Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

FAS is a series of malformations and abnormalities occurring in children exposed to ethanol in utero and is the leading nongenetic cause of mental retardation in the world (West et al., 1994). The primary features of children with FAS are growth deficiencies, including microencephaly, craniofacial dysmorphology and central nervous system dysfunction such as deficits in memory, attention and IQ. In rodents, ethanol exposure during development significantly reduces the size of the brain as well as brain/body weight ratios (Grant and Samson, 1982; Pierce and West, 1986). Similarly, magnetic resonance imaging of individuals exposed prenatally to alcohol show a reduction in the size of several brain areas including the diencephalon, basal ganglia and anterior vermis of the cerebellum (Mattson et al., 1992; Sowell et al., 1996). The ethanol-induced microencephaly appears to be due to a decrease in cell number as exposure of rats to ethanol in utero or during early postnatal development reduces the number of neurons in the inferior olive (Napper and West, 1995), and principal sensory nucleus (Miller, 1995), as well as decreasing the number of cerebellar Purkinje and granule cells (Bonthius and West, 1990; West et al., 1990). Similarly, the addition of in vitro ethanol to primary cultures of cerebellar granule cells (Pantazis et al., 1993), neural crest cells (Chen and Sulik, 1996) and various transformed cell lines (Pantazis et al., 1992; Klein et al., 1996) results in a reduction in cell number.

The above reductions in cell number could be due to decreased proliferation, altered migration or increased cell death. Ethanol treatment has been shown to cause a concentration-dependent decrease in proliferation of the neuron-like PC12 cells (Pantazis et al., 1992) as well as human astroglia in culture (Kane et al., 1996). Although neuronal proliferation in vivo also appears to be altered by prenatal exposure to ethanol (Miller, 1986, 1988), at least cerebellar Purkinje cells are more vulnerable to ethanol during differentiation rather than during neurogenesis (Marcussen et al., 1994). In addition to alterations in cell proliferation, in utero ethanol exposure of rats has been reported to delay neuronal migration and cause neuronal movement to ectopic locations (Miller, 1993).

The ethanol-mediated reductions in neuronal number could be due to increased cell death; however, there is a paucity of data on this subject. During normal development as many as 50% of the neurons in some regions of the central nervous system are lost via a form of cell death referred to as apoptosis (Oppenheim, 1991; Raff et al., 1993). This apoptotic death appears to be primarily controlled by a limiting supply of target-derived growth factors and thus allows a matching of neuronal populations to target size. In addition to this developmental programmed cell death, apoptotic death also is observed in response to insults such as ethanol. An increase in the number of apoptotic cells in the livers of micropigs and rats was reported after chronic administration of ethanol (Yacoub et al., 1995; Halsted et al., 1996). Similarly, acute treatment of hepatocytes, lymphocytes and thymocytes with ethanol increased the number of apoptotic cells (Slukvin and Jerrells, 1995; Kurose et al., 1997). In a preliminary report, Hamby-Mason et al. (1996) noted an increase in the number of apoptotic cells in various brain areas after postnatal exposure of rats to ethanol.

Our investigation was undertaken to determine the effects of ethanol on neuronal cell death using rat PC12 pheochromocytoma as a model system. These studies demonstrate that ethanol can induce apoptosis in neuronal cells. In addition, pharmacologically relevant concentrations of ethanol potentiate the apoptosis induced by removal of serum factors. This is in contrast to the ethanol-induced reduction in cell death caused by H₂O₂.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. Cell culture media and agarose was purchased from Gibco BRL (Grand Island, NY). Fetal calf serum was purchased from HyClone Laboratories (Logan, UT) and horse serum from JRH Biosciences (Lenexa, KS). The ethidium homodimer was purchased from Molecular Probes (Eugene, OR). For TUNEL assays, the biotin-16-dUTP was purchased from Boehringer Mannheim (Indianapolis, IN), TdT and TdT buffer from USB (Cleveland, OH) and strepavidin-peroxidase and hematoxylin from Zymed Laboratories (San Francisco, CA). Proteinase K was purchased from Fisher Biotech (Fair Lawn, NJ) and RNase A from Boehringer Mannheim. All other chemicals were purchased from common commercial suppliers.

Tissue culture. PC12 cells were maintained at 37°C in DMEM containing 25 mM glucose, 4 mM glutamine and supplemented with 15% serum (i.e., 5% heat-inactivated horse serum, 10% fetal calf serum), 100 U/ml penicillin and 100 µg/ml streptomycin. For experiments, PC12 cells were removed from the culture plates by trituration in DMEM and centrifuged at 800 × g for 3 min. The cell pellet was resuspended in DMEM and the cells again centrifuged at 800 × g for 3 min. This procedure was repeated and the washed cells were plated onto collagen-coated tissue culture plates in either serum-free or serum-containing medium. Ethanol exposure was carried out by including ethanol in the media and incubating the plates at 37°C in plastic desiccators containing an atmosphere of 95% air, 5% CO₂ that was saturated with the appropriate concentration of ethanol. This system has been shown to result in an approximate 20% loss of ethanol from the medium after 2 days (Rabin, 1988).

Evaluation of cell death. PC12 cells, which were plated onto collagen-coated 24-well plates (0.8-1 × 10⁶ cells/well), were incubated for 1 hr in PBS containing 3 µg/ml of the DNA-binding fluorophore ethidium homodimer. Ethidium fluorescence was measured using a fluorescent plate reader (535 nm excitation, 645 nm emission). Preliminary studies had established that a 1-hr loading time resulted in maximum ethidium-DNA fluorescence.

Evaluation of DNA fragmentation. Cells were incubated in lysis buffer (5 mM Tris, 20 mM EDTA, 0.025% Triton X-100, pH 7.4) for 30 min on ice and then centrifuged at 15,000 × g for 15 min. The resulting pellet was resuspended in 2 ml lysis buffer and briefly sonicated. The DNA contents of the 15,000 × g supernatant (fragmented DNA) and pellet were measured using the DNA-binding fluorophore Hoechst-33342 (2 µg/ml) as previously described (Labarca and Paigen, 1980). Salmon sperm DNA was used as a standard, and DNA fragmentation was expressed as percent total DNA [i.e., DNA_sup/(DNA_sup + DNA_pellet)].

In situ detection of DNA fragmentation by TUNEL. PC12 cells, which were plated onto collagen-coated glass coverslips (8 × 10⁴ cells/cm²), were fixed in 4% formaldehyde for 15 min and then incubated with 100% ethanol for 5 min. The TUNEL reaction was performed by incubating the coverslips at 37°C in a final volume of 50 µl TdT reaction buffer (100 mM Na cacodylate, 0.2 mM 2-mercaptoethanol, 2 mM CoCl₂ · 6 H₂O, 16 units TdT, 1.25 nM biotin-16-dUTP, pH 7.2). After 60 min, cells were incubated for 15 min with 30 mM Na citrate containing 300 mM NaCl. The cells were then rinsed in distilled H₂O and covered with a 2% bovine serum albumin blocking solution for 10 min. The biotinylated 3'-OH DNA ends were visualized by reaction with strepavidin-peroxidase, and cells were counter-stained with hematoxylin following manufacturer directions (Zymed Laboratories). The extent of apoptosis was quantified by counting the number of apoptotic events in a minimum of 20 randomly chosen fields on each coverslip using 400× magnification. Nonspecific biotinylation was evaluated by incubating sister coverslips in the above assay buffer but in the absence of TdT. Additional sister coverslips also were used for a determination of both protein and DNA content.

Evaluation of DNA laddering. Cells were incubated on ice in lysis buffer (20 mM Tris, 5 mM EDTA, 0.025% Triton X-100, pH 7.5) for 30 min, and an aliquot of the cell lysate removed for determination of protein and DNA content. The lysate was centrifuged at 15,000 × g for 20 min, and the resulting supernatant incubated with RNase A (250 µg/ml) and Proteinase K (100 µg/ml) for 2 hr at 37°C. The samples were extracted twice with phenol/chloroform (1:1) and once with chloroform. DNA was precipitated by incubating the sample for 30 min at 20°C in 2 volumes 95% ethanol containing 0.3 M Na acetate. The sample was then centrifuged at 20,000 × g for 15 min at 4°C, and the resulting pellet was either air dried or dried under vacuum in a Savant SpeedVac. The DNA pellet was resuspended in loading buffer (10 mM Tris, 1 mM EDTA, 5% glycerol, 0.25% bromphenyl blue, pH 7.5), and the entire sample was then subjected to electrophoresis on a 2% agarose gel using a TBE running buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA, pH 8.3) containing 0.5 µg/ml ethidium bromide. The cumulative fluorescence intensity of each of the lower bands (i.e., intensity of the DNA ladder) was measured using a Bio-Rad Molecular Imaging system (Bio-Rad Lab., Hercules, CA).

Total cell protein/DNA determinations. For the determination of total cellular protein, cells were incubated in 0.1 N NaOH and protein content measured using the colorimetric Bio-Rad protein dye binding procedure with bovine serum albumin (fraction V) as a standard. For the determination of total cell DNA, cells were incubated in 5 mM Tris, 20 mM EDTA, 0.025% Triton X-100 (pH 7.4) and DNA content measured as described above using salmon sperm DNA as a standard.

Statistical analysis. The effect of ethanol on serum withdrawal-induced DNA laddering was evaluated using a paired t test. The effect of ethanol on cell death at the various serum concentrations was evaluated using two-way analysis of variance. The remaining data were analyzed using analysis of variance and the post hoc Student-Newman-Keuls pairwise comparison procedure.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Because during development some 50% of the neurons normally undergo cell death due, at least in part, to a loss of neurotrophic factors (Oppenheim, 1991; Raff et al., 1993), the effects of ethanol on death induced by serum withdrawal were studied. A 24-hr incubation in serum-free media significantly (P < .05) increased the number of cells labeled with the cell impermeable DNA-binding fluorophore ethidium homodimer (fig. 1). Inclusion of 50 mM ethanol in the serum-free media caused a significant 39% potentiation in the number of ethidium-labeled cells (fig. 1). Similarly, treatment with 150 mM ethanol significantly potentiated the amount of cell death after serum withdrawal by 125% (data not shown). Ethanol, however, does not appear to be accelerating cell death. Compared to control cells, serum removal for 6 hr increased the amount of ethidium-labeled cells 190 ± 15.7 and 223 ± 20.3% in the absence and presence of 150 mM ethanol, respectively (N = 8). Although a 24-hr inclusion of 150 mM ethanol in the presence of 15% serum did not alter cell viability, a 3-day exposure resulted in a statistically significant 42 ± 9.6% increase in the amount of ethidium-labeled cells (P < .01; N = 7). The above increases in ethidium fluorescence were not due to a direct effect of ethanol on membrane permeability as addition of 150 mM ethanol during the 1 hr loading with the dye did not alter ethidium labeling of cells (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Ethanol potentiates cell death due to serum withdrawal. PC12 cells were treated for 24 hr with 50 mM ethanol in the presence or absence of 15% serum. Cell death was determined using the fluorescent DNA-binding probe ethidium homodimer. Data are expressed as arbitrary fluorescent units/mg protein and are plotted as mean ± S.E.M. (*P < .05, N = 10).

The effect of ethanol on cells maintained in reduced serum also was evaluated using ethidium homodimer. Two-way analysis of variance of the cell death indicated both a significant effect of serum concentration (F_3,40 = 67.2, P < .0001), and a significant ethanol treatment effect (F_1,40 = 67.2, P < .0001), as well as a significant ethanol/serum concentration interaction (F_3,40 = 14.2, P < .0001). A posteriori statistical analysis indicated that lowering the serum concentration to 4 or 6% for 24 hr did not alter cellular viability. Furthermore, a 3-day incubation in 4% serum did not lead to an increase in cell death; 4 ± 9.3% (N = 3) more cells were labeled with ethidium after 3 days in 4% serum compared to sister plates maintained in the normal 15% serum for 3 days. The addition of 150 mM ethanol to cells maintained in 4 or 6% serum, however, caused a significant induction of cell death (fig. 2).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Ethanol exposure induces cell death at low serum concentrations. PC12 cells were treated with 150 mM ethanol in the presence of 15, 6, 4 or 1% serum (fetal calf serum and heat-inactivated horse serum at a ratio of 2:1) for 24 hr. Cell death was determined using the fluorescent DNA-binding probe ethidium homodimer. Data are expressed as arbitrary fluorescent units/mg protein and are plotted as mean ± S.E.M. (**P < .01, ***P < .001, N = 6). , 0 mM ethanol, , 150 mM ethanol.

To determine whether the increases in cell death by ethanol were specific to serum withdrawal, we investigated the effect of ethanol on free-radical induced cell death. PC12 cells exposed to 2 mM H₂O₂ in the presence of 15% serum displayed a significant increase in cell death. However, contrary to cell death induced by serum withdrawal, cotreatment with 150 mM ethanol significantly reduced the amount of H₂O₂-mediated cell death by 34% (fig. 3). Cotreatment with a lower ethanol concentration (50 mM), however, did not significantly alter the free-radical induced cell death (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Ethanol reduces cell death due to H2O2 exposure. PC12 cells were treated for 30 min with 150 mM ethanol in the presence or absence of 2 mM H2O2, and cell death was determined using the fluorescent DNA-binding probe ethidium homodimer. Data are expressed as arbitrary fluorescent units/mg protein and are plotted as mean ± S.E.M. (*P < .05, **P < .01, N = 8).

Ethanol has been reported to increase both apoptosis and necrosis (Koop et al., 1997; Zhong et al., 1993). Experiments were therefore undertaken to determine which form of death was involved in the above studies. Because cells undergoing apoptosis characteristically contain low molecular-weight DNA fragments due to internucleosomal cleavage of the chromatin into 180- to 200-bp fragments (Evans and Cidlowski, 1995), the effect of ethanol on DNA fragmentation was evaluated. Serum withdrawal increased DNA fragmentation as indicated by a 149% increase in the relative amount of DNA in the 15,000 × g supernatant of the cell lysate. This extent of DNA fragmentation was further potentiated 74% by cotreatment with 150 mM ethanol (fig. 4). Ethanol exposure in the presence of 15% serum did not alter DNA fragmentation.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Ethanol potentiates DNA fragmentation after serum withdrawal. Cells were treated for 24 hr with 150 mM ethanol in the presence or absence of 15% serum. DNA fragmentation was determined using the fluorescent DNA-binding probe Hoechst-33342. Data are expressed as percent total DNA [i.e., DNAsup/(DNAsup + DNApellet)] and are plotted as mean ± S.E.M. (**P < .01, N = 6).

The effects of ethanol on DNA fragmentation in situ were determined using the TUNEL technique which labels free 3'-OH ends of the DNA in the cells. Evaluation of the number of labeled cells by phase-contrast microscopy demonstrated that serum withdrawal caused a significant (P < .05) increase in the number of apoptotic events (fig. 5). This increase was further potentiated 213% by the addition of 150 mM ethanol. Ethanol treatment in the presence of 15% serum did not alter the frequency of apoptotic events. This enhancement by ethanol of apoptosis induced by serum withdrawal was observed whether the data were normalized to total cellular protein (fig. 5) or total cellular DNA (data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Ethanol increases in situ DNA fragmentation after serum withdrawal. After a 24-hr treatment with 150 mM ethanol in the presence or absence of 15% serum, in situ DNA fragmentation was determined by the TUNEL method. Data are expressed as number of apoptotic events/mg protein and are plotted as mean ± S.E.M. (*P < .05, N = 5).

Because necrosis also can cause DNA fragmentation and lead to TUNEL-stained cells (Yasuda et al., 1995; Grasl-Kraupp et al., 1995; Charriaut-Marlangue et al., 1995), the fragmented DNA from the 15,000 × g supernatant was electrophoresed on agarose gels to determine whether the characteristic internucleosomal cleavage of the DNA into 180- to 200-bp fragments (i.e., DNA laddering) was present. Cells maintained in the absence of serum displayed the ~200-bp DNA ladder (fig. 6a) which is suggestive of apoptosis (Evans and Cidlowski, 1995). Quantification of the DNA laddering by summing the intensity of the DNA bands in each lane of the gel revealed that 150 mM ethanol significantly increased the laddering present after serum withdrawal 142% (fig. 6b). This enhancement by ethanol of DNA laddering induced by serum withdrawal was seen whether the data was normalized to total cellular protein (fig. 6b) or total cellular DNA (data not shown). Also, DNA from cells treated with 2 mM H₂O₂ did not exhibit any laddering (fig. 6a) indicating that this death was not apoptotic. Additional experiments were carried out to confirm that DNA laddering also occurred with 50 mM ethanol. For these studies, we used a phosphate-citrate DNA extraction method (Gong et al., 1994) which involves less tissue handling than extraction with phenol/chloroform. Quantification of the DNA laddering using this technique demonstrated that ethanol caused a concentration-dependent potentiation of apoptosis. A low ethanol concentration (50 mM) significantly potentiated DNA laddering 49% although a high ethanol concentration (150 mM) caused a 184% increase (fig. 7).

View larger version (47K): [in this window] [in a new window]   Fig. 6.   Ethanol treatment increases internucleosomal DNA cleavage after serum withdrawal. a, Cells were treated with 150 mM ethanol either for 24 hr in the presence or absence of 15% serum or for 30 min in the presence or absence of 2 mM H2O2. The DNA was then electrophoresed and visualized with ethidium bromide and UV trans-illumination. b, The extent of DNA laddering was determined by summing the intensity of the DNA bands in each lane of the gel. Data are expressed as arbitrary fluorescent units/mg protein and are plotted as mean ± S.E.M. (*P < .05, N = 6).

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Ethanol potentiates apoptosis due to serum withdrawal. After a 24-hr treatment with 50 or 150 mM ethanol in the absence of serum, the extent of DNA laddering was determined by summing the intensity of the DNA bands in each lane of the gel. Data are expressed as arbitrary fluorescent units/mg protein and are plotted as mean ± S.E.M. (*P < .05, N = 6).

To determine whether the ethanol-mediated cell death at low serum concentrations was apoptotic, cells were treated with 150 mM ethanol in the presence of 4% serum and the DNA in the 15,000 × g supernatant subjected to agarose gel electrophoresis. A portion of the cells maintained in 4% serum were undergoing apoptosis as indicated by cleavage of DNA into ~200-bp fragments (fig. 8a). Quantification of the DNA laddering showed that ethanol significantly increased apoptosis 66% (fig. 8b). Bhave and Hoffman (1997) reported that ethanol increased apoptosis in cerebellar granule cells by inhibition of the trophic action of NMDA. Inclusion of 100 µM NMDA in the treatment conditions, however, did not alter PC12 cell viability (data not shown).

View larger version (56K): [in this window] [in a new window]   Fig. 8.   Ethanol promotes apoptosis in cells maintained in the presence of 4% serum. a, After a 24-hr treatment with 150 mM ethanol in the presence of 4% serum, the DNA was electrophoresed and visualized with ethidium bromide and UV-transillumination. b, The extent of DNA laddering was determined by summing the intensity of the DNA bands in each lane of the gel. Data are expressed as arbitrary fluorescent units/mg protein and are plotted as mean ± S.E.M. (**P < .01, N = 6).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Neuronal survival requires the presence of neurotrophic factors. For example, during development a limiting supply of target-derived neurotrophic factors appears to be responsible for the normal loss of many of the neurons (Oppenheim, 1991; Raff et al., 1993). Removal of neurotrophic factors by withdrawal of serum in vitro causes cell death by apoptosis (see Deshmukh and Johnson, 1997). In our study the extent of PC 12 cell death induced by serum withdrawal was increased by ethanol. The ethanol-induced potentiation of cell death by serum deprivation does not appear to represent a generalized response resulting from imposing multiple insults on the PC 12 cells. Rather, the effects of ethanol on cell viability display specificity as inclusion of ethanol inhibited cell death induced by H₂O₂.

The enhancement of cell death by ethanol was concentration dependent and was observed with an ethanol concentration as low as 50 mM. This ethanol concentration is comparable to or less than the blood alcohol level in neonatal rodents that induces behavioral deficits (Thomas et al., 1996), alters neurotransmitter levels (Maier et al., 1996), retards brain growth (Maier et al., 1997) and causes loss of cerebellar Purkinje cells and granule cells (Napper and West, 1995; Goodlett et al., 1997). Human fetuses also may be exposed to a comparable concentration of alcohol. In one study at a community hospital emergency room, women who admitted recent drinking had a mean blood alcohol level of 260 ± 13 mg/dl (i.e., 56 mM), but showed little or no signs of intoxication (Urso et al., 1981). One mother whose child was diagnosed with the FAS had a blood alcohol concentration of 345 mg/dl (i.e., 81 mM) at the time of delivery (Church and Gerkin, 1988). Furthermore, much higher blood alcohol concentrations have been reported in women of child-bearing age (Hammond et al., 1973; Wells and Barnhill, 1996) including one report of a 24-yr-old woman with a blood alcohol level of 1510 mg/dl (i.e., 328 mM) who was described as "... agitated and slightly confused but alert, responsive to questioning, and oriented to person and place (though unclear as to time)" (Johnson et al., 1982).

In addition to increasing the extent of cell death caused by serum withdrawal, ethanol also induced PC 12 cell death. Similarly, Chen and Sulik (1996) observed that a 48-hr exposure to ethanol resulted in a concentration-dependent decrease in the viability of mouse neural crest cells in culture. Conversely, Luo and Miller (1997) did not observe any affect of ethanol (400 mg/dl; 87 mM) on cell death in three neuroblastoma cell lines. The effects of ethanol on cell viability, however, appear to involve a complex antagonistic interaction between the concentrations of ethanol and serum factors. Thus, no ethanol-induced death was observed in PC 12 cells with 150 mM ethanol in the presence of 15% serum, but when the serum levels were reduced to 4%, which maintains the cells without a loss of viability, 100 mM ethanol caused a significant increase in cell death. Also as shown in figure 2, the amount of cell death induced by ethanol displayed a dramatic increase as the serum levels were reduced to 1%. Others also have reported opposing actions of ethanol and trophic factors. Luo et al. (1997) reported that nerve growth factor and basic fibroblast growth factor, but not epidermal growth factor or insulin-like growth factor 1, diminished the ethanol-induced loss of 1-day-old cerebellar granule cells in culture. Cui et al. (1997) found that ethanol blocked the protective effect of insulin-like growth factor I on the cytotoxic action of tumor necrosis factor in BALB/c3T3 fibroblasts. Ethanol also blocked the proliferative effect of various mitogenic growth factors on neuroblastoma cells (Luo and Miller, 1997).

The ethanol-induced enhancement of cell death by serum deprivation involves an increase in the extent of apoptosis. Inclusion of ethanol increased the amount of DNA fragmentation and the number of TUNEL-stained cells after serum withdrawal. Furthermore, the extent of DNA laddering by serum deprivation was increased by ethanol. Similarly, ethanol initiates cell death by inducing apoptosis as shown by the DNA laddering in cells exposed to ethanol in the presence of 4% serum. An increase in DNA fragmentation and the number of TUNEL-stained cells was reported in hypothalamic cell cultures exposed to 200 mM ethanol (De et al., 1994). Addition of 100 mM ethanol also was reported to increase the amount of TUNEL-stained cells in cultures of cerebellar granule cells exposed to a nondepolarizing concentration of KCl (Bhave and Hoffman, 1997). This effect in the granule cells was secondary to an ethanol-induced inhibition of the trophic action of NMDA (Bhave and Hoffman, 1997). A different mechanism, however, is responsible for the ethanol-induced apoptotic death of the PC12 cells as inclusion of NMDA did not alter PC12 cell death. Ethanol also appears to induce apoptosis in vivo. Postnatal exposure of rats to ethanol was noted in a preliminary report to increase the number of TUNEL-stained cells in various brain regions (Hamby-Mason et al., 1996). Similarly, Bannigan and Burke (1982) reported the appearance of pyknotic nuclei and other morphology evidence of apoptosis in the brains of fetal mice that were exposed to ethanol.

The microencephaly observed in the FAS is due to a loss of neurons. Ethanol can reduce the number neurons by various mechanisms including an inhibition of proliferation (Miller, 1986, 1988; Pantazis et al., 1992; Kane et al., 1996) and an alteration in migration (Miller, 1993). The above discussion indicates that the ethanol-induced neuronal loss in vivo also may involve an increase in neuronal apoptosis. Compounding the ability of ethanol to elicit cell death by counteracting the effects of neurotrophic factors are reports that ethanol reduces the levels of some neurotrophic factors in vivo (Walker et al., 1992; Breese et al., 1993; MacLennan et al., 1995) although this has not been a consistent finding (Baek et al., 1994). Ethanol also has been reported to alter expression of some neurotrophic factor receptors including decreasing the amount of the low affinity neurotrophin receptor, p75 (Luo et al., 1996; Luo and Miller, 1997; Dohrman et al., 1997). However, because p75 that is not bound by ligand induced cell death (Barrett and Georgiou, 1996), the ethanol-induced increase in apoptosis is not consistent with a reduction in p75 levels. The induction of apoptosis by ethanol could involve the neurotrophin tyrosine kinase receptor, trkA, which promotes cell survival (Fagan et al., 1996), as this receptor was reported in some (Dohrman et al., 1997), but not all cases (Luo and Miller, 1997), to be decreased by ethanol. Also, because ethanol inhibits tyrosine kinase activity of the insulin-like growth factor 1 receptor in C6 rat glioblastoma cells and Balb/c 3T3 fibroblasts (Resnicoff et al., 1993, 1994), it is possible that a similar action on neurotrophin signaling could be involved in the ethanol-induced increase in apoptosis. However, neither a reduction in trkA expression nor an inhibition of receptor tyrosine kinase activity would appear to explain how ethanol potentiates cell death caused by serum withdrawal.