- Volume 141, Issue 12, 1995

Unlike the CytA toxin from native Bacillus thuringiensis subsp. israelensis (Bti) crystals, the inclusions of cloned CytA produced by Bti IPS 78/11 in the presence of the 20 kDa ‘helper’ protein require a reducing agent in addition to a highly alkaline pH for complete solubilization. Activation of the solubilized CytA with a range of proteases produced 25-22 kDa products. SDS-PAGE analysis and N-terminal amino acid sequencing revealed that CytA was processed very similarly at both termini by proteinase K or by Anopheles or Culex gut extracts. Trypsin, by contrast, processed CytA predominantly at the N-terminus. In vitro cytolytic assays against Aedes aegypti cells, and haemolytic assays against rat erythrocytes, showed that CytA processed at both termini by proteinase K was the most active form. Thus CytA, like other Bt δ-endotoxins, is processed to a well-defined protease-resistant product and this processing enhances the toxicity in vitro and possibly in vivo.

A DNA sequence homologous to a Gas DNA probe, and the corresponding Gas protein (stimulatory a-subunit of GTP-binding protein) were detected in Phycomyces blakesleeanus. The protein was demonstrated in membrane fractions of dormant spores of this fungus using three different experimental approaches. Photoaffinity-labelling experiments with [a-32P]GTP of the membrane fraction revealed two bands, of 56 and 32 kDa. The 56 kDa GTP-binding protein was detected by this method in all the stages of early development and growth investigated. Also, a spore protein of 56 kDa was ADP-ribosylated by cholera toxin, and a 56 kDa protein was detected by Western blotting with a specific antibody against mammalian Gas. These results indicate that Gas (56 kDa) is present in dormant spores of P. blakesleeanus. Using the ADP-ribosylation and Western blotting assays, Gas was detected during all stages of spore germination before the hyphae became highly branched, but it was not detected in the branched hyphae that formed 18 h after the initiation of spore germination. Therefore, Gas is expressed differentially during Phycomyces development.

Siderophore-mediated iron uptake systems play a central role in the pathogenesis of infection for many bacterial pathogens. Campylobacter species are not thought to produce siderophores, yet they are able to utilize both ferrichrome and enterochelin as sources of iron. Part of an operon named ceuBCDE, encoding components of a periplasmic binding-protein-dependent transport (PBT) system for the uptake of a ferric siderophore from Campylobacter coli, was cloned directly into Escherichia coli using a plasmid rescue technique. Phenotypic and genetic analyses of this system showed it to comprise two hydrophobic integral membrane proteins, CeuB (35-5 kDa) and CeuC (34-8 kDa), which may form the cytoplasmic membrane permease, an ATP-binding protein, CeuD (28-8 kDa), and a periplasmic substrate-binding protein, CeuE (34-5 kDa). In vivo labelling studies using [3H]palmitate demonstrated that CeuE, the periplasmic binding protein, is expressed as a lipoprotein in C. coli, which is unusual for a Gram-negative PBT system. Mutants of C. coli, defective in components of the transport mechanism, were severely impaired in the ability to utilize enterochelin as an iron source suggesting that this siderophore is a substrate for the transport system. This is the first molecular characterization of a PBT system in Campylobacter species.

There are difficulties with all published models that attempt to explain how rod-shaped bacteria locate their midpoint in preparation for the next cell division. Many bacteria find their middle quite accurately. This evenness of partition has been measured cytologically and is implied by the persistence of synchrony under certain circumstances. Previously, a number of models for the control of cell division have been proposed based on aspects of molecular genetics, ultrastructure measurements, or biochemical kinetics. This paper points out that none of the models in spite of their quite different natures can explain the precision of the location of the division site. Here, the ‘Central Stress Model’ is proposed which depends on the partition of wall tension between the cytoplasmic membrane (CM) and the murein layer in such a way that the CM at the centre of the rod experiences a higher stress than near the poles and that this peak stress increases through the cell cycle. The model assumes that: (i) murein is not incorporated at an established pole but is incorporated diffusely over the sidewall and intensely at sites of cell constriction; (ii) CM is synthesized over the entire cell wall; (iii) the murein and CM layers are attached non-covalently to each other, and interact physically with each other; (iv) this differential location of synthesis leads to a ‘tug-of-war’ that creates differential stresses that peak at the cell centre. Because of the fluid nature of the phospholipid bilayer there is a flux of lipid from the established poles towards the cell centre as the murein sidewall elongates. The flux from the pole lowers the tension in the CM at the ends of the sidewalls and creates a peak tension in the centre. Cells also have a discontinuity in the stresses in the murein at the junction of the curved pole with the cylindrical region of the cell wall (a doubling of the hoop stress above the axial stress). Thus in addition to the midpoint of the cell, these junctions between the polar caps and the cylindrical part of the cell wall are characterized by an abrupt change in the surface stress and we suggest that this can trigger cell division at these junctions to form a chromosome-less minicell. Two other assumptions of the model are that the cell has a membrane-associated system to sense the stress, and to trigger cell division locally when a threshold has been reached. It is suggested that there is a special two-component sensory system responding to tension in the CM. As the cell cycle progresses, and the stress in the centre of the cell exceeds some threshold, a system molecule triggers a cell division event at that site. Like other two-component systems, the sensory component of this two-component system is assumed to be distributed all over the CM. It is also assumed that when the sensory component is triggered it also causes local changes that ensure that division occurs at that site. Consequently, this model can explain why sister cells have very nearly the same size (length, volume, or biomass) and why genes that control a mechanism that senses cell size and initiates cell division have never been identified because they may not exist.

The Vi antigen is a capsular polysaccharide expressed by Salmonella typhi, the agent of human typhoid fever. Expression of this antigen is controlled by the viaA and viaB chromosomal loci. The viaB locus is composed of 11 genes designated tviA-tviE (typhi Vi), vexA-vexE (Vi antigen export) and ORF11. We constructed S. typhi Ty2 strains carrying non-polar mutations in ten genes located at the viaB locus and examined the individual contribution of each gene to Vi phenotype. Phenotypes of the mutants and complementation experiments suggested that synthesis of Vi antigen monomer was catalysed by the TviB and TviC polypeptides. Subsequent polymerization of the polysaccharide might be catalysed by the TviE protein, but required functional TviD product. Proteins encoded by vexA, vexB and vexC directed transport of the polymer to the bacterial cell surface. Anchoring of the Vi antigen at the bacterial cell surface was dependent of the VexE protein. The TviA protein was not essential for Vi polymer synthesis. However, disruption of the tviA gene on S. typhi Ty2 chromosome strongly decreased expression of Vi antigen. This defect was fully complemented by providing tviA in trans on a recombinant plasmid. By using lacZ transcriptional fusions, it was shown that the TviA product positively regulated co-transcription of the tviA and tviB genes from a promoter located upstream of tviA. Moreover, we showed that a tviAB-lacZ fusion was not expressed in a viaA (rcsB) mutant of S. typhi. However, expression of the tviAB-lacZ fusion was restored in this viaA mutant either by the rcsB gene of Escherichia coli, or by the tviA gene of S. typhi when present in high copy number. This suggested that the tviA and viaA products could be involved in the same regulatory pathway modulating Vi antigen expression in S. typhi. Together these results demonstrated that proteins encoded by theviaB locus are not only involved in Vi polymer synthesis and translocation of the polysaccharide to the bacterial cell surface, but also in regulation of Vi antigen expression in S. typhi.

A gene (phoV) encoding an alkaline phosphatase fromSynechococcus sp. strain PCC 7942 was isolated by screening a plasmid gene bank for expression of alkaline phosphatase activity in Escherichia coli JM103. Two independent clones carrying the same alkaline-phosphatase-encoding gene were isolated. One of these clones (pKW1) was further analysed and the nucleotide sequence of a contiguous 3234 bp DNA fragment was determined. Two complete open reading frames (ORF1 and phoV) and an incomplete ORF3 were identified reading in the same direction. The deduced phoV gene product showed 34% identity to the alkaline phosphatase PhoA from Zymomonas mobilis, and the N-terminal part of the putative ORF3 protein exhibited 57% identity to a protein of unknown function from Frankia sp. Insertional inactivation of the Synechococcus PCC 7942 phoV gene failed, indicating an essential role for either the phoV or the ORF3 gene product. PhoV consists of 550 amino acid residues, resulting in a molecular mass of 61.3 kDa. To overexpress the Synechococcus PCC 7942 phoV gene in E. coli, plasmid pKW1 was transformed into a phoA mutant of E. coli (CC118). In E. coli strain CC118(pKW1) PhoV was expressed constitutively with high rates of activity, and was shown to be membrane associated in the periplasmic space. After partial purification of the recombinant PhoV, it was shown that, like other alkaline phosphatases, the Synechococcus PhoV had a broad pH optimum in the alkaline region and a broad substrate specificity for phosphomonoesters, required Zn2† for activity, and was inhibited by phosphate. In contrast to several other alkaline phosphatases, PhoV was inhibited by Mn2†. Due to the lack of a Synechococcus PCC 7942 phoV mutant strain, the function of PhoV remains uncertain. However, the present results show thatSynechococcus PCC 7942 has a second, probably phosphate-irrepressible, alkaline phosphatase (PhoV, 61.3 kDa) in addition to the phosphate-repressible enzyme (PhoA, 145 kDa) already described.

The ß-galactosidase of Lactobacillus sake DSM 20017 is encoded by two genes located on its chromosome. These genes designated lacL and lacM were cloned in Escherichia coli NM 554 on an 8.65 kbp Hindlll fragment inserted in vector pRB473. Deletion analysis of the originally cloned fragment revealed that both genes are required for the formation of a functional ß-galactosidase. lacL and lacM are transcribed as a single transcript of approximately 2.9 kbp starting 34 bp upstream of the translational start codon. The proteins derived from lacL and lacM share only 18-59% homology with other ß-galactosidases. The genes encoding the ß-galactosidase are scattered with multiple direct and inverted repeats of 9-12 bp. However, comparison with the plasmid-encoded Leuconostoc lactis ß-galactosidase revealed equal distribution of conserved amino acid residues and suggests that the genes have a common origin. Specific deletions or insertions resulting from the presence of the repeats were not observed. The L. sake ß-galactosidase was phenotypically expressed in E. coli NM 554 and Lactobacillus curvatus LTH 1432. Its two genes can be used to replace antibiotic reporter genes to develop food-grade vectors and á-complementation systems for self-cloning in meat lactobacilli.

The X-prolyl dipeptidyl aminopeptidase gene (pepX) of an industrially used Lactobacillus helveticus strain has been detected by nucleic acid hybridization, cloned, characterized and sequenced. One ORF of 2379 bp with coding capacity for a 90.6 kDa protein (PepX) was found. The ORF was preceded by a typical prokaryotic promoter region. An inverted repeat structure with δG of −84.1 kJ mol−1 was found downstream of the coding region. The deduced amino acid sequence of the 90.6 kDa protein showed 49.3, 49.4 and 77.7% homology with the PepX proteins from Lactococcus lactis subsp. lactis, Lc. lactis subsp. cremoris and Lactobacillus delbrueckii subsp. lactis, respectively. Northern blotting revealed a 2.6 kb transcript and one transcription start site was identified via primer extension analysis using an A.L.F. sequencer. In a bioreactor study, the expression of pepX in Lb. helveticus was studied as a function of growth. Transcription of pepX was typical of exponential growth phase expression. The pepX gene has been cloned into pKK223-3 and expressed at a high level in Escherichia coli JM105. PepX was purified to homogeneity by ion-exchange and hydrophobic interaction chromatography. Optimum PepX activity was observed at pH 6.5 and 45 °C. According to gel filtration analysis, PepX is a dimer of 165 kDa. The enzyme was inactivated by heavy metal ions such as Cu2+, Cd2+ and Zn2-. EDTA and 1,10-phenanthroline did not decrease PepX activity significantly. It was completely inhibited by p-hydroxy-mercuribenzoate and reactivated by adding DTT, and strongly inhibited by PMSF. PepX is thus a metal-independent serine peptidase having functional sulfhydryl groups at or near the active site.

The cattle protozoan parasite Tritrichomonas foetus has multiple forms of cysteine proteinases. To investigate their diversity, PCR and reverse transcriptase PCR were used to isolate genomic DNA and cDNA fragments, respectively, encoding different cysteine proteinases. Seven genes have been identified, TFCP3-6 from amplification of genomic DNA and TFCP7-9 from amplification of cDNA. Comparison of the predicted amino acid sequences indicates that the T. foetus enzymes are cathepsin-L-like rather than cathepsin-B-like in structure. However, there is considerable diversity among the proteinases. TFCP7 and TFCP8 are most similar to one another (78% identity), while TFCP3 and TFCP9 are the least closely related (30% identity). All but one of the genes are single-copy, the exception being TFCP3, which was present in multiple copies in one of the three isolates examined. Single transcripts were detected for each of the seven genes. TFCP8 was expressed at the highest levels, while transcripts for TFCP4 were only just detectable. In T. foetus F2, the strain from which the genomic DNA and mRNA were isolated, transcripts of the five other genes were present at intermediate levels. When two other isolates were compared with F2, differences in the expression of individual genes were apparent, with either one or two of them not expressed. In spite of these differences the major cysteine proteinases detected in the three isolates using substrate-SDS-PAGE appeared identical. The data show that the multiplicity of cysteine proteinases in T. foetus is due, in part at least, to the presence of multiple genes and that some of the genes encode cysteine proteinases which are not among the high-activity enzymes detected previously.

A PCR-based detection system was developed for Xanthomonas albilineans, a pathogen of sugarcane, and other related xanthomonads, using the conserved sequence of two adjacent tRNA genes and the variable length and sequence of the spacer region between them. An appropriate region was identified as follows: tRNA genes with the same anticodon from a wide variety of bacteria were aligned and the most frequent base at each position was chosen to derive primers that would anneal to the gene in either orientation. Pairs of such primers were screened against various Xanthomonas species and members of related genera using PCR at low to moderate annealing stringency. A subset of these pairs of tRNA consensus primers gave one or more PCR products which generally displayed interspecific length variability. The primer pair 5'-3' tRNAalaand 3'-5' tRNAileshowed interspecific length polymorphisms between X. albilineans and all other Xanthomonas species examined. These PCR products were cloned and sequenced from four isolates of X. albilineans and four isolates from different pathovars of X. campestris, and the spacer length variation confirmed. Specific tRNA gene primers were derived from the tRNA gene sequences. These primers yielded a PCR product of a characteristic length within most Xanthomonas species and pathovars tested. When a primer that projected from tRNAalainto the 3' end of the variable intergenic spacer was used with a tRNAile-specific primer, PCR was a very sensitive diagnostic test for X. albilineans-infected sugarcane and gave no product or only a faint product with other species of bacteria. The specificity of this PCR-based detection system was further enhanced by a nested PCR reaction that took advantage of the fact that the tRNAala-tRNAileregion was found to be embedded in a 16S rRNA-23S rRNA gene spacer. By amplifying the region between the 16S rRNA gene and tRNAileor between the tRNAalaand the 23S rRNA gene, the subsequent nested PCR product was shown to be X. albilineans-specific.

The bfeA (Bordetella ferric enterobactin) receptor gene was cloned from a Bordetella pertussis chromosomal library by using a screen in Escherichia coli to detect iron-repressed genes encoding exported proteins translationally fused to the E. coli phoA gene. The bfeA gene encoded a protein with a molecular mass of approximately 80 kDa and about 50% amino acid sequence identity to both the fepA- and pfe-encoded enterobactin receptors of E. coli and Pseudomonas aeruginosa, respectively. Enterobactin prepared from iron-starved E. coli cultures supported growth of B. pertussis and Bordetella bronchiseptica in the presence of the iron chelator ethylenediamine-di-(o-hydroxyphenylacetic acid) (EDDA). Expression of the bfeA gene was induced by low iron availability, and iron-regulated expression appeared to be dependent upon the presence of the sequence contained within 370 bp upstream of the bfeA structural gene. An internal fragment of the bfeA structural gene and flanking regions were shown by Southern analysis to be highly conserved among Bordetella species. Insertional inactivation of bfeA in both B. pertussis and B. bronchiseptica greatly impaired their ability to grow in the presence of enterobactin and EDDA. These findings suggest that enterobactin produced by other respiratory flora could aid in the colonization of the respiratory tract by Bordetella species.

Surface antigenic variation was investigated in Mycoplasma arthritidis, an agent that produces chronic arthritis in rats which shares several features with many mycoplasma-induced diseases and thus defines a well-characterized model system. Hyperimmune rabbit antisera (anti-ISR1, anti-PG6, anti-H606 and anti-158p10) to whole M. arthritidis organisms were used as immunological probes in Western immunoblots of four M. arthritidis prototype strains (ISR1, PG6, H606 and D263) and five rat-passaged substrains (ISR1p1, ISR1p7, ISR1p8, 158p10 and D263p1). Several prominent antigens were identified that varied in expression. By Triton X-114 phase fractionation and treatment of whole cells with trypsin and carboxypeptidase Y, these strain-variant antigens were shown to be integral membrane proteins with C-termini and portions of the polypeptide chains oriented outside the membrane. Western blot immunoscreening of a large number of randomly selected clonal isolates and well-established clonal lineages from stock cultures of M.arthritidis ISR1p7, 158p10, PG6 and H606 revealed an expanded repertoire of variant membrane proteins whose expression was subject to independent, reversible phase variation. Colony immunoblots of these clonal populations with a hyperimmune rabbit antiserum to a gel-purified variant membrane protein (P36) showed that this phase switching occurred at a high frequency (10-4to 10-2per generation). Detailed immunological and biochemical characterization of the phase-variant membrane proteins demonstrated that they are: (i) antigenically related or distinct; (ii) apparently specific to particular strain populations; (iii) proteins or lipoproteins; (iv) major immunogens of M. arthritidis, recognized by serum antibodies from convalescent rat; and (v) able to undergo variation in expression during in vivo passage. Thus, M. arthritidis possesses a complex system capable of creating large repertoires of cell surface phenotypes which may affect the multiple interactions of this organism with its host and dictate its potential as a successful infectious agent and pathogen.

Strains of Mycoplasma mycoides subsp. mycoides small colony (SC) type, the agent of contagious bovine pleuropneumonia (CBPP), were analysed with respect to the polymorphism of distribution of a newly discovered insertion element, IS1296, on the chromosome. Analysis of 64 strains isolated from Europe, Africa and Australia, including four vaccine strains and the type strain PG1, revealed ten different IS patterns, forming two main clusters. The European strains originated from outbreaks of CBPP in different countries, and from various other sources such as semen and preputial washings from cattle, lungs from goats and buffalo, and milk from sheep. They showed identical IS1296 patterns, except one strain which had an additional IS1296 element, but the pattern belonged to the same cluster. This shows that the strains from Europe form a clonal lineage. The strains originating from different geographical parts of the African continent and from Australia showed four closely related IS1296 patterns which belong to a separate cluster. This indicates that strains from Africa and Australia form a clonal lineage different from that of the European strains, suggesting that the sporadic cases of CBPP that have re-emerged in Europe almost 15 years after the last declared endemic case in 1967 arose from an established reservoir within Europe rather than being the result of repeated importation from Africa and Australia. While most strains from Africa and Australia had the same IS1296 pattern, all vaccine strains could be distinguished by an individual pattern. The type strain PG1 also had a particular IS1296 pattern which belongs to the cluster of the strains from Africa and Australia. The molecular definition of clonality of M. mycoides subsp. mycoides SC strains with IS1296 represents a rapid and reproducible method for subtyping and differentiation of vaccine strains. It permits at the present time the definition of two main clonal lines, one including the strains from the European continent and a second with strains from Africa and Australia.