The determination of proline in protein hydrolysates with Lactobacillus brevis.

Proline was first determined in 1901 by Fischer (2), who isolated copper Land nn-proline dihydrates from an ester fraction of an acid hydrolysate of casein. During the ensuing three decades proline was determined essentially by this procedure in the hydrolysates of more than 50 proteins by Fischer (1901-ll), Abderhalden (1902-36), Osborne (1906-ll), and other workers. That data of satisfactory analytical value were probably unobtainable by the ester method was demonstrated by Osborne and Jones (3), who were able to recover only about 73 per cent of the proline from a mixture containing eleven pure amino acids. Because of the difficulties encountered in the ester fractionation method, proline has been determined by isolation as the hydantoin (4), cadmium chloride salt (5), picrate (6), rhodanilate (7), and the mercury salt of its N-methyl derivative (8) from butanol extracts, copper salts, or protein hydrolysate fractions of other types. Dakin (9) determined proline by extracting protein hydrolysates with butanol and crystallizing the proline from ethanol solutions which had been treated with mercuric acetate and barium hydroxide to precipitate other monoamino acids. Jones and Johns (10) determined proline by calculation from the total and amino nitrogen values of hydrolysates which had been freed from the basic amino acids by precipitation with phosphotungstic acid. Guest (11) proposed a method based on the calorimetric determination of pyrrole obtained by oxidation of proline; however, proline and hydroxyproline could not be differentiated by this procedure. The solubility product principle was adapted by Bergmann and Stein (12, 13) to the quantitative determination of proline in gelatin hydrolysates. Although the method appears to yield reliable data when the numerous precautions emphasized by these authors are observed, it has not found wide application. Extensive data on the recovery of N-acetylproline from mixtures of acetylated amino acids have been reported by Synge et al. (14-16) and by Tristram (17). These authors stated that the proline values may be accurate to f5 per cent. Polson et &. (18) have de-