Human erythrocyte antigen expression: its molecular bases.

This review summarises the considerable body of information now available on the molecular bases of human blood group-antigen expression. The elucidation of this information has only been possible since the identification, purification, and subsequent cDNA cloning of the mRNAs which encode the blood group active proteins. The surface components which are responsible for antigen expression are divided into two types: carbohydrate and protein. All carbohydrate structures are attached covalently to either glycolipids or glycoproteins, and are synthesised in the Golgi apparatus of erythropoietic cells (or in other cell lines in secreted fluids). As a consequence, the molecular bases of carbohydrate antigens lie in polymorphic variation seen in the genes which synthesise these carbohydrate structures (i.e. glycosyltransferase enzymes). The structural differences in the ABO, Hh and Lewis transferase genes, which alter the substrate specificities of these glycosyltransferases and hence generate the different antigens, have been defined. The molecular bases underlying the P blood glycosyltransferases are unknown. Polymorphism in the remaining 19 blood group systems is defined by amino acid sequence changes in erythrocyte membrane proteins, which are generated by sequence variation at the DNA level (largely by point mutation). Blood group active erythrocyte membrane proteins can be categorised broadly into six functional groups: (1) membrane transporters or channels: Rh, Diego, Colton, Kidd, KX; (2) Membrane bound enzymes: Kell and Cartwright; (3) Structural or assembly proteins: Gerbich and MNSs; (4) Chemokine receptors: Duffy; (5) Cell adhesion molecules: Lutheran, LW, Xg, Indian; (6) Complement regulatory proteins: Cromer, Knops. The Chido/Rodgers blood group system is defined by polymorphic variation in C4 of the complement cascade, and is adsorbed passively onto the surface of erythrocytes. This system is not considered here. Only two remaining blood group systems defy molecular identity: Dombrock and Scianna. The potential clinical applications of such a rapid accumulation of data on the molecular bases of blood group antigens is discussed.