Brownian adhesive dynamics (BRAD) for simulating the receptor-mediated binding of viruses

Current viral docking models have relied upon the assumption that bond formation and breakage are independent of viral and docking surface geometry, as well as the forces exerted on the bonds. This assumption, known as the equivalent site hypothesis (ESH), is examined in detail using a newly developed simulation technique-Brownian adhesive dynamics (BRAD). The simulation couples the thermal motion of viral particles with adhesive dynamics models to characterize the effect of bonding on viral motion. We use the binding of HIV-like particles to CD4 expressing cells as a model system to illustrate the utility of BRAD. Comparison of the transition rates between bound states predicted by ESH and the rates resulting from BRAD simulations show dramatic differences; at values of the equilibrium crosslinking constant, K(x)R(T), where ESH suggests all virus adhesion proteins will be bound (K(x)R(T) = 10(6)), BRAD predicts not all virus adhesion proteins will be bound. At values of the equilibrium crosslinking constant used in typical ESH calculations of virus docking (K(x)R(T) = 1) we find BRAD simulations predict no binding. The mean bond density from BRAD models is often much lower than that predicted by ESH for equivalent parameter values. BRAD suggests that the viruses are much less well bound than ESH predicts. The differences suggest that binding models for viruses need to be reexamined closely. BRAD is a simulation technique that will be useful for quantifying the receptor-mediated binding of a wide variety of viruses to cells.     

The bulky and the sweet: How neutralizing antibodies and glycan receptors compete for virus binding

Numerous viruses rely on glycan receptor binding as the initial step in host cell infection. Engagement of specific glycan receptors such as sialylated carbohydrates, glycosaminoglycans, or histo‐blood group antigens can determine host range, tissue tropism, and pathogenicity. Glycan receptor‐binding sites are typically located in exposed regions on viral surfaces—sites that are also generally prone to binding of neutralizing antibodies that directly interfere with virus‐glycan receptor interactions. In this review, we examine the locations and architecture of the glycan‐ and antibody‐binding sites in four different viruses with stalk‐like attachment proteins (reovirus, influenza virus, norovirus, and coronavirus) and investigate the mechanisms by which antibodies block glycan recognition. Those viruses exemplify that direct molecular mimicking of glycan receptors by antibodies is rare and further demonstrate that antibodies often partly overlap or bind sufficiently close to the receptor‐binding region to hinder access to this site, achieving neutralization partially because of the epitope location and partly due to their sheer size.     

General analysis of receptor-mediated viral attachment to cell surfaces.

Viruses are multivalent particles that attach to cells through one or more bonds between viral attachment proteins (VAP) and specific cellular receptors. Three modes of virus binding are presented that can explain the diversity in binding data observed among viruses. They are based on multivalency of attachment and spatial versus receptor saturation effects which are easily distinguished based upon simple criteria. Mode 1 involves only monovalent virus/receptor binding. Modes 2 and 3 involve multivalent bonds between the virus and cell; however, in mode 3 space on the cell surface becomes saturated before receptors. A model is developed for viral attachment that accounts for nonspecific binding, receptor/virus interactions, and spatial saturation effects. The model can describe each mode in different limits and can be applied to virus binding data to extract key physical information such as receptor number and affinity. These values are used to postulate the type of VAP/receptor interaction involved and to predict binding at different parameter values. For the mode 2 binding of Adenovirus 2, the model predicts a receptor number of 4-15 x 10(3) on HeLa cells and an affinity of 2-6 x 10(7) M-1 which closely approximate experimental estimates. For the binding of three, broad-host-range, enveloped viruses, Semliki Forest virus, Vesicular Stomatitis virus, and the baculovirus, Autographa californica nuclear polyhedrosis virus, the model predicts receptor numbers of 10(5) or greater and affinities in the range of 10(4) to 10(5) M-1. These values are indicative of a VAP/oligosaccharide interaction which has been documented for a number of other viruses. Experimental evidence is presented that is the first to demonstrate that baculovirus binding is mediated by a cell surface receptor.     

Only two residues are responsible for the dramatic difference in receptor binding between swine and new pandemic H1 hemagglutinin

In view of its critical role in influenza A virus (IAV) tropism and pathogenesis, we evaluated the receptor binding properties of HA proteins of the closely related swine and new pandemic human IAVs. We generated recombinant soluble trimeric H1 ectodomains of several IAVs and analyzed their sialic acid binding properties using fetuin-binding and glycan array analysis. The results show that closely related swine and new pandemic H1 proteins differ dramatically in their ability to bind these receptors. Although new pandemic H1 protein exhibited hardly any binding, swine H1 bound efficiently to a number of α2-6-linked sialyl glycans. The responsible amino acids were identified by analyzing chimeric H1 proteins and by performing systematic site-directed mutagenesis of swine and new pandemic human H1 proteins. The difference was found to map to residues at positions 200 and 227. Although substitution of either residue significantly affected the binding phenotype, substitution of both was found to act synergistically and reverse the phenotype almost completely. Modeling of the T200A and E227A substitutions into the crystal structure of the new pandemic human H1 protein revealed the loss of potential hydrogen bond formation with Gln(191), which is part of the 190-loop of the receptor binding site, and with the penultimate galactose, respectively. Thus, a residue not belonging to the receptor binding site may affect the interaction of HA with its receptor. Interestingly, whereas alanine at position 200 is found in most new pandemic human viruses, the residue at position 227 in these viruses is invariably a glutamic acid.     

Triggering of the newcastle disease virus fusion protein by a chimeric attachment protein that binds to Nipah virus receptors

The fusion (F) proteins of Newcastle disease virus (NDV) and Nipah virus (NiV) are both triggered by binding to receptors, mediated in both viruses by a second protein, the attachment protein. However, the hemagglutinin-neuraminidase (HN) attachment protein of NDV recognizes sialic acid receptors, whereas the NiV G attachment protein recognizes ephrinB2/B3 as receptors. Chimeric proteins composed of domains from the two attachment proteins have been evaluated for fusion-promoting activity with each F protein. Chimeras having NiV G-derived globular domains and NDV HN-derived stalks, transmembranes, and cytoplasmic tails are efficiently expressed, bind ephrinB2, and trigger NDV F to promote fusion in Vero cells. Thus, the NDV F protein can be triggered by binding to the NiV receptor, indicating that an aspect of the triggering cascade induced by the binding of HN to sialic acid is conserved in the binding of NiV G to ephrinB2. However, the fusion cascade for triggering NiV F by the G protein and that of triggering NDV F by the chimeras can be distinguished by differential exposure of a receptor-induced conformational epitope. The enhanced exposure of this epitope marks the triggering of NiV F by NiV G but not the triggering of NDV F by the chimeras. Thus, the triggering cascade for NiV G-F fusion may be more complex than that of NDV HN and F. This is consistent with the finding that reciprocal chimeras having NDV HN-derived heads and NiV G-derived stalks, transmembranes, and tails do not trigger either F protein for fusion, despite efficient cell surface expression and receptor binding.     

Binding of transmissible gastroenteritis coronavirus to cell surface sialoglycoproteins

The surface glycoprotein S of transmissible gastroenteritis virus (TGEV) has two binding activities. (i) Binding to porcine aminopeptidase N (pAPN) is essential for the initiation of infection. (ii) Binding to sialic acid residues on glycoproteins is dispensable for the infection of cultured cells but is required for enteropathogenicity. By comparing parental TGEV with mutant viruses deficient in the sialic acid binding activity, we determined the contributions of both binding activities to the attachment of TGEV to cultured cells. In the presence of a functional sialic acid binding activity, the amount of virus bound to two different porcine cell lines was increased sixfold compared to the binding of the mutant viruses. The attachment of parental virus was reduced to levels observed with the mutants when sialic acid containing inhibitors was present or when the cells were pretreated with neuraminidase. In virus overlay binding assays with immobilized cell surface proteins, the mutant virus only recognized pAPN. In addition, the parental virus bound to a high-molecular-mass sialoglycoprotein. The recognition of pAPN was sensitive to reducing conditions and was not dependent on sialic acid residues. On the other hand, binding to the sialic acid residues of the high-molecular-mass glycoprotein was observed regardless of whether the cellular proteins had been separated under reducing or nonreducing conditions. We propose that binding to a surface sialoglycoprotein is required for TGEV as a primary attachment site to initiate infection of intestinal cells. This concept is discussed in the context of other viruses that use two different receptors to infect cells.     

Purified feline and canine transferrin receptors reveal complex interactions with the capsids of canine and feline parvoviruses that correspond to their host ranges

The cell infection processes and host ranges of canine parvovirus (CPV) and feline panleukopenia virus (FPV) are controlled by their capsid interactions with the transferrin receptors (TfR) on their host cells. Here, we expressed the ectodomains of wild-type and mutant TfR and tested those for binding to purified viral capsids and showed that different naturally variant strains of the viruses were associated with variant interactions with the receptors which likely reflect the optimization of the viral infection processes in the different hosts. While all viruses bound the feline TfR, reflecting their tissue culture host ranges, a naturally variant mutant of CPV (represented by the CPV type-2b strain) that became the dominant virus worldwide in 1979 showed significantly lower levels of binding to the feline TfR. The canine TfR ectodomain did not bind to a detectable level in the in vitro assays, but this appears to reflect the naturally low affinity of that interaction, as only low levels of binding were seen when the receptor was expressed on mammalian cells; however, that was sufficient to allow endocytosis and infection. The apical domain of the canine TfR controls the specific interaction with CPV capsids, as a canine TfR mutant altering a glycosylation site in that domain bound FPV, CPV-2, and CPV-2b capsids efficiently. Enzymatic removal of the N-linked glycans did not allow FPV binding to the canine TfR, suggesting that the protein sequence difference is itself important. The purified feline TfR inhibited FPV and CPV-2 binding and infection of feline cells but not CPV-2b, indicating that the receptor binding may be able to prevent the attachment to the same receptor on cells.     

Antibodies to the vitronectin receptor (integrin alpha V beta 3) inhibit binding and infection of foot-and-mouth disease virus to cultured cells.

The amino acid sequence Arg-Gly-Asp (RGD) is highly conserved on the VP1 proteins of different serotypes and subtypes of foot-and-mouth disease virus (FMDV) and is essential for cell attachment. This sequence is also found in certain extracellular matrix proteins that bind to a family of cell surface receptors called integrins. Within the Picornaviridae family, enterovirus coxsackievirus A9 also has an RGD motif on its VP1 capsid protein and has recently been shown to utilize the vitronectin receptor integrin alpha V beta 3 as a receptor on monkey kidney cells. Competition binding experiments between type A12 FMDV and coxsackievirus A9 using BHK-21 and LLC-MK2 cells revealed shared receptor specificity between these two viruses. Polyclonal anti-serum to the vitronectin receptor and a monoclonal antibody to the alpha V subunit inhibited both FMDV binding and plaque formation, while a monoclonal antibody to the beta 3 subunit inhibited virus binding. In contrast, antibodies to the fibronectin receptor (alpha 5 beta 1) or to the integrin (alpha V beta 5) had no effect on either binding or plaque formation. These data demonstrate that the alpha V beta 3 vitronectin receptor can function as a receptor for FMDV.     

Characterization of saturable binding sites for rabies virus.

A specific, saturable receptor for rabies virus was analyzed on cultured cells of neural or non-neural origin. Viral attachment kinetics were enhanced by DEAE-dextran, an effect which in turn enhanced the apparent infectivity of the virus inoculum. Under optimized conditions, the attachment of metabolically labeled ERA strain rabies virus obeyed the laws of mass action, whereby the amount of virus bound to cells varied proportionally with the concentration of cells or virus. Attachment was sensitive to changes of temperature and pH, did not require divalent cations such as Mg2+ or Ca2+, and occurred despite prior treatment of cells with proteolytic or sialic acid-specific enzymes. Saturation of the cell surface with rabies virus could be accomplished with 3 X 10(3) to 15 X 10(3) attached virions per cell. Competition for the rabies receptor occurred with rabies nonpathogenic variant virus, RV194 -2, and vesicular stomatitis virus. Reovirus type 3, another neurotropic virus, failed to inhibit rabies virus binding, and West Nile virus only slightly inhibited rabies virus binding, suggesting independent cellular receptors were recognized by these viruses. Isolated rabies virus glycoprotein failed to compete in an equivalent manner. However, solubilization of BHK-21 cells with octylglucoside yielded a chloroform-methanol-soluble extract which blocked rabies virus attachment. The binding inhibition activity of this extract was resistant to proteases but could be destroyed by phospholipases and neuraminidase, suggesting a phospholipid or glycolipid component at the receptor site. These data provide evidence for a rhabdovirus-common mechanism for cellular attachment to cells in culture.     

Characterization of a common high-affinity receptor for reovirus serotypes 1 and 3 on endothelial cells.

During viremia, viruses may be cleared from the bloodstream and taken up by specific organs. The uptake of virus from the bloodstream is dependent on the association of viral particles with endothelial cells that line the luminal surfaces of large and small blood vessels. To understand the nature of this interaction, we have studied the binding of reovirus serotypes 1 and 3 to these cells in vitro. Both serotypes of reovirus productively infected endothelial cells. By using [35S]methionine-biolabeled reovirus as a tracer ligand, we found that both viruses rapidly bind to endothelial cells and that equilibrium is reached after 4 h. The binding of the radiolabeled viruses was saturable and mediated by a homogeneous population of cellular receptors with very high affinity (Kd = 0.5 nM) for the virus ligands. Both serotypes bind to the same receptor, since the attachment of each radiolabeled serotype is inhibited by both the homologous and heterologous unlabeled virus. Exposure of labeled virus to monoclonal antibodies directed against the viral hemagglutinin (sigma 1 protein) inhibited binding, demonstrating that the attachment of reovirus to endothelial cells is mediated by the hemagglutinin for both serotypes. By using a novel ligand-blotting assay, the binding of both viruses to a 54,000-dalton protein could be demonstrated. The binding of each radiolabeled serotype to this protein was inhibited by the homologous and heterologous unlabeled serotype. By using cell fractionation after homogenization, we demonstrated that this 54-kilodalton protein is a membrane protein, in agreement with its proposed role as a cell surface receptor for reovirus serotypes 1 and 3.     

Identification and chemical synthesis of a host cell receptor binding site on hepatitis B virus

Hepatitis B virus (HBV) has not yet been propagated in vitro, and knowledge concerning its reaction with receptors on target cells remained scant. We have located within the HBV envelope proteins a sequence mediating the attachment of HBV to human hepatoma HepG2 cells. A synthetic peptide analog (PLGFFPDHQLDPAFGANSNNPDWDFNP) is recognized by both cell receptors and anti-HBV antibodies and elicits antibodies reacting with native HBV. The synthetic peptide is a promising immunogen expected to elicit protective antibodies based on the concept of the attachment blockade pathway of virus neutralization. The approach described here, based on anti-peptide antisera and the binding of peptide analogs to cell receptors is generally applicable for the delineation of cell receptor binding sites on viruses with known gene sequences.     

Attachment of VLDL receptors to an icosahedral virus along the 5-fold symmetry axis: multiple binding modes evidenced by fluorescence correlation spectroscopy

Human rhinoviruses (HRVs) are composed of 60 identical subunits, each comprising one copy of the viral capsid proteins VP1, 2, 3, and 4. Consequently, 60 symmetry-related epitopes are available for binding of antibodies or receptors. The minor receptor group of HRVs uses members of the low-density lipoprotein receptor family for cell entry. The ligand binding domains of these receptors are composed of various numbers of ligand binding repeats, and several of these modules within a single molecule are believed to attach simultaneously to the star-shaped dome at the 5-fold symmetry axis of the virus. Using fluorescence correlation spectroscopy (FCS), we have now determined the equilibrium binding constants and the mode of attachment of recombinant concatemers of ligand binding module 3 of the human very-low-density lipoprotein receptor to HRV2. We demonstrate that the avidity of the interaction drastically increases with the number of concatenated modules. For the trimer, the binding isotherm was biphasic, indicating that attachment of two and of three modules within the same molecule was resolved. The receptor consisting of seven repeats was found to bind most strongly, but a complete binding isotherm could not be established due to cross-linking of virions. The values of the dissociation constants were about 1 order of magnitude higher than those previously determined by using surface plasmon resonance techniques reflecting the different presentation of the binding partners. As compared to the concatemers, the natural receptors are composed of similar but not identical repeats; thus, cooperativity and different specificity of the ligand-binding modules allow for recognition of many ligands and viral serotypes. Due to the low concentrations and amounts of sample required, FCS is ideally suited for the determination of receptor binding parameters of viruses difficult to produce in high quantities and/or concentrations.     

Identification of Echovirus 1 and coxsackievirus A9 receptor molecules via a novel flow cytometric quantification method

BACKGROUND: Virus-receptor binding is an essential step in every virus infectious process. Many viruses employ more than one receptor molecule or even receptor complexes for attachment. In this study, we investigate the binding of Echovirus 1 (Echo1) and Coxsackievirus A9 (CAV-9) on cell surface molecules. CAV-9 has been reported to utilize integrin alpha v beta(3) in binding to cells, whereas Echo1 has been known to utilize integrin alpha 2 beta(1). METHODS AND RESULTS We directly test whether the presence of these molecules alone was sufficient for virus binding. We devised a novel flow cytometric binding assay that enables us to quantify virus particles bound on host cells and to further determine the extent to which viruses utilize specific receptors. CONCLUSIONS: By quantifying virus particles and possible receptor molecules, we found that Echo1 utilizes mainly integrin alpha 2 beta(1). CAV-9 utilizes integrin alpha v beta(3) to a much lesser extent (40%), indicating that CAV-9 also utilizes other receptor(s).     

Single-virus assay reveals membrane determinants and mechanistic features of Sendai virus binding

Sendai virus (SeV, formally murine respirovirus) is a membrane-enveloped, negative-sense RNA virus in the Paramyxoviridae family and is closely related to human parainfluenza viruses. SeV has long been utilized as a model paramyxovirus and has recently gained attention as a viral vector candidate for both laboratory and clinical applications. To infect host cells, SeV must first bind to sialic acid glycolipid or glycoprotein receptors on the host cell surface via its hemagglutinin-neuraminidase (HN) protein. Receptor binding induces a conformational change in HN, which allosterically triggers the viral fusion (F) protein to catalyze membrane fusion. While it is known that SeV binds to α2,3-linked sialic acid receptors, and there has been some study into the chemical requirements of those receptors, key mechanistic features of SeV binding remain unknown, in part because traditional approaches often convolve binding and fusion. Here, we develop and employ a fluorescence microscopy-based assay to observe SeV binding to supported lipid bilayers (SLBs) at the single-particle level, which easily disentangles binding from fusion. Using this assay, we investigate mechanistic questions of SeV binding. We identify chemical structural features of ganglioside receptors that influence viral binding and demonstrate that binding is cooperative with respect to receptor density. We measure the characteristic decay time of unbinding and provide evidence supporting a “rolling” mechanism of viral mobility following receptor binding. We also study the dependence of binding on target cholesterol concentration. Interestingly, we find that although SeV binding shows striking parallels in cooperative binding with a prior report of Influenza A virus, it does not demonstrate a similar sensitivity to cholesterol concentration and receptor nanocluster formation.     

Evidence that binding of cucumber necrosis virus to vector zoospores involves recognition of oligosaccharides

Despite the importance of vectors in natural dissemination of plant viruses, relatively little is known about the molecular features of viruses and vectors that permit their interaction in nature. Cucumber necrosis virus (CNV) is a small spherical virus whose transmission in nature is facilitated by zoospores of the fungus Olpidium bornovanus. Previous studies have shown that specific regions of the CNV capsid are involved in transmission and that transmission defects in several CNV transmission mutants are due to inefficient attachment of virions to the zoospore surface. In this study, we have undertaken to determine if zoospores contain specific receptors for CNV. We show that in vitro binding of CNV to zoospores is saturable and that vector zoospores bind CNV more efficiently than nonvector zoospores. Further studies show that treatment of zoospores with periodate and trypsin reduces CNV binding, suggesting the involvement of glycoproteins in zoospore attachment. In virus overlay assays, CNV binds to several proteins, whereas CNV transmission mutants either fail to bind or bind at significantly reduced levels. The possible involvement of specific sugars in attachment was investigated by incubating CNV with zoospores in the presence of various sugars. Two mannose derivatives (methyl alpha-D-mannopyranoside and D-mannosamine), as well as three mannose-containing oligosaccharides (mannotriose, alpha3,alpha6-mannopentaose, and yeast mannan) and L-(-)-fucose, all inhibited CNV binding at relatively low concentrations. Taken together, our studies suggest that binding of CNV to zoospores is mediated by specific mannose and/or fucose-containing oligosaccharides. This is the first time sugars have been implicated in transmission of a plant virus.     

Identification of an extracellular domain within the human PiT2 receptor that is required for amphotropic murine leukemia virus binding

Human PiT2 (PiT2) is a multiple-membrane-spanning protein that functions as a type III sodium phosphate cotransporter and as the receptor for amphotropic murine leukemia virus (A-MuLV). Human PiT1 (PiT1), another type III sodium phosphate cotransporter, is a highly related protein that functions as a receptor for gibbon ape leukemia virus but not for A-MuLV. The ability of PiT1 and PiT2 to function as discrete viral receptors with unique properties presumably is reflected in critical residue differences between these two proteins. Early efforts to map the region(s) within PiT2 that is important for virus binding and/or entry relied on infection results obtained with PiT1-PiT2 chimeric cDNAs expressed in Chinese hamster ovary (CHOK1) cells. These attempts to localize the PiT2 virus-binding site were hampered because they were based on infectivity, not binding, assays, and therefore, receptors that bound but failed to facilitate virus entry could not be distinguished from receptors that did not bind virus. Using a more accurate topological model for PiT2 as well as an A-MuLV receptor-binding assay, we have identified extracellular domain one (ECD1) of the human PiT2 receptor as being important for A-MuLV binding and infection.     

Residues in the apical domain of the feline and canine transferrin receptors control host-specific binding and cell infection of canine and feline parvoviruses

Canine parvovirus (CPV) and feline panleukopenia virus (FPV) capsids bind to the transferrin receptors (TfRs) of their hosts and use these receptors to infect cells. The binding is partially host specific, as FPV binds only to the feline TfR, while CPV binds to both the canine and feline TfRs. The host-specific binding is controlled by a combination of residues within a raised region of the capsid. To define the TfR structures that interact with the virus, we altered the apical domain of the feline or canine TfR or prepared chimeras of these receptors and tested the altered receptors for binding to FPV or CPV capsids. Most changes in the apical domain of the feline TfR did not affect binding, but replacing Leu221 with Ser or Asp prevented receptor binding to either FPV or CPV capsids, while replacing Leu221 with Lys resulted in a receptor that bound only to CPV but not to FPV. Analysis of recombinants of the feline and canine TfRs showed that sequences controlling CPV-specific binding were within the apical domain and that more than one difference between these receptors determined the CPV-specific binding of the canine TfR. Single changes within the canine TfR which removed a single amino acid insertion or which eliminated a glycosylation site gave that receptor the expanded ability to bind to FPV and CPV. In some cases, binding of capsids to mutant receptors did not result in infection, suggesting a structural role for the receptor in cell infection by the viruses.     

Identification of two surface proteins from C6/36 cells that bind dengue type 4 virus.

Dengue viruses infect cells by attaching to a surface receptor, probably through the envelope (E) glycoprotein, located on the surface of the viral membrane. However, the identity of the dengue virus receptor in the mosquito and in mammalian host cells remains unknown. To identify and characterize the molecules responsible for binding dengue virus, overlay protein blot and binding assays were performed with labeled virus. Two glycoproteins of 40 and 45 kDa located on the surface of C6/36 cells bound dengue type 4 virus. Virus binding by total and membrane proteins obtained from trypsin-treated cells was inhibited, while neuraminidase treatment did not inhibit binding. Periodate treatment of cell proteins did not reduce virus binding, but it modified the molecular weight of the polypeptide detected by overlay assays. Preincubation of C6/36 cells with electroeluted 40- and 45-kDa proteins or with specific antibodies raised against these proteins inhibited virus binding. These results strongly suggest that the 40- and 45-kDa surface proteins are putative receptors or part of a receptor complex for dengue virus.     

Identification of a putative coreceptor on Vero cells that participates in dengue 4 virus infection

Dengue virus infects target cells by attaching to a cell surface receptor through the envelope (E) glycoprotein, located on the surface of the viral membrane. On Vero and BHK cells, heparan sulfate (HS) moieties of proteoglycans are the receptors for dengue virus; however, additional proteins have also been described as putative dengue virus receptors on C6/36, HL60, and BM cells. HS can also act as a receptor for other types of viruses or as an attachment molecule for viruses that require additional host cell molecules to allow viral penetration. In this study we searched for molecules other than HS that could participate in dengue virus infection of Vero cells. Labeled dengue 4 virus bound with high affinity to two molecules of 74 and 44 kDa. Binding of dengue virus to the 74-kDa molecule was susceptible to protease and sodium periodate treatment and resistant to heparinase treatments. Lectins such as concanavalin A and wheat germ agglutinin prevented dengue virus binding to both the 74- and the 44-kDa protein in overlay assays, while phytohemagglutinin P did not affect binding, suggesting that carbohydrate residues (alpha-mannose or N-acetylglucosamine) are important in virus binding to host cells. Protease susceptibility, biotin labeling, and immunofluorescence with a polyclonal antibody raised against the 74-kDa protein consistently identified the protein on the surfaces of Vero cells. Moreover, the antibody against the 74-kDa protein was able to inhibit dengue virus infection. These data suggest that HS might serve as a primary receptor, probably concentrating virus particles on the surfaces of Vero cells, and then other molecules, such as the 74-kDa protein, might participate as coreceptors in viral penetration. The 74-kDa protein possibly constitutes part of a putative receptor complex for dengue virus infection of Vero cells.