Identification of AcnR, a TetR-type Repressor of the Aconitase Gene acn in Corynebacterium

12 Pages·2004·0.89 MB·English

THEJOURNALOFBIOLOGICALCHEMISTRY Vol.280,No.1,IssueofJanuary7,pp.585–595,2005
©2005byTheAmericanSocietyforBiochemistryandMolecularBiology,Inc. PrintedinU.S.A.
Identification of AcnR, a TetR-type Repressor of the
Aconitase Gene acn in Corynebacterium glutamicum*
Receivedforpublication,July21,2004,andinrevisedform,October5,2004
Published,JBCPapersinPress,October19,2004,DOI10.1074/jbc.M408271200
AndreasKrug,VolkerF.Wendisch,andMichaelBott‡
FromtheInstitutfu¨rBiotechnologie1,ForschungszentrumJu¨lich,D-52425Ju¨lich,Germany
InCorynebacteriumglutamicum,theactivityofacon- theaspartateandglutamatefamily.Thus,itisnotsurprising
itase is 2.5–4-fold higher on propionate, citrate, or ace- thatseveraltricarboxylicacidcycleenzymesofC.glutamicum
tatethanonglucose.Hereweshowthatthisvariationis havebeenstudiedbiochemicallyand/orgeneticallyinthepast,
caused by transcriptional regulation. In search for pu- i.e. citrate synthase (gltA, Ref. 7), isocitrate dehydrogenase
tative regulators, a gene (acnR) encoding a TetR-type (icd,Ref.8),2-oxoglutaratedehydrogenase(9),malate:quinone
transcriptional regulator was found to be encoded im- oxidoreductase (mqo, Ref. 10), and malate dehydrogenase
mediately downstream of the aconitase gene (acn) in (mdh, Ref. 11). However, no information is available hitherto
C.glutamicum.DeletionoftheacnRgeneledtoa5-fold on the genetic regulation of this pathway in C. glutamicum,
increased acn-mRNA level and a 5-fold increased acon-
althoughthereisevidencethatitdiffersfromthatofthemodel
itaseactivity,suggestingthatAcnRfunctionsasrepres-
bacteria Escherichia coli and Bacillus subtilis, since e.g.
sor of acn expression. DNA microarray analyses indi-
glucose and acetate are consumed in parallel rather than
catedthatacnistheprimarytargetgeneofAcnRinthe
successively(12).
C.glutamicumgenome.PurifiedAcnRwasshowntobe
No studies have been performed yet in C. glutamicum on
a homodimer, which binds to the acn promoter in the
region from (cid:1)11 to (cid:1)28 relative to the transcription aconitase (EC 4.2.1.3), which catalyzes the stereospecific and
reversibleisomerizationofcitratetoisocitrateviacis-aconitate
start. It thus presumably acts by interfering with the
inthetricarboxylicacidcycleandintheglyoxylatecycle.Itis
bindingofRNApolymerase.Theacn-acnRorganization
is conserved in all corynebacteria and mycobacteria anunusualenzymeinthatitcontainsa[4Fe-4S]cluster,which
withknowngenomesequenceandaputativeAcnRcon- isnotinvolvedinelectrontransfer,butinbindingofthesub-
sensusbindingmotif(CAGNACnnncGTACTG)wasiden- strate (13). Besides their catalytic function, a certain class of
tifiedinthecorrespondingacnupstreamregions.Muta- aconitases can also have a regulatory function by binding to
tionswithinthismotifinhibitedAcnRbinding.Because certain mRNAs and inhibiting or increasing their translation
theactivitiesofcitratesynthaseandisocitratedehydro- intoprotein(14–18).Thisregulatoryfunctioniscarriedoutby
genase were previously reported not to be increased acatalyticallyinactiveformofaconitase,whichisformedun-
duringgrowthonacetate,ourdataindicatethataconi- derconditionsofironstarvationoroxidativestress,whenthe
taseisamajorcontrolpointoftricarboxylicacidcycle iron-sulfur cluster is disassembled and the apoprotein is
activityinC.glutamicum,andtheyidentifyAcnRasthe formed.
first transcriptional regulator of a tricarboxylic acid In many bacteria, expression of the aconitase genes is con-
cyclegeneintheCorynebacterianeae. trolled by transcriptional regulators (see “Discussion”). Here,
we provide evidence that this is also the case for the C. glu-
tamicumaconitasegene(acn)1anddescribetheidentification
Corynebacterium glutamicum is a non-pathogenic, aerobic andcharacterizationoftheTetR-typerepressorproteinAcnR.
Gram-positivesoilbacteriumthatwasdescribedin1957(1)as
anL-glutamate-excretingbacterium.Ithasgainedconsiderable EXPERIMENTALPROCEDURES
interest because of its use in the large scale biotechnological BacterialStrainsandCultureConditions—Allstrainsandplasmids
plyrsoidnuect(i(cid:2)on0.6ofmLi-lglliuontamtoantsep(e(cid:1)r1ymeairll)io(n2–t5o)nasnpderbeyceaaurs)eaonfditLs- AusTeCdC1in303th2iswawsourksedaraeslwisitldedtyipne iTnabthleisIs.tuCd.y.gSlutrtaaminic(cid:3)uamcnRstriasina
derivativecontaininganin-framedeletionoftheacnRgene.Forculti-
emergingroleasamodelorganismfortheCorynebacterineae,a
vationofC.glutamicuminliquidmedia,5mlofCGIIImedium(19)or
suborder of the actinomycetes, which also includes the genus brain heart infusion (BHI) medium (Difco Laboratories, Detroit, MI),
Mycobacterium(6). bothsupplementedwith2%(w/v)glucose,wereinoculatedwithcolonies
The tricarboxylic acid cycle is of central importance for the fromafreshLuria-Bertani(LB)agarplate(20)andincubatedovernight
metabolism of C. glutamicum because it provides energy and at30°Cand170rpm.Thisfirstpreculturewasusedtoinoculate60ml
ofCGXIIminimalmedium(21)ina500-mlshakeflaskwithtwobaffles.
biosynthetic precursors and therefore the flux through this
Thesecondpreculturewasincubatedovernightat30°Cand120rpm
cycleisanimportantaspectfortheproductionofaminoacidsof
andthenusedtoinoculatethemainculture,forwhichthesamecon-
ditionswereappliedasforthesecondpreculture.TheCGXIIminimal
mediumcontainedasacarbonsourceeither222mMglucose,244mM
*ThisworkwassupportedbytheEuropeanUnionwithintheframe- sodiumacetate,50mMsodiumcitrate,or104mMsodiumpropionate.
workoftheVALPANproject(QLK3-2000-00497).Thecostsofpubli- When citrate was used as carbon source, the medium contained in
cation of this article were defrayed in part by the payment of page addition 100 mM MgCl2. When appropriate, the medium was supple-
charges.Thisarticlemustthereforebeherebymarked“advertisement” mentedwith25(cid:1)gofkanamycin/ml.C.glutamicumstrainscarrying
inaccordancewith18U.S.C.Section1734solelytoindicatethisfact. plasmidpEKEx2andderivativesthereofwerecultivatedinthepres-
ThisarticleisdedicatedtoProf.RudolfK.Thauerontheoccasionof
his65thbirthday.
‡Towhomcorrespondenceshouldbeaddressed:Institutfu¨rBiotech- 1Theabbreviationsusedare:acn, C.glutamicumaconitasegene;
nologie1,ForschungszentrumJu¨lich,D-52425Ju¨lich,Germany.Tel.: RT-PCR, reverse transcriptase-PCR; AcnR, TetR-type repressor
49-2461-61-5515;Fax:[phone];E-mail:[email protected]. protein.
Thispaperisavailableonlineathttp://www.jbc.org 585
This is an Open Access article under the CC BY license.
586 AcnR, a Repressor of the Aconitase Gene in Corynebacterium
TABLE I
Bacterialstrainsandplasmidsusedinthisstudy
Strainsorplasmids Relevantcharacteristics SourceorRef.
Strains
C.glutamicum
ATCC13032 Biotin-auxotrophicwildtype (1)
13032(cid:3)acnR In-framedeletionoftheacnRgene Thiswork
E.coli
DH5(cid:3) supE44(cid:3)lacU169((cid:1)80lacZ(cid:3)M15)hsdR17recA1endA1gyrA96thi-1relA1 Invitrogen
BL21(DE3) ompThsdS (r (cid:4)m (cid:4))galdcm(DE3) (22)Novagen
B B B
Plasmids
pK19mobsacB KanR;vectorforallelicexchangeinC.glutamicum;(pK18oriV ,sacB,lacZ(cid:3)) (27)
pK19mobsacB-(cid:3)acnR KanR;pK19mobsacBderivativecontainingacrossoverPCRproEd.cu.ctcoveringtheup- Thiswork
anddownstreamregionsofacnR
pET24b KanR;vectorforoverexpressionofgenesinE.coli,addingaC-terminal Novagen
hexahistidineaffinitytagtothesynthesizedprotein(pBR322oriV ,P ,lacI)
E.c. T7
pET24b-Streptag KanR;pET24bderivativeaddingaC-terminalStrepTag-IItothesynthesizedprotein MeyerandBotta
pET24b-acnR-C KanR;pET24b-Streptagderivativeforover-productionofAcnRwithaC-terminal Thisstudy
StrepTag-II
pET28a KanR;vectorforoverexpressionofgenesinE.coli,addinganN-terminaloraC- Novagen
terminalhexahistidinetagtothesynthesizedprotein(pBR322oriV ,P ,lacI)
E.c. T7
pET28a-Streptag KanR;pET28aderivativeaddinganN-terminalStrepTag-IItothesynthesized (29)
protein
pET28a-acnR-N KanR;pET28a-Streptagderivativeforover-productionofAcnRwithanN-terminal Thisstudy
StrepTag-II
pEKEx2 KanR;C.glutamicum/E.colishuttlevectorforregulatedgeneexpression(P ,lacIQ, (63)
tac
pBL1oriV ,pUC18oriV )
C.g. E.c.
pEKEx2-acnR KanR;pEKEx2derivativecontainingtheacnRgenefromC.glutamicumunderthe Thisstudy
controlofthetacpromoter
pEKEx2-acnR60 KanR;pEKEx2derivativecontainingashortenedversionoftheacnRgene(first180 Thisstudy
bp)fromC.glutamicumunderthecontrolofthetacpromoter
pEKEx2-acnR89 KanR;pEKEx2derivativecontainingashortenedversionoftheacnRgene(first267 Thisstudy
bp)fromC.glutamicumunderthecontrolofthetacpromoter
pEKEx2-acnR-C-Strep KanR;pEKEx2derivativecontainingtheacnRgenewithaC-terminalStrepTag-II Thisstudy
fromC.glutamicumunderthecontrolofthetacpromoter
aUnpublisheddata.
enceof1mMisopropyl-1-thio-(cid:2)-D-galactopyranoside.Forcloningpur- productof(cid:2)1kbwasdigestedwithEcoRIandHindIIIandclonedinto
poses,E.coliDH5(cid:3)(InvitrogenLifeTechnologies,Inc.)wasusedand pK19mobsacB(27)cutwiththesameenzymes.DNAsequenceanalysis
grownat37°CinLBmedium(20).E.coliBL21(DE3)(22)wasusedfor confirmedthattheclonedPCRproductdidnotcontainspuriousmuta-
overproduction of the C. glutamicum AcnR protein with expression tions. Transfer of the resulting plasmid pK19mobsacB-(cid:3)acnR into
plasmidsbasedonpET24orpET28(Novagen,Madison,WI). C. glutamicum and screening for the first and second recombination
Determination of Aconitase Activity—For the determination of the eventwasperformedasdescribedpreviously(26).Thegenomicstruc-
aconitase activity, cells from the CGXII main culture ((cid:2)25 ml) were tureofthemutantwasfirstcheckedbyPCRanalysisofchromosomal
harvestedwith(cid:2)25gofcrushedice(precooledto(cid:4)20°C)bycentrifu- DNA with the primer pair tetR-amp-for/tetR-amp-rev (Table II). The
gationat4,000(cid:5)gfor5min.Thecellpelletwasresuspendedin900(cid:1)l PCRproductsobtainedwithwild-typeDNA(1.6kb)and(cid:3)acnRmutant
ofTris-HCl(90mM;pH8.0),andthecellsweremechanicallydisrupted DNA (1.1 kb) had the expected sizes. Furthermore, the mutant was
by3(cid:5)20sbeadbeatingwith1gofzirconia-silicabeads(diameter0.1mm; controlled by Southern blot analysis (26) using EcoRI-digested DNA
Roth, Karlsruhe, Germany) using a Silamat S5 (Vivadent, Ellwangen, andadigoxigenin-labeledPCRproductcoveringtheacnRgeneandthe
Germany).Aftercentrifugation(5min;18,320(cid:5)g;4°C),thesupernatant 500-bpup-anddownstreamDNAasprobe.Theexpectedfragmentsizes
wasusedimmediatelyfortheenzymeassay.Aconitaseactivitywasas- were obtained for wild type (1.8 kb) and the (cid:3)acnR mutant (1.3 kb),
sayedbyfollowingtheformationofcis-aconitatefromisocitrate(23)at confirmingthesuccessfulacnRdeletion(datanotshown).
20°C.Theassaymixturecontained950–995(cid:1)lof90mMTris-HClpH8.0 ForthepurificationofAcnRwithaC-terminalStrepTag-II(28),the
containing20mMDL-trisodiumisocitrate.Thereactionwasstartedbythe acnR coding region was amplified using the Expand High Fidelity
additionof5–50(cid:1)lofcellextract,andcis-aconitateformationwasdeter- polymerase mix (Roche Diagnostics) and oligonucleotides that intro-
minedbymeasuringtheabsorbanceincreaseat240nmusingaJasco ducedanNdeIrestrictionsiteoverlappingthestartcodon(acnR-FP-for)
V560spectrophotometer.Anextinctioncoefficientforcis-aconitateof3.6 andanXhoIrestrictionsiteprecedingthestopcodon(acnR-FP-24rev).
mM(cid:4)1cm(cid:4)1at240nmwasused.Oneunitofactivitycorrespondsto1 The purified PCR product was cloned into the expression vector
(cid:1)molofisocitrateconvertedtocis-aconitatepermin. pET24b-StreptagcutwithNdeIandXhoI.TheAcnRproteinencodedby
RecombinantDNAWork—TheenzymesforrecombinantDNAwork the pET24 derivative (AcnR-C) contains ten additional amino acid
wereobtainedfromRocheDiagnostics(Mannheim,Germany)orNew residues (LEWSHPQFEK) at the C terminus. For the purification of
England Biolabs (Frankfurt, Germany). The oligonucleotides used in AcnR with an N-terminal StrepTag-II, the acnR coding region was
thisstudywereobtainedfromMWGBiotech(Ebersberg,Germany)and amplified with the primers acnR-FP-for and acnR-FP-28rev, and the
arelistedinTableII.RoutinemethodslikePCR,restriction,orligation PCRproductwasclonedintopET28a-Streptag(29)cutwithNdeIand
were carried out according to standard protocols (20). Chromosomal XhoI. The AcnR protein encoded by the pET28 derivative (AcnR-N)
DNAfromC.glutamicumwaspreparedasdescribed(7).Plasmidsfrom contains14additionalaminoacidresidues(MASWSHPQFEKGAH)at
E. coli were isolated with the QIAprep spin miniprep kit (Qiagen, theNterminus.DNAsequenceanalysisconfirmedthattheAcnRcod-
Hilden,Germany).E.coliwastransformedbytheRbClmethod(24), ing sequence of the resulting plasmids pET24b-acnR-C and pET28a-
C.glutamicumbyelectroporation(25).DNAsequencingwasperformed acnR-Ndidnotcontainspuriousmutations.Fortheoverproductionof
with a Licor 4200 sequencer (Licor Inc., Lincoln, NB). Sequencing thestreptaggedAcnRderivatives,theplasmidsweretransferredinto
reactions were carried out with the Thermo Sequenase Primer Cycle E.coliBL21(DE3).
SequencingKit(AmershamBiosciences,Freiburg,Germany). In order to construct plasmids pEKEx2-acnR, pEKEx2-acnR60,
Anin-frameacnRdeletionmutantofC.glutamicumwasconstructed pEKEx2-acnR89,andpEKEx2-acnR-C-Strep,theacnRfragmentswere
viaatwo-stephomologousrecombinationprocedureasdescribedpre- amplified with oligonucleotide acnR-for as forward primer and acnR-
viously (26). The acnR up- and downstream regions ((cid:2)500 bp each) rev, acnR-rev-60, acnR-rev-89, or acnR-strep-rev, respectively, as re-
were amplified using the oligonucleotide pairs tetR-A-for/tetR-B-rev verseprimers.AfterdigestionwithBamHIandEcoRI,thepurifiedPCR
and tetR-C-for/tetR-D-rev, respectively, followed by a crossover PCR products were cloned into the vector pEKEx2 cut with the same en-
with oligonucleotides tetR-A-for and tetR-D-rev. The resulting PCR zymes.AllPCR-derivedpartsoftheresultingplasmidswerecheckedby
AcnR, a Repressor of the Aconitase Gene in Corynebacterium 587
TABLE II
Oligonucleotidesusedinthisstudy
Oligonucleotide Sequence(5(cid:6)33(cid:6))andpropertiesa
tetR-A-for TAT ATA GAA TTC TGC AGT TCC CTG CAG GCG AAT C (EcoRI)
tetR-B-rev CCC ATC CAC TAA ACT TAA ACA GCC TGC CGC TAC GGA CAC AAT G
tetR-C-for TGT TTA AGT TTA GTG GAT GGG GAT CTG GTC GAG GGA ACT GTC CG
tetR-D-rev TAT ATA AAG CTT GTT AGC AAC ACT GTG GTG GCG CC (HindIII)
tetR-amp-for CTA AGG GCA CTA ACC TGC TCG G
tetR-amp-rev TTG CCA ACA CCA CGT GTC CAG G
acnR-FP-for TAT ATA CAT ATG TCC GTA GCG GCA GGC GAC AAA C (NdeI)
acnR-FP28-rev TAT ATA CTC GAG TTA GTC GCG TTT ACG GAC AGT TC (XhoI)
acnR-FP24-rev TAT ATA CTC GAG GTC GCG TTT ACG GAC AGT TC (XhoI)
acn-PEK-for CCT TCA TTA GTG TCG GGC TCA CG
acn-PEK-rev GGA TTT CGA TGC TTG GAT CGG AAG
acn-PE1* IRD800-GGG TGC TCT TAG CAT TGA AGG AG
acn-PE3* IRD800-GTG AGA TTT CGC GAC GGC GTC TG
acn-FP1* IRD800-GAA AGT CAC ATC ACG CAC GTA CC
RT-1-for CAG GCG ACA AAC CAA CAA ATA GCC
RT-1-rev CTC GAC CAG ATC CAA TAC TTC GG
RT-2-for CGA GAC TCC TAA GAC TGT CAA GG
RT-2-rev GTT GCT TCT TCC AGT CGG CGT AC
RT-3-for CAG AAC CAG CTG GTT GAC ATC GC
RT-3-rev CAT CTG ACG CAG CAC GTA CTG C
acnR-for TAT ATA GGA TCC AAG GAG ATA TAG ATG TGT CCG TAG CGG CAG GCG AC (BamHI)
acnR-rev TAT ATA GAA TTC GTC ACG TCG CTC AGC TGC GGC (EcoRI)
acnR-strep-rev TAT ATA GAA TTC TTA CTT CTC GAA CTG CGG GTG GCT CCA CTC GAG GTC GCG TTT ACG GAC AGT TCC
(EcoRI)
acnR-rev-60 TAT ATA GAA TTC TAG GAA CAG GTT TTC TTT GTC ACC G (EcoRI)
acnR-rev-89 TAT ATA GAA TTC ATC TTC CAG CAT TCC TCG CAT C (EcoRI)
RT-dnaE-fw TGC CCT TCC GGC GAT GTG CAA
RT-dnaE-rv CTG GAA CCA TGT CGT CCC AGA G
BS-1-for TCC GAC AAA ACC GCT GCC TAG G
BS-1-rev GGC GTC TGC TTC GGC TGA GCC
BS-2-for AGA CGT AGG GTC CTT TTC CAC AG
BS-2-rev TCA TTA TCC TAA CAG TAC AAG CGT T
BS-3-for CAA CTT TCC CGC CAG AAC GCT TG
BS-3-rev GGG TTC TTG GCG CGT CAA TAA CG
BS-4*-for CAA CTT TCC CGC CAG ATG GCT TCA ACT GTT AGG A
BS-5*-for CAA CTT TCC CGC GTG AAC GCT TGT ACA CTT AGG ATA AT
BS-6*-for CAA CTT TCG GGC CAG AAC GCT TGT ACT GTT TCG ATA ATG AAG
BS-7*-for CAA CTT TCC CGC CAG AAC CGA AGT ACT GTT AG
aIn some cases oligonucleotides were designed to introduce recognition sites for restriction endonucleases (recognition sites underlined,
restrictionendonucleasesindicatedinparentheses)orcomplementary21mersequencesforgeneratingcrossoverPCRproducts(printedinitalics).
TABLE III
Influenceofthecarbonsourceontheaconitaseactivityof
C.glutamicumwildtypeandthe(cid:3)acnRmutantstrain
C.glutamicumwildtypeand(cid:3)acnRmutantweregrowninCGXII
minimal medium containing either 222 mM glucose, 50 mM sodium
citrate((cid:7)100mMMgCl2),244mMsodiumacetate,or104mMsodium
propionate as carbon source. In the exponential growth phase at an
OD ofabout5,cellswereharvested,washed,anddisruptedbybead
bea6t0in0g.Aftercentrifugationat15,000(cid:5)g,thesupernatant(cellex-
tract)wasimmediatelyusedfortheaconitaseassay.Aconitaseactivity
wasdeterminedwithDL-trisodiumisocitrateassubstratebymeasuring
theformationofcis-aconitateat240nm.
Aconitaseactivity
Carbonsource
13032wildtype 13032(cid:3)acnR
units/mgofprotein
Glucose 0.20(cid:8)0.05 1.04(cid:8)0.06
Citrate 0.53(cid:8)0.08 1.78(cid:8)0.40
Acetate 0.82(cid:8)0.08 1.54(cid:8)0.08
Propionate 0.48(cid:8)0.13 2.38(cid:8)0.21
DNAsequenceanalysisinordertoexcludeunwantedmutations. FIG.1.GenomiclocusofacnfromC.glutamicumandother
PreparationofTotalRNA—Culturesofthewildtypeandthe(cid:3)acnR species of the suborder Corynebacterineae. In all cases except
Mycobacterium leprae, a gene encoding a transcriptional repressor
mutant were grown in CGXII minimal medium containing 4% (w/v)
was identified downstream of acn. This gene (shown in dark gray)
glucose.IntheexponentialgrowthphaseatanOD600of5–6,25mlof wasnamedacnR.InthecaseofRhodococcussp.,onlythe3(cid:6)-terminal
thecultureswereusedforthepreparationoftotalRNAasdescribed
partoftheacngenewasavailable.Othergeneswerenumberedand
previously(30,31).IsolatedRNAsampleswereanalyzedforquantity annotatedasfollows:1,invasin;2,3,4,and8,hypotheticalproteins;
and quality by UV spectrophotometry and denaturing formaldehyde 5, hypothetical cytosolic protein; 6, putative GMP synthase-gluta-
agarose gel electrophoresis (20), respectively, and stored at (cid:4)70°C mine amidotransferase domain; 7, ABC transporter, ATP-binding
untiluse. protein; 9, putative NAD(cid:7)-dependent dehydrogenase. Data were
DNA Microarray Analyses—The generation of whole genome DNA taken from the National Center for Biotechnology Information
microarrays (32), synthesis of fluorescent-labeled cDNA from total (NCBI) and the bioinformatics software ERGO (Integrated
RNA,microarrayhybridization,washing,anddataanalysiswereper- Genomics).
588 AcnR, a Repressor of the Aconitase Gene in Corynebacterium
FIG.2.SequencealignmentofAcnRproteinsfromdifferentCorynebacterineaespecies.Aminoacidsidenticalinallsequencesare
shadedinblack,otherconservedaminoacidsingray.TheN-terminalregionoftheAcnRproteins(aminoacids16–62)showsclearsimilarityto
theN-terminalpartoftheTetRfamilyofbacterialregulatoryproteins(TetR-N;PFAM00440).TheconsensussequenceofTetR-Nandtheamino
acidsidenticaltotheconsensusintheAcnRproteinsareshownabovethealignment.
formed as described previously (33–35). Genes that exhibited signifi- toapparenthomogeneity.Thesameprocedureasdescribedabovewas
cantlychangedmRNAlevels(p(cid:9)0.05inaStudent’sttest)byatleast used for the isolation of AcnR with an N-terminal StrepTag-II using
a factor of three were determined in four series of DNA microarray E.coliBL21(DE3)/pET28a-acnR-N.
experiments:(i)10comparisonsofthewildtypeandthe(cid:3)acnRmutant Size Exclusion Chromatography—The size of purified AcnR-C and
cultivated in regular CGXII minimal medium with 4% (w/v) glucose AcnR-NwasestimatedbysizeexclusionchromatographyusingaHi-
(contains36(cid:1)Mironbeforeautoclavingand(cid:2)8(cid:1)Mironafterautoclav- Load26/60Superdex200prepgradecolumn(AmershamBiosciences)
ing); (ii) four comparisons of the wild type and the (cid:3)acnR mutant integratedintoanA¨ktaExplorersystem(AmershamBiosciences).The
cultivatedinCGXII-glucosemediumwithanelevatedironconcentra- columnwasequilibratedwith20mMHEPES-bufferpH8.0containing
tion (addition of 100 or 500 (cid:1)M FeSO4 after autoclaving); (iii) three 300mMNaCland1mMdithiothreitol.Afterapplicationof0.5–1mgof
comparisonsofthewildtypegrownwith(cid:2)8(cid:1)Mironandwith508(cid:1)M purified protein, elution was performed at 4°C with a flow rate of 2
iron; (iv) three comparisons of the(cid:3)acnR mutant grown with(cid:2)8 (cid:1)M ml/min.Thecolumnwascalibratedwithapremixedproteinmolecular
iron and with 508 (cid:1)M iron. The 34 genes that were differentially massmarker(MWGF-200,Sigma).
expressedinoneormoreoftheseseriesofexperimentsweresubjected DNase I Footprinting Assays—The binding site of purified AcnR
toahierarchicalclusteranalysis(36)asdescribedpreviously(33,35). proteinattheacnpromoterwasanalyzedbyDNaseIfootprintingas
PrimerExtensionAnalysis—Non-radioactiveprimerextensionanal- described previously (29). PCR products covering the acn promotor
ysis was performed using the IRD800-labeled oligonucleotides acn- region from position (cid:4)346 to (cid:7)52 or from position (cid:4)268 to (cid:7)241
PE1* and acn-PE3* (Table II) and 15 (cid:1)g of total RNA as described relativetotheproposedTTGstartcodonwereobtainedwiththeprimer
previously(29).Thelengthoftheprimerextensionproductswasdeter- pairs acn-PEK-for/acn-PE1* (labeled non-template strand) and acn-
minedbyrunningthefourlanesofaDNAsequencingreactionsetup FP1*/acn-PEK-rev(labeledtemplatestrand)respectively.Theprimers
withthesameoligonucleotideasusedforprimerextensionalongside with an asterisk were IRD800-labeled at the 5(cid:6)-end. The DNA-AcnR
theprimerextensionproductsonthedenaturingpolyacrylamidegel.A mixturesweretreatedwithDNaseI,andthereactionproductswere
PCRproductobtainedwiththeprimerpairacn-PEK-fw/acn-PEK-rev separated by denaturing PAGE. The regions, which were protected
wasusedastemplatefortheDNAsequencingreaction. fromDNaseIdigestion,werelocalizedbyrunningbesidesthefootprint-
Overproduction and Purification of Streptagged AcnR—E. coli ingreactionsthefourlanesofaDNAsequencingreactionsetupwith
BL21(DE3) carrying pET24b-acnR-C was cultivated in 100 ml of LB theoligonucleotideacn-PE1*oracn-FP1*andaPCRproductobtained
mediumcontaining50(cid:1)g/mlkanamycinina500-mlErlenmeyerflask withtheprimerpairacn-PEK-for/acn-PEK-revastemplate.
at30°CuntiltheOD600reachedavaluebetween0.3and0.5.Synthesis GelShiftAssays—PurifiedAcnR-C(500nM)wasmixedwith(cid:2)50nM
oftheAcnR-Cwasinducedbyadditionofisopropyl-1-thio-(cid:2)-D-galacto- DNAfragments(80–300bp)coveringtheacnpromoterregioninatotal
pyranosidetoafinalconcentrationof1mM,andtheculturewasincu- volumeof20(cid:1)l.Thebuffercontained10mMTris-HClpH7.5,0.5mM
bated for another 3 h. Cells (final OD600 (cid:2)1) were washed once and EDTA,5%(v/v)glycerol,0.5mMdithiothreitol,0.005%(v/v)TritonX-100,
resuspended in 10 ml of buffer W (100 mM Tris-HCl, pH 8.0, 1 mM 50mMNaCl,5mMMgCl2,and2.5mMCaCl2.Afterincubationfor20min
EDTA). After addition of 1 mM diisopropylfluorophosphate and 1 mM atroomtemperature,thesampleswereseparatedbyagarosegelelectro-
phenylmethylsulfonyl fluoride, the cell suspension was passed three phoresiswithTrisacetate/EDTAbuffer(20)at4°Cand75V.Thegelwas
timesthroughaFrenchpressurecell(SLMAminco,SpectronicInstru- subsequentlystainedwithethidiumbromideandphotographed.
ments,Rochester)at207MPa.Intactcellsandcelldebriswereremoved
bycentrifugation(20min,5,000(cid:5)g,4°C).Thecell-freeextractwas RESULTS
subjectedtoultracentrifugation(1h,150,000(cid:5)g,4°C)andthesuper- Influence of the Carbon Source on Aconitase Activity of
natantappliedtoaStrepTactinSepharosecolumnwithabedvolumeof
C. glutamicum—Previous studies revealed that in C. glutami-
1ml(IBA,Go¨ttingen,Germany).Thecolumnwaswashedwith15mlof
bufferWandstreptaggedAcnRwaselutedwith10(cid:5)1mlbufferW cum the carbon flux through the tricarboxylic acid cycle varies
containing15mMdesthiobiotin(Sigma-Aldrich).Fractionscontaining with the carbon source (12). In cells growing on acetate, the
AcnR-Cwerepooled,andtheelutionbufferwasexchangedagainst40 tricarboxylicacidcyclefluxwasabout0.4(cid:1)molmin(cid:4)1 mgpro-
mMTris-HClbuffer,pH7.5,containing10%(v/v)glycerolbygelfiltra- tein(cid:4)1 and thus 4-fold higher than in cells growing on glucose.
tionwithSephadexG-25(PD-10column,AmershamBiosciences).Pro- Remarkably,theactivitiesofcitratesynthase(0.5–0.8units(mg
tein concentrations were determined with the BCA protein assay kit ofprotein)(cid:4)1;Ref.7)andisocitratedehydrogenase(0.9–1.1units
(Pierce) using bovine serum albumin as standard. The purity of the (mg of protein)(cid:4)1; Ref. 8) as measured in cell extracts were
protein preparation was assessed by SDS-polyacrylamide gel electro-
phoresis(37)andsubsequentproteindetectionwithGelCodebluestain reportedtobesimilaringlucose-andacetate-growncellsandare
reagent(Pierce).Usingthisprotocol,about1mgofAcnR-Cwaspurified sufficient to allow the maximally measured carbon fluxes. We
AcnR, a Repressor of the Aconitase Gene in Corynebacterium 589
FIG.3.Transciptionalorganizationoftheacn-acnRlocusinC.glutamicumanalyzedbyRT-PCR.A,schemeshowingtheacn-acnR
intergenicregioncontainingtwoputativestemloopstructures.BelowtheRT-PCR,reactionsusedtodetermineco-transcriptionofacnandacnR
areshownschematically.RNAfromC.glutamicumwildtypewastranscribedintocDNAwiththreedifferentprimersinthereversetranscriptase
reactionsA–C.AfterwardthesecDNAswereusedastemplatesforthePCRreactionslabeled1–5.B,resultsfromtheRT-PCRanalysesdescribed
above.ThelowerDNAfragmentvisibleinlanes1–5representsdnaE,andRT-PCRofdnaEservedaspositivecontrolinallreactions.Theupper
bandscorrespondtotheproductsofthePCRreactions1–5indicatedinA.Reactions6–10representcontrolsconfirmingtheabsenceofDNAinthe
RNA preparation. The reactions were identical to the PCR reactions as shown in lanes 1–5 except that reverse transcriptase was omitted in
reactions(A–C).
thereforebecameinterestedtodeterminewhethertheactivityof dratase(PrpD)andaconitase(AcnB)(40).Thiscouldexplainthe
aconitasedoesordoesnotvarywithdifferentcarbonsourcesand increasedaconitaseactivityinC.glutamicumduringgrowthon
measured its activity in extracts of cells cultivated on glucose, propionate.
citrate,acetate,orpropionate.AsshowninTableIII,theaconi- Identification of a Transcriptional Repressor Gene Down-
taseactivityofglucose-growncellswas0.2units/mgofprotein, streamoftheAconitaseGene—SincepreviousDNAmicroarray
whereas the activity of citrate-, acetate-, and propionate-grown experiments had shown that the mRNA level of acn is 4-fold
cellswas0.5,0.8,and0.5units/mgprotein,respectively,i.e.upto higher in acetate-grown cells (41), we assumed that the acn
4-fold higher. Thus, aconitase activity is regulated and might geneistranscriptionallyregulated.Asinbacteriagenesencod-
representarate-limitingstepinthetricarboxylicacidcycleflux. ingtranscriptionalregulatorsareoftenlocatedinthevicinity
The increased activity on acetate correlates with the increased oftheirtargetgenes,weinspectedtheC.glutamicumgenome
tricarboxylic acid cycle flux on this carbon source (12), which region encoding aconitase. In contrast to e.g. E. coli, C. glu-
might also occur with citrate. Propionate is metabolized in C. tamicumcontainsonlyasingleaconitasegene(42)encodinga
glutamicum via the methylcitrate cycle (38), in the course of proteinof943aminoacidresidueswithacalculatedmolecular
whichmethylcitrateisconvertedtomethylisocitrate.InE.coli, mass of 102,168 Da (43). It shows 58.8% sequence identity to
whichalsodegradespropionateviathemethylcitratecycle(39), AcnAofE.coli(44)(45)butlessthan20%identitytoAcnBof
this conversion requires two enzymes, i.e. methylcitrate dehy- E. coli (46) and thus belongs to the bacterial AcnA group of
590 AcnR, a Repressor of the Aconitase Gene in Corynebacterium
(13032(cid:3)acnR)inwhichcodons7–177ofacnRwerereplacedby
a21-bpsequencetag.Thegenomicstructureofthemutantwas
confirmedbyPCRandSouthernblotanalysis(51;see“Exper-
imental Procedures”). In a first set of experiments, the influ-
enceoftheacnRdeletionontheaconitaseactivitywasdeter-
mined. Cells were cultivated in CGXII minimal medium with
different carbon sources and used in the early exponential
growth phase (OD (cid:2)6) for the determination of enzyme
600
activity. As shown in Table III, the aconitase acitivity in glu-
cose-growncellswasabout1unit/mgproteinandthus5.3-fold
higherthaninthewildtype.Thisresultclearlysupportedthe
proposed function of AcnR as repressor of acn expression. In
(cid:3)acnR mutant cells grown on citrate, acetate, or propionate,
theaconitaseactivitywas3.4-fold,1.9-fold,and5.0-foldhigher
FIG.4. Growth of C. glutamicum wild type (circles) and the thaninwild-typecellsgrownwiththesamecarbonsourcesand
(cid:2)acnR mutant (triangles) on glucose or acetate minimal me-
dium. The strains were cultivated at 30°C and 120 rpm in CGXII 1.7-fold,1.5-fold,and2.3-foldhigherthanin(cid:3)acnRcellsgrown
minimalmediumcontainingeither222mMglucose(opensymbols)or onglucose.Thisindicatesthattheincreasedaconitaseactivity
244mMsodiumacetate(filledsymbols)ascarbonsource. of wild-type cells grown on acetate is partially because of a
lower repression by AcnR and that an additional transcrip-
tional regulator besides AcnR or another regulatory mecha-
aconitases, which are homologous to the iron-responsive pro-
nismforaconitasemightexist.
teins(IRPs)ofeukaryotes(47).
Inthecultivationsforthedeterminationofaconitaseactivity
AsshowninFig.1,immediatelydownstreamofacn,agene it was observed that in most cases the (cid:3)acnR mutant grew
encodingaputativetranscriptionalregulatorof188aminoacid
somewhat faster than the wild type on glucose and acetate
residues (21,184 Da) was identified, which we named acnR.
minimalmedium,sometimesalsoincitrateminimalmedium.
Interestingly,thesamegeneorganizationwasalsofoundinC.
Inglucosemedium,forexample,thegrowthrateofthe(cid:3)acnR
efficiens, Corynebacterium diphtheriae, Mycobacterium tuber- mutantwas0.40(cid:8)0.02h(cid:4)1andthatofthewildtype0.36(cid:8)
culosis,M.bovis,M.marinum,andRhodococcusstrainIG124 0.01h(cid:4)1(valuesfromsevenreplicates).Thisbehaviorwasalso
(Fig.1).InM.leprae,theacnRgeneisdisruptedbyframeshifts
evidentwhenthecellswereallowedtogrowtostationaryphase
and stop codons, as found for many other genes of this orga-
(Fig.4)andmightbecausedbyanincreasedtricarboxylicacid
nism (48). An amino acid sequence alignment of the AcnR
cycleactivityinthemutant.
homologs (Fig. 2) revealed that the primary sequences are
Transcriptome Analyses—To determine the effect of the
highly conserved, particularly in the N-terminal part from
acnR deletion on global gene expression and in particular on
position 16 to 62, where nearly 75% of the amino acids are
acn expression, the transcriptomes of the (cid:3)acnR mutant and
identical. In silico analysis revealed that this conserved N-
thewildtypewereanalyzedbyDNAmicroarrayanalyses.RNA
terminalpartoftheAcnRproteinsbelongstotheTetRfamily
was isolated from cells growing exponentially in CGXII mini-
of bacterial regulatory proteins (PFAM family PF00440; (49).
malmediumwith4%(w/v)glucose(OD (cid:2)6).Intencompar-
Members of this family contain a multihelical DNA binding 600
isons with RNA from five independent cultivations (experi-
domainattheirN-terminalendandahighlydivergentCter-
mentsA01–A10,Fig.5)theacnmRNAlevelwasalwayshigher
minus,whichisinvolvedinthebindingofaninducermolecule
in the (cid:3)acnR mutant than in the wild type (4.2-fold (cid:8)2.3),
(50).
confirmingthattheincreasedaconitaseactivityofthemutant
Operon Structure of the acn-acnR Genes—The transcrip-
is caused by increased acn expression. Besides acn, several
tional organization of the acn-acnR locus was analyzed by
othergenesshowedmRNAlevelsthatwereonaverage(cid:4)3-fold
RT-PCR (see Fig. 3 and “Experimental Procedures”). Total
increasedinthe(cid:3)acnRmutant,mostofwhicharepresumably
RNA isolated from C. glutamicum wild type was transcribed
into cDNA using three different primers (RT-1-rev, RT-2-rev, involvedintheuptakeofhemeandironsiderophores(Fig.5).
andRT-3-rev)inthereversetranscriptasereactionsA,B,and This indicates that the (cid:3)acnR mutant faces iron limitation, a
C.TheresultingproductswerethenusedforthePCRtestsno. situation which might be due to the increased production of
1–5(Fig.3A).AsshowninFig.3B,evidencewasobtainedthat aconitase,a[4Fe-4S]protein.Indeed,aftersupplementationof
acnandacnRareco-transcribed,sinceacDNAcreatedwitha themediumwith100(cid:1)Mor500(cid:1)MFeSO4thesegenesshowed
primerannealingatthe3(cid:6)-endofacnR(RTreactionA)allowed no longer an increased mRNA level in the (cid:3)acnR mutant,
notonlytheamplificationofanacnRfragment(no.1),butalso whereastheacnmRNAratiowasstillraised(5.8-fold(cid:8)0.6)in
of an acn fragment (no. 2). As an internal control in the RT- themutant(experimentsB01–B04,Fig.5).Thesedataindicate
PCR assays we used dnaE, which encodes a subunit of DNA that acn is the only gene whose mRNA level is significantly
polymerase.Besidesthecontrolreactionno.3,wherenoacnR increasedinthe(cid:3)acnRmutantofC.glutamicum.
productcouldbeobtainedwithacDNAcoveringonlyacn,five Further microarray experiments, in which the influence of
additional control reactions (no. 6–10) were performed which different iron concentrations (508 (cid:1)M versus 8 (cid:1)M) on global
were identical to reactions no. 1–5, respectively, except that gene expression was analyzed separately for the wild type
reverse transcriptase was omitted from the initial reactions (experiments C01–C03, Fig. 5) and the (cid:3)acnR mutant (experi-
A-C. The fact that no PCR products were obtained in these mentsD01–D03,Fig.5),revealedthattheacnmRNAlevelwas
reactionsconfirmedthattheRNAwasnotcontaminatedwith increased under iron excess, both in the wild type (3.0-fold
chromosomalDNA. (cid:8)1.0)andinthe(cid:3)acnRmutant(4.8-fold(cid:8)0.7).Thus,AcnRis
Aconitase Activity and Growth of a (cid:3)acnR Mutant—Since not directly involved in iron regulation. Besides acn, several
AcnR was found to be a member of the TetR family, we sug- othergenesencodingiron-containingproteins(52)showedin-
gested that it functions as transcriptional repressor of the creased mRNA levels under iron excess, indicating that the
neighboringaconitasegene.Inordertotestthishypothesis,an synthesis of these iron-containing proteins is to some extent
in-frame deletion mutant of C. glutamicum was constructed controlledbytheavailabilityofironinC.glutamicum.
AcnR, a Repressor of the Aconitase Gene in Corynebacterium 591
FIG.5. Hierarchical cluster analysis of gene expression changes observed in four series of DNA microarray experiments. The
clusterincludesthosegenes,whoseaveragemRNAlevelwaschanged(cid:4)3-foldor(cid:5)3-foldinatleastoneoffourseriesofmicroarrayexperiments
(totally20).Thefollowingseriesofexperimentswereperformed:A01–A10,(cid:3)acnRversuswildtypecultivatedinCGXII-glucosemediumwith(cid:2)8
(cid:1)Miron;B01–B04,(cid:3)acnRversuswildtypecultivatedinCGXII-glucosemediumwith108(cid:1)Miron(B01,B02)or508(cid:1)Miron(B03,B04);C01–C03,
wildtypewith508(cid:1)Mironversuswildtypewith8(cid:1)Miron;D01–D03,(cid:3)acnRwith508(cid:1)Mironversus(cid:3)acnRmutantwith8(cid:1)Miron.Thescalebar
indicatesthecolorcodingoftherelativeRNAratios.
FIG.6.DeterminationofthetranscriptionalstartsitesoftheC.glutamicumacngeneandexpressionanalysis.12.5(cid:1)g(AandB)
or15(cid:1)g(C)oftotalRNAisolatedfromcellsofC.glutamicumwildtypeandstrain(cid:3)acnRcultivatedinCGXIIminimalmediumwithdifferent
carbonsourceswasusedforprimerextensionanalysiswiththeoligonucleotidesacn-PE1*(AandC)andacn-PE3*(B).Thetranscriptionalstart
sitesareindicatedbyasterisks.ThesequencingreactionsweregeneratedwithaPCRproductcoveringthecorrespondingDNAregionastemplate
andthesameIRD-800-labeledoligonucleotideasintheprimerextensionreaction.
PrimerExtensionAnalysisoftheacnPromoter—Inorderto addition,aweakertranscriptionalstartpointwasobserved113
confirm the results of the DNA microarray experiments with bp upstream of the TTG start codon. The intensity of the
analternativemethodandtolocatetheacnpromoter,primer primerextensionproductswasstrongerinthe(cid:3)acnRmutant
extensionwasperformedusingtotalRNAsamplesofwildtype thaninthewildtype,inaccordancewiththeDNAmicroarray
and (cid:3)acnR mutant, reverse transcriptase and two different data (Fig. 6B). This was true not only for RNA isolated from
primers, acn-PE1* (Fig. 6, A and C) and acn-PE3* (Fig. 6B). glucose-growncells,butalsoforRNAisolatedfromcitrate-and
Withbothprimers,themajortranscriptionalstartpointofacn acetate-grown cells (Fig. 6C), which qualitatively agrees with
waslocalized110bpupstreamoftheproposedstartcodon.In theaconitaseactivitymeasurements(TableIII).
592 AcnR, a Repressor of the Aconitase Gene in Corynebacterium
FIG.7.DeterminationofthenativemolecularmassofAcnRby
size exclusion chromatography. Purified AcnR protein containing
eitheraC-terminaloranN-terminalStrepTag-IIwasseparatedona
HiLoad26/60Superdex200prepgradecolumnusing20mMHEPES
buffer (pH 8.0) containing 300 mM NaCl, and 1 mM dithiothreitol for
elution. For calibration, a premixed protein molecular mass marker
containing the following proteins was used: cytochrome c (12.4 kDa),
carbonicanhydrase(29kDa),albumin(66kDa),alcoholdehydrogenase
(150kDa),(cid:2)-amylase(200kDa).V wasdeterminedwithbluedextran FIG.8.DNaseIfootprintinganalysiswithAcnR-Candtheacn
0
(2,000kDa). promoter region. 0.95 nM labeled acn template strand (panel A) or
0.75nMlabeledacnnon-templatestrand(panelB)wereincubatedwith
TABLE IV increasingconcentrationsofAcnR-C: lane0,noprotein;lane1,5nM
AconitaseactivityindifferentC.glutamicumstrains monomericAcnR;lane2,10nM;lane3,25nM;lane4,50nM;lane5,100
All strains were cultivated under the same conditions in CGXII nM;lane6,250nM;lane7,500nM(notinpanelA);lane8,1.0(cid:1)M;lane
mediumwith222mMglucose.ForstrainsharboringpEKEx2orderiv- 9,2.5(cid:1)M;lane10,5(cid:1)M.RegionsprotectedfromdigestionbyDNaseI
ativesthemediumcontainedinaddition50(cid:1)g/mlkanamycinand1mM areindicatedbyablackbar.TheDNAsequencingreactionsweresetup
isopropyl-1-thio-(cid:2)-D-galactopyranoside. withtheIRD-800-labeledoligonucleotidesusedforgeneratingthela-
beledPCRfragmentsforfootprinting.UnlabeledPCRfragments,iden-
C.glutamicumstrain Aconitaseactivity tical in sequence to the footprinting probes, were used as template
DNA.
units/mgofprotein
13032(wildtype) 0.20(cid:8)0.05
13032/pEKEx2 0.24(cid:8)0.01 from position (cid:4)11 to (cid:4)28 relative to the first transcription
13032/pEKEx2-acnR 0.09(cid:8)0.01 start site (Fig. 8A), whereas on the non-template strand, pro-
13032/pEKEx2-acnRC-Strep 0.08(cid:8)0.02 tectionwasobservedonlyintheregion(cid:4)18to(cid:4)20.Inspection
13032/pEKEx2-acnR60 0.24(cid:8)0.02
13032/pEKEx2-acnR89 0.25(cid:8)0.01 of the upstream regions of the acn genes in C. diphtheriae,
(cid:3)acnR 1.04(cid:8)0.06 C.efficiens,andfourMycobacteriumspeciesrevealedthatthey
(cid:3)acnR/pEKEx2 1.17(cid:8)0.21 all contained sequence motifs similar to the one protected by
(cid:3)acnR/pEKEx2-acnR 0.16(cid:8)0.13 AcnRinC.glutamicum.Accordingtothealignmentshownin
Fig. 9, a putative AcnR consensus sequence could be derived,
PurificationofAcnRandDeterminationofItsNativeMass— whichcontainsanimperfectinvertedrepeatseparatedbyfour
The increased level of acn mRNA and of aconitase activity in nucleotides: CAGNACaagcGTACTG. A binding motif of this
the (cid:3)acnR mutant indicated that AcnR represses acn tran- typeandthissizeistypicalforTetR-typetranscriptionalreg-
scription. In order to identify the operator region, the AcnR ulators like TetR or CamR, although larger motifs have also
protein containing either an N-terminal or a C-terminal beenfound,e.g.forQacR(56)orEthR(55).
StrepTag-IIwasoverproducedinE.coliandpurifiedbyStrep- InordertoconfirmthefootprintdataandtheproposedAcnR
Tactin affinity chromatography to apparent homogeneity as consensus sequence, gel shift assays were performed. Experi-
described under “Experimental Procedures.” The apparent mentswithavaryingAcnR/DNAratiorevealedthata4–6-fold
massonanSDSgelcorrespondedtothecalculatedmassof22.5 molarexcessofAcnRissufficientforashift(datanotshown).
kDa. Similar results were obtained for AcnR-N (data not IntheexperimentshowninFig.10,theshiftofninedifferent
shown). The native molecular mass of the purified AcnR pro- DNA fragments was analyzed, in each case with a 10-fold
teins was determined by size exclusion chromatography. excessofAcnR.Theresultsobtainedwithfragments1–5con-
AcnR-C(1mg)orAcnR-N(0.5mg)werechromatographedona firmed that the binding site is located within a 40-bp region,
Superdex200column(AmershamBiosciences)calibratedwith which includes the binding site proposed from the footprints.
five proteins. As shown in Fig. 7, both AcnR proteins were Fragmentscontainingthisregion(nos.1,2,and5)werecom-
found to have a molecular mass of 44 kDa, indicating a ho- pletelyshifted,whereasthoselackingthisregion(nos.3and4)
modimericstructure,whichwasalsofoundforothermembers werenotshiftedatall.Fragments6–9representderivativesof
of the TetR family, e.g. TetR (53), CamR (50), QacR (54), or fragment 5 with mutations within or outside the proposed
EthR (55). Complementation studies with the streptagged binding motif. Exchange of the two inner (fragment 6) or the
AcnRshowedthatthistagdidnotinterferewiththerepressing two outer (fragment 7) bases of the imperfect inverted repeat
activityofAcnR(TableIV). completely inhibited the shift, as did an exchange of the four
Identification of the AcnR Binding Site within the acn Pro- basesseparatingtheinvertedrepeat(fragment9).Incontrast,
moter Region—The binding site of AcnR at the acn promoter exchangeoffourbasesoutsidetheproposedbindingsite(frag-
was determined with DNase I footprint assays. As shown in ment 8) did not prevent the shift. These data provide strong
Fig. 8, the protected region on the template strand extended supportfortheAcnRconsensusbindingsiteproposedabove.
AcnR, a Repressor of the Aconitase Gene in Corynebacterium 593
FIG.9.SequenceoftheacnpromoterregionofC.glutamicumalignedtoputativeacnpromoterregionsfromotherCoryne-
bacterium and Mycobacterium species. Indicated are the start of the acn coding region (start codons bold and underlined), the
transcriptionalstartsites(*)oftheC.glutamicumacngene,putative(cid:4)10(underlinedandbold)and(cid:4)35regions(bold),andtheAcnRbinding
siteasdeterminedbyDNaseIfootprints(shadedingray)fortheC.glutamicumacngene.Asshownbythealignment,alsotheotherspecies
possessputativeAcnRbindingsitesintheacnupstreamregion.Thebindingsiterepresentsanimperfectinvertedrepeatwiththeconsensus
sequenceCAGNACnnncGTACTG.
FIG.10.Gelshiftexperimentswith
purified AcnR-C. A, DNA fragments
usedforgelshiftanalysis(1–9)areindi-
catedbythewhitenumbersinblackcir-
cles. The gray bar indicates the region
requiredforashiftandtheblackbarin-
dicatestheproposedpositionoftheAcnR
binding site. Fragments 6–9 are deriva-
tivesoffragment5containingmutations
thatwereintroducedwiththeprimers4*-
for, 5*-for, 6*-for, and 7*-for. The 24-bp
sequence shown beside these fragments
includes the proposed AcnR consensus
binding site (bold and underlined) and
fourbasepairsup-anddownstream.The
mutatedbasesareindicatedinboldand
double underlined. B, approximately 50
nM DNA fragments nos. 1–9 were incu-
bated either without AcnR-C (lanes la-
beled with (cid:4)) or with 500 nM AcnR-C
(laneslabeledwith(cid:7))for20minatroom
temperature. Subsequently, the samples
wereseparatedona2.5%agarosegelat
4°C and 75 V in 1(cid:5) Tris acetate/EDTA
buffer, and the gel was stained with
ethidium bromide. The lanes labeled M
containaDNAsizestandard(50-bplad-
der,Fermentas,St.Leon-Rot,Germany).
EffectofOverexpressionofacnRandDerivativesonAconitase DISCUSSION
Activity—Since deletion of acnR had caused derepression of Inthisstudythefirstregulatorofatricarboxylicacidcycle
acn, overexpression of acnR was supposed to lead to an en- gene in C. glutamicum was identified. The acnR gene located
hancedacnrepression.AsshowninTableIV,overexpressionof downstream of the aconitase gene acn of C. glutamicum was
the unmodified acnR gene by the expression plasmid pEKEx- showntoencodearepressorofacnexpression.DeletionofacnR
acnRcausedan(cid:2)2-folddecreasedaconitaseactivity.Acompa- led to a 5-fold increased acn mRNA level and to a 5-fold in-
rable decrease was observed upon overexpression of an acnR creasedaconitaseactivity.Repressionismostlikelycausedby
derivative encoding an AcnR protein with a C-terminal bindingofthedimericAcnRproteintotheacnpromoterinthe
StrepTag-II. Therefore, the tag does not interfere with the regionlocated14to29bpupstreamofthetranscriptionalstart
repressing activity of AcnR. If two shortened derivatives of point,whereitinterfereswiththebindingofRNApolymerase.
acnRwereoverexpressed,whichcontainedonlytheN-terminal By comparing the experimentally determined AcnR binding
60 or 89 amino acid residues of AcnR, there was no effect on sitewithintheC.glutamicumacnpromoter(CAGAACGCTT-
aconitaseactivity.Thus,theN-terminalDNAbindingregionof TGTACTG)withtheacnupstreamregionsofotherCorynebac-
AcnR is not sufficient for its repressing activity, but requires terium and Mycobacterium species, a putative consensus se-
the C-terminal effector binding region. Most likely, the C- quence CAGNACnnnnGTACTG containing an imperfect
terminal domain is essential for the dimerization of AcnR, as inverted repeat was found. Analysis of the C. glutamicum
shownforTetR(53)andQacR(56).WhenunmodifiedacnRwas genome sequence revealed that this motif occurs only three
overexpressedinthe(cid:3)acnRmutant,wild-typelevelsofaconi- timesintheentiregenome,oncewithintheacnpromoter,once
taseactivitywereobtained(TableIV). within the coding region for the ribosomal protein L31, and
594 AcnR, a Repressor of the Aconitase Gene in Corynebacterium
once within the coding region of a gene of unknown function. REFERENCES
ThissuggeststhatacnistheonlytargetgeneofAcnR.Further 1. Kinoshita,S.,Udaka,S.,andShimono,M.(1957)J.Gen.Appl.Microbiol.3,
193–205
support for this assumption was obtained by comparing the
2. Eggeling,L.,andSahm,H.(1999)Appl.Microbiol.Biotechnol.52,146–153
transcriptomes of the (cid:3)acnR mutant and the wild type with 3. Kimura,E.(2003)inMicrobialProductionofL-AminoAcids(Faurie,R.,and
wholegenomeDNAmicroarrays.Inthesestudies,onlytheacn Thommel,J.,eds)Vol.79,pp.37–57,Springer,Heidelberg
4. Pfefferle,W.,Mo¨ckel,B.,Bathe,B.,andMarx,A.(2003)inMicrobialProduc-
gene showed reproducibly a (cid:4)3-fold increased mRNA level in tionofL-AminoAcids(Faurie,R.,andThommel,J.,eds)Vol.79,pp.59–112,
the(cid:3)acnRmutant.Experimentswithdifferentironconcentra- Springer,Heidelberg
5. Hermann,T.(2003)J.Biotechnol.104,155–172
tions revealed that the acn mRNA level is higher under iron 6. Stackebrandt,E.,Rainey,F.A.,andWardRainey,N.L.(1997)Int.J.Syst.
sufficiency than under iron deficiency. A regulation of the ac- Bacteriol.47,479–491
7. Eikmanns,B.J.,Thum-Schmitz,N.,Eggeling,L.,Lu¨dtke,K.U.,andSahm,H.
onitase level by the iron availability was previously also re- (1994)Microbiology140,1817–1828
portedforM.tuberculosis(57). 8. Eikmanns,B.J.,Rittmann,D.,andSahm,H.(1995)J.Bacteriol.177,774–782
9. Usuda, Y., Tujimoto, N., Abe, C., Asakura, Y., Kimura, E., Kawahara, Y.,
OverexpressionoftheacnRgenebyaplasmid-encodedcopy
Kurahashi,O.,andMatsui,H.(1996)Microbiology142,3347–3354
underthecontrolofthetacpromotercauseda2-folddecrease 10. Molenaar,D.,vanderRest,M.E.,andPetrovic,S.(1998)Eur.J.Biochem.254,
395–403
of the aconitase activity in glucose-grown wild-type cell ex-
11. Molenaar,D.,vanderRest,M.E.,Drysch,A.,andYu¨cel,R.(2000)J.Bacteriol.
tracts.Previousexperiencewithothergenesthatwerecloned 182,6884–6891
12. Wendisch,V.F.,deGraaf,A.A.,Sahm,H.,andEikmanns,B.J.(2000)J.
intothesamevector(pEKEx2)andcultivatedunderthesame
Bacteriol.182,3088–3096
growthconditionsusedforacnRoverexpressionindicatesthat 13. Beinert,H.,Kennedy,M.C.,andStout,C.D.(1996)Chem.Rev.96,2335–2374
atleast10-foldhigherAcnRlevelsshouldbeobtained.There- 14. Alen, C., and Sonenshein, A. L. (1999) Proc. Natl. Acad. Sci. U.S.A. 96,
10412–10417
fore,thequestionariseswhythereisonlya2-foldreductionof 15. Beinert,H.,andKiley,P.J.(1999)Curr.Opin.Chem.Biol.3,152–157
aconitaseactivity.Besidesotherexplanationsoneobviousrea- 16. Tang,Y.,andGuest,J.R.(1999)Microbiology145,3069–3079
17. Tang,Y.,Quail,M.A.,Artymiuk,P.J.,Guest,J.R.,andGreen,J.(2002)
soncouldbethatincreasedacnrepressionleadstoanincrease Microbiology148,1027–1037
of the concentration of a metabolite, which binds to AcnR, 18. Tang,Y.,Guest,J.R.,Artymiuk,P.J.,Read,R.C.,andGreen,J.(2004)Mol.
Microbiol.51,1817–1826
triggers the dissociation of the AcnR-operator complex and 19. Menkel,E.,Thierbach,G.,Eggeling,L.,andSahm,H.(1989)Appl.Environ.
thuscounteractstheincreasedrepression.Suchamechanism Microbiol.55,684–688
20. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning. A
would ensure that aconitase expression is never completely
Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring
repressedbyAcnR,whichwouldbeharmfultothecell.Inorder Harbor,NY
21. Keilhauer,C.,Eggeling,L.,andSahm,H.(1993)J.Bacteriol.175,5595–5603
toidentifytheproposedmetabolitewetestedavarietyofme-
22. Studier,F.W.,andMoffatt,B.A.(1986)J.Mol.Biol.189,113–130
tabolites of central carbon metabolism with respect to their 23. Henson,C.P.,andCleland,W.W.(1967)J.Biol.Chem.242,3833–3838
effectsonAcnRbindingtotheacnpromoterusingthegelshift 24. Hanahan,D.(1985)inDNACloning(Glover,D.M.,ed)Vol.1,pp.109–135,
IRL-Press,Oxford/WashingtonD.C.
assay.However,wecouldnotyetfindametabolitethatinhibits 25. van der Rest, M. E., Lange, C., and Molenaar, D. (1999) Appl. Microbiol.
bindingofAcnRtoitsoperator.Thus,thequestionoftheligand Biotechnol.52,541–545
26. Niebisch,A.,andBott,M.(2001)Arch.Microbiol.175,282–294
recognizedbyAcnRisstillopen. 27. Scha¨fer,A.,Tauch,A.,Ja¨ger,W.,Kalinowski,J.,Thierbach,G.,andPu¨hler,A.
Regulation of aconitase gene expression is widespread in (1994)Gene(Amst.)145,69–73
28. Skerra,A.,andSchmidt,T.G.(2000)MethodsEnzymol.326,271–304
bacteria. As mentioned before, E. coli contains two distinct 29. Engels,S.,Schweitzer,J.E.,Ludwig,C.,Bott,M.,andSchaffer,S.(2004)Mol.
aconitases (46), AcnB, which is the major citric acid cycle en- Microbiol.52,285–302
30. Wendisch,V.F.,Zimmer,D.P.,Khodursky,A.,Peter,B.,Cozzarelli,N.,and
zyme and AcnA, an aerobic stationary phase enzyme induced
Kustu,S.(2001)Anal.Biochem.290,205–213
by iron and redox stress. The acnB gene is activated by CRP 31. Mo¨ker,N.,Brocker,M.,Schaffer,S.,Kra¨mer,R.,Morbach,S.,andBott,M.
(2004)Mol.Microbiol.54,420–438
and repressed by ArcA, FruR and Fis and acnA expression is
32. Wendisch,V.F.(2003)J.Biotechnol.104,273–285
activateddirectlyorindirectlybyCRP,FruR,Fur,andSoxRS 33. Polen,T.,Rittmann,D.,Wendisch,V.F.,andSahm,H.(2003)Appl.Environ.
Microbiol.69,1759–1774
andrepressedbyArcAandFNR(58).InB.subtilis,expression
34. Lange,C.,Rittmann,D.,Wendisch,V.F.,Bott,M.,andSahm,H.(2003)Appl.
oftheaconitasegenecitBissubjecttorepressionbytheLysR- Environ.Microbiol.69,2521–2532
typeregulatorCcpCandrepressioncanberelievedbycitrate 35. Ishige, T., Krause, M., Bott, M., Wendisch, V. F., and Sahm, H. (2003) J.
Bacteriol.185,4519–4529
(59, 60). Besides CcpC, also CodY and AbrB were reported to 36. Eisen,M.B.,Spellman,P.T.,Brown,P.O.,andBotstein,D.(1998)Proc.Natl.
be involved in citB expression (61). Similar to E. coli and Acad.Sci.U.S.A.95,14863–14868
37. Laemmli,U.K.(1970)Nature227,680–685
B. subtilis, we also obtained evidence that acn expression in 38. Claes, W. A., Pu¨hler, A., and Kalinowski, J. (2002) J. Bacteriol. 184,
C.glutamicummightbecontrolledbymorethanoneregulator. 2728–2739
As shown in Table III, the aconitase activity of the (cid:3)acnR 39. TexDto.,ra,nSd.,BWuecnkdeils,cWh,.V(1.9F9.,7D)AerGchra.aMf,icAr.oAbi.o,lM.1u¨6ll8e,r4,2U8.,–L4i3n6der,M.I.,Linder,
mutantwas1.5–2.3-foldhigheronacetate,citrate,andpropi- 40. Brock,M.,Maerker,C.,Schutz,A.,Volker,U.,andBuckel,W.(2002)Eur.
J.Biochem.269,6184–6194
onate compared with glucose. Regulators other than AcnR or
41. Gerstmeir,R.,Wendisch,V.F.,Schnicke,S.,Ruan,H.,Farwick,M.,Reinsc-
other regulatory mechanisms might be responsible for this heid,D.,andEikmanns,B.J.(2003)J.Biotechnol.104,99–122
42. Kalinowski,J.,Bathe,B.,Bartels,D.,Bischoff,N.,Bott,M.,Burkovski,A.,
increase. Recently, RamB was identified as a negative tran-
Dusch, N., Eggeling, L., Eikmanns, B. J., Gaigalat, L., Goesmann, A.,
scriptionalregulatorofgenesinvolvedinacetatemetabolismof Hartmann,M.,Huthmacher,K.,Kra¨mer,R.,Linke,B.,McHardy,A.C.,
C.glutamicum(62).Theacngenebelongstotheacetatestimu- Meyer,F.,Mo¨ckel,B.,Pfefferle,W.,Pu¨hler,A.,Rey,D.A.,Ru¨ckert,C.,
Rupp,O.,Sahm,H.,Wendisch,V.F.,Wiegrabe,I.,andTauch,A.(2003)
lon(41)andtwoputativeRamBbindingsiteswereidentifiedat J.Biotechnol.104,5–25
positions(cid:4)492to(cid:4)480((cid:7)strand)and(cid:4)456to(cid:4)444((cid:4)strand) 43. Schaffer,S.,Weil,B.,Nguyen,V.D.,Dongmann,G.,Gunther,K.,Nickolaus,
M.,Hermann,T.,andBott,M.(2001)Electrophoresis22,4404–4422
upstream of the proposed acn start codon (62). Thus, RamB 44. Prodromou,C.,Haynes,M.J.,andGuest,J.R.(1991)J.Gen.Microbiol.137,
might represent a second regulator of acn expression besides 2505–2515
45. Prodromou,C.,Artymiuk,P.J.,andGuest,J.R.(1992)Eur.J.Biochem.204,
AcnR.Ourfuturestudieswillfocusonthesearchforadditional 599–609
regulators of acn expression, for the inducer molecule recog- 46. Gruer,M.J.,andGuest,J.R.(1994)Microbiology140,2531–2541
47. Gruer,M.J.,Artymiuk,P.J.,andGuest,J.R.(1997)TrendsBiochem.Sci.22,
nizedbyAcnR,andfortheanswertothequestionwhetherthe
3–6
C.glutamicumaconitasealsohasaregulatoryfunction. 48. Cole,S.T.,Eiglmeier,K.,Parkhill,J.,James,K.D.,Thomson,N.R.,Wheeler,
P. R., Honore, N., Garnier, T., Churcher, C., Harris, D., Mungall, K.,
Basham,D.,Brown,D.,Chillingworth,T.,Connor,R.,Davies,R.M.,Dev-
Acknowledgments—WethankC.LangeandT.Polenforhelpwith
lin, K., Duthoy, S., Feltwell, T., Fraser, A., Hamlin, N., Holroyd, S.,
DNAmicroarraysanddataanalysisandS.Engelsforassistancewith Hornsby,T.,Jagels,K.,Lacroix,C.,Maclean,J.,Moule,S.,Murphy,L.,
DNaseIfootprintassays. Oliver,K.,Quail,M.A.,Rajandream,M.A.,Rutherford,K.M.,Rutter,S.,