Monoclonal antibodies (mAb) against the envelope (E) protein of Zika virus (ZIKV) have shown great potential as therapeutics against the Zika epidemics. However, their use as a therapy may predispose treated individuals to severe infection by the related dengue virus (DENV) via antibody-dependent enhancement of infection (ADE). Here, we generated a broadly neutralizing flavivirus mAb, ZV1, with an identical protein backbone but different Fc glycosylation profiles. The three glycovariants, produced in wild-type (WT) and glycoengineered ΔXF Nicotiana benthamiana plants and in Chinese hamster ovary cells (ZV1WT, ZV1ΔXF, and ZV1CHO), respectively, showed equivalent neutralization potency against both ZIKV and DENV. By contrast, the three mAb glycoforms demonstrated drastically different ADE activity for DENV and ZIKV infection. While ZV1CHO and ZV1ΔXF showed ADE activity upon DENV and ZIKV infection, ZV1WT totally forwent its ADE. Importantly, all three glycovariants exhibited antibody-dependent cellular cytotoxicity (ADCC) against virus-infected cells, with increased potency by the fucose-free ZV1ΔXF glycoform. Moreover, the in vivo efficacy of the ADE-free ZV1WT was demonstrated in a murine model. Collectively, we demonstrated the feasibility of modulating ADE by Fc glycosylation, thereby establishing a novel approach for improving the safety of flavivirus therapeutics. Our study also underscores the versatile use of plants for the rapid expression of complex human proteins to reveal novel insight into antibody function and viral pathogenesis.
Antibody-dependent enhancement of infection (ADE) is clinically relevant to Dengue virus (DENV) infection and poses a major risk to the application of monoclonal antibody (mAb)-based therapeutics against related flaviviruses such as the Zika virus (ZIKV). Here, we tested a two-tier approach for selecting non-cross-reactive mAbs combined with modulating Fc glycosylation as a strategy to doubly secure the elimination of ADE while preserving Fc effector functions. To this end, we selected a ZIKV-specific mAb (ZV54) and generated three ZV54 variants using Chinese hamster ovary cells and wild-type (WT) and glycoengineered ΔXF Nicotiana benthamiana plants as production hosts (ZV54CHO, ZV54WT, and ZV54ΔXF). The three ZV54 variants shared an identical polypeptide backbone, but each exhibited a distinct Fc N-glycosylation profile. All three ZV54 variants showed similar neutralization potency against ZIKV but no ADE activity for DENV infection, validating the importance of selecting the virus/serotype-specific mAbs for avoiding ADE by related flaviviruses. For ZIKV infection, however, ZV54CHO and ZV54ΔXF showed significant ADE activity while ZV54WT completely forwent ADE, suggesting that Fc glycan modulation may yield mAb glycoforms that abrogate ADE even for homologous viruses. In contrast to the current strategies for Fc mutations that abrogate all effector functions along with ADE, our approach allowed the preservation of effector functions as all ZV54 glycovariants retained antibody-dependent cellular cytotoxicity (ADCC) against the ZIKV-infected cells. Furthermore, the ADE-free ZV54WT demonstrated in vivo efficacy in a ZIKV-infection mouse model. Collectively, our study provides further support for the hypothesis that antibody–viral surface antigen and Fc-mediated host cell interactions are both prerequisites for ADE, and that a dual-approach strategy, as shown herein, contributes to the development of highly safe and efficacious anti-ZIKV mAb therapeutics. Our findings may be impactful to other ADE-prone viruses, including SARS-CoV-2.