signalingadsence

ReducedPericellularSensitivitytoIGF-IinFibroblastsFromGirlsWithTurnerSyndrome:AMechanismtoImpairClinicalResponsestoGHMELISSAWESTWOOD,SHAHINH.
TAJBAKHSH,KIRKW.
SIDDALS,ANDREWJ.
WHATMORE,ANDPETERE.
CLAYTONMaternalandFetalHealthResearchCentre[M.
W.
],Endocrinology[S.
H.
T.
,A.
J.
W.
,P.
E.
C.
],andImaging,GenomicsandProteomics[K.
W.
S.
],UniversityofManchester,ManchesterM139WL,UnitedKingdomABSTRACT:GirlswithTurnersyndrome(TS)aretreatedwithsupraphysiologicaldosesofgrowthhormone(GH)toimprovenalheight;howeverinsomegirls,thegrowthresponsecanbepoor.
ThismayreectaberrationsinGHand/orIGF-Iactionsatthecellularlevel,andthusthisstudycomparedtheresponseofskinbroblastsfromnormalchildren(n5)andgirlswithTS(n8)toGH,IGF-I,oracombination,byassessingtheIGFbindingprotein(IGFBP)proleofconditionedmediumharvestedover7d.
ThetwocelltypeshadacomparableIGFBPprole;IGFBP-3andIGFBP-4werethemostabundantspecies.
TSbroblastsproducedmoreIGFBP-3(d7,51.
445ng/mLversus2022ng/mL;p0.
05)thancontrolcells;levelsofIGFBP-4weresimilar(2112ng/mLversus3021ng/mL).
GHdidnotinuenceIGFBPproduction.
IGF-ItreatmentdidnotaffectIGFBP-4levelsbutenhancedtheproductionofIGFBP-3bybothcelltypes(p0.
05).
However,theresponseofTSbroblaststoIGF-Iwasapproximatelyhalfthatobservedinnormalcells(p0.
05).
AlteredIGF-Iactivity,becauseofreducedbioavail-abiltyand/orreducedsensitivity,couldcontributetotheneedforhighGHdosesinTSandforthepoorresponsetoGHinsomegirlswithTS.
(PediatrRes70:25–30,2011)Turnersyndrome(TS;completeorpartialabsenceofanXchromosome)affects1in2500livebirthsandisassoci-atedwithpre-andpostnatalgrowthfailure(1),sothatifuntreated,girlswithTSare20cmshorterthanthenormalfemalepopulationatadulthood(2).
Consequently,childrenwiththisconditionaretreatedwithgrowthhormone(GH).
ThereissomecontroversyaboutwhethertheGH/IGFaxisofchildrenwithTSisnormal.
SomestudieshaveshownthatsecretionofGHisdiminishedinTS(3),althoughthemajorityhavefoundthatGHlevelsarenormalincomparisonwithage-matchedcontrolsubjects(4,5).
Similarly,thedataregardingIGF-IlevelsinTSareconictingasbothhigher(3)andlower(6)thannormallevelshavebeenreported.
RegardlessofendogenousGH/IGFlevels,numerousstud-ieshaveshownthattreatmentwithGHhasabenecialeffectontheheightofchildrenwithTS(7);however,supraphysi-ologicaldosesarerequiredandtheresponsetotreatmentishighlyvariable—determined,atleastinpart,byGHdose,andageandheightatthestartoftherapy(8)—andalthoughtreatedwomenaretallerthanuntreatedwomen,theirnalheightismorethan2SDscoresbelowthemeanfornormalwomen(7).
TheseclinicaldatasuggestthatalthoughtherearenomajoraberrationsinthelevelsofGHand/orIGF-IinchildrenwithTS,theircellularsensitivitytothesehormonesmaybere-duced.
Indeed,ithasbeenreportedthatIGF-I-stimulateddegradationofLDLbymonocytesisolatedfromchildrenwithTSisreducedincomparisonwiththatofcellsobtainedfromcontrolsubjects(9).
ItisalsopossiblethatinTS,impairedIGF-IactionmayreectareductioninbioavailabilitybecausebroblastcellsfromchildrenwithTShavebeenshowntoproducemoreIGFbindingprotein(IGFBP)-3—anegativeregulatorofIGFfunction—thanbroblastsfromnormalchil-dren(10).
Consequently,thisstudyaimedtocharacterizeafunctionalresponse(IGFBPgeneration)toGHandIGF-IinbroblastsisolatedfromchildrenwithTS.
MATERIALSANDMETHODSCellculture.
Fibroblastcellswereisolatedfromskinbiopsiesobtained,withinformedparentalconsentandapprovalfromboththelocalNHS(SalfordandTrafford)andtheUniversityofManchesterresearchethicscommittees,fromeightchildrenwithTS(conrmedbykaryotypeanalysis)andvenormalchildrenintheprepubertalagerangeattimeofbiopsy.
AllcellsweremaintainedinDMEMcontaining10%FCS,2mMglutamine,50IUpenicillin,50IUstreptomycin,and2.
5g/mLamphotericinBat37°C,5%CO2.
CharacterizationofIGFBPproduction.
NormalandTurnerbroblastcellswereseededat1104cells/wellinDMEMcontaining1%FCSandeitherGH(200ng/mL;basedondose-responseexperimentsperformedinourpreviousstudiesofhumanbroblastproliferation)(11),IGF-I(100ng/mL),GHandIGF-I,ornoadditionalsupplement.
Thecellswereculturedfor7dwiththefurtheradditionofhormones[GH(200ng/mL),IGF-I(100ng/mL),orGHandIGF-Iond2,4,and6;conditionedmedium(CM)washarvestedond1,3,5,and7.
Eachtreatmentandtime-pointwasanalyzedintriplicateintwoindependentexperimentsoneachcellline.
Westernligandblotting.
TheproteincontentofeachsampleofCMwasdetermined,andthenaliquotscontaining100gproteinweresubjectedtotrichloroaceticacidprecipitation.
TheresultingproteinpelletwasresuspendedinSDSloadingbuffer(without-mercaptoethanol)andboiledfor5min.
TheproteinineachsamplewasthenresolvedbySDSPAGE,transferredtonitrocellulose,andprobedwith125I-IGF-I(12).
DensitometricanalysisofthescannedautoradiographswasperformedusingImageJsoftware(http://rsbweb.
nih.
gov/ij/).
Westernimmunoblotting.
Sampleswereelectophoresedandblottedasdescribedaboveandthenprobedovernightat4°Cwithananti-IGFBP-3polyclonalantibody(1:1000dilution;Upstate,MiltonKeynes,UnitedKing-dom)followedbyananti-rabbitIgGantibodylinkedtohorseradishperoxi-ReceivedSeptember10,2010;acceptedJanuary12,2011.
Correspondence:MelissaWestwood,Ph.
D.
,MaternalandFetalHealthResearchCentre,UniversityofManchester,StMary'sHospital,OxfordRoad,ManchesterM139W,UnitedKingdom;e-mail:melissa.
westwood@manchester.
ac.
ukThisresearchwasfundedbyanunrestrictedgrantfromNovoNordisk.
TheGroupissupportedbyfundingfromtheNIHRManchesterBiomedicalResearchCentre.
Abbreviations:CM,conditionedmedium;IGFBP,IGFbindingprotein;ISS,idiopathicshortstature;TS,Turnersyndrome0031-3998/11/7001-0025PEDIATRICRESEARCHVol.
70,No.
1,2011Copyright2011InternationalPediatricResearchFoundation,Inc.
PrintedinU.
S.
A.
25dase(1:5000dilution;AmershamLifeSciences,Buckinghamshire,UnitedKingdom)for1hatroomtemperature.
Proteinsweredetectedbyenhancedchemiluminescence(AmershamLifeSciences).
IGFBP-3immunoassay.
TheconcentrationofIGFBP-3presentineachsampleofCMwasdeterminedusingtheImmulite(SiemensHealthcare,Surrey,UnitedKingdom)assay,whichhasaninter-andintra-assaycoef-cientofvariation(CV)of5.
6%and10%,respectively.
IGFBP-4immunoassay.
ThesamplesofCMwereanalyzedforIGFBP-4byusingtheELISAdevelopedbyDSL(Texas),whichhasaninter-andintra-assayCVof5%.
CharacterizationofGH-stimulatedactivationofJAK-2.
NormalandTurnerbroblastcellswereserumstarvedfor24handthenincubatedintheabsenceorpresenceofGH(200ng/mL)for0,2,5,15,30,or60min.
CelllysateswerepreparedusingRIPAbufferandafterproteinquantication,30gofproteinfromeachsamplewasdilutedinSDSloadingbuffercontaining-mercaptoethanol,electrophoresedandthenelectroblottedontonitrocellu-losemembranes.
ActivationofGHsignalingpathwayswasassessedbyprobingwithanantibodythatspecicallyrecognizesthephosphorylatedisoformofJAK-2(1:500;CellSignalingTechnology).
ImmunecomplexesweredetectedusinganHRP-anti-rabbit-IgGantibodyfollowedbyenhancedchemilumines-cence.
AnantibodythatrecognizesallisoformsofJAK-2(anti-JAK-2,1:500,Upstate)wasusedtocontrolforproteinloadingofthegels.
Statisticalanalysis.
ThelevelofIGFBP-3andIGFBP-4presentinmediaconditionedbybroblasts,isolatedfromnormalchildrenorchildrenwithTS,andculturedwithandwithouthormonesfor7dwaslog10transformedandanalyzedbyathree-way(celltype,time,andtreatment)independentANOVAwithBonferroniposthoctesting,andthen,thedataobtainedfromeachtime-point/hormonecombinationwereanalyzedbyaone-wayanalysisofcovariance(ANCOVA)tocontrolforthedifferenceinbasalhormonepro-ductionbythetwocelltypes.
ThereportedresultsincludetheF-ratio(theratioofsystematicvariancetounsystematicvariance),thedegreesoffreedomfromwhichitwascalculated,andthesignicancevalue.
Thefold-change(versuscontrol)inIGFBP-3levelsinducedbytreatmentofnormalversusTurnerbroblastsfor7dwithGH,IGF-I,orbothhormoneswasanalyzedbyMann-WhitneyUtest.
Resultswereconsideredsignicantwhenp0.
05(IGFBP-3)orp0.
01(IGFBP-4)asdeterminedbyLevene'stestofequalityoferrorvariances.
Untransformeddataarepresentedgraphi-callyasmeanSEM.
RESULTSFibroblastsfromchildrenwithTSproduceasimilarIGFBPproletocellsfromnormalchildren.
LigandblotanalysisofconditionedmediaobtainedfromnormalandTurnerbroblastcellsmaintainedincultureunderbasalcon-ditionsfor7drevealedthatthetwocelltypesproduceasimilarproleofIGFBPs.
Bandswereevidentat45,34,30,and24kD(Fig.
1A)andcomparisonoftheseproteinswiththemigratorypatternofrecombinanthumanIGFBP-3andtheIGFBPswithinhumanserum,whichwereincludedintheanal-ysisaspositivecontrols,suggeststhatbothnormalandTurnerbroblastcellsproduceIGFBPs3,2,5,and4(Fig.
1A).
IGFBPproductionbytheindividualcelllineswithinboththenormalandTurnercellcohortsvaried.
However,densitometricanalysisofallligandblotssuggestedthatbothcelltypespredominantlyproduceIGFBP-3althoughnormalbroblastsalsoproduceanabundanceofIGFBP-4(Fig.
1B);consequently,thesebindingproteinswereselectedforfurtheranalysis.
FibroblastsfromchildrenwithTSproducemoreIGFBP-3butsimilarlevelsofIGFBP-4incomparisonwithcellsfromnormalchildren.
ThegenerationoftheseIGFBPs,bybothnormalandTurnerbroblastcells,seemedtoincreaseoverthe7dinculture(Fig.
1AandB),whichwasconrmedbymeasur-ingtheconcentrationofthesebindingproteinsusingspecicimmunoassays(Fig.
2andTable1;IGFBP-3:p0.
05;IGFBP-4:p0.
01).
AnalysisoftheIGFBP-3levelspresentinCMalsorevealedthatincomparisonwithbroblastsobtainedfromnormalchildren,cellsfromchildrenwithTSproducesignicantlymoreIGFBP-3underbasalconditions(p0.
05).
IGFBP-3issubjecttoproteolyticcleavageandastudyofcircu-latingIGFBP-3foundthisprocesstobeenhancedinwomenwithTS(13).
WethereforeusedWesternimmunoblottingtoinvesti-gatewhethertheIGFBP-3presentinCMofnormalandTurnerbroblastswasalsofragmented.
However,asshowninFig.
2C,wewereabletodetectonlyintactIGFBP-3.
ThelevelofIGFBP-4inthemediumconditionedbybroblastsfromnormalandTSsubjectswassimilar(pnonsignicant).
FibroblastsfrombothnormalchildrenandchildrenwithTunersyndromeproduceIGFBP-3inresponsetoIGF-IbutnotGH.
WenextinvestigatedwhetherGHand/orIGF-ItreatmentalterstheIGFBPprolegeneratedbynormalandTurnerbroblastcells.
WesternligandblotanalysisofmediaharvestedfromcellstreatedwithGH,IGF-I,oracombinationofthetwohormonesdemonstratedthatthepatternofIGFBPsproducedbythetwocelltypeswasnotaffectedbyhormonetreatmentasIGFBP-3andIGFBP-4werestillthemostabun-dantIGFBPspresent(datanotshown).
Analysisofthecon-ditionedmediabyspecicimmunoassayrevealedthatonlythegenerationofIGFBP-3wasinuencedbyhormonetreat-ment(p0.
05;Fig.
3andTable1)becausethelevelofIGFBP-4presentinmediaharvestedfromhormone-treatedcellswassimilartothatmeasuredinmediafromcellsmain-tainedunderbasalconditions(pnonsignicant;Fig.
3).
IGFBP-3productionwasonlyenhancedbystimulationwithFigure1.
NormalandTurnerbroblastsproduceasimilarIGFBPprole.
TheIGFBPproleofbroblastsfromnormalchildren(n5)andchildrenwithTS(n8)wasassessedbyWesternligandblottingwith125I-IGF-I.
Arepresentativeblotisshownin(A);humanrecombinantIGFBP-3andhumanserumwereincludedaspositivecontrols.
Thedataobtainedfromdensitometricanalysisofallblotsareshownin(B)(normalbroblasts)and(C)(Turnerbroblasts);columnsrepresentmeanSEMdensitometryunitsofIGFBPspresentinCMharvestedond1(Ⅺ),3(f),5(u),and7(o)ofculture.
26WESTWOODETAL.
IGF-I(Fig.
3).
TheBonferroniposthoctestrevealedthatthelevelofIGFBP-3producedbycellstreatedwithGHwasnodifferenttothatsecretedbycontrolcells,whereastherewasa2-to4-foldincrease(dependingonday)inthelevelofIGFBP-3presentinmediaconditionedbycellstreatedwitheitherIGF-IoracombinationofGHandIGF-I(p0.
05).
TheresponseofcellsfromchildrenwithTStotreatmentwitheitherIGF-IorGHplusIGFwascomparedwiththatofcontrolcellsaftercontrollingforthedifferenceinthebasallevelofIGFBP-3producedbythetwocelltypes.
Interest-ingly,thisanalysisrevealedthattheTurnerbroblastsdidnotrespondaswellascontrolcellstostimulationwithacombi-nationofGHandIGF-I(p0.
05)asevenafter7dinculture,therewasonlya2-foldincreaseinthelevelofIGFBP-3detectedinCM,whereasIGFBP-3levelshadincreased4-foldinmediaconditionedbycontrolcells(Fig.
4;p0.
05).
AsimilartrendwasobservedwhencellsweretreatedwithonlyIGF-I.
Indeed,thelevelofIGFBP-3producedinresponsetoincubationofcellswithGHandIGF-IwasnodifferenttothatgeneratedbycellstreatedwithonlyIGF-I,whichsuggeststhattheincreaseinIGFBP-3levelsobservedafterdualhormonetreatmentwasprimarilybecauseoftheactionsofIGF-I.
However,toensurethattheGHpreparationusedforthesestudieswasbioactive,weusedWesternblottingtoassessthephosphorylationstatusofJAK-2,thesignalingmoleculeim-mediatelydownstreamoftheGHreceptor.
Figure5demon-stratesthatalthoughtheremaybesubtledifferencesinthetempoofactivation,GHdoesstimulatephosphorylationofJAK-2inbothtypesofbroblast,whichsuggeststhatthesecellsarecapableofrespondingtoGH.
NoneofthehormonetreatmentsresultedinthedetectionofIGFBP-3fragments(datanotshown).
DISCUSSIONGirlswithTSaretreatedwithGHtoalleviatetheshortstatureassociatedwiththiscondition;however,theclinicalbenet,atleastintermsofgrowthpromotion,canbedisap-pointingdespitetheuseofsupraphysiologicaldoses.
Theseclinicalobservationsalongwithexvivostudiesofcellsiso-latedfromaffectedgirls(9,10)haveledtothehypothesisthatalthoughtherearenomajorabnormalitiesintheGH/IGFaxisofchildrenwithTS,cellularsensitivitytooneorbothofthesehormonesisdecreased.
Theimportanceofautocrine/paracrineIGF-Ilevelsforgrowthwashighlightedbythegenerationofmicewithaliver-specicdeletionoftheigf1gene(14),whichhavenormalpostnatalgrowthdespitea75%reductionincirculatingIGF-Ilevels.
Therefore,thisstudysoughttoinvestigate,byassessingIGFBPgenerationasamarkeroffunction,whetherTScellsdisplayanalteredcellularresponsetoGHandIGF-I.
WesternligandblottingofmediaconditionedbybroblastsisolatedfromnormalchildrenandgirlswithTSrevealedthatbothcelltypesproduceIGFBP-3,-2,-4,and-5,whichsupportsthendingsofastudyanalyzingIGFBPmRNAexpressioninhumanskin(15)andalsotheIGFBPproleofcelllinesderivedfromadulthumanbroblasts(16)whichwerefound,likethebroblastsusedinthisstudy,topredom-inantlyproduceIGFBP-3andIGFBP-4.
Figure2.
TurnerbroblastsproducemoreIGFBP-3butasimilarlevelofIGFBP-4incomparisonwithnormalbroblasts.
MediaconditionedbybroblastsfromnormalchildrenandchildrenwithTSwereanalyzedforIGFBP-3(A)andIGFBP-4(B)byimmunoassay.
DataarepresentedasmeanSEM;whitebars—normalbroblasts,blackbars—Turnerbroblasts.
*denotessignicantdifferencebetweennormalandTurnerbroblasts(p0.
05);§denotessignicantdifferenceincomparisonwithd3(IGFBP-3—p0.
05;IGFBP-4—p0.
01);denotessignicantdifferencesincomparisonwithd5.
IGFBP-3andIGFBP-4levelsarereportedasng/mL;datacanbeconvertedtoSIunits(nmol/L)bymultiplyingby0.
035and0.
038,respectively.
(C)TheproteolyticstatusofIGFBP-3presentinharvestedmedia(100gtotalprotein)wasanalyzedbyWesternimmunoblottingwithananti-IGFBP-3polyclonalantibody;arepresentativeblotisshown.
IntactrecombinanthumanIGFBP-3wasincludedasacontrol.
Table1.
AnalysisofIGFBP-3andIGFBP-4levelsbythree-wayindependentANOVAIGFBP-3IGFBP-4VariableFdfSigFdfSigCell27.
3710.
003.
9411NSTime41.
9220.
0055.
2020.
00Treatment18.
5430.
002.
433NSCellTime0.
142NS1.
152NSCellTreatment0.
143NS0.
683NSTimeTreatment0.
166NS0.
776NSCellTimeTreatment0.
186NS0.
596NSThelevelofIGFBP-3andIGFBP-4presentinmediaconditionedbybroblasts,isolatedfromnormalchildrenorchildrenwithTS,culturedwithandwithouthormonesfor7dwaslog10transformedandanalyzedbyathree-wayindependentANOVA.
CellreferstonormalorTurnerbroblasts;timereferstothelengthofculture—d3,5,or7;treatmentreferstothetreatmentconditions—control,GH(200ng/mL),IGF-I(100ng/mL),orbothGHandIGF-I.
determineswhetherthereisaninteractioneffectbetweentwoormoreofthevariables(e.
g.
celltreatmentdetermineswhethertreatmenthasadifferenteffectinnormalvsTSbroblasts).
Resultswereconsideredsignicantwhenp0.
05(IGFBP-3)orp0.
01(IGFBP-4)asdeterminedbyLevene'stestofequalityoferrorvariances.
F,Fratio;df,degreesoffreedom;Sig,signicance.
27CELLULARIGFAXISINTURNERSYNDROMETherewasmarkedvariationinthelevelofIGFBP-3andBP-4producedbybothcohortsofcells;nonethelessitwasapparentthatthebasalproductionofIGFBP-3byTSbro-blastswassignicantlyhigherthanthatofnormalbroblasts,whichsupportspreviousndings(10).
WeinvestigatedwhetherthisincreaseinIGFBP-3levelswascounteredbyachangeintheproteolysisofthebindingproteinbecauseastudyofIGFBP-3presentintheserumofadultfemalesdemonstratedincreasedfragmentationinTS(13).
However,therewasnoevidencetosuggestthattheIGFBP-3presentwithintheCMofeithernormalorTurnerbroblastcellshadundergoneproteolysis.
Interestingly,cellsfromchildrenwithanothergrowthdisorder,idiopathicshortstature(ISS),alsohaveenhancedIGFBP-3productionincomparisonwithcon-trolcells(17).
ThelevelofIGFBP-3presentinthecirculationiswellknowntoberegulatedbyGH(16,18)andastudyofnormalandTurnerbroblastsreportedastimulatory,albeitnonsigni-cant,affectofGHonIGFBP-3productionbycontrolcells(10).
Inourstudy,onlytreatmentofcellswithIGF-IaffectedthelevelofIGFBP-3presentintheculturemedia.
TheresponseofcellstreatedwithbothGHandIGF-IwassimilartothatofcellstreatedwithIGF-IalonesuggestingthatGHdoesnotregulateIGFBP-3productionbythesebroblasts.
WeexcludedthepossibilitythatthisndingwasduetoaninactivepreparationofGHusedforthestudiesbymonitoringtheabilityofGHtophosphorylatethesignalingmoleculeJAK-2;theGHreceptorlacksintrinsickinaseactivityanddependsontherecruitmentofJAK-2forphosphorylationoftheGH/GHRcomplexandsubsequentactivationofdown-streamsignalingpathways(19).
Consequently,GHisunabletoactivateJAK-2inskinbroblastsderivedfromindividualswithLaronsyndrome(20,21).
JAK-2waspresentandbasallyactiveinbothcelltypes,supportingpreviousobservationsinnormalbroblasts(20).
GHstimulatedthephosphorylationofJAK2inbothnormalandTScells,therebyconrmingGHbioactivity;however,inthelatter,activeJAK-2wasstillpresent60minaftertreatment,whichsuggeststhatthetempoofJAK-2activationmightdifferbetweenthetwocelltypes.
RecentstudieshaveidentieddefectswithintheGH/IGFsignalingcascadesinindividualswithgrowthdisorders;mu-tationsinSTAT-5bresultinGHinsensitivityandreducedgrowth(22)andshortstatureisalsoassociatedwithmutationsinthegenecodingforthetype1IGFreceptor(23).
TheseareFigure3.
IGF-I,butnotGH,stimulatestheproductionofIGFBP-3bybothnormalandTurnerbroblasts.
FibroblastswereculturedintheabsenceorpresenceofGH,IGF-I,oracombinationofbothhormonesforupto7dinculture.
Mediawasharvestedond3(AandD),5(BandE),and7(CandF)ofcultureandanalyzedforIGFBP-3(A–C)andIGFBP-4(D–F)byimmunoassay.
DataarepresentedasmeanSEM;whitebars—normalbroblasts,blackbars—Turnerbroblasts.
*denotessignicantdifferencebetweennormalandTurnerbroblasts(p0.
05);§denotessignicantdifferencesfromuntreated(basal)cells(p0.
05);denotessignicantdifferencesfromGH-treatedcells(p0.
05).
IGFBP-3andIGFBP-4levelsarereportedasng/mL;datacanbeconvertedtoSIunits(nmol/L)bymultiplyingby0.
035and0.
038,respectively.
Figure4.
TurnerbroblastsdonotrespondascontrolcellstotreatmentwithGHplusIGF-I.
FibroblastsfromnormalchildrenandchildrenwithTSwereculturedintheabsenceorpresenceofGH,IGF-I,oracombinationofbothhormonesforupto7d.
Mediaharvestedond7ofculturewasanalyzedforIGFBP-3byimmunoassay.
Thefoldchange,relativetothelevelofIGFBPmeasuredinthemediafromuntreatedcells,wascalculatedforeachcellline,anddataarepresentedasthemeanfoldincreasefornormal(Ⅺ)andTSbroblasts(f).
*denotessignicantdifferencebetweennormalandTurnerbroblasts(p0.
05);§denotessignicantdifferencesfromuntreated(basal)cells(p0.
05).
28WESTWOODETAL.
severegeneticabnormalitiesthatresultinextremephenotypesbutsuchobservationsdosupportthehypothesisthatlessseveresignalingdefectsmaybeassociatedwithgrowthre-strictioninconditionssuchasTS.
OurdatademonstratingthatIGFBP-3productionisregu-latedbyIGF-Iisconsistentwiththendingsofotherinvitro(24,25)andinvivo(26,27)studies,andIGF-IhasalsobeenshowntostimulateIGFBP-3inbroblastcellsisolatedfromindividualswithISS(17).
Inourstudy,IGF-IseemedtobemoreeffectiveinnormalbroblastsbecausethelevelofIGFBP-3producedbyTurnerbroblasts,particularlywhentreatedincombinationwithGHwassignicantlylessthanthatproducedbycontrolcells.
ThesendingsareincontrasttotheeffectofbroblastsfromchildrenwithISS,whereIGF-Iwashyperstimulatory(17)andsuggestthatinTS,theactivityofautocrine/paracrineIGFiscompromised.
Thereducedsen-sitivitytolocalIGF-Imay,inpart,beexplainedbyadecreaseinIGF-IbioavailabilitybecauseoftheincreasedbasalIGFBP-3productionbythesecells;however,ourdatasuggestthatTScellsmusthaveasecondabnormality,forexampledysfunctioninthesignalingpathwaysevokedbyIGF-I,be-causeANCOVAdemonstratedthatincomparisonwithcon-trolcells,theTScellresponsetoIGF-IwasattenuatedevenaftertheirexcessIGFBP-3productionhadbeentakenintoaccount.
Ourndingssupporttheconceptofend-organresis-tancetoIGF-IandperhapsexplainthenecessityfortreatinggirlswithTSwithsupraphysiologicaldosesofGH.
Interest-ingly,inastudyofchildrenbornsmall-for-GA,IGFBP-3levelsatage2–3ywerenegativelyrelatedtoadultheight,andGHefcacywasreducedinthepresenceofhighIGFBP-3levels(28).
ThebasalproductionofIGFBP-4byTSbroblastswassimilartothatofcellsobtainedfromnormalchildren,andthelevelofIGFBP-4presentinconditionedmediawasnotaf-fectedbyhormonestimulationineithercelltype.
OtherstudieshaveshownthatIGF-IdoesnotinuenceIGFBP-4mRNAexpression(29);however,therearedatatosuggestthatIGF-ImightreducethepresenceofIGFBP-4indermalbroblast-conditionedmedia(16,25)bystimulatingthepro-teolysisofthisbindingprotein.
Inourstudy,therewasnoevidenceofIGFBP-4proteolysisfromWesternligandblotanalysisoftheconditionedmediaobtainedfromcellstreatedwithGHand/orIGF-I.
However,IGFBP-3isthoughttoinhibitIGF-inducedproteolysisofIGFBP-4inamannerdependentontherelativeproportionsofIGFBPsandthelevelofIGF(25);asIGF-IstimulatedIGFBP-3productionbythecellsusedinthisstudy,thismayexplainwhywedidnotobserveanyIGFBP-4proteolysis.
Insummary,ourdatademonstratethatbroblastsisolatedfromchildrenwithTSproduceasimilarIGFBPproletocellsfromnormalchildren,butseemtohaveimpairedresponsive-nesstotreatmentwithIGF-I.
Thissuggeststhatalteredauto-crine/paracrineIGFbioavailabilitymaycontributetotheneedforhigh-doseGHtreatmentandtotherelativelypoorclinicalresponsetoGHofsomegirlswithTS.
Acknowledgments.
WethankDrJennyJonesforperform-ingtheIGFBP-4assaysreportedinthisarticle.
REFERENCES1.
SaengerP1996Turner'ssyndrome.
NEnglJMed335:1749–17542.
RochiccioliP,DavidM,MalpuechG,ColleM,LimalJM,BattinJ,MarianiR,SultanC,NivelonJL,SimoninG,BostM,ChatelainP,BertrandAM,FreyconR,BouquierJJ,PierronP1994StudyofnalheightinTurner'ssyndrome:ethnicandgeneticinuences.
ActaPaediatr83:305–3083.
ZadikZ,LandauH,ChenM,AltmanY,LiebermanE1992Assessmentofgrowthhormone(GH)axisinTurner'ssyndromeusing24-hourintegratedconcentrationsofGH,insulin-likegrowthfactor-I,plasmaGH-bindingactivity,GHbindingtoIM9cells,andGHresponsetopharmacologicalstimulation.
JClinEndocrinolMetab75:412–4164.
LanesR,BritoS,SuniagaM,MoncadaG,BorgesM1990GrowthhormonesecretioninpubertalagepatientswithTurner'ssyndrome.
JClinEndocrinolMetab71:770–7725.
WitJM,MassaranoAA,KampGA,HindmarshPC,vanEsA,BrookCG,PreeceMA,MatthewsDR1992GrowthhormonesecretioninpatientswithTurner'ssyndromeasdeterminedbytimeseriesanalysis.
ActaEndocrinol(Copenh)127:7–126.
MassaG,BouillonR,Vanderschueren-LodeweyckxM1992Serumgrowthhormone(GH)-bindingproteinandinsulin-likegrowthfactor-IlevelsinTurner'ssyndromebeforeandduringtreatmentwithrecombinanthumanGHandethinylestradiol.
JClinEndocrinolMetab75:1298–13027.
BaxterL,BryantJ,CaveCB,MilneR2007RecombinantgrowthhormoneforchildrenandadolescentswithTurnersyndrome.
CochraneDatabaseSystRev1:CD0038878.
RankeMB,LindbergA,Ferrandez-LongasA,DarendelilerF,Albertsson-WiklandK,DungerD,CuteldWS,TauberM,WiltonP,WollmannHA,ReiterEO2007MajordeterminantsofheightdevelopmentinTurnersyndrome(TS)patientstreatedwithGH:analysisof987patientsfromKIGS.
PediatrRes61:105–1109.
HochbergZ,AviramM,RubinD,PollackS1997Decreasedsensitivitytoinsulin-likegrowthfactorIinTurner'ssyndrome:astudyofmonocytesandTlymphocytes.
EurJClinInvest27:543–54710.
BarrecaA,LarizzaD,DamonteG,ArvigoM,PonzaniP,CesaroneA,LoCF,SeveriF,GiordanoG,MinutoF1997Insulin-likegrowthfactors(IGF-IandIGF-II)andIGF-bindingprotein-3productionbybroblastsofpatientswithTurner'ssyndromeinculture.
JClinEndocrinolMetab82:1041–104611.
FreethJS,AylingRM,WhatmoreAJ,TownerP,PriceDA,NormanMR,ClaytonPE1997Humanskinbroblastsasamodelofgrowthhormone(GH)actioninGHreceptor-positiveLaron'ssyndrome.
Endocrinology138:55–6112.
WestwoodM,GibsonJM,PennellsLA,WhiteA1999Modicationofplasmainsulin-likegrowthfactorsandbindingproteinsduringoralcontraceptiveuseandthenormalmenstrualcycle.
AmJObstetGynecol180:530–53613.
GravholtCH,FrystykJ,FlyvbjergA,OrskovH,ChristiansenJS2001ReducedfreeIGF-IandincreasedIGFBP-3proteolysisinTurnersyndrome:modulationbyfemalesexsteroids.
AmJPhysiolEndocrinolMetab280:E308–E31414.
YakarS,LiuJ-L,StannardB,ButlerA,AcciliD,SauerB,LeRoithD1999Normalgrowthanddevelopmentintheadsenceofhepaticinsulin-likegrowthfactorI.
ProcNatlAcadSciUSA96:7324–7329Figure5.
GHstimulatesthephosphorylationofJAK-2inbroblastcellsfrombothnormalandTurnerbroblasts.
Fibroblastsobtainedfromnormalchildren(A)andchildrenwithTS(B)wereserumstarvedfor24handthenstim-ulatedwithGHfor0–60min.
LysateswereanalyzedbyWesternblottingusingantibodiesthatspecicallyrecognizethephosphorylatedortotalisoformsofJAK-2.
Lysatesfrom3T3–F442AbroblastsincubatedintheabsenceorpresenceofGHwereincludedaspositivecontrols.
29CELLULARIGFAXISINTURNERSYNDROME15.
BatchJA,MercuriFA,EdmondsonSR,WertherGA1994Localizationofmessen-gerribonucleicacidforinsulin-likegrowthfactorbindingproteinsinhumanskinbyinsituhybridization.
JClinEndocrinolMetab79:1444–144916.
NeelyEK,RosenfeldRG1992Insulin-likegrowthfactors(IGFs)reduceIGF-bindingprotein-4(IGFBP-4)concentrationandstimulateIGFBP-3independentlyofIGFreceptorsinhumanbroblastsandepidermalcells.
Endocrinology130:985–99317.
KampGA,OuwensDM,HoogerbruggeCM,ZwindermanAH,MaassenJA,WitJM2002SkinbroblastsofchildrenwithidiopathicshortstatureshowanincreasedmitogenicresponsetoIGF-IandsecretemoreIGFBP-3.
ClinEndocrinol(Oxf)56:439–44718.
GargoskySE,WilsonKF,FielderPJ,VaccarelloMA,Guevara-AguirreJ,DiamondFB,BaxterRC,RosenbloomAL,RosenfeldRG1993Thecompositionanddistri-butionofinsulin-likegrowthfactors(IGFs)andIGF-bindingproteins(IGFBPs)intheserumofgrowthhormonereceptor-decientpatients:effectsofIGF-ItherapyonIGFBP-3.
JClinEndocrinolMetab77:1683–168919.
HansenLH,WangX,KopchickJJ,BoucheloucheP,NielsenJH,GalsgaardED,BillestrupN1996IdenticationoftyrosineresiduesintheintracellulardomainofthegrowthhormonereceptorrequiredfortranscriptionalsignalingandStat5acti-vation.
JBiolChem271:12669–1267320.
SilvaCM,KlothMT,WhatmoreAJ,FreethJS,AndersonN,LaughlinKK,HuynhT,WoodallAJ,ClaytonPE2002GHandepidermalgrowthfactorsignalinginnormalandLaronsyndromebroblasts.
Endocrinology143:2610–261721.
FreethJS,SilvaCM,WhatmoreAJ,ClaytonPE1998Activationofthesignaltransducersandactivatorsoftranscriptionsignalingpathwaybygrowthhormone(GH)inskinbroblastsfromnormalandGHbindingprotein-positiveLaronSyndromechildren.
Endocrinology139:20–2822.
KofoedEM,HwaV,LittleB,WoodsKA,BuckwayCK,TsubakiJ,PrattKL,BezrodnikL,JasperH,TepperA,HeinrichJJ,RosenfeldRG2003GrowthhormoneinsensitivityassociatedwithaSTAT5bmutation.
NEnglJMed349:1139–114723.
AbuzzahabMJ,SchneiderA,GoddardA,GrigorescuF,LautierC,KellerE,KiessW,KlammtJ,KratzschJ,OsgoodD,PfafeR,RaileK,SeidelB,SmithRJ,ChernausekSD2003IGF-Ireceptormutationsresultinginintrauterineandpostnatalgrowthretardation.
NEnglJMed349:2211–222224.
ConoverCA1990Regulationofinsulin-likegrowthfactor(IGF)-bindingproteinsynthesisbyinsulinandIGF-Iinculturedbovinebroblasts.
Endocrinology126:3139–314525.
DonnellyMJ,HollyJM1996RheroleofIGFBP-3intheregulationofIGFBP-4proteolysis.
JEndocrinol149:R1–R726.
JrgensenJO,BlumWF,MllerN,RankeMB,ChristiansenJS1991Short-termchangesinseruminsulin-likegrowthfactors(IGF)andIGFbindingprotein3afterdifferentmodesofintravenousgrowthhormone(GH)exposureinGH-decientpatients.
JClinEndocrinolMetab72:582–58727.
VillafuerteBC,ZhangWN,PhillipsLS1996Insulinandinsulin-likegrowthfactor-Iregulatehepaticinsulin-likegrowthfactorbindingprotein-3bydifferentmecha-nisms.
MolEndocrinol10:622–63028.
deRidderMA,StijnenT,Hokken-KoelegaAC2008Predictionmodelforadultheightofsmallforgestationalagechildrenatthestartofgrowthhormonetreatment.
JClinEndocrinolMetab93:477–48329.
Camacho-HubnerC,BusbyWHJr,McCuskerRH,WrightG,ClemmonsDR1992Identicationoftheformsofinsulin-likegrowthfactor-bindingproteinsproducedbyhumanbroblastsandthemechanismsthatregulatetheirsecretion.
JBiolChem267:11949–1195630WESTWOODETAL.