Abstract
Loss of gastrointestinal (GI) barrier integrity has been implicated in a wide range of inflammatory illnesses, including alcoholic cirrhosis. Using monolayers of Caco-2 (intestinal) cells as a model, we showed that the ability of ethanol (EtOH) to disrupt intestinal barrier integrity depends on damage to the microtubule (MT) cytoskeleton, especially oxidative injury. One drug that prevented both the MT damage and barrier disruption wasl-N6-1-iminoethyl-lysine, a selective inhibitor of the inducible form of nitric-oxide synthase (iNOS). Because of this finding and because overproduction of nitric oxide (NO) and generation of peroxynitrite (ONOO−) have been proposed to be responsible for mucosal injury in other GI disorders, we sought to determine whether NO overproduction and ONOO− formation mediates EtOH-induced MT damage and loss of intestinal barrier function. To this end, Caco-2 monolayers were exposed to EtOH or to authentic ONOO− or ONOO− generators with or without pretreatment with iNOS inhibitors or antioxidants. We found that EtOH caused 1) iNOS activation, 2) NO overproduction, 3) increases in oxidative stress and superoxide anion production (superoxide dismutase quenchable fluorescence of dichlorofluorescein), 4) nitration and oxidation of tubulin (immunoblotting), 5) decreased levels of stable polymerized tubulin, and 6) increased levels of disassembled tubulin. EtOH also 7) extensively damaged the MT cytoskeleton and 8) disrupted barrier function. Authentic ONOO− or ONOO− donors had similar effects. Pretreatment with a selective iNOS inhibitor,l-N6-1-iminoethyl-lysine, or with antioxidants (ONOO− scavengers urate orl-cysteine; superoxide anion scavenger superoxide dismutase) attenuated damage due to EtOH or to ONOO−generators. We conclude that EtOH-induced MT damage and intestinal barrier dysfunction require iNOS activation followed by NO overproduction and ONOO− formation. These findings provide a rationale for the development of novel therapeutic agents for alcohol-induced GI disorders that inhibit this mechanism.
The gastrointestinal (GI) epithelium is a highly selective barrier that normally prevents the passage of harmful molecules across the mucosa and into the circulation (Bode et al., 1987; Hollander, 1992). An abnormal GI barrier, in contrast, can allow the penetration of normally excluded luminal substances (e.g., endotoxin) across the mucosa and can lead to the initiation and/or perpetuation of inflammatory processes and mucosal damage. This damage, and the ensuing loss of GI barrier integrity, has been implicated in a wide range of inflammatory illnesses, including alcoholic cirrhosis (Bode et al., 1987; Hollander, 1992; Keshavarzian et al., 1994). The underlying difficulty in managing these inflammatory disorders is due in large part to our limited understanding of their pathophysiology.
For example, alcohol [ethanol (EtOH)] intake injures the functional and structural integrity of the intestinal mucosa (Bjorkman and Jessop, 1994) and causes loss of intestinal barrier function (Talbot et al., 1984; Keshavarzian et al., 1994, 1999). This is thought to be important in the development of alcoholic cirrhosis (Bode et al., 1987;Keshavarzian et al., 1999). Little is known, however, as to the underlying mechanisms. While investigating this mechanism, we showed (Banan et al., 1998b, 1999a,b,c, 2000), using monolayers of intestinal cells as a model, that EtOH-induced disruption of barrier integrity requires damage to and disruption of the microtubule cytoskeleton. Oxidative damage appeared to be key because microtubules became oxidized and because several agents, including antioxidants, prevented these effects of EtOH. One of these drugs wasl-N6-1-iminoethyl-lysine(l-NIL), a selective inhibitor of inducible nitric-oxide synthase (iNOS). iNOS was further implicated by the observations that EtOH increased iNOS activity and thatl-NIL prevented not only the iNOS up-regulation but also the EtOH-induced microtubule damage and barrier disruption.
iNOS up-regulation is predicted to lead to NO overproduction (Chen et al., 1996; Salzman et al., 1996; Unno et al., 1997), and many of the toxic effects of NO overproduction are mediated by the peroxynitrite (ONOO−), a product of the reaction of NO with superoxide anions (Radi et al., 1991; Ischiropoulos et al., 1992, 1995;Rachmilewitz et al., 1995; Haddad et al., 1994; Muijsers et al., 1997). Indeed, overproduction and uncontrolled generation of ONOO− have been proposed in several recent studies to be an important factor in tissue damage during inflammation (Radi et al., 1991; Rachmilewitz et al., 1993, 1995; Ischiropoulos et al., 1995). Nevertheless, the precise pathogenic mechanism by which oxidative stress leads to mucosal abnormalities in the intestine, especially after EtOH insult, is not known. Accordingly, in the present study, we investigated the possibility that NO overproduction and ONOO− generation cause the oxidative damage to microtubules that mediates EtOH-induced intestinal barrier dysfunction.
Materials and Methods
Cell Culture.
Caco-2 cells (from a human colonic cell line) were obtained from American Type Culture Collection (Rockville, MD) at passage 15. Although of colonic origin, these widely studied cells resemble small intestinal cells in that they have defined apical brush borders, form highly tight junctions, and exhibit a highly organized microtubule network on differentiation (Gilbert et al., 1991). These cells also express markers of mature enterocytes such as small intestinal hydrolases (sucrase-isomaltase and alkaline phosphatase) and nutrient transporters. In addition, these cells resemble small bowel epithelium in having receptors for prostaglandins; growth factors [e.g., epidermal growth factor (EGF), insulin-like growth factor-I] and insulin; receptors for vasoactive intestinal peptide and low-density lipoprotein, and transporters such as dipeptides, fructose, glucose, other hexoses, and vitamin B12 (Gilbert et al., 1991). Accordingly, Caco-2 cell model provides a suitable in vitro model for our barrier function studies. Cells were maintained at 37°C in Dulbecco's modified Eagle's medium in an atmosphere of 5% CO2 and 100% relative humidity. Cells were split at a ratio of 1:6 on reaching confluency every 6 days and set up in either 6-, 24-, or 48-well plates for experiments or in T-175 flasks for the maintenance of stocks. Cells grown for barrier integrity work were split at a ratio of 1:2 and seeded at a density of 200,000 cells/cm2 into 0.4-μm Biocoat Collagen I Cell Culture Inserts (0.3-cm2 growth surface; Becton Dickinson Labware, Bedford, MA), and experiments were performed at least 7 days postconfluence. The utility, maintenance, and characterization of this cell line have been previously published (Gilbert et al., 1991).
Experimental Design.
In the first series of experiments, we evaluated the effect of 30-min exposure of cells to injurious (2.5 and 15%) and noninjurious (1%) concentrations of EtOH (v/v) or vehicle (isotonic saline/Dulbecco's modified Eagle's medium) on cell monolayer barrier integrity, iNOS activity, NO production, ONOO− generation (i.e., nitration and oxidation), oxidative stress, and microtubule disassembly and instability as described later. The concentrations of EtOH used in the present investigation are clinically relevant (Dinda et al., 1996;Bjorkman and Jessop, 1994) and have previously been shown by us to cause loss of monolayer barrier function without cell death in this cell line (Banan et al., 1998a,b).
In a second series of experiments, we assessed the ability of authentic ONOO− (Alexis Corp., San Diego, CA) or ONOO− donors (Sigma Chemical Co., St. Louis, MO) to mimic the effects of EtOH on barrier function and on the microtubule cytoskeleton. Monolayers were incubated with NOS substrate or NO-related agents (listed here) (Sigma Chemical Co.) and, in selected experiments, with EtOH or vehicle. Agents and incubation times included 1) an NOS substrate, l-arginine (l-Arg, 3 mM, 48 h), 2) authentic ONOO− (0.01–1.0 mM, see later) or ONOO−-generating systems: 1,3-morpholinosydnonymine (SIN-1; 0.1–5 mM), or a combination ofS-nitroso-N-acetyl penicillamine (SNAP; 1 mM) plus xanthine (X; 1 mM) plus xanthine oxidase (XO; 100 mU/ml). These doses of agents were determined to be effective in our pilot studies and in previously published studies (Salzman et al., 1996; Kennedy et al., 1998, 1999). To promote stability of ONOO−in solution, ONOO− (180 mM stock in 0.3 M NaOH; Alexis Corp. San Diego, CA) was added to the cell culture media to a final pH 7.6. Aliquots of ONOO− and HCl were added just above the surface of the cell culture solution on the side of the culture dish immediately followed by a gentle swirl. The concentration of ONOO− in the cell culture media was monitored by removing a small aliquot and measuring the increase in absorbance at a wavelength of 302 nm (E302 nm = 1.67 mM−1 cm−1). Because ONOO− does degrade over time due to its reaction with water, we normalized ONOO− values to that of an internal ONOO− standard incubated under identical conditions and expressed the results as a percentage. In preliminary studies, we noted no adverse effects due to the use of pH of 7.6 on monolayer barrier function, on cell viability, or on the cytoskeleton.
In a third series of experiments, we investigated the effects of pretreatment with either an isoform-selective NOS inhibitor or with various antioxidants (listed here) that prevent ONOO− formation or scavenge ONOO−. Outcome measures were, again, cell oxidative state, barrier function, and microtubule assembly and stability. We anticipated that these agents would protect monolayers exposed to EtOH or ONOO−. All pretreatment drugs were left in the incubation media during the subsequent steps: 1) NO scavengers urate (0.5 mM) or l-cysteine (3 mM) (or, as control, d-cysteine) preincubated for 30 min before EtOH or ONOO− compounds; 2) a superoxide scavenger, superoxide dismutase (SOD, 300 U/ml) [or heat-inactivated SOD (iSOD)] (Ferro et al., 1997) preincubated for 30 min before EtOH or ONOO− compounds; or 3) the selective iNOS inhibitorl-N6-1-iminoethyl-lysine (l-NIL, 1 mM) or the nonselectiveNG-nitro-l-arginine (l-NNA, 1 mM) orNG-monomethyl-l-arginine (l-NMMA, 1 mM) (Salzman et al., 1996; Banan et al., 1999a) 1 h before EtOH. To determine whether the effects of NOS inhibitors was specific to inhibition of NOS,l-arginine (l-Arg; 3 mM) was added to selected assays to reverse the inhibition by NOS inhibitors. Inhibitors or scavengers were purchased from Sigma Chemical Co.
In a final series of experiments, we investigated the role of tubulin nitrotyrosination or carbonylation and of tubulin disassembly in microtubule cytoskeletal instability and epithelial barrier dysfunction. Monomeric and polymerized fractions of tubulin (structural protein of the microtubules) were isolated and then analyzed using immunoblotting. In all experiments, microtubule integrity was assessed by 1) immunofluorescent labeling/fluorescence microscopy to determine the percentage of cells with normal microtubules, 2) detailed analysis by high-resolution laser scanning confocal microscopy (LSCM), and 3) quantitative immunoblot analysis of monomeric and polymerized tubulin and oxidation and nitration of these tubulin fractions as described below.
Determination of Cell Integrity.
Live/Dead kits (Molecular Probes, Eugene, OR) were used. This assay measures parameters of cell death: nuclear membrane integrity and chromatin condensation by ethidium homodimer-1 probe, as we described previously (Banan et al., 1999a).
Determination of Epithelial Barrier Function by Fluorometry.
Barrier integrity was determined by measuring apical-to-basolateral flux of a fluorescent marker [fluorescein sulfonic acid (FSA), 200 μg/ml, 478 Da); Molecular Probes] as previously described (Unno et al., 1996; Banan et al., 1999a). After pharmacological treatments, fluorescent signals from samples were quantified using a fluorescence multiplate reader. The excitation and emission spectra for FSA were excitation = 485 nm and emission = 530 nm. Clearance (CL) was calculated using the following formula: CL (nl/h/cm2) =Fab/([FSA]a × S), whereFab is the apical to basolateral flux of FSA (light units/h), [FSA]a is the concentration at baseline (light units/nl), and S is the surface area (0.3 cm2) (Unno et al., 1996). Simultaneous controls were performed with each experiment.
Assay of iNOS Activity.
Cells grown to confluence were removed by scraping, centrifuging, and homogenizing on ice in a buffer containing 50 mM Tris-HCl, 0.1 mM EDTA, 0.1 mM EGTA, 12 mM 2-mercaptoethanol, and 1 mM phenylmethylsulfonyl fluoride (pH 7.4). The homogenates were incubated with a cation-exchange resin (AG 50W-X8, Na+ form; Sigma Chemical Co.) for 5 min on ice to deplete endogenous l-Arg. Conversion ofl-[3H]Arg (Amersham Corp., Arlington Heights, IL) tol-[3H]citrulline was measured in the homogenates by scintillation counting, as previously described (Salzman et al., 1996; Banan et al., 1999a). Experiments in the absence of NADPH or in the presence of NOS inhibitor l-NMMA (1 mM) determined the extent ofl-[3H]citrulline formation independent of NOS activity. Experiments in the presence of NADPH, without Ca2+ and with 5 mM EGTA, determined Ca2+-independent NOS (iNOS) activity. In selected experiments, the isoform-selective iNOS inhibitor l-NIL was also present (see earlier). Protein concentrations were determined according to the Bradford method (Bradford, 1976).
Western Blot Analysis of Level of iNOS Protein.
After treatment with EtOH or vehicle, the cells were washed once with cold PBS, scraped in 1 ml of cold PBS, and harvested in an antiprotease cocktail (2 μg/ml aprotinin, 2 μg/ml pepstatin, 2 μg/ml leupeptin, and 2 μg/ml phenylmethylsulfonyl fluoride). Protein content was determined according to the Bradford method (Bradford, 1976). For immunoblotting, samples (25 μg of protein/lane) were added to SDS buffer (250 mM Tris-HCl, pH 6.8, 2% glycerol, 5% mercaptoethanol), boiled for 5 min, and then separated on 7.5% SDS-polyacrylamide gel electrophoresis. Subsequently, proteins were transferred to nitrocellulose membranes and then blocked in 3% BSA for 1 h followed by several washes (Tris-buffered saline). The immunoblotted proteins were incubated for 2 h in Tween 20 and Tris-buffered saline and 1% BSA with the primary antibody (mouse monoclonal anti-human iNOS, 1:3000 dilution; Santa Cruz Biotech, Santa Cruz, CA). A horseradish peroxidase (HRP)-conjugated goat anti-mouse antibody (Molecular Probes, Eugene, OR) was used as a secondary antibody at 1:3000 dilution. Membranes were visualized by enhanced chemiluminescence (ECL; Amersham) and autoradiography (Singer et al., 1996).
Chemiluminescence Analysis of NO Concentration in Cultures.
NO production was assessed by a novel and sensitive chemiluminescence procedure (Al-Mufti et al., 1998). Briefly, cells were homogenized by sonication, and the endogenous nitrate (NO3−) and nitrite (NO2−), the metabolic degradation products of NO, were then reduced to NO using vanadium(III) (Sigma Chemical Co.) and HCl at 90°C before the measurement of NO concentration by chemiluminescence analysis. Chemiluminescence was measured using a Seivers NOA 280 analyzer (Sievers, Boulder, CO). NO was expressed in micromolar and calculated by comparison with the chemiluminescence of a standard solution of NaNO2. The absolute NO values were reported as micromoles per 1 × 106 cells.
Determination of Cell Oxidative Stress and Superoxide Anion Production.
Oxidative stress and superoxide anion (O⨪2) production were assessed by measuring the conversion of a nonfluorescent compound, 2′,7′-dichlorofluorescein diacetate (DCFD; Molecular Probes) into a fluorescent dye, dichlorofluorescein (DCF) as previously described (Wakulich and Tepperman, 1997; Banan et al., 1999a). The dependence of the assay on monolayer O⨪2 generation was shown by adding an active superoxide radical scavenger, SOD (300 U/ml), or an inactive superoxide radical scavenger, iSOD. Briefly, monolayers grown in 96-well plates were preincubated with the membrane-permeable DCFD (10 μg/ml for 30 min) before the subsequent treatments (see experimental series 1 and 3). After treatments, fluorescent signals (i.e., DCF fluorescence) from samples were quantified using a fluorescence multiplate reader set at an excitation wavelength of 485 nm and an emission wavelength of 530 nm.
Immunofluorescent Staining and High-Resolution LSCM of Microtubule Cytoskeleton.
Cells from monolayers were fixed in cytoskeletal stabilization buffer and then postfixed in 95% EtOH as previously described (Allen, 1985; Banan et al., 1998a,b). Cell monolayers were subsequently processed for incubation with primary monoclonal mouse anti-β-tubulin antibody (IgG1, rat/human reactive; Sigma Chemical Co.) at 1:200 dilution for 1 h at 37°C. Slides were washed three times in Dulbecco's PBS (D-PBS) and then incubated with a secondary antibody (fluorescein isothiocyanate-conjugated goat anti-mouse; Sigma Chemical Co.) at 1:50 dilution for 1 h at room temperature, washed three times in D-PBS and once with deionized H2O, and subsequently mounted in Aquamount. All antibodies were diluted with D-PBS containing 0.1% BSA. Samples were stored in the dark at −20°C and were examined by both standard fluorescent and LSCM (Zeiss, Munich, Germany). Cell monolayers on slides were observed in a blinded fashion with LSCM using a 63× oil immersion plan-Apochromat objective, NA 1.4 (Zeiss). An argon laser (λ = 488 nm) was used to examine fluorescein isothiocyanate-labeled cells, and the cytoskeletal elements were examined for their overall morphology, orientation, and disruption as previously described (Banan et al., 1998b).
Microtubule (Tubulin) Fractionation and Quantitative Immunoblotting of Tubulin.
Polymerized (S2) and monomeric (S1) fractions of tubulin were isolated as we previously described (Banan et al., 1998a,b). Fractionated S1 and S2 samples were flash frozen in liquid N2 and then stored at −70°C until immunoblotting. For immunoblotting, samples (5 μg) were placed in SDS sample buffer (250 mM Tris-HCl, pH 6.8, 2% glycerol, 5% mercaptoethanol), boiled for 5 min, and then subjected to electrophoresis on 7.5% polyacrylamide gels. Procedures for Western blotting were performed at room temperature (Banan et al., 1998b). To quantify the relative levels of tubulin, the absorbance of the bands corresponding to immunoradiolabeled tubulin was measured with a laser densitometer. We ensured equal loading of proteins in all Western blots by always loading 5 μg of protein/lane assessed according to the Bradford method (Bradford, 1976). Experimental variations were further minimized by the loading of 5 μg of standard tubulin, which was tested concurrently with each gel. To further ensure reproducibility, each treatment group was run in duplicate and/or triplicate on different days.
Immunoblotting Determination of Tubulin Oxidation and Tubulin Nitration.
Oxidation and nitration of the tubulin backbone of microtubules were assessed by measuring protein carbonyl and nitrotyrosine formation, respectively (Banan et al., 1999c, 2000a;Ferro et al., 1997). Carbonylation and nitrotyrosination of tubulin were determined in a similar manner as the quantitative blotting of tubulin (Banan et al., 1998b) except for differences in primary antibodies and buffers. To avoid unwanted oxidation of tubulin samples, all buffers contained 0.5 mM dithiothreitol and 20 mM 4,5-dihydroxy-1,3-benzene sulfonic acid (Sigma Chemical Co.). To determine the carbonyl content, samples were blotted onto a polyvinylidene difluoride membrane, followed by successive incubations in 2 N HCl and 2,4-dinitrophenylhydrazine (DNP; 100 μg/ml in 2 N HCl; Sigma Chemical Co.) for 5 min each. Membranes were then washed three times in 2 N HCl and subsequently washed seven times in 100% methanol for 5 min each, followed by blocking for 1 h in 5% BSA in 10× PBS/Tween 20 (PBS-T). Immunologic evaluation of carbonyl formation was performed for 1 h in 1% BSA/PBS-T buffer containing anti-DNP (1:25,000 dilution; Molecular Probes). Membranes were then incubated with an HRP-conjugated secondary antibody (1:4000 dilution, 1 h; Molecular Probes, Eugene, OR). To determine nitrotyrosine content, after the blocking step earlier (i.e., BSA/PBS-T buffer), membranes were probed for nitrotyrosine by incubation with 2 μg/ml monoclonal anti-nitrotyrosine antibody for 1 h (Upstate Biotechnology, Lake Placid, NY) followed by the HRP-conjugated secondary antibody (as earlier). Wash steps and film exposure were as previously described (Banan et al., 1998b). The relative levels of oxidized or nitrated tubulin were then quantified by measuring, with a laser densitometer, with the absorbance (A) of the bands corresponding to anti-DNP or anti-nitrotyrosine immunoreactivity. Immunoreactivity was expressed as the fraction (ratio) of carbonyl or nitrotyrosine formation in the treatment group to that in the oxidized or nitrated tubulin standard run concurrently.
Statistical Analysis.
Data are presented as mean ± S.E. All experiments were carried out with a sample size of at least four to six observations per group. Statistical analysis between or among groups was carried out using ANOVA followed by Dunnett's multiple range test (Harter, 1960). Correlational analyses were performed using the Pearson test for parametric analysis or, when applicable, the Spearman test for nonparametric analysis. For all analyses, a value of P < .05 was deemed to represent statistical significance.
Results
Injurious Effects of EtOH on Monolayer Barrier Integrity and Its Prevention by a Selective iNOS Inhibitor.
Caco-2 monolayers exposed to a range of concentrations of EtOH (1, 2.5, and 15%) for 30 min showed a dose-dependent loss of epithelial barrier function as demonstrated by increased clearance of FSA (Fig.1). The lowest dose of EtOH that significantly increased FSA clearance was 2.5%. Under these condition, EtOH does not cause cell death as determined by ethidium homodimer-1 probe (Banan et al., 1999a). Pretreatment with a selective iNOS inhibitor (l-NIL) significantly attenuated the EtOH-induced disruption of barrier integrity by up to 65%. Preincubation with an NOS substrate, l-Arg (48 h), synergized with 1% EtOH to create an injurious effect and potentiated the monolayer barrier dysfunction induced by 2.5 and 15% EtOH.l-Arg by itself did not significantly affect monolayer barrier function. Pretreatment with l-NIL completely abolished the “potentiating” interaction between l-Arg and EtOH (Fig. 1). Similar to l-NIL, 1-h preincubation with nonselective NOS inhibitors (l-NMMA orl-NNA) significantly protected against barrier dysfunction by EtOH (FSA clearance = 1223 ± 53 nl/h/cm2 for l-NMMA or 1309 ± 93 for l-NNA versus 2548 ± 105 for 15% EtOH). One-hour incubation with l-NMMA or l-NNA by itself did not significantly affect FSA clearance compared with vehicle (not shown).
NO-Dependent Oxidative Mechanisms in Injurious Effects of EtOH.
Similar to l-NIL, the ONOO− scavengers urate andl-cysteine and the O⨪2 scavenger SOD significantly attenuated the loss of barrier integrity induced by EtOH (Table1; cysteine and SOD shown). The analogsd-cysteine and iSOD were ineffective.
Figure 2A shows that Caco-2 monolayers exposed to 2.5 or 15% EtOH exhibited increases in calcium-independent iNOS activity as early as 30 min after exposure. In Fig. 2B, a representative Western blot of iNOS protein shows significant increases in iNOS protein levels after exposure to injurious EtOH (compared with basal levels of iNOS protein seen in controls). Densitometry analysis of six per treatment group revealed the following absorbency values: control/vehicle (712 ± 101), noninjurious 1% EtOH (778 ± 74), and 2.5% EtOH (5211 ± 97).
Chemiluminescence analysis of cell lysates of EtOH-exposed cells (Fig.3) shows that this iNOS up-regulation is associated with the overproduction of NO, the product of the iNOS-catalyzed reaction. Increases in both iNOS and NO were prevented by preincubation with l-NIL (Figs. 2A and 3). A noninjurious dose of EtOH (1%) did not have any significant effect on iNOS activity (0.49 ± 0.11 pmol/min/mg of protein), nor did it cause NO overproduction (2.23 ± 0.31 μmol/106 cells) compared with vehicle (0.40 ± 0.03 and 1.9 ± 0.240, respectively).
To determine whether iNOS activation and NO overproduction induced by EtOH result in oxidative stress, we assessed increased fluorescence of DCF. As expected, DCF was increased by EtOH, and this was prevented byl-NIL (Fig. 4A). To determine whether O⨪2 is a key radical for this oxidative stress, we preincubated monolayers with SOD. We showed (Fig. 4B) that SOD quenched the DCF signal to control levels, whereas iSOD did not. This confirms O⨪2 generation by EtOH.
Role of NO-Dependent Mechanisms in Deleterious Effects of EtOH on Microtubule Cytoskeleton.
EtOH dose-dependently decreased microtubule stability (to 27% of control) as determined by laser confocal microscopy and as indicated by the reduced percentage of cells displaying normal microtubules (see Table2). Similar to effects on barrier function, the lowest EtOH dose that significantly induced instability of microtubules was 2.5%. Additionally, there was a significant (P < .05) positive correlation (r = 0.98) between EtOH doses and percentage of cells with abnormal microtubules. Preincubation with the selective iNOS inhibitorl-NIL or with antioxidants (urate,l-cysteine, or SOD) markedly and significantly blunted the EtOH-induced microtubule instability (Table 2).
Figure 5 shows that control cells from the monolayer exhibit a normal, stellate distribution of the microtubule network as revealed by immunofluorescent staining (Fig.5a). After exposure to EtOH (Fig. 5b), fragmentation and disruption of the microtubules are seen. Preincubation with l-NIL before EtOH prevented the injurious effects of EtOH (Fig. 5c).
We then measured “footprints” of ONOO−generation, namely nitrotyrosine and carbonyl moieties, on the tubulin backbone of the microtubules using immunoblotting. Figure6A shows that EtOH caused polymerized tubulin (S2) nitration and oxidation. Quantitative immunoblotting shows the fraction of polymerized tubulin (S2) that was nitrated (0.71 ± 0.03%) or oxidized (0.75 ± 0.01%) after exposure to 2.5% EtOH. Preexposure of monolayers to the same iNOS inhibitor or antioxidants as shown in Table 2 significantly prevented the nitration and oxidation of tubulin (Fig. 6B, nitration shown). Analogues and/or inactive forms of these antioxidant did not protect. Representative Western immunoblots of anti-DNP (Fig. 7A) and antinitrotyrosine (Fig. 7B) show the aforementioned treatment regimens. There was a significant (P < .05) positive correlation between microtubule instability and tubulin oxidation (r = 0.98) and tubulin nitration (r = 0.95).
Using Western immunoblotting, we showed (Fig.8) that EtOH (2.5% shown) elicits a significant reduction in the stable (polymerized) S2 tubulin fraction (51 ± 0.29% versus 66 ± 0.25% for vehicle) and, in concert, increases the unstable (monomeric) S1 tubulin fraction, indicating overall microtubule disassembly and disruption. Preincubation with either l-NIL or antioxidants enhanced the stable S2 tubulin, while concomitantly decreasing the monomeric S1 tubulin in monolayers exposed to damaging EtOH. Correlation analysis showed a significant (P < .05) positive correlation (r = 0.94) between EtOH-induced barrier dysfunction (FSA) and microtubule depolymerization. Not surprisingly, the two markers for microtubule status (i.e., percentage of polymerization and percentage of normal microtubules) correlate robustly with each other (r = 0.977). These data are consistent with the protective effects of these antioxidants against tubulin oxidation, microtubule destabilization, and loss of barrier integrity.
Cytoskeletal and Barrier Disruption Induced by ONOO−Compounds.
Our hypothesis predicts that authentic ONOO−- or ONOO−-generating systems will, similar to EtOH, induce cytoskeleton and monolayer injury by themselves. Indeed, the disruption of monolayer barrier function (increased FSA clearance; Fig.9A) was caused by incubation 1) with a range of concentrations of authentic added ONOO−or 2) with known ONOO− donors [e.g., SIN-1 (an NO and O anion donor) or SNAP (an NO donor) in combination with xanthine (X) plus xanthine oxidase (XO) (an O anion donor)]. These effects were prevented by the antioxidants urate,l-cysteine, or SOD but not by d-cysteine or iSOD (Fig. 9B, data for ONOO− and SIN-1 shown). At the same concentration (3 mM), d-cysteine was much less protective than l-cysteine (4 versus 94%, respectively). To confirm that this difference in the protective capabilities of cysteine isomers is related to their ability to scavenge ONOO−, we measured their scavenging ability in vitro in the presence of added ONOO− (absorbance change at 302 normalized to percentages). l-Cysteine (3 mM) removed more than 95% of exogenously added ONOO−, whereas d-cysteine (3 mM) removed less than 15% of added ONOO−. Furthermore, ONOO− compounds did not significantly affect cell viability assessed by ethidium homodimer-1 (not shown). SNAP by itself did not induce barrier dysfunction (not shown).
Consistent with the above findings, quantitative immunoblotting showed that these ONOO− compounds not only induced a dose-dependent oxidation and nitration of the tubulin-based cytoskeleton (Table 3) but also disassembled the microtubules as shown by decreased stable S2 tubulin and increased unstable S1 tubulin (Fig.10A). Antioxidants almost completely abolished both the microtubule oxidation and nitration (Table 3) and microtubule disassembly (Fig. 10A) that was induced by the aforementioned ONOO− compounds. Immunofluorescent staining demonstrated (Fig. 10B) that ONOO− causes microtubule fragmentation, kinking, and collapse (panel b) and pretreatment with l-cysteine (panel c) protected the microtubules against ONOO−-induced damage as indicated by their normal array, resembling the controls (panel a).
Discussion
Characterizing the pathophysiological mechanism for ethanol-induced barrier dysfunction, as we tried to do in this study, is clinically important because the leaky gut has been proposed to be one of the underlying mechanisms of alcohol-mediated endotoxemia in patients with alcoholic liver disease (Bode et al., 1987; Hollander, 1992; Keshavarzian et al., 1994, 1999). We hypothesized that the cytoskeletal disruption (and barrier dysfunction) that is induced by EtOH requires iNOS activation, NO overproduction, and ONOO− formation and that it is these free radical reactions that lead to the oxidative injury to the tubulin-based backbone of the microtubule cytoskeleton. We conclude that all three separate lines of our investigation confirm the hypothesis: 1) measurement of oxidative reactions and cytoskeletal injury after EtOH administration, 2) mimicking EtOH-induced damage using ONOO−-generating systems, and 3) preventing EtOH-induced cytoskeletal instability using NOS selective inhibitors and antioxidants. The following section elaborates on this conclusion.
First, we found that under conditions where EtOH oxidizes tubulin, disrupts the microtubule cytoskeleton, and diminishes barrier integrity in cell monolayers, it also creates oxidative stress, including increased levels of O⨪2 (SOD quenchable DCF fluorescence), NO, and ONOO−. These associations strongly suggest that the underlying cause of the injury to the cytoskeletal network (and loss of barrier function) is the oxidation and nitration of its structural tubulin subunits. This mechanism is further supported by a highly significant correlation between increases in NO and nitrotyrosine levels (r = 0.985); increases in nitrotyrosine levels, and either abnormal microtubules (r = 0.95) or microtubule depolymerization (r = 0.94). Also, oxidation in general (as measured by DCF or carbonyl moieties) predicts increases in abnormal microtubules (r = 0.91 and 0.98, respectively).
These data suggest that the reaction between O⨪2 and NO to form ONOO− is important in EtOH-induced damage. This is consistent with previous studies that have documented that O⨪2 reacts rapidly with NO to form ONOO−(Hue and Padmaja, 1993). In our studies, evidence of ONOO− generation by EtOH was confirmed by using a O⨪2 scavenger, SOD, to prevent the interaction of O⨪2and NO. This conclusion was further supported by the fact that ONOO− scavengers (urate andl-cysteine) also inhibited EtOH-induced cytoskeletal damage (and barrier dysfunction). The demonstration that EtOH administration increases NO and O⨪2 and at the same time leads to increases in stable ONOO− footprints (tubulin-associated nitrotyrosine and carbonyl) further corroborates our interpretation.
Our findings are consistent with earlier reports. For example, previous in vivo studies also implicated oxidative stress and generation of free radicals as key to the pathogenesis of a variety of GI disorders, including EtOH-mediated injury (Kvietys et al., 1990; Keshavarzian et al., 1992; Dinda et al., 1996; McKenzie et al., 1996). ONOO−-mediated damage is not limited to the GI tract and has been proposed for other organs. For example, alterations in lung function mediated by tumor necrosis factor (TNF)-α were proposed to stem from nitrated proteins such as SOD, antiproteases, and glutathione (Phelps et al., 1995).
Also, studies of systemic diseases and inflammatory GI disorders have shown that oxidation and nitration of proteins occur in vivo and that they can serve as markers of oxidative injury (Haddad et al., 1994;Ischiropoulos et al., 1995; McKenzie et al., 1996; Singer et al., 1996;Ferro et al., 1997). For example, recent studies have shown increased nitrotyrosine and iNOS up-regulation in the inflamed intestinal epithelium in vivo (Salzman et al., 1996; Singer et al., 1996; Kimura et al., 1998). Moreover, it appears that ONOO−can oxidize a variety of essential molecules (e.g., sulfhydryls, thiols, ascorbate) and trigger injurious processes, including lipid peroxidation (Muijsers et al., 1997). Indeed, a major product of the reaction of ONOO− with proteins (e.g., in macrophages, in lung tissue) is the addition of a nitro group in theortho position of tyrosine to form nitrotyrosine (Ischiropoulos et al., 1992, 1995).
Second, we showed using ONOO− or two different ONOO−-generating systems that ONOO− mimics the ability of EtOH to disrupt barrier function and to oxidize and damage microtubules. These findings are in accord with previous pharmacological studies in endothelial cells, in test-tube models, and in inflammatory processes (Ischiropoulos et al., 1992, 1995; Radi et al., 1993; Kooy and Royall, 1994; Haddad et al., 1994; Rachmilewitz et al., 1995). For example, ONOO− not only can cause chemical oxidation of luminol in the test-tube but also oxidatively injures endothelial cells (Radi et al., 1993; Phelps et al., 1995). Our finding that ONOO− can disrupt intestinal barrier function is in accord with finding by Ferro et al. (1997) for endothelial barrier dysfunction. They showed that p42 oxidation and barrier dysfunction induced by TNF-α were both mediated by NO and ONOO−.
Third, we showed that agents that scavenge ONOO−or diminish the formation of ONOO− from NO and O⨪2 attenuate the deleterious effects of both EtOH- and ONOO−-generating systems. Our findings parallel a previous study (Phelps et al., 1995) in which urate and SOD protected against TNF-α-induced, ONOO−-mediated lung endothelial injury. Interestingly, some of these same antioxidant enzymes (e.g., SOD) and free radical scavengers (e.g., uric acid) are normal constituents of all tissues (Muijsers et al., 1997). Similarly, a previous study in vivo in rat lung showed that inhibition of NO synthesis abolished the increases in protein nitrotyrosine and protein carbonyl levels (Ischiropoulos et al., 1995).
In our studies, we checked the specificity of the protective effects of antioxidants. First, we showed that analogs that lack the biological activity of these protective agents (iSOD and d-cysteine) were neither capable of protecting against EtOH-induced loss of epithelial barrier function nor had any stabilizing effects on tubulin or on microtubules. iSOD lacks the ability to scavenge O⨪2 and thus cannot inhibit the reaction of O⨪2 + NO → ONOO−. We surmise that d-cysteine does not protect against EtOH-induced damage because it is much more poorly transported into the cell interior than l-cysteine, which is the natural substrate for the amino acid carrier (Hopfer, 1987). We presume that intracellularly, l-cysteine interacts directly with intracellular ONOO−, such as is generated during EtOH exposure. However, it is also possible that l-cysteine enhances the antioxidant defenses of the cell because it is a precursor of the natural antioxidant glutathione (Van Klaveren et al., 1997).
Second, we directly measured ONOO− footprints (nitrotyrosine and carbonyl). Not only does this oxidation indicate that added ONOO− reaches the interior of our cells but also attenuation of this oxidation indicates that urate andl-cysteine are effective ONOO−scavengers. This is consistent with the conclusions of a previous study (Radi et al., 1991).
The protective superiority of 3 mM l-cysteine over 3 mMd-cysteine against added ONOO−obviously cannot be attributed to differential transport. It can be attributed to our finding that d-cysteine (15% scavenging) is not equally effective as l-cysteine (95%) in scavenging ONOO− in solution.
Other observations we have made (Banan et al., 1999a, 2000a,c) showing that growth factors protect against the injurious mechanisms induced by EtOH or oxidants are consistent with the findings of this study. The long-term goal of our laboratory is to clarify mechanisms of EtOH-induced intestinal barrier dysfunction and then to find a means of preventing this injury. Our current working hypothesis is that chemical agents that are injurious to the intestinal tract (e.g., EtOH, H2O2) cause iNOS up-regulation. The activity of the iNOS enzyme leads to intracellular increases in NO and ONOO− that directly damage (oxidize) cellular proteins such as tubulin. These effects on tubulin disrupt the microtubule cytoskeleton, causing a loss in integrity of the intestinal barrier, and predispose the organism to inflammation. Clearly, other factors are involved in intestinal injury due to these agents such as increases in intracellular calcium (Banan et al., 1999b) and damage to the actin cytoskeleton (Banan et al., 2000b,c), but microtubule damage appears to be responsible for a significant part of this injury. The effect of growth factors (e.g., EGF) is to counteract these deleterious effects (Banan et al., 1999a, 2000a,c) and to prevent or reverse iNOS induction and its sequelae. In this sense, chemically induced intestinal injury, on the one hand, and its prevention by endogenous defense and repair mechanisms, on the other hand, appear to modulate a single (iNOS-driven) pathway, in opposite directions. Aspects of this model have recently received support from studies using non-GI cell models in which EGF apparently inhibited iNOS up-regulation (Heck et al., 1992; Schini et al., 1992; Asano et al., 1994). Further studies are needed to elucidate the underlying mechanism through which EtOH up-regulates iNOS and protective agents such as EGF may prevent this up-regulation in the GI tract.
In conclusion, our present findings strongly suggest that the underlying mechanism of intestinal epithelial barrier dysfunction induced by EtOH is due to oxidative injury to the cytoskeleton. Our findings also provide avenues for development of novel therapies for alcohol-induced GI disorders such as ONOO−scavengers, iNOS inhibitors, or antioxidants.
Acknowledgment
We thank Dr. Rick Hutte at Sievers Inc. (Boulder, CO) for generous help with NO analysis.
Footnotes
-
Send reprint requests to: Ali Banan, Ph.D., Rush University Medical Center, Division of Digestive Diseases, 1725 W. Harrison, Suite 206, Chicago, IL 60612. E-mail: ali_banan{at}rush.edu
-
↵1 This work was supported in part by a grant from Rush University Medical Center. Portions of this work will be presented in the abstract form at the annual meeting of the American Gastroenterological Association in San Diego, CA, 2000.
- Abbreviations:
- GI
- gastrointestinal
- MT
- microtubule
- l-NIL
- l-N6-1-iminoethyl-lysine
- NO
- nitric oxide
- NOS
- nitric-oxide synthase
- iNOS
- inducible NOS
- EtOH
- ethanol
- EGF
- epidermal growth factor
- FSA
- fluorescein sulfonic acid
- SOD
- superoxide dismutase
- iSOD
- heat-inactivated SOD
- LSCM
- laser scanning confocal microscopy
- l-NNA
- NG-nitro-l-arginine
- HRP
- horseradish peroxidase
- l-Arg
- l-arginine
- l-NMMA
- NG-monomethyl-l-arginine
- DNP
- 2,4-dinitrophenylhydrazine
- DCF
- dichlorofluorescein
- DCFD
- 2′,7′-dichlorofluorescein diacetate
- X
- xanthine
- XO
- xanthine oxidase
- SIN-1
- 1,3-morpholinosydnonymine
- SNAP
- S-nitroso-N-acetyl penicillamine
- D-PBS
- Dulbecco's PBS
- Received February 9, 2000.
- Accepted May 31, 2000.
- The American Society for Pharmacology and Experimental Therapeutics