Structural characterization of protective non-neutralizing antibodies targeting Crimean-Congo hemorrhagic fever virus

Abstract

Crimean-Congo Hemorrhagic Fever Virus (CCHFV) causes a life-threatening disease with up to a 40% mortality rate. With no approved medical countermeasures, CCHFV is considered a public health priority agent. The non-neutralizing mouse monoclonal antibody (mAb) 13G8 targets CCHFV glycoprotein GP38 and protects mice from lethal CCHFV challenge when administered prophylactically or therapeutically. Here, we reveal the structures of GP38 bound with a human chimeric 13G8 mAb and a newly isolated CC5-17 mAb from a human survivor. These mAbs bind overlapping epitopes with a shifted angle. The broad-spectrum potential of c13G8 and CC5-17 and the practicality of using them against Aigai virus, a closely related nairovirus were examined. Binding studies demonstrate that the presence of non-conserved amino acids in Aigai virus corresponding region prevent CCHFV mAbs from binding Aigai virus GP38. This information, coupled with in vivo efficacy, paves the way for future mAb therapeutics effective against a wide swath of CCHFV strains.

Fig. 1. c13G8 interactions with nairovirus GP38.

a BLI binding kinetics of c13G8 with nairovirus GP38s. b Overlay of CCHFV IbAr10200 GP38 (PDBID 6VKF) (wheat) with our solved CCHFV Hoti GP38 Structure (PDBID 8DC5) (blue) c, c13G8 Fab complexed with GP38 Hoti d, LCDR3 (red) interaction with Hoti GP38 (blue) e, HCDR3 (orange) interaction with Hoti GP38 (blue).

Fig. 2. Generation of human-derived anti-GP38 mAbs.

a Six CCHF survivors in Trabzon (Turkey) were recruited, and their plasma were collected to determine their ELISA titers against GP38 Turkey2004. b Patient CC5 was selected, and its plasma was used to mine 11 mAbs which were then retested against GP38. ELISA experiments were performed in duplicate (n = 2.) c Neutralization potential of anti-GP38 mAbs. Foci reduction neutralization assays were performed with CCHFV-ZsG. These data represent one independent experiment out of 2. The average values of replicate wells (n = 4) are plotted, and error bars represent the associated standard deviation of the replicates. d Seven anti-GP38 CC5 antibodies were then tested against known Group 1, Group 2, and Group 3 antibodies. <33% binding indicates competition (red), 34–66% indicate partial competition (blue), >67% indicate no competition (green). e BLI binding kinetics of CC5-17 with nairovirus GP38s f, CC5-17 Fab (Yellow and Green) complexed with GP38 Hoti (Blue) (PDBID 8DDK) g, HCDR3 (Green) binding and LCDR3 (Yellow) binding interactions with GP38 strain Hoti (blue).

Fig. 3. GP38 site I investigation.

a Comparison of c13G8 (magenta) vs. CC5-17 (lime) engagement with GP38 site I epitope. b Surface area comparison of site I epitope, magenta showing surface area of 13G8 engagement, lime denoting surface area of CC5-17 engagement, and orange denoting the shared surface area of the two Fabs. c Homology model of GP38 from Aigai virus (orange) overlay on complex structure. d Sequence alignment of relevant residues on GP38, purple stars denote sites of interest. e Single mutations were conducted at GP38 Hoti sites 292 and 296 and Aigai virus GP38 sites 299, 301, and 303. These mutant proteins were then tested via BLI to obtain their binding kinetics.

Fig. 4. Efficacy of site I antibodies in mice against CCHFV infection.

Groups of IFNAR^−/− mice were infected subcutaneously with CCHFV Turkey-2004 (target dose: 100 TCID50) and treated intraperitoneally with c13G8 (blue) or CC5-17 (light blue) (n = 6) at indicated timepoints and doses and followed for 21 days post-challenge. For this study, both antibodies were generated on the human IgG1 and kappa light chain backbone. Mean weight change (baseline at day 0), survival and clinical scores in mice following: a 1 mg mAb treatment at 30 min or +1/+4 days after challenge or b, 0.25 mg mAb dose treatment at 30 min or +1/+4 days after challenge. In parallel, groups of mice were treated intraperitoneally with IgG1 isotype control (1 mg dose only, n = 4) at indicated timepoints (mean weight loss and survival indicated by gray line). In contrast to mice treated with 1 mg of 13G8 or CC5-17, all mice treated with 1 mg of isotype control (at indicated timepoints, c developed severe clinical illness and succumbed to infection. Clinical scores ≥10 indicate end-point criteria. Gray boxes in clinical scores indicate animals removed from study due to fatal disease.