• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. Aug 2006; 188(15): 5428–5438.
PMCID: PMC1540026

Role of Histone-Like Proteins H-NS and StpA in Expression of Virulence Determinants of Uropathogenic Escherichia coli

Abstract

The histone-like protein H-NS is a global regulator in Escherichia coli that has been intensively studied in nonpathogenic strains. However, no comprehensive study on the role of H-NS and its paralogue, StpA, in gene expression in pathogenic E. coli has been carried out so far. Here, we monitored the global effects of H-NS and StpA in a uropathogenic E. coli isolate by using DNA arrays. Expression profiling revealed that more than 500 genes were affected by an hns mutation, whereas no effect of StpA alone was observed. An hns stpA double mutant showed a distinct gene expression pattern that differed in large part from that of the hns single mutant. This suggests a direct interaction between the two paralogues and the existence of distinct regulons of H-NS and an H-NS/StpA heteromeric complex. hns mutation resulted in increased expression of alpha-hemolysin, fimbriae, and iron uptake systems as well as genes involved in stress adaptation. Furthermore, several other putative virulence genes were found to be part of the H-NS regulon. Although the lack of H-NS, either alone or in combination with StpA, has a huge impact on gene expression in pathogenic E. coli strains, its effect on virulence is ambiguous. At a high infection dose, hns mutants trigger more sudden lethality due to their increased acute toxicity in murine urinary tract infection and sepsis models. At a lower infectious dose, however, mutants lacking H-NS are attenuated through their impaired growth rate, which can only partially be compensated for by the higher expression of numerous virulence factors.

Uropathogenic Escherichia coli (UPEC) is the leading cause of uncomplicated urinary tract infections, the most common bacterial infection in industrialized countries (28). The ability of these pathogenic variants to colonize the urinary tract is linked to the presence of virulence factors encoded by horizontally acquired genes not present in their nonpathogenic relatives. These factors include adhesins, cytotoxins, iron uptake systems, capsular polysaccharide, and serum resistance. In E. coli strain 536 (O6:K15:H31), a clinical isolate from acute pyelonephritis, most of the virulence factors known so far are encoded on large, unstable regions of the chromosome, so-called pathogenicity islands (PAIs) (16). PAIs are considered to have been acquired by horizontal gene transfer and therefore represent “foreign” DNA, which is now under the influence of cellular regulators. Interestingly, PAIs often encode putative regulators themselves (11). Therefore, the interaction between PAI-encoded genes and the regulatory network of the “host cell” is thought to be highly complex.

One important regulator in Enterobacteriaceae is the histone-like nucleoid structuring protein H-NS (for reviews, see references 2 and 10). The hns gene, also known as osmZ, drdX, or bglY, encodes a 137-amino-acid protein with a molecular mass of 15.4 kDa. The protein consists of two domains: an N-terminal oligomerization domain and a C-terminal DNA-binding domain (36, 46). H-NS binds to DNA in a non-sequence-specific manner but with a preference for intrinsically curved, double-stranded DNA and results in the compactation of the chromosome (43, 44) and in transcriptional repression of many target genes. Mutations in the hns gene affect several cellular processes including recombination, transposition, and deletion events (13, 38) and leads to a highly pleiotropic phenotype. Expression of hns is activated by the FIS protein and is repressed by autoregulation (12). Furthermore, hns expression is induced at low temperatures by the cold shock protein CspA (23), but the H-NS protein also mediates temperature regulation itself, as observed, for instance, for the pap operon encoding digalactoside binding pyelonephritis-associated pili (15). H-NS responds to environmental conditions and also affects many genes involved in adaptation to a changing environment, such as osmolarity, pH, and anaerobiosis, thereby indicating a central role for H-NS in the adaptability of bacterial cells (17).

In addition to H-NS, Escherichia coli cells express a paralogous protein called StpA, which was first discovered as a multicopy suppressor of a thymidilate synthase-negative phenotype in a T4 phage (48). StpA is a 133-amino-acid protein that shares 58% sequence identity to the H-NS protein. It can bind DNA with even higher affinity than H-NS (41) and displays an additional RNA chaperone activity (48). Overexpression of StpA can in some parts complement the lack of H-NS, as also shown for the regulatory effect on proU and pap genes as well as the hns gene itself (41). Mutations in the stpA gene, however, do not result in a notable phenotype under standard growth conditions. This can be partly explained by the low level of stpA expression in wild-type cells, yielding approximately 200 monomers of the StpA protein per cell, whereas H-NS is usually present in 20,000 monomers per cell (43). The level of StpA increases during growth in minimal medium due to the action of Lrp or after a temperature upshift. Furthermore, cross-regulation between H-NS and StpA exists, resulting in higher StpA levels in an hns mutant, which can compensate for some of the effects of the mutation (40). In addition to its role as a molecular backup for H-NS, StpA might play a role in cooperation with H-NS, since the two proteins were found to form heteromers (19). As part of this heteromeric complex, StpA is protected from degradation by the Lon protease (20), and this way is also embedded in the regulatory network of DNA-binding proteins.

Although the effects of hns and stpA mutations are well characterized in nonpathogenic E. coli K-12 strains (6, 17), the effects in pathogenic strains have been studied only for some selected virulence traits. Using DNA macroarrays, reverse transcription-PCR (RT-PCR), and different phenotypic tests, we monitored the effects of H-NS and StpA in a uropathogenic E. coli isolate on a larger scale and assessed their role in overall virulence in an experimental murine model of ascending urinary tract infection and sepsis.

Here, we provide evidence that H-NS is a global regulator of gene expression in E. coli strain 536, affecting not only the “core” genome but also many genes encoded on the pathogenicity islands of this strain, thereby linking these genes to the cellular regulatory network. The classical virulence factors were found to be regulated exclusively by the H-NS protein, whereas no role for StpA could be observed under these conditions. However, since the expression pattern of an hns stpA double mutant differed from that of the hns single mutant, we suggest the existence of a second regulon that is dependent on the presence of both regulators, presumably in the form of a heteromeric complex.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The clinical UPEC strain 536 (5) and its isogenic hns, stpA, and hns stpA mutants were used (Table (Table1).1). Bacteria were grown with aeration at 37°C in Luria broth (LB) until the optical density at 600 nm reached 0.6. For the detection of siderophores, cultures were grown overnight in MM9 medium (37), and free iron was removed during 3 h by the addition of 0.1 mM of the chelator 2,2′-dipyridyl.

TABLE 1.
Bacterial strains and plasmids used

Mutagenesis.

The construction of the mutants was performed using linear DNA for recombination, as described previously by Datsenko and Wanner (8). Briefly, a linear DNA fragment containing a cat cassette flanked by FLP recognition target (FRT) sites and 45-nucleotide homologous extensions to the target genes (up- and downstream of hns and stpA, respectively) was amplified by PCR using plasmid pKD3 as a template. Strain 536 wild-type cells were first transformed with the pKD46 helper plasmid and afterwards with the linear PCR fragment in the presence of arabinose in order to induce the recombinase from the helper plasmid, thereby mediating the exchange of the wild-type allele with the cat insertions. The cat cassette was removed with the help of the FLP recombinase (encoded on plasmid pCP20), which mediates recombination between the two FRT sites flanking the cat cassette, thus leaving behind a complete deletion of the open reading frame. The double mutant was constructed in the same manner by introducing the ΔstpA::cat fragment into the newly constructed Δhns mutant. All mutations were confirmed by both PCR and Southern hybridization.

RNA isolation.

Total RNA was extracted from mid-log-phase cultures as described previously (7). Contaminating DNA was removed by DNase I (Roche Diagnostics, Mannheim, Germany) digestion for 1 h at 37°C, followed by RNA cleanup using the RNeasy Mini kit (QIAGEN, Hilden, Germany). For macroarray experiments, mRNA was enriched from 10 μg total RNA using the MICROBExpress kit (Ambion, United Kingdom) according to the manufacturer's instructions.

Semiquantitative RT-PCR.

For cDNA synthesis, 4 μg of total RNA was mixed with 1 μg of random hexamer primers (GE Healthcare, Amersham Biosciences, Freiburg, Germany). After primer annealing, a mix of deoxynucleotides, dithiothreitol, first-strand buffer, 40 U RNase OUT recombinant RNase inhibitor (Invitrogen, Germany), and 200 U Superscript III reverse transcriptase (Invitrogen) was added according to the manufacturer's recommendations. cDNA synthesis was performed at 52°C for 60 min, followed by heat inactivation at 70°C for 15 min. cDNA samples were 20× diluted in water and directly used for PCR amplification. As a control for DNA contaminations, a second PCR was performed using total RNA without any reverse transcription reaction. For the adjustment of cDNA amounts, 16S rRNA (rrsA) was used as an internal standard. The sequences of all oligonucleotides used are listed in Table Table22.

TABLE 2.
Oligonucleotide sequences

Macroarray hybridizations.

For expression profiling, both K-12-specific arrays (Panorama E. coli gene arrays from Sigma-Genosys, Cambridge, United Kingdom) and a 536-specific pathoarray (9) were used. Total RNA was prepared from mid-log-phase cultures grown in LB at 37°C, and at least four independent hybridizations were performed for each mutant.

cDNA synthesis was performed using 1.5 μg of enriched mRNA together with 750 ng random hexamer primers (Amersham Biosciences) in a total volume of 26 μl. After primer annealing, 12 μl of a mixture containing 0.5 mM of the deoxynucleotides dCTP, dGTP, and dTTP; 20 μCi [α-33P]dATP (Amersham Biosciences); 80 U RNase OUT recombinant RNase inhibitor (Invitrogen); and 8 U Omniscript reverse transcriptase (QIAGEN) in 1× reaction buffer was added to the RNA to a total volume of 40 μl. The reaction mixtures were incubated at 42°C for 2 h. Unincorporated nucleotides were removed from labeled cDNA using Microspin G-50 columns (Amersham Biosciences). Incorporation rates ranged from 58 to 65%. Following 10 min of denaturation at 93°C in hybridization buffer, labeled cDNA was added to the prehybridized membranes and incubated for 16 h at 65°C in a hybridization oven. After washing, the membranes were sealed in plastic bags and exposed to a PhosphorImager screen (Molecular Dynamics) for 48 h. Buffer composition and conditions for hybridization, washes, and stripping are described in the Gene-Array manual (Sigma-Genosys).

Data analysis.

Exposed PhosphorImager screens were scanned on a Typhoon 8600 variable-mode imager (Molecular Dynamics) at a 50-μm resolution. Spot intensities were measured with the ArrayVision software (Imaging Research, St. Catharines, Canada) using the overall spot normalization function of the program. Mean values were calculated from the duplicate spots of each gene. Only values with intensities that were twofold higher than the background were included in the analysis.

For each mutant, four independent hybridizations were performed, with each one paired with the wild type. For the calculation of intensity ratios, spot intensities of the mutants were directly compared to the intensities of the wild type for the corresponding hybridization experiment. As an indicator for the reproducibility of the procedure, the four wild-type experiments from each set of data were compared to each other, showing a high degree of regression in all three cases (data not shown), thereby validating the quality of the hybridizations.

Statistical analysis was performed using the SAM software (Statistical Analysis of Microarrays; Stanford University) as described previously by Beloin et al. (4). Data sets were composed as two-class, unpaired data in four blocks for each experiment. After the analysis was finished, all genes that were significantly deregulated by a factor of at least ±2 were listed by the program. Here, a significant deregulation is defined by a false discovery rate of at least less than 10% for each gene. In addition, a two-tailed Student's t test was performed on the significant called genes, comparing the four intensity readings of each gene between the mutants and the wild type and thus giving rise to the level of significance (P values).

Phenotypic testing.

Several classical tests were performed to study the phenotypes of the mutants. Motility was assessed on 0.3% swarming agar plates. Hemolytic activity was studied both on 5% blood agar plates and in a liquid hemolysis assay as described previously by Nagy and coworkers (29), with minor modifications. Briefly, 500 μl supernatants from mid-log-phase cultures grown at 37°C was mixed with an equal volume of washed bovine erythrocytes in a 150 mM NaCl-20 mM CaCl2 solution. Samples were incubated for 30 min at 37°C prior to centrifugation. The amount of hemoglobin released during hemolysis was measured at A545 by using a saponin-treated sample as a reference for 100% hemolysis. The production of siderophores was assessed in a chrome azurol S (CAS) liquid assay as described previously by Schwyn and Neilands (37). Supernatants from mid-log-phase cultures grown in iron-depleted MM9 medium were mixed with an equal volume of CAS assay solution containing a blue Fe-CAS complex, which turns red when siderophores are present. The degradation of the complex was monitored photometrically at A630. Survival at high osmolarity was assessed by determining of the number of CFU before and after a dilution of cultures grown overnight in LB with 1,000× fresh LB medium supplemented with 3 M NaCl and incubation of cells at 37°C for 2 h.

Western blotting.

Whole-cell extracts from bacterial cultures were prepared by pelleting the cells and adjusting optical densities in an appropriate volume of buffer. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed as described previously by Laemmli (22) using 15% polyacrylamide gels. Samples were transferred onto nitrocellulose membranes using a semidry blotting apparatus. After blocking the membranes overnight in Tris-buffered saline containing 0.1% Tween 20 (TBS-T) and 5% skim milk, separate membranes were incubated for 1 h at room temperature with Prf- and SfaI-specific antisera in 2,000× dilutions in TBS-T containing 5% skim milk. After three 15-min washes in TBS-T, membranes were incubated for 1 h with 5,000×-diluted anti-rabbit immunoglobulin-horseradish peroxidase conjugate (Dianova, Hamburg, Germany). After further washing, membranes were developed using the enhanced chemiluminescence kit (Amersham Biosciences).

Atomic force microscopy.

Bacteria grown on LB agar plates were resuspended in water and immediately dropped onto freshly cleaved ruby red mica. After a 5-min incubation at room temperature, excess bacteria were washed off with filtered water, and the samples were dried in argon gas overnight before they were scanned with an atomic force microscope (NanoscopeIII; Digital Instruments, Santa Barbara, Calif.) using the tapping mode as described previously (3).

Animal experiments.

Animal experiments were conducted according to the principles set forth in the Guide for the Care and Use of Laboratory Animals in a laboratory authorized by Hungarian rule (decree no. XXVII. 1998) and by subsequent regulation (government order no. 243/1998).

As a model for ascending urinary tract infections, 3- to 4-day-old NMRI mice were infected intravesically, as described in detail previously (29). In order to determine 50% lethal dose (LD50) values for E. coli strain 536 and its derivatives, groups of mice were inoculated with log dilutions of washed bacteria. Lethality was determined daily for 3 weeks. LD50 values were calculated by using the method of Reed and Muench (35).

In the murine model of sepsis, groups of 10 female 8-week-old NMRI mice were infected intravenously through the tail veins with 0.5 ml of inocula containing either 109 CFU (4 LD50 for the wild-type strain) or 108 CFU (0.4 LD50 for the wild-type strain) of washed bacteria. Deaths were recorded every 30 min within the first 6 h postinfection and twice daily afterwards.

The mouse lung toxicity assay was performed as described previously (29). Briefly, groups of five NMRI mice were instilled with 50 μl of cultures grown overnight under superficial ether anesthesia. Lethality was monitored for 24 h (every 30 min during the first 6 h, hourly between 6 and 12 h, and at 24 h postinfection).

The log-rank test (Mantel-Haenszel test) was used for statistical analysis of survival curves obtained from the animal experiments. Kaplan-Meier survival curves were created and analyzed using Prism 4.0 software (GraphPad).

RESULTS

The lack of H-NS has a severe impact on the gene expression pattern of E. coli strain 536.

To date, no comprehensive study on the role of H-NS in gene expression in a pathogenic E. coli strain has been carried out. The same is true for the H-NS paralogue, StpA, whose mutants still lack a determined phenotype. Therefore, hns and stpA single mutants as well as an hns stpA double mutant were constructed using the uropathogenic E. coli isolate 536. The expression of the K-12-like “core genome” was monitored using a commercially available K-12 DNA macroarray. At the same time, a significant fraction of the “patho gene” pool, represented on an E. coli “pathoarray” containing most genes encoded on pathogenicity islands I to V of strain 536, was monitored in these strains. The complete DNA array data set can be found at the website for the E. coli research group of the Institute for Molecular Infection Biology (http://www.infektionsforschung.uni-wuerzburg.de/).

In the hns single mutant, 551 of the 4,290 translatable open reading frames on the K-12 array were determined to be significantly deregulated compared to the wild type. Regarding the genes specific for E. coli strain 536 on the pathoarray, 37 of the 391 open reading frames were significantly deregulated in the mutant (Fig. (Fig.1A).1A). In detail, 515 of the K-12 genes were upregulated, whereas 36 genes were downregulated in the absence of H-NS. For the E. coli pathoarray, all 37 differently expressed genes in the hns mutant were significantly upregulated compared to the wild type. These results confirm the role of H-NS as a global regulator in E. coli acting mostly as a repressor of gene expression.

FIG. 1.
(A) Total numbers of altered genes in the hns mutant (shaded) and the hns stpA mutant (white) determined by expression profiling using a K-12-specific array (left) and an E. coli pathoarray (right). ORF, open reading frame. (B) Distribution of the deregulated ...

In contrast, not a single gene appeared differently regulated in the stpA single mutant. Therefore, genes that may be regulated by StpA alone and that finally might result in a notable phenotype remain to be identified.

In the hns stpA mutant, 389 genes of the K-12 genome were deregulated compared to the wild type. As in the hns mutant, most of the affected genes had a higher expression level in the mutant (323 upregulated versus 66 downregulated genes on the K-12 array). This was also true for the virulence-associated genes on the E. coli pathoarray, where all of the 40 genes with altered expression were upregulated compared to the wild type.

From these results, one could speculate that all differences in gene expression in the hns stpA mutant might be due to the lack of H-NS. However, when the data set of the hns single mutant was compared to that of the hns stpA mutant, only a subset of genes was commonly altered in both mutants, especially with regard to the core genome (Fig. (Fig.1A).1A). This led us to take a closer look at the obvious differences between the hns and the hns stpA mutants.

hns and hns stpA mutants differ in their expression patterns.

Even though the expression pattern of both the hns single mutant and the hns stpA double mutant were very similar with regard to their total numbers of deregulated genes, their distributions into functional groups exhibited some differences (Fig. (Fig.1B).1B). The distribution of the upregulated genes differed only slightly in both mutants, with one exception: significantly more genes involved in translation and posttranslational modifications were upregulated in the hns mutant than in the hns stpA mutant (86 versus 4 genes). This group included genes encoding ribosomal proteins and tRNA synthetases, thereby indicating a high level of protein synthesis in the absence of H-NS.

The most prominent difference in the hns stpA double mutant was the occurrence of a large number of downregulated genes. Most of these genes clustered in the group of transport and binding proteins (10 genes) as well as genes involved in all kinds of metabolic processes like energy metabolism and central intermediary metabolism (12 genes). This effect on metabolism was also reflected in the reduced growth rate of the double mutant. Growth experiments with strain 536 and its mutants at 37°C in LB medium revealed an impaired growth rate in the hns stpA mutant, with a generation time of 43 min compared to about 32 min in the wild type and the stpA single mutant. On the other hand, despite a prolonged lag phase, the generation time of the hns mutant during exponential growth turned out to be the same as that of the wild type, thereby suggesting some compensatory effects of StpA in the hns mutant.

Taken together, these results suggest the existence of two different regulons, depending on either the H-NS protein alone or a concerted action of H-NS and StpA.

H-NS acts as a repressor of all major virulence factors.

Despite the differences in metabolism and translation, the hns and the hns stpA mutants behaved very similarly with regard to the expression of virulence-associated genes. The expression pattern on the E. coli pathoarray revealed major overlaps in both mutants (Fig. (Fig.1A).1A). All 33 commonly affected genes on the E. coli pathoarray were upregulated compared to the wild type and included the major virulence factors of strain 536.

Motility and chemotaxis.

The effect of H-NS on motility is a rare example of an activating role of H-NS and has already been described for several strains (42). This was also true for strain 536, where both the hns and the hns stpA mutants were completely nonmotile on swarming agar plates (data not shown). Expression profiling supported the experimental data and further revealed that a lack of motility coincided with a downregulation of genes involved in chemotaxis, like cheW, cheA, cheZ, and cheY. In total, 32 genes of this functional group were downregulated. The strongest effect was observed for the flagellin FliC itself, where the expression was downregulation by a factor of 12.6 (P = 0.028) in the hns mutant and by a factor of 9.3 (P = 0.003) in the hns stpA mutant, respectively. Taken together, our results were in good agreement with previous observations of other E. coli strains, including K-12, and further confirm the positive regulation of motility by H-NS.

Adhesins.

Strain 536 encodes at least three types of fimbrial adhesins, namely, the chromosomally encoded type 1 fimbriae (fim), P-related fimbriae (prf) encoded on PAIII, and S-fimbrial adhesin (sfa) encoded on PAIIII.

The effect of H-NS on type 1 fimbria expression has already been described in the literature, where H-NS was shown to repress the recombinases fimB and fimE (33). Both recombinases are involved in the phase variation of an invertible DNA fragment containing the fimA promoter, thereby mediating switching between a fimbriated and a nonfimbriated state. In strain 536, the same effect of H-NS on the expression of the recombinases could be observed, resulting in a significant upregulation of fimB by a factor of >4 and of fimE by a factor of >2 in both hns and hns stpA mutants (hns-fimB, P = 0.04; fimE, P = 0.10; hns stpA-fimB, P = 0.002; fimE, P = 0.001) (Fig. (Fig.2A).2A). The expression of the structural genes was not significantly altered at the transcriptional level.

FIG. 2.
Expression levels of the major virulence factors of UPEC strain 536. Signal intensities derived from the array hybridizations for the three different fimbrial determinants fim (A), sfa (B), and prf (C) as well as for the alpha-hemolysin determinant hly ...

S fimbriae are found mainly in uropathogenic E. coli strains and in strains causing newborn meningitis (26). These fimbriae bind to α-sialyl-2,3-β-lactose-containing receptors and are encoded in a cluster consisting of nine genes, where sfaA encodes the major subunit and sfaS encodes the specific adhesin. SfaC and SfaD are regulatory proteins that are required for the expression of the structural genes (26). By using the E. coli pathoarray, we were able to monitor the expression level for each gene of the sfa cluster in an hns mutant strain. All structural genes of the cluster were significantly upregulated in the mutants compared to the wild type (Fig. (Fig.2B).2B). The most prominent derepression could be observed for the major subunit SfaA. The expression of the sfaA gene was 27 times higher in the hns mutant than in the wild type (P = 0.07) and 12 times higher in the hns stpA mutant (P = 0.004), which ranked sfaA among the 10 most strongly affected genes sorted according to their expression levels. When the expression of SfaA was monitored at the protein level by using specific antisera, we also saw a massive derepression in both hns mutants (Fig. (Fig.3A),3A), thereby confirming the results from transcriptional profiling.

FIG. 3.
Hyperfimbriated phenotype of hns mutants. (A) Protein levels of major subunits of S and P fimbriae in mid-log-phase cultures monitored by Western blotting. (B) Morphology of a wild-type cell (upper panel) and an hns mutant cell (lower panel) from agar ...

In contrast to type 1 and S fimbriae, the regulation of the P-related fimbriae (prf) in strain 536 is not well characterized to date. These fimbriae are encoded by a cluster comprising 12 genes, which is very similar to the pap determinant of other UPEC strains. PrfI and PrfB are regulators with homology to SfaC and SfaB, thereby being able to cross-react and activate transcription of the sfa cluster (27). We now found that, like for the pap genes, H-NS also affects the expression of P-related fimbriae. Western blotting using PrfA-specific antiserum showed that the expression of the major subunit was increased in H-NS-deficient strains compared to the wild type (Fig. (Fig.3A).3A). When gene expression levels were compared using the E. coli pathoarray, the derepression of prfA was not as drastic as that seen at the protein level (Fig. (Fig.2C).2C). However, when signal intensities on the array were compared, it was clear that the overall level of expression of P-related fimbriae was very low under these conditions, i.e., mid-log-phase samples from liquid cultures. Only the prfX gene, encoding a putative regulator, was significantly upregulated in both mutants, with expression ratios of 5.2 in the hns mutant (P = 0.021) and 3.7 in the double mutant (P = 0.009). This gene encodes a 17-kDa protein that is considered to be a putative regulator of the prf operon.

To verify whether the derepression of the genes for both S and P fimbriae at the transcriptional level also resulted in increased biogenesis of fimbriae, cells grown on agar plates were fixed onto mica plates and analyzed by atomic force microscopy. As seen in Fig. Fig.3B,3B, the hns mutant cell was heavily fimbriated, in contrast to the wild type.

Taken together, our results infer a major role for H-NS in regulating the expression of both P and S fimbriae that are believed to be important virulence factors of strain 536 during urinary tract infections.

Hemolytic activity.

Strain 536 possesses two hly determinants encoding alpha-hemolysin. Both gene clusters are located on pathogenicity islands (PAII and PAIII), and the biological function of having two copies of hlyCABD is still unknown. Both clusters encode the RTX toxin HlyA, the activator HlyC, and the HlyB and HlyD proteins, involved in toxin secretion. On the E. coli pathoarray, both transport proteins HlyB and HlyD were significantly upregulated in H-NS-deficient strains, and even HlyC showed a trend towards higher expression in the mutants (Fig. (Fig.2D).2D). The level of the hlyA transcript was not altered under these conditions. However, when hemolytic activity from culture supernatants was measured, the values were six times higher in the hns mutant than in the wild type and even 12 times higher in the hns stpA mutant (P < 0.003 each) (data not shown). This effect of H-NS on hemolysin expression is consistent with previous findings described in the literature (25, 31), where H-NS was shown to mediate thermoregulation by binding to the regulator Hha. In our hns mutants, the expression of hha itself was also increased by a factor of 4.7 (hns, P = 0.058) and a factor of 2.1 (hns stpA, P = 0.017), respectively.

Iron uptake systems.

Strain 536 encodes five types of iron uptake systems. The salmochelin system (iro) and the yersiniabactin system (ybt) are both encoded on large pathogenicity islands, and the hemin receptor (chuA) locus resides on a small islet not present in K-12, whereas the determinants coding for the ferrichrome (fhu) and the enterobactin system (fep) are located within the E. coli core genome. Each system consists of proteins involved in the binding of the siderophore/hemin by a receptor and distinct proteins for the transport of the complexes into the cell. When the expression levels of the various gene products were monitored using the DNA arrays, expression of almost all genes stayed below the detection limit, which is due to the noninducing conditions. Only the gene coding for the salmochelin receptor, iroN, was clearly expressed, with 3.4-times-higher levels in the hns mutants (hns, P = 0.045; hns stpA, P = 0.002). When the more sensitive method of semiquantitative RT-PCR, which enables the detection of low transcript amounts, was used, we also found a significantly higher amount of product from the salmochelin receptor iroN in the case of the hns-deficient strains compared to the wild type. On the contrary, transcript amounts of the hemin receptor chuA appeared much lower in these strains (Fig. (Fig.4A).4A). The expression of the enterobactin receptor fepA was not altered by the mutations, and expression of the yersiniabactin receptor fyuA stayed undetectable under these conditions. These results indicate that H-NS has divergent effects on the expression of the different siderophore receptor genes, playing both an activating role and a repressing role.

FIG. 4.
H-NS-dependent siderophore production in E. coli strain 536. (A) Expression levels of the three different siderophore receptor genes of the H-NS-dependent siderophore systems determined by semiquantitative RT-PCR using RNA extracted from LB cultures and ...

To further study this effect, a chromatographic assay for the detection of siderophores was performed (see Material and Methods). In this assay, a blue iron complex in the assay solution is degraded gradually if siderophores have accumulated in the supernatants of liquid cultures and consequently results in decreased absorbance. As shown in Fig. Fig.4B,4B, all strains reach the same absorbance at a state of equilibrium, thereby indicating an equal amount of siderophores in the supernatants. However, the decrease in absorbance happened much faster in H-NS-deficient strains. When RNA from cultures grown under the same conditions as those for the liquid assay (i.e., minimal medium overnight) was prepared and RT-PCR analysis of the expression of the receptor genes was performed, a similar result as that for the mid-log-phase cultures grown in LB was obtained, with iroN exhibiting the highest expression level (data not shown). This result suggests that different types of siderophores with different affinities for iron might be involved. Taken together, our data infer an altered composition of secreted siderophores in strains lacking H-NS.

Stress resistance.

H-NS is known to be involved in the adaptation to various types of stress situations (2, 17). This role was also found when the hns mutants of strain 536 was analyzed. Many genes involved in stress adaptation and multidrug resistance were found among the list of upregulated genes on the K-12 array. In particular, many of these genes were involved in acid resistance, like the glutamate decarboxylases/antiporters gadA, gadB, and xasA and their activators gadX (yhjX) and gadE (yhjE); the lysine decarboxylases cadA and cadB; and the acid chaperones hdeA and hdeB, which were all among the 10 most strongly upregulated genes, and many genes that are located within the “acid cluster,” like yhiQ, yhiD, and yhjHJKL. Derepressed genes involved in multidrug resistance included mdtEF and emrKY, as was previously found by Nishino and Yamaguchi (32). Our data also supported the well-known role of H-NS in osmoprotection. The glycine-betaine/proline transporters proVWX were significantly upregulated in the hns mutants, with proV ranking at positions 4 (a factor of 16.97 for hns; P = 0.05) and 15 (a factor of 10.42 for hns stpA; P = 0.003) among the 50 most strongly upregulated genes. Other osmotically inducible proteins like those encoded by osmC, osmB, and osmY as well as the chaperones dnaKJ showed an increased expression in the mutants compared to the wild type.

To test whether this upregulation of osmoprotection genes also resulted in an increased survival rate at high osmolarity, LB cultures were incubated with 3 M NaCl for 2 h, and the amount CFU was determined before and after the treatment (17). Only 4% (±2.3%) of the wild-type cells and 5% (±3.9%) of the stpA mutant cells were still alive after 2 h of incubation at a high osmolarity, whereas the survival rates were significantly increased in both the hns single mutant (61% ± 17.8%) and the hns stpA double mutant (69% ± 13.7%).

Taken together, our data show that H-NS is regulating many of the known virulence factors of strain 536, acting mostly as a repressor. Therefore, mutants deficient for H-NS are highly hemolytic, hyperfimbriated, and very stress resistant and have a different siderophore expression kinetic but are nonmotile and have a growth deficiency in the early growth phases. To examine the combined effect of these changes on virulence-associated gene expression, the actual virulence properties of the strains were tested using animal infection models.

Lack of H-NS affects in vivo virulence.

To assess whether the loss of H-NS and/or StpA has any impact on the urovirulence of strain 536, mortality caused by isogenic mutants was compared to that caused by the parental wild-type strain in the infant mouse model of ascending urinary tract infection. In agreement with previously reported data (30), we determined an LD50 value of approximately 5 × 103 CFU/ml for the wild-type strain using this model. Neither the hns and the stpA mutants nor the hns stpA double mutant showed drastic deviations in virulence (Table (Table3),3), especially when the high level of natural variations of this approach was taken into account. There was, nevertheless, a clear difference in the time course of lethality elicited by the different mutants.

TABLE 3.
Virulence of E. coli strain 536 and its isogenic derivatives in the infant murine model of ascending urinary tract infection

At a lethal infectious dose (107 CFU/ml), all mice infected with H-NS-deficient strains (hns and hns stpA mutants) died within the first 24 h postinfection, whereas only 17% lethality was observed within the same time period among mice infected with the wild-type strain, while most (83%) of these mice died only after 48 h. Deletion of stpA, on the other hand, did not principally influence the lethality time course. Since alpha-hemolysin is a key virulence factor in this model (30), overexpression of hly determinants upon the loss of H-NS might be responsible for the more expeditious lethality observed in hns mutants. This hypothesis is supported by the highly elevated LD50 value exhibited by the nonhemolytic variant of the hns mutant of strain 536 (Table (Table33).

Acute mortality resulting from urinary tract infection, in most cases, is a consequence of disseminating infection leading to urosepsis. To test whether the lack of H-NS and/or StpA affect this final stage of the infectious process, mice were challenged intravenously. In a pilot study, the LD50 value of the wild-type strain in this model was determined to be 2.5 × 108 CFU (data not shown). All mice challenged with 109 CFU (four times the LD50) of the wild-type strain as well as the hns, stpA, and hns stpA mutants died within 6 h postinfection (Fig. (Fig.5A).5A). The time course of lethality, however, showed great differences in this model as well. Mutants lacking hns elicited severe symptoms very rapidly. As early as 30 min postinfection, 90% and 100% mortality rates were detected among mice infected with the hns and the hns stpA mutant, respectively. On the contrary, both the stpA mutant and the parental wild-type strain caused a more lingering infectious process (Fig. (Fig.5A).5A). This difference may originate from the upregulation of alpha-hemolysin in the hns mutants, as a nonhemolytic variant of 536Δhns elicited a highly significantly different (P < 0.001) time course of lethality from all investigated strains (Fig. (Fig.5A5A).

FIG. 5.
Survival curves of mice infected intravenously with E. coli strain 536 or its isogenic mutants. Groups of 10 mice were infected with either 109 CFU (A) or 108 CFU (B) of washed bacteria through the tail veins.

Mice challenged with 108 CFU (0.4 times the LD50) of the same bacteria showed major differences in mortality rates (Fig. (Fig.5B).5B). Sixty percent and 80% of mice infected with the wild-type strain and the stpA mutant, respectively, died. Infection with the same number of the hns and hns stpA mutants, however, resulted in only 20% and 10% mortality, respectively. The survival curve of the double mutant was found to be statistically different, whereas that of the hns mutant was on the borderline of being significantly different in comparison to the survival curve of the wild-type strain (P = 309, P = 0.058, and P = 0.017 in case of the stpA, hns, and hns stpA mutants, respectively).

The results described above suggested that the hns mutant is able to elicit a more expeditious lethality when injected at a high number, due to its increased alpha-hemolysin production. To test this possibility, an HlyA-dependent in vivo toxicity assay was performed. As shown in Fig. Fig.6,6, the hns mutant caused a significantly more rapid lethality than its parental wild-type strain (P = 0.027) in the mouse lung toxicity assay. Loss of StpA, on the other hand, had no significant influence on this virulence characteristic.

FIG. 6.
Time course of lethality in the mouse lung toxicity assay. Groups of five mice were instilled intranasally with E. coli strain 536 or its isogenic mutants.

DISCUSSION

In this study, we have explored the role of the DNA-binding proteins H-NS and StpA in virulence by using expression profiling as well as classical phenotypic tests and in vivo experiments. By the use of a specific pathoarray, we identified some previously unknown members of the H-NS regulon encoded on the pathogenicity islands of strain 536. We obtained proof that H-NS is playing a major role in the regulation of PAI-encoded genes, thereby linking those genes to the cellular regulatory network.

H-NS was found to be a global regulator in UPEC strain 536, affecting the transcription of more than 500 genes, among them all classical virulence factors tested. Many of the genes of E. coli K-12 whose transcription has been described to be affected upon hns mutation were also identified in our study (see also the complete DNA array data set at http://www.infektionsforschung.uni-wuerzburg.de/). Mutants deficient for H-NS expressed more fimbriae and hemolysin and differed in their siderophore expression levels. Furthermore, these mutants were nonmotile and showed increased resistance to high osmolarity.

These results are in good agreement with data from other E. coli strains. The downregulation of genes involved in motility and chemotaxis is a well-known phenotype of hns mutants and is one of the relatively few examples of an activating role of H-NS (42). In a recent study, Snyder and coworkers (39) demonstrated for the UPEC isolate CFT073 that flagella are not needed for a successful colonization of the urinary tract. Therefore, the lack of motility does not imply a disadvantage for the mutants during infection. Those same authors postulated a major role for type 1 fimbriae in the in vivo colonization process of the mouse urinary tract. In our strain, we could not find upregulation of the fim operon, which is probably due to the unfavorable growth conditions for expression of type 1 fimbriae. However, H-NS affected expression of both recombinases fimB and fimE, as previously described by Olsen and Klemm (33). The consequences for the in vivo situation still need to be assessed.

Strain 536 possesses at least two additional types of fimbriae, namely, SfaI and Prf. At least for the S pili, regulation by H-NS has been described previously (26). In this study, we were able to analyze the mechanism of regulation. Repression by H-NS acts on the transcription at both the sfaB and sfaA promoters, resulting in the derepression of the activator SfaB and the structural genes in the hns mutant, as seen on the array. The SfaB protein itself activates both the sfaC and sfaB promoters, thereby further enhancing transcription of the structural genes. Hence, H-NS is acting both directly and indirectly via SfaB.

For the P-related fimbriae, we could not demonstrate increased transcription of the structural genes on the array, even though the protein levels were strongly increased in the mutant. Interestingly, only the prfX gene was significantly upregulated at the transcriptional level. This gene encodes a 17-kDa protein that is believed to be a putative regulator of the prf operon. Therefore, one could speculate that one of the effects exerted by H-NS might be indirect by decreasing the expression of PrfX, which would then activate the prf structural genes. Another important aspect is cross talk between both types of fimbriae: the regulators of both gene clusters are very homologous and can complement each other (27). Therefore, upregulation of the regulator SfaB in the hns mutant might also indirectly result in the activation of the prf operon. Finally, we conclude that the increased amount of (so-far-unidentified) fimbrial adhesins on the cell surface of the mutants, as verified by atomic force microscopy, presumably contributes during the first stages of infection by virtue of increased adhesion to the urothelium.

An increased production of alpha-hemolysin in strains deficient for H-NS has previously been described for E. coli K-12 (31). We could confirm these results in strain 536 both by RT-PCR and in a liquid assay. Interestingly, we also found upregulation of the hemolysin repressor Hha, thereby confirming that Hha alone is not able to mediate repression of the hemolysin operon, but in contrast, a functional H-NS-Hha complex is required (31). Furthermore, derepression of the hly determinant, as well as the amount of secreted alpha-hemolysin, was twice as high in the hns stpA double mutant than in the hns single mutant, thereby suggesting that StpA is able to partially substitute H-NS function as a repressor of hly genes.

In addition to alpha-hemolysin, some enterobacteria express another type of pore-forming toxin, namely, ClyA (also known as SheA/HlyE), which could also contribute to the hemolytic phenotype (34), and clyA expression was found to be silenced by H-NS in nonpathogenic E. coli K-12 (14, 47). However, in UPEC strain 536 as well as in most other uropathogenic isolates, a large portion (217 bp) of the clyA gene is deleted (21, 24). Therefore, we assume that alpha-hemolysin is solely mediating the hyperhemolytic phenotype of the hns mutants.

To date, no effect of H-NS on iron uptake systems has been reported. In this study, we now demonstrate that the effects on the five siderophore receptor genes of strain 536 are divergent: H-NS is playing both an activating and a repressing role, as seen, e.g., for the chuA and iroN genes encoding the hemin receptor and the salmochelin receptor, respectively. When accumulated siderophores in culture supernatants were analyzed, the total amount of secreted siderophores was the same in both the wild type and the mutants. Because of the altered kinetic in the mutant, however, one could speculate that the composition of siderophores differs from that of the wild type. For instance, the upregulation of a siderophore with a high affinity for iron in the hns mutant could explain the faster decrease in absorbance detected in the photometric assay. However, it is not clear whether these observations result from an altered expression of the iron uptake systems per se or are due to specific effects of H-NS on the secretion machinery, since many transport and binding proteins are affected by the mutation.

Another aspect of the role of H-NS in the virulence and fitness of strain 536 is stress resistance. We could confirm the array data by demonstrating a survival rate of the mutant that is more than 10 times higher at high osmolarity, as has previously been described for E. coli K-12 (17). These results suggest that in uropathogenic E. coli strains, H-NS plays a major role in the adaptation to various stress situations like high osmolarity and low pH. Since uropathogenic bacteria are even more likely to be confronted with these stress conditions, these findings are of particular importance and might also contribute to urovirulence.

Although our results suggested some benefits for the hns mutant for in vivo virulence, the animal experiments revealed an ambiguous role of H-NS in pathogenesis. At infective doses exceeding the LD50 values, the loss of H-NS resulted in more expeditious lethality in both the urinary tract infection and the sepsis models. This was proposed to be the result of increased alpha-hemolysin production by the hns mutants, which caused acute toxicity of mice, leading to a rapid fatal outcome. In support of this hypothesis, greatly reduced virulence was detected for an isogenic nonhemolytic variant of 536Δhns, implying a prominent role for alpha-hemolysin in both models. Moreover, we showed that the inactivation of hns resulted in a significantly higher in vivo toxicity in the alpha-hemolysin-dependent mouse lung assay.

At lower infectious doses, the situation changes. In the case of the urinary tract infection model, no significant difference in the virulence of hns mutants could be shown in comparison to the wild-type strain. Moreover, when injected intravenously, the loss of H-NS significantly reduced the chance of eliciting lethal infection in mice. This contradiction in the role of H-NS at various infectious doses could be resolved by taking into consideration the altered growth ability of an hns mutant, which is even more pronounced in minimal medium or at lower temperatures. Although usually not considered to be a classical virulence characteristic, an optimal rate of bacterial replication in vivo may highly influence the outcome of an infection. We suppose that a slower growth rate of hns mutants in the bladder could be compensated for by the upregulation of various “classical” virulence factors, such as fimbriae, alpha-hemolysin, and osmotic resistance. Upon reaching the bloodstream, however, bacteria have to face severe menace by the immune system, which could be overcome only through rapid bacterial growth. Since the role of the upregulated virulence factors seems to be minor within the bloodstream, hns mutants become attenuated compared to their wild-type strain. These observations emphasize the role of housekeeping genes in bacterial fitness, which is a prerequisite of virulence. The impact of H-NS on virulence appears to be divergent along the infectious process. This may explain the complex regulation of hns reviewed elsewhere previously by Atlung and Ingmer (2). Nevertheless, a role for H-NS in virulence regulation is without doubt.

Interestingly, most changes in expression of virulence-associated genes seem to be solely dependent upon the lack of H-NS. The paralogous protein StpA does not seem to be important in the context of virulence under the conditions tested. However, StpA seemed to compensate for some effects of the hns mutation, as seen for alpha-hemolysin production, and also has some modulating effects, e.g., for expression of adhesins. Moreover, the expression patterns of the hns single mutant and the hns stpA double mutant differ for many genes, especially those of the core genome. Therefore, our data suggest the existence of two different regulons: one subset of genes is regulated by the H-NS protein alone, whereas a second group seems to be regulated by a concerted action of H-NS and StpA. It has been shown previously that H-NS can exist as homodimers or as a heteromer together with StpA (19, 40), and a distinct function for each complex has been proposed. By using DNA arrays, we were able to identify some of the target genes of the two types of DNA-binding complexes.

Finally, the observed RNA chaperone activity of StpA also has to be taken into account, which suggests a role for StpA in mRNA folding and posttranscriptional regulation (48). If some transcripts of other regulators (for instance, the small RNA DsrA, which was shown to counteract H-NS) are affected by StpA in their conformation and activity, the absence of StpA could have indirect effects on H-NS-regulated genes.

Such indirect effects might also result by other means, since the lack of H-NS was shown to affect several other regulators. Johansson and coworkers (18) previously reported that mutations in the hns and/or the stpA gene affect the levels not only of the factor for inversion stimulation, Fis, but also of the cyclic AMP receptor protein and the guanine nucleotide derivative (p)ppGpp. Moreover, there was a direct link between each of these global regulators, resulting in strong interrelationships. Using semiquantitative RT-PCR, we could at least observe lower crp transcript levels in strains lacking H-NS (data not shown). Array data confirmed a 1.3-fold-reduced expression level of crp in both hns mutants compared to the wild type.

Furthermore, we could identify more than 20 known or putative transcriptional regulators with altered expression in the hns mutants and about 50 proteins with other regulatory functions. Among them were many two-component systems such as EvgAS or the PhoPQ system or other pleiotropic regulators like the Hha protein mentioned above, all of which have been described to have a broad spectrum of target genes.

These results suggest that the effects seen by transcriptional profiling are not all ascribable to H-NS-mediated regulation alone but most likely also result from indirect effects from other regulators. Therefore, the exact role of H-NS in the virulence regulation cascade could not be determined in this study. Nevertheless, it clearly demonstrates a variety of possible interactions between the two DNA-binding proteins H-NS and StpA and the complexity of the regulatory network with H-NS as a major player.

Acknowledgments

The Würzburg group was supported by the German Research Foundation (SFB479, TP A1, and the international graduate college, IGK 587/2). L. Emődy and G. Nagy were supported by OTKA grants T037833 and F048526. B. E. Uhlin was supported by the Swedish Research Council and the Swedish Foundation for International Cooperation in Research and Higher Education (STINT).

REFERENCES

1. Reference deleted.
2. Atlung, T., and H. Ingmer. 1997. H-NS: a modulator of environmentally regulated gene expression. Mol. Microbiol. 24:7-17. [PubMed]
3. Balsalobre, C., J. Morschhäuser, J. Jass, J. Hacker, and B. E. Uhlin. 2003. Transcriptional analysis of the sfa determinant revealing multiple mRNA processing events in the biogenesis of S fimbriae in pathogenic Escherichia coli. J. Bacteriol. 185:620-629. [PMC free article] [PubMed]
4. Beloin, C., K. Michaelis, K. Lindner, P. Landini, J. Hacker, J. Ghigo, and U. Dobrindt. 2006. The transcriptional antiterminator RfaH represses biofilm formation in Escherichia coli. J. Bacteriol. 188:1316-1331. [PMC free article] [PubMed]
5. Berger, H., J. Hacker, A. Juarez, C. Hughes, and W. Goebel. 1982. Cloning of the chromosomal determinants encoding hemolysin production and mannose-resistant hemagglutination in Escherichia coli. J. Bacteriol. 152:1241-1247. [PMC free article] [PubMed]
6. Bertin, P., F. Hommais, E. Krin, O. Soutourina, C. Tendeng, S. Derzelle, and A. Danchin. 2001. H-NS and H-NS-like proteins in gram-negative bacteria and their multiple role in the regulation of bacterial metabolism. Biochimie 83:235-241. [PubMed]
7. Chuang, S. E., D. L. Daniels, and F. R. Blattner. 1993. Global regulation of gene expression in Escherichia coli. J. Bacteriol. 175:2026-2036. [PMC free article] [PubMed]
8. Datsenko, K., and B. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645. [PMC free article] [PubMed]
9. Dobrindt, U., F. Agerer, K. Michaelis, A. Janka, C. Buchrieser, M. Samuelson, C. Svanborg, G. Gottschalk, H. Karch, and J. Hacker. 2003. Analysis of genome plasticity in pathogenic and commensal Escherichia coli isolates by use of DNA arrays. J. Bacteriol. 185:1831-1840. [PMC free article] [PubMed]
10. Dorman, C. 2004. H-NS: a universal regulator for a dynamic genome. Nat. Rev. Microbiol. 2:391-400. [PubMed]
11. Elliott, S., V. Sperandio, J. Giron, S. Shin, J. Mellies, L. Wainwright, S. Hutcheson, T. McDaniel, and J. Kaper. 2000. The locus of enterocyte effacement (LEE)-encoded regulator controls expression of both LEE- and non-LEE-encoded virulence factors in enteropathogenic and enterohemorrhagic Escherichia coli. Infect. Immun. 68:6115-6126. [PMC free article] [PubMed]
12. Falconi, M., N. Higgins, R. Spurio, C. Pon, and C. Gualerzi. 1993. Expression of the gene encoding the major bacterial nucleotide protein H-NS is subject to transcriptional auto-repression. Mol. Microbiol. 10:273-282. [PubMed]
13. Falconi, M., V. McGovern, D. Hillyard, and N. Higgins. 1991. Mutations altering chromosomal protein H-NS induce mini-Mu transposition. New Biol. 3:615-625. [PubMed]
14. Gomez-Gomez, J., J. Blazquez, F. Baquero, and J. Martinez. 1996. hns mutant unveils the presence of a latent haemolytic activity in Escherichia coli K-12. Mol. Microbiol. 19:909-910. [PubMed]
15. Göransson, M., B. Sonden, P. Nilsson, B. Dagberg, K. Forsman, K. Emanuelsson, and B. E. Uhlin. 1990. Transcriptional silencing and thermoregulation of gene expression in Escherichia coli. Nature 344:682-685. [PubMed]
16. Hacker, J., G. Blum-Oehler, I. Mühldorfer, and H. Tschäpe. 1997. Pathogenicity islands of virulent bacteria: structure, function and impact on microbial evolution. Mol. Microbiol. 23:1089-1097. [PubMed]
17. Hommais, F., E. Krin, C. Laurent-Winter, O. Soutourina, A. Malpertuy, J. Le Caer, A. Danchin, and P. Bertin. 2001. Large-scale monitoring of pleiotropic regulation of gene expression by the prokaryotic nucleoid-associated protein, H-NS. Mol. Microbiol. 40:20-36. [PubMed]
18. Johansson, J., C. Balsalobre, S. Wang, J. Urbonaviciene, D. Jin, B. Sonden, and B. E. Uhlin. 2000. Nucleoid proteins stimulate stringently controlled bacterial promoters: a link between the cAMP-CRP and the (p)ppGpp regulons in Escherichia coli. Cell 102:475-485. [PubMed]
19. Johansson, J., S. Eriksson, B. Sonden, S. Wai, and B. E. Uhlin. 2001. Heteromeric interactions among nucleoid-associated bacterial proteins: localization of StpA-stabilizing regions in H-NS of Escherichia coli. J. Bacteriol. 183:2343-2347. [PMC free article] [PubMed]
20. Johansson, J., and B. E. Uhlin. 1999. Differential protease-mediated turnover of H-NS and StpA revealed by a mutation altering protein stability and stationary-phase survival of Escherichia coli. Proc. Natl. Acad. Sci. USA 96:10776-10781. [PMC free article] [PubMed]
21. Kerenyi, M., H. E. Allison, I. Batai, A. Sonnevend, L. Emody, N. Plaveczky, and T. Pal. 2005. Occurrence of hlyA and sheA genes in extraintestinal Escherichia coli strains. J. Clin. Microbiol. 43:2965-2968. [PMC free article] [PubMed]
22. Laemmli, U. K. 1970. Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227:680-685. [PubMed]
23. La Teana, A., A. Brandi, M. Falconi, R. Spurio, C. Pon, and C. Gualerzi. 1991. Identification of a cold shock transcriptional enhancer of the Escherichia coli gene encoding nucleoid protein H-NS. Proc. Natl. Acad. Sci. USA 88:10907-10911. [PMC free article] [PubMed]
24. Ludwig, A., C. von Rhein, S. Bauer, C. Hüttinger, and W. Goebel. 2004. Molecular analysis of cytolysin A (ClyA) in pathogenic Escherichia coli strains. J. Bacteriol. 186:5311-5320. [PMC free article] [PubMed]
25. Madrid, C., J. M. Nieto, S. Paytubi, M. Falconi, C. O. Gualerzi, and A. Juárez. 2002. Temperature- and H-NS-dependent regulation of a plasmid-encoded virulence operon expressing Escherichia coli hemolysin. J. Bacteriol. 184:5058-5066. [PMC free article] [PubMed]
26. Morschhäuser, J., B. E. Uhlin, and J. Hacker. 1993. Transcriptional analysis and regulation of the sfa determinant coding for S fimbriae of pathogenic Escherichia coli strains. Mol. Gen. Gent. 238:97-105. [PubMed]
27. Morschhäuser, J., V. Vetter, L. Emödy, and J. Hacker. 1994. Adhesin regulatory genes within large, unstable DNA regions of pathogenic Escherichia coli: cross-talk between different adhesin gene clusters. Mol. Microbiol. 11:555-566. [PubMed]
28. Mühldorfer, I., W. Ziebuhr, and J. Hacker. 2001. Escherichia coli in urinary tract infections, p. 1739-1748. In M. Sussmann (ed.), Molecular medical microbiology. Academic Press, New York, N.Y.
29. Nagy, G., U. Dobrindt, G. Schneider, A. Khan, J. Hacker, and L. Emödy. 2002. Loss of regulatory protein RfaH attenuates virulence of uropathogenic Escherichia coli. Infect. Immun. 70:4406-4413. [PMC free article] [PubMed]
30. Nagy, G., A. Altenhoefer, O. Knapp, E. Maier, U. Dobrindt, G. Blum-Oehler, R. Benz, L. Emödy, and J. Hacker. Both alpha-haemolysin determinants contribute to full virulence of uropathogenic Escherichia coli strain 536. Microbes Infect., in press. [PubMed]
31. Nieto, J., C. Madrid, A. Prenafeta, E. Miquelay, C. Balsalobre, M. Carrascal, and A. Juarez. 2000. Expression of the hemolysin operon in Escherichia coli is modulated by a nucleoid-protein complex that includes the proteins Hha and H-NS. Mol. Gen. Genet. 263:349-358. [PubMed]
32. Nishino, K., and A. Yamaguchi. 2004. Role of histone-Like protein H-NS in multidrug resistance of Escherichia coli. J. Bacteriol. 186:1423-1429. [PMC free article] [PubMed]
33. Olsen, P. B., and P. Klemm. 1994. Localization of promoters in the fim gene cluster and the effect of H-NS on the transcription of fimB and fimE. FEMS Microbiol. Lett. 116:95-100. [PubMed]
34. Oscarsson, J., Y. Mizunoe, L. Li, X. H. Lai, A. Wieslander, and B. E. Uhlin. 1999. Molecular analysis of the cytolytic protein ClyA (SheA) from Escherichia coli. Mol. Microbiol. 32:1226-1238. [PubMed]
35. Reed, J., and H. Muench. 1938. Simple method of determining fifty percent endpoints. Am. J. Hyg. 27:493-497.
36. Renzoni, D., D. Esposito, M. Pfuhl, J. Hinton, C. Higgins, P. Driscoll, and J. Ladbury. 2001. Structural characterization of the N-terminal oligomerization domain of the bacterial chromatin-structuring protein, H-NS. J. Mol. Biol. 306:1127-1137. [PubMed]
37. Schwyn, B., and J. B. Neilands. 1987. Universal chemical assay for the detection and determination of siderophores. Anal. Biochem. 160:47-56. [PubMed]
38. Shiga, Y., Y. Sekine, Y. Kano, and E. Ohtsubo. 2001. Involvement of H-NS in transpositional recombination mediated by IS1. J. Bacteriol. 183:2476-2484. [PMC free article] [PubMed]
39. Snyder, J., B. Haugen, E. Buckles, C. Lockatell, D. Johnson, M. Donnenberg, R. Welch, and H. Mobley. 2004. Transcriptome of uropathogenic Escherichia coli during urinary tract infection. Infect. Immun. 72:6373-6381. [PMC free article] [PubMed]
40. Sonden, B., and B. E. Uhlin. 1996. Coordinated and differential expression of histone-like proteins in Escherichia coli: regulation and function of the H-NS analog StpA. EMBO J. 15:4970-4980. [PMC free article] [PubMed]
41. Sonnenfield, J., C. Burns, C. Higgins, and J. Hinton. 2001. The nucleoid-associated protein StpA binds curved DNA, has a greater DNA-binding affinity than H-NS and is present in significant levels in hns mutants. Biochimie 83:243-249. [PubMed]
42. Soutourina, O., A. Kolb, E. Krin, C. Laurent-Winter, S. Rimsky, A. Danchin, and P. Bertin. 1999. Multiple control of flagellum biosynthesis in Escherichia coli: role of H-NS protein and the cyclic AMP-catabolite activator protein complex in transcription of the flhDC master operon. J. Bacteriol. 181:7500-7508. [PMC free article] [PubMed]
43. Spassky, A., S. Rimsky, H. Garreau, and H. Buc. 1984. H1a, an E. coli DNA-binding protein which accumulates in stationary phase, strongly compacts DNA in vitro. Nucleic Acids Res. 12:5321-5340. [PMC free article] [PubMed]
44. Spurio, R., M. Dürrenberger, M. Falconi, A. La Teana, C. Pon, and C. Gualerzi. 1992. Lethal overproduction of the Escherichia coli nucleoid protein H-NS: ultramicroscopic and molecular autopsy. Mol. Gen. Gent. 231:201-211. [PubMed]
45. Reference deleted.
46. Ueguchi, C., C. Seto, T. Suzuki, and T. Mizuno. 1997. Clarification of the dimerization domain and its functional significance for the Escherichia coli nucleoid protein H-NS. J. Mol. Biol. 274:145-151. [PubMed]
47. Uhlin, B. E., and Y. Mizunoe. 1994. Expression of a novel contact-hemolytic activity by E. coli. J. Cell. Biochem. 18A:71-73.
48. Zhang, A., S. Rimsky, M. Reaban, H. Buc, and M. Belfort. 1996. Escherichia coli protein analogs StpA and H-NS: regulatory loops, similar and disparate effects on nucleic acid dynamics. EMBO J. 15:1340-1349. [PMC free article] [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...