Ventral Striatopallidal Regulation of Cocaine Addiction

Addiction is characterized by a recurrent pattern of relapse to drug seeking and drug use. This behavior is driven by neuronal activity in the nucleus accumbens and ventral pallidum (VP). Nucleus accumbens dopamine D1 and D2 receptor expressing medium spiny neurons (D1-/D2-MSN) both project to the VP, and we investigated the effects of cocaine self administration on D1-VP and D2-VP projections (see: Chapter III). After cocaine self-administration, we observed a synaptic loss of function in the D2-VP projection, and further reducing D2-VP functioning augmented the motivation to seek cocaine. Activating D1-VP projections also increased reinstatement. These data indicate that cocaine induced weakening of D2-VP pathway predisposes animals to relapse. D2-MSN exclusively project to the VP, whereas D1-MSN also project to the ventral midbrain (VM). We investigated whether D1-MSN collateralize between the VP and VM, and tested the involvement of D1-VP and D1-VM projections during reinstatement of cocaine seeking (see: Chapter IV). Surprisingly, nucleus accumbens D1-MSN heavily collateralize between the VP and VM, but only D1-VP projections regulate cocaine seeking. These data further implicate the D1-VP pathway as a critical regulator of relapse to cocaine seeking. Besides a major population of GABAergic neurons (VP-GABA), the VP also contains a substantial number of glutamatergic neurons (VP-Glu). We used retrograde rabies tracing to investigate the innervation of VP-Glu and VPGABA neurons by D1- and D2-MSN, and used chemogenetics to see whether these cells differentially drive reinstatement. We also investigated whether cocaine alters the synaptic inputs onto these neurons (see: Chapter V). VP-Glu neurons inhibit reinstatement. Furthermore, VP-GABA neurons drive drug seeking and taking behavior, but do not mediate reinstatement. Instead, a small subset of VP-GABA neurons that co-express enkephalin (VP-Penk) drive reinstatement. Inhibitory inputs onto VP-GABA neurons originate from both D1- and D2-MSN and are reduced after cocaine self-administration. Meanwhile VP-Penk and VP-Glu neurons are preferentially innervated by D1-MSN, and inhibitory inputs onto VP-Penk neurons, but not VP-Glu neurons, are increased following cocaine self-administration. Combined these data indicate that cued reinstatement is driven by D1-VP projections onto VP-Glu and VP-Penk neurons, and they place the VP as a central component of basal ganglia circuits controlling addiction