Abstract
Glycolytic enzymes have been recently shown to exist as multi-enzyme complexes in association with the cytoplasmic domain of band 3 at the inner surface of the human erythrocyte membrane. Because several of the glycolytic enzyme binding sites have been mapped to sequences near the NH2-terminus of band 3 (DDYED and EEYED) that are not conserved in mice (EEVLE and EELEN), the question naturally arose whether the existence of glycolytic enzyme complexes on erythrocyte membranes might be only a product of recent evolution. To test this hypothesis, fresh murine erythrocytes were fixed and stained with antibodies to glyceraldehyde-3-phosphate dehydrogenase (GAPDH), aldolase, phosphofructokinase (PFK), pyruvate kinase (PK), lactate dehydrogenase (LDH) and carbonic anhydrase II (CA II was used as a control, since it binds to a distant site near the COOH-terminus of band 3). Importantly, analysis of intact murine erythrocytes by confocal microscopy demonstrated that all of the above enzymes are localized to the membrane in oxygenated cells. In contrast, upon deoxygenation of the intact cells, release of the glycolytic enzymes (but not CA II) from the erythrocyte membrane and their uniform redistribution throughout the cytoplasm is observed. Because deoxyhemoglobin has been shown in human erythrocytes to compete with glycolytic enzymes (but not with CA II) for a common binding site at the NH2-terminus of band 3, these data argue that murine band 3, despite its weak homology to human band 3, still constitutes an organization center for glycolytic enzymes on the erythrocyte membrane. To further test this hypothesis, erythrocytes from band 3 knockout mice were similarly examined by confocal microscopy. Not surprisingly, all of the enzymes in all of the cells were evenly distributed throughout the cytoplasm, regardless of the oxygenation state of the cell. Further, immunoblot analyses demonstrated that glycolytic enzyme content of the band 3 knockout erythrocytes was measurably reduced compared to healthy mice, suggesting that the anion transporter may also contribute to enzyme stabilization during the lifetime of the erythrocyte. Finally, to determine whether the integrity of other membrane structures might impact the assembly of glycolytic enzyme complexes on the erythrocyte membrane, α-spectrin deficient mice were also examined for their enzyme distributions. Curiously, > 50% of the cells in any field exhibited glycolytic enzyme staining throughout the cytoplasm, with the remainder showing mainly membrane staining. Conceivably, the stabiity of glycolytic enzyme complexes on the membrane may also depend on the integrity of the membrane skeleton. Taken together, these data argue that glycolytic enzymes assemble in an oxygenation-dependent manner into complexes on murine erythrocyte membranes and that the stability of these complexes depends on the presence of band 3 and to a lesser extent α-spectrin. Supported by NIH grant GM24417.
Author notes
Corresponding author