Wnt antagonist sclerostin and Dickkopf-1 in gestational diabetes

(1)

Pleasecitethisarticleinpressas:CatalanoA,etal.WntantagonistsclerostinandDickkopf-1ingestationaldiabetes.DiabetesMetab(2016),

http://dx.doi.org/10.1016/j.diabet.2016.09.009

ARTICLE IN PRESS

+Model

DIABET-821; No.ofPages3

Availableonlineat

ScienceDirect

www.sciencedirect.com

Diabetes&Metabolismxxx(2016)xxx–xxx

Research

letter

Wnt antagonist sclerostin and Dickkopf-1 in gestationaldiabetes

1. Introduction

Wnt secretoryglycoproteins are a family of

developmen-tallyimportantsignallingmoleculesthat playimportantroles

inembryonicinduction,generationofcellpolarityand

specifi-cationofcellfate.Canonicalandnon-canonicalWntpathways

havebeenidentified [1].TheimportanceofWntsignallingin

diabetes arose after the observation by Kanazawa et al. [2],

whoshowed thatasinglepolymorphism locusintheWNT5B

genemaycontributetosusceptibilitytotype2diabetesmellitus

(T2DM)andmaybeinvolvedinthepathogenesisofthe

disor-derthroughregulationofadipocytefunction.Emergingevidence

alsosupportstheeffectsofalteredWntsignallingon

cardiovas-cularrisk factors[1–3].Inanimalmodels, theWnt/␤-catenin

signallingpathwayhasbeenshowntocontributetomodulation

of insulin secretion,␤-cell function andinsulin signalling in

skeletalmuscle[4].Accordingtothecanonicalpathway,

bind-ing of the appropriate Wntligands to aco-receptor complex

involvingFrizzledandlow-densitylipoproteinreceptor-related

protein(LRP)-5or-6stabilizescytoplasmic␤-cateninprotein,

whichtranslocatestothe nucleus andactivatesthe

transcrip-tionof targetgenes.TheWnt/␤-catenincanonical pathwayis

modulatedbyanumberoffactors,includingsecretedproteins

suchasDickkopf-1(Dkk-1)andsclerostin,whichprevent

for-mationoftheWnt–Frizzled–LRP5complexbyinternalization

of theLRP5/6co-receptor andcompetitivebinding toLRP5,

respectively[1].Also,Nuche-Berengueretal.[5]demonstrated

the upregulation of gene expressionof Dkk-1 and sclerostin

inT2DMrats,andrevealedthatsclerostinoverexpressionwas

associated withincreased mRNAlevels of activatorLRP5 in

insulin-resistant rats. Consistent with these findings, human

studieshavereportedsignificantlyhigherserumsclerostin

lev-elsinT2DMpatientsthanincontrols;beyondT2DM,sclerostin

ispositivelyassociatedwiththemainfeaturesofthemetabolic

syndrome(MetS),includingobesityanddyslipidaemia,aswell

asatheroscleroticdisease[3,6].

Gestationaldiabetes(GDM)isdefinedasdiabetesinduced

bypregnancy,but whichresolvesat the endof pregnancy.It

usually developsinlate pregnancywhen insulinantagonistic

hormonespeak,leading toinsulinresistance,glucose

intoler-ance and hyperglycaemia. However, GDM may also lead to

severalseriousmaternalandfoetalcomplications,and

consti-tutesasignificantriskfactorforthesubsequentdevelopmentof

T2DMandcardiovasculardiseaseinlaterlife[7].

Todate,fewdatahavebeenreportedforthelevelsand

asso-ciationsofsclerostininwomendiagnosedwithGDM.AsGDM

canserveasamodelofpre-T2DMandbecausesclerostin

lev-elsare increased inthosewithprediabetes, ourpresentstudy

investigatedbothsclerostinandDkk-1serumlevelsinwomen

withGDMcomparedwithhealthy,non-diabeticpregnant

con-trols, andalso lookedfor possibleassociationsbetweenthese

Wntantagonistsandcertainmaternal/foetaloutcomes.

2. Materialsandmethods

This was acase–control study involving pregnant women

attending the Diabetes Outpatient Unitof the Departmentof

Internal Medicine at the G. Martino University Hospital in

Messina, Italy. Women were referred tothe Unitfor an oral

glucosetolerancetest(OGTT)forthedetectionofGDM.They

were included inthe study if theywere aged ≥18 yearsand

willingtogivetheirinformedconsent.Exclusioncriteriawere

the presenceof renalorliverfailure,severeheartfailureora

psychiatricdisorder.Pregnantwomenwithestablishedrisk

fac-tors for GDM at gestational weeks 24–28 underwent a75-g

OGTT, withcut-off valuesof 5.1mmol/Lfor fastingglucose,

and10.0mmol/Land8.5mmol/Lfor1-hand2-hpost-load

glu-coselevels,respectively,andwereconsideredeligibleaccording

toInternationalAssociationof DiabetesandPregnancyStudy

Groups(IADPSG)criteria[6].

Overaperiodof6months(fromOctober2014toApril2015),

35consecutivewomenwithGDMwererecruited,whileagroup

ofpregnantwomenwhowerenegativeonthescreeningtestwere

randomlyselected,usingacomputer-generatedrandomization

table,toserveasthecontrolgroup.

Bloodsamplesweredrawntocheckinsulin,sclerostinand

Dkk-1levelsatthesametimeastheOGTT.Insulinresistance

wascalculatedbyhomoeostasismodelassessment forinsulin

resistance (HOMA-IR).Enzyme immunoassays wereused to

measurelevelsofinsulin(BeckmanCoulter,Brea,CA,USA),

sclerostin and Dkk-1 (BiomedicaMedizinprodukte GmbH&

CoKG,Vienna,Austria),whichhadintra-andinterassay

coef-ficientsofvariation(CVs)<7%forallanalyses.Therecruited

womenwerefolloweduntildelivery.Informationonthe

follow-ingparameterswascollected:age;height;pregestationalweight;

(2)

ARTICLE IN PRESS

+Model

2 Researchletter/Diabetes&Metabolismxxx(2016)xxx–xxx

Table1

Mainclinicalcharacteristicsofthestudiedpopulation.

Overall Normalglucosetolerance Gestationaldiabetes Pvalue

n 71 36 35

Age(years) 33.8±4.9 33.4±5.3 34.1±4.5 0.55

Height(cm) 162.7±6.1 161.6±6.0 163.9±5.9 0.11

Pregestationalweight(kg) 66.4±13.5 62.5±11.2 70.5±14.7 0.01

PregestationalBMI(kg/m2₎ _25.1_±_5.0 _24.1_±_4.5 _26.2_±_5.4 _0.08

Familyhistoryofdiabetes(%) 59.5 33.3 66.7 0.12

Previousgestationaldiabetes(%) 14.3 0.0 20.0 0.53

BaselineBGonOGTT(mg/dL) 84.4±10.1 76.5±5.0 92.8±6.8 <0.0001

1-hpost-loadBGonOGTT(mg/dL) 134.5±36.2 113.7±24.4 156.5±33.7 <0.0001 2-hpost-loadBGonOGTT(mg/dL) 112.8±28.8 96.1±19.1 130.6±26.8 <0.0001

InsulinonOGTT(uU/mL) 18.2±11.0 13.1±6.4 23.5±12.1 <0.0001

HOMA-IR 3.5±2.3 2.3±1.2 4.8±2.5 <0.0001

Sclerostinlevels(pmol/L) 24.5±15.2 25.0±14.8 24.0±15.8 0.79

Dkk-1levels(pmol/L) 10.9±3.9 11.4±3.9 10.3±4.0 0.26

Weightatdelivery(kg) 77.1±12.7 74.3±10.9 80.1±13.9 0.06

BMIatdelivery(kg/m2₎ _29.1_±_4.7 _28.5_±_4.2 _29.6_±_5.2 _0.32

Weightgain(kg) 10.5±3.2 11.7±2.8 9.1±3.2 0.001

Gestationalweekatdelivery(weeks) 38.6±1.5 38.9±1.6 38.3±1.5 0.13

Caesareansection(%) 40.3 44.4 37.1 0.63

Femalenewborn(%) 52.8 50.0 48.6 0.99

Neonatalweight(g) 3112.4±437.2 3144.4±487.2 3078.5±381.6 0.53

Apgarscoreat1min 9.2±0.9 9.2±0.8 9.2±1.1 0.78

Apgarscoreat5min 9.8±0.4 9.7±0.4 9.8±0.4 0.23

Dataarepresentedasmeans±SDoraspercentages(%).

BMI,bodymassindex;BG,bloodglucose;OGTT,oralglucosetolerancetest;HOMA-IR,homoeostasismodelassessmentforinsulinresistance.

familyhistoryofdiabetes;historyofpreviousGDM;weightat

delivery;weightgain;gestationalweekatdelivery;caesarean

sectionrate;genderofthenewborn;neonatalweight;andApgar

1-and5-minscores.

The studywas conducted inaccordancewiththe

Declara-tionofHelsinki,andallparticipantsgavetheirwritteninformed

consent.

2.1. Statisticalanalyses

Data wereexpressed as means±SD for continuous

varia-bles and as percentages for categorical variables. The

Kolmogorov–Smirnov test was used to test the normality of

distributionofcontinuousvariables.Clinicalanddemographic

characteristicswerecomparedusingthechi-squaretestfor

cat-egoricalvariables andthe Kruskal–Wallistestfor continuous

variables. Pearson’s correlation coefficient was employed to

testcorrelationsbetweensclerostin, Dkk-1,otherwell-known

risk factorsfor GDM and pregnancy outcomes. Multivariate

regression modelswere performed toanalyzethe association

of sclerostin and Dkk-1 withGDM onsetas well as the

fol-lowingcovariates:age;pregestationalbodymassindex(BMI);

familyhistoryofdiabetes;HOMA-IR;sclerostin;andDkk-1.A

Pvalue<0.05wasconsideredstatisticallysignificant.Analyses

wereperformedusingIBMSPSSversion21.0software(IBM

Corp.,Armonk,NY,USA).

3. Results

Overall,71womenwereincluded inourstudy;their main

characteristicsarepresentedinTable1.Oncomparingwomen

withnormal glucosetolerance(NGT) withthosewithGDM,

thelatterhadsignificantlyhigherpregestationalweight,blood

glucoselevelsatallOGTTpoints,insulinlevelsandHOMA-IR

scores, and less weight gain (Table 1). No between-group

differences were detected for age, height, family history of

diabetes, history of previous GDM, or sclerostin or Dkk-1

levels. As for pregnancy outcomes,the two groupswere not

statistically differentintermsof gestationalweekatdelivery,

caesarean section rate, newborn gender,neonatal weightand

Apgarscoresat1and5min(Table1).

Correlationanalysesshowedthatsclerostincorrelatedonly

withpregestationalBMI;incontrast,Dkk-1correlatedwithnone

ofthetestedvariables.Themultivariateregressionmodelalso

showednoassociationbetweeneithersclerostin(OR:0.98,95%

CI: 0.87–1.10)or Dkk-1 (OR:1.73, 95% CI:0.91–3.30)and

GDMonset.

4. Discussion

Pregnancy is a challenging period for the mother’s bones

becauseofthedevelopmentofthefoetalskeleton;consequently,

surrogatemarkersofboneresorptionandboneformationmaybe

observedtovaryovertime,withapredominanceofbone

resorp-tion during thefirst andsecond trimesters.To date,however,

maternal placental–foetal mineralhomoeostasishasremained

largelypoorlyunderstood,andtherearefewdataregardingthe

roleofWntantagonists,especiallyinGDM.

Ourpresentstudyinvestigatedserumlevelsofsclerostinand

Dkk-1inpregnantwomenwithGDMandinNGTcontrols.No

significantdifferenceswerefoundinmaternalserumlevelsof

(3)

ARTICLE IN PRESS

+Model

Researchletter/Diabetes&Metabolismxxx(2016)xxx–xxx 3

noassociationscouldbefoundbetweensclerostinorDkk-1with

themainmetabolicmaternalfeaturesorwithfoetaloutcomes.

The close connection between bone tissue and glucose

homoeostasiswasrecentlyhighlighted,andosteocalcin(BGP),

oneoftheveryfewosteoblast-specificproteinsand,thus,a

sur-rogatemarker ofbone formation,emergedas ahormonethat

canregulate␤-cellproliferation, insulinsecretionandinsulin

sensitivity.BGPhasalsobeenreportedlydecreasedinpatients

withT2DMandnegativelycorrelatedwithfastingplasma

glu-cose,HbA1c,HOMA-IRandBMI,butincreasedwithimproved

glycaemiccontrol.Accordingly,thepossibleroleofbonein

con-trollingglucosehomoeostasishasbeenhypothesized,andithas

beenevenreportedthatBGPwassignificantlyhigherinwomen

withGDMthaninNGTwomen[6,8].

BothsclerostinandDkk-1interactwithLRP5andLRP6,and

antagonizethecanonicalWnt/␤-cateninsignallingpathway,the

activationofwhichleadstoanincreasedproliferationand

differ-entiationofosteoblastprecursorcellsandreducedapoptosisof

matureosteoblasts,whilepromotingtheabilityofdifferentiated

osteoblaststoinhibitosteoclastdifferentiation[1].Therefore,it

maybespeculatedthatWntantagonistscouldactasmodulators

ofBGPproductionthroughregulationofosteoblastactivity.

The Wnt/␤-catenin signalling pathway is involved in the

pathogenesisofobesityandT2DM[1,3].Itwasalsoobserved

thattheTCF7L2gene,whichencodesanuclear-bindingfactor

for␤-catenin,wasassociatedwithfeaturesoftheMetS,

includ-ingelevatedsystolicanddiastolicbloodpressures,andraised

levelsofglucose,cholesterol,triglycerideanduricacid[3].

However,inpregnantwomen,Platzetal.[9]recentlyreported

thatsclerostindoesnotcorrelatewithanyfeatureoftheMetS,

incontrasttowhathasbeenobservedinothercohorts[6,10].

Infact,consistentwithpreviousdataobtainedduringpregnancy

[9],ourpresentstudyalsoobservednoassociationbetweenWnt

antagonistlevelsandourmainmaternalandfoetaloutcomes.

Althoughourstudyhasafewlimitations(smallsamplesize,

singletimepointofmeasurements),itisthefirsttoinvestigate

bothsclerostinandDkk-1inacohortofpregnantwomenwhose

deliveryoutcomeswererecorded.Ourpresentfindingssuggest

thatsclerostinandDkk-1donotplayasignificantroleinthe

pathophysiologyofGDM,althoughfurtherresearchisneeded

toexploretheirassociationswithhormonalstatusandperhaps

indifferenttrimesters.

Inconclusion,inourcohortof pregnantwomen,sclerostin

and Dkk-1 were not associated with any adverse metabolic

profile,and possiblydo not play relevant roles inthe

patho-physiologyofGDM.

Funding

Thisresearchreceivednospecificgrantsfromfunding

agen-ciesinthepublic,commercialornot-for-profitsectors.

Disclosureofinterest

Theauthorsdeclarethattheyhavenocompetinginterest.

References

[1]MacDonaldBT,TamaiK,HeX.Wnt/beta-cateninsignaling:components, mechanisms,anddiseases.DevCell2009;17:9–26.

[2]KanazawaA,TsukadaS,SekineA,TsunodaT,TakahashiA,Kashiwagi A,etal.Associationofthegeneencodingwingless-typemammarytumor virusintegration-sitefamilymember5B(WNT5B)withtype2diabetes. AmJHumGenet2004;75:832–43.

[3]GrantSF,ThorleifssonG,ReynisdottirI,BenediktssonR,ManolescuA, SainzJ,etal.Variantoftranscriptionfactor7-like2(TCF7L2)geneconfers riskoftype2diabetes.NatGenet2006;38:320–3.

[4]ElghaziL,GouldAP,WeissAJ,BarkerDJ,CallaghanJ,OplandD,etal. Importance of ␤-cateninin glucose and energyhomeostasis. SciRep 2012;2:693.

[5]Nuche-Berenguer B, Moreno P, Portal-Nu˜nez S, Dapía S, Esbrit P, Villanueva-Pe˜nacarrillo ML. Exendin-4 exerts osteogenic actions in insulin-resistantandtype2diabeticstates.RegulPept2010;159:61–6. [6]DanieleG,WinnierD,MariA,BruderJ,FourcaudotM,PengouZ,etal.

Sclerostinandinsulinresistanceinpre-diabetes:evidenceofacrosstalk betweenboneandglucosemetabolism.DiabetesCare2015;38:1509–17. [7]InternationalAssociationofDiabetesPregnancyStudyGroupsConsensus

Panel,MetzgerBE,GabbeSG,PerssonB,BuchananTA,CatalanoPA, etal.Internationalassociationofdiabetesandpregnancystudygroups recommendationsonthediagnosisandclassificationofhyperglycemiain pregnancy.DiabetesCare2010;33:676–82.

[8]WinhoferY,HandisuryaA,TuraA,BittighoferC,KleinK,SchneiderB, etal.Osteocalcinisrelatedtoenhancedinsulinsecretioningestational diabetesmellitus.DiabetesCare2010;33:139–43.

[9]PlatzM,StepanH,SchreyS,KralischS,WurstU,LossnerU,etal. Serumlevelsofsclerostinincardiometabolicdisordersduringpregnancy. Cytokine2015;76:591–3.

[10]CatalanoA,PintaudiB,MorabitoN,DiViesteG,GiuntaL,BrunoML, etal.Genderdifferencesinsclerostinandclinicalcharacteristicsintype1 diabetesmellitus.EurJEndocrinol2014;171:293–300.

A.Catalanoa,∗,1 B.Pintaudib,1 N.Morabitoa L.Giuntaa S.Loddoa F.Corradoc R.D’Annac A.Lascoa,2 A.DiBenedettoa,2

a_Department_of_Clinical_and_Experimental_Medicine,

UniversityHospitalofMessina,Messina,Italy

b_SSD_Diabetology,_Ca’Granda_Niguarda_Hospital,_Milan,

Italy

c_Department_of_Pediatric,_{Gynecological,}_{Microbiological}_and

BiomedicalSciences,Messina,Italy

∗_{Corresponding}_author_at:_Department_of_Clinical_and

ExperimentalMedicine,UniversityHospitalofMessina,Via

C.Valeria,98125Messina,Italy.Tel.:+390902213946;fax:

+390902935162.

E-mailaddress:catalanoa@unime.it(A.Catalano)

1_These_authors_contributed_equally.

2_Senior_authors.

Received16September2016

Receivedinrevisedform26September2016