Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Tight junctions (TJs) play a central role in sealing the intercellular space between the apical and basolateral compartments of epithelial and endothelial cells. Modulation of the barrier function of the TJ has been investigated as a strategy to enhance transmucosal drug absorption, particularly to increase the bioavailability of drugs after rectal administration (Morimoto et al., 1989; Yamazaki et al., 1990; Anderberg et al., 1993; Kinouchi and Yata, 1996; Yamamoto et al., 1996; Shimazaki et al., 1998; Soderholm et al., 1998). The sodium salts of several medium chain fatty acids (MCFAs), particularly capric (C10) and lauric (C12) acids, have been shown to increase rectal drug absorption, presumably by causing alterations in intestinal TJ barrier function. C10 has also been shown to lead to profound alterations in the barrier function of the airway TJ and has been investigated as an agent to enhance viral-mediated airway gene transfer (Coyne et al., 2000). These agents are attractive as potential therapies to enhance absorption of gene transfer vectors due to the rapid onset of action (within minutes) and their relatively rapid recovery (within hours) after treatment (Lindmark et al., 1995, 1998; Tomita et al., 1995, 1996; Coyne et al., 2000).

Although MCFAs have been widely investigated as agents to increase the delivery of therapeutic agents, relatively little is known regarding their primary mechanism of action. The ability of intercellular TJs to function as a barrier to the diffusion of macromolecules is dynamically regulated by numerous intracellular signals and the permeability properties of the TJ vary in response to changes in physiological state. Therefore, agents that regulate the TJ likely do so by direct or indirect actions on intracellular signals and/or the protein components of the TJ.

The integrity of the intercellular junctions requires a finite concentration of extracellular Ca²⁺. Removal of Ca²⁺ by chelators such as EGTA leads to significant increases in paracellular permeability and gene transfer efficiency (Duan et al., 1998; Wang et al., 1998; Coyne et al., 2000). The role of intracellular Ca²⁺ (Ca_i²⁺) in TJ regulation has also been well studied. Studies in Caco-2 cells have suggested that the increase in C10-induced permeability may be a result of the elevation of Ca_i²⁺ via a phospholipase C (PLC)-dependent pathway (Tomita et al., 1995, 1996; Lindmark et al., 1997, 1998). Pharmacological inhibition of PLC, myosin light chain kinase, and calcium calmodulin have been shown to reduce the C10- and C12-induced alterations in TJ integrity in Caco-2 cells (Lindmark et al., 1998).

These inhibition studies, implicating a role for PLC in the effect of C10 on permeability, focused on the relatively long-term effect of exposure (12-60 min) (Lindmark et al., 1998). However, the effects of C10 on paracellular permeability occurred within seconds of exposure in both Caco-2 and WD primary HAE cells (Tomita et al., 1995; Coyne et al., 2000), and inhibition of PLC in Caco-2 cells did not prevent the acute increase in permeability. The role that the increase in Ca_i²⁺ levels plays in the acute mechanism of C10 and C12, which have the similar acute effects on the TJ, remains unclear but both agents likely mediate their initial effects via a similar path (Lindmark et al., 1995). Because these compounds remain attractive as enhancers of therapeutic drug or vector absorption, a clearer understanding of the acute mechanism of action of these compounds is essential.

To determine the acute mechanism of action of C10 and C12 in the airway epithelium, we incubated primary human airway epithelial (HAE) cells with C10 and C12 and determined the effects of this treatment on transepithelial resistance (R_T), paracellular permeability, and luminal gene transfer efficiency. We then examined the effects of MCFA exposure on the level of Ca_i²⁺ and determined whether buffering Ca_i²⁺ to high and low levels affected the C10-induced changes in permeability. We then characterized the effects of C10 and C12 on structural components of the intercellular junction, including JAM, claudins-1 and -4, actin, and -catenin.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Antibodies. The sodium salts of C10 and C12 were purchased from Sigma-Aldrich (St. Louis, MO). Rabbit polyclonal antibodies to ZO-1, claudin-1, and -catenin and mouse monoclonal antibody to claudin-4 were purchased from Zymed Laboratories (South San Francisco, CA). The 3D8 mouse monoclonal antibody to JAM was kindly provided by Dr. Kenji Ishii (Department of Geriatric Medicine, Kyoto University Graduate School of Medicine, Kyoto, Japan). Fura-2 acetoxymethyl ester (AM) and Fluo-4 were purchased from Molecular Probes (Eugene, OR). BAPTA-AM, thapsigargin, UTP, and inhibitors were purchased from Sigma-Aldrich.

Cell Culture. Primary airway cells from human subjects were isolated in accordance with guidelines approved by the Committee on the Protection of the Rights of Human Subjects. Bronchial cells of normal type were isolated from surgical specimens, plated at a density of 2 × 10⁵ cells/12 mm (0.4-µm pore size) on Transwell-Col inserts, and maintained in a 50:50 mixture of LHC basal medium (Biofluids, Rockville, MD) and Dulbecco's modified Eagle's medium-H medium supplemented with growth factors, retinoic acid, and bovine serum albumin as described previously (Gray et al., 1996). Upon reaching confluence, medium was aspirated from the apical surface and cells maintained at an air-liquid interface for 2 to 4 weeks. Cultures with >10% cilia as determined by microscopy, and an R_T of >600 -cm² was selected for experiments.

Measurement of R_T. The R_T of primary HAE cells was monitored with an ohmmeter (EVOM; WPI, Sarasota, FL). Medium was added to the apical and basolateral surfaces and R_T was measured between electrodes. All R_T values have been corrected for background contributed by the Transwell.

Recombinant Adenovirus (Ad) Infections. Recombinant, first-generation E1, E3-delected adenovirus serotype-5 vectors containing Escherichia coli lacZ (AdlacZ) complementary DNA (cDNA) were prepared by the University of North Carolina at Chapel Hill Gene Therapy Vector Core. The vector titers were approximately 10¹¹ transducing units/ml. Cultures of primary HAE cells were treated with MCFAs and when R_T values had decreased to <90% of their initial levels (2-5-min treatment) cells were washed three times with phosphate-buffered saline (PBS) and infected from the apical surface with AdlacZ at a multiplicity of infection of 100. After infection for 2 h at 37°C, cells were washed with PBS and incubated for an additional 48 h. Transduction efficiency was measured with the Galactostar light assay (Tropix, Bedford, MA). The His6-tagged version of the adenovirus type-5 fiber knob containing the 22nd repeat of the fiber shaft was expressed from the pQE30 plasmid (QIAGEN, Valencia, CA) and was purified as described previously (Henry et al., 1994).

Transepithelial Cell Permeability. For visualization of paracellular permeability in live primary HAE cultures, cultures were treated with 150 µl of vehicle, C10, or C12 in nominally Ca²⁺-free HEPES-buffered Ringer (HBR) (130 mM NaCl, 5 mM KCl, 1.3 mM MgCl₂, 10 mM HEPES pH 7.4, 1.3 mM CaCl₂, and 5 mM glucose) containing 2 mg/ml Texas Red (TR)-labeled dextrans of 10 and 2000 kDa while mounted on a glass coverslip over an objective coupled to a confocal microscope. xz-scans were recorded at the indicated time points.

Measurements of Ca_i²⁺. Primary HAE cells were loaded through the mucosal and serosal surfaces with Fura-2 acetoxymethyl ester (5 µM) for 30 min in tissue culture medium (F-12) at 37°C and then mounted in a miniature Ussing chamber over an objective of a microscope coupled to a microfluorometer as described previously (Paradiso et al., 1995). Cells were bathed in nominally Ca²⁺-free HBR and the Fura-2 fluorescence intensity ratio (excitation 340/380; emission >450 nm) collected from a field of 30 cells. Data are presented as the 340/380 ratio and individual wavelengths.

ATP Depletion. ATP depletion was accomplished by inhibition of both oxidative and glycolytic pathways. Briefly, glycolysis was inhibited by 3-h incubation at 37°C of primary HAE cells in a modified glucose-free Ringer's solution containing 2 mM glutamine in the apical and basolateral compartments as described previously (Bacallao et al., 1994). After this incubation, ATP was rapidly depleted by bathing cells apically in 10 µM antimycin A and 2-deoxyglucose (10 µM) (Bacallao et al., 1994), and R_T was determined at indicated time points.

Immunofluorescent Labeling and Confocal Microscopy. Cells were permeabilized with methanol at 20°C for 30 min. Antibodies to ZO-1, claudin-1, -catenin, and claudin-4, and JAM diluted to 1:1000 were added to the luminal surface for 1 h. Cells were washed with PBS, and TR-labeled secondary antibodies (Amersham Biosciences Inc., Piscataway, NJ) diluted 1:600 in 10% goat serum/PBS were added to the luminal surface. For JAM and ZO-1 double labeling, 3D8 mouse anti-JAM and rabbit anti-ZO-1 were added to the cultures. After washing, anti-rabbit TR and anti-mouse fluorescein isothiocyanate antibodies were incubated in 10% goat serum/PBS for 1 h at room temperature. For actin staining, cultures were permeabilized in 1% Triton X-100, washed three times with PBS, and incubated with 1:250 dilution of Oregon Green phalloidin (Molecular Probes) for 30 min at room temperature and washed in PBS. Cells were postfixed with 4% paraformaldehyde. Transwell-Col inserts were excised and mounted on slides with 100 µl of Vectashield (Vector laboratories, Burlingame, CA) containing 4,6-diamidino-2-phenylindole. Images were captured with a confocal laser scanning microscope (Leica, Exton, PA).

Statistics. Data are presented as mean ± S.E.M. A one-way analysis of variance and Bonferroni's correction for multiple comparisons were used to determine statistical significance (p < 0.05).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

MCFA-Induced Alterations in TJ Barrier Function. To evaluate the acute effects of MCFAs on permeability, we measured the effect of 30 mM caprylic (C8), 30 mM capric (C10), and 10 mM lauric (C12) treatment on R_T at 0, 15, 30, 60, and 300 s (Fig. 1A). By 15 s post-treatment, R_T in cultures exposed to C10 and C12 already showed a pronounced decrease in R_T, corresponding to R_T values of 46.9 ± 5.5% (C10) and 32.7 ± 3.1% (C12) of vehicle-treated controls. By 60 s, R_T had decreased to 19.4 ± 2.3% (C10) and 14.3 ± 1.5% (C12) of control levels. However, C8 has no effect on R_T at any time point. Upon removal of C10 and C12, R_T increased over 12 h, and returned to vehicle-treated levels by 24 h post-treatment (Fig. 1B).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Effect of MCFAs on RT and gene transfer efficiency. A, acute (<5 min) effects of C8, C10, C12 on RT post-treatment. B, recovery of RT following removal of C10 and C12. C, AdlacZ transduction following treatment of cultures with vehicle, C8, C10, or C12 48 h postinfection with AdlacZ at an MOI of 100. D, effect of adenovirus serotype-5 fiber knob on AdlacZtransduction following vehicle or MCFA treatment. *, significantly different than vehicle-treated controls (C) and significantly different than cultures treated without AdlacZ fiber knob (D) (p < 0.05), n = 4.

C10-and C12-Induced Alterations in Permeability. The histological appearance of primary WD HAE is shown in Fig. 2A. These cells develop as a well differentiated pseudostratified epithelium consisting of ciliated columnar cells that face the lumen and basal cells. As cultures age, goblet cells may also become apparent, and some multilayering may be seen in the basal cell layer. Note that C10 treatment caused no significant change in the histological appearance of the epithelium, whereas C12 treatment induced intermittent breaks or disruption in the luminal surface (Fig. 2A, see arrows). Measurements of lactate dehydrogenase levels after C10 and C12 exposure revealed no significant increase in lactate dehydrogenase release after C10 treatment, but an increase after C12, indicative of cellular injury (L. G. Johnson, M. K. Vanhook, C. B. Coyne, N. Haykal-Coates, and S. H. Gavett, manuscript submitted for publication).

View larger version (48K): [in this window] [in a new window]   Fig. 2.   Imaging of 10-kD Texas Red dextran permeability in live primary HAE cells. A, primary HAE cultures were exposed to vehicle (medium), 30 mM C10, or 10 mM C12 in nominally Ca2+-free HBR containing 2 mg/ml Texas Red-labeled 10-kD dextran and scans taken along the xz-axis at 0, 15, 30, 60, and 300 s following apical application. Images were captured from at least four cultures in each treatment type. Left, H&E staining of a representative section of vehicle, C10, or C12-treated WD primary HAE cells. B, fate of 10-kD TR-dextran internalized following vehicle or MCFA treatment at 1, 4, 12, and 24 h postexposure. Images are representative of at least six cultures each.

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Imaging of 2000-kD Texas Red dextran permeability in live primary HAE cells. A, primary HAE cultures were exposed to vehicle (medium), 30 mM C10, or 10 mM C12 in nominally Ca2+-free HBR containing 2 mg/ml Texas Red-labeled 2000-kD dextran and scans taken along the xz-axis at 0, 15, 30, 60, and 300 s following apical application. Images are representative of at least four cultures in each treatment type. B, fate of 10-kD TR-dextran internalized following vehicle or MCFA treatment at 1, 4, 12, and 24 h postexposure. Images are representative of at least six cultures each.

View larger version (44K): [in this window] [in a new window]   Fig. 4.   Imaging of 10- and 2000-kD Texas Red dextran permeability in live primary HAE cells. Primary HAE cultures were exposed to vehicle (medium), 30 mM C10, or 10 mM C12 in nominally Ca2+-free HBR containing 2 mg/ml Texas Red-labeled 10- or 2000-kD dextran and scans taken along the xy- and xz-axis at 60 s following apical application. Middle panel, 10-kD dextran; right panel, 2000-kD dextran. Images were captured from of at least four cultures in each treatment group.

Effect of C10 and C12 on Intracellular Calcium Levels. A rapid and sustained increase in Ca_i²⁺ measured by Fura-2 fluorescence has been reported as the potential mechanism of action for C10-induced changes in R_T in Caco-2 cells (Tomita et al., 1995). To determine whether C10 mobilizes calcium from Ca²⁺ stores in the endoplasmic reticulum (ER), we added apically 5 µM thapsigargin (an ER Ca²⁺-ATPase inhibitor) to deplete stores. Subsequent C10 (10 mM) application led to a rapid increase in Ca_i²⁺ levels by microfluorometer anlaysis (data not shown), suggesting that C10 may increase Ca_i²⁺ via thapsigargin-insensitive Ca²⁺ stores. However, the pattern of the sustained increase in Ca_i²⁺ induced by C10, which did not return to baseline levels, raised concerns about possible adverse effects of the MCFA on the Fura-2 compound.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Measurement of C10- and C12-induced alterations in Cai2+ by Fluo-4 fluorescence. A, primary HAE cultures were preloaded with Fluo-4 AM and changes in fluorescence (indicating increased Cai2+) determined by confocal microscopy following apical addition of 30 mM C10 or 10 mM C12 in nominally Ca2+-free HBR containing 200 µM EGTA at the indicated time points. B, fluorescence intensity of images in A. *, significantly greater than vehicle and C12-treated cultures (p < 0.05).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Measurement of C10-induced alterations in Cai2+ via thapsigargin-sensitive stores. A, primary HAE cultures were preloaded with Fluo-4 AM and fluorescence intensity determined following application of thapsigargin, followed by UTP and 30 mM C10 in nominally Ca2+-free HBR containing 200 µM EGTA at the indicated time points. Alternatively, vehicle (HBR) was applied before C10.

Effect of Buffering Ca_i²⁺ on Permeability. If the action of C10 depends on the rapid increase in Ca_i²⁺ levels, then chelation of Ca_i²⁺ should inhibit its effects on the TJ. To chelate Ca_i²⁺, cells were loaded with BAPTA-AM (100 µM) in HBR containing 300 µM Ca²⁺ in the basolateral compartment and nominally Ca²⁺-free HBR in the apical compartment. Cells were subsequently exposed to 30 mM C10 in nominally Ca²⁺-free buffer containing 100 µM BAPTA-AM in the apical and basolateral compartments. The R_T of BAPTA-AM-loaded cultures did not differ significantly from those treated with vehicle before addition of C10. C10 reduced R_T by 94.9 ± 2.5% in vehicle (no BAPTA-AM)-treated cultures, which was not different from the reduction in BAPTA-loaded cells (94.7 ± 1.8%) (Fig. 7A). These data suggested that buffering Ca_i²⁺ to low levels had no effect on C10-mediated changes in R_T.

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Effect of buffering Cai2+ levels on C10-induced alterations in RT. A, Cai2+ levels were depleted with Bapta-AM, and RT was measured following application of 30 mM C10 at 5 min. Alternatively, Cai2+ was increased with application of LaCl3 and thapsigargin alone or sequentially. Cells were also pretreated with LaCl3 and thapsigargin or vehicle (HBR) before addition of 30 mM C10, and RT was measured at 5 min. B, top panel, effect of pretreatment of primary HAE cells with Bapta-AM on C10-induced changes in Fluo-4 fluorescence intensity. B, bottom panel, effect of La3+ and thapsigargin treatment on primary HAE cells of Fluo-4 intensity. Images are representative of three cultures. *, significantly different than vehicle-treated controls and Bapta-AM treated cells (A, left) or from vehicle and LaCl3, thapsigargin-treated (A, right) (p < 0.05).

Effect of Signaling Inhibitors on MCFA-Induced Reduction in R_T. Previously published data suggested that C10-induced increase in Ca_i²⁺ levels may result from PLC activation, which is a necessary step in its long-term effects on R_T (Lindmark et al., 1998). To determine the role of signaling components in the acute effects of C10 on airway permeability, we pretreated primary HAE cultures with pharmacological inhibitors of signaling components known to be effectors in TJ regulation in Caco-2 cells: a nonspecific PLC inhibitor (U73122, 50 µM) and phospholipase A2 inhibitor (48/80, 50 µg/ml), an inhibitor of calmodulin (W7, 50 µM), Ca²⁺/calmodulin-dependent protein kinase inhibitor (KN62, 20 µM), and a nonspecific PKC inhibitor (H7, 50 µM). There was no effect of any inhibitor tested on the C10-induced decreases in R_T (Fig. 8A). Although activation of these signaling cascades cannot be ruled out as a consequence of C10 treatment, they do not seem to play a functional role in the rapid changes in R_T (1-5 min) induced by C10.

View larger version (11K): [in this window] [in a new window]   Fig. 8.   Effect of signaling inhibitors and ATP depletion on RT. A, effect of several pharmacological inhibitors of signaling components on the C10-induced reduction in RT at 5 min post-treatment. Cultures were pretreated with the indicated inhibitors, and then 30 mM C10 was added in the presence of inhibitors. B, oxidative and glycolytic energy pathways were depleted, and RT was measured at the indicated time points.

Effect of ATP Depletion on R_T. Because treatment of Caco-2 cells with C10 has been shown to decrease cellular ATP levels, we investigated whether ATP depletion could induce the same effect on R_T as C10 (Lindmark et al., 1998). We depleted ATP levels by inhibiting both the oxidative and glycolytic metabolic pathways in primary HAE cells. Although this depletion did lead to a reduction in R_T, the kinetics of these changes was distinct from that of either C10 or C12. C10 and C12 both exert their maximal effects on R_T by 2 to 5 min post-treatment, whereas the maximal reduction in R_T induced by depletion of ATP did not occur until after 90 min of treatment (Fig. 8B). Therefore, the depletion of ATP by C10 and C12 does not seem to play a functional role in the acute effects on TJ barrier function.

Effect of C10 and C12 on Structural Components of the TJ. We evaluated whether MCFAs that modulate airway permeability do so by their effects on key structural components of the TJ. Because JAM has been speculated to play a role in maintaining junctional integrity, we investigated the effect of MCFA exposure of primary HAE cells on JAM distribution. Although JAM was clearly redistributed after exposure to either C10 or C12, ZO-1 distribution remained relatively unchanged (Fig. 9). C10 and C12 also induced a redistribution of actin (Fig. 10A), which has been shown previously to interact directly with the TJ via its interaction with ZO-1 (Fanning et al., 1998, 2002). Image analysis of relative fluorescence intensity revealed that both C10 and C12 caused redistribution of actin, with 39 ± 3.3% less intensity for C10 and 38 ± 5% for C12 in comparison with vehicle-treated controls. This redistribution was not blocked by preincubation of cultures with BAPTA-AM and could not be replicated by treatment of naïve cells with La³⁺ and thapsigargin (Fig. 10A). The data showing no changes in actin distribution after incubation with La³⁺ and thapsigargin suggested that the effects of C10 and C12 on actin are Ca²⁺-independent.

View larger version (58K): [in this window] [in a new window]   Fig. 9.   Localization of ZO-1 and JAM following MCFA exposure. Immunofluorescent staining for ZO-1 and JAM in primary HAE cells exposed to vehicle, 30 mM C10, or 10 mM C12 for 5min. JAM staining is in green, ZO-1 is in red, and the merged image is at the bottom. Colocalization appears as yellow. Images are representative of at least three cultures each.

View larger version (51K): [in this window] [in a new window]   Fig. 10.   MCFA-induced redistribution of actin and -catenin. A, primary HAE cultures were treated with vehicle, 30 mM C10, or 10 mM C12 and stained for actin. Additionally, cells were pretreated with Bapta-AM before C10 exposure or La3+ and thasigargin and stained for actin. B, immunofluorescent localization of -catenin in primary HAE cells exposed to vehicle, 30 mM C10, or 10 mM C12 for 5 min. Blue staining represents DAPI-stained nuclei, and red staining represents -catenin staining. Data are representative of imaging of at least three cultures each.

View larger version (65K): [in this window] [in a new window]   Fig. 11.   Redistribution of claudin-1 and -4 induced by MCFA exposure. Immunofluorescent localization of claudin-1 and -4 in primary HAE cells exposed to vehicle, 30 mM C10, or 10 mM C12 for 5 min. Claudin-4 staining is in green, claudin-1 is in red, and the merged image is at the bottom. Colocalization appears as yellow. Data are representative of imaging of at least three cultures each.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The ability of the MCFAs C10 and C12 to enhance drug absorption and in more recent studies, absorption of Ad vectors, has emphasized the therapeutic potential of these compounds. A role for Ca²⁺ regulation of transepithelial permeability has been suggested as a mechanism for these compounds. Alterations in either extracellular or intracellular Ca²⁺ levels have a pronounced effect on TJ barrier function and may contribute to the relative permeability of the junction. It is therefore not surprising that agents that lead to alterations in Ca_i²⁺ may have effects on both R_T and solute permeability. However, the data presented here indicate that although C10 and C12 have similar effects on the protein components of the TJ, their effects on Ca_i²⁺ levels diverge. This incongruity would suggest that C10 and C12 have dissimilar mechanisms, and/or that the effect of C10 on Ca_i²⁺ is a secondary response not associated with its effects on permeability.

The R_T of primary HAE cells was significantly reduced within seconds after exposure to either C10 or C12, whereas treatment of cultures with C8 did not have any effect on R_T (Fig. 1A). The decrease in R_T induced by C10 and C12 correlated with a significant enhancement of AdlacZ transduction, indicating that the barrier function of the TJ was altered after MCFA exposure, whereas C8 had no effect on this parameter (Fig. 1C). The increase in transduction efficiency after pretreatment of cultures with C10 and C12 indicate considerable effects of these agents on the airway permeability to macromolecules, because Ad vectors are approximately 2000 to 4000 kDa. C8 had no effects on either R_T or paracellular permeability and thus was unable to enhance diffusion of an Ad vector to the basolateral surface. The significant effects of both C10 and C12 on permeability were reversible and returned to baseline levels within 24 h of treatment (Fig. 1B).

We further explored the effects of C10 and C12 on airway epithelial permeability with live cell imaging of fluorescently labeled dextrans by confocal microscopy (Figs. 2 and 3). Surprisingly, C10 and C12 induced cellular uptake of both low (10 kDa) and high (2000 kDa) molecular weight dextrans, which took greater than 4 h to clear for C10 and greater than 12 h to clear when induced by C12. C10 induced a relatively specific uptake in columnar cells within the epithelium, whereas C12-induced uptake into both columnar and basal cells. The cellular uptake of dextran in C12-treated cultures may be exaggerated by the relative toxicity induced by this agent. Paracellular flow of dextrans was also increased after C10 and C12 treatments (Figs. 2A and 3A, see arrows). Although the images would suggest that the majority of dextran uptake seemed to be cellular, the retention of dextrans within cells prevents an accurate assessment of cellular uptake relative to the amount of paracellular flow (Fig. 4). Therefore, measurements of mannitol or dextran flux immediately after MCFA-treatment would not detect uptake into cellular compartments, rather these measurements of solute permeability would exclusively reflect paracellular flow.

Because of the increased cellular uptake of dextrans across the apical membrane, we tested whether CAR-dependent entry of Ad was required for transduction of WD primary HAE cells after treatment with C10 or C12. If increased apical membrane permeability increased Ad uptake, competition with Ad fiber knob for CAR binding should not inhibit Ad transduction. As shown in Fig. 1D, incubation of C10- and C12-treated cells with Ad vector in the presence of excess fiber knob protein inhibited Ad-mediated transduction. Because CAR has been localized to the basolateral membrane of WD HAE cells (Pickles et al., 1998), these data suggest that the enhancement of Ad gene transfer occurred predominantly by enhancing paracellular permeability. Increased cellular permeability played a much smaller role in enhancing gene transfer.

Our data indicated that measurements of the effects of C10 on Ca_i²⁺ by Fura-2 were complicated by interference of C10 and C12 (data not shown) with Fura-2 fluorescence, particularly its effects on the 380-nm wavelength even at doses as low as 0.3 mM (data not shown). However, our studies with Fluo-4-loaded cells support previous reports that C10 application leads to a rapid and sustained increase in Ca_i²⁺ (Tomita et al., 1995, 1996; Lindmark et al., 1998), resulting from the release of thapsigargin-sensitive ER Ca²⁺ stores (Figs. 5 and 6). Although C10 does lead to changes in Ca_i²⁺, correlating temporally with changes in R_T, our data demonstrated that this effect may be dissociated from C10 effects on permeability in primary HAE cells. When Ca_i²⁺ is chelated with BAPTA-AM, C10 and C12 maintain their ability to rapidly decrease R_T (Fig. 7A), consistent with a Ca²⁺-independent mechanism. Additionally, mimicking the rapid, sustained increase in Ca_i²⁺ induced by C10 by treatment of cells with thapsigargin and La³⁺ had no effect on R_T (Fig. 7A). These studies combined confirm that the effects of C10 and C12 on permeability are independent of changes in Ca_i²⁺.

Although ATP levels have been shown to be decreased after application of C10, this is likely a secondary response, which may contribute to the long-term, but not acute, effects of C10 on R_T and permeability. This is supported by the differences in the effects of ATP depletion and C10 on R_T, where greater than 15 min was required for ATP depletion to decrease R_T compared with C10-induced effects on R_T occurring within seconds and reaching maximum effects in minutes (Fig. 8). These data support previously published data showing reversible long-term decreases in R_T after ATP depletion (Canfield et al., 1991).

The similar effects of C10 and C12 on protein components of the TJ provide further evidence that C10 acts via a Ca²⁺-independent pathway and likely acts upon the same protein components as C12 in airway epithelia. Although C12 did not lead to increases in Ca_i²⁺, both C10 and C12 caused redistribution of the same TJ protein components, the claudins and JAM. The claudin family of transmembrane proteins seems to exert effects on TJ integrity independently of Ca²⁺ (Kubota et al., 1999). The reorganization of claudins-1 and -4 may account for the almost immediate effects of both C10 and C12 on R_T and permeability (Fig. 11). This would suggest that these proteins are unaffected by agents that alter junctional integrity via a Ca²⁺-dependent cascade.

Although C10 and C12 do not significantly affect the distribution of ZO-1, both led to significant relocalization of JAM, which has been shown to interact with ZO-1 (Fig. 9) (Bazzoni et al., 2000; Ebnet et al., 2000). Because JAM has been implicated in the resealing of the junction (Liu et al., 2000), this suggests that the redistribution of JAM may contribute to the MCFA-induced effects on paracellular permeability.

Actin filaments play a central role in cellular architecture and have been shown to bind directly to TJ-associated proteins. Both C10 and C12 caused redistribution of actin in primary HAE cultures. Previously published data in Caco-2 cells suggested that the initial increase in Ca_i²⁺ levels induced by C10 causes the reorganization of actin filaments (Sakai et al., 1998). However, when primary HAE cultures were treated with lanthanum and thapsigargin to mimic the rapid and sustained increase in Ca_i²⁺ levels induced by C10, there was no effect on actin distribution (Fig. 10A). Furthermore, although C12 had minimal effects on Ca_i²⁺ levels, it induced a similar level of reorganization as C10, indicating Ca²⁺-independent reorganization. C10 and C12 also affected a component of the adherens junction, -catenin, which has also been shown to regulate claudin-1 gene expression (Miwa et al., 2001).

The results of this study suggest that both C10 and C12 exert their acute effects on permeability via a Ca²⁺-independent mechanism. Although exposure of primary HAE cells to C10 caused a rapid increase in Ca_i²⁺ levels, exposure to C12 had little to no effect on Ca_i²⁺ levels. In addition, inhibitors of Ca²⁺-dependent signaling cascades had no effect on the C10-induced effects on R_T, nor did blocking the rise in Ca_i²⁺ with BAPTA-AM, or mimicking the increase with thapsigargin and lanthanum. However, both C10 and C12 caused reorganization of JAM, actin, claudins-1 and -4, suggesting that both agents exert their effects on the TJ via direct or indirect effects on these protein components. Thus, C10 and C12 may act via similar Ca²⁺-independent mechanisms in human primary airway cells to enhance permeability.

The effects of C10 and C12 on apical membrane permeability and cellular uptake are more of an enigma. Although disruption of apical membrane lipids with loss and collapse of the apical domain might occur, preservation of ZO-1 staining (Fig. 9) and previous data demonstrating that C10 does not disrupt occludin distribution in WD HAE cultures (Coyne et al., 2000) suggest that the polarity of these cells remains intact after C10 and C12 treatment. Clearly, an immediate increase in apical membrane permeability occurs that is restricted to the apical pole by a functional TJ. The cellular retention of dextrans for extended periods (4-12 h) in lumen-facing columnar cells is consistent with this notion. When this observation is coupled with the inhibition of Ad-mediated transduction by excess fiber knob and the immediate changes in claudins-1, -4, and JAM localization, these data suggest that an increase in paracellular permeability is the primary method of C10- and C12-induced enhancement of Ad-mediated gene transfer.

Understanding how MCFAs enhance permeability to permit more efficient transduction of airway cells is important because it may allow for more accurate predictions of toxicities induced by this approach in human subjects.