Lipid Raft- and Src Family Kinase-Dependent Entry of Coxsackievirus B into Human Placental Trophoblasts

Abstract

Maternal-fetal transmission of group B coxsackieviruses (CVB) during pregnancy has been associated with a number of diverse pathological outcomes, including hydrops fetalis, fetal myocarditis, meningoencephalitis, neurodevelopmental delays, congenital skin lesions, miscarriage, and/or stillbirth. Throughout pregnancy, the placenta forms a critical antimicrobial protective barrier at the maternal-fetal interface. Despite the severity of diseases accompanying fetal CVB infections, little is known regarding the strategies used by CVB to gain entry into placental trophoblasts. Here we used both a trophoblast cell line and primary human trophoblasts to demonstrate the mechanism by which CVB gains entry into polarized placental trophoblasts. Our studies revealed that the kinetics of CVB entry into placental trophoblasts are similar to those previously described for polarized intestinal epithelial cells. Likewise, CVB entry into placental trophoblasts requires decay-accelerating factor (DAF) binding and involves relocalization of the virus from the apical surface to intercellular tight junctions. In contrast, we have identified a divergent mechanism for CVB entry into polarized trophoblasts that is clathrin, caveolin-1, and dynamin II independent but requires intact lipid rafts. In addition, we found that members of the Src family of tyrosine kinases were required for CVB entry. Our studies highlight the complexity of viral entry into human placental trophoblasts and may serve as a model for mechanisms used by diverse pathogens to penetrate the placental barrier.

INTRODUCTION

Group B coxsackieviruses (CVB) are nonenveloped, positive-sense, single-stranded RNA viruses belonging to the enterovirus genus of the family Picornaviridae. CVB infections are generally asymptomatic or cause mild flu-like symptoms in most healthy individuals. However, in some cases, CVB infections result in severe pathologies, such as myocarditis, pancreatitis, and/or aseptic meningitis (1, 2).

There are six serotypes of CVB (CVB1 to 6), all of which require the coxsackievirus and adenovirus receptor (CAR), a tight junction (TJ)-localized type I transmembrane protein, to enter and infect cells (3–5). A subset of CVB serotypes (CVB1, CVB3-RD, and CVB5) also utilize decay-accelerating factor (DAF or CD55) as an attachment factor (6–10). DAF, a glycosylphosphatidylinositol (GPI)-anchored surface membrane protein, localizes to the apical surfaces of polarized intestinal epithelial cells (6, 9). We previously demonstrated that a DAF-binding isolate of CVB, CVB3-RD, exploits a unique mechanism to enter polarized intestinal epithelial cells (6, 11) whereby CVB binding to DAF on the apical surface promotes the relocalization of the CVB-DAF complex to the TJ, where CVB subsequently interacts with CAR to initiate events required for entry into the cytoplasm (6, 12). In these cells, CVB enters by a lipid raft-dependent, dynamin-independent mechanism that combines features of caveolar endocytosis and macropinocytosis (6, 11). In contrast, CVB entry into polarized endothelial cells requires caveolar endocytosis, dynamin II, and unidentified Src family tyrosine kinases (SFKs) (13). In nonpolarized HeLa cells, dynamin II and lipid rafts are necessary for CVB entry, while clathrin and caveolin-1 are not (14). Taken together, these studies highlight the diverse pathways utilized by CVB to enter both polarized and nonpolarized cells.

The placenta is an organ unique to pregnancy and constitutes the major physical barrier at the maternal-fetal interface. The hemochorial placental villi are composed of terminally differentiated, multinucleated syncytiotrophoblasts, which are in direct contact with maternal blood and therefore regulate critical functions such as the exchange of gases, nutrients, and wastes between the mother and fetus. Syncytiotrophoblasts also produce essential hormones and mediate fetal immunological defense. Progenitor cytotrophoblasts underlie the syncytiotrophoblasts and contact fetal endothelial cells. Together, these cells form an innate feto-placental defense unit against viral invasion.

Maternal CVB infection and consequential fetal transmission have been associated with severe pathological outcomes, including congenital skin lesions (15), the development of type I diabetes (16, 17) and thyroiditis (18) later in life, hydrops fetalis (19), fetal myocarditis (20–22), meningoencephalitis (23), neurodevelopmental delays (24), fetal sepsis (25), miscarriage (26), and stillbirth (27, 28). Currently, testing for enterovirus infections during pregnancy is not routine, and thus there is a lack of data regarding their prevalence. However, the consequential detrimental fetal pathologies of undetected and untreated in utero enterovirus infections further underscore the need to advance our understanding of host-pathogen interactions at the maternal-fetal interface.

In this study, we examined the mechanism(s) by which CVB gains entry into polarized placental trophoblasts. As the trophoblast is likely the initial site of pathogen invasion across the placenta (29–31; further reviewed in reference 32), we modeled CVB entry into cytotrophoblasts using an undifferentiated human BeWo trophoblast cell line and complemented our findings using cultures of primary human trophoblasts. We show that the kinetics of CVB entry and uncoating in placental trophoblasts are similar to those previously described in polarized intestinal epithelial cells. Similarly, CVB entry into placental trophoblasts requires DAF binding and involves the relocalization of virus from the apical surface to intercellular TJs. In contrast, we have identified a divergent mechanism for CVB cytoplasmic entry into polarized trophoblasts that is clathrin, caveolin-1, and dynamin II independent but requires lipid rafts and SFK signaling. These studies highlight the strategies associated with viral entry into human placental trophoblasts and may serve as a model for how other viruses have evolved to circumnavigate the placental barrier.

MATERIALS AND METHODS

Cells and viruses.

BeWo cells were obtained from the ATCC and cultured in Ham's F-12K medium with Kaighn's modification containing 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cell monolayers were cultured on collagen-coated 8-well chamber slides (LabTek) at a density of 1.5 × 10⁵ cells/well, on 24-well plates at 2.4 × 10⁵ cells/well, on 12-mm Transwell-Col inserts (0.4-μm pore size) (Costar) at 1.25 × 10⁶ cells/well, or on collagen-coated 6-well plates at 1.7 × 10⁶ cells/well. Cells were grown a minimum of 48 h prior to study.

PHT cells were isolated from healthy singleton term placentas using the trypsin-DNase-dispase/Percoll method as described by Kliman et al., with previously published modifications (33, 34), using a protocol approved by the Institutional Review Board at the University of Pittsburgh. Cells were maintained in Dulbecco's modified Eagle medium (DMEM) (Sigma) containing 10% fetal bovine serum (FBS, Gibco) and antibiotics. Media were changed every 24 h, and cells were maintained for 72 h after plating. Cell quality was routinely monitored both morphologically (by microscopy) and by medium human chorionic gonadotropin (hCG) levels, determined by enzyme-linked immunosorbent assay (ELISA; DRG International) (33, 35). All cells were maintained at 37°C in a 5% carbon dioxide (CO₂)-air atmosphere.

Experiments were performed with coxsackievirus B3-Nancy (non-DAF binding), coxsackievirus B3-RD isolate (CVB3-RD; DAF binding) as described previously (6), green fluorescent protein (GFP)-tagged vesicular stomatitis virus (VSV) as described previously (36), or neutral red CVB3-RD (NR-CVB) as described below. CVB was propagated in HeLa cells and purified as previously described (6). GFP-VSV was expanded by growth on Vero cells, and medium was harvested. Viral titers were determined by plaque assays. Confluent monolayers were exposed to serial dilutions of virus for 1 h at 37°C (VSV) or at room temperature (CVB). Cells were then overlaid with agarose and incubated for 24 (VSV) or 48 h (CVB). Plaques were visualized by crystal violet staining and plaques enumerated.

Preparation of neutral red (NR)-labeled CVB.

CVB3-RD was grown on HeLa cells as previously described (6) in medium containing neutral red dye (10 μg/ml; Sigma) (37, 38). NR-CVB3-RD was purified, and plaque assays were performed as described above. All procedures were performed under semidark conditions.

Virus infection and entry assays.

For reverse transcription-quantitative PCR (RT-qPCR)-based infection assays, cells were infected with CVB3-RD at a multiplicity of infection (MOI) of 5 PFU per cell for 6 to 7 h. For electron microscopy (EM)-based infection assays, cells were infected with CVB3-RD (MOI = 10) for 14.5 h, fixed with 2.5% glutaraldehyde in phosphate-buffered saline (PBS) for 1 h at room temperature (RT), and then processed for electron microscopy as previously described (39). For immunofluorescence-based infection assays, virus infections were performed with CVB3-RD, CVB3-Nancy, or GFP-VSV (MOI = 5). Cellular infection occurred for 7 to 8 h at 37°C. CVB-infected cells were fixed with ice-cold 3:1 methanol-acetone for 5 min, and GFP-VSV-infected cells were fixed with 4% paraformaldehyde (PFA). For virus infection assays, cells were stained for markers of virus infection [CVB3 (VP1)] or assessed for GFP expression (GFP-VSV) as described below (see “Immunofluorescence microscopy and serial staining for virus entry”). A minimum of three independent fields per condition were captured and counted (>600 cells per field) per replicate. Infection levels are reported as the percentage of total cells that were virus positive relative to the total number of cells, determined by DAPI staining. Results obtained for inhibitor-exposed cells were normalized to results obtained for controls (no inhibitor). Infection of cells expressing wild-type or dominant negative mutants of endocytic pathways was quantified by determining the extent of colocalization between virus positive and wild-type or mutant-positive cells (as determined by GFP or HA-positive fluorescence) normalized to the total amount of transfected cells. Quantification of percent virus positive cells and colocalization was performed using ImageJ.

Virus entry assays were performed with CVB3-RD (MOI = 100 to 250 as indicated). Virus reconstituted in binding buffer (see below) was adsorbed to cells for 1 h at RT. For BeWo cells, binding buffer was F12K with 20 mM HEPES. For PHTs, binding buffer was minimal essential media (MEM) with 20 mM HEPES. Unbound virus was washed off, complete medium was added, and cells were placed at 37°C to initiate virus particle entry. Virus particle entry was stopped by fixation with 4% PFA for 10 min at RT. PFA was removed, and cells were washed with PBS. Cells were quenched with 50 mM NH₄Cl in PBS for 10 min. Virus was visualized with anti-VP1 antibody following a serial staining procedure as further detailed in “Immunofluorescence microscopy and serial staining for virus entry” below.

Neutral red virus infectious center (NRIC) assay.

Cells were pretreated with various inhibitors and then infected with NR-CVB (MOI = 5) in the presence of inhibitor for 2 h. At this time, cells were illuminated on a light box for 20 min. In parallel, monolayers were maintained in the dark to control for nonspecific effects of the pharmacological inhibitor on events unrelated to entry. Cells were then washed, trypsinized, pelleted, resuspended in drug-free medium, and plated onto naive cells (preseeded on collagen-coated 8-well chamber slides). Cells were infected for approximately 20 to 24 h, fixed, and stained for VP1. Infection levels were assessed by immunofluorescence microscopy as described in “Virus infection assays” and “Immunofluorescence microscopy and serial staining for virus entry” below.

Antibodies.

Mouse anti-enterovirus VP1 monoclonal antibody (NCL-Entero) was purchased from Novacastra Laboratories. Affinity-purified CAR-specific rabbit antibody (9) and mouse anti-DAF (clone IF7) (40) have been described and were provided by Jeffrey Bergelson (Children's Hospital of Philadelphia). Mouse anti-clathrin heavy chain (CHC) and mouse anti-caveolin-1 (Cav1) antibodies were obtained from BD Transduction Laboratories. Rabbit anti-dynamin II (Dyn II) was purchased from Abcam. Rabbit anti-GAPDH antibodies conjugated to horseradish peroxidase (HRP) were purchased from Santa Cruz Biotechnology.

Small interfering RNAs (siRNAs), plasmids, and transfections.

BeWo cells were transfected with DharmaFECT 1 (Thermo-Fisher Scientific) according to the manufacturer's protocol (final siRNA concentration was 25 to 50 nM per well) or X-treme gene HP DNA (Roche) transfection reagents. siRNAs targeting human dynamin II, CHC, and caveolin 1 (Cav-1) were obtained from Thermo-Fisher Scientific. CAR siRNA was obtained from IDT and has been described previously (6). Eps15 siRNA (5′-AAACGGAGCUACAGAUUAU-3′) was obtained from Sigma. Scrambled control siRNAs were purchased from Ambion. Dominant negative mutants of HA-tagged Eps15 and GFP-tagged dynamin II have been described previously (6). Cells were assayed 48 to 72 h posttransfection.

Pharmacological agents.

For pharmacological inhibitor studies, cells were pretreated with inhibitor for 60 min in complete medium. In contrast, for experiments involving dynasore or drugs targeting lipid rafts (cholesterol oxidase, filipin, nystatin, nystatin/progesterone, MβCD, and simvastatin), cells were incubated in complete medium containing 10% NuSerum (BD Transduction Laboratories) rather than FBS, as the presence of serum can negatively impact the efficacy of these drugs (14, 41, 42). Dynasore (25 μM), chlorpromazine (12.5 μg/ml), monodansylcadaverine (MDC; 100 μM), filipin (3 μg/ml), MβCD (5 mM), EIPA [5-(N-ethyl-N-isopropyl) amiloride; 102 μM] rottlerin (10 μM), cholesterol oxidase (4 U/ml), progesterone (20 μg/ml), simvastatin (20 μM), and PMA (phorbol 12-myristate 13-acetate; 1 μg/ml) were purchased from Sigma. Nystatin (25 μg/ml) and cytochalasin D (2.5 μg/ml) were obtained from MP Biomedicals. Genistein (74 μM), PP2 (30 μM), wortmannin (2.5 μM), nocodazole (10 μg/ml), latrunculin A (1 μM), blebbistatin (50 μM), and toxin B (1 ng/ml) were purchased from Calbiochem. A419259 (1 μM) was provided by Thomas Smithgall (University of Pittsburgh).

Immunofluorescence microscopy and serial staining for virus entry.

For CVB infections, fixed monolayers were incubated with VP1 primary antibody (1:500 in PBS), washed twice with PBS, incubated with 1:1,000 Alexa Fluor-488 secondary antibodies (Invitrogen) in PBS, washed three times with PBS, and then mounted with Vectashield (Vector Laboratories) containing 4′,6-diamidino-2-phenylindole (DAPI). For GFP-VSV infections, fixed monolayers were permeabilized with 0.25% Triton X-100 for 5 min and then mounted with Vectashield containing DAPI.

For CVB virus entry, a serial staining procedure was used to distinguish surface-associated virus under nonpermeabilized conditions from internalized virus under permeabilized conditions (43). Following fixation with 4% PFA and quenching with NH₄Cl (50 mM in PBS) (nonpermeabilizing conditions), surface-associated virus was detected with VP1 primary antibody (1:500 in PBS), washed twice with PBS, incubated with 1:1,000 Alexa Fluor-594 secondary antibodies (Invitrogen) in 10% species-specific serum–PBS, and washed three times with PBS. Cells were refixed with 4% PFA, washed, and permeabilized with 0.25% Triton X-100 for 10 min (permeabilizing conditions). Internalized virus was detected with VP1 primary antibody (1:500 in PBS), washed twice with PBS, incubated with 1:1,000 Alexa Fluor-488 secondary antibodies (Invitrogen) in 10% species-specific serum–PBS, and washed three times with PBS. Slides were mounted with Vectashield containing DAPI. The extent of virus internalization was calculated as described previously using the equation (VP1_in − VP1_out)/(VP1_out + VP1_in), where VP1_out is the relative fluorescence from the red channel and VP1_in is the relative fluorescence from the green channel.

Infection images were captured with an IX81 inverted microscope equipped with a motorized stage or with an Olympus Fluoview 1000 laser scanning confocal microscope. Images of infected cells were taken using an Olympus UPlanApo 10× 0.4 NA dry or UApo 20× 0.75 NA dry objective, whereas all other images (entry, transferrin, and dextran uptake) were taken with an Olympus UPlanFLN 40× 1.30 NA oil or UPlanSApo 60× 1.35 NA oil objective. For three-dimensional analysis, xz or yz series stacks were acquired at 0.35- to 0.5-μm intervals through the cell monolayer.

Transferrin uptake assay.

Transferrin uptake was performed essentially as previously described (14). Cells were transfected with various siRNAs and ∼48 to 72 h posttransfection were serum starved for 30 min and then incubated with 10 μg/ml transferrin conjugated to Alexa Fluor 594 (Invitrogen) for 30 min at 37°C. Cells were then washed, and surface-associated transferrin was stripped with two 5-min acid washes (0.2 N acetic acid, 0.5 M NaCl; pH 2.5) at 4°C.

Dextran uptake assay.

CVB was adsorbed to BeWo cells as described in “Virus infection and entry assays.” Following binding, cells were washed, and incubated with virus-free medium containing 70,000-MW dextran conjugated to Oregon green 488 (0.5 mg/ml; Invitrogen) and/or PMA as appropriate for various time points until fixation.

Electron microscopy.

Following CVB infection, cells were washed, fixed, and then processed for electron microscopy as previously described (39). Sections were imaged, and digital TEM images were captured using a JEOL JEM 1011 transmission electron microscope at 80 kV fitted with a bottom-mount AMT 2k digital camera (Advanced Microscopy Techniques).

Immunoblots.

Cells were grown in 24-well plates and lysates prepared with radioimmunoprecipitation assay (RIPA) buffer [50 mM Tris-HCl (pH 7.4); 1% NP-40; 0.25% sodium deoxycholate; 150 mM NaCl; 1 mM EDTA; 1 mM phenylmethanesulfonyl fluoride; 1 mg/ml aprotinin, leupeptin, and pepstatin; 1 mM sodium orthovanadate], and insoluble material was precipitated by brief centrifugation. Protein concentration of lysates was determined by BCA protein assay (Thermo Scientific). Lysates containing equal amounts of protein were loaded onto 4 to 20% Tris-HCl gels (Bio-Rad) and transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were blocked in 5% nonfat dry milk, probed with the indicated antibodies, and developed with horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology), and SuperSignal West Pico or Dura chemiluminescent substrates (Pierce Biotechnology).

RNA isolation and RT-qPCR.

For cellular mRNA analysis, total RNA was extracted using TRI reagent (MRC) according to the manufacturer's protocol. RNA samples were treated with RNase-free DNase (Qiagen). Total RNA was reverse transcribed using iScript cDNA synthesis kit (Bio-Rad). For each sample, 1 μg RNA was used for cDNA synthesis. RT-qPCR was performed using iQ SYBR green Supermix (Bio-Rad) in an Applied Biosystems StepOnePlus real-time PCR machine according to the manufacturer's instructions. Gene expression was calculated using the 2^−ΔΔCT method, normalized to human β-actin. RNA from PBMC was provided by Jeffrey Bergelson (Children's Hospital of Philadelphia). Primers used were as follows: for CVB3, 5′-ACGAATCCCAGTGTGTTTTGG-3′ and 5′-TGCTCAAAAACGGTATGGACAT-3′; for CAR, 5′-ACCTGCTAAGTTCAAGTACC-3′ and 5′-CTTCAGATTCTTCAGTAGTCC-3′; for DAF, 5′-CCCTAATCCGGGAGAAATAC-3′ and 5′-TGTACCCTGTTACATGAG-3′; for Src, 5′-CTATGACTATGAGTCTAGGACG-3′ and 5′-CTG TGTTGTTGACAATCTGG-3′; for Yes, 5′-TATGGCTGCTCAGATTGCTG-3′ and 5′-TTCAGGAGCTGTCCATTTGA-3′; for Lyn, 5′-TGTGAGAGATCCAACGTCCA-3′ and 5′-TTT GCTTTCCACCATTCTCC-3′; for Fyn, 5′-TGAACAGCTCGGAAGGAGAT-3′ and 5′-GGTTTCACTCTCGCGGATAA-3′; for Blk, 5′-ACTACACCGCTATGAATGATCGG-3′ and 5′-CTGTGACGAGTGACCTGGC-3′; for Fgr, 3′-GGACTGCAGGTACTCGAAGG-3′ and 5′-AACTACATTCACCGCGACCT-3′; for Lck, 5′-CTTCCCCACTGCAAGACAAC-3′ and 5′-GCCACCGTTGTCCAGATTAC-3′; for Hck, 5′-CCCTGTATGATTACGAGGCCA-3′ and 5′-CACTCCCCGGATTCCTCTAGG-3′; for Eps15, 5′-GATCTCTTCTCTGAAAGCTG-3′ and 5′-CTGTTGTGAATCTTGTAGGTG-3′; and for actin, 5′-ACTGGGACGACATGGAGAAAA-3′ and 5′-GCCACACGCAGCTC-3′).

Statistical analysis.

All experiments were performed at least three times. Data are presented as means ± standard deviations. Student's t test was used to determine statistical significance for virus infection assays when two sets were compared. A P value of <0.05 was considered significant.

RESULTS

BeWo cells are an appropriate model to study CVB entry into polarized trophoblasts.

To determine whether human trophoblast BeWo cells were a suitable in vitro model for investigating CVB entry into polarized placental trophoblasts, we stained for cell markers to demonstrate polarized apical-basolateral localization and appropriate asymmetric distribution of DAF and CAR. We observed distinct localization of zonula occludens-1 (ZO-1) at the apical TJ and the basolateral localization of the Na⁺/K⁺ ATPase (Fig. 1A), supporting the idea that BeWo cells are polarized epithelial cell models with the capability to form intercellular junctions (44, 45). We also found that DAF was localized at the apical surface and CAR was localized to the TJ, similar to other polarized cell types (Fig. 1B) (5, 6). In contrast to polarized intestinal epithelial cells (Caco-2), BeWo and primary human trophoblast (PHT) cells express very high levels of DAF relative to CAR, as determined by RT-qPCR (Fig. 1C) and immunoblotting (data not shown). Additionally, we found that CVB readily infects BeWo cells and forms characteristic double-membrane replication complexes in BeWo trophoblasts (Fig. 1D), similar to those detected in other cell types following picornavirus infection (46–48). These data reveal that BeWo trophoblasts are an appropriate polarized cell model for studying CVB interactions at the maternal-fetal interface.

Fig 1.

DAF is required for efficient CVB apical infection of placental trophoblasts. (A) Confocal micrographs of BeWo cells stained for the TJ marker ZO-1 (red) and the basolateral marker Na⁺/K⁺ ATPase (green) and DAPI-stained nuclei (blue). Bar = 10 μm. (B) Confocal micrographs of BeWo cells fixed and stained for DAF (red) and CAR (green). Bar = 10 μm. (C) Shown are the levels of DAF expression normalized to CAR expression within the indicated cell types as assessed by RT-qPCR. (D) CVB-infected BeWo cells demonstrate numerous double membrane vesicles by electron microscopy. (Left) Magnification, ×30,000. Bar = 500 nm. (Right) Enlarged view of double membrane structures shown on left. (E) BeWo cells were grown on collagen-coated chamber slides and then infected with either CVB3-RD or CVB3-Nancy (5 PFU/cell). Shown is the percent infected cells (VP1⁺/DAPI as assessed by IF; *, P < 0.0005). (F) BeWo cells were grown on collagen-coated Transwells and infected with CVB3-RD from either the apical or basolateral surfaces. Shown is the percent infected cells (VP1⁺/DAPI as assessed by IF; *, P < 0.0001). (G) BeWo cells were incubated with either control (mock or nonspecific MAb) or anti-DAF MAb (IF7) antibody for 60 min and then infected with CVB. RNA was isolated, and RT-qPCR was performed. Shown is the percent infection assessed by relative RNA (*, P < 0.005). (C, E to G) Data are means ±standard deviations (SD).

DAF is required for efficient apical infection of CVB in human placental trophoblasts.

We next determined whether DAF was required to facilitate apical CVB infection of BeWo cells, as has been shown for other polarized cell types (9, 13). First, we compared the efficiency of apical infection by a non-DAF binding CVB isolate (CVB3-Nancy) versus a DAF-binding isolate (CVB3-RD) (a single amino acid difference exists between CVB3-RD and CVB3-Nancy [49]). We found that CVB3-Nancy infected BeWo cells poorly from the apical surface (<10%) compared to CVB3-RD, indicating that DAF was necessary for efficient apical infection of these cells (Fig. 1E). (Unless otherwise stated, all subsequent CVB infections were performed with CVB3-RD.) Next, to determine the efficiency of CVB3-RD infection in a polarized manner, we cultured BeWo cells on collagen-coated Transwells and then initiated infection at either the apical or basolateral domain. We found CVB infection initiated at the apical surface to be less efficient (>55%) than infection at the basolateral surface (Fig. 1F). This is consistent with previous work in polarized endothelial cells (13). Finally, to further investigate the requirement of DAF in CVB apical infection of trophoblasts, we incubated cells with an anti-DAF blocking antibody (IF7) (40), a control antibody, or no antibody and then infected with CVB. We found that IF7 significantly inhibited apical CVB infection of BeWo cells (Fig. 1G). Taken together, these data indicate that DAF is critical for mediating the apical infection of CVB in placental trophoblasts.

Kinetics of CVB entry into BeWo trophoblasts.

To examine the kinetics of CVB entry into BeWo cells, we performed an immunofluorescence-based virus internalization assay that distinguishes between surface-associated and internalized virus particles. Virus was adsorbed to BeWo cells at room temperature, shifted to 37°C to synchronize subsequent entry, and fixed at various time points postentry. At 0 min postinfection (p.i.), CVB was observed localized at the apical surface (Fig. 2A). By 30 min p.i. CVB had relocated from the apical surface to the apical TJ complex (Fig. 2A). Finally, at 90 min p.i., the virus had internalized and localized to vesicular and/or perinuclear compartments (Fig. 2A). We observed ∼30% total CVB internalization into BeWo cells (Fig. 2B), which is lower than that in polarized epithelial and endothelial cells (6, 13) but consistent with the high level of DAF relative to CAR expression in these cells (Fig. 1C). Thus, the expression of CAR is likely rate-limiting for the entry of CVB into trophoblasts, as would be expected.

Fig 2.

Kinetics of CVB entry into placental trophoblasts. (A) CVB (100 to 150 PFU/cell) was bound to BeWo cells at RT, unbound virus was removed, and cells were incubated at 37°C to facilitate virus entry. Cells were fixed at the indicated time points and serially stained for virus [prior to permeabilization (VP1_out) or after permeabilization (VP1_in)]. Red or colocalized (red and green overlapping) fluorescence denotes virus bound on the cell surface; distinctively green fluorescence (no red) denotes internalized virus. DAPI is shown in blue. Insets show ×3 magnifications. (B) Quantification of the extent of CVB internalization (shown as percent total internalization) at various times following a shift to 37°C. (C) Cells were bound with CVB (100 to 150 PFU/cell), washed, and incubated at 37°C to facilitate virus entry for the indicated times prior to fixation. Cells were permeabilized and stained for virus (green) or CAR (red). (D) BeWo cells were transfected with either control or CAR siRNAs for a minimum of 48 h. Cells were bound with CVB (100 to 150 PFU/cell), washed, and incubated at 37°C to facilitate virus entry for 90 min prior to fixation. Cells were permeabilized and stained for virus (green). Confocal micrographs are shown. (E) BeWo cells were transfected with either control or CAR siRNAs for a minimum of 48 h and then infected with CVB (5 PFU/cell). Shown is the percent infected cells (*, P = 0.0001). Inset, immunoblot analysis for CAR expression (top) and GAPDH (bottom) in BeWo cells transfected with either control or CAR siRNA.

We next investigated whether CAR internalized with CVB during its entry into BeWo cells, as CAR cointernalization with CVB appears to be cell-type dependent (6, 14). We found no evidence of CAR internalization by 90 min p.i. (Fig. 2C), consistent with CVB entry in polarized intestinal cells (6). RNA interference (RNAi)-mediated silencing of CAR inhibited CVB entry (Fig. 2D) and infection (>90%) (Fig. 2E). Taken together, these data indicate that the kinetics of CVB entry into BeWo trophoblasts and the role of CAR in facilitating entry are similar to those observed in other polarized epithelial cell types (6).

Clathrin endocytosis is not required for CVB entry.

To dissect the entry mechanism by which CVB enters BeWo cells, we generated neutral-red (NR) CVB. NR is an RNA-binding dye that, when cultured in the presence of virus, becomes readily incorporated into the viral RNA during viral propagation and renders the resulting NR-containing virions light sensitive (37, 50, 51). Upon virus entry and subsequent uncoating, the NR dye disengages from the viral RNA, and replication ensues in a light-insensitive manner. This method is thus a powerful approach to separate pathways and/or molecules with specific functions in viral entry from those that also might play a role(s) in replication.

We performed modified neutral-red infectious-center (NRIC) assays (37) to assess the propensity of known pharmacological inhibitors of the clathrin endocytic pathway to inhibit CVB entry and infection in BeWo cells (a schematic of the NRIC assay is shown in Fig. 3A). Using this assay, we found that chlorpromazine, a pharmacological drug known to induce the loss of clathrin and the adaptor AP2 from the plasma membrane (52; reviewed in reference 53) significantly inhibited CVB replication under illuminated (>80%) but not nonilluminated conditions (Fig. 3B). We also tested whether monodansylcadaverine (MDC), an inhibitor of clathrin endocytosis that functions by stabilizing clathrin-coated pits at the plasma membrane (reviewed in reference 53) would affect CVB entry using the NRIC assay. Under illuminated conditions, MDC had no effect on CVB infection; conversely, we observed a partial reduction of CVB infection under nonilluminated conditions (∼35%) (Fig. 3B). As a control, we also tested the effects of chlorpromazine and MDC on vesicular stomatitis virus (VSV) infection, as VSV utilizes a clathrin-dependent pathway to enter and infect host cells (54–56). We found that both chlorpromazine and MDC significantly inhibited VSV infection, as expected (∼70%) (Fig. 3C).

Fig 3.

Clathrin-mediated endocytosis is not required for CVB entry into BeWo trophoblasts. (A) Shown is a schematic of the modified NRIC assay, which is described in detail in Materials and Methods. (B) NRIC assay in cells treated with either chlorpromazine (12.5 μg/ml) or MDC (100 μM). Shown is the percent infected cells (*, P < 0.005; **, P < 0.0001). (C) Cells were pretreated for 60 min prior to and during GFP-VSV infection with either chlorpromazine (12.5 μg/ml) or MDC (100 μM). Shown is the percent infected cells (*, P < 0.001). (D) Cells were transfected with either control or clathrin heavy-chain (CHC) siRNAs for a minimum of 48 h and then infected with CVB (5 PFU/cell). Shown are the percent infected cells (left; VP1⁺/DAPI as assessed by IF; not significant). (Right) Immunoblot analysis for CHC expression (top) and GAPDH (bottom). (E) Transferrin (red) internalization into cells transfected with control or CHC siRNAs. DAPI-stained nuclei are shown in blue. (F) Cells were transfected with either control or Eps15 siRNAs for a minimum of 48 h and infected with CVB (5 PFU/cell). Shown are the levels of Eps15 (as assessed by RT-qPCR) and the level of CVB infection both shown as the control siRNA-transfected cells (*, P < 0.005). (G) Cells were transfected with either the wild type (WT) or a dominant negative mutant of Eps15 (DN, EH21) for 48 h prior to infection with CVB (5 PFU/cell). Shown is the percent transgene-positive cells infected with CVB, shown as a percent of WT-expressing cells.

Because both chlorpromazine and MDC exhibit known pleiotropic effects (53), we confirmed our findings above using RNAi-mediated silencing of clathrin heavy chain (CHC). Transfection with the CHC siRNA had no effect on CVB infection in BeWo cells (Fig. 3D). As a control, we detected inhibition of transferrin uptake when cells were transfected with CHC siRNA (Fig. 3E). Similarly, we found that RNAi-mediated silencing of Eps15 or transfection of cells with a dominant negative mutant of Eps15, a component in the formation of clathrin-coated vesicles (57), also had no effect on CVB infection (Fig. 3F and G) but inhibited transferrin uptake (data not shown). Taken together, these data indicate that clathrin-mediated endocytosis is not required for CVB entry and infection into BeWo cells and that the inhibitory effects of chlorpromazine occur nonspecifically and/or via inhibition of other non-clathrin-dependent processes involved in CVB entry.

Dynamin II is not required for CVB entry into placental trophoblasts.

We also investigated whether the small GTPase dynamin II (Dyn II), which is critical for the endocytosis of clathrin (58) and caveolar (59, 60) vesicles, was involved in CVB entry into BeWo cells. We found that dynasore, a pharmacological inhibitor of dynamin (61), had no effect on CVB infection, while it potently inhibited VSV infection (Fig. 4A). To further validate our findings, we transfected cells with a siRNA targeting Dyn II. Transfection with Dyn II siRNA had no effect on CVB infection in BeWo cells (Fig. 4B). As a control, we also detected inhibition of transferrin uptake when cells were transfected with Dyn II siRNA (Fig. 4C). In addition, we found that transfection of cells with a well-characterized dominant negative mutant of Dyn II (K44A) (58, 60, 62) had no effect on CVB infection (Fig. 4D) but inhibited transferrin uptake (data not shown). Taken together, these data indicate that dynamin II is not required for CVB entry and infection into BeWo cells.

Fig 4.

Dynamin II is not required for CVB entry into placental trophoblasts. (A) BeWo cells were pretreated with dynasore (25 μM) for 60 min prior to and during infection with either CVB or GFP-VSV. Shown is the percent infected cells (*, P < 0.0005). (B) BeWo cells were transfected with either control or dynamin II (Dyn II) siRNAs for a minimum of 48 h and then infected with CVB (5 PFU/cell). (Left) Percent infected cells; (right) immunoblot analysis for Dyn II expression (top) and GAPDH (bottom). (C) Transferrin (red) internalization into cells transfected with control or DYNII siRNAs. DAPI-stained nuclei are shown in blue. (D) Cells were transfected with either the wild type (WT) or a dominant negative mutant (K44A) of Dyn II for 48 h prior to infection with CVB (5 PFU/cell). Shown is the percentage of transgene-expressing cells that were infected (normalized to WT).

Lipid rafts, but not caveolae, are required for CVB entry into placental trophoblasts.

To determine whether lipid rafts were required for CVB entry into placental BeWo trophoblasts, we tested the effects of a panel of lipid raft inhibitors and/or destabilizing agents in our modified NRIC assay (Fig. 5A). These agents included cholesterol oxidase, filipin, nystatin alone or in combination with progesterone, methyl-β-cyclodextrin (MβCD), and simvastatin. We found no significant inhibition of CVB infection by filipin or nystatin under either illuminated or nonilluminated conditions (Fig. 5A). In contrast, we detected significant inhibition of CVB infection under illuminated conditions in cells treated with cholesterol oxidase, nystatin/progesterone, MβCD, and simvastatin (Fig. 5A). The inhibitory effects of all of these agents were lost under nonilluminated conditions, supporting a role for lipid rafts in CVB entry (Fig. 5A).

Fig 5.

Lipid rafts, but not caveolin-1, are required for CVB entry. (A) NRIC assay in BeWo cells treated with cholesterol oxidase (4 U/ml), filipin (3 μg/ml), nystatin (25 μg/ml), a combination of nystatin (25 μg/ml) and progesterone (20 μg/ml), MβCD (5 mM), or simvastatin (20 μM). Shown is the percentage of infected cells (*, P < 0.0001). (B) Cells transfected with either control or caveolin-1 (Cav-1) siRNAs for a minimum of 48 h were infected with CVB (5 PFU/cell). Shown is the percent infected cells (left). (Right) Immunoblot analysis for Cav-1 expression (top) and GAPDH (bottom). (C) Cells were pretreated with MβCD (5 mM) for 60 min prior to and during CVB (100 to 150 PFU/cell) binding at RT. Unbound virus was removed, and cells were incubated at 37°C for 90 min, fixed, and serially stained for virus [prior to permeabilization (VP1_out) and with a green fluorophore after permeabilization (VP1_in)]. Red or colocalized (red and green overlapping) fluorescence denotes virus bound on the cell surface; distinctively green fluorescence (no red) denotes internalized virus. (D) Quantification of the extent of CVB internalization (shown as a percentage of total internalization) in mock- or MβCD-treated cells 90 min p.i.

To investigate whether caveolae were involved in CVB entry, we transfected BeWo cells with either control or caveolin-1 (Cav-1) siRNAs, and infected with CVB. Caveolae are a unique lipid raft subdomain which form morphologically distinct invaginations (50 to 100 nm) at the plasma membrane and are dependent on the integral membrane protein caveolin-1 (Cav-1) (reviewed in reference 63). We observed no significant effect of Cav-1 silencing on CVB infection (Fig. 5B), but a significant reduction of cholera toxin B (CTB) internalization under these conditions (data not shown).

We next performed a virus entry assay to examine whether MβCD specifically blocked CVB virion endocytosis. At 90 min p.i., internalized virus was observed within the cytoplasm under mock treatment conditions; however, in the presence of MβCD, virus was detected predominantly extracellularly (Fig. 5C) and there was very little detectable internalization of CVB (Fig. 5D). Based on both the NRIC and CVB entry assays, lipid raft disrupting and cholesterol-depleting agents were found to inhibit CVB internalization, suggesting that lipid rafts are necessary for CVB entry into trophoblasts. Taken together, these data indicate that CVB entry in BeWo trophoblasts is independent of caveolin-1 and caveolae. Thus, CVB entry into placental trophoblasts is clathrin, dynamin II, and caveolin-1 independent but is dependent on cholesterol-enriched lipid rafts.

Macropinocytosis is not required for CVB entry.

Macropinocytosis is a non-clathrin-, non-caveola-, dynamin-independent mechanism that requires actin for the nonspecific uptake of fluid and other cargo (reviewed in references 64 and 65). Previously, it was shown that CVB entry requires features of macropinocytosis in polarized human intestinal epithelial cells (11). Therefore, we also investigated the possible role of macropinocytosis in facilitating CVB entry into polarized trophoblasts. To examine whether macropinocytosis was involved in facilitating CVB entry into placental trophoblasts, cells were treated with EIPA and rottlerin, pharmacological inhibitors of macropinocytosis, in the NRIC assay under both illuminated and nonilluminated conditions (Fig. 6A). The pharmacological agent EIPA (ethyl isopropyl amiloride) and other amiloride derivatives are inhibitors of the epithelial Na⁺/H⁺ exchanger and are indirect inhibitors of the Rho GTPases Rac1 and Cdc42 (66–68). Rottlerin is a nonspecific inhibitor of PKC (protein kinase C) (69, 70). Rac1, Cdc42, and PKC are all necessary components of macropinocytosis. We found that EIPA and rottlerin both inhibited CVB replication under illuminated conditions using the NRIC assay (Fig. 6A). However, rottlerin also inhibited replication under nonilluminated conditions, suggesting that it also inhibited steps in the virus life cycle that occur postentry (Fig. 6A).

Fig 6.

Macropinocytosis is not involved in CVB entry into BeWo placental trophoblasts. (A) NRIC assay in BeWo cells treated with EIPA (102 μM) or rottlerin (10 μM). Shown is the percent infected cells (*, P < 0.0001). (B) BeWo cells were pretreated with cytochalasin D (CytoD; 2.5 μg/ml), latrunculin A (LatA; 1 μM), or toxin B (1 ng/ml) for 60 min prior to and during neutral red CVB infection, then shifted to 37°C to facilitate entry. Cells were illuminated 2 h p.i. to inactivate virions that had not yet entered. Shown is the percent infected cells (*, P < 0.0005). (C) Shown are confocal micrographs of BeWo cells incubated with dextran (0.5 mg/ml) in the absence or presence of CVB or in cells treated with PMA (1 μg/ml) for 60 min. Blue, DAPI-stained nuclei; green, dextran. Shown at bottom and right are xz and yz cross sections.

The actin cytoskeleton is a key mediator of macropinocytosis, particularly in the formation and closure of membrane ruffles (71). To determine if the actin cytoskeleton played a role in mediating CVB entry into BeWo placental trophoblasts, we also tested the actin-polymerizing inhibitors cytochalasin D (CytoD) (reviewed in reference 72) and latrunculin A (LatA) (72, 73). We performed a standard infection assay using NR-CVB and light illumination 2 h p.i. and found that these drugs had no significant effect on CVB infection in BeWo cells (Fig. 6B). We also investigated whether Rho GTPases were involved in CVB entry into placental trophoblasts by assessing the effects of toxin B (derived from Clostridium difficile), an inhibitor of Rho GTPases (74, 75). The Rho family of GTPases (Rho, Rac, Cdc42) mediates critical coordination and spatiotemporal regulation of cellular actin dynamics, which enable macropinocytosis (76, 77). We found that instead of inhibiting CVB, treatment with toxin B enhanced CVB infection of BeWo cells (>200%) (Fig. 6B).

Finally, we determined whether CVB internalization increased the uptake of dextran, which accompanies the induction of macropinocytosis. We did not see any evidence of increased dextran uptake in response to CVB entry (Fig. 6C), and we did not observe any evidence of macropinosome formation (as assessed by actin localization) in response to CVB entry (data not shown). In contrast, treatment of cells with PMA led to robust dextran uptake (78, 79) as expected (Fig. 6C). Taken together, these data indicate that macropinocytosis is not involved in facilitating CVB entry into placental trophoblasts and that the inhibitory effects of EIPA occur nonspecifically and/or via inhibition of other nonmacropinocytic processes involved in CVB entry.

CVB entry is dependent on Src family tyrosine kinases.

In addition to examining the endocytic route by which CVB enters, we also investigated potential signaling mechanisms utilized by CVB to enter and infect trophoblasts. We performed an initial pharmacological drug screen using the standard virus infection assay with NR-CVB and light illumination 2 h p.i. and found that the PI3K inhibitor wortmannin or the tyrosine phosphatase inhibitor pervanadate had no effect on CVB infection (Fig. 7A). In contrast, the pan-tyrosine kinase inhibitor genistein (>60%), the SFK inhibitor PP2 (∼70%), the microtubule-depolymerizing agent nocodazole (>65%), and the myosin II inhibitor blebbistatin (>85%) all inhibited infection (Fig. 7A). To further assess whether these drugs acted on CVB entry, we applied the CVB NRIC assay (Fig. 7B). Using the NRIC assay, we detected near-complete inhibition of CVB with nocodazole and blebbistatin under illuminated conditions but also observed significant inhibition (>80%) without light illumination (Fig. 7B), suggesting that the effects of these agents may occur due to inhibition of postentry events. In contrast, we found no effect of genistein or PP2 on CVB infection under nonilluminated conditions (Fig. 7B). Because PP2 may target tyrosine kinases in addition to SFKs (80, 81), we tested the effect of A419259, a specific inhibitor of SFKs (82) in the NRIC assay. We found that A419259 potently inhibited CVB replication under illuminated, but not nonilluminated, conditions (Fig. 7B). Taken together, these data point to a direct role for SFKs in facilitating CVB entry into placental trophoblasts.

Fig 7.

CVB entry is dependent on Src family tyrosine kinases. (A) BeWo cells were pretreated with genistein (74 μM), PP2 (30 μM), wortmannin (2.5 μM), pervanadate (50 μM), nocodazole (10 μg/ml), or blebbistatin (50 μM) for 60 min prior to and during neutral-red CVB infection. Cells were illuminated 2 h p.i. to inactivate virions that had not yet entered. Shown is the percent infected cells (*, P < 0.001). (B) NRIC assay in BeWo monolayers treated with nocodazole (10 μg/ml), blebbistatin (50 μM), genistein (74 μM), PP2 (30 μM), or A419259 (1 μM). Shown is the percent infected cells (*, P < 0.05). (C) Relative expression levels (ΔCt to actin) of the SFK members Src, Yes, Lyn, Fyn, Blk, Fgr, Lck, and Hck in Caco-2, BeWo, or PHT cells, or from PBMC as assessed by RT-qPCR. ND, not detected. (D) Cells were pretreated with genistein (74 μM) or PP2 (30 μM) for 60 min prior to and during binding with CVB (100 to 150 PFU/cell) at RT. Unbound virus was removed, and cells were incubated at 37°C (in the presence of drug) to facilitate virus entry. Cells were fixed at 90 min and serially stained for virus [prior to permeabilization (VP1_out) and after permeabilization (VP1_in)]. Red or colocalized (red and green overlapping) fluorescence denotes virus bound on the cell surface; distinctively green fluorescence (no red) denotes internalized virus. Insets show 2× magnifications. (E) Quantification of the extent of CVB internalization (shown as percent total internalization) in mock-, genistein-, or PP2-treated cells 90 min p.i.

SFK members exhibit cell-type-specific expression profiles. Thus, we applied RT-qPCR to determine the expression profile of the SFKs Src, Yes, Lyn, Fyn, Blk, Fgr, Lck, and Hck in trophoblasts. Similar to intestinal epithelial Caco-2 cells, we found that BeWo cells express high levels of Src, Yes, Lyn, and Fyn with no detectable expression of Blk, Fgr, Lck, and Hck, which are generally more restricted in their expression profiles (83) (Fig. 7C). In contrast, we found that peripheral blood mononuclear cells (PMBCs) expressed Blk, Fgr, Lck, and Hck, consistent with their known patterns of expression (84–86). Interestingly, we found that the pattern of SFK expression was distinct between BeWo and PHT cells, with PHT cells also expressing Fgr and Hck (Fig. 7C). We next verified that tyrosine kinases and SFKs acted specifically on CVB entry by performing an immunofluorescence-based virus internalization assay in the presence of either genistein or PP2. We found that both drugs prevented CVB entry into BeWo trophoblasts (Fig. 7D and E), supporting a role for SFKs in CVB entry into BeWo cells.

CVB entry into primary human trophoblasts.

As the studies detailed above utilized a human trophoblast cell line, we next investigated whether CVB entry into PHT cells occurred by a similar mechanism. We found that the entry kinetics of CVB into PHT cells were comparable to that in BeWo trophoblasts (Fig. 2A), with virus appearing in the cytoplasm within 90 min p.i. (Fig. 8A and B). Similar to BeWo cells, the ratio of DAF to CAR expression in PHT cells was very high (Fig. 1C) and the level of overall CVB internalization (Fig. 8B) was lower than is observed in other cells, such as polarized endothelial and epithelial cells (6, 13). We also investigated whether the cholesterol-depleting agent MβCD blocked CVB entry into PHT cells, as occurs with BeWo cells. Indeed, MβCD treatment greatly reduced CVB entry into PHT cells (Fig. 8A and B). Thus, CVB utilizes similar entry kinetics and possibly even similar endocytic pathways in both primary and a trophoblast line.

Fig 8.

CVB entry into primary human trophoblasts requires lipid rafts. (A) CVB (250 PFU/cell) was bound to primary human trophoblast (PHT) cells for 1 h at RT, unbound virus removed, and cells were shifted to 37°C to facilitate virus entry. Cells were fixed at the indicated time points and serially stained for virus [prior to permeabilization (VP1_out) and after permeabilization (VP1_in)]. Red or colocalized (red and green overlapping) fluorescence denotes virus bound on the cell surface; distinctively green fluorescence (no red) denotes internalized virus. DAPI is shown in blue. (Bottom row) In parallel, cells were pretreated with MβCD (5 mM) for 60 min prior to and during binding with CVB (250 PFU/cell) at RT. Unbound virus was removed, and cells were incubated at 37°C (in the presence of drug) to facilitate virus entry. Cells were fixed and serially stained. Insets show 3× magnifications. (B) Quantification of the extent of CVB internalization (shown as percent total internalization) into mock-treated PHT cells at 0 or 90 min p.i. or in cells treated with MβCD at 90 min p.i.

DISCUSSION

Enteroviruses have evolved mechanisms to subvert the placental protective barrier, as incidences of fetal disease and death have been reported (15–28). Here we utilized the BeWo trophoblast cell line as a model of placental trophoblasts to delineate the key cellular components that facilitate CVB entry into the trophoblast monolayer. However, as BeWo is a cell line derived from a choriocarcinoma (87, 88), which may not fully capture the biology of primary trophoblasts, we validated our key findings in PHT cells, and found strikingly similar results both in the kinetics of entry and in the lipid raft-dependent pathway utilized by the virus. Our findings therefore reveal insights into how CVB, and possibly other enteroviruses, cross the placental barrier to initiate fetal infection in utero.

We found that BeWo cells were an excellent in vitro model for studying virus-host interactions in polarized trophoblasts, as they exhibit characteristic junctional complexes, unique polarized localization of apical, junctional, and basolateral proteins, and asymmetric localization of the CVB attachment factor and receptor DAF and CAR, respectively. Similar to other polarized cell types (6, 9, 13), we found that DAF was required for apical CVB infection of BeWo trophoblasts. Furthermore, CAR was required for both CVB entry and infection in BeWo trophoblasts but did not internalize with the virus. This was also comparable to results in other polarized cell types (6) but distinct from those in nonpolarized cells (14). Analogously, we found that CVB entry into both BeWo cells and primary trophoblasts is similar to what has been observed in intestinal epithelial cells (6), with virus internalization and perinuclear localization requiring 90 min. Taken together, our data show that the requirement for DAF and the kinetics of CVB entry into polarized trophoblasts are similar to those described for polarized intestinal cells (6) but divergent from those described for nonpolarized cells (14).

To dissect the route by which CVB enters placental trophoblasts, we performed a variety of assays and techniques, including NRIC and virus entry assays, and the use of pharmacological inhibitors, dominant negative mutants, and/or siRNAs targeting cellular components involved in endocytosis and signaling. We found that CVB required a clathrin-, caveolar-, and dynamin-independent, endocytic pathway that was not consistent with macropinocytosis. This route of entry is distinct from that which has been described previously for both polarized endothelial (13) and nonpolarized cells (14), in which dynamin II is required. We also found that in contrast to CVB entry into polarized intestinal epithelial and endothelial cells (6, 13), CVB entry into polarized trophoblasts required lipid rafts (but not caveolin-1). By screening pharmacological agents that disrupt or deregulate lipid rafts (cholesterol oxidase, filipin, nystatin, nystatin/progesterone, MβCD, and simvastatin) in a modified NRIC assay, we determined that cholesterol oxidase, MβCD, nystatin/progesterone, and simvastatin all inhibited CVB entry. Notably, the pharmacological activities of filipin and nystatin alone differ from those of cholesterol oxidase, MβCD, nystatin/progesterone, and simvastatin (53, 89), which may potentially explain why these agents had no effect on CVB entry. Indeed, treatment with nystatin alone, which sequesters lipid rafts similarly to filipin, had no effect on CVB entry and inhibited entry only when used in combination with the cholesterol synthesis inhibitor progesterone. Unlike other cell types, trophoblasts exhibit specialized lipid raft domains at their apical surfaces (90), which may at least partially contribute to the inability of the lipid sequestering agents filipin or nystatin alone to disrupt CVB entry. In addition, transfection of a siRNA targeting caveolin 1, the transmembrane protein component critical for the formation of caveolae (63), had no effect on CVB infection in these cells, suggesting that lipid rafts, but not caveolae, were necessary for CVB entry into BeWo cells. This mechanism differs from CVB entry into polarized intestinal epithelial and human brain microvascular endothelial cells in which caveolin-1 was required (6, 13). Interestingly, the mechanism of CVB entry into placental trophoblasts is very similar to HIV entry into these cells, which was shown to be independent of clathrin, caveolae, dynamin II, and macropinocytosis but required free-cholesterol and lipid rafts (91).

Previous work has demonstrated a significant role for SFK members in mediating various aspects of CVB entry into polarized intestinal epithelial and human brain microvascular endothelial cells (6, 13). Additionally, SFKs appeared to be important for early stages of CVB infection in nonpolarized cells but were not required specifically for virus entry (14). By applying three pharmacological inhibitors—genistein (a pan-tyrosine kinase inhibitor) (92), PP2 (a SFK inhibitor that exhibits activity against other tyrosine kinases such as Abl) (80, 81), and the broad-spectrum SFK inhibitor A419259 (82)—we found that SFKs were critical for CVB entry into BeWo trophoblasts. Clearly, the inclination of CVB to usurp this cellular signaling pathway to promote its invasion into the host is a common strategy used across multiple polarized and nonpolarized cell types and may function as a consequence of DAF-mediated signals generated by receptor clustering (6). Due to the association between DAF, SFKs, and lipid-rich raft domains, this signaling pathway may be commonly manipulated by DAF-binding viruses such as CVB (93, 94).

BeWo cells express the ubiquitous SFKs Src, Yes, Lyn, and Fyn but not the more restricted members of the SFK family Blk, Lck, Fgr, and Hck. This pattern of expression is distinct from that in PHT cells, which also express Fgr and Hck. The precise cellular molecules targeted by SFKs to promote CVB entry into trophoblasts remain unclear, but SFKs target a diverse array of cellular substrates involved in a myriad of cellular processes. In addition, SFKs exhibit redundancy and/or opposing functions, which complicates the identification of the specific SFK family member(s) responsible for the entry of CVB into trophoblasts. Further studies are thus required to dissect the specific host cell components, and specific SFK family members, involved in CVB entry into polarized trophoblasts.

At this time, our understanding of the molecular events associated with pathogen entry into placental trophoblasts is limited. Elucidation of the endocytic and/or host signaling pathways associated with virus entry into trophoblasts is important for modeling pathogen transmission and infection of the fetus. Furthermore, microbial infections during pregnancy are associated with a variety of serious pathologies. Specific targeting of molecular mediators of CVB entry in the placenta is critical for reducing the incidence of diseases during pregnancy. Further study into virus-host interactions at the maternal-fetal interface is necessary to effectively prevent and reduce the incidences of prenatal infections, birth defects, and fetal death.