Dual promoters of the major catalase (KatA) govern distinct survival strategies of Pseudomonas aeruginosa

KatA is the major catalase required for hydrogen peroxide (H2O2) resistance and acute virulence in Pseudomonas aeruginosa PA14, whose transcription is driven from the promoter (katAp1) located at 155 nucleotide (nt) upstream of the start codon. Here, we identified another promoter (katAp2), the +1 of which was mapped at the 51 nt upstream of the start codon, which was responsible for the basal transcription during the planktonic culture and down-regulated upon H2O2 treatment under the control by the master regulator of anaerobiosis, Anr. To dissect the roles of the dual promoters in conditions involving KatA, we created the promoter mutants for each -10 box (p1m, p2m, and p1p2m) and found that katAp1 is required for the function of KatA in the logarithmic growth phase during the planktonic culture as well as in acute virulence, whereas katAp2 is required for the function of KatA in the stationary phase as well as in the prolonged biofilm culture. This dismantling of the dual promoters of katA sheds light on the roles of KatA in stress resistance in both proliferative and growth-restrictive conditions and thus provides an insight into the regulatory impacts of the major catalase on the survival strategies of P. aeruginosa.

Scientific RepoRts | 6:31185 | DOI: 10.1038/srep31185 account for the involvement of Anr and the elevated expression of KatA upon anaerobic respiration for its roles to ensure the distinct infection strategies of P. aeruginosa.
In the present study, we have explored the regulatory module underlying the anaerobic regulation of KatA, by newly identifying a second promoter (katAp2) of the katA gene, which is responsible for the basal expression of KatA during the planktonic growth of P. aeruginosa strain PA14. This second promoter is regulated primarily by Anr. More importantly, the mutation analyses of each promoter revealed the separate roles of the dual promoters to cope with reactive oxygen and/or nitrogen species (ROS and RNS), which is dependent on the growth states.

Results
Identification of a second promoter (katAp2) of the katA gene. A transcription start site of the katA gene was identified at the 155 nucleotide (nt) upstream (T-155) of the katA initiation codon, whose transcription is induced upon H 2 O 2 treatment. Despite over 5-fold induction of katAp1 transcription upon H 2 O 2 exposure, the katA-lacZ fusion displayed a minor (~2-fold) induction 8 , which may be attributed to another level of regulation. Based on this speculation, we have identified another promoter of the katA gene, katAp2: from the high-resolution S1 nuclease mapping data, a second protective band was observed at the 51 nt upstream (T-51) of the translation initiation codon (Fig. 1A). The upstream of the position of T-51 could be a potential promoter by its context. It is of marked interest that the putative Anr box consensus (TTGac-N 4 -gtCAA) was identified, which overlaps with the potential − 35 box of katAp2 (Fig. 1B), since KatA expression is deemed reduced in the anr mutant as previously assessed by the transcriptome and proteome analyses 12 . Identification of the katAp2 promoter and the potential Anr box suggests that the anr-dependent regulation of KatA might be most likely attributed to this second promoter. Moreover, the katAp2 transcript was slightly reduced upon H 2 O 2 treatment ( Fig. 1A; see below), which is in a good agreement with the fact that the Fe-S center of Anr is vulnerable to ROS and RNS with its transactivation activity compromised in the presence of O 2 − , H 2 O 2 , and NO 13-15 . Regulation of katAp2 by Anr during aerobic NO 3 − respiration. To verify the involvement of Anr in the regulation of katAp2, we examined the katA transcription in the anr null mutant, by using low-resolution S1 nuclease protection assay ( Fig. 2A). The katAp1 transcription was highly elevated, whereas the katAp2 transcription was slightly reduced upon H 2 O 2 treatment in the wild type bacteria. In contrast, however, the katAp2 transcription in the anr mutant was significantly reduced even in the absence of H 2 O 2 treatment and the katAp1 High-resolution S1 nuclease mapping. The 5′ end of the katA mRNA transcript was determined by high-resolution S1 nuclease mapping. Total RNA (50 μ g) that had been prepared from the cells with (+ ) or without (− ) 1 mM H 2 O 2 treatment for 10 min at OD 600 of 1.0 in LB were subjected to S1 nuclease mapping. The sequencing reactions (lanes C, T, A and G) were carried out as described in Methods. The two transcriptional start sites of the katA gene are indicated by open (katAp1) and closed (katAp2) arrows. The detailed transcriptional start site of katAp2 promoter is shown on the left by solid line, and its putative − 10 box is enclosed in box. (B) Promoter elements of the katA gene. The sequence of the katA promoter region of the katA gene is shown, with the two 5′ ends of katAp1 and katAp2 promoters indicated by the bent arrows. The Anr binding consensus (TTGac-N4-gtCAA) is designated, whose center is located at − 29 position from the katAp2 promoter. The putative − 10 and − 35 boxes are underlined. The N-terminal amino acid sequences of the katA gene are designated.
transcription was not affected in the anr mutant. Furthermore, the H 2 O 2 -mediated decrease in katAp2-driven transcription is not associated with OxyR, in that the katAp2 transcription pattern in the oxyR null mutant did not significantly differ from that in the wild type bacteria. This result suggests that the dual promoters of katA were regulated separately by two global regulators, OxyR and Anr.
Since the katAp2 transcription is abolished in the anr mutant, accounting for the basal expression of KatA during the planktonic growth, we have tested if the Anr-dependent NO 3 − respiration during the planktonic growth may be associated with the katAp2 transcription as well. We have exploited the mutants for the genes encoding another downstream regulator, Dnr and the four nitrogen oxide reductases (Nar, Nir, Nor, and Nos) involved in dissimilatory denitrification 16 : NO 3 − is reduced to nitrite (NO 2 − ) by Nar; NO 2 − is reduced to NO by Nir; NO is reduced to N 2 O by Nor; N 2 O is reduced to N 2 by Nos. Denitrification is vital for growth and survival under microaerobic and anaerobic conditions as found in biofilms and microcolonies.
We found that only the nirS mutant displayed reduced katAp2 transcription (Figs 2B and S1). This is in good agreement with the previous observation by Su et al. 11 that NO and Anr might be required for the increased KatA activity under anaerobic conditions. Creation of the katA promoter mutants. To explore the roles of the dual promoters in the expression of KatA that is required for several phenotypes such as stress resistance and virulence in P. aeruginosa 6 , we created the promoter mutants for each presumable − 10 box, as described in Fig. 3A. The substitution mutations were introduced into the chromosome by pEX18T-based allelic exchange, which was confirmed by PCR followed by restriction enzyme digestions. The functional verification of the created promoter mutants (p1m, p2m and p1p2m) was performed based on the katA transcription and catalase activity profiles in the mutants (Fig. 3B,C). The katAp1 and katAp2 transcripts from the promoter mutants were analyzed by S1 nuclease protection assay, which verified that the introduced substitutions at the proposed − 10 box resulted in the specific loss of the corresponding transcripts. A slower migrating band that was newly observed by the p2m mutation did not affect the catalase profiles in the mutants (Fig. 3C). As a result, the basal expression of KatA during the planktonic growth was abolished by the p2m mutation throughout the entire growth phases (Fig. 3D), whereas the p1m mutation contributed only to the induced expression upon H 2 O 2 treatment, which was more evident in the p2m mutant (Figs. 3C and S2). These results indicate that the katAp2 promoter is involved in the basal as well as increased expression during the planktonic growth, whereas katAp1 is responsible for the H 2 O 2 -inducible expression of KatA. Because the katAp2 requires the transactivator, Anr for its transcription, we have postulated that Anr is somehow active even in the aerobic planktonic culture in complex media such as PIA that contains ~63 μ M NO 3 − used for NO 3 − respiration 17 (Fig. 2). However, as shown in Fig. S3, the stationary phase-induced transcription of katAp2 was not completely abolished in the double mutant for anr and rpoS, indicating that it involves more complicated regulatory networks such as quorum sensing and so on 18 . katAp1 is critical in the well-growing state, whereas katAp2 is critical in the growth-restricting state.
We first investigated the several stress-related phenotypes of the promoter mutants in comparison with the katA null mutant, which is hypersusceptible to H 2 O 2 or acidified nitrite (aNO 2 ) 6,11,19 . As shown in Fig. 4A, when the cells were tested during the logarithmic growth phase, the p1m mutant and the p1p2m mutant were hypersusceptible to H 2 O 2 or aNO 2 treatment, whereas the p2m mutant was no more susceptible than the wild type. In contrast, however, when cells were taken during the stationary growth phase or from the biofilm culture, the outcome did clearly differ: the p2m mutant as well as the p1p2m mutant was hypersusceptible to H 2 O 2 or aNO 2 treatment, whereas the p1m mutant was no more susceptible than the wild type ( Fig. 4B,C). These results may be associated with the inducible expression of KatA by katAp1 when the basal expression of KatA by katAp2 might be The transcription patterns were assessed by low-resolution S1 nuclease protection assay with H 2 O 2 treatments in oxyR and anr mutants grown in LB (A) and in nirS and norB mutants grown in LB amended with 15 mM KNO 3 (B). Total RNA (50 μ g) that had been prepared from the wild type (WT) and the mutant (oxyR, anr, nirS, and norB) cells with (+ ) or without (− ) 1 mM H 2 O 2 treatment for 10 min at OD 600 of 0.5 were subjected to S1 nuclease assay. The two transcriptional start sites of the katA gene are indicated by open (katAp1) and closed (katAp2) arrows.
insufficient, and also with the higher basal expression of KatA by katAp2 upon the stationary growth phase when the stress-induced expression of KatA through katAp1 is likely compromised in this general stress condition 20 .
katAp1, but not katAp2, contributes to acute virulence. We next measured the virulence potential of the promoter mutants using Drosophila and murine acute infection models, since the katA null mutant of PA14 strain was virulence-attenuated in the same infection models 6 . It is of marked interest that p1m and p1p2m, but not p2m displayed virulence attenuation in both infection models (Fig. 5), suggesting that the inducible KatA expression driven by katAp1, is critical in the host environments. Although we do not have any direct evidence on whether the katAp2 transcription would occur during the infection conditions, it is clear that the katAp2-mediated KatA expression is likely dispensable in the virulence pathways at least under our experimental conditions. katAp2 contributes to lowering endogenously generated RNS. We hypothesized that KatA might have some surveillance function for homeostatic maintenance of cellular states regarding ROS and RNS by lowering the levels of endogenously generated ROS and RNS. We speculated that ROS and/or RNS could be accumulated in the p2m mutant, because katAp2 is required for the basal expression of KatA throughout the entire growth phases (Fig. 3D). We were able to measure the steady-state level of NO 2 − accumulated during the planktonic growth in LB broth supplemented with 15 mM NO 3 − , which can be used as an alternative electron acceptor through dissimilatory NO 3 − reduction involving three consecutive intermediates (NO 2 − , NO, and N 2 O) to the end product (N 2 ) 16 . Among the intermediates in the denitrification process, NO is highly unstable and vulnerable to spontaneous oxidation to NO 2 − 21 . Therefore, the amount of NO 2 − (either periplasmic or extracellular) may reflect the degree of NO 3 − respiration and the concomitant level of periplasmic RNS including NO 22 . Figure 6 shows that the katA null mutant as well as the katAp2 mutants (p2m and p1p2m) exhibited significantly higher levels (56.6~64.3 μ M) of NO 2 − than the wild type and the p1m mutant (5.7~6.0 μ M) during the planktonic growth. This result indicates the role of KatA as well as its katAp2-driven expression to lower the endogenous generation of RNS, under assumption that the NO 3 − respiration activity of the katA mutant does not significantly differ from that of the wild type bacteria, which requires more extensive investigation.

Discussion
KatA is the major catalase of P. aeruginosa, whose "adaptive" function is to dismutate H 2 O 2 , an ROS either sporadically generated during the normal respiration on O 2 or externally provided by redox-cycling agents or host factors. However, KatA exhibited higher activity under anaerobic conditions than aerobic conditions 23 , where no ROS could be potentially generated. This "enigmatic" expression profile of KatA under anaerobic conditions has been granted with the evidence that the KatA expression is lower in the anr mutant than in the wild type, with the potential Anr box located upstream of the katA coding region 12 . However, it had not been fully substantiated until the study about KatA function in anaerobiosis was reported 11 . They found that KatA contributes to the resistance to aNO 2 that generates RNS like NO. They also proposed that KatA binds to NO to sequester and prevent it from its harmful effect, a new role of KatA in buffering free NO to ensure P. aeruginosa anaerobiosis, based on the investigation of direct NO-KatA interaction under anaerobic condition, as first demonstrated in bovine liver catalase able to complex with NO at the heme 24 . Although the anaerobic KatA expression could be justified by its proposed function, the exact regulatory mechanisms have remained elusive. In this study, we first connect the dots between the previous results regarding the expression and the function of KatA under anaerobic conditions, by identifying the second promoter, katAp2. This promoter is positively regulated by Anr and required to lower the endogenous generation of RNS during planktonic culture, which we now call the "surveillance" function of KatA.
In our previous study, P. aeruginosa KatA exhibits unusual properties ascribed to its metastability, high specific activity, and/or extracellular presence, unlike the other clade 3 monofunctional bacterial catalases such as Streptomyces coelicolor CatA and Bacillus subtilis KatA (BsKatA) 7 . At least one of those properties has been proposed to be implicated in biofilm life style of P. aeruginosa. P. aeruginosa respires on NO 2 − and/or NO 3 − under anaerobic conditions, which is in marked contrast to the two bacterial species: S. coelicolor is unable to grow under anaerobic conditions; B. subtilis is capable of fermentation under anaerobic conditions. Interestingly, BsKatA is subjected to activity modulation called "instant adaptation" by NO in response to ROS 11,25 . Although the careful comparison between KatA and the related catalases needs to be furthered in regards to the physiological traits of the cognate bacteria, the functions of KatA in the microaerobic and anaerobic life style of P. aeruginosa may be attributed to the unusual biochemical properties of KatA, most likely the extracellular presence, as previously suggested 7 . This aspect will open a new venue for an insight into molecular evolution of a clade 3 bacterial catalase to fulfill the new needs that the bacterium is faced with, by delving into the roles of the particular amino acid residues of KatA associated with the extracellular and/or periplasmic presence and the protective function to buffer the NO, because NO is normally generated in the periplasm by Nir during during microaerobic and/or biofilm growth of P. aeruginosa 16 .
Another important contribution of this work is to provide a new picture to associate the protective functions of KatA with its regulation governed by OxyR and Anr. It should be noted that OxyR and Anr work apparently under the opposite conditions: for example, OxyR is activated by H 2 O 2 or NO 26 , while Anr is inactivated by H 2 O 2 or NO, because its Fe-S cluster is vulnerable to H 2 O 2 and NO [13][14][15] . Nevertheless, the dual promoters ensure the katA transcription in both balanced (i.e. surveillance function-requiring) and perturbed (i.e. adaptive function-requiring) conditions of ROS and/or RNS homeostasis. It is also likely that OxyR and Anr may cooperate for proper KatA expression during infection, due to the complicated respiration mode for P. aeruginosa growth that would be microaerophilic along the continuum between aerobic and anaerobic conditions in the human airways 27,28 .
Finally, the present study clearly demonstrates the discrete roles of the dual promoters in the KatA-associated phenotypes. The creation of the chromosomal mutations for the presumable − 10 boxes worked well to disrupt each promoter activity and the subsequent KatA expression specifically from the corresponding promoters. Use of these promoter mutants helped clarify the roles of each promoter in the functions of KatA in P. aeruginosa physiology. Based on these properties regarding regulation and function, the pivotal roles of KatA during the growth and survival of P. aeruginosa warrant further verification. This can be done by using the promoter mutants created in the present study. Moreover, the complex involvement of multiple regulatory systems will be comprehensively  Steady-state level of extracellular NO 2 − was measured in the wild type (WT) and the katA null and promoter mutant (katA, p1m, p2m, and p1p2m) cells as well as anr mutant cells that had been grown to OD 600 of 0.7. The amount (μ M) of NO 2 − is calculated using the standard curve (r 2 = 0.999) and the average values measured from the three independent experiments, with the error bars representing the one positive value of the standard deviations. Statistical significance based on the Student's t-test is indicated (* p < 0.01; * * p < 0.005).
Scientific RepoRts | 6:31185 | DOI: 10.1038/srep31185 explored at various regulatory levels, especially in response to stationary phase, quorum-sensing, and/or iron availability etc. All these aspects that will be more directly elucidated may provide a new insight into the therapeutic targets of this and the related bacteria associated with the stress responses toward ROS and RNS.

Methods
Bacterial strains and culture conditions. The bacterial strains and plasmids used in this study are listed in Table 1. E. coli and P. aeruginosa strains were grown at 37 °C using Luria-Bertani (LB) broth or on 1.5% Bacto-agar solidified LB plates. For anaerobic growth, bacteria were grown in LB medium supplemented with 15 mM KNO 3 in an anaerobic jar with AnaeroPack (MGC). Overnight-grown cultures were used as inoculum (1.6 × 10 7 cfu/ml) into fresh LB broth and grown at 37 °C to the OD 600 as indicated and then used for the experiments described herein.
DNA oligonucleotide primers. The information on DNA oligonucleotide primers used for gene deletion, expression, and detection in this study are listed in Table S1.
RNA isolation and S1 nuclease protection analysis. P. aeruginosa strains were grown in LB or LB containing 15 mM KNO 3 media and then the half of the culture was treated with 1 mM H 2 O 2 for 10 min, with the remaining half used as the untreated control. Total RNA was isolated from 10 9 cells by using RNeasy kit (Qiagen) and 50 μ g of RNA samples were used for S1 nuclease protection experiment as described elsewhere 8 . Briefly, the PCR-generated probe using the oligonucleotide primer pairs (katA-N10: 5′ end at − 133 and katA-S1C1: 5′ end at + 264) was labeled with [γ -32 P] ATP by T4 polynucleotide kinase. For hybridization, the mixtures of RNA samples and labeled probe were incubated at 90 °C for 10 min for denaturation, and then slowly cooled down to 55 °C for hybridization. After overnight hybridization, S1 nuclease digestion was performed at 30 °C for 30 min by adding 8 units of S1 nuclease for each sample. The reaction was stopped and precipitated with 100% ethanol. The samples were dissolved and denatured at 90 °C for 5 min in formamide-dye solution, and then analyzed by 6% PAGE containing 7 M urea. For high resolution S1 mapping, the unlabeled katA-S1C1 was used to generate the nucleotide sequence ladder using Sequenase Version 2.0 DNA Sequencing Kit (USB) with [α -32 P] dATP and pUCP-katA as the template 7 .
Creation of the katA promoter mutants. The katA promoter mutant allele for the katAp1 (− 10 box; CATCCT to GGTACC, the KpnI site) or for the katAp2 (− 10 box; CACGCT to GGATCC, the BamHI site) were generated by SOEing (splicing by overlap extension) PCR using 4 oligonucleotide primers. The resulting PCR products were cloned into pEX18T. These katA promoter mutant alleles were introduced into PA14 chromosome as described elsewhere 29 .
Creation of the lacZ fusions and β-galactosidase assay. All the lacZ fusions of the katA promoter regions were created by SOEing PCR using 4 oligonucleotide primers (Table S1). The primers were designed to generate transcriptional fusions of the katA promoter regions without its own ribosome-binding site (RBS) at the upstream of the lacZ RBS. Briefly, the katA promoter regions were prepared by using the primers pairs (katA-N3 and katA-lacZ-UC) and each of the chromosomes from the WT and the promoter mutant (p1m, p2m and p1p2m) bacteria as the templates. The lacZ coding region was generated by using the primer pairs (katA-lacZ-DN and pQF50-lacZ-C1) and the pQF50 plasmid as the template. The PCR products were fused by SOEing PCR using katA-N3 and pQF50-lacZ-C1 primers and the amplified fragments were cloned into pQF50. These promoter fusion constructs were introduced by electroporation and LacZ (β -galactosidase) activity was determined using the bacterial culture aliquots taken at the indicated time points as previously described 30 . Catalase activity staining. Catalase activity staining was performed as described previously 7 . Briefly, cell extracts (40 μ g) were applied to a 7% native polyacrylamide gel and electrophoresed. The gel was washed in distilled water and then treated with 1 mM H 2 O 2 for 10 min. After treatment, the gel was rinsed and transferred to 1% (w/v) ferric chloride and 1% (w/v) potassium ferricyanide solution. The reaction was stopped by washing in distilled water.
Stress susceptibility measurement. Stress susceptibility was measured based on the survival of the P. aeruginosa cells upon 24 h treatment to 15 mM H 2 O 2 or 200 μ M NaNO 2 for prolonged exposure. For pulse-treatment, cells were exposed to 100 mM H 2 O 2 or 1.2 M NaNO 2 in acidified LB (pH 6.5). After exposure, 10-fold serial dilutions (3 μ l) of the cells were spotted onto an LB agar plate to enumerate the survivor bacteria.
Virulence measurement. Determination of mortality from Drosophila systemic infections and mouse peritonitis-sepsis caused by P. aeruginosa cells that had been grown to the OD 600 of 3.0 was performed as previously described 30 . For Drosophila systemic infections, 3-to 6-day-old adult female flies (Oregon R) were infected by pricking at the dorsal thorax with a 10 mm needle (Ernest F. Fullam, Inc.). The needle dipped into bacterial suspensions diluted in 10 mM MgSO 4 containing 10 7 cfu. Fly mortality was monitored for 54 h postinfection. For murine peritonitis-sepsis, bacterial cells were harvested, washed twice with phosphate buffered saline (PBS) (2.7 mM KCl, 137 mM NaCl, 10 mM Na 2 HPO 4 , and 2 mM KH 2 PO 4 , pH 7.0) and diluted to 2 × 10 6 cfu in 100 ml of PBS containing 1% mucin as an adjuvant. Anesthetized 4-week-old mice (ICR) were intraperitoneally infected according to the institutional protocol approved by the Institutional Animal Care and Use Committee at CHA University. Kaplan-Meier analysis and log-rank tests were used to compare the virulence difference between the groups 30,31 . A p value of less than 0.01 was considered to be significant.
Scientific RepoRts | 6:31185 | DOI: 10.1038/srep31185 NO 2 − measurement. The steady-state level of NO 2 − that had been endogenously generated and transported during the normal growth in LB amended with 15 mM KNO 3 was measured based on Griess reaction. Culture aliquots (300 μ l) of the cells that had been grown to OD 600 of 0.7 were mixed with the equal volume of Griess reagent (Sigma, USA) and incubated at 37 °C for 10 min. The NO 2 − level in the supernatant was measured by the absorbance at 550 nm. Statistical significance between the groups is indicated, based on a p value of less than 0.01 by the Student's t test.