Molecular basis of hepatitis B virus serotype variations within the four major subtypes.

Amino acid residues 101 to 180 of hepatitis B surface antigen (HBsAg) were predicted by sequencing the corresponding part of the S gene of hepatitis B virus (HBV) DNA in 46 HBsAg-positive sera, which had been subtyped by immunodiffusion with respect to d/y, w/r, w1 to w4 and q. The sequences of the nine different HBV serotypes defined by these specificities were found to be homogeneous proving that they represent consistent variations of HBV at the genomic level. Residue 127 was found to be important as were Pro, Thr and Leu for w1/w2, w3 and w4, respectively. Five residues were found to differ between ayw1 and ayw2. These were at positions 134 (Phe instead of Tyr), 143 (Thr instead of Ser), 159 (Ala instead of Gly), 161 (Tyr instead of Phe) and 168 (Val instead of Ala). However, all these residues were shared by ayw1 and adw2, implying that Arg122 was also important for w1 expression. All genomes expressing r, apart from one ayr strain, had an Ile126, which might explain the pseudo-allelism of w1 to w4 in relation to r, since this substitution might influence the w epitope. There were two regions where adw4q- and adrq- differed from all the q+ subtypes. These were located at residues 158 and 159, and at residues 177 and 178, where both the q- subtypes had amino acid substitutions in adjacent positions. The mapping of the epitopes defining these antigenic specificities will help to link information on the world-wide distribution of HBsAg subtypes to future molecular epidemiology with regard to HBV.