The activity-regulated cytoskeleton-associated protein (Arc) functions in a cell type- and sex-specific manner in the adult nucleus accumbens to regulate non-contingent cocaine behaviors

Abstract

Repeated cocaine use produces adaptations in brain function that contribute to long-lasting behaviors associated with cocaine use disorder (CUD). In rodents, the activity-regulated cytoskeleton-associated protein (Arc) can regulate glutamatergic synaptic transmission, and cocaine regulates Arc expression and subcellular localization in multiple brain regions, including the nucleus accumbens (NAc)-a brain region linked to CUD-related behavior. We show here that repeated, non-contingent cocaine administration in global Arc KO male mice produced a dramatic hypersensitization of cocaine locomotor responses and drug experience-dependent sensitization of conditioned place preference (CPP). In contrast to the global Arc KO mice, viral-mediated reduction of Arc in the adult male, but not female, NAc (shArc^NAc) reduced both CPP and cocaine-induced locomotor activity, but without altering basal miniature or evoked glutamatergic synaptic transmission. Interestingly, cell type-specific knockdown of Arc in D1 dopamine receptor-expressing NAc neurons reduced cocaine-induced locomotor sensitization, but not cocaine CPP; whereas, Arc knockdown in D2 dopamine receptor-expressing NAc neurons reduced cocaine CPP, but not cocaine-induced locomotion. Taken together, our findings reveal that global, developmental loss of Arc produces hypersensitized cocaine responses; however, these effects cannot be explained by Arc's function in the adult mouse NAc since Arc is required in a cell type- and sex-specific manner to support cocaine-context associations and locomotor responses.

Keywords: AMPA; Arc/Arg3.1; NMDA; cocaine; conditioned place preference; knockout mouse; medium spiny neurons; nucleus accumbens; sensitization; sex differences.

FIGURE 1

Global Arc KO mice show enhanced cocaine locomotor sensitization. (A) Experimental timeline for locomotor testing. Mice received 4 days of habituating saline injections before a week of daily cocaine injections (15 mg/kg). After 7 days of home‐cage abstinence, mice received individual cocaine challenge doses separated by home cage abstinence periods (indicated above challenge doses). (B) Arc KO mice show enhanced locomotor responses to cocaine across the sensitizing regimen and following challenge injections after 7 days withdrawal (n = 11–13/condition). (C) Enhanced Arc KO locomotor behavior is not associated with increased stereotypy behavior (n = 7–14/condition). Data shown are mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 2

Arc KO mice show drug experience‐dependent enhancement of cocaine reward. (A) Experimental timeline for cocaine sensitization and subsequent conditioned place preference. Mice received 4 days of habituating saline injections before a week of daily cocaine injections (15 mg/kg). After 7 days of home‐cage abstinence, mice were re‐exposed to the locomotor chamber in the drug‐free state (black arrow) and then underwent cocaine CPP (C = cocaine, 5 mg/kg; S = saline). (B) Arc KO mice show enhanced locomotor responses to cocaine across the sensitizing regimen, with no differences in locomotor activity in repeated‐saline exposed mice (n = 17–19/condition). (C) Cocaine‐sensitized Arc KO mice show increased cocaine conditioned place preference to low dose cocaine (5 mg/kg; i.p.) compared with saline‐treated Arc KO mice. CPP score is post‐pre (n = 9–11/condition). Data shown are mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 3

Arc reduction in the adult male NAc reduces cocaine‐related behavior without altering basal synaptic activity. (A) Experimental timeline for shRNA studies. AAV2 vectors containing shRNA expression constructs were infused 14 days prior to the start of cocaine CPP (5 mg/kg; i.p.). Mice were then tested for locomotor sensitization (two‐injection protocol; 20 mg/kg; i.p.). (B) shArc^NAc mice show reduced conditioned place preference (n = 8–12/condition). (C) shArc^NAc mice show reduced cocaine‐induced locomotor activity (n = 9–14/condition). (D) shArc^NAc does not alter mEPSC amplitude (left) or frequency (middle) in drug naive mice (n = 5–6 mice/condition, 24–40 cells/condition). (Right) Representative traces. (E) shArc^NAc does not alter AMPAR (left) nor NMDAR‐mediated (middle) currents (n = 5–6 mice/condition, 24–40 cells/condition). (Right) Representative traces. Data shown are mean ± SEM. *p < 0.05.

FIGURE 4

Cell type‐specific Arc reduction in male NAc produces distinct cocaine‐related behaviors. (A) Experimental timeline for shRNA studies. Mice were treated with the same schedule as in Figure 3. (B) Arc reduction in NAc of male D1‐Cre mice (shArc^D1) does not alter cocaine CPP behavior (n = 9‐10/condition). (C) shArc^D1 mice show reduced cocaine locomotor activity compared with shCntrl^D1 mice (n = 8/condition). (D) Arc reduction in NAc of male D2‐Cre mice (shArc^D2) reduces cocaine CPP behavior compared with shCntrl^D2 mice (n = 10–15/condition). (E) shArc^D2 mice do not show altered locomotor responses to cocaine (n = 8–9/condition). Data shown are mean ± SEM. *p < 0.05; **p < 0.01.