Differentiation of mesenchymal stem cells to osteoblasts and chondrocytes: a focus on adenosine...

Differentiation of mesenchymal stem

cells to osteoblasts and chondrocytes:

a focus on adenosine receptors

Shannon H. Carroll1 and Katya Ravid2,3,4,*

Skeletogenesis, either during development, post-injury or for maintenance, is acarefully coordinated process reliant on the appropriate differentiation ofmesenchymal stem cells. Some well described, as well as a new regulator ofthis process (adenosine receptors), are alike in that they signal via cyclic-AMP(cAMP). This review highlights the known contribution of cAMP signalling tomesenchymal stem cell differentiation to osteoblasts and to chondrocytes.Focus has been given to how these regulators influence the commitment ofthe osteochondroprogenitor to these separate lineages.

The development of the skeletal system, as well asits repair, are dependent on the differentiation ofboth chondrocytes and osteoblasts from theircommon progenitor, the mesenchymal stem cell(MSC). There are two processes from whichbone is formed. During intramembranousosteogenesis, bone is formed directly from MSCdifferentiation to osteoblasts. This process givesrise to flat bones, such as those of the skull andclavicle. In contrast, endochondral osteogensisrequires MSC differentiation to chondrocytesand the formation of a cartilage template, whichis followed by ossification by osteoblasts. Thisprocess is responsible for the formation of thelong bones of the skeleton, and remains active inthe growth plates of growing bones (reviewed inRefs 1, 2). Bone fracture repair recapitulates theevents of skeletogenesis and is, therefore, used

as an experimental model of bone formation(Ref. 3). These processes are regulated byparacrine actions between osteoblasts andchondrocytes, and thus, proper skeletogenesisrequires precise control over the differentialdifferentiation of the MSC to these lineages.Failure of chondrocyte or osteoblast precursorsto proliferate and differentiate leads to varioustypes of skeletal dysplasias, depending on thepathway involved (reviewed in Ref. 4).

Skeletogenesis: an overviewThe osteochondroprogenitorThe majority of what is known about MSCdifferentiation along the skeletal lineage comesfrom embryology. Endochondral andintramembranous skeletogenesis begins with theproliferation and migration of mesenchymal

1Department of Biochemistry, Boston University School of Medicine, Boston, MA, USA2Department of Medicine, Boston University School of Medicine, Boston, MA, USA3Whitaker Cardiovascular Institute, Boston University School of Medicine, Boston, MA, USA4The Evans Center for Interdisciplinary Biomedical Research, Boston University School of Medicine,Boston, MA, USA

*Corresponding author: Katya Ravid, Boston University School of Medicine, 700 Albany St RoomW602, Boston, MA 02191, USA. E-mail: [email protected]

expert reviewshttp://www.expertreviews.org/ in molecular medicine

1Accession information: doi:10.1017/erm.2013.2; Vol. 15; e1; February 2013

© Cambridge University Press 2013

Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

mailto:[email protected]

cells to form condensations. These condensationsare characterised as tightly packed cells thatexpress specific condensation markers, and willdetermine the position, shape and size of theskeletal elements (Ref. 5). The transforminggrowth factor-β and Hox family of molecules,derived from the epithelium, have beenimplicated in directing the formation ofcondensations. These molecules, along with thecell–cell and cell–matrix interactions that resultfrom the compact nature of these cells, arethought to trigger MSC differentiation (Refs 5, 6).As the osteochondroprogenitor is multipotent,

lineage fate decisions must be made andmaintained. The transcription factor runt-relatedtranscription factor 2 (Runx2) is believed todetermine osteoblast versus chondrocytedifferentiation. Various studies have shown thatupregulation of Runx2 induces osteoblasticdifferentiation, whereas its persistent expressionin chondrocytes causes premature maturationand mineralisation (Ref. 7). Runx2 is expressed inprechondrogenic and preosteogeniccondensations, but during early embryogenesis,outside signals regulate its expression (Ref. 5).Homeobox protein A2 (Hoxa-2) and Wnt(wingless-type mouse mammary tumor virus(MMTV) integration site) signalling have bothbeen implicated in this process. Through specificinactivation in the developing limbs of mice, β-catenin, a downstream effector protein of Wntsignalling, was found to be essential to bonedevelopment as these embryos had impairedosteoblast differentiation (Ref. 8). Also, it wasfound that if β-catenin is activated, it upregulatesRunx2 expression and MSCs differentiate toosteoblasts, rather than chondrocytes (Refs 8, 9,10). These results implicate β-catenin in thedetermination of osteo- versus chondrogenesis.

ChondrogenesisChondrocytes are the first skeletal cells to arise,and their differentiation depends on theexpression of the transcription factor sry (sexdetermining region Y)-box 9 (Sox9). Theimportance of Sox9 was discovered when agenetic mutation in humans was found to be thecause of campomelic dysplasia, characterised bysevere cartilage abnormalities (Refs 11, 12).Genetic deletion of Sox9 in mice blockschondrocyte differentiation at the point ofmesenchymal condensation, indicating that it isnecessary for the induction of chondrocyte

differentiation (Ref. 13). In the prechondrogenicmesenchyme, Runx2 is expressed along withSox9 (Ref. 14). However, studies show Sox9 tobe dominant over Runx2 (Ref. 15) by promotingRunx2 protein degradation as well as inhibitingits transcriptional activity (Ref. 16). In addition,Nkx3.2 inhibits Runx2 transcription, enforcingthe differentiation to chondrocytes (Ref. 17).Sox9, along with co-activators Sox5 and Sox6,binds and activates promoters of chondrocyte-specific genes. These include collagen 2α1(Col2α1) and aggrecan (Ref. 5).

Also, important to chondrogenesis ischondrocyte maturation. This involves thetransition of chondrocytes from proliferating, tononproliferating and hypertrophic, andeventually apoptotic (Ref. 2). The expression ofSox9, along with the activation of cAMPresponse element binding protein (CREB) and c-Fos, maintains the chondrocytes in aproliferative state (Ref. 2). In order for thechondrocytes to exit the cell cycle and becomehypertrophic, Runx2 must be upregulated andSox9 suppressed. The exit from the cell cycleand apoptosis is necessary for the eventualinvasion by osteoblasts and ossification of thebone matrix (Ref. 2).

OsteoblastogenesisThe transcription factor Runx2 is absolutelynecessary for osteoblast differentiation. Runx2knockout (KO) mice show a total absence ofdifferentiated osteoblasts and, therefore, lackany bone (Refs 18, 19). As Sox9 suppressesRunx2 activity, osteoblast differentiation occurswhen Runx2 is upregulated and stabilised. Twistproteins, transcription factors with importantroles in embryogenesis, also negatively regulateRunx2, by blocking its DNA binding domainand inhibiting its ability to upregulateosteoblast-specific genes (Ref. 20). Therefore, theinhibition of Twist proteins, specifically Twist-1and/or Twist-2, is thought to initiateosteoblastogenesis (Ref. 20). Members of thedistal-less (Dlx) family of homeobox proteins(Dlx5 and Dlx6) are important for endochondralossification in the developing appendicular andaxial skeleton (Ref. 21). Dlx5 has been shown toact directly upstream of Runx2, by binding itsgene promoter, (Ref. 22) and Dlx5/Dlx6 doubleKO mice exhibit delayed ossification (Ref. 23).Other homeobox transcription factors are mshhomeobox 1 and 2 (Msx1/Msx2), although




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

whether their regulation of osteoblastdifferentiation is positive or negative is stillcontroversial (Ref. 24). Msx1/Msx2 may benecessary for the expression of Runx2, as theMsx1/Msx2 double KO fails to express Runx2(Ref. 25), although more direct evidence isneeded (Fig. 1 summarises major transcriptionalregulators of osteoblastogenesis). Wnt signallinghas also been implicated in triggering osteoblastdifferentiation (reviewed in Refs 26, 27). There isa β-catenin responsive TCF/Lef binding site onthe Runx2 gene promoter, and therefore, mayupregulate Runx2 expression (Ref. 28). Onceupregulated, Runx2 triggers osteoblastdifferentiation and bone development bybinding promoters of osteoblast-specific genes,including Osterix, alkaline phosphatase andOsteocalcin. Once its expression is stabilised,Runx2 inhibits Sox9 transcriptional activity(Ref. 29), further pushing the MSC to the

osteoblast lineage. In addition to regulatingosteoblast differentiation, Wnt signalling alsoregulates osteoblast number and function. Post-natal deletion of β-catenin in Osterix-expressingcells causes osteopenia, however the number ofosteoblasts is increased (Ref. 30). Therefore acontinued investigation of the role of Wntsignalling in osteoblast function, particularlyosteoclast regulation, is needed.

In summary, it appears that the differentiation tochondrocytes may occur by default, when Runx2and osteoblast differentiation are suppressed. It isalso possible that signals that upregulate Sox9 areresponsible for the suppression of the osteoblastdifferentiation programme. Figure 2 illustratesthese two paths of differentiation.

Cyclic-AMP as a regulator ofosteoblastogenesis and chondrogenesisAlthough there are master regulators that directMSC differentiation, various other moleculeshave been demonstrated to modulate theprocess. Here, we will focus on those that signalthrough cyclic-AMP (cAMP). cAMP is aubiquitous second messenger that is synthesisedfrom ATP by adenylyl cyclases. cAMP levels areregulated by different stimuli, one major effectorbeing G-protein-coupled receptors. Thesereceptors are classified either as stimulatory(Gαs) or inhibitory (Gαi) of adenylyl cyclase.Changing levels of cAMP is translated to the cellthrough cAMP’s action on cyclic nucleotide-gated ion channels, on exchange proteins knownas Epacs, and on protein kinase A (PKA)(reviewed in Ref. 31). PKA activation furtherperpetuates the signal by phosphorylatingdifferent target proteins. Ultimately, cellulartranscription can be modified through thecAMP-dependent transcription factors CREB,cAMP response element modulator (CREM) andATF-1 (reviewed in Refs 31, 32, 33). cAMP isdegraded by phosphodiesterases, which removea phosphodiester bond and produce AMP. Theseenzymes play a major role, not only interminating the signal, but also in regulating theamplitude and duration of the signal (Ref. 34).

The role of Gαs signalling in osteoblastdifferentiation was spurred by the finding thatindividuals with mutations in the Gαs genehave a bone phenotype. Inactivating mutationsin the Gαs gene (GNAS) cause Albrighthereditary osteodystrophy (Ref. 35), whereasactivating mutations cause fibrous dysplasia of

Collagen2a1aggrecan

Chondrocyte

MSC

Osteoblast

Osteochondro-progenitor

Sox9

Nkx3.2

Dlx5

Msx1/Msx2

Sox9

Runx2 Runx2Osterix

Collagen 1osteoclcin

Transcriptional regulators of MSCdifferentiation to skeletal lineagesExpert Reviews in Molecular Medicine © 2013 Cambridge University Press

Figure 1. Transcriptional regulators ofmesenchymal stem cell (MSC) differentiation toskeletal lineages. Osteochondroprogenitors arisefrom MSCs and express both Runx2 and Sox9.Multiple transcriptional regulators (including Dlx5,Nkx3.2 and Msx1/Msx2) have been found to modifythe expression of Runx2 and Sox9 and, therefore,drive differentiation to chondrocytes or osteoblasts.Supportive references are included in the text.




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

bone (Ref. 36). To explore its role in bonedevelopment, Hsaio et al. created a mouseexpressing an engineered Gαs receptor inosteoblasts and found it to have drasticallyincreased bone mass (Ref. 37). Conversely,complete KO of Gαs receptors in osteoblastsreduced the number of osteoblasts and impairedbone formation (Refs 38, 39). Althoughinterpretation of these types of experiments islimited, they support a potential role for cAMPsignalling in MSC differentiation. In accordance,parathyroid hormone (PTH), a hormone criticalfor bone development and homoeostasis, signalsthrough a G-protein-coupled receptor (reviewedin Ref. 40), reinforcing the importance of cAMPsignalling in skeletogenesis. The PTH receptorbinds PTH, as well as PTH-related protein(PTHrP), and is coupled to Gαs as well as Gαq(Ref. 41). There is a long history of PTH’s effecton osteoblast differentiation. Studies in PTHreceptor KO mice show that the signalling is notrequired for osteoblast differentiation (Ref. 42),however PTH has been demonstrated toenhance osteoblast differentiation (reviewed inRef. 43). Recently, it was found that PTHinteracts with canonical Wnt signalling andFGF-2 signalling and that this interactionenhances osteoblast differentiation (reviewed inRef. 44). Also, PTH signalling was found tointeract with bone morphogenetic protein

signalling, which additionally enhancesosteoblast differentiation (Refs 45, 46, 47, 48).

Thedirect effect of cAMPon chondrogenesis hasnot been fully investigated. In one study, treatmentof rabbit chondrocyte cultures with cAMPanalogues was found to suppress terminaldifferentiation and hypertrophy of chondrocytes.Similarly to osteoblasts, PTH is not required forchondrocyte differentiation (Ref. 42). However,PTH treatment of C3H10T1/2 cells can enhancethe early stages of chondrocyte differentiationwhile suppressing chondrocyte maturation(Ref. 49). In addition, limb explants from PTHrPKO mice displayed accelerated chondrocytematuration (Ref. 50).

cAMP effect on transcriptional regulators ofosteoblastogenesisStudies have shown that increasing intracellularlevels of cAMP in cell lines or in primary mouseand human MSCs, through cAMP analogues orforskolin (a direct adenylyl cyclase activator),has a positive effect on Runx2 expression(Refs 51, 52, 53, 54). In addition, activation ofreceptors that signal through cAMP increasesRunx2 expression in MSC cell lines and inprimary rodent MSCs. These include the PTHreceptor (Ref. 55) and the A2B adenosinereceptor (Refs 54, 56). This increase inexpression may be owing to cAMP activation of

Osteoblast-chodrocyte precursor(exprsses Runx2)

Runx2

Osterix

Transcriptionb-cateninHoxa-2

Transcription

SOX9

Chondrocytes

NKX3.2

Wnt signallingActivates

Stabilzes

CAMP/PKA

osteoblasts

Targeted differentiationHallmarks of differentiation as default (upon Runx2 downregulation)

Osteoblast versus chondrocyte differentiationExpert Reviews in Molecular Medicine © 2013 Cambridge University Press

Figure 2. Osteoblast versus chondrocyte differentiation. Differentiation to the chondrocyte lineage mayoccur, by default, with suppression of Runx2 and osteoblastogenesis. Conversely, osteoblast differentiationmay be actively suppressed by Sox9 and/or its regulators.




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

CREB, as the Runx2 promoter contains afunctional CREB domain (Ref. 57). Anotherlikely effect of cAMP is its modulation of Runx2activity. Runx2 has putative PKAphosphorylation sites in its activation domain,and treatment with a PKA-specific cAMPanalogue increases its phosphorylation (Refs 58,59) and activation of the collagenase-3 promoter(Ref. 58) and the matrix metalloproteinase-13promoter (Ref. 59). The role that PKAphosphorylation of Runx2 has in vivo duringosteoblast differentiation remains to bedetermined.As mentioned earlier, Osterix is another

transcription factor essential for osteoblastdifferentiation and bone development. As Runx2KO mice do not express Osterix (Refs 18, 19) butOsterix KO mice do express Runx2, it wasdeduced that Osterix is downstream of Runx2(Ref. 60). Analysis of the Osterix gene promoteridentified a Runx2 binding site, and Runx2 wasreported to upregulate Osterix expression(Ref. 61). A putative CREB site was also foundin this gene promoter, suggesting cAMP may beable to regulate Osterix expression (Ref. 61). Insupport of this, treatment of cells lines orprimary mouse MSCs with cAMP analoguesincreases Osterix expression (Refs 51, 54, 62).However, exposure to relatively high cAMP(1 mM) inhibits its expression in UM-106-01 cellline or in mouse primary osteoblasts (Ref. 63).Similarly to Runx2 expression, Gαs receptoractivation causes an increase in Osterixexpression (Ref. 54), the mechanism of whichremains to be demonstrated.Dlx5, an upregulator of Runx2 (Ref. 22), is

phosphorylated by PKA, which increases Dlx5protein levels by augmenting its stability. PKAsignalling also increases Dlx5 transcriptionalactivity. Therefore, PKA signalling enhancesDlx5-induced osteoblast differentiation (Ref. 64).Of note, some have reported a negative effect of

cAMP on osteoblast differentiation. Yang et al.found that the PKA inhibitor, PKI, increased theexpression of Runx2 in a human MSC-derivedcell line whereas forskolin, 3-isobutyl-1-methylxantine and a cAMP analogue decreasedthe expression of osteopontin (Ref. 65). Kohet al. found forskolin to decrease osteocalcinexpression in MC3T3-E1 cells. Forskolin alsodecreased the number of mineralised nodulesformed by rat primary calvarial cells (Ref. 66).Tintut et al. reported that treatment of MC3T3-E1

cells with forskolin decreased the activity ofalkaline phosphatase and inhibited mineralisation,as well as decreased the expression of alkalinephosphatase, bone sialoprotein, osteocalcin andosteopontin (Ref. 67). The discrepancy in reportedeffects of cAMP on osteoblastogenesis may bedue, in part, to the relatively high concentrationof forskolin or cAMP analogue used (10–100 μM).For instance, Turksen et al. found that treatmentwith 10 μM of forskolin inhibited osteoblastdifferentiation whereas 1 nM increased it (Ref. 68).

cAMP effect on Sox9 and chondrogenesisMultiple studies have suggested synergismbetween cAMP signalling and Sox9 expression.The Sox9 promoter contains a CRE site (Ref. 69)and, using a Sox9 gene promoter reporterconstruct in various MSC cell lines, it wasreported that binding of this site by CREBincreases Sox9 promoter activity (Ref. 70).

Sox9 interacts with CREB binding protein (CBP)and p300 to increase its transcriptional activity.Using a Col2α1 gene promoter reporterconstruct in a chondrocyte cell line, as well as aGal4-Sox9 fusion protein, it was reported thatco-transfection with CBP and/or p300 increasedSox9 activityas a transcriptional activator (Ref. 71).

A yeast two-hybrid screen of Sox9 bindingpartners identified the PKA catalytic subunit asinteracting with Sox9 (Ref. 72). This promptedthe investigation of PKA-dependent Sox9phosphorylation and two serine residues werefound to be phosphorylated (Ref. 72).Phosphorylation of these sites by PKA increasedSox9 activation of the Col2α1 gene promoter(Ref. 72). Zhao et al. investigated this findingfurther by mutating the serine residues. Theabsence of PKA phosphorylation partiallyreversed PKA enhancement of Sox9 activity,suggesting that PKA augmented Sox9 activitythrough multiple mechanisms. The investigatorsfound an interaction between Sox9 and CBP inchondrogenic differentiating C3H10T1/2 cells byusing co-immunoprecipitation, and thisinteraction increased Sox9 transcriptionalactivity. The importance of this interaction wasconfirmed by mutating a CREB site in theregion of the Sox9 interaction, which reversedthe effect (Ref. 73). Therefore, cAMP signalling,by means of PKA, enhances Sox9 expressionand transcriptional activity through directphosphorylation and activation of CREB. Inaddition to findings in a cell line, differentiation




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

of chick limb bud chondrocytes in the presence ofthe PKA inhibitor H89 blocked chondrogenesis.Yoon et al. found PKA to promotechondrogenesis by downregulating N-Cadherinthrough a PKC-dependent pathway (Ref. 74).Taken together, it appears that cAMP signalling

is capable of upregulating Runx2, a masterregulator of osteoblasts, and Sox9, a masterregulator of chondrocytes. Since Sox9 is aninhibitor of Runx2 expression, it is possible thatthe sum effect of cAMP on lineagedetermination depends on its concentration,cellular localisationandapossible regulatory loop.

Adenosine receptorsAdenosine is a regulatory metabolite and itsreceptors are comprised of a family of seventransmembrane domain G protein-coupledreceptors. A1 and A3 adenosine receptors(A1AR and A3AR) are coupled to Gαi and are,therefore, adenylyl cyclase inhibiting, whereasA2A and A2B adenosine receptors (A2AAR andA2BAR) are coupled to Gαs, which stimulatesadenylyl cyclase and produces a cAMP signal.The A2BAR may also be coupled to Gαq, whichactivates phospholipase C (Ref. 75). A2BAR hasa relatively low affinity for adenosine. Highextracellular concentrations of adenosine can beachieved during cell injury or stress (Ref. 76).A2AAR and A2BARs are widely expressed tovarying degrees, with high expression in thevasculature (Ref. 77). Our laboratory and othershave shown that A2BAR expression is inducibleunder stress conditions such as inflammationand hypoxia (reviewed in Ref. 78). A2AAR andA2BAR are expressed in MSCs (Refs 56, 79), andinterestingly, high expression of the A2AARand A2BAR receptors can be found in cartilage(Ref. 80).

Adenosine receptors and osteoblast andchondrocyte differentiationOnly recently has the role of adenosine receptors inosteoblast differentiation been examined. In abroad sense, purinergic signalling was looked atin the context of bone, and ATP receptors (P2Xand P2Y) were found to promote differentiationand proliferation in an osteoblast cell line(MC3T3-E1) (Ref. 81). It was found that afterbone injury and when exposed to hypoxicconditions, rat osteoblasts secrete ATP in thehigh nM to μM range (Ref. 82). This extracellularATP is available for catabolism to adenosine by

ectonucleotidases, which are expressed onosteoblasts (Ref. 79). In accordance, it wasrecently demonstrated that genetic ablation ofthe ectonucleotidase CD73 results in osteopeniaand decreased osteoblast differentiation in mice(Ref. 83).

In vitro studies found adenosine receptors to beexpressed in both human (Ref. 79) and rodentMSCs (Ref. 56). Based on its expression andmeasurements of cAMP levels after agonisttreatments of rat MSCs, Gharibi et al. concludedthat the A2BAR is the dominant receptor,relative to other adenosine receptors, and that itsexpression increases during osteoblastdifferentiation (Ref. 56). In human MSCs,A2BAR activation increased osteoblastdifferentiation, as determined by an increase inalkaline phosphatase activity (Ref. 79). Similarly,in rat MSCs, activation of the A2BAR increasedRunx2 and alkaline phosphatase expression, aswell as the number of mineralised nodules(Ref. 56). We found bone marrow-derived MSCsfrom A2BAR KO mice to have decreasedosteoblast differentiating potential, withdiminished expression of Runx2 and Osterix.Activation of the A2BAR with pharmacologicalagonists increased the expression of thesetranscription factors, as well as caused anincrease in the number of mineralised nodules.Treatment with a cAMP analogue also increasedthe expression of Osterix, suggesting that cAMPmay be the mechanism of action for A2BAReffect on differentiation. In addition, the A2BARKO mouse had mild osteopenia, and a delayedor impaired bone fracture healing response(Ref. 54).

There have been few reports on the effect of thecAMP inhibitory adenosine receptors, A1ARand A3AR on osteoblast differentiation.Overexpression of the A1AR in an osteoblastprecursor cell line led to inhibition of osteoblastdifferentiation, and instead promoteddifferentiation to adipocytes (Ref. 84). However,treatment of human MSCs with the A1ARagonist, N6-cyclopentyladenosine, caused anincrease in osteoblast differentiation (Ref. 79). Inrat MSCs, very little A3AR was found (Ref. 56) andin human MSCs, treatment with the A3AR agonist,1-Deoxy-1-[6-[[(3-iodophenyl)methyll] amino]-9H-purin-9-yl]-N-methyl-β-D-ribofuranuronamide, hadno effect on osteoblast differentiation (Ref. 79).

A role for adenosine receptors in chondrocytedifferentiation has not yet been examined and/




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

or reported. However, as A1- andA2-type ARs areknown to be expressed in MSCs (Refs 56, 79), andgiven the published influence of cAMPon Sox9 (asdescribed above), it is likely that a role for thesereceptors on the differentiation of this lineagewill be found.

Inflammation and osteoblast andchondrocyte differentiation: implication

for adenosine receptorsInflammation plays a significant role in bonedevelopment (reviewed in Ref. 85) andregeneration (reviewed in Ref. 86). Althoughadenosine receptor signalling through cAMP hasthe potential to directly influence osteo- orchondrogenesis, cAMP signalling through thesereceptors can also affect the level ofinflammatory cytokines systemically or at thecellular level (reviewed in Ref. 87). Here, we willfocus on a brief survey of the effects ofinflammatory processes on bone celldifferentiation, followed by a summary of A2-type adenosine receptors effects oninflammation and its potential influence on bonecell lineages.

The effect of inflammation on osteoblastdifferentiationThe effect of inflammation on the skeletal system,including osteoblast differentiation, has been wellstudied, particularly the effects of tumour necrosisfactor-α (TNF-α) (reviewed in Ref. 88).Experiments by Gilbert et al. in both fetalcalvarial cells and the osteoblast precursor cellline MC3T3-E1 show that TNF-α inhibitsosteoblast differentiation, specifically at the earlystage of lineage commitment (Ref. 89). Furtherexperimentation showed this inhibition to beassociated with downregulation of Runx2transcription and a subsequent decrease innuclear Runx2 (Ref. 90). Although these authorsdid not find NFκB signalling to be involved,Huang et al. found that inhibition of NFκBsignalling by overexpression of IκB in ST2 cells,a MSC line, abolished the inhibitory effect ofTNF-α on Runx2 gene expression (Ref. 91). Theinhibitory effect of TNF-α on mRNA expressionand osteoblast differentiation was confirmed inprimary mouse MSCs. Here, Lacey et al. alsofound interleukin-1β (IL-1β) to have a similareffect (Ref. 92).These studies on mRNA expression were

complimented by Kaneki et al. who found

TNF-α to promote Runx2 protein degradation inC2C12 and 2T3 osteoblast precursor cells byupregulating the E3 ligases Smurf1 and Smurf2(Ref. 93). In order to examine this phenomenon inprimary MSCs, Zhao et al. isolated an MSC-enriched fraction from the bone marrow of TNF-αoverexpressing mice. In these cells they foundupregulation of the E3 ligase Wwp1 relative tocells from wildtype (WT) mice. Furtherexperimentation found Wwp1 to be upregulatedby TNF-α and responsible for inhibiting osteoblastdifferentiation. However, in these primary cells,the inhibition was caused by the degradation ofJunB, a promoter of osteoblast differentiation,rather than affecting Runx2 directly (Ref. 94).

In addition to their studies on Runx2, Lu et al.found TNF-α treatment to inhibit OsterixmRNA expression, and claim this to be a directeffect of TNF-α signalling through mitogenactivated (MEK) and inhibition of the Osterixgene promoter (Ref. 95). In addition, they foundTNF-α-stimulated binding of paired mesodermhomeobox protein 1(Prx1), causing inhibition ofOsterix gene promoter activity and transcriptionin MC3T3 and C3H10T1/2 cells (Ref. 96).Interestingly, Prx1 is a developmental regulatorof skeletogenesis that was previously thought tobe silenced after embryogenesis (Ref. 96).

Other regulators of osteoblast differentiation areaffected by TNF-α. In culture, Msx2 has beenshown to inhibit osteoblast differentiation(Ref. 97). Treatment of C2C12 or fetal calvarialcells with TNF-α caused increased expression ofMsx2 and reduced expression of alkalinephosphatase. This effect was independent ofRunx2, as the phenomenon was maintained inRunx2 null cells, and overexpression ofdominant negative IκB showed NFκB signallingto be involved. To model inflammation inducedby wear of artificial limbs, macrophages wereactivated with titanium particles and theirmedia, enriched in inflammatory cytokines, wasused to treat MC3T3-E1 cells. Treatment withthis conditioned media or TNF-α inhibitedRunx2 expression and osteoblast differentiationand this was attributed to NFκB activation andincreased expression of sclerostin, an inhibitor ofthe Wnt pathway (Ref. 98).

The effect of inflammation on chondrocytedifferentiationRelative to osteoblastogenesis, little is knownabout the effect of inflammation on chondrocyte




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

differentiation. IL-1 and TNF-α decreased Sox9mRNA and protein expression in a chondrocyticcell line, MC615, and in primary mousechondrocytes through the NFkB signallingpathway (Refs 99, 100). However, in thesestudies differentiation was not directlyaddressed. To study chondrocyte differentiation,Nakajima et al. differentiated a chondrocyteprogenitor cell line, ATDC5, in the presence ofIL-6. IL-6 decreased collagen 2α1 and collagen10 mRNA expression in a dose-dependentmanner and inhibited the formation of cartilagenodules (Ref. 101). Wehling et al. differentiatedhuman MSCs to chondrocytes in the presence ofeither IL-1 or TNF-α. They found that bothcytokines decreased the size of the cartilagepellet and lowered the amount of

glycosaminoglycan accumulation. IL-1 treatmentdecreased the expression of Col2α1 andaggrecan mRNA (Ref. 102).

Taken together, it is then quite possible thatadenosine receptor modulation of inflammatorycytokines affects directly or indirectly theprocess of MSC differentiation into bone celllineages. Figure 3 illustrates a proposedmechanism of A2BAR effects on osteoblastdifferentiation involving CREB and/or TNF-α.

Adenosine receptors and inflammationOne of the principal attributes of adenosine and itsreceptors is its pro- or anti-inflammatory effects,most of which are mediated via cAMPsignalling. The concentration of our laboratoryhas been the A2BAR, and thus will be the focusof this section. Both pro- and anti-inflammatoryeffects of the A2BAR have been described, andthese depend on the cell type and stimulus. Inaddition, these effects have been ascribed eitherprotective or deleterious roles, depending on thecontext, e.g. chronic versus acute pathology(reviewed in Ref. 103). Complete KO of theA2BAR gene results in a slight systemicinflammation, as KO animals have elevatedplasma levels of TNF-α at baseline, and elevatedlevels of TNF-α and IL-6 expression inmacrophages (Ref. 104). These differences incytokine levels are exacerbated upon stress orinjury. Treatment of A2BAR KO mice withlipopolysaccharide causes an exceptionalincrease in TNF-α and IL-6 plasma levels and inmacrophage expression, relative to WT mice(Ref. 104).

The ability of the A2BAR to dampeninflammation in response to stimuli highlightsits protective role during stress or injury. Theimportance of this role is confirmed by findingsthat the expression of the receptor is alsoinduced by these stimuli. Treatment of vascularsmooth muscle cells with TNF-α causes anupregulation of A2BAR expression. Further, thisincrease was shown to be mediated by NADPHoxidase 4 (Nox4) signalling (Ref. 105).

AsNFκB has an important role in the regulationof cellular inflammation it has been a target ofinvestigation in relation to A2BAR signalling.Recently, we found the A2BAR to directly bindp105, an inhibitor of NFκB, stabilising it andpreventing its degradation (Ref. 106). Thismechanism helps explain the inflammatoryphenotype in the A2BAR KO mice and the

Injury/stress

A2BARMSC

cAMP

PKA

Runx2-

CREB Runx2

OsterixOsteoblast

differentiation

TNF-a

Adenosine

P

Proposed mechanism for A2BARaction on osteoblast differentiationExpert Reviews in Molecular Medicine © 2013 Cambridge University Press

Figure 3. Proposed mechanism for A2BARaction on osteoblast differentiation. Activationof the mesenchymal stem cell A2BAR triggerscAMP signalling, which may modulate theexpression and/or activity of a key osteoblasttranscription factor, Runx2, and promoteosteoblast differentiation. A2BAR activationalso decreases tumour necrosis factor-α level,which has an inhibitory effect on osteoblastdifferentiation.




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

ability of the A2BAR to downregulateinflammation.Considering the above described effects of TNF-

α, IL-1 and IL-6 on bone cell differentiation, it ispossible that some of the A2BAR’s recentlydescribed protective effect in a mouse boneinjury model (Ref. 54) is related to changes incytokine levels.

Adenosine receptors as therapeutics forosteoporosis, skeletal injury and arthritisA role for adenosine receptors in osteoporosis isgaining increasing momentum. In mouse bonemarrow-derived cells, antagonism of the A1ARwith 8-cyclopentyl-1,3-dipropylxanthine inhibitedthe differentiation of osteoclasts (bonereabsorbing cells) (Ref. 107). Also in these cells,treatment with an A2AAR agonist, 4-[2-[[6-Amino-9-(N-ethyl-β-D-ribofuranuronamidosyl)-9H-purin-2-yl]amino]ethyl]benzene propanoicacid hydrochloride (CGS21680), inhibited thedifferentiation of osteoclasts, as well as inhibitedtheir activity. In addition, the A2AAR KO mousehas an increased number of osteoclasts, as well asreduced bone volume (Ref. 108). Similarly,A2BAR KO mice display a mild reduction inbone density (Ref. 54). In our study, we alsofound that A2BAR KO mice have changes inbone fracture healing with decreased callus boneformation and an apparent delay in healing(Ref. 54). Although the anti-inflammatory effectsof adenosine likely have an important role inimproving skeletal injury repair, we contend thatadenosine, through the A2BAR, may alsoimprove healing by promoting osteoblastdifferentiation (Ref. 54). Therefore, agonism ofthe A2-type ARs may be useful as a therapeuticfor osteoporosis and bone injury.Ascaffeineisanantagonistofadenosinereceptors,

its effects on bone may be relevant here. Inepidemiological studies of risk factors forosteoporosis in humans, caffeine was either foundto be negatively associated with (Refs 109, 110) ornot associated with bone mineral density(Refs 111, 112, 113, 114). It is possible that effects ofcaffeine are confounded by factors such as age,oestrogen levels and calcium intake. In a study ofover 3000 individuals using the FraminghamCohort, it was found that caffeine intake wasassociated with a higher relative risk of hipfracture (Ref. 115). In experimental animals,caffeine has been shown to inhibit bone formation.When demineralised bone particles were

implanted subcutaneously, rats that were treatedwith caffeine had decreased chondrogenesis anddecreased mineralisation (Ref. 116), suggestingthat caffeine impairs new endochondral boneformation by inhibiting the proliferation anddifferentiation of chondroprogenitor cells. Inaddition, chick osteoblasts treated with caffeinehad decreased collagen expression and alkalinephosphatase activity, resulting in reduced matrixformation (Ref. 117). Finally, differentiation ofosteoclasts from mouse bone marrow-derived cellswas enhanced with caffeine treatment (Ref. 118).Whether any or all of these negative effects ofcaffeine on bone formation and maintenance issolely because of antagonism of adenosinereceptors has not been determined.

Adenosine receptors have been found to beprotective against a variety of injuries, includingbut not limited to cardiovascular (reviewed inRef. 119), kidney (reviewed in Ref. 120), lung(reviewed in Ref. 121) and gastrointestinal(reviewed in Ref. 122). As arthritis is aninflammatory disease of the joints, adenosinereceptors have been investigated in the contextof this disease. Direct infusion of adenosine intothe joint in a rat arthritis model reduced thepathogenesis of the disease (Ref. 123).Methotrexate is a drug commonly used for thetreatment of rheumatoid arthritis andameliorates the condition by decreasing theinflammatory response in the joints. At leastpart of its inflammatory action is attributed toits ability to increase adenosine release in thejoints (Ref. 124). Currently, the principal receptorto be implicated is the A3AR. An A3AR receptoragonist, CF-101, has been undergoing clinicaltrials for the treatment of rheumatoid arthritis;however, the improvement in rheumatoidarthritis has not yet reached statisticalsignificance (Ref. 125).

ConclusionsThere are a multitude of signalling molecules andpathways that converge during chondrocyte andosteoblast differentiation. The coordination ofthe spatial and temporal pattern of thesemolecules is necessary for correct boneformation. Through exploration of thedownstream signalling of G-protein-coupledreceptors, cAMP was found to be an importantcomponent of the signalling pathways neededfor full differentiation of MSCs along theskeletal lineage. Continued investigation into the




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

contribution of these receptors, adenosinereceptors among them, to bone homoeostasisand regeneration could lead to importantdiscoveries with clinically therapeuticimplications.

References1 Olsen, B.R., Reginato, A.M. and Wang, W. (2000)

Annual Reviewof Cell andDevelopmental Biology16, 191-220

2 Hartmann, C. (2009) CurrentOpinion inGenetics&Development 19, 437-443

3 Schindeler, A. et al. (2008) Seminars in Cell &Developmental Biology 19, 459-466

4 Baldridge, D. et al. (2010) Annual Review ofGenomics and Human Genetics 11, 189-217

5 DeLise, A.M., Fischer, L. and Tuan, R.S. (2000)Osteoarthritis and Cartilage 8, 309-334

6 Hall, B.K. and Miyake, T. (2000) Bioessays 22,138-147

7 Enomoto, H. et al. (2000) Journal of BiologicalChemistry 275, 8695-8702

8 Hill, T.P. et al. (2005) Developmental Cell 8, 727-7389 Day, T.F. et al. (2005) Developmental Cell 8, 739-75010 Hu, H. et al. (2005) Development 132, 49-6011 Foster, J.W. et al. (1994) Nature 372, 525-53012 Wagner, T. et al. (1994) Cell 79, 1111-112013 Augello, A. and De Bari, C. (2010) Human Gene

Therapy 21, 1226-123814 Smith,N. et al. (2005) Journal ofCellular Physiology

203, 133-14315 Zhou, G. et al. (2006) Proceedings of the National

Academy of Sciences of the United States ofAmerica 103, 19004-19009

16 Cheng, A. and Genever, P.G. (2010) Journal of Boneand Mineral Research 25, 2680-2689

17 Lengner, C.J. et al. (2005) Journal of BiologicalChemistry 280, 15872-15879

18 Otto, F. et al. (1997) Cell 89, 765-77119 Komori, T. et al. (1997) Cell 89, 755-76420 Bialek, P. et al. (2004) Developmental Cell 6, 423-43521 Kraus, P. and Lufkin, T. (2006) American Journal of

Medical Genetics. Part A 140, 1366-137422 Lee, M.H. et al. (2005) Journal of Biological

Chemistry 280, 35579-3558723 Robledo, R.F. et al. (2002)Genes&Development 16,

1089-110124 Komori, T. (2006) Journal of Cellular Biochemistry

99, 1233-123925 Han, J. et al. (2007) Mechanisms of Development

124, 729-74526 Lin, G.L. and Hankenson, K.D. (2011) Journal of

Cellular Biochemistry 112, 3491-3501

27 Milat, F. and Ng, K.W. (2009) Molecular andCellular Endocrinology 310, 52-62

28 Dong, Y.F. et al. (2006) Journal of CellularPhysiology 208, 77-86

29 Topol, L. et al. (2009) Journal of BiologicalChemistry 284, 3323-3333

30 Chen, J. and Long, F. (2012) Journal of Bone andMineral Research doi: [10.1002/jbmr.1834] [epubahead of print]

31 Sands, W.A. and Palmer, T.M. (2008) CellularSignalling 20, 460-466

32 Mayr, B. andMontminy,M. (2001)Nature Reviews.Molecular Cell Biology 2, 599-609

33 Lee, K.A. and Masson, N. (1993) Biochimica etBiophysica Acta 1174, 221-233

34 Beavo, J.A. (1995) Physiological Reviews 75, 725-74835 Wilson, L.C. and Hall, C.M. (2002) Seminars in

Musculoskeletal Radiology 6, 273-28336 Riminucci, M. et al. (2010) Journal of Molecular

Endocrinology 45, 355-36437 Hsiao, E.C. et al. (2008) Proceedings of the National

Academy of Sciences of the United States ofAmerica 105, 1209-1214

38 Sakamoto, A. et al. (2005) Journal of BiologicalChemistry 280, 21369-21375

39 Wu, J.Y. et al. (2011) Journal ofClinical Investigation121, 3492-3504

40 Datta, N.S. and Abou-Samra, A.B. (2009) CellularSignalling 21, 1245-1254

41 Abou-Samra, A.B. et al. (1992) Proceedings of theNational Academy of Sciences of the United Statesof America 89, 2732-2736

42 Lanske, B. et al. (1996) Science 273, 663-66643 Jilka, R.L. (2007) Bone 40, 1434-144644 Fei, Y. and Hurley, M.M. (2012) Journal of Cellular

Physiology 227, 3539-354545 Yu, B. et al. (2012) Journal of Bone and Mineral

Research 27, 2001-201446 Takase, H. et al. (2009) Hormone and Metabolic

Research 41, 861-86547 Nakao, Y. et al. (2009) Bone 44, 872-87748 Chan, G.K. et al. (2003) Endocrinology 144,

5511-552049 Hollnagel, A., Ahrens, M. and Gross, G. (1997)

Journal of Bone andMineral Research 12, 1993-200450 Vortkamp, A. et al. (1996) Science 273, 613-62251 Ghayor,C. et al. (2009) Biochemical andBiophysical

Research Communications 381, 247-25252 Siddappa, R. et al. (2008) Proceedings of the

National Academy of Sciences of the United Statesof America 105, 7281-7286

53 Lo, K.W. et al. (2012) Journal of Tissue Engineeringand Regenerative Medicine 6, 40-48




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

54 Carroll, S.H. et al. (2012) Journal of BiologicalChemistry 287, 15718-15727

55 Krishnan, V. et al. (2003) Molecular Endocrinology17, 423-435

56 Gharibi, B. et al. (2011) Journal of Bone andMineralResearch 26, 2112-2124

57 Xiao, Z.S. et al. (2001) Journal of CellularBiochemistry 82, 647-659

58 Selvamurugan, N. et al. (2000) Journal of BiologicalChemistry 275, 5037-5042

59 Selvamurugan, N. et al. (2009) FEBS Letters 583,1141-1146

60 Nakashima, K. et al. (2002) Cell 108, 17-2961 Nishio, Y. et al. (2006) Gene 372, 62-7062 Wang, B.L. et al. (2006) Journal of Endocrinological

Investigation 29, 101-10863 Hong, S.H. et al. (2009) Journal of Molecular

Endocrinology 43, 197-20764 Han, Y. et al. (2011) Biochemical and Biophysical

Research Communications 407, 461-46565 Yang, D.C. et al. (2008) PLoS One 3, e154066 Koh, A.J. et al. (1999) Endocrinology 140, 3154-316267 Tintut, Y. et al. (1999) Journal of Biological

Chemistry 274, 28875-2887968 Turksen, K. et al. (1990) Journal of Cellular

Physiology 142, 61-6969 Kanai, Y. and Koopman, P. (1999) Human

Molecular Genetics 8, 691-69670 Piera-Velazquez, S. et al. (2007) Experimental Cell

Research 313, 1069-107971 Tsuda, M. et al. (2003) Journal of Biological

Chemistry 278, 27224-2722972 Huang, W. et al. (2000) Molecular and Cellular

Biology 20, 4149-415873 Zhao, L., Li, G. and Zhou, G.Q. (2009) Journal of

Bone and Mineral Research 24, 826-83674 Yoon, Y.M. et al. (2000) Journal of Bone andMineral

Research 15, 2197-220575 Fredholm, B.B. et al. (2001) Pharmacological

Reviews 53, 527-55276 Fredholm, B.B. (2007) Cell Death and

Differentiation 14, 1315-132377 St Hilaire, C. et al. (2009) Journal of Cellular

Physiology 218, 35-4478 Hasko, G. et al. (2009) Trends in Immunology 30,

263-27079 Costa, M.A. et al. (2011) J Journal of Cellular

Physiology 226, 1353-136680 Koolpe, M., Pearson, D. and Benton, H.P. (1999)

Arthritis and Rheumatism 42, 258-26781 Shimegi, S. (1996) Calcified Tissue International 58,

109-11382 Orriss, I.R. et al. (2006) Bone 39, 300-309

83 Takedachi, M. et al. (2011) Journal of CellularPhysiology 227, 2262-2631

84 Gharibi, B. et al. (2012) International Journal ofObesity 36, 397-406

85 DeBenedetti, F. (2009)HormoneResearch 72(Suppl1), 26-29

86 Mountziaris, P.M. and Mikos, A.G. (2008) TissueEngineering Part B: Reviews 14, 179-186

87 Blackburn, M.R. et al. (2009) Handbook ofExperimental Pharmacology 195, 215-269

88 Nanes, M.S. (2003) Gene 321, 1-1589 Gilbert, L. et al. (2000) Endocrinology 141, 3956-396490 Gilbert, L. et al. (2002) Journal of Biological

Chemistry 277, 2695-270191 Huang,H. et al. (2011) Cell Proliferation 44, 420-42792 Lacey,D.C. et al. (2009)Osteoarthritis andCartilage

17, 735-74293 Kaneki, H. et al. (2006) Journal of Biological

Chemistry 281, 4326-433394 Zhao, L. et al. (2011) Stem Cells 29, 1601-161095 Lu, X. et al. (2006) Journal of Biological Chemistry

281, 6297-630696 Lu, X. et al. (2011) Journal of Bone and Mineral

Research 26, 209-21997 Lee, H.L. et al. (2010) Experimental & Molecular

Medicine 42, 437-44598 Lee, S.S. et al. (2012) Biomaterials 33, 4251-426399 Murakami, S., Lefebvre, V. and deCrombrugghe, B.

(2000) Journal of Biological Chemistry 275,3687-3692

100 Sitcheran, R., Cogswell, P.C. and Baldwin, A.S., Jr(2003) Genes & Development 17, 2368-2373

101 Nakajima, S. et al. (2009) Cytokine 47, 91-97102 Wehling, N. et al. (2009) Arthritis and Rheumatism

60, 801-812103 Johnston-Cox, H.A., Koupenova, M. and Ravid, K.

(2012) Arteriosclerosis, Thrombosis, and VascularBiology 32, 870-878

104 Yang, D. et al. (2006) Journal of ClinicalInvestigation 116, 1913-1923

105 St Hilaire, C. et al. (2008) Biochemical andBiophysical Research Communications 375,292-296

106 Sun, Y. et al. (2012) Journal of Cell Science 125,4507-4517

107 He,W. andCronstein, B.N. (2012) Purinergic Signal8, 327-337

108 Mediero, A. et al. (2012) American Journal ofPathology 180, 775-786

109 Rapuri, P.B. et al. (2001) American Journal ofClinical Nutrition 74, 694-700

110 Ilich, J.Z. et al. (2002) Journal of the AmericanCollege of Nutrition 21, 536-544




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

111 Wetmore, C.M. et al. (2008) OsteoporosisInternational 19, 519-527

112 Waugh, E.J. et al. (2009) Osteoporosis International20, 1-21

113 Hannan, M.T. et al. (2000) Journal of Bone andMineral Research 15, 710-720

114 Conlisk, A.J. and Galuska, D.A. (2000) PreventiveMedicine 31, 562-568

115 Kiel, D.P. et al. (1990) American Journal ofEpidemiology 132, 675-684

116 Barone, L.M. et al. (1993) Journal of CellularBiochemistry 52, 171-182

117 Tassinari, M.S. et al. (1991) Journal of Bone andMineral Research 6, 1029-1036

118 Liu, S.H. et al. (2011) Journal of OrthopaedicResearch 29, 954-960

119 Riksen, N.P. and Rongen, G.A. (2012)Expert Review of Clinical Pharmacology 5,199-218

120 Bauerle, J.D. et al. (2011) Journal of the AmericanSociety of Nephrology 22, 14-20

121 Eckle, T., Koeppen, M. and Eltzschig, H.K. (2009)Physiology (Bethesda) 24, 298-306

122 Colgan, S.P. and Eltzschig, H.K. (2012) AnnualReview of Physiology 74, 153-175

123 Green, P.G. et al. (1991) Proceedings of the NationalAcademy of Sciences of the United States ofAmerica 88, 4162-4165

124 Cronstein, B.N., Naime, D. and Ostad, E. (1993)Journal of Clinical Investigation 92, 2675-2682

125 Silverman, M.H. et al. (2008) Journal ofRheumatology 35, 41-48

Features associated with this article

FiguresFigure 1. Transcriptional regulators of mesenchymal stem cell (MSC) differentiation to skeletal lineages.Figure 2. Osteoblast versus chondrocyte differentiation.Figure 3. Proposed mechanism for A2BAR action on osteoblast differentiation.

Citation details for this article

Shannon H. Carroll and Katya Ravid (2013) Differentiation of mesenchymal stem cells to osteoblasts andchondrocytes: a focus on adenosine receptors. Expert Rev. Mol. Med. Vol. 15, e1, February 2013,doi:10.1017/erm.2013.2




Differen

tiationofmes

ench

ymal

stem

cells

toosteo

blastsan

dch

ond

rocy

tes:

afocu

sonad

enosine

rece

ptors

Differentiation of mesenchymal stem cells to osteoblasts and chondrocytes: a focus on adenosine...

Documents

Transcript of Differentiation of mesenchymal stem cells to osteoblasts and chondrocytes: a focus on adenosine...