Hb in Diabetic Patients with Normal ( Hb A ) and Abnormal ( Hb S and Hb C ) Hemoglobins , as Measured

In a recent study (1 ) of diabetic patients with hemoglobins S and C (Hb S and Hb C), we showed a significant decrease in the proportion of glycosylated Hb, as measured by cation-exchange chromatography. We suggested ( I ) that these low values for glycosylated Hb might be caused by the inability of either Hb S or Hb C to undergo glycosylation or by their failure to be eluted from the column with the usual buffer solution. To explore these two possibilities, we used an alternative, colorimetric method, in which glycosylated Hb is measured without first separating total hemoglobin into its various components. The method, first described by Fluckiger and Winterhalter (2), and recently modified by Gabbay et al. (3) to increase its sensitivity and reproducibility, is based on the fact that ketoamine-linked hexoses are liberated on heating under acidic conditions as furfural compounds, which are then quantitated colorimetrically as an add uct of 2-thiobarbituric acid. Using a combination of colorimetric and microcolumn chromatographic methods, we have found evidence that both Hb S and Hb C undergo glycosylation, and that the rates of their glycosylation are similar to that for Hb A. The patients we studied were from the Diabetic Center of The Montreal General Hospital. Blood specimens were drawn into an evacuated collection tube, with ethylenediaminetetraacetate as anticoagulant. The percentage of abnormal hemoglobins was estimated by electrophoresis on acrylamide gel. The proportions of glycosylated Hb in the erythrocytes of each patient were determined by the modified colorimetric method (2,3) and by a commercial cation-exchange microcolumn chromatographic procedure (Isolab, Akron, OH 44321) carried out at 23 ± 0.3 #{176}C. Table 1 shows the colorimetric and chromatographic values forglycosylated Hb in four patients with abnormal hemoglobin. For comparison, the results of a group of patients with normal Hb (Hb A), who had similar colorimetric values, are shown. Note that although the colorimetric values in each pair of patients (normal vs abnormal Hb) were similar, the chromatographic values were consistently lower in patients with abnormal hemoglobins. Our subsequent studies have shown that in subjects with normal hemoglobin, the colorimetric values increased linearly with increasing chromatographic values. Similar observations were made by Gabbay et al. (3), who found that the colorimetric values determined directly on total hemolysate correlated significantly and linearly (r = 0.87, p <0.001) with the percentage of Hb A, determined by macrocolumn chromatography. This linear correlation between results by the two methods did not, however, apply to the abnormal hemoglobins. Table 2 compares the glycosylated Hb values obtained by column chromatographic and colorimetric methods in seven patients with Hb S and Hb C. The absorbance readings by the colorimetric method were converted to glycosylated Hb values by reference to a standard curve correlating chromatographic and colorimetric values in subjects with normal hemoglobin. As is shown in Table 2, the column-chromatographic method underestimates glycosylated Hb in patients with Hb S and Hb C. To determine whether this decrease was proportional to the percentage of the abnormal hemoglobins, we corrected the chromatographic values by using the following formula (1):