Manipulation of immunodominant variable epitopes of norovirus capsid protein elicited cross-blocking antibodies to different GII.4 variants despite the low potency of the polyclonal sera

Figures

• 

  Fig 1. Development of 12∆AG VLPs. (A) Sequence of antigenic sites A and G of wild-type Sydney 2012 variant strain, WT2012 (SY), and 12∆AG where the residues on both antigenic sites were mutated to alanine. The colors on the cell indicate the chemical properties of the amino acid. (B) The three-dimensional structure of the VP1 protein of GII.4 norovirus (PDB ID: 7K6V). Antigenic sites that are involved in viral neutralization and HBGA blockade, including those found mutated in this study (sites A and G), were indicated. (C) The pictures indicate SDS-PAGE gel (top left), Western blot (top right), and the electron microscopy images (bottom) of purified WT2012 (SY) and 12∆AG VLPs. Norovirus VP1-specific cross-reactive mAb, 30A11 (22), was used to confirm the expression of the VP1 protein in the Western blot analysis. (D) ELISA data to confirm the binding (or not binding) of previously characterized WT2012 (SY) mAbs (18) to 12∆AG VLPs. The heatmap shows OD405nm values in ELISA, using 10 µg/mL of purified mAbs and 0.5 µg/mL of VLPs. (E) MAbs that bind to 12∆AG VLPs were further tested for binding by 10-fold serial dilution of mAbs. The bar graph shows the difference of EC50 titers between…

• 

  Fig 2. Characterization of mAbs produced from mice immunized with 12∆AG VLPs. (A) Ten mAbs were newly developed after immunization of 12∆AG VLPs to mice. They were selected based on their binding patterns to P and S domains of WT2012 (SY) and homologous and heterologous reactivities to the VLPs of WT2012 (SY) and WT2004 (FH) viruses. The heatmap shows OD405nm values in ELISA, using 10 µg/mL of purified mAbs and 0.5 µg/mL of proteins or VLPs. (B) Three WT2012 (SY)-specific mAbs were characterized using antigenic sites C- or D-swapped mutants, in which residues on individual antigenic sites were swapped between WT2012 (SY) and WT2004 (FH) viruses (18). Sequences of antigenic sites C and D of WT2012 (SY) and WT2004 (FH) are shown on the top panel. The bar graph on the bottom indicates ELISA OD405nm values, using 10 µg/mL of purified mAbs and 0.5 µg/mL of VLPs. The dashed line indicates the cutoff value calculated using OD405nm values taken from negative control wells (mean + 2 × standard deviation). (C) Breadth of the cross-reactivity from seven mAbs was confirmed using VLPs from viruses detected from 1974 to 2019. The heatmap shows OD405nm values in ELISA, using 10 µg/mL of mAbs and 0.5…

• 

  Fig 3. Blockade potential of cross-reactive mAbs produced from mice immunized with 12∆AG VLPs. (A) The homologous and heterologous HBGA-blockade antibodies were measured using blockade assays using PGM III as a source of carbohydrates. The line graphs indicate best-fit linear regression curves and normalized OD405nm values in duplicate wells at twofold dilution of mAbs. (B) The heatmap summarizes the blockade EC50 titers of four cross-reactive mAbs against historical GII.4 variants that emerged between 1974 and 2019. (C) The three-dimensional structure of the VP1 dimers of GII.4 norovirus (PDB ID: 7K6V) indicates the position of mutations that were detected only in the Hong Kong 2019 variant in red. Position of antigenic sites is indicated by the same color palette used in Fig. 1A.

• 

  Fig 4. HBGA-blockade assays of sera from mice immunized with wild-type WT2012 (SY) or the mutant 12∆AG VLPs. (A) ELISA to measure IgG titers in WT2012 (SY)- and 12∆AG-immunized mouse sera against immunogens—WT2012 (SY) and 12∆AG mutant VLPs. The IgG titers were summarized in a bar graph using EC50 values. *P < 0.05; **P < 0.01 in One-way ANOVA and Šídák multiple comparison test. (B) The HBGA-blockade antibodies in the wild-type VLP-immunized (left) and 12∆AG VLP-immunized mouse sera (right) were measured against homologous WT2012 (SY) VLPs and heterologous VLPs including WT1996 (GR), WT2004 (FH), WT2006 (DH), WT2018 (SF), and WT2019 (HK), using blockade assays using PGM III as a source of carbohydrates. The top line graphs show normalized OD405nm values from four mice in duplicate wells at twofold dilution of serum. The bottom bar graph summarizes the EC50 titers of individual sera against different GII.4 variants in the blockade assays. The error bars represent the 95% confidence intervals of estimates in nonlinear regression analyses. The dot indicates EC50 titers of four mice. The horizontal dashed line indicates the limit of detection in the blockade assays (EC50 = 50). *P < 0.05;…

• 

  Fig 5. Follow-up analyses of antibodies and 12∆AG VLPs. (A) ELISA using the supernatant from 34 parental hybridoma clones obtained after the initial fusion step from mice immunized with 12∆AG VLPs. The heatmap shows OD405nm values in ELISA, using 1:2 dilution of the supernatant from hybridoma clones and 0.5 µg/mL of proteins or VLPs. (B) The line graph presents site-by-site B-cell epitope prediction scores generated using DiscoTope-3.0- and AlphaFold-predicted three-dimensional structures of wild-type WT2012 (SY) (red solid line) and 12∆AG mutant VP1 (black dashed lines). The horizontal dashed lines indicate suggested thresholds of high (= 1.5) or moderate (= 0.9) confidence of epitope prediction (33).