THE PURIFICATION AND SOME PROPERTIES OF INSULIN

¹ From the Biological Laboratories of E. R. Squibb and Sons, New Brunswick, New Jersey

1. A scheme of purification by fractional precipitation of insulin with alcohol has been presented. The procedure yields a product as potent as crystalline insulin.

2. The solubility of insulin and associated inert protein in alcoholic solution has been described. Neutral salts increase the solubility at the pH of minimum solubility. The increase is dependent upon the concentration of the neutral salts present in the alcoholic system. Sulphates displace the pH of minimum solubility to a region quite acid to isoelectric point whereas chlorides and acetates do not alter markedly the precipitation zone. Possibly displacement by chlorides is to a region slightly alkaline to the isoelectric point.

3. Preparations of different physiological activity have been analyzed for tyrosine and tryptophane. Three of the four methods employed showed tryptophane absent in crystalline insulin. The highly potent insulin obtained by fractional precipitation with alcohol carried from less than 0.1 to 0.56 per cent tryptophane.

4. A comparison of procedures for crystallization of insulin was made. Abel's brucine-pyridine-ammonium acetate method was considered to be the most satisfactory for the majority of samples. New modifications were described, based upon the conception that crystallization will occur in systems in which insulin is held in solution at its isoelectric by agents which do not cause chemical change. All methods yield crystalline insulin of essentially the same activity.

5. Insulin was crystallized from material which has been previously coagulated by heat in acid solution. Its potency was essentially the same as crystalline insulin prepared from uncoagulated insulin.

6. The influence of purity, pH and temperature upon the rate of coagulation was studied in some detail. No conclusive evidence could be obtained which showed that highly purified insulin would coagulate at a faster rate than the cruder preparations. It was shown that presence of certain types of inert protein actually retarded the rate of coagulation. Retardation also occurred in the presence of formaldehyde.

The rate of coagulation conformed to that calculated for a first order reaction. In dilute H₂SO₄ solutions, it was found to be slowest between pH 2.6 and 3.3. Between 70 and 100°C. the temperature coefficient, µ, was 36,000.

7. Coagulated insulin was tentatively regarded to be the anhydride form.

8. Comments have been made as to the probable molecular weight of insulin protein.