Various ammonium sulfate concentrations and reaction conditions were employed in the fractionation of sera from rabbits, sheep, horses, and goats. Precipitates and supernatant fluids were analyzed by electrophoresis to study the effects of the controlled variables. At room temperature, the third precipitate in 35% saturated (NH)2SO4 was the best fraction from both rabbit and sheep sera; 80 to 90% of the gamma globulins were recovered. The second and third precipitates of horse sera proteins in 30% saturated (NH),2SO0 were both satisfactory, but only 44% of the gamma globulin was recovered after three precipitations. Goat sera yielded a very satisfactory fraction; 80% of the gamma globulin was recovered after two precipitations-the first in 30% and the second in 45% saturated (NH,)2SO. The composition of these fractions was not MATERIALS AND METHODS Antisera. The antisera used in these studies were produced in rabbits, sheep, horses, and goats against a variety of bacterial antigens including Bacillus anthracis, Bordetella bronchiseptica, Bordetella pertussis, Escherichia coli, Pseudomonas pseudomallei, Salmonella, Shigella dysenteriae, Shigella sonnei, and Yersinia pestis. Some normal sera were also used. Ammonium sulfate. A stock solution of saturated ammonium sulfate (SAS) was prepared and stored at room temperature (approximately 25 C). Working solutions of 50, 60, 70, 80, and 90% SAS were prepared (v/v) fresh as needed from the stock saturated solution. Equal volumes of these solutions and various antisera resulted in reaction mixtures of 25, 30, 35, 40, 45, and 50% SAS (6). Fractionation. The following procedure for frac

Summary Experimental data showed that nonspecific staining (NSS) as measured by the tissue culture cell technique increased in a linear fashion as the fluorescein to protein ratios (F/P) increased. Similar results were found at levels of conjugated protein ranging from 0.5 to 2.5%. When the same data were plotted without regard to protein concentration, it was evident that NSS was related directly to the fluorescein isothiocyanate (FITC) content of the conjugates. The F/P ratios were misleading when used to judge the NSS characteristics of conjugates since reagents having the same F/P ratios may differ widely in FITC concentration. It was shown that the NSS values given by conjugates prepared with some impure dyes were considerably higher than those of conjugates labeled with pure dyes. Globulins labeled by the dialysis method for short periods of time with the purer dye preparations yielded conjugates which gave markedly reduced NSS but whose specific staining titers were comparable to those of reagents prepared by routine methods. If conjugates of sufficiently high specific titer can be obtained, dilution is the simplest and most effective way of reducing or eliminating NSS as measured in these studies. Isomer I and II of FITC vary in respect to the nonspecific staining given by their respective conjugates depending upon the filter system employed.