Antigenicity of influenza virus hemagglutinin following chemical modification

Abstract

Hemagglutinin (HA) molecules from a number of different strains of type A influenza virus were reacted with 1-fluoro 2–4-dinitrobenzene (FDNB) at pH values ranging from 8.4 to 10.3. HA was also reacted with tetranitromethane (TNM) or diazotized sulfanilic acid (DSA). Sequence studies on HA1 from HA molecules treated with FDNB, TNM, or DSA showed that certain lysine, tyrosine, or histidine residues were 100% substituted after the reaction, while others apparently did not react at all. DNP-substituted hemagglutinin molecules, isolated from FDNB-treated A/Memphis/1/71 H-BEL N(H3N1) virus particles, had up to 58% of lysines substituted with DNP. These molecules, nevertheless, retained hemagglutinin activity and, as far as could be measured, the same capacity as the unsubstituted hemagglutinin to react with heterogeneous antiserum or a panel of monoclonal antibodies. These results suggest that those amino acid side chains able to react with FDNB (lysine, histidine, tyrosine, and cysteine) are either not present in the antigenic sites on the HA, or if they are, then either the side chains which bind antibody do not react with DNP, or the presence of DNP in the site does not affect its ability to combine with antibody. The results also suggest that substitution of more than half of the lysine in the hemagglutinin molecule does not cause any marked conformational changes, for such changes would be expected to affect the ability of the HA to combine with both cell receptors and antibody molecules. Similar findings with TNM-treated HA suggest that tyrosine is not an essential part of any antigenic site on H3 type HA. HA treated with diazotized sulfanilic acid lost HA activity, but its antigenicity was similar to that of untreated HA when tested with heterogeneous antisera, suggesting that histidine was not present in the antigenic sites. However, when HA from a monoclonal variant of A/Mem/1/71 (H3N2) virus with a sequence change from wild-type in HA1 of proline (143) to histidine ( Laver, Air, and Webster, 1981) was reacted with diazotized sulfanilic acid, the histidine at position 143 in HA1 reacted completely and the HA lost the ability to bind antibody specific for the new antigenic site on this variant. However, treatment of this variant with FDNB did not lead to substitution of histidine 143.