Abstract
Effects of chronic ethanol exposure on N-methyl-d-aspartate (NMDA) receptor function were examined in hippocampal neurons. Rat hippocampal neurons grown in culture were chronically exposed to 100 mM ethanol to examine mechanisms that could underlie ethanol-induced changes in receptor function and excitotoxicity. NMDA-stimulated, but not kainic acid-stimulated, increases in intracellular calcium were enhanced after 1-, 2- and 7-day exposures to 100 mM ethanol. Chronic exposure to ethanol for 7 days duration increased the magnitude of cell death mediated by NMDA application, but not that mediated by α-amino-3-hydroxy-5-methylisoxazole-4-propionate or kainic acid exposure. In addition, NMDA-induced excitotoxicity after chronic ethanol exposure (CEE) was not altered in the presence of nifedipine. The enhancement of NMDA-induced neuronal cell death was evident after 2 days of CEE, but not significantly different after a 1-day exposure to 100 mM ethanol. The enhancement of NMDA-induced calcium responses and excitotoxicity could be mimicked by a chronic 7-day exposure to aminophosphonovaleric acid. However, a concomitant chronic exposure of ethanol/aminophosphonovaleric acid did not enhance NMDA-induced calcium responses or excitotoxicity. Chronic exposure paradigms did not consistently alter basal intracellular calcium levels nor total cell number in the absence of exposure to glutamate receptor agonist. These findings support the hypothesis that NMDA receptor function is enhanced after CEE, and this predisposes hippocampal neurons to excitotoxicity.
The major excitatory neurotransmitter in the brain is glutamate, which exerts its effects through two classes of receptors, ionotropic and metabotropic. Several subtypes of the iGluRs have been defined based on their selective activation by agonists, NMDA, AMPA and KA. Excessive activation of these receptors has been implicated in the neuropathology associated with certain disease states, such as Parkinson’s disease, Huntington’s disease, Alzheimer’s disease and epilepsy (see Bennett and Huxlin, 1996; Whetsell, 1996 for review). It is also likely that overstimulation of iGluRs participates in the neuronal loss observed in alcoholics (Krill et al., 1995).
Chronic alcohol abuse has been shown to be associated with a selective neuronal loss in various regions of the brain including the mamillary bodies, cerebellum, hippocampus and cortex. Several studies have shown that CEE may sensitize central neurons to excitotoxic damage by glutamate (Ahern et al., 1994; Blevins et al., 1995; Chandler et al., 1993; Davidson et al., 1993). The NMDA subtype of glutamate receptor is a principal conduit for neuronal death given the high permeability of the NMDA-associated ion channel to calcium (Mayer and Westbrook, 1987). Furthermore, the NMDA receptor is also the most efficacious of the iGluRs in mediating cell death (Hartley et al., 1993; Tymianski et al., 1993). Hence, NMDA receptors likely play a major role in the neuronal toxicity observed with alcohol abuse or as a consequence of alcohol withdrawal (Lovinger, 1993).
The NMDA receptor is particularly sensitive to inhibition by acute ethanol exposure (Bhave et al., 1996; Lovinger et al., 1989) which may vary regionally in the brain (Randollet al., 1996). Conversely, the chronic presence of ethanol appears to induce an adaptive increase in the function of NMDA receptors. CEE has been reported to enhance NMDA-mediated increases in [Ca++]i and excitotoxicity in cortical and cerebellar neurons (Ahern et al., 1994; Blevins et al., 1995; Iorio et al., 1992, 1993). However, whether this functional enhancement occurs in NMDA receptors on hippocampal neurons is unknown, despite the fact that the hippocampus is among the most severely affected brain regions after chronic ethanol treatment (see Walker et al., 1993 for review). Recent evidence has shown that non-NMDA receptors (i.e., AMPA/kainate) may be more sensitive to inhibition by ethanol than originally thought (Dildy-Mayfield and Harris, 1995;Martin et al., 1995). In addition, VSCCs are also inhibited by ethanol and have been implicated in the hyperexcitability after CEE (Leslie et al., 1990; Ripley et al., 1996). The sensitivity of VSCCs and iGluRs to CEE and the contribution of these channels to increases in intracellular calcium concentration and/or neurotoxicity after CEE in hippocampal neurons remain unclear.
In the present study, we have characterized the role in agonist-induced excitotoxicity after CEE of iGluRs and VSCCs in cultured hippocampal neurons. To assess whether an increase in receptor function underlies enhanced excitotoxicity, we measured agonist- or KCl-induced increases the level of intracellular calcium in individual hippocampal neurons. The specificity of CEE effects on NMDA receptors was assessed by examining calcium transients induced by non-NMDA iGluRs and KCl-induced depolarization after CEE. The results obtained in this study indicate selective CEE-induced changes in NMDA receptor function.
Methods
Neuronal cell culture.
Primary hippocampal mixed neuronal/glial cultures were established according to the methods ofBanker and Cowan (1977) with modifications described by Mattsonet al. (1988). Hippocampi were excised from 18-day-old Sprague-Dawley rat fetuses, trypsinized (0.125% trypsin for 30 min), triturated with a narrow-bore pipette and the resulting dissociated cells distributed onto collagen/poly-d-lysine-coated substrates. Cultures used for intracellular calcium measurements were plated onto glass coverslips, whereas cultures used for excitotoxicity were plated onto multiwell plastic plates at a density of 2 × 105 cells per well. The plating medium for these cells consisted of Eagle’s MEM (adjusted to 320–330 mOsmol) supplemented with 10% fetal calf serum, 10% heat-inactivated horse serum and 2 mM l-glutamine. After 24 h, the plating medium was replaced with feeding medium which consisted of MEM supplemented with 5% heat-inactivated horse serum and 2 mMl-glutamine. Cultures were maintained at 37°C in a humidified 10% CO2/90% air mixture. The mitotic inhibitor, 5-fluoro-2-deoxyuridine, was added to the medium after 7 DIV to retard the proliferation of non-neuronal cells. Culture medium was replenished every 7 days. All excitotoxicity treatments or measurements of intracellular calcium were performed on cultures at 14 DIV.
CEE.
Hippocampal neurons in culture were exposed chronically to 100 mM ethanol for 7 DIV. Distilled, absolute ethanol purchased from Midwest Grain Products (Perkin, IL) was used in all experiments. Ethanol was added directly to the feeding medium and the culture was placed in a plastic container along with an isomolar concentration of ethanol in an open well. The container was equilibrated with 10% CO2/90% air and sealed to prevent evaporative loss of ethanol. Control cultures were sealed likewise in plastic containers not containing ethanol. The method of ethanol addition, maintenance and preparation of control cultures was identical in all subsequent paradigms. The day at which CEE was begun and the duration of CEE varied in some paradigms. One- and two-day ethanol treatment paradigms were initiated in hippocampal cultures beginning on day 13 and day 12 in vitro, respectively. A 100 mM ethanol concentration was used for all 1- and 2-day ethanol treatment experiments.
Chronic APV and ethanol + APV paradigm.
Chronic APV cultures were exposed to 70 μM APV for 7 days starting at 7 DIV. Control and APV-treated cultures were equilibrated with 10% CO2/90% air and sealed in plastic containers not containing ethanol. In the chronic ethanol + APV paradigm, 70 μM APV and 100 mM ethanol were added to cultures at 7 DIV and maintained for 7 days. Chronic ethanol + APV cultures were equilibrated with 10% CO2/90% air and sealed in plastic containers with an isomolar concentration of ethanol in an open well. Control cultures were sealed likewise in plastic containers not containing ethanol.
Ethanol determination.
The determination of ethanol concentration was performed by an assay based on the ability of the enzyme alcohol dehydrogenase to catalyze the oxidation of alcohol to acetaldehyde with simultaneous reduction of nicotinamide adenine dinucleotide (NAD) to NADH, which was measured spectrophotometrically with absorbance at 340 nm. This assay was performed with a commercial kit (332-UV) supplied by Sigma Chemical Company, St Louis, MO.
Calcium determination.
Fluorescence ratio imaging with the calcium indicator fura-2 was used to measure free intracellular calcium concentrations ([Ca++]i). Cells were loaded at 37°C for 20 min with 4–6 μM fura-2 AM (Molecular Probes Inc. Eugene, OR) dissolved in an external perfusion buffer of the following composition (in mM): NaCl, 150; KCl, 2.5; CaCl2, 2.5; HEPES, 10; glucose, 10; tetrodotoxin, 0.001; and glycine, 0.003 (adjusted to 320–340 mOsmol with sucrose; pH 7.4). After fura-2 loading, cells were washed with perfusion buffer and allowed to incubate an additional 30 to 50 min before determinations of [Ca++]i were attempted.
Determination of [Ca++]iin single cells was performed in a perfusion chamber mounted on the stage of a Zeiss IM-35 inverted microscope. Fluorescent emission images produced by 340 nm and 380 nm wavelength excitation were acquired at 0.5 Hz with an ISIT camera and fed into Image-1/Fluor (Universal Imaging, West Chester, PA) image processing system. Background fluorescence images taken from regions of the coverslip not containing cells were subtracted at each wavelength. Values for [Ca++]i were determined by averaging all pixel values within a region of interest outlining the cell soma. The increase in [Ca++]i (Δ [Ca++]i), was calculated as the difference between the peak agonist-induced [Ca++]i and resting (basal) [Ca++]i. The change in [Ca++]i was determined on a neuron-by-neuron basis, because this method controlled for variability in basal [Ca++]i among neurons before drug exposure. Measurements from neurons with basal [Ca++]i greater than 160 nM or that failed to maintain a stable base line were not included for analysis.
Free [Ca++]i was calculated from a calibration curve generated by the equation [Ca++]i =Kd (Fmin/Fmax)[(R− Rmin)/(Rmax− R)], in which the Kd = 135 (Grynkiewicz et al., 1985). TheRmin term is the minimum ratio corresponding to totally unbound dye and theRmax term is the maximum ratio for totally bound dye. The terms Fmin andFmax represent the fluorescence value of unbound and bound dye at 380 nm excitation, respectively. The values for Rmin andRmax were determined by use of the in vitro titration method (Grynkiewicz et al., 1985) supplied in kit form (Molecular Probes Inc. Eugene, OR).
Drug and solution application for calcium determination.
A perfusion chamber was constructed by sandwiching the neurons between two coverslips separated by spacers. The volume of this chamber was ≈150 μl and allowed for the complete exchange of solutions within 5 s. Solution flow was gravity driven, and different solutions could be switched via a multiport distribution valve. All drug solutions for [Ca++]i determination were dissolved in external perfusion buffer. NMDA and APV were purchased from Research Biochemicals International, Natick, MA. Kainic acid, glycine and all other chemicals were purchased from Sigma Chemical Company, St. Louis, MO.
Excitotoxicity.
Exposure to excitatory amino acids was similar to that described by Koh and Choi (1988). Before exposure to test solutions, cells were washed twice with a Tris-buffered CSS, consisting of the following (in mM): NaCl, 120; KCl, 5.4; CaCl2, 1.8; Tris-HCl (pH 7.4), 25; and glucose, 15. Glycine (3 μM) was included in the test solutions of all experiments involving NMDA excitotoxicity, but not in excitotoxicity experiments involving other agents. In experiments involving the continued presence of ethanol, ethanol was included in the pretest wash and during the excitotoxic drug exposure. Test solutions of excitotoxic agents dissolved in CSS were applied to cells for 5 min at room temperature. After the 5-min exposure, the test solutions were washed out thoroughly. Feeding medium lacking serum replaced the CSS and the cells were returned to the incubator. Twenty to 24 h after exposure to the test solutions, the cells were exposed to 0.4% trypan blue (5 min) which stained nonviable cells. Neurons were identified by morphology. Cell death was determined by counting viable (phase bright, lacking staining) and nonviable (stained) neurons in three randomly selected, nonoverlapping fields in each well.
Statistical analysis.
Sister cultures (controlvs. experimental) from a single culture preparation were compared in all analyses with control for interculture variability within treatment paradigms. Data are reported as the mean values ± S.E.M. from the indicated number of cells and cultures. For multiple comparisons, data were analyzed by appropriate ANOVA procedures.Post hoc analysis was performed with the Scheffe test for multiple comparisons. For single comparisons, data were analyzed by Student’s t test. A level of P < .05 was considered statistically significant. Statistical analysis was performed with Statview II (Abacus Concepts, Berkeley, CA) or SAS (Statistical Analysis System, Cary, NC).
Results
Exposure of hippocampal neurons to a 5-min application of NMDA produced a dose-dependent increase in percent cell death as measured by trypan blue exclusion (fig. 1). The NMDA-induced excitotoxicity in cultures treated chronically with 100 mM ethanol for 7 days was greater than that in control cultures (F = 10.32, P = .0014, ANOVA). The enhancement of NMDA-induced excitotoxicity after CEE was not caused by an inability to induce cell death in control cultures because maximal excitotoxicity, 96.6 ± 1.4% (mean ± S.E.M.) in control and 98.2 ± 0.9% in CEE cultures, was produced by 500 μM NMDA with no significant difference between control and CEE cultures (two-tailed t test, P > .1). The NMDA-induced excitotoxicity was mediated by NMDA receptors because it could be effectively blocked by coapplication during 5 min exposure to the competitive NMDA antagonist, APV (70 μM). The continued presence of ethanol reduced NMDA excitotoxicity to the same extent in both control and CEE cultures. The percent cell death induced by NMDA (50 μM) exposure in the absence of ethanol was 22.6 ± 2.6 (mean ± S.E.M.) in control cultures and 40.4 ± 4.7 in CEE, whereas excitotoxicity in presence of 100 mM ethanol was 4.8 ± 1.1 in control and 6.8 ± 1.9 in CEE cultures. Thus, enhanced excitotoxicity was only observed after ethanol withdrawal.
The duration of ethanol exposure necessary for the enhancement of NMDA-induced excitotoxicity was examined in hippocampal cultures exposed to 100 mM ethanol for 2 days. NMDA-induced excitotoxicity was significantly greater in ethanol-treated cultures than controls after the 2-day CEE treatment (F = 15.65, P = .0001, ANOVA). Significant increases in excitotoxicity were observed in the 2-day CEE group in to 50 and 75 μM NMDA. A 1-day, 100 mM ethanol exposure did not significantly enhance NMDA-induced excitotoxicity (F = 3.74, P = .0549, ANOVA).
Recent evidence has shown that antagonists of NMDA receptors can prevent the enhancement of NMDA receptor function if given during the chronic ethanol treatment period (Hu and Ticku, 1995). To examine possible mechanisms for this enhancement and to determine whether NMDA antagonists would prevent the enhancement of NMDA receptor function, 70 μM APV was included in a 100 mM ethanol/7-day chronic exposure paradigm. No enhancement of NMDA-induced excitotoxicity was found (F = 0.52, P = .47, ANOVA) with concomitant ethanol/APV exposure (fig. 2A). However, when hippocampal cultures were treated chronically with 70 μM APV alone for 7 days, a significant increase in NMDA-induced excitotoxicity was observed (F = 8.16, P = .005, ANOVA) (fig. 2B). The lack of enhanced NMDA receptor function in the chronic ethanol + APV paradigm was not evident in positive control test wells treated with either chronic ethanol or APV alone for this group of neurons.
Specificity of enhanced channel function after CEE.
In contrast to the enhanced NMDA-induced excitotoxicity, no enhancement of KA-induced excitotoxicity was observed after a chronic exposure to 100 mM ethanol for 7 days (F = 0.23, P = .63, ANOVA). KA exposure produced a robust dose-dependent increase in neuronal cell death that was of similar magnitude in both ethanol-treated and control cultures (fig. 3A). In test sample wells serving as positive controls, a significant enhancement of NMDA-induced excitotoxicity was evident (F = 6.05, P = .018, ANOVA). The KA-induced excitotoxicity was sensitive to block by the NMDA antagonist, APV (70 μM), which reduced the percent cell death induced by 250 μM KA from 95.9 ± 1.0 to 49.7 ± 3.8 (mean ± S.E.M.) in control and 97.0 ± 0.9 to 49.4 ± 6.6 in CEE cultures. A combination of APV (70 μM) and the non-NMDA antagonist, 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo(F)quinoxaline (6 μM), further reduced KA-induced excitotoxicity to 12.7 ± 2.9 (mean ± S.E.M.) and 7.9 ± 3.6 in control and CEE cultures, respectively. No differences were observed in the ability of antagonists to reduce KA-induced excitotoxicity in control and CEE cultures.
The non-NMDA glutamate receptor agonist, AMPA, also induced neuronal cell death in hippocampal cultures. In four of five experiments, a 5-min exposure to AMPA resulted in a concentration-dependent increase in neuronal cell death; however, AMPA-induced excitotoxicity was not enhanced as a result of CEE (F = 2.19, P = .14, ANOVA) (fig.3B). The excitotoxicity induced by AMPA (250 μM), like that of KA, appeared to be mediated by NMDA receptors because it was completely blocked by 70 μM APV. The percent cell death induced by AMPA exposure in the absence of APV was 28.9 ± 8.5 (mean ± S.E.M.) in control cultures and 33.0 ± 11.8 in CEE cultures, whereas excitotoxicity in the presence of APV was 4.1 ± 0.8 in control and 1.6 ± 0.4 in CEE cultures.
Contribution of VSCCs to enhanced NMDA excitotoxicity after CEE.
Chronic ethanol treatment has been reported to enhance the function of L-type VSCCs (Whittington et al., 1995;Whittington and Little, 1993). Such an enhancement could contribute to the excitotoxic damage produce by NMDA. To examine the role of VSCCs in cell death, control and CEE cultures were exposed to 50 mM KCl to activate VSCCs. KCl exposure produced no neuronal cell death in either control or CEE cultures (table 1). However, it is possible that a contribution of VSCCs is only evident in cooperation with NMDA receptor activation. Nifedipine, an antagonist of L-type VSCCs, was coapplied along with 50 μM NMDA to control and CEE cultures. Nifedipine (5 μM) inclusion resulted in a neuroprotective effect in control cultures, but not in CEE cultures (fig.4). Neuronal cell death mediated by NMDA exposure was inhibited in the presence of nifedipine in control cultures, whereas nifedipine had no effect on NMDA-mediated cell death in CEE cultures. The percent cell death induced in control cultures in the absence and presence of nifedipine was 22.6 ± 2.6 (mean ± S.E.M.) and 14.3 ± 2.6, respectively. In CEE cultures, the percent cell death in the absence and presence of nifedipine was 40.4 ± 3.9 (mean ± S.E.M.) and 32.7 ± 4.6, respectively. Nifedipine or the vehicle, 0.5% DMSO, were not neurotoxic by themselves because no cell death was observed in the absence of NMDA. However, a small increase in excitotoxicity above control was observed with vehicle in the presence of NMDA in control neurons. Nifedipine did not produce a significant reduction in excitotoxicity in the CEE group when compared with either the NMDA or NMDA + DMSO group (fig. 4). The differences that were observed in excitotoxicity in the presence of nifedipine indicate that involvement of L-type VSCCs might be decreased after CEE because nifedipine reduced cell death in control but not CEE cultures. L-type channels did not contribute to CEE-enhanced excitotoxicity induced by NMDA because inhibition by nifedipine was not observed in CEE cultures.
NMDA-induced calcium transients after CEE.
To determine whether enhanced NMDA receptor function is associated with the enhanced excitotoxicity observed after chronic antagonism, we examined NMDA-stimulated increases in intracellular calcium. The ability of ionotropic glutamate receptor activation to increase intracellular calcium was assessed after a chronic exposure to 100 mM ethanol for 7 days. A 3-s application of NMDA produced transient increases in the [Ca++]i, which returned to baseline after agonist removal in both control and CEE neurons (fig.5, A and B, respectively). Peak NMDA-induced calcium responses occurred at NMDA concentrations of ∼20 to 30 μM. When NMDA-induced calcium responses were compared in control versus CEE cultures, the NMDA-induced increases in [Ca++]i were found to be larger in CEE neurons than in control neurons (F = 7.96, P = .0051, ANOVA) at NMDA concentrations of 5 and 10 μM (fig. 5C). Conversely, calcium responses induced by KA (25 μM) were larger in control neurons than in CEE neurons treated with 100 mM ethanol for 7 days (table 2). However, the KA-induced calcium response was not consistently altered across several treatment paradigms. Considerable variability in the KA-induced calcium response was observed between paradigms (table 2). Because the experimental measurements within any given paradigm were made with sister cultures derived from a single culture preparation (five separate culture preparations for the 7-day CEE paradigm, table 2) the variability in the magnitude of kainate responses likely reflects interculture preparation variability.
The duration of ethanol exposure needed to elicit enhanced NMDA-induced calcium responses was examined after 1- and 2-day chronic treatments with 100 mM ethanol (F = 13.74, P = .0003, ANOVA; F = 15.86, P = .0001, ANOVA, respectively). In contrast to the CEE-induced change in the NMDA response at 7-day CEE, the enhancement of NMDA-induced increases in [Ca++]i was observed at the higher NMDA concentrations (>20 μM NMDA). KA-induced calcium responses were not altered in either the 1- or 2-day paradigms (table2). Overall, the evidence from the time-course experiments indicate that alterations in NMDA receptor function can occur within 24 h of ethanol exposure.
In contrast to the enhancement of [Ca++]i observed with CEE (100 mM/7 days), NMDA-induced calcium responses after a chronic exposure to 100 mM ethanol + APV (70 μM) were found to be significantly smaller than those observed in control neurons (F = 9.71, P = .0025, ANOVA). This finding indicates that a concomitant ethanol + APV exposure blocks the enhancement of NMDA receptor function (i.e., receptor adaptation) associated with CEE (fig.6A). This effect was specific to calcium responses induced by NMDA because the KA-induced increase in [Ca++]i was not altered (table 2; two-tailed t test, P = 0.3). A chronic 7-day exposure of 70 μM APV alone produced an enhancement of NMDA-induced calcium responses (F = 38.66, P = .0001, ANOVA) (fig. 6B). In contrast, KA-induced increases in [Ca++]i were not altered by chronic APV (table 2; two-tailed t test, P = .094).
Increased [Ca++]iinduced by activation of VSCCs.
Because all VSCCs can be activated by membrane depolarization, 50 mM KCl was used to elicit activation of VSCCs. KCl application produced increases in [Ca++]i of a magnitude similar to that of 10 to 20 μM NMDA (data not shown). When KCl-stimulated responses were examined in control versus CEE neurons, no differences were observed, which indicates that a 100 mM ethanol/7-day CEE does not enhance the VSCC-mediated calcium response (two-tailed t test, P = .54).
Differences in total cell number and basal intracellular calcium.
The range of values for the mean total cell number was 98.0 ± 5.0 to 204.2 ± 7.6 in control cultures and 97.9 ± 4.6 to 179.1 ± 6.4 in CEE cultures. Significant differences in total cell number were found in two (2-day CEE and chronic ethanol + APV) of six treatment paradigms (two-tailed t test, P < .05). The largest difference was observed in the ethanol + APV paradigm, with a mean difference in total cell number of 38.1. However, no correlation between total cell number and percent cell death was evident in either paradigm (r2 = 0.01, 2-day CEE; r2 = 0.0009, chronic ethanol + APV). The differences in cell number most likely reflect variability in plating density rather than neurotoxicity during the chronic treatment.
Higher levels of basal [Ca++]i from those of sister control neurons were observed in the following chronic exposure paradigms: 1-day CEE, 2-day CEE, chronic APV and chronic ethanol + APV (two-tailed t test, P < .05). The largest mean difference in basal [Ca++]i between control and treated neurons that was observed in any paradigm was ∼17 nM [24.2 ± 1.7 (mean ± S.E.M.) in control neurons and 41.4 ± 3.9 in treated neurons] in the chronic ethanol + APV paradigm. The range of basal calcium values was 20.8 ± 2.1 (mean ± S.E.M.) to 30.4 ± 2.5 in control neurons and 22.0 ± 1.6 to 43.1 ± 3.0 in CEE neurons. The differences in basal [Ca++]i do not affect the overall interpretation of the relative increases in calcium, because basal calcium values were subtracted from peak calcium values for each individual neuron.
Discussion
This study demonstrates that CEE sensitizes hippocampal neurons to excitotoxicity mediated by NMDA receptors. We have shown that the effect of CEE on NMDA-induced excitotoxicity does not involve AMPA/kainate or L-type VSCCs. The enhancement of excitotoxicity that was observed in hippocampal neurons is in agreement with other studies which demonstrate an enhancement of NMDA-induced excitotoxicity after CEE (Ahern et al., 1994; Blevins et al., 1995;Chandler et al., 1993; Iorio et al., 1993;Davidson et al., 1993). Our observations extend these earlier findings by demonstrating that no enhancement of excitotoxicity is observed when non-NMDA receptors are explicitly activated with selective agonists. The antagonist, APV, blocked NMDA-induced excitotoxicity and was equally effective in both the control and CEE cultures, which indicates that CEE does not alter the ability of the receptor to be blocked by this antagonist. Furthermore, the ability of ethanol to reduce cell death was not altered by CEE, which indicates that a loss of ethanol inhibition of NMDA receptors is not involved in the adaptation of NMDA receptors to CEE, as indicated previously byChandler et al. (1993).
The enhancement of NMDA-induced excitotoxicity could be observed after 2 days of chronic exposure to 100 mM ethanol. This indicates that a relatively short time is necessary for adaptive mechanisms to occur leading to enhanced excitotoxicity after CEE. This finding could have implications for repetitive “binge” intoxication. It has been reported that short, repetitive ethanol administration will result in neuronal cell loss (Collins et al., 1996; Lundqvist et al., 1995).
Excessive stimulation of AMPA/KA receptors results in neurotoxicity (reviews, Coyle and Puttfarcken, 1993; Whetsell, 1996). Although these receptors are sensitive to inhibition by ethanol (Dildy-Mayfield and Harris, 1995; Lovinger et al., 1989; Martin et al., 1995) we did not observe evidence of an adaptive response to CEE with respect to excitotoxicity because there was no enhancement of AMPA- or KA-induced excitotoxicity after a 7-day chronic exposure to 100 mM ethanol. Furthermore, a large proportion of the neurotoxicity induced by non-NMDA agonists (i.e., AMPA, KA) appeared to involve activation of NMDA receptors, because APV could effectively reduce the percent cell death. The ability of APV to reduce KA-induced excitotoxicity likely reflects secondary synaptic glutamate release, because receptor activation can stimulate neuronal firing which ultimately results in transmitter release. Although AMPA and KA exposure produced excitotoxicity which appeared to be partly mediated by NMDA receptors, CEE did not result in an increased AMPA- or KA-induced excitotoxicity from enhanced NMDA receptor expression. It is possible that this reflects a less than full expression of NMDA receptor function, because glycine was not included during excitotoxic AMPA or KA exposure. It is also possible that the population of NMDA receptors activated by the secondary synaptic glutamate release is different than that which would be activated by direct NMDA application (e.g., synaptic vs. synaptic and somal receptors).
Antagonists to VSCCs decrease the severity of ethanol withdrawal hyperexcitability, which suggests that VSCCs are altered after chronic ethanol treatment (Huang and McArdle, 1993; Ripley et al., 1996; Whittington et al., 1995; Whittington and Little, 1993). However, we did not observe an alteration of KCl-induced increases in [Ca++]iafter CEE. This agrees with other studies which have also reported no perturbation of depolarization-induced increases in [Ca++]i after CEE (Hu and Ticku, 1995; Iorio et al., 1992, 1993).
The role of VSCCs in neuronal cell death after CEE is unknown. We observed that KCl-induced depolarization did not result in neuronal cell death in either control or CEE cultures. The absence of cell death resulting from prolonged depolarization has been reported by other investigators (Churn et al., 1995; Tymianski et al., 1993). The failure of KCl to induce cell death was not the result of an inability to increase [Ca++]i, because a 50 mM KCl exposure produced increases in [Ca++]i similar in magnitude to that of low NMDA concentrations. Although during the course of the 5-min KCl exposure which was used to induce excitotoxicity, other factors such as channel inactivation could influence the total [Ca++]i. The lack of neuronal damage associated with KCl exposure is consistent with the idea that the route of calcium entry in hippocampal neurons is an important factor in determining neuronal cell death (Tymianski et al., 1993). Furthermore, this relationship is not altered after CEE.
Although activation of VSCCs does not induce cell death it is possible that VSCCs could play an associative role in NMDA-mediated excitotoxicity. When nifedipine was applied along with NMDA to control neurons a decrease in excitotoxicity was observed. However, no effect of nifedipine on NMDA-induced excitotoxicity was observed in CEE cultures. The lack of an effect of nifedipine on NMDA receptor-mediated excitotoxicity after CEE is consistent with what has been shown in cerebellar granule cells; however, nifedipine did not decrease excitotoxicity in untreated granule cells (Iorio et al., 1993). Thus, in hippocampal neurons a portion of the NMDA-induced excitotoxicity is mediated by activation of L-type VSCCs in control cultures. The decrease in the involvement of these channels in excitotoxicity after CEE to hippocampal neurons may result from a decrease in the numbers of channels or alterations in channel pharmacological properties or function. The contribution of L-type VSCCs to NMDA-induced excitotoxicity after CEE could be masked by the enhanced NMDA receptor-mediated excitotoxicity.
The ability of concomitant ethanol/APV exposure to prevent an enhancement of NMDA-induced cell death does suggest that the enhanced excitotoxicity after CEE requires active NMDA receptors. Furthermore, the lack of enhanced excitotoxicity observed after chronic ethanol + APV exposure is associated with the failure of this paradigm to produce a functional enhancement of the NMDA-induced calcium response. This finding suggests that functional NMDA receptors are necessary for the enhancement of receptor function resulting from CEE. Overall, these findings indicate that the enhanced excitotoxicity in hippocampal neurons that resulted after CEE selectively depends on alterations in functional NMDA receptors and not non-NMDA glutamate receptors or calcium channels.
The mechanisms that underlie enhancement of NMDA-induced excitotoxicity after CEE may involve alterations in the function of NMDA receptors. Enhancement of NMDA-induced increases in [Ca++]i was observed in all treatment paradigms except chronic ethanol + APV, whereas KA-induced increases in [Ca++]i were not altered consistently. The enhancement of NMDA-induced [Ca++]i agrees with evidence from other studies which have reported similar results (Ahernet al., 1994; Blevins et al., 1995; Hu and Ticku, 1995; Iorio et al., 1992, 1993). Evidence from receptor binding studies indicates that there are increases in receptor number with no change in affinity after CEE or chronic ethanol abuse (Freund and Anderson, 1996; Grant et al., 1990; Hoffman et al., 1995; Michaelis et al., 1990; Snell et al., 1993). However, the NMDA dose-response curve that we observed indicated no increase in the maximal NMDA-induced [Ca++]i increase, which would have been expected from a simple increase in receptor number. Comparison of the calcium dose-response curve after 7-day CEE with that from chronic APV exposure shows that enhancement of [Ca++]i in the 7-day CEE paradigm was present predominately at low NMDA concentrations (<20 μM), whereas in the chronic APV paradigm enhancement of [Ca++]i was present at all NMDA concentrations, which is consistent with a general increase in receptor number (Williams et al., 1992). These findings suggest a complex alteration of NMDA receptor function with CEE which may involve changes other than an increased density of receptors. It is possible that changes in NMDA receptor subunit composition could underlie these effects because receptor subunit composition can affect NMDA receptor properties such as agonist affinity, calcium permeability and increases in [Ca++]i(Grant et al., 1997; Koltchine et al., 1996;Kuner and Schoepfer, 1996; Monyer et al., 1994). Furthermore, several studies have reported that mRNA levels and protein expression of only certain NMDA subunits are elevated after chronic ethanol paradigms (Follesa and Ticku, 1995, 1996; Snell et al., 1996; Trevisan et al., 1994). However, evidence for changes in NMDA receptor subunit composition which could mediate the enhanced calcium response that we observed after CEE has yet to be demonstrated in cultured hippocampal neurons.
A correlation between the concentration of intracellular calcium and cell death was not observed under all of the conditions used in this study, which is consistent with evidence that the major determinants of cell death are the duration of agonist exposure, the route of calcium entry and the buffering capability of the neuron, and not the peak calcium concentration (Mattson et al., 1991; Tymianskiet al., 1993). The alteration of the NMDA calcium response was correlated with the alteration of excitotoxicity in all but two paradigms. Enhancement of the NMDA-induced increase in [Ca++]i was observed in the 1-day CEE paradigm in which no enhancement of excitotoxicity was observed. In the chronic ethanol + APV paradigm a decrease in the NMDA calcium response was observed in treated neurons relative to control neurons; however, the excitotoxicity curve for both treated and control was not different. It is probable that in the former case a threshold level of intracellular calcium needed to enhance NMDA-induced excitotoxicity was not exceeded after the 1-day CEE paradigm. In the latter case, no change in excitotoxicity would necessarily be expected because peak intracellular calcium concentration is not tightly correlated with neuronal death.
Overall, the observations reported in this study indicate that in cultured hippocampal neurons the effect of CEE on agonist-induced increases in [Ca++]i and excitotoxicity are specific to NMDA receptors. Furthermore, the effect of CEE likely involves alterations in NMDA receptor properties. The enhanced excitotoxicity observed after CEE is associated with an enhanced NMDA receptor calcium response and little if any contribution from other ion channels.
Footnotes
-
Send reprint requests to: David M. Lovinger, Ph.D., Department of Molecular Physiology and Biophysics, Vanderbilt University Medical School, 702 Light Hall, Nashville, TN 37232-0615.
-
↵1 This research was supported by AA05361, National Institute on Alcohol Abuse and Alcoholism predoctoral fellowship (C.T.S.); AA08986 (D.M.L.); RR03032 Research Centers and Minority Institutions, RI18714805 RCMI, F49620 Air Force Office of Scientific Research (J.J.M.).
- Abbreviations:
- AMPA
- α-amino-3-hydroxy-5-methylisoxazole-4-propionate
- ANOVA
- analysis of variance
- APV
- aminophosphonovaleric acid
- CEE
- chronic ethanol exposure
- CSS
- control salt solution
- DIV
- days in vitro
- DMSO
- dimethyl sulfoxide
- iGluRs
- ionotropic glutamate receptors
- KA
- kainic acid
- MEM
- minimum essential medium
- mOsmol
- milliosmole
- NMDA
- N-methyl-d-aspartate
- VSCCs
- voltage-sensitive calcium channels
- Received January 24, 1997.
- Accepted August 18, 1997.
- The American Society for Pharmacology and Experimental Therapeutics