INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The outer membrane (OM) of gram-negative bacteria consists of an inner monolayer of phospholipids linked to an outer leaflet predominantly composed of negatively charged lipopolysaccharide molecules which are noncovalently bridged by divalent cations. In addition, a certain number of porins form nonselective channels allowing free diffusion of small hydrophilic molecules (<700 Da), while more selective porins facilitate diffusion of particular nutrients (18). This overall structure enables the OM to exclude potentially harmful molecules and to serve as a selective permeability barrier.

S-type pyocins and type A colicins are bacterial toxins which exert their lethal activity upon intracellular targets (26, 28). These bacteriocins are bulky molecules (5, 26) exceeding the diffusion limit of the OM and consequently cannot freely diffuse to the cell periplasm. Escherichia coli type A colicins parasitize the high-affinity OM ferric siderophore receptors and the inner membrane proteins TolQRA in order to circumvent the OM permeability barrier and gain entry into bacterial cells (5, 28). The E. coli TolQRA proteins are also involved in maintaining the integrity of the E. coli cell envelope, and they have been proposed to form a structural bridge between the inner and outer membranes in concert with the tolB and pal gene products (3, 28).

The tolQRA homologs have been identified in Pseudomonas aeruginosa and shown to be involved in pyocin transport (8). The P. aeruginosa tolQ and tolR genes were found to be cotranscribed as a 1.5-kb transcript, while transcription of the 1.2-kb tolA mRNA was shown to occur from its own promoter (8). Interestingly, a P. aeruginosa DNA fragment encoding an open reading frame (ORF) (pig6) exhibiting a high degree of homology to E. coli orf1, which is located upstream of E. coli tolQ, was isolated as an iron-regulated gene directly bound by the P. aeruginosa ferric uptake regulator (Fur) (21). The ORF immediately downstream of pig6 had homology to E. coli tolQ. These studies suggest that tolQ expression as well as orf1 (pig6) might be regulated by Fur in P. aeruginosa.

There are no reports which suggest that in E. coli the tol genes may be iron regulated by Fur; the only environmental condition reported to influence expression of the E. coli tol genes has been temperature. Clavel et al. (6) demonstrated that expression of chromosomal tolR::lacZ and tolA::lacZ operon fusions was approximately threefold higher when cells were grown at temperatures lower than 37°C. The investigators were able to correlate tol expression with capsule synthesis. The objectives of the present study were to determine if expression of the tolQRA genes in P. aeruginosa is regulated by either iron or temperature.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains, culture conditions, and -galactosidase assays. Strains and plasmids described in this study are listed in Table 1. For -galactosidase assays, cultures were grown in either LB broth (Difco Laboratories, Detroit, Mich.) or TSBDC broth (22) supplemented in some instances with either 400 µg of ethylenediamine di(o-hydroxy)phenylacetic acid (EDDHA) per ml or 50 µM FeCl₃. The initial iron concentration of TSBDC has previously been determined to be 1 µM (22). Unless indicated otherwise, cultures were grown at 37°C at 200 rpm to mid-log phase of growth (optical density at 600 nm [OD₆₀₀] = 0.5). This OD represented mid-log phase of growth in all medium conditions used. -Galactosidase assays were performed as previously described and are expressed in Miller units (17). Values represent the mean ± standard deviation of triplicate experiments. All cultures were assayed during mid-log phase at the same OD. No growth effects on -galactosidase activity were observed during mid-log phase of growth (data not shown). No differences in -galactosidase activity of tolQR or tolA::lacZ fusions were observed between LB and TSBDC broth supplemented with 400 µg of EDDHA per ml (compare Fig. 1 and 3). For RNA isolation, cultures were grown in TSBDC medium as parallel Northern hybridization experiments were performed with P. aeruginosa toxA as a control iron-regulated transcript (data not shown) (16). All reagents and media were prepared with H₂O purified by the Milli Q system (Millipore Corp., Bedford, Mass.). Antibiotics were added to the growth medium at the following concentrations where appropriate: for E. coli, 15 µg of gentamicin or tetracycline per ml; for P. aeruginosa, 200 µg of gentamicin or 250 µg of tetracycline per ml.

DNA probes. DNA fragments used as probes in Northern hybridization experiments were purified by using Gene-Clean II (Bio/Can Scientific, Mississauga, Ontario, Canada) according to the manufacturer's recommendations. DNA fragments were labeled with [-³²P]dCTP (Amersham Life Sciences, Oakville, Ontario, Canada), using an oligolabeling kit (Pharmacia Biotech Inc., Baie d'Urfe, Quebec, Canada) as recommended by the manufacturer. Synthetic oligonucleotides used in primer extension experiments were end labeled with [-³²P]dATP (Amersham Life Sciences), using T4 phosphorylase nucleotide kinase (Life Technologies, Burlington, Ontario, Canada) as recommended by the manufacturer.

RNA isolation and Northern hybridization. Total RNA was isolated by using a hot-phenol extraction procedure (27) from 10-ml cultures at an OD₆₀₀ of 0.5 or 1.0. Aliquots of 15 µg of total RNA were denatured at 55°C for 15 min in the presence of 2 M formaldehyde and 50% formamide, and duplicate samples of RNA were electrophoresed on a 1.5% agarose gel containing 2 M formaldehyde in morpholinepropanesulfonic acid buffer (24). After electrophoresis, half of the agarose gel was stained with ethidium bromide and the RNA was visualized under UV light. The other half was transferred onto nitrocellulose membranes (Bio-Rad Laboratories Ltd., Mississauga, Ontario, Canada) and baked for 2 h at 80°C, and hybridization was performed overnight at 65°C in 15 ml of 1% sodium dodecyl sulfate (SDS)-1 M NaCl-10% dextran sulfate-0.1 mg of salmon sperm DNA per ml. The membranes were washed twice for 15 min each time at room temperature in 1× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% SDS and twice for 15 min at 65°C in 0.25× SSC-0.1% SDS. The membranes were dried and autoradiographed with Kodak X-Omat AR 100 film.

Nucleotide sequencing and primer extension experiments. Sequencing was performed by the dideoxynucleotide chain termination method (25) with T7 DNA polymerase (Sequenase 2.0 kit; United States Biochemical Corp., Oakville, Ontario, Canada), [³⁵S]dATP (Amersham Life Sciences), and a synthetic oligonucleotide primer (5'-GGCCAGAAATAACTCTCGG-3') as recommended by the manufacturer. Terminal deoxynucleotidyltransferase (Life Technologies) was added after termination of the sequencing-labeling reactions to resolve ambiguities in GC-rich templates (13).

DNA mobility shift assays. DNA mobility shift assays were performed as described by Ochsner et al. (20), with minor modifications. DNA fragments were labeled as above and mixed with 20 µl of binding buffer, composed of 20 mM Tris borate (pH 7.5), 40 mM KCl, 0.1 mM MnSO₄, 1 mM MgSO₄, 0.1 mg of bovine serum albumin per ml, 50 µg of poly(dI-dC) (Sigma Chemical Co., St. Louis, Mo.) per ml, and 10% glycerol. Fur protein was added to give final concentrations ranging from 0 to 1,000 nM, and the mixtures were incubated at 37°C for 30 min. A 5% nondenaturing polyacrylamide gel in running buffer (20 mM Tris borate, 0.1 mM MnSO₄) was prerun for 10 min at 200 V and loaded with 10 µl of the binding reaction mixture. The gel was electrophoresed at 90 V for 4 h at 4°C, dried, and autoradiographed. A DNA fragment containing a synthetic Fur box consensus sequence was used as a positive control (20).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effect of iron on expression of the tolQRA genes. To determine if expression of the tol genes is iron regulated, strain PAO harboring pRKAz (tolA::lacZ) or pRKQz (tolQR::lacZ) was grown in LB broth supplemented either with 400 µg of EDDHA per ml (low iron) or with various concentrations of FeCl₃ to mid-log phase of growth (A₆₀₀ = 0.5), and the -galactosidase activity was determined. -Galactosidase activity was a direct measure of tolQR and tolA expression since the lacZ gene is under the transcriptional control of the tolQR or tolA promoter (8). tolQR and tolA expression was highest from cultures in low-iron medium (Fig. 1A and B, lanes 1). A 21% decrease in expression of tolA (Fig. 1A, lane 2) and an 18% decrease in expression of tolQR (Fig. 1B, lane 2) were observed when cells were grown in LB broth without EDDHA. Addition of 10 µM FeCl₃ resulted in a 51% reduction in expression of tolA (Fig. 1A, lane 3) and a 47% reduction in expression of tolQR (Fig. 1B, lane 3), while addition of 25 µM FeCl₃ decreased expression of tolA by 63% (Fig. 1A, lane 4) and tolQR by 60% (Fig. 1B, lane 4). Growth in the presence of higher concentrations of iron did not further reduce tolQR or tolA expression (Fig. 1A and B, lanes 5).

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Effects of iron on expression of plasmid-encoded tolA::lacZ (A) and tolQR::lacZ (B) operon fusions in PAO. -Galactosidase activity is expressed in Miller units. The plasmid control (pPA3.5RK) is represented in lanes 6 and 7 in both panels. Values represent the mean (± standard deviation) of triplicate experiments. Strains were grown in LB broth supplemented with 400 µg of EDDHA per ml (lanes 1 and 6), LB broth (lanes 2), and LB broth supplemented with 10 µM (lanes 3), 25 µM (lanes 4), or 100 µM (lanes 5 and 7) FeCl3. A significant difference in -galactosidase activity was observed between lanes 1, 2, 3, and 4 (P  0.05, Student's t test with Bonferroni correction). Addition of 100 µM FeCl3 did not result in a significant decrease in -galactosidase activity from that noted with 25 µM added FeCl3.

Effect of Fur on the expression of tolQRA. Total RNA isolated from PAO(pPA3.5RK) was used in primer extension experiments with RT and an oligonucleotide primer (5'-GGCCAGAAATAACTCTCGG-3') complementary to the tolA coding strand and located 25 bp downstream of the tolA predicted translational start codon (8). Figure 2A shows the primer extended with RT (lane 1). The size of the product suggests that transcription begins at a thymine residue located 65 bases upstream of the predicted translational start codon of tolA (Fig. 2). RT has a tendency to stop or pause in regions of RNA exhibiting a high degree of secondary structures (24), and such prematurely aborted extension products can be observed (Fig. 2A, lane 1). A potential Fur box (9-of-19-bp homology to the consensus sequence) was observed 20 bases upstream of the proposed transcriptional start site of tolA (Fig. 2B). Primer extension experiments designed to determine the transcriptional start site of tolQ yielded ambiguous results, resulting in pausing/stopping of RT along the mRNA strand (data not shown), and were not investigated further.

View larger version (94K): [in this window] [in a new window]   FIG. 2.   Primer extension analysis of tolA mRNA. (A) Radioactively labeled oligonucleotide primer (5'-GGCCAGAAATAACTCTCGG-3') extended with RT by using total RNA isolated from PAO(pPA3.5RK) (lane 1) and nucleotide sequence analysis of tolA performed with Sequenase 2.0, a synthetic oligonucleotide primer (5'-GGCCAGAAATAACTCTCGG-3'), and plasmid pEL3.5 (lanes A, C, G, and T). Arrowhead, proposed tolA transcriptional start site; lines, prematurely aborted extension products. The autoradiogram was turned over to represent the tolA coding strand and scanned with a Hewlett Packard Scan Jet 4c and Hewlett Packard Deskscan II software (Hewlett Packard Co.). (B) DNA sequence of the region upstream of the tolA predicted translational start codon (boldface and italic). Arrowhead, proposed tolA transcriptional start site; underlined bases, putative 35 and 10 promoter regions; shaded bases, putative Fur box.

View larger version (27K): [in this window] [in a new window]   FIG. 3.   Effects of iron on expression of plasmid-encoded tolA::lacZ (A) and tolQR::lacZ (B) operon fusions in P. aeruginosa fur mutants. -Galactosidase activity is expressed in Miller units. The plasmid control (pPA3.5RK) is represented in both panels. Values represent the mean (± standard deviation) of triplicate experiments. Cultures were grown in LB broth supplemented with 400 µg of EDDHA per ml (closed bars) or 50 µM FeCl3 (hatched bars). A significant decrease in -galactosidase activity was observed in all strains with the addition of FeCl3 (P  0.05, Student's t test with Bonferroni correction).

Effect of temperature on expression of tolQRA. tolQRA expression in E. coli has been previously shown to be temperature dependent (6). To determine if expression of the P. aeruginosa homologs is also regulated by temperature, we determined the -galactosidase activity of PAO cultures harboring pRKAz or pRKQz grown at different temperatures during the same stages of growth. Expression of the tol genes was optimal at 37°C, and increasing the temperature of growth to 42°C resulted in a 40% reduction in tol expression during mid-log phase of growth (Fig. 4). Expression of the tol genes was also approximately 30 to 40% lower when cells were grown at 30°C compared to 37°C. Growth at 25°C reduced expression of tolA by 57% (Fig. 4A) and expression of tolQR by 49% (Fig. 4B) compared to expression at 37°C. Unlike the E. coli tol genes, which are optimally expressed at 25°C (6), tol expression in P. aeruginosa is optimal at 37°C. Varying the growth temperature had no effect on expression of the P. aeruginosa tol::lacZ fusions in E. coli (data not shown).

View larger version (16K): [in this window] [in a new window]   FIG. 4.   Effects of temperature on expression of plasmid-encoded tolA::lacZ (A) and tolQR::lacZ (B) operon fusions in PAO. -Galactosidase activity is expressed in Miller units. The plasmid control (pPA3.5RK) is represented in both panels. Values represent the mean (± standard deviation) of triplicate experiments. Strains were grown in LB broth at 25°C (25), 30°C (30), 37°C (37 and control), or 42°C (42). A significant decrease in -galactosidase activity was observed when strains were grown at 25, 30, and 42°C (P  0.05, Student's t test with Bonferroni correction).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Plasmid-encoded tolQR::lacZ or tolA::lacZ operon fusions were used to demonstrate that expression of P. aeruginosa tolQR and tolA is regulated by iron and temperature. PAO(pRKAz) and PAO(pRKQz) exhibited no differences in pyocin sensitivity compared to PAO(pRK415), indicating that the plasmid-encoded copies of the tol genes did not have a marked effect on membrane integrity (data not shown). Omission of the iron-chelating agent EDDHA from the growth media is sufficient to reduce the -galactosidase activity of PAO(pRKAz) and PAO(pRKQz) by 20%, and the addition of increasing concentrations of iron further decreases the enzymatic activity. The fur gene product has previously been shown to act as a global modulator of iron-regulated genes in P. aeruginosa (21). It was therefore not surprising to discover that iron regulation of tol expression occurs at the level of transcription and was altered in fur mutants of P. aeruginosa. A potential Fur box (9-of-19-bp homology to the consensus sequence) was identified 20 bases upstream of the proposed transcriptional start site of tolA (Fig. 2B). Promoter regions of iron-regulated genes generally show homology to the Fur box consensus sequence with a 13- to 17-bp match (14); however, sequences exhibiting homology of only 10 to 11 bp to the Fur-binding consensus sequence have been identified as being directly bound by the P. aeruginosa Fur (21). Purified Fur, however, was unable to bind to either tolQR or tolA promoter region, suggesting that perhaps 10 bp is the minimum homology necessary.

Since Fur does not bind directly, it appears that Fur mediates iron regulation of the tol genes through an intermediate transcriptional regulator. Expression of the intermediate activator would be directly repressed at the level of transcription by Fur-Fe²⁺. Under conditions of iron starvation, repression by Fur would be relieved and the activator would be expressed, which in turn would activate transcription of the tol genes. Such a hierarchical organization of iron regulation by Fur has been reported in P. aeruginosa for expression of the ferripyochelin receptor, FptA, and pyochelin through the activator PchR (10, 11), and for production of pyoverdine (7) and exotoxin A (19) through the alternate sigma factor PvdS. Fur has previously been shown to bind to an ORF (orf1) of unknown function immediately upstream of tolQ (21). It is possible that this gene is involved in the intermediate regulation of tolQRA, although it exhibits no homology to known regulatory genes. E. coli Fur does not mediate iron repression of P. aeruginosa tol::lacZ expression, and -galactosidase activity is actually increased when cultures are grown in iron-supplemented medium (data not shown). These data are consistent with the hypothesis that an intermediate transcriptional regulator might be involved, as this gene would presumably be absent in E. coli.

The iron regulation of tolA appears to be similar to that of P. aeruginosa tonB (23), a similar protein which is required for siderophore-mediated iron transport. The exbB and exbD homologs, which are functionally related to tolQ and tolR, respectively, in E. coli (4), have not yet been identified in P. aeruginosa, and so their iron regulation has not been examined. Exotoxin A and siderophore production in P. aeruginosa are more tightly regulated by iron than either the tol or tonB gene (1, 2, 23). Yields of exotoxin A and siderophores are generally decreased by approximately 90% in cultures grown in high-iron medium. Because of their possible role in maintaining membrane structure, it would likely not be possible to decrease expression of tolQRA to the same extent as extracellular factors.

The PAO A2 and A4 mutants have a mutation in the same amino acid residue resulting in a change of His-86 to Arg and Tyr, respectively (1). The mutation in A4 has previously been shown to have a greater effect on the binding ability of Fur. For example, pyoverdine yields are decreased by 66% in PAO A2 grown at 37°C in 50 µM iron compared to growth in 0 µM iron, whereas yields of pyoverdine are decreased by only 25% in PAO A4 in the same conditions (1). Purified Fur from PAO A4 was shown to have reduced binding activity to the Fur box of the pvdS promoter compared to purified Fur from PAO A2 (1). Our data are consistent with those of Barton et al. (1) in that iron regulation of tolQRA expression was less in PAO A4 than in PAO A2.

Clavel et al. (6) demonstrated that expression of chromosomal E. coli tolR::lacZ and tolA::lacZ operon fusions was approximately threefold higher when cells were grown at temperatures lower than 37°C. These investigators determined that this increase in tol expression was mediated by the RcsC transcriptional regulator of E. coli capsular polysaccharide colanic acid. A decrease in tol expression correlated with an increase in capsule synthesis, which might serve to protect the cell from increased sensitivity to deleterious agents as a result of the tol mutations.

In P. aeruginosa the optimal temperature of expression for tolQRA was determined to be 37°C. There was no difference in expression of the tol::lacZ fusions in E. coli, which suggests that there may be a specific regulatory factor present in P. aeruginosa involved in the temperature regulation of tolQRA. Pyoverdin and exotoxin A are produced in greater yields at 25 to 27°C (1) and 32°C (15), respectively. Therefore, there does not appear to be a consistent optimal temperature for expression of iron-regulated genes in P. aeruginosa. Factors mediating temperature-dependent expression of genes have not yet been identified in this organism.

In conclusion, expression of P. aeruginosa tolQR and tolA is iron regulated at the level of transcription, and this iron regulation may involve Fur in addition to an intermediate regulator. Identification of regulators involved in modulating tol expression in response to environmental cues may yield important information on the role(s) of the tolQRA gene products in P. aeruginosa.