当前位置: 首页 > 医学版 > 期刊论文 > 基础医学 > 感染与免疫杂志 > 2006年 > 第1期 > 正文
编号:11254917
Iron and pH Homeostasis Intersect at the Level of Fur Regulation in the Gastric Pathogen Helicobacter pylori
     Department of Microbiology and Immunology, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Rd., Bethesda, Maryland 20814

    IRIS-Chiron Vaccines, Via Fiorentina 1, 53100, Siena, Italy

    ABSTRACT

    Helicobacter pylori persistently colonizes the stomach of the majority of the world's population and is a tremendous medical burden due to its causal role in diverse gastric maladies. Since the stomach is a constantly changing environment, successful colonization of H. pylori within this niche requires regulation of bacterial gene expression to cope with the environmental fluctuations. In H. pylori, the ferric uptake regulator (Fur) has been shown to play an intricate role in adaptation of the bacterium to two conditions known to oscillate within the gastric mucosa: iron limitation and low pH. To extend our knowledge of the process of regulation and adaptation in H. pylori, we show that Fur is required for efficient colonization of the Mongolian gerbil: the mutant strain exhibits a 100-fold increase in the 50% infectious dose, as well as a 100-fold defect in competitive colonization, when coinfected with wild-type bacteria. Furthermore, we used DNA microarrays to identify genes whose expression was altered in a Fur-deficient strain. We show that the Fur regulon of H. pylori consists of approximately 30 genes, most of which have been previously annotated as acid stress associated. Finally, we investigate the role of Fur in acid-responsive modulation of gene expression and show that a large number of genes are aberrantly expressed in the Fur mutant specifically upon acid exposure. This fact likely explains the requirement for this regulator for growth and colonization in the stomach.

    INTRODUCTION

    Helicobacter pylori colonizes the stomach of more than 50% of the world's population and is causally associated with the development of disease symptoms that range in severity from inflammation to gastric cancer (8, 17, 36). Within the stomach, H. pylori predominantly persists in the mucous layer overlaying the gastric epithelium and colonization typically lasts for the lifetime of infected individuals unless they receive specific anti-H. pylori treatment (7). The ability to survive in this tumultuous environment is an amazing feat when one considers that the pH in the gastric mucus normally ranges from 4.0 to 6.5 and may plummet to 2.0 during perturbation of the mucus layer (10, 40). The high rate of H. pylori chronic infections and the ability of the bacterium to persist for the lifetime of the infected individual suggest that H. pylori is particularly proficient at adaptation and modification of gene expression in response to changes in the host environment. As such, a number of studies have investigated the ability of H. pylori to alter gene expression upon exposure to a range of diverse environmental parameters (1, 3, 9, 31-33, 57).

    Adaptation to fluctuating environmental conditions requires the ability to tightly regulate gene expression. Gene regulatory hierarchies in H. pylori currently remain poorly defined. However, it is clear that, compared to other sequenced microbial genomes, the H. pylori genome is predicted to encode relatively few transcriptional regulators. This is true for both standard transcriptional regulators and two-component systems that are typically used by bacteria to monitor environmental changes. With regard to regulation and adaptation, previous work identified the ferric uptake regulator (Fur) as an essential component for adaptation to low pH (6) and iron limitation (5).

    Fur is best known for its role in regulation of a complicated system that governs iron uptake and storage. This is typically accomplished as Fur binds to promoter regulatory elements, known as Fur boxes, and represses expression of Fur-regulatory genes. As shown in other bacteria, this binding is typically regulated in an iron-dependent manner; in iron-depleted conditions Fur is not iron bound and is not competent for binding at traditional Fur boxes. However, in H. pylori, this paradigm is considerably more complicated, since Fur has been shown to have the ability to bind to some 80-like sequences only upon iron restriction (15); Fur regulates gene expression in both its iron-bound and apo forms (19). Thus, Fur plays a crucial role in iron-dependent regulation and is likely partially responsible for the adaptive response that H. pylori has been shown to mount upon iron limitation (33).

    The Fur protein is autoregulatory (14). In addition, other members of the Fur regulon are known to include frpB and pfr, which are involved in iron uptake and storage, respectively (13, 15). Also, expression of the genes fecA1, fecA2, feoB, and frpB2/3 have been shown to be iron regulated in a Fur-dependent manner, but it is currently unclear whether this regulation is direct or indirect (13, 54). Recent work utilizing DNA microarrays (18) identified 48 genes whose expression was affected in a fur mutant strain. A similar study utilized two-dimensional proteomics to identify 29 proteins showing aberrant expression in a fur mutant (24). Surprisingly, comparison of the array and proteomic studies reveals only four genes in common: HP0115 (flaB), HP0631 (hydA), HP0653 (pfr), and HP1238 (amiF).

    Although the exact mechanism is not well understood, Fur is also essential for growth at low pH (6). The critical role of Fur in adaptation to acidic conditions may be partially linked to regulation of amiE and amiF, which degrade amides to produce ammonia and the corresponding carboxylic acid (53). In addition, the requirement for Fur under acid stress is likely linked to the fact that the regulatory protein, NikR, which directly regulates Fur expression, is involved in the acid response (9, 11). As such, the acid response in H. pylori is a complex process that is governed by more than one regulatory protein. Acid response in H. pylori, has been studied intensively utilizing a wide range of H. pylori strains, techniques, and experimental designs (1, 3, 9, 29, 32, 50, 57). However, direct comparison of these studies shows that there is only partial overlap between the acid-regulated genes identified. The absence of substantial overlap between those studies suggests that the differences in the experimental design, strains, and conditions used are critical factors affecting the defined acid regulon.

    Since Fur represents a regulatory intersection between adaptation to low pH and iron availability, we sought to expand our knowledge of the Fur regulon by assessing the importance of Fur in vivo, as well as in vitro, by examining the genes affected under conditions of both normal growth and acidic shock. First, we exemplify the importance of Fur as a colonization factor by demonstrating that a fur mutant shows a competitive defect in colonization compared to the wild type in the Mongolian gerbil model of colonization. In addition, the fur mutant strain displays a 2-order-of-magnitude difference in the number of bacteria required to establish a successful infection. To dissect this marked effect at the level of transcription, we compared the transcription profiles of the wild type and isogenic fur mutants across a growth curve. This comparison revealed deregulation of a number of factors. Altered gene expression was most pronounced in exponentially growing H. pylori cultures and included the majority of the known Fur-regulated genes, as well as a number of new genes. Since this comparison overlooks the acidic conditions H. pylori encounters in vivo, we finally sought to determine the role of Fur in adaptation to low pH, a condition the mutants must have encountered in the Mongolian gerbil infection model. We thus compared the adaptive response of our wild-type and fur strains to low pH. As expected, acid-adaptive expression of many iron-regulated and acid-affected genes was altered in the Fur-deficient bacterial cells.

    MATERIALS AND METHODS

    Bacterial strains and growth. The H. pylori strains G27 (12) and B128 (38) were maintained as frozen stocks at –80°C in brain heart infusion medium supplemented with 20% glycerol and 10% fetal bovine serum (FBS). Bacteria were grown on horse blood agar (HBA) plates containing 4% Columbia agar base (Oxoid), 5% defibrinated horse blood (HemoStat Labs, Dixon, CA), 0.2% -cyclodextrin (Sigma), 10 μg of vancomycin (Sigma)/ml, 5 μg of cefsulodin (Sigma)/ml, 2.5 U of polymyxin B (Sigma)/ml, 5 μg of trimethoprim (Sigma)/ml, and 8 μg of amphotericin B (Sigma)/ml under microaerophilic conditions at 37°C. A microaerobic atmosphere was generated by using a CampyGen sachet (Oxoid) in a gas pack jar. For liquid culture, H. pylori was grown in brucella broth (Difco) containing 10% FBS (Gibco-BRL) with shaking in a microaerobic environment.

    fur mutant construction and characterization. An isogenic fur (HP1027) mutation in the G27 and B128 H. pylori strains was constructed by using a strategy based on "gene splicing by overlap extension" (2). Briefly, amplicons encompassing the upstream and downstream regions of fur were generated by PCR with primer pairs 1027-F1-1027-R2 and 1027-F5-1027-R6, respectively (Table 1). In addition, A 1.4-kb fragment was amplified over the Campylobacter coli aphA-3 kanamycin resistance gene (23) by using the Kan-F3-Kan-R4 primer pair. Primer sequences incorporate the indicated restriction sites (Table 1). The PCR conditions were 94°C for 2 min; followed by 10 cycles of 94°C for 20 s, 50°C for 20 s, and 72°C for 1 min; followed in turn by 20 cycles of 94°C for 20 s, 60°C for 20 s, and 72°C for 1 min; with a final extension at 72°C for 7 min.

    The HP1027 PCR primers were designed to allow sticky end tails with complementary sequences for the extremities of the antibiotic resistance cassette (Table 1, underlined sequences). Two consecutive "fusion PCRs" were conducted to fuse the fur upstream and downstream fragments such that they flanked the kanamycin cassette. The resulting PCR chimera was digested with SalI-NotI and cloned into pBluescript vector (Stratagene) to create pHP1027-K7. The resulting plasmid was introduced into H. pylori by natural transformation, and deletion of the Fur coding sequence in the resulting kanamycin-resistant colonies was confirmed by PCR with the primers 1027-F1 and 1027-R6.

    Acid exposure and RNA isolation. Stationary-phase and exponential-phase liquid cultures of G27 wild type and two G27 fur isogenic mutants (24 h, optical density at 600 nm [OD600] of 0.85 to 1.0 for stationary-phase cultures; 12 h, OD600 of 0.42 to 0.56 for exponential-phase cultures) were split such that a portion was harvested to serve as the T = 0 time point, and the remainder was resuspended in pH 5.0 brucella broth supplemented with 10% FBS. The pH of the medium was adjusted by using concentrated hydrochloric acid prior to resuspension. Bacterial samples from each growth phase and each strain were harvested 30, 60, and 90 min after the initiation of the acid exposure on a 0.45-μm-pore-size cellulose filter by vacuum filtration. At each time point, aliquots were removed from the pH 5.0 media and plated to determine the CFU count. Measurement of the media pH after the 90-min time point showed that the pH remained 5.0 for all strains, and the colony counts showed full survival of both wild-type and fur strains for the duration of the acid exposure. Filters containing the H. pylori cells were immediately frozen in liquid nitrogen and subsequently used for isolation of bacterial RNA. H. pylori RNA was isolated by using TRIzol reagent (Gibco-BRL) as previously described (31). The RNA concentration was quantitated by determination of the absorbance at 260 and 280 nm, and RNA integrity was verified by visualization on a 1% agarose gel.

    Microarray hybridization and analysis. To determine genes whose expression was altered in the G27 fur strain in comparison to the wild-type G27 strain under normal growth conditions, RNA harvested from the fur strain at T = 0 was compared to RNA harvested from the wild-type H. pylori strain at T = 0 by microarray analysis (for stationary- and exponential-phase cultures, respectively). To examine the differential gene expression of both mutant and wild-type bacteria in response to acid exposure as a function of growth phase, RNA from wild-type and fur strains was extracted from stationary-phase and logarithmic acid-shifted cultures at T = 0, 30, 60, and 90 min. The RNA was later singularly analyzed by microarray with wild-type T = 0 RNA (harvested from the relevant growth phase) in the reference channel. The biological replicates for the identification of Fur-regulated genes were attained by using two different mutants. For the acid-regulated genes we used a time-dependent analysis. Technical replicas were attained, and each gene was represented twice on each slide. In all cases an equal concentration of each RNA was used for cDNA synthesis in a standard reverse transcriptase reaction using Superscript II (Invitrogen) and Panorama H. pylori cDNA labeling primers (SigmaGenosys). Synthesized cDNAs were singularly purified by using QiaQuick PCR purification columns (QIAGEN) according to the manufacturer's instructions and subsequently indirectly labeled with either Cy5 or Cy3 fluorophores, as previously described (42). Individual Cy5 and Cy3 reactions were properly combined, and unincorporated dyes were removed by using a QiaQuick PCR column according to the manufacturer's instructions. The eluates from the columns were concentrated by evaporation in a Speed Vac and resuspended in 11 μl of Tris-EDTA. Then, 1 μl of 25 mg of yeast tRNA/ml, 2.55 μl of 20x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), and 0.45 μl of 10% sodium dodecyl sulfate was added to each of the labeled probes. Samples were heated to 99°C for 2 min, cooled briefly, and then added to the H. pylori microarray for hybridization and stringency washes as previously described (42). The array used contained 1,660 unique PCR-amplified sequences printed in duplicate and representing 98.9% of all H. pylori open reading frames from strains 26695 and JHP99 (42). The hybridized slides were scanned and analyzed by using the Axon 4000A scanner and GenePix 3.0 software (Axon).

    Data were collated by using the Stanford University Microarray Database (45). Spots showing obvious abnormalities were excluded from analysis. For the wild-type versus fur mutant comparison, genes were filtered by requiring that the log2 of the red/green normalized ratio have an absolute value of 0.926 (1.9-fold) in both experimental sets. For the acid analysis, the data for each growth phase were normalized by mathematical transformation such that the abundance of each gene transcript represented by a given spot was relative to the level of transcript from the same strain (wild type or fur) at the T = 0 time point. The resulting data were analyzed by two-class analysis using the significance analysis of microarrays (SAM) program (38). For each growth phase, T = 30, 60, and 90 min wild-type arrays represented a single class, whereas arrays encompassing the same time point for each of the fur strains represented the second class. For both exponential- and stationary-phase analyses, a false discovery rate of <5% was used as a cutoff for significance. All microarray data generated by the present study is publicly available (http://genome-www.stanford.edu/microarray/).

    Gerbil infections. Four- to eight-week-old male Mongolian gerbils (Harlan, Indianapolis, IN) were fasted for 20 h prior to infection. Animals were subsequently infected with 0.5 ml of liquid grown H. pylori strain B128 via oral gavage. In the case of single infections, 20 animals were infected with approximately 109 total bacteria of either the wild-type or the fur strain. For competition assays, 10 animals were infected with a 1:1 mixture of both wild-type and fur bacteria, representing an input ratio of approximately 108 total CFU. For 50% infectious dose (ID50) assays, 61 animals were infected with serial dilutions of wild type or fur mutant ranging in concentration from 108 to 103 input bacteria. Infections were allowed to proceed for 3, 7, 14, and 45 days in the case of single infections; for 6 and 20 days in the case of competition assays; and for 17 days for the ID50 assays. At the indicated times, animals were sacrificed, the glandular portion of the stomach was excised and homogenized with a mechanical homogenizer, and the CFU counts were determined by plating on HBA. For competition assays, homogenates were additionally plated on HBA supplemented with kanamycin, and the relative number of wild-type bacteria was determined by subtraction of the number of kanamycin-resistant colonies from the total number of colonies growing on the plain HBA plates. Competitive indices were determined by division of the number of mutant bacteria by the number of wild-type bacteria, followed by corrections for deviations from an input ratio of 1:1. ID50 determinations were done by using the Reed-Muench calculation (31).

    Electrophoretic mobility shift assay (EMSA). Recombinant H. pylori Fur protein was expressed in Escherichia coli and purified as previously described (55). Promoter regions of the genes of interest were PCR amplified by using the primers listed in Table 1. Each promoter region was gel purified and end labeled with [32P]ATP and T4 polynucleotide kinase. After labeling, 3.3 pmol of promoter DNA was mixed with recombinant Fur protein at concentrations ranging from 0 to 3 μM. Protein and DNA were mixed in binding buffer (10% glycerol, 15 mM Tris-Cl [pH 8.0], 60 mM KCl, 0.5 mM MnCl2, 8 mM dithiothreitol, 240 μg of bovine serum albumin, and 300 ng of shredded salmon sperm DNA [final concentrations]) in a 20-ml final volume, followed by incubation at 37°C for 30 min. Samples were subsequently separated on a 5% polyacrylamide gel in running buffer (25 mM Tris, 190 mM glycine) for 30 min at 200 V. The gel was then exposed to a Kodak phosphor screen, and the autoradiograph was captured by using a Typhoon 9410 scanner and analyzed with the ImageQuant image analysis software (version 5.2; Molecular Dynamics).

    RPAs. RNA for RNase protection assays (RPAs) was isolated as previously described (30). For verification of growth-phase-dependent changes in gene expression that occurred in the fur strain compared to the wild-type strain, a new liquid growth curve was determined as previously described (48). Samples were harvested at T = 0, 4, 8, 12, 20, and 27 h, and the resulting RNA was used for the RPA. For verification of changes in gene expression that occurred after acid exposure in the fur strain compared to the wild-type strain, the same RNA was used for the RPA that was used for the microarrays. All RPAs were conducted as previously described (48). Riboprobe templates for amiE, hydA, oorD, and pdxJ were generated by using the primer pairs listed in Table 1. The amplified products were ligated to pGEM-T (Promega), proper orientation was confirmed, and riboprobes were synthesized with a Maxiscript kit (Ambion) and 50 μCi of [32P]UTP (Perkin-Elmer) as instructed by the manufacturer. RPAs were performed according to the manufacturer's instruction with the RPAIII kit (Ambion) and 1 μg of H. pylori RNA. Protected products were resolved on a 5% acrylamide-1x Tris-borate-EDTA-8 M urea denaturing gel and exposed to phosphor screens (Kodak). Quantification and peak analysis of bands was conducted by using a Typhoon 9410 scanner and ImageQuant image analysis software (version 5.2; Molecular Dynamics).

    RESULTS

    Role of Fur in gerbil infection. Since Fur plays an important role in in vitro growth at low pH, as well as the modulation of iron homeostasis, it is perhaps not surprising that it has also been shown to play a role in colonization of mice; the fur mutant shows a 2-log difference in the number of recoverable bacteria after 1 month (9). To confirm and expand on this observation in an additional animal model, we examined the role of Fur in colonization of the Mongolian gerbil model of H. pylori infection. We first conducted single-strain infections of male Mongolian gerbils with either the B128 wild type or the fur derivative of B128 and monitored the colonization rate as a function of time. At day 3, we noted a >50-fold decrease in the number of CFU/g of stomach tissue recoverable with the fur strain. This decrease was less pronounced at 7 days postinfection (25-fold difference) and was not apparent by days 14 and 45, at which point we recovered virtually identical numbers of the wild-type and fur strains (Fig. 1A). These data suggest that in this animal model Fur is involved in efficient colonization during the early stages of infection but is not absolutely required to establish a persistent infection in the Mongolian gerbil.

    To more clearly define the role of Fur in colonization of this model, we next conducted ID50 assays to determine whether there was a difference in the number of bacteria required by the wild type and fur mutant to establish infection. As shown in Table 2, the wild type had an ID50 of 7.9 x 103 bacteria, whereas the fur mutant ID50 was 7.9 x 105 (Table 2). Finally, we conducted competition assays, simultaneously infecting identical numbers of both wild-type and mutant strains to examine the importance of Fur in colonization in a competitive environment. Determination of the competitive index (CI) at 6 and 20 days showed CIs of 0.010 and 0.013, respectively (Fig. 1B). This represents a 100-fold defect in the numbers of fur mutant bacteria recovered. Taken together, these data indicate that Fur plays a significant role in colonization of the Mongolian gerbil, although the dynamics of this role appear different than in the murine model.

    Microarray identification of Fur-regulated genes. Since Fur plays a role in colonization and since there is little overlap between the two previous genomic scale studies that sought to determine Fur-regulated genes (18, 24), we wanted to identify genes controlled by this important regulator. To accomplish this, RNA was harvested from G27 wild type and two independently constructed isogenic fur mutants in the G27 strain. RNA was harvested in the exponential and stationary phases of growth. Two independent fur mutants were constructed and analyzed to decrease the possibility of erroneous results due to the occurrence of spontaneous mutations in a single strain. In addition, the G27 strain was used to facilitate comparison to previous studies conducted by us with this strain (32). We chose to look at both growth phases due to the fact that H. pylori has been shown to greatly vary gene expression across a growth curve (48). The isolated RNA was reverse transcribed and subsequently used for hybridization to an H. pylori microarray as described in Materials and Methods. Only genes whose expression was reproducibly altered in both of the examined fur strains are discussed further.

    Direct comparison of the expression profiles of the wild-type and fur mutant strains in the exponential phase of growth revealed 29 genes whose expression was altered by at least 1.9-fold in the fur mutants (Table 3). The 1.9-fold threshold was chosen as pfr, the known Fur-regulated iron storage protein, was notably absent from our list when the common arbitrary twofold cutoff was used. The list of Fur-regulated genes included amiE, fecA1, fecA2, frpB1, exbB, and pfr, which comprise a subset of the known Fur-regulated genes. In addition to these factors, we found altered expression of the following: vacA, which encodes the vacuolating cytoxin; oorA, oorC, and oorD, which encode subunits of the ferredoxin oxidoreductase; pdxA and pdxJ, which encode pyridoxal phosphate biosynthetic proteins; and nine genes predicted to encode hypothetical or conserved hypothetical proteins. Notably, in keeping with the canonical role of Fur as a repressor, all of the genes whose expression was altered were overexpressed in the fur mutant, with the exception of the ferredoxin subunit genes and JHP0947.

    Analysis of the genes that were differentially expressed in the Fur mutant in the stationary growth phase revealed 12 genes whose expression was reproducibly altered by 1.9-fold or more (Table 3). Of note, 10 of these genes were also altered in the exponential phase of growth. The stationary list includes the previously mentioned known Fur-regulated genes, as well as exbD, ruvC (which encodes a Holliday junction resolvase), HP1432 (which is predicted to encode a histidine- and glutamine-rich metal-binding protein), and tonB and aspB (which are predicted to encode a siderophore-mediated iron transport protein and an aspartate aminotransferase, respectively).

    Fur regulation and acid-dependent gene expression. Since Fur is required for growth at low pH (6), we reasoned that we should be able to identify acid-regulated changes in gene expression that require Fur. To accomplish this, both exponential- and stationary-phase cultures of H. pylori G27 were shifted to pH 5.0. At this pH, bacterial replication stops (32), but the G27 strain remains viable (data not shown). RNA was harvested in a time course fashion 30, 60, and 90 min after acid exposure. The resulting samples were hybridized to the H. pylori microarray as described in Materials and Methods, and data were analyzed by using the SAM algorithm (51). Two-class SAM analysis comparing the wild-type expression profiles to the Fur mutant expression profiles revealed that 95 genes showed altered expression in the fur mutant (see supplemental material). Of the 118 genes that were originally identified by us as acid regulated (32), the expression of 12 was altered in the fur mutant in either the exponential or the stationary phase of growth. Of these 12, 6 were also aberrantly regulated in the Fur mutant under normal growth conditions (amiE, frpB1, exbB, exbD, pfr, and HP1432). The other six affected genes included four predicted to encode proteins of unknown function (HP0119, HP0219, HP0318, and HP1457), one putative iron-regulated outer membrane protein (HP0876), and ansB (HP0723).

    This list of 95 genes includes 21 of the 29 genes identified as showing altered expression in the fur mutant in normal growth conditions. As with the standard growth in rich media, we found that growth phase affected the genes that showed altered expression upon exposure to low pH. We found that expression of 57 genes in the exponential phase and 67 genes in the stationary phase was altered in the fur mutant (Fig. 2; see also the supplemental material). Of these, 29 genes were altered in both growth phases and 66 were growth phase unique. Included on the list of genes affected in both growth phases were the ferredoxin-oxidoreductase subunits (oorABCD), the quinine-reactive Ni/Fe hydrogenase subunits (hyaABC), catalase (katA), superoxide dismutase (sodB), and the Holliday junction nuclease (ruvC). As with the standard comparison between the wild-type and fur strains, a significant number of the known Fur-regulated genes (amiE, pfr, fecA1, fecA2, and frpB1) were differentially affected, regardless of the growth phase of the culture. The one exception was the exbB gene, which was only affected in the exponential phase of growth. Of the factors affected in the stationary phase of growth, it is notable that a number of known virulence factor or virulence factor related genes appear on the list. These include cagG (HP0542) and cagH (HP0541), which are encoded on the cag pathogenicity island, the neutrophil-activating protein; napA, HP0609, which encodes a paralog of the vacuolating cytotoxin; HP0009 (hopZ), which is suspected to be involved in bacterial adhesion (37); HP0685, which is involved in flagellar biosynthesis and was found to be essential in vivo (22, 35); and HP0723 (ansB), which has been shown to have a role in colonization (32).

    Microarray data validation. To validate the ability of our H. pylori microarray to determine significant changes in gene expression, RPAs were performed for a select number of genes. Moreover, to confirm the reproducibility of the expression patterns we found for the fur mutant in standard growth conditions, we repeated the growth curve of the wild type and one of the fur strains and harvested RNA at time points that ranged from 4 to 27 h. As discussed earlier, the previously generated array data utilized single time points in the exponential- and stationary-growth phases to analyze gene expression. Based on the OD600 and the relative length of culture, the comparable time points for the RPA data are represented by the 12- and 20-h time points, respectively, in Fig. 3. We chose to investigate expression of four genes; amiE and pdxJ, both of which displayed increased expression by our microarray analysis, oorD, which showed decreased expression by microarray and hydA, which showed no change in expression (Table 3). The results for RPAs with riboprobes specific for these genes are presented in Fig. 3. At the earliest time point, amiE, pdxJ, and hydA all showed >1.9-fold higher expression in the fur strain, whereas oorD expression was roughly equivalent to the wild type. Increased expression of amiE and pdxJ was maintained across the entire growth curve, but the relative levels of hydA message in the fur strain fell below our 1.9-fold cutoff and was maintained at roughly equivalent levels as the wild type for the remainder of the time points. Expression of oorD, which was repressed in the fur strain by our microarray analysis, consistently showed lower levels of expression than the wild type but never to the extent of our 1.9-fold cutoff (Fig. 3). Looking specifically at the 12- and 20-h time points, our RPA data are in good agreement with our array data with the exception that the decreased expression of oorD appears to not be as significant as suggested by the arrays.

    To investigate this discrepancy and to determine whether oorD expression really is affected in a fur strain, we also performed RPAs with the RNA samples harvested from strains that had been exposed to low pH. RPA data for oorD and hydA are presented in Fig. 4 and show that patterns of expression of both genes are indeed altered in the fur strain. Taken together, the RPA data indicate that the changes in gene expression identified by our microarray analysis can be validated by an independent approach and show the ability of the H. pylori microarray to detect changes in gene expression.

    Specific binding of Fur with promoters of regulated genes. Although Fur is a regulatory protein that typically affects gene expression by directly binding specific sequences in the promoters of genes that it controls, altered expression of genes in the fur strain that we identified by microarray could also be due to indirect effects. To investigate this, we purified the Fur protein and tested its ability to bind to the promoters of a subset of our identified Fur-regulated genes by using EMSAs. For these assays, we chose the oorD, pdxJ, HP1432, and vacA promoters, which we identified to be affected by the fur mutation under normal growth conditions (Table 3) and the porG promoter, which we identified as Fur dependent under acidic conditions (see supplemental material). We also incorporated the amiE promoter, which has been shown to be directly regulated by Fur (53), as a positive control and the promoter of the constitutively expressed gene rpoB and a portion of the coding region of flaA as negative controls. As expected, we found that the amiE promoter fragment was shifted by Fur (data not shown), whereas rpoB and flaA migration was not affected (Fig. 5). In addition, oorD, HP1432, porG, and pdxJ showed a marked shift when incubated with the recombinant Fur protein. Moreover, the specificity of this binding was shown by the fact that the addition of excess unlabeled promoter fragment was able to compete away Fur binding for each of these promoters (Fig. 5, lane 6). The promoter of vacA did not show any change in electrophoretic mobility, strengthening our assumption that altered expression of a portion of the genes affected in the fur mutant is due to indirect affects.

    Identification of putative Fur-binding sites. To determine the proportion of our Fur-regulated genes that might be directly bound by Fur, we conducted a bioinformatics analysis that examined the 250-bp region encompassing the predicted promoter for each gene. Based on previous footprinting analysis that identified high-affinity binding sites for Fur in the frpB and fur promoters (13, 14), we previously predicted a conserved Fur-binding consensus sequence of AATAATNNTNA (33). Visual analysis of the genome organization of the genes that displayed altered expression in the fur strain in normal growth media revealed that the 29 genes could be arranged into 22 predicted operons or single gene transcriptional units. Of these 22 predicted transcriptional units, 50% have a conserved theoretical fur box contained in the predicted promoter region in one or both of the sequenced H. pylori strains (Table 3).

    A similar analysis was also conducted on the genes we found to be differentially expressed upon acid exposure in the fur strain. Visual analysis revealed that these 95 genes could be arranged into 73 predicted operons or single gene transcriptional units (71 in the H. pylori strain 26695 genome and two unique to H. pylori strain JHP99). Of these 73 predicted transcriptional units, 14 (19%) contained a conserved Fur box in the 250 bp encompassing the predicted promoter region (see supplemental material). Taken together, this bioinformatic analysis indicates that the 103 total genes altered in our fur strain are predicted to be contained within 80 operons or single gene transcriptional units, of which 19 (24%) are predicted to be located downstream of a theoretical Fur box found less than 250 bp upstream of the first ATG.

    DISCUSSION

    The process of survival and persistence for pathogenic bacteria requires the ability to adapt to constantly fluctuating changes in the in vivo environment. For each microbe, the environmental parameters encountered depend upon the site of colonization. Undoubtedly, tropism for these particular colonization sites is controlled by the ability of the organism to properly modulate gene expression within this niche. H. pylori, which colonizes and persists within the gastric mucosa, is faced with the necessity to react to constant fluctuations in pH. As such, H. pylori is armed with acid resistance defenses that have been shown to include a powerful urease, which buffers acidity via the production of ammonia (34); wbcJ, which is involved in the assembly of the lipopolysaccharide layer (28); recA, which is involved in DNA repair (49); attpD and attpF, which encode subunits of the F1F0-ATPase (27); rocF, which is required for arginase activity (26); and the ferric uptake regulator, fur. Indeed, Fur was identified as the first H. pylori regulatory protein required for growth in acidic conditions (6).

    Fur is known to play a crucial role in modulation of iron homeostasis in H. pylori (54). This is true for Fur homologs in many other gram-negative and gram-positive microbes. In addition, Fur has a role in acid survival in H. pylori (6), as well as in other bacteria (4). Fur classically functions by binding to promoter regulatory elements, called Fur boxes, in an iron-dependent manner; Fur-regulated promoters are repressed under iron-replete conditions and derepressed under iron-deplete conditions. However, this canonical mechanism of regulation appears to be significantly more complicated in H. pylori: Fur has been shown to act as a repressor at the pfr and sodB promoters during iron-deplete conditions (15, 19), has been suggested to act as an activator of expression of some genes (19, 24), and has been shown to affect formamidase enzyme activity via a posttranscriptional mechanism (53). The complexity of these many modes of regulation, combined with the fact that the H. pylori genome encodes relatively few transcriptional regulators, leads to the intriguing possibility that the Fur protein has acquired compensatory roles in gene regulation to facilitate bacterial adaptation and survival in the tumultuous gastric environment (52).

    We show here that although a fur mutant is able to colonize the gastric mucosa of Mongolian gerbils if given at a large enough infectious dose, it does so inefficiently at early stages of infection. We further show that the oral ID50 for the fur mutant strain is 100-fold higher than that of the wild type and demonstrate that the fur mutant displays a CI of 0.01 (a 100-fold defect) upon coinfection with wild type in competition assays. Our finding that Fur is required for efficient colonization in Mongolian gerbils is somewhat in keeping with the recent results of Bury-Mone et al. (9). These investigators showed that a fur mutant strain was cleared or showed decreased colonization efficiency when singly infected in the murine model of infection (9). However, it should be noted that, unlike in the murine model, the fur mutant did not show a decrease in numbers in the Mongolian gerbil model after prolonged single infection. This result might partially be explained by the fact that Bury-Mone used a lower infectious dose than we did in our single-strain colonization assays (108 versus 109). However, this seems unlikely based on the fact that we achieved 100% colonization efficiency in our ID50 assays with an infectious dose of 108 (Table 2). In addition, it is possible that the different H. pylori strains used in the studies affect colonization in the different models. A perhaps more intriguing possibility is that differences in the two animal models affect the absolute requirement of certain genes of H. pylori for various stages of colonization. A systematic comparison of the murine and gerbil models will be required to determine whether this is in fact the case.

    We found that the regulon of genes whose expression is controlled by Fur is affected by bacterial growth phase (Table 5). This is in keeping with our previous studies that showed that H. pylori drastically modulates gene expression across the growth curve and shows a growth-phase-dependent response to iron starvation (33, 48). We identified 29 genes whose expression was altered by at least 1.9-fold in our fur strain under standard growth conditions. These genes include the iron transport genes fecA1, fecA2, and frpB1, as well as amiE, which encodes an aliphatic amidase. Each of these genes has previously been shown to be Fur regulated (13, 53, 54). We additionally identified a number of genes encoding metabolic enzymes. These include oorD, oorA, and oorC, which encode ferredoxin oxidoreductase subunits, and pdxA and pdxJ, which encode the pyridoxal phosphate biosynthetic protein subunits that are involved in vitamin B6 production. The oxygen-labile nature of the oor components has been suggested to be partially responsible for H. pylori's microaerobic nature. Other identified genes include the Holliday junction RNase, ruvC, a putative strain-specific XerCD family recombinase (JHP0951), and the vacuolating cytotoxin, vacA.

    VacA is an H. pylori-secreted protein that is internalized by host cells and results in vacuolar degeneration. A role in colonization for this toxin has previously been shown (43), and recent evidence suggests that it plays a key role in H. pylori persistence via killing of immunomodulatory cells (20, 46). It has been demonstrated by our, and other groups that vacA expression is upregulated upon iron limitation (33, 47); thus, one could hypothesize that our identification of vacA as a Fur-regulated gene was due to a deficiency in iron uptake in the fur strain. However, van Vliet et al. previously showed that Fur-deficient cells actually display a fivefold increase in intracellular iron content (54). Taken together with our findings that we could not identify a putative Fur box in the vacA promoter sequence and the lack of direct binding of recombinant Fur to the vacA promoter (Fig. 5), these data suggest that Fur regulation of vacA expression is indirect. This indirect path of regulation is further supported by the results of a recent microarray study that showed that vacA expression is regulated independently by both iron concentration and Fur (18). We therefore hypothesize that an additional regulator that affects vacA transcription exists. Future research will focus on the identification of this regulatory factor as well as the role of Fur in expression of this important toxin.

    We were struck by the fact that we observed increased expression of two genes predicted to play roles in recombination, ruvC and the putative XerCD family recombinase JHP0951. The RuvC endonuclease functions as a resolvase of Holliday junctions via dual-strand incision across the DNA branch point and, thus, releases nicked duplexes that are subsequently closed by DNA ligase (16, 21). Mutation of H. pylori ruvC was recently shown to result in increased sensitivity to DNA-damaging agents, oxidative stress, and killing by macrophages (25). In addition, the ruvC mutant was shown to exhibit decreased efficiency of homologous recombination and a 100-fold increase in ID50 and to be absolutely required for long-term colonization of the murine gastric mucosa (25). The requirement for recombination in persistence is intriguing due to the panmictic population structure and natural competence of H. pylori; as much as 22% of the genome varies between examined strains (42). Our finding that Fur affects expression of ruvC, combined with the role of the resolvase in persistence, suggests that H. pylori could use iron concentration as a signal to increase natural competence within the human stomach. However, we have currently been unable to demonstrate any difference in the frequency of competence and homologous recombination in our fur strain (data not shown). Thus, it is likely that an increased intracellular iron concentration in the fur mutant results in increased DNA damage via the production of oxygen radicals; this damage subsequently serves as a signal leading to increased ruvC expression.

    Twenty-four genes of our list of twenty-nine Fur-dependent genes are found in the H. pylori strain 26695 genome sequence. Since most of the previously published microarray studies looked specifically at strain 26695 genes, we could directly compare only those 24 genes to previous work. Of these genes, 14 were identified in previous studies that investigated iron and/or Fur-dependent gene expression (18, 24, 33). Moreover, the transcriptional structure of the 24 genes can be arranged into 17 predicted operons and/or single genes. We found that components of 14 of these 17 transcriptional units have previously been identified in Fur/iron studies (18, 24, 33). The remaining three transcriptional units include: HP1520-HP1516, which encodes components that belong to the iron-sulfur-dependent L-serine dehydratase family; HP0991, which encodes a predicted protein with no homologue in the databases, and; HP0877, which encodes the Holliday junction endodeoxyribonuclease ruvC.

    A recent microarray study by Ernst et al. (18) specifically examined genes regulated by iron and Fur under both iron-restricted and iron-replete conditions. Comparison of our list of 24 Fur-regulated genes to the genes identified in their study showed 10 genes in common. These include the previously known Fur-regulated genes: amiE, pfr, fecA1, fecA2, and frpB. Of our 17 predicted transcriptional units, Ernst et al. found components of 11 of these as Fur/iron dependent in their study. The fact that ca. 50% of our genes were also identified by the Ernst et al. study increases our confidence that the genes we identified are indeed Fur regulated. Furthermore, the larger list of genes found by Ernst et al. is not surprising since their experimental design was such that they should identify genes regulated by both iron-bound and apo-Fur.

    As mentioned earlier, of the 24 genes found in H. pylori strain 26695, 14 were previously identified as Fur/iron regulated. In addition, 14 of these genes were previously identified as acid stress related (1, 3, 9, 32, 57), and 10 appear on both an acid and Fur/iron-regulated list (Table 3). This is in keeping with the fact that Fur plays a pivotal role in the acid stress response (6). Because of this, we wondered which genes are regulated by Fur during the course of exposure of H. pylori to low pH Comparison of gene expression profiles after shift to acidic pH revealed that a total of 95 genes showed statistically significant difference in gene expression in the fur mutant in comparison to the wild type (Fig. 2 and see supplemental material). This altered pattern of transcription of so many factors likely accounts for the defect in growth of the fur mutant strain at low pH (6). Our list of genes contains a substantial number of the known Fur-regulated genes and is significantly enriched for genes encoding factors that are involved in iron uptake and storage or require iron as a cofactor (supplemental material). We additionally identified genes involved in ammonia production (amiE, ansB, and hypB), detoxification (katA and sodB), metabolism (oorABCD, hyaABC, porBD, pdxAJ, etc.) pathogenicity (cagG, cagH, a vacA paralog, and napA), and regulation (HP0166). Of note, HP0166, which encodes an essential member of the OmpR subfamily of transcriptional regulators, was previously shown to be regulated by iron limitation (33) and has recently been shown to modulate acid-induced expression of the urease genes ureAB (39), as well as the acid-regulated genes HP0119 and HP1432 (32, 39). This suggests that, like Fur, HP0166 may prove to be another point of convergence of iron and pH homeostasis.

    Of the 29 genes that showed altered expression in the fur strain under standard growth conditions, 21 were also identified in the acid conditions. The remaining eight genes include four JHP specific genes, HP0991 and HP0261, which have not previously been suggested to be regulated by low pH and vacA and HP1520. Expression of vacA is known to be repressed by low pH (9, 32), whereas expression of HP1520 is acid induced (57). The absence of these two known pH-regulated genes from the acid portion of our study suggests two possibilities. First, since array studies are not perfect, these genes may not have been identified due to the inability of our array to resolve every regulated gene. Second, additional regulators may control expression of these two genes. This hypothesis is based on the fact that our 95 genes were identified due to altered expression in the fur mutant. If an additional regulatory factor modulated expression of vacA or HP1520 under acidic conditions, this could account for their absence from our list. As mentioned, the existence of such an additional regulator of vacA is supported by the findings of Ernst et al., which showed that both iron and Fur regulate vacA expression independently (18). Alternate control of vacA under acidic conditions is further strengthened by the recent work of Bury-Mone et al., which showed that vacA expression is still affected by low pH in a nikR-fur double mutant (9). Interestingly, we did identify a vacA paralog (HP0609) whose expression was affected by the fur mutation under acidic shock. Although the function of the vacA paralogs is currently unclear, one of them (HP0289) was shown to be essential for colonization in the Mongolian gerbil model (22). Our finding that HP0609 expression is affected by Fur and acid shock implies that H. pylori may activate different vacA paralogs in response to alternating environmental signals. This leads to the question of what is the role of each of these paralogs in H. pylori

    Based on the fact that there is often little overlap between microarray studies, we performed several broad comparisons to determine the extent of overlap of our study with previous studies that identified iron-regulated or acid-regulated genes of H. pylori (3, 9, 29, 32, 57). Of our list of 95 genes that were differentially regulated in the fur strain in acid conditions, 93 of the genes are found in the H. pylori strain 26695 genome sequence and could be directly compared to the previous work. A gene-by-gene comparison revealed the following: 43 of our 93 genes have been directly identified in a previous acid study, 41 of our genes were previously found to be regulated in an iron or Fur study, and 23 genes were identified in both an acid and a Fur/iron study (supplemental material). Moreover, these 93 genes are predicted to be arranged in 71 operons or single gene transcriptional units (supplemental material). Of these, 45 transcriptional units (either the same gene we identified or another gene in the predicted transcriptional unit) were previously identified in an acid-regulated gene study; 44 were found in an iron or Fur regulation study (18, 24, 33), and 29 appear in both an acid and an Fur/iron study. Thus, of our list, only 11 transcriptional units have not previously been shown to be correlated to either stress condition. Of these, three are related to acid or iron by annotation (HP0132-HP0133, HP0139-HP0136, and HP0373). Taken together, this indicates that 89% of our Fur/acid-regulated transcriptional units have been previously identified as Fur, iron, or acid regulated.

    The overall theme of cross talk between metal ions and acid is further highlighted by recent studies linking acid-responsive gene induction to a metal-responsive repressor cascade involving Fur (9, 52). Both of these studies revealed that the nickel-responsive regulatory protein, NikR, acts upstream of Fur. Upon sensing acidic conditions, nikR is upregulated, and increased NikR, which mediates nickel uptake, then directly represses expression of fur (11). This role for NikR in pH-regulated gene expression is in keeping with the previous finding that nickel is important for acid resistance and also demonstrates that this role in acid resistance is significantly more complex than just provision of nickel as a cofactor for the urease enzyme.

    Bury-Mone et al. previously identified 101 genes that showed altered expression during growth of H. pylori in acidic conditions (9). Of these 101 genes, 36 required NikR and/or Fur for pH-regulated changes in gene expression. This was shown as acid-induced changes in expression were negated in a nikR-fur double mutant (9). Six of these genes were also found to show altered expression in our single fur mutant (supplemental material). These include HP0119, a hypothetical protein; the response regulator HP0166; amiE; and several outer membrane and transport proteins, including HP0869, which is involved in incorporation of nickel ions into the hydrogenase and urease proteins, and HP1339-HP1340, which are predicted biopolymer transport accessory proteins. We additionally found two genes that showed altered expression in our fur mutant that were still regulated properly in the nikR-fur double mutant (9). These include HP1432, a histidine and glutamine-rich metal-binding protein, which is known to be Fur regulated (18), and HP1457, which is predicted to encode a hypothetical protein. Interestingly, HP1432 was previously shown to be NikR regulated (11). Moreover, HP0229 (hopA), HP1177 (hopQ), and HP1501 (hopK) were all previously shown to be NikR regulated and still show normal levels of gene expression in the nikR-fur double mutant (53). This once again strongly suggests that there may be additional regulators that are active when the bacteria are grown in an acidic environment. Differences in the sets of genes identified by us and those identified by Bury-Mone et al. (9) may be attributed to the different experimental designs (growth under mild acidic condition versus acidic shock causing an arrest in replication), as well as to the use of different mutant strains (nikR-fur versus fur). Based on the findings of both studies, it would be interesting to determine the transcriptome of a single nikR mutant upon shift to acidic pH and compare the results with the present study, as well as with the results of the acid response of a double nikR-fur mutant (9), to facilitate our understanding of the role of each of these factors, independently, in mediating acid resistance.

    Comparison of our list of 93 Fur/acid-regulated genes to the Fur-dependent (48 genes) and/or iron-dependent (80 genes) previously identified by Ernst et al. revealed 23 genes in common (18). Comparison to our 71 predicted transcriptional units revealed 27 U (38%) in common, 10 of which have been shown to be iron dependent (supplemental material). Once again, the fact that somewhat different sets of genes were found by the groups is no surprise when one considers the very different experimental designs used by the two studies. Moreover, as mentioned previously, H. pylori apparently utilizes a complex regulatory circuits to respond to fluctuations in environmental conditions. Thus, the genes regulated by Fur upon exposure to low pH are likely very different from those regulated by Fur during standard laboratory growth.

    Our bioinformatics analysis of putative Fur-binding sights suggests that 19% of our predicted transcriptional units contain a theoretical Fur box. A much larger percentage of putative Fur boxes was found in the genes identified under normal growth conditions than in acid conditions. This may indicate either a mostly indirect effect on gene transcription or altered Fur-binding properties under acid conditions. Interestingly, in H. pylori, apo-Fur is known to be competent for binding to alternative sequences that do not resemble the previously identified Fur boxes bound by iron-bound Fur (19). Therefore, perhaps upon exposure to low pH, Fur-binding properties are altered and a different set of genes is regulated as a result.

    Along this line, it should be noted that our EMSAs showed that the promoter regions of oorD, pdxJ, porG, and HP1432 were all specifically bound by Fur (Fig. 5), although it should be noted that there were differences in the affinity of Fur for the different fragments, at least in vitro. Interestingly, only oorD contains a conserved consensus Fur-binding sequence. Close examination of the promoter regions of porG, HP1432, and pdxJ reveals a single base modification of the theoretical sequence present in all three promoters in both published genomes. This strongly suggests that the conserved binding site, which was defined based on high-affinity binding of Fur to the fur and frpB promoters (13, 14, 33), needs to be redefined. This will require an in-depth investigation of the actual Fur-binding consensus of H. pylori on multiple promoters.

    Finally, our finding that Fur modulates expression and changes in transcription in response to acid stress of such a large number of genes likely explains the in vivo colonization defect of a fur mutant strain we observed in the gerbil model. The fact that we observed a slight decrease in colonization of the fur mutant at early time points, which became nonapparent by day 14 in single-strain infections, suggests that Fur is important for the initial stages of colonization but is not absolutely required to establish infection in this model. The reason for this is not immediately evident, but perhaps the fur bacteria are killed by environmental conditions present in the animal in regions different from the site of final colonization; the low numbers of the fur mutant bacteria retrieved could be the result of significant numbers of bacteria killed in the stomach during the course of colonization establishment at the mucus layer. Exactly which of the Fur-regulated genes are responsible for the requirement for Fur in vivo is unclear. However, there are a number of genes on the Fur-regulated list that have been previously shown to play a role in colonization. These include hopZ (HP0009), which is believed to be involved in adhesion (37); ansB (HP0723), which encodes an L-asparaginase that has been shown to have a role in colonization (32); and fliP (HP0685), which plays a role in flagellar biosynthesis (35) and was found to be essential in vivo (22). The altered expression of these factors, as well as of others, upon acid stress may account for the reduced infectivity of the fur strain in both animal models. Thus, the Fur regulator is an important mediator of H. pylori colonization. Future studies to determine whether this is a general defect due to deregulation of the large Fur-regulon described herein or due to aberrant expression of a particular gene (or set of genes) will be the focus of future research and should shed insight into the ability of this pathogen to persist within the human stomach.

    ACKNOWLEDGMENTS

    We thank D. Israel and R. Peek for providing the B128 gerbil-colonizing strain; A. Camilli, S. Falkow, and L. Thompson for critical review of the manuscript; and members of the Falkow lab for invaluable discussion. In addition, we thank S. Bereswill and M. Kist for providing the recombinant H. pylori Fur expression plasmid (derived from pASKIBA-7, IBA), A. H. M van Vliet for providing plasmids and sharing unpublished data, and J. Whitmire for technical assistance.

    We are grateful to S. Falkow and L. S. Tompkins for generous monetary support of this work. This study was also supported by the Damon Runyon Cancer Research Fund (D.S.M.) and NIH grants CA92229 and AI38459 (S.F. and L.S.T.).

    Supplemental material for this article may be found at http://iai.asm.org/.

    REFERENCES

    1. Allan, E., C. L. Clayton, A. McLaren, D. M. Wallace, and B. W. Wren. 2001. Characterization of the low-pH responses of Helicobacter pylori using genomic DNA arrays. Microbiology 147:2285-2292.

    2. Amberg, D. C., D. Botstein, and E. M. Beasley. 1995. Precise gene disruption in Saccharomyces cerevisiae by double fusion polymerase chain reaction. Yeast 11:1275-1280.

    3. Ang, S., C. Z. Lee, K. Peck, M. Sindici, U. Matrubutham, M. A. Gleeson, and J. T. Wang. 2001. Acid-induced gene expression in Helicobacter pylori: study in genomic scale by microarray. Infect. Immun. 69:1679-1686.

    4. Bearson, S., B. Bearson, and J. W. Foster. 1997. Acid stress responses in enterobacteria. FEMS Microbiol. Lett. 147:173-180.

    5. Bereswill, S., S. Greiner, A. H. van Vliet, B. Waidner, F. Fassbinder, E. Schiltz, J. G. Kusters, and M. Kist. 2000. Regulation of ferritin-mediated cytoplasmic iron storage by the ferric uptake regulator homolog (Fur) of Helicobacter pylori. J. Bacteriol. 182:5948-5953.

    6. Bijlsma, J. J., B. Waidner, A. H. Vliet, N. J. Hughes, S. Hag, S. Bereswill, D. J. Kelly, C. M. Vandenbroucke-Grauls, M. Kist, and J. G. Kusters. 2002. The Helicobacter pylori homologue of the ferric uptake regulator is involved in acid resistance. Infect. Immun. 70:606-611.

    7. Blaser, M. J. 1990. Epidemiology and pathophysiology of Campylobacter pylori infections. Rev. Infect. Dis. 12(Suppl. 1):S99-S106.

    8. Blaser, M. J. 1998. Helicobacter pylori and gastric diseases. BMJ 316:1507-1510.

    9. Bury-Mone, S., J. M. Thiberge, M. Contreras, A. Maitournam, A. Labigne, and H. De Reuse. 2004. Responsiveness to acidity via metal ion regulators mediates virulence in the gastric pathogen Helicobacter pylori. Mol. Microbiol. 53:623-638.

    10. Chu, S., S. Tanaka, J. D. Kaunitz, and M. H. Montrose. 1999. Dynamic regulation of gastric surface pH by luminal pH. J. Clin. Investig. 103:605-612.

    11. Contreras, M., J. M. Thiberge, M. A. Mandrand-Berthelot, and A. Labigne. 2003. Characterization of the roles of NikR, a nickel-responsive pleiotropic autoregulator of Helicobacter pylori. Mol. Microbiol. 49:947-963.

    12. Covacci, A., S. Censini, M. Bugnoli, R. Petracca, D. Burroni, G. Macchia, A. Massone, E. Papini, Z. Xiang, N. Figura, et al. 1993. Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. USA 90:5791-5795.

    13. Delany, I., A. B. Pacheco, G. Spohn, R. Rappuoli, and V. Scarlato. 2001. Iron-dependent transcription of the frpB gene of Helicobacter pylori is controlled by the Fur repressor protein. J. Bacteriol. 183:4932-4937.

    14. Delany, I., G. Spohn, A. B. Pacheco, R. Ieva, C. Alaimo, R. Rappuoli, and V. Scarlato. 2002. Autoregulation of Helicobacter pylori Fur revealed by functional analysis of the iron-binding site. Mol. Microbiol. 46:1107-1122.

    15. Delany, I., G. Spohn, R. Rappuoli, and V. Scarlato. 2001. The Fur repressor controls transcription of iron-activated and -repressed genes in Helicobacter pylori. Mol. Microbiol. 42:1297-1309.

    16. Dunderdale, H. J., F. E. Benson, C. A. Parsons, G. J. Sharples, R. G. Lloyd, and S. C. West. 1991. Formation and resolution of recombination intermediates by Escherichia coli RecA and RuvC proteins. Nature 354:506-510.

    17. Dunn, B. E., H. Cohen, and M. J. Blaser. 1997. Helicobacter pylori. Clin. Microbiol. Rev. 10:720-741.

    18. Ernst, F. D., S. Bereswill, B. Waidner, J. Stoof, U. Mader, J. G. Kusters, E. J. Kuipers, M. Kist, A. H. van Vliet, and G. Homuth. 2005. Transcriptional profiling of Helicobacter pylori Fur- and iron-regulated gene expression. Microbiology 151:533-546.

    19. Ernst, F. D., G. Homuth, J. Stoof, U. Mader, B. Waidner, E. J. Kuipers, M. Kist, J. G. Kusters, S. Bereswill, and A. H. van Vliet. 2005. Iron-responsive regulation of the Helicobacter pylori iron-cofactored superoxide dismutase SodB is mediated by Fur. J. Bacteriol. 187:3687-3692.

    20. Gebert, B., W. Fischer, E. Weiss, R. Hoffmann, and R. Haas. 2003. Helicobacter pylori vacuolating cytotoxin inhibits T lymphocyte activation. Science 301:1099-1102.

    21. Iwasaki, H., M. Takahagi, T. Shiba, A. Nakata, and H. Shinagawa. 1991. Escherichia coli RuvC protein is an endonuclease that resolves the Holliday structure. EMBO J. 10:4381-4389.

    22. Kavermann, H., B. P. Burns, K. Angermuller, S. Odenbreit, W. Fischer, K. Melchers, and R. Haas. 2003. Identification and characterization of Helicobacter pylori genes essential for gastric colonization. J. Exp. Med. 197:813-822.

    23. Labigne-Roussel, A., P. Courcoux, and L. Tompkins. 1988. Gene disruption and replacement as a feasible approach for mutagenesis of Campylobacter jejuni. J. Bacteriol. 170:1704-1708.

    24. Lee, H. W., Y. H. Choe, D. K. Kim, S. Y. Jung, and N. G. Lee. 2004. Proteomic analysis of a ferric uptake regulator mutant of Helicobacter pylori: regulation of Helicobacter pylori gene expression by ferric uptake regulator and iron. Proteomics 4:2014-2027.

    25. Loughlin, M. F., F. M. Barnard, D. Jenkins, G. J. Sharples, and P. J. Jenks. 2003. Helicobacter pylori mutants defective in RuvC Holliday junction resolvase display reduced macrophage survival and spontaneous clearance from the murine gastric mucosa. Infect. Immun. 71:2022-2031.

    26. McGee, D. J., F. J. Radcliff, G. L. Mendz, R. L. Ferrero, and H. L. Mobley. 1999. Helicobacter pylori rocF is required for arginase activity and acid protection in vitro but is not essential for colonization of mice or for urease activity. J. Bacteriol. 181:7314-7322.

    27. McGowan, C. C., T. L. Cover, and M. J. Blaser. 1997. Analysis of F1F0-ATPase from Helicobacter pylori. Infect. Immun. 65:2640-2647.

    28. McGowan, C. C., A. Necheva, S. A. Thompson, T. L. Cover, and M. J. Blaser. 1998. Acid-induced expression of an LPS-associated gene in Helicobacter pylori. Mol. Microbiol. 30:19-31.

    29. McGowan, C. C., A. S. Necheva, M. H. Forsyth, T. L. Cover, and M. J. Blaser. 2003. Promoter analysis of Helicobacter pylori genes with enhanced expression at low pH. Mol. Microbiol. 48:1225-1239.

    30. Merrell, D. S., and A. Camilli. 1999. The cadA gene of Vibrio cholerae is induced during infection and plays a role in acid tolerance. Mol. Microbiol. 34:836-849.

    31. Merrell, D. S., and S. Falkow. 2003. Expression profiling in Helicobacter pylori infection, p. 273-303. In K. Appasani (ed.), Perspectives in gene expression. Eaton Publishing, Westborough, United Kingdom.

    32. Merrell, D. S., M. L. Goodrich, G. Otto, L. S. Tompkins, and S. Falkow. 2003. pH-regulated gene expression of the gastric pathogen Helicobacter pylori. Infect. Immun. 71:3529-3539.

    33. Merrell, D. S., L. J. Thompson, C. C. Kim, H. Mitchell, L. S. Tompkins, A. Lee, and S. Falkow. 2003. Growth phase-dependent response of Helicobacter pylori to iron starvation. Infect. Immun. 71:6510-6525.

    34. Mobley, H. L., M. D. Island, and R. P. Hausinger. 1995. Molecular biology of microbial ureases. Microbiol. Rev. 59:451-480.

    35. Niehus, E., H. Gressmann, F. Ye, R. Schlapbach, M. Dehio, C. Dehio, A. Stack, T. F. Meyer, S. Suerbaum, and C. Josenhans. 2004. Genome-wide analysis of transcriptional hierarchy and feedback regulation in the flagellar system of Helicobacter pylori. Mol. Microbiol. 52:947-961.

    36. Parsonnet, J., G. D. Friedman, D. P. Vandersteen, Y. Chang, J. H. Vogelman, N. Orentreich, and R. K. Sibley. 1991. Helicobacter pylori infection and the risk of gastric carcinoma. N. Engl. J. Med. 325:1127-1131.

    37. Peck, B., M. Ortkamp, K. D. Diehl, E. Hundt, and B. Knapp. 1999. Conservation, localization, and expression of HopZ, a protein involved in adhesion of Helicobacter pylori. Nucleic Acids Res. 27:3325-3333.

    38. Peek, R. M., Jr., S. A. Thompson, J. P. Donahue, K. T. Tham, J. C. Atherton, M. J. Blaser, and G. G. Miller. 1998. Adherence to gastric epithelial cells induces expression of a Helicobacter pylori gene, iceA, that is associated with clinical outcome. Proc. Assoc. Am. Physicians 110:531-544.

    39. Pflock, M., P. Dietz, J. Schar, and D. Beier. 2004. Genetic evidence for histidine kinase HP165 being an acid sensor of Helicobacter pylori. FEMS Microbiol. Lett. 234:51-61.

    40. Quigley, E. M., and L. A. Turnberg. 1987. pH of the microclimate lining human gastric and duodenal mucosa in vivo: studies in control subjects and in duodenal ulcer patients. Gastroenterology 92:1876-1884.

    41. Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty per cent endpoints. Am. J. Hyg. 27:493-497.

    42. Salama, N., K. Guillemin, T. K. McDaniel, G. Sherlock, L. Tompkins, and S. Falkow. 2000. A whole-genome microarray reveals genetic diversity among Helicobacter pylori strains. Proc. Natl. Acad. Sci. USA 97:14668-14673.

    43. Salama, N. R., G. Otto, L. Tompkins, and S. Falkow. 2001. Vacuolating cytotoxin of Helicobacter pylori plays a role during colonization in a mouse model of infection. Infect. Immun. 69:730-736.

    44. Salama, N. R., B. Shepherd, and S. Falkow. 2004. Global transposon mutagenesis and essential gene analysis of Helicobacter pylori. J. Bacteriol. 186:7926-7935.

    45. Sherlock, G., T. Hernandez-Boussard, A. Kasarskis, G. Binkley, J. C. Matese, S. S. Dwight, M. Kaloper, S. Weng, H. Jin, C. A. Ball, M. B. Eisen, P. T. Spellman, P. O. Brown, D. Botstein, and J. M. Cherry. 2001. The Stanford Microarray Database. Nucleic Acids Res. 29:152-155.

    46. Sundrud, M. S., V. J. Torres, D. Unutmaz, and T. L. Cover. 2004. Inhibition of primary human T-cell proliferation by Helicobacter pylori vacuolating toxin (VacA) is independent of VacA effects on IL-2 secretion. Proc. Natl. Acad. Sci. USA 101:7727-7732.

    47. Szczebara, F., L. Dhaenens, S. Armand, and M. O. Husson. 1999. Regulation of the transcription of genes encoding different virulence factors in Helicobacter pylori by free iron. FEMS Microbiol. Lett. 175:165-170.

    48. Thompson, L. J., D. S. Merrell, B. A. Neilan, H. Mitchell, A. Lee, and S. Falkow. 2003. Gene expression profiling of Helicobacter pylori reveals a growth-phase-dependent switch in virulence gene expression. Infect. Immun. 71:2643-2655.

    49. Thompson, S. A., and M. J. Blaser. 1995. Isolation of the Helicobacter pylori recA gene and involvement of the recA region in resistance to low pH. Infect. Immun. 63:2185-2193.

    50. Toledo, H., M. Valenzuela, A. Rivas, and C. A. Jerez. 2002. Acid stress response in Helicobacter pylori. FEMS Microbiol. Lett. 213:67-72.

    51. Tusher, V. G., R. Tibshirani, and G. Chu. 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 98:5116-5121.

    52. van Vliet, A. H., E. J. Kuipers, J. Stoof, S. W. Poppelaars, and J. G. Kusters. 2004. Acid-responsive gene induction of ammonia-producing enzymes in Helicobacter pylori is mediated via a metal-responsive repressor cascade. Infect. Immun. 72:766-773.

    53. van Vliet, A. H., J. Stoof, S. W. Poppelaars, S. Bereswill, G. Homuth, M. Kist, E. J. Kuipers, and J. G. Kusters. 2003. Differential regulation of amidase- and formamidase-mediated ammonia production by the Helicobacter pylori fur repressor. J. Biol. Chem. 278:9052-9057.

    54. van Vliet, A. H., J. Stoof, R. Vlasblom, S. A. Wainwright, N. J. Hughes, D. J. Kelly, S. Bereswill, J. J. Bijlsma, T. Hoogenboezem, C. M. Vandenbroucke-Grauls, M. Kist, E. J. Kuipers, and J. G. Kusters. 2002. The role of the ferric uptake regulator (Fur) in regulation of Helicobacter pylori iron uptake. Helicobacter 7:237-244.

    55. Waidner, B., S. Greiner, S. Odenbreit, H. Kavermann, J. Velayudhan, F. Stahler, J. Guhl, E. Bisse, A. H. van Vliet, S. C. Andrews, J. G. Kusters, D. J. Kelly, R. Haas, M. Kist, and S. Bereswill. 2002. Essential role of ferritin Pfr in Helicobacter pylori iron metabolism and gastric colonization. Infect. Immun. 70:3923-3929.

    56. Waidner, B., K. Melchers, I. Ivanov, H. Loferer, K. W. Bensch, M. Kist, and S. Bereswill. 2002. Identification by RNA profiling and mutational analysis of the novel copper resistance determinants CrdA (HP1326), CrdB (HP1327), and CzcB (HP1328) in Helicobacter pylori. J. Bacteriol. 184:6700-6708.

    57. Wen, Y., E. A. Marcus, U. Matrubutham, M. A. Gleeson, D. R. Scott, and G. Sachs. 2003. Acid-adaptive genes of Helicobacter pylori. Infect. Immun. 71:5921-5939.(Hanan Gancz, Stefano Cens)