Simultaneous synthesis and biological applications of DNA fragments: an efficient and complete methodology

Four-helix bundle topology re-engineered: monomeric Rop protein variants with different loop arrangements

We converted the small homodimeric four-helix bundle repressor of primer protein (Rop) into a monomeric four-helix bundle by introduction of connecting loops. Both left- and right-handed four-helix bundles were produced. The left-handed bundles were more stable and were used to introduce biologically interesting peptides in one of the loops.

Simultaneous and combinatorial chemical synthesis techniques for the generation and screening of molecular diversity

By skillful 'mixing and sorting', thousand millions of different compounds, so-called molecular repertoires or molecular libraries, can be produced and screened for biological activity in a very short time. Today, both sophisticated as well as surprisingly simple procedures for simultaneous and combinatorial chemical synthesis open fascinating new perspectives for basic research, biotechnology and drug development. This contribution will give a brief survey of current strategies in the chemical synthesis of oligonucleotide and peptide repertoires with special emphasis on techniques employing membrane-type solid supports developed in this laboratory.

A novel multispecific competitor fragment for quantitative PCR analysis of cytokine gene expression in rats

Competitive polymerase chain reaction (PCR) is a sensitive method for quantification of cytokine mRNA expression. Co-amplification of an internal standard serves as control for comparing the efficiency of PCR in different samples. We have developed a novel control fragment for multiple analyses of rat cytokine gene expression containing primers for IL-1 beta, IL-2, IL-4, IL-5, IL-6, IL-10, TNF-alpha, TGF-beta 1, IFN-gamma and MIP-2. Additional primers were incorporated to analyse the content of T cells (CD3), activated T cells (CD25) and housekeeping genes (beta-actin and HPRT). As an example we demonstrate analysis of IL-2 mRNA expression in small pieces of kidney tissue obtained from rats after kidney allotransplantation. The IL-2 expression decreased tenfold during treatment with an anti-rat CD4 monoclonal antibody as compared to untreated animals.

Synthetic cryIIIA gene from Bacillus thuringiensis improved for high expression in plants

A 1974 bp synthetic gene was constructed from chemically synthesized oligonucleotides in order to improve transgenic protein expression of the cryIIIA gene from Bacillus thuringiensis var. tenebrionis in transgenic tobacco. The crystal toxin genes (cry) from B. thuringiensis are difficult to express in plants even when under the control of efficient plant regulatory sequences. We identified and eliminated five classes of sequence found throughout the cryIIIA gene that mimic eukaryotic processing signals and which may be responsible for the low levels of transcription and translation. Furthermore, the GC content of the gene was raised from 36% to 49% and the codon usage was changed to be more plant-like. When the synthetic gene was placed behind the cauliflower mosaic virus 35S promoter and the alfalfa mosaic virus translational enhancer, up to 0.6% of the total protein in transgenic tobacco plants was cryIIIA as measured from immunoblot analysis. Bioassay data using potato beetle larvae confirmed this estimate.

Expression of a synthetic gene encoding potato carboxypeptidase inhibitor using a bacterial secretion vector

A synthetic gene encoding the 39-amino-acid (aa) potato carboxypeptidase inhibitor IIa (PCI-IIa) has been constructed and expressed using the secretion vector, pIN-III-ompA-3, fused in frame to the OmpA signal peptide-encoding sequence. Recombinant Escherichia coli secreted a PCI with 10 additional aa at the N terminus (rePCI + 10). These extra aa were removed by site-directed mutagenesis giving a PCI with no additional aa (rePCI), as shown by fast atom bombardment mass spectrometry (M(r) 4295). The two forms of rePCI were found almost exclusively in the culture medium, not in the periplasmic space, as would be expected from OmpA signal peptide fusions. Both rePCI + 10 and rePCI are biologically active and react strongly with serum raised against PCI from potato. A method for the purification of rePCI to homogeneity has been developed. The purified rePCI shows a Ki for carboxypeptidase A within the range of the natural PCI-IIa (1.5-2.7 nM). These results indicate that both rePCI + 10 and rePCI are properly folded and that their three disulfide bridges are correctly formed. Together with previous reports, our results show that fusion to a secretion signal peptide is an effective way of producing small proteins containing disulfide bridges in a biologically active form.

Restructuring the translation initiation region of the human parathyroid hormone gene for improved expression in Escherichia coli

Overexpression of native human parathyroid hormone in Escherichia coli was achieved by a modification of the 5' end of the genomic gene sequence, thereby adapting this part of the translation initiation region to the bacterial host. Some simple rules abstracted from optimization studies of translation initiation of a beta-interferon gene were applied. These included (a) extending complementarity of the mRNA to the anticodon loop of tRNAfMet by use of a codon with a purine nucleotide directly following the ATG, (b) avoidance of stable secondary structure in the mRNA by use of synonymous A/U-rich codons, (c) elimination of a potential second Shine-Dalgarno sequence. The appropriate silent changes led to a 20-fold increase in parathyroid hormone production resulting in 4.3% of total soluble protein. This result proves the validity of our simple approach for optimization of foreign gene expression in E. coli.

Effect of a single 3'-methylene phosphonate linkage on the conformation of an A-DNA octamer double helix

The three-dimensional structure of the self-complementary DNA octamer d(GCCCGpGGC) has been determined in the crystalline state using X-ray diffraction data to a nominal resolutoin of 2.12 measured from a very small crystal at DESY, Hamburg. The structure was refined with stereochemical restraints to an R value of 17.1%. d(GCCCGpGGC), containing one single 3'-methylene phosphonate linkage (denoted p), forms an A-DNA double helix with strict dyad symmetry, that is distinct from canonical A-DNA by a wide open major groove and a small average base-pair inclination against the helix axis. The conformation of the unmodified control d(GCCCGGGC) is known from an X-ray analysis of isomorphous crystals (Heinemann et al. (1987) Nucleic Acids Res. 15, 9531-9550). Comparison of the two structures reveals only minor conformational differences, most notably in the pucker of the reduced deoxyribose. It is suggested that oligonucleotides with charged 3'-methylene phosphonate groups may form stable duplexes with complementary DNA or RNA strands rendering them candidates for use as gene-regulatory antisense probes.

A phasmid optimised for protein design projects: pMAMPF

Protein design requires the rapid production of recombinant genes and active recombinant proteins, the latter in sufficient amounts for functional and physical studies. We present here the construction and application of a new phasmid vector system, using the fd phage origin, lambda pL promoter, ompA-leader sequence and pMB1 origin, which allows the preparation of secretable proteins in active form, mutagenesis and gene sequencing, without subcloning steps. The vector can be used in plasmid form in a stably transformed culture to induce product formation, or as a packaged single-stranded phasmid, which, via batch transduction in a growing culture, leads directly to recombinant protein formation. This latter method has the advantage that, during the short period required for phasmid amplification, little counterselection against clones with high rDNA-protein synthesis potential occurs. The total sequence of pMAMPF-1 and pMAMPF-3 can be assembled from known sequences of constituent fragments. Mutated regions were directly sequenced.

Expression of the chemically synthesized coding region for the cytotoxin alpha-sarcin in Escherichia coli using a secretion cloning vector

The coding region for the cytotoxin alpha-sarcin from Aspergillus giganteus has been chemically synthesized by the ligation of 19 overlapping oligodeoxyribonucleotides. An Escherichia coli clone producing the cytotoxin was constructed by inserting the synthesized gene directly downstream to the region coding for the signal peptide of the OmpA protein (a major outer membrane protein of E. coli), using the secretion cloning vector pIN-III-OmpA2. The enzyme encoded by the chemically synthesized gene expressed in E. coli displayed properties identical to those of native alpha-sarcin isolated from A. giganteus with respect to its chemistry, antigenicity and ribonucleolytic activity in qualitative assays.

Temporal fluctuations in HIV quasispecies in vivo are not reflected by sequential HIV isolations

A genetic study has been made of the HIV tat gene from sequential HIV-1 isolates and the corresponding infected peripheral blood mononuclear cells. DNA was amplified by polymerase chain reaction (PCR) and cloned into a eukaryotic expression vector. Twenty clones were sequenced from each sample. Comparing the sequential HIV isolates, abrupt differences were seen between the major forms of each isolate. These progressive changes were not reflected at all among the in vitro samples. The fluctuation in the quasispecies in vivo may suggest a much more dynamic role for latently infected mononuclear cells. High frequencies of functionally defective tat genes were identified. Given such complexity and the evident differences between quasispecies in vivo and in vitro, the task of defining HIV infection in molecular terms will be difficult.

An efficiently transcribed human tRNA(Val) gene variant produces a stable pre-tRNA: repair of the processing defect by in vitro mutations

A tRNA(CACVal) gene variant, pHtV4, was cloned from human placenta genomic DNA. This gene differs from a closely related, functional tRNA(CACVal) gene by four base exchanges: T residues in place of C25, C62, and C66 create G:U pairs, and an A instead of G65 creates an A:C mismatch in the corresponding RNA transcript. The tRNA(Val) gene variant in pHtV4 is efficiently transcribed in HeLa cell nuclear extracts; however, the resulting pre-tRNA is processing-deficient, i.e., neither its 5'- nor its 3'-flanking sequences are removed to generate mature tRNA. Reversion of all four point mutations in pHtV4 by oligonucleotide-directed mutagenesis yielded a functional tRNA(CACVal) gene within the flanks of pHtV4, the pre-tRNA of which was processed to mature tRNA. Construction of a chimeric tRNA(Val) gene and site-directed mutagenesis of the tRNA(Val) gene in pHtV4, respectively, followed by transcription and processing studies showed that each of the four mutations contributes to the processing defect of the pHtV4-derived pre-tRNA. Moreover, this revealed that G:U pairs, which are common in all tRNAs, can impair pre-tRNA processing and therefore do not occur in certain positions in eukaryotic tRNAs.

Human pancreatic secretory trypsin inhibitor (PSTI) produced in active form and secreted from Escherichia coli

As a basis for a protein design project, we decided to produce the human pancreatic secretory trypsin inhibitor (PSTI) in its active form. Total gene synthesis was carried out efficiently by (i) computer design of the gene fragments, (ii) synthesis of the oligodeoxynucleotides by the segmental support method, and (iii) assembly of double strands under optimized ligation conditions. Fusion to the ompA gene signal peptide led to secretion of processed PSTI in various constructions, with or without additional amino acids (aa) at the N-terminus. The secreted proteins (56 to 63 aa) were biologically active, suggesting that the three cysteine bridges were correctly formed. Surprisingly, after induction the product was found almost exclusively in the culture medium. Variants of PSTI with Asp or Asn at aa positions 21 and 29 [sequences published by Greene et al., Methods Enzymol. (1976) 813-825, and by Yamamoto et al., Biochem. Biophys. Res. Commun. (1985) 605-612] showed the same Ki for both human and porcine trypsin.

Expression of the chemically synthesized gene for ribonuclease T1 in Escherichia coli using a secretion cloning vector

The gene for ribonuclease T1 from Aspergillus oryzae has been chemically synthesized using the segmental support technique. An Escherichia coli clone producing the ribonuclease at high levels was constructed by linking the gene downstream to the region coding for the signal peptide of the OmpA protein (a major outer membrane protein of E. coli), using the secretion cloning vector pIN-III-ompA2. This strategy was employed in order to circumvent a possible toxic effect of the gene product on the host cell. Active ribonuclease containing four additional amino acids at the N-terminus could be isolated from the periplasmic fraction of the host. The final yield after purification was 20 mg enzyme/l liquid culture. With respect to immunological, catalytic and specific behaviour, no qualitative differences could be detected between the enzyme from the over-producing E. coli strain and ribonuclease T1 isolated from A. oryzae.

Complete synthesis and transcription in vitro of a gene coding for human ribosomal 5S RNA

The gene coding for the major human ribosomal 5S RNA was chemically synthesized and cloned into a pUC13 vector. This approach was taken, because attempts to isolate the human 5S gene have thus far yielded either pseudogenes or variant 5S genes of unknown function. The synthetic human gene was transcribed by RNA polymerase III either in a crude HeLa cell extract or in a system reconstituted from partially purified transcription factors. Comparative studies with the Xenopus laevis somatic 5S gene show that the human gene is transcribed with similar fidelity and an efficiency of about 80% under optimal conditions. The time-course of transcription and optimal concentrations of template and transcription factors were found to be similar for both genes studied. The synthetic gene described may prove useful to study its interaction with human transcription factors in a homologous system.