1. A cell-free system prepared from Bacillus brevis cells, harvested in the late phase of growth and consisting of the 11000g supernatant, has been shown to incorporate into gramicidin S the five constituent amino acids added in labelled form. The results are consistent with complete synthesis and not merely a completion of pre-existing intermediate peptides. 2. The incorporation of (14)C-labelled amino acids by the 11000g supernatant into gramicidin S requires an energy source. Omission of phosphoenolpyruvate and pyruvate kinase from the incubation mixture prevents incorporation into gramicidin S. The cell-free system incorporates [(14)C]-leucine, -proline and -phenylalanine over a period of 4hr. With [(14)C]leucine, incorporation into gramicidin S takes place in the range pH6-9 with maximum incorporation at pH7.0. High concentrations of chloramphenicol or puromycin decreased the incorporation into gramicidin S by only about 20%. 3. The 50000g supernatant exhibited no decrease in ability of incorporating [(14)C]valine into gramicidin S as compared with the 11000g supernatant. About 40% of the incorporating ability remained in the 105000g supernatant after 3hr. centrifugation. When recombining the 105000g sediment with the 105000g supernatant, some increase in incorporation over that obtained with the supernatant alone was obtained. The findings tend to support the view that gramicidin S is synthesized in a different manner from that of proteins.

A 20 fold purification of the gramicidin S synthesizing activity in a 11,OOOxg supernatant from Bacillus brevis has been achieved. The purification procedure involves precipitation of undesired material with streptomycin sulphate, followed by precipitation of gramicidin S synthesizing activity a t 40°/, (NH,),SO, saturation and fractionation on DEAE-Sephadex. All gramicidin S synthesizing activity appeared in one peak. It incorporated labelled phenylalanine, proline, valine, ornithine and leucine to the same extent into gramicidin S. The gramicidin S synthesizing activity was found to be stable in a 0.05 M potassium phosphate buffer (pH 7.1) containing glutathione (0.01 M) and glycerol (20°/,,, v/v) for a t least 3 days a t + 4" and four weeks at -20".Experiments with cell-free extracts from Bacillus brevis has established that gramicidin S is synthesized by a mechanism which is different from that of proteins [l-31. The next stage, naturally, is the isolation of the individual enzymes and the elucidation of the mechanism of synthesis. Recently, fractionation of the gramicidin S synthesizing activity from an 11,000~g supernatant of B. brevis was briefly described [4]. The present report is a more detailed account of and also an extension of these experiments. Since the gramicidin S synthesizing activity was found to be unstable, attempts were made to stabilize it. It was found that the gramicidin S synthesizing activity was stable a t $-4" in the presence of glycerol, glutathione (GSH) and phosphate buffer. Purification of the gramicidin S synthesizing activity from an 11,000 x g supernatant using streptomycin sulphate, (NH,),SO, and fractionation on a DEAE-Sephadex column resulted in a 20 fold purification with the activity in one peak. MATERIALS AND METHODS Growth of the OrganismThe Bacillus brevis strain ATCC 9999 was grown under aeration (5 1 airlmin) in 10 1 of the synthetic medium described by Eikhom, Jonsen, Laland