OprE is a channel-forming outer membrane protein of Pseudomonas aeruginosa, the expression of which is induced under anaerobic conditions. We constructed various mutants and observed the effects on oprE expression. Deficiency in RpoN, an alternative sigma factor for RNA polymerase, abolished oprE expression under aerobic conditions, but did not affect the expression under anaerobic conditions. One mutation on the putative RpoN recognition site also caused reduction of oprE expression. The region 500 nucleotides upstream of the mRNA start site was required for optimal oprE transcription, which contains an AT-rich region including a putative integration host factor binding site. These results indicate that OprE production is directly or indirectly controlled by RpoN but also require some other regulatory proteins bound to the upstream region.
OprE of Pseudomonas aeruginosa is an outer membrane protein, which forms a channel of quite small pore size . The physiological role of OprE is unclear because a deficiency in OprE did not affect the phenotype of strain PAO1, including the growth rate or susceptibility to various antibiotics . We have previously reported that OprE production is induced under anaerobic conditions, which is a unique characteristic of porins . A putative RpoN recognition site was found upstream of the oprE transcription start site, but a putative −10/−35 promoter sequence was apparently absent. RpoN is an alternative sigma factor for RNA polymerase, displaying no significant homology to the major sigma family such as the sigma70 protein [4, 5]. RpoN has been reported to have two common features in different bacteria. First, RpoN is not essential for cell growth, but is required for expression of a wide variety of genes involved in many diverse functions, including nitrogen fixation, utilization of alternative carbon sources, production of extracellular structures, and virulence determinants, which are needed in order to adapt to various environmental conditions. Second, RpoN-dependent expression requires positive regulators which usually exert their influence from a distance by binding a specific sequence upstream of the promoter. Since there has been no report demonstrating that RpoN controls the expression of the channel-forming major outer membrane proteins, we tried to clarify the contribution of RpoN to OprE production.
In this study, we found that RpoN and a far upstream region of the oprE gene are required for optimal oprE transcription under aerobic conditions.
Materials and methods
Media and culture conditions
P. aeruginosa PAO1 was grown in LB broth at 37°C supplemented with 0.2% KNO3, which is essential for growth under anaerobic conditions. The strains were grown in 50 ml of the broth in 500-ml flasks with vigorous shaking until the mid-exponential phase. Growth under anaerobic conditions was performed in an anaerobic chamber using a gas mixture of N2, CO2, and H2 (8:1:1) .
Escherichia coli strains were also grown in LB broth. Appropriate antibiotics were added to the medium when necessary: 50 μg ml−1 ampicillin for E. coli; 250 μg ml−1 carbenicillin for P. aeruginosa; 50 μg ml−1 chloramphenicol for E. coli; 12.5 μg ml−1 and 125 μg ml−1 tetracycline for E. coli and P. aeruginosa, respectively.
Isolation of RpoN-deficient mutants
RpoN-deficient mutants were constructed by introducing a mutated rpoN gene into the PAO1 chromosome by allelic gene replacement techniques . The 1.5-kb rpoN gene fragment ranging from its ATG initiation codon to its termination codon was amplified by polymerase chain reaction (PCR), and inserted into pNOT19, yielding pRPON5. Next, a 0.35-kb NaeI fragment was deleted from the rpoN structural gene and replaced by a 2.4-kb ΩTc fragment conferring tetracycline resistance , yielding pRPON51 (Fig. 1). A 5.8-kb NotI MOB cassette was then inserted into the NotI site of pRPON51 to yield pRPON511. Conjugal transfer was performed between the E. coli mobilizer strain S17-1 carrying pRPON511 and the recipient P. aeruginosa strain PAO1 as described previously . Next, tetracycline-resistant colonies were picked, and streaked onto LB agar medium containing 125 μg ml−1 of tetracycline and 5% sucrose. The resultant RpoN-deficient mutants were selected by confirming the replacement of the mutated rpoN gene in place of the normal rpoN gene by PCR and Southern hybridization. Further, RpoN deficiency was confirmed by glutamine requirement and lack of pilin and flagellin [9, 10]. One of the RpoN-deficient strains derived from strain PAO1 was designated strain YY151.
Construction of oprE promoter deletion mutants
Plasmid pOPRE2002 was constructed by introducing 6.0-kb BamHI-HindIII fragments of pOPRE200  into the broad-host-range vector pKT240 . The insert contains the oprE structural gene with 2.2 kb of the 5′ flanking region and 2.4 kb of the 3′ flanking region . Plasmid pOPRE4081 was constructed to carry the oprE structural gene with 0.5 kb of the 5′ flanking region and 140 bp of the 3′ flanking region on pKT240 (Fig. 2). A series of plasmids pOPRE4041–4071 were constructed to have deletions in the oprE upstream region of pOPRE4081 by utilizing the restriction enzyme recognition site shown in Fig. 2. The resultant plasmids were transferred into an OprE-deficient strain YY100  by utilizing E. coli mobilizer strain S17-1, and their effects on oprE expression were determined by Western and Northern blotting analyses.
Mutagenesis at the RpoN recognition site
A putative RpoN recognition site on plasmid pOPRE4081 was mutagenized using PCR with the mutagenized primers in conjunction with the neighboring restriction enzyme cleavage site. The GG nucleotide at −23 (the nucleotide residue 23 bp upstream of the oprE mRNA start site) or the CC nucleotide at −11 were replaced with TT nucleotides, yielding pOPRE2421 and -2431, respectively. Their effects on oprE expression were analyzed as described above.
Analysis of outer membrane proteins and Western blotting using anti-OprE polyclonal antiserum were performed as described previously .
Oligodeoxyribonucleotides were synthesized on Oligo1000 (Beckman), and extracted as described in the manufacturer's instructions, followed by precipitation by vacuum centrifugation.
DNA sequencing, analysis of the nucleotide sequence determination, PCR, extraction of total RNA, Northern blotting, and other standard recombinant DNA techniques were performed as described previously .
oprE expression in RpoN-deficient mutants
The effect of RpoN deficiency on oprE expression is shown in Fig. 3. Northern blotting analysis showed that oprE expression almost disappeared in the RpoN-deficient mutant YY151 under strict aerobic conditions, although its parental strain PAO1 produced a large amount of oprE transcript. Outer membrane protein profiles of the mutant YY151 and its parental strain PAO1 also demonstrated the greatly decreased production of OprE under aerobic conditions. On the other hand, RpoN deficiency did not affect oprE transcription under anaerobic conditions. These results indicate that oprE expression under aerobic conditions is controlled by RpoN at the transcriptional level. We therefore turned our attention to various mutations of the upstream region and observed their effects on oprE expression as described below.
Determination of the upstream region required for oprE expression
RpoN-dependent transcription usually requires additional regulatory proteins, which bind to the region upstream of the promoter. We therefore tried to establish whether the region upstream of the mRNA start site was required for oprE expression.
Deletion of the far 5′ upstream region of oprE significantly affected oprE expression. The results in Fig. 4 indicate that the region starting from −484 is required for optimal OprE production, because only strain YY100 harboring pOPRE4081 produced a similar amount of OprE to that harboring pOPRE2002. Northern blot analysis also showed similar results (data not shown). Strain YY100 harboring pOPRE4071 produced less OprE under aerobic conditions, suggesting that the region spanning −457 to −484 is involved in OprE production. OprE production under both aerobic and anaerobic conditions greatly decreased in strain YY100 harboring pOPRE4061, indicating the significance of the region spanning −347 to −457.
Effect of mutation in the RpoN recognition site on oprE expression
A putative RpoN recognition site (GGGCCC-N5-TTCCA) is found in the region 9–24 bp upstream of the oprE mRNA start site (Fig. 2). Of the 11 nucleotides, eight were identical to the consensus RpoN recognition site (TGGCAC-N5-TTGCt/a), although the almost universally conserved −12 GC motif was not found and the distance from the mRNA start site was rather short. In order to observe whether this site is needed for expression, we constructed plasmids pOPRE2421 and -2431 to have an alteration in GG/CC nucleotides at −23/−11, corresponding to the most conserved region of the consensus RpoN recognition site, respectively.
As shown in Fig. 5, the mutation of the GG nucleotide at −23, which is invariant among the known RpoN recognition sites, significantly decreased OprE expression under both aerobic and anaerobic conditions, while the mutation of the CC nucleotides at −11 had no significant effect.
This is the first report that RpoN is required for optimal expression of a major outer membrane protein which forms a channel such as OprE. We found that RpoN deficiency, deletion of the oprE far upstream region and one mutation of the putative RpoN recognition site significantly decreased oprE expression under aerobic conditions. However, another alteration in the putative RpoN recognition site (the CC nucleotide at −11) did not decrease oprE expression under aerobic conditions, although it has been reported that the −12 GC motif is almost universally conserved and mutations in this motif significantly decrease the expression levels . Another study reported that the distorted region around the −12 GC motif appeared not to be tightly in contact with the RpoN protein . Thus, it is uncertain whether or not RpoN binds to the promoter region of oprE. The possibility still remains that RpoN indirectly affects oprE expression by reducing the production of a regulatory protein.
The regulation mechanism of OprE production under anaerobic conditions remains unclear. Although RpoN deficiency did not reduce OprE production under anaerobic conditions, alterations in the putative RpoN recognition site or deletion of the upstream region greatly reduced OprE production. These results indicate that the region spanning −347 to −453 and the GG nucleotides at −23 play significant roles in OprE production under anaerobic conditions. On the other hand, ANR deficiency (data not shown) did not affect anaerobic induction of OprE although ANR is a positive regulator under anaerobic conditions [15, 16] and the latter half of the ANR box was found upstream of the oprE gene .
Three candidates for contribution to regulation of oprE expression were found in the region spanning −347 to −484, which seemed to be important for optimal oprE expression. First, the region from −453 to −479, the deletion of which caused partially decreased oprE expression only under aerobic conditions (see pOPRE4071 in Fig. 5), contains only three mismatches with a consensus integration host factor (IHF) binding site, which has been reported to contain poly-A and -T sequences located immediately upstream of the central WATCAANNNNTTR core . IHF, which has also been reported to act in P. aeruginosa[18–20], is known to serve as an accessory protein which influences the binding between DNA and RNA polymerase sigma factor including RpoN [4, 21]. Second, the region spanning −447 to −415 contains two consecutive TTTTAN2-3TTTATT sequences. Such an AT-rich region might cause inherent DNA bending, which has been considered to be involved in the expression of various genes . These results suggest that the DNA bending caused by IHF or AT-rich sequences might affect oprE expression, although it is unusual for bending sites to be this far from the mRNA start site. Third, two series of inverted repeats appeared at the regions from −408 to −395 and from −393 to −380. These sequences may also be involved in oprE expression, presumably affecting the secondary structure of DNA or functioning as a binding target of other regulator proteins.
The results obtained in this study suggest that RpoN and some DNA binding factors, perhaps IHF, affect optimal oprE expression, although it is still uncertain whether or not RpoN directly binds to the oprE upstream region. Further research utilizing a reporter gene should help clarify this systems and identify the proteins regulating oprE transcription.