Science in medicine

Pathogenesis of osteoporosis: concepts, conflicts, and prospects

Lawrence G. Raisz
University of Connecticut Health Center, Musculoskeletal Institute, Farmington, Connecticut, USA.

Osteoporosis is a disorder in which loss of bone strength leads to fragility fractures. This review examines the fundamental pathogenetic mechanisms underlying this disorder, which include: (a) failure to achieve a skeleton of optimal strength during growth and development; (b) excessive bone resorption resulting in loss of bone mass and disruption of architecture; and (c) failure to replace lost bone due to defects in bone formation. Estrogen deficiency is known to play a critical role in the development of osteoporosis, while calcium and vitamin D deficiencies and secondary hyperparathyroidism also contribute. There are multiple mechanisms underlying the regulation of bone remodeling, and these involve not only the osteoblastic and osteoclastic cell lineages but also other marrow cells, in addition to the interaction of systemic hormones, local cytokines, growth factors, and transcription factors. Polymorphisms of a large number of genes have been associated with differences in bone mass and fragility. It is now possible to diagnose osteoporosis, assess fracture risk, and reduce that risk with antiresorptive or other available therapies. However, new and more effective approaches are likely to emerge from a better understanding of the regulators of bone cell function.
Osteoporosis,characterizedbythelossofbonemassandstrength thatleadstofragilityfractures,hasprobablyexistedthroughout humanhistorybutonlyrecentlybecameamajorclinicalproblemas humanlifespanincreased.Intheearly19thcentury,SirAstleyCooper,adistinguishedEnglishsurgeon,notedthelightnessandsoftnessthat(bones)acquireinthemoreadvancedstagesoflifeand thatthisstateofbone...favorsmuchtheproductionoffractures (1).ThetermosteoporosiswascoinedbyJohannLobsteinatabout thesametime,butthedisorderhedescribedwasprobablyosteogenesisimperfecta(2).In1940,theAmericanphysicianandendocrinologistFullerAlbrightdescribedpostmenopausalosteoporosis andproposedthatitwastheconsequenceofimpairedboneformationduetoestrogendeficiency(3).Subsequently,theconceptthat thereare2formsofosteoporosis,onerelatedtoestrogendeficiency atthemenopauseandtheothertocalciumdeficiencyandagingof theskeleton,wasproposed(4).Thishasbeenreplacedbythecurrent conceptthatosteoporosisrepresentsacontinuum,inwhichmultiplepathogeneticmechanismsconvergetocauselossofbonemass andmicroarchitecturaldeteriorationofskeletalstructure.These factors,coupledwithanincreasedriskoffalls,contributetoahigh incidenceoffragilityfracturesinosteoporoticpatients. Thereisarapidlyexpandingamountofinformation,based onlaboratorystudies,thatindicatesthatosteoporosisislikely
Nonstandard abbreviations used:BMD,bonemineraldensity;BMP2,bonemorphogeneticprotein2;BMU,bonemulticellularunit;COX2,cyclooxygenase2;ER, estrogenreceptor;LRP5,LDLreceptorrelatedprotein5;OPG,osteoprotegerin; PGE2,prostaglandinE2;PTH,parathyroidhormone;RANK,receptoractivatorof NF-B;RANKL,RANKligand;SNP,singlenucleotidepolymorphism;VDR,vitamin Dreceptor. Conflict of interest:LawrenceG.RaiszchairsDataSafetyandMonitoringBoards forNovartis,whichmanufacturesanddistributesbisphosphonatesforthetreatment ofbonemetastases.HereceivesresearchsupportfromServierInternational,which manufacturesanddistributesstrontiumranelateforthetreatmentofosteoporosis. HeservedasScientificEditorfortheUSSurgeonGenerals2004reportBoneHealth andOsteoporosis(111). Citation for this article:J. Clin. Invest.115:33183325(2005). doi:10.1172/JCI27071.
3318

tobecausedbycomplexinteractionsamonglocalandsystemic regulatorsofbonecellfunction.Theheterogeneityofosteoporosismaybeduenotonlytodifferencesintheproductionof systemicandlocalregulators,butalsotochangesinreceptors, signaltransductionmechanisms,nucleartranscriptionfactors, andenzymesthatproduceorinactivatelocalregulators.Withinthelastdecade,theidentificationofmanyoftheregulatory mechanismsthathavebeenlinkedtoosteoporosishasbeenthe resultofgeneticstudies.Sincethefirsthumanosteoporosisstudy indicatedanassociationamongbonemass,fragility,andpolymorphismsinthevitamin D receptor (VDR)gene,morethan30 candidategeneshavebeenreportedthatmightinfluenceskeletal massandfragility(5,6).However,thesestudieshavepresented conflictingdata,dueinparttosmallsamplesizeanddifferences inthegeneticbackgroundofcontrolanddiseasesubjects.Since osteoporosisisacomplex,polygenicdisorder,thecontributions ofspecificgenepolymorphismsarelikelytoberelativelysmall, butmaystillbeclinicallyimportant.Largecohortstudiesusing standardizedgenotypingmethodologyandgeneticallysimilar controlandaffectedindividualsareneededtobetterdefinethe roleofspecificgenesinosteoporosispathogenesis,asexemplified byarecentstudyonestrogenreceptorpolymorphisms(7). Basic pathogenetic mechanisms Skeletalfragilitycanresultfrom:(a)failuretoproduceaskeletonof optimalmassandstrengthduringgrowth;(b)excessiveboneresorptionresultingindecreasedbonemassandmicroarchitecturaldeteriorationoftheskeleton;and(c)aninadequateformationresponse toincreasedresorptionduringboneremodeling.Inaddition,the incidenceoffragilityfractures,particularlyofthehipandwrist,is furtherdeterminedbythefrequencyanddirectionoffalls. Tounderstandhowexcessiveboneresorptionandinadequate formationresultinskeletalfragility,itisnecessarytounderstand theprocessofboneremodeling,whichisthemajoractivityofbone cellsintheadultskeleton.Theboneremodelingorbonemulticellularunits(BMUs)describedmanyyearsagobyFrostandothers

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

science in medicine
(8)canoccureitheronthesurfaceoftrabecularboneasirregular Howshiplacunaeorincorticalboneasrelativelyuniformcylindricalhaversiansystems.AsillustratedinFigure1,theprocessbegins withtheactivationofhematopoieticprecursorstobecomeosteoclasts,whichnormallyrequiresaninteractionwithcellsofthe osteoblasticlineage.Becausetheresorptionandreversalphases ofboneremodelingareshortandtheperiodrequiredforosteoblasticreplacementoftheboneislong,anyincreaseintherateof boneremodelingwillresultinalossofbonemass.Moreover,the largernumberofunfilledHowshiplacunaeandhaversiancanals willfurtherweakenthebone.Excessiveresorptioncanalsoresult incompletelossoftrabecularstructures,sothatthereisnotemplateforboneformation.Thus,therearemultiplewaysinwhich anincreaseinosteoclasticresorptioncanresultinskeletalfragility. However,highratesofresorptionarenotalwaysassociatedwith boneloss;forexample,duringthepubertalgrowthspurt.Hence aninadequateformationresponseduringremodelingisanessentialcomponentofthepathogenesisofosteoporosis. one-quarterofthedoseofestrogenthatstimulatesthebreastand uterusissufficienttodecreaseboneresorptionandincreasebone massinolderwomen(13).Thisgreatersensitivityoftheskeleton maybeagerelated.In3-month-oldmice,theuterusappearedto bemoreresponsivetoestrogenthanbone,whereasin6-month-old mice,thereversewasfound(14). Estrogeniscriticalforepiphysealclosureinpubertyinbothsexes andregulatesboneturnoverinmenaswellaswomen.Infact,estrogenhasagreatereffectthanandrogenininhibitingboneresorptioninmen,althoughandrogenmaystillplayarole(15).Estrogen mayalsobeimportantintheacquisitionofpeakbonemassinmen (16).Moreover,osteoporosisinoldermenismorecloselyassociated withlowestrogenthanwithlowandrogenlevels(17). Estrogendeficiencyincreasesandestrogentreatmentdecreases therateofboneremodeling,aswellastheamountofbonelost witheachremodelingcycle.Studiesinanimalmodelsandincell culturehavesuggestedthatthisinvolvesmultiplesitesofestrogen action,notonlyonthecellsoftheBMU,butonothermarrow cells(Figure1).Estrogenactsthrough2receptors:estrogenreceptor(ER)andER.ERappearstobetheprimarymediator ofestrogensactionsontheskeleton(10).Osteoblastsdoexpress

Central role of estrogen Theconceptthatestrogendeficiencyiscriticaltothepathogenesis of osteoporosis was based initially onthefactthatpostmenopausalwomen, whose estrogen levels naturally decline, areatthehighestriskfordevelopingthe disease.Morphologicstudiesandmeasurementsofcertainbiochemicalmarkershave indicatedthatboneremodelingisacceleratedatthemenopause,asbothmarkersof resorptionandformationareincreased(8, 9).Hence,contrarytoAlbrightsoriginal hypothesis,anincreaseinboneresorption, andnotimpairedboneformation,appears to be the driving force for bone loss in the setting of estrogen deficiency. But therapidandcontinuousbonelossthat occurs for several years after the menopause must indicate an impaired bone formationresponse,sinceinyoungerindividualsgoingthroughthepubertalgrowth spurt,evenfasterratesofboneresorption canbeassociatedwithanincreaseinbone mass.However,theincreasedboneformationthatnormallyoccursinresponseto mechanicalloadingisdiminishedinestro- Figure 1 gendeficiency,suggestingestrogenisboth The BMU with possible sites of estrogen action. Bone remodeling on the surface of trabecular bone is illustrated here. The process is similar in the haversian systems in cortical bone. Osteoanti-catabolicandanabolic(10). clast activation is ordinarily initiated by an interaction of hematopoietic precursors with cells of Estrogen deficiency continues to play the osteoblast lineage but may also be initiated by inflammatory cells, particularly T cells (see aroleinbonelossinwomenintheir70s Figure 2 for further details). Once osteoclasts are formed, there is a resorption phase of limited and80s,asevidencedbythefactthatestro- duration and a brief reversal phase, during which the bone surface is covered by mononuclear gentreatmentrapidlyreducesbonebreak- cells but bone formation has not yet begun. The formation phase is then initiated, possibly by downintheseolderwomen(11).Moreover, factors produced by the osteoclast or reversal cells or released from the bone matrix. The forrecentstudiesinhumanshaveshownthat mation phase, which is substantially longer than the first 3 phases, involves the production of thelevelofestrogenrequiredtomaintain matrix by progressive waves of osteoblasts. These then become flat lining cells, are embedded relativelynormalboneremodelinginolder in the bone as osteocytes, or undergo apoptosis. Potential sites of action of estrogen include effects on T cell cytokine production (i); effects on stromal or osteoblastic cells to alter their postmenopausalwomenislowerthanthat production of RANKL or OPG (ii); direct inhibition of differentiated osteoclasts (iii); and effects requiredtostimulateclassictargettissues on bone formation mediated by osteoblasts or osteocytes to enhance the response to mechanisuchasthebreastanduterus(12).Fracture cal forces initiated by these cells (iv). Note that the BMU is shown as being compartmentalized riskisinverselyrelatedtoestrogenlevels by an overlying layer of cells. It has been proposed that these are separated lining cells, but inpostmenopausalwomen,andaslittleas they may be of vascular origin.
The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005 3319

science in medicine
Figure 2
Regulation of osteoclast formation and activity. In physiologic remodeling, activation of bone resorption requires contact between cells of the osteoblast and osteoclast lineages. M-CSF, which may be either membrane bound or secreted, interacts with its receptor, c-fms, to stimulate differentiation and proliferation of hematopoietic progenitors, which then express RANK as preosteoclasts. Osteoclast differentiation and activity are stimulated by RANK/ RANKL interaction, and this interaction can be blocked by soluble OPG. Bone-resorbing factors can also stimulate COX2 activity, which may amplify responses to RANKL and OPG by producing prostaglandins. In pathologic conditions, inflammatory and malignant cells can increase osteoclastogenesis by producing soluble or membranebound M-CSF and RANKL as well as PTH-related protein (PTHrP), cytokines, and prostaglandins.

ER,buttheactionsofERagonistsonbonearelessclear.Some studiessuggestthattheeffectsofestrogensignalingthroughER andERareinopposition,whileotherstudiessuggestthatactivationofthese2receptorshassimilareffectsonbone(18,19). Singlenucleotidepolymorphisms(SNPs)ofERmayaffect bonefragility.Inthelargeststudytodate,1oftheSNPsforthis receptorwasassociatedwithasignificantreductioninfracture risk,independentofbonemineraldensity(BMD)(7).OtherstudieshavesuggestedthatSNPsofERcanaffectBMDandratesof bonelossaswellasfractureriskinbothmenandwomen(20,21). Anorphannuclearreceptor,estrogenreceptorrelatedreceptor(ERR),withsequencehomologytoERandER,isalso presentinbonecells(22).Despiteitsinabilitytobindestrogens, thisreceptormayinteractwithERandERoractdirectlyto alterbonecellfunction.AregulatoryvariantofthegeneencodingERRwasrecentlyfoundtobeassociatedwithasignificant difference between lumbar spine and femoral neck BMD in premenopausalwomen(23). Sexhormonebindingglobulin(SHBG),themajorbindingproteinforsexsteroidsinplasma,maynotonlyalterthebioavailability ofestrogentohormone-responsivetissuesbutalsoaffectitsentry intocells.EpidemiologicstudiessuggestthatSHBGmayhavean effectonbonelossandfractureriskindependentoftheeffectasa bindingprotein(24).Localformationofestradiolfromaromatase couldplayanadditionalrole(25). Whileestrogencanactoncellsoftheosteoblasticlineage,its effectsonbonemayalsobedependentonactionsoncellsofthe hematopoieticlineage,includingosteoclastprecursors,mature osteoclasts,andlymphocytes.Localcytokinesandgrowthfactors maymediatetheseeffects.Bonelossafterovariectomyinrodent modelscanbepreventedbyinhibitingIL-1orTNF-anddoes notoccurinmicedeficientintheIL-1receptororTNF-(26,27). Theeffectsofestrogenoncytokineproductionmaybemediated byTcells(28).Adirecteffectofestrogeninacceleratingosteoclast apoptosishasbeenattributedtoincreasedTGF-production(29).
3320

Anotherpossibilityisthatestrogenexertsitsbeneficialeffectsby suppressingROS(30).Inestrogendeficiency,thiolantioxidant defensesmaybediminished,andtheresultantincreaseinROS mayinduceTNF-(31).Therelevanceofthesefindingsforhuman osteoporosishasyettobedetermined. Calcium, vitamin D, and parathyroid hormone Theconceptthatosteoporosisisdueprimarilytocalciumdeficiency,particularlyintheelderly,wasinitiallyputforwardas a counterproposal to Albrights estrogen deficiency theory. Decreasedcalciumintake,impairedintestinalabsorptionofcalciumduetoagingordisease,aswellasvitaminDdeficiencycan resultinsecondaryhyperparathyroidism.Theactivehormonal form,1,25dihydroxyvitaminD(calcitriol),isnotonlynecessary foroptimalintestinalabsorptionofcalciumandphosphorus, butalsoexertsatonicinhibitoryeffectonparathyroidhormone (PTH)synthesis,sothattherearedualpathwaysthatcanlead tosecondaryhyperparathyroidism(32).VitaminDdeficiency andsecondaryhyperparathyroidismcancontributenotonlyto acceleratedbonelossandincreasingfragility,butalsotoneuromuscularimpairmentthatcanincreasetheriskoffalls(33, 34).ClinicaltrialsinvolvingolderindividualsathighriskforcalciumandvitaminDdeficiencyindicatethatsupplementationof bothcanreversesecondaryhyperparathyroidism,decreasebone resorption,increasebonemass,decreasefracturerates,andeven decreasethefrequencyoffalling(32).However,inalargerecent study,calciumandvitaminDsupplementationdidnotreduce fractureincidencesignificantly,perhapsbecausethispopulation waslessdeficientinvitaminD(35). PolymorphismsoftheVDRhavebeenstudiedextensively,but theresultshavebeenvariable.Thismaybeinpartbecausetheeffect ofagivenpolymorphisminthisreceptorisdependentonaninteractionwiththeenvironment,particularlywithcalcium(36).VDR polymorphismsarealsoassociatedwithdifferencesintheresponse totherapywithcalcitriol(37).Thereisalsoevidenceforaneffect

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

science in medicine
onfractureriskindependentofbonedensityandboneturnover, whichmightbeduetoanalterationinthefrequencyoffalls(38). Secondaryhyperparathyroidismpresentswhenthereisrelative insufficiencyofvitaminD,thatis,wherethelevelsofthecirculatingform25-hydroxyvitaminDfallbelow30ng/ml,suggestingthatthetargetforvitaminDsupplementationshouldbeat thislevelorhigher(39).TheseasonaldecreaseinvitaminDlevel andincreaseinPTHlevelduringthewintermonthsisassociated withanincreaseinfractures,independentoftheincreaseinrate offalls(40).Inaddition,increasedPTHlevelsareassociatedwith increasedmortalityinthefrailelderly,independentofbonemass andvitaminDstatus.Theprecisemechanismsunderlyingthis relationshiphavenotyetbeendetermined,buttheriskofcardiovasculardeathwasincreased(41).Polymorphismsofthecalciumsensingreceptor,whichregulatescalciumsecretionbysuppression ofPTHtranslationandPTHsecretion,havenotyetbeenassociatedwithanyalterationinbonephenotype(42,43). Receptor activator of NF-B, its ligand, and osteoprotegerin Theconceptthatstimulationofboneresorptionrequiresaninteractionbetweencellsoftheosteoblasticandosteoclasticlineages wasputforwardmanyyearsago,butitsmolecularmechanism wasonlyidentifiedrecently(43,44).ThreemembersoftheTNF andTNFreceptorsuperfamilyareinvolved(Figure2);osteoblasts produceRANKL,aligandforthereceptoractivatorofNF-B (RANK)onhematopoieticcells,whichactivatesthedifferentiation ofosteoclastsandmaintainstheirfunction.Osteoblastsalsoproduceandsecreteosteoprotegerin(OPG),adecoyreceptorthatcan blockRANKL/RANKinteractions.Stimulatorsofboneresorption havebeenfoundtoincreaseRANKLexpressioninosteoblasts, andsomealsodecreaseOPGexpression(43).Bonecellsappearto expressthemembrane-boundformofRANKL,andthus,osteoblastsmustphysicallyinteractwithosteoclastsprecursorsinorder toactivateRANK.SolubleRANKLcanbeproducedbyactivatedT lymphocytesandisasactiveasmembrane-boundRANKLinbindingtoRANK(45).StudiesintransgenicmiceshowedthatoverexpressionofOPGproducedosteopetrosis,whileOPG-knockout micehadaphenotypeofsevereosteoporosiswithahighincidence offractures(46).RecentlyamonoclonalantibodyagainstRANKL wasshowntoproduceprolongedinhibitionofboneresorptionin postmenopausalwomen(47).ItwasalsoshownthatRANKLlevels wereincreasedonthesurfaceofbonemarrowcellsfromearlypostmenopausalwomenwhoareestrogendeficient(48).However,it hasbeendifficulttodemonstratearoleforOPGdeficiencyinthe pathogenesisofosteoporosis,sinceOPGlevelsarenotconsistently altered.OPGlevelsincreasewithage,anditispossiblethatOPG productionrisesasahomeostaticresponsetolimittheboneloss thatoccurswithanincreaseinotherbone-resorbingfactors(21, 49).PolymorphismsintheOPG genehavebeenassociatedwith osteoporoticfracturesanddifferencesinBMD(50).OPGpolymorphismshavealsobeenlinkedtocoronaryarterydisease(51). ThisisconsistentwithrecentinformationlinkingtheRANKL/ RANK/OPGsystemtovascularcalcification(52). TheRANKL/RANKinteractioniscriticalforbothdifferentiation andmaintenanceofosteoclastactivityandhencerepresentsafinal commonpathwayforanypathogeneticfactorinosteoporosisthat actsbyincreasingboneresorption.Whileitisassumedthatcellsof thestromal/osteoblasticlineagearethemajorsourceofRANKL inphysiologicboneremodeling,othercellsmayactasasourceof

RANKLinpathologicstates;forexample,Tcellproductionmay playaroleinosteoporosisaswellasinflammatoryboneloss(53). Recently a second system that might affect the interaction betweenosteoblastsandosteoclastshasbeenidentified(54).This involvesthemembraneadapterDNAX-activatingprotein12and theFcreceptorcommonchain.Deletionofthesemolecules resultsinsevereosteopetrosisinmice.Themoleculesareinvolved insignalingthroughtheimmunoreceptortyrosine-basedactivation motif(ITAM).CooperationbetweenRANKLandITAMsignaling maybeessentialforosteoclastogenesis,forwhichnuclearfactorof activatedTcells(NFAT)isthemastertranscriptionfactor. Genes determining osteoblast differentiation and function Therecentdiscoveriesofsignaltransductionpathwaysandtranscriptionfactorscriticalforosteoblastdifferentiationandfunctionhaveopenedupnewapproachestounderstandingthepathogenesisofosteoporosis.Genedeletionstudieshaveshownthat absenceofrunt-relatedtranscriptionfactor2(Runx2)oradownstreamfactor,osterix,arecriticalforosteoblastdifferentiation(55, 56).Interestingly,overexpressionofRunx2leadstoadecreasein bonemass(57).Aroleforpolymorphismsofthesetranscription factorsinosteoporosishasnotyetbeenidentified. TherecentidentificationofthecriticalrolefortheWntsignalingpathwayinregulatingosteoblastfunctionisofparticularinterest,sinceithasbeenshowntoplayanimportantrole indeterminingbonemassandstrength(5861)(Figure3).LDL receptorrelated protein 5 (LRP5) interacts with the frizzled receptortotransducesignalingbyWntligands.Amutationof LRP5thatleadstoconstitutiveactivationcanresultinanincrease inbonedensity(58,59).ThephenotypeoffamilieswithLRP5 activatingmutationsvariesconsiderably,althoughallshowa strikingabsenceoffractures.Somehavenormalskeletalarchitecture,whileothersshowabnormalitiesduetoskeletalovergrowth (60).DeletionofLRP5resultsinasevereosteoporoticsyndrome

Figure 3
Interaction of the Wnt, BMP, and sclerostin pathways. Differentiation of osteoblasts during both development and remodeling is dependent on the activity of both the Wnt and BMP pathways. Wnt signaling requires the interaction of the LRP5 and frizzled receptors (Frz) and can be inhibited by Dickkopf (DKK; an inhibitor of LRP5) and secreted frizzledrelated protein (SFRP). Antagonists such as sclerostin can block both BMP and Wnt signaling. The mediator of the canonical Wnt pathway, -catenin, can synergize with BMP2 to enhance osteoblast differentiation and bone formation. Consistent with these interactions are the findings that high bone mass can result both from activating mutations of the Wnt pathway and deletion of SOST, the gene encoding sclerostin.
3321

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

science in medicine
associatedwithabnormaleyedevelopment(61).Polymorphisms ofLRP5havebeenassociatedwithdifferencesinbonemassand fractures(62,63).MutationsinLRP5havebeenidentifiedinafew patientswithidiopathicjuvenileosteoporosis(64). AnimalandinvitrostudiesindicatethattheWntsignalingpathwayiscriticalforosteoblastdifferentiationandfunction.Studies inmicesuggestthattheincreaseinbonemassinanimalswith activatingmutationsofLRP5isduetoanincreasedresponseto mechanicalloading(65).Thefactthatfluidshearstressactivates -cateninsignalingfurthersupportstheconceptthatWntsignalingiscriticalintheresponsetomechanicalloading(66).However, Wntsignalingisalsocriticalinbonedevelopmentandcanaffect peakbonemass(67).TheinhibitionofskeletalgrowthbyglucocorticoidsmaybemediatedbyeffectsonWntsignaling(68). TheprecisemechanismswherebyWntsignalingaltersosteoblastfunctionarenotfullyunderstood,butthereisevidencethat thecanonical-cateninpathwayisinvolvedandthatthereisan interactionwithbonemorphogeneticprotein2(BMP2)(69)(Figure3).Thereareanumberofinhibitorsthathavebeenshownto interactwithBMP2andwiththeWntsignalingpathway.Oneof these,sclerostin,theproductoftheSOSTgene,hasbeenshown toinhibitbothBMP2andWntsignaling(70,71).Inactivating mutationsofthisgenecanproducethehigh-bone-massdisorder VanBuchemdiseaseorsclerosteosis(72,73).Anotherpotential inhibitoryfactoristheproductionofsecretedfrizzled-related protein(SFRP)byosteoblasts(74). Local and systemic growth factors Remodelingimbalance,characterizedbyanimpairedboneformationresponsetoincreasedactivationofboneremodeling,isan essentialcomponentofthepathogenesisofosteoporosis(8,75). Thismaybedue,inpart,toanage-relateddecreaseinthecapacity ofosteoblaststoreplicateanddifferentiate.However,itseemslikely thatspecificdefectsintheproductionoractivityoflocalandsystemicgrowthfactorswillalsocontributetoimpairedboneformation.BMPsaswellasothermembersoftheTNFfamilyhavebeen implicated.IGF,whichisbothasystemicandlocalregulator,as wellasTGF-,canalsoalterboneformation.ThereissomeassociationbetweenBMDand theincidenceofosteoporoticfracturesand polymorphismsinthegenesencodingIGF-1andTGF-(7678), butthelargeststudytodate,inIcelandicandDanishcohorts,suggeststhatpolymorphismsoftheBMP2genearelinkedtolowBMD andfracturerisk(79).InhibitionoflocalIGF-1productionmaybe animportantcomponentofglucocorticoid-inducedosteoporosis aswellastheinhibitionofgrowthinchildhood(80). Cytokines, prostaglandins, NO, and leukotrienes TheconceptthatlocallyproducedcytokinessuchasIL-1and prostaglandinssuchasprostaglandinE2(PGE2)canaffectbone ismorethan30yearsold(81,82).Subsequently,manycytokines werefoundtoeitherstimulateorinhibitboneresorptionandformation(83).Prostaglandinshavebothstimulatoryandinhibitory actions;however,thepredominanteffectofPGE2,whichisthe majorprostaglandinproducedbybonecells,istostimulateboth resorptionandformation(84).Thepossibilitythatthesefactorsmightalsobeinvolvedinthepathogenesisofosteoporosis isbasedlargelyonanimalstudiesofbonelossafterovariectomy (26,27,53,85,86);however,thereisevidencethatpolymorphisms ofIL-1,IL-6,TNF-,andtheirreceptorscaninfluencebonemass inhumans(8789).
3322

Prostaglandins,particularlyPGE2,areproducedbybonecells largelythroughtheactionofinduciblecyclooxygenase2(COX2). COX2isinducedbymostofthefactorsthatstimulateboneresorptionandthusmayenhancetheresponsetotheseagents(84). TreatmentwithCOXinhibitorsbluntstheresponsetoimpact loadingandfluidshearstress,indicatingthatprostaglandinsplay animportantroleintheresponsemechanicalforces,andthismay beenhancedbyestrogen(90,91).Inepidemiologicstudies,small increasesinBMDanddecreasesinfractureriskhavebeenreported inindividualsusingNSAIDS(92,93). NOisproducedbybonecellsandisacofactorfortheanabolic responsetomechanicalloading(91,94).However,unlikeprostaglandins,NOmayinhibitboneresorption,perhapsbyincreasing OPGproduction(95).Thiseffectmayberesponsiblefortheincrease inBMDthathasbeendemonstratedinpatientstreatedwithisosorbidemononitrateandotheractivatorsoftheNOpathway(96). Leukotrienes,theproductsoflipoxygenase,canaffectboneby stimulatingresorptionandinhibitingformation(97).Recently, arachidonate15-lipoxygenase(encodedbyAlox15),wasidentified asanegativeregulatorofbonedensityinmice(98),andpolymorphismsinthehumangene,ALOX15,werefoundtobeassociated withdifferencesinpeakBMDinpostmenopausalwomen(99). Collagen abnormalities ApolymorphismofthefirstintronofthegenecodingforthetypeI collagen1chainandincreasedlevelsofhomocysteinecaninfluencefractureriskindependentofBMD(100102).Thismaybe duetodifferencesinhelixformationorcross-linkingofcollagen, challengingtheconceptthatmineralandmatrixcompositionare normalinosteoporosisandthatonlystructuralabnormalities accountforskeletalfragility. Leptin and neural pathways LeptindeficiencyorresistanceisassociatedwithhighBMDin micedespitethefactthatgonadalfunctionisdiminished(103). Thishasbeenattributedtoacentraleffectonadrenergicsignaling.Increased-adrenergicactivitycandecreasebonemass,but otherneuralpathwaysmaybeinvolved(104,105).Some,butnot all,epidemiologicstudiessuggestthat-adrenergicblockerscan decreasefractureriskandincreaseBMD(106,107).Anotherneuralpathwayhasrecentlybeenimplicatedbythefindingthatmice inwhichthecannabinoidtype1receptorisinactivated,aswellas micetreatedwithantagonistofthisreceptor,areprotectedfrom ovariectomy-inducedboneloss(108). Implication for the diagnosis and treatment of osteoporosis PriortotheintroductionofBMDmeasurements,thediagnosisof osteoporosiswasonlymadewhenfragilityfracturesoccurredinan appropriateclinicalsetting,largelyinpostmenopausalwomenand oldermen.TodayweuseBMDtodiagnoseosteoporosisandosteopeniabeforefracturesoccur,aswellastoconfirmthediagnosisin patientswithfragilityfractures(109).Ourgoalistoidentifythose individualsmostlikelytohavefracturesinthefutureandtotreat themappropriatelysothattheirfractureriskwillbereduced. WhilelowBMDisanimportantriskfactor,manyotherfactorshavebeenidentifiedinepidemiologicstudies,includingage, lowbodyweight,andsmoking.Manydrugsanddiseasesthatcan aggravatebonelossandfragilityarealsoimportantinpredicting futurefracturerisk,suchasglucocorticoidorthyroidhormone

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

science in medicine
excessandchronicinflammatorydisorders(109).Anotherkeypredictoristherateofboneremodeling.Increasedratesofosteoclasticboneresorption,measuredbythelevelofcollagenbreakdown products,aswellasincreasedboneformation,measuredbybonespecificalkalinephosphatase,osteocalcin,orprocollagenpeptide levels,areassociatedwithanincreaseinriskforbonelossandfragilityfractures.Techniquesthatcananalyzethemicroarchitecture ofbonearebeingdevelopedthatcouldfurtherimprovediagnosis (110).Inthefuture,itmaybepossibletoidentifygeneticprofiles thatpredictgreaterorlesserriskoffractures. Atpresentthetherapyofosteoporosisisdirectedatthemajor pathogeneticmechanismsoutlinedhere,withastrongemphasis onpreventionofbonelossandfractures(111).Whilethefirst majorpathogeneticmechanism,failuretoachieveoptimalpeak bonemass,isdeterminedlargelybyourgenes,nutritionandlifestylecanhaveaneffectduringgrowthanddevelopment.Moreover, adequatecalciumandvitaminDintakeandappropriatephysical activitymaynotonlyincreasepeakbonemassbutalsoslowbone lossandreducefractureriskthroughoutlife. Pharmacotherapyforosteoporosishasbeenfocusedmainlyon interventionsthatcouldreversethesecondpathogeneticmechanism,excessiveboneresorption(112).Inthepast,estrogenreplacementwasthemostwidelyusedtherapy.Ironically,thefirststudy thattrulyestablishedtheantifractureefficacyofestrogenhasalso resultedinitsbeinglargelyabandonedastheprimarytherapyfor osteoporosis(113).DatafromtheWomensHealthInitiativeindicatedthattheestrogen,particularlywhencombinedwithprogestins,increasedtheriskofbreastcancerandcardiovasculardisease, whichoutweighedthebenefitstotheskeleton.Thefactthatmuch lowerdosesofestrogencanpreventbonelossinpostmenopausal women(13)suggeststhatthecost/benefitratiomightbereversed withsuchtherapy,butappropriatelong-termstudieshavenotyet beencarriedout.Onealternativeistheuseofselectiveestrogen receptormodulators.Oneofthese,raloxifene,hasbeenshown todecreasetheriskofvertebralfractureanddoesnotappearto increasetheriskofbreastcancer(114). Thediscoverythatcalcitoninisaselectiveinhibitorofosteoclasticactivitymorethan40yearsagoledtotheconceptthatcalcitonindeficiencymightbethecauseofosteoporosisandthatits administrationmightbethecure.However,calcitonindeficiency wasnotfoundinosteoporoticpatients,andcalcitonintherapy hasbeenlesseffectivethanotherantiresorptiveagents,possibly becauseosteoclastscanescapecalcitonininhibition(112,114). Theclinicaltrialsshowingthatbisphosphonatescouldprovide effectiveandrelativelysafeantiresorptivetherapyhavehadan enormousimpactonthemanagementofosteoporosis.Thecur 1.Cooper,A.,andCooper,B.B.1822.Atreatiseondislocations,andonfracturesofthejoints.Churchill. London,UnitedKingdom.425pp. 2.Schapira,D.,andSchapira,C.1992.Osteoporosis: theevolutionofascientificterm.Osteoporos. Int. 2:164167. 3.Albright,F.,Bloomberg,E.,andSmith,P.H.1940. Postmenopausal osteoporosis. Trans. Assoc. Am. Physicians.55:298305. 4.Riggs,B.L.,etal.1982.Changesinbonemineral density of the proximal femur and spine with aging.Differencesbetweenthepostmenopausal andsenileosteoporosissyndromes.J. Clin. Invest. 70:716723. 5.Liu,Y.Z.,Liu,Y.J.,Recker,R.R.,andDeng,H.W. 2003.Molecularstudiesofidentificationofgenes forosteoporosis:the2002update.J. Endocrinol.

rentlyFDA-approvedbisphosphonatesalendronate,risedronate,andibandronatehavebeenshowntoreducetheincidence offractures,includingbothvertebralandnonvertebralfractures, particularlyhipfractures(112,114).Theseagentsbindtothebone surfaceandthenaretakenupbyosteoclasts,leadingtotheirinactivationandprogrammedcelldeath.Potentbisphosphonateshave nowbeeninuseformorethan10years,andthereissomeconcern thatprolongedusemaycauseexcessiveinhibitionofboneremodelingandslowtherepairoffracturesandofmicrodamage(115). Effectivetreatmentforthethirdpathogeneticmechanism, impairedboneformation,hasonlybecomeavailablerecently. Althoughitwasshownmorethan60yearsagothatintermittent administrationoflowdosesofPTHcouldincreasebonemassin animals,aneffectiveclinicalapproachwasonlydevelopedand approvedbytheFDAin2002.Dailysubcutaneousinjections ofthesyntheticN-terminalportionofPTH,teriparatide,can increasebonemassandstrengthbymarkedlyincreasingbone formation, while only modestly increasing bone resorption (116).Thistherapyisaseffectiveasbisphosphonatesinfracture reduction.ExtensivestudiesonthemechanismofactionofPTH areunderwayandshouldhelpusinidentifyingthecriticalpathwaysthatregulateboneformationandareabnormalinosteoporoticpatients.Anotherputativeanabolicagent,strontium ranelate,hasrecentlybeenapprovedfortreatmentofosteoporosisinseveralcountriesoutsidetheUS(117).Itsmechanismof actionisnotwellunderstood. Newantiresorptiveapproachesarebeinginvestigated,includingselectiveinhibitorsofosteoclastichydrogeniontransportand cathepsinKaswellasantagoniststothev3integrinsthatare necessaryforosteoclastadherenceandmotility(118).Newanabolicapproachesarelikelytoemergefromfurtherstudyofthe transcriptionalregulationofosteoblasts. Whileimprovementsindiagnosisandtherapyareimportant,it isequallytruethatwehavesufficienttoolstodaytoassessfracture riskandpreventortreatosteoporosistoreducethatrisk.Thus, thereare2challenges:(a)todevelopbetterdiagnostictoolsand treatment;and(b)toapplyourcurrentknowledgemorebroadly inthecommunity.Thelatterisfarfrombeingmet.Moreover,itis criticalthatwemeetboththesechallengesinthecomingdecades ifwearetodealeffectivelywiththeincreaseinosteoporosisand fracturesthatisprojectedasourpopulationages(111). Addresscorrespondenceto:LawrenceG.Raisz,UniversityofConnecticutHealthCenter,MusculoskeletalInstitute,263FarmingtonAvenue,Farmington,Connecticut06032,USA.Phone:(860) 679-2129;Fax:(860)679-1258;E-mail:raisz@nso.uchc.edu.
tion requires oestrogen receptor-alpha. Nature. 424:389. 11.Prestwood,K.M.,etal.1994.Theshort-termeffects ofconjugatedestrogenonboneturnoverinolder women.J. Clin. Endocrinol. Metab.79:366371. 12.Prestwood, K.M., Kenny, A.M., Unson, C., and Kulldorff,M.2000.Theeffectoflowdosemicronized17ss-estradiolonboneturnover,sexhormone levels,andsideeffectsinolderwomen:arandomized,doubleblind,placebo-controlledstudy.J. Clin. Endocrinol. Metab.85:44624469. 13.Prestwood, K.M., Kenny, A.M., Kleppinger, A., andKulldorff,M.2003.Ultralow-dosemicronized17beta-estradiolandbonedensityandbone metabolisminolderwomen:arandomizedcontrolledtrial.JAMA.290:10421048. 14.Modder,U.I.,etal.2004.Dose-responseofestrogen
3323

177:147196. 6.Baldock, P.A., and Eisman, J.A. 2004. Genetic determinantsofbonemass.Curr. Opin. Rheumatol. 16:450456. 7.Ioannidis, J.P., et al. 2004. Differential genetic effectsofESR1genepolymorphismsonosteoporosisoutcomes.JAMA.292:21052114. 8.Parfitt,A.M.,Villanueva,A.R.,Foldes,J.,andRao, D.S.1995.Relationsbetweenhistologicindices ofboneformation:implicationsforthepathogenesisofspinalosteoporosis.J. Bone Miner. Res. 10:466473. 9.Ebeling,P.R.,etal.1996.Boneturnovermarkers andbonedensityacrossthemenopausaltransition.J. Clin. Endocrinol. Metab.81:33663371. 10.Lee,K.,Jessop,H.,Suswillo,R.,Zaman,G.,and Lanyon, L. 2003. Endocrinology: bone adapta-

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

science in medicine
onboneversustheuterusinovariectomizedmice. Eur. J. Endocrinol.151:503510. 15.Falahati-Nini,A.,etal.2000.Relativecontributions oftestosteroneandestrogeninregulatingbone resorptionandformationinnormalelderlymen. J. Clin. Invest.106:15531560. 16.Khosla,S.,Melton,L.J.,3rd,Atkinson,E.J.,and OFallon,W.M.2001.Relationshipofserumsex steroidlevelstolongitudinalchangesinbonedensityinyoungversuselderlymen.J. Clin. Endocrinol. Metab.86:35553561. 17.VanPottelbergh,I.,Goemaere,S.,Zmierczak,H., andKaufman,J.M.2004.Perturbedsexsteroidstatusinmenwithidiopathicosteoporosisandtheir sons.J. Clin. Endocrinol. Metab.89:49494953. 18.Windahl,S.H.,etal.2001.Femaleestrogenreceptorbeta/micearepartiallyprotectedagainst age-relatedtrabecularboneloss.J. Bone Miner. Res. 16:13881398. 19.Sims,N.A.,etal.2002.Deletionofestrogenreceptorsrevealsaregulatoryroleforestrogenreceptors-betainboneremodelinginfemalesbutnotin males.Bone.30:1825. 20.Albagha,O.M.,etal.2005.Associationofoestrogenreceptoralphagenepolymorphismswithpostmenopausalboneloss,bonemass,andquantitativeultrasoundpropertiesofbone.J. Med. Genet. 42:240246. 21.Khosla,S.,etal.2004.Relationshipofestrogen receptorgenotypestobonemineraldensityandto ratesofbonelossinmen.J. Clin. Endocrinol. Metab. 89:18081816. 22.Bonnelye,E.,andAubin,J.E.2005.Estrogenreceptor-relatedreceptoralpha:amediatorofestrogenresponseinbone.J. Clin. Endocrinol. Metab. 90:31153121. 23.Laflamme,N.,etal.2005.Afrequentregulatory variantoftheestrogen-relatedreceptoralphagene associatedwithBMDinFrench-Canadianpremenopausalwomen.J. Bone Miner. Res.20:938944. 24.Goderie-Plomp,H.W.,etal.2004.Endogenoussex hormones,sexhormone-bindingglobulin,andthe riskofincidentvertebralfracturesinelderlymen andwomen:theRotterdamStudy.J. Clin. Endocrinol. Metab.89:32613269. 25.VanPottelbergh,I.,Goemaere,S.,andKaufman, J.M.2003.Bioavailableestradiolandanaromatase genepolymorphismaredeterminantsofbonemineraldensitychangesinmenover70yearsofage. J. Clin. Endocrinol. Metab.88:30753081. 26.Kimble,R.B.,etal.1995.Simultaneousblockof interleukin-1andtumornecrosisfactorisrequired tocompletelypreventbonelossintheearlypostovariectomyperiod.Endocrinology.136:30543061. 27.Lorenzo,J.A.,etal.1998.MicelackingthetypeI interleukin-1receptordonotlosebonemassafter ovariectomy.Endocrinology.139:30223025. 28.Gao, Y., et al. 2004. Estrogen prevents bone loss throughtransforminggrowthfactorbetasignalingin Tcells.Proc. Natl. Acad. Sci. U. S. A.101:1661816623. 29.Hughes,D.E.,etal.1996.EstrogenpromotesapoptosisofmurineosteoclastsmediatedbyTGF-beta. Nat. Med.2:11321136. 30.Lean,J.M.,etal.2003.Acrucialroleforthiolantioxidantsinestrogen-deficiencyboneloss.J. Clin. Invest.112:915923.doi:10.1172/JCI200318859. 31.Lean,J.M.,Jagger,C.J.,Kirstein,B.,Fuller,K.,and Chambers,T.J.2005.Hydrogenperoxideisessentialforestrogen-deficiencybonelossandosteoclast formation.Endocrinology.146:728735. 32.Lips,P.2001.VitaminDdeficiencyandsecondary hyperparathyroidismintheelderly:consequences forbonelossandfracturesandtherapeuticimplications.Endocr. Rev.22:477501. 33.Bischoff-Ferrari,H.A.,etal.2004.Effectofvitamin Donfalls:ameta-analysis.JAMA.291:19992006. 34.Sambrook,P.N.,etal.2004.Serumparathyroid hormonepredictstimetofallindependentofvitaminDstatusinafrailelderlypopulation.J. Clin. Endocrinol. Metab.89:15721576. 35.Grant,A.M.,etal.2005.OralvitaminD3andcalciumforsecondarypreventionoflow-traumafracturesinelderlypeople(RandomisedEvaluationof CalciumOrvitaminD,RECORD):arandomised placebo-controlledtrial.Lancet.365:16211628. 36.Ferrari,S.L.,Rizzoli,R.,Slosman,D.O.,andBonjour, J.P.1998.Dodietarycalciumandageexplainthe controversysurroundingtherelationshipbetween bonemineraldensityandvitaminDreceptorgene polymorphisms?J. Bone Miner. Res.13:363370. 37.Morrison,N.A.,etal.2005.VitaminDreceptor genotypesinfluencethesuccessofcalcitrioltherapyforrecurrentvertebralfractureinosteoporosis. Pharmacogenet. Genomics.15:127135. 38.Garnero,P.,Munoz,F.,Borel,O.,Sornay-Rendu,E., andDelmas,P.D.2005.VitaminDreceptorgene polymorphismsareassociatedwiththeriskoffracturesinpostmenopausalwomen,independently ofbonemineraldensity.TheOFELYstudy.J. Clin. Endocrinol. Metab.90:48294835. 39.Lips,P.2004.Whichcirculatinglevelof25-hydroxyvitaminDisappropriate?J. Steroid Biochem. Mol. Biol.8990:611614. 40.Pasco, J.A., et al. 2004. Seasonal periodicity of serumvitaminDandparathyroidhormone,bone resorption,andfractures:theGeelongOsteoporosisStudy.J. Bone Miner. Res.19:752758. 41.Sambrook,P.N.,etal.2004.Serumparathyroidhormoneisassociatedwithincreasedmortalityindependentof25-hydroxyvitamindstatus,bonemass, andrenalfunctioninthefrailandveryold:acohort study.J. Clin. Endocrinol. Metab.89:54775481. 42.Bollerslev,J.,etal.2004.Calcium-sensingreceptor genepolymorphismA986Sdoesnotpredictserum calciumlevel,bonemineraldensity,calcanealultrasoundindices,orfracturerateinalargecohortof elderlywomen.Calcif. Tissue Int.74:1217. 43.Suda,T.,etal.1999.Modulationofosteoclastdifferentiationandfunctionbythenewmembersof thetumornecrosisfactorreceptorandligandfamilies.Endocr. Rev.20:345357. 44.Rodan,G.A.,andMartin,T.J.1982.Roleofosteoblastsinhormonalcontrolofboneresorptiona hypothesis[letter].Calcif. Tissue Int.34:311. 45.Kanamaru,F.,etal.2004.ExpressionofmembraneboundandsolublereceptoractivatorofNF-kappaBligand(RANKL)inhumanTcells.Immunol. Lett.94:239246. 46.Bucay,N.,etal.1998.Osteoprotegerin-deficient micedevelopearlyonsetosteoporosisandarterial calcification.Genes Dev.12:12601268. 47.Bekker,P.J.,etal.2004.Asingle-doseplacebo-controlledstudyofAMG162,afullyhumanmonoclonalantibodytoRANKL,inpostmenopausal women.J. Bone Miner. Res.19:10591066. 48.Eghbali-Fatourechi,G.,etal.2003.RoleofRANK ligand in mediating increased bone resorption in early postmenopausal women. J. Clin. Invest. 111:12211230.doi:10.1172/JCI200317215. 49.Khosla,S.,etal.2002.Correlatesofosteoprotegerinlevelsinwomenandmen.Osteoporos. Int. 13:394399. 50.Langdahl, B.L., Carstens, M., Stenkjaer, L., and Eriksen,E.F.2002.Polymorphismsintheosteoprotegeringeneareassociatedwithosteoporotic fractures.J. Bone Miner. Res.17:12451255. 51.Soufi,M.,etal.2004.Osteoprotegeringenepoly morphismsinmenwithcoronaryarterydisease. J. Clin. Endocrinol. Metab.89:37643768. 52.Collin-Osdoby,P.2004.RegulationofvascularcalcificationbyosteoclastregulatoryfactorsRANKL andosteoprotegerin.Circ. Res.95:10461057. 53.Weitzmann, M.N., et al. 2001. T cell activation induceshumanosteoclastformationviareceptor activatorofnuclearfactorkappaBligand-dependentand-independentmechanisms.J. Bone Miner. Res.16:328337. 54.Takayanagi, H. 2005. Mechanistic insight into osteoclast differentiation in osteoimmunology. J. Mol. Med.83:170179. 55.Ducy, P., Zhang, R., Geoffroy, V., Ridall, A.L., andKarsenty,G.1997.Osf2/Cbfa1:atranscriptionalactivatorofosteoblastdifferentiation.Cell. 89:747754. 56.Nakashima,K.,etal.2002.Thenovelzincfingercontainingtranscriptionfactorosterixisrequired forosteoblastdifferentiationandboneformation. Cell.108:1729. 57.Geoffroy,V.,Kneissel,M.,Fournier,B.,Boyde,A., andMatthias,P.2002.Highboneresorptionin adultagingtransgenicmiceoverexpressingcbfa1/ runx2incellsoftheosteoblasticlineage.Mol. Cell. Biol.22:62226233. 58.Little,R.D.,etal.2002.AmutationintheLDL receptor-relatedprotein5generesultsintheautosomaldominanthigh-bone-masstrait.Am. J. Hum. Genet.70:1119. 59.Boyden,L.M.,etal.2002.Highbonedensitydue toamutationinLDL-receptor-relatedprotein5. N. Engl. J. Med.346:15131521. 60.VanWesenbeeck,L.,etal.2003.Sixnovelmissense mutationsintheLDLreceptor-relatedprotein5 (LRP5)geneindifferentconditionswithanincreased bonedensity.Am. J. Hum. Genet.72:763771. 61.Gong,Y.,etal.2001.LDLreceptor-relatedprotein 5(LRP5)affectsboneaccrualandeyedevelopment. Cell.107:513523. 62.Koay,M.A.,etal.2004.InfluenceofLRP5polymorphismsonnormalvariationinBMD.J. Bone Miner. Res.19:16191627. 63.Bollerslev,J.,etal.2005.LRP5genepolymorphisms predictbonemassandincidentfracturesinelderly Australianwomen.Bone.36:599606. 64.Hartikka,H.,etal.2005.Heterozygousmutations intheLDLreceptor-relatedprotein5(LRP5)gene areassociatedwithprimaryosteoporosisinchildren.J. Bone Miner. Res.20:783789. 65.Akhter,M.P.,etal.2004.BonebiomechanicalpropertiesinLRP5mutantmice.Bone.35:162169. 66.Norvell, S.M., Alvarez, M., Bidwell, J.P., and Pavalko, F.M. 2004. Fluid shear stress induces beta-cateninsignalinginosteoblasts.Calcif. Tissue Int.75:396404. 67.Kronenberg,H.,andKobayashi,T.2004.Transcriptionalregulationindevelopmentofbone.Endocrinology.146:10121017. 68.Ohnaka,K.,Tanabe,M.,Kawate,H.,Nawata,H., and Takayanagi, R. 2005. Glucocorticoid suppresses the canonical Wnt signal in cultured humanosteoblasts.Biochem. Biophys. Res. Commun. 329:177181. 69.Mbalaviele,G.,etal.2005.beta-CateninandBMP-2 synergizetopromoteosteoblastdifferentiationand newboneformation.J. Cell. Biochem.94:403418. 70.Li,X.,etal.2005.SclerostinbindstoLRP5/6and antagonizescanonicalWntsignaling.J. Biol. Chem. 280:1988319887. 71.Winkler,D.G.,etal.2004.Sclerostininhibitionof Wnt-3a-inducedC3H10T1/2celldifferentiation isindirectandmediatedbyBMPproteins.J. Biol. Chem.280:24982502. 72.Loots,G.G.,etal.2005.GenomicdeletionofalongrangeboneenhancermisregulatessclerostininVan Buchemdisease.Genome Res.15:928935. 73.Balemans,W.,etal.2001.Increasedbonedensityinsclerosteosisisduetothedeficiencyofa novelsecretedprotein(SOST).Hum. Mol. Genet. 10:537543. 74.Bodine, P.V., et al. 2005. The Wnt antagonist secretedfrizzled-relatedprotein-1controlsosteoblast and osteocyte apoptosis. J. Cell. Biochem. doi:10.1002/jcb.20599. 75.Eriksen,E.F.,etal.1990.CancellousboneremodelingintypeI(postmenopausal)osteoporosis:quan-

3324

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

science in medicine
titativeassessmentofratesofformation,resorption,andbonelossattissueandcellularlevels. J. Bone Miner. Res.5:311319. 76.Lau, E.M., et al. 2004. Osteoporosis and transforminggrowthfactor-beta-1genepolymorphism inChinesemenandwomen.J. Bone Miner. Metab. 22:148152. 77.Langdahl,B.L.,Carstens,M.,Stenkjaer,L.,andEriksen,E.F.2003.Polymorphismsinthetransforming growthfactorbeta1geneandosteoporosis.Bone. 32:297310. 78.Kim,J.G.,Roh,K.R.,andLee,J.Y.2002.Therelationshipamongseruminsulin-likegrowthfactor-I,insulin-likegrowthfactor-Igenepolymorphism,and bonemineraldensityinpostmenopausalwomenin Korea.Am. J. Obstet. Gynecol.186:345350. 79.Styrkarsdottir,U.,etal.2003.Linkageofosteoporosistochromosome20p12andassociationto BMP2.PLoS Biol.1:E69. 80.Mehls,O.,Himmele,R.,Homme,M.,Kiepe,D.,and Klaus,G.2001.Theinteractionofglucocorticoids withthegrowthhormone-insulin-likegrowthfactoraxisanditseffectsongrowthplatechondrocytesandbonecells.J. Pediatr. Endocrinol. Metab. 14(Suppl.6):14751482. 81.Klein,D.C.,andRaisz,L.G.1970.Prostaglandins: stimulationofboneresorptionintissueculture. Endocrinology.86:14361440. 82.Raisz,L.G.,etal.1975.Effectofosteoclastactivatingfactorfromhumanleukocytesonbonemetabolism.J. Clin. Invest.56:408413. 83.Horwitz, M.C., and Lorenzo, J.A. 2002. Local regulatorsofbone:IL-1,TNF,lymphotoxin,interferon-,IL-8,IL-10,IL-4,theLIF/IL-6family,and additionalcytokines.InPrinciples of bone biology.J.P. Bilezikian,L.G.Raisz,andG.A.Rodan,editors.AcademicPress.SanDiego,California,USA.961977. 84.Pilbeam,C.C.,Harrison,J.R.,andRaisz,L.G.2002. Prostaglandinsandbonemetabolism.InPrinciples of bone biology.J.P.Bilezikian,L.G.Raisz,andG.A. Rodan,editors.AcademicPress.SanDiego,California,USA.979994. 85.Kawaguchi,H.,etal.1995.Ovariectomyenhances andestrogenreplacementinhibitstheactivityof bonemarrowfactorsthatstimulateprostaglandin productioninculturedmousecalvariae.J. Clin. Invest.96:539548. 86.Ammann,P.,etal.1997.Transgenicmiceexpressingsolubletumornecrosisfactor-receptorareprotectedagainstbonelosscausedbyestrogendeficiency.J. Clin. Invest.99:16991703. 87.Tasker,P.N.,Albagha,O.M.,Masson,C.B.,Reid, D.M.,andRalston,S.H.2004.Associationbetween TNFRSF1B polymorphisms and bone mineral density, bone loss and fracture. Osteoporos. Int. 15:903908. 88.Chung,H.W.,etal.2003.Associationofinterleukin-6 promotervariantwithbonemineraldensityinpremenopausalwomen.J. Hum. Genet.48:243248. 89.Ferrari,S.L.,etal.2004.Interactionsofinterleukin-6promoterpolymorphismswithdietaryand lifestylefactorsandtheirassociationwithbone massinmenandwomenfromtheFramingham OsteoporosisStudy.J. Bone Miner. Res.19:552559. 90.Forwood,M.R.1996.Induciblecyclo-oxygenase (COX-2)mediatestheinductionofboneformationbymechanicalloadinginvivo.J. Bone Miner. Res.11:16881693. 91.Bakker,A.D.,etal.2004.Additiveeffectsofestrogen and mechanical stress on nitric oxide and prostaglandinE(2)productionbybonecellsfrom osteoporoticdonors.Osteoporos. Int.16:983989. 92.Raisz, L.G. 2001. Potential impact of selective cyclooxygenase-2 inhibitors on bone metabolisminhealthanddisease.Am. J. Med.110(Suppl. 3A):43S45S. 93.Carbone,L.D.,etal.2003.Associationbetween bonemineraldensityandtheuseofnonsteroidalanti-inflammatorydrugsandaspirin:impact of cyclooxygenase selectivity. J. Bone Miner. Res. 18:17951802. 94.Chow,J.W.,Fox,S.W.,Lean,J.M.,andChambers, T.J.1998.Roleofnitricoxideandprostaglandins inmechanicallyinducedboneformation.J. Bone Miner. Res.13:10391044. 95.Wang, F.S., et al. 2004. Nitric oxide donor increasesosteoprotegerinproductionandosteoclastogenesisinhibitoryactivityinbonemarrow stromalcellsfromovariectomizedrats.Endocrinology.145:21482156. 96.Jamal, S.A., Cummings, S.R., and Hawker, G.A. 2004. Isosorbide mononitrate increases bone formationanddecreasesboneresorptioninpostmenopausalwomen:arandomizedtrial.J. Bone Miner. Res.19:15121517. 97.Traianedes,K.,Dallas,M.R.,Garrett,I.R.,Mundy, G.R., and Bonewald, L.F. 1998. 5-Lipoxygenase metabolitesinhibitboneformationinvitro.Endocrinology.139:31783184. 98.Klein,R.F.,etal.2004.Regulationofbonemass inmicebythelipoxygenasegeneAlox15.Science. 303:229232. 99.Urano,T.,etal.2005.AssociationofasinglenucleotidepolymorphisminthelipoxygenaseALOX15 5-flankingregion(-5229G/A)withbonemineral density.J. Bone Miner. Metab.23:226230. 1 00.Mann,V.,andRalston,S.H.2003.Meta-analysisof COL1A1Sp1polymorphisminrelationtobone mineraldensityandosteoporoticfracture.Bone. 32:711717. 1 01.McLean,R.R.,etal.2004.Homocysteineasapredictive factor for hip fracture in older persons. N. Engl. J. Med.350:20422049. 1 02.vanMeurs,J.B.,etal.2004.Homocysteinelevels andtheriskofosteoporoticfracture.N. Engl. J. Med. 350:20332041. 1 03.Ducy,P.,etal.2000.Leptininhibitsboneformation throughahypothalamicrelay:acentralcontrolof bonemass.Cell.100:197207. 1 04.Elefteriou,F.,etal.2005.Leptinregulationofbone resorptionbythesympatheticnervoussystemand CART.Nature.434:514520. 1 05.Takeda,S.2005.Centralcontrolofboneremodeling.Biochem. Biophys. Res. Commun.328:697699. 1 06.Pasco,J.A.,etal.2004.Beta-adrenergicblockers reduce the risk of fracture partly by increasing bonemineraldensity:GeelongOsteoporosisStudy. J. Bone Miner. Res.19:1924. 1 07.Reid,I.R.,etal.2005.beta-Blockeruse,BMD,and fracturesinthestudyofosteoporoticfractures. J. Bone Miner. Res.20:613618. 1 08.Idris,A.I.,etal.2005.Regulationofbonemass, bonelossandosteoclastactivitybycannabinoid receptors.Nat. Med.11:774799. 1 09.Raisz,L.G.2005.Clinicalpractice.Screeningfor osteoporosis.N. Engl. J. Med.353:164171. 1 10.Benito,M.,etal.2003.Deteriorationoftrabecular architectureinhypogonadalmen.J. Clin. Endocrinol. Metab.88:14971502. 1 11.OfficeoftheSurgeonGeneral.2004.Bonehealth andosteoporosis:areportoftheSurgeonGeneral. USDepartmentofHealthandHumanServices. Rockville,Maryland,USA.404pp. 1 12.Rosen,C.J.2005.Clinicalpractice.Postmenopausal osteoporosis.N. Engl. J. Med.353:595603. 1 13.Anderson,G.L.,etal.2004.Effectsofconjugated equineestrogeninpostmenopausalwomenwith hysterectomy:theWomensHealthInitiativerandomizedcontrolledtrial.JAMA.291:17011712. 1 14.Cranney,A.,etal.2002.Meta-analysesoftherapies forpostmenopausalosteoporosis.IX:summary ofmeta-analysesoftherapiesforpostmenopausal osteoporosis.Endocr. Rev.23:570578. 1 15.Ott,S.M.2005.Long-termsafetyofbisphosphonates.J. Clin. Endocrinol. Metab.90:18971899. 1 16.Neer,R.M.,etal.2001.Effectofparathyroidhormone(1-34)onfracturesandbonemineraldensityinpostmenopausalwomenwithosteoporosis. N. Engl. J. Med.344:14341441. 1 17.Meunier,P.J.,etal.2004.Theeffectsofstrontium ranelateontheriskofvertebralfractureinwomen withpostmenopausalosteoporosis.N. Engl. J. Med. 350:459468. 1 R 18. odan, G.A., and Martin, T.J. 2000. Therapeutic approachestobonediseases.Science.289:15081514.

The Journal of Clinical Investigation http://www.jci.org Volume115 Number12 December2005

3325