The mesolimbic reward pathway is necessary for disruptions in cocaine-seeking behavior following mediated devaluation

Abstract

We developed an approach to disrupt cocaine-seeking behaviors using mediated devaluation. Male rats underwent cocaine self-administration training in which active lever responses led to cocaine infusions and the presentation of a tone-light conditioned stimulus (CS). Subsequently, during mediated devaluation rats received non-contingent presentations of the cocaine-associated CS in a second distinct context, which led to the cue-evoked retrieval of associated memories. This was immediately followed by an intraperitoneal injection of lithium chloride (LiCl) and served to pair the memory of cocaine reward with gastric malaise. Consequently, this led to a substantial reduction in cocaine-seeking behavior during extinction training, relative to rats that received CS-saline or LiCl alone during mediated devaluation. Cue- and cocaine-evoked reinstatement testing indicated that the manipulations did not devalue the CS or the reinforcing properties of cocaine. A separate cohort of rats received a dual-viral chemogenetic strategy that permitted circuit-specific inactivation of midbrain ventral tegmental area (VTA) cells projecting to the nucleus accumbens (NAc). Inactivation of VTA→NAc circuitry during mediated devaluation prevented the subsequent reduction of cocaine-seeking behavior during extinction training. Overall, these findings suggest that intact mesolimbic signaling is required to enable disruptions in cocaine-seeking behavior following mediated devaluation.

Fig. 1. Whole cell patch clamp recordings in VTA→NAc hM4Di-expressing cells.

a Bath application with CNO significantly attenuated the number of APs generated following step-wise increases in depolarization current (pA). b, c CNO treatment significantly reduced the (b) total number and (c) percentage increase in APs following the depolarization stimulus. ** significant current step X drug interaction, p < 0.01. *Overall reduction in APs, p’s<0.05.

Fig. 2. Immunohistochemical verification of Cre-dependent mCherry and hM4Di in VTA→NAc circuitry.

a Heat maps revealing the extent of bilateral mCherry and hM4Di expression. b Heat maps revealing the extent of bilateral eGFP-Cre expression in NAc. Light shading reflects minimal spread, darker shading reflects maximal spread. c–e Colocalization between (c) TH+ neurons (green), (d) hM4Di (red) and (e) merged images. Arrows indicate somatic expression. f Representative micrographs of mCherry-hM4Di positive fibers in the NAc (arrows indicate visible fibers in NAc). g Quantification of colocalization between TH+ and hM4Di-expressing cells throughout the rostral-caudal axis of the VTA. Overall colocalization was ~ 64%. AC anterior commissure.

Fig. 3. Cocaine self-administration training, mediated devaluation and extinction.

a Active lever responses, b inactive lever responses, c number of infusions did not differ between viral or upcoming mediated devaluation conditions. Main effect of session, ****p < 0.0001, ***p < 0.001. d Simplified schematic for mediated devaluation testing adapted with permission [75]. e Active lever responses in the initial 2 h extinction training session in mCherry rats previously treated with either saline or LiCl during mediated devaluation. Significant reduction in cocaine-seeking behavior as a result of mediated devaluation, **p = 0.01, *p’s<0.05, #p = 0.06. Mediated devaluation did not impact inactive lever responses. f Inactivating VTA→NAc cells during mediated devaluation prevented the subsequent attenuation in cocaine-seeking behavior in hM4Di rats. Mediated devaluation did not impact inactive lever responses in hM4Di rats.

Fig. 4. Cue and cocaine-primed reinstatement.

a Cue-induced reinstatement led to substantial increases in active lever responses relative to previous extinction session. Overall hM4Di rats displayed greater reinstatement than mCherry group. **** main effect of testing phase, p’s<0.001. * main effect of virus, p < 0.05. b Time course of active lever responding during reinstatement test. c Overall inactive lever responses during reinstatement test. d Mediated devaluation did not impact cocaine-primed cue-induced reinstatement as similar levels of active lever responses were noted across saline and LiCl conditions. **** main effect of testing phase, p’s<0.001. e Time course of active lever responding during cocaine-primed reinstatement test. (f) Inactive lever responding during cocaine-primed reinstatement test.

Fig. 5. CS evoked retrieval of cocaine reward is necessary for mediated devaluation.

a Active lever responses, b inactive lever responses, c number of infusions did not differ between upcoming mediated devaluation condition. Main effect of session, ***p = 0.001, ****p < 0.001. d Active lever responses in the 2 h extinction test in rats that received CS-LiCl pairing (paired) or LiCl alone (LiCl alone) during mediated devaluation. Significant reduction in cocaine-seeking behavior as a result of CS-LiCl pairing. Main effect of condition, *p < 0.05. e Time course of active lever responding during extinction test. ★Condition x time interaction (F(5,50) = 2.73, p < 0.05), reduction in cocaine-seeking as a result of mediated devaluation condition, #p = 0.05, *p < 0.05. f Inactive lever responses during extinction did not differ as a function of condition.