Restoration of Kv7 Channel-Mediated Inhibition Reduces Cued-Reinstatement of Cocaine Seeking

Abstract

Cocaine addicts display increased sensitivity to drug-associated cues, due in part to changes in the prelimbic prefrontal cortex (PL-PFC). The cellular mechanisms underlying cue-induced reinstatement of cocaine seeking remain unknown. Reinforcement learning for addictive drugs may produce persistent maladaptations in intrinsic excitability within sparse subsets of PFC pyramidal neurons. Using a model of relapse in male rats, we sampled >600 neurons to examine spike frequency adaptation (SFA) and afterhyperpolarizations (AHPs), two systems that attenuate low-frequency inputs to regulate neuronal synchronization. We observed that training to self-administer cocaine or nondrug (sucrose) reinforcers decreased SFA and AHPs in a subpopulation of PL-PFC neurons. Only with cocaine did the resulting hyperexcitability persist through extinction training and increase during reinstatement. In neurons with intact SFA, dopamine enhanced excitability by inhibiting Kv7 potassium channels that mediate SFA. However, dopamine effects were occluded in neurons from cocaine-experienced rats, where SFA and AHPs were reduced. Pharmacological stabilization of Kv7 channels with retigabine restored SFA and Kv7 channel function in neuroadapted cells. When microinjected bilaterally into the PL-PFC 10 min before reinstatement testing, retigabine reduced cue-induced reinstatement of cocaine seeking. Last, using cFos-GFP transgenic rats, we found that the loss of SFA correlated with the expression of cFos-GFP following both extinction and re-exposure to drug-associated cues. Together, these data suggest that cocaine self-administration desensitizes inhibitory Kv7 channels in a subpopulation of PL-PFC neurons. This subpopulation of neurons may represent a persistent neural ensemble responsible for driving drug seeking in response to cues.SIGNIFICANCE STATEMENT Long after the cessation of drug use, cues associated with cocaine still elicit drug-seeking behavior, in part by activation of the prelimbic prefrontal cortex (PL-PFC). The underlying cellular mechanisms governing these activated neurons remain unclear. Using a rat model of relapse to cocaine seeking, we identified a population of PL-PFC neurons that become hyperexcitable following chronic cocaine self-administration. These neurons show persistent loss of spike frequency adaptation, reduced afterhyperpolarizations, decreased sensitivity to dopamine, and reduced Kv7 channel-mediated inhibition. Stabilization of Kv7 channel function with retigabine normalized neuronal excitability, restored Kv7 channel currents, and reduced drug-seeking behavior when administered into the PL-PFC before reinstatement. These data highlight a persistent adaptation in a subset of PL-PFC neurons that may contribute to relapse vulnerability.

Keywords: Kv7 ion channels; afterhyperpolarization; cocaine; dopamine; prefrontal cortex; spike-frequency adaptation.

Figure 1.

Experimental outline for behavioral training and physiological measurements. A, Timeline shows the different behavioral phases, time points of electrophysiological measurements (e-phys), and the color-coded treatment groups. B, After behavioral end points, brain slices containing PL-PFC were made for current-clamp measurements of SFA (measured during the final 500 ms of the step) and AHPs generated in response to five APs (expressed as the area under the curve). C, In some experiments, after measuring firing to check for SFA with a 1 nA step, the I_Kv7 was determined in voltage clamp. Currents were evoked with voltage steps (−60 to −20 mV; 800 ms; in 200 nm TTX). Currents generated after 8 min of incubation of the Kv7 channel antagonist, XE-991 (XE; 20 μm, 8 min) was subtracted from control (predrug) currents. To determine the effect of dopamine (DA; 10 μm) or the Kv7 channel stabilizer retigabine (retig; 20 μm) on I_Kv7, the currents generated after bath application of DA + XE-991 (20 μm, 8 min) or retigabine + XE-991 (20 μm, 8 min) were subtracted from currents after DA or retigabine alone, respectively. sucr, Sucrose.

Figure 2.

Behavioral responding in cocaine self-administration, yoked-cocaine, and yoked-saline rats. A, Rats were presented with levers that resulted in the activation of a light and tone cue paired with the infusion of cocaine (cocaine-SA, n = 22; left), passive infusion of saline (saline-yoked, n = 39; middle), or passive infusion of cocaine (cocaine-yoked, n = 16; right). The vertical dashed line on time courses indicates the switch from self-administration to extinction. In extinction sessions, lever pressing produced no infusions or cue presentations. Behavioral responses for voluntary cocaine self-administration extinguished in 14 d. B, Re-exposure to drug-conditioned cues (without cocaine) significantly increased lever pressing in rats with a history of cocaine-SA. Numbers in italics represent the total number of animals. For these and all other figures, error bars indicate the mean ± SEM. ***p ≤ 0.0001 compared with extinction using a Sidak test for multiple comparisons.

Figure 3.

Experiment 1: training to self-administer cocaine (but not passive exposure to cocaine) suppressed intrinsic inhibition and inhibitory Kv7 channel currents. A, Timeline showing behavioral training ended 24 h before killing for electrophysiological recordings. B1, Scatter plots show firing responses of individual cells at 1 nA. Matching I–O curves (inset) show increased mean firing responses for the cocaine-SA relative to cocaine-yoked or saline-yoked treatment groups. Firing exceeded the upper limit for SFA (operationally defined as five or fewer spikes at 1 nA; dashed gray line) only in the cocaine-SA group. Black segments in the pie charts indicate the percentage of sampled cells that were SFA⁻. B2, Scatter plots show decreased average AHPs in the cocaine-SA group only. C1, I–O curves show reduced I_Kv7 currents in the cocaine-SA group. Before I_Kv7 measurements, cells in the cocaine-SA and saline-yoked groups were confirmed to be SFA⁻ and SFA⁺, respectively. C2, Reductions in I_Kv7 current amplitudes correlated with AP firing at the 1 nA step. D1, D2, Dot plot graphs show dopamine increased firing (D1) and decreased AHP (D2) in cocaine-SA/ SFA⁺ and saline-yoked/SFA⁺ cells. Dopamine produced no measurable change in cocaine-SA/SFA⁻ cells. E, Dopamine reduced I_Kv7 in the saline-yoked/SFA⁺ but not in the cocaine-SA/SFA⁻ cells. All insets show sample traces collected at 1 nA step for firing (I-clamp) or at −20 mV for Kv7 (V-clamp). Italicized numbers represent the number of cells/animals. Calibration: B1, 20 mV, 0.2 s; B2, 3 mV, 0.4 s; C1, 0.1 nA, 0.4 s; D1, 20 mV, 0.2 s; D2, 3 mV, 0.5 s; E, 0.1 nA, 0.4 s. *p ≤ 0.01 compared with saline-yoked animals using a Sidak test (B, C, E) or a Tukey's test (D1, D2) for multiple comparisons. DA, Dopamine; e-phys, electrophysiological measurements.

Figure 4.

Extinguishing the behavioral responding for cocaine does not restore intrinsic inhibition or Kv7 channel currents. A, Timeline showing cocaine-SA (14 d) and extinction (14 d) before killing for electrophysiological recordings at 15 min after the final extinction session. B1–B3, Average firing (B1, I–O curves) and individual firing (B2, scatter plots at 1 nA) were increased in the cocaine-SA group, while AHP (B3, scatter plots) was decreased relative to saline-yoked controls. Black segments in the pie charts indicate the percentage of sampled cells that were SFA⁻. C1, I–O curves show reduced I_Kv7 currents in the cocaine-SA/SFA⁻ cells, but not the saline-yoked/SFA⁺ or cocaine-SA/SFA⁺ cells. C2, Reductions in I_Kv7 current amplitudes correlated with AP firing at the 1 nA step. D1, D2, Dot plot graphs show that dopamine increased firing (D1) and decreased AHPs (D2) in SFA⁺ cells from both saline-yoked and cocaine-SA groups. Dopamine did not measurably alter cocaine-SA/SFA⁻ cells. E, Dopamine reduced I_Kv7 in SFA⁺ cells of saline-yoked and cocaine-SA rats, but not in SFA⁻ cells. All insets showing sample traces were collected at 1 nA step for firing (I-clamp) or at −20 mV for Kv7 (V-clamp). Italicized numbers represent the number of cells/animals. Calibration: B1, 20 mV, 0.2 s; B3, 3 mV, 0.4 s; C1, 0.1 nA, 0.4 s; D1, 20 mV, 0.2 s; D2, 3 mV, 0.5 s; E, 0.1 nA, 0.4 s. *p ≤ 0.001 compared with saline-yoked animals using a Sidak test (B, C, E) or a Tukey test (‡; D) for multiple comparisons. DA, Dopamine; e-phys, electrophysiological measurements.

Figure 5.

Differences in intrinsic inhibition following withdrawal from cocaine-SA and behavioral extinction from sucrose-SA. A, Top, Timeline showing cocaine-SA with 14 d of home-cage withdrawal, ending immediately before the preparation of brain slices for electrophysiological recordings (e-phys). Cells from rats withdrawn from cocaine-SA (without extinction training) showed elevated average firing beyond the maximal range for SFA (I–O curve with inset sample trace collected at 1 nA), heterogeneous individual firing responses (scatter plots at 1 nA), and reduced AHPs (inset, sample trace and scatter plots). B, Responses from unhandled, experimentally naive rats showed robust SFA across the I–O curve and homogeneous individual firing (scatter plots at 1 nA) and AHP (scatter plots) responses. C1, Timeline showing 14 d of sucrose-SA followed by extinction (14 d). Preparation of brain slices occurred at either 24 h after the last SA session (day 15) or 15 min after the final extinction session (day 29). C2, Rats were presented with levers that resulted in the activation of a light and tone cue paired with the delivery of a sucrose pellet. The vertical dashed line on time courses indicates the switch from self-administration to extinction. In extinction sessions, lever pressing had no consequence. Behavioral responses for sucrose extinguished in 14 d. C3–C5, Comparison of responses showing the increased excitability associated with sucrose-SA resolves with extinction, including average firing (C3, I–O curves), individual firing responses (C4, scatter plot, at 1 nA), and AHPs (C5, scatter plot). The black segments in the pie charts indicate the percentage of sampled cells that were SFA⁻. Italicized numbers indicate the number of cells/animal. Calibration: A, B, C, firing traces, 20 mV, 0.2 s; A, B, C, AHP traces, 3 mV, 0.4 s. *p ≤ 0.001 compared with sucrose-extinction using a Sidak test for multiple comparisons or unpaired t test.

Figure 6.

Cues that reinstate cocaine seeking rekindle the suppression of intrinsic inhibition: Restoration of inhibition by stabilization of Kv7 channels reduces cocaine seeking. A, Timeline showing the preparation of brain slices, occurring 15 min after the 2 h cued reinstatement test. Diagram shows proposed retigabine (retig) action on I_Kv7 in SFA⁻ cells during reinstatement. B1–B3, Comparison of responses showing cued reinstatement reduces intrinsic inhibition in cocaine-SA treatments relative to saline-yoked controls. Shown are average firing (B1, I–O curves), individual firing responses (B2, scatter plot, at 1 nA), and AHPs (B3, scatter plot). Black segments in pie charts below indicate the percentage of sampled cells that were SFA⁻. C, Relative to SFA⁺ (cocaine-SA or saline-yoked) cells, I_Kv7 currents are reduced in SFA⁻ (cocaine-SA) cells. D, E, Summary graphs showing retigabine (20 μm) reduced firing (D) and increased I_Kv7 (E) only in SFA⁻ (cocaine-SA) cells, and not in SFA⁺ (saline-yoked and cocaine-SA) cells. F, Bilateral microinjection of retigabine into PL-PFC 10 min before reinstatement testing resulted in a dose-dependent reduction of cue-induced reinstatement. G, Western blots of PL-PFC punch lysates showing Kv7 subunit expression (7.2, 7.3, 7.5) as a percentage of actin control did not differ between saline-yoked and cocaine-SA, either after extinction or after cued reinstatement. Calibration: B1, D, 20 mV, 0.2 s; B3, 3 mV, 0.4 s; C1, E, 0.1 nA, 0.4 s. All insets showing sample traces were collected at the 1 nA step for firing (I-clamp) or at −20 mV for Kv7 (V-clamp). Italicized numbers indicate the cells/animal. Sidak test for multiple comparisons: p ≤ 0.0001 compared with saline-yoked (*; B1, C) or preretigabine (‡; D); p ≤ 0.01 compared with preretigabine (†; E). F, ***p ≤ 0.0001, vs ext; #p = 0.0299, vehicle vs retig (300 μm). e-phys, Electrophysiological recordings; vehicle, veh.

Figure 7.

Characterization of intrinsic inhibition in cFos-GFP⁺ and cFos-GFP⁻ cells in the PL-PFC after extinction and cued reinstatement of cocaine seeking. A, Schematic of the transgene containing a c-fos promoter that induces GFP in strongly activated Fos-expressing neurons, which can be sampled in electrophysiological slice preparations. B, Lever-pressing behavior for Long–Evans rats providing the PL-PFC electrophysiology recordings and immunohistochemistry. Rats were presented with levers that resulted in the activation of a light and tone cue paired with the infusion of cocaine (cocaine-SA, right) or yoked passive infusion of saline (saline-yoked, left). The vertical dashed line on the time courses indicates the switch from SA to Ext. In Ext sessions, lever pressing had no consequence. C, Re-exposure to cues (without saline or rewards) significantly increased lever pressing in rats with a history of cocaine-SA (Sidak post hoc; ***p ≤ 0.0001, cue vs extinction). D1, Confocal photomicrographs showing colabeling (white arrows) of GFP (green) and Fos (red) using immunofluorescence histochemistry. Micrographs are from a c-Fos-GFP rat that underwent cocaine-SA and a single Ext training session before perfusion (day 15). D2, Table summarizing GFP/Fos colabeling in transgenic rats with different treatments: Rat 1, saline-yoked+14 d EXT(killing on day 28); Rat 2-cocaine-SA (killing on day 14); Rat 3 cocaine-SA+1 d Ext (killing on day 15). A high degree of colocalization between fluorophores was evident from the Pearson's correlation coefficient (F_P; summarizing the ratio of cells that colocalize) and Manders' overlap coefficient (r; showing the overlap between channels). E1, Confocal photomicrographs show similarly low levels of GFP-IR (open arrowheads) in the PFC of a saline-yoked (left) and cocaine-SA (right) rat immediately after the final Ext session (day 28). E2, Photomicrographs of pyramidal neurons in a living brain slice from a cFos-GFP rat killed after the final Ext session (day 28). Cells were visualized in the electrophysiology recording chamber by contrast (Dodt) optics, and overlayed with pseudocolor fluorescence to show a GFP⁺ (open arrowhead) and a GFP⁻ (closed arrowhead) nucleus. Star indicates the recording electrode. All subjects were killed immediately following their final 90 min behavioral session. F, G, Comparison of responses from GFP⁺ and GFP⁻ cells from cocaine-SA-treated (red) and saline-yoked-treated (blue) rats following Ext (F) or cue reinstatement (G). Closed and open symbols denote responses in GFP⁺ and GFP⁻ cells, respectively. Measurements included average firing (inset, I–O curves), individual firing (scatter plots of responses at 1 nA), and AHPs (scatter plots). Where possible, equal numbers of GFP⁺/GFP⁻ cells in animals were sampled. Note the elevated firing and reduced AHPs only in a subset of GFP⁺ cells from cocaine-SA rats, but not from saline-yoked rats. Italicized numbers in B indicate the number of animals, whereas those in C, F, and G indicate the number of cells/animal. Tukey multiple comparison post-test: ***p ≤ 0.0001 (current × treatment and current × GFP^+/−) comparing I–O (insets) after extinction or cue; **p = 0.0218 comparing AHPs from GFP⁺ and GFP⁻ cells from the cocaine-SA group after extinction; *p ≤ 0.05 effect of treatment comparing AHPs after cue.