Joanna Peris, Kevin J. Anderson, Thomas W. Vickroy , Michael. A. King, Bruce E. Hunter and Don W. Walker

5. CET-INDUCED CHANGES IN GABA TRANSMISSION IN HIPPOCAMPUS

While the NMDA receptor serves as a critical trigger in the induction of LTP, alterations in NMDA receptor function appear to play little role in the expression of established LTP. Considerable controversy exists over the mechanism and site involved in the expression of LTP (70-72). Presynaptic GABA_B receptors are also proposed as a mechanism affecting LTP, since a decrease in GABA release may contribute to an increased NMDA response (73-75). During induction of LTP, GABA_A-mediated inhibition is decreased thereby allowing NMDA-mediated excitation to increase (75). Blockade of GABA_B receptors prevents the reduction in GABA inhibition, the increase in NMDA excitation and the induction of LTP (75).

It is possible that CET produces an enduring increase in pre- and/or postsynaptic elements of GABAergic synaptic transmission. This increase in GABA transmission could counteract the depolarizing effects of the LTP-induced NMDA receptor activation. This hypothesis is supported by the fact that the difference in LTP between CET and sucrose-control groups is abolished by bicuculline blockade of GABAergic synaptic transmission (45). These data indicate that an increase in postsynaptic GABA_A receptor activation is involved in CET inhibition of LTP.

The GABA_A receptor/chloride (Cl^-) ionophore is a hetero-oligomer composed of a total of 4-5 polypeptide subunits of at least thirteen different types (alpha_1-6, beta_1-3, gamma_1-3 and delta), each displaying a unique regional expression even within hippocampal subregions (76, 77). The gamma₂ subunit is necessary for benzodiazepine modulation of function (78) and the alternatively spliced gamma_2L form of this subunit must be appropriately phosphorylated by protein kinase C before ethanol sensitivity is conferred (79-81). In receptors that are sensitive to ethanol, there is an enhancement of GABA-stimulated Cl^- conductance (see 82). Subchronic ethanol exposure (5-10 days) results in a loss of in vitro enhancement of channel function by ethanol (83, 84). Although there has not been any consistent evidence for a change in the number or affinity of the GABA_A receptor as a whole (85, 86), subchronic ethanol exposure decreases both mRNA and peptide levels for alpha₁, alpha₂ and alpha₃ subunits in cortex, increases alpha₆ in cerebellum (87-92) and causes long-term increases in both mRNA and peptides for the beta₂ and beta₃ subunits in both regions (93). Thus, subchronic ethanol exposure changes the subunit composition and very likely, the functional status of the receptor. However, most of the changes described above are transient in nature, returning to control levels within 48 hrs after ethanol withdrawal.

CET appears to increase the number of [³H]bicuculline binding sites in hippocampus (94) which may indicate an increase in GABA_A receptor number or an alteration in receptor subunit composition. However, CET does not change the efficacy or potency of muscimol to activate postsynaptic GABA_A-controlled Cl^- ionophores nor is there a difference in the efficacy or potency of bicuculline to block this agonist stimulation (95). However, changes in the number of binding sites may not always be accompanied by a functional change if spare receptors are present. Therefore, the functional significance of the increase in [³H]bicuculline binding sites in hippocampal subregions of CET rats remains to be determined.

On the other hand, CET significantly increases electrically-stimulated [³H]GABA release from superfused hippocampal slices (95) which could also explain why bicuculline could abolish the effects of CET on LTP. It is not likely that the increase in GABA release is due to hyper innervation of pyramidal cells by GABAergic interneurons since previous studies have found that CET reduces functional inhibition of CA1 pyramidal cells as well as the number of GABAergic interneurons in CA1 (33-35). Instead the mechanism for an increase in GABA release seems to involve changes in GABA_B receptor-mediated presynaptic inhibition of GABA release (94). The effects of maximally effective doses of GABA_B receptor agents on [³H]GABA release are significantly decreased in CET rats (94). Results from binding experiments in similarly-treated rats suggest that CET has no overt effect on the number of GABA_B receptors in hippocampus which are predominantly postsynaptic in nature (94). In support of this, CET decreases the presynaptically-mediated hyperpolarizing responses of CA1 pyramidal cells to bath-applied baclofen (96) without affecting the postsynaptic effect of baclofen to inhibit EPSPs. Similar effects of CET on baclofen-induced inhibition of GABA release have been found in cortex (97) although there is no effect of CET on baclofen-induced inhibition of isoproterenol-stimulated cAMP formation (98).

GABA_B autoreceptors are capable of regulating induction of LTP via a decrease in GABA release thereby permitting sufficient NMDA receptor activation (99). On the other hand, postsynaptic GABA_B receptors can cause hyperpolarization of the postsynaptic membrane resulting in an enhanced blockade of NMDA channels by Mg²⁺ thereby inhibiting NMDA-mediated EPSPs (100). Thus, the involvement of GABA_B receptors in LTP may be quite complex since presynaptic disinhibition of GABA neurons would enhance LTP but postsynaptic activation of GABA_B receptors could diminish NMDA-mediated EPSPs and decrease LTP.