By ROSALYN S. YALOW and SOLOMON A. BERSON (From the Radioisotope Service, Veterans Administration Hospital, New York, N. Y.)
(Submitted for publication March 7, 1960; accepted March 22, 1960)
For years investigators have sought an assay for insulin which would combine virtually absolute specificity with a high degree of sensitivity, sufficiently exquisite for measurement of the minute insulin concentrations usually present in the circulation. Methods in use recently depend on the ability of insulin to exert an effect on the metabolism of glucose in vivo or in excised muscle or adipose tissue. Thus, the insulin concentration in plasma has been estimated: a) from the degree of hypoglycemia produced in hypophysec-tomized, adrenalectomized, alloxan-diabetic rats (1) ; b) from the augmentation of glucose uptake by isolated rat hemidiaphragm (2) ; or c) from the increased oxidation of glucose- 1-C14 by the rat epididymal fat pad (3). Since there have been reports indicating the presence, in plasma, of inhibitors of insulin action (4) and of noninsulin substances capable of inducing an insulinlike effect (5, 6), these procedures, while yielding interesting information regarding the effects of various plasmas on glucose metabolism in tissues, are of doubtful specificity for the measurement of insulin per se (5).
Recently it has been shown (7, 8) that insulins from various species (pork, beef, horse and sheep) show quantitative differences in reaction and cross reaction with antisera obtained from human subjects treated with commercial insulin preparations (beef, pork insulin mixtures). An immunoassay method for beef insulin has been reported in which the insulin content is determined from the degree of competitive inhibition which the insulin offers to the binding of beef insulin-I131 by human antisera (9-12). Although human insulin reacts with human antibeef, pork insulin antiserum and displaces beef insulin-I131 by competitive inhibition (7, 8, 10), the reaction is too weak to permit measurement of the low insulin concentrations present in human plasma (7, 8, 11-13). In preliminary communications we have reported that the competitive inhibition by human insulin of binding of crystalline beef insulin-I131 to guinea pig antibeef insulin antibodies is sufficiently marked to permit measurement of plasma insulin in man (11, 12, 14), and to be capable of detecting as little as a fraction of a microunit of human insulin (12, 14). Preliminary data on insulin concentrations in man before and after glucose loading have been reported (12, 14, 15). The present communication describes in detail the methods employed in the immunoassay of endogenous insulin in the plasma of man, and reports plasma insulin concentrations during glucose tolerance tests in nondiabetic and in early diabetic subjects and plasma insulin concentrations in subjects with functioning islet cell tumors or leucine-sensitive hypoglycemia.
Was this article helpful?