- Volume 147, Issue 8, 2001

The Escherichia coli glycine cleavage enzyme system, encoded by the gcvTHP operon, catalyses the oxidative cleavage of glycine to CO2, NH3 and a one-carbon methylene group. Transcription of the gcv operon is positively regulated by GcvA and negatively regulated by GcvA and GcvR. Using a LexA-based system for analysing protein heterodimerization, it is shown that GcvR interacts directly with GcvA in vivo to repress gcvTHP expression. Several mutations in either gcvA or gcvR that result in a loss of gcv repression also result in a loss of GcvA/GcvR heterodimerization. Finally, it is shown that the C-terminal half of GcvA is involved in its interaction with GcvR, whilst the entire GcvR protein appears to be necessary for heterodimerization.

Lipoamide dehydrogenase (LPD) is an essential component of the pyruvate dehydrogenase and 2-oxoglutarate dehydrogenase complexes, both playing a crucial role within the central metabolism of aerobic organisms. Using oligonucleotides designed according to conserved regions of LPD amino acid sequences from several organisms, the lpd gene from Corynebacterium glutamicum was identified and subsequently subcloned. The cloned lpd gene expressed in C. glutamicum cells harbouring the gene on a plasmid showed a 12-fold higher specific LPD activity when compared to the wild-type strain. DNA sequence analysis of a 4524 bp segment containing the lpd gene and adjacent regions revealed that the lpd gene is not flanked by genes encoding other subunits of the pyruvate or 2-oxoglutarate dehydrogenase complexes and predicted an LPD polypeptide of 469 amino acids with an M r of 50619. The amino acid sequence of this polypeptide shows between 26 and 58% identity when compared to LPD enzymes from other organisms. Transcriptional analyses revealed that the lpd gene from C. glutamicum is monocistronic (1·45 kb mRNA) and that its transcription is initiated exactly at the nucleotide defined as the translational start. LPD was purified and biochemically characterized. This analysis revealed that the enzyme catalyses the reversible reoxidation of dihydrolipoic acid and NADH:NAD+ transhydrogenation, and is able to transfer electrons from NADH to various redox-active compounds and quinones. An in vivo participation of C. glutamicum LPD in facilitation of quinone redox cycling is proposed.

Alternative sigma factors have been detected in the myxobacterium Stigmatella aurantiaca during indole-induced sporulation, fruiting body formation and heat shock using an antiserum raised against sigma factor SigB. The time course of sigB gene expression was analysed by RT-PCR and by determining β-galactosidase activity during development in a merodiploid strain that harboured a sigB–lacZ fusion gene. Inactivation of the sigB gene by insertion of the neo gene resulted in the loss of one sigma factor as shown by Western analysis. Neither fruiting body formation nor sporulation, nor the production of possible SigB targets, such as DnaK, GroEL or HspA, were affected.

Identifying genes that are differentially expressed by Mycobacterium bovis BCG after phagocytosis by macrophages will facilitate the understanding of the molecular mechanisms of host cell–intracellular pathogen interactions. To identify such genes a cDNA–total RNA subtractive hybridization strategy has been used that circumvents the problems both of limited availability of bacterial RNA from models of infection and the high rRNA backgrounds in total bacterial RNA. The subtraction products were used to screen a high-density gridded Mycobacterium tuberculosis genomic library. Sequence data were obtained from 19 differential clones, five of which contained overlapping sequences for the gene encoding mycocerosic acid synthase (mas). Mas is an enzyme involved in the synthesis of multi-methylated long-chain fatty acids that are part of phthiocerol dimycocerosate, a major component of the complex mycobacterial cell wall. Northern blotting and primer extension data confirmed up-regulation of mas in intracellular mycobacteria and also revealed a putative extended −10 promoter structure and a long untranslated upstream region 5′ of the mas transcripts, containing predicted double-stranded structures. Furthermore, clones containing overlapping sequences for furB, groEL-2, rplE and fadD28 were identified and the up-regulation of these genes was confirmed by Northern blot analysis. The cDNA–RNA subtractive hybridization enrichment and high density gridded library screening, combined with selective extraction of bacterial mRNA represents a valuable approach to the identification of genes expressed during intra-macrophage residence for bacteria such as M. bovis BCG and the pathogenic mycobacterium, M. tuberculosis.

Pathogenesis of Mycobacterium tuberculosis is closely connected to its survival and replication within the host. Some pathogenic bacteria employ protein kinases that interfere with the cellular signalling network of host cells and promote bacterial survival. In this study, the pknF and pknG genes, which encode two putative protein kinases of M. tuberculosis H37Rv, protein kinase F (PknF) and protein kinase G (PknG), respectively, were cloned and expressed in Escherichia coli. Purified PknF phosphorylated the peptide substrate myelin basic protein (MBP) at serine and threonine residues, while purified PknG phosphorylated only at serine residues. The activity of the two kinases was abrogated by mutation of the codon for the predicted ATP-binding-site lysine residue. Southern blot analysis revealed that homologues of the genes encoding the two kinases are present in M. tuberculosis H37Ra and Mycobacterium bovis BCG, but not in Mycobacterium smegmatis. Immunoblot analysis of various cellular fractions of M. tuberculosis H37Rv revealed that PknF is a transmembrane protein and that PknG is predominantly a cytosolic enzyme. The present study should aid in elucidating the role of these protein kinases in the pathogenesis of mycobacteria.

Helicobacter pylori is unique among bacterial pathogens in its ability to persist in the acidic environment of the human stomach. To identify H. pylori genes responsive to low pH, the authors assembled a high-density array of PCR-amplified random genomic DNA. Hybridization of radiolabelled cDNA probes, prepared using total RNA from bacteria exposed to buffer at either pH 4·0 or pH 7·0, allowed both qualitative and quantitative information on differential gene expression to be obtained. A previously described low-pH-induced gene, cagA, was identified together with several novel genes that may have relevance to the survival and persistence of H. pylori in the gastric environment. These include genes encoding enzymes involved in LPS and phospholipid synthesis and secF, encoding a component of the protein export machinery. A hypothetical protein unique to H. pylori (HP0681) was also found to be acid induced. Genes down-regulated at pH 4·0 include those encoding a sugar nucleotide biosynthesis protein, a flagellar protein and an outer-membrane protein. Differential gene expression was confirmed by total RNA slot-blot hybridization.

Erp (exported repeated protein) was originally characterized as a virulence factor in Mycobacterium tuberculosis and was thought to be present only in Mycobacterium leprae and members of the TB complex. Here it is shown that Erp is a ubiquitous extracellular protein found in all of the mycobacterial species tested. Erp proteins have a modular organization and contain three domains: a highly conserved amino-terminal domain which includes a signal sequence, a central variable region containing repeats based on the motif PGLTS, and a conserved carboxy-terminal domain rich in proline and alanine. The number and fidelity of PGLTS repeats of the central region differ considerably between mycobacterial species. This region is, however, identical in all of the clinical M. tuberculosis strains tested. In addition, it is shown here that a Mycobacterium smegmatis erp::aph mutant displays altered colony morphology which is complemented by all the Erp orthologues tested. The genome sequence flanking the erp gene includes cell-wall-related ORFs and displays extensive conservation between saprophytic and pathogenic mycobacteria.

Previously, a complete genome analysis of Neisseria meningitidis strain MC58 revealed the largest repertoire of putative phase-variable genes described in any species to date. Initial comparisons with two incomplete Neisseria spp. genome sequences available at that time revealed differences in the repeats associated with these genes in the form of polymorphisms, the absence of the potentially unstable elements in some alleles, and in the repertoire of the genes that were present. Analyses of the complete genomes of N. meningitidis strain Z2491 and Neisseria gonorrhoeae strain FA1090 have been performed and are combined with a comprehensive comparative analysis between the three available complete genome sequences. This has increased the sensitivity of these searches and provided additional contextual information that facilitates the interpretation of the functional consequences of repeat instability. This analysis identified: (i) 68 phase-variable gene candidates in N. meningitidis strain Z2491, rather than the 27 previously reported; (ii) 83 candidates in N. gonorrhoeae strain FA1090; and (iii) 82 candidates in N. meningitidis strain MC58, including an additional 19 identified through cross-comparisons with the other two strains. In addition to the 18 members of the opa gene family, a repertoire of 119 putative phase-variable genes is described, indicating a huge potential for diversification mediated by this mechanism of gene switching in these species that is central to their interactions with the host and environmental transitions. Eighty-two of these are either known (14) or strong (68) candidates for phase variation, which together with the opa genes make a total of 100 identified genes. The repertoires of the genes identified in this analysis diverge from the different species groupings, indicating horizontal exchange that significantly affects the species and strain complements of these genes.

Pilus-mediated adherence to mucosal epithelial cells is a critical step for Neisseria gonorrhoeae to establish an infection in the human host. CD46, the defined receptor for the gonococcal pilus, is a complement-regulatory protein that is expressed on all human nucleated cells. It was observed that a piliated, Opa− variant of gonococcal strain FA1090 adhered with different efficiencies to the human epithelial cell lines tested (Chang, ME180, HEC-1B and PC-3). Surprisingly, these differences in adherence levels did not correlate with levels of CD46 expressed by these cell lines. In fact, there was an inverse relationship between total surface-exposed CD46 and gonococcal adherence. Four major isoforms of CD46 are produced due to alternative RNA splicing of a surface-exposed region and the cytoplasmic tail. The relative isoform surface expression of each cell line was determined, and each was found to express different ratios of the four CD46 isoforms. No correlation could be derived between CD46 isoform surface expression and pilus-mediated gonococcal adherence, indicating that CD46 does not act as a classic receptor for gonococcal pili.

Attaching and effacing Escherichia coli (AEEC) strains isolated from diarrhoeic lambs and goat kids were characterized for intimin (eae) and EspB (espB) gene subtypes by PCR and sequencing, and for genetic relatedness by PFGE. Fifty (23 ovine and 27 caprine) AEEC strains of 398 (246 ovine and 152 caprine) analysed were detected by colony blot hybridization. These strains were epidemiologically unrelated since they were isolated from different outbreaks of neonatal diarrhoea over a long period. Ovine AEEC strains belonged to serogroups O2, O4, O26, O80, O91 or were untypable, and caprine strains belonged to serogroups O3, O153 and O163. Two intimin subtypes were detected among the ovine and caprine strains studied. Most of the strains (43/50) had the β type intimin gene, but seven ovine strains possessed a variant γ type intimin gene (γV). Analysis of deduced amino acid sequences of the eae gene revealed that the sequences of β intimin of ovine and caprine strains were virtually identical to those of β intimin of rabbit EPEC, human EPEC clone 2 and swine AEEC, whereas the γV intimin present in seven ovine strains had 75–76% identity with γ intimin of human EHEC clone 1 strains, and 96% of identity with intimin of the human EHEC strain 95NR1 of serotype O111:H–. A PCR test was developed to identify the three different espB gene subtypes, espB of human EPEC clone 1 (espBα), espB of human EHEC clone 1 (espBγ) and espB of rabbit EPEC and human EPEC clone 2 (espBβ). There was close correlation between the intimin β type and the espBβ gene subtype in the ovine and caprine AEEC strains. The seven ovine strains possessing the γV intimin gene possessed the espBα gene subtype. None of the strains studied possessed the espBγ gene found in human O157:H7 EHEC strains. PFGE analysis of genomic DNA of selected strains showed a great diversity among strains. Cluster analysis of PFGE patterns showed greater divergence between strains with the γV intimin gene than between strains with the β intimin gene. This study showed that most of the AEEC strains isolated from diarrhoeic lambs and goat kids possessed β intimin and espB genes identical to those of rabbit EPEC, and they may be associated with enteric disease in small ruminants.

The pathogenesis of Cryptococcus neoformans infection has been studied extensively with respect to inflammatory and pathological changes, but very little information is available regarding the morphology of yeast cells during the course of infection. Electron microscopy of Cryptococcus neoformans in murine pulmonary infection revealed increased cell wall thickness with time, but this difference was only partially accounted for by increases in cell diameter. Cell walls of melanized cells were thicker than those of nonmelanized cells 2 h after infection, and the cell wall of yeast became blacker with time, suggesting that melanization contributes to the increased cell wall thickness. Heterogeneous cell populations emerged, with the appearance of giant forms. While for C. neoformans ATCC strain 24067 (serotype D) the full spectrum of cell sizes were observed, for strains H99 (serotype A) and 3501 (serotype D) cells were divisible into two populations, giant and micro forms. In contrast to cellular heterogeneity, the epitope recognized by a protective mAb on the capsular glucuronoxylomannan (GXM) was found at all times of infection. Immunoelectron microscopy using mAbs to GXM demonstrated reactivity with intracellular structures, suggesting that synthesis of capsular polysaccharide occurs, at least in part, in the cytoplasm. In summary, the results indicate that: (i) the infection is dynamic with respect to yeast cell morphology; (ii) giant cell forms arise in tissue during the course of infection; (iii) cell walls blacken and thicken during the course of infection, consistent with melanin synthesis during infection; and (iv) GXM epitopes are found in the capsule, cell wall and cytoplasm, consistent with intracellular polysaccharide synthesis. The results indicate that the population of C. neoformans cells in tissue is in a highly dynamic state, implying that the immune system must confront cells with varying characteristics during the course of infection.

Lipooligosaccharide (LOS) is a critical virulence factor of Neisseria meningitidis. A Tn916 insertion mutant, designated 469, was found to exhibit a markedly truncated LOS of 2·9 kDa when compared by Tricine/SDS-PAGE to the parental LOS (4·6 kDa). Electrospray mass spectrometry analysis of 469 LOS revealed that it consisted of the deep rough, heptose-deficient structure, Kdo2-lipid A. Sequencing of chromosomal DNA flanking the Tn916 insertion in mutant 469 revealed that the transposon had inserted into an ORF predicted to encode a 187 aa protein with sequence homology to the histidinol-phosphate phosphatase domain of Escherichia coli HisB and to a family of genes of unknown function. The gene, designated gmhX, is part of a polycistronic operon (ice-2) containing two other genes, nlaB and orfC. nlaB encodes a lysophosphatidic-acid acyltransferase and orfC is predicted to encode a N-acetyltransferase. Specific polar and non-polar gmhX mutations in the parental strain, NMB, exhibited the truncated LOS structure of mutant 469, and repair of gmhX mutants by homologous recombination with the wild-type gmhX restored the LOS parental phenotype. GmhX mutants demonstrated increased sensitivity to polymyxin B. GmhX mutants and other Kdo2-lipid A mutants also demonstrated increased sensitivity to killing by normal human serum but were not as sensitive as inner-core mutants containing heptose. In the genomes of Helicobacter pylori and Synechocystis, gmhX homologues are associated with heptose biosynthesis genes; however, in N. meningitidis, gmhX was found in a location distinct from that of gmhA, rfaD, rfaE, aut and rfaC. GmhX is a novel enzyme required for the incorporation of L-glycero-D-manno-heptose into meningococcal LOS, and is a candidate for the 2-D-glycero-manno-heptose phosphatase of the heptose biosynthesis pathway.

Azorhizobium caulinodans mutant 62004 carries a null allele of pdhB, encoding the E1β subunit of pyruvate dehydrogenase, which converts pyruvate to acetyl-CoA. This pdhB mutant completely lacks pyruvate oxidation activities yet grows aerobically on C4 dicarboxylates (succinate, L-malate) as sole energy source, albeit slowly, and displays pleiotropic growth defects consistent with physiological acetyl-CoA limitation. Temperature-sensitive (ts), conditional-lethal derivatives of the pdhB mutant lack (methyl)malonate semialdehyde dehydrogenase activity, which thus also allows L-malate conversion to acetyl-CoA. The pdhB mutant remains able to fix N2 in aerobic culture, but is unable to fix N2 in symbiosis with host Sesbania rostrata plants and cannot grow microaerobically. In culture, A. caulinodans wild-type can use acetate, β-D-hydroxybutyrate and nicotinate – all direct precursors of acetyl-CoA – as sole C and energy source for aerobic, but not microaerobic growth. Paradoxically, acetyl-CoA is thus a required intermediate for microaerobic oxidative energy transduction while not itself oxidized. Accordingly, A. caulinodans energy transduction under aerobic and microaerobic conditions is qualitatively different.

The effect of acetylene (14C2H2) on aerobic and anaerobic ammonia oxidation by Nitrosomonas eutropha was investigated. Ammonia monooxygenase (AMO) was inhibited and a 27 kDa polypeptide (AmoA) was labelled during aerobic ammonia oxidation. In contrast, anaerobic, NO2-dependent ammonia oxidation (NO2/N2O4 as oxidant) was not affected by acetylene. Further studies gave evidence that the inhibition as well as the labelling reaction were O2-dependent. Cells pretreated with acetylene under oxic conditions were unable to oxidize ammonia with O2 as oxidant. After these cell suspensions were supplemented with gaseous NO2, ammonia oxidation activity of about 140 μmol NH4 + (g protein)−1 h−1 was detectable under both oxic and anoxic conditions. A significantly reduced acetylene inhibition of the ammonia oxidation activity was observed for cells incubated in the presence of NO. This suggests that NO and acetylene compete for the same binding site on AMO. On the basis of these results a new hypothetical model of ammonia oxidation by N. eutropha was developed.

The fundamental question in this study is concerned with whether the increase of unsaturated fatty acids in the cell membrane is a general response of certain thermotolerant strains or species when exposed to superoptimal temperatures, and in combination with other stresses, especially oxidative stress. A strain of Lactobacillus helveticus, a species widely used as a starter in the dairy industry and able to tolerate high temperature and NaCl concentrations as well as acidic conditions, was chosen for this study. Cells of strain CNBL 1156, grown in its natural medium (i.e. milk whey), were exposed for 100 min to sublethal combinations of temperature, NaCl, H2O2 and pH, modulated according to a Central Composite Design. The fatty acid composition of cell lipid extract was identified by GC/MS. Polynomial equations, able to describe the individual interactive and quadratic effects of the independent variables on cell fatty acid composition, were obtained. The results and the mathematical models relative to the individual fatty acids indirectly suggest that desaturase activation or hyperinduction play an important role in the response to heat stress. In fact, the relative proportions of oleic, linoleic and palmitic acids increased with temperature in a range between 38 and 54 °C. The fatty acid profiles included vernolic acid (up to 37% of total fatty acids), an epoxide of linoleic acid not previously reported in microbial cells. In particular, this epoxide was present in cells exposed to low pH in combination with high temperatures and oxidative stress. In conclusion, these results provide experimental support to the hypothesis that the increase of an oxygen-consuming desaturase system, with a consequent increase in fatty acid desaturation, is a cellular response to environmental stresses able to protect the cells of this anaerobic micro-organism from toxic oxygen species and high temperatures.

The starvation survival response (SSR) of Listeria monocytogenes EGD is induced under glucose- or multiple-nutrient-, but not amino-acid limitation. 0·01–0·2% of the population remain viable even after 20 d and the survivors show a reduced cell size and increased cross-protection to several environmental stresses. The development of the SSR may therefore be important in L. monocytogenes survival in the food environment. The initiation, but not the maintenance, of the SSR involves both protein and cell wall biosynthesis. It is also likely that nutrients released from dead cells are recycled to allow survival of the remaining population. To define the molecular mechanisms involved in the initiation, maintenance and release from the SSR the role of known, and novel, components was examined. The well-characterized regulators SigB and PrfA are both required for the full SSR and effect stress resistance during growth and starvation. A transposon mutagenesis screen identified two novel loci with roles in the SSR and stress resistance. Characterization of the transposon insertion sites revealed a putative homologue of the gene yulB from Bacillus subtilis and a gene of unknown function. The potential individual and combined roles of the SSR components are discussed.

Among species of the heterobasidiomycetous yeasts, Filobasidiella neoformans is the only serious pathogen that causes fatal infections in both immunocompromised as well as immunocompetent patients. Three phenotypic characteristics, including growth at 37 °C, extracellular polysaccharide capsule and laccase activity, of F. neoformans are known to play major roles in the pathogenicity of the fungus. Several CAP genes involved in polysaccharide capsule formation, as well as the CNLAC1 gene encoding a laccase, have previously been cloned and characterized. To analyse the presence of these Cryptococcus neoformans virulence factors in other heterobasidiomycetous yeasts, numerous species of heterobasidiomycetous yeasts were screened for the presence of laccase activity and a polysaccharide capsule. Species exhibiting laccase activity and possessing a glucuronoxylomannan (GXM) capsule were screened for homologues of both the CAP59 gene and the CNLAC1 gene of F. neoformans. Southern blots of genomic DNA from GXM capsule-producing species exhibited no discernible hybridization to the CAP59 DNA sequence except for the two varieties of F. neoformans and Cryptococcus podzolicus. Although discernible, the hybridization band observed with the DNA of C. podzolicus was faint. Oligonucleotide primers constructed using the CAP59 gene sequence also failed to yield PCR products from DNAs of these yeasts except for the two varieties of F. neoformans. These results, coupled with the absence of a CAP59 homologue in the database, suggested the CAP59 gene to be unique to F. neoformans. C. podzolicus was the only species besides F. neoformans that possessed a capsule and expressed strong laccase activity on various media containing phenolic compounds. A CNLAC1 homologue was isolated from C. podzolicus while it was not detected in the species producing beige to faint tan colonies on media with phenolic compounds. Compared to the CNLAC1 sequence of four serotypes of F. neoformans, the CNLAC1 homologue of C. podzolicus showed the highest homology to that of serotype B/C strains and the lowest homology to that of serotype A strains.

Quorum sensing is a widespread form of bacterial communication in which individual cells produce and respond to specific N-acyl homoserine lactone signal metabolites. The different autoinducer synthases that generate these signals and the receptor/activator proteins that mediate the cell’s response to them constitute evolutionarily conserved families of regulatory proteins known as the LuxI and LuxR families, respectively. We have performed a phylogenetic analysis of 76 individual LuxI and LuxR homologues present in diverse members of the Gram-negative Proteobacteria. The results were consistent with an early origin for these regulators during the evolution of the Proteobacteria, with functional pairs of luxI and luxR genes possibly coevolving as regulatory cassettes. In many cases, specific LuxI and LuxR family members appeared to have been inherited horizontally. In particular, those species containing multiple LuxI and/or LuxR homologues usually appeared to have obtained each individual homologue or functional pair of homologues from an independent source. Because multiple homologues interact to form regulatory cascades, this finding suggests that hierarchical signalling pathways can potentially evolve by the sequential integration of pre-existing regulatory circuits acquired from diverse sources.