Doublet preference and gene evolution

Summary Doublet preference analysis was carried out on coding and noncoding regions ofEscherichia coli, Saccharomyces cerevisiae, and human mitochondrial and nuclear DNA. The preference pattern in 1–2 and 2–3 doublets inE. coli andS. cerevisiae correlated with that in noncoding regions. The 3-1 doublet preference inE. coli genes with low optimal codon frequency and inS. cerevisiae genes also showed a correlation with each of their noncoding doublet preference. A mechanism to explain these double preference correlations in doublet preference is presented: mutational biases, the origin of the noncoding region doublet preference, evolved so as to maintain the 1–2 and 2–3 doublet preference, which is determined by codon usage. These biases then acted on the 3-1 doublet, which was almost free of coding constraints, resulting in a similar preference in this doublet.     

Cloning and sequencing of a cyclodextrin glucanotransferase gene from Bacillus ohbensis and its expression in Escherichia coli

Summary A cyclcodextrin glucanotransferase (CGTase) gene of Bacillus ohbensis was cloned in Escherichia coli and the nucleotide sequence was determined. A single open reading frame (2112 bp) with a TTG codon as an initiator was identified that encodes a typical signal peptide of 29 amino acids followed by the mature enzyme (675 amino acids), of which the partial amino acid sequences of the N-terminal region and some lysyl-endopeptidase fragments were determined by Edman degradation. The CGTase gene was expressed in E. coli under control of the lac promoter only when the upstream region containing a long inverted repeat structure (located at –108 to –67 bp from the initiation codon) was deleted. Substitution of an ATG codon for the initiation TTG triplet doubled the expression of the CGTase gene in E. coli. Enzyme preparations purified from the culture supernatant of B. ohbensis and from the periplasmic fraction of the E. coli transformant exhibited the same molecular weight (M _r) and enzymatic properties as follows: M _r, 80 000; optimum pH for activity, 5.0 (and a suboptimum at 10.0); stability between pH 6.5 and 10.0; optimum temperature for activity, 55°C; and stability below 45°C. The yields of the products from starch as the substrate were 25% for -and 5% for -cyclodextrin.The nucleotide and deduced amino acid sequence data reported in this paper have been deposited in the DDBJ, EMBL and Genbank Nucleotide Sequence Databases under the accession number D90243 Offprint requests to: T. Uozumi     

Effect of 3′ Terminal Codon Pairs with Different Frequency of Occurrence on the Expression of <Emphasis Type="Italic">cat</Emphasis> Gene in <Emphasis Type="Italic">Escherichia coli</Emphasis>

In a previous study, we have identified four types of 3 terminal codon pairs depending on their frequency of occurrence in the Escherichia coli genome: overrepresented, moderately represented, underrepresented, and missing. In this study, the influence of eight codon pairs belonging to these four groups on the efficiency of chloramphenicol acetyltransferase (cat) gene expression in E. coli is examined. Our results show that the missing codon pairs CCU:UAG (Pro:Stop) and CCC:UAG (Pro:Stop) had decreasing effect, whereas another missing pair CCU:AGG (Pro:Arg) had an opposite effect on the yield of CAT protein in comparison with the wild-type cat gene.     

A bacteriocinogenic factor of Enterobacter cloacae

Summary Enterobacter cloacae strain DF13 produces a bacteriocin which is able to kill other strains of Enterobacter and Klebsiella. This property can be transmitted to Enterobacter cloacae strain O 2 (up to 90% of the acceptor population became bacteriocinogenic), to E. coli K12F^- and E. coli K 12 Hfr. Transfer of chromosome material was never observed, suggesting that the production of the bacteriocin is determined by a plasmid. However all attempts to eliminate this plasmid failed. The plasmid F trp cys Col B Col V could be transferred from E. coli into Ent. cloacae DF13 and subsequently it could be eliminated by acridine orange treatment. Ent. cloacae DF13 harbours in addition two independently transferable R-factors, one determining resistance against streptomycin and sulfanilamide and the other resistance against penicillin.Most but not all Ent. cloacae O2 recombinants which have received only the bacteriocinogenic factor upon conjugation with Ent. cloacae DF 13, can transfer this property to Ent. cloacae O2 but not to E. coli. E. coli F^- recombinants, which have received only the bacteriocinogenic factor cannot transfer this property. The results suggest that the bacteriocinogenic factor cannot mediate its own transfer, but can be transferred when another transmissible plasmid is present. This may be the R(str sul) factor, the F-factor in E. coli Hfr or a transfer factor () in Ent. cloacae O2.Closed circular DNA molecules were selectively isolated from these strains and investigated by electron microscopy and velocity sedimentation. Ent. cloacae DF13 harbours small closed circular DNA molecules ranging from 0.5 to 3.2 in contour length, 98% of which corresponds to a size class of 0.7±0.1 . Ent. cloacae O2 also harbours closed circular DNA ranging from 0.8 to 3.0 in contour length, with major size classes of 0.9 and 1.4 respectively. Circular DNA of a contour length of 3.0±0.2 (S_20,w=26 S) corresponding to a molecular weight of 6.0×10⁶ daltons was transferred to E. coli and Ent. cloacae O 2 concomitantly with the ability to produce the bacteriocin. A significant difference was observed in the number of copies of the plasmid between Ent. cloacae and E. coli.     

A comparative study of variation in codon 33 of the<Emphasis Type="Italic"> rpoS</Emphasis> gene in<Emphasis Type="Italic"> Escherichia coli</Emphasis> K12 stocks: implications for the synthesis of σ<Superscript>s</Superscript>

The Escherichia coli rpoS gene encodes an RNA polymerase sigma factor (sigma S or ^S) required for the expression of stationary-phase genes. In the first published rpoS sequence from E. coli K-12 codon 33 is given as CAG. However, several subsequent independent studies found the amber codon TAG at this position ( rpoSAm). Besides this amber codon, other codons such as TAT have also been found at this location in rpoS. Comparative genome analysis now leads us to propose TAG as the parental codon 33 in rpoS in E. coli K-12. Five different stocks of the strain W3110, which differ in the levels of ^S protein they express, were investigated. We sequenced the rpoS gene from these, and found a T at nucleotide position 97 in four out of the five stocks and a G at position 99 in three out of the five. W1485, a parental strain of W3110, and W3350, a derivative of W3110, are also rpoSAm mutants. Such rpoSAm mutants would be expected to show no RpoS activity. The retention of partial or intermediate ^S activity by suppressor-free rpoSAm mutants is therefore puzzling. We propose that a functional, N-terminally truncated, ^S (1–53^S) can be translated from a Secondary Translation Initiation Region (STIR) located downstream of the amber codon 33. It has recently been reported that a fragment of RpoS (1–53^S) that lacks the first 53 amino acids is functional when synthesized in vivo. Taken together, our results support the hypothesis that the original codon 33 of the rpoS gene in E. coli K-12 strains is the amber codon TAG.Communicated by W. Goebel     

The ribosomal protein gene cluster of Mycoplasma capricolum

Summary The DNA sequence of the part of the Mycoplasma capricolum genome that contains the genes for 20 ribosomal proteins and two other proteins has been determined. The organization of the gene cluster is essentially the same as that in the S10 and spc operons of Escherichia coli. The deduced amino acid sequence of each protein is also well conserved in the two bacteria. The G+C content of the M. capricolum genes is 29%, which is much lower than that of E. coli (51%). The codon usage pattern of M. capricolum is different from that of E. coli and extremely biased to use of A and U(T): about 91% of codons have A or U in the third position. UGA, which is a stop codon in the universal code, is used more abundantly than UGG to dictate tryptophan.     

Plasmid mediated silver resistance in Acinetobacter baumannii

Acinetobacter baumannii BL88, an environmental isolate, was resistant to 13 metals and 10 antibiotics. Plumbagin cured resistance to silver, cadmium, antimony, streptomycin and ampicillin at varying frequencies. However, only silver resistance transferred (1 × 10^–6 recepient^–1) to Escherichia coli K12 during conjugation. Correspondingly there was transfer of a 54 kb plasmid (pUPI199) from A. baumannii BL88. The plasmid transformed E. coli DH5 cells at a frequency of 1 × 10^–8 recepient^–1. The growth rate of E. coli DH5; (pUPI199) was slower as compared with E. coli DH5. Plasmid pUPI199 was 76 and 9.6% stable in the host A. baumannii BL88 in the presence and absence of selection pressure, respectively. A. baumannii BL88 was found to accumulate and retain silver whereas E. coli DH5 (pUPI199) effluxed 63% of the accumulated silver ions.     

Expression in Bacillus subtilis of the gene for human urogastrone using synthetic ribosome binding sites

Summary A chemically synthesised gene coding for human urogastrone which was earlier cloned in E. coli (Smith et al. 1982) has now been cloned into expression vectors for Bacillus subtilis Two types of constructs have been made, one giving production of methionylurogastrone and the other giving rise to a methionyl-urogastrone- galactosidase fusion polypeptide facilitating quantification of expression levels.The ribosome binding sites used in the expression plasmids are synthetically made oligonucleotides residing on short restriction fragments to allow easy replacement by other ribosome binding sites.Using shuttle vectors and constitutive promoters from Bacillus phages 105 and SPP1, we were able to detect levels of expression amounting to a few thousand molecules per cell during logarithmic growth in both E. coli and B. subtilis.     

An operational RNA code for amino acids and variations in critical nucleotide sequences in evolution

An operational RNA code relates specific amino acids to sequences/structures in RNA hairpin helices which reconstruct the seven-base-pair acceptor stems of transfer RNAs. These RNA oligonucleotides are aminoacylated by aminoacyl tRNA synthetases. The specificity and efficiency of aminoacylation are generally determined by three or four nucleotides which are near the site of amino acid attachment. These specificity-determining nucleotides include the so-called discriminator base and one or two base pairs within the first four base pairs of the helix. With three examples considered here, nucleotide sequence variations between the eubacterial E. coli tRNA acceptor stems and their human cytoplasmic and mitochondrial counterparts are shown to include changes of some of the nucleotides known to be essential for aminoacylation by the cognate E. coli enzymes. If the general locations of the specificity-determining nucleotides are the same in E. coli and human RNAs, these RNA sequence variations imply a similar covariation in sequences/structures of the E. coli and human tRNA synthetases. These covariations would reflect the integral relationship between the operational RNA code and the design and evolution of tRNA synthetases.Based on part of a presentation made at a workshop- Aminoacyl-tRNA Synthetases and the Evolution of the Genetic Code-held at Berkeley, CA, July 17–20, 1994     

Transgenic mice expressing a chimaeric anti-E. coli immunoglobulin α heavy chain gene

To induce constitutive immunity against a pathogenic strain ofEscherichia coli (K99), a rearranged immunoglobulin (Ig) heavy chain (HC) gene was constructed. Because the route ofE. coli infection is enteric, an IgA transgene was desirable. A chimaeric gene construct was cloned that coded for a HC that recognized a specificE. coli pilus antigen. The construct comprised a gene promoter, murine VDJ, and bovine -HC constant region. Following microinjection of the HC construct into murine zygotes, of 50 liveborn mice, three were identified as transgenic. In all three transgenic founders, transgene-encoded mRNA expression was detected by northern blot. The transgenic founders were analysed for transgene-encoded RNA expression in splenic tissue before and after challenge with pathogenicE. coli. Founder 4-3 expressed transgene-encoded RNA both before and after challenge; expression was detected in the other two founders only post-challenge. As no differences were found when sera were analysed for bovine IgA in control and transgenic mice, protein expression was assessed by challenge of HC founders with K99E. coli by gavage. Control mice challenged with K99E. coli were moribund within 24 h post-gavage, but there was no observable affect in the three transgenic founders. Unfortunately, after obtaining offspring from all founders, no transgenic offspring were identified (0/108). The low yield of transgenic founders, coupled with the apparent germ-line mosaicism may point to either mechanical or critical developmental anomalies. However, the production of transgenic mice harbouring an Ig HC gene construct confirmed that an Ig transgene coding for an antibody to an animal pathogen could function in a tissue-specific and protective manner in a mammalian system.     

Cloning, sequencing and expression of the amylase isozyme gene from Pseudomonas sp. KFCC 10818

Summary A gene coding for amylase was cloned and sequenced from an alkalophilic Pseudomonas sp. KFCC 10818. The coding region for the amylase precursor contained 1,692 nucleotides. The presumed Shine-Dalgarno sequence, AAGG, was located at 8 nucleotides upstream from the ATG initiation codon. The precursor protein had a putative signal peptide of 25 amino acid residues at its amino terminus. The Pseudomonas amylase had four highly conserved regions as other -amylases. The amylase expressed from E. coli harboring the Pseudomonas gene produced maltose and maltotriose from soluble starch.The nucleotide sequence reported in this paper has been submitted in the GenBank/EMBL database with accession number(s) U40056.     

Sequence and activity of an endogenous promoter of the assimilatory sulfite reductase gene from Desulfovibrio vulgaris (Hildenborough)

Summary The 5-flanking region of the assimilatory-type sulfite reductase-encoding gene from the anaerobic bacterium, Desulfovibrio vulgaris (Hildenborough), has been isolated and sequenced. The promoter element has been identified and compared with the consensus sequence for Escherichia coli and the sequences of promoter elements identified in other Desulfovibrio strains. The utility of such promoter sequences in E. coli-based expression systems is discussed.     

Electroporation of Bradyrhizobium japonicum

Summary Electroporation offers a fast, efficient and reproducible way to introduce DNA into bacteria. We have successfully used this technique to transform two commercially important strains of Bradyrhizobium japonicum, the nitrogen-fixing soybean symbiont. Initially, electroporation conditions were optimized using plasmid DNA which had been prepared from the same B. japonicum strain into which the{imDNA was to b}e transformed. Efficiencies of 10⁵-10⁶ transformants/g DNA were obtained for strains USDA 110 and 61A152 with ready-to-use frozen cells. Successful electroporation of B. japonicum with plasmid DNA prepared from Escherichia coli varied with the E. coli strain from which the plasmid was purified. The highest transformation efficiencies (10⁴ transformants/g DNA) were obtained using DNA prepared from a dcm ^– dam ^– strain of E. coli. This suggests that routine isolation of DNA from an E. coli strain incapable of DNA modification should help in increasing transformation efficiencies for other strains of bacteria where DNA restriction appears to be a significant obstacle to successful transformation. We have also monitored the rate of spontaneous mutation in electroporated cells and saw no significant difference in the frequency of streptomycin resistance for electroporated cells compared to control cells.     

In vivo effect of DNA repair on the transition frequency produced from a single O6-methyl- or O6-n-butyl-guanine in a T:G base pair

Summary We have previously reported some effects of DNA repair on the transition frequencies produced by an O⁶-methyl-guanine (MeG) or an O⁶-n-butyl-guanine (BuG) paired with C at the first position of the third codon in gene G of bacteriophage X174 form I'DNA (Chambers et al. 1985). We now report experiments in which the transition is produced from T:MeG or T:BuG, instead of C:MeG or C:BuG, located at this site. The site-modified DNAs were transfected into cells with normal DNA repair as well as into cells with repair defects (uvrA, uvrB, uvrC, recA, uvrArecA). The lysates were screened for phage carrying the expected transition using a characteristic change in phenotype. The data demonstrate that the transition frequency from T:BuG is low (0.3% of total phage progeny) in cells with normal repair (Escherichia coli AB1157) and increases 7-fold in uvrA cells (E. coli AB1886). A similar increase is seen in uvrB and uvrC cells (AB1885, AB1884). These data, like our previous data, indicate BuG is repaired primarily by excision. In contranst to this, the transition frequency from T:MeG is high (5±2%) in cells with normal repair. After induction of alkyl transfer repair in E. coli AB1157, the transition frequency goes up 5-fold. Compared with cells with normal repair, the transition frequency goes up 2-fold in uvrA, uvrB and uvrC cells; it goes up 1.5-fold in recA cells (E. coli AB2463). The data reinforce our earlier conclusion that MeG is repaired primarily by alkyl transfer, but the ABC excinuclease as well as RecA protein inhibit this repair process. Using the BuG data reported here and in our previous paper, we calculate that BuG pairs with a thymine residue 0.5%–0.62% of the time during replication in vivo, and that BuG markedly inhibits replication of the strand that contains it. Because of the complication introduced by alkyl transfer repair, similar calculations for MeG cannot be made from the current data.Abbreviations MeG and BuG O⁶-methyl-or O⁶-n-butyl-guanine moiety in X DNA (in each case, the plus strand nucleotide is specified first) - form I'DNA relaxed, covalently closed, circular, double-stranded DNA - Wt wild-type phenotype - Am amber phenotype - pfu plaque forming units - MNNG N-methyl-N'-nitro-N-nitrosoguanidine X mutants are named by designating the gene, the type of mutation (e.g. ms=missense), the codon number, the mutant codon and the new amino acid (where pertinent) in that order (e.g. XGam3) carries an amber in the third codon of gene G, and should not be confused with the classical am3 mutant used in the older literature to designate what is now known to be XEam7     

The Vitreoscilla hemoglobin gene: Molecular cloning, nucleotide sequence and genetic expression in Escherichia coli

Summary Vitreoscilla hemoglobin is involved in oxygen metabolism of this bacterium, possibly in an unusual role for a microbe. We have isolated the Vitreoscilla hemoglobin structural gene from a pUC19 genomic library using mixed oligodeoxy-nucleotide probes based on the reported amino acid sequence of the protein. The gene is expressed in Escherichia coli from its natural promoter as a major cellular protein. The nucleotide sequence, which is in complete agrecment with the known amino acid sequence of the protein, suggests the existence of promoter and ribosome binding sites with a high degree of homology to consensus E. coli upstream sequences. In the case of at least some amino acids, a codon usage bias can be detected which is different from the biased codon usage pattern in E. coli. The down-stream sequence exhibits homology with the 3 end sequences of several plant leghemoglobin genes. E. coli cells expressing the gene contain greater than fivefold more heme than controls.     

In vitro template activity of 0.3 mRNA from wild type and initiation mutants of bacteriophage T7

Summary Bacteriophage T7 0.3 mRNA synthesised and processed in vitro has been purified starting from the DNA of T7⁺ as well as from that of two initiation mutants of T7 (CR17 with a U C transition in the initiation codon and CR35b whose potential Shine and Dalgarno (S-D) interaction is interrupted by a G A transition). These mRNAs were used as templates to direct the binding of fMet-tRNA and the synthesis of 0.3 protein in both E. coli and wheat germ cell-free systems. The initiation codon mutant displayed approximately 50% inhibition of fMet-tRNA binding and 0.3 protein synthesis in both systems. The S-D sequence mutant, on the other hand, was found to be less affected than the initiation triplet mutant (20%–40% inhibition) in both fMet-tRNA binding and template activity in the E. coli system. In the wheat germ system, which does not make use of the S-D interaction, however, this mutant displayed normal template activity suggesting that the inhibition obtained in the E. coli system, albeit slight, is due to the impairment of the S-D interaction and not to an alteration of the mRNA secondary or tertiary structure caused by the base substitution.     

Optimizing the promoter and ribosome binding sequence for expression of human single chain urokinase-like plasminogen activator in Escherichia coli and stabilization of the product by avoiding heat shock response

Summary The expression of recombinant single-chain urokinase-like plasminogen activator (rscuPA) in Escherichia coli was optimized by fusing the puk gene to different promoters and ribosome binding sequences. Comparison of the tac, trp and P _L promoters showed that expression was maximal under tac control. Variation in the ribosome binding sequence and its distance to the AUG start codon yielded a further slight improvement of expression. The largest increase in rscuPA expression was achieved by variations in the host strain and growth conditions. In E. coli DG75 grown at 37°C maximal expression was achieved 30 min after induction and decreased gradually until 240 min after induction. Growth at 30°C yielded maximal expression 60 min after induction and resulted in reduced activity at longer times. Western blot analysis of the products showed that degradation of rscuPA was much larger at 37°C than at 30°C. Using E. coli CAG630 carrying the htpR mutation, which avoids heat shock response, for expression of rscuPA eliminated the instability of the product at both temperatures. Expression in this strain was even more efficient than in E. coli JM101 carrying the lon mutation. It is concluded that induction of the general heat-shock response in E. coli must be avoided to obtain stabilization of rscuPA. This drastically improves the overall yield of rscuPA from recombinant E. coli strains.     

Involvement of β-Carotene in the Antioxidant Defense of the Bacterial Cell

Effects of recombinant -carotene on the resistance of E. coli culture to menadione and paraquat were studied. The presence of -carotene in E. coli cells prevented, to a considerable extent, an increase in superoxide dismutase activity (induced by redox mediators) without affecting the culture growth. These findings suggest that -carotene is involved in the defense of cells against oxidative stress.     

Lethal and mutagenic lesions induced by ionizing radiations inE. coli and DNA strand breaks

Summary The inactivation doses forE. coli exposed to alpha particles and protons of different LETs and to gamma rays have been measured. Strains derived fromE. coli B/r showed a maximum sensitivity at LETs of 30 keV/ whilst B_s–1 and other strains known to be deficient in repair capacity had sensitivities which decreased monotonically with increasing LET. These results can be interpreted in terms of two types of lethal damage to the bacterial genome. Damage of type 1 affects only one strand of the DNA macromolecule and is partially reparablein vivo whilst damage of type 2, inflicted by one track intersection of the DNA, is irreparable. The identity of both types of damage is uncertain but type 1 probably gives rise to a lesion recognizablein vitro as a single strand break. Type 2 damage probably corresponds to double strand scission of DNA as observedin vitro.Mutations to prototrophy of three auxotrophic strains ofE. coli are induced with an effectiveness which decreases steadily with increasing LET. This form of LET dependence implies that these mutations involve damage to one target only, probably one strand of the DNA duplex.Paper read at the 6th Annual Meeting of the European Society for Radiobiology, Interlaken, 5.–8. June, 1968. Round Table: Radiation Effectsin vitro andin vivo. Correlations and Discrepancies.     

Isolation and functional identification of a novel cDNA for astaxanthin biosynthesis from Haematococcus pluvialis,and astaxanthin synthesis in Escherichia coli

We succeeded in isolating a novel cDNA involved in astaxanthin biosynthesis from the green alga Haematococcus pluvialis, by an expression cloning method using an Escherichia coli transformant as a host that synthesizes -carotene due to the Erwinia uredovora carotenoid biosynthesis genes. The cloned cDNA was shown to encode a novel enzyme, -carotene ketolase (-carotene oxygenase), which converted -carotene to canthaxanthin via echinenone, through chromatographic and spectroscopic analysis of the pigments accumulated in an E. coli transformant. This indicates that the encoded enzyme is responsible for the direct conversion of methylene to keto groups, a mechanism that usually requires two different enzymatic reactions proceeding via a hydroxy intermediate. Northern blot analysis showed that the mRNA was synthesized only in the cyst cells of H. pluvialis. E. coli carrying the H. pluvialis cDNA and the E. uredovora genes required for zeaxanthin biosynthesis was also found to synthesize astaxanthin (3S, 3S), which was identified after purification by a variety of spectroscopic methods.