Deep mutational scanning comprehensively maps how Zika envelope protein mutations affect viral growth and antibody escape

Abstract

Functional constraints on viral proteins are often assessed by examining sequence conservation among natural strains, but this approach is relatively ineffective for Zika virus because all known sequences are highly similar. Here we take an alternative approach to map functional constraints on Zika virus’s envelope (E) protein by using deep mutational scanning to measure how all amino-acid mutations to the protein affect viral growth in cell culture. The resulting sequence-function map is consistent with existing knowledge about E protein structure and function, but also provides insight into mutation-level constraints in many regions of the protein that have not been well characterized in prior functional work. In addition, we extend our approach to completely map how mutations affect viral neutralization by two monoclonal antibodies, thereby precisely defining their functional epitopes. Overall, our study provides a valuable resource for understanding the effects of mutations to this important viral protein, and also offers a roadmap for future work to map functional and antigenic selection to Zika virus at high resolution. Importance Zika virus has recently been shown to be associated with severe birth defects. The virus’s E protein mediates its ability to infect cells, and is also the primary target of the antibodies that are elicited by natural infection and vaccines that are being developed against the virus. Therefore, determining the effects of mutations to this protein is important for understanding its function, its susceptibility to vaccine-mediated immunity, and its potential for future evolution. We completely mapped how amino-acid mutations to E protein affected the virus’s ability to grow in cells in the lab and escape from several antibodies. The resulting maps relate changes in the E protein’s sequence to changes in viral function, and therefore provide a valuable complement to existing maps of the physical structure of the protein.

Identification and characterization of key residues in Zika virus envelope protein for virus assembly and entry

“…Notably, Mut 4, Mut 9, Mut 10 and Mut 37 nearly completely abolished the infectivity of the ZIKV TCPs HA , albeit substantial E was released as evidenced by western blotting analysis ( Figure 2 (e–f) and Supplementary Figure 2(a, e)), suggesting that ZIKV TCPs bearing these mutants fail to enter into cells (see below). The observation that most E mutants impaired ZIKV TCPs egress is consistent with the previously proposed, participation of E in virion assembly [ 41 ].…”

“…Previous study has identified some E mutations abolished virus rescue in the reverse genetic systems by using deep mutational scanning, which is important for understanding its function, its susceptibility to immunity and its future evolution [ 41 ]. In our research, we found that except for the loop regions in DI, most mutants impaired the egress of the ZIKV TCPs ( Figure 2 (d–f) and Supplementary Figures 2 and 5).…”

Integrated protein-protein interaction and RNA interference screens reveal novel restriction and dependency factors for a tick-borne flavivirus in its human host

“…The Hypr strain of TBEV (GenBank accession number U39292.1), isolated in the Czech Republic from a child with TBE 53 , was propagated and titrated in Vero cells grown in DMEM supplemented with antibiotics as described above and 2% FCS. Viral stocks were prepared as described in Lemasson et al 20 and quantified by endpoint dilution assay on Vero cells, as visualized by the virus-induced cytopathic effect (CPE), as previously described 54 . Briefly, 1.5 × 10 4 Vero cells were seeded in 100 μL of DMEM supplemented with antibiotics and 2% FCS in wells of a 96-well plate the day prior to infection with 25 μL of serial dilutions of virus in medium as above (6 wells per dilution).…”

B-Cell Epitope Discovery: The First Protein Flexibility-Based Algorithm – Zika Virus Conserved Epitope Demonstration

“…[5] Examination of epitope discovery performance against the top two most conserved flavivirus epitopes shows isolated protein flexibility also maintained almost perfect capacity to 1) pinpoint residues from the top two conserved epitopes (a PRAUC mean of 0.85 ± 0.30) and 2) rank residues from the top two conserved epitopes above all other residues (a ROCAUC mean of 1.00 ± 0.01). Closer inspection reveals 1) isolated protein flexibility of seven structurally related flaviviruses retained comparable capacity to pinpoint residues from the top seven conserved epitopes (a mean PRAUC over flavivirus structures of 0.41 ± 0.04; S2 Table ), 2) isolated protein flexibility from a ZIKV protein simulation without disulfides performed with a lower capacity to pinpoint residues from the top seven conserved epitopes (a lower PRAUC mean, 0.35 ± 0.22) than is isolated protein flexibility with disulfides, supporting separate studies suggesting disulfides play a structural role in flavivirus immunogenicity, [37][38][39][40][41] 3) IUPred, a sequencebased measure of intrinsically disordered protein regions (which one might associate with protein flexibility), [42,43] performed no better than chance against the top seven conserved epitopes (baseline ROCAUC and PRAUC means), and 4) ElliPro had overlapping performance with isolated protein flexibility in terms of capacity to pinpoint residues from the top seven conserved epitopes (despite a lower PRAUC mean).…”