Homologous response regulators KvgA, KvbA and KvhR regulate the synthesis of capsular polysaccharide in Klebsiella pneumoniae CG43 in a coordinated manner

(1)

Homologous Response Regulators KvgA, KvhA and KvhR Regulate

the Synthesis of Capsular Polysaccharide in Klebsiella pneumoniae

CG43 in a Coordinated Manner

Ching-Ting Lin, Teng-Yi Huang, Wan-Chun Liang and Hwei-Ling Peng*

Department of Biological Science and Technology, National Chiao Tung University, 75 Po-Ai Street, Hsin Chu 30050, Taiwan, Republic of China

Received May 25, 2006; accepted July 24, 2006

On the basis of phenotypic analysis, the Klebsiella pneumoniae CG43 derived mutants with deletions of the gene encoding respectively the response regulators KvgA, KvhA,

and KvhR were classified into two groups. Group I bacteria carrying either kvgA-or

kvhR-_{exhibited less mucoidy, lower level of capsular polysaccharide (CPS) synthesis}

and higher LD50than the parental strain. No apparent change of the group II, including

kvhA-and kvhA-kvhR-mutants, was observed. However, the mucoidy of kvhA-kvhR

-mutant was found to be diminished after introducing into a kvhA-expressing plasmid.

Via promoter-lacZ fusion analysis, kvhA deletion was found to reduce kvhR expression. A regulatory role of KvhA for the expression of kvhR was supported further by EMSA showing a specific binding of KvhA to the putative promoter of kvhR. The promoter activity measurement and EMSA also revealed that KvgA acted as an autoregulator and an activator for the expression of kvhAS and kvhR. In addition, deletion of kvgA suppressed slightly the promoter activity of the cps-orf16-17, and the expression of

all three cps transcripts orf1-2, orf3-15, and orf16-17 were reduced in the kvhR-_mutant.

These suggest that the three homologous response regulators interact to control, in coordination, the bacterial cps expression.

Key words: cps expression, homologous response regulator, Klebsiella pneumoniae CG43, KvgA, KvhA, KvhR.

Bacterial two-component systems (2CSs), consisting of a sensor histidine kinase and a response regulator, recognize specific signals and convert this information into specific transcriptional or behavioral responses (1). Upon sensing the input signals, the sensor protein catalyzes an autopho-sphorylation reaction in transmitter domain, which trans-fers a phosphate from ATP to a conserved histidine residue. The phosphate group is subsequently transferred from the histidine residue to a specific aspartate residue on the receiver domain of the cognate response regulator. The phosphorylated response regulator begins to perform an appropriate regulatory function, possibly by conforma-tional change (2, 3). The number of 2CS varies dramati-cally among bacterial genomes. For instance, Bacillus subtilis encodes 70 2CS proteins (4), whereas Helicobacter pylori and Haemophilus influenzae contained only 11 and nine 2CS protein–encoding genes, respectively (2). As is widely believed, the 2CS proteins function as components of a signal transduction network, enabling bacteria to respond to complex environmental stimuli. Indeed, the presence of 2CS regulatory circuits in various bacteria has been recently acknowledged (5, 6).

As a common nosocomial pathogen, Klebsiella pneumo-niae causes suppurative lesions, septicemia, and urinary and respiratory tract infections in immunocompromised patients (7–10). The increasing prevalence of extended

spectrum b-lactamase, which produces K. pneumoniae (ESBLKp), prompted the search for new drugs to intervene in the bacterial infections (11). Most recently, the develop-ment of an antimicrobial drug that targets bacterial 2CSs has been evaluated (12). In an earlier study, we have iso-lated 2CS genes by PCR-supported genomic subtractive hybridization from K. pneumoniae CG43, a highly virulent clinical isolate of K2 serotype (13). On the basis of the sequence similarity to that of the Bordetella pertussis BvgAS, which plays an important role in pathogenesis (14), the bvgAS-like genes were named kvgAS (15). The analysis using dot-blotting hybridization revealed that kvgAS is present in approximately 15% of the laboratory collected clinical isolates, suggesting an accessory role of the 2CS in the bacterial pathogenesis (15). Downstream to kvgAS, a gene encoding KvgA homolog (53.8% similarity), namely kvhR, was later identified (16). Interestingly, BLASTP analysis was unable to identify either kvgAS or kvhR in the K. pneumoniae MGH78578 genome

(http://genome.wustl.edu/projects/bacterial/). In stead,

kvgAS-homologous genes were found and the genes subse-quently isolated from K. pneumoniae CG43 and designated kvhAS as kvgAS homolog. In contrast to kvgAS, kvhAS is present in all the strains collected in the laboratory, as analyzed by dot-blotting hybridization using kvhA as a probe (data not shown), suggesting a important role of the 2CS in K. pneumoniae. A stress-responsive role of KvgAS has been proposed since kvgAS expression was acti-vated in LB medium adding with 0.2% paraquat or 0.2 mM

2,20_{-dipyridyl (16). However, functional roles of KvhAS and}

*To whom correspondence should be addressed. Phone: +886 3 5712121 (Ext. 56916), Fax:+886 3 5729288, E-mail: hlpeng@cc. nctu.edu.tw

at National Chiao Tung University Library on April 26, 2014

http://jb.oxfordjournals.org/

(2)

KvhR have not yet identified. In this study, by generation of a series of the response regulator mutants and measure-ment of their promoter activities, we were able to show an interactive regulation of the 2CSs and also demonstrate their regulatory roles on cps gene expression in the bacteria.

MATERIALS AND METHODS

Bacterial Strains, Plasmids, and Growth Conditions— Bacterial strains and plasmids used in this study are listed in Table 1. All bacterial strains were routinely cultured

at 37_{C in Luria-Bertani (LB) medium or M9 minimal}

medium supplemented with appropriate antibiotics. Sequence Analysis—Approximately 2 kb and 3.3 kb DNA, located upstream respectively to kvgA and kvhR, were subjected to sequence determination. The sequence analysis including Open reading frames (ORFs) identifica-tion and annotaidentifica-tion were carried out using BLAST (NCBI database). The presence of tRNA sequence was identified by the program tRNAscan-SE (17). The G+C content analysis was performed by the program of GEECEE in EMOBSS.

Construction of a LacZ Reporter System—In order to assess each of the promoter activities, a promoter-trap vec-tor with LacZ as the reporter, and CG43S3-Z01, derived from K. pneumoniae CG43S3 (18) with a deletion of lacZ gene, were constructed. Briefly, a promoterless lacZ gene was PCR amplified from K. pneumoniae CG43S3 with the primer set lac01/lac02 (Table 2) and then inserted into the promoter-trap vector pYC016 (18). The resulting reporter plasmid was designated as placZ15 (Table 1). In addition, two 1 kb DNA fragments flanking the lacZ gene were PCR amplified using specific primer sets lac05/lac06 and lac03/lac07 (Table 2). The generated DNA fragments were ligated and subcloned into a suicide vector pKAS46 (19). The resulting plasmid placZ16 was transformed into Escherichia coli S17-1lpir and then mobilized to the streptomycin-resistant strain K. pneumoniae CG43S3 by conjugation. A kanamycin resistant transconjugant was initially picked, grown overnight, and then spread onto a

LB plate supplemented with 500 mg ml-1 streptomycin.

After the occurrence of double crossover, the streptomycin resistant colonies were further ascertained for their sus-ceptibility to kanamycin. The lacZ mutation was confirmed by plating the bacteria onto a X-gal containing medium and by Southern hybridization (data not shown), and the mutant was designated as K. pneumoniae CG43S3-Z01 (Table 1).

Construction of kvgA, kvhA, and kvhR Deletion Mutants—The mutants with specific deletion of either of kvgA, kvhA, and kvhR genes were also constructed by the allelic exchange strategy described above. The primer sets used for the construction of the deletions are listed in Table 2. The gene-specific deletion mutants derived from K. pneumoniae CG43S3-Z01 were generated through homologous recombination and the resulting strains

were designated AZ18 (kvgA-), AhZ01 (kvhA-) and RZ01

(kvhR-). For the construction of kvhR-kvgA- or kvhR

-kvhA-double mutant, the pKAS46 derivative containing

either a kvgA or kvhA deletion was delivered respectively from E. coli S17-1 lpir into RZ01 by conjugation. The plas-mid carrying a kvhA deletion was also mobilized from

E. coli S17-1 lpir to K. pneumoniae AZ18 by conjugation

to generate kvgA-kvhA-. For kvhR-kvgA-kvhA- triple

mutant, the pKAS46 derivative containing kvgA deletion

was delivered from E. coli S17–1 lpir into kvhA-kvhR

-mutant by conjugation. The selections for the -mutants were carried out likewise. The resulting mutants were

designated as AAh01 (kvgA-kvhA-), AR01 (kvgA-kvhR-),

AhR01 (kvhA-kvhR-) and AAhR01 (kvgA-kvhA-kvhR-).

The mutant with deletion of rcsB, which has been demon-strated to encode a K2 cps activator in K. pneumoniae CG43S3 (20), was also generated and named RcsBZ01

(rcsB-).

Determination of Promoter Activity—The putative pro-moter regions of kvgAS, kvhAS, kvhR, rcsB, and the three cps transcriptional units (21, 22) were PCR amplified from K. pneumoniae CG43S3 by the designed primer sets (Table 2) and subcloned into placZ15 to fuse them with the lacZ reporter gene. One-tenth overnight culture of the bacteria carrying each of the plasmids were refreshly grown in M9 medium to an optical density at wavelength of

600 nm (OD600) about 0.6 to 0.7. The b-galactosidase

activ-ity assay was carried out essentially as described by Miller (23). The data presented were derived from a single experiment which is representative of at least three inde-pendent experiments. Every sample was assayed in tripli-cate, and the average activity and standard deviation were presented.

Preparation of the Recombinant KvgAt, KvhA, and KvhRt—The coding region of kvgA, kvhA, and kvhR were PCR amplified from K. pneumoniae CG43S3 with the specific primers (Table 2), and the PCR products cloned into pUC-T vector (MDBio). The resulting plasmids were designated as pkvgA1, pHP4004, and pR28, respectively. The plasmid pHP4004 was digested with BamHI and the entire kvhA fragment subcloned into pET30c, and the resultant plasmid was named as pHP4005. While overex-pression of either kvgA or kvhR resulted in largely inso-luble proteins. In order to resolve the problem, the plasmid pkvgA1 was digested with ClaI to remove the receiver domain. The remaining DNA binding domain of

approxi-mately 200-bp, KvgAt, was subcloned into the SalI–NotI

sites of pET30c, which resulted in the expression plasmid pkvgA4. Likewise, pR28 was digested with EcoRV and HindIII to remove the receiver domain and the

remain-ing DNA bindremain-ing domain, KvhRt, was subcloned into the

EcoRV–HindIII sites of pET30a, which resulted in the

expression plasmid pR31. The plasmids pkvgA4,

pHP4005, and pR31 were then transformed into E. coli BL21-RIL. The transformants carrying either pkvgA4, pHP4005, or pR31 were cultured in LB medium to log phase, and expression of either the recombinant

His-KvgAt, His-KvhA, or His-KvhRt protein was induced

with 1 mM IPTG for 3 h at 37_{C. The overexpressed}

His-KvgAt and His-KvhRt protein formed an inclusion

body, but the His-KvhA appeared to be in soluble form. The bacteria carrying pkvgA4 and pR31 respectively were lysed by sonication and the pellet was resuspended

and denatured with 6Nurea. After purification by affinity

chromatography with His-Bind resin (Novagen), the

dena-tured His-KvgAt and His-KvhRt protein were refolded

respectively through dialysis against a gradient of decreas-ing concentrations of urea in the reaction buffer (20 mM

Tris-HCl, pH 8.0, 4 mM MgCl2, 50 mM KCl, 1 mM CaCl2

(3)

Table 1. Bacterial strains and plasmids used in this study.

Strains or plasmids Descriptions Reference or source

Strains K. pneumoniae

CG43S3 CG43 Smr (18)

CG43S3-Z01 CG43S3 DlacZ This study

CG43S3-AZ18 CG43S3-Z01 DkvgA This study

CG43S3-AhZ01 CG43S3-Z01 DkvhA This study

CG43S3-RZ01 CG43S3-Z01 DkvhR This study

CG43S3-AAh01 CG43S3-Z01 DkvgA DkvhA This study

CG43S3-AAR01 CG43S3-Z01 DkvgA DkvhR This study

CG43S3-AhR01 CG43S3-Z01 DkvhA DkvhR This study

CG43S3-AAhR01 CG43S3-Z01 DkvgA DkvhA DkvhR This study

CG43S3-RcsBZ01 CG43S3-Z01 DrcsB This study

E. coli

JM109 RecA1 supE44 endA1 hsdR17 gyrA96 relA1 thi D [lac-proAB] Laboratory stock BL21-RIL F-_{ompT hsdS}

B[rB-mB-]gal dcm [DE3] Laboratory stock

S 17-1 l pir hsdR recA pro RP4-2 [Tc::Mu; Km::Tn7] [lpir] (19)

Plasmids

pKAS46 Positive selection suicide vector, rpsL Apr_Kmr ₍₁₉₎

pET30a-c His-tagging protein expression vector, Kmr _Novagen

pUC-T TA cloning vector MDBio

pYC016 Promoter selection vector, LuxAB+Cmr (18)

placZ15 A derivative of pYC016, containing K. pneumoniae CG43S3 lacZ as a reporter, Cmr _{This study}

pETm-c A derivative of pET30C, containing malonate promoter, Kmr This study placZ16 2-kb fragment containing an internal 1.5-kb deletion in lacZ cloned into pKAS46,

AprKmr

This study pA13 2-kb fragment containing an internal 0.6-kb deletion in kvgA cloned into pKAS46,

Apr_Kmr This study

pAhm1 2-kb fragment containing an internal 0.7-kb deletion in kvhA cloned into pKAS46,

Apr Kmr This study

pR14 pKAS46 carring a DkvhR fragment This study

pYC220 2.0-kb fragment containing a 763-bp deletion in rcsB locus cloned into pKAS46 (20) pRcsB2 The DNA fragment carrying entire rcsB coding sequence cloned into the EcoRV/SalI

site of pETm-c

This study pkvgA1 A fragment of K. pneumoniae CG43S3 kvgA gene generated by PCR, and cloned

into pUC-T, Apr

This study pkvgA4 Deletion of the receiver domain of kvgA gene digested by ClaI, and cloned into

pET30c, Kmr

This study pHP4004 A fragment of K. pneumoniae CG43S3 kvhA gene generated by PCR, and cloned into

pUC-T, Apr

This study pHP4005 A fragment of K. pneumoniae CG43S3 kvhA gene digested by BamHI, and cloned into

pET30c, Kmr

This study pR28 A fragment of K. pneumoniae CG43S3 kvhR gene generated by PCR, and cloned into

pUC-T, Apr This study

pR31 Deletion of the receiver domain of kvhR gene digested by EcoRV/HindIII, and cloned

into pET30a, Kmr This study

pA16 399-bp BamHI/BglII fragment containing the putative kvgAS promoter, cloned into BamHI site of placZ15

This study pAh01 516-bp BamHI/BglII fragment containing the putative kvhAS promoter, cloned into

BamHI site of placZ15

This study pRP05 500-bp BamHI fragment containing the putative kvhR promoter, cloned into placZ15 This study pA415 A 1.3 kb EcoRI fragment containing kvhA locus with the putative promoter cloned into

pRK415

This study pAHm A BamHI fragment of pHP4005 carrying entire kvhA coding sequence cloned into the

BamHI site of pETm-c

This study pRC01 A 1.2 kb BamHI/EcoRI fragment containing kvhR locus with the putative promoter

cloned into pACYC184

This study pRC02 A 1.2 kb BamHI/EcoRI fragment containing kvhR locus with the putative promoter

cloned into pRK415

This study pOrf12 500-bp BamHI fragment containing the putative orf1-2 promoter, cloned into placZ15 This study pOrf315 900-bp BamHI fragment containing the putative orf3-15 promoter, cloned into placZ15 This study pOrf1617 300-bp BamHI fragment containing the putative orf16-17 promoter, cloned into placZ15 This study

(4)

and 1 mM dithiothreitol). The His-KvhA protein was purified from the soluble fractions of the IPTG-induced

bacteria carrying pHP4005. The purified His-KvgAt,

His-KvhA and His-KvhRt were then concentrated with

polyethylene glycol 20,000 and the concentration of protein was determined by the Bradford assay (24). Finally, mole-cular weight and purity of the proteins were analyzed by SDS–polyacrylamide gel electrophoresis.

Electrophoretic Mobility Shift Assay (EMSA)—DNA fragments comprising a series of the putative promoter regions were obtained by PCR amplification with respec-tive primer sets as described in Table 2, and then labeled

with [g-32P]ATP using T4 polynucleotide kinase. The

pur-ified His-KvgAt, His-KvhA, or His-KvhRt was incubated

with the radioactively labeled DNA in a 20 ml solution

containing 20 mM Tris-HCl (pH 8.0), 4 mM MgCl2,

50 mM KCl, 1 mM CaCl2 and 1 mM dithiothreitol at

37_{C for 20 min. Excess amount (approximately 10 times}

more than the labeled DNA) of each of the unlabeled DNA was used in the competition assay. The samples were then loaded onto a running gel of 5% nondenaturing polyacry-lamide in 0.5· TBE (45 mM Tris-HCl, pH 8.0, 45 mM boric acid, 1 mM EDTA). Gels were electrophoresed with a

20-mA current at 4_{C and detected by InstantImager}TM

(Packard Instrument Company).

Extraction and Quantification of CPS—CPS was

extracted as described previously (25). Five hundred microliters of bacteria cultured in LB broth overnight were mixed with 100 ml of 1% Zwittergent 3-14 detergent (Sigma-Aldrich) in 100 mM citric acid (pH 2.0), and the

mixture was incubated at 50_{C for 20 min. After}

centrifu-gation at 13,500 rpm for 10 min, 250 ml of the supernatant was transferred to a new tube, and the CPS was

precipi-tated with 1 ml of absolute ethanol at 4_{C for 20 min and}

then centrifuged at 13,500 rpm for 25 min. The pellet was

dried at 37_{C and dissolved in 200 ml of distilled water, and}

1.2 ml of 12.5 mM borax (Sigma-Aldrich) in H2SO4 was

added. The mixture was vigorously vortexed, boiled for 5 min, and cooled, and then 20 ml of 0.15% (v/v) 3-hydroxydiphenol (Sigma-Aldrich) was added and the absorbance at 520 nm was measured. The uronic acid con-tent was determined from a standard curve of glucuronic acid (Sigma-Aldrich) and expressed as micrograms per

109CFU (26).

Mouse Lethality Assay—Female BALB/c mice with an average age of four weeks were acclimatized in an animal house for 7 days. The tested bacterial strains were cultured

in LB medium at 37_{C overnight. Four mice of a group were}

injected intraperitoneally with bacteria suspended in

0.2 ml of saline in 10-fold steps graded doses. The LD50,

Table 2. Primers used in this study.

Primer no. Sequence Complementary position

lac01 50_{-GCGAACGACAAGATCTGACTTA-3}0 _{-24 relative to the lacZ start codon}

lac02 50_{-ATTATGCCGTTCTAGAGGCG-3}0 _{+103 relative to the lacZ stop codon}

lac03 50_{-TGAAACGCAAGGATCCGAGC-3}0 _{+1444 of the lacZ coding region}

lac05 50_{-CAGGTGGAGGAGCTCGAAAG-3}0 _{-907 relative to the lacZ start codon}

lac06 50_{-AAACGGGATCCGCTGGCA-3}0 _{+117 of the lacZ coding region}

lac07 50_{-GCAGTGCGCCTCTAGATCGT-3}0 _{+2498 of the lacZ coding region}

a02 50_{-CAATATCATAGCCAGCA-3}0 _{+45 relative to the kvgA stop codon}

a03 50_{-ATTGCTTCACTCACCCT-3}0 _{-32 relative to the kvgA start codon}

a08 50_{-GAGAGCTCGATTATTTCATCGA-3}0 _{-834 relative to the kvgA start codon}

a09 50_{-CATATTGTGGATCCTGCTGTTC-3}0 _{+22 of the kvgA coding region}

a10 50_{-CGATGCGGGATCCAATGCCTTTA-3}0 _{+296 of the kvgA coding region}

a11 50_{-AACAAGATCTAGCTTTTGAT-3}0 _{+699 relative to the kvgA stop codon}

a14 50_{’-ATTTTCAGGATCCACCACCTT-3}0 _{-409 relative to the kvgA start codon}

AS02 50_{-CAGCCATGCTTTCTCCTT-3}0 _{+156 relative to the kvhA stop codon}

AS07 50_{-ATCAGGATCCACGCCCCC-3}0 _{-18 relative to the kvhA start codon}

AS04 50_{-ATCTGCAGAATATCCCGT-3}0 _{+1532 of the kvhS coding region}

AS12 50_{-TCCTGCAATGCTGGAATT-3}0 _{-1245 relative to the kvhA start codon}

AS16 50_{-GCCCGGGTTATTTTTATC-3}0 _{-52 relative to the kvhA start codon}

AS23 50_{-CATGGCGGTTCGTCTTAT-3}0 _{-1 relative to the kvhS start codon}

A201 50_{-GTGAAAAAGCTTCGTTCA-3}0 _{-516 relative to the kvhA start codon}

A203 50_{-CAACGACAGCTCTTCCAA-3}0 _{+69 of the kvhA coding region}

R01 50_{-CTTTTTAAGCTTAAATGA-3}0 _{-469 relative to the kvhR start codon}

R02 50_{-TTCGGGTACCTCTCCATC-3}0 _{+62 relative to the kvhR start codon}

R04 50_{-AGGCCTTCAATCCCACAC-3}0 _{+23 relative to the kvhR stop codon}

R07 50_{-AGGTTAAGAGCTCCAGCGCC-3}0 _{-1097 relative to the kvhR start codon}

R09 50_{-TGGATCCGTTTGTATGAATGTA-3}0 _{+353 of the kvhR coding region}

P074 50_{-ACTGGATCCACGATCATGGATAAGAT-3}0 _{-724 relative to the orf1 start codon}

P075 50_{-ACTGGATCCTGCGACCGGAATAACC-3}0 _{+42 of the orf1 coding region}

P040 50_{-ACTGGATCCAGGCCTGGTAATAGCCATT-3}0 _{-890 relative to the orf3 start codon}

P041 50_{-ACTGGATCCCGCTGTCGTATCTCAATG-3}0 _{+60 of the orf3 coding region}

P045 50_{-GGTGCGCAGATCTATAAGC-3}0 _{-307 relative to the orf16 start codon}

P046 50_{-ACTGGATCCAGACGGAGGAACTGTTTC-3}0 _{+89 of the orf16 coding region}

(5)

based on the number of survivors after 10 days, was cal-culated and expressed as CFU as described (27).

RESULTS AND DISCUSSION

Sequence Comparisons of KvgA, KvhA, and KvhR— Increasing studies have acknowledged that, excluding sequences of closely related homologues, the transmitter domain from any two sensors typically share 20% to 50% sequence identity (average sequence identity, 25%). On the other hand, receiver domains from any two response reg-ulators share sequence identity at only 20% to 30% (28–30). Sequence analysis of the receiver domains revealed a 47.9% amino acid identity between the two response regulators KvgA and KvhA, and KvhR shares 43.8% and 46.3% amino acid identity with KvgA and KvhA respec-tively. The results, together with the high sequence iden-tity of the transmitter domain, which is 45.8%, between KvgS and KvhS strongly suggested that KvgS and KvhS are paralogous sensors, and KvgA, KvhA, and KvhR are paralogous response regulators.

We have previously shown by BLASTX sequence analy-sis that KvgAS is highly homologous to B. pertusanaly-sis BvgAS (31, 32) and Escherichia coli EvgAS (Utsumi et al., 1992). The bvg system controls the expression of major virulence factors in B. pertussis, such as filamentous haremaggulti-nin (fha), pertactin (prn), adenylate cyclase toxin (cya), and pertussis toxin (ptx) (33). While in E. coli, EvgAS has also been shown to be involved in regulating the gene expression of virulence-related property such as multi-drug resistance and acid resistance (34–36). As shown in Fig. 1A, phylogenetic analysis, on the basis of the comparison of overall amino acid sequence of the sensor and response regulator, revealed that KvgAS and BvgAS are relatively distant from the branches of KvhAS and EvgAS, that appeared to be clustered together. This

implies that KvhAS and EvgAS are most likely to be orthologous 2CS.

Sequence Analysis of the DNA Fragments That Contain kvhAS and kvgAS—Figure 1B shows a comparative ana-lysis of the genes near kvhAS with that of E. coli evgAS revealing a YfdX homologue (49% amino acid sequence identity), a hypothetical protein for acid resistance in E. coli (37). Moreover, flanking both sides of kvhAS, homolo-gues of putative acid resistance proteins HdeB and HdeD, which are positively regulated by EvgA (35), including HdeB1 (23% amino acid sequence identity), HdeB2 (38% amino acid sequence identity) and HdeD (26% amino acid sequence identity), were also identified. In analogy to the regulatory role of EvgAS, which modulates expression of the flanking genes, including putative efflux pump emrKY (36, 38, 39) and the acid-resisting gene yfdX (37, 40), we speculate that KvhAS controls expression of the nearby genes, hdeB, hdeD and yfdX. This possibility remains to be validated.

Dot-blotting hybridization using the probe of either orfX or kvhR gene, which is located downstream of the kvgAS operon, shows that only about 70% of the kvgAS-carrying isolates also harbored the orfX and kvhR genes (data not shown), suggesting that the kvgAS operon and kvhR were not acquired concurrently. Subsequently, 3.3 kb DNA upstream of kvhR and 2 kb DNA upstream of kvgA were sequenced (the sequences deposited in the GenBank database under accession number AJ250891) and the sequences analyzed to confirm whether mobile elements are present. As shown in Fig. 1B, only the sequences 3 kb beyond kvhR could be identified as the counterpart in the K. pneumoniae MGH78578 genome. Analysis of the sequences upstream of kvhR identifies an ORF, namely orfY, encoding a putative exported lipase and a partial sequence of IS911. Intriguingly, the 177-bp intergenic sequence between kvgS and orfX, 640-bp intergenic

Fig. 1. Evolutionary relation-ship of KvhAS, KvgAS, and KvhAS gene clusters. (A) The phylogenetic tree of KvgAS, KvhAS, EvgAS, and BvgAS. The Neighbor-Joining tree was built by CLUSTAL W 1.81 (52) with the deduced amino acid sequences. The resultant tree was visualized by MEGA2 (53). The black bar represents 10% sequence divergence. (B) Com-parison of the respective gene clusters flanking kvhAS, kvgAS, and evgAS. The respective ORFs flanking kvgAS in K. niae CG43, kvhAS in K. pneumo-niae MGH78578 and evgAS in E. coli K12 are shown. The amino acid identities are indicated.

(6)

sequence between orfX and kvhR, and 2 kb sequences upstream of kvgAS, revealed neither ORF nor mobile element. The G+C content of the 12 kb DNA, containing kvgAS-orfX-kvhR and the flanking sequences, was 43%, which is somewhat lower than that of the K. pneumoniae MGH78578 genome (55%). The lower G+C content of the DNA fragment, which can be identified only in some of the clinical isolates (15), implied that the gene cluster had been recently acquired by horizontal transfer.

Phenotype Analysis of the Mutants kvgA-, kvhA-,

kvhR-_, _kvgA-_kvhA-_, _kvgA-_kvhR-_, _kvhA-_kvhR- _and

kvgA-kvhA- kvhR-—The mutants, including AZ18

(kvgA-), AhZ01 (kvhA-), RZ01 (kvhR-), AAh01

(kvgA-kvhA-), AR01 (kvgA-kvhR-), AhR01 (kvhA-kvhR-)

and AAhR01 (kvgA-kvhA-kvhR-) displayed a relatively

large, glistening colony on LB agar. The morphology was indistinguishable from that of the wild-type strain. Never-theless, a reduction in the mucoid characteristics was noted when the bacteria cultures were subjected to low-speed centrifugation. The sedimentation test to assess bac-terial mucoidy allowed these mutants to be classified into two groups. Group I bacteria, carrying either kvgA or kvhR mutation, exhibited faster precipitation than that of the

parental strain Z01. Group II bacteria, including kvhA

-and kvhA-kvhR- mutants, exhibit precipitation that is

similar to that exhibited by the parental strain

Z01 (Fig. 2A). As determined by the string test (41), the viscous colony nature of the group I bacteria appeared to be considerably diminished suggesting a reduction of the CPS (Fig. 2A). It is of interest to note that the

kvhA-kvhR-mutant of group II exhibited a less mucoidy

than either wild type or kvhA-mutant of the same group.

While the kvhA-kvhR-mutant supplied with the plasmid

pRC01, containing a kvhR locus, exerted no effect on the

bacterial phenotype indicating that the deleting effect of kvhR was suppressed by kvhA deletion. On the other hand,

transformation of kvhA-kvhR- with the plasmid pA415

carrying a kvhA locus, converted the phenotype from group II to group I (Fig. 2B). This suggests an upstream regulation of KvhA for a proper expression of kvhR.

Promoter Activity Measurements of kvgAS, kvhR and kvhAS—The interacting regulation of 2CS network has been reported, which showed that some of the sensor pro-teins can conditionally transfer the phosphoryl molecules to non-cognate response regulators as well as to their cog-nate regulators (3). The possibility that if the signal is relayed from KvgS or KvhS to the orphan response regu-lator KvhR, remains to be examined. Nevertheless, the promoter-lacZ fusion constructs of kvgAS, kvhAS, and kvhR were generated to investigate whether the three homologous regulators regulate each other. The

b-galacto-sidase activity of PkvgAS (pA16) measured in wild-type

(Z01), kvgA mutant (AZ18), kvhA mutant (AhZ01) and kvhR mutant (RZ01) was found to be higher in M9 minimal medium than in LB (data not shown). Hence, the bacteria were grown in M9 minimal medium to enable the promoter activity to be measured. Table 3 shows that the activity of PkvgAS-pA16, which contains a 399-bp noncoding region of

the kvgA start codon, in the kvgA deletion mutant AZ18, was approximately 50% that of Z01, indicating a positive auto-regulatory role of KvgA. The activity of pA16 measured in AhZ01 and RZ01 were similar, revealing that neither kvhA nor kvhR deletion affected the

expres-sion of kvgAS. Interestingly, the activity of PkvhAS(pAh01)

and PkvhR (pRP05), contained respectively a 500-bp

non-coding region upstream of the start codon of kvhA and kvhR, were found to be lower in the kvgA mutant AZ18, suggesting that KvgA is probably an activator for the

Fig. 2. Comparison of precipitation speed of the mutants derived from K. pneumoniae CG43S3-Z01. (A) The strains tested were cultured overnight in LB broth at 37_{C and subjected}

to centrifugation at 4,000 rpm (1,500· g) for 3 min. According to collective analysis of both assays, two groups can be identified, in which the group I includes: AZ18 (kvgA-_{), RZ01 (kvhR}-_{), AAh01}

(kvgA-kvhA-), AR01 (kvgA-kvhR-), AAhR01 (kvgA-kvhA-kvhR-), and the group II includes: Z01, AhZ01 (kvhA-_), _AhR01

(kvhA-_kvhR-_{). Bacterial mucoidy was assessed by string formation}

test after the bacteria grown on LB plates for 48 h. Symbols:-, negative;+, positive; ++, strong. (B) Effect of the complementation of AhR01 (kvhA-kvhR-) with either an intact kvhA or kvhR. The precipitation speed of the strains was also shown. AhR01 [pA415] was indicated AhR01 (kvhA-_kvhR-_{) with an intact kvhA. AhR01}

[pRC01] was indicated AhR01 (kvhA-kvhR-) with an intact kvhR.

(7)

expression of kvhAS and kvhR. Although the deletion of

kvhA or kvhR did not affect the expression of PkvhAS, both

mutations appeared to reduce PkvhRactivity, implying that

KvhA positively regulates the kvhR expression and KvhR is an auto-regulator of its own expression. The finding is consistent with the above-mentioned notion that KvhA is likely an upstream regulator for kvhR expression. The BPROM program (http://www.softberry.com) used to

ana-lyze the sequences of the PkvgAS, PkvhAS, and PkvhRdid not

identify any cis-element, indicating that more studies are required toward understanding the regulatory

mechan-isms for the expression of PkvgAS, PkvhAS, and PkvhRin K.

pneumoniae CG43.

EMSA—Subsequently, EMSA was performed using

pur-ified KvgAt protein and DNA fragments that contained

PkvgAS, PkvhAS and PkvhR, to verify that KvgA, as a

tran-scriptional activator, indeed binds directly to either of its

own promoter PkvgAS, PkvhASand PkvhR. Figure 3A shows

that KvgAt which comprises the DNA binding domain

could bind to it own promoter and that the DNA-protein

interaction was specific, as the formation of the His6

-KvgAt-promoter complex could only be inhibited by the

presence of the unlabelled DNA. Furthermore, the two binding complexes, C1 and C2, observed when the amount

of His6-KvgAtwas increased from 0.3 mg to 0.6 mg to bind

PkvgA. This could indicate a higher order complex of the

protein either to the same site or to distinct sites. The His6

-KvgAtwas shown also to bind PkvhASDNA and PkvhR, and

the bindings were demonstrated to be specific since the bindings could only be inhibited by the unlabelled specific DNA (Fig. 3, B and C). The results verified that KvgA

positively regulated the expressions of PkvhAS and PkvhR

by direct binding. The assay further established that

His6-KvhA bound specifically to the [g-32P]ATP-labeled

PkvhR and that the DNA-protein complex could only be

competed in the presence of an excess of unlabelled

PkvhR(Fig. 3D). Finally, as shown in Fig. 3E, specific

bind-ing of KvhR to the DNA fragment PkvhR, containing its own

putative promoter, was also demonstrated. Deletion of kvgA or kvhR Affect the

CPS Expression—An extremely thick CPS is characteristic of the genus Klebsiella, which provides the bacteria a glis-tening and mucoid phenotype. Diminished mucoidy of the group I bacteria could be attributed to the reduction of their CPS. The amount of CPS produced in these mutants was determined by measuring the glucuronic acid content, an indicator of Klebsiella K2 CPS (42). Like E. coli group I CPS biosynthesis, Klebsiella K2 cps expression is regulated by the 2CS RcsAB at the transcriptional level (43, 44).

A CG43S3Z01-derived rcsB- mutant was therefore

con-structed and the CPS content was also determined and compared. Table 4 reveals that the group I bacteria as

well as the rcsB-mutant, synthesized less CPS than the

wild-type strain, respectively from 0.51- to 0.68-fold of that of wild type, suggesting a positive regulation by KvgA and KvhR on cps expression. In the mouse peritonitis model,

the deletion of either kvgA or kvhR increased LD50 by a

factor of 90 to 100 (Table 4). It is most likely that the

Table 3. Effect of kvgA, kvhA, and kvhR gene deletion on expressions of PkvgAS, PkvhAS, and PkvhR.

Strains

b-Galactosidase activity (Miller units) [mean– SD (folda_)] pA16 (PkvgAS::lacZ) pAh01 (PkvhAS::lacZ) pRP05 (PkvhR::lacZ) Z01 44– 3 (1.00) 226– 13 (1.00) 374– 6 (1.00) AZ18 (kvgA-) 232– 4 (0.52) 166– 8 (0.74) 250– 8 (0.67) AhZ01 (kvhA-₎ ₄₃₁_{– 2 (0.96)} ₂₂₈_{– 11 (1.00)} ₂₃₇_{– 5 (0.63)} RZ01 (kvhR-₎ ₄₃₈_{– 9 (0.98)} ₂₃₂_{– 6 (1.02)} ₁₇₄_{– 10 (0.47)} a_{Compared with Z01 carrying each the detected plasmid.}

Fig. 3. EMSA assessment of the specific DNA binding activity of KvgA, KvhA, and KvhR. The DNA fragments of the PkvgA,

PkvhA, and PkvhRwere labeled with [g-32P]ATP and used as probes.

The recombinant KvgAt, KvhA, and KvhRt added to the binding assay mixture. The amounts of protein used are indicated at the top

of each lane (lanes 1 to 3). Specific competition was performed by adding the unlabelled DNA fragments into the mixture (lane 4). The DNA and protein complexes formed are indicated as C and the free forms are indicated as F.

Table 4. Characterization of the K. pneumoniae CG43S3-Z01 derived mutants. Strains CPS amounts LD50 (CFU) (mean quantity– SDa ) Foldb Z01 22.8– 3.8 1.00 3· 103 RcsBZ01 (rcsB-₎ _11.6_{– 2.8} _0.51 _NDc AZ18 (kvgA-) 15.7– 0.3 0.68 2.75· 105 AhZ01 (kvhA-) 24.4– 2.4 1.07 3· 103 RZ01 (kvhR-₎ _13.6_{– 1.5} _0.59 ₃_{· 10}5

AAh01 (kvgA-kvhA-) 11.9– 2.2 0.52 4· 105 AR01 (kvgA-_kvhR-₎ _12.9_{– 0.8} _0.56 ₄_{· 10}5

AhR01 (kvhA-kvhR-) 17.6– 0.9 0.77 3· 105 AAhR01 (kvgA-_kvhA-_kvhR-₎ _15.1_{– 0.3} _0.66 ₄_{· 10}5 a_{Values are the averages of triplicate samples and are given as}

micrograms of uronic acid per 109 CFU.bCompared with Z01.

c_{ND means not determined.}

(8)

reduction of the bacterial CPS, an important virulence factor that affects bacterial resistance to phagocytosis by polymorphonuclear cells (44–46), leads to decrease of the virulence. The kvhA deletion mutant, AhZ01, of group II

exhibited unchanged LD50and a slight increase of

glucu-ronic acid content in comparing with that of wild type bacteria Z01. Although classified into the same group as

the kvhA- mutant (Fig. 2A), the LD50 of kvhA-kvhR

-mutant appeared to be comparable with those of group I

bacteria (Table 4). In addition, kvhA-kvhR-mutant

pro-duced less amount of CPS than either kvhA-_{mutant or}

the wild type bacteria. In comparing with the kvhR mutant

of group I, however, the kvhA-kvhR- mutant produced

more CPS. Consistent with the result of string test as shown in Fig. 2A, this suggests a negative role of KvhA on cps expression and deletion of kvhA released the rep-ression of cps exprep-ression, and hence more CPS were produced.

Regulation of KvgA, KvhR, and KvhA on cps

Expression—In order to validate the role of each of the response regulators on cps expression, a series of lacZ fusion constructs, containing each of the putative cps

pro-moters were generated. These include Porf1-2, which

com-prises the non-translated sequence 724-bp upstream of

orf1-2; Porf3-15, which comprises the non-translated

sequence 890-bp upstream of the operon orf3-15, and Porf16-17, which comprises the 244-bp non-translated

sequence upstream of orf16-17 (Fig. 4A). These plasmids were then transformed into wild type bacteria, the deletion

mutants kvgA-_{, kvhA}-_{, kvhR}-_{, and rcsB}-_{and also the wild}

type strain carrying a multicopy plasmid expressing with kvhA, pAHm, and the b-galactosidase activities were mea-sured. Figure 4B (a, b, and c) shows that the activity of

Porf1-2, Porf3-15, and Porf16-17in the kvhR deletion mutant

RZ01 were approximately 50% lower than those of Z01, implying a positive regulatory role of KvhR. Transforma-tion of these bacteria with a kvhR expressing plasmid pRC02 complemented the deleting effects, which confirmed the positive regulation of KvhR on cps expression. The

activity of Porf1-2 was eliminated in the rcsB deletion

mutant [Fig. 4B (a)], which could be explained by the

pre-sence of a typical RcsAB box 50_{-TAAGATTATTCTCA-3}0_(G)

in the region from 168 to 181 nucleotides upstream of K2 orf1-2. As shown in Fig. 4B (b and c), despite the lack of a

typical RcsAB box in Porf3-15 and Porf16-17, both promoter

activities were still affected by rcsB mutation. No apparent

change for either activity of Porf1-2or Porf3-15was observed

in the kvgA deletion strain. A comparison with the wild-type strain showed that the deletion of kvgA reduced Porf16-17activity by approximately 30% which could also be

complemented by supplying the mutant bacteria with a kvgA expression plasmid pA14. This reveals that the response regulator KvgA is also involved in the regulation of the expression of transcript orf16-17. The ORF16 and ORF17, encoding ManC, GDP-mannose pyrophosphory-lase, and ManB, phosphomannomutase, respectively, have been demonstrated to be required in the synthesis of Klebsiella K2 sugar nucleotide precursor (48). The question of why the particular step of the CPS biosynthetic pathway in the bacteria involves complex regulation remains to be answered. As shown in Fig. 4C, the activity

of either Porf1-2, Porf3-15, or Porf16-17 in the kvhA

dele-tion strain was indistinguishable from that in the wild type strain Z01. However, in the presence of pHAm,

activity of Porf1-2, Porf3-15and Porf16-17reduced by

approxi-mately 5.5-, 3- and 2.5-fold, which further supported

(A)

(B)

Fig. 4. A: Organization of the K. pneumoniae K2 cps gene cluster. Putative promoters of the three cps transcripts are also indicated. The horizontal arrows that begin with a solid circle represent the putative transcriptional units. B: Expression of K2 cps gene in various genetic back-grounds. The plasmids carrying Porf1-2(a), Porf3-15(b), and Porf16-17

(c) promoter fused with lacZ gene and transferred into wild type, kvgA-, kvhR-, and rcsB- respec-tively by conjugation and shown as open bar. The complementa-tion test was performed and shown as black bar. C: The plas-mids, pOrf12, pOrf315, and pOrf1617, were transferred into wild type (open bar), kvhA- (black bar), and wild type strain carrying pAHm (gray bar). The cps-promoter carrying cells were grown in M9 medium to an OD600of 0.7

and the b-galactosidase activities were measured and presented in Miller units as described in ‘‘MATERIALS AND METHODS.’’

(9)

the negative role of KvhA in regulation of the cps expression.

Regulation in Coordination—As shown in Fig. 5, a model is proposed for a coordinate regulation of cps expression in the bacteria. Under a stress environment, the response regulator KvgA exhibits an auto-regulatory activity as well as a positive regulation on the expression of kvhAS, kvhR, and cps-orf 16-17. With a relatively low level of pro-moter activity (Table 4), however, KvhA also affects posi-tively the expression of kvhR. The increasing expression of kvhR hence stimulates the transcription of K2 cps. On the other hand, an overexpression of kvhA under a not yet identified condition, in turn, suppressed the synthesis of K2 CPS at transcriptional level.

A complex 2CS regulatory system has been identified in E. coli CPS synthesis in responding to the environmental changes (43). PhoPQ, PmrAB and the Rcs regulatory system have been shown to regulate expression of ugd

encoding UDP-glucose dehydrogenase, an enzyme

required for the synthesis of polysaccharide on the in

coor-dination in E. coli (49). While the activity of PkvgAS, PkvhAS

and PkvhRwere measured, rcsB deletion appeared no effect

on the expression of either kvgAS, kvhAS, or kvhR.

More-over, no apparent change of PrcsBactivity was observed in

either of kvgA, kvhA, and kvhR mutants (data not shown). This suggests an independent regulation of RcsB and the three response regulators on cps expression in K. pneumo-niae CG43 (Fig. 5). We and others have observed that para-logous 2CS proteins may regulate similar functions, probably at different levels (50, 51). By using mutagenesis analysis, promoter activity measurement and EMSA, we are able to demonstrate an interacting regulation among the three paralogous response regulators. In addition, they are all responsible for modulation of the mucoidy and virulence of K. pneumoniae CG43, most likely through a transcriptional regulation of the cps expression.

This work was supported by National Science Council of the Republic of China (NSC92-2311-B009-001 to H. L. P.)

REFERENCES

1. Stock, A.M., Robinson, V.L., and Goudreau, P.N. (2000) Two-component signal transduction. Annu. Rev. Biochem. 69, 183–215

2. Mizuno, T. (1998) His-Asp phosphotransfer signal transduc-tion. J. Biochem. 123, 555–563

3. West, A.H. and Stock, A.M. (2001) Histidine kinases and response regulator proteins in two-component signaling systems. Trends Biochem. Sci. 26, 369–376

4. Fabret, C., Feher, V.A., and Hoch, J.A. (1999) Two-component signal transduction in Bacillus subtilis: How one organism sees its world. J. Bacteriol. 181, 1975–1983

5. Wosten, M.M. and Groisman, E.A. (1999) Molecular char-acterization of the PmrA regulon. J. Biol. Chem. 274, 27185–27190

6. Kato, A., Latifi, T., and Groisman, E.A. (2003) Closing the loop: the PmrA/PmrB two-component system negatively controls expression of its posttranscriptional activator PmrD. Proc. Natl. Acad. Sci. USA 100, 4706–4711

7. Chou, F.F. and Kou, H.K. (1996) Endogenous endophthalmitis associated with pyogenic hepatic abscess. J. Am. Coll. Surg. 182, 33–36

8. Han, S.H. (1995) Review of hepatic abscess from Klebsiella pneumoniae. An association with diabetes mellitus and septic endophthalmitis. West. J. Med. 162, 220–224

9. Lau, Y.J., Hu, B.S., Wu, W.L., Lin, Y.H., Chang, H.Y., and Shi, Z.Y. (2000) Identification of a major cluster of Klebsiella pneumoniae isolates from patients with liver abscess in Taiwan. J. Clin. Microbiol. 38, 412–414

10. Peng, H.L., Wang, P.Y., Wu, J.L., Chiu, C.T., and Chang, H.Y. (1991) Molecular epidemiology of Klebsiella pneumoniae. Zhonghua Min Guo Wei Sheng Wu Ji Mian Yi Xue Za Zhi 24, 264–271

11. Wiener, J., Quinn, J.P., Bradford, P.A., Goering, R.V., Nathan, C., Bush, K., and Weinstein, R.A. (1999) Multiple antibiotic-resistant Klebsiella and Escherichia coli in nursing homes. JAMA 281, 517–523

12. Matsushita, M. and Janda, K.D. (2002) Histidine kinases as targets for new antimicrobial agents. Bioorg. Med. Chem. 10, 855–867

13. Chang, H.Y., Lee, J.H., Deng, W.L., Fu, T.F., and Peng, H.L. (1996) Virulence and outer membrane properties of a galU mutant of Klebsiella pneumoniae CG43. Microb. Pathog. 20, 255–261

14. Konig, J., Bock, A., Perraud, A.L., Fuchs, T.M., Beier, D., and Gross, R. (2002) Regulatory factors of Bordetella pertussis affecting virulence gene expression. J. Mol. Microbiol. Biotech-nol. 4, 197–203

15. Lai, Y.C., Yang, S.L., Peng, H.L., and Chang, H.Y. (2000) Identification of genes present specifically in a virulent strain of Klebsiella pneumoniae. Infect. Immun. 68, 7149–7151 16. Lai, Y.C., Lin, G.T., Yang, S.L., Chang, H.Y., and Peng, H.L.

(2003) Identification and characterization of KvgAS, a two-component system in Klebsiella pneumoniae CG43. FEMS Microbiol. Lett. 218, 121–126

17. Lowe, T.M. and Eddy, S.R. (1997) tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Res, 25, 955–964

18. Lai, Y.C., Peng, H.L., and Chang, H.Y. (2001) Identification of genes induced in vivo during Klebsiella pneumoniae CG43 infection. Infect. Immun. 69, 7140–7145

19. Skorupski, K. and Taylor, R.K. (1996) Positive selection vectors for allelic exchange. Gene 169, 47–52

20. Lai, Y.C., Peng, H.L., and Chang, H.Y. (2003) RmpA2, an activator of capsule biosynthesis in Klebsiella pneumoniae Fig. 5. The model is shown for the regulatory circuit of

KvgAS, KvhAS, and KvhR that is responsible for controlling transcriptionally the expression of Klebsiella K2 cps gene cluster.

(10)

CG43, regulates K2 cps gene expression at the transcriptional level. J. Bacteriol. 185, 788–800

21. Whitfield, C. and Paiment, A. (2003) Biosynthesis and assem-bly of Group 1 capsular polysaccharides in Escherichia coli and related extracellular polysaccharides in other bacteria. Carbo-hydr. Res. 338, 2491–2502

22. Whitfield, C. and Roberts, I.S. (1999) Structure, assembly and regulation of expression of capsules in Escherichia coli. Mol. Microbiol. 31, 1307–1319

23. Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor, New York

24. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254

25. Domenico, P., Schwartz, S., and Cunha, B.A. (1989) Reduction of capsular polysaccharide production in Klebsiella pneumo-niae by sodium salicylate. Infect. Immun. 57, 3778–3782 26. Blumenkrantz, N. and Asboe-Hansen, G. (1973) New method

for quantitative determination of uronic acids. Anal. Biochem. 54, 484–489

27. Reed, L.J. and Muench, H. (1938) A simple method of estimating the fifty percent endpoints. Am. J. Hyg. 27, 493–497

28. Chang, C. and Stewart, R.C. (1998) The two-component system. Regulation of diverse signaling pathways in prokar-yotes and eukarprokar-yotes. Plant Physiol. 117, 723–731

29. Stock, J.B., Stock, A.M., and Mottonen, J.M. (1990) Signal transduction in bacteria. Nature 344, 395–400

30. Volz, K. (1993) Structural conservation in the CheY superfam-ily. Biochemistry 32, 11741–11753

31. Utsumi, R., Katayama, S., Ikeda, M., Igaki, S., Nakagawa, H., Miwa, A., Taniguchi, M., and Noda, M. (1992) Cloning and sequence analysis of the evgAS genes involved in signal transduction of Escherichia coli K-12. Nucleic Acids Symp. Ser. 149–150

32. Roy, C.R. and Falkow, S. (1991) Identification of Bordetella pertussis regulatory sequences required for transcriptional activation of the fhaB gene and autoregulation of the bvgAS operon. J. Bacteriol. 173, 2385–2392

33. Cotter, P.A. and Jones, A.M. (2003) Phosphorelay control of virulence gene expression in Bordetella. Trends Microbiol. 11, 367–373

34. Nishino, K., Inazumi, Y., and Yamaguchi, A. (2003) Global analysis of genes regulated by EvgA of the two-component regulatory system in Escherichia coli. J. Bacteriol. 185, 2667–2672

35. Masuda, N. and Church, G.M. (2003) Regulatory network of acid resistance genes in Escherichia coli. Mol. Microbiol. 48, 699–712

36. Kato, A., Ohnishi, H., Yamamoto, K., Furuta, E., Tanabe, H., and Utsumi, R. (2000) Transcription of emrKY is regulated by the EvgA-EvgS two-component system in Escherichia coli K-12. Biosci. Biotechnol. Biochem. 64, 1203–1209

37. Masuda, N. and Church, G.M. (2002) Escherichia coli gene expression responsive to levels of the response regulator EvgA. J. Bacteriol. 184, 6225–6234

38. Nishino, K. and Yamaguchi, A. (2001) Overexpression of the response regulator evgA of the two-component signal transduction system modulates multidrug resistance

conferred by multidrug resistance transporters. J. Bacteriol. 183, 1455–1458

39. Tanabe, H., Yamasaki, K., Katoh, A., Yoshioka, S., and Utsumi, R. (1998) Identification of the promoter region and the transcriptional regulatory sequence of the evgAS operon of Escherichia coli. Biosci. Biotechnol. Biochem. 62, 286–290 40. Zhou, L., Lei, X.H., Bochner, B.R., and Wanner, B.L. (2003)

Phenotype microarray analysis of Escherichia coli K-12 mutants with deletions of all two-component systems. J. Bacteriol. 185, 4956–4972

41. Fang, C.T., Chuang, Y.P., Shun, C.T., Chang, S.C., and Wang, J.T. (2004) A novel virulence gene in Klebsiella pneumoniae strains causing primary liver abscess and septic metastatic complications. J. Exp. Med. 199, 697–705

42. Park, S.H., Eriksen, J., and Henrikden, S.D. (1967) Structure of the capsular polysaccharide of Klebsiella pneumoniae type 2 (B). Acta Pathil. Microbiol. Scand. 69, 431–436

43. Majdalani, N. and Gottesman, S. (2005) The Rcs Phosphorelay: A Complex Signal Transduction System. Annu. Rev. Microbiol 44. Kabha, K., Nissimov, L., Athamna, A., Keisari, Y., Parolis, H., Parolis, L.A., Grue, R.M., Schlepper-Schafer, J., Ezekowitz, A.R., Ohman, D.E. et al. (1995) Relationships among capsular structure, phagocytosis, and mouse virulence in Klebsiella pneumoniae. Infect. Immun. 63, 847–852

45. Highsmith, A.K. and Jarvis, W.R. (1985) Klebsiella pneumo-niae: selected virulence factors that contribute to pathogeni-city. Infect. Control. 6, 75–77

46. Held, T.K., Jendrike, N.R., Rukavina, T., Podschun, R., and Trautmann, M. (2000) Binding to and opsonophagocytic activ-ity of O-antigen-specific monoclonal antibodies against encap-sulated and nonencapencap-sulated Klebsiella pneumoniae serotype O1 strains. Infect. Immun. 68, 2402–2409

47. Wehland, M. and Bernhard, F. (2000) The RcsAB box. Characterization of a new operator essential for the regulation of exopolysaccharide biosynthesis in enteric bacteria. J. Biol. Chem. 275, 7013–7020

48. Arakawa, Y., Wacharotayankun, R., Nagatsuka, T., Ito, H., Kato, N., and Ohta, M. (1995) Genomic organization of the Klebsiella pneumoniae cps region responsible for serotype K2 capsular polysaccharide synthesis in the virulent strain Chedid. J. Bacteriol. 177, 1788–1796

49. Mouslim, C., Latifi, T., and Groisman, E.A. (2003) Signal-dependent requirement for the co-activator protein RcsA in transcription of the RcsB-regulated ugd gene. J. Biol. Chem. 278, 50588–50595

50. Ochsner, U.A. and Vasil, M.L. (1996) Gene repression by the ferric uptake regulator in Pseudomonas aeruginosa: cycle selection of iron-regulated genes. Proc. Natl. Acad. Sci. USA 93, 4409–4414

51. Chen, Y.T., Chang, H.Y., Lu, C.L., and Peng, H.L. (2004) Evo-lutionary analysis of the two-component systems in Pseudo-monas aeruginosa PAO1. J. Mol. Evol. 59, 725–737

52. Thompson, J.D., Higgins, D.G., and Gibson, T.J. (1994) CLUS-TAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids. Res. 22, 4673–4680

53. Kumar, S., Tamura, K., Jakobsen, I.B., and Nei, M. (2001) MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17, 1244–1245