miR-497 and miR-302b Regulate Ethanol-induced Neuronal Cell Death through BCL2 Protein and Cyclin D2*

Background: Ethanol-induced neuronal apoptosis causes brain shrinkage and cognitive defects. Results: Exposure to ethanol (0.5% v/v for 72 h) in SH-SY5Y cells induced expression of miR-497 and miR-302b and down-regulated expression of BCL2 and/or cyclin D2. Conclusion: Ethanol-induced neuronal apoptosis follows both the mitochondria-mediated and non-mitochondria-mediated pathways. Significance: Our study shows that miRNAs are involved in regulation of ethanol neurotoxicity. In chronic alcoholism, brain shrinkage and cognitive defects because of neuronal death are well established, although the sequence of molecular events has not been fully explored yet. We explored the role of microRNAs (miRNAs) in ethanol-induced apoptosis of neuronal cells. Ethanol-sensitive miRNAs in SH-SY5Y, a human neuroblastoma cell line, were identified using real-time PCR-based TaqMan low-density arrays. Long-term exposure to ethanol (0.5% v/v for 72 h) produced a maximum increase in expression of miR-497 (474-fold) and miR-302b (322-fold). Similar to SH-SY5Y, long-term exposure to ethanol induced miR-497 and miR-302b in IMR-32, another human neuroblastoma cell line. Using in silico approaches, BCL2 and cyclin D2 (CCND2) were identified as probable target genes of these miRNAs. Cotransfection studies with 3′-UTR of these genes and miRNA mimics have demonstrated that BCL2 is a direct target of miR-497 and that CCND2 is regulated negatively by either miR-302b or miR-497. Overexpression of either miR-497 or miR-302b reduced expression of their identified target genes and increased caspase 3-mediated apoptosis of SH-SY5Y cells. However, overexpression of only miR-497 increased reactive oxygen species formation, disrupted mitochondrial membrane potential, and induced cytochrome c release (mitochondria-related events of apoptosis). Moreover, ethanol induced changes in miRNAs, and their target genes were substantially prevented by pre-exposure to GSK-3B inhibitors. In conclusion, our studies have shown that ethanol-induced neuronal apoptosis follows both the mitochondria-mediated (miR-497- and BCL2-mediated) and non-mitochondria-mediated (miR-302b- and CCND2-mediated) pathway.

In chronic alcoholism, brain shrinkage and cognitive defects because of neuronal death are well established, although the sequence of molecular events has not been fully explored yet. We explored the role of microRNAs (miRNAs) in ethanol-induced apoptosis of neuronal cells. Ethanol-sensitive miRNAs in SH-SY5Y, a human neuroblastoma cell line, were identified using real-time PCR-based TaqMan low-density arrays. Longterm exposure to ethanol (0.5% v/v for 72 h) produced a maximum increase in expression of miR-497 (474-fold) and miR-302b (322-fold). Similar to SH-SY5Y, long-term exposure to ethanol induced miR-497 and miR-302b in IMR-32, another human neuroblastoma cell line. Using in silico approaches, BCL2 and cyclin D2 (CCND2) were identified as probable target genes of these miRNAs. Cotransfection studies with 3-UTR of these genes and miRNA mimics have demonstrated that BCL2 is a direct target of miR-497 and that CCND2 is regulated negatively by either miR-302b or miR-497. Overexpression of either miR-497 or miR-302b reduced expression of their identified target genes and increased caspase 3-mediated apoptosis of SH-SY5Y cells. However, overexpression of only miR-497 increased reactive oxygen species formation, disrupted mitochondrial membrane potential, and induced cytochrome c release (mitochondriarelated events of apoptosis). Moreover, ethanol induced changes in miRNAs, and their target genes were substantially prevented by pre-exposure to GSK-3B inhibitors. In conclusion, our studies have shown that ethanol-induced neuronal apoptosis follows both the mitochondria-mediated (miR-497-and BCL2-mediated) and non-mitochondria-mediated (miR-302b-and CCND2-mediated) pathway. 4 are small regulatory RNA molecules that down-regulate the expression of protein coding genes in a sequence-specific manner. MiRNAs regulate the expression of several genes that are important for apoptosis, senescence, aging, neuronal differentiation, neurogenesis, and neurodegeneration (1,2). Knockout studies of dicer (the endoribonuclease responsible for the formation of mature miRNAs) have shown that miRNAs are required for the survival of neural progenitor cells and differentiation and maturation of postmitotic neurons (3,4). Ethanol is a well known neurotoxin for both the developing and the adult brain (5). In the developing brain, only a single dose of ethanol can induce widespread apoptotic neurodegeneration (6). Over-expression of antiapoptotic BCL2 can protect the neonatal cerebellum from ethanol-induced apoptosis (7). In adults, chronic use of ethanol decreases neurogenesis of the hippocampus, which is the major site of adult neurogenesis and ethanol neurotoxicity (8 -10).

MicroRNAs (miRNAs)
Recent studies have clearly shown the role of miRNAs in adult and embryonic neurogenesis (11,12). Levels of cyclin D2 (CCND2), which plays a crucial role in the maintenance of the neurogenesis potential of the adult brain, are directly regulated by miR-302b (13,14). Pietrzykowski et al. (15) have shown that neuronal adaptation to ethanol is regulated by miR-9. Studies have also shown that competing interactions between a set of miRNAs can decide the fate of ethanol-exposed neural progenitor cells (16). Ethanol-induced neurotoxicity can be prevented substantially by pre-exposure to glycogen synthase kinase 3 ␤ (GSK-3〉) inhibitors (17). The GSK-3 isoforms, GSK-3A and GSK-3B are constitutively active serine/threonine protein kinases of the WNT/␤-catenin signaling cascade. GSK-3 regulates protein translation, promotion of mitochondrial apoptosis, and levels of other signaling elements like cyclin D1 and D2 by activation of different transcription factors (18 -20). In the brain, GSK-3B plays a crucial role in the regulation of neurogenesis, neuronal survival, and neurite outgrowth (21). Moreover, exposure to GSK-3B inhibitors (LiCl 3 or valproate) has been shown to alter the expression profile of selective miRNAs (down-regulated: let-7b, let-7c, miR-128a, miR-24a, miR-30c, miR-34a, and miR-221; up-regulated: miR-144) that target proteins involved in neurite outgrowth; neurogenesis; and the phosphatase and tensin homolog (PTEN), extracellular signalregulated kinases (ERK), and Wnt/␤-catenin pathways (22).
Studies were initiated to explore the involvement of miRNAs in ethanol-induced neuronal death by silencing dicer in SH-SY5Y, a human neuroblastoma cell line. Increased sensitivity of dicer-silenced SH-SY5Y cells toward ethanol exposure has prompted us to identify ethanol-sensitive miRNAs. Further studies were concentrated on two miRNAs that showed maximum alterations after long-term exposure to ethanol. Potential target genes of these miRNAs were identified and validated by 3Ј-UTR binding assays, and their role in ethanol-induced neuronal apoptosis was investigated. Moreover, the effect of preexposure to two well known GSK-3〉 inhibitors (lithium and DZD-8) was also studied on ethanol-induced alterations in miRNAs and their target genes.
Cell Culture and Exposure to Chemicals-SH-SY5Y cells were grown in a 1:1 mixture of Eagle's minimum essential medium with non-essential amino acids and Ham's F12 medium. IMR-32 human neuroblastoma cells were grown in Eagle's minimum essential medium with 2 mM L-glutamine and non-essential amino acids. The growth media of both cell lines were supplemented with 10% fetal bovine serum and 1% antibiotic and antimycotic solution. Cells were kept in 5% CO 2 95% atmosphere with high humidity at 37°C. Considering the volatile properties of ethanol, exposure to ethanol in a 96-well plate was given as described earlier (23). For short-term exposure studies, cells were exposed to 0.5, 1.5, 2.5, 3.5, 4.5, or 5.5% v/v ethanol for 4 h. For long-term exposure studies, cells were exposed to 0.125, 0.25, 0.5, 1.0, 2.0, or 4.0% v/v ethanol for 72 h. A non-cytotoxic concentration of ethanol was selected by car-rying out MTT and neutral red uptake (NRU) assays as described in our earlier paper (24). GSK-3B inhibition was carried out by exposing cells to either TDZD-8 (10 M) or LiCl 3 (20 mM) as reported earlier (17,25).
Dicer Silencing Studies-SH-SY5Y cells were transfected with siRNA for dicer (target sequence: anti-Dicer 1, ACCACT-GATTCTGCATATGAA or anti-Dicer 2, AATGTGCTATCT-GGATCCTAG) or NTC siRNA using the Cell Line Nucleofactor Kit V. After 24 h of transfection, a decrease of more than 80% in expression of dicer mRNA and protein was achieved by using anti-Dicer 1 siRNA (supplemental Fig. S1). After 24 h of transfection, cells were exposed with different concentrations of ethanol, and cell viability was assayed by MTT and NRU assays.
Expression Profiling of miRNAs-The total RNA, which also contained small RNAs, was isolated by means of the mirVana miRNA isolation kit as described by the manufacturer. The expression of 667 unique miRNAs was studied by TaqMan lowdensity arrays pool A and pool B (TaqMan Human MicroRNA Set Cards v2.0, Part no. 4400238). Primers of 667 miRNAs were present on a set of two plates named pool A (containing primers for 380 unique miRNAs) and pool B (containing primers for 287 unique miRNAs). After RNA isolation, RT was carried out with megaplex RT primers of pool A (part no. 4399966) and pool B (part no. 4399968), which are a set of two predefined pools containing stem-looped reverse transcription primers for mature miRNAs. RT was performed with the TaqMan miRNA RT kit from ABI as described by the manufacturer. In brief, the reaction mixture contained 1ϫ megaplex RT primers, 2.6 mM deoxynucleotide triphosphates (dNTPs) with deoxythymidine triphosphate (dTTP), 75 units of MultiScribe reverse transcriptase, 1ϫ buffer, 3 mM MgCl 2 , 2 units of RNase inhibitor, and 100 ng of total RNA with a final volume of 7.5 l. Thermal cycling conditions used for reverse transcription involved 40 cycles of 16°C for 2 min, 42°C for 1 min, and 50°C for 1 s. Finally, the reaction mixture was denatured by incubating it at 85°C. After reverse transcription, preamplification was carried out by means of the TaqMan preamp master mix and megaplex preamp primers of pool A (part no. 4399233) and pool B (4399201). 12.5 l of preamp master mixture was mixed with 2.5 l of preamp primers and 2.5 l of reverse transcription product with a total volume of 25 l. Thermal cycling conditions used for preamplification were 10 min at 95°C, 2 min at 55°C, 2 min at 72°C, and 12 cycles of 15 s at 95°C, and 4 min at 60°C. Finally, the preamp products were incubated for 10 min at 99.9°C. The preamplification products were diluted up to four times with 0.1 ϫ Tris-EDTA buffers. For running TLDA plates, 450 l of the TaqMan universal PCR master mixture was mixed with 9 l of the diluted preamplification product in a total volume of 900 l. In each port of the TLDA plates, 100 l from the above-mentioned master mixture was added, and after sealing and spinning, the plates were loaded on a thermal cycler. Relative quantification was done by means of the -⌬⌬ cycle threshold method, considering the levels of mammalian U6 RNA as endogenous control. The expression of miR-497 and miR-302b were studied using individual TaqMan miRNA assays.
Luciferase Assay-The effect of miR-497 or miR-302b binding on 3Ј-UTR of BCL2 or CCND2 was studied by cotransfection of 3Ј-UTR and miRNA mimics. 3Ј-UTR of BCL2 (1-2751 bp after the stop codon of NM_000633) and CCND2 (536 -2137 bp after the stop codon of NM_001759) were cloned in the pEZX-MTO1 vector, which contains firefly luciferase as the reporter gene (controlled by the SV40 promoter) and Renilla luciferase as the tracking gene (controlled by the CMV promoter). Both firefly and Renilla luciferase activities were measured by means of the Dual-Glo luciferase assay kit at 10-min intervals. Maximal luciferase activity was calculated by normalizing firefly luciferase activity with Renilla luciferase activity.

Flow Cytometry Studies
Cell Death-Cell death in SH-SY5Y and IMR-32 cells was measured by means of FITC-labeled annexin V. Cells were transfected with either NTC or syn-miR-497 or syn-miR-302b. After 72 h of transfection, the cells were detached from the culture plate, washed twice with PBS, and suspended in 100 l of binding buffer. 5 l of annexin V was added, and the cells were incubated in the dark for 15 min. Subsequently, 5 l of propidium iodide (PI) was added, and cell death was measured using flow cytometry (BD FACScanto II).
Reactive Oxygen Species (ROS) and Mitochondrial Membrane Potential (MMP) Measurement in SH-SY5Y -ROS formation was measured using 2Ј,7Ј-dichlorodihydrofluorescein diacetate, a non-fluorescent probe that is converted into a highly fluorescent (2Ј,7Ј-dichlorofluorescein) molecule in the presence of ROS. In brief, cells were transfected with miRNA mimics or NTC and cultured for 72 h. Subsequently, cells were detached from wells and incubated in phosphate buffer saline containing 20 M 2Ј,7Ј-dichlorodihydrofluorescein diacetate in the dark at 37°C for 30 min. After 30 min. the cells were pelleted, resuspended in PBS, and analyzed by flow cytometry. Changes in MMP were studied using JC-1, a cationic dye that shows potential dependent accumulation in mitochondria, indicated by a fluorescence emission shift from green to red. In brief, cells were detached from 6-well plates and incubated in an incomplete medium containing 10 M JC-1 in the dark at 37°C for 20 min. After 20 min, cells were pelleted and resuspended in PBS and analyzed by flow cytometry or a fluorescence microscope.
Caspase 3 Activity and Cytochrome c Release-Total cell lysates of SH-SY5Y cells transfected with miRNA mimics or NTC were prepared using Cell Lytic M reagent. Caspase 3 activity was measured in cell lysates using fluorogenic substrate Ac-DEVD-AMC of caspase 3 from BD Pharmingen as described by the manufacturer. Cytochrome c release was studied by immunoblotting of cytochrome c in the cytosolic and mitochondrial fraction of SH-SY5Y cells and transfected with miRNA mimics or NTC. Mitochondrial fractions were prepared using mitochondrial protein isolation buffer from AMRESCO supplemented with protease inhibitor mixture and DTT. Prior to immunoblotting of cytochrome c, immunoblotting of ␤-actin and VDAC-1 were performed to establish the purity of mitochondrial and cytosolic fractions. Levels of ␤-actin and VDAC-1 were also used for normalization of protein loading.
Real-time PCR of BCL2, CCND2, and ␤-Actin-RT was performed with the high-capacity cDNA kit from ABI. The reac-

FIGURE 1. Effect of either short-term (4 h, A and B) or long-term (72 h, C and D) ethanol exposure on viability of SH-SY5Y cells transfected with NTC
siRNA or dicer siRNA. Cell viability is measured by MTT (A and C) and NRU (B and D) assays. The viability of cells transfected with NTC siRNA and exposed to different concentration of ethanol (either short-term or long-term) is compared with the viability of cells transfected with NTC siRNA and not exposed to ethanol (control ϭ 100%). Similarly, the viability of cells transfected with dicer siRNA and exposed to ethanol (either short-term or long-term) is compared with the viability of cells transfected with dicer siRNA and not exposed to ethanol. All values are mean of four different experiments, and the error bars represent ϩ S.E. of the mean. Statistical significance between control and different ethanol exposed groups was calculated by one-way analysis of variance followed by a post hoc Tukey test. Values showing p Ͻ 0.05 in the Tukey test are considered statistically significant and are marked a, b, c, d, e, and f. (Dicer K D : dicer knockdown (KD)).
tion mixture contains 1ϫ RT buffer, 8 mM dNTP mix, 1ϫ RT random primers, and 1 l each of MultiScribe reverse transcriptase and RNase inhibitor, with a total volume of 20 l, which also includes 1 g of total RNA. The thermal cycling conditions used for RT were 10 min at 25°C, 120 min at 37°C, and 5 s at 85°C. Real-time PCR was carried out using SYBR Green chemistry. In brief, the reaction mixture contained 1ϫ SYBR Green master mixture, 1 l of cDNA, and 0.50 m of forward and reverse primers, with a total volume of 20 l. Primers of BCL2 and CCND2 were designed with Primer Express software from ABI. The sequence of primers used was as follows. BCL2: forward, 5Ј-CTGAGTACCTGAACCGGC-ACC-3Ј and reverse, 5Ј-GAGCAGAGTCTTCAGAGACAG-3Ј. CCND2: forward, 5Ј-GTTCCTGGCCTCCAAACTCA-3Ј and reverse, 5Ј-CTTGATGGAGTTGTCGGTGTAAAT-3Ј.
Immunoblotting-The total cell lysates were prepared with CellLytic M cell lysis reagent from Sigma and supplemented with a protease inhibitor mixture and DTT. Immunoblotting was performed as described in our earlier paper (26). In brief, protein samples were subjected to SDS-PAGE (3% acrylamide stacking gel and 12 or 15% acrylamide separating gel). After electrophoresis, proteins were transferred to a nitrocellulose membrane. The membrane was incubated overnight with a primary antibody (1:1000 dilution) at 4°C and for 1 h at room temperature. HRP-conjugated secondary antibodies were used at a 1:5000 dilution for protein detection. Blots were developed with the Pierce Supersignal West Pico horse-peroxidase reagent.
Statistical Analysis-Where comparisons had to be made between only two groups, Student's t test was employed to calculate the statistical significance. p Ͻ 0.05 in Student's t test was considered significant. A one-way analysis of variance followed by a post hoc Tukey test were conducted for pairwise comparisons made between more than two groups.

Dicer Silencing Increased Ethanol Toxicity in SH-SY5Y Cells-
Dicer-silenced cells showed a higher rate of cell death compared with cells transfected with NTC after either short-term or long-term exposure to ethanol (Fig. 1). In NTC-transfected cells, short-term exposure to 5.5% v/v ethanol and long-term exposure to 4 and 2% of ethanol reduced cell viability below 50% (Fig. 1). In dicer-silenced cells, short-term exposure to 5.5 and 4.5% v/v ethanol and long-term exposure to 4, 2, 1, and 0.5% v/v ethanol reduced cell viability below 50% (Fig. 1, A-D). For further studies, 2.5% (for 4 h) and 0.5% (for 72 h) of ethanol were selected as short-term and long-term exposure concentrations, respectively (Fig. 1).  Fig. 1, A and C, represents the set of miRNAs present in pool A of the TLDA plates. Fig. 1, B and D, represents the set of miRNAs present in pool B of the TLDA plates. For short-term exposure, cells are exposed to 2.0% v/v ethanol for 4 h, and for long-term exposure, cells were exposed to 0.
Identification of BCL2 and CCND2 as Potential Target Genes Involved in Ethanol-induced Neuronal Apoptosis-After observing a significant increase (p Ͻ 0.05) in apoptosis by ectopic expression of miR-497 or miR-302b, we generated a list of potential target genes of miR-497 and mir-302b. After scanning of in silico-predicted targets, we identified BCL2 and CCND2 as possible regulators of ethanol-induced neuronal apoptosis. Interestingly, scanning of 3Ј-UTR of BCL-2 and CCND2 (supplemental Fig. 2) showed the presence of target sites for several miRNAs (BCL2: miR-497, miR-204, miR-181, miR-182, miR-449, and miR-365; CCND2: miR-497, miR-302, miR-204, miR-182, miR-503, and let 7) that were induced at least 2-fold by long-term exposure to ethanol. Transfection of the miR-497 mimic in SH-SY5Y cells reduced the levels of BCL2 and CCND2 mRNAs up to 34 and 41%, respectively (Fig.   4B). While transfection of the miR-302b mimic in SH-SY5Y cells reduced the expression of CCND2 by up to 32%, no significant change was observed in expression of BCL2 (Fig. 4B). In line with real-time PCR results, ectopic expression of miR-497 reduced the levels of BCL2 and CCND2 proteins, and ectopic expression of miR-302b reduced the levels of CCND2 protein (Fig. 4, C and D). Moreover, in SH-SY5Y cells, cotransfection of Syn-miR-497 with 3Ј-UTR of BCL2 significantly decreased (up to 72%) maximal luciferase activity in comparison to cells cotransfected with NTC (Fig. 4, E-I). As expected, cotransfection of Syn-miR-302b with 3Ј-UTR of BCL2 did not produce a significant change in maximal luciferase activity (Fig. 4, E-I). However, cotransfection of 3Ј-UTR of CCND2 with mimics of either miR-497 or miR-302b significantly reduced maximal luciferase activity by up to 41 and 26%, respectively (Fig. 4, E-II).
Overexpression of miR-497 Increased ROS Formation, Disrupted MMP, and Induced Cytochrome c Release-Flow cytometry analysis of SH-SY5Y cells transfected with the miR-497 mimic showed a significant increase (6.6%) in ROS formation (Fig. 5, A and B). In accordance with ROS formation, overexpression of miR-497 also increased the MMP loss by up to 6%, which was 1.5% in cells transfected with NTC (Fig. 5, C and D).

. Ectopic expression of miR-497 or miR-302b induced caspase 3 activity (A) and down-regulated expression of BCL2 and/or CCND2 (B-D) by targeting 3Ј-UTR of these genes (E-I and E-II). A, caspase 3 activity measured in lysates of cells transfected with NTC (control ϭ 100%), Syn-miR-497, or
Syn-miR-302b using the fluorogenic substrate Ac-DEVD-AMC. Fluorescence was measured using a multiple-well plate reader at an excitation wavelength of 380 nm and an emission wavelength of 450 nm. B, effect of ectopic expression of miR-497, miR-302b, or NTC (control ϭ 1) on transcript levels of BCL2 or CCND2 as measured by real-time PCR. Expression of BCL2 and CCND2 in cells transfected with Syn-miR-497 or Syn-miR-302b is compared with their expression in cells transfected with NTC (control ϭ 1). C, effect of ectopic expression of miR-497, miR-302b, or NTC on protein levels of BCL2 or CCND2 measured by immunoblotting in cell lysates of SH-SY5Y cells. Fluorescence imaging of cells transfected with mimics of miR-497 clearly showed the presence of the green JC-1 dye, and this validated our flow cytometry results (Fig. 5E). Overexpression of miR-302b in SH-SY5Y cells did not produce any significant change in ROS formation or MMP loss (Fig. 5, A-E). Similarly, although overexpression of only miR-497 induced the release of cytochrome c from mitochondria to the cytosol, overexpression of miR-302b did not produce any significant change in the localization of cytochrome c (Fig. 6, B and C).

Effect of GSK-3〉 Inhibition on Ethanol-mediated Regulation of miR-497, miR-302b, and Their Target Genes in SH-SY5Y
Cells-Pre-exposure to GSK-3〉 inhibitors significantly (p Ͻ 0.05) prevented ethanol-mediated up-regulation of miR-497 and miR-302b (Fig. 8, A and B). Short-term and long-term exposure to ethanol produced a 20-and 380-fold increase in miR-497 expression, respectively. Pre-exposure with LiCl 3 or TDZD-8 prevented approximately 90% of the increase found in expression of miR-497 after either short-term exposure (20fold was reduced to 3-fold) or long-term exposure (380-fold was reduced to 40-fold) to ethanol (Fig. 8A). Although longterm exposure to ethanol induced the expression of miR-302b up to 280-fold, short-term exposure did not alter the expression of miR-302b (Fig. 8B). Pre-exposure with LiCl 3 or TDZD-8 decreased the ethanol-induced expression of miR-302b up to 22-fold (LiCl 3 ) and 27-fold (TDZD-8) (Fig. 8B). In accordance with induction of miR-497, expression of BCL2 (the target gene of miR-497) decreased by 72 and 38% after long-term or short- term exposure to ethanol, respectively (Fig. 8C). Pre-exposure with LiCl 3 and TDZD-8 significantly prevented the ethanolmediated decrease in mRNA expression of BCL2 after either short-term (increased from 72 to 89%) or long-term (increased from 38 to 93%) exposure to ethanol (Fig. 8C). Similarly, when cells were pre-exposed with TDZD-8, expression of BCL2 decreased significantly after either short-term or long-term exposure to ethanol (Fig. 8C). Short-term exposure to ethanol did not produce any significant alteration in expression of CCND2, but long-term exposure reduced the expression of CCND2 up to 45%. When these cells were pre-exposed with LiCl 3 or TDZD-8, expression of CCND2 decreased up to 85 and 90%, respectively (Fig. 8D). In line with real-time PCR results, our immunoblotting studies also showed a significant decrease in the expression of BCL2 (53%) and CCND2 (66%) after longterm exposure to ethanol (Fig. 8, E and F). Further pre-exposure to LiCl 3 and TDZD-8 significantly (p Ͻ 0.05) prevented an ethanol-meditated decrease in the levels of BCL2 and CCND2 (Fig.  8, E and F).

DISCUSSION
Increased Sensitivity of dicer-silenced SH-SY5Y cells to ethanol indicated the involvement of miRNAs in ethanol neuro-toxicity. In dicer-silenced SH-SY5Y cells, long-term exposure to low dosages of ethanol has shown more toxicity than shortterm exposure to higher dosages of ethanol. Similarly, longterm exposure to a low dose of ethanol (0.5% v/v for 72 h) altered the expression of more miRNAs (28 miRNAs, 3-fold and p Ͻ 0.05) than short-term exposure (13 miRNAs, 3-fold and p Ͻ 0.05) of a high dose (2.5% v/v for 4 h) of ethanol in SH-SY5Y cells. MiR-369 -3p, miR-493, miR-518f, and miR-370 were found to be specifically up-regulated by short-term exposure to ethanol, whereas expression of miR-302b, miR-373, miR204, miR-208b, miR-432, miR-375, miR-936, and miR-483-5p were up-regulated specifically by long-term exposure to ethanol. Long-term exposure to ethanol also induced the expression of miR-204 (23.17-fold) and miR-375 (7.4-fold), which are known to express at higher levels in the brain (27,28). Sathyan et al. (16) have shown that ethanol at levels found in alcoholics alters the expression of miR-21, miR-335, miR-9, and miR-153 in neurospheres derived from neuroepithelial cells of a fetal mouse cerebral cortex at gestational day 12.5. Similar to SH-SY5Y cells, long-term exposure to ethanol substantially induced expression of miR-497 and miR-302b in IMR-32 cells.
In SH-SY5Y cells, mir-497 was found to be particularly sensitive to ethanol exposure and exhibited a 21-and 474-fold increase in expression after short-term and long-term exposure, respectively.
Ectopic Expression of miR-497 in SH-SY5Y cells increased the rate of early and late apoptosis up to 2.37-and 5.90-fold, respectively. Similarly, ectopic expression of miR-302b increased the rate of early and late apoptosis up to 2.12-and 3.3-fold, respectively. Moreover, overexpression of either miR-497 or miR-302b in SH-SY5Y cells induced the activity of caspase 3, an effector caspase of apoptosis. Interestingly, several miRNAs (miR-497, miR-449a, miR-192, miR-139 -5p, miR-146a, miR-328, let-7b, and miR-342-3p) that were induced by ethanol have been reported previously to target apoptosis-related genes in other systems or cells (29,30). Using the Target-Scan web portal, we generated list of potential target genes of miR-497 or miR-302b and selected BCL2 and CCND2 as possible candidate genes involved in the regulation of ethanol-induced neuronal apoptosis. The presence of target sites for several miRNAs in 3Ј-UTR of BCL2 (miR-497, miR-204, miR-181, miR-182, miR-449, and miR-365) and CCND2 (miR-497, miR-302, miR-204, miR-182, miR-503, and let 7) that are induced by long-term exposure to ethanol have supported our selection. A significant decrease in maximal luciferase activity in neuronal cells cotransfected with 3Ј-UTR of BCL2 and Syn-miR-497 provided further evidence that BCL2 is directly targeted by miR-497. The maximum luciferase activity of plasmids containing 3Ј-UTR of CCND2 decreased because of cotransfection of either miR-497 or miR-302b mimics. Also, the decreased expression of BCL2 mRNAs and proteins in cells transfected with miR-497 mimics has confirmed the direct role of miR-497 in down-regulation of BCL2. Similarly, the decreased expression of CCND2 mRNA and protein in SH-SY5Y cells transfected with mimics of either miR-497 or miR-302b has proved that levels of CCND2 are regulated by both miR-497 and miR302b. Overexpression of miR-497, which targets BCL2, also induced ROS formation and MMP loss. Down-regulation of antiapoptotic protein BCL2 along with increased ROS forma-tion and MMP loss indicates involvement of a mitochondriamediated pathway of apoptosis (31). Also, the increased release of cytochrome c in cells overexpressing miR-497 confirms the For calculating fold change in total cell death, the fold change in early apoptosis, late apoptosis, and necrosis is calculated individually and added. The fold change in early apoptosis, late apoptosis, and necrosis of E-, LϩE-, or TϩE-exposed groups is calculated against the vehicle control. Significant changes are calculated by one-way analysis of variance followed by a pairwise Tukey test. Values showing p Ͻ 0.05 in the Tukey test are considered statistically significant. a, p Ͻ 0.05 when compared with vehicle control; b, p Ͻ 0.05 when compared with ethanol-exposed cells). , and CCND2 (D) measured by real-time PCR in SH-SY5Y cells exposed with GSK-3B inhibitors and ethanol. E, immunoblotting of BCL2, CCND2, and ␤-actin in total cell lysates of SH-SY5Y cells exposed with GSK-3B inhibitors and ethanol. F, densitometry of immunoblots. The concentrations used for exposure are ethanol (E), 0.5% ϫ 72 h; lithium (L), 20 mM; and TDZD-8 (T), 10 M. The expression of miR-497 and miR-302b was measured by using TaqMan assays, whereas expression of BCL2 and CCND2 was measured using SYBR Green chemistry as mentioned under "Experimental Procedures." The fold change in expression of miR-497 and miR-302b was calculated by considering their levels as 1-fold in cells exposed to vehicle control. Significant changes are calculated by one-way analysis of variance followed by a pairwise Tukey test. a, p Ͻ 0.05 versus vehicle control; b, p Ͻ 0.05 versus ethanol. involvement of a mitochondrial pathway of apoptosis. Studies carried out by Young et al. (32) have shown that ethanol-induced neuronal apoptosis is an intrinsic pathway-mediated phenomenon that involves disruption of mitochondrial membrane potential, cytochrome c release, and caspase 3 activation. Similar to our results, a recent study has shown that induction of miR-497 in response to ischemia and oxygen-glucose deprivation induces neuronal cell death by down-regulating BCL2 (33). BCL2-mediated neuronal apoptosis is also known to be regulated by miR-34a (34,35). Although no alteration was observed in expression of miR-34a, a significant and substantial increase was found in the expression of miR-34a * (a minor strand of pre-miR-34) after either short-term (25.11-fold) or long-term (36.9-fold) exposure to ethanol. Overexpression of miR-302b also induced the caspase 3-mediated apoptosis of SH-SY5Y cells. However, overexpression of miR-302b did not produce significant alterations in ROS formation, MMP loss, and cytochrome c localization, which indicates that miR-302binduced neuronal apoptosis is not mitochondria-mediated.
Overexpression of either miR-497 or miR-302b decreased the expression of CCND2, a cyclin also known to regulate neurogenesis in the adult brain (13). CCND2 regulates the cell cycle by exporting p27 (a cell cycle inhibitor known to control cell cycle progression at G 1 ) from the nucleus to the cytoplasm, which results in degradation of p27 at the G 0 -G 1 transition of the cell cycle (36). Our cotransfection studies have shown that miR-302b directly targets 3Ј-UTR of CCND2 in neuronal cells. Earlier studies that used gain and loss of function approaches showed that miR-302b is involved in the regulation of stemness and inhibits differentiation of neuronal cells by targeting CCND2 (14). Moreover, CCND2 is the only cyclin isoform known to express in the hippocampus and regulates neurogenesis in the hippocampus of the adult brain (13). Chronic use of ethanol is known to target the hippocampus and its associated physiological functions (37). Several studies have shown that miRNAs play an important role in the regulation of neurogenesis (11,38,39), and interestingly, our in silico studies (supplemental Fig. 3) have also identified neurogenesis as the primary cellular process targeted by ethanol-sensitive miRNAs. As in vitro systems alone may not be sufficient for studying adult neurogenesis, animal studies are required to further identify the role of miR-302b in ethanol-mediated alterations in adult neurogenesis.
Cellular stress in the brain modulates several events of neurogenesis by activating GSK-3B (40). In dicer knockdown mice (or in the absence of mature miRNAs), GSK-3B is inactivated because of increased phosphorylation at the serine 9 position (41). Moreover, during neuronal differentiation, ethanol inhibits neurite outgrowth by activating GSK-3B through dephosphorylation at serine 9 (42). Prevention of ethanol-induced neuronal apoptosis at substantial levels by GSK-3B inhibition supports earlier reports about the involvement of GSK-3B signaling in ethanol neurotoxicity (17). Earlier studies have also shown that inhibition of GSK-3B with LiCl 3 down-regulates the expression of CCND2 (18). Prevention of ethanol-mediated induction of miRNAs and their target genes because of inhibition of GSK-3B suggests that the modulation in miRNA levels is preceded by GSK-3B activity. In line with our findings, the recent studies of Ji et al. (43) have shown that consistent activation of the Wnt/␤-catenin signaling pathway by inhibition of GSK-3B induces transcriptional activation of the miR-181 family (43). Inhibition of GSK-3B can result in activation of several transcription factors that are probably involved in the induction of miRNAs (21).
In summary, our studies have shown that the absence of miR-NAs in SH-SY5Y cells increases ethanol-induced apoptosis, particularly after long-term exposure to ethanol. We have identified miRNAs that are regulated by both short-term and longterm exposure to ethanol (miR-34a * , miR-523, miR-497, and miR-618) and miRNAs regulated by either short-term (miR-369 -3p, miR-656, miR-606, and miR-493) or long-term (miR-302b, miR-373, miR204, miR-208b, and miR-432) exposure to ethanol. Our studies have also shown that in neuronal cells, miR-497 regulates expression of both BCL2 and CCND2, whereas miR-302b regulates expression of CCND2. Moreover, although sour experiments have demonstrated involvement of mitochondria in miR-497-induced apoptosis of neuronal cells, miR-302b-induced apoptosis is independent of mitochondria. Also, our studies with LiCl3 and TCZD-8 indicated that ethanol-induced modulation in miRNAs and their target genes are dependent on GSK-3B activity. In conclusion, our studies suggest that miR-497 and miR-302b regulate ethanol-induced neuronal apoptosis by both mitochondria-dependent and mitochondria-independent pathways, respectively. Further in vivo studies are required to explore the role of miR-302b and CCND2 in ethanol-mediated inhibition of adult neurogenesis.