The interaction of bovine serum albumin and its chicken antibodies.

In 1937 Heidelberger and Pedersen (1) mixed assayed amounts of antigen (G) and antibody (A), and used ultracentrifugal analysis to determine the amount of free component, antigen or antibody, in the system. Somewhat later these experiments were extended by Pappenheimer, Lundgren, and Williams (2) with diphtheria toxin and its equine antibody to the point at which it could be demonstrated that the empirical composition of the dissolved complex in the region of antigen excess varies from AG to AG2, approaching the latter value in the region of high antigen excess. In short, the “valence” of the precipitating equine antibody is d. More recently, Singer, and Campbell (3) devised an alternative method for analyzing soluble complexes which depends upon the fact that at low pH, such as 2.4, the antigen-antibody bonds are broken. In this way it becomes possible to determine, ordinarily by either electrophoresis or sedimentation analysis, the amount of antigen from a total, measured at the low pH, which is bound to a given measured amount of antibody at pH 8.6. With the bovine serum albumin-rabbit anti-bovine serum albumin system and the ovalbumin-rabbit anti-ovalbumin system, it could be established with these methods that the valence of the precipitin is in each case 2. The reports to which reference is made are representative of one approach to the problem. In other experiments, which have to do with hapten binding, the number of active sites on rabbit antibodies is again found to be two (4). Thus, even though the study of the combining power of antibodies formed in other species is now in its early stages, it was a matter of unusual interest to find a paper by Banovitz, Singer, and Wolfe (5) in which for the system BSAl-chicken anti-BSA antibody, the mole ratio, r, of antigen to antibody in the complex, as determined largely by electrophoretic a,nalysis, turned out to be relatively constant and not greater than approximately 0.9. No explanation is given of why r is relatively constant, but its low value is described as being the result of one of two situations: (a) either the antibody is a mixture of Iand X-valent species so that the degree of binding of BSA is less than in the rabbit system where all the antibody is bivalent, or (b) the antibody is essentially all bivalent but with