Senescent cell antigen (SCANT) is a "neo antigen" that appears on the surface of normal old cells and initiates IgG binding and cellular removal. To investigate the mechanism by which SCANT is generated from its parent molecule, band 3, we subjected intact human erythrocytes to treatments that have been reported to result in changes in band 3 and/or to mimick aging in vitro. The validity of these treatments as model systems for erythrocyte aging was evaluated using a "red cell aging panel" that provides a biochemical profile of a senescent red cell. Treatments were assessed for their ability to induce in vitro the following changes observed in normal erythrocytes aged in vivo: 1 increased breakdown of band 3 as detected by immunoblotting, 2 decrease in anion transport efficiency as detected with a sulfate self-exchange assay, 3 decrease in total glyceraldehyde 3-phosphate dehydrogenase activity with an increase in membrane-bound activity, and 4 increase in the binding of autologous IgG as detected with a protein A binding assay. Neither incubation with the free radical-generating xanthine oxidase/xanthine system, nor treatment with malondialdehyde, and end product of free radical-initiated lipid (per)oxidation, results in age-specific changes. Loading of the cells with calcium and oxidation with iodate results in increased breakdown of band 3, but does not lead to increased binding of autologous IgG. Only erythrocytes that have been stored for 3-4 weeks show the same structural and functional changes as observed during aging in vivo.