Complement protein C1q reduces the stoichiometric threshold for antibody-mediated neutralization of West Nile virus

Abstract

Virus neutralization is governed by the number of antibodies that bind a virion during the cellular entry process. Cellular and serum factors that interact with antibodies have the potential to modulate neutralization potency. Although the addition of serum complement can increase the neutralizing activity of antiviral antibodies in vitro, the mechanism and significance of this augmented potency in vivo remain uncertain. Herein, we show that the complement component C1q increases the potency of antibodies against West Nile virus by modulating the stoichiometric requirements for neutralization. The addition of C1q does not result in virolysis but instead reduces the number of antibodies that must bind the virion to neutralize infectivity. For IgG subclasses that bind C1q avidly, this reduced stoichiometric threshold falls below the minimal number of antibodies required for antibody-dependent enhancement (ADE) of infection of cells expressing Fc-gamma receptors (CD32) and explains how C1q restricts the ADE of flavivirus infection.

Figure 1. C1q augments the neutralization potency of anti-WNV antibodies

(A) Serial dilutions of E16 (murine IgG_2b) were incubated with WNV RVP in the presence of media, 5% fresh or heat-inactivated mouse serum prior to infection of Raji cells that express the attachment factor DC-SIGNR (Davis et al., 2006). Forty hours later, cells were fixed and analyzed by flow cytometry for GFP expression. (B) Experiments were performed as described in panel (A) except mixtures of RVP and E16 were incubated with C1q^-/- or C3^-/- fresh mouse serum, or purified C1q (50 mg/ml). (C-D) The impact of purified C1q on the neutralizing activity of affinity purified IgG from naïve or WNV-immune mouse serum (C) or convalescent heat-inactivated human serum from WNV-infected patients (D) was assayed as described in panel (A). In all cases, data are expressed as infectivity relative to conditions in the absence of antibody. Error bars represent the standard error of the mean. The number of independent replicates is indicated in the text.

Figure 2. C1q reduces the stoichiometric requirements for neutralization

(A) WNV RVPs that differ with respect to the number of DIII-LR epitopes incorporated into the average virion were produced by genetic complementation. WNV RVPs that incorporate different proportions of WT E proteins (indicated on the right) were incubated with serial dilutions of murine E16 in the absence (left panel) or presence (right panel) of 50 μg/ml purified human C1q protein prior to infection of Raji DC-SIGNR cells. Infection was monitored using flow cytometry as described in Fig. 1. (B) The size of the population of RVPs resistant to neutralization by saturating concentrations of murine E16 is summarized for RVPs composed of 100%, 50%, 25%, and 10% WT E proteins. The average of data from 4-10 independent experiments performed using 4-6 different preparations of RVPs is shown; error bars display the standard error of the mean. (C) The size of the resistant fraction observed following the incubation of RVPs composed of 25% WT E proteins with humanized isotypes of E16 in the presence or absence of purified C1q is summarized as described for panel (B). The average of 5 experiments using at least 4 independent preparations of RVPs is presented. (D) To measure the ability of E16 to support ADE in the presence or absence of C1q, complexes of virus, antibody, and complement (where indicated) produced for the experiments described in (A) were also used to infect Fc-γR-expressing K562 cells. Infection was expressed as infectivity relative to conditions in the absence of antibody. Error bars represent the standard error of the mean. Asterisks indicate statistical significance as judged by a paired Students T-test (**, P ≤ 0.01; ***, P < 0.001).

Figure 3. C1q augments neutralization of antibodies that recognize poorly accessible determinants on WNV

The efficiency of the maturation of WNV RVPs was manipulated as described previously (Nelson et al., 2008). To reduce the efficiency of virion maturation (immature WNV), WNV RVPs were produced in the presence of ammonium chloride (NH₄Cl, 20 mM); an increase in maturation efficiency (mature WNV) was achieved by transfection of RVP-producing cells with a plasmid that expresses the human furin protease. Serial dilutions of the fusion loop-reactive mAb E53 were mixed with RVPs produced using standard approaches (standard WNV, left panel), mature WNV preparations (center panel), and immature WNV preparations (right panel) in the presence (red curves) or absence (blue curves) of purified human C1q protein prior to infection of Raji-DC-SIGNR cells. Forty hours later, cells were fixed and analyzed by flow cytometry for GFP expression. The results are representative to 2-3 independent experiments.

Figure 4. Cross-linking of antibody bound to WNV is sufficient to lower the stoichiometric threshold for neutralization

WNV RVPs composed of 25% WT E proteins were incubated with serial four-fold dilutions of murine E16 in the absence or presence of anti-murine antibody-specific IgG, Fab₂, or Fab (2 μg/ml). RVP-antibody complexes were used to infect duplicate wells containing Raji DC-SIGNR cells. Infection was monitored using flow cytometry as described in Fig. 1. The size of the resistant fraction observed at high concentrations of antibody was enumerated by non-linear regression analysis and is shown on the y-axis. Error bars display the standard error of the mean from 2-3 independent assays.

Figure 5. C1q augments hu-E16-mediated protection in vivo

Wild-type and C1q^-/- mice were passively transferred with serial 10-fold reductions in dose (ranging from 67 to 0.067 mg/kg) of E16 subclass-switch variants hu-IgG₁ (n ≥ 13 mice/dose), hu-IgG₂ (n ≥ 15 mice/dose), hu-IgG₃ (n ≥ 14 mice/dose), and an A330L variant of E16 hu-IgG₃ (IgG₃ A330L)(n ≥ 11 mice/dose) one day prior to infection with 10² PFU of WNV and subsequently monitored for morbidity. Survival data from at least three independent experiments were analyzed by log-rank test, and IC50s were calculated by non-linear regression of survival percentage at each mAb dose. Error bars represent the standard error. Asterisks indicate statistical significance as judged by analysis of variance and an F Test (**, p = 0.001).