Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Mar 17.
Published in final edited form as: Cell Host Microbe. 2007 Apr 19;1(2):135–145. doi: 10.1016/j.chom.2007.03.002

The stoichiometry of antibody-mediated neutralization and enhancement of West Nile virus infection

Theodore C Pierson 1,*, Qing Xu 1, Steevenson Nelson 1, Theodore Oliphant 3, Grant E Nybakken 4, Daved H Fremont 4, Michael S Diamond 2,3,4
PMCID: PMC2656919  NIHMSID: NIHMS28912  PMID: 18005691

Abstract

Viruses with icosahedral symmetry display a large number of repetitive antigens on their surface that can be recognized by antibodies. Antibody binding to flaviviruses, such as West Nile virus (WNV), can lead to neutralization or enhancement of infection on cells expressing Fc-γ-receptors. In this study, we evaluate directly how many antibodies for a given epitope must be bound to a virion to achieve neutralization or enhancement of infection. Monoclonal antibodies (mAbs) that recognize a conserved and accessible determinant in domain III of WNV envelope protein block infection at concentrations that result in a low occupancy of the available sites on the virion. In contrast, weakly neutralizing mAbs recognize fewer sites on the virion, and require almost complete occupancy to inhibit WNV infection. Our studies suggest that neutralization by the most potent mAbs occurs when as few as 30 of 180 sites are occupied. When fewer antibodies are bound, enhancement of infection is possible in cells bearing activating Fc-γ receptors. For icosahedral flaviviruses, neutralization is best described by a model requiring ‘multiple-hits’ with the cumulative functional outcome determined by interplay between antibody affinity and epitope accessibility.

Introduction

WNV is a single-stranded positive-sense RNA virus of the Flavivirus genus. The ~11 kilobase WNV genomic RNA is translated in the cytoplasm as a polyprotein and then cleaved into three structural (capsid (C), pre-membrane/membrane (prM/M) and envelope (E)) and seven non-structural proteins by virus and host-encoded proteases (Brinton, 2002; Lindenbach and Rice, 2001). In nature, WNV is maintained in an enzootic cycle between mosquitoes and birds, but can also infect and cause disease in horses and other vertebrate animals (reviewed in reference (Hayes et al., 2005)). WNV infection of humans is associated with a febrile illness that can progress to lethal encephalitis, particularly in the elderly and immunocompromised (Ceausu et al., 1997; Petersen et al., 2002; Sejvar et al., 2003). Since the mid-1990’s, outbreaks of WNV fever and encephalitis have occurred annually throughout the world (Dauphin et al., 2004). Following its introduction into the United States in 1999, WNV rapidly disseminated across North America and has now been reported in Mexico, South America, and the Caribbean (Deardorff et al., 2006; Komar and Clark, 2006; Lanciotti et al., 1999). At present, treatment is supportive and no vaccine or therapy exists for human use.

Humoral immunity is an essential aspect of host protection against WNV (Ben-Nathan et al., 2003; Camenga et al., 1974; Diamond et al., 2003a; Diamond et al., 2003b; Oliphant et al., 2005; Tesh et al., 2002; Wang et al., 2001) and other flaviviruses (Roehrig et al., 2001). B cell deficient mice die after WNV infection, but are protected by passive transfer of immune sera (Diamond et al., 2003a; Diamond et al., 2003b). Antibody-mediated control of flavivirus infection in vivo (Ben-Nathan et al., 2003; Diamond et al., 2003a; Diamond et al., 2003b; Engle and Diamond, 2003; Gould et al., 2005; Oliphant et al., 2005) has been correlated with in vitro neutralizing activity (Kaufman et al., 1987; Phillpotts et al., 1987; Roehrig et al., 2001). The majority of neutralizing antibodies against flaviviruses are directed against the E protein, although some likely recognize the prM/M protein (Colombage et al., 1998; Falconar, 1999; Pincus et al., 1992; Vazquez et al., 2002).

The crystal structure of the E protein ectodomain has been solved for several flaviviruses including dengue virus (DENV), tick-borne encephalitis virus (TBEV), and WNV (Kanai et al., 2006; Modis et al., 2004; Modis et al., 2005; Nybakken et al., 2006; Rey et al., 1995; Zhang et al., 2004). E is composed of three domains that mediate viral attachment, entry, and assembly. Domain III (DIII) contains the putative receptor binding domain (Bhardwaj et al., 2001; Chu et al., 2005), domain II (DII) encodes the putative fusion loop involved in pH-dependent fusion of virus and host cell membranes (Allison et al., 2001), and domain I (DI) participates in E protein structural rearrangements required for fusion (reviewed in (Mukhopadhyay et al., 2005)). Crystallography, NMR, and epitope mapping studies have established that E protein-specific neutralizing antibodies map to all three domains of the WNV E protein (Beasley and Barrett, 2002; Nybakken et al., 2005; Oliphant et al., 2005; Oliphant et al., 2006; Sanchez et al., 2005; Volk et al., 2004). The most potent inhibitory antibodies recognize a single neutralizing epitope on the lateral face of DIII that constitute the amino-terminal region and three loops of the immunoglobulin-like fold (Beasley and Barrett, 2002; Oliphant et al., 2005; Sanchez et al., 2005; Volk et al., 2004). Recent studies in rodent models of WNV infection demonstrate that antibodies that bind this DIII epitope are protective or therapeutic when passively administered (Morrey et al., 2006; Oliphant et al., 2005).

The structure of the WNV virion has been determined by cryoelectron microscopy (Kuhn et al., 2002; Mukhopadhyay et al., 2003). The mature WNV is an icosahedral particle that lacks conventional T = 3 symmetry (reviewed in (Mukhopadhyay et al., 2005)). As a result, the E protein exists in three chemically distinct environments (shown in Fig. 1c). Crystallographic modeling and cryo-electron microscopy studies have shown that only 120 of the available 180 E protein epitopes can be occupied by the DIII-specific neutralizing antibody E16 due to steric hindrance of binding of the 60 DIII epitopes that lie closest to the five-fold symmetry axs (shown in yellow in Fig. 1c)(Kaufmann et al., 2006; Nybakken et al., 2005). These studies provide a structural demonstration that the accessibility of an epitope is determined in part by its location on the surface of the virion, and suggest that engagement of all the E proteins on the virion is not a requirement for neutralization. In this study, we evaluate how the number of antibodies bound to DIII-specific epitopes on flaviviruses impacts the infectivity of the virus particle.

Figure 1. mAb epitope location on WNV DIII.

Figure 1

Open in a new tab

(a) Ribbon diagram of the DIII portion of the WNV E protein as viewed down the long axis of the E protein monomer. Residues that comprise antibody-binding determinants are indicated. (b) Surface representation diagrams of DIII indicating the binding determinants for the six mAbs studied. Red color indicates residues that when mutated, abrogate 100% of mAb binding as measured by a quantitative yeast mapping method. Yellow color identifies residues that result in greater than 50% reduction in mAb binding. (c) Molecular modeling of epitope accessibility on the WNV virion. Residues that form the epitope for each antibody studied were previously identified using yeast-display technology. Residues involved in binding for mAb E9, E16, E22, and E49 (4.0 Å radius magenta atoms) were mapped onto the crystal structure of the E protein and then docked onto the pseudoatomic model of the mature WNV virion. mAbs E24 and E47 bind an epitope that overlaps the E16 epitope. Virions are depicted as 2.0 Å radius Ca atoms colored according to the symmetry axis to which they are closest, two-fold (cyan), three-fold (green) and five-fold (yellow).

Results

The stoichiometric requirements of antibody-mediated neutralization of WNV infection were investigated using monoclonal antibodies (mAbs) that bind epitopes on DIII of the E protein. A group of six DIII-reactive murine mAbs were selected from a recently described panel produced by immunization with purified recombinant E protein or infectious WNV (Oliphant et al., 2005). Epitope mapping data for each antibody have been reported, and were used as part of our selection criteria to include mAbs with a variety of functional properties and binding specificities (Oliphant et al., 2005)(Fig. 1a and b and Table 1).

Table 1.

Neutralization potency and avidity of DIII-specific WNV mAbs

EC50 (nM) St. Dev. (nM) Slope St. Dev. N Avidity (nM) St. Dev. (nM) N
E9 39.5 25.5 −0.35 0.07 4 2.33 0.94 9
E16 0.043 0.016 −2.09 0.41 15 0.15 0.08 22
E22 1.39 0.9 −0.34 0.02 6 1.02 0.081 3
E24 0.018 0.011 −1.01 0.2 10 0.1 0.07 12
E47 0.031 0.015 −0.8 0.12 5 0.26 0.2 9
E49 0.98 0.85 −0.61 0.06 6 1.07 0.65 9
Open in a new tab

To investigate the relative accessibility of each epitope on the virion, molecular modeling studies were performed in which residues that comprise the epitope of each antibody were mapped onto the E protein crystal structure and modeled on the pseudo-atomic reconstruction of the mature WNV virion (Fig 1c). This approach predicts noticeable differences in the accessibility of several of the antibody epitopes. E16, E24, E47 and E49 bind determinants on the upper lateral surface of DIII that project away from the surface of the virion, whereas the residues bound by E22 and E9 are found on lateral surfaces partially obscured by neighboring E proteins. Moreover, while the residues constituting the E16 epitope are solvent exposed in all three different symmetry environments, steric constraints prevent simultaneous occupancy of more than 120 sites on the virion (Kaufmann et al., 2006; Nybakken et al., 2006). Thus, this modeling approach predicts the upper limit for occupancy at each epitope on the mature virion. A more quantitative assessment of the number of antibodies that bind the E9, E22, and E49 epitopes under saturating conditions awaits further structural studies.

Antibody-mediated neutralization

The neutralization potency of each DIII-specific antibody was characterized using a validated highly quantitative approach (Pierson et al., 2006). Reporter virus particles (RVP) are virus-like particles composed of the structural proteins of WNV and a sub-genomic replicon encoding a reporter gene. RVPs are capable of only a single round of infection and allow virus infection to be measured as a function of reporter gene activity. An important feature of this approach is that the functional outcome of antibody engagement is measured under conditions that satisfy the law of mass action as it applies to virus neutralization, increasing precision and accuracy (Andrewes and Elford, 1933; Klasse and Sattentau, 2002; Pierson et al., 2006). In this study, neutralization experiments were performed primarily using a B-lymphoblastoid cell line that stably expresses the attachment factor DC-SIGNR (CD209L) (Davis et al., 2006b).

Neutralization profiles for each antibody were generated using 19 two-fold dilutions of antibody performed in triplicate (Fig. 2a). The potency of each antibody was calculated using a curve fitting approach and expressed as the concentration that inhibits 50% of infection (EC50). Analysis of dose-response curves obtained from each antibody revealed a wide range in neutralization activity (Table 1). As predicted from preliminary studies performed using the traditional plaque reduction neutralization test (PRNT50) (Oliphant et al., 2005), antibodies that bind an overlapping cluster of residues on an upper lateral ridge of DIII centered around residues 307 and 332 were the most potent and could neutralize infection at concentrations as low as 20 pM (Fig. 2a and Table 1). In contrast, the two antibodies that bind lateral surfaces on DIII (E22 and E9) inhibit infection only at much higher concentrations (Fig. 2a and 2b and Table 1).

Figure 2. Neutralization of WNV by DIII-specific mAbs.

Figure 2

Open in a new tab

Nineteen dilutions of each mAb were incubated with WNV RVPs and used to infect Raji cells that express the c-type lectin DC-SIGNR. Infection was monitored as a function of GFP expression using flow cytometry. Error bars display the standard error of triplicate infections (present even when not visible due to low variation), while dotted lines illustrate the 95% confidence interval. Non-linear regression analysis used to calculate the EC50. (d) Identification of a neutralization-resistant fraction of WNV. Dose-response curves with mAb E22 do not extend to the y-axis, revealing a population of RVPs that are resistant to E22-mediated neutralization at high concentrations of antibody. Shown is the regression analysis describing the neutralization potency of E22 focusing only on the fraction of RVPs that can be neutralized by mAb. Error bars display the standard error of triplicate infections, while dashed lines illustrate the 95% confidence interval. Non-linear regression analysis was used to calculate the EC50. Considering the entire RVP population (including the 31% of virions in the resistant fraction) reveals a significantly less potent response (approximately 2 nM).

In addition to contributing to the precision of our measurements, the use of a large number of data points also provides insight into the nature of the interactions that result in neutralization (Fig. 2a). Differences in the shapes of the dose-response curves among the antibodies studied were observed. Dose-response profiles of the most potently neutralizing antibodies (E16, E24 and E47) were relatively steep when compared to the relatively flat shape of the neutralization profile of less potent mAbs such as E9 (Table 1). The steepness of these profiles indicates a greater capacity to neutralize virus at lower concentrations of antibody, and reflects differences in the fraction of epitopes on the virion that must be bound by antibody for neutralization to occur (discussed below).

We also noted that not all virions in the population were equally sensitive to neutralization by mAbs. Inspection of the dose response curves performed with the mAb E22 indicates that approximately 31% (± 8%, n=6) of the virions were insensitive to neutralization at even the highest concentrations of antibody tested (Fig 2b). As a result, calculating the concentration of this antibody required to completely neutralize infection with confidence may not be possible (greater than 2 nM). In contrast, if only those particles sensitive to E22-mediated neutralization are considered, the EC50 of this antibody is much lower (0.29 nM ± 0.07, n=6). As RVPs are produced by transfection of DNA plasmids and are thus identical with respect to E protein sequence, this finding suggests that heterogeneity exists with respect to the accessibility of the E22 epitope in the RVP population.

Relationships between antibody-dependent neutralization and enhancement of infection

Under some circumstances, the engagement of Flavivirus virions by antibody facilitates the infection of Fc-γ or complement receptor-expressing cell lines in vitro. Both processes are thought to reflect an opsonic phenomenon by which the efficiency of virus attachment is augmented in the presence of antibody and/or complement, respectively. Antibody-dependent enhancement of infection (ADE) has been suggested to contribute to the pathogenesis of secondary dengue virus (DENV) infections (reviewed in (Halstead, 2003)), and is implicated in the adverse effects following challenge of individuals vaccinated with some formalin-inactivated vaccines (Barrett and Gould, 1986; Gould and Buckley, 1989), including measles and respiratory synycytial viruses (Polack et al., 2002). To explore the quantitative relationships between neutralization and ADE, the functional properties of antibodies were investigated using K562 cells: a human erythroleukemia cell line that expresses high levels of the activating Fc-γ receptor FcγRIIa (CD32a) and has been used extensively to study ADE of flaviviruses in vitro (Klein et al., 1976; Littaua et al., 1990; Mady et al., 1991). The utility of this cell line in measuring ADE is due in part to an inability to efficiently bind WNV in the absence of antibody (Davis et al., 2006b). To compare neutralization and ADE on cells capable of efficiently binding virus, studies were performed in parallel on a K562 cell line that stably expresses the attachment factor DC-SIGNR (K562-R).

RVP-antibody complexes were created by pre-incubation of 19 two-fold dilutions of mAb E47 and used to infect Raji-R, K562, and K562-R cells (Fig 3a). Analysis of infections with Raji-R cells revealed the expected sigmoid shaped dose response curve (red circles) as described above. In contrast, the pattern of infection of K562 cells in the presence of antibody approximated a biphasic distribution (green circles) characteristic of ADE (Morens et al., 1987): the highest level of infection was observed at an intermediate concentration of antibody (approximately 100 pM). Infection of K562-R cells (blue circles) with antibody revealed a dose-response curve that largely overlapped with data obtained with the parental K562 cell line. The only significant difference between these two curves was the basal level of infection of the target cell types by WNV (0.08% vs. 7.5% for K562 and K562-R cells respectively). It is important to note that the peak enhancement titer for E47 on K562 or K562-R cells corresponds to the concentration of antibody close to the EC50 on Raji-R cells, reflecting the enhanced infection of the 50% of the virions bound by antibody at a stoichiometry that is not sufficient to neutralize infection. Similar relationships were observed when neutralization and ADE profiles of each antibody in our panel were compared (Fig. 3b–e, see also Fig. 6a–c for E16), as well as in a more extensive analysis of a larger panel of antibodies recognizing determinants throughout the E protein (Oliphant et al., 2006) (Nelson et. al, manuscript in preparation).

Figure 3. Analysis of the relationships between neutralization and ADE of WNV infection.

Figure 3

Open in a new tab

(a.) Complexes of mAb E47 and WNV RVPs were formed by incubation at room temperature and divided into three aliquots for addition to Raji-R (red circles), K562 (green circles) or K562-R cells (blue circles). Error bars display the standard error of triplicate infections, while dashed lines illustrate the 95% confidence interval. Nonlinear regression analysis was used to calculate the EC50, revealing a potency of 47.6 pM (95% CL: 40–56.4 pM) on Raji-R cells. For K562 and K562-R cells, the presence of ADE prevents analysis of antibody potency using curve-fitting approaches. Additional dose response curves of mAbs (b.) E9, (c.) E22, and (d.) E49 were also performed using Raji-R and K562 cells as described above. (e.) To determine the impact of decreasing the strength of the antibody-virion interaction on both neutralization and ADE, RVPs were produced that incorporate a T330I mutation and used in neutralization studies on Raji-R and K562 cells with mAb E24.

Figure 6. Estimating the stoichiometry of ADE of WNV infection.

Figure 6

Open in a new tab

Dose response curves for chE16 and chE16 N297Q were performed on (a.) Raji-R, (c.) K562, and (c.) K562-R cells. (d). Analysis of ADE using mixtures of mAbs ch E16 and chE16 N297Q. Antibodies were mixed at the indicated ratios (by mass) and incubated with WNV RVPs at various concentrations prior to addition to K562 cells. Infection was monitored as a function of GFP expression using flow cytometry. Error bars display the standard error of triplicate infections.

Together, these results suggest that neutralization and ADE are two phenomena related by the number of antibodies bound to the virion irrespective of the epitope recognized. To confirm this conclusion, we investigated the impact of reducing the number of antibodies bound to a single determinant by manipulating the strength of the antibody-virion interaction. Previous studies demonstrate that a T330I mutation in DIII dramatically reduces E24 binding (Oliphant et al., 2005). Dose response profiles obtained with RVPs that incorporate this single amino acid change reveal a shift in both the neutralization and ADE dose response profiles towards higher concentrations of antibody without altering the relationship between the two phenomena (Fig. 3f as compared to Fig. 3e).

Computational approach to determine occupancy requirements for antibody-mediated neutralization

As a first step towards determining the stoichiometry of antibody binding during virus neutralization, we measured the strength of the interaction between antibody and virus particles. This allows an estimate of the percent occupancy of accessible determinants on the virion for each concentration of antibody used in our functional studies. As multivalent interactions may occur during antibody engagement and neutralization, virus particles were captured in the solid phase for use as antigen and incubated with serial dilutions of anti-WNV antibodies under conditions of antibody-excess. Functional avidity of each antibody (AV50) was computed by curve fitting (Table 1).

Using the binding constants and concentrations of half-maximal neutralization we calculated the fraction of accessible epitopes on the virion that are bound by antibody when virus infection is neutralized (Klasse and Moore, 1996; Parren et al., 1998)(Fig. 4a). Engagement of all of the accessible epitopes on the virion was not required to achieve significant levels of neutralization. For example, at the EC50 of E16, on average less than 25% of the available epitopes are bound by antibody, and complete (99%) neutralization occurs at non-saturating concentrations (occupancy of 45% of the accessible epitopes). By comparison, engagement of a larger fraction of the potential E49 epitopes is required to achieve neutralization, with approximately half the available epitopes engaged at the EC50. Finally, neutralization by E9 is observed almost exclusively at concentrations at which the full occupancy of available epitopes is expected (Fig. 4b).

Figure 4. Computational approach to estimate the occupancy requirements for antibody-mediated neutralization of WNV.

Figure 4

Open in a new tab

(a.) Occupancy requirements for neutralization were estimated by plotting data from mAb dose response curves (y-axis) against the percentage of accessible epitopes bound by antibody at each concentration of antibody. Percent binding was computed using the avidity data obtained above by solving the equation: percent bound = [Ab]/([Ab] + KD). Error bars display the standard error of triplicate infections. (b.) Percent occupancy at the EC50 and EC99. Error bars reflect the variability arising when the calculations are performed using the calculated avidity of each antibody plus the standard deviation. This analysis was not performed for E22 as complete neutralization with this antibody is not achieved even at the highest concentrations tested.

Estimating the occupancy required for neutralization by altering the number of epitopes present on West Nile virions

The confidence of our computational approach for estimating the antibody occupancy requirements of virus neutralization is dependent in part on the accuracy of the avidity constant measurements. This may be limited by the impact of distortions in virion architecture that occur during virus capture and the potential for heterogeneity in the virus population with respect to particle geometry, maturation state, and epitope accessibility in each symmetry environment. To provide an independent estimate, we changed the number of available epitopes on the virion and determined their impact on neutralization by E24.

RVPs were produced that contain of a mixture of both wild-type E proteins and a variant incorporating a single point mutation (T332K) that abrogates the binding of E24 to WNV (Oliphant et al., 2005)(data not shown). The relative proportion of each form of the E protein on the virion was controlled by changing the ratio of two plasmids encoding wild type and T332K E proteins during transfection. Analogous approaches have been employed to investigate the stoichiometry of HIV neutralization (Schonning et al., 1999; Yang et al., 2005). RVPs produced using the indicated proportions of wild type and T332K E plasmids were pre-incubated with two-fold dilutions of E24 mAb and used to infect Raji-R and K562 cells (Fig. 5). As expected, E24 potently neutralized RVPs composed of wild type E protein (EC50 of 24 pM), while RVPs composed of E proteins with the T332K mutation were not neutralized on Raji-R cells (Fig. 5a). Reducing the proportion of wild type E protein in the virion significantly changed the dose-response profile. Populations of RVPs composed of both wild type and T332K E proteins contained a significant proportion of virions resistant to neutralization. The lower plateaus of the dose-response curves of “mixed” RVPS identified approximately 2%, 11%, 40%, and 77% of RVPs in the resistant fraction (for the 75%, 50%, 25%, and 10% wild type E protein RVP population, respectively). The appearance of a resistant fraction as a function of the reduction in the number of antibody binding determinants suggests virions in this sub-population no longer bind a sufficient number of antibodies for neutralization. In addition, regression analysis of the fraction of virus sensitive to neutralization revealed a small yet significant (P < 0.0001) increase in the EC50 associated with a reduction in the fraction of E24 binding sites (EC50 of 21, 32, 42 and 62 pM for the 75%, 50%, 25%, and 10% wild type RVPs respectively). This shift indicates a requirement for engagement of a larger fraction of the epitopes on the virion concomitant with the reduction in the total number of accessible sites. In agreement with our computational prediction (Fig. 4), these mixing studies indicate that a large percentage of WNV virions (90%) are neutralized when a relatively low percentage of the available E24 epitopes (50%) are occupied by antibody. Half of the virions are neutralized when roughly 25% of the accessible sites are bound by mAb. As E16 and E24 bind essentially the same epitope (Oliphant, 2005), on mature virions only 120 E proteins are accessible to E24, and therefore we estimate approximately 30 mAbs sufficient for neutralization.

Figure 5. Impact of manipulating the number of antibody-binding determinants on the neutralization potency of mAb E24.

Figure 5

Open in a new tab

The T332K mutation blocks antibody-recognition of a conserved neutralization determinant on the upper lateral surface of DIII (Oliphant et al., 2005). This change reduces E24 binding to WNV to levels below the limits of detection by ELISA. RVPs incorporating varying proportions of both wild type E protein and E protein encoding the T332K mutation were produced using a modified complementation approach and incubated with the indicated concentrations of E24. (a.) Dose-response curves of each RVP population using Raji-R cells. Error bars display the standard error of triplicate infections, while dotted lines illustrate the 95% confidence interval using non-linear regression analysis. The fraction of RVPs resistant to E24 infection was calculated as a function of the bottom of each dose response curve. (b.) ADE of “mixed” RVP populations was measured on K562 cells as described above.

Dose response profiles of the same RVP-E24 complexes on K562 cells also allowed us to determine the impact of changing the number of antibody binding sites on ADE. As expected, enhancement of wild type RVPs was observed on K562, whereas ADE with E24 was not observed with T332K RVPs at any concentration of antibody tested (Fig. 5b). The ADE profile for the “mixed” RVPs composed of both wild type and mutant proteins was shifted to the right, indicating enhancement at higher antibody concentrations than observed with wild type RVPs. In addition, the breadth of the biphasic profile was significantly increased. Together, these changes are a consequence of reducing the number of epitopes present on virions to levels that do not exceed a threshold required for neutralization, and thus permit enhancement even at complete occupancy. These experiments are an important control for the transfection approach used to generate the “mixed” particles. Had transfection of cells with mixtures of wild type and mutant plasmids produced RVPs composed entirely of wild type or T332K protein, rather than mixtures of each protein species within the same particles, a reduction of the amplitude of the ADE curve, rather than a shift to the right in the dose response curve (as seen in Fig. 6d described below) would have been expected.

Both the computational approach and experiments performed with mixed RVPs suggest that neutralization by the most potent mAbs occurs at relatively low occupancy of the number of accessible epitopes on the virion. The precision of this analysis, however is constrained by an inability to directly measure the distribution of wild type and mutant E proteins in the RVP populations studied. While the approximate ratio of the two species in the population was validated by Western blot (data not shown), interpretation is limited because RVPs are analyzed as a population rather than as individual virions. Higher resolution requires an understanding of how wild type and mutant proteins are distributed within the population of individual virions, and segregate within a single particle into the different symmetry environments of the virion.

Relationships between antibody-dependent neutralization and enhancement of infection

To estimate the stoichiometry requirements for ADE, studies were performed using two different forms of the mAb E16 that differ with respect to their ability to interact with Fc-γ receptors. A chimeric form of mAb E16 (chE16) was originally generated as an intermediate of a humanization process and contains the mouse VH and VL sequences and the human γ1 and κ constant region sequences (Oliphant et al., 2005). chE16 N297Q is a variant with a single amino acid substitution that abolishes an asparagine-linked glycosylation site essential for the interaction of this antibody with Fc-γ receptors (Tao and Morrison, 1989). Wild type and N297Q chE16 bind and neutralize WNV with identical efficiency when assayed on BHK-21 (Oliphant et al., 2005) or Raji-R cells (Fig. 6a). In contrast, these two antibodies behave differently when assayed on activating Fc-γRIIa+ K562 cells. Incubation of RVPs with chE16, but not chE16 N297Q, enhanced infection on K562 cells (Fig 6b), illustrating the importance of interactions between antibody and the Fc-γ receptor during ADE. When studied with K562-R cells (Fig. 6c), chE16 displayed a pattern similar to E47 (Fig. 3a) characterized by both neutralization and enhancement of infection. In contrast, chE16 N297Q neutralized WNV RVP infection on K562-R cells but did not support ADE.

Previous studies have observed that ADE occurs in the presence of “sub-neutralizing quantities” of antibody or sera (Morens et al., 1987). In agreement, our analyses of neutralization and ADE on a single cell type suggest that these two phenomena are related simply by differences in the number of antibodies bound to a single virion (Fig. 3). Enhancement of infection occurs once the virion is bound by antibody with a stoichiometry that is no longer sufficient for neutralization. While the neutralization threshold governs the upper limit of the stoichiometry of ADE the minimum number of antibodies capable of mediating ADE remains to be established. To explore this, we performed experiments in which equal concentrations of chE16 and N297Q chE16 were mixed at different proportions and used in dose-response experiments on K562 cells. Surprisingly, a relatively large fraction of the antibodies (50%) bound to the virion must be capable of interacting with FcγRIIa to mediate ADE (Fig. 6d). As the peak enhancement titer corresponds to occupancy of approximately 25% (30 of 120 accessible sites for E16), these studies suggest that ADE occurs efficiently when approximately 15–29 mAbs are bound to a virion at this epitope. Engagement by a single mAb does not appear to allow for ADE. Assuming an equal distribution of mAb on the virion surface, these studies suggest that several mAbs must be present at the interface of the virus and cell, highlighting a stoichiometry that is bordered at the lower end by requirements for stable attachment to the cell surface or cross-linking of more than a single Fc-γ receptor (Fig 6d).

Discussion

In this study, we investigated the impact of the number of antibodies bound to a particular neutralizing determinant on the infectivity of WNV. The functional properties of antibodies that recognized six partially overlapping epitopes on DIII were characterized. Two independent approaches suggest that for the most potent antibodies studied, neutralization was achieved upon engagement of a relatively small fraction of the available epitopes on the virion. Relative to their strength of binding, analysis of the dose response curves of the three most potent antibodies revealed that most of virions in the population (99%) were neutralized when an average of 55% of the epitopes were bound by antibody. For E24, this conclusion was supported by a second approach in which the number of epitopes on the virion was reduced and the impact on neutralization and ADE examined. E24 completely neutralized populations of RVPs that contain, on average, half the number of wild type epitopes. Moreover, neutralization of half of the virions in the population was achieved when an intact epitope was present on approximately 25% of the E proteins.

Quantitative analysis of the stoichiometric requirements for WNV neutralization is complicated by the heterogeneity of epitopes displayed on the surface of the virion. Differences in epitope accessibility shape the occupancy requirements for neutralization among antibodies that bind distinct epitopes on DIII of the E protein and highlight the significance of the shape and steepness of the dose-response curve. In this context, accessibility reflects not only differences in the solvent exposure of an epitopes on the virion, but also the dynamics of how bound antibodies impact subsequent antibody binding events on the surface of the virion. The three most potent antibodies displayed relatively steep dose-response curves, indicating greater neutralization per antibody molecule on the lower portion of the sigmoidal dose-response curve. These antibodies neutralize infection when a relatively low fraction of the accessible epitopes is bound. By comparison, neutralization by antibodies that recognize epitopes that are predicted to be less solvent-accessible structures exhibit relatively flat neutralization profiles and require engagement of a larger fraction (>90%) of epitopes on the virion. Together, these observations are consistent with a threshold requirement for neutralization. The occupancy requirements for different antibodies reflect a requirement to engage a larger fraction of less abundant epitopes to reach the threshold required for neutralization. For antibodies like E22, not all virions in the population appear to be capable of achieving this threshold, resulting a fraction resistant to neutralization even at saturating concentrations of antibody.

Two models for neutralization of viruses have been favored (reviewed in references (Burton et al., 2001; Della-Porta and Westaway, 1978; Klasse and Sattentau, 2002). “Single hit” models, based in large part upon kinetic studies, suggest that virus neutralization can be achieved following engagement by a single antibody molecule which may trigger conformational changes that result in a loss of virion infectivity. Alternatively, “multiple hit” models require docking of multiple antibodies to a single virion with neutralization occurring only after the number of antibodies bound to an individual virion surpasses a threshold required for virion inactivation. Our results support a multiple hit mechanism for WNV neutralization with the most potent DIII-specific antibodies blocking infection at a relatively low occupancy of the accessible epitopes. If the overlapping epitope shared by E16, E24 and E47 is accessible on the virion at a maximum of 120 locations as suggested (Kaufmann et al., 2006; Nybakken et al., 2005), neutralization of half the virions occurs when an average of 30 antibodies are bound to the virion at this location (25% of accessible sites). As antibodies that bind this epitope inhibit a post-attachment step during virus entry (Nybakken et al., 2005) and are thought to directly block fusion (Kaufmann et al., 2006), the stoichiometric requirement of antibodies that bind other epitopes on the virion, or neutralize via different mechanisms, may differ.

While the stoichiometry of antibody-mediated neutralization has been examined for several RNA and DNA viruses (reviewed in (Burton et al., 2001; Klasse and Sattentau, 2002) and references within), no information has been reported for enveloped viruses with icosahedral symmetry or those with class II fusion glycoproteins. Burton and colleagues have suggested the stoichiometry of neutralization reflects the number of antibodies required to “coat” the surface of the virion, which can be estimated as a function of the surface area of the virion (Burton et al., 2001). By comparison, extrapolating from the reported stoichiometry of four different viruses, the “coating theory” for neutralization predicts that 26 antibodies are sufficient to neutralize a virion with the 50 nM diameter of a flavivirus. While this number is in close agreement with the estimates for E24 reported within, we note that the binding footprint of 30 antibodies (30 x approximately 1,550 Å2 of surface area) at this epitope occludes a modest fraction (6%) of the total surface area of the 50 nM diameter virion highlighting the potential for steric effects associated with the relatively large size of the intact antibody on the mechanism of neutralization (Nybakken et al., 2005).

An unanswered question has been whether antibodies that promote ADE recognize unique epitopes or bind to common viral determinants with a specific stoichiometry. Considering ADE in the context of the stoichiometric requirements for neutralization suggests criteria for defining “enhancing” epitopes on the virion. Our studies of over 160 mAbs suggest that when a virion is bound by a number of antibodies that is not sufficient for neutralization, enhancement in Fc-γ receptor expressing cells is possible irrespective of the epitope (Oliphant et al., 2006)(T. Pierson, T. Oliphant, and M. Diamond, unpublished results). Not all epitopes on surface of the virion are equally accessible to antibody (Oliphant et al., 2006; Stiasny et al., 2006). In instances where the abundance of accessible epitopes on individual virions does not exceed the threshold for neutralization, enhancement of infection is possible even at saturating concentrations of antibody. Enhancement in Fc-γ receptor expressing cells of the “resistant fraction” of E22 and several mAbs that bind epitopes in DII was readily observed under saturating conditions (Oliphant et al., 2006)(S. Nelson and T. Pierson, manuscript in preparation). Similarly, mutant RVPs that contain a relatively small number of E24 epitopes exhibit a fraction of virions resistant to neutralization, a shift of the dose-response curve to the right, and a broadening of the concentrations that promote ADE on cells bearing Fc-γ receptors.

Overall, our studies have explored the quantitative relationships between antibody binding, neutralization, and enhancement of WNV infection. Flaviviruses may be a particularly unique case with respect to the stoichiometry of antibody function due to heterogeneity in epitope accessibility that stems from the quasi-icosahedral nature of the virion. The finding that antibody epitope specificity determines occupancy and neutralization will likely be relevant to the generation of vaccines and antibody-based therapeutics against flaviviruses, including WNV. Antibodies that recognize exposed neutralizing epitopes on the lateral ridge of DIII are highly effective in vivo (Morrey et al., 2006; Oliphant et al., 2005) because complete neutralization occurs at a lower occupancy, and thus, lower antibody concentration. Based on these findings, we suggest that vaccines against flaviviruses should be modified to target specific epitopes on E protein that elicit antibodies that bind with high occupancy at low concentrations for optimal protection and increased safety.

Experimental procedures

Cell lines

BHK-21 cells that stably propagate a lineage II WNV sub-genomic replicon (BHK-21 WNIIrep-G/Z cells) were created by transduction of BHK-21 (clone 15) cells with RVPs that encapsidate a replicon encoding both GFP and a gene product conferring resistance to the zeocin antibiotic (ShBle). These cells were maintained in Dulbecco’s Modified Eagles Media (DMEM) supplemented with 10% fetal bovine serum (FBS), Glutamax, 300 μg/ml zeocin, and a 1% penicillin/streptomycin (PS) solution. HEK-293T cells were maintained in DMEM, 10% FBS and 1% PS. K562, K562-DCSIGNR, and Raji-DCSIGNR cells were propagated in RPMI containing 10% FCS, 25mM HEPES, and 1% PS. All cells were cultured at 37°C in 7% CO2 and suspension cultures split daily to maintain cells at a density of approximately 3×105 cells/ml and in log phase growth. All cell culture reagents were purchased from Invitrogen (Carlsbad, CA).

Plasmids

Plasmids encoding the WNV prM-E and capsid structural genes have been described previously (Pierson et al., 2006). Neutralization escape mutations were introduced into the WNV E protein using an overlap extension PCR strategy with primers specific for the 5′ and 3′ end of the prM-E gene. PCR products were introduced into a pcDNA 3.1 vector using directional topoisomerase-mediated cloning. The sequence of the entire prM-E gene was confirmed by sequencing.

Epitope mapping

Epitope identification for E22 and E9 was performed using a previously described yeast-display method. Mapping data for E16, E24, E47 and E49 has been reported previously (Oliphant et al., 2005). Yeast expressing a random library of WNV DIII mutants were stained with Alexa-Fluor-647 conjugated mAbs E22 and E9 to identify loss of binding mutants as described (Nybakken et al., 2005). Three rounds of sorting were performed and individual colonies were tested for loss of binding by flow cytometry using a Becton Dickinson FACSCaliber flow cytometer. Plasmid was recovered using the Zymoprep Yeast Miniprep kit (Zymo Research, Orange, CA), transformed into DH5α competent cells (Stratagene, La Jolla, CA) and sequenced.

Production of WNV RVPs

WNV RVPs for neutralization studies were produced by transfection of BHK-21 WNIIrep-G/Z cells with DNA plasmids encoding the structural genes of WNV as recently described previously. Transfection was performed in T75 flasks using a total of 40 μg of DNA and Lipofectamine 2000 (Invitrogen, Carlsbad, CA). After removal of lipid-DNA complexes, cells were cultured in a low glucose formulation of DMEM containing 25mM HEPES, 10% FBS, and 1% PS. RVPs were harvested at 24 and 48 hours post transfection, filtered through a 0.22 μM filter, and frozen in aliquots at −80°C. Prior to use in neutralization studies, the infectious titer of RVP stocks was measured by infection of Raji-R cells with serial two-fold dilutions of virus particles. Infection was measured two days later by flow cytometry using a Becton Dickinson (Franklin Lakes, NJ) FACs Calibur. RVPs used during creation of the BHK-21 WNIIrep-G/Z cell line were produced by transfection of HEK-293T cells with plasmids encoding the WNIIrep-G/Z replicon and the WNV structural genes.

To produce RVPs incorporating both wild type E proteins and a variant encoding the T332K mutation, a modified complementation approach was employed. BHK-21 WNIIrep-G/Z cells were transfected with a plasmid encoding WNV capsid and varying proportions (by mass) of either wild type prM-E or the T332K variant. To estimate the proportion of both forms of E proteins present in the RVP population, RVP containing supernatants were concentrated by ultracentrifugation and analyzed by Western blot using two previously characterized antibodies. Total E protein was measured by using the DII-specific mAb E60 (Oliphant et al., 2005; Oliphant et al., 2006), whereas the fraction of wild type E protein in the RVP was approximated as a function of mAb E24-dependent signal (data not shown).

Neutralization and enhancement of WNV RVP infection

Neutralization studies were performed using the B-lymphoblastoic cell line Raji that stably expresses the c-type lectin DC-SIGNR (Raji-R). DC-SIGNR is a tetrameric c-type lectin that promotes more efficient WNV infection of cells via high-avidity interactions with complex or high mannose sugars present on the envelope proteins prM and E (Davis et al., 2006a; Davis et al., 2006b). Raji-R cells were selected for study because they bind WNV avidly, facilitating studies using the small quantities of virus particles (and thus viral antigen) required to ensure antibody remains in excess over all of the informative points of a neutralization profile.

WNV RVP stocks were diluted and incubated with mAb for 60 minutes at room temperature (in a 200 μl volume). Control experiments performed after a longer incubation (120 minutes) resulted in similar neutralization profiles, suggesting that one-hour is sufficient for the establishment of steady-state conditions. For antibody dose-response experiment, 19 two-fold dilutions of antibody were studied. Antibody-RVP complexes were then added to pre-plated cells in triplicate (5×104 cells per well of 96 well plate in 300 μl total volume). Infection was measured by flow cytometry at 48 hours post-infection. At least 5×104 events were collected at each point. The EC50 of each antibody was predicted by nonlinear regression analysis using a variable slope. Statistical comparisons of the neutralization sensitivity of different RVP populations were performed using the F-test (GraphPad Prism 4, GraphPad Software Inc., San Diego CA). Appropriate RVP dilutions for neutralization studies were established for each RVP preparation empirically. Initially, neutralization studies were performed with serial dilutions of RVPs to establish that the concentration of RVPs does not affect the observed potency (EC50) of a given antibody. Once established and used to characterize the potency of E16, this antibody was then used as a reference to characterize RVP stocks prior to inclusion in neutralization studies.

Antibody avidity measurements

Antibody avidity was measured as described previously (Sanchez et al., 2005). WNV subviral particles composed solely of prM and E, or RVPs were produced by transfection of HEK-293T cells. Twenty-four hours after transfection, media was replaced with a low-glucose formulation of DMEM without FBS. Virus particles were harvested 24 hours later, filtered, and captured overnight to the bottom of high protein-binding plates under alkaline conditions. Virus particle-coated plates were blocked and incubated in the presence of serial dilutions of anti-WNV antibodies under conditions of antibody excess. Bound antibody was detected using a horseradish peroxidase-conjugated goat anti-murine kappa chain antibody followed by colormetric analysis. To estimate the apparent avidity (AV50), the resulting data were fit to both a one-site binding model and a non-linear dose response method that does not assume engagement of a single uniform site on the virus particle (allowing for cooperative binding). Both methods yielded similar results (data not shown).

Acknowledgments

This work was supported by the Intramural Research Program of the NIH, National Institutes of Allergy and Infectious Diseases (NIAID). Additional support was provided by grants NIH F31 RR05074 and NIH U01 AI061373 (M.S.D), and the Pediatric Dengue Vaccine Initiative (T.C.P., M.S.D, and D.H.F).

M.S.D. is a consultant for MacroGenics, Inc, which has licensed the E16 antibody for commercial use.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Allison SL, Schalich J, Stiasny K, Mandl CW, Heinz FX. Mutational evidence for an internal fusion peptide in flavivirus envelope protein E. J Virol. 2001;75:4268–4275. doi: 10.1128/JVI.75.9.4268-4275.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andrewes CH, Elford WJ. Observations on anti-phage sera. I. ‘The percentage law’. British Journal of Experimental Pathology. 1933;14:367–376. [Google Scholar]
  3. Barrett AD, Gould EA. Antibody-mediated early death in vivo after infection with yellow fever virus. J Gen Virol . 1986;67(Pt 11):2539–2542. doi: 10.1099/0022-1317-67-11-2539. [DOI] [PubMed] [Google Scholar]
  4. Beasley DW, Barrett AD. Identification of neutralizing epitopes within structural domain III of the West Nile virus envelope protein. J Virol. 2002;76:13097–13100. doi: 10.1128/JVI.76.24.13097-13100.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ben-Nathan D, Lustig S, Tam G, Robinzon S, Segal S, Rager-Zisman B. Prophylactic and therapeutic efficacy of human intravenous immunoglobulin in treating west nile virus infection in mice. J Infect Dis. 2003;188:5–12. doi: 10.1086/376870. [DOI] [PubMed] [Google Scholar]
  6. Bhardwaj S, Holbrook M, Shope RE, Barrett AD, Watowich SJ. Biophysical characterization and vector-specific antagonist activity of domain III of the tick-borne flavivirus envelope protein. J Virol. 2001;75:4002–4007. doi: 10.1128/JVI.75.8.4002-4007.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brinton MA. The molecular biology of West Nile Virus: a new invader of the western hemisphere. Annu Rev Microbiol. 2002;56:371–402. doi: 10.1146/annurev.micro.56.012302.160654. [DOI] [PubMed] [Google Scholar]
  8. Burton DR, Saphire EO, Parren PW. A model for neutralization of viruses based on antibody coating of the virion surface. Curr Top Microbiol Immunol. 2001;260:109–143. doi: 10.1007/978-3-662-05783-4_7. [DOI] [PubMed] [Google Scholar]
  9. Camenga DL, Nathanson N, Cole GA. Cyclophosphamide-potentiated West Nile viral encephalitis: relative influence of cellular and humoral factors. J Infect Dis. 1974;130:634–641. doi: 10.1093/infdis/130.6.634. [DOI] [PubMed] [Google Scholar]
  10. Ceausu E, Erscoiu S, Calistru P, Ispas D, Dorobat O, Homos M, Barbulescu C, Cojocaru I, Simion CV, Cristea C, et al. Clinical manifestations in the West Nile virus outbreak. Rom J Virol. 1997;48:3–11. [PubMed] [Google Scholar]
  11. Chu JJ, Rajamanonmani R, Li J, Bhuvanakantham R, Lescar J, Ng ML. Inhibition of West Nile virus entry by using a recombinant domain III from the envelope glycoprotein. J Gen Virol. 2005;86:405–412. doi: 10.1099/vir.0.80411-0. [DOI] [PubMed] [Google Scholar]
  12. Colombage G, Hall R, Pavy M, Lobigs M. DNA-based and alphavirus-vectored immunisation with prM and E proteins elicits long-lived and protective immunity against the flavivirus, Murray Valley encephalitis virus. Virology. 1998;250:151–163. doi: 10.1006/viro.1998.9357. [DOI] [PubMed] [Google Scholar]
  13. Dauphin G, Zientara S, Zeller H, Murgue B. West Nile: worldwide current situation in animals and humans. Comp Immunol Microbiol Infect Dis. 2004;27:343–355. doi: 10.1016/j.cimid.2004.03.009. [DOI] [PubMed] [Google Scholar]
  14. Davis CW, Mattei LM, Nguyen HY, Ansarah-Sobrinho C, Doms RW, Pierson TC. The location of asparagine-linked glycans on West Nile virions controls their interactions with CD209 (DC-SIGN) J Biol Chem. 2006a doi: 10.1074/jbc.M605429200. [DOI] [PubMed] [Google Scholar]
  15. Davis CW, Nguyen HY, Hanna SL, Sanchez MD, Doms RW, Pierson TC. West Nile virus discriminates between DC-SIGN and DC-SIGNR for cellular attachment and infection. J Virol. 2006b;80:1290–1301. doi: 10.1128/JVI.80.3.1290-1301.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Deardorff E, Estrada-Franco J, Brault AC, Navarro-Lopez R, Campomanes-Cortes A, Paz-Ramirez P, Solis-Hernandez M, Ramey WN, Davis CT, Beasley DW, et al. Introductions of West Nile virus strains to Mexico. Emerg Infect Dis. 2006;12:314–318. doi: 10.3201/eid1202.050871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Della-Porta AJ, Westaway EG. A multi-hit model for the neutralization of animal viruses. J Gen Virol. 1978;38:1–19. doi: 10.1099/0022-1317-38-1-1. [DOI] [PubMed] [Google Scholar]
  18. Diamond MS, Shrestha B, Marri A, Mahan D, Engle M. B cells and antibody play critical roles in the immediate defense of disseminated infection by West Nile encephalitis virus. J Virol. 2003a;77:2578–2586. doi: 10.1128/JVI.77.4.2578-2586.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Diamond MS, Shrestha B, Mehlhop E, Sitati E, Engle M. Innate and adaptive immune responses determine protection against disseminated infection by West Nile encephalitis virus. Viral Immunol. 2003b;16:259–278. doi: 10.1089/088282403322396082. [DOI] [PubMed] [Google Scholar]
  20. Engle M, Diamond MS. Antibody prophylaxis and therapy against West Nile Virus infection in wild-type and immunodeficient mice. J Virol. 2003;77:12941–12949. doi: 10.1128/JVI.77.24.12941-12949.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Falconar AK. Identification of an epitope on the dengue virus membrane (M) protein defined by cross-protective monoclonal antibodies: design of an improved epitope sequence based on common determinants present in both envelope (E and M) proteins. Arch Virol. 1999;144:2313–2330. doi: 10.1007/s007050050646. [DOI] [PubMed] [Google Scholar]
  22. Gould EA, Buckley A. Antibody-dependent enhancement of yellow fever and Japanese encephalitis virus neurovirulence. J Gen Virol . 1989;70(Pt 6):1605–1608. doi: 10.1099/0022-1317-70-6-1605. [DOI] [PubMed] [Google Scholar]
  23. Gould LH, Sui J, Foellmer H, Oliphant T, Wang T, Ledizet M, Murakami A, Noonan K, Lambeth C, Kar K, et al. Protective and therapeutic capacity of human single chain Fv-Fc fusion proteins against West Nile virus. J Virol. 2005;79:14606–14613. doi: 10.1128/JVI.79.23.14606-14613.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Halstead SB. Neutralization and antibody-dependent enhancement of dengue viruses. Adv Virus Res. 2003;60:421–467. doi: 10.1016/s0065-3527(03)60011-4. [DOI] [PubMed] [Google Scholar]
  25. Hayes EB, Komar N, Nasci RS, Montgomery SP, O’Leary DR, Campbell GL. Epidemiology and transmission dynamics of West Nile virus disease. Emerg Infect Dis. 2005;11:1167–1173. doi: 10.3201/eid1108.050289a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kanai R, Kar K, Anthony K, Gould LH, Ledizet M, Fikrig E, Marasco WA, Koski RA, Modis Y. Crystal structure of west nile virus envelope glycoprotein reveals viral surface epitopes. J Virol. 2006;80:11000–11008. doi: 10.1128/JVI.01735-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kaufman BM, Summers PL, Dubois DR, Eckels KH. Monoclonal antibodies against dengue 2 virus E-glycoprotein protect mice against lethal dengue infection. Am J Trop Med Hyg. 1987;36:427–434. doi: 10.4269/ajtmh.1987.36.427. [DOI] [PubMed] [Google Scholar]
  28. Kaufmann B, Nybakken GE, Chipman PR, Zhang W, Diamond MS, Fremont DH, Kuhn RJ, Rossmann MG. West Nile virus in complex with the Fab fragment of a neutralizing monoclonal antibody. Proc Natl Acad Sci U S A. 2006;103:12400–12404. doi: 10.1073/pnas.0603488103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Klasse PJ, Moore JP. Quantitative model of antibody- and soluble CD4-mediated neutralization of primary isolates and T-cell line-adapted strains of human immunodeficiency virus type 1. J Virol. 1996;70:3668–3677. doi: 10.1128/jvi.70.6.3668-3677.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Klasse PJ, Sattentau QJ. Occupancy and mechanism in antibody-mediated neutralization of animal viruses. J Gen Virol. 2002;83:2091–2108. doi: 10.1099/0022-1317-83-9-2091. [DOI] [PubMed] [Google Scholar]
  31. Klein E, Ben-Bassat H, Neumann H, Ralph P, Zeuthen J, Polliack A, Vanky F. Properties of the K562 cell line, derived from a patient with chronic myeloid leukemia. Int J Cancer. 1976;18:421–431. doi: 10.1002/ijc.2910180405. [DOI] [PubMed] [Google Scholar]
  32. Komar N, Clark GG. West Nile virus activity in Latin America and the Caribbean. Rev Panam Salud Publica. 2006;19:112–117. doi: 10.1590/s1020-49892006000200006. [DOI] [PubMed] [Google Scholar]
  33. Kuhn RJ, Zhang W, Rossmann MG, Pletnev SV, Corver J, Lenches E, Jones CT, Mukhopadhyay S, Chipman PR, Strauss EG, et al. Structure of dengue virus: implications for flavivirus organization, maturation, and fusion. Cell. 2002;108:717–725. doi: 10.1016/s0092-8674(02)00660-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lanciotti RS, Roehrig JT, Deubel V, Smith J, Parker M, Steele K, Crise B, Volpe KE, Crabtree MB, Scherret JH, et al. Origin of the West Nile virus responsible for an outbreak of encephalitis in the northeastern United States. Science. 1999;286:2333–2337. doi: 10.1126/science.286.5448.2333. [DOI] [PubMed] [Google Scholar]
  35. Lindenbach BD, Rice CM. Flaviviridae: The Viruses and Their Replication. 4. Philadelphia: Lippincott, Wiliams & Wilkins; 2001. [Google Scholar]
  36. Littaua R, Kurane I, Ennis FA. Human IgG Fc receptor II mediates antibody-dependent enhancement of dengue virus infection. J Immunol. 1990;144:3183–3186. [PubMed] [Google Scholar]
  37. Mady BJ, Erbe DV, Kurane I, Fanger MW, Ennis FA. Antibody-dependent enhancement of dengue virus infection mediated by bispecific antibodies against cell surface molecules other than Fc gamma receptors. J Immunol. 1991;147:3139–3144. [PubMed] [Google Scholar]
  38. Modis Y, Ogata S, Clements D, Harrison SC. Structure of the dengue virus envelope protein after membrane fusion. Nature. 2004;427:313–319. doi: 10.1038/nature02165. [DOI] [PubMed] [Google Scholar]
  39. Modis Y, Ogata S, Clements D, Harrison SC. Variable surface epitopes in the crystal structure of dengue virus type 3 envelope glycoprotein. J Virol. 2005;79:1223–1231. doi: 10.1128/JVI.79.2.1223-1231.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Morens DM, Halstead SB, Marchette NJ. Profiles of antibody-dependent enhancement of dengue virus type 2 infection. Microb Pathog. 1987;3:231–237. doi: 10.1016/0882-4010(87)90056-8. [DOI] [PubMed] [Google Scholar]
  41. Morrey JD, Siddharthan V, Olsen AL, Roper GY, Wang H, Baldwin TJ, Koenig S, Johnson S, Nordstrom JL, Diamond MS. Humanized Monoclonal Antibody against West Nile Virus Envelope Protein Administered after Neuronal Infection Protects against Lethal Encephalitis in Hamsters. J Infect Dis. 2006;194:1300–1308. doi: 10.1086/508293. [DOI] [PubMed] [Google Scholar]
  42. Mukhopadhyay S, Kim BS, Chipman PR, Rossmann MG, Kuhn RJ. Structure of West Nile virus. Science. 2003;302:248. doi: 10.1126/science.1089316. [DOI] [PubMed] [Google Scholar]
  43. Mukhopadhyay S, Kuhn RJ, Rossmann MG. A structural perspective of the flavivirus life cycle. Nat Rev Microbiol. 2005;3:13–22. doi: 10.1038/nrmicro1067. [DOI] [PubMed] [Google Scholar]
  44. Nybakken GE, Nelson CA, Chen BR, Diamond MS, Fremont DH. Crystal structure of the west nile virus envelope glycoprotein. J Virol. 2006;80:11467–11474. doi: 10.1128/JVI.01125-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Nybakken GE, Oliphant T, Johnson S, Burke S, Diamond MS, Fremont DH. Structural basis of West Nile virus neutralization by a therapeutic antibody. Nature. 2005;437:764–769. doi: 10.1038/nature03956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Oliphant T, Engle M, Nybakken GE, Doane C, Johnson S, Huang L, Gorlatov S, Mehlhop E, Marri A, Chung KM, et al. Development of a humanized monoclonal antibody with therapeutic potential against West Nile virus. Nat Med. 2005;11:522–530. doi: 10.1038/nm1240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Oliphant T, Nybakken GE, Engle M, Xu Q, Nelson CA, Sukupolvi-Petty S, Marri A, Lachmi BE, Olshevsky U, Fremont DH, et al. Antibody Recognition and Neutralization Determinants on Domains I and II of West Nile Virus Envelope Protein. J Virol. 2006 doi: 10.1128/JVI.01732-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Parren PW, Mondor I, Naniche D, Ditzel HJ, Klasse PJ, Burton DR, Sattentau QJ. Neutralization of human immunodeficiency virus type 1 by antibody to gp120 is determined primarily by occupancy of sites on the virion irrespective of epitope specificity. J Virol. 1998;72:3512–3519. doi: 10.1128/jvi.72.5.3512-3519.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Petersen LR, Roehrig JT, Hughes JM. West Nile virus encephalitis. N Engl J Med. 2002;347:1225–1226. doi: 10.1056/NEJMo020128. [DOI] [PubMed] [Google Scholar]
  50. Phillpotts RJ, Stephenson JR, Porterfield JS. Passive immunization of mice with monoclonal antibodies raised against tick-borne encephalitis virus. Brief report. Arch Virol. 1987;93:295–301. doi: 10.1007/BF01310983. [DOI] [PubMed] [Google Scholar]
  51. Pierson TC, Sanchez MD, Puffer BA, Ahmed AA, Geiss BJ, Valentine LE, Altamura LA, Diamond MS, Doms RW. A rapid and quantitative assay for measuring antibody-mediated neutralization of West Nile virus infection. Virology. 2006;346:53–65. doi: 10.1016/j.virol.2005.10.030. [DOI] [PubMed] [Google Scholar]
  52. Pincus S, Mason PW, Konishi E, Fonseca BA, Shope RE, Rice CM, Paoletti E. Recombinant vaccinia virus producing the prM and E proteins of yellow fever virus protects mice from lethal yellow fever encephalitis. Virology. 1992;187:290–297. doi: 10.1016/0042-6822(92)90317-i. [DOI] [PubMed] [Google Scholar]
  53. Polack FP, Teng MN, Collins PL, Prince GA, Exner M, Regele H, Lirman DD, Rabold R, Hoffman SJ, Karp CL, et al. A role for immune complexes in enhanced respiratory syncytial virus disease. J Exp Med. 2002;196:859–865. doi: 10.1084/jem.20020781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Rey FA, Heinz FX, Mandl C, Kunz C, Harrison SC. The envelope glycoprotein from tick-borne encephalitis virus at 2 A resolution. Nature. 1995;375:291–298. doi: 10.1038/375291a0. [DOI] [PubMed] [Google Scholar]
  55. Roehrig JT, Staudinger LA, Hunt AR, Mathews JH, Blair CD. Antibody prophylaxis and therapy for flavivirus encephalitis infections. Ann N Y Acad Sci. 2001;951:286–297. doi: 10.1111/j.1749-6632.2001.tb02704.x. [DOI] [PubMed] [Google Scholar]
  56. Sanchez MD, Pierson TC, McAllister D, Hanna SL, Puffer BA, Valentine LE, Murtadha MM, Hoxie JA, Doms RW. Characterization of neutralizing antibodies to West Nile virus. Virology. 2005;336:70–82. doi: 10.1016/j.virol.2005.02.020. [DOI] [PubMed] [Google Scholar]
  57. Schonning K, Lund O, Lund OS, Hansen JE. Stoichiometry of monoclonal antibody neutralization of T-cell line-adapted human immunodeficiency virus type 1. J Virol. 1999;73:8364–8370. doi: 10.1128/jvi.73.10.8364-8370.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Sejvar JJ, Haddad MB, Tierney BC, Campbell GL, Marfin AA, Van Gerpen JA, Fleischauer A, Leis AA, Stokic DS, Petersen LR. Neurologic manifestations and outcome of West Nile virus infection. Jama. 2003;290:511–515. doi: 10.1001/jama.290.4.511. [DOI] [PubMed] [Google Scholar]
  59. Stiasny K, Kiermayr S, Holzmann H, Heinz FX. Cryptic properties of a cluster of dominant flavivirus cross-reactive antigenic sites. J Virol. 2006;80:9557–9568. doi: 10.1128/JVI.00080-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Tao MH, Morrison SL. Studies of aglycosylated chimeric mouse-human IgG. Role of carbohydrate in the structure and effector functions mediated by the human IgG constant region. J Immunol. 1989;143:2595–2601. [PubMed] [Google Scholar]
  61. Tesh RB, Travassos da Rosa AP, Guzman H, Araujo TP, Xiao SY. Immunization with heterologous flaviviruses protective against fatal West Nile encephalitis. Emerg Infect Dis. 2002;8:245–251. doi: 10.3201/eid0803.010238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Vazquez S, Guzman MG, Guillen G, Chinea G, Perez AB, Pupo M, Rodriguez R, Reyes O, Garay HE, Delgado I, et al. Immune response to synthetic peptides of dengue prM protein. Vaccine. 2002;20:1823–1830. doi: 10.1016/s0264-410x(01)00515-1. [DOI] [PubMed] [Google Scholar]
  63. Volk DE, Beasley DW, Kallick DA, Holbrook MR, Barrett AD, Gorenstein DG. Solution structure and antibody binding studies of the envelope protein domain III from the New York strain of West Nile virus. J Biol Chem. 2004;279:38755–38761. doi: 10.1074/jbc.M402385200. [DOI] [PubMed] [Google Scholar]
  64. Wang T, Anderson JF, Magnarelli LA, Wong SJ, Koski RA, Fikrig E. Immunization of mice against West Nile virus with recombinant envelope protein. J Immunol. 2001;167:5273–5277. doi: 10.4049/jimmunol.167.9.5273. [DOI] [PubMed] [Google Scholar]
  65. Yang X, Kurteva S, Lee S, Sodroski J. Stoichiometry of antibody neutralization of human immunodeficiency virus type 1. J Virol. 2005;79:3500–3508. doi: 10.1128/JVI.79.6.3500-3508.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zhang Y, Zhang W, Ogata S, Clements D, Strauss JH, Baker TS, Kuhn RJ, Rossmann MG. Conformational changes of the flavivirus E glycoprotein. Structure (Camb) 2004;12:1607–1618. doi: 10.1016/j.str.2004.06.019. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES