Abstract
A radioisotopic assay for serum vitamin B12 is described based on the fact the unlabeled B12 will competitively inhibit the binding of Co57B12 to a specific B12 binding protein, transcobalamin I. This protein is stable in dilute solution and at the concentration of Co57B12 used for the assay, its binding capacity increases much less than IF or normal serum as the concentration of B12 increases in the reaction mixture. Because of the stability of the binding capacity of stored frozen TC-I containing serum, only a single standard curve need be established with each lot of binding protein. When a limited quantity of TC-I is incubated with 60 pg of CO57B12 and crystalline B12 concentrations ranging from 10 to 160 pg/ml, a linear standard curve is obtained when the reciprocal of the fraction of Co57B12 bound to TC-I is plotted as a function of unlabeled B12 concentration. This permits use of a simple arithmetic expression to determine the B12 concentration of unknown extracts after experimentally determining the fraction of Co57B12 bound to TC-I. The serum B12 concentration of 25 normal subjects ranged from 154 to 674 pg/ml while the range in 15 B12 deficient patients was 0-126 pg/ml. The B12 concentration in 4 patients with chronic myelogenous leukemia was greater than 1000 pg/ml. The mean variance of 15 serum extracts frozen and reassayed several weeks later was 7.4 per cent. The range of recovery of B12 added to serum extracts was 84-118 per cent.