competence repression under oxygen limitation through the ... · ppt14 0.98-kb xbai/xhoi amplicon...

JOURNAL OF BACTERIOLOGY,0021-9193/01/$04.0010 DOI: 10.1128/JB.183.15.4599–4608.2001

Aug. 2001, p. 4599–4608 Vol. 183, No. 15

Copyright © 2001, American Society for Microbiology. All Rights Reserved.

Competence Repression under Oxygen Limitation through theTwo-Component MicAB Signal-Transducing System in

Streptococcus pneumoniae and Involvementof the PAS Domain of MicB

JOSE R. ECHENIQUE† AND MARIE-CLAUDE TROMBE*

Laboratoire de Genetique et Physiologie Bacterienne, E.A. 3036, Centre Hospitalo Universitaire deRangueil, Universite Paul Sabatier, 31403 Toulouse Cedex, France

Received 21 February 2001/Accepted 14 May 2001

In Streptococcus pneumoniae, a fermentative aerotolerant and catalase-deficient human pathogen, oxidaseswith molecular oxygen as substrate are important for virulence and for competence. The signal-transducingtwo-component systems CiaRH and ComDE mediate the response to oxygen, culminating in competence. Inthis work we show that the two-component MicAB system, whose MicB kinase carries a PAS domain, is alsoinvolved in competence repression under oxygen limitation. Autophosphorylation of recombinant MicB andphosphotransfer to recombinant MicA have been demonstrated. Mutational analysis and in vitro assaysshowed that the C-terminal part of the protein and residue L100 in the N-terminal cap of its PAS domain areboth crucial for autokinase activity in vitro. Although no insertion mutation in micA was obtained, expressionof the mutated allele micA59DA did not change bacterial growth and overcame competence repression undermicroaerobiosis. This was related to a strong instability of MicA59DA-PO4 in vitro. Thus, mutations whicheither reduced the stability of MicA-PO4 or abolished kinase activity in MicB were related to competencederepression under microaerobiosis, suggesting that MicA-PO4 is involved in competence repression whenoxygen becomes limiting. The micAB genes are flanked by mutY and orfC. MutY is an adenine glycosylaseinvolved in the repair of oxidized pyrimidines. OrfC shows the features of a metal binding protein. We did notobtain insertion mutation in orfC, suggesting its requirement for growth. It is proposed that MicAB, with itsPAS motif, may belong to a set of functions important in the protection of the cell against oxidative stress,including the control of competence.

In the catalase-negative pathogen Streptococcus pneumo-niae, which has essentially fermentative metabolism, oxygen lim-itation in a microaerobic atmosphere abolishes developmentalcompetence. Nox, an NADH oxidase that produces water byreducing O2 as it recycles NADH, has been shown to contributeto competence regulation by oxygen (3, 8). Studies of oxygen-independent mutant strains demonstrated the involvement ofthe two-component systems (TCSs) CiaRH and ComDE inthis regulation (7, 8). To characterize in more detail the regu-latory network facilitating bacterial adaptation to oxygen avail-ability, we searched for amino acid sequences correspondingto motifs putatively involved in O2 and redox sensing, in thepublicly available pneumococcal genome sequence (http://www.tigr.org).

The PAS domain may perceive cell energetic status by sens-ing oxygen, redox potential, ligands, proton motive force, andlight (22; for review, see reference 25). PAS domains have beenfound in bacterial, archaeal, and eukaryotic proteins. Redoxsensing and the corresponding signal transduction via two-component systems carrying PAS domains is one strategy used

by bacteria and archaea for adaptation to variations in ambientoxygen concentration (4). PAS domains are frequently foundupstream from the kinase transmitter domain. A heme-con-taining domain in a sensor histidine kinase from Sinorhizobiummeliloti directly detects oxygen (11); however, in most caseslittle is known about the primary signals triggering the kinaseactivity. The presence of a PAS motif in the essential two-component HK02-RR02 system was recently described forS. pneumoniae (15), but its role in competence was not eluci-dated. The HK02-RR02 TCS has been detected in virulentstrains of S. pneumoniae, in which it was designated 492HK-RR. Insertional mutagenesis of the kinase did not impair growthin mouse lungs (26). By genome analysis we found this proteinto be the only one containing a PAS domain in S. pneumoniaeand have established its involvement in competence regulationby oxygen. The kinase has therefore been designated MicB forits role under microaerobiosis, and its cognate response regu-lator was named MicA. Mutational and biochemical studiesallowed us to demonstrate that MicB autophosphorylation re-quires the L100 residue of PAS and that the D59 residue ofMicA is involved in the stability of the phosphorylated form ofthe response regulator. Genetic dissection and biochemicalanalysis suggest that MicA-PO4 functions upstream of ComDE torepress competence when oxygen is limiting.

MATERIALS AND METHODS

Bacterial strains, growth conditions, antibiotic susceptibility, and S. pneu-moniae transformation. The bacterial strains used in this study are listed in Table

* Corresponding author. Mailing address: Laboratoire de Genet-ique et Physiologie Bacterienne, E.A. 3036, Centre Hospitalo Univer-sitaire de Rangueil, Universite Paul Sabatier, 31403 Toulouse Cedex,France. Phone: (33)61-322974. Fax: (33)61-322620. E-mail: [email protected].

† Present address: Departmento de Bioquımica Clınica, Facultad deCiencias Quımicas, Universidad Nacional de Cordoba, Pabellon Ar-gentina, Ciudad Universitaria, CP 5000 Cordoba, Argentina.

4599

on July 18, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

1. Escherichia coli was grown and induced by 1 mM isopropyl-b-D-thiogalacto-pyranoside (IPTG) under standard conditions (23). S. pneumoniae was grown,transformed, and tested for competence development under aerobic or mi-croaerobic conditions as described previously (7). Antibiotic susceptibility toerythromycin and lincomycin was determined on plates containing various con-centrations of these antibiotics. Plates were incubated for 12 h at 37°C underaerobic and microaerobic conditions. For strain screening, antibiotics were usedat the following concentrations; for E. coli, 100 mg of ampicillin ml21 and 50 mgof spectinomycin ml21; for S. pneumoniae, 2 mg of erythromycin ml21, 2 mg ofrifampin ml21, 40 mg of kanamycin ml21, and 10 mg of spectinomycin ml21.

DNA manipulation, plasmid construction, and sequencing. Standard recom-binant DNA techniques were used, as described by Sambrook et al. (23). Re-

striction enzyme digestions were conducted according to the manufacturers’instructions, and digestion products were separated by electrophoresis in agarosegels in Tris-borate-EDTA buffer (23). All PCRs were performed in a GeneAmpPCR System 9600 (Perkin-Elmer Cetus). The cloning vectors were pWSK29 andpAM239 (Table 1). The nucleotide sequence of all the constructs was establishedby dye-terminator cycle sequencing with an automated 373 DNA Sequencer(Perkin-Elmer Applied Systems).

Cloning the micAB-orfC genes. The nucleotide sequence of the open readingframe (ORF) containing the PAS domain (Fig. 1A) was obtained from theserotype 4 strain of S. pneumoniae, the genome of which has been published (seeThe Institute for Genomic Research website at http://www.tigr.org). An ORFwas identified and micB was designated, flanked by micA and orfC (Fig1A).

TABLE 1. Bacterial strains and plasmids used in this study

Strain or plasmid Genotype or description Phenotypea Source or reference

StrainS. pneumoniae

Cp1015 Rx derivate; str1 hexA Smr 19Cp6602 comE38KE micB::km Kanr This workCp7002 micB::km Kanr This workCp7003 micB100LR This workCp7009 micA59DA This workCp7039 micA59DA comE::km Kanr This workCp7089 micA59DA comA::ery Eryr This work

E. coliDH5a supE44 DlacU169 (f80 lacZDM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1 Bethesda Research

LaboratoriesTOPI0 F2 mcrA D(mrr-hsdRMS-mcrBC) f80lacZDM15 DlacX74 deoR recA1 araD139 D(ara-leu)

7697 galU galK rpsL endA nupGInvitrogen

PlasmidpAM239 Derived from pBR322; ColE1 origin; lacZa selection Sptr 10pFS16 0.68-kb HindIII amplicon from Cp1015 containing the 16S gene inserted into pAM239 Sptr 7pPFMD52 1.92-kb KpnI/HindIII amplicon (primers FMD52 and Rmic) from pPT12 inserted into

pWSK29Ampr This work

pPFMD59 1.90-kb KpnI/HindIII amplicon (primers FMD59 and Rmic) from pPT12 inserted intopWSK29

Ampr This work

pPFMD105 1.0-kb BamHI/HindIII amplicon (primers FMD105 and Rmic) from pPT12 inserted intopWSK29

Ampr This work

pPFMG128 1.0-kb BamHI/HindIII amplicon (primers FMG128 and Rmic) from pPT12 inserted intopWSK29

Ampr This work

pPFML100 1.0-kb BamHI/HindIII amplicon (primers FML100 and Rmic) from pPT12 inserted intopWSK29

Ampr This work

pPFMN120 1.0-kb BamHI/HindIII amplicon (primers FMN120 and Rmic) from pPT12 inserted intopWSK29

Ampr This work

pPJ1 pUC derivative containing an HincII fragment carrying the kanamycin resistance gene Ampr Kanr 21pPMAE 0.7-kb BamHI/EcoRI amplicon (primers FMAE and RMAE) from pPT12 containing the

micA gene inserted into pTrcHis2AAmpr This work

pPMAE59 0.7-kb BamHI/EcoRI amplicon (primers FMAE and RMAE) from Cp7009 containing themicA59DA gene inserted into pTrcHis2A

Ampr This work

pPMAI 0.4-kb EcoRI/HindIII amplicon (primers FMAI and RMAI) from pPT12 containing aninternal fragment of micA inserted into pAM239

Sptr This work

pPMBE1 1.23-kb BamHI/EcoRI amplicon (primers FMBE and RMBE) from pPT12 containing themicB gene inserted into pTrcHis2A

Ampr This work

pPMBE100 1.23-kb BamHI/EcoRI amplicon (primers FMBE and RMBE) from Cp7003 containingthe micB100LR gene inserted into pTrcHis2A

Ampr This work

pPMBEN 0.88-kb BamHI/SalI DNA fragment from pPMBE1 containing the N terminal of micBpinserted into pTrcHis2A

Ampr This work

pPMCI 0.5-kb EcoRI/HindIII amplicon (primers ForfCin and RorfCin) from pPT14 containing aninternal fragment of orfC inserted into pAM239

Sptr This work

pPT12 2.1-kb KpnI/HindIII amplicon (primers Fmic and Rmic) from Cp1015 containing themicAB genes inserted into pWSK29

Ampr This work

pPT12-Km pPT12 micB::km Ampr Kanr This workpPT14 0.98-kb XbaI/XhoI amplicon (primers ForfC and RorfC) from Cp1015 containing the orfC

gene inserted into pWSK29Ampr This work

pPT14-Km pPT14 orfC::km Ampr Kanr This workpPT18 2.9-kb XbaI-PstI fragment from Cp1015 amplified by PCR containing the comCDE

operon, inserted into pWSK29Ampr 7

pPT18-KCl pPT18 comE::km Ampr Kanr This workpTrcHis2A pUC-derived vector for the production and purification of recombinant proteins in E. coli Ampr InvitrogenpWSK29 Cloning vector derived from pBlueScript Ampr 29pXF208 Plasmid containing an internal fragment of comA Emr 6

a Amp, ampicillin; Ery, erythromycin; Kan, kanamycin; Spt, spectinomycin.

4600 ECHENIQUE AND TROMBE J. BACTERIOL.


.org/D

ownloaded from

http://jb.asm.org/

Primers Fmic (59-CGATATGGTACCGAATTACCCACTTGCCAAACCC)and Rmic (59-CACACAAGCTTCTAGTCTTCTACTTCATCCTCCCATA)were designed to amplify by PCR the micAB genes from DNA of Cp1015. Theresulting amplicon was digested with KpnI and HindIII and cloned to give pPT12(Fig. 1C and Table 1). For orfC, primers FmicC (59-GGCGCGTCTAGACCAAGAGTGAATACGGCAAGGG) and RmicC (59-CGCTCGCTCGAGGCTCCCTTTTTTAATGGTAACACC) were used to amplify orfC from Cp1015 DNA,and the PCR product, digested with XbaI and XhoI, was cloned to give pPT14(Fig. 1C and Table 1). The complete nucleotide sequence of micAB-orfC fromstrain Cp1015 is available in GenBank (accession number AF219111).

Insertional mutagenesis. For gene disruption two strategies were set up. Ei-ther the kanamycin-resistance cassette, aphA-3, with a candidate transcriptionterminator at the 59 extreme (DG° 5 226.5 kcal/mol) from pPJ1 (21), wasinserted into relevant genes, previously cloned in pWSK29 (29), or internalamplicons of the relevant genes were cloned into pAM239 (10) to be used forplasmid insertion mutagenesis. The wild-type recipient Cp1015 was transformedby the mutagenic plasmids (10 mg/ml), and, accordingly, kanamycin or spectino-mycin selection was applied to select for allelic exchange (kanamycin) or plasmidinsertion (spectinomycin). Transformation efficiency of the recipient strain,Cp1015, was determined using chromosomal DNA (1 mg/ml) from the rifampin-resistant Cp1016 strain. Routinely the transformant recovery for Rifr transfor-mants was .2% of the total population. For micB mutagenesis the km cassette

was inserted into the SalI site of micB in pPT12 to generate pPT12-Km (Table1). Transformation of strain Cp1015 with pPT12-Km yielded 1% Kmr transfor-mants. The putative recombinants were checked by PCR with primers Fmic andRmic, and one recombinant clone was retained and designated strain Cp7002(micB::km). For micA mutagenesis, an internal amplicon was obtained withprimers FmaI (59-CATCACGAATTCTTATTATTCTGGATTTGATGCTTCC) and RmaI (59-GGTCACAAGCTTGCAAGTGTTCGCGCGTGATGACTTG) and inserted between the EcoRI and HindIII sites of pAM239 to obtainpPMAI (Table 1). Strain Cp1015 was transformed with pPMAI, and spectino-mycin selection was used. Mutagenesis of orfC was attempted with pPT14-Kmcontaining orfC::km (Table 1) and pPMCI containing the amplicon obtained withprimers ForfC in 59-GGGCATGGATCCGCTGAAATTAACCGTAAGCCAGand RorfC in 59-AGCGCCGAATTCCGATTTCCTAGCGTCCGAAT insertedbetween the EcoRI and HindIII sites of pAM239 (Table 1). Strain Cp1015 wastransformed with pPMCI, and spectinomycin selection was applied accordingly.

Point mutagenesis. Missense mutations were obtained by PCR amplificationof the relevant genes using mutagenic primers carrying a single restriction site tofacilitate the selection of recombinant clones after transformation, as describedelsewhere (3). The resulting amplicons were cloned into pWSK29, and theresulting plasmids were used to transform the wild-type strain, Cp1015. Recom-binant clones were screened by restriction analysis of the amplicons obtainedwith primers Fmic and Rmic. The mutations were confirmed by DNA sequencing

FIG. 1. Organization of the mutY-micAB-orfC locus with the conserved modules in MicA and MicB and strategy for producing the recombinantproteins. (A) Gene organization with significant conserved motifs in MicA, MicB, and OrfC. RD, receiver domain; HTH, helix turn helix; TM,transmembrane domain; HK, histidine kinase; MbL, metallo-b-lactamase. (B) Site-specific mutagenesis of MicA and MicB. (C) Cloning strategyfor the nucleotide sequence determination of micAB and orfC and the expression of recombinant proteins in E. coli (the sizes of the recombinantproteins were calculated with the addition of 2.5 kDa for the His tag). (D) Sequence alignment of candidate PAS domains in MicB orthologuesfrom B. subtilis (YYCG Bacsu, GenBank no. D78193), L. lactis (L1KINC Llact, GenBank no. AF178425), S. aureus (YYCG Staur, GenBank no.AF136709), FixL from S. meliloti (FIXL Rhime, GenBank no. J03174), and the PAS consensus sequence (http://smart.embl-heidelberg.de/).Identical residues are in bold. Residues chosen as mutagenesis targets are underlined.

VOL. 183, 2001 COMPETENCE REGULATION BY O2 INVOLVES THE MicAB SYSTEM 4601


.org/D

ownloaded from

http://jb.asm.org/

using primers FMBT (59-TTGTGACCCTCTTATTACTG) for micB and RMAT(59-CAACTCACGATTGGAG) for micA. We obtained 2 to 4% recombinantsby this procedure. Mutagenesis of residues L100 to R, D105 to A, N120 to I, andG128 to Q in MicB was achieved using the mutagenic primers FML100 (59-TGGATCCGAGGCTAAATAGTATCCGGTTTTATATGACAG), FMD105 (59-GCCGGATCCTGTTTTATATGACAGCTGGGGTTCTTGCGAC), FMN120(59-CCGGATCCAGATTATCATGATTATCGATACAGCCAAGAAG) andFMG128 (59-CCGGGATCCAAGAAGCAACTGCAGTTGGTTAAGGAAGATG), respectively, and the complementary primer Rmic. The resulting plasmidswere pPFML100, pPFMD105, pPFN120, and pPFMG128, respectively. Mu-tagenesis of residues D52 to Q and D59 to A in MicA was achieved with themutagenic primers FMD52 (59-CGATATGGTACCGATATTATTATTCTGCAGTTGATGCTTCCAGAA) and FMD59 (59-CGATATGGTACCATGCTTCCAGAAATTGCCGGTTTAGAAGTTGCT), respectively, and the complementaryRmic oligonucleotide. This gave pPFMD52 and pPFMD59, respectively (Table 1).

Northern blot analysis. Total RNA preparation, Northern blots, and probelabeling were carried out as described elsewhere (7). For studies of the micABregion, we used freshly prepared RNA that had not been kept frozen. SpecificmRNA species were detected by hybridization with the following 32P-labeledprobes: a 540-bp comE fragment from pPT18 (GenBank sequence U33315; po-sition, 2450 to 2990), a 730-bp micB fragment from pPT12 (GenBank sequenceAF219111; position, 1100 to 1830); a 490-bp orfC fragment from pPmicCin (Gen-Bank sequence AF219111; position, 2342 to 2832); and a 650-bp 16S rRNA frag-ment from pP16S (GenBank sequence X58312; position, 166 to 816) from Cp1015.

Overproduction and isolation of His6-tagged MicA and MicB proteins inE. coli. The DNA fragment encoding the cytoplasmic region of MicB, with thesequence encoding the first 38 N-terminal amino acids deleted, designatedMicB*, was obtained by PCR amplification using primers FMBE (59-GGTGAGGATCCGCGTGATAATATTCAGTTGAAGCAAGTCAAT) and RMBE(59-GCGCGAATTCTTGTCTTCTACTTCATCCTCCCATACTTCTTC) con-taining EcoRI and BamHI restriction sites, respectively, facilitating insertion intothe expression vector pTrcHis2A (Invitrogen) to give pPMBE1. To delete theC-terminal domain of MicB, which carries the kinase and the ATPase module, aSalI-EcoRI DNA fragment from pPMBE1 was inserted into pTrcHis2A to givepPMBEN. Primers FMAE (59-GGCACGGATCCGATGAAAAAAATACTAATTGTAGATGATGAG) and RMAE (59-CGCGAATTCTTAGCATTATTTCTCATGTAATACCCTACACC), containing EcoRI and BamHI restriction sites,respectively, were used to clone micA for insertion into pTrcHis2A to givepPMAE. For MicB*100LR and MicA59DA, the DNA targets for amplificationwere obtained from S. pneumoniae Cp7003 and Cp7009, respectively (Table 1).The corresponding amplicons were inserted into pTrcHis2A as described forMicBp to give pPMBE100 and pPMAE59, respectively.

E. coli TOP10 cells carrying the relevant plasmid were induced with 1 mMIPTG and lysed under denaturing conditions, according to Invitrogen’s recom-mendations. The lysate was loaded onto a 3-ml Ni-nitrilotriacetic acid column.The column was washed with 0.5 M NaCl and 10 mM imidazole in 20 mMsodium phosphate buffer (pH 6). Refolding was achieved in a linear 7 to 0 Murea gradient in 20 mM sodium phosphate buffer (pH 6.0) containing 0.5 MNaCl. Proteins were eluted in 0.5 M imidazole and 0.5 M NaCl in 20 mM sodiumphosphate buffer (pH 6.0). The samples were dialyzed against 50 mM Tris (pH7.2), 1 mM EDTA, 100 mM KCl, 1 mM dithiothreitol (DTT) and 30% glycerol(vol/vol). Aliquots were frozen and stored at 280°C. Protein concentration wasdetermined as described by Bradford (5). The mass of the recombinant proteinswas estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis(SDS-PAGE) using mid-range protein molecular mass markers (Promega) andwas compared to their sizes predicted from the length of the coding fragmentswith the addition of 2.5 kDa from His tags.

Protein phosphorylation assays. Proteins were phosphorylated in vitro byincubation with [g-32P]ATP (Amersham Pharmacia Biotech) at 37°C for 10 minand analyzed by electrophoresis in 12% polyacrylamide gels after the addition of100 ml of 23 loading buffer (pH 8.0; 0.2 M Tris, 50 mM EDTA, 0.1 M DTT, 8%SDS, 40% glycerol, 0.1% bromophenol blue). The 32P-labeled bands were de-tected by autoradiography, and the signals were quantified by densitometry usinga Personal Densitometer SI (Molecular Dynamics) with the ImageQuant version3.0 Fast Scan software package. As a control, samples were incubated with[a-32P]ATP in the same conditions as the assays. For MicB* autophosphoryla-tion, 5 mg (100 pmol) of protein was incubated at 23°C in 100 ml of phosphor-ylation buffer (pH 7.5; 50 mM Tris buffer, 0.1 mM [g-32P]ATP, 200 mCi/mmol, 5mM DTT, 50 mM KCl, 0.1 mM EDTA, 10% [vol/vol] glycerol). Aliquots wereremoved at intervals, kept at 4°C, and analyzed as described above. For phos-photransfer from MicB*-PO4 to MicA and to MicA59DA, MicB* phosphoryla-tion reaction was allowed for 4 min, 300 pmol (9 mg) of the wild-type andmutated MicA were added to the assay mixture, and 16-ml aliquots were removed

at intervals, transferred at 4°C, and analyzed as described above. The negativecontrol was a sample in which MicBp was omitted. To investigate the stability ofMicA-PO4 and MicA59DA-PO4, isotopic dilution of [g-32P]ATP in the reactionmixture was achieved by adding unlabeled ATP to a final concentration of 10mM. Aliquots were removed at intervals and analyzed as previously described.MicA-PO4 was treated with acid or alkali by incubation with 0.1 N HCl or 0.1 NNaOH for 20 min at 23°C. Samples were analyzed as described above.

RESULTS

Cloning and nucleotide analysis of the micAB-orfC locusfrom the RX strain Cp1015. We searched for proteins thatmight be involved in bacterial adaptation to oxygen by screen-ing the published genome of S. pneumoniae (http://www.tigr.org), using the WU-BLAST (version 2.0) program and thePAS motif of the FixL protein of S. meliloti as a probe (12). Asingle ORF containing the PAS motif was found in contig4189. According to further functional characterization, thisORF was named micB. ORFs flanking micB were designatedmicA and orfC. The micAB genes are located close to theantimutator, mutY (2, 16, 24) (Fig. 1A).

As a prerequisite to mutational analysis, the nucleotide se-quence of micAB and orfC from the reference strain Cp1015cloned in plasmids pPT12 and pPT14, respectively, has beenestablished (Fig. 1C and Materials and Methods). The de-duced amino acid sequences (GenBank accession numberAF219111) were identical with the sequence published byLange et al. (15; GenBank accession number AJ006392) exceptfor the conservative V44 to I change and the R270 to C mis-sense mutation in the kinase domain of MicB.

Localized mutagenesis of micAB-orfC and phenotypic char-acterization of the mutant strains. The physiological role ofmicAB (492RR-492HK; HK02-RR02) and orfC has been as-sessed in vegetative growth, antibiotic susceptibility, and com-petence development in response to oxygen availability by us-ing a mutational strategy.

The plasmids pPMAI, pPT12-Km, pPT14-Km, and pPMCI(Table 1) were used in the transformation of strain Cp1015 toobtain insertion mutations in micA, micB, and orfC, respec-tively. Recombinants were obtained only with pPT12-Km,resulting in strain Cp7002 carrying a micB insertion muta-tion (Fig. 1). For orfC mutagenesis, plasmids pPT14-Km andpPMCI (Table 1) were used to transform Cp1015 (see Mate-rials and Methods). No recombinants were obtained in threeindependent experiments with each of these plasmids as donorDNA. This suggests that orfC cannot be disrupted and is prob-ably required for in vitro growth in the Cp1015 genetic back-ground. Indeed, Northern blotting of total RNA with an orfC-specific DNA probe allowed verification of the presence oforfC mRNA in all the mutant strains studied (data not shown).Virulent strains of S. pneumoniae deleted for orfC were selec-tively impaired in growth in vivo (26). Functional OrfC is likelyrequired for S. pneumoniae growth in specific conditions, de-pending on the genetic background. We investigated furtherthe function of MicA and MicB by introducing missense mu-tations at critical sites.

The conserved N120 and G128 residues in the PAS core andL100 and D105 in the N-terminal cap (see Fig. 1B and D) (22,25) were chosen as targets for MicB mutagenesis for technicalreasons inherent to the mutagenesis strategy (see Materialsand Methods). Strain Cp1015 was transformed with the muta-



.org/D

ownloaded from

http://jb.asm.org/

genic plasmids pPFML100, pPFMG128, pPFMD105, andpPFMN120 to introduce the mutations into the resident chro-mosome by allelic exchange (Table 1 and Materials and Meth-ods). Cultures were plated without selection, and for eachtransformation 100 colonies were checked. The presence of themutation vas verified by restriction analysis of the relevantamplicons (see Materials and Methods). Transformants wereobtained following transformation with pPFML100 only, giv-ing Cp7003 carrying the micB100LR mutation.

In MicA, we mutated putative phosphorylation sites (seeMaterials and Methods). Plasmids pPFMD52 and pPFMD59,carrying the mutated alleles micA52DQ and micA59DA, re-spectively (Table 1), were used to transform Cp1015. Recom-binant clones were obtained only with pPFMD59. Alignmentof the sequence of MicA with that of the well studied CheY ofE. coli showed that the D59 residue corresponds to residueD64 of CheY. In CheY, D64 is part of a very rigid “g-turnloop,” which has not been mutated (28). In S. pneumoniae,

mutation of this loop did not diminish bacterial growth andone clone carrying the micA59DA mutation gave strain Cp7009(Table 1 and Materials and Methods).

Mutant strains Cp7002, Cp7003, and Cp7009 had growthcharacteristics similar to those of the wild type. They werefurther analyzed for developmental competence and suscepti-bility to erythromycin and lincomycin. None of these mutationsaffected the level of resistance of the bacteria to these antibi-otics (data not shown). This finding was in contrast with theobserved hypersensitivity to macrolide and lincosamide antibi-otics under aerobiosis in yycFG temperature-conditional mu-tants of Staphylococcus aureus (17).

In competence tests for both strains Cp7003 (micB100LR)and Cp7009 (micA59DA), comCDE mRNA (7) was detected incultures grown under microaerobic conditions, and transformantswere obtained (Fig. 2 and 3). In contrast, the micB::km mutantstrain, Cp7002, was not transformable under microaerobiosis de-spite the presence of significant amounts of comCDE mRNA in

FIG. 2. Effect of the micB100LR and micB::km mutations on competence control by O2. Strains Cp1015 (A), Cp7003 (B), and Cp7002 (C) weregrown under aerobiosis or microaerobiosis. Aliquots were withdrawn at 30-min intervals and analyzed for change in biomass by determinations ofthe optical density (OD) at 400 nm (black bars), transformability (grey bars), or by Northern blotting. For Northern blotting, a DNA probe specificfor comE (see Materials and Methods) was used to detect comCDE transcripts. We used 16S rRNA as a qualitative and quantitative internalcontrol. Quantitative densitometry of 16S rRNA gave less than 25% variability between samples taken throughout the period of growth for a givenculture. The experiment was repeated with independent cultures to test reproducibility. The columns are aligned with the corresponding signalsin the Northern blots.



.org/D

ownloaded from

http://jb.asm.org/

these bacteria (Fig. 2C). MicA59DA and MicB100LR there-fore allow comCDE transcription and bacterial transformabil-ity, while the micB::km mutation dissociated comCDE tran-script levels and bacterial transformability under oxygenlimitation, leading to transformant recovery below the detec-tion level (Fig. 3B). This possibly reveals a strong developmen-tal checkpoint after comCDE transcription. In the comE38KEgenetic background, in which the ComE38KE hyperactive re-sponse regulator is expressed (7), the micB::km mutation wassilent because in cultures grown microaerobically Cp6600(comE38KE) and Cp6602 (comE38KE, micB::km) gave 20 and15% Rifr transformants, respectively, in the same experiment.Thus, comE38KE is epistatic to micB::km. This suggests post-transcriptional ComE regulation by MicB.

To investigate further this point, production of competencestimulating factor (CSF) by micB::km mutants and their re-

sponse to the synthetic competence stimulating peptide (CSP)has been addressed. Media conditioned by bacteria from strainCp7002 grown microaerobically during 2 and 4 h contained aCSF activity equivalent to that of media conditioned by thewild-type strain Cp1015 grown in the same conditions, as mea-sured by complementation of strain Cp1008 defective for CSFproduction. Indeed, these media were 103-fold less activethan those produced by the oxygen-independent strain Cp6600grown microaerobically. Taken as controls, media conditionedby aerobically growing bacteria from these three strains showedsimilar activity with regard to Cp1008 activation. This indicatesthat CSF concentrations in cultures of strains Cp7002 andCp1015 growing under oxygen limitation did not reach thethreshold level to trigger competence development. It has beenverified that in the CSP-induced competence regime, strainCp7002 (micB::km) showed the same activation pattern as

FIG. 3. Effect of the micA59DA mutation on competence control by O2. Strains Cp7009 and Cp1015 were grown under aerobiosis ormicroaerobiosis. Aliquots of cultures were withdrawn at 30-min intervals and analyzed as described in the legend to Fig. 1. For Northern blotting,probes specific for comE and micB (see Materials and Methods) were used to detect comCDE and micAB RNA, respectively. The signal for 16SrRNA differed by less than 24% between wells. The experiment was repeated with independent cultures to check reproducibility.



.org/D

ownloaded from

http://jb.asm.org/

strain Cp1015 with a threshold CSP requirement at 1 ng/ml,indicating that expression of the micB::km mutation specifi-cally affects developmental competence but not competencetriggered by addition of extra CSP to the culture. Despite asignificant amount of comCDE mRNA in bacteria from strainCp7002 grown microaerobically, its CSF production remainedunder the threshold level for competence activation, indicatinga posttranscriptional control by full-size MicB.

We investigated further the relationship between micA andthe competence autoregulation network encoded by comCDEand comAB. The response regulator ComE and the ComABsystem facilitating the maturation and export of the compe-tence-stimulating peptide are essential for competence devel-opment (for review, see reference 13). Genetic dissectionshowed that alleles comA::ery and comE::km, both of whichabolished competence, were epistatic to micA59DA because notransformants were obtained in bacteria from microaerobiccultures of strains Cp7089 (comA::ery miA59DA) and Cp7039(comE::km micA59DA), respectively. Thus, competence con-trol by micA occurred upstream from the CSP-ComDE circuit.It should be noticed that mutation micA59DA did not signifi-cantly change the cellular levels of micB mRNA during com-petence development (Fig. 3B).

In order to obtain more precise insight into the mechanismof competence regulation by MicAB, biochemical analysis ofthe recombinant MicA and MicB and their corresponding mu-tated proteins has been undertaken.

Biochemical characterization of the two-component MicABsystem and impact of the mutations which alter competenceregulation on phosphorylation and phosphotransfer. Analysisof the primary sequences of MicA and MicB strongly suggestedthat they form a two-component signaling system. Signal trans-duction systems involving phosphotransfer have been exten-sively described from genome analysis of S. pneumoniae. How-ever, to date, no biochemical evidence has been obtained ofautokinase activity and phosphotransfer in these bacteria.

We expressed genes encoding recombinant proteins with a(His)6 tag fused to the C terminus in E. coli. The full-size MicAand MicA59DA products, as well as MicB*, with the trans-membrane domains deleted (to facilitate expression in E. coli),MicB*100LR, and truncated MicB*-N were purified accordingto procedures described in Materials and Methods (Fig. 1Cand 4). MicB* was autophosphorylated in the presence of[g-32P]ATP (Fig. 5A). Autophosphorylation was abolished inthe truncated product, MicB*-N (Fig. 5B). This demonstratesthat MicB* autokinase activity is in the C-terminal part of theprotein as suggested by computer-assisted prediction (Fig. 1A).MicB*100LR displayed no autokinase activity (Fig. 5C), con-sistent with residue L100 in the N-terminal cap of PAS beingrequired for MicB* autokinase activity.

We also analyzed phosphotransfer from MicB*-PO4 toMicA. Both wild-type MicA and mutant MicA59DA were rap-idly phosphorylated by MicB*-PO4 because MicB-PO4 was nolonger detected after incubation for 30 s in the presence ofMicA. The amount of MicA-PO4 increased with time from 30to 1,200 s whereas the amount of MicA59DA-PO4 remainedconstant (Fig. 6A). Both phosphorylated proteins were labilein alkali and stable in acid (Fig. 6B), which indicates that anaspartate residue was likely phosphorylated in both (1, 11).However, MicA59DA-PO4 had a half-life of 12 min whereas

MicA-PO4 had a half-life of over 20 min (Fig. 6C), indicatingthat the phosphorylated form of MicA59DA was unstabledue to the missense mutation. It is possible that the phosphor-ylated residue in MicA59DA-PO4 is labile. Alternatively, theMicA59DA mutation may induce conformational changes inthe protein resulting in higher suceptibility to phosphatases asdescribed for other response regulators (14, 30).

DISCUSSION

In S. pneumoniae, oxygen limitation induces competencerepression, thereby abolishing gene transfer by transformation.Previous work has provided evidence for an O2 signaling path-way encoded by nox, ciaRH, and comCDE (3, 7, 8). We providehere evidence that the autokinase MicB, which has a PASmotif, is involved in this regulation. Autophosphorylation ofMicB and phosphotransfer to the response regulator, MicA,demonstrated that they form a signal transduction system. Mu-tations which abolish the kinase activity of MicB or lower thestability of MicA-PO4 both relieve competence repression un-der microaerobiosis. Thus, one function for the kinase MicBand the essential response regulator MicA is the perception ofsignals produced in response to oxygen deprivation and theirtransduction, culminating in competence repression.

Recombinant MicB was autophosphorylated with [g-32P]ATP, and phosphotransfer to MicA was observed in vitro,providing the first biochemical evidence for signaling via phos-photransfer in S. pneumoniae. Mutational and biochemicalanalysis showed that residue L100 in the N-terminal cap of thePAS domain in recombinant MicB was important for C-termi-nal kinase activity in vitro. Although the N-terminal cap ofPAS is the less conserved part in this motif (22, 25), residueL100 in cap is important for the kinase activity of MicB in vitroand for competence repression under microaerobiosis in vivo.

Insertion mutations in the gene encoding the response reg-ulator, MicA, were not obtained in any genetic background(15, 26; this work), which suggests that this protein is essentialin S. pneumoniae. Instability of recombinant MicA59DA-PO4

compared to MicA-PO4 was related to competence expressionunder oxygen limitation in mutant strains expressing themicA59DA allele but did not significantly affect growth. Takenas a whole, these data show that mutations which impact thelevel of phosphorylated MicA, either by abolishing the kinaseactivity of MicB or by lowering the stability of MicA-PO4 invitro, allow accumulation of comCDE mRNA in bacteria grownunder microaerobiosis. This indicates that the essential re-sponse regulator MicA represses competence when phosphor-ylated by MicB under oxygen limitation. The TCS CiaRH is astrong repressor of competence development both in aerobic-and microaerobic-grown cultures (7), and the NADH oxidasebelongs to the competence signaling network (8). MicAB mightexert its control through CiaRH regulation or via a comple-mentary pathway. The signal recognized by the PAS domaincontrolling MicB phosphorylation remains to be characterized,as well as the MicA targets.

Oxic growth is required for induction of competence but notfor transformation of competent bacteria. Indeed, when cul-tures were incubated under oxygen limitation, bacteria alreadycommitted to competence (competent or precompetent) re-



.org/D

ownloaded from

http://jb.asm.org/

mained transformable in contrast to noncompetent bacteria,which grew without expressing any competence and thereforedid not transform. Occasionally, stored cultures from strainCp1015 grown aerobically at low pH (19) contain some com-petent bacteria which account for residual transformation ob-served under oxygen limitation (3 and data not shown). Al-though the oxygen-dependent checkpoint for competencedevelopment is not yet characterized, the contribution of theNADH oxidase, Nox, which has O2 as substrate, has beendemonstrated. Loss of function mutation in nox results in earlyand limited competence expression in cultures growing aero-bically (3–8).

For the purpose of this study we focused our screening ofmutant strains transformable under oxygen limitation andshowing similar growth characteristics to the wild type. How-ever, some mutated alleles of micA and micB were not toler-ated in pneumococcus whatever the oxygen status of the cul-ture. This identifies residues in MicA and also in MicB whichare probably crucial for bacterial growth.

The micAB genes are flanked by mutY and orfC. OrfC isimportant for the growth of strains 23F and 0100993 in vivo(26) and Cp1015 in vitro, regardless of ambient oxygen con-centration. OrfC shows the characteristic signature of a metalbinding protein (15) and sequence similarity to YycJ of S. au-

reus, Lactococcus lactis and Bacillus subtilis (9, 17, 20). InB. subtilis, insertion mutation in yycJ is silent (9), and this issuehas not been addressed for the yycJ of S. aureus (17). MutY ofS. pneumoniae was recently identified by computer-aided stud-ies and by complementation of a mutY mutant of E. coli (24).This antimutator gene in E. coli and S. pneumoniae encodes anadenine glycosylase specific for A/G miss-match pairs. The ma-jor in vivo substrate for this enzyme is probably the adeninefrom A/7,8-dihydro-8-oxoguanine, a product of oxidative dam-age to DNA (2, 16, 18). The MutY of E. coli has the highlyconserved stretch of four cysteines, Cys-X6-Cys-X2-Cys-X5-Cys, that coordinates the (4Fe-4S)21 cluster loop (FCL).These (4Fe-4S)21 clusters have been extensively described forproteins that sense the redox status of the cells (27). MutY ofS. pneumoniae has no FCL domain (24) and lies near micAB,a signaling system involved in the cellular response to oxygen.

In conclusion, this work provides a functional characteriza-tion of the essential signal-transducing two-component MicABsystem. Phosphotransfer through MicAB represses compe-tence under oxygen limitation. Although competence develop-ment during growth is not essential, competence regulation byMicAB might indicate nodes that are shared between the reg-ulatory networks controlling essential biological processes andcompetence development. MicAB orthologues have been foundin the transformable B. subtilis (9) and in bacteria in whichtransformation has never been described, such as L. lactis (20),and the pathogen S. aureus (17). When investigated by muta-tional analysis, the response regulator was shown to be essen-tial for bacteria growth and involved in cell division and mem-brane integrity. Corresponding mutant strains exhibited a morepronounced phenotype in aerobic cultures (9, 17). It is pro-posed as a hypothesis that MicAB and its orthologues areinvolved in the bacterial adaptation to oxygen, whatever themechanism specifically used by the different species. InS. pneumoniae, competence expression requires high oxygenlevels (7) and the NADH oxidase which has O2 as a secondsubstrate (3, 8). MicAB with its PAS signature contributes tocompetence regulation. Utilization of soluble DNA may havethe advantage of making it possible to correct oxidative dam-age to chromosomal DNA, in addition to other repair enzymessuch as MutY. Work is in progress to determine the role ofMicAB in the relationship between MutY and competenceregulation by oxygen in this pathogen.

FIG. 4. Purification of MicA and MicB* recombinant proteinsfrom E. coli extracts. Cultures of strain TOP10 of E. coli producing therecombinant pneumococcal proteins (see Fig. 1) were induced byIPTG and lysed, and the recombinant proteins were purified as de-scribed in Materials and Methods. The proteins were analyzed bySDS-PAGE, and the gels were stained with Coomassie blue. Lane M,molecular mass standard; lanes 1 and 6, total protein from IPTG-induced cultures of E. coli TOP10 containing pTrcHis2A; lanes 2, 4, 7,9, and 11, total protein from IPTG-induced bacteria containingpPMAE, pPMAE59, pPMBE1, pPMBE100, and pPMBEN, respec-tively; lanes 3, 5, 8, 10, and 12, purified MicA, MicA59DA, MicB*,MicB*100LR, and MicB*-N, respectively.

FIG. 5. MicB* autokinase activity lies in the C-terminal part of the protein and is under the control of PAS. MicB*, MicB*-N, and MicB*100LRwere autophosphosphorylated with [g-32P]ATP at 23°C in 48 ml of phosphorylation buffer (see Materials and Methods) containing 12 pmol ofpurified protein. At intervals, 16-ml aliquots were mixed with 23 loading buffer and run on SDS-12% PAGE gels for analysis as described inMaterials and Methods. (C) Control in which [a-32P]/ATP replaced [g-32P]/ATP in the reaction mixture containing MicB*.



.org/D

ownloaded from

http://jb.asm.org/

ACKNOWLEDGMENTS

This work was supported by Universite Paul Sabatier Toulouse andRhone-Poulenc Rorer, France. J.R.E. was supported by an RPR post-doctoral fellowship.

We thank Franck Pasta for discussion. We are grateful to DelphineDos Santos, Saliha Mimar, and Suzanne Eychenne for technical assis-tance. We thank the Technical Department of Institut Louis Bugnard/INSERM, Toulouse, for providing access to certain pieces of appara-tus. We thank the Institute for Genomic Research website at http://www.tigr.org for the S. pneumoniae genome sequence.

REFERENCES

1. Appleby, J. L., and R. B. Bourret. 1999. Activation of CheY mutant D57N byphosphorylation at an alternative site, Ser-56. Mol. Microbiol. 34:915–925.

2. Au, K. G., M. Cabrera, J. H. Miller, and P. Modrich. 1988. Escherichia colimutY gene product is required for specific A-G/C-G mismatch correction.Proc. Natl. Acad. Sci. USA 85:9163–9166.

3. Auzat, I., S. Chapuy-Regaud, G. Le Bras, D. Dos Santos, A. D. Ogunniyi,I. Le Thomas, J.-R. Garel, J. C. Paton, and M.-C. Trombe. 1999. The NADHoxidase of Streptococcus pneumoniae: its involvement in competence andvirulence. Mol. Microbiol. 34:1018–1028.

4. Bauer, C. E., S. Elsen, and T. H. Bird. 1999. Mechanisms for redox controlof gene expression. Annu. Rev. Microbiol. 53:495–523.

5. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye bind-ing. Anal. Biochem. 72:248–254.

6. Chandler, M. S., and D. A. Morrison. 1987. Competence for genetic trans-formation in Streptococcus pneumoniae: molecular cloning of com, a com-petence control locus. J. Bacteriol. 169:2005–2011.

7. Echenique, J. R., S. Chapuy-Regaud, and M.-C. Trombe. 2000. Competenceregulation by oxygen in Streptococcus pneumoniae: involvement of ciaRH andcomCDE. Mol. Microbiol. 36:688–696.

8. Echenique, J. R., and M.-C. Trombe. 2001. Competence modulation by theNADH oxidase of Streptococcus pneumoniae involves signal transduction.J. Bacteriol. 183:768–772.

9. Fabret, C., and J. A. Hoch. 1998. A two-component signal transductionsystem essential for growth of Bacillus subtilis: implications for anti-infectivetherapy. J. Bacteriol. 180:6375–6383.

10. Gil, D. 1990. Elaboration et caracterisation d’un nouveau type de vecteur declonage a nombre de copies regulable. These de l’Universite, Universite PaulSabatier, Toulouse, France.

11. Gilles-Gonzalez, M. A., G. S. Ditta, and D. R. Helinski. 1991. A haemopro-tein with kinase activity encoded by the oxygen sensor of Rhizobium meliloti.Nature 350:170–172.

12. Gong, W., B. Hao, S. S. Mansy, G. Gonzalez, M. A. Gilles-Gonzalez, andM. K. Chan. 1998. Structure of a biological oxygen sensor: a new mechanismfor heme-driven signal transduction. Proc. Natl. Acad. Sci. USA 95:15177–15182.

13. Havarstein, L. S. 1998. Identification of a competence regulon in Strepto-coccus pneumoniae by genomic analysis. Trends Microbiol. 6:297–299.

14. Hess, J. F., K. Oosawa, N. Kaplan, and M. I. Simon. 1988. Phosphorylationof three proteins in the signaling pathway of bacterial chemotaxis. Cell53:79–87.

FIG. 6. Phosphotransfer from MicB-PO4 to MicA and MicA59DA and properties of the phosphorylated proteins. (A) The MicA andMicA59DA proteins (9 mg corresponding to 300 pmol) were added to the MicB* phosphorylation mixture at 240 s, and aliquots were analyzedat intervals. Ct, control sample containing MicA in the absence of MicB*. (B) MicA-PO4 and MicA59DA-PO4 were incubated with 0.1 N HCl or0.1 N NaOH for 20 min at 23°C, and proteins were analyzed and compared to the control incubated with H2O. (C) MicA-PO4 and MicA59DA-PO4were incubated at 23°C in phosphorylation buffer supplemented with 10 mM ATP. Aliquots were analyzed at intervals. Proteins were subjectedto electrophoresis in 12% polyacrylamide gels. Signals on autoradiographs were quantified by densitometry (see Materials and Methods) and areexpressed in arbitrary units. The bars are aligned with the corresponding signals.



.org/D

ownloaded from

http://jb.asm.org/

15. Lange, R., C. Wagner, A. de Saizieu, N. Flint, J. Molnos, M. Stieger, P.Caspers, M. Kamber, W. Keck, and K. E. Amrein. 1999. Domain organiza-tion and molecular characterization of 13 two-component systems identifiedby genome sequencing of Streptococcus pneumoniae. Gene 237:223–234.

16. Lu, A. L., M. J. Cuipa, M. S. Ip, and W. G. Shanabruch. 1990. SpecificA/G-to-C.G mismatch repair in Salmonella typhimurium LT2 requires themutB gene product. J. Bacteriol. 172:1232–1240.

17. Martin, P. K., T. Li, D. Sun, D. P. Biek, and M. B. Schmid. 1999. Role in cellpermeability of an essential two-component system in Staphylococcus aureus.J. Bacteriol. 181:3666–3673.

18. Michaels, M. L., J. Tchou, A. P. Grollman, and J. H. Miller. 1992. A repairsystem for 8-oxo-7,8-dihydrodeoxyguanine. Biochemistry 31:10964–10968.

19. Morrison, D. A., M.-C., Trombe, M. K. Hayden, G. A. Waszak, and J. D.Chen. 1984. Isolation of transformation-deficient Streptococcus pneumoniaemutants defective in control of competence, using insertion-duplication mu-tagenesis with the erythromycin resistance determinant of pAM beta 1.J. Bacteriol. 159:870–876.

20. O’Connell-Motherway, M., D. van Sinderen, F. Morel-Deville, G. F. Fitzger-ald, S. D. Ehrlich, and P. Morel. 2000. Six putative two-component regula-tory systems isolated from Lactococcus lactis subsp. cremoris MG1363. Mi-crobiology 146:935–947.

21. Peeters, B. P., J. H. de Boer, S. Bron, and G. Venema. 1988. Structuralplasmid instability in Bacillus subtilis: effect of direct and inverted repeats.Mol. Gen. Genet. 212:450–458.

22. Pellequer, J. L., K. A. Wager-Smith, S. A. Kay, and E. D. Getzoff. 1998.

Photoactive yellow protein: a structural prototype for the three-dimensionalfold of the PAS domain superfamily. Proc. Natl. Acad. Sci. USA 95:5884–5890.

23. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

24. Samrakandi, M. M., and F. Pasta. 2000. Hyperrecombination in Streptococ-cus pneumoniae depends on an atypical mutY homologue. J. Bacteriol. 182:3353–3360.

25. Taylor, B. L., and I. B. Zhulin. 1999. PAS domains: internal sensors ofoxygen, redox potential, and light. Microbiol. Mol. Biol. Rev. 63:479–506.

26. Throup, J. P., K. K. Koretke, A. P. Bryant, K. A. Ingraham, A. F. Chalker,Y. Ge, A. Marra, N. G. Wallis, J. R. Brown, D. J. Holmes, M. Rosenberg, andM. K. Burnham. 2000. A genomic analysis of two-component signal trans-duction in Streptococcus pneumoniae. Mol. Microbiol. 35:566–576.

27. Unden, G., and J. Schirawski. 1997. The oxygen-responsive transcriptionalregulator FNR of Escherichia coli: the search for signals and reactions. Mol.Microbiol. 25:205–210.

28. Volz, K. 1993. Structural conservation in the CheY superfamily. Biochemis-try 32:11741–11753.

29. Wang, R. F., and S. R. Kushner. 1991. Construction of versatile low-copy-number vectors for cloning, sequencing and gene expression in Escherichiacoli. Gene 100:195–199.

30. Weiss, V., and B. Magasanik. 1988. Phosphorylation of nitrogen regulator I(NRI) of Escherichia coli. Proc. Natl. Acad. Sci. USA 85:8919–8923.



.org/D

ownloaded from

http://jb.asm.org/

competence repression under oxygen limitation through the ... · ppt14 0.98-kb xbai/xhoi amplicon...

Documents