Understanding the Effects of Dopants on Calcium Phosphate Ceramics: Bone Cell

Home , DMP1, Osteonectin, Regucalcin

Differentiation and Bone Growth in vitro and in vivo

By:

Gary Fielding

A dissertation submitted in partial fulfillment of

the requirements for the degree of:

DOCTOR OF PHILOSOPHY

WASHINGTON STATE UNIVERSITY

Engineering Science Program

December 2013

To the Faculty of Washington State University:

The members of the Committee appointed to examine the dissertation of GARY FIELDING find it satisfactory and recommend that it be accepted.

Susmita Bose, Ph.D., Chair

Amit Bandyopadhyay, Ph.D.

Anita Vasavada, Ph.D.

Acknowledgements

At the foremost of the long list of people that have made this dissertation possible, I would like to extend my gratitude to my advisor Dr. Susmita Bose. Without her vision, drive and expertise, this body of work as well as my professional and personal development could not be a reality.

I would also thank Dr. Amit Bandyopadhyay for all of the valuable lessons that were passed on to me over the course of my PhD career. It has been a great learning experience for me.

I would also like to give a very special thank you to Dr. Anita Vasavada, who has been offering me advice and guidance since nearly the beginning of my undergraduate career at

Washington State University. Over the past 8 years, her guidance has been invaluable.

Thank you to all of the graduate students that I have had the privilege and pleasure to work with, with special recognition to Dr. Solaiman Tarafder and Dr. Mangal Roy, who spent many hours passing down their knowledge to me and laid the foundation for my career as a graduate student. I can’t thank Subhadip Bodhak, Felix Espana, Joe Edgington, Abhimanyu

Bhat, Stan Dittrick, Sahar Vahabzadeh, Dongxu Ke and Khalid Emshadi enough for the company and fellowship that made this experience a truly memorable one. Also, a special thanks to Pasha Rudenko, a fellow entrepreneurial spirit, who has provided hours of interesting conversation as well as a lasting friendship. I look forward to the continuing successes in everyone’s lives.

And finally, but certainly not least, I would like to give my deepest gratitude to my family. To my mother, Sherry, and my father, Gary, who taught me some of life’s most valuable

iii lessons and continue to do so. To my sisters Melissa, Amanda and Andrea who have helped shape the person I am today. And to the two most important people in my life, my wife, Tricia and our beautiful daughter, Stormy, I would like to sincerely thank you. Without Tricia’s support, understanding, patience and everlasting love, I could not have succeeded in this. Our daughter has been the most precious gift I could have asked for and has given me an entirely new perspective on life. Thank you again.

Financial support for this work was provided by the National Institutes of Health, NIBIB under the Grant # NIH-R01-EB-007351. M. J. Murdock charitable trust provided the funding to acquire the Ex One system.

Understanding the Effects of Dopants on Calcium Phosphate Ceramics: Bone Cell

Differentiation and Bone Growth in vitro and in vivo

Abstract

Gary Fielding, Ph.D.

Washington State University

December, 2013

Chair: Susmita Bose

The objective of this research is to further develop the understanding of trace elements in bone biology and, specifically, their potential osteogenic effects in calcium phosphate bone substitute ceramics. Trace elements including magnesium (Mg2+), silicon (Si4+), strontium (Sr2+) and zinc (Zn2+) were incorporated into β tricalcium phosphate (β –TCP) as

dopants into various form factors including dense ceramic compacts, 3D printed scaffolds and

porous scaffolds fabricated by an oil emulsion method. The hypothesis of this research is that the

addition of these trace element dopants into β –TCP can significantly alter its phase stability,

microstructure, mechanical strength and in vitro and in vivo biocompatibility.

3D printed scaffolds containing 0.5% SiO2 and 0.25% ZnO increased the average density

of pure TCP from 90.8 ± 0.8% to 94.1 ± 1.6% which resulted in an average 2.5 fold increase in

compressive strength. Scaffolds that contained the smallest amount of designed porosity had

compressive strengths of 5.48 ± 0.04 MPa and 10.21 ± 0.33 for pure and doped scaffolds,

respectively. Doped samples demonstrated increased cellular proliferation. In vivo results in a

v murine model demonstrated that the presence of SiO2 and ZnO increased osteocalcin and collagen production as well as increased osteogenesis and angiogenesis over the course of 16 weeks.

Further studies examining 1% SrO, 1%MgO, 0.5% SiO2 and 0.25% ZnO single dopant

TCP systems on osteoblastic differentiation markers alkaline phosphatase (ALP) and runt related transcription factor 2 (Runx2) demonstrated the ability of dopants to increase cellular proliferation in the early stages of the osteoblastic lifecycle, while down regulating the Runx2 expression at later time points, allowing for faster terminal differentiation. SiO2 and ZnO dopants were also analyzed using quantitative polymerase chain reactions for the following targets in osteoblast cells: bone morphogenic protein 2 (BMP2), Runx2, receptor activator of nuclear factor kappa-B ligand (RANKL) and osteoprotegerin. At day 21, all doped samples expressed 2-4 times less BMP2, 1.5-2.5 times greater OPG and 2-4 times less RANKL when compared to pure TCP.

Mg and Sr doped samples expressed 2.5 times more Runx2, Si 1.5 times more and zinc similar amounts of Runx2 to pure TCP. These results demonstrate an affinity for the production of signaling molecules that favor increased osteoblastogenesis and decreased osteoclastogenesis.

Contents Acknowledgements ...... iii Abstract ...... v List of Tables ...... x List of Figures ...... xi 1. CHAPTER ONE ...... 1 Introduction ...... 1 1.1 The biology of bone ...... 1 1.2 Role of bone substitutes ...... 11 1.3 Porous calcium phosphates ...... 15 1.5 Scaffold fabrication techniques...... 16 1.6 The importance of trace elements in bone biology and their role in calcium phosphates ...... 22 1.7 Bioreactors for cellular population in scaffolds ...... 29 2. CHAPTER TWO ...... 35

Effects of SiO2 and ZnO doping on mechanical and biological properties of 3D printed TCP scaffolds ...... 35 2.1 Introduction ...... 35 2.2 Materials and Methods ...... 37 2.3 Results ...... 42 2.4 Discussion ...... 49 Conclusion ...... 53 3.CHAPTER THREE ...... 54

SiO2 and ZnO Dopants in 3D Printed TCP Scaffolds Enhances Osteogenesis and Angiogenesis in vivo ...... 54 3.1. Introduction ...... 54 3.2.0 Materials and Methods ...... 57 3.3.0 Results ...... 64 3.4.0 Discussion ...... 72 3.5.0 Conclusion ...... 78

vii

4. CHAPTER FOUR ...... 79

Effects of SiO2, SrO, MgO and ZnO dopants in TCP on osteoblastic Runx2 expression...... 79 4.1. Introduction ...... 79 4.2. Methods...... 82 4.3. Results ...... 86 4.4. Discussion ...... 91 4.5. Conclusions ...... 98 5. CHAPTER FIVE ...... 100

Effects of SiO2 and ZnO dopants in porous TCP scaffolds on osteoblastic differentiation .... 100 5.1 Introduction ...... 100 5.2 Methods...... 103 5.3 Results ...... 107 5.4 Discussion ...... 115 5.5 Conclusions ...... 123 6. CHAPTER SIX ...... 124

Effects of SiO2 and ZnO binary dopant system on osteoblasts in porous TCP scaffolds cultured in a flow perfusion bioreactor...... 124 6.1 Introduction ...... 124 6.2 Methods...... 126 6.3 Results ...... 132 6.4 Discussion ...... 144 6.5 Conclusions ...... 149 Summary ...... 151 References ...... 154 Appendix A ...... 179 Journal Publications ...... 179 To Be Submitted ...... 180 Conference Proceedings...... 180 Presentations ...... 180

viii

Appendix B ...... 182 MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) Calculations ...... 182 Alamar Blue Assay ...... 184 In Vivo Sample Preparation ...... 186 TRAP Staining Staining of Decalcified Tissue Sections ...... 194 Antibody Staining Protocol (Confocal or Light) ...... 196 Histomorphometry ...... 202 AAS Ion Concentration Determination ...... 207 Fluorescent Area Scan Calculations ...... 211 qRT-PCR Analysis...... 214 Flow Simulation Analysis from SolidWorks ...... 218

List of Tables

Table 1.1. Osteoclastic resorption cycle [19]. Table 1.2. Types of biomaterials, characteristics and applications [61]. Table 1.3. Properties of different calcium phosphates [62]. Table 1.4 Effects of Dopants on Mechanical Propties of TCP [104–107].

Table 2.1. A list of important 3D printing parameters for fabrication of CaP scaffolds.

Table 2.2. Percent designed porosity and density of 3D printed scaffolds of pure and doped composition.

Table 2.3. Measurements of post sintered pore size in pure and SiO2/ZnO doped TCP scaffolds.

Table 4.1. Relative densities of sintered samples.

List of Figures

Figure 1.1. Structure of mineralized bone [3]. Figure 1.2. (a) Diagrammatic view of a pie-shaped segment of compact bone. (b) Close-up of a portion of one osteon. Note the position of osteocytes in the lacunae. (c) Photomicrograph of a cross-sectional view of an osteon (140×) [3]. Figure 1.3. The osteon is drawn as if pulled out like a telescope to illustrate the individual lamellae. The slanted lines in each lamella indicate the direction of its collagen fibers [3]. Figure 1.4. Figure demonstrating relationship between osteoclasts, osteoblasts, lining cells and osteocytes [16]. Figure 1.5. Typical μCT scans of blocks created using varying concentrations of Kolliphor EL emulsifier and paraffin [82].

Figure 1.6. Fused deposition modeling: 1 – nozzle ejecting molten plastic, 2 – deposited material (modeled part), 3 – build bed. [83].

Figure 1.7. Various constructs created using FDM. [84,85]. Figure 1.8. Selective laser sintering process [86]. Figure 1.9. (a) Schematic diagram of the scaffold model in diametric view; (b) side view of the scaffold model; (c) scaffolds produced by SLS of varying compositions; (d) MicroCT image of composite scaffold [87,88]. Figure 1.10. (a) The 3DP process; (b) a CAD file of a designed scaffold; (c) final sintered scaffolds. Figure 1.11. General and detailed view of the implant bearing skull. Implants created using a 3DP system [89].

Figure 1.12. Typical spinner flask bioreactor Figure 1.13. Schematic of a rotating wall bioreactor. Outside wall rotates to circulate media [142]. Figure 1.14. Schematic of perfusion bioreactor. Media is directly perfused through porous scaffold sealed into a growth chamber.

Figure 2.1. CAD image of a cross section of a porous scaffold. Square channels are oriented 0˚/90˚ for subsequent layers.

Figure 2.2. Schematic drawing representing the 3D printing process.

Figure 2.3. XRD patterns of doped and pure scaffolds sintered at 1250 ˚C. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP). Figure 2.4. Sintered SiO2/ZnO doped and pure scaffolds of each designed pore volume. Figure 2.5. Surface morphology of sintered scaffolds of (a) pure TCP composition and (b) SiO2/ZnO doped composition.

Figure 2.6: Compressive strengths of sintered 3D printed scaffolds for both pure and doped β- TCP. Statistical analysis shows that the differences are significant (* P << 0.05, where n = 10).

Figure 2.7. MTT assay of hFOB on TCP scaffolds with 500 μm, 750 μm, and 1000 μm 3D interconnected channels after 3, 7, and 11 days. Data is normalized to the pure TCP controls for each channel size and day point. (*p < 0.05, n = 3).

Figure 2.8. SEM micrographs of hFOB cells showing the cell adhesion morphology on the scaffold surface and inside the 3D interconnected macropores: (a) doped TCP at 3 days, (b) doped TCP at 7 days, (c) pure TCP at 3 days, (d) pure TCP at 7 days. Arrows are indicating hFOB cells. Figure 3.1: (a) Rendered cross section of typical CAD file used for the 3D printing process. (b) Schematic diagram demonstrating 3D printing process. Large arrows indicate moving direction. (c) Actual scaffolds created by the Ex One printer. Numbers below the scaffold represent designed pore size.

Figure 3.2. (a) Implant sections stained using Goldner’s trichrome. The CaP implants are identified by gray/brown color, mineralized implants by blue and osteoid formation by orange. (b) Histomorphometry performed on trichrome sections. (* P < 0.1, where n = 3.)

Figure 3.3. (a) Confocal micrographs showing collagen I formation in sectioned implants over the course of 16 weeks. Green indicates collgen I, while blue indicates counterstain for cell nuclei. (b) Histomorphometry for collagen I labeled sections. (* P < 0.1, where n = 3)

Figure 3.4. (a) Confocal micrographs showing osteocalcin formation in sectioned implants over the course of 16 weeks. Green indicates osteocalcin, while blue indicates counterstain for cell nuclei. (b) Histomorphometry for osteocalcin labeled sections. (* P < 0.1, where n = 3)

Figure 3.5. Cartoon depicting effect of dopants on collagens and osteocalcin production. Zn2+ and Si4+ ions increase osteoblastic collagen matrix production. Increased collagen matrix leads to increased binding of osteoblasts to matrix integrins which activates the MAPK pathway. The MAPK pathways can eventually lead to the upregulation of osteocalcin expression.

xii

Figure 3.6. BSE imaging of longitudinal cross sections of samples after 6, 8 and 12 weeks implantation. The lighter color indicates the calcium phosphate scaffold, while the darker areas indicate mineralized tissue.

Figure 3.7. (a) Light micrographs depicting vWF staining. The dark red spots within the sections are blood vessel. (b) Graph showing blood vessel density for each group over the course of 16 weeks. (* P < 0.1, where n = 3)

Figure 3.8. FESEM micrographs showing blood vessel formation. Arrows indicate blood vessels. Dotted lines show vascular branching pathways formed in the samples. Figure 3.9. TRAP staining in sectioned samples over the course of 16 weeks. The red color is indicative of TRAP activity and the green counterstain shows the tissue/scaffold.

Figure 3.10. AAS results for ion concentration in urine collected over 16 weeks. Figure 4.1. MTT measurements for cellular proliferation and ALP measurements for cellular differentiation at day 7 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. ALP data is normalized to MTT measurements. *P < 0.05 (n=3) Figure 4.2. MTT measurements for cellular proliferation and ALP measurements for cellular differentiation at day 11 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. ALP data is normalized to MTT measurements. *P < 0.05 (n=3) Figure 4.3. FESEM micrographs depicting hFOB cell morphology after 3, 7 and 11 days in culture Figure 4.4. MTT measurements for cellular proliferation and Runx2 fluorescent measurements for cellular differentiation at day 3 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. Runx2 data is normalized to MTT measurements. *P < 0.05 (n=3) Figure 4.5. MTT measurements for cellular proliferation and Runx2 fluorescent measurements for cellular differentiation at day 11 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. Runx2 data is normalized to MTT measurements. *P < 0.05 (n=3) Figure 4.6. Confocal micrographs showing Runx2 expression in hFOB cells at 3 and 11 days. Green fluorescence indicates active Runx2; red fluorescence indicates propidium iodide bound to the nuclei of cells.

Figure 4.7. Diagram depicting the osteoblast cell lifecycle and influence of Runx2 [228,229].

xiii

Figure 4.8. Activation of the Wnt/Fz ligand-receptor leads to the production of the second messengers IP3 and DAG from membrane-bound PIP2 via the action of membrane-bound enzyme PLC. IP3 causes release of Ca2 from the ER and extracellular Ca2+ influx through transmembrane Ca2+ channels; CaN and CamKII are activated which in turn activate NFAT and NFkB. DAG is also activated by increased inctracellular Ca2+, which activates PKC. PKC activates NFkB and CREB. NFAT, NFkB, and CREB translocate to the nucleus and transcribe downstream regulatory genes such as Runx2.

Figure 5.1. XRD patterns of doped and pure scaffolds sintered at 1250 ˚C. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP). Figure 5.2. Bulk density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3).

Figure 5.3. Apparent density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3). Figure 5.4. Compressive strength of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 5).

Figure 5.5. Surface morphology of sintered scaffolds of pure TCP composition, Si-TCP and Zn-TCP compositions after sintering and after 28 days in PBS

Figure 5.6. AAS results for Ca2+ ion concentration in PBS collected over 28 days. Figure 5.7. FESEM micrographs depicting hFOB cell morphology after 3, 7 and 11 days in culture.

Figure 5.8. RT-qPCR data for day 21 (* P < 0.05, where n = 3; **P<<0.01). Figure 5.9. RT-qPCR data for day 28 (* P < 0.05, where n = 3; **P<<0.01). Figure 5.10. Activation of the Wnt/Fz ligand-receptor leads to the production of the second messengers IP3 and DAG from membrane-bound PIP2 via the action of membrane-bound enzyme PLC. IP3 causes release of Ca2 from the ER and extracellular Ca2+ influx through transmembrane Ca2+ channels; CaN and CamKII are activated which in turn activate NFAT and NFkB. DAG is also activated by increased inctracellular Ca2+, which activates PKC. PKC activates NFkB and CREB. NFAT, NFkB, and CREB translocate to the nucleus and transcribe downstream regulatory genes such as Runx2. Elevated Runx2 activity increases OPG production while simultaneously decreasing RANKL production. BMP2, upregulated by extracellular Ca2+, can affect this pathway by binding to BMP2 receptors which activate PI3k. PI3k activates the gamma subunit of PLC increasing its conversion of IP3 and DAG.

Figure 6.1. CAD drawing of single bioreactor chamber containing sample holder and sample.

xiv

Figure 6.2. XRD patterns of doped and pure scaffolds sintered at 1250 ˚C. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP). Figure 6.3. Bulk density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3). Figure 6.4. Bulk density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3).

Figure 6.5. Surface morphology of sintered scaffolds of pure TCP composition and SiO2/ZnO doped composition before and after 28 days exposure in PBS.

Figure 6.6. AAS results for Ca2+ ion accumulation in PBS over 28 days.

Figure 6.7. Results of simulated shear stress in bioreactor on sample surface. Blue indicates low shear stress, while color gradient up to red indicates increasing shear stresses.

Figure 6.8. Results of simulated pressure in bioreactor on sample surface. Blue indicates low pressure, while color gradient up to red indicates increasing pressures.

Figure 6.9. Results of simulated fluid flow in bioreactor. Blue indicates low fluid velocity, while color gradient up to red indicates increasing velocities.

Figure 6.10. Alamar Blue measurements for cellular proliferation at days 3,7 and 14 for pure and Si/Zn-TCP samples in both culture conditions. *P < 0.05 (n=3) Figure 6.11. Cell morphologies for samples after 14 days of culture. SC indicates standard culture while BR indicates bioreactor. Arrows in SC indicate single cells, while arrows in BR indicate sheet-like cell formations.

Figure 6.12. ALP measurements for cellular differentiation for days 3, 7 and 14. Data was adjusted to reflect differences in cellular proliferation as obtained by the Alamar Blue assay. *P < 0.05 (n=2). Figure 6.13. Confocal micrographs showing ALP expression in hFOB cells at 3, 7 and 14 days. Green fluorescence indicates active ALP; blue fluorescence indicates DAPI bound to the nuclei of cells.

1. CHAPTER ONE

Introduction

1.1 The biology of bone

At the surface, many understand bone to be a simple rigid organ that provides support for movement as well as protection of vital organs. While both of these functions are an important aspect of bone, the tissue is actually an incredibly complex system that plays a vital role in human health ranging from red and white blood cell production to the storage of essential minerals. Mineralized bone tissue can be divided into two distinct types: 1.) Dense cortical bone that generally comprises the outer shell of bone and 2.) Spongy cancellous/trabecular bone as an inner layer (Figure 1.1). Each has their own distinct role and function.

Cancellous bone has a honeycombed architecture that is composed of small needle-like, flat pieces called trabeculae. In living tissues, the open gaps that the trabeculae form are home to either red or yellow bone marrow. Red marrow is the factory where most of the body’s red blood cells, platelets and white blood cells are produced [1]. In adults, most of the red marrow is found in flat bones such as the pelvis, sternum, cranium and vertebrae as well as the epiphyseal ends of the long bones such as the femur and humerus. Yellow bone marrow consists mainly of adipose, or fatty, cells and acts as a natural reserve chamber for energy. In extreme hunger conditions, the body will use the fatty tissue in the yellow marrow as an emergency energy

1 reserve. Also, in trauma cases, where there is a large amount of blood loss, the yellow marrow will convert to red marrow to produce more blood cells [2].

Figure 1.1. Structure of mineralized bone [3].

Cortical bone, at initial glance, gives the appearance of being dense, solid and smooth. A microscopic look, however, reveals a complex system of passageways that serve as channels nerves, blood vessels and lymphatic vessels (Figure 1.2). Each of the individual structural units are synonymously referred to as either Haversian systems or osteons. Each Haversian system is roughly 400 mm long and 200 mm wide and form lateral branching called Volkmann’s canals

[4]. Surrounding the central canal are the lamella (Figure 1.3), which consist of bone matrix and single directional collagen fibers. The collagen fibers in adjacent lamellar layers run in opposite directions and are a perfect design to withstand torsional stresses. Each osteon usually is comprised of 4-20 concentric lamellar rings[1,3]. Between lamellae, are a number of oblong

2 spaces called lacunae that are home to osteocytes. Osteocytes are connected to each other between lacunae through tiny channels called canaliculi.

Figure 1.2. (a) Diagrammatic view of a pie-shaped segment of compact bone. (b) Close-up of a portion of one osteon. Note the position of osteocytes in the lacunae. (c) Photomicrograph of a cross-sectional view of an osteon (140×) [3].

Figure 1.3. The osteon is drawn as if pulled out like a telescope to illustrate the individual lamellae. The slanted lines in each lamella indicate the direction of its collagen fibers [3]. The outside of the cortical bone is sheathed in a fibrous connective tissue called the periosteum. The periosteum is tightly bound to the bone via thick collagenous fibers, called

Sharpey’s fibers, which extend into the underlying bone tissue [5]. The periosteum contains blood vessels, nerve fibers, osteoblasts and osteoclasts. The inside of the cortical tissue as well as the central Haversian canals, Volkmann’s canals, and cancellous bone are covered in the endosteum. The endosteum also contains blood vessels, osteoblasts and osteoclasts. The periosteal surface is known to function in appositional growth and fracture repair and is characteristically defined by an affinity for new bone formation, rather than bone resorption, so bones normally increase in diameter with aging [5,6]. The endosteal surface has an increased rate of remodeling, likely due to greater biomechanical strain and greater cytokine exposure from the adjacent bone marrow compartment [5,7].

There are three major types of cells that are responsible for the building and remodeling of bone: osteoblasts, osteocytes and osteoclasts. They each have specific functions in the homeostasis of bone tissue. If one type of cell is not operating as needed, several disease states may arise including osteoporosis, metastatic bone disease and Paget's disease. It is important, then, to understand the basic healthy physiological behavior of these cells.

Osteoblasts are single nucleated cells that originate from osteoprogenitor cells such as mesenchymal stem cells (MSCs). MSCs also have the ability to differentiate into chondrocytes as well as adipocytes and are of particular importance because they can be fairly easily harvested from a live host and manipulated into differentiating into a patient-specific osteoblast cell [8]. Preosteoblastic cells demonstrate a phenotypical spindle shape. As preosteoblastic proliferation ceases, differentiation occurs and the cell morphs into a phenotypically osteoblastic cuboidal shape with many filopodia for mobility and communication. There are two primary functions of the osteoblast cell in bone. First, is to lay the foundation for new bone by manufacturing and deposition of the largely collagen based extracellular matrix. The release other non-collagenous proteins and enzymes that help in the mineralization of the ECM such as alkaline phosphatase (ALP), osteopontin (OPN), osteocalcin (OCN) and bone sialoprotein

(BSP) [9]. Secondly, osteoblasts are needed for the differentiation of osteoclasts, so they play a vital role in bone resorption (Figure 1.4). Osteoblasts secrete receptor activator of nuclear factor-κB ligand (RANKL), which will bind to receptor activator of NF-κB (RANK) receptors on pre-osteoclasts and stimulate monocyte differentiation into multinucleated osteoclast cells that can function in the bone remodeling process. This process can be further modulated due to the production of osteoprotegerin (OPG) by osteoblasts. OPG is a decoy receptor for RANKL

5 and can decrease the amount of free RANKL to bind to RANK on pre-osteoclastic cells, therefore stifling the differentiation process [10].

At the end of the mineralization phase of the osteoblastic life cycle, there are three terminal directions for the cells. First they can differentiate into bone lining cells. Bone lining cells have a flattened morphology and are generally considered inert in function. They are thought to form the form the endosteum on trabecular and endosteal surfaces and underlie the periosteum on the mineralized surface [5]. Functionally, the may play a role in the regulation of mineral ions in and out of the bone extracellular fluid, acting as a blood-bone barrier, and are able to dedifferentiate into osteoblasts upon exposure to parathyroid hormone or mechanical stresses [11]. Second, they can undergo apoptosis, or programmed cellular death. The third option is to differentiation into an osteocyte. Osteocytes are settled in bone lacunae and maintain a complex signaling network through filipodial processes though the canaliculi. The primary function of the syncytium of the osteoblast-bone lining cell- osteocyte system is mechanosensation [12]. Osteocytes can transduce mechanical stress signals into biological activities and are known to play a role in several signaling mechanisms including prostaglandin

E2, cyclo-oxygenase 2, Runx2 and nitrous oxide [13–15].

Figure 1.4. Figure demonstrating relationship between osteoclasts, osteoblasts, lining cells and osteocytes [16]. Osteoclasts are large, multinucleated cells that are formed via fusion of their monocyte precursors [17]. They differentiate from hematopoietic stem cells in the bone marrow. As discussed previously, they rely on osteoblastic production of RANKL to differentiate from their monocyte phase to mature osteoclast. Although it is still unknown what causes the recruitment of osteoclasts and the beginning of the remodeling process, some believe that it is triggered by the formation of microcracks in the lacunae [18]. Microdamage in the lacunae may trigger an alarm response by osteocytes, which encourage osteoblasts to produce more RANKL and less

OPG, resulting in the stimulation of osteoclastogenesis. Once mature osteoclasts are recruited to the resorption site, there is a very specific process that they undergo (Table 1.1). Briefly,

7 podosomes attach to the resorption area and form a tight bond. Once the bond is formed, the area under the sealing zone undergoes an acidic reaction that begins the dissolution of the apatite crystals. Vesicles remove dissolution products via a transcytotic process. Ultimately, the resorption activity is thought to cease when a certain concentration of Ca2+ in the microenvironment is reached and the osteoclasts detach and go through an apoptosis process or move to another resorption site.

Table 1.1. Osteoclastic resorption cycle [19]

The bone remodeling process, although very complex, can be broken down into eight distinct stages: origination, osteoclast recruitment, resorption, osteoblast recruitment, osteoid formation, mineralization, mineral maturation and quiescence. Each state utilizes a variety of important signaling mechanisms that are of interest to researchers. A more complete understanding of these cascades can lead to better treatments for degenerative bone disorders.

The origination phase is thought to take place following microdamage caused by mechanical stresses or a trigger event to satisfy calcium homeostasis [20]. In either case, signals from osteocytes are secreted causing the formation of basic multicellular unit (BMU) consisting of osteoblasts, osteoclasts, bone lining cells and osteocytes. The bone lining cells form a circulatory canopy or microenvironment for the remodeling to take place [21]. It has been observed that during this process there is an increase in parathyroid hormone (PTH), insulin like growth factor (IGF), interleukins 1 and 6, prostaglandin E2 (PGE), calcitriol, tumor necrosis factor (TNF) and nitric oxide (NO) and a decrease in estrogen [22–27].

During the osteoclast recruitment stage, stromal cells reacting to signals in the origination phase will begin producing macrophage colony stimulating factor (M-CSF) and will also divide to produce preosteoblastic cells that produce RANKL on their surface [28,29]. When

RANKL activates the RANK receptors, the preosteoblast cell will fuse to form multinucleated osteoclasts. In this phase, RAKL and M-CSF production is upregulated, while OPG and granulocyte-macrophage colony-stimulating factor (GM-CSF), a molecule known to inhibit osteoclastogenesis, are downregulated [22,30].

In the resorption phase, the mature osteoclasts attach to the matrix surface and begin to secrete hydrogen ions and cathepsin. The BMU will roam around the resorption site

9 continuously forming new osteoclast and will continue to undergo the resorption process [31].

Apoptosis is delayed by decreased estrogen levels [32]. This phase is characterized by low levels of estrogen, calcitonin, interferon, TGF and increased integrin levels in osteoclasts [33–

35].

During the osteoblast recruitment phase, preosteoblastic cells differentiate in the presence of bone-derived growth factors such as BMPs, insulin like growth factors (IGFs) and

Platelet-derived growth factors (PDGFs) [36–38]. At this point, upregulation of the runt related transcription factor (Runx2) in osteoblasts is absolutely necessary and is likely accomplished via the Wnt signaling pathway, which plays several roles in this process (Figure 1.4) [16,39,40].

Osteocytes also decrease the production of sclerostin during this phase, which is an inhibitor of the Wnt signaling mechanism [41]. During this phase, Wnt, BMPs, FGFs, IGFs, PTH and

Runx2 are all upregulated and are considered very important in initial osteoblastic differentiation [42–44].

Osteoid formation phase is characterized by mature osteoblasts producing ECM and filling in the resorption areas. During this stage osteoblasts secrete growth factors, osteopontin, osteocalcin, osteonectin and other matrix maturation proteins [45–47]. In this phase, FGFs and

PDGFs are decreased, while TGFs, BMPs and IGFs remain increased. [48,49].

The mineralization phase is initiated when the osteoid is roughly 6 µm thick [50].

Osteoblast phosphate metabolism is seemingly regulated via phosphate regulating endopeptidase homolog (PHEX) and FGF-23, although the mechanisms are still not well understood [51].

During this phase, ALP is continued to be produced by osteoblasts. ALP will catalyze the

10 hydrolysis of phosphomonoesters into inorganic phosphates, that will reprecipitate to the surface of the matrix actually resulting in the mineralization process [52].

The mineral maturation phase is a long phase where over the course of months, the apatite crystals continue to be packed more closely and the density of the new bone will continue to increase [53,54].

The final step is quiescence of the osteoblast cells. This step is probably the most poorly understood of all of the stages. It is known, as mentioned previously, that at this point osteoblasts have three options: differentiation into bone lining cells, differentiation into osteocytes, or apoptosis. This entire process, at any given point on the bone, is thought to be repeated around once every two years [55].

1.2 Role of bone substitutes

Musculoskeletal impairments have been ranked number one in the United States for chronic impairments and surveys from around the globe indicate that one in four people suffer from musculoskeletal pain [56,57]. The World Health Organization and the United Nations dubbed the decade of 2000-2010 as the “bone and joint” decade with the specific goal of raising awareness and improving the quality of life of people with musculoskeletal conditions. In the

US, it is estimated that a $250 billion annual cost can be associated with various musculoskeletal disorders arising from disability, need for assistance and reduced independence, out of pocket medical costs and loss of potential earnings [58].

Some of the key ailments that are often treated in patients include osteoporosis, osteoarthritis, age-related spinal stenosis and herniated disks. All of these conditions are most

11 common in the aging population. It is estimated that the number of people over the age of 50 will double by the year 2020 according to the US National Institute on Aging, so there is a pressing need for safe, reliable and cost effective treatment options. With spinal fusion surgeries and joint replacement surgeries as well as tumor and fracture treatment, bone grafting materials are often used to fill bony voids. Until a few years ago, autografts and allografts were most commonly used to fill these bone defects. Autografts, while an excellent natural material, require that a second surgery be performed to remove bone from a donor site, usually on the iliac crest. In addition to the added cost of surgery and increased pain at the harvest site, there may also be complications causing donor site morbidity [59]. Allografts are materials taken from a cadaver source. While these materials have also demonstrated excellent results in bone defect and fracture nonunion healing, personal reservations, and source variability may turn surgeons and patients to seek alternative sources [60]. Research and clinical data have demonstrated that synthetics are a viable alternative to autografts and allografts.

Vast arrays of synthetic biomaterials are used in a wide variety of applications throughout the human body in medicine ranging from metals and plastics to ceramics (Table

1.2). Of particular interest in bone defect repair are calcium phosphate (CaP) ceramics. The inorganic phase of bone is typically made up of around 70% biological apatite, which synthetic

CaPs have a compositional similarity to. Depending on several factors including the presence of water, the Ca:P ratio and crystallinity, CaPs exist in many forms (Table 1.3). Not all CaPs are bioactive, however. Monocalcium phosphate monohydrate (MCPM) and its amorphous form monocalcium phosphate (MCP) are too acidic and soluble to be useful in bone applications. The most commonly used and studies CaPs for bone implant applications are tetracalcium phosphate

(TTCP), tricalcium phosphate (TCP) and hydroxyapatite (HA). These have all demonstrated excellent biocompatibility and good host-implant tissue integration. Because they are all soluble at biological conditions, these ceramics are dubbed resorbable.

Table 1.2. Types of biomaterials, characteristics and applications [61]

Table 1.3. Properties of different calcium phosphates [62]

HA and TTCP have extremely low solubilities and are believed to be resorbed over the course of years, while TCP may be resorbed over the course of several months [63,64]. Although, biphasic mixtures of TCP and HA have been investigated as bone grafting materials, it is generally believed that TCP sits in the “Goldilocks” zone, where the implant is degrading at a comparable rate as new bone is being formed at most sites.

TCP has three polymorphs: α-TCP, α'-TCP and β-TCP. β-TCP transforms to α-TCP at temperatures greater than 1125 ˚C. Studies have demonstrated that this phase transformation is closely related to the expansion of sample volume, declining shrinkage rate and is generally considered to prevent TCP from further densification [65,66]. The theoretical density of α-TCP

(2.83 g/cm3) is lower, then, than that of β-TCP (3.07 g/cm3) and also has a higher solubility than

β-TCP. Due to these differences, α-TCP has a resorption rate of several weeks, but it is still used and studied considerably as a bone replacement material. α'-TCP is not stable at temperatures below 1430 ˚C and will transform back to α-TCP.

1.3 Porous calcium phosphates

TCP commercial products come in a variety of forms including pastes, putties, granules, and blocks. One of most critical and most researched aspects of these forms is the effects porosity. The amount, size, morphology and interconnectivity play extremely important roles not only on the physical aspects of the implant, but on the biological. New bone growth favors high porosity as the vacant spaces allow bone to colonize immediately, without going through the resorption process [67]. A porous surface can also play a role in interdigitation or mechanical interlocking between the implant and host tissue [68]. Having an interconnected porous structure facilitates the introduction and migration of new bone forming cells, naturally produced osteotropic agents and vascularization, allowing faster bone growth [69]. As porosity increases, however, there is a significant tradeoff in mechanical strength. Calcium phosphate ceramics exhibit fairly weak mechanical strength and are susceptible to crack propagation under cyclic conditions [70], so there is a delicate balance that has to be satisfied. It has been demonstrated that microporosity (<5 μm pores) as well as macroporosity (>100 μm pores) can both play an important role in the healing process [71,72]. Microporosity, due to its higher surface area, is believed to be a contributing factor in bone forming protein adsorption as well as ion exchange and apatite formation [73]. In addition, microporosity creates surface roughness that has demonstrated increased cellular attachment, proliferation and differentiation [74]. The minimum effective macropore size seems to be around 100 μm, but as pore size increases to between 300um - 1000μm, the trend favors direct osteogenesis, since they allow vascularization and high oxygenation [75–77].

Pore interconnectivity is defined as the amount of open pathways between pores in a construct. Properties that seem to have some importance are the throat size, throat length and overall interconnectedness. Maximum tissue formation has been noted if pore throat sizes are greater than 100 μm and throat length as short as possible [78]. In addition to new bone tissue formation, interconnected pores have a minimum threshold permeability for vascularization and

11 2 mineralization to occur within an implant. This threshold ( ~ 3 x 10 m ) can be achieved with a minimum throat size of about 50 μm [78,79].

1.5 Scaffold fabrication techniques

Various fabrication techniques have been developed to make porous architectures. Most standard methods utilize some sort of sacrificial material or gas releasing agent to create a pore shape. Porogens in the form of sucrose, naphthalene and Poly(D,L-lactic-co-glycolic acid)

(PLGA), among others, are incorporated into the CaP powder before processing to final shapes

[80,81]. Porogens will sublimate during setting or be burned out during sintering leaving the overall porous design. These methods are often commercially used in synthetic grafting material due to their scalability and somewhat predictable behavior and reproducible results. Figure 1.5 demonstrates a method of producing various pore sizes using a single oil emulsion method with varying concentrations of emulsifier.

Figure 1.5. Typical μCT scans of blocks created using varying concentrations of Kolliphor EL emulsifier and paraffin [82].

While these standard methods of producing porous scaffolds are currently in wide use, there has been an explosion of interest in newer, cutting edge techniques utilizing solid freeform fabrication (SFF) techniques. SFF offers the promise of complete control of macropore design as well as potential in patient specific applications, where standard materials will simply not work. The most widely used SFF methods for CaPs are fused deposition modeling (FDM), selective laser sintering (SLS) and three-dimensional printing (3DP). All SFF techniques essentially start the same way. A computer aided design (CAD) file is created using commercially available software and the file is converted to a stereolithography (STL) file,

17 where only the surface geometry information is kept. This file is then loaded into the SFF system where it is sliced into many layers and processed.

With fused deposition modeling (Figure 1.6), a plastic filament is fed into an extrusion head, where the feed wire is melted and deposited, layer by layer, onto the surface according to the CAD geometry input. After each layer is printed, a CaP slurry is made and poured into the mold. During sintering the plastic mold is burned off and the remaining CaP slurry takes on the inverse shape of the sacrificial support structure. Figure 1.7 shows some examples of the complexity of geometry achieved using this method.

Figure 1.6. Fused deposition modeling: 1 – nozzle ejecting molten plastic, 2 – deposited material (modeled part), 3 – build bed [83].

Figure 1.7. Various constructs created using FDM [84,85].

SLS is a method where a feed bed and a build bed are utilized to print a CAD file (see

Figure 1.8). A laser is used (usually a CO2 laser) to sinter loose powder particles together in a layer by layer pattern designated by the CAD file design. The feed bed provides fresh loose powder, usually via a roller, to the build bed. After one layer is sintered, the build

Figure 1.8. Selective laser sintering process [86]. layer will decrease by one layer thickness (varies) and a spreader will roll new powder from the feed bed onto the build platform. This layer by layer approach is repeated until the CAD file is completely printed. This process is often used to create plastic-ceramic composites shown in

Figure 1.9 [87,88].

Figure 1.9. (a) Schematic diagram of the scaffold model in diametric view; (b) side view of the scaffold model; (c) scaffolds produced by SLS of varying compositions; (d) MicroCT image of composite scaffold [87,88].

Similarly to SLS, the 3DP process generally consists of a build bed and a feed bed. In place of the laser sintering source, however, is a printer head much like would be found in your home printer. The printing process is also a layer by layer build method, where the printer head sprays a pattern on binder onto fresh powder in the build bed. After this layer is “printed”, the build bed lowers by one layer and the feed bed is raised up and new powder spread onto the build bed (Figure 1.10).

Figure 1.10. (a) The 3DP process; (b) a CAD file of a designed scaffold; (c) final sintered scaffolds.

The 3DP process is probably the most researched in the bone grafting space due to its relative ease of use and lack expensive parts. The process has been investigated in the manufacturing of patient specific implants for craniomaxillofacial applications with a high degree of success. This is an area of interest because the skull features curves that are not easily replicated or reproduced during conventional machining processes. In a study by Klammert et. al., they produced patient specific CaP pieces for implantation in a human skull for various defects [89]. This study demonstrated a high degree of success using this method as seen in

Figure 1.11.

Figure 1.11. General and detailed view of the implant bearing skull. Implants created using a 3DP system [89].

1.6 The importance of trace elements in bone biology and their role in calcium phosphates

Recent trends in technology focus on the incorporation of materials that extends the performance of CaPs to have osteoinductive properties, or the ability to actually stimulate new bone formation. It is important to recognize that the ability to stimulate osteogenesis, or bone growth, has to also couple with angiogenesis, or blood vessel formation. The two processes are intricately linked and osteogenesis would not be possible without angiogenesis [90,91]. The first commercial synthetic product to offer the promise of osteoinductivity and angiogenesis was

Metronic’s recombinant human bone morphogenic protein-2 (rhBMP-2) Infuse product. While fantastic results have been achieved in many applications ranging from lumbar (lower) spinal fusion, craniomaxilofacial and dental applications and long bone grafting applications, the

22 apparent off label use in cervical (upper) spinal fusion applications has led to troubling issues regarding the safety and efficacy of the product. rhBMP is a growth factor that has been proven to induce bone and cartilage growth [92,93]. The major concerns have related to ectopic, or unwanted bone formation where in certain situations can lead to the very serious side effects

[94–96]. The clinical successes of rhBMP products have led to a flurry of research into the incorporation of other potentially beneficial growth hormones and biologics into orthopedic materials. Compounds of interest include transforming growth factor β (TGF-β), fibroblast growth factor (FGF), insulin like growth factor (IGF), vascular endothelial growth factor

(VEGF) and a variety of bisphosphonates [62]. With all of the negative attention drawn to these biological compounds, however, the FDA has taken a harder stance on the approval of such devices making commercial viability much more difficult and expensive.

An alternative and potentially safer strategy has been to look past the addition of biologics and investigate a more natural approach. The major CaP component of bone is not a homogenous material. Bone incorporates various nutrients in the form of trace elements. These trace elements have been found to play absolutely vital roles in the formation, growth and repair of bone. Various studies have also demonstrated that the addition of trace elements to CaP materials can lend controlled degradation, increase the mechanical strength of the materials and positively influence the biological response [97–100]. Table 1.4 demonstrates the changes found in mechanical strength and strength degradation properties of various dopants and combinations. Much research has gone into the study of how additives affect the physicochemical materials properties of calcium phosphates. Silicon is an additive that has been used extensively to improve the densification of calcium phosphates during sintering. It does

23 this by impeding the phase transformation of of β-TCP to α-TCP at temperatures over 1050˚C

[101]. α-TCP is known to be a phase that is associated rapid grain growth and has an adverse effect on densification. Other additives have demonstrated similar behaviors, but also can have adverse effects at increasing weight percents and combinations [97]. The balance, then, has to be an additive that can be incorporated in sufficient amounts to achieve increased biological performance, while maintaining the physical integrity of the implant initially as well as the degradation kinetics over extended periods of time.

Because these dopants are incorporated directly into the CaP material and CaPs are biodegradable, there is a sustained release of these additive ions over a long period of time directly at the defect site, so you get a more targeted release of these nutrients. Pharmacologics and biologics are all surface loaded and exhibit a burst release effect with very little long term benefit. While the specific action of many of these trace elements is still largely a mystery, it has been a hotbed for research and the potential for commercialization is high. In fact, many studies have shown that the addition of trace elements will naturally increase production of important growth factors such as BMP-2 and VEGF [102,103]. Because these inorganic additives are largely incorporated into the body on a dietary basis and are used in many important processes in the body, there are distinctive regulatory mechanisms that inherently deem their use as additives in CaPs safer than the addition of pharmacologics or biologics.

Table 1.4 Effects of Dopants on Mechanical Propties of TCP [104–107].

SBF Study Ceramics Relative Density (%) Compressive Strength (MPa) 2 Weeks 4 Weeks 8 Weeks 12 Weeks 16 Weeks Pure TCP 96.17 ± 1.15% 418.19 ± 28 MPa 207.3 ± 24.8 MPa 219.8 ± 37.5 MPa 216 ± 9.0 MPa 162.3 ± 22.0 MPa 158 ± 50.0 MPa Tricalcium Phosphate with Single Dopants 0.25% ZnO 93.88 ± 0.15% 386.61 ± 69.22 MPa 183.45 ± 26.02 MPa 199.79 ± 46.1 MPa 132.4 ± 29.69 MPa -- -- 1.0% SrO 94.47 ± 1.27% 180.47 ± 23.58 MPa 179.93 ± 40.15 MPa 164.75 ± 6.09 MPa 148.44 ± 16.55 MPa -- -- 0.5% SiO2 91.46 ± 1.06% 365.91 ± 30.34 MPa 353.5 ± 37.02 MPa 267.22 ± 31.75 MPa 250.41 ± 55.87 MPa 238.36 ± 20.54 MPa 215.7 ± 29.3 MPa Tricalcium Phosphate with Binary Dopants 0.5% SiO2 + 1.0% SrO 95.46 ± 0.95% 253.4 ± 11.6 MPa 246.6 ± 35.3 MPa 242.7 ± 40.2 MPa 237.0 ± 25.0 MPa 140.6 ± 12.2 MPa 168.9 ± 41.9 MPa 1.0% MgO + 1.0% SrO 97.26 ± 1.11% 359.0 ± 10.0 MPa 311.5 ± 42.3 MPa 350.33 ± 49.1 MPa 284.6 ± 28.0 MPa 234.3 ± 37.1 MPa 233.92 ± 32.4 MPa 0.5% SiO2 + 0.25% ZnO 91.69 ± 1.21% 272.53 ± 23.84 MPa 250.84 ± 20.55 MPa 239.84 ± 24.51 MPa 286.57 ± 47.18 MPa 274.49 ± 19.24 MPa 243.6 ± 3.73 MPa 1.0% SrO + 0.25% ZnO 95.14 ± 0.48% 208.18 ± 72.07 MPa 211.09 ± 51.08 MPa 257.98 ± 51.08 MPa 180.88 ± 36.9 MPa -- -- Tricalcium Phosphate with Ternary Dopants 1.0% MgO + 1.0% SrO + 0.5% SiO2 96.88 ± 1.0% 172.56 ± 12.85 MPa 137.3 ± 11.3 MPa 136.5 ± 28.0 MPa 118 ± 5.4 MPa 149.6 ± 15.2 MPa 149.1 ±41.3 MPa 1.0% MgO + 0.5% SiO2 + 0.25% ZnO 85.10 ± 1.54% 127.64 ± 6.52 MPa 135.63 ± 8.57 MPa 126.43 ± 15.18 MPa 122.45 ± 16.73 MPa 137.25 ± 26.80 MPa 157.4 ± 15.2 MPa

1.6.1 Silicon

Silicon, while not labeled an essential trace element to human health, has some very important functions in the human body. It has been found to be beneficial to cartilage and glycosaminoglycan formation or function, which may affect bone formation and maintenance, cardiovascular health, and wound healing [108]. In the early stages of the biomineralization process, silicon can be found at active calcification sites [109]. During the later stages of calcification, silicon plays a direct role in mineralization in aqueous orthosilicic acid (Si(OH)4) form by inducing the precipitation of HA from electrolyte solutions[109]. Because silicon is the second most abundant element on earth, deficiencies are rare and difficult to indicate in humans, but in experimental animal models dietary silicon deficiencies can result in deformed development of bone, low collagen formation and stunted growth [110,111]. In silicon substituted CaP materials, the presence of silicon can stimulate biological activity by increasing the solubility of the material, generating a more electronegative surface and creating a finer microstructure resulting in transformation of the material surface to a biologically equivalent apatite [112]. Early uses of silicon in CaPs were investigated as a way to stabilize β-TCP at high

25 temperatures to improve densification and prevent the formation of α-TCP, a more soluble and unstable form of CaP [66]. In a study where silicon substituted HA scaffolds were implanted in rabbits at various levels, results showed that silicon-substituted HA had increased amount of bone growth compared to the pure HA composition and also affected the osteoclastic resorption rates. Another study investigated the effects of silicon substituted HA granules in rabbit model and found similar results. [113,114], Aside from the potential osteogenic benefits of silicon addition to CaPs, several recent studies have noted that silicon may have angiogenic capabilities as well. In a study by Li and Chang, a simple calcium silicate materials was shown to induce

VEFG expression in human dermal fibroblasts, which in turn upregulated nitric oxide synthase and nitric oxide production in human endothelial cells [115]. Calcium silicate was used in an rabbit femur defect and results also indicated angiogenic effects [116]. Commercially, silicon is found in various orthopedic applications from silicate stabilized CaPs to the main component of

Bioglass and a mix of the two.

1.6.2 Zinc

Zinc is also considered an important essential trace element in human development.

There are several important metalloenzymes that utilize zinc for structure, catalytic or regulatory actions. One such enzyme that is absolutely vital for the maturation of new bone formation is alkaline phosphatases (ALP). ALPs are glycosyl-phosphatidylinositol anchored, Zn2+ metallated glycoproteins that are released during the maturation process of the osteoblastic cell life cycle and help to catalyze the hydrolysis of phosphomonoesters into inorganic phosphates [52].

Essentially, their role is to create an alkaline environment that favors the precipitation and

26 subsequent mineralization of these inorganic phosphates onto the extracellular matrix that the osteoblasts produce. Dietary deficiencies in zinc are related to several skeletal developmental defects and may have a role in the prevention of osteoporosis. During the natural bone remodeling process, zinc is released from the bone [117]. The excess amount of zinc in the microenvironment due to this release is thought to stop the osteoclastic resorption process and stimulate the osteoblastic bone building process [118,119], which makes it an ion of high interest. Zinc doped CaP materials have shown increase osteoblastic response in vitro as well as increase new bone formation in vivo [98,120]. Zinc phosphate cements are some of the oldest and most widely used cements in the dental industry.

1.6.3 Magnesium

Magnesium is an essential element and the 10th most abundant element in the human body, with about 65% of total body magnesium contained in bone and teeth [121]. It is no wonder, then, that it has had generated particular interest in bone health. In a report by Maier et al., high doses of magnesium in vitro were correlated to responses that suggested magnesium plays a direct and vital role in maintaining vascular function. Furthermore, they confirmed that the presence of magnesium induced nitric oxide production in endothelial cells which is essentially the same mechanism that VEGF uses to induce angiogenesis [122,123]. In a yearlong animal deficiency study, a group reported that prolonged magnesium deficiency directly resulted in osteoporosis [124]. Researchers have also found that by doping CaP materials with magnesium, the densification is improved and well as osteoblastic cellular attachment, proliferation and ALP production [125]. In vivo studies have noted that hydroxyapatite doped

27 with magnesium phosphate in a femoral bone defect showed greater osteogenic properties when compared to a pure control [126]. Magnesium has been used clinically in magnesium phosphate bone cements and in several different bioglass compositions.

1.6.4 Strontium

Strontium is a non-essential element that has bone seeking behavior. Essentially, because it is very similar in size and charge to Ca2+, it is thought to displace Ca2+ ions in osteoblastic mediated processes. Researchers have identified that strontium likely stimulates bone formation by a dual mode of action. First is activates the calcium sensing receptor (CaSR) in osteoblasts

[127,128] which simultaneously increases osteoprotegerin (OPG), production and decreases receptor activator of nuclear factor kappa beta ligand (RANKL) expression [129]. OPG is a protein that inhibits RANKL induced osteoclastogenesis by operating as a decoy receptor for

RANKL [130]. The OPG/RANKL ratio, then, can be a powerful regulator of bone resorption and osteoclastogenesis. In the UK, strontium is utilized in the form of strontium ranelate as a prescriptive treatment for osteoporosis in post-menopausal women. Phase III clinical trials that began in 2000 investigated the efficacy of strontium ranelate in reducing vertebral fractures and peripheral fractures, including hip fractures. After 3 years, patients treated with strontium ranelate showed significant reduction in vertebral fractures (41%) and hip fractures (36%) compared with patients treated with placebo [131]. Several studies have shown positive effects of the addition of strontium to CaP materials both in vitro and in vivo. In a study by Ni et al., a strontium doped HA based bone cement was used in a goat model to investigate the outcomes of revisional total hip arthroplasty surgeries compared to a conventional Poly(methyl methacrylate)

(PMMA) cement. After 9 months, the strontium containing HA cement was found to have better mechanical bonding than the PMMA [132]. Weak bonding is a primary reason for failure of a revisional THA surgery. Another group studies the in vitro angiogenic properties of a strontium doped CaP material and found increased endothelial cellular proliferation and tubule formation, both characteristics of angiogenic abilities[133]. The same group published an in vivo study showing that their strontium doped material also showed between 5% and 10% increased new bone growth over 16 weeks when compared to a similar materials without strontium [134].

While strontium is not use in any US devices, a company RepRegen is already marketing a grafting product StronBone that contains strontium in the UK.

1.7 Bioreactors for cellular population in scaffolds

In addition to adding porous structures, dopants and biologics to increase the success and healing rate of grafting applications, a new trend is emerging utilizing the hosts’ individual biology. Adult mesenchymal stem cells can be extracted from the fatty tissue in the bone marrow with relative ease. These cells can then be cultured and grown onto scaffolds before implantation, where they can differentiate into bone building osteoblast cells and help facilitate the healing process. Standard cell culture methods, however, are not very effective at propagating the infiltration of these cells fully into the scaffold. Various dynamic culture systems have been developed to assist in the cell infiltration process and offer superior mechanical stimulation, cytokine and other important signal recirculation as well as gas exchange [135]. There are essentially three classes of bioreactors that are widely used in bone tissue engineering: spinner flasks, rotating wall and perfusion [136–140].

The spinner flask bioreactor is perhaps the simplest of the three. Spinner flasks are generally composed of a media reservoir with side arms for media and scaffold removal and a porous cap that allows gas exchange (Figure 1.12). Scaffolds are suspended from the cap into the media. A stir bar is placed on the bottom of the flask and the entire system set on a stir plate.

Cells can be seeded directly onto the sample or suspended in media for infiltration. While this method has been effective with smaller scaffolds and has been shown to increase ALP and osteocalcin production in osteoblasts, it has been speculated that for larger scaffolds the mass transport may not be sufficient resulting in a sharp nutrient gradient that leaves cells in the center of the construct starved [136,141,142].

Stir Plate

Figure 1.12. Typical spinner flask bioreactor [142].

The rotating wall bioreactor offers similar gas exchange properties and optimal shear stresses compared to spinner flask systems [143]. In this type of reactor, two concentric cylinders are utilized. The inner concentric cylinder provides gas exchange and is stationary, while the outer cylinder rotates. The media and scaffolds are confined in the space between the two cylinders (Figure 1.13). The free movement of the scaffolds in the media results in a microgravity environment, while the fluid is caused to flow due to the centrifugal forces of the cylinder [144]. While this system has the benefits of optimal shear stresses caused by fluid flow and enhanced gas exchange, its performance is hindered by the free movement of scaffolds within the vessel, colliding chaotically against the vessel walls [142]. This system is best used when the density of scaffolds is less than that of the cell media [145].

Figure 1.13. Schematic of a rotating wall bioreactor. Outside wall rotates to circulate media [142].

Perhaps the most versatile and best option for circulation of media throughout the scaffold is the flow perfusion bioreactor. While there are many different designs, most systems use the same basic principles. They consist of a peristaltic pump, a tubing circuit and a sample holder (Figure 1.14). Using a perfusion system allows for control of media flow rate as well as the shear stresses experienced on the surface and pores of the scaffold. The complexity of this system is its only drawback. Sample holders have to be specifically designed for each type of scaffold to ensure that ample flow of media through the sample is achieved. Also, with varying sample holders and sample types, the fluid flow rate has to be optimized so that the shear stresses experienced by the surfaces of the scaffolds are not too extreme for cell growth.

Figure 1.14. Schematic of perfusion bioreactor. Media is directly perfused through porous scaffold sealed into a growth chamber [142].

The proposed research has the following specific aims:

Specific Aim 1. To study the effects of ZnO and SiO2 dopants and volume fraction porosity in β- tricalcium phosphate (β-TCP) scaffolds on the mechanical strength and cell material interaction in vitro.

Porous scaffolds with 500μm, 750μm and 1000μm interconnected channel sizes will be fabricated using a 3-dimensional printing technique (3DP). These scaffolds will be sintered at 1250 ˚C in a conventional electric muffle furnace for 2 hours. A comparative study was carried out by evaluating the differences between the pure composition and doped composition. Phase, microstructure and mechanical strength was tested. In vitro cell-material interaction was assessed with human fetal osteoblast cells to determine the effects of dopants and designed scaffold channel size on osteogenic capabilities.

Specific Aim 2. To study the effects of SiO2, ZnO, SrO and MgO dopants in CaP materials on osteogenesis in vivo and the cellular response of osteoblasts in vitro.

SiO2, ZnO, SrO and MgO were physically mixed with commercially purchased β-TCP at optimized values and dense compact samples were uniaxially pressed then cold isostatically pressed.

Samples were sintered at 1250 ˚C and seeded with human fetal osteoblasts and then tested for alkaline phosphatase (ALP) activity and runt-related transcription factor 2 (Runx2) production.

Samples containing SiO2 and ZnO were physically mixed with synthesized β-TCP powders.

Porous samples were made using an oil emulsion method and cultured with osteoblasts for 28 days. RT-

PCR was used to evaluate various differentiation markers. Samples were alto evaluated for density, compressive strength and microstructure.

SiO2/ZnO doped β-TCP were used to make scaffolds using 3DP. Impact of the addition of dopants on in vivo osteogenesis was tested in the cortical defect of rat distal femur model. New bone growth, collagen I, osteocalcin, angiogenesis and TRAP activity were evaluated.

Specific Aim 3. To study the effects of SiO2/ ZnO dopant combination in porous β-TCP and dynamic cell culture on the proliferation and differentiation of osteoblast cells.

SiO2/ZnO was physically mixed with synthesized β-TCP powder at optimized values and porous samples were made using an oil emulsion method. Samples were sintered at 1250 ˚C and seeded with human pre-osteoblast cells. The effects of dopants as well as a dynamic culture system on proliferation and differentiation will be measured by the activities of alkaline phosphatase, alamar blue and SEM.

Physical characteristics such as density and microstructure were also evaluated.

2. CHAPTER TWO

Effects of SiO2 and ZnO doping on mechanical and biological properties of 3D printed

TCP scaffolds

2.1 Introduction

Currently in orthopedic and dental applications, autografts (bone material obtained from another anatomic site in the same subject) and allografts (bone material obtained from an outside source, such as processed cadaver bones) are the standard materials for bone repair, substitution and augmentation. Autografts require a second operation at the site of tissue harvest, which increases the time and cost of surgery, post-operative pain and may lead to donor site morbidity [146]. There is also a limited supply of autograft sites that may be harvested. Studies have shown that at the 10 year mark, as many as 30% of allografts generate complications that could affect the structural integrity of the implant [146]. In the past 15 years, due in part to the shortcomings of natural graft materials, there has been a growing interest in the investigation of synthetic bone alternatives as well manufacturing methods. Advances in rapid prototyping give rise to the possibility of designing patient specific implants with complex geometry that can be tailored to the patient’s defect site [147].

Bone substitution materials should encompass many different characteristics for consideration in clinical use. Not only should the material be porous to serve as a scaffold for capillary growth, the material should have excellent biocompatibility, osteoconductivity, and a

35 complete lack of [148]. Ideally, a scaffold will offer mechanical support and will be resorbed as new bone and tissue growth occur. β- tricalcium phosphate materials have been widely studied for their use in orthopedic implant applications and tissue scaffolds due to their excellent biocompatibility and controllable bioresorbability [97,149–151]. For optimal tissue ingrowth, mechanical interlocking at the implant-bone interface and transport of nutrients and metabolic waste, parameters such as density, pore shape, pore size and pore interconnection pathways are important [79,152–154]. Many methods for creating complex calcium phosphate (CaP) scaffolds have been investigated including direct and indirect extrusion freeforming, selective laser sintering, stereolithography and ink-jet printing [155]. Direct ink-jet printing has demonstrated very good resolution with great control over the complexity of geometry [156–

159]. Other advantages of 3D printing are that macroporosity is induced into the scaffold during the fabrication process without the need for a support structure [112,160]. Moreover, these scaffolds are fabricated directly using the 3D printer with designed materials unlike many other processes where multiple post processing steps are needed.

One of the major disadvantages of ceramic scaffolds is their low strength, especially at high volume percentages of porosity [161]. Recent studies have shown that by doping β-TCP with trace elements commonly found in bone, the dissolution/resorption rates, densification, cell-material interaction and mechanical strength can be controlled [109,162–165]. One study showed that in osteoporotic subjects, supplemental silicon and monomethyl trisilanol resulted in increases in femoral and lumbar spine bone mass density [166]. The same study demonstrated that this treatment was more effective than a bisphosphonate (Etidronate) and sodium fluoride treatment. Silicon has been proven to be an important trace element in bone and connective

36 tissue formation and can stimulate biological activity by increasing the solubility of the material, generating a more electronegative surface and creating a finer microstructure resulting in transformation of the material surface to a biologically equivalent apatite [109]. Zinc, also an important trace element in bone formation, has been shown to control grain growth, increase density and have stimulatory effects on bone formation when added to β-TCP [99]. Zinc is released during skeletal breakdown and has been shown to inhibit osteoclastic bone resorption

[29]. It does this by inhibiting osteoclast-like cell formation from bone marrow cells and inducing apoptosis of mature osteoclast [167]. Osteoporotic patients have also been shown to have lower levels of skeletal zinc than that of control groups [168]. This study examines the effects of silica (SiO2) and zinc oxide (ZnO) addition, as a binary dopant system, on the physical, mechanical and biological properties of 3D interconnected porous TCP scaffolds created by the 3-D printing, a solid freeform fabrication method.

2.2 Materials and Methods

2.2.1 Fabrication of Porous Scaffolds

High purity oxide based sintering additives, silicon dioxide (SiO2) (99%+ purity) and zinc oxide (ZnO) (99.9%+ purity) were purchased from Fisher Scientific (Fair Lawn, NJ).

Synthetic β-tricalcium phosphate powder was obtained from Berkley Advanced Biomaterials

Inc. (Berkeley, CA) with an average particle size of 550 nm and specific average surface area of

10-50 m2g-1. Powder was made in batches of 100g and doped with 0.25 wt % ZnO and 0.5 wt %

SiO2 as a binary dopant system. Dopant amounts were chosen based on previous research from our group [97,169]. Powder was added to and mixed in 500 mL polypropylene Nalgene bottles,

37 with 300 g of 5 mm diameter zirconia milling media. 150mL of ethanol was added and ball milling was carried out for 6 h at 70 rpm to minimize the formation of agglomerates and increase the homogeneity of the powders. After milling, the Nalgene bottle was placed in an oven at 60° C for 24 h for drying. Milling media were removed and the powder was further dried at 60° C for another 24 h. Finally, agglomerates were removed using a mortar and pestle.

Cylindrical scaffold CAD files (diameter 7 mm and height 10.5 mm) were created with interconnected square channels of 1000µm, 750µm, and 500µm sides (Fig. 2.1). Scaffolds were fabricated using a 3-D printer (R-1 R&D printer by ProMetal). The printing system consists of a deposition bed that starts with only a small base layer of powder, a feed bed that is filled with powder, a powder spreader, an ink jet print head and a drying unit. A schematic representation of operation is given in Figure 2.2. The printer head sprays binder onto the loose powder in the deposition bed according to the CAD pattern, the deposition bed is lowered by 20 µm and the feeder bed is raised by 60 µm. The spreader will then push the excess powder from the feeder bed and evenly spread it onto the lowered deposition bed. The process is repeated, layer by layer, until the CAD image is completely printed. Parts were built in 20µm layers thickness with commercial organic water based binder obtained from ProMetal (Irwin, Pa). Once finished, parts were dried at 150°C for 1.5 h. Green scaffolds were then gently brushed clean and the remaining powder was removed by using a compressed air blower. After cleaning, green scaffolds were sintered in a muffle furnace at 1250° C for 2 h. Important fabrication parameters that were optimized include binder selection, binder drop size and amount deposited, powder particle size, drying time between layers, powder spread rate, build layer thickness and powder feed to build ratio.

Figure 2.1. CAD image of a cross section of a porous scaffold. Square channels are oriented 0˚/90˚ for subsequent layers

Figure 2.2. Schematic drawing representing the 3D printing process.

2.2.2 Microstructure, Phase Analysis and Mechanical Properties

The surface morphologies of the sintered scaffolds were observed under a field emission scanning electron microscope (FESEM) (FEI Inc., OR, USA). Phase analysis of sintered β-TCP samples with and without dopants was carried out by X-ray diffraction (XRD) using a Siemens

D-500 X-ray powder Diffractometer (Siemens AG, Karlsruhe, Germany) with CuKα radiation and a Ni-filter. Each run was performed with 2θ values between 10° and 60° at a step size of

0.02° and a count time of 0.5 s per step. Density was measured using Archimedes method.

Samples were weighed initially dry and then submerged in boiling water for 3 minutes to remove any excess air that may be trapped in the porous structure. The samples were then transferred from the boiling water to room temperature water, where the weight was recorded again (n=3). Compressive strengths of undoped and doped TCP scaffolds were determined using a screw-driven universal testing machine (AG-IS, Shimadzu, Japan) with a constant crosshead speed of 0.33 mm/min. Compressive strength was calculated using the maximum load recorded and the sample dimensions. Compressive strength was tested on ten samples for each composition.

2.2.3 In Vitro Osteoblast Cell-Material Interactions

Pure and doped TCP samples were studied for their cell-materials interactions using established human fetal osteoblast (hFOB) cells (hFOB 1.19, ATCC, Manassas, VA). The base medium for this cell line was a 1:1 mixture of Ham's F12 Medium and Dulbecco's Modified

Eagle's Medium (DMEM/F12, Sigma, St. Louis, MO), with 2.5 mM L-glutamine (without phenol red). This base medium was supplemented with 10% fetal bovine serum (HyClone,

Logan, UT) and 0.3 mg/mL G418 (Sigma, St. Louis, MO).

All scaffolds were sterilized by autoclaving at 121oC for 30 min and then placed in 24- well plates. Cells were then seeded onto the samples. Cultures were maintained at 34oC under a

5% CO2 atmosphere as recommended by ATCC for hFOB 1.19. The medium was changed every 2 to 3 days for the duration of the experiment.

An MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay was used to evaluate cell proliferation. The MTT (Sigma, St. Louis, MO) solution of 5 mg/ml was prepared by dissolving MTT in PBS and the solution then filter-sterilized (0.2 µm). The MTT was diluted (100 µL into 900 µL) in DMEM/F12 medium and 1 mL of the diluted solution was then added to each of the samples. After 2 h of incubation, 1 mL of solubilization solution made up of 10% Triton X-100, 0.1N HCl, and isopropanol was added to dissolve the formazan crystals. 100 µL of the resulting supernatant was transferred into a 96-well plate, and read by a plate reader at 570 nm. Cultures were evaluated at 3, 7 and 11 days. Data was normalized to pure TCP control scaffolds with corresponding pore parameters.

Samples for SEM observation were collected at days 3, 7 and 11 and fixed with 2% paraformaldehyde/2% glutaraldehyde in 0.1M cacodylate buffer overnight at 4 oC. Post-fixation was performed with 2% osmium tetroxide (OsO4) for 2 hours at room temperature. Fixed samples were then dehydrated in an ethanol series (30%, 50%, 70%, 95% and 100% three times), followed by a hexamethyldisilane (HMDS) drying procedure. After gold sputtering, samples were observed under FESEM.

2.3 Results

2.3.1 Scaffold Fabrication

The Ex One R1 3D printer was commercially designed and optimized for the use of metal powders. Our goal in this study was to produce scaffolds using CaP ceramic powders.

One of the major challenges was the optimization of processing parameters to effectively produce these scaffolds using commercially available binder. Our process optimization research approach was first focused on how to produce a green TCP part followed by how to improve green strength of the part for mechanical handleability and depowderization to produce a 3D interconnected porous scaffold. Processing parameters used are given in Table 2.1. The drop volume is a fixed parameter that is measured by the printer and is dictated by the viscosity and density of the binder. It is a measurement of the volume of each drop of binder that is released from a nozzle on the print head. The drop volume measured for the commercial binder used was

76 pL. The saturation percentage is based on the drop volume, packing efficiency of the powder layer and binder viscosity. The 100% value is a good starting point to work with because it is essentially a calculated estimate of how the binder will spread through the powder. If the saturation is too high, the binder will bleed into the surrounding powder, while too little saturation will result in extremely weak green scaffolds due to poor bonding between layers.

The values that worked best were 110% for pure TCP powder and 100% for doped TCP powder.

The pure TCP needed a little extra binder to maintain strength while handling and depowdering.

The number of passes that the print head used to reach the desired saturation was 7 at a speed of

140 mm/s. The layer thickness was an important parameter because it dictated the depth resolution of the scaffold. The primary concern that was taken into consideration when choosing

42 a layer thickness was the size of the powder particles. By using very fine powder, 550 nm, layer thicknesses of 20 µm for pure TCP powder and 30 µm for doped TCP powder were achieved while maintaining very smooth spreading. The doped powder had less depth resolution due to slight agglomeration of particles during the ball milling process. The feed powder to layer thickness ratio chosen was 3. This means that for every 20 or 30 µm that the build bed is lowered, the feed bed is raised by 60 or 90 µm, respectively. To help ensure that the powder is spread smoothly onto the build bed, a spread speed of 0.5 mm/s was used. This is the speed that the beds pass under the rolling spreader. As the beds move, loose powder from the raised feed layer is spread evenly onto the lowered build layer. By understanding the fixed processing parameters and using this information to adjust the variable processing parameters, we were able to successfully produce high resolution scaffolds with both doped and pure TCP powders.

Table 2.1. A list of important 3D printing parameters for fabrication of CaP scaffolds.

2.3.2 Physical and Mechanical Characterization

XRD was performed to determine the CaP phases of the sintered scaffolds and confirmed that in both samples, β-TCP was the primary phase present after sintering at 1250 °C.

α-TCP phase was also present in small amounts in both samples, which is expected at sintering temperatures above 1150 ˚C. The amount of α phase present in the doped samples was smaller than in the pure composition as shown in Figure 2.3. The characteristic peaks of β-TCP and α-

TCP match well with JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP), but show a small peak shift in the doped TCP samples. Such peak shifts due to dopant addition are common and reported by other researchers as well [97,125,169].

Figure 2.3. XRD patterns of doped and pure scaffolds sintered at 1250 ˚C. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP)

Figure 2.4 shows printed, cleaned and sintered SiO2/ZnO doped and pure TCP scaffolds.

Density and designed percent porosity for sintered scaffolds are shown in Table 2.2. The average apparent strut density of the SiO2/ZnO doped scaffolds was found to be 94.1 ± 1.6% and for pure TCP scaffolds strut density was measured to be 90.8 ± 0.8%. Figure 2.5 shows the surface morphology of samples sintered at 1250 oC. The microstructures of both pure and doped scaffolds showed some residual porosity, 9% and 5% by volume, respectively. Both samples also showed evidence of liquid phase sintering. Table 2.3 gives data collected from these micrographs, demonstrating that there was more shrinkage in the doped samples than in their pure counterparts. Compressive strength comparisons between doped and pure scaffolds are presented in Figure 2.6. The doped samples, which had less total open pore volume than the pure samples, showed the greatest compressive strength with 1000µm, 750µm and 500µm green channel sizes at 10.21 ± 0.11 MPa, 8.2 ± 0.4 MPa and 4.34 ± 0.3 MPa, respectively. The pure samples with green channel sizes 1000µm, 750µm and 500µm had average compressive strengths of 5.48 ± 0.04 MPa, 2.68 ± 0.2 MPa and 1.75 ± 0.2 MPa, respectively.

Figure 2.4. Sintered SiO2/ZnO doped and pure scaffolds of each designed pore volume.

Table 2.2. Percent designed porosity and density of 3D printed scaffolds of pure and doped composition.

Figure 2.5. Surface morphology of sintered scaffolds of (a) pure TCP composition and (b) SiO2/ZnO doped composition.

Table 2.3. Measurements of post sintered pore size in pure and SiO2/ZnO doped TCP scaffolds.

Figure 2.6: Compressive strengths of sintered 3D printed scaffolds for both pure and doped β- TCP. Statistical analysis shows that the differences are significant (* P << 0.05, where n = 10)

2.3.3 In Vitro Biological Characterization

MTT results depict a clear distinction between the SiO2/ZnO doped scaffolds and the pure TCP control scaffolds, given in Figure 2.7. The amount of viable cells was significantly greater in the doped samples than on the pure scaffolds. Data ranged from a 3% increase as a low on day 3, to 92% more viable cells as a high on day 11. SEM micrographs confirmed data gathered from the MTT assay. The micrographs show on day 3 that cellular attachment characteristics were good in both doped and pure scaffolds, with osteoblast cells extending and attaching to various anchor points. By day 7, there was a clear proliferation distinction between the doped and pure samples, with doped samples exhibiting significantly larger groups of cells

47 attaching to the surface. In both cases, complex filopodial communication networks were beginning to emerge between nearby cells, as shown in Figure 2.8.

2.4 Discussion

One of the major challenges of this study was the optimization of the processing parameters of the 3D printer to produce high quality scaffolds. While there are several factors that need to be taken into consideration, as mentioned previously, the ability to create conditions that result in uniform and smooth powder spreading and high green strength are what made this

49 study successful. A smooth powder build layer is extremely important to the mechanical strength of the scaffolds. If the build layer is rough, defects are incorporated into the structure during the printing process resulting in decreased mechanical strength of sintered scaffolds as well as diminished ability to handle greens scaffolds. Binder choice was also an important factor. The binder had to fluid enough to penetrate through the powder build layer to give good bonding between layers, but viscous enough that it did not bleed into areas surrounding the print pattern. Also, due to complex geometry, it had to offer enough mechanical support to withstand the vigorous cleaning and depowdering process to ensure interconnectivity between the pores in the scaffolds.

Present results show that addition of SiO2 and ZnO to 3-D interconnected porous TCP scaffolds can influence both biological and mechanical properties of the scaffolds. Sintering additives have been investigated in the past for dense and porous TCP and hydroxyapatite compacts to increase the density and mechanical strength [65,125], however there have been no studies examining the effects on 3D printed scaffolds. XRD patterns shown in Figure 2.3 confirm that the α-TCP phase formation is reduced in the SiO2/ZnO doped samples when compared to pure TCP scaffolds sintered at the same temperature and time. In the case of TCP ceramics, dopants are used as a sintering additive to retard the transformation of β-TCP to α-

TCP at temperatures above 1150 oC. Studies have indicated that the presence of α-phase is closely related with the expansion of sample volume, declining shrinkage rate and is generally considered to prevent TCP from further densification [66,170]. Both compositions also showed evidence of some liquid-phase sintering, demonstrated in Figure 2.5, which is beneficial because it results in a more uniform densification than that of solid state sintering. Channel size

50 analysis, given in Table 2.2, in sintered samples demonstrates overall volume shrinkage. When compared to pure TCP scaffolds, the doped samples revealed increased density, specified in

Table 2.1, which can be attributed to reduced α-TCP formation. Another possible mechanism of increased densification would be enhanced wettability during liquid-phase sintering in the doped samples. Greater wettability during sintering would lead to more efficient capillary action and particle rearrangement. Ultimately, density and designed porosity play the most important role to increase the maximum strength of the samples [161]. Pores act as stress concentrators and may lead to premature failure if in the near vicinity of other defects such as inclusions, agglomerates and surface flaws. The average compressive strength for the doped scaffolds was higher than their pure counterparts with a mean 250% increase in strength for each channel size.

The pure scaffolds matched well with the lower end strength of cancellous bone, while the doped scaffolds matched well with the higher end strength of cancellous bone, which has a strength range of 0.5 – 14.6 MPa. These results indicate that with the addition of dopants, TCP scaffolds may be tailored to match specific strength criteria.

Additional issues to be considered when evaluating the performance of a viable scaffold for clinical applications are the initial fixation period and healing time after a surgical procedure.

While CaP materials alone are excellent for bone replacement applications, there is merit in investigating the use of dopants to enhance proliferation and differentiation of bone building osteoblasts. MTT data shows that cells proliferate at a greater rate within the doped samples when compared a pure TCP control group as seen in Figure 2.7. SEM micrographs taken of hFOB cells seeded and incubated on the scaffolds, presented in Figure 2.8, show similar cellular morphology between the doped and pure scaffolds. At day 3, the cells are elongated and

51 flattened and are growing into the abundance of micropores. By day 7, complex filipodial communication networks are set up between neighboring cells and there is a clear increase in the amount of living cells on the surface of the doped scaffolds when compared to the pure TCP scaffolds. Zinc plays an important role in bone metabolism, where deficiency may result in bone loss. Studies have shown that zinc has a stimulatory effect on bone formation in vitro and in vivo and increases alkaline phosphatase activity, an enzyme marker for osteoblastic differentiation

[119,171]. It has also been shown that zinc also plays a critical role in the production and release of vascular endothelial growth factor (VEGF) by osteoblasts [102]. VEGF stimulates the proliferation of vascular endothelial cells which, during bone remodeling, provide microvasculature. Increased silicon dietary uptake has been linked to better bone health and increased bone mineral density while lab studies have shown positive effects on osteoblast proliferation and differentiation [172,173]. While much research has demonstrated that silicon has positive effects on bone metabolism, the exact mechanism of action is poorly understood.

The effects of designed macropore size on tissue-material interaction would best be determined in an animal model. For such a study it would be important to consider the performance tradeoffs in tissue integration vs. mechanical strength. There is enough compelling evidence that

SiO2/ZnO doped scaffolds perform better, mechanically and biologically, than pure TCP scaffolds to justify in vivo studies in the future. 3D printing is shown to be a viable manufacturing tool for tailored porosity TCP scaffolds with and without dopants for future clinical studies and applications in dental and orthopedic applications.

Conclusion

The present study examined the influence of a SiO2/ZnO binary dopant system on the mechanical and biological properties of designed macroporous TCP scaffolds fabricated using commercial 3D printing technology. The addition of dopants decreased the β to α phase transformation of TCP sintered at 1250 oC, increased densification and as a result, showed up to

250% increase in compressive strength when compared to pure TCP scaffolds. A maximum compressive strength of 10.21 ± 0.33 MPa was achieved in doped scaffolds with 500 µm interconnected macropores. MTT assay and SEM analysis of cellular interaction suggests that the doped samples increased cellular attachment and proliferation. These results establish

SiO2/ZnO doped TCP scaffolds with interconnected channels fabricated via 3D printing as viable candidates for further research in clinical applications in non-load bearing bone tissue replacements.

3.CHAPTER THREE

SiO2 and ZnO Dopants in 3D Printed TCP Scaffolds Enhances Osteogenesis and

Angiogenesis in vivo

3.1. Introduction

Currently in the orthopedic and dental industry, calcium phosphate (CaP) materials are amongst the most widely used and studied resorbable bone simulate due to their compositional similarity to natural bone and excellent biocompatibility. Presently, clinical utilization of CaPs can be found in a wide array of products including coatings on metallic implants, bone cements and as grafting materials [174]. β-tricalcium phosphate (β-TCP) is a CaP material that has been

-29 of particular interest because its solubility product (Ksp = 1.25 x 10 ) allows the material to

-26 degrade over the course of months rather than several weeks such as α-TCP (Ksp = 3.16 x 10 )

-59 or several years for hydroxyapatite (HA) (Ksp = 2.35 x 10 ) [63,64]. This puts β-TCP in an ideal category for scaffolding material. New bone growth can occur faster than the dissolution of the scaffold, allowing the material to act as mechanical support for cellular migration, attachment and proliferation while the healing process takes place, but still achieving complete resorption relatively quickly so the healing process can complete.

Aside from biocompatibility, there are three major considerations that play a role in the success of a scaffold material for bone and dental implants: osteoconduction, osseointegration and osteoinduction. β-TCP is considered an osteoconductive material, or a material that permits bone growth onto its surface or throughout its pores . Osseointegration describes the ability to

54 encourage mechanical interlocking between living tissue and the implanted device that can withstand functional loading [175]. This consideration is usually addressed by the incorporation of a complex porous structure within an osteoconductive implant. Many methods have been used to create scaffolding materials with controlled porosity including direct and indirect extrusion freeforming, selective laser sintering, stereolithography and ink-jet printing [176].

Direct ink jet printing has the benefits of very fine resolution, high degree of control over complex geometry and introduction of micro/macro porosity without the need for a sacrificial support structure [100,156–158]. Osteoinduction is generally described as the ability to stimulate primitive, undifferentiated and pluripotent cells to develop into the bone-forming cell lineage, or to put more simply, the ability to induce osteogenesis [177]. β-TCP, on its own, lacks osteoinductive capabilities and much time and effort has been spent improving this quality.

Many researchers have noted that physical material properties, such as the presence of microporosity, macroporosity and pore interconnectivity, may be enough to introduce osteoinductive capabilities [71,178–180], but most of the focus has been on incorporating pharmaceuticals and biologics to achieve this effect with varying degrees of success[63,181–

184]. While most studies have noted markedly positive effects with pharmaceutical incorporation, the compounds tend to be fragile, making commercial sterilization difficult, and usually only exhibit short term release profiles. It is also important to note extreme hesitation by the FDA to approve combination products containing biologics. An alternative of interest to this strategy is to incorporate small amounts of biologically relevant metal oxide dopants into TCP in place of the pharmaceuticals. Our research and others have noted that the use of dopants such as silicon, zinc, strontium and magnesium have the ability to not only tailor strength and

55 strength degradation, but also enhance the biological response in vitro and in vivo[97,99,105,106,185]. Silicon has been noted to be an important trace element in osteogenesis, with research results indicating a strong stimulatory effect on cellular activities such as proliferation, differentiation, and mineralization of osteoblast cells as well as facilitating osteogenic differentiation of mesenchymal stem cells [109,186,187]. Other studies have observed that with supplemental silicon and monomethyl trisilanol in osteoporotic patients increased the femoral and lumbar spine bone mass density and was found to be more effective than a bisphosphonate and sodium fluoride treatment [166]. Botelho, et al. also noted that the presence of silicon in calcium phosphates increased alkaline phosphatase activity and total protein production of osteoblasts in vitro [188]. Zinc, also considered an essential trace element, is released during the skeletal breakdown process and has demonstrated the ability to inhibit osteoclastic bone resorption [117,119]. It also boosts osteogenic characteristics by inducing osteoblastogenesis as well as osteoblastic differentiation and mineralization [189,190]. Other studies have demonstrated that post-menopausal osteoporotic patients exhibit lower levels of skeletal zinc than control groups [168].

In this study we have combined the osteoconductive capabilities of β-TCP, added the capacity for osseointegration through complex geometry design and 3D printing and supplemented the material with metal oxide dopants silica (SiO2) and zinc oxide (ZnO) to promote osteoinduction. Scaffolds were implanted into a murine femoral defect model and analyzed over the course of 16 weeks for osteogenic properties.

3.2.0 Materials and Methods

3.2.1 Fabrication of Porous Scaffolds

High purity oxide based sintering additives, silicon dioxide (SiO2) (99%+ purity) and zinc oxide (ZnO) (99.9%+ purity) were purchased from Fisher Scientific (Fair Lawn, NJ).

Synthetic β-tricalcium phosphate powder was obtained from Berkley Advanced Biomaterials

Inc. (Berkeley, CA) with an average particle size of 550 nm and specific average surface area of

10-50 m2g-1. Powder was made in batches of 100g and doped with 0.25 wt % ZnO and 0.5 wt %

SiO2 by adding dopant powder to β-TCP precursor powder. Dopant amounts were chosen based on previous research from our group that optimized concentration based on cell interaction as well as physicochemical properties [97,125]. Powder was added to and mixed in 500 mL polypropylene Nalgene bottles, with 300 g of 5 mm diameter zirconia milling media. 150mL of ethanol was added and ball milling was carried out for 6 h at 70 rpm to minimize the formation of agglomerates and increase the homogeneity of the powders. After milling, the Nalgene bottle was placed in an oven at 60° C for 24 h for drying. Milling media were removed and the powder was further dried at 60° C for another 24 h. Finally, agglomerates were removed using a mortar and pestle.

Cylindrical scaffold CAD files (diameter 3.3 mm and height 5 mm) were created with interconnected square channels of 423µm, with the expectation of roughly 30% shrinkage after sintering for a final pore size of 300µm (Figure 3.1). Scaffolds were fabricated using a 3-D printer (R-1 R&D printer by ProMetal) by a process described previously [100]. Briefly, the printer head sprays binder onto loose powder in the deposition bed according to a CAD pattern, the deposition bed is lowered by 20 µm and the feeder bed is raised by 60 µm. The spreader will

57 then push the excess powder from the feeder bed and evenly spread it onto the lowered deposition bed. The process is repeated, layer by layer, until the CAD image is completely printed. After cleaning, green scaffolds were sintered in a muffle furnace at 1250° C for 2 h.

Sintered samples generally had a shrinkage rate of around 25%, giving final dimensions of

2.5mm diameter and 3.75 mm height with a final pore size of 317 μm. A previous study demonstrated that the addition of SiO2 and ZnO decreased α-phase formation in the sintered scaffolds and had as much as a 2.5 times increase in compressive strength when compared to the pure samples [100]. Samples were sterilized by autoclaving at 121 ˚C for 20 minutes.

Figure 3.1. (a) Rendered cross section of typical CAD file used for the 3D printing process. (b) Schematic diagram demonstrating 3D printing process. Large arrows indicate moving direction. (c) Actual scaffolds created by the Ex One printer. Numbers below the scaffold represent designed pore size.

3.2.2 Implantation Procedure

Sprague–Dawley rats (280–300 g, Charles Rivers Laboratories International, Inc.,

Wilmington, MA, USA) were used as an animal model for this study. Prior to surgery, the rats were housed in individual cages with alternating 12 h cycles of light and dark in temperature-

58 and humidity-controlled rooms. Ethics approval for animal experimentation was obtained from

Washington State University. Following acclimatization, all animals underwent bilateral surgery to create a bicortical defect in the distal femur on the popliteal plane (2.5 mm diameter). This defect model was designed by trained veterinary surgeons and found previously to be critical in size [106]. Rats were anesthetized using IsoFlo (isoflurane, USP, Abbott Laboratories, North

Chicago, IL, USA) coupled with an oxygen (Oxygen USP, A-L Compressed Gases Inc.,

Spokane, WA, USA) regulator, and monitored by pedal reflex and respiration rate to maintain proper surgical anesthesia. The defect was created in either femur, proximal to the lateral condyle, by means of 1–3 mm drill bits. The cavity was rinsed with physiological saline to wash away remaining bone fragments. Each animal received a control implant (pure TCP) and in addition, a doped implant in the contralateral leg. Following implantation, undyed braided- coated polyglycolic acid synthetic absorbable surgical suture (Surgical Specialties Corporation,

Reading, PA, USA) was used for stitching. Disinfectant was applied to the wound site to prevent infection. At 4,6, 8, 12 and 16 weeks post-surgery, rats with implants were euthanized by overdosing with halothane in a bell jar, followed by administration of a lethal injection of potassium chloride (70%) into the heart. For a control group, similar surgical procedures were undertaken but only one implant was placed (either pure or doped). Controls were used for urine analysis throughout the 16 week study. Overnight urine samples were collected in metabolic cages at: 0, 2, 8, 14, 21, 42, 63, 84 and 112 days after surgery. A total of 24 rats were used for the experimental group (4 for each time point for 4, 6, 8 and 12 weeks) and 8 rats for the control group (4 of each composition at week 16).

3.2.3 Mechanical Pushout Analysis

Mechanical testing was performed at 4 weeks after implantation. At test jig was manufactured to press the samples out of the bone using a screw driven universal testing machine (AG-IS, Shimadzu, Japan) at a constant crosshead speed of 0.33 mm/min.

3.2.4 Histomorphology

The bone–implant specimens were fixed in 10% neutral buffered formalin solution and dehydrated in graduated ethanol (70%, 95% and 100%) series. After embedding samples in

Spurr’s resin, each undecalcified implant block was sectioned perpendicular to the implant surface using a low-speed diamond blade. After polishing, the sections were stained by

Goldner–Masson trichrome stain and observed by light microscopy (Olympus BH-2, Olympus

America Inc., USA).

3.2.5 Immunohistochemistry

For the immunohistochemical work scaffolds, along with surrounding tissue, were excised and fixed in 4% paraformaldehyde followed by decalcification in 10% EDTA. Samples were then dehydraded in a graded ethanol series and embedded in paraffin and sectioned to

10μm thickness with a standard microtome and placed on glass slides. Afterwards, sections were deparaffinized with xylene and rehydrated to water in a graded ethanol series. Antigen retrieval was performed in a Tris-EDTA Buffer (10mM Tris Base, 1mM EDTA Solution, 0.05% Tween

20, pH 9.0) by use of a vegetable steamer for 20 minutes followed by rinsing for 10 minutes in cold running tap water. Slides were then washed 2 x 5 minutes in TBS (Tris-buffered saline, 250

60 mM NaCl, pH 8.3) plus 0.025% Triton X-100 and blocked in 10% normal serum with 1% BSA in TBS for 2 hours at room temperature. Slides were drained and primary antibody, either for collagen type 1 (COL1), osteocalcin (OCN) (Abcam, Cambridge, MA), or vonWillebrand

Factor (vWF) (Millipore, Billerica, MA) diluted in TBS with 1% BSA (1:100 dilution) was added and set to incubate overnight at 4°C. The samples were then rinsed 2 x 5min in TBS with

0.025% Triton. The secondary antibody, goat anti-rat Oregon green 488 (Molecular Probes,

Eugene, OR), was diluted 1:100 in TBS and was used to incubate the samples with COL1 and

OCN for 1h. After rinsing three times for 10 minutes each with TBS the samples were mounted with ProLong Gold Antifade Reagent with DAPI (Life Technologies, Grand Island, NY), coverslipped and set to cure. Afterwards, confocal micrographs were taken using a Zeiss 510 laser scanning microscope (LSM 510 META, Carl Zeiss MicroImaging, Inc., NY, USA) with excitation wavelength of 488 nm and emission wavelength of 540 nm. The vWF samples were further processed with a blood vessel staining kit (Millipore, Billerica, MA) by incubating with premade secondary goat anti-rabbit antibody for 15 m, rinsing with TBS rinse buffer and reacting with streptavidin-HRP for 15 minutes. Samples were rinsed again and incubated in pre- prepared chromogen reagent for 10 minutes. Slides were then dehydrated through an ethanol grade mounted in a xylene based mounting fluid and a coverslip was applied. Samples were observed by light microscopy (Olympus BH-2, Olympus America Inc., USA).

3.2.6 FESEM

After sample excision, specimens were fixed in 10% neutral buffered formalin solution.

Post-fixation was performed with 2% osmium tetroxide (OsO4) for 2 h at room temperature. The

61 fixed samples were then dehydrated in an ethanol series (30%, 50%, 70%, 95% and 100% three times), followed by a hexamethyldisilane (HMDS) drying procedure. After drying, the samples were immersed in liquid nitrogen and the implant was freeze fractured. Samples were then gold sputter coated and observed under field emission scanning electron microscope (FESEM) (FEI

200F, FEI Inc., OR, USA). Additionally, another set of samples was extracted and fixed in 10% neutral buffered formalin solution for 24 hours. After fixation, samples were cut longitudinally in half using a low speed diamond saw at 600 RPM. Halves were placed in glass vials with

1mg/mL Proteinase K in a 10mg/mL sodium dodecyl sulfate (SDS) in distilled water solution to digest any organic material. Vials were placed in an incubator at 37˚C with gentle agitation for 1 week. The Proteinase K solution was changed every 24 hours. After one week, samples were rinsed with distilled water 3 times to remove any remaining debris. Acetone was then added to the samples and let sit for 3 hours. After 3 hours, acetone was replaced and let sit overnight.

Acetone was removed and samples gently shaken to remove excess solution. Diethyl ether was then added to the sample and let sit for 8 hours, then subsequently changed with fresh ether and let sit overnight. The ether solution was then discarded and samples shaken to remove any residue, then sample were placed in oven at 60˚C for 8 hours on a glass dish to evaporate any remaining residue. Finally, samples were gold sputtered and imaged using back scattered electrons (BSE) on the FESEM.

3.2.7 TRAP Staining

Samples were evaluated for tartrate resistant acid phosphatase (TRAP) activity. Paraffin embedded sections were deparaffinized and rehydrated as described previously and slides were

62 placed in 1% Naphthol AS-BI Phosphate (0.05M Napthol AS-BI phosphate in 2-ethoxyethanol) stock incubation medium (3% glacial acetic acid, 0.1M sodium acetate, 0.05M sodium tartrate) for 1 hour at 37 ˚C. Next slides were transferred to the developing solution (2% 0.5M sodium nitrate solution, 2% basic fuchsin solution (0.15M basic fuchsin in 2N HCL) in stock incubation medium) and kept at 37 ˚C for 5-12 m. Slides were then rinsed with water 3 times and counterstained with 0.02% Fast Green for 30 s. Finally, slides were dehydrated through an ethanol grade and cleared in xylene and observed by light microscopy (Olympus BH-2,

Olympus America Inc., USA).

3.2.8 Ca2+, Zn2+, and Si4+ ion excretion

Ca2+, Zn2+ and Si4+ content in urine samples was measured using a Shimadzu AA-6800 atomic absorption spectrophotometer (Shimadzu, Kyoto, Japan). Standard solutions were freshly prepared in ionization buffer to obtain a final concentration of 1-20 μg ml-1. Calcium, zinc, silicon and ionization buffer standards were purchased from High-Purity Standards (Charleston,

SC, USA).

3.2.9 Histomorphometry

Histomorphometry for samples stained with Goldner’s Trichrome was performed according to a protocol developed by Egan, et al. [191] with 3 random areas of interest in 3 different samples. For samples that were examined using fluorescent microscopy, images were first converted to grayscale. Image J (http://imagej.nih.gov/ij/) was then used to measure the area, integrated density and mean gray value after calibrating for microscope used to take the

63 pictures. The corrected total fluorescence (CTCF) was taken as: CTCT = Integrated Density –

(Area of micrograph X mean fluorescence of background readings). At least 3 random areas of interest in 3 different samples were used for each group.

3.2.10 Statistical Analysis

A one sided paired t-test was used to evaluate statistical differences between samples from the control group and doped group. Due to the limited availability of samples that is associated with an animal model study, results were found to be significant at p < 0.1, with at least n=3 samples.

3.3.0 Results

3.3.1 Mechanical Pushout Analysis

Mechanical pushout testing was performed at 4 weeks. Full interlocking was achieved at this time and the bone was broke in each case before the implant loosened from the defect site for both the pure and doped samples. Due to this result, further time points were not considered for testing.

3.3.2 New Bone Growth

At 6 and 8 weeks, Goldner’s trichrome staining and subsequent histomorphological measurements reveal enhanced bone growth in samples doped with SiO2/ZnO (Figure 3.2). By

12 weeks, samples had nearly complete infiltration of new bone and a distinction between the

64 samples was difficult to discern. Neither sample indicated a difference in degradation kinetics of the calcium phosphate structure over the course of 12 weeks.

3.3.3 Collagen I Formation

Samples doped with SiO2/ZnO developed significantly increased collagen I formation at

week 6 (Figure 3.3). The collagen levels measured in doped samples remained relatively stable

until week 16, where a noticeable increase was observed. The pure samples also had a relatively

stable amount of collagen formation over the first 12 weeks with an increase at week 16.

Although not statistically significant, the average collagen I formation in doped samples was

greater than that of the pure samples at weeks 12 and 16.

3.3.4 Osteocalcin Expression

The expression of osteocalcin was measured at weeks 6, 8, 12 and 16 (Figure 3.4). The

doped samples, on average, had increased amounts of osteocalcin protein present in samples at

weeks 6, 8 and 12 with statistical significance for weeks 8 and 12 when compared to pure TCP

scaffolds. For both sets of samples, levels remained relatively high and constant over the first 12

weeks and then there was a noticeable drop in activity at week 16.

3.3.6 Scaffold Dissolution

Figure 3.6 shows BSE imaging of longitudinal halves of implants after 6, 8 and 12 weeks. At the 6 week time point, both pure and doped samples retained most of their structural characteristics. By 8 weeks, however, the pure samples began to demonstrate increased degradation properties when compared to the doped samples, a trend that continued through the

12 week observation perior. At 12 weeks, pure samples exhibited significant degradation when compared to the SiO2/ZnO doped samples.

3.3.7 New Blood Vessel Formation

Figure 3.7 shows results from vWF immunohistostaining. Samples containing SiO2/ZnO

had remarkably higher number of new blood vessel formation than their pure counterparts at

weeks 6, 8 and 12. By the 16th week, however, samples revealed nearly identical total blood

vessel formation. Staining also suggest larger vessel and more complex vessel formation in the

doped samples. FESEM results (Figure 3.8) confirm what was observed with the vWF staining.

Blood vessel formation in the doped samples was more prevalent than in the pure samples and

vascular branching morphogenesis was more notable.

Figure 3.8. FESEM micrographs showing blood vessel formation. Arrows indicate blood vessels. Dotted lines show vascular branching pathways formed in the samples.

3.3.8 Osteoclast Activity

TRAP staining (Figure 3.9) was performed to discern differences in osteoclastic activity between the pure and doped samples. Both pure and doped samples demonstrated similar TRAP activity through the 16 week study.

Figure 3.9. TRAP staining in sectioned samples over the course of 16 weeks. The red color is indicative of TRAP activity and the green counterstain shows the tissue/scaffold.

3.3.9 Ionic Excretion

AAS results (Figure 3.10) indicate no perceivable difference in scaffold degradation

2+ throughout the 16 week study. Urine Ca levels measured indicate no trend over the course of

the study, but animals containing pure or doped samples essentially had equal urine

concentrations of Ca2+ throughout. Results from Si4+ analysis also indicate no difference between the control and doped implant, but demonstrate a slight increase in urine Si4+ levels immediately

after surgery. Zn2+ concentration in urine again did not have any differences between the control

and doped group, but showed a slight decrease in concentration immediately following

implantation.

Figure 3.10. AAS results for ion concentration in urine collected over 16 weeks.

3.4.0 Discussion

The key reasons for introducing porosity into calcium phosphate bone void fillers are for

enhanced nutrient delivery and mechanical interlocking at the bone-implant interface as well as

to facilitate bone ingrowth. Results from the pushout tests at 4 weeks indicate a very rapid

interlocking effect between the implant and the bone. While it was expected that, at 4 weeks,

there would have been some resistance to push the implant free, it was unanticipated that

complete fusion would have already taken place. Microporosity, macroporosity, pore size and

pore shape all play important roles in the joining of the bone-implant boundary [179,192,193].

Studies have demonstrated that anywhere between 100μm and 1000μm is sufficient for bone ingrowth [82]. Previous studies by our group show that the 3D printing method used incorporates microporosity as well as the designed macroporosity into the samples [100,194]. These previous results indicated that the pure samples had an average of 15.64% microporosity (open pores only) and the Si/Zn doped samples 28.25% microporosity (open pores only) along with the

35.5% designed macroporosity. The availability of micropores (<50 μm) in a macroporous scaffold has been shown to drastically increase bone infiltration [195]. The micropores on the surface of the implant are faster for bone to fill (less void area), while the macropores take much longer (much larger void area). Essentially, the micropores allow for rapid mechanical interlocking, while interconnected macropores allow for nutrient delivery and cell migration and proliferation to the center of the scaffold. In this case, the micropores allowed this effect, acting as adhesion points at the bone-implant boundary causing the rapid interlocking effect between the host tissue and CaP implant. Trichrome staining (Figure 3.2) results at 6 weeks confirm that both pure and doped samples had complete infiltration of mineralized bone into the micropores

72 at the bone-implant interface. The doped samples at week 6, however, also showed new bone tissue growth into some of the macropores as well and histomorphometry showed a significant increase in new bone formation when compared to the pure samples. At week 12, both samples demonstrated the progression of tissue integration throughout the scaffold, filling both the micropores as well as the macropores, with doped samples having significantly higher bone formation. By 12 weeks, both samples showed nearly complete infiltration of mineralized bone tissue and differences were difficult to detect. Because micropore and macropore characteristics between the doped and pure samples were essentially the same, the increased rate of bone regeneration is likely to be linked to the incorporation of the SiO2/ZnO dopant combination.

In a study performed by Hadley et al. [118], dietary zinc was linked to inhibition of osteoclastic resorption and increased markers for osteoblast differentiation, matrix maturation and mineralization in long bones of growing rats. Zinc deficiency, in other studies, has been linked to the suppression of matrix mineralization and osteoblastic differentiation [189,196].

Silicon has been linked to increased bone mineralization [197], increased bone growth [99,198] as well as increased osteoblastic cell differentiation [186,199,200]. By incorporating ZnO and

2+ 4+ SiO2 into the CaP scaffolds, an effective delivery system of Zn and Si has been achieved as indicated by results from this study. Collagen type I is the dominant fibrous protein in hard tissues such as bone and dentin. Osteoblasts, during the proliferation and maturation phase of their lifecycle, will produce a collagen extracellular matrix. This matrix will then become mineralized at the behest of various signaling mechanisms, most notably bone sialoprotien, ostopontin and osteocalcin [201]. It has been reported that collagen I mRNA levels are not effected by dietary zinc administration [118,202] in various animal models, but higher levels of

collagen I production were seen in this study, significantly so at week 6 (Figure 3.3). It is

possible that greater amount of collagen I was due to increased bone infiltration into the

scaffolds, rather than increased production of collagen. However, if correlated with Trichrome

data (Figure 3.2), the data does not seem to match this scenario well. Seaborn and Nielsen [203]

4+ demonstrated that Si can play an important role at different stages of bone healing and, more

specifically, silicon deprivation decreases the formation of collagen associated with wound

healing. Results from this study indicate a high level of collagen I during the initial healing phase

in samples containing SiO2/ZnO and a steady production up until the 12 week time point and

then a large increase. The pure samples saw a steady increase in collagen I over the course of the

study. Overall, then, the pure samples seem to have more collagen I formation as more new bone

tissue was forming, while the doped samples exhibited behavior of a Si4+ modulated response.

Osteocalcin is considered to be the latest expression marker in mature osteoblasts and is thought to trigger the mineralization process [204]. In this study, both the pure and doped scaffolds exhibited similar trends in osteocalcin expression during the 16 week study (Figure

3.4). At 6 weeks, levels were high and remained so through 12 weeks. At 16 weeks there was a noticeable drop in osteocalcin presence. The doped samples maintained higher levels at 6, 8 and

12 weeks and identical levels when compared to pure samples at week 16. Although studies have revealed that both Si4+ and Zn2+ can effectively upregulate osteocalcin activity [200,205], the

mechanism of action is poorly understood. Varanasi et al. [206] proposed that it may be a

downstream effect of increased collagen I levels. Briefly, osteoblasts will bind to the mature

collagen matrix, which is α2β1 integrin mediated process. Integrin binding at the osteoblast cell

membrane upregulates the expression of mitogen-activated protein kinases (MAPK). MAPK

transduces signals to the cell nucleus and phosphorylate to activate alkaline phosphatase (ALP)

and runt related transcription factor 2 (Runx2). ALP and Runx2 will then bind to the promoter

region of genes such as osteocalcin. A simplified cartoon is presented in Figure 3.5. Previous

studies have demonstrated that the addition of SiO2 and ZnO in these amounts actually increase

densification by significantly decreasing the alpha phase formation at high sintering temperatures

when compared to pure samples, leading to slower degradation kinetics [97,100,106].

Observation utilizing backscatter SEM, Figure 3.6. also noted significantly higher degradation in

the pure samples over 12 weeks. As osteoblasts have demonstrated significant positive response

3- 2+ to the dissolution products of CaPs (PO4 and Ca ) [200,207] and the pure samples are

significantly more soluble due to alpha phase formation, the increased osteogenic effects

observed in the doped samples, then, is very likely due to the presence of silicon and zinc.

Another important factor that affects the rate of new bone growth is angiogenesis. In

order to meet the nutritional demands of new bone growth, a vascular network has to be present.

In the present study, there is a stark difference in new blood vessel formation as observed by

vWF staining (Figure 3.7) between the pure and doped scaffolds over the first 12 weeks of the

study. Not only are there more blood vessels forming, FESEM results (Figure 3.8) show

increased occurrence of vascular branching morphogenesis at earlier time points in samples

containing SiO2 and ZnO. Recently, silicon has been shown to play an important role in angiogenesis. One study demonstrated that silicate containing bioceramics can induce expression of kinase insert domain receptor (a receptor for vascular endothelial growth factor (VEGF)),

basic fibroblast growth factor receptor 1 (a receptor for fibroblast growth factor (FGF)) and

activin A receptor type II-like 1 (a receptor for transforming growth factor β (TGFβ)) in human

aortic endothelial cells [208]. VEGF, FGF and TGFβ are thought to be the main cytokines

involved in angiogenesis. In the same study, a rabbit model was used to evaluate the angiogenic

effects of the bioceramics where results demonstrated more and larger vessel formation as well

as more branching. Zinc, has also demonstrated the ability to induce vascularization. In a study

performed by Hanai et al. [102] zinc upregulated VEGF release from osteoblasts via increased

FGF production.

The other half of the bone remodeling process is performed by osteoclasts. Osteoclasts

produce TRAP, which creates an acidic microenvironment that favors the dissolution of calcium

phosphate resulting in bone breakdown. In a situation where the expedited repair of a defect site

is essential, it is beneficial to mitigate the activity of osteoclasts. Results from TRAP staining

(Figure 3.9) indicate equal TRAP production over the course of 16 weeks were relatively similar

in both the pure and doped samples. While counter intuitive, other studies have shown zinc

supplementation in vivo and in vitro actually causes TRAP activity to increase while still

decreasing the amount of osteoclastic bone resorption [118,209]. It was found that zinc facilitates

an inverse relationship between TRAP and carbonic anhydrase II (CAII) production in osteoclast

cells. CAII is also a vital enzyme in the osteoclastic resorption of the extracellular matrix (ECM).

− + It acts by catalyzing hydrolysis of CO2 to generate HCO3 and H for solubilizing the ECM

inorganic phase of bone ECM. So, as Zn2+ osteoclast cells produce more TRAP, they are producing less CAII. Another study, however, showed that silica based nanoparticles were able to significantly reduce TRAP activity in osteoclasts [210], so it is possible that there is a dual

effect taking place. On the Zn2+ side, TRAP activity is increasing while decreasing the CAII

activity, and on the other side, Si4+ is decreasing the TRAP activity back to normal levels

resulting in an overall net osteoclastic resorption decrease. Although this phenomenon can be

explained in terms of past research, it would be difficult to definitively define the relationship

between SiO2 and ZnO on osteoclastic bone resorption from this study alone. More research is needed to better understand these relationships.

In addition to understanding the biological effects of the scaffolds in vivo, it is also

important to understand the degradation kinetics of the material. AAS was performed (Figure

3.10) on urine collected from the control group during the course of the 16 week study to

determine if either scaffold had degradation rates that were too high. It was found that urine

levels of Ca2+, Zn2+ and Si4+ were nearly identical, indicating that if there were an excess of ions due to scaffold degradation, they were not excreted renally. These results would indicate that degradation of the scaffolds did not reach a level that the body could not effectively utilize.

3.5.0 Conclusion The goal of this research was to demonstrate that the addition of bone-important trace elements, such as zinc and silicon, to calcium phosphate scaffolds could add the supplementary benefit of osteoinduction and angiogenesis. In this study, 3D printing was used to fabricate pure and SiO2/ZnO doped tricalcium phosphate scaffolds with a designed porosity. Scaffolds were then implanted into a bicortical femur defect in a murine model. After 4 weeks, mechanical interlocking between the implant and the host tissue was stronger than the force it took to fracture the bone for both pure and doped samples. Goldner’s trichrome histology, collagen I and osteocalcin immunostaining all indicated increased bone formation and maturation in doped samples over the course of 16 weeks. vonWillebrand Factor staining and FESEM micrographs also indicated an affinity for increased neovascularization in doped samples when compared to their pure counterparts. The addition of dopants did not affect the dissolution properties in vivo of the scaffold. Overall results, then, confirm that the addition of SiO2 and ZnO may be able to provide robust osteoinductive capabilities to CaP bone replacement materials without the need for biologics or pharmacologics.

4. CHAPTER FOUR

Effects of SiO2, SrO, MgO and ZnO dopants in TCP on osteoblastic Runx2 expression

4.1. Introduction

Calcium phosphate (CaP) ceramics are amongst the most widely studied materials for use in orthopedic implant applications due to their excellent biocompatibility, tailorable bioresorbability and compositional similarity to natural bone [62,211,212]. Current clinical uses

include, but are not limited to, coatings on metallic implants, grafting materials and bone cement

applications. [174]. The physical properties of CaPs are largely dependent on composition and

phases present in the material. Of particular interest are the CaPs with a calcium-to-phosphorus ratio of 1.5:1, known as tricalcium phosphates (TCP). TCP has a solubility product (Ksp = 2.83 x

10-30) that allows it to degrade over the course of months under physiological conditions,

allowing it to provide mechanical support while the natural healing process can take place

[62,213]. Our group and others have shown that these degradation kinetics as well as the overall

material strength can be further tailored by the addition of dopants or sintering additives

[97,99,125]. Additionally, by choosing dopants that can play important biological roles in the

bone healing process, such as strontium, magnesium, silicon and zinc, the biological interaction

between the material and natural bone can be enhanced [105,106,214].

Strontium is a non-essential element that has the ability to replace calcium in natural bone

tissue, so it is said to have bone seeking behavior[215]. It is thought to displace Ca2+ ions in

osteoblastic mediated processes. Researchers have identified that strontium likely stimulates

bone formation by activating the calcium sensing receptor (CaSR) [127,128] and simultaneously increasing osteoprotegerin (OPG) production and inhibiting receptor activator of nuclear factor kappa B ligand (RANKL) expression in osteoblasts [129] . OPG is a protein that inhibits

RANKL induced osteoclastogenesis by operating as a decoy receptor for RANKL [130]. The

OPG/RANKL ratio, then, can be a powerful regulator of bone resorption and osteoclastogenesis.

Phase III clinical trials that began in 2000 investigated the efficacy of strontium ranelate in reducing vertebral fractures and peripheral fractures, including hip fractures. After 3 years, patients treated with strontium ranelate showed significant reduction in vertebral fractures (41%) and hip fractures (36%) compared with patients treated with placebo [131].

Magnesium is one of the most abundant cations in the human body, with about 65% of it being contained in bone and teeth [121]. It has the ability to influence intracellular Ca2+

homeostasis via the parathyroid hormone pathway [216]. Furthermore, studies have shown that

magnesium deficiency can result in reduced bone growth, bone formation and mineralization

[124,166].

Silicon has been proven to be an important trace element in bone and connective tissue

formation and can stimulate biological activity by increasing the solubility of the material,

generating a more electronegative surface and creating a finer microstructure resulting in

transformation of the material surface to a biologically equivalent apatite [217]. Recent studies

have also demonstrated that silicate containing bioceramics have the added benefit of inducing

angiogenesis, which is a crucial aspect of bone defect healing [208,218].

Zinc, also an important trace element in bone formation, has been shown to have

stimulatory effects on bone formation when added to CaP [219]. Zinc is released during skeletal

breakdown and has shown to inhibit osteoclastic bone resorption [117]. It does this by inhibiting

osteoclast-like cell formation from bone marrow cells and inducing apoptosis of mature

osteoclast [167]. Osteoporotic patients have also been shown to have lower levels of skeletal zinc

than that of control groups [168]. Zinc has also been shown to induce osteoblastogenesis,

osteoblastic differentiation and mineralization [189,220].

While many benefits have been attributed to these particular elements in healthy bone biology, the mechanisms of action to which they may influence bone and the healing process is still largely a mystery. The aim of this study is to identify potential signaling pathways by which strontium, silicon, magnesium and zinc may play a role in osteoblastic differentiation. Runt related transcription factor 2 (Runx2) is essential for osteoblastic differentiation and skeletal morphogenesis and acts as a scaffold for nucleic acids and regulatory factors involved in skeletal gene expression [221]. It is vital for the maturation of osteoblasts and both intramembranous and endochondral ossification [40]. Alkaline phosphatase (ALP) has long been identified as an

osteoblast differentiation marker, signaling the maturation of the extracellular matrix (ECM)

produced by osteoblasts; the final step before terminal differentiation or apoptosis for osteoblast

cells. By studying these two targets, which are seemingly unrelated in the differentiation process,

insight may be gained as to which signaling pathways should be studied further.

4.2. Methods

4.2.1 Sample Preparation

β-tricalcium phosphate powder (β-TCP) was obtained from Berkeley Advanced

Biomaterials (Berkeley, CA) with an average particle size of 550 nm. High purity strontium

oxide (SrO) (99.9% purity) was purchased from Sigma Aldrich ( St. Louis, MO) and magnesium

oxide (MgO, 99.998%) was procured from Alfa Aesar, MA, USA. Silicon dioxide (SiO2) (99%+

purity) and zinc oxide (ZnO) (99.9%+ purity) were purchased from Fisher Scientific (Fair Lawn,

NJ). Samples were prepared by mixing 50 g of β-TCP powder and appropriate amounts of dopants (1 wt% SrO, 1 wt% MgO, 0.5 wt% SiO2 and 0.25 wt% ZnO) in 250 mL polypropylene

Nalgene bottles containing 75 mL of anhydrous ethanol and 100 g zirconia milling media with

5mm diameter. Dopant concentrations were chosen based on previous optimization research

[97,105,106,125] The mixtures were then milled for 6 h at 70 rpm to minimize the formation of

agglomerates and increase homogeneity. After milling, powder was dried in an oven at 60 °C for

72 h and pressed to discs (12 mm diameter and 2.5 mm thickness) using a uniaxial press at 145

MPa. Green compacts were then cold isostatically pressed at 414 MPa for 5 min and sintered at

1250 °C for 2 h in a muffle furnace.

4.2.2 Sample Density

Density was measured using Archimedes method. Samples were weighed initially dry

and then submerged in boiling water for 3 minutes to remove any excess air that may be trapped

in the porous structure. The samples were then transferred from the boiling water to room

temperature water, where the weight was recorded again (n=3).

4.2.3 Cell Culture

All samples were sterilized by autoclaving at 121 °C for 20 min. In this study, established human preosteoblast cell line hFOB 1.19 (ATCC, Manassas, VA) were used. Cells were seeded onto the samples in 24-well plates at a density of 105 cells/sample. The base medium for this cell

line was a 1:1 mixture of Ham's F12 Medium and Dulbecco's Modified Eagle's Medium

(DMEM/F12, Sigma, St. Louis, MO), with 2.5 mM L-glutamine (without phenol red). The

medium was supplemented with 10% fetal bovine serum (HyClone, Logan, UT) and 0.3 mg/ml

G418 (Sigma, St. Louis, MO). Cultures were maintained at 34 °C under an atmosphere of 5%

CO2 as recommended by ATCC for this particular cell line. Medium was changed every 2 days

for the duration of the experiment.

4.2.4 Cellular Proliferation

MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay was used to evaluate cell proliferation. The MTT (Sigma, St. Louis, MO) solution of 5 mg/ml was prepared by dissolving MTT in sterile filtered PBS. 10% MTT solution was then added to each sample in

24-well plates. After 2 h of incubation, 1 ml of solubilization solution made up of 10% Triton X-

100, 0.1N HCl and isopropanol was added to dissolve the formazan crystals. 100 μl of the

resulting supernatant was transferred into a 96-well plate, and read by a plate reader at 570 nm.

Triplicate samples were used in MTT assay experiments to insure reproducibility. Data was normalized to the absorbance values for the pure TCP control.

4.2.5 Cellular Morphology

Samples for testing were removed from culture after 3 days of incubation. All samples for SEM observation were fixed with 2% paraformaldehyde/2% glutaraldehyde in 0.1M cacodylate buffer overnight at 4 °C. Post-fixation was performed with 2% osmium tetroxide

(OsO4) for 2 h at room temperature. The fixed samples were then dehydrated in an ethanol series

(30%, 50%, 70%, 95% and 100% three times), followed by a hexamethyldisilane (HMDS) drying procedure. After gold coating, the samples were observed under field emission scanning electron microscope (FESEM) (FEI 200F, FEI Inc., OR, USA) for cell morphologies.

4.2.6 Alkaline Phosphatase Activity

The alkaline phosphatase (ALP) activity of the cells was determined by a spectrophotometric endpoint assay that determined the conversion of colorless p-nitrophenyl phosphate to colored p-nitrophenol. Cell media was aspirated from each sample at 7 and 11 days and samples were subsequently transferred to new 24 well plates. 350 μl 2-amino-2-methyl- propanol buffer and 500 μl p-nitrophenyl phosphate (4 mg/ml) were added. Samples were incubated for 15 min at 37˚ C and stopped by the addition of 500 μl of 2M NaOH to each well.

The solution was then transferred to a 96 well plate in triplicate and absorbance was read at 405 nm. Triplicates of each sample were examined to ensure reproducibility. Data was normalized to

MTT data to account differences in cellular proliferation/expression.

4.2.7 Runx2 Activity

Samples were fixed in 3.7% paraformaldehyde/ phosphate buffered solution with a pH of

buffered saline with 1% bovine serum albumin, 250 mM NaCl, pH 8.3) blocking solution for 1h

at room temperature. Primary antibody against Runx2 (Abcam, Cambridge, MA) was added at a

1:100 dilution and incubated at room temperature for 2 h and kept at 4 °C overnight. Samples

were then washed with TBST 3 times (10 min each). The secondary antibody, goat anti-mouse

(GAM) Oregon green 488 (Molecular Probes, Eugene, OR), was diluted 1:100 in TBST and was

used to incubate the cells for 1h. After rinsing three times for 10 minutes each with TBST the

samples placed in 24 well plates with Vectashield mounting medium (Vector Labs, Burlingame,

CA) with propidium iodide (PI). Samples were immediately analyzed by use of fluorescent area

scans using a 5x5 grid pattern on a Synergy Hybrid multi-mode microplate reader (Biotek,

Winooski, VT) with excitation wavelength of 488 nm and emission wavelength of 540 nm. Data

was normalized to account for differences in cellular proliferation/expression. Afterwards,

samples were placed on glass coverslips and confocal micrographs were taken using a Zeiss 510

laser scanning microscope (LSM 510 META, Carl Zeiss MicroImaging, Inc., NY, USA).

4.2.8 Statistical Analysis

Statistical analysis was performed using a one way ANOVA and p<0.05 was considered

statistically significant.

4.3. Results

4.3.1 Density

All samples exhibited very high density, >95%, as seen in Table 1. None of the samples

demonstrated a significant difference in density in the group.

Table 4.1. Relative densities of sintered samples.

4.3.2 Cellular proliferation

At day 3 (Figure 4.4), all samples showed decreased cellular proliferation when compared to the pure TCP control with ZnO-TCP samples having significantly lower living cells. At day 7 (Figure 4.1), both ZnO-TCP and MgO-TCP samples had significantly greater amount of live cells when compared to the pure TCP controls, while the SiO2-TCP and SrO-TCP

samples where comparable in live cell density to the pure TCP controls. By day 11 (Figure

4.2/4.5) SiO2-TCP and SrO-TCP had lower live cell density, ZnO-TCP had comparable cell

density and MgO-TCP showed increased cellular density when compared to pure TCP controls.

Figure 4.1. MTT measurements for cellular proliferation and ALP measurements for cellular differentiation at day 7 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is ormalized to pure TCP control. ALP data is normalized to MTT measurements. *P < 0.05 (n=3)

Figure 4.2. MTT measurements for cellular proliferation and ALP measurements for cellular differentiation at day 11 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. ALP data is normalized to MTT measurements. *P < 0.05 (n=3)

Figure 4.4. MTT measurements for cellular proliferation and Runx2 fluorescent measurements for cellular differentiation at day 3 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. Runx2 data is normalized to MTT measurements. *P < 0.05 (n=3)

Figure 4.5. MTT measurements for cellular proliferation and Runx2 fluorescent measurements for cellular differentiation at day 11 for SiO2, ZnO, SrO, MgO doped and pure TCP samples. Data is normalized to pure TCP control. Runx2 data is normalized to MTT measurements. *P < 0.05 (n=3)

4.3.3 Cellular morphology

Micrographs depicting cellular morphology are given in Figure 4.3. At day 3 all samples demonstrated a barely visible confluent monolayer of cells. At day 7, a second layer of cells was reaching confluency on all samples. By day 11, all samples showed multiple layers of cells with evidence of early stages of mineralization.

Figure 4.3. FESEM micrographs depicting hFOB cell morphology after 3, 7 and 11 days in culture

4.3.4 Alkaline phosphatase activity

Expression of ALP was evaluated for hFOB cells cultured on SiO2-TCP, ZnO-TCP, SrO-

TCP, MgO-TCP and pure TCP samples. Results after 7 and 11 days of culture are shown in

Figure 4.1 and Figure 4.2, respectively. At day 7, ALP activity was significantly lower in all doped samples when compared to the pure TCP samples. At day 11, ZnO-TCP and SrO-TCP

samples demonstrated higher ALP activity, while SiO2-TCP samples had decreased activity and

MgO-TCP samples had comparable activity to the pure TCP control samples.

4.3.5 Runx2 activity

Expression of Runx2 in osteoblast cells cultured on doped and pure TCP samples were

evaluated at days 3 and 11. Results for fluorescent area scans for day 3 are given in Figure 4.4

and results for day 11 are given in Figure 4.5. Confocal micrographs for each day are shown in

Figure 4.6, where cellular nuclei fluoresce in red and green indicates Runx2 expression. At day

3, SiO2-TCP, ZnO-TCP and SrO TCP all showed increased Runx2 expression when compared to

pure TCP samples, with SiO2-TCP and ZnO-TCP samples being significantly greater. MgO doped samples showed less Runx2 activity, but not at a significant level. At day 11, samples containing SiO2 and ZnO had comparable Runx2 expression to pure TCP control samples, while both SrO-TCP and MgO-TCP samples had decreased Runx2 expression. The MgO-TCP samples

demonstrated significantly lower Runx2 expression than the pure TCP samples. In general, when

compared with day 3 data, Runx2 expression decreased for all samples with respect to pure TCP.

Confocal micrographs confirm the results of the quantitative fluorescent area scans, showing

similar patterns of Runx2 expression.

Figure 4.6. Confocal micrographs showing Runx2 expression in hFOB cells at 3 and 11 days. Green fluorescence indicates active Runx2; red fluorescence indicates propidium iodide bound to the nuclei of cells.

4.4. Discussion

Previous studies have well characterized the inherent material properties associated with these dopants [6,7,9,10]. The use of isostatic pressing, eliminates most of the inherent density differences associated with dopant addition, resulting in a very high average density ranging from 95% - 98% (Table 4.1) with no statistically significant differences detected. Studies have also shown that the addition of any single dopant used in this study has the ability to reduce α phase formation associated with a high sintering temperature when compared to the pure TCP samples [97,105,106,125]. These two properties are extremely important when thinking about potential differences in dissolution between sample types. The pure control composition has higher alpha phase so its dissolution rate will be higher than its doped experimental samples if the densities are similar, however, dissolution rate will be minimal considering the time frame of

91 the study as seen in previous studies . It is well documented that increased Ca2+ and P5+ can positively affect cellular differentiation and proliferation [222–224], but the results observed in this study demonstrated that samples containing dopants had significant positive effects when compared to the pure composition. It is then very likely, then, that the cells are responding to the dopants rather than Ca2+ or P5+ due to differences in dissolution.

The osteoblastic cell cycle is a process comprised of three distinct stages (Figure 4.7): proliferation, maturation and termination. During the proliferation stage and part of the maturation stage, Runx2 plays a stimulatory role in differentiation of mesenchymal stem cells

(MSCs), osteochondrogenic precursors and preosteoblasts. During the latter portion of the maturation stage and the terminal differentiation stage, Runx2 plays an inhibitory role in osteoblastic differentiation [43]. It is, therefore, necessary for the upregulation of Runx2 initially and the downregulation of Runx2 during the latter part of the cell cycle to obtain mature bone growth. Results from the MTT assay at day 3 (Figure 4.4) suggest an expeditive transition from the proliferation stage of the osteoblast life cycle to the maturation stage in samples containing

SiO2, ZnO and MgO as evidenced by the decrease in proliferation rates. Samples doped with SrO are comparable to that of the pure TCP samples. Research has shown that Sr is able to augment proliferation of osteoblast cells while maintaining increased differentiation rates [225], so it is possible that cells seeded on the SrO doped samples are also in the maturation phase. SEM micrographs (Figure 4.3) of cells at day 3 show characteristics of mature osteoblasts in all samples manifested by the broadening and flattening of cells as well as the extension of cellular processes and cellular aggregation. At day 7 (Figure 4.1), MTT results show increased proliferation in samples containing ZnO and MgO compared to pure TCP. The other samples had

92 similar values to pure TCP, indicating that cells on the pure TCP samples had reached advanced maturity and the resultant slowing of proliferation. The increased level of viable cells in the

MgO and ZnO doped samples can be attributed to the fact that Zn2+ and Mg2+ are both mitogenic factors and may prevent terminal apoptosis of osteoblast cells [226,227]. SEM micrographs show the formation of multiple layers of mature osteoblast cells on all samples. Day 11 MTT results

(Figure 4.2/4.5) indicate an affinity for active terminal differentiation in samples containing

SiO2 and SrO demonstrated by decreased cellular viability (increased terminal apoptosis).

Samples containing MgO had increased proliferation rates, while ZnO doped samples were comparable to pure TCP indicating a lasting effect of Mg2+ as a mitogenic factor. SEM images

(Figure 4.3) for day 11 show all doped samples demonstrate advanced characteristics of cellular differentiation when compared to pure TCP samples, embodied by the sheet-like formation of bone lining cells, where single cells are often difficult to distinguish, as well as apoptosis and mineralization.

Figure 4.7. Diagram depicting the osteoblast cell lifecycle and influence of Runx2 [228,229].

Results from the ALP assay reveal that at day 7 (Figure 4.1), all samples show

significantly lower activity when compared to pure TCP. At day 11 (Figure 4.2), ALP activity in

SiO2 doped samples was less than pure TCP, MgO doped samples comparable and ZnO and SrO

doped samples significantly greater. Alkaline phosphatases are glycosyl-phosphatidylinositol

anchored, Zn2+metallated glycoproteins that are released during the beginning of the termination

phase of the osteoblastic cell life cycle and help to catalyze the hydrolysis of phosphomonoesters

into inorganic phosphates [52]. They create an alkaline environment that favors the mineralization of these inorganic phosphates. These results suggest a prolonged maturation phase in the ZnO and MgO doped samples through day 7, likely caused by increased Runx2 activity in the samples through this time period. Researchers have also shown that the Zn2+ binding sites in

ALP may be easily substituted by other cations, where differing cationic radii and charge can act

as destabilizing forces in the enzyme causing decreased functionality [230]. This phenomenon

becomes apparent when considering day 11 results. ZnO and SrO have a large increase in

functional enzymatic ALP activity while SiO2 is continuing to show destabilization when

compared to the pure TCP samples. Zn2+, being the native cation in ALP, is expected to show increased activity. While the ionic radii of Sr2+ is dissimilar to Zn2+ (132pm and 88 pm,

respectively), Sr2+ seems to act as an agonist for ALP activity. Si4+, likely due to vast differences

in ionic radii (54 pm) and charge, may cause destabilization of the ALP enzyme. Mg2+ is very similar in size (86 pm) and charge to Zn2+ and has been demonstrated to have very little effect on

the stability of ALP [230], so it was expected that results would be similar to samples doped with

ZnO. This was not the observation, however. It is apparent that MgO doped samples are only producing as much ALP as the pure samples at day 11where we expected to see an increase in

activity. Two different mechanisms may be responsible for these findings. The first, we still see

significant increase in proliferation rates with samples doped with MgO, so it may be that

samples remain in the early maturation stage. Runx2 findings, however, do not correlate well

with this. They should remain at high levels during states where proliferation is prevailing. The

second mechanism would suggest that Mg2+ is well regulated along the differentiation pathways

(as we still see increases in cellular proliferation). This seems to be the more likely scenario as

Mg2+ is one of the most abundantly available cations in the human body and is expected to be

more highly regulated.

Results from this study indicate that at day 3 (Figure 4.4), Runx2 expression is

significantly upregulated in samples containing SiO2 and ZnO. SrO doped samples also showed

upregulation of Runx2, but not quite at a significant level. At day 11 (Figure 4.5), samples containing SiO2 and ZnO show Runx2 expression to be about the same as pure TCP. Samples containing MgO and SrO show downregulation of Runx2, with MgO doped samples giving significant results. Confocal microscopy (Figure 4.6) confirms these findings. Overall results signify that samples containing SiO2, ZnO and SrO can appropriately modulate Runx2 activity

leading to accelerated differentiation of osteoblast cells when compared to pure TCP samples. A

proposed signaling mechanism for the effect of dopants on Runx2 expression is offered in

Figure 4.8. One of the major regulators of osteoblastic cellular differentiation, and in particular

Runx2, is the canonical Wnt signaling pathway[40,231]. The non-canonical Wnt5 calcium-

dependent signaling pathway may also play an important role in the osteoblastic differentiation

process and has been shown to cross paths with other Wnt isoform signaling pathways [232].

Several studies have presented that dopants, such as Sr2+, can interact with the calcium sensing

receptor (CaSR) in osteoblastic cells to moderate cellular processes [127,233]. In the presence of high extracellular Ca2+, Wnt5 is activated through the CaSR and binds to the Frizzled G protein- coupled receptor and starts the intracellular Ca2+ signaling cascade [234]. There are extensive reviews available concerning this cascade [39,231]. Briefly, once Wnt is bound, the cytosolic G protein breaks off into its α and β/γ subunits. The β/γ then activates membrane bound phospholipase C (PLC) which, in turn, hydrolyzes phosphatidylinositol 4,5-bisphosphate (PIP2) into second messengers inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 then diffuses through the cytosol and activates calcium channels located in the cellular membrane and endoplasmic reticulum causing an influx of intracellular calcium. Studies have shown that calcium channels can be non-selective to certain ions, including Sr2+, which may allow entrance

of the dopants into the cell[235]. Other studies also suggest that the endoplasmic reticulum may

also allow for the storage of cations such as Mg2+, Sr2+ and Zn2+ [236–238]. The influx of Ca2+

into the cell leads to the activation of calcineurin (CaN) and Ca2+-calmodulin-dependent protein kinase II (CamKII), while DAG activates protein kinase C (PKC). PKC activates the translocation of cyclic adenosine monophosphate (cAMP) response element-binding (CREB) and nuclear factor kappa B (NFκB) from the cytoplasm to the nucleus. CamKII also plays a role in the translocation of NFκB. CaN is responsible for the permutation of Nuclear factor of activated T-cells (NFAT) into the nucleus. CREB, NFκB and NFAT have all been shown to play important roles in the regulation of Runx2 [239–241]. The ability of cations to act as a

competitive agonist with Ca2+ for the activation of CaN or CamKII has been established with strontium. Fromique et al. demonstrated that strontium ranelate induced NFATc1 nuclear translocation and was completely abrogated with CaN inhibitors[242]. It is likely that other

dopants may interact in a similar manner, although confirmation studies are still required to

confirm this.

A recapitulating theme throughout this study is the apparent inactivity of Mg2+ in the osteoblastic cellular differentiation process as evidenced by results from Runx2 and ALP analysis. In the human body, Mg2+ is the second most abundant cation and is known to be a

cofactor in over 300 enzymatic reactions ranging from energy metabolism to nucleic acid

synthesis [243]. Because it is so abundant, it is also highly regulated and can even act as a Ca2+

ion channel blocker [244]. Another study demonstrated that while some calcium ion channels

may be non-selective to some cations, Mg2+ efflux remained low through these same channels

[235]. While dietary deficiencies have been linked to osteoporosis, this study provides no

evidence that an abundance of Mg2+ will help in the maturation and differentiation of osteoblastic

cells. On the other hand, significant results were seen in cellular proliferation at days 7 and 11,

2+ confirming Mg as a valuable mitogenic element. SrO, ZnO and SiO2 all seemed to have

positive effects on Runx2 regulation, increasing early expression and decreasing at the latter

stages of the cell life cycle indicating the ability for accelerated differentiation when compared to

pure TCP samples. The mineralization capabilities of the osteoblast cells were also positively

affected by the ZnO and SrO dopants, demonstrated by ALP activity at day 11. Overall results

suggest that ZnO, SrO and SiO2 dopants can have profound effects on the differentiation

capabilities of osteoblast cells in vitro and likely act through the intracellular Ca2+ signaling

pathway as a competitive agonist with calcium to activate key proteins in this pathway. While

Si4+ was found to cause functional destabilization of ALP, a key enzyme in the mineralization process of osteoblasts, it has also been shown that it is an important trace element in bone and connective tissue formation and can stimulate biological activity by increasing the solubility of the CaP material, generating a more electronegative surface and it can also create a finer microstructure resulting in a nucleation site that is favorable for apatite formation [217,245].

4.5. Conclusions

CaP materials have been widely studied for use as an osteogenic bone replacement

material and are effectively used in several medical and dental applications today. In this study,

by incorporating the dopants SiO2, ZnO and SrO into TCP, the important transcription factor

Runx2 was able to be modulated in a beneficial manner as to induce accelerated bone cell differentiation, with high expression in the early stages of cell maturation and low expression in the terminal stage. ALP activity was increased with the introduction of ZnO and SrO into the samples when compared to pure TCP indicating increased affinity for matrix mineralization.

ZnO and MgO dopants both demonstrated favorable mitogenic properties, but Mg did not have any significant effect on the differentiation markers. In addition to these results a signaling mechanism was proposed as to the action of these dopants for consideration of further study and understanding of the system.

5. CHAPTER FIVE

Effects of SiO2 and ZnO dopants in porous TCP scaffolds on osteoblastic

differentiation

5.1 Introduction

Calcium phosphate ceramics have been widely used in the orthopedic industry as

synthetic grafting, coating and cement materials. They have a compositional similarity to natural

bone and excellent biocompatibility that make them ideal as synthetics [213,246]. Two of the

major classes of calcium phosphates that are most often used and researched are hydroxyapatite

(HA) and β-tricalcium phosphate (β-TCP). Because the solubility of hydroxyapatite is extremely

-59 low in physiological conditions (Ksp = 2.35 x 10 ), it is most often used in coating applications, where a long functional lifetime is necessary, or as a biphasic system as a mixture with β-TCP to

-29 increase the strength and longevity of grafting materials [62,247,248]. β-TCP (Ksp=1.25 x 10 )

is about 30 orders of magnitude more soluble than HA, which makes it considered to be

bioresorbable [62]. While HA may last in the human body for over 10 years, β-TCP has a

functional lifetime of 12-16 months and is generally completely replaced by natural bone after 3

years [249–251]. β-TCP, then, is an ideal material to fill small bony voids created during

surgery or for spinal fusion and non-union fracture healing.

100

β-TCP is considered by many to be a osteoconductive material. It provides support for new tissue formation and the migration of bone forming cells during the healing process. Recent trends in technology, however, demonstrate a paradigm shift from designing a simply functional material to a fully bioactive material. It has been the goal of many researchers and industry leaders to further develop β-TCP grafting materials to give them osteoinductive properties, or the ability to actually stimulate new bone formation. In a study by Lindhorst, et. al., VEGF was loaded onto β-TCP scaffolds and was found to increase neovascularization in vivo, a vital characteristic of the bone healing process [252]. Several studies have outlined the efficacy of recombinant human bone morphogenic protein 2 (rhBMP-2) loaded TCP scaffolds, demonstrating enhancements of human adipose derived stem cells in vitro in as few as 15 minutes and significantly increased new bone formation in vivo [253,254]. Other growth hormones often investigated include fibroblast growth factors (FGFs) and transforming growth factors (TGFs) all with similar positive results [255,256]. While much of these results have been extremely positive, growth hormone products have come under severe scrutiny of the public and the FDA due to several serious possible side effects experienced in clinical off label uses [257–

259].

An alternative approach to adding osteoinductive capabilities to β-TCP has been through the use of trace element nutrients that are vital to bone development and health such as silicon, zinc, magnesium and others. The use of such additives can alter the physicochemical properties of β-TCP such as compressive strength, strength degradation, grain size and density

[104,106,260]. Furthermore, biological response has been shown to be enhanced in vitro and in vivo [100,261–263].

101

Silicon has long been understood to play an important role bone and connective tissue

biology [172]. In a study by Seaborn and Nielsen, silicon deprived rats were shown to have significantly decreased collagen formation in bone and wound healing [203]. Another study

noted positive correlation with dietary silicon uptake and bone mineral density in the lumbar of

men and premenopausal women [264]. In a similar study, 53 osteoporotic women demonstrated

significantly increased femoral bone mass density in the silicon treated group [166]. Silicon has

been widely studied as a sintering additive to β-TCP. It is known to increase densification by

impeding the β to α phase transformation in TCP [66]. α-TCP is often an undesirable phase that

-26 appears when sintering at temperatures greater than 1125 ˚C. Because α-TCP (Ksp = 3.16 x 10 )

has a much lower solubility product than β-TCP, it degrades much faster in physiological conditions [62]. In addition to the osteogenic effects of silicon, it has also demonstrated angiogenic capabilities. Human dermal fibroblasts indicated upregulated VEGF production in response to calcium silicates, which had a downstream effect on the nitric oxide synthase and nitric oxide production in human endothelial cells [115]. Nitric oxide synthase and nitric oxide

production are key regulators of new blood vessel formation.

Another trace element of importance in bone biology is zinc. It is one of the ions that is

released during the bone remodeling process and plays a key role in the stability and activity of

alkaline phosphatase (ALP) [117]. ALP is an enzyme that is released from osteoblasts after depositing the collagen based extracellular matrix. Its key function is to catalyze the hydrolysis of phosphomonoesters, released during osteoclastic breakdown of bone, into inorganic phosphates that can mineralize the matrix [265]. Excess zinc released during the remodeling

process is believed to play a role in the regulation of osteoclast activity as well as stimulation of

102

osteoblast activity [118,119]. Zinc doped calcium phosphate materials have also been shown to

increase new bone formation in animal models [98,120].

In this study, highly porous TCP constructs were doped with SiO2 or ZnO in order to

investigate late stage differentiation markers in human osteoblast cells as well as the physicochemical properties of the scaffolds. While the beneficial effects of silicon and zinc have been demonstrated in various studies, both in vitro and in vivo, little is known as to their primary

mechanism of action when compared to commonly used growth factors.

5.2 Methods

5.2.1 Sample Preparation

β-tricalcium phosphate powder (β-TCP) was synthesized in house via a solid state

method. Briefly, 2 moles of dicalcium phosphate (CaHPO4) were mixed with 1 mole of calcium

carbonate (CaCO3). Powders were mixed for 2 hours in a 5:1 milling media to powder weight

ratio, then heated to 1050 ˚C for 24 hours let cool to room temperature.

High purity silicon dioxide (SiO2) (99%+ purity) and zinc oxide (ZnO) (99.9%+ purity)

were purchased from Fisher Scientific (Fair Lawn, NJ). Doped powders were prepared by

mixing 20 g of β-TCP powder and appropriate amounts of dopants (0.5 wt% SiO2 and 0.25 wt%

ZnO) in 250 mL polypropylene Nalgene bottles containing 30 mL of anhydrous ethanol and 100

g zirconia milling media with 5mm diameter. Dopant concentrations were chosen based on

103

previous optimization research [97,105,106,125]. The mixtures were then milled for 6 h at 70

rpm to minimize the formation of agglomerates and increase homogeneity. After milling, powder

was dried in an oven at 70 °C for 72 h. Samples were prepared using an oil emulsion method

previously described [82]. Briefly, 20g of β-TCP powder (doped or pure) were mixed with 20g

paraffin oil purchased from Sigma Aldrich (St. Louis, MO) and 13.4 mL of an emulsifier

solution of 0.14g/L Kolliphor EL purchased from Sigma Aldrich (St. Louis, MO) in 0.2M

Na2HPO4. The mixture was stirred at 2000 rpm for 45 seconds to yield a stable emulsion slurry.

The slurry was then pipetted into disk (11.43 mm diameter and 6.35 mm thickness) or cylinder shaped molds (6 mm diameter and 12 mm height) and let dry at 70 °C overnight. Samples were then removed from the molds and sintered at 1250 °C for 2 h in a muffle furnace.

5.2.2 Microstructure, Phase Analysis and Mechanical Properties

The surface morphologies of the sintered scaffolds were observed under a field emission scanning electron microscope (FESEM) (FEI Inc., OR, USA) before and after 28 days in PBS.

Phase analysis of sintered TCP samples with and without dopants was carried out by X-ray diffraction (XRD) using a Siemens D-500 X-ray powder Diffractometer (Siemens AG,

Karlsruhe, Germany) with CuKα radiation and a Ni-filter. Each run was performed with 2θ

values between 10° and 60° at a step size of 0.1° and a count time of 1 s per step. Density was

measured using Archimedes method. Samples were weighed initially dry and then submerged in

boiling water for 3 minutes to remove any excess air that may be trapped in the porous structure.

The samples were then transferred from the boiling water to room temperature water, where the

weight was recorded again (n=3). Bulk densities were also measured for the same samples for

104

comparison. Compressive strengths of undoped and doped TCP scaffolds were determined using

a screw-driven universal testing machine (AG-IS, Shimadzu, Japan) with a constant crosshead

speed of 0.33 mm/min. Compressive strength was calculated using the maximum load recorded

and the sample dimensions. Compressive strength was tested on at least five samples for each

composition.

5.2.3 Ca2+ Ion Release

Ca2+ content of samples incubated in PBS was analyzed over 28 days using a Shimadzu

AA-6800 atomic absorption spectrophotometer (Shimadzu, Kyoto, Japan). Standard solutions

were freshly prepared in ionization buffer to obtain a final concentration of 0-1000 μg/ml.

Calcium ionization buffer standards were purchased from High-Purity Standards (Charleston,

SC, USA).

5.2.4 Cell Culture

All samples were sterilized by autoclaving at 121 °C for 20 min. In this study, established

human preosteoblast cell line hFOB 1.19 (ATCC, Manassas, VA) were used. Cells were seeded

onto the samples in 24-well plates at a density of 2 x 105 cells/sample. The base medium for this cell line was a 1:1 mixture of Ham's F12 Medium and Dulbecco's Modified Eagle's Medium

(DMEM/F12, Sigma, St. Louis, MO), with 2.5 mM L-glutamine (without phenol red). The medium was supplemented with 10% fetal bovine serum (HyClone, Logan, UT) and 0.3 mg/ml

G418 (Sigma, St. Louis, MO). Cultures were maintained at 34 °C under an atmosphere of 5%

105

CO2 as recommended by ATCC for this particular cell line. Medium was changed every 2-3 days for the duration of the experiment.

5.2.5 Cellular Morphology

Samples for testing were removed from culture after 21 and 28 days of incubation. All samples for SEM observation were fixed with 2% paraformaldehyde/2% glutaraldehyde in 0.1M phosphate buffer overnight at 4 °C. Post-fixation was performed with 2% osmium tetroxide

(OsO4) for 2 h at room temperature. The fixed samples were then dehydrated in an ethanol series

(30%, 50%, 70%, 95% and 100% three times), followed by a hexamethyldisilane (HMDS)

drying procedure. After gold coating, the samples were observed under field emission scanning

electron microscope (FESEM) (FEI 200F, FEI Inc., OR, USA) for cell morphologies.

5.2.6 RNA Extraction and real time RT-PCR

RNA was extracted from samples using Aurum Total RNA Mini Kit spin columns from

BioRad (Hercules, CA) and the manufacturer’s recommended procedure. Three biological replicates were used in this study with each having three technical replicates. First strand cDNA was synthesized using iScript Advanced cDNA Synthesis Kit for RT-qPCR (Biorad, Hercules,

CA) in 20 µL reactions according to manufacturer’s recommended procedure. RT-qPCR was performed under standard enzyme and cycling conditions on a CFX Connect Real Time PCR

Detection System (Biorad, Hercules, CA). Primer sets were pre-validated PrimePCR SYBR

Green Assays from Biorad (Hercules, CA) for BMP-2 (assay ID: qHsaCID0015400), runt related

106

transcription factor 2 (Runx2) (assay ID: qHsaCID0006726), osteoprotegerin (OPG) (assay ID:

qMmuCID0027158) and Receptor activator of nuclear factor kappa-B ligand (RANKL) (assay

ID: qHsaCID0015585). Two housekeeping genes were also used in this study β-actin (assay ID: qHsaCED0036269 ) and ribosomal protein, large, P0 (assay ID: qHsaCED0036271). Data analysis was performed using the BioRad CFX Manager Software 3.0 (Hercules, CA)

Expression levels of the gene of interest were normalized to the housekeeping genes and data reported is the normalized expression given by 2 . −∆∆퐶푡

5.2.7 Statistical analysis

Statistical analysis was performed using one way ANOVA and p<0.05 was considered

statistically significant.

5.3 Results

5.3.1 Physical and Mechanical Characterization

XRD (Figure 5.1) was performed to determine the CaP phases of the sintered scaffolds

and confirmed that in both samples, β-TCP was the primary phase present after sintering at 1250

°C. α-TCP phase was found to be present in the pure samples, but drastically reduced in the

samples containing dopants. Zn-TCP had the lease amount of α-TCP, while pure TCP the most.

The characteristic peaks of β-TCP and α-TCP match well with JCPDS # 09-0169 (β-TCP) and

09-0348 (α-TCP), but show about a 1.57 degree peak shift in Zn-TCP samples and a 1.79 degree

peak shift in the Si-TCP samples.

107

Figure 5.1. XRD patterns of doped and pure scaffolds sintered at 1250 ˚C. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP)

Bulk density (Figure 5.2) of the samples demonstrated that all samples were over 65% porous. Pure TCP had a relative bulk density of 45.14 ± 0.03 %, Si-TCP of 32.81 ± 0.02 % and

Zn-TCP of 32.4 ± 0.02 %. Both of the doped samples had significantly lower density than the

pure samples. Apparent density (Figure 5.3), as measured by Archimedes’ method showed that

the estimated pore wall density in the doped samples were significantly higher than that of the

pure TCP samples. Zn-TCP had the highest apparent density at 82.36 ± 0.02%, then Si-TC at

79.64 ± 0.02% and Pure TCP with the lowest at 66.05 ± 0.12%.

108

Figure 5.2. Bulk density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3).

Figure 5.3. Apparent density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3).

109

Pure TCP and Si-TCP had similar compressive strength at 3.31 ± 0.85 MPa and 3.50 ±

1.61 MPa, respectively. Zn-TCP had lower compressive strength, 2.06 ± 0.64 MPa, than Pure

TCP at a significant level. Data is shown in Figure 5.4.

Figure 5.4. Compressive strength of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 5).

FESEM micrographs taken before and after 28 days (Figure 5.5) in PBS demonstrate the

highly porous nature of the scaffolds. Both of the Pure-TCP and Zn-TCP samples had identifiable grains, but the Si-TCP samples did not. All samples did show evidence of liquid phase sintering, characterized by the flowing particle structures. All samples afte r 28 days in

PBS, showed considerable surface degradation, but the microstructure remained largely

unchanged. All samples also had significant plate-like apatite formation on the surface of the

110

samples. Si-TCP samples showed the most apatite formation and the pure TCP samples had the

least amount of apatite.

Figure 5.5. Surface morphology of sintered scaffolds of pure TCP composition, Si-TCP and Zn-TCP compositions after sintering and after 28 days in PBS.

Ca2+ ion release results (Figure 5.6) indicated both of the doped compositions had higher

degradation rates than the pure scaffolds. In total, Si-TCP samples released 1.74 mg of Ca2+ over

28 days. Zn-TCP samples had an average 1.83mg Ca2+ over 28 days, while the pure samples had

only 1.15 mg Ca2+ release. All samples had a steady state of release over the testing period.

111

Figure 5.6. AAS results for Ca2+ ion concentration in PBS collected over 28 days.

5.3.2 Cellular Morphology

Cells on all samples demonstrated sheet-like structure after 21 days of culture (Figure

5.7). While single cells were difficult to identify on doped samples, the pure samples had many indicating a higher proliferation rate in the Si-TCP and Zn-TCP samples. The sheet-like phenotype is consistent with bone lining cell morphology, indicating that after 21 days, many cells had already reached terminal differentiation. Results for day 28 were similar.

112

Figure 5.7. FESEM micrographs depicting hFOB cell morphology after 3, 7 and 11 days in culture

5.3.3 Gene Expression Analysis

At day 21 (Figure 5.8) Si-TCP and Zn-TCP expressed significantly less BMP-2 than the

pure samples. OPG expression was upregulated significantly in the doped samples, while a

significant downregulation of RANKL was noted when compared to the pure samples. Runx2

expression was similar at day 21 for all samples, with slightly elevated expression measured in

the Si-TCP samples. At day 28 (Figure 5.9), similar BMP-2 expression was seen in all samples and OPG remained elevated only in the Zn-TCP samples. RANKL expression was elevated for

Si-TCP samples, while Zn-TCP was similar to the pure composition. Runx2 expression in Si-

TCP was similar to Pure TCP, while Zn-TCP experienced a significant increase.

113

** ** **

Figure 5.8. RT-qPCR data for day 21 (* P < 0.05, where n = 3; **P<<0.01).

** *

Figure 5.9. RT-qPCR data for day 28 (* P < 0.05, where n = 3; **P<<0.01

114

5.4 Discussion

The use of silicon and its result as a stabilizing agent in CaPs has been well documented

[266–268]. In this study, XRD results (Figure 5.1) show that α-TCP phase formation is significantly reduced in samples containing silicon. The same result is identified in samples doped with ZnO, but the defects created in the TCP lattice structure are not similar. It is believed that Si4+ will replace P5+ in the TCP lattice with charge compensation created by either O2-

vacancies or the presence of excess Ca2+ [269]. Zn2+, however, is most likely to replace Ca2+ in

Ca(5) position in the lattice structure where it is believed that the Ca(5)O6 polyhedron may be

strained due to over-bonding and would favor the incorporation of smaller cations such as zinc

[270,271]. Both dopants seem to stabilize the crystal structure while sintering at high

temperatures (>1150 ˚C) where α-TCP is known to form. A slight peak shift is detectable in both

doped samples, verifying that substitution defects are likely taking place.

Density analysis (Figures 5.2 and 5.3) is in agreement with the XRD results. α-TCP

phase formation is known to coincide with a stage of rapid grain growth and overall decrease in

the densification, so decreased amounts present would conceivably result in higher densification

[66,170]. Results from this study indicate that the pore wall density (estimated from Archimedes

density) of the doped samples were over 79%, while the pure composition was about 66%, with a

trend of increased α-TCP phase present resulting in decreased densification. Microstructural

analysis by SEM (Figure 5.5) verifies that all the samples had greater than 55% open pores with

most of the porosity being micropores. While all samples had evidence of liquid phase

formation, Si-TCP samples exhibited a significantly increased amount. Silicon (111 pm) has a

115

slightly larger atomic radius than that of phosphorous (98 pm), which causes bond elongation and strain in the crystal structure that could result in a decrease in liquidus temperature and increase in liquid phase sintering [272].

The mechanical strength of the Pure TCP and Si-TCP compositions were similar (~3

MPa) with Zn-TCP significantly lower than either (~ 2MPa). Although previous studies have reported decrease in initial compressive strength of ZnO doped compact TCP samples [97], this decrease is more likely due to slight differences in the emulsion formation during processing and resultant variances in pore size and distribution rather than an inherent physicochemical property associated with ZnO doping in this system. The driving factor in the compressive strength in this study is far more weighted by the large amount of porosity than physical effects of the dopants.

After 28 days in PBS (Figure 5.5) solution at 37 ˚C, all samples showed signs of surface

degradation, but not major changes to microstructure. Cumulative Ca2+ ion release over the 28

day period showed significantly higher Ca2+ initially in the doped samples, but average weekly

release rates were not significantly greater than the pure samples. It is hypothesized that initial

calcium release rates were greater in the doped samples based on increased open porosity and

consequently increased surface area in contact with the PBS (Figure 5.2). The only calcium

phosphate product that is stable in solution at pH greater than 4.2 is hydroxyapatite (HA) [273].

The dissolution-precipitation process of β-TCP has been described as a diffusion limited process,

which will decrease as an interfacial apatite layer is formed on the surface of the sample [274]. In

this study, all samples showed a relatively steady degradation rate, with no sign of decreased

kinetics. While all samples had apatite formation, Si-TCP and Zn-TCP had significantly more

116 surface apatite formation than the pure samples indicating a faster rate towards the dissolution- precipitation equilibrium.

The cell culture performed in this study was designed so that all samples would reach a confluent monolayer before testing. This is an important factor when studying differentiation.

The osteoblast cells go through three stages in their lifecycle: proliferation, maturation and differentiation. The early stage is focused on population and recruitment which is required for the maturation stage. During the maturation stage cells are laying the foundation for new bone called the extracellular matrix (ECM). Once the ECM is created, cells will begin terminal differentiation or undergo apoptosis [228,229]. The osteoblastic cell lifecycle SEM images at day

21 (Figure 5.7) confirm that a monolayer was achieved in all samples, though differentiation to bone lining cells seemed to be occurring in the doped compositions with more regularity. Bone lining cells have a flattened morphology and are generally considered inert in function. They are characterized by their sheet-like appearance and difficulty determining one cell from the next.

They are thought to form the endosteum on trabecular and endosteal surfaces and underlie the periosteum on the mineralized surface [5]. Functionally, the may play a role in the regulation of mineral ions in and out of the bone extracellular fluid, acting as a blood-bone barrier, and are able to dedifferentiate into osteoblasts upon exposure to parathyroid hormone or mechanical stresses [11]. Bone lining cells also play a key role in the remodeling process of bone by forming a circulatory canopy or microenvironment for the remodeling to take place [21]. Results were similar for Day 28, with distinguishable phenotypical osteoblast visible only in the Pure TCP samples.

117

RT-PCR results at day 21 (Figure 5.8) confirm visual analysis from SEM images at day

21. Low levels of BMP-2 in the doped samples indicate that the differentiation process has already likely occurred in the osteoblasts. BMPs are potent osteoblastic differentiation factors that make up the largest family in the transforming growth factor – β (TGF- β) superfamily

[275]. BMPs are known to activate specific bone morphogenic protein receptors (BMPRs).

Receptor regulated Smad proteins are then recruited by the activated receptors, where they help to propagate the BMP signal to target genes such as alkaline phosphatase (ALP), bone sialoprotein, osteocalcin and Runx2 [276,277]. In addition to the canonical Smad activation of

BMPs, they are also subject to several intracellular signaling molecules such as extracellular signal-regulated protein kinase (Erk), p38 mitogen-activated protein kinase (MAPK), phosphatidylinositol 3-kinase (PI-3K), and protein kinases C and D (PKC, PKD) [278–280].

BMP-2 has been shown to be upregulated by increases in extracellular Ca2+ [281].

Day 21 results also show that Runx2 transcriptional activity is similar with the Pure TCP

control group. Runx2 is a transcription factor that is absolutely vital to healthy bone growth. It is

a master regulator of mesenchymal stem cell (MSC) differentiation into osteoblasts and also

regulates the expression of genes that are necessary for osteogenesis such as ALP, osteopontin,

collagen type I, and osteocalcin [43,282]. In addition, studies have shown that mice with induced

Runx2 deficiencies completely lack mineralized bone [283]. It has also been demonstrated that

Runx2 expression needs to be upregulated for differentiation of MSCs into osteoblast, but

downregulated for terminal differentiation of osteoblasts into bone lining cells and osteocytes

[228]. Known to be regulated by the Wnt singaling pathway as well as intracellular calcium

signaling, Runx2 has been shown to be modified post-translationally by phosphorylation,

118

acetylation and ubiquitination which alters the transcriptional activity and stability of Runx2

[231,284,285].

Osteoblasts play dual role in the bone remodeling process. Aside from depositing the

ECM, they can regulate osteoclastogenesis via OPG and RANKL production. After successfully

depositing the new matrix, osteoblasts need to send a signal to monocytes to begin the formation

of merged and multinucleated osteoclasts. This signal has been identified as RANKL and is

considered the primary cause of osteoclastogenesis [10]. OPG is a decoy receptor for RANKL

that is produced by osteoblasts. Generally, when RANKL is upregulated it is associated with a

downregulation of OPG, so as to favor osteoclastogenesis and the ratio is considered to be a

major determinant of bone mass [10]. Day 21 results from this study indicate that osteoblasts

cultured on the doped samples caused a delay in the shift of the OPG:RANKL ratio that would

result in continued bone building and the delay of osteoclastogenesis. The Pure TCP samples at

day 21, however, indicate the opposite response where high RANKL expression and low OPG expression would result in the induction of bone remodeling by osteoclast cells. It has been verified that OPG and RANKL are both regulated by the Wnt signaling pathway, which is one of the main pathways in osteogenesis, as well as by intracellular calcium signaling [130,284,286].

Runx2 has also been discovered to be an important regulator of RANKL and OPG, where high

levels of Runx2 will suppress OPG and upregulate RANKL [287].

Day 28 results (Figure 5.9) show a decrease in BMP-2 and RANKL expression for pure

TCP as well as an increase in RANKL expression for Si-TCP and Zn-TCP had a slightly smaller

119

OPG:RANKL ratio. These results indicate that the osteoblasts on all samples were mostly differentiated into relatively inactive bone lining cells.

When all of the results are put together, they begin to tell an interesting story. The cells on the Pure TCP samples are undergoing the differentiation process, indicated by high BMP-2 levels, moderate Runx2 expression and a low OPG:RANKL ratio and the phenotypical osteoblast from micrographs. The osteoblasts cultured on both the Si-TCP and Zn-TCP seemed to already have gone through differentiation with phenotypical bone lining cells as seen in the

SEM micrographs and as a result have much lower BMP-2 levels and stable Runx2 expression.

Interestingly, though, we see that the OPG:RANKL ratio remains relatively high indicating an affinity for prolonged bone growth processes and delayed remodeling processes. By day 28, samples seemed relatively inactive, which is typical of bone lining cells. In a previous study, we found that the use of dopants in compact TCP samples effectively modulated post-translational

Runx2 activity, having elevated levels during the proliferation stage and decreased levels during the maturation stage of the osteoblast lifecycle [260]. In the same study a hypothesis was made based on extensive literature review as well as results from the study that the effect of dopants is likely to act through Ca2+ mediated pathways such as the non-canonical Wnt5 calcium dependent pathway. The results from this study also fit well into this hypothesis. Briefly, in the presence of high extracellular Ca2+, Wnt5 is activated through the calcium sensing receptor (CaSR) and initiates the intracellular Ca2+ signaling cascade [234]. Once Wnt is bound, membrane bound phospholipase C (PLC) and hydrolyzes phosphatidylinositol 4,5-bisphosphate (PIP2) into second messengers inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 will then activate the calcium channels in the cellular membrane causing an influx of intracellular calcium and

120

dopant ions. The influx of ions into the cell leads to the activation of calcineurin (CaN) and Ca2+-

calmodulin-dependent protein kinase II (CamKII), while DAG activates PKC. Because OPG has

been shown to be regulated by the Wnt pathway, intracellular Ca2+ signaling and Runx2,

RANKL by Runx2 and intracellular Ca2+ singaling, BMP-2 by PI-3K and PKC and Runx2 by intracellular Ca2+ signaling, they all have a common relationship with intracellular calcium

(Figure 5.10). Dopants may act as competitive agonists with Ca2+ in many of these pathways

and disrupt normal function due to size or and charge differences causing altered, but beneficial

expression of important targets. Although Ca2+ release from the doped samples were elevated

initially, the average weekly Ca2+ release from the samples up to 28 days were not statistically

different. Along with changing media in the cell cultures every 2-3 days, this gives fairly strong

evidence that the osteoblasts are responding to dopants rather than just elevated calcium levels.

While not part of this particular study, strontium is a good example because it is fairly well

researched due to its use as a pharmaceutical in the form of strontium ranelate. Research

performed on strontium have all indicated a relationship with the CaSR and the OPG/RANKL

ratio, but have also noted that ERK1/2 is upregulated by strontium in osteoblasts cultured from

CaSR null mice [233,288,289].

While these collective results seem to fit well into our previous hypothesis, much work is

still needed in this area to gain a better understanding of the importance and role of trace

elements in bone biology.

121

Figure 5.10. Activation of the Wnt/Fz ligand-receptor leads to the production of the second messengers IP3 and DAG from membrane-bound PIP2 via the action of membrane-bound enzyme PLC. IP3 causes release of Ca2 from the ER and extracellular Ca2+ influx through transmembrane Ca2+ channels; CaN and CamKII are activated which in turn activate NFAT and NFkB. DAG is also activated by increased inctracellular Ca2+, which activates PKC. PKC activates NFkB and CREB. NFAT, NFkB, and CREB translocate to the nucleus and transcribe downstream regulatory genes such as Runx2. Elevated Runx2 activity increases OPG production while simultaneously decreasing RANKL production. BMP2, upregulated by extracellular Ca2+, can affect this pathway by binding to BMP2 receptors which activate PI3k. PI3k activates the gamma subunit of PLC increasing its conversion of IP3 and DAG.

122

5.5 Conclusions

CaP materials are some of the most widely used and studied materials for use as

synthetics in orthopedic medicine. In this study, by incorporating the dopants SiO2 and ZnO into highly porous calcium phosphate scaffolds, density was increased and soluble α-TCP phase formation was mitigated at high sintering temperatures. Dopants also did not affect the dissolution properties of the scaffolds. SEM and RT-PCR demonstrated that dopants caused cells to differentiate more quickly in samples containing either dopant as well as maintain an increased OPG:RANKL ratio, indicating an increased affinity for osteogenesis.

123

6. CHAPTER SIX

Effects of SiO2 and ZnO binary dopant system on osteoblasts in porous TCP scaffolds

cultured in a flow perfusion bioreactor

6.1 Introduction

In 2005, over 21% of US adults were reported as having osteoarthritis and nearly 27 million people as having clinical arthritis [290,291]. With such a large proportion of the population suffering from orthopedic disorders, there has been an increasing interest in new treatment methods. Bone grafts are the second most transplanted tissue in the world, with more than 1.5 million transplantations occurring annually in the US alone [292]. Until recently, the standard has been the use of autografts, or bone harvested from a second site on the host. While autografts have demonstrated excellent osteoconduction and osteoinduction, there is limited availability and potential for donor site morbidity [59,293]. Allografts, or tissues harvested from a cadaver source, have also been widely used, but can elicit an immunogenic response and has the potential to transmit disease [294]. Synthetic alternatives that have gained much clinical traction over the past few years come from the family of calcium phosphate (CaP) materials.

Most widely used are hydroxyapatite (HA) and β-tricalcium phosphate (β-TCP). These materials have a compositional similarity to natural bone that gives them excellent bioactivity [295].

β-TCP is considered biodegradable and has a functional lifetime usually around 1 year and is completely replaced by natural bone after three years, whereas HA may remain in the

124 body for over 10 years [249–251]. β-TCP is generally considered to be an osteoconductive material, or a material that easily allows bone to grow on and into. The goal of much recent research has been to increase the biological response of CaP materials so that they may be osteoconductive, or actively promote osteogenesis. There have been many methods investigated to support osteoinductivity in β-TCP, ranging from the addition of growth hormones and pharmaceuticals to the incorporation of other biologics such as collagen [62,256]. While positive results have been achieved using these methods, drug eluting and biologic containing CaPs have long and costly development times before clinically acceptable to use. Our group, as well as others, have investigated the use of trace element nutrients into CaP materials and have catalogued numerous benefits ranging from stable strength degradation, increased compressive strength, increased biological response both in vitro and in vivo[97,134,217,260,263]. Silicon and zinc have both been widely studied in bone biology. Silicon, while not an essential trace element, is believed to play a positive role in the biomineralization process and has been found at active calcification in bone healing [109]. It has also been found that deficiencies in silicon can result in a number of disease states such as deformed bone development, stunted growth and low collagen levels [110,111]. Zinc is an important essential element that is released from bone during osteoclastic resorption [117]. The released ion plays a vital role in the mineralization process, as it is a key stabilizer in alkaline phosphatase (ALP) [230], an enzyme responsible for the catalysis of the hydrolysis of phosphomonoesters into inorganic phosphates [52]. Zinc is also thought to play an important function in regulating osteoclastic resorption process and stimulating the osteoblastic bone building process [118,119]. In a previous study, the combination of silicon and zinc increased early stage osteogenesis and angiogenesis in vivo[262].

125

Another key factor in stimulating osteogenesis in β-TCP materials is to incorporate a high

degree of micro and macroporosity. Having a porous structure with sufficient interconnectivity

results in early bone-implant fixation and increased bone growth rates [78,179,296]. Further still,

if these porous scaffolds can be infiltrated with the host’s bone cells prior to implantation, the

osteogenic effects can be significantly increased [297,298]. The challenge is in the infiltration

and expansion process, however. Standard culture methods do not offer optimal mass exchange

for cells towards the center of the porous constructs [144]. Many studies have demonstrated that

the use of a bioreactor can stimulate three dimensional seeding efficiency, proliferation and

differentiation of bone cells in vitro [299–301]. While several types of bioreactors have been developed and tested, it is generally believed that the flow perfusion type create the most ideal

microenvironments for bone tissue engineering applications [143].

In this study, highly porous TCP constructs were doped with SiO2 and ZnO as a binary

dopant system. Scaffolds were seeded with human pre-osteoblast cells and cultured in a static

standard environment as well as in a dynamic flow perfusion bioreactor in order to investigate

the effects on cell growth, differentiation and morphology.

6.2 Methods

6.2.1 Sample Preparation

β-tricalcium phosphate powder (β-TCP) was synthesized in house via a solid state

method. Briefly, 2 moles of dicalcium phosphate (CaHPO4) were mixed with 1 mole of calcium

126 carbonate (CaCO3). Powders were mixed for 2 hours in a 5:1 milling media to powder weight ratio, then heated to 1050 ˚C for 24 hours let cool to room temperature.

High purity silicon dioxide (SiO2) (99%+ purity) and zinc oxide (ZnO) (99.9%+ purity) were purchased from Fisher Scientific (Fair Lawn, NJ). Doped powders were prepared by mixing 20 g of β-TCP powder and appropriate amounts of dopants (0.5 wt% SiO2 and 0.25 wt%

ZnO) in 250 mL polypropylene Nalgene bottles containing 30 mL of anhydrous ethanol and 100 g zirconia milling media with 5mm diameter. Dopant concentrations were chosen based on previous optimization research [97,105,106,125] The mixtures were then milled for 6 h at 70 rpm to minimize the formation of agglomerates and increase homogeneity. After milling, powder was dried in an oven at 70 °C for 72 h. Samples were prepared using an oil emulsion method previously described [82]. Briefly, 20g of β-TCP powder (doped or pure) were mixed with 20g paraffin oil purchased from Sigma Aldrich (St. Louis, MO) and 13.4 mL of an emulsifier solution of 0.14g/L Kolliphor EL purchased from Sigma Aldrich (St. Louis, MO) in 0.2M

Na2HPO4. The mixture was stirred at 2000 rpm for 45 seconds to yield a stable emulsion slurry.

6.2.2 Microstructure and Phase Analysis

The surface morphologies of the sintered scaffolds were observed under a field emission scanning electron microscope (FESEM) (FEI Inc., OR, USA) before and after 28 days in PBS.

Phase analysis of sintered TCP samples with and without dopants was carried out by X-ray

127

diffraction (XRD) using a Siemens D-500 X-ray powder Diffractometer (Siemens AG,

Karlsruhe, Germany) with CuKα radiation and a Ni-filter. Each run was performed with 2θ values between 10° and 60° at a step size of 0.1° and a count time of 1s per step. Density was measured using Archimedes method. Samples were weighed initially dry and then submerged in

boiling water for 3 minutes to remove any excess air that may be trapped in the porous structure.

The samples were then transferred from the boiling water to room temperature water, where the

weight was recorded again (n=3). Bulk densities were also measured for the same samples for

comparison.

6.2.3 Ca2+ Ion Release

Ca2+ content of samples incubated in PBS was analyzed over 28 days using a Shimadzu

AA-6800 atomic absorption spectrophotometer (Shimadzu, Kyoto, Japan). Standard solutions

were freshly prepared in ionization buffer to obtain a final concentration of 0-1000 μg/ml.

Calcium ionization buffer standards were purchased from High-Purity Standards (Charleston,

SC, USA).

6.2.4 Bioreactor Design and Flow Analysis

The bioreactor was designed and analyzed with Solidworks and Solidworks Flow Analysis

softwares (Waltham, MA).

128

6.2.5 Static Culture

All samples were sterilized by autoclaving at 121 °C for 20 min. In this study, established

human preosteoblast cell line hFOB 1.19 (ATCC, Manassas, VA) were used. Cells were seeded

onto the samples in 24-well plates at a density of 2 x 105 cells/sample. The base medium for this cell line was a 1:1 mixture of Ham's F12 Medium and Dulbecco's Modified Eagle's Medium

(DMEM/F12, Sigma, St. Louis, MO), with 2.5 mM L-glutamine (without phenol red). The medium was supplemented with 10% fetal bovine serum (HyClone, Logan, UT) and 0.3 mg/ml

G418 (Sigma, St. Louis, MO). Cultures were maintained at 37 °C under an atmosphere of 5%

CO2 as recommended by ATCC for this particular cell line. Medium was changed every 2-3 days for the duration of the experiment.

6.2.6 Bioreactor Culture

The flow perfusion system is a modification of the system described by Bancroft et al.

[140]. The system consists of 12 separate polycarbonate flow chambers (Figure 6.1) each supplied by medium from a common reservoir by a Cole Parmer Masterflex (Vernon Hills, IL)

peristaltic pump at a flow rate of 500mL/min. The cell medium flows through the scaffolds from top to bottom. All tubing, chambers, sample holders and samples were autoclaved prior to culture at 121 °C for 20 min. Cells were seeded onto samples at a density of 2 x 105 cells/sample.

Cultures were also maintained at 37 °C under an atmosphere of 5% CO2 and media was changed

once per week.

129

Figure 6.1. CAD drawing of single bioreactor chamber containing sample holder and sample.

6.2.7 Cellular Proliferation & Morphology

Samples for testing were removed from culture after 14 days of incubation. All samples

for SEM observation were fixed with 2% paraformaldehyde/2% glutaraldehyde in 0.1M

cacodylate buffer overnight at 4 °C. Post-fixation was performed with 2% osmium tetroxide

(OsO4) for 2 h at room temperature. The fixed samples were then dehydrated in an ethanol series

(30%, 50%, 70%, 95% and 100% three times), followed by a hexamethyldisilane (HMDS)

drying procedure. After gold coating, the samples were observed under field emission scanning

electron microscope (FESEM) (FEI 200F, FEI Inc., OR, USA) for cell morphologies.

130

Cellular proliferation was monitored with Alamar Blue purchased from Life

Technologies (Carlsbad, CA). The manufacturer’s recommended protocol was used for the

experiments. Briefly, at 3, 7 and 14 days of culture, sample media was changed and 100 µL of

Alamar blue reagent was added to each culture well. Samples were then incubated for 8 hours

and fluorescence detected with a Synergy Hybrid multi-mode microplate reader (Biotek,

Winooski, VT) with excitation wavelength of 540 nm and emission wavelength of 590 nm.

Samples were then processed for downstream applications.

6.2.8 Alkaline Phosphatase Activity

Samples were fixed in 3.7% paraformaldehyde/ phosphate buffered solution with a pH of

7.4 at room temperature for 10 min. Samples were then washed in PBS 3 times (5 min each) and cells were permeabilized with 0.1% Triton X-100 (in PBS) for 4 min at room temperature. Next, samples were rinsed in TBST 3 times (5 minutes each) and incubated in TBST-BSA (Tris- buffered saline with 1% bovine serum albumin, 250 mM NaCl, pH 8.3) blocking solution for 1h at room temperature. Primary antibody against ALP (Abcam, Cambridge, MA) was added at a

1:100 dilution and incubated at room temperature for 2 h and kept at 4 °C overnight. Samples were then washed with TBST 3 times (10 min each). The secondary antibody, goat anti-mouse

(GAM) Oregon green 488 (Molecular Probes, Eugene, OR), was diluted 1:100 in TBST and was used to incubate the cells for 1h. After rinsing three times for 10 minutes each with TBST the samples placed in 24 well plates with Vectashield mounting medium (Vector Labs, Burlingame,

CA) with 4',6-diamidino-2-phenylindole (DAPI). Samples were immediately analyzed by use of fluorescent area scans using a 5x5 grid pattern on a Synergy Hybrid multi-mode microplate

131

reader (Biotek, Winooski, VT) with excitation wavelength of 488 nm and emission wavelength

of 540 nm. Data was corrected to account for differences in cellular proliferation/expression.

Afterwards, samples were placed on glass coverslips and confocal micrographs were taken using

a Zeiss 510 laser scanning microscope (LSM 510 META, Carl Zeiss MicroImaging, Inc., NY,

USA).

6.2.9 Statistical analysis

Statistical analysis was performed using a one way ANOVA and p<0.05 was considered

statistically significant.

6.3 Results

6.3.1 Physical Properties

XRD (Figure 6.2) performed on the samples indicated that β-TCP was the primary phase

present after sintering at 1250 °C. α-TCP was present in the pure samples, but was barely

detectable in the doped samples. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP) were used to match significant peaks. The doped samples also demonstrated a peak shift of 1.832 2θ indicating substitution of Si4+ and Zn2+ into the lattice structure.

132

Figure 6.2. XRD patterns of doped and pure scaffolds sintered at 1250 ˚C. JCPDS # 09-0169 (β-TCP) and 09-0348 (α-TCP)

Bulk density (Figure 6.3) of the samples demonstrated that both compositions were over

55% porous. Pure TCP had a relative bulk density of 45.14 ± 0.03 %, while Si/Zn-TCP had a relative bulk density of 41.37 ± 1.61%. The doped samples had significantly lower bulk density than the pure samples. Apparent density (Figure 6.4), as measured by Archimedes’ method, showed that the pore wall density (estimated from Arhimedes’ method) in the doped samples were significantly higher than that of the pure TCP samples. Si/Zn-TCP had a relative apparent density of 92.33 ± .02% and pure TCP with 66.05 ± 0.12%.

133

Figure 6.3. Bulk density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3).

Figure 6.4. Bulk density of sintered porous samples. Statistical analysis shows that the differences are significant (* P < 0.05, where n = 3).

134

FESEM micrographs taken before and after 28 days (Figure 6.5) in PBS demonstrate the

highly porous nature of the scaffolds. The Pure-TCP scaffolds had visible grain structures, while the Si/Zn-TCP samples did not. All samples did show evidence of liquid phase sintering, characterized by the flowing particle structures. All samples after 28 days in PBS, showed considerable surface degradation, but the microstructure remained largely unchanged. All samples also had significant plate-like apatite formation on the surface of the samples. Si/Zn-

TCP samples showed considerably more apatite formation than the Pure-TCP samples.

Figure 6.5. Surface morphology of sintered scaffolds of pure TCP composition and SiO2/ZnO doped composition before and after 28 days exposure in PBS.

135

Ca2+ ion release (Figure 6.6) in the Si/Zn-TCP composition demonstrated higher

degradation rates than the pure scaffolds. Si/Zn-TCP samples have an average cumulative release

of 1.57mg, while the Pure-TCP samples had an average cumulative release of 1.15 mg. Both

samples had a steady state of release over the 28 day testing period.

Figure 6.6. AAS results for Ca2+ ion accumulation in PBS over 28 days.

136

6.3.2 Bioreactor Design and Flow Analysis

Flow analysis indicated that a flow rate of 500 mL/min would be required to reach

physiologically relevant pressures and fluid shear stresses at the surface of the sample. While

many others have indicated much lower flow rates in their research, the sample holder was

designed with outlet holes to reduce shear stresses and pressures at flow rates for our specific

pump and tubing size of 1/4 inch inner diameter. Figure 6.7 is the simulation results of fluid shear stress experienced across the surface of the sample. Shear stresses across the samples were an average of 24.39 ± 0.6 dyn/cm2 with maxima of 42.6 dyn/cm2 and minima of 0. The highest

pressures were seen in a thin ring near the edge of the samples where flow became turbulent.

Lowest occurred in the direct center of the sample where flow was minimum. Pressures (Figure

6.8) experienced by the cells on the surface of the sample were an average of 101.45 ± 0.32 kPa with the highest in the center of the sample where the bulk of the flow was located. Pressure maxima and minima were 101.55 kPa and 101.3 kPa, respectively. Fluid flow velocity and pattern is given in Figure 6.9.

137

Figure 6.7. Results of simulated shear stress in bioreactor on sample surface. Blue indicates low shear stress, while color gradient up to red indicates increasing shear stresses.

138

Figure 6.8. Results of simulated pressure in bioreactor on sample surface. Blue indicates low pressure, while color gradient up to red indicates increasing pressures.

139

Figure 6.9. Results of simulated fluid flow in bioreactor. Blue indicates low fluid velocity, while color gradient up to red indicates increasing velocities.

6.3.3 Cellular Proliferation and Morphology

Beginning at day 3 cellular proliferation (Figure 6.10) was increased for all pure samples

when compared to doped samples, but remained relatively constant between the standard culture

and the bioreactor cultures. Live cell levels in the doped samples remained steady throughout the

experiment, while pure samples saw an increase at day 7, then a decrease by day 14. Cells in

standard cultures demonstrated single cell morphology (Figure 6.11) on both the pure and doped samples with little sign of mineralization or apatite formation. Samples in the bioreactors

140 demonstrated sheet-like structures indicating differentiation from osteoblasts to lining cells. Both pure and doped samples in the bioreactor had large amounts of plate-like apatite formation.

Figure 6.10. Alamar Blue measurements for cellular proliferation at days 3,7 and 14 for pure and Si/Zn-TCP samples in both culture conditions. *P < 0.05 (n=3)

141

SC BR

Pure TCP

50 µm 50 µm

Si/Zn TCP

50 µm 50 µm

Figure 6.11. Cell morphologies for samples after 14 days of culture. SC indicates standard culture while BR indicates bioreactor. Arrows in SC indicate single cells, while arrows in BR indicate sheet-like cell formations.

6.3.4 Alkaline Phosphatase Activity

Quantitative analysis of ALP (Figure 6.12) demonstrated an increase in activity in the

doped samples up to day 7 and then a decline at day 14. Pure samples remained relatively

constant over the two week study. While no significant pattern was observed between pure and

doped samples, the bioreactor samples remained equal to or greater than the standard cultures

throughout the study. Confocal images (Figure 6.13) reflected the results from the fluorescent

area scans.

142

143

Figure 6.13. Confocal micrographs showing ALP expression in hFOB cells at 3, 7 and 14 days. Green fluorescence indicates active ALP; blue fluorescence indicates DAPI bound to the nuclei of cells.

6.4 Discussion

Previous studies by our group have investigated the effects of SiO2 and ZnO binary doping on the physicochemical properties of 3D printed scaffolds [100]. While the emulsion process used in this study has less control over the geometric placement of the macroporosity of the scaffolds, it can ultimately incorporate a much higher percentage of porosity overall. Similar to the results in 3D printed scaffolds, the XRD (Figure 6.2) results showed a decrease in α-TCP phase formation in the doped samples and a slight shift in 2θ indicating substitution of Si4+ and

Zn2+ into the TCP lattice structure. It has been found that Si4+ can replace P5+ in the TCP lattice with charge compensation created by either O2- vacancies or the presence of excess Ca2+ [269].

Zn2+, because of its similar size and charge with Ca2+, has an affinity for the Ca(5) position in the lattice structure where it is believed that the Ca(5)O6 polyhedron may be strained due to over- bonding and would favor the incorporation of smaller cations such as zinc [270,271]. The

144

combination of Si/Zn dopants stabilizes the crystal structure while sintering at high temperatures

(>1150 ˚C) where α-TCP is known to form.

Apparent density measurements (Figure 6.4) coincide well with XRD results where α-

TCP presence resulted in increased densification of pore walls in the samples containing the

dopants. α-TCP phase formation has been well documented as a stage of rapid grain growth and

overall decrease in the densification [66,170]. Si/Zn-TCP had a relative apparent density of about

92%, while Pure TCP samples achieved only 66% density. Relative bulk density measurements, however, indicated that there were more pores incorporated into the Si/Zn-TCP scaffolds. Pure-

TCP had a total porosity of about 55%, with Si/Zn TCP scaffolds had a total porosity of about

59% as determined from bulk density measurements (Figure 6.3). It is believed that the decreased bulk density is a result of increased densification of the pore walls during sintering resulting in a more porous structure. Microstructural analysis by SEM (Figure 6.5) verified a highly porous structure similar to results observed in the study this method was adapted from

[82]. SEM results also indicate that both sets of samples showed signs of liquid phase sintering, though the doped samples to a much higher extent. Individual grains are visible in many places in the Pure TCP microstructure, but not apparent in the Si/Zn TCP samples. Silicon (111 pm) has a slightly larger atomic radius than that of phosphorous (98 pm), which causes bond elongation and strain in the crystal structure that could result in a decrease in liquidus temperature and increase in liquid phase sintering [272].

Surface degradation on both samples was apparent after 28 days in PBS solution at 37 ˚C, but overall microstructure remained unchanged (Figure 6.5). AAS results for Ca2+ release

145

(Figure 6.6) indicated that the Si/Zn-TCP samples had a significantly higher initial Ca2+ release with slightly elevated kinetics throughout the study. Average weekly release, however, was not significantly different between the two groups. Both groups showed relatively stable release kinetics indicating a continued trend towards controlled degradation. Initial Ca2+ released rates in

the doped samples were likely due to increased surface area resulting from a higher porosity of

the samples. Previous studies have noted that similar Si/Zn co-doped TCP samples resulted in decreased dissolution kinetics, likely as a result of decreased α-TCP phase formation [245]. After

28 days in PBS, the Si/Zn TCP samples had nearly complete coverage with plate-like apatite formation. Pure samples also had a fair amount of apatite formation, but much of the sample remained uncovered. In the early stages of the biomineralization process, silicon can be found at active calcification sites [109] and during the later stages of calcification, silicon plays a direct role in mineralization in aqueous orthosilicic acid (Si(OH)4) form by inducing the precipitation of HA from electrolyte solutions[109]. Other studies have also suggested that the incorporation of Si4+ onto CaP materials creates a more electronegative surface that favors apatite formation

and protein binding [302]. The dissolution-precipitation process is understood to be a diffusion

limited process, where the more apatite formation on the surface and pores of the samples the

slower the dissolution becomes [274]. While a decrease in dissolution rate was not observed in

the timeframe of this study, the apatite formation on the doped samples would indicate a trend

towards stabilization at a faster rate than the pure TCP samples.

The bioreactors utilized in this study were designed for optimum flexibility in sample

type, flow rate, shear stress and pressure requirements. The reactor chamber itself consists of

three concentric platforms that can hold either multiple samples or accommodate various sample

146

sizes and sample holders. Sample holders then can be designed based on equipment constraints

for optimal flow, shear stress and pressure for an individual cell type. This system utilized a

Masterflex pump and pump head that was readily available for use utilizing ¼” tubing. Because

the tubing diameter was relatively large, sample holders had to be designed to allow high flow

rates while maintaining physiological shear stresses on the surface of the scaffold samples.

Modeling results indicated that a flow rate of 500 mL/min resulted in an average shear stress on

the surface of the sample (Figure 6.7) of 24.39 ± 0.6 dyn/cm2 with maxima of 42.6 dyn/cm2 and

minima of 0. The maxima and minima were limited in area. It is estimated that in vivo,

osteoblasts experience shear stresses between 8 and 30 dyn/cm2 [303,304], although several studies have demonstrated that increasing shear stresses (up to 24 dyn/cm2) results in increased

cellular proliferation and differentiation [299,305–307]. Average pressure (Figure 6.8) experience by the cells due to the fluid flow was only slightly above atmospheric pressure at

101.45 ±32.42 kPa. Several studies have noted that increased pressures up to 50 kPa above atmospheric pressure can have positive effects on osteoblast differentiation by promoting extracellular signal-regulated kinases 1 and 2 (ERK 1/2) and collagen type 1 formation

[308,309]. Although additional pressure was not included in this study, the chamber pressure can be controlled by a combination of sample holder design and adding filters to the flow outlet.

Simulation of fluid flow (Figure 6.9) in the chamber verified that flow across sample was mostly laminar, but turbulent flows resulted from outlets in the sample holder. This scenario is ideal because you can get predictable shear stresses across the surface of the sample in the laminar region, while maintaining the superior heat and mass transfer capabilities of turbulent flow.

147

Cellular proliferation as determined by the Alamar Blue assay (Figure 6.10)

demonstrated no difference between the bioreactor and static culture conditions, but pure TCP

had slightly increased cell populations. These results were expected from this study due to the

incubation temperature of the cultures. The hFOB 1.19 pre-osteoblast cells will proliferate rapidly at permissive temperatures of around 33.4 ˚C, but little cell division occurs at restrictive temperatures of 39.5 ˚C. The culture conditions used (37 ˚C) were to induce differentiation characteristics, while maintaining some cell division. As osteoblast cells move through their normal lifecycle, they undergo a proliferation phase, where cell division is rampant, then a maturation phase and a differentiation phase [229]. As cells advance into the maturation phase proliferation slows down until it halts at the differentiation phase. These results suggest that maturation occurred in both culture conditions in as little as 7 days. Both Si4+ and Zn2+ have been

extensively studied and have been shown to stimulate the differentiation of osteoblast cells

[120,217,220,260]. It is likely then, that the osteoblasts cultured on the Si/Zn-TCP samples were

further along the cell lifecycle than the cells on the pure TCP samples, resulting in decreased

proliferation rates. While there was no noticeable proliferation rate differences between the

standard culture and bioreactor, there were apparent morphological differences (Figure 6.11).

Cells cultured in standard conditions all had osteoblastic morphology, single elongated cells on

the surface of the samples. Osteoblasts cultured in the bioreactor, however, exhibited areas of

flattened sheet-like cells phenotypical of osteoblasts differentiated to relatively inert bone lining

cells. Alkaline phosphatase activity (Figure 6.12) at day 3 was similar for both pure and doped

samples as well as static and dynamic culture conditions. At day 7, however the Si/Zn-TCP

samples had significantly more ALP activity in the bioreactor and slightly more activity in the

148 standard culture than the Pure TCP compositions. The Si/Zn-TCP samples cultured in the bioreactor also had more ALP production than measured in the standard culture, though not significantly so. At day 14 the trend was reversed, with pure TCP demonstrating more ALP production in both culture conditions than the Si/Zn-TCP samples as well as a significantly increased ALP production in the bioreactor compared to static conditions. These results suggest that the shear stresses in the bioreactor may increase the ALP production of cells in the maturation phase when compared to static culture conditions. ALP is known to be most expressed during the maturation stage of the osteoblastic lifecycle [229]. The results also indicate a more rapid differentiation of the cells cultured on the Si/Zn-TCP samples, with ALP levels spiking a week previous to the Pure TCP samples. Although these results are not definitive of osteoblastic differentiation, this preliminary study shows an appropriate window of study and culture conditions to investigate further.

6.5 Conclusions

In this study, a flexible bioreactor was designed for use with a wide variety of sample sizes and conditions. Predictive modeling of the system allowed a sample holder and flow rate to be optimized for human osteoblast cells. Highly porous scaffolds created using an emulsion method, demonstrated decreased α-TCP phase formation in samples containing SiO2 and ZnO dopants. Decreased α-TCP resulted in increased pore wall density, but overall increase in bulk density due to pore wall shrinkage leading to pore size expansion. Degradation kinetics were similar for both samples, but with an initial increase in Ca2+ likely due to increased porosity in

149 the doped samples. Cell proliferation between the bioreactor and static cultures remained relatively constant, but the culture was designed to focus on differentiation. ALP results indicated that the Si/Zn-TCP samples went through cellular maturation faster than the Pure TCP samples. Both Pure-TCP and Si/Zn-TCP demonstrated higher ALP levels in the bioreactor cultures at peak production.

150

Summary

1. 3D printing was successfully used to make highly controlled porous architecture in

calcium phosphate scaffolds with SiO2/ZnO dopant combinations.

2. The addition of SiO2/ZnO dopants in 3D printed scaffolds decreased α phase formation at

high sintering temperatures (1250 °C), which led to increased densification and an

average of 2.5 fold increase in compressive strength when compared to pure TCP

scaffolds.

3. SiO2/ZnO dopants in 3D printed scaffolds also lead to increased cellular proliferation

over the course of a preliminary 11 day cell culture study.

4. SiO2/ZnO addition to 3D printed scaffolds implanted into Sprague Dawley rats

demonstrated increased bone formation over the first twelve weeks of implantation as

well as increased angiogenesis at early time points. Collagen and osteocalcin levels were

also increased in samples containing dopants, also confirming increased healing.

5. SiO2/ZnO doped scaffolds showed decreased degradation kinetics after 12 weeks

implantation in a murine model.

6. Incorporating the dopants SiO2, ZnO and SrO into TCP compacts, the important

transcription factor Runx2 was able to be modulated in a beneficial manner as to induce

accelerated bone cell differentiation, with high expression in the early stages of cell

maturation and low expression in the terminal stage.

7. ZnO and SrO dopants in compact TCP samples increased ALP activity when compared

to pure TCP indicating increased affinity for matrix mineralization.

151

8. ZnO and MgO dopants both demonstrated favorable mitogenic properties, but Mg did not

have any significant effect on the differentiation markers.

9. It was hypothesized that dopants can affect Runx2 expression and other differentiation

markers via the intracellular Ca2+ signaling pathway and the Wnt signaling pathway.

10. Incorporating the dopants SiO2 and ZnO into highly porous calcium phosphate scaffolds,

density was increased and soluble α-TCP phase formation was mitigated at high sintering

temperatures.

11. The SiO2 and ZnO dopants in porous TCP scaffolds also did not affect the dissolution

properties of the scaffolds.

12. SEM and RT-PCR demonstrated that the SiO2 and ZnO dopants caused cells to

differentiate more quickly in samples containing either dopant as well as maintain an

increased OPG:RANKL ratio, indicating an increased affinity for osteogenesis. These

results also fit well into the previously proposed intracellular signaling mechanism.

13. A flexible bioreactor was designed for use with a wide variety of sample sizes and

conditions.

14. Predictive modeling of the bioreactor system allowed a sample holder and flow rate to be

optimized for human osteoblast cells.

15. Highly porous scaffolds created using an emulsion method, demonstrated decreased α-

TCP phase formation in samples containing a binary system of SiO2 and ZnO dopants.

Decreased α-TCP resulted in increased pore wall density, but overall increase in bulk

density due to pore wall shrinkage leading to pore size expansion.

152

16. Degradation kinetics were similar for SiO2/ZnO doped samples as the pure TCP samples,

but with an initial increase in Ca2+ likely due to increased porosity in the doped samples.

17. Cell proliferation between the bioreactor and static cultures remained relatively constant,

but the culture was designed to focus on differentiation.

18. ALP results indicated that the Si/Zn-TCP samples went through cellular maturation

faster than the Pure TCP samples. Both Pure-TCP and Si/Zn-TCP demonstrated higher

ALP levels in the bioreactor cultures at peak production.

153

References

[1] Seeley RR, Tate P, Stephens TD. Anatomy & physiology. Dubuque, IA: McGraw-Hill; 2008.

[2] Gray H. Gray’s anatomy: the anatomical basis of clinical practice. [Edinburgh U.a.: Churchill Livingstone Elsevier; 2008.

[3] Marieb EN. Human anatomy & physiology. San Francisco: Pearson Benjamin Cummings; 2007.

[4] Eriksen EF, Axelrod DW, Melsen F. Bone histomorphometry. New York: Raven Press; 1994.

[5] Clarke B. Normal Bone Anatomy and Physiology. Clin J Am Soc Nephrol 2008;3:S131– S139.

[6] Colnot C, Zhang X, Knothe Tate ML. Current insights on the regenerative potential of the periosteum: molecular, cellular, and endogenous engineering approaches. J Orthop Res Off Publ Orthop Res Soc 2012;30:1869–78.

[7] Kollet O, Dar A, Lapidot T. The multiple roles of osteoclasts in host defense: bone remodeling and hematopoietic stem cell mobilization. Annu Rev Immunol 2007;25:51–69.

[8] Studer D, Millan C, Öztürk E, Maniura-Weber K, Zenobi-Wong M. Molecular and biophysical mechanisms regulating hypertrophic differentiation in chondrocytes and mesenchymal stem cells. Eur Cell Mater 2012;24:118–135; discussion 135.

[9] Murshed M, Harmey D, Millán JL, McKee MD, Karsenty G. Unique coexpression in osteoblasts of broadly expressed genes accounts for the spatial restriction of ECM mineralization to bone. Genes Dev 2005;19:1093–104.

[10] Boyce BF, Xing L. Biology of RANK, RANKL, and osteoprotegerin. Arthritis Res Ther 2007;9 Suppl 1:S1.

[11] Dobnig H, Turner RT. Evidence that intermittent treatment with parathyroid hormone increases bone formation in adult rats by activation of bone lining cells. Endocrinology 1995;136:3632–8.

[12] Rubin CT, Lanyon LE. Osteoregulatory nature of mechanical stimuli: Function as a determinant for adaptive remodeling in bone. J Orthop Res 1987;5:300–10.

154

[13] Klein-Nulend J, Bacabac RG, Mullender MG. Mechanobiology of bone tissue. Pathol Biol (Paris) 2005;53:576–80. [14] Klein-Nulend J, Bacabac RG, Bakker AD. Mechanical loading and how it affects bone cells: the role of the osteocyte cytoskeleton in maintaining our skeleton. Eur Cell Mater 2012;24:278–91.

[15] Mullender M, Huiskes R. Osteocytes and bone lining cells: Which are the best candidates for mechano-sensors in cancellous bone? Bone 1997;20:527–32.

[16] Baron R, Kneissel M. WNT signaling in bone homeostasis and disease: from human mutations to treatments. Nat Med 2013;19:179–92.

[17] Takahashi N, Udagawa N, Kobayashi Y, Takami M, Martin TJ, Suda T. Chapter 9 - Osteoclast Generation. In: John P. Bilezikian, Lawrence G. Raisz and T. John MartinA2 - John P. Bilezikian LGR, T. John Martin, editors. Princ. Bone Biol. Third Ed., San Diego: Academic Press; 2008, p. 175–92.

[18] Martin RB. Fatigue microdamage as an essential element of bone mechanics and biology. Calcif Tissue Int 2003;73:101–7.

[19] Väänänen K. Mechanism of osteoclast mediated bone resorption--rationale for the design of new therapeutics. Adv Drug Deliv Rev 2005;57:959–71.

[20] Martin R. Is all cortical bone remodeling initiated by microdamage? Bone 2002;30:8–13.

[21] Hauge EM, Qvesel D, Eriksen EF, Mosekilde L, Melsen F. Cancellous Bone Remodeling Occurs in Specialized Compartments Lined by Cells Expressing Osteoblastic Markers. J Bone Miner Res 2001;16:1575–82.

[22] Eriksen EF, Eghbali-Fatourechi GZ, Khosla S. Remodeling and vascular spaces in bone. J Bone Miner Res Off J Am Soc Bone Miner Res 2007;22:1–6.

[23] Kular J, Tickner J, Chim SM, Xu J. An overview of the regulation of bone remodelling at the cellular level. Clin Biochem 2012;45:863–73.

[24] Tian XY, Zhang Q, Zhao R, Setterberg RB, Zeng QQ, Ma YF, et al. Continuous infusion of PGE2 is catabolic with a negative bone balance on both cancellous and cortical bone in rats. J Musculoskelet Neuronal Interactions 2007;7:372–81.

[25] Costa AFP, dos Reis LM, Ribeiro MC, Moysés RMA, Jorgetti V. Effects of calcitriol on parathyroid function and on bone remodelling in secondary hyperparathyroidism. Nephrol Dial Transplant Off Publ Eur Dial Transpl Assoc - Eur Ren Assoc 2003;18:743–9.

155

[26] Chae HJ, Park RK, Chung HT, Kang JS, Kim MS, Choi DY, et al. Nitric oxide is a regulator of bone remodelling. J Pharm Pharmacol 1997;49:897–902.

[27] Hadjidakis DJ, Androulakis II. Bone Remodeling. Ann N Y Acad Sci 2006;1092:385–96. [28] Fan X, Biskobing DM, Fan D, Hofstetter W, Rubin J. Macrophage colony stimulating factor down-regulates MCSF-receptor expression and entry of progenitors into the osteoclast lineage. J Bone Miner Res Off J Am Soc Bone Miner Res 1997;12:1387–95.

[29] Yu B, Zhao X, Yang C, Crane J, Xian L, Lu W, et al. Parathyroid hormone induces differentiation of mesenchymal stromal/stem cells by enhancing bone morphogenetic protein signaling. J Bone Miner Res Off J Am Soc Bone Miner Res 2012;27:2001–14.

[30] Lee MS, Kim HS, Yeon J-T, Choi S-W, Chun CH, Kwak HB, et al. GM-CSF regulates fusion of mononuclear osteoclasts into bone-resorbing osteoclasts by activating the Ras/ERK pathway. J Immunol Baltim Md 1950 2009;183:3390–9.

[31] Boyce BF, Rosenberg E, de Papp AE, Duong LT. The osteoclast, bone remodelling and treatment of metabolic bone disease. Eur J Clin Invest 2012;42:1332–41.

[32] Kameda T, Mano H, Yuasa T, Mori Y, Miyazawa K, Shiokawa M, et al. Estrogen Inhibits Bone Resorption by Directly Inducing Apoptosis of the Bone-resorbing Osteoclasts. J Exp Med 1997;186:489–95.

[33] Kuo Y-J, Tsuang F-Y, Sun J-S, Lin C-H, Chen C-H, Li J-Y, et al. Calcitonin Inhibits SDCP-Induced Osteoclast Apoptosis and Increases Its Efficacy in a Rat Model of Osteoporosis. PLoS One 2012;7.

[34] Cheng J, Liu J, Shi Z, Jules J, Xu D, Luo S, et al. Molecular Mechanisms of the Biphasic Effects of Interferon-? on Osteoclastogenesis. J Interferon Cytokine Res 2012;32:34–45.

[35] Tang Y, Wu X, Lei W, Pang L, Wan C, Shi Z, et al. TGF-?1-induced Migration of Bone Mesenchymal Stem Cells Couples Bone Resorption and Formation. Nat Med 2009;15:757–65.

[36] Matsumoto T, Yamada A, Aizawa R, Suzuki D, Tsukasaki M, Suzuki W, et al. BMP-2 Induced Expression of Alx3 That Is a Positive Regulator of Osteoblast Differentiation. PloS One 2013;8:e68774.

[37] Li J, Jin D, Fu S, Mei G, Zhou J, Lei L, et al. Insulin-like growth factor binding protein-3 modulates osteoblast differentiation via interaction with vitamin D receptor. Biochem Biophys Res Commun 2013;436:632–7.

156

[38] Yoshida S, Iwasaki R, Kawana H, Miyauchi Y, Hoshi H, Miyamoto H, et al. PDGFBB promotes PDGFRα-positive cell migration into artificial bone in vivo. Biochem Biophys Res Commun 2012;421:785–9.

[39] De A. Wnt/Ca2+ Signaling Pathway: A Brief Overview. Acta Biochim Biophys Sin 2011;43:745–56.

[40] Lin GL, Hankenson KD. Integration of BMP, Wnt, and notch signaling pathways in osteoblast differentiation. J Cell Biochem 2011;112:3491–501.

[41] Canalis E, Adams DJ, Boskey A, Parker K, Kranz L, Zanotti S. Notch Signaling in Osteocytes Differentially Regulates Cancellous and Cortical Bone Remodeling. J Biol Chem 2013.

[42] Nagayama T, Okuhara S, Ota MS, Tachikawa N, Kasugai S, Iseki S. FGF18 accelerates osteoblast differentiation by upregulating Bmp2 expression. Congenit Anomalies 2013;53:83–8.

[43] Komori T. Regulation of bone development and extracellular matrix protein genes by RUNX2. Cell Tissue Res 2010;339:189–95.

[44] Brown EM. Role of the calcium-sensing receptor in extracellular calcium homeostasis. Best Pract Res Clin Endocrinol Metab 2013;27:333–43.

[45] Mathews S, Bhonde R, Gupta PK, Totey S. Extracellular matrix protein mediated regulation of the osteoblast differentiation of bone marrow derived human mesenchymal stem cells. Differ Res Biol Divers 2012;84:185–92.

[46] Beck GR, Sullivan EC, Moran E, Zerler B. Relationship between alkaline phosphatase levels, osteopontin expression, and mineralization in differentiating MC3T3-E1 osteoblasts. J Cell Biochem 1998;68:269–80.

[47] Perrien DS, Brown EC, Aronson J, Skinner RA, Montague DC, Badger TM, et al. Immunohistochemical study of osteopontin expression during distraction osteogenesis in the rat. J Histochem Cytochem Off J Histochem Soc 2002;50:567–74.

[48] Kumar A, Salimath BP, Stark GB, Finkenzeller G. Platelet-derived growth factor receptor signaling is not involved in osteogenic differentiation of human mesenchymal stem cells. Tissue Eng Part A 2010;16:983–93.

[49] Park J-B. Effects of fibroblast growth factor 2 on osteoblastic proliferation and differentiation by regulating bone morphogenetic protein receptor expression. J Craniofac Surg 2011;22:1880–2.

157

[50] Adler C-P. Bone Diseases: Macroscopic, Histological, and Radiological Diagnosis of Structural Changes in the Skeleton. Springer; 2000.

[51] Rowe PSN. Regulation of bone-renal mineral and energy metabolism: the PHEX, FGF23, DMP1, MEPE ASARM pathway. Crit Rev Eukaryot Gene Expr 2012;22:61–86.

[52] Fielding GA, Roy M, Bose S, Bandyopadhyay A. Antibacterial and biological characteristics of plasma sprayed silver and strontium doped hydroxyapatite coatings. Acta Biomater 2012;In Press. [53] Kirsch T. Biomineralization--an active or passive process? Connect Tissue Res 2012;53:438–45.

[54] Gorski JP. Biomineralization of bone: a fresh view of the roles of non-collagenous proteins. Front Biosci Landmark Ed 2011;16:2598–621.

[55] Eriksen EF. Cellular mechanisms of bone remodeling. Rev Endocr Metab Disord 2010;11:219–27.

[56] White KP, Harth M. The occurrence and impact of generalized pain. Baillières Best Pr Res Clin Rheumatol 1999;13:379–89.

[57] Woolf AD, ?kesson K. Understanding the burden of musculoskeletal conditions. BMJ 2001;322:1079–80.

[58] March LM, Bachmeier CJM. 10 Economics of osteoarthritis: a global perspective. Baillières Clin Rheumatol 1997;11:817–34.

[59] Pollock R, Alcelik I, Bhatia C, Chuter G, Lingutla K, Budithi C, et al. Donor site morbidity following iliac crest bone harvesting for cervical fusion: a comparison between minimally invasive and open techniques. Eur Spine J 2008;17:845–52.

[60] Sammarco VJ, Chang L. Modern issues in bone graft substitutes and advances in bone tissue technology. Foot Ankle Clin 2002;7:19–41.

[61] Sáenz A, Rivera E, Brostow W, Castano VM. Ceramic biomaterials: an introductory overview. J Mater Educ 1999;21:267–76.

[62] 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.

[63] Bose S, Tarafder S. Calcium Phosphate Ceramic Systems in Growth Factor and Drug Delivery for Bone Tissue Engineering-A Review. Acta Biomater 2011.

158

[64] Yamada M, Shiota M, Yamashita Y, Kasugai S. Histological and histomorphometrical comparative study of the degradation and osteoconductive characteristics of α- and β- tricalcium phosphate in block grafts. J Biomed Mater Res B Appl Biomater 2007;82B:139–48.

[65] Suchanek W, Yashima M, Kakihana M, Yoshimura M. Hydroxyapatite ceramics with selected sintering additives. Biomaterials 1997;18:923–33.

[66] Kanazawa T, Umegaki T, Yamashita K, Monma H, Hiramatsu T. Effects of additives on sintering and some properties of calcium phosphates with various Ca/P ratios. J Mater Sci 1991;26:417–22. [67] Bignon A, Chouteau J, Chevalier J, Fantozzi G, Carret J-P, Chavassieux P, et al. Effect of micro- and macroporosity of bone substitutes on their mechanical properties and cellular response. J Mater Sci Mater Med 2003;14:1089–97.

[68] Blokhuis TJ, Termaat MF, den Boer FC, Patka P, Bakker FC, Haarman HJ. Properties of calcium phosphate ceramics in relation to their in vivo behavior. J Trauma 2000;48:179– 86.

[69] Yoshikawa H, Myoui A. Bone tissue engineering with porous hydroxyapatite ceramics. J Artif Organs Off J Jpn Soc Artif Organs 2005;8:131–6.

[70] Giannoudis PV, Dinopoulos H, Tsiridis E. Bone substitutes: An update. Injury 2005;36:S20–S27.

[71] Chan O, Coathup MJ, Nesbitt A, Ho C-Y, Hing KA, Buckland T, et al. The effects of microporosity on osteoinduction of calcium phosphate bone graft substitute biomaterials. Acta Biomater 2012;8:2788–94.

[72] Dellinger JG, Eurell JAC, Jamison RD. Bone response to 3D periodic hydroxyapatite scaffolds with and without tailored microporosity to deliver bone morphogenetic protein 2. J Biomed Mater Res A 2006;76A:366–76.

[73] Hing KA. Bioceramic Bone Graft Substitutes: Influence of Porosity and Chemistry. Int J Appl Ceram Technol 2005;2:184–99.

[74] LeGeros RZ. Calcium Phosphate-Based Osteoinductive Materials. Chem Rev 2008;108:4742–53.

[75] Kuboki Y, Jin Q, Kikuchi M, Mamood J, Takita H. Geometry of artificial ECM: sizes of pores controlling phenotype expression in BMP-induced osteogenesis and chondrogenesis. Connect Tissue Res 2002;43:529–34.

159

[76] Tsuruga E, Takita H, Itoh H, Wakisaka Y, Kuboki Y. Pore size of porous hydroxyapatite as the cell-substratum controls BMP-induced osteogenesis. J Biochem (Tokyo) 1997;121:317–24.

[77] Karageorgiou V, Kaplan D. Porosity of 3D biomaterial scaffolds and osteogenesis. Biomaterials 2005;26:5474–91.

[78] Jones AC, Arns CH, Hutmacher DW, Milthorpe BK, Sheppard AP, Knackstedt MA. The correlation of pore morphology, interconnectivity and physical properties of 3D ceramic scaffolds with bone ingrowth. Biomaterials 2009;30:1440–51.

[79] Hui P, Leung P, Sher A. Fluid conductance of cancellous bone graft as a predictor for graft-host interface healing. J Biomech 1996;29:123–32. [80] Lopez-Heredia MA, Sariibrahimoglu K, Yang W, Bohner M, Yamashita D, Kunstar A, et al. Influence of the pore generator on the evolution of the mechanical properties and the porosity and interconnectivity of a calcium phosphate cement. Acta Biomater 2012;8:404– 14.

[81] Le Ray A-M, Gautier H, Bouler J-M, Weiss P, Merle C. A new technological procedure using sucrose as porogen compound to manufacture porous biphasic calcium phosphate ceramics of appropriate micro- and macrostructure. Ceram Int 2010;36:93–101.

[82] Bohner M, van Lenthe GH, Grünenfelder S, Hirsiger W, Evison R, Müller R. Synthesis and characterization of porous beta-tricalcium phosphate blocks. Biomaterials 2005;26:6099–105. [83] Fused deposition modeling. Wikipedia Free Encycl 2013.

[84] Chen Z, Li D, Lu B, Tang Y, Sun M, Xu S. Fabrication of osteo-structure analogous scaffolds via fused deposition modeling. Scr Mater 2005;52:157–61.

[85] Tellis BC, Szivek JA, Bliss CL, Margolis DS, Vaidyanathan RK, Calvert P. Trabecular scaffolds created using micro CT guided fused deposition modeling. Mater Sci Eng C 2008;28:171–8.

[86] Selective laser sintering. Wikipedia Free Encycl 2013.

[87] Duan B, Wang M, Zhou WY, Cheung WL, Li ZY, Lu WW. Three-dimensional nanocomposite scaffolds fabricated via selective laser sintering for bone tissue engineering. Acta Biomater 2010;6:4495–505.

[88] Tan K., Chua C., Leong K., Cheah C., Cheang P, Abu Bakar M., et al. Scaffold development using selective laser sintering of polyetheretherketone–hydroxyapatite biocomposite blends. Biomaterials 2003;24:3115–23.

160

[89] Klammert U, Gbureck U, Vorndran E, Rödiger J, Meyer-Marcotty P, Kübler AC. 3D powder printed calcium phosphate implants for reconstruction of cranial and maxillofacial defects. J Cranio-Maxillofac Surg 2010;38:565–70.

[90] Carano RAD, Filvaroff EH. Angiogenesis and bone repair. Drug Discov Today 2003;8:980–9.

[91] Gerber HP, Ferrara N. Angiogenesis and bone growth. Trends Cardiovasc Med 2000;10:223–8.

[92] Geiger M. Collagen sponges for bone regeneration with rhBMP-2. Adv Drug Deliv Rev 2003;55:1613–29.

[93] Tamai N, Myoui A, Hirao M, Kaito T, Ochi T, Tanaka J, et al. A new biotechnology for articular cartilage repair: subchondral implantation of a composite of interconnected porous hydroxyapatite, synthetic polymer (PLA-PEG), and bone morphogenetic protein-2 (rhBMP-2). Osteoarthritis Cartilage 2005;13:405–17.

[94] Shields LBE, 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.

[95] Boraiah S, Paul O, Hawkes D, Wickham M, Lorich DG. Complications of recombinant human BMP-2 for treating complex tibial plateau fractures: a preliminary report. Clin Orthop 2009;467:3257–62.

[96] Luca L, Rougemont A-L, Walpoth BH, Gurny R, Jordan O. The effects of carrier nature and pH on rhBMP-2-induced ectopic bone formation. J Controlled Release 2010;147:38– 44.

[97] Bandyopadhyay A, Bernard S, Xue W, Bose S. Calcium Phosphate-Based Resorbable Ceramics: Influence of MgO, ZnO, and SiO2 Dopants. J Am Ceram Soc 2006;89:2675– 88.

[98] Li X, Sogo Y, Ito A, Mutsuzaki H, Ochiai N, Kobayashi T, et al. The optimum zinc content in set calcium phosphate cement for promoting bone formation in vivo. Mater Sci Eng C Mater Biol Appl 2009;29:969–75.

[99] Bohner M. Silicon-substituted calcium phosphates–a critical view. Biomaterials 2009;30:6403–6.

[100] Fielding GA, Bandyopadhyay A, Bose S. Effects of SiO2 and ZnO doping on mechanical and biological properties of 3D printed TCP scaffolds. Dent Mater 2011.

161

[101] Mestres G, Le Van C, Ginebra M-P. Silicon-stabilized α-tricalcium phosphate and its use in a calcium phosphate cement: Characterization and cell response. Acta Biomater 2012;8:1169–79.

[102] Hanai Y, Tokuda H, Yasuda E, Noda T, Ohta T, Takai S, et al. Up-regulation by zinc of FGF-2-induced VEGF release throughenhancing p44/p42 MAP kinase activation in osteoblasts. Life Sci 2006;80:230–4.

[103] Gao T, Aro HT, Ylänen H, Vuorio E. Silica-based bioactive glasses modulate expression of bone morphogenetic protein-2 mRNA in Saos-2 osteoblasts in vitro. Biomaterials 2001;22:1475–83.

[104] Bandyopadhyay A, Petersen J, Fielding G, Banerjee S, Bose S. ZnO, SiO2, and SrO doping in resorbable tricalcium phosphates: Influence on strength degradation, mechanical properties, and in vitro bone-cell material interactions. J Biomed Mater Res B Appl Biomater 2012;100:2203–12.

[105] Banerjee SS, Tarafder S, Davies NM, Bandyopadhyay A, Bose S. Understanding the influence of MgO and SrO binary doping on the mechanical and biological properties of [beta]-TCP ceramics. Acta Biomater 2010;6:4167–74.

[106] Bose S, Tarafder S, Banerjee SS, Davies NM, Bandyopadhyay A. Understanding in vivo response and mechanical property variation in MgO, SrO and SiO2 doped β-TCP. Bone 2011;48:1282–90.

[107] Rewinkel SE, Washington State University, School of Mechanical and Materials Engineering. The effects of SiO , ZnO, and MgO doping on the mechanical and biological properties of beta-tricalcium phosphate bioceramics for bone tissue engineering, in vitro and in vivo analysis. Pullman, Wash.:₂ Washington State University; 2009.

[108] Nielsen FH. Micronutrients in Parenteral Nutrition: Boron, Silicon, and Fluoride. Gastroenterology 2009;137:S55–S60.

[109] Pietak AM, Reid JW, Stott MJ, Sayer M. Silicon substitution in the calcium phosphate bioceramics. Biomaterials 2007;28:4023–32.

[110] Carlisle EM. Biochemical and morphological changes associated with long bone abnormalities in silicon deficiency. J Nutr 1980;110:1046–56.

[111] Jugdaohsingh R, Calomme MR, Robinson K, Nielsen F, Anderson SHC, D’Haese P, et al. Increased longitudinal growth in rats on a silicon-depleted diet. Bone 2008;43:596–606.

[112] Bergmann C, Lindner M, Zhang W, Koczur K, Kirsten A, Telle R, et al. 3D printing of bone substitute implants using calcium phosphate and bioactive glasses. J Eur Ceram Soc 2010;30:2563–7.

162

[113] Hing KA, Revell PA, Smith N, Buckland T. Effect of silicon level on rate, quality and progression of bone healing within silicate-substituted porous hydroxyapatite scaffolds. Biomaterials 2006;27:5014–26.

[114] Patel N, Best SM, Bonfield W, Gibson IR, Hing KA, Damien E, et al. A comparative study on the in vivo behavior of hydroxyapatite and silicon substituted hydroxyapatite granules. J Mater Sci Mater Med 2002;13:1199–206.

[115] Li H, Chang J. Bioactive Silicate Materials Stimulate Angiogenesis in Fibroblast and Endothelial Cell Co-culture System through Paracrine Effect. Acta Biomater 2013.

[116] Wang C, Lin K, Chang J, Sun J. Osteogenesis and angiogenesis induced by porous β- CaSiO3/PDLGA composite scaffold via activation of AMPK/ERK1/2 and PI3K/Akt pathways. Biomaterials 2013;34:64–77.

[117] Yamaguchi M. Role of nutritional zinc in the prevention of osteoporosis. Mol Cell Biochem 2009;338:241–54.

[118] Hadley KB, Newman SM, Hunt JR. Dietary zinc reduces osteoclast resorption activities and increases markers of osteoblast differentiation, matrix maturation, and mineralization in the long bones of growing rats. J Nutr Biochem 2010;21:297–303.

[119] Yamaguchi M, Weitzmann MN. Zinc stimulates osteoblastogenesis and suppresses osteoclastogenesis by antagonizing NF-κB activation. Mol Cell Biochem 2011;355:179– 86.

[120] Ito A, Kawamura H, Otsuka M, Ikeuchi M, Ohgushi H, Ishikawa K, et al. Zinc-releasing calcium phosphate for stimulating bone formation. Mater Sci Eng C 2002;22:21–5.

[121] Rude RK, Gruber HE, Norton HJ, Wei LY, Frausto A, Kilburn J. Dietary magnesium reduction to 25% of nutrient requirement disrupts bone and mineral metabolism in the rat. Bone 2005;37:211–9.

[122] Maier JA., Bernardini D, Rayssiguier Y, Mazur A. High concentrations of magnesium modulate vascular endothelial cell behaviour in vitro. Biochim Biophys Acta BBA - Mol Basis Dis 2004;1689:6–12.

[123] Cooke JP, Losordo DW. Nitric Oxide and Angiogenesis. Circulation 2002;105:2133–5.

[124] Stendig-Lindberg G, Koeller W, Bauer A, Rob P. Experimentally induced prolonged magnesium deficiency causes osteoporosis in the rat. Eur J Intern Med 2004;15:97–107.

[125] Xue W, Dahlquist K, Banerjee A, Bandyopadhyay A, Bose S. Synthesis and characterization of tricalcium phosphate with Zn and Mg based dopants. J Mater Sci Mater Med 2008;19:2669–77.

163

[126] Landi E, Logroscino G, Proietti L, Tampieri A, Sandri M, Sprio S. Biomimetic Mg- substituted hydroxyapatite: from synthesis to in vivo behaviour. J Mater Sci Mater Med 2008;19:239–47.

[127] Coulombe J, Faure H, Robin B, Ruat R. In vitro effects of strontium ranelate on the extracellular calcium-sensing receptor. Biochem Biophys Res Commun 2004;323:1184– 90.

[128] Brown EM. Is the calcium receptor a molecular target for the actions of strontium on bone? Osteoporos Int 2003;14:25–34.

[129] Tat SK, Pelletier JP, Mineau F, Caron J, Martel-Pelletier J. Strontium ranelate inhibits key factors affecting bone remodeling in human osteoarthritic subchondral bone osteoblasts. Bone 2011.

[130] Kostenuik PJ, others. Osteoprotegerin A Physiological and Pharmacological Inhibitor of Bone Resorption. Curr Pharm Des 2001;7:613–35.

[131] Meunier PJ, Roux C, Seeman E, Ortolani S, Badurski JE, Spector TD, et al. The effects of strontium ranelate on the risk of vertebral fracture in women with postmenopausal osteoporosis. N Engl J Med 2004;350:459–68.

[132] Ni G, Chiu K, Lu W, Wang Y, Zhang Y, Hao L, et al. Strontium-containing hydroxyapatite bioactive bone cement in revision hip arthroplasty. Biomaterials 2006;27:4348–55.

[133] Chen YW, Feng T, Shi GQ, Ding YL, Yu XX, Zhang XH, et al. Interaction of endothelial cells with biodegradable strontium-doped calcium polyphosphate for bone tissue engineering. Appl Surf Sci 2008;255:331–5.

[134] Tian M, Chen F, Song W, Song Y, Chen Y, Wan C, et al. In vivo study of porous strontium-doped calcium polyphosphate scaffolds for bone substitute applications. J Mater Sci Mater Med 2009;20:1505–12.

[135] Yeatts AB, Choquette DT, Fisher JP. Bioreactors to influence stem cell fate: augmentation of mesenchymal stem cell signaling pathways via dynamic culture systems. Biochim Biophys Acta 2013;1830:2470–80.

[136] Meinel L, Karageorgiou V, Fajardo R, Snyder B, Shinde-Patil V, Zichner L, et al. Bone tissue engineering using human mesenchymal stem cells: effects of scaffold material and medium flow. Ann Biomed Eng 2004;32:112–22.

[137] Stiehler M, Bünger C, Baatrup A, Lind M, Kassem M, Mygind T. Effect of dynamic 3-D culture on proliferation, distribution, and osteogenic differentiation of human mesenchymal stem cells. J Biomed Mater Res A 2009;89:96–107.

164

[138] Wang T-W, Wu H-C, Wang H-Y, Lin F-H, Sun J-S. Regulation of adult human mesenchymal stem cells into osteogenic and chondrogenic lineages by different bioreactor systems. J Biomed Mater Res A 2009;88:935–46.

[139] Zhang Z-Y, Teoh SH, Chong W-S, Foo T-T, Chng Y-C, Choolani M, et al. A biaxial rotating bioreactor for the culture of fetal mesenchymal stem cells for bone tissue engineering. Biomaterials 2009;30:2694–704.

[140] Bancroft GN, Sikavitsas VI, Mikos AG. Design of a flow perfusion bioreactor system for bone tissue-engineering applications. Tissue Eng 2003;9:549–54.

[141] Sikavitsas VI, Bancroft GN, Mikos AG. Formation of three-dimensional cell/polymer constructs for bone tissue engineering in a spinner flask and a rotating wall vessel bioreactor. J Biomed Mater Res 2002;62:136–48.

[142] Yeatts AB, Fisher JP. Bone tissue engineering bioreactors: dynamic culture and the influence of shear stress. Bone 2011;48:171–81.

[143] Zhang Z-Y, Teoh SH, Teo EY, Khoon Chong MS, Shin CW, Tien FT, et al. A comparison of bioreactors for culture of fetal mesenchymal stem cells for bone tissue engineering. Biomaterials 2010;31:8684–95.

[144] Godara P, McFarland CD, Nordon RE. Design of bioreactors for mesenchymal stem cell tissue engineering. J Chem Technol Biotechnol 2008;83:408–20.

[145] Botchwey EA, Pollack SR, Levine EM, Johnston ED, Laurencin CT. Quantitative analysis of three-dimensional fluid flow in rotating bioreactors for tissue engineering. J Biomed Mater Res A 2004;69:205–15.

[146] Khan Y, Yaszemski MJ, Mikos AG, Laurencin CT. Tissue engineering of bone: material and matrix considerations. J Bone Joint Surg Am 2008;90 Suppl 1:36–42.

[147] Darsell J, Bose S, Hosick HL, Bandyopadhyay A. From CT Scan to Ceramic Bone Graft. J Am Ceram Soc 2003;86:1076–80.

[148] Neamat A, Gawish A, Gamal-Eldeen AM. $\beta$-Tricalcium phosphate promotes cell proliferation, osteogenesis and bone regeneration in intrabony defects in dogs. Arch Oral Biol 2009;54:1083–90.

[149] Enderle R, Götz-Neunhoeffer F, Göbbels M, Müller FA, Greil P. Influence of magnesium doping on the phase transformation temperature of beta-TCP ceramics examined by Rietveld refinement. Biomaterials 2005;26:3379–84.

[150] Bandyopadhyay A, Bose S. Orthopedic Devices, Materials and Design for. Encycl. Med. Devices Instrum., John Wiley & Sons, Inc.; 2006.

165

[151] Kondo N, Ogose A, Tokunaga K, Ito T, Arai K, Kudo N, et al. Bone formation and resorption of highly purified beta-tricalcium phosphate in the rat femoral condyle. Biomaterials 2005;26:5600–8.

[152] Hulbert SF, Young FA, Mathews RS, Klawitter JJ, Talbert CD, Stelling FH. Potential of ceramic materials as permanently implantable skeletal prostheses. J Biomed Mater Res 1970;4:433–56.

[153] Kuboki Y, Jin Q, Takita H. Geometry of carriers controlling phenotypic expression in BMP-induced osteogenesis and chondrogenesis. J Bone Jt Surg Am 2001;83:S105–S115.

[154] Bose S, Suguira S, Bandyopadhyay A. Processing of controlled porosity ceramic structures via fused deposition. Scr Mater 1999;41:1009–14.

[155] Hollister SJ. Porous scaffold design for tissue engineering. Nat Mater 2005;4:518–24.

[156] Franco J, Launey AP, Tomsia ES. Direct write assembly of calcium phosphate scaffolds using a water-based hydrogel. Acta 2010;6:218–28.

[157] Vorndran E, Klarner M, Klammert U, Grover LM, Patel S, Barralet JE, et al. 3D Powder Printing of β-Tricalcium Phosphate Ceramics Using Different Strategies. Adv Eng Mater 2008;10:B67–B71.

[158] Khalyfa A, Vogt S, Weisser J, Grimm G, Rechtenbach A, Meyer W, et al. Development of a new calcium phosphate powder-binder system for the 3D printing of patient specific implants. J Mater Sci Mater Med 2007;18:909–16.

[159] Ahmad Z, Huang J, Thian ES, Edirisinghe MJ, Jayasinghe SN, Best SM, et al. Freeform Fabrication of Nano-Biomaterials Using 3D Electrohydrodynamic Print-Patterning. J Biomed Nanotechnol 2008;4:185–95.

[160] Habibovic P, Sees TM, van den Doel MA, van Blitterswijk CA, de Groot K. Osteoinduction by biomaterials—Physicochemical and structural influences. J Biomed Mater Res A 2006;77A:747 – 762.

[161] Bose S, Darsell J, Kintner M, Hosick H, Bandyopadhyay A. Pore size and pore volume effects on alumina and TCP ceramic scaffolds. Mater Sci Eng C 2003;23:479–86.

[162] Kalita SJ, Bose S, Hosick HL, Bandyopadhyay A. CaO--P2O5--Na2O-based sintering additives for hydroxyapatite (HAp) ceramics. Biomaterials 2004;25:2331–9.

[163] Hattiangadi A, Bandyopadhyay A. Strength Degradation of Nonrandom Porous Ceramic Structures under Uniaxial Compressive Loading. J Am Ceram Soc 2000;83:2730–6.

[164] Seeley Z, Bandyopadhyay A, Bose S. Tricalcium phosphate based resorbable ceramics: Influence of NaF and CaO addition. Mater Sci Eng C 2008;28:11–7.

166

[165] Seeley Z, Bandyopadhyay A, Bose S. Influence of TiO2 and Ag2O addition on tricalcium phosphate ceramics. J Biomed Mater Res A 2007;82A:113–21.

[166] Eisinger, J C D. Effects of silicon, fluoride, etidronate and magnesium on bone mineral density: a retrospective study. Magnes Res 1993;6:247–9.

[167] Murray EJ, Messer HH. Turnover of bone zinc during normal and accelerated bone loss in rats. J Nutr 1981;111:1641–7.

[168] Reginster J, Saltman O, Franchimont P. Trace elements and postmenopausal osteoporosis: a preliminary study of decreased serum manganese. J Nutr 1981;111:1641–7.

[169] Bandyopadhyay A, Withey EA, Moore J, Bose S. Influence of ZnO doping in calcium phosphate ceramics. Mater Sci Eng C 2007;27:14–7.

[170] Ryu H-S, Youn H-J, Hong KS, Chang B-S, Lee C-K, Chung S-S. An improvement in sintering property of beta-tricalcium phosphate by addition of calcium pyrophosphate. Biomaterials 2002;23:909–14.

[171] Hashizume M, Yamaguchi M. Effect of beta-alanyl-L-histidinato zinc on differentiation of osteoblastic MC3T3-E1 cells: increases in alkaline phosphatase activity and protein concentration. Mol Cell Biochem 1994;131:19–24.

[172] JUGDAOHSINGH R. SILICON AND BONE HEALTH. J Nutr Health Aging 2007;11:99–110.

[173] Thian ES, Huang J, Best SM, Barber ZH, Brooks RA, Rushton N, et al. The response of osteoblasts to nanocrystalline silicon-substituted hydroxyapatite thin films. Biomaterials 2006;27:2692–8.

[174] Brown PW. Calcium Phosphates in Biomedical Engineering. Encycl. Mater. Sci. Technol. Second Ed., Oxford: Elsevier; 2001, p. 893–7.

[175] Davies JE. Mechanisms of endosseous integration. Int J Prosthodont 1998;11:391–401.

[176] Hollister SJ. Porous scaffold design for tissue engineering. Nat Mater 2005;4:518–24.

[177] Albrektsson T, Johansson C. Osteoinduction, osteoconduction and osseointegration. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 2001;10 Suppl 2:S96–101.

[178] Habibovic P, Yuan H, van der Valk CM, Meijer G, van Blitterswijk CA, de Groot K. 3D microenvironment as essential element for osteoinduction by biomaterials. Biomaterials 2005;26:3565–75.

167

[179] Otsuki B, Takemoto M, Fujibayashi S, Neo M, Kokubo T, Nakamura T. Pore throat size and connectivity determine bone and tissue ingrowth into porous implants: Three- dimensional micro-CT based structural analyses of porous bioactive titanium implants. Biomaterials 2006;27:5892–900.

[180] Li B, Liao X, Zheng L, Zhu X, Wang Z, Fan H, et al. Effect of nanostructure on osteoinduction of porous biphasic calcium phosphate ceramics. Acta Biomater 2012;8:3794–804.

[181] Matsumoto T, Okazaki M, Nakahira A, Sasaki J, Egusa H, Sohmura T. Modification of apatite materials for bone tissue engineering and drug delivery carriers. Curr Med Chem 2007;14:2726–33.

[182] Yoshinari M, Oda Y, Inoue T, Matsuzaka K, Shimono M. Bone response to calcium phosphate-coated and bisphosphonate-immobilized titanium implants. Biomaterials 2002;23:2879–85.

[183] Martin I, Muraglia A, Campanile G, Cancedda R, Quarto R. Fibroblast Growth Factor-2 Supports ex Vivo Expansion and Maintenance of Osteogenic Precursors from Human Bone Marrow. Endocrinology 1997;138:4456 –4462.

[184] Lan Levengood SK, Polak SJ, Poellmann MJ, Hoelzle DJ, Maki AJ, Clark SG, et al. The effect of BMP-2 on micro- and macroscale osteointegration of biphasic calcium phosphate scaffolds with multiscale porosity. Acta Biomater 2010;6:3283–91.

[185] Ito A, Kawamura H, Miyakawa S, Layrolle P, Kanzaki N, Treboux G, et al. Resorbability and solubility of zinc‐containing tricalcium phosphate. J Biomed Mater Res 2002;60:224– 31.

[186] Obata A, Tokuda S, Kasuga T. Enhanced in vitro cell activity on silicon-doped vaterite/poly (lactic acid) composites. Acta Biomater 2009;5:57–62.

[187] Mastrogiacomo M, Papadimitropoulos A, Cedola A, Peyrin F, Giannoni P, Pearce S, et al. Engineering of bone using bone marrow stromal cells and a silicon-stabilized tricalcium phosphate bioceramic: evidence for a coupling between bone formation and scaffold resorption. Biomaterials 2007;28:1376–84.

[188] Botelho CM, Brooks RA, Best SM, Lopes MA, Santos JD, Rushton N, et al. Human osteoblast response to silicon-substituted hydroxyapatite. J Biomed Mater Res A 2006;79:723–30.

[189] Kwun IS, Cho YE, Lomeda RAR, Shin HI, Choi JY, Kang YH, et al. Zinc deficiency suppresses matrix mineralization and retards osteogenesis transiently with catch-up possibly through Runx 2 modulation. Bone 2010;46:732–41.

168

[190] Yamaguchi M, Goto M, Uchiyama S, Nakagawa T. Effect of zinc on gene expression in osteoblastic MC3T3-E1 cells: enhancement of Runx2, OPG, and regucalcin mRNA expressions. Mol Cell Biochem 2008;312:157–66.

[191] Egan KP, Brennan TA, Pignolo RJ. Bone histomorphometry using free and commonly available software. Histopathology 2012:no–no.

[192] Wang Q, Chen Q, Zhu J, Huang C, Darvell BW, Chen Z. Effects of pore shape and porosity on the properties of porous LNKN ceramics as bone substitute. Mater Chem Phys 2008;109:488–91.

[193] Van Bael S, Chai YC, Truscello S, Moesen M, Kerckhofs G, Van Oosterwyck H, et al. The effect of pore geometry on the in vitro biological behavior of human periosteum- derived cells seeded on selective laser-melted Ti6Al4V bone scaffolds. Acta Biomater 2012;8:2824–34.

[194] Tarafder S, Balla VK, Davies NM, Bandyopadhyay A, Bose S. Microwave-sintered 3D printed tricalcium phosphate scaffolds for bone tissue engineering. J Tissue Eng Regen Med 2012:n/a–n/a.

[195] Polak SJ, Levengood SKL, Wheeler MB, Maki AJ, Clark SG, Johnson AJW. Analysis of the roles of microporosity and BMP-2 on multiple measures of bone regeneration and healing in calcium phosphate scaffolds. Acta Biomater 2011;7:1760–71.

[196] Hie M, Iitsuka N, Otsuka T, Nakanishi A, Tsukamoto I. Zinc deficiency decreases osteoblasts and osteoclasts associated with the reduced expression of Runx2 and RANK. Bone 2011;49:1152–9.

[197] Matsko NB, Žnidaršič N, Letofsky-Papst I, Dittrich M, Grogger W, Štrus J, et al. Silicon: The key element in early stages of biocalcification. J Struct Biol 2011;174:180–6.

[198] Hing KA, Revell PA, Smith N, Buckland T. Effect of silicon level on rate, quality and progression of bone healing within silicate-substituted porous hydroxyapatite scaffolds. Biomaterials 2006;27:5014–26.

[199] Bjerre L, Bünger CE, Kassem M, Mygind T. Flow perfusion culture of human mesenchymal stem cells on silicate-substituted tricalcium phosphate scaffolds. Biomaterials 2008;29:2616–27.

[200] Gupta G, Kirakodu S, El-Ghannam A. Effects of exogenous phosphorus and silicon on osteoblast differentiation at the interface with bioactive ceramics. J Biomed Mater Res A 2010;95A:882–90.

[201] Wiesmann HP, Meyer U, Plate U, Höhling HJ. Aspects of Collagen Mineralization in Hard Tissue Formation. Int. Rev. Cytol., vol. 242, Elsevier; 2004, p. 121–56.

169

[202] Hie M, Tsukamoto I. Administration of zinc inhibits osteoclastogenesis through the suppression of RANK expression in bone. Eur J Pharmacol 2011;668:140–6.

[203] Seaborn CD, Nielsen FH. Silicon deprivation decreases collagen formation in wounds and bone, and ornithine transaminase enzyme activity in liver. Biol Trace Elem Res 2002;89:251–61.

[204] Sommer B, Bickel M, Hofstetter W, Wetterwald A. Expression of matrix proteins during the development of mineralized tissues. Bone 1996;19:371–80.

[205] Yamaguchi M, Hashizume M. Effect of β-alanyl-L-histidinato zinc on protein components in osteoblastic MC3T3-El cells: Increase in osteocalcin, insulin-like growth factor-I and transforming growth factor-β. Mol Cell Biochem 1994;136:163–9.

[206] Varanasi V, Saiz E, Loomer P, Ancheta B, Uritani N, Ho S, et al. Enhanced osteocalcin expression by osteoblast-like cells (MC3T3-E1) exposed to bioactive coating glass (SiO 2–CaO–P 2 O 5–MgO–K 2 O–Na 2 O system) ions. Acta Biomater 2009;5:3536–47.

[207] Barradas AMC, Fernandes HAM, Groen N, Chai YC, Schrooten J, van de Peppel J, et al. A calcium-induced signaling cascade leading to osteogenic differentiation of human bone marrow-derived mesenchymal stromal cells. Biomaterials 2012;33:3205–15.

[208] Zhai W, Lu H, Chen L, Lin X, Huang Y, Dai K, et al. Silicate bioceramics induce angiogenesis during bone regeneration. Acta Biomater 2012;8:341–9.

[209] Holloway WR, Collier FM, Herbst RE, Hodge JM, Nicholson GC. Osteoblast-mediated effects of zinc on isolated rat osteoclasts: Inhibition of bone resorption and enhancement of osteoclast number. Bone 1996;19:137–42.

[210] Beck GR, Ha S-W, Camalier CE, Yamaguchi M, Li Y, Lee J-K, et al. Bioactive silica- based nanoparticles stimulate bone-forming osteoblasts, suppress bone-resorbing osteoclasts, and enhance bone mineral density in vivo. Nanomedicine Nanotechnol Biol Med 2012;8:793–803.

[211] Bai X, Sandukas S, Appleford MR, Ong JL, Rabiei A. Deposition and investigation of functionally graded calcium phosphate coatings on titanium. Acta Biomater 2009;5:3563– 72.

[212] Ginebra MP, Espanol M, Montufar EB, Perez RA, Mestres G. New processing approaches in calcium phosphate cements and their applications in regenerative medicine. Acta Biomater 2010;6:2863–73.

170

[213] Saikia K, Bhattacharya T, Bhuyan S, Talukdar D, Saikia S, Jitesh P. Calcium phosphate ceramics as bone graft substitutes in filling bone tumor defects. Indian J Orthop 2008;42:169–72.

[214] Castellani C, Lindtner RA, Hausbrandt P, Tschegg E, Stanzl-Tschegg SE, Zanoni G, et al. Bone–implant interface strength and osseointegration: Biodegradable magnesium alloy versus standard titanium control. Acta Biomater 2011;7:432–40.

[215] Bärenholdt O, Kolthoff N, Nielsen SP. Effect of long-term treatment with strontium ranelate on bone strontium content. Bone 2009;45:200–6.

[216] Paunier L. Effect of magnesium on phosphorus and calcium metabolism. Monatsschrift Kinderheilkd Organ Dtsch Ges Für Kinderheilkd 1992;140:S17–20.

[217] Pietak AM, Reid JW, Stott MJ, Sayer M. Silicon substitution in the calcium phosphate bioceramics. Biomaterials 2007;28:4023–32.

[218] Alt V, Kögelmaier DV, Lips KS, Witt V, Pacholke S, Heiss C, et al. Assessment of angiogenesis in osseointegration of a silica-collagen biomaterial using 3D-nano-CT. Acta Biomater 2011;7:3773–9.

[219] Porter AE, Patel N, Skepper JN, Best SM, Bonfield W. Effect of sintered silicate- substituted hydroxyapatite on remodelling processes at the bone–implant interface. Biomaterials 2004;25:3303–14.

[220] Yamaguchi M, Goto M, Uchiyama S, Nakagawa T. Effect of zinc on gene expression in osteoblastic MC3T3-E1 cells: enhancement of Runx2, OPG, and regucalcin mRNA expressions. Mol Cell Biochem 2008;312:157–66.

[221] Long F. Building strong bones: molecular regulation of the osteoblast lineage. Nat Rev Mol Cell Biol 2011;13:27–38.

[222] Maeno S, Niki Y, Matsumoto H, Morioka H, Yatabe T, Funayama A, et al. The effect of calcium ion concentration on osteoblast viability, proliferation and differentiation in monolayer and 3D culture. Biomaterials 2005;26:4847–55.

[223] Eklou-Kalonji E, Denis I, Lieberherr M, Pointillart A. Effects of extracellular calcium on the proliferation and differentiation of porcine osteoblasts in vitro. Cell Tissue Res 1998;292:163–71.

[224] Ma S, Yang Y, Carnes DL, Kim K, Park S, Oh SH, et al. Effects of dissolved calcium and phosphorous on osteoblast responses. J Oral Implantol 2005;31:61–7.

171

[225] Sila-Asna M, Bunyaratvej A, Maeda S, Kitaguchi H, Bunyaratavej N. Osteoblast differentiation and bone formation gene expression in strontium-inducing bone marrow mesenchymal stem cell. Kobe J Med Sci 2007;53:25–35.

[226] Leidi M, Dellera F, Mariotti M, Banfi G, Crapanzano C, Albisetti W, et al. Nitric oxide mediates low magnesium inhibition of osteoblast-like cell proliferation. J Nutr Biochem n.d.

[227] Guo B, Yang M, Liang D, Yang L, Cao J, Zhang L. Cell apoptosis induced by zinc deficiency in osteoblastic MC3T3-E1 cells via a mitochondrial-mediated pathway. Mol Cell Biochem 2012;361:209–16.

[228] Krause C, de Gorter DJJ, Karperien M, ten Dijke P, Asbmr. Chapter 2. Signal Transduction Cascades Controlling Osteoblast Differentiation. Prim. Metab. Bone Dis. Disord. Miner. Metab., John Wiley & Sons, Inc.; 2009, p. 10–6.

[229] Beck GR, Zerler B, Moran E. Gene Array Analysis of Osteoblast Differentiation. Cell Growth Differ 2001;12:61–83.

[230] Chlebowski JF, Mabrey S, Falk MC. Calorimetry of Alkaline Phosphatase. Stability of the Monomer and Effect of Metal Ion and Phosphate Binding on Dimer Stability. J Biol Chem 1979;254:5745–53.

[231] Yavropoulou MP, Yovos JG, others. The role of the Wnt signaling pathway in osteoblast commitment and differentiation. Horm-ATHENS- 2007;6:279.

[232] Bradley EW, Drissi MH. WNT5A Regulates Chondrocyte Differentiation through Differential Use of the CaN/NFAT and IKK/NF- B Pathways. Mol Endocrinol 2010;24:1581–93.

[233] Chattopadhyay N, Quinn SJ, Kifor O, Ye C, Brown EM. The calcium-sensing receptor (CaR) is involved in strontium ranelate-induced osteoblast proliferation. Biochem Pharmacol 2007;74:438–47.

[234] MacLeod RJ, Hayes M, Pacheco I. Wnt5a Secretion Stimulated by the Extracellular Calcium-Sensing Receptor Inhibits Defective Wnt Signaling in Colon Cancer Cells. Am J Physiol - Gastrointest Liver Physiol 2007;293:G403–G411.

[235] Chen X-Z, Vassilev PM, Basora N, Peng J-B, Nomura H, Segal Y, et al. Polycystin-L is a calcium-regulated cation channel permeable to calcium ions. Nature 1999;401:383–6.

[236] Hogstrand C, Kille P, Nicholson RI, Taylor KM. Zinc transporters and cancer: a potential role for ZIP7 as a hub for tyrosine kinase activation. Trends Mol Med 2009;15:101–11.

172

[237] Rasgado-Flores H, Sanchez-Armass S, Blaustein MP, Nachshen DA. Strontium, barium, and manganese metabolism in isolated presynaptic nerve terminals. Am J Physiol 1987;252:C604–610.

[238] Günther T. Functional compartmentation of intracellular magnesium. Magnesium 1986;5:53–9.

[239] Fan Q-M, Yue B, Bian Z-Y, Xu W-T, Tu B, Dai K-R, et al. The CREB-Smad6-Runx2 axis contributes to the impaired osteogenesis potential of bone marrow stromal cells in fibrous dysplasia of bone. J Pathol 2012.

[240] Cho HH, Shin KK, Kim YJ, Song JS, Kim JM, Bae YC, et al. NF-κB activation stimulates osteogenic differentiation of mesenchymal stem cells derived from human adipose tissue by increasing TAZ expression. J Cell Physiol 2010;223:168–77.

[241] Koga T, Matsui Y, Asagiri M, Kodama T, Crombrugghe B de, Nakashima K, et al. NFAT and Osterix cooperatively regulate bone formation. Nat Med 2005;11:880–5.

[242] Fromigué O, Haÿ E, Barbara A, Marie PJ. Essential Role of Nuclear Factor of Activated T Cells (NFAT)-Mediated Wnt Signaling in Osteoblast Differentiation Induced by Strontium Ranelate. J Biol Chem 2010;285:25251–8.

[243] Elin RJ. Magnesium: the fifth but forgotten electrolyte. Am J Clin Pathol 1994;102:616– 22.

[244] Kubota T, Shindo Y, Tokuno K, Komatsu H, Ogawa H, Kudo S, et al. Mitochondria are intracellular magnesium stores: investigation by simultaneous fluorescent imagings in PC12 cells. Biochim Biophys Acta BBA - Mol Cell Res 2005;1744:19–28.

[245] Wei X, Ugurlu O, Ankit A, Acar HY, Akinc M. Dissolution behavior of Si,Zn-codoped tricalcium phosphates. Mater Sci Eng C 2009;29:126–35.

[246] Beswick A, Blom AW. Bone graft substitutes in hip revision surgery: a comprehensive overview. Injury 2011;42 Suppl 2:S40–46.

[247] Aulakh TS, Jayasekera N, Kuiper JH, Richardson JB. Long-term clinical outcomes following the use of synthetic hydroxyapatite and bone graft in impaction in revision hip arthroplasty. Biomaterials 2009;30:1732–8.

[248] Li S, de Wijn JR, Li J, Layrolle P, de Groot K. Macroporous Biphasic Calcium Phosphate Scaffold with High Permeability/Porosity Ratio. Tissue Eng 2003;9:535–48.

[249] Tanaka T, Kumagae Y, Saito M, Chazono M, Komaki H, Kikuchi T, et al. Bone formation and resorption in patients after implantation of beta-tricalcium phosphate blocks

173

with 60% and 75% porosity in opening-wedge high tibial osteotomy. J Biomed Mater Res B Appl Biomater 2008;86:453–9.

[250] Nakashima Y, Sato T, Yamamoto T, Motomura G, Ohishi M, Hamai S, et al. Results at a minimum of 10 years of follow-up for AMS and PerFix HA-coated cementless total hip arthroplasty: impact of cross-linked polyethylene on implant longevity. J Orthop Sci Off J Jpn Orthop Assoc 2013.

[251] Iezzi G, Malchiodi L, Quaranta A, Ghensi P, Piattelli A. Peri-implant Bone Response Around a Human Hydroxyapatite-Coated Implant Retrieved After a 10-Year Loading Period: A Case Report. Int J Oral Maxillofac Implants 2013;28:e190–194.

[252] Lindhorst D, Tavassol F, von See C, Schumann P, Laschke MW, Harder Y, et al. Effects of VEGF loading on scaffold-confined vascularization. J Biomed Mater Res A 2010;95:783–92.

[253] Overman JR, Farré-Guasch E, Helder MN, ten Bruggenkate CM, Schulten EAJM, Klein- Nulend J. Short (15 minutes) bone morphogenetic protein-2 treatment stimulates osteogenic differentiation of human adipose stem cells seeded on calcium phosphate scaffolds in vitro. Tissue Eng Part A 2013;19:571–81.

[254] Kim J-W, Jung I-H, Jeong I-H, Lee K-I, Jung U-W, Kim C-S, et al. Volumetric bone regenerative efficacy of biphasic calcium phosphate-collagen composite block loaded with rhBMP-2 in vertical bone augmentation model of a rabbit calvarium. J Biomed Mater Res A 2012;100:3304–13.

[255] Ripamonti U, Teare J, Ferretti C. A Macroporous Bioreactor Super Activated by the Recombinant Human Transforming Growth Factor-β(3). Front Physiol 2012;3:172.

[256] Komaki H, Tanaka T, Chazono M, Kikuchi T. Repair of segmental bone defects in rabbit tibiae using a complex of beta-tricalcium phosphate, type I collagen, and fibroblast growth factor-2. Biomaterials 2006;27:5118–26.

[257] McKay WF, Peckham SM, Badura JM. A comprehensive clinical review of recombinant human bone morphogenetic protein-2 (INFUSE Bone Graft). Int Orthop 2007;31:729–34.

[258] Poynton AR, Lane JM. Safety profile for the clinical use of bone morphogenetic proteins in the spine. Spine 2002;27:S40–48.

[259] Carragee EJ, Hurwitz EL, Weiner BK. A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: emerging safety concerns and lessons learned. Spine J Off J North Am Spine Soc 2011;11:471–91.

[260] Fielding GA, Smoot W, Bose S. Effects of SiO2 , SrO, MgO and ZnO dopants in TCP on osteoblastic Runx2 expression. J Biomed Mater Res A 2013.

174

[261] Tarafder S, Davies NM, Bandyopadhyay A, Bose S. 3D printed tricalcium phosphate bone tissue engineering scaffolds: effect of SrO and MgO doping on in vivo osteogenesis in a rat distal femoral defect model. Biomater Sci 2013.

[262] Fielding G, Bose S. SiO2 and ZnO dopants in three-dimensional printed tricalcium phosphate scaffolds enhance osteogenesis and angiogenesis in vivo. Acta Biomater 2013.

[263] Roy M, Fielding GA, Bandyopadhyay A, Bose S. Effects of zinc and strontium substitution in tricalcium phosphate on osteoclast differentiation and resorption. Biomater Sci 2012;1:74–82.

[264] Jugdaohsingh R, Tucker KL, Qiao N, Cupples LA, Kiel DP, Powell JJ. Dietary silicon intake is positively associated with bone mineral density in men and premenopausal women of the Framingham Offspring cohort. J Bone Miner Res Off J Am Soc Bone Miner Res 2004;19:297–307.

[265] Buchet R, Millán JL, Magne D. Multisystemic functions of alkaline phosphatases. Methods Mol Biol Clifton NJ 2013;1053:27–51.

[266] Reid JW, Pietak A, Sayer M, Dunfield D, Smith TJN. Phase formation and evolution in the silicon substituted tricalcium phosphate/apatite system. Biomaterials 2005;26:2887– 97.

[267] Ming CQ, Greish Y, El-Ghannam1 A. Crystallization behavior of silica-calcium phosphate biocomposites: XRD and FTIR studies. J Mater Sci Mater Med 2004;15:1227– 35.

[268] Gillespie P, Wu G, Sayer M, Stott MJ. Si complexes in calcium phosphate biomaterials. J Mater Sci Mater Med 2010;21:99–108.

[269] Sayer M, Stratilatov A., Reid J, Calderin L, Stott M., Yin X, et al. Structure and composition of silicon-stabilized tricalcium phosphate. Biomaterials 2003;24:369–82.

[270] Schroeder LW, Dickens B, Brown WE. Crystallographic studies of the role of Mg as a stabilizing impurity in β-Ca3(PO4)2. II. Refinement of Mg-containing β-Ca3(PO4)2. J Solid State Chem 1977;22:253–62.

[271] Wei X, Akinc M. Crystal Structure Analysis of Si- and Zn-Codoped Tricalcium Phosphate by Neutron Powder Diffraction. J Am Ceram Soc 2007;90:2709–15.

[272] Pichavant M. Effects of B and H 2 O on Liquidus Phase Relations in the Haplogranite System at L Kbar. Am Miner 1987;72:1056–70.

[273] Driessens FCM, Verbeeck RMH. Biominerals. Boca Raton: CRC Press; 1990.

175

[274] Bohner, Lemaître, Ring. Kinetics of Dissolution of beta-Tricalcium Phosphate. J Colloid Interface Sci 1997;190:37–48.

[275] Wan M, Cao X. BMP signaling in skeletal development. Biochem Biophys Res Commun 2005;328:651–7.

[276] Nohe A, Keating E, Knaus P, Petersen NO. Signal transduction of bone morphogenetic protein receptors. Cell Signal 2004;16:291–9.

[277] Lee M-H, Kim Y-J, Kim H-J, Park H-D, Kang A-R, Kyung H-M, et al. BMP-2-induced Runx2 expression is mediated by Dlx5, and TGF-beta 1 opposes the BMP-2-induced osteoblast differentiation by suppression of Dlx5 expression. J Biol Chem 2003;278:34387–94.

[278] Celil AB, Campbell PG. BMP-2 and Insulin-like Growth Factor-I Mediate Osterix (Osx) Expression in Human Mesenchymal Stem Cells via the MAPK and Protein Kinase D Signaling Pathways. J Biol Chem 2005;280:31353–9.

[279] Yakymovych I, Ten Dijke P, Heldin CH, Souchelnytskyi S. Regulation of Smad signaling by protein kinase C. FASEB J Off Publ Fed Am Soc Exp Biol 2001;15:553–5.

[280] Aubin J, Davy A, Soriano P. In vivo convergence of BMP and MAPK signaling pathways: impact of differential Smad1 phosphorylation on development and homeostasis. Genes Dev 2004;18:1482–94.

[281] Tada H, Nemoto E, Kanaya S, Hamaji N, Sato H, Shimauchi H. Elevated extracellular calcium increases expression of bone morphogenetic protein-2 gene via a calcium channel and ERK pathway in human dental pulp cells. Biochem Biophys Res Commun 2010;394:1093–7.

[282] Komori T. Regulation of osteoblast differentiation by Runx2. Adv Exp Med Biol 2010;658:43–9.

[283] Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K, et al. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 1997;89:755–64.

[284] Takami M, Takahashi N, Udagawa N, Miyaura C, Suda K, Woo JT, et al. Intracellular calcium and protein kinase C mediate expression of receptor activator of nuclear factor- kappaB ligand and osteoprotegerin in osteoblasts. Endocrinology 2000;141:4711–9.

[285] Xiao Z, Zhang S, Magenheimer BS, Luo J, Quarles LD. Polycystin-1 Regulates Skeletogenesis through Stimulation of the Osteoblast-specific Transcription Factor RUNX2-II. J Biol Chem 2008;283:12624–34.

176

[286] Boyce BF, Xing L, Chen D. Osteoprotegerin, the bone protector, is a surprising target for beta-catenin signaling. Cell Metab 2005;2:344–5.

[287] Enomoto H, Shiojiri S, Hoshi K, Furuichi T, Fukuyama R, Yoshida CA, et al. Induction of Osteoclast Differentiation by Runx2 through Receptor Activator of Nuclear Factor-κB Ligand (RANKL) and Osteoprotegerin Regulation and Partial Rescue of Osteoclastogenesis in Runx2–/– Mice by RANKL Transgene. J Biol Chem 2003;278:23971–7.

[288] Marie PJ. Strontium ranelate: New insights into its dual mode of action. Bone 2007;40:S5–S8.

[289] Caverzasio J. Strontium ranelate promotes osteoblastic cell replication through at least two different mechanisms. Bone 2008;42:1131–6.

[290] Helmick CG, Felson DT, Lawrence RC, Gabriel S, Hirsch R, Kwoh CK, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part I. Arthritis Rheum 2008;58:15–25.

[291] Lawrence RC, Felson DT, Helmick CG, Arnold LM, Choi H, Deyo RA, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part II. Arthritis Rheum 2008;58:26–35.

[292] Salgado AJ, Coutinho OP, Reis RL. Bone tissue engineering: state of the art and future trends. Macromol Biosci 2004;4:743–65.

[293] Banwart JC, Asher MA, Hassanein RS. Iliac crest bone graft harvest donor site morbidity. A statistical evaluation. Spine 1995;20:1055–60.

[294] Graham SM, Leonidou A, Aslam-Pervez N, Hamza A, Panteliadis P, Heliotis M, et al. Biological therapy of bone defects: the immunology of bone allo-transplantation. Expert Opin Biol Ther 2010;10:885–901.

[295] Shuai C, Li P, Liu J, Peng S. Optimization of TCP/HAP ratio for better properties of calcium phosphate scaffold via selective laser sintering. Mater Charact 2013;77:23–31.

[296] Guda T, Walker JA, Singleton B, Hernandez J, Oh DS, Appleford MR, et al. Hydroxyapatite scaffold pore architecture effects in large bone defects in vivo. J Biomater Appl 2013.

[297] Dahl M, Jørgensen NR, Hørberg M, Pinholt EM. Carriers in mesenchymal stem cell osteoblast mineralization-State-of-the-art. J Cranio-Maxillo-Facial Surg Off Publ Eur Assoc Cranio-Maxillo-Facial Surg 2013.

177

[298] Arinzeh TL, Tran T, Mcalary J, Daculsi G. A comparative study of biphasic calcium phosphate ceramics for human mesenchymal stem-cell-induced bone formation. Biomaterials 2005;26:3631–8.

[299] Alvarez-Barreto JF, Linehan SM, Shambaugh RL, Sikavitsas VI. Flow perfusion improves seeding of tissue engineering scaffolds with different architectures. Ann Biomed Eng 2007;35:429–42.

[300] Sikavitsas VI, Bancroft GN, Holtorf HL, Jansen JA, Mikos AG. Mineralized matrix deposition by marrow stromal osteoblasts in 3D perfusion culture increases with increasing fluid shear forces. Proc Natl Acad Sci 2003;100:14683–8.

[301] Grayson WL, Bhumiratana S, Cannizzaro C, Chao P-HG, Lennon DP, Caplan AI, et al. Effects of initial seeding density and fluid perfusion rate on formation of tissue-engineered bone. Tissue Eng Part A 2008;14:1809–20.

[302] Vandiver J, Dean D, Patel N, Botelho C, Best S, Santos JD, et al. Silicon addition to hydroxyapatite increases nanoscale electrostatic, van der Waals, and adhesive interactions. J Biomed Mater Res A 2006;78:352–63.

[303] Rubin J, Rubin C, Jacobs CR. Molecular pathways mediating mechanical signaling in bone. Gene 2006;367:1–16.

[304] Weinbaum S, Cowin SC, Zeng Y. A model for the excitation of osteocytes by mechanical loading-induced bone fluid shear stresses. J Biomech 1994;27:339–60.

[305] Kamel MA, Picconi JL, Lara-Castillo N, Johnson ML. Activation of β-catenin signaling in MLO-Y4 osteocytic cells versus 2T3 osteoblastic cells by fluid flow shear stress and PGE2: Implications for the study of mechanosensation in bone. Bone 2010;47:872–81.

[306] Reich KM, Frangos JA. Effect of flow on prostaglandin E2 and inositol trisphosphate levels in osteoblasts. Am J Physiol 1991;261:C428–432.

[307] Kapur S, Baylink DJ, William Lau K-H. Fluid flow shear stress stimulates human osteoblast proliferation and differentiation through multiple interacting and competing signal transduction pathways. Bone 2003;32:241–51.

[308] Nagatomi J, Arulanandam BP, Metzger DW, Meunier A, Bizios R. Cyclic Pressure Affects Osteoblast Functions Pertinent to Osteogenesis. Ann Biomed Eng 2003;31:917– 23.

[309] Liu J, Zhao Z, Li J, Zou L, Shuler C, Zou Y, et al. Hydrostatic pressures promote initial osteodifferentiation with ERK1/2 not p38 MAPK signaling involved. J Cell Biochem 2009;107:224–32.

178

Appendix A

Journal Publications

1. Fielding GA, Bandyopadhyay A, Bose S. Effects of SiO2 and ZnO doping on mechanical and biological properties of 3D printed TCP scaffolds. Dent Mater 2012;28:113-22

2. Bandyopadhyay A, Feuerstein J, Fielding GA, Banerjee S, Bose S. ZnO, SiO2 and SrO doping in resorbable tricalcium phosphates: Influence on strength degradation, mechanical properties and in vitro bone cell material interactions. JBMR B 2012;28:113- 22

3. Roy M, Fielding GA, Bandyopadhyay A, Bose S. Effects of zinc and strontium substitution in tricalcium phosphate on osteoclast differentiation and resorption. Biomaterials Science 2013;1:74-82

4. Roy M, Fielding GA, Bandyopadhyay A, Bose S. Mechanical, In Vitro Antimicrobial and Biological Properties of Plasma Sprayed Silver-Doped Hydroxyapatite Coating. ACS: App Mater & Inter 2012;4:1341-9

5. Fielding GA, Roy M, Bandyopadhyay A, Bose S. Mechanical, Antibacterial and biological characteristics of plasma sprayed silver and strontium doped hydroxyapatite coatings. Acta Biomaterialia 2012;8:3144-52

6. Dhal J, Fielding G, Bose S, et al. Understanding bioactivity and polarizability of hydroxyapatite doped with tungsten. JBMR B 2012;100B:1836-45

7. Fielding GA, Bandyopadhyay A, Bose S. SiO2 and ZnO dopants in three-dimensionally printed tricalcium phosphate bone tissue engineering scaffolds enhances osteogenesis and angiogenesis in vivo. Acta Biomaterialia. http://dx.doi.org/10.1016/j.actbio.2013.07.009,

8. Bose S, Fielding GA, Tarafder SM. Mechanistic understanding of trace element induced osteogenesis and angiogenesis in calcium phosphate ceramics. Featured Review in Trends In Biotechnology. http://dx.doi.org/10.1016/j.tibtech.2013.06.005

9. Fielding GA, Smoot W, Bose S. Effects of SiO2 , SrO, MgO and ZnO dopants in TCP on osteoblastic Runx2 expression. JBMR B 2013. doi: 10.1002/jbm.a.34909

179

To Be Submitted

1. Nandi SK, Fielding GA, Bandyopadhyay A, Bose S. Effects of SiO2 and ZnO doping on in vivo biological properties of 3D printed β-TCP scaffolds. Under Revision

2. Fielding GA, Bose S. Effects of SiO2 and ZnO dopants in porous TCP scaffolds on osteoblastic differentiation. Under Preparation

3. Fielding GA, Bandyopadhyay A, Bose S. Effects of SiO2 and ZnO binary dopant system on osteoblasts in porous TCP scaffolds cultured in a flow perfusion bioreactor. Under Preparation

Conference Proceedings

1. Fielding G, Feuerstein J, Bandyopadhyay A, Bose S. SiO2 and SrO Doped β-TCP: Influence of Dopants on Mechanical and Biological Properties. In: Narayan R, Bose S, Bandyopadhyay A, editors. Biomater. Sci. Process. Prop. Appl. II, John Wiley & Sons, Inc.; 2012, p. 171–81.

2. Roy M, Fielding G, Bose S. Influences of Sr, Zn and Mg Dopants on Osteoclast Differentiation and Resorption. In: Narayan R, Bose S, Bandyopadhyay A, editors. Biomater. Sci. Process. Prop. Appl. II, John Wiley & Sons, Inc.; 2012, p. 227–37.

Presentations

1. Fielding GA, Bandyopadhyay A, Bose S. “The effect of SiO2 and ZnO doping on the mechanical and biological properties of β-TCP scaffolds fabricated using three-dimensional printing”. Podium presentation, MS&T, Columbus, OH (2011)

2. Fielding GA, Roy M, Bandyopadhyay A, Bose S. “Antimicrobial Effects and Cell-Materials Interactions of Sr/Ag Doped Hydroxyapatite Coatings.” Podium Presentation. MS&T, Columbus, OH (2011)

3. Fielding GA, Feuerstein J, Bandyopadhyay A, Bose S. “SiO2 and SrO Doped β-TCP: Influence of Dopants on Mechanical and Biological Properties.” Poster Presentation. MS&T, Columbus, OH (2011)

180

4. Nandi, SK, Fielding GA, Bose S. “Effects of SiO2 and ZnO doping on in vivo biological properties of 3D printed β-TCP scaffolds”. Northwest Biomechanics Symposium, Moscow, ID (2013)

181

Appendix B

MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) Calculations

MTT (Sigma, St. Louis, MO) solution of 5 mg/ml is prepared by dissolving MTT in PBS and the

solution then filter-sterilized (0.2 µm). The MTT is then diluted (100 µL into 900 µL) in DMEM

medium and 1 mL of the diluted solution was then added to each of the samples. After 2 h of

incubation at 34 °C, solution is aspirated from each of the wells. 1 mL of solubilization solution

made up of 10% Triton X-100, 0.1N HCl, and isopropanol is added to dissolve the formazan

crystals. 200 µL of the resulting supernatant is then transferred into a 96-well plate, and read by a

plate reader at 570 nm.

Absorbance values are then averaged and either plotted or normalized to the control.

Example Calculations:

Average Control Absorbance: 1.35 ± 0.26, n = 3

Average Experimental Absorbance: 2.64 ± 0.84, n = 3

182

To normalize to control, divide averages by control: 2.64 = = 1.96 1.35 푁표푟푚푎푙푖푧푒푑 퐸푥푝푒푟푖푚푒푛푡푎푙 퐴푏푠표푟푏푎푛푐푒 1.35 = = 1.00 1.35

To propagate the standard푁표푟푚푎푙푖푧푒푑 deviation 퐶표푛푡푟표푙we use the퐴푏푠표푟푏푎푛푐푒 following equation where deltas equal standard deviations and variables are averages:

= ( ) + ( ) + ∆푧 ∆푥 2 ∆푦 2 � ⋯ For Experimental SD: 푧 푥 푦

0.26 0.84 = ( ) + ( ) 1.96 1.35 2.64 ∆푧 2 2 � = 0.729

For Control SD: ∆푧

0.26 0.84 = ( ) + ( ) 1 1.35 2.64 ∆푧 2 2 � = 0.372

So, our normalized values then become % ∆Viability푧 values:

Experimental: 1.96 ± 0.729

Control: 1 ± 0.372

183

Where the experimental samples is demonstrating 196% more cellular proliferation than the control sample.

Alamar Blue Assay

Cellular proliferation can be monitored with Alamar Blue purchased from Life Technologies

(Carlsbad, CA). The manufacturer’s recommended protocol should be used for the experiments.

Briefly, at predetermined day points, sample media is replaced with 1mL of fresh media and 100

µL of Alamar blue reagent is added to each culture well. Samples are then incubated for 8 hours and then 200µL of media is extracted from each well and placed in triplicate in a 96 well plate.

Fluorescence can be detected using our Synergy Hybrid multi-mode microplate reader (Biotek,

Winooski, VT) with excitation wavelength of 540 nm and emission wavelength of 590 nm.

Samples can next be rinsed with fresh media and placed back in incubator for further downstream applications. Fluorescence intensity is plotted. Example data are as follows:

184

1 2 3 4 5 6 7 8 9 10 11 12 P-BR 513 433 516 519 362 361 540/25,620/40 495.25 41.57223 Pure 575 470 461 258 308 315 540/25,620/40 502 63.37981 D-BR 379 357 356 367 322 378 540/25,620/40 359.8333 20.98968 D 311 457 377 329 359 336 540/25,620/40 361.5 52.19866 540/25,620/40

540/25,620/40

540/25,620/40 540/25,620/40

The raw data are in the software generated table. To the right are averages and standard deviations for each sample for each composition.

Averages and SDs are plotted and statistical analysis performed to demonstrate any significant differences:

Critical Issues:

1. The incubation time really has to be optimized for each experiment. Generally 8 hours should be sufficient, but may need up to 24 hours. Check on the sample every two hours after initially putting in the incubator to look for a pink color change.

185

In Vivo Sample Preparation

Fixation

All in vivo bone sample with scaffolds needs to be kept immediately after surgery into 10% neutral buffered formalin for fixation. Fixation should be done for at least 72 h. Fixation step is the same whether it is for undecalcified or decalcified tissue sections. All procedures after fixation are different for undecalcified and decalcified tissue sections.

Undecalcified Tissue Sections Processing

Fixation:

Keep all in vivo samples in 10 % neutral buffered for 72 h formalin immediately. Dehydration:

1. Wash all formalin fixed tissue samples 2-3 times by DI water.

2. Keep tissue samples in 70 % ethanol for 6-8 h (or overnight).

3. Keep tissue samples in 95 % ethanol for 6-8 h.

4. Keep tissue samples in 100 % ethanol for for 6-8 h (2 times).

5. Keep tissue samples in 1:1 ethanol: acetone for 6-8 h

6. Keep tissue samples in 100% acetone for 6h (2 times). Infiltration:

186

1. Acetone : Spurrs (2:1) for 8 h-12 h (or overnight).

2. Acetone : Spurrs (1:1) for 8 h-12 h (or overnight).

3. Now fill the tissue sample tube/vial with acetone : Spurrs (1:1) and keep for 8 h-12 h

(or overnight) with lid open. This will cause acetone evaporate and a better infiltration

of SPURRs into specimen. Make sure there is enough acetone: Spurrs (1:1) in the

tube/vial so that when acetone evaporates, the sample get fully covered with SPURRs.

4. Now keep the tissue sample in 100 % spurrs for 6 to 8 h (2 times).

5. Putt the specimen in the mold and fill that up with 100 % spurs, and keep this at 60

°C overnight.

Warning and precautions:

1. Formalin and SPURRs is toxic and carcinogenic. So work with proper knowledge of

their handling and disposal. Always wear protective clothing and gloves.

2. SPURRs is a hydrophobic polymer. Therefore, the dehydration has to be done with

absolute perfection. Any water left in the tissue sample will leave void while

infiltrating and embedding. More time in dehydration will not harm the sample, but

less time can have a detrimental effect.

3. Don’t leave sample in SPURRs (without changing) for more than one night

as polymerization may occur.

4. Up to the infiltration step, work should be done in glass vials (as acetone may

damage a plastic container). Embedding has to be done in a plastic mould (NO

GLASS VIAL), as removing the polymerized SPURRs is extremely difficult from a

187

glass vial.

Plastic embedded tissue sample cutting and slide preparation

Cutting:

Samples can be cut into thick tissue sections using a diamond saw. Tissue sections cut

using a diamond saw needs to be polished to make it thin before staining.

Slide preparation

Affix the tissue sections on a glass slide using heavy duty super glue. Let it dry, and

then polish it. Tissue slide is then ready for staining.

Note:

Make sure you use very strong glue to affix the tissue section on the glass slide.

Otherwise, tissue section will come off the glass slide during polishing in contact with

water. However, do not use a very strong glue to affix the glass slide with the holder

during polishing. The best super glue to affix the tissue slide with the holder is Hot Stuff

(Cyanoacrylate Adhesive) [Satellite City, Catalog # HS-4, www.caglue.com) instant glue.

This super glue can also be used to affix tissue sections on glass slides. The glue remover

solvent that acts best with this Hot Stuff is Super Solvent (Golden West, Woodland Hills,

CA). This Super Solvent can be used to detach the glass slides from the holder).

Modified Masson Goldner's trichrome of undecalcified tissue sections

Solution A

188

--Haematoxylin 10g

--Distilled water 1000ml

Ripen for at least 2 weeks before use

Solution B

--Ferric chloride (hydrated) 11.6g

--Distilled water 1000ml

--2% hydrochloric acid 10ml

For use mix equal points of A and B immediately before required. Do not keep working solution pre-made.

Ponceau de xylidine/acid fuchsin

--Ponceau de xylidine 1.5g

--Acid fuchsin 0.5g

--Acetic acid (conc) 2ml

--Distilled Water 98ml

--Azophloxine 0.5g

--Acetic acid (conc) 0.6ml

--Distilled water 99.4ml

189

Ponceau/acid fuchsin/azophloxine (working solution)

--Ponceau de Xylidine/acid fuchsin 12ml

--Azophloxine 8ml

--0.2% Acetic acid 80ml

--Re-use and keep made up.

Phosphomolybdic acid/orange G

--Phosphomolybolic acid 6g

--Orange G 4g

--Distilled water 1000ml Light green

--Light green 2g

--Acetic acid (conc.) 2ml

--Distilled water 1000ml Staining

Procedure:

1. Sections to distilled water x 2 15 minutes

2. Place sections in Weigert's haematoxylin 20 minutes

3. Wash in water.

4. Differentiate with 0.5% acid alcohol.

5. Wash in water 20 minutes

190

6. Ponceau/acid fuchsin/azophloxine 5 minutes

7. Rinse in 1% acetic acid 10 seconds

8. Phosphomolybdic acid/orange G 20 minutes

9. Repeat step 7

10. Light green 5 minutes

11. Rinse in water

12. Blot dry

13. Rinse in alcohol x 2

14. Wash in methyl cyclohexane x 2

15. Mount in Picomount

Interpretation of result:

Nuclei: Blue/black, Mineralized bone/collagen: Green, Osteoid: Orange/Reddish

Decalcified Tissue Section Processing

Fixation:

Keep all in vivo samples in 10 % neutral buffered for 72 h formalin immediately.

191

Making 14% EDTA solution

• Add 140 g EDTA to 700 ml distilled H2O

• On stir plate under the fume hood, add conc. ammonium hydroxide (around 30 mL)

drop wise until solution clears (if it is free acid/ disodium salt) or glacial acetic acid (if

it is tetra sodium salt) and check the solution pH. The final pH of this solution should

be around 7.4-7.6.

• Add H2O to almost 1L. Check pH and adjust with ammonium hydroxide/glacial

acetic acid drop wise up to pH 7.4-7.6, then adjust final volume to 1L.

Decalcification:

• Wash all formalin fixed tissue samples 2-3 times by DI water

• After washing, keep fixed tissue in at least 15 volumes of 14 % EDTA, and change

every 2 days, with mixing

• Time of decalcification varies with tissue size, species, etc (a rat femur might take 8 to

12 weeks).

• Rinse with DI H2O (4 times).

Paraffin Embedding:

• Place in 30%, 50%, 70% and 100% (2 times) ethanol for at least 60 min each. (Note:

Thick tissue specimen might need longer times. 4 to 6 h is good for rat/rabbit bone

samples. Keeping longer time in ethanol will not cause any detrimental effect, but less

time can cause a detrimental effect during embedding).

192

• 1:1 Xylene:Ethanol (4 to 6 h or overnight)

• 100 % Xylene (overnight)

• 100 % Xylene (4 to 6 h)

• Keep the sample in 1:1 xylene : paraffin at 37 °C until the paraffin dissolves into xylene

(occasional swirling/shaking accelerates the process).

• After the xylene is saturated with paraffin, place the vials at 60 °C for 4 h to overnight.

• Pour off half of the solution and replace with hot paraffin (this must be done quickly so

that the paraffin in the vial does not solidify). – 2 to 3 times in 4 to 12 h interval.

• Now embed the tissue sample in paraffin.

Notes:

o EDTA tetra-sodium salt is readily soluble in water.

o The pH of EDTA tetra-sodium salt is pH 7.6. The pH is adjusted down with glacial acetic acid using continuous stirring and a pH electrode in the solution.

EDTA disodium salt is sparingly soluble in water. That’s why it is necessary to

add 15-30 ml

o ammonium hydroxide initially to dissolve in water. EDTA tetrasodium salt is thus a better choice than EDTA disodium salt for making decalcification solution.

Paraffin embedded tissue sample cutting using microtome

Cutting:

193

a. Trim paraffin blocks using a scalpel blade before sectioning.

b. Cut tissue sections of 5-15 µm thick using a microtome.

c. Place paraffin ribbon in water bath set at about 40-45 ºC.

d. Break apart sections at intersections as necessary using a forceps.

e. Holding the slide about 90 degree angle, scoop up the section/s onto the glass slide.

f. Gently shake off the excess water from the slide.

g. Bake the slides at about 45-50 ºC

overnight.

Note:

Paraffin embedded decalcified bone tissue sections with ceramic scaffold can still be little

crunchy. Try to use the blade that gives you nice slides. If your samples are still hard and

gets rolled during cutting due to crunchiness, you may dip the paraffin embedded tissue

block for an hour into water facing the tissue upside down. This might give you enough

softness to cut using a microtome to obtain nice tissue sections.

TRAP Staining Staining of Decalcified Tissue Sections For TRAP staining, procedure provided in the following reference was followed: Ref: Jiang et al. J of Histochem Cytochem 2005; 53:593-602).

Buffer I Solution:

Sodium Acetate Anhydrous: 9.2g

194

L- (+)Tartaric Acid: 11.4g

Distilled Water: 950ml

Glacial Acetic Acid: 2.8ml

• Dissolve and adjust pH to 4.7-5.0 with 5M Sodium Hydroxide

• Bring total volume to 1L

Buffer II Solution:

Naphthol AS-BI Phosphate (store at –20º C): 0.1g

Ethylene Glycol Monoethyl Ether: 5ml

Buffer III Solution:

Sodium Nitrite: 1g

Distilled Water: 20ml

Buffer IV Solution:

Pararosaniline Chloride: 1g

2N HCL: 20ml

• Heat to 60º; C for 5 min do not boil.

• Filter through kimwipe.

195

Staining Procedure:

1. Preheat to 37°C 2 Coplin jars with 50ml of Buffer I Solution each.

2. Take one Coplin jar and add 0.5 ml of Buffer II Solution, add slides and incubate at

37°C for 45 minus.

3. A few min before the time is up, mix 1 ml of Buffer III Solution and 1 ml of Buffer IV

solution for 30 sec and let it sit for 2 min.

4. Add this mixed solution to the other Coplin jar of 50 ml Buffer I Solution, mix and add

the slides without rinsing.

5. Incubate at room temp ~5 min.

6. Rinse, counterstain with hematoxylin for 40 sec and put slides in 0.05% Ammonia

water.

7. Dehydrate through alcohol, clear in xylene and mount.

Antibody Staining Protocol (Confocal or Light)

A. Specimen Collection

Specimens should be from fresh autopsy or biopsy specimens, and fixed as soon as

possible. Fixed, and paraffin-embedded specimens are recommended. It is recommended

that proper in-house quality control and assay verification measures be taken to determine

suitability of specimen types for immunohistochemical methodologies.

196

B. Specimen Preparation

Specimens fixed in 10.0% neutral buffered formalin for 6-12 hours at room temperature

(18-25ºC) are optimal (9). Fixation should not exceed 24hours. Fixatives other than neutral buffered formalin, such as B5 (Formal sublimate), zinc formalin or Bouin’s may be used. It is recommended that the user verify optimal conditions. The following are some factors to be considered when evaluating the suitability of the users’ tissue processing.

Procedures:

1. Formaldehyde-BasedFixation(NeutralBufferedFormalinandBouin’s)

Most formaldehyde-based fixatives contain 10% formalin, a neutral salt to maintain tonicity, and a buffering system to maintain pH. These fixatives are well tolerated by tissues, exhibit good histological penetration and are well suited for labeled streptavidin- biotin immunostains. Specimens should be fixed from 6-12 hours depending on tissue thickness. Bouin’s solution is an alternative formaldehyde-based fixative, which contains picric acid and is suitable for use on all tissues except kidney. Excessively fixed tissues become brittle and adversely affect the appearance and quantity of lipids. After fixation, remove yellow color by treating with a saturated solution of lithium carbonate in 70% ethanol for 2 minutes. Follow fixation with dehydration, clearing, infiltration and embedding.

C. Tissue Sectioning, Handling, and Placement on Microscope Slides.

197

1. Section the tissues approximately four microns thick for optimum resolution with staining.

2. Place the sections into a container with deionized or distilled water preheated to 40-44ºC to relax the sections from compression due to sectioning. Do not add gelatin or polylysine, or any other agent to the water.

3. Place the sections on the slides as flat and wrinkle free as possible to optimize stain contact with the tissues.

4. Place the tissue on the slide with the painted portion side up.

Note: After placing the tissue sections on the slides, store slides at 18 - 25°C in the dark up to one month. Prior to use, the tissue sections should be dried (with a conventional or microwave oven) onto the slides.

5. Dry the tissue sections on the slides by heating in an oven at 60°C for a minimum of 60 minutes. Ensure that any moisture trapped under the tissue section is completely eliminated by melting the paraffin and evaporating of the water droplets. This eliminates the "water spotting" phenomenon when reading the slides.

6. Deparaffinize/Rehydrate tissues by dipping into the specified containers containing the followings:

1) Xylene 2 x 5 minutes

2) 100% Ethanol 2 x 2 minutes

198

3) 95% Ethanol 2 x 2 minutes

4) 70% Ethanol 1 x 2 minutes

5) 50% Ethanol 1 x 2 minutes

6) 1X Rinse Buffer 2 x 5 minutes

Note: Tissue sections should be used the same day they are deparaffinized.

D. Tissue Pretreatment.

1. Use one of the recognized retrieval methods (i.e., steam, microwave, pressure cooker or hotplate) to recover tissue antigenicity.

2. Blot dry the area around the tissue and then use a PAP Pen to draw a circle around each section.

3. Allow the PAP Pen barrier to dry for at least 2 minutes at room temperature but ensure that the tissue remains hydrated.

Note: Leave space between the tissue and the PAP Pen barrier as the PAP Pen

solution may adversely affect staining of the tissue, causing false negative results.

4. Using a Superfrost Pen, mark on the top of each slide which antibody will be used to stain the slide.

5. Apply enough 3% hydrogen peroxide (in distilled water) to the specimen so that it is completely covered.

199

6. Incubate 10-30 minutes in a hydrated incubation enclosure.

7. Rinse tissue section 2 x 5 minutes with TBST.

Note: Do not allow tissues to dry out during the staining procedure.

E. Staining Procedure

• Blocking Step

1. Apply about 3 drops (1 drop = ~45uL) of TBST-BSA to cover the specimen.

2. Incubate 15-30 minutes in an enclosed container.

3. While holding the slide at a 45o angle, gently rinse the specimen with 1X TBST for a minimum of 15 seconds. Tap the end of the slide onto a paper towel to remove excess

Rinse Buffer.

• Primary Antibody

1. Apply ~150 µL of the diluted Primary Antibody or a negative control reagent to cover the specimen.

Note: Optimal antibody titer must be determined prior to running assay. Recommended is a starting dilution for both antibodies of 1:200 in diluted blocking reagent (dilute 2 drops of ready-to-use blocking reagent into 5 mL of 1xPBS and use this diluted blocking reagent fordilution of antibody).

2. Incubate 1½ - 2hours in an enclosed container.

200

Note: Incubation time of primary antibody will depend on final dilution and specimen and should therefore be determined by end user.

3. Rinse tissue section 3 x 5 minutes with TBST.

• Secondary Antibodies (For Confocal)

1. Apply about 3 drops of the appropriate secondary antibody to slide

2. Incubate for one hour in an enclosed container at room temperature

3. Rinse tissue section 3 x 5 minutes with 1X TBST.

• Streptavidin-HRP (For Light Microscopy)

1. Apply about 3 drops of Streptavidin-HRP solution to cover the specimen.

2. Incubate 10-15 minutes in an enclosed container.

3. Rinse tissue section 3 x 5 minutes with TBST.

• Fast Green Counterstain (Optional)

1. Dip tissue slides 10-15 times in container with Fast Green

2. Rinse tissue sections under running tap water for 2 minutes.

3. Place the slides directly into a container filled with deionized water. Hold here until the next step.

201

• Mounting Fluid and Coverslipping

1. Remove slides from the water, and apply 1-2 drops of an aqueousbased mounting fluid

to the tissue and apply a coverslip.

2. Dehydrate through an inverse rehydration process (as described previously), apply 1-2

drops Xylene-based mounting fluid (e.g., Permount) and apply a coverslip.

Histomorphometry

The histomorphometrical analysis from light microscopy used during the course of this research was covered exhaustively by Egan et al. (Egan KP, Brennan TA, Pignolo RJ, Bone histomorphometry using free and commonly available software. Histopathology 61 (2012)

1168-1173. DOI: 10.1111/j.1365-2559.2012.04333.x)

202

Briefly, The steps are as follow.

1. A mask is created using readily available photo manipulation software selecting the color

of interest. In this case, mineralized bone formation (blue color) is of interest.

2. Once all of the blue color is selected, a black fill is inserted into each of the selected

regions.

3. The inverse is then selected and filled with white. What you end up with is a black and

white photo, where black represents new bone formation as seen above.

4. The figure is then imported to Imaje J, where the software is calibrated according to the

original scale bar.

5. The magic wand tool is then used to select all black areas and the total is measured

6. The output will give you the area in micrometers (based on the scale bar for calibration)

and will need to be converted to mm2. The data can then be averaged and plotted as seen

below.

203

Histomorphometry for confocal microscopy is a little different, but a very simple method. The steps are as follows:

2. Crop and convert the color of interest to grayscale. The original confocal image

already has the green and blue colors separated. For this particular micrograph,

the green color representing osteocalcin is the fluorophore of interest.

3. Next, import the grayscale image into ImageJ and select measure from the

Analyze toolbar.

4. Select the elliptical or rectangular selection tool and get three measurements from

the background. Ideal readings should have an intensity density close to zero.

5. Results should show up as in figure below.

6. Repeat for at least three samples of each type and day.

7. Once all of the data is collected, export to an Excel spreadsheet

204

Background area selected

Total Area Reading Background Readings

205

8. Take the average of the mean background readings

9. Calculate the Corrected total intensity for each sample by the following equation:

= ( 푇표푡푎푙 퐶표푟푟푒푐푡푒푑× 퐼푛푡푒푛푠푖푡푦 ) 퐼푛푡푒푔푟푎푡푒푑 퐷푒푛푠푖푡푦 − 퐴푟푒푎 Here is an example of one data set:퐴푣푒푟푎푔푒 퐵푎푐푘푔푟표푢푛푑 퐼푛푡푒푔푟푎푡푒푑 퐷푒푛푠푖푡푦

Background 0.1

6 Week Doped Area Mean IntDen RawIntDen Corrected Total Intensity Average SD 1 260099 23.713 6167717 6167717 6141707.1 8361684.7 2148510 2 260610 40.124 10456833 10456833 10430772 3 260100 32.828 8538585 8538585 8512575

For sample 1:

Total Corrected Intensity = 6,167,717 – (260099*0.1) = 6141707.1

10. Next take the average and standard deviation of the total corrected intensities for

each sample and plot

206

Critical Issues:

1. Make sure that you are being fair to what you are choosing for each sample. You cannot

conceivable select every piece of color in the trichrome staining for masking, so it is

important that you remain unbiased on your methods of selection.

AAS Ion Concentration Determination

Our lab has several AAS bulbs that are tuned to specific element wavelengths. Any user should first consult the Shimadzu AAS Cookbook to determine the correct settings of the instrument, mode to run in and matrix modifiers needed for possible interfering ions. For the purpose of this document, one of the most commonly read elements by our lab, Ca2+, will be explained.

Standard Preparation

1. Prepare a 50,000 ppm (50 g/L) La/Sr solution or 2% SrCl2 modifier solution.

2. Prepare serial standard dilutions with final volume of 5 mL.

3. The stock standard solutions are 1000 ppm of Ca2+

2+ Final Concentration Ca Standard Modifier Solution Nanopure H2O Ca2+ (µL) (µL) (mL) 0.5 ppm 50 500 4.45 1.0 ppm 100 500 4.40 2.0 ppm 200 500 4.30 3.0 ppm 300 500 4.20 4.0 ppm 400 500 4.10 5.0 ppm 500 500 4.00

Determination of calcium from rat urine

207

1. Centriguge each raw urine sample for 5 minutes at 5,000 RPM

2. Make each sample according to the following:

500 μL urine + 500 μL 2% SrCl2 solution + 4.00 mL of Nanopure H2O

3. Vortex each sample for 10 seconds before reading

Determination of calcium release from scaffold (in PBS)

1. Make each sample according to the following

500 μL sample solution + 4.50 mL of Nanopure H2O

2. Vortex each sample for 10 seconds before reading

The following are a set of example caclulations:

Standard Ca2+ Absorbance at concentration 422.67 nm (ppm) 0 0.0128 0.5 0.023 1 0.0316 2 0.0496 3 0.0659 4 0.0845 5 0.1016

1. Plot the standard concentrations vs. the absorbance values and fit with a linear line.

Display the equation as well as the R2 value

208

2. Analyze your experimental samples.

3. The software will automatically calculate the concentrations if dilution and all

appropriate data was entered. Alternatively the standard curve line equation can be used

to find the concentrations of ions in the sample by the following:

Sample Day Average Absorbance 1 0.128 2 0.123 3 0.316 5 0.423 10 0.329 15 0.148 20 0.352

Concentration can be found by utilizing the line equation:

= .0176 + .0137

Where y is the average absorbance reading푦 and x푥 is the concentration of the unknown. By rearranging the equation to solve for x, we get: 0.0137 = × ( 10) . 0176 푦 − 푥 푑푖푙푢푡푖표푛 푓푎푐푡표푟 푖푛 표푢푟 푐푎푠푒

209

After calculations we would plot the cumulative release as follows:

Sample Day Concentration 1 64.94318

2 62.10227 3 171.7614 5 232.5568 10 179.1477 15 76.30682 20 192.2159

210

Critical Issues:

1. Make sure that the lamps are warmed up before you start running your experiment. Sometimes they will read at a high level and the sensitivity changes as they warm up. Give them about 20 minutes before starting. Run a standard sample a couple of times and record the readings to make sure they are stable.

Fluorescent Area Scan Calculations

This method allows for quantitation of fluorescence from samples fixed for confocal microscopy.

1. Fix and apply primary and secondary antibodies for protein of interest as normal

procedure. The secondary antibody should be conjugated with a fluorophore that the

Biotek instrument has the capability of reading (Oregon Green, Texas Red, DAPI, etc.).

2. Immediately following the antibody labeling procedures, place samples in fresh 24 well

plate along with a “blank” sample that did not get antibody treatment. Samples should

have fluorescent side up.

3. Setup microplate reader for fluorescent reading with a 5x5 area scan. For Oregon Green,

excitation filter should be 485/20 and the emission filter should be 528/20.

4. Run the program and export the data to excel for analysis. Results should look similar to

example below:

211

485/20,528/20 A2 485/20,528/20 A4

1800 1800

1600 1600

1400 1400

1200 1200

1000 1000

800 800

600 600

400 400

200 200

Blank Pure SC Day 7 1 2 3 4 5 1 2 3 4 5 1 126 785 1 2 164 169 119 2 1192 1678 1413 122 160 159 115 79 3 219 1192 1655 1672 1361 3 4 666 1191 1388 4 120 127 113 5 306 5 87

5. Data can be analyzed by taking the average fluorescent intensity for each sample and

applying a scalar factor based on MTT data or Alamar Blue data and while discounting

any readings near or below average background readings. A scalar factor is necessary to

determine accurate expression levels. An easy way to think about it is: If sample one has

two times the amount of cells as sample two and twice as much average fluorescent

intensity, then the cells are really expressing similar amounts of protein on each sample.

The easiest way to correct this is to apply a scalar factor based on cell number.

An example calculation is as follows:

Fluorescent Area Scan Data

Control Average: 830.91 ± 250.21

Experimental Average: 798 ± 352.65

To apply the scalar factor, divide average area scan data by the normalized Alamar data.

212

830.91 = = 830.91 1 퐶표푛푡푟표푙 798 = = 961.53 0.82 퐸푥푝푒푟푖푚푒푛푡푎푙 Propagate standard deviation as previously explained in MTT section:

= ( ) + ( ) + ∆푧 ∆푥 2 ∆푦 2 � ⋯ 푧 푥 푦 Final results for corrected fluorescents intensity are then:

Control: 830.91 ± 250.21

Experimental: 961.53 ± 428.52

Plot the data in a bar chart as below:

213

Critical Issues:

1. You have the option to “turn off” fluorescent spots individually. Take care that you are

turning off only spots that are not likely directly on the sample rather than turning off

spots that are pointing our areas of low fluorescence on the sample.

qRT-PCR Analysis

qRT PCR is a relatively complex process, but is made easy through the use of RNA

extraction/purification kits, pre-validated primer sets and user friendly software for analysis. It is

recommended that RNA extraction and purification are done with commercially available spin

columns for confidence in extract and purity levels. Pre-validated high quality primer pairs are

also available for purchase from many companies for many different genes. Recommended

housekeeping genes for osteoblast type cells are: β-Actin and RPLP0.

The steps for the entire PCR process are as follows:

1. Culture cells normally on samples

2. At predetermined time points, extract RNA using spin columns or TRIzol method. Very

easy to follow protocols are readily available and generally come with the extraction kit

purchased.

3. Convert extracted RNA using Reverse Transcriptase enzyme. This protocol is usually

specific to the Reverse Transcriptase enzyme being purchased. Follow included

instructions for handling as well as thermocycling.

214

4. Once cDNA is obtained via reverse transcription, they may be stored in -80 °C or -20 °C

for several months.

5. cDNA is then amplified based on specific genetic targets. If pre-validated primer pairs

are purchased, a detailed protocol will be included. Follow handling and thermocycling

instructions.

6. Data is collected during the amplification process and is recorded. Details regarding the

calculation of genetic expression are below:

Cv values

• Number of cycles until minimum reading

• Higher concentrations of target genes will have lower Cv values

• The average Cv values are what is recorded for beginning the calculations

Step 1: Averaging all Cv values

Take the arithmetic mean for all technical and biological replicates for a particular target

The standard deviation has to be calculated from the pooled deviation method:

( ) ( ) ( ) = ( 2 ) ( )2 ( ) 2 ; 푛1−1 푆1 + 푛2−1 푆2 +⋯+ 푛푘−1 푆푘 푆푝 � 푛1−1 + 푛2−1 +⋯+ 푛푘−1 n= number of replicates for a certain sample

S = standard deviation for sample set

Step 2: Geometric mean of housekeeping genes (HKG)

215

Take the geometric mean for each biological replicate for each sample

= … 푛 The standard deviation is propagated퐺푒표푚푒푡푟푖푐 by the푚푒푎푛 following� equation:푎1푎2 푎 푛

a. = + + + 1 2 1 2 1 2 � 2 1 2 2 2 푛 Next푆 take� the∗ arithmetic푆 � � ∗mean푆 � of the⋯ geometric� ∗ 푆 � averages for each sample set

Propagate standard deviation by using the pooled standard deviation method in step 1

Step 3: Calculate ∆Ct

Delta Ct is the relative quantity of cDNA present in the sample

Subtract average Cq of reference genes from the average Cq of target gene:

Propagate standard deviation by∆ the퐶푡 following:퐶푞 푎푣푔 푡푎푟푔푒푡 − 퐶푞 푎푣푔 퐻퐾퐺

= + 2 2 푆 �푆푡푎푟푔푒푡 푆퐻퐺퐾

Step 4: Calculate ∆∆Ct

Subtract average ∆Ct of control sample target gene from the ∆Ct of the experimental sample target gene

푡 푡 exp 푡푎푟푔푒푡 푡 cont 푡푎푟푔푒푡 Propagate standard deviation∆ by:∆퐶 ∆퐶 − ∆퐶

= +

2 2 푆 �푆 ∆퐶푡 푡푎푟푔푒푡 푆∆퐶푡 퐻퐺퐾

216

Step 5: Calculate normalized expression ratio

This step gives the fold change between pure and doped samples

Determine the efficiency of the primer assays. This data is available from companies that you purchase your pre validated primer sets from.

Ex: For OPG is the efficiency is 99%, for Runx2 it is 109% and for BMP2 it is 100%

Multiply efficiency by 2

Ex: OPG = 2 * 0.99 = 1.98; Runx2 = 2*1.09 = 2.18; BMP2 = 2*1.00 = 2

The idea is that for 100% efficiency, you get double the amount of cDNA each cycle

Use the efficiency (E) in the following equation for each sample and target:

= −∆∆퐶푡 Propagate퐹표푙푑 퐶ℎ푎푛푔푒 the standard퐸 deviation (S) by the following equation:

( ) ( ) 2 ) 2 ) = ∆∆퐶 ∆∆퐶 − 퐹표푙푑 퐶ℎ푎푛푔푒 − 푆 푡 푡푎푟푔푒푡 2 − 퐹표푙푑 퐶ℎ푎푛푔푒 + 푆 푡 푡푎푟푔푒푡 − 푆

Step 6: Statistical Analysis

Use the normalized expression ratio and corresponding standard deviation to compare control groups (pure) to experimental groups (doped) for each target using an unpaired t-test

Alternatively, a one way ANOVA analysis can be done to compare similar targets across all samples

Step 7: Plot data and label significant differences

Sample calculations for Zn -TCP sample for BMP for day 21 results from this thesis are as follows:

217

Average Cq SD

34.17 0.801476285

Geometric Mean of Reference Genes

Average SD

19.68460166 0.224006085

∆Ct = 34.17 – 19.68 = 14.49 ± 0.83; ∆Ct for pure TCP BMP2 = 12.33 ± .70

∆∆Ct = 14.49 – 12.33 = 2.157 ± 1.09

Normalized Expression Ratio = 2- (2.157) = 0.224 ± 0.708

Data is normally plotted as ∆∆Ct or the normalized expression ratio. Generally a 4 fold increase as seen in the normalized expression ratio is considered to be biologically significant.

Critical Issues:

1. The most critical issue is determining what primer sets you are going to use. There are software that are available to help design primers, but generally it is much easier to use predesigned primer pairs for genetic targets. All of the large companies offer premade sets at various prices with different fluorescent labels. Our lab uses SYBR green labelling. 2. Make sure to follow directions exactly for thermocylcing temperatures. If buying premade primer sets, then they should all be using the same cycle (assuming all bought from same company)

Flow Simulation Analysis from SolidWorks

1. Open the fully assembled bioreactor model 2. Place caps on flow input and output ends of the assembly. There is a function in the program that allows you to do this or you can do this in assembly mode. 3. Choose the appropriate boundary conditions for the flow input and output. Generally flow rate in and atmospheric pressure at the outlet will be sufficient.

218

219

4. Next use run function to solve the system 5. Insert specific results such as fluid flow plots, surface plots for pressure or shear stress and point plots to get average pressures and stresses.

220

Critical Issues:

1. The flow rate of the bioreactor can me made smaller or larger with the use of different tubing sizes and/or the use of a filter on the end of each bioreactor and of course changing the pump settings. The most critical factor in determining flow rate is having the chamber always be completely full of culture media. 2. The temperature in which the bioreactor is running is important for the cell ines. Generally, the lower temperatures are good for proliferation, while higher temperatures are good for differentiation 3. Cell seeding has to be optimized. Cells should be seeded on samples in static conditions and let sit for at least several hours up to several days before starting the bioreactor experiment. 4. Media reservoir changes should also be optimized. During this study the media was changed every 7 days, but may need to be changed sooner or less. 5. Sterility is key. Make sure that everytime you handle the bioreactor use sterile techniques. If one reactor gets infected, they all will until a time where a 12 channel pump can be used.

221