The Determination of the Albumin and Globulin Contents of Human Serum by Methanol Precipitation

Although the determination of the albumin and globulin contents of human serum by neutral salt fractionation, as in Howe’s method (I), has proved useful because of its simplicity, such methods have definite limitations. Butler (2) has shown that the albumin-globulin separation by neutral salt precipitation is not’ sharp and that these fractions overlap one another grossly. Leutscher (3) has also pointed out that salting-out methods are most inaccurate in cases in which they are of greatest clinical interest, and that a change in the technique of the salting-out methods is indicated. While it is impossible to defend any of these techniques as necessarily yielding absolute albumin and globulin values, accumulated evidence indicates that electrophoretic data have more meaning than have results obtained by any salting-out method. The method described in the present paper has been found to yield albumin and globulin values which check closely with the results obtained by the electrophoretic method. The method is based on the observation of the present authors that normal human serum may be satisfactorily separated into its globulin and albumin components by appropriate treatment with methanol. Electrophoretic analyses of whole serum, and of the separated albumin and globulin fractions, showed that at 0” almost all of the serum albumin remains soluble at a concentration of 42.5 per cent methanol, in the pH range 6.7 to 6.9, and at ionic strength of about 0.03, while the globulins are almost quantitatively precipitated. A comparison of the albumin-globulin ratios obtained electrophoretically and by the methanol technique as well as by sodium sulfate separation is given in Table I and indicates that the results obtained with the first two methods agree within 5 per cent for normal serums, and within 5 to 10 per cent for abnormal serums. The results by the sodium sulfate method are in far less satisfactory agreement with the electrophoretic analysis, confirming the recent observation of Dole (4) that albuminglobulin ratios, measured electrophoretically, are roughly two-thirds the ratio found by chemical fractionation.