Cyclic-AMP and bacterial cyclic-AMP receptor proteins revisited: adaptation for different ecological niches.

(1)

promoting access to White Rose research papers

White Rose Research Online

eprints@whiterose.ac.uk

Universities of Leeds, Sheffield and York

http://eprints.whiterose.ac.uk/

This is a copy of the final published version of a paper published via gold open access

in

Current Opinion in Microbiology

.

This open access article is distributed under the terms of the Creative Commons

Attribution Licence (

http://creativecommons.org/licenses/by/3.0

), which permits

unrestricted use, distribution, and reproduction in any medium, provided the

original work is properly cited.

White Rose Research Online URL for this paper:

http://eprints.whiterose.ac.uk/78741

Published paper

(2)

Cyclic-AMP

and

bacterial

cyclic-AMP

receptor

proteins

revisited:

adaptation

for

different

ecological

niches

§

Jeffrey

Green

1

,

Melanie

R

Stapleton

1

,

Laura

J

Smith

1

,

Peter

J

Artymiuk

1

,

Christina

Kahramanoglou

2

,

Debbie

M

Hunt

2

and

Roger

S

Buxton

2

Escherichiacolicyclic-AMPreceptorprotein(CRP)represents oneoftheparadigmsofbacterialgeneregulation.Yetdespite decadesofintensivestudy,newinformationcontinuesto emergethatpromptsreassessmentofthisclassicregulatory system.Moreover,inrecentyearsCRPsfromseveralother bacterialspecieshavebeencharacterized,allowingthegeneral applicabilityoftheCRPparadigmtobetested.Herethe propertiesoftheE.coli,Mycobacteriumtuberculosisand

PseudomonasputidaCRPsareconsideredinthecontextofthe ecologicalnichesoccupiedbythesebacteria.Itappearsthat thecyclic-AMP-CRPregulatorysystemhasbeenadaptedto respondtodistinctexternalandinternalinputsacrossabroad sensitivityrangethatis,atleastinpart,determinedbybacterial lifestyles.

Addresses

1

TheKrebsInstitute,DepartmentofMolecularBiologyand Biotechnology,UniversityofSheffield,SheffieldS102TN,UK

2_Division_of_{Mycobacterial}_Research,_MRC_National_Institute_for_Medical

Research,MillHill,LondonNW71AA,UK

Correspondingauthor:Green,Jeffrey(jeff.green@sheffield.ac.uk)

CurrentOpinioninMicrobiology2014,18:1–7

ThisreviewcomesfromathemedissueonCellregulation

EditedbyCeciliaArraianoandGregoryMCook

1369-5274/$–seefrontmatter,#2014TheAuthors.Publishedby ElsevierLtd.Allrightsreserved.

http://dx.doi.org/10.1016/j.mib.2014.01.003

The

Escherichia

coli

paradigm

Thecyclic-AMPreceptor protein(CRP)anditseffector cyclic-AMP (cAMP) were discovered in E. coli during investigations to explain the phenomenon of diauxic growth more than 40 years ago. Since then, E. coli CRP–cAMPhasbecomeaparadigmofgeneregulation, providing insights into signal perception and transduc-tion,DNArecognitionbyregulatoryproteins,regulator– polymerase interactionsandpromoterarchitecture[1].

TheformationofthesecondmessengercAMPfromATP is catalyzed bya group of enzymes known as adenylyl cyclases. These enzymes are classified into six groups basedontheirprimarystructures.E.colipossessesasingle ClassIadenylylcyclase(Cya)whoseactivityiscontrolled by glucose availability, such that growth at micromolar concentrations of glucose increases intracellular cAMP concentrations10-fold(20mMto180mM)compared

toexcessglucoseconditions[2].Theconsensusviewhas been that when the preferred carbon source glucose is available, it is transported into the cell by the glucose phosphotransferasesystem(PTS)andglucoseentersthe cytoplasmasglucose-6-phosphate[3,4].The phosphoryl-ationstateofthePTSthus actsasareporterofglucose availability—the phosphorylation state of the PTS is lowerwhentheglucosetransporterissaturated,whereas when glucose is absent, phosphorylated PTS proteins accumulate. The phosphorylated PTS interacts with Cyatoenhanceadenylylcyclaseactivity[3].Thus,when thebacteria areglucose-starved,theintracellular cAMP concentrationincreases as aconsequence ofthe altered phosphorylationstateofthePTS;howeverthisisdifficult toreconcilewithobservationsthattheglucosePTSisstill saturated when intracellular cAMP concentrations increase [2]. Consequently, it has been suggested that cAMPconcentrationsincreaseinresponsetolowenergy charge, such that when ATP is low, cAMP is high, promoting catabolism and inhibiting anabolism by CRP–cAMP-mediatedgeneregulationtobridgethe per-ceivedenergydeficit[5].

Degradationof cAMPismediatedbya phosphodiester-ase,CpdA,butthisenzymehasaratherhighKmforcAMP

(500mM)relativeto theconcentrationofcAMPin the

cell,andacpdAmutantexhibitedonlyatwofoldincrease in intracellular cAMP concentration [6]. Consequently, theobservationthatcAMPisoftenfoundextracellularly (0.03–0.5mM) ledto thefindingthatE. colicanquench

intracellular levels of cAMP by TolC-mediated efflux, althoughthecAMPtransporter(s)thatlinkstothe outer-membranepore TolChasnotyetbeen identified[7].

Changes in intracellular cAMP concentration are per-ceivedbythetranscriptionfactorCRP.CRPisa homo-dimer in which each subunit possesses three major structural features (Figure 1). The N-terminal region housesthehigh-affinitycAMP-bindingdomainandthe C-terminal region consists of a DNA-binding domain with a canonical helix-turn-helix motif. These two

§_This_is_an_open-access _article_distributed_under_the_terms_of_the

CreativeCommonsAttributionLicense,whichpermitsunrestricteduse,

distribution, andreproduction inanymedium, providedtheoriginal

(3)

domains are connected by a long helix (C-helix) that forms a coiled-coil at the dimer interface and a short linker followed by another helix (D-helix) (Figure 1). Cyclic-AMP binding to the sensory domain is initially exothermic (DH1=16.3kJmol1; DS1=

41JK1mol1) followed by an endothermic phase (DH2=25.1kJmol

1

;DS2=176JK 1

mol1)andcAMP interactionswiththetwoprotomersthatmakeuptheCRP dimer are cooperative (DG2DG1=2.7kJmol1) with

binding constants of 8104M1 for site 1 and 6

104M1forsite2[8].

In theapo-CRP dimer, thetwo DNA-binding domains interact to form arigid body in which the DNA-recog-nition helices areburied suchthat they cannot interact withDNA[9].BindingofcAMPtoCRPinitiates struc-turalrearrangementsabouta‘hinge’regionallowingthe DNA-bindingdomainstorelocaterelativetothe cAMP-bindingdomaininaprocessmediatedthroughhydrogen bondsbetweentheN(6)positionofadenosinewith Ser-128 of the dimerization helices (C-helices; Figure 1) [9,10].Thisallostericconversioncriticallyinvolves exten-sionoftheC-helicesbysixresiduesandshorteningofthe D-helicesbyfourresidues,suchthatAsp138becomesthe N-terminalcappingresidueoftheD-helixinCRP–cAMP

(Figure1),butisaninternalpartofthelongerD-helixin

apo-CRP(Figure 2aand b).The helix-capping propen-sity of residue 138 is correlated to the degree of co-operative cAMP-binding and hence this property of

Asp138isakeyfeatureoftheinterdomainconformational changes that modulate the apo-CRP$CRP–cAMP equilibrium[8].NMRspectroscopyandthermodynamic analysesofseveralCRPvariantsrevealedhowchangesin conformationalentropymodulateDNA-bindingactivity [11].Thesestructuralrearrangementsexposethe DNA-recognition helices (highlighted in green in Figure 1) suchthattheyareabletoparticipateinsequencespecific (consensus sequenceTGTGAnnnnnnTCACA) binding attwoadjacent majorgroovesof DNA(Figure 2b)[12].

Cyclic-AMPbindinghasabiphasiceffectonsite-specific DNA-bindingbyCRP.High-affinitybindingofcAMPin

theanti-conformationatthesensorydomainsoftheCRP

subunits (1:1 cAMP/CRP protomer) enhances DNA-binding 1000-fold. This is followed by decreased DNA-bindingwhencAMPinthesyn-conformation inter-actswithweakbindingsitesformedbycomponentsofthe helix-turn-helix,ab-hairpinfromtheregulatorydomain and DNA (2:1 cAMP/CRP protomer) [13]. However, proof that cAMP-binding to these low affinity sites is ofphysiological significancehasnotyetbeen provided.

CRP–cAMPbindingtointergenicCRPsitesisassociated withclassicalgeneregulation[14,15];however,itisnow recognized that CRP–cAMP also binds at many intra-genic sites where it is thought to fulfill a role as a chromosome organizer or nucleoid associated protein (NAP)[14].GenomeSELEXscreeningidentifiedatleast

2 Cellregulation

Figure1

C-helix Recognition

helix Asp 138

Ser

128 _{high affinity}

cAMP site D-helix

low affinity cAMP site

Current Opinion in Microbiology

(4)

254 CRP-binding sites across the E. coli genome and becauseCRPiscapableofcontrollingexpressionof diver-gentpromoters fromasinglebindingsiteitisestimated that CRP–cAMP directly controls a minimum of 378 promoters, and perhaps >500 genes in E. coli [15]. Amongsttheseoperons,CRP–cAMPactsasthe‘master’ regulatorfor70‘slave’transcriptionfactorsfurther expand-ingtheprofoundinfluenceofCRPonglobalgene

expres-sionin E. coli,in which itplays akeyrole in managing

catabolism,includingthetransportofsubstrates, glycoly-sis,theKrebscycleandaerobicrespiration[15,16].

Variation

1 —

Mycobacterium

tuberculosis

CRP,

Rv3676,

a

regulator

evolved

to

operate

at

high

cAMP

concentrations?

Unlike E. coli,which hasonly one adenylylcyclase, M. tuberculosisH37Rvpossessesatleast16ClassIIIadenylyl cyclase-like proteins, including soluble and membrane-associated multidomain proteins, suggesting that their catalyticactivities(10ofthe16havebeenshowntoact as adenylyl cyclases) can be regulated by extracellular

and/orintracellularsignals,reviewedbyChakraborti[17]. Accordingly,adenylylcyclaseactivityofM.tuberculosisis affected bypH, CO2,and fatty acids. Ithas long been

recognizedthatmycobacteriasecretecAMP,butitisonly morerecentlythatthecapacitytointoxicatemacrophages with cAMP has been recognized as a contributor to virulence [18,19]. Thus, the synthesis (in particular by Rv0386)andsecretionofcAMParecentralfeaturesofM. tuberculosispathogenesisandresultinthebacteriumbeing exposedtorelativelyhighconcentrationsofcAMP;there arereportsofintracellularconcentrationofcAMPashigh

as 4mMfor M.tuberculosisH37Rvand 3mMfor

Myco-bacteriumsmegmatis,whichfarexceedvaluesreportedfor

E. coli [20,21]. However, it is wise to offer a note of

cautionhere;becausethedifferentgrowthconditionsand methods usedto measure cAMP,it isdifficult to make direct comparisons. Nevertheless, the important role that cAMP plays in tuberculosis pathogenesis exposes the need for careful investigation of both intracellular and extracellular cAMP concentrations using modern approaches.

Figure2

(a) (b)

(d) (c)

Current Opinion in Microbiology

(5)

TherelativelyhighconcentrationsofcAMPreportedfor mycobacteriaareconsistentwiththepresenceofmultiple adenylylcyclasesbutofonlyonecAMP phosphodiester-ase (Rv0805) in M. tuberculosis H37Rv. Moreover, the cAMP phosphodiesterase activity of Rv0805 is poor, and like its E. coli counterpart, it hasa relatively high Km for cAMP (150mM) [22]. Thisrather poor in vitro

activity is reflected in vivo, where overproduction of Rv0805 resulted in only a 30% decrease in cAMP (a

90%decreasewasobservedforoverproductionofCpdA in E. coli), perhaps indicating alternative roles for this enzyme,whichalso possessestheabilityto hydrolyzea rangeofcNMPandlinearphosphodiesters[23].Inthe lightofthesedata,ithasbeensuggestedthatintracellular cAMPlevelsmightbecontrolledbyexcretionratherthan conversiontoAMPbut,asisthecaseforE.coli,thereisa need to establishthe mechanism(s) of cAMP secretion andhowthis mightberegulated.

TheM.tuberculosisCRP(Rv3676;32%aminoacididentity

toE.coliCRPover189aminoacids,includingfourofthe sixkeycAMP-interactingresiduesinthesensorydomainof

E.coliCRP;Table1)differsfromtheE.coliparadigmat

severallevels.TheRv3676homodimerexhibitsrelatively weak(Kb=1.7104M1)bindingofcAMPtotwo

inde-pendentsites(1:1cAMP/protomer).Furthermore, cAMP-binding is endothermic (DH=30.7kJmol1; DS= 183JK1mol1;DG=23.7kJmol1)andthusbinding is entropically driven [24]. The independent nature of cAMP-binding to Rv3676comparedto E. coli CRPwas accountedforbythereplacementofasingleaminoacid residue(Ser-128ofCRP,whichisrequiredforthedramatic conformationalchangesthatoccuruponcAMP-binding,is replacedbyAsninRv3676)thathastheeffectof reorga-nizingahydrogen-bondingnetworkinvolvingcAMPsuch thatthecAMP-bindingsitesinRv3676areuncoupled[24]. Ithasbeenarguedthattherelativelyweakand indepen-dentbindingofcAMPatthesensorydomainofRv3676 hasevolved to allow the protein to be atleast partially

cAMP-responsiveagainstthebackgroundof highcAMP concentrations required to intoxicate the host during infection.

Thecrystalstructuresofapo-Rv3676andRv3676-cAMP reveal that cAMP-binding is associated with much less dramaticstructural rearrangementsthan those observed

for E. coli CRP [25–27] (Figure 2c and d). The major

alterationthatoccursuponcAMP-bindingisweakening of theinteractions betweentheDNA-binding and sen-sorydomains,resultinginincreasedspatialfreedomofthe DNA-binding domain that is apparently sufficient to permitbindingtotargetDNAsequences byaninduced fitmechanism[27].

Consistent withthe relatively mild structural rearrange-ments that occur upon cAMP-binding by Rv3676, the formationoftheRv3676–cAMPcomplexhasarelatively small effect (2-fold) on DNA-binding to a consensus sequence that is very similar to that of E. coli (GTGnnAnnnnnCACA)[28].Furthermore,unlikeE.coli CRP,apo-Rv3676iscapableofsite-specificDNA-binding andtranscriptionregulation [24].Theseobservationsare consistentwiththelimitedoverlapbetweengenes dysre-gulatedinthecrpmutantand thoseaffectedbyRv0805 overproduction[23]andsuggeststhattheprimaryroleof Rv0805mightnotbetoactasacAMPphosphodiesterase and/orthatRv3676cansignificantlyinfluencegene expres-sionwithouttheneedtobindcAMP.

Like E. coli CRP, M. tuberculosis Rv3676 is a global

regulator but, perhaps unsurprisingly in the context of the very different lifestyles of these two bacteria, the correspondingCRPregulonsdiffer.ThusRv3676appears tobeinvolvedinregulatingthetransitionbetween repli-cating and nonreplicating states by exerting influence over virulence-critical functions, including phthiocerol dimycocerosate(DIM)synthesis,resuscitationpromoting factors, the ESX-1 type VII secretion system, carbon

4 Cellregulation

Table1

ComparisonoffeaturesofcAMP-signalinginthreebacteria

Bacterium M.tuberculosis E.coli P.putida

Niche Lungmacrophage Mammalianintestine Soil

Numberofadenylylcyclases 16 1 2

IntracellularcAMPconcentrations High Moderate Low

CRP Rv3676 CRP PP_0424

cAMP–CRPinteractions Independentbinding Cooperativebinding Independentbinding

KDforcAMP 60mM 13–16mM 23nM

MotifforcAMPinteractiona _E_._._._TS_._._._R_._._._T_N _E_._._._RS_._._._R_._._._T_S _E_._._._RS_._._._R_._._._T_T

Numberofphosphodiesterases 1 1 1

Numberofchromosomalbindingsitesb

>70 >378 >30

a_Amino_acids_involved_in_direct_interaction_with_cAMP_in_E._coli_CRP_as_single_letter_code_with_dots₍

...)representinginterveningregionsofvarious lengths.TheaminoacidatthepositionequivalenttoSer-128inE.coliCRPthatmakesacross-subunitcontactwithcAMPisshowninboldfont.

b_The_binding_site_numbers_represent:_matches_to_the_Rv3676_consensus_sequence_identified_by_Rickman_et_al._[₂₈_]_in_the_M._tuberculosis_H37Rv

genomesequence;E.coliCRP-bindingsitessuggestedbyShimadaetal.[15_]_following_genomic_SELEX_analysis;_and_a_minimum_value_based_on

(6)

metabolism,energyconservationand‘slave’transcription factors, such as the nitric oxide-responsive regulator WhiB1[28,29].Thisdegreeofcontroloverthe transcrip-tome is consistent with the attenuated state of the M. tuberculosiscrpmutantin modelsof infection[28].

Variation

2 —

Pseudomonas

putida

CRP,

PP

_

0424,

a

regulator

evolved

to

operate

at

low

cAMP

concentrations?

P.putidapossessesaCyaA-typeadenylylcyclasecapable

of cAMP synthesis (and there is also a second protein PP_5187 annotated as an adenylyl cyclase), but never-thelesscAMPconcentrationsarebelowthelevelof detec-tioninbioassays[30].P.putidaKT2440possessesacAMP phosphodiesterase (PP_4917) equivalent to the E. coli CpdA protein.Thus, theverylow levelsof cAMPin P. putidacouldarisefrompoorsynthesisorrapiddegradation, butbasedoncomplementationexperimentswithE.colicya mutants the former is the more likely. Thus, P. putida seemstorepresenttheoppositeendofthe‘cAMP spec-trum’toM.tuberculosis.TheP.putidaCRPis63%amino acid identical to E. coli CRP over 208 amino acids, in-cludingfiveoutofthesixcAMP-interactingresiduesinthe sensory domain of E. coli CRP—interestinglythe mis-matchisagainlocatedatthepositionequivalentto128in

E. coli CRP (Table 1). Consistent with the very low

concentrationsofcAMP,theP.putidaCRPexhibitsvery highaffinity(Kb=4.4107M1)bindingofcAMPtotwo

independent sites (1:1 cAMP/protomer). Furthermore, cAMP-binding is exothermic (DH=25kJmol1; DS= 63JK1mol1;DG=10.5kJmol1),andbindingisboth enthalpy and entropy driven [31]. Although detailed structuralinformationisnotyetavailable,this hypersensi-tive binding of cAMP invokes large conformational changes thatcan bedetected bysizeexclusion chroma-tographyandresultinenhancedDNA-bindingtoatypical CRPinvertedrepeatsequenceby>10-fold[31].

Although P. putida exhibits catabolite repression, this behaviorisnotmediatedbyCRP–cAMP,asacrpmutant wasunaffectedinitsabilitytoutilizeafullrangeofsugars as carbon sources [32]. Rather, CRP–cAMP appears to control theutilizationofL-phenylalanineandof various

dipeptidesasnitrogensourcesinP.putida[30,33].The fullrangeofgenesregulatedbyCRPinP.putidahasnot been establishedbut aconservativeanalysisof the gen-omesequenceforlikelybindingsitesindicatesthat>30 genesmightberegulatorytargets,fewofwhichappearto haveametabolicrole[30].Theseobservationsledtothe suggestion that P.putida and M. tuberculosisCRPshave evolved to control different biological processes com-pared to the E. coli paradigm, a possible example of regulatoryexaptation[30].

Perspectives

and

outstanding

questions

InrecentyearsmanyaspectsofthecAMP-signaling CRP-regulatory paradigm that has emerged from intensive

studiesofcataboliterepressioninE.colihavecomeunder closerscrutiny. Almostevery stepfrom therelationship between the activity of the glucose PTS and cAMP synthesis to the role of cAMP–CRP in gene regulation and chromosomeorganization hasbeen reassessed.The picturethatisemergingisoneinwhichintracellularand extracellular cAMP concentrations are modulated by adenylyl cyclases, phosphodiesterasesand cAMP efflux systemssomeofwhichrespondtoexternaland/orinternal signals.ChangesinintracellularcAMPconcentrationare perceived by CRP proteins that react with different sensitivities related to the niches occupied bythe bac-teria.Thus,thepathogenM.tuberculosisRv3676isalow sensitivity CRP evolved to maintain some degree of responsiveness at the high cAMP concentrations used to intoxicate host macrophages; the commensal enteric bacterium E. coli possesses a mid-sensitivity CRP to regulatecataboliterepressionandchromosomestructure, probably in response to energy charge; and the soil bacteriumP.putidahasahypersensitiveCRP,reflecting the very low concentrations of cAMPproduced bythis bacterium.Nevertheless,despitedecadesofstudythere are still many outstanding questions that need to be addressed to complete ourunderstanding of cAMP-sig-naling inE. coliandother bacteria.Forexample,atthe rootofcAMP-mediatedsignalinginbacteriaistheability tosynthesizeanddegradecAMPinresponseto environ-mentalandmetabolicsignals.Toplaceinvestigationsof cAMP-signaling on a sound footing there is a need to apply the latest metabolite quenching, extraction and analysis techniques to accurately measure intracellular and extracellular cAMP concentrations under diverse growthconditions.

Further characterization of the phosphodiesterases involvedincAMPdegradationandtheprocessesrequired for cAMP excretion is required. Signal-dependent syn-thesis of cAMP is only one component in controlling bacterialresponsestothissecondmessenger;therehave tobemechanismsforremovingcAMPfromthesystem.It appears that the cAMP phosphodiesterases are rather poor enzymes, leading to the suggestion that secretion ofcAMPmightbethemajorroutetoloweringcAMPin thecell.However,thusfarnocAMPeffluxsystemshave been identified beyond the recognition that TolC is involved in facilitating cAMP crossing the outer mem-braneofE.coli.IdentifyingcAMPsecretionsystemsand definingtheirroleinbacterialsignalingandpathogenesis will fillamajordeficitinourcurrentknowledge.

Itwillbeinformativetoseekphysiologicaland evolution-aryexplanationsforthedifferencesincAMP-bindingto the regulatory CRP proteins in different bacteria, in particular, establishing whether CRPs, like that of M.

tuberculosis,which exhibitsonly a mildenhancement in

(7)

mechanisticexplanationsfortheextremelyavid cAMP-bindingbyCRPs,asexemplifiedbytheP.putida CRP, shouldbesought.Atpresentitappearsthatrespondingto thesecondmessengercAMPallowsCRPtobeco-opted to control different regulons in bacteria that occupy distinct niches. Thus, because cAMP intoxication of the host is an important component of M. tuberculosis pathogenicity, its CRP has become desensitized to cAMP,whereas theCRPofthesoilbacteriumP.putida has become hypersensitive to cAMP. Hence througha commonmechanismofCRP-mediatedRNApolymerase recruitment and signal-dependent cAMP synthesis/ degradation(e.g. glucose availability in E. coli;pH and othervirulence-relatedsignalsinM.tuberculosis;unknown signals possibly related to the utilization of aromatic amino acids and nitrogen sources by P. putida), CRP hasbeenco-optedtocontroldistinct regulonsaccording totheparticularnichesoccupiedbythebacteria. Devel-oping a mechanistic framework that accounts for the shifts in cAMP-binding affinities observed in different CRPswouldthenallowquestionsaboutthephysiological rolesofcAMP–CRPcomplexes withalternative stoichi-ometriestobeaddressed.

Acknowledgements

Our work has been generously supported over several years by The

BiotechnologyandBiologicalSciencesResearchCouncilUK,TheMedical

ResearchCouncilUK(grantnumberU117585867),EuropeanUnionFP7

program(SysteMTbHEALTH-F4-2010-241587)andTheWellcome

Trust.

References

and

reading

Papersofparticularinterest,publishedwithintheperiodofreview, havebeenhighlightedas:

ofspecialinterest

ofoutstandinginterest

1. BusbyS,EbrightRH:Transcriptionactivationbycatabolite activatorprotein(CAP).JMolBiol1999,293:199-213.

2. Notley-McRobbL,DeathA,FerenciT:Therelationshipbetween externalglucoseandconcentrationofcAMPlevelsin

Escherichiacoli:implicationsformodelsof

phosphotransferase-mediatedregulationofadenylate cyclase.Microbiology1997,143:1909-1918.

3. PostmaPW,LengelerJW,JacobsonGR:

Phospho-enolpyruvate:carbohydratephosphotransferasesystemsof bacteria.MicrobiolRev1993,57:543-594.

4. BettenbrockK,SauterT,JahreisK,KremlingA,LengelerJW, GillesED:Correlationbetweengrowthrates,EIIACrr phosphorylationandintracellularcyclicAMPlevelsin

EscherichiacoliK12.JBacteriol2007,189:6891-6900.

5. NarangA:Quantitativeeffectandregulatoryfunctionofcyclic adenosine50_-phosphate_in_Escherichia_coli_._J_Biosci_2009,

34:445-463.

6. InamuraR,YamanakaK,OguraT,HiragaS,FujitaN,IshihamaA, NikiH:IdentificationofthecpdAgeneencodingcyclic30_,50

-adenosinemonophosphatephosphodiesteraseinEscherichia coli.JBiolChem1996,271:25423-25429.

7.

HantkeK,WinklerK,SchultzJE:Escherichiacoliexportscyclic AMPviaTolC.JBacteriol2011,193:1086-1089.

Thisstudyprovidesevidencethatthemosteffectivemechanismtoswitch offcAMP-signalinginE.coliisnotbydegradationofcAMPbutviaits effluxinaprocessthatinvolvesTolC.

8.

YuS,MaillardRA,GribenkoAV,LeeJC:TheN-terminalcapping propensitiesoftheD-helixmodulatetheallostericactivation oftheEscherichiacolicAMPreceptorprotein.JBiolChem

2012,287:39402-39411.

Thispaper describes the quantitative characterization of the roleof residue138inmodulatingtheequilibriumofapo-CRPandCRP–cAMP andthusprovidesthethermodynamicbasisfortheallosteryrevealedby structuralanalyses.

9. SharmaH,YuS,KongJ,WangJ,SteitzTA:Structureofapo-CAP revealsthatlargeconformationalchangesarenecessaryfor DNA-binding.ProcNatlAcadSciUSA2009,106:16604-16609.

10. PopovychN,TzengSR,TonelliM,EbrightRH,KalodimosCG:

StructuralbasisforcAMP-mediatedallostericcontrolofthe cataboliteactivatorprotein.ProcNatlAcadSciUSA2009,

106:6927-6932.

11.

TzengSR,KalodimosCG:Proteinactivityregulationby conformationalentropy.Nature2012,488:236-240.

Thispaperprovidesevidencethattheaffinityofprotein–ligand interac-tionsismodulatedbyglobalinternaldynamicsbystudyingDNA-binding byCRP.

12. WeberIT,SteitzTA:Structureofacomplexofcatabolitegene activatorproteinsandcyclicAMPrefinedat2.5A˚ resolution.J MolBiol1987,198:311-326.

13. PassnerJM,SteitzTA:ThestructureofaCAP–DNAcomplex havingtwocAMPmoleculesboundtoeachmonomer.Proc NatlAcadSciUSA1997,94:2843-2847.

14. GraingerDC,HurdD,HarrisonM,HoldstockJ,BusbySJ:Studies ofthedistributionofEscherichiacolicAMP-receptorprotein andRNApolymerasealongtheE.colichromosome.ProcNatl AcadSciUSA2005,102:17693-17698.

15.

ShimadaT,FujitaN,YamamotoK,IshihamaA:Novelrolesof cAMPreceptorprotein(CRP)inregulationoftransportand metabolismofcarbonsources.PLoSONE2011,6:e200081.

Thisstudyidentifies183newCRP–cAMP-regulatedpromotersinE.coli usingaGenomicSELEXapproach,expandingthenumberofoperons underCRPcontroltoaminimumof378.

16. HaverkornvanRijsewijkBR,NanchenA,NalletS,KleijnRJ, SauerU:Largescale13_C-flux_analysis_reveals_distinct

transcriptionalcontrolofrespiratoryandfermentative metabolisminEscherichiacoli.MolSystBiol2011,7:477.

17. ChakrabortiPK,MatangeN,NandicooriVK,SinghY,TyagiJS, VisweswariahSS:Signallingmechanismsinmycobacteria.

Tuberculosis2011,91:432-440.

18. AgarwalN,LamichhaneG,GuptaR,NolanS,BishaiWR:Cyclic AMPintoxicationofmacrophagesbyaMycobacterium tuberculosisadenylatecyclase.Nature2009,460:98-102.

19. BaiG,SchaakDD,McDonoughKA:cAMPlevelswithin

MycobacteriumtuberculosisandMycobacteriumbovisBCG increaseuponinfectionofmacrophages.FEMSImmunolMed Microbiol2009,55:68-73.

20. PadhH,VenkitasubramanianTA:Cyclicadenosine30_,50

-monophosphateinmycobacteria.IndJBiochemBiophys1976,

13:413-414.

21. DassRKM,SharmaR,ShenoyAR,MattooR,VisweswariahSS:

CyclicAMPinmycobacteria:characterizationandfunctional roleoftheRv1647orthologinMycobacteriumsmegmatis.J Bacteriol2008,190:3824-3834.

22. ShenoyAR,SreenathN,PodobnikM,KovacevicM, VisweswariahSS:TheRv0805genefromMycobacterium tuberculosisencodesa30_,50_-cyclic_nucleotide

phosphodiesterase:biochemicalandmutationalanalysis.

Biochemistry2005,44:15695-15704.

23.

MatangeN,HuntDM,BuxtonRS,VisweswariahSS:

OverexpressionoftheRv0805phosphodiesteraseelicitsa cyclic-AMP-independenttranscriptionalresponse.

Tuberculosis2013http://dx.doi.org/10.1016/j.tube.2013.05.004. ThispapershowsthedissociationoftheM.tuberculosis phosphodies-teraseRv0805fromcAMP-regulatedgeneexpressionconsistentwiththe smallenhancementinRv3676DNA-bindingactivityelicitedbycAMPand suggestingalternativerolesforRv0805.

(8)

24. StapletonMR,HaqI,HuntDM,ArnvigKB,ArtymiukPJ, BuxtonRS,GreenJ:MycobacteriumtuberculosiscAMP receptorprotein(Rv3676)differsfromtheEscherichiacoli

paradigminitscAMP-bindingandDNA-bindingproperties andtranscriptionactivationproperties.JBiolChem2010,

285:7016-7027.

25. GallagherDT,SmithN,KimSK,RobinsonH,ReddyPT:Profound asymmetryinthestructureofthecAMP-freecAMPreceptor protein(CRP)fromMycobacteriumtuberculosis.JBiolChem

2009,284:8228-8232.

26. ReddyMC,PalaninathanSK,BruningJB,SmithD,SacchettiniJC:

Structuralinsightsintothemechanismoftheallosteric transitionsofMycobacteriumtuberculosiscAMPreceptor protein.JBiolChem2009,284:36581-36591.

27. KumarP,JoshiDC,AkifM,AkhterY,HasnainSE,MandeSC:

MappingconformationalchangesincyclicAMPreceptor protein:crystalstructureandnormal-modeanalysisof

Mycobacteriumtuberculosisapo-cAMPreceptorprotein.

BiophysJ2010,98:305-314.

28. RickmanL,ScottC,HuntDM,HutchinsonT,Mene´ndezMC, WhalanR,HindsJ,ColstonMJ,GreenJ,BuxtonRS:Amemberof thecAMPreceptorproteinfamilyoftranscriptionregulatorsin

Mycobacteriumtuberculosisisrequiredforvirulenceinmice andcontrolstranscriptionoftherpfAgenecodingfora resuscitationpromotingfactor.MolMicrobiol2005,56 :1274-1286.

29. SmithLJ,StapletonMR,FullstoneGJM,CracjJC,ThomsonAJ,Le BrunNE,HuntDM,HarveyE,AdinolfiS,BuxtonRS,GreenJ:

MycobacteriumtuberculosisWhiB1isanessential DNA-bindingproteinwithanitricoxide-sensitiveiron–sulfurcluster.

BiochemJ2010,432:417-427.

30.

MilanesioP,Arce-RodriguezA,MunozA,CallesB,deLorenzoV:

Regulatoryexaptationofthecataboliterepressionprotein (Crp)–cAMPsysteminPseudomonasputida.EnvironMicrobiol

2011,13:324-339.

ThisstudyinvestigatedwhetherthepresenceofCyaandCRPproteinsin P.putidaretainedthefunctionsoftheE.coliorthologsandconcludedthat CRP–cAMPhasbeenco-optedtocontroladistinctsetofgenesinP. putida.

31.

Arce-RodriguezA,Durante-RodriguezG,PlateroR,KrellT, CallesB,deLorenzoV:TheCrpregulatorofPseudomonas putida:evidenceofanunusuallyhighaffinityforits physiologicaleffector,cAMP.EnvironMicrobiol2012,14 :702-713.

ThispapershowsthatalthoughtheP.putidaCRPretainsmanyofthe basicfeaturesoftheE.coliparadigmithasaremarkablyhighaffinityfor cAMP.

32. RojoF:CarboncataboliterepressioninPseudomonas

optimizingmetabolicversatilityandinteractionswiththe environment.FEMSMicrobiolRev2010,34:658-684.

33. HerreraMC,DaddaouaA,Fernandez-EscamillaA,RamosJL:

InvolvementoftheglobalCrpregulatorincyclic AMP-dependentutilizationofaromaticaminoacidsby

Pseudomonasputida.JBacteriol2012,194:406-412.