Analysis of the Networks Controlling the Antimicrobial-Peptide … · also be effective against host APs, thus contributing to bacterial resistance and survival in host tissues. To

INFECTION AND IMMUNITY, Sept. 2011, p. 3718–3732 Vol. 79, No. 90019-9567/11/$12.00 doi:10.1128/IAI.05226-11Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Analysis of the Networks Controlling the Antimicrobial-Peptide-DependentInduction of Klebsiella pneumoniae Virulence Factors�†

Enrique Llobet,1,2 Miguel A. Campos,1 Paloma Gimenez,1,2 David Moranta,1,2

and Jose A. Bengoechea1,2,3*Laboratory Microbial Pathogenesis, Fundacio d’Investigacio Sanitaria de les Illes Balears, Recinto Hospital Joan March, 07110,

Bunyola, Spain1; Program Host-Pathogen interactions, Centro de Investigacion Biomedica en Red Enfermedades Respiratorias,Bunyola, Spain2; and Consejo Superior de Investigaciones Científicas, Madrid, Spain3

Received 12 April 2011/Returned for modification 2 May 2011/Accepted 18 June 2011

Antimicrobial peptides (APs) impose a threat to the survival of pathogens, and it is reasonable to postulatethat bacteria have developed strategies to counteract them. Polymyxins are becoming the last resort to treatinfections caused by multidrug-resistant Gram-negative bacteria and, similar to APs, they interact with theanionic lipopolysaccharide. Given that polymyxins and APs share the initial target, it is possible that bacterialdefense mechanisms against polymyxins will be also effective against host APs. We sought to determine whetherexposure to polymyxin will increase Klebsiella pneumoniae resistance to host APs. Indeed, exposure of K.pneumoniae to polymyxin induces cross-resistance not only to polymyxin itself but also to APs present in theairways. Polymyxin treatment upregulates the expression of the capsule polysaccharide operon and the locirequired to modify the lipid A with aminoarabinose and palmitate with a concomitant increase in capsule andlipid A species containing such modifications. Moreover, these surface changes contribute to APs resistanceand also to polymyxin-induced cross-resistance to APs. Bacterial loads of lipid A mutants in trachea and lungsof intranasally infected mice were lower than those of wild-type strain. PhoPQ, PmrAB, and the Rcs systemgovern polymyxin-induced transcriptional changes, and there is a cross talk between PhoPQ and the Rcssystem. Our findings support the notion that Klebsiella activates a defense program against APs that iscontrolled by three signaling systems. Therapeutic strategies directed to prevent the activation of this programcould be a new approach worth exploring to facilitate the clearance of the pathogen from the airways.

Antimicrobial peptides (APs) are ubiquitous in nature, andin vertebrates they belong to the arsenal of weapons of theinnate immune system against infections. There are four struc-tural classes of APs: the disulfide-bonded �-sheet peptides, theamphipathic �-helical peptides, the extended peptides, and theloop-structured peptides (8, 32, 50). Despite their diverse sizeand structures, nearly all APs have a net positive charge, andthe three-dimensional folding results in an amphipathic struc-ture (8, 32, 50). In most cases, the action of APs is initiatedthrough electrostatic interaction with the bacterial surface (8,32, 50, 62) and, in the case of Gram-negative bacteria, APsinteract with the anionic lipid A moiety of the lipopolysaccha-ride (LPS) (8, 32, 50, 62).

APs impose a threat to the survival of pathogens, and there-fore it is reasonable to postulate that bacteria have developedmeans to sense the presence of APs in order to activate coun-termeasures to limit their effectiveness. Furthermore, given theimportance of APs in host defense, it is likely that these coun-termeasures will be important virulence factors. Bacteria uti-lize phosphorelay signaling cascades in the form of two-com-ponent systems to respond and adapt to different hostile

environments. The sensors of these two-component systemsrespond to particular cues by modulating the phosphorylationstatus of their cognate regulators which are often transcriptionfactors. As a result, genes necessary for growth and survival areupregulated, whereas genes deleterious for infectivity might bedownregulated. It can be speculated that bacteria may utilizetwo-component systems to transduce AP-mediated signals,leading to the activation of bacterial defense mechanisms. Sup-porting this idea, the Salmonella PhoPQ two-component sys-tem regulates genes necessary for intracellular survival andcellular invasion, and it is required for resistance to a subset ofAPs (6, 20, 21, 25, 26).

Polymyxin B (PxB) and PxE (colistin) are two antibioticsoriginally derived from Bacillus polymyxa and made availablefor clinical use in the late 1950s and early 1960s. Polymyxinsare pentacationic amphipathic lipopeptide antibiotics charac-terized by a heptapeptide ring and a fatty acid tail (63). Poly-myxins are active against Gram-negative bacteria and, similarto APs, they do interact with the anionic LPS. Soon after theirintroduction, the clinical use was limited due to perceived toxicside effects and the emergence of new antimicrobials (17, 35).However, the occurrence of multidrug-resistant Gram-nega-tive bacteria has prompted researchers to reconsider poly-myxin therapies (24, 63). Nevertheless, the pharmacokineticsand pharmacodynamics of polymyxins are poorly understood,making it possible that bacteria are exposed to sublethal con-centrations during treatment. Consequently, the possibility ex-ists that bacteria may activate defense mechanisms againstpolymyxins. Furthermore, given that polymyxins and APs sharethe initial target, it is possible that PxB countermeasures will

* Corresponding author. Mailing address: Laboratory MicrobialPathogenesis, Fundacio d’Investigacio Sanitaria de les Illes Balears,Recinto Hospital Joan March, Carretera Soller Km 12, 07110 Bunyola,Spain. Phone: 34 971 011780. Fax: 34 971 011797. E-mail: [email protected].

† Supplemental material for this article may be found at http://iai.asm.org/.

� Published ahead of print on 27 June 2011.

3718

on March 11, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

also be effective against host APs, thus contributing to bacterialresistance and survival in host tissues.

To study this hypothesis, we used the Gram-negative humanpathogen K. pneumoniae. The frequent isolation of K. pneu-moniae multidrug-resistant strains makes polymyxins a thera-peutic option (24). There is a paucity of information on themechanisms of resistance of this pathogen against polymyxinsand APs. By mass spectrometry and genetic methods, we dem-onstrate that PxB indeed induces the expression of loci con-ferring resistance against PxB but also against host APs. Wedemonstrate that these loci play an important role in K. pneu-moniae virulence. Finally, we show that at least three signalingtransduction systems govern PxB-induced changes.

MATERIALS AND METHODS

Bacterial strains and growth conditions. The bacterial strains and plasmidsused in the present study are listed in Table 1. Strains were grown in Luria-Bertani (LB) medium at 37°C. When appropriate, antibiotics were added to thegrowth medium at the following concentrations: rifampin at 25 �g/ml, ampicillinat 100 �g/ml for K. pneumoniae and 50 �g/ml for Escherichia coli; kanamycin(Km) at 100 �g/ml; and chloramphenicol (Cm) at 12.5 �g/ml.

K. pneumoniae 52145 mutant construction. Primers for mutant construction(Table 2) were designed based on the available genome sequence of K. pneu-moniae subsp. pneumoniae MGH78578 (GenBank accession no. CP000647.1).DNA fragments for phoPQ, rcsB, and pagP were PCR amplified, gel purified, andcloned into pGEM-T Easy (Promega) to obtain pGEMTphoPQ, pGEMTrcsB,and pGEMTpagP, respectively. These plasmids were amplified by inverse PCRusing the method described by Byrappa et al. (9) to delete internal coding regionsof phoQ, rcsB, and pagP. A Km resistance cassette, obtained as a 1.4-kb PstIblunt-ended fragment from pUC-4K (Pharmacia), was cloned into the plasmidsobtained by inverse PCR to generate pGEMT�phoPQGB and pGEMT�pagPGB.�phoPQ::GB, �rcsB, and �pagP::GB alleles were obtained by PvuII digestion ofpGEMT�phoPQGB, pGEMT�rcsB, and pGEMT�pagPGB, respectively, gel-puri-fied and cloned into SmaI-digested pMAKSACB. pMAKSACB is a suicide vector thatcarries a rep101ts origin of replication, an oriT sequence for conjugational transfer, anda Cm resistance marker (19). It also carries the sacB gene that mediates sucrose sensi-tivity as a positive selection for the excision of the vector after double crossing-over (19).pMAKSAC�phoPQGB, pMAKSAC�rcsB, and pMAKSAC�pagPGB were electro-porated into E. coli S17-1�pir, from which the plasmids were mobilized into K.pneumoniae 52145. Transconjugants were selected after growth on LB platessupplemented with Cm at 30°C. Bacteria from 10 individual colonies were pooledin 500 �l of phosphate-buffered saline (PBS), serially diluted in PBS, and spreadonto LB plates with Cm that were incubated at 42°C in order to select mero-diploids in which the suicide vector was integrated into the chromosome byhomologues recombination. A total of 5 to 10 merodiploids were serially dilutedin PBS, and dilutions were spread in LB plates without NaCl containing 10%sucrose and were incubated at 30°C. The recombinants that survived 10% su-crose were checked for their antibiotic resistance, and the appropriate replace-ment of the wild-type alleles by the mutant ones was confirmed by PCR (data notshown). Recombinants selected were named 52145-�phoQGB; 52145-�rcsB, and52145-�pagPGB. To confirm that pagP mutation has no polar effects, the ex-pression of the downstream gene, cpsE, was analyzed by real-time quantitativePCR (RT-qPCR). Briefly, 200 ng of cDNA, obtained by retrotranscription of 2�g of total RNA using a commercial RT2 first-strand kit (Superarray BioscienceCorp.), were used as a template in a 25-�l reaction mixture containing 1� SYBRgreen RT2 qPCR master mix (Superarray Bioscience Corp.) and primer mix.rpoD was amplified as control. RT-qPCR analyses were performed as previouslydescribed (48). The expression of cpsE by 52145-�pagPGB was similar to that bythe wild-type strain (data not shown).

DNA fragments for pmrAB and pmrF were PCR amplified, gel purified, andcloned into pGEM-T Easy (Promega) to obtain pGEMTpmrAB and pGEMTpmrF, respectively. These plasmids were amplified by inverse PCR to delete theinternal coding regions of pmrAB and pmrF, respectively. A Km resistancecassette, obtained as a 1.5-kb PCR fragment from pKD4 (14), was cloned intothe plasmids obtained by inverse PCR to generate pGEMT�pmrABKm andpGEMT�pmrFKm. �pmrAB::Km and �pmrF::Km alleles were obtained byPvuII digestion of pGEMT�pmrABKm, and pGEMT�pmrFKm, respectively,and cloned into SmaI-digested pKOV (42). Recombinants in which the wild-typeallele was replaced by the mutant one were selected as described previously and

named 52145-�pmrABKm, and 52145-�pmrFKm. The kanamycin cassette wasexcised by Flp-mediated recombination using plasmid pFLP2 (37), and the gen-erated mutants were named 52145-�pmrAB and 52145-�pmrF. RT-qPCR anal-ysis revealed that the expression of pmrI, the pmrF downstream gene, was notsignificantly different between pmrF mutant and the wild-type strain (data notshown).

To obtain a pmrD mutant, two sets of primers (Table 2) were used to obtaintwo different pmrD fragments, PmrDUP and PmrDDown. These fragments wereannealed at their overlapping region and amplified by PCR as a single fragment,which was cloned into pGEM-T Easy to obtain pGEMT�pmrD. A Km cassettewas PCR amplified from pKD4 and cloned into pGEM-T Easy to give pGEMTFRTKM. The cassette was obtained as a BamHI fragment, which was cloned intoBamHI-digested pGEMT�pmrD to generate pGEMT�pmrDKm. The �pmrD::Kmallele was gel purified after PvuII digestion of pGEMT�pmrDKm and cloned intoSmaI-digested pKOV (42). Recombinants in which the wild-type allele wasreplaced by the mutant one were selected as described previously and named52145-�pmrDKm. The kanamycin cassette was excised by Flp-mediated recom-bination using plasmid pFLP2 (37), and the generated mutant was named 52145-�pmrD.

The 52145-�phoQGB-�pmrAB and 52145-�phoQGB-�rcsB double mutantswere obtained mobilizing the pMAKSAC�phoPQGB plasmid into 52145-�pmrAB and 52145-�rcsB, respectively. The replacement of the wild-type allelesby the mutant ones was done as described above and confirmed by PCR (data notshown).

Construction of reporter fusions. DNA fragments containing the promoterregions of the cps, pmrH, pagP, mgtA, pmrD, phoP, ugd, rcsD, and rcsC geneswere amplified by PCR using Vent polymerase, EcoRI digested, gel purified, andcloned into EcoRI-SmaI-digested pGPL01 suicide vector (29). This vector con-tains a promoterless firefly luciferase gene (lucFF) and an R6K origin of repli-cation. Plasmids in which lucFF was under the control of the Klebsiella promoterswere identified by restriction digestion analysis and named pGPLKpnPcps,pGPLKpnPmrH, pGPLKpnPagP, pGPLKpnMgtA, pGPLKpnPmrD, pGPLKpnPhoP,pGPLKpnPugd, pGPLKpnPrcsD, and pGPLKpnPrcsC, respectively. Plasmidswere electroporated into the different Klebsiella strains used in the present study.Strains in which the suicide vector was integrated into the genome by homolo-gous recombination were selected. This was confirmed by Southern blot (datanot shown).

Luciferase activity. The reporter strains were grown on an orbital incubatorshaker (180 rpm) until late log phase and, when required, PxB (1 �g/ml) wasadded, and the culture was incubated for 1 h more. At the end of the incubationthe optical density at 540 nm (OD540) was recorded. A 100-�l aliquot of thebacterial suspension was mixed with 100 �l of luciferase assay reagent (1 mMD-luciferin [Synchem] in 100 mM citrate buffer [pH 5]). The luminescence wasimmediately measured with a LB9507 Luminometer (Berthold) and expressed asrelative light units/OD540. All measurements were carried out in quintuplicate onat least three separate occasions.

Antimicrobial peptide susceptibility assay. Bacteria were grown at 37°C in 5ml of LB medium and harvested (2,500 � g, 20 min, 24°C) in the exponentialgrowth phase (OD600 � 0.6). When required, PxB (1 �g/ml) was added, and theculture incubated for 1 h more. Bacteria were washed once with PBS, and asuspension containing �106 CFU/ml was prepared in 10 mM PBS (pH 6.5), 1%tryptone soy broth (Oxoid), and 100 mM NaCl. Aliquots (5 �l) of this suspensionwere mixed in 1.5-ml microcentrifuge tubes with various concentrations of AP. Inall cases, the final volume was 30 �l. After 1 h of incubation at 37°C, the contentsof the tubes were plated on LB agar. Colony counts were determined, and theresults were expressed as percentages of the colony count of bacteria not exposedto antibacterial agents. All experiments were performed with duplicate sampleson at least four independent occasions.

The 50% inhibitory concentration (IC50) of AP was defined as the concentra-tion producing a 50% reduction in the colony counts compared to bacteria notexposed to the antibacterial agent. According to guidelines of the NationalInstitutes of Health Chemical Genomics Center (www.ncgc.nih.gov), the IC50 ofa given AP was determined from dose-response curve data fit using a standardfour-parameter logistic nonlinear regression analysis. Dose-response experi-ments were performed on four independent occasions. The results are reportedas means standard deviations.

Isolation and analysis of lipid A. Lipid A’s were extracted using an ammoniumhydroxide/isobutyric acid method and subjected to negative-ion matrix-assistedlaser desorption ionization time-of-flight (MALDI-TOF) mass spectrometryanalysis (16, 53). Briefly, lyophilized bacteria (10 mg) were resuspended in 400 �lof isobutyric acid–1 M ammonium hydroxide (5:3 [vol/vol]) and incubated in ascrew-cap test tube at 100°C for 2 h, with occasional vortexing. Samples werecooled in ice water and centrifuged (2,000 � g for 15 min). The supernatant was

VOL. 79, 2011 POLYMYXIN INDUCES K. PNEUMONIAE COUNTERMEASURES 3719


http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

transferred to a new tube, diluted with an equal volume of water, and lyophilized.The sample was then washed twice with 400 �l of methanol and centrifuged(2,000 � g for 15 min). The insoluble lipid A was solubilized in 100 to 200 �l ofchloroform-methanol-water (3:1.5:0.25 [vol/vol/vol]). Analyses were performedon a Bruker Autoflex II MALDI-TOF mass spectrometer (Bruker Daltonics,Inc.) in negative reflective mode with delayed extraction. The ion-acceleratingvoltage was set at 20 kV. To analyze the samples, few microliters of lipid Asuspension (1 mg/ml) were desalted with a few grains of ion-exchange resin(Dowex 50W-X8; H) in a 1.5-ml microcentrifuge tube. Then, a 1-�l aliquot of

the suspension (50 to 100 �l) was deposited on the target and covered with thesame amount of dihydroxybenzoic acid matrix (Sigma Chemical Co., St. Louis,MO) dissolved in 0.1 M citric acid. Different ratios between the samples anddihydroxybenzoic acid were used when necessary. Alternatively, lipid A wasmixed with 5-chloro-2-mercapto-benzothiazole (Sigma Chemical Co., St. Louis,MO) at 20 mg/ml in chloroform-methanol (1:1 [vol/vol]) at a ratio of 1:5. Eachspectrum was an average of 300 shots. A peptide calibration standard (BrukerDaltonics) was used to calibrate the MALDI-TOF. Further calibration for lipidA analysis was performed externally using lipid A extracted from E. coli strain

TABLE 1. Bacterial strains and plasmids used in this study

Bacterial strain or plasmid Genotype or commentsa Source orreference(s)

StrainsEscherichia coli

C600 thi thr leuB tonA lacY supE 2XL1-Blue recA1 endA1 gyrA96 thi-1 hsdR17 supE44S17-1�pir recA thi pro hsd(r� m) RP4::2-Tc::Mu::Km, Tn7� pir

Klebsiella pneumoniaeKp52145 Clinical isolate (serotype O1:K2); Rifr 13, 4952145-�wcaK2 Kp52145, �manC; Rifr; the manC gene of the wca gene cluster was inactivated; no CPS

expression43

52145-�phoQGB Kp52145, �phoQ::Km-GenBlock; Rifr Kmr; the phoQ gene was inactivated This study52145-�pmrAB Kp52145, �pmrAB; Rifr; the pmrAB genes were inactivated This study52145-�rcsB Kp52145, �rcsB; Rifr; the rcsB gene was inactivated This study52145-�pmrDKm Kp52145, �pmrD::Km; RifrKmr; the pmrD gene was inactivated This study52145-�pmrD Kp52145, �pmrD; Rifr, the pmrD gene was inactivated This study52145-�pmrF Kp52145, �pmrF; Rifr; the pmrF gene was inactivated This study52145-�pagPGB Kp52145, �pagP::Km-GenBlock; Rifr Kmr; the pagP gene was inactivated This study52145-�pmrAB-�phoQGB 52145-�pmrAB, �phoQ::Km-GenBlock; Rifr Kmr; the phoQ and pmrAB genes were inactivated This study52145-�rcsB-�phoQGB 52145-�rcsB, �phoQ::Km-GenBlock; Rifr Kmr; the phoQ and rcsB genes were inactivated This study52145-�wcaK2-�pmrF 52145-�wcaK2; �pmrF; Rifr Kmr; the pmrF gene was inactivated in a CPS mutant background This study52145-�wcaK2-�pagPGB 52145-�wcaK2; �pagP::Km-GenBlock; Rifr Kmr; the pagP gene was inactivated in a CPS

mutant backgroundThis study

PlasmidspGEM-T Easy Cloning plasmid; Ampr PromegapGPL01 Firefly luciferase transcriptional fusion suicide vector; Ampr 29pMAKSACB Suicide vector, Psc101 replication origin, Mob, sacB gene; Cmr 19pKOV Suicide vector, Psc101 replication origin, sacB gene; Cmr 42pUC-4K Source GenBlock; Ampr Kmr PharmaciapKD4 Km cassette source for one-step mutagenesis protocol 14pGEMTFRTKm Km cassette source for mutagenesis flanked by BamHI-FRT sites This studypFLP2 Plasmid encoding FLP to remove cassettes between FRT sites; mobilizable; sacB for

counterselection37

pGEMT�phoQGB pGEM-T Easy containing �phoQ::Km-GenBlock; Ampr Kmr This studypGEMT�pagPGB pGEM-T Easy containing �pagP::Km-GenBlock; Ampr Kmr This studypGEMT�rcsB pGEM-T Easy containing �rcsB; Ampr This studypGEMT�pmrABKm pGEM-T Easy containing �pmrAB::Km; Ampr Kmr This studypGEMT�pmrD pGEM-T Easy containing �pmrD; Ampr This studypGEMT�pmrDKm pGEM-T Easy containing �pmrD::Km; Ampr Kmr This studypGEMT�pmrFKm pGEM-T Easy containing �pmrF::Km; Ampr Kmr This studypMAKSACB�phoQGB pMAKSACB containing �phoQ::Km-GenBlock; Cmr Kmr This studypMAKSACB�pagPGB pMAKSACB containing �pagP::Km-GenBlock; Cmr Kmr This studypMAKSACB�rcsB pMAKSACB containing �rcsB; Cmr This studypKOV�pmrAB pKOV containing �pmrAB::Km; Cmr Kmr This studypKOV�pmrD pKOV containing �pmrD; Cmr This studypKOV�pmrF pKOV containing �pmrF::Km; Cmr Kmr This studypGPLKpnPcps pGPL01 containing cps promoter region; Ampr This studypGPLKpnPmrH pGPL01 containing pmrH promoter region; Ampr This studypGPLKpnPagP pGPL01 containing pagP promoter region; Ampr This studypGPLKpnMgtA pGPL01 containing mgtA promoter region; Ampr This studypGPLKpnPmrD pGPL01 containing pmrD promoter region; Ampr This studypGPLKpnPhoP pGPL01 containing phoP promoter region; Ampr This studypGPLKpnRcsD pGPL01 containing rcsD promoter region; Ampr This studypGPLKpnRcsC pGPL01 containing rcsC promoter region; Ampr This study

a Cmr, chloramphenicol resistance; Rifr, rifampin resistance; Ampr, ampicillin resistance; Kmr, kanamycin resistance. The wcaK2 gene is manC.

3720 LLOBET ET AL. INFECT. IMMUN.


http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

MG1655 grown in LB medium at 37°C. Interpretation of the negative-ion spectrais based on earlier studies showing that ions with masses higher than 1,000 gavesignals proportional to the corresponding lipid A species present in the prepa-ration (3, 41, 52, 58). Important theoretical masses for the interpretation of peaksfound in the present study were as follows: C14:OH, 226; C12, 182, C14, 210;aminoarabinose (Ara4N), 131; and C16, 239.

Capsule polysaccharide (CPS) purification and quantification. Cell-associ-ated CPSs from K. pneumoniae strains, grown in 5 ml of LB, were obtained usingthe hot phenol-water method exactly as previously described (10). CPS was

quantified by determining the concentration of uronic acid in the samples, usinga modified carbazole assay (7), exactly as described by Rahn and Whitfield (57).

Intranasal infection model. Six- to seven-week-old virus-free female C57BL/6JOlaHsd mice (Harlan) were anesthetized by intraperitoneal injection with amixture containing ketamine (50 mg/kg) and xylazine (5 mg/kg). Bacteria weregrown at 37°C in 5 ml of LB medium, harvested (2,500 � g, 20 min, 24°C) in theexponential growth phase, resuspended in PBS, and adjusted to 106 CFU/ml.Portions (20 �l) of the bacterial suspension were inoculated intranasally in four5-�l aliquots. To facilitate consistent inoculations, mice were held vertically

TABLE 2. Primers used in this study

Method and target gene Primer Sequence (5�–3�)a

MutagenesisphoP KpnphoPQmutr CCGGAATTCCGAACATCTCCCGGATATCG

KpnphoPQmutf GTTCGATAAAGTCGGGCCAGrcsBF CGGGATCCCTGACGGTCCGCCGCCATGC

rcsB rcsBR CGGGATCCCCAGCTGGAGAGCTGGGGAGrcsBinvF GACAACTGTCCATCCCCGATTCrcsBinvR TAATCAGCGTGATCCCGTCGKpnPagPF GACATTCACCATTACCGGGAAC

pagP KpnPagPR TGCGCTGGTGGCTCGGCATGKpnPagPinvL CATTCCAGGTCTGCGCTACGKpnPagPinvR GGTTTCGGCGTCTCGCGCTGKpnpmrABR CAAGCTTGTGGCCAAAGCCATTGGCGAG

pmrAB KpnpmrABF GGAATTCCAACGATAACGACGGCGGCTGKpnpmrABinvR GGGTACCCGCTACAGCCCTGAAGGGTCGKpnpmrABinvF GGGTACCCTCCCGCCGCATGCGGGAGAGMutpmrDupF TGGGCAAAGGTCGCCTGGTC

pmrD MutpmrDupR CGGATCCGCGGCAGAGGCTGGCCGAAGCMutpmrDdownF CGGATCCGCTGAATAATAATCCGACGCAAACMutpmrDdownR TCCAGCAAATAGTTGCGGAACKpnpmrFF CGGATCCACCTGCGCGAGCTGGCGGAC

pmrF KpnpmrFR CGGATCCCGGCGTCATCCGCGCCAATCKpnpmrFinvF TCTCCTCCGGCGGGTTTTGCKpnpmrFinvR CAAATACAGCTTTATGCGCCTG

Promoter regioncps K2ProcpsF GGAATTCCTGCTGGGACAAATTGCCACC

K2ProcpsR AGATGGATGACCCCGCGATCphoP ProPhoPF GGGTACCCTCTTCATCCGGCAGGCCGAG

ProPhoPR GGAATTCCGGCCGCCGAAGAAGGCTTCGrcsC ProrcsCF GGAATTCCATGAAGAGCTGGATGCCATCG

ProrcsCR GGGTACCCGGGGTTCAAAGTTGGGCACCrcsD KpnPYojNF ATTTCCGCGCGCCGCGACTC

KpnPYojNR GGAATTCCCGTTCGAGATCTCCGATTTGGpagP PKpnPagpF AGATAATGGCCGCGATGGAG

KpnPagPinvR CATTCCAGGTCTGCGCTACGpmrD PropmrDF GGAATTCCTTTTATCTTCCTCCGGCAAAG

PropmrDR TCTGAAGCACGACGCGGCAGpmrH PknppmrHF GGAATTCCGCGATGCCGGCCCGGCCTAC

PkpnpmrHR GGGTACCGTCGCATGACGGTTGCCGGTCugd PromkpnugdF GATCGTAGTCGGTCGGCGTG

PromKpnugdR GGAATTCCAGTCCCAGAAAGGCGTATTGCmgtA ProKpnMgtAF GGAATTCCTCTTTGATGGTCAGCCGGTTC

ProKpnMgtAR AGGTCCACAGATGCACCCACC

Km cassetteCassette-F1 CGCGGATCCGTGTAGGCTGGAGCTGCTTCGCassette-R1 CGCGGATCCCATGGGAATTAGCCATGGTCC

RT-qPCRpmrI KpnyfbGF1 CGCTGGATCTACTCGGTCTC

KpnyfbGR1 TCTTTGTTCTCGATGATGCGcspE KpncspEF1 TCATTACTCCGGAAGATGGC

KpncspER1 CCTTTGGCACCGTTAGTGATrpoD KpnrpoDLEFT CCGGAAGACAAAATCCGTAA

KpnrpoDRIGHT CGGGTAACGTCGAACTGTTT

a The restriction site is indicated by underlining.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

during inoculation and placed on a 45° incline while recovering from anesthesia.At the indicated times after infection, mice were euthanized by cervical disloca-tion. The trachea and lungs were aseptically removed, weighed, and homoge-nized in 500 �l of PBS for bacterial load determination. The results werereported as the log CFU per g of tissue. Mice were treated in accordance with theEuropean Convention for the Protection of Vertebrate Animals used for Exper-imental and other Scientific Purposes (Directive 86/609/EEC) and in agreementwith the Bioethical Committee of the University of the Balearic Islands.

Statistical analysis. Statistical analyses were performed using one-way analysisof variance with Bonferroni contrasts or the two-tailed t test or when, therequirements were not met, by the Mann-Whitney U test. A P value of 0.05 wasconsidered statistically significant. The analyses were performed using Prism4 forPC (GraphPad Software).

RESULTS

Exposure of K. pneumoniae to PxB increases the resistanceto antimicrobial peptides. We sought to determine whetherexposure to PxB increases the resistance of Kp52145, the wild-type strain, to PxB. The results shown in Fig. 1A demonstratethat 1 h of treatment with PxB (1 �g/ml) resulted in a markedlyincreased resistance to killing by this agent. To test whether

exposure to PxB also induces cross-resistance to other APs,killing assays were performed with human �-defensin 1(hBD1) and hBD2, HNP-1, and magainin II. hBD1 is consti-tutively expressed in the airways (46). hBD2 is produced byairway epithelial cells upon induction by cytokines or by thepresence of pathogens (33, 34, 46, 59), and its levels increaseseveralfold in the lungs during pneumonia (36). HNP-1 is pro-duced by neutrophils and released to the medium after degran-ulation (22). Magainin II is an AP produced by frogs, and it iswidely used as a model AP (45). Bacteria treated with PxBwere also more resistant against hBD1, hBD2, HNP-1, andmagainin II than untreated bacteria (Fig. 1). In summary, theseobservations demonstrate that K. pneumoniae resistance toAPs can be induced by exposure to PxB.

Exposure of K. pneumoniae to PxB increases capsule expres-sion and LPS modifications. Recently, we have shown that K.pneumoniae CPS acts as a protective shield against APs (10),whereas released CPS traps APs, thereby blocking their bac-tericidal activity (43). Moreover, there is a correlation between

FIG. 1. Exposure of K. pneumoniae 52145 to PxB increases the resistance to antimicrobial peptides. After PxB treatment, bacteria were washed,and the susceptibility to PxB (A), �-defensin 1 (B), �-defensin 2 (C), HNP-1 (D), or magainin II (E) was tested by the survival assay. Each pointrepresents the mean and standard deviation of eight samples from four independently grown batches of bacteria, and significant survival differences(P 0.05 [two-tailed t test]) between PxB-pretreated (solid symbols) and untreated (open symbols) bacteria are indicated by asterisks.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

the amount of CPS and the resistance to PxB (10). Therefore,the observed PxB-induced resistance could be mediated by anincrease in CPS expression. Indeed, exposure of Kp52145 toPxB (1 �g/ml) upregulated the transcription of the cps operon(Fig. 2A) with a concomitant increased in the amount of cell-bound CPS (64.4 3 �g/104 CFU versus 113 5 �g/104 CFU,respectively; P 0.05 [two-tailed t test]). To test whetherPxB-induced resistance was solely due to the increased expres-sion of CPS, we analyzed the effect of PxB on 52145-�wcaK2 aCPS mutant. Figure 2 demonstrates that exposure of 52145-�wcaK2 to PxB also induced cross-resistance to PxB (Fig. 2B),hBD1 (Fig. 2C), hBD2 (Fig. 2D), HNP-1 (Fig. 2E), andmagainin II (Fig. 2F). To compare the PxB-induced levels ofresistance against the different APs between Kp52145 and52145-�wcaK2, we determined the IC50s of APs for thesestrains pretreated with PxB. The IC50s of PxB, hBD1, hBD2,HNP-1, and magainin II for PxB-treated Kp52145 were 4.5 0.8, 19 1.1, 10.2 0.6, 48 2.1, and 56 �g/ml, respectively,values which were significantly higher than those of PxB,hBD1, hBD2, HNP-1, and magainin II for PxB-treated 52145-�wcaK2 (2.6 0.7, 12.5 0.9, 7.0 0.5, 14 3.1, and 36.8�g/ml, respectively; P 0.05 for each comparison versusKp52145 values [two-tailed t test]). Taken together, these ob-servations suggest that PxB-induced resistance is in part CPS-dependent, but there are CPS-independent mechanisms oper-ating as well.

Bacteria can modify the lipid A part of LPS by addingaminoarabinose, phosphoethanolamine, or palmitate to re-duce the interaction of the peptides with the lipid A (28, 30, 31,40, 51). We speculated that the CPS-independent PxB-inducedmechanism(s) of resistance could involve changes in the lipidA structure. To explore this notion, we determined the struc-ture of lipid A extracted from Kp52145 after exposure to PxBby MALDI-TOF mass spectrometry (Fig. 3). Lipid A fromKp52145 grown without PxB contained predominantly hexa-acylated species (m/z 1,824) corresponding to two gluco-samines, two phosphates, four 3-OH-C14, and two C14. Otherpeaks (m/z 1,840) may represent a hexa-acylated lipid A con-taining two glucosamines, two phosphates, four 3-OH-C14, oneC14, and one C14:OH (hydroxymyristate). Minor species (m/z1,797) may correspond to a hexa-acylated lipid A containingfour 3-OH-C14, one C12, and one C14. Other minor speciesdetected were consistent with the addition of aminoarabinose(m/z 1,955) to the hexa-acylated form (m/z 1,824) or palmitateto the hexa-acylated species (m/z 1,797 and 1,824), hence pro-ducing hepta-acylated lipid A’s (m/z 2,036 and 2,063) (Fig. 3A).PxB induced an increase in the relative abundance of theminor lipid A species containing aminoarabinose (m/z 1,955)and palmitate (m/z 2,036 and 2,063) (Fig. 3B). Collectively,these results might indicate that exposure to PxB increases K.pneumoniae lipid A modifications containing aminoarabinoseand palmitate.

FIG. 2. Treatment of the K. pneumoniae 52145 cps mutant with PxB increases resistance to antimicrobial peptides. (A) Analysis of theexpression of the cps operon by Kp52145 carrying the fusion cps::lucFF. The strain was treated with PxB for 1 h (f) or not treated (�). The dataare presented as means the standard deviations (n � 6). *, Results are significantly different (P 0.05; two-tailed t test) from the results fornontreated bacteria. (B to F) After treatment of 52145-�wcaK2 with PxB for 1 h, the bacteria were washed and exposed to different concentrationsof PxB (B), �-defensin 1 (C), �-defensin 2 (D), HNP-1 (E), and magainin II (F). Each point represents the mean and standard deviation of eightsamples from four independently grown batches of bacteria, and significant survival differences (P 0.05 [two-tailed t test]) between bacteriapretreated with PxB (solid symbols) and untreated bacteria (open symbols) are indicated by asterisks.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

Perusal of the literature shows that the products of ugd andpmrHFIJKLM (arnBCADTEF) (hereafter referred to as thepmrF operon) loci are required for the synthesis and additionof aminoarabinose to lipid A. Ugd converts UDP-D-glucoseinto UDP-D-glucuronic acid, which is next modified by pmrFoperon-encoded enzymes to generate aminoarabinose (56).The gene coding for the acyltransferase pagP is required forthe addition of palmitate to lipid A (31). To verify that theseloci were indeed implicated in PxB-triggered lipid A modifica-tions with aminoarabinose and palmitate, we analyzed the lipidA structure from 52145-�pmrF, 52145-�pagPGB, and 52145-�pmrF-�pagPGB mutants. The three strains expressed thesame amounts of cell-bound CPS (65.6 8 �g/104 CFU,69.5 7 �g/104 CFU, and 61.6 9 �g/104 CFU, respectively)as Kp52145 (64.4 3 �g/104 CFU; P � 0.05 for each compar-ison versus Kp52145 value [two-tailed t test]). The resultsshown in Fig. 4 demonstrate that the pmrF mutant grown withPxB lacked lipid A species containing aminoarabinose,whereas the pagP mutant grown with PxB did not containspecies containing palmitate. 52145-�pmrF-�pagPGB lackedspecies containing aminoarabinose and palmitate. It should benoted that lipid A species containing hydroxymyristate (C14:OH;m/z 1,840) and palmitate (m/z 2,036 and 2,063) were not af-fected in 52145-�pmrF, whereas lipid A species containinghydroxymiristate and aminoarabinose (m/z 1,955) were notaffected in 52145-�pagPGB.

These findings led us to study whether PxB upregulatesthe expression of ugd, pmrF operon, and pagP. To monitortranscription of these loci quantitatively, three transcrip-

tional fusions were constructed in which a promoterlesslucFF was under the control of the loci promoter regions.The fusions were introduced into Kp52145, and the lucifer-ase activity was determined. PxB upregulated the expressionof ugd::lucFF, pmrH::lucFF, and pagP::lucFF transcriptionalfusions (Fig. 5), thereby giving experimental support to ourhypothesis.

Aminoarabinose and palmitate lipid A substitutions con-tribute to K. pneumoniae antimicrobial peptide resistance. Wesought to determine the contribution of aminoarabinose andpalmitate lipid A substitutions to resistance against PxB andmagainin II. 52145-�pmrF was more susceptible to PxB thanthe wild-type strain but as susceptible as 52145-�wcaK2,whereas 52145-�wcaK2-�pmrF was the most susceptible strain(Fig. 6A). 52145-�pmrF was as resistant as Kp52145 tomagainin II, indicating that the lipid A decoration with ami-noarabinose is not implicated in the resistance to this AP (Fig.6A). On the other hand, lipid A substitution with palmitatedoes not play any role in PxB resistance but does play a role inresistance to other peptides, such as magainin II (31). Asexpected, 52145-�pagPGB was as resistant as Kp52145 to PxB(data not shown). However, 52145-�pagPGB was more suscep-tible than Kp52145 to magainin II (Fig. 6B). 52145-�wcaK2-�pagPGB, lacking CPS and palmitate, was the most susceptiblestrain to magainin II (Fig. 6B).

We assessed the contribution of aminoarabinose or palmi-tate lipid A modifications to PxB-induced resistance to PxBand magainin II. The IC50 of PxB for PxB-treated Kp52145 was4.5 0.8 �g/ml, which is significantly higher than those for

FIG. 3. Exposure of K. pneumoniae 52145 to PxB affects the lipid A structure. Negative ion MALDI-TOF mass spectrometry spectra of lipidA isolated from Kp52145 that was treated with 65 ng of PxB/ml for 12 h (B) or not treated (A). The results in the panels are representative ofthree independent lipid A extractions. (C) Proposed structures corresponding to major peaks and acyl group positions follow previously reportedstructures for Klebsiella (11) and other Gram-negative bacteria.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

52145-�pmrF, 52145-�wcaK2 and 52145-�wcaK2-�pmrF (1.3 0.2, 2.6 0.5, and 0.3 0.2 �g/ml, respectively; P 0.05 foreach comparison versus Kp52145 value [two-tailed t test]). PxBtreatment also increased the resistance to magainin II of

52145-�pagPGB (IC50 � 39.4 2.3 �g/ml) but the level waslower than that observed in the wild-type strain (IC50 � 56 4.3 �g/ml; P 0.05 [two-tailed t test]) and similar to that of52145-�wcaK2 (IC50 � 41.3 4.2 �g/ml; P � 0.05 [two-tailedt test]). The lowest IC50 of magainin II was observed for 52145-�wcaK2-�pagPGB (IC50 � 30.3 1.7 �g/ml).

Collectively, these data support the notion that CPS and thelipid A substitutions with aminoarabinose and palmitate con-tribute to AP resistance in K. pneumoniae and also to PxB-induced cross-resistance to APs.

Virulence of K. pneumoniae lipid A mutans. To determinethe ability of 52145-�pmrF, 52145-�pagPGB, and 52145-�pmrF-�pagPGB to cause pneumonia, C57BL/6JOlaHsd micewere infected intranasally, and bacterial loads at 24 and 96 hpostinfection in the trachea and lung homogenates were de-termined (Fig. 7). At 24 h postinfection, all strains colonizedtrachea and lungs, although the bacterial loads of mutantstrains were lower than those of the wild type in both organs(Fig. 7A). A similar picture was observed at 96 h postinfection(Fig. 7B). The bacterial loads of mutants were not significantlydifferent in either trachea or lungs at 24 and 96 h postinfection(Fig. 7).

FIG. 4. Lipid A analysis from K. pneumoniae lipid A mutants. Negative-ion MALDI-TOF mass spectrometry spectra of lipid A isolated fromthe indicated K. pneumoniae strains treated with 65 ng of PxB/ml for 12 h (B, D, and F) or not treated (A, C, and E) are shown. The results inall panels are representative of three independent lipid A extractions.

FIG. 5. PxB induces the expression of K. pneumoniae 52145 ugd,pmrF, and pagP loci. The expression of the loci implicated in lipid Aremodeling was analyzed by measuring the luciferase activity ofKp52145 carrying ugd::lucFF, pmrF::lucFF or pagP::lucFF transcrip-tional fusions, which were either treated with PxB for 1 h (f) or nottreated (�). The data are presented as means the standard devia-tions (n � 5). *, Results are significantly different (P 0.05 [two-tailedt test]) from the results for nontreated bacteria.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

Signaling networks controlling PxB induced capsule andLPS lipid A modifications. Having established that PxB up-regulated the expressions of cps, pmrF, ugd, and pagP, wesought to identify the regulatory architecture that mediatesPxB-triggered upregulation of them. The Rcs (Regulator ofcapsule synthesis) phosphorelay system consists of three pro-teins; RcsC, RcsD (also called YojN), and RcsB, the latterbeing a cytoplasmic response regulator (44). The Rcs systemfine-tunes the expression of CPS in several Enterobacteriaceae,and it mediates AP resistance (15, 65). PhoPQ and PmrABtwo-component systems mediate AP resistance by activatingloci, leading to lipid A remodeling, including pagP, pmrFoperon, and ugd (25, 27). In Salmonella enterica, the expressionof pmrH and ugd is controlled by PmrAB, whose activity can be

modulated by the PhoPQ-dependent PmrD connector proteinat the posttranscriptional level (25, 27, 39).

To define the contribution of these systems to PxB-inducedupregulation of cps, pmrF, ugd, and pagP, we investigated thetranscription of these loci in isogenic mutants with or withoutPxB treatment. Basal levels of the cps transcriptional fusionwere lower in 52145-�rcsB and 52145-�pmrAB than inKp52145 (Fig. 8A). PxB induced the fusion only in the 52145-�rcsB, 52145-�pmrD and 52145-�pmrAB backgrounds and tothe same levels obtained in Kp52145 (Fig. 8A). PxB treatmentupregulated the expression of the pmrH transcriptional fusionin 52145-�rcsB, 52145-�phoQ, 52145-�pmrAB, and 52145-�pmrD backgrounds, although the levels obtained in the52145-�rcsB background were significantly higher than those

FIG. 6. Roles of K. pneumoniae 52145 capsule and lipid A modifications on the resistance to antimicrobial peptides. (A) pmrF mutants wereexposed to different concentrations of PxB; and magainin II. (B) pagP mutants were exposed to different concentrations of magainin II. Each pointrepresents the mean and the standard deviation of eight samples from four independently grown batches of bacteria. Symbols: �, Kp52145; F,52145-�wcaK2; f, 52145-�pmrF; ‚, 52145-�wcaK2-�pmrF; ƒ, 52145-�pagPGB; �, 52145-�wcaK2-�pagPGB.

FIG. 7. Virulence of K. pneumoniae 52145 lipid A mutants. The bacterial counts in mouse organs at 24 h (A) or 96 h (B) postinfection weredetermined. Mice were infected intranasally with a bacterial mixture containing 4.1 � 104 bacteria of the wild-type strain (Kp52145, F), 4.5 � 104

bacteria of 52145-�pmrF (pmrF, E), 5.1 � 104 bacteria of 52145-�pagPGB (pagP, �), and 4.8 � 104 bacteria of 52145-�pmrF-�pagPGB(pmrF-pagP, ‚), respectively. The results were reported as log CFU per gram of tissue (log CFU/g). *, Results are significantly different (P 0.05[two-tailed t test]) from the results for Kp52145.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

observed in the other strains (Fig. 8B). A similar picture wasobserved for ugd::lucFF (Fig. 8C). PxB did not upregulate theexpression of pmrH and ugd fusions in the mutant lacking bothphoQ and pmrAB (Fig. 8B and C). PxB upregulated thepagP::lucFF fusion only in Kp52145, 52145-�rcsB, 52145-�pmrAB, and 52145-�pmrD (Fig. 8D), and the levels foundin the 52145-�rcsB background were the highest obtained(Fig. 8D).

To further sustain the role of PhoPQ in PxB-induced tran-scriptional effects, we analyzed whether PxB upregulates the

expression of mgtA, whose expression is PhoPQ dependent(60). As expected, PxB upregulated the expression of mgtA inKp52145, 52145-�pmrAB, and 52145-�pmrD backgrounds tosimilar levels but not in 52145-�phoQ mutant (Fig. 8E). Inter-estingly, mgtA expression in the 52145-�rcsB background wassignificantly higher than those in the other strains (Fig. 8E).Finally, we tested whether PxB induced the expression of thepmrD connector in a PhoPQ-dependent manner. Indeed, thiswas the case (Fig. 8F). Furthermore, PxB-induced levels of thepmrD fusion were similar in Kp52145 and 52145-�pmrAB

FIG. 8. K. pneumoniae PhoPQ and PmrAB two-component systems control PxB-induced transcriptional changes. Analysis of the expression ofcps, pmrH, ugd, pagP, mgtA, and pmrD loci by Kp52145 (WT), 52145-�rcsB (rcsB), 52145-�phoQGB (phoQ), 52145-�pmrAB (pmrAB), 52145-�pmrD (pmrD), and 52145-�phoQGB-�pmrAB (phoQ-pmrAB) carrying the transcriptional fusions cps::lucFF (A), pmrF::lucFF (B), ugd::lucFF (C),pagP::lucFF (D), mgtA::lucFF (E), and pmrD::lucFF (F) treated with PxB for 1 h (f) or not treated (�). The data are presented as means thestandard deviations (n � 3). *, Results are significantly different (P 0.05 [two-tailed t test]) from the results for nontreated bacteria. ‚, Resultsare significantly different (P 0.05 [two-tailed t test]) from the results for Kp52145 treated in the same manner.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

backgrounds, and the highest levels were observed again in the52145-�rcsB mutant background (Fig. 8F).

In summary, these data indicate that PhoPQ is necessary forPxB-triggered induction of cps and pagP. PxB induction ofpmrH and ugd was only abolished in the double-mutant phoQ-pmrAB, suggesting that both two-component systems can pro-mote PxB-induced lipid A modification with aminoarabinose.Confirming this hypothesis, lipid A species containing amino-arabinose were only absent in the PxB-treated phoQ-pmrABdouble mutant background (see Fig. S1 in the supplementalmaterial).

Cross talk between Rcs and PhoPQ systems. Consideringthat pmrH, ugd, pagP, mgtA, and pmrD loci were overexpressedin the rcsB mutant background and that PhoPQ regulates theirexpression, we hypothesized that the expression of phoPQcould be upregulated in the rcsB mutant background. To ex-plore this, the expression of phoP::lucFF was measured indifferent genetic backgrounds upon PxB treatment. The datashown in Fig. 9 indicate that phoP was overexpressed in 52145-�rcsB but downregulated in 52145-�phoQ (Fig. 9A). PxB up-regulated phoP transcription in Kp52145 and 52145-�rcsBbackgrounds, with the highest levels being those found in thelatter. This was dependent on PhoPQ because phoP transcrip-tion was not upregulated in a double mutant lacking rcsB andphoQ (Fig. 9A). phoP expression was not affected in 52145-�pmrAB background (data not shown). Altogether, these datagave experimental support to the hypothesis that phoPQ isupregulated in the rcsB mutant background. To further sustainthis notion, we sought to determine whether the expressions ofpagP, mgtA, and pmrD are upregulated in the 52145-�rcsB-phoQ mutant. As expected, this was not the case (Fig. 9B to D).Furthermore, PxB did not upregulate these loci in the 52145-�rcsB-phoQ background (Fig. 9B to D), which is in good agree-ment with the findings showing that PhoPQ is necessary fortheir PxB-mediated induction. Finally, we sought to determinewhether the upregulation of pmrH and ugd obtained in the rcsBmutant background was also dependent on PhoPQ. Indeed,the expression of pmrH and ugd was not upregulated in 52145-�rcsB-phoQ (Fig. 9E and F). In sharp contrast to pagP, mgtA,and pmrD, the expressions of pmrH and ugd were still inducedby PxB in 52145-�rcsB-phoQ (Fig. 9E and F), which is consis-tent with our data showing that PmrAB also promotes PxB-induced upregulation of these loci (Fig. 7).

We explored whether the expression of the Rcs system isaffected in phoQ, pmrAB, phoQ-pmrAB, and pmrD mutantbackgrounds. To monitor transcription of the Rcs system, weanalyzed the expression of rcsD::lucFF and rcsC::lucFF tran-scriptional fusions. Basal levels of the rcsD fusion were lower in52145-�phoQ and 52145-�phoQ-pmrAB than those obtainedin Kp52145, 52145-�pmrAB, and 52145-�pmrD, which werenot significantly different between them (Fig. 9G). A similarpicture was observed for the rcsC fusion (Fig. 9H). rcsD tran-scription was downregulated in 52145-�rcsB (Fig. 9G),whereas the expression of the rcsC fusion was abolished in52145-�rcsB (Fig. 9H). The former result is in good agreementwith the autoregulation of the Rcs system, whereas the latterone is consistent with the fact that RcsB is essential for rcsCexpression (44). PxB treatment upregulated the expression ofrcsD; however, PxB-induced rcsD expression was lower in52145-�phoQ and 52145-�phoQ-pmrAB than in Kp52145,

52145-�pmrAB, and 52145-�pmrD (Fig. 9G). PxB did not in-duce the fusion in 52145-�rcsB (Fig. 9G). Similar results wereobserved when the expression of rcsC was analyzed upon PxBtreatment (Fig. 9H).

On the whole, these data support the notion that there iscross talk between the Rcs and PhoPQ systems. Whereas theRcs system downregulates phoP, the PhoPQ system promotesthe expression of rcsD and rcsC.

Finally, we tested the susceptibility to PxB of the three trans-duction systems. The results shown in Fig. 10 demonstrate thatthe most susceptible strain was 52145-�rcsB-phoQ (IC50 �0.42 0.2 �g/ml), followed by 52145-�phoQ-pmrAB (IC50 �0.77 0.1 �g/ml), the two single mutants 52145-�phoQ(IC50 � 1.17 0.2 �g/ml) and 52145-�pmrAB (IC50 � 1.25 0.3 �g/ml), and Kp52145 (IC50 � 2.1 0.3 �g/ml). 52145-�rcsB (IC50 � 2.3 0.4 �g/ml) was as susceptible as the wildtype.

DISCUSSION

In the present study, we provide new insights into how abacterial pathogen activates countermeasures to fight againstAPs. Our findings revealed that brief treatment of a virulentisolate of K. pneumoniae with PxB induces cross-resistance toAPs found in humans, as well as to magainin II and PxB.Mechanistically, PxB triggers changes in K. pneumoniae sur-face that contribute to AP resistance and PxB-induced cross-resistance. Finally, we demonstrate that lipid A modificationsare important for K. pneumoniae survival in the airways.

Our data showed that PxB-induced cross-resistance to APsnot structurally related (32, 62), thereby indicating that PxB-triggered resistance is not specific for the compound used.Since APs share the initial electrostatic interaction with theanionic bacterial surface, we hypothesized that PxB treatmentshould affect the bacterial surface. Indeed, PxB treatment up-regulated the expression of the cps operon and the loci re-quired to modify the lipid A with aminoarabinose and palmi-tate with a concomitant increase in CPS and lipid A speciescontaining such modifications. Moreover, these surfacechanges were linked to AP resistance. Previous data (10, 43),further confirmed here, had shown that K. pneumoniae CPSmediates resistance to several APs. In the present study, wealso demonstrate that lipid A modifications with aminoarabi-nose and palmitate are required for AP resistance. It should benoted that these Klebsiella countermeasures are not redundantsince double mutants lacking CPS and lipid A modificationswere more susceptible to APs than the single mutants. Al-though here we have just considered surface changes, we donot rule out that other bacterial systems could also be affected.In fact, we put forward the notion that upon challenge withPxB, or other APs such as defensins, K. pneumoniae may alterglobal gene expression. It is tempting to speculate that thesebacterial global changes could be a “molecular pattern” de-voted to counteract the innate immune system, including theactivation of inflammatory responses and the microbicidal ac-tion of professional phagocytes (macrophages and neutro-phils). Studies are ongoing to confirm this hypothesis.

Lipid A analysis revealed the presence of a species (m/z1,840) consistent with the presence of hydroxymyristate(C14:OH). This species has been previously reported for K. pneu-



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

moniae (11) and S. enterica serovar Typhimurium (23). For thelatter, the dioxygenase responsible for 2-hydroxylation, namedLpxO, has been identified (23). This enzyme generates 2-hy-droxymyristate by hydroxylation of the myristate fatty acidtransferred to lipid A by the acyltransferase MsbB/LpxM (23).In silico analysis of the available K. pneumoniae genomes re-

vealed that this pathogen may encode an orthologue of LpxO.Studies are ongoing to characterize K. pneumoniae LpxO andwhether this lipid A modification plays any role in the resis-tance to APs. A remaining question is to explain at the molec-ular level the species (m/z 1,744) found only in strains treatedwith PxB. This species is consistent with elimination of one

FIG. 9. There is cross talk between the Rcs and PhoPQ systems in K. pneumoniae 52145. Analysis of the expression of phoP, pagP, mgtA, pmrD,pmrH, ugd, rcsD, and rcsC loci by Kp52145 (WT), 52145-�rcsB (rcsB), 52145-�phoQGB (phoQ), and 52145-�rcsB-�phoQGB (rcsB phoQ) carryingthe transcriptional fusions phoP::lucFF (A), pagP::lucFF (B), mgtA::lucFF (C), pmrD::lucFF (D), pmrH::lucFF (E), ugd::lucFF (F), rcsD::lucFF (G),and rcsC::lucFF (H) treated with PxB for 1 h (f) or not treated (�). The data are presented as means the standard deviations (n � 3). *, Resultsare significantly different (P 0.05 [two-tailed t test]) from the results for nontreated bacteria. ‚, Results are significantly different (P 0.05[two-tailed t test]) from the results for Kp52145 treated in the same manner.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

phosphate from the molecular ion m/z 1,824. Among otherpossibilities, it can be speculated that PxB treatment activatesa lipid A phosphatase. In fact, lipid A dephosphorylation me-diates AP resistance in Porphyromonas, Rhizobium, Francisella,and Helicobacter spp. (12, 38, 61, 64). Future studies will at-tempt to identify this putative K. pneumoniae lipid A phospha-tase.

Our analysis of the regulatory architecture governing PxB-induced changes revealed that the two-component systemPhoPQ is necessary for the PxB induction of cps, pagP, mgtA,and pmrD. These findings further support the notion that brieftreatment with APs activates the PhoPQ regulon, as first re-ported by Bader et al. in S. enterica serovar Typhimurium (4).However, the fact that PxB treatment increased the expres-sions of rcsC, rcsD, pmrH, and ugd in the phoQ mutant suggeststhat PhoPQ is not the only sensor/regulator governing PxB-induced AP resistance in K. pneumoniae. Indeed, our resultsdemonstrate that PmrAB also transduces the PxB-dependentregulatory signal. Thus, only in the phoQ pmrAB double-mu-tant background did PxB not upregulate the expressions ofpmrH and ugd, a finding consistent with the presence of PhoPand PmrA boxes in the promoter regions of both loci (47). Thisis in contrast to what happens in Salmonella since, under theconditions used here, the PxB-dependent upregulation ofpmrH and ugd is dependent on PmrAB via the activation ofPhoPQ and PmrD (27, 28, 30, 39). Interestingly, PxB treatmentactivated the Rcs system, as detected by the upregulation ofthe rcsC and rcsD transcriptional fusions. Despite the fact thatthe levels of both fusions were 20% lower in phoQ and phoQpmrAB mutants than in Kp52145, PxB still increased the ex-pression of rcsC and rcsD, suggesting that, at least, an addi-tional regulatory system may govern PxB-induced changes.Our data suggest that indeed the Rcs system could be thisother system since PxB no longer induces the rcsD fusion in thercsB mutant. Furthermore, our observation that the most sus-ceptible strain to PxB was the rcsB phoQ double mutant sug-gests that there are Rcs-dependent responses implicated in the

resistance to APs. This does not contradict the fact that thercsB mutant was as susceptible as the wild-type strain to PxBbecause we have demonstrated that PhoPQ-dependent re-sponses were upregulated in this genetic background. There-fore, it is tempting to conclude that these putative Rcs-depen-dent AP countermeasures play a role only in the absence of thePhoPQ-dependent ones.

An important issue is to understand how these signalingsystems sense APs to activate a transcriptional program. It hasbeen proposed that Salmonella PhoQ binds APs in an acidicpatch of the periplasmic domain, which leads to PhoQ auto-phosphorylation and the subsequent phosphotransfer to PhoP(5, 54, 55). K. pneumoniae PhoQ also contains this acidic patch,and therefore a similar mechanism of PhoQ activation by APsmight be expected. At present, we can only speculate on howthe PmrAB and Rcs systems are activated by APs. To date, nostructural study has analyzed in-depth PmrB proteins, althoughthe activation of Salmonella PmrB has been linked to thebinding of cations to a periplasmic domain (27). Alternatively,activation of the membrane sensors could result from outermembrane perturbations and not involve a direct interactionwith APs. Considering that all APs disorganize the outer mem-brane (62), it would be logical that Gram-negative pathogenshave evolved means to detect membrane integrity through theuse of membrane-located proteins. In support of this model, ithas been shown recently that different APs activate the Rcssystem through increased accessibility of the RcsF protein tothe inner membrane or periplasm (18). It is possible that APsmay activate each signaling system in a different way, which willgive Gram-negative bacteria the opportunity to integrate dif-ferent signals to promote phenotypic changes leading to APresistance. This could also explain why the differences betweenthe wild type and the single mutants are not dramatic, since theresponse is mediated by a collective effort of at least threesignaling systems.

We demonstrate here for the first time, using a pneumoniamouse model, that lipid A modifications are important forbacterial survival in the lung. Previously, we demonstrated thatCPS mediates resistance against APs (10, 43), and it is knownthat CPS is an important Klebsiella virulence factor (13). Col-lectively, it can be postulated that there is a correlation be-tween resistance to APs and the ability to cause pneumonia.There are several APs in the airway liquid, including lysozyme,lactoferrin, �-defensins, �-defensins, and cathelicidins (1).Therefore, the mutants’ susceptibility to APs could explain thedecreased bacterial loads of these strains in airways. However,the in vivo scenario is complex, and the final outcome of pneu-monia is a combination of the action of antimicrobial factors(among others complement and APs) and several types ofcells, including alveolar macrophages, epithelial cells, and neu-trophils. It should be noted that cytokines and chemokinesreleased by epithelial cells do upregulate the expression of APsand also increase the bactericidal activity of professionalphagocytes. Studies are ongoing to determine whether lipid Amodifications play any role in the interplay of K. pneumoniaewith airway cells.

Polymyxins are considered the “last hope” for treating in-fections caused by multidrug-resistant Gram-negative bacteriaincluding K. pneumoniae. Our results demonstrate that expo-sure to PxB induces cross-resistance not only to PxB itself but

FIG. 10. Role of K. pneumoniae Rcs, PhoPQ, and PmrAB systemsin bacterial susceptibility to PxB. Wild-type (Kp52145), 52145-�rcsB(rcsB), 52145-�phoQGB (phoQ), 52145-�pmrAB (pmrAB), 52145-�phoQGB-�pmrAB (phoQ-pmrAB), and 52145-�rcsB-�phoQGB(rcsB-phoQ) strains were exposed to different concentrations of PxB.Each point represents the mean and standard deviation of eight sam-ples from four independently grown batches of bacteria.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

also to APs present in the airways. Interestingly, the resultsfrom our laboratory indicate that K. pneumoniae prevents theexpression of APs by airway cells (48). Collectively, these find-ings are consistent with a scenario in which the setting ofpneumonia by K. pneumoniae will be facilitated by, on the onehand, preventing the expression of APs and, on the other hand,activating countermeasures against them. In turn, we put for-ward the idea that therapeutic strategies directed to preventthe activation of this program could be a new approach worthexploring to facilitate the clearance of the pathogen from theairways.

ACKNOWLEDGMENTS

We are grateful to members of Bengoechea lab for helpful discus-sions and to Christian Frank for critically reading the manuscript.

This study has been funded by grants from the Fondo de Investi-gacion Sanitaria (PI06/1629) and the Biomedicine Program (SAF2009-07885) from the Ministerio de Ciencia e Innovacion (Spain) to J.A.B.CIBERES is an initiative from Instituto de Salud Carlos III.

REFERENCES

1. Agerberth, B., et al. 1999. Antibacterial components in bronchoalveolarlavage fluid from healthy individuals and sarcoidosis patients. Am. J. Respir.Crit. Care Med. 160:283–290.

2. Appleyard, R. K. 1954. Segregation of new lysogenic types during growth ofdoubly lysogenic strain derived from Escherichia coli K-12. Genetics 39:440–452.

3. Aussel, L., et al. 2000. Novel variation of lipid A structures in strains ofdifferent Yersinia species. FEBS Lett. 465:87–92.

4. Bader, M. W., et al. 2003. Regulation of Salmonella typhimurium virulencegene expression by cationic antimicrobial peptides. Mol. Microbiol. 50:219–230.

5. Bader, M. W., et al. 2005. Recognition of antimicrobial peptides by a bac-terial sensor kinase. Cell 122:461–472.

6. Behlau, I., and S. I. Miller. 1993. A PhoP-repressed gene promotes Salmo-nella typhimurium invasion of epithelial cells. J. Bacteriol. 175:4475–4484.

7. Bitter, T., and H. M. Muir. 1962. A modified uronic acid carbazole reaction.Anal. Biochem. 4:330–334.

8. Brogden, K. A. 2005. Antimicrobial peptides: pore formers or metabolicinhibitors in bacteria? Nat. Rev. Microbiol. 3:238–250.

9. Byrappa, S., D. K. Gavin, and K. C. Gupta. 1995. A highly efficient proce-dure for site-specific mutagenesis of full-length plasmids using Vent DNApolymerase. Genome Res. 5:404–407.

10. Campos, M. A., et al. 2004. Capsule polysaccharide mediates bacterial resis-tance to antimicrobial peptides. Infect. Immun. 72:7107–7114.

11. Clements, A., et al. 2007. Secondary acylation of Klebsiella pneumoniaelipopolysaccharide contributes to sensitivity to antibacterial peptides. J. Biol.Chem. 282:15569–15577.

12. Coats, S. R., et al. 2009. Human Toll-like receptor 4 responses to Porphy-romonas gingivalis are regulated by lipid A 1- and 4�-phosphatase activities.Cell Microbiol. 11:1587–1599.

13. Cortes, G., et al. 2002. Molecular analysis of the contribution of the capsularpolysaccharide and the lipopolysaccharide O side chain to the virulence ofKlebsiella pneumoniae in a murine model of pneumonia. Infect. Immun.70:2583–2590.

14. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromo-somal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad.Sci. U. S. A. 97:6640–6645.

15. Detweiler, C. S., D. M. Monack, I. E. Brodsky, H. Mathew, and S. Falkow.2003. virK, somA, and rcsC are important for systemic Salmonella entericaserovar Typhimurium infection and cationic peptide resistance. Mol. Micro-biol. 48:385–400.

16. El, H. A., A. Tirsoaga, A. Novikov, A. Hussein, and M. Caroff. 2005. Micro-extraction of bacterial lipid A: easy and rapid method for mass spectrometriccharacterization. J. Lipid Res. 46:1773–1778.

17. Evans, M. E., D. J. Feola, and R. P. Rapp. 1999. Polymyxin B sulfate andcolistin: old antibiotics for emerging multiresistant gram-negative bacteria.Ann. Pharmacother. 33:960–967.

18. Farris, C., S. Sanowar, M. W. Bader, R. Pfuetzner, and S. I. Miller. 2010.Antimicrobial peptides activate the Rcs regulon through the outer mem-brane lipoprotein RcsF. J. Bacteriol. 192:4894–4903.

19. Favre, D., and J. F. Viret. 2000. Gene replacement in gram-negative bacteria:the pMAKSAC vectors. Biotechniques 28:198–200.

20. Fields, P. I., E. A. Groisman, and F. Heffron. 1989. A Salmonella locus thatcontrols resistance to microbicidal proteins from phagocytic cells. Science243:1059–1062.

21. Fields, P. I., R. V. Swanson, C. G. Haidaris, and F. Heffron. 1986. Mutantsof Salmonella typhimurium that cannot survive within the macrophage areavirulent. Proc. Natl. Acad. Sci. U. S. A. 83:5189–5193.

22. Ganz, T. 1987. Extracellular release of antimicrobial defensins by humanpolymorphonuclear leukocytes. Infect. Immun. 55:568–571.

23. Gibbons, H. S., S. Lin, R. J. Cotter, and C. R. Raetz. 2000. Oxygen require-ment for the biosynthesis of the S-2-hydroxymyristate moiety in Salmonellatyphimurium lipid A: function of LpxO, a new Fe2/�-ketoglutarate-depen-dent dioxygenase homologue. J. Biol. Chem. 275:32940–32949.

24. Gootz, T. D. 2010. The global problem of antibiotic resistance. Crit. Rev.Immunol. 30:79–93.

25. Groisman, E. A. 2001. The pleiotropic two-component regulatory systemPhoP-PhoQ. J. Bacteriol. 183:1835–1842.

26. Groisman, E. A., C. Parralopez, M. Salcedo, C. J. Lipps, and F. Heffron.1992. Resistance to host antimicrobial peptides is necessary for Salmonellavirulence. Proc. Natl. Acad. Sci. U. S. A. 89:11939–11943.

27. Gunn, J. S. 2008. The Salmonella PmrAB regulon: lipopolysaccharide mod-ifications, antimicrobial peptide resistance and more. Trends Microbiol. 16:284–290.

28. Gunn, J. S., et al. 1998. PmrA-PmrB-regulated genes necessary for 4-ami-noarabinose lipid A modification and polymyxin resistance. Mol. Microbiol.27:1171–1182.

29. Gunn, J. S., and S. I. Miller. 1996. Pho-PhoQ activates transcription ofpmrAB, encoding a two-component regulatory system involved in Salmonellatyphimurium antimicrobial peptide resistance. J. Bacteriol. 178:6857–6864.

30. Gunn, J. S., S. S. Ryan, J. C. Van Velkinburgh, R. K. Ernst, and S. I. Miller.2000. Genetic and functional analysis of a PmrA-PmrB-regulated locus nec-essary for lipopolysaccharide modification, antimicrobial peptide resistance,and oral virulence of Salmonella enterica serovar Typhimurium. Infect. Im-mun. 68:6139–6146.

31. Guo, L., et al. 1998. Lipid A acylation and bacterial resistance againstvertebrate antimicrobial peptides. Cell 95:189–198.

32. Hancock, R. E. W., and D. S. Chapple. 1999. Peptide antibiotics. Antimicrob.Agents Chemother. 43:1317–1323.

33. Harder, J., J. Bartels, E. Christophers, and J. M. Schroder. 2001. Isolationand characterization of human beta-defensin-3, a novel human induciblepeptide antibiotic. J. Biol. Chem. 276:5707–5713.

34. Harder, J., et al. 2000. Mucoid Pseudomonas aeruginosa, TNF-�, and IL-1�,but not IL-6, induce human beta-defensin-2 in respiratory epithelia. Am. J.Respir. Cell Mol. Biol. 22:714–721.

35. Hermsen, E. D., C. J. Sullivan, and J. C. Rotschafer. 2003. Polymyxins:pharmacology, pharmacokinetics, pharmacodynamics, and clinical applica-tions. Infect. Dis. Clin. N. Am. 17:545–562.

36. Hiratsuka, T., et al. 1998. Identification of human beta-defensin-2 in respi-ratory tract and plasma and its increase in bacterial pneumonia. Biochem.Biophys. Res. Commun. 249:943–947.

37. Hoang, T. T., R. R. Karkhoff-Schweizer, A. J. Kutchma, and H. P. Schweizer.1998. A broad-host-range Flp-FRT recombination system for site-specificexcision of chromosomally located DNA sequences: application for isolationof unmarked Pseudomonas aeruginosa mutants. Gene 212:77–86.

38. Karbarz, M. J., S. R. Kalb, R. J. Cotter, and C. R. Raetz. 2003. Expressioncloning and biochemical characterization of a Rhizobium leguminosarumlipid A 1-phosphatase. J. Biol. Chem. 278:39269–39279.

39. Kox, L. F., M. M. Wosten, and E. A. Groisman. 2000. A small protein thatmediates the activation of a two-component system by another two-compo-nent system. EMBO J. 19:1861–1872.

40. Lee, H., F. F. Hsu, J. Turk, and E. A. Groisman. 2004. The PmrA-regulatedpmrC gene mediates phosphoethanolamine modification of lipid A and poly-myxin resistance in Salmonella enterica. J. Bacteriol. 186:4124–4133.

41. Lindner, B. 2000. Matrix-assisted laser desorption/ionization time-of-flightmass spectrometry of lipopolysaccharides. Methods Mol. Biol. 145:311–325.

42. Link, A. J., D. Phillips, and G. M. Church. 1997. Methods for generatingprecise deletions and insertions in the genome of wild-type Escherichia coli:application to open reading frame characterization. J. Bacteriol. 179:6228–6237.

43. Llobet, E., J. M. Tomas, and J. A. Bengoechea. 2008. Capsule polysaccharideis a bacterial decoy for antimicrobial peptides. Microbiology 154:3877–3886.

44. Majdalani, N., and S. Gottesman. 2005. The Rcs phosphorelay: a complexsignal transduction system. Annu. Rev. Microbiol. 59:379–405.

45. Matsuzaki, K. 1998. Magainins as paradigm for the mode of action of poreforming polypeptides. Biochim. Biophys. Acta 1376:391–400.

46. McCray, P. B., Jr., and L. Bentley. 1997. Human airway epithelia express abeta-defensin. Am. J. Respir. Cell Mol. Biol. 16:343–349.

47. Mitrophanov, A. Y., M. W. Jewett, T. J. Hadley, and E. A. Groisman. 2008.Evolution and dynamics of regulatory architectures controlling polymyxin Bresistance in enteric bacteria. PLoS Genet. 4:e1000233.

48. Moranta, D., et al. 2010. Klebsiella pneumoniae capsule polysaccharide im-pedes the expression of �-defensins by airway epithelial cells. Infect. Immun.78:1135–1146.

49. Nassif, X., J. M. Fournier, J. Arondel, and P. J. Sansonetti. 1989. Mucoidphenotype of Klebsiella pneumoniae is a plasmid-encoded virulence factor.Infect. Immun. 57:546–552.



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

50. Nicolas, P., and A. Mor. 1995. Peptides as weapons against microorganismsin the chemical defense system of vertebrates. Annu. Rev. Microbiol. 49:277–304.

51. Nizet, V. 2006. Antimicrobial peptide resistance mechanisms of human bac-terial pathogens. Curr. Issues Mol. Biol. 8:11–26.

52. Oertelt, C., B. Lindner, M. Skurnik, and O. Holst. 2001. Isolation andstructural characterization of an R-form lipopolysaccharide from Yersiniaenterocolitica serotype O:8. Eur. J. Biochem. 268:554–564.

53. Perez-Gutierrez, C., E. Llobet, C. M. Llompart, M. Reines, and J. A. Ben-goechea. 2010. Role of lipid A acylation in Yersinia enterocolitica virulence.Infect. Immun. 78:2768–2781.

54. Prost, L. R., M. E. Daley, M. W. Bader, R. E. Klevit, and S. I. Miller. 2008.The PhoQ histidine kinases of Salmonella and Pseudomonas spp. are struc-turally and functionally different: evidence that pH and antimicrobial peptidesensing contribute to mammalian pathogenesis. Mol. Microbiol. 69:503–519.

55. Prost, L. R., et al. 2007. Activation of the bacterial sensor kinase PhoQ byacidic pH. Mol. Cell 26:165–174.

56. Raetz, C. R., C. M. Reynolds, M. S. Trent, and R. E. Bishop. 2007. Lipid Amodification systems in gram-negative bacteria. Annu. Rev. Biochem. 76:295–329.

57. Rahn, A., and C. Whitfield. 2003. Transcriptional organization and regula-tion of the Escherichia coli K30 group 1 capsule biosynthesis (cps) genecluster. Mol. Microbiol. 47:1045–1060.

58. Rebeil, R., R. K. Ernst, B. B. Gowen, S. I. Miller, and B. J. Hinnebusch. 2004.Variation in lipid A structure in the pathogenic yersiniae. Mol. Microbiol.52:1363–1373.

59. Singh, P. K., et al. 1998. Production of beta-defensins by human airwayepithelia. Proc. Natl. Acad. Sci. U. S. A. 95:14961–14966.

60. Soncini, F. C., V. E. Garcia, F. Solomon, and E. A. Groisman. 1996.Molecular basis of the magnesium deprivation response in Salmonellatyphimurium: identification of PhoP-regulated genes. J. Bacteriol. 178:5092–5099.

61. Tran, A. X., et al. 2006. The lipid A 1-phosphatase of Helicobacter pylori isrequired for resistance to the antimicrobial peptide polymyxin. J. Bacteriol.188:4531–4541.

62. Vaara, M. 1992. Agents that increase the permeability of the outer mem-brane. Microbiol. Rev. 56:395–411.

63. Vaara, M. 2010. Polymyxins and their novel derivatives. Curr. Opin. Micro-biol. 13:574–581.

64. Wang, X., S. C. McGrath, R. J. Cotter, and C. R. Raetz. 2006. Expressioncloning and periplasmic orientation of the Francisella novicida lipid A 4�-phosphatase LpxF. J. Biol. Chem. 281:9321–9330.

65. Wehland, M., and F. Bernhard. 2000. The RcsAB box. Characterization ofa new operator essential for the regulation of exopolysaccharide biosynthesisin enteric bacteria. J. Biol. Chem. 275:7013–7020.

Editor: F. C. Fang



http://iai.asm.org/

Dow

nloaded from

http://iai.asm.org/

Documents

Analysis of the Networks Controlling the Antimicrobial-Peptide … · also be effective against host APs, thus contributing to bacterial resistance and survival in host tissues. To