Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The airway epithelium functions as a barrier that protects the underlying tissue against allergens, irritants, viruses, and microbial pathogens (1). The effectiveness of this barrier relies on epithelial integrity and this is dependent on intercellular adhesion. In airways disease such as asthma, the epithelial barrier is significantly damaged; inflammation, cell shedding, and increased paracellular permeability are common features (4). Increasing evidence suggests that viral infections may play an important role in the development of asthma and also the initiation of acute asthma attacks (7). Mechanisms for viral initiation and provocation of asthma are not fully understood, but it is likely that epithelial cell damage, enhanced production of proinflammatory cytokines, and increased inflammatory cell recruitment are contributory factors (7). Rhinovirus, in particular, along with respiratory syncytial virus (RSV) and parainfluenza virus are most frequently associated with virus-induced asthma exacerbations (7). Moreover, the damaged airway epithelium in asthmatics might predispose to infection because the barrier is compromised.

Intercellular adhesion of epithelial cells is regulated by cell junctions located on the basolateral surface of the cell. The tight junction (TJ), most apical of the junctions in the junctional complex, forms a continuous permeability barrier regulating flux of solutes through the paracellular space and creates a boundary between the apical and basolateral membrane domains that maintains cell polarity (11, 12). Immediately below the TJ is the cadherin-rich adherens junction. Cadherins are a family of Ca²⁺-dependent adhesion molecules essential for the induction and maintenance of intercellular contacts (13, 14). E-cadherin is the major cadherin expressed in epithelial cells, including those of the airway. Intercellular adhesion is achieved when E-cadherin dimers bind homotypically to equivalent dimers on adjacent cells. The cytoplasmic domain of E-cadherin binds directly to - or -catenin, which then associates with -catenin. The complex is anchored to the actin cytoskeleton by linkage with several of the actin-binding proteins, and this is vital for full adhesive function (15, 16). The essential function of E-cadherin in the junctional complex has been demonstrated in Madin-Darby canine kidney (MDCK) epithelial cells using E-cadherin function- blocking antibodies or Fab fragments (17). However, disruption of established junctions was achieved only by depletion of extracellular Ca²⁺ (18) or by exposure of the basolateral epithelial surface to E-cadherin function-blocking antibodies (21). Both of these methods cause functional loss of the epithelial barrier, disorganization of junctional proteins, and enhanced paracellular permeability.

In our studies we used a specific function-blocking antibody to E-cadherin that disrupts the TJ permeability barrier. Using replication-incompetent adenovirus that contains a lacZ reporter gene, we show that infection increases dramatically, concomitant with loss of epithelial integrity in both confluent 16HBE14o- epithelia and primary human bronchial epithelial (HBE) cell cultures.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Antibodies

Mouse antihuman E-cadherin antibodies SHE78-7 and HECD-1 were purchased from BioWhittaker (Wokingham, Berkshire, UK). Mouse immunoglobulin (IgG2a) was purchased from ICN Biomedicals Ltd. (Thame, Oxfordshire, UK). These antibodies were diluted in defined media (described later) for function-blocking studies. For immunoblotting, sheep antimouse Ig secondary antibody linked to horseradish peroxidase (HRP) was purchased from Amersham Pharmacia Biotech (Little Chalfont, Buckinghamshire, UK).

Cell Culture

16HBE14o cells, an HBE cell line (obtained from D. Gruenert, University of California-San Francisco, San Francisco, CA) were maintained in growth medium consisting of Eagle's minimum essential medium (MEM) with Earle's salts (Hyclone Europe, Crammington, Northumberland, UK) supplemented with 5% fetal calf serum (FCS) (Life Technologies Ltd., Paisley, UK), 2 mM glutamine, and antibiotics. After trypsinization, 16HBE14o cells were seeded into six-well plates (Corning Costar, High Wycombe, Buckinghamshire, UK) at 10⁵ cells/well or 96-well microplates (Nunc brand products; Nalge Nunc Int., Hereford, UK) at 1 × 10⁴ cells/well, or on 6.5 mm Transwell Clear membranes (Corning Costar) at 4 × 10⁴ cells/membrane. Transwell Clear membranes were previously coated with MEM with Earle's salts containing 1 mg/ml fatty acid-free bovine serum albumin (BSA) (Sigma Chemical Co., Poole, Dorset, UK), 10 µg/ml fibronectin (Sigma), and 1% Vitrogen 100 (Imperial Laboratories, Andover, Hampshire, UK). Cells grown on membranes were cultured for 7 to 10 d in 200 µl of growth medium in the apical compartment and 500 µl in the basal compartment. Culture medium was changed every 2 to 3 d. When confluent, cell cultures were incubated with defined media with or without antibody, 24 h before assay. Defined media consisted of MEM with Earle's salts containing 1 mg/ml fatty acid-free BSA (Sigma), 5 µg/ml transferrin (Sigma), 0.1 µM selenium ions (Sigma), and no FCS.

Human primary airway epithelial cells were isolated and cultured by procedures previously described, with minor modifications (22). Epithelial cells were removed from normal airway by digestion in 1 mg/ml pronase E (Sigma) for 24 h at 4°C. Cells were scraped (larger airways) or flushed (small airways) into Dulbecco's modified Eagle's medium (DMEM) (Life Technologies) containing 10% FCS. After centrifugation at 1,000 rpm, the cell pellet was resuspended in Clonetics bronchial epithelial cell growth medium (BEGM) supplied by BioWhittaker and seeded on 6.5-mm precoated Transwell Clear membranes (described earlier) at 4 × 10⁴ cells/membrane. Cells were subsequently cultured in growth media; a 1:1 mixture of BEGM/DMEM, 200 µl in the apical compartment and 500 µl in the basal compartment. Media was changed every 2 to 3 d. When cells became confluent and high resistances had developed, growth media were replaced with bronchial epithelial cell basal medium (BioWhittaker) 24 h before assay.

Disruption of E-Cadherin Function

After 4 to 14 d of seeding, 16HBE14o- and primary HBE cells grown on Transwell membranes and 96-well plates were confluent. To disrupt E-cadherin function, cultures were exposed to anti-E-cadherin antibody SHE78-7 or subjected to low Ca²⁺.

Antibody method. Defined medium (200 µl) containing either mouse control IgG2a or SHE78-7 (0.3 to 4 µg/ml) was added only to the apical compartment of 16HBE14o- cells cultured on Transwell membranes. Defined medium (300 µl) was added to the basal compartment. Primary cells were exposed to both apical and basal antibody (4 µg/ml). Cells cultured on microplates were incubated with 100 µl defined medium containing either IgG2a or SHE78-7. Antibody incubations were carried out for 24 h and antibodies remained in the media during the infection period.

Ca²⁺ switch method. Culture media were removed from the apical and basal compartments of 16HBE14o- cultures. Cells were washed three times with phosphate-buffered saline (PBS) and replaced with S-MEM (Life Technologies) containing 20 µM CaCl₂. Cultures were incubated in low Ca²⁺ for 1 h, after which CaCl₂ was added to normalize the Ca²⁺ concentration to 1.8 mM and cells were exposed to virus.

Measurement of Transepithelial Resistance

Transepithelial electrical resistance (TER) of cells grown on Transwell membranes was monitored using a Millipore Minicell ERS meter. When resistances were 300 .cm² or greater, cultures were placed overnight in defined medium in the presence or absence of antibodies for 24 h or low Ca²⁺ for 1 h (described earlier). TERs were measured throughout the experiment, before and after adenoviral infection.

Permeability Measurements

Permeability studies were carried out using the method of Wong and Gumbiner (23) with minor modification. Briefly, cultures incubated with antibodies as described earlier were used to study the flux of mannitol (molecular weight [mol wt], 184), inulin (mol wt 5,200), and fluorescein isothiocyanate (FITC)-labeled dextran (Sigma; mol wt 77,000). Final concentrations of unlabeled tracers, 5 mM mannitol and 1 mM inulin, were added to both apical and basal compartments in 0.1 and 0.7 ml, respectively. In addition, 1- and 2-µM final concentrations of [¹⁴C]mannitol and [³H]inulin, respectively, were added to the apical compartment and cultures were allowed to equilibrate at 37°C for 30 min. The Transwell filters were then transferred to a new 24-well plate containing 0.7 ml defined medium in each well, and after 1 h the entire basal medium from each well was analyzed in a counter. Flux experiments using fluoresceinated dextran were performed in phenol red-free defined medium, where 50 µg of dextran were added to the apical compartment of cells grown on Transwells. At the end of the flux period, FITC fluorescence of basal medium samples was measured in a Wallac plate reader.

Western Blot Analysis

Confluent cultures of 16HBE14o- cells grown on six-well plates were washed twice with PBS containing protease inihibitors (complete protease inhibitor tablets; Boehringer Mannheim, Lewes, East Sussex, UK). The cells were lysed in sodium dodecyl sulfate (SDS) sample buffer (Novex, Gebäude, Frankfurt, Germany) containing 5% dithiothreitol, then boiled for 4 min. Samples were separated on 6% SDS-polyacrylamide gel electrophoresis (PAGE) gels under reducing conditions and transferred onto nitrocellulose membranes (Novex). Membranes were blocked in 5% low-fat dried milk in TBS-T (20 mM Tris [pH 7.6], 137 mM NaCl, and 0.1% Tween 20) overnight at 4°C, then incubated with 2 µg/ml SHE78-7 for 1 h at room temperature. This was followed by incubation with antimouse Ig secondary antibody linked to HRP for 1 h. Immunoreactive bands were detected on Hyperfilm ECL film (Amersham Pharmacia Biotech) after incubation of the blot with chemiluminescent Supersignal CL-HRP substrate (Pierce Warriner UK Ltd, Chester, UK).

Adenovirus Propagation and Purification

Recombinant replication-incompetent type 5 adenovirus, RAd35, containing cytomegalovirus-driven -galactosidase (lacZ) reporter gene was a kind gift from Gavin Wilkinson (University of Wales College of Medicine, Cardiff, UK) (24).

Propagation and CsCl density gradient purification procedure of RAd35 has been previously described (25). Briefly, RAd35 was cultivated in A293 cells grown in DMEM containing 2% FCS, 2 mM glutamine, and antibiotics. Infected cells were harvested and subjected to five freeze-thaw cycles by alternating between 70°C and 37°C. The lysed cellular material was removed by centrifugation at 5,000 rpm for 15 min. The supernatant was overlayed onto a double-layered cushion of CsCl densities of 1.25 and 1.4 g/ml and spun in a Beckman SW40Ti rotor at 35,000 rpm for 1 h at 15°C. The virus band at the 1.25 and 1.4 g/ml CsCl interface was collected, resuspended in 1.35 g/ml CsCl, and centrifuged in a SW40Ti rotor at 40,000 rpm for 16 h at 15°C. The visible virus band at ~ 1.35 g/ml was recovered and dialyzed against 16HBE14o- defined media containing 10% glycerol.

Determination of the 50% tissue culture infective dose per milliliter was carried out on HER911 cells. Briefly, cells suspended in DMEM containing 2% FCS, 1 mM glutamine, and antibiotics were seeded on a 96-well plate at 1 × 10⁴ cells/well. Serial dilutions of RAd35 were added to the subconfluent cells the following day. At 7 d later the cell sheet in each well was checked for cytopathic effect and infectious units per milliliter were calculated (26).

Adenovirus Infection

After incubation of cell cultures with either control IgG2a or SHE78-7 for 24 h, an aliquot of virus (2 to 5 µl) was added to 100 µl of the existing media in the apical compartment of Transwell membranes (or 50 µl of existing media in microplate wells). Cells were exposed to a final RAd35 titer of 1.58 × 10¹⁰ infectious units/100 µl for 3 to 24 h unless otherwise stated. After viral exposure, cultures were washed in PBS and infection was assessed by evaluating reporter gene expression, either by measurement of -galactosidase activity or by staining with 5-bromo-4-chloro-3-indolyl--D-galactoside (X-Gal), as described later.

For the Ca²⁺ switch assay, at the end of the low-Ca²⁺ incubation for 1 h, CaCl₂ was added to normalize the Ca²⁺ concentration to 1.8 mM and an aliquot of virus (3.95 × 10⁹ infectious units) was added to 50 µl of existing media for 1 h. After viral exposure for 1 h, virus was removed from the wells and cultures were washed in PBS. Culture media were added and cells were incubated for a further 5 h. Cells were then washed and lysed, and infection was assessed by measurement of -galactosidase activity. For the comparative experiment, cells treated with SHE78-7 (for 24 h) and exposed to virus for 1 h were processed in a similar manner.

-Galactosidase Expression

Measurement of -galactosidase activity. -Galactosidase activity was measured using a commercially available assay kit (Promega, Oxford, Oxfordshire, UK). After viral infection, cell cultures were washed in PBS and solubilized in 50 µl of reporter lysis buffer (Promega) for 15 min at 4°C. Cell extracts were spun at 13,000 rpm for 3 min at 4°C. The supernatants were assayed in a microplate for -galactosidase activity. Absorbances were read on a Molecular Devices spectrophotometer (Crawley, West Sussex, UK). -Galactosidase activity was measured as units per milligram of protein, where 1 U of enzyme will hydrolyse 1 µmol of o-nitrophenyl--D-galactopyranoside per min at pH 7.3 and 37°C. Protein concentrations were determined using the Bio-Rad DC-protein assay kit (Bio-Rad, Hemel Hempstead, Hertfordshire, UK).

X-Gal staining. After viral infection, cells were washed in PBS and fixed with 2% formaldehyde and 0.2% glutaraldehyde for 10 min, rinsed twice in PBS for 5 min, and stained with X-Gal solution consisting of PBS, 1 mg/ml X-Gal (Boehringher Mannheim), 2 mM MgCl₂, 5 mM K₃Fe(CN)₆, and 5mM K₄Fe(CN)₆, overnight at 37°C as described previously (27).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

E-Cadherin Expression in 16HBE14o- Cells

Because 16HBE14o- cells form functional TJs, we investigated E-cadherin expression in these cells. Cell extracts were separated by SDS-PAGE, blotted onto nitrocellulose membranes, and then probed with anti-E-cadherin antibody SHE78-7 (2 µg/ml). The single strong immunoreactive band detected at ~ 120 kD corresponds to full length E-cadherin (Figure 1). Presence of E-cadherin in 16HBE14o- cells was confirmed by a second monoclonal antibody, HECD-1 (data not shown). No signal was obtained on the blot probed with the control IgG2a antibody (data not shown).

View larger version (35K): [in this window] [in a new window]   Figure 1.   Western blot analysis of E-cadherin in 16HBE14o- cells. Total cell lysate of confluent cells (40 µg of protein) was separated by SDS-PAGE. After transfer onto nitrocellulose membrane the blot was probed with 2 µg/ml SHE78-7. Full-length E-cadherin (~ 120 kD) is indicated by the arrow.

E-Cadherin Function Is Essential for 16HBE14o Intercellular Adhesion

The role of E-cadherin in 16HBE14o- intercellular adhesion was investigated by exposure to specific function-blocking antibody SHE78-7. Incubation of isotype control IgG2a with nonconfluent 16HBE14o- cultures for 24 h had no effect on colony morphology (Figure 2A). In contrast, the addition of 3 µg/ml of SHE78-7 to growing cultures for 24 h caused colony disruption (Figure 2B) where complete loss of all intercellular contacts was observed in some areas. A similar disruption was seen after exposure of 16HBE14o- cells to 10 µg/ml HECD-1, another E-cadherin specific function- blocking antibody (data not shown). Removal and washout of SHE78-7 resulted in re-formation of the cell colony within 24 h and also cell growth (data not shown).

View larger version (93K): [in this window] [in a new window]   Figure 2.   Effect of SHE78-7 on intercellular contacts of subconfluent 16HBE14o- cells. Cells were incubated with 3 µg/ml of either (A) IgG2a or (B) SHE78-7 for 24 h. IgG2a had no effect on colony morphology, whereas SHE78-7 caused disruption of the colony with extensive loss of intercellular adhesion.

Effect of Anti-E-Cadherin Antibody on TJ Function of Confluent Epithelia

SHE78-7 effectively disrupted cell contacts in subconfluent cell cultures (Figure 2). To determine whether SHE78-7 could perturb intercellular adhesion in confluent cell cultures where TJs had formed, studies were carried out on confluent 16HBE14o cells. Cells grown on Transwell membranes were exposed to increasing concentrations of either control IgG2a or SHE78-7. The TER was monitored as a measure of TJ function. In 24 h, SHE78-7 (0.3 to 4 µg/ml) caused a concentration-related decline in TER (Figure 3A) with a maximum fall of 93%. In the presence of 2.5 and 4 µg/ml SHE78-7, TERs were close to the resistance of the membrane alone, indicating severe loss of epithelial integrity. The time-dependent effect of SHE78-7 (4 µg/ml) on TER is shown in Figure 3B, where SHE78-7 caused a fall in TER with the resistances only 15 and 8% of their initial value by 14 and 24 h, respectively. In these confluent cultures, incubation with SHE78-7 still induced marked morphologic changes; the cells had a more rounded appearance as intercellular contacts were weakened or lost. Incubation of the cells with propidium iodide showed no alteration in cell viability on exposure to IgG2a or SHE78-7 for 24 h. After removal and washout of SHE78-7, the cells regained resistances of > 300 .cm² in 24 h (data not shown). Control IgG2a-treated cell cultures had no effect on TERs, which were maintained between 350 and 380 .cm².

View larger version (14K): [in this window] [in a new window]   Figure 3.   Effect of SHE78-7 on TER of confluent polarized 16HBE14o- epithelia. Control IgG2a (open circles), SHE78-7 (filled circles), and membrane resistance (open squares). (A) Dose-dependent effect of SHE78-7. TERs were recorded 24 h after addition of 0.3 to 4 µg/ml of either IgG2a or SHE78-7. (B) Time-dependent effect of SHE78-7. TERs were recorded over a 24-h period after addition of 4 µg/ml of either IgG2a or SHE78-7. SHE78-7 effectively decreased TERs, with maximal effect at 24 h. In contrast, IgG2a had no effect on TER. Values are means ± SEM of three to five experiments. Membrane resistances of 46 .cm2 (A) and 48 .cm2 (B) were recorded. This value was taken before addition of cells to the Transwell membrane and is the mean of four to eight wells.

E-Cadherin Antibody Increases Paracellular Permeability of 16HBE14o- Epithelia

To ensure that SHE78-7-mediated reduction in TER was a good indicator of increased paracellular permeability rather than activation of transcellular ion transport, we assessed the paracellular flux of membrane-impermeant tracers across 16HBE14o- epithelia. Hydrophilic tracer molecules [¹⁴C]mannitol (mol wt, 184), [³H]inulin (mol wt 5,200), and FITC-labeled dextran (mol wt 77,000) were used. Dextran, the largest tracer of the three, would give a better indication of paracellular transport of macromolecules. In these experiments, SHE78-7 caused a concentration-related decline in TER with 94% reduction in the presence of 4 µg/ml. In the same cultures, SHE78-7 caused concentration-related increases in paracellular flux of mannitol, inulin, and dextran (Table 1). Thus, decrease in TER caused by SHE78-7 was associated with an increase in paracellular permeability of the TJ. The maximal rates of mannitol, inulin, and dextran transport occurred in the presence of 4 µg/ml SHE78-7 and were 166, 38, and 0.01 nmol/cm²/h, respectively, indicating that flux rates decreased with increasing size of tracer.

RAd35 Infection of 16HBE14o- Epithelia

To investigate whether perturbation of TJ function of 16HBE14o- epithelia results in increased susceptibility to adenoviral infection, Transwell cultures were treated with 4 µg/ml of either control IgG2a or SHE78-7 for 24 h, then incubated for a further 6 h with replication-incompetent RAd35 (2.1 × 10¹⁰ infectious units/100 µl). TERs after 24 h incubation with IgG2a or SHE78-7 were 305 ± 25 and 62 ± 3 .cm² (n = 3, ± standard deviation [SD]), respectively, and were not affected by incubation with virus. Infected cells were identified by staining with X-Gal, where -galactosidase-expressing cells stained blue. Figure 4 shows a substantial increase in adenovirus infection of SHE78-7-treated cell cultures, where ~ 90% of cells were stained blue. In contrast, a very low level of infection was detected in cultures exposed to control IgG2a, where epithelial integrity was unaffected.

View larger version (31K): [in this window] [in a new window]   Figure 4.   Effect of SHE78-7 on RAd35 infectivity of 16HBE14o- epithelia. Cells cultured on Transwell membranes were incubated with 4 µg/ml of either control IgG2a or SHE78-7 for 24 h, then exposed to RAd35 (2.1 × 1010 infectious units/100 µl). Infection was assessed 6 h later by staining with X-Gal for transgene expression. SHE78-7-treated cells stained intensely blue, indicating that disruption of E-cadherin function of 16HBE14o epithelia caused a marked increase in adenoviral infection.

To quantitate changes of infection associated with SHE78-7 treatment of 16HBE14o- cells, we exposed cells that had been incubated with either 4 µg/ml control IgG2a or 0.3 to 4 µg/ml SHE78-7 to RAd35 for 6 h and measured the -galactosidase activity. The dose-response effect of SHE78-7 on adenovirus infection is shown in Figure 5A. The data are also plotted as fold increase in infection (Figure 5B; data are means of three independent experiments ± standard error of the mean [SEM]) and show that for cultures incubated with 4 µg/ml SHE78-7, the infection was 30-fold greater than in controls.

View larger version (13K): [in this window] [in a new window]   Figure 5.   SHE78-7 dose-dependent effect on adenoviral infection of 16HBE14o- epithelia. Cells cultured on Transwell membranes were incubated with either 4 µg/ml control IgG2a or 0.3 to 4 µg/ml SHE78-7 for 24 h, then exposed to adenovirus (1.58 × 1010 infectious units/100 µl) for 6 h. SHE78-7 caused a concentration-related increase in adenoviral infection of 16HBE14o- epithelia (A). Values are means ± SD of triplicate determinations. Fold increases of RAd35 infection are shown in (B). Values are means ± SEM of three experiments.

Comparison of SHE78-7 with Low-Ca²⁺ Treatment on Infection

To address whether the enhanced infection caused by SHE78-7 treatment for 24 h is simply due to disruption of the epithelial barrier, we compared this method of perturbing junction function with lowering extracellular Ca²⁺ to 20 µM. Low Ca²⁺ for only 1 h resulted in TERs that were similar to SHE78-7 treatment. Moreover, this method should not allow for expression of new proteins that may aid viral infection. Exposure of treated cells to virus (3.95 × 10⁹ infectious units/ 50 µl) was limited to 1 h. Assay of reporter gene expression showed that infection of cultures treated with SHE78-7 and low Ca²⁺ were increased by 23- and 7-fold, respectively (Figure 6). Therefore, SHE78-7-facilitated infection was ~ 3-fold greater than infection enhanced by low Ca²⁺.

View larger version (36K): [in this window] [in a new window]   Figure 6.   Comparison of SHE78-7 with low-Ca2+ treatment on infection of 16HBE14o- epithelia. Cells cultured on membranes were incubated with SHE78-7 for 24 h or low Ca2+ for 1 h, then exposed to adenovirus (3.95 × 109 infectious units/50 µl) for 1 h. Both treatments resulted in similar TERs but significantly greater infection was measured in cells treated with the SHE78-7. Values are means ± SD of triplicate determinations.

Time Course of Infection of 16HBE14o- Epithelia

To simplify experimental procedures, assays were adapted to a 96-well-plate format. RAd35 infectivity of 16HBE14o- cells after antibody treatment was similar to that seen on Transwell membranes (data not shown). We used the microplate assay to investigate the time course of infection. Confluent cells were incubated with 4 µg/ml of either control IgG2a or SHE78-7 for 24 h before RAd35 infection (7.88 × 10⁹ infectious units/50 µl) for 3 to 24 h. Adenoviral infection of 16HBE14o- epithelia incubated with SHE78-7 continued to increase with time (Figure 7). At 3 h, infection of control IgG2a cultures was just detectable. After 12 h exposure to RAd35, there was no further increase of infection in cultures incubated with IgG2a. For a 3-h incubation with the virus, the difference in infection between control IgG2a- and SHE78-7 was ~ 6-fold but this increased considerably with time, up to 30-fold at 24 h.

View larger version (13K): [in this window] [in a new window]   Figure 7.   Time course of RAd35 infection of 16HBE14o- epithelia in the presence of control IgG2a or SHE78-7. Cells cultured on a microplate were incubated with 4 µg/ml of either IgG2a (open circles) or SHE78-7 (filled circles) for 24 h, then exposed to RAd35 (7.88 × 109 infectious units/50 µl) for 6 h. Disruption of cadherin function resulted in a marked increase in infection rate (A). (B) is an enlargement of graph (A). Values are means ± SD of triplicate determinations.

Titration of RAd35 on 16HBE14o- Epithelia

To determine the differences in susceptibility of 16HBE14o epithelia to infection, RAd35 (20 to 315 × 10⁸ infectious units/50 µl) was titrated on cells pretreated with 4 µg/ml of either control IgG2a or SHE78-7. As before, RAd35 infection for 6 h was greatly enhanced in the SHE78-7-treated cell cultures at all viral titers tested (Figure 8). Equivalent levels of infection of control IgG2a and SHE78-7-treated cultures were achieved with viral inocula of 315 × 10⁸ infectious units/50 µl and 20 × 10⁸ infectious units/50 µl, the -galactosidase activities being 30.5 ± 3.8 and 27.3 ± 0.57 U/mg protein (n = 4, ± SD), respectively. Thus, infection of the same magnitude between control IgG2a- and SHE78-7-treated cultures occurred using approximately 16-fold lower viral titers in the latter.

View larger version (19K): [in this window] [in a new window]   Figure 8.   Titration of RAd35 on 16HBE14o- cells. Cells cultured on a microplate were incubated with 4 µg/ml of either IgG2a or SHE78-7 for 24 h, then exposed to RAd35 (20 to 315 × 108 infectious units/50 µl) for 6 h. Infection with the same viral titer resulted in a much-enhanced infection of SHE78-7-treated cell cultures. Values are means ± SD of quadruplicate determinations.

Adenovirus Infectivity of Anti-E-Cadherin-Treated Primary Human Bronchial Epithelia

To determine whether SHE78-7 could disrupt TJ function and enhance adenoviral infection in cultures of primary HBE cells, confluent cells cultured on Transwell membranes were incubated apically and basally with 4 µg/ml of either control IgG2a or SHE78-7 for 24 h, then exposed to RAd35 for 6 h. Consistent with the effect observed on 16HBE14o- cells, SHE78-7 disrupted TJ function of primary HBE epithelia with a decline in TER of 93%. An associated ~ 23-fold increase of adenoviral infection was obtained (Figure 9). These data show that disruption of cadherin function using SHE78-7 and the associated effects are not restricted to 16HBE14o- epithelia.

View larger version (17K): [in this window] [in a new window]   Figure 9.   Effect of perturbing cadherin function on RAd35 infection of primary HBE cell cultures. Cells cultured on Transwell membranes were incubated with 4 µg/ml of either IgG2a (open bars) or SHE78-7 (shaded bars) for 24 h, then exposed to RAd35 (1.58 × 1010 infectious units/100 µl) for 6 h. Compared with cells treated with IgG2a, incubation with SHE78-7 resulted in a large fall in TER (A) and a corresponding increase in adenoviral infection (B). Values are means ± SD of triplicate determinations.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The respiratory epithelium is fundamental in protecting the underlying tissue against environmental pollutants and pathogens. In this study we demonstrate the dependence of the epithelial barrier generated by 16HBE14o- and primary HBE cells on E-cadherin function, and how the efficacy of this barrier controls adenoviral infection.

In agreement with previous studies using other cell lines (17, 21, 28), we found that disruption of E-cadherin function in nonconfluent 16HBE14o- cell cultures using a specific function-blocking antibody, SHE78-7, caused colony fragmentation and complete loss of cell-cell adhesion in some areas (Figure 2). However, it was not clear whether SHE78-7 would disrupt confluent 16HBE14o- epithelia with functional TJs, inasmuch as previous studies with polarized MDCK epithelia showed that prolonged loss of epithelial integrity was achieved only after basal E-cadherin antibody addition, or by initial disruption using low-Ca²⁺ medium to facilitate antibody penetration (17, 21). To our surprise, we found that efficient disruption of the 16HBE14o- epithelial barrier occurred after apical exposure to SHE78-7 without the need to lower extracellular Ca²⁺ (Figure 3). On further investigation, we discovered occasional small, discrete defects in the 16HBE14o- permeability barrier that allowed rapid access of antibody from the apical surface (data not shown), and this probably explains the different responses of 16HBE14o- and MDCK epithelia. However, efficiency of antibody-mediated TJ disruption might be determined by SHE78-7 and the antibodies used in the MDCK studies recognizing different epitopes on E-cadherin.

Monitoring TERs showed that SHE78-7-mediated loss of epithelial integrity was concentration-related (Figure 3A) and maximal at 24 h with 4 µg/ml (Figure 3B). Maximal SHE78-7 effect resulted in TERs that were just above membrane resistances, indicating severe disruption of epithelial integrity. Loss of TER was not caused by antibody toxicity because cell viability was not affected and washout experiments showed complete restitution of TERs. These observations were confirmed using another E-cadherin- blocking antibody, HECD-1, that caused a similar decline in TER (data not shown). Although our data and previous reports demonstrate E-cadherin involvement in maintaining TJ function, it is not known whether control of E-cadherin function is a mechanism for physiologic regulation of TJ function, and if so, under what circumstances.

The disruptive effect of SHE78-7 on the TJ barrier was not 16HBE14o- cell-specific but could also be reproduced in primary HBE cells. After seeding, these cells rapidly developed TERs of ~ 3000 .cm². These values were ~ 9-fold greater than that attained by 16HBE14o- cells and, reflecting the greater barrier efficacy of these cells, it was necessary to add SHE78-7 to the apical and basal compartments to reduce the TER. In vivo, physiologic TERs of rabbit bronchial epithelia are ~ 266 .cm² (29) and those of human bronchial epithelia are thought to be similar. In comparison, primary HBE TERs are high, but despite this, 4 µg/ml SHE78-7 effectively decreased their resistances by 93%.

Several studies have used ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid (EGTA) to deplete extracellular Ca²⁺ when disrupting confluent polarized epithelia (18). Although lacking specificity in that other Ca²⁺-dependent processes could be affected, this method, like the use of E-cadherin function-blocking antibodies, effectively reduces TER, increases paracellular permeability, and results in disorganization of intercellular junctional complexes and loss of intercellular adhesion (17, 18).

Although a fall in TER following E-cadherin antibody exposure could result from ion channel activation rather than loss of barrier function, our studies with three different hydrophilic tracer moleculesmannitol, inulin, and dextranshow that paracellular permeability is increased in a dose-related manner after treatment with SHE78-7. As expected, fluxes were inversely related to the size of the tracer. Although these tracers are much smaller than adenovirus, it is probable that SHE78-7 exposure also increases adenoviral access to the lateral intercellular space. Since the adenoviral receptors Coxsackie/adenovirus receptor (CAR) (30) and major histocompatibility complex (MHC) Class I (31) and the _v3 and _v5 integrins (32) involved in cellular uptake of the virus are all confined to the basolateral membrane, penetration through the apical epithelial barrier will facilitate infection, as has been shown in EGTA-mediated barrier disruption (20). Our data show that dose-responsive increases in adenoviral infection, concomitant with increased paracellular permeability, occurred after SHE78-7 exposure, with infection being up to 30-fold greater than in the controls (Figure 5). Similarly, the decline in junction function of primary HBE cells was associated with a marked increase in adenovirus infection. Although improved access to basolateral adenoviral receptors remains the most likely explanation for much of the SHE78-7-stimulated infection, the E-cadherin antibody could also be influencing infection by other mechanisms. Indeed, this is probable, inasmuch as in a direct comparison between SHE78-7, or low Ca²⁺-treated cultures, where barrier disruption was equivalent as measured by TER, infection was 3-fold higher in 16HBE14o- cultures exposed to the antibody. It is known that cadherins have a profound effect on cell phenotype and that loss of cell polarity after SHE78-7 exposure could induce membrane protein redistribution, resulting in apical display of adenoviral receptors. Alternatively, receptor number might be increased or integrin internalization upregulated. These possibilities are currently being investigated. Our preliminary observations show that SHE78-7 causes a small but significant increase in nonspecific pinocytosis in 16HBE14o cells (data not shown), and because this has been shown to be important in adenoviral infection (33) it could contribute to the SHE78-7 stimulation of infection.

Exposure to the E-cadherin antibody could simply reduce time taken to achieve maximal infection of susceptible cells and/or it could increase the multiplicity of infection (MOI). If SHE78-7 reduces the time to maximal infection without affecting MOI, then control cultures exposed to the virus for longer periods would eventually reach equivalent infection levels. In time-course experiments, infection in SHE78-7-treated cultures was greater than in the controls at all time points and continued to increase up to 24 h, at which point infection levels were ~ 30-fold higher than in the controls (Figure 7). In contrast, control culture infection was maximal at 12 h. These data suggest that SHE78-7 increased the percentage of cells infected and/or the MOI. Further studies are necessary to define the time of maximal infection in the SHE78-7-treated cultures and how the antibody affects the MOI.

In experiments designed to measure susceptibility to infection, control IgG2a- or SHE78-7-treated cultures were exposed to various viral concentrations. Comparisons show that equivalent infection levels between the two sets of cultures required a 16-fold higher viral titer in the controls and that for an equivalent viral load, infection was always higher in the SHE78-7-treated cultures, where the epithelial barrier is compromised (Figure 8). If, as is thought to be the case, airway epithelial barrier disruption in vivo increases viral infection, then exposure to viral titers that only lead to subclinical infections in healthy individuals might induce a far more severe clinical response in patients with lung disease where the airway epithelium is damaged. This could be particularly important in asthma, chronic obstructive pulmonary disease (COPD), and cystic fibrosis, where viral infection can cause severe exacerbation of the disease (7, 8, 34, 35). Adenoviral infections are responsible for only 3 to 5% of respiratory infections in children, but if epithelial integrity is also important for protection against more relevant viruses such as RSV and rhinovirus, then the defective nature of this barrier in diseases such as asthma may be more serious than currently thought. Moreover, although incidence of adenoviral infection is low, it has been suggested to be involved in the etiology of COPD because E1A adenoviral DNA is more prevalent in the lungs of patients with COPD (36).

Although the airway epithelial barrier is important for, and efficient in, preventing infection, it poses a major problem for gene therapy treatment of lung diseases. One of the major limiting factors for gene delivery by adenoviral vectors is the basolateral disposition of the CAR and MHC Class I receptors (30). Even modest gene transfer to epithelial cells necessitates high adenoviral titers that are associated with adverse inflammatory responses. In an attempt to circumvent this problem, several laboratories have depleted extracellular Ca²⁺ to disrupt the apical barrier, thereby enhancing viral infection 2- to 9-fold (19, 20), and promising results have recently been achieved by a similar approach in vivo (37). Our in vitro studies using a more specific method for barrier disruption resulted in a 30-fold enhancement of gene transfer. Also, when we directly compared SHE78-7 exposure with the low-Ca²⁺ treatment, antibody-mediated disruption of E-cadherin function resulted in a 3-fold greater infection. Therefore, specific modification of cadherin function to alter the state of the intercellular junctional complex might be a potentially useful approach for enhancing delivery of therapeutic genes to patients with lung disease.

In summary, loss of epithelial integrity induced by antibody-mediated disruption of E-cadherin function results in a concomitant increase in adenoviral infection. As a corollary to these findings, manipulation of E-cadherin function may provide a means of altering epithelial integrity and, if enhanced, this could provide greater resistance to infection with adenovirus and possibly other pathogens. In contrast, our work also provides a possible way of improving adenoviral-mediated gene transfer, which is critical if gene therapy for lung diseases is to be successful.