The yellow protein crystallized by Kunitz and McDonald (7) from yeast has been examined for some of its properties. The crystalline preparations employed were kindly placed at my disposal by these authors. The yellow color of the preparation appears to be largely due to a flavin prosthetic group, though another yellowish prosthetic group also appears to be present. No enzymatic function for this protein preparation has as yet been found.
EXPERIMENTALThe preparations of the yellow protein were received as a paste containing ammonium sulfate. In the first preparation received, attempts to dialyze away this salt resulted in a loss of color and denaturation of the protein. All subsequent experiments were, therefore, performed on solutions of the protein made by dissolving the paste in water. A dear lemon-yellow-colored solution was thus obtained.The absorption spectrum of such a solution, as measured in a Beckman spectrophotometer, is shown in Fig. 1 by the curve labeled "Oxidized." A peak may be observed in the region of k 450 m/~. If, to this solution, a little solid dithionite (Na~S=O4) is added, the yellow color fades pronouncedly but does not entirely disappear. This solution was overlaid with mineral oil in order to prevent its reoxidation by air and its absorption spectrum was measured. The curve obtained is the one shown in Fig. 1 labeled "Reduced." Readings below 400 m~ were not taken since Na~S204 itself absorbs in this region. If the curve thus obtained is now subtracted from that of the oxidized form, there is obtained the curve labeled "Difference" which is given in Fig. 1. This curve with a definite peak in the region of 440 to 450 m~ is suggestive of those given by various flavin compounds (cf. Ball (2)). If air is admitted to the reduced solution, the yellow color is restored, the density reading at 435 m/, returning to within 95 per cent of its original value. The reduction and oxidation process may be repeated indicating that a truly reversible oxidation-reduction system is involved in the color change.The protein content of the solution employed in these spectrophotometric experiments was determined. The protein was denatured by heat and by the addition of alcohol and the precipitate obtained washed with water until free from ammonium sulfate as indicated by Nessler's reagent. The dry weight indicated 94.6 mg. of protein per ml. of solution. If we are dealing with a flavoprotein, as seems likely, then using the density readings given in Fig.