Observations on the crystallizable albumin fraction of horse serum

Abstract

THE advantage of sodium sulphate as a reagent for use in the precipitation and crystallization of proteins has been pointed out in a previous paper [Kekwick & Cannan, 1936] and it was considered of value to attempt to extend this procedure to the serum proteins, and in particular to the crystallization of the albumin fraction.Extensive investigations on the fractionation of serum proteins with sodium sulphate have been made by Howe [1921], but in no case was a crystalline product prepared.It soon appeared that the crystalline albumin fraction of horse serum was probably not homogeneous, a fact which has been observed a number of times, and most extensively investigated by S0rensen [1930].Two crystalline fractions have been prepared and evidence with regard to their chemical individualities is presented, together with a survey of some of their physico-chemical characteristics.The fractionation of horse serum proteins with sodium sulphate Globulin precipitation.In Howe's [1921] investigations diluted serum was always used and Homer [1919] stated that in undiluted serum sharp separation of the protein fractions did not occur either with sodium sulphate or other precipitating salts.The avoidance of large dilutions is an obvious advantage for large scale preparations and consequently the precipitation curve of undiluted serum with sodium sulphate was reinvestigated.Horse serum for this purpose, and for all subsequent preparations described in the paper, was obtained from healthy horses by collecting blood in a receptacle in the bottom of which had been poured sufficient 3 % sodium oxalate solution to make the final concentration 0 I% when the jar was filled with blood.This concentration of sodium oxalate did not inhibit clotting, but a very loose clot was formed which shrank to about one-third of its initial volume in less than 12 hr. in the cold.Practically no haemolysis occurred during this period.After siphoning off the supernatant serum and centrifuging lightly to remove sus- pended particles a water-clear pale yellow serum was always obtained.Such serum occasionally contained minimal quantities of fibrinogen.The precipitation curve was investigated at 300, 50 ml.samples of a batch of serum being added to weighed quantities of anhydrous sodium sulphate, which was dissolved by the aid of mechanical stirring.After 2 days' standing the precipitates were filtered off and washed with 25 ml.sodium sulphate solution of the concentration used for the precipitation.The filtrate and washings were combined and made up to volume and the N content determined directly by the micro-Kjeldahl method.