A large number of studies have suggested a direct involvement of the preS1 domain of the hepatitis B virus large envelope protein (L-HBsAg) (in particular amino acids 21 to 47) in a virus attachment to hepatocytes [[4, 6, 7], reviewed in ]. Recently, it was demonstrated that a mutant L-HBsAg bearing a deletion in the 26–30 amino acid sequence of the preS1 receptor binding site was non-infectious and hence deficient in viral entry , further supporting the importance of this domain as an infectivity determinant in hepatitis B. The aim in this study was to identify preS1 antibodies reacting with two human antibody binding sites, p (21–32), containing an infectivity determinant and p (32–47), including a previously defined HBV-neutralisation epitope comprising amino acids 37–45 of preS1 that have been identified within this sequence  as well as preS1 antibodies reacting with the C-terminal part (94–117) in sera from newborns completely immunised with the third-generation recombinant vaccine (Sci-B-Vac™).
We found that immunised newborns exhibited anti-preS1 (21–32) and anti-preS1 (32–47) and anti-preS1 (94–117) antibody reactivity in 50% (14/28), 54% (15/28) and 43% (12/28), respectively. The prevalence of vaccine-induced anti-peptide antibodies in this study is identical to the prevalence of antibodies reacting with p (21–32), p (32–47) and p (94–117) in sera obtained during convalescence from natural HBV infection . The third-generation vaccine used in this study, Sci-B-Vac™, was developed using mammalian cells, which provide envelope proteins similar to those isolated from infected patients and used for preparation of some of the first-generation HBV vaccines . This similarity could explain the identical immunogenicity and specificity induced by natural infection or complete vaccination with a preS1-, preS2- and S-containing hepatitis B vaccine.
Five of 28 (18%) newborns lacked detectable levels of anti-preS1 (21–32), anti-preS1 (32–47) or anti-preS1 (94–117) antibodies. One explanation for this observation could be that the preS1 (21–32; 32–47 and 94–117) non-reactive patients have conformation dependent anti-preS1 antibodies in the circulation that are not reactive with the linear peptide analogues of preS1 (21–32, 32–47 and 94–117) used in this study. In most cases synthetic peptides are considered unsuitable for studying the discontinuously conformational epitopes because most peptide-specific antibodies are continuous sequence-specific . The three-dimensional structure of the HBV surface protein, including preS1, has not yet been determined, though Lian et al  recently demonstrated that amino acids 31–36 were pivotal for the folding and conformational stability of the preS protein.
We demonstrated that those newborns that developed antibody reactivity against two or three B-cell epitopes had a significantly stronger anti-HBs response than those individuals that showed a narrower preS1 response. In congenic mice inclusion of preS regions elicits a broader spectrum of protective antibodies that augment the anti-HBs response and circumvent non-responsiveness to the S protein .
The preS1 region in humans has been shown to be a particularly efficient immunogen at the T-cell level. A dominant T-cell recognition site was identified in the N-terminal residues 21–28 (serotype adw) of the preS1 sequence [29, 30] and in vitro binding studies have defined preS1 (25–33) as a promising helper T-cell epitope . Further studies will show whether preS1 T-cell epitopes contained within the third-generation hepatitis B vaccine can mediate helper T-cell functions on the human antibody response to the particulate HBsAg similarly to what has been shown in the mouse system.