Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Nitric oxide plays many different roles in the mammalian nervous system. It functions as a neurotransmitter, a regulator of blood flow, a mediator of host defense mechanisms, and as a neurotoxin (Bredt and Snyder, 1994). The specific role it plays depends on spatial, temporal, and environmental conditions. Three genes encode for a mammalian nitric-oxide synthase [L-arginine, NADPH:oxygen oxidoreductase (nitric-oxide forming); EC 1.14.13.39]. Two genes encode constitutively expressed synthases, whereas one encodes an inducible synthase. The inducible nitric-oxide synthase (iNOS) is primarily responsible for production of nitric oxide used during host defense. However, it also produces nitric oxide in amounts that are injurious to host cells, leading to neurotoxicity and disease (Yun et al., 1997). Thus, there is a great need to understand drug interactions with regulatory mechanisms governing iNOS expression in the brain.

iNOS regulation occurs primarily through transcriptional controls (Xie and Nathan, 1994). The 5'-untranslated region of the murine Nos2 gene consists of a proximal core promoter module and a distal classical enhancer element (Alley et al., 1995). Analogous regions are found in the rat iNOS gene promoter (Eberhardt et al., 1996), suggesting conserved regulatory mechanisms. The distal enhancer region mediates the potentiating actions of interferon- through the binding of IRF-1 (Martin et al., 1994), whereas the core promoter appears to require NF-B binding for minimal activity (Xie and Nathan, 1994), perhaps in conjunction with a nearby putative octamer site (Alley et al., 1995). A number of chemical stimuli that activate or inhibit NF-B activity can modulate iNOS induction. However, within the iNOS gene promoter there remain additional trans-activation sites, the functions of which are not yet understood. Some of these additional sites may be involved in the cell-specific regulation of iNOS expression observed for rat brain astrocytes and glial cell lines (Galea and Feinstein, 1999).

The biochemical mechanisms involved in glial iNOS gene induction have been widely studied using primary cell cultures and the astrocyte-derived C6 cell line. Rat C6 cells possess extensive chemical and functional analogy to normal rat astrocytes and recapitulate numerous features of iNOS expression by rat brain astrocytes. These include gene induction by proinflammatory cytokines (Murphy et al., 1993), sensitivity of expression to tyrosine kinase inhibition (Feinstein et al., 1994a), cytokine regulation of L-arginine-dependent cyclic GMP production (Simmons and Murphy, 1993), potentiation of iNOS expression by lithium chloride (Feinstein, 1998a), and inhibition of iNOS expression by elevated intracellular cAMP (Galea and Feinstein, 1999).

C6 cells also share remarkable similarity with astrocytes in their sensitivity to the commonly abused drug ethanol. For instance, ethanol dose dependently inhibits proliferation of C6 glial cells (Isenberg et al., 1992) at the same concentrations that inhibit astrocyte proliferation (Davies, 1992). Other shared ethanol effects include inhibition of glucose uptake (Singh et al., 1999), enhanced free radical production (Gonthier et al., 1997), and reduction of glutamine synthase activity (Davies and Vernadakis, 1986). Our previous studies demonstrated that iNOS induction by activated rat astrocytes and C6 glial cells is dose dependently inhibited to similar degrees by acute and chronic ethanol exposure (Syapin, 1995, 1996). The acute effect on C6 cells is specific and shared by other short-chain alkanols (Syapin et al., 1999). The present study investigated the responsiveness of rat iNOS gene induction by lipopolysaccharide (LPS) and proinflammatory cytokines in C6 cells exposed to acute ethanol and determined the molecular mechanism underlying the ethanol suppression of iNOS expression. The data indicate a selective effect of ethanol on the potency of interleukin-1 and tumor necrosis factor- to stimulate iNOS expression, with no apparent effect on the enhancing activity of interferon-. Consistent with an earlier report on iNOS induction by lipopolysaccharide combined with a phorbol ester (Militante et al., 1997), cytokine-induced iNOS protein and mRNA levels are also reduced by ethanol. Reporter gene studies confirm that the ethanol inhibition is transcriptional. Furthermore, a promoter deletion mutation verifies that the enhancer region and several additional trans-activation sites of the iNOS gene are not required for the ability of ethanol to suppress rat iNOS gene promoter activity. The current in vitro studies expand prior work that demonstrated reductions in lipopolysaccharide-induced nitrite production and ex vivo iNOS expression following in vivo ethanol exposure (Greenberg et al., 1994; Kimura et al., 1996; Syapin and Huron, 1998). These results along with recent findings (Ren et al., 2000) suggest that transcription of inducible genes can be reduced in the brain during ethanol intoxication and may be an important mechanism related to alcoholic brain damage.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture. Stock cultures of C6 glioma cells (Benda et al., 1968) were propagated and maintained using supplies and methods previously described (Syapin, 1995; Militante et al., 1997). In brief, cultures were grown in high glucose containing Dulbecco's modified Eagle's medium with 5% (v/v) fetal bovine serum added. Experimental cultures were plated and grown in medium containing 2.5% (v/v) fetal bovine serum. All cultures were maintained at 37°C inside humidified incubators with 5% CO₂:95% air. Culture medium was replenished 2 to 3 days after seeding and every other day thereafter, unless otherwise noted.

iNOS Induction Protocols. C6 cells were activated using conditions previously shown conducive to induction of iNOS activity (Syapin, 1995; Militante et al., 1997; Syapin et al., 1999). Activation occurred in serum-free culture medium by treatment with combinations of lipopolysaccharide (Escherichia coli serotype 055:B5; Sigma, St. Louis, MO), rat recombinant interferon- (Invitrogen, Gaithersburg, MD), human recombinant tumor necrosis factor- (Collaborative Biomedical Products, Bedford, MA), and human recombinant interleukin-1 (Sigma). The length of exposure was typically 24 h. The concentrations used were determined as explained in the following paragraph and are described under Results. Experimental controls included water- and ethanol-exposed cultures that were not induced to express iNOS. Cells were treated with ethanol and other alkanols as previously described (Syapin, 1995; Militante et al., 1997 Syapin et al., 1999).

iNOS Activity Measurements. The iNOS activity of intact cells was assayed by measuring nitrite accumulation in the culture medium as previously described (Syapin, 1995; Militante et al., 1997), except for experiments with transfected cells. To measure nitrite production by transfected cells, duplicate aliquots of culture medium (0.1 ml) were mixed with an equal volume of a 1:1 (v/v) mixture of 1% sulfanilamide in 5% phosphoric acid and 0.1% N-(1-naphthyl)ethylenediamine (Green et al., 1982). Absorbance was read at 546 nm using a PerkinElmer HTS 7000 plate reader (PerkinElmer Instruments, Norwalk, CT). Sodium nitrite diluted into serum-free culture medium was used to prepare quantitative standards. The total amount of cell protein per culture dish or well was determined as previously described (Militante et al., 1997) for normalization of data as nanomoles of nitrite per milligram of protein. In the case of transfected cells, normalization was to protein content of cell lysates and not to total cell protein (see Preparation of Cell Lysates). The determination of total nitrite plus nitrate in the culture medium was performed using the Pseudomonas oleovorans method (Granger et al., 1996).

RT-PCR Analysis. These procedures, including isolation of whole cell RNA, were conducted according to standard laboratory protocols already described (Militante et al., 1997; Ren et al., 2000) with previously described primers for rat iNOS cDNA (Militante et al., 1997) and S12 ribosomal protein cDNA (Ren et al., 2000). PCR was carried out on a RapidCycler PCR machine (Idaho Technology, Idaho Falls, ID). Initial denaturation was at 93°C for 2 min, which was followed by 40 cycles at 93°C for 30 s, 53°C for 45 s, 72°C for 45 s, and a final elongation period of 10 min at 72°C. Quantification was achieved by band densitometry as previously described (Ren et al., 2000) using an Alpha Imager (Alpha Innotech Corp., San Leandro, CA) with normalization to S12 ribosomal protein cDNA band density. S12 mRNA served as an internal standard to verify whether equal quantities of cDNA were amplified and allowed for an estimation of the integrity of the extracted RNA. This abundant RNA does not appear to change upon addition of lipopolysaccharide, cytokines, or ethanol to cultured C6 cells.

Western Blot Analysis. Procedures established in our laboratory were used as previously described (Militante et al., 1997; Syapin et al., 1999). Anti-iNOS mouse monoclonal antibody was obtained from Transduction Laboratories (Lexington, KY) and horseradish peroxidase-conjugated anti-mouse polyclonal antibody was obtained from Amersham Pharmacia Biotech (Arlington Heights, IL). The chemiluminescence reagents were obtained from Pierce Chemical (Rockford, IL).

Plasmids. The rat iNOS promoter/enhancer reporter plasmids piNOS-Luc-3/2 and piNOS-Luc-3/1, also called pGL-3/2 (Beck et al., 1998) and pGL-3/1, were kindly provided by Josef Pfeilschifter and Wolfgang Eberhardt. Plasmid piNOS-Luc-3/2 contains a 1846-bp fragment of the rat iNOS gene promoter region (Eberhardt et al., 1996), whereas piNOS-Luc-3/1 contains a 5'-deleted 526-bp fragment identical to that contained in the piNOS-CAT-II plasmid (Eberhardt et al., 1998). Both plasmids were constructed in the pGL3-basic vector and drive expression of a modified firefly luciferase gene (Promega, Madison, WI).

Stable Transfections. The coprecipitation of calcium phosphate and DNA method was used to prepare stable transfectants of C6 cells. We followed the standard protocol for adherent cells with some modifications for C6 cells (Feinstein et al., 1996). Cells in 6-well plates at 30 to 50% confluence were used. Fresh 2.5% fetal bovine serum medium was added 30 min prior to addition of the cDNA/calcium phosphate coprecipitate. The cDNA was prepared at a 1:4 (w/w) ratio of pBK-CMV selection plasmid (Stratagene, La Jolla, CA) to expression plasmid in HEPES-buffered saline and 125 mM CaCl₂. Following precipitation, 1 µg of expression plasmid along with 0.25 µg of pBK-CMV was added to each well and the cells placed at 37°C in a 5% CO₂ incubator. After 4 h, the cells were shocked with 15% cell culture-tested grade dimethyl sulfoxide (Sigma) in HEPES-buffered saline for exactly 3 min at room temperature, followed by extensive rinsing to remove the dimethyl sulfoxide and the addition of fresh warmed medium (3 ml/well). The plates were then returned to the incubator. Geneticin (1.2 mg/ml or 0.6 mg/ml) was added on the 3rd day post-transfection to begin selection for cells that stably incorporated the pBK-CMV plasmid into chromosomal DNA. Cultures were left untouched for 3 weeks except for weekly renewal of fresh medium and geneticin. After at least 3 weeks, well developed subclones were isolated with cloning rings and expanded through two passages to provide frozen stocks and cultures for screening. Screening was done in two stages. The first stage looked for nitrite production and luciferase expression at 24 h following stimulation with 5 µg/ml lipopolysaccharide and 50 units/ml interferon-. The second-stage screening verified that the cells and expression plasmid performed as expected (cf., Fig. 6, A and B). Clones that did not react positive after three initial or two secondary screens, each at least 7 days apart, were discarded. Stable transfectants were maintained by subculturing in complete medium containing the concentration of geneticin used for initial selection. Stably transfected clones were maintained in the presence of geneticin at all times except during exposure to iNOS-inducing agents.

Preparation of Cell Lysates. Transfected cells grown in 12-well plates were extracted with 250 µl per well of 1X passive lysis buffer (Promega) according to the manufacturer's instructions, after 24 h of exposure to either serum-free medium (unstimulated activity) or medium containing 5 µg/ml lipopolysaccharide plus 50 units/ml interferon- (total activity). Luciferase activity in the lysate was assayed on duplicate 20-µl aliquots by addition of 50 µl of luciferase assay substrate (Promega) according to manufacturer's instructions using a Monolight 2010 luminometer (Analytical Luminescence Laboratory, San Diego, CA). Luciferase activity was linear over 4 logs of activity under these conditions (data not shown). Protein content of the lysate was determined (Smith et al., 1985) on duplicate 10-µl aliquots for normalization of luciferase activity as relative light units per microgram of protein and nitrite production as nanomoles of nitrite per milligram of protein. Stimulated luciferase activity represents total activity minus unstimulated activity, determined after protein normalization.

Data Analysis. Data transformations and plotting, linear and nonlinear regression, and statistical analyses were performed using GraphPad Prism version 3.02 for Windows (GraphPad Software, San Diego, CA). The nonlinear regression variables were adjusted as previously described (Syapin et al., 1999). A probability (P) of <0.05 was accepted as demonstrating statistically significant differences between groups.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Characterization of Cytokine Induction of iNOS Activity in C6 Glial Cells. Although nitrite is a commonly used indicator of nitric oxide production by the iNOS, nitrite is also further oxidized to nitrate. Thus, drugs that can alter the ratio of nitrite to nitrate may give misleading results as to their effects on iNOS expression. An experiment was performed to determine whether exposure to ethanol or other alkanols known to inhibit nitrite production in C6 cells (Syapin et al., 1999) had any effect on the ratio of nitrite to nitrate measured in the culture medium (Fig. 1). The results showed a strong correlation between the amount of measured nitrite and the amount of total nitrite plus nitrate over a 15- to 30-fold range, irrespective of treatment with alcohols (slope = 1.71 ± 0.07; r² = 0.97). Therefore, the easier to measure nitrite level was routinely used to estimate iNOS activity.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Correlation between nitrite measurements and nitrite plus nitrate measurements in cultures exposed to alcohols. C6 cell cultures were stimulated with lipopolysaccharide plus PMA in the presence or absence of primary alcohols for 24 h followed by collection of culture medium for determination of just nitrite levels (abscissa) or nitrite plus nitrate levels (ordinate). Correlation line was determined by least-square means linear regression analysis. Each symbol is the mean of duplicate determinations from a separate dish of cells. Combined results of three independent experiments.

1

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Comparison of iNOS activity after stimulation with cytokines and lipopolysaccharide alone and in combination. Stimulation with a combination of lipopolysaccharide plus interferon- led to maximal iNOS activity. Concentrations used were 60 ng/ml TNF-, 100 units/ml IFN, 2 ng/ml IL-1, and 10 µg/ml LPS. CM = interferon- + tumor necrosis factor- + interleukin-1. Values are mean ± S.E.M. of duplicate determinations on three or four samples at each condition. Results are pooled data from two independent experiments. *P < 0.05 versus treatment with TNF-, IFN, or IL-1. **P < 0.001 versus treatment with TNF-, IFN, IL-1, LPS, or IL-1 + IFN. P < 0.05 versus CM and P < 0.001 versus all other treatments.

Effect of Ethanol on Cytokine-Induced iNOS Activity. Previously, a concentration-dependent ethanol inhibition of iNOS activity induced by stimulation with lipopolysaccharide plus PMA was observed (Syapin, 1995). In the current study, those results were extended to determine the effect of ethanol exposure on iNOS activity stimulation with interleukin-1 plus interferon-, lipopolysaccharide plus interferon-, lipopolysaccharide alone, and the cytokine mixture.

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Ethanol inhibition of iNOS activity with induction by different treatments. Simultaneous exposure to ethanol inhibits iNOS activity induced by various cytokine and lipopolysaccharide mixtures without appreciably altered total cellular protein content. Values are mean ± S.E.M. for percentage of respective control. A, interleukin-1 (2 ng/ml) + interferon- (50 units/ml). Control values = 24.6 ± 16.6 nmol · mg of protein1 and 1.14 ± 0.25 mg of protein per dish; B, lipopolysaccharide (10 µg/ml) + interferon- (50 units/ml). Control values = 62.9 ± 14.8 nmol · mg of protein1 and 2.98 ± 0.44 mg of protein per dish; C, lipopolysaccharide (10 µg/ml). Control values = 33.4 ± 10.2 nmol · mg of protein1 and 3.36 ± 0.44 mg of protein per dish; D, cytokine mixture (60 ng/ml tumor necrosis factor- + 100 units/ml interferon- + 2 ng/ml interleukin-1). Control values = 50.9 ± 11.0 nmol · mg of protein1 and 1.32 ± 0.41 mg of protein per dish. Results are pooled data from four independent experiments at each condition. *P < 0.05 versus control (0 mM ethanol), **P < 0.001 versus control (0 mM ethanol).

Determination of Cytokine Potencies for iNOS Induction in the Presence of Ethanol. The inhibition of iNOS activity by ethanol could result from any of several possible mechanisms. Previously, acute ethanol exposure was found to significantly reduce the potency of lipopolysaccharide to stimulate iNOS activity, in the presence of a maximally enhancing concentration of PMA, without affecting the potency of the phorbol ester to enhance the lipopolysaccharide (Syapin, 1995). Whether ethanol worked through a similar mechanism to inhibit cytokine-induced iNOS activity was examined in a series of experiments.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Cytokine response curves in the presence and absence of ethanol. Simultaneous exposure to 200 mM ethanol inhibits maximum obtainable iNOS activity for all three cytokines and decreases the potency of interleukin-1 and tumor necrosis factor-. Top, interleukin-1 response determined with coadministration of 50 units/ml interferon-; middle, tumor necrosis factor- response determined with coadministration of both 2 ng/ml interleukin-1 and 50 units/ml interferon-; bottom, interferon- response determined with coadministration of 2 ng/ml interleukin-1. Values are mean ± S.E.M. of duplicate determinations on two samples per cytokine and ethanol concentration. Results are from a paired set of sister cultures and representative of at least two additional experiments.

Effect of Ethanol on Cytokine-Stimulated Rat iNOS Gene Expression. If ethanol reduces the promoting and enhancing activities of interleukin-1, tumor necrosis factor-, and interferon- at the iNOS gene, a reduction in both iNOS mRNA and iNOS protein expression would be expected. Experiments were performed using stimulation of iNOS activity with the standardized triple cytokine mixture and the lipopolysaccharide + PMA mixture used previously to test this prediction (Fig. 5). Ten hours after simultaneous exposure to the cytokine mixture or 500 ng/ml lipopolysaccharide plus 400 ng/ml PMA, iNOS mRNA levels, normalized to the amount of message for the abundant ribosomal protein S12, were significantly reduced compared with levels in cells not exposed to ethanol (Fig. 5, A and B). These results are consistent with our previous quantitative PCR analysis using a competitive standard (Militante et al., 1997). Twenty-four hours after stimulation, the amount of iNOS immunoreactive protein was likewise reduced by ethanol treatment in concert with reduced iNOS activity (Fig. 5, C and D). Although highly suggestive of transcriptional inhibition by ethanol, reductions in both mRNA and protein can occur through nontranscriptional mechanisms. Therefore, a luciferase reporter gene system controlled by rat iNOS gene promoter/enhancer sequences was used to examine ethanol effects on iNOS gene promoter and enhancer activity.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Ethanol effect on cytokine-stimulated iNOS mRNA and protein content. Ethanol treatment diminishes both iNOS mRNA and protein content in cytokine-stimulated C6 glial cells. A, agarose gel separation of iNOS mRNA and S12 mRNA cDNA products generated by RT-PCR. M, size markers; lanes 1 and 5, unstimulated cells; lanes 3 and 7, 150 mM ethanol-exposed unstimulated cells; lanes 2 and 6, cells stimulated with LPS + PMA or the CM, respectively; lanes 4 and 8, 150 mM ethanol-exposed cells stimulated with LPS + PMA or CM, respectively. B, normalization of iNOS PCR products to S12 PCR products as integrated density values (IDV) ratios (means with S.E.M., n = 3 per group). Ethanol treatment significantly reduced normalized iNOS mRNA expression in both lipopolysaccharide + PMA-stimulated (LPS + PMA group) and CM-stimulated (CM group) cells. C, immunoblot analysis of equal amounts of cytosol protein for iNOS immunoreactivity in four separate groups of CM-stimulated control (Cx) and 200 mM ethanol-exposed (Ex) cells; +, iNOS protein positive control. D, quantification of immunoreactive iNOS protein immunoblots with comparison to iNOS activity determined in the same experiment. Values are mean ± S.E.M. for percentage of respective control. Results are from four independent experiments. *P < 0.05, t test versus respective control condition.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   Ethanol effect on cytokine-induced iNOS gene promoter activity. Representative data on subclones of stable transfected C6 L3/1T cells, expressing vector piNOS-Luc-3/1, and L3/2T cells, expressing vector piNOS-Luc-3/2 demonstrate interferon--dependent enhancement of lipopolysaccharide-induced endogenous iNOS activity in both cell lines (A), whereas L3/1T cells with a deletion mutation from position 527 to 1846 do not show interferon--dependent enhancement of lipopolysaccharide-induced iNOS promoter-driven luciferase activity (B). A and B results are characteristic of data obtained on at least three additional L3/1T and L3/2T subclones. Concentration-response experiments demonstrate that ethanol exposure inhibits luciferase activity stimulated by 5 µg/ml lipopolysaccharide + 50 units/ml interferon- in C6 subclones expressing either the full-length promoter (L3/2T cells) (C), or the 5'-deletion mutated promoter (L3/1T cells) (D). Values are means of duplicate determinations on triplicate samples for each clone; C and D, results are presented as percent of control-stimulated activity for cells not exposed to ethanol.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study has investigated the effect of acute 24-h ethanol exposure on the ability of proinflammatory cytokines to induce iNOS expression in C6 glial cells and performed experiments to understand the mechanism responsible for the observed concentration-dependent suppression by ethanol. To make comparisons more meaningful, the maximal iNOS response of C6 cells to individual or mixtures of cytokines and lipopolysaccharide was first determined under uniform experimental conditions. This was done because previous studies on C6 cells and iNOS activity have used several different assay conditions and cytokine and lipopolysaccharide concentrations, but the rationale for choosing a condition, and whether it measured a maximal response, was not always apparent. Thus, our study provides new information on the responsiveness of C6 cells to combinations of cytokines and lipopolysaccharide not before examined. In particular, having observed a concentration-dependent increase in iNOS activity with stimulation by lipopolysaccharide alone at concentrations 1.5 µg/ml (Militante, 1996), we demonstrated that maximal iNOS activity in C6 glial cells tested under identical growth and assay conditions occurred by the synergistic action of 50 units/ml interferon- on 10 µg/ml lipopolysaccharide.

Using optimal conditions for iNOS induction by lipopolysaccharide and/or cytokines, acute 24-h ethanol exposure was found to cause a concentration-dependent inhibition of iNOS activity, similar to our previous studies using lipopolysaccharide coadministered with a phorbol ester (Syapin, 1995; Syapin et al., 1999). Ethanol exposure was equally effective against most induction stimuli, being somewhat more effective against 2 ng/ml interleukin-1 coadministered with 100 units/ml interferon-, suggesting that ethanol inhibition is not dependent on the inducing agent. This conclusion is corroborated by other data (Greenberg et al., 1997; Colton et al., 1998) and would be expected if ethanol acted downstream of receptor binding to diminish signal transduction to the iNOS gene and inhibit gene expression. A recent study suggests pretreatment with ethanol can elicit post-transcriptional effects on cytokine-induced astrocyte iNOS expression (Wang and Sun, 2001). Although there is nothing to say this is not possible, the only evidence is a lack of ethanol effect on -actin normalized iNOS mRNA at 16 h after induction. As is evident from our current normalized data compared with our previous quantitative PCR data (Militante et al., 1997), common normalization procedures only approximate the true mRNA level. Thus, conclusions based on normalized mRNA values should be viewed with caution without corroborative data such as gene promoter studies.

To our knowledge, the present study is the first to examine the type of inhibition by ethanol observed with regard to different cytokine-inducing agents. This was done to gain insight into cytokine signal transduction pathways that might be targeted by the ethanol. The data are consistent with an intracellular signal dampening of the interleukin-1 and tumor necrosis factor- pathways by ethanol, leading to a reduced maximal response and decreased potency for induction of iNOS activity. A similar pattern was observed previously whereby ethanol exposure significantly reduced the maximal response and decreased the potency of lipopolysaccharide for induction of iNOS activity (Syapin, 1995). One feature shared by the lipopolysaccharide, interleukin-1, and tumor necrosis factor- signal transduction pathways is activation of NF-B. Preliminary data indicate that acute ethanol exposure can inhibit lipopolysaccharide plus interferon-- and interleukin-1-stimulated NF-B activation in C6 glial cells (D. K. Garrett and P. J. Syapin, unpublished results). Whether this is involved in suppression of iNOS expression by ethanol is currently being investigated.

Interestingly, the interferon- pathway was not affected by ethanol exposure in the present study, which may shed light on the lack of ethanol effect on PMA-mediated iNOS expression observed previously (Syapin, 1995). PMA was recently reported to increase IRF-1 activity (Momose et al., 2000), which is important for interferon- enhancement of murine iNOS expression (Martin et al., 1994). If this is the mechanism behind PMA enhancement of lipopolysaccharide in C6 glial cells, than a lack of effect on IRF-1 activity by ethanol could explain the lack of effect on both PMA and interferon- enhancement of rat iNOS activity. Consistent with our results, ethanol exposure selectively lacked an effect on interferon--induced intercellular adhesion molecule-1 expression by astrocytoma cells, but it inhibited intercellular adhesion molecule-1 expression induced by tumor necrosis factor- in the same cells (DeVito et al., 2000). It appears, therefore, that ethanol exposure can have selective effects on proinflammatory cytokine signal transduction pathways in glial cell types.

The selective effect of ethanol exposure on proinflammatory cytokine signaling of iNOS gene expression may be further explained based on study of the 5'-untranslated regulatory region of the iNOS gene. In the mouse gene, the enhancer element resides at position 913 to 951 and contains the cis-acting sequence that mediated the synergistic effect of interferon- on lipopolysaccharide-stimulated iNOS expression (Martin et al., 1994). The present results demonstrate for the first time the need for the region upstream of position 526 for interferon- enhancement of lipopolysaccharide-induced iNOS gene expression in rat cells, even though several -IRE sites are present downstream of position 526. Furthermore, our results indicate that cytokine signals that activate the promoter region upstream of 526, such as interferon-, are not involved in the ethanol suppression of rat glial iNOS expression. Whether any -IRE sites downstream of position 526 are involved in the ethanol suppression is not known at this time.

We also did not observe any requirement for sequences upstream of position 526 to elicit maximal reporter gene induction by lipopolysaccharide. This observation differs from data reported on rat vascular smooth muscle cells (Zhang et al., 1998). In addition to the obvious difference in cell types, our experiments used stable transfections, whereas the latter experiments (Zhang et al., 1998) used transient transfections. Other studies using transient transfection of rat iNOS gene promoter sequences into rat cells have found significant differences between cell types as well. Vascular smooth muscle cells required at least 3.2 kilobases of upstream sequence for maximal stimulation by lipopolysaccharide or a triple cytokine mix (Zhang et al., 1998, 2000), whereas insulin-producing cells required at least 1002 upstream bases for induction by interleukin-1 and various cytokine mixtures (Darville and Eizirik, 1998). In contrast, studies show that the 526 promoter fragment used in our experiments contains all necessary information for full induction by interleukin-1 or dibutyryl cAMP in rat glomerular mesangial cells (Eberhardt et al., 1998). These authors went on to demonstrate that for mesangial cells, the B site at position 96 to 106 mediates interleukin-1-stimulation of iNOS gene promoter activity, whereas CCAAT/enhancer-binding protein sites from position 111 to 526 are necessary for dibutyryl cAMP-stimulated activity. These results highlight the complex regulation of iNOS expression within different cell types of the same species. Using 1588 and 85 base fragments of the mouse iNOS promoter stably transfected into rat C6 cells, Feinstein observed a decrease in cytokine-induced transcriptional activity after elevation of intracellular cAMP by norepinephrine pretreatment only in the cells with the 1588 fragment (Feinstein, 1998b). It will be interesting to determine whether the same holds true for the rat promoter and if the CCAAT/enhancer-binding protein sites identified by Eberhardt et al. (1998) also mediate the inhibiting effect of cAMP on rat glial iNOS expression.

Although the rat iNOS promoter contains two B sites, it is the site at position 96 to 106 that most likely mediates interleukin-1, tumor necrosis factor-, and lipopolysaccharide-stimulated iNOS induction in C6 glial cells. This conclusion is drawn from the present data that lipopolysaccharide alone was able to activate the luciferase reporter gene in C6 cells transfected with the piNOS-Luc-3/1 plasmid that lacks the B site at position 955 to 965. Whether ethanol exposure affects NF-B binding to the B site at position 96 to 106 remains to be shown. We can conclude, however, that the B site present at position 955 to 965, the cyclic AMP response element site at 1168 to 1174, the two activator protein-1 sites at 668 to 675 and 1120 to 1126, and the numerous -IRE, interferon-stimulated response element and -activated sites located upstream of position 526 are not involved in ethanol inhibition of rat C6 cell iNOS gene transcription. This conclusion is consistent with our pharmacological studies that found no effect of ethanol exposure on the potency of interferon- to stimulated iNOS activity and suggests that ethanol exposure does not influence interferon- signaling and expression of IRF-1. In fact, removal of sequences upstream of position 526 appeared to enhance the ability of ethanol to inhibit iNOS gene expression in the presence of lipopolysaccharide and interferon-, judging by the lower IC₅₀ values for ethanol inhibition of luciferase activity in L3/1T cells.

Ethanol suppression of inducible genes involved in host defense and responses to injuries such as stroke, infection, and brain trauma may have dire consequences during acute and chronic alcohol abuse. It is well known that alcohol abusers experience high blood ethanol levels and are more prone to opportunistic pathogens, as well as to the human immunodeficiency virus. Unfortunately, there is a paucity of information regarding the effects of alcohol on central nervous system immunity and the consequences to alcohol abusers. The present study showed that in vitro ethanol exposure inhibits cytokine- and lipopolysaccharide-induction of rat glial iNOS gene transcription. Given the importance of iNOS induction as a universal response of innate immune systems, whether ethanol suppression of glial iNOS expression in vivo impedes innate immune responses in the brain requires consideration.