Role of root hydrophobic barriers in salt exclusion of a mangrove plant Avicennia officinalis

Plant,CellandEnvironment(2014)37,1656–1671doi:10.1111/pce.12272

OriginalArticle

RoleofroothydrophobicbarriersinsaltexclusionofamangroveplantAvicenniaofficinalis

PannagaKrishnamurthy1,2,PavithraA.Jyothi-Prakash1,2,LinQin3,JieHe3,QingsongLin1,Chiang-ShiongLoh1,2&PrakashP.Kumar1

DepartmentofBiologicalSciences,NationalUniversityofSingapore,117543Singapore,2NUSEnvironmentalResearch

Institute(NERI),NationalUniversityofSingapore,117411Singaporeand3NaturalSciencesandScienceEducationAcademicGroup,NationalInstituteofEducation,NanyangTechnologicalUniversity,637616Singapore

1

ABSTRACT

Saltexclusionattherootsandsaltsecretionintheleaveswereexaminedinamangrove,Avicenniaof cinalis.Thenon-secretormangroveBruguieracylindricawasusedforcom-parativestudyofhydrophobicbarrierformationintheroots.Bypass owwasreducedwhenseedlingswerepreviouslytreatedwithhighsaltconcentration.Abiseriateexodermiswasdetectedinthesalt-treatedroots,alongwithanenhanceddepositionofhydrophobicbarriersintheendodermis.ThesebarriersreducedNa+loadingintothexylem,accountingfora90–95%saltexclusioninA.of cinalis.ProminentbarrierswerefoundintherootsofB.cylindricaevenintheabsenceofsalttreatment.AcytochromeP450genethatmayregulatesuberinbiosynthesiswasup-regulatedwithinhoursofsalttreatmentinA.of cinalisrootsandleaves,correspondingwithincreasedsuberindeposition.X-raymicroanalysisshowedpreferentialdepositionofNa+andCl intherootcortexcomparedwiththestele,suggestingthattheendodermisistheprimarysiteofsaltexclusion.Enhancedsaltsecretionandincreasedsuberindepositionsurroundingthesaltglandswereseenintheleaveswithsalttreatment.Overall,thesedatashowthatthedepositionofapoplasticbarriersincreasesresistancetobypass owleadingtoef cientsaltexclusionattherootsinmangroves.

Key-words:bypass ow;Casparianbands;mangroves;saltglands;saltsecretion;suberin;xylemsap.

INTRODUCTION

Salinityisamajorenvironmentalstressfactoraffectingplantgrowthandproductivityworldwide(Flowers&Yeo1995;Munns2002).Na+andCl aretheprincipaltoxicionsinsalinesoilswhichimpartbothosmoticandionicstresstoplants(Blumwald2000;Hasegawaetal.2000;Munns&Tester2008).SurvivalofplantsundertheseconditionsmainlydependsontheshootNa+levels(Yeo&Flowers1982;Coskunetal.2013a).Mangrovesareoneofthemajorgroupsofsalt-tolerantplantsthatthriveunderhighsalineenvironments.

Correspondence.P.P.Kumar.Fax:+6567792486;e-mail:dbskumar@nus.edu.sg1656

Theyhaveevolvedvarioussurvivalstrategiessuchas:accumulationofosmolytes,saltexclusionattheroots,saltsecretionthroughaerialpartsoftheplants,andef cientsaltsequestrationintosenescentleavesand/ortothebark(Tomlinson1986;Zhengetal.1999;Parida&Jha2010).Basedontheseadaptivemechanisms,mangroveshavebeenbroadlyclassi edassaltsecretors(e.g.speciesbelongingtoAvicennia,AegicerasandAcanthus)andnon-secretors(ultra ltrators)(e.g.speciesbelongingtoBruguiera,RhizophoraandCeriops)(Scholanderetal.1962;Scholander1968;Tomlinson1986;Parida&Jha2010).

Saltsecretors lter85–95%ofsaltattheroots,andtheexcessNaClreachingtheleafiseliminatedbysecretionthroughspecializedsaltglandspresentintheleaves(Shimonyetal.1973;Drennan&Pammenter1982;Sobrado2001),therebymaintaininglowNa+levelsinthemesophyllcells.Manystudiesdealingwithsaltsecretioninvariousman-grovespecieshaveshownanincreasedsecretioninresponsetoincreasingexternalNaCl(Ball1988;Sobrado2001;Suarez&Medina2008).Ontheotherhand,ultra ltratorsthatlacksaltglandsareknowntomaintainlowlevelsofNa+inthexylemsapby ltering～99%ofsaltfromthesurroundingseawater,thuseliminatingtheneedforsaltsecretion(Scholander1968).Althoughtheseprocessesinmangroveshavereceivedtheattentionofresearchersfordecades(Drennan&Pammenter1982;Sobrado2001;Mukherjeeetal.2012),themechanismofsaltsecretionattheleavesandultra ltrationattherootsarestillnotwellunderstood.

Rootultra ltrationisawell-knownprocessadaptedbymangroveplantsforexcludingthesaltfromseawater.Theroot ltrationef cienciesofvariousmangrovespecieshavebeenmeasuredusingthepressurebombtechnique(Scholanderetal.1962,1966;Scholander1968).However,allthesestudiesused eld-grownplantsamples,whichwouldhavehadpriorexposuretosaltstressfordifferentdurations.Incontrast,speci ceffectsofsaltstressonroot ltrationcanbestudiedusinggreenhouse-grownseedlings,astheyarenotexposedtopriorsaltstress.Suchstudiesarelackinginmangroves.

Despiteyearsofresearch,themajorpathofNa+entryintothexylemisstillnotclear.Thereisgrowingevidenceintheliteraturetoshowthattheapoplasticbypass owisapromi-nentpathofNa+uptakeandentryintothexyleminmany

©2014JohnWiley&SonsLtd

Saltexclusionandsecretioninmangroves

cropplants,suchasrice(Yeoetal.1987;Malagolietal.2008;Faiyueetal.2010;Krishnamurthyetal.2011;Kronzucker&Britto2011),althoughthispathwasdismissedinthemodelplantArabidopsis(Essahetal.2003).Somestudiesthatattemptedtoexaminethebypass owinmangroveshavereportedcontrastingresults(Lawtonetal.1981;Moonetal.1986).However,itisestablishedthatthemangroveshavedeployedultra ltrationasamajorstrategytoexcludeNa+attheroots.Althoughthemechanismofultra ltrationisnotwellcharacterized,itisunderstoodtobeaprocessinvolvingphysicalbarriers(Scholander1968;Parida&Jha2010).Physi-calbarriers(apoplasticbarriers)intherootsfunctionbyblock-ingtheapoplasticbypass owofionsandwaterintoxylem,thushelpingtoreduceleakageofionsintothesteleinvariousplants(Ma&Peterson2003;Schreiber&Franke2011).

Casparianbands(CBs)andsuberinlamellae(SL)arethecommonapoplasticbarriersintheendodermisandexodermisofroots(Perumulla&Peterson1986;Schreiberetal.1999).TheCBsareformedbythedepositionofhydro-phobicsuberinand/orligninintheporesoftheradialandanticlinalwalls,whereas,SLandligninaredepositedassec-ondarywallthickeningsontheinnerfaceoftheprimarycellwalls(Schreiberetal.1999;Naseeretal.2012).CBsin uencetheradialapoplastictransportofsubstances,whereasSLin uencethemovementofsubstancesfromtheapoplasttotheinterioroflivingcells.Whenthecellsaredead,thelamellaewouldthenfunctionbyin uencingradialapoplasticmovement(Enstoneetal.2003;Ma&Peterson2003).Therearenumerousstudiesreportingtherolesofsuchbarriersinbioticandabioticstressresponsesofhigherplants(Enstoneetal.2003;Chenetal.2011;Schreiber&Franke2011),withtheprimaryemphasisonNa+uptakeandradialoxygenloss(Colmeretal.1998;Zimmermann&Steudle1998;Armstrong&Armstrong2001,2005;Kotulaetal.2009;Krishnamurthyetal.2009,2011;Ranathunge&Schreiber2011;Ranathungeetal.2011a).However,thepropertiesofsuchbarriersinmangrovespeciesarenotwellstudied.There-fore,adetailedandsystematicstudytounderstandtheimportanceofapoplasticbypass owandtheroleofhydro-phobicbarriersinsaltexclusionofmangrovesisdesirable.Suberinisabiopolymermadeupofaliphaticandaromaticdomains(Kolattukudy1984).Thealiphaticcomponentisknowntobethemajorcontributortoitsbarrierproperties(Schreiberetal.1999).Elongases,hydroxylasesandperoxidasesarethemajorclassesofenzymesinvolvedinthebiosynthesisofsuberin(Bernardsetal.2004;Frankeetal.2005;Hoferetal.2008).Amongthese,peroxidases(e.g.cytochromeP450)arewellcharacterizedandtheyareknowntobeimportantintheformationofthemajoraliphaticconstituent,namelyω-hydroxyacids(Hoferetal.2008;Compagnonetal.2009;Sauveplaneetal.2009).However,tothebestofourknowledge,therearenostudiesconcerningtheexpressionandregulationoftheseenzymesinmangroveplants.

Inthecurrentstudy,themechanismofsaltbalancewasstudiedinasalt-secretingmangrovespecies,Avicenniaof cinalis.Formostofthetreatments,weusedgreenhouse-grown,2-month-oldseedlingswithoutpriorexposureto

1657

salinity,whichallowedustospeci callyexaminetheeffectofexposuretoaknownconcentrationofsalt.Theeffectsofsalttreatmentonnetphotosynthesis,chlorophyllcontent,tissueNa+distributionandsaltsecretionweremeasured.Uponsalttreatment,rapidup-regulationofacytochromeP450genethatcodesforasuberinbiosyntheticenzyme,andearlydevel-opmentofapoplasticbarriersintheendodermisandexodermisofA.of cinalisrootsaswellasleafsaltglandswereseen.Therolesofthesebarriersinregulatingsaltuptakeandbypass owofNa+wereexamined.Anon-secretor(ultra ltrator)Bruguieracylindricawasusedsolelyforcomparativehistologicalstudiesofhydrophobicbarrierformationinroots.Apossiblestructuraldifferenceintherootbetweenasalt-secretormangroveandanon-secretorisdiscussed.

MATERIALSANDMETHODS

Plantmaterialsandgrowthconditions

StemcuttingsofmangroveplantsandsaltsecretionswerecollectedfrommaturetreesgrowingintheBerlayerCreekarea,Singapore(1.27°N;103.80°E).ThepropagulesofA.of cinalisandB.cylindricawerecollectedduringfruitingseasonsfromthemangroveswampsinSingapore(BerlayerCreekandSungeiBulohWetlandReserve,Singapore).SeedlingswereestablishedandmaintainedinNaCl-freeconditionsbygrowingtheminpottingmixture(FarEastFlora,Singapore),untiltheyreachedthefour-nodestage(～2months)inagreenhouse(25–35°C,～90%relativehumidity;12hphotoperiod).Theplantswerewateredonalternatedays.

Salttreatmentprotocol

Themangroveseedlingsbearingfournodeswerecarefullytransferredtopotscontainingsandandwereallowedtoadaptfor2dbywateringwithhalf-strengthHoagland’ssolu-tion(SupportingInformationFig.S1b).Theplantswerethensubjectedtosalttreatment(inhalf-strengthHoagland’ssolu-tion)usingthreemethodsforvariousexperiments:(1)with200or500mmNaClforaperiodof2weeks;(2)with200or500mmNaClforaperiodof3weeks;and(3)plantsweregivenaninductionsalttreatmentwith200or500mmNaClfor3weeksafterwhichtheywereallowedtorecoverfor1weekinnormalhalf-strengthHoagland’ssolution,followedbyatreatmentwith200or500mmNaClfor1week,tomeasuretheionconcentrationinthexylemsap,withandwithoutaninductiontreatment.Attheendofeachtreatmentprotocol,plantswereharvestedforsubsequentanalysis.

Estimationoftotalionconcentration(Na+,K+andCl )fromplants

Seedlingsexposedtosalttreatmentwereharvestedandrinsedbrie ywithdistilledwatertoremovesurfacecontami-natingNa+.Oneplanteachwastakenasonereplicate,and vetosixindependentreplicateswereusedtogeneratethe

©2014JohnWiley&SonsLtd,Plant,CellandEnvironment,37,1656–1671

1658P.Krishnamurthyetal.

NaClfor2weekswereused.Theplantswereremovedcarefullyfromthesandandallowedtoacclimatizeinhalf-strengthHoagland’ssolutionfor～24hbeforetheactualmeasurementsweredonetoallowthemtorecoverfromanyosmoticshock.Preliminarystudiesdidnotshowanyincreaseinthebypass owofplantshandledinthismanner,suggest-ingthattherootswerenotdamagedduringthetransferofplantsfromsandculturetothesolution(Krishnamurthyetal.2011).Therefore,thismethodwasusedinallsubsequentexperiments.Theseplantswithintactrootsystemwereplacedinatube lledwithhalf-strengthHoagland’ssolutioncontaining0.01%(w/v)oftheapoplastictracerPTSand50mmNaCl.Theentireset-upwasplacedinapressurechamber.Priortostartingthemeasurements,theshoots(withleaves)thatwereprotrudingoutofthechamberwerecutoffusingrazorblades.About40–70mmofthestemwithintactrootsystemthatwasimmersedinthesolutionremainedwithinthechamber(Ranathungeetal.2011a).Apressureof1.0–3.0MPawasappliedupto10minandthexylemsapexudingfromthecutsurfaceofthestemwascollectedusingamicropipette.TheconcentrationsofPTSintheexternalmediumandintheexudedxylemsapweremeasuredusingaspectro uorophotometer(TECANIn niteM200PRO,TECAN,Switzerlandexcitationwavelength:405nm;emis-sionwavelength:515nm).A5-pointcalibrationcurvewasconstructedovertheconcentrationrangefrom2×10 8to2×10 6mPTS.Thecalibrationcurvewaslinearoverthisrange.Na+concentrationwasestimatedusingICP-OES.Therootswereplacedonascanner(HPScanjet3970,Hewlett-Packard,Palo,Alto,CA,USA)andthesurfaceareawasmeasuredfromthescannedimagesusingAdobePhotoshopsoftware.Thevaluesobtainedforasinglesurfacescanweretakenasrepresentative‘unitrootarea’.Thebypass ow(%ofexternalNa+orPTSenteringthexylembypassingthehydrophobicbarriersintheroots)wascalculatedasfollows:

meanvaluesreported.Leaves,stemandrootsfromplantswereseparatedandlefttodryat50°Cfor3–4d.Thedriedtissuewasgroundintoapowderinliquidnitrogen,and1.0gofthispowderwasaciddigestedbysuspendingitin3mLofconcentratedsulphuricacidfor2hfollowedbyaddinganequalvolumeofhydrogenperoxideandheatingat350°Cfor30min.Thiswasallowedtocoolandthedigestionwascom-pletedbyadding1mLofhydrogenperoxideandheatingforanother30minat350°C.Thedigestedsolutionwascooledandmadeupto20mLwithMilli-Q(Millipore,Billerica,MA,USA)water.TheNa+andK+levelsintheacid-digestedsamples(representingthetotalNa+andK+concentrationsinthetissuesample)wereestimatedbyinductivelycoupledplasma-opticalemissionspectrometry(ICP-OES;PerkinElmerDualViewOptima5300DV,Waltham,MA,USA)andtheCl concentrationwasestimatedbyanionchromatogra-phy(820ICSeparationCentre;MetrohmIonChromatogra-phySystem,Riverview,FL,USA).

Measurementofionconcentrationsfromleafsecretionsandxylemsap

Fiveleaveswerepooledforeachreplicatefrom eld-growntrees,andsecretionswereobtainedfromthreesuchrepli-catesbycollectingtheleafwashingsin50mLMilli-Qwater.Inthecaseofgreenhouse-grownplants,asingleplant(withfourpairsofleaves)wasusedasonereplicate.Theleafsecre-tionswereobtainedbycollectingtheleafwashingsin20mLofMilli-Qwaterfromninesuchreplicates.Leafareaswererecordedforeachreplicateandtheionconcentrationinthecollectedsecretionswasexpressedasμmolcm 2.

Forxylemsapcollection,twigsharvestedfromthe eld-growntreesandthestemcuttingsofthegreenhouse-grownplants(controlandtreated)wereused.Thecuttingsweremountedwithinthepressurechamberwiththecutendspro-trudingout(Model610,PressureChamberInstrument;PMSInstruments,Albany,OR,USA).Pressurewasapplied(from0.5to3.0MPabyincreasingthechamberpressurewithinter-valsof0.5MPa)untilthexylemsapstartedtoexudefromthecutendofthestem,andenoughexudedsapwascollected(～100μL).TheconcentrationofNa+andK+inthexylemsapwasmeasuredusingICP-OES,whileCl wasestimatedbyanionchromatography.Theinitialionconcentrationintheexternalsolution(500mmNaCl)wassetat100%andtheroot ltrationef ciency(oftheseedlingstreatedwith500mmNaClandthe eld-collectedsamples)wasobtainedbysub-tractingthepercentagefractioncorrespondingtothemeas-uredionconcentrationinthexylemsapfromitasfollows:

% bypass flow=[(xylem sap Na+or PTS concentration÷external Na+or PTS concentration)÷unit root area]×100

Histochemicaldetectionofhydrophobicbarriersinleavesandroots

Healthyleavesofknownsurfaceareafromthesecondnodeofseedlingswerechosenforanatomicalstudy.Leafdiscsweremadefromallpartsoftheleaflaminaexcludingthemid-veinregionand xedinethanol:aceticacid(3:1v/v)toremovethepigments.Freehandcrosssectionswerepreparedfromthefreshrootsofcontrolandtreated(500mmNaClfor2weeks)seedlingsandthesectionswereclearedusinglacticacidsaturatedwithchloralhydrate(Brundrettetal.1988).Crosssectionsweretakenatthefollowingdistancesfromtheroottip:5,10,20,30and50mm.TocheckforCBs,rootsectionswerestainedfor1hwith0.1%berberinehemisulfateandforanotherhourwith0.5%(w/v)anilineblue(Brundrettetal.1988),andleafdiscswerestainedwith0.1%berberinehemisulfatefor1h(Luxetal.2005).TocheckforSL,rootsectionsandleafdiscswerestainedfor1hwithFluorolYellow088(Brundrettetal.1991).Stainedroot

% root filtration efficiency=100% [(mMxylem sap ionconcentration÷500mM)×100]

Measurementoftrisodium

3-hydroxy-5,8,10-pyrenetrisulfonate(PTS)andNa+bypass ow

Forthebypassmeasurements,2-month-oldplantsgrowninthegreenhousewithorwithoutasalttreatmentof500mm

©2014JohnWiley&SonsLtd,Plant,CellandEnvironment,37,1656–1671

Saltexclusionandsecretioninmangroves

sectionsandleafdiscswereviewedunderanepi uorescencemicroscope(NikonEclipse80i,Tokyo,Japan)equippedwithanultraviolet lterset.Quantitativeanalysisofhydrophobicbarrier(CBandSL)formationinendodermisandexodermiswasperformedbyscoringtheimagesforthepresenceandabsenceoftheCBandSL.Sixseedlingseachrepresentingonebiologicalreplicateand～10freehandcrosssectionsfromeachreplicatewereusedforthequanti cation.

1659

Field-emissionscanningelectronmicroscope(FESEM)andX-raymicroanalysisinrootsofA.officinalisseedlings

Rootsfromcontrolandtreated(500mmNaClfor2weeks)plantswerechosenforenergydispersiveX-rayspectroscopy(EDS)analysis.Twomethodswereemployedtogeneratefreehandcrosssectionsofapicalandbasalpartsoftherootsinordertoruleoutthepossibilityofartefactualdataduetothesamplepreparation:(1)therootswereharvestedandquicklyfrozeninliquidnitrogenfollowedbylyophilizationand(2)freehandcrosssectionsweremadefromfreshlyhar-vestedrootsandthesectionswereairdriedbeforetheanaly-sis.Thesectionsmadefrombothmethodsexhibitedasimilarpatternofiondistribution,andhencesectionsfrommethod1wereusedinallsubsequentexperiments.Thesectionsweremountedonthesamplestub,Ptcoated(JEOLJFC-1600AutoFineCoater)andviewedwithaFESEM(JEOLJSM-6701FFESEM,Tokyo,Japan)equippedforEDS(JEOLJED2300).Imageswerecaptured,andareaanalyses,EDSspectraanddigitalmapswererecordedat15keV.

primersequencesandpredictedampliconsizeswere5′-CGGGATGTTGCCTTCTCTCA-3′(forward)and5′-TGAGATTCGTGAACCAGGGC-3′(reverse)forcytochromeP450(AoCYP86B1)(119bp)and5′-CGCCGGCAAGCAGCTAGAGG-3′(forward)and5′-ACCACGGAGCCTGAGGACCA-3′(reverse)forUbiquitin10(AT4G05320)(250bp).ThehomologousgenesequencestoAoCYP86B1werealignedusingClustalWwebtoolandaphylogenetictreewasgeneratedusingphylogeny.fronlinetool(Dereeperetal.2008).

Statisticalanalysis

Signi cantdifferencesatP<0.01andP<0.05betweenthecontrolandtreatmentswereestimatedbytheStudent’st-test.Datawerepresentedinthe guresasthemeanvalues±SE.

RESULTS

IonuptakeanddistributioninA.officinalisseedlings

Theiondistributionvariedintheleaves,stemsandrootsofA.of cinalisseedlings(Fig.1a–c).Therewasasigni cantincreaseintheconcentrationsofNa+andCl inallthetissuesamplesmeasureduponincreasingNaClintheexternalmedium,whileK+levelsremainedconstant.TheNa+levelsincreasedtoagreaterextentintheroots(Fig.1c)thanin

the

Real-timequantitativeRT-PCRanalysis

TotalRNAwasisolatedfromleaves,stemsandroots(apical,middleandbasal)ofgreenhouse-grownplantsundercontrolconditions,aswellasfromleavesandrootstreatedwith500mmNaClforvarioustimeperiods(0–48h,1and2weeks),usingQiagenRNeasykit(Qiagen,Valencia,CA,USA)andwasDNasetreated(DNA-freeTMreagent;Ambion,Austin,TX,USA)accordingtomanufacturer’sinstructions.AnaliquotofthisRNA(1g)wasusedtosynthesizecDNAusingThermoScriptTMreversetranscriptase(Invitrogen,Carlsbad,CA,USA).Real-timeqRT-PCRforcytochromeP450(AoCYP86B1)wasperformedusingtheStepOneReal-TimePCRmachine(AppliedBiosystems,FosterCity,CA,USA)withthefollowingprogramme:20sat95°Cfollowedby40cyclesof3sat95°Cand30sat60°C.TheSYBRFastABIPrismPCRkitfromKAPA(Biosystems,Wilmington,MA,USA)wasusedforreal-timeqRT-PCRanalysis.Thereactionmixtureconsistedof5.2μLofmastermix(providedinthekit),0.2μmforwardprimer,0.2μmreverseprimer,3.4μLofnuclease-freewaterand1μLofsamplecDNAtemplateina nalvolumeof10μL.ThePCRdatawereanalysedusingtheStepOneTMsoftware(v2.1,ABI).TheprimersweredesignedusingtheA.of cinalisCYPgenesequenceobtainedbytranscriptomeanalysisofA.of cinalisrootsinourlaboratory(unpublishedresults).ConstitutivelyexpressedUbiquitin10wasusedastheinternalcontrol.The

Figure1.IonuptakeanddistributioninAvicenniaof cinalis

seedlings.Seedlingssubjectedtosalttreatmentof200and500mmNaClfor2weeks(method1)werewashed,driedandaciddigestedtoestimateionlevelsbyinductivelycoupledplasma-opticalemissionspectrometry(ICP-OES)andanionchromatography.(a)Na+,K+andCl iondistributionintheleavesofcontrolandtreatedseedlings.(b)Na+,K+andCl iondistributioninthestemsofcontrolandtreatedseedlings.(c)Na+,K+andCl iondistributionintherootsofcontrolandtreatedseedlings.

(d)Na+/K+ratiosplottedfromallthetissuesamplesmeasured.Datarepresentmeans±SEfrom vetosixbiologicalreplicates.Asterisksindicatestatisticallysigni cantdifferencesbetweencontrolandtreatedat*P≤0.05and**P≤0.01.

©2014JohnWiley&SonsLtd,Plant,CellandEnvironment,37,1656–1671

1660P.Krishnamurthyetal.

leaves(Fig.1a),andtheincreasewastheleastinstems(Fig.1b).TheCl concentrationsincreasedtoagreaterextentinstems(3.6times)andleaves(2.4times)comparedwithroots(twotimes),althoughtheabsoluteconcentrationswerehigherintheroots.TheNa+/K+ratiosofleaves,stemsandrootsincreasedwithincreasingsaltconcentrationinalltissuesamples,withahigherdegreeofincreaseinrootscomparedwithleavesandstems(Fig.1d).TheK+concentrationsintheleavesandrootswerecomparable,whereastheNa+levelsinrootsincreasedtohigherconcentrationsleadingtoahigherNa+/K+ratio(Fig.1a,c).

EffectofsalttreatmentonleafchlorophyllcontentandnetphotosynthesisinA.officinalisseedlings

Thetotalchlorophyllcontentsofleaveswerenotaffectedwhentheseedlingsweretreatedwith200mmNaCl.However,asigni cantreductionwasseenwiththesalttreatmentof500mmNaCl(SupportingInformationFig.S1a).ControlandtreatedseedlingsgrowninthegreenhouseareshowninSup-portingInformationFig.S1b.Therewasnovisualdifferenceinthegrowthofseedlingsuponsalttreatment(200and500mmNaCl)duringthe2weekexperimentalperiod.Salt-treatedplantslookedsimilartothecontrolplantswithhealthy,greenleaves.Reductioninnetphotosynthesiswasseenwithasalttreatmentof200and500mmNaCl.Thereductioninphoto-syntheticrate,stomatalconductanceandtranspirationwas43,39and31%,respectively,with200mmNaCltreatment,and74,63and61%,respectively,with500mmNaCltreatment(Sup-portingInformationTableS1).

Figure2.IoncontentsinthexylemsapofAvicenniaof cinalis

seedlingsandbypass ow.Xylemsapwascollectedfromtwigsof eldsamplesandstemcuttingsofgreenhouse-growncontrolandsalt-treated[200and500mmNaClfor3weeks(method1)]

seedlingsusingapressurechamber.Theioncontentsinthexylemsapweremeasuredbyinductivelycoupledplasma-opticalemissionspectrometry(ICP-OES)andanionchromatography.(a)Na+,K+andCl ionconcentrationinthexylemsapoftwigsharvestedfrom eld-growntrees.(b)Na+,K+andCl ionconcentrationinthexylemsapofstemcuttingsofgreenhouse-growncontrolandtreatedseedlings.Asterisksindicatestatisticallysigni cantdifferencesbetweencontrolandtreatedat*P≤0.05and

**P≤0.01.Tomeasurethexylemsapionconcentrationswithsaltinduction,seedlingsweretreated(induced)with200and500mmNaClfor3weeksfollowingarecoveryperiodof1weekinnormalhalf-strengthHoagland’ssolutionafterwhichtheywereagain

treatedwith200and500mmNaClfor1weekbeforecollectingthexylemsap(method3).(c)Na+ioncontentfromthexylemsapofgreenhouse-growncontrolandtreatedseedlingswithoutandwithaninductiontreatment.Filledbarswithoutinduction,openbarswithinduction.Datarepresentmeans±SEfrom vebiologicalreplicates.Asterisksindicatestatisticallysigni cantdifferencesofionconcentrationinxylemsapwithoutandwithinduction,

*P≤0.05.Forbypass owmeasurements,seedlingswithfournodesgrowinginsandweretreatedwith200and500mmNaClfor3weeks(method2)andallowedtorecoverinnormalhalf-strengthHoagland’ssolutionfor～24hbeforestartingthemeasurements.Bypass owmeasurementswerecarriedoutusingapressurechamberfromuntreatedandsalt-treatedplants.(d)Apoplasticbypass owisgivenasthepercentage(w/v)ofexternallyappliedPTS/Na+.Datarepresentmeans±SEfromsixbiologicalreplicates.Asterisksindicatestatisticallysigni cantdifferencesbetweencontrolandtreatedat*P≤0.05and**P≤0.01.

Effectofsalttreatmentonthexylemsapionconcentration

Theionconcentrationsintheshootxylemsapweremeasuredundercontrolandtreatedconditions.Thepredominantionsfoundinthexylemsapoftwigsharvestedfromthe eldandthestemcuttingsoftheseedlingsgrowninthegreenhousewereNa+,K+andCl (Fig.2a,b).Inthe eldsamples,48mmNa+,62mmCl and24mmK+werefound(Fig.2a).Asigni -cantincreaseinthesapionconcentrationwithincreasedexternalNaClwasseeninthegreenhouse-grownseedlings(Fig.2b).Therewasabouttwo-andthreefoldincreaseintheNa+levelswith200and500mmNaCltreatment,respectively,whiletherewasanincreaseofaboutthree-andsixfoldintheCl levels.Therewasnosigni cantincreaseintheK+levelsuponsalttreatment.Uponsubjectingthesalt-inducedseed-lingsto500mmNaCltreatment,asigni cantreductioninthexylemsapNa+concentrationwasseen(Fig.2c).

Apoplasticbypass owinA.officinalisseedlings

Bypass ow(bulk ow)wasmeasuredfromtheA.of cinalisseedlingsthatwerepretreatedfor2weekswith200and500mmNaCltocheckthemajorpathofsodiumdeliveryinto

thexylemintheseplants.PTSisalargeanionthathasbeenusedasatracerformovementofapoplastic uidthroughtherootapoplastandintothexylembypassingthehydrophobicbarriers,namelyCBsintheendodermisandexodermis.PTSandNa+concentrationsinthexylemsaphavebeenexpressedasapercentageofthatpresentintheexternalmediuminFig.2d.Signi cantreductionofbypass owofPTSwasobservedwhenplantsweretreatedwith200and500mmNaCl.Similarly,therewasareductionintheNa+bypass

ow

©2014JohnWiley&SonsLtd,Plant,CellandEnvironment,37,1656–1671