Fig 1.
Chikungunya virus glycoprotein architecture and overview of human monoclonal antibodies.
(A) The CHIKV glycoprotein spike consists of three copies each of E1 and E2 each in the prefusion form (modified from [8]). The inset shows the arrangement of E1 (comprised of DI, DII, and DIII) and E2 (comprised of domains A, B, and C). The location of E3 and the furin cleavage site also are shown. The hybrid protein “p62-E1”, consisting of p62 ectodomain linked to the E1 ectodomain by a polypeptide linker, was used for binding and sorting experiments. (B) X-ray crystallographic structure of CHIKV p62-E1 (PDB ID: 3N40) [9] with domains colored according to panel A, and with the β-connector colored magenta. (C) Reactivity of plasma from convalescent patients DC1 and DC2 toward CHIKV p62-E1 in comparison to negative control wells (3% BSA). A representative dataset is shown for the DC2 ELISA from two experiments, each performed in triplicate (points represent mean ± SD). Sera for DC1 were limited and thus data presented here are from a single experiment with no replicates. (D) Volcano plot of ELISA (OD450) for 46 of the isolated mAbs at 30 nM and 300 nM. Each data point represents the mean from 2 or more replicates. Distribution of IGHV families (E), CDR-H3 lengths (F), and IGKV families (G) for the mAbs.
Fig 2.
MAb binding profiles and epitope binning.
(A) Distribution of specificities for mAbs isolated from DC1 and DC2, based on IP, ELISA and/or BLI studies. “p62-E1” specificity refers to mAbs that were confirmed to bind p62-E1 by ELISA and/or BLI, or that immunoprecipitated p62-E1 or E2. These mAbs likely have epitopes contained in E2 or shared epitopes across E1 and E2. “E1” specificity refers to mAbs that were confirmed to bind E1' by ELISA and/or BLI. (B) Binding of E1-specific mAbs to E1' by BLI. A representative dataset from two experiments is shown. (C) Full ELISA binding curves for DC2.271B and DC2.429 against p62-E1; a representative dataset from two experiments each performed in triplicate is shown (points represent mean ± SD). (D) Two-phase binding by BLI of E1 mAbs DC1.56, DC2.82, and DC2.85 against E1 mAb chCHK-166pMAZ. In all cases, the human mAbs were able to engage p62-E1 simultaneously as CHK-166, regardless of order of addition, thus indicating that they do not share epitopes with CHK-166. A representative dataset from two experiments is shown.
Table 1.
Summary of immunoprecipitation (IP) and E1’ or p62-E1 ELISA resultsa.
Fig 3.
Neutralization of CHIKV by human mAbs.
(A) Volcano plot of 46 human mAbs for their ability to inhibit infection of CHIKV 181/25 at 30 and 300 nM. (B) IC50 values for 19 of the mAbs against CHIKV 181/25. The error bars represent 95% confidence interval from data fitting, all IC50 values were measured twice independently with similar results. (C) Neutralization of ESCA genotype LR2006_OPY1 and Asian genotype AF15561 by human mAbs. (D) Fusion assay in which CHKV 181/25 was first bound to cells, the mAb added, and the pH lowered to pH 5.5 for 2 min at 37 oC to trigger virus fusion at the plasma membrane. (E) Post-attachment assay, which was similar to (D), but following mAb addition, the virus was allowed to enter cells by endocytosis. (F) Cross-neutralization of MAYV by human CHIKV mAbs. A representative dataset from two experiments is shown; points represent mean ± SD.
Fig 4.
Viral escape studies for DC2.271B and DC2.429 using rVSV-CHIKV.
(A) Schematic for the rVSV-CHIKV genome. Infection of Vero cells by rVSV-CHIKV, infection could be tracked by eGFP expression. chCHK-152pMAZ inhibits this infection but SUDV-F4 does not. (B) Neutralization assay for rVSV-CHIKV WT and E2 viral escape mutations by DC2.271B, DC2.429, and chCHK-152pMAZ. Data are pooled from two experiments, each performed in duplicate, (points represent mean ± SD). (C) Location of rVSV-CHIKV escape mutations for DC2.429 and DC2.271B (orange and red Cα spheres, respectively) mapped onto the p62-E1 X-ray structure. E2 domains as well as E3 are colored as in Fig 1A; E1 is colored gray. Also shown are previously reported alanine scanning or viral escape mutations that ablate binding for 4N12 (parent of SVIR001, yellow Cα sphere), 4J21 (green Cα spheres), C9 (magenta Cα sphere), and IMCV-063 (blue Cα sphere). The structural epitope for murine mAb CHK-152, as mapped by cryo-EM is shown as cyan Cα spheres. The alanine scanning mutations that reduced Mxra8 binding are shown as magenta side chains. (D) Comparison of neutralizing potency for DC2.271B, DC2.429, C9pMAZ, IM-CKV063pMAZ, and 4N12pMAZ. (E) Neutralization of DC2.271B viral escape mutant rVSV-CHIKVE2-K233T by C9pMAZ, IM-CKV063pMAZ, and 4N12pMAZ. For panels B and C, data are pooled from two experiments, each performed in duplicate (points represent mean ± SD). Panel D, two independent experiments were performed in triplicate with similar results; a representative dataset is shown.
Fig 5.
Viral escape studies with E1-targeting mAbs.
Location of A286 and R289 on the p62-E1 X-ray structure (A, PDB ID: 3N40) [9] or on the E1/E2 cryoEM heterohexamer (B, PDB ID: 3J2W) [11]. For clarity, p62 or E2 subunits are colored gray whereas E1 domains DI, DII, and DIII colored as per Fig 1. In panel B, a complete prefusion E1/E2 hexameric spike (outlined with dotted line) is illustrated, along with an E1/E2 heterodimer from an adjacent spike, to depict relative orientation within adjacent spikes. (C) Neutralization studies with WT VSV-CHIKV and viral escape mutants. Data are pooled from two experiments, each performed in duplicate or triplicate (points represent mean ± SD).
Fig 6.
In vivo properties of human mAbs in mice.
(A) Protective efficacy using a prophylactic dosing regime (day -1) against CHIKV LR2006_OPY1 in 3-week old mice rendered immunodeficient with anti-Ifnar1 mAb. Survival curves were compared using the log-rank test with a Bonferroni correction. Results were combined from two independent experiments of five mice per treatment group (n = 10). (B) Serum mAb levels 48 hours after mAb administration in infected mice from (A) (“I”, closed symbols) and 60 hours after mAb administration in uninfected mice (“U”, open symbols). There were three mice per group points represent mean ± SD). Serum mAb levels were compared by unpaired t-test. (C) Therapeutic dosing regimen, similar to (A) but CHIKV mAbs were administered at day +1. Analysis as in (A).
Fig 7.
Recognition requirements for DC2.271B.
CryoEM visualization (A) and single particle three-dimensional reconstruction (B) of p62-E1 alone and in complex with DC2.271B Fab. (C) Neutralization of CHIKV 181/25 by DC2.271B IgG1 and Fab. Data are pooled from two experiments each performed in triplicate (points represent mean ± SD).