Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Accumulation of inflammatory cells within the airways can be influenced by expression of adhesion molecules on airway epithelium. Intercellular adhesion molecule 1 (ICAM-1; CD54) is a transmembrane glycoprotein of the immunoglobulin supergene family expressed on multiple cell types throughout the body (1). ICAM-1 functions to regulate infiltration of leukocytes within the lung, and its expression on airway epithelial cells has been implicated in the pathogenesis of inflammatory lung diseases such as asthma (2) and chronic bronchitis (3), as well as in hyperoxic lung injury (4) and airway hyperresponsiveness (5). ICAM-1 is expressed constitutively, and its level of expression increases in response to cytokines such as tumor necrosis factor (TNF)-, interferon (IFN)-, or interleukin (IL)-1 (6, 7).

TNF- is a pleiotropic cytokine associated with asthma and airway inflammation (3). It can exert multiple biologic effects on airway epithelium, including enhanced secretion of other cytokines (e.g., IL-6) (8), enhanced secretion of mucin (9), and alterations in lung permeability (10). In addition, TNF- has been shown to increase expression of adhesion molecules, including ICAM-1, on airway epithelial cells (10, 11). In this report, mechanisms associated with TNF--induced enhanced expression of ICAM-1 on epithelial cell surfaces and at the messenger RNA (mRNA) level were investigated. Specifically, the intracellular signaling pathway(s) leading to enhanced ICAM-1 expression in well-differentiated normal human bronchial epithelial (NHBE) cells maintained in primary air-liquid interface culture were examined.

TNF- increased steady-state mRNA levels of ICAM-1 and enhanced surface expression. The results further suggest that this expression occurs via a pathway involving the TNF- 55-kD receptor (TNF-RI), activation of a phospholipase C that hydrolyzes phosphatidylcholine (PC-PLC), leading to production of diacylglycerol (DAG), and activation of protein kinase C (PKC). This pathway appears to mediate the TNF--induced activation of the transcription factor nuclear factor (NF)-B (measured by electrophoretic mobility shift assay [EMSA]) in NHBE cells, as this effect was inhibitable by the PC-PLC inhibitor D609. These results suggest TNF--induced ICAM-1 gene expression may be regulated via binding of NF-B to the consensus binding site in the ICAM-1 gene promoter in NHBE cells in addition to post-transcriptional mechanisms.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Human recombinant TNF-, primary monoclonal antibody mouse antihuman ICAM-1, and anti-TNF-RI and -RII neutralizing monoclonal antibodies were purchased from R&D Systems (Minneapolis, MN). Polyclonal secondary antibody and goat antimouse immunoglobulin (Ig) G-fluorescein isothiocyanate (FITC) were purchased from Boehringer Mannheim (Indianapolis, IN). All- trans retinoic acid and actinomycin D were purchased from Sigma Chemical Co. (St. Louis, MO). Calphostin C and D609 were purchased from Calbiochem (La Jolla, CA). All experiments were done with doses that were nontoxic to NHBE cells as determined by lactate dehydrogenase (LDH) assay.

Methods

Primary culture of human bronchial epithelial cells. Expansion and cryopreservation. One aliquot of primary NHBE cells (catalog no. CC-2540, lot no. 1194; Clonetics, San Diego, CA) was thawed in a 37°C water bath for 3 min and seeded into vented T75 tissue culture flasks (500 cells/cm²) until cells reached 75 to 80% confluency. Cells were maintained at 37°C in an atmosphere of 5% CO₂ and air. The expansion medium used was bronchial epithelial basal medium (Clonetics) containing human recombinant epidermal growth factor (EGF) (25 ng/ml; Intergen, Purchase, NY), 65 ng/ml bovine pituitary extract that was prepared by the method of Bertolero and coworkers (12), 5 × 10⁸ M all- trans retinoic acid, 1.5 µg/ml bovine serum albumin (BSA) (Intergen), 20 IU/ml nystatin (GIBCO BRL, Grand Island, NY), 0.5 µg/ml hydrocortisone, 5 µg/ml insulin, 10 µg/ml transferrin, 0.5 µg/ml epinephrine, 6.5 ng/ml triiodothyronine, 50 µg/ml gentamicin, and 50 µg/ml amphotericin-B (Clonetics). Once confluent, cultures were dissociated with trypsin ethylenediaminetetraacetic acid (EDTA) and frozen as passage 2 according to methods described by Clonetics Corporation. Air-liquid interface culture of NHBE cells. After the expansion, NHBE cells were cultured in an air-liquid interface system similar to that described by Gray and colleagues with several modifications (13). The air-liquid interface culture was initiated by seeding NHBE cells (passage 2, 2 × 10⁴ cells/cm²) on Trans-well-clear culture inserts (24.5 mm, 0.45 mm pore size; Costar, Cambridge, MA) that were thin-coated with rat-tail collagen, type I (Collaborative Research, Bedford, MA). Cells were cultured submerged for the first 5 to 7 d in medium containing a 1:1 mixture of bronchial epithelial cell growth medium (BEGM) (Clonetics):Dulbecco's modified Eagle's medium with high glucose (BEGM:DMEM-H), containing the same concentrations of supplements as described previously with the exception of EGF (0.5 ng/ml). When cultures were 70% confluent (Days 5 to 7), the air-liquid interface was created by removing the apical medium and exposing cells only to medium on their basal surface. The apical surface of the cells was exposed to a humidified 97% air/3% CO₂ environment. Medium beneath the cells (BEGM:DMEM-H) was changed daily thereafter. Cells were cultured for an additional 14 d in air-liquid interface, for a total of 21 d in culture. Morphologic assessment. All samples were prepared so that one dimension was 1 mm or less in thickness before being placed into McDowell's and Trump's 4F:1G primary fixative (14). After 1 h at room temperature, samples were stored at 4°C. Next, samples were rinsed twice in Sorenson's sodium phosphate buffer, pH 7.2 to 7.4, over 30 min and placed in 1% osmium tetroxide in the same buffer for 1 h at room temperature. The samples were then rinsed twice in distilled water before being dehydrated in an ethanolic series, culminating in two changes of 100% acetone. Samples were infiltrated with a 1:1 mixture of acetone and Spurr resin for 30 min, followed by 2 h in 100% resin with two changes. Next, the samples were placed in fresh resin in appropriate molds and polymerized at 70°C from 8 h to 3 d. Semithin (0.25 to 0.5 mm thick) sections were cut with glass knives and stained with toluidine blue O in 1% sodium borate. Ultrathin (70 to 90 nm thick) sections were cut with a diamond knife, stained with methanolic uranyl acetate followed by Reynold's lead citrate, and examined with a transmission electron microscope.

Analysis of ICAM-1 surface expression by flow cytometric analy-sis. NHBE cells were exposed to TNF- over a range of concentrations (0 to 150 ng/ml) or to culture medium (controls) for the specified time periods (0 to 24 h). After exposure, cells were dissociated with 0.5% trypsin and 0.02% EDTA in Hanks' balanced salt solution (HBSS) at 37°C for 10 min. Trypsin was neutralized by the addition of 0.25 mg/ml soybean trypsin inhibitor (Sigma Chemical Co), and cells were rinsed in HBSS and resuspended in phosphate-buffered saline (PBS) 1% BSA (PBS/ BSA) to a final concentration of 5 × 10⁶ cells/ml. All procedures were carried out at 4°C. Each sample, containing approximately 500,000 cells, was incubated with the primary monoclonal antibody mouse antihuman ICAM-1 (1:1,000 final dilution), lightly vortexed, and incubated at 4°C for 45 min. Each sample was then washed, resuspended in PBS/BSA containing the polyclonal secondary antibody goat antimouse IgG-FITC (1:500 final dilution), and incubated at 4°C for 45 min. After this incubation, each sample was then washed, resuspended in PBS/BSA, and immediately analyzed with a FACScan (Becton Dickinson, Mountain View, CA). Three parameter list mode data were collected on 10,000 cells per sample. Mean index of fluorescence was determined on all cells for each sample after debris was excluded by forward angle and side scatter gating.

Measurement of DAG production by thin-layer chromatography (TLC). Total cellular lipids were isolated and extracted by a modification of the method of Bligh and Dyer (15). Briefly, after the appropriate incubation period, media were removed and the cells were scraped in 1 ml of ice-cold PBS. This isolate was added to 3 ml of ice-cold methanol containing 0.5 mg/ml of boric acid in silanized borsilicate glass tubes. Silanization of the tubes and the addition of boric acid were used to help prevent acyl migration of 1,2 DAG (physiologically relevant form) to 1,3 DAG. This step was followed by addition of 3 ml of chloroform and 500 µl of TLC quality water resulting in a final ratio of 1:1:0.5 of methanol:cholorform:water (or PBS) (vol/vol/vol). Samples were centrifuged to separate the phases. The lower organic (chloroform) phase containing the lipids was removed and set aside. A second extraction was performed by addition of 1.5 ml of chloroform to the remaining aqueous (upper) phase, and the centrifugation and separation steps were repeated and the organics were pooled with those from the first extraction. Lipid-containing organics (in chloroform) were dried completely under nitrogen at 25°C in a TurboVap LV evaporator (Zymark, Hopkinton, MA). Dried samples were stored desiccated at 70°C (up to 72 h) before running on TLC plates.

TLC plates were impregnated with 0.25 M boric acid and developed in 100% anhydrous diethyl ether before use. This helped to prevent acyl migration of samples while on the TLC plate and to remove impurities from the silica gel. Standards for cholesterol, 1,2 DAG, and 1,3 DAG were included on each TLC plate. Samples were redissolved in chloroform, and both samples and standards were spotted on the TLC plate under a stream of nitrogen. TLC was performed in a short bed-continuous development (SB/CD) TLC chamber by the method of Welsh and Schmeichel (16). The principle of this type of TLC chamber is that the mobile-phase solvent continuously evaporates off the upper edge of the solvent front. Both the mobile-phase solvent velocity and chromatographic run times can be varied in this system. This allows for continuous solvent migration and optimizes visualization of the lipids of interest. Briefly, the plate was first developed for 2 min in anhydrous diethyl ether in position 1 of the SB/CD chamber, allowing for separation of neutral lipids and phospholipids, as the phospholipids remain at the origin while the neutral lipids, which include DAG, migrate to the ether front. Next, the plate was developed in the same direction in a solvent system consisting of benzene/hexane/diethyl ether/formic acid (65/50/2/ 0.2, vol/vol/vol/vol). The plate was developed in position 3 for 50 to 90 min. In this solvent system, one can effectively separate triglycerides, fatty acids, cholesterol, and DAG from each other. Other neutral lipids remain at the ether front and phospholipids remain at the origin. To visualize the lipids, plates were then sprayed with 10% copper sulfate in phosphoric acid and charred at 170°C. To analyze the plates, an image of the plate was captured using a digital camera and Adobe Photoshop software (Adobe Systems, Inc., San Jose, CA), and then the bands were analyzed densitometrically using NIH Image software on a Macintosh computer. To account for variations in cell numbers between wells and for variations in loading, cholesterol was used as an internal control.

Detection of mRNA by Northern hybridization. Total cellular RNA was collected by the method of Chomczynski and Sacchi (17) using acid guanidinium thiocyanate-phenol-chloroform extraction. The purified RNA pellet was resuspended in 0.5 ml water. A total of 10 µg of total RNA per sample was loaded per lane on a 1% agarose/formaldehyde denaturing gel. The RNA was separated by electrophoresis and transferred to a nitrocellulose filter, cross-linked, and allowed to air-dry overnight. The ICAM-1 complementary DNA probe was generously supplied by Dr. Timothy Springer (Harvard Medical School, Center for Blood Research, Boston, MA). This probe was labeled using a commercially available random priming DNA labeling kit (GIBCO BRL) and 50 µCi of [-³²P]deoxycytidine triphosphate. Hybridization of Northern blots was carried out at 65°C and subsequent washes were done by standard methods (18). Quantity of the hybridized probe was measured (in cpm) using an electronic autoradiography system and associated software (Instant Imager; Packard-Canberra, Rockville, MD). To account for lane variation, samples were normalized to the 28S ribosomal RNA.

EMSA. Nuclear extracts were harvested from NHBE cells grown to confluence in T75 tissue culture flasks with the same media (BEGM:DMEM-H) and supplements as described previously. Cells for the EMSA studies were grown on plastic because of the large numbers of cells required to provide adequate amounts of nuclear proteins. Nuclear proteins were extracted by the method of Dignam and coworkers (19). Harvested nuclear protein concentrations were determined using the Bradford dye-binding procedure (Bio-Rad Protein Assay; Bio-Rad, Richmond, CA), standardized with BSA.

Nuclear extracts were analyzed by EMSA. Oligonucleotides for the NF-B recognition site within the TNF--responsive region of the ICAM-1 promoter (5' AGC TTG GAA ATT CCG GAG 3') were labeled with [-³²P]adenosine triphosphate, using polynucleotide kinase (50,000 cpm/ng). Binding reaction mixtures were incubated at room temperature for 20 min, and contained 0.5 ng DNA probe and 10 µg nuclear extract in 10 mM Tris (pH 7.5) buffer with 50 mM NaCl, 1 mM dithiothreitol, 1 mM EDTA, 5% glycerol, and 1 mg of poly (dI-dC) to inhibit nonspecific interactions of the labeled probe with the nuclear extract proteins. DNA-protein complexes were resolved by electrophoresis through 4% polyacrylamide gels containing 50 mM Tris, 0.38 M glycine, and 2 mM EDTA. Specific competition was accomplished using 20-fold excess of unlabeled probe. In addition, a mutant control probe (NF-B) that did not contain the correct binding sequence was used to demonstrate specificity. Oligonucleotides used in the gel shifts were as follows: ICAM-1 NF-B, 5' AGC TTG GAA ATT CCG GAG 3', and ICAM-1 NF-Bm, 5' AGC TTG tAA ATT aCG GAG 3'. (Sequence motifs within the oligonucleotides are underlined and the mutations are in lower case.) The gel supershift assays were performed as described previously with the exception that subsequent to incubation of oligonucleotide probes with nuclear extracts, 2.0 ml of TransCruz (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) gel supershift antibody was added to the reaction mixture and incubated for 15 to 45 min at room temperature. The gels were dried and autoradiographed with intensifying screens at 70°C. The autoradiograms were examined for differential migration.

Experimental Protocols

Effects of TNF- on ICAM-1 surface expression. Based on results of time and concentration studies (illustrated in Figures 3 and 4), all experiments using flow cytometric analysis of ICAM-1 surface expression involved exposing the cells to TNF- (50 ng/ ml; added apically and basally) for 2 h. At the end of the 2-h exposure period, the cells were washed and placed in fresh medium for an additional 16-h period, at which time they were removed and processed for FACS analysis. In studies using other reagents (e.g., D609, calphostin C, anti-TNF receptor antibodies), cells were preincubated with the specific reagent at the indicated concentration(s) for 15 min, then coincubated with the reagent (or solvent control) plus TNF- (or control medium) for the 2-h exposure period. Values of n are given in figure legends.

View larger version (39K): [in this window] [in a new window]   Figure 3.   ICAM-1 surface expression on NHBE cells in response to TNF-. (A) Concentration response: Cells were exposed to a range of concentrations of TNF- for 2 h, then incubated for an additional 16 h and harvested. TNF- increased ICAM-1 surface expression in a concentration-dependent manner. "Heat inact." refers to 15 ng/ml TNF- incubated at 100°C for 45 min before use. Results are expressed as mean ± SD, n = 6; *significantly different from control, P < 0.05; significantly different from cells treated with 1.5 ng/ml TNF-, P < 0.05; §significantly different from cells treated with 15 ng/ml TNF-, P < 0.05. (B) Time course: Cells were exposed to 50 ng/ml TNF- for time points ranging from 8 to 24 h, then harvested. ICAM-1 surface expression increased with increasing time of exposure. Results are expressed as percent of control fluorescence. All values are mean ± SD, n = 6 for all points. All samples were significantly different from control, P < 0.05; *significantly different from cells treated for 8 h, P < 0.05; significantly different from cells treated for 12 h, P < 0.05; §significantly different from cells treated for 16 h, P < 0.05.

View larger version (50K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 4.   ICAM-1 steady-state mRNA levels in NHBE cells in response to TNF-. (A) Concentration response: Cells were exposed to a range of concentrations of TNF- for 1 h, and RNA was harvested for Northern blot analysis. TNF- increased ICAM-1 mRNA levels in a concentration-dependent manner. (B) Time course: Cells were exposed to 50 ng/ml TNF- for time points ranging from 1 to 24 h, and RNA was harvested for Northern blot analysis. TNF- increased steady-state ICAM-1 mRNA levels at 1 h; however, steady-state mRNA levels returned to control levels within 4 h of exposure. Results are representative of three separate experiments.

Effects of TNF- on DAG production by TLC. NHBE cells were exposed to TNF- (50 ng/ml) for 2 h with pre-incubation and coincubation with the indicated reagents (e.g., D609) at the given concentrations, or with solvent controls. These experiments were repeated three times.

Effects of TNF- on ICAM-1 steady-state mRNA levels by Northern hybridization. NHBE cells were exposed to TNF- at 50 ng/ml (or TNF- plus inhibitors) as described previously for 1, 4, 8, or 24 h. At the end of the time period, the cells were harvested and the RNA isolated. Wells of cells were pooled to generate appropriate amounts of RNA for analysis. Each study was performed at least three times, and the results illustrate a typical experiment.

Effects of TNF- on transcription factor activation by EMSA. For EMSAs, cells were grown in T75 flasks until they reached confluence, and they then were exposed to TNF- at 50 ng/ml (or TNF- plus inhibitors) as described previously for 1 h. Nuclear proteins then were isolated and prepared for EMSA as described. Each study was performed at least three times, and the results illustrate a typical experiment.

Statistical Analysis

Data were analyzed for significance using a one-way analysis of variance with Bonferroni post-test correction for multiple comparisons (20). Data were considered significant at P < 0.05. In addition, all experiments were done with noncytotoxic concentrations of TNF- and other related agents, as determined by LDH release/retention assay (data not shown).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

NHBE cells maintained in the air-liquid interface culture system were well differentiated, containing a heterogenous population of both ciliated and secretory cells remarkably similar to their in situ appearance (Figure 1). As illustrated in Figure 2, NHBE cells expressed a basal, constitutive level of ICAM-1 on their surfaces. Both surface and gene expression of ICAM-1 were enhanced by exposure to TNF- (Figures 3 and 4). Surface expression was maximally expressed between 16 and 24 h of TNF- exposure. In contrast, steady-state mRNA levels of ICAM-1 were maximally expressed between 1 and 2 h of TNF- exposure, with a relatively rapid return to basal expression (within 4 h). Both surface and gene expression in response to TNF- were blocked by the RNA polymerase II inhibitor actinomycin D (Figure 5). As illustrated in Figure 6, TNF--induced ICAM-1 surface expression was attenuated after preincubation and coincubation with a commercially available neutralizing/blocking monoclonal antibody against the TNF-RI, whereas a similar antibody directed against TNF-RII had no effect.

View larger version (142K): [in this window] [in a new window]   View larger version (136K): [in this window] [in a new window]   Figure 1.   Morphologic assessment of NHBE cells maintained in air-liquid interface culture for 21 d. (A) Whole mount of NHBE cells stained with alcian blue (pH 2.5)/periodic acid-Schiff. Secretory cells containing granules of acidic glycoprotein (mucin) appear purple (original magnification: ×400). (B) Transmission electron micrograph of NHBE culture illustrating goblet cells containing mucin granules (hollow arrow) with electron-dense "cores." Notice adjacent cilia with basal bodies (double arrows) and 9 + 2 microtubule doublets in axonemes (single arrow) (original magnification: ×3,500).

View larger version (28K): [in this window] [in a new window]   Figure 2.   Measurement of ICAM-1 surface expression on NHBE cells by flow cytometry. ICAM-1 on cell surfaces is expressed as a logarithmic increase in mean index of fluorescent intensity. (A) 1 = autofluorescence of cells; 2 = nonspecific binding of secondary antibody. (B) 1 = control (untreated) cells (constitutive ICAM-1 expression); 2 = cells treated with 50 ng/ml (429 U/ml) TNF- for 24 h. The x-axis represents log increase in fluorescence, and the y-axis represents relative cell number, as described in text.

View larger version (42K): [in this window] [in a new window]   Figure 5.   Effects of actinomycin D on TNF--induced expression in NHBE cells. Both surface expression (A) and gene expression (B) in response to 50 ng/ml TNF- for 2 h were significantly inhibited by actinomycin D (0.1 µg/ml). Results of surface expression are expressed as percentage of control; n = 6 for each point. *Significantly different than control, P < 0.05; significantly different than TNF- alone, P < 0.05. Results of gene expression are representative of three separate Northern blot experiments.

View larger version (39K): [in this window] [in a new window]   Figure 6.   Effect of antibodies against TNF- receptors on TNF--induced ICAM-1 surface expression on NHBE cells. A neutralizing monoclonal antibody against human soluble TNF receptor I (sTNF-RI [55 kD], neutralizing dose [ND50] = 15 µg/ ml) inhibited TNF- stimulation of ICAM-1 surface expression at 16 h following a 2-h exposure. A similar antibody against soluble TNF receptor II (sTNF-RII [75 kD], ND50 = 15 µg/ml) had no effect. Neither antibody affected ICAM-1 levels in the absence of TNF-. All values are mean ± SD, n = 6 at each point. *Significantly different than control, P < 0.05; significantly different than TNF- alone.

TNF- stimulated formation of DAG in NHBE cells, a response inhibited by the PC-PLC inhibitor D609 (Figure 7). To determine whether activation of PC-PLC was involved in TNF--induced ICAM-1 surface and gene expression, cells were preincubated and coincubated with D609 plus TNF-. As shown in Figure 8, both surface and gene expression of ICAM-1 in response to TNF- were inhibited by D609 (1 to 20 µg/ml). In addition, the PKC inhibitor calphostin C (0.3 and 0.5 µM) also attenuated TNF--induced ICAM-1 surface and gene expression (Figure 9).

View larger version (60K): [in this window] [in a new window]   Figure 7.   Effects of TNF- on DAG formation in NHBE cells, as assessed by TLC. TNF- induced formation of DAG, an effect inhibited by the PC-PLC inhibitor D609 (10 µg/ml). DAG values are normalized to their corresponding cholesterol values and then expressed as percentage of unpaired controls. Results are expressed as percentage of control. All values are mean ± SD, n = 6 for all points. *Significantly different than control, P < 0.05; significantly different than TNF- alone.

View larger version (56K): [in this window] [in a new window]   Figure 8.   Effects of the PC-PLC inhibitor D609 on TNF--induced ICAM-1 expression in NHBE cells. D609, in a concentration-dependent manner, attenuated TNF- enhancement of (A) surface expression of ICAM-1 and (B) steady-state mRNA levels of ICAM-1. For surface expression (A), all values are mean ± SD, n = 6 for all points. *Significantly different than control, P < 0.05; significantly different than TNF- alone. For gene expression (B), results are representative of three separate experiments. D609 did not affect control levels of ICAM-1 surface or gene expression.

View larger version (50K): [in this window] [in a new window]   Figure 9.   Effect of PKC inhibitor calphostin C on TNF--induced ICAM-1 expression in NHBE cells. The PKC inhibitor calphostin C (Cal-C) inhibited TNF- enhancement of both surface (A) and gene (B) expression of ICAM-1. For surface expression (A), all values are mean ± SD, n = 6 for all points. *Significantly different than control, P < 0.05; significantly different than TNF- alone; §significantly different from cells treated with TNF- + 0.3 µM Cal C for 16 h, P < 0.05. For gene expression assessed by Northern hybridization (B), results are representative of three separate experiments. Cal-C did not affect control levels of ICAM-1 surface or gene expression.

Because transcriptional regulation may contribute to TNF--induced ICAM-1 expression, we investigated the binding of transcription factors to the TNF-responsive region within the ICAM-1 promoter by EMSA. TNF- stimulated nuclear protein binding to the ICAM-1 NF-B binding site (5' AGC TTG GAA ATT CCG GAG 3') located within the TNF-responsive region of the promoter. In addition, when the shifted complexes were assessed by EMSA supershift assays using antibodies against known NF-B proteins, they were found to contain p65 and/or c-rel (Figure 10A). To determine whether PC-PLC was involved in the regulation of these complexes, cells stimulated with TNF- were preincubated and coincubated with D609. As illustrated in Figure 10B, the PC-PLC inhibitor D609 inhibited activation of the NF-B complex.

View larger version (106K): [in this window] [in a new window]   View larger version (94K): [in this window] [in a new window]   Figure 10.   Effects of TNF- on activation of NF-B complexes in NHBE cells as assessed by EMSA. (A) Effects of TNF-: lane 1 = free probe; lane 2 = unstimulated; lane 3 = TNF- (50 ng/ml); lane 4 = TNF- (50 ng/ml) + 20 × cold NF-B probe; lane 5 = TNF- (50 ng/ml) + 20 × cold mutant NF-B probe; lane 6 = TNF- (50 ng/ml) + p65 antibody; lane 7 = TNF- (50 ng/ml) + c-rel antibody; lane 8 = TNF- (50 ng/ml) + p50 antibody; lane 9 = TNF- (50 ng/ml) + p52 antibody. TNF- activated NF-B in these cells. (B) Effects of D609 on TNF--induced NF-B activation: lane 1 = free probe; lane 2 = unstimulated; lane 3 = TNF- (50 ng/ml); lane 4 = TNF- (50 ng/ml) + 20 × cold NF-B probe; lane 5 = TNF- (50 ng/ml) + 20 × cold mutant NF-B probe; lane 6 = D609 (10 µg/ml) alone; lane 7 = TNF- (50 ng/ml) + D609 (10 µg/ml). D609 inhibited TNF- activation of NF-B, implicating PC-PLC in the response. The band in lane 4 becomes apparent with longer length of exposure.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In diseases such as asthma, chronic bronchitis, and acute respiratory distress syndrome, adhesion molecules on the surface of the airway epithelium may play a major role in recruitment of leukocytes and their subsequent retention in the airways. Adhesion molecule expression can be upregulated by a number of cytokines, such as IFN- (7) or TNF- (21), but little is known about the intracellular signaling mechanisms involved in these responses. In these studies, we investigated potential signal transduction pathways involved in TNF--regulated ICAM-1 surface and gene expression in well differentiated primary cultures of human airway epithelial cells. The results indicate that these cells express basal levels of ICAM-1 message and surface protein and that expression at both the mRNA and protein levels is increased by pathophysiologically relevant concentrations of TNF- (22). Transcriptional control may contribute to the regulation of TNF--enhanced ICAM-1 surface and gene expression, as the RNA polymerase II inhibitor actinomycin D blocked both responses. However, several lines of evidence suggest that a number of post-transcriptional events probably are involved in regulating ICAM-1 expression. First, increases in surface expression were not apparent until 16 h after exposure to TNF-, whereas, as illustrated in Figure 4B, increases in ICAM-1 steady-state mRNA in response to TNF- were very short-lived. Second, as illustrated in Figure 8, ICAM-1 surface expression returned to control levels after D609 treatment at the highest concentration (10 µg/ml), whereas the message still was 5 to 6-fold higher than control levels. These data suggest that regulation of ICAM-1 expression by TNF- involves additional post-transcriptional and/or translational regulation.

Even with this apparent post-transcriptional regulation of ICAM-1 expression, it is unlikely that post-transcriptional regulation alone accounted for the low level of ICAM-1 message observed in actinomycin D-treated cells. The actinomycin D studies were performed with actinomycin D present only during exposure of the cells to TNF-. Thus, before addition of actinomycin D and TNF-, both control and treated cells would have the same level of ICAM-1 mRNA. Therefore, if TNF- were to affect transcription of ICAM-1, the actinomycin D treatment would block increases in ICAM-1 steady-state mRNA, which it did (as illustrated in Figure 5B). If, on the other hand, TNF- were to enhance ICAM-1 message via increases in steady-state mRNA levels, actinomycin D would not block the increase in ICAM-1 steady-state mRNA in response to TNF-. While there is still the slight possibility that mRNA stability could require transcription of an RNA stability factor that would also not be produced in the presence of actinomycin D, the results, taken together, suggest both transcriptional and post-transcriptional regulation of ICAM-1 expression in response to TNF-.

Relatedly, the concentration of actinomycin D used in these studies was lower than that reported by others, but most of those reports refer to studies done with cell lines, not primary cells as used in our studies. In preliminary experiments, the concentration of actinomycin D used herein (0.1 µg/ml) was the highest concentration that did not generate a cytotoxic response in NHBE cells (data not shown). While we are not certain that this concentration of actinomycin D blocked all the RNA synthesis in these cells, use of higher concentrations that may cause cytotoxicity was precluded.

TNF-RI appears to be involved in the ICAM-1 response in NHBE cells because TNF- stimulation of ICAM-1 surface expression was inhibited by preincubation of TNF- with (human recombinant) soluble TNF-RI (23). It is known that TNF- exerts many of its biologic effects on cells by binding to one of two different transmembrane receptors, TNF-RI or TNF-RII. These receptors are multifunctional and, after binding TNF-, can activate multiple genes through various intermediates such as protein kinases, protein phosphatases, phospholipases, proteases, sphingomyelinases, and transcription factors (24). Intra-cellular signaling appears to occur when a TNF trimer binds two or three receptors in an extracellular complex that permits aggregation and activation of the cytoplasmic domains (25). Additional reports have suggested that upon activation, receptors can activate different kinases and phosphatases via recruitment of different cytoplasmic adaptor proteins, which bind to these receptor complexes during activation (26). TNF-RI, the receptor apparently involved here, is associated with activation of protein kinases, sphingomyelinases, and phospholipases in many different cell types (27). Although the precise signaling pathway whereby TNF-RI is involved in enhanced ICAM-1 expression in NHBE cells is not known, the pathway downstream from the receptor appears to involve activation of PC-PLC.

TNF- has been shown to directly activate PC-PLC in a number of different cell types, including the human monocytic cell line U-937 (30) and murine 70Z/3 pre B cells (31). Recent studies from this laboratory have shown that TNF- activation of PC-PLC is involved integrally in the stimulatory effect of TNF- on mucin secretion by rodent airway epithelium (9). PC-PLC catalyzes hydrolysis of phosphatidylcholine within cell membranes to form two products: phosphocholine and DAG. Because, as illustrated in Figure 7, D609, the PC-PLC inhibitor, blocked formation of DAG in NHBE cells in response to TNF-, PC-PLC hydrolysis of phosphatidylcholine is implicated in this response. Additional evidence for a role for PC-PLC in the TNF--induced ICAM-1 response is provided by studies in which NHBE cells were radioactively labeled with [¹⁴C]mytistic acid, which preferentially labels the phosphatidylcholine lipid pool within cell membranes (16, 32) before exposure to TNF-. The DAG generated after TNF- exposure was also labeled with [¹⁴C]mytistic acid, indicating it was generated from PC-PLC hydrolysis of phosphatidylcholine (33). Thus, PC-PLC appears to be activated in NHBE cells in response to TNF-, and to play an important role in the ensuing enhanced expression of ICAM-1.

DAG, the product of PC-PLC hydrolysis of phosphatidylcholine, is a potent signaling molecule that in turn activates several intracellular pathways, the most prominent being PKC. PKC has been implicated previously in the actions of TNF- in a number of cell types (9, 34). In these studies, PKC appears to be involved in the signaling pathway by which TNF- induces ICAM-1 surface and gene expression. The PKC inhibitor calphostin C, which binds to the DAG binding site on PKC to inactivate it, attenuated both enhanced surface and gene expression of ICAM-1 in response to TNF- (Figure 9). The stimulatory effect of TNF- on ICAM-1 surface expression was also mimicked by the phorbol ester phorbol-12-myristate-13-acetate, which causes persistent stimulation of PKC in intact cells (37) (data not shown). Thus, the results suggest that PC-PLC is activating PKC, which, in turn, is acting to enhance ICAM-1 expression. Interestingly, this signaling pathway is quite similar to that by which TNF- stimulates mucin secretion in differentiated rodent airway epithelial cells (9). This response appears to be limited to stimulation by TNF-, as there is no evidence (to our knowledge) that other stimuli that upregulate expression of ICAM-1 in airway epithelial cells, such as IFN- or IL-1, activate PC-PLC or PKC. Activation of PC-PLC and PKC, followed by phosphorylation of a number of intracellular targets, could be a common mechanism by which TNF- exerts many of its effects on airway epithelium and, perhaps, other cell types.

PKC enzymes are composed of at least 11 isoforms that differ in structure and enzymatic properties (38). We did not attempt to elucidate which isoform(s) of PKC were involved in the ICAM-1 response to TNF- in these studies. However, because DAG activated by PC-PLC does not change cellular calcium levels, the PKC isoforms involved in this pathway would be the calcium-independent, novel, or atypical PKC isoforms (i.e., PKC ,  , , , , , ) rather than the classic calcium-dependent PKC isoforms, such as PKC , I, II, and  (39).

A common therapy for airway inflammation in respiratory diseases such as allergic asthma is administration of inhaled corticosteroids. One of the suggested molecular mechanisms by which corticosteroids exert their beneficial effects is through inhibition of adhesion molecules (40). In previous studies, we have reported that corticosteroids can attenuate TNF--induced ICAM-1 surface and gene expression in NHBE cells (23). Corticosteroids have been shown to alter transcriptional regulation of genes either by inhibiting the binding of the transcription factor NF-B to DNA or by inhibiting activation of several other transcription factors that are known to regulate inflammation, such as activator protein 1 (AP-1) (41, 42). NF-B has been implicated in regulation of expression of adhesion molecules in several cell types (43, 44), and sequence analysis of the 5' regulatory region of the ICAM-1 gene has revealed several transcription factor consensus binding sites, including NF-B and AP-1 sites (45). The human endothelial cell ICAM-1 promoter contains a TNF--responsive region consisting of a 92-base pair sequence. This TNF--responsive region contains multiple transcription factor consensus binding sites, including an NF-B site that, when mutated, completely abolishes activation of the ICAM-1 promoter (46).

We investigated the effects of TNF- on this NF-B site on the ICAM-1 promoter in NHBE cells. In EMSAs using nuclear protein extracts from TNF--stimulated cells, in vitro binding of p65 and c-rel to the NF-B consensus binding sites found in the ICAM-1 promoter was observed. Because binding of these protein complexes to this site in response to TNF- was inhibited by D609 (Figure 9B), it would appear that activation of NF-B also was dependent on upstream activation of PC-PLC by TNF-.

In summary, differentiated NHBE cells express basal levels of ICAM-1 on their surfaces. TNF- stimulates both surface and gene expression of ICAM-1 in these cells. The following signaling pathway appears to be involved in the enhanced expression of ICAM-1 in response to TNF-: TNF-  TNF-RI  PC-PLC  DAG  PKC  (NF-B?)  ICAM-1 mRNA  ICAM-1 surface expression.