The entire structural gene for tyrocidine synthetase 1 from Bacillus brevis ATCC 8185 has been cloned and expressed in Escherichia coli. Transformed E. coli cells were screened for their ability to produce tyrocidine synthetase 1 by in situ immunoassay using antibodies against gramicidin S synthetase 2 which cross-react with tyrocidine synthetase 1. The cloned gene is within a 5.2 kb fragment of B. brevis genomic DNA and requires no external promoter for its expression in E. coli. It was also observed that cloning of the 5.2 kb insert in the opposite orientation still resulted in a high level of tyrocidine synthetase 1 expression in transformed E. coli cells. In addition, protein blotting and partial purification of the gene product by gel filtration revealed a major protein of molecular weight about 100,000 with specific D-phenylalanine dependent ATP-32PPi and 2'deoxy ATP-32PPi exchange activities. These unique activities of tyrocidine synthetase 1 were not detected in protein extracts of E. coli strains carrying the vector.

When cultivated at 37°C in static broth, human clinical isolates of Yersinia enterocolitica (serogroups 0:3, 0:8, and 0:9) and Yersinia pseudotuberculosis (serogroup 0:11I) produced numerous nonflagellar surface appendages, which appeared as a lawn of fine fibrillae, each having a diameter of 1.5 to 2.0 nm and a length of 50 to 70 nm. Cultivation at 22°C resulted in complete disappearance of the fibrillae. The phenotypic expression of these appendages was correlated with the presence of the 40to 48-megadalton virulence plasmid and was strongly affected by the growth medium. Evidence is presented which suggests that these plasmidmediated, temperature-inducible surface fibrillae are responsible for autoagglutination and are related to production of one prominent, Sarkosyl-insoluble polypeptide of ca. 180 kilodaltons in the bacterial outer membrane.

The multienzyme gramicidin S synthetase 2 was treated with trypsin to obtain fragments capable of activating proline. Three different active fragments were detected. The course of proteolysis was simulated by using a concentration range of trypsin; the cleavage pattern indicated that one of the fragments was particularly stable. This fragment was purified and shown to have a molecular mass of 115 kDa. It was compared chromatographically, by SDS/PAGE, and enzymatically to a Pro-activating fragment produced by a gramicidin-S-negative mutant. It can be concluded that the proteolytic fragment represents a structure which is contained on a continuous part of the polypeptide chain of gramicidin S synthetase 2 and has a relatively compact structure. This provides evidence that the multienzyme gramicidin S synthetase 2 is, at least in part, constructed from functional domains. An approach towards extending these studies to other parts of the gramicidin S synthetase 2 molecule has also been devised. This work complements recombinant DNA studies in the area, providing stable functional fragments.

The multienzyme gramicidin S synthetase 2, composed of one polypeptide chain, was treated with trypsin and chymotrypsin to give fragments retaining partial enzyme activities. Previously, a tryptic fragment of this multi‐enzyme has been identified as a structural and functional domain. In this study two more fragments, activating Leu and Val, respectively, are shown to represent domains. Careful inspection of the data on limited proteolysis, from this study as well as from previous work, suggests that domains are not simply connected like pearls on a string, and a model for the structure of gramicidin S synthetase, with implications for other peptide synthetase multienzymes, is presented. It is suggested that gramicidin S synthetase 2 is constructed from core catalytic domains and intervening framework. Such an interpretation is in accordance with all published data on limited proteolysis of peptide synthetases, but needs an interplay with gene structural studies in order to be validated and refined.