Weiner, J. L., C. Gu, and T. V. Dunwiddie. Differential ethanol sensitivity of subpopulations of GABAA synapses onto rat hippocampal CA1 pyramidal neurons. J. Neurophysiol. 77: 1306–1312, 1997. The actions of ethanol on γ-aminobutyric acid-A (GABAA) receptor-mediated synaptic transmission in rat hippocampal CA1 neurons remain controversial. Recent studies have reported that intoxicating concentrations of ethanol (10–100 mM) can potentiate, inhibit, or have no effect on GABAA receptor-mediated synaptic responses in this brain region. The essential determinants of ethanol sensitivity have not been defined; however, GABAA receptor subunit composition, as well as posttranslational modifications of these receptors, have been suggested as important factors in conferring ethanol sensitivity to the GABAA receptor complex. Multiple types of GABAA receptor-mediated synaptic responses have been described within individual hippocampal CA1 neurons. These responses have been shown to differ in some of their physiological and pharmacological properties. In the present study we tested the hypothesis that some of the disparate findings concerning the effects of ethanol may have resulted from differences in the ethanol sensitivity of GABAA receptor-mediated synapses on single CA1 pyramidal cells. Electrical stimulation adjacent to the stratum pyramidale (proximal) and within the stratum lacunosum-moleculare (distal) activated nonoverlapping populations of GABAA receptors on rat hippocampal CA1 neurons. Proximal inhibitory postsynaptic currents (IPSCs) decayed with a single time constant and were significantly potentiated by ethanol at all concentrations tested (40, 80, and 160 mM). Distal IPSCs had slower decay rates that were often described better by the sum of two exponentials and were significantly less sensitive to ethanol at all concentrations tested. Three other allosteric modulators of GABAA receptor function with well-defined GABAA receptor subunit requirements, pentobarbital, flunitrazepam, and zolpidem, potentiated proximal and distal GABAA IPSCs to the same extent. These results demonstrate that the ethanol sensitivity of GABAA receptors can differ, not only between brain regions but within single neurons. These findings offer a possible explanation for the conflicting results of previous studies on ethanol modulation of GABAA receptor-mediated synaptic transmission in rat hippocampal CA1 neurons.
The γ-aminobutyric acid-A (GABAA) receptor mediates the majority of fast inhibitory synaptic transmission in the mammalian CNS (Krnjevic 1991; Thompson 1994). This heteromeric complex is an important target for a variety of sedative-hypnotic drugs such as benzodiazepines and barbiturates (Macdonald and Olsen 1994) that allosterically modulate GABAA receptor function. A number of reports have suggested that ethanol intoxication may also be mediated, in part, via modulation of GABAA receptor activity. In support of this hypothesis, both biochemical (Allan and Harris 1987; Suzdak et al. 1986; Ticku et al. 1986) and electrophysiological (Aguayo 1990; Celantano et al. 1988; Reynolds and Prasad 1991) studies have demonstrated that intoxicating concentrations of ethanol (10–100 mM) can potentiate GABAA receptor function. However, in a number of studies in which similar methods were used, this interaction was not observed (Mihic et al. 1992; Morelli et al. 1988; Osmanovic and Shefner 1990). In fact, it has proven difficult to identify a specific neuronal population or receptor activity measure that can be used to reliably demonstrate the ethanol/GABAA receptor interaction. Although the factors underlying the variability of this interaction are unclear, recent studies have suggested that factors such as GABAA receptor subunit composition, phosphorylation state of the receptor, and the methods for GABA and ethanol application may play important roles. For example, Xenopus oocyte and mammalian cell recombinant studies have demonstrated that a long variant of the γ2 subunit (γ2L) is required to confer ethanol sensitivity to heteromeric GABAA receptors (Harris et al. 1995; Wafford and Whiting 1992; Wafford et al. 1990, 1991). This finding has not been replicated in all studies (Kleingoor et al. 1991; Sigel et al. 1993), suggesting that the presence of the γ2L subunit alone is not sufficient to confer significant ethanol sensitivity to the GABAA receptor complex. Differences in the subunit composition of GABAA receptors cannot readily explain conflicting results obtained from the same neuronal population. For example, hippocampal CA1 neurons have been shown to contain both γ2L mRNA (Zahniser et al. 1992) and protein (Miralles et al. 1994; Ruano et al. 1994); however, intoxicating concentrations of ethanol have been reported to potentiate (Wan et al. 1996; Weiner et al. 1994), inhibit (Siggins et al. 1987) or have no effect on (Proctor et al. 1992) GABAA receptor-mediated synaptic responses in this brain region.
Hippocampal pyramidal neurons, which express mRNA for ≥12 different GABAA receptor subunits (Wisden et al. 1992), also have GABAA responses that differ in terms of their kinetics, reversal potentials, and pharmacological properties (Alger and Nicoll 1982; Pearce 1993, 1996; Pearce et al. 1995). Because multiple GABAA receptor subtypes can be studied simultaneously in the same cell, this provides an excellent test system in which to determine the relative importance of such factors as receptor subunit composition and phosphorylation, as opposed to procedural factors that may differ between laboratories but that may also be important in determining ethanol sensitivity (Weiner et al. 1995). One approach that appears to differentially activate subpopulations of GABAA receptors in hippocampal pyramidal neurons is to electrically stimulate somatic and dendritic regions of the hippocampal CA1 region; such stimulation evokes GABAA receptor-mediated inhibitory postsynaptic currents (IPSCs) with distinct physiological and pharmacological properties (Pearce 1993, 1996; Pearce et al. 1995). In the present study we determined whether a number of putative GABAA modulators, including ethanol, had differential effects on GABAA receptor-mediated responses evoked by these two kinds of stimulation.
Transverse hippocampal slices (400 μm) were prepared from 4- to 6-wk-old male Sprague-Dawley rats with the use of a McIlwain tissue chopper. Slices were incubated for ≥2 h before recording in a submersion chamber at 32°C in artificial cerebrospinal fluid containing (in mM) 126 NaCl, 3 KCl, 1.5 MgCl2, 2.4 CaCl2, 1.2 NaH2PO4, 11 glucose, and 26 NaHCO3, saturated with 95% O2-5% CO2.
Electrophysiological recordings and drugs
Slices were transferred to a submersion recording chamber and superfused with aerated artificial cerebrospinal fluid at 2 ml/min. Whole cell patch-clamp recordings were made from CA1 neurons at 31–33°C with the use of an Axoclamp-2A amplifier operating in the continuous voltage-clamp mode. Cells were not included in this study if their membrane potential was more depolarized than −60 mV, if their access resistance was >40 MΩ, or if this value changed by >20% over the course of the recording. Cells were voltage clamped near −55 mV to increase the chloride driving force and permit the use of lower stimulation intensities. Recording electrodes were constructed from thin-walled borosilicate glass (1.5 mm diam, Sutter Instrument, Novato, CA) and had resistances of 6–9 MΩ. The patch pipette solution contained (in mM) 125 potassium gluconate (Fluka, Buchs, Switzerland), 15 KCl, 10 N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (Fluka), 0.1 CaCl2, 1 potassium ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid (Fluka), 2 Mg-ATP, and 0.2 tris(hydroxymethyl)aminomethane guanosine 5′-triphosphate, pH adjusted to 7.25 with KOH, 290 ± 5 mosM, and was kept on ice until immediately before use. All drugs were purchased from Sigma (St. Louis, MO) unless otherwise indicated and were added to the artificial cerebrospinal fluid in known concentrations immediately before administration to the slice recording chamber via calibrated syringe pumps (Razel, Stamford, CT): dl(−)-2-amino-5-phosphonovaleric acid; 6,7-dinitroquinoxaline-2,3-dione; bicuculline methiodide; pentobarbital sodium; flunitrazepam (RBI, Natick, MA); and zolpidem (RBI). A 2 M solution of ethanol (diluted in deionized water) was prepared immediately before each experiment from a 95% stock solution kept in a glass storage bottle (Aaper, Shelbyville, KT).
Proximal and distal GABAA IPSCs
Evoked GABAA IPSCs were pharmacologically isolated by superfusion with the excitatory amino acid antagonists dl(−)-2-amino-5-phosphonovaleric acid (50 μM) and 6,7-dinitroquinoxaline-2,3-dione (20 μM). Synaptic stimulation was delivered with the use of two bipolar twisted tungsten wire electrodes (0.2-ms pulses of 5–20 V) placed in the stratum lacunosum-moleculare (distal stimulation) and within 250 μm of the recording pipette, near the CA1 somatic region (proximal stimulation), with an interstimulus interval of 20 s. Stimulation intensities were adjusted to evoke proximal and distal IPSCs of similar amplitude with no detectable GABAB receptor-mediated component. Drug effects were quantified as the percent change in amplitude or area under the curve of IPSCs following drug application relative to the mean of control and washout values. Statistical analyses were carried out with the use of two-tailed Student's paired and unpaired t-tests as indicated, with a minimal level of significance of P < 0.05. Decay time constants were calculated with the use of Prism software (Graphpad, San Diego, CA). Proximal and distal IPSC decays were first fit to a single exponential. This fit was accepted unless the sum of squared errors (R 2 value) was <0.95. For all such currents, addition of a second exponential improved the fit such that the R 2 value was >0.95.
Properties of proximal and distal GABAA IPSCs
We recorded from 56 CA1 pyramidal neurons with an average resting membrane potential of −68.6 ± 3.6 mV and input resistance of 136.5 ± 5.8 MΩ (measured at −55 mV). Proximal stimulation evoked GABAA IPSCs that decayed with a single time constant in 54 of 56 neurons analyzed (mean time constant = 59.3 ± 4.8 ms). However, there was considerable variability in the duration of these responses (range = 13.2–136.2 ms). Of the 56 distal IPSCs examined, 30 were fit well by a single decay time constant (mean time constant = 87.3 ± 7.1 ms) that was significantly slower than that of proximal IPSCs (unpaired t-test, P < 0.01). The decay phase of the remaining 26 distal IPSCs was better described by the sum of two exponentials (fast time constant = 56.5 ± 9.7 ms; slow time constant = 241.7 ± 14.9 ms). Despite the slow decay time constant of many distal and some proximal synaptic currents, both types of IPSCs were blocked essentially completely by 20 μM bicuculline methiodide, and therefore did not contain a significantGABAB component (Fig. 1). Proximal and distal IPSCs also exhibited similar reversal potentials (−70.3 ± 5.3 mV and −73.8 ± 6.1 mV, n = 6, respectively).
Because the kinetic properties of the IPSCs measured in this study were considerably slower than those described previously (Pearce 1993), we wanted to ascertain that the electrode placements and stimulation intensities employed in this study were in fact activating two independent, nonoverlapping populations of GABAA receptors. To test this, the amplitude of a single, unconditioned proximal IPSC was compared with a second proximal IPSC that had been conditioned by a preceding prepulse from either the proximal or distal stimulating electrode. At an interstimulus interval of 50 ms, a conditioning proximal prepulse significantly reduced the amplitude of a test proximal IPSC (36.1 ± 6.6%; n = 9; paired t-test, P < 0.05; Fig. 2). In contrast, a conditioning distal prepulse had no effect on the amplitude of a test proximal IPSC (Fig. 2). These results provide evidence that under our experimental conditions, proximal and distal stimulation activated nonoverlapping populations of GABAA receptors on CA1 neurons.
Effects of ethanol on proximal and distal GABAA IPSCs
The effect of ethanol was determined on pairs of proximal and distal GABAA IPSCs evoked under the conditions already described. Bath superfusion with ethanol for 7–10 min potentiated proximal IPSCs in a concentration-dependent manner. This potentiation was typically maximal 2–4 min after the onset of ethanol perfusion, was maintained for the duration of the ethanol superfusion, and recovered fully within 5–15 min (Fig. 3). The most prominent change was typically in the area under the IPSC waveform rather than in the amplitude of these responses, although both parameters were significantly enhanced (Fig. 4). The greater increase in IPSC area resulted from a significant ethanol-mediated prolongation of the decay time constant of these responses at 80 mM (69.9 ± 12.0%; n = 15) and 160 mM (128.7 ± 27.4; n = 9; paired t-tests, P < 0.01).
The minimum concentration of ethanol that produced significant enhancement of proximal IPSC area under our recording conditions was 40 mM (26.7 ± 5.3%; n = 10; paired t-test, P < 0.003; Fig. 4). In contrast, distal IPSC area was significantly potentiated only at the highest ethanol concentration tested (160 mM; 59.5 ± 11.7%; n = 9; paired t-test, P < 0.03), and distal IPSCs were significantly less sensitive to ethanol than proximal IPSCs at all three concentrations tested (unpaired t-tests, P < 0.05). Ethanol did prolong the decay of distal IPSCs; however, a quantitative comparison between ethanol effects on the decay of proximal and distal IPSCs was precluded because of the differences in their basal kinetics. Interestingly, for those distal IPSCs with decay time constants that were better fit by the sum of two exponentials, ethanol did not have a selective effect on the fast component.
Robust potentiation of the proximal response with little or no change in the distal IPSC was often observed simultaneously in the same cell (e.g., Fig. 3). Ethanol did not alter the reversal potential of proximal or distal IPSCs at any concentration tested (data not shown). However, at 160 mM it did produce a modest, but significant, decrease in input resistance (37.4% ± 8.5%; n = 9; P < 0.05).
Effects of other modulators of GABAA receptor function on proximal and distal GABAA IPSCs
Responses evoked by proximal and distal stimulation have been shown to differ in some of their pharmacological properties; in particular, furosemide and volatile anesthetics differentially affect the two kinds of responses (Pearce 1993, 1996). However, the specific GABAA receptor subunit requirements that underlie these differences are not clearly defined. To characterize in more detail possible subunit differences between the GABAA receptors mediating the proximal and distal IPSCs, we examined the effects of other allosteric modulators of GABAA receptor function whose subunit dependence has been relatively well established on the basis of recombinant receptor expression studies. Pentobarbital is a classical barbiturate that can potentiate all combinations of GABAA receptor subunits that form functional channels, except those formed by homomeric ρ-subunits (Seighart 1995). Bath application of 80 μM pentobarbital produced a small increase in IPSC amplitude and a large increase in IPSC area. Most importantly, the pentobarbital sensitivity of proximal and distal IPSCs was not significantly different (n = 12; Fig. 5). Flunitrazepam is a relatively nonspecific benzodiazepine agonist that has greatest efficacy at recombinant GABAA receptors expressing some combination of an α-, β-, and either variant of the γ2 subunit (Seighart 1995). Under our recording conditions, 1 μM flunitrazepam was equally effective at potentiating proximal and distal IPSCs (n = 9; Fig. 5). The most prominent effect of this concentration of flunitrazepam was to increase IPSC amplitude.
Zolpidem is a benzodiazepine type 1 receptor agonist, and has the highest affinity for recombinant GABAA receptors composed of α1, β2/3, and γ2 subunits. Moreover, it has been suggested that ethanol may share the same GABAA receptor subunit requirements as zolpidem (Criswell et al. 1993). We tested a series of zolpidem concentrations and found that 1 μM was the lowest concentration that produced significant potentiation of either proximal or distal IPSC amplitude. At this concentration, there was no significant difference between zolpidem potentiation of proximal and distal IPSCs (n = 14; Fig. 5).
The results of this study demonstrate that IPSCs gated by distinct subpopulations of GABAA receptors on individual CA1 pyramidal neurons differ markedly in their sensitivity to ethanol. Electrical stimulation in close proximity to stratum pyramidale evoked GABAA IPSCs that were significantly more sensitive to ethanol than IPSCs evoked by stimulation in dendritic regions of the CA1 field (stratum lacunosum-moleculare). The differential sensitivity of proximal and distal IPSCs observed with ethanol was not seen with pentobarbital, flunitrazepam, or zolpidem, three other drugs known to allosterically enhance GABAA receptor function.
A series of recent studies (Pearce 1993, 1996; Pearce et al. 1995) has characterized “fast” and “slow” GABAA synaptic currents in CA1 neurons with the use of stimulation protocols similar to those described in this study. The properties of these currents are similar to the proximal and distal IPSCs described in this report except that the decay time constant of proximal IPSCs was slower than that reported for fast IPSCs. The difference in decay kinetics may reflect differences in the populations of GABAA receptor-mediated synapses underlying fast and proximal IPSCs. In fact, fast IPSCs were evoked by stimulating directly in the CA1 somatic layer, whereas the proximal IPSCs tested in the present experiments were evoked by stimulating adjacent to (within 250 μm) the CA1 pyramidal cell layer.
The predominant effect of ethanol on both proximal and distal IPSCs was to increase IPSC area rather than amplitude. This was due to a significant prolongation of the decay time constant of these responses in the presence of ethanol. A similar ethanol-mediated change in IPSC decay was also noted in another recent study (Wan et al. 1996) and is at least consistent with a postsynaptic mechanism of ethanol action (Otis and Mody 1992). These results do not rule out a presynaptic mechanism of ethanol potentiation of GABAA synaptic responses. However, ethanol does not potentiate GABAB responses in CA1 neurons (Wan et al. 1996), suggesting that it does not nonspecifically increase GABA release. In addition, biochemical studies have shown that ethanol, at concentrations similar to those used in this study, inhibits GABA release (Strong and Wood 1984) and enhances GABA reuptake (Foley and Rhodes 1992). These observations would suggest that ethanol is more likely to inhibit rather than potentiate GABAA IPSCs. Nevertheless, to date there have been no reports of ethanol potentiation of currents evoked by exogenous GABA application in hippocampal slices.
At least three major explanations have been proposed for the striking lack of concordance in the reported literature concerning effects of ethanol on GABAA receptor function. One possibility is that differences in the primary structure of the receptor, related most probably to the receptor subunit composition (e.g., Wafford et al. 1991), determine ethanol sensitivity. Another possibility is that posttranslational modification of the receptor, such as by receptor phosphorylation, determines ethanol sensitivity (Freund and Palmer 1996; Weiner et al. 1994); such mechanisms might also require subunits that possess specific phosphorylation sites to be present (Wafford and Whiting 1992). Finally, there might be unknown factors related to procedural differences between different laboratories that could also affect apparent ethanol sensitivity (e.g., Weiner et al. 1995). In the present case, the latter explanation can be completely ruled out as the basis for sensitivity differences, because recordings of synaptic responses with differential ethanol sensitivity were made concurrently in the same cells, although this explanation may still account for some of the differences that have been reported previously in the literature.
Differentiating between the first two alternatives is more difficult, primarily because the populations of synapses that mediate the proximal and distal responses have not been identified, and their subunit composition and phosphorylation state are unknown as well. Evidence from anatomic and electrophysiological studies suggests that the interneurons that generate proximal and distal IPSCs synapse on anatomically segregated regions of CA1 pyramidal neurons. Interneurons located within or near the somatic layer of the CA1 region have been shown to synapse primarily on cell bodies and initial axonal segments of CA1 neurons (Knowles and Schwartzkroin 1981; Miles et al. 1996; Somogyi et al. 1983), whereas more distal interneurons synapse preferentially on dendritic processes of these cells (Lacaille and Schwartzkroin 1988; Miles et al. 1996). Proximal (but not distal) IPSCs can be selectively blocked by focal application of bicuculline to the CA1 pyramidal cell layer (Pearce 1993), and the time course of IPSCs evoked by proximal stimulation appears to be consistent with localization to sites that are electrotonically closer to the soma than are distal responses (Miles et al. 1996; Pearce 1993). In addition,we demonstrated, with the use of a paired-pulse paradigm, that the distal and proximal stimulation protocols used in this study activate nonoverlapping populations of GABAA receptors.
Given the likely anatomic segregation of the synapses mediating proximal and distal IPSCs, it is possible that distinct types of GABAA receptor subunits may mediate these two responses. The general pattern of ethanol-sensitive and-insensitive populations of GABAA receptors both being expressed in individual neurons is further supported by studies of cerebral cortical neurons (Soldo et al. 1993; B. L. Soldo and T. V. Dunwiddie, unpublished data) in which hyperpolarizing responses to local application of GABA to somatic GABAA receptors were enhanced by ethanol whereas depolarizing responses (presumably dendritic) were not. If pharmacologically distinct subtypes of GABAA receptors mediate the proximal and distal responses, then it would seem likely that drugs other than ethanol would share its ability to differentiate between these receptors. A candidate for such a drug is zolpidem, a selective BZ1 receptor agonist, which has been postulated to share GABAA receptor subunit requirements similar to those of ethanol (Criswell et al. 1993). However, under our recording conditions, proximal and distal IPSCs were potentiated to the same extent by 1 μM zolpidem. This result suggests that either the differential ethanol sensitivity of proximal and distal IPSCs does not arise from a difference in the distribution of GABAA receptor subunits at these synaptic loci or that, at least in the hippocampus, zolpidem and ethanol do not share the same GABAA receptor subunit requirements.
The fact that proximal and distal IPSCs were not differentially sensitive to zolpidem and flunitrazepam also rules out the possibility that the differential ethanol sensitivity arose from reduced voltage control of distal dendritic synapses. Because the magnitude of the potentiation induced by the benzodiazepines was similar to that of ethanol, any limitations in recording small enhancement of currents from remote synapses would have affected the benzodiazepine potentiation to the same extent as ethanol.
The possibility that receptor phosphorylation might also regulate ethanol potentiation of GABAA receptor function has been suggested by several groups (protein kinase A, Freund and Palmer 1996; protein kinase C, Wafford and Whiting 1992; Weiner et al. 1994). If this were the case, then differential ethanol sensitivity could arise from differences in the localization of receptors with the appropriate phosphorylation sites; alternatively, it is possible that a homogeneous population of GABAA receptors is distributed throughout CA1 cell bodies and dendrites, but that kinase/phosphatase activity is compartmentalized such that the phosphorylation state of GABAA receptors varies between subcellular regions of these neurons. Although it has been shown that different isozymes of protein kinase C have specific subcellular distributions in hippocampal tissue (Tanaka and Saito 1992), there is currently no evidence as to whether somatic and dendritic GABAA receptors are differentially regulated by protein kinase C in any neuronal population.
Another related possibility stems from a recent report in which it was demonstrated that GABAA IPSCs recorded from CA1 neurons were normally insensitive to ethanol, but in the presence of the competitive GABAB receptor antagonist CGP 35348, ethanol significantly potentiated GABAA responses (Wan et al. 1996). Our experiments were not carried out in the presence of GABAB receptor antagonists, but we did use stimulation conditions that did not generate significant GABAB receptor-gated currents (i.e., minimal stimulation intensity). GABAB receptor activation can inhibit both protein kinase A (Gerber and Gahwiler 1994) and protein kinase C (Tremblay et al. 1995) activity in the hippocampus. Therefore it is possible that the differential ethanol sensitivity of proximal and distal IPSCs observed in this study may reflect differences in GABAB receptor-mediated posttranslational modification of GABAA receptors at these two populations of synapses.
As noted in the results, almost half of the distal IPSCs recorded in this study decayed with two exponential components, one fast and one slow. Because the fast component was very similar in duration to that of proximal IPSCs, it is possible that these distal responses actually contained a small proximal component. This possibility appears unlikely, because we did not find a significant correlation between IPSC decay time constant under control conditions and the degree of subsequent ethanol potentiation of either proximal or distal IPSCs, suggesting that the rate of IPSC decay was not a good predictor of the ethanol sensitivity of the GABAA receptors underlying these responses. Two-component decay rates have been observed for homogeneous populations of recombinant GABAA receptors (Verdoorn et al. 1990) and for GABAA currents recorded from small outside-out patches of cultured hippocampal neurons (Jones and Westbrook 1995). Therefore the complex decay of some distal responses recorded in this study could still be consistent with a homogeneous population of GABAA receptors that, for one of the reasons discussed above, possessed relatively low ethanol sensitivity.
In summary, we have shown that subpopulations of GABAA receptor-gated synapses within individual hippocampal CA1 neurons possess differential sensitivity to intoxicating concentrations of ethanol. This finding suggests that the inconsistent results of previous electrophysiological studies characterizing ethanol modulation of GABAA-mediated synaptic transmission in hippocampal CA1 neurons may have been due to differences in stimulation electrode placement and/or stimulation intensity. In addition, this result may also account for the lack of ethanol potentiation of GABA-mediated chloride flux in microsacs prepared from hippocampal tissue (Proctor et al. 1992), because it is not known whether all populations of GABAA synapses are represented equally in the microsacs and because, at best, only a fraction of the total chloride flux measured would have been carried by “ethanol-sensitive” GABAA receptors. It will be interesting to see whether the synapse specificity of ethanol sensitivity of GABAA receptors observed in hippocampal CA1 and cortical neurons is seen in other neuronal populations in which ethanol/GABAA receptor interactions have been inconsistently observed.
Finally, the results of this study may also have functional implications concerning the physiological consequences of ethanol's action in the hippocampus. Proximal IPSCs have been shown to inhibit repetitive firing of sodium spikes in CA1 neurons (Miles et al. 1996) and therefore are likely to potently modulate the ultimate output of CA1 neurons in response to excitatory synaptic input. In contrast, distal IPSCs are thought to limit the efficacy of local afferent inputs, possibly at the level of individual dendritic spines (Miles et al. 1996; Pearce et al. 1995). Therefore a selective effect of ethanol on proximal GABAA receptor-mediated synapses is unlikely to change the relative weighting of discrete excitatory inputs, but will exert a more generalized inhibitory control over the net output of these neurons.
This work was supported by the Veterans Affairs Medical Research Service, and by National Institute on Alcohol Abuse and Alcoholism Grant AA-03527 and Fellowship AA-05425 to J. L. Weiner.
Address for reprint requests: J. L. Weiner, Dept. of Pharmacology, Box C-236, University of Colorado Health Sciences Center, 4200 E. Ninth Ave., Denver, CO 80262.