Human Norovirus Epitope D Plasticity Allows Escape from Antibody Immunity without Loss of Capacity for Binding Cellular Ligands

Figures

• 

  FIG 1. Human norovirus GII.4 phylogenetic tree. A radial phylogram for norovirus GII isolates, including strains from the donor here (Hu_NoV_PT1-9, -11, and -12), was created using the neighbor-joining method with 100 replicates based on a multiple sequence alignment of amino acids. Numbers on the tree correspond to consensus support values (%). The tree is rooted to a human GII.5 norovirus, which is 63% identical in sequence to the other viruses in the tree. All other viruses are ∼93% identical in sequence to each other. Consensus sequences GII.4 MC4 and MC12 were designed for this study.

• 

  FIG 2. GII.4 MC4 and MC12 capsid protein. (A) Capsid protein amino acid sequences of GII.4 MC4 and MC12 compared to GII.4 2009 New Orleans and 2012 Sydney. Residues known to interact with carbohydrate ligand (*) and antibodies that block VLP-ligand interaction (epitopes A [blue], D [orange], E [purple], and F [green]) are indicated. Residues that vary between the compared sequences are in boldface. (B) Side view of GII.4 MC12 P domain dimer homology model based on GII.4 2012 with one chain color-coded by epitopes as in panel A. The carbohydrate binding domain is not visible in this orientation. Residues that vary in MC12 compared to the consensus sequence of GII.4 2009 and 2012 are colored red. (C) Electron micrograph of VLP created for this study.

• 

  FIG 3. GII.4 MC4 and MC12 are antigenically distinct. GII.4 MC4 and MC12 VLP were screened for reactivity to MAbs to epitopes A (green shading), D (gray shading), and G (blue shading), and half-maximum binding titers and 95% CI values were calculated. The MAb reactivities to GII.4 2009 and GII.4 2012 are reported both here and elsewhere (16, 19, 24, 25, 27).

• 

  FIG 4. Carbohydrate binding profiles of GII.4 MC4 and MC12. GII.4 MC4 and MC12 bind pig gastric mucin III and human type B saliva. The dashed line is the limit of detection. Markers denote the mean, and error bars indicate the standard errors of the mean (SEM) based on eight replicates from four independent experiments. *, PGM binding significantly different from B saliva binding.

• 

  FIG 5. The number and identity of epitope D residues impacts VLP interaction with blockade antibodies by multiple mechanisms. (A) GII.4 MC12, MC12.AAA, and MC12.STT binding to MAbs NVB 97, NO66, and NO224 was analyzed by single-site binding curve fitting, and maximum binding (Bmax) and dissociation constants (Kd) were calculated to differentiate the effects of sequence changes on antibody-epitope access and binding strength, respectively. Boldface numbers are significantly different from GII.4 MC12.A-A. (B to D) Antibody function, as measured by blockade of ligand binding potency, the IC50 titer, was determined by dose-response nonlinear curve fit analysis of the mean percent control binding with 95% CI values (error bars).

• 

  FIG 6. Blockade antibodies anchor at residues 393 and 396 of the protruding loop of epitope D. (A) GII.4 MC12 P domain dimer model with epitope D (cyan) circled and other features color coded as in Fig. 2B. (B) Epitope D enlarged with residues 393 and 396 identified. (C) Half-maximal binding titers with 95% confidence intervals (95% CI) were determined for MAbs NVB 97, NO66, and NO224 to VLPs with substitutions reflective of natural variation between pandemic strains GII.4 2009 and 2012 within epitope D at residues 393 and 396. Boldface numbers are significantly different from GII.4 2009.

• 

  FIG 7. Residue number and sequence within epitope D modulates ligand binding. GII.4 MC12.A-A, MC12.AAA, and MC12.STT binding to synthetic B (A), A (B), H type III (C), and Lewis Y (D) was analyzed by single-site binding curve fitting, and the Bmax and Kd values were determined (E). Markers denote the mean, and error bars indicate the SEM (minimum of four replicates from two independent experiments). Orange shading indicates significantly different from GII.4 MC12. Boldface values denote significantly different from GII.4.MC12.STT. ND, not determined.