Stenotrophomonas maltophilia PhoP, a Two-Component Response Regulator, Involved in Antimicrobial Susceptibilities

Stenotrophomonas maltophilia, a gram-negative bacterium, has increasingly emerged as an important nosocomial pathogen. It is well-known for resistance to a variety of antimicrobial agents including cationic antimicrobial polypeptides (CAPs). Resistance to polymyxin B, a kind of CAPs, is known to be controlled by the two-component system PhoPQ. To unravel the role of PhoPQ in polymyxin B resistance of S. maltophilia, a phoP mutant was constructed. We found MICs of polymyxin B, chloramphenicol, ampicillin, gentamicin, kanamycin, streptomycin and spectinomycin decreased 2–64 fold in the phoP mutant. Complementation of the phoP mutant by the wild-type phoP gene restored all of the MICs to the wild type levels. Expression of PhoP was shown to be autoregulated and responsive to Mg2+ levels. The polymyxin B and gentamicin killing tests indicated that pretreatment of low Mg2+ can protect the wild-type S. maltophilia from killing but not phoP mutant. Interestingly, we found phoP mutant had a decrease in expression of SmeZ, an efflux transporter protein for aminoglycosides in S. maltophilia. Moreover, phoP mutant showed increased permeability in the cell membrane relative to the wild-type. In summary, we demonstrated the two-component regulator PhoP of S. maltophilia is involved in antimicrobial susceptibilities and low Mg2+ serves as a signal for triggering the pathway. Both the alteration in membrane permeability and downregulation of SmeZ efflux transporter in the phoP mutant contributed to the increased drug susceptibilities of S. maltophilia, in particular for aminoglycosides. This is the first report to describe the role of the Mg2+-sensing PhoP signaling pathway of S. maltophilia in regulation of the SmeZ efflux transporter and in antimicrobial susceptibilities. This study suggests PhoPQ TCS may serve as a target for development of antimicrobial agents against multidrug-resistant S. maltophilia.


Introduction
Stenotrophomonas maltophilia is a nonfermentative gram-negative bacillus that may cause nosocomial infections, especially affecting immunocompromised patients who have been hospitalized for prolonged periods and received broad-spectrum antibiotic therapy [1][2][3][4]. Therapy of these infections presents a significant challenge because of the intrinsic resistance of S. maltophilia to most of the currently used antimicrobial agents including carbapenems [1][2][3][4]. Isolation of multidrug-resistant S. maltophilia in intensive care settings has also been noted with increasing frequency [1,2,4]. Several molecular mechanisms contribute to multi-drug resistance of S. maltophilia, including multidrug efflux pumps and plasmids or integrons carrying resistant determinants such as L1/L2 β-lactamases and aminoglycoside-modifying enzymes [1][2][3]. The resistance-nodulation-division (RND)-type efflux pump is a common cause of the multidrug resistance of S. maltophilia. Eight putative RND-type efflux systems (SmeABC, Sme-DEF, SmeGH, SmeIJK, SmeMN, SmeOP, SmeVWX, and SmeYZ) were reported in the S. maltophilia [5]. In particular, it was noted that the SmIJK and SmeYZ pumps are constitutively expressed, and both are redundant in extrusion of aminoglycosides [6,7].
Cationic antimicrobial polypeptides (CAPs) are increasingly used to treat infections caused by multidrug-resistant bacteria. One of the important mechanisms of resistance to CAPs in gram-negative bacteria is modification of lipopolysaccharide (LPS) to remodel the composition of the outer membrane [8,9]. A polycistronic unit (arnBCADTEF or pmrHFIJKLM, called arn or pmr operon for short) is involved in LPS modification. Genes of the arn operon are necessary for biosynthesis and addition of 4-aminoarabinose (Ara4N) to the 4' phosphate of lipid Aa modification contributing to a reduction in the net negative charge of LPS and consequently decreasing binding of CAPs to the outer membrane [8,9]. In a large number of gram-negative species, the genes involved in LPS modification, are regulated by the bacterial two-component systems (TCSs) [10][11][12][13][14]. The Salmonella PmrAB and PhoPQ TCSs are involved in resistance to the CAPs such as polymyxin B. Transcriptional activation of the pmr operon requires the PmrAB, PmrB the sensor kinase and PmrA the cognate response regulator [9]. The PmrAB of Salmonella is activated by Fe 3+ , which is sensed by the PmrB protein, and by low Mg 2+ , which is sensed by the PhoQ sensor protein. The low Mg 2+ activation requires pmrD, a PhoPQ-regulated gene [13]. PhoP positively controls the pmr operon by increasing production of PmrD at the transcriptional level [13], which in turn activates the PmrA protein post-translationally, resulting in modification of LPS [12,13].
Seeing S. maltophilia polymyxin B resistance rate is up to 57.7% in the Asian-Pacific region [15] and MIC ranges of polymyxin B for clinical S. maltophilia isolates from National Taiwan University Hospital were 32-256 μg/ml (unpublished data), we investigated the role of PhoP in polymyxin B resistance. PhoP was shown to mediate polymyxin B resistance as expected. Surprisingly, we found PhoP is also associated with susceptibilities of S. maltophilia to other antimicrobials. Moreover, we found expression of SmeZ, an efflux transporter protein, was regulated by PhoP. This is the first report to describe the role of Mg 2+ -sensing PhoPQ TCS of S. maltophilia in regulation of the SmeZ efflux protein and in antimicrobial susceptibilities.

Bacterial strains, plasmids and growth condition
The bacterial strains and plasmids used in this study are listed in S1 Table. S. maltophilia S22 is a blood culture isolate from National Taiwan University Hospital. Bacteria were routinely cultured at 37°C in Luria-Bertani (LB) medium.

Gene-knockout by homologous recombination
For construction of the phoP mutant, sequences flanking the phoP were amplified by PCR using primer pairs PhoPQout up971-F/PhoPQout up971-R and PhoPQout down1022-F/ PhoPQout down1022-R (S2 Table), respectively and cloned into pGEM 1 -T Easy (Promega, USA) to generate pGphoP-up and pGphoP-dn. The pGphoP-up was digested with SalI/XbaI and the phoP upstream sequence-containing fragment was ligated to the SalI/XbaI-digested pGphoP-dn to produce the pGphoP-updn plasmid which contains both upstream and downstream sequences of phoP. The DNA fragment containing the combined upstream and downstream sequence of phoP was cleaved from pGphoP-updn by SalI/SphI and ligated into SalI/ SphI-cleaved pEX18Tc [7] to generate pEX18phoP-updn. The plasmid (pEX18phoP-updn) was mobilized into S. maltophilia S22 via conjugation [7,16]. The deleted allele was transferred to the S. maltophilia chromosome by double-crossover homologous recombination through sequential selection of tetracycline (30 μg/ml) and then 10% sucrose. The resultant mutants with correct double-crossover events were verified by PCR and sequencing as described previously [7,16]. smeZ mutant was constructed in the same way using primer pairs SmeZout up-F/ SmeZout up-R and SmeZout down-F/SmeZout down-R.

Construction of the PhoP-complemented strain
Since phoP knockout lowered the expression of phoQ mRNA (data not shown), full length phoPQ was used to construct the PhoP-complemented strain. Full length phoPQ was amplified by PCR using primer pair PhoPQcomplementation-F and PhoPQcomplementation-R (S2 Table) and cloned into PstI/KpnI digested pRK415 [7] to generate the phoP complementation plasmid, pRKphoPQ. The pRKphoPQ was then transformed into the phoP-knockout mutant to generate the PhoP-complemented strain.

MIC assay
In vitro determination of MICs for various drugs was performed by the broth microdilution method according to the guidelines proposed by Clinical and Laboratory Standard Institute [17].

Reporter assay
The promoter region of phoP was amplified by the primer pair, phoPpromoterF/phoPpromo-terR (S2 Table) and cloned into pGEM1-T Easy to generate pGphoPp. pGphoPp was cut with SacI/PstI and the promoter-containing fragment was ligated with the xylE containg pRK415 to construct the phoP-xylE reporter plasmid. The overnight cultures of the reporter plasmid-containing wild-type and phoP mutant were diluted to an OD 600 of 0.1 in the same medium and the XylE activity was measured as described previously [18] at the time points indicated after incubation at 37°C.
Purification of Glutathione S-transferase (GST)-tagged recombinant protein and electrophoretic mobility shift assay (EMSA) The full-length phoP gene was cloned into plasmid pGEX-4T-1 by BamHI and XhoI to generate plasmid pGEX-PhoP. Over-expression and purification of the GST-tagged PhoP was performed as described (GE Healthcare Life Sciences, USA). The purity of PhoP preparation was confirmed by SDS-PAGE. The promoter regions of phoP and smeZ were amplified by the primer pairs, phoPpromoterF/phoPpromoterR and smeZ EMSA-F/smeZ EMSA-R, respectively (S2 Table). The amplified promoter DNA fragments (0.1 μg each for the promoter DNA of phoP and smeZ) were incubated with 0, 0.25 or 0.5 μM GST-tagged PhoP protein in a 10-μl binding buffer [18]. After incubation for 30 min at room temperature, the reaction mixtures were loaded onto 5% non-denaturing polyacrylamide gels and the gel was electrophoresed at 100 V for 1.5-2 h before staining with ethidium bromide [18]. An EMSA using purified GSTtagged PhoP and the IRDye-labeled phoP promoter DNA was also performed. The unlabeled phoP promoter DNA fragments were included in the assay as the competitors to demonstrate the binding specificity of the phoP promoter DNA and GST-PhoP. The phoP promoter sequence was amplified from the pGphoPp using the IRDye1 700-labeled M13F/M13R primers (S2 Table, Integrated DNA Technologies). The DNA fragment amplified from the pGEM1-T Easy without the phoP promoter sequence by the same primer pair was used as the negative control. The amplified product (0.1 μg) was incubated with 0-0.5 μM GST-PhoP protein in a 10-μl binding buffer. After incubation for 30 min at room temperature, the reaction mixtures were analyzed as described above except that the gel image was obtained by the quantitative infrared fluorescent imaging system (Odyssey, LI-COR). For the competitive assay, the un-labeled phoP promoter DNA fragments (0.5, 1, 2 μg) amplified from pGphoPp using unlabeled M13 primers were included in the incubation solution.

Real-time reverse transcription (RT)-PCR
To investigate whether Mg 2+ is a signal of the PhoPQ TCS, we performed real-time RT-PCR to assess the expression of phoP using primers, phoP realtime-F and phoP realtime-R (S2 Table). Overnight cultures were diluted 100-fold, grown to OD 600 0.5 and supplemented with various concentrations of Mg 2+ in N-minimal medium (5 mM KCl, 7.5 mM (NH 4 ) 2 SO 4 , 0.5 mM K 2 SO 4 , 1 mM KH 2 PO 4 , 38 mM glycerol, 0.1% casamino acid, 0.1 M Tris-HCl, pH7.4). After incubation at 37°C for 5 h, total RNA was extracted and real-time RT-PCR was performed as described [18]. The mRNA levels were normalized against 16S rRNA. In addition, the effect of phoP mutation on expression of the nine efflux transporter genes (smeB, smeE, smeH, smeJ, smeK, smeN, smeP, smeW, and smeZ) was also examined in the same way without Mg 2+ treatment using primers previously described [19].
The polymyxin B and gentamicin killing tests for S. maltophilia in response to Mg 2+ levels We first tested the effect of pretreatment with various concentrations of Mg 2+ on the survival of the wild-type after challenge with polymyxin B [18]. The overnight 3-ml LB culture was washed with N-minimal medium without Mg 2+ three times and the pellet was resuspended in the same medium with Mg 2+ at 0.01, 10 and 20 mM. After incubation at 37°C for 5 h, cells were washed twice, resuspended in 1-ml LB to OD 600 0.5 and incubated with polymyxin B (64 μg/ml) at 37°C for 1 h. The numbers of CFU were determined after serial dilution and plating on LB agar plates. The percent survival in polymyxin B was calculated as follows: (CFU of polymyxin B-challenged culture/CFU of no challenge culture) x 100%.
To further test the effects of pretreatment with Mg 2+ on the survival of the wild-type, phoP mutant and phoP-complemented strain in drugs, we performed the polymyxin B and gentamicin killing tests after pretreatment with low (10 μM) and high (10 mM) concentrations of Mg 2+ . In the polymyxin B killing test, bacterial cells were incubated for 1 h in the presence or absence of polymyxin B (64 and 128 μg/ml) after pretreatment with Mg 2+ , while the wild-type, phoP mutant and phoPc strain were challenged or not with Gm at concentrations of 0.5x and 1x MICs for each strains after the same pretreatment in the gentamicin killing test. The relative survival (low Mg 2+ to high Mg 2+ ) was from the percent survival in low Mg 2+ divided by the percent survival in high Mg 2+ . The percent survival was calculated as described above.
ANS (8-anilino-1-naphthylenesulfonic acid) membrane permeability assay [20] To assess the integrity of bacterial cell membranes, the fluorescent probe, 8-anilino-1-naphthylenesulfonic acid (Sigma-Aldrich, USA) was used. ANS is a hydrophobic probe that exhibits enhanced fluorescence in nonpolar/hydrophobic environments. Overnight cultures were diluted 100-fold in 5 ml fresh LB medium and grown to an OD 600 of 0.5 at 37°C. A 1-ml cell culture was harvested by centrifugation and washed with phosphate buffered saline (PBS). Cells were then resuspended in 1-ml PBS containing ANS (final concentration 3 μM). The solution was kept at room temperature for 10 min in the dark and the fluorescence intensity of cells was quantified using a microplate reader with 375 nm excitation and 510 nm emission.
Triton X-100 sensitivity assay [21] Overnight cultures were diluted 100-fold in fresh LB medium and grown to an OD 600 of around 0.2 at 37°C. Triton X-100 (0.5%) was added or not into the bacterial culture and the growth was monitored by OD 600 determination at 30-min intervals after further incubation at 37°C.

Detection of aminoglycoside-modifying enzymes by the agar diffusion method
We detected the presence of aminoglycoside-modifying enzymes in wild-type S. maltophilia and phoP mutant using a protocol modified by Gad et al. [22]. Briefly, crude enzyme extracts were prepared after sonication of bacterial suspensions and centrifugation to obtain the supernatant. For a total of 50 μl reaction mixture, 25 μl of the enzyme extract (containing 1 mg/ml protein) or phosphate buffer saline (PBS, as the control) was incubated for 3 h in the presence of 22 nmoles of the tested aminoglycoside antibiotics, 50 nmoles dithiothreitol, 2.5 μmoles Tris pH 8.1, 120 nmoles ATP and 0.4 μmoles MgCl 2 . To monitor the extent of inactivation of the antibiotics, the reaction mixture was spotted on filter paper disks (10 μl per disk), which were placed on Mueller-Hinton agar plates seeded with a sensitive E. coli DH5α strain. Plates were incubated at 37°C for 24 h. In case of inactivation of the antibiotics, the presence of a crude enzyme extract reduces the inhibition zone (diameter) of the disk relative to that of the control disk (PBS). For method validation, two clinical isolates of E. coli (resistant to gentamicin and amikacin, respectively) and E. coli DH5α carrying a kanamycin resistance plasmid were included in the assay.

Nucleotide sequence accession numbers
The phoP and phoQ nucleotide sequences of S. maltophilia S22 have been deposited in Gen-Bank under accession no. KT001239.

Identification of PhoP and PhoQ in S. maltophilia
Many TCSs, including PhoPQ and RppAB, are known to regulate tolerance to CAPs in pathogenic bacteria [12,23,24]. Knowing MICs of polymyxin B, a naturally occurring CAP, may range from 32-256 μg/ml in clinical isolates of S. maltophilia (unpublished data), we investigated whether S. maltophilia PhoP protein is involved in tolerance to polymyxin B. First, we searched for phoP and phoQ homologues in the released S. maltophilia K279a genome and found the phoP and phoQ counterparts. We then amplified phoPQ genes using the genomic DNA of a clinical polymyxin B-resistant isolate (S22, MIC 128 μg/ml) with primers designed from the S. maltophilia K279a genome sequence and analyzed the product sequences. S1 Fig shows the alignment of the S. maltophilia S22 PhoP and PhoQ proteins with those of other bacteria. S. maltophilia S22 PhoP and PhoQ proteins share sequence identity with their counterparts in S. maltophilia K279a (92% and 99%), Xanthomonas campestris NCPPB 2251 (91% and 77%), Xanthomonas oryzae pv. oryzae KACC 10331 (91% and 76%), Pseudomonas aeruginosa PAO1 (55% and 34%) and Salmonella enterica serovar Typhimurium LT2 (41% and 33%).
Increased susceptibilities of S. maltophilia phoP mutant to polymyxin B and other antimicrobial agents To demonstrate the role of phoP gene in regulating polymyxin B susceptibility, we constructed phoP mutant through allelic exchange mutagenesis (see Materials and Methods). Growth was not affected in the phoP mutant (S2 Fig). We found phoP mutant was 64-fold more susceptible to polymyxin B than the wild-type S. maltophilia S22 (MIC 2 vs 128 μg/ml) ( Table 1). MICs of other drugs (except amikacin) tested for phoP mutant also decreased 2-16 fold (Table 1) relative to the wild-type, with an 8-and a 16-fold decrease in MIC for kanamycin and gentamicin, respectively. The complemented strain had restored MIC levels as the wild-type (Table 1).
Autoregulation and the response to Mg 2+ levels of S. maltophilia PhoP Salmonella PhoPQ [12] and Proteus RppAB [24], both involved in PB susceptibility, have been shown to sense Mg 2+ levels and to be autoregulated. To study whether PhoP is autoregulated by binding to the S. maltophilia phoP putative promoter, an EMSA was performed. The data in Fig 1A indicated that the purified GST-tagged PhoP proteins could bind specifically to the phoP promoter DNA fragment but not to the smeZ promoter DNA fragment (compare lanes 1-2 to lanes 3-5). Another EMSA with purified GST-tagged PhoP and the IRDye-labeled phoP promoter DNA fragment showed that a band shift was visible when the labeled phoP promoter DNA was incubated with enough GST-PhoP protein (lanes 3-5 vs 7 in Fig 1B) and the unlabeled phoP promoter DNA can act as a competitor to reduce the binding of the IRDye-labeled phoP promoter DNA with GST-PhoP (lane 7 vs 8-10 in Fig 1B). No shift band was observed when the IRDye-labeled negative control DNA was incubated with GST-PhoP (lanes 1-2 in Fig 1B). In addition, the wild-type carrying the phoP-xylE reporter plasmid also displayed higher XylE activity compared to the phoP mutant carrying the same reporter plasmid (Fig 2). Therefore, expression of S. maltophilia phoP gene is likely autoregulated by itself.
Knowing PhoP is likely autoregulated, we then investigated the expression of phoP in the wild-type S. maltophilia in the presence of different concentrations of Mg 2+ (in N-minimal medium) by the real-time RT-PCR assay. The phoP mRNA level of S. maltophilia was reduced  Fig  3). Thus, low Mg 2+ may serve as a signal to activate the expression of S. maltophilia PhoP.
Low Mg 2+ pretreatment protecting the wild-type S. maltophilia but not phoP mutant from killing by polymyxin B or gentamicin Knowing S. maltophilia PhoP regulates drug susceptibility and PhoP expression is responsive to Mg 2+ , we performed the MIC assay of polymyxin B and gentamicin after overnight pretreatment of S. maltophilia with low (10 μM) and high Mg 2+ (10 mM) in N-minimal medium. Compared to high Mg 2+ , low Mg 2+ pretreatment resulted in an increase in MICs of both drugs for the wild-type but not the phoP mutant (Table 2). Moreover, we found polymyxin B MIC for the wild-type increased 2-fold in the presence of the PhoP-expressing plasmid. To further demonstrate if PhoP induced by low Mg 2+ can protect S. maltophilia from antibiotic killing, the polymyxin B killing assay was performed after pretreating wild-type cells with Mg 2+ at 0.01, 10, and 20 mM. The survival rate in polymyxin B increased about 4-fold after pretreating with Mg 2+ at 0.01 mM relative to that of pretreating with Mg 2+ at 20 mM (left axis in Fig 3). Moreover, we tested the effect of pretreatment of low Mg 2+ on the survival of wild-type S. maltophilia, phoP mutant and phoP-complemented strain (phoPc) in either polymyxin B or gentamicin. The percent survival of phoP mutant pretreated with low and high Mg 2+ for 5 h was almost the same (i.e. relative survival rate of low to high Mg 2+ pretreatment near 1) after exposure to polymyxin B (64 or 128 μg/ml), while the wildtype and phoPc strains pretreated in the same way exhibited an increase in survival in low Mg 2+ , particularly in polymyxin B of 128 μg/ml (Fig 4A). In the gentamicin killing assay, the same pretreated-bacterial cells were challenged or not with Gm at concentrations of 0.5x and 1x MICs for each strains and the relative survival rate (low to high Mg 2+ ) were determined. Again, the relative survival rate in gentamicin for phoP mutant was near 1 and more than 1 for the wild-type and the phoPc strain ( Fig 4B). Together, these data indicate low Mg 2+ may protect S. maltophilia from killing by polymyxin B and gentamicin through a PhoP-dependent pathway. Regulation of S. maltophilia smeZ, a gene encoding an efflux pump protein, by PhoP It has been reported that bacteria can use the TCS in disposing chemicals, including antibiotics, through regulating the efflux pumps [11,[25][26][27][28]. Therefore, we investigated if S. maltophilia PhoP would regulate efflux pumps to affect antimicrobial susceptibilities. Inner membrane transporter genes (smeB, smeE, smeH, smeJ, smeK, smeN, smeP, smeW, and smeZ) of each RND efflux system were selected as PhoP-regulated targets for real-time RT-PCR to evaluate the expression of each pump in the wild-type, phoP mutant and complemented strain (phoPc).
The results revealed only smeZ expression was significantly low in the phoP mutant relative to the wild-type and phoPc strain (Fig 5). Since SmeZ has been known to participate in extrusion of aminoglycosides [7], we examined aminoglycoside MICs for S. maltophilia S22 smeZ mutant. As expected, aminoglycoside (except spectinomycin) MICs for smeZ mutant were 2-4 fold lower than those of the wild-type. Loss of smeZ didn't affect MICs of ampicillin and chloramphenicol but did alter polymyxin B MIC (Table 1). Together, the data suggest that PhoP could affect aminoglycoside susceptibilities through regulating expression of SmeZ transporter. We observed MICs of gentamicin, kanamycin and streptomycin decreased both for phoP mutant and smeZ mutant relative to the wild-type (Table 1). It is worth noting that except amikacin and streptomycin, all drugs tested showed increased potency for the phoP mutant relative to the smeZ mutant (Table 1).

Increased membrane permeability in phoP mutant
PhoPQ-mediated regulation was shown to produce a more robust permeability barrier in the outer membrane of S. enterica Serovar Typhimurium and Neisseria [14,29]. They found that phoP null mutant produces an outer membrane severely compromised in its barrier function that is important in making the pathogen more resistant to the stresses, including antimicrobial agents. In view of the increased drug susceptibilities of S. maltophilia phoP mutant, the fluorescent probe, 8-anilino-1-naphthylenesulfonic acid (ANS), was used to assess the integrity of the bacterial outer membrane. We found the outer membrane permeability of phoP mutant increased compared to that of the wild-type, phoPc strain and smeZ mutant (Fig 6A). It seems that somehow the barrier function of the outer membrane in the mutant is disturbed. To further characterize the change in permeability, Triton X-100, a nonionic surfactant, was used to test its effect on the wild-type, phoP mutant and phoPc strain. Similarly, we found the phoP mutant was more sensitive to the treatment of Triton X-100 than the wild-type and smeZ mutant (Fig 6B). Together, the data indicate the permeability change may be also one of the reasons for the susceptibilities of phoP mutant to the drugs tested. Comparison of aminoglycoside MICs for the phoP mutant and the smeZ mutant (Table 1), it is likely that both the cell  . We performed the PB and Gm killing tests for the wild-type S. maltophilia (wt), phoP mutant (phoP) and phoPcomplemented strain (phoPc) after pretreatment with low (10 μM) and high (10 mM) concentrations of Mg 2+ . In the PB killing assay, bacterial cells were challenged with PB (64 and 128 μg/ml) or not. In the gentamicin killing test, bacterial cells were challenged or not with Gm at concentrations of 0.5x and 1x MICs for each strains. The relative survival was calculated as described in Materials and Methods. The relative survival for the wild-type, phoP mutant and phoPc strain without challenge with PB or Gm was set at 1. The data represent the averages and standard deviations of three independent experiments. Significant difference between the relative survival of the drug challenge and no challenge for the wild-type and the phoPc strain was observed by Student's t-test analysis (*, p<0.05; **, p<0.005).
doi:10.1371/journal.pone.0153753.g004 The increase of aminoglycoside susceptibility in phoP mutant was not related to the absence of aminoglycoside-modifying enzymes Because aminoglycoside-modifying enzyme is an important consideration for aminoglycoside resistance in S. maltophilia, we assessed if aminoglycoside-modifying enzymes exist in phoP mutant by the agar diffusion method. Table 3 shows that the extent of inactivation of gentamicin, amikacin and kanamycin by enzyme extracts from wild-type S. maltophilia and phoP mutant was similar to that of PBS control. In contrast, the extracts from a gentamicin and an amikacin resistant clinical isolates of E. coli exhibited a significant inactivation of gentamicin

Discussion
The PhoPQ TCS governs the adaptation to low Mg 2+ environments and other stress conditions by regulating expression of as much as 1% of the genes in many gram-negative bacterial species [11,12,30]. For example, CAP resistance is known to be regulated by the PhoPQ system in response to Mg 2+ [10][11][12][13]. In this study, we described that PhoP could regulate antimicrobial susceptibilities in S. maltophilia. On one hand, the PhoP-mediated SmeZ expression could extrude aminoglycosides; on the other the cell permeability barrier regulated by PhoP also may contribute to drug susceptibilities. Several lines of evidence suggest that S. maltophilia PhoP should be the PhoP counterpart in other bacteria. First, sequence analysis of S. maltophilia PhoP realveals homology with other PhoP proteins, up to 91% identity and 92% similarity to Xanthomonas oryzae PhoP which has been shown to confer tolerance to CAPs in response to Mg 2+ [23]. Second, S. maltophilia phoP mutant exhibited increased sensitivity to PB. Third, S. maltophilia PhoP expression is autoregulated and can be activated by low Mg 2+ . It is becoming increasingly clear that the TCS-mediated stress responses are linked to antimicrobial resistance [31]. In this regard, a number of TCSs of important bacterial pathogens, such as CesRK of Listeria monocytogenes [32], CpxAR of Klebsiella pneumoniae [28], AdeRS and BaeSR of Acinetobacter baumannii [25,26], and PprBA and PhoPQ of P. aeruginosa [11,33] were shown to regulate antibiotic susceptibility. Seeing PhoPQ mediates responses to stresses, such as growth-limiting Mg 2+ , acids and antimicrobial peptides [12,23], it seems reasonable for us to discover S. maltophilia PhoP is involved in drug susceptibility.
How could a TCS affect drug tolerance? First, several efflux pumps have been shown to be regulated by the TCS [11,[25][26][27][28]. In E. coli, CpxR-induced MarA activates expression of the AcrAB/TolC tripartite multidrug pump, thus leading to drug resistance [34]. For the first time, we found PhoP regulates expression of the efflux transporter SmeZ (Fig 5). Although the EMSA strongly suggested the lack of PhoP interaction with smeZ promoter region of S. maltophilia (Fig 1A), the result may be due to low affinity between them or a co-factor needed. Second, the TCS-mediated surface changes may alter cell permeability and thus affect drug susceptibility [29,33,35,36]. In view of PhoPQ-mediated regulation producing a more robust permeability barrier in the outer membrane of S. enterica serovar Typhimurium [29], the finding of increased membrane permeability in S. maltophilia phoP mutant (Fig 6) may explain the increased susceptibility of the mutant to the drugs tested in this study. The increased permeability of the outer membrane (by ANS test) may be indicative of a general decrease in outer membrane integrity in the S. maltophilia phoP mutant and thus a greater susceptibility to membrane perturbation by Triton X-100, lending support to the view of a membrane stabilization role for S. maltophilia PhoP. Cell membrane permeability-associated drug susceptibilities may arise from changes in outer membrane proteins to limit the drug access to bacterial cells [33,36,37] and TCSs may influence the expression and/or membrane localization of a number of membrane proteins to regulate bacterial cell permeability [33]. Nevertheless, our preliminary data indicated no obvious difference in the profile of outer membrane proteins between wildtype and S. maltophilia phoP mutant (data not shown). In addition, PhoPQ-mediated LPS modification is known to be associated with outer membrane permeability and CAP susceptibility [35]. More efficient permeability barrier in the PhoP-constitutive mutant has also been confirmed by higher drug resistance and the known PhoP-dependent lipid A modifications explain most of the alterations in the outer membrane permeability [29]. In this regard, analysis of the LPS profile of S. maltophilia phoP mutant is ongoing.
One susceptibility profile in Table 1 is affected by smeZ mutation only (amikacin) and the other by both phoP and smeZ mutation (streptomycin, gentamicin, kanamycin, polymyxin B). Among the latter, streptomycin tolerance is possibly through PhoP-mediated SmeZ extrusion only (the same MICs for phoP and smeZ mutants). However, PhoP-mediated both SmeZdependent and -independent pathways may confer tolerance to gentamicin and kanamycin (MIC for phoP mutant lower than smeZ mutant). In view of the antibacterial mechanism of PB and constitutive expression of S. maltophilia SmeZ [7], increased PB susceptibility in smeZ mutant should be due to the membrane disturbance [38]. As expected, S. maltophilia smeZ mutant (defective in aminoglycoside extrusion) has increased susceptibilities to streptomycin, gentamicin, kanamycin and amikacin. The limited contribution of SmeZ to kanamycin and amikacin susceptibilities is similar to the report of Crossman et al. [5] but much smaller than that reported by Lin et al. [7]. Interestingly, we found loss of S. maltophilia phoP decreased MICs of all aminoglycosides tested except amikacin. Based on the PhoP-mediated SmeZ expression shown in Fig 5, it doesn't make sense to see amikacin susceptibility is affected by smeZ mutaion but not phoP mutation. In this regard, QseEF and QseBC TCSs were shown to be able to sense the same signal (epinephrine) to regulate formation of attaching and effacing lesion in Enterohemorrhagic E. coli [39]. Hence it is feasible that amikacin may induce other systems to compensate for the defect of SmeZ in the S. maltophilia phoP mutant. Regarding to this issue, a TCS operon is located upstream of smeZ, work is under way to disclose the role of the system in SmeZ-mediated drug resistance.
Several lines of evidence support that the Mg 2+ -sensing PhoPQ system can regulate expression of the LPS-modification operon, arn, leading to polymyxin B resistance [11,12]. Both pmrD gene and arn operon are not present in the released S. maltophilia genome. In S. maltophilia S22, we did find homologues of ugd and galU which have been shown to be involved in LPS modification and polymyxin B resistance [40] but both are not regulated by PhoP (data not shown). Systematic approaches to dissect the PhoP regulon relating to susceptibility to polymyxin B are under way.
In conclusion, for the first time, we reported S. maltophilia PhoP was involved in drug susceptibilities. It is likely that S. maltophilia PhoP regulates expression of downstream genes which may affect membrane permeability and expression of SmeZ efflux protein in response to a certain environmental cue, low Mg 2+ , thus resulting in drug tolerance. Further studies aimed at defining genes that fall into the PhoP regulon, as well as unravelling their complex interaction, should greatly aid our understanding of antibiotic resistance in S. maltophilia. This study highlights the possibility to develop antimicrobial agents against multidrug-resistant S. maltophilia by targeting the PhoPQ TCS.