Neutralizing monoclonal antibodies against the Gc fusion loop region of Crimean-Congo hemorrhagic fever virus

Abstract

Crimean-Congo hemorrhagic fever virus (CCHFV) is a highly pathogenic tick-borne virus, prevalent in more than 30 countries worldwide. Human infection by this virus leads to severe illness, with an average case fatality of 40%. There is currently no approved vaccine or drug to treat the disease. Neutralizing antibodies are a promising approach to treat virus infectious diseases. This study generated 37 mouse-derived specific monoclonal antibodies against CCHFV Gc subunit. Neutralization assays using pseudotyped virus and authentic CCHFV identified Gc8, Gc13, and Gc35 as neutralizing antibodies. Among them, Gc13 had the highest neutralizing activity and binding affinity with CCHFV Gc. Consistently, Gc13, but not Gc8 or Gc35, showed in vivo protective efficacy (62.5% survival rate) against CCHFV infection in a lethal mouse infection model. Further characterization studies suggested that Gc8 and Gc13 may recognize a similar, linear epitope in domain II of CCHFV Gc, while Gc35 may recognize a different epitope in Gc. Cryo-electron microscopy of Gc-Fab complexes indicated that both Gc8 and Gc13 bind to the conserved fusion loop region and Gc13 had stronger interactions with sGc-trimers. This was supported by the ability of Gc13 to block CCHFV GP-mediated membrane fusion. Overall, this study provides new therapeutic strategies to treat CCHF and new insights into the interaction between antibodies with CCHFV Gc proteins.

Fig 1. Preliminary screening of neutralizing antibodies.

(A). Flow chart of nAb screening by CCHFV pseudotyped virus. *G-VSVΔG/GFP was used to infect HEK293T cells that were pre-transfected with CCHFV GP 24 h earlier. After incubation at 37°C for 24 h, the supernatant containing pseudotyped CCHFV was collected. For neutralization assay, 3000 TCID₅₀ pseudotyped CCHFV was mixed with mAbs at 37°C for 1 h, and the mixture was added to Vero E6 cells for 1 h. After an additional 24 h incubation at 37°C, GFP expressing cells were counted under a fluorescence microscope. (B). Thirty-seven mAbs with 10, 40, and 160 dilution fold were tested for neutralization effects against pseudotyped CCHFV as mentioned above. Twenty-five mAbs could inhibit > 50% pseudotyped CCHFV infection at the dilution of 1:160. (C) Representative images at the antibody dilution of 1:160.

Fig 2. Neutralizing activities of Gc mAbs against live CCHFV-YL16070 in vitro.

IC₅₀ values of Gc8, Gc13, and Gc35 against live CCHFV-YL16070 at an MOI of 0.01. After incubation with serially diluted mAbs, the virus-antibody mixture was used to infect Vero E6 cells. Three days later, the viral copies in supernatant were detected by qRT-PCR (A) and the infectious progeny in the supernatant was determined using end-point dilution assay (B). The inhibition rates to DMSO control were calculated and graphed using GraphPad Prism. The experiments were performed in triplicates.

Fig 3. In vivo protective efficacy of Gc mAbs against CCHFV challenge.

(A). Weight loss and survival curves of infected mice treated with Gc8, Gc13 and Gc35. The C57BL/6J-IFNAR^-/- mice (7–8 mice per group) were pre-treated with 50 mg/kg mAbs or equivalent volume of PBS control via intraperitoneal route. After 24 h, mice were infected with 5,000 TCID₅₀ of CCHFV-YL16070 via the intraperitoneal route. Weight loss and clinical symptoms were monitored daily. Statistical analysis was performed using Simple survival analysis (Kaplan-Meier). *P < 0.05. (B). Viral copies in the liver and spleen of infected mice. Total RNA in the infected tissues were extracted and transcribed into cDNA. Viral RNA was quantified by qRT-PCR and transformed into RNA copies per gram of tissues.

Fig 4. Antibody recognition of CCHFV-infected cells detected by western blot and immunofluorescent analysis.

(A). Protein samples of CCHFV-YL16070 infected Vero E6 cells were subjected to western blot analysis. Gc8, Gc13, Gc35, and negative mouse serum were used as primary antibodies at a dilution of 1:5,000. The bands, approximately 75 kDa, indicated by the black arrow suggested that all 3 mAbs could recognize linear epitopes of Gc. (B). Vero E6 cells were infected with CCHFV-YL16070 at an MOI of 1. At 72 h p.i., samples were collected and subjected to indirect immunofluorescence assay using Gc8, Gc13, and Gc35 as the primary antibodies at a dilution of 1:5,000. Fluorescent signals represented Gc proteins detected by the three mAbs.

Fig 5. Binding characteristics of Gc8, Gc13, and Gc35.

(A). Competitive ELISA of Gc mAbs for epitope determination. Biotin-labeled antibodies and serially-diluted unlabeled antibodies were mixed and added to sGc-S-tag protein-coated plates for ELISA and the OD values were determined using a microplate reader. The results showed that Gc8 and Gc13 could compete with each other as the OD values decreased by increasing another antibody concentration. (B). The recognizing regions in Gc proteins by mAb were determined using western blot analysis. Different domains (DI, DII, DIII, Stem region, and fusion loops) of Gc were classified according to the Gc structure information (PDB: 7FGF) and shown in different colors. Four truncated Gc fragments (gray bars) were cloned into pET32a vector and protein expression was induced by IPTG. Protein samples were collected and subjected to western blot using the mAbs as primary antibodies at a dilution of 1:5,000. (C). Binding affinity assay of Gc8, Gc13, and Gc35 to sGc-S-tag by BLI. The binding curves and kinetics of association and dissociation were analyzed using ForteBio Data Analysis Software.

Fig 6. Gc8 and Gc13 block CCHFV GP-mediated membrane fusion.

(A). Flow chart for fusion inhibition assay. Huh-7 cells were transfected with a pT7-eGFP plasmids and mixed with BSR-T7/5 cells transfected with pCAGGS-CCHFV-GPdel53aa. After induction by low-pH medium, these cells would fuse with each other to form syncytia with GFP expression, and the addition of nAbs, but not medium control, would inhibit this process. (B). Quantification of syncytium inhibition rate of Gc8, Gc13, and Gc35 at 10 and 100 ng/mL. Statistical analysis was performed using unpaired T-test. **P < 0.01, ***P < 0.001. (C). Representative fluorescent images of each nAb at 100 ng/mL were shown and the edges of syncytia are outlined using a white line.

Fig 7. Binding stoichiometry between sGc-trimers and Fab fragments.

(A). Representative 2D class averages showing that the sGc-trimer can bind to one Gc8 Fab (left) or simultaneously interact with two Gc8 Fabs (right). The arrows indicate the bound Fab molecules. (B) Representative 2D class averages of sGc-trimer in complex with one Gc13 (left) or two Gc13 Fab (right). (C) Orthogonal views of molecular surface representation for sGc-trimers without bound Fab (EMD-31579). The three sGc protomers are presented in red, purple and green. The three fusion loops from the green sGc molecule are presented in cartoon form in the inset. (D–F) Orthogonal views of the sGc-trimer in complex with a single Gc8 Fab (D), two Gc13 Fab, (E) and three Gc13 Fab (F). Density corresponding to Fabs was presented in gray and roughly outlined by elliptical dashed lines.

Fig 8. Interactions of Gc8/13 Fabs with CCHFV Gc.

(A–B) Close-up view of the hydrogen bonding networks between fusion loops of CCHFV Gc and Gc8 (A) or Gc13 (B). Variable heavy and light (VH and VL) chains are presented in distinct colors for both Gc8 and Gc13. (C) Fab–Gc complexes, focusing on their interface, are superimposed on fusion loops of Gc. Color codes are similar to those in (A) and (B). (D) Structural comparison between Gc13–Gc and ADI-37801 complexes. Fusion loops are superimposed.