|Home | About | Journals | Submit | Contact Us | Français|
Internalization of the obligate intracellular apicomplexan parasite, Cryptosporidium parvum, results in the formation of a unique intramembranous yet extracytoplasmic niche on the apical surfaces of host epithelial cells, a process that depends on host cell membrane extension. We previously demonstrated that efficient C. parvum invasion of biliary epithelial cells (cholangiocytes) requires host cell actin polymerization and localized membrane translocation/insertion of Na+/glucose cotransporter 1 (SGLT1) and of aquaporin 1 (Aqp1), a water channel, at the attachment site. The resultant localized water influx facilitates parasite cellular invasion by promoting host-cell membrane protrusion. However, the molecular mechanisms by which C. parvum induces membrane translocation/insertion of SGLT1/Aqp1 are obscure. We report here that cultured human cholangiocytes express several nonmuscle myosins, including myosins IIA and IIB. Moreover, C. parvum infection of cultured cholangiocytes results in the localized selective aggregation of myosin IIB but not myosin IIA at the region of parasite attachment, as assessed by dual-label immunofluorescence confocal microscopy. Concordantly, treatment of cells with the myosin light chain kinase inhibitor ML-7 or the myosin II-specific inhibitor blebbistatin or selective RNA-mediated repression of myosin IIB significantly inhibits (P < 0.05) C. parvum cellular invasion (by 60 to 80%). Furthermore ML-7 and blebbistatin significantly decrease (P < 0.02) C. parvum-induced accumulation of SGLT1 at infection sites (by approximately 80%). Thus, localized actomyosin-dependent membrane translocation of transporters/channels initiated by C. parvum is essential for membrane extension and parasite internalization, a phenomenon that may also be relevant to the mechanisms of cell membrane protrusion in general.
Cryptosporidium parvum, a coccidian organism of the phylum Apicomplexa, infects the gastrointestinal and biliary epithelia and occasionally the respiratory epithelia of humans and animals. C. parvum infects the apical surfaces of these epithelia and remains in an intramembranous yet extracytoplasmic niche (i.e., the parasitophorous vacuole). Infection typically causes a self-limited diarrheal disease in immunocompetent individuals; however, infection in immunocompromised patients causes potentially life-threatening disease (13, 54). Indeed, C. parvum is the single most common identifiable pathogen in the biliary tracts of immunosuppressed patients, found in up to 65% of patients with AIDS cholangiopathy, a disease state characterized by biliary obstruction resulting from infection-related strictures of the biliary ducts (9, 14). The molecular mechanism of C. parvum entry into host cells is not fully understood; however, it has become increasingly clear that the host actin cytoskeleton and membrane protrusions are involved.
Several studies have investigated the role of actin and actin-binding proteins during host cell invasion by C. parvum sporozoites. Initially, it was shown that C. parvum induces actin polymerization at sites of infection by utilizing the actin branching and nucleation machinery of the Arp2/3 complex of proteins (12, 21). Additionally, it was determined that several intracellular signaling cascades are involved in C. parvum-induced actin reorganization, including c-Src/cortactin (11) and phosphoinositol-3 kinase (PI3K)/Cdc42 pathways (12). Indeed, functional inhibition of these pathways diminishes C. parvum-induced actin reorganization, host cell membrane protrusion, and, consequently, parasite internalization and parasitophorous-vacuole formation. Thus, it appears that parasite interactions with the apical surfaces of host cells initiate signaling cascades culminating in N-WASP activation with resultant Arp2/3 nucleation and branching of actin filaments, which facilitate efficient membrane protrusion and parasite entry into host cells.
In addition to actin remodeling, a rapid increase of local cell volume occurs at the region of membrane protrusions (i.e., lamellipodia, membrane ruffles, and filopodia). Cells regulate volume primarily by adjusting membrane permeability to water and ions (36). Aquaporins (Aqps) are channel proteins selective for the movement of water and other small nonionic molecules across membranes (1). In response to gradients established by transport of osmotically active molecules across cell membranes, Aqps permit water to rapidly cross the plasma membranes of a wide variety of cell types, including cholangiocytes (41). This rapid movement of water has been implicated in lamellar extensions (37) as well as directed cell migration (8, 28, 47, 48, 55). Additionally, we have recently shown that both sodium glucose cotransporter 1 (SGLT1) and Aqp1 accumulate at C. parvum invasion sites and participate in a localized, glucose-driven, Aqp1-mediated water influx that promotes efficient membrane protrusion events during C. parvum invasion (16). The molecular mechanisms regulating the trafficking and insertion of these transporters and channels at C. parvum invasion sites has not been addressed.
Nonmuscle myosin II localizes to the cortical actin cytoskeleton and is required for cell motility, cytokinesis, morphology maintenance, and vesicular trafficking in several cell types. Myosin II functions as a heterohexamer consisting of two heavy chains and two pairs of light chains (19, 39). The heavy chain consists of a globular head domain that binds and hydrolyzes ATP and an extended tail that interacts with other heavy chain tails to form a rigid bipolar rod-like structure. Phosphorylation of the myosin light chains (MLCs) plays a critical regulatory role in the contractile forces generated by the myosin head domain in nonmuscle cells (17). Three nonmuscle myosin II isoforms have been identified in epithelial cells: IIA, IIB, and IIC (26, 46). These isoforms share sequence similarity (64% to 80%) but have different enzymatic/biochemical properties (32, 35). As a result, the different nonmuscle myosin II heavy chains may have unique (7, 32, 33, 49, 52) or overlapping (3, 4) roles in regulating cell motility, cytokinesis, morphology maintenance, and vesicular trafficking.
In the present article, we clarify a role for nonmuscle myosin II during C. parvum invasion of cholangiocytes. We demonstrate that (i) the myosin II inhibitor blebbistatin diminishes C. parvum invasion, (ii) human cholangiocytes (i.e., H69 cells) express myosins IIA and IIB but not myosin IIC, (iii) myosin IIB and phosphorylated myosin light chain localize to C. parvum invasion sites, (iv) selective depletion of myosin IIB diminishes C. parvum invasion, and (v) myosin IIB inhibition with blebbistatin inhibits SGLT1 accumulation and parasitophorous-vacuole formation at invasion sites. Thus, we have defined a novel role for myosin IIB in the aggregation of SGLT1 at regions of C. parvum internalization, a process that facilitates the formation of the parasitophorous vacuole on the apical surfaces of cholangiocytes and is likely relevant to epithelial membrane protrusion in general.
H69 cells (a gift from D. Jefferson, Tufts University, Boston, MA) are simian virus 40 (SV40)-transformed normal human cholangiocytes originally derived from normal liver harvested for transplant (27). C. parvum oocysts of the Iowa strain were purchased from a commercial source (Bunch Grass Farm, Deary, ID).
To assay the attachment and invasion of C. parvum to H69 cells, two in vitro models were used as previously described: an attachment model and an attachment/invasion model (15). For the attachment model (i.e., no invasion), H69 cells were seeded onto 4-well chamber slides, grown to 70% to 80% confluence, and fixed with 4% paraformaldehyde (Electron Microscopy Sciences) in phosphate-buffered saline (PBS) before exposure to C. parvum sporozoites. Infection with C. parvum was done in culture medium containing Dulbecco's modified Eagle's medium-Ham's F-12 medium (DMEM-F12), 100 U/ml penicillin, and 100 μg/ml streptomycin (Invitrogen). Before infection of cells, oocysts were treated with a 10% bleach solution, washed, and excysted to release infective sporozoites. Cells were infected using a 1:1 sporozoite/H69 cell ratio for 2 h. For the attachment/invasion model, H69 cells were seeded onto 4-well chamber slides and grown to 70% to 80% confluence. For myosin inhibition assays, the H69 cells were preincubated for 1 h with the myosin II inhibitor blebbistatin (50 μM), a high-affinity noncompetitive inhibitor of myosin II, or the myosin light chain kinase (MLCK) inhibitor 1-(5-iodonaphthalene-sulfonyl)-1H-hexahydro-1,4-diazepine (ML-7; 20 μM). Both inhibitors were used within their effective doses and in accordance with the manufacturers' recommendations. To assess for the cytotoxic effects of these inhibitors, we performed trypan blue exclusion assays and monitored nuclear condensation, indicative of apoptosis, using DAPI (4′,6-diamidino-2-phenylindole) staining. The concentrations used exhibited no cytotoxic effect on the H69 cells throughout the duration of our experiments. Attachment and attachment/invasion assays were performed after a wash with DMEM-F12. Following the 2-h infection, parasites were detected by immunofluorescence with a polyclonal antibody (Ab) against C. parvum (45). Parasites attached to prefixed cells or nonfixed cells were counted, and the results were expressed as attachment rate or attachment/invasion rate (number of parasites/total number of cells × 100), respectively. Experiments were performed in triplicate, and up to 1,000 cells were counted for each assay. All values are given as means ± standard errors (SE). Means of results for groups were compared by analysis of variance (ANOVA). P values of <0.05 were considered statistically significant.
Sporozoites were resuspended in infection medium at 107 sporozoites/ml in the presence or absence of blebbistatin (50 μM) or 2,3-butanedione monoxime (2,3-BDM; 10 mM) and then cytospun onto poly-l-lysine-coated glass slides at 1,000 rpm for 3 min at room temperature. The slides were maintained for 5 min at 37°C and were fixed immediately with a solution containing 0.1 M 1,4-piperazinediethanesulfonic acid (pH 6.95), 1 mM EGTA, 3 mM magnesium sulfate (Sigma-Aldrich), and 2% paraformaldehyde at 37°C for 10 min. After membrane permeabilization with 0.2% (vol/vol) Triton X-100 in PBS, the fixed sporozoites were incubated with the CP2 antibody to label the sporozoites, and short motility trails were monitored by use of a fluorescein-labeled secondary antibody (Molecular Probes). Labeled slides were assessed with a Zeiss 510 confocal microscope (Carl Zeiss, Inc.). The length of the motility trail of each sporozoite was then measured with an LSM 510 analysis system (Carl Zeiss, Inc.). Up to 100 randomly selected sporozoites were measured for each group. Gliding motility was expressed in length of motility trail (μM). Morphological analyses of the parasites were also performed in the presence or absence of blebbistatin (50 μM) or 2,3-BDM (10 mM), a low affinity, noncompetitive inhibitor of multiple myosins with an effective dose in the low millimolar range. Sporozoites were cytospun onto poly-l-lysine-coated glass slides and immediately fixed. Phase-contrast images of the parasites were obtained at ×400 magnification, and both length and width were measured using a micrometer and Adobe Photoshop software. Data are presented as mean length/width ratios ± SE of results from more than 100 sporozoites.
Reverse transcription-PCR (RT-PCR) analyses were performed to detect nonmuscle myosin II isoforms expressed in H69 cells. Total RNA from cultured HT-29 (control) and H69 cells were isolated using the conventional method of acid phenol-chloroform extraction using TRI reagent (Sigma-Aldrich) and subsequent alcohol precipitation. Total RNA (1 μg) was reverse transcribed to cDNA by using a Moloney murine leukemia virus (M-MLV) reverse transcriptase kit (Invitrogen). Isoform-specific amplification was performed using the following primer pairs: for nonmuscle myosin IIA, 5′-AAGGATGATGTGGGCAAGAG-3′ (forward) and 5′-TCCTTCAGGTCCATCTCCAG-3′ (reverse); for nonmuscle myosin IIB, 5′-TCGCACTGAGAAGAAGCTGA-3′ (forward) and 5′-TCTGTGTCATCGTCGGAGAG-3′ (reverse); and for nonmuscle myosin IIC, 5′-ATTCTGGAGGAGAAGCGTCA-3′ (forward) and 5′-CCAGCTTTCCAGAGAGGATG-3′ (reverse). Amplification was performed at 94°C for 1 min, 52°C for 1 min, and 72°C for 1 min for 30 cycles. PCR products were separated by electrophoresis and confirmed by sequencing (Mayo Clinic Molecular Core Facility).
H69 cells were grown either to 70% confluence or to confluence in the presence or absence of C. parvum as described above. Cells were then fixed with 2% paraformaldehyde and permeabilized with 0.2% (vol/vol) Triton X-100 in PBS. A monoclonal antibody against myosin IIB (CMII-23; Developmental Studies Hybridoma Bank, Iowa City, IA), a polyclonal antibody against myosin IIA (PRB-440P; Covance, Inc.), and subsequent incubation with fluorophore-conjugated anti-mouse or anti-rabbit secondary antibodies (Molecular Probes) were used for colocalization of these myosin isoforms in subconfluent cells. For studies addressing accumulation of myosin isoforms and serine-19-phosphorylated myosin light chain (MLC2) to C. parvum invasion sites, infected cells were fixed as described above and incubated with monoclonal antibodies against myosin IIB (CMII-23), pMLC2 (sc-19849-R; Santa Cruz Biotechnology), and the polyclonal anti-C. parvum antibody Cp2, followed by incubation with fluorophore-conjugated secondary anti-mouse and anti-rabbit antibodies. Localization of myosin IIA and Sglt1 in infected cells was performed with the polyclonal antibodies PRB-440P and SG13-A (Alpha Diagnostics), respectively. Infection sites were identified either by phase-contrast microscopy and verified with DAPI staining of parasite and host cell nuclei or by use of the monoclonal C. parvum antibody 2H2 (Immucell). Slides were then mounted with mounting medium (H-1000; Vector Laboratories) and analyzed with a Zeiss LSM510 confocal microscope.
For scanning electron microscopy (SEM) and transmission electron microscopy (TEM), cells were grown to 70 to 80% confluence on cell culture inserts (Becton Dickinson). The cells were treated with dimethyl sulfoxide (DMSO) as a vehicle control or with blebbistatin as described above, washed, and exposed to viable C. parvum sporozoites for 2 h. Cells were prepared for electron microscopy (EM) as described in the product literature (technical bulletins 405 and 406), with critical-point drying for SEM and embedding and sectioning for TEM. For immunogold labeling, cells were fixed and processed according to a previously described protocol (21, 22). The relative distribution of Sglt1 was determined by counting gold particles over cell profiles. Invasion site-associated labeling for both control and blebbistatin-treated cells was quantitated by counting the gold particles in the area within a 0.2-μm distance along the dense-band or vacuole membrane; totals were described as numbers of gold particles per invasion site. Gold particles were counted from a total of >25 parasite invasion sites for each condition. Data represent three independent experiments.
ON TARGETplus SMARTpool short interfering RNAs (siRNAs) to nonmuscle myosins IIA (myh9) and IIB (myh10) were purchased (Thermo Scientific, Dharmacom RNAi Technologies). For siRNA transfection, H69 cells were grown to 60 to 70% confluence in 10-cm dishes (for siRNA validation) or in a 4-well chamber slide (for attachment/invasion assays) and were transfected at a final concentration of 100 nM with siPORT lipid transfection agent (Ambion). A siRNA with limited sequence homology to human transcripts was used as a control (Ambion). Knockdown of respective myosin isoforms was validated by Western blotting as described below, and attachment/invasion assays were performed by immunofluorescence using the polyclonal anti-C. parvum antibody as described above.
H69 cells were lysed with the M-PER mammalian protein extraction reagent (Pierce), and protein concentrations were determined using Bradford reagent according to the instructions of the supplier (Sigma-Aldrich). Ten micrograms of lysate protein per lane was separated by SDS-PAGE under reducing conditions and blotted onto nitrocellulose membranes. For siRNA validation experiments, membranes were incubated with the primary antibodies to nonmuscle myosins IIA and IIB (PRB-440P and PRB-445P, respectively; Covance) and then with 0.2 μg/ml horseradish peroxidase (HRP)-conjugated secondary Ab and revealed with enhanced chemiluminescence (ECL) light substrate (ECL; Amersham Biosciences). For detection of phospho-MLC2, H69 cells were grown to 95% confluence and exposed to C. parvum sporozoites. Cells were lysed as described above in the presence of PhosStop phosphatase inhibitor cocktail (Roche). Western blotting was performed as described above, using the phospho-MLC2 specific monoclonal antibody sc-19849-R (Santa Cruz Biotechnology).
It was previously demonstrated that pretreatment of bovine fallopian tube epithelial cells with the myosin light chain kinase inhibitor ML-7 or the broad-range myosin inhibitor 2,3-BDM diminished C. parvum infection (24). Here, we assessed the effects of pretreatment of cultured human cholangiocytes with ML-7 (20 μM) or the selective myosin II inhibitor blebbistatin (50 μM). DMSO-treated control cells were readily infected with C. parvum sporozoites (Fig. (Fig.11 A). In contrast, pretreatment of cultured cholangiocytes with either blebbistatin (Fig. (Fig.1B)1B) or ML-7 (Fig. (Fig.1C)1C) significantly diminished the number of parasites detected by fluorescence after a 2-hour infection. An attachment assay, where the cells were treated as described above and fixed with 4% paraformaldehyde prior to exposure to parasite (10), was performed to assess the effects of these cytoskeletal inhibitors on parasite attachment to the apical membranes of cholangiocytes. Pretreatment of cholangiocytes with blebbistatin had no effect on parasite attachment, while ML-7 pretreatment reduced attachment by nearly 50% (Fig. (Fig.1D),1D), suggesting that residual ML-7 may affect parasite processes required for attachment to the host cell membrane. While blebbistatin did not affect attachment to the cholangiocyte membrane, invasion was reduced nearly 10-fold (Fig. (Fig.1D).1D). Others have demonstrated that the broad-range, low-affinity myosin inhibitor 2,3-butanedione monoxime (2,3-BDM) and the myosin light chain kinase inhibitor ML-7 inhibit C. parvum gliding motility and alter sporozoite morphology, processes that require parasite actin and myosin (25). We therefore assessed whether blebbistatin (50 μM), a small molecule inhibitor with specificity for myosin II, affected the parameter values of C. parvum sporozoite morphology or motility, both characteristics dependent on a functioning parasite actomyosin cytoskeleton. As shown previously, incubation of sporozoites with 2,3-BDM (10 mM) resulted in rounded sporozoites while blebbistatin had no effect on parasite morphology (presented as length/width ratio) (Fig. (Fig.22 A and B). Correspondingly, 2,3-BDM inhibited parasite motility, while blebbistatin-treated parasites exhibited motility trails, indicative of gliding cell motility, that were no different form control, untreated parasites (Fig. 2C and D). Therefore, while treatment of host cells with 50 μM blebbistatin inhibited invasion, this same concentration does not appear to inhibit parasite myosin, which is categorized within the highly divergent class XIV myosins (30).
Given that blebbistatin does not inhibit the contractility of parasite myosins required to maintain sporozoite morphology and motility and that this molecule preferentially inhibits myosin II molecules, we reasoned that host cell myosin II was required for parasite internalization. Three nonmuscle myosin II molecules have been identified in human epithelial cells: myosins IIA, IIB, and IIC. RT-PCR was utilized to identify which myosin II molecules are expressed in cultured human cholangiocytes (H69). H69 cells express both myosin IIA and myosin IIB, yet we were unable to detect myosin IIC. The human colon adenocarcinoma cell line HT-29, which express all three myosin II isoforms, was used as a positive control (Fig. (Fig.33 A). Immunofluorescence confocal microscopy demonstrates that myosins IIA and IIB exhibit differential distributions within subconfluent cells. Within these subconfluent cells, myosin IIB localized to the periphery and throughout the cells, while myosin IIA preferentially localized to the periphery (Fig. (Fig.3B).3B). However, within confluent cells, myosin IIA preferentially localized to stress fibers, while myosin IIB localized to the peripheries of cells (Fig. (Fig.44 A and C). Furthermore, within C. parvum-infected H69 cells, myosin IIA remained associated with stress fibers (Fig. (Fig.4A),4A), while myosin IIB exhibited localized aggregation at regions of parasite internalization (Fig. (Fig.4C).4C). Furthermore, phosphorylation of myosin light chain plays a critical regulatory role in the contractile forces generated by nonmuscle myosins (17). We therefore assessed the intracellular localization of phosphorylated myosin light chain (phospho-S19) in infected H69 cells. A localized aggregation of phospho-MLC was detected at more than 90% of the invasion sites, as assessed by confocal colocalization. As determined by optical sectioning and confocal x-z plane analysis, the phospho-MLC aggregated to a region subjacent to regions of parasite interactions with the plasma membrane (Fig. (Fig.55 A to C). Furthermore, infection of cultured cholangiocytes resulted in an increase in the total detectable levels of myosin light chain phosphorylation (Fig. (Fig.5D5D).
Having demonstrated that the myosin II inhibitor blebbistatin diminishes C. parvum invasion and that myosin IIB localizes to infection sites, we assessed whether selective depletion of either myosin IIA or myosin IIB inhibited C. parvum infection of cholangiocytes. Western blots were performed on siRNA-transfected H69 cells. H69 cells transfected with the control siRNA did not diminish myosin IIA or IIB expression, while individual siRNAs targeting either myosin IIA or myosin IIB isoforms exhibited selective depletion of the respective targeted myosin II isoforms (Fig. (Fig.66 A). An immunofluorescence infection assay was performed on those cells transfected with the control siRNA (Fig. (Fig.6B)6B) or cells transfected with the myosin IIA (not shown) or IIB (Fig. (Fig.6C)6C) siRNA. In support of the Western blot data, the H69 cells transfected with the nonspecific siRNA did not diminish myosin IIB expression, and cells were readily infected. In contrast, those cells transfected with the myosin IIB siRNA exhibited decreased myosin IIB-specific immunofluorescence and fewer parasites were detected (Fig. (Fig.6C).6C). The number of parasites detected by immunofluorescence was quantified and is presented as number of parasites per 100 cells (Fig. (Fig.6D).6D). Myosin IIA depletion did not significantly diminish the number of C. parvum invasion sites, while myosin IIB depletion reduced infectivity by approximately 60% (P < 0.05).
We have previously demonstrated that a localized water influx associated with C. parvum induced aggregation of sodium glucose cotransporter 1 (SGLT1) and that aquaporin 1 promotes the efficient internalization of C. parvum sporozoites (16). We therefore asked whether inhibition of myosin II contractility with blebbistatin diminishes the localized aggregation of SGLT1. Confocal microscopy was used to colocalize SGLT1 and C. parvum (Fig. (Fig.77 A) in blebbistatin (50 μM)-treated or control H69 cells. SGLT1 localized to over 90% of the invasion sites in vehicle-treated control cholangiocytes, while minimal SGLT1 localized to invasion sites of blebbistatin-treated cholangiocytes (Fig. (Fig.7B).7B). Optical sectioning and confocal x-z plane analysis were used to demonstrate that SGLT1 abuts the parasite host cell interface in vehicle-treated cholangiocytes (Fig. (Fig.7A,7A, inset). Electron microscopic analyses of infection sites from control and blebbistatin (50 μM)-treated cholangiocytes demonstrates differences in the composition of the parasitophorous vacuole membrane (PVM). Following a 2-hour infection of control cells, the parasite is encapsulated in a bimembrane parasitophorous vacuole on the surfaces of H69 cells (Fig. (Fig.88 A). In blebbistatin-treated cells, however, a single, discontinuous membrane was evident (Fig. 8B and F). Indeed, 76% (35/46) of the observed infection sites under control conditions were fully internalized with a bimembrane parasitophorous vacuole membrane, while fewer than 10% (2/27) of the infection sites in blebbistatin-treated cells were fully internalized. A well-formed dense band was evident in over 50% of the invasion sites in both control cells and blebbistatin-treated cells, suggesting that myosin II is not required for the formation of this structure. No obvious ultrastructural differences in the region of the plasma membrane between vehicle (Fig. (Fig.8C)-8C)- and blebbistatin (Fig. (Fig.8D)-treated8D)-treated uninfected cells were noted. Immunogold electron microscopy was used to localize SGLT1 at invasion sites in control and blebbistatin-treated cells. In accordance with our confocal images, SGLT1 localized to invasion sites in control infected cells and was associated with dramatic host cell membrane rearrangements (Fig. (Fig.8E),8E), while few immunogold particles were observed at infection sites in blebbistatin-treated cells (Fig. (Fig.8F).8F). Quantitation of the gold particles (Fig. (Fig.8G)8G) revealed a significant decrease in the aggregation of SGLT1 at sites of C. parvum invasion following blebbistatin treatment.
The results of our study provide the first direct evidence that myosin IIB is required for efficient C. parvum invasion of cholangiocytes and parasitophorous-vacuole formation. Using an in vitro model of biliary cryptosporidiosis, we have shown that (i) the myosin II inhibitor blebbistatin diminishes C. parvum invasion, (ii) H69 cells express myosins IIA and IIB but not myosin IIC, (iii) myosin IIB and phosphorylated myosin light chain localize to C. parvum invasion sites, (iv) selective depletion of myosin IIB diminishes C. parvum invasion, and (v) myosin IIB inhibition with blebbistatin inhibits SGLT1 accumulation and parasitophorous-vacuole formation at invasion sites. Our findings demonstrate a novel role for myosin IIB during C. parvum invasion involving the localized accumulation of SGLT1 and parasitophorous-vacuole formation, processes that may be associated with apical-membrane protrusion events in general.
The luminal surfaces of epithelial cells are often the point of entry for intracellular pathogens. A common mechanism of entry into these nonphagocytic cells is through modulation of the host actin cytoskeleton. Immediately following oocyst excystation, C. parvum sporozoites exhibit parasite-actin-dependent gliding cell motility, a method of locomotion conserved in the Apicomplexa. Through sequential gliding, attachment, and host membrane alterations, C. parvum sporozoites are ultimately encapsulated in the host-derived membrane. The molecular mechanism of C. parvum entry into host cells is not fully understood; however, in addition to parasite-specific processes (56), it has become increasingly clear that the host actin cytoskeleton and membrane protrusions are involved.
Previous C. parvum invasion studies, using the myosin light chain kinase inhibitor ML-7 as well as the myosin II ATPase inhibitor 2,3-BDM, demonstrated a potential role for myosin function in parasite internalization processes (24). Our current studies support the observations of the inhibitory effects on invasion of those myosin inhibitors and add that pretreatment of host H69 cells with the myosin II-specific inhibitor blebbistatin significantly inhibits parasite internalization; sporozoite motility, which depends on the parasite actomyosin cytoskeleton, is unaffected. Blebbistatin has a high affinity for class II myosins and blocks myosin motor function by maintaining the myosin head in an actin-detached state (35). Three nonmuscle myosin II isoforms have been characterized; these isoforms are composed of myosin IIA-C heavy chains and are encoded by the genes MYH9, MYH10, and MYH14. Our cultured cholangiocytes expressed only myosins IIA and IIB. Despite a high degree of homology in amino acid sequence (26), myosins IIA and IIB have been shown to have differences in ATPase activity and subcellular localization (6). Indeed, within our subconfluent cells, myosin IIB localized throughout the cell as well as at the cell periphery, while myosin IIA preferentially localized to the periphery. However, in confluent cells, myosin IIB preferentially localized to the periphery, while myosin IIA localized predominantly to stress fibers. On the basis of this observation, we speculate that these highly similar proteins perform distinct cellular processes within cholangiocytes. In our model of biliary cryptosporidiosis, cholangiocyte myosins IIA and IIB again exhibited differential localization patterns, and only myosin IIB, with associated phosphorylated myosin light chain, localized to sites of C. parvum internalization. Moreover, selective depletion of myosin IIB diminishes C. parvum invasion, while depletion of myosin IIA had no effect on parasite internalization. It seems likely, therefore, that C. parvum induced actin reorganization, and the recruitment of actin-binding proteins, including myosin IIB, is required for parasitophorous-vacuole formation.
In a previous study, we determined that SGLT1 and Aqp1, a channel protein selective for the movement of water and other small nonionic molecules (1), accumulate at C. parvum invasion sites and participate in efficient membrane protrusion events induced by the parasite (16). Concordantly, the region of attachment displays localized glucose-driven water influx associated with membrane protrusion events and rapid internalization of the parasite. The overall rate of membrane protrusion depends on both the actin polymerization rate and the increase in localized cell volume (36). Thus, an important role for Aqps and solute transporters in membrane remodeling associated with invasion of host cells by C. parvum is likely. Cholangiocytes exposed to blebbistatin, with subsequent exposure to C. parvum, exhibited decreased accumulation of SGLT1 at infection sites (Fig. (Fig.4B),4B), suggesting that the contractile activity of myosin IIB is required for transporter aggregation and membrane remodeling at sites of C. parvum invasion. Nonmuscle myosin II localizes to the cortical actin cytoskeleton and is required for cell motility, cytokinesis, and morphology maintenance in several cell types. Moreover, several recent reports describe a functional role for nonmuscle myosin II during (i) regulated exocytosis in chromaffin cells (42), (ii) glucose uptake in adipocytes (51), and (iii) trafficking of bile salt exporters to the apical domain of hepatocytes, a process involving a direct interaction with the myosin II regulatory light chain MLC (10). Furthermore, myosin IIB was implicated in insulin-like growth factor 1-induced membrane ruffling and macropinocytosis in a neuroblastoma cell line (34). Therefore, in addition to the classical roles of myosin II, this molecule is likely involved in numerous process associated with plasma membrane alterations, including localized protrusive events. The mechanism by which C. parvum induces myosin II aggregation and contractility remains to be determined. A recent study has implicated protein kinase C (PKC) in C. parvum entry into primary bovine and human intestinal cells as well as Cryptosporidium hominis invasion of primary human intestinal cells (29). In the same study, inhibition of MLCK had no effect on parasite invasion. A likely explanation for the differences in our studies is the use of different cell types and differences in the duration of our invasion assays. Interestingly, PKC activation has been shown to promote the phosphorylation of MLC, likely through the inhibition of myosin phosphatase activity (22, 23, 50). Given that the phosphorylation status of MLC depends on the concerted effects of kinases and phosphatases (57), future investigations into the molecular mechanisms of localized activation of myosin IIB and the way that this activation promotes invasion for both C. parvum and C. hominis are merited. Our recent findings provide a unique model for testing the role of myosins during C. parvum invasion as well as localized water influx in microbial-stimulated membrane protrusions, a phenomenon relevant to the molecular mechanisms of apical-membrane modifications in general.
The electron microscopic ultrastructural details of zoite attachment and invasion are well characterized (2, 31, 38, 40, 44). Upon internalization, the zoites are encapsulated in a bimembrane structure on the surfaces of epithelial cells. The bimembrane structure encapsulating the parasite is composed of the outer and inner parasitophorous vacuole membranes (PVMs), which are derived from the host, and between which is a thin layer of cytoplasm (18, 38, 40). The bimembrane structure links over the top of an invading parasite (53) and surrounds the parasitophorous vacuole that separates the outer parasite pellicle membrane from the host-derived membranes. As the invasion process ensues, a unique structure is formed at the base of the host-parasite interface, containing electron-dense material (dense band) with an adjacent filamentous actin network (5, 20). Ultrastructural analyses of infected cholangiocytes following blebbistatin treatment revealed that the bimembrane structure comprising the PVM is formed at few invasion sites (<10%). Rather, a discontinuous membranous structure localizes around attached parasites, suggesting that the host cell-derived membranes have failed to surround the parasite or the membrane has lost structural integrity. While the precise role of myosin IIB during C. parvum invasion is currently unknown, it is reasonable, given the demonstrated roles of myosin II in the regulated exocytosis in chromaffin cells, that myosin IIB may function in the exocytic insertion of transporters and channels that facilitate membrane protrusion events. Additionally, C. parvum invades cholangiocytes via sphingolipid-enriched microdomains (43). Given that myosin IIB localizes to regions of parasite internalization and that myosin II is a component of the cortical membrane cytoskeleton, we propose that inhibition of this molecular motor results in membrane microdomain disorganization and hence inhibits parasitophorous-vacuole formation. Interestingly, myosin II is not required for parasite attachment or dense-band formation, as demonstrated by the attachment studies and EM analyses. Hence, myosin II inhibition blocks the invasion process at a point after parasite attachment and before internalization, a characteristic that can be utilized to deduce host cell processes and signaling cascades that are initiated following parasite internalization.
In summary, using a cell culture model of biliary cryptosporidiosis, we have identified a functional role for myosin II during the C. parvum invasion process. We have demonstrated that myosin IIB is required for efficient parasite internalization and associated accumulation of solute transporters involved in efficient membrane protrusion events. It will be of interest to extend these studies to address the signaling cascades required for myosin IIB localization and activation and determine how these processes promote transporter aggregation at sites of C. parvum internalization. Furthermore, identification and characterization of host molecules required for C. parvum internalization not only may identify novel therapeutic targets for cryptosporidiosis but will provide insight into dynamic cholangiocyte apical-membrane remodeling in general.
We thank Deb Hintz for secretarial assistance. The monoclonal antibody against cytoplasmic nonmuscle myosin IIB (CMII 23) developed by G. W. Conrad and A. H. Conrad was obtained from the Developmental Studies Hybridoma Bank, developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biology (Iowa City, IA).
This work was supported by National Institutes of Health grants DK76922 (S.P.O.) and DK57993 (N.F.L.) and by the Optical Microscopy Core of the Mayo Clinic Center for Cell Signaling in Gastroenterology (P30DK084567).
Editor: J. H. Adams
Published ahead of print on 10 May 2010.