Model of calcium oscillations due to negative feedback in olfactory cilia.

We present a mathematical model for calcium oscillations in the cilia of olfactory sensory neurons. The underlying mechanism is based on direct negative regulation of cyclic nucleotide-gated channels by calcium/calmodulin and does not require any autocatalysis such as calcium-induced calcium release. The model is in quantitative agreement with available experimental data, both with respect to oscillations and to fast adaptation. We give predictions for the ranges of parameters in which oscillations should be observable. Relevance of the model to calcium oscillations in other systems is discussed.

Is a NAD pyrophosphatase activity necessary for Haemophilus influenzae type b multiplication in the blood stream?

Haemophilus influenzae has an absolute requirement for factor V because it lacks all the biosynthetic enzymes necessary for the de novo synthesis of NAD. Factor V can be provided as either nicotinamide adenosine dinucleotide (NAD), nicotinamide mono-nucleotide (NMN) or nicotinamide riboside (NR) in vitro, but little is known about the source or the mechanism of uptake for factor V in vivo. Recently, a hypothetical open reading frame (ORF), termed nadN, was identified to encode a gene product essential for H. influenzae growth on NAD. Here, we report its role in the virulent H. influenzae serotype b strain Eagan. Our results indicate that NadN of type b Eagan strains is involved in NAD uptake and in processing NAD to NR, which appears to be the substrate for an as yet unidentified cytoplasmic membrane NR transport system. Furthermore, we present data showing that H. influenzae type b nadN mutants are able to survive as well as Eagan, in vivo in the five-day-old infant rat model of human invasive disease. NAD pyrophosphatase and NMN 5'-nucleotidase activities were present in rat and human serum, implying that under infection conditions H. influenzae may obtain NR directly from its host.

NadN and e (P4) are essential for utilization of NAD and nicotinamide mononucleotide but not nicotinamide riboside in Haemophilus influenzae.

Haemophilus influenzae has an absolute requirement for NAD (factor V) because it lacks almost all the biosynthetic enzymes necessary for the de novo synthesis of that cofactor. Factor V can be provided as either nicotinamide adenosine dinucleotide (NAD), nicotinamide mononucleotide (NMN), or nicotinamide riboside (NR) in vitro, but little is known about the source or the mechanism of uptake of these substrates in vivo. As shown by us earlier, at least two gene products are involved in the uptake of NAD, the outer membrane lipoprotein e (P4), which has phosphatase activity and is encoded by hel, and a periplasmic NAD nucleotidase, encoded by nadN. It has also been observed that the latter gene product is essential for H. influenzae growth on media supplemented with NAD. In this report, we describe the functions and substrates of these two proteins as they act together in an NAD utilization pathway. Data are provided which indicate that NadN harbors not only NAD pyrophosphatase but also NMN 5'-nucleotidase activity. The e (P4) protein is also shown to have NMN 5'-nucleotidase activity, recognizing NMN as a substrate and releasing NR as its product. Insertion mutants of nadN or deletion and site-directed mutants of hel had attenuated growth and a reduced uptake phenotype when NMN served as substrate. A hel and nadN double mutant was only able to grow in the presence of NR, whereas no uptake of NMN was observed.

Heat-inducible surface stress protein (Hsp70) mediates sulfatide recognition of the respiratory pathogen Haemophilus influenzae.

The in vitro glycolipid binding specificity of clinical strains of nontypeable Haemophilus influenzae is altered to include sulfated glycolipids following a brief heat shock. We have constructed, expressed, and purified a recombinant protein of H. influenzae Hsp70, which showed significant specific binding to sulfated galactolipids in vitro. Furthermore, indirect immunofluorescence demonstrates that Hsp70 proteins are surface exposed in H. influenzae only after heat shock and are contained in the outer membrane protein fractions.

Characterization of Vibrio cholerae O1 El tor galU and galE mutants: influence on lipopolysaccharide structure, colonization, and biofilm formation.

Recently we described the isolation of spontaneous bacteriophage K139-resistant Vibrio cholerae O1 El Tor mutants. In this study, we identified phage-resistant isolates with intact O antigen but altered core oligosaccharide which were also affected in galactose catabolism; this strains have mutations in the galU gene. We inactivated another gal gene, galE, and the mutant was also found to be defective in the catabolism of exogenous galactose but synthesized an apparently normal lipopolysaccharide (LPS). Both gal mutants as well as a rough LPS (R-LPS) mutant were investigated for the ability to colonize the mouse small intestine. The galU and R-LPS mutants, but not the galE mutant, were defective in colonization, a phenotype also associated with O-antigen-negative mutants. By investigating several parameters in vitro, we could show that galU and R-LPS mutants were more sensitive to short-chain organic acids, cationic antimicrobial peptides, the complement system, and bile salts as well as other hydrophobic agents, indicating that their outer membrane no longer provides an effective barrier function. O-antigen-negative strains were found to be sensitive to complement and cationic peptides, but they displayed significant resistance to bile salts and short-chain organic acids. Furthermore, we found that galU and galE are essential for the formation of a biofilm in a spontaneous phage-resistant rugose variant, suggesting that the synthesis of UDP-galactose via UDP-glucose is necessary for biosynthesis of the exopolysaccharide. In addition, we provide evidence that the production of exopolysaccharide limits the access of phage K139 to its receptor, the O antigen. In conclusion, our results indicate involvement of galU in V. cholerae virulence, correlated with the observed change in LPS structure, and a role for galU and galE in environmental survival of V. cholerae.

Pathogenicity islands and phage conversion: evolutionary aspects of bacterial pathogenesis.

Horizontal gene transfer plays a key role in the generation of novel bacterial pathogens. Besides plasmids and bacteriophages, large genomic regions termed pathogenicity islands (PAIs) can be transferred horizontally. All three mechanisms for DNA exchange or transfer may be important for the evolution of bacterial pathogens.

Characterization of vibrio cholerae O1 antigen as the bacteriophage K139 receptor and identification of IS1004 insertions aborting O1 antigen biosynthesis.

Bacteriophage K139 was recently characterized as a temperate phage of O1 Vibrio cholerae. In this study we have determined the phage adsorption site on the bacterial cell surface. Phage-binding studies with purified lipopolysaccharide (LPS) of different O1 serotypes and biotypes revealed that the O1 antigen serves as the phage receptor. In addition, phage-resistant O1 El Tor strains were screened by using a virulent isolate of phage K139. Analysis of the LPS of such spontaneous phage-resistant mutants revealed that most of them synthesize incomplete LPS molecules, composed of either defective O1 antigen or core oligosaccharide. By applying phage-binding studies, it was possible to distinguish between receptor mutants and mutations which probably caused abortion of later steps of phage infection. Furthermore, we investigated the genetic nature of O1-negative strains by Southern hybridization with probes specific for the O antigen biosynthesis cluster (rfb region). Two of the investigated O1 antigen-negative mutants revealed insertions of element IS1004 into the rfb gene cluster. Treating one wbeW::IS1004 serum-sensitive mutant with normal human serum, we found that several survivors showed precise excision of IS1004, restoring O antigen biosynthesis and serum resistance. Investigation of clinical isolates by screening for phage resistance and performing LPS analysis of nonlysogenic strains led to the identification of a strain with decreased O1 antigen presentation. This strain had a significant reduction in its ability to colonize the mouse small intestine.

NADP and NAD utilization in Haemophilus influenzae.

Exogenous NAD utilization or pyridine nucleotide cycle metabolism is used by many bacteria to maintain NAD turnover and to limit energy-dependent de novo NAD synthesis. The genus Haemophilus includes several important pathogenic bacterial species that require NAD as an essential growth factor. The molecular mechanisms of NAD uptake and processing are understood only in part for Haemophilus. In this report, we present data showing that the outer membrane lipoprotein e(P4), encoded by the hel gene, and an exported 5'-nucleotidase (HI0206), assigned as nadN, are necessary for NAD and NADP utilization. Lipoprotein e(P4) is characterized as an acid phosphatase that uses NADP as substrate. Its phosphatase activity is inhibited by compounds such as adenosine or NMN. The nadN gene product was characterized as an NAD-nucleotidase, responsible for the hydrolysis of NAD. H. influenzae hel and nadN mutants had defined growth deficiencies. For growth, the uptake and processing of the essential cofactors NADP and NAD required e(P4) and 5'-nucleotidase. In addition, adenosine was identified as a potent growth inhibitor of wild-type H. influenzae strains, when NADP was used as the sole source of nicotinamide-ribosyl.

Characterization of ferrochelatase (hemH) mutations in Haemophilus influenzae.

Haemophilus influenzae lacks most of the biosynthetic enzymes for hemin synthesis. However, the organism has retained ferrochelatase activity, which we identified to be encoded by a hemH-homologous gene. In this report we characterize the growth physiology conferred by hemH mutations under infection and laboratory conditions.

Genetic rearrangements of the regions adjacent to genes encoding heat-labile enterotoxins (eltAB) of enterotoxigenic Escherichia coli strains.

One of the most common bacterially mediated diarrheal infections is caused by enterotoxigenic Escherichia coli (ETEC) strains. ETEC-derived plasmids are responsible for the distribution of the genes encoding the main toxins, namely, the heat-labile and heat-stable enterotoxins. The origins and transfer modes (intra- or interplasmid) of the toxin-encoding genes have not been characterized in detail. In this study, we investigated the DNA regions located near the heat-labile enterotoxin-encoding genes (eltAB) of several clinical isolates. It was found that the eltAB region is flanked by conserved 236- and 280-bp regions, followed by highly variable DNA sequences which consist mainly of partial insertion sequence (IS) elements. Furthermore, we demonstrated that rearrangements of the eltAB region of one particular isolate, which harbors an IS91R sequence next to eltAB, could be produced by a recA-independent but IS91 sequence-dependent mechanism. Possible mechanisms of dissemination of IS element-associated enterotoxin-encoding genes are discussed.

Characterization of the major control region of Vibrio cholerae bacteriophage K139: immunity, exclusion, and integration.

The temperate bacteriophage K139 is highly associated with pathogenic O1 Vibrio cholerae strains. The nucleotide sequence of the major control region of K139 was determined. The sequences of four (cox, cII, cI, and int) of the six deduced open reading frames and their gene order indicated that K139 is related to the P2 bacteriophage family. Two genes of the lysogenic transcript from the mapped promoter PL encode homologs to the proteins CI and Int, with deduced functions in prophage formation and maintenance. Between the cI and int genes, two additional genes were identified: orf2, which has no significant similarity to any other gene, and the formerly characterized gene glo. Further analysis revealed that Orf2 is involved in preventing superinfection. In a previous report, we described that mutations in glo cause an attenuation effect in the cholera mouse model (J. Reidl and J. J. Mekalanos, Mol. Microbiol. 18:685-701, 1995). In this report, we present strong evidence that Glo participates in phage exclusion. Glo was characterized to encode a 13.6-kDa periplasmic protein which inhibits phage infection at an early step, hence preventing reinfection of vibriophage K139 into K139 lysogenic cells. Immediately downstream of gene int, the attP site was identified. Upon analysis of the corresponding attB site within the V. cholerae chromosome, it became evident that phage K139 is integrated between the flagellin genes flaA and flaC of O1 El Tor and O139 V. cholerae lysogenic strains.

In vivo transposon mutagenesis in Haemophilus influenzae.

In order to devise an in vivo insertion mutagenesis scheme for Haemophilus influenzae, a novel set of transposons has been constructed. These are Tn10-based minitransposons carried on pACYC184- and pACYC177-based replicons, which are suitable for in vivo transposition in H. influenzae. The transposon delivery system was designed to contain an H. influenzae-specific uptake signal sequence which facilitates DNA transformation into H. influenzae. The following mini-Tn10 elements have been made suitable for specific tasks in H. influenzae: (i) Tn10d-cat, which can be used to generate chloramphenicol-selectable insertion mutations; (ii) Tn10d-bla, an ampicillin-selectable translational fusion system allowing the detection of membrane or secreted proteins; and (iii) Tn10d-lacZcat, a chloramphenicol-selectable lacZ transcriptional fusion system. For the rapid identification of the transposon insertions, a PCR fragment enrichment method was developed. This report demonstrates that this in vivo mutagenesis technique is a convenient tool for the analysis of biochemical and regulatory pathways in the human pathogen H. influenzae.

In vivo transduction with shiga toxin 1-encoding phage.

To facilitate the study of intestinal transmission of the Shiga toxin 1 (Stx1)-converting phage H-19B, Tn10d-bla mutagenesis of an Escherichia coli H-19B lysogen was undertaken. Two mutants containing insertions in the gene encoding the A subunit of Stx1 were isolated. The resultant ampicillin-resistant E. coli strains lysogenic for these phages produced infectious H-19B particles but not active toxin. These lysogens were capable of transducing an E. coli recipient strain in the murine gastrointestinal tract, thereby demonstrating that lysogens of Shiga toxin-converting phages give rise to infectious virions within the host gastrointestinal tract.

Lipoprotein e(P4) is essential for hemin uptake by Haemophilus influenzae.

Heme uptake is a common means of iron and porphyrin acquisition by many pathogenic bacteria. The genus Haemophilus includes several important pathogenic bacterial species that characteristically require hemin-, protoporphyrin-, or heme-substituted proteins as essential growth factors under aerobic conditions. However, the mechanism of heme transport is not understood for Haemophilus. We have cloned a DNA fragment from H. influenzae that allows an Escherichia coli hemA mutant to employ exogenous hemin or protoporphyrin IX as sole sources of porphyrin. DNA sequencing of the cloned DNA fragment suggested that a previously characterized gene (hel) encoding an antigenic, outer membrane lipoprotein e(P4) was responsible for the complementation activity. Construction of hel insertion mutations in strain H. influenzae Rd demonstrated that hel is essential for growth under aerobic conditions but not under anaerobic conditions. The aerobic growth defect of hel mutants could be reversed by providing exogenous hemin in the presence of outer membrane. The analysis of hybrids between e(P4) and beta-lactamase demonstrated that a domain of e(P4) near its NH2' terminus was required for its function in hemin use. Within this domain is a short amino acid sequence that displays similarity to H. influenzae hemin binding protein HbpA, hemin-binding motifs present in eukaryotic transcription activator heme-activated protein, and the heme containing proteins hemoglobin (alpha-chain) and cytochrome C3, suggesting that this region may be involved in hemin binding and/or transport.

Characterization of Vibrio cholerae bacteriophage K139 and use of a novel mini-transposon to identify a phage-encoded virulence factor.

Temperate bacteriophage K139 was isolated from a Vibrio cholerae O139 isolate and characterized in this study. The phage genome consists of a 35 kbp, double-stranded, linear DNA molecule that circularizes and integrates into the chromosome in a site-specific manner. DNA sequences that cross-hybridize with K139 phage DNA are present in all strains of V. cholerae serogroup O1 of the classical biotype examined and in some strains of the El Tor biotype. Phage K139 produces plaques on El Tor O1 strains that do not carry the K139-related sequences but does not plaque on O139 strains that lack detectable phage DNA. This results suggests that O139 strains arose in part by horizontal gene transfer of the O139 antigen genes into an El Tor O1 strain that harboured a K139 prophage. Consistent with this interpretation, the morphology of K139 phage particles is identical to that displayed by the widely distributed family of O1 phages referred to as 'kappa'. In order to test whether K139 phage is involved in lysogenic conversion of V. cholerae, we constructed a novel mini-transposon, Tn10d-bla, which was designed to produce beta-lactamase fusions to phage-encoded, exported proteins. All Tn10d-bla insertions obtained were closely linked to one location on the K139 phage genome. DNA sequence determination of the fusion joints revealed an open reading frame (ORF1), encoding a gene product of 137 amino acids with a typical N-terminal hydrophobic signal sequence. ORF1 was designated the glo gene (G protein-like ORF) because its amino acid sequence shows similarity to eukaryotic Gs(alpha) protein (34.5% identity over an 81-amino-acid overlap) and its C-terminus displays the consensus motif (CAAX) which is found in many small eukaryotic GTP-binding proteins. LD50 assays with isogenic Glo+ and Glo- K139 lysogens suggest that glo encodes a secreted virulence determinant of V. cholerae.

MalY of Escherichia coli is an enzyme with the activity of a beta C-S lyase (cystathionase).

The Escherichia coli maltose system consists of a number of genes whose products are involved in the uptake and metabolism of maltose and maltodextrins. MalT is the central positive gene activator of the regulon and is, together with the cyclic AMP-catabolite gene activator protein system, necessary for the expression of the maltose genes. Expression of malY, a MalT-independent gene, leads to the repression of all MalT-dependent genes. We have purified MalY to homogeneity and found it to be a pyridoxal-5-phosphate-containing enzyme with the enzymatic activity of a beta C-S lyase (cystathionase). MalY is a monomeric protein of 42,000 to 44,000 Da. Strains expressing MalY constitutively abolish the methionine requirement of metC mutants. The enzymatic activity of MetC, the cleavage of cystathionine to homocysteine, ammonia, and pyruvate, can be catalyzed by MalY. However, the cystathionase activity is not required for the function of MalY in repressing the maltose system. By site-directed mutagenesis, we changed the conserved lysine residue at the pyridoxal phosphate binding site (position 233) of MalY to isoleucine. This abolished beta C-S lyase activity but not the ability of the protein to repress the maltose system. Also, the overexpression of plasmid-encoded metC did not affect mal gene expression, nor did the deduced amino acid sequence of MetC show homology to that of MalY.

Maltose and maltotriose can be formed endogenously in Escherichia coli from glucose and glucose-1-phosphate independently of enzymes of the maltose system.

The maltose system in Escherichia coli consists of cell envelope-associated proteins and enzymes that catalyze the uptake and utilization of maltose and alpha,1-4-linked maltodextrins. The presence of these sugars in the growth medium induces the maltose system (exogenous induction), even though only maltotriose has been identified in vitro as an inducer (O. Raibaud and E. Richet, J. Bacteriol., 169:3059-3061, 1987). Induction is dependent on MalT, the positive regulator protein of the system. In the presence of exogenous glucose, the maltose system is normally repressed because of catabolite repression and inducer exclusion brought about by the phosphotransferase-mediated vectorial phosphorylation of glucose. In contrast, the increase of free, unphosphorylated glucose in the cell induces the maltose system. A ptsG ptsM glk mutant which cannot grow on glucose can accumulate [14C]glucose via galactose permeases. In this strain, internal glucose is polymerized to maltose, maltotriose, and maltodextrins in which only the reducing glucose residue is labeled. This polymerization does not require maltose enzymes, since it still occurs in malT mutants. Formation of maltodextrins from external glucose as well as induction of the maltose system is absent in a mutant lacking phosphoglucomutase, and induction by external glucose could be regained by the addition of glucose-1-phosphate entering the cells via a constitutive glucose phosphate transport system. malQ mutants, which lack amylomaltase, are constitutive for the expression of the maltose genes. This constitutive nature is due to the formation of maltose and maltodextrins from the degradation of glycogen.

The malX malY operon of Escherichia coli encodes a novel enzyme II of the phosphotransferase system recognizing glucose and maltose and an enzyme abolishing the endogenous induction of the maltose system.

Mutants lacking MalK, a subunit of the binding protein-dependent maltose-maltodextrin transport system, constitutively express the maltose genes. A second site mutation in malI abolishes the constitutive expression. The malI gene (at 36 min on the linkage map) codes for a typical repressor protein that is homologous to the Escherichia coli LacI, GalR, or CytR repressor (J. Reidl, K. Römisch, M. Ehrmann, and W. Boos, J. Bacteriol. 171:4888-4899, 1989). We now report that MalI regulates an adjacent and divergently oriented operon containing malX and malY. MalX encodes a protein with a molecular weight of 56,654, and the deduced amino acid sequence of MalX exhibits 34.9% identity to the enzyme II of the phosphototransferase system for glucose (ptsG) and 32.1% identity to the enzyme II for N-acetylglucosamine (nagE). When constitutively expressed, malX can complement a ptsG ptsM double mutant for growth on glucose. Also, a delta malE malT(Con) strain that is unable to grow on maltose due to its maltose transport defect becomes Mal+ after introduction of malI::Tn10 and the plasmid carrying malX. MalX-mediated transport of glucose and maltose is likely to occur by facilitated diffusion. We conclude that malX encodes a phosphotransferase system enzyme II that can recognize glucose and maltose as substrates even though these sugars may not represent the natural substrates of the system. The second gene in the operon, malY, encodes a protein of 43,500 daltons. Its deduced amino acid sequence exhibits weak homology to aminotransferase sequences. The presence of plasmid-encoded MalX alone was sufficient for complementing growth on glucose in a ptsM ptsG glk mutant, and the plasmid-encoded MalY alone was sufficient to abolish the constitutivity of the mal genes in a malK mutant. The overexpression of malY in a strain that is wild type with respect to the maltose genes strongly interferes with growth on maltose. This is not the case in a malT(Con) strain that expresses the mal genes constitutively. We conclude that malY encodes an enzyme that degrades the inducer of the maltose system or prevents its synthesis.

MalI, a novel protein involved in regulation of the maltose system of Escherichia coli, is highly homologous to the repressor proteins GalR, CytR, and LacI.

The maltose regulon of Escherichia coli comprises several operons that are under common regulatory control of the MalT activator protein. Five mal genes, organized in two divergent operons, code for a binding-protein-dependent transport system specific for maltose and maltodextrins. MalK, one of the subunits of this transport system, not only is essential for transport but also plays a role in regulation. Mutations abolishing MalK function not only result in inability to transport maltose but also cause constitutive expression of the maltose regulon. For this constitutivity to be exerted, the function of an additional gene product, MalI, is necessary. Using the constitutive expression of a malK-lacZ fusion as a signal, we cloned the malI gene, expressed it in minicells, and determined its DNA sequence. The sequence predicted a protein of 34,729 molecular weight, in agreement with the apparent molecular weight of the protein (35,000) when expressed in minicells and analyzed by polyacrylamide gel electrophoresis and autoradiography. MalI exhibited high homology to the repressor proteins GalR, CytR, and LacI. When the amino acid sequences were appropriately aligned, MalI showed 28% identity to GalR, 21% to CytR, and 24% to LacI. Including conservative amino acid exchanges, these numbers increased to 69, 56, and 58%, respectively. The regions of high homology were clustered in particular at the N-terminal portion of the protein that includes the helix-turn-helix motif thought to be involved in DNA binding. The protein contained a short stretch of 30 amino acids that was surprisingly homologous to a sequence in MalT. The amino-terminal half of the protein exhibited significant homology with MalK. The transcriptional start of malI was determined by reverse transcriptase and by S1 nuclease mapping. We found a possible binding site for cyclic AMP receptor protein in the promoter region of malI as well as two perfect direct repeats of 14 base pairs with twofold symmetry indicating their possible role as operator sites. Upstream to malI we observed a divergent open reading frame that extended to the end of the sequenced DNA.