13 pages
English

Mitochondrial NAD+-dependent malic enzyme from Anopheles stephensi:a possible novel target for malaria mosquito control

-

Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus

Description

Anopheles stephensi mitochondrial malic enzyme (ME) emerged as having a relevant role in the provision of pyruvate for the Krebs' cycle because inhibition of this enzyme results in the complete abrogation of oxygen uptake by mitochondria. Therefore, the identification of ME in mitochondria from immortalized A. stephensi (ASE) cells and the investigation of the stereoselectivity of malate analogues are relevant in understanding the physiological role of ME in cells of this important malaria parasite vector and its potential as a possible novel target for insecticide development. Methods To characterize the mitochondrial ME from immortalized ASE cells (Mos. 43; ASE), mass spectrometry analyses of trypsin fragments of ME, genomic sequence analysis and biochemical assays were performed to identify the enzyme and evaluate its activity in terms of cofactor dependency and inhibitor preference. Results The encoding gene sequence and primary sequences of several peptides from mitochondrial ME were found to be highly homologous to the mitochondrial ME from Anopheles gambiae (98%) and 59% homologous to the mitochondrial NADP + -dependent ME isoform from Homo sapiens . Measurements of ME activity in mosquito mitochondria isolated from ASE cells showed that ( i ) V max with NAD + was 3-fold higher than that with NADP + , ( ii ) addition of Mg 2+ or Mn 2+ increased the V max by 9- to 21-fold, with Mn 2+ 2.3-fold more effective than Mg 2+ , ( iii ) succinate and fumarate increased the activity by 2- and 5-fold, respectively, at sub-saturating concentrations of malate, ( iv ) among the analogs of L-malate tested as inhibitors of the NAD + -dependent ME catalyzed reaction, small (2- to 3-carbons) organic diacids carrying a 2-hydroxyl/keto group behaved as the most potent inhibitors of ME activity (e.g., oxaloacetate, tartronic acid and oxalate). Conclusions The biochemical characterization of Anopheles stephensi ME is of critical relevance given its important role in bioenergetics, suggesting that it is a suitable target for insecticide development.

Sujets

Informations

Publié par
Publié le 01 janvier 2011
Nombre de lectures 10
Langue English
Poids de l'ouvrage 4 Mo

Mitochondrial

NAD+-dependent

camil

enzyme

fromAnophelesstephensi:apossiblenovel

targetformalariamosquitocontrol

Pon
etal
.

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318(26October2011)

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

RESEARCHOpenAccess
MitochondrialNAD
+
-dependentmalicenzyme
from
Anophelesstephensi:
apossiblenoveltarget
formalariamosquitocontrol
JenniferPon
1
,EleonoraNapoli
1
,ShirleyLuckhart
3
andCeciliaGiulivi
1,2*

Abstract
Background:
Anophelesstephensi
mitochondrialmalicenzyme(ME)emergedashavingarelevantroleinthe
provisionofpyruvatefortheKrebs

cyclebecauseinhibitionofthisenzymeresultsinthecompleteabrogationof
oxygenuptakebymitochondria.Therefore,theidentificationofMEinmitochondriafromimmortalized
A.stephensi
(ASE)cellsandtheinvestigationofthestereoselectivityofmalateanaloguesarerelevantinunderstandingthe
physiologicalroleofMEincellsofthisimportantmalariaparasitevectoranditspotentialasapossiblenoveltarget
forinsecticidedevelopment.
Methods:
TocharacterizethemitochondrialMEfromimmortalizedASEcells(Mos.43;ASE),massspectrometry
analysesoftrypsinfragmentsofME,genomicsequenceanalysisandbiochemicalassayswereperformedto
identifytheenzymeandevaluateitsactivityintermsofcofactordependencyandinhibitorpreference.
Results:
TheencodinggenesequenceandprimarysequencesofseveralpeptidesfrommitochondrialMEwere
foundtobehighlyhomologoustothemitochondrialMEfrom
Anophelesgambiae
(98%)and59%homologousto
+themitochondrialNADP-dependentMEisoformfrom
Homosapiens
.MeasurementsofMEactivityinmosquito
++mitochondriais
2
o
+
latedfr
2
o
+
mASEcellsshowedthat(
i
)
V
max
withNADw
2
a
+
s3-foldhigherthanthatwithNA
2
D
+
P,(
ii
)
additionofMgorMnincreasedthe
V
max
by9-to21-fold,withMn2.3-foldmoreeffectivethanMg,(
iii
)
succinateandfumarateincreasedtheactivityby2-and5-fold,respectively,atsub-saturatingconcentrationsof
malate,(
iv
)amongtheanalogsofL-malatetestedasinhibitorsoftheNAD
+
-dependentMEcatalyzedreaction,
small(2-to3-carbons)organicdiacidscarryinga2-hydroxyl/ketogroupbehavedasthemostpotentinhibitorsof
MEactivity(e.g.,oxaloacetate,tartronicacidandoxalate).
Conclusions:
Thebiochemicalcharacterizationof
Anophelesstephensi
MEisofcriticalrelevancegivenitsimportant
roleinbioenergetics,suggestingthatitisasuitabletargetforinsecticidedevelopment.
Keywords:
malaria,mitochondria,bioenergetics,metabolism,inhibitors,mosquitoes

Background
forflightmetabolisminthetsetsefly[3],themosquito
Recently,severalpathwaysforenergyproductionhave
Aedesaegypti
[4]aswellasotherinsects[5].About20%
beenidentifiedinmitochondriafrom
Anophelesste-
oftheglutamateproducedbyprolineoxidationisin
phensi
[1],awell-studied
Anopheles
speciesintheinves-turnoxidizedbyglutamatedehydrogenase[6],whereas
tigationofmalariatransmission[2].Themitochondria-theremainderundergoestransaminationbyreaction
dependentenergypathwaysmainlyuseproline,pyru-withpyruvateandtheresultingalanineaccumulatesas
vate,
a
-glycerophosphate,andacyl-carnitinederivativestheprolineisutilized.The2-oxoglutarateformedby
assuitablesubstrates.ProlineisalsothemainsubstratetransaminationisfurthermetabolizedbytheKrebs

cycle.Originallypyruvatewasthoughttobeproduced
*Correspondence:cgiulivi@ucdavis.edu
fromoxaloacetatebyanoxaloacetatedecarboxylase[7],
1
DepartmentofMolecularBiosciences,SchoolofVeterinaryMedicine,
butthisenzymewaslaterlocalizedinthecytoplasm
UniversityofCaliforniaDavis,Davis,CA95616,USA
whereasprolineoxidationandsubsequentreactionsall
Fulllistofauthorinformationisavailableattheendofthearticle
©2011Ponetal;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommons
AttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductionin
anymedium,providedtheoriginalworkisproperlycited.

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

takeplaceinthemitochondria[6],consistentwithpre-
viousstudies[1].Mitochondriaofculturedcells[ASE
cellline(
A.stephensi
Mos.43cellline)]from
A
.
ste-
phensi
,aswellasflightmusclemitochondriaofabeetle
(
Popilliajaponica
),whichalsohavetheabilitytooxidize
prolineatahighrate,havebeenshowntocontainan
unusuallyactivemalicenzyme[8].Thelatterspecies
utilizesNAD
+
preferentiallyasacoenzymeandpresum-
ablyproducespyruvatebytheoxidativedecarboxylation
ofmalate[8].Thismitochondrialenzymeininsectsmay
haveacriticalroleinthereplenishmentofpyruvatefor
eithertransaminationorKrebs

cycle.
Malicenzyme(ME;EC1.1.1.39)catalysesthereversi-
bleoxidativedecarboxylationof
L
-malatetopyruvate
andCO
2
withtheconcomitantreductionofthecofac-
torNAD
+
orNADP
+
[9-11].Theenzymerequires
divalentcations(Mg
2+
,Mn
2+
,orothers)inthecataly-
sisofthisreaction.MEactivitywasfirstisolatedfrom
pigeonliver[12]andhassincebeenfoundinmostliv-
ingorganisms,frombacteriatohumans.MostMEsare
homotetramers,withmonomerscontaining550amino
acidsandhavingmolecularweightsof60kDa.The
aminoacidsequencesofMEsarehighlyconserved
acrossallstudiedorganisms,buttheylackrecognizable
homologytootherproteins,includingotheroxidative
decarboxylases.ThewidedistributionofMEactivityin
natureandthehighdegreeofsequenceconservation
areconsistentwiththeimportantbiologicalfunctions
oftheseenzymes,suchasphotosynthesisinC4plants
andevensomeC3plants[13]andbiosynthesisoffatty
acidsandsteroidsinliverandadiposetissuesinani-
mals.Inmammals,threeisoformsofMEhavebeen
identified

cytosolicNADP
+
-dependentME(ME-1;
[14]),mitochondrialNADP
+
-dependentME(ME-3;
[15]),andmitochondrialNAD(P)
+
-dependentME
(ME-2;[10]),whichcanuseeitherNAD
+
andNADP
+
asacofactor(dualspecificity),butprefersNAD
+
underphysiologicalconditions.Ininvertebrates,andin
particularininsects,unusuallyhighactivityofNAD
+
-linkedmalicenzymehasbeenreportedinflightmus-
clemitochondriaofthebeetle
Popilliajaponica
[8]
andfromthetsetseflyandotherinsects[16].
Basedonpreviousreports[1],ASEmitochondrialME
emergedashavingarelevantroleintheprovisionof
pyruvatefortheKrebs

cyclebecausethechemicalinhi-
bitionofthisenzymeresultedinthecompleteabroga-
tionofoxygenuptakebymitochondria.Therefore,the
identificationofMEinASEmitochondriaandtheinves-
tigationofthestereoselectivityofmalateanaloguesare
relevantinunderstandingthephysiologicalroleofME
incellsofthisimportantmalariaparasitevectorandits
potentialasapossiblenoveltargetforinsecticide
development.

Page2of12

Methods
Chemicals
OrganicacidswerepurchasedfromSigmaChemicalCo.
(St.Louis,USA).Allreagentswereofanalyticalgrade.
Cellmaintenance
Theimmortalized
A.stephensi
ASEcelllinewasgrown
inmodifiedEagle

sminimalessentialmedium("E5

)
supplementedwithglucose,L-glutamine,vitaminsolu-
tion,nonessentialaminoacids,penicillinandstreptomy-
cin,and5%heat-inactivatedfetalbovineserumat28°C
with5%CO
2
asdescribed[1].Thepopulationdoubling
timeofthesecellsisapproximately18-20h.Thecells
weresplit1:10intoE5mediumandgrownin50mlcul-
tureflasksuntilconfluent.Theseflaskswereusedto
seed500-mlcultureflaskstoprepare~2billioncellsfor
mitochondriapreparation.Cellsharvestedformitochon-
driapreparationweregentlypipetted,resuspendedin
themedium,andtransferredto50-mltubes.Cellswere
pelletedbycentrifugationat800gfor5min.Thesuper-
natantwasremoved,andthecellswereresuspendedin
asmallamountofmediumbygentlepipettingand
transferredtoasterileholdingtubeonice.Thiscycle
wasrepeated,withcollectionoftheconcentratedcells
intoonetube,untilallflaskswereprocessed.
Isolationofmitochondria
Cellswerecentrifugedfor1minat500gat4°Cand
mitochondriawereisolatedfrompelletedcellsas
described[1].ThepelletwasweighedandMSHEbuffer
wasaddedataratioof3mlofMSHEbuffer(220mM
mannitol,70mMsucrose,0.5mMEGTA,0.1%fatty
acid-freebovinealbumin,and2mMHEPES,pH7.4)
per1gofcells.Thecellsweredisruptedbygentle
homogenization,centrifugedat600gfor5minat4°C,
thepelletwasdiscarded,andthesupernatantwascentri-
fugedat10,300gfor10minat4°C.Thepellet,whichis
richinmitochondria,wasresuspendedinasmall
volumeofMSHE.Usingthisproceduretheyieldwas
7.5±0.5
μ
gmitochondrialprotein/10
6
cells.Protein
concentrationwasdeterminedbyusingtheBCAProtein
Assay(Pierce).
Enzymaticassays
TheMEenzymaticassaywasperformedusingamethod
outlinedby[17]withthefollowingmodifications.Mos-
quitomitochondriawerehomogenizedinMSHEcon-
taining2mMmercaptoethanol.Ina1mlcuvette,2-
μ
g/
mlantimycin,1mML-malate,0.3mMNAD
+
,50mM
HEPES(pH7.8),and3mMofMnCl
2
(unlessindicated
otherwise)wereadded.Thereactionwasinitiatedwith
theadditionof40
μ
gofmosquitomitochondriallysate
protein.Thechangeinabsorbancewasmeasuredusing

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

theCary1ESpectrophotometerat340nmfor2-3min-
utes(control).TheMEspecificactivitywascalculated
foreachtraceutilizingtheextinctioncoefficientfor
NADHat340nm(6.22mM
-1
cm
-1
).EachL-malate
analoguewasaddedtothereactionmixtureattheindi-
catedconcentrationsandthechangeinabsorbancewas
measuredforanother3-5minutes.
Massspectrometryanalysis,proteinidentification,and
confirmationof
A.stephensi
encodingsequence
LC-MS/MSanalyseswereperformedattheProteomics
FacilityoftheUniversityofCaliforniaDavisGenome
Center.TandemmassspectrawereextractedbyBio-
Worksversion3.3.Chargestatedeconvolutionanddei-
sotopingwerenotperformed.AllMS/MSsampleswere
analysedusingX!Tandem[18,19].X!Tandemwasset
uptosearchtheEnsemble
A.gambiae
proteindatabase
(13,740entries)assumingthedigestionenzymetrypsin.
X!Tandemwassearchedwithafragmentionmasstol-
eranceof0.40Daandaparentiontoleranceof1.8Da.
IodoacetamidederivativeofcysteinewasspecifiedinX!
Tandemasafixedmodification.DeamidationofAsn
andGln,oxidationofMetandTrp,sulphoneofMet,
TrpoxidationtoformylkynureninofTrpandacetylation
ofthe
N
-terminuswerespecifiedinX!Tandemasvari-
ablemodifications.Scaffold(versionScaffold-3_00_08)
wasusedtovalidateMS/MSbasedpeptideandprotein
identifications.Peptideidentificationswereacceptedif
theycouldbeestablishedatgreaterthan90.0%probabil-
ityasspecifiedbythePeptideProphetalgorithm[20].
Proteinidentificationswereacceptediftheycouldbe
establishedatgreaterthan99.0%probabilityandcon-
tainedatleast2identifiedpeptides.Proteinprobabilities
wereassignedbytheProteinProphetalgorithm[21].
Proteinsthatcontainedsimilarpeptidesandcouldnot

Page3of12

bedifferentiatedbasedonMS/MSanalysisalonewere
groupedtosatisfytheprinciplesofparsimony.
Toidentifythefull-lengthcodingsequencefor
A.ste-
phensi
ME,the
A.gambiae
MEaminoacidsequence
(Q7QB64)wasusedasaquerybyDr.Zhijian(Jake)Tu
(VirginiaTech)toidentifyhomologoussequenceinthe
June2010unpublisheddraftofthe
A.stephensi
assem-
blybyTBLASTN(e-valuecutoff1e
-7
).Theregionwith
thebestmatchplus1-kbflankingsequencesoneither
sidewereretrieved.
Statisticalanalyses
Theexperimentswereruninduplicateortriplicateand
repeatedtwotimesinindependentexperiments.Data
wereexpressedasmean±SEM.Thedatawereevalu-
atedbyusingthe
t
-test(StatSimplev2.0.5;NidusTech-
nologies,Toronto,Canada)with
p

0.05consideredas
statisticallysignificant.
Resultsanddiscussion
IdentificationofthemitochondrialMEfrom
A.stephensi
TocharacterizetheMEin
A.stephensi
cells,Multidi-
mensionalProteinIdentificationTechnology(MudPIT)
wasused.Thistechniqueinvolvesdigestingmosquito
proteinswithtrypsinandseparatingtheresultingpep-
tideswithtwoliquidcolumnchromatographysteps:the
firstbeingastrongcationicexchange,andthesecond
beingreversed-phaseHPLC.Asthepeptideselutefrom
thesecondcolumn,theyaresprayedintoalinearion
trapmassspectrometer.ThefirstMSscanassigned
eachpeptideamass/chargeratio.Themostintensepep-
tidesignalsarethenfragmentedinasecondMS/MS
scan,whichassignseachpeptideaunique

fingerprint

(Figure1).Thefingerprintsarethenanalysedagainst

Figure1
MassspectrometryanalysesofASEME
.MassspectrometryresultsfromMudPITanalysesindicatingthepeptidesequences,
identificationprobabilityto
A.gambiae
Q7QB64,peptidemodificationsidentifiedbyspectrum(carboxymethylatedorCM,methioninesulfoxide
orOX,anddeaminatedorDA),calculatedandactualpeptidemasses,andpeptidestartandendaminoacidfromtheaboveindicated
A.
gambiae
sequence.

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

bioinformaticsdatabases(NCBI)thatrevealedthepro-
tein

sidentity.
Sixoutofsevenuniquepeptideshada

90%match
withoneoftwoMEsreportedfor
A.gambiae
(Q7QB64;
Figure1);onlyoneofthesepeptidesshowedapeptide
identificationprobabilityofapproximately51%.These
peptidesprovided26%proteincoverage(148amino
acidsoutof572;Figure2).Giventhehighhomologyof
thesepeptidesto
A.gambiae
Q7QB64(Figure2),this
sequencewasusedtoquerythe
A.stephensi
genome
assembly.Thisanalysisretrievedthehomologous
A.ste-
phensi
MEsequence(Additionalfile1).Thepredicted
A.stephensi
proteinsequence(Additionalfile1)revealed
a98%homology(99%positives)with
A.gambiae
ME
(Q7QB64;Additionalfile1).
BLASTqueryofthecompleteaminoacidsequenceof
A.stephensi
MEtothenon-redundantNCBIdatabase
revealedhighidentitytoMEsfrommosquitoes(
A.gam-
biae96%,Anophelesdarlingi
94%,
Aedesaegypti
88%,
Culexpungens
87%),flies(
Drosophila
spp.75%,
Glossina

Page4of12

spp.62%),ants(
Harpegnathos
spp.74%;
Camponotus
spp.74%),lice(
Pediculus
spp.73%),moth(
Bombyx
spp.
64%),andanematode(
Ascarissuum
54%).BLAST
queryagainstthehumandatabaserevealeda59%iden-
titytothehumanmitochondrialNADP
+
-dependentME
(orME-3),followedbya57%anda55%identitytothe
humancytosolicME-1andthehumanmitochondrial
ME-2,respectively.Thelowerhomologyofthe
A.stepe-
phensi
MEwiththehumanmitochondrialcounterpart
wasnotduetothemitochondrialtargetingsequence
becauseitsremovalfromthehumanisoformandrea-
lignmentofthematureproteinwiththemosquitopro-
teindidnotchangetheidentity(59%),similarity(75%)
orscore(1742).Ofnote,thetwohumanmitochondrial
MEisoformsshareonly54%oftheiraminoacids,which
isthesamerangeofhomologywhencomparinghuman
MEswith
A.stephensi
ME(thisstudy)orfrommaize
chloroplasts(47%).Bycomparisonthehighlyhomolo-
gous
A.gambiae
ME(Q7QB64)is59%identicaltothe
humanME-3,57%tohumanME-1,and55%tohuman

Figure2
Proteinalignmentsof
A.gambiae
MEs
.SequencealignmentofthetwoMEproteinsfrom
Anophelesgambiae
(Q7PRY4andQ7QB64).
SequencealignmentsofthetwoMEproteins(sequencesasreportedintheSwissProtdatabase;lastvisitedonApril13,2011).Thealignments
weregeneratedusingCLUSTAL[55].BoxedaminoacidsindicatethepeptidesidentifiedbymassspectrometryfromFigure1.

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

ME-2,whereastheotherMEfrom
A.gambiae
(Q7PRY4)has60%identitytothehumanME-1,58%to
ME-3,and53%toME-2.Asin
A.gambiae
,twoMEs
havebeenidentifiedfor
A.darlingi
,inwhichoneof
them(E3WUS4)is62%identicaltothehumanME-1,
whiletheotherME(E3WW73)is60%identicaltothe
humanME-3.Thus,forthesemosquitospecies,noME-
2ortholog(mitochondrialandNAD
+
-dependent)is
apparent.Inagreementwiththesefindings,onlyone
mitochondrialMEfrom
A.stephensi
wasdetectedinthe
performedMudPITanalysesandgenomesequenceana-
lyses.However,apossible
A.stephensi
MEparalogwith
approximately60%identityto
A.gambiae
ME
(Q7QB64)wasdetected(Dr.ZhijianTu,personalcom-
munication).Whilethisencodedproteinispotentially
interesting,wehavefocusedonthehighlyhomologous
A.stephensi
MEhere.
Withtheprimarysequenceofthe
A.stephensi
ME,a
high-qualitypredictionofthe3Dstructureandbiologi-
calfunctionoftheMEwasperformedbyusingatem-
plate-basedmodellingplatform[22,23].Theresultsof
thismodelling(Additionalfile2)revealedthatthe3D
structureof
A.stephensi
ME(withaC-scoreof1.985)
wascloselyrelatedtothatof
Ascarissuum
(1o0sA;TM
score0.9645;RMSD1.43,identity0.53;coverage99%;
Figure3).This3Dmodelof
A.stephensi
MEwasbased
ontemplatesfromhumanME-1(2aw5B),humanME-2
(1pj3Aand1gz4A)and
Ascarissuum
(1llqAand1o0sA;
Additionalfile2).Themodellingresultsalsopredicted
thebindingsiteforvariousligands(Additionalfile2).In
particular,thepredictedbindingsitesforMn
2+
,oxalate
andNADP
+
resemblethoseinpigliverME-1,whereas
thosefortartronate,malate,fumarateandMgresembled

Figure3
Three-dimensionalstructuresof
A.stephensi
and
Ascarissuum
MEs
.The3Dstructureof
A.stephensi
ME(A)was
obtainedthroughmodellingfromtheprimarysequencefrom
Additionalfile1andtheuseofthemodelingsoftwareI-TASSER
(resultsshowninAdditionalfile2).The3Dstructureof
Ascarissuum
MEwasobtainedfromtheproteindatabaseandcorrespondstothe
accessionnumber1o0sA.Bothpdbfileswerevisualizedusing
PyMol1.4.1[56].

Page5of12

thoseinhumanME-2.Thebindingsitefortartronate
wasalsosimilartothatdefinedfor
Ascarissuum
ME.
TogaininsightintheactivityofME,enzymaticpara-
meterswereevaluatedinmosquitomitochondriaiso-
latedfromASEcells.TheproductionofNAD(P)Hwas
followedinthepresenceofmalateandNAD(P)
+
.The
K
m
valueforL-malatewascalculatedasafunctionof
totalconcentrationsofthesubstrateanditwasfoundto
be0.12mMatthepHof7.8(Table1).Alowactivity
wasobtainedintheabsenceofcofactors(Mg
2+
orMn
2
+),buttheadditionofeithermetalincreasedsignifi-
cantlythemaximumactivity(9to21-fold);however,
Mn
2+
wasabettercofactorthanMg
2+
(2.3-foldofMg
2+
alone;Table1).ThespecificactivityoftheASEME
enzymewassignificantlyhigherthanthatoftheother
twohumanmitochondrialisoforms:thespecificactiv-
itiesofthehumanmitochondrialNADP
+
-MEandNAD
+
-MEhavebeenreportedas12U/mgprotein(1Uor
unit=1
μ
mol/min)and35U/mgprotein,respectively,
+whereastheactivityofthehumancytosolicNADP-ME
is40to55U/mg[15].ASEMEcouldutilizeeither
NAD
+
orNADP
+
;however,
V
max
withNAD
+
was3-
timeshigherthanthatwithNADP
+
(Table1),consistent
withthepredictionsmadefromtheproteinsimilarity
search(
videsupra
)andstructure(
videinfra
).
Ithasbeenshownthatthepresenceofbasicresidues
atpositions67and91ofhumanME-2areapparently
criticalforfumarate-dependentactivation[24](Figure
4A,residuesmarkedwithasterisks).Allthreehuman
isoformscontainArgattheequivalentpositions[24]
butonlyME-2isactivatedbyfumarate.Thus,otherfac-
torsmaycontributetothisactivation[24].Themutation
oftheaminoacidresidueAsp-102hadasignificant
effectonthefumarate-mediatedactivationofhuman
ME-2[25].Atthisposition,thehumanME-1andME-3
isoformshaveaSerresidueandtheydonotshowany
increaseinactivitywithfumarate.ThemosquitoME
hasaGluatthisposition,andthe
Ascarissuum
MEhas
anAsp,suggestingthattheseisoforms,likethehuman
ME-2,maybeactivatedbyfumaratebyconservingthe
negativechargeatthisposition.Ontheotherhand,in
themosquitoME,Leuoccupiestheequivalentposition
ofArg-67(allhumanisoformsandthenematodeME
haveanArginthisposition),suggestingthatmosquito
MEshouldnotbeactivatedbyfumarate.Inadditionto
Arg-67,ithasbeenreportedthatinthenematodemito-
chondrialME,theresiduesthatareinvolvedinthe
bindingoffumarateareAla-78,Leu-81andLeu-105
(Figure4A,rowsindicatedwitharrows),whichin
A.ste-
phensi
MEareoccupiedbyIle,LeuandLeuinagree-
mentwiththenematodeisoform,whichisactivatedby
fumarate.Inagreementwiththislastprediction,succi-
nateandfumarateincreasedtheactivityby2-to5-fold,
respectively,atsub-saturatingconcentrationsofmalate

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

Page6of12

Table1EffectofcofactorsonV
max
ofMEactivity
Addition
K
m
formalate
V
max
p
-valuetocontrols
(mM)nmol×(min×mgprotein)
-1
None-3±1-
MgCl
2
(3mM)-24±20.02
MnCl
2
(3mM)0.1256±3(202)0.01
AllkineticparameterswereobtainedwithNAD
+
exceptthe
V
max
obtainedwithNADP
+
(valueinparentheses).The
V
max
valueswereexpressedasmean±SEM.
AllotherexperimentaldetailsaredescribedintheMethods.

(Figure5B),showingnoeffectatsaturatingones(not
shown).Otheraminoacids,suchasPro,hadamoderate
effectonenzymeactivityuptoconcentrationsof10
mM(1.4-fold).Surprisingly,Gluwasfoundtobea
potentactivatorincreasingASEMEactivityby4-foldat
saturatingconcentrationsofmalate(Figure5B).
Oneveryhighlyconservedregiondefinedasapartofa
malate-bindingdomainisshowninFigure4B.This
regionhasbeenidentifiedonthebasisofinhibition
experimentscarriedoutonNADP
+
-dependentMEwith
thecompetitiveinhibitorbromopyruvate[15].Indeed,a
highlyconservedCys(Figure4B,asterisk)andthepre-
cedingsequenceVYTPTVG(Figure4B,bar)arepresent
inthisdomaininhuman,nematode,plantand
A.gam-
biae
and
A.stephensi
mosquitoisoforms(Figure4B),
suggestingthatthesametypeofbindingoccursinthe
mosquitoisoforms.However,recentlytheroleofthis
Cyshasbeenchallengedbasedoncrystallographydata
[24].Thisstudyindicatedthatthisresidueisabout13Å
awayfromthesubstrateanalogoxalatesuggestingthat
theinhibitionofsubstratebindinguponchemicalmodi-
ficationofthisCysisanindirecteffect.
TheproposedADPbindingbeta-alpha-betafoldis
shownwhichcontainsacharacteristicarrangementof
Glyresiduesandnonpolarandhydrophilicaminoacids
atspecificpositions(Figure4C,boxedaminoacids;
[26]).AhighlyconservedCysresidue(orthehighly
similarSerinplantNADPME-2;Figure4C,asterisk)
observedinthisregionispresentinallNADP
+
-depen-
dentmalicenzymes[27]butisabsentfromtheNAD
+
-dependentmalicenzymes(humanME-2,nematode
andmosquitoisoforms).Thereforethepresenceor
absenceofthisCysorSerresidueinthisbindingfold
+apparentlydistinguishesthepreferenceforNADPover
NAD
+
assuggestedbeforebyothers[28].
ThefingerprintregionofNAD(P)
+
-bindingsitesis
characterizedbyaGly-richsequence(Gly-X-Gly-X-X-
Gly;Figure4D,boxedaminoacids),whichisthephos-
phatebindingconsensus[29,30].ThefirstGlyallows
thetightturnofthemainchain,thesecondallowsa
closecontacttothediphosphateofNAD(P)
+
,andthe
thirdisimportantfortheclosepackingofthesecondary
structure[29,31].ThisthirdGlyisreplacedbylarger
aminoacids(AlaorSer)inenzymesthatutilizeNADP
+

(e.g.,glutathionereductase,thioredoxinreductase)
becauseitisbelievedthatdisruptingtheclosepacking
ofthesecondarystructureallowsthestructureto
accommodatethephosphatemoietyofNADP
+
[29].
Furthermore,proteinengineeringofglutathionereduc-
tase,inwhichAla-179wasreplacedbyaGlyresidueat
anequivalentpositioninthisenzyme,bothinsingleas
+wellasinmultiplesubstitutions,causedNADbinding
preferenceoverthatofNADP
+
[32].However,thisrea-
soningdoesnotseemtoapplytoallMEslistedinFig-
ure4D(positionmarkedwithasterisk).Inparticular,all
humanMEshaveathirdGlyregardlessofthecofactor
preference,indicatingthatotherfactorsarecontributing
tothecofactorpreference.Insupportofthisstatement,
GlyresiduesintheboxedsegmentsshowninFigures
4C&4DwhenreplacedbyValresiduesrenderedabor-
tivemutantsofmaizeC
4
NADP
+
-ME[31].Thisiscon-
sistentwiththethree-dimensionalmodelobtainedfor
themaizeC
4
NADP
+
-MEshowingthatresiduesatboth
sitesarepartoftheNADP
+
-bindingsitejustifyingthe
+highdegreeofconservationamongallNAD(P)-ME
13[.]IthasbeensuggestedthattwoAlaresiduesspacedby
threeaminoacids(Gly-X-Gly-X-X-Ala-X-X-X-Ala;Fig-
ure4Dindicatedwithabar)aremorecharacteristicfor
NADP
+
-bindingdomainsrelativetoNAD
+
-binding
domains,whereasGlyresiduesweremoreconsistently
identifiedinNAD
+
-dependentenzymes[31,32].How-
ever,allhumanMEisoforms,nematodemitochondrial
MEand
A.stephensi
MEhaveanAla-X-X-X-Alaseg-
ment,regardlessofwhethertheyuseNADP
+
orNAD
+
.
Otherresiduesmayalsoaffectbindingspecificityof
NAD(P)
+
inMEs.IthasbeennotedthatallNADP
+
-
dependentMEisoformshaveaconservedLysatGln-362
inhumanME-2(Figure4E,asterisk;[33]),whereasdual
specificityMEshaveheterogeneousresiduesatthisposi-
tion[34](Figure4E,asterisk).Thisobservation,combined
withmutationalandmodellingstudiesusingpigeonME-1,
ledtothesuppositionthatinNADP
+
-dependentMEs,
NADP
+
specificityisconferredbyLysatthisposition.
MitochondrialNAD
+
-MEsof
Ascaris
andotherspecies
mayhavenoneedtoexhibitstrictspecificityforNAD
+
duetothehighconcentrationofNAD
+
relativetoNADP
+
withinmitochondria,andthustheresidueatthe

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

Page7of12

Figure4
AnalysesofproteinsequencealignmentsofvariousMEs
.Sequencealignmentsofmaize(NADP
+
-ME-2,chloroplasticNADP
+
-ME,
andNADP
+
-ME-4),human(ME-1,ME-2,andME-3),nematode(NAD
+
-ME-2),bacterial(NAD
+
-ME)and
A.stephensi
ME(mosquitoME).Theamino
acidsequencesoftheMEisoformswereanalyzedbyBLASTagainsttheSwissProtdatabase,andthealignmentsweregeneratedusingCLUSTAL
[55].Theaminoacidresidueshighlightedingreysharehighhomologywhereasthoseinboldareidentical.Keyaminoacidsdiscussedinthe
textareshownwithasterisks,arrowsorbars(seetextforfulldescription).(A)Sitesinvolvedinfumarateactivation;(B)Malatebindingsite;(C)
ADPbindingsite;(D)NAD(P)
+
bindingdomainandsitesassociatedwithNAD
+
versusNADP
+
preference;(E)Othersitesalsoassociatedwith
++NADversusNADPpreference.ForcompleteproteinnomenclatureseeAdditionalfile3.

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

Page8of12

Figure5
EffectofL-malateanaloguesonMEactivity
.A.ChemicalstructureofL-malateandstructuralanalogues.B.EffectofL-malate
analoguesonASEmitochondrialMEenzymaticactivity.TheactivityofMEwasevaluatedintheabsenceandpresenceofthefollowing
inhibitors/compoundsatsaturatingconcentrationsofallothercofactorsandsubstratesexceptforsuccinateandfumaratewhichareshownat
subsaturatingconcentrationsofmalate(0.5mM):1mMoxaloacetate;5mMtartronicacid;1mMoxalate;2.5mMalpha-ketobutyrate;2.5mM
beta-hydroxybutyrate;2.5mMalpha-hydroxybutyrate;2.5mMacetoacetate;2.5mMmalonate;10mMpyruvate;10mMproline;10mM
succinate;10mMglutamate;and5mMfumarate.

correspondingpositionwouldnotnecessarilybecon-malatewith
K
i
valuesfrom80to430
μ
M[24,35].
strainedtoLys.Indeed,inmosquitoME(thisstudy),AccordingtoHsieh
etal
[36],thepresenceofLys-346
nematodeME,humanME-2andbacteriaME,thisinmammalianME-2iscriticalforATPinhibition(Fig-
sequencepositioncorrespondstoaLys,His,GlnandGlu,ure4E,arrow)becausesite-directedmutagenesisofthis
respectively(Figure4E,asterisk).Recentdataclearlyindi-residuetoAlaorSerdiminishesATP-mediatedinhibi-
catethattheGln-362-LysmutantofhumanME-2isation.Thisresultsuggestedthatthepositivechargeis
non-allosteric,non-cooperativeandNADP
+
-specificcriticalforATPbinding.Asindicatedabove,inthe
enzyme,asisME-1[33].Additionally,Gln-362-LysishumanME-1andME-3aswellastheplantMEs,iniso-
moresensitivetoATP,andtheinhibitionconstantisformsthatarenotinhibitedbyATP,theequivalent
smallerthanthatofwild-type[33].SequencealignmentspositionsforLys-346areoccupiedbySer.Inthemito-
ofthenucleotide-bindingregionamongMEshavechondrialnematodeandmosquitoMEs,Ileresiduesare
revealedthat,inadditiontoGln-362,Lys-346isconservedpresentinthesepositions,thusnosignificanteffectof
amongtheNAD
+
-dependentmalicenzymes(humanME-ATPwouldbeexpectedasithasbeenshownforthe
2;Figure4Earrow),butinNADP
+
-dependentMEs
Ascaris
ME[37,38].Inagreementwiththisprediction,
(maize,human),thisLysisreplacedbySer(Figure4EkineticresultsshowedthatATPwasnotaninhibitor
arrow).Inthecaseof
A.stephensi
ME,thisresidueiswithrespecttoNAD
+
orL-malatein
A.stephensi
mito-
replacedbyIle,asitisinthecaseofthenematodeME.chondrialMEattheconcentrationstested(20to1,000
Thus,inthoseisoformswithNADP
+
-onlyspecificity,a
μ
M;datanotshown).TheArg-197,Arg-542andArg-
Lysinposition362(Figure4Easterisk)andaSerin346556residuesinhumanME-2appeartobeinvolvedin
(Figure4Earrow)areapparentlynecessary.thebindingofATPattheexosite[24].Theseresidues
ATPactsasanactive-siteinhibitorofmitochondrialdifferfromthoseinhumanME-3ormosquitoME
ME-2followingacompetitivemechanismforNAD
+
and(humanME-3=Gln,TyrandLeuandmosquitoME=

Pon
etal
.
MalariaJournal
2011,
10
:318
http://www.malariajournal.com/content/10/1/318

Gln,ThrandHis),suggestingthatthisbindingisnot
operationalintheseisoforms.
Inhibitorsofthe
A.stephensi
mitochondrialME
SeveralanaloguesofL-malateweretestedasinhibitors
oftheNAD
+
-dependentMEcatalyzedreaction(Figure
5A).Oxalate,oxaloacetate,tartronateandalpha-ketobu-
tyratewerefoundtobeinhibitorsoftheME-catalyzed
reaction,throughacompetitivemechanismwithrespect
tothesubstrate(malate)concentration.Their
K
i
values
rangedfrom0.1(oxaloacetateandoxalate)to0.6
μ
M
(tartronate;Table2).Alpha-hydroxybutyrateandthe
ketonebodiesbeta-hydroxybutyrateandacetoacetatedid
notexhibitanyeffectontheMEactivity.Thesameout-
comewasobtainedwitheithermalonateorpyruvate
(Figure5B).
ThecatalysisbyMEgenerallyproceedsintwosteps,
namelythedehydrogenationofmalatetoproduceoxa-
loacetate,andthenthedecarboxylationtoproducepyru-
vate.Inaproposedenzymaticcatalysisbasedon
structuraldatathecrystallographydata,thelatterstepis
believedtooccurthroughtheformationofaketo-enol
intermediatestructurallysimilartooxalate[39].The
strongestMEinhibitorsforwhich
K
i
wereevaluated(i.
e.,oxalate,oxaloacetate,andtartronate)areconsistent
withthismechanism,forthesecompoundsareeither
intermediates(i.e.,oxaloacetate,oxalate)orstructurally
similarintermediates(tartronateorketomalonate)that
fitthetightactivesiteoftheenzyme[39].Besidesthe
roleofsize,theseinhibitorsarediacids.Thus,thepre-
senceofanextracarboxylgroupappearstobeneces-
saryforbindingthesubstrate(orinhibitor)tothe
enzyme,probablythroughaninduceddipolartypebond
withthesidechainsofGlnandAsn,assuggestedby
others[40].Insupportofthisrationale,thepresenceof
theextracarboxylateinmalateandoxaloacetatecom-
paredtoalpha-hydroxybutyrateandalpha-ketobutyrate,

Table2Inhibitionconstantsofvariousmalateanalogues
Compound
K
i
(
μ
M)
Oxaloacetate0.12
Oxalate0.15
Tartronate0.61
Alpha-ketobutyraten.d.
InhibitionstudieswereperformedbyvaryingtheconcentrationsofL-malate
from0.1mM(subsaturating)toupto1mM(saturating)whiletheanalogue
(chosenbasedonitsinhibitoryactivityfromFigure3B)washeldatseveral
fixedconcentrationsarounditsinhibitionconstant.Concentrationsofthe
othercomponentsintheassaymixturewereheldatsaturation.Reciprocal
velocitieswereplottedasafunctionofreciprocalsubstrateconcentrations
andallplotswerelinearintherangeofsubstrateconcentrationsassayed(0.1-
0.5mM).Thelineswerecalculatedfromthefitsoftheexperimentally
determinedvaluestotheappropriatekineticparameters.Abbreviations:n.d,
notdetermined.

Page9of12

respectively,yieldedstrongerinhibitionofME(Figure
.)5Withinthediacids,thepresenceofaketogroupin
position2seemedtobeanimportantfactorincontrol-
lingtheinhibitoryeffectofthecompoundsstudied,for
lackofthismoietyin,forexample,succinateorfuma-
rate,resultedinalossoftheinhibitoryeffect.The
changeofthe2-ketogroupofthealpha-ketobutyricacid
to2-hydroxy(alpha-hydroxybutyrate)orthetransferto
the3-hydroxy(beta-hydroxybutyrate)resultedinnegligi-
bleeffectontheMEactivity(Figure3B).Thus,theketo
groupinposition2seemsessentialforeffectiveinhibi-
tionofME,possiblybyfavoring
π
-
π
interactionswith
thenicotinamideringofNAD
+
and/orhydrogenbond-
ingwiththesidechainamideofAsnandthe2

-hydro-
xylofthenicotinamideribose[39].Theseresults
suggestthattheextentofinhibitionwasdependenton
thesizeoftheanalogues(2to4-carbons),thepresence
oftwocarboxylgroupsalongwitha2-hydroxylor2-
ketomoietyimportantforbindingofthesubstrateana-
loguetotheenzyme.Inastudyperformedwiththe
enzymefrom
Flaveriatrinervia
,thepresenceofagroup
withalow
p
Kvalue(6to6.6),probablyanHresidue,
wasresponsibleforthebindingofthe2-hydroxylof
malateandtransferofthehydridetoformtheinter-
mediateoxaloacetate[41].Theseresultssuggestthatthe
extentofinhibitionwasdependentonthesizeofthe
analogues(2to3-carbons),thepresenceoftwocarboxyl
groupsalongwitha2-hydroxylor2-ketomoietyimpor-
tantforbindingofthesubstrateanaloguetothe
enzyme.
TheME-3geneisconservedinhuman,chimpanzee,
dog,cow,mouse,rat,avarietyoffliesandmosquito
species,
Caenorhabditiselegans,Arabidopsisthaliana
,
andrice.ThemammalianME-3isoformhasastrong
tissue-specificexpression,mostlyinorganswithalow
divisionrate(e.g.,heart,skeletalmuscleandbrain;[15])
orwithaninvolvementinsteroidhormonebiosynthesis
(ovaryandtestes).HighactivityofME-3hasalsobeen
reportedinmuscleofcrustaceansandfish[42].Ithas
beenpreviouslyreportedthattheimmortalizedcells
from
A.stephensi
,fromwhichtheMEactivityinthis
studywasevaluated,resembledskeletalmusclecells[1].
Thus,thefindingofahighactivityofthismitochondrial
enzymeisconsistentwiththispreviousstudy.Given
thatthereversereaction(i.e.,thecarboxylationofpyru-
vate)was12-timesslowerthantheforwardreaction
(notshown),theroleforME-3intheanapleroticreac-
tion(pyruvatetomalate)oftheKrebs

cycleisexpected
tobenegligible.
Apreviousreportindicatedthatadditionoftartronic
acidcompletelyinhibitedtheoxygenconsumptionof
Pro-supplementedphosphorylatingmitochondria[1].
ThissuggestsacriticalroleforMEinapathwayto