg/kg) (top panel) and an animal given IP cocaine HCl (10 mg/kg) (bottom panel).

The green line shows a chromatogram from a standard solution containing free cocaine, methylated cocaine, and the primary cocaine metabolite bonzoylecgonine.