Secondary Logo

Journal Logo


Increased iNOS Activity is Essential for Intestinal Epithelial Tight Junction Dysfunction in Endotoxemic Mice

Han, Xiaonan*; Fink, Mitchell P.*,‡; Yang, Runkuan*; Delude, Russell L.*,†

Author Information
doi: 10.1097/01.shk.0000112346.38599.10
  • Free



The normal functioning of the lungs, liver, kidneys, and intestine, among other organs, depends on the establishment and maintenance of compositionally distinct compartments that are lined by sheets of epithelial cells (1). An essential element in this process is the formation of tight junctions (TJs) between adjacent cells making up the epithelial sheet. The TJ serves as a fence that differentiates the cytosolic membrane into apical and basolateral domains. This fence function is essential for establishing and maintaining cellular polarity. Transcellular vectorial transport processes, which depend upon proper cellular polarization, generate distinct internal environments in the opposing compartments defined by the epithelial sheet. In addition, the TJ acts as a regulated semipermeable barrier that limits the passive diffusion of solutes across the paracellular pathway between adjacent cells. Thus, the barrier function of the TJ is necessary to prevent dissipation of the concentration gradients that exist between the two compartments defined by the epithelium. In some organs, notably the gut and the lung, this barrier function is also important to prevent systemic contamination by microbes and toxins that are present in the external environment (1).

It is becoming increasingly apparent that inflammatory mediators can alter the expression and/or localization of a number of TJ proteins in various epithelia. In many cases, these changes in protein expression and/or localization are associated with functional alterations in epithelial barrier function. For example, our laboratory recently reported that incubating monolayers of Caco-2 human enterocyte-like cells with a mixture of proinflammatory cytokines decreases the expression of the TJ proteins, zonula occludens (ZO)-1 and occludin, and increases apical-to-basolateral permeation across this model epithelium by the hydrophilic macromolecule, fluorescein isothiocyanate (FITC)-labeled dextran (Mr = 4 kDa:FD4) (2). We followed up this in vitro study with two studies using mice challenged with a sublethal dose of lipopolysaccharide (LPS) as an in vivo model of systemic inflammation. In one of these studies, we showed that ZO-1 and occludin localization and expression is markedly altered within the pulmonary epithelium in endotoxemic mice (3). These changes in TJ protein expression were associated with evidence of increased bronchoalveolar epithelial permeability. In another study, we showed that injecting mice with LPS leads to alterations in the expression and localization of ZO-1 and occludin in the liver as well as evidence of hepatobiliary epithelial dysfunction (4). Interestingly, in all three of these studies, we obtained data in support of the view that cytokine- or LPS-induced alterations in epithelial TJ structure and function are dependent on increased expression and activity of inducible nitric oxide (NO·) synthase (iNOS) and increased production of NO· on this basis (2–4).

Herein, we sought to extend our studies of the effects of systemic inflammation on TJ structure and function by assessing the effects of acute endotoxemia in mice on ileal mucosal permeability as well as ZO-1, ZO-2, ZO-3, occludin, and claudin-1 expression and localization in samples of ileal and colonic mucosa. To test the hypothesis that LPS-induced changes in gut epithelial TJ structure and function are iNOS-dependent, we carried out studies using genetic or pharmacological approaches to ablate or inhibit iNOS activity.



All chemicals were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO) unless otherwise noted. l-N(6)-(1-iminoethyl)lysine (L-NIL) was purchased from A.G. Scientific (San Diego, CA). Anti-claudin-1, anti-occludin, and anti-ZO-1 polyclonal antibodies (pAb) were from Zymed Laboratories (South San Francisco, CA). Anti-actin mAb was from Sigma-Aldrich. Anti-β-actin mAb and rabbit pAb against ZO-2 and ZO-3 were from Santa Cruz Biotechnology (Santa Cruz, CA). Unconjugated and FITC-conjugated anti-iNOS mAb were from TransLabs (Lexington, KY). All secondary antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA).


This research complied with regulations regarding animal care as published by the National Institutes of Health and was approved by the Institutional Animal Use and Care Committee of the University of Pittsburgh. Male 7- to 8-week-old C57B1/6J mice weighing 20 to 25 g were from Jackson Laboratories (Bar Harbor, ME). iNOS knockout (iNOS−/−) mice on a C57B1/6J background and weighing 22 to 25 g were generously provided by Dr. T.R. Billiar (University of Pittsburgh). Construction of the iNOS−/− mice was previously described (5). All animals were maintained in the University of Pittsburgh Animal Research Facility on a 12-h light/dark cycle with free access to standard laboratory chow and water. Animals were not fasted before experiments. Animals were anesthetized before surgical procedures by sodium pentobarbital (60-90 mg/kg s.c.).

To induce a systemic inflammatory response, mice were injected intraperitoneally with Escherichia coli (strain O111:B4) LPS (40–50 μg/mouse; 2 mg/kg) dissolved in 1.0 mL of phosphate-buffered saline (PBS). Control animals were injected with a similar volume of PBS without LPS. Some mice were treated with two 5-mg/kg doses of L-NIL administered by intraperitoneal injection 2 and 8 h after the injection of LPS or PBS. Some mice were treated with L-NIL according to the same schedule in the absence of prior injection of LPS. Groups of mice were anesthetized 6, 12, or 18 h after injection of LPS or PBS for measuring ileal mucosal barrier function (see below) or harvesting tissue specimens for various biochemical, histological, or physiological assays.

Measurement of intestinal mucosal permeability

Ileal mucosal permeability to FITC-Dextran with an Mr of 4,000 Da (FD4) was determined using an everted gut sac method as previously described (6). Permeability is expressed as the mucosal-to-serosal clearance of FD4. Bacterial translocation was determined by culturing mesenteric lymph node (MLN) homogenates on MacConkey agar plates exactly as described (7). Results are expressed as colony-forming units per gram of tissue.

NP-40-insoluble and total protein extracts

An 8- to 9-cm segment of ileum or the entire length of the colon was gently scraped twice with a glass microscope slide to obtain the mucosal tissue. The scrapings were homogenized on ice with a Polytron tissue homogenizer in 1 mL of NP-40 lysis buffer (25 mM HEPES, pH 7.4, 150 mM NaCl, 4 mM EDTA, 25 mM NaF, 1% NP-40, 1 mM Na3 VO4, 1 mM 4-amidinophenylmethanesulfonyl fluoride [APMSF], 10 μg/mL leupeptin, and 10 μg/mL aprotinin). The samples were gently rocked at 4°C for 30 min. The sample was centrifuged at 12,000 g for 30 min at 4°C and the pellets were resuspended in SDS-dissolving buffer (25 mM HEPES, pH 7.5, 4 mM EDTA, 25 mM NaF, 1% SDS, and 1 mM Na3VO4) using five strokes with a Dounce homogenizer (pestle B) followed by sonication with a 0.1 W Sonic Dismembrator fitted with a microtip on power setting 3 (Fisher Scientific, Pittsburgh, PA). Sonication continued until the precipitate was completely dissolved. Samples prepared in this manner were designated the “NP-40-insoluble fractions.”

For total cellular protein extracts, mucosal specimens were homogenized in cold RIPA buffer (PBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.1 mg/mL APMSF, 1.0 mM sodium orthovanadate, and 1× mammalian protease inhibitor cocktail [Sigma-Aldrich catalog no. P 8340]) and were sonicated until the sample was completely dissolved. These samples are referred to as “total protein” extracts in the manuscript.


One hundred micrograms of NP-40 insoluble fraction was immunoprecipitated with either occludin or ZO-1 pAb. The lysate was precleared by adding 0.25 μg of normal mouse IgG, together with 20 μL of suspended protein A/G agarose (Santa Cruz Biotechnologies). After incubation at 4°C for 30 min, the beads were collected by centrifugation at 2,500 rpm for 5 min at 4°C. The supernatant was transferred to a fresh tube, and 3 μg of anti-occludin or anti-ZO-1 antibody was added and the tube was incubated on a rocker platform for 2 h at 4°C. Twenty microliters of resuspended agarose A/G was added to the tube, and the incubation was continued overnight at 4°C with gentle shaking. The agarose beads were washed five times with 1 mL of NP-40 lysis buffer. Proteins were eluted by boiling in 1× Laemmli buffer (10% glycerol, 5% β-mercaptoethanol, 2.5% SDS, 0.1 M Tris, pH 6.8, and 0.2% bromphenol blue) for 10 min. Equal volumes of sample were separated by gel electrophoresis followed by immunoblotting.


Equal amounts of total protein extract were mixed in 1× Laemmli buffer, boiled for 3 min, and centrifuged for 10 s. The supernatants were electrophoresed on 7.5% or 12% precast SDS-polyacrylamide gels (Bio-Rad, Hercules, CA). The proteins were electroblotted onto Hybond-P PVDF membranes (Amersham Pharmacia Biotech, Leicester, UK) and were blocked with Blotto (1× TBS, 5% milk, 0.05% Tween-20, and 0.2% NaN3) or 1:10 normal donkey serum (ZO-2 and ZO-3 only) for 60 min. The filter was incubated at room temperature for 1 h with anti-ZO-1 or anti-actin antibody at a 1:4000 dilution or anti-claudin-1, anti-ZO-2, anti-ZO-3, or anti-iNOS antibody diluted 1:2000 in PBS and 0.02% Tween-20 (PBST). After washing three times in PBST, immunoblots were exposed for 1 h to a 1:20,000 dilution of the appropriate horseradish peroxidase-conjugated secondary antibody. After three washes in PBST and two washes in PBS, the membrane was impregnated with the enhanced chemiluminescence substrate (ECL; Amersham Pharmacia Biotech) and was used to expose x-ray film. Autoradiographs were captured using a Hewlett Packard (Palo Alto, CA) ScanJet 6300s. Band intensities were quantified by densitometry and expressed as mean area density using GelExpert 3.5 software (Nucleotech Corporation, San Mateo, CA). For total protein blots, mean area density is expressed relative to actin expression.


Frozen tissue sections (6 μm) were prefixed in cold acetone and then air dried. The sections were fixed with 4% paraformaldehyde and then washed three times with cold PBS. The sections were blocked with 10% donkey serum. Tissue sections were incubated with primary antibodies as follows: rabbit anti-ZO-1 pAb and rabbit anti-occludin pAb. Some sections were double stained with FITC-conjugated iNOS mAb. Some sections were incubated with PBS alone in place of primary antibody as a negative control (data not shown). After a 1-h incubation at room temperature, the sections were washed three times with PBS. TRITC-conjugated goat anti-rabbit secondary antibody was added and incubated for 45 min at room temperature. The sections were washed three times with PBS, and the nuclei were stained with 4´,6-diamidino-2-phenylindole dihydrochloride (DAPI; Molecular Probes, Eugene, OR). Coverslips were mounted using a Antifade kit (Molecular Probes). Thin sections were obtained from paraffin-embedded tissue specimens and were stained with hematoxylin and eosin. Images were captured using an Provis fluorescence microscope (Olympus, Melville, NY).

Statistical analyses

Results are presented as means ± SEM. Data were analyzed using analysis of variance (ANOVA) followed by Fisher’s LSD test. P values of <0.05 were considered significant.


Endotoxemia increases intestinal mucosal permeability

Ileal mucosal permeability to FD4 increased in a time-dependent manner in mice injected with LPS (Fig. 1A). Treatment of endotoxemic mice with L-NIL, an isoform-selective iNOS inhibitor (8), significantly modulated the LPS-induced increase in ileal mucosal permeability (Fig. 1B). Basal ileal mucosal permeability in control (PBS-treated) iNOS−/− mice was greater than that measured in control iNOS+/+ mice (two open bars in Fig. 1C), a finding that is consistent with reports that basal levels of iNOS-derived NO· are required for normal gut homeostasis (9,10). Despite the apparent basal defect in intestinal barrier function in iNOS−/− mice, permeability to FD4 did not increase further when these mice were challenged with LPS (Fig. 1C).

Fig. 1
Fig. 1:
LPS decreases mucosal barrier function in an NO·-dependent manner. (A) Everted gut sacs were prepared from mice treated with LPS (2 mg/kg; n = 8/group) or PBS (n = 7/group) alone at the times indicated. Mucosal-to-serosal clearance was determined by measuring FD4 clearance. (B) The everted gut sac technique was used to measure ileal mucosal permeability in mice treated with PBS (n = 5), L-NIL alone (2 × 5 mg/kg; n = 5), LPS (2 mg/kg; n = 5), or LPS + L-NIL (n = 6). Gut sacs were prepared 12 h after administration of LPS. (C) Everted gut sacs were prepared from iNOS+/+ or iNOS−/− mice 12 h after injection with LPS or PBS (n = 5 per group). Values are means ± SEM; * P < 0.05; ** P < 0.01.

Bacterial translocation from the gut lumen to MLN is another measure of in vivo mucosal barrier function (1). Endotoxemia increased the number of bacteria that were recovered from MLN obtained from iNOS+/+ mice (Fig. 2A). Treatment of endotoxemic mice with L-NIL decreased LPS-induced bacterial translocation (Fig. 2B). Similarly, LPS failed to induce bacterial translocation in iNOS−/− mice (Fig. 2C).

Fig. 2
Fig. 2:
Increased bacterial translocation caused by LPS is NO· dependent. (A) The MLN complex was harvested from the animals used for Figure 1. The tissue was homogenized and plated on MacConkey agar at the times indicated. Bacterial translocation was prevented in mice treated with L-NIL (B) or in iNOS−/− mice (C). * P < 0.05; ** P < 0.01.

TJ protein expression decreases in ileal mucosa of endotoxemic mice

Ileal tissue was used to prepare total and NP-40-insoluble protein extracts, the latter being enriched for cytoskeletal and associated TJ proteins (11). Total protein extracts were prepared and subjected to immunoblotting. NP-40-insoluble proteins were first solubilized with SDS buffer and were concentrated by immunoprecipitation before immunoblotting.

The expression of occludin in NP-40-insoluble extracts was decreased in samples obtained 6 h after injecting mice with LPS (Fig. 3). Occludin expression in NP-40-insoluble extracts (i.e., TJ-associated occludin) decreased still further 12 h after the induction of endotoxemia, and was starting to return toward normal at 18 h after LPS challenge. In total protein extracts, the changes in occludin levels were less dramatic, and the maximal decrease was observed at 12 h. ZO-1 expression decreased slightly in total protein extracts from ileal mucosa of mice exposed to LPS. However, there was a greater than 50% decrease in ZO-1 levels in the NP-40-insoluble fraction. Total extractable ZO-2 decreased slightly 12 h after the induction of endotoxemia, and further decreased over the following 6-h period. ZO-3 expression decreased 6 h after LPS administration and decreased still further at 12 h. Claudin-1 expression increased in total protein extracts prepared from ileal mucosa. Immunoblotting total protein extracts for actin revealed equivalent loading of the samples in these gels. iNOS protein expression increased in total protein extracts from ileal mucosa of LPS-treated mice.

Fig. 3
Fig. 3:
TJ protein expression decreases in ileum of endotoxemic mice. Mice were injected intraperitoneally with LPS or PBS (Con), and sections of ileum were removed at the times indicated. Total protein- and NP-40-insoluble protein extracts were subjected to immunoblotting. OCC, occludin; cldn1, claudin-1. The results are representative of four to five blots using different mice. The bar graphs are relative mean area density of the total protein blots ± SEM.

Endotoxemia is associated with derangements in ileal mucosal TJ protein localization

All of the immunohistochemical studies of tissues from endotoxemic mice were performed using samples harvested 12 h after injection of LPS. Both occludin and ZO-1 formed a continuous staining pattern around the enterocyte layer near the apical region of the lateral membrane of crypt and villous cells of the epithelium and the endothelium of the lamina propria from normal mice (see panels labeled “Con” in Fig. 4). Light punctate and diffuse staining of occludin and ZO-1 was observed throughout the cytoplasm using confocal microscopy (insets). After injection of mice with LPS, ZO-1 and occludin staining was maintained in the crypts, but staining progressively decreased over the tips of the villi (Fig. 4, panels labeled “LPS”). In sections from endotoxemic mice, the staining patterns for ZO-1 and occludin were disrupted only in focal regions of the ileum; approximately 60% of the villi in a given section stained normally (data not shown). Examination of hematoxylin and eosin (H&E) stained thin sections did not reveal significant differences between the different treatment groups (Fig. 4, panel labeled “H&E”).

Fig. 4
Fig. 4:
Indirect immunofluorescence of ileum from control mice (Con), mice exposed to LPS for 12 h (LPS), and endotoxemic mice that were treated with L-NIL (LPS + L-NIL). In ileum from control mice occludin and ZO-1 staining was continuous along the villous epithelium (TRITC-conjugated goat anti-rabbit secondary antibody, red fluorescence). LPS caused marked down-regulation of both proteins in the epithelium. L-NIL treatment ameliorated the deleterious effects of LPS on the localization of both proteins. Arrows point to regions within the lower magnification fields (fluorescence microscopy) that appear in confocal images (insets). All occludin images and the LPS + L-NIL panels are at 400x, and the Con and LPS ZO-1 panels are at 600x. The bar is 50 μm. Sections from at least five mice were examined for each condition.

Blockade of iNOS activity modulates decreased ZO-1 and occludin expression in ileum of endotoxemic mice

NP-40-insoluble proteins were prepared from ileal mucosal samples from control and endotoxemic mice exposed to L-NIL or vehicle alone. These extracts were subjected to immunoprecipitation followed by immunoblotting. Treatment of endotoxemic mice with L-NIL significantly modulated LPS-induced decreases in occludin and ZO-1 expression (Fig. 5). However, this treatment did not completely block decreased expression of these proteins. L-NIL preserved the correct targeting of these proteins to TJs as revealed using immunohistochemistry (Fig. 4, panels labeled “LPS + L-NIL”). However, the intensity of staining was still somewhat less than that observed in control mice.

Fig. 5
Fig. 5:
Pharmacologic inhibition and genetic ablation of iNOS activity modulates decreased TJ protein expression in ileum of endotoxemic mice. Mice were injected with LPS or PBS (Con), sections of ileum were removed 12 h later, and NP-40-insoluble protein extracts were subjected to immunoblotting. The results are representative of four to six blots on different mice. The bar graphs show relative mean area density of the ZO-1 blots ± SEM.

Similar experiments were performed using iNOS−/− mice. These studies were complicated by differences in the basal levels of expression of occludin and ZO-1 in control iNOS−/− versus iNOS+/+ mice. The levels of occludin and ZO-1 in ileal mucosa from control iNOS−/− mice were reproducibly lower than the levels of these proteins in control iNOS+/+ mice. However, treatment of iNOS−/− mice with LPS did not lead to a further decrease in expression of ZO-1 or occludin in ileal mucosa (Fig. 5). The localization of ZO-1 and occludin was preserved in ileal sections prepared from LPS-treated iNOS−/− mice, being essentially unchanged from what was observed in sections from iNOS−/− animals injected with vehicle (Fig. 6). No differences in mucosal architecture were apparent between wild-type and iNOS−/− mice when H&E-stained paraffin-embedded sections were examined using light microscopy.

Fig. 6
Fig. 6:
Indirect immunofluorescence of ileum from control iNOS−/− mice (iNOS−/−) and iNOS−/− mice exposed to LPS for 12 h (iNOS−/− + LPS). There was comparable occludin and ZO-1 staining observed in the villi and crypts of both groups. Original magnification is 400× (OCC) and 200× (ZO-1). Bar = 50 μm.

TJ protein expression decreases in colon of endotoxemic mice

Endotoxemia increased iNOS and claudin-1 expression and decreased occludin and ZO-1 expression in total protein extracts from colonic mucosa (Fig. 7). Note that NP-40-insoluble extracts from colon were not prepared due to the small amount of material that was available from each mouse.

Fig. 7
Fig. 7:
TJ protein expression decreases in colon of endotoxemic mice. Mice were injected with LPS or PBS (Con), sections of colon were removed at the times indicated, and total protein extracts were subjected to immunoblotting. The results are representative of four to six blots using different mice. The bar graphs present relative mean area density of the ZO-1 blots ± SEM.

Endotoxemia is associated with derangements in colonic mucosal TJ protein localization

In normal colon, occludin and ZO-1 appeared as continuous bands along the epithelial sheet from the crypt to the villous tip (Fig. 8, panel labeled “Con”). Staining of endothelium was also evident within the lamina propria. Tissues obtained from endotoxemic mice showed extensive decreases in the staining for both proteins over the entire colonic epithelium with some residual staining remaining in crypts (Fig. 8, panel labeled “LPS”). Note that staining of the endothelium in the lamina propria was well preserved. The sections prepared from control and LPS-treated mice were double stained for ZO-1 (red) and iNOS (green). Whereas iNOS immunofluorescence was undetectable in control specimens, diffuse cytoplasmic staining for iNOS in the epithelium was observed in samples from endotoxemic mice.

Fig. 8
Fig. 8:
Indirect immunofluorescence of colon from control mice (Con), mice exposed to LPS for 12 h (LPS), and endotoxemic mice that were treated with L-NIL (LPS + L-NIL). Continuous occludin and ZO-1 immunostaining were observed in epithelium of the Con group. In the LPS group, occludin and ZO-1 staining was globally down-regulated, and iNOS expression (FITC-conjugated iNOS mAb, green staining) was increased. Occludin and ZO-1 expression was largely preserved in LPS + L-NIL mice. Original magnification is 200× (OCC) and 600× (ZO-1), inset is confocal analysis at 1000x. Bar = 50 μm.

Pharmacologic and genetic blockade of iNOS activity modulates decreased ZO-1 and occludin expression in colonic mucosa of endotoxemic mice

Treatment with L-NIL preserved normal occludin expression and ameliorated changes in ZO-1 expression in total protein extracts obtained 12 h after the onset of endotoxemia (Fig. 9). Although basal levels of colonic mucosal occludin and ZO-1 expression were higher in vehicle-treated iNOS−/− as compared with vehicle-treated iNOS+/+ mice, the change in expression of these proteins in colonic mucosa induced by LPS was much less in iNOS−/− mice as compared with iNOS+/+ mice (Fig. 9). L-NIL preserved the proper localization of ZO-1 and occludin in animals exposed to LPS (Fig. 8, panel labeled “LPS + L-NIL”). The localization of ZO-1 and occludin was also similar in colonic sections from iNOS−/− and iNOS+/+ animals (data not shown).

Fig. 9
Fig. 9:
Pharmacological and genetic inhibition of iNOS activity modulates decreased TJ protein expression in colon of endotoxemic mice. Mice were injected with LPS or PBS (Con), sections of ileum were removed 12 h later, and NP-40 insoluble protein extracts were prepared and subjected to immunoblotting. The results are representative of four to six blots using different mice.


In the present study, we used several techniques to document that epithelial barrier function was deranged in the intestines of endotoxemic animals. Specifically, two different assays (permeability to FD4 and enumeration of viable bacteria in MLN) showed that injecting mice with LPS impaired gut barrier function. Using immunoblotting and immunofluorescence, we showed that injecting mice with LPS markedly down-regulated the expression of several key TJ proteins in ileal mucosa and colonic mucosa. H&E-stained sections of ileum (Fig. 3) and colon (data not shown) from mice injected 12 h earlier with LPS did not show obvious morphological changes. Finally, by using a combination of genetic and pharmacological approaches, we obtained evidence that these changes in the structure and function of TJs in the epithelial lining of the intestine were iNOS-dependent.

Combined with other data recently reported by our laboratory (2–4), the findings reported here support the view that the structural and functional changes in TJ formation that are induced by LPS in the ileum and colon are part of an apparently generalized phenomenon that involves the epithelial components in multiple organs. It is noteworthy that the organs involved, namely, the lung (3), the liver (4), and the gut (present data), are all commonly involved in critically ill patients with the multiple organ dysfunction syndrome (MODS), a common life-threatening complication of sepsis and other conditions (e.g., acute pancreatitis or severe trauma) that are associated with dysregulation of the systemic inflammatory response (12–17).

We previously reported that plasma concentrations of nitrite plus nitrate (NOx) are increased in mice injected with the dose of E. coli O111:B4 LPS used for the studies reported here (2 mg/kg) (3). Moreover, in this previous report, we documented that circulating NOx concentration is significantly decreased when endotoxemic mice are treated with the same dosing schedule for L-NIL used for the studies reported here (3). Accordingly, these studies were not repeated here.

We also previously reported that pharmacologic iNOS inhibition ameliorates LPS-induced bacterial translocation and intestinal epithelial hyperpermeability in rats (18). Mishima et al. (19) showed that LPS induces bacterial translocation to MLN in iNOS+/+ but not iNOS−/− mice. Cuzzocrea et al. (20) showed that intraperitoneal administration of zymosan increases gut mucosal permeability to FD4 in wild-type but not iNOS−/− mice and resulted in discontinuous ZO-1, occludin, and β-catenin immunostaining in ileum of wild-type but not iNOS−/− mice. Thus, some of the findings reported here confirm findings that have been previously reported by our laboratory or other investigators. Some of our observations are novel and add to our growing understanding of inflammation-induced alterations in epithelial TJ structure and function. Specifically, we showed herein that both ZO-2 and ZO-3 expression transiently decrease in ileal mucosa after the injection of LPS. Moreover, we documented changes in TJ protein expression and localization not only in ileal mucosa but also colonic mucosa as well.

Prior in vitro studies have shown that the permeability of cultured enterocytic monolayers is increased when the cells are incubated with various NO· donors, such as S-nitroso-N-acetylpenicillamine (SNAP) or sodium nitroprusside (21–24). This effect does not appear to be mediated by NO· directly, but rather by ONOO, a potent oxidizing and nitrosating agent that is formed when NO· reacts with O2·. Support for this view comes from studies showing that SNAP-induced hyperpermeability is augmented by addition of diethyldithiocarbamate, a superoxide dismutase inhibitor, or pyrogaliol, an O2· generator (24). Furthermore, SNAP-induced hyperpermeability is blocked by Tiron, an agent that scavenges O2·, as well as various ONOO scavengers such as urate and deferoxamine (24). When Caco-2 monolayers are incubated with IFN-γ or a combination of IFN-γ, TNF-α, and IL-1β, iNOS expression is induced and permeability is increased (25,26). The increase in permeability can be blocked by inhibiting NO· production or scavenging ONOO (25,26). These findings support the view that cytokine-induced intestinal epithelial hyperpermeability is mediated, at least in part, by the formation ONOO. However, further studies are required to determine if ONOO, or some other reactive nitrogen intermediate, is the actual moiety responsible for the deleterious effects of NO· on TJ protein expression that we observed in organs of endotoxemic animals.

We recently carried out an extensive analysis of cytokine-induced barrier dysfunction using Caco-2 cells in an attempt to better understand the effects of NO· on the expression and localization of several TJ proteins (2). In these studies, we showed that NO· generated endogenously as the result of iNOS activity or exogenously from the NO· donor DETA-NONOate {(Z)-1-[2-(2-aminoethyl)-N-(2-ammonioethyl) amino]diazen-1-ium-1,2-diolate]} decreased the expression and impaired proper localization of ZO-1, ZO-3, and occludin. With a few of exceptions, the findings reported herein recapitulate our recent in vitro observations using Caco-2 monolayers. In the in vivo studies reported here, treatment with the isoform-selective iNOS inhibitor, L-NIL, or genetic ablation of functional iNOS expression largely prevented the structural and functional epithelial changes induced by LPS. L-NIL probably has both local (enterocyte) and systemic effects (inhibiting iNOS in various parenchymal cells in other tissues and in circulating and resident myeloid cells). Nevertheless, we suspect that at least part of the protection observed by iNOS inhibition in this model is due to the effect of L-NIL on iNOS-dependent NO· production in enterocytes. In our recent cell culture study, coincubation of immunostimulated Caco-2 cells with 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide, an NO· scavenger, similarly ameliorated changes in TJ protein expression and localization as well as the functional read-out (i.e., permeability of the monolayers to FD4). These data strongly suggest that the mechanism(s) responsible for the alterations in TJ protein expression and localization are similar in the two model systems. Taken together, our current and recently reported (2–4) observations support the view that induction of iNOS and increased production of NO· contributes to the development of epithelial dysfunction in multiple organs in endotoxemic mice.

We observed down-regulation of ZO-2 expression in our in vivo studies, but ZO-2 expression did not change appreciably when Caco-2 cells were incubated with proinflammatory cytokines. The reasons for these differences are unknown, but may relate to species (mouse versus human), timing of observations, cell type (nontransformed versus transformed), or stimulus (LPS versus cytokine cocktail).

Basal ileal mucosal permeability to FD4 was greater in iNOS−/− as compared with iNOS+/+ mice. Interestingly, we also observed this difference in mucosal permeability to FD4 in mice lacking the IL-6 gene when compared with wild-type mice (27). In another study, we similarly observed that basal circulating levels of conjugated bilirubin, but not bile salts, were significantly higher in control iNOS−/− mice as compared with control iNOS+/+ mice (4). This finding might reflect increased paracellular diffusion of conjugated bilirubin from the bile into blood, a process that is normally limited by hepatobiliary TJs. Although we can only speculate about the mechanism responsible for these observations, the possibility exists that the low background rate of NO· production from constitutively expressed iNOS in normal enterocytes is necessary for the proper assembly of TJs. We (Fig. 3) and others (28,29) have detected iNOS protein expression in ileal mucosa of normal control animals by Western blotting. Others have also documented detectable iNOS-dependent NO· production from resting normal enterocytes (29) and ileal mucosa (28). However, Morin et al. (30) failed to detect iNOS in ileum of normal rats, which may be the result of the different species used or the sensitivity of the assays performed. In support of a role for basal NO· production in normal ileal homeostasis, we cite data reported a number of years ago by Kubes (9), who showed that mucosal permeability is increased when autoperfused segments of cat ileum are infused intra-arterially with the isoform-nonselective NOS inhibitor, NG-nitro-L arginine methyl ester.

The mechanism(s) responsible for the NO·-dependent decrease in TJ protein expression and consequent epithelial barrier dysfunction remains to be determined. One hypothesis was recently suggested by Sugi et al. (31), who sought to explain the results of studies using T84 enterocyte-like cells exposed to IFN-γ. These investigators reported that intracellular Na+ concentration and cell volume increase after exposure to IFN-γ. Moreover, these changes preceded paracellular barrier dysfunction. Remarkably, growing cells in low Na+ medium significantly blocked the effects of IFN-γ on these parameters and also ameliorated decreased occludin expression caused by chronic IFN-γ exposure. Their studies also were consistent with NO· production being responsible for the observed epithelial dysfunction. Endotoxemia is associated with decreased function of epithelial Na+, K+-ATPase activity in liver (32), intestine (33), and lung (34) during endotoxemia. Therefore, it is plausible that increased NO· production is (somehow) responsible for decreased Na+ transport activity, and the resulting imbalance in Na+ homeostasis leads to cell swelling with subsequent dysregulation of tight junction protein expression and targeting in these tissues. Further studies will be required to determine whether this mechanism is applicable to epithelia other than T84 cells (e.g., hepatocytes or alveolar epithelial cells) and, particularly, whether it pertains to alterations in epithelial TJ structure and function induced by inflammation in vivo.

Contemporary strategies aimed at treating sepsis patients rely mainly on providing support for the circulation and failing organs. Until the recent approval of drotrecogin α (activated), specific therapies aimed at the underlying pathophysiological mechanisms in sepsis were unavailable to clinicians (4). Drotrecogin α (activated), as well as a number of other failed therapeutic strategies, such as pharmacological neutralization of TNF-α or IL-1β, target the inflammatory response per se and/or sepsis-induced alterations in the microcirculation. Collectively, the findings reported previously by Han et al. (3,4) with the data presented herein suggest that another therapeutic strategy might be to interrupt the downstream pathophysiological mechanisms that are responsible for organ system dysfunction that are caused by a dysregulated inflammatory response. Specifically, our findings suggest that pharmacological inhibition of iNOS using an isoform-selective agent might ameliorate many of the features of MODS. Further studies to evaluate this concept are warranted.


The authors thank Meaghan E. Killeen and Peter S. Williams for their technical assistance, and Dr. Simon S. Watkins and Sean Alber for assistance with the immunohistochemical staining and imaging.


1. Stevenson BR: Understanding tight junction clinical physiology at the molecular level. J Clin Invest 104:3–4, 1999.
2. Han X, Fink MP, Delude RL: Proinflammatory cytokines cause NO-dependent and independent changes in expression and localization of tight junction proteins in intestinal epithelial cells. Shock 19:229–237, 2003.
3. Han X, Fink MP, Uchiyama T, Delude RL: Increased iNOS activity is essential for pulmonary epithelial tight junction dysfunction in endotoxemic mice. Am J Physiol Lung Cell Mol Physiol 286:L259–L267, 2004.
4. Han X, Fink MP, Uchiyama T, Yang R, Delude RL: Increased iNOS activity is essential for hepatic epithelial tight junction dysfunction in endotoxemic mice. Am J Physiol Gastrointest Liver Physiol 286:G126–G136, 2004.
5. MacMicking JD, Nathan C, Horn G, Chartrain N, Fletcher DS, Trumbauer M, Stevens K, Quia-wen X, Sokol K, Hutchinson N, Chen H, Mudgett JS: Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide synthase. Cell 81:641–650, 1995.
6. Wattanasirichaigoon S, Menconi MJ, Delude RL, Fink MP: Lisofylline ameliorates intestinal mucosal barrier dysfunction caused by ischemia and ischemia/reperfusion. Shock 11:269–275, 1999.
7. Yang R, Gallo DJ, Baust JJ, Uchiyama T, Watkins SK, Delude RL, Fink MP: Ethyl pyruvate modulates inflammatory gene expression in mice subjected to hemorrhagic shock. Am J Physiol Gastrointest Liver Physiol 283:G212–G221, 2002.
8. Moore WM, Webber RK, Jerome GM, Tjong FS, Misko TP, Currie MG: l -N6-(1-iminoethyl)lysine: a selective inhibitor of inducible nitric oxide synthase. J Med Chem 37:3886–3888, 1994.
9. Kubes P: Nitric oxide modulates epithelial permeability in the feline small intestine. Am J Physiol 262:G1138–G1142, 1992.
10. Payne D, Kubes P: Nitric oxide donors reduce the rise in reperfusion-induced intestinal mucosal permeability. Am J Physiol 265:G189–G195, 1993.
11. Sakakibara A, Furuse M, Saitou M, Ando-Akatsuka Y, Tsukita S: Possible involvement of phosphorylation of occludin in tight junction formation. J Cell Biol 137:1393–1401, 1997.
12. Tenner S, Sica G, Hughes M, Noordhoek E, Feng S, Zinner M, Banks PA: Relationship of necrosis to organ failure in severe acute pancreatitis. Gastroenterology 113:899–903, 1997.
13. Doig CJ, Sutherland LR, Sandham JD, Fick GH, Verhoef M, Meddings JB: Increased intestinal permeability is associated with the development of multiple organ dysfunction syndrome in critically ill ICU patients. Am J Respir Crit Care Med 158:444–451, 1998.
14. Moore FA, Sauaia A, Moore EE, Haenel JB, Burch JM, Lezotte DC: Postinjury multiple organ failure: a bimodal phenomenon. J Trauma 40:501–512, 1996.
15. Russell JA, Singer J, Bernard GR, Wheeler A, Fulkerson W, Hudson L, Schein R, Summer W, Wright P, Walley KR: Changing pattern of organ dysfunction in early human sepsis is related to mortality. Crit Care Med 28:3405–3411, 2000.
16. Franson TR, LaBrecque DR, Buggy BP, Harris GJ, Hoffmann RG: Serial bilirubin determinations as a prognostic marker in clinical infections. Am J Med Sci 297:149–152, 1989.
17. Vincent JL, Angus DC, Artigas A, Kalil A, Basson BR, Jamal HH, Johnson III, G Bernard GR: Recombinant human activated protein C worldwide evaluation in severe sepsis (PROWESS) Study Group: effects of drotrecogin alfa (activated) on organ dysfunction in the PROWESS trial. Crit Care Med 31:834–840, 2003.
18. Unno N, Wang H, Menconi MJ, Tytgat SHAJ, Larkin V, Smith M, Morin MJ, Chavez A, Hodin RA, Fink MP: Inhibition of inducible nitric oxide synthase ameliorates lipopolysaccharide-induced gut mucosal barrier dysfunction in rats. Gastroenterology 113:1246–1257, 1997.
19. Mishima S, Xu D, Lu Q, Deitch EA: Bacterial translocation is inhibited in inducible nitric oxide synthase knockout mice after endotoxin challenge but not in a model of bacterial overgrowth. Arch Surg 132:1190–1195, 1997.
20. Cuzzocrea S, Mazzon E, De Sarro A, Caputi AP: Role of free radicals and poly(ADP-ribose) synthetase in intestinal tight junction permeability. Mol Med 6:766–778, 2000.
21. Xu DZ, Lu Q, Deitch EA: Nitric oxide directly impairs intestinal barrier function. Shock 17:139–145, 2002.
22. Forsythe RM, Xu DZ, Lu Q, Deitch EA: Lipopolysaccharide-induced enterocyte-derived nitric oxide induces intestinal monolayer permeability in an autocrine fashion. Shock 17:180–184, 2002.
23. Salzman AL, Menconi MJ, Unno N, Ezzell RM, Casey DM, Gonzalez PK, Fink MP: Nitric oxide dilates tight junctions and depletes ATP in cultured Caco-2BBe intestinal epithelial monolayers. Am J Physiol 268:G361–G373, 1995.
24. Menconi MJ, Unno N, Smith M, Aguirre DE, Fink MP: The effect of nitric oxide donors on the permeability of cultured intestinal epithelial monolayers: role of superoxide, hydroxyl radical, and peroxynitrite. Biochem Biophys Acta 1425:189–203, 1998.
25. Unno N, Menconi MJ, Smith M, Fink MP: Nitric oxide mediates interferon-gamma-induced hyperpermeability in cultured human intestinal epithelial monolayers. Crit Care Med 23:1170–1176, 1995.
26. Chavez A, Menconi MJ, Hodin RA, Fink MP: Cytokine-induced epithelial hyperpermeability: role of nitric oxide. Crit Care Med 27:2246–2251, 1999.
27. Yang R, Han X, Uchiyama T, Watkins SK, Yaguchi A, Delude RL, Fink MP: IL-6 is essential for development of gut barrier dysfunction after hemorrhagic shock and resuscitation in mice. Am J Physiol Gastrointest Liver Physiol 285:G621–G629, 2003.
28. Hoffman RA, Zhang G, Nüssler NC, Gleixner SL, Ford HR, Simmons RL, Watkins SC: Constitutive expression of inducible nitric oxide synthase in the mouse ileal mucosa. Am J Physiol 272:G383–G392, 1997.
29. Wu G, Knabe DA, Flynn NE, Yan W, Flynn SP: Arginine degradation in developing porcine enterocytes. Am J Physiol 271:G913–G919, 1996.
30. Morin MJ, Unno N, Hodin RA, Fink MP: Differential expression of inducible nitric oxide synthase messenger RNA along the longitudinal and crypt-villus axes of the intestine in endotoxemic rats. Crit Care Med 26:1258–1264, 1998.
31. Sugi K, Musch MW, Field M, Chang EB: Inhibition of Na+/K+-ATPase by interferon-γ down-regulates intestinal epithelial transport and barrier function. Gastroenterology 120:1393–1403, 2001.
32. Green RM, Beier D, Gollan JL: Regulation of hepatocyte bile salt transporters by endotoxin and inflammatory cytokines in rodents. Gastroenterology 111:193–198, 1996.
33. Abad B, Mesonero JE, Salvador MT, Garcia-Herrera J, Rodriguez-Yoldi MJ: Effect of lipopolysaccharide on small intestinal l -leucine transport in rabbit. Dig Dis Sci 46:1113–1119, 2001.
34. Matthay MA, Folkesson HG, Verkman AS: Salt and water transport across alveolar and distal airway epithelia in the adult lung. Am J Physiol 270:L487–L503, 1996.

Multiple organ dysfunction syndrome; nitric oxide; lipopolysaccharide; intestine; small; colon; ZO-1 protein; ZO-2 protein; ZO-3 protein; occludin protein

©2004The Shock Society