- Volume 153, Issue 9, 2007

Type IV secretion systems are common bacterial macromolecule transporters that have been adapted to various functions, such as effector protein translocation to eukaryotic cells, nucleoprotein transfer to bacterial or eukaryotic cells, and DNA transport into and out of bacterial cells. Helicobacter pylori, the causative agent of bacterial gastritis, peptic ulcers, gastric adenocarcinoma and mucosa-associated lymphoid tissue (MALT) lymphoma, uses the Cag type IV secretion system to inject the CagA protein into host cells, thereby altering gene expression profiles and the host cell cytoskeleton. The molecular mechanism of CagA recognition as a type IV substrate is only poorly understood, but seems to be more complex than that of other type IV secretion systems. Apart from 14 essential components of the secretion apparatus, CagA translocation specifically requires the presence of four additional Cag proteins. Here we show that the CagA-binding protein CagF is a secretion chaperone-like protein that interacts with a 100 aa region that is adjacent to the C-terminal secretion signal of CagA. The interaction between CagA and CagF takes place at the bacterial cytoplasmic membrane, and is independent of a functional type IV secretion apparatus and other cag-encoded factors. Our data indicate that CagF binding precedes recognition of the C-terminal CagA translocation signal, and that both steps are required to recruit CagA to the type IV translocation channel.

Sporotrichosis, a mycosis caused by Sporothrix schenckii, is characterized by lymphocutaneous lesions. In immunocompromised hosts, this fungus may invade the bloodstream and disseminate to other tissues, such as lung and bone. Our group previously showed that S. schenckii yeasts adhere to endothelial monolayers and that this interaction is modulated by cytokines. Using 3.0 μm-pore culture inserts, the present work shows that transforming growth factor (TGF)-β1 led to a 80±26 % increase in fungal migration across endothelial monolayers and inhibited fungus internalization by 55±23.5 %, when compared to untreated cells. The major surface endothelial molecules recognized by S. schenckii were not modulated by TGF-β1. These data suggested that a paracellular route is preferentially used by S. schenckii during the transmigration of cultured endothelial cells. It was further observed that TGF-β1 increased the subendothelial matrix exposure and that anti-fibronectin (anti-FN) and anti-laminin (anti-LM) antibodies abolished the increase in S. schenckii association with endothelial monolayers induced by TGF-β1. These antibodies also inhibited (38.2±4.29 % and 50.8±17.3 %, respectively) the adhesion of S. schenckii to freshly prepared native endothelial matrices. Furthermore, transendothelial migration of S. schenckii was blocked by anti-FN and anti-LM antibodies. These data indicate that TGF-β1-induced S. schenckii adhesion to endothelial monolayers results from the increased exposure of the subendothelial extracellular matrix and that this event may contribute to the enhancement of transendothelial migration.

The goal of this study was to characterize the Yersinia pestis KIM OmpX protein. Yersinia spp. provide a model for studying several virulence processes including attachment to, and internalization by, host cells. For Yersinia enterocolitica and Yersinia pseudotuberculosis, Ail, YadA and Inv, have been implicated in these processes. In Y. pestis, YadA and Inv are inactivated. Genomic analysis of two Y. pestis strains revealed four loci with sequence homology to Ail. One of these genes, designated y1324 in the Y. pestis KIM database, encodes a protein designated OmpX. The mature protein has a predicted molecular mass of 17.47 kDa, shares approximately 70 % sequence identity with Y. enterocolitica Ail, and has an identical homologue, designated Ail, in the Y. pestis CO92 database. The present study compared the Y. pestis KIM6+ parental strain with a mutant derivative having an engineered disruption of the OmpX structural gene. The parental strain (and a merodiploid control strain) expressed OmpX at 28 and 37 °C, and the protein was detectable throughout all phases of growth. OmpX was required for efficient adherence to, and internalization by, cultured HEp-2 cell monolayers and conferred resistance to the bactericidal effect of human serum. Deletion of ompX resulted in a significantly reduced autoaggregation phenotype and loss of pellicle formation in vitro. These results suggest that Y. pestis OmpX shares functional homology with Y. enterocolitica Ail in adherence, internalization into epithelial cells and serum resistance.

Induction of host cell death is thought to play an important role in bacterial pathogenesis. Campylobacter jejuni is a prevalent cause of bacterial enteritis; however, its effects on enterocytes remain unclear. The present study indicates for the first time that C. jejuni induces oncotic, rather than apoptotic death of T84 enterocytes. C. jejuni-treated enterocytes exhibited extensive cytoplasmic vacuolation, rapid (3–6 h) loss of plasma membrane integrity (‘cytotoxicity’), loss of mitochondrial transmembrane potential, and ATP depletion. Enterocytes also exhibited increased oligonucleosomal DNA fragmentation, a feature characteristic of apoptosis. However, consistent with a non-apoptotic process, DNA fragmentation and cytotoxicity were not caspase dependent. During apoptosis, caspases mediate cleavage of poly(ADP-ribose) polymerase; however, cleavage was not observed in C. jejuni-treated monolayers. Cytotoxicity, ATP depletion and DNA fragmentation were not prevented by the deletion of the cytolethal distending toxin (CDT) gene, indicating that C. jejuni causes enterocyte oncosis via a mechanism that is CDT independent. The ability to cause oncosis was significantly decreased in a FlaAFlaB mutant (CDT+) that was defective in the ability to adhere and invade enterocytes. Analysis of clinical isolates revealed that oncosis was strain dependent and correlated with increased invasive ability. These observations offer new insights into the pathogenesis of C. jejuni infection.

Vibrio cholerae is the causative agent of cholera, which continues to be a major public health concern in Asia, Africa and Latin America. The bacterium can persist outside the human host and alternates between planktonic and biofilm community lifestyles. Transition between the different lifestyles is mediated by multiple signal transduction pathways including quorum sensing. Expression of the Zn-metalloprotease haemagglutinin (HA)/protease is subject to a dual regulation which involves the quorum-sensing regulator HapR and the cAMP receptor protein. In a previous study, we observed that a mutant defective in the cAMP-receptor protein (CRP) expressed lower levels of HapR. To further investigate the role of CRP in modulating HapR and other signal transduction pathways, we performed global gene expression profiling of a Δcrp mutant of El Tor biotype V. cholerae. Here we show that CRP is required for the biosynthesis of cholera autoinducer 1 (CAI-1) and affects the expression of multiple HapR-regulated genes. As expected, the Δcrp mutant produced more cholera toxin and enhanced biofilm. Expression of flagellar genes, reported to be affected in ΔhapR mutants, was diminished in the Δcrp mutant. However, an epistasis analysis indicated that cAMP–CRP affects motility by a mechanism independent of HapR. Inactivation of crp inhibited the expression of multiple genes reported to be strongly induced in vivo and to affect the ability of V. cholerae to colonize the small intestine and cause disease. These genes included ompU, ompT and ompW encoding outer-membrane proteins, the alternative sigma factor σ E required for intestinal colonization, and genes involved in anaerobic energy metabolism. Our results indicate that CRP plays a crucial role in the V. cholerae life cycle by affecting quorum sensing and multiple genes required for survival of V. cholerae in the human host and the environment.

Yersinia pestis is the aetiologic agent of plague. Without appropriate treatment, the pathogen rapidly causes septicaemia, the terminal and fatal phase of the disease. In order to identify bacterial genes which are essential during septicaemic plague in humans, we performed a transcriptome analysis on the fully virulent Y. pestis CO92 strain grown in either decomplemented human plasma or Luria–Bertani medium, incubated at either 28 or 37 °C and harvested at either the mid-exponential or the stationary growth phase. Y. pestis genes involved in 12 iron-acquisition systems and one iron-storage system (bfr, bfd) were specifically induced in human plasma. Of these, the ybt and tonB genes (encoding the yersiniabactin siderophore virulence factor and the siderophore transporter, respectively) were induced at 37 °C, i.e. under conditions mimicking the mammalian environment. Growth in human plasma also upregulated genes involved in the synthesis of five fimbrial-like structures (including the Psa virulence factor), and in purine/pyrimidine metabolism (the nrd genes). Genes known to play a role in the virulence of several bacterial pathogens (such as those encoding the Lpp lipoprotein and non-iron metal-uptake proteins) were induced in human plasma, during either the exponential or the stationary phase. Finally, 120 genes encoding proteins of unknown function were upregulated in human plasma. Eleven of these genes were specifically transcribed at 37 °C and may thus represent new virulence factors that are important during the septicaemic phase of human plague.

This study used atomic force microscopy (AFM) to probe the local cell-surface interactions associated with the glucan polymers of Streptococcus mutans, the macromolecules most commonly attributed to the virulence of this microbe. In situ force spectroscopy was used to quantitatively probe and correlate cell-surface adhesion and dynamics with S. mutans UA140 wild-type and five glucosyltransferase mutants. Adhesion between the tooth surface and S. mutans is largely mediated by glucan production from sucrose via three glucosyltransferases (Gtfs; GtfB, GtfC and GtfD). To monitor the contribution of these particular Gtfs, isogenic mutants of S. mutans were constructed by specific gene inactivation and compared to the wild-type under sucrose and non-sucrose conditions. We report direct measurement of the mechanical properties associated with glucan macromolecules demonstrating that the local adhesion strength increases in a time-dependent process, with a decrease in the average number of rupture events. This finding suggests that S. mutans attaches mainly through glucans to surfaces in the presence of sucrose. In addition, a possible role of the Gtf proteins in sucrose-independent attachment is supported by the decreased adhesion properties of the GtfBCD mutant compared to the wild-type.

Alterations to the composition or architecture of the mycobacterial cell envelope can affect the macrophage infectivity of the bacillus. To further characterize the mycobacterial gene products that modulate the interaction with host cells, we employed transposon mutagenesis and screened for mutants that demonstrated an enhanced binding affinity toward macrophages. After successive rounds of mutant selection and enrichment, a total of five mutants were isolated that harboured gene disruptions within loci involved in lipid synthetic pathways as well as genes coding for putative hypothetical proteins. One mutant in particular, with a disruption in the Rv3826 gene (fadD23), was repeatedly isolated during library screening. Analysis of the cell envelope constituents of the Tn : : fadD23 strain revealed a lack of sulfolipid production which was restored following complementation with the wild-type gene.

Francisella tularensis is a zoonotic, Gram-negative coccobacillus that causes tularemia in humans and animals. F. tularensis subspecies tularensis (type A) and F. tularensis subspecies holarctica (type B) are antigenically similar and more virulent than Francisella novicida in humans. The genetic locus that encodes the LPS O antigen was found to be substantially different between the type B live vaccine strain (LVS) and F. novicida. One LVS-specific gene with homology to a galactosyl transferase was selected for allelic replacement using a sacB–chloramphenicol expression suicide plasmid, and recombinants were screened for colony morphology on Congo red agar that matched that of F. novicida. Two mutants (WbtIS187Y and WbtIG191V) were isolated that contained substitutions in conserved motifs in the sugar transamine/perosamine synthetase (WbtI) of the O-antigen locus, and the latter mutant was extensively tested and characterized. WbtIG191V grew at the same rate as the parent strain in Chamberlain's defined medium, completely lacked O antigen, was serum-sensitive but could grow in a mouse macrophage cell line, had increased resistance to sodium deoxycholate, and was highly attenuated in mice. Complementation of WbtIG191V with the wild-type wbtI gene in trans restored normal LPS synthesis, phenotypic properties similar to the parent, and virulence in mice. Immunization with WbtIG191V protected mice against a relatively low-dose intraperitoneal challenge with LVS, but was less protective against a high-dose challenge. These results indicate that complete loss of O antigen alters the surface phenotype and abrogates virulence in F. tularensis, but also compromises the induction of full protective immunity against F. tularensis infection in mice.

Heat-shock proteins are essential for stress tolerance and allowing organisms to survive conditions that cause protein unfolding. The role of the Staphylococcus aureus DnaK system in tolerance of various stresses was studied by disruption of dnaK by partial deletion and insertion of a kanamycin gene cassette. Deletion of dnaK in S. aureus strain COL resulted in poor growth at temperatures of 37 °C and above, and reduced carotenoid production. The mutant strain also exhibited increased susceptibility to oxidative and cell-wall-active antibiotic stress conditions. In addition, the mutant strain had slower rates of autolysis, suggesting a correlation between DnaK and functional expression of staphylococcal autolysins. Deletion of dnaK also resulted in a decrease in the ability of the organism to survive in a mouse host during a systemic infection. In summary, the DnaK system in S. aureus plays a significant role in the survival of S. aureus under various stress conditions.

Sequencing of the DNA region on the left fringe of the pimaricin gene cluster revealed the presence of a 579 bp gene, pimM, whose deduced product (192 aa) was found to have amino acid sequence homology with bacterial regulatory proteins. Database comparisons revealed that PimM combines an N-terminal PAS domain with a C-terminal helix–turn–helix (HTH) motif of the LuxR type. Gene replacement of pimM from the Streptomyces natalensis chromosome with a mutant version lacking the HTH DNA-binding domain resulted in complete loss of pimaricin production, suggesting that PimM is a positive regulator of pimaricin biosynthesis. Complementation of the ΔpimM mutant with a single copy of pimM integrated into the chromosome restored pimaricin production. The insertion of a single copy of pimM, with its own promoter, into the S. natalensis wild-type strain boosted pimaricin production. Gene expression analyses in S. natalensis wild-type and ΔpimM by reverse transcriptase PCR (RT-PCR) of the pimaricin gene cluster revealed the targets for the PimM regulatory protein. According to these analyses, the genes responsible for initiation and first elongation cycles of polyketide chain extension are among the major targets for regulation. Other pim genes are differentially affected. Interestingly, our results indicate that PimM plays its regulatory role independently of PimR, the first pathway-specific regulator of pimaricin biosynthesis.

Strains of Rhizobium leguminosarum bv. trifolii (Rlt) able to form effective nodules on Trifolium ambiguum (Caucasian clover, CC) form ineffective nodules on Trifolium repens (white clover, WC), whereas strains that form effective nodules on WC usually do not nodulate CC. Here, we investigate the genetic basis of the host-specific nitrogen-fixation phenotype of CC rhizobia. A cosmid library of the symbiotic plasmid from the WC rhizobium strain Rlt NZP514 was introduced into the CC rhizobium strain Rlt ICC105. An 18 kb Asp718 fragment containing the nifABHDKEN and fixABCX genes of NZP514 that imparted the Fix+ phenotype was identified. Tn5 mutagenesis of this region revealed that the nifHDKEN, fixABC and nifB genes were required for the Fix+ phenotype, but that the nifA gene was not. Introduction of several plasmids containing NZP514 nif/fix genes into an ICC105 nifA mutant strain demonstrated that the NifA protein of ICC105 was able to activate expression of the NZP514 nif/fix genes but not the ICC105 nif/fix genes in WC nodules. Reporter gene fusion studies showed that the host-specific regulation of the nif/fix genes depended on the DNA region between the promoters of the divergently transcribed nifH and fixA genes. We hypothesize that a protein acting either in response to a host-specific signal or in the absence of such a signal is able to bind upstream of the NifA-binding sites and interact with NifA to prevent it activating nif/fix gene expression.

We have been using a mutational approach to determine how plant-beneficial bacteria such as Enterobacter cloacae 501R3 obtain carbon and energy for colonization of subterranean portions of cucumber and other plants. Reduced carbon detected in cucumber root exudate consisted of 73.3 % amino acids, 22.2 % organic acids and 4.4 % carbohydrate. Ent. cloacae M2, a mini-Tn5 Km transposon mutant of strain 501R3, was severely reduced in in vitro growth relative to strain 501R3 on the mixture of amino acids and organic acids detected in cucumber root exudate when these compounds were supplied as the sole source of carbon and energy, but was similar in growth on the mixture of carbohydrates detected in this exudate. Molecular and biochemical characterization of Ent. cloacae M2 indicated that the transposon was inserted in sdhA, which encodes a subunit of succinate dehydrogenase. Ent. cloacae A-11, a mutant of strain 501R3 with a mini-Tn5 Km insertion in pfkA, was severely reduced in in vitro growth relative to strain 501R3 on the mixture of carbohydrates detected in cucumber root exudate, but similar in growth on the mixture of amino acids and organic acids. When strains A-11 and M2 were coapplied with strain 501R3 to cucumber seeds above carrying capacity in competitive root colonization assays, populations of strains A-11 and M2 were roughly one order of magnitude lower than those of strain 501R3 in cucumber rhizosphere, while populations of strains A-11 and M2 were similar to one other when coapplied to cucumber seeds. When Ent. cloacae strains were coapplied to cucumber seeds below carrying capacity, populations of A-11 and M2 were roughly two to three orders of magnitude lower than those of 501R3 in cucumber rhizosphere, and populations of A-11 were significantly lower than those of M2 when these two strains were coapplied to cucumber seed. The experiments reported here indicate an important role for pfkA and sdhA and the catabolism of carbohydrates, and of amino acids and organic acids, respectively, in the colonization of cucumber roots by Ent. cloacae. The results reported here also indicate that catabolism of carbohydrates by this bacterium is more important than catabolism of amino acids and organic acids at lower population densities, despite the much higher relative quantities of amino acids and organic acids detected in cucumber root exudate.