Transport characteristics of intact albumin were investigated using primary cultured rat alveolar epithelial cell monolayers. The apical-to-basolateral (ab) flux of intact fluorescein isothiocyanate (FITC)-labeled albumin (F-Alb) is greater than basolateral-to-apical (ba) flux at the same upstream [F-Alb]. Net absorption of intact F-Alb occurs with half-maximal concentration of ∼1.6 μM and maximal transport rate of ∼0.15 fmol · cm−2 · s−1. At 15 and 4°C, both ab and ba F-Alb fluxes are not different from zero, collapsing net absorption. The presence of excess unlabeled albumin (but not other macromolecule species) in either the apical or basolateral fluid significantly reduces bothab and ba unidirectional F-Alb fluxes. Photoaffinity labeling of apical cell membranes revealed an ∼60-kDa protein that exhibits specificity for albumin. These data indicate that net absorption of intact albumin takes place via saturable receptor-mediated transcellular endocytotic processes recognizing albumin, but not other macromolecules, that may play an important role in alveolar homeostasis in the mammalian lung.
- protein transport
- air-blood barrier
- saturable transcytosis
- albumin-binding sites
- alveolar fluid balance
alveolar epithelium lines the distal air spaces of the lung and provides high resistance to the leak of solutes and fluid from the surrounding interstitial and vascular spaces, as suggested by a number of physiological and morphological investigations (for reviews see Refs.13 and 19). Moreover, alveolar epithelium exhibits many solute transport processes, including active Na+ resorption from apical to basolateral fluid (6,22). The presence of transport processes for removal of solutes (and water) from apical fluid by active mechanisms is thought to be of major importance for the maintenance of alveolar fluid balance in healthy lungs and may also participate in the removal of alveolar fluid during recovery from lung injury.
The presence of osmotically active solutes (e.g., serum proteins) in interstitial and alveolar lining fluid may be a contributing factor to normal alveolar fluid balance. Although immunocytochemical and biochemical approaches have been used to demonstrate the presence of serum proteins (e.g., albumin, transferrin, and IgG) on the epithelial surface of the distal air space and in bronchoalveolar lavage fluid (1-3), the ability of serum proteins to traverse the alveolar epithelium remains a subject of debate. The air-blood barrier of sheep lung was reported to be impermeable to a serum protein, albumin, within the physiological range of lung inflation volumes (11, 12, 31). Horseradish peroxidase (HRP) and cytochrome c were shown not to cross epithelial junctions but are found in interstitial spaces when these markers were vascularly perfused (32-34). However, finite albumin transport rates were more recently reported for lungs of dogs, guinea pigs, sheep, rabbits, and rats (for reviews see Refs. 13 and19). Of particular interest is the fact that ∼1% of intratracheally instilled albumin tracer is translocated into the vascular space per hour in all species studied. Apparent permeability index for albumin is ∼5 × 10−10 cm/s for sheep lung (18), about an order of magnitude smaller than that estimated for canine lung (14, 16, 17, 29). These data suggest that the distal respiratory epithelial tract allows slow (perhaps species-specific) but finite translocation of proteins into the interstitial and vascular spaces from the air spaces of the lung. The specific mechanisms, however, remain largely unknown.
In this study, we investigated albumin transport properties of the alveolar epithelium, utilizing an in vitro model of primary cultured rat pneumocyte monolayers grown on tissue culture-treated polycarbonate filters (6, 22, 25, 26). These monolayers comprise alveolar epithelial type I-like cells (5, 8), develop high barrier resistance (>2,000 Ω-cm2), and actively reabsorb Na+ (∼0.2 μeq · cm−2 · h−1) from apical fluid. Transepithelial fluxes of intact fluorescein isothiocyanate (FITC)-labeled bovine serum albumin (F-Alb) across the monolayers as a function of [F-Alb] were studied. Effects of temperature, specificity of albumin absorption processes, and cellular metabolism of F-Alb were also investigated. It was determined that intact albumin translocation across the alveolar epithelial barrier occurs predominantly via transcellular saturable processes (e.g., transcytosis) mediated by specific albumin-binding sites.
Primary culture of rat alveolar epithelial cell monolayers.
We have previously reported the routine generation of tight monolayers of rat alveolar epithelial cells in primary culture (6, 22, 25,26). Briefly, lungs of male specific pathogen-free Sprague-Dawley rats (∼150 g) were perfused, lavaged, treated with porcine pancreatic elastase (2.5 U/ml; Worthington, Lakewood, NJ), and chopped into ∼1-mm3 tissue blocks (devoid of large airways). The crude cell mixtures obtained by triturating the small tissue blocks were filtered (150-, 35-, and 10-μm Nitex; Tetko, Depew, NY) and further purified by IgG panning methods (10) to yield >90% type II pneumocytes. The purified pneumocytes were resuspended in culture medium (1:1 mixture of Dulbecco's modified Eagle's medium and Hanks' F-12 supplemented with 10% newborn bovine serum, 100 U/ml penicillin, and 100 ng/ml streptomycin) and plated onto tissue culture-treated polycarbonate filters (1.13 cm2, Transwell; Costar, San Francisco, CA) at 1.2 × 106 cells/cm2 (on day 0). After 2 days of culture in 5% CO2 in air at 37°C, these cells-on-filter preparations were fed on both sides with fresh culture medium. Monolayers develop spontaneous potential difference (PD) >10 mV (apical negative) and transepithelial electrical resistance (TEER) >2,000 ohm-cm2 by day 4 of culture. We measured PD and TEER of monolayers before and after flux experiments using a volt-ohmmeter (MilliCell; Millipore, Marlborough, MA).
Measurements of unidirectional fluxes of F-Alb and estimation of kinetic constants.
Primary cultured monolayers of rat alveolar epithelial cells were used on day 4 for all experiments. Cell monolayers contained in 12-well clusters were washed twice with preequilibrated (37°C, pH 7.4) Ringer solution on both sides and allowed to equilibrate with the new bathing medium for 2 h in a humidified incubator (37°C). Ringer solution contains (in mM) 126.4 NaCl, 5.4 KCl, 0.78 NaH2PO4, 1.8 CaCl2, 0.81 MgSO4, 15N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES), 5.55 glucose, and 0.075 dextran (of average mol wt of ∼70,000 Da). The osmolarity of Ringer solution, when measured with an osmometer (Micro-Osmette; Precision Systems, Natick, MA), is ∼300 mosM.
We initiated unidirectional flux measurements by adding F-Alb to either apical or basolateral upstream fluid, followed by monitoring accumulation of the protein marker in contralateral downstream fluid as a function of time at 37°C. Bathing fluid volumes were 0.6 and 1.5 ml for the apical and basolateral reservoirs, respectively, and the transmonolayer hydrostatic pressure gradient was zero. Samples ([1/10] of the corresponding volume of the bathing fluid) of the downstream fluid were taken at 0, 0.5, 1, 2, 3, 4, and 5 h after the addition of F-Alb to upstream fluid. An aliquot of upstream fluid was taken at 5 min and 5 h during the flux experiments to determine upstream F-Alb concentration. An equal volume of fresh Ringer solution was replenished after each sampling to keep the volumes of the bathing fluids constant.
These timed samples were diluted in phosphate (50 mM)-buffered saline (PBS, pH 7.4) to minimize the pH effect on FITC fluorescence. Fluorescence is measured in a fluorometer (Perkin Elmer 650–10S; Norwalk, CT) at excitation and emission wavelengths of 490 and 520 nm, respectively. Fluorescence in these samples was estimated from a calibration curve generated using six serially diluted, known concentrations of F-Alb at pH 7.4. Unidirectional F-Alb fluxes across the monolayers were estimated from the rates of F-Alb appearing in downstream fluid. Net albumin flux is estimated as the difference between the ab and ba unidirectional fluxes at the same [F-Alb]. Kinetic constants of unidirectional fluxes are estimated by nonlinear curve fitting of the observed fluxes to yield the maximal saturable flux (J max) and transport constant (K t) at which concentration the saturable process attains half-maximal saturable flux.
Determination of intact albumin in bathing fluids.
Gel permeation chromatography (GPC) was performed on upstream and downstream fluids that were pooled separately at the end of F-Alb flux experiments. The pooled bathing fluids were concentrated in a rotary evaporator and resuspended with a small volume of deionized water (MilliQ; Millipore). The stock F-Alb solution was similarly treated and analyzed. One hundred fifty microliters of these concentrated samples were injected into a GPC system comprising a dual pump, a controller, and a variable wavelength UV detector (all from Waters, Milford, MA). A Protein Pack 300WS column (Waters) was used with a flow rate of 1 ml/min. The mobile phase was PBS diluted ten times with deionized water. One-milliliter fractions of the eluant were assayed for fluorescence.
Effects of temperature on unidirectional F-Alb fluxes and estimation of activation energies.
We investigated the temperature sensitivity of albumin transport by measuring unidirectional F-Alb fluxes with upstream [F-Alb] of 10 μM at 37, 15, and 4°C. Apparent activation energies (E a) and Q10 for albumin transport processes in the ab and ba directions are estimated from the relation between observed fluxes and experimental temperature.
Effects of unlabeled macromolecules on unidirectional albumin fluxes.
We studied specificity of albumin for the putative albumin transport process by measuring unidirectional fluxes of intact F-Alb at 5-μM upstream concentrations with excess unlabeled macromolecules (e.g., 25 μM each of unlabeled albumin, transferrin, or ovalbumin) in the apical or basolateral upstream fluid. We assessed specificity by comparing intact F-Alb fluxes in the absence and presence of excess unlabeled macromolecules in upstream fluid.
Measurements of F-Alb binding to cell membranes.
Cell monolayers were incubated with various concentrations of F-Alb in apical or basolateral fluid (pH 7.4) for 2 h at 4°C to assess binding characteristics of albumin to cell membranes of alveolar epithelium. After the incubation period, monolayers were washed five times with ice-cold PBS, and the cells on filters were lysed with 1% Triton X-100 in PBS (pH 7.4) at room temperature for 2 h. All cell lysate samples were assayed for fluorescence and quantified using a calibration curve generated with known concentrations of F-Alb (which is treated the same way as cell monolayers). The observed data for steady-state F-Alb binding as a function of [F-Alb] were analyzed for maximal binding sites (Bmax , mol/monolayer) and dissociation constant (K d, mol/l). In one series of experiments, excess (100 μM) unlabeled bovine serum albumin or ovalbumin was included with 10 μM F-Alb in the apical or basolateral incubation medium, and F-Alb binding properties were determined. We assessed specificity by comparing F-Alb binding in the presence and absence of excess unlabeled macromolecules in the respective incubation medium.
Photoaffinity labeling of albumin-binding sites in alveolar epithelial cells.
A photoaffinity labeling procedure (15) was used to identify albumin binding proteins in apical cell membranes of pneumocyte monolayers. Briefly, a heterobifunctional affinity probe, sulfosuccinimidyl-2-(p-azidosalicylamido)-ethyl-1-3′-dithiopropionate (SASD; Pierce, Rockford, IL), was radioiodinated by the Iodogen method and conjugated to native bovine serum albumin (fatty acid-free, crystalline). SASD is photoactivatable by UV light and cleavable by reducing agent. Cell monolayers were kept at 4°C throughout the following procedures, unless noted otherwise. Monolayers were washed gently three times with phosphate-buffered Ringer solution (PBR, Ringer solution whose HEPES was replaced equimolarly with phosphates). Each wash consisted of a 5-min incubation followed by suction to remove the medium. Washed cell monolayers were incubated in the dark with the albumin-specific photoaffinity probe (125I-ASD-Alb, dissolved in PBR) in apical fluid for 2 h, with PBR alone bathing the basolateral side. After incubation, these monolayers were washed gently with PBR three times in the dark and subjected to 5-min photolysis using a 500-watt UV lamp. The photolyzed monolayers were washed with PBR three times, and cells were lysed with a sample buffer for sodium-dodecyl-sulfate polyacrylamide gel electrophoresis. Cell proteins were fractionated using 8% Tris-glycine polyacrylamide gel under reducing conditions. The electrophoresed gel was dried for 2 h in a vacuum dryer, followed by exposure to Kodak X-O-Mat film for 2 days at −80°C. The size of albumin binding proteins in apical cell membranes of pneumocyte monolayers was deduced from the position of the125I-labeled protein relative to the positions of various standard proteins of known molecular masses. Cell monolayers treated as above except for the photolysis step served as negative controls. We studied specificity of albumin binding sites by using the125I-ASD-Alb photoprobe in the presence of excess (20×, molar basis) unlabeled albumin, ovalbumin, or dextran (70 kDa) in apical fluid during the affinity labeling period.
Data are presented as means ± SE (n), wheren is the number of observations. Differences among more than two group means were determined by one-way analyses of variance with modified post hoc Newman-Keuls procedures. Where appropriate, unpaired Student's t-tests were performed to estimate statistical significance between two group means. P < 0.05 is taken as the level of significance.
Figure 1 shows representative chromatographic data obtained for upstream and downstream bathing fluids collected at the end of 5-h flux measurements with upstream [F-Alb] of 1 μM. A single peak corresponding to 100% intact F-Alb was noted for upstream (i.e., donor) fluid collected after 5-h flux experiments. A chromatographic profile of stock F-Alb solution was similar to that for upstream fluid (data not shown). The chromatogram for basolateral and apical downstream fluids collected at the end of 5-h flux experiments showed ∼90% and 70% of total fluorescence was associated with intact F-Alb for basolateral (ab) and apical(ba) downstream fluids, respectively. These data indicate that a majority of F-Alb is translocated intact across the alveolar epithelial cell monolayers at this upstream concentration.
Figure 2 depicts a summary of fractions estimated via GPC for intact F-Alb vs. [F-Alb] in either apical or basolateral upstream fluid. As seen, fraction of intact F-Alb decreased as a function of [F-Alb], falling to ∼50% at 10 μM upstream [F-Alb]. These data suggest that alveolar epithelial cells possess proteolytic capacity for albumin at higher [F-Alb].
On the basis of the information on fractions of intact protein in respective downstream fluids at the end of 5-h flux experiments (Fig.2), we determined unidirectional fluxes of intact albumin at 5 h (Fig. 3). Two processes appear to be responsible for intact albumin fluxes in both directions. At ∼5 μM [F-Alb], both ab and ba fluxes (J ab and J ba) of intact albumin began to saturate (Fig. 3 A). At >10 μM [F-Alb], intact F-Alb fluxes increased linearly with [F-Alb]. This increasing flux at higher [F-Alb] is equal in each direction with an apparent permeability coefficient (P app) of ∼4.9 × 10−10 cm/s. As a result, net absorption of intact F-Alb (Fig. 3 B) exhibits a simple saturable process withK t ∼1.6 μM and J max∼0.15 fmol · cm−2 · s−1.
When the experimental temperature was lowered to 15°C, unidirectional fluxes of intact albumin measured at 10 μM upstream [F-Alb] decreased by 92 and 83% in the ab and badirection, respectively (Fig. 4), collapsing net albumin absorption. Interestingly, the respectiveJ ab or J ba observed at 15°C was not different from that at 4°C, indicating that net absorptive flux also collapsed at this more moderate temperature. Moreover, the fluxes observed at both 15 and 4°C are not significantly different from zero, indicating near cessation of albumin transport across monolayers at these temperatures. ApparentE a for J ab andJ ba of intact albumin between 37 and 15°C are 21.0 and 14.4 kcal/mol (corresponding to Q10 of 3.3 and 2.3), respectively. These observations of a strong temperature dependency of albumin absorption are consistent with albumin transport being primarily mediated by transcellular (but not diffusion-limited) process(es).
The effects of unlabeled excess bovine serum albumin on intact F-Alb fluxes (measured at 5 μM upstream [F-Alb]) are shown in Fig.5. As the concentration of unlabeled albumin increases, commensurate decreases in intact F-Alb fluxes are observed in either ab or ba directions. Half-maximal concentrations for inhibitory effects of unlabeled albumin on intact F-Alb fluxes in ab and ba directions are about the same (∼6 μM). The maximal inhibition of intact F-Alb fluxes due to the presence of unlabeled albumin approaches 100% at 200 μM in either ab or ba directions. By contrast, when fivefold excess unlabeled ovalbumin (40 kDa) or transferrin (70 kDa) was coincubated with 5 μM F-Alb in upstream fluid, no significant inhibition of intact F-Alb fluxes in either direction was found (Fig. 6), indicating the net absorptive transport process for F-Alb is specific for albumin, but not ovalbumin or transferrin.
Alveolar epithelial cell monolayers appear to express binding sites for F-Alb. A Scatchard plot for apical and basolateral binding of F-Alb at 4°C is shown in Fig. 7. An apparentK d of ∼113 nM and an F-Alb Bmax of ∼0.2 pmol/cell monolayer are estimated at apical fluid pH of 7.4, whereas basolateral binding exhibits K d and Bmax of 246 nM and 2.1 pmol/monolayer. One can estimate the number of albumin-binding sites per cell as follows. On days 4–6 in culture, there are ∼2 × 105alveolar epithelial cells present in the monolayer. Thus apical and basolateral albumin-binding sites are 1.0 × 10−18and 1.1 × 10−17 mol/cell, respectively. Assuming a one-to-one interaction between albumin and its cognate binding site, the number of apically and basolaterally expressed albumin-binding sites is therefore 6.0 × 105 and 6.6 × 106 sites/cell, respectively. Density of albumin binding sites in basolateral cell membranes is about tenfold greater than that for apical cell membranes, when estimated at 4°C.
Figure 8 illustrates the effect of tenfold excess unlabeled albumin or ovalbumin on F-Alb binding (measured with 10 μM [F-Alb] in apical or basolateral fluid) to cell membranes. Excess ovalbumin did not significantly alter the magnitude of F-Alb binding to either apical or basolateral membranes of cell monolayers, whereas unlabeled albumin significantly decreased (>90% and ∼60%, respectively) apical and basolateral binding of F-Alb. These data indicate that the putative binding site(s) specifically recognize(s) albumin, but not ovalbumin.
In support of these physiological data on mechanisms of intact albumin transport across the alveolar epithelial barrier, photoaffinity labeling yields an albumin-binding molecule of ∼60 kDa on apical cell membranes, as shown in lane 2 of Fig.9. As a control, lane 1 shows the sample not photolyzed. Excess unlabeled albumin competes successfully for this binding site, as seen in lane 3. By contrast, excess ovalbumin or dextran (70 kDa) results in no changes in the 60-kDa band intensity (data not shown), suggesting that the 60-kDa molecule is specifically interacting with albumin.
We have shown that the transport of intact albumin across the rat alveolar epithelial cell monolayer is asymmetric (i.e., net absorption), highly sensitive to temperature, and saturable. This net absorption of intact albumin appears to occur by a receptor-mediated transcytotic process that involves binding sites of ∼60 kDa, specific for albumin but not other macromolecules. Net saturable albumin absorption mediated by receptors expressed in alveolar epithelial cells is likely to participate in alveolar homeostasis in normal mammalian lungs.
Unidirectional fluxes of intact albumin across alveolar epithelium in both the ab and ba directions consist of asymmetric saturable fluxes at lower upstream [albumin], linearly (and equally) increasing with increasing upstream [albumin]. These data suggest the presence of a net (J ab >J ba) saturable absorption process of intact albumin that is highly likely to be mediated by specific albumin binding proteins, superimposed on which is a symmetric and linear component that is likely nonspecific. The symmetric and linear component(s) of the unidirectional fluxes at higher upstream [albumin] may take place via adsorptive and/or fluid-phase transcytosis, although these symmetric fluxes do not contribute to net absorption of intact albumin. The P app of ∼5 × 10−10 cm/s, compared with a theoretically predicted P app for ∼3.5-nm-radius albumin of ∼6 × 10−9 cm/s across an alveolar epithelial barrier with equivalent pores of ∼6 nm in radius (27), is very small and not compatible with restricted passive diffusion.
The observation of saturable transport of intact albumin at lower upstream [albumin] led us to delineate the receptor-mediated process(es) further. We confirmed by photoaffinity labeling approaches that alveolar epithelial cells express an ∼60-kDa plasma membrane protein that specifically binds to albumin, but not other macromolecules. In this regard, pulmonary endothelial cells have been reported to express albondin [∼60-kDa glycoprotein (gp60)] for unmodified albumin (35), gp18/gp30 for albumins modified chemically (e.g., formaldehyde treatment, gold adsorption) (36), and receptors for heavily glycated end products of albumin (30). Of these endothelial albumin-binding proteins, albondin-like proteins have been reported recently to be expressed in rat lung alveolar epithelium (20). We also reported preliminary data showing that a polyclonal antibody against rat endothelial albondin cross-reacts with a 60-kDa cellular protein in primary cultured rat alveolar epithelial cell monolayers (24). Recently, albondin antibody conjugation has been reported to stimulate albumin internalization into rat alveolar epithelial cells in culture and in situ (20). These reports suggest that the albumin-binding site of 60 kDa found in our study by photoaffinity labeling techniques represents albondin.
The 60-kDa protein observed for albumin binding in alveolar epithelium does not appear to be the same as those identified as renal epithelial albumin binding proteins (e.g., cubilin and megalin) (4). These latter proteins have much greater molecular masses of 460 and 600 kDa, respectively, and are known to mediate internalization of albumin into renal tubular epithelial cells, followed by lysosomal degradation of all internalized albumin (7). No transport of intact albumin has been documented across renal proximal tubules, indicating that renal albumin handling via receptor-mediated endocytosis is mainly for processing of internalized albumin to amino acids and exhibits characteristics different from alveolar epithelial transcytosis of intact albumin.
Significant decreases (i.e., 83 and 92%) in unidirectional fluxes of intact albumin were observed with 10 μM upstream [albumin], when temperature was reduced from 37 to 15 or 4°C. The Q10estimated for unidirectional fluxes of intact albumin in theab and ba directions at 15 and 37°C are 3.3 and 2.3, respectively. It was recently reported that, when 0.05% (∼7.5 μM) unlabeled albumin was instilled together with tracer amount of125I-labeled albumin into distal air spaces of the intact rat lung, Q10 of 2.1 between 27 and 37°C was observed (20). These data are consistent with receptor-mediated albumin transport across alveolar epithelium at low upstream [albumin]. However, identification of mechanisms underlying decreased protein transport at lower temperature is difficult, due in part to currently unknown relationships between temperature and specific elements of the transcytosis pathway (e.g., membrane fluidity).
Albumin fluxes observed at lower upstream [albumin] in this study are rapid and asymmetric. A similar order of magnitude for14C-labeled albumin fluxes was observed in bullfrog alveolar epithelium (23). Compared with the symmetric fluxes exhibited by macromolecules (e.g., 40-kDa HRP and 70-kDa dextrans) known to be translocated via nonspecific endocytosis, saturable albumin fluxes are about one to two orders of magnitude greater (27, 28). By contrast, a paracellular diffusion marker, mannitol, exhibits about one order of magnitude greater permeability than 14C-labeled albumin (27).
Absorption of albumin across various pulmonary epithelial barriers has been reported previously (9, 21, 23). We found net absorption of albumin across the bullfrog alveolar epithelium, showing that the downstream fluid contains intact albumin and that albumin transport is not dependent on an electrical gradient imposed across the alveolar epithelium (23). Net absorption of labeled proteins (e.g., albumin and ovalbumin) across primary cultured monolayers of guinea pig tracheal epithelial cells has been shown, although such protein transport appears to take place via adsorptive endocytosis without requiring specific receptors (9). Radiolabeled albumin instilled into bronchial luminal fluid was reported to be exhaustively processed in bronchial epithelial cells, resulting in the appearance of smaller radiolabeled species (e.g., amino acids) in downstream fluid when excised canine bronchial tissue was studied in Ussing chambers (21).
The saturable process(es) for albumin translocation in bothab and ba directions is (are) likely to be mediated by binding sites specific for albumin. The degree of competition by unlabeled albumin for these apically and basolaterally localized specific binding sites is about the same at each surface. The twofold greater K d for albumin binding to basolateral vs. apical membranes is somewhat unexpected sinceJ ab exceeds J ba, suggesting that internalization and transcellular movement of albumin must be considerably more rapid (and/or more intact) in theab compared with the ba direction. For example, albumin internalized via basolateral binding sites may (to a greater extent than via apical binding sites) be directed to a degradative (rather than transcytotic) pathway. Whether or not the differentK d reflect different albumin binding sites also remains to be determined.
In summary, saturable process(es) for trafficking of intact albumin across the alveolar epithelial barrier have been demonstrated. Net absorption of intact albumin across alveolar epithelium appears to take place via transcytosis mediated by ∼60-kDa proteins expressed in alveolar epithelial cell membranes. Net saturable absorption of intact albumin across the respiratory epithelial tract lining the distal air spaces may play an important role in alveolar homeostasis in normal mammalian lungs. As a corollary, future strategies and design of protein drugs for systemic delivery via pulmonary routes may benefit from utilization of alveolar epithelial transcytosis pathways mediated by specific receptors.
The authors appreciate the technical assistance of Dr. Vinod Rattan [Department of Biochemistry, University of Southern California (USC)] on photoaffinity labeling approaches and many helpful discussions with Dr. Zea Borok (Department of Medicine, USC).
↵* K.-J. Kim and Y. Matsukawa contributed equally to this work.
This work was supported in part by American Heart Association-National Center Grant-in-Aid 9950442N; National Institutes of Health Grants GM-59297, HL-38578, HL-38621, HL-38658, and HL-64365; and the Hastings Foundation. V. H. L. Lee is Herbert Gavin Professor of Pharmaceutical Sciences. E. D. Crandall is Hastings Professor of Medicine and K. T. Norris Jr. Chair of Medicine.
Present address for Y. Matsukawa and H. Yamahara: Tanabe Seiyaku Co., Ltd., Discovery Research Laboratory, 16-89, Kashima 3-chome, Yodogawa-ku, Osaka 532-8505, Japan.
Address for reprint requests and other correspondence: K.-J. Kim, Rm. HMR 914, Dept. of Medicine, USC Keck School of Medicine, 2011 Zonal Ave., Los Angeles, CA 90033 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published November 15, 2002;10.1152/ajplung.00237.2002
- Copyright © 2003 the American Physiological Society