The Projection Targets of Medium Spiny Neurons Govern Cocaine-Evoked Synaptic Plasticity in the Nucleus Accumbens

Abstract

We examine synaptic connectivity and cocaine-evoked plasticity at specific networks within the nucleus accumbens (NAc). We identify distinct subpopulations of D1+ medium spiny neurons (MSNs) that project to either the ventral pallidum (D1+^VP) or the ventral tegmental area (D1+^VTA). We show that inputs from the ventral hippocampus (vHPC), but not the basolateral amygdala (BLA), are initially biased onto D1+^VTA MSNs. However, repeated cocaine exposure eliminates the bias of vHPC inputs onto D1+^VTA MSNs, while strengthening BLA inputs onto D1+^VP MSNs. Our results reveal that connectivity and plasticity depend on the specific inputs and outputs of D1+ MSNs and highlight the complexity of cocaine-evoked circuit level adaptations in the NAc.

Keywords: cocaine sensitization; medium spiny neuron; nucleus accumbens; spiny projection neuron; synaptic plasticity.

Figure 1.. D1+ MSNs Projecting to VP and VTA Are Distinct Cell Populations

(A) Left: schematic and representative images for injections of CTB-647 into the VP (green) and CTB-488 into the VTA (blue) of D1-tdTomato mice (red) (scale bar, 500 μm). Right: confocal images of retrogradely labeled neurons in the NAcMS (arrows indicate dual-labeled MSNs; scale bar, 100 μm). (B) Left: quantification of CTB and tdTomato overlap in the NAcMS (total cell count of three slices per mouse, five mice). Right: summary of the number of D1+^VTA, D1+^VP and dual-projecting MSNs in the NAcMS (average number of labeled cells of three slices per mouse, five mice). (C) Two-photon images of D1+^VTA, D1+^VP, and D1− MSNs (left) (scale bar, 20 μm) and quantification of total dendritic length for all reconstructed cells (right), showing no differences. (D) Examples of physiological responses of D1+^VTA, D1+^VP, and D1− MSNs to current injections of 250 and −50 pA. Right: summary of current step-evoked firing, showing similar number of action potentials (APs) for D1+^VTA and D1+^VP MSNs. Box-and-whisker plots represent median and minimum to maximum. F-I curve is presented as mean ± SEM. *p < 0.05. See also Figure S1.

Figure 2.. vHPC Inputs Are Stronger onto D1+^VTA Cells in Naive Mice

(A) Left: schematic for injections of AAV-ChR2-eYFP into the vHPC and CTB-647 into the VP or the VTA. Right: representative image of viral expression in the vHPC (scale bar, 500 μm) and of dense vHPC axon labeling in the NAcMS (scale bar, 100 μm). (B) Left: vHPC inputs evoke larger EPSCs at D1+^VTA MSNs compared with neighboring D1− MSNs. Middle: summary of the absolute amplitude of vHPC-evoked AMPAR EPSCs at D1− and D1+^VTA MSNs, where lines indicate pairs of recorded neurons. Right: summary of D1+^VTA/D1− amplitude ratios for AMPAR EPSCs at −70 mV and NMDAR EPSCs at +40 mV. (C) Similar to (B), showing no bias of vHPC inputs onto pairs of D1+^VP and D1− MSNs. (D) Left: schematic for injections of AAV-ChR2-eYFP into the BLA and CTB-647 into the VP or the VTA. Right: representative image of viral expression in the BLA (scale bar, 500 μm) and of BLA axon labeling in the NAcMS (scale bar, 100 μm). (E) Left: there is no bias of BLA-evoked EPSCs at D1+^VTA MSNs compared to neighboring D1− MSNs. Middle: summary of the absolute amplitude of BLA-evoked AMPAR EPSCs at D1− and D1+^VTA MSNs, where lines indicate pairs of recorded neurons. Right: summary of D1+^VTA/D1− amplitude ratios for AMPAR EPSCs at −70 mV and NMDAR EPSCs at +40 mV. (F) Similar to (E) showing no bias of BLA-evoked EPSCs at D1+^VP MSNs compared to neighboring D1− MSNs. Average traces are presented as mean ± SEM. Ratio data are presented as geometric mean with 95% CI on logarithmic axes. *p < 0.05. See also Figure S1.

Figure 3.. Cocaine-Induced Plasticity Depends on Projection Target

(A) Sensitization protocol, with 3 days of habituation, 5 days of either cocaine or saline injections, and 1 day of withdrawal, followed by either cocaine challenge to assess behavior or slice recordings to study synaptic physiology. (B) Left: typical movement of saline and cocaine mice to a challenge dose of cocaine on day 6 of the protocol. Middle: summary of distance traveled in the cocaine challenge session. Right: summary of distance traveled per experiment day. (C) Left: vHPC inputs evoke larger EPSCs at D1+^VTA MSNs in saline mice, but this bias is abolished in cocaine mice. Right: summary of D1+^VTA/D1− amplitude ratios. (D) Similar to (C), showing equivalent vHPC-evoked EPSCs at D1+^VP and D1− MSNs in both saline and cocaine mice. (E) Left: BLA-evoked EPSCs are similar at D1+^VTA and D1− MSNs in saline and cocaine mice. Right: summary of D1+^VTA/D1− amplitude ratios. (F) Similar to (E) but showing a large increase in EPSCs at D1+^VP MSNs in cocaine mice. Average traces are presented as mean ± SEM. Ratio data are presented as geometric mean with 95% CI on logarithmic axes. *p < 0.05. See also Figure S3.