Scanning immune electron microscopy using a monoclonal anti-A antibody
which reacts with all type A oligosaccharide chains revealed A antige
ns on less than 5% of A(m) and A(el) cells, some of which showed extre
mely strong labelling. This explains why A(m) and A(el) cells can abso
rb significant amounts of anti-A without being agglutinated. A(3) may
be a heterogenous subgroup, since A antigens were found on 82 and 58%,
respectively, of the cells of 2 A(3) individuals. A antigens were fou
nd on 75% or more of A(x) cells. In many weak A individuals A-positive
cells are apparently best detected if an anti-A is used which reacts
strongly with other A oligosaccharide chains than type 2. From hyperim
mune pregnancy sera A(x), A(m) and A(el) erythrocytes absorbed antibod
ies which seemed to have other fine specificities than those absorbed
by A(2) cells. We conclude that weak subgroups of A may deviate from A
(2) both by number of erythrocytes expressing A antigens and the bioch
emical nature of the antigens.