Voltage-gated potassium channels control extended access cocaine seeking: a role for nucleus accumbens astrocytes

Abstract

Dopaminergic signaling in the nucleus accumbens shell (NAc) regulates neuronal activity relevant to reward-related learning, including cocaine-associated behaviors. Although astrocytes respond to dopamine and cocaine with structural changes, the impact of dopamine and cocaine on astrocyte functional plasticity has not been widely studied. Specifically, behavioral implications of voltage-gated channel activity in the canonically non-excitable astrocytes are not known. We characterized potassium channel function in NAc astrocytes following exposure to exogenous dopamine or cocaine self-administration training under short (2 h/day) and extended (6 h/day) access schedules. Electrophysiological, Ca²⁺ imaging, mRNA, and mass spectrometry tools were used for molecular characterization. Behavioral effects were examined after NAc-targeted microinjections of channel antagonists and astroglial toxins. Exogenous dopamine increased activity of currents mediated by voltage-gated (K_v7) channels in NAc astrocytes. This was associated with a ~5-fold increase in expression of Kcnq2 transcript level in homogenized NAc micropunches. Matrix-assisted laser desorption/ionization mass spectrometry revealed increased NAc dopamine levels in extended access, relative to short access, rats. K_v7 inhibition selectively increased frequency and amplitude of astrocyte intracellular Ca²⁺ transients in NAc of extended access rats. Inhibition of K_v7 channels in the NAc attenuated cocaine-seeking in extended access rats only, an effect that was occluded by microinjection of the astrocyte metabolic poison, fluorocitrate. These results suggest that voltage-gated K⁺ channel signaling in NAc astrocytes is behaviorally relevant, support K_v7-mediated regulation of astrocyte Ca²⁺ signals, and propose novel mechanisms of neuroglial interactions relevant to drug use.

Fig. 1. Dopamine decreases astrocyte sensitivity to 4-AP.

A Current-voltage relationship of whole cell currents in NAc astrocytes from control cells in regular aCSF (open circles, n = 10/N = 4) or in aCSF containing 4-AP (closed circles, n = 6/N = 4). Inset: representative traces at −120 mV and at +40 mV (black trace: aCSF; blue trace: aCSF + 4-AP; scale bars, 100 ms by 2000 pA). *p < 0.05; **p < 0.01; ***p < 0.001, Sidak’s post-hoc comparisons of Control-aCSF vs. Control-4AP at indicated voltage steps. B Current-voltage responses from astrocytes pre-incubated with dopamine (DA)-containing aCSF (0.1 μM, n = 13/N = 9; 75 μM, n = 12/N = 7). Note that dopamine was included in pre-incubation solution, but was absent from aCSF during electrophysiological recordings. *p < 0.05; **p < 0.01; ***p < 0.001, (Control-aCSF vs. DA75μM-aCSF). ^#p < 0.05 (DA0.1μM-aCSF vs. DA75μM-aCSF). Sidak’s post-hocs at indicated voltage steps. C Pre-incubation with dopamine reduced 4-AP inhibition of astrocyte currents (DA75μM-aCSF, n = 12/N = 7; DA75μM-4AP, n = 7/N = 4). Inset: representative traces at −120 mV and +40 mV (black trace: aCSF; red trace: aCSF + 4-AP; scale bars, 100 ms by 2000 pA). D Concentration-dependence of dopamine effects on 4-AP sensitivity. Current-voltage relationship from slices pre-incubated with 0.1 μM dopamine (pink, DA0.1μM-4AP, n = 6/N = 4) are overlaid on responses under control-4AP (blue, from panel A) and DA75μM-aCSF (red, from panel C) conditions. *p < 0.05; **p < 0.01; ***p < 0.001 (Control-4AP vs. DA75μM-4AP). ^#p < 0.05; ^##p < 0.01; ^###p < 0.001 (DA0.1μM-4AP vs. DA75μM-4AP). Sidak’s post-hocs at indicated voltage steps.

Fig. 2. Dopamine upregulates Kcnq transcript levels and astrocyte K_v7 activity.

A qPCR analysis of Kcnq2-5 mRNA levels in the NAc homogenates with and without DA exposure (NAc punches were pooled from 4–8 brain slices for each transcript). *p < 0.05; Mann–Whitney U. B Representative traces from astrocytes pre-incubated in regular aCSF or with DA (75 μM) before and during bath application of the K_v7 antagonist, XE 991 (10 μM). Scale bar: 100 ms by 2000 pA. C XE 991 effects are normalized to baseline current without XE 991 in the bath at indicated voltage steps. Note prominent inhibition of whole-cell currents by XE 991 in cells pre-incubated with dopamine, but potentiation of whole-cell currents by XE 991 in cells maintained in regular aCSF. D Current-voltage plots illustrate mean (non-normalized) amplitude ±SEM of whole-cell currents in NAc astrocytes during XE 991 application after pre-incubation with control (n = 12/N = 4) or dopamine aCSF (n = 8/N = 4).

Fig. 3. NAc dopamine metabolism differences after short and extended access cocaine.

A Mass spectrum from one of the imaged slices highlighting the dopamine peak (152.068 m/z ± 44.1ppm). B Representative MALDI-MSI image of dopamine localization in brain slices from short (top) and extended (bottom) access cocaine rats. C Relative abundance of dopamine and its metabolites, DOPAC and HVA, in the NAc. **p < 0.01, ***p = 0.001, Mann–Whitney U. D Relative abundance of dopamine in hippocampal area CA1 and ventral tegmental area (VTA). *p < 0.05, Mann–Whitney U.

Fig. 4. K_v7 regulates astrocyte Ca²⁺ after short- and extended-access cocaine.

A Median intensity projections and representative traces of GCamp6f fluorescence in NAc astrocytes from short (left) and extended access (right) groups before (top) and during (bottom) exposure to XE 991 (10 μM). B–E Bar histograms of active ROI numbers (per field of view), amplitude, frequency and duration of Ca²⁺ events in NAc astrocytes (n = 31 (short-aCSF), 17 (short-XE 991), 8 (extended-aCSF), and 9 (extended-XE 991) slices, from N = 6 animals in each group, contributing a total of ~2000–4000 Ca²⁺ events). *p < 0.05, **p < 0.01; ***p < 0.001; ****p < 0.0001; Sidak’s post-hocs.

Fig. 5. NAc astrocytes are required for XE 991-mediated attenuation of cocaine seeking after extended access to cocaine.

A Schematic of the experimental protocol. B–D Effects of FC and XE 991 microinjections on cocaine infusion numbers, total active lever presses, and total inactive lever presses (N = 4–6/group). In each histogram, the bars represent measurements at day 10 (B. = baseline), day 13 (last day of FC or saline microinjections) and day 14 (co-administration of saline or FC with XE 991). Note the different axes for short and extended access groups. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Sidak’s post-hocs. Significant differences are only illustrated within each of the four behavioral groups for clarity. All statistically significant differences resulting from post-hoc comparisons can be found in Supplementary Fig. 4.