VU Research Portal

Osteoinductive bone substitutes Liu, T.

2013

document version Publisher's PDF, also known as Version of record

Link to publication in VU Research Portal

citation for published version (APA) Liu, T. (2013). Osteoinductive bone substitutes.

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?

Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

E-mail address: [email protected]

Download date: 10. Oct. 2021

Osteoinductive Bone Substitutes

Tie Liu

The following institutions generously funded printing of this thesis: Academic Centre for Dentistry Amsterdam VU University Amsterdam

Tie Liu Osteoinductive bone substitutes

Thesis Amsterdam – With ref. – With Summary in Dutch

ISBN: 978-90-5383014-7

Copyright © 2013 by Tie Liu. All Rights Reserved. No part of this book may be reproduced, stored in a retrievable system, or transmitted in any form or by any means, mechanical, photo-copying, recording or otherwise, without the prior written permission of the holder of copyright.

VRIJE UNIVERSITEIT

Osteoinductive bone substitutes

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. F.A. van der Duyn Schouten, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de Faculteit der Tandheelkunde op woensdag 18 september 2013 om 11.45 uur in de aula van de universiteit, De Boelelaan 1105

door

Tie Liu geboren te Zhejiang, China

promotoren: prof.dr. D. Wismeijer prof.dr. Z. Gu copromotor: dr. Y. Liu

Dedicated to my wife Qian Lu

to my parents

Publications 1. Tie Liu, Bing Xia, Zhiyuan Gu. Inferior alveolar canal course: a radiographic study. Clinical Oral Implant Research, 2009; 20: 1212–1218.

2. Tie Liu, Gang Wu, Daniel Wismeijer, Zhiyuan Gu and Yuelian Liu. Deproteinized bovine bone functionalized with the slow delivery of BMP-2 for the repair of critical-sized bone defects in sheep. Bone. 2013, 56: 110–118.

3. Xin Zhang&, Tie Liu&, Yuanliang Huang, Daniel Wismeijer, and Yuelian Liu. Icariin: Does It Have An Osteoinductive Potential for Bone Tissue Engineering? Phytotherapy Research. 2013 Jul 4. doi: 10.1002/ptr.5027. [Published online] (& contributed equally)

4. Tie Liu, Gang Wu, Yuanna Zheng, Daniel Wismeijer, Vincent Everts, and Yuelian Liu. Cell-mediated BMP-2 release from a novel dual drug delivery system promotes bone formation. Clinical Oral Implant Research 2013. [under revision]

5. Yuanna Zheng, Gang Wu, Tie Liu, Yi Liu, Daniel Wismeijer, and Yuelian Liu. A novel BMP2-coprecipitated, layer-by-layer assembled biomimetic calcium phosphate particle: a biodegradable and highly-efficient osteoinducer. Clinical Implant Dentistry and Related Research, 2013 Mar 4. doi: 10.1111/cid.12050. [Published online]

6. Jingxiao Wang, Yuanna Zheng, Juan Zhao, Tie Liu, Lixia Gao, Zhiyuan Gu and Gang Wu. Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation. Journal of Clinical Periodontology 2012 Jan;39(1):98-105.

7. Tie Liu, Gang Wu, Yuanna Zheng, Daniel Wismeijer, and Yuelian Liu. A biomimetic osteoinducer enhances the therapeutic effects of deproteinized bovine bone in a sheep critical-sized bone defect (Ø8×13mm) model. 2013 [submitted].

8. Tie Liu, Gang Wu, Daniel Wismeijer, and Yuelian Liu. Osteoinductive biomimetic bone substitute for the repair of critical-sized bone defects in sheep. 2013 [submitted].

9. Tie Liu, Gang Wu, Yuanna Zheng, Afsheen Tabassum, Daniel Wismeijer, Vincent Everts, and Yuelian Liu. A single biomimetic calcium phosphate granule as a model to deliver proteins. 2013 [submitted].

10. Tie Liu, Sven Bakx, Gang Wu, Leo van Ruijven, Daniel Wismeijer, and Yuelian Liu. Cone-beam CT and micro-CT analysis of deproteinized bovine bone for the repair of critical-sized bone defects in sheep. 2013 [in preparation].

11. Yuanan Zheng, Tie Liu, Zhiyuan Gu. Investigation of changes of articles on China national academic stomatological conferences in last two decades. Stomatology 2007 (12): 643-645. (Chinese)

12. Qian Lu, Tie Liu. Progressive Studies on Effects of Traditional Chinese Medicines on Differentiation, Proliferation and Bone Formation Gene Expression of Osteoblasts. Journal of Zhejiang Chinese Medical University 2012(5): 609-612. (Chinese)

CONTENT

Chapter 1 General introduction ……………………………………………….…1

Chapter 2 Cell-mediated BMP-2 release from a novel dual drug delivery system promotes bone formation ...…………………………….….………….9

Chapter 3 Preparation and characteristics of osteoinductive biomimetic calcium phosphate material: in vitro and in vivo study ……………………...29

Chapter 4 Osteoinductive biomimetic bone substitute for the repair of critical-sized bone defects in sheep ...……………………………….47

Chapter 5 A novel BMP2-coprecipitated, layer-by-layer assembled biomimetic calcium phosphate particle: a biodegradable and highly-efficient osteoinducer ..…………………………………...…………………...65

Chapter 6 A biomimetic osteoinducer enhances the therapeutic effects of deproteinized bovine bone in a sheep critical-sized bone defect (Ø8×13mm) model ……………………………………………….…83

Chapter 7 Deproteinized bovine bone functionalized with the slow delivery of BMP-2 for the repair of critical-sized bone defects in sheep ……………………………………………………………….101

Chapter 8 Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation ……………………….…………....123

Chapter 9 Icariin: does it have an osteoinductive potential for bone tissue engineering? …………………………………………………….….137

Chapter 10 General discussion ...……………………………………………….161

Chapter 11 General summary ….……………………………………………….167

Dutch summary….………………………………………...…….….172

Acknowledgements ………..………………………………………177

Curriculum vitae ………………………...…………………………179

Abbreviations ACP amorphous calcium phosphate BCP biphasic calcium phosphate BioCaP biomimetic calcium phosphate BMP bone morphogenetic protein BMP2-cop.BioCaP BMP-2-coprecipitated biomimetic calcium phosphate BMSCs bone marrow stem cells BSA bovine serum albumin CaP calcium phosphate CDHA calcium deficient hydroxyapatite CPS calcium phosphate supersaturated solution CSBD critical-sized bone defect DBB deproteinized bovine bone EDX energy-dispersive x-ray spectroscopy FBGC foreign body giant cell HA hydroxyapatite ICA icariin MNC multinucleated giant cell OCP octacalcium phosphate OPG panoramic radiograph PBMCs peripheral blood mononuclear cell PMMA poly methylene methacrylate RANKL receptor activator for nuclear factor-κB ligand SBF supersaturated body fluids SEM scanning electron microscopy TB trabecular bone TCM traditional Chinese medicine TCP tricalcium phosphate TRACP tartrate-resistant acid phosphatase VEGF vascular endothelial growth factor XRD X-ray diffraction

Chapter 1

General Introduction

1

Chapter 1

GENERAL INTRODUCTION

The treatment of bone fractures and defects requires adequate volume of bone tissue which is of paramount importance to achieve an excellent restoration. When the bone defects are too large to be self-healed, bone grafting is required in order to fill the defect [1, 2]. Bone grafts fill voids and serve as scaffolds to provide support, and therefore may enhance the biological repair of the defect. Critical-sized bone defect (CSBD) is defined as the intraosseous wound with the smallest size, which cannot spontaneously heal completely without intervention [3]. Bone healing heals through the generation of new bone rather than by forming fibrotic tissue. Usually, the fibrous connective tissue regenerates faster than bone tissue and becomes dominant within the CSBD because of the faster migration mechanism of fibroblasts compared to osteoblasts. Bone grafting, as a common surgical procedure, is carried out in approximately 10% of all skeletal reconstructive surgery cases [4]. Worldwide, more than 2.2 million grafting procedures are performed annually [1, 5]. In most patients, the intervention therapies can be unproblematically executed and the outcome is generally excellent [6, 7]. However, there are still a significant number of eligible individuals with the existence of well-recognized risk factors such as diabetes, local osteoporosis and metabolic bone disorder. These risk factors are associated with poor activity of bone formation [2, 8]. Nevertheless, the expectations of patients and surgeons alike are continually rising, both aspiring to a curtailment of the recovery phase and the postoperative period of functional incapacity [9]. Consequently, these clinical, social and economic pressures make it absolutely necessary to develop a simple, efficacious and cost-effective bone substitute to expedite and augment bone formation.

Bone regeneration Bone regeneration in large bone defects requires four critical elements: (i) osteogenic cells (e.g. progenitor cells or osteoblasts); (ii) osteoinductive signals (growth factors); (iii) an biocompatible, biodegradable and osteoconductive matrix (scaffold); and (iv) adequate blood and nutrient supply [10]. Therefore, bone grafts are often associated with the terms biocompatibility, biodegradability, osteoconductivity and osteoinductivity. Good biocompatibility refers to the ability of a biomaterial to perform its desired function with respect to a medical therapy, without eliciting any undesirable local or systemic effects in the recipient or beneficiary of that therapy, but generating the most appropriate beneficial cellular or tissue response in that specific situation, and optimizing the clinically relevant performance of that therapy [11]. Ideal biodegradability refers to that bone substitute can be degraded in short time, enabling bone remodeling, and concomitantly replaced by bone tissue [12]. Osteoconductivity is the ability of the graft to function as a scaffold to permit bone growth on its surface or for ingrowth of new bone [13]. Osteoinductivity is the ability of a graft to stimulate primitive, undifferentiated and pluripotent cells to develop into the bone-forming cell lineage, and consequently to promote bone formation [13, 14]. Therefore, the perfect bone substitute should be osteoconductive, osteoinductive, biocompatible and biodegradable. It should induce minimal or no fibrotic reaction,

2

General introduction undergo remodeling and support new bone formation. From a mechanical point of view bone substitutes should have similar strengths to that of the bone being replaced. Finally, it should be cost-effective and ought to be available in the amount required.

Bone substitutes Autografts Autologous bone, mostly harvested from the iliac crest, is regarded as the gold standard since it provides an osteoconductive 3-demensional scaffold for bone ingrowth, osteogenic cells and osteoinductive growth factors [15]. However, the harvesting of autologous bone prolongs the surgery and the graft amount may be insufficient. Autograft harvesting is also associated with an 8-39% risk of complications, e.g. infection, hematoma, nerve injury, cosmetic disadvantages, pain and morbidity of the donor site [16]. Moreover, the irregular rate of resorption of the autologous bone may require secondary corrective surgery or compromise the restoration rate [17]. It has been reported that resorption rates for endochondral bone is up to 75% and rates of 20%-30% for membranous bone autografts [18, 19].

Allografts The allogeneic bone graft is obtained either from cadavers or living individuals from the same species [20, 21]. It provides a good, natural, and bony scaffold. However, allogeneic bone is still associated with risks such as disease transmission [22], variable host immune response [23], toxicity associated with sterilization [24], and limited supplies [25]. In some countries, the allografts are culturally unacceptable.

Xenografts Xenografts are composed of tissue taken from another species (i.e. from an animal source, usually bovine). The use of xenografts has the potential to reduce morbidity as harvest of autogenous bone is unnecessary. The antigenic potential of xenografts can be diminished or eliminated by chemical treatment. One of the most widely used xenograft in clinical dentistry is deproteinized bovine bone (DBB, Bio-Oss®, Geistlich, Switzerland). It is derived from a bovine source and is treated by a chemical extraction process to remove all the organic components and pathogens [17]. In terms of its inorganic composition and its isomeric crystalline dimensions, DBB has a physical and chemical structure similar to that of natural bone [26]. It shows osteoconductive properties when it is in close contact with the newly formed bone [27]. However, it was reported that DBB delays the early bone formation [28] and lacks sufficient intrinsic osteoinductivity [29].

Synthetic calcium phosphate bone substitute Technological evolution and better understanding of bone-healing mechanism resulted in the development of numerous alternative bone substitutes. Calcium phosphate (CaP) grafts have been widely used for bone regeneration in most trauma and orthopedic surgery procedures when grafting is necessary to restore bone defects. The calcium phosphate materials are available with different application forms, e.g. pastes, granules, blocks, composites. Based upon their chemical composition, calcium phosphates can be

3

Chapter 1 classified as either hydroxyapatite (HA), beta-tricalcium phosphate (β-TCP), biphasic calcium phosphate (BCP), amorphous calcium phosphate (ACP), carbonated apatite (CA) or calcium deficient HA (CDHA) [4]. A further subdivision can be made between ceramics and cements. A ceramic is defined as an inorganic, non-metallic solid prepared by sintering [30]. The sintering process removes volatile chemicals and increases crystal size, resulting in a porous and solid material. Cements consist of a mixture of calcium phosphates which can be applied as a paste and harden in situ due to precipitation reactions. Calcium phosphate cements/ceramics can be applied as carriers for drugs. In general, ceramics show a higher initial release than cements that have a more sustained release pattern. It has been aware that the low degradability of sintered CaP material, and in particular HA or CA forms a problem. It has been well known that the high porosity of implants has benefits for bone formation inside the implant and increases degradation [12]. Porosity in the CaP material can be introduced by leaching and sintering out of salt crystals or polymeric microparticles/mold after which the CaP material remains. Complete resorption in most cases is very difficult due to the crystalline architecture. The combination of CaP and polymer can form a suitable scaffold for cells and serve as a delivery vehicle of osteoinductive drugs. The addition of biodegradable polymers can improve the degradability of the CaP materials and alter their mechanical/physical properties.

The ideal osteoinductive bone substitute Although most bone substitutes are osteoconductive to bone-forming osteoblasts, only a limited number of osteoinductive materials are currently available on the market with FDA approval [31]. It was reported that tricalcium phosphate (TCP) is osteoinductive as a synthetic alternative to autologous bone grafting [32]. However, a fundamental understanding of the term, osteoinductivity, is of critical importance. Osteoinductivity is the ability of the material to induce de novo bone formation. The osteoinduction phenomenon could be divided into 3 principles [31]: (1) mesenchymal cell recruitment; (2) mesenchymal differentiation to bone-forming osteoblasts; and (3) ectopic bone formation in vivo. The ability for a bone graft to induce new bone formation in an intraosseous defect does not fully reflect its true osteoinductive property. The osteoinductive property of a material is usually demonstrated by bone formation after implantation in ectopic/nonosseous sites. There are two kinds of nonosseous sites that can be used to test osteoinduction in vivo. One is to implant subcutaneously, and the other is to implant into intramuscular site [33]. However, the subcutaneous and intramuscular sites are different. The difference in osteoinduction could be related to the partial pressure of oxygen or the blood supply in the intramuscular and subcutaneous sites, and that immature mesenchymal cells in the muscle could more easily differentiate into osteoblasts, leading to osteoinduction [34]. It was also demonstrated that even a small amount (5μg) of recombinant human bone morphogenetic protein-2 (rhBMP-2) induces new bone in the subcutaneous tissue, which has a lesser blood flow than the muscle [34].

4

General introduction

Biomimetic calcium phosphate for bone regeneration Recently, the biomimetic calcium phosphate coating has been developed very well for the slow delivery of growth factors. This coating usually included two layers: an amorphous layer and a crystalline layer, both of which are calcium phosphate. The amorphous layer, serving as a seeding layer, can strongly deposit on the underlying materials. Thereafter, the crystalline layer, octacalcium phosphate serving as a three-dimensional reservoir for carrying protein/drugs, can grow on the seeding layer. Therefore, this biomimetic coating can be applied on a variety of materials such as titanium implant [35], polymer [36], ceramic [37], zirconia [38], and deproteinized bovine bone [39]. At the same time, growth factor or drugs have been incorporated into this coating, such as antibiotics, bone morphogenetic protein-2 (BMP-2), and vascular endothelial growth factor (VEGF). These bioactive agents incorporated in the latticework of crystalline calcium phosphate of the coating presented a slow and sustained release manner [39, 40]. Such a slow release has been shown to be beneficial for the effect of different growth factors such as BMP-2 and VEGF. The slow delivery of BMP-2 from the coating enhances osteoinduction [41], and the slow delivery of VEGF promotes vascularisation [40].

Objectives of the thesis The general aim of this thesis includes 5 aspects: 1. To develop a biomimetic calcium phosphate (BioCaP) bone substitute as a dual delivery model with two protein-delivery modes: one mode by which protein was incorporated in the interior of BioCaP; and one by which protein was coated on the outside of BioCaP. We hypothesize that using this model the release of the protein can be sequential and slow, and that the two delivery modes of BMP-2 could efficiently accelerate bone formation. 2. To develop particles of biomimetic BMP-2-coprecipitated calcium phosphate (BMP2-cop.BioCaP). We hypothesize that these particles could serve as an independent and biodegradable osteoinducer. 3. To evaluate the therapeutic effect of the deproteinized bovine bone functionalized with coating-incorporated BMP-2 in the repair of critical-sized bone defect in sheep. 4. To delineate the dynamic micro-architectures of bone induced by low-dose bone morphogenetic protein (BMP)-2/7 heterodimer in peri-implant bone defects compared to BMP2 and BMP7 homodimer. 5. To determine the present evidence of the osteoinductive potential of a Chinese traditional medicine, icariin.

Outline of this thesis As an alternative of autograft, a biomimetic calcium phosphate (BioCaP) bone substitute was developed with two protein-delivery modes: 1) internally-incorporated mode: protein was incorporated in the interior of BioCaP; and 2) coating-incorporated mode: protein was coated on the outside of BioCaP. Slow release of bioactive agents from bone substitutes plays an important role in the treatment of bone defects. Resorbing cells such as osteoclasts may accelerate the degradation of bone substitutes so as to elevate the protein release. The cell-mediated protein release of the two modes of BioCaP was

5

Chapter 1 investigated. The in vivo bone formation and cell response were evaluated by histological and histomorphometric analysis in an ectopic rat model (Chapter 2 and 3). In addition, we evaluated the physical and chemical properties of BioCaP and the ability for protein loading and release in a long period. A Micro-CT method for the evaluation of graft material and bone has been applied by using a unique “onion-peeling” algorithm and specific threshold settings (Chapter 3). Furthermore, we hypothesized that BioCaP can be a synthetic alternative to autologous bone grafting. The aim of Chapter 4 is to investigate the therapeutic effectiveness of BioCaP with or without the two modes of BMP-2 in repairing a large cylindrical bone defect in sheep. To repair large-size bone defects, most bone-defect-filling materials in clinic need to obtain osteoinductivity either by mixing them with particulate autologous bone or adsorbing BMP-2. However, both approaches encounter various limitations. In Chapter 5, we hypothesized that our novel particles of biomimetic BMP-2-coprecipitated calcium phosphate (BMP2-cop.BioCaP) could serve as an independent and biodegradable osteoinducer to induce bone formation efficiently for the bone-defect-filling materials, e.g. deproteinized bovine bone (DBB). To enhance the therapeutic effect of DBB for bone defect repair, BMP2-cop.BioCaP particles was mixed with DBB. In Chapter 6, we investigated the therapeutic effect of BMP2-cop.BioCaP mixed with DBB in the treatment of critical-sized bone defects in sheep. As an alternative to an autologous bone graft, deproteinized bovine bone (DBB) is widely used in clinical dentistry. Although DBB provides an osteoconductive scaffold, it is not capable of enhancing bone regeneration because it is not sufficiently osteoinductive. In order to render DBB osteoinductive, BMP-2 has previously been incorporated into a three dimensional reservoir (a biomimetic calcium phosphate coating) on DBB, because it can effectively promote the osteogenic response by the slow delivery of BMP-2. We investigated the therapeutic effectiveness of such BMP-2/coating functionalized DBB granules in repairing large cylindrical (critical-sized) bone defects in sheep (Chapter 7). Heterodimeric BMPs exhibited several- or dozens-fold more effect than the respective homodimers in inducing in vitro osteoblastogenesis. In Chapter 8, we hypothesized that BMP2/7 heterodimer could facilitate more rapid bone regeneration in better quality than BMP2 and BMP7 homodimers in a peri-implant bone defect model in minipigs. Traditional Chinese Medicines (TCMs) have been recommended for bone regeneration and repair for thousands of years. Icariin, a typical flavonol glycoside, has been extracted from the Herba Epimedii (a native herb). Icariin can be locally delivered by biomaterials and has an osteoinductive potential for bone tissue engineering. The review in Chapter 9 focuses on the performance of icariin in bone tissue engineering and blended the information from icariin with the current knowledge relevant to molecular mechanisms and signal pathways. The osteoinductive potential and low price of icariin make it a very attractive candidate as a substitute of expensive osteoinductive protein − BMPs, or as a promoter to enhance the therapeutic effects of BMPs.

6

General introduction

REFERENCES 1. Lewandrowski KU, Gresser JD, Wise DL, Trantol DJ. Bioresorbable bone graft substitutes of different osteoconductivities: a histologic evaluation of osteointegration of poly(propylene glycol-co-fumaric acid)-based cement implants in rats. Biomaterials 2000;21:757-64. 2. Hamann C, Rauner M, Hohna Y, Bernhardt R, Mettelsiefen J, Goettsch C, et al. Sclerostin antibody treatment improves bone mass, bone strength, and bone defect regeneration in rats with type 2 diabetes mellitus. J Bone Miner Res 2012. 3. Schmitz JP, Hollinger JO. The critical size defect as an experimental-model for craniomandibulofacial nonunions. Clin Orthop Relat R 1986;299-308. 4. Van der Stok J, Van Lieshout EM, El-Massoudi Y, Van Kralingen GH, Patka P. Bone substitutes in the Netherlands - a systematic literature review. Acta Biomater 2011;7:739-50. 5. Muschler GF, Negami S, Hyodo A, Gaisser D, Easley K, Kambic H. Evaluation of collagen ceramic composite graft materials in a spinal fusion model. Clin Orthop Relat Res 1996;250-60. 6. Einhorn TA. Enhancement of fracture-healing. J Bone Joint Surg Am 1995;77:940-56. 7. Mordenfeld A, Hallman M, Johansson CB, Albrektsson T. Histological and histomorphometrical analyses of biopsies harvested 11 years after maxillary sinus floor augmentation with deproteinized bovine and autogenous bone. Clin Oral Implants Res 2010;21:961-70. 8. He YX, Zhang G, Pan XH, Liu Z, Zheng LZ, Chan CW, et al. Impaired bone healing pattern in mice with ovariectomy-induced osteoporosis: A drill-hole defect model. Bone 2011;48:1388-400. 9. Rustemeyer J, Bremerich A. Patients ' knowledge and expectations regarding dental implants: assessment by questionnaire. Int J Oral Max Surg 2007;36:814-17. 10. Hing KA. Bone repair in the twenty-first century: biology, chemistry or engineering? Philos Transact A Math Phys Eng Sci 2004;362:2821-50. 11. Williams DF. On the mechanisms of biocompatibility. Biomaterials 2008;29:2941-53. 12. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 13. Albrektsson T, Johansson C. Osteoinduction, osteoconduction and osseointegration. Eur Spine J 2001;10 Suppl 2:S96-101. 14. Urist MR. Bone: formation by autoinduction. Science 1965;150:893-9. 15. Cypher TJ, Grossman JP. Biological principles of bone graft healing. J Foot Ankle Surg 1996;35:413-7. 16. Younger EM, Chapman MW. Morbidity at bone graft donor sites. J Orthop Trauma 1989;3:192-5. 17. Thaller SR, Hoyt J, Dart A, Borjeson K, Tesluk H. Repair of experimental calvarial defects with Bio-Oss particles and collagen sponges in a rabbit model. J Craniofac Surg 1994;5:242-6. 18. Smith JD, Abramson M. Membranous vs endochondrial bone autografts. Arch Otolaryngol 1974;99:203-5. 19. Vuyk HD, Adamson PA. Biomaterials in rhinoplasty. Clin Otolaryngol Allied Sci 1998;23:209-17. 20. Ayerza MA, Aponte-Tinao LA, Abalo E, Muscolo DL. Continuity and function of patellar tendon host-donor suture in tibial allograft. Clin Orthop Relat Res 2006;450:33-8. 21. Muscolo DL, Ayerza MA, Aponte-Tinao LA, Ranalletta M. Use of distal femoral osteoarticular allografts in limb salvage surgery. J Bone Joint Surg Am 2005;87A:2449-55. 22. Buck BE, Malinin TI, Brown MD. Bone transplantation and human immunodeficiency virus. An estimate of risk of acquired immunodeficiency syndrome (AIDS). Clin Orthop Relat Res 1989;129-36. 23. Lewandrowski KU, Rebmann V, Passler M, Schollmeier G, Ekkernkamp A, Grosse-Wilde H, et al. Immune response to perforated and partially demineralized bone allografts. J Orthop Sci 2001;6:545-55.

7

Chapter 1

24. Moreau MF, Gallois Y, Basle MF, Chappard D. Gamma irradiation of human bone allografts alters medullary lipids and releases toxic compounds for osteoblast-like cells. Biomaterials 2000;21:369-76. 25. Carson JS, Bostrom MP. Synthetic bone scaffolds and fracture repair. Injury 2007;38 Suppl 1:S33-7. 26. Accorsi-Mendonca T, Conz MB, Barros TC, de Sena LA, Soares Gde A, Granjeiro JM. Physicochemical characterization of two deproteinized bovine xenografts. Braz Oral Res 2008;22:5-10. 27. Simion M, Fontana F, Rasperini G, Maiorana C. Vertical ridge augmentation by expanded-polytetrafluoroethylene membrane and a combination of intraoral autogenous bone graft and deproteinized anorganic bovine bone (Bio Oss). Clin Oral Implants Res 2007;18:620-9. 28. Araujo M, Linder E, Lindhe J. Effect of a xenograft on early bone formation in extraction sockets: an experimental study in dog. Clin Oral Implants Res 2009;20:1-6. 29. Schwartz Z, Weesner T, van Dijk S, Cochran DL, Mellonig JT, Lohmann CH, et al. Ability of deproteinized cancellous bovine bone to induce new bone formation. J Periodontol 2000;71:1258-69. 30. Dorozhkin SV. Bioceramics of calcium orthophosphates. Biomaterials 2010;31:1465-85. 31. Miron RJ, Zhang YF. Osteoinduction: a review of old concepts with new standards. J Dent Res 2012;91:736-44. 32. Yuan H, Fernandes H, Habibovic P, de Boer J, Barradas AM, de Ruiter A, et al. Osteoinductive ceramics as a synthetic alternative to autologous bone grafting. Proc Natl Acad Sci U S A 2010;107:13614-9. 33. Urist MR, Mc LF. Osteogenetic potency and new-bone formation by induction in transplants to the anterior chamber of the eye. J Bone Joint Surg Am 1952;34-A:443-76. 34. Yoshida K, Bessho K, Fujimura K, Kusumoto K, Ogawa Y, Tani Y, et al. Osteoinduction capability of recombinant human bone morphogenetic protein-2 in intramuscular and subcutaneous sites: an experimental study. J Craniomaxillofac Surg 1998;26:112-5. 35. Liu Y, Layrolle P, de Bruijn J, van Blitterswijk C, de Groot K. Biomimetic coprecipitation of calcium phosphate and bovine serum albumin on titanium alloy. J Biomed Mater Res 2001;57:327-35. 36. Wu G, Liu Y, Iizuka T, Hunziker EB. Biomimetic coating of organic polymers with a protein-functionalized layer of calcium phosphate: the surface properties of the carrier influence neither the coating characteristics nor the incorporation mechanism or release kinetics of the protein. Tissue Eng Part C Methods 2010;16:1255-65. 37. Wernike E, Hofstetter W, Liu Y, Wu G, Sebald HJ, Wismeijer D, et al. Long-term cell-mediated protein release from calcium phosphate ceramics. J Biomed Mater Res A 2010;92:463-74. 38. Stefanic M, Krnel K, Pribosic I, Kosmac T. Rapid biomimetic deposition of octacalcium phosphate coatings on zirconia ceramics (Y-TZP) for dental implant applications. Appl Surf Sci 2012;258:4649-56. 39. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 40. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 41. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106.

8

Chapter 2

Cell-mediated BMP-2 release from a novel dual drug delivery system promotes bone formation

Tie Liu, Gang Wu, Yuanna Zheng, Daniel Wismeijer, Vincent Everts, and Yuelian Liu.

Clinical Oral Implant Research, under revision, 2013.

9

Chapter 2

ABSTRACT

Objectives: In this study, a novel biomimetic calcium phosphate bone substitute (BioCaP) is introduced as a dual drug release system with two drug/protein delivery modes: protein is incorporated into (i) the interior of BioCaP (an internal depot); and (ii) a superficial calcium phosphate coating on BioCaP (a surface coated depot). Our aim is to investigate each of the two delivery modes of BioCaP. Our hypotheses are that (i) both of the drug delivery modes, in in vitro as well as in vivo environment, can achieve a sustained cell-mediated protein release; and (ii) BioCaP with these two delivery modes with incorporated bone morphogenetic protein-2 (BMP-2) promotes bone formation. Materials and Methods: Tablets of BioCaP were prepared with different carrying modes using bovine serum albumin (BSA) as model protein. The release of this protein was analyzed. Next, granules of BioCaP with different carrying modes of BMP-2 were implanted subcutaneously in rats. Samples were collected after five weeks for histomorphometric analysis. Results: In vitro data showed that the internal and surface coated depots of BSA resulted in a sustained osteoclast-mediated release, while the adsorbed BSA was rapidly released and this release was not affected by osteoclasts. In vivo data showed that the volume densities of bone, bone marrow, and blood vessels were significantly higher in samples where BMP-2 was incorporated internally or in the coating compared with granules with adsorbed growth factor. Osteoclast-like cells were associated with the granules and resorption lacunae were frequently observed. Conclusion: It is shown that different modes of incorporation of BMP-2 on and in BioCaP granules have a beneficial effect on the formation of ectopic bone. This dual drug release system makes BioCaP granule a promising tool for delivering multiple therapeutic agents for different clinical applications.

Keywords: Biomimetic; Calcium phosphate; Protein release; Osteoclast; Bone regeneration; BMP-2.

INTRODUCTION

Calcium phosphate (CaP)-based biomaterials are widely used for the regeneration of bone defects because of their similarity to bone, good biocompatibility, osteoconductivity and unlimited availability [1-5]. Currently, the major focus is to utilize CaP biomaterials as a drug delivery system by integration of different bioactive agents [6, 7]. By integrating bioactive agents, CaP biomaterials can acquire additional properties such as anti-infection [8], osteoinduction [9, 10], and anti-cancer properties [11, 12]. The therapeutic effect of these bioactive agents is highly dependent on their release kinetics [12, 13]. Usually, superficial adsorption results in a rapid and passive release which limits the effectiveness of many bioactive agents [7, 14-16], while a controlled release can optimize the therapeutic effects [11, 17, 18]. A controlled release of protein from CaP biomaterials can be achieved in various ways. Biomimetic coating is one of the attractive approaches for achieving a controlled release of protein/drug [17] and this coating seems to provide better results than materials, e.g. (bio) polymers that are physically or chemically mixed with a

10

Cell-mediated BMP-2 release from BioCaP ceramic/cement compound [5, 19, 20]. This biomimetic approach is to precipitate protein and CaP together in simulated body fluids under physiological conditions (37°C) [21, 22]. Consequently, a thin layer of biomimetic CaP coating with incorporated protein is formed on biomaterials. The protein release kinetics from carriers has usually been investigated by incubating them in physiological solutions such as cell culture media, phosphate buffered saline or simulated body fluid [5, 15, 23]. However, once a biomaterial is introduced into the body multiple factors might affect the protein release, such as cellular invasion and interstitial body fluid flow [24]. Therefore, in vivo, apart from the solubility of the biomaterials (physicochemical dissolution), cell-mediated resorption also plays a critical role in the degradation of the biomaterials [25-27]. This influences the protein release [28]. The cells involved in the degradation are mainly osteoclasts, foreign body giant cells, macrophages, and monocytes [24, 29]. It is important to understand how and to what extent these cells might influence the protein release from a drug-delivery system [28, 30]. Such a study may provide a guideline to predict the in vivo protein release kinetics. Recently, we have made a breakthrough in modifying the biomimetic coating approach. We have for the first time developed a novel biomimetic CaP bone substitute (BioCaP) as a dual release system. In this system protein and calcium phosphate were precipitated together to form BioCaP granules in which a depot of protein was incorporated in the center of the granules as an internal depot. Next, protein and calcium phosphate were co-precipitated onto the surface of these granules, thus creating a surface coated depot. This dual system provides an ideal model for delivery of different protein/drugs in two phases, an initial slow delivery phase (surface coated depot) and a delayed phase (internal depot). By adopting this system, a single drug can be administered in a more consistent manner due to this dual phase release or two different drugs can be administered simultaneously. Therefore, BioCaP granule might be considered as a promising tool for the orderly delivery of multiple therapeutic agents, such as antibiotics, osteogenic agents, and anti-cancer drugs for different clinical applications. There is, however, a need to study the delivery modes of this biomaterial to evaluate the amount and extent of cell mediated drug release during each phase. Our aim in the present study is to investigate each of the two delivery modes of BioCaP. Our hypotheses are that (i) both of the drug delivery modes, in in vitro as well as in vivo environment, can achieve a sustained cell-mediated protein release; and (ii) BioCaP with these two delivery modes with incorporated bone morphogenetic protein-2 (BMP-2) promotes bone formation. For this purpose, BioCaP tablets will be coated with labelled bovine serum albumin (BSA) to analyze protein release in vitro and BioCaP granules will be coated (incorporated) with BMP-2, and implanted subcutaneously in rats to analyse their capacity to induce bone formation in vivo.

MATERIALS AND METHODS

In vitro investigation Fabrication of BioCaP According to the biomimetic coating principle [21, 22], a supersaturated CaP

11

Chapter 2

solution [200 mM HCl, 20 mM CaCl2·2H2O, 680 mM NaCl, 10 mM Na2HPO4, and 250 mM Tris (pH 7.4)] was incubated in a shaking water bath (50 agitations/min) at 37°C. Protein was added to this CaP solution and co-precipitated (incorporated) into the interior of BioCaP (viz., internal depot of protein). After 24 hours of incubation, the precipitation was retrieved, gently washed by Milli-Q water, filtered and compressed to form a tablet (diameter: 5mm; thickness: 0.4 mm) using a vacuum exhaust filtering method with a vacuum filter (0.22-μm pore, Corning, NY, USA) and an air pump. After drying at room temperature, BioCaP tablets were used as such or ground into granules of different sizes. For sterilization, all the solutions were filtered with the vacuum filter (0.22-μm pore) before co-precipitation. All the procedures were performed under aseptic conditions. In this study, BioCaP tablets were used to investigate the in vitro cell-mediated release kinetics because cell seeding was easy on these tablets with their flat surface, and BioCaP granules were used for the in vivo investigation. The tablets and granules have the same physicochemical properties. It has been proven that the tablets and granules have the same surface structure by scanning electron microscopy, and the protein loading did not affect the surface structure of BioCaP.

Biomimetic coating procedure To introduce the surface coating of protein, the BioCaP was immersed in the CaP coating solution [40 mM HCl, 4 mM CaCl2·2H2O, 136 mM NaCl, 2 mM Na2HPO4, and 50 mM Tris (pH 7.4); total volume of 20 ml] for 24 hours at 37°C according to a biomimetic coating protocol [21, 31]. Protein was added to this CaP solution and thereafter co-precipitated in the CaP coating on the surface of BioCaP (viz., surface coated depot of protein).

Distribution of protein To confirm that the protein was incorporated into BioCaP, four BioCaP tablets were prepared with an internal or surface coated depot of bovine serum albumin which was labelled with fluorescein-isothiocyanate (FITC-BSA, 5.0μg/ml, Sigma, St. Louis, MO, USA). The distribution of FITC-BSA in BioCaP tablets was studied by analysing cross sections of the tablets. The samples were embedded in methylmethacrylate, sectioned, and ground [23]. 80-µm -thick sections were prepared and examined with a fluorescence microscopy (Leica, Wetzlar, Germany). The surface and cross-section morphology of BioCaP tablets was also investigated by scanning electron microscopy (SEM, XL20, FEI Company, The Netherlands) at an accelerating voltage of 10 kV.

Cell-mediated protein release kinetics Two different concentrations of BSA were incorporated into BioCaP tablets to study the protein release. Apart from FITC-BSA with a concentration of 5.0μg/ml, BSA labelled with Alexa Fluor® 555 (Alexa-BSA, invitrogen, Carlsbad, CA, USA) was used at a concentration of 0.5μg/ml, since FITC-BSA release from the samples with 0.5μg/ml is out of the detection range. According to the manufacture’s protocol Alexa-BSA is more easily detected at lower concentrations than FITC-BSA. Six groups were established for in vitro cell-mediated release: (1) BioCaP tablets with an internal depot of FITC-BSA; (2) BioCaP tablets with an internal depot of

12

Cell-mediated BMP-2 release from BioCaP

Alexa-BSA; (3) BioCaP tablets with a surface coated depot of FITC-BSA; (4) BioCaP tablets with a surface coated depot of Alexa-BSA; (5) BioCaP tablets with adsorbed FITC-BSA; and (6) BioCaP tablets with adsorbed Alexa-BSA. To adsorb BSA, BioCaP tablets were immersed in an aqueous protein solution (total volume of 20 ml per tablet) for 24 h at 37°C in plastic tubes. Passive and cell-mediated release of the variously labelled BSAs from the BioCaP tablets was monitored over a period of 16 days. For a cell-mediated release, bone marrow cells (BMC) were harvested from femurs and tibias of 6-week-old male mice and 1×106 BMCs were seeded on the tablets. The tablets were cultured in duplicate in α-MEM (Gibco BRL) containing 10% fetal calf serum (FCS) and 1% penicillin/streptomycin (100 U/ml and 100 μg/ml, respectively) with M-CSF (25 ng/ml, R&D Systems, Minneapolis, MN, USA) and RANKL (40 ng/ml, PreProtech, Rocky Hill, NJ, USA) [32]. BioCaP tablets without BMCs were incubated in α-MEM culture media to investigate the passive (spontaneous) release of BSAs. The culture medium was refreshed at 3-day intervals and used for spectrophotometric analysis (n=6 per time point) in a Fluorimeter (Spectramax M2, Molecular Devices, CA, USA), using 490 nm excitation and 504 nm emission wavelengths for FITC-BSA and using 540 nm excitation and 570 nm emission wavelengths for Alexa-BSA. Fluorescence readings were converted into the amount of protein by using a standard curve that was generated from a dilution series of labelled BSA prepared in 2 mL PBS. In this study, we chose osteoclasts for the analysis, since it has been demonstrated that the monocytes/macrophages has no significant effects on the release of coating-incorporated protein (Wernike et al. 2010a). At the end of the release experiments, the residual BSA in the BioCaP tablets was determined by dissolving the materials in 0.5 M ethylenediamine tetraacetic acid (EDTA, pH 8.0). The percentage of BSA released from the BioCaP tablets was calculated using the formula: [amount of the released fraction of BSA / total amount of BSA (amount of the released fraction + amount of the residual BSA of BioCaP tablets) ×100]. All cell culture experiments were performed at least three times. Tartrate resistant acid phosphatase (TRACP), which is a marker enzyme for osteoclasts, was used to identify the presence and number of these cells in the cultures. The tablets with cultured cells were washed with phosphate buffered saline (PBS) and fixed in 4% PBS buffered formaldehyde for 5 min and then stained for TRACP activity using the leucocyte acid phosphatase kit (Sigma). The nuclei were stained by incubating the cell cultures with diamidino-2-phenylindole-dihydrochloride (DAPI) in PBS. The number of TRACP+ cells with three or more DAPI+ nuclei was counted. Tablets with cells were also fixed, dried, and sputter-coated for SEM investigation [28]. To monitor the dissolution of BioCaP, the BioCaP tablets with or without cultured cells were gently washed with water, dried and weighted at each time point.

In vivo investigation Experimental animal model Adult male wistar rats (200–220g) were used as an animal model for ectopic bone formation.[33] A total of 30 rats were used, which was approved by Ethical Committee of School of Stomatology, Zhejiang Chinese Medical University. All the animal experiments were carried out according to the ethics laws and regulations of China. Throughout the

13

Chapter 2 study, the rats were treated following the guidelines of animal care established by Zhejiang Chinese Medical University

Grouping Two experimental and three control groups were analyzed (n=6 animals per group). The groups were: (1) BioCaP granules with an internal depot of BMP-2 (BioCaP int. BMP-2, experimental); (2) BioCaP granules with a surface coated depot of BMP-2 (BioCaP surf. BMP-2, experimental); (3) BioCaP granules with adsorbed BMP-2 (BioCaP ads. BMP-2, control); (4) BioCaP granules without a CaP coating or BMP-2 (BioCaP, control); and (5) BioCaP granules with a CaP coating but no BMP-2 (BioCaP CaP, control). Human recombinant BMP-2 (INFUSE® Bone Graft, Medtronic, USA) was introduced into the CaP solution or the coating solution at a concentration of 1 μg/ml. The amount of incorporated BMP-2 was determined using the ELISA technique [23]. About 35-μg of BMP-2 was finally incorporated into each sample of group (1) and (2). Hence, for group (3) as a control, 35-μg of BMP-2 was likewise loaded (0.22g of BioCaP granules) by adsorption [23].

Surgery and histology The surgery was performed under conditions of general anaesthesia using Sumianxin II (purchased from the Military Veterinary Institute, Quartermaster University of PLA, Chang Chun, China). Two samples of 0.22g of BioCaP granules were implanted in dorsal subcutaneous pockets in each rat, one on the left side and one on the right according to a random protocol as used in the previous studies [23, 34]. The samples were trapped by suturing the incision. . Five weeks after implantation, the samples were collected, fixed and embedded as previously reported [13, 35]. Applying a systematic random sampling [36], the samples were sawn vertical to the short axis, into 10–12 slices of 600 μm-thickness, 1 mm apart. All the slices of each sample were separately mounted on Plexiglas holders and polished. The thickness of the final histologic sections after polishing is about 500μm. Then they were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue [23] for the histological and histomorphometric analysis.

Histomorphometric analysis In this study, the space under the fibrous capsule that embraced the whole block of implants (subcapsular space) was taken as the reference space, as described in a previous study [23]. The reference space was estimated using Cavalieri’s methodology [37]. This involves measuring the cross sectional area of a defined number of tissue sections separated at a fixed distance through the reference volume. The cross sectional area of each section was estimated using a point-counting technique [38]. The volume densities of both the unmineralized and mineralized newly formed bone, bone marrow, blood vessels, fibrous capsular tissue, and multinucleated giant cells (on the surface of BioCaP) within the reference space were assessed using the point-counting technique [38]. The unmineralized bone is defined as the new bone with less density than the mineralized bone [13]. The volume density of a component (Va) is defined as its volume (Vb) per unit volume of reference space (Vc): Va=Vb/Vc.

14

Cell-mediated BMP-2 release from BioCaP

To evaluate the degradation of BioCaP, the volume of BioCaP before implantation (Da) and after 5 weeks of implantation (Db) was measured by the same histological method. Six samples were used for the measurement of the BioCaP volume before implantation. The percentage of non-degraded BioCaP (Dc) was defined as: Dc=(Da/Db)×100%.

Statistical analysis All data were presented as mean values and standard deviation (SD). The data were evaluated statistically using by a one way analysis of variance (ANOVA) using SPSS statistical software (version 16.0 for Windows). Post-hoc comparisons were made using Bonferroni's corrections with the level of significance set at p< 0.05.

RESULTS

In vitro results Surface topography and protein distribution There was no significant difference of the surface topography between BioCaP granule and table (Fig. 1). The cross section of BioCaP tablets showed that the FITC-BSA incorporated internally was distributed throughout the whole volume of the tablet in a net like configuration (Fig. 2A), while the surface coated FITC-BSA was found only in the crystalline surface coating of the tablet (Fig. 2B). The representative SEM micrographs depicted the cross section of BioCaP tablet with the internally incorporated FITC-BSA (Fig. 2C) and the crystalline coating with incorporated FITC-BSA (Fig. 2D).

Figure 1. SEM micrographs of the surface of BioCaP granule (A) and tablet (B).

15

Chapter 2

Figure 2. Fluorescence micrographs of BioCaP tablets with two different protein-carrying modes: A: FITC-BSA (green) distributed throughout the tablets (an internal depot); B: FITC-BSA distributed in the crystalline coating layer (a surface-coated depot); C: Scanning electron micrographs of sectioned BioCaP tablet (the center part); D: Scanning electron micrographs of a section of the crystalline coating layer on the surface of BioCaP tablet. Bar= 100μm in (A). Bar= 20μm in (B) and (D). Bar= 10μm in (C).

Passive release of BSAs from BioCaP tablets The passive release, that is without cells, of BSAs [FITC-BSA (5.0μg/ml) and Alexa-BSA (0.5μg/ml)] from BioCaP tablets was monitored over a period of 16 days (Fig. 3). A burst release occurred in all groups within the first 4 days of incubation. The internal or the surface coated depot had a significantly lower burst release than the adsorbed fraction in the 4 day period. No significant difference was found between the internal and the surface coated depots of protein. The adsorbed depot of BSAs was released rapidly within 16 days, whereas the release of the internal or surface coated depots occurred at a steady rate after the 4th day up to the 16th day. Alexa-BSA had a significantly lower burst release than FITC-BSA within 4 days in all groups.

16

Cell-mediated BMP-2 release from BioCaP

Figure 3. Graphs show the release percentage profiles of FITC-BSA (5 μg/ml) and Alexa-BSA (0.5 μg/ml) with/without cells from an internal (A and D), a surface-coated (B and E), and an adsorbed depot (C and F) respectively. Mean values are represented ± SD (n = 6 for each group). #p< 0.001.

Cell-mediated release of BSAs from BioCaP tablets The cell-mediated release of BSAs from BioCaP tablets was monitored over a period of 16 days. The samples with the internal or the surface coated depot showed a sustained cell-mediated of BSAs release (Fig. 3A, B, D, E), while the samples with the adsorbed depot showed a rapid release (Fig. 3C, F). The cell-mediated release of BSAs from an internal depot was significantly higher than the passive (without cells) release from day 7 until day 16 (Fig. 3A, D). At the 16 day time point, the initial amount of FITC-BSA and Alexa-BSA had been decreased by 55% and 25%, respectively. The cell-mediated release of BSAs from a surface coated tablet was also significantly higher than the passive release from day 4 until day 16 (Fig. 3B, E). After 16 days the initial amount of FITC-BSA and Alexa-BSA was decreased by 45% and 40%, respectively. Cells appeared to have no influence on the protein release from the adsorbed depot (Fig. 3C, F). At day 16, the initial adsorbed amount of FITC-BSA and Alexa-BSA was decreased by 90% and 80%, respectively.

Osteoclast formation and the degradation of BioCaP tablet Representative cultures of cells on BioCaP without and with coating are shown in Fig. 4A and B after staining for TRACP and DAPI respectively. No significant differences were found between the two differently labelled BSAs as regards osteoclast formation or the degradation of BioCaP tablets. The number of TRACP positive multinucleated osteoclasts on the BioCaP tablet with FITC-BSA is shown in Fig. 4C. The number of osteoclasts on the uncoated BioCaP tablet, with an internal depot of BSA, was significantly lower than on the BSA coated BioCaP tablet on day 4 and 7 (Fig. 4C).

17

Chapter 2

From day 7, osteoclast numbers gradually decreased on uncoated or coated BioCaP tablet. The weight of coated and uncoated BioCaP tablets was monitored over a period of 16 days (Fig. 4D). During the 16 day incubation period, the weight of uncoated BioCaP tablets in the absence of cells significantly decreased from 35.23 ± 0.33mg to 28.86 ± 0.68mg, while the weight of the coated BioCaP tablet showed no significant changes. When cells were seeded on the uncoated BioCaP tablets, there was a significant decline of the weight of BioCaP from day 7 compared with the group without cells. The cells did not influence the weight of the coated BioCaP. Scanning electron microscopy (SEM) showed the surface morphology of BioCaP tablet with internal or surface coated FITC-BSA in Fig. 5A and 5C respectively. The incorporation of protein did not significantly change the surface morphology of BioCaP. Actively resorbing osteoclasts associated with typical resorption lacunae were observed on BioCaP with internal or surface coated FITC-BSA in Fig. 5B and 5D respectively.

Figure 4. Tartrate-resistant acid phosphatase (TRACP) -positive and multinucleated osteoclasts (arrows) generated from murine bone marrow cells (BMCs) are visible on A: BioCaP tablet and B: the CaP coating of BioCaP tablet. C: The graph depicts that the number of TRACP+ multinucleated cells were significantly higher for BioCaP with coating at day 4 and day 7 as compared to non-coated BioCaP tablets. D: The graph shows the weight of BioCaP tablets during the cell culture. The presence of cells resulted in a significant decline of the weight of BioCaP

18

Cell-mediated BMP-2 release from BioCaP without coating from day 7 compared to the no cell group. The weight of BioCaP with coating showed no significant changes. Mean values are represented ± SD (n = 6 for each group). *p< 0.05; #p< 0.001. Bar=50μm in (A) and (B).

Figure 5. Scanning electron microscopy was performed to analyze the surface morphology of BioCaP tablets and resorbing osteoclasts. A: BioCaP tablet bearing internal depot of FITC-BSA; B: Osteoclasts seemed to stick in their resorption lacunae on BioCaP (arrow head); C: BioCaP tablet bearing a surface-coated depot of FITC-BSA; D: Osteoclasts seemed to stick in their resorption lacunae on the coating of BioCaP (arrow head).

In vivo results Formation of bone, bone marrow, blood vessels, and fibrous tissue Five weeks after implantation in rats, bone and bone marrow were associated only with BMP-2-functionalized BioCaP granules. This was found for all types of granules that contained BMP-2 (see Figs 6A-F). In the two groups without BMP-2 neither bone nor bone marrow was found (Fig. 6G, H). The volume density of unmineralized and mineralized bone (Fig. 7A), bone marrow (Fig. 7B) and blood vessels (Fig. 7C) was significantly higher in association with BioCaP granules with internal or surface coated BMP-2 as compared to granules with adsorbed BMP-2. The volume density of fibrous capsular tissue, however, was significantly higher in the adsorption group (Fig. 7D).

19

Chapter 2

Figure 6. Light micrographs of cross-sections through BioCaP granules, 5 weeks after subcutaneous implantation in rats. BioCaP granules with an internal depot of BMP-2 (A), a surface-coated depot of BMP-2 (B), or an adsorbed depot of BMP-2 (C). Newly formed mineralized bone (reddish, white asterisk) and unmineralized bone (purple, black asterisk) has been formed between the BMP-2-functionalized BioCaP granules or deposited on these BioCaP granules. The bone quantity in (C) was smaller than in (A) and (B). Higher-magnification of BioCaP granules with an internal depot of BMP-2 (D), a surface-coated depot of BMP-2 (E), or an adsorbed depot of BMP-2 (F). BioCaP was observed in close contact with bone (white asterisk) and bone marrow (M) in (D) and (E). Osteoblasts (arrow), fibrous capsular tissue (FT) and blood vessels (arrow head) were also observed. The bone in (F) displayed unmineralized appearance (black asterisk). Neither bone nor bone marrow was formed on or around the non-functionalized BioCaP granules that had a CaP coating but no BMP-2 (G) or those without coating or BMP-2 (H). The sections were stained with McNeal’s Tetrachrome, basic Fuchsine, and Toluidine Blue. Bar= 500μm in (A), (B), (C), (G), and (H). Bar= 100μm in (D), (E), and (F).

BioCaP degradation The histomorphometric evaluation of the degradation of BioCaP revealed that the degradation was higher in the groups with adsorbed or in those without BMP-2 (Fig. 7E). About 60% of BioCaP granules were degraded in these two groups compared with about 20% for BioCaP in the other three groups.

20

Cell-mediated BMP-2 release from BioCaP

Figure 7. Graphs depicting the volume density of unmineralized and mineralized bone (A), the bone marrow (B), blood vessels (C), and fibrous capsular tissue (D), and the percentage of non-degraded BioCaP granules (E). Mean values are represented ± SD (n = 6 for each group). *p< 0.05; +p< 0.01; #p< 0.001.

The response of the cells Multinucleated osteoclast-like cells were observed not only on the surface of BioCaP granules (Fig. 8A ,B), but also on the newly-formed bone (Fig. 8A, C). The cells were associated with resorption pits on the surface of BioCaP (Fig. 8B). The volume density of multinucleated osteoclast-like cells on the surface of BioCaP granules was significantly higher in association with BioCaP with adsorbed BMP-2 compared with

21

Chapter 2

BioCaP with internal or surface coated BMP-2 (Fig. 8D).

Figure 8. Light micrographs of a cross-section through BioCaP granules with internal depot of BMP-2 after 5 weeks of subcutaneous implantation in rats (A). Multinucleated giant cells could be observed on the BioCaP surface (arrow, B) or on the bone surface (arrow, C) at higher magnification. Newly formed bone (asterisk); bone marrow (M); blood vessels (V), and fibrous capsulate tissues (FT). The sections were stained with McNeal’s Tetrachrome, basic Fuchsine, and Toluidine Blue. Bar= 100μm in (A). Bar= 50μm in (B) and (C). Graphs depicting the volume density of multinucleated giant cells on the BioCaP surface (D). Mean values are represented ± SD (n = 6 for each group). +p< 0.01; #p< 0.001.

DISCUSSION

The main objective of the present study is to investigate BioCaP as a potential dual drug/protein delivery system for the slow and sustained delivery of different bioactive agents. The data from the present study confirmed that a gradual, sustained and cell-mediated release of bioactive agents can be achieved in vitro as well as in vivo using BioCaP. Moreover, we showed that the BioCaP granules containing BMP-2 are osteoinductive. In addition, the histological analysis confirmed the excellent biocompatibility of BioCaP. In order to analyse in vitro cell-mediated protein release we used labelled BSA instead of the very costly BMP-2 [28, 39]. BSA has been often utilized as a substitute for BMP-2 [22, 23] because the release kinetics of BSA and BMP-2 are similar [40-42]. In the present study, BSA presented a sustained cell-mediated release from the two delivery modes of BioCaP, thus suggesting that the BMP-2 release from BioCaP might be sustained as well. It is known that the adsorption of BMP-2 on materials is always associated with a high-dose burst release. This results in a poor osteoinduction [23, 43]. In the current study, the histomorphometric analysis demonstrated that the two modes in which BMP-2 was incorporated into the BioCaP granules resulted in a better osteoinduction than BMP-2 adsorbed onto the granules. The two groups not only induced more bone, bone marrow and blood vessels, but also less fibrous capsular tissue was found. Since bone marrow and blood vessels are important sources of oxygen, nutrients, signalling

22

Cell-mediated BMP-2 release from BioCaP molecules, and pluripotent progenitor cells for osseous tissue [44], the presence of these structures and cells in the experimental groups helps to enhance bone formation. Previous studies have demonstrated that the slow release of BMP-2 enhances osteoinductivity [13], and leads to a higher osteoinductive efficiency than the adsorption [23]. Therefore, given the in vivo results, we speculate that the two modes resulted in a sustained release of BMP-2, since they achieved a better osteoinduction than the adsorbed mode. However, when BioCaP was implanted in this ectopic rat model, endogenous protein including BMPs can be adhered on the graft surface and thus induce different cell-mediated resorption processes. The BMP-2 release may result in a completely different protein release pattern. The detection of the BMP-2 release in vivo was studied apart from this study (unpublished data). The degradability of a CaP-based material is very important for the in vivo longevity and efficacy of its biological effects [45]. In the present study, the findings indicate that BioCaP is biodegradable. The material degradation is associated with its dissolubility and the cell-mediated resorption [30]. The degradation rate of BioCaP granules was significantly lower for those in which BMP-2 was incorporated internally compared with those with adsorbed BMP-2 or with those with no BMP-2. A reason for this could be that newly formed bone tissues covered the surface of the granules, thereby preventing their degradation. In addition, both in vitro and in vivo findings indicate that the coating can prevent and delay the degradation of underlying BioCaP, even though the coating is biodegradable [13, 34]. The coating and BioCaP were all biomimetically formed by precipitation of calcium phosphate. However, their surface structures were totally different. The coating had a crystalline surface, while the surface of BioCaP seemed to be amorphous. A higher number of osteoclasts were found on the coated BioCaP tablets compared with the BioCaP without coating. This indicates that the different physicochemical properties of BioCaP and the coating may affect the formation of osteoclasts. Moreover, in vivo the two different modes of BMP-2 delivery resulted in a significantly lower number of multinucleated cells compared with the adsorption mode. It has been shown that BMP-2 may exert a dual concentration dependent effect [13, 46]. At low doses, it stimulates the recruitment, proliferation and differentiation of osteoprogenitor cells, whereas at high doses, it induces the recruitment, formation and activation of osteoclasts. In a previous study we have demonstrated that a slow and steady release of BMP-2 from the coating suppresses the formation of multinucleated cells [23]. We assume that not only the physicochemical property of materials affects the formation of multinucleated cells but also the way BMP-2 is released and its local dose influences this process. Our previous study has demonstrated that BMP-2 incorporated into biomimetic coatings can retain its biological activity [31]. However, the BMP-2 activity might be influenced by the digestion of multinucleated osteoclast-like cells. The previous study showed that the growth factors released through osteoclast-mediated manner could significantly promoted osteoblastogenesis in vitro [47]. Therefore, we may assume the proteinaceous BMP-2 may maintain its activity in this study. However, further investigations are needed to prove the percentages of BMP-2 activity. In this study, the rodent ectopic model may be sensitive to BMP-2. Therefore, larger animals may need more BMP-2 to respond appropriately. In our on-going study, we

23

Chapter 2 implanted BioCaP with the two delivery modes of BMP-2 into the bone defects in sheep compared with the therapeutic effect of autologous bone. One concern associated with the use of BioCaP is its biocompatibility. Histological analysis revealed that the newly formed bone was deposited directly on the BioCaP surface and bone marrow was in close contact with BioCaP. This suggests that the used BioCaP granules are highly biocompatible. Our previous study has demonstrated that the coating was highly biocompatible and osteoconductive [34]. Our findings strongly suggest that BioCaP granules can be used as an attractive protein delivery vehicle. This seems particularly true for the granules in which the growth factor was incorporated internally. In addition, the use of the coating can offer an alternative for slow release. By combining the two protein carrying modes, BioCaP can be a dual release system for a sequential delivery of different proteins/drugs. This combination could be applicable for a variety of clinical applications. For example, osteogenic agents can be incorporated into the interior of BioCaP, and at the same time antibiotics can be incorporated into the surface coating. This could be considered as a new strategy for the treatment of bone defects caused by peri-implantitis. In conclusion, it was shown that BioCaP with an internal or surface coated depot of protein has the capacity to maintain a slow and sustained protein release in the presence of osteoclasts in vitro. Both modes of delivering of BMP-2 with the use of BioCaP make these granules efficient osteoinductive compounds and suppress the formation of multinucleated giant cells in vivo. The in vivo detection of the BMP-2 release by using a radioactive labelling method was studied apart from this study (unpublished data). The dual drug release system renders BioCaP granules promising tools for an orderly delivery of multiple therapeutic agents, such as antibiotics, osteogenic agents, and anti-cancer drugs for different clinical applications.

ACKNOWLEDGMENTS

We would like to thank Dr. Afsheen Tabassum for the assistance and thank Prof. Dr. Tony Hearn as a native speaker for editing the grammar.

24

Cell-mediated BMP-2 release from BioCaP

REFERENCES 1. Bauer TW, Muschler GF. Bone graft materials. An overview of the basic science. Clin Orthop Relat Res 2000;10-27. 2. Bohner M, Galea LG, Lemaitre J, Kohler T, Muller R. Bone substitute: Transforming beta-tricalcium phosphate porous scaffolds into monetite. Biomaterials 2008;29:3400-07. 3. Dorozhkin SV. Bioceramics of calcium orthophosphates. Biomaterials 2010;31:1465-85. 4. Kurashina K, Kurita H, Wu Q, Ohtsuka A, Kobayashi H. Ectopic osteogenesis with biphasic ceramics of hydroxyapatite and tricalcium phosphate in rabbits. Biomaterials 2002;23:407-12. 5. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 6. Verron E, Khairoun I, Guicheux J, Bouler JM. Calcium phosphate biomaterials as bone drug delivery systems: a review. Drug Discov Today 2010;15:547-52. 7. Lode A, Wolf-Brandstetter C, Reinstorf A, Bernhardt A, Konig U, Pompe W, et al. Calcium phosphate bone cements, functionalized with VEGF: release kinetics and biological activity. J Biomed Mater Res A 2007;81:474-83. 8. Sudo A, Hasegawa M, Fukuda A, Uchida A. Treatment of infected hip arthroplasty with antibiotic-impregnated calcium hydroxyapatite. J Arthroplasty 2008;23:145-50. 9. Seeherman H, Azari K, Bidic S, Rogers L, Li XJ, Hollinger JO, et al. rhBMP-2 delivered in a calcium phosphate cement accelerates bridging of critical-sized defects in rabbit radii. J Bone Joint Surg Am 2006;88:1553-65. 10. Habraken WJ, Boerman OC, Wolke JG, Mikos AG, Jansen JA. In vitro growth factor release from injectable calcium phosphate cements containing gelatin microspheres. J Biomed Mater Res A 2009;91:614-22. 11. Lebugle A, Rodrigues A, Bonnevialle P, Voigt JJ, Canal P, Rodriguez F. Study of implantable calcium phosphate systems for the slow release of methotrexate. Biomaterials 2002;23:3517-22. 12. Uchida A, Shinto Y, Araki N, Ono K. Slow release of anticancer drugs from porous calcium hydroxyapatite ceramic. J Orthop Res 1992;10:440-5. 13. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106. 14. Lasserre A, Bajpai PK. Ceramic drug-delivery devices. Crit Rev Ther Drug Carrier Syst 1998;15:1-56. 15. Autefage H, Briand-Mesange F, Cazalbou S, Drouet C, Fourmy D, Goncalves S, et al. Adsorption and release of BMP-2 on nanocrystalline apatite-coated and uncoated hydroxyapatite/beta-tricalcium phosphate porous ceramics. J Biomed Mater Res B Appl Biomater 2009;91:706-15. 16. Liu Y, Hunziker EB, Randall NX, de Groot K, Layrolle P. Proteins incorporated into biomimetically prepared calcium phosphate coatings modulate their mechanical strength and dissolution rate. Biomaterials 2003;24:65-70. 17. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 18. Ferraz MP, Mateus AY, Sousa JC, Monteiro FJ. Nanohydroxyapatite microspheres as delivery system for antibiotics: Release kinetics, antimicrobial activity, and interaction with osteoblasts. J Biomed Mater Res A 2007;81A:994-1004. 19. Bose S, Tarafder S. Calcium phosphate ceramic systems in growth factor and drug delivery for bone tissue engineering: a review. Acta Biomater 2012;8:1401-21. 20. Felix Lanao RP, Leeuwenburgh SC, Wolke JG, Jansen JA. Bone response to fast-degrading, injectable calcium phosphate cements containing PLGA microparticles. Biomaterials 2011;32:8839-47. 21. Liu Y, Layrolle P, de Bruijn J, van Blitterswijk C, de Groot K. Biomimetic coprecipitation of

25

Chapter 2

calcium phosphate and bovine serum albumin on titanium alloy. J Biomed Mater Res A 2001;57:327-35. 22. Wu G, Liu Y, Iizuka T, Hunziker EB. Biomimetic coating of organic polymers with a protein-functionalized layer of calcium phosphate: the surface properties of the carrier influence neither the coating characteristics nor the incorporation mechanism or release kinetics of the protein. Tissue Eng Part C Methods 2010;16:1255-65. 23. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 24. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20:86-100. 25. Teitelbaum SL. Bone resorption by osteoclasts. Science 2000;289:1504-08. 26. Leeuwenburgh S, Layrolle P, Barrere F, de Bruijn J, Schoonman J, van Blitterswijk CA, et al. Osteoclastic resorption of biomimetic calcium phosphate coatings in vitro. J Biomed Mater Res A 2001;56:208-15. 27. Schilling AF, Linhart W, Filke S, Gebauer M, Schinke T, Rueger JM, et al. Resorbability of bone substitute biomaterials by human osteoclasts. Biomaterials 2004;25:3963-72. 28. Wernike E, Hofstetter W, Liu Y, Wu G, Sebald HJ, Wismeijer D, et al. Long-term cell-mediated protein release from calcium phosphate ceramics. J Biomed Mater Res A 2010;92:463-74. 29. Liu Y, Wu G, de Groot K. Biomimetic coatings for bone tissue engineering of critical-sized defects. J R Soc Interface 2010;7 Suppl 5:S631-47. 30. Zhang Z, Egana JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al. Cell-based resorption assays for bone graft substitutes. Acta Biomater 2012;8:13-9. 31. Liu Y, Hunziker EB, Layrolle P, De Bruijn JD, De Groot K. Bone morphogenetic protein 2 incorporated into biomimetic coatings retains its biological activity. Tissue Eng Part A 2004;10:101-8. 32. Azari A, Schoenmaker T, de Souza Faloni AP, Everts V, de Vries TJ. Jaw and long bone marrow derived osteoclasts differ in shape and their response to bone and dentin. Biochem Biophys Res Commun 2011;409:205-10. 33. Wu G, Liu Y, Iizuka T, Hunziker E. The effect of a slow mode of BMP-2 delivery on the inflammatory response provoked by bone-defect-filling polymeric scaffolds. Biomaterials 2010;31:7485-93. 34. Liu Y, de Groot K, Hunziker EB. BMP-2 liberated from biomimetic implant coatings induces and sustains direct ossification in an ectopic rat model. Bone 2005;36:745-57. 35. Liu Y, Huse RO, de Groot K, Buser D, Hunziker EB. Delivery mode and efficacy of BMP-2 in association with implants. J Dent Res 2007;86:84-9. 36. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc 1987;147:229-63. 37. Cavalieri B. Geometria Indivisibilibus Continuorum. 1635. Reprinted as Geometria degli Indivisibili. Torino: Unione Tipografico-Editorice Torinese 1966. 38. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey. Am J Physiol 1990;258:L148-56. 39. Lee M, Chen TT, Iruela-Arispe ML, Wu BM, Dunn JC. Modulation of protein delivery from modular polymer scaffolds. Biomaterials 2007;28:1862-70. 40. Yilgor P, Tuzlakoglu K, Reis RL, Hasirci N, Hasirci V. Incorporation of a sequential BMP-2/BMP-7 delivery system into chitosan-based scaffolds for bone tissue engineering. Biomaterials 2009;30:3551-9. 41. Yilgor P, Sousa RA, Reis RL, Hasirci N, Hasirci V. Effect of scaffold architecture and BMP-2/BMP-7 delivery on in vitro bone regeneration. J Mater Sci Mater Med 2010;21:2999-3008.

26

Cell-mediated BMP-2 release from BioCaP

42. Basmanav FB, Kose GT, Hasirci V. Sequential growth factor delivery from complexed microspheres for bone tissue engineering. Biomaterials 2008;29:4195-204. 43. Schwarz F, Rothamel D, Herten M, Ferrari D, Sager M, Becker J. Lateral ridge augmentation using particulated or block bone substitutes biocoated with rhGDF-5 and rhBMP-2: an immunohistochemical study in dogs. Clin Oral Implants Res 2008;19:642-52. 44. Le Nihouannen D, Saffarzadeh A, Gauthier O, Moreau F, Pilet P, Spaethe R, et al. Bone tissue formation in sheep muscles induced by a biphasic calcium phosphate ceramic and fibrin glue composite. J Mater Sci-Mater M 2008;19:667-75. 45. Tanuma Y, Anada T, Honda Y, Kawai T, Kamakura S, Echigo S, et al. Granule size-dependent bone regenerative capacity of octacalcium phosphate in collagen matrix. Tissue Eng Part A 2012;18:546-57. 46. Zara JN, Siu RK, Zhang X, Shen J, Ngo R, Lee M, et al. High doses of bone morphogenetic protein 2 induce structurally abnormal bone and inflammation in vivo. Tissue Eng Part A 2011;17:1389-99. 47. Howard GA, Bottemiller BL, Turner RT, Rader JI, Baylink DJ. Parathyroid hormone stimulates bone formation and resorption in organ culture: evidence for a coupling mechanism. Proc Natl Acad Sci U S A 1981;78:3204-8.

27

Chapter 2

28

Chapter 3

Preparation and characteristics of osteoinductive biomimetic calcium phosphate material: in vitro and in vivo study

Tie Liu, Gang Wu, Yuanna Zheng, Afsheen Tabassum, Daniel Wismeijer, Vincent Everts, and Yuelian Liu.

Submitted, 2013.

29

Chapter 3

ABSTRACT

Objectives: A biomimetic calcium phosphate (BioCaP) bone substitute was developed with two protein-delivery modes: one mode by which protein was incorporated in the interior of BioCaP (internally-incorporated mode); and one by which protein was coated on the outside of BioCaP (coating-incorporated mode). The aim of this study was to prepare and evaluate the physical and chemical properties of BioCaP, and the ability for protein loading and release in a long period, and using micro-CT analysis to evaluate the degradation of BioCaP and bone formation. Material and methods: BioCaP was prepared by refining a well-established biomimetic protocol. The compressive strength of BioCaP was assessed by a compressive strength machine. The structure and morphology of BioCaP was analyzed by X-ray diffraction (XRD) analysis and scanning electron microscope (SEM). The two phases of precipitated BioCaP were visualized using SEM. An energy dispersive X-ray (EDX) source was used for the chemical composition analysis. The protein release was analyzed in vitro (35 days). Human osteoclasts were used for testing the cell-based degradation of BioCaP. A micro-CT method for the evaluation of graft material and bone was applied by using an “onion-peeling” algorithm and specific threshold settings in an ectopic rat model. Results: BioCaP exhibited bone-like mechanical strength and the characteristics of calcium-deficient apatite. The Ca/P molar ratio of BioCaP was about 1.48. The granules with internally- or coating-incorporated protein exhibited a slow release in vitro. Osteoclasts seeded on the granules were shown to resorb the BioCaP. Micro-CT analysis showed three-dimensional reconstructions of BioCaP and new bone formation in vivo. Conclusion: BioCaP granules were developed as an osteoinductive bone substitute and a vehicle for protein/drug slow release. The micro-CT method provides an alternative for the evaluation of bone substitute and bone formation.

Keywords: Biomimetic calcium phosphate material; Protein slow release; Human osteoclasts; Micro-CT; BMP-2

INTRODUCTION

Large-size bone defects, which exceed the self-healing capacity of bone tissue, can be treated using autografts, allografts, xenografts, and synthetic materials [1, 2]. Autografts (gold standard) can be seen as an osteoconductive scaffold, containing osteoinductive cytokines and osteogenic cells. However, it is always associated with limited availability as well as with donor-site pain and morbidity [3]. At present, synthetic calcium phosphate (CaP)-based bone substitutes have become widely used in the clinic [4, 5]. However, most of these clinically used CaP bone substitutes lack intrinsic osteoinductivity, which is an essential property to realize osseous restoration of large-size bone defects [2]. The osteoinductivity of such CaP material can be conferred by using osteogenic growth factors such as bone morphogenetic protein-2 (BMP-2). However, the rather harsh and non-physiological conditions during the production process preclude the incorporation of biological active proteinaceous molecules into the interior of CaP material. The usual way to circumvent this is to adsorb osteogenic agents onto the CaP surface [6, 7]. However, such adsorbed molecules are frequently associated with a rapid and burst

30

Preparation and characteristics of BioCaP release, which results in a poor osteogenic potential [8-10]. Using a biomimetic mineralization approach, growth factors were co-precipitated into the latticework of crystalline calcium phosphate on titanium implants or other biomaterials and were shown to be released locally in a slow manner [2, 11]. Such a slow release has been shown to be beneficial for the effect of different growth factors such as BMP-2 and vascular endothelial growth factor (VEGF). The slow delivery of BMP-2 enhanced osteoinduction [12], and the slow delivery of VEGF promoted vascularization [13]. Furthermore, since bone regeneration is a coordinated cascade of events regulated by several growth factors, the local sequential delivery of VEGF and BMP-2 could enhance bone formation compared with BMP-2 alone [14]. Therefore, there is a need to develop a carrier material that has the capacity to sequentially and slowly deliver growth factors. Previous study has introduced the calcium-phosphate (BioCaP) granular bone substitute that can act as a dual delivery model [15]. We incorporated protein in two ways: in the interior of the granules and in their surface coating. These two modes of incorporation of BMP-2 rendered BioCaP osteoinductive efficiently. However, the physical and chemical properties of BioCaP is unclear. The purpose of the present study was i) to evaluate the BioCaP physicochemical properties such as mechanical strength, dissolution and degradability; ii) to investigate the in vitro release kinetics of the two protein-delivery modes of BioCaP during a period of 35 days; iii) using human osteoclasts to evaluate the cell-based resorption of BioCaP; and iv) using micro-CT to distinguish BioCaP and new bone and determine whether BioCaP with these two delivery modes of BMP-2 can efficiently induce ectopic bone formation in rats.

MATERIALS AND METHODS

Fabrication of biomimetic calcium phosphate (BioCaP) bone substitute BioCaP was fabricated by refining a well-established biomimetic mineralization technique, which has been described in a previous study [15]. Briefly, a CaP solution (200 mM HCl, 20 mM CaCl2·2H2O, 680 mM NaCl, and 10 mM Na2HPO4) buffered by TRIS (250 mM) to a pH of 7.4. Rapid precipitation appeared at pH of 6.25. In order to sterilize the CaP solution it was filtered with a vacuum filter (0.22-μm pore) before buffering. All the following procedures were performed under aseptic conditions. After buffering, the solution was incubated in a shaking water bath (50 agitations/min) at 37°C for 24 hours. Thereafter the solution was removed. The precipitated material was gently washed by Milli-Q water, filtered and compressed to a block using a vacuum exhaust filtering method with a filter (0.22-μm pore, Corning, NY, USA) and a vacuum pump. Before drying, the block can be shaped differently such as in a tablet or cylinder shape. After drying at room temperature for 2 hours, the hardened block can be ground and filtered to obtain different sizes of granules using metallic mesh filters. The protein introduced into the CaP solution can be co-precipitated into the interior of BioCaP, viz., the internally-incorporated depot of protein (Fig. 1).

Biomimetic coating procedure and protein incorporation The superficial coating was deposited on BioCaP according to the procedure described before [16]. Briefly, BioCaP was incubated in the coating solution [40 mM HCl, 4 mM

31

Chapter 3

CaCl2·2H2O, 136 mM NaCl, 2 mM Na2HPO4, and 50 mM TRIS (pH 7.4); total volume of 20 ml] in a shaking water bath (50 agitations/min) at 37°C for 24 hours. The protein present in this coating solution can be co-precipitated into the coating on the BioCaP surface, viz., the coating-incorporated depot of protein (Fig. 1).

Figure 1. Schematic illustration of BioCaP granules with two cytokine-carrying (delivery) modes: the internally- or the coating- incorporated modes.

Physicochemical properties of BioCaP bone substitute The compressive strength of BioCaP was assessed. BioCaP was made into cylinders (diameter, 5mm; height, 8 mm; n=6). Each cylinder was crushed at a crosshead speed of 1 mm/min, using a compressive strength machine (Instron 6022, High Wycombe, Bucks, U.K). BioCaP granules with no protein or those with internally-incorporated bovine serum albumin (BSA, 1µg/ml or 20µg/ml, Sigma, St. Louis, MO, USA) or BMP-2 (1µg/ml, INFUSE® Bone Graft, Medtronic, USA) were ground into fine powder for X-ray diffraction (XRD) analysis. XRD patterns of the samples were recorded with a vertically mounted diffractometer system (Bruker D8 Advance, Bruker AXS, Germany), using Ni-filtered Cu Kβ radiation generated at 40 kV and 40 mA. Specimens were scanned from 5° to 60° 2θ (where θ is the Bragg angle) in continuous mode.

32

Preparation and characteristics of BioCaP

The morphology of precipitated BioCaP (BioCaP dried on a glass plate at room temperature) and granules with or without protein was visualised using a scanning electron microscope (SEM, XL20, FEI Company, the Netherlands), under an accelerating voltage of 10 kV after being sputter-coated with gold. An energy dispersive X-ray (Voyager, Eindhoven, The Netherlands) source was attached to the apparatus for the chemical composition analysis. To confirm whether protein was incorporated into the interior of the BioCaP granules or in its coating, the presence and distribution of BSA labelled with fluorescein-isothiocyanate (FITC-BSA, Sigma, St. Louis, MO, USA) in BioCaP granules was analysed by using cross sections of the granules. The samples were embedded in methylmethacrylate, sectioned, and ground [11]. A series of 80-µm -thick sections were prepared for analysis by fluorescence microscopy. Micrographs were taken with a digital camera (Leica, Wetzlar, Germany) mounted on an inverted light microscope (Leica) equipped with a fluorescence lamp.

Protein and Ca2+ release in vitro To study the protein release from BioCaP granules, bovine serum albumin (BSA) labelled with fluorescein-isothiocyanate (FITC-BSA, Sigma, St. Louis, MO, USA) or with Alexa Fluor® 555 (Alexa-BSA, invitrogen, Carlsbad, CA, USA) was used as cost effective alternative of BMP-2. Previous studies have indicated the similarity of the release kinetics of BSA and BMP-2 [17-19]. The BSA and Ca2+ release kinetics from BioCaP granules with the internally-incorporated, coating-incorporated, or adsorbed delivery modes (n=6 per group) were investigated by soaking them in phosphate-buffered saline (PBS) at pH of 7.4. BioCaP granules with adsorbed BSA (as control) were prepared by immersing them in an aqueous protein solution (total volume of 20 ml) for 24 h at 37°C in plastic tubes. According to the manufacturer’s protocol, Alexa-BSA is more easily detected at lower concentration than FITC-BSA. Therefore, 20μg/ml of FITC-BSA and 1μg/ml of Alexa-BSA were used for loading. Each sample (0.05 g BioCaP granules per sample) was placed in a 2-ml sealed Eppendorf tube containing 2 ml PBS. The tubes were incubated for up to 35 days in a shaking water bath (50 agitations/min) at 37°C. The PBS was refreshed at each time point (hour 3, 6, day 1, 2, 3, 5, 7, 10, 13, 17, 22, 28, and 35) and triplicate 200-µl aliquots of the PBS were withdrawn for spectrophotometric analysis in a Fluorimeter (Spectramax M2, Molecular Devices, CA, USA). Fluorescence readings were converted into amounts of protein using a standard curve that was generated from a dilution series of BSA prepared in 2 ml PBS. Meanwhile, Ca2+ release was monitored by measuring Ca2+ concentration in the 200-µl aliquots of the PBS using atomic adsorption spectrometry (Analyst 100, PerkinElmer, USA). At the end of the release experiments, the residual BSA in BioCaP was determined by dissolving the materials in 20 ml of 0.5 M ethylene diaminetetraacetic acid (EDTA, pH 8.0) for spectrophotometric analysis. The percentage of BSA released from the BioCaP was calculated according to the formula: % BSA= (amount of the released BSA fraction / total loaded BSA)×100 %

Osteoclast-based resorption assay The resorbability of BioCaP was tested using a cell-based resorption assay [20]. BioCaP tablets were used to investigate the resorption. Briefly, human peripheral blood

33

Chapter 3 mononuclear cells (PBMCs), isolated from whole blood with Ficoll–Paque density gradient, were seeded on 800-μm-thick BioCaP tablets with or without coating (without protein) in 96-well plates at a density of 106 cells per well. The PBMCs were cultured with RANKL (40 ng/ml, PreProtech, Rocky Hill, NJ, USA) and M-CSF (25 ng/ml, R&D Systems, Minneapolis, MN, USA) as described preciously [21]. Cells were cultured in duplicate, and the culture media were refreshed twice a week. The formation of osteoclasts was assessed by analyzing tartrate-resistant acid phosphatase (TRACP) positive multinucleated cells and by SEM after 21 days of culturing on BioCaP tablets [21, 22]. The cells were washed with PBS and fixed in 4% PBS-buffered formaldehyde for 5 min and stained for TRACP activity using the leucocyte acid phosphatase kit (Sigma). The cells were fixed, dried, and sputter-coated for SEM investigation. After 24 days of culturing, cells were removed by using demineralised water and 10% NH3OH and the formation of lacunae was detected by SEM. All in vitro cell culture experiments were performed at least three times.

In-vivo investigation As an experimental animal model, we used adult male wistar rats (200–220g). The animal experiments were approved by the Ethics Committee of Zhejiang Chinese Medical University, China. Four groups were established (n=6 animals per group): (1) BioCaP with internally-incorporated BMP-2; (2) BioCaP with coating-incorporated BMP-2; (3) BioCaP with directly-adsorbed BMP-2 (as control); and (4) BioCaP without BMP-2 (BioCaP material only; as control). Each sample consisted of 0.22g BioCaP granules (0.25-1mm). The samples were prepared as described in a previous study [15]. Two samples per rat were randomly implanted in dorsal subcutaneous pockets (one on the left side and one on the right) [15]. Five weeks after implantation, the samples were retrieved, chemically fixed and embedded as previously reported [23]. Three-dimensional (3D) reconstructions of the samples were obtained using a high-resolution micro-CT system (μCT 40, Scanco Medical AG, Bassersdorf, Switzerland). To this end, samples were fixed in synthetic foam and placed vertically in a polyetherimide holder and scanned at a 18 μm isotropic voxel size, 70 kV source voltage, and 113 μA current. Grey values, depending on radiopacity of the scanned material, were converted into corresponding values of degree of mineralization by the analysis software (Scanco Medical AG). A distinction could be made between the newly formed bone and BioCaP, since the mineralization degree of BioCaP, was significantly higher than the mineralization degree of bone. A method for the separation of graft material and newly formed bone was applied according to the previous study by using an “onion-peeling” algorithm (Scanco Medical AG) and specific threshold settings [24]. By using this method, the micro-CT results were comparable with histomorphometrical results [24]. Briefly, a low threshold of 467.1 mg hydroxyapatite (HA)/cm3 to distinguish bone tissue from connective tissue and bone marrow, and the grey values were scaled from 1 to 1000 and the threshold was set at 158 to distinguish BioCaP from bone tissue. These two thresholds were calculated by averaging the thresholds determined in 3 slices of three samples by two independent observers. The samples were analysed for bone volume

34

Preparation and characteristics of BioCaP

(BV), bone density (BV/TV), material volume (MV), and bone mineral density (BMD). For the measuring of the volume of BioCaP before implantation (time 0), six chemically fixed and plastic-embedded samples (0.22g of BioCaP granules per sample) were specifically reserved and evaluated by micro-CT.

Statistical analysis All data are presented as a mean ± SD. The data were statistically evaluated by a one-way analysis of the variance (ANOVA) using SPSS statistical software (version 16.0 for Windows). Post-hoc comparisons were made using Bonferroni's corrections. The significance level was set at p < 0.05.

RESULTS

In vitro characterization of BioCaP granules The biomimetic precipitation of BioCaP consisted of two different phases: the initial quick precipitation forming super-fine crystalline structures (Fig. 2A), and the subsequent slow precipitation forming micro-crystalline spheres with 2-10 µm in diameter (Fig. 2B). The BioCaP granules were shown in Fig. 2C. The compressive strength value of BioCaP was 4.58±0.31 MPa. The energy dispersive X-ray analysis revealed that the Ca/P molar ratio of BioCaP was about 1.48. The contents of BioCaP include Ca (16.7%), P (24.8%), O (52.8%), Na (1.9%), and Cl (3.8%) (Fig. 3). BioCaP powder exhibited a unique diffraction peak at 2θ = 26°, 32°, and 46° in the XRD spectra (Fig. 4). The incorporation of BSA or BMP-2 elicited no profound change in the major diffraction spectrum. The surface morphology of the BioCaP and the coating was shown in Fig. 5A and Fig.5B, respectively. The loading of 1μg/ml BMP-2, 1μg/ml Alexa-BSA, or 20μg/ml FITC-BSA had no effect on the surface structure of BioCaP or the crystalline coating. ELISA results showed that BMP-2 with the internally-incorporated mode had a significantly higher loading efficiency (40.3 ± 0.9%) than with the coating-incorporated mode (35.1 ± 1.3%). By the internally-incorporated mode, protein (green signal) was distributed homogeneously throughout the BioCaP granules (Fig. 6Aa and Ab). By the coating-incorporated mode, protein (green signal)was distributed uniformly throughout the coating layer (Fig. 6Ac). The coating showed at higher magnification a crystalline structure (Fig. 6Ad). The mean thickness of the coating was 21.02±7.53 μm.

Figure 2. SEM micrographs of two phases of the precipitation of BioCaP (A and B) and BioCaP granules (C). The first phase was the initial precipitation (A). The second phase of precipitation (B; white pane) came from the incubation period containing small particles with tiny crystals (particle size of 2-10 µm).

35

Chapter 3

Figure 3. Energy Dispersive X-Ray Spectroscopy analysis of the chemical content of BioCaP.

Figure 4. XRD patterns showed BioCaP bearing no protein (a), 1µg/ml of BSA (b), 20µg/ml of BSA (c), and 1µg/ml of BMP-2 (d). The protein-carrying mode was the internally-incorporated mode.

36

Preparation and characteristics of BioCaP

Figure 5. SEM micrographs of the surface of BioCaP granules bearing an internally-incorporated (A) or a coating-incorporated (B) depot of FITC-BSA. After 35 days immersion in PBS (pH 7.4, 37°C), SEM micrographs showed the surface of BioCaP granules bearing an internally-incorporated (C) or a coating-incorporated (D) depot of FITC-BSA

The adsorbed FITC- and Alexa-BSA was released rapidly, being completely exhausted after 10 or 13 days (Fig. 6B). However, both FITC- and Alexa-BSA revealed a low burst release and subsequently a sustained release from the internally- or coating-incorporated depot in BioCaP granules. In the internally-incorporated depot, about 10-14% of BSA released from BioCaP granules in an initial burst release stage (within the first 24 hours). Subsequently BSA was gradually released at a steady rate until the 35th day (Fig. 6B). The initial burst release of coating-incorporated depot of BSA was about 18-20% and was also followed by a sustained release (Fig. 6B). Meanwhile, Ca2+ also revealed a low burst release and subsequently a sustained release from BioCaP with or without protein (Fig. 6C). The presence of BSA did not affect the Ca2+ release (Fig. 6C in which only Ca2+ release from BioCaP with FITC-BSA is shown). After 35 days, the total amount of released Ca2+ from BioCaP with coating was significantly lower than from granules without coating (p<0.05). In the end, the surface morphology of BioCaP without coating had not changed (Fig. 5C), while BioCaP with coating showed that most crystals of the coating had been dissolved (Fig. 5D).

37

Chapter 3

Figure 6. Fluorescence micrographs of cross-sections of BioCaP granules at low and high magnifications: the internally-incorporated depot of FITC-BSA (Aa and Ab, green signal) or the coating-incorporated depot of FITC-BSA (Ac and Ad, green signal). FITC-BSA has been incorporated into the interior of BioCaP granules (black arrow) or the coating (white arrow). Graphs depicting the cumulative protein release kinetics from BioCaP granules (FITC-BSA: 20μg/ml; Alexa-BSA: 1μg/ml) soaked in PBS (pH 7.4, 37°C) for 35 days (B), and the cumulative Ca2+ release kinetics of BioCaP granules with or without protein (C). Mean values (n=6 per group) are represented together with the standard deviation. Time points: hour 3, 6, day 1, 2, 3, 5, 7, 10, 13, 17, 22, 28 and 35.

Figure 7. TRACP staining of osteoclasts (white arrows) on BioCaP (A) or its coating (D); SEM

38

Preparation and characteristics of BioCaP micrographs of osteoclasts on BioCaP (B) or its coating (E), and the resorption by osteoclasts on BioCaP (C) or its coating (F). Tartrate-resistant acid phosphatase-positive osteoclasts generated from human peripheral blood mononuclear cells were grown on BioCaP tablets (Fig. 7A and D). Representative SEM micrographs of osteoclasts showed their attachment to the material (Fig. 7B and E). Resorption pits were clearly discernible after 24 days of culture (Fig. 7C and F).

In-vivo investigation At the time of their sacrifice (5-week juncture), all animals were in good health, and no complications had become manifest during the postoperative period. The histological analysis showed no inflammatory activity in all the samples. Three-dimensional (3D) reconstructions of samples with BioCaP and newly formed bone were shown in Fig. 8. New bone was only found in the sample containing BioCaP with BMP-2. Bone and BioCaP were separated by the analysis software (Fig. 9). Micro-CT analysis revealed that the volume and the volume density of newly formed bone for BioCaP with internally- or a coating-incorporated BMP-2 was significantly higher than around an adsorbed depot (Table 1), but no significant differences were found between the internally- and the coating- incorporated depots. 5 weeks after implantation, the volume of BioCaP of all the four groups decreased significantly, compared with the volume before implantation (298.48±8.87 mm3) (p<0.05). The volume of BioCaP with internally- or a coating-incorporated BMP-2 was significantly higher than those with adsorbed or no BMP-2 (p<0.05). No significant difference was found in the bone mineral density among the three BMP-2 groups.

39

Chapter 3

Figure 8. Three-dimensional (3D) reconstructions of BioCaP (left column) and bone (right column) by micro-CT. Group (1), BioCaP granules bearing an internally-incorporated depot of BMP-2 (A, B); Group (2), BioCaP granules bearing a coating-incorporated depot of BMP-2 (C, D); Group (3), BioCaP granules bearing an adsorbed depot of BMP-2 (E, F); and Group (4), BioCaP granules bearing no BMP-2 (G, H). In the Group (4), no bone was found (H).

40

Preparation and characteristics of BioCaP

Figure 9. Two-dimensional (2D) image of BioCaP and bone by Micro-CT (A). Bone (red) and BioCaP (white) were separated by the analysis software (B).

Table 1 Micro-CT Evaluation on BioCaP and bone 5 weeks after implantation

BMD Groups BV (mm3) BV/TV MV (mm3) (mg HA/cm3) BioCaP + internally-incorporated (1) 11.78±2.67* 0.05±0.01* 229.69±16.13*† 649.10±26.69 BMP-2

BioCaP + coating-incorporated (2) 13.76±4.92# 0.06±0.03# 236.43±29.82#§ 605.84±28.77 BMP-2

(3) BioCaP + adsorbed BMP-2 1.22±1.87 0.01±0.01 124.97±66.93 648.10±41.43

(4) BioCaP - - 118.18±35.29 -

BV, bone volume; BV/TV, bone volume/total volume; MV, material volume; BMD, bone mineral density *Significant difference (p < 0.05) between Group (1) and Group (3). #Significant difference (p < 0.05) between Group (2) and Group (3). †Significant difference (p < 0.05) between Group (1) and Group (4). §Significant difference (p < 0.05) between Group (2) and Group (4).

DISCUSSION

In this study, the mechanical strength of BioCaP granules is comparable with that of trabecular bone. These granules have a series of characteristics that make them attractive for bone regeneration purposes. First, BioCaP granules as a bone-defect filling material are easy to handle. Second, both delivery modes showed a slow release of protein in vitro. Third, the BioCaP granules can be resorbed as shown by the human osteoclasts which

41

Chapter 3 confirmed the previous results by using mouse osteoclasts [15]. Forth, micro-CT analysis showed 3D construction and 2D image of newly formed bone around the granules containing BMP-2 after the subcutaneous implantation. This micro-CT analysis with an “onion-peeling” algorithm successfully distinguished BioCaP and bone. Micro-CT results showed that BioCaP granules were biodegradable and bone formation proved to be higher with the granules with the two delivery modes of BMP-2 compared with the adsorbed mode, which confirmed the histological results in the previous study [15]. Our findings demonstrate that the BioCaP granules in which BMP-2 had been incorporated are osteoinductive and we propose the use of this implant material if efficient bone formation is needed. Compressive strength is most often used to characterize the mechanical behaviour of bone substitutes [25]. The compressive strength of BioCaP is about 4.58 MPa, thus being similar to that of human trabecular bone which ranges from 0.22 to 10.44 MPa, with a mean value of 3.9 MPa [26]. The XRD and EDX analysis showed that BioCaP had the characteristics of calcium-deficient apatite with a low crystallinity [27], also indicating similarities to native bone [28, 29]. The bone-like characteristics of BioCaP could be caused by the mixture of the two phases of precipitations under a biomimetic environment. In addition, the biomimetic environment used in this study can retain BMP-2 activity [30]. The hardness of BioCaP granules makes it easy of handling for filling bone defects. A slow and sustained delivering mode is critical to maximize the functional efficiency of a cytokine [12, 31]. The protein release from the internally- and the coating- incorporated modes is highly depending on the degradation rate of BioCaP or the coating. In the present study, these two modes exhibited a low burst release within the first 24 hours and subsequently a slow release period in vitro. On the other hand, the resorption assay revealed that human osteoclasts can resorb BioCaP and the coating in vitro. Resorbing cells such as osteoclasts may increase the degradation rate of material to speed up the agent's rate of delivery [32], and consequently to affect the functional efficiency of the agent. The cell-mediated protein release from BioCaP has been investigated in another study [15]. In the present study, micro-CT analysis revealed that the two delivery modes induced more bone formation compared to the adsorption mode. The findings suggest that the two delivery modes of BMP-2 have the capacity to maintain the slow release in vivo, whereas the adsorbed mode resulted in a burst release. In the internally-incorporated mode, protein distributes throughout the whole volume of BioCaP granules, which means that the protein can be gradually liberated as long as the granules are degraded. After 5-week implantation BioCaP granules with the internally-incorporated BMP-2 had not been completely degraded. This probably implies that a certain amount of BMP-2 is still present in the BioCaP granules, thus increasing the duration of BMP-2 activity. In contrast, about 80-100% of the coating (about 20-µm thickness) can be degraded within 5 weeks in vivo [12, 33]. However, micro-CT could not distinguish BioCaP and its coating. Histological observation indicated the coating of BioCaP was degraded [15]. Moreover, the in vitro Ca2+ release represents the self-dissolubility of BioCaP and its coating, and indicates that the solubility of BioCaP is higher than the coating. The degradation of BioCaP with coating-incorporated BMP-2 would be delayed because of the coating. The micro-CT analysis revealed that the degradation of BioCaP with

42

Preparation and characteristics of BioCaP internally- or coating-incorporated BMP-2 was slower than that of BioCaP with adsorbed BMP-2 or without BMP-2. These results are is consistent with the previous study [15]. As a dual protein-delivery model, BioCaP can be used for different clinical applications. Growth factors, anticancer drugs and antibiotics can be candidates for this model. For example, BioCaP granules with internally-incorporated BMP-2 and in addition antibiotics incorporated into the coating may be considered to treat bone defects in peri-implantitis. However, the potential applications of this dual protein-delivery model need to be evaluated further.

CONCLUSION

The biomimetic mineralization approach confers BioCaP bone-like mechanical strength. BioCaP showed the characteristics of calcium-deficient apatite. The two delivery modes of BioCaP showed slow release of protein in vitro in 35 days. However, the in vivo protein release may more complicated, which needs to be investigated further. Human osteoclasts assay indicated the good biodegradability of BioCaP which is also confirmed in vivo by micro-CT analysis. Micro-CT results showed more bone formation by the two protein-delivery modes, suggesting these two modes can achieve a highly efficient delivery of BMP-2. This results confirmed the histological result in the previous study [15]. This micro-CT method provides an alternative for the evaluation of bone substitute and bone formation.

ACKNOWLEDGEMENTS

The authors acknowledge Cees Kleverlaan, Arie Werner, Ben Norder, Teun J. de Vries, and Ineke Jansen for their assistance with the operation of SEM, XRD, and cell culture.

43

Chapter 3

REFERENCES 1. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 2. Liu Y, Wu G, de Groot K. Biomimetic coatings for bone tissue engineering of critical-sized defects. J R Soc Interface 2010;7 Suppl 5:S631-47. 3. Ahlmann E, Patzakis M, Roidis N, Shepherd L, Holtom P. Comparison of anterior and posterior iliac crest bone grafts in terms of harvest-site morbidity and functional outcomes. J Bone Joint Surg Am 2002;84-A:716-20. 4. Bohner M, Galea LG, Lemaitre J, Kohler T, Muller R. Bone substitute: Transforming beta-tricalcium phosphate porous scaffolds into monetite. Biomaterials 2008;29:3400-07. 5. Dorozhkin SV. Bioceramics of calcium orthophosphates. Biomaterials 2010;31:1465-85. 6. Seeherman H, Wozney JM. Delivery of bone morphogenetic proteins for orthopedic tissue regeneration. Cytokine Growth Factor Rev 2005;16:329-45. 7. Bose S, Tarafder S. Calcium phosphate ceramic systems in growth factor and drug delivery for bone tissue engineering: a review. Acta Biomater 2012;8:1401-21. 8. Koempel JA, Patt BS, O'Grady K, Wozney J, Toriumi DM. The effect of recombinant human bone morphogenetic protein-2 on the integration of porous hydroxyapatite implants with bone. J Biomed Mater Res 1998;41:359-63. 9. Kuboki Y, Takita H, Kobayashi D, Tsuruga E, Inoue M, Murata M, et al. BMP-induced osteogenesis on the surface of hydroxyapatite with geometrically feasible and nonfeasible structures: topology of osteogenesis. J Biomed Mater Res 1998;39:190-9. 10. Lasserre A, Bajpai PK. Ceramic drug-delivery devices. Crit Rev Ther Drug Carrier Syst 1998;15:1-56. 11. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 12. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106. 13. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 14. Kempen DH, Lu L, Heijink A, Hefferan TE, Creemers LB, Maran A, et al. Effect of local sequential VEGF and BMP-2 delivery on ectopic and orthotopic bone regeneration. Biomaterials 2009;30:2816-25. 15. Liu T, Wu G, Zheng Y, Wismeijer D, Everts V, Liu Y. Cell-mediated BMP-2 release from a novel dual drug delivery system promotes bone formation. Clin Oral Implants Res 2013;submitted. 16. Liu Y, Layrolle P, de Bruijn J, van Blitterswijk C, de Groot K. Biomimetic coprecipitation of calcium phosphate and bovine serum albumin on titanium alloy. J Biomed Mater Res 2001;57:327-35. 17. Basmanav FB, Kose GT, Hasirci V. Sequential growth factor delivery from complexed microspheres for bone tissue engineering. Biomaterials 2008;29:4195-204. 18. Yilgor P, Tuzlakoglu K, Reis RL, Hasirci N, Hasirci V. Incorporation of a sequential BMP-2/BMP-7 delivery system into chitosan-based scaffolds for bone tissue engineering. Biomaterials 2009;30:3551-9. 19. Yilgor P, Sousa RA, Reis RL, Hasirci N, Hasirci V. Effect of scaffold architecture and BMP-2/BMP-7 delivery on in vitro bone regeneration. J Mater Sci Mater Med 2010;21:2999-3008. 20. Zhang Z, Egana JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al. Cell-based resorption assays for bone graft substitutes. Acta Biomater 2012;8:13-9. 21. Olivier BJ, Schoenmaker T, Mebius RE, Everts V, Mulder CJ, van Nieuwkerk KM, et al.

44

Preparation and characteristics of BioCaP

Increased osteoclast formation and activity by peripheral blood mononuclear cells in chronic liver disease patients with osteopenia. Hepatology 2008;47:259-67. 22. Faust J, Lacey DL, Hunt P, Burgess TL, Scully S, Van G, et al. Osteoclast markers accumulate on cells developing from human peripheral blood mononuclear precursors. J Cell Biochem 1999;72:67-80. 23. Liu Y, Huse RO, de Groot K, Buser D, Hunziker EB. Delivery mode and efficacy of BMP-2 in association with implants. J Dent Res 2007;86:84-9. 24. Schulten EA, Prins HJ, Overman JR, Helder MN, ten Bruggenkate CM, Klein-Nulend J. A novel approach revealing the effect of a collagenous membrane on osteoconduction in maxillary sinus floor elevation with beta-tricalcium phosphate. Eur Cell Mater 2013;25:215-28. 25. Hannink G, Arts JJ. Bioresorbability, porosity and mechanical strength of bone substitutes: what is optimal for bone regeneration? Injury 2011;42 Suppl 2:S22-5. 26. Misch CE, Qu Z, Bidez MW. Mechanical properties of trabecular bone in the human mandible: implications for dental implant treatment planning and surgical placement. J Oral Maxillofac Surg 1999;57:700-6. 27. Dorozhkin SV. Amorphous calcium (ortho)phosphates. Acta Biomater 2010;6:4457-75. 28. Boskey AL. Amorphous calcium phosphate: The contention of bone. J Dent Res 1997;76:1433-36. 29. Habibovic P, Barrere F, van Blitterswijk CA, de Groot K, Layrolle P. Biomimetic hydroxyapatite coating on metal implants. J Am Ceram Soc 2002;85:517-22. 30. Liu Y, Hunziker EB, Layrolle P, De Bruijn JD, De Groot K. Bone morphogenetic protein 2 incorporated into biomimetic coatings retains its biological activity. Tissue Eng Part A 2004;10:101-8. 31. Su Y, Su Q, Liu W, Lim M, Venugopal JR, Mo X, et al. Controlled release of bone morphogenetic protein 2 and dexamethasone loaded in core-shell PLLACL-collagen fibers for use in bone tissue engineering. Acta Biomater 2012;8:763-71. 32. Wernike E, Hofstetter W, Liu Y, Wu G, Sebald HJ, Wismeijer D, et al. Long-term cell-mediated protein release from calcium phosphate ceramics. J Biomed Mater Res A 2010;92:463-74. 33. Liu Y, de Groot K, Hunziker EB. BMP-2 liberated from biomimetic implant coatings induces and sustains direct ossification in an ectopic rat model. Bone 2005;36:745-57.

45

Chapter 3

46

Chapter 4

Osteoinductive biomimetic bone substitute for the repair of critical-sized bone defects in sheep

Tie Liu, Gang Wu, Daniel Wismeijer, and Yuelian Liu.

Submitted, 2013.

47

Chapter 4

ABSTRACT

As a synthetic alternative to autologous bone grafting, a biomimetic calcium phosphate bone substitute (BioCaP) was developed with two protein delivery modes: 1) protein in the interior of BioCaP (internally-incorporated mode), 2) protein in the coating on the surface of BioCaP (coating-incorporated mode). The aim of this study is to investigate the therapeutic effectiveness of BioCaP with each delivery mode of BMP-2 in the repair of a large cylindrical bone defect (8mm in diameter and 13mm in depth) in sheep. Six groups were estabilshed: (i) BioCaP only; (ii) BioCaP with coating-incorporated BMP-2; (ii) BioCaP with internally-incorporated BMP-2; (iv) no graft material; (v) autologous bone; (vi) deproteinized bovine bone (DBB, a commercial product). 4 and 8 weeks after implantation, samples were withdrawn for histological and histomorphometric analysis. BioCaP with BMP-2 showed equal efficacy as autologous bone in the bone defect repair at 8 weeks post-implantation. Both delivery modes of BMP-2 accelerated the bone formation in an early period of 4 weeks. The internally-incorporated mode enhanced bone formation after 8 weeks, showing more efficient than DBB. Within 8 weeks, about half of BioCaP with either internally-incorporated BMP-2 or without BMP-2 was degraded, which was significantly higher than that of BioCaP with coating-incorporated BMP-2. In conclusion, both two delivery modes of BMP-2 enhance bone formation. Benefiting from these two delivery modes, BioCaP can be a promising alternative to the autografts.

Keywords: Biomimetic calcium phosphate; Protein delivery; Bone repair; Critical-sized bone defect; BMP-2

INTRODUCTION

The treatment of bone defects requires adequate volume of bone tissue, which is of paramount importance to achieve an excellent aesthetic and functional restoration. When the bone defects are too large to be self-healed, it requires bone grafting in order to fill the defect [1, 2]. Autografts (gold standard), allografts, xenografts, and synthetic materials are available for the repair of bone defects in the fields of dentistry, orthopedics and traumatology [3]. Synthetic calcium phosphate (CaP) biomaterials are widely used in the regeneration of bone defects because of their chemical similarity to native bone tissue. Currently, there is an increased interest in biomimetic calcium-phosphate materials because of their capacity to carry (delivery) bioactive agents without compromising their bioactivity [4-8]. For example, the delivery of bone morphogenetic protein 2 (BMP-2) can enhance bone regeneration [9, 10]. Biomimetic materials are capable of eliciting specific cellular responses and directing new tissue formation [11]. Biomimetic CaP coating has been developed to deliver growth factors for bone regeneration [4]. However, the increase of the thickness of this coating is highly dependent on underlying substrates as a scaffold such as dental titanium implants, polymers, and deproteinized bovine bone [4]. To overcome this limit, we recently developed a biomimetic calcium-phosphate bone substitute (BioCaP) using a refined biomimetic coating approach [12]. The BioCaP granule can be used as a dual protein-delivery model which possesses two delivery modes: 1) an

48

Osteoinductive BioCaP for the repair of bone defects in sheep

internally-incorporated mode, protein can be incorporated into the interior of BioCaP; 2) a coating-incorporated mode, protein can be incorporated into the coating on the surface of BioCaP [13]. Our previous study has demonstrated that BioCaP with the two delivery modes resulted in a slow release of protein in vitro, and an efficient osteoinduction in rats when BMP-2 was delivered. The aim of this study was to evaluate the therapeutic effects of BioCaP bone substitute with two delivery modes of BMP-2 in the repair of the critical-sized bone defect in sheep. New bone formation and the degradation of BioCaP were evaluated histologically and histomorphometrically after a 4- and 8-week implantation.

MATERIALS AND METHODS

Fabrication of biomimetic calcium phosphate (BioCaP) bone substitute BioCaP was fabricated by refining a well-established biomimetic mineralization approach [14]. A CaP solution (200 mM HCl, 20 mM CaCl2·2H2O, 680 mM NaCl, and 10 mM Na2HPO4) buffered by TRIS (250 mM) to a pH of 7.4. The whole solution was incubated in a shaking water bath (50 agitations/min) at 37°C for 24 hours. Thereafter all precipitations were retrieved and gently washed by Milli-Q water, strongly filtered and compressed to a block using a vacuum exhaust filtering method with a vacuum filter (0.22-μm pore, Corning, NY, USA) and an air pump. After drying in air circulation at room temperature for 2 hours, the hardened block can be ground and filtered to obtain granules with a size of 0.25-1.0mm using metallic mesh filters. For sterilization, the CaP solution was filtered with the vacuum filter (0.22-μm pore) before buffering. All the following procedures were performed under aseptic conditions. BMP-2 (INFUSE® Bone Graft, Medtronic, USA) can be introduced into the CaP solution at a final concentration of 0.2μg/ml before buffering as mentioned above and thereafter was co-precipitated into the interior of BioCaP, viz., the internally-incorporated mode.

Biomimetic coating procedure The superficial coating was deposited on BioCaP according to the well-established biomimetic mineralization approach [14]. Briefly, 0.58g of BioCaP with a size of 0.25-1mm was incubated in the coating solution [40 mM HCl, 4 mM CaCl2·2H2O, 136 mM NaCl, 2 mM Na2HPO4, and 50 mM TRIS (pH 7.4); total volume of 150 ml] in a shaking water bath (50 agitations/min) at 37°C for 24 hours. BMP-2 present in the coating solution at a final concentration of 0.2μg/ml can be co-precipitated into the coating on BioCaP, viz., the coating-incorporated mode. The granules were then freeze-dried. The entire procedure was conducted under sterile conditions.

Quantification of the amount of the incorporated BMP-2 The amount of incorporated BMP-2 was determined by a commercially available enzyme-linked immunosorbent assay (ELISA) kit (PeproTech, London, UK). 0.05g of BioCaP with the two carry (delivery) modes of BMP-2 (n=6) was dissolved in 1ml 0.5M EDTA (pH 8.0). The ELISA assay was performed according to the manufacturer's

49

Chapter 4 instructions. The BioCaP without BMP-2 was evaluated as a control. According to the ELISA result, each sample of BioCaP granule that bore internally-incorporated or coating-incorporated BMP-2 contained 10.5±1.27μg and 18.56±0.88μg of BMP-2, respectively.

The morphology of BioCaP The morphology of BioCaP granules with or without BMP-2 was visualised using a scanning electron microscope (SEM, XL20, FEI Company, the Netherlands), under an accelerating voltage of 10 kV after being sputter-coated with gold.

Surgical procedure Twelve sheep were anesthetized by administering Sumianxin II (0.3 ml/kg, purchased from the Military Veterinary Institute, Quartermaster University of PLA, Chang Chun, China) with the addition of Penicillium (5 × 104 U/kg) and atropine (0.03 mg/kg) at 30 min before surgery. After applying local anaesthesia (1% lidocaine with 1:100,000 adrenaline) and skin disinfection (0.5% iodophor solution) to the implantation sites, the surgery and animal care was performed and the cylinder-shaped defects were created (8 mm in diameter and 13 mm in depth) as described in a previous study [15]. The implantation sites were the proximal part of the diaphysis and distal epiphysis of humerus and femur of 12 adult female sheep. Each sheep provided 8 standardized implantation sites. Six implantation sites were randomly chosen. These implantation sites were assigned to the experiment groups according to a randomization protocol [16]. Membranes (Bio-Gide®, Geistlich Biomaterials, Wolhuser, Switzerland) were used to cover the defects. Samples with surrounding tissues were retrieved at 4 weeks and 8 weeks post-operation.

Experimental groups Six groups were established to treat CSBD (n=6 animals per group per time point, Table 1): (i) BioCaP bearing neither a coating nor BMP-2 (experimental); (ii) BioCaP bearing a coating-incorporated BMP-2(experimental); (iii) BioCaP bearing an internally-incorporated BMP-2(experimental); (iv) No graft material (negative control); (v) Autologous bone (positive control); and (vi) Deproteinized bovine bone (DBB, Bio-Oss®, control, a bovine xenograft, is one of the most widely used commercial bone substitutes used in bone repair and augmentation in clinical dentistry).

In the case of autograft, the bone was harvested at the same time as the creation of the defect and reduced to 0.25-1mm particles using a rongeur.

Histological procedures Samples with surrounding tissues chemically were fixed and embedded into a block as previously reported [17, 18]. Applying a systematic random-sampling strategy [19], the samples were sawed vertical to the long axis of the cylindrical defect, into 10-12 slices of 600-μm thickness, 1 mm apart (interval). All slices of each sample were separately

50

Osteoinductive BioCaP for the repair of bone defects in sheep mounted on plexiglass holders and polished. The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue [18] and examined with a light microscope with a digital camera (Leica, Wetzlar, Germany).

Table 1 Experimental groups

Graft Total BMP-2 materials Loading

Granule Volume Groups Abbreviation Absence size (amount) Dose of of graft (−) BMP-2 material Presence (per per (+) sample) sample

BioCaP bearing neither 0.65cm3 (i) BioCaP 0.25-1.0mm - - a coating nor BMP-2 (0.58g)

BioCaP bearing BioCaP 0.65cm3 (ii) coating-incorporated 0.25-1.0mm + 10.5μg BMP inc. (0.58g) BMP-2

BioCaP bearing BioCaP 0.65cm3 (iii) internally-incorporated 0.25-1.0mm + 18.56μg BMP int. (0.58g) BMP-2

No graft material (iv) NGM - - - - (negative control)

Autologous bone (v) AB 0.25-1.0mm 0.65cm3 - - (positive control)

Deproteinized bovine 0.65cm3 (vi) DBB 0.25-1.0mm - - bone (Bio-Oss®) (0.35g)

Histomorphometric analysis In addition to a subjective histological description, 10 slices of each sample was used for quantitative histomorphometric analysis including the volume of newly formed bone, bone marrow (fat), BioCaP and the volume density of multinucleated giant cells (MGCs) on BioCaP or DBB. Using the point-counting technique [20], the surface area (S) of a component per slice was obtained. The interval between two slices was 1mm. Therefore, the volume (V) of a component is defined as: V ∑ ( ) The volume density of MGCs was normalized to the volume of BioCaP or DBB. The volume density of MGCs (Va) is defined as its volume (Vb) per unit volume of graft materials (Vc): Va= Vb/Vc. To evaluate the degradation of BioCaP granules, the volume of BioCaP before implantation (time 0, as control) were evaluated by the same

51

Chapter 4 histological method. Six chemically fixed and plastic-embedded samples (0.58g of BioCaP granules per samples) were specifically reserved for this purpose.

Statistical analysis All data are presented as mean values together with the standard deviation (SD). Data were compared using a one way analysis of variance (ANOVA), and post-hoc comparisons were made using Tukey's corrections. The significance level was set at p < .05.

RESULTS

In-vitro characterization A scanning electron microscopy (SEM) micrograph of BioCaP granules is depicted in Fig. 1A. BioCaP granule that bore internally-incorporated BMP-2 showed rough surface (Fig. 1B and C). BioCaP granule with a layer of coating was shown in Fig. 1D. BioCaP granule that bore coating-incorporated BMP-2 showed crystalline surface (Fig. 1E and F). The incorporation of BMP-2 did not affect the surface morphology of BioCaP or the coating [13].

Clinical observations After 4- and 8-week implantation, a total of 72 implants were harvested (36 implants at 4 weeks and 36 implants at 8 weeks). All the sheep exhibited good health and all the surgical implant sites healed well without any significant wound complication. No visual signs of inflammatory or adverse tissue reaction were observed.

Figure 1. SEM micrographs of the BioCaP granules (A). BioCaP with internally-incorporated BMP-2 showed a rough surface (B) with calcium phosphate microspheres (C). A cross section of BioCaP granules with coating-incorporated BMP-2 showed the coating layer on the surface of BioCaP (D). This coating displayed a crystalline rough surface (E) with tiny crystalline plates (F).

52

Osteoinductive BioCaP for the repair of bone defects in sheep

Descriptive light microscopy Histological images of each group at 4 and 8 weeks of implantation are depicted in Fig. 2. At 4 week, newly formed bone was observed in close contact with the BioCaP granules in the three BioCaP groups (see Figs 2A-C). BioCP with BMP-2 showed more bone formation than BioCaP without BMP-2. The new bone started to form a network in some areas. The empty group (negative control) confirmed that the defect was critically sized (Fig. 2D and Fig. 2d). Bone defect was not healed after 8 weeks. New bone only appeared sporadically near the defect boundary and fibrous tissues were constantly appeared in the center part of the defect. The DBB group (Fig. 2F) showed significantly less bone formation compared with the BioCaP with incorporated BMP-2 or autologous groups (Fig. 2E). Images at higher magnification of the three BioCaP groups showed a thin layer of bone which capsulated BioCaP granules (Fig. 3A and C), and new bone was also observed between the granules (Fig. 3E). At 8 weeks, all the graft groups showed significant bone formation (see Figs 2a-f). BioCaP with internally-incorporated BMP-2 (Fig. 2c) displayed significantly more bone formation compared with BioCaP without BMP-2 (Fig.2a). In areas where BioCaP with or without BMP-2 (Figs 2a-c) was still present, a complete interconnected bone network could be observed. The deposited bone was constantly in close contact with these BioCaP granules, most of which were entirely encapsulated in the new bone. The newly formed bone had a woven appearance including the presence of remodeling lacunae. DBB also showed a similar trabecular bone structure (Fig. 2f). The autologous bone group showed uneven woven bone formation (Fig. 2e). Images at higher magnification of the three BioCaP groups showed that the areas mainly consisting of bone, bone marrow and BioCaP with or without BMP-2 can be observed indicating a healthy bone environment (Fig. 3B, D, and F). The group containing BioCaP with internally-incorporated BMP-2 showed that BioCaP had been replaced by newly formed bone in some areas with a trabecular bone appearance (Fig. 4A). It was displayed that the woven bone was undergoing remodeling to be the lamellar bone due to the osteoblasts and multinucleated osteoclasts (Fig. 4B). At 4 weeks, a few BioCaP granules were surrounded by multinucleated giant cells, which were apparently osteoclastic cells (Fig. 5A). No bone formation occurred in this area. Multinucleated giant cells were observed sporadically on BioCaP at 8 weeks (Fig. 5B). In both implantation time points, it was difficult to recognize or find the coating on BioCaP, which indicated the coating to be degraded completely.

53

Chapter 4

Figure 2. Representative histological micrographs at higher magnification of bone defect of each group at 4 and 8 weeks after implantation. Groups: (i) BioCaP (asterisk) bearing neither a coating nor BMP-2 (A; a); (ii) BioCaP bearing a coating-incorporated BMP-2 (B; b); (iii) BioCaP bearing an internally-incorporated BMP-2 (C; c); (iv) No graft material (D; d); (v) Autologous bone (E; e); and (vi) Deproteinized bovine bone (#) (F; f). At 4 weeks, the newly formed bone (unmineralized) was dark purple (arrow). At 8 weeks, the newly formed bone (mineralized) was reddish (arrow). The newly formed bone and the autologous (+) bone can be separated in the group (v). The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Scale bar = 1 mm.

54

Osteoinductive BioCaP for the repair of bone defects in sheep

Figure 3. Representative histological micrographs at higher magnification of bone defect of the three BioCaP groups at 4 and 8 weeks after implantation. At 4 weeks, BioCaP (A), BioCaP with coating-incorporated BMP-2 (C), BioCaP with internally-incorporated BMP-2 (E). New bone (white arrow) was observed in close contact with BioCaP (asterisk) or encapsulating BioCaP. At 8 weeks, BioCaP (B), BioCaP with coating-incorporated BMP-2 (D), BioCaP with internally-incorporated BMP-2 (F). Most areas of the bone defect were filled with bone, bone marrow (M) and BioCaP. Most residual BioCaP granules were entirely encapsulated in the new bone. Bone marrow was in close contact with BioCaP. The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Scale bar = 200 µm.

55

Chapter 4

Figure 4. Representative histological micrographs of trabecular bone structure with bone remodeling in the group containing BioCaP (asterisk) with internally-incorporated BMP-2 (A). At higher magnification (B), osteoblasts (white arrow) and osteoclasts (black arrow) were observed. The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Scale bar = 200µm in A. Scale bar = 50µm in B.

Histomorphometric results Quantitative evaluation of the amount of bone formation after 4 and 8 weeks of implantation (Fig. 6) revealed that bone formation significantly increased with increasing implantation time (p<0.05). At 4 weeks post-implantation, the bone formation in the samples containing autologous bone was the highest. Significantly more bone formation was observed in the groups containing BioCaP with internally- or coating-incorporated BMP-2 than the group containing BioCaP without BMP-2 or containing DBB. No significant difference was found between the two groups containing BioCaP with BMP-2. At 8 weeks implantation time, no significant difference in bone formation was found between autologous bone and BioCaP with internally- or coating-incorporated BMP-2. Significantly more bone formation was observed in the group containing BioCaP with internally-incorporated BMP-2 compared with the group containing DBB. Significantly more bone formation was observed in the group containing BioCaP with internally-incorporated BMP-2 than the group containing BioCaP without BMP-2. Quantitative evaluation of the amount of bone marrow (Fig. 7) revealed that at 4 weeks bone marrow only appeared in the groups containing BioCaP with or without BMP-2 and autologous bone. Significantly more bone marrow was observed in the group containing BioCaP with coating-incorporated BMP-2 compared with the group containing BioCaP with internally-incorporated BMP-2 or without BMP-2. At 8 weeks, no significant difference was found between the groups containing BioCaP with or without BMP-2 and autologous bone. Significant more bone marrow formation was observed in the groups containing BioCaP with or without BMP-2 compared with the group containing DBB. BioCaP degradation significantly increased with increasing implantation time (p<0.05) (Fig. 8). BioCaP with internally-incorporated BMP-2 or without BMP-2 showed significantly more BioCaP degradation after 4 and 8 weeks of implantation than the BioCaP with coating-incorporated BMP-2.

56

Osteoinductive BioCaP for the repair of bone defects in sheep

The volume density of multinucleated giant cells (MGCs) at 4 weeks was lowest in association with samples containing BioCaP with internally or coating-incorporated BMP-2 (Fig. 9). At 8 weeks, no significant difference in the volume density of MGCs was found among the four groups contain BioCaP or DBB.

Figure 5. Representative histological micrographs of multinucleated giant cell (black arrow) on the surface of BioCaP at 4 weeks (A) and 8 weeks (B) from the group containing BioCaP with internally-incorporated BMP-2. At 4 weeks, multinucleated giant cells were easily observed on a few BioCaP granules in all the three BioCaP groups, whereas at 8 weeks, multinucleated giant cells were occasionally found, since most BioCaP granules had been covered by new bone. Osteocytes (white arrow). The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Scale bar = 50 µm.

57

Chapter 4

Figure 6. Graph depicting the volume of newly formed bone within the bone defect at 4 and 8 weeks after implantation for each group (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05.

Figure 7. Graph depicting the volume of bone marrow within the bone defect at 8 weeks after implantation for each group (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05.

58

Osteoinductive BioCaP for the repair of bone defects in sheep

Figure 8. Graph depicting the percentage degradation of BioCaP within the bone defect at 4 and 8 weeks after implantation for each of the BioCaP groups. See Table 1 for an explanation of the abbreviations. Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05.

Figure 9. Graph depicting the volume density of multinucleated giant cells on the surface of BioCaP or DBB at 4 and 8 weeks after implantation. See Table 1 for an explanation of the abbreviations. Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05; +p<0.01.

DISCUSSION

Recently, a biomimetic calcium phosphate bone substitute (BioCaP) was developed as a dual delivery model with two delivery modes, viz., an internally-incorporated mode and a coating-incorporated mode. In current study, BioCaP with the two delivery modes of BMP-2 proved to be equally efficient as autologous bone in the repair of critical-sized bone defects at 8 weeks after implantation. The findings indicate that BioCaP with each delivery mode of BMP-2 can be a promising alternative to the autografts.

59

Chapter 4

Critical-sized bone defect is defined as the intraosseous wound with the smallest size, which cannot spontaneously heal completely without intervention [21]. It is usually used as an experimental model to test bone repair materials [22]. The critical-sized bone defect (CSBD) model in this study was created by drilling holes in the humerus and femur of sheep according to a widely published protocol by Nuss et al. [15]. This drill hole model in sheep has proved to be an excellent animal model for testing biomaterials for use in orthopedics, maxillofacial and dental surgery [23]. It allowed the intraosseous implantation of up to 8 different test materials within one animal due to the standardization of the bone defect, while at the same time it can reduce the overall suffering of animals and give the necessary numbers to satisfy statistical requirements [15, 24]. Because of the similarities with humans in weight, bone and joint structure and bone regeneration, the results from sheep are more valid than those obtained from small laboratory animals [25]. Although rodents may be less expensive, they have a different bone morphology and they are often are too small for testing bone substitute. Positive results in rodents may have to be repeated and verified in larger species before human clinical trials can be initiated. In the current study, we showed that the coating-incorporated mode of BMP-2 significantly induced more bone formation for BioCaP at 4 weeks but not 8 weeks. The accelerated bone formation is attributed to the gradual degradation of the coating with the slow release of BMP-2 [26]. However, the coating could be totally degraded within 5 weeks [17]. The coating on BioCaP cannot be recognized or distinguished from BioCaP using the histological staining. The reason could be the similar chemical property of the coating and BioCaP, since both of which are obtained by precipitating calcium phosphate. The new bone observed wrapping the BioCaP granules may cover the coating and thereby delay its degradation. The internally-incorporated mode of BMP-2 significantly accelerated more bone formation for BioCaP at both implantation time points (4 and 8 weeks), which indicated that the release of BMP-2 may be sustain during the 8 weeks. The internally-incorporatedBMP-2 could be continually released as long as BioCaP has not been totally degraded. The coating has been demonstrated as a three-dimensional reservoir from which BMP-2 can be gradually liberated [4], while BioCaP using the granular form can offer a larger reservoir than the coating. Moreover, BioCaP resulted in more bone marrow compared with DBB. The abundance of bone marrow bodes well for the health and the endurance of the newly formed bone, since it is an important source of nutriments and pluripotent progenitor cells for osseous tissue [27]. Both osteoblasts and osteoclasts are derived from progenitors that reside in the bone marrow. Therefore, bone marrow is critical for bone remodeling. The further degradation of BioCaP may be dependent on the bone remodeling. Histological and histomorphological analysis revealed the good biodegradability of BioCaP. Ideal biodegradability refers to that bone substitute can be degraded in short time, enabling bone remodeling, and concomitantly replaced by bone tissue [28]. The degradation of calcium phosphate materials depended on the self-dissolubility and cell-based resorption [29]. The use of coating can delay the degradation of BioCaP. When the induced bone covered or even encapsulated BioCaP, the degradation of BioCaP could also be delayed. In this case, the encapsulated residual BioCaP can be degraded in the following bone remodeling process. The observed degradation of BioCaP

60

Osteoinductive BioCaP for the repair of bone defects in sheep was associated with the multinucleated osteoclast-like cells. There cells were not only observed on bone but also on BioCaP. However, the two modes of BMP-2 resulted in a lower volume density of these osteoclast-like cells on material surface. This might also prevent the degradation of BioCaP. Our previous study has proven that the slow release of BMP-2 from the biomimetic coating suppress the formation of multinucleated giant cells [30]. In other studies, BMP-2 has been shown to stimulates the recruitment, proliferation and differentiation of osteoprogenitor cells at low doses, whereas it induces the recruitment, formation and activation of osteoclasts at high doses [31, 32]. In this study, BioCaP bone substitute was produced by the precipitation of calcium phosphate in a biomimetic environment which can retain the protein biological activity [7]. Both of BioCaP and the biomimetic calcium phosphate coating are produced in this environment using similar calcium phosphate solutions. The biomimetic coating has been demonstrated to be highly biocompatible and osteoconductive [4]. Good biocompatibility refers to the ability of a biomaterial to perform its desired function with respect to a medical therapy, without eliciting any undesirable local or systemic effects in the recipient or beneficiary of that therapy, but generating the most appropriate beneficial cellular or tissue response in that specific situation, and optimizing the clinically relevant performance of that therapy [33]. Osteoconductivity is the ability of the graft to function as a scaffold to permit bone growth on its surface or for ingrowth of new bone [34]. In the present study, BioCaP granules were observed in close contact with bone or bone marrow. These findings indicated the good biocompatibility and osteoconductivity of BioCaP, This may be attributed to its biomimetic chemical property. It has been known that the slow delivery of BMP-2 enhances osteoinduction [9], and the slow delivery of vascular endothelial growth factor (VEGF) promotes vascularization [35]. Furthermore, since bone regeneration is a coordinated cascade of events regulated by several growth factors, the local sequential delivery of VEGF and BMP-2 could enhance bone formation compared to BMP-2 alone [36]. Therefore, BioCaP can just turn to this application for the local, sequential and slow delivery of VEGF and BMP-2. Other candidates could be antibiotics and anti-cancer drugs. There is a need for further investigation into this dual release model.

CONCLUSION

In conclusion, the findings indicate that BioCaP with the two delivery modes of BMP-2 can be a promising alternative to the autografts. The coating-incorporated mode of BMP-2 accelerated the bone formation for BioCaP in an early period (4 weeks); the internally-incorporated mode enhanced bone formation in a longer period (8 weeks), which is more efficient for the large bone defect repair. BioCaP with the two delivery modes of BMP-2 give better bone formation compared with the deproteinized bovine bone (a commercial product). BioCaP showed good biodegradability. It might not be too earlier to conclude that BioCaP with the two delivery modes of BMP-2 would be the substitute for autologouse bone for clinic applications. Benefiting from the two delivery modes, BioCaP could be used as a dual delivery vehicle for the sequential delivery of different protein/drugs.

61

Chapter 4

REFERENCES 1. Lewandrowski KU, Gresser JD, Wise DL, Trantol DJ. Bioresorbable bone graft substitutes of different osteoconductivities: a histologic evaluation of osteointegration of poly(propylene glycol-co-fumaric acid)-based cement implants in rats. Biomaterials 2000;21:757-64. 2. Hamann C, Rauner M, Hohna Y, Bernhardt R, Mettelsiefen J, Goettsch C, et al. Sclerostin antibody treatment improves bone mass, bone strength, and bone defect regeneration in rats with type 2 diabetes mellitus. J Bone Miner Res 2012. 3. Cypher TJ, Grossman JP. Biological principles of bone graft healing. J Foot Ankle Surg 1996;35:413-7. 4. Liu Y, Wu G, de Groot K. Biomimetic coatings for bone tissue engineering of critical-sized defects. J R Soc Interface 2010;7 Suppl 5:S631-47. 5. Tanase CE, Popa MI, Verestiuc L. Biomimetic chitosan-calcium phosphate composites with potential applications as bone substitutes: preparation and characterization. J Biomed Mater Res B Appl Biomater 2012;100:700-8. 6. Panzavolta S, Torricelli P, Bracci B, Fini M, Bigi A. Functionalization of biomimetic calcium phosphate bone cements with alendronate. J Inorg Biochem 2010;104:1099-106. 7. Liu Y, Hunziker EB, Layrolle P, De Bruijn JD, De Groot K. Bone morphogenetic protein 2 incorporated into biomimetic coatings retains its biological activity. Tissue Eng Part A 2004;10:101-8. 8. Buschmann J, Harter L, Gao S, Hemmi S, Welti M, Hild N, et al. Tissue engineered bone grafts based on biomimetic nanocomposite PLGA/amorphous calcium phosphate scaffold and human adipose-derived stem cells. Injury 2012;43:1689-97. 9. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106. 10. Lee DD, Tofighi A, Aiolova M, Chakravarthy P, Catalano A, Majahad A, et al. alpha-BSM: a biomimetic bone substitute and drug delivery vehicle. Clin Orthop Relat Res 1999;S396-405. 11. Shin H, Jo S, Mikos AG. Biomimetic materials for tissue engineering. Biomaterials 2003;24:4353-64. 12. Zheng Y, Wu G, Liu T, Liu Y, Wismeijer D, Liu Y. A novel BMP2-coprecipitated, layer-by-layer assembled biomimetic calcium phosphate particle: a biodegradable and highly-efficient osteoinducer. Clin Implant Dent Relat Res 2013;inprint. 13. Liu T, Wu G, Zheng Y, Wismeijer D, Everts V, Liu Y. Cell-mediated BMP-2 release from a novel dual drug delivery system promotes bone formation. Clin Oral Implants Res 2013;submitted. 14. Liu Y, Layrolle P, de Bruijn J, van Blitterswijk C, de Groot K. Biomimetic coprecipitation of calcium phosphate and bovine serum albumin on titanium alloy. J Biomed Mater Res A 2001;57:327-35. 15. Nuss KM, Auer JA, Boos A, von Rechenberg B. An animal model in sheep for biocompatibility testing of biomaterials in cancellous bones. BMC Musculoskelet Disord 2006;7:67. 16. Wang J, Zheng Y, Zhao J, Liu T, Gao L, Gu Z, et al. Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation. J Clin Periodontol 2012;39:98-105. 17. Liu Y, de Groot K, Hunziker EB. BMP-2 liberated from biomimetic implant coatings induces and sustains direct ossification in an ectopic rat model. Bone 2005;36:745-57. 18. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 19. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc 1987;147:229-63. 20. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey.

62

Osteoinductive BioCaP for the repair of bone defects in sheep

Am J Physiol 1990;258:L148-56. 21. Schmitz JP, Hollinger JO. The critical size defect as an experimental-model for craniomandibulofacial nonunions. Clin Orthop Relat R 1986;299-308. 22. Hollinger JO, Kleinschmidt JC. The critical size defect as an experimental model to test bone repair materials. J Craniofac Surg 1990;1:60-8. 23. Theiss F, Apelt D, Brand B, Kutter A, Zlinszky K, Bohner M, et al. Biocompatibility and resorption of a brushite calcium phosphate cement. Biomaterials 2005;26:4383-94. 24. Apelt D, Theiss F, El-Warrak AO, Zlinszky K, Bettschart-Wolfisberger R, Bohner M, et al. In vivo behavior of three different injectable hydraulic calcium phosphate cements. Biomaterials 2004;25:1439-51. 25. Nunamaker DM. Experimental models of fracture repair. Clin Orthop Relat Res 1998;S56-65. 26. Liu Y, Huse RO, de Groot K, Buser D, Hunziker EB. Delivery mode and efficacy of BMP-2 in association with implants. J Dent Res 2007;86:84-9. 27. Manolagas SC, Jilka RL. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med 1995;332:305-11. 28. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 29. Zhang Z, Egana JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al. Cell-based resorption assays for bone graft substitutes. Acta Biomater 2012;8:13-9. 30. Liu T, Wu G, Wismeijer D, Gu Z, Liu Y. Deproteinized bovine bone functionalized with the slow delivery of BMP-2 for the repair of critical-sized bone defects in sheep. Bone 2013;56:110-18. 31. Pham L, Beyer K, Jensen ED, Rodriguez JS, Davydova J, Yamamoto M, et al. Bone morphogenetic protein 2 signaling in osteoclasts is negatively regulated by the BMP antagonist, twisted gastrulation. J Cell Biochem 2011;112:793-803. 32. Paul S, Lee JC, Yeh LC. A comparative study on BMP-induced osteoclastogenesis and osteoblastogenesis in primary cultures of adult rat bone marrow cells. Growth Factors 2009;27:121-31. 33. Williams DF. On the mechanisms of biocompatibility. Biomaterials 2008;29:2941-53. 34. Albrektsson T, Johansson C. Osteoinduction, osteoconduction and osseointegration. Eur Spine J 2001;10 Suppl 2:S96-101. 35. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 36. Kempen DH, Lu L, Heijink A, Hefferan TE, Creemers LB, Maran A, et al. Effect of local sequential VEGF and BMP-2 delivery on ectopic and orthotopic bone regeneration. Biomaterials 2009;30:2816-25.

63

Chapter 4

64

Chapter 5

A novel BMP2-coprecipitated, layer-by-layer assembled biomimetic calcium phosphate particle: a biodegradable and highly-efficient osteoinducer

Yuanna Zheng, Gang Wu, Tie Liu, Daniel Wismeijer, and Yuelian Liu.

Clinical Implant Dentistry and Related Research, 2013 Mar 4. doi: 10.1111/cid.12050. [Published online]

65

Chapter 5

ABSTRACT

Statements of the problem: To repair large-size bone defects, most bone-defect-filling materials in clinic need to obtain osteoinductivity either by mixing them with particulate autologous bone or adsorbing bone morphogenetic protein-2 (BMP-2). However, both approaches encounter various limitations. In this study, we hypothesized that our novel particles of biomimetic BMP-2-coprecipitated calcium phosphate (BMP2-cop.BioCaP) could serve as an independent and biodegradable osteoinducer to induce bone formation efficiently for these bone-defect-filling materials, e.g. deproteinized bovine bone (DBB). Method of study: We alternately layer-by-layer assembled amorphous and crystalline CaP triply to enable a “bamboo-like” growth of the particles. We functionalized BioCaP by coprecipitating BMP2 into the most outer layer of BioCaP. We monitored the degradation, osteoinductivity and foreign-body reaction of either BMP2-cop.BioCaP or its combination with DBB in an ectopic site in rats. Results: After 5 weeks, the BMP2-cop.BioCaP significantly induced new bone formation not only alone but also when mixed with DBB. Its osteoinductive efficiency was 10-fold higher than the adsorbed BMP2. Furthermore, BMP2-cop.BioCaP also reduced significantly the host foreign body reaction to DBB in comparison with the adsorbed BMP2. After a 5-week implantation, more than 90% of BMP2-cop.BioCaP degraded. Conclusions: These findings indicate a promising clinical potential for BMP2-cop.BioCaP in the repair of large-size bone defects.

Keywords: Biomimetic; Bone morphogenetic protein; Bone regeneration; Calcium phosphate; Layer-by-layer; Osteoinducer

INTRODUCTION

Large-size bone defects exceed the self-healing capacity of bone tissue and often a pro-fibrotic microenvironment is formed in the defects [1]. To realize their osseous restoration, bone-defect-filling materials are indispensible. Although autografts are still regarded as the “gold-standard” bone-defect-filling materials, their application is still limited because of the low available quantity as well as donor-site pain and morbidity [2]. Consequently, allografted, xenografted and synthetic calcium phosphate (CaP)-based materials (e.g. deproteinized bone and biphasic CaP) are widely adopted clinically for the treatment of large-size bone defects. These materials are also highly osteoconductive which enhances the migration of osteogenic cells. However, such an enhancement is still too limited to realize osseous restoration. They intrinsically lack osteoinductivity for inducing bone regeneration in a pro-fibrotic environment. One common approach used clinically is to combine bone-defect-filling materials with ground autografts [3] — which supplies the necessary osteogenic elements — for repairing large-size bone defects. In this case, the limitations above mentioned of autografts ensue. One promising approach to this problem is to confer osteoinductivity to these CaP-based materials by using an osteogenic agent, such as bone morphogenetic protein 2 (BMP2). BMP2 is a dimeric disulfide-linked polypeptide growth factor under transforming growth factors- superfamily. BMP2 has been approved by FDA and shown

66

A highly-efficient osteoinducer to induce bone formation in animal studies and clinical trials [4-7]. A consensus has been reached that the in-vivo osteoinductive efficiency of BMP2 is highly dependent on its release kinetics. The present mode of delivery in clinic ─ superficial adsorption of BMP2 onto bone-defect-filling materials [8] ─ is associated with a rapid and burst release [9, 10]. Most of such delivered BMP2 is released too rapidly to induce a sustained osteogenic response at the site of the implantation. This difficulty cannot be overcome satisfactorily merely by increasing the loading dose of BMP2. Apart from the tremendous expense that would be incurred, the transient high local concentration of BMP2, which would be generated, could induce deleterious side effects, such as an excessive stimulation of local bone resorption and the induction of bone formation at unintended sites [11-13]. To be optimally osteoinductive, BMP2 needs to be delivered to target sites at low level concentrations in a sustained manner. One such approach is to coprecipitate BMP2 into a thin layer of biomimetic CaP coating that is prepared on the surfaces of biomaterials [4]. We have recently shown that coating-coprecipitated BMP2 induced a significantly higher volume of new bone surrounding the biomaterials than the superficially adsorbed BMP2 [14]. In addition, the coating-coprecipitated BMP2 could also suppress significantly the host foreign-body reaction to the biomaterials, while the superficially adsorbed BMP2 could not [15]. On the other hand, although the biomimetic coating technique is broadly applicable to a series of bone-defect-filling materials [16], its application is not unlimited because of the dependence of coating growth on the physicochemical properties of the underlying biomaterials as well as the need to prepare the coatings on these materials. Recently, we made a breakthrough in modifying the biomimetic coating procedure. Thereby, we have for the first time alternately layer-by-layer assembled BMP2-coprecipitated biomimetic CaP particles (BMP2-cop.BioCaP) that could serve as an independent “osteoinducer”. This novel BMP2-cop.BioCaP was designed to be mixed directly with clinically-used bone-defect-filling materials to induce bone formation. In this study, we monitored the biological properties of BMP2-cop.BioCaP such as degradation, osteoinductivity and foreign-body reaction. We also ascertained whether BMP2-cop.BioCaP could efficiently induce bone formation surrounding, and suppress the host foreign-body reaction to a clinically-used bone-defect-filling material ─ deproteinized bovine bone (DBB).

MATERIALS AND METHODS

In-vitro investigation Preparation of Layer-by-layer assembled biomimetic calcium phosphate (BioCaP) particles with or without coprecipitated BMP2 The protocol (Fig. 1) to produce the Layer-by-layer assembled BioCaP particles was derived from our well established biphasic biomimetic coating protocols [4, 17, 18]. Briefly, micro-particles of amorphous CaP were obtained in 2000ml of a five-fold-concentrated simulated body fluid [684mM NaCl; 12.5mM CaCl2·2H2O; 21mM NaHCO3; 5mM Na2HPO4·2H2O and 7.5mM MgCl2·2H2O (Sigma, St. Louis, USA)] for 24 hours at 37C. Thereafter, the amorphous CaP micro-particles were immersed in 1000ml of a supersaturated calcium phosphate solution [40mM HCl; 2mM

67

Chapter 5

Na2HPO4·2H2O; 4mM CaCl2·2H2O; 50mM TRIS base (Sigma, St. Louis, USA) (pH 7.4)] for 48 hours at 37C. Thereby, a thick layer of crystalline CaP was deposited on amorphous CaP micro-particles. After drying at room temperature, these particles were then immersed in the 5-fold simulated body fluid (24 hours) and the supersaturated calcium phosphate solution (48 hours) alternately for a total of three cycles. During the preparation of the final crystalline CaP layer, BMP2 (INFUSE® Bone Graft, Medtronic, USA) was introduced into this supersaturated calcium phosphate solution at a final concentration of 2µg/ml and coprecipitated with the crystalline CaP layer. The samples were then air-dried. The entire procedure was conducted under sterile conditions.

Figure 1. Schematic graphs demonstrate the layer-by-layer assembling process of biomimetic (CaP). Micro-particles of amorphous CaP that were initially obtained from a 5-fold simulated body fluid were immersed into supersaturated CaP solution for 48 hours and 5-fold simulated body fluid for 24 hours alternately. Thereby, amorphous CaP and crystalline CaP were layer-by-layer assembled. Then, the particles were immersed into a supersaturated CaP solution with 2µg/ml BMP2. After 48 hours, the particles were air-dried and ready for use. The increase of particle size was attributed to both the layer-by-layer growth of coatings and the aggregation of particles by the growing coatings.

Surface characterization of the BioCaP The surface characteristics of BioCaP were evaluated in a scanning electron microscope (XL 30, Philips, the Netherlands). For this purpose, samples of the material were mounted on aluminium stubs and sputtered with gold particles to a thickness of 10-15nm.

Determination of the amount of the coprecipitated BMP2 The amount of coprecipitated BMP2 was determined using a commercially available enzyme-linked immunosorbent assay (ELISA) kit (PeproTech, London, UK). 0.05g of

68

A highly-efficient osteoinducer

BMP2-cop.BioCaP was dissolved in 1ml 0.5M EDTA (pH 8.0). The ELISA assay was performed according to the manufacturer's instructions. Three samples were used for this purpose.

Confirmation of the homogeneous distribution of the coprecipitated protein by fluorescence microscopy To investigate the distribution of the coprecipitated protein within BioCaP, BMP2 was substituted by a model protein ─ bovine serum albumin that had been conjugated with fluorescein isothiocyanate [19] [FITC-BSA (Sigma, St. Louis, USA)]. FITC-BSA was introduced into the supersaturated calcium phosphate solution at a final concentration of 2µg/ml. After freeze-drying, the coated samples were embedded in methylmethacrylate. 600m-thick sections were prepared and affixed to Plexiglas holders. These sections were then ground down to a thickness of 80 m for an inspection in a fluorescence microscope.

In-vitro monitoring of the release kinetics of the coprecipitated protein in BioCaP To monitor the release kinetics of the coprecipitated protein in BioCaP, FITC-BSA (2µg/ml) was introduced into supersaturated calcium phosphate solution for the final immersion. Six samples were used to determine the total amount of coprecipitated FITC-BSA. These samples were immersed in 1ml of 0.5% EDTA (pH 8.0) and vortexed twice for 5 minutes to ensure complete dissolution of coatings. The supernatants were withdrawn for analysis of total loading of FITC-BSA. To monitor the release kinetics, six samples of DBB mixed with FITC-BSA-cop.BioCaP at a volume ratio of 4:1 and six samples of DBB bearing an equivalent amount of adsorbed FITC-BSA (included for the purpose of comparison and prepared likewise as the adsorption of BMP2) were incubated in sealed 10-ml glass tubes containing 2ml of phosphate-buffered 0.9% saline (pH 7.4). The tubes were incubated for up to 35 days in a shaking waterbath (60 agitations/min), which was maintained at 37C. The sampling and measurement with spectrophometer were performed following the protocol as previously published [15]. Fluorescence readings were converted to amounts of protein using a standard curve, which was generated by preparing a dilution series of FITC-BSA in 5ml of phosphate-buffered 0.9% saline. The temporal release of FITC-BSA was expressed as a percentage of the total amount that had been coprecipitated into the crystalline layer of the BioCaP or that had been adsorbed directly onto the DBB particles.

In-vivo investigation We adopted a subcutaneous bone induction model in rats to further evaluate the BMP2-cop.BioCaP in vivo in aspects of degradation, osteoinductivity and foreign-body reactivity. We measured the following parameters: 1) volume density of newly formed bone; 2) volume density of foreign-body giant cells; 3) volume density of BioCaP and 4) osteoinductive efficiency of BMP2.

Grouping As an experimental animal model, we used adult male Wistar rats (200-220g). Six groups were established (n=6 animals per group): (1) BioCaP; (2) BMP2-cop.BioCaP; (3) DBB alone (4) DBB bearing adsorbed BMP2; (5) DBB mixed with 0.07cc BioCaP; (6) DBB

69

Chapter 5 mixed with 0.07cc BMP2-cop.BioCaP. The amount of BMP2-cop.BioCaP (0.07) was determined according to our previous study [15]. It showed that about 10-15µg of the coating-coprecipitated BMP2 could sufficiently induce bone formation. 0.07cc BMP2-cop.BioCaP contains 10.29±1.94µg BMP2 according to the ELISA result. 0.15g of DBB (about 0.35cc) per sample was used. The samples of DBB bearing adsorbed BMP2 (about 13.5µg) were prepared as described previously [15]. The loading process was achieved by introducing a 75-µl aliquot of stock solution (0.18µg/µl) into 1-ml Eppendorf tubes containing 0.15g of DBB particles.

Surgery and histology Animal experiments were conducted with the permission of and in accordance with the regulations laid down by the Animal Protection Commission of the State of Bern (Switzerland). 18 rats were used in this study. Each rat received two samples from two different groups: they are either non-BMP-2-containing discs (group 1, 3, 5) or BMP-2-containing discs (group 2, 4, 6). To the end, 6 animals were used for each group. (n=6 animals per group). The rats were acclimatized to their new surroundings for 5 days. Housing is in compliance with the national guidelines for animal experimentation. Surgery was performed under conditions of general anesthesia [using Vetalar® (ketamine hydrochloride) (Boehringer Ingelheim Vetmedica, Inc., St. Joseph, USA] [4]. Two samples per rat were surgically implanted within lateral dorsal subcutaneous pockets (one on the left side and one on the right), and were trapped therein by suturing the incision site. After surgery, the rats were kept in cages of Animal facility of Bern University for 5 weeks. The animals were fed ad libitum with hay, granulated food and water. Five weeks after surgery, the samples were retrieved, chemically fixed and embedded in methylmethacrylate as previously reported [4, 18]. By applying a systematic random-sampling strategy [20], the samples were sawed vertical to the short axis, into 10-12 slices of 600µm-thickness with 1mm apart. Odd- or even-numbered slices of each sample were separately mounted on Plexiglas holders and polished. The odd-numbered slices were surface-stained with McNeal’s Tetrachrome, basic Fuchsine and Toluidine Blue O [21] for the histomorphometric analysis of various parameters (see below). The even-numbered slices were subjected to the tartrate-resistant acid phosphatase (TRAP) reaction [4, 22] and counterstained with Methyl Green. They were used to estimate the volume density of multinucleated osteoclasts. Applying a two-step systematic random-sampling strategy, 25-30 images at a final magnification of 320 were recorded in a Nikon-Eclipse light microscope and printed in color for the histomorphometric analysis.

Histomorphometric analysis In the present study, the space under the fibrous capsule that embraced the whole block of implants (subcapsular space) was taken as the reference space. The reference space was estimated using Cavalieri’s methodology [23]. This involves measuring the cross-sectional area of a defined number of tissue sections at a fixed distance apart through the reference volume. The cross-sectional area of each section was estimated using the point-counting technique [24]. The volume densities of bone, of the multinucleated cells and of the remaining

70

A highly-efficient osteoinducer materials were determined stereologically from its area density on tissue sections by the point-counting technique [24]. The volume density of foreign-body giant cells (FBGCs) was obtained by subtracting that of TRAP-positive osteoclasts from that of multinucleated cells [4]. To compare the foreign-body reaction to either BMP2-cop.BioCaP or BioCaP or DBB in different groups, the volume dentisity of FBGCs was normalized to the corresponding volume density of BMP2-cop.BioCaP or BioCaP or DBB. The total volume of bone and the remaining BioCaP material were estimated by multiplying the volume densities of each parameter by the corresponding subcapsular reference volume. The osteoinductive efficiency of BMP2 was estimated by dividing the total volume of bone by the amount of BMP2. Since more than 90% of BMP2-cop.BioCaP was degraded, we assumed all the coprecipitated BMP2 in its outer layer was completely used. The BMP2 that was adsorbed onto DBB should also be exhausted after 5 weeks since it exhibited a burst release. Therefore, we use the total loading of BMP2 to estimate the osteoinductive efficiency.

Statistical analysis All data are presented as mean values together with the standard deviation (Mean±SD). Data were compared using a one-way analysis of variance (ANOVA) with the significance level being set at p<0.05. Post-hoc comparisons were made using Bonferroni’s corrections.

RESULTS

In-vitro characterization: In this study, we assembled 3-dimensional particles using this novel biomimetic layer-by-layer assembling technique. Under scanning electron microscopy, the amorphous CaP microparticles that were derived from the 5-fold simulated body fluid in the first cycle showed morphology of irregular clusters of microspheres with a diameter of 1.5-3µm (Fig. 2A). After immersing these amorphous CaP microparticles in supersaturated calcium phosphate solution for 48 hours, a crystalline CaP deposited on their surfaces and showed plate or needle-like crystals (Fig. 2B). After three cycles of alternate immersion, the particle size increased from the initial 5-20 µm up to 100-1000 µm (Fig. 2D) with a crystalline outer layer (Fig. 2C). The coprecipitated protein is located within the whole outer crystalline CaP layer (Fig. 3A). As anticipated, the FITC-BSA that was adsorbed onto DBB was released rapidly, being completely exhausted after 13 days (Fig. 3B). In contrast, protein that was coprecipitated into BioCaP was released gradually and at a steady rate after the 3rd day until the 35th day, at which juncture the initial depot had been depleted by no more than 50.1% (Fig. 3B). The total loading of BMP2 in 0.35cc BMP2-cop.BioCaP is 51.13±9.68 µg with a coprecipitation rate of 30.1±5.7%.

71

Chapter 5

Figure 2. Scanning electron micrographs depicting the morphologies of the initial amorphous CaP particles (A), the initial layer of crystalline CaP (B), and the final BMP2-cop.BioCaP (C&D). Bars=5m in A, B, and C. Bar=200m in D.

Figure 3. (A) Fluorescence micrographs depicting the distribution of coprecipitated protein in the outer layer of BioCaP. FITC-BSA (green signal) was used to as a substitute for bone morphogenetic protein-2. Bar=100m. (B) Graph depicting the in-vitro release kinetics of BMP2 from DBB with BMP2-cop.BioCaP and DBB with adsorbed BMP2.

72

A highly-efficient osteoinducer

In-vivo study: Histological results Five weeks after the subcutaneous implantation, BioCaP distributed compactly and did not induce new bone formation (Fig. 4A). Connective tissue infiltrated into the BioCaP particles, on the surfaces of which multinucleated FBGCs was frequently found (Fig. 4A1). In contrast, BMP2-cop.BioCaP distributed loosely and induced a large volume of new bone (Fig. 4B). BMP2-cop.BioCaP was tightly integrated into the new bone (Fig. 4B1).

Figure 4. Light micrographs of the cross-sections through BioCaP (A&A1) and BMP2-cop.BioCaP (B&B1) after a 5-week implantation in subcutaneous site in rats. The sections were stained with McNeal’s Tetrachrome, basic Fuchsine and Toluidine Blue O. Yellow arrows points to the foreign-body giant cells (FBGCs) lying on BioCaP. Black arrows points to the remaining BioCaP. Asterisks indicates the newly formed bone. Bars=200m in A and B. Bars=30m in A1 and B1.

Five weeks after implantation, new bone was only found surrounding the DBB either with adsorbed BMP2 (Fig. 5B) or mixed with BMP2-cop.BioCaP (Fig. 5D). No new bone formation was found surrounding the DBB either alone (Fig. 5A) or mixed with BioCaP (Fig. 5B). Bone was deposited abundantly surrounding the DBB with BMP2-cop.BioCaP (Fig. 5D), but only sporadically surrounding the DBB with an adsorbed BMP2 (Fig. 5B). The remaining BioCaP showed clusters of round or ellipse or irregular microparticles (Fig. 5C1). They distributed among the DBB particles (Fig. 5C)

73

Chapter 5 with connective tissue filling in between (Fig. 5C1). For the DBB with BMP2-cop.BioCaP, bone tissue formed with BMP2-cop.BioCaP as centre and deposited on the surfaces of both BMP2-cop.BioCaP and DBB (Fig. 5D). Bone tissue was found tightly integrated with BMP2-cop.BioCaP and DBB without intervening tissue (Fig. 5D1).

Histomorphometric results Volume density of bone surrounding BMP2-cop.BioCaP was [0.36±0.07 (mm3/mm3)] (Fig. 6A). The remaining percentage of BMP2-cop.BioCaP (5.6±2.1%) was significantly lower than that of BioCaP (34.2±6.4%) (Fig. 6B). The volume ratio of FBGCs to BMP2-cop.BioCaP [0.063±0.0198 (mm3/mm3)] was also significantly lower than that to BioCaP [0.110±0.0188 (mm3/mm3)] (Fig. 6C). Volume density of bone surrounding DBB mixed with BMP2-cop.BioCaP [0.06±0.03 (mm3/mm3)] was significantly higher than that surrounding DBB with adsorbed BMP2 [0.007±0.009 (mm3/mm3)] (Fig. 7A). The osteoinductive efficiency of BMP2 in the group of DBB mixed with BMP2-cop.BioCaP was 10-fold higher than that in the group of DBB with adsorbed BMP2 (Fig. 7B). The remaining percentage of BioCaP in the group of DBB with BMP2-cop.BioCaP (8.4±5.5%) was significantly lower than that in the group of DBB with BioCaP (24.9±6.1%) (Fig. 6B). The mixture with DBB did not significantly influence the remaining percentage of BioCaP regardless of the coprecipitation of BMP2 (Fig. 6). The volume ratio of FBGCs to BMP2-cop.BioCaP [0.013±0.018 (mm3/mm3)] was also significantly lower than that to BioCaP [0.155±0.019 (mm3/mm3)] at the presence of DBB (Fig. 6C). The volume ratio of FBGCs to the DBB mixed with BMP2-cop.BioCaP [0.009±0.005 (mm3/mm3)] was significantly lower than that to the DBB either alone [0.039±0.012 (mm3/mm3)] or with adsorbed BMP2 [0.038±0.006 (mm3/mm3)] or with BioCaP [0.043±0.004 (mm3/mm3)] (Fig. 7C).

74

A highly-efficient osteoinducer

Figure 5. Light micrographs of the cross-sections through DBB alone (A), DBB with adsorbed BMP2 (B), DBB with BioCaP (C&C1) and DBB with BMP2-cop.BioCaP (D&D1) after a 5-week implantation in subcutaneous site in rats. The sections were stained with McNeal’s Tetrachrome, basic Fuchsine and Toluidine Blue O. Asterisks indicates the newly formed bone. Bars=200m in A, B, C, and D. Bars=30m in C1 and D1.

75

Chapter 5

Figure 6. Graph depicting the volume density of new bone (A), percentage of remaining BioCaP (B) and volume ratio of foreign-body giant cells (FBGCs) to BioCaP that were associated with BioCaP within the subcapsular space (reference volume) for the four groups, 5 weeks after subcutaneous implantation in rats. Mean values (n=6 animals per group) are represented together with the standard deviation. *: p< 0.05; **: p< 0.01; ***: p< 0.001.

Figure 7. Graph depicting the volume density of new bone (A), osteoinductive efficiency of BMP2 (B) and volume ratio of foreign-body giant cells (FBGCs) to BioCaP that were associated with BioCaP within the subcapsular space (reference volume) for the four groups, 5 weeks after subcutaneous implantation in rats. Mean values (n=6 animals per group) are represented together with the standard deviation. *: p< 0.05; **: p< 0.01; ***: p< 0.001.

DISCUSSION

In this study, we have for the first time developed 3-dimensional biomimetic CaP (BioCaP) particles (100-1000µm) by modifying the principle for preparing the thin (10-50µm), and substrate-dependent biomimetic CaP coatings. In this novel particle, the advantage of the coatings in coprecipitating and slowly releasing proteinaceous cytokines was maintained. We showed that this novel BMP2-cop.BioCaP, serving as an independent “osteoinducer”, could induce bone formation efficiently and suppress the host foreign-body reaction when it was mixed with DBB ─ a clinically-used bone-defect-filling material. In addition, BMP2-cop.BioCaP also exhibited a proper degradation rate in vivo. In our previous studies, we have already shown that the BMP2-coprecipitated biomimetic coating is very broadly applicable to bone-defect-filling materials and dental implants. This was proven by the success in the preparation of this coating on a broad range of biomaterials (e.g. metallic [4, 25], inorganic [15], polymeric materials [26]) that

76

A highly-efficient osteoinducer have completely different geometries, topographies and surface chemistries [16]. Albeit so, this type of biomimetic coating on bone-defect-filling materials has the limitation that their growth relies still highly on the proper surface roughness and/or active surface chemistry of the bone-defect-filling materials [16]. In this study, we modified the biomimetic coating technique and developed this BMP2-cop.BioCaP with an aim of completely breaking through these limitations. The BMP2-cop.BioCaP exhibited no dependence on the physiochemical properties of bone-defect-filling materials and thus can possibly be applied with any kind of granular bone-defect-filling materials used clinically. Meanwhile, this BMP2-cop.BioCaP is also easily handled clinically, which will significantly favor its clinical application. The alternate assembling of the amorphous and crystalline layer was indispensible to increase significantly the volume of BioCaP particles. This is because the amorphous CaP layer is very thin (1.5-10 µm) and the crystalline CaP is hardly beyond 100 µm. By this alternate layer-by-layer approach, we use the amorphous CaP layer as a connection and seeding layer for the growth of another layer of crystalline CaP. The BioCaP grows in a “bamboo-like” pattern with the amorphous CaP as the nodes and the crystalline CaP as the internodes. After three cycles of alternate soaking in 5-fold simulated body fluid and supersaturated calcium phosphate solution alternately (Fig. 1), the size of the BioCaP significantly increased from the initial 5-20µm to 100µm-1mm (Fig. 2). The increase in size was attributed both to the “bamboo-like” layer-by-layer growth of coatings and to the aggregation of underlying particles by the growing coatings (Fig. 1). The current size of BMP2-cop.BioCaP seemed correct for sustaining the osteoinductive effect of coprecipitated BMP2, since a large amount of new bone was induced with high efficiency (Fig. 7B). Besides the size, the degradability of a CaP-based biomaterial is very important for the in-vivo longevity and efficacy of its biological effects [27]. After 5 weeks, 60-82% of BioCaP degraded (Fig. 6B), which indicated a significantly higher degradability of BioCaP than most of the clinically-used, CaP-based bone-filling materials. Such a rapid degradation is associated with its high dissolubility of BioCaP. This is because BioCaP was prepared in biomimetic principle without the involvement of non-physiological conditions (e.g. high temperature) and was composed of both amorphous CaP and crystalline calcium-deficient hydroxiapatite with a low crystallity [26]. In contrast, most of the clinically-used bone-defect-filling materials are sintered, which leads to the significantly increased crystallinity and thus decreased dissolubility [28]. Apart from the spontaneous dissolution, the degradation of a material is also accelerated by many types of cells (e.g. fibroblasts, monocytes/macrophages) through phagocytotic mechanisms [29]. When their phagocytic capacity is exceeded, macrophages can also fuse to form FBGCs. In contrast, these multinucleated FBGCs had a significantly higher resorptive efficiency [30] and played a major role in the degradation of BioCaP. Interestingly, although the volume ratio of FBGCs to BMP2-cop.BioCaP was significantly decreased (Fig. 6C), the degradation rate of BMP2-cop.BioCaP was significantly increased in comparison with BioCaP (Fig. 6B). In fact, the suppression of FBGCs to CaP coatings in the presence of coprecipitated BMP2 could be found from 2-3 weeks [4]. These findings suggested that other resorption mechanisms played key roles in the degradation of BMP2-cop.BioCaP. The activities of osteoblasts and osteoclasts during the osteogenesis may account for this phenomenon.

77

Chapter 5

Besides, except phagocytic activity [29], osteoblasts-mediated mineralization can generate many protons [31] that may promote the degradation of BMP2-cop.BioCaP. Conventionally, these protons have to be neutralized by an extracellular buffering system to prevent their accumulation [32]. CaP materials with a high dissolubility may directly neutralize the protons, which promotes the activities of osteoblasts. The calcium and phosphate ions generated in this way can greatly support the process of osteogenesis. Consequently, a CaP material that bears the greater solubility shows the higher osteoconductivity [33]. In this study, the osteogenesis was significantly promoted by the coprecipitated BMP2, which also increased significantly the osteoblast-mediated degradation and reuse of BioCaP. On the other hand, the mixture with DBB did not significantly influence the degradation rate of either BioCaP or BMP2-cop.BioCaP (Fig. 6B), which indicated that the degradation property of BMP2-cop.BioCaP was not influenced by the targeting bone-defect-filling materials. The release kinetics is a crucial factor for the osteoinductive efficiency of BMP2. In a clinical application, BMP2 is simply adsorbed superficially onto the bone-defect-filling materials, which is associated with a high-dose burst release and thus low osteoinductive efficiency [8]. In contrast, the coating-coprecipitated BMP2 showed a slow and sustained release and thus a significantly higher osteoinductive efficiency than the adsorbed BMP2 [14, 15]. In line with this principle, DBB with BMP2-cop.BioCaP induced significantly higher volume density of bone than the DBB with adsorbed BMP2 (Fig. 7A). Accordingly, the osteoinductive efficiency of BMP2 in the group of DBB with BMP2-cop.BioCaP was 10-fold higher than that in the group of DBB with adsorbed BMP2 (Fig. 7B). These findings indicated that BMP2-cop.BioCaP could act as a powerful “osteoinducer” to induce efficiently new bone formation for other granular clinically-used bone-defect-filling materials. Although the newly formed bone originated from the BMP2-cop.BioCaP, it did not stay unattached but integrated tightly onto the DBB (Fig. 5D1) without the intervening of connective tissues. Thereby, DBB, BMP2-cop.BioCaP and the new bone form an interconnected bony network (Fig. 5D). In contrast, for the BioCaP without the coprecipitation of BMP2, BioCaP and DBB were isolated by fibrous connective tissues (Fig. 5C1) and no bone tissue was detected (Fig. 5C). One concern associated with the use of DBB is its biocompatibility. Although DBB can integrate with bone in a pro-osteogenic environment such as in non-critical-sized bone defects and/or in the presence of a sufficiency of autologous bone chips [34], it can provoke significant foreign-body reactions in a pro-fibrotic environment such as at a subcutaneous site [35] or in critical-sized bony defects [1]. Foreign-body reactivity is histologically characterized by the local accumulation of macrophages, their fusion to form FBGCs, and the deposition of dense fibrous connective tissue [30]. FBGCs begin to appear between the 2nd and the 10th day after implantation [36]. They often persist for the whole lifetime of the implant [37] and their presence is known to be associated with the failure of biomaterials [30]. The foreign-body reaction may significantly hinder the regeneration of bone and the osseointegration of DBB. In this study, we found that the volume ratio of FBGCs to DBB was significantly lower in the group of DBB with BMP2-cop.BioCaP than that in the group of either DBB alone, or DBB with adsorbed BMP2, or DBB with BioCaP (Fig. 7C). This finding indicated that BMP2-cop.BioCaP could not only induce bone formation efficiently but also significantly suppress the host

78

A highly-efficient osteoinducer foreign-body reaction to DBB. Such suppression was most probably attributed to the extensive osteogenesis [15]. The suppression of osteogenic activity on the formation of FBGCs may be partially mediated by the elevated levels of osteopontin, that is enriched during bone regeneration. Osteopontin was previously shown to suppress the fusion of macrophages into FBGCs both in vitro and in vivo [38]. The volume ratio of FBGCs to DBB in the group of DBB with adsorbed BMP2 is similar with that in the group of DBB alone. This finding indicated that the transient high local concentration of BMP2 that was generated by its burst release did not influence the formation and accumulation of FBGCs at the 5-week juncture. Since bone-formation activity cannot be sustained when BMP2 was liberated in a single high-dose burst, the volume density of osseous tissue that was laid down was low (Fig. 7A) and insufficient to hinder the formation of FBGCs (Fig. 7C).

CONCLUSION

In this study, we developed a novel BMP2-cop.BioCaP as an independent slow delivery system for BMP2. BMP2-cop.BioCaP can serve as “osteoinducer” to induce bone formation efficiently and to suppress the foreign-body reaction to a clinically-used bone-defect-filling material. In addition, this material also exhibited proper degradability. All these properties confer this BMP2-cop.BioCaP a very promising potential for the application clinically for the repair of large-size bone defects.

AUTHOR DISCLOSURE STATEMENT

All authors have no conflicts of interest. We sincerely thank Prof. Dr. Tony Hearn for editing the English.

79

Chapter 5

REFERENCES 1. Park JW, Jang JH, Bae SR, An CH, Suh JY. Bone formation with various bone graft substitutes in critical-sized rat calvarial defect. 2009;20:372-8. 2. Sokolsky-Papkov M, Agashi K, Olaye A, Shakesheff K, Domb AJ. Polymer carriers for drug delivery in tissue engineering. 2007;59:187-206. 3. Bannister SR, Powell CA. Foreign body reaction to anorganic bovine bone and autogenous bone with platelet-rich plasma in guided bone regeneration. Journal of periodontology 2008;79:1116-20. 4. Liu Y, de Groot K, Hunziker EB. BMP-2 liberated from biomimetic implant coatings induces and sustains direct ossification in an ectopic rat model. Bone 2005;36:745-57. 5. Lou J, Xu F, Merkel K, Manske P. Gene therapy: adenovirus-mediated human bone morphogenetic protein-2 gene transfer induces mesenchymal progenitor cell proliferation and differentiation in vitro and bone formation in vivo. 1999;17:43-50. 6. Starr AJ. Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures. 2003;85-A:2049. 7. Zhao M, Zhao Z, Koh JT, Jin T, Franceschi RT. Combinatorial gene therapy for bone regeneration: cooperative interactions between adenovirus vectors expressing bone morphogenetic proteins 2, 4, and 7. 2005;95:1-16. 8. Schwarz F, Rothamel D, Herten M, Ferrari D, Sager M, Becker J. Lateral ridge augmentation using particulated or block bone substitutes biocoated with rhGDF-5 and rhBMP-2: an immunohistochemical study in dogs. 2008;19:642-52. 9. Haidar ZS, Hamdy RC, Tabrizian M. Delivery of recombinant bone morphogenetic proteins for bone regeneration and repair. Part B: Delivery systems for BMPs in orthopaedic and craniofacial tissue engineering. 2009;31:1825-35. 10. Haidar ZS, Hamdy RC, Tabrizian M. Delivery of recombinant bone morphogenetic proteins for bone regeneration and repair. Part A: Current challenges in BMP delivery. 2009;31:1817-24. 11. Shields LB, Raque GH, Glassman SD, Campbell M, Vitaz T, Harpring J, et al. Adverse effects associated with high-dose recombinant human bone morphogenetic protein-2 use in anterior cervical spine fusion. Spine 2006;31:542-7. 12. Smith DM, Cooper GM, Mooney MP, Marra KG, Losee JE. Bone morphogenetic protein 2 therapy for craniofacial surgery. The Journal of craniofacial surgery 2008;19:1244-59. 13. Toth JM, Boden SD, Burkus JK, Badura JM, Peckham SM, McKay WF. Short-term osteoclastic activity induced by locally high concentrations of recombinant human bone morphogenetic protein-2 in a cancellous bone environment. 2009;34:539-50. 14. Wu G, Liu Y, Iizuka T, Hunziker EB. The effect of a slow mode of BMP-2 delivery on the inflammatory response provoked by bone-defect-filling polymeric scaffolds. 2010;31:7485-93. 15. Wu G, Hunziker EB, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 16. Liu Y, Wu G, de Groot K. Biomimetic coatings for bone tissue engineering of critical-sized defects. 2010;7 Suppl 5:S631-47. 17. Liu Y, Hunziker EB, Layrolle P, De Bruijn JD, De Groot K. Bone morphogenetic protein 2 incorporated into biomimetic coatings retains its biological activity. Tissue Eng Part A 2004;10:101-8. 18. Liu Y, Huse RO, de Groot K, Buser D, Hunziker EB. Delivery mode and efficacy of BMP-2 in association with implants. Journal of dental research 2007;86:84-9. 19. De Lumen BO, Tappel AL. Fluorescein-hemoglobin as a substrate for cathepsin D and other proteases. 1970;36:22-9. 20. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its

80

A highly-efficient osteoinducer

prediction. 1987;147:229-63. 21. Schenk RK, Olah AJ, Herrmann W. Preparation of calcified tissues for light microscopy. In: GR D, editor. Methods of Calcified Tissue Preparation. Amsterdam: Elsevier Science Publishers B.V.; 1984. p. 1-56. 22. Ballanti P, Minisola S, Pacitti MT, Scarnecchia L, Rosso R, Mazzuoli GF, et al. Tartrate-resistant acid phosphate activity as osteoclastic marker: sensitivity of cytochemical assessment and serum assay in comparison with standardized osteoclast histomorphometry. 1997;7:39-43. 23. Cavalieri B. Geometria Indivisibilibus Continuorum. Bononi: Typis Clemetis Feronij1635. Reprinted as Geometria degli Indivisibili. Torino: Unione Tipografico-Editorice Torinese, 1966. 24. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey. 1990;258:L148-56. 25. Liu Y, Enggist L, Kuffer AF, Buser D, Hunziker EB. The influence of BMP-2 and its mode of delivery on the osteoconductivity of implant surfaces during the early phase of osseointegration. 2007;28:2677-86. 26. Wu G, Liu Y, Iizuka T, Hunziker EB. Biomimetic coating of organic polymers with a protein-functionalized layer of calcium phosphate: the surface properties of the carrier influence neither the coating characteristics nor the incorporation mechanism or release kinetics of the protein. 2010;16:1255-65. 27. Tanuma Y, Anada T, Honda Y, Kawai T, Kamakura S, Echigo S, et al. Granule size-dependent bone regenerative capacity of octacalcium phosphate in collagen matrix. 2012;18:546-57. 28. Bose S, Tarafder S. Calcium phosphate ceramic systems in growth factor and drug delivery for bone tissue engineering: a review. 2012;8:1401-21. 29. Heymann D, Pradal G, Benahmed M. Cellular mechanisms of calcium phosphate ceramic degradation. 1999;14:871-7. 30. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. 2008;20:86-100. 31. Blair HC, Schlesinger PH, Huang CL, Zaidi M. Calcium signalling and calcium transport in bone disease. 2007;45:539-62. 32. Kohn DH, Sarmadi M, Helman JI, Krebsbach PH. Effects of pH on human bone marrow stromal cells in vitro: implications for tissue engineering of bone. 2002;60:292-9. 33. Nagano M, Nakamura T, Kokubo T, Tanahashi M, Ogawa M. Differences of bone bonding ability and degradation behaviour in vivo between amorphous calcium phosphate and highly crystalline hydroxyapatite coating. 1996;17:1771-7. 34. Araujo MG, Lindhe J. Socket grafting with the use of autologous bone: an experimental study in the dog. 2011;22:9-13. 35. Zambuzzi WF, Oliveira RC, Pereira FL, Cestari TM, Taga R, Granjeiro JM. Rat subcutaneous tissue response to macrogranular porous anorganic bovine bone graft. 2006;17:274-8. 36. Ratner BD, Bryant SJ. Biomaterials: where we have been and where we are going. 2004;6:41-75. 37. Salthouse TN. Some aspects of macrophage behavior at the implant interface. 1984;18:395-401. 38. Tsai AT, Rice J, Scatena M, Liaw L, Ratner BD, Giachelli CM. The role of osteopontin in foreign body giant cell formation. 2005;26:5835-43.

81

Chapter 5

82

Chapter 6

A biomimetic osteoinducer enhances the therapeutic effects of deproteinized bovine bone in a sheep critical-sized bone defect (Ø8×13mm) model

Tie Liu, Gang Wu, Yuanna Zheng, Daniel Wismeijer, and Yuelian Liu.

Submitted, 2013

83

Chapter 6

ABSTRACT

Purpose: Most materials used clinically for filling bone defects [e.g. deproteinized bovine bone (DBB)] lack osteoinductivity so that their therapeutic effects are far from satisfactory. We have recently developed a novel biomimetic “osteoinducer” to provide an alternative viable approach. We hypothesize that this osteoinducer could enhance the therapeutic effect of DBB significantly in the repair of critical-sized bone defects (CSBD). Materials and Methods: The osteoinducer as an alternative of autograft was synthesised by assembling a triple layer of amorphous and crystalline calcium phosphate into which BMP2 was co-precipitated (BMP2-cop.BioCaP). DBB mixed with BMP2-cop.BioCaP was tested. These samples and proper positive (autologous bone and DBB mixed with autologous bone) and negative (DBB alone and DBB mixed with BioCaP without BMP2) controls were implanted in the critical-sized bone defects in sheep for 4 and 8 weeks. We assessed the degradability, foreign body reaction and osteoinductivity of the materials to evaluate the efficacy of BMP2-cop.BioCaP of changing the therapeutic effects of DBB. Results: The volume of newly formed bone associated with the test group (BMP2-cop.BioCaP/DBB) was significantly higher than the negative controls and the positive control which is DBB mixed with autologous bone after 4 and 8 weeks; The newly formed bone of the test group was comparable with the autologous bone group after 8 weeks. About 95% BMP2-cop.BioCaP had been degraded and replaced by newly formed bone after 8 weeks. A significantly lower foreign-body reaction was found with BMP2-cop.BioCaP/DBB than the other groups. Conclusions: BMP2-cop.BioCaP significantly induced bone formation and thus enhanced the therapeutic effect of DBB. DBB mixed with this osteoinducer may reduce the use of autograft in the repair of critical-sized bone defects.

Key words: Osteoinducer, Deproteinized bovine bone, Biomimetic calcium phosphate, BMP, Osteoinductive, Critical-sized bone defect, Bone repair

INTRODUCTION

The essence for treating bone fractures and defects is to achieve an adequate volume of bone tissue. When the bone defects are too large to heal by themselves, bone grafting is needed to fill the defect [1, 2]. An autograft is still regarded as the gold standard in treatment since it provides an osteoconductive 3-dimensional scaffold for bone growth, osteogenic cells and osteoinductive growth factors [3]. However, autogenous bone is often associated with limitations such as the need for additional surgical intervention, pain in the donor site, morbidity and a high and unpredictable resorption [4, 5]. These limitations have led to a continual search for alternatives [6, 7]. A better understanding of the biology of the healing of bones and technological development have resulted in the development of numerous alternative materials for filling bone defects, such as allografts, xenografts, and synthetic materials. Most of these materials that are used clinically are highly osteoconductive. This enhances the migration of osteogenic cells. However, most of them lack intrinsic osteoinductivity so that their therapeutic effects on large bone defects are far from satisfactory [8].

84

A highly-efficient osteoinducer for the repair of bone defects

One of the materials is deproteinized bovine bone (DBB) and it is widely used clinically. It is a bovine xenograft [9]. The use of DBB has the potential of reducing morbidity since taking an autograft is unnecessary. Because DBB shows a physical chemical structure similar to that of natural bone [10], it has excellent osteoconductive properties for serving as a scaffold for bone formation [11]. Previous studies have demonstrated its efficiency in the repair of critical-sized bone defect (CSBD) and the augmentation of the maxillary sinus compared with other materials [12, 13]. However, in some cases DBB delays early bone formation which probably results from the lack of osteoinductivity [14, 15]. On the other hand, both surgeons and patients would like to see a shortening of the recovery phase. The use of DBB in combination with a particulate autograft has been suggested for inducing an adequate volume of bone tissue for an excellent restoration. This can provide osteogenic elements [11, 16], but the limitations mentioned above follow. The application of osteogenic growth factors such as bone morphogenetic protein 2 (BMP2) by superficially adsorbing them onto DBB did not promote new bone formation in CSBD [17, 18], since such a delivery mode only gives a short term burst release of BMP2. Consequently, DBB had been considered as an unfeasible carrier for BMP [19]. We have shown previously that the biomimetic calcium phosphate (CaP) coating with incorporated BMP2 can functionalize DBB and render the material efficiently osteoinductive in an ectopic rat mode [20]. The sustained release of BMP2 from the carrier coating enhances osteoinductivity [21]. However, the application of such a coating is also limited since the whole procedure needs several days before the operation [20]. To provide a viable alternative, we recently developed a novel “osteoinducer” by biomimetically assembling calcium phosphate layer by layer and co-precipitating BMP2 into it (BMP2-cop.BioCaP). In the previous study, we have shown that this novel osteoinducer can be directly mixed with DBB and it enhances the bone formation highly efficiently in a rat ectopic model (subcutaneously) [22]. Whether this osteoinducer can work in an osseous environment is still unknown. In this study, we hypothesized that the osteoinducer could enhance the therapeutic effect of DBB significantly in a sheep critical-sized bone defect model, the bone defect size is 8 mm in diameter and 13 mm depth (Ø8×13mm). DBB mixed with BMP2-cop.BioCaP was tested. These samples and proper controls were implanted in the critical-sized bone defects in sheep for 4 and 8 weeks. The degradability, foreign body reaction and osteoinductivity of the materials were evaluated.

MATERIALS AND METHODS

In-vitro preparation and characterization Preparation of layer-by-layer assembled biomimetic calcium phosphate (BioCaP) particles with or without incorporated BMP2 The layer-by-layer assembled BioCaP particles were produced according to our recent publication [22]. Briefly, micro-particles of amorphous CaP were formed and deposited by incubating in a beaker containing 2000ml of a five-fold concentrated simulated body fluid (684mM NaCl; 12.5mM CaCl2·2H2O; 21mM NaHCO3; 5mM Na2HPO4·2H2O and 7.5mM MgCl2·2H2O) for 24 hours at 37°C. These particles served as the cores for the

85

Chapter 6 subsequent layer-by-layer assembly. They were immersed in 1000ml of a supersaturated calcium phosphate solution [40mM HCl; 2mM Na2HPO4·2H2O; 4mM CaCl2·2H2O; 50mM TRIS base (pH 7.4)] for 48 hours at 37°C. Thereby, the first layer of crystalline CaP coating deposited/grew on the amorphous CaP micro-particles. Thereafter, as mentioned above, these particles were immersed in the five-fold concentrated simulated body fluid for 24 hours and subsequently the supersaturated calcium phosphate solution for 48 hours for the second layer of coating. Consequently, the size of BioCaP particles was enlarged by assembling layer-by-layer. In this study, the BioCaP particle was assembled in three cycles. During the preparation of the outer layer, BMP2 (INFUSE® Bone Graft, Medtronic, USA) was introduced into this supersaturated calcium phosphate solution at a final concentration of 2μg/ml and co-precipitated into the outermost crystalline CaP layer (BMP2-cop.BioCaP). The BMP2-cop.BioCaP particles were then freeze-dried and retrieved with the size of 0.25-1mm. The entire procedure was conducted under sterile conditions.

Quantification of the amount of the incorporated BMP2 The amount of incorporated BMP2 was determined by a commercially available enzyme-linked immunosorbent assay (ELISA) kit (PeproTech, London, UK). 0.05g of BMP2-cop.BioCaP (n=6) was dissolved in 1ml 0.5M EDTA (pH 8.0). The ELISA assay was performed according to the manufacturer's instructions.

In-vivo investigation Experimental groups We used an experimental animal model in sheep with drill holes with 8mm in diameter and 13mm in depth within the proximal and distal humerus and femur. Deproteinized bovine bone granules (DBB, size: 0.25–1mm, Bio-Oss®, Geistlich, Switzerland) was used in this study. One experimental and four control groups were established (n=6 sheep per group, Table 1): (1) Autologous bone particles (positive control, taken from the same sheep); (2) DBB granules mixed with autologous bone particles (positive control); (3) DBB granules mixed with BMP2-cop.BioCaP particles (experimental group); (4) DBB granules mixed with BioCaP particles (negative control for the effects of BMP2); (5) DBB granules alone (negative control for the effects of BioCaP and of BMP2).

In Group 2, a 1 : 1 ratio of DBB and autologous bone particles was determined according to previous studies [11, 23-25]. Autologous bone was harvested from the cylindrical bone defects during the surgery and ground to 0.25-1mm particles under sterile conditions. These bone chips were reserved for Group 1 and 2. In Group 3 and 4, 0.59cm3 of DBB (size: 0.25-1.0mm) and 0.07cm3 of BMP2-cop.BioCaP or BioCaP (size: 0.25-1.0mm) per sample were placed into 1-ml Eppendorf tubes and homogeneously mixed by manually shaking. The amount of BMP2-cop.BioCaP was determined according to our previous study [22].

86

A highly-efficient osteoinducer for the repair of bone defects

Table 1 Experimental groups Total Graft materials Loading Groups Abbreviation Dose of Volume of graft BMP2 (per material per sample sample)

(1) Autologous bone AB 0.66cm3 -

3 DBB granules mixed with AB 0.33cm (2) AB DBB - autologous bone DBB 0.33cm3

DBB granules mixed with BMP2-cop.BioCaP DBB BioCaP 3 (3) BMP2-cop.BioCaP particles 0.07cm ; 10.3μg BMP (experimental) DBB 0.59cm3

3 DBB granules mixed with BioCaP 0.07cm ; (4) DBB BioCaP - BioCaP particles DBB 0.59cm3

(5) DBB granules DBB 0.66cm3 -

Experimental animal model A total of 12 adult (2- to 4-year-old) female Australia sheep (40-50 kg in weight) were used in the present study, which was approved by Ethical Committee of School of Stomatology in Zhejiang University. All the animal experiments were carried out according to the ethics laws and regulations of China. Throughout the study, the sheep were treated following the guidelines of animal care established by Zhejiang University. The sheep were subjected to anaesthesia by administering Sumianxin II (0.3 ml/kg, purchased from the Military Veterinary Institute, Quartermaster University of PLA, Chang Chun, China) with the addition of Penicillium (5 × 104 U/kg) and atropine (0.03 mg/kg) at 30 min before surgery. A local anaesthesia (1% lidocaine with 1:100,000 adrenaline) and skin disinfection (0.5% iodophor solution) were applied to the implantation sites. The implantation sites were the proximal part of the diaphysis and distal epiphysis of humerus and femur of sheep.[26] Each sheep can provide 8 totally standardized implantation sites. 5 implantation sites were randomly chosen, and these sites were assigned to the five groups (n=6 sheep per group) according to a randomization protocol [27]. The surgery procedures are shown in Fig. 1. The surgery and animal care were performed and the cylinder-shaped defects (8mm in diameter and 13mm in depth) were created as described in a previous study [26]. Membranes (Bio-Gide®, Geistlich, Switzerland) were used to cover the defects. Samples with surrounding tissues were retrieved at 4 weeks and 8 weeks post-operation.

87

Chapter 6

Figure 1. Surgical procedures: perforation of the cylinder-shaped defects (8mm in diameter and 13mm in depth) (A), and the cylinder bone was shown in the pane; filling materials (B); cover the defects with membranes (C); suture of inner soft tissue (D), and suture of outside soft tissue (E).

Histological procedures Samples were chemically fixed and embedded into a block as previously reported.[20, 28] Applying a systematic random sampling strategy,[29] the samples were sawn vertically to the long axis, into 10 slices of 600-μm thickness, 1 mm apart (interval). All the slices of each sample were mounted separately on plexiglass holders and polished. The slices were surface stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue [20] and examined with a light microscope (Leica).

Histomorphometric analysis In addition to a subjective histological description, 10 slices of each sample was used for quantitative histomorphometric analysis. The volume of newly formed bone, BioCaP and DBB and the volume density of multinucleated giant cells (MGC) on DBB were measured using the point-counting technique.[22, 30] The volume density of MGC on DBB was normalized to the volume of DBB. The volume density of MGC (D) is defined as its volume (Va) per unit volume of DBB (Vb): D = Va / Vb. To evaluate the degradation of BioCaP, the volume of BioCaP before implantation (Vtime0) and after 5 weeks of implantation (V5weeks) was evaluated by using the same histological method. Six samples containing BioCaP (0.07cm3) were reserved for ‘time 0’. Therefore, the percentage of non-degraded BioCaP (P) is defined as: P= V5weeks / Vtime0 × 100%.

Statistical analysis All data are presented as mean values with the standard deviation (SD). Data were compared using a one way analysis of variance (ANOVA), and post hoc comparisons were made using Bonferroni’s corrections. The significance level was set at p< .05.

RESULTS

In-vitro characterization: After three cycles of alternate immersion, the particle size increased up to 100-1000μm with a crystalline outer layer. According to our previous results [22], BMP2 has been successfully incorporated in the outermost coating layer of BioCaP, and an efficient ectopic bone formation was induced by about 10 μg of BMP2 in BMP2-cop.BioCaP. In

88

A highly-efficient osteoinducer for the repair of bone defects the current study, each sample with BMP2-cop.BioCaP contains 10.3±1.9μg BMP2 with an incorporation rate of 30.1±5.7% according to the ELISA result.

Clinical observations At the end of the two-implantation periods (6 sheep per time point), a total of 60 implants were harvested (30 implants at 4 weeks and 30 implants at 8 weeks). All the sheep exhibited good health and all the surgical implant sites were healed well without any complications. No visual signs of inflammation or adverse tissue reaction were observed.

Histological results Representative histological images of each group are shown in Fig. 2. In the positive controls, plenty of autografts were found at 4 weeks (Fig. 2 A, B). After 8 weeks, most autografts had been replaced by newly formed bone (Fig. 2 A1, B1). In the samples containing DBB (Fig. 2 C-E, C1-E1), the whole bone defect was filled uniformly with DBB granules. The newly formed bone was always in close contact with the DBB surface and more mature bone presented at 8 weeks than at 4 weeks. At 4 weeks of implantation, in the samples containing BMP2-cop.BioCaP or BioCaP, newly formed bone with a woven appearance was observed between DBB granules or deposited on DBB (Fig. 3A and B), while in the samples with only DBB, most DBB did not have bone deposition (Fig. 3C). At 8 weeks, an interconnected bone DBB network was observed (Fig. 3D-F). DBB granules were always encapsulated in bone. More bone growth was observed throughout the space between DBB granules in the samples containing BMP2-cop.BioCaP (Fig. 3D) compared with those containing BioCaP (Fig. 3E) and DBB only (Fig. 3F). At a higher magnification, BMP2- cop.BioCaP and BioCaP particles were observed constantly in close contact with new bone or completely encapsulated in the new bone at both time points (Fig. 4). BMP2-cop.BioCaP and BioCaP particle consisted of many calcium phosphate microspheres. Representative images at high magnification of BMP2-cop.BioCaP are shown in Fig. 5A and B. Mononuclear cells were observed in close contact with BMP2-cop.BioCaP particle or in the interior of the particle between the small calcium-phosphate spheres at both time points (Fig. 5A, B). Moreover, multinucleated giant cells (MGCs) were occasionally observed on the surface of BMP2-cop.BioCaP particles at 4 weeks (Fig. 5A), whereas no MGC was found in contact with BMP2-cop.BioCaP at 8 weeks (Fig. 5B). At 4 weeks, MGCs were observed on the surface of DBB granules (Fig. 5C), while MGCs were sporadically found on DBB granules at 8 weeks (Fig. 5D). A light zone of DBB was often observed beneath these MGCs (Fig. 5C and D). The light zone in Fig. 5D seemed to be a sign of resorption.

89

Chapter 6

Figure 2. Representative histological micrographs of the whole bone defect (in the white circle) of each group at 4 weeks (A-E) and 8 weeks (A1-E1) after material placement. Group 1 (A, A1); Group 2 (B, B1); Group 3 (C, C1); Group 4 (D, D1); and Group 5 (E, E1) (see Table 1 for an explanation of the groups). Autograft (#); Newly formed bone (arrow); DBB granules (asterisk). The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue.

Figure 3. Representative histological sections of DBB granules and newly formed bone (arrow) in Group 3, 4 and 5 after 4 and 8 weeks of implantation at higher magnification (see Table 1 for an explanation of the groups). After 4 weeks of implantation, newly formed bone was observed in close contact with DBB granules in Group 3 (A) and Group 4 (B), but not in Group 5 (C). After 8 weeks of implantation, more bone growth was observed throughout the space between DBB granules in Group 3 (D) compared to Group 4 (E) and Group 5 (F). The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Bar=500µm.

90

A highly-efficient osteoinducer for the repair of bone defects

Figure 4. Representative histological sections of BioCaP (asterisk), DBB, and bone (arrow) in Group 3 and 4 after 4 and 8 weeks of implantation at higher magnification (see Table 1 for an explanation of the groups). Group 3 at 4 weeks (A); Group 4 at 4 weeks (B); Group 3 at 8 weeks (C); Group 4 at 8 weeks (D). Most BMP2-cop.BioCaP particles and BioCaP particles without BMP2 were in close contact in newly formed bone at both time points. The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Bar=100µm.

91

Chapter 6

Figure 5. Mononuclear cells (yellow arrow) appeared in the interior of BMP2-cop.BioCaP particles at 4 weeks (A) and 8 weeks (B). Osteocytes (white arrow). Multinucleated giant cells (MGC; arrow head) were occasionally observed on BMP2-cop.BioCaP at 4 weeks (A), while no MGCs were found on BMP2-cop.BioCaP at 8 weeks. MGCs were observed on DBB granules at 4 weeks (C), while MGCs were sporadically observed on DBB at 8 weeks (D). A light zone (black arrow) commonly observed when MGCs could be observed adjacent to DBB. The slices were surface-stained with McNeal's Tetrachrome, basic Fuchsine and Toluidine Blue. Bar=50µm.

Histomorphometric results For each treatment, the volume of new bone at 8 weeks was significantly higher (p<0.05) than that at 4 weeks (Fig. 6). At 4 weeks, the volume of new bone in the group of BioCaP (no BMP2)/DBB was significantly higher than with DBB alone. The volume of new bone associated with BMP2-cop.BioCaP/DBB was comparable with an autograft at 8 weeks, and it was significantly higher than autograft/DBB, BioCaP (no BMP2)/DBB, and DBB alone at both 4 and 8 weeks. The degradation of BMP2-cop.BioCaP and BioCaP increased with increasing implantation time (p<0.05). BMP2-cop.BioCaP showed significantly more degradation than BioCaP at both time points (Fig. 7A). After 8 weeks, about 95% of BMP2-cop.BioCaP was degraded. The volume of DBB at 4 and 8 weeks after implantation revealed that there was no significant difference between the two time points in all the four groups (Fig. 7B). The volume density of multinucleated giant cells (MGC) on the surface of DBB revealed that bone formation decreased significantly with increasing time after

92

A highly-efficient osteoinducer for the repair of bone defects implantation (p<0.05). At 4 weeks after implantation, the volume density of MGC on DBB was the lowest in samples containing BMP2-cop.BioCaP (Fig. 8), while at 8 weeks there were no significant differences among the four groups with DBB.

Figure 6. Graph depicting the volume of newly formed bone within the bone defect at 4 and 8 weeks after implantation for each of the 5 groups (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05.

Figure 7. Graph depicting the percentage of non-degraded BioCaP (A) and the volume of DBB (B) within the bone defect at 4 and 8 weeks after implantation (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05.

93

Chapter 6

Figure 8. Graph depicting the volume density of multinucleated giant cells on the surface of deproteinized bovine bone at 4 and 8 weeks after implantation (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05.

DISCUSSION

The purpose of the present study is to test our hypothesis that our novel osteoinducer ─ BMP2-cop.BioCaP could significantly enhance the therapeutic effects of DBB on the repair of critical-sized bone defects. BMP2-cop.BioCaP is a biodegradable and highly efficient osteoinducer. It resulted in more bone formation than the single use of DBB and it significantly suppressed foreign-body reaction not only in an orthotopic environment, but also in an ectopic environment [22]. The volume of new bone associated with BMP2-cop.BioCaP/DBB was significantly higher than autograft/DBB. These findings indicate that BMP2-cop.BioCaP could significantly enhance the therapeutic effects of DBB on critical-sized bone defects. An ectopic model (subcutaneous) of ossification is useful for testing the principle of an osteoinductive system [28]. However, the osseous environment is different from a non-osseous environment. The critical-sized bone defect (CSBD) model in this study was created by drilling holes in the humerus and femur of sheep according to the well published protocol by Nuss et al [26]. This drill hole model in sheep has proved to be an excellent animal model for testing biomaterials for use in orthopedics, maxillofacial and dental surgery [31]. It allowed the intraosseous implantation of up to 8 different test materials within one animal due to the standardization of the bone defect, while at the same time it can reduce the overall suffering of animals and give the necessary numbers to satisfy statistical requirements [26, 32]. Our previous study has confirmed that these critical-sized bone defects cannot heal by themselves [33]. In addition, the results from sheep are more convincing than those obtained with small laboratory animals because of the similarities in the bone structure of humans [34]. In the current study, all BioCaP particles (with or without BMP2) were observed to be in close contact with bone or entirely encapsulated in the bone, and mononuclear cells (osteocyte-like cells) were observed to be in close contact with BioCaP. This finding

94

A highly-efficient osteoinducer for the repair of bone defects suggests that BioCaP is highly biocompatible, which has been proved in our previous study, since each layer of BioCaP is produced using the biomimetic coating technique [22]. During the preparation of BioCaP, the five-fold concentrated simulated body fluid provided the CaP microspheres which are formed under the nucleation inhibitory 2+ 2- influence of Mg and HCO3 [35, 36]. These microspheres then serve as seed for the subsequent growth of a crystalline latticework of octacalcium phosphate under conditions that are conducive to nucleation [37]. Therefore, the physicochemical property of BioCaP might be similar to the biomimetic coating. A number of studies showed that the coating has excellent biocompatibility, biodegradability, osteoconductivity and the capability of slow delivery of growth factors such as BMP2 and vascular endothelial growth factor (VEGF) [20, 21, 28, 38]. The slow delivery of BMP2 plays a very important role in bone formation [20, 21]. It has been shown that the protein incorporated in BioCaP resulted in a sustained release of protein in vitro and BMP2 delivered by BioCaP led to a highly osteoinductive efficiency in vivo in our previous study [22]. The results in the current study have confirmed the high efficiency of BMP2-cop.BioCaP with 10.3ug of BMP2, which is significantly less than the clinic applications [39]. Moreover, the results revealed that BioCaP without BMP2 is also conducive to bone formation at 4 weeks after implantation. One possible mechanism for this might be related to the degradation of BioCaP which provides calcium for the process of osteogenesis. Moreover, mononuclear cells were observed in the interior of BioCaP particle between the small CaP spheres. This suggests that BioCaP has a porous structure. The porosity also contributes to the degradability of the material [40]. Histomorphometric analysis revealed that BMP2-cop.BioCaP had high degradability and resulted in a significant increase in bone formation. A proper degradability is an essential property for biomaterials [41]. The degradation mechanism of BMP2-cop.BioCaP is associated with its dissolubility (the spontaneous dissolution), and the cell based resorption [21, 22, 28]. There are two possible mechanisms for the cell based degradation of material and they are a phagocytosis mechanism and an acidic mechanism [42]. The phagocytosis mechanism is associated with fibroblasts and monocytes/macrophages and the acidic mechanism is related to multinucleated giant cells (FBGCs, osteoclasts) and osteoblasts. FBGCs and osteoclasts can produce at the surfaces of biomaterials an acidic microenvironment that exists between the cell membrane and the surface of the biomaterial [43, 44]. The mineralization mediated by osteoblasts (osteogenesis) can generate many protons [45] which have to be neutralized conventionally by an extracellular buffering system to prevent their accumulation [46]. Therefore, these protons may promote the degradation of BMP2-cop.BioCaP, since the osteogenesis was significantly promoted in this study by BMP2-cop.BioCaP. Moreover, MGCs were shown to be in close contact with DBB in this study. Histomorphometric analysis revealed that BMP2-cop.BioCaP significantly decreased the formation of MGCs on DBB. This finding coincides with a previous study which showed that BMP2-cop.BioCaP suppressed these cells in an ectopic environment [22]. We assume that these MGCs could be regarded as foreign-body reaction in agreement with our previous studies [20, 22]. In addition, a light zone of DBB appeared beneath the MGCs. This finding is in line with a previous study [16], indicating a sign of resorption of DBB. It should also be noted that MGCs play a critical role in the surface treatment

95

Chapter 6 and the degradation of material.[44] In the present study, DBB served as an osteoconductive scaffold for bone formation [11]. However, the low degradability of DBB is one of its drawbacks [10, 16]. Although the presence of cell based demineralization of DBB was observed, the histomorphometric analysis of the DBB volume demonstrated the low degradability of DBB. The ideal bone regeneration requires that the material can be gradually replaced by new bone in a short period of time [41]. Therefore, we propose that the combination of BMP2-cop.BioCaP and a biodegradable material for filling a bone defect may result in a better regeneration of the bone in a shorter period. Our on-going study is developing a biodegradable biomimetic calcium phosphate bone substitute for filling bone defects.

CONCLUSION

It was shown that BMP2-cop.BioCaP can serve as a highly efficient osteoinducer for inducing bone formation with DBB and for suppressing the foreign-body reaction in a critical-sized bone defect. BMP2-cop.BioCaP also showed good degradability. This novel material has a very promising clinical potential as an osteoinducer which can be a substitute for an autograft and which can enhance significantly the therapeutic effects of materials for filling bone defects.

ACKNOWLEDGMENT

Authors would like to thank Prof. Dr. Tony Hearn for editing the grammar.

96

A highly-efficient osteoinducer for the repair of bone defects

REFERENCES 1. Lewandrowski KU, Gresser JD, Wise DL, Trantol DJ. Bioresorbable bone graft substitutes of different osteoconductivities: a histologic evaluation of osteointegration of poly(propylene glycol-co-fumaric acid)-based cement implants in rats. Biomaterials 2000;21:757-64. 2. Hamann C, Rauner M, Hohna Y, Bernhardt R, Mettelsiefen J, Goettsch C, et al. Sclerostin antibody treatment improves bone mass, bone strength, and bone defect regeneration in rats with type 2 diabetes mellitus. J Bone Miner Res 2012. 3. Cypher TJ, Grossman JP. Biological principles of bone graft healing. J Foot Ankle Surg 1996;35:413-7. 4. Schlegel KA, Fichtner G, Schultze-Mosgau S, Wiltfang J. Histologic findings in sinus augmentation with autogenous bone chips versus a bovine bone substitute. Int J Oral Maxillofac Implants 2003;18:53-8. 5. Widmark G, Andersson B, Ivanoff CJ. Mandibular bone graft in the anterior maxilla for single-tooth implants. Presentation of surgical method. Int J Oral Maxillofac Surg 1997;26:106-9. 6. Younger EM, Chapman MW. Morbidity at bone graft donor sites. J Orthop Trauma 1989;3:192-5. 7. Banwart JC, Asher MA, Hassanein RS. Iliac crest bone graft harvest donor site morbidity. A statistical evaluation. Spine (Phila Pa 1976) 1995;20:1055-60. 8. Van der Stok J, Van Lieshout EM, El-Massoudi Y, Van Kralingen GH, Patka P. Bone substitutes in the Netherlands - a systematic literature review. Acta Biomater 2011;7:739-50. 9. Thaller SR, Hoyt J, Dart A, Borjeson K, Tesluk H. Repair of experimental calvarial defects with Bio-Oss particles and collagen sponges in a rabbit model. J Craniofac Surg 1994;5:242-6. 10. Accorsi-Mendonca T, Conz MB, Barros TC, de Sena LA, Soares Gde A, Granjeiro JM. Physicochemical characterization of two deproteinized bovine xenografts. Braz Oral Res 2008;22:5-10. 11. Simion M, Fontana F, Rasperini G, Maiorana C. Vertical ridge augmentation by expanded-polytetrafluoroethylene membrane and a combination of intraoral autogenous bone graft and deproteinized anorganic bovine bone (Bio Oss). Clin Oral Implants Res 2007;18:620-9. 12. Schmitt JM, Buck DC, Joh SP, Lynch SE, Hollinger JO. Comparison of porous bone mineral and biologically active glass in critical-sized defects. J Periodontol 1997;68:1043-53. 13. Schmitt CM, Doering H, Schmidt T, Lutz R, Neukam FW, Schlegel KA. Histological results after maxillary sinus augmentation with Straumann(R) BoneCeramic, Bio-Oss(R), Puros(R), and autologous bone. A randomized controlled clinical trial. Clin Oral Implants Res 2012. 14. Schwartz Z, Weesner T, van Dijk S, Cochran DL, Mellonig JT, Lohmann CH, et al. Ability of deproteinized cancellous bovine bone to induce new bone formation. J Periodontol 2000;71:1258-69. 15. Araujo M, Linder E, Lindhe J. Effect of a xenograft on early bone formation in extraction sockets: an experimental study in dog. Clin Oral Implants Res 2009;20:1-6. 16. Mordenfeld A, Hallman M, Johansson CB, Albrektsson T. Histological and histomorphometrical analyses of biopsies harvested 11 years after maxillary sinus floor augmentation with deproteinized bovine and autogenous bone. Clin Oral Implants Res 2010;21:961-70. 17. Schwarz F, Rothamel D, Herten M, Ferrari D, Sager M, Becker J. Lateral ridge augmentation using particulated or block bone substitutes biocoated with rhGDF-5 and rhBMP-2: an immunohistochemical study in dogs. Clin Oral Implants Res 2008;19:642-52. 18. Schmitt C, Lutz R, Doering H, Lell M, Ratky J, Schlegel KA. Bio-Oss(R) blocks combined with BMP-2 and VEGF for the regeneration of bony defects and vertical augmentation. Clin Oral Implants Res 2011;1-11.

97

Chapter 6

19. Tuominen T, Jamsa T, Tuukkanen J, Marttinen A, Lindholm TS, Jalovaara P. Bovine bone implant with bovine bone morphogenetic protein in healing a canine ulnar defect. Int Orthop 2001;25:5-8. 20. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 21. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106. 22. Zheng Y, Wu G, Liu T, Liu Y, Wismeijer D, Liu Y. A novel BMP2-coprecipitated, layer-by-layer assembled biomimetic calcium phosphate particle: a biodegradable and highly-efficient osteoinducer. Clin Implant Dent Relat Res 2013;doi: 10.1111/cid.12050. 23. Jensen T, Schou S, Stavropoulos A, Terheyden H, Holmstrup P. Maxillary sinus floor augmentation with Bio-Oss or Bio-Oss mixed with autogenous bone as graft in animals: a systematic review. Int J Oral Maxillofac Surg 2012;41:114-20. 24. Jensen T, Schou S, Stavropoulos A, Terheyden H, Holmstrup P. Maxillary sinus floor augmentation with Bio-Oss or Bio-Oss mixed with autogenous bone as graft: a systematic review. Clin Oral Implants Res 2012;23:263-73. 25. Urban IA, Nagursky H, Lozada JL. Horizontal Ridge Augmentation with a Resorbable Membrane and Particulated Autogenous Bone With or Without Anorganic Bovine Bone-Derived Mineral: A Prospective Case Series in 22 Patients. Int J Oral Max Impl 2011;26:404-14. 26. Nuss KM, Auer JA, Boos A, von Rechenberg B. An animal model in sheep for biocompatibility testing of biomaterials in cancellous bones. BMC Musculoskelet Disord 2006;7:67. 27. Wang J, Zheng Y, Zhao J, Liu T, Gao L, Gu Z, et al. Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation. J Clin Periodontol 2012;39:98-105. 28. Liu Y, de Groot K, Hunziker EB. BMP-2 liberated from biomimetic implant coatings induces and sustains direct ossification in an ectopic rat model. Bone 2005;36:745-57. 29. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc 1987;147:229-63. 30. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey. Am J Physiol 1990;258:L148-56. 31. Theiss F, Apelt D, Brand B, Kutter A, Zlinszky K, Bohner M, et al. Biocompatibility and resorption of a brushite calcium phosphate cement. Biomaterials 2005;26:4383-94. 32. Apelt D, Theiss F, El-Warrak AO, Zlinszky K, Bettschart-Wolfisberger R, Bohner M, et al. In vivo behavior of three different injectable hydraulic calcium phosphate cements. Biomaterials 2004;25:1439-51. 33. Liu T, Wu G, Wismeijer D, Gu Z, Liu Y. Deproteinized bovine bone functionalized with the slow delivery of BMP-2 for the repair of critical-sized bone defects in sheep. Bone 2013;56:110-18. 34. Nunamaker DM. Experimental models of fracture repair. Clin Orthop Relat Res 1998;S56-65. 35. Barrere F, van Blitterswijk CA, de Groot K, Layrolle P. Influence of ionic strength and carbonate on the Ca-P coating formation from SBFx5 solution. Biomaterials 2002;23:1921-30. 36. Barrere F, van BC, de GK, Layrolle P. Nucleation of biomimetic Ca-P coatings on ti6A14V from a SBF x 5 solution: influence of magnesium. Biomaterials 2002;23:2211-20. 37. Wu G, Liu Y, Iizuka T, Hunziker EB. Biomimetic coating of organic polymers with a protein-functionalized layer of calcium phosphate: the surface properties of the carrier influence neither the coating characteristics nor the incorporation mechanism or release

98

A highly-efficient osteoinducer for the repair of bone defects

kinetics of the protein. Tissue Eng Part C Methods 2010;16:1255-65. 38. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 39. Block MS, Achong R. Bone morphogenetic protein for sinus augmentation. Atlas Oral Maxillofac Surg Clin North Am 2006;14:99-105. 40. Karageorgiou V, Kaplan D. Porosity of 3D biomaterial scaffolds and osteogenesis. Biomaterials 2005;26:5474-91. 41. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 42. Heymann D, Pradal G, Benahmed M. Cellular mechanisms of calcium phosphate ceramic degradation. Histol Histopathol 1999;14:871-7. 43. Everts V, de Vries TJ, Helfrich MH. Osteoclast heterogeneity: lessons from osteopetrosis and inflammatory conditions. Biochim Biophys Acta 2009;1792:757-65. 44. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20:86-100. 45. Blair HC, Schlesinger PH, Huang CL, Zaidi M. Calcium signalling and calcium transport in bone disease. Subcell Biochem 2007;45:539-62. 46. Kohn DH, Sarmadi M, Helman JI, Krebsbach PH. Effects of pH on human bone marrow stromal cells in vitro: implications for tissue engineering of bone. J Biomed Mater Res 2002;60:292-9.

99

Chapter 6

100

Chapter 7

Deproteinized bovine bone functionalized with the slow delivery of BMP-2 for the repair of critical-sized bone defects in sheep

Tie Liu, Gang Wu, Daniel Wismeijer, Zhiyuan Gu and Yuelian Liu.

Bone. 2013, 56: 110–118.

101

Chapter 7

ABSTRACT

As an alternative to an autologous bone graft, deproteinized bovine bone (DBB) is widely used in the clinical dentistry. Although DBB provides an osteoconductive scaffold, it is not capable of enhancing bone regeneration because it is not osteoinductive. In order to render DBB osteoinductive, bone morphogenetic protein 2 (BMP-2) has previously been incorporated into a three dimensional reservoir (a biomimetic calcium phosphate coating) on DBB, which effectively promoted the osteogenic response by the slow delivery of BMP-2. The aim of this study was to investigate the therapeutic effectiveness of such coating on the DBB granules in repairing a large cylindrical bone defect (8mm diameter, 13mm depth) in sheep. Eight groups were randomly assigned to the bone defects: (i) no graft material; (ii) autologous bone; (iii) DBB only; (iv) DBB mixed with autologous bone; (v) DBB bearing adsorbed BMP-2; (vi) DBB bearing a coating but no BMP-2; (vii) DBB bearing a coating with adsorbed BMP-2; and (viii) DBB bearing a coating-incorporated depot of BMP-2. 4 and 8 weeks after implantation, samples were withdrawn for a histological and a histomorphometric analysis. Histological results confirmed the excellent biocompatibility and osteoconductivity of all the grafts tested. At 4 weeks, DBB mixed with autologous bone or functionalized with coating-incorporated BMP-2 showed more newly-formed bone than the other groups with DBB. At 8 weeks, the volume of newly-formed bone around DBB that bore a coating-incorporated depot of BMP-2 was greatest among the groups with DBB, and was comparable to the autologous bone group. The use of autologous bone and BMP-2 resulted in more bone marrow formation. Multinucleated giant cells were observed in the resorption process around DBB, whereas histomorphometric analysis revealed no significant degradation of DBB. In conclusion, it was shown that incorporating BMP-2 into the calcium phosphate coating of DBB induced strong bone formation around DBB for repairing a bone defect.

Keywords: Deproteinized bovine bone, Biomimetic calcium phosphate coating, BMP-2, Critical-sized bone defect, Bone repair, Drug delivery

INTRODUCTION

In recent years, a critical-sized bone defect (CSBD) is defined as an intraosseous wound that will not spontaneously heal completely without intervention [1, 2]. Autograft is regarded as the ‘gold standard’ because of its excellent combination of osteoconduction and osteoinduction. However, it is always associated with irregular rates of resorption, pain and morbidity of the donor site, and requires additional surgical procedures. These limitations have already led to the pursuit of alternatives including allografts, xenografts and synthetic alloplasts [3, 4]. Most of them are osteoconductive, while the lack of an intrinsic property of osteoinductivity is always the main problem [5]. To solve this problem, one of the strategies in bone tissue engineering has been the introduction of bone growth factors into a suitable scaffold [2, 6]. The scaffold that acts as a template for cell interactions and provides a structural support for the newly formed tissue is a key component for bone regeneration [7]. Deproteinized bovine bone (DBB), a bovine xenograft of which there is an unlimited supply, is one of the most widely used scaffolds used in bone repair and augmentation in

102

Deproteinized bovine bone functionalized with the slow delivery of BMP-2 clinical dentistry. DBB is derived from a bovine source and is treated by a chemical extraction process to remove all the organic components and pathogens [4]. In terms of its inorganic composition and its isomeric crystalline dimensions, DBB has a physical and chemical structure similar to that of natural bone [8]. It shows osteoconductive properties when it is in close contact with the newly formed bone [9]. However, it was reported that DBB delays the early bone formation [10] and lacks intrinsic osteoinductivity [11]. This suggested using DBB in combination with autogenous bone chips which can then provide osteogenic elements [9, 12]. But this still has the limitations of autografts mentioned above. The addition of platelet-rich plasma which contains various growth factors on DBB did not enhance early and late healing of the bone [13]. The local delivery of mesenchymal stem cells (MSC) by DBB offers the promising potential of augmenting the healing of CSBD [14, 15], but it needs harvesting a cell from a secondary site, which is then expanded in vitro and seeded onto DBB directly prior to implantation. Another simple option is to use osteogenic agents such as bone morphogenetic protein 2 (BMP-2) adsorbed on DBB [16]. However, due to the high burst release of BMP-2, the adsorption mode has not promoted de novo bone formation [17]. To be effective, this mode usually needs very high doses (in the milligram range) [18, 19], and is neither efficient nor cost effective [20, 21]. Therefore, DBB was once considered not to be a feasible carrier for BMPs [22]. It is well known that a slow delivery of BMP-2 plays a crucial role in bone formation [23, 24]. For the slow delivery of BMP-2, technique of a biomimetic calcium phosphate (CaP) coating has been developed and applied on different materials such as titanium implants [25], polymers [26] and DBB [27]. This biomimetic coating deposited on the surface of carrier materials can serve as a three-dimensional reservoir for growth factors. BMP-2 incorporated into the crystalline latticework of this biomimetic coating during its growth (deposition) can retain its biological activity [28]. In the previous studies, the coating incorporated with BMP-2 has been applied on dental titanium implants to improve osteoconductivity and osteoinductivty, especially for the early bone formation (1-3 weeks) in an orthotopic site [29]. Recently, we found that the functionalization of DBB granules with the BMP-2-incorporated biomimetic coating induced efficient bone formation at an ectopic (subcutaneous) site in rats [27]. Moreover, the BMP-2-incorporated biomimetic coating can suppress foreign body reaction which may significantly hinder the regeneration of bone and the osseointegration of DBB. However, Whether this BMP-2-incorporated coating on DBB can facilitate bone formation for the repair of CSBD is not clear. The aim of the current investigation was to study the therapeutic effectiveness of DBB functionalized with coating-incorporated BMP-2 for the repair of critical-sized bone defect. To this end, autologous bone was used as a positive control; DBB granules that bore either a directly or a coating-adsorbed depot of BMP-2, or a coating-incorporated depot of this agent, were filled into the bone defects as testing groups in the adult sheep. There were 8 experimental groups in total. The volumes of newly formed bone within the bone defect 4 and 8 weeks after surgery were estimated histomorphometrically.

MATERIALS AND METHODS

103

Chapter 7

Biomimetic calcium phosphate coating procedure The DBB granules (size: 0.25–1mm, Bio-Oss®, Geistlich, Switzerland) were biomimetically coated, according to a well-established protocol [25, 28, 30, 31], with a layer of crystalline calcium phosphate in the absence or presence of BMP-2 [27]. Briefly, 0.35g of the DBB granules were immersed in 300ml of five-fold-concentrated simulated body fluid (684 mM NaCl; 12.5 mM CaCl2·2H2O; 21 mM NaHCO3; 5 mM Na2HPO4·2H2O; 7.5 mM MgCl2·2H2O) for 24 h at 37 °C and afterwards in 130 ml of a supersaturated solution of calcium phosphate [40 mM HCl; 2 mM Na2HPO4·2H2O; 4 mM CaCl2·2H2O; 50 mM TRIS base (pH 7.4)] for 48 h at 37 °C described as the precious study [27]. The entire procedure was conducted under sterile conditions.

Incorporation of BMP-2 into the calcium phosphate coating BMP-2 (INFUSE® Bone Graft, Medtronic, USA) was present in the supersaturated solution of calcium phosphate at a final concentration of 1μg/ml, and was subsequently coprecipitated into the biomimetic calcium phosphate coating of the DBB granules. The samples were then freeze-dried. The entire procedure was conducted under sterile conditions. The quantification of the amount of BMP-2 encapsulated in the coating was determined using an enzyme linked immunosorbent assay (ELISA) kit (PeproTech EC, London, UK), as described the previous studies [26, 31].

Adsorption of BMP-2 onto DBB granules with or without the coating According to the results of ELISA, 35.0 ± 0.62 μg (mean ± SD) of BMP-2 were incorporated into the coating of each sample. Hence, 35.0 μg of BMP-2 was likewise adsorbed onto each 0.35 g sample of DBB granules with or without the coating as described in the previous study [27]. Briefly, the loading process was achieved by introducing a 200-μl aliquot of a stock solution with 175 μg/ml of BMP-2 into 1-ml Eppendorf tubes containing 0.35g of DBB granules. Finally the DBB granules were homogeneously mixed and wetted. Afterwards, the samples were freeze dried for 24 h. The entire procedure was conducted under sterile conditions

Surface characterization of the coating on DBB The surface characteristic of the calcium phosphate coating with or without incorporated BMP-2 on DBB was evaluated with a scanning electron microscope (SEM, XL 30, Philips, The Netherlands). For this purpose, samples of the material were mounted on aluminum stubs and sputtered with gold particles to a thickness of 10–15 nm.

Confirmation of the homogeneous distribution of a coating-incorporated depot of proteins To confirm the homogeneous distribution of the protein in the crystalline latticework of the coating on DBB granules, BMP-2 was substituted by the model protein bovine serum albumin labeled with fluorescein-isothiocyanate (FITC-BSA, Sigma, St. Louis, MO, USA). FITC-BSA (green signal) was introduced into the supersaturated calcium phosphate solution at a final concentration of 1 μg/ml. The coated samples were embedded in methylmethacrylate, sectioned, and ground [27]. A series of 50-µm -thick sections were prepared for analysis by fluorescence microscopy. Micrographs were taken with a digital camera (Leica, Wetzlar, Germany) mounted on an inverted light

104

Deproteinized bovine bone functionalized with the slow delivery of BMP-2 microscope (Leica) equipped with a fluorescence lamp.

Experimental animal model A total of 12 adult (2- to 4-year-old) female Australia sheep (40-50 kg in weight) were used in the present study, which was approved by Ethical Committee of School of Stomatology, Zhejiang University. All animal experiments were carried out according to the ethic laws and regulations of China. Throughout the study, the sheep were treated following the guidelines of animal care established by Zhejiang University. The sheep were anesthetized by administering Sumianxin II (0.3 ml/kg, purchased from the Military Veterinary Institute, Quartermaster University of PLA, Chang Chun, China) with the addition of Penicillium (5 × 104 U/kg) and atropine (0.03 mg/kg) 30 min before surgery. After applying local anesthesia (1% lidocaine with 1:100,000 adrenaline) and skin disinfection (0.5% iodophor solution) to the implantation sites, the surgery and animal care was performed and the cylinder shaped defects were created (8mm in diameter and 13mm in depth) as described in a previous study [32]. The implantation sites were the proximal part of the diaphysis and distal epiphysis of humerus and femur of 12 adult female sheep. Eight implantation sites per animal were assigned to the eight groups according to a randomization protocol [33]. Membranes (Bio-Gide®, Geistlich Biomaterials, Wolhuser, Switzerland) were used to cover the defects after filling materials. The sheep were sacrificed at 4 weeks and 8 weeks post-operation, and samples with surrounding tissues were retrieved.

Experimental groups Eight groups were established to treat CSBD (n=6 animals per group per time point, Table 1): (i) No graft material; (ii) Autologous bone; (iii) Deproteinized bovine bone (DBB, Bio-Oss®) bearing neither a calcium phosphate coating nor a depot of BMP-2; (iv) DBB bearing neither a calcium phosphate coating nor a depot of BMP-2, but mixed with autologous bone (1:1); (v) DBB bearing no coating but a superficially adsorbed depot of BMP-2; (vi) DBB bearing a calcium phosphate coating but no BMP-2; (vii) DBB bearing a calcium phosphate coating upon which BMP-2 was superficially adsorbed; and (viii) DBB bearing a calcium phosphate coating into which BMP-2 was incorporated.

Autogenous bone was harvested from the cylindrical bone defects and ground to chips under sterile conditions. These bone chips were reserved for groups (ii) and (iv).

Histological procedures Samples with surrounding tissues were fixed chemically and embedded into a block as previously reported [27, 31]. Applying a systematic random sampling strategy [34], the samples were sawn vertically to the long axis into 10-12 slices of 600-μm thickness, 1 mm apart (interval). All slices of each sample were separately mounted on plexiglass holders and polished. The surfaces of the slices were stained with McNeal's

105

Chapter 7

Tetrachrome, basic Fuchsine and Toluidine Blue [27] and examined with a light microscope (Leica).

Table 1 Experimental groups Total DBB Coating BMP-2 Loading Absence Absence Dose of Groups Abbreviation Amount of (−) (−) BMP-2 DBB (per Presence Presence (per sample) (+) (+) sample) No graft material (i) NGM _ _ _ _ (negative control)

Autologous bone (ii) AB _ _ _ _ (positive control)

Deproteinized bovine bone (Bio-Oss®) bearing 0.35g (iii) DBB _ _ _ neither a coating (0.65cm3) nor a depot of BMP-2

DBB mixed with 0.175g (iv) autologous bone DBB+AB (0.325 _ _ _ (1:1) cm3)

DBB bearing no coating but a 0.35g (v) superficially DBB+BMP ads. _ + 35µg (0.65 cm3) adsorbed depot of BMP-2

DBB bearing a 0.35g (vi) coating but no DBB+CaP + _ _ (0.65 cm3) BMP-2

DBB bearing a coating upon DBB+CaP+BMP 0.35g (vii) which BMP-2 was + + 35µg ads. (0.65 cm3) superficially adsorbed

DBB bearing a coating into which 0.35g (viii) DBB+BMP inc. + + 35µg BMP-2 was (0.65 cm3) incorporated

Histomorphometric analysis In addition to a subjective histological description, 10 slices of each sample were used for quantitative histomorphometric analysis. The volume of newly formed bone, bone marrow, DBB and the volume density of multinucleated giant cells (MGC) on DBB were measured using the point counting methodology [35]. During analysis, the evaluator was always blinded for the groups.

106

Deproteinized bovine bone functionalized with the slow delivery of BMP-2

The volume density of MGC was normalized to the volume of DBB. The volume density of MGC (Va) is defined as its volume (Vb) per unit volume of DBB (Vc): Va= Vb/Vc. To evaluate the degradation of DBB granules, the volume of DBB before implantation (time 0, as control) was evaluated by the same histological method. Six samples (0.35g of DBB granules per sample) which had been chemically fixed and embedded in plastic were specifically reserved for this purpose.

Statistical analysis All data are presented as mean values together with the standard deviation (SD). Data were compared using a one way analysis of variance (ANOVA), and post-hoc comparisons were made using Tukey's corrections. The significance level was set at p < .05.

RESULTS

In-vitro investigation In the scanning electron microscope, the biomimetic coating of calcium phosphate with incorporated BMP-2 on DBB displayed a uniform crystalline surface (Fig .1A). The incorporation of BMP-2 did not affect the coating morphology. Fluorescence microscopy revealed that protein (green signal) was homogeneously distributed in the crystalline latticework of the coating (Fig. 1B). The thickness of the coating was 21.2±13.8μm. When the coating has no fluorescently tagged protein, it was not visible in the fluorescence micrograph (Fig. 1C). There was 35µg of BMP-2 incorporated and/or added per sample of 0.35 g of DBB (Table 1).

Figure 1. SEM micrographs of the crystalline calcium phosphate coating with incorporated BMP-2 on deproteinized bovine bone (A). Fluorescence micrographs illustrating the even distribution of a depot of protein (green signal) in the coating (B). The coating without fluorescently tagged protein was not visible in the fluorescence micrograph (C).

Clinical observations All 12 sheep exhibited good health and all the surgical implant sites healed well without any complications in the wound. At the end of the two implantation periods, a total of 96 implants were harvested (48 implants at 4 weeks and 48 implants at 8 weeks). No visual sign of inflammation or adverse tissue reaction was observed.

107

Chapter 7

Descriptive light microscopy Representative histological images of each group at a low magnification are depicted in Fig. 2. In the negative control group (i) without graft material, newly formed bone was found sporadically close to the defect border at 4 and 8 weeks. There was no new bone formation in the center part of bone defect. In group (ii) containing autologous bone, plenty of autologous bone chips were observed at 4 weeks, while at 8 weeks most autologous bone chips had been replaced by newly formed bone which had encapsulated the residual autologous bone chips. Bone formation was irregular in most sections due to the irregular resorption rate of autologous bone. Large regions filled with bone marrow or fibrous tissues were usually observed in the bone defect. In group (iii) which contained only DBB, the bone defect was mainly filled with four components: newly formed bone, bone marrow, fibrous tissues, and DBB granules. In general, DBB granules were distributed uniformly in the bone defect. Bone formation always started from the borders of the bone defect. The newly formed bone was always in close contact with the DBB surface and presented a more mature appearance at 8 weeks than at 4 weeks. At 4 weeks, unmineralized bone was observed but not uniformly between the DBB granules. At 8 weeks, mineralized trabecular structures were observed uniformly within the bone defect and most DBB granules were encapsulated in the bone. The trabecular appearance can be also observed in the other groups with DBB. In group (iv) which contained DBB mixed with autologous bone, newly formed bone was observed at 4 weeks between autologous bone chips and DBB. At 8 weeks, residual autologous bone chips were observed sporadically and encapsulated in the new bone. Group (v), containing DBB with adsorbed BMP-2, group (vi), containing DBB with coating but without BMP-2, and group (vii) containing DBB with coating-adsorbed BMP-2 did not show a significantly histological difference compared with group (iii) containing only DBB at both implantation times. Group (viii), which contains DBB with coating-incorporated BMP-2, showed a different behavior compared to the other groups with DBB. An interconnected bone network, which had a woven appearance, can be easily observed at 4 weeks. At 8 weeks, the bone growth was observed throughout the space between DDB granules in group (viii), and thus formed compact bone areas within the defects. This kind of area was rarely found in other DBB groups (iii-vi). A representative image at higher magnification of the unmineralized bone at 4 weeks from group (v) containing DBB with adsorbed BMP-2 is shown in Fig. 3A. The unmineralized new bone with deep purple can be observed in all the groups at 4 weeks. At this time point, new bone in group (viii) containing DBB with coating-incorporated BMP-2 appeared more mature (Fig. 3B) compared with other DBB groups such as group (v) (Fig. 3A). Moreover, a development stage of bone marrow was observed in groups (ii), (iv) and (viii) containing autograft or the coating-incorporated BMP-2 (Fig. 3B), while bone marrow was rarely found in other DBB groups. Multinucleated giant cells were found on the bone or DBB surface in each group with DBB at both times. A representative multinucleated giant cell enveloping a very small DBB granule at 4 weeks is shown in Fig. 3A. Light demineralized regions of DBB were always observed under these cells (Fig. 3B). With a further resorbing process, the resorption lacunae were clearly created by the multinucleated giant cells (Fig. 4A). However, these cells were found sporadically at 8 weeks because most of the DBB granules had been encapsulated

108

Deproteinized bovine bone functionalized with the slow delivery of BMP-2 in the new bone. The calcium phosphate coating can be found at 4 weeks but not at 8 weeks (Fig. 4B). Representative images at higher magnification showed different appearances of bone tissues in group (viii) containing DBB with coating-incorporated BMP-2 at 8 weeks of implantation (Fig. 5). A compact bone structure with the presence of small porous structure was observed (Fig. 5 A and B). In this compact bone area, bone marrow was observed in some porous structure (Fig. 5C). Also, trabecular-like bone was visible as characterized by the presence of an open porous structure with bone marrow formation (Fig. 5D).

Figure 2. Representative histological sections of bone defect of each group at 4 and 8 weeks after material placement. (i) No graft material; (ii) Autologous bone (#); (iii) Deproteinized bovine bone (DBB, asterisk) bearing neither a coating nor BMP-2; (iv) DBB mixed with autologous bone (1:1); (v) DBB bearing an adsorbed depot of BMP-2; (vi) DBB bearing a calcium phosphate coating but no BMP-2; (vii) DBB bearing a calcium phosphate coating upon which BMP-2 was superficially adsorbed; and (viii) DBB bearing a calcium phosphate coating into which BMP-2 was incorporated. At 4 weeks, the newly formed bone (unmineralized) was purple (black arrow). At 8 weeks, the newly formed bone (mineralized) was reddish (black arrow). The newly formed bone and the autologous bone can be separated in groups (ii and iv). Scale bar = 500 µm.

109

Chapter 7

Figure 3. Representative histological sections at 4 weeks of DBB granules that bore an adsorbed depot of BMP-2 (A) in group (vi) and those that bore a coating-incorporated depot of this agent (B) in group (viii). The newly formed bone was more in contact with the DBB surface and presented more mature in group (viii) than those in group (vi). At this juncture, a development stage of bone marrow was found in group (viii) (M). Multinucleated giant cells presented on the surface of DBB (arrow) and the underlying DBB presented demineralized region (light region). Capsular fibrous tissues (F) were also found around DBB. Scale bar = 100 µm.

Figure 4. Representative histological sections at 4 weeks of multinucleated giant cells (black arrow) and the resorption lacunae on DBB granules that bore an adsorbed depot of BMP-2 from group (v) (A) and the calcium phosphate coating (white arrow) from group (vii) (B). Scale bar = 100 µm.

110

Deproteinized bovine bone functionalized with the slow delivery of BMP-2

Figure 5. Representative histological sections at 8 weeks of DBB granules that bore a coating-incorporated depot of BMP-2 with different bone tissues in group (viii). A stable compact bone (CB) area (A); a compact bone area in an active phase with osteoblasts (arrow, B); Bone marrow in close contact with DBB (C); and trabecular bone appearance (D). Scale bar = 100 µm for (A-C); Scale bar = 200 µm for (D).

Histomorphometry Bone formation Quantitative evaluation of the amount of bone formation 4 and 8 weeks after implantation (Fig. 6) revealed that bone formation increased significantly with increasing time after implantation (p < 0.05). The statistical data revealed that 4 weeks or 8 weeks after implantation group (viii) with coating-incorporated BMP-2 especially had a significant effect on new bone formation among the groups with DBB. At 4 weeks, the volume of newly formed bone was significantly the highest in group (ii) with only autologous bone, and the lowest in group (i) without treatment. Significantly more bone formation was found in group (iv) containing DBB mixed with autologous bone and group (viii) compared with other groups with DBB. No significant difference was found in bone formation between groups (iv) and (viii). Moreover, significantly more bone formation was found in group (v) with directly adsorbed BMP-2 compared with group (iii) with only DBB at 4 weeks. At 8 weeks, the volume of newly formed bone was significantly the highest in groups (ii) and (viii), and significantly the lowest in group (i). There were no significant

111

Chapter 7 differences in bone formation between groups (ii) and (viii) and there were no significant differences among the other five groups with DBB. There was significantly more bone formation in group (iii DBB only) compared with group (i).

Figure 6. Graph depicting the volume of newly formed bone within the bone defect at 4 and 8 weeks after implantation for each of the 8 groups (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05; **p<0.01; ***p<0.001.

Bone marrow Quantitative evaluation of the amount of bone marrow revealed that there were no significant differences among groups (ii), (iv) and (viii) containing autologous bone or the coating-incorporated BMP-2 at 4 weeks, while bone marrow was not found in the other five groups (Fig. 7). At 8 weeks, the volume of bone marrow was significantly the highest in group (ii) with autologous bone, while bone marrow was still not found in group (i) with no graft materials. Significantly more bone marrow was observed in groups (vii) and (viii) with coating-adsorbed or coating-incorporated BMP-2 compared with groups (iii) and (vi) without BMP-2. Furthermore, a trend was observed: the groups containing autologous bone or BMP-2 resulted in more bone marrow than the other groups.

DBB volume Quantitative evaluation of the DBB volume after 4 and 8 weeks of implantation (Fig. 8) revealed that no significances were found in the DBB volume among groups (iii), (v), (vi), (vii), and (viii). As anticipated, the DBB volume of group (iv) containing DBB mixed with autologous bone was about half that compared with other groups. Compared with the start (time 0), no significant decrease of the DBB volume in each group was found.

112

Deproteinized bovine bone functionalized with the slow delivery of BMP-2

Figure 7. Graph depicting the volume of bone marrow within the bone defect at 8 weeks after implantation for each of the 8 groups (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05; **p<0.01; ***p<0.001.

Figure 8. Graph depicting the volume of DBB within the bone defect at 4 and 8 weeks after implantation for each of the DBB-relevant groups (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation.

Volume density of multinucleated giant cells on DBB The multinucleated giant cells at 8 weeks were too few to count. Therefore, only data at 4 weeks were obtained (Fig. 9). At 4 weeks, the volume density of multinucleated giant cells was the highest in DBB that bore an adsorbed depot of BMP-2, and the lowest in

113

Chapter 7 those that bore a coating-incorporated depot of the agent. No significant differences were found among the other 4 DBB groups.

Figure 9. Graph depicting the volume density of multinucleated giant cells on the surface of deproteinized bovine bone at 4 weeks after implantation (see Table 1 for an explanation of the abbreviations). Mean values (n=6 samples per group) are represented together with the standard deviation. *p<0.05; **p<0.01; ***p<0.001.

DISCUSSION

In the current study, the histomorphometric analysis of bone formation 4 and 8 weeks after implantation revealed that the functionalization of DBB with coating-incorporated BMP-2 in group (viii) induced significant bone formation in the treatment of critical-sized bone defects in sheep. This mode of drug delivery was more efficient osteoinductively compared with the adsorption modes (directly adsorbed depot of BMP-2; coating-adsorbed depot of BMP-2). After 8 weeks of implantation, group (viii) exhibited a volume of induced new bone similar to the positive control group (ii) containing autologous bone. Group (viii) also showed significantly more bone formation compared with group (iv) containing DBB mixed with autologous bone. This study provided evidence in support of the functionalization of DBB with the coating-incorporated BMP-2 for optimizing the osteoinductivity for treatment of critical-sized bone defect. The critical-sized bone defect (CSBD) model in this study was created by drilling holes in the humerus and femur of sheep according to a widely published protocol by Nuss et al. [32]. This drill hole model in sheep has proved to be an excellent animal model for testing biomaterials for use in orthopedics, maxillofacial and dental surgery [36]. It allowed the intraosseous implantation of up to 8 different test materials within one animal due to the standardization of the bone defect, while at the same time it can reduce the overall suffering of animals and give the necessary numbers to satisfy statistical requirements [32, 37]. Because of the similarities with humans in weight, bone and joint structure and bone regeneration, the results from sheep are more valid than those obtained from small laboratory animals [38]. Although rodents may be less

114

Deproteinized bovine bone functionalized with the slow delivery of BMP-2 expensive, they have a different bone morphology and they are often are too small for testing bone substitute. Positive results in rodents may have to be repeated and verified in larger species before human clinical trials can be initiated. Our previous study has demonstrated that DBB with coating-incorporated BMP-2 can induce bone formation efficiently in rats subcutaneously [27]. Whilst the ectopic model of ossification is useful for testing the osteoinduction of materials, it is not suitable for its optimization [29]. To improve the osteoinductive effect substantially, BMPs need to be delivered to target sites gradually at a sustained and low level [29, 39]. Biomimetic calcium phosphate coating has been proven as a simple and effective tool for delivering growth factors slowly, such as BMP-2 and vascular endothelial growth factor (VEGF) [23, 40]. In the present study, the histomorphometric analysis of bone formation revealed a coating-incorporated depot of BMP-2 to be more efficient compared with the adsorption mode. This is reflected in the volume of bone that had been deposited by the end of both the sampling times: the value was higher in group (viii) with slow delivery of BMP-2 compared with groups (v) and (vii) with a rapid delivery of BMP-2. Moreover, no significant difference was found between groups (v) and (vii). The value in group (v) with directly adsorbed BMP-2 was significantly higher compared with that in group (iii) with DBB at 4 weeks only, while no significant differences were found between these two groups at 8 weeks. These findings were anticipated and can be readily accounted for by the relatively rapidly release of BMP-2 [41], which was highly water-soluble within a biological milieu, and speedily borne away from DBB, despite the fact that BMP-2 has a strong affinity for DBB [42]. Hence, the osteoinductive effect of the agent is exerted within a short time span. On the contrary, group (viii) with coating-incorporated BMP-2 resulted in a sustained BMP-2 release at a low level [2, 27], and thereby led to more bone formation. The way of BMP-2 release of the coating-incorporated mode includes: 1) controlled low burst release (initial diffusion), 2) release controlled by dissolution based on the solubility of the coating itself, and 3) cell (osteoclast) mediated release based on the digestive activity of cells [12]. It was demonstrated that the degradation of calcium phosphate coating was enhanced by osteoclasts and thus resulted in an elevated protein release, but it was still maintained in a sustained manner in vitro [43]. The efficiently induced bone formation at an ectopic site confirmed in turn the gradual, sustained and cell-mediated release of BMP-2 from the calcium phosphate coating in vivo [44]. Therefore, the lifetime of the coating determined the duration of the protein release. In the present study, the BMP-2-incorporated coating can be observed at 4 weeks but not at 8 weeks, which indicates that its BMP-2 release can last at least 4 weeks. A previous study showed that the coating (thickness: about 20μm) had not been completely degraded during a 5 week period at an ectopic site [31]. The incorporation of BMP-2 could increase the degradation of the coating on titanium implants and it resulted in the complete degradation within 3 weeks in an orthotopic (maxillary) site [29]. The coating may result in a faster degradation in the orthotopic site compared with the ectopic site, because any material placed in the soft tissue may stimulate the formation of a soft tissue capsule in an attempt to wall it off. Even the local cellular mechanisms of degradation may be different in soft and osseous tissue. Therefore, different environments could affect the degradation of the coating. The abundance of bone marrow bodes well for the health and the endurance of the

115

Chapter 7 newly formed bone, since it is an important source of nutriments and pluripotent progenitor cells for osseous tissue [45]. Both osteoblasts and osteoclasts are derived from progenitors that reside in the bone marrow. DBB granules were observed in close contact with bone marrow, which indicates the excellent biocompatibility of DBB. The histomorphometric data revealed a trend indicating that the groups containing autologous bone or BMP-2 resulted in more bone marrow than the rest groups. These findings indicate that autologous bone and BMP-2 may be conducive in developing a healthy bone environment [27], since autologous bone also provides osteogenic cells and growth factors for osteoinduction. Moreover, no significant differences were found in bone marrow among the groups (v, vii, and viii) with slow or rapid delivery of BMP-2 after 8 weeks of implantation, whereas in a previous study the slow delivery of BMP-2 significantly induced more bone marrow than the rapid delivery mode at an ectopic site [27]. In all the groups with DBB, most DBB granules had become an integral part of the bone network 8 weeks after the implantation. DBB can continually serve as a scaffold because it degrades slowly. There is a controversy about the degradation of DBB. It was reported that the resorption of DBB was very minor within 11 years after grafting with no significant changes in the DBB particle size [12]. However, a severe resorption of DBB was observed recently in a porcine calvaria augmentation model under a certain experimental condition [46]. On the basis of current knowledge, the biodegradation of biomaterials is based on the dissolubility of the material itself and the cell (osteoclast) based resorption [47]. It was reported that an average of 4.7% (± 1.61) for DDB was dissolved in Tris-HCl (120 h, pH 7.3, 37°C) [8]. Osteoclasts can be found on the surface of DBB within 2 weeks at an orthotopic site [48]. In the present study, although a certain amount of DBB was observed in a very small size and the resorption lacunae by cells were clearly observed, the quantitative evaluation of the DBB volume in each group supports the notion of the low degradability of DBB. The formation of the capsular fibrous tissue and the bone deposited on the surface of DBB could prevent the dissolution and resorption of DBB. Moreover, it was shown that the coating with incorporated BMP-2 might increase the degradation of underlying materials [26]. The adsorbed mode of BMP-2 was shown to increase the degradation of ceramic [49]. However, in the current study, the different delivery modes of BMP-2 did not significantly influence the degradation of DBB. When DBB was implanted into an extraction wound healing model (not a critical-sized bone defect model), a series of different processes were involved: 1) innate inflammation; 2) formation of granulation tissue and provisional matrix; 3) surface cleaning and resorption; and 4) de novo bone formation and hard tissue integration of the material [48]. In the present study, the release of coating-incorporated BMP-2 from DBB resulted in the lowest volume density of multinucleated giant cells on DBB, whereas the adsorbed BMP-2 led to the highest one. These multinucleated cells could be osteoclasts or foreign body giant cells (FBGCs), both of which can degrade biomaterials. It was reported that the coating-incorporated BMP-2 with slow release can surpress the foreign body reaction [27]. The foreign body reaction producing mcrophages and FBGCs is the end stage of the inflammatory responses following the implantation of biomaterial. FBGCs can release mediators of degradation such as reactive oxygen intermediates (ROIs, oxygen free radicals), degradative enzymes, and acid into this privileged zone

116

Deproteinized bovine bone functionalized with the slow delivery of BMP-2 between the cell membrane and biomaterial [50]. FBGCs can also release inflammatory cytokines which stimulate circulating stem cells to become osteoprogenitors [50, 51]. On the other hand, BMP-2 (endogenous or released from the carrier biomaterial) has been shown to regulate both osteoblasts and osteoclasts [52, 53]. At low doses, BMP-2 stimulates the recruitment, proliferation and differentiation of osteoprogenitor cells, whereas at high doses, it induces the recruitment, formation and activation of osteoclasts. Osteoclast has become the common term to denote any cell that has a unique function to break down mineralized matrices [54]. Osteoclasts exhibited tartrate resistant acid phosphatase (TRAP) positivity and a well defined ruffled border, and they were observed at the surface of both newly formed bone and biomaterials [55]. The precise identification of the osteoclasts and the FBGCs seems to be difficult in vivo, since both can be TRAP-positive [56]. It has also been shown that osteoblasts could be positively stained with TRAP [57]. Therefore, staining with TRAP would not be specific for osteoclast [58]. However, it should be noted that all these multinucleated giant cells play a critical role in the surface treatment and the degradation of material and the bone formation and remodeling. The precise distinction of them in vivo needs to be investigated further. Clinically it is necessary to accelerate bone formation for a curtailment of the recovery phase in the bone defect repair or bone augmentation, since the expectations of surgeons and patients alike are continually rising [59]. The calcium phosphate coating incorporated with growth factors has been widely studied in animal models in an ectopic or an orthotopic site. More studies need to devote to exploring its clinical performance with titanium implants or bone substitutes. Meanwhile, the coating technique has been continually modified and developed. By adjusting the ratio of the coating solution volume to the surface area of the substrate, the coating thickness and the incorporation rate of BMP-2 can be customized for a more precisely controlled release [60]. All in all, we are well on the way to developing a simple and effective tool to optimize the commercial products of bone substitute by giving them osteoinduction, and ultimately to achieve a more satisfactory therapeutic effectiveness.

CONCLUSION

Our findings show the excellent biocompatibility and osteoconductivity of DBB. This material can undergo cell-mediated resorption, but still showed slow degradation. The capacity of BMP-2 to induce and sustain local bone formation in critical-sized bone defects can be influenced by its mode of delivery. The osteogenic response can be more efficiently promoted by its sustained release from a three dimensional reservoir, which is a calcium phosphate coating with incorporated BMP-2, than by its rapid release from an adsorption way. At the same time, the coating-incorporated BMP-2 on DBB led to an excellent therapeutic effect which is comparable with that of autograft. This functionalization approach could greatly enhance the clinical potential of DBB to be an alternative to bone autografts in the repair of large or critical-sized bone defects.

ACKNOWLEDGMENTS

We would like to thank Prof. Dr. Tony Hearn for his scientific input and English editing

117

Chapter 7 as a native speaker for this publication. This project was supported by Osteology grant (2008-015 / Dr. Liu) and KNAW grants (08CDP043, 09CDP036 and 11CDP011).

118

Deproteinized bovine bone functionalized with the slow delivery of BMP-2

REFERENCES

1. Hollinger JO, Kleinschmidt JC. The critical size defect as an experimental model to test bone repair materials. J Craniofac Surg 1990;1:60-8. 2. Liu Y, Wu G, de Groot K. Biomimetic coatings for bone tissue engineering of critical-sized defects. J R Soc Interface 2010;7 Suppl 5:S631-47. 3. Carson JS, Bostrom MP. Synthetic bone scaffolds and fracture repair. Injury 2007;38 Suppl 1:S33-7. 4. Thaller SR, Hoyt J, Dart A, Borjeson K, Tesluk H. Repair of experimental calvarial defects with Bio-Oss particles and collagen sponges in a rabbit model. J Craniofac Surg 1994;5:242-6. 5. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 6. Felix Lanao RP, Leeuwenburgh SC, Wolke JG, Jansen JA. Bone response to fast-degrading, injectable calcium phosphate cements containing PLGA microparticles. Biomaterials 2011;32:8839-47. 7. Karageorgiou V, Kaplan D. Porosity of 3D biomaterial scaffolds and osteogenesis. Biomaterials 2005;26:5474-91. 8. Accorsi-Mendonca T, Conz MB, Barros TC, de Sena LA, Soares Gde A, Granjeiro JM. Physicochemical characterization of two deproteinized bovine xenografts. Braz Oral Res 2008;22:5-10. 9. Simion M, Fontana F, Rasperini G, Maiorana C. Vertical ridge augmentation by expanded-polytetrafluoroethylene membrane and a combination of intraoral autogenous bone graft and deproteinized anorganic bovine bone (Bio Oss). Clin Oral Implants Res 2007;18:620-9. 10. Araujo M, Linder E, Lindhe J. Effect of a xenograft on early bone formation in extraction sockets: an experimental study in dog. Clin Oral Implants Res 2009;20:1-6. 11. Schwartz Z, Weesner T, van Dijk S, Cochran DL, Mellonig JT, Lohmann CH, et al. Ability of deproteinized cancellous bovine bone to induce new bone formation. J Periodontol 2000;71:1258-69. 12. Mordenfeld A, Hallman M, Johansson CB, Albrektsson T. Histological and histomorphometrical analyses of biopsies harvested 11 years after maxillary sinus floor augmentation with deproteinized bovine and autogenous bone. Clin Oral Implants Res 2010;21:961-70. 13. Mooren RE, Dankers AC, Merkx MA, Bronkhorst EM, Jansen JA, Stoelinga PJ. The effect of platelet-rich plasma on early and late bone healing using a mixture of particulate autogenous cancellous bone and Bio-Oss: an experimental study in goats. Int J Oral Maxillofac Surg 2010;39:371-8. 14. Fuerst G, Tangl S, Gruber R, Gahleitner A, Sanroman F, Watzek G. Bone formation following sinus grafting with autogenous bone-derived cells and bovine bone mineral in minipigs: preliminary findings. Clin Oral Implants Res 2004;15:733-40. 15. Gutwald R, Haberstroh J, Kuschnierz J, Kister C, Lysek DA, Maglione M, et al. Mesenchymal stem cells and inorganic bovine bone mineral in sinus augmentation: comparison with augmentation by autologous bone in adult sheep. Br J Oral Maxillofac Surg 2010;48:285-90. 16. Schwarz F, Rothamel D, Herten M, Ferrari D, Sager M, Becker J. Lateral ridge augmentation using particulated or block bone substitutes biocoated with rhGDF-5 and rhBMP-2: an immunohistochemical study in dogs. Clin Oral Implants Res 2008;19:642-52. 17. Schmitt C, Lutz R, Doering H, Lell M, Ratky J, Schlegel KA. Bio-Oss(R) blocks combined with BMP-2 and VEGF for the regeneration of bony defects and vertical augmentation. Clin Oral Implants Res 2011;1-11.

119

Chapter 7

18. Dickerman RD, Reynolds AS, Morgan BC, Tompkins J, Cattorini J, Bennett M. rh-BMP-2 can be used safely in the cervical spine: dose and containment are the keys! Spine J 2007;7:508-09. 19. Barnes B, Boden SD, Louis-Ugbo J, Tomak PR, Park JS, Park MS, et al. Lower dose of rhBMP-2 achieves spine fusion when combined with an osteoconductive bulking agent in non-human primates. Spine (Phila Pa 1976) 2005;30:1127-33. 20. Garrison KR, Donell S, Ryder J, Shemilt I, Mugford M, Harvey I, et al. Clinical effectiveness and cost-effectiveness of bone morphogenetic proteins in the non-healing of fractures and spinal fusion: a systematic review. Health Technol Assess 2007;11:1-150, iii-iv. 21. Glassman SD, Carreon LY, Campbell MJ, Johnson JR, Puno RM, Djurasovic M, et al. The perioperative cost of Infuse bone graft in posterolateral lumbar spine fusion. Spine J 2008;8:443-8. 22. Tuominen T, Jamsa T, Tuukkanen J, Marttinen A, Lindholm TS, Jalovaara P. Bovine bone implant with bovine bone morphogenetic protein in healing a canine ulnar defect. Int Orthop 2001;25:5-8. 23. Liu Y, Huse RO, de Groot K, Buser D, Hunziker EB. Delivery mode and efficacy of BMP-2 in association with implants. J Dent Res 2007;86:84-9. 24. Bodde EW, Boerman OC, Russel FG, Mikos AG, Spauwen PH, Jansen JA. The kinetic and biological activity of different loaded rhBMP-2 calcium phosphate cement implants in rats. J Biomed Mater Res A 2008;87:780-91. 25. Liu Y, Layrolle P, de Bruijn J, van Blitterswijk C, de Groot K. Biomimetic coprecipitation of calcium phosphate and bovine serum albumin on titanium alloy. J Biomed Mater Res A 2001;57:327-35. 26. Wu G, Liu Y, Iizuka T, Hunziker E. The effect of a slow mode of BMP-2 delivery on the inflammatory response provoked by bone-defect-filling polymeric scaffolds. Biomaterials 2010;31:7485-93. 27. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 28. Liu Y, Hunziker EB, Layrolle P, De Bruijn JD, De Groot K. Bone morphogenetic protein 2 incorporated into biomimetic coatings retains its biological activity. Tissue Eng Part A 2004;10:101-8. 29. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106. 30. Liu Y, Hunziker E, Randall N, de Groot K, Layrolle P. Proteins incorporated into biomimetically prepared calcium phosphate coatings modulate their mechanical strength and dissolution rate. Biomaterials 2003;24:65-70. 31. Liu Y, de Groot K, Hunziker EB. BMP-2 liberated from biomimetic implant coatings induces and sustains direct ossification in an ectopic rat model. Bone 2005;36:745-57. 32. Nuss KM, Auer JA, Boos A, von Rechenberg B. An animal model in sheep for biocompatibility testing of biomaterials in cancellous bones. BMC Musculoskelet Disord 2006;7:67. 33. Wang J, Zheng Y, Zhao J, Liu T, Gao L, Gu Z, et al. Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation. J Clin Periodontol 2012;39:98-105. 34. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc 1987;147:229-63. 35. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey. Am J Physiol 1990;258:L148-56. 36. Theiss F, Apelt D, Brand B, Kutter A, Zlinszky K, Bohner M, et al. Biocompatibility and resorption of a brushite calcium phosphate cement. Biomaterials 2005;26:4383-94.

120

Deproteinized bovine bone functionalized with the slow delivery of BMP-2

37. Apelt D, Theiss F, El-Warrak AO, Zlinszky K, Bettschart-Wolfisberger R, Bohner M, et al. In vivo behavior of three different injectable hydraulic calcium phosphate cements. Biomaterials 2004;25:1439-51. 38. Nunamaker DM. Experimental models of fracture repair. Clin Orthop Relat Res 1998;S56-65. 39. Sokolsky-Papkov M, Agashi K, Olaye A, Shakesheff K, Domb AJ. Polymer carriers for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:187-206. 40. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 41. Sachse A, Wagner A, Keller M, Wagner O, Wetzel WD, Layher F, et al. Osteointegration of hydroxyapatite-titanium implants coated with nonglycosylated recombinant human bone morphogenetic protein-2 (BMP-2) in aged sheep. Bone 2005;37:699-710. 42. Hanseler P, Jung UW, Jung RE, Choi KH, Cho KS, Hammerle CH, et al. Analysis of hydrolyzable polyethylene glycol hydrogels and deproteinized bone mineral as delivery systems for glycosylated and non-glycosylated bone morphogenetic protein-2. Acta Biomater 2012;8:116-23. 43. Wernike E, Hofstetter W, Liu Y, Wu G, Sebald HJ, Wismeijer D, et al. Long-term cell-mediated protein release from calcium phosphate ceramics. J Biomed Mater Res A 2010;92:463-74. 44. Hagi TT, Wu G, Liu Y, Hunziker EB. Cell-mediated BMP-2 liberation promotes bone formation in a mechanically unstable implant environment. Bone 2010;46:1322-7. 45. Manolagas SC, Jilka RL. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med 1995;332:305-11. 46. Busenlechner D, Tangl S, Arnhart C, Redl H, Schuh C, Watzek G, et al. Resorption of deproteinized bovine bone mineral in a porcine calvaria augmentation model. Clin Oral Implants Res 2012;23:95-9. 47. Zhang Z, Egana JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al. Cell-based resorption assays for bone graft substitutes. Acta Biomater 2012;8:13-9. 48. Araujo MG, Liljenberg B, Lindhe J. Dynamics of Bio-Oss Collagen incorporation in fresh extraction wounds: an experimental study in the dog. Clin Oral Implants Res 2010;21:55-64. 49. Kessler S, Mayr-Wohlfart U, Ignatius A, Puhl W, Claes L, Gunther KP. The impact of bone morphogenetic protein-2 (BMP-2), vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (b-FGF) on osseointegration, degradation and biomechanical properties of a synthetic bone substitute. Z Orthop Ihre Grenzgeb 2003;141:472-80. 50. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20:86-100. 51. Le Nihouannen D, Saffarzadeh A, Gauthier O, Moreau F, Pilet P, Spaethe R, et al. Bone tissue formation in sheep muscles induced by a biphasic calcium phosphate ceramic and fibrin glue composite. J Mater Sci-Mater M 2008;19:667-75. 52. Pham L, Beyer K, Jensen ED, Rodriguez JS, Davydova J, Yamamoto M, et al. Bone morphogenetic protein 2 signaling in osteoclasts is negatively regulated by the BMP antagonist, twisted gastrulation. J Cell Biochem 2011;112:793-803. 53. Paul S, Lee JC, Yeh LC. A comparative study on BMP-induced osteoclastogenesis and osteoblastogenesis in primary cultures of adult rat bone marrow cells. Growth Factors 2009;27:121-31. 54. Everts V, de Vries TJ, Helfrich MH. Osteoclast heterogeneity: lessons from osteopetrosis and inflammatory conditions. Biochim Biophys Acta 2009;1792:757-65. 55. Basle MF, Chappard D, Grizon F, Filmon R, Delecrin J, Daculsi G, et al. Osteoclastic resorption of Ca-P biomaterials implanted in rabbit bone. Calcif Tissue Int 1993;53:348-56. 56. Kadoya Y, al-Saffar N, Kobayashi A, Revell PA. The expression of osteoclast markers on

121

Chapter 7

foreign body giant cells. Bone Miner 1994;27:85-96. 57. Bianco P, Ballanti P, Bonucci E. Tartrate-resistant acid phosphatase activity in rat osteoblasts and osteocytes. Calcif Tissue Int 1988;43:167-71. 58. Athanasou NA. Cellular biology of bone-resorbing cells. J Bone Joint Surg Am 1996;78:1096-112. 59. Pommer B, Zechner W, Watzak G, Ulm C, Watzek G, Tepper G. Progress and trends in patients' mindset on dental implants. II: implant acceptance, patient-perceived costs and patient satisfaction. Clin Oral Implants Res 2011;22:106-12. 60. Yu X, Qu H, Knecht DA, Wei M. Incorporation of bovine serum albumin into biomimetic coatings on titanium with high loading efficacy and its release behavior. J Mater Sci Mater Med 2009;20:287-94.

122

Chapter 8

Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation

Jingxiao Wang, Yuanna Zheng, Juan Zhao, Tie Liu, Lixia Gao, Zhiyuan Gu, and Gang Wu

Journal of Clinical Periodontology, 2012 Jan;39(1): 98-105.

123

Chapter 8

ABSTRACT

Objectives: To delineate the dynamic micro-architectures of bone induced by low-dose bone morphogenetic protein (BMP)-2/7 heterodimer in peri-implant bone defects compared to BMP2 and BMP7 homodimer. Material and Methods: Peri-implant bone defects (8mm-in-diameter, 4mm-in-depth) were created surrounding SLA-treated titanium implants (3.1mm-in-diameter, 10mm-in-length) in minipig’s calvaria. We administrated collagen sponges with adsorbed low-dose (30ng/mm3) BMP2/7 to treat the defects using BMP2, BMP7 or no BMP as controls. 2, 3, and 6 weeks after implantation, we adopted micro-computer tomography to evaluate the micro-architectures of new bone using the following parameters: relative bone volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), connectivity density (Conn.D), and structure mode index (SMI). Bone implant contact (BIC) was also revealed histologically. Results: Consistent with 2 and 3 weeks, after 6 weeks post-operation, BMP2/7 resulted in significantly higher BV/TV (63.033±2.055%) and significantly lower SMI (-4.405±0.500) than BMP2 (BV/TV: 43.133±2.001%; SMI: -0.086±0.041) and BMP7 (BV/TV: 41.467±1.850%; SMI: -0.044±0.016) respectively. Significant differences were also found in Tb.N, Tb.Th and Tb.Sp at all the time points. At 2 weeks, BMP2/7 resulted in significantly higher BIC than the controls. Conclusions: Low-dose BMP2/7 heterodimer facilitated more rapid bone regeneration in better quality in peri-implant bone defects than BMP2 and BMP7 homodimers.

Keywords: Bone morphogenetic protein, Heterodimer, Bone regeneration, MicroCT, Bone defect

INTRODUCTION

More rapid repair of peri-implant bone defects has been pursued for years in the field of dental implantology. The treatment with autograft is not satisfactory due to its high resorption rate and limited availability [1, 2]. Such a situation has engendered vigorous efforts to develop alternative materials. Albeit so, most of the commercially available bone-defect-filling materials, such as collagen and deproteinized bovine bone, are not intrinsically osteoinductive. Consequently, they have to be premixed with particulate autologous bone to obtain osteogenicity when they are applied to repair the large-volume bone defects. In this process, the limitations of autografts ensue. As a viable option, homodimeric bone morphogenetic proteins (BMPs) can confer osteoinductivity and expedite osteogenesis [3]. Absorbable collagen sponges with adsorbed BMP2 or BMP7 homodimers have been approved by FDA for clinical use [1]. However, the effective doses of BMP homodimers are very high (e.g. 12 milligrams) [4], which results in not only a substantial economic burden, but also a series of potential side-effects, such as the overstimulation of osteoclastic activity [5]. One alternative approach to overcome this dilemma is to adopt more potent BMPs [6]. Heterodimeric BMPs exhibited several- or dozens-fold more effects than the respective homodimers in inducing in-vitro osteoblastogenesis [7]. Most of previous studies were based on the BMP heterodimers through the technology of combined BMP2 and BMP7

124

Low-dose rhBMP2/7 heterodimer gene transfer, which is still far away from clinical applications [8]. However, hitherto, as a promising cytokine therapy, the effects of purified recombinant human BMP heterodimers on in-vivo osteogenesis were merely reported. We hereby designed this experiment using a peri-implant bone defect model in minipigs in order to identify the dynamic changes in micro-architectures of BMP2/7-induced bone. We hypothesized that BMP2/7 could facilitate more rapid bone regeneration in better quality than BMP2 and BMP7. In the field of bone tissue engineering, a suitable carrier is important to optimize osteoinductive effects of BMPs. Although many controlled-release carriers have been well developed recently, collagen sponge is the only FDA-approved BMP-delivery carrier for clinical use [9]. Collagen sponge has been proved to be a acceptable carrier of BMPs in numerous clinical trials [10]. In the present study, we selected collagen sponge as the carrier of BMP2/7 heterodimer with a view to giving a direct relevance and significance for clinical practice. The minimal dose of BMP homodimers to induce bone regeneration in minipig’s calvaria was not documented. A previous study showed that BMP-2 of 30 to 240ng/mm3 could induce bone regeneration in a dose-dependent-increasing manner in the critical-size defects [11]. Consequently, we adopted 5.0ug (equivalent to 30ng/mm3 in bone defects) as test concentration. Although histological analyses provide unique information on cellularity and dynamic indices of bone remodeling, they have limitations in assessing bone micro-architectures. Histological analyses are derived from stereological analysis of a few 2D sections, usually assuming that the underlying structure is plate-like [12]. The inclusion of dental implants may further lessen the histomorphometric information of bone because only one central section of each implant can be used for analysis [13]. In contrast, micro-computed tomography (microCT) can directly measure bone micro-architectures independent of stereological models [14]. In this study, we adopted microCT in order to clarify 1) the dynamic 3D micro-architectures of newly regenerated bone induced by BMP2/7 heterodimer compared to BMP2 and BMP7 homodimers, and 2) the efficacy of low-dose BMP2/7 to repair peri-implant bone defects.

MATERIALS AND METHODS

Preparation of collagen sponges containing BMP2/7, BMP2 or BMP7 According to the manufacturer’s instruction, recombinant human BMP2/7 heterodimer, BMP2 homodimer or BMP7 homodimer (R&D System Inc., Minneapolis, USA) was reconstituted to a final concentration of 0.05μg/μl in a sterile 4mM HCl solution containing 0.1% bovine serum albumin (BSA). The sterile 4mM HCl solution containing 0.1% BSA without BMP was used as control (non-BMP-containing suspension). The collagen sponges (collagen type I, Helistat®, Integra, USA) were adapted into uniform small pieces (15mm×4mm×2.5mm) under sterile condition. 100μl of either BMP-containing or non-BMP-containing suspension was then homogeneously adsorbed onto each collagen sponge piece. The final loading of BMP was 5μg per collagen sponge piece. The sponge pieces were then dried under sterile condition.

125

Chapter 8

Animal models We chose mini-pigs as experimental animals because they bear a comparable rate of bone regeneration to human [15]. The calvaria bone defects were selected by the following considerations: 1) calvaria bone does not depend on central blood supply [16]; 2) calvaria has a more stable mechanical and chemical environment than in mouth, which will significantly increase the successful rate and decrease analysis complexity. Eighteen 9-month-old Guangxi Bama minipigs (9 male, 9 female, weighing from 16.50 to 19.80kg), purchased from and kept in Animal Research Centre of Zhejiang University. Throughout the study, the minipigs were treated following the guidelines of animal care established by Zhejiang University.

Group set-up Four groups were set up (n=6 animals per group per time point): 1) Collagen with BMP2/7 heterodimer (experimental group); 2) Collagen with BMP2 homodimer; 3) Collagen with BMP7 homodimer; 4) Collagen without BMPs. To balance the influence from the gender of animals and defect sites (Fig. 1c, d), the samples were subjected to different defects of different animals following a randomization protocol (Table 1).

Table 1. A randomization protocol for the distribution of collagen sponges with adsorbed BMP2/7, BMP2 or BMP7 or control (no BMP) to the different defects of different animals at one of the three time points.

126

Low-dose rhBMP2/7 heterodimer

Figure 1. Images depicting the schematic model (a, b) and surgical images (c, d) of peri-implant bone defects that were surgically created in minipig’s calvariae and were treated using collagen with adsorbed BMPs. (a) A defect (8mm-in-diameter, 4mm-in-depth) was created using a trephine drill. An implant (8mm-long fixture) was implanted in the center defects with 4mm fixture within the bone defect. (b) The peri-implant bone defect was fulfilled with collagen sponge with or without BMPs, and covered with Bio-Gide® membrane before suture. (c) The borders of four defects (8mm-in-diameter and 4mm-in-depth) were created using trephine drills in the calvarial bone of minipigs. (d) Four implants were centrally implanted into the four defects with interval spaces filling with collagen sponges. From the left upper, the defects were numbered as 1, 2, 3 and 4 respectively in the anticlockwise direction.

Surgeries The pigs were subjected to anesthesia by administrating Sumianxin II (0.3ml/kg, purchased from the Military Veterinary Institute, Quartermaster University of PLA, Chang Chun, China) with the addition of Penicillium (5×104 Unit/kg) and Atropine

127

Chapter 8

(0.03mg/kg) at 30 minutes before surgery. After applying a local anesthesia (1% Lidocaine with 1:100000 Adrenaline) to the frontal calvariae of the minipig, a 10-cm-long sagittal incision was made on the forehead region. The calvarial bone was exposed after lifting a full thickness flap. Four bone defects (8mm-in-diameter, 4mm-in-depth) with a 1cm interval distance were prepared on minipig’s calvaria using a trephine drill. The cylinder-shaped bone tissue in the centre of each circular defect was thereafter removed (Fig. 1a). We adopted the titanium implants (Zhejiang Guangci Medical Appliance co., ltd., Cixi, China) with large-grit sand-blasted, acid-etched (SLA) fixtures (3.1mm-in-diameter and 8mm-in-length) in this experiment. They were implanted in the center of the bone defects with 4mm-long fixture within the defect (Fig. 1a). The total volume of each bone defect after implantation was 166mm3. The remaining circular bone defect around the implant was filled with collagen sponges with or without adsorbed BMPs (5000ng/166mm3≈30ng/mm3) (Fig. 1b). The bone defects and the implants were covered with a piece of Bio-Gide® membrane (40mm×50mm, Geistlich PhamaAG, Switzerland, Fig. 1b). The soft tissue was sutured layer by layer. In addition, Penicillium (50000U/kg) was administered for 3 days postoperatively to protect the minipig from any inflammation. The suture was removed on the 7th day after surgery.

Sample retrievement and preparation At 2, 3 and 6 weeks post-operation, minipigs were sacrificed by intramuscular injection of overdose of Sumianxin II. All the calvarial blocks of the sacrificed animals were harvested and immediately immersed into the 10% neutrally buffered formalin for fixation. After a 7-day fixation, each individual bone sample including the implant and the bone defect around it was separated using a gypsum saw from each calvarial block. The specimens and uninjured bone (obtained in the same location as the defect region) were dehydrated with alcohol and embedded in methyl methacrylate (MMA).

MicroCT evaluation Embedded specimens and uninjured bone were scanned by micro-CT (micro-CT80, ScancoMedical, Bassersdorf, Switzerland) with a resolution of 10µm followed by off-line reconstruction. After image acquisition, the titanium and mineralized tissue were segmented from each other by applying a multilevel thresholding procedure [17, 18]. We evaluated the micro-architectures of bone within the defects using following parameters: 1) Relative bone volume (bone volume/tissue volume, BV/TV, %); 2) Trabecular number (Tb.N, 1/mm); 3) Trabecular thickness (Tb.Th, mm); 4) Trabecular separation (Tb.Sp, mm); 5) Connectivity density (Conn.D, 1/mm3) and 6) Structure mode index (SMI).

Histomorphometric analysis The histomorphometric analysis was subsequently performed to measure bone implant contacts (BIC) in histological sections. Details can be seen in supplementary data.

Statistical analysis All data were presented as mean values together with the standard deviations (Mean±SD).

128

Low-dose rhBMP2/7 heterodimer

Data were analyzed by one-way analysis of variance (ANOVA). SPSS software (version 18 for windows, SPSS Inc., Chicago, IL, USA) was employed for the statistical analysis. Post Hoc comparisons were made using Bonferroni’s corrections with the level of significance at p<0.05.

RESULTS

All implants achieved good primary stability. The healing period was uneventful and the surgical sites healed well during the 6 weeks. Significant increases in BV/TV, Tb.N and Tb.Th as well as significant decreases in Tb.Sp were found for each group at a later time point than at an earlier time point (Table 2), except for Tb.Th in the non-BMP-treated group at 3 weeks and in the BMP2/7 group at 6 weeks. At each time point, significantly higher BV/TV, Tb.N and Tb.Th as well as significant lower Tb.Sp were found in the three BMP-treated groups than that in the non-BMP-treated group. Significantly higher BV/TV, Tb.N and Tb.Th as well as significant lower Tb.Sp were also found in the BMP2/7-treated group than those in the BMP2- and BMP7-treated group. However, no significant difference in BV/TV, Tb.N, Tb.Th and Tb.Sp was detected between the BMP2- and BMP7-treated groups at each time point. BMP2/7-treatment for 6 weeks restored 87.3% BV/TV of the uninjured bone, which was significantly superior to BMP2- (59.8%), BMP7- (57.5%), and non-BMP-treatment (31.1%). Tb.N and Tb.Sp reached the equivalent level to that of the uninjured bone only in the BMP2/7-treated group at 6 weeks. On the two-dimensional images that were perpendicular to the implants, the Tb.N after 6-week BMP2/7-treatment was significantly higher than those after 6-week BMP2-, BMP7- and non-BMP-treatments (Fig. 2). The border between the defects and the surrounding uninjured bone was also less distinct in the BMP2/7-treated groups than those in the other three groups (Fig. 2). For the non-BMP-treated group, significant increases in Tb.Th were only found at 6 weeks. For the BMP2/7-treated group, Tb.Th reached the equivalent level to that of the uninjured bone at 3 weeks and maintained in this level in the monitoring span. For each time point, significantly lower SMI was found in the BMP2/7-treated groups than that in the either BMP2- or BMP7- or non-BMP-treated group (Table 2). No significant differences in SMI were found among BMP2-, BMP7- and non-BMP-treated groups at each time point. At 6 weeks, BMP2/7-treatment resulted in the least difference (2.45) in SMI from the uninjured bone than the treatments of BMP2 (6.76), BMP7 (6.80) and non-BMP (7.47). Conn.D in non-BMP-treated group significantly increased at 3 weeks and thereafter maintained in the equivalent level to that in the uninjured bone (Table 2). Conn.D in the three BMP-treated groups significantly increased at 3 weeks and then significantly decreased at 6 weeks. Conn.D in the three BMP-treated groups was significantly higher than the non-BMP-treated group for each time point. Conn.D in BMP2/7-treated group was significantly higher at 2 weeks but significantly lower at 3 weeks than the BMP2- and BMP7-treated groups. At 6 weeks, Conn.D in BMP2/7-treated group reached the equivalent level to that in the uninjured bone. In contrast, Conn.D in BMP2- and BMP7-treated group was significantly higher than that in the uninjured bone. At as early as 2 weeks post-operation, BMP2/7 resulted in a significantly higher BIC

129

Chapter 8 than the groups of BMP2, BMP7 and non-BMPs. And BIC in BMP2/7 group remained in this level in the following monitoring span. (Details can be seen in supplementary data.)

Figure 2. Graph depicting 2-dimensional microCT images of peri-implant bone defects in the minipigs’ calvaria at 6 weeks post-operation. (a) collagen without BMPs (non-BMP-treated); (b) collagen with BMP2 homodimer; (c) collagen with BMP7 homodimer; (d) collagen with BMP2/7 heterodimer. White arrows pointed to the borders of the bone defects.

130

Low-dose rhBMP2/7 heterodimer

DISCUSSION

Grafts of collagen sponge with adsorbed BMP homodimers, in particular of BMP2 and BMP7, were shown to accelerate bone formation [19]. However, the use of the BMP

131

Chapter 8 homodimers are associated with a high cost [4, 20] and potential side-effects [5]. BMP heterodimers can be a promising approach to solve this dilemma. The study, for the first time, showed that purified recombinant human BMP2/7 heterodimer in a low dose could induce in-vivo bone regeneration more rapidly and in a significantly higher dose-efficiency than rhBMP2 and rhBMP7 homodimers. The quality of newly formed bone is crucial to the long-term stability of implants. It depends on both its volumetric properties such as bone mineral density, and its geometric properties such as bone structure and micro-architecture [21, 22]. In contrast to conventional histological evaluation, microCT is superior to show not only bone mineral density but also bone micro-architectures [21, 23]. Among the six parameters, BV/TV, Tb.N and Tb.Th directly reflected the amount of new bone. This study confirmed that BMP homodimers could significantly promote bone regeneration in such a low dose (30ng/mm3). We also found that the BV/TV in BMP2/7-treated group was significantly higher than that in BMP2- or BMP7-treated group for all time points. This finding indicated that BMP heterodimer induced bone regeneration in a significantly higher dose-efficiency than the homodimers. This specificity of BMP heterodimers was previously attributed to their higher osteoinductive potency than the respective homodimers [24]. Recently, we systematically delineated the functional characteristics of BMP2/7 heterodimer in inducing bone regeneration in a time-course and dose-dependent study [25]. We found that the maximum effect of BMP2/7 heterodimer on promoting in-vitro osteoblastogenesis was not superior to BMP2 or BMP7 homodimers; instead, the effective concentration of BMP2/7 heterodimer for each osteoblastogenetic event was significantly lower than that of the two homodimers. These findings suggested that the advantages of BMP2/7 over BMP2 and BMP7 were significant when they were applied in and probably only in, relatively lower doses. However, this hypothesis needs to be clarified in in-vivo dose-dependent studies. In the present study, the volume density of newly formed bone tissue (BV/TV) induced by BMP2/7 was 1.163- or 1.379-fold of those induced by BMP2 or BMP7 at 3 weeks, and 1.489- or 1.512-fold of those at 6 weeks, which were not as high as previously reported (2-3-fold) [8, 24, 26]. We supposed that this inconsistency might be due to the different animal models and different BMP concentrations, etc. The time-course and BMP-dependent patterns of BV/TV that were obtained from microCT were identical to those of area percentage of bone (APB) that was obtained from histological sections (Supplementary data). The high positive correlation (Pearson coefficient=0.992, p<0.001) between these two parameters validated the reliability of microCT. This finding is consistent with previous studies in animal [27] and human specimens [28]. Albeit so, APB was relatively lower than the corresponding BV/TV. This may be attributable to several factors: inadequate resolution of microCT images relative to the trabecular size and the use of a plate model to estimate trabecular size, etc [28]. After 6 weeks of BMP2/7-treatment, the Tb.N reached the equivalent level to that in the uninjured bone, while either BMP2 or BMP7 or non-BMP treatment failed to do so. Tb.Th in BMP2/7-treated group reached the same level as the uninjured bone at as early as 3 weeks (Fig. 2). In contrast, it was obtained in neither BMP2- nor BMP7- nor non-BMP-treated groups even after 6 weeks. All together, these findings indicated that BMP2/7 heterodimer could more rapidly increase Tb.Th than the respective homodimers.

132

Low-dose rhBMP2/7 heterodimer

The effect of BMP2/7 in increasing Tb.Th was also more rapid than its effect on increasing Tb.N. Consequently, the significant increase of BV/TV in BMP2/7-treated group from 3 to 6 weeks should be attributed to the increase of not Tb.N but to Tb.Th. These findings may add new knowledge to the specificity of BMP2/7 heterodimer’s function. BIC in each group increased gradually. At 6 weeks post-operation, no significant difference could be found among the four groups (Supplementary data). The difference in BIC among different groups was not as significant as other parameters. This may be due to the short vertical depth (4mm) of the bone defects and the high conductivity of the dental implants. Bone tissue could easily grow along the surface of dental implants. In BMP2, BMP7 and non-BMP groups, a thin layer bone could be seen on the surface of dental implants with few contacts at 2 weeks, while the surrounding space remained unfilled with bone. On the other hand, BMP2/7 could significantly enhance the BIC at as early as 2 weeks with surrounding space fulfilled with newly regenerated bone. This finding suggested that BMP2/7 in the selected dose could facilitate significantly earlier functioning of implants than BMP2 and BMP7. Conn.D and SMI reflect the network and structure of trabecular bone tissue [29] and they cannot be obtained from 2-D histological sections. Conn.D in non-BMP-treated group significantly increased 1.6-fold at 3 weeks than at 2 weeks, and thereafter maintained at about 8.2 which was equivalent to that in the uninjured bone till the end of the monitoring span. In contrast, Conn.D increased 3.60-fold for BMP2 and 4.08-fold for BMP7 at 3 weeks than at 2 weeks. Conn.D in the two BMP homodimer-treated groups decreased at 6 weeks to 14.1 which was higher than the uninjured bone. These results indicated BMP2 and BMP7 heterodimers could significantly induce de novo bone regeneration at about 3 weeks post-operation. The Conn.D in the two BMP homodimer-treated groups were about 4.2- to 4.5-fold of that in non-BMP treated group at 3 weeks and reached 34.9-37.4. Such a high Conn.D suggested that the BMP homodimers resulted in a dense stellate-reticulum network of newly formed trabeculae. The thereafter decrease of Conn.D may be due to the re-organization of trabecular by osteoclastic activity [30]. Unexpectedly, although a significant increase was also detected for BMP2/7 heterodimer at 3 weeks, Conn.D in this group was significantly lower than those of BMP2 or BMP7 homodimers. Two possible mechanisms might account for this phenomenon: 1) the peak of BMP2/7-induced de novo bone regeneration appeared earlier and was not detected in the selected time point; 2) BMP2/7 may facilitate a rapid osteoclastic activity to simultaneously re-organizing the trabeculae [30]. Further studies need to be performed to clarify this mechanism. At 6 weeks post-operation, Conn.D in BMP2/7-treated group was significantly lower than those of BMP2 or BMP7 group and it was equivalent to that in the uninjured bone. This finding indicated that BMP2/7 can favor more rapid bone maturation than BMP2 and BMP7 homodimers. At 6 weeks, SMI significantly decreased to -4.4, which was the nearest to that in the uninjured bone (-6.8). In contrast, SMI were at about 0 in the two BMP homodimer-treated groups and was 0.6 in the non-BMP-treated group. Negative values are the result of pores within bones that have high bone volume fractions, which presented a type of Swiss cheese-like structure with a concave surface [21, 31]. In consistent with other parameters, SMI indicated BMP2/7 heterodimer resulted in the more rapid maturation of bone than BMP2 and BMP7 homodimers. This finding was also

133

Chapter 8 confirmed by the indistinct border between the new and the old bone in BMP2/7 group at 6 weeks (Fig. 2). One limitation in this study was that only one concentration of BMPs was adopted in this study. Further studies with serial concentrations could provide more complete delineation of the functional characteristics of BMP heterodimer in inducing in-vivo bone regeneration. Caution should be taken when extrapolating the findings to alveolar bone defects due to the different biological properties between the calvarial bone and alveolar bone. Although we adopted a protocol to enable the homogeneous distribution of BMPs within collagen sponges, the direct adsorption of BMP2 onto collagen sponges can be still clinician-dependent and the potential influence of incubation protocol could, therefore, totally excluded. Release profile of BMPs was also a key determinant of osteoinductive efficacy of BMPs [32]. This study was based on a delivery system of collagen sponge, which may present a different release profile of BMPs and different results from other delivery systems such as biomimetic coatings [33, 34] and polymers [35]. In summary, this study indicated that collagen sponge-delivered rhBMP2/7 heterodimer could repair peri-implant bone defects more rapidly and in a significantly higher dose-efficiency than rhBMP2 and rhBMP7 homodimers when they were applied in the same low dose (30ng/mm3).

ACKNOWLEDGEMENTS

We thank Dr. Yuelian Liu from Academic Center for Dentistry Amsterdam, VU University, the Netherlands for giving technical assistance during the experiment and we also thank Geistlich Phama AG Company for providing Bio-Gide® membrane.

134

Low-dose rhBMP2/7 heterodimer

REFERENCES 1. Gautschi OP, Frey SP, Zellweger R. Bone morphogenetic proteins in clinical applications. Anz J Surg 2007;77:626-31. 2. Carson JS, Bostrom MPG. Synthetic bone scaffolds and fracture repair. Injury 2007;38:S33-S37. 3. Barr T, McNamara AJ, Sandor GK, Clokie CM, Peel SA. Comparison of the osteoinductivity of bioimplants containing recombinant human bone morphogenetic proteins 2 (Infuse) and 7 (OP-1). 2010;109:531-40. 4. Boden SD, Kang J, Sandhu H, Heller JG. Use of recombinant human bone morphogenetic protein-2 to achieve posterolateral lumbar spine fusion in humans: a prospective, randomized clinical pilot trial: 2002 Volvo Award in clinical studies. 2002;27:2662-73. 5. Toth JM, Boden SD, Burkus JK, Badura JM, Peckham SM, McKay WF. Short-term osteoclastic activity induced by locally high concentrations of recombinant human bone morphogenetic protein-2 in a cancellous bone environment. 2009;34:539-50. 6. Zhu W, Kim J, Cheng C, Rawlins BA, Boachie-Adjei O, Crystal RG, et al. Noggin regulation of bone morphogenetic protein (BMP) 2/7 heterodimer activity in vitro. 2006;39:61-71. 7. Aono A, Hazama M, Notoya K, Taketomi S, Yamasaki H, Tsukuda R, et al. Potent ectopic bone-inducing activity of bone morphogenetic protein-4/7 heterodimer. 1995;210:670-7. 8. Zhao M, Zhao Z, Koh JT, Jin TC, Franceschi RT. Combinatorial gene therapy for bone regeneration: Cooperative interactions between adenovirus vectors expressing bone morphogenetic proteins 2, 4, and 7. J Cell Biochem 2005;95:1-16. 9. Haidar ZS, Hamdy RC, Tabrizian M. Delivery of recombinant bone morphogenetic proteins for bone regeneration and repair. Part B: Delivery systems for BMPs in orthopaedic and craniofacial tissue engineering. 2009;31:1825-35. 10. McKay WF, Peckham SM, Badura JM. A comprehensive clinical review of recombinant human bone morphogenetic protein-2 (INFUSE Bone Graft). 2007;31:729-34. 11. Cowan CM, Aghaloo T, Chou YF, Walder B, Zhang XL, Soo C, et al. MicroCT evaluation of three-dimensional mineralization in response to BMP-2 doses in vitro and in critical sized rat calvarial defects. Tissue Eng 2007;13:501-12. 12. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, et al. Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. 1987;2:595-610. 13. Sennerby L, Dasmah A, Larsson B, Iverhed M. Bone tissue responses to surface-modified zirconia implants: A histomorphometric and removal torque study in the rabbit. 2005;7 Suppl 1:S13-20. 14. Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ, Muller R. Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. 2010;25:1468-86. 15. Hönig JH, Merten HA. Das Go¨ ttinger Miniaturschwein (GMS) als Versuchstier in der humanmedizinischen osteologischen Grundlagenforschung. 1993;2:237-41. 16. Wiltfang J, Kloss FR, Kessler P, Nkenke E, Schultze-Mosgau S, Zimmermann R, et al. Effects of platelet-rich plasma on bone healing in combination with autogenous bone and bone substitutes in critical-size defects. An animal experiment. 2004;15:187-93. 17. Gabet Y, Muller R, Levy J, Dimarchi R, Chorev M, Bab I, et al. Parathyroid hormone 1-34 enhances titanium implant anchorage in low-density trabecular bone: a correlative micro-computed tomographic and biomechanical analysis. 2006;39:276-82. 18. Muller R, Ruegsegger P. Micro-tomographic imaging for the nondestructive evaluation of trabecular bone architecture. 1997;40:61-79. 19. Jung RE, Thoma DS, Hammerle CH. Assessment of the potential of growth factors for localized alveolar ridge augmentation: a systematic review. 2008;35:255-81. 20. Govender S, Csimma C, Genant HK, Valentin-Opran A, Amit Y, Arbel R, et al. Recombinant

135

Chapter 8

human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. 2002;84-A:2123-34. 21. Jiang YB, Zhao J, Liao EY, Dai RC, Wu XP, Genant HK. Application of micro-CT assessment of 3-D bone microstructure in preclinical and clinical studies. J Bone Miner Metab 2005;23:122-31. 22. Mosekilde L. Age-Related-Changes in Vertebral Trabecular Bone Architecture - Assessed by a New Method. Bone 1988;9:247-50. 23. Thomsen JS, Laib A, Koller B, Prohaska S, Mosekilde L, Gowin W. Stereological measures of trabecular bone structure: comparison of 3D micro computed tomography with 2D histological sections in human proximal tibial bone biopsies. J Microsc-Oxford 2005;218:171-79. 24. Koh JT, Zhao Z, Wang Z, Lewis IS, Krebsbach PH, Franceschi RT. Combinatorial gene therapy with BMP2/7 enhances cranial bone regeneration. J Dent Res 2008;87:845-49. 25. Zheng YN, Wu G, Zhao J, Wang LH, Sun P, Gu ZY. rhBMP2/7 Heterodimer: An Osteoblastogenesis Inducer of Not Higher Potency but Lower Effective Concentration Compared with rhBMP2 and rhBMP7 Homodimers. Tissue Eng Pt A 2010;16:879-87. 26. Zhu W, Rawlins BA, Boachie-Adjei O, Myers ER, Arimizu J, Choi E, et al. Combined bone morphogenetic protein-2 and -7 gene transfer enhances osteoblastic differentiation and spine fusion in a rodent model. 2004;19:2021-32. 27. Waarsing JH, Day JS, Weinans H. An improved segmentation method for in vivo microCT imaging. 2004;19:1640-50. 28. Chappard D, Retailleau-Gaborit N, Legrand E, Basle MF, Audran M. Comparison insight bone measurements by histomorphometry and microCT. 2005;20:1177-84. 29. Borah B, Gross GJ, Dufresne TE, Smith TS, Cockman MD, Chmielewski PA, et al. Three-dimensional microimaging (MR mu I and mu CT), finite element modeling, and rapid prototyping provide unique insights into bone architecture in osteoporosis. Anat Record 2001;265:101-10. 30. Efeoglu C, Burke JL, Parsons AJ, Aitchison GA, Scotchford C, Rudd C, et al. Analysis of calvarial bone defects in rats using microcomputed tomography: potential for a novel composite material and a new quantitative measurement. Brit J Oral Max Surg 2009;47:616-21. 31. Hildebrand T, Ruegsegger P. Quantification of Bone Microarchitecture with the Structure Model Index. 1997;1:15-23. 32. Haidar ZS, Hamdy RC, Tabrizian M. Delivery of recombinant bone morphogenetic proteins for bone regeneration and repair. Part A: Current challenges in BMP delivery. 2009;31:1817-24. 33. Wu G, Liu Y, Iizuka T, Hunziker EB. Biomimetic coating of organic polymers with a protein-functionalized layer of calcium phosphate: The surface properties of the carrier influence neither the coating characteristics nor the incorporation mechanism or release kinetics of the protein. 2010; Apr 20. [Epub ahead of print]. 34. Wu G, Liu Y, Iizuka T, Hunziker EB. The effect of a slow mode of BMP-2 delivery on the inflammatory response provoked by bone-defect-filling polymeric scaffolds. 2010;31:7485-93. 35. Yu NY, Schindeler A, Peacock L, Mikulec K, Baldock PA, Ruys AJ, et al. In vivo local co-delivery of recombinant human bone morphogenetic protein-7 and pamidronate via poly-D, L-lactic acid. 2010;20:431-41; discussion 41-2.

136

Chapter 9

Icariin: does it have an osteoinductive potential for bone tissue engineering?

Xin Zhang&, Tie Liu&, Yuanliang Huang, Daniel Wismeijer, and Yuelian Liu.

&Xin Zhang and Tie Liu share first authorship. Phytotherapy Research, 2013 Jul 4. doi: 10.1002/ptr.5027.

137

Chapter 9

ABSTRACT

Traditional Chinese Medicines (TCMs) have been recommended for bone regeneration and repair for thousands of years. Currently, the Herba Epimedii and its multi-component formulation are the attractive native herbs for the treatment of osteoporosis. Icariin, a typical flavonol glycoside, is considered to be the main active ingredient of the Herba Epimedii from which icariin has been successfully extracted. Most interestingly, it has been reported that icariin can be delivered locally by biomaterials and that it has an osteoinductive potential for bone tissue engineering. This review focuses on the performance of icariin in bone tissue engineering and on blending the information from icariin with the current knowledge relevant to molecular mechanisms and signal pathways. The osteoinductive potential of icariin could be attributed to its multiple functions in the musculoskeletal system which is involved in the regulation of multiple signaling pathways in anti-osteoporosis, osteogenesis, anti-osteoclastogenesis, chondrogenesis, angiogenesis, and anti-inflammation. The osteoinductive potential and the low price of icariin make it a very attractive candidate as a substitute of osteoinductive protein − bone morphogenetic proteins (BMPs), or as a promoter for enhancing the therapeutic effects of BMPs. However, the effectiveness of the local delivery of icariin needs to be investigated further.

Keywords: icariin, osteoinductive, BMPs, bone regeneration, bone tissue engineering

Abbreviations: ALP, alkaline phosphatase; BMD, bone mineral density; BMP, bone morphogenetic protein; BR, bone resorption; BSP, bone sialoprotein; BMSCs, bone marrow mesenchymal stem cells; Cbfa1, core-binding factor alpha 1; CD14/TLR4, cluster of differentiation 14/toll-like receptor 4; CPC, calcium phosphate cement; EGF-EGFR, epidermal growth factor-epidermal growth factor receptor; ERK, extracellular regulated protein kinases; GAGs, glycosaminoglycans; LPS, lipopolysaccharide; MEK, mitogen-activated protein kinase; OCG, osteoclastogensis; OCN, osteocalcin; OPG, osteoprotegerin; PA, proliferative activity; PGE2, prostaglandin E2; RANKL, receptor activator of nuclear factor-kB ligand; Runx2, runt-related transcription factor 2; Smad, drosophila mothers against decapentaplegic protein; Sox9, SRY (sex determining region Y)-box 9; TBA, trabecular bone area; TCP, tricalcium phosphate

138

Icariin: does it have an osteoinductive potential

Contents 1. Introduction 1.1 Osteoinduction and osteoconduction in bone regeneration 1.2 Clinical and economic backgrounds 1.3 Traditional Chinese Medicine in bone regeneration

2. What is icariin?

3. Icariin-based multi- or single- component formulation 3.1 Xian Ling Gu Bao 3.2 Herba Epimedii

4. Icariin applications in bone tissue engineering

5. Underlying mechanisms of icariin for bone regeneration 5.1 Anti-osteoporosis 5.2 Osteogenesis 5.3 Anti-osteoclastogenesis 5.4 Chondrogenesis 5.5 Angiogenesis 5.6 Anti-inflammation

6. Toxicity of icariin

7. Concluding remarks and perspectives

Acknowledgment References

1. Introduction 1.1. Osteoinduction and osteoconduction in bone regeneration A satisfactory bone regeneration of bone defects which are so large that they cannot heal by themselves remains a big problem for surgeons (Otto and Rao 2004). An ideal osteoinductive bone graft is still desired. The osteoinductive property of such a graft has become the most important issue for bone substitutes. Autografts are the gold standard due to their osteoconductive and osteoinductive properties, while they are unfortunately associated with a limited availability as well as with pain and morbidity at the donor site (Ahlmann et al., 2002). The use of allografts or xenografts can overcome these problems but they are associated with possible infections and immune responses (Bauer and Muschler 2000; Stevenson 1998; Donos et al., 2004). At present, synthetic bone substitutes such as calcium phosphate based biomaterials have become widely used in clinics because of their high osteoconductivity (Dorozhkin 2010), but most of them lack an intrinsic osteoinductivity. Consequently, bone growth factors and mesenchymal stem cells are usually introduced into the system to render these synthetic biomaterials osteoinductive followed by protein or gene delivery (Cowan et al., 2004; Franceschi et al., 2004; Byers et al., 2004; Yamamoto et al., 2000). Osteoinductive growth factors such

139

Chapter 9 as bone morphogenetic proteins (BMPs) have been widely studied in bone tissue engineering (Langer 2009). In particular BMP-2 and BMP-7 were usually carried (delivered) by bone substitutes, and thus confer osteoinduction on the bone substitutes (Liu et al., 2010b; Magin and Delling 2001). They have been applied in clinically. BMP-2 was used to expedite and augment spinal fusion (Shimer et al., 2009), to heal open tibia fractures (Garrison et al. 2010; Alt et al.,2006), and to augment the alveolar bone (Casado et al., 2010; Tonetti and Haemmerle et al., 2008); BMP-7 was used to promote the healing of bony non-unions (Schmidmaier et al., 2009).

1.2. Clinical and economic backgrounds Nowadays the implantation of bone substitutes for bone repair and augmentation is fairly routine clinically and the postoperative healing follows a predictable course in most patients resulting in a good long-term functional outcome (Mordenfeld et al., 2010; Ozkan et al., 2011). However, the expectations of surgeons and patients are continually increasing and aspiring to a shortening of the recovery. (Rustemeyer and Bremerich 2007). Large amounts of BMPs are required in some cases for osteoinduction and for a further improvement in bone formation (Seeherman et al., 2006, Dickerman et al., 2007). The devices containing BMPs tend to fail in a certain percentage of cases, and thereby raise concerns about costs and safety (Geesink et al., 1999; Lieberman et al., 2002; Bridwell et al., 2004). The high price and the rapid degradation of BMPs are its major shortcomings and limit its use clinically. (Urist 1965; Zhao et al., 2006). Therefore, there is an impending need to develop alternative methods to overcome these limitations (Zhao et al., 2008). Attempts have been made to reduce the dose of BMP and so raise the efficiency, such as the use of biomimetic calcium-phosphate coating (Liu et al., 2010b) and polymers mixed with calcium-phosphate cements (Ruhe et al., 2005) which give a sustained release of BMP. Most of these attempts have been effective but remain in a preclinical stage. There remains the need for improving the loading efficiency of BMP to reduce the amount of BMP used. All in all, this research is still on the way to developing a simple, efficient and cost effective method.

1.3. Traditional Chinese Medicine in bone regeneration Traditional Chinese medicines (TCMs) are considered as good alternatives for bone regeneration (Shang et al., 1987). It becomes of great interest to combine bone substitutes with TCMs used for bone regeneration (Zhao et al., 2010). TCMs are divided into single component and multi-component formulations. One multi-component formulation contains many kinds of herbs, whereas one single component contains only one herb. Some of TCMs have been recommended for bone regeneration for hundreds of years (Putnam et al., 2007). A variety of TCMs for bone regeneration have been widely studied (see Table 1). They have shown positive effects on the treatment of osteoporosis, and can stimulate the proliferative activity of osteoblasts, inhibit the formation of osteoclasts, prevent bone loss, and increase the bone mineral density (Zhu et al., 2012; Qin et al., 2005; Xu and Lawson 2004; Lee et al., 2005).

140

Icariin: does it have an osteoinductive potential

Table 1. Multi-component formulations and single components for bone health Multi-component Experimental Product Name formulations (M) or Main effects References subjects Single components (S) Xian ling gu bao M OVX rats; BMD↑ Zhu et al., 2012; Postmenopausal Qin et al., 2005 women Shu di shan zha M Menopausal BR↓ Xu and Lawson women 2004 Shen gu M Osteoporotic BMD↑; BR↓ Mingyue et al., patients 2005 Yang huo gu bao M Osteoporotic BMD↑ Liao et al., 2001 male rats Hachimi-jio-gan M OVX rats BR↓ Hidaka et al., 1997 Kami-kihi-to M OVX rats BMD↑ Kanai et al., 2005 Jian gu M OVX rats BMD↑; TBA↑ Lin et al., 2004 BushenNingxin M Osteoblasts; PA↑ Wang et al., OVX mice 2001 Dang-gui-ji-hwang-yeum M OVX rats TBA↑; OCG↓ Chae et al., 2004 Hochu-ekki-to M Rats BMD↑ Sakamoto et al., 2000 Herba Epimedii S Postmenopausal BR↓; PA↑; mRNA Zhang et al., women; of OPG↑; 2007; UMR-106 cells; RANKL↓; cbfa1 Meng et al., Rat osteoblasts; mRNA↑; OCN↑ 2005a; OVX rats Liu et al., 2005; Qian et al., 2006 Sambucus williamsii S OVX rats; BMD↑; ALP↑; Xie et al., 2005b UMR106 cells OCN↑; OCG↓ Cistanche salsa S OVX rats BR↓ Yamaguchi et al., 1999 Red sage S Osteoclasts OCG↓ Lee et al., 2005 Drynariae rhizoma S Rats and mice; Cathepsins K and Jeong et al., Human L↓ 2004; Jeong et osteoprecursor al., 2005 cells Puerariae radix S Castrated mice BMD↑; TBA↑ Wang et al., 2001 Astragalus S OVX rats BR↓ Kim et al., 2003 membranaceous Abelmoschus manihot (L.) S OVX rats BR↓ Shirwaikar et Medik al., 2003; Puel et al., 2005

Wedelia calendulacea S OVX rats BR↓ Annie et al., Less. 2006 Sophorae fructus S OVX rats BR↓ Joo et al., 2004

Cimicifuga racemosa S OVX rats BR↓ Nisslein and Freudenstein 2003

Among these TCMs, the Herba Epimedii and its multi-component formulation ‘Xian Ling Gu Bao’ (XLGB) have icariin as their main ingredient. Recently, it was reported that icariin is safe, non-toxic, inexpensive and osteoinductive (Wu et al., 2009b; Zhao et al., 2010), and this makes it a very attractive potential agent for bone tissue engineering.

141

Chapter 9

It was demonstrated that icariin enhanced the osteogenic induction activity of BMP-2 in a fibroblastic cell line (Zhao et al., 2010) and induced osteogenic differentiation of preosteoblastic cells (Zhao et al., 2008). After the intramuscular implantation in the backs of rats for three months, new bone formation was observed in β- tricalcium phosphate (TCP) ceramic loaded with icariin but not in the β-TCP ceramic alone (Zhang et al., 2011a). All these studies indicate the highly positive effects of icariin on bone formation. Especially, the ectopic bone formation strongly proved the potential of osteoinduction of icariin (Zhang et al., 2011a). Therefore, it is of great interest that icariin may be used as a substitute for BMP, or as a promoter to enhance the therapeutic effects of BMP and so reduce the dose of BMP. Potentially there is an application for icariin in bone tissue engineering. Hereby, we review the performance of icariin for bone tissue engineering with the current knowledge relevant to molecular mechanisms and signal pathways. The aim of this review is to clarify whether icariin has osteoinductive potential. The publications in the regard of icariin and bone tissue regeneration were selected using following keywords: icariin AND (bone* OR osteoblasts* OR osteoclasts* OR chondrocytes*). Databases were searched from the earliest date available until 1 April 2013. The initial literature search, resulted in 52 articles from PubMed, 104 from ISI, and 2 from Cochrane. After screening all titles and abstracts, 32 articles from PubMed and 41 from ISI were considered to be eligible for this study. The exclusion of 32 duplicates resulted in a total of 41 articles, as shown in Fig. 1. All references in the selected manuscripts were reviewed in order to ensure that no papers had been missed with the chosen search strategy.

2. What is icariin? Icariin (C33H40O15, molecular weight: 676.67) was recorded in the Chinese pharmacopoeia for the purpose of anti-rheumatics (anti-inflammation), tonics (health promotion), and aphrodisiacs (Hsieh et al., 2010). It is a prenyl flavonoid glycoside with a glucosyl group on C-3; a rhamnosyl group on C-7; a methoxyl group on C-4; and a prenyl group on C-8 position (Fig. 2). This prenyl group on C-8 could be the active group that takes part in osteoblastic differentiation and explains its greater potency in osteogenesis and mechanisms of action (Ma et al., 2011). The metabolites profiles in plasma revealed that glucuronide conjugates of isoflavonoids and flavonoid aglycones were the major circulating forms of icariin (Qian et al., 2012). Through the development of modern separation techniques, Icariin has been extracted successfully as a bone active ingredient from Herba Epimedii (Nian et al., 2009, Hsieh et al., 2010). A rapid and accurate reversed-phase liquid chromatography-tandem mass spectrometry method has been developed and validated for the quantitative determination of the flavonoid glycosides in Herba Epimedii (Islam et al., 2008). Icariin can also be extracted and purified by an ultrasonic technique (Zhang et al., 2008a, Jia et al., 2011) and by Dual-Mode HSCCC (Li and Chen 2009).

142

Icariin: does it have an osteoinductive potential

Figure 1. Flow diagram of literature selection process. *E.g. reviews, letters.

Figure 2. Chemical structure of icariin (Ma et al., 2011).

143

Chapter 9

3. Icariin-based multi- or single- component formulation 3.1 Xian Ling Gu Bao Xian Ling Gu Bao (XLGB) is a phytoestrogen-rich multi-component formulation containing Epimedin B, Epimedin C and icariin (Guan et al., 2011). These three flavonoids were all from the Herba Epimedii. XLGB is one such herbal medication officially approved by the Chinese Food and Drug Administration and is orally administered intermittently in the treatment of osteoporosis (Zhu et al., 2012). The ingredients of XLGB consist of six non-leguminous herbs with percentages in weight as follows: Herba Epimedii (70%), Radix Dipsaci (10%), Radix Salviae Miltiorrhizae (5%), Rhizoma Anemarrahenae (5%), Psoralea Corylifolia L. (5%), and Rehmannia Glutinosa (5%) (Guan et al. 2011). Recently, both qualitative and quantitative methods were established for the comprehensive quality control of XLGB. Using high performance liquid chromatography coupled with diode array detection and electrospray ionization tandem mass spectrometry, a total of 47 compounds were identified from XLGB (Guan et al., 2011). XLGB prevented a deterioration of musculoskeletal tissues induced by ovariectomy (OVX). (Qin et al., 2005). The treatment over one year with the conventional dose of XLGB demonstrated a safe and a statistically significant increase in bone mineral density in the lumbar spine after 6 months in postmenopausal women (Zhu et al., 2012).

3.2 Herba Epimedii Icariin is the main pharmacological component of Herba Epimedii. Herba Epimedii is a centuries old traditional medicine herb and its formulation is one of the most frequently prescribed herbs (Pei and Guo 2007). It is recorded in the Chinese pharmacopoeia as ‘yin yang huo’ and was used to cure bone diseases such as osteoporosis and bone fracture in ancient China. Herba Epimedii can be considered as a complementary and alternative medicine for treatment of postmenopausal osteoporosis (Xie et al., 2005a; Zhang et al., 2007). It was shown that the Herba Epimedii can promote the proliferation, the differentiation and the expression of osteoprotegerin (OPG) mRNA of the osteoblasts cultured in vitro (Liu et al., 2006; Meng et al., 2005a). Core binding factor alpha1 (Cbfa1) is a member of the runt family of transcription factors, which appears to play a pivotal role in regulating the differentiation of osteoblastic precursors and the activity of mature osteoblasts. Herba Epimedii could increase the expression of Cbfa1 mRNA in the bone of ovariectomized rats depending on the dose. Furthermore, a high dose of Herba Epimedii of 160 mg/kg administered for 12 weeks in vivo stimulated osteocalcin expression (Qian et al., 2006).

4. Icariin applications in bone tissue engineering The applications of icariin in bone tissue engineering are summarized in Table 2. In order to enhance bone formation for the repair of bone defects, icariin, was loaded into porous beta-tricalcium phosphate ceramic (ICA/beta-TCP) disks (Zhang et al., 2011a). It was revealed that loading icariin in Ica/beta-TCP disks hardly affected the attachment and morphology of rat osteoblast-like (Ros17/28) cells, supporting the proliferation and differentiation of the cells at a higher level than the porous beta-TCP ceramic (beta-PTCP) disks. After intramuscular implantation in the back of rats for three months, no obvious osteogenic evidence was detected in beta-PTCP disks, but new bone formation was

144

Icariin: does it have an osteoinductive potential observed in ICA/beta-TCP disks. These results indeed prove a potential of osteoinductive property of icariin. More and more studies reported the application of icariin combined with calcium phosphate biomaterials. Calcium phosphate cement (CPC) loaded with Icariin filled in the mouse calvarial bone defect induced significant new bone formation and increased bone thickness (Zhao et al., 2010). Obvious blood vessel formation was also observed in the icariin induced new bone in the calvarial bone defect. Moreover, by the thorough mixing of icariin and chitosan/hydroxyapatite (ICA-CS/HA) using a freeze-drying technique, a new bone repair scaffold was generated (Wu et al., 2009b). The results showed that ICA-CS/HA had favorable cell compatibility and promoted osteogenic differentiation of human bone marrow stem cells (hBMSCs). The controlled release of icariin was satisfactory and the release retained after 90 days in vitro. Most interestingly, ICA-CS/HA scaffolds showed favorable osteoconduction and osteoinduction in vivo. They could fill bone segment defects and stimulate new born bone tissues formation at early stage. Recently, another icariin-loaded chitosan/nano-sized hydroxyapatite system was developed, which also controls the release kinetics of icariin to enhance bone repair (Fan et al., 2012). The in vitro bioactivity assay revealed that the loaded icariin was biologically active. Due to the development of the carrier as mentioned above, icariin administered locally can be more efficient for the local bone repair than a systemic administration. For example, the gastrointestine may reduce the therapeutic effect of icariin given orally. Therefore, the use of icariin for bone tissue engineering should concentrate on administration locally rather than systemically.

5. Underlying mechanisms of icariin for bone regeneration 5.1 Anti-osteoporosis Icariin has a definite anti-osteoporotic effect which is similar to estrogen and it is especially effective for the prevention of bone fractures induced by an estrogen deficiency (Nian et al., 2009; Liu et al., 2012). The anabolic effects of icariin in bone possibly result from activating the estrogen receptor in a ligand-independent manner. Research delineates the mechanism by which icariin prevented bone loss after ovariectomy. Icariin suppressed the loss of bone mass and increased the strength in distal femur and the mRNA expression ratio of OPG/RANKL in tibia (Mok et al., 2010). Oral administration of icariin could promote bone formation during mandibular distraction osteogenesis and might be a promising method for shortening the course of distraction osteogenesis (Wei et al., 2011). OVX rats treated orally with icariin could improve the degree of bone mineralization and bone strength and also prevent the suppression of serum calcium phosphorus and 17β-oestradiol (Nian et al., 2009). The oral administration of icariin, limited the metabolism of the medicine due to the gastrointestine. Icariin propylene glycol-liposome suspension (ICA-PG-liposomes) injected intraperitoneally in mice changed the pharmacokinetic behavior (Yang et al., 2012). With improved pharmacokinetics, ICA-PG-liposomes might be developed as promising carriers for icariin injection. Consequently, the use of icariin locally should be considered for future clinical applications.

145

Chapter 9

Table 2. Icariin in bone tissue engineering Carrier for loading Experiments Main effects References icariin In vivo Rabbits BMD↑ Chitosan/hydroxyapatite Wu et al., 2009b

Cartilage formation↑ Cell-hydrogel constructs Li et al., 2012 BMD↑; Volumes of new bone↑; - Wei et al., 2011 TBA↑; Trabecular separation↓

Rats Bone formation↑ Porous β-TCP ceramic Zhang et al., 2011a BMD↑; BR↓; Biomechanical - Xue et al., 2012a; strength↑; Serum estrogen, Xue et al., 2012b; calcium and phosphorus↑; Root Wang et al., 2012; resorption index↓; Collagen↑; Liu et al., 2012; Osteoclast number and activity↓; Bian et al., 2012; OCN↑; OPG/RANKL↑; Cbfa1↑; Nian et al., 2009; Osterix↑ Qin et al., 2008

Mice Bone formation↑; Calcium phosphate Zhao et al., 2010 Bone thickness↑ cement TBA↑; OPG/RANKL↑ - Zheng et al., 2012; Mok et al., 2010

In vitro BMSCs ALP↑; Mineralized nodules↑; Chitosan/nano-size Fan et al., 2012 Proliferation↑ hydroxyapatite

ALP↑; Proliferation↑ Chitosan/hydroxyapatite Wu et al., 2009b

Mineralized nodules↑; - Fan et al., 2011; Proliferation↑; Differeatition↑; Chen et al., 2007b; Calcium deposition↑; ALP↑; Chen et al., 2005; OCN↑; OPN↑; Bone sialoprotein↑; Bian et al., 2012 TGF-β1↑; IGF-I↑; Cbfa1↑; Collagen I↑ Osteoblasts Proliferation↑ PHBV coatings Dai et al., 2011

ALP↑; Runx2↑; BSP↑; OCN↑; Calcium phosphate Zhao et al., 2010 Mineralization↑ cement

Proliferation↑; - Zhang et al., 2011a; Mineralization↑; Zheng et al., 2012; Osteoblast colonies↑; Mok et al., 2010; Cell viability↑; Qin et al., 2008; Calcified nodules↑; Liang et al., 2012; ALP↑; Cbfa1↑; Cao et al., 2012; BSP↑; Zhang et al., 2011b; BMP-2↑; Ma et al., 2011; OPG↑; Hsieh et al., 2011; OPG/RANKL↑; Hsieh et al., 2010; RANKL↑; Zhao et al., 2008; Smad4↑; Zhang et al., 2008b; NO↑; Collagen I↑; Zhang et al., 2008c; OCN↑; Osterix↑; Yin et al., 2007; OPN↑; ERK1/2↓; Xiao et al., 2005; IκBα↓; p38↑ Meng et al., 2005a; Meng et al., 2005b; Huang et al., 2007a; Yang et al., 2013

146

Icariin: does it have an osteoinductive potential

Osteoclasts Apoptosis and cell cycle arrest ↑; - Zhang et al., 2012b; Osteoclastogenesis↓; Pit areas↓; Xue et al., 2012a; Superoxide anion↓; Hsieh et al., 2011; TRAP↓; MMP-9↓; RANKL↓; Huang et al., 2007a; OPG↑; IL-6↓; TNF-α↓; COX-2↓; Huang et al., 2007b; PGE2↓; HIF-1α↓; p38↓; JNK↓ Qin et al., 2008; Chen et al., 2007a

Chondrocytes Aggrecan↑; Sox9↑; Collagen II↑; Cell-hydrogel constructs Li et al., 2012 GAGs↑ Viability↑; Extracellular matrix↑; - Zhang et al., 2012a; NO↑; MMP-1,3,13↓; COX-2↓; Liu et al., 2010a; iNOS↓; Aggrecan↑; Sox9↑; Collagen II↑; GAGs↑

5.2 Osteogenesis The investigation of icariin on rat bone marrow stroma cells revealed an enhancement of the osteogenic differentiation of these cells. A higher concentration of icariin in the extract caused more mineralized bone nodules and higher levels of calcium deposition. The gene expression involved in osteogenesis was also improved, including alkaline phosphatase, bone matrix protein (osteocalcin, osteopontin, bone sialoprotein) and cytokines (TGF-β1 and IGF-I) (Chen et al., 2007b). The effect of icariin on the proliferation of human marrow stroma cells was found to be dependent on the dose and it could also enhance the osteogenic differentiation of these cells in a suitable range of concentrations (Fan et al., 2011). Icariin may strengthen the bone by enhancing the osteogenic differentiation of bone marrow stroma cells, which partially explains the anti-osteoporotic action of the Epimedium herb. When icariin was added to osteoblasts, it promoted the proliferation of human osteoblast and MC3T3-E1 cell lines (Guo et al., 2011; Cao et al., 2012). However, a certain concentration of icariin showed no effect on the proliferation of rat osteoblasts (He et al., 2009). Osteoprotegerin (OPG) plays an essential role in beneficial effects of icariin on bone (Zheng et al., 2012). It was also reported that icariin significantly promoted the expression of type I collagen and osteopontin (OPN) mRNA in rat osteoblasts, and the expression was strengthened gradually with increasing concentration of Icariin (Xiao et al., 2005). Icariin with final concentration of 1 x 10-5 mol/L, which was the best concentration, significantly enhanced the osteogenic differentiation and maturation of rat osteoblasts. It improved significantly the secretion of collagen I, CFU-F(ALP) amounts and mineralized nodules and it also enhanced the mRNA level of Cbfa1 and Osterix (Zhai et al., 2011; Ming et al., 2011). Furthermore, the Cbfa1, BMP2, BMP4 and mRNA were significantly up-regulated after icariin treatment (He et al., 2009). It was suggested that icariin exerts its potent osteogenic effect through the induction of Cbfa1 expression, the production of BMP-4 and the activation of BMP signaling (Zhao et al., 2008). The osteogenic effect was inhibited by the introduction of Smad6 or dominant-negative Cbfa1, as well as Noggin treatment. It was demonstrated that icariin is a bone anabolic agent that may exert its osteogenic effects through the induction of BMP-2 and nitric oxide (NO) synthesis, subsequently regulating Cbfa1/Runx2, OPG, and RANKL gene expressions (Fig. 3) (Hsieh et al., 2010). NO regulates the Cbfa1/Runx2 gene expression, and these effects may contribute to the induction of osteoblasts proliferation and differentiation. Meanwhile BMP-2/Smad suppresses capsase-3 activities

147

Chapter 9 and thus inhibits apoptosis of osteobalsts and hence improves the survival of osteoblasts. In a recent study, icariin up-regulated the expression of BMP-2, Smad4, Cbfa1/Runx2, OPG, RANKL and the OPG/RANKL ratio, indicating that icariin can modulate the process of bone formation via the BMP-2/Smad4 signal transduction pathway in human osteoblastic cell line (Liang et al., 2012).

Figure 3. Molecular mechanism of the anabolic effect of icariin on osteoblasts (Hsieh et al., 2010).

5.3 Anti-osteoclastogenesis Icariin inhibited osteoclastic differentiation in both osteoblast-preosteoclast co-culture and osteoclast progenitor cell culture, and reduced the motility and bone resorption activity of isolated osteoclasts (Huang et al., 2007a). It can be concluded that icariin has the ability to inhibit the formation and bone resorption activity of osteoclasts (Chen et al., 2007a). This in turn, supports the use icariin as an effective component for strengthening bone. In a recent study, icariin decreased osteoclast numbers and activity levels, and increased OPG/RANKL expression ratios, evoking a reparative effect on rapid palatal expansion induced root resorption in rats (Wang et al., 2012). The detail molecular mechanisms of icariin on anti-osteoclastogenesis were further examined (Hsieh et al., 2011). It was demonstrated that a low dose of icariin inhibited LPS-induced osteoclastogenesis without losing cell viability. Icariin can also inhibit LPS-induced pro-inflammatory cytokines synthesis and scavenge LPS-induced RANKL up-regulation and OPG down-regulation. Icariin decreased LPS-mediated prostaglandin E2 (PGE2) production by inhibiting the cyclooxygenase-2 (COX-2) synthesis of osteoblasts and osteoclasts. In osteoclasts, icariin suppressed LPS-mediated activation of the IκB, Jun N-terminal kinase (JNK), extracellular regulated protein kinases (ERK1/2), p38, and

148

Icariin: does it have an osteoinductive potential

Hypoxia-inducible factor 1α (HIF-1α) pathways. While in osteoblasts, only IκB and ERK1/2 pathways were involved. It can be concluded that Icariin inhibited LPS-induced osteoclastogenesis by suppressing the activation of the p38 and JNK pathway (Fig. 4).

Figure 4. Molecular mechanism of icariin on the LPS-induced osteoclastogensis (Hsieh et al., 2011).

5.4 Chondrogenesis Icariin is a safe anabolic agent for chondrogenesis (Liu et al., 2010a). When rabbit chondrocytes isolated from articular cartilage were cultured in vitro with different concentrations of icariin, the higher concentration of icariin produced more extracellular matrix synthesis and expression of chondrogenesis genes of chondrocytes (Zhang et al., 2012a). The effect of icariin on the synthesis of glycosaminoglycans (GAGs) and collagen of chondrocytes, and its potent chondrogenic effect, might be due to its ability to up-regulate the expression of aggrecan, collagen II and Sox9 genes and to down-regulate the expression of the collagen I gene of chondrocytes (Li et al., 2012). It also improves the efficiency of restoring of supercritical-sized osteochondral defects in adult rabbit model, and enhances the integration of newly formed cartilage with subchondral bone (Li et al., 2012). These preliminary studies imply that icariin might be an effective accelerant for chondrogenesis and a substitute for the use of some growth factors. The biomaterials loaded with icariin might have a potential in bone and cartilage tissue engineering. It is known that there are two mechanisms for bone formation. They are intramembranous, which is direct bone formation, and endochondral ossification which is indirect bone formation on a cartilage intermediate. (Einhorn TA, 1998). The potential

149

Chapter 9 different mechanisms between chondrogenic and osteogenic differentiation are associated with two transcription factors, Sox9 and Cbfa1. The transcription factor Sox9 acts during early chondrogenic differentiation (Bi et al., 1999), while Cbfa1 is essential for osteoblast differentiation (Qian et al., 2006). Icariin can up-regulate the expression of Sox9 and Cbfa1 in controlling osteogenesis and chondrogenesis. However, more mechanisms need to be investigated in detail.

5.5 Angiogenesis Vascularization is considered to be a crucial step in bone formation (Wernike et al., 2010). Icariin stimulated in vitro endothelial cell proliferation, migration, and tubulogenesis, as well as increasing in vivo angiogenesis (Chung et al., 2008). It was shown that Icariin has the protective effect on injured vascular endothelial cells, which may be related to its anti-apoptosis effect (Ji et al., 2005, Wang and Huang 2005). Icariin increases the endothelial nitric oxide synthase (eNOS) expression through activating the EGF-EGFR pathway in porcine aorta endothelial cells, by which the endothelial cell function could be regulated (Liu et al., 2011). Moreover, icariin activated the angiogenic signal modulators, ERK, phosphatidylinositol 3-kinase (PI3K), Akt, and eNOS, and increased NO production, without affecting the expression of vascular endothelial growth factor. This indicates that icariin may stimulate angiogenesis directly (Xu and Huang 2007, Chung et al., 2008). Therefore, it should be noted that Icariin stimulated angiogenesis by activating the MEK/ERK- and PI3K/Akt/eNOS-dependent signal pathways and it may also have a potential as a drug in angiogenic therapy (Koizumi et al., 2010, Chung et al. 2008).

5.6 Anti-inflammatory Anti-inflammation plays an important role in bone healing. For example, the treatment of bone defects in peri-implantitis in dentistry particularly needs anti-inflammation (Park 2011). Icariin has displayed its anti-inflammatory potential (Wu et al., 2011). The partial mechanism could be the multiple link intervention on pro-inflammatory cytokines (TNF-α, IL-6), inflammatory mediators (NO) and adhesion molecules (CD11b) (Wu et al., 2009a). Research on the anti-inflammatory effects of icariin on LPS-induced acute inflammatory and its molecular mechanism, suggests that activation of the PI3K/Akt pathway and the inhibition of NF-kappaB are involved in the protective effects of icariin on lipopolysaccharide (LPS)-induced acute inflammatory responses (Xu et al., 2010). Icariin may exert its protective effects through the inhibition of nitric oxide and matrix metalloproteinase (MMP) synthesis, and it may then reduce the destruction of the extracellular matrix (Liu et al., 2010a). Recently, researchers have found an anti-inflammatory property of a novel derivative of icariin, 3, 5, 7-Trihydroxy-4'-methoxy-8-(3-hydroxy-3-methylbutyl)-flavone (ICT) (Wu et al., 2012). It was reported that Icariin and ICT exert anti-inflammatory and anti-tumor effects, and modulate myeloid derived suppressive cell (MDSC) functions (Zhou et al., 2011). The anti-inflammatory effects of ICT were mediated, at least partially, via inhibition of the CD14/TLR4 signaling pathway. ICT reduced NO and PGE2 levels by inhibiting inducible NO synthase and cyclooxygenase-2 protein expression (Wu et al., 2011). This icariin derivative inhibits tumor necrosis factor-alpha (TNF-α) production, inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) mRNA expression, and

150

Icariin: does it have an osteoinductive potential protein expression in LPS stimulated macrophages. Furthermore, ICT suppresses the activation of mitogen-activated protein kinase and inhibits translocation of nuclear factor (NF)-kappaB p65 to the nucleus through decreasing the phosphorylation of IkappaBalpha (Chen et al., 2010). As a result of all these properties, icariin and its derivative can be considered as a potential drug for inflammatory diseases.

6. Toxicity of icariin There was no cytotoxicity toward hBMSCs when the concentration of icariin was smaller than 10−6M, whereas icariin can limit the cell viability when the concentration was larger than 10-5M (Fan et al., 2011). The cytotoxicity test of icariin on MC3T3-E1 cells (a pre-osteoblastic cell line) revealed that the cell viabilities varied from 88% to 98% on both days 1 and 3 after treating with different concentrations of icariin (range from 10-10M to 10-5M) for 72 hours (Zhao et al., 2008). Icariin at a concentration of 5×10-5M strongly inhibited the proliferation of osteoblast-like (Ros17/28) cells (Zhang et al., 2011a). However, many studies demonstrated that icariin with concentration of 10-5M had positive effect on the proliferation of UMR106 cell and human osteoblast (Meng et al., 2005a; Huang et al., 2007a; Yin et al., 2005). Therefore, the optimal concentration of icariin with low cytotoxicity toward osteoblasts was equal to or less than 10-5 M (6.8 μg/ml) (Zhang et al., 2011a). Additionally, more than 90% of murine macrophages (ANA-1) can survive at concentrations up to 80 μg/ml icariin (Li et al., 2011). In general, icariin is safe and non-toxic at low doses (Wu et al., 2009b; Zhao et al., 2010). At doses up to 120 mg/kg in rats given orally administration, icariin has low toxicity, but without overt toxic effects (Luo et al., 2007)

7. Concluding remarks and perspectives It is well known that BMPs induce a sequential cascade of events leading to chondrogenesis, osteogenesis, angiogenesis and the controlled synthesis of extracellular matrix (Kang et al., 2004). BMP-2 and BMP-7 are the most extensively evaluated BMPs with a very high price (Wang et al., 2011). Although BMPs have an outstanding performance in bone formation, they also could result in some cases in negative effects (Kao et al., 2012). Research has developed a variety of methods in bone tissue engineering to reduce the use of BMPs and to improve the osteoinductive effects of BMPs by a slow delivery (Liu et al., 2010b; Ruhe et al., 2005). However, one of the simplest ways could be to search for an effective and low cost substitute for the expensive BMPs. The perspectives discussed herein demonstrate the importance of exploiting an inexpensive osteoinductive drug. More and more researches show that icariin has an osteoinductive potential, due to its properties of inducing osteogenesis, chondrogenesis and angiogenesis. The multiple function of icariin, especially the induction of osteogenesis, is remarkable. The loading of icariin in calcium phosphate biomaterials provides a good alternative to for delivering icariin locally for bone repair, since the calcium phosphate materials have been used as osteoconductive scaffolds. It has been known that the local use of icariin demonstrated positive effects in bone formation at an early stage (Wu et al., 2009b). Several studies have tried to clarify the molecular mechanisms underlying the osteogenic effects. In summary, icariin may exert its osteogenic effects through the induction of BMP-2 and NO synthesis and the BMP-2/Smad4 signal transduction pathway, by up-regulating the

151

Chapter 9 expression of BMP-2, BMP-4, Smad4, Cbfa1/Runx2, OPG, RANKL and the OPG/RANKL ratio (Hsieh et al., 2010; Liang et al., 2012). Icariin can inhibit LPS-induced osteoclastogenesis by suppressing the activation of the p38 and JNK pathway (Hsieh et al., 2011), which in turn contributes to strengthening the bone. The positive effects of icariin on a potent chondrogenic effect might be the up-regulation of the expression of aggrecan, collagen II and Sox9 genes and down-regulation the expression of the collagen I gene of chondrocytes (Zhang et al., 2012a). However, the more detailed osteoinductive mechanisms and the clinical applications of icariin need to be investigated further. Compared with BMPs, icariin is cheaper and has low adverse effects (Zhao et al., 2010; Wu et al., 2009b). The extremely low cost and the high abundance of icariin and its excellent function for bone regeneration make it very appealing for clinical applications. Therefore, it could be candidate for an assistant of BMPs or as a substitution. Nevertheless, the effects of local use of icariin still need to be continually investigated and there is also a need for an appropriate carrier for the most effective delivery. According to the current studies and knowledge, it can be concluded that icariin can be a potential osteoinductive agent. We would like to prove that icariin indeed has a potential for bone tissue engineering. Several projects are running in our lab, both in vitro and in vivo. One of our studies was to use icariin that was incorporated into a biomimetic calcium phosphate bone substitute for the repair of critical-sized bone defects in the rat calvaria. On the whole, the developing techniques give us the confidence to believe that icariin might have a very bright future in bone tissue engineering.

Acknowledgments We would like to thank Prof. Dr. Tony Hearn for his scientific input and English editing as a native speaker for this publication.

Conflict of Interest The authors declare that there are no conflicts of interest.

152

Icariin: does it have an osteoinductive potential

REFERENCES 1. Ahlmann E, Patzakis M, Roidis N, Shepherd L, Holtom P. 2002. Comparison of anterior and posterior iliac crest bone grafts in terms of harvest-site morbidity and functional outcomes. J Bone Joint Surg Am 84: 716-720. 2. Alt V, Bitschnau A, Schnettler R. 2006. Cost-benefit analysis of the use of rhBMP-2 in open tibial fractures. Savings from a health insurer's perspective. Unfallchirurg 109: 463-470 3. Annie S, Prabhu RG, Malini S. 2006. Activity of Wedelia calendulacea Less. in post-menopausal osteoporosis. Phytomedicine: international journal of phytotherapy and phytopharmacology 13: 43-48. 4. Bauer TW, Muschler GF. 2000. Bone graft materials. An overview of the basic science. Clinical Orthopaedics and Related Research 371: 10-27. 5. Bi W, Deng JM, Zhang Z, Behringer RR, De Crombrugghe B.1999. Sox9 is required for cartilage formation. Nature Genetics 22: 85-89. 6. Bian Q, Huang JH, Liu SF, et al. 2012. Different molecular targets of Icariin on bMSCs in CORT and OVX -rats. Frontiers in bioscience (Elite edition) 4: 1224-1236. 7. Bridwell KH, Anderson PA, Boden SD, Vaccaro AR, Zigler JE. 2004. What's new in spine surgery. Journal of Bone and Joint Surgery-American Volume 86: 1587-1594. 8. Byers BA, Guldberg RE, Garcia AJ. 2004. Synergy between genetic and tissue engineering: Runx2 overexpression and in vitro construct development enhance in vivo mineralization. Tissue Engineering 10: 1757-1766. 9. Cao H, Ke Y, Zhang Y, Zhang CJ, Qian W, Zhang GL. 2012. Icariin stimulates MC3T3-E1 cell proliferation and differentiation through up-regulation of bone morphogenetic protein-2. International Journal of Molecular Medicine 29: 435-439. 10. Casado PL, Duarte MEL. Carvalho W, da Silva LE, Barboza EP. 2010. Ridge bone maintenance in human after extraction. Implant Dentistry 19: 314-322. 11. Chae HJ, Yun DH, Chin HY, et al. 2004. Effect of dang-gui-ji-hwang-yeum on osteoporosis in ovariectomized rats. The American journal of Chinese medicine 32: 351-360. 12. Chen KM, Ge BF, Ma HP, Liu XY, Bai MH, Wang Y. 2005. Icariin, a flavonoid from the herb Epimedium enhances the osteogenic differentiation of rat primary bone marrow stromal cells. Pharmazie 60(12): 939-942. 13. Chen KM, Ge BF, Liu XY, et al. 2007a. Icariin inhibits the osteoclast formation induced by RANKL and macrophage-colony stimulating factor in mouse bone marrow culture. Die Pharmazie 62: 388-391. 14. Chen KM, Ma HP, Ge BF, et al. 2007b. Icariin enhances the osteogenic differentiation of bone marrow stromal cells but has no effects on the differentiation of newborn calvarial osteoblasts of rats. Die Pharmazie 62: 785-789. 15. Chen SR, Xu XZ, Wang YH, et al. 2010. Icariin derivative inhibits inflammation through suppression of p38 mitogen-activated protein kinase and nuclear factor-kappaB pathways. Biological and Pharmaceutical Bulletin 33: 1307-1313. 16. Chung BH, Kim JD, Kim CK, et al. 2008. Icariin stimulates angiogenesis by activating the MEK/ERK- and PI3K/Akt/eNOS-dependent signal pathways in human endothelial cells. Biochemical and Biophysical Research Communications 376: 404-408. 17. Cowan CM, Shi YY, Aalami OO, et al. 2004. Adipose-derived adult stromal cells heal critical-size mouse calvarial defects. Nature Biotechnology 22: 560-567. 18. Dai Y, Liu H, Xia L, Zhou Z. 2011. Preparation and characterization of icariin/PHBV drug delivery coatings on anodic oxidized titanium. Transactions of Nonferrous Metals Society of China 21(11): 2448-2453. 19. Dickerman RD, Reynolds AS, Morgan BC, Tompkins J, Cattorini J, Bennett M. 2007. rh-BMP-2 can be used safely in the cervical spine: dose and containment are the keys! Spine Journal 7: 508-509. 20. Donos N, Lang NP, Karoussis IK, Bosshardt D, Tonetti M, Kostopoulos L. 2004. Effect of

153

Chapter 9

GBR in combination with deproteinized bovine bone mineral and/or enamel matrix proteins on the healing of critical-size defects. Clinical Oral Implants Research 15: 101-111. 21. Dorozhkin SV. 2010. Bioceramics of calcium orthophosphates. Biomaterials 31: 1465-1485. 22. Einhorn TA. 1998. The cell and molecular biology of fracture healing. Clinical Orthopaedics and Related Research 355: S7-21. 23. Fan J, Bi L, Wu T, et al. 2012. A combined chitosan/nano-size hydroxyapatite system for the controlled release of icariin. Journal of materials science. Materials in medicine 23: 399-407. 24. Fan JJ, Cao LG, Wu T, et al. 2011. The dose-effect of icariin on the proliferation and osteogenic differentiation of human bone mesenchymal stem cells. Molecules 16: 10123-10133. 25. Franceschi RT, Yang S, Rutherford RB, Krebsbach PH, Zhao M, Wang D. 2004. Gene therapy approaches for bone regeneration. Cells, tissues, organs 176: 95-108. 26. Garrison KR, Shemilt I, Donell S, et al. 2010. Bone morphogenetic protein (BMP) for fracture healing in adults. Cochrane Database of Systematic Reviews 6: 1-51. 27. Geesink RGT, Hoefnagels NHM, Bulstra SK. 1999. Osteogenic activity of OP-1 bone morphogenetic protein (BMP-7) in a human fibular defect. Journal of Bone and Joint Surgery-British Volume 81: 710-718. 28. Guan XY, Li HF, Yang WZ, et al. 2011. HPLC-DAD-MS(n) analysis and HPLC quantitation of chemical constituents in Xian-ling-gu-bao capsules. Journal of Pharmaceutical and Biomedical Analysis 55: 923-933. 29. Guo HL, Zhao YF, Wang X, Xu Y, Zhan HS. 2011. Experimental study on the mechanism of icariin improving human osteoblasts proliferation and the expression of OPG protein. China journal of orthopaedics and traumatology 24: 585-588. 30. He W, Li ZL, Cui YL, et al. 2009. Effect of icariin on the mRNA expressions of Cbfalpha1, BMP2, BMP4 in rat osteoblasts. Beijing Da Xue Xue Bao 41: 669-673. 31. Hidaka S, Okamoto Y, Nakajima K, Suekawa M, Liu SY. 1997. Preventive effects of traditional Chinese (Kampo) medicines on experimental osteoporosis induced by ovariectomy in rats. Calcified Tissue International 61: 239-246. 32. Hsieh TP, Sheu SY, Sun JS, Chen MH. 2011. Icariin inhibits osteoclast differentiation and bone resorption by suppression of MAPKs/NF-kappaB regulated HIF-1alpha and PGE(2) synthesis. Phytomedicine: international journal of phytotherapy and phytopharmacology 18: 176-185. 33. Hsieh TP, Sheu SY, Sun JS, Chen MH, Liu MH. 2010. Icariin isolated from Epimedium pubescens regulates osteoblasts anabolism through BMP-2, SMAD4, and Cbfa1 expression. Phytomedicine: international journal of phytotherapy and phytopharmacology 17: 414-423. 34. Huang J, Yuan L, Wang X, Zhang TL, Wang K. 2007a. Icaritin and its glycosides enhance osteoblastic, but suppress osteoclastic, differentiation and activity in vitro. Life Sciences 81:832-840. 35. Huang J, Zhang J, Zhang TL, Wang K. 2007b. Icariin suppresses bone resorption activity of rabbit osteoclasts in vitro. Chinese Science Bulletin 52(7): 890-895. 36. Islam NM, Yoo HH, Lee MW, et al. 2008. Simultaneous quantitation of five flavonoid glycosides in Herba Epimedii by high-performance liquid chromatography-tandem mass spectrometry. Phytochemical analysis : PCA 19: 71-77. 37. Jeong JC, Lee BT, Yoon CH, Kim HM, Kim CH. 2005. Effects of Drynariae rhizoma on the proliferation of human bone cells and the immunomodulatory activity. Pharmacological research: the official journal of the Italian Pharmacological Society 51: 125-136. 38. Jeong JC, Yoon CH, Jeong CW, Lee YC, Chang YC, Kim CH. 2004. Inhibitory activity of Drynariae rhizoma extracts on cathepsin having bone resorption activity. Immunopharmacology and Immunotoxicology 26: 373-385. 39. Ji RR, Li FY, Zhang XJ. 2005. Effect of icariin on hypoxia induced vascular endothelial cells injury. Chinese journal of integrated traditional and Western medicine 25: 525-530.

154

Icariin: does it have an osteoinductive potential

40. Jia HH, Gao JL, Zeng W, et al. 2011. Study on extraction process for the icariin of hugu capsule with ultrasonic technology. Journal of Chinese medicinal materials 34: 1604-1607. 41. Joo SS, Won TJ, Kang HC, Lee DI. 2004. Isoflavones extracted from Sophorae fructus upregulate IGF-I and TGF-beta and inhibit osteoclastogenesis in rat bone marrow cells. Archives of Pharmacal Research 27: 99-105. 42. Kanai S, Taniguchi N, Higashino H. 2005. Effect of kami-kihi-to (jia-wei-gui-pi-tang) for experimental osteopenia. The American journal of Chinese medicine 33: 41-48. 43. Kang Q, Sun MH, Cheng H, et al. 2004. Characterization of the distinct orthotopic bone-forming activity of 14 BMPs using recombinant adenovirus-mediated gene delivery. Gene Therapy 11: 1312-1320. 44. Kao DW, Kubota A, Nevins M, Fiorellini JP. 2012. The Negative Effect of Combining rhBMP-2 and Bio-Oss on Bone Formation for Maxillary Sinus Augmentation. The International journal of periodontics & restorative dentistry 32: 61-67. 45. Kim C, Ha H, Lee JH, Kim JS, Song K, Park SW. 2003. Herbal extract prevents bone loss in ovariectomized rats. Archives of Pharmacal Research 26: 917-924. 46. Koizumi H, Yu J, Hashimoto R, Ouchi Y, Okabe T. 2010. Involvement of androgen receptor in nitric oxide production induced by icariin in human umbilical vein endothelial cells. FEBS Letters 584: 2440-2444. 47. Langer R. 2009. Perspectives and challenges in tissue engineering and regenerative medicine. Advanced Materials 21: 3235-3236. 48. Lee SY, Choi DY, Woo ER. 2005. Inhibition of osteoclast differentiation by tanshinones from the root of Salvia miltiorrhiza bunge. Archives of Pharmacal Research 28: 909-9013. 49. Li D, Yuan T, Zhang X, Xiao Y, Wang R, Fan Y. 2012. Icariin: a potential promoting compound for cartilage tissue engineering. Osteoarthritis and cartilage / OARS, Osteoarthritis Research Society 20: 1647-1656. 50. Li HB, Chen F. 2009. Separation and Purification of Epimedin A, B, C, and Icariin from the Medicinal Herb Epimedium brevicornum Maxim by Dual-Mode HSCCC. Journal of Chromatographic Science 47: 337-340. 51. Li L, Peng L, Miao J, et al. 2011. Icariin induces the expression of toll-like receptor 9 in ana-1 murine macrophages. Phytotherapy research: PTR 25:1732-1735. 52. Liang W, Lin M, Li X, et al. 2012. Icariin promotes bone formation via the BMP-2/Smad4 signal transduction pathway in the hFOB 1.19 human osteoblastic cell line. International Journal of Molecular Medicine 30: 889-895. 53. Liao S, Kao YH, Hiipakka RA. 2001. Green tea: biochemical and biological basis for health benefits. Vitamins and Hormones 62: 1-94. 54. Lieberman JR, Daluiski A, Einhorn TA. 2002. The role of growth factors in the repair of bone - Biology and clinical applications. Journal of Bone and Joint Surgery-American Volume 84: 1032-1044. 55. Lin YP, Zhou RX, Guo SM. 2004. Effect of jiangu granule on quality of bone in model rats with osteoporosis induced by ovariectomy. Chinese journal of integrated traditional and Western medicine 24: 431-434. 56. Liu M, Zhong C, He RX, Chen LF. 2012. Icariin associated with exercise therapy is an effective treatment for postmenopausal osteoporosis. Chinese Medical Journal 125: 1784-1789. 57. Liu MH, Sun JS, Tsai SW, Sheu SY, Chen MH. 2010a. Icariin protects murine chondrocytes from lipopolysaccharide-induced inflammatory responses and extracellular matrix degradation. Nutrition research 30: 57-65. 58. Liu T, Qin XC, Li WR, et al. 2011. Effects of icariin and icariside II on eNOS expression and NOS activity in porcine aorta endothelial cells. Beijing Da Xue Xue Bao 43: 500-504. 59. Liu Y, Wu G, De Groot K. 2010b. Biomimetic coatings for bone tissue engineering of critical-sized defects. Journal of the Royal Society, Interface / the Royal Society 7: S631-647.

155

Chapter 9

60. Liu Y, Zang H, Zhang H, Chen J. 2005. Effect of Herba Epimedii Flavone on expression of OPG and RANKL in rat osteoblasts. Journal of Chinese medicinal materials 28: 1076-1078. 61. Liu YH, Zhang HY, Zang HM, Cheng JC. 2006. Effect of Herba Epimedii flavone on the osteoblasts metabolism in vitro. China journal of Chinese materia medica 31: 487-490. 62. Luo Y, Nie J, Gong QH, Lu YF, Wu Q, Shi JS. 2007. Protective effects of icariin against learning and memory deficits induced by aluminium in rats. Clinical and experimental pharmacology & physiology 34: 792-795. 63. Ma HP, Ming LG, Ge BF, et al. 2011. Icariin is more potent than genistein in promoting osteoblast differentiation and mineralization in vitro. Journal of Cellular Biochemistry 112: 916-923. 64. Magin MN, Delling G. 2001. Improved lumbar vertebral interbody fusion using rhOP-1: a comparison of autogenous bone graft, bovine hydroxylapatite (Bio-Oss), and BMP-7 (rhOP-1) in sheep. Spine 26: 469-478. 65. Meng FH, Li YB, Xiong ZL, Jiang ZM, Li FM. 2005a. Osteoblastic proliferative activity of Epimedium brevicornum Maxim. Phytomedicine: international journal of phytotherapy and phytopharmacology 12: 189-193. 66. Meng FH, Xiong ZL, Jiang ZM, Li FM. 2005b. Osteoblastic proliferation stimulating activity of Epimedium koreanum Nakai extracts and its flavonol glycosides. Pharmaceutical Biology 43(1): 92-95. 67. Ming L, Chen K, Ge B, Ma H, Zhou J. 2011. Comparative study on effect of icariin and genistein on proliferation and mineralization of osteoblasts in vitro. China journal of Chinese materia medica 36: 2240-2245. 68. Mingyue W, Ling G, Bei X, Junqing C, Peiqing Z, Jie H. 2005. Clinical observation on 96 cases of primary osteoporosis treated with kidney-tonifying and bone-strengthening mixture. Journal of traditional Chinese medicine 25: 132-136. 69. Mok SK, Chen WF, Lai WP, et al. 2010. Icariin protects against bone loss induced by oestrogen deficiency and activates oestrogen receptor-dependent osteoblastic functions in UMR 106 cells. British Journal of Pharmacology 159: 939-949. 70. Mordenfeld A, Hallman M, Johansson CB, Albrektsson T. 2010. Histological and histomorphometrical analyses of biopsies harvested 11 years after maxillary sinus floor augmentation with deproteinized bovine and autogenous bone. Clinical Oral Implants Research 21: 961-970. 71. Nian H, Ma MH, Nian SS, Xu LL. 2009. Antiosteoporotic activity of icariin in ovariectomized rats. Phytomedicine: international journal of phytotherapy and phytopharmacology 16: 320-326. 72. Nisslein T, Freudenstein J. 2003. Effects of an isopropanolic extract of Cimicifuga racemosa on urinary crosslinks and other parameters of bone quality in an ovariectomized rat model of osteoporosis. Journal of Bone and Mineral Metabolism 21: 370-376. 73. Otto WR, Rao J. 2004. Tomorrow's skeleton staff: mesenchymal stem cells and the repair of bone and cartilage. Cell Proliferation 37: 97-110. 74. Ozkan Y, Akoglu B, Kulak-Ozkan Y. 2011. Maxillary sinus floor augmentation using bovine bone grafts with simultaneous implant placement: a 5-year prospective follow-up study. Implant Dentistry 20: 455-459. 75. Park JB. 2011. Treatment of peri-implantitis with deproteinised bovine bone and tetracycline: a case report. Gerodontology 29: 145-149. 76. Pei LK, Guo BL. 2007. A review on research of raw material and cut crude drug of Herba epimedii in last ten years. China journal of Chinese materia medica 32: 466-471. 77. Puel C, Mathey J, Kati-Coulibaly S, et al. 2005. Preventive effect of Abelmoschus manihot (L.) Medik. on bone loss in the ovariectomised rats. Journal of Ethnopharmacology 99: 55-60. 78. Putnam SE, Scutt AM, Bicknell K, Priestley CM, Williamson EM. 2007. Natural products as

156

Icariin: does it have an osteoinductive potential

alternative treatments for metabolic bone disorders and for maintenance of bone health. Phytotherapy Research 21: 99-112. 79. Qian G, Zhang X, Lu L, Wu X, Li S, Meng J. 2006. Regulation of Cbfa1 expression by total flavonoids of Herba epimedii. Endocrine Journal 53: 87-94. 80. Qian Q, Li SL, Sun E, et al. 2012. Metabolite profiles of icariin in rat plasma by ultra-fast liquid chromatography coupled to triple-quadrupole/time-of-flight mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis 66: 392-398. 81. Qin L, Han T, Zhang Q, et al. 2008. Antiosteoporotic chemical constituents from Er-Xian Decoction, a traditional Chinese herbal formula. Journal of Ethnopharmacology 118(2): 271-279. 82. Qin L, Zhang G, Hung WY, et al. 2005. Phytoestrogen-rich herb formula "XLGB" prevents OVX-induced deterioration of musculoskeletal tissues at the hip in old rats. Journal of Bone and Mineral Metabolism 23: 55-61. 83. Ruhe PQ, Boerman OC, Russel FG, Spauwen PH, Mikos AG, Jansen JA. 2005. Controlled release of rhBMP-2 loaded poly(dl-lactic-co-glycolic acid)/calcium phosphate cement composites in vivo. Journal of controlled release: official journal of the Controlled Release Society 106: 162-171. 84. Rustemeyer J, Bremerich A. 2007. Patients ' knowledge and expectations regarding dental implants: assessment by questionnaire. International Journal of Oral and Maxillofacial Surgery 36: 814-817. 85. Sakamoto S, Sassa S, Kudo H, Suzuki S, Mitamura T, Shinoda H. 2000. Preventive effects of a herbal medicine on bone loss in rats treated with a GnRH agonist. European Journal of Endocrinology / European Federation of Endocrine Societies 143: 139-142. 86. Schmidmaier G, Capanna R, Wildemann B, Beque T, Lowenberg D. 2009. Bone morphogenetic proteins in critical-size bone defects: what are the options? Injury 40: S39-43. 87. Seeherman H, Azari K, Bidic S, et al. 2006. rhBMP-2 delivered in a calcium phosphate cement accelerates bridging of critical-sized defects in rabbit radii. The Journal of bone and joint surgery. American volume 88: 1553-1565. 88. Shang TY, Gu YW, Dong FH. 1987. Treatment of forearm bone fractures by an integrated method of traditional Chinese and Western medicine. Clinical Orthopaedics and Related Research 215: 56-64. 89. Shimer AL, Oner FC, Vaccaro AR. 2009. Spinal reconstruction and bone morphogenetic proteins: open questions. Injury 40: S32-38. 90. Shirwaikar A, Khan S, Malini S. 2003. Antiosteoporotic effect of ethanol extract of Cissus quadrangularis Linn. on ovariectomized rat. Journal of Ethnopharmacology 89: 245-250. 91. Stevenson S. 1998. Enhancement of fracture healing with autogenous and allogeneic bone grafts. Clin Orthop Relat Res 355: S239-246. 92. Tonetti MS and Haemmerle CH. 2008. Advances in bone augmentation to enable dental implant placement: Consensus Report of the Sixth European Workshop on Periodontology. Journal of Clinical Periodontology 35: 168-172. 93. Urist MR. 1965. Bone - Formation by Autoinduction. Science 150: 893-899. 94. Wang D, Li F, Jiang Z. 2001. Osteoblastic proliferation stimulating activity of Psoralea corylifolia extracts and two of its flavonoids. Planta Medica 67: 748-749. 95. Wang F, Liu Z, Lin S, Lu H, Xu J. 2012. Icariin enhances the healing of rapid palatal expansion induced root resorption in rats. Phytomedicine: international journal of phytotherapy and phytopharmacology 19: 1035-1041. 96. Wang J, Zheng Y, Zhao J, et al. 2011. Low-dose rhBMP2/7 heterodimer to reconstruct peri-implant bone defects: a micro-CT evaluation. Journal of clinical periodontology 39: 98-105. 97. Wang YK, Huang ZQ. 2005. Protective effects of icariin on human umbilical vein endothelial cell injury induced by H2O2 in vitro. Pharmacological research: the official journal of the

157

Chapter 9

Italian Pharmacological Society 52: 174-182. 98. Wei H, Zili L, Yuanlu C, et al. 2011. Effect of icariin on bone formation during distraction osteogenesis in the rabbit mandible. International Journal of Oral and Maxillofacial Surgery 40: 413-418. 99. Wernike E, Montjovent MO, Liu Y, et al. 2010. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. European Cells & Materials 19: 30-40. 100. Wu J, Du J, Xu C, et al. 2011. In vivo and in vitro anti-inflammatory effects of a novel derivative of icariin. Immunopharmacology and Immunotoxicology 33: 49-54. 101. Wu J, Zhou J, Chen X, et al. 2012. Attenuation of LPS-induced inflammation by ICT, a derivate of icariin, via inhibition of the CD14/TLR4 signaling pathway in human monocytes. International immunopharmacology 12: 74-79. 102. Wu JF, Dong JC, Xu CQ. 2009a. Effects of icariin on inflammation model stimulated by lipopolysaccharide in vitro and in vivo. Chinese journal of integrated traditional and Western medicine 29: 330-334. 103. Wu T, Nan KH, Chen JD, et al. 2009b. A new bone repair scaffold combined with chitosan/hydroxyapatite and sustained releasing icariin. Chinese Science Bulletin 54: 2953-2961. 104. Xiao Q, Chen A, Guo F. 2005. Effects of Icariin on expression of OPN mRNA and type I collagen in rat osteoblasts in vitro. Journal of Huazhong University of Science and Technology. Medical sciences 25: 690-692. 105. Xie F, Wu CF, Lai WP, et al. 2005a. The osteoprotective effect of Herba epimedii (HEP) extract in vivo and in vitro. Evidence-based complementary and alternative medicine: eCAM 2: 353-361. 106. Xie F, Wu CF, Zhang Y, et al. 2005b. Increase in bone mass and bone strength by Sambucus williamsii HANCE in ovariectomized rats. Biological and Pharmaceutical Bulletin 28: 1879-1885. 107. Xu CQ, Liu BJ, Wu JF, et al. 2010. Icariin attenuates LPS-induced acute inflammatory responses: involvement of PI3K/Akt and NF-kappaB signaling pathway. European Journal of Pharmacology 642: 146-153. 108. Xu H, Lawson D. 2004. Effects of Chinese herbal medicine on bone structure and function. Journal of traditional Chinese medicine 24: 224-232. 109. Xu HB, Huang ZQ. 2007. Icariin enhances endothelial nitric-oxide synthase expression on human endothelial cells in vitro. Vascular pharmacology 47: 18-24. 110. Xue L, Jiao L, Wang Y, et al. 2012a. Effects and interaction of icariin, curculigoside, and berberine in er-xian decoction, a traditional chinese medicinal formula, on osteoclastic bone resorption. Evidence-Based Complementary and Alternative Medicine: doi:10.1155/2012/490843. 111. Xue L, Wang Y, Jiang Y, et al. 2012b. Comparative effects of er-xian decoction, epimedium herbs, and icariin with estrogen on bone and reproductive tissue in ovariectomized rats. Evidence-Based Complementary and Alternative Medicine: doi: 10.1155/2012/241416. 112. Yamaguchi K, Shinohara C, Kojima S, Sodeoka M, Tsuji T. 1999. (2E,6R)-8-hydroxy-2,6-dimethyl-2-octenoic acid, a novel anti-osteoporotic monoterpene, isolated from Cistanche salsa. Bioscience, Biotechnology, and Biochemistry 63: 731-735. 113. Yamamoto M, Tabata Y, Hong L, Miyamoto S, Hashimoto N, Ikada Y. 2000. Bone regeneration by transforming growth factor beta 1 released from a biodegradable hydrogel. Journal of Controlled Release 64: 133-142. 114. Yang L, Lu D, Guo J, Meng X, Zhang G, Wang F. 2013. Icariin from Epimedium brevicornum Maxim promotes the biosynthesis of estrogen by aromatase (CYP19). Journal of Ethnopharmacology 145(3): 715-721. 115. Yang W, Yu XC, Chen XY, Zhang L, Lu CT, Zhao YZ. 2012. Pharmacokinetics and tissue

158

Icariin: does it have an osteoinductive potential

distribution profile of icariin propylene glycol-liposome intraperitoneal injection in mice. The Journal of pharmacy and pharmacology 64: 190-198. 116. Yin XX, Chen ZQ, Dang GT, Ma QJ, Liu ZJ. 2005. Effects of Epimedium pubescens epimedium on proliferation and differentiation of human osteoblasts. China J Chin Mat Med (in Chinese) 30: 289-291. 117. Yin XX, Chen ZQ, Liu ZJ, Ma QJ, Dang GT. 2007. Icariine stimulates proliferation and differentiation of human osteoblasts by increasing production of bone morphogenetic protein 2. Chinese Medical Journal 120(3): 204-210. 118. Zhai YK, Li ZZ, Chen KM, Zhang N, Cheng GZ, Zhu RQ. 2011. Effects of icariin on the proliferation, differentiation and maturation of rat calvarial osteoblasts in vitro. Journal of Chinese medicinal materials 34: 917-922. 119. Zhang D, Zhang J, Fong C, Yao X, Yang M. 2012b. Herba epimedii flavonoids suppress osteoclastic differentiation and bone resorption by inducing G2/M arrest and apoptosis. Biochimie 94(12): 2514-2522. 120. Zhang DW, Cheng Y, Wang NL, Zhang JC, Yang MS, Yao XS. 2008b. Effects of total flavonoids and flavonol glycosides from Epimedium koreanum Nakai on the proliferation and differentiation of primary osteoblasts. Phytomedicine: international journal of phytotherapy and phytopharmacology 15: 55-61. 121. Zhang DW, Cheng Y, Zhang JC, et al. 2008c. Synergistic effect of trace elements and flavonoids from Epimedium koreanum Nakai on primary osteoblasts. Chinese Science Bulletin 53(3): 347-356. 122. Zhang G, Qin L, Shi Y. 2007. Epimedium-derived phytoestrogen flavonoids exert beneficial effect on preventing bone loss in late postmenopausal women: a 24-month randomized, double-blind and placebo-controlled trial. Journal of bone and mineral research: the official journal of the American Society for Bone and Mineral Research 22: 1072-1079. 123. Zhang HF, Yang TS, Li ZZ, Wang Y. 2008a. Simultaneous extraction of epimedin A, B, C and icariin from Herba Epimedii by ultrasonic technique. Ultrasonics Sonochemistry 15: 376-385. 124. Zhang J, Li Y, Sun J, Liu C, Zhang D. 2011b. Synergistic or Antagonistic Effect of MTE plus TF or Icariin from Epimedium koreanum on the Proliferation and Differentiation of Primary Osteoblasts In Vitro. Biological Trace Element Research 143(3): 1746-1757. 125. Zhang L, Zhang X, Li KF, et al. 2012a. Icariin promotes extracellular matrix synthesis and gene expression of chondrocytes in vitro. Phytotherapy research: PTR 26: 1385-1392. 126. Zhang X, Guo Y, Li DX, et al. 2011a. The effect of loading icariin on biocompatibility and bioactivity of porous beta-TCP ceramic. Journal of materials science. Materials in medicine 22: 371-379. 127. Zhao BH, Katagiri T, Toyoda H, et al. 2006. Heparin potentiates the in vivo ectopic bone formation induced by bone morphogenetic protein-2. Journal of Biological Chemistry 281: 23246-23253. 128. Zhao J, Ohba S, Komiyama Y, Shinkai M, Chung UI, Nagamune T. 2010. Icariin: a potential osteoinductive compound for bone tissue engineering. Tissue engineering. Part A 16: 233-243. 129. Zhao J, Ohba S, Shinkai M, Chung UI, Nagamune T. 2008. Icariin induces osteogenic differentiation in vitro in a BMP- and Runx2-dependent manner. Biochemical and Biophysical Research Communications 369: 444-448. 130. Zheng D, Peng S, Yang SH, et al. 2012. The beneficial effect of Icariin on bone is diminished in osteoprotegerin-deficient mice. Bone 51: 85-92. 131. Zhou J, Wu J, Chen X, et al. 2011. Icariin and its derivative, ICT, exert anti-inflammatory, anti-tumor effects, and modulate myeloid derived suppressive cells (MDSCs) functions. International immunopharmacology 11: 890-898. 132. Zhu HM, Qin L, Garnero P, et al. 2012. The first multicenter and randomized clinical trial of herbal Fufang for treatment of postmenopausal osteoporosis. Osteoporosis international: a

159

Chapter 9

journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 23: 1317-1327.

160

Chapter 10

General discussion

161

Chapter 10

GENERAL DISCUSSION

The preparation procedure of biomimetic calcium phosphate material In this thesis (Chapters 2-4), a biomimetic calcium phosphate material (BioCaP) was developed as a biodegradable bone substitute which is based on the protocol of biomimetic calcium phosphate coating. This is a breakthrough in modifying the biomimetic coating approach. BioCaP is prepared in a biomimetic environment which can retain the bioactivity of growth factors [1, 2]. This biomimetic environment [(200 mM HCl, 20 mM CaCl2·2H2O, 680 mM NaCl, and 10 mM Na2HPO4) buffered by TRIS (250 mM) to a pH of 7.4] is the key point in the preparation of BioCaP. It was found that BioCaP has two phases of precipitations with different crystalline morphologies during the preparation in Chapter 3. It was shown that BioCaP is a compound including different crystalline structures of calcium phosphate. The biomimetic preparation methods resulted in the bone-like mechanical strength (enough hardness) of BioCaP, whereas other calcium phosphate biomaterials need sintering at high temperature to achieve enough hardness [3]. More importantly, the use of BioCaP can simply and slowly deliver proteins/drug without requiring other materials such as polymers and chitosan. During the preparation, the sterility of biomedical materials is very important for clinical trials. Accordingly, we took strict measures to guarantee the sterility of our biomaterials. The whole processing of this material, including vacuum filtering, was carried in a biological safety cabinet. For sterile vacuum filtering, we used the sterile “Bottle Top Filter”. Therefore, the whole process was performed under sterile condition. In our cell experiments we never encountered the presence of bacteria in the culture media. Consequently, our in vitro as well as in vivo results in Chapter 2-4 indicated the sterility of BioCaP materials.

The characteristics of BioCaP In BioCaP, protein and calcium phosphate were precipitated together to form BioCaP granules in which a depot of protein was incorporated in the center of the granules as an internal depot. Next, protein and calcium phosphate were co-precipitated onto the surface of these granules, thus creating a surface coated depot. This dual system provides an ideal vehicle for the delivery of different protein/drugs in two phases, an initial slow delivery phase (surface coated depot) and a delayed phase (internal depot). This thesis evaluated the BioCaP with incorporated bone morphogenetic protein-2 (BMP-2). The two delivery modes of BMP-2 were studied separately in Chapters 2-4. The simultaneous delivery of two or multiple growth factors in vivo is needed to be investigated further in the future. Since bone regeneration is a coordinated cascade of events regulated by several growth factors, the local sequential delivery of vascular endothelial growth factor (VEGF) and BMP-2 could enhance bone formation compared with BMP-2 alone [4]. Therefore, it will be interesting to study the dual release of BMP-2 and VEGF from BioCaP.

162

General discussion

The work in this thesis brought us the following highlights of BioCaP:

 Enough hardness / mechanical stability / easy of handling  Good biodegradability  A slow release system for the delivery of single or multiple proteins/drugs  Osteoconductivity  Osteoinductive potential  Micro-porosity

These characteristics make BioCaP an attractive or idea bone substitute. However, the limitation is that it has no macro-porosity which can enable cell ingrowth into the composite. The use of granular form of BioCaP may overcome this disadvantage, since the space between the granules could be equal to the macro-porosity.

Osteoinductivity of BioCaP The osteoinductivity of BioCaP was mainly conferred by using osteogenic growth factors such as bone morphogenetic protein-2 (BMP-2). The findings in Chapters 2-4 imply that the in vivo BMP-2 release from BioCaP may be in a slow release pattern, since the adsorption way always has a burst release which results in poor osteoinduction [5, 6]. However, the in vivo BMP-2 release kinetic needs to be investigated further. In previous studies, growth factors were co-precipitated into the latticework of crystalline calcium phosphate coating and then were shown to be released locally in a slow manner [2, 5, 7]. The slow release of BMP-2 can enhance the osteoinductivity of BioCaP [1]. In Chapter 4, we implanted BioCaP with the two delivery modes of BMP-2 into the bone defects in sheep. The findings demonstrated that the therapeutic effect of this material was excellent, which was significantly better than the deproteinized bovine bone (DBB, Bio-Oss®) and was comparable with the autologous bone. All the finding indicated that osteoinducive BioCaP by carrying BMP-2 can be a good alternative to autologous bone (gold standard).

Osteoinductive promoter Biomimetic BMP2-coprecipitated calcium phosphate (BMP2-cop.BioCaP) particles were developed and evaluated in Chapters 4 and 5. It was shown that this material serves as a biodegradable and efficient “osteoinducer” for DBB which need effective osteoinduction [8, 9]. The directly mixing BMP2-cop.BioCaP with other bone substitutes is one of the simplest operations for clinicians. However, this calcium phosphate material may not be used as an independent bone substitute, since its mechanical property still needs to be improved and the preparation time also needs to be shortened.

Multinucleated giant cells The cells involved in the degradation of calcium phosphate materials are mainly multinucleated giant cells (osteoclasts or foreign body giant cells), macrophages, and monocytes [7, 10]. It is important to understand how and to what extent these cells might influence the protein release from a drug-delivery system [11, 12]. It has been demonstrated that the monocytes/macrophages has no significant effects on the protein release from calcium phosphate ceramics, but osteoclasts do [11]. When the resorbing cells digest BioCaP, the protein release from BioCaP could be elevated. The findings in

163

Chapter 10

Chapter 2 indicate osteoclasts are able to increase the protein release from BioCaP. However, when BioCaP delivers BMP-2, this growth factor could also influence the proliferation and differentiation of these cells. This needs further investigated. When biomaterials were implanted into body, multinucleated giant cells (MGCs) formed on biomaterials could be osteoclasts or foreign body giant cells (FBGCs). However, FBGCs is the end stage of the inflammatory responses following the implantation of biomaterial. FBGCs can also release inflammatory cytokines which stimulate circulating stem cells to become osteoprogenitors [10, 13]. Osteoclast has become the common term to denote any cell that has a unique function to break down mineralized matrices [14, 15]. However, it is difficult to precisely identify FBGCs and osteoclasts, since both of them are tartrate resistant acid phosphatase (TRAP) positive. This is very interesting to investigate further.

Icariin: an osteoinductive traditional Chinese medicine Traditional Chinese Medicines (TCMs) have been recommended for bone regeneration and repair for thousands of years. Icariin, a typical flavonol glycoside, is considered to be the main active ingredient of the Herba Epimedii from which icariin has been successfully extracted. Because of its osteoinductive potential and the low price, it can be a very attractive candidate as a substitute of BMP-2 (Chapter 8). We thought about that icariin can be slowly delivered by BioCaP, since icariin has been incorporated into the interior of BioCaP. In our on-going study, we are going to investigate the therapeutic effects of icariin-incorporated BioCaP with or without BMP-2 in the treatment of critical-sized bone defects in rats.

Micro-CT and histological analysis Histological analyses are derived from stereological analysis of a few 2D sections, usually assuming that the underlying structure is plate-like [16]. The inclusion of dental implants may further lessen the histomorphometric information of bone because only one central section of each implant can be used for analysis [17]. In contrast, micro-computed tomography (micro-CT) can directly measure bone micro-architectures independent of stereological models [18]. The high positive correlation (Pearson coefficient=0.992, p<0.001) between these two parameters was found (Chapter 9). This validated the reliability of micro-CT. This finding is consistent with previous studies in animal [19] and human specimens [20]. However, when new bone was deposited on bone substitute, single threshold setting cannot distinguish bone and materials. The micro-CT method in Chapter 3 successfully separated BioCaP and new bone by using an “onion-peeling” algorithm (Scanco Medical AG) and specific threshold settings. The results were confirmed by the histological analysis in Chapter 2. However, the correlation between the two parameters needs to be evaluated further.

164

General discussion

CONCLUSION

The main scope of the work conducted in this thesis was to develop osteoinductive bone substitutes.

1. It was shown that BioCaP with an internal or surface coated depot of protein has the capacity to maintain a slow and sustained protein release. Both modes of delivering of BMP-2 with the use of BioCaP make these granules efficient osteoinductive compounds. Benefiting from these two delivery modes, BioCaP with BMP-2 can be a promising alternative to the autografts. The findings also showed that BioCaP has good biocompatibility and degradability.

2. The findings indicate BMP2-cop.BioCaP can serve as a highly efficient osteoinducer for inducing bone formation with DBB and for suppressing the foreign-body reaction in a critical-sized bone defect. BMP2-cop.BioCaP also showed good biocompatibility and degradability. This material has a very promising clinical potential to enhance significantly the therapeutic effects of bone substitutes for filling bone defects.

3. Our findings show the excellent biocompatibility and osteoconductivity of DBB. Incorporating BMP-2 into the calcium phosphate coating of DBB induced strong bone formation around DBB in critical-sized bone defects. This functionalization approach renders DBB efficiently osteoinductive and could greatly enhance the clinical potential of DBB to be an alternative to bone autografts in the repair of large or critical-sized bone defects.

4. Low-dose BMP2/7 heterodimer facilitated more rapid bone regeneration in better quality in peri-implant bone defects than BMP2 and BMP7 homodimers. Micro-CT results were confirmed by histological analysis.

5. The osteoinductive potential and the low price of icariin make it a very attractive candidate as a substitute of osteoinductive protein − BMPs, or as a promoter for enhancing the therapeutic effects of BMPs. However, the effectiveness of the local delivery of icariin needs to be investigated further.

FUTURE SCOPE We are on the way to developing an ideal osteoinductive bone substitute. Osteoinductive BioCaP with incorporated BMP-2 is expected to be evaluated in clinical trials in the future. BioCaP granule might also be considered as a promising tool for the orderly delivery of multiple therapeutic agents, such as antibiotics, osteogenic agents, and anti-cancer drugs for different clinical applications.

165

Chapter 10

REFERENCES 1. Hunziker EB, Enggist L, Kuffer A, Buser D, Liu Y. Osseointegration: The slow delivery of BMP-2 enhances osteoinductivity. Bone 2012;51:98-106. 2. Wernike E, Montjovent MO, Liu Y, Wismeijer D, Hunziker EB, Siebenrock KA, et al. VEGF incorporated into calcium phosphate ceramics promotes vascularisation and bone formation in vivo. Eur Cell Mater 2010;19:30-40. 3. Habraken WJ, Wolke JG, Jansen JA. Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:234-48. 4. Kempen DH, Lu L, Heijink A, Hefferan TE, Creemers LB, Maran A, et al. Effect of local sequential VEGF and BMP-2 delivery on ectopic and orthotopic bone regeneration. Biomaterials 2009;30:2816-25. 5. Wu G, Hunziker E, Zheng Y, Wismeijer D, Liu Y. Functionalization of deproteinized bovine bone with a coating-incorporated depot of BMP-2 renders the material efficiently osteoinductive and suppresses foreign-body reactivity. Bone 2011;49:1323-30. 6. Schwarz F, Rothamel D, Herten M, Ferrari D, Sager M, Becker J. Lateral ridge augmentation using particulated or block bone substitutes biocoated with rhGDF-5 and rhBMP-2: an immunohistochemical study in dogs. Clin Oral Implants Res 2008;19:642-52. 7. Liu Y, Wu G, de Groot K. Biomimetic coatings for bone tissue engineering of critical-sized defects. J R Soc Interface 2010;7 Suppl 5:S631-47. 8. Schwartz Z, Weesner T, van Dijk S, Cochran DL, Mellonig JT, Lohmann CH, et al. Ability of deproteinized cancellous bovine bone to induce new bone formation. J Periodontol 2000;71:1258-69. 9. Araujo M, Linder E, Lindhe J. Effect of a xenograft on early bone formation in extraction sockets: an experimental study in dog. Clin Oral Implants Res 2009;20:1-6. 10. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol 2008;20:86-100. 11. Wernike E, Hofstetter W, Liu Y, Wu G, Sebald HJ, Wismeijer D, et al. Long-term cell-mediated protein release from calcium phosphate ceramics. J Biomed Mater Res A 2010;92:463-74. 12. Zhang Z, Egana JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al. Cell-based resorption assays for bone graft substitutes. Acta Biomater 2012;8:13-9. 13. Le Nihouannen D, Saffarzadeh A, Gauthier O, Moreau F, Pilet P, Spaethe R, et al. Bone tissue formation in sheep muscles induced by a biphasic calcium phosphate ceramic and fibrin glue composite. J Mater Sci-Mater M 2008;19:667-75. 14. Basle MF, Chappard D, Grizon F, Filmon R, Delecrin J, Daculsi G, et al. Osteoclastic resorption of Ca-P biomaterials implanted in rabbit bone. Calcif Tissue Int 1993;53:348-56. 15. Everts V, de Vries TJ, Helfrich MH. Osteoclast heterogeneity: lessons from osteopetrosis and inflammatory conditions. Biochim Biophys Acta 2009;1792:757-65. 16. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, et al. Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. 1987;2:595-610. 17. Sennerby L, Dasmah A, Larsson B, Iverhed M. Bone tissue responses to surface-modified zirconia implants: A histomorphometric and removal torque study in the rabbit. 2005;7 Suppl 1:S13-20. 18. Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ, Muller R. Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. 2010;25:1468-86. 19. Waarsing JH, Day JS, Weinans H. An improved segmentation method for in vivo microCT imaging. 2004;19:1640-50. 20. Chappard D, Retailleau-Gaborit N, Legrand E, Basle MF, Audran M. Comparison insight bone measurements by histomorphometry and microCT. 2005;20:1177-84.

166

Chapter 11

General summary Algemene samenvatting

167

Chapter 11

GENERAL SUMMARY

In the treatment of bone defects an adequate volume of bone tissue is of paramount importance to achieve bone regeneration. When the bone defects are too large to allow self-healing or post-traumatic complications occur such as delayed union, non-union or malunion, bone grafting is indicated in order regenerate the defect. Autografts (gold standard), allografts, xenografts, and synthetic materials are available to assist in the repair of bone defects. Synthetic calcium phosphate (CaP) biomaterials are widely used in the regeneration of bone defects because of their chemical similarity to native bone tissue. To achieve bone regeneration in large or critical sized bone defects, (i) osteogenic cells (e.g. progenitor cells or osteoblasts); (ii) osteoinductive signals (growth factors); (iii) a biocompatible, biodegradable and osteoconductive matrix (scaffold); and (iv) adequate blood and nutrient supply are required. Bone grafts are often associated with the terms biocompatibility, biodegradability, osteoconductivity and osteoinductivity. This thesis discusses the research on a biomimetic calcium phosphate bone substitute (BioCaP). For the first time we have developed a dual protein release system using this material. In this system protein and calcium phosphate were precipitated together to form BioCaP granules in which, in the center of the granules a depot of protein was incorporated, a so called internal depot. Next, protein and calcium phosphate were co-precipitated onto the surface of these granules, thus creating a surface coated depot. This dual system provides an ideal model for delivery of different proteins/drugs in two phases, an initial slow delivery phase (surface coated depot) and a delayed phase (internal depot). By adopting this system, a single drug can be administered in a more consistent manner or two different drugs can be administered simultaneously. Therefore, BioCaP granules might be considered as a promising tool for the systematic delivery of multiple therapeutic agents, such as antibiotics, osteogenic agents, and anti-cancer drugs for different clinical applications. Moreover, we developed particles of biomimetic BMP-2-coprecipitated calcium phosphate (BMP2-cop.BioCaP) as an osteoinductive promoter for commercial bone substitutes such as deproteinized bovine bone (DBB). We also functionalized DBB with a coating-incorporated depot of BMP-2 for the repair of critical-sized bone defects. All the products developed and discussed in this thesis were based on the principle of biomimetic calcium phosphate coating.

The general aim of this thesis includes 5 aspects: 1. To develop a biomimetic calcium phosphate (BioCaP) bone substitute as a dual delivery model with two protein-delivery modes: one mode by which protein was incorporated in the interior of BioCaP; and one by which protein was coated on the outside of BioCaP. We hypothesize that using this model the release of the protein can be sequential and slow, and that the two delivery modes of BMP-2 could efficiently accelerate bone formation. 2. To develop particles of biomimetic BMP-2-coprecipitated calcium phosphate (BMP2-cop.BioCaP). We hypothesize that these particles could serve as an independent and biodegradable osteoinducer. 3. To evaluate the therapeutic effect of the deproteinized bovine bone functionalized with coating-incorporated BMP-2 in the repair of critical-sized bone defect in sheep.

168

General summary

4. To delineate the dynamic micro-architectures of bone induced by low-dose bone morphogenetic protein (BMP)-2/7 heterodimer in peri-implant bone defects compared to BMP2 and BMP7 homodimer. 5. To determine the present evidence of the osteoinductive potential of a Chinese traditional medicine, icariin.

Cells responsible for bone resorption such as osteoclasts may accelerate the degradation of bone substitutes and so increase the protein release. In Chapter 2, mouse osteoclasts were used to test the cell-mediated protein release from BioCaP which was produced by refining a well-established biomimetic protocol. It was shown that BioCaP with the proteins retained in the internal and surface coatings resulted in a sustained osteoclast-mediated release, while the adsorbed protein was rapidly released. This release of the adsorbed protein was not affected by osteoclasts seeded on BioCaP. Next, granules of BioCaP with an internal or a surface coated depot of BMP-2 were implanted subcutaneously in rats. Histological analysis showed that the volume densities of bone, bone marrow, and blood vessels were significantly higher in samples where BMP-2 was incorporated internally or in the coating compared with granules with adsorbed growth factor. In the latter samples fibrous capsular tissue proved to be significantly higher. Osteoclast-like cells and the resorption lacunae were observed on BioCaP granules in vivo. Different modes of incorporation of BMP-2 on and in BioCaP granules had a beneficial effect on the formation of ectopic bone. This dual drug release system makes the BioCaP granule a promising tool for delivering multiple therapeutic agents, such as osteogenic agents, antibiotics, and anti-cancer drugs for different clinical applications. In Chapter 3, the physicochemical properties of BioCaP were investigated. Two phases of precipitation of BioCaP were observed by scanning electron microscopy. BioCaP exhibited bone-like mechanical strength and the characteristics of calcium-deficient apatite. The granules with internally- or coating-incorporated protein exhibited a slow release in vitro (35 days). Human osteoclasts seeded on the granules were shown to resorb the BioCaP. This finding is consistent with the results in the previous study using mouse osteoclasts in Chapter 2. Micro-CT analysis using an “onion-peeling” algorithm can distinguish between BioCaP and newly formed bone. In a rat ectopic model Micro-CT results showed that significantly more bone formation was present in the samples containing BioCaP with internally- or coating- incorporated BMP-2 than those with adsorbed BMP-2. BioCaP with BMP-2 showed slower degradation than that without BMP-2. These results confirmed the histological results in Chapter 2. In Chapter 4 we investigated the therapeutic effect of BioCaP with two delivery modes of BMP-2 in the repair of a large cylindrical bone defects (Ø8×13mm) in sheep. Both delivery modes of BMP-2 accelerated the bone formation within a period of 4 weeks. The internally-incorporated protein mode enhanced bone formation after 8 weeks, showing to be more efficient than DBB. BioCaP with BMP-2 showed equal efficacy as autologous bone in the bone defect repair at 8 weeks post-implantation. BioCaP with BMP-2 showed significant degradation at both time points. Benefiting from these two delivery modes, BioCaP might be a promising alternative to autografts. Novel particles of biomimetic BMP2-coprecipitated calcium phosphate (BMP2-cop.BioCaP) were developed to serve as an independent and biodegradable

169

Chapter 11 osteoinducer with the idea to induce bone formation more efficiently for bone-defect-filling materials such as DBB. To prepare BMP2-cop.BioCaP, we alternately layer-by-layer assembled amorphous and crystalline CaP triply to enable a “bamboolike” growth of the particles. BMP2 was incorporated into the outermost layer of BioCaP. We monitored the degradation, osteoinductivity, and foreign-body reaction of both BMP2-cop.BioCaP and its combination with DBB in an ectopic site in rats. After 5 weeks, the BMP2-cop.BioCaP significantly induced new bone formation not only when applied as a solitary product but also when mixed with DBB. Its osteoinductive efficiency was 10-fold higher than the BioCaP with adsorbed BMP2. More than 90% of BMP2-cop.BioCaP degraded. Moreover, BMP2-cop.BioCaP also significantly reduced the host foreign-body reaction to DBB in comparison with the adsorbed BMP2. These findings indicate a promising clinical potential for BMP2-cop.BioCaP in the repair of large (critical)-sized bone defects (Chapter 5). To enhance the therapeutic effect of DBB, we mixed DBB with BMP2-cop.BioCaP in critical-sized bone defects in sheep. Histological results confirmed the excellent biocompatibility and osteoconductivity of BMP2-cop.BioCaP and DBB. DBB mixed with BMP2-cop.BioCaP displayed significantly more bone formation when compared to DBB, and the induced bone formation was comparable with autologous bone at 8 weeks post-implantation. At this time point, about 95% BMP2-cop.BioCaP was degraded. It was shown that the BMP2-cop.BioCaP, as an osteoinductive promoter, has excellent biocompatibility, biodegradability, osteoconductivity, and a strong capacity to induce bone formation. It might be possible to substitute an autograft by DBB mixed with BMP2-cop.BioCaP in the repair of critical-sized bone defects (Chapter 6). In order to render DBB osteoinductive, BMP-2 has previously been incorporated into a three dimensional reservoir (a biomimetic calcium phosphate coating) on DBB, which effectively promoted the osteogenic response by the slow delivery of BMP-2 as described in Chapter 7. The aim of this chapter was to investigate the therapeutic effectiveness of such coatings on the DBB granules in repairing large cylindrical bone defects in sheep. Histological results confirmed the excellent biocompatibility and osteoconductivity of the coated DBB. Incorporating BMP-2 into the calcium phosphate coating of DBB induced strong bone formation around DBB when repairing bone defects. 8 Weeks after implantation, the volume of newly-formed bone around DBB that bore a coating-incorporated depot of BMP-2 was comparable to that of autologous bone. Multinucleated giant cells were observed in the resorption process around DBB, whereas histomorphometric analysis revealed no significant degradation of the DBB. More rapid repair of peri-implant bone defects has been pursued for years. Heterodimeric BMPs exhibited several- or even multiple- effect than the respective homodimers in inducing in vitro osteoblastogenesis. We administrated collagen sponges with adsorbed low-dose (30 ng/mm3) BMP2/7 to treat the peri-implant defects in the calvaria of minipig. After 6 weeks post-operation, BMP2/7 showed a significantly higher relative bone volume and significantly lower structure mode index than BMP-2 and BMP-7 respectively. The findings indicate that low-dose BMP2/7 heterodimer facilitated more rapid bone regeneration in better quality in peri-implant bone defects than BMP-2 and BMP-7 homodimers (Chapter 8). BMP-2 and BMP-7 are the most extensively evaluated BMPs and they are very costly. Although BMPs have an outstanding performance in bone formation, in some cases when

170

General summary used in a high concentration could also cause some negative effects. Research has developed a variety of methods in bone tissue engineering to reduce the use of BMPs and to improve the osteoinductive effects of BMPs by a slow delivery. Icariin, a typical flavonol glycoside, is considered to be the main active ingredient of the Herba Epimedii (a Traditional Chinese Medicine) from which icariin has been extracted. Most interestingly, it has been reported that icariin can be delivered locally by biomaterials and that it has an osteoinductive potential for bone tissue engineering. Therefore, we reviewed the performance of icariin in bone tissue engineering and blended this information with the current knowledge relevant to molecular mechanisms and signal pathways. The osteoinductive potential of icariin could be attributed to its multiple functions in the musculoskeletal system which is involved in the regulation of multiple signaling pathways in anti-osteoporosis, osteogenesis, anti-osteoclastogenesis, chondrogenesis, angiogenesis, and anti-inflammation. The osteoinductive potential and the low price of icariin make it a very attractive candidate as a substitute of osteoinductive proteins − BMPs, or as a promoter for enhancing the therapeutic effects of BMPs (Chapter 9).

171

Algemene samenvatting

ALGEMENE SAMENVATTING

Bij de behandeling van botdefecten is een voldoende hoeveelheid botweefsel van belang om botregeneratie te bereiken. In het geval van botdefecten van een dusdanige grootte dat spontaan herstel niet optreedt, of in geval van post-traumatische complicaties als delayed union, malunion en nonunion. is transplantatie van bot geïndiceerd om het defect te regenereren. Autograft (de gouden standaard), allografts, xenografts en synthetische materialen zijn hiervoor beschikbaar. Dankzij de chemische overeenkomst met menselijk bot, worden synthetische calciumfosfaat (CaP) biomaterialen wijdverspreid gebruikt bij de regeneratie van botdefecten. Om botregeneratie te bewerkstelligen in grote of critical-size botdefecten zijn (i) osteogenetische cellen (progenitor cellen of osteoblasten); (ii) osteoinductieve signalen (groeifactoren); (iii) een biocompatibele, biologisch afbreekbare en osteoconductieve matrix (geraamte); en (iv) een voldoende toevoer van blood en voedingsstoffen vereist. Bottransplantaten zijn veelal geassocieerd met de termen biocompatibiliteit, biologische afbreekbaarheid, osteoconductiviteit en osteoinductiviteit. Dit proefschrift behandelt het onderzoek naar een biomimetisch calciumfosfaat botsubstituut. Voor de eerste keer is dit materiaal gebruikt om een dual-protein-release-systeem te ontwikkelen. In dit systeem, een zogenoemd intern depot bestaande uit proteïnen, is gebruikt als basis om de proteïne en calciumfosfaat, samen BioCaP granulaat vormend, op te laten neerslaan, wat tot incorporatie van het depot leidt. Vervolgens is dit proces herhaalt maar met het hiervoor gevormde granulaat als basis, leidend tot een depot met oppervlakte coating. Dit tweeledige systeem geeft een ideaal model om proteïnen/stoffen in twee fases toe te dienen, een langzame initiële fase (oppervlakte coating) en een vertraagde fase (intern depot). Door dit systeem te gebruiken kan een enkele stof geleidelijker worden toegediend, of twee verschillende stoffen tegelijkertijd. Hierom kan BioCaP granulaat gezien worden als een veelbelovend systeem voor de systematische toediening van diverse therapeutische middelen zoals antibiotica, osteogenetische middelen, en kanker medicatie voor diverse klinische doeleinden. Daarbij hebben we biomimetische BMP-2-coprecipitated calciumfosfaat (BMP2-cop.BioCaP) ontwikkeld als een osteoinductieve promotor voor commerciële botsubstituten zoals deproteinized bovine bone (DBB). Tevens hebben we DBB gefunctionaliseerd met een coating-incorporated depot van BMP-2 voor de regeneratie van critical-sized botdefecten. Alle ontwikkelde en in dit proefschrift besproken producten zijn gebaseerd op het principe van biomimetische calciumfosfaat coating.

Het doel van dit proefschrift bestaat uit 5 aspecten: 1. De ontwikkeling van een biomimetisch calciumfosfaat (BioCaP) botsubstituut als een dual delivery model met twee proteïne afgifte methoden: een methode waarbij het proteïne is geïncorporeerd in de kern van de BioCaP; en een methode waarbij het proteïne als een coating aan de buitenzijde van de BioCaP was aangebracht. De hypothese luidt dat het gebruik van dit model de afgifte van het proteïne geleidelijk en langzaam is, en dat de twee methoden van BMP-2 toediening botvorming efficiënt kan versnellen. 2. De ontwikkeling van biomimetische BMP-2-coprecipitated calciumfosfaat (BMP2-cop.BioCaP) granules. De hypothese luidt dat deze granules kunnen dienen als

172

Algemene samenvatting een op zichzelf staande en biologisch afbreekbare osteoinducer. 3. De evaluatie van het therapeutische effect van het deproteinized bovine bone gefunctionaliseerd met coating-incorporated BMP-2 bij het herstel van critical-size botdefecten in schapen. 4. Het in kaart brengen van de dynamische micro architectuur van bot geïnduceerd door lage doses bone morphogenetic protein (BMP)-2/7 heterodimeer in peri implantaire botdefecten in vergelijking met BMP2 en BMP7 homodimeer. 5. Het vaststellen van het reeds bestaande bewijs van het osteoinductieve potentieel van een Chinees traditioneel medicijn, icariin.

De cellen die verantwoordelijk zijn voor bot resorptie, zoals osteoclasten, kunnen de degeneratie van bot substituten versnellen en zo de proteïne afgifte vergroten. In hoofdstuk twee zijn osteoclasten afkomstig van muizen gebruikt om de cell-mediated protein release van BioCaP te testen, welke is geproduceerd door het verfijnen van een goed gegrond biomimetisch protocol. Het is aangetoond dat BioCaP, met de proteïnen vastgehouden als kern en als oppervlakte coating, resulteert in een stabiele osteoclast-gemedieerde afgifte, terwijl de geadsobeerde proteïnen snel worden afgegeven. Het loslaten van de geadsorbeerde proteïnen is niet beïnvloed door osteoclasten op de BioCaP. Vervolgens zijn granules van BioCaP met een intern depot of coating van BMP-2 subcutaan aangebracht in ratten. Histologische analyse toont aan dat de volume dichtheid van bot, merg en bloedvaten significant hoger is in de samples waarin BMP-2 intern of in de coating is geïncorporeerd in vergelijking met granules met geadsorbeerde groei factoren. In verdere samples werd een fibreus kapsel significant vaker aangetroffen. Osteoclast-achtige cellen en resorptielagunes zijn in vivo aangetroffen op BioCaP granules. Verschillende methode van incorporatie van BMP-2 op en in BioCaP granules hebben een positief effect op ectopische botformatie. Dit dual drug delivery system maakt de BioCaP granule een veelbelovend middel om verschillende therapeutische middelen, zoals osteogenetische middelen, antibiotica en kankermedicatie voor diverse klinische doeleinden toe te dienen. In hoofdstuk drie worden de fysiochemische eigenschappen van BioCap onderzocht. Twee fasen van het neerslaan van BioCaP zijn geobserveerd met behulp van een scanning electronen microscoop. BioCaP vertoont mechanische sterkte vergelijkbaar met bot en de karakteristieken van calciumdeficiënt apatiet. De granules met interne of coating-incorporated proteïnen laten een langzame afgifte zien in vitro (35 dagen). De menselijke osteoclasten gezeten op de granules resorberen de BioCaP. De bevinding komt overeen met de resultaten van een eerdere studie uit hoofdstuk twee waarbij osteoclasten afkomstig van muizen zijn gebruikt. Micro-CT analyse waarbij gebruik is gemaakt van een “union-peeling” algoritme kan onderscheid maken tussen BioCaP en nieuw gevormd bot. Micro-CT resultaten uit een ectopisch rat model laten zien dat er significant meer botformatie is in samples die BioCaP met intern of coating-incorporated BMP-2 bevatten, in vergelijking met samples met geadsorbeerd BMP-2. BioCaP met BMP-2 vertoont een tragere degeneratie dan BioCaP zonder BMP-2. Deze resultaten bevestigen de histologische resultaten uit hoofdstuk 2. In hoofdstuk vier onderzoeken we het therapeutisch effect van BioCaP met twee afgifte methoden van BMP-2 bij de reparatie van grote cilindrische botdefecten (Ø8×13mm) in schapen. Beide methoden versnellen de botformatie over een periode van

173

Chapter 11 vier weken. De intern geïncorporeerde proteïne methode versnelt botformatie over een periode van acht weken, waarmee is aangetoond dat dit efficiënter is dan DBB. Acht weken na implantatie vertoont BioCaP met BMP-2 een gelijke efficiëntie als autoloog bot in de reparatie van het botdefect. BioCaP met BMP-2 vertoont significante degeneratie op beide tijdstippen. Gezien het voordeel dat BioCaP heeft bij beide methoden, zou dit een veelbelovend alternatief kunnen zijn voor autografts. Nieuwe deeltjes van biomimetisch BMP2-coprecipitated calciumfosfaat (BMP2-cop.BioCaP) zijn ontwikkeld om te dienen als een zelfstandig en biologisch afbreekbare osteoinducer met het idee de efficiëntie van botformatie door botsubstituten als DBB te vergroten. Om BMP2-cop.BioCaP te creëren hebben we om en om cq laag voor laag amorf en gekristalliseerd calciumfosfaat in drievoud aangebracht om zo een “bamboe-achtige” groeimethode te realiseren. BMP-2 is geïncorporeerd in de buitenste laag van de BioCaP. De degeneratie, osteoinductiviteit en vreemd-lichaam-reactie is gevolgd van zowel BMP2-cop.BioCaP als BMP2-cop.BioCaP gecombineerd met DBB in ectopische locaties in ratten. Na vijf weken vertoont de BMP2-cop.BioCaP een significante inductie van nieuwe botformatie, niet alleen als solitair product, maar ook gemixt met DBB. De osteoinductieve efficiëntie is een tienvoud hoger dan BioCaP met adsorbed BMP-2. Meer dan 90% van de BMP2-cop.BioCaP is gedegenereerd. BMP2-cop.BioCaP zorgt tevens voor een reductie in de vreemd-lichaam reactie tegen DBB in vergelijking met het geadsorbeerde BMP-2. Deze bevindingen indiceren een veelbelovend klinisch potentieel voor BMP2-cop.BioCaP bij de reparatie van grote (critical-sized) botdefecten (Hoofdstuk vijf). Om het therapeutisch effect van DBB te vergroten hebben we DBB gemixt met BMP2-cop.BioCaP in critical-sized botdefecten in schapen. Histologische resultaten hebben de perfecte biocompatibiliteit en osteoconductiviteit van BMP2-cop.BioCaP en DBB bevestigd. DBB gemixt met BMP2-cop.BioCaP laat een significant hogere botformatie zien in vergelijking met DBB, en de geïnduceerde botformatie is vergelijkbaar met autoloog bot acht weken na aanbrengen. Op dat moment is ongeveer 95% BMP2-cop.BioCaP gedegenereerd. Het is aangetoond dat BMP2-cop.BioCaP als een osteoinductieve promotor een perfecte biocompatibiliteit, biologische afbreekbaarheid en osteoconductiviteit heeft, evenals een grote capaciteit tot het induceren van botformatie. Het is bij de reparatie van critical-sized botdefecten wellicht mogelijk om een autograft te vervangen door een mix van BMP2-cop.BioCaP en DBB. (Hoofdstuk zes). Om DBB osteoinductief te maken is BMP-2 al eerder geïncorporeerd in een driedimensionaal reservoir (een biomimetische calciumfosfaat coating) op DBB, wat de osteogenetische respons van de langzame afgifte van BMP-2, zoals beschreven in hoofdstuk zeven, effectief bevordert. Het doel van dit hoofdstuk is het onderzoeken van de therapeutische effectiviteit van dit soort coating op de DBB granules bij de reparatie van grote cilindrische bot defecten in schapen. Histologische resultaten bevestigen de goede biocompatibiliteit en osteoconductiviteit van de gecoate DBB. Het incorporeren van BMP-2 in de calciumfosfaatcoating van DBB induceert een duidelijke botformatie rond DBB bij de reparatie van botdefecten. Acht weken na implantatie is het volume van nieuw gevormd bot rond DBB dat een coating-incorporated depot van BMP-2 droeg, vergelijkbaar met dat van autoloog bot. Multinucleaire reuscellen zijn aangetroffen in het resorptieproces rond DBB, terwijl histomorphometrische analyse geen significante

174

Algemene samenvatting afbraak van de DBB aantoont. Een sneller herstel van peri implantaire botdefecten wordt al jaren nagestreefd. Heterodimeric BMP’s bevatten verschillende – of zelfs meer – effecten dan de respectievelijke homodimeren met betrekking tot induceren van in vitro osteoblastogenese. We hebben sponsen aangebracht met geadsorbeerde BMP2/7 in lage doses om de peri implantaire defecten in caviaschedels te behandelen. Zes weken post operatief laat de BMP2/7 een significant hoger relatief bot volume zien en een significant lagere structure mode index in vergelijking met BMP-2 en BMP-7 respectievelijk. Deze bevindingen indiceren dat een lage dosis BMP2/7 heterodimeer een snellere bot regeneratie van betere kwaliteit in peri implantaire botdefecten faciliteert dan BMP-2 en BMP-7 homodimeren (Hoofdstuk acht). BMP-2 en BMP-7 zijn de meest uitgebreid geëvalueerde BMP’s en zijn erg kostbaar. Hoewel BMP’s een buitengewone staat van dienst hebben qua botformatie, kunnen ze in sommige situaties als ze in hoge concentraties gebruikt worden negatieve effecten hebben. Onderzoek heeft geleid tot de ontwikkeling van tal aan methoden in bone tissue engineering om het gebruik van BMP’s te beperken en de osteoinductieve effecten van BMP’s te vergroten door middel van een vertraagde toediening. Icariin, een typisch flavonol glycoside wordt gezien als het voornaamste actieve ingrediënt van de Herba Epimedii (een traditioneel Chinees medicijn), waar deze icariin uit gewonnen wordt. Het is interessant dat er wordt gemeld dat icariin lokaal kan worden afgegeven door biomaterialen en dat het een osteoinductief potentieel heeft voor bone tissue engineering. Daarom hebben we de werking van icariin in de bone tissue engineering beoordeeld en hebben we deze informatie samengevoegd met de huidige kennis met betrekking tot moleculaire mechanismen en signaal routes. Het osteoinductieve potentieel van icariin kan worden toegedicht aan zijn vele functies in het musculoskeletale systeem dat is betrokken bij de regulatie van verschillende signalering routes in anti-osteoporose, osteogenese, anti-osteoclastogenese, chondrogenese, angiogenese en anti-inflammatie. Het osteoinductieve potentieel en de lage prijs van icariin maken het een erg aanlokkelijke kandidaat als vervanging voor osteoinductieve proteïnen – BMP’s, of als een promotor om de therapeutische effecten van BMP’s te versterken (Hoofdstuk negen).

175

Chapter 11

176

Acknowledgments

ACKNOWLEDGEMENTS

First and foremost, I would like to express my sincerest gratitude to my supervisors, prof. dr. Daniel Wismeijer and dr. Yuelian (Maria) Liu, who kindly provided me with the opportunity to do this research in ACTA.

Dear Daniel, it is my great honor to have you as my supervisor. Thank you for fully and strongly supporting me to complete my research and thesis. Your concern, guidance and encouragement always gave me great power to keep working. I am deeply inspired by your intellectual elegance and amiable personality. Despite your role as the head of the department of oral implantology and all other obligations, you always had time to assist me, making everything go silky smooth.

Dear Maria, thank you for leading me through my whole Ph.D. pursuit. Words have not been able to express the gratitude that I feel for your tireless help and encouragement. You are a wonderful guider, providing me so many ideas. You always stood beside me, encourage and guide me, whenever I met with difficulties. I greatly appreciate your efforts to make my doctorial journey productive.

Dear prof. Zhiyuan Gu, thank you for introducing me to ACTA. It is so lucky that I have you as my supervisor. I have had great respect for your noble personality, since I was a master student.

Dear prof. Vincent Everts, thank you for being a knowledgeable advisor. You are an amazing advisor. I always asked questions, and you always gave me the right answers. I really learned a lot from you. Thank you for your insightful questions and comments on those manuscripts, and thus make me know how to write a good article.

Dear Ton Broncker, thank you for teaching me about histological studies. You always provided very nice ideas to me.

Dear Gang, thank you for your company. We had lived together for almost two years. Everything seems like just happened yesterday. I enjoyed discussing research issues with you. I could see your attitude on research is very serious.

Dear Afsheen, thank you for your suggestions on the manuscripts. I learned many writing skills from you.

Dear Sven, I would like to thank you for the translation. Without your help, I cannot complete the thesis.

In regards to the cell biology studies, I especially would like to thank Jolanda, Cor, Dirk-Jan, Teun, Ton, Ineke, Behrouz, and Marion. I also would like to thank Leo and Jan Harm for your kind help. All of you provided me techniques and supports to complete my research.

177

Acknowledgments

Jenny, Janak, Alejandra, Marjolein, Ceylin, Nawal, Greetje, Sepanta, I am lucky to have your help and to share Ph.D. life with you.

Chenfeng, Qilong, Lei, Xingnan, Xiao, Dongyun, thank you for the passionate friendship. We had a very good time.

Special appreciate should be given to all my Chinese and Dutch friends in ACTA or once in ACTA.

Lastly, I would like to thank my parents, my parents-in-low, and my sweet wife for your endless love and support.

至此，再次衷心感谢我的导师刘月莲老师，多谢您在这四年来的悉心指导。

最后，我希望对我的父母，岳父岳母，以及我最爱的妻子表达深深的感谢。 你们是我心中的动力。没有你们的支持，我无法完成这四年的学业。 特别感谢我可爱的妻子陆倩，感谢你的理解，支持和奉献。

178

Curriculum vitae

CURRICULUM VITAE

Name: Tie Liu Date of birth: 17th February, 1983, Place of birth: Yuyao, Zhejiang province, China

Contact Mobile phone: (86)13588135231 (China) Email: [email protected] [email protected]

Scientific education 2009.09-2013.09 PhD-student. Department of Oral Implantology and Prosthetic Dentistry, Academic Centre for Dentistry Amsterdam (ACTA), Research Institute MOVE, University of Amsterdam and VU University Amsterdam, The Netherlands. 2007.09 Exchange student. School of Dentistry, Medical College of Georgia, USA 2007.08-2009.07 Master degree. Oral and Maxillofacial Surgery / Oral Implantology, Hospital/School of Stomatology, Zhejiang University. 2002.09-2007.07 Bachelor degree. General Dentistry, Hospital/School of Stomatology, Zhejiang University.

Clinical education 2007.05-2009.07 Hospital/School of Stomatology, Zhejiang University. Internship. 2007.02-2007.04 Department of Oral and Maxillofacial Surgery. The First Affiliated Hospital of College of Medicine, Zhejiang University (First Hospital of Zhejiang Province). Internship. 2006.06-2007.01 Hospital/School of Stomatology, Zhejiang University. Internship.

179