Structure/function studies on the signal sequence of organophosphorus hydrolase of P̲s̲e̲u̲d̲o̲m̲o̲n̲a̲s̲ d̲i̲m̲i̲n̲u̲t̲a̲ in E̲s̲c̲h̲e̲r̲i̲c̲h̲i̲a̲ c̲o̲l̲i̲

Typescript (photocopy).The soil bacteria, Pseudomonas diminuta MG and Flavobacterium sp. 27551 both contain identical plasmid-borne opd (organophosphorous degradation) genes, but on dissimilar plasmids. Other paraoxonase strains did not appear to contain the sameopd gene in their DNA. The amino terminus of the gene product, organophosphorus hydrolase (OPH), contained characteristics of a typical signal sequence which were hypothesized to function in the subcellular transport of OPH. Alkaline phosphatase activity from opd-phoA fusions demonstrated that this amino terminal region functioned in the export of these fusions in E. coli. DNA sequencing of these fusions provided functional verification of the opd gene sequence. Western blot analysis of opd+ cell lysates of P. diminuta, Flavobacterium and Escherichia coli identified the presence of a 38,000 dalton full length protein as well as a 35,000 dalton processed OPH. This processing occurred at stationary phase and not during logarithmic growth in E. coli. The observed processing did not involve signal peptidase II as indicated by a cysteine24 to glycine mutation and treatment with globomycin. While deletion (d2-29) or defective (L17R) mutation of the signal sequence prevented export in the opd-phoA fusions, these mutations increased the specific activity of OPH by approximately two fold. The Km values of these mutants and the full length OPH were very similar to that of the processed form and the full length protein appeared to fold into a stable, enzymatically active conformation in the cytoplasm. Subcellular fractionation following cell lysis and spheroplast formation indicated that the full length and processed forms of OPH are not exported, but remained soluble in the cytoplasm. A site-directed mutant (LL) which was predicted to enhance the hydrophobicity of the signal sequence, improved export in opd-phoA fusions by approximately twenty fold as estimated by alkaline phosphatase activity, but caused OPH to be rapidly proteolyzed. Pulse-labeling studies suggested the presence of alternative start codons for the synthesis of wild type, d2-29, and LL mutants resulting in multiple molecular weight sizes. SecA did not affect either proteolytic stability or processing of LL. A model for the in vivo folding and processing of OPH in E. coli is discussed