Mapping of immunodominant B-cell epitopes and the human serum albumin-binding site in natural hepatitis B virus surface antigen of defined genosubtype

Institute of Medical Virology 1 , and Institute of Biochemistry 4 , Clinics of the Justus-Liebig-University, Frankfurter Str. 107, D-35392 Giessen, Germany Institute of Biomedical Research, Riga, Latvia 2 Institute of Virology, Humboldt University, D-10098 Berlin, Germany 3 Institute of Medical Microbiology, Georg-August University, D-37075 Göttingen, Germany 4 Author for correspondence: Wolfram H. Gerlich. Fax +49 641 99 41209.e-mail Wolfram.H.Gerlich{at}viro.med.uni-giessen.de Twelve MAbs were generated by immunization of BALB/c mice with plasma-derived hepatitis B virus surface spherical antigen particles subtype ayw2 (HBsAg/ayw2 genotype D). Their epitopes were mapped by analysis of reactivity with plasma-derived HBsAg/ayw2 and HBsAg/adw2 (genotype A) in enzyme immunoassays and blots. Mapping was supported by nested sets of truncated preS2 proteins and preS2 peptides. Five antibodies were S domain-specific, seven were preS2-specific and 11 had a preference for genotype D. According to our data, group I of the three known epitope groups of preS2 has to be divided into IA and IB. Three preS2-specific MAbs forming the new group IA reacted with genotype D residues 3–15 which have not yet been described as an epitope region. IA antibodies strongly inhibited the binding of polymerized human serum albumin. Two antibodies (group II) reacted with the glycosylated N-terminal region of preS2 in plasma-derived HBsAg, but not with a preparation from transfected murine cells. One group III antibody was subtype-specific and reacted with the highly variable preS2 sequence 38–48. Only one antibody (group IB) mapped to the region (old group I) which was believed to be immunodominant and genotype-independent. Geno(sub)type-specific epitopes of preS2 are obviously the immunodominant components of natural HBsAg in BALB/c mice, but these epitopes may be masked by serum albumins in humans. The data may explain why it is difficult to detect anti-preS2 antibodies in human recipients of preS2-containing vaccines, in spite of the preS2 immunodominance in mice.