- Volume 142, Issue 4, 1996

A prolinase (pepR) gene was cloned from an industrial Lactobacillus helveticus strain (53/7). Three clones, hybridizing with a gene probe specific for a peptidase shown to have activity against di- and tripeptides, were detected from a L. helveticus genomic library constructed in Escherichia coli . None of the three clones, however, showed enzyme activity against the di- or tripeptide substrates tested. One of the clones, carrying a vector with a 5.5 kb insert, was further characterized by DNA sequencing. The sequence analysis revealed the presence of two ORFs, ORF1 and ORF2 of 912 and 1602 bp, respectively, ORF2, located upstream of and in the opposite orientation to ORF1, had a promoter region overlapping that of ORF1. ORF1 had the capacity to encode a 35083 Da protein. When amplified by PCR, ORF1 with its control regions specified a 35 kDa protein in E. coli that was able to hydrolyse dipeptides, with highest activity against Pro-Leu, whereas from the tripeptides tested, only Leu-Leu-Leu was slowly degraded. By the substrate-specificity profile and protein homologies, the 35 kDa protein was identified as a prolinase. The activity of the cloned prolinase was inhibited by p-hydroxymercuribenzoate. Northern and primer-extension analyses of ORF1 revealed a 1.25 kb transcript and two adjacent transcription start sites, respectively, thus confirming the DNA sequence data. ORF2 had encoding capacity for a 59.5 kDa protein that showed significant homology to several members of the family of ABC transporters. Determination of the mRNA levels at different growth phases revealed that the pep gene and ORF2 are transcribed in L. helveticus at the exponential and stationary phases of growth, respectively. Furthermore, two ORF2 deletion constructs, carrying the intact pepR gene, showed that this upstream operon adversely affected PepR activity in E. coli, which explains the enzymic inactivity of the original clones.

Biochemical analysis of the soluble hydrogenase from the thermophilic organism Acetomicrobium flavidum revealed that the enzyme is an a2ß2 tetramer, with the a and ß subunits having a molecular mass of 50 kDa and 25 kDa, respectively. The most important biochemical properties of the enzyme are a specific activity of 77 μmol min-1 (mg protein)-1 a K m for methylviologen of 0.2 mM, a pH optimum of 7.5 and a T 50 of about 70 †C. In addition, the enzyme is remarkably stable to oxygen inactivation, retaining full activity after 24 h exposure to air. By using oligodeoxynucleotides designed on the basis of the N-terminal sequences of the two subunits, the corresponding genes have been isolated and sequenced. When compared to the other hydrogenases so far characterized, the A. flavidum hydrogenase appears to be a typical [Ni-Fe] enzyme. The hydrogenase was expressed in Escherichia coli at high levels in a soluble form but it was not active. The analysis of the recombinant large subunit showed that it was not post-translationally processed at its C-terminus.

The algJ gene from Azotobacter vinelandii was cloned using a labelled RNA probe representing the coding region of the algE gene from Pseudomonas aeruginosa. DNA sequencing revealed an ORF of 1452 bp encoding a protein of 484 amino acid residues with a calculated molecular mass of 54611 Da. An RNA probe corresponding to algE was also used for Southern hybridization of chromosomal DNA, which showed that algE-related DNA sequences are also present in the alginate-producing phytopathogen species Pseudomonas marginalis and Pseudomonas syringae pv. glycinea. The coding region of algJ was subcloned in the expression vector pT7-7, leading to a corresponding gene product with an apparent molecular mass of 54 kDa which could be identified in the outer membrane (OM) of Escherichia coli BL21(DE3). Additionally, a cross-reacting protein with the same molecular mass was also found in the OM of A. vinelandii using an anti-AlgE antiserum. The derived amino acid sequence of AlgJ shared approximately 52% identity with AlgE from P. aeruginosa. The hydrophilicity profile as well as the amphipathicity of regions in the amino acid sequence of AlgJ showed significant similarities to AlgE. Based on these data, a topological model of AlgJ was created with the aid of known structures of outer-membrane proteins. This model presents AlgJ as a ß-barrel containing 18 ß-strands inserted in the OM.

Southern hybridization analysis revealed that there were three rrn loci within the genome of Dichelobacter nodosus, the causative organism of ovine footrot. These loci (rrnA, rrnB and rrnC) were isolated on recombinant lambda clones, and comprised 16S, 23S and 5S rRNA genes closely linked in that order. Sequence and primer extension analysis revealed the presence of putative genes encoding tRNAlle and tRNAAla within the 16S-23S spacer region, as well as a number of potential regulatory features. These elements included a single promoter, which was mapped upstream of the 16S rRNA gene and which was similar to Escherichia coli consensus promoter sequences, an AT-rich upstream region, a GC-rich motif that may be involved in stringent control, leader and spacer antitermination sequences, sites for ribonuclease processing, and a putative factor-independent terminator sequence. Potential open reading frames (ORFs) were identified within the regions flanking the rrn loci, with identical copies of the 3′ terminal ORF present downstream of each rRNA operon. Determination of the complete sequence of the 5S rRNA gene, and derivation of the 5S rRNA secondary structure, further substantiated the 16S rRNA-based placement of D. nodosus within the gamma division of the Proteobacteria.

The sigma factor σ73 of the obligate intracytoplasmic bacterium Rickettsia prowazekii was overexpressed and purified from Escherichia coli . The rickettsial rpoD gene encoding σ73 was cloned into a Ndel-BamHI-cleaved pET15b vector under control of T7 transcription and translation signals. The recombinant plasmid encoded a 75 kDa fusion protein that was overproduced in E. coli BL21(DE3) and purified from inclusion bodies after solubilization with guanidine hydrochloride and using His. Bind metal chelation resin. The N-terminal His. Tag sequence of the 75 kDa fusion protein was removed by thrombin treatment to obtain R. prowazekii σ73T. The R. prowazekii σ73T as well as the 75 kDa fusion protein had the ability to bind to core DNA-dependent RNA polymerase of both R. prowazekii and E. coli and to stimulate their interaction with a rickettsial promoter.

Almost all clinical enterohaemorrhagic Escherichia coli (EHEC) O157: H7 isolates harbour a large virulence plasmid designated pO157. In this study, pO157 of EHEC O157:H7 reference strain EDL 933 was characterized at the molecular level. A restriction map was constructed by using seven restriction enzymes, with appropriate gel electrophoretic and hybridization methods. The molecular size of pO157 was determined to be 93.6 kb. By sequencing the DNA region extending in the 3′-direction of the previously described EHEC hlyC and hlyA genes, two further genes were discovered and analysed; these were termed EHEC hlyB and EHEC hlyD. The newly discovered genes together with the EHEC hlyC and hlyA genes constitute a typical RTX (Repeats in ToXin) determinant (EHEC hly operon) with the gene order CABD. The map position of the operon was determined by hybridization experiments. Analysis of a DNA fragment carrying the downstream flanking region of the EHEC hly operon revealed an open reading frame which was highly homologous to orf1 of RepFIB, a basic replicon of IncF plasmids. It was located close to the EHEC hly operon.

The usage of alternative synonymous codons in Mycobacterium tuberculosis (and M. bovis) genes has been investigated. This species is a member of the high-G + C Gram-positive bacteria, with a genomic G + C content around 65 mol%. This G + C-richness is reflected in a strong bias towards C- and Gending codons for every amino acid: overall, the G + C content at the third positions of codons is 83%. However, there is significant variation in codon usage patterns among genes, which appears to be associated with gene expression level. From the variation among genes, putative optimal codons were identified for 15 amino acids. The degree of bias towards optimal codons in an M. tuberculosis gene is correlated with that in homologues from Escherichia coli and Bacillus subtilis. The set of selectively favoured codons seems to be quite highly conserved between M. tuberculosis and another high-G + C Gram-positive bacterium, Corynebacterium glutamicum, even though the genome and overall codon usage of the latter are much less G + C-rich.

The CelF-encoding sequence was isolated from Clostridium cellulolyticum genomic DNA using the inverse PCR technique. The gene lies between cipC (the gene encoding the cellulosome scaffolding protein) and celC (coding for the endoglucanase C) in the large cel cluster of this mesophilic cellulolytic Clostridium species. Comparisons between the deduced amino acid sequence of the mature CelF (693 amino acids, molecular mass 77626) and those of other ß-glycanases showed that this enzyme belongs to the recently proposed family L of cellulases (family 48 of glycosyl hydrolases). The protein was overproduced in Escherichia coli using the T7 expression system. It formed both cytoplasmic and periplasmic inclusion bodies when induction was performed at 37 °. Surprisingly, the protein synthesized from the cytoplasmic production vector was degraded in the Ion protease-deficient strain BL21(DE3). The induction conditions were optimized with regard to the concentration of inductor, cell density, and temperature and time of induction in order to overproduce an active periplasmic protein (CelFp) which was both soluble and stable. It was collected using the osmotic shock method. The enzymic degradation of various cellulosic substrates by CelFp was studied. CelFp degraded swollen Avicel more efficiently than substituted soluble CM-cellulose or crystalline Avicel and was not active on xylan. Its activity is therefore quite different from that of endoglucanases, which are most active on CM-cellulose.

Using the lacZ operon fusion technique, the transcriptional control of the Acinetobacter calcoaceticus recA gene was studied. A low (approximately twofold) inductive capacity was observed for compounds that damage DNA and/or inhibit DNA replication, e.g. methyl methanesulfonate, mitomycin C, UV light and nalidixic acid. Induction of the recA gene by DNA damage was independent of functional RecA. The presence of the recA promoter region on a multicopy plasmid had the same effect on recA transcription as the presence of DNA-damaging agents. Thus, recA expression in A. calcoaceticus appears to be regulated in a novel fashion, possibly involving a non-LexA-like repressor. Regulation of the recA gene in A. calcoaceticus appears not to be part of a regulon responsible for competence for natural transformation: in cells exhibiting extremely low transformation frequencies, the level of transcription of the recA gene was found to be comparable to the level found in cells in the state of maximal competence.

The responses of the autochthonous soil and aquatic organism, Pseudomonas aeruginosa to UV radiation wavelengths (UVA, 320-400 nm, and UVB, 280-320 nm) has been investigated in this study. P. aeruginosa recA mutants were found to be more sensitive to both UVA and UVB radiation than were their isogenic RecA+ parents. Introduction of a low-copy-number plasmid containing the cloned wild-type P. aeruginosa recA gene restored UVA and UVB resistance to recA mutants. The concentration of RecA protein increased twofold 120 min after exposure to either UVA or UVB radiation, suggesting induction of expression of the recA gene by these wavelengths. In this study, we found that a functional RecA protein is required for activation of D3 prophage in lysogenic cells following exposure to UVB radiation. Prophage were not induced by exposure of their hosts to UVA radiation. Induction of damage-inducible (din) genes in response to UVA or UVB irradiation was also shown to be RecA dependent. These data indicate that the recA gene plays a role in the response of P. aeruginosa to exposure to wavelengths of UV radiation found in the solar spectrum.

Five spontaneous mutants of Streptomyces lividans TK64 resistant to 5 or 15 μM ethidium bromide (EB) were isolated, and the corresponding mutations were mapped to two different chromosomal locations. Both types of mutations conferred unselected resistance to several basic dyes and norfloxacin. The strain with the low-level resistance exhibited wild-type levels of EB uptake and energy-dependent efflux, and the resistance mechanism is unclear. The highly resistant mutants, which additionally were resistant to phosphonium ions, had a reduced accumulation and an increased efflux of EB, reminiscent of a mammalian multidrug resistance efflux pump.

Cryptococcus neoformans produces large amounts of the acyclic hexitol mannitol in culture and infected animals, but the functional and pathogenic significance of mannitol production by this fungus is not known. We exposed C. neoformans H99 (Cn H99) to UV irradiation (1 × LD50) and screened survivors for mannitol production. A mutant, Cn MLP (Mannitol Low Producer), synthesized less mannitol from glucose (2.7 vs 8.2 nmol per 108 cells min−1 at 37 °) and contained less intracellular mannitol (1 vs 11 μmol per 106 cells at 37 °) than did Cn H99. Cn MLP and Cn H99 were similar with respect to carbon assimilation patterns, rates of glucose consumption, growth rates at 30 °, urease and phenoloxidase activities, morphology, capsule formation, mating type, electrophoretic karyotype, rapid amplification of polymorphic DNA (RAPD) patterns and antifungal susceptibility. However, Cn MLP was more susceptible than was Cn H99 to growth inhibition and killing by heat and high NaCl concentrations. Also, the LD50 values in mice injected intravenously were 3.7 × 106 c.f.u. for Cn MLP compared to 6.9 × 102 c.f.u. for Cn H99. Moreover, 500 c.f.u. Cn H99 intravenously killed 12 of 12 mice by 60 d, whereas all mice given the same inoculum of Cn MLP survived. Classical genetic studies were undertaken to determine if these differences were due to a single mutation, but the basidiospores were nonviable. These results suggest that the abilities of C. neoformans to produce and accumulate mannitol may influence its tolerance to heat and osmotic stresses and its pathogenicity in mice.

Bacterial proteases may participate in the pathogenesis of periodontal diseases through their action on host proteins. In the present study, the ability of selected periodontopathogens, as well as two proteases isolated from Porphyromonas gingivalis and Treponema denticola, to degrade host protease inhibitors was evaluated. The activation of human plasminogen by the two bacterial proteases was also investigated. Proteolytic breakdown of host protease inhibitors (α-1-antitrypsin, antichymotrypsin, α2-macroglobulin, antithrombin III, antiplasmin and cystatin C) was evaluated by SDS-PAGE. The 80 kDa trypsin-like protease of P. gingivalis completely digested the six protease inhibitors under investigation, whereas the 95 kDa chymotrypsin-like protease of T. denticola was slightly less active, more particularly on α2-macroglobulin and cystatin C. When whole cells from a number of oral bacterial species were tested, the most significant degradation was obtained with P. gingivalis, T. denticola, Prevotella intermedia, Prevotella nigrescens and Capnocytophaga spp. Peptostreptococcus micros and Propionibacterium acnes had only some degradative activity on selected inhibitors, whereas three bacterial species, Actinobacillus actinomycetemcomitans, Bacteroides forsythus and Fusobacterium nucleatum, had no effect on the protease inhibitors. The 80 kDa protease of P. gingivalis demonstrated strong plasminogen activation, whereas no such activity was associated with the 95 kDa protease of T. denticola. This study indicates the high potential of some periodontal pathogens to destroy protease inhibitors and activate plasminogen. This may result in an uncontrolled degradation of periodontal tissues and a rapid progression of the disease.

To optimize bacteriocin production processes, the relationships between growth, bacteriocin production and factors affecting the occurrence and intensity of the activity peak during the growth cycle must be understood. Amylovorin L471, a bacteriocin produced by Lactobacillus amylovorus DCE 471, displays primary metabolite kinetics with a peak activity during the midexponential phase. Because of this growth association, only conditions favouring a drastic increase in biomass improve the volumetric bacteriocin titre. Specific bacteriocin production is enhanced under unfavourable growth conditions such as low temperatures (30°), and the presence of potentially toxic compounds such as ethanol (1.0%, v/v) and oxygen (80%, v/v, air saturation). Whereas volumetric biomass formation and growth-associated bacteriocin production are dependent on the amount of glucose and nitrogen supplied, slow growth rates stimulate specific bacteriocin production. Bacteriocin inactivation can be ascribed to protein aggregation and adsorption phenomena. It may be overcome by switching the pH to 2.0 during the fermentation run after having reached the peak activity. Thus, manipulation of the cell environment can stimulate bacteriocin production. The latter can be induced by unfavourable growth conditions, so-called stress factors. The specific growth rate seems to play an important role in the control of bacteriocin production.

HPr is a low-molecular-mass phosphocarrier protein of the bacterial phosphoenolpyruvate (PEP): sugar phosphotransferase system (PTS) found in the cytoplasm or associated with the inner surface of the cytoplasmic membrane. Treatment of Streptococcus suis cells with a Sorvall Omnimixer, a technique used to extract cell surface components, resulted in the extraction of a major protein with a molecular mass of 9 kDa. Several lines of evidence suggested that this protein was HPr: (i) the S. suis protein showed homology over the first 35 N-terminal amino acid residues with the HPrs of Streptococcus salivarius and Streptococcus mutans, including the signature sequence for the site of PEP-dependent phosphorylation; (ii) it cross-reacted with the S. salivarius anti-HPr antibody preparation; (iii) it could be phosphorylated by enzyme 1 at the expense of PEP, and by a membrane-associated kinase at the expense of ATP; and (iv) it possessed phosphocarrier activity when used as a source of HPr in an in vitro PTS assay. The data suggested that a portion of the cellular HPr is associated with the external cell surface in S. suis, a result that was confirmed by immunogold electron microscopy. The cellular HPr of S. suis consisted of two forms that could be distinguished by the presence or the absence of the N-terminal methionine. Amino acid sequence analysis indicated that the cell-surface-associated HPr of S. suis lacked the N-terminal methionine residue.

Pseudomonas aeruginosa M60, a mucoid strain, was grown in continuous culture (D 0-05 h−1) under ammonia limitation with glucose as the carbon source. Steady-state alginate production occurred for only 1-2 d under these conditions [q alginate 0.097 g alginate h−1 (g dry wt cells)−1], after which time the percentage of mucoid cells and the alginate concentration in the culture decreased in parallel and approached zero after approximately 10 d. These changes were accompanied by similar decreases in the activities of the alginate biosynthetic enzymes (represented by phosphomannomutase and GDP-mannose dehydrogenase) and by a large increase in the activity of the first enzyme of the ‘external’ non-phosphorylative pathway of glucose metabolism, glucose dehydrogenase. In contrast, the activities of other enzymes associated with this pathway (gluconate dehydrogenase, 2-ketogluconate kinase plus 2-ketogluconate-6-phosphate reductase) or with the ‘internal’ phosphorylative pathway of glucose metabolism (glucokinase and glucose-6-phosphate dehydrogenase) remained essentially unchanged. The loss of mucoidy and alginate production was accompanied by the appearance of low concentrations of intracellular polyhydroxyalkanoate (PHA) and of extracellular gluconate and 2-ketogluconate (partly at the expense of alginate production and partly as a result of increased glucose consumption). It is suggested that ammonia-limited, glucose-excess cultures of P. aeruginosa growing at low dilution rate are unable fully to regulate the rate at which glucose and/or its ‘external’ pathway metabolites are taken up by the cell, and therefore form copious amounts of alginate in order both to overcome the potentially deleterious osmotic effects of accumulating surplus intracellular metabolites and to consume the surplus ATP generated by the further oxidation of these metabolites. The loss of mucoidy invokes the use of an alternative, but analogous, strategy via which non-mucoid cells produce an osmotically inactive intracellular product (PHA) plus increased amounts of the extracellular metabolites gluconate and 2-ketogluconate via the low-energy-yielding and, under these conditions, largely dead-end ‘external’ metabolic pathway.

Actinobacillus actinomycetemcomitans, a Gram-negative periodontopathic bacterium, produces a leukotoxin belonging to the RTX family. The production of leukotoxin varies greatly among different strains of this species. In this paper the effects of growth rate and bicarbonate on the leukotoxin production by a toxin-production-variable strain (301-b) during growth in a chemostat were examined. When the bacterium was grown in anaerobic fructose-limited chemostat cultures (pH 7.0 and 37 °) at dilution rates (D) ranging from 0.04 to 0.20 h−1 in the absence and presence of 10 mM bicarbonate, it produced leukotoxin as a cluster of two polypeptides (M r 113000 and 120000) and complexed with nucleic acids on the bacterial cell surface. The relationship between leukotoxin production and specific growth rate was analysed by plotting the specific rate of leukotoxin production [q LT, in μg (mg dry wt)−1 h−1] against D. The plots were approximated to the linear relationships q LT 2.7D − 0.058 and q LT − 9.3D − 0.407 without and with bicarbonate, respectively. These relationships suggest that the apparent leukotoxin production is a result of both growth-rate-dependent production and growth-rate-independent decomposition. The cellular leukotoxin level was also followed after the change from chemostat to batch culture in the same fermenter. In batch culture leukotoxin production stopped immediately and the cellular toxin level rapidly decreased, suggesting toxin decomposition. From the slopes of the approximated linear relationships between q LT and D, a theoretical maximum leukotoxin yield (Y LT) was estimated as 2.7 and 9.3 μg (mg dry wt)−1 in the absence and presence of 10 mM bicarbonate, respectively. The increased Y LT value in the cultures containing bicarbonate indicated that the addition stimulated the efficiency of leukotoxin synthesis up to about threefold. Further increases of bicarbonate concentration to between 20 and 40 mM had no effect on the total leukotoxin production, but the amount of extracellular leukotoxin increased with higher bicarbonate concentrations.