A Spectrophotometric Assay for Avidin and Biotin Based on Binding of Dyes by Avidin.

The binding of aniohic dyes by serum albumin has been studied by many workers (reviewed by Foster, 1960). The evidence suggests several cationic sites situated in hydrophobic regions of the molecule. For example, the bound dye shows spectral changes that are in accord with its location in a non-polar environment. Since it was suggested (Green, 1963b) that the biotin-binding sites of avidin were in part non-polar, the interaction of avidin with several dyes was tested by spectroscopic methods. The most striking effects were shown by 4'-hydroxyazobenzene-2-carboxylic acid, a dye that has been used for the estimation of serum albumin (Rutstein, Ingenito & Reynolds, 1953) and studied in detail by Baxter (1963, 1964). When bound by albumin the dye has a decreased extinction at 348m,u and a new absorption band at 480m,u. The effects observed with avidin were even greater. Addition of the dye to an excess of avidin under conditions where almost all of it was bound gave a new absorption band at 500m,u (c500 increased from 600 to 34500) and a change of colour from yellow to red. At the same time the 348m,u band of the free azo-dye anion (e 20700; Baxter, 1964) almost disappeared. These changes were reversed by the addition of 4moles of biotin/mole of avidin. Similar spectral changes could be produced by diluting a neutral aqueous solution of the dye into lOOvol. of dimethylformamide. The results of titrating the dye-binding sites at two different avidin concentrations are shown in Fig. 1 (0, A), where they are compared with curves calculated from the law of mass action. Extrapolation of the initial linear portion of the cuirve obtained at high concentration of avidin (0) showed that 1 dye mol. was bound/biotinbinding site (4/mol. of avidin; Green, 1964). The shape of both curves agreed fairly well with that calculated assuming a single set of sites (K 5.8 x 10-6M) with no interactions between them. The slight departure from the calculated curve could be due either to the presence of intrinsically different types of avidin in the preparation used (D. 31; Melamed & Green, 1963) or to a dependence of K on the number of sites occupied. * Visiting Scientist of the U.S. Public Health Service, 1963-64. Present address: National Institute for Medical Research, London, N.W. 7. The displacement ofthe dye by biotin is shown by the third set of points (0). The sharp end point, which was to be expected from the low dissociation constant of the avidin-biotin complex (Green, 1963a), permits an accurate titration of biotinbinding sites by using the dye as an indicator. The following procedure was found convenient. The avidin (2ml. of 10,UM solution in 0-02m-sodium phosphate or -tris-HCl buffer, pH7-0) was mixed with dye at a concentration (100,UM) sufficient almost to saturate the binding sites. The decrease in extinction at 500mu was then measured after successive addition of a standard solution of biotin from a pipette or micro-syringe, until no further