- Volume 148, Issue 7, 2002

Mycobacterium vaccae represents an alternative mycobacterial cloning host that has been largely overlooked to date. The main reason for this may be the reported non-transformability of this species, specifically the so-called Stanford strain (NCTC 11659), with expression vectors that use kanamycin resistance as a selection method. However, this strain can be transformed using hygromycin resistance as an alternative selectable phenotype. The present study has shown that in contrast to previous reports, M. vaccae (ATCC 15483) is capable of being transformed with a range of vectors encoding kanamycin resistance as the selectable marker. Thereafter, the expression of the lacZ reporter gene in M. vaccae, Mycobacterium bovis BCG and Mycobacterium smegmatis mc2155 was evaluated using a range of characterized mycobacterial promoter sequences (hsp60, hsp70, PAN, 18kDa and 16S rRNA) cloned in the same promoter probe vector. In general, the promoters showed similar levels of activity in the three species, demonstrating that existing expression systems can readily be employed with M. vaccae (ATCC 15483). This was further confirmed by the observation that M. vaccae was capable of stable, in vitro expression of recombinant S1 subunit of pertussis toxin at levels equivalent to those obtained with BCG and M. smegmatis. Analysis of structural and functional stability of a range of vectors demonstrated that the incidence of instability noted for M. vaccae was lower than that recorded for M. smegmatis. Taken together, the results indicate that M. vaccae is an additional cloning host which may prove useful for specific aspects of mycobacterial biology and provide increased flexibility to the field of recombinant protein technology for mycobacteria.

Despite the introduction of mass vaccination in 1953 in The Netherlands, pertussis is currently an endemic disease with regular epidemic outbreaks. Changes in the Bordetella pertussis population in the first 20 years after the introduction of vaccination were studied by indexing IS1002 fingerprint types, fimbrial serotypes and 15 genes encoding surface proteins. Three periods were compared, the pre-vaccination period (1949–1952) and two subsequent periods, 1953–1958 and 1965–1972. Except for fimbrial serotypes, no changes were observed in the B. pertussis population between the first two periods. Mortality decreased fivefold and 543-fold in the periods 1953–1958 and 1965–1972, respectively, compared to the pre-vaccination period. The largest decrease in mortality coincided with significant changes in the B. pertussis population with respect to the frequencies of fimbrial serotypes, fingerprint types and ptxS1 alleles. A new fingerprint type (ft29), associated with the novel ptxS1 allele ptxS1A was observed in 50% of the isolates in the period 1965–1972. Of the 15 investigated genes, only ptxS1 showed a mismatch between the vaccine strains and clinical isolates, suggesting that it may have played a role in driving the observed changes. It is proposed that, within 10–20 years after the introduction of mass vaccination, an adaptive response occurred consisting of clonal expansion of strains, which expressed a pertussis toxin variant distinct from the vaccine variants. This adaptation had very little, if any, effect on mortality, however.

The significance of charged residues for the target-cell binding, potency and specificity of pediocin-like bacteriocins has been studied by site-directed mutagenesis of sakacin P. Most of the charged residues are located in the N-terminal half, which is thought to mediate the initial binding of these bacteriocins to target cells through electrostatic interaction. All the mutated peptides in which the net positive charge was reduced by one (by replacing a charged residue with threonine) exhibited reduced binding to target cells and a 2–15-fold reduction in potency. The least deleterious of these mutations was the removal of the positive charge in position 8 (H8T). This mutation was, in fact, less deleterious than the conservative His to Lys mutation, indicating that the positive charge in position 8 per se is not of major importance. Somewhat more deleterious was the removal of positive charges at the N- and C-terminal ends (K1T, K43T). Most deleterious was the elimination of the positive charge at positions 11 and (but to a lesser extent) 12, demonstrating the importance of the cationic patch in the middle of the N-terminal half of pediocin-like bacteriocins. Mutated peptides in which the net positive charge was increased by one were also constructed. Some of these exhibited increased cell binding and a potency that was the same as (44K, i.e. an extra positive charge at the C-terminus), or somewhat greater (T20K) than, that of sakacin P, whereas others (0K, i.e. an extra positive charge at the N-terminus) had reduced potency. Sakacin P contains only one negatively charged residue (Asp17). This negative charge and its orientation in space were crucial for activity, since the Asp to Asn mutation and (especially) the conservative Asp to Glu mutation were deleterious. Mutations that made the peptide less cationic had, overall, less effect on the potency toward the Carnobacterium piscicola strain than on the potency toward the three other strains tested, whereas the opposite was the case for mutations that made the peptide more cationic. Thus, charged residues in the N-terminal half may – apparently via the initial electrostatic binding of the bacteriocin to target cells – influence the target-cell specificity.

The patC gene encoding the cystathionine β-lyase (CBL) of Lactobacillus delbrueckii subsp. bulgaricus NCDO 1489 was cloned and expressed in Escherichia coli. Overexpression of CBL complemented the methionine auxotrophy of an E. coli metC mutant, demonstrating in vivo that this enzyme functions as a CBL. However, PatC is distinguishable from the MetC CBLs by a low identity in amino acid sequence, a sensitivity to iodoacetic acid, greater thermostability and a lower substrate affinity. Homologues of patC were detected in the 13 Lb. delbrueckii strains studied, but only seven of them showed CBL activity. In constrast to CBL+ strains, all CBL-deficient strains analysed were auxotrophic for methionine. This supports the hypothesis that CBLs from lactobacilli are probably involved in methionine biosynthesis. Moreover, the results of this study suggest that post-transcriptional mechanisms account for the differences in CBL activities observed between strains of Lb. delbrueckii.

When a shotgun phage-display library of Staphylococcus aureus Newman was affinity selected (panned) against recombinant von Willebrand factor (vWf), a novel von Willebrand factor binding protein (vWbp) was found. Experimental data indicate that the interaction between vWbp and vWf is very specific and mediated by a region of 26 aa residues in the C-terminal part of vWbp. vWbp has an N-terminal secretory signal sequence but no cell wall anchoring motif, suggesting a soluble extracellular location. Mature vWbp could be purified from the culture supernatant and the identity of the protein was confirmed by N-terminal sequencing. vWbp migrates with an apparent molecular mass of 66 kDa and the deduced protein consists of 482 aa. The gene encoding vWbp, named vwb, was present in all S. aureus strains investigated.

The polysaccharide capsule of Streptococcus pneumoniae and several well-characterized virulence proteins are known to contribute to the pathogenesis of pneumococcal disease. However, there is a paucity of data on the expression of their respective genes in vivo. In this study, the relative abundance of the mRNA transcripts of the genes encoding pneumolysin (ply), pneumococcal surface protein A (pspA), pneumococcal surface antigen A (psaA) and choline-binding protein A (cbpA), and of the first gene of the capsular polysaccharide biosynthesis locus (cps2A), was measured in virulent type 2 pneumococci harvested from the blood of BALB/c mice at 12 h and 24 h following intraperitoneal infection. The mRNA levels were then compared, using relative quantitative RT-PCR, with those present in organisms grown in serum broth. The expression of ply was upregulated threefold at 12 h, and 10-fold at 24 h post-infection; the expression of pspA and psaA was upregulated threefold and fivefold, respectively, at 12 h post-infection. Interestingly, the expression of pspA was 36-fold higher at 24 h post-infection whereas the expression of cps2A was upregulated approximately fourfold at 12 and 24 h post-infection. However, cbpA mRNA levels remained comparable in vivo and in vitro. When organisms were grown in whole blood or THY broth, the relative expression of these genes in the two growth media also differed markedly. This work provides direct molecular evidence that known virulence-associated genes of S. pneumoniae are differentially expressed in vivo. Data on the relative expression of these genes in different growth media also suggests that the regulation of expression of these genes is highly complex and multifactorial.

The Mig protein of Streptococcus dysgalactiae is a type III immunoglobulin G (IgG)-binding protein, expressing IgG- and α2-macroglobulin (α2-M)-binding receptors. This study showed that the Mig protein also displays binding activities to bovine immunoglobulin A (B-IgA). Biotin-labelled bovine serum IgA bound immobilized recombinant Mig and α2-M receptors derived from Mig, as well as the native Mig extracted from the surface of S. dysgalactiae strain SDG8 and the α2-M receptor released from the isogenic mig mutant strain Mig8-Mt, as determined by Western blotting and ELISA. There was no B-IgA binding activity to the immobilized IgG receptor derived from Mig or the proteins in the culture supernatant from the mig mutant strain Mig7-Mt, in which expression of Mig or Mig-related peptides on the cell surface was completely abolished. In a reciprocal experiment, biotin-labelled Mig was found to bind immobilized bovine serum IgA but not human IgA (H-IgA). The binding of Mig to bovine serum IgA was competitively inhibited by unlabelled Mig, intact and truncated α2-M receptors, and bovine serum IgA, but not by the Mig-IgG receptor, H-IgA or B-IgG. The binding of Mig and partially purified bovine secretory IgA (B-sIgA) was also characterized by Western blotting. Membrane-immobilized B-sIgA did not react with the biotin-labelled Mig, whereas soluble B-sIgA showed binding activity to the immobilized α2-M receptor of Mig. It is therefore concluded that the 11 kDa N-terminal region of the α2-M receptor of the S. dysgalactiae Mig protein specifically binds soluble and immobilized bovine serum IgA, as well as soluble B-sIgA. This is believed to be the first report of a B-IgA-binding protein in S. dysgalactiae.

N-terminal lipidation is a major mechanism by which bacteria can tether proteins to membranes and one which is of particular importance to Gram-positive bacteria due to the absence of a retentive outer membrane. Lipidation is directed by the presence of a cysteine-containing ‘lipobox’ within the lipoprotein signal peptide sequence and this feature has greatly facilitated the identification of putative lipoproteins by gene sequence analysis. The properties of lipoprotein signal peptides have been described previously by the Prosite pattern PS00013. Here, a dataset of 33 experimentally verified Gram-positive bacterial lipoproteins (excluding those from Mollicutes) has been identified by an extensive literature review. The signal peptide features of these lipoproteins have been analysed to create a refined pattern, G+LPP, which is more specific for the identification of Gram-positive bacterial lipoproteins. The ability of this pattern to identify probable lipoprotein sequences is demonstrated by a search of the genome of Streptococcus pyogenes, in comparison with sequences identified using PS00013. Greater discrimination against likely false-positives was evident from the use of G+LPP compared with PS00013. These data confirm the likely abundance of lipoproteins in Gram-positive bacterial genomes, with at least 25 probable lipoproteins identified in S. pyogenes

Sequence homologies suggest that the Bacillus subtilis 168 tagO gene encodes UDP-N-acetylglucosamine:undecaprenyl-P N-acetylglucosaminyl 1-P transferase, the enzyme responsible for catalysing the first step in the synthesis of the teichoic acid linkage unit, i.e. the formation of undecaprenyl-PP-N-acetylglucosamine. Inhibition of tagO expression mediated by an IPTG-inducible Pspac promoter led to the development of a coccoid cell morphology, a feature characteristic of mutants blocked in teichoic acid synthesis. Indeed, analyses of the cell-wall phosphate content, as well as the incorporation of radioactively labelled precursors, revealed that the synthesis of poly(glycerol phosphate) and poly(glucosyl N-acetylgalactosamine 1-phosphate), the two strain 168 teichoic acids known to share the same linkage unit, was affected. Surprisingly, under phosphate limitation, deficiency of TagO precludes the synthesis of teichuronic acid, which is normally induced under these conditions. The regulatory region of tagO, containing two partly overlapping σA-controlled promoters, is similar to that of sigA, the gene encoding the major σ factor responsible for growth. Here, the authors discuss the possibility that TagO may represent a pivotal element in the multi-enzyme complexes responsible for the synthesis of anionic cell-wall polymers, and that it may play one of the key roles in balanced cell growth.

Bacillus cereus 569 (ATCC 10876) endospores germinate in response to inosine or L-alanine, the most rapid germination response being elicited by a combination of these germinants. The gerI operon has already been characterized as a homologue of the gerA spore-germination receptor family of operons found in all Bacillus spp. examined; the primary defect in gerI mutant spores is in the inosine germination response, although spores were also slower to germinate in L-alanine. Additional transposon-insertion mutants, from similar Tn917-LTV1 mutagenesis and enrichment experiments, now define two more operons, also members of the family of gerA homologues, important in L-alanine and inosine germination. Transposon insertions were identified in an alanine-specific germination locus, named gerL, which represents an operon of three genes, termed gerLA, gerLB and gerLC. By examining the residual germination response to L-alanine in gerI and gerL mutants, it was deduced that the GerL proteins contribute most strongly to the L-alanine germination response, and that the GerI proteins, required primarily in inosine germination, mediate only much slower germination responses to alanine. The L-alanine germination responses mediated by GerL and GerI proteins differ in their germination rates, temperature optima and germinant concentration dependence. The gerQ locus, again identified by transposon insertion, is a second inosine-related germinant-receptor operon. GerQ and GerI proteins are both required for the germination response to inosine as sole germinant, but GerQ has no role in L-alanine germination. Although near-identical homologues of gerI and gerL operons are evident in the Bacillus anthracis genome sequence, there is no evidence of a close homologue of gerQ.

Several Bacillus strains belonging to the B. subtilis/amyloliquefaciens group isolated from plant-pathogen-infested soil possess plant-growth-promoting activity [Krebs, B. et al. (1998) R26 J Plant Dis Prot 105, 181–197]. Three out of the four strains investigated were identified as B. amyloliquefaciens and were able to degrade extracellular phytate (myo-inositol hexakisphosphate). The highest extracellular phytase activity was detected in strain FZB45, and diluted culture filtrates of this strain stimulated growth of maize seedlings under phosphate limitation in the presence of phytate. The amino acid sequence deduced from the phytase phyA gene cloned from FZB45 displayed a high degree of similarity to known Bacillus phytases. Weak similarity between FZB45 phytase and B. subtilis alkaline phosphatase IV pointed to a possible common origin of these two enzymes. The recombinant protein expressed by B. subtilis MU331 displayed 3(1)-phytase activity yielding D/L-Ins(1,2,4,5,6)P5 as the first product of phytate hydrolysis. A phytase-negative mutant strain, FZB45/M2, whose phyA gene is disrupted, was generated by replacing the entire wild-type gene on the chromosome of FZB45 with a km::phyA fragment, and culture filtrates obtained from FZB45/M2 did not stimulate plant growth. In addition, the growth of maize seedlings was promoted in the presence of purified phytase and the absence of culture filtrate. These genetic and biochemical experiments provide strong evidence that phytase activity of B. amyloliquefaciens FZB45 is important for plant growth stimulation under phosphate limitation.

The DUP240 gene family of Saccharomyces cerevisiae is composed of 10 members. They encode proteins of about 240 amino acids which contain two predicted transmembrane domains. Database searches identified only one homologue in the closely related species Saccharomyces bayanus, indicating that the DUP240 genes encode proteins specific to Saccharomyces sensu stricto. The short-flanking homology PCR gene-replacement strategy with a variety of selective markers for replacements, and classical genetic methods, were used to generate strains deleted for all 10 DUP240 genes. All of the knock-out strains were viable and had similar growth kinetics to the wild-type. Two-hybrid screens, hSos1p fusions and GFP fusions were carried out; the results indicated that the Dup240 proteins are membrane associated, and that some of them are concentrated around the plasma membrane.

Glycosylphosphatidylinositol (GPI)-anchoring represents a mechanism for attaching proteins to the cell surface of all eukaryotic cells. Two localizations of GPI proteins have been observed in the yeasts Saccharomyces cerevisiae and Candida albicans: plasma membrane and cell wall. The signals and the mechanisms involved in this differential targeting are presently not well understood. Here several cell-wall-related phenotypes of a gpi7/las21 deletion are described, where GPI7/LAS21 encodes a GPI-anchor-modifying activity. In both organisms, the structure and composition of the cell wall was modified, with a clear increase in chitin deposition. Cell-wall-targeted proteins accumulated in the growth medium, whereas the protein content of the cell wall decreased significantly, suggesting inefficiency of the covalent linkage. The level of plasma-membrane-targeted GPI proteins was not affected. Sequence analyses revealed that gene families involved in the addition of phosphoethanolamines to the core GPI anchor are highly conserved between eukaryotes, with the exception of the Gpi7 family which seems to be fungus-specific. These data are compatible with the notion that the phosphoethanolamine added by Gpi7 protein to the GPI anchor is a key factor in the covalent linkage of cell-wall proteins to fungal cell-wall components.

A 0·5 kb fragment of Streptomyces venezuelae ISP5230 genomic DNA was amplified by PCR using primers based on consensus sequences of cysteine synthase isozyme A from bacteria. The deduced amino acid sequence of the PCR product resembled not only cysteine synthase sequences from prokaryotes and eukaryotes but also eukaryotic cystathionine β-synthase sequences. Probing an Str. venezuelae genomic library with the PCR product located a hybridizing colony from which pJV207 was isolated. Sequencing and analysis of the Str. venezuelae DNA insert in pJV207 detected two ORFs. The deduced amino acid sequence of ORF1 matched both cysteine synthase and cystathionine β-synthase sequences in GenBank, but its size favoured assignment as a cystathionine β-synthase. ORF2 in the pJV207 insert was unrelated in function to ORF1; in its sequence the deduced product resembled acetyl-CoA transferases, but disruption of the ORF did not cause a detectable phenotypic change. Disruption of ORF1 failed to elicit cysteine auxotrophy in wild-type Str. venezuelae, but in the cys-28 auxotroph VS263 it prevented restoration of prototrophy with homocysteine or methionine supplements. The change in phenotype implicated loss of the transsulfuration activity that in the wild-type converts these supplements to cysteine. This study concludes that disruption of ORF1 inactivates a cbs gene, the product of which participates in cysteine synthesis by transsulfuration. Enzyme assays of Str. venezuelae mycelial extracts confirmed the formation of cysteine by thiolation of O-acetylserine, providing the first unambiguous detection of this activity in a streptomycete. Enzyme assays also detected cystathionine γ-synthase, cystathionine β-lyase and cystathionine γ-lyase activity in the extracts and showed that the substrate for cystathionine γ-synthase was O-succinyl-homoserine. Based on assay results, the cys-28 mutation in Str. venezuelae VS263 does not inactivate the cysteine synthase gene but impairs expression in cultures grown in minimal medium.

Two pectin lyase genes, designated pnl-1 and pnl-2, were cloned from Colletotrichum gloeosporioides f. sp. malvae, a pathogen of round-leaved mallow (Malva pusilla). pnl-1 was isolated using cDNA from infected plant material; pnl-2 was isolated using cDNA from 3-day-old mycelia grown in mallow-cell-wall extract (MCWE) broth. pnl-1 is the first pectinase gene described thus far to encode a cellulose-binding domain (CBD), which is common in cellulases and xylanases, whereas pnl-2 encodes a pectin lyase that lacks a CBD. In pure culture, pnl-1 expression could be detected when purified pectin or glucose was the sole carbon source, but not when MCWE was the sole carbon source. The lack of pnl-1 expression appeared to be due to gene repression by some unknown factor(s) in the cell-wall extract. In contrast, expression of pnl-2 was detected in cultures when MCWE, but not when purified pectin or glucose, was the sole carbon source. In infected tissue, detection of pnl-1 expression by Northern-blot hybridization and by RT-PCR began with the onset of the necrotrophic phase of infection. Expression ofpnl-2 was not detectable by Northern-blot hybridization, but was observed byRT-PCR in both the biotrophic and necrotrophic phases of infection. The differences between pnl-1 and pnl-2 (i.e. pnl-1 encoding a CBD and differences in the expression patterns of both genes) may be related to the requirements of C. gloeosporioides f. sp. malvae to be able to grow in host tissue under the different conditions present during the biotrophic and necrotrophic phases of infection.

The major laccase isoenzyme LAP2 secreted by the white-rot basidiomycete Trametes pubescens in response to high copper concentrations was purified to apparent electrophoretic homogeneity using anion-exchange chromatography and gel filtration. The monomeric protein has a molecular mass of 65 kDa, of which 18% is glycosylation, and a pI value of 2·6. The pH optima of the laccase depend on the substrates oxidized and show bell-shaped pH activity profiles with an optimum of 3–4·5 for phenolic substrates such as 2,6-dimethoxyphenol or syringaldazine, while the non-phenolic substrates ABTS [2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid)] and ferrocyanide show a monotonic pH profile with a rate increasing with decreasing pH. The catalytic efficiencies k cat/K m determined for some of its substrates were 48×106, 47×106, 20×106 and 7×106 M−1 s−1 for ABTS, syringaldazine, ferrocyanide and oxygen, respectively. Furthermore, the gene lap2 encoding the purified laccase was cloned and its nucleotide sequence determined. The gene consists of 1997 bp, with the coding sequence interrupted by eight introns and flanked by an upstream region in which putative CAAT, TATA, MRE and CreA consensus sequences were identified. Based on Northern analysis containing total RNA from both induced and uninduced cultures, expression of lap2 is highly induced by copper, which is also corroborated by an increase in laccase activity in response to copper. A stimulating effect of various other heavy metal ions on laccase synthesis was also observed. In addition to induction, a second regulatory mechanism seems to be repression of lap2 transcription by glucose.

The gene encoding a haemagglutinin of H. paragallinarum, hagA, has been identified and the full-length nucleotide sequence determined. A ∼39 kDa protein, recognized by an anti-haemagglutinin monoclonal antibody, mAb4D, was purified from H. paragallinarum strain 0083 and the N-terminal sequence obtained. The full-length nucleotide sequence was obtained by inverse PCR and the deduced amino acid sequence of the protein encoded was shown to be similar to other outer-membrane proteins of closely related organisms in the HAP group (Haemophilus, Actinobacillus, Pasteurella), especially the P5 protein of Haemophilus influenzae. The hagA gene was cloned into a His-tag expression vector and overexpressed in Escherichia coli strain M15(pREP4). The identity of the purified recombinant protein as a H. paragallinarum haemagglutinin was confirmed by haemagglutination of chicken red blood cells and reactivity, in a Western blot, with the monoclonal antibody specific for the serovar A haemagglutinin.