Sensitivity of isoelectric focusing, ion exchange, and affinity chromatography to labile glycated hemoglobin

We examined the effect of labile glycated hemoglobin on measurements of glycated hemoglobin by several commercial procedures. Erythrocytes from diabetic and nondiabetic patients were incubated in vitro with various concentrations of glucose, to generate labile glycated hemoglobin, and the species of glycated and nonglycated hemoglobin in each sample were identified by isoelectric focusing. Glycated hemoglobin was then assayed by the Bio-Rad A1 column method (I), the Bio-Rad A1c column method (II), and the Pierce affinity column method (III). I was sensitive to the labile (aldimine) fraction of glycated hemoglobin, and percentages of glycated hemoglobin so determined represented the sum of the labile fraction plus hemoglobin A1c and other stable glycated species. This spurious increase in glycated hemoglobin concentration by the aldimine could be obviated by any of three wash procedures, which eliminated the labile fraction from the samples: incubating the erythrocytes (a) in phosphate-buffered saline (pH 7.4) for 18 h at 22 degrees C, (b) in 100 mmol/L citrate (pH 5.0) for 30 min at 37 degrees C, or (c) in saline containing, per liter, 30 mmol of semicarbazide and 12 mmol of aniline hydrochloride (pH 5.0) for 30 min at 37 degrees C. Methods II and III did not detect the labile fraction. However, treatments a-c decreased the concentrations of stable glycated hemoglobin as determined by all three column-chromatographic methods as compared with unwashed sample. By isoelectric focusing we determined that blood with high glucose content had concentrations of aldimine roughly proportional to the blood glucose concentration. The kinetics of formation of labile glycated hemoglobin in these cells were consistent with the reported rate constants determined by using purified hemoglobin preparations in vitro.