Report

Ethanol Augments GABAergic Transmission in the Central Amygdala via CRF1 Receptors

See allHide authors and affiliations

Science  05 Mar 2004:
Vol. 303, Issue 5663, pp. 1512-1514
DOI: 10.1126/science.1092550

Abstract

The central amygdala (CeA) plays a role in the relationship among stress, corticotropin-releasing factor (CRF), and alcohol abuse. In whole-cell recordings, both CRF and ethanol enhanced γ-aminobutyric acid–mediated (GABAergic) neurotransmission in CeA neurons from wild-type and CRF2 receptor knockout mice, but not CRF1 receptor knockout mice. CRF1 (but not CRF2) receptor antagonists blocked both CRF and ethanol effects in wild-type mice. These data indicate that CRF1 receptors mediate ethanol enhancement of GABAergic synaptic transmission in the CeA, and they suggest a cellular mechanism underlying involvement of CRF in ethanol's behavioral and motivational effects.

CRF and its family members are implicated in stress-related disorders such as anxiety and depression (1). Generally, CRF plays biologically diverse roles in generating stress responses by acting at CRF1 and CRF2 receptors (CRFR1s, CRFR2s) that are differentially expressed in diverse areas of the central nervous system (CNS) (24). Overactivation of CRFR1s induces a hormonal and behavioral stress-like response (2). Mice lacking CRFR1s display reduced anxiety (5), and selective CRFR1 antagonists inhibit the anxiogenic action of CRF (2). Hyperactivation of the CRF system through CRFR1 is implicated in several stress-related psychiatric disorders such as alcoholism (6, 7) and is a current target of medication development for affective disorders (8). Certainly, stressful life events and maladaptive responses influence alcohol drinking and relapse behavior (9, 10). CRF has been strongly linked to behavioral changes associated with alcohol withdrawal in laboratory animals (11, 12).

Stress can increase alcohol drinking in humans, perhaps as an attempt to cope with this state (13). However, the molecular and cellular mechanisms underlying stress-induced alcohol drinking are still unknown. The central amygdala (CeA) is a brain region prominently involved in alcohol dependence and reinforcement (14, 15) and contains an abundance of CRF and its receptors. In the CeA, CRF is localized and cosynthesized within GABAergic neurons (16, 17), which suggests that CRF may have a modulatory role on GABAergic intrinsic circuits of the CeA. Injection of a GABAA receptor agonist into the amygdala decreases ethanol self-administration in dependent rats (18), and CRF systems in the amygdala may be activated during ethanol withdrawal (11, 19). Ethanol enhances GABAergic transmission at both pre- and postsynaptic sites in rat CeA (20). This suggests that the CeA is an important area involved in alcohol reward and dependence, and that CRF and GABA play roles in these phenomena. To assess a possible cellular role and mechanism of CRF in the interaction of ethanol with the GABAergic system in the CeA, we used whole-cell recording of neurons in slices of mouse CeA.

We prepared amygdala slices as described (20) from CRFR1/CRFR2 knockout (KO) mice and their wild-type littermates on a mixed C57BL/6J × 129 background, as well as C57BL/6J mice (21). Because both the CRF and ethanol effects were equivalent in wild-type littermates and C57BL/6J mice, we have pooled the summary data for these two groups (hereafter, WT). GABAA inhibitory postsynaptic currents (IPSCs) were isolated in all experiments by pharmacological blockade of NMDA (N-methyl-D-aspartate), non-NMDA, and GABAB receptors (21). We first studied the effect of CRF on IPSCs in CeA slices by superfusing 50 or 100 nM CRF. In neurons of WT mice, application of 50 nM CRF for 8 to 16 min enhanced IPSC amplitudes to 128 ± 12% of control (n = 5). At a concentration of 100 nM, CRF significantly [F(1,14) = 6.72, P < 0.05; n = 8] enhanced IPSC amplitudes to 139 ± 8% of control (range 130 to 146%; Fig. 1, A and C). We then assessed the effects of ethanol on IPSCs. In neurons from WT mice, superfusion of 44 mM ethanol significantly [F(1,12) = 6.91, P < 0.05; n = 7] enhanced IPSCs to 139 ± 13% of control (range 131 to 142%; Fig. 1, D and F). Neither CRF nor ethanol affected holding currents. We tested different concentrations of ethanol (5 to 66 mM) on IPSCs in WT mice. Whereas 5 mM had no effect (n = 6), 11 mM increased IPSC amplitudes to 116% of control (n = 6), 22 mM to 131% (n = 5), 44 mM to 139% (n=7), and 66 mM to 135% (n = 6). As we reported previously (20, 22), the highest ethanol concentration of 66 mM actually enhanced IPSCs to a lesser extent than did 44 mM ethanol. Therefore, we used 44 mM ethanol in subsequent experiments.

Fig. 1.

Both CRF and ethanol enhance GABAA IPSCs in CeA neurons from WT mice, but not those of CRFR1 KO mice. (A) 100 nM CRF enhanced IPSCs to 146% of control in this neuron from a WT littermate mouse. (B) Superfusion of 100 nM CRF had no effect on IPSCs from a CRFR1 KO mouse. (C) Pooled data of the effect of CRF on the mean amplitudes of GABAA IPSCs. 100 nM CRF significantly enhanced IPSCs to 139 ± 8% of control in neurons from WT mice (n = 8) but had no significant effect on IPSC amplitudes in neurons from CRFR1 KO mice (n = 6). (D) Superfusion of 44 mM ethanol enhanced IPSCs to 136% of control in this neuron from a WT littermate mouse. (E) Ethanol had no effect on IPSCs from a CRFR1 KO mouse. (F) Pooled data of the effect of 44 mM ethanol on mean GABAA IPSC amplitudes. Ethanol significantly enhanced IPSCs to 137 ± 13% of control in neurons from WT mice (n = 7) but had no significant effect in those from CRFR1 KO mice (n = 11).

In contrast to neurons of WT mice, 100 nM CRF had no significant (P > 0.1) effect on IPSC amplitudes in neurons of CRFR1 KO mice (n = 6; Fig. 1, B and C). In another 11 neurons of CRFR1 KO mice, 44 mM ethanol also had no significant (p > 0.1) effect on IPSC amplitudes (Fig. 1, E and F). By contrast, in neurons from CRFR2 KO mice, 44 mM ethanol enhanced IPSCs to 137 to 142% of control (n = 3). In another three neurons from CRFR2 KO mice, 100 nM CRF also enhanced IPSCs to 134 to 144% of control. There was no significant (p > 0.1) difference in the amplitude or shape of baseline IPSCs evoked from neurons of WT mice compared with those of CRFR1/CRFR2 KO mice (Fig. 1).

To further identify the CRF receptors implicated in the ethanol enhancement of IPSCs, we used CRFR1/CRFR2 antagonists. In seven neurons from WT mice, superfusion of the peptide CRF receptor antagonist D-Phe-CRF12-41 (200 nM) completely blocked the effect of 44 mM ethanol (Fig. 2, A and D). D-Phe-CRF12-41 alone slightly enhanced IPSC amplitudes, but the difference was not significant (P > 0.1). In another six neurons from WT mice, D-Phe-CRF12-41 also blocked the effect of 100 nM CRF (Fig. 2, C and D). The nonpeptide CRFR1 antagonist NIH-3 (LWH-63; 10 μM) (23) also totally blocked the ethanol effects in seven neurons from WT mice (Fig. 2B). In another three neurons from WT mice, NIH-3 also blocked the CRF effect. We also tested the selective nonpeptide CRFR2 antagonist astressin2-B on the ethanol and CRF effects. In WT mice, 3 μM astressin2-B failed to alter the ethanol effect on IPSCs (n = 5; Fig. 2D). In another three neurons from WT mice, astressin2-B also did not alter the CRF effect (Fig. 2, C and D). Astressin2-B alone did not affect IPSCs (Fig. 2, C and D). These data further suggest that CRFR2s are not involved in ethanol enhancement of GABAAergic synaptic transmission in the CeA.

Fig. 2.

CRFR1, but not CRFR2, antagonists block the ethanol and CRF effects. (A) Superfusion of the peptide CRF antagonist D-Phe-CRF12-41 (D-Phe, 200 nM) alone slightly enhanced IPSCs but prevented ethanol enhancement of the IPSCs. After washout of the antagonist, 44 mM ethanol enhanced the IPSCs; the IPSCs were subsequently obliterated by the GABAA receptor blocker bicuculline (Bic, 30 μM). (B) Ethanol (44 mM) enhanced IPSCs in this neuron, with recovery on washout. Subsequent addition of 10 μM NIH-3 (the selective nonpeptide CRFR1 antagonist) did not affect IPSCs but totally blocked ethanol effects. (C) D-Phe-CRF12-41 (200 nM) prevented the CRF effect, but 3 μM astressin2-B (Ast2-B, the selective nonpeptide CRFR2 antagonist) did not alter CRF effects. (D) Pooled data of the effect of D-Phe-CRF12-41 and astressin2-B on GABAA IPSCs in neurons from WT mice. D-Phe-CRF12-41 (200 nM) alone slightly enhanced mean IPSC amplitudes, although the effect was not significant (P > 0.1, n = 6). However, D-Phe-CRF12-41 completely blocked the ethanol and CRF effects. Astressin2-B (3 μM) had no effect on ethanol (44 mM; n = 5) or CRF (100 nM; n = 5) enhancement of IPSC amplitudes in CeA from WT mice.

Because ethanol or CRF could act at either pre- or postsynaptic sites to enhance IPSCs (20), we examined paired-pulse facilitation (PPF) of IPSCs by using an interstimulus interval of 50 ms. A reduction of PPF is associated with an increased probability of release (24). In seven neurons from WT mice, both 100 nM CRF and 44 mM ethanol significantly decreased the PPF ratio of IPSCs (Fig. 3, A and B). These data, combined with findings that both CRF (25) and ethanol (20) enhance the frequency of spontaneous IPSCs and inhibitory postsynaptic potentials in rat CeA slices, suggest that CRF increases GABA release from presynaptic sites as ethanol does in rat CeA.

Fig. 3.

Both CRF and ethanol decrease the PPF ratio of GABAA IPSCs (paired-pulse IPSCs evoked with interstimulus intervals of 50 ms). (A) Superfusion of 44 mM ethanol decreased the PPF ratio of IPSCs (IPSC2 to IPSC1) with recovery on washout and block of the IPSCs by 30 μM bicuculline. (B) 100 nM CRF also decreased the PPF ratio of IPSCs with recovery on washout. (C) Pooled data of the effect of 44 mM ethanol (n = 6) and 100 nM CRF (n = 7) on mean PPF ratios of IPSCs. PPF ratio is expressed as percentage of mean baseline value (100%). The asterisk denotes statistical significance (P < 0.05, analysis of variance).

Our data indicate that the CRF system and CRFR1s are involved in the ethanol enhancement of GABAergic transmission in CeA neurons. Anxiolytic or intoxicating concentrations (11 to 44 mM) of ethanol enhance GABAergic synaptic transmission in the CeA of WT or CRFR2 KO mice but not of CRFR1 KO mice. Furthermore, both a peptide CRF receptor antagonist and a selective nonpeptide CRFR1 antagonist completely blocked the ethanol enhancement of GABAergic neurotransmission in WT mice, whereas a CRFR2 antagonist had no effect. Because CRF is relatively selective for CRFR1 over CRFR2 (2), and both CRF and ethanol enhanced IPSCs in WT and CRFR2 KO mice but not in CRFR1 KO mice, we conclude that the enhancement of IPSCs in the CeA by CRF and ethanol involves CRFR1s. Our PPF data suggest that the CRFR1s involved in this effect are probably located on presynaptic terminals of GABAergic neurons. Because CRF is colocalized with GABA in inhibitory interneurons of the CeA (17), CRF1Rs may play a role as positive feedback “autoreceptors” enhancing GABA release. Ethanol may then augment this autoinhibition, perhaps ultimately leading to excitation by disinhibiting downstream neurons.

GABA is the major inhibitory neurotransmitter in the CNS. Enhancement of the activation of GABAA receptors is a common feature of many sedative and hypnotic drugs (13, 26). Behavioral and electrophysiological studies have shown that the GABA system is an important target for ethanol action in the CNS, especially in the CeA (18, 20). Ethanol can enhance GABAergic transmission by acting at presynaptic sites in rat CeA neurons (20). Here, we found that ethanol significantly enhanced GABAA IPSCs in CeA neurons of WT mice, consistent with data from the rat. Thus, the GABAergic synapse may be an important target for ethanol action in the CeA across species.

Recently, several studies have shown that stress-induced alcohol drinking and relapse behavior have a significant genetic component (2729). Although our data pertain only to nondependent animals, growing evidence points to an interaction among stress, brain CRF, and alcohol drinking, and both CRF and GABAergic neurotransmission are facilitated during the development of alcohol dependence (6, 7, 12, 19). Our combined findings indicate that these two neuropharmacological systems are linked and suggest that CRFR1s mediate ethanol enhancement of GABAA IPSCs in CeA, thus providing a cellular mechanism for the reported involvement of CRF in ethanol's behavioral effects. These findings also support the hypothesis that CRF plays a crucial role in the motivational effects of ethanol. Thus, as previously suggested (11), CRFR1s could represent an important therapeutic target for the treatment of stress-induced alcohol drinking.

Supporting Online Material

www.sciencemag.org/cgi/content/full/303/5663/1512/DC1

Materials and Methods

References and Notes

View Abstract

Navigate This Article