Microbial production of spider silk proteins.

The remarkable properties of spider dragline silk and related protein polymers will find many applications if the materials can be produced economically. We have demonstrated the production of high molecular weight spider dragline silk analog proteins encoded by synthetic genes in several microbial systems, including Escherichia coli and Pichia pastoris. In E. coli, proteins of up to 1000 amino acids in length could be produced efficiently, but the yield and homogeneity of higher molecular weight silk proteins were found to be limited by truncated synthesis, probably as a result of ribosome termination errors. No such phenomenon was observed in the yeast P. pastoris, where higher molecular weight silk proteins could be produced without heterogeneity due to truncated synthesis. Spider dragline silk analog proteins could be secreted by P. pastoris when fused to both the signal sequence and N-terminal pro-sequence of the Saccharomyces cerevisiae alpha-mating factor gene.

Synthetic spider dragline silk proteins and their production in Escherichia coli.

Synthetic genes were designed to encode analogs of the two proteins of Nephila clavipes dragline silk, spidroins 1 and 2. The genes were constructed of tandem repeats of relatively long (more than 300 bp) DNA sequences assembled from synthetic oligonucleotides, and encoded proteins of high molecular mass (65-163 kDa). Both analogs were produced efficiently in Escherichia coli. The yield and homogeneity of the products of longer genes were limited by premature termination of synthesis, probably as a result of processivity errors in protein synthesis. Average termination rates were determined to be 1 in 1100 codons to 1 in 300 codons, depending on the length and synonymous codon choices of the gene. Both analog proteins could be induced to form stable aqueous solutions without denaturants. Circular dichroism spectra of the purified proteins in dilute solution resembled spectra of redissolved natural dragline silk in reflecting a largely disordered structure in water and more ordered structures in mixed solvents with methanol and trifluoroethanol.

Production of synthetic spider dragline silk protein in Pichia pastoris.

The methylotrophic yeast Pichia pastoris was tested as a host for the production of long, repetitive protein polymers. Synthetic genes for a designed analog of a spider dragline silk protein were readily expressed at high levels under control of the methanol-inducible AOX1 promoter. Transformants containing multiple gene copies produced elevated levels of silk protein, but of a variety of altered sizes as a result of gene rearrangements at the time of transformation. Genes up to 3000 codons in length or longer could be expressed with no evidence of the prevalent truncated synthesis observed for similar genes in Escherichia coli, though genes longer than 1600 codons were expressed less efficiently than shorter genes. Silk-producing P. pastoris strains were stable without selection for at least 100 doublings.

Protease-deficient Bacillus subtilis host strains for production of Staphylococcal protein A.

We constructed strains of Bacillus subtilis which produced very low levels of extracellular proteases. These strains carried insertion or deletion mutations in the subtilisin structural gene (apr) which were constructed in vitro by using the cloned gene. The methods used to construct the mutations involved the use of a plasmid vector which allowed the selection of chromosomal integrates and their subsequent excision by homologous recombination to effect replacement of the chromosomal apr gene by a derivative carrying an inactivating insert with a selectable marker (a cat gene conferring chloramphenicol resistance). The strains produced no subtilisin, no detectable extracellular metalloprotease activity, and residual extracellular serine protease levels as low as 0.5% of that of the standard strain from which they were derived. The strains proved to be superior host strains for the production of staphylococcal protein A, accumulating higher levels of intact protein than do previously available B. subtilis strains.

Gene for an immunoglobulin-binding protein from a group G streptococcus.

The gene (spg) for an immunoglobulin G (IgG)-binding protein from a Streptococcus clinical isolate of Lancefield group G was cloned and expressed in Escherichia coli. The complete nucleotide sequence of the gene and 5'-flanking sequences was determined. The DNA sequence includes an open reading frame which encodes a hypothetical protein of 448 amino acid residues (Mr = 47,595). The 5' end of this open reading frame encodes a sequence resembling a typical secretion signal sequence, and the remainder of the encoded protein has features reminiscent of staphylococcal protein A and of streptococcal M6 protein, including repeated sequences and a similar C-terminal structure. Aside from this C-terminal structure, the encoded protein has little direct amino acid sequence homology to either protein A or M6 protein. In E. coli, the cloned gene directs the synthesis of a protein which binds to immunoglobulins, including rabbit immunoglobulin, goat IgG, and human IgG3(lambda). Its binding properties are similar to those of the protein G described by Björck and Kronvall (L. Björck and G. Kronvall, J. Immunol. 133:969-974, 1984), a type III Fc receptor from a group G streptococcus.

Expression of the staphylococcal protein A gene in Bacillus subtilis by integration of the intact gene into the B. subtilis chromosome.

Staphylococcal protein A was synthesized at high levels and was secreted efficiently into the culture medium by strains of Bacillus subtilis in which the cloned gene (spa) from Staphylococcus aureus 8325-4 was inserted into the chromosome. The spa gene could not be established in B. subtilis on multicopy plasmids.

Expression of the staphylococcal protein A gene in Bacillus subtilis by gene fusions utilizing the promoter from a Bacillus amyloliquefaciens alpha-amylase gene.

Gene fusions of DNA sequences encoding protein A from Staphylococcus aureus (spa) with expression elements from an alpha-amylase gene from Bacillus amyloliquefaciens (amyEBamP) directed the synthesis and efficient secretion of protein A in Bacillus subtilis. The fusions were established on multicopy pUB110-based plasmid vectors, in contrast to the intact spa gene, which could not be stably established on plasmids in B. subtilis. Some of the resulting B. subtilis strains secreted protein A at levels in excess of 1 g/liter, demonstrating that a foreign protein encoded by an engineered gene can be secreted by B. subtilis at levels comparable to endogenous exoproteins.

Functional organization of the large ribosomal subunit of Bacillus stearothermophilus.

Bacillus stearothermophilus 50 S ribosomal subunits active in polyphenylalanine (polyPhe) synthesis were reconstituted from a mixture of purified proteins and RNA. Proteins were omitted one at a time, and the resulting particles were examined by sucrose gradient sedimentation and assayed for polyPhe synthesis, peptidyltransferase activity, and in some cases binding of elongation factor EF-G and GTP, and association with a (20 S . Phe-tRNA . poly(U)) complex. Based on their effect on polyPhe synthesis and peptidyltransferase activity, the proteins were grouped into four functional categories. The set of proteins most strongly required for peptidyltransferase activity, which must include the protein or proteins most directly involved in the active center, consists of proteins (probable Escherichia coli homologs in parentheses) B-L3 (E-L2), B-L4 (E-L4), B-L5 (E-L5), B-L6 (E-L3 or E-L6), B-L18 (E-L14), B-L20b (E-L16), and B-L25 (E-L20). Several proteins affected both polyPhe synthesis and peptidyltransferase activity more weakly. Only four proteins were required for polyPhe synthesis but not for peptidyltransferase activity, B-L2 (E-L1), B-L8 (E-L10), B-L13 (E-L7/L12), and B-L11(E-L11). The results indicate that the peptidyltransferase center is tightly integrated into the cooperative body of the 50 S subunit and that the (B-L8 . B-L13) complex is relatively independent of this cooperative domain.

Immunochemical evidence of homologies among 50 S ribosomal proteins of Bacillus stearothermophilus and Escherichia coli.

Antibodies prepared against individual 50 S ribosomal subunit proteins from Escherichia coli were reacted with 70 S ribosomal proteins from Bacillus stearothermophilus in order to identify homologous protein pairs. B. stearothermophilus proteins were separated by two-dimensional polyacrylamide gel electrophoresis and transferred electrophoretically to diazobenzyloxymethyl paper to which they became covalently attached. The paper was then washed with antiserum followed by radioactive protein A, and the resulting antigen-antibody-protein A complexes were located by autoradiography. Seventeen cross-reacting protein pairs were identified.

Chloramphenicol resistance mutation in Escherichia coli which maps in the major ribosomal protein gene cluster.

Localized mutagenesis and selection for streptomycin resistance were utilized to isolate a chloramphenicol resistance mutation in Escherichia coli K-12 linked to the strA (rpsL) locus. Bacteriophage P1 transduction verified the map position of the new resistance mutation at 72 min, placing it within a dense cluster of ribosomal protein genes. The map position differs from that of known cmlA and cmlB mutations, which map at 18 and 21 min, respectively. Ribosomes prepared from chloramphenicol-resistant and -sensitive isogenic transductants were analyzed in vitro for activity in formation of N-formylmethionyl-puromycin, polyphenylalanine, and polylysine in the presence of inhibitory concentrations of chloramphenicol. Comparisons were also made of 14C-chloramphenicol binding to 70S ribosomes and of the two-dimensional polyacrylamide gel electrophoresis pattern of ribosomal proteins from each strain. There was no detectable difference between ribosomes from sensitive and resistant strains as measured by these assays. Enzymatic modification by chloramphenicol acetyltransferase is not responsible for the observed phenotype.

Covalent cross-linking of ribosomal RNA and proteins by methylene blue-sensitized photooxidation.

Ribosomal proteins are covalently cross-linked to ribosomal RNA by irradiation with visible light in the presence of methylene blue and O2. Proteins S3, S4, S5 and S7 from the 30 S subunit of Escherichia coli ribosomes and L2 and L3 from the 50 S subunit are among the cross-linked proteins. S3 and S5 had not previously been identified as RNA-binding proteins.