Eb

La

b

c

d

a

AA

KCGCOB

1

cpaecrchwtitr

0h

Applied Surface Science 262 (2012) 200– 206

Contents lists available at SciVerse ScienceDirect

Applied Surface Science

jou rn al h om epa g e: www.elsev ier .com/ locate /apsusc

ctopic bone formation cannot occur by hydroxyapatite/�-tricalcium phosphateioceramics in green fluorescent protein chimeric mice

ijia Chenga,1, Xin Duanb,c,1, Zhou Xiangc, Yujun Shia, Xiaofeng Lua, Feng Yea, Hong Bua,d,∗

Key Laboratory of Transplant Engineering and Immunology, Ministry of Health, West China Hospital, Sichuan University, Chengdu, People’s Republic of ChinaDepartment of Orthopaedics, Chengdu Second People’s Hospital, Chengdu, People’s Republic of ChinaDepartment of Orthopaedics, West China Hospital, Sichuan University, Chengdu, People’s Republic of ChinaDepartment of Pathology, West China Hospital, Sichuan University, Chengdu, People’s Republic of China

r t i c l e i n f o

rticle history:vailable online 4 May 2012

eywords:alcium phosphate ceramicsFPhimeric modelsteoinductionM transplantation

a b s t r a c t

Many studies have shown that calcium phosphate ceramics (CP) have osteoconductive and osteoinductiveproperties; however, the exact mechanism of bone induction has not yet been reported. This study wasperformed to investigate if destroying immunological function will influence osteogenesis, to explain themechanism which is unclear. In this study, twenty C57BL/6 mice were divided into two groups (n = 10),in group 1, a hydroxyapatite/�-tricalcium phosphate (HA/�-TCP) ceramic was implanted into both theleft and right leg muscles of each mouse; in group 2, ten mice experienced lethal irradiation, then wereinjected bone marrow (BM) cells from green fluorescent protein (GFP) transgenic mice by tail veil, afterbone marrow transplantation (BMT), heart, liver, spleen, lung, kidney, and muscle were harvested forbiological analysis, after the GFP chimera model was established successfully, the same HA/�-TCP ceramicwas implanted into both leg muscles of each mouse immediately after irradiation. 45 and 90 days afterimplantation, the ceramics of the two groups were harvested to perform with hematoxylin and eosin

(HE) and immunohistochemistry (IHC) staining; the results showed that there was no bone formation ingroup 2, while new bone tissues were detected in group 1. Our findings suggest that the BM cell from GFPtransgenic mice is a good biomarker and it could set a good platform for chimera model; it also showsthat BM cell is one of cell resources of bone induction, and destruction of immune function will impedeosteoinduction by CP. Overall, our results may shed light on clear mechanism study of bone induction in the future.

. Introduction

Bone autograft is the gold standard in some clinical appli-ations [1]. However, it is associated with some limitations:ain, infection, loss of sensibility and hematoma in donor sites well as limited availability [2]. Synthetic materials have beenxtensively studied as substitutes for bone autografts [3]. Cal-ium phosphate ceramics (CP) have been widely used in boneeconstructive surgery due to their osteoconductive and bio-ompatibility properties [4,5]. In addition, these biomaterialsave been demonstrated to be osteoinductive in muscular sitesithout the addition of seed cells or osteogenic growth fac-

ors after 6–12 weeks [6–13]. This phenomenon reflects the

ntrinsic osteoinductivity of CP. Osteoinductivity is consideredo be a clinical advantage for biomaterials used in the boneepair and bone tissue engineering. Although osteoinduction by

∗ Corresponding author. Tel.: +86 028 85164030.E-mail address: hongbu@scu.edu.cn (H. Bu).

1 Co-first author.

169-4332/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2012.04.168

© 2012 Elsevier B.V. All rights reserved.

biomaterials has been widely observed, its mechanism is stilllargely unknown. A proposed mechanism suggests that osteoin-ductive biomolecules, such as bone morphogenetic protein 2(BMP-2), are adsorbed on the surface of CP after implanta-tion, and then these adsorbed biomolecules involve in boneformation, which appears as osteoinduction [10]. Green fluo-rescent protein (GFP) is a 27-kD protein, originally discoveredin jellyfish Aequorea Victoria. Because the fluorescence activ-ity of GFP requires no substrates or cofactors, cells expressingthe intracellular GFP marker can be observed directly by flu-orescent microscopy without the requirement for antibodystaining. Therefore, bone marrow cell of GFP transgenic micewould be helpful and used as a biomarker in bone marrowtransplantation.

The objective of this study was to investigate the mechanisms ofintrinsic osteoinduction of CP. In this study, we compared osteoin-duction by the same calcium phosphate material in the same

genotype mice, but implanted in two different models. The resultsshow that the difference models resulted in significant differencesin osteoinduction, and the mechanism of intrinsic osteoinductionby CP will be analyzed and discussed.

L. Cheng et al. / Applied Surface Sc

F

2

2

sRt5poasow

2

wStpriaibmpi

ig. 1. A SEM micrograph showing the microstructure of porous HA/�-TCP implants.

. Materials and methods

.1. Calcium phosphate samples

Cylindrical hydroxyapatite/� -tricalcium phosphate (HA/�-TCP)amples (Fig. 1) were provided by the National Engineeringesearch Center for Biomaterials (Chengdu, Sichuan, China). Briefly,he sintering was conducted at 1100 ◦C with a heating rate of◦C min−1 and a holding time of 2 h to obtain ceramics. The sam-les were cut into cylinder with a diameter of 3 mm and a heightf 5 mm. The samples contained 60 wt.% HA and 40 wt.% �-TCP,nd the volumetric porosity was approximately 50% and the poreize ranged from 300 to 500 �m. Subsequently, samples were thor-ughly washed with ultrapure water several times. Finally, samplesere dried and autoclaved prior to use.

.2. Chimera

The outline of the whole protocol is presented in Fig. 2. 8-eek-male C57BL/6 mice from genetic engineering mice center of

ichuan University (Chengdu, China) were used as recipients. GFPransgenic mice (GFP mice) of C57BL/6 background were kindlyrovided by Dr. Mo (Stem Cell Laboratory of National Key Labo-atory of Sichuan University, Chengdu, Sichuan) and maintainedn our animal facility. Male GFP mice aged 6–8 weeks were useds bone marrow (BM) donors. 10 recipient C57BL/6 mice wererradiated at dose of 13 Gy. BM cells of GFP mice were collectedy flushing the bone shafts of the femora and tibia with DMEMedium. The cells were then washed once in this medium and sus-

ended in phosphate-buffered saline (PBS) and 1 × 107 cells werenjected via the tail vein.

Fig. 2. Protocol of chimeric mice production.

ience 262 (2012) 200– 206 201

2.3. Western blotting

One month after bone marrow transplantation (BMT), proteinswere collected from fresh skeletal muscle, heart, liver, spleen,lung, kidney of wild type mice, GFP mice and chimeric mice, andseparated using a 10% polyacrylamide gel. After transferring theproteins on a nitrocellulose membrane, the proteins were blockedwith a 5% defatted milk solution. The membrane was then probedwith mouse monoclonal antibody against GFP (1:2500, Chemi-con) and Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH,1:5000, Chemicon), and then probed with a secondary antibodyusing a 1 in 5000 dilution of alkaline phosphatase conjugatedanti-mouse IgG.

2.4. Fluorescence observation

Frozen sections of various tissues were cut at 6-mm thickness,and the nucleus were stained with 4′,6-diamidino-2-phenylindole(DAPI). Frozen sections were examined under a fluorescence micro-scope for detection of GFP-positive cells.

2.5. Implants

20 wild type mice (group 1, n = 10) and chimeric mice (group 2,n = 10) were both intraperitoneally anesthetized and their legs weredisinfected. The operations were done immediately after BMT. OneHA/�-TCP sample was implanted into both the left and right legmuscles of each animal. The wounds were closed by single inter-rupted suturing. The study was approved by the Animal Care andUse Committee of Sichuan University. The operative proceduresand animal care were performed in compliance with NIH guidelineson the care and use of laboratory animals, under the supervision ofa licensed veterinarian.

2.6. Histology

The HA/�-TCP samples were harvested at 45 and 90 post-operative days. They were immediately fixed in 10% neutralformalin buffer solution for approximately 24 h at room tem-perature, decalcified in 10% ethylenediaminetetraacetic acid(EDTA), pH 7.0, for about 20 days at room temperature, washedwith diethyl pyrocarbonate (DEPC), dehydrated and embeddedin paraffin (melting point 56–58 ◦C). The embedded sampleswere cut into 5 �m thick histological sections and trans-ferred onto 3-aminopropyrytrietoxy silane-coated glass slides.The ceramic sections were stained with hematoxylin and eosin(HE).

2.7. Immunohistochemistry

The slides were deparaffinized, rehydrated and rinsed withdouble-distilled water. They were treated with 3% H2O2 for15 min in the dark to block endogenous peroxidase, rinsed threetimes with double-distilled water, immersed in a Tris–EDTAbuffer (pH 9.0) and kept in a 95 ◦C water bath for 45 min.After cooling to room temperature they were rinsed three times(5 min each) with PBS and incubated in mouse monoclonal anti-body against GFP (1:1000, Chemicon) at 4 ◦C overnight. Thenthe slides were rinsed three times with PBS and incubated

in a horseradish peroxidase (HRP)-labeled secondary antibodyat room temperature for 30 min. Finally, they were devel-oped with 3,3′-diaminobenzidine (DAB) and counterstained withhematoxylin.

202 L. Cheng et al. / Applied Surface Sc

Fmd

3

3

sGm(

siwshhwHfimciwofwoamGsncl

fttm

3

l

ig. 3. GFP protein levels analyzed by western blotting in wild type mice, chimericice and GFP mice. GFP protein was found in chimeric mice and GFP mice but not

etected in wild type mice. Blots for GAPDH are included as internal loading controls.

. Results

.1. Chimeric model

Western Blot was used to detect the GFP in muscle, heart, liver,pleen, lung and kidney from wild type mice, chimeric mice andFP mice. Positive GFP were detected in chimeric mice and GFPice, however, GFP was not observed in wild type mice as control

Fig. 3).Immunohistochemistry (IHC) was also used to detect GFP. Serial

ections of wild type mice, chimeric mice and GFP mice weremmunostained. The brown-colored area was a positive result

hich indicated the expression of the specific antigen GFP. Ashown in microscopic pictures, GFP was positive within muscle,eart, liver, spleen, lung and kidney from GFP mice (Fig. 4M–R). Ineart muscle and skeletal muscle of chimeric mice, GFP positive cellere found in interstitial space, but not in muscle fibers (Fig. 4G and); while in GFP mice, GFP positive cell were found both in muscleber and interstitial space (Fig. 4A and B). In the liver of chimericice, positive cells were scattered diffusely within the lobule, espe-

ially along the sinusoids and no GFP-positive cells were observedn hepatocytes (Fig. 4I); and in GFP mice, the GFP positive cells

ere found both in sinusoids and hepatocytes (Fig. 4O). In spleenf chimeric mice, GFP-positive cells were observed in red pulp andew in white pulp (Fig. 4J); and in GFP mice, GFP-positive cellsere observed both in red pulp and white pulp (Fig. 4P). In lung

f chimeric mice, GFP-positive cells were found in the peripherallveoli but not in bronchioles (Fig. 4K); while in GFP mice, the sameicrostructure was observed (Fig. 4Q). In kidney of chimeric mice,FP-positive cells could be found in any glomeruli and interstitialpace (Fig. 4L); and in GFP mice, GFP cells were found in glomeruli,ephridial tubule and interstitial space (Fig. 4R). However, GFPould not be found in skeletal muscle, heart muscle, liver, spleen,ung and kidney of wild type mice (Fig. 4A–F).

To detect the fluorescence generated by GFP, the frozen sectionrom chimeric mice were stained with DAPI and examined underhe fluorescence microscope. The fluorescence was found in skele-al muscle, heart muscle, liver, spleen, lung and kidney in chimeric

ice (Fig. 5).

.2. Bone formation

In group 1, 45 days after implantation, a large number of irregu-ar arrangements of bone tissues were found in ceramics (Fig. 6A).

ience 262 (2012) 200– 206

Ninety days implantation, bone tissue was observed in the micro-pores of the bioceramics in group 1, the arrangement was moreregular compared with the bone tissues harvested at 45 days. Somecalcification lines and bone marrow could be observed in the bonematrix (Fig. 6B). However, bone tissue was not found 45 days or 90days after implantation in group 2 (Fig. 6C and D). Various inflam-matory cells, such as monocytes, macrophages and lymphocyteswere observed in the implanted biomaterials (Fig. 6C and D). Somevessels were found in day 45 (Fig. 6C). More and larger vessels werefound in day 90 after implantation (Fig. 6D).

Immunostaining of GFP showed that GFP was identified in vas-cular cells in group 2 at 45 and 90 days after implantation (Fig. 7Aand B). These findings showed that bone marrow cell was one of cellsources of vascular formation in bioceramics implanted in chimericmice.

4. Dicussion

The migration of bone marrow cells after transplantation hasbeen studied previously [12]. This animal model has been devel-oped to mimic status after BMT. External radiation was deliveredby a single fraction of 13 Gy. Irradiation was used for ablation of BMand to create space for infused BM cells as previously suggested byCui et al. [13]. In our study, we found that mice could not surviveand died several days after lethal irradiation without BMT, whileanimals with BMT could survive for more than one year withoutany obvious complications. Under lethal irradiation, bone marrowcells with GFP could be monitored and tracked after BMT. The greenfluorescent of GFP mice could be observed under UV light; afterbeing killed, the green muscles could be examined by the nakedeye, however, other tissues were not green by the naked eye. Fromour IHC results we can see that GFP was positive in almost all cellsin heart muscle and skeletal muscle of GFP mice, while GFP was notpositive in some cells in liver, spleen, lung and kidney; this is consis-tent with our western blot’s results, in which the expression of GFPprotein in muscle was more than that in other tissues. However,the expression of GFP protein was almost the same in all tissuesin chimeric mice from our results of western blot, IHC and directfluorescence observation. Originally, this model was built to revealprocess that the BM cells have contribution to ectopic bone forma-tion. However, the new bone was not found in chimeric mice basedon our results. Several studies have recently reported that specificmaterials induced bone formation 6–12 weeks after implantationin non-bony sites [6,7,9,14–16]. For instance, bone induction hasbeen observed with synthetic hydroxyapatite ceramics, biphasiccalcium phosphate ceramics, tricalcium phosphate ceramics, cal-cium phosphate cements, bioglass, and even with titanium oxideand alumina ceramics implanted into the muscles of rabbits, dogs,goats, baboons, and pigs [6–10,14–17]. As indicated in the reportsof Fellah’s team, osteoinduction by bioceramic materials is a com-plex biological phenomenon which is not fully understood, but hasshown that ectopic bone formation is both material and animaldependant [18]. The majority animals were considered quite largeand included very few rodent animals. Cheng reported osteoinduc-tion in fractured fibula mice in certain calcium phosphate ceramics[19]. However, the bone could be induced in soft tissue with a spe-cial treatment. The osteoinduction of biphasic calcium phosphateceramics in mice was confirmed based on our work.

Chimeric mice did not show any bone formation. Cheng believedthat the HA/�-TCP samples with poor vascularization might leadto few bone formation in ectopic sites [19]. Poor vascularization

might have limited exposure to osteogenic factors, as well as otherbiological substances (e.g. angiogenic factors and nutrients). Ripa-monti deemed that angiogenic and osteogenic proteins played animportant role in tissue induction and bone morphogenesis within

L. Cheng et al. / Applied Surface Science 262 (2012) 200– 206 203

Fig. 4. Immunohistochemistry of GFP in wild type mice, chimeric mice and GFP mice. Skeletal muscle, heart muscle, liver, spleen, lung and kidney in wild type mice (A–F).Skeletal muscle, heart muscle, liver, spleen, lung and kidney in chimeric mice (G–L). Skeletal muscle, heart muscle, liver, spleen, lung and kidney in GFP mice (M–R) (400×).

204 L. Cheng et al. / Applied Surface Science 262 (2012) 200– 206

Fig. 5. GFP fluorescence from GFP-positive cell in skeletal muscle, heart muscle, liver, spleen, lung and kidney from chimeric mice (A–F). DAPI in skeletal muscle, heartmuscle, liver, spleen, lung and kidney (G–L). An overlay of GFP fluorescence and DAPI in skeletal muscle, heart muscle, liver, spleen, lung and kidney (M–R) (100×).

L. Cheng et al. / Applied Surface Science 262 (2012) 200– 206 205

F serveM

tpacmbowm

aiccbtahr

ig. 6. HE staining of HA/�-TCP harvested from mice. Histological features were ob, HA/�-TCP; B, bone; BM, bone marrow (400×).

he porous spaces of biomaterials [20]. Angiogenic and osteogenicroteins bind to type IV collagen of the invading capillaries, andre presented in an immobilized form to responding mesenchymalells to initiate osteogenesis and angiogenesis. Initiators and pro-oters of angiogenesis and bone morphogenesis are thus modeling

one formation by induction in angiogenesis. However, based onur study, vascularization was found obviously and osteoinductionas not detected at day 45 and 90 after implantation in the BMTodel.The BMT procedure is associated with phenotypic changes and

reduced ability of the host to produce immunoglobulin of var-ous isotypes after antigen stimulation [21]. Deficits in various Tell-mediated functions, such as mitogen responsiveness, helper-ell function, cytotoxicity, and contact hypersensitivity have alsoeen found. Moreover, natural killer cell activity appears to be

ransiently depressed in BMT recipients [22]. While some of theselterations in immune function reverse with time others, such aselper T cell function, contact hypersensitivity, and IgA secretion,emain depressed for years after successful engraftment [23]. As

Fig. 7. Immunohistochemistry of GFP in decalcified sections of the group 2 at

d in group 1 (A, C) and group 2 (B, D) at days 45 and 90. Arrow head, blood vessels;

we know, the immune system and skeletal system share variousmolecules, including signaling molecules, membrane receptors andtranscription factors [24], so the immune system should have a linkwith osteoinduction. It has been proved that osteoblasts composethe niche required for the maintenance of hematopoietic stem cells[25], when the mice went through lethal irradiation, the generatingof hematopoietic stem cells was affected; therefore, ectopic boneformation cannot occur in GFP chimeric mice. Though new vas-cularization was detected in the ceramics, it has suggested thatplatelet-derived growth factor decreased osteogenic differentia-tion of mesenchymal progenitor cells [26]. All these immunologicalchanges may block the cascade of bone differentiation by induc-tion, despite angiogenic and osteogenic proteins that providesthe conceptual framework of the supramolecular assembly of theextracellular matrix of bone attached to the implanted matrices.

Which types of cells are capable of transforma-tion/differentiation into active secreting osteoblastic-like cells atthe biomaterial interface? Bone marrow has cells with the potentialto differentiate into mature cells of various organs including the

45 (A) and 90 (B) days. Arrow head, blood vessels; M, HA/�-TCP (400×).

2 face Sc

hcemcdurrpceb

5

Tiwrmd

A

o

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

06 L. Cheng et al. / Applied Sur

eart, liver, kidney, lungs, gastrointestinal tract, skin, bone, muscle,artilage, fat, endothelium, and brain, in addition to hematopo-tic stem cells and supportive stromal cells [27]. Multipotentyoendothelial cells residing in stem cells respond to endothelial

ell mitogens including angiogenic factors. Myoendothelial cellsifferentiate into myogenic, chondrogenic and osteogenic cellsnder appropriate conditions [28]. Myoendothelial cells may alsoespond to bone morphogenetic proteins previously inducing theipple-like cascade of the induction of bone formation within theorous spaces of the biomimetic matrices, mimicking the cortico-ancellous remodeling cycle of the primate bone. Myogenic andndothelial cells derived from bone marrow stem cells were foundased on our findings (Fig. 7).

. Conclusion

This study compared the osteoinduction of porous HA/�-CP samples implanted in the same genotype mice. HA/�-TCPmplanted in wild type mice showed osteoinductivity compared

ith those implanted in GFP Chimeric mice. The differences wereelated to immunological function. A better understanding of theechanisms may provide valuable information for the rational

esign and improvement of osteoinductive biomaterials.

cknowledgments

This work was funded by the National Basic Research Programf China (973 program) (no. 2005CB623900-G).

eferences

[1] C. Laurencin, Y. Khan, S.F. El-Amin, Bone graft substitutes, Expert Review ofMolecular Diagnostics 3 (January (1)) (2006) 49–57.

[2] P.V. Giannoudis, I. Pountos, Tissue regeneration. The past, the present and thefuture, Injury 36 (November (Suppl. 4)) (2005) S2–S5.

[3] P.D. Costantino, C.D. Friedman, Synthetic bone graft substitutes, Otolaryngo-logic Clinics of North America 27 (October (5)) (1994) 1037–1074.

[4] G. Daculsi, O. Laboux, O. Malard, P. Weiss, Current state of the art of bipha-sic calcium phosphate bioceramics, Journal of Materials Science Materials inMedicine 14 (March (3)) (2003) 195–200.

[5] S. Fujibayashi, J. Shikata, C. Tanaka, M. Matsushita, T. Nakamura, Lumbar pos-terolateral fusion with biphasic calcium phosphate ceramic, Journal of SpinalDisorders 14 (June (3)) (2001) 214–221.

[6] U. Ripamonti, J. Crooks, L. Khoali, L. Roden, The induction of bone formation bycoral-derived calcium carbonate/hydroxyapatite constructs, Biomaterials 30(March (7)) (2009) 1428–1439.

[7] U. Ripamonti, Osteoinduction in porous hydroxyapatite implanted in hetero-topic sites of different animal models, Biomaterials 17 (January (1)) (1996)31–35.

[8] H. Yuan, M. Van Den Doel, S. Li, C.A. Van Blitterswijk, K. De Groot, J.D. De Bruijn, Acomparison of the osteoinductive potential of two calcium phosphate ceramics

implanted intramuscularly in goats, Journal of Materials Science Materials inMedicine 13 (December (12)) (2002) 1271–1275.

[9] H. Yuan, Z. Yang, J.D. De Bruij, K. De Groot, Zhang X, Material-dependent boneinduction by calcium phosphate ceramics: a 2.5-year study in dog, Biomaterials22 (October (19)) (2001) 2617–2623.

[

[

ience 262 (2012) 200– 206

10] Z. Yang, H. Yuan, W. Tong, P. Zou, W. Chen, X. Zhang, Osteogenesis inextraskeletally implanted porous calcium phosphate ceramics: variabilityamong different kinds of animals, Biomaterials 17 (November (22)) (1996)2131–2137.

11] H. Yuan, C.A. van Blitterswijk, K. de Groot, J.D. de Bruijn, Cross-speciescomparison of ectopic bone formation in biphasic calcium phosphate (BCP)and hydroxyapatite (HA) scaffolds, Tissue Engineering 12 (June (6)) (2006)1607–1615.

12] A. Satoh, T. Saito, Y. Sato, T. Tsuchiya, A. Kenjo, T. Kimura, et al., Traffic of infusedbone marrow cells after genetically-labeled syngeneic bone marrow trans-plantation following lethal irradiation in mice, Fukushima Journal of MedicalScience 54 (June (1)) (2008) 11–24.

13] Y.Z. Cui, H. Hisha, G.X. Yang, T.X. Fan, T. Jin, Q. Li, et al., Optimal protocol fortotal body irradiation for allogeneic bone marrow transplantation in mice, BoneMarrow Transplantation 30 (December (12)) (2002) 843–849.

14] M. Takemoto, S. Fujibayashi, M. Neo, J. Suzuki, T. Matsushita, T. Kokubo, et al.,Osteoinductive porous titanium implants: effect of sodium removal by diluteHCl treatment, Biomaterials 27 (May (13)) (2006) 2682–2691.

15] S. Fujibayashi, M. Neo, H.M. Kim, T. Kokubo, T. Nakamura, Osteoinductionof porous bioactive titanium metal, Biomaterials 25 (February (3)) (2004)443–450.

16] H. Yuan, J.D. de Bruijn, X. Zhang, C.A. van Blitterswijk, K. de Groot, Bone induc-tion by porous glass ceramic made from Bioglass (45S5), Journal of BiomedicalMaterials Research 58 (May (3)) (2001) 270–276.

17] Ripamonti U, Bone induction in nonhuman primates. An experimental studyon the baboon, Clinical Orthopaedics and Related Research (Aug (269)) (1991)284–294.

18] B.H. Fellah, O. Gauthier, P. Weiss, D. Chappard, P. Layrolle, Osteogenicity ofbiphasic calcium phosphate ceramics and bone autograft in a goat model, Bio-materials 29 (March (9)) (2008) 1177–1188.

19] L. Cheng, F. Ye, R. Yang, X. Lu, Y. Shi, L. Li, et al., Osteoinduc-tion of hydroxyapatite/beta-tricalcium phosphate bioceramics inmice with a fractured fibula, Acta Biomaterialia 6 (April (4)) (2010)1569–1574.

20] U. Ripamonti, Soluble osteogenic molecular signals and the induction of boneformation, Biomaterials 27 (February (6)) (2006) 807–822.

21] L. Pan, S. Bressler, K.R. Cooke, W. Krenger, M. Karandikar, Ferrara JL, Long-term engraftment, graft-vs.-host disease, and immunologic reconstitution afterexperimental transplantation of allogeneic peripheral blood cells from G-CSF-treated donors, Biology of Blood and Marrow Transplantion 2 (October (3))(1996) 126–133.

22] M. Bengtsson, T.H. Totterman, B. Smedmyr, R. Festin, G. Oberg, B. Simonsson,Regeneration of functional and activated NK and T sub-subset cells in the mar-row and blood after autologous bone marrow transplantation: a prospectivephenotypic study with 2/3-color FACS analysis, Leukemia 3 (January (1)) (1989)68–75.

23] A.R. Zander, J.M. Reuben, D. Johnston, L. Vellekoop, K.A. Dicke, J.C. Yau, et al.,Immune recovery following allogeneic bone marrow transplantation, Trans-plantation 40 (August (2)) (1985) 177–183.

24] H. Takayanagi, Osteoimmunology: shared mechanisms and crosstalk betweenthe immune and bone systems, Nature Reviews Immunology 7 (April) (2007)292–304.

25] O. Kollet, A. Dar, T. Lapidot, The multiple roles of osteoclasts in host defense:bone remodeling and hematopoietic stem cell mobilization, Annual Review ofImmunology 25 (2007) 51–69.

26] R. Gruber, F. Karreth, B. Kandler, G. Fuerst, A. Rot, M.B. Fischer, et al., Platelet-released supernatants increase migration and proliferation, and decreaseosteogenic differentiation of bone marrow-derived mesenchymal progen-itor cells under in vitro conditions, Platelets 15 (February (1)) (2004)29–35.

27] D.S. Krause, N.D. Theise, M.I. Collector, O. Henegariu, S. Hwang, R. Gardner, et al.,

Multi-organ, multi-lineage engraftment by a single bone marrow-derived stemcell, Cell 105 (May (3)) (2001) 369–377.

28] B. Zheng, B. Cao, M. Crisan, B. Sun, G. Li, A. Logar, et al., Prospective identificationof myogenic endothelial cells in human skeletal muscle, Nature Biotechnology25 (September (9)) (2007) 1025–1034.