in rats with STZ-induced diabetes antioxidant activation distinct and of apoptotic markermolecules Gut region-speciﬁc of accumulation reactive oxygen species leads toregionally Cell Biology The International Journal of Biochemistry&

(1)

ContentslistsavailableatScienceDirect

The International Journal of Biochemistry

& Cell Biology

j ou rn a l h o m ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / b i o c e l

Gut region-speciﬁc accumulation of reactive oxygen species leads to regionally distinct activation of antioxidant and apoptotic marker molecules in rats with STZ-induced diabetes

Zsanett Jancsó

^a

, Nikolett Bódi

^b

, Barbara Borsos

^a

, Éva Fekete

^b

, Edit Hermesz

^a,∗

aDepartmentofBiochemistryandMolecularBiology,FacultyofScienceandInformatics,UniversityofSzeged,Szeged,Hungary

bDepartmentofPhysiology,AnatomyandNeuroscience,FacultyofScienceandInformatics,UniversityofSzeged,Szeged,Hungary

a r t i c l e i n f o

Articlehistory:

Received19December2014 Receivedinrevisedform3March2015 Accepted9March2015

Availableonline18March2015

Keywords:

Apoptosis Diabetes Digestivetract Peroxynitrite Necrosis

a b s t r a c t

Theaimofthisstudywastoseekpossiblelinksbetweentheregionalityalongthedigestivetractand theaccumulationofreactiveoxygenspecies,theeffectivenessoftheantioxidantdefensesystemandthe sensitivitytothetypesofdemiseindifferentgutregionsofratswithstreptozotocin-induceddiabetes.

Significantchangeswereobservedintheoxidativestatusandintheactivityoftheantioxidantdefense systeminthediabeticcolon;theperoxynitriteproductionwasdoubled,thelevelofhemoxygenase- 2proteinwasincreased11-foldandtheexpressionofanti-apoptoticbcl-2wasalso increased.The segment-specificvulnerabilityofthegastrointestinaltractinducedbyhyperglycemiawasalsoconfirmed byelectronmicroscopy,demonstratingthepresenceofseverenecrosisinthecolonofthediabeticrats.No remarkablehistopathologicalalterationswereseenintheduodenumofthediabeticanimalsandthere werelikewisenosignificantchangesintheproductionofperoxynitriteintheirduodenum,whereasthe levelofthefreeradicalscavengermetallothionein-2wasincreased∼300-fold.

Conclusion:Thespatiallyrestrictedvulnerabilityobservedalongthedigestivetractcouldoriginatefrom ahighlevelofoxidativestressviaperoxynitriteproduction.

1. Introduction

Type 1 diabetes mellitus (T1D) results in severe metabolic imbalancesandpathological changesinmanytissues,and com- monly affects the entire gastrointestinal (GI) tract, from the esophagustotheanorectalregion(WolosinandEdelman,2000;

Zhaoet al.,2002).T1D involvesastateof highoxidativestress generated as a result of hyperglycemia-induced reactive oxy- genspecies(ROS)(Wolff,1993).Oxidativestressisanimbalance betweentheproductionofROS,andtheabilityofabiologicalsys- temtoachieve theready detoxification ofROSor torepairthe resultingdamage. WhileROS areimportant secondmessengers atlowconcentrationsandareinvolvedintheregulationofapo- ptosisandtheactivationoftranscriptionfactors,theycancause significantcellulardamagewhenpresentinexcess.Theycaninflict damage on all classes of cellular macromolecularcomponents,

∗ Correspondingauthorat:DepartmentofBiochemistryandMolecularBiology, FacultyofScienceandInformatics,UniversityofSzeged,P.O.Box533,H-6701 Szeged,Hungary.Tel.:+3662544887;fax:+3662544887.

E-mailaddress:hermesz@bio.u-szeged.hu(E.Hermesz).

eventually leading to tissue injury or even cell death, which canoccuressentiallyby twomechanisms,necrosisandapopto- sis (Bergamini et al., 2004). Although numerous reports have provideddetailsofthemolecularmechanismsresponsibleforROS- inducedapoptosis,littleisknownconcerningthemechanismsand signaltransductionpathwaysunderlyingROS-mediatednecrotic cell death. Necrosis has long been considered to be a passive modeofcelldeath(Kanducetal.,2002)andmuchmoreharmful thanapoptosisbecauseitcausesarobustinﬂammatoryresponse (Proskuryakovetal.,2003).

Toeliminatetheharmfuleffectsofreactivespecies,cellsare equippedwithanefﬁcientantioxidantdefensesystem,including enzymessuchassuperoxidedismutase(SOD),catalase(CAT),and hemeoxygenases(HOs),andlow-molecularweightantioxidants suchasglutathione(GSH)andmetallothioneins(MTs)(Kruidenier et al.,2003; Inoueet al.,2008).SOD catalyzesthereduction of thesuperoxideanion(^•O₂⁻)tohydrogenperoxide(H₂O₂),which isthendetoxiﬁedtowaterbyCATinthelysosomes(Wangand Ballatori,1998).TheHOsplayrolesinhemedegradation,yield- ingequimolaramountsofbiliverdin,andcarbonmonoxidewith importantfreeradical-scavengingpropertiesandfreeiron.HO-2is expressedconstitutively,contributingtocellhomeostasis,whereas http://dx.doi.org/10.1016/j.biocel.2015.03.005

(2)

HO-1isaninducibleenzymeexpressedatarelativelylowlevelin mosttissues(Maines,1997),recentlyidentiﬁedasanimportantcel- lulardefensemechanismagainstoxidativestress(Abrahametal., 2009).HO-1andHO-2areregulatedbystrikinglydifferentmech- anisms,whichmayreﬂectdifferentphysiologicalandpathological roles(GibbonsandFarrugia,2004).

GSHalso plays a critical role in this system, as an antioxidant,enzymecofactorandmajorredoxbuffer(WangandBallatori, 1998).TheMTsarepresentinallcellsthroughoutthebody.They haveacardinalroleinmetalhomeostasisandheavymetaldetox- iﬁcationthroughtheirhighmetal-bindingcapacity, theyplay a roleintheimmunefunction,andtheyareinvolved inavariety ofGItractfunctions(Thirumoorthyetal.,2011).Theyalsoplayan importantpartinthepreventionofdevelopmentofT1D,thecom- plicationsandthesubsequentpathogenictoxicity(Cai,2004).The overexpressionofMTsinvariousmetabolicorganshasbeenshown toreducehyperglycemia-inducedoxidativestress,organ-speciﬁc diabeticcomplications,andDNAdamageinexperimentaldiabetes (IslamandLoots,2007).

Inanearlierstudy,wedemonstratedspatially-restricteddam- ageofthegutcapillaryendotheliuminratswithstreptozotocin (STZ)-induceddiabetesincomparisonwithcontrolanimals(Bódi etal.,2012).Metagenomicanalysisoftheluminalcontentsofduo- denum,ileumandcolonofdiabeticratsalsofurnishesevidence oftheregionalityofthegutmicrobiota(Wirthetal.,2014).The twostudiesareingoodagreementasconcernstheadvantageous statusoftheduodenumofthediabeticratascomparedwiththe colon.

Thoseresultsledustofocusinthepresentstudyonthespatially- restricted differences in ROS production and activation of the antioxidantdefensesystemintheduodenumandcolonofratswith STZ-induceddiabetes.Theaimof thestudywastocharacterize thepossiblelinksbetweentheantioxidantstatusandthemacro- moleculardamageinselectedgutsegmentsinthediabeticrat.We reportdataontheaccumulationofapowerfuloxidant,peroxynitrite(ONOO⁻),theactivitiesofantioxidantenzymes(SODandCAT), andtheexpressionsofasetofgenescodingformembersofantiox- idantdefensesystem(mt-1,mt-2,ho-1andho-2),togetherwiththe detectionofpro-apoptoticandanti-apoptoticmarkers(bax,bcl-2 andcaspase-9).

2. Materialsandmethods 2.1. Animalmodel

AdultmaleWistarratsweighing300–400g,keptonstandard laboratorychow(BioplanKft.,Hungary)and withfreeaccessto drinkingwater,wereusedthroughouttheexperiments.Therats weredividedrandomlyintothreegroups:STZ-induceddiabetics (n=14),insulin-treateddiabetics(n=12)andsex-andage-matched controls(n=6).

Hyperglycemiawas induced as described previously (Izbéki et al.,2008).The animalswere considereddiabetic ifthe non- fastingbloodglucoseconcentrationwashigherthan18mM.From thistimeon,onegroupofhyperglycemicratsreceivedasubcuta- neousinjectionofinsulin(HumulinM3,EliLillyNederland)each morning(4U)andafternoon(2U).Thenon-fastingbloodglucose concentrationandweightofeachanimalweremeasuredweekly.

Thececumsizeofthesacriﬁcedratwasanalyzedbymeansofthe ImageJ1.48vprogram(http://imagej.nih.gov/ij/).Inallprocedures involvingexperimentalanimals,theprinciplesoflaboratoryanimal care(NIHpublicationno.85-23,revised1985)werefollowedand alltheexperimentsreceivedapprovalinadvancefromtheLocal EthicsCommitteeforAnimalResearchStudiesattheUniversityof Szeged.

2.2. Tissuehandling

Tenweeksaftertheonsetofdiabetes,theanimalswerekilledby cervicaldislocationunderchloralhydrateanesthesia(375mg/kg i.p.).The gutsegmentsof thecontrol,STZ-induceddiabeticand insulin-treateddiabeticratsweredissectedandrinsedin0.05M phosphatebuffer, pH7.4.Samplesweretakenfromtheduode- num(1cmdistaltothepylorus)andthemiddlepartofthecolon andprocessedforbiochemical,molecularbiologicalandelectron microscopystudy.

2.3. Biochemicalassays

0.5gduodenumandcolonofeachindividualrats,tissueswere homogenizedin4volumeofice-cold0.9%serumphysiologicby meansofaglasshomogenizerimmersedinanicewaterbath,cen- trifugedat17,000×gfor15minat4^◦C,andtheclearsupernatants usedforGSH,ONOO⁻,proteinanalysis,andmeasuringtheactivities ofantioxidantenzymes.

TotalproteinlevelsmeasuredbythemethodofLowryetal.

(1951)usingbovineserum albuminasa standard.The concen- trationsoftotal andreducedGSHin thetissuesweremeasured asdescribed by SedlakandLindsay(1968). Spectrophotometric measurementswerecarriedoutbyGENESYS10SUV-Vis(Thermo Scientiﬁc)spectrophotometer.

ONOO⁻wasassayedbydilutingsamplesinto1MNaOH(60:1) andmeasuringtheincreaseinabsorbance at302nm.Asa control,sampleswereaddedto100mMpotassiumphosphate(pH7.4) (60:1).ThedecreaseinabsorbancewasmeasuredatneutralpHas ONOO⁻decomposes(HuieandPadmaja,1993).

Catalase activity was determined spectrophotometrically at 240nmbythemethodofBeersandSizer(1953)andexpressedin Bergmeyerunits(1BU=decompositionof1gH₂O₂/minat25^◦C).

SOD activity was determined onthe basis of the inhibition ofepinephrine–adrenochromeautoxidation(MisraandFridovich, 1972). Spectrophotometric measurement was carried out at 480nm.TheresultswereexpressedinU/mgprotein.

2.4. Post-embeddingimmunohistochemistry

Forpost-embeddingimmuno-electromicroscopy,smallpieces (2–3mm)ofthegutsegmentswereﬁxedovernightat4^◦Cin2%

paraformaldehydeand2%glutaraldehydesolution,bufferedwith 0.1MPB(pH 7.4).Thesampleswerethenwashedin 0.05MPB andfurtherfixedfor1hin1%OsO₄.Afterfixation,thegutseg- mentswererinsedin0.1MPB,dehydratedinincreasingalcohol concentrations (50,70, 96%and absoluteethanol) and acetone, andembeddedinEpon(ElectronMicroscopySciences,Hatfield,PA, USA).TheEponblockswereusedtoprepareultrathin(70nm)sections,which weremountedonFormvar-coatednickelgridsand processedforimmunogoldlabeling.

Ultrathin sections from each block were pre-incubated in 1% bovine serum albumin in TRIS-buffered saline (TBS) for 30min, incubated overnight in the primary antibodies (heme oxygenase-2 mousemonoclonalIgG (Santa CruzBiotechnology, USA;workingdilution1:50)andcaspase-9rabbitpolyclonalIgG (Sigma–Aldrich, USA;workingdilution 1:50)),followed bypro- teinA-gold-conjugatedanti-mouse(18nmgoldparticles,Jackson ImmunoResearch,WestGrove,PA,USA;ﬁnaldilution1:20)sec- ondaryantibodiesfor3h,withextensivewashinginbetween.All stepswereperformedatroomtemperature.Thespeciﬁcityofthe immunoreactionwasassessedinallcasesbyomittingtheprimary antibodiesfromthelabelingprotocolandincubatingthesections onlyintheproteinA-gold-conjugatedsecondaryantibodies.Sec- tionswerecounterstainedwithuranylacetate(Merck,Darmstadt, Germany)andleadcitrate(Merck,Darmstadt,Germany),andwere

(3)

examinedandphotographedwithaPhilipsCM10electronmicro- scopeequipped witha MEGAVIEWII camera.The quantitative propertiesofgoldparticlescodingforHO-2 andcaspase-9were determinedinthemyentericgangliaandin theendotheliumof capillariesinthevicinityofthesegangliainallexperimentalgroup.

Countingwasperformedondigital photographsof fiveganglia, andtheentireendothelialprofileoffivewell-orientedcapillaries, whichwerecutperpendicularlytotheirlongaxisandvisualized atamagnificationof5800×,perintestinalsegmentpercondition atamagnificationof34,000×withtheAnalySIS3.2program(Soft ImagingSystemGmbH,Münster,Germany).Theintensityofthe labelingwasexpressedasthetotalnumberofgoldparticlesper unitarea.

2.5. RNAextraction,reversetranscriptionandPCRampliﬁcation

IntestinalsampleswerehomogenizedinRNABeereagent(Tel- Test Inc., Friendswood, TX, USA) and total RNA was prepared accordingtotheproceduresuggestedbythemanufacturer.Total RNAwasroutinelytreatedwith100URNAse-freeDNAseI(Thermo Scientific)toavoidanyDNAcontamination.ForassessingRNAcon- centrationandpuritytheabsorbanceofadilutedRNAsampleswere measuredat260and280nmusingNanoDrop1000UV/VISSpec- trophotometer(Thermo Scientific).The RNA concentrationwas calculatedusingtheA260=1.0isequivalentto∼40␮g/mlsingle- strandedRNAequation.TheA₂₆₀/A₂₈₀ratiowasusedtoassessRNA purityandratio∼2wasacceptedforpurifiedRNA.

First-strand cDNA was synthesizedby using 3␮g total RNA as template, 200pmolof each dNTP (Thermo Scientific), 200U MaximaHMinusReverseTranscriptase(Thermo Scientific)and 500pmolrandomhexamerprimers(Sigma)inafinalvolumeof 20␮L,andincubatedfor10minat37^◦C,followedby1hat52^◦C.

Real-timeqPCRwasdoneforgeneexpressionstudies,usingLumi- narisColorHiGreenLowROXqPCRMasterMix(ThermoScientiﬁc) in Applied Biosystems 7500 Real-Time PCR System (Life Tech- nologies).TheqPCRreactionswereperformedwithatemperature program of 10minat 95^◦C (initialdenaturing),followed by 40 cyclesof15sat95^◦C;1minattheannealingtemperature63^◦Cfol- lowedbyameltingcurvestagewithtemperaturerampingfrom60 to95^◦Candaﬁnalcoolingfor30sat40^◦C.Thequantitiesofexam- inedmRNAswerenormalizedtothatofˇ-actin,ahousekeeping gene,andgeneexpressionwascalculatedintermsofddCtmethod (LivakandSchmittgen,2001).

Table1

Primersequenceswithaccessionnumber.

Gene Primers(5→3)

ho-1(NM012580) GCTGCTGGTGGCCCACGCTT ACAGTCCAATGTTGAGCAGG ho-2(NM024387) GCTGCTGGTGGCCCACGCTT

AGGGTTTCTTTTGTTAGCATGGA

mt-1(M11794) ATGGACCCCAACTGCTCCTG

TGGAGGTGTACGGCAAGACT mt-2(AY341880) ATGGACCCCAACTGCTCCTG

GAAAAAAGTGTGGAGAACCG

Caspase-9(NM031632) AGCCAGATGCTGTCCCATAC

CAGGAACCGCTCTTCTTGTC

bax(RRU49729) GGAGGCGGCGGGCCCACCAG

CACGTCAGCAATCATCCTCTGC

bcl-2(NM016993) GGAAGGATGGCGCAAGCCGG

CGCAGGCCCAGCGTTGGCGAC

ˇ-Actin(M24113) GCAAGAGAGGTATCCTGACC

CCCTCGTAGATGGGCACAGT

For theampliﬁcationofrat mRNAs,isoform-speciﬁcprimers were designed on the basis of thedata bank entries.For nor- malization of the amounts of mt, ho, caspase-9, bax and bcl-2 mRNAs, theˇ-actin mRNAlevel wasused as internal standard (Table1).

2.6. Statisticalanalysis

RT-qPCRreactionsforeachanimalwereperformedintriplicate toincreasethereliabilityofthemeasurements.Statisticalanalysis wasperformedwithone-wayANOVAandtheNewman–Keulstest.

AllanalyseswerecarriedoutwithGraphPadPrism4.0(GraphPad Software,LaJolla,CA,USA)andMedCalcStatisticalSoftwarever- sion9.4.2.0(MedCalcSoftware,Mariakerke,Belgium).Aprobability levelofp<0.05wassetasthelevelofstatisticalsigniﬁcance.Alldata wereexpressedasmeans±SD.

3. Results

3.1. Ischemicandinﬂammatoryhallmarksandperoxynitritelevel indifferentgutsegments10weeksaftertheonsetofdiabetes

Ratsweresacriﬁcedandsignsofintestinalinﬂammationand severeischemia(bluish-purpleintestinesandanenlargedcecum (1.5–2-fold))wereobservedinthediabeticanimalsascompared withthecontrols(Fig.1AandB).Intestinalischemiawasalsovisible intheinsulin-treatedrats,but,thececumwasnotenlarged(not shown).

Tenweeksaftertheonsetofdiabetestherewasnosigniﬁcant changeinONOO⁻levelintheduodenuminanyoftheexamined groups.However,theONOO⁻levelinthecolonwassigniﬁcantly increased(1.7–2-fold)inthediabeticrats,whereasintheinsulin- treateddiabeticanimalsitwassimilartothecontrollevel(Fig.1C).

3.2. Activationoftheantioxidantdefensesystem

TheGSHlevelintheduodenumwaselevated2.5–3-foldinthe diabeticratsand1.5-foldintheinsulin-treateddiabeticratsrelative tothecontrolgroup(Fig.2A).TheelevationintheratioGSH/GSSG wasevenhigherinthisintestinalsegment:a6-foldincreasewas measuredinthediabeticanimals(datanotshown).Inthecolon, theGSHandGSSGcontentswerenotsigniﬁcantlychangedinany oftheexperimentalgroups.

TheSODactivitywasunaltered intheduodenum,regardless ofthetreatment,whileitwassigniﬁcantlydecreasedinthecolon in the diabetic groups (Fig. 2B). The activity of CAT was not signiﬁcantly affectedby eithertheSTZ-induceddiabetes or the insulinreplacementinanyoftheintestinalsegmentsexamined (Fig.2C).

TenweeksaftertheSTZinjection,therewasa4-foldincreasein theexpressionoftheho-1geneintheduodenum,butnosigniﬁcant changeinthecolon.Immediateinsulinreplacementmaintainedthe mRNAlevelclosetothecontrolvalueinbothintestinalsegments (Table2).

Theexpressionofthemt-1genewasupregulatedinbothofthe examinedgutsegmentsofthediabeticanimals:themt-1mRNA levelwasincreasedby6-and7-foldintheduodenumandthecolon, respectively.However,nosigniﬁcantchangeinmt-1expression wasobservedintheseintestinalsegmentsintheinsulin-treated diabeticanimals.Themt-2mRNAcontentwasmorethan300-fold thecontrollevelintheduodenumofthediabeticanimals,while insulintreatmentkepttheexpressionlevel closetothecontrol.

Theexpressionofmt-2didnotchangeinthecoloninthediabetic andinsulin-treatedanimals(Table2).

(4)

Fig.1. Inﬂammatoryhallmarksofgastrointestinalsystem(A,B)andaccumulationofperoxynitrite(ONOO⁻)indifferentgutsegments(C).Representativeimageofthefreshly dissectedintestineofratsfromcontrol(A)anddiabetic(B)animals.Bluish-purpleintestinalcolorandenlargedcecumwasobservedindiabeticanimals.Dataareexpressed asmeans±S.D.*p<0.05(relativetocontrols).C:control,n=6;D:diabetics,n=14;ID:insulin-treateddiabetics,n=12.

Fig.2.Concentrationoftheantioxidantglutathione(GSH)(A),activitiesofsuperoxidedismutase(SOD)(B)andcatalase(C)indifferentgutsegmentsinallexperimental groups.Dataareexpressedasmeans±S.D.*p<0.05(relativetocontrols).C:control,n=6;D:diabetics,n=14;ID:insulin-treateddiabetics,n=12.

Table2

Foldofinductionoftheho-1,mt-1andmt-2genesintheduodenumandcolonof ratswithSTZ-induceddiabetes.

ho-1 mt-1 mt-2

Duodenum

C 1 1 1

D 3.93±1.49^*** 5.94±2.32^*** 324.74±167.21^*

ID 1.28±0.06^‡ 1.41±0.51^# 1.06±0.01^†

Colon

C 1 1 1

D 0.93±0.06 7.5±4.95^* 1.4±0.02

ID 0.74±0.19 1.5±0.99^† 1.08±0.01

Dataaremeans±S.D.

Analysisofvariance(ANOVA):

* p<0.05(relativetocontrols).

***p<0.001(relativetocontrols).

† p<0.05(relativetodiabeticanimals).

‡ p<0.01(relativetodiabeticanimals).

#p<0.001(relativetodiabeticanimals).

3.3. MeasurementofHO-2expressionindifferentgutsegments

Theexpressionoftheho-2genewasupregulatedsigniﬁcantly onlyinthecoloninthediabeticandtheinsulin-treateddiabetic rats (Fig. 3A). The presence of HO-2 protein was followed by post-embeddingimmunohistochemistry.A∼4-foldincreasewas measuredin the number of gold particles signingHO-2 in the duodenumofthediabeticanimalsrelativetothecontrols.Inthe diabeticcolon,therewasamarked11-foldelevationintheHO-2 labelsresultingina∼5-foldhigherproteinlevelofHO-2inthecolon versusduodenum.Inbothexaminedregionsoftheinsulin-treated diabeticratsthecontrollevelofHO-2 expressionwasobserved (Fig.3B).

3.4. Effectsofdiabetesonapoptoticmarkersintheduodenum andcolon

Weexaminedthelevelsofthepro-apoptoticmarkerbaxandthe anti-apoptoticmarkerbcl-2.Intheduodenumofthediabeticrat,a 40%increasewasdetectedinthelevelofbaxmRNAascompared

(5)

Fig.3.Levelsofhemeoxygenase-2mRNA(ho-2)andprotein(HO-2).Expressionofho-2gene(A)indifferentgutsegmentsinallexperimentalgroups.Dataareexpressedas foldofinduction.mRNAsarenormalizedtothatofˇ-actin.QuantitativeevaluationofthenumberofgoldparticleslabelingHO-2intheduodenumandcolon(B).Allvalues arepresentedasmeans±S.D.**p<0.01,***p<0.001(relativetocontrols).C:control,n=6;D:diabetics,n=14;ID:insulin-treateddiabetics,n=12.

withthecontrolandinsulin-treatedgroups.However,inthecolon ofthediabeticrats,theexpressionofthebaxgenewasdownregu- latedby15–20%(datanotshown).Thediabetes-inducedalterations intheexpressionoftheanti-apoptoticmarkerblc-2weretheoppo- site of those in bax expression; nochangeor a non-signiﬁcant decreaseintheduodenum,anda40–45%increaseinthecolon.As aconsequence,theratiobax/bcl-2differedevenmoredramatically inthetwointestinalsegments:a50%increaseintheduodenum, anda40%decreaseinthecolon(Fig.4A).

Theexpressionpatternofcaspase-9intheduodenumwassimi- lartothatofbax.Therewasasigniﬁcant2.5–3-foldincreaseinthe caspase-9mRNAlevelinthediabeticrats.Theexpressionpattern ofthisgenewasunchangedinthecoloninthediabeticanimalsand thelevelwasthesameasthecontrolintheanimalstreatedwith insulin(Fig.4B).Theintestinalcaspase-9proteinexpressionwas alsodemonstratedbyimmunohistochemistry.Thenumberofgold particleslabelingcaspase-9wassigniﬁcantlyhigherintheduode- numofdiabeticratsthan thecontrol(Fig.4C andD),whilethe numberoflabelsinthecolonwasunchanged.

Thesegment-speciﬁcvulnerabilityofthegastrointestinaltract induced by hyperglycemia was also conﬁrmed by electron microscopy, demonstrating the presence of severe necrosis in thecolon ofthediabeticrats.Thehyperglycemiccolonsamples frequentlyexhibitednecrotic smoothmusclecells withabarely recognizablecytoplasm,andthestructuralintegrityoftheplasma membranewaslost(Fig.5).

4. Discussion

Thisstudyhaspresenteddataongutsegmentspeciﬁcoxidative stress,theeffectivenessoftheantioxidantdefensesystemandthe tissuedamageintheduodenumandcoloninratswithdiabetes.

Thesegment-speciﬁcvulnerabilityoftheGItractwasconﬁrmed byRT-qPCR,immunohistochemistryandelectronmicroscopy(EM).

EMdemonstratedthepresenceofseverenecrosisinthecolonofthe diabeticrats.ChangesinROSproduction,aspresumptivecandi- datestriggeringnecrosiswerealsodemonstratedinthecolon:the ONOO⁻productionwasdoubled.ElevatedlevelofONOO⁻serves asindirectevidenceofincreased^•O₂⁻andNOproductioninthe colon.AnincreasingnumberofstudieshaveimplicatedONOO⁻in thedevelopmentofT1D-associatedcomplications(Pacheretal., 2007).Astudyofthepathogenesisofinﬂammatoryboweldisease yieldedevidencethatintrarectallyadministeredONOO⁻induced

inﬂammation and transmucosal necrosis in the rat colon (Rachmilewitzetal.,1993).TheactivityofSODinthediabeticcolon wassigniﬁcantlyreduced,whichstatusalsocouldbeaconsequence ofthefastdepletionof^•O₂⁻throughitsreactionwithNO,which is3timesfasterthantheSOD-catalyzedreductionof^•O2−(Walsh, 1997).

Orstudyhasalsodemonstratedthattheintestinepossessessev- eraldefensemechanismsinasegment-speciﬁcmanner:maintains highconcentrationsoftheantioxidantGSHandupregulatesthe expressionsofhosandmtssoastopreservecellularintegrity.

GSHispresentinhighconcentrationsinthetissuesandpar- ticipatesinthecellulardefensebyscavengingROS(Nicoteraand Orrenius,1986;Flechneret al.,1990).Ourstudyhasrevealeda markedincreaseinGSHlevelintheduodenaltissues,butnotin thecolon,indiabeticrats.HighGSHlevelsprotectcellularproteins againstoxidationthroughtheGSHredoxcycle,andalsodirectly detoxifyROSinducedbyexposuretoSTZ(Razaetal.,2000).

BesidestheGSHsystem,theMTsalsoplayanoteworthyrole inthemaintenanceofthephysiologicalthiol/redoxbalance.The MTsmayserveaspotentantioxidantspreventingdiabeticcom- plicationsthroughthesuppressionofdiabeticoxidativedamage (Dabrowiak,2009).Ourstudyrevealedtheaccumulationofahuge amountofmt-2mRNAinthediabeticduodenum.Theexpressionof mt-1wasinducedtoaboutthesamelevelinthediabeticduodenum andcolon.Inthecolon,this7-foldelevationinmt-1mRNAlevel isoneofthefewpositivesignsofanactiveantioxidantdefense.

IthasbeendocumentedthatZn-inducedmtsynthesisinratpan- creaspreventedSTZ-induceddiabetes(YangandCherian,1994).

ElevatedlevelsofMTsintheliverandkidneyofdiabeticratshave alsobeenreported(CraftandFailla,1983;FaillaandKiser,1981;

UriuHareetal.,1988),andoverexpressedmtinthemouseheart signiﬁcantlypreventeddiabetes-inducedcardiomyopathy(Caiand Kang,2001;KangandCai,2001).

We also demonstrated gut segment-speciﬁc changes in the expressionoftheHOsduringdiabetes.Anumber ofinvivoand invitro studies have indicated theinduction of theHO system in response to a wide array of oxidative and cellular stresses (Applegateetal.,1991;Nath,1994;Vileetal.,1994;Otterbeinetal., 1995).HO-1andHO-2sharesimilarphysicalandkineticproper- ties,buthavedifferentphysiologicalrolesandtissuedistributions (Maines,2005).Inoursystem,hyperglycemiainducedho-1expres- sionintheduodenum,whileinthecolonincreasedamountsofho-2 mRNAandHO-2proteinweredetected.UnlikeHO-1,whichlacks

(6)

Fig.4. Expressionsofapoptoticmarkersintheduodenumandcolon.Ratiobax/bcl-2(A)andexpressionofcaspase-9gene(B)indifferentgutsegmentsofcontrol(C)n=6, diabetic(D)n=14andinsulin-treateddiabetic(ID)n=12rats.Dataareexpressedasfoldofinduction.Representativeelectronmicrographofultrathinsectionoftheenteric smoothmusclecells(smc)intheduodenumofdiabeticratafterpost-embeddingimmunohistochemistry(C),usingacaspase-9-speciﬁcprimaryantibody.Themajorityof the18nmgoldparticles(arrows)labelingcaspase-9accumulatedabovethemitochondria(m).Bar:500nm.Quantitativeevaluationofthenumberofgoldparticleslabeling caspase-9(D).Allvaluesarepresentedasmeans±S.D.*p<0.05,**p<0.01,***p<0.001(relativetocontrols).

cysteineresidues,HO-2containsthreeCys-Prosignaturemotifs, knownashemeregulatorymotifs(HRMs).Ithasbeenproposed (RagsdaleandYi,2011)thattheHRMsactasa“molecularrheo- stat”thatrespondstotheintracellularredoxpotential,controlling theHO-2activity.ThelevelofHO-2proteinincreased11-fold,and theactivitycouldbefurtherenhancedbyfreeradicalsviatheCys- Prosignaturemotifs.ThereleaseofFeionduringhemedegradation beforetheirsequestrationbyferritinmaymakethemavailablefor thecatalysisofharmfuloxidationreactions(Rouault,2009).

HOoverexpressionresultsinananti-apoptoticphenotypeasso- ciatedwithanincreasedexpressionofbcl-2indiabeticrats(Cao etal.,2008).Thesurvivalfunctionofbcl-2dependsontheextent ofbindingtoproteinssuchasBaxthat seemtoantagonizebcl- 2activity(AshkenaziandDixit,1998;Kroemer,1997).Ourstudy haspresentedevidence ofthegutregion-speciﬁcexpression of bcl-2,baxandcaspase-9thekeyelementsofapoptoticpathways (ThornberryandLazebnik,1998).Intheduodenaltissuesofthe diabeticrats,thelevelofbaxexpressionwasincreased,resultingin majorchangesintheratiobax/bcl-2mRNA.Thesechanges,along

withanincreasedlevelofcaspase-9,aninitiatorcaspaseintheapo- ptosispathway,indicateanenhancedpro-apoptoticenvironment, triggeringtheeventofprogrammedcelldeathintheduodenum.In thediabeticcolontheratiobax/bcl-2isloweredasaconsequence ofupregulationofthebcl-2expression.Theobservedincreasein bcl-2expressioncouldpossiblybeattributedtothehighlyelevated HO-2level.

Allof thesedata suggest that thecolon is more vulnerable thantheduodenumtooxidativestress.Thefactthattheantioxi- dantprotectionismoreefficientintheproximalintestinalsections than in the distal sections may bea consequence of theposi- tionaland functional differences(Blázovicset al.,2004).Earlier findingssupportedthis(Sandersetal.,2004):thecolongenerates moreendogenousROSthandoesthesmallintestine,andthisbasic pro-oxidantenvironmentofthecolonmayleadtoitsinabilityto handleoxidativestressaseffectivelyasthesmallintestine.Arecent studyonintestinalbacterialpopulationsinT1Dlikewiseempha- sizedregionalityalongtheGItract;T1Daffectedthecomposition ofthemicrobiotainagutregion-specificmanner:thecomposition

(7)

Fig.5.Necrotichallmarkindiabeticrats.Representativeelectronmicrographofan ultrathinsectiononthecolonofadiabeticrat.Smoothmusclecell(smc)withlocal membraneinjuriesandleakycytoplasm(arrows)werefrequentlyseen.Bar:10␮m.

oftheduodenalmicrobiotadidnotindicatethedevelopmentofa pathologicalentericmicroenvironment.Inthediabeticcolonhow- ever,theincreasedleveloftheGram-negativeKlebsiellacouldbe associatedwithsevereintestinalinﬂammation(Wirthetal.,2014).

Conﬂictofintereststatement

Theauthorsdeclarethattherearenoconﬂictsofinterest.

References

AbrahamNG,CaoJ,SacerdotiD,LiX,DrummondG.Hemeoxygenase:thekeyto renalfunctionregulation.AmJPhysiol2009;297:1137–42.

Applegate LA, Luscher P, Tyrrell RM. Induction of heme oxygenase: a gen- eral response to oxidant stress in culturedmammalian cells.Cancer Res 1991;51:974–8.

Ashkenazi A, Dixit VM. Death receptors: signaling and modulation. Science 1998;281:1305–8.

BeersRF,SizerIW.Catalaseassaywithspecialreferencetomanometricmethods.

Science1953;117:710–2.

BergaminiCM,GambettiS,DondiA,CervellatiC.Oxygen,reactiveoxygenspecies andtissuedamage.CurrPharmDes2004;10:1611–26.

BlázovicsA,SzentmihályiK,FehérE.Alterationsofredox-homeostasisinbowel parts.ZGastroenterol2004:42–9.

BódiN,BattonyaiI,TalapkaP,HermeszE,JancsóZs,KatarovaZ,etal.Diabetes- relatedstructural,molecularandfunctionalalterationsincapillariessupplying themyentericplexusinthegutofthestreptozotocininduceddiabeticrats.

Microcirculation2012;19:316–26.

CaiL.Metallothioneinasanadaptiveproteinpreventsdiabetesanditstoxicity nonlinearity.BiolToxicolMed2004;2:89–103.

CaiL, KangYJ.Metallothioneinpreventsdiabeticcardiomyopathy. ToxicolSci 2001;60:13.

CaoJ, DrummondG, InoueK,SodhiK,LiXY,OmuraS.Upregulationofheme oxygenase-1combinedwithincreasedadiponectinlowersbloodpressurein diabeticspontaneouslyhypertensiveratsthroughareductioninendothelial celldysfunction,apoptosisandoxidativestress.IntJMolSci2008;9:2388–406.

CraftNE,FaillaML.Zinc,iron,andcopperabsorptioninthestreptozotocin-diabetic rat.AmJPhysiol1983;244:E122–8.

Metallo-drugsandtheiraction.In:Metalsinmedicine.1sted.Chicheste,UK:Wiley

&Sons;2009.p.49–72.

FaillaML,KiserRA.Alteredtissuecontentandcytosoldistributionoftracemetals inexperimentaldiabetes.JNutr1981;11:1900–9.

FlechnerI,MarutaK,BurkartV,KawaiK,KolbH,KieselU.Effectsofradicalscavengers onthedevelopmentofexperimentaldiabetes.DiabetesRes1990;13:67–73.

GibbonsSJ,FarrugiaG.Theroleofcarbonmonoxideinthegastrointestinaltract.J Physiol2004;556:325–36.

HuieRE,PadmajaS.ThereactionofNOandsuperoxide.FreeRadicResCommun 1993;18:195–9.

InoueK,TakahashiT,UeharaK,ShimuzuH,IdoK,MorimatsuH,etal.Protectiverole ofhemeoxygenase1intheintestinaltissueinjuryinhemorrhagicshockinrats.

Shock2008;29:252–61.

IslamMS,LootsduT.Diabetes,metallothionein,andzincinteractions:areview.

Biofactors2007;29:203–12.

IzbékiF,WittmanT,RosztóczyA,LinkeN,BódiN,FeketeE,etal.Immediateinsulin treatmentpreventsgutmotilityalterationsandlossofnitrergicneuronsinthe ileumandcolonofratswithstreptozotocin-induceddiabetes.DiabetesResClin Pract2008;80:192–8.

KanducD,MittelmanA,SerpicoR,SinigagliaE,SinhaAA,NataleC,etal.Celldeath:

apoptosisversusnecrosis.IntJOncol2002;21:165–70.

KangYJ,CaiL.Metallothioneinsuppressionofdiabeticcardiomyopathybyinhibition ofhyperglycemiainducedoxidativestress.FreeRadicBiolMed2001;31:S33.

KroemerG.Theproto-oncogeneBcl-2anditsroleinregulatingapoptosis.NatMed 1997;3:614–20.

KruidenierL,KuiperI,VanDuijnW,Mieremet-OomsMA,vanHogezandRA,Lamers CB,etal.Imbalancedsecondarymucosalantioxidantresponseininﬂammatory boweldisease.JPathol2003;201:17–27.

LivakKJ,SchmittgenTD.Analysisofrelativegeneexpressiondatausingrealtime quantitativePCRandthe2-[delta][delta]CTmethod.Methods2001;25:402–8.

LowryOH,RosebroughNJ,FarrAL,RandallRJ.ProteinmeasurementwiththeFolin phenolreagent.JBiolChem1951;193:265.

MainesMD.Thehemeoxygenasesystem:aregulatorofsecondmessengergases.

AnnRevPharmacol1997;37:517–54.

MainesMD.Thehemeoxygenasesystem:update2005.AntioxidRedoxSignal 2005;7:1761–6.

MisraHP,FridovichI.Theroleofsuperoxideanionintheautoxidationofepinephrine andasimpleassayforsuperoxidedismutase.JBiolChem1972;247:3170–5.

NathKA.Thefunctionalsigniﬁcanceofinductionofhemeoxygenasebyoxidant stress.JLabClinMed1994;123:461–3.

NicoteraP,OrreniusS.Roleofthiolsinprotectionagainstbiologicalreactiveinter- mediates.AdvExpMedBiol1986;197:41–51.

OtterbeinL,SylvesterSL,ChoiAM.Hemoglobinprovidesprotectionagainstlethal endotoxemiainrats:theroleofhemeoxygenase-1.AmJRespirCellMolBiol 1995;13:595–601.

PacherP,BeckmanJS,LiaudetL.Nitricoxideandperoxynitriteinhealthanddisease.

PhysiolRev2007;87:315–424.

ProskuryakovSY,KonoplyannikovAG,GabaiVL.Necrosis:aspeciﬁcformofpro- grammedcelldeath.ExpCellRes2003;283:1–16.

RachmilewitzD,StamlerJS,KarmeliF,MullinsME,SingelDJ,LoscalzoJ,etal.

Peroxynitrite-inducedratcolitis–anewmodelofcolonicinﬂammation.Gas- troenterology1993;105:1681–8.

RagsdaleSW,YiL.Thiol/disulﬁderedoxswitchesintheregulationofhemebinding toproteins.AntioxidRedoxSignal2011;14:1039–47.

Raza H, AhmedI, JohnA, SharmaAK. Modulation of xenobiotic metabolism andoxidativestressinchronicstreptozotocin-induceddiabeticratsfedwith Momordicacharantiafruitextract.JBiochemMolToxicol2000;14:131–9.

RouaultTA.Cellbiology.Anancientgaugeforiron.Science2009;326:676–7.

SandersLM,HendersonCE,HongMY,BarhoumiR,BurghardtRC,CarrollRJ,etal.

Pro-oxidantenvironmentofthecoloncomparedtothesmallintestinemay contributetogreatercancersusceptibility.CancerLett2004;208:155–61.

SedlakJ,LindsayRH.Estimationoftotal,protein-bound,andnonproteinsulfhydryl groupsintissuewithEllman’sreagent.AnalBiochem1968;25:192–205.

ThirumoorthyN,SunderAS,KumarKTM,KumarMS,GaneshGNK,ChatterjeeM.A reviewofmetallothioneinisoformsandtheirroleinpathophysiology.WorldJ SurgOncol2011;9:54.

ThornberryNA,LazebnikY.Caspases:enemieswithin.Science1998;281:1312–6.

UriuHareJY,StemJS,KeenCL.Theeffectofdiabetesonthemolecularlocalization ofmaternalandfetalzincandcoppermetalloproteinintherat.BiolTraceElem Res1988;18:71–9.

VileGF,Basu-ModakS,WaltnerC,TyrrellRM.Hemeoxygenase1mediatesanadap- tiveresponsetooxidativestressinhumanskinﬁbroblasts.ProcNatlAcadSciU SA1994;91:2607–10.

WalshSW.Theroleofoxidativestressandantioxidantsinpreeclampsia.Contem- poraryOB/GYN1997;42:113–24.

WangW,BallatoriN.Endogenousglutathioneconjugates:occurrenceandbiological functions.PharmacolRev1998;50:335–56.

WirthR,BódiN,MarótiG,BagyánszkiM,TalapkaP,FeketeE,etal.Regionallydistinct alterationsinthecompositionofthegutmicrobiotainratswithstreptozotocin- induceddiabetes.PLOSONE2014;9:e110440.

WolffSP.Diabetesmellitusandfreeradicals.BrMedBull1993;49:642–52.

WolosinJD,EdelmanSV.Diabetesandthegastrointestinaltract.ClinDiabetes 2000;18(4):148–51.

YangJ,CherianMG.Protectiveeffectsofmetallothioneinonstreptozotocin-induced diabetesinrats.LifeSci1994;55:43–51.

ZhaoJ,ShaH,ZhouS,TongX,ZhuangFY,GregersenH.Remodellingofzero-stress stateofsmallintestineinstreptozotocin-induceddiabeticrats.Effectofgli- clazide.DigLiverDis2002;34:707–16.