Phosphatidylinositol 3-kinase and frabin mediate Cryptosporidium parvum cellular invasion via activation of Cdc42.

Cryptosporidium parvum invades target epithelia via a mechanism that involves host cell actin reorganization. We previously demonstrated that C. parvum activates the Cdc42/neural Wiskott-Aldrich syndrome protein network in host cells resulting in actin remodeling at the host cell-parasite interface, thus facilitating C. parvum cellular invasion. Here, we tested the role of phosphatidylinositol 3-kinase (PI3K) and frabin, a guanine nucleotide exchange factor specific for Cdc42 in the activation of Cdc42 during C. parvum infection of biliary epithelial cells. We found that C. parvum infection of cultured human biliary epithelial cells induced the accumulation of PI3K at the host cell-parasite interface and resulted in the activation of PI3K in infected cells. Frabin also was recruited to the host cell-parasite interface, a process inhibited by two PI3K inhibitors, wortmannin and LY294002. The cellular expression of either a dominant negative mutant of PI3K (PI3K-Deltap85) or functionally deficient mutants of frabin inhibited C. parvum-induced Cdc42 accumulation at the host cell-parasite interface. Moreover, LY294002 abolished C. parvum-induced Cdc42 activation in infected cells. Inhibition of PI3K by cellular overexpression of PI3K-Deltap85 or by wortmannin or LY294002, as well as inhibition of frabin by various functionally deficient mutants, decreased C. parvum-induced actin accumulation and inhibited C. parvum cellular invasion. In contrast, the overexpression of the p85 subunit of PI3K promoted C. parvum invasion. Our data suggest that an important component of the complex process of C. parvum invasion of target epithelia results from the ability of the organism to trigger host cell PI3K/frabin signaling to activate the Cdc42 pathway, resulting in host cell actin remodeling at the host cell-parasite interface.

ites. Mediated by uncharacterized but specific ligands on the sporozoite surface and unidentified receptors on the host cell, the sporozoite attaches to the apical membrane of the host epithelial cell, inducing host cell actin cytoskeleton remodeling and protrusion of the host cell membrane around the sporozoite to form a parasitophorous vacuole. At the base of each vacuole, the parasite establishes a unique "electron-dense band" of unknown composition, which separates the organism from the host cell cytoplasm. Thus, the parasitophorous vacuole and the dense-band keep the internalized parasite intracellular but extracytoplasmic (2)(3)(4).
Host cell actin is a common molecular target of many pathogenic microbes including viruses and bacteria such as vaccinia virus, Listeria monocytogenes, Escherichia coli, Salmonella enterica, and Shigella flexneri (5). These microbes utilize host cell actin for multiple reasons including cell attachment and entry, movement within and between cells, vacuole formation and remodeling, and avoidance of phagocytosis (6). Recent studies by us (7) and others (8 -10) have demonstrated that host cell actin remodeling at the host cell-parasite interface is necessary for the cellular invasion of C. parvum. Inhibition of host cell actin polymerization by pharmacological inhibitors, such as cytochalasin B and cytochalasin D (7,8), or by cellular expression of specific inhibitory fragments of actin-associated proteins, such as Scar-WA (10), block C. parvum cellular invasion. Several host cell proteins necessary for the actin-based motility of Listeria and Shigella within the infected host cells (11) also have been identified at the C. parvum attachment site, including cortactin, Arp2/3 complex, neural Wiskott-Aldrich syndrome protein (N-WASP), 1 and vasodilator-stimulated phosphoprotein (VASP) (10). More recent studies have revealed that several intracellular signaling pathways are involved in C. parvum-induced actin remodeling. Whereas the accumulation of cortactin at the C. parvum host cell-parasite interface appears to be dependent upon the activation of the c-Src signaling pathway (12), the activation of Cdc42, one member of the Rho family of small GTPases, is required for the accumulation of N-WASP and the Arp2/3 complex at the host cell-parasite interface during C. parvum cellular invasion (13).
Activation of Cdc42 requires guanine nucleotide exchange factors (GEFs). These proteins stimulate the dissociation of GDP from the GDP-bound inactive form, resulting in the bind-ing of GTP to form the GTP-bound active form, the activity of which is often regulated by an upstream signal. One of the GEFs for Cdc42 is frabin, a recently identified protein ubiquitously expressed and implicated in the mechanisms of Cdc42associated actin remodeling (14,15). Frabin contains an Factin binding (FAB) domain and two pleckstrin homology (PH) domains (16). Proteins with PH domains can bind to phosphoinositides such as phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P 3 ) (17,18). One important signaling kinase implicated in actin polymerization and activated upon membrane stimulation by a variety of ligands is the class IA phosphatidylinositol 3-kinase (PI3K). Activation of PI3K catalyzes the production of PtdIns (3,4,5)P 3 at cell membranes, thus resulting in the recruitment and activation of various signaling components to the cell membrane, some of which have been implicated in the regulation of the cytoskeleton, e.g. growth factor-induced membrane ruffling (19). Using an in vitro model of intestinal cryptosporidiosis, Forney et al. (8) have demonstrated previously that a selective PI3K inhibitor, wortmannin, blocked C. parvum infection of cultured bovine fallopian tube epithelial cells. Thus, in this study, we tested the role of PI3K and frabin in C. parvum-induced Cdc42 activation in cellular invasion.
We found that C. parvum infection of cultured human biliary epithelial cells induces the accumulation of PI3K at the host cell-parasite interface and results in the activation of PI3K in infected cells. Frabin is also recruited to the host cell-parasite interface, a process that is dependent upon PI3K. Both PI3K and frabin appear to be required for Cdc42 accumulation at the host cell-parasite interface. In addition, the inhibition of PI3K by either selective pharmacologic inhibitors or host cell overexpression of functionally deficient mutants of PI3K and frabin was associated with a reduction of C. parvum-induced actin remodeling and ultimately C. parvum invasion of biliary epithelia. These findings demonstrate that C. parvum invasion of biliary epithelial cells is facilitated by the recruitment and activation of PI3K in host cells, a process that activates the Cdc42 pathway via frabin and is required for C. parvum-induced actin remodeling and cellular invasion.

EXPERIMENTAL PROCEDURES
C. parvum and H69 Cells-C. parvum oocysts of the Iowa strain were purchased from a commercial source (Pleasant Hill Farms, Troy, ID). Before infecting cells, oocysts were excysted to release infective sporozoites as described previously (7). H69 cells (a gift of Dr. D. Jefferson, Tufts University, Boston, MA) are SV40-transformed human bile duct epithelial cells originally derived from a normal liver harvested for transplant and have been characterized extensively (20). For experiments, H69 cells were used between passages 23 and 30 and maintained for three passages without co-culture cells to ensure that the culture was free of 3T3 fibroblasts.
In Vitro Models and Infection Assay-Two in vitro models, an attachment model and an attachment/invasion model, were employed to assay the attachment and invasion of C. parvum to H69 cells as described previously (7). H69 cells were seeded into 4-well chamber slides or 6-well Costar tissue culture plates (BD Biosciences Labware, Franklin Lakes, NJ) and grown to 70 -80% confluence. For the attachment model, H69 cells were fixed with 4% paraformaldehyde (Electron Microscopy Sciences, Fort Washington, PA) in PBS before exposure to C. parvum sporozoites. In this model, the organism can only attach to the fixed cell surface. For the attachment/invasion model, live cells (without pre-fixation) were exposed directly to C. parvum sporozoites, thus allowing the organism to both attach to and enter into host cells. Infection with C. parvum was done in a culture medium consisting of Dulbecco's modified Eagle's-Ham's F-12 medium, 100 units/ml penicillin, and 100 g/ml streptomycin (Invitrogen) and freshly excysted C. parvum sporozoites (1 ϫ 10 6 sporozoites/slide well or culture plate). Inactivated organisms (treated at 65°C for 30 min) were used for sham infection experiments (1). Infection assays (attachment rate or attachment/invasion rate) were carried out after a 2-h incubation with the parasite employing an indirect immunofluorescent technique as described previously (7). Parasites attaching to pre-fixed cells or invading non-fixed cells were counted, and the results were expressed, respectively, as attachment rate or attachment/invasion rate (the number of parasites/total number of cells ϫ 100). Approximately 2,000 cells were counted for each assay. For the inhibitory experiments, two selective inhibitors of PI3K, LY294002 and wortmannin (21), were added in the medium at the time C. parvum was added. A concentration of 10 M LY294002 or 100 nM wortmannin showed no cytotoxic effects on H69 cells or on C. parvum sporozoites and was selected for the study.
Transfection of Cells-H69 cells for transient transfection were grown to 40 -60% confluency on 8-well chamber slides and transfected with 1 g/well plasmid DNA using the LipofectAMINE Plus TM reagent kit according to the manufacturer's recommendations. The plasmid constructs for transient transfection included the following: pEF-BOS⌬RI-PI3K-⌬p85-HA (a dominant negative mutant of p85 regulatory subunit that cannot interact with the catalytic p110 subunit) (22)(23)(24) and pEF-BOS⌬RI-PI3K-p85-HA (wild type of the p85 subunit of PI3K) (gifts from Dr. L. M. Karnitz, Mayo Medical School, Rochester, MN); pKR5-Cdc42(17N)-Myc (a dominant negative mutant and a gift from Dr. A. Hall, University College London, London, United Kingdom) (25); and various constructs of frabin (14,15,26) such as pCMV-Frabin-GFP, pCMV-Frabin-dead-Full-Myc (an internal deletion of amino acids 353-362 in the Dbl homology (DH) domain that abolishes Cdc42-activating activity), and pCMV-Frabin-dead-DH-PH1-Myc (a truncated mutant lacking of the FAB domain but containing the DH and PH1 domains with an internal deletion of amino acids 353-362 in the DH domain) (gifts from Dr. Y. Takai, Osaka University Medical School, Osaka, Japan). Empty vectors were used as controls. Twenty-four h after transfection, cells were exposed to C. parvum sporozoites for the attachment and invasion assays.
Immunofluorescent Microscopy-H69 cells were exposed to C. parvum sporozoites as described above. After 2 h of incubation, cells were fixed (0.1 mol/liter PIPES, pH 6.95, 1 mmol/liter EGTA, 3 mmol/liter magnesium sulfate (Sigma-Aldrich), and 2% paraformaldehyde) at 37°C for 20 min and then permeabilized with 0.2% (v/v) Triton X-100 in PBS. For double immunofluorescent labeling, fixed cells were incubated with primary monoclonal antibodies to associated proteins mixed with a polyclonal antibody against C. parvum sporozoite membrane proteins (a generous gift from Drs. Guan Zhu and Janet Keithly, Wadsworth Center, Albany, NY) followed by rhodamine-labeled anti-mouse and fluorescein-labeled anti-rabbit antibodies (Molecular Probes, Eugene, OR). Some cells were incubated with polyclonal antibodies to associated proteins mixed with a monoclonal antibody against C. parvum (2H2, ImmunuCell, Portland, ME) followed by rhodamine-labeled anti-rabbit and fluorescein-labeled anti-mouse antibodies. To confirm the specificity of the staining, multiple antibodies from different sources were used for some of the proteins including two antibodies against Cdc42 (clone B8, Santa Cruz Biotechnology and a polyclonal antibody from Calbiochem), one monoclonal antibody and one polyclonal antibody to PI3K (Upstate Biotechnology), and a monoclonal antibody to frabin (clone 43, BD Transduction Laboratories). For localization of actin with C. parvum, rhodamine-phalloidin (Sigma-Aldrich) was co-incubated with the secondary antibody. For triple immunofluorescent labeling in experiments with the transient transfected cells, monoclonal antibodies to the c-Myc epitope tag (clone 9B11, Cell Signaling) or HA epitope tag (clone 262K, Cell Signaling) were used to identify the transfected cells. The cells were co-stained with another antibody against the associated protein plus rhodamine-phalloidin to stain actin or 4Ј,6-diamidino-2phenylindole (DAPI) (5 mol/liter) to label the parasite (which stains the nuclei of both the host cells and the parasite in blue) (8). Labeled cells were rinsed three times with PBS and once with distilled water, mounted with mounting medium (H-1000, Vector Laboratories), and assessed by confocal laser-scanning microscopy. The numbers of parasites with and without accumulation of associated proteins at the host cell-parasite interface were counted separately for quantitative analysis. Those with the obvious accumulation of each associated protein were counted as positive, and the results were expressed as the accumulation percentage (the number of parasites with an accumulation of the molecules at the host cell-parasite interface/total number of parasites ϫ 100). Between 500 and 1000 C. parvum cells were counted randomly for each assay. Images obtained from the Zeiss 510 confocal microscope (Carl Zeiss, Inc. Oberkochen, Germany) were manipulated uniformly for contrast and intensity using the Adobe Photoshop (Mountain View, CA) software.
Immunoprecipitation and Phosphotyrosine of PI3K-H69 cells were grown in 10-cm dishes to 95% confluence and exposed to C. parvum sporozoites at 37°C for 1 and 2 h. Cells then were rinsed with PBS and scraped into 1 ml of cold lysis buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid, and 0.1% SDS) supple-mented with 1 mM phenylmethylsulfonyl fluoride, leupeptin, and pepstatin at 20 g/ml and tyrosine phosphatase inhibitors, sodium orthovanadate and sodium fluoride, at 1 mM. The cell lysates were centrifuged at 13,000 ϫ g for 10 min and were assayed for protein concentration using the Bradford reagent (Sigma-Aldrich). The cell lysates (1 mg of protein) then were incubated with 10 l of anti-PI3K (Upstate Biotechnology) at 4°C overnight to immunoprecipitate PI3K. Immune complexes were collected by direct binding to protein A-Sepharose. The immunoprecipitated protein then was released with Western sample buffer (20% glycerol, 4% SDS, 10% ␤-mercaptoethanol, 0.05% bromphenol blue, and 1.25 M Tris, pH 6.8), incubated at 95°C for 5 min, separated by SDS-PAGE, and immunoblotted with an antibody to phosphotyrosine (Clone 4G10, Upstate Biotechnology). Immunoreactive bands were developed using a chemiluminescent substrate (ECL Plus, Amersham Biosciences).
PI3K and Cdc42 Activity Assay-PI3K activity was detected using an enzyme-linked immunosorbent assay (ELISA). H69 cells were grown on 10-cm dishes to 95% confluency and exposed to C. parvum sporozoites at 37°C for 1 and 2 h. PI3K protein was immunoprecipitated, and the activity of PI3K was determined using a commercial ELISA kit (Echelon Biosciences Inc.). The activity of PI3K was expressed as the picomole of product phosphatidylinositol 3,4,5-trisphosphate/1 mg protein.

PI3K Is Recruited to the Host Cell-Parasite Interface and Is
Activated-To test the role of PI3K in the mechanism of host cell actin-dependent invasion of epithelia by C. parvum, we used double immunofluorescent labeling for C. parvum and PI3K. Very strong staining of PI3K was observed at the C. parvum host cell-parasite interface (Fig. 1, A and B). No accumulation of PI3K was detected in sham-infected control cells, and no positive staining of PI3K was found in the parasite itself (data not shown). Quantitative analysis showed 96.7 Ϯ 2.5% of the host cell-parasite interfaces with an accumulation of PI3K.
Ligand-stimulated activation of class IA PI3K is dependent upon the phosphorylation of Tyr 688 of its p85 subunit (29). To test whether C. parvum infection induces the phosphorylation of this p85 subunit in host cells, biliary epithelia were exposed to C. parvum, cell lysates were immunoprecipitated using an antibody to PI3K, and the immunoprecipitates were blotted for phosphotyrosine. As shown in Fig. 1C, phosphotyrosine in the p85 subunit was detected at 1 and 2 h after exposure to C. parvum. Activation of PI3K was determined further by ELISA using the immunoprecipitate as a substrate. As shown in Fig. 1D, cells exposed to C. parvum for 1 and 2 h showed a marked increase in PI3K activity compared with the shaminfected control cells.
PI3K-dependent Accumulation of Frabin at the Host Cell-Parasite Interface-To test whether frabin is involved in C. parvum cellular invasion, accumulation of frabin at the host cell-parasite interface was measured first by double immunofluorescent labeling using a monoclonal antibody to frabin and a polyclonal antibody to C. parvum. Strong staining of frabin was observed at the C. parvum host cell-parasite interface (Fig.  2, A1 and A2). No accumulation of frabin in the sham-infected cells or positive staining in the parasite itself was found with the procedure used in the experiments (data not shown). To further confirm the specificity of frabin staining with the antibody, a GFP-frabin construct was used to transfect cells before exposure to C. parvum sporozoites followed by confocal microscopy. As shown in Fig. 2, B1 and B2, GFP-frabin also was found to be recruited to the host cell-parasite interface.
To define whether C. parvum-induced frabin accumulation at the host cell-parasite interface is dependent upon PI3K

FIG. 1. C. parvum infection of biliary epithelial cells accumulates host cell PI3K at the host cell-parasite interface and results in PI3K activation in infected cells.
A and B, representative confocal micrographs demonstrating that C. parvum accumulates PI3K at the host cell-parasite interface. Cells were exposed to C. parvum for 2 h and then co-stained for C. parvum and PI3K followed by confocal microscopy. Accumulation of host cell PI3K at the host cell-parasite interface was observed in infected cells (arrowheads in A and B). Labels indicate staining of C. parvum or PI3K. C, Western blot of tyrosinephosphorylated p85 subunit. Cells were harvested after 0 -2 h of exposure to C. parvum. The p85 subunit of PI3K was immunoprecipitated from cell protein isolates, separated by SDS-PAGE, and then probed for phosphotyrosine. The same total cell lysates also were separated by SDS-PAGE and probed for p85 to demonstrate unchanged total p85 protein level. D, activity of PI3K determined by ELISA. Cells were exposed to C. parvum and harvested at the indicated time points. PI3K was immunoprecipitated from cell protein lysates, and PI3K activity in the immunoprecipitates then was measured with a commercially available ELISA kit. Results are expressed as picomole of product phosphatidylinositol 3,4,5-trisphosphate/1 mg protein and are means Ϯ S.E. of three independent experiments. Asterisks indicate p Ͻ 0.05 versus sham-infected controls. Infect, infection. Bar ϭ 5 m.
activation, LY294002 and wortmannin, two selective PI3K inhibitors (21), and PI3K-⌬p85, a dominant negative mutant p85 regulatory subunit of PI3K unable to interact with the catalytic p110 subunit thus blocking PI3K activity (22-24), were used. As shown in Fig. 2, C1-D2, a significant decrease in frabin accumulation was found at the host cell-parasite interface in cells treated with LY294002 or transfected with PI3K-⌬p85. A similar inhibition of frabin accumulation was detected in cells treated with wortmannin (data not shown). Quantitative anal-ysis showed a significant (p Ͻ 0.01) decrease in frabin accumulation in the presence of PI3K inhibitors and in cells transfected with PI3K-⌬p85 compared with cells in the absence of inhibitors or cells transfected with the empty vector (Fig. 2E). These data indicate that the recruitment of a Cdc42-specific GEF to the host cell-parasite interface requires the activation of PI3K.
To test the possibility that frabin may be required for PI3K accumulation at the host cell-parasite interface, we tested the effects of functionally deficient mutates of frabin on PI3K accumulation. H69 cells were transfected transiently with Frabin-dead-Full or Frabin-DH-PH1, two functionally deficient mutants of frabin (14,15,26), and then exposed to C. parvum followed by co-staining of PI3K. Accumulation of PI3K was found both in cells transfected with Frabin-dead-Full (arrowheads in Fig. 3A2) or Frabin-DH-PH1 (arrowheads in Fig. 3, B2) and in the non-transfected cells (arrows in Fig. 3, A2 and  B2). Quantitative analysis showed no significant difference in PI3K accumulation between transfected and non-transfected cells (Fig. 3C), suggesting that frabin is not required for the accumulation of PI3K during C. parvum invasion of biliary epithelial cells.
Accumulation/Activation of Cdc42 at the Host Cell-Parasite Interface Requires PI3K and Frabin-Our previous studies indicated that C. parvum cellular invasion activates the Cdc42-Arp/N-WASP network to induce actin remodeling at the attachment site, thus facilitating C. parvum cellular invasion (13). To clarify the relationship between PI3K/Frabin signaling pathways and the Cdc42-Arp/N-WASP network during C. parvum cellular invasion, C. parvum-induced Cdc42 accumulation was explored in cells treated with PI3K inhibitors or in cells transfected with PI3K-⌬p85 or various mutants of frabin. Consistent with our previous studies (13), a strong accumulation of Cdc42 was observed at the host cell-parasite interface in the absence of PI3K inhibitors (Fig.   4, A1 and A2) with 93.7 Ϯ 1.3% of the attachment sites accumulated with Cdc42. A significant decrease in Cdc42 accumulation at the host cell-parasite interface was observed in cells treated with LY294002 (Fig. 4, B1 and B2) or transfected with PI3K-⌬p85 (Fig. 4, C1 and C2). A decrease in Cdc42 accumulation also was found in cells transfected with the Frabin-DH-PH1 (Fig. 4, D1 and D2). Quantitative analysis showed a significant decrease in Cdc42 accumulation in the transfected cells or cells treated with PI3K inhibitors compared with the controls (Fig. 4E). To test whether the activation of Cdc42 in infected cells is also dependent on PI3K, cells were exposed to C. parvum in the absence or presence of selective PI3K inhibitors followed by the GST pull-down assay. As shown in Fig. 4F, a strong band for Cdc42 was detected in infected cells compared with shaminfected controls. This band was not detected in cells exposed to C. parvum in the presence of LY294002, indicating that C. parvum-induced activation of Cdc42 in infected cells requires PI3K activity.
PI3K also has been implicated as one of the downstream effectors of the Cdc42 signaling pathway in some cell types (30,31). To clarify whether PI3K activation is dependent upon the Cdc42 signaling pathway during C. parvum cellular invasion, we observed the effects of Cdc42 inhibition on PI3K/frabin accumulation during the C. parvum invasion of biliary epithelial cells. H69 cells were transfected transiently with Cdc42(17N), a dominant negative mutant that blocks the activation of Cdc42, and then exposed to C. parvum followed by co-staining of PI3K or frabin. Accumulation of PI3K (Fig. 5, A1 and A2) and frabin (Fig. 5, B1 and B2) was found both in cells transfected with Cdc42(17N) (arrowheads in Fig. 5, A2 and B2) and in the non-transfected cells (arrows in Fig. 5, A2 and B2).

FIG. 4. PI3K and frabin are required for both C. parvum-induced accumulation and activation of Cdc42 in infected cells. A1
and A2, representative confocal micrographs demonstrating Cdc42 accumulation at the host cell-parasite interface in the absence of PI3K inhibitors (arrows in A1 and A2). B1 and B2, a significant decrease in Cdc42 accumulation was observed in the presence of LY294002 (arrows in B1 and B2). C1-D2, cellular expression of PI3K-⌬p85 (C1 and C2) or Frabin-DH-PH1 (D1 and D2) also inhibited C. parvum-associated Cdc42 accumulation at the host cell-parasite interface. C. parvum was identified with DAPI staining of its nucleus (arrowheads in C1 and D1), and transfected cells were identified by antibodies to the c-Myc/HA epitope tag (inset in C2 and D2). Similar inhibition of Cdc42 accumulation was found in cells treated with wortmannin or transfected with Frabin-dead-Full (data not shown). E, quantitative analysis of Cdc42 accumulation. F, LY294002 inhibited C. parvum-induced Cdc42 activation in infected cells by GST pull-down assay. Cells were exposed to C. parvum sporozoites for 1 h in the absence or presence of LY294002. The whole lysates showed similar bands to Cdc42 suggesting no change at the total protein levels in infected cells either in the absence or presence of LY294002. GST pull-down assay using GST-Cdc42/Racinteractive binding (CRIB) domain (which specifically binds to the GTP-bound form of Cdc42) showed a much stronger band in C. parvuminfected cells than in the sham-infected cells, suggesting the activation of Cdc42 in biliary epithelial cells after C. parvum infection. A much weaker band of activated Cdc42 was detected in cells exposed to C. parvum in the presence of LY294002. p Ͻ 0.05 versus no inhibitor or empty vector-transfected controls. LY29, LY294002; Wort, wortmannin; Infect, infection; Bar, 5 m. Quantitative analysis showed no significant difference in PI3K and frabin accumulation between Cdc42(N17)-transfected and non-transfected cells (Fig. 5C), further suggesting that PI3K and frabin are the upstream activators but not the downstream effectors of Cdc42 during C. parvum cellular invasion of biliary epithelial cells.
Activation of PI3K and Frabin Is Required for C. parvum Induced Actin Remodeling at the Host Cell-Parasite Interface-C. parvum-induced actin remodeling at the host cellparasite interface in cells treated with PI3K inhibitors or transfected with mutants of PI3K and frabin was observed by immunofluorescent microscopy. Whereas a strong actin accumulation was found at the host cell-parasite interface in cells in the absence of PI3K inhibitors (Fig. 6, A1 and A2), a significant decrease in actin accumulation was detected in cells in the presence of LY294002 (Fig. 6, B1 and B2) or wortmannin (data not shown). A similar pattern of strong actin accumulation was observed in cells transfected with the empty vectors compared with non-transfected cells (data not shown). In contrast, cells transfected with PI3K-⌬p85 (Fig. 6, C1 and C2), Frabin-dead-Full (Fig. 6, D1 and D2), or Frabin-DH-PH1 (Fig. 6, E1 and E2) showed a significant decrease in actin accumulation at the host cell-parasite interface compared with non-transfected cells. Quantitative analysis showed a significant decrease in actin accumulation in the transfected cells or cells treated with LY294002 or wortmannin compared with the controls (Fig. 6F). Interestingly, cells transfected with Frabin-dead-Full (with the FAB domain) and Frabin-DH-PH1 (without the FAB domain) showed a similar decrease in actin accumulation (Fig. 6F), suggesting that the FAB domain of frabin is not involved directly in the mechanism of C. parvum-induced actin remodeling. Thus, it appears that a functional PI3K Cdc42-GEF pathway is required for the accumulation of actin at the host cell-parasite interface.
C. parvum Cellular Invasion Depends upon Host Cell PI3K/ Frabin-To define whether PI3K/frabin pathways are involved in the attachment or invasion process of C. parvum into host epithelial cells, C. parvum attachment and invasion of cells in the presence of PI3K inhibitors or cells transfected with various mutants of PI3K and frabin were observed by immunofluorescent microscopy. When biliary epithelia were pre-fixed and then exposed to C. parvum sporozoites (a model in which the organism can attach to but not invade the fixed cell surface), no significant change in attachment rate was found among treated cells, including cells transfected with PI3K-⌬p85, PI3K-p85, Frabin-dead-Full, and Frabin-DH-PH as well as the cells treated with inhibitors (Fig. 7A). These results suggest that C. parvum sporozoite attachment to biliary epithelial cells is not dependent upon host cell PI3K/frabin pathways. In contrast, when unfixed living biliary epithelial cells were exposed to C. parvum sporozoites (a model in which the organism can both attach to and enter host cells), a significant decrease in the attachment/invasion rate was found in the presence of PI3K inhibitors compared with controls (Fig. 7B). Moreover, a marked decrease in the attachment/invasion rate also was found in cells transfected with PI3K-⌬p85, Frabin-dead-Full, or Frabin-DH-PH (Fig. 7B). Frabin-dead-Full and Frabin-DH-PH1 showed a similar inhibition of C. parvum invasion of host FIG. 6. PI3K and frabin are required for C. parvum-induced actin accumulation at the host cell-parasite interface. Cultured biliary epithelial cells, as well as cells transfected with various constructs of PI3K or frabin, were exposed to C. parvum sporozoites for 2 h in the absence or presence of LY294002 or wortmannin followed by immunofluorescent microscopy. A1 and A2, obvious actin accumulation was found at the host cell-parasite interface in the absence of PI3K inhibitors. B1 and B2, a significant decrease in actin accumulation at parasite-epithelial cell interface was detected in the presence of LY294002 (arrowhead in B2). Similar inhibition was observed in the presence of wortmannin (data not shown). C1 and C2, a decrease in actin accumulation also was found in cells transfected with PI3K-⌬p85 (arrowhead in C2). D1 and E2, actin accumulation at the host cellparasite interface after cellular expression of Frabin-dead-Full or  Fig. 7, D-F). Coupled with the observations of using the pre-fixed cells, the above data suggest that C. parvum sporozoite invasion of biliary epithelial cells requires host cell activation of PI3K and frabin. DISCUSSION In the work described here, we show that C. parvum not only recruits host cell PI3K to the host cell-parasite interface but also activates host cell PI3K during infection of human biliary epithelial cells. The activation of host cell PI3K is associated with frabin accumulation at the C. parvum host cell-parasite interface. Moreover, frabin accumulation is related to the C. parvum-induced Cdc42 activation in infected cells and consequently involved in the actin remodeling at the host cell-parasite interface during C. parvum invasion of epithelial cells. These results provide direct evidence that C. parvum activates host cell PI3K/frabin/ Cdc42 signaling pathways and induces host cell actin remodeling to facilitate C. parvum invasion of target epithelial cells. Our results are consistent with the PI3K-dependent host cell invasion that has been reported in infections by other microbes such as E. coli, Chlamydia pneumoniae, Helicobacter pylori, Listeria monocytogenes, and Toxoplasma cruzi (32)(33)(34)(35)(36).
PI3K catalyzes the production of PtdIns(3,4,5)P 3 at cellular membranes and contributes to the recruitment and activation of different intracellular signaling components in response to a variety of stimuli (19,37). We show here that the activation of PI3K during C. parvum infection is required for the accumulation of frabin at the host cell-parasite interface. Frabin, a recently identified GEF specific for Cdc42, contains other domains including one FAB domain, one DH domain, two PH domains, and an FYVE-finger (FYVE) domain (14,15). The PH domains are protein motifs of ϳ100 amino acids involved in the regulated targeting of signaling molecules to plasma membranes by protein-protein and/or protein-lipid interactions. Several studies have indicated that PH domains can bind to specific phosphoinositides as well as to ␤␥ subunits of heterotrimeric G-proteins and protein kinase C (17). Indeed, recent studies (17,18) have demonstrated that the binding of the PH domain to the PI3K substrate and products regulates GEF activity. Therefore, it is not surprising that the inhibition of PI3K by wortmannin or LY294002 or cellular expression of a mutant p85 regulatory subunit of PI3K blocked C. parvuminduced frabin accumulation at the host cell-parasite interface. Additionally, because functionally deficient mutants of frabin do not affect C. parvum-induced PI3K accumulation, it appears that frabin is not required for PI3K accumulation.
Our previous studies showed that C. parvum recruits Cdc42 to the parasite host cell-parasite interface, resulting in the recruitment of downstream effectors of Cdc42, and host cell actin remodeling, processes that induce membrane protrusion and dense-band formation, parasitophorous vacuole formation, and C. parvum invasion. The activation of Cdc42 depends upon GDP/GTP exchange proteins that stimulate the dissociation of GDP from the GDP-bound inactive form of Cdc42 followed by the binding of GTP (38). Here we show that the activation of Cdc42 is dependent upon the PI3K/frabin signaling activated by C. parvum. Inhibition of host cell PI3K by wortmannin or FIG. 7. C. parvum invasion of biliary epithelial cells requires the activity of host cell PI3K/frabin signaling pathways. Cells were either exposed to C. parvum in the presence of PI3K inhibitors or transfected with wild type or functional inhibitory mutants of PI3K or frabin and then exposed to C. parvum followed by immunofluorescence confocal microscopy. A, attachment assay in pre-fixed cells after a 2-h exposure to C. parvum sporozoites shows no significant difference in C. parvum attachment in all treated cells. B, attachment/invasion assay in non-fixed cells after a 2-h exposure to C. parvum sporozoites. A significant increase in infection rate was observed in cells transfected with PI3K-p85, and a significant decrease in infection rate was detected in cells treated with PI3K inhibitors, wortmannin and LY294002, or transfected with functionally inhibitory mutants of PI3K or frabin. A similar decrease in infection rate was observed in cells transfected with Frabindead-Full (with the FAB domain) or Frabin-DH-PH1 (without the FAB domain). C-F, representative confocal images of cells exposed to C. parvum for 2 h. More C. parvum parasites were detected in cells transfected with PI3K-p85 (stained in red using an antibody to the HA epitope tag in C), and much fewer C. parvum parasites were found in cells transfected with PI3K-⌬p85 (stained in red in D), Frabin-dead-Full (stained in red in E), or Frabin-DH-PH1 (stained in red in F) compared with non-transfected cells (as outlined in C-E). p Ͻ 0.05 versus no inhibitor or empty vector-transfected controls. LY29, LY294002; Wort, wortmannin; Bar, 5 m.
LY294002 or by cellular expression of a mutant p85 regulatory unit of class IA PI3K significantly inhibited both accumulation and activation of Cdc42 during C. parvum infection. Moreover, functionally truncated mutants of the DH domain of frabin (which abolish its Cdc42-activating activity) (14,15,26) also significantly inhibited Cdc42 accumulation at the C. parvum host cell-parasite interface. This observation is consistent with previous studies in fibroblasts, which showed that the DH domain of frabin is essential for the activation of Cdc42 (14,15,26). Direct binding of Cdc42 to PtdIns(3,4,5)P 3 has not been identified (39). However, because functionally deficient mutants of host cell PI3K and frabin did not block C. parvuminduced Cdc42 accumulation/activation completely, other signaling pathways and/or GEFs probably are involved in the process. On the other hand, PI3K also has been implicated as one of the downstream effectors of the Cdc42 signaling pathway in some cell types (30,31). This appears not to be the case in epithelial cells during C. parvum infection, because cellular expression of a dominant negative mutant of Cdc42 did not inhibit PI3K accumulation at the host cell-parasite interface.
In this study, we also found that both the inhibition of PI3K by the inhibitors and cellular expression of functionally inhibitory mutants of PI3K and frabin significantly blocked C. parvum-induced actin accumulation and inhibited C. parvum cellular invasion. We found no change in C. parvum attachment in cells transfected with PI3K-⌬p85 and mutants of frabin or in cells treated with the PI3K inhibitors. In contrast, we observed a significant decrease in C. parvum invasion. Moreover, the overexpression of the p85 subunit of PI3K significantly increased C. parvum invasion. Thus, C. parvum attachment does not require host cell PI3K/frabin signaling, probably because the attachment process is mediated by ligands on the sporozoite surface and receptors on the external surface of the host cell plasma membrane as others and we (1-4) have reported previously. Frabin can bind directly to F-actin with its FAB domain (14,15). Interestingly, both Frabin-dead-Full (with the FAB domain) and Frabin-DH-PH1 (without the FAB domain) showed a similar inhibition of C. parvum-induced actin accumulation and C. parvum cellular invasion, suggesting that the direct binding of frabin to F-actin via its FAB domain may not be involved in the host cell actin-dependent C. parvum cellular invasion. In a previous study, Forney et al. (8) show that wortmannin, an irreversible PI3K inhibitor, blocked C. parvum infection of cultured bovine fallopian tube epithelial cells. Together, these findings suggest that PI3K/frabin pathway facilitates C. parvum invasion (not attachment) of host epithelial cells via the activation of the Cdc42 signaling pathway to induce host cell actin remodeling at the host cell-parasite interface. How C. parvum activates PI3K/frabin at the host cellparasite interface remains unclear. A possible mechanism we favor is that the attachment of C. parvum sporozoites to the host cell surface would activate membrane receptors to trigger phosphorylation of PI3K, a scenario supported by the increase of tyrosine phosphorylation of PI3K p85 subunit detected in C. parvum-infected cells, because most membrane receptors activate PI3K via tyrosine phosphorylation of the p85 subunit (17).
In conclusion, using an in vitro model of biliary cryptosporidiosis, we demonstrated that C. parvum recruits PI3K to the host cell-parasite interface, an event that results in the activation of Cdc42 via frabin, leading to the recruitment of downstream effectors of Cdc42 and host cell actin remodeling, thus facilitating C. parvum invasion. Future studies should focus attention to define the molecular mechanisms by which C. parvum activates PI3K and the role of other actinassociated proteins.