Human liver cDNA coding for protein C has been synthesized, cloned and sequenced. The abundance of protein C message is approximately 0.02% of total mRNA. Three overlapping clones contain 1,798 nucleotides of contiguous sequence, which approximates the size of the protein's mRNA, based upon Northern hybridization. The cDNA sequence consists of 73 5'-noncoding bases, coding sequence for a 461 amino acid nascent polypeptide precursor, a TAA termination codon, 296 3'-noncoding bases, and a 38 base polyadenylation segment. The nascent protein consists of a 33 amino acid "signal", a 9 amino acid propeptide, a 155 amino acid "light" chain, a Lys-Arg connecting dipeptide, and a 262 amino acid "heavy" chain. Human protein C and Factor IX and X precursors possess about one third identical amino acids (59% in the gamma-carboxyglutamate domain), including two forty-six amino acid segments homologous to epidermal growth factor. Human protein C also has similar homology with prothrombin in the "leader", gamma-carboxyglutamate and serine protease domains, but lacks the two "kringle" domains found in prothrombin.

Human liver cDNA encoding a protein S precursor was isolated from two cDNA libraries by two different techniques. Based upon the frequency of positive clones, the abundance of mRNA for protein S is =0.01%. Blot hybridization of electrophoretically fractionated poly(A)+ RNA revealed a major mRNA =4 kilobases long and two minor forms of -3.1 and =2.6 kilobases. One of the cDNA clones contains a segment encoding a 676 amino acid protein S precursor, as well as 108 and 1132 nucleotides of 5' and 3' noncoding sequence, respectively, plus a poly(A) region at the 3' end. The cDNAs are adenosine plus thymidine-rich (60%) except for the 5' noncoding region, where 78% of the nucleotides are guanosine or cytosine. The protein precursor consists of a 41 amino acid "leader" peptide followed by 635 amino acids corresponding to mature protein S. Comparison of the mature protein region with homologous vitamin K-dependent plasma proteins shows that it is composed of the following domains: an amino-terminal y-carboxyglutamic acid-rich region of 37 amino acids; a 36 amino acid linker region rich in hydroxy amino acids; four epidermal growth factor-like segments, each -45 amino acids long; and a 387 amino acid carboxyl-terminal domain of unrecognized structure and unknown function.

A hybrid cefE gene was constructed by juxtaposing promoter sequences from the Penicillium chrysogenum pcbC gene to the open reading frame of the Streptomyces clavuligerus cefE gene. In S. clavuligerus the cefE gene codes for the enzyme penicillin N expandase [also known as deacetoxycephalosporin C synthetase (DAOCS)]. To insert the hybrid cefE gene into P. chrysogenum the vector pPS65 was constructed; pPS65 contains the hybrid cefE gene and the Aspergillus nidulans amdS gene. The amdS gene encodes acetamidase and provides for dominant selection in P. chrysogenum. Protoplasts of P. chrysogenum were transformed with pPS65 and selected for the ability to grow on acetamide medium. Extracts of cells cultivated in penicillin production medium were analyzed for penicillin N expandase activity. Penicillin N expandase activity was detected in approximately one-third of the transformants tested. Transformants WG9-69C-01 and WG9-61L-03 had the highest specific activities of penicillin N expandase: 4.3% and 10.3%, respectively, relative to the amount of penicillin N expandase in S. clavuligerus. Untransformed P. chrysogenum exhibited no penicillin N expandase activity. Analysis of the penicillin V titer revealed that WG9-61L-03 produced titers equal to that of the recipient strain while the amount of penicillin V produced in WG9-69C-01 was reduced by five fold.

A hybrid cefE gene, encoding penicillin N expandase, was constructed by fusing the promoter sequences, Pcp, and terminator sequences, Pct from the Penicillium chrysogenum pcbC gene to the open reading frame (orf), cefEorf, from the Streptomyces clavuligerus cefE gene. The resulting hybrid gene, Pcp/cefE'orf/Pct, differed from a previously reported hybrid cefE gene contained on plasmid pPS65. The latter gene, Pcp/cefE'orf/Sct, contained the Pcp sequences fused to the S. clavuligerus cefE orf still attached to the S. clavuligerus terminator sequences, Sct. The new hybrid gene was transformed into P. chrysogenum on plasmid vector pRH6. Transformants were selected by phleomycin resistance conferred by a hybrid ble gene present on plasmid pRH6. The hybrid ble gene was formed by attaching Pcp sequences to the ble orf. Among transformants obtained with pRH6, one exhibited a 70-fold higher level of activity of penicillin N expandase than the best transformant previously obtained from a 10-fold larger population of pPS65 transformants. The penicillin N expandase activity in pRH6 transformant, 9EN-5-1, was fourfold higher than the activity in the S. clavuligerus strain used as the source of the cefE orf and 75% of the activity observed in an industrial strain of Cephalosporium acremonium. Sequencing of the junctions of the heterologous DNA in Pcp/cefEorf/Pct uncovered a modification of the cefE open reading frame introduced during construction of the hybrid gene; the modified open reading frame is designated cefE'orf.

Deacetoxycephalosporin C (DAOC), a precursor of cephalosporins excreted by Cephalosporium and Streptomyces species, has been produced in Penicillium chrysogenum transformed with DNA containing a hybrid penicillin N expandase gene (cefEh) and a hybrid isopenicillin N epimerase gene (cefDh). DAOC from a P. chrysogenum transformant was identified by ultraviolet light (UV), high performance liquid chromatography (HPLC), nuclear magnetic resonance (NMR) and mass spectrum analyses. P. chrysogenum transformed with DNA containing cefEh without cefDh did not produce DAOC. Untransformed P. chrysogenum produced penicillin V (phenoxymethylpenicillin) but not DAOC. Transformants also produced penicillin V but, in general, less than untransformed P. chrysogenum. The cefEh and cefDh genes were constructed by replacing the open reading frame (ORF) of cloned P. chrysogenum pcbC and penDE genes with the ORF of the Streptomyces clavuligerus expandase gene, cefE, and the ORF of the Streptomyces lipmanii epimerase gene, cefD, respectively. Analyses of representative transformants suggested that production of DAOC occurred via cefEh and cefDh genes stably integrated in the P. chrysogenum genome. DNA from untransformed P. chrysogenum did not hybridize to cefE or cefD gene probes.