The halophilic archaebacterium Halobacterium strain GRB harbours a multicopy plasmid of 1.7 kb which is negatively supercoiled. After addition of novobiocin to culture medium all 1.7 kb plasmid molecules become positively supercoiled. Positive supercoiling occurs at the same dose of novobiocin inhibiting the eubacterial DNA gyrase in vitro. Novobiocin also induces positive supercoiling of pHV2, a 6.3 kb plasmid from Halobacterium volcanii. These results indicate the existence of a mechanism producing positive superturns in halobacteria. The 1.7 kb plasmid from Halobacterium GRB could be used to produce high amounts of pure positively supercoiled DNA for biophysical and biochemical studies. INTRODUCTION Superhelicity is an essential feature of DNA shaping and determines a variety of biological events (for reviews, see ref. 1-3). Circular DNA duplexes isolated from cells and viruses are negatively supercoiled. Negative supercoiling facilitates DNA melting whereas fine adjustments of the level of supercoiling may control the strength of DNA-protein interactions. In contrast to negative supercoiling, positive supercoiling might be unfavourable to replication, transcription and recombination by increasing the stability of the DNA double helix. As expected from such considerations, no positively supercoiled DNA has been isolated from cells in physiological conditions. Therefore, it came as a surprise that the thermophilic archaebacterium Sulfolobus acidocaldarius contains an ATP-dependent type I DNA topoisomerase which introduces positive superturns in covalently closed circular © I R L Press Limited, Oxford, England. 1 379 Nucleic Acids Research DNA: reverse gyrase (4-6). In addition, the DNA of a virus-like particle (SSVI) discovered in Sulfolobus solfataricus B12 is positively supercoiled (7,8). The archaebacteria are a group of prokaryotes including extreme thermophiles, methanogens and extreme halophiles (halobacteria) (9). The halobacteria are mesophilic organisms which require 3-4 M NaCl in their culture medium and contain 3-4 M KCI in their cytoplasm (10). in contrast to the DNA of Sulfolobus virus-like particle SSVl, the 1.7 kb plasmid of the halophilic archaebacterium Halobacterium GRB is negatively supercoiled (11). Halobacteria probably contain a classical type II DNA topoisomerase since they are sensitive to inhibitors of these enzymes such as the epipodophyllotoxins (VP 16 and VM26) and the coumarines (novobiocin and relatives) (8,11-13). In particular, halobacteria are inhibited by novobiocin concentrations which are otherwise specific for the inhibition of negative supercoiling generated by eubacterial type II DNA topoisomerase (DNA gyrase). It has been suggested that positive supercoiling by reverse gyrase could prevent the genomic DNA of Sulfolobus from denaturation at high temperature (4). In this hypothesis, positive supercoiling could be restricted to thermophilic archaebacteria. Nevertheless, we report here that novobiocin induces positive supercoiling of small plasmids in halobacteria. The different mechanisms which could produce positive supercoiling in these mesophilic archaebacteria are discussed. MATERIALS AND METHODS Bacterial strains, growth and novobiocin treatment. Halobacterium strain GRB (14)(a gift from U. Rdest, Wurrzburg University, FRG) and H. volcanii (a gift from I. Rosenshein, Tel Aviv University) were grown at 370C in liquid shaken cultures. Halobacterium GRB was grown in classical halophile medium (12) and H. volcanii was grown in the medium of Mullakhanbhai and Larsen (15). Novobiocin was added when the cultures reached an optical density of 0.3 at 600 nm.

Synechocystis PCC6803 displays two inorganic carbon‐uptake processes, a low‐affinity one (apparent K m : 300–400 µM) functional in cells grown under standard or limiting inorganic carbon concentrations, and one with a higher affinity (60±12 µM), detected only in cells adapted to limiting inorganic carbon conditions. A mutational and screening procedure allowed the isolation of a mutant deficient in the high‐affinity system, but only slightly impaired in its growth capacities. The mutated genomic region revealed two open reading frames (ORFs), possibly belonging to an operonic structure. A clone in which the downstream ORF, hatR ( h igh‐ a ffinity t ransport), had been inactivated showed a phenotype close to that of the original mutant. Inactivation of the other ORF, hatA , yielded a clone unable to grow in limiting inorganic carbon conditions. The deduced HatA protein showed no homology with any registered protein. It possessed three hydrophobic domains, including a putative signal peptide. Several hypotheses are considered as to its role. The deduced HatR protein, which possessed the features characteristic of the response regulators of the two‐component regulatory systems ubiquitous in bacteria, might be a regulator controlling the activity of the high‐affinity transport process. It would belong to the subclass of these molecules lacking the DNA‐binding domain.

As an approach in the study of the evolution of threonine biosynthetic pathways throughout various organisms, the sequences of three enzymes, namely homoserine dehydrogenase, homoserine kinase and threonine synthase, originating from six organisms, namely Escherichia coli, Bacillus subtilis, Corynebacterium glutamicum, Brevibacterium lactofermentum, Pseudomonas aeruginosa and Saccharomyces cerevisiae, were compared. As a general trend all three enzymatic activities were carried out by proteins sharing sequence relatedness (except for the homoserine kinase of P aeruginosa). Unexpectedly however, for each step one or two enzymes stood out of the main stream: i) for homoserine dehydrogenase, the yeast protein is atypically similar to the E coli enzyme; ii) for homoserine kinase, the P aeruginosa protein shares no similarity with any other species; and iii) for threonine synthase, the B subtilis protein is far distant from the enzymes of other species. Hence in contrast to other biosynthetic pathways such as the tryptophan one, the threonine pathway seems not to have evolved as a whole throughout different organisms but rather each step seems to have been subjected to multiple constraints including substrate-mediated ones and host-specific ones.

Three genes from Pseudomonas aeruginosa involved in threonine biosynthesis, hom, thrB and thrC, encoding homoserine dehydrogenase (HDH), homoserine kinase (HK) and threonine synthase (TS), respectively, have been cloned and sequenced. The hom and thrC genes lie at the thr locus of the P. aeruginosa chromosome map (31 min) and are likely to be organized in a bicistronic operon. The encoded proteins are quite similar to the Hom and TS proteins from other bacterial species. The thrB gene was located by pulsed-field gel electrophoresis experiments at 10 min on the chromosome map. The product of this gene does not share any similarity with other known ThrB proteins. No phenotype could be detected when the chromosomal thrB gene was inactivated by an insertion. Therefore the existence of isozymes for this activity is postulated. HDH activity was feedback inhibited by threonine; the expression of all three genes was constitutive. The overall organization of these three genes appears to differ from that in other bacterial species.

A genetic locus implicated in the synthesis and secretion of alkaline protease (APR) in Pseudomonas aeruginosa has been previously described [Guzzo et al. J. Bacteriol. 172 (1990) 942–948]. The nucleotide sequence of the DNA fragment encoding these functions was determined and revealed the existence of five open reading frames: aprA, the structural gene encoding APR; aprI, which encodes a protease inhibitor; and aprD, aprE, aprF whose products are involved in protease secretion. The AprD, AprE and AprF proteins share significant homology with proteins implicated in secretion of Erwinia chrysanthemi proteases and Escherichia coli α-haemolysin. These results provide further evidence for the existence of a specialized secretory system widespread among Gram − bacteria.

The Zymomonas mobilis phoA gene, encoding a phosphate-irrepressible alkaline phosphatase (ZAPase), was cloned and its expression was studied in phoA mutants of Escherichia coli. The ZAPase was recovered in the soluble fraction of E. coli. The enzyme was synthesized constitutively and its synthesis not repressed by phosphate, unlike the phoA gene of E. coli. The phoA gene of Z. mobilis was mutagenized by Mini Mu PR13 and the mutated gene crossed into Z. mobilis in order to obtain phoA mutants by reverse genetics. Although Z. mobilis mutants with Mini Mu PR13 integrated in the chromosome were obtained, none had an allele replacement for none was defective in ZAPase.

A L-delta 1-pyrroline-5-carboxylate reductase activity has been detected in crude extracts of Desulfovibrio desulfuricans Norway. This P5C reductase activity is also found when a 2.5 kb D. desulfuricans DNA fragment is introduced into an Escherichia coli proC mutant. Although it restores growth of the proC mutant, the ProDd enzyme might be detrimental to the E. coli host since the plasmid carrying the cognate proDd gene is segregated at high rate by the cells but is stabilized by small deletions which lead to a loss of the P5C reductase activity.

The Erwinia chrysanthemi (strain 3937) celY gene encoding the minor endoglucanase (EGY) was sequenced. The analysis of the upstream region allowed us to identify an in vivo active promoter recognized by the NtrA (sigma-54) holoenzyme. No similarity was found between the predicted amino acid (aa) sequences of EGY and either the Er. chrysanthemi major endoglucanase, EGZ, or the Er. carotovora CelS endoglucanase. In contrast, a very high level of identity, both at the nucleotide and the predicted aa levels, was found between celY and an EG-encoding gene from Cellulomonas uda, a Gram+ bacterium taxonomically distant from Er. chrysanthemi. By comparing the molar G + C% of the cellulase-encoding genes and that of Er. chrysanthemi and C. uda chromosomal DNAs, we speculate that celY was transferred from Er. chrysanthemi to C. uda.

The gene from Bacillus licheniformis coding for a thermostable α-amylase was subcloned into the broad-host-range plasmid pKT210 in Escherichia coli. The recombinant plasmid pGNB6 was transferred into Zymomonas mobilis ATCC 31821 by conjugation. Plasmid pGNB6 was stably maintained in E. coli and unstable in Z. mobilis. The amylase gene was expressed in Z. mobilis at a lower level (25%) than in E. coli and regulation of enzyme biosynthesis was different in the host cells. Almost all the α-amylase activity was recovered in the culture medium of Z. mobilis. This enzyme localization seemed to be the result of protein secretion rather than cell lysis. Integration of the amylase gene into a cryptic plasmid of Z. mobilis was observed. The amylase gene was still expressed, although at a lower level, and the α-amylase activity, associated with a protein of molecular mass 62,000 daltons, was immunologically identical in Z. mobilis, E. coli and B. licheniformis.

The Zymomonas mobilis gene (sacA) encoding a protein with sucrase activity has been cloned in Escherichia coli and its nucleotide sequence has been determined. Potential ribosome-binding site and promoter sequences were identified in the region upstream of the gene which were homologous to E. coli and Z. mobilis consensus sequences. Extracts from E. coli cells, containing the sacA gene, displayed a sucrose-hydrolyzing activity. However, no transfructosylation activity (exchange reaction or levan formation) could be detected. This sucrase activity was different from that observed with the purified extracellular protein B46 from Z. mobilis. These two proteins showed different electrophoretic mobilities and molecular masses and shared no immunological similarity. Thus, the product of sacA (a polypeptide of 58.4-kDa molecular mass) is a new sucrase from Z. mobilis. The amino acid sequence, deduced from the nucleotide sequence of sacA, showed strong homologies with the sucrases from Bacillus subtilis, Salmonella typhimurium, and Vibrio alginolyticus.

SummaryAn open reading frame (ORF) of 141 bp was observed upstream from the Pseudomonas aeruginosa lysA gene. The translation product of this ORF contains a signal peptide with a lipoprotein box, fle‐Ala‐Ala‐Cys, at the predicted signal peptidase cleavage site. The Escherichia coli phoA gene without its signal sequence was fused in frame to this ORF in a broad host‐range plasmid. The resulting construct expressed a hybrid protein exhibiting alkaline phosphatase activity in phoA mutants of both E. coli and P. aeruginosa. This indicates that the ORF encodes a peptide, part of which acts as an export signal. The hybrid peptide was identified by immunoblotting with alkaline phosphatase antiserum. The accumulation of a precursor form was observed when P. aeruginosa cells carrying this gene fusion on a plasmid were treated with globomycin. Moreover, the mature form could be labelled with 2‐[3H]‐glycerol, indicating that lipidic residues may be linked to the hybrid protein. Taken together, these results strongly suggest that the ORF encodes a lipopeptide. We propose that the gene is called IppL.

The cellulase gene from Erwinia chrysanthemi coding for endoglucanase Z was subcloned into a broad-host-range plasmid pGSS33 in Escherichia coli. The recombinant pNB20 was transferred into Zymomonas mobilis ATCC 10988 by mobilization using the helper plasmid RP4. Plasmid pNB20 was stably maintained in E. coli and Z. mobilis hosts. The endoglucanase gene celZ was expressed efficiently and the level of expression was higher in Z. mobilis than in E. coli. The specific activity of the enzyme was comparable to that of the parent strain of Er. chrysanthemi. The proteins produced by Z. mobilis and Er. chrysanthemi presented identical immunological and electrophoretic properties. Biosynthesis of endoglucanase occurred during the exponential growth phase of Z. mobilis and about 35% of the enzyme was released into the medium in the absence of detectable cell lysis. The endoglucanase appeared to be located in the periplasmic space in Z. mobilis.

Nucleotide sequencing of the celZ gene encoding the extracellular endoglucanase Z of Erwinia chrysanthemi indicated the presence of an open reading frame encoding 428 amino acids. The mature protein appeared to be extended by a signal peptide of 43 amino acids; this sequence is unusually long and positively charged (+5). It was shown to function as a signal peptide by fusing it to a truncated phoA gene encoding Escherichia coli alkaline phosphatase. Comparison of the encoded sequence with those of the endoglucanases of Bacillus subtilis and alkalophilic Bacillus revealed the existence of a region of extensive homology occurring in all three proteins at about the same distance from the NH2-terminal end. These regions may be involved in substrate binding and/or catalytic sites.

The lysA gene encodes meso-diaminopimelate (DAP) decarboxylase (E.C.4.1.1.20), the last enzyme of the lysine biosynthetic pathway in bacteria. We have determined the nucleotide sequence of the lysA gene from Pseudomonas aeruginosa. Comparison of the deduced amino acid sequence of the lysA gene product revealed extensive similarity with the sequences of the functionally equivalent enzymes from Escherichia coli and Corynebacterium glutamicum. Even though both P. aeruginosa and E. coli are Gram-negative bacteria, sequence comparisons indicate a greater similarity between enzymes of P. aeruginosa and the Gram-positive bacterium C. glutamicum than between those of P. aeruginosa and E. coli enzymes. Comparison of DAP decarboxylase with protein sequences present in data bases revealed that bacterial DAP decarboxylases are homologous to mouse (Mus musculus) ornithine decarboxylase (E.C.4.1.1.17), the key enzyme in polyamine biosynthesis in mammals. On the other hand, no similarity was detected between DAP decarboxylases and other bacterial amino acid decarboxylases.

Besides lignin, plant cell walls are made of polysaccharides: pectin, cellulose and hemicellulose. Most members of the Erwinia chrysanthemi group are secreting hydrolytic enzymes such as pectinases, proteases, cellulases and phospholipase (Starr and Chatterjee). Soft-rot disease and plant tissue maceration probably result from concerted activity of some or all of these enzymes.

The broad host range vectors, pKT210, pKT212, pKT248, pKT240, pGSS33 were mobilized into Zymomonas mobilis with the help of conjugative plasmids belonging to various incompatibility groups. The vectors pKT210, pKT212, pKT248, pGSS33 were stably maintained in Zymomonas mobilis in contrast to the conjugative plasmids. Based on these results we suggest the potential usefulness of these vectors as cloning vehicles in Zymomonas mobilis.

The structural gene encoding the major endoglucanase (EGZ) of Erwinia chrysanthemi strain 3665 has been isolated by hybridization using a DNA fragment of E. chrysanthemi strain 3937 previously cloned in the λL47-1 vector and expressing an endoglucanase activity. Furthermore, genes homologous to celZ and celY of strain 3665 were also isolated and subcloned from strain 3937 into Escherichia coli. They were shown to encode enzymes similar to EGZ and EGY of strain 3665.

The structural gene coding for a new endo-beta-1,4-glucanase of Erwinia chrysanthemi strain 3665, previously identified in a cosmid library, was subcloned into pUC18. The gene is expressed from a 1.9 X 10(3)-base-pair insert and its direction of transcription was determined. The properties of the gene product purified from cell-free extracts of Escherichia coli have been studied. The purified protein has an endoglucanase activity but is significantly different from the major endoglucanase Z secreted by E. chrysanthemi strain 3665. The new enzyme was designated as endoglucanase Y and the related gene celY. In E. coli, most of the endoglucanase activity was found in the periplasmic space.