Improvement of the thiobarbituric acid assay for serum glycosylprotein determination.

This study shows that the thiobarbituric acid (TBA) reaction as applied to the measurement of non-enzymatically glycosylated serum proteins(s) [1], yields erroneous results unless strictly standardized conditions are followed. Critical points are in particular (a) the concentrations of NaBH4 required for the preparation of the blanks by reduction of the ketoamine linkages; (b) the amounts of protein which should be identical in all serum samples to be compared; (c) dialysis for removal of glucose and NaBH4 prior to hydrolysis; (d) increase in the yield of 5-hydroxymethylfurfural (HMF) by appropriate conditions of hydrolysis. From our data obtained with an improved TBA-assay it appears that about 90% of total glycosylated serum protein is accounted for by glycosylated albumin. Thus changes in the latter used for diagnostic purposes in diabetes, may well be reflected by total serum glycosylprotein measurement.