A r t i c l e s

454 VOLUME 17 | NUMBER 4 | APRIL 2011 nAture medicine

Heterotopic ossification is a potentially severe pathology in which ectopic bone forms within muscles and connective tissues and near blood vessels or nerves. More than 10% of patients who undergo invasive surgery can develop heterotopic ossification1–3, and 65% of seriously wounded soldiers develop the disease as well4, caus-ing chronic pain, problems with fitting prostheses, deep venous thrombosis, limited motion and other complications. Although the pathogenesis of this condition remains unclear, it is generally agreed that the inciting events—trauma, surgery, deep burns or protracted immobilization—induce local inflammation5,6. This is followed by the recruitment of skeletal progenitor cells that differentiate into chondrocytes, undergo hypertrophy and are replaced by endochon-dral bone. Individuals with the rare congenital condition fibrodys-plasia ossificans progressiva (FOP) develop extensive heterotopic ossification that is also induced by trauma and inflammation7, is aggressive and often fatal and is propelled by an activating mutation in the BMP type I receptor ALK2 R206H (ref. 8). Given the etiology of heterotopic ossification, current therapeutic treatments aim to target different pathogenic steps, but they have had varied success and are far from ideal9,10.

A recent study tested a selective inhibitor of BMP type I receptor kinases in transgenic mice11 that express a strong, constitutively active ALK2 Q207D mutant12 and found some inhibition of muscle- associated heterotopic ossification13. Using a standard mouse model of subcutaneous heterotopic ossification, we tested a selective agonist

for the nuclear RAR-α14. The rationale was based on the fact that retinoid signaling is a strong inhibitor of chondrogenesis15 and that unliganded RAR transcriptional repressor activity—possibly involv-ing RAR-α—is needed for chondrogenic differentiation16,17. The RAR-α agonist did inhibit heterotopic ossification, but not com-pletely. To further explore the idea of a retinoid agonist–based therapy for heterotopic ossification, we considered the fact that another RAR family member with distinct characteristics18, RAR-γ, is expressed in chondrogenic cells17 and chondrocytes19 where it also operates as an unliganded transcriptional repressor20. Hence, a therapy for heterotopic ossification that was based on RAR-γ agonists could be more effective because it would target both chondrogenic cells and chondrocytes. The data presented here support this possibility.

RESULTSRAR-g agonists block chondrogenesisTo test whether RAR-γ agonists inhibit chondrogenesis, we treated micromass cultures of mouse embryo limb mesenchymal cells isolated on embryonic day (E) 11.5 with the synthetic selective RAR-γ ago-nist NRX204647 (refs. 20,21). For comparison, we treated compan-ion cultures with all-trans-retinoic acid, the active natural vitamin A derivative that is a pan-agonist of the RAR family (RAR-α, RAR-β and RAR-γ) and is well known for its antichondrogenic action15. Whereas numerous Alcian blue–positive cartilaginous nodules formed in control cultures (Fig. 1a), few if any formed in cultures treated with

1Department of Orthopaedic Surgery, Thomas Jefferson University College of Medicine, Philadelphia, Pennsylvania, USA. 2Io Therapeutics, Irvine, California, USA. 3School of Dentistry, University of Michigan, Ann Arbor, Michigan, USA. 4Present address: The Children’s Hospital of Philadelphia, Division of Orthopaedic Surgery, Philadelphia, Pennsylvania, USA. Correspondence should be addressed to M.I. (iwamotom@email.chop.edu) or M.P. (pacificim@email.chop.edu).

Received 15 September 2010; accepted 18 February 2011; published online 3 April 2011; doi:10.1038/nm.2334

Potent inhibition of heterotopic ossification by nuclear retinoic acid receptor-γ agonistsKengo Shimono1,4, Wei-en Tung1,4, Christine Macolino1, Amber Hsu-Tsai Chi1, Johanna H Didizian1,4, Christina Mundy1, Roshantha A Chandraratna2, Yuji Mishina3, Motomi Enomoto-Iwamoto1,4, Maurizio Pacifici1,4 & Masahiro Iwamoto1,4

Heterotopic ossification consists of ectopic bone formation within soft tissues after surgery or trauma. It can have debilitating consequences, but there is no definitive cure. Here we show that heterotopic ossification was essentially prevented in mice receiving a nuclear retinoic acid receptor-g (RAR-g) agonist. Side effects were minimal, and there was no significant rebound effect. To uncover the mechanisms of these responses, we treated mouse mesenchymal stem cells with an RAR-g agonist and transplanted them into nude mice. Whereas control cells formed ectopic bone masses, cells that had been pretreated with the RAR-g agonist did not, suggesting that they had lost their skeletogenic potential. The cells became unresponsive to rBMP-2 treatment in vitro and showed decreases in phosphorylation of Smad1, Smad5 and Smad8 and in overall levels of Smad proteins. In addition, an RAR-g agonist blocked heterotopic ossification in transgenic mice expressing activin receptor-like kinase-2 (ALK2) Q207D, a constitutively active form of the receptor that is related to ALK2 R206H found in individuals with fibrodysplasia ossificans progressiva. The data indicate that RAR-g agonists are potent inhibitors of heterotopic ossification in mouse models and, thus, may also be effective against injury-induced and congenital heterotopic ossification in humans.

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

A r t i c l e s

nAture medicine VOLUME 17 | NUMBER 4 | APRIL 2011 455

NRX204647; inhibition was greater than that elicited by comparable doses of all-trans-retinoic acid (Fig. 1a).

To test the relevance of RAR-γ in regulating chondrogenesis, we prepared micromass cultures with limb mesenchymal cells from E11.5 RAR-γ–null embryos that still expressed RAR-α and RAR-β (ref. 20). Treatment with all-trans-retinoic acid failed to inhibit chondrogenesis in these RAR-γ–null cultures (Fig. 1b). To verify the apparent centrality of RAR-γ, we prepared micromass cultures with limb mesenchymal cells isolated from E11.5 double RAR-α–null, RAR-β–null embryos (which still expressed RAR-γ) and treated them with all-trans-retinoic acid (for analysis of conditional gene deletion and genotyping, see Supplementary Fig. 1). Chondrogenesis was fully suppressed by all-trans-retinoic acid in these double-mutant cultures as well as in com-panion wild-type cultures (Fig. 1c). Thus, RAR-γ agonists can block chondrogenic differentiation, and RAR-γ seems to exert a strong role in the control of chondrogenesis.

RAR-g agonists block trauma-induced heterotopic ossificationTo determine whether RAR-γ agonists inhibit heterotopic ossification, we used a standard mouse model that combines local trauma and implantation of a skeletogenic factor22. In such models14, progenitor cells are recruited to the injured site and form cartilage by day 5 or 6; cartilage undergoes hypertrophy and is replaced by endochondral bone by days 10–14. We created a small microsurgical pouch in the calf muscles of 2-month-old female mice and inserted a collagen sponge

containing 1 µg of recombinant bone morphogenetic protein-2 (BMP-2) as a skeletogenic factor. Starting on day 1, mice received daily doses of NRX204647 (1.2 mg per kg body weight per day), all-trans-retinoic acid (12 mg per kg body weight per day) or corn oil (vehicle control) by gavage for 12–14 d. Large ectopic mineralized tissue masses formed in vehicle-treated mice at the operated site near the tibia and fibula by day 14 (Fig. 1d), but formation of such tissues was markedly reduced in mice treated with all-trans-retinoic acid and essentially prevented in mice treated with the selective RAR-γ agonist (Fig. 1d). Staining with Masson’s trichrome and alcian blue showed that ectopic masses in control mice contained abundant cartilage, endochondral bone and marrow (Fig. 1d) and proliferative cells (Supplementary Fig. 2), but all these tissues were diminished in mice treated with all-trans- retinoic acid and were essentially absent in mice treated with the selec-tive RAR-γ agonist (Fig. 1d); the number of proliferative cells was also lower (Supplementary Fig. 2). Consistent with these results, amounts of the bone markers osteocalcin and tartrate-resistant alkaline phos-phatase (TRAP) were prominent in control ectopic tissues, were clearly reduced in mice treated with all-trans-retinoic acid and were unde-tectable in mice treated with the selective RAR-γ agonist (Fig. 1d). We obtained identical results in a model of subcutaneous heterotopic ossification (Fig. 1e) used previously14.

Because the drugs are given systemically, they could have unwanted side effects. However, we found no obvious changes in food intake and blood chemistry (Supplementary Fig. 3) or in trabecular

a

e

0 nM Vehicle

Vehicle RA NRX204647

d3 nM 10 nM 30 nM b RA

None

RA

WT γ-null c NRX204647

None

RA

Cont αβ-def

RA

NRX204647

MT

ABAB

OC/MHC

TRAP

µCT

µCT

Figure 1 RAR agonists block chondrogenesis and intramuscular rBMP- 2-driven heterotopic ossification. (a) Micromass cultures of mesenchymal cells from E11.5 mouse limb treated with increasing doses of all-trans-retinoic acid (RA) or the RAR-γ agonist NRX204647 and stained with alcian blue on day 8. (b) Similar micromass cultures prepared from E11.5 RAR-γ–null or wild-type (WT) limb mesenchymal cells and treated with all-trans-retinoic acid (100 nM) for 8 d. (c) Micromass cultures prepared from dual RAR-α– and RAR-β–deficient (αβ-def) or wild-type limb mesenchymal cells and treated with all-trans-retinoic acid as above. (d) Heterotopic ossification masses induced by implantation of rBMP-2-filled collagen sponge into a microsurgically created pocket in the calf muscles. The mice were treated with vehicle, all-trans-retinoic acid (12 mg per kg body weight per day) or NRX204647 (1.2 mg per kg body weight per day) by gavage. Large round mineralized heterotopic ossification masses were visible by µCT in mice that received vehicle, markedly decreased in mice treated with retinoic acid and absent in NRX204647-treated mice. Ectopic tissues were sectioned and examined by Masson trichrome (MT) and alcian blue (AB) staining, immunofluorescence for myosin heavy chain (MHC: red) and osteocalcin (OC: green) and TRAP staining. Scale bar for µCT, 5 mm; for histology, 200 µm. (e) Heterotopic ossification tissue induced by subcutaneous implantation of rBMP-2/Matrigel mixture. Treatment and analysis of ectopic tissue were as above. Scale bar for µCT, 5 mm; for alcian blue, 2.5 mm.

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

A r t i c l e s

456 VOLUME 17 | NUMBER 4 | APRIL 2011 nAture medicine

bone density (Supplementary Fig. 4). The mice that received the highest does of all-trans-retinoic acid (12 mg per kg body weight per day) showed skin redness and loss of some whiskers, and those that received the highest dose of NRX204647 (1.2 mg per kg body weight per day) showed a transient delay in long-bone fracture repair (Supplementary Fig. 3).

Drug class effect of RAR-g agonistsSynthetic RAR-γ agonists differ in their chemical and physical char-acteristics, backbone structure and effectiveness21,23–25. We compared the ability of NRX204647 to inhibit heterotopic ossification with that of the RAR-γ agonists CD1530 and R667 (Fig. 2a) to determine whether inhibition of heterotopic ossification is a drug class effect. We implanted 2-month-old mice with 1 µg recombinant BMP-2 (rBMP-2) in 250 µl Matrigel subcutaneously to induce heterotopic ossification14 and administered the indicated daily doses of NRX204647, CD1530 (ref. 26) or R667 (ref. 27) by gavage. For comparison, companion groups of mice received daily doses of: corn oil (vehicle control), retinol (the largely inert precursor of natural active retinoids)28, all-trans-retinoic acid, 13-cis-retinoic acid, which has been tested for the prevention of heterotopic ossification in patients with FOP29, or the RAR-α agonist NRX195183, which has previously been tested in a mouse model of heterotopic ossification14. Large subcutaneous ectopic masses of vascularized and mineralized tissues formed in control mice, but their formation was dose-dependently inhibited

by each RAR-γ agonist tested, as indicated by soft X-ray, histology and microcomputed tomography (µCT)-based quantification of bone volume/total volume (BV/TV) ratios (Fig. 2b,c). Because all of the RAR-γ agonists were effective, we used them interchangeably in subsequent experiments. In line with the above data and previous work, the pan-agonist all-trans-retinoic acid and the RAR-α agonist NRX185183 also inhibited heterotopic ossification, but their effective-ness was partial and required high doses30 (Fig. 2b). Retinol had no effect, but 13-cis-retinoic acid had a slight and consistent stimulatory effect (Fig. 2b), which might account for the occurrence of flare ups in certain patients with FOP that were treated with it29.

To verify the specificity of RAR-γ agonist action, we induced subcutaneous heterotopic ossification in RAR-γ–null mice, mice heterozygous for the gene encoding RAR-γ and wild-type mice and treated them with CD1530 or NRX204647 for 12–14 d. The agonists blocked heterotopic ossification in wild-type and heterozygous mice as expected, but not in RAR-γ–null mice (Fig. 2d).

To determine whether heterotopic ossification would restart after drug treatment ended (a phenomenon referred to ‘rebound effect’), we induced subcutaneous heterotopic ossification and treated the mice with CD1530 or corn oil (control) for 10 d. A few control and treated mice were killed at that point, and the remain-ing mice were withdrawn from drug treatment and examined over time. Ectopic tissues grew in control mice, but only minimally in mice treated with CD1530, indicating that there was no rebound

13-cis-retinoic acid(pan-agonist)

NRX195183(RAR-α agonist)

CD1530(RAR-γ agonist)

R667(RAR-γ agonist)

NRX204647(RAR-γ agonist)

a 200

Mea

n of

BV

/TV

(%

)

150

100

Contro

l

50

1.2–36 1.2–36 1.2–36 4–36 1.2–12 0.12–1.2 0.12–4

Retinol 13-cis NRX195183

CD1530 NRX204647

R667

* * * * * * *

RA

0

b

Vehicle CD1530c d 0.075

BV

/TV

0.050

0.025

0WT WT HT KO

CD1530

Contro

l

**

0 10 20

Time (d)

0

e 0.15Control

NRX195183

CD15300.10

0.05

BV

/TV

f 0.075 Control

NRX195183

CD1530

0 6 12

Time (d)

0.050

0.025

0

BV

/TV

Figure 2 Effectiveness of different retinoids against heterotopic ossification. (a) Chemical structures of synthetic RAR-γ agonists and natural 13-cis-retinoic acid. (b) Suppression of heterotopic ossification by retinoids evaluated by measuring BV/TV in ectopic masses in control versus retinoid-treated mice by µCT on day 12. Values from control groups were set at 100% and used to calculate relative values in experimental groups. Each group had ≥ 8 samples; data shown as means ± s.e.m. (*P < 0.01 versus control). (c) Macro views (top), soft X-ray radiograms (middle) and histological sections stained with H&E (bottom) of ectopic masses collected from mice treated with vehicle or CD1530. Scale bar in top and middle images, 1 cm; bottom images, 100 µm. (d) Selectivity of RAR-γ agonist action. Subcutaneous heterotopic ossification was triggered in wild-type, heterozygous RAR-γ (HT) and RAR-γ–null (KO) mice, and mice were treated with RAR-γ agonist CD1530 (4 mg per kg body weight per day) or vehicle for 12 d. *P < 0.01 versus control. (e) Evaluation of rebound effects. Mice implanted with rBMP-2–Matrigel mixture subcutaneously were treated with vehicle, CD1530 or RAR-α agonist NRX195183 for 10 days. Treatment was stopped, and heterotopic ossification was evaluated by µCT at later time points. (f) Window of opportunity tests. Mice implanted as above were left untreated up to day 6 and were then treated with CD1530 or NRX195183 until day 12.

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

A r t i c l e s

nAture medicine VOLUME 17 | NUMBER 4 | APRIL 2011 457

effect (Fig. 2e). We tested the RAR-α agonist NRX195183 and observed a rebound effect in this case (Fig. 2e).

Last, we tested whether the RAR-γ agonists would be effective against heterotopic ossification even if treatment did not start imme-diately, providing insights into the ‘window of opportunity’. We started treatment with CD1530 or corn oil on day 6 (corresponding to the chondrogenic phase of heterotopic ossification) and continued it for 6–8 d. The RAR-γ agonist was still able to block heterotopic ossifi-cation (Fig. 2f). However, when treatment was delayed until day 12 (corresponding to the osteogenic phase), the RAR-γ agonist had little or no effect (Supplementary Fig. 5). Together, the data indicate that the agonists block heterotopic ossification by preventing chondrogenic cell differentiation and cartilage formation but have minimal effects once bone has formed.

Prevention of heterotopic ossification in ALK2 Q207D miceCongenital heterotopic ossification in individuals with FOP is driven by the ALK2 R206H mutant protein8, which has mild basal constitutive activity but becomes hyperactive upon ligand binding31. To determine whether RAR-γ agonists could be effective against congenital heterotopic ossification, we prepared an expression vector encoding the constitutively active ALK2 Q207D mutant, which induces chondrogenesis and endochondral ossification in chicks12, and expressed it in ATDC5 chondrogenic cells along with the BMP signaling reporter plasmid Id1-luc32. We treated the cells with dif-ferent doses of CD1530 and measured luciferase activity after 24 h. Reporter activity increased markedly in ALK2 Q207D–expressing cells compared to empty vector control cells, but this increase was inhibited by CD1530 treatment (Fig. 3a).

To test the effectiveness of CD1530 in vivo, we used conditional ALK2 Q207D–transgenic mice in which the effectiveness of a BMP type I receptor kinase inhibitor against FOP-like heterotopic ossifica-tion has been tested13. We injected mice intramuscularly with adeno-Cre in a hind limb and then treated them with CD1530 or vehicle. Massive heterotopic ossification developed in vehicle-treated mice, but heterotopic ossification was essentially absent in CD1530-treated mice (Fig. 3b,c). To verify these data, we injected mice with alizarin complexon before they were killed. Tissue from mice treated with vehicle showed abundant alizarin complexon staining, but there was no obvious alizarin complexon–positive tissue in CD1530-treated mice (Fig. 3d).

RAR-g agonists reduce BMP signalingTo clarify the mechanisms by which the RAR-γ agonists inhibit het-erotopic ossification, we co-transfected ADTC5 cell cultures with the Id1-1uc reporter plasmid and treated them with various CD1530 concentrations (0–100 nM) in the absence or presence of exogenous rBMP-2 (100 ng ml−1) for 24 h. Luciferase activity was stimulated more than tenfold by rBMP-2 treatment, but this increase was counteracted by simultaneous treatment with CD1530 (Fig. 4a) or NRX204647 (data not shown). BMPs transmit signals by transient phosphorylation of Smad1, Smad5 and Smad8 proteins (p-Smads) (ref. 33). Immunoblot analysis with antibodies to p-Smad1, 5 and 8 showed that in control cells Smad phosphorylation was already strong after 30 min of BMP-2 treatment (100 ng ml−1), peaked at 1 h and returned to basal levels by 3 h (Fig. 4b). Concurrent treatment with CD1530 reduced the amounts of p-Smad proteins by more than 80%, as estimated by imaging quanti-fication and normalization to α-tubulin amounts (Fig. 4b). Treatment with CD1530 could have provoked such a drop in p-Smad abundance by affecting mechanisms involved in signal transduction and Smad phosphorylation or by reducing the overall steady-state levels of Smad proteins. To test the latter possibility, we processed identical immu-noblots with antibodies to Smad1. There was a marked reduction in overall Smad1 abundance in CD1530-treated cells, whereas these levels remained essentially unchanged in companion control cells (Fig. 4b). The decreases in Smad1 were appreciable by immunofluorescence staining (Fig. 4c). In control cells, Smad1 was largely if not exclusively confined to the cytoplasm, but as expected it relocated to the nucleus upon treatment with rBMP-2 (Fig. 4c). In cells treated with CD1530 for 24 h (in the absence or presence of rBMP-2), there was little if any Smad1 in the cytoplasm or nucleus (Fig. 4c). In addition, there were similar decreases in the overall amounts of Smad5 and regulatory Smad4 in CD1530-treated cells (Fig. 4d).

We wondered whether the reductions in Smad protein resulted from shunting toward degradation34. To test this idea, we treated cells with 1 µM CD1530 in the absence or presence of proteasome inhibitors AW9155 or PI-108 for 6 h and probed the resulting cell lysates by immunoblotting. Whereas treatment with CD1530 led

a

d

b

ALK2 Q207D + Ad.Cre

Vehicle CD1530

ALK2 Q207D + Ad.Cre

Vehicle CD1530

P7

P21

P35

ALK2 Q207D

No virus Vehicle CD1530

ALK2 Q207D + Ad.Cre

BF

GFP

AR

c

40

30

20

10

Control

CD1530

ALK2 Q207D0

Rel

ativ

e Id

rep

orte

r ac

tivity

0 100 0 100 (nM)

Figure 3 RAR-γ agonists block FOP-like heterotopic ossification. (a) ADTC5 cells expressing the strong constitutively active ALK2 Q207D or control empty vector were transfected with the BMP signaling reporter Id1-luc and then treated with 0, 10, 30 or 100 nM CD1530. Reporter activity was normalized to phRG-TK encoding Renilla luciferase. Error bars represent s.d. for triplicate samples. (b) ALK2 Q207D–transgenic mouse model of FOP-like heterotopic ossification. Consecutive soft X-ray images of the same mice, injected with Ad-Cre and cardiotoxin, taken at P7, P21 and P35 showing massive heterotopic ossification in vehicle-treated mice (double arrowheads), but markedly less in companion mice treated with CD1530 (4.0 mg per kg body weight per day). Scale bar, 1 cm. (c) µCT images showing massive heterotopic ossification in vehicle-treated ALK2 Q207D–transgenic mice injected with Ad-Cre and cardiotoxin (double arrowheads) but not in companion CD1530-treated mice. Scale bar, 1 cm. (d) Bright-field (BF) and fluorescent images showing ectopic skeletal tissue (positive for alizarin complexon, AR) in vehicle-treated ALK2 Q207D–transgenic mice injected with Ad-Cre and cardiotoxin but not in those treated with CD1530. Positive GFP fluorescence confirmed that Ad-Cre had activated transgene expression in all mice. There was minimal GFP and AR fluorescence in mice that had not been injected with Ad-Cre and cardiotoxin. Scale bar, 5 mm.

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

A r t i c l e s

458 VOLUME 17 | NUMBER 4 | APRIL 2011 nAture medicine

to the expected drop in Smad1, this response was mitigated by treatment with either proteasome inhibitor (Fig. 4e).

An RAR-g agonist alters cell differentiation potentialThe marked reduction in steady-state amounts of Smad1, Smad4 and Smad5 caused by RAR-γ agonists suggested that the progenitor cells could have lost their lineage assignment and skeletogenic potentials. To test this possibility, we treated ADTC5 cells in monolayer with 1 µM CD1530 or vehicle for 2 d. Control and CD1530-treated cells

were then detached, rinsed, reseeded in multiwell cultures, grown in the absence or presence of 100 ng ml−1 rBMP-2 for an additional 7 d (without further CD1530 treatment) and assessed for chondro-genic differentiation. As expected, control cells underwent differ-entiation upon rBMP-2 treatment and stained strongly with alcian blue (a measure of chondrocyte differentiation and cartilage matrix accumulation) and for alkaline phosphatase activity (a measure of chondrocyte maturation; Fig. 5a,b). However, CD1530-pretreated cells did not (Fig. 5a,b). To verify these data, we repeated the

60a c

e

dbp-Smad1,5,8

Smad1

0 1 3 24 (h)

(h)

0 1 3 24

DMSO

Smad1

Smad4

Smad3

Smad5

BMP-2 (100 ng ml–1)

CD1530 (1 µM)BMP-2 (100 ng ml–1)

CD1530 (1 µM)

AW9155 (1 µM)

Pl-108 (1 µM)

– –

––

+

+

+

+

–

– –

– – – – –

– – –

–+ + +

+

+ + + + +

+

++

CD1530 (1 µM)

0 1 30.5DMSO CD1530 (1 µM)

BMP-2 (100 ng ml–1)

BMP-2 (100 ng ml–1)

0 1 30.5

α-tubulin

α-tubulin

Rel

ativ

e Id

repo

rter

act

ivity

40

20

0

0 0100CD1530

Control

Control

α-tubulin

α-tubulin

α-tubulin

α-tubulin

α-tubulinSmad1

Mer

geS

mad

1

BMP-2CD1530BMP-2

BMP-2(100 ng ml–1)

100 (nM)

Figure 4 Mechanisms of RAR-γ agonist action. (a) Id1-luc reporter activity in ATDC5 cells. Error bars represent s.d. for triplicate samples. (b) Immunoblots showing phosphorylated Smad1, Smad5 and Smad8 (p-Smad1,5,8) protein amounts in control (DMSO) ATDC5 cells treated with rBMP-2 and in cells treated with CD1530 (top) and overall Smad1 amounts in control cells and in CD1530-treated cells (bottom). Membranes

a

µCT

H&E

GFP

OC

Merge

ControlCD1530PretreatfNone

Control

CD1530Pretreat

BMP-2

Control

CD1530Pretreat

None BMP-2

Rel

ativ

e Id

repo

rter

act

ivity

15

10

5

BMP-2CD1530

0– + – +– – + +

Smad1

CD1530 CD1530+– +–

α–tubulin α–tubulin

Smad4

Control

CD1530Pretreat

None BMP-2

b

c

d

e

Figure 5 RAR-γ agonists reprogram the differentiation potentials of skeletal progenitor cells. (a,b) ATDC5 cells were grown in the presence of vehicle (control) or 1 µM CD1530 for 2 d, rinsed, replated and then grown for an additional 7 d with or without 100 ng ml−1 rBMP-2. Control cells responded well to rBMP-2 and underwent chondrogenic differentiation revealed by strong alcian blue (a) and alkaline phosphatase (b) staining, but the cells pretreated with CD1530 did not and failed to stain. (c) Similar experiments with GFP-expressing MSCs derived from mouse bone marrow show that the cells underwent differentiation upon rBMP-2 treatment (top) but CD1530 pretreated MSCs did not and failed to stain with alizarin red (bottom). (d) Immunoblots showing the steady-state amounts of Smad1 and Smad4 in MSCs treated without (−) or with (+) 1 µM CD1530 for 12 h. (e) Id1-luc reporter activity in control MSCs treated with or without rBMP-2 in the presence or absence of CD1530. Error bars represent s.d. for four wells. (f) Control and CD1530-pretreated GFP-expressing MSCs were mixed with rBMP-2– Matrigel and implanted in nude mice subcutaneously, and ectopic tissue masses were analyzed by µCT, H&E staining, osteocalcin (OC) immunostaining and GFP fluorescence signal. Bottom, merged GFP and OC images. Scale bars, 5 mm (top row); 250 µm (second row); 150 µm (third row).

were reblotted with α-tubulin–specific antibodies for normalization. (c) Dual Smad1 and β-actin immunofluorescence staining and nuclear DAPI staining showing that Smad1 was largely cytoplasmic in control cells and relocated to the nucleus during rBMP-2 treatment, but there was minimal detectable Smad1 signal in cytoplasm or nucleus of cells concurrently treated with CD1530. Red, Smad1; green, β-actin; blue, nuclei. (d) Immunoblots similar to those in b showing that the overall amounts of Smad1, Smad4 and Smad5 with or without CD1530 and rBMP-2 treatment. (e) Immunoblots showing that the decreases in Smad1 levels elicited by CD1530 treatment were counteracted by co-treatment with the proteasome inhibitors AW9155 or PI-108.

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

A r t i c l e s

nAture medicine VOLUME 17 | NUMBER 4 | APRIL 2011 459

experiments with bone marrow–derived mesenchymal stem cells (MSCs) isolated from a reporter mouse line35. Whereas control cells treated with rBMP-2 underwent differentiation and stained strongly with alizarin red (Fig. 5c) and for alkaline phosphatase (data not shown), CD1530-pretreated cells did not (Fig. 5c). In ADTC5 cells (Fig. 4d), CD1530 treatment lowered steady-state levels of Smad proteins and counteracted the response to rBMP-2, as indicated by Id1-luc activity (Fig. 5d,e).

To test whether a similar loss of skeletogenic potential occurred in vivo, the bone marrow–derived MSCs were first grown for 2–3 d in monolayer in the presence or absence of 100 nM CD1530. The cells were then detached, mixed with Matrigel–rBMP-2 and implanted subcutaneously in nude mice. Because the cells express GFP, they were readily distinguishable from host cells. Control cells produced conspicuous masses of endochondral bone by 2 weeks that were obvious by µCT and histology and contained numerous GFP- and osteocalcin-positive cells (Fig. 5f). However, the CD1530-pretreated cells largely did not produce ectopic bone, although they were present in large numbers, as revealed by their strong GFP signal (Fig. 5f). Bone volume/total volume quantification showed that the reduction in ectopic bone formation was more than 95%.

DISCUSSIONHere we have shown that RAR-γ agonists are potent inhibitors of intramuscular and subcutaneous heterotopic ossification and can even mitigate ectopic skeletal tissue formation induced by the constitutively active ALK2 Q207D mutation. The inhibition of hetero-topic ossification seems to be largely irreversible, because there were minimal rebound effects when we stopped drug treatment; in addi-tion, the drugs seem to have minimal side effects. As chondrogen-esis normally requires a drop in retinoid signaling and a concurrent upregulation of pro-chondrogenic pathways including BMP signal-ing15,36, it is likely that the RAR-γ agonists inhibit chondrogenesis and heterotopic ossification by maintaining active retinoid signal-ing and liganded RAR activity while dampening BMP signaling. Our data also indicate that the latter effect involves not only inhibition of Smad phosphorylation, but also a marked proteasome-mediated drop in overall Smad abundance and a concurrent reprogramming of the progenitor cells into a nonskeletogenic lineage. A recent study of ALK2 Q207D–transgenic mice found no substantial decrease in overall Smad levels after treatment of skeletogenic cells with BMP type I receptor kinase inhibitors13, indicating that the RAR-γ agonists use distinct and broader modes of action and might therefore be more potent. Liganded RAR-γ has been shown to interact directly with Smad proteins and to inhibit their function37, and liganded RAR-γ increases Wnt–β-catenin signaling38. Because Wnt–β-catenin sig-naling is a potent inhibitor of chondrogenesis39,40, it follows that the RAR-γ agonists probably block chondrogenesis and heterotopic ossification by reciprocally increasing Wnt–β-catenin signaling while decreasing signaling through BMP and possibly other pathways41. In summary, the RAR-γ agonists have the biological properties needed to interfere with the specific processes and mechanisms that are needed for the initiation and progression of heterotopic ossifica-tion and might therefore represent effective remedies, probably the most effective reported so far, for this condition and related ectopic ossification conditions42,43.

Several notes of caution must be considered before retinoid ago-nists can be tested in humans. More in-depth studies need to be carried out to exclude side effects. The transient delay in fracture repair seen in mice treated with RAR-γ agonists could be of concern

for patients such as wounded soldiers who may have severe bone fractures. However, these fractures are often subjected to stabilization and traction before healing and as inflammation subsides44, which could provide a window of opportunity for retinoid treatment and the prevention of heterotopic ossification. The agonists could also have negative effects on the growth plate in developing skeletal ele-ments, either directly or through excess Wnt–β-catenin signaling. If so, this would be a concern for long-term treatment of children with FOP. However, because the RAR-γ agonists act rapidly and at least in part by diverting the skeletogenic potentials of progenitor cells, it is possible that treatment could be short or intermittent, which would allow the growth plate to recover in between treatments. Lastly, although all the RAR-γ agonists tested here are very effective, we need to identify the leading candidate for clinical development. Given its particularly strong activity20, NRX204647 should be at the top of the list, but this drug has not been tested in humans. In contrast, R667 (also known as Palovarotene) is already in a phase 2 clinical trial for emphysema27 and could therefore move toward a clinical trial for heterotopic ossification in an efficient manner.

METHODSMethods and any associated references are available in the online version of the paper at http://www.nature.com/naturemedicine/.

Note: Supplementary information is available on the Nature Medicine website.

ACKnoWlEDgMEnTSWe thank T. Katagiri (Saitama) for the Id1-luc reporter plasmid; S. Otsuru and E. Horwitz (Children’s Hospital of Philadelphia) for the GFP-expressing MSCs and culture methods; and N. Ghyselinck and P. Chambon (Institut National de la Santé et de la Recherche Médicale, Illkirch Cedex) for the RAR floxed and null mouse lines. This work was supported by contract no. W81XWH-07-1-0212 from the Department of the Army, United States Army Medical Research Acquisition Activity (M.P.) and US National Institutes of Health RO1 grant AR056837 (M.I. and M.P.).

AUTHoR ConTRIBUTIonSM.I. and M.P. directed the project. K.S., W.T., C.M., A.H.-T.C., J.H.D., C.M. and M.E.-I. performed experiments, analyzed data and participated in experimental design. R.A.C. provided expertise on retinoid biology. Y.M. provided expertise in mouse genetics. M.P., K.S. and M.I. wrote the manuscript.

CoMPETIng FInAnCIAl InTERESTSThe authors declare no competing financial interests.

Published online at http://www.nature.com/naturemedicine/. Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/.

1. Chalmers, J., Gray, D.H. & Rush, J. Observation on the induction of bone in soft tissues. J. Bone Joint Surg. Br. 57, 36–45 (1975).

2. Garland, D.E. A clinical prospective on common forms of acquired heterotopic ossification. Clin. Orthop. Relat. Res. 263, 13–29 (1991).

3. Nilsson, O.S. & Persson, P.E. Heterotopic bone formation after joint replacement. Curr. Opin. Rheumatol. 11, 127–131 (1999).

4. Forsberg, J.A. et al. Heterotopic ossification in high-energy wartime extremity injuries: prevalence and risk factors. J. Bone Joint Surg. Am. 91, 1084–1091 (2009).

5. McCarthy, E.F. & Sundaram, M. Heterotopic ossification: a review. Skeletal Radiol. 34, 609–619 (2005).

6. Vanden Bossche, L. & Vanderstraeten, G. Heterotopic ossification: a review. J. Rehabil. Med. 37, 129–136 (2005).

7. Pignolo, R.J., Suda, R.K. & Kaplan, F.S. The fibrodysplasia ossificans progressiva lesion. Clin. Rev. Bone Min. Metab. 3, 195–200 (2005).

8. Shore, E.M. et al. A recurrent mutation in the BMP type I receptor ACVR1 causes inherited and sporadic fibrodysplasia ossificans progressiva. Nat. Genet. 38, 525–527 (2006).

9. Chao, S.T., Joyce, M.J. & Suh, J.H. Treatment of heterotopic ossification. Orthopedics 30, 457–464 (2007).

10. Kaplan, F.S. The medical management of fibrodysplasia ossificans progressiva: current treatment considerations. Clin. Proc. Intl. Clin. Consort. FOP 3, 1082 (2008).

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

A r t i c l e s

460 VOLUME 17 | NUMBER 4 | APRIL 2011 nAture medicine

11. Fukuda, T. et al. Generation of a mouse with conditionally activated signaling through the BMP receptor, ALK2. Genesis 44, 159–167 (2006).

12. Zhang, D. et al. ALK2 functions as a BMP type I receptor and induces Indian hedgehog in chondrocytes during skeletal development. J. Bone Miner. Res. 18, 1593–1604 (2003).

13. Yu, P.B. et al. BMP type I receptor inhibition reduces heterotopic ossification. Nat. Med. 14, 1363–1369 (2008).

14. Shimono, K. et al. Inhibition of ectopic bone formation by a selective retinoic acid receptor α-agonist: a new therapy for heterotopic ossification? J. Orthop. Res. 28, 271–277 (2010).

15. Pacifici, M., Cossu, G., Molinaro, M. & Tato, F. Vitamin A inhibits chondrogenesis but not myogenesis. Exp. Cell Res. 129, 469–474 (1980).

16. Weston, A.D., Chandraratna, R.A.S., Torchia, J. & Underhill, T.M. Requirement for RAR-mediated gene repression in skeletal progenitor differentiation. J. Cell Biol. 158, 39–51 (2002).

17. Weston, A.D., Hoffman, L.M. & Underhill, T.M. Revisiting the role of retinoid signaling in skeletal development. Birth Defects Res. C. Embryo Today 69, 156–173 (2003).

18. Rochette-Egly, C. & Germain, P. Dynamic and combinatorial control of gene expression by nuclear retinoic acid receptors (RARs). Nucl. Recept. Signal. 7, e005 (2009).

19. Koyama, E. et al. Retinoid signaling is required for chondrocyte maturation and endochondral bone formation during limb skeletogenesis. Dev. Biol. 208, 375–391 (1999).

20. Williams, J.A. et al. Retinoic acid receptors are required for skeletal growth, matrix homeostasis and growth plate function in postnatal mouse. Dev. Biol. 328, 315–327 (2009).

21. Thacher, S.M., Vasudevan, J. & Chandraratna, R.A.S. Therapeutic applications for ligands of retinoic receptors. Curr. Pharm. Des. 6, 25–58 (2000).

22. O’Connor, J.P. Animal models of heterotopic ossification. Clin. Orthop. Relat. Res. 346, 71–80 (1998).

23. Géhin, M. et al. Structural basis for engineering of retinoic acid receptor isotype-selective agonists and antagonists. Chem. Biol. 6, 519–529 (1999).

24. Keidel, S., LeMotte, P. & Apfel, C. Different agonist- and antagonist-induced conformational changes in retinoic acid receptors analyzed by protease mapping. Mol. Cell Biol. 14, 287–298 (1994).

25. Klein, E.S. et al. Identification and functional separation of retinoic acid receptor neutral antagonists and inverse agonists. J. Biol. Chem. 271, 22692–22696 (1996).

26. Bernard, B.A. et al. Identification of synthetic retinoids for selectivity for human nuclear retinoic acid receptor γ. Biochem. Biophys. Res. Commun. 186, 977–983 (1992).

27. Hind, M. & Stinchcombe, S. Palovarotene, a novel retinoic acid receptor gamma agonist for the treatment of emphysema. Curr. Opin. Investig. Drugs 10, 1243–1250 (2009).

28. Blaner, W.S. & Olson, J.A. in The Retinoids: Biology, Chemistry, and Medicine (eds. Sporn, M.B., Roberts, A.B. & Goodman, D.S.) 229–255 (Raven, 1994).

29. Zasloff, M.A., Rocke, D.M., Crofford, L.J., Hahn, G.V. & Kaplan, F.S. Treatment of patients who have fibrodysplasia ossificans progressiva with isotretinoin. Clin. Orthop. Relat. Res. 346, 121–129 (1998).

30. DeSimone, D.P. & Reddi, A.H. Influence of vitamin A on matrix-induced endochondral bone formation. Calcif. Tissue Int. 35, 732–739 (1983).

31. Shen, Q. et al. The fibrodysplasia ossificans progressiva R206H ACVRa mutation activates BMP-independent chondrogenesis and zebrafish embryo ventralization. J. Clin. Invest. 119, 3462–3472 (2009).

32. Korchynskyi, O. & ten Dijke, P. Identification and functional characterization of distinct critically important bone morphogenetic protein-specific response elements in the Id1 promoter. J. Biol. Chem. 277, 4883–4891 (2002).

33. Massagué, J. TGF-β signal transduction. Annu. Rev. Biochem. 67, 753–791 (1998).

34. Derynck, R. & Zhang, Y.E. Smad-dependent and Smad-independent pathways in the TGF-β family signaling. Nature 425, 577–584 (2003).

35. Otsuru, S. et al. Osteopoietic engraftment after bone marrow transplantation: effect of imbred strain of mice. Exp. Hematol. 38, 836–844 (2010).

36. Hoffman, L.M. et al. BMP action in skeletogenesis involves attenuation of retinoid signaling. J. Cell Biol. 174, 101–113 (2006).

37. Pendaries, V., Verrecchia, F., Michel, S. & Mauviel, A. Retinoic acid receptors interfere with the TGF-β/Smad signaling pathway in a ligand-specific manner. Oncogene 22, 8212–8220 (2003).

38. Yasuhara, R. et al. Wnt/β-catenin and retinoic acid receptor signaling pathways interact to regulate chondrocyte function and matrix turnover. J. Biol. Chem. 285, 317–327 (2010).

39. Day, T.F., Guo, X., Garrett-Beal, L. & Yang, Y. Wnt/β-catenin signaling in mesenchymal progenitors controls osteoblast and chondrocyte differentiation during vertebrate skeletogenesis. Dev. Cell 8, 739–750 (2005).

40. Tamamura, Y. et al. Developmental regulation of Wnt/β-catenin signals is required for growth plate assembly, cartilage integrity, and endochondral ossification. J. Biol. Chem. 280, 19185–19195 (2005).

41. Nozaki, Y. et al. Anti-inflammatory effect of all-trans-retinoic acid in inflammatory arthritis. Clin. Immunol. 119, 272–279 (2006).

42. Benglis, D., Wang, M.Y. & Levi, A.D. A comprehensive review of the safety profile of bone morphogenetic protein in spine surgery. Neurosurgery 62, ONS423–ONS431 (2008).

43. Shao, J.-S., Cai, J. & Towler, D.A. Molecular mechanisms of vascular calcification: lessons learned from the aorta. Arterioscler. Thromb. Vasc. Biol. 26, 1423–1430 (2006).

44. Bone, L.B., Johnson, K.D., Weigelt, J. & Scheinberg, R. Early versus delayed stabilization of femural fractures. A prospective randomized study. J. Bone Joint Surg. Am. 71, 336–340 (1989).

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

nAture medicinedoi:10.1038/nm.2334

ONLINE METHODSIn vitro chondrogenesis. Limb bud mesenchymal cells isolated from E11.5 wild-type, compound RAR-α, RAR-β–deficient or RAR-γ–null mouse embryos20 were seeded in high-density micromass culture and maintained in DMEM and F12 medium (1:1) containing 3% FBS as previously described15. Cultures were treated with the indicated concentrations (Fig. 1a) of all-trans-retinoic acid (Sigma) or RAR-γ agonist NRX 204647 (NuRx Pharmaceuticals); fresh drugs were given with medium change every other day. Parallel control cultures received equal volume of DMSO (vehicle). Cultures were stained with alcian blue (pH 1) to monitor chondrogenic cell differentiation. Floxed RAR-α and RAR-β mice were mated with Prx1-Cre mice to create the compound RAR-α, RAR-β-deficient limbs. Genotyping procedures and results are shown in Supplementary Figure 1.

Mouse models of heterotopic ossification. The standard muscle trauma– associated heterotopic ossification model was as previously described22,45. Briefly, 3 mm × 1 mm–thick collagen discs were prepared by punched out of collagen sponge sheets (Ultraform, Davol). A 5 microliter aliquot containing 1 µg of recombinant human BMP-2 (GenScript) was adsorbed onto each disc. Longitudinal skin incisions were made on the medial surface of calf muscles in 2-month-old CD-1 female mice. An intramuscular pocket was created microsurgi-cally, and one rBMP-2–collagen disc was placed in it; the skin was then closed with 4.0 nylon sutures. The subcutaneous model of heterotopic ossification was created as described14. Only female mice were used because they are more easily handled during drug treatment by gavage and can be kept in larger numbers per cage. For the model of FOP-like heterotopic ossification, we used transgenic mice that carry a Cre-inducible constitutively active ALK2 Q207D mutant and EGFP (CAG-Z-EGFP-caALK2)13. To induce heterotopic ossification, 108 plaque-forming units per 10 µl of Ad-Cre (Vector Biolabs) and 0.3 µg per 10 µl cardiotoxin (Sigma) were injected into the left anterior tibial muscles of P7 ALK2 Q207D mice. To monitor bone formation, 35 µg per kg body weight alizarin complexon (Sigma) was injected intraperitoneally on P29, P31 and P33. GFP and alizarin signals were captured with a fluorescence stereomicroscope (SZX, Olympus). To test the skeletogenic potentials of RAR-γ agonist–treated MSCs in vivo, we cultured bone marrow–derived GFP-expressing MSCs35 in monolayer for 2–3 d in the absence (control) or presence of 100 nM CD1530. Control and CD1530-pretreated GFP-expressing MSCs (2.5 × 105) were then mixed with rBMP-2 and Matrigel and implanted subcutaneously in host nude mice. All surgical procedures were carried out under anesthesia by inhalation of 1.5% isofluorane in 98.5% oxygen. All pro-cedures were approved by the Institutional Animal Care and Use Committee at Thomas Jefferson University.

Systemic retinoid administration. Stock solutions of retinoids prepared in DMSO were stored at −30 °C under argon. Just before administration, 30 µl

of each stock solution was mixed with 70 µl corn oil (Sigma) to produce a 100 µl dose per mouse delivered by gavage. Unless otherwise noted, treatment with retinoids started on day 1 after surgery. Companion control mice received vehicle alone (DMSO in corn oil). With regard to the FOP-like transgenic mice, treatment with vehicle or 4 mg per kg body weight per day of CD1530 was started 3 days after injection of Ad-Cre and cardiotoxin and continued every other day.

Soft X-ray and µCT analyses. Tissue samples were fixed in 4% paraformalde-hyde at 4 °C and subjected to µCT analysis using a CT40 scanner (SCANCO) at 55 kV and 70 mA. Data were analyzed at threshold 244 for detection of min-eralized components and calculation of bone volume and then re-analyzed at threshold 110 to measure total sample volume. Normalized BV/TV ratios were calculated by dividing bone volume by tissue volume and used to calculate dif-ferences between experimental and control values. Soft X-ray images were taken using piXarray100 (Bioptics) in auto-exposure mode.

Blood tests. Blood samples were collected from the heart apex at the moment of death and stored on ice before centrifugation at 1,000g at 4 °C for 10 min to collect serum samples. Samples were then analyzed by VetScan Comprehensive Diagnostic Profile (Abaxis).

Bone fracture healing. Fibulas were transected with a bone cutter in adult mice under anesthesia and allowed to heal without stabilization. Immediately after surgery and at 2, 4 and 8 weeks after surgery, lateral radiographs of limbs were taken to monitor callus formation and bone healing. which were evaluated with a typical fracture healing scoring system: 0, lack of healing; 1, callus formation; 2, onset of bony union; 3, disappearance of fracture line; and 4, complete bony union. Radiographs were scored by two examiners with a randomized blinded selection of films.

Statistical analysis. The statistical significance of all experiments was deter-mined by one-way factorial analysis of variance followed by Bonferroni/Dunn post hoc multiple comparison tests (Prism 5, GraphPad Software.). P values less than 0.01 were considered as statistically significant versus control. All statistical data are presented as means ± s.e.m. (n = 8).

Additional methods. Detailed methodology of histological and immuno-histochemical analyses, reporter assay, protein analysis and immunocyto-chemistry are described in the Supplementary Methods.

45. Kakudo, N., Kausumoto, K., Wang, Y.B., Iguchi, Y. & Ogawa, Y. Immunolocalization of vascular endothelial growth factor on intramuscular ectopic osteoinduction by bone morphogenetic protein-2. Life Sci. 79, 1847–1855 (2006).

© 2

011

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.