GP38-targeting monoclonal antibodies protect adult mice against lethal Crimean-Congo hemorrhagic fever virus infection

Abstract

Crimean-Congo hemorrhagic fever virus (CCHFV) is an important human pathogen. Limited evidence suggests that antibodies can protect humans against lethal CCHFV disease but the protective efficacy of antibodies has never been evaluated in adult animal models. Here, we used adult mice to investigate the protection provided against CCHFV infection by glycoprotein-targeting neutralizing and non-neutralizing monoclonal antibodies (mAbs). We identified a single non-neutralizing antibody (mAb-13G8) that protected adult type I interferon-deficient mice >90% when treatment was initiated before virus exposure and >60% when administered after virus exposure. Neutralizing antibodies known to protect neonatal mice from lethal CCHFV infection failed to confer protection regardless of immunoglobulin G subclass. The target of mAb-13G8 was identified as GP38, one of multiple proteolytically cleaved glycoproteins derived from the CCHFV glycoprotein precursor polyprotein. This study reveals GP38 as an important antibody target for limiting CCHFV pathogenesis and lays the foundation to develop immunotherapeutics against CCHFV in humans.

Fig. 1. mAb-13G8 protects against lethal disease in IFNAR^−/−mice.

(A) IFNAR^−/− mice (n = 10 per group) were infected with CCHFV strain IbAr 10200 by the subcutaneous route. Mice were injected with mAb-13G8, mAb-8A1, or a combination of the two on day −1 (1 mg of total mAb concentration) by the intraperitoneal route. As a control, mice were treated with PBS. Survival and percent group weight change were recorded. Log-rank test, ****P < 0.0001. (B) IFNAR^−/− mice (n = 10 per group) were infected with virus as in (A). Mice were treated with two doses of the mAb-13G8 (1 mg per dose) or PBS on days −1/+3, +1/+4, or + 2/+5, and survival and weight were monitored. Log-rank test, ****P < 0.0001, **P = 0.0030. (C) IFNAR^−/− mice (n = 10 per group) were infected as in (A). Mice were treated with two doses of mAb-13G8, mAb-11E7, or a combination of both (1 mg of total mAb concentration per dose), and survival and weight were monitored. A nonspecific murine antibody was used as a control. Log-rank test, ****P < 0.0001. (D) Hematoxylin and eosin staining of the livers and spleens from mice infected with CCHFV strain IbAr 10200 and treated with mAb-13G8 or an isotype control antibody. Uninfected mice treated with mAb-13G8 serve as a negative control. (E) In situ hybridization (ISH) staining showing the presence of CCHFV RNA (red) in the liver and spleen of mice taken on day 4 after virus exposure. Cells were counterstained with hematoxylin.

Fig. 2. mAb-13G8 interacts with the GP38 molecule.

(A) Simplified schematic depicting CCHFV M-segment glycoprotein processing, cleaved by signal peptidases such as subtilisin kexin isozyme-1/site-1 protease (SKI-1/S1P). (B) The indicated mAbs were serially diluted 10-fold (starting at 1:100) and incubated with purified G_N ectodomain. H3C8 is an irrelevant murine mAb. Sera from a CCFHV-infected human were used as a positive control along with a negative control human sample. ND, not detected. (C) Capture ELISA was developed using mAb-13G8 or a negative control (CNTL) (mAb-QC03) as capture antibody and anti–M-segment polyclonal rabbit antisera as detection antibody. Two-way ANOVA, ****P < 0.0001. (D) 293T cells were transfected with constructs expressing the indicated proteins and protein expression analyzed by Western blot at 24 and 48 hours. G_N was detected using anti-G_N 4093 (1:1000), and G_C was detected using mAb-11E7 (1:1000). Numbers on the left refer to molecular weight standard (kilodaltons). (E) Capture antibody for GP38 presence in the supernatant. For bait, 293T cells were transfected with plasmids expressing full-length strain IbAr 10200 M segment (M-seg), ΔMucΔGP38, tPA-GP38¹⁰²⁰⁰, or a negative control (NEG CNTL) plasmid. Two-way ANOVA, **P < 0.001, ***P < 0.0001.

Fig. 3. Heterologous protection of mice by mAb-13G8.

(A) Percent identical and percent divergence of CCHFV GP38 amino acid sequences from the indicated strains were determined using MegAlign Ver. 13.0.0 (DNASTAR software). (B) IFNAR^−/− (n = 10 per group) mice were treated with two doses of the mAb-13G8 or an isotype control antibody on day −1/+3 and challenged subcutaneously with either IbAr 10200 or Afg09-2990. Survival and percent weight change were monitored. Log-rank test, ****P < 0.0001, ***P = 0.0002. (C) ISH staining from the livers and spleens from strain Afg09-2990–infected mice harvested on day 4 (n = 3). ISH was conducted as in Fig. 1E. (D) Hematoxylin and eosin of the liver and spleen from Afg09-2990–infected mice taken on day 4 (n = 3). Top left: Focal area of mild inflammation (black arrow) and occasional Kupffer cell hypertrophy (arrowhead). Middle left: A vessel is occluded by a fibrin thrombus (white arrow) and surrounded by an area of coagulative necrosis (circled). These mice were infected at the same time as those in Fig. 1.

Fig. 4. Fc domains play a role in mAb-13G8–mediated protection.

(A) FcR^−/−, C3^−/−, or B6;129 (wild-type) mice (n = 8 per group) were injected subcutaneously with two doses of the mAb-13G8 day −1/+3 or an isotype control (1 mg per dose). IFN-I was blocked on day +1 using mAb-5A3 (1 mg/ml) injected intraperitoneally. Mice were challenged with 100 PFU of CCHFV IbAr 10200 by the intraperitoneal route. Survival and weight were plotted. Log-rank test, ****P < 0.0001, ***P = 0.0004, *P = 0.0201. (B) B6 mice (n = 8) were treated with mAb-10E11, mAb-11E7, or isotype control on day −1/+3 (1 mg per dose) and infected as in (A). Log-rank test, *P = 0.0433.

Fig. 5. GP38 localization to the viral envelope and plasma membrane.

(A) CCHF_VLP or VEE_VLP were incubated with the indicated antibodies and then stained with an anti-mouse secondary antibody conjugated to 10-nm gold particles. Samples were then examined by electron microscopy for the presence of GP38 or G_C. (B) 293T cells were transfected with full-length M segment (M-seg), ΔMucΔGP38, or an irrelevant protein (arenavirus glycoprotein precursor, NEG). To detect surface GP38 and G_C, nonpermeabilized cells were incubated with mAb-13G8 or mAb-11E7, respectively. Cells were then stained with an anti-mouse secondary antibody conjugated with Alexa Fluor 488 and analyzed. Ten thousand cells were counted per sample, and data were plotted with FlowJo software. (C) Vero E6 cells were transfected with the indicated plasmids and incubated with MAb-10E11, then stained with an anti-mouse secondary antibody conjugated with Alexa Fluor 488, and analyzed by confocal microscopy.

Fig. 6. Characterization of mAb and human convalescent sera interactions with the GP38 molecule.

(A) GP38his was plated in the wells of a 96-well plate (500 ng per well). The indicated mAbs were serially diluted 10-fold (from 1:100) and incubated with purified protein. End point titers were calculated as described in Materials and Methods. Data were plotted as a mean titer for each group. Asterisk indicates that mAb-6B12 was run in a separate assay. The dashed line indicates the limit of detection. (B) Sera from CCHF human survivors or a negative control sample were serially diluted in a purified GP38, G_N, or N protein ELISA. The dashed line indicates the limit of detection. (C) Epitope grouping of GP38-reactive mAbs based on competitive ability. Antibodies are grouped on the basis of binding permissivity relative to their competing partner. Colors represent the degree of permissivity, with green denoting low competition (≥76), yellow weak competition (51 to 75), orange intermediate competition (26 to 51), and red high competition (≤25). (D) Schematic depicting putative GP38 antibody competition groups using antibody from (A).