Pseudomonas Invasion of Type I Pneumocytes Is Dependent on the Expression and Phosphorylation of Caveolin-2*

Pseudomonas aeruginosa is a major cause of pneumonia in patients with cystic fibrosis and other im-muncompromising conditions. Here we showed that P. aeruginosa invades type I pneumocytes via a lipid raft-mediated mechanism. P. aeruginosa invasion of rat primary type I-like pneumocytes as well as a murine lung epithelial cell line 12 (MLE-12) is inhibited by drugs that remove membrane cholesterol and disrupt lipid rafts. Confocal microscopy demonstrated co-lo-calization of intracellular P. aeruginosa with lipid raft components including caveolin-1 and -2. We generated caveolin-1 and -2 knockdowns in MLE-12 cells by using RNA interference techniques. Decreased expression of caveolin-2 significantly impaired the ability of P. aeruginosa to invade MLE-12 cells. In addition, the lipid raft-dependent tyrosine phosphorylation of caveolin-2 appeared to be a critical regulator of P. aeruginosa invasion. ). Most interestingly, the combination of OA and genistein had no effect on P. aeruginosa invasion. Pretreatment with methyl- (cid:2) -cyclodextrin (3 m M ) for 30 min prior to P. aeruginosa infection in the presence of OA negated the previously noted increased invasion associated with OA. Neither OA nor genistein affected the invasion of L. monocytogenes . B , OA and genistein had no effect on the adherence of P. aeruginosa. C , immunoprecipitation of caveolin-2 was performed to examine the effects of P. aeruginosa infection on the tyrosine phosphorylation of caveolin-2. The bar graph shows the densitometry measured from Western blots obtained from three separate experiments. P. aeruginosa infection significantly increases cav-2 tyrosine phosphorylation. Treatment with genistein before and during P. aeruginosa infection decreased caveolin-2 phosphorylation. On the other hand, Cav-2 tyrosine phosphorylation was significantly increased by infection in the presence of OA. Finally, tyrosine phosphorylation of caveolin-2 is dependent on lipid raft integrity as pretreatment with MCD prior to P. aeruginosa infection with OA negated the effects of OA (data are presented as mean (cid:7) S.E., *, p (cid:8) 0.05).

Pseudomonas aeruginosa is an important cause of nosocomial pneumonia, as well as a major pulmonary pathogen in patients with cystic fibrosis and other immunocompromising conditions (1)(2)(3)(4). A large number of patients develop P. aeruginosa colonization of the upper airways; however, only a small percentage of these patients develop clinically significant P. aeruginosa pneumonia. The morbidity and mortality of P. aeruginosa respiratory infections usually result from the dissemination of P. aeruginosa to the alveolar space prior to the establishment of pneumonia (2,5,6). However, the pathogenesis of P. aeruginosa pneumonia is not completely understood.
The alveolar epithelium is the largest host epithelial surface exposed to the external environment with an area roughly equal to the size of a tennis court. Approximately 95% of that surface area is lined by specialized type I pneumocytes (7). Once bacteria escape the mucociliary host defenses of the upper airways and enter the alveolar space, they are likely to come in contact with type I pneumocytes. The role of type I pneumocytes in the pathogenesis of P. aeruginosa pneumonia is not well understood. Type I pneumocytes were initially believed to serve only simple barrier functions separating the alveolar space from the pulmonary capillaries. More recently, type I cells have been implicated in a wide range of functions including lipid metabolism, cell signaling, remodeling, and host defense (8 -10). The cell membrane of type I pneumocytes has a high concentration of specialized lipid rafts and caveolae, which occupy nearly 70% of the plasma membrane (11,12). Invasion of type 1 pneumocytes would protect P. aeruginosa from phagocytosis by alveolar macrophages and offer a protected environment for replication. P. aeruginosa invasion of type I cells may facilitate dissemination throughout the host because of the single cell thickness of the alveolar epithelium. We hypothesized that the virulence of P. aeruginosa in the development of pneumonia is at least due in part to its ability to invade type I pneumocytes.
P. aeruginosa is able to invade nasal and bronchial epithelial cells via a lipid raft-dependent mechanism (13,14). Lipid rafts are specialized areas of the plasma membrane that are detergent-insoluble, low density membrane fractions that are enriched in cholesterol and sphingolipids (15). In addition, these regions contain a number of important signaling molecules and structural proteins such as caveolins (16). Lipid rafts have been implicated in a wide range of cellular functions, including lipid metabolism (17), cell signaling (18), and endocytosis (19). The pathway of lipid raft-dependent endocytosis is distinguished from other clathrin-mediated endocytosis and other endocytic pathways by its sensitivity to cholesterol depletion (20). An increasing number of pathogens have been recognized to co-opt the mechanism of lipid raft-mediated endocytosis in order to invade host cells (21)(22)(23)(24). P. aeruginosa invasion of nasal and bronchial epithelial cells induces apoptosis and shedding of the superficial layer epithelial cells (13). P. aeruginosa may co-opt lipid raft-mediated endocytosis to invade the alveolar epithelium during the pathogenesis of P. aeruginosa pneumonia. However, the single cell thickness of the alveolar epithelium may lead to markedly different consequences for the host.
Caveolin proteins are key components of lipid rafts and caveolae (16). Caveolin proteins have many important functions. In addition to serving as key structural proteins that organize caveolae platforms, caveolin proteins are important in regulating endocytosis and cell signaling (16,25,26). The caveolin family consists of three 21-24-kDa integral membrane proteins (27). Caveolin-1 and -2 are co-expressed on most cell types, including type I pneumocytes (26,28). A majority of the research on caveolin proteins have focused on the functions of caveolin-1. We have shown previously that caveolin-1 expression is an important determinant in the uptake of Escherichia coli by bladder epithelial cells (29). Although caveolin-1 has been extensively studied, very little is known about the function of caveolin-2. Caveolin-2 interacts with caveolin-1 to form a hetero-oligomeric complex within lipid rafts (30,31). Unlike caveolin-1, caveolin-2 is not required for the formation of caveo-lae and has not been implicated previously in cell signaling (32). Caveolin-1 is required as a chaperone to transport caveolin-2 to the cell surface (33). In the absence of caveolin-1, caveolin-2 is degraded, and its expression is markedly decreased (34). Thus, it is possible that many of the functions previously attributed to caveolin-1 may be due to caveolin-2.
In this paper we demonstrate that P. aeruginosa invades type I pneumocytes via a lipid raft-dependent mechanism. In addition, P. aeruginosa invasion of alveolar epithelial cells is dependent on expression and tyrosine phosphorylation of caveolin-2.

MATERIALS AND METHODS
Animals-Male Sprague-Dawley rats weighing 150 -200 g were obtained from Taconic Laboratories.
Type II Cell Isolation-Isolation of primary rat type II cells was accomplished by using methods published previously (35). Approximately 20 -30 million type II cells were isolated per rat with Ͼ90% viability as assessed by trypan blue exclusion. Purity was greater than 85% type II cells as confirmed by Papanicolaou staining. Cells were seeded on tissue culture plastic plates at a density of 6 ϫ 10 5 cells/cm 2 and maintained in Dulbecco's modified Eagle's medium (Invitrogen) with dexamethasone and 10% fetal bovine serum (Hyclone). At 36 h cell cultures were washed three times to remove nonadherent cells and to enrich the percentage of type II cells. All experiments were performed on days 6 -7 to allow differentiation into the type I-like phenotype (35).
Bacterial Strains and Cell Lines-P. aeruginosa strain 27853 and strain PAO-1 were obtained from the American Type Culture Collection (ATCC, Manassas, VA). P. aeruginosa strains were grown in 5 ml of static LB broth for 18 h. Listeria monocytogenes was grown for 18 h in 5 ml of brain heart infusion broth (BD Biosciences). After static incubation for 18 h, bacterial cultures were diluted 1:50 and placed on shaker at 37°C for 2-3 h until they reached mid-log growth phase with an A 600 0.4 -0.8.
Creation of Caveolin Knockdowns Using RNA Interference-RNA interference vectors were generated using pQCXIN retroviral vector (BD Biosciences). Briefly, pQCXIN was digested by BamHI and EcoRI and then was religated to generate pQCXIN1. Human U6 small nuclear RNA promoter was PCR-amplified from pTZ U6 ϩ 1 (gift from John Rossi, Beckman Research Institute of the City of Hope, Duarte, CA) with added BglII site (5Ј ends), BamHI, and XbaI sites (3Ј ends). The PCR product was cloned to the BglII and XbaI sites of pQCXIN1 to generate pOCXIN-U6. The following oligonucleotides were ordered from Integrated DNA Technologies, Inc. Coralville, IA: C1a, 5Ј-GATC-CGGAGATTGACCTGGTCAACTTCAAGAGAGTTGACCAGGTCAA-TCTCCTTTTTT-3Ј, and C1b, 5Ј-CTAGAAAAAAGGAGATTGACCTG-GTCAACTCTCTTGAAGTTGACCAGGTCAATCTCCG-3Ј; C2a, 5Ј-GA-TCCGGCCGATGTGCAGCTCTTCTTCAAGAGAGAAGAGCTGCAC-ATCGGCCTTTTTT-3Ј, and C2b, 5Ј-CTAGAAAAAAGGCCGATGTGC-AGCTCTTCTCTCTTGAAGAAGAGCTGCACATCGGCCG-3Ј. The boldface and underlined sequences are forward and reverse sequences, respectively, which correspond to nucleotides 221-239 of the mouse caveolin-1 gene (C1a and C1b, GenBank TM accession number NM_007616) and nucleotides 41-59 of the mouse caveolin-2 gene (C2a and C2b, GenBank TM accession number NM_016900). C1a and C1b and C2a and C2b were annealed to form double-stranded DNA and cloned into BamHI and XbaI sites of pQCXIN-U6 to generate pSi-Cav1and pSi-Cav2. Amphopack-293 Cell Line (BD Biosciences) was used to produce the viral particles. Production of viral particles, infection of target cell line (MLE-12), and selection of viral infected cells were performed as recommended by the vendor of the pQCXIN vector (BD Biosciences). The neomycin-resistant stable-transfected cell lines were named pQCcav-1 and pQC-cav-2. In the same way, oligonucleotides Lu1 5Ј-GATC-CGTACGCGGAATACTTCGAATTCAAGAGATTCGAAGTATTCCG-CGTACTTTTTT-3Ј and Lu1a 5Ј-CTAGAAAAAAGTACGCGGAATAC-TTCGAATCTCTTGAATTCGAAGTATTCCGCGTACG-3Ј (the forward and reverse sequences) correspond to nucleotides 423-441 of the firefly luciferase in the pGL2-control vector. GenBank TM accession number X65324 was used to generate the control cell line pQC-LUC.
Rat Model of P. aeruginosa Pneumonia-150-g Sprague-Dawley rats were anesthetized with halothane and then orotracheally intubated. ATCC 27853 bacteria were grown to mid-log phase and labeled with PKH2 dye (Sigma) per the manufacturer's protocol. P10 tubing was inserted into the left lower lobe, and 350 l of bacteria in PBS with 1% bovine serum albumin was instilled. Animals were sacrificed after 90 min. The lungs were lavaged 10 times with PBS/EDTA to remove any nonadherent bacteria. The lungs were inflated to total lung capacity with optimal cutting temperature (OCT) medium (Miles Scientific) and frozen at Ϫ80°C. Ten-micron frozen sections were cut using a cryostat (Leica). Sections were fixed in 100% acetone at Ϫ20°C for 5 min prior to labeling for confocal microscopy. They were then washed with PBS prior to blocking with 10% goat serum in PBS for 1 h. The following primary antibodies were diluted 1:100 in 10% goat serum and added to the sections for 1 h; rabbit polyclonal caveolin-1 (BD Biosciences), rabbit polyclonal caveolin-2 (Abcam), and RT140 (courtesy of Leland Dobbs, University of California, San Francisco). Cells were then washed three times with 10% goat serum. Goat anti-rabbit or goat anti-mouse secondary antibodies conjugated to Alexa Fluor 660 (Molecular Probes) were diluted 1:250 and added to cells for 1 h. Cells were washed three times with 10% goat serum and mounted using Prolong Gold Anti-fade (Molecular Probes) for examination by confocal microscopy.
Confocal Microscopy-Primary rat type II cells were grown on Per-manox© plastic chamber slides for 6 days. MLE-12 cells were seeded onto collagen coated 12-mm diameter glass coverslips placed into the wells of a 24-well plate and grown for 36 h. The cells were infected with ATCC 27853 strain of P. aeruginosa conjugated to the amine reactive Alexa Fluor 660 (Molecular Probes). The infected cells were incubated at 37°C for 45 min, washed four times with PBS (Invitrogen) to remove unbound bacteria, and fixed overnight in 2.0% paraformaldehyde in PBS. To examine caveolin-1 and caveolin-2 labeling, cells were grown and infected as above. After removing the fixative, the cells were permeabilized with 0.1% Triton X-100 (Sigma). Caveolin-1 was then labeled using polyclonal anti-caveolin-1 antibody (BD Biosciences); caveolin-2 was labeled with mouse polyclonal caveolin-2 antibody (Abcam), and GM1 was labeled using a biotinylated cholera toxin. Type I cells labeled with RT140 antibody (1:500). Alexa 488 secondary antibodies (Molecular Probes) were used for each experiment. Coverslips were examined using a Nikon Eclipse TE200 microscope. Confocal microscopy was also used to quantify the percentage of intracellular bacteria that co-localize with caveolin proteins. Co-localization was determined by identifying bacteria within the center of cells that co-localize with intracellular vesicle staining for caveolin-1 or -2. Twenty five intracellular bacteria per slide were counted for analysis.
Electron Microscopy-Rat type II cells were isolated as described above and cultured on Thermanox© plastic coverslips for 6 days. Cells were infected with P. aeruginosa for 45 min. Cells were then washed four times with PBS and fixed with glutaraldehyde, dehydrated with cold acetone, and processed using standard methods. The cells were examined by transmission electron microscopy.
Bacterial Invasion and Adherence Assays-Type I-like pneumocytes or MLE-12 cells were seeded into 96-well plates. Type I-like cells were seeded at a density of 6 ϫ 10 5 cells/cm 2 . MLE-12 cells were grown to ϳ80% confluency for all experiments. For infections with strains of P. aeruginosa, cells were infected at a multiplicity of infection (m.o.i.) of 300 -600 bacteria per host cell by the addition of 100 l of bacteria diluted in serum-free cell culture medium containing 10 mg/ml bovine serum albumin (Sigma), A 600 1.0. Due to the increased cytotoxicity associated with L. monocytogenes, cells were infected at a m.o.i. 50 -100, A 600 0.1. Plates were then incubated at 37°C for 45 min. The medium was replaced with fresh culture medium containing 50 g/ml of the membrane-impermeable antibiotic gentamicin (Invitrogen) to kill extracellular bacteria and incubated for 45 min. Each well was washed three times with PBS. In order to remove and lyse the cells, 30 l of 0.25% trypsin (Invitrogen) was added to each well for 5 min, and 70 l of 0.1% Triton X-100 in PBS was then added. Cells were scraped and transferred to 0.65-ml tubes, and each well was washed with an additional 100 l of Triton that was also transferred to the tubes. The cell lysates were diluted and plated onto LB agar plates for Pseudomonas strains or brain heart infusion agar plates for L. monocytogenes. In order to quantify the number of adherent bacteria, the cells were exposed to P. aeruginosa for 45 min at 4°C, which inhibits endocytosis. They were then washed five times with PBS to remove nonadherent bacteria, serially diluted, and plated on LB agar, and the number of colony-forming units were counted to quantify adherent bacteria.
To test the effect of pretreatment by specific lipid rafts disrupters/ usurpers on bacterial invasion, methyl-␤-cyclodextrin (MCD, 3 mM, Sigma), nystatin (20 g/ml, Sigma), or filipin (0.5 g/ml, Sigma) in serum-free medium was added to the cells for 30 min and then washed off prior to infection. To reinsert cholesterol back into the plasma membrane after MCD treatment, cells were incubated for 1 h with a cholesterol-cyclodextrin complex (0.2 mM) (36). To test the effect of altering tyrosine phosphorylation on bacterial invasion, the cells were pretreated with genistein (100 g/ml, Sigma) for 30 min prior to infection and throughout the period of infection. The effects of okadaic acid (OA, 10 nM, Sigma) were examined by having OA present throughout the infection without any pretreatment. After the preincubation step, the cells were washed, and the medium was replaced with bacteria in serum-free medium with or without the agent being tested as required. A gentamicin protection assay was performed as detailed above to quantify intracellular bacteria. The viability of the cells was not affected by any of the treatments used as determined by trypan blue exclusion.
An MTT adherence assay was performed as follows. Cells were plated 96-well plates and then incubated with drugs for 30 min as stated above and then fixed overnight with 2.0% paraformaldehyde. The monolayers were washed three times with sterile PBS and pretreated for 1 h at room temperature with blocking buffer (3% bovine serum albumin in PBS). 100 l of P. aeruginosa, ATCC 27853 (A 600 1.0) in PBS, was incubated with cells for 1 h at 37°C. Nonadherent bacteria were removed by washing the cell monolayers three times with PBS. Fifty microliters of LB was applied to each monolayer and incubated for 15 min at 37°C. Fifty microliters of 2 mg/ml MTT (Sigma) was added, and the plates were incubated for 30 min at 37°C to allow reduction of MTT to formazan by live bacteria. Next, 150 l of isopropyl alcohol and hydrochloric acid (24:1) was added to solubilize the formazan, and the absorbance was measured at 450 nm using a Tecan Sunrise remote microplate reader.
Western Blotting and Immunoprecipitation-Cells were grown in a 6-well plate to greater than 90% confluency. Cells were lysed with lysis buffer (1:10, Upstate Biotechnology, Inc.), phenylmethylsulfonyl fluoride (1 mM), and protease inhibitor mixture (1:100, Sigma). Total protein was quantified using Bradford dye (Bio-Rad) and measured at A 595 . Cell lysates were normalized for protein and volume. Samples were diluted in Laemmli sample buffer (Bio-Rad) with mercaptoethanol and boiled for 5 min. Proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes. Western blotting was then performed using antibodies against tubulin (1:2000, Sigma), caveolin-1 (1:500, BD Biosciences), or caveolin-2 (1:500, BD Biosciences), and the proteins were detected using Super Signal West Pico Chemiluminescent kit (Pierce).
Caveolin-1 and caveolin-2 were immunoprecipitated from whole cell lysates of MLE-12 cells as follows. MLE-12 cells were grown to 90% confluency in 6-well plates. The cells were washed twice with cold PBS. The cells were then lysed in 500 l of lysis buffer (lysis buffer (Upstate Biotechnology, Inc., 1:10), n-octyl-␤-D-glucopyranoside (60 mM, Sigma), phenylmethylsulfonyl fluoride (1 mM), protease inhibitor, phosphatase mixture (Sigma, 1:100)). The sample was precleared with 30 l of a 50% suspension of protein G-Sepharose (Sigma) in PBS for 1 h at 4°C. The samples were then incubated for 16 h at 4°C with 50 l of a 50% suspension of protein G-Sepharose in PBS plus 8 g of rabbit polyclonal anti-caveolin-1 (BD Biosciences) or 6 g of rabbit polyclonal caveolin-2 (Abcam) was added. The immunoprecipitates were washed three times with 1 ml of PBS and resuspended in 50 l of Laemmli sample buffer (Bio-Rad) with mercaptoethanol and boiled for 5 min. Immunoprecipitates were assayed with antibodies for caveolin-1 (BD Biosciences), caveolin-2 (BD Biosciences and Abcam), or phosphotyrosine (Upstate Biotechnology, Inc. 4G10) by Western blotting as described above. In addition, Image J software (National Institutes of Health) was used for densitometry to quantify protein expression for statistical analysis.
Statistics-Data were compiled and analyzed using Microsoft Ex-cel©. Data for gentamicin protection assays are reported as a mean of at least three experiments with error bars set as S.E. Significance was calculated using a two-tailed t test with significance defined and a p value of less than 0.05. Graphs were created using Graph Pad© Prism 3.0.

RESULTS
P. aeruginosa Invades Type I Pneumocytes during the Pathogenesis of Pneumonia-Although P. aeruginosa frequently col-onizes the upper airways of susceptible hosts, the development of P. aeruginosa pneumonia requires dissemination to the alveolar space prior to the development of epithelial cell damage and alveolar filling. We sought to determine whether P. aeruginosa invades type I pneumocytes during the pathogenesis of pneumonia by using a rat model of P. aeruginosa pneumonia. Male Sprague-Dawley rats were intubated, and fluorescently labeled P. aeruginosa was instilled directly into the left lower lobe. Animals were sacrificed 90 min after instillation to examine the early stages of pneumonia prior to the development of severe lung injury. We examined frozen thin sections of rat lung to examine the location of P. aeruginosa within alveolar spaces. In Fig. 1A, the type I pneumocytes are labeled red with the type I cell-specific antibody, RT140. The yellow staining demonstrates areas of co-localization where P. aeruginosa is directly adherent to and possibly within type I pneumocytes. Thus, type I pneumocytes appear to be a main site of P. aeruginosa infection in a rat model of P. aeruginosa pneumonia.
P. aeruginosa Invades Type I-like Pneumocytes in an in Vitro Model of the Alveolar Epithelium-We used a common in vitro model of primary type I-like pneumocytes (35) to determine whether P. aeruginosa is able to invade type I cells. Type II cells are easily isolated and known to be the progenitor cells of the alveolar epithelium. When cultured on plastic tissue culture surface for 5-7 days, the type II cells differentiate into type I-like cells with features typical of primary type I cells.
Type I-like cells have a similar membrane lipid content to primary type I pneumocytes (35) and also express type I cellspecific proteins such as aquaporin-5 (8), caveolin-1 (11,12), and RT140 (37). This model is frequently used in vitro to study the behavior of type I cells for studies of fluid homeostasis (38), signal transduction (39), and bacterial pathogenesis (40). Western blots of whole cell lysates confirmed that after culture for 6 days, type I-like cells express type I cell-specific proteins including caveolin-1 and RT140 (data not shown).
Type I-like pneumocytes were exposed to fluorescently labeled P. aeruginosa and examined using confocal microscopy. Fig. 1B shows intracellular P. aeruginosa within the center of type I-like pneumocytes. Next, we used transmission electron microscopy (TEM) to confirm the intracellular location of P. aeruginosa within type I-like pneumocytes. P. aeruginosa was identified within type I-like cells as shown in Fig. 1C. The bacteria appeared intact and were located within vacuolar membranes. We next sought to quantify the number of adherent and intracellular P. aeruginosa after exposure to type I-like cells in vitro. After exposure to P. aeruginosa at a m.o.i. of 500, 8.6% of P. aeruginosa are able to adhere to the cell surface, and 4.5% of adherent P. aeruginosa invade type I-like cells (Fig.  1D). Thus, it appears that during the initial stages of pneumonia, P. aeruginosa is able to enter the alveolar space and adhere to and invade type I pneumocytes.
MLE-12 Cells Are Good Models for P. aeruginosa Invasion of Type I Pneumocytes-Primary type I-like cells have several limitations that make them unsuitable for studying the molecular mechanisms of P. aeruginosa invasion. These primary cells are terminally differentiated and cannot be transfected to examine the role of specific proteins in P. aeruginosa invasion. Therefore, we chose to use a murine lung epithelial cell line (MLE-12) as a model for P. aeruginosa invasion of type I cells. MLE-12 cells were initially described as a model of type II pneumocytes because of their capacity to produce surfactant proteins (41). However, they are likely immortalized in the process of differentiation into type I pneumocytes, and they possess many type I cell-specific features. MLE-12 cells lack lamellar bodies that are characteristics of type II cells (41), have visible caveolae on their cell surface typical of type I cells (39), and also have high levels of type I cell-specific proteins including aquaporin-5 (42) and caveolin-1 (39). MLE-12 cells were exposed to fluorescently labeled P. aeruginosa and examined with confocal microscopy. Similar to type I-like cells, P. aeruginosa was clearly visualized intracellularly (data not shown). In addition, MLE-12 cells were exposed to P. aeruginosa, and the number of adherent and intracellular bacteria was quantified (Fig. 1E). The frequency of adherence and invasion was nearly identical between type I-like cells and MLE-12 cells. Therefore, MLE-12 cells are a good model for invasion of type I pneumocytes and can be used for future molecular studies to examine in detail the mechanism of P. aeruginosa invasion.
P. aeruginosa Invasion of Type I-like Cells Is Dependent on Lipid Rafts-We next sought to determine the mechanism by which P. aeruginosa invades type I-like pneumocytes. To determine whether P. aeruginosa invasion of type I-like pneumocytes occurs via lipid rafts, the cells were pretreated with cholesterol-disrupting agents prior to quantifying invasion with a gentamicin protection assay. Drugs that disrupt membrane cholesterol are known to specifically inhibit lipid raftmediated endocytosis (20). Fig. 2A shows that P. aeruginosa invasion of type I-like cells is significantly inhibited by low doses of nystatin, filipin, and cyclodextrin. To show that the inhibition of P. aeruginosa invasion was due to lipid raft disruption and not any other unintended effect of these drugs, cells were treated with cyclodextrin to remove membrane cholesterol, and the cholesterol was then reinserted back into the plasma membrane by using a cholesterol-cyclodextrin complex (36). As shown in Fig. 2A, reinsertion of cholesterol back into the cell membrane restores the ability of P. aeruginosa to invade type I-like cells. Although disruption of lipid rafts impairs the entry of P. aeruginosa into type I-like cells, it has no effect on the invasion of type I-like cells by L. monocytogenes, which is known to invade host cells in a lipid raft independent manner (43) (Fig. 2A). To ensure that the decreased invasion was not due to decreased bacterial adherence to the cell surface, an adherence assay was performed to quantify extracellular adherent bacteria. As shown in Fig. 2B, pretreatment with nystatin, filipin, cyclodextrin, or the cholesterol-cyclodextrin complex had no effect on bacterial adherence to type I-like cells. Similar to the invasion of type I-like pneumocytes, P. aeruginosa invades MLE-12 cells in a lipid raft-dependent manner (Fig. 2C). Finally, the disruption of lipid rafts does not alter the adherence of P. aeruginosa to the MLE-12 cells (Fig.  2D). Because the majority of experiments are performed with the ATCC 27853 isolate of P. aeruginosa, we wanted to ensure that lipid raft-dependent invasion was not limited to this strain. PAO-1, a common lab isolate of P. aeruginosa, was also found to invade MLE-12 cells in a lipid raft-dependent manner (data not shown). Therefore, removal of membrane cholesterol to inhibit lipid raft-mediated endocytosis causes a significant reduction in the number of intracellular P. aeruginosa after infection of both type I-like cells and MLE-12 cells.
Intracellular P. aeruginosa Co-localizes with Lipid Raft Components in Type I-like Cells-A number of proteins and glycolipids are predominantly located within lipid rafts including the caveolin proteins, GM1, and flotillin. Confocal microscopy of permeabilized type I-like cells was used to determine whether the vacuolar membranes that surrounded intracellu-

FIG. 1. P. aeruginosa is able to invade type I pneumocytes.
A, rat lung frozen sections were examined after intratracheal instillation of fluorescently labeled P. aeruginosa. Confocal microscopy shows labeled P. aeruginosa (red) that co-localizes with type I pneumocytes (green), which are labeled with the type I cell-specific antibody, RT 140. B, rat primary type II cells were isolated and cultured for 6 days to differentiate into type I-like pneumocytes. Cells were labeled with the type I cell-specific antibody, RT140 (green). These cells were infected with fluorescently labeled P. aeruginosa (red) in vitro and examined using confocal microscopy. Sections through the center of type I-like cells demonstrate intracellular P. aeruginosa within type I-like cells. C, type I-like pneumocytes infected with P. aeruginosa were examined with TEM. TEM shows intracellular P. aeruginosa within vacuolar membranes of type I-like cells (arrow). The P. aeruginosa visualized here appear intact and to be in the process of cell division. D, absolute numbers of adherent and intracellular P. aeruginosa (Pa) after exposure to type I-like cells are shown from a representative experiment. After 45 min, 8.6% of the added P. aeruginosa adhere to type I-like cells, and 4.5% of adherent P. aeruginosa are able to invade. E, MLE-12 cells were chosen as a cell line model of type I cells in order to further examine the molecular mechanisms of P. aeruginosa (Pa) invasion. Absolute numbers of adherent and intracellular bacteria after exposure to MLE-12 cells are shown from a representative experiment. Similar to type I-like cells, 11.9% of added P. aeruginosa adhere to MLE-12 cells, and 2.9% of adherent P. aeruginosa invade (data are presented as mean Ϯ S.E.). lar P. aeruginosa were enriched in lipid raft components. Fig. 3 shows that fluorescently labeled P. aeruginosa co-localizes with GM1, caveolin-1, and caveolin-2. Similar co-localization of intracellular P. aeruginosa with lipid raft components was also observed in MLE-12 cells (data not shown). The co-localization of P. aeruginosa with lipid raft components such as caveolin proteins led us to question what role these proteins play in the lipid raft-mediated uptake of P. aeruginosa.
The Expression of Caveolin Proteins Is Required for Optimal P. aeruginosa Invasion of MLE-12 Cells-Because type I pneumocytes express high levels of both caveolin-1 and 2 (Fig. 4, A  and B), we sought to determine whether caveolin-1 and -2 play a role in the lipid raft-mediated uptake of P. aeruginosa. Caveolin-1 has been identified as a determinant of lipid raft-mediated uptake of E. coli by bladder epithelial cells (29). The role of caveolin-2 in bacterial invasion has not been studied previously. MLE-12 cells were used to create stable knockdowns of caveolin-1 (pQC-cav-1) and caveolin-2 (pQC-cav-2) by using RNA interference. Western blotting after several passages confirmed the generation of stable knockdowns of both caveolin proteins (Fig. 4C). As a control, MLE-12 cells were infected with the viral vector containing a 19-amino acid sequence unique to the luciferase gene that did not alter caveolin expression (data not shown). pQC-cav-1 cells have not only greater than 60% decreased expression of caveolin-1 but also have significantly decreased expression of caveolin-2 (Fig. 4C). These findings are consistent with previous studies (33) showing that caveolin-1 is required to transport caveolin-2 to the plasma membrane. In the absence of caveolin-1, caveolin-2 is degraded within the Golgi apparatus and not expressed on the cell membrane (33). On the other hand, pQC-cav-2 cells have greater than 80% decreased expression of caveolin-2 and normal levels of caveolin-1 (Fig. 4C).
The adherence and invasion of P. aeruginosa after exposure

FIG. 2. P. aeruginosa invasion of type I-like pneumocytes is dependent on lipid rafts.
A, type I-like cells were exposed to P. aeruginosa for 45 min in vitro. A gentamicin protection assay was used to quantify intracellular P. aeruginosa. In order to disrupt lipid rafts, cells were pretreated for 30 min with drugs known to remove membrane cholesterol including methyl-␤-cyclodextrin (MCD, 3 mM), filipin (0.5 g/ml), or nystatin (20 g/ml). Removal of membrane cholesterol prior to exposure to P. aeruginosa significantly decreased the number of intracellular bacteria. After MCD treatment, cholesterol was reinserted back into the cell membrane using a cholesterol-cyclodextrin complex (chol, 0.2 M), and P. aeruginosa invasion returned back to base-line levels. On the other hand treatment with MCD had no effect on L. monocytogenes invasion. B, MTT adherence assay was used to determine whether disruption of lipid rafts altered P. aeruginosa adherence to type I-like cells. Neither filipin, nystatin, MCD, or cholesterol affected the adherence of P. aeruginosa to type I-like cells. C, P. aeruginosa invasion of MLE-12 cells is also dependent on lipid rafts similar to type I-like cells. The number of intracellular P. aeruginosa was significantly decreased by pretreatment with filipin, nystatin, or MCD. Restoration of membrane cholesterol normalized invasion. In addition, pretreatment with MCD had no effect on L. monocytogenes invasion. D, disruption of lipid rafts has no effect on the adherence of P. aeruginosa to MLE-12 cells (data are presented as mean Ϯ S.E., *, p Ͻ 0.05).
to pQC-LUC cells was quantified and found to be similar to the behavior of type I-like cells and MLE-12 cells (Fig. 4D). Next, P. aeruginosa invasion of pQC-cav-1 and pQC-cav-2 was quantified to determine the possible role of caveolin proteins in P. aeruginosa invasion. pQC-cav-1 cells with decreased expression of both caveolin-1 and -2 were resistant to P. aeruginosa invasion with 50% less intracellular P. aeruginosa after infection (Fig. 4E). In order to determine whether the inhibition of P. aeruginosa invasion was because of decreased expression of caveolin-1 or -2, we then examined P. aeruginosa invasion in the selective caveolin-2 knockdown (pQC-cav-2). After infection with P. aeruginosa, pQC-cav-2 cells had 50% fewer intracellular bacteria similar to pQC-cav-1 cells with decreased expression of caveolin-1 and -2 (Fig. 4E). Down-regulation of caveolin-1 and -2 has no effect on P. aeruginosa adherence (Fig. 4F) or the invasion of L. monocytogenes (Fig. 4E). Thus, it appears that caveolin-2 expression is required for optimal P. aeruginosa invasion of MLE-12 cells.
P. aeruginosa Invasion of MLE-12 Cells Is Dependent on Tyrosine Phosphorylation of Caveolin-2-Because caveolin-2 expression is required for maximal P. aeruginosa invasion, we next sought to determine the mechanism by which caveolin-2 may be involved in controlling lipid raft-mediated endocytosis. Caveolin-1 and -2 have been recognized to have important sites for tyrosine phosphorylation, and several pathogens are known to influence host cell signaling pathways involving tyrosine phosphorylation (44). MLE-12 cells were infected with P. aeruginosa in the presence of a phosphatase inhibitor, okadaic acid (10 nm), and intracellular bacteria were quantified. A marked increase in the number of intracellular P. aeruginosa was observed (Fig. 5A). Treatment with the tyrosine kinase inhibitor, genistein (100 g/ml), before and during infection with P. aeruginosa caused a significant decrease in bacterial invasion (Fig. 5A). Most interestingly, the combination of okadaic acid and genistein appeared to neutralize the effects of either drug alone and had no effect on P. aeruginosa invasion (Fig. 5A). In contrast, changes in tyrosine phosphorylation had no impact on L. monocytogenes invasion (Fig. 5A). Also, an MTT adherence assay confirmed that treatment with okadaic acid or genistein does not impair the adherence of P. aeruginosa to MLE-12 cells (Fig. 5B). Treatment with okadaic acid and genistein indicate that P. aeruginosa invasion of MLE-12 cells can be regulated by changes in phosphorylation.
Next, we determined whether caveolin-1 or -2 is the target of tyrosine phosphorylation that influences P. aeruginosa infection. Immunoprecipitation of caveolin-1 and caveolin-2 was performed to measure any changes in phosphorylation during infection with okadaic acid and genistein. Immunoblotting for tyrosine phosphorylation after immunoprecipitation of caveolin-2 revealed that P. aeruginosa infection leads to significantly increased tyrosine phosphorylation of caveolin-2 (Fig. 5C). A cell-free extract of P. aeruginosa had no effect on caveolin-2 tyrosine phosphorylation (data not shown). Therefore, it appears that caveolin-2 tyrosine phosphorylation is not a result of the multiple secreted toxins that are produced by P. aeruginosa but as a result of P. aeruginosa binding to the cell surface. P. aeruginosa infection in the presence of genistein leads to a marked reduction in tyrosine phosphorylation of caveolin-2 (Fig. 5C). On the other hand, treatment with okadaic acid during P. aeruginosa infection caused a significant increase in the tyrosine phosphorylation of caveolin-2. The phosphorylation was dependent on lipid raft integrity because pretreatment with cyclodextrin prior to P. aeruginosa exposure caused a marked decrease in caveolin-2 phosphorylation. In contrast, tyrosine phosphorylation of caveolin-1 was unchanged during infection and also by treatment with okadaic acid, genistein, or cyclodextrin (data not shown). Thus, tyrosine phosphorylation of caveolin-2 is dependent on lipid raft integrity and appears to be a critical regulator of P. aeruginosa invasion.
In summary, type I pneumocytes are a main site of infection during P. aeruginosa pneumonia. P. aeruginosa is able to invade type I-like cells in vitro via a lipid raft-dependent mechanism. The lipid raft-dependent entry of P. aeruginosa appears to be regulated by expression and phosphorylation of caveolin-2.

DISCUSSION
The alveolar epithelium is the largest surface area exposed to the external environment and likely plays an important yet under-recognized role in the pathogenesis of bacterial pneumonia. In this paper, we have shown in vitro that P. aeruginosa has evolved the ability to invade type I pneumocytes. P. aeruginosa invasion of type I pneumocytes is markedly inhibited by disruption of lipid rafts. Once internalized, P. aeruginosa colocalizes with lipid raft components within intracellular vacuoles. An increasing number of viral and bacterial pathogens have been recognized to use this mechanism of invasion in order to avoid host defenses (21,22). Although previous work has shown that P. aeruginosa invades nasal and bronchial epithelium during the colonization of the upper airways (13,14), we have shown for the first time that P. aeruginosa is able to use lipid raft-mediated endocytosis to invade type I pneumocytes. We believe that P. aeruginosa has co-opted the pathway of lipid raft-mediated endocytosis as a means of surviving within the protected intracellular environment. Once P. aeruginosa is intracellular, it is safe from clearance by alveolar macrophages and may be able to replicate within this protected environment. Alternatively, P. aeruginosa may leave type I cells via exocytosis from the basolateral surface. This penetration of the alveolar capillary barrier would allow dissemination into the bloodstream and throughout the host. P. aeruginosa invasion of type I pneumocytes via lipid rafts may be a critical event in the pathogenesis of P. aeruginosa pneumonia.
The role of caveolin proteins in the regulation of lipid raftmediated endocytosis is controversial. Several authors (19,45) have shown that caveolin-1 serves to stabilize the plasma membrane and decrease membrane fluidity. They concluded that decreased caveolin-1 expression led to increased caveolae-mediated uptake. On the other hand, previous work from our laboratory has shown that decreased expression of caveolin-1 impairs E. coli invasion of bladder epithelial cells through lipid rafts (29). It is possible that the role of caveolin proteins differs depending on the specific cell type and also the endocytic cargo. In addition, studies of caveolin-1 knockdowns have been lim-FIG. 3. Intracellular P. aeruginosa co-localizes with lipid raft components. Type I-like cells were cultured on Permanox© plastic coverslips and infected with fluorescently labeled P. aeruginosa (red). After fixation, the cells were labeled with antibodies directed against typical lipid raft components (green) and examined using confocal microscopy. Confocal sections through the center of type I-like cells reveals that intracellular P. aeruginosa co-localizes (yellow) with GM1 (A), caveolin-1 (B), and caveolin-2 (C). Using confocal microscopy to quantify the percent of co-localization with lipid raft proteins, we found that 80% of intracellular P. aeruginosa co-localized with a caveolin-1, and 84% of intracellular P. aeruginosa co-localized with caveolin-2. Similar colocalization of P. aeruginosa with lipid raft components was also seen after infection of MLE-12 cells (data not shown).
ited due to the concomitant decrease in caveolin-2 expression because caveolin-1 is required as a chaperone for the transport of caveolin-2 (33,34). It is possible that many of the functions previously attributed to caveolin-1 may be due to caveolin-2. For example, the caveolin-2 knock-out mice have a similar pulmonary phenotype to caveolin-1-deficient mice despite ade- FIG. 4. Caveolin-2 expression is required for the optimal invasion of P. aeruginosa. In vivo type I pneumocytes express high levels of caveolin-1 and 2. Frozen sections of rat lung were fixed and stained with anti-caveolin-1 (A) and caveolin-2 (B) antibodies (green). MLE-12 cells also express high levels of caveolin proteins (data not shown). C, RNA interference was used to create stable knockdowns of caveolin-1 and -2. Immunoblotting of whole cell lysates was performed after several passages to examine the relative amount of caveolin expression relative to control (pQC-LUC). Caveolin-1 knockdowns (pQC-cav1) have significantly decreased expression of caveolin-1 and -2. In contrast, caveolin-2 knockdowns (pQC-cav2) have decreased caveolin-2 expression but have resulted in normal expression of caveolin-1. D, similar to type I-like cells and wild type MLE-12 cells, 7.9% of added P. aeruginosa (Pa) adhere and 4.2% of adherent P. aeruginosa are able to invade pQC-LUC cells. E, intracellular P. aeruginosa was quantified in pQC-cav1 and pQC-cav2 using gentamicin protection assays. MLE-12 cells transfected with a non-sense sequence from the luciferase gene (pQC-LUC) were used as controls. P. aeruginosa invasion is inhibited in both pQC-cav-1 and pQC-cav-2 cells. In contrast, there is no change in the number of intracellular L. monocytogenes. F, there was no change in P. aeruginosa adherence to either pQC-cav1 or pQC-cav2 cells compared with controls (data are presented as mean Ϯ S.E., *, p Ͻ 0.05). quate expression of caveolin-1 and the formation of visible caveolae on the cell membrane (46).
In our model of P. aeruginosa invasion, decreased expression of caveolin-1 and -2 significantly inhibits P. aeruginosa invasion of MLE-12 cells. Our data suggest that caveolin-2 may be a more important determinant of P. aeruginosa invasion of alveolar epithelial cells because a selective knockdown of caveolin-2 expression has the identical susceptibility to P. aeruginosa invasion as the combined caveolin-1 and -2 knockdown. Although we cannot rule out a significant contribution of caveolin-1 to the regulation of P. aeruginosa invasion, we can conclude for the first time that caveolin-2 plays a critical role in mediating this endocytic process.
Lipid rafts have been implicated in a wide range of cellular functions including endocytosis and cell signaling. We have shown that the endocytosis of P. aeruginosa not only requires the integrity of lipid raft platforms and expression of caveolin proteins but also the tyrosine phosphorylation of key lipid raft proteins including caveolin-2. In our model, tyrosine phosphorylation of caveolin-2 appears to be a key regulator of lipid raft-mediated endocytosis. Caveolin-2 has a recognized conserved site for tyrosine phosphorylation at its N terminus on FIG. 5. Lipid raft-dependent tyrosine phosphorylation of caveolin-2 appears to regulate P. aeruginosa invasion of MLE-12 cells. A, gentamicin protection assays were used to compare the number of intracellular P. aeruginosa after manipulating phosphorylation signaling pathways. Treatment with the tyrosine kinase inhibitor genistein (100 g/ml) for 30 min prior to infection and during infection significantly decreased the number of intracellular P. aeruginosa. In contrast, the number of intracellular P. aeruginosa bacteria was significantly increased when MLE-12 cells were infected in presence of OA (10 nM). Most interestingly, the combination of OA and genistein had no effect on P. aeruginosa invasion. Pretreatment with methyl-␤-cyclodextrin (3 mM) for 30 min prior to P. aeruginosa infection in the presence of OA negated the previously noted increased invasion associated with OA. Neither OA nor genistein affected the invasion of L. monocytogenes. B, OA and genistein had no effect on the adherence of P. aeruginosa. C, immunoprecipitation of caveolin-2 was performed to examine the effects of P. aeruginosa infection on the tyrosine phosphorylation of caveolin-2. The bar graph shows the densitometry measured from Western blots obtained from three separate experiments. P. aeruginosa infection significantly increases cav-2 tyrosine phosphorylation. Treatment with genistein before and during P. aeruginosa infection decreased caveolin-2 phosphorylation. On the other hand, Cav-2 tyrosine phosphorylation was significantly increased by infection in the presence of OA. Finally, tyrosine phosphorylation of caveolin-2 is dependent on lipid raft integrity as pretreatment with MCD prior to P. aeruginosa infection with OA negated the effects of OA (data are presented as mean Ϯ S.E., *, p Ͻ 0.05). tyrosine 19 (QLFMADDApY, where pY is phosphotyrosine) and can be phosphorylated by c-Src, which is located within lipid rafts (47). We are currently trying to identify the site and mechanism of caveolin-2 phosphorylation after P. aeruginosa infection. Multiple pathogens have been recognized to induce phosphorylation of host proteins and alter host cell signaling during infection (44). It has been recognized that P. aeruginosa infection of airway epithelial cells is able to induce phosphorylation of other host proteins including MUC1 and also activate mitogen-activated protein kinase signaling pathways (48). Bacterial pathogens may have evolved mechanisms of altering host cell signaling pathways in order to invade host cells and develop a survival advantage. In addition, it is not known how caveolin-2 phosphorylation may mediate the downstream events that appear to regulate lipid raft-mediated uptake of P. aeruginosa. Phosphorylation of caveolin-2 causes the disassociation of caveolin-2 from caveolin-1 hetero-oligomers. After phosphorylation caveolin-2 remains with the lipid rafts and is now able to bind Src homology 2 domain containing proteins, including RAS-gap (47). Therefore, the phosphorylation of caveolin-2 may activate host cell signaling pathways that regulate lipid raft-mediated endocytosis. Alternatively, caveolin-2 may interact directly with important cytoskeletal proteins depending on its phosphorylation state. Although endocytosis and signaling are often considered distinct functions of lipid rafts, we have shown here that they are interdependent. The integrity of lipid rafts brings together key signaling molecules that coordinate the complex cellular process of endocytosis of P. aeruginosa by alveolar epithelial cells.
In summary we have shown that P. aeruginosa has evolved the ability to invade type I pneumocytes by co-opting the pathway of lipid raft-mediated endocytosis. The uptake of P. aeruginosa by alveolar epithelial cells is dependent on the lipid raft integrity as well as tyrosine phosphorylation. Specifically, P. aeruginosa invasion requires the adequate expression and phosphorylation of caveolin-2. Although P. aeruginosa has evolved the ability to invade type I pneumocytes through lipid rafts, the evolutionary response of the host is still unknown. Previous studies (13) have shown that nasal and bronchial epithelial cells undergo apoptosis after P. aeruginosa invasion. Similarly, bladder epithelial cells undergo apoptosis after E. coli invasion via lipid rafts (49). This host defense mechanism allows these epithelial cells to shed infected cells and clear bacterial pathogens. The bronchial, nasal, and bladder epithelia are several layers thick and able to tolerate superficial apoptosis. The alveolar epithelium is a single layer thick and an important regulator of permeability within the lung. Apoptosis may not be protective in this compartment but rather may have detrimental effects on alveolar permeability leading to alveolar filling and worsening oxygenation. This process may contribute to the observation that invasive P. aeruginosa infections in the alveolar space frequently lead to pneumonia, sepsis, and death (2). Future studies to examine the role of lipid rafts and caveolin in an animal model will help to elucidate the host response in the ongoing evolutionary battle.