Peptides designed to spatially depict the Epstein-Barr virus major virion glycoprotein gp350 neutralization epitope elicit antibodies that block virus-neutralizing antibody 72A1 interaction with the native gp350 molecule

Abstract

Epstein-Barr virus (EBV) is the etiologic agent of infectious mononucleosis and the root cause of B-cell lymphoproliferative disease in individuals with a weakened immune system, as well as a principal cofactor in nasopharyngeal carcinoma, various lymphomas, and other cancers. The EBV major virion surface glycoprotein gp350 is viewed as the best vaccine candidate to prevent infectious mononucleosis in healthy EBV-naive persons and EBV-related cancers in at-risk individuals. Previous epitope mapping of gp350 revealed only one dominant neutralizing epitope, which has been shown to be the target of the monoclonal antibody 72A1. Computer modeling of the 72A1 antibody interaction with the gp350 amino terminus was used to identify gp350 amino acids that could form strong ionic, electrostatic, or hydrogen bonds with the 72A1 antibody. Peptide DDRTTLQLAQNPVYIPETYPYIKWDN (designated peptide 2) and peptide GSAKPGNGSYFASVKTEMLGNEID (designated peptide 3) were designed to spatially represent the gp350 amino acids predicted to interact with the 72A1 antibody paratope. Peptide 2 bound to the 72A1 antibody and blocked 72A1 antibody recognition of the native gp350 molecule. Peptide 2 and peptide 3 were recognized by human IgG and shown to elicit murine antibodies that could target gp350 and block its recognition by the 72A1 antibody. This work provides a structural mapping of the interaction between the EBV-neutralizing antibody 72A1 and the major virion surface protein gp350. gp350 mimetic peptides that spatially depict the EBV-neutralizing epitope would be useful as a vaccine to focus the immune system exclusively to this important virus epitope.

Importance: The production of virus-neutralizing antibodies targeting the Epstein-Barr virus (EBV) major surface glycoprotein gp350 is important for the prevention of infectious mononucleosis and EBV-related cancers. The data presented here provide the first in silico map of the gp350 interaction with a virus-blocking monoclonal antibody. Immunization with gp350 peptides identified by in silico mapping generated antibodies that cross-react with the EBV gp350 molecule and block recognition of the gp350 molecule by a virus-neutralizing antibody. Through its ability to focus the immune system exclusively on the gp350 sequence important for viral entry, these peptides may form the basis of an EBV vaccine candidate. This strategy would sidestep the production of other irrelevant gp350 antibodies that divert the immune system from generating a protective antiviral response or that impede access to the virus-blocking epitope by protective antibodies.

FIG 1

72A1 heavy- and light-chain variable-region sequences. (Top) Anti-gp350 chimeric antibody amino acid sequence with predicted CDRs (bold) and heavy-chain (double underlined) and light-chain (underlined) amino acids that interact strongly with gp350. Amino-terminal peptide sequences were determined by Edman sequencing (gray). The 72A1 variable-region and human Ig constant-region junctions are identified (|). (Bottom) Depiction of 72A1 variable-region light chain (blue) and heavy chain (green). Predicted CDR (space filled) and framework amino acids in the light chain or heavy chain predicted to H bond with gp350 are shown in yellow and purple, respectively (Table 1). Heavy-chain amino acid His₁₀₄ predicted to form a salt bridge with gp350 Glu₂₁₁ is shown in red.

FIG 2

Chimeric anti-gp350 antibody recognizes native gp350 and blocks EBV infection of B cells. (A) Chimeric anti-gp350 antibody recognizes gp350 expressed on the surface of CEM cells (CEMgp350), as measured by immunofluorescence microscopy (left) or by FACS analysis (right). (B) Log₁₀ doses of chimeric anti-gp350 or of the murine 72A1 monoclonal antibody that block P3HR1 virus superinfection of Raji cells. SSC, side scatter; FSC, forward scatter; mAb, monoclonal antibody; chAb, chimeric antibody; Ab, antibody.

FIG 3

Depiction of the amino acids on the surface of gp350 predicted to couple with the 72A1 variable region. (Top) Interfacing amino acids (bold) predicted to H bond or form a salt bridge with the 72A1 heavy-chain (double underlined) or light-chain (underlined) variable region in domain 1 (green), domain 2 (cyan), domain 3 (red), and domain 4 (orange) (Table 1). The peptide 1 sequence, which was used in the assays whose results are presented in Fig. 4 to 6 and which lies outside the gp350-neutralizing domain, is shown in blue. (Bottom) Front view (left) and side view (right) of the gp350 amino terminus (tan ribbon) with glycosylation (ball-and-stick representation) showing domain 1 to 4 peptides (space filled) predicted to interact with the 72A1 antibody.

FIG 4

The gp350 mimetic peptide binds to 72A1 antibody to block gp350 recognition. (A) Optical density at 450 nm (OD_450nm) ± SE for 72A1 (closed bars) or anti-LMP1 S-12 (open bars) binding to peptides 1 to 3 (Table 2). Plots were derived from duplicate samples and from three separate experiments. (B and C) Average percent inhibition ± SEM for 72A1 antibody recognition of gp350 in an ELISA-based competition assay by peptides 1 to 3. Plots were derived from duplicate samples and from three separate experiments. P values of ≤0.01 for the differences in the results between peptide 2-positive serum and peptide 1-positive serum are indicated (*). D1 and D2, domains 1 and 2, respectively.

FIG 5

gp350 peptides generate anti-gp350 antibodies that block 72A1 recognition of gp350. (A) Box plot of the antipeptide antibody concentrations found in peptide-immunized mice (n = 4). (B) Box plot of the anti-gp350 concentrations found in peptide-immunized mice (n = 4). (C) Histogram of biotinylated 72A1 binding to a gp350 protein that was preexposed to mouse preimmune (Preimm) serum, pooled serum from peptide-immunized mice (1:50 dilution), 1 μg/ml 72A1, or 0.5 mg/ml BSA. Preimmune serum recognition of peptide or gp350 is indicated (○). The average optical density at 450 nm (OD_450nm) ± SEM was derived from three separate experiments. P values of ≤0.05 for the differences in the results between peptide 2- and 3-positive sera and peptide 1-positive serum are indicated (*).

FIG 6

Human immune globulin recognizes gp350 mimetic peptides. Antipeptide antibody concentrations are depicted as a box-and-whiskers plot, with sample values outside the 1st percentile and 99th percentile being identified (●). Plot values were calculated for duplicate samples from three separate experiments using 39 different lots of intravenous immunoglobulin.