Glutathione content as a potential mediator of the vulnerability of cultured fetal cortical neurons to ethanol-induced apoptosis

Abstract

Ethanol ingestion during pregnancy elicits damage to the developing brain, some of which appears to result from enhanced apoptotic death of neurons. A consistent characteristic of this phenomenon is a highly differing sensitivity to ethanol within specific neuron populations. One possible explanation for this "selective vulnerability" could be cellular variations in glutathione (GSH) homeostasis. Prior studies have illustrated that ethanol elicits apoptotic death of neurons in the developing brain, that oxidative stress may be an underlying mechanism, and that GSH can be neuroprotective. In the present study, both multiphoton microscopy and flow cytometry demonstrate a striking heterogeneity in GSH content within cortical neuron populations. Ethanol differentially elicits apoptotic death and oxidative stress in these neurons. When neuron GSH content is reduced by treatment with butathione sulfoxamine, the ethanol-mediated enhancement of reactive oxygen species is exacerbated. Sorting of cells into high- and low-GSH populations further exemplifies ethanol-mediated oxidative stress whereby apoptotic indices are preferentially elevated in the low-GSH population. Western blot analysis of the low-GSH subpopulations shows higher ethanol-mediated expression of active caspase 3 and 24-kDa PARP-1 fragments compared with the high-GSH subpopulation. In addition, neuronal content of 4-hydroxynonenal adducts is higher in low-GSH neurons in response to ethanol. These studies suggest that GSH content is an important predictor of neuronal sensitivity to ethanol-mediated oxidative stress and subsequent cell death. The data support the proposition that the differences in proapoptotic responses to ethanol within specific neuron populations reflect a heterogeneity of neuron GSH content.

Fig. 1

Variable apoptotic responses to ethanol in the intact developing brain. A postnatal rearing model was utilized for comparisons of the percentage of apoptotic cells from controls (ad lib suckle control and isocaloric maltose-dextrin pair-fed control) and ethanol (5.6 g/kg/24 hr on postnatal days 4 and 5) in neuroepithelium (A; left) and choroid plexus (right), layers 2–3 of the cortex (B; left), CA3 region of the hippocampus (center) and dentate gyrus (right), brainstem (C; left) and cerebellum (right), dorsomedial hypothalamus (DMH; D) and medial habenular nucleus (MHN). Bars represent the means ± SEM. *Statistically significant difference from both controls (P < 0.05).

Fig. 2

Variability of ethanol-mediated oxidative stress in cultured fetal cortical neurons. Live neurons plated on coverslips were preloaded with DCF-DA and treated with ethanol (4 mg/ml). A: A representative confocal microscopy image collected at 18.3 min after ethanol exposure shows a wide variability in DCF fluorescence. B: This variability is represented in a bar graph format. The average fluorescence intensity per unit area was calculated for each of the seven cells and plotted, and a sixfold variation between the brightest and the faintest stained cell can be observed. C: This variability is illustrated in large populations (10,000 events) of neurons by flow cytometry. There was a broad range in intracellular ROS generation and the ROS content shifted in a positive direction with ethanol treatment (4 mg/ ml, 2 hr).

Fig. 3

Variability of apoptotic response to ethanol in cultured fetal cortical neurons. Apoptosis-related effects of ethanol on neuronal DNA in three random fields. Cells were treated with 4 mg/ml ethanol for 24 hr and assessed for DNA damage with the TUNEL assay as seen by FITC (green) stain. Cells were counterstained with DAPI (blue) to label all nuclei. Not all nuclei, as visualized by DAPI, showed TUNEL staining. The arrows in the middle column identify cells that are healthy, not susceptible to ethanol treatment, and do not colocalize with TUNEL staining as illustrated in the third column (merge).

Fig. 4

Heterogeneity of GSH content in cultured fetal cortical neurons. Neurons grown on coverslips were treated with ethanol (4 mg/ ml) for various times and labeled with 50 µM MCB (a GSH-sensing probe) for the final 15 min of incubation. A: Random images across the coverslips were collected using a multiphoton microscope. Representative images show the random distribution of GSH-related fluorescence among the neurons in control and in ethanol-treated neurons (15-min exposure). B: The graph shows the percentage of cells counted as a function of cellular GSH mean fluorescence intensity (arbitrary units). With ethanol treatment, a decrease in fluorescence (GSH) is observed along with transient redistribution of GSH at three early time points. The total number of cells in each group was as follows: control (n = 289) and ethanol exposure for 15 min (n = 284), 30 min (n = 321), and 60 min (n = 272).

Fig. 5

Ethanol- and BSO-related reductions of GSH content determined by flow cytometry. A: Representative flow cytometry analysis of neurons: control, GSH-depleted (200 µM BSO), ethanol-treated (4 mg/ml for 2 hr), or combined GSH-depleted and ethanol-treated. There is a broad range of intensity along the X-axis ranging from 10² to 10⁵ units. GSH depletion, ethanol exposure, or the two in combination caused a peak shift toward the left (lower GSH content). B: The bar graph depicts decreases in mean intensities (GSH content) of peak P1 (% gated peak) with respect to the controls. The values are an average from three independent experiments and are represented as mean ± SEM. ^*P < 0.01 compared with controls. ^**P < 0.05 for ethanol compared with ethanol combined with BSO.

Fig. 6

Ethanol-mediated oxidative stress is enhanced by GSH depletion. Flow cytometry experiments (10,000 events) illustrated the effects of GSH depletion, alone and in combination with ethanol, on ROS production. Neurons were depleted of GSH by overnight treatment with 200 µM BSO and/or were treated with ethanol (4 mg/ml, 2 hr). The values are an average of n = 6, represented as mean ± SEM. ^*Difference from controls (P < 0.05). ^**P < 0.05 for ethanol compared with ethanol combined with BSO.

Fig. 7

Ethanol-mediated DNA damage is enhanced by GSH depletion. Cortical neurons were grown on coverslips and assessed by TUNEL assay for DNA damage. The fluorescence intensity of each field was assessed in ImageJ and divided by the number of cells per field. Treatments consisted of control, GSH-depleted (200 µM BSO), ethanol-treated (4 mg/ml, 2 hr), and ethanol-treated cells that had been pretreated with BSO. In total, 180–225 cells were counted for each group. Values are mean fluorescence intensities/cell ± SEM from four image scans (^*P < 0.05 compared with controls). ^**P < 0.05 for ethanol compared with ethanol combined with BSO.

Fig. 8

HNE-protein adduct formation is enhanced in ethanol-exposed neurons with low GSH content. A: Neurons treated with ethanol (4 mg/ml, 2 hr) and stained for GSH with MCB (green) and HNE-conjugated to Alexa 633 (red) were imaged using confocal microscopy. Neurons with high GSH content typically had less HNE adduct staining compared with those expressing low GSH-related fluorescence. B: Neurons treated with ethanol (4 mg/ml, 2 hr) were sorted into two subpopulations based on their GSH content: “low” for cells in the lowest 15% and “high” for cells with the 15% highest GSH content. Expression of neuronal HNE adducts (an estimate of lipid peroxidation) is amplified in the low-GSH subpopulation. Tubulin expression served as internal controls to ensure correct protein loading between lanes. This blot represents one of three similar experiments.

Fig. 9

Induction of apoptosis markers is enhanced in ethanol-exposed neurons with low GSH content. A: Neurons were treated with ethanol (4.0 mg/ml, 2 hr), stained with MCB, and sorted by FACS. Two subpopulations based on their low and high GSH content were collected and processed for Western blotting. Compared with the high-GSH neurons, there is an increased expression of “active” caspase-3 (11-kDa fragment) and the 24-kDa PARP-1 along with a decrease in procaspase-3 in the subpopulation with low GSH content. Tubulin expression levels served as internal controls to check protein loading. This blot represents one of three similar experiments. B: Cultured neurons (not sorted by FACS) were treated with BSO (200 µM, 24 hr) and ethanol (4.0 mg/ml, 2 hr), then processed for Western blotting. Expression of the 11-kDa caspase-3 fragment was increased in the ethanol-treated cells (4 mg/ml, 2 hr) and in ethanol-exposed GSH-depleted neurons. BSO, BSO treatment alone; ethanol treatment alone; BSO + Ethanol, ethanol treatment (4 mg/ml, 2 hr) and BSO (200 µM, 24 hr). Tubulin was used as the loading control. This blot represents one of three identical experiments.