Corticotropin releasing factor-induced amygdala gamma-aminobutyric Acid release plays a key role in alcohol dependence

Abstract

Background: Corticotropin-releasing factor (CRF) and gamma-aminobutyric acid (GABA)ergic systems in the central amygdala (CeA) are implicated in the high-anxiety, high-drinking profile associated with ethanol dependence. Ethanol augments CeA GABA release in ethanol-naive rats and mice.

Methods: Using naive and ethanol-dependent rats, we compared electrophysiologic effects and interactions of CRF and ethanol on CeA GABAergic transmission, and we measured GABA dialyzate in CeA after injection of CRF(1) antagonists and ethanol. We also compared mRNA expression in CeA for CRF and CRF(1) using real-time polymerase chain reaction. We assessed effects of chronic treatment with a CRF(1) antagonist on withdrawal-induced increases in alcohol consumption in dependent rats.

Results: CRF and ethanol augmented CeA GABAergic transmission in naive rats via increased GABA release. Three CRF1 receptor (CRF(1)) antagonists decreased basal GABAergic responses and abolished ethanol effects. Ethanol-dependent rats exhibited heightened sensitivity to CRF and CRF(1) antagonists on CeA GABA release. Intra-CeA CRF(1) antagonist administration reversed dependence-related elevations in GABA dialysate and blocked ethanol-induced increases in GABA dialyzate in both dependent and naive rats. Polymerase chain reaction studies indicate increased expression of CRF and CRF(1) in CeA of dependent rats. Chronic CRF(1) antagonist treatment blocked withdrawal-induced increases in alcohol drinking by dependent rats and tempered moderate increases in alcohol consumption by nondependent rats in intermittent testing.

Conclusions: These combined findings suggest a key role for specific presynaptic CRF-GABA interactions in CeA in the development and maintenance of ethanol dependence.

Figure 1

Corticotropin releasing factor (CRF) dose-dependently increases gamma-aminobutyric acid (GABA)ergic transmission in rat central amygdala (CeA) neurons, and the maximal CRF effect is increased in CeA neurons from dependent rats. (A) Black line, filled squares: concentration–response data from 4 to 7 CeA neurons from naive rats for each point, representing percent increase in mean (± SEM) inhibitory postsynaptic currents (IPSC) peak amplitudes plotted against a log scale of CRF concentration. The black solid curve is a logistic fit of the data plotted by origin software (OriginLab Corporation, Northampton, Massachusetts), using y = (A1 − A2)/[1 + (x/xo)p + A2, where A1 is the initial value of IPSC increase (5%), A2 the estimated final maximum value (49%), xo is the center × value (77 nmol/L; unfixed), and p is the rate or power (2.3). All values were fixed except center. Dashed lines: apparent EC₅₀ for CRF. Gray line, open squares: concentration–response data from 3 to 9 CeA neurons from ethanol-dependent rats; each point represents percent change in mean (± SEM) IPSC peak amplitude pooled and plotted versus log of CRF concentration. The gray solid curve is a logistic fit, where A1 is the initial value of IPSC change (−3%), A2 the estimated final maximum value (66%), xo is the center × value (82 nmol/L; unfixed), and p is the rate or power (3.7). All values fixed except center. Dashed lines: apparent EC₅₀ for CRF. The logistic fit here is a simulated because the highest CRF concentration (400 nmol/L) actually elicits a weaker IPSC effect than 200 nmol/L CRF. * Indicates significance (p < .05) between the two groups. (B) Time course of CRF on IPSPCs. Top panel: representative IPSCs in a CeA neuron from a naive rat recorded before and during 200 nmol/L CRF (20 min) and washout (30 min). Bottom panel: averaged responses obtained from 7 AA neurons. CRF significantly increased mean IPSC amplitudes with partial recover on washout. (C) IPSC input-output (I-O) curves: CRF (200 nmol/L) superfusion (at 10–15 min) significantly (p < .05) increased the mean evoked IPSCs over three stimulus intensities in CeA from naive rats. Top insert: representative recordings of evoked IPSCs. (D) I-O curves: CRF (200 nmol/L, 10- to 15-min application) superfusion significantly (p < .05) increased the mean evoked IPSCs over all three stimulus strengths, with recovery on washout in CeA from ethanol-dependent rats. Top insert: representative recordings of evoked IPSCs.

Figure 2

Corticotropin releasing factor (CRF) augments inhibitory postsynaptic currents (IPSCs) in CeA via increased gamma-aminobutyric acid release. (A) Two hundred nanomolar CRF (12- to 15-min application) significantly (*p < .05) decreased the mean PPF of IPSCs. (B) CRF also significantly (*p < .05) decreased the mean paired-pulse facilitation (PPF) of IPSCs in central amygdala (CeA) neurons from ethanol-dependent rats. (C) Top: representative recordings of miniature IPSCs (mIPSCs) from a CeA neuron from a naive rat. Bottom: CRF significantly (*p < .05) increased the mean frequency but not the amplitude of the mIPSCs in CeA of naive rats. Top insert: representative recordings of mIPSCs. (D) In CeA of dependent rats, CRF significantly (*p < .001) increased (to 200.9 ± 17% of control) the mean mIPSC frequency but not the mean amplitude of mIPSCs in CeA of ethanol-dependent rats. Top insert: representative recordings of mIPSCs.

Figure 3

Corticotropin releasing factor-1 (CRF₁) antagonists block the ethanol-induced increase of evoked inhibitory postsynaptic currents (IPSCs) in central amygdala (CeA) neurons of naive rats. (A) Ethanol (44 mmol/L, 10- to 12-min application) significantly (*p < .05) increases evoked IPSCs with recovery in washout. (B) The CRF1 antagonists slightly decrease the baseline gamma-aminobutyric acid (GABA)ergic transmission and prevent the ethanol-induced increase of IPSCs in naive CeA neurons from naive rats. Representative evoked IPSCs recorded before and during superfusion of CRF₁ antagonists 10 μmol/L antalarmin (Anta), 10 μmol/L NIH3 and 1 μmol/L R121919) and during and after the coapplication of the CRF₁ antagonists with ethanol and washout. (C) Pooled data of the experiments from Figure 3B: superfusion of CRF¹ antagonists prevented the ethanol-induced increase of IPSCs in naive CeA neurons. (D) IPSC recordings from CeA of dependent rats. As in CeA of naive rats, ethanol (44 mmol/L, 10-to 12-min application) significantly (*p < .05) augments evoked IPSCs with recovery on washout. (E) The CRF₁ antagonists significantly decrease basal GABA transmission and block ethanol effects in CeA neurons from ethanol-dependent rats. Representative recordings of evoked IPSCs recorded before and during superfusion of CRF₁ antagonists, during coapplication of the CRF₁ antagonists and ethanol, and washout. (F) Pooled data of the experiments in Figure 3E: superfusion of the three CRF₁ antagonists significantly (*p < .05) decreased the mean amplitudes of evoked IPSCs and also blocked the ethanol-induced augmentation of IPSCs. * Indicates that the CRF₁ antagonist effects in ethanol-dependent rats are significantly larger (p < .05) than the antagonist effects in naive rats (Panel F vs. Panel C).

Figure 4

Acute and chronic ethanol increase dialysate levels of gamma-aminobutyric acid (GABA) in rat central amygdala (CeA) in vivo. (A) In ethanol-dependent rats, the CRF₁ antagonist R121919 administration into the CeA significantly (**p < .005) decreased mean local dialysate GABA levels. In both naive and ethanol-dependent rats, R121919 administration significantly (*p < .05) blocked the ethanol-induced increase in GABA release. Furthermore, the mean baseline dialysate GABA level was significantly (#p < .001) increased in ethanol-dependent rats, compared with naive rats. (B) In ethanol-dependent (n = 8) rats, the levels of CRF mRNA, normalized to cyclophilin A, were significantly increased in CeA punches (*p < .05), compared with naive (n = 11) controls, as measured by quantitative real-time polymerase chain reaction.

Figure 5

(A) Chronic R121919 treatment blocks the development of alcohol dependence-induced increases in alcohol consumption (g/kg) and also blocks moderate increases in alcohol drinking by nondependent rats over time. Rats were injected with R121919 (10 mg/kg subcutaneous) or vehicle on even-numbered days of vapor exposure (indicated by arrows), and tested for responding at 6- to 8-hour withdrawal on Days 3, 7, 11, 15, 19, and 23 of vapor exposure. (B) Cumulative ethanol intake (g/kg) across all operant test sessions during the vapor exposure period. * Indicates significance (p < .05) suppression by R121919 relative to vehicle, and # indicates significantly (p < .05) higher ethanol intake by dependent rats relative to nondependent controls.