Activity Defined Neuronal Ensembles in Substance Use Disorder: Reward Competition and Transient Synaptic Plasticity

Abstract

Substance use disorder (SUD) is characterized by maladaptive learning in which drug-associated cues gain disproportionate motivational value compared to natural rewards, impacting behavior long after abstinence from substance use. Across human and preclinical studies, this bias reflects enduring adaptation within the mesolimbic dopamine pathway. Yet, the cellular mechanisms by which neurons associated with drug and natural reward compete and express themselves during relapse remain poorly understood. This dissertation investigates these questions across two complementary domains: reward competition and reward-seeking plasticity in ensemble neurons, defined here as sparsely distributed populations of neurons that are co-activated during a specific behavioral event.First, using reward competition in a non-contingent conditioned place preference (CPP) model, the size of neuronal ensembles was assessed for a single reward of cocaine and chocolate, and in a dual-reward paradigm using both rewards conditioned in the two distinct chambers. Activity-dependent c-Fos-TRAP2 tagging revealed that ensemble size in the prelimbic cortex (PL) and nucleus accumbens core (NAcore) was comparable across single- and dual-reward conditions. Despite robust preference for the reward chambers in single and cocaine in dual-reward conditions, ensemble size did not correlate with time spent in the reward chamber. This indicates that non-contingent CPP ensembles encode 2 contextual associations with rewards rather than motivational value. These findings establish that reward competition does not alter ensemble recruitment in non-contingent CPP and that there is likely a bias in ensemble neurons within the dual-reward group towards cocaine encoding. The second section tested whether transient synaptic plasticity, a feature of cue-induced reinstatement that includes dendritic spine enlargement and an increased AMPA/NMDA ratio, peaking at 30 minutes and returning to baseline by 120 minutes, is observed specifically within drug-seeking ensembles. Using cFos-TRAP2 tagging during cue-induced cocaine reinstatement in male and female mice, combined with structural and functional assays, identified dissociable plasticity between neuronal populations. Ensemble neurons exhibit a structural increase in spine head diameter at 30 minutes that persists through 120 minutes, in contrast to prior non-selective studies. Assessing these changes using spine morphology revealed that mature spines in both ensemble and non-ensemble neurons are destabilized, indicating that remodeling of established spines rather than new spine formation occurs in both populations. In contrast, non-ensemble neurons show functional potentiation, with increased AMPA/NMDA ratios at 30 minutes, whereas ensemble neurons are already potentiated across timepoints following cocaine exposure. These results demonstrate that transient synaptic plasticity is not a uniform property of NAcore neurons; rather, structural and functional components differ between neuronal populations. Together, these findings advance a unified model in which cocaine-associated ensembles retain a persistent pre-potentiated synaptic state and are reengaged during cue-induced reinstatement with structural remodeling. Non-ensemble neurons over the same time span increase their functional activity during cue-induced reinstatement. By integrating reward competition and ensemble-specific plasticity, this dissertation reframes how neuronal ensembles encode drug-associated memories and identifies ensemble-targeted mechanisms for future studies aimed at developing relapse-prevention strategies.