The Influenza Virus Receptor and Blood Group Antigens of Human Erythrocyte Stroma

Abstract

Summary The virus receptor and blood group antigens of human erythrocytes have been investigated through the quantitative fractionation of the post-hemolytic residue or stroma, before and after dry ether extraction. Comminution of whole erythrocyte ghosts by various means involving chiefly the manipulation of protein concentration, pH and ionic strength, showed that the major portion of the stroma protein remained bound with lipid and carbohydrate in larger fragments readily sedimented at alkaline pH by centrifugation at 25,000 G. This was designated as lipid rich (LR) protein. A smaller proportion of the stroma protein could be separated either as a lipid poor (LP) fraction (no cholesterol, and a phosphorus content 10–20% that of the LR fraction) or a lipid free (LF) fraction (no cholesterol or phosphorus). The LP fraction contained less carbohydrate and receptor activity than the LR fraction, and was devoid of Rh antigen. The LF fraction, with only a trace of carbohydrate, showed no receptor activity of any kind. The LP fraction probably represented incompletely separated LR and LF portions of the stroma complex. Neither LP nor LF fractions were homogeneous, both showing more than three components electrophoretically. By the removal of certain ether-soluble lipids from lyophilized stroma, the relatively large protein-lipid-carbohydrate complex of the LR fraction was broken apart into smaller particles, not of uniform size, which were of essentially similar gross chemical composition with respect to protein, carbohydrate, and phospholipid. The enhancement of potency in the inhibition of virus and specific blood group agglutination resulting from dry ether extraction may have been a reflection of this reduction in particle size. The ether-soluble lipids were devoid of detectable virus receptor or blood group antigens. Some of the solubility characteristics of stroma before and after ether extraction were discussed. The number of proteins present in the stroma other than hemoglobin could not be established from the data. The potency of these stroma fractions as inhibitors of virus hemagglutination bore no relation to the type of cell from which they had been derived. The virus inhibitor was heat stable, and in the presence of certain of the structural lipids in crude stroma (i.e. in LR fractions) was resistant to proteolytic enzymes. In preparations freed of cholesterol, i.e. either in ether-extracted stroma, or in the LP fraction of unextracted stroma, it remained heat stable, but became sensitive to the action of chymotrypsin. It thus appeared that part, if not all, of the protein to which carbohydrate and lipid were attached was necessary for interaction with the virus. The A antigen in ether-extracted stroma and in unextracted stroma was heat stable, and was insensitive to both chymotrypsin and active influenza virus. The Rh antigen, in those fractions where it was detectable, was heat labile, and sensitive to the action of proteolytic enzymes. The virus receptor in soluble stroma fractions was inactivated by native influenza virus which was capable of eluting almost completely from fowl cells; but not by influenza virus which did not elute from fowl cells.