The Osteogenic Properties of Novel Bioactive and Nanostructured Biomaterials for Mesenchymal Stromal Cell Differentiation

THE OSTEOGENIC PROPERTIES OF NOVEL BIOACTIVE AND NANOSTRUCTURED BIOMATERIALS FOR MESENCHYMAL STROMAL CELL DIFFERENTIATION

MATTHEW ILLSLEY

A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS OF THE UNIVERSITY OF BRIGHTON FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

APRIL 2011

THE UNIVERSITY OF BRIGHTON IN COLLABORATION WITH THE BLOND MCINDOE RESEARCH FOUNDATION

Abstract.

Soybeans have been in the human diet since the first millennium BC. They contain three major types of isoflavone, genistein, daidzein and glycitein. All three have been shown to have a powerful antioxidant effect and bind to oestrogen receptor beta with high specificity. It has been demonstrated that dietary soybeans favourably impact bone health in postmenopausal women by lowering the incidence of osteoporosis. The literature shows that genistein supplementation reduces marrow stromal cell (MSC) differentiation to an adipocytic pathway and may favourably stimulate osteogenesis. More recently soybean biomaterials (SB) have been developed which contain significant levels of isoflavones and have shown bone regeneration potential in vivo.

The work undertaken in this thesis aims to study the effects of soy extract on the differentiation of MSCs to an osteogenic phenotype and to test in vitro the potential of novel soybean based biomaterials to differentiate MSCs. These aims were studied by A) identifying and characterising a differentiating primary MSC model, B) developing multiple differentiation markers, C) characterising the effect of pure genistein supplementation on MSCs, D) observing differentiation after addition of extracted crude soy extract, E) identifying the mechanism by which soy extract initiates osteoblast differentiation, F) analysing the biomaterials with and without MSCs and G) analysing the differentiation of MSCs when grown on biomaterials.

The results described in this thesis show that both pure genistein and crude soy extract are osteogenic. This osteogenic effect can be modulated dose dependently and by addition of a competitive inhibitor that specifically interrupts isoflavone binding to oestrogen receptors. The assessment of the in vitro effects on cell types, including an osteosarcoma cell line and several primary MSC strains, indicates that it is important to choose the cell type carefully when undertaking work in differentiation models. This work demonstrates that isoflavones can dose dependently modulate MSC differentiation, proliferation and cell death. Using the characterised differentiating MSC model the soybean biomaterial was shown to be significantly osteogenic.

To conclude, this thesis demonstrates that soy extracted isoflavones and soy extract containing biomaterials induce the differentiation of a characterised MSC model into an osteoblast phenotype and that this class of novel biomaterials has a potential use in regenerative medicine. Treatments based on the use of soy extract biomaterials or soy extract biomaterial derived tissue engineering constructs may lead to enhanced clinical outcomes in maxillofacial, orthopaedic and periodontal surgery.

Table of contents

ABSTRACT i

TABLE OF CONTENTS ii

LIST OF TABLES xiii

LIST OF FIGURES xv

ACKNOWLEDGEMENTS xix

AUTHORS DECLARATION xx

ABBREVIATIONS xxi

CHAPTER 1 GENERAL INTRODUCTION 1.1 Motivation 1 1.2 Bone formation 1 1.3 Bone mineral composition 2 1.4 Bone types 3 1.4.1 Cortical bone 3 1.4.2 Trabecular bone 3 1.4.3 Bone marrow 4 1.4.3.1 Bone marrow stroma 4 1.5 Bone maintenance 5 1.5.1 Bone turnover 5 1.5.2 Bone digestion 6 1.5.2.1 Control of bone resorption 6 1.6 Genetic regulation of bone 7 1.6.1 BMP 9 1.6.2 Cbfa1 10 1.6.3 Alkaline phosphatase 11 1.7 Isolation of marrow stromal cell osteoprogenitors 12 1.7.1 Culture of BMSC 14

1.7.2 in vitro differentiation 14 1.8 BMSC in vivo 15 1.8.1 BMSC transplants 16 1.9 Bone replacements 16 1.10 Soy phytoestrogens 19 1.11 Measurements of isoflavone concentrations in vivo 23 1.12 Hypothesis 25 1.13 The aim of the work 25

CHAPTER 2 MATERIALS AND METHODS 2.1 Materials 26 2.1.1 Equipment for extraction of BMSCs 26 2.1.2 Extraction consumables 26 2.1.3 Cell culture plastic 26 2.1.4 Cell culture consumables 27 2.1.5 Wash buffers 27 2.1.6 Isolation and culturing medium 27 2.1.7 Osteogenic medium 28 2.1.8 Adipogenic medium 28 2.1.9 Chondrogenic medium 28 2.1.10 Experimental media 29 2.1.10.1 Growth medium 29 2.1.10.2 -glycerophosphate medium 29 2.1.10.3 Dex medium 30 2.1.10.4 Osteogenic medium 30 2.1.11 Staining solutions 30 2.1.11.1 Alizarin red stain, for calcium anions 30 2.1.11.2 Von Kossa stain. 31 2.1.11.3 Oil red O, lipid stain. 31 2.1.11.4 Cartilage stain, for the assessment of proteoglycans 31 2.1.11.5 Collagen staining 31 2.1.11.6 HPI stain for membrane integrity 31

iii

2.1.12 Cell lysate buffer 32 2.1.13 Solutions for determination of ALP activity 32 2.1.14 Total protein determination by Bradford reagent 32 2.1.15 LDH cytotoxicity testing 32 2.1.16 Immunohistology 32 2.1.17 Immunohistology antibodies 33 2.1.18 Antibody diluent 33 2.1.19 Cryostat sectioning 33 2.2 Methods 34 2.2.1 BMSC Extraction protocol 34 2.2.2 Cell maintenance 35 2.2.3 Cell passaging 35 2.2.4 Cell freezing 36 2.2.5 Cell thawing 36 2.2.6 Quantification of cell number by haemocytometer 37 2.2.7 Growth curves 37 2.2.8 Cambrex BMSC maintenance protocols 37 2.2.9 Characterisation of putative BMSCs 38 2.2.9.1 Immunohistochemistry 38 2.2.9.2 Single cell cloning 38 2.2.9.3 Differentiation 39 2.2.9.3.1 Attached monolayer cultured differentiation 39 2.2.9.3.2 Pellet culture for the differentiation on chondrocytes 39 2.2.9.3.3 Qualitative assessment of mineralisation by Von Kossa 39 2.2.9.3.4 Qualitative assessment of adipocytes by Oil red O staining 40 2.2.9.3.5 Qualitative assessment of chondrocyte differentiation 40 2.2.9.3.5.1 Preparation of chondrocyte pellets for cryostat sectioning 40 2.2.9.3.5.2 Cryostat sectioning and staining of pellet cultures 41 2.2.9.3.5.3 Qualitative assessment of chondrocyte differentiation by cartilage 41 staining 2.2.10 Quantitative analysis of experimental cultures 41 2.2.10.1 Quantification of mineralisation, osteoblast differentiation staining 41

2.2.10.2 Quantitative analysis of osteoblast differentiation by ALP 42 2.2.10.3 Determination of total protein content by Bradford reagent 43 2.2.10.4 Quantification of collagen by Sirius staining 43 2.2.10.5 Quantification of cell proliferation and death 44 2.2.10.5.1 LDH determination 44 2.2.10.5.1.1 LDH standard curve 44 2.2.10.5.1.2 Quantification of cell number by LDH assay 44 2.2.10.5.1.3 Relative quantification of cell death by LDH release 45 2.2.10.5.2 Measurement of proliferation by Brd-U incorporation 46 2.2.10.5.3 Quantification of proliferation by the uptake of radio labelled thymine 46 2.2.10.5.4 Live and dead staining with Hoechst Propidium iodide 47 2.2.10.6 Quantification of levels of intracellular Adenine triphosphate 47 2.2.10.7 Molecular biological techniques for the determination of mRNA 47 quantities 2.2.10.7.1 Total RNA extraction from monolayers 47 2.2.10.7.2 Quantification of total RNA 48 2.2.10.7.3 Production of cDNA from total RNA 48 2.2.10.7.4 qPCR primer design 49 2.2.10.7.5 Experimental design for qPCR 50 2.2.10.7.6 Analysis of qPCR results 51

2.2.10.7.7 The Ct analysis method 51 2.2.11 Extraction and characterisation of isoflavones from soy 52 2.2.11.1 The extraction of isoflavones from soy flour 52 2.2.11.2 The determination of the soy extract composition by HPLC 52 2.2.12 Production of biomaterials 53 2.2.12.1 Production of calcium phosphate cement containing soy extract 53 2.2.12.2 Production of calcium phosphate cement containing gelatine 53 2.2.12.3 Components required for production of the gelatine and soy cement 54 biomaterial 2.2.13 Experimental method for the application of isoflavones to cell 54 cultures 2.2.13.1 Initial seeding densities and media 54 2.2.13.2 Experimental setup for analysis of the effect of pure genistein 55

2.2.13.3 Experimental setup for the culture of soy extract with CBMSCs 55 2.2.13.4 Experimental seeding densities for production of a confluent 55 monolayer 2.2.13.5 Quantity of genistein in the soy extract 56 2.2.14 Experimental method for the application of BMSCs to cement 56 biomaterials 2.2.15 Preparation and visualisation of biomaterials by SEM 56

CHAPTER 3. THE DEVELOPMENT AND CHARACTERISATION OF A CONFLUENT DIFFERENTIATING IN VITRO MODEL FOR THE ANALYSIS OF OSTEOGENESIS DURING CULTURE WITH GENISTEIN. 3.1 Introduction 57 3.2 Hypothesis 59 3.2.1 Aims. 59 3.3 Methods. 59 3.3.1 Experimental procedure. 59 3.3.2 Statistics 60 3.4 Results 60 3.4.1 The effect of genistein on MG63 cell number as derived by analysis 60 of lactate dehydrogenase (LDH) content after culture in media for 2, 4 and 6 days. 3.4.1.1 The LDH derived cell number after culture in growth media and 61 genistein. 3.4.1.2 The LDH derived cell number after culture in beta-glycerophosphate 62 media and genistein. 3.4.1.3 The LDH derived cell number after culture in dexamethasone media. 63 3.4.1.4 The LDH derived cell number after culture in osteogenic media. 64 3.4.1.5 Summary of LDH derived cell number found after treatment of 65 MG63s with genistein. 3.4.2 Relative cell death measured by release of supernatant LDH. 65 3.4.2.1 Relative LDH release after culture in growth media with genistein. 66 3.4.2.2 Relative LDH release after culture in beta media with genistein. 67 3.4.2.3 Relative LDH release after culture in dex media with genistein. 68 3.4.2.4 Relative LDH release after culture in osteo media with genistein. 69 3.4.2.5 Summary of the relative LDH release after culture with genistein. 70

3.4.3 Measurements of ALP activity after treatment with genistein. 70 3.4.3.1 ALP activity after culture with genistein in growth media. 71 3.4.3.2 ALP activity after culture with genistein in beta media. 72 3.4.3.3 ALP activity after culture with genistein in dex media 73 3.4.3.4 ALP activity after culture with genistein in osteogenic media. 74 3.4.3.5 Summary of the activation of ALP with the addition of genistein. 74 3.4.4 Measurements of total collagen production after treatment with 75 genistein. 3.4.4.1 Collagen production after culture with genistein in growth media. 75 3.4.4.2 Collagen production after culture in beta media with genistein. 76 3.4.4.3 Collagen production after culture with genistein in dex media. 77 3.4.4.4 Collagen production after culture with genistein in osteogenic media. 78 3.4.4.5 Summary of collagen production after culture in media containing 79 genistein. 3.4.5 Alizarin Red staining for quantification of mineralisation. 79 3.4.5.1 Alizarin red staining after culture with genistein in growth media. 80 3.4.5.2 Alizarin red staining after culture with genistein in beta media. 81 3.4.5.3 Alizarin red staining after culture with genistein in dex media. 82 3.4.5.4 Alizarin red staining after culture with genistein in osteogenic media. 83 3.4.5.5 Summary of the effect of genistein on alizarin red staining. 83 3.5 Results summary. 84 3.5.1 Cell death. 84 3.5.2 ALP activity 84 3.5.3 Collagen synthesis. 85 3.5.4 Mineralisation. 85 3.6 Discussion. 85 3.7 Conclusions 91

CHAPTER 4. AN IN VITRO EVALUATION OF THE OSTEOGENIC EFFECT OF GENISTEIN ON A CONFLUENT DIFFERENTIATING BMSC MODEL. 4.1 Introduction 92 4.2 Hypothesis 98 4.2.1 Aims 99

vii

4.3 Methods 99 4.3.1 Experimental procedure 99 4.3.2 Statistical analysis. 100 4.4 Results 100 4.4.1 Characterisation of BMSC cultures. 100 4.4.2 BMSC Growth curves in isolation and culturing medium. 100 4.4.3 Antibody staining for the characterisation of surface epitope 101 phenotype. 4.4.4 Characterisation of the multipotentiality of isolated populations. 102 4.4.4.1 Characterisation of the BMSC culture for osteogenic potential 102 4.4.4.2. Characterisation of the BMSC culture for adipogenic potential 103 4.4.4.3 Characterisation of the BMSC culture for chondrogenic potential. 103 4.4.5. The differentiation potential of BMSCs with time in culture and the 104 analysis of the culture potential for multipotent single cells. 4.4.6 Temporal characterisation of differentiation markers 106 4.4.7. Summary of BMSC cultures. 106 4.4.8 Osteogenic effect of genistein after culture with primary human 107 BMSCs in several medium types for 2, 4 and 6 days. 4.4.8.1. Genistein effect on total cell number in the monolayer 107 4.4.8.1.1 Cell number after culture with genistein in growth media. 107 4.4.8.1.2 Cell number after culture with genistein in beta media. 108 4.4.8.1.3 Cell number after culture with genistein in dex media. 109 4.4.8.1.1 Cell number after culture with genistein in osteogenic media. 110 4.4.8.2 The effect of genistein on LDH release compared with total LDH 111 4.4.8.2.1 LDH release after culture with genistein in growth media. 111 4.4.8.2.2 LDH release after culture with genistein in beta media. 112 4.4.8.2.3 LDH release after culture with genistein in dex media. 113 4.4.8.2.2 LDH release after culture with genistein in osteogenic media. 114 4.4.8.3 ALP activity after addition of genistein to BMSC cultures. 115 4.4.8.3.1 ALP activity after culture with genistein in growth media. 115 4.4.8.3.2 ALP activity after culture with genistein in beta media. 116 4.4.8.3.2 ALP activity after culture with genistein in dex media. 117 4.4.8.3.3 ALP activity after culture with genistein in osteogenic media. 118

viii

4.4.8.3.4 Summary of the effect of genistein on ALP activity. 119 4.4.8.4 Total collagen observed after the addition of genistein. 119 4.4.8.4.1 The collagen observed after culture with genistein in growth media. 119 4.4.8.4.2 Quantity of collagen after culture with genistein in beta media. 120 4.4.8.4.3 Quantity of collagen after culture with genistein in dex media. 121 4.4.8.4.4 Quantity of collagen after culture with genistein in osteogenic media. 122 4.4.8.5 Alizarin Red staining for the quantification of mineralisation. 123 4.4.8.5.1 Alizarin red staining after culture with genistein in growth media 123 4.4.8.5.2 Alizarin red staining after culture with genistein in beta media 124 4.4.8.5.3 Alizarin red staining after culture with genistein in dex media 125 4.4.8.5.4 Alizarin red staining after culture with genistein in osteogenic media 126 4.5. Discussion 127 4.5.1 Characterisation. 127 4.5.2 Genistein mediated differentiation 131 4.6 Conclusions 135

CHAPTER 5. THE USE OF CULTURED CBMSCS IN A CHARACTERISED CONFLUENT DIFFERENTIATING MODEL OF OSTEOGENESIS TO DETERMINE THE IN VITRO EFFECT OF GENISTEIN IN SEVERAL MEDIA. 5.1 Introduction 136 5.2 Hypothesis 140 5.2.1 Aim 140 5.3 Experimental procedure 140 5.4 Results 140

5.4.1 T0 samples and confirmation of cell attachment. 140 5.4.2 LDH measurement of cell number. 140 5.4.2.1 LDH derived cell number after culture in growth media. 141 5.4.2.2 LDH derived cell number after culture in β-glycerophosphate media. 142 5.4.2.3 LDH derived cell number after culture in dex media. 143 5.4.2.4 LDH derived cell number after culture in osteogenic media. 144 5.4.2.5 LDH derived cell number summary. 144 5.4.3 Summary of cell counts and live/dead measurements. 145 5.4.4 Summary of Brd-U Incorporation into replicating DNA. 146

5.4.5 LDH in the supernatant as a measurement of cell death. 147 5.4.5.1 Supernatant LDH after culture in growth media. 147 5.4.5.2 Supernatant LDH after culture in beta-glycerophosphate media. 148 5.4.5.3 Supernatant LDH after culture in dex media. 149 5.4.5.4 Supernatant LDH after culture in osteogenic media. 150 5.4.5.5 Summary of relative LDH release. 150 5.4.6 Measurements of ATP availability. 151 5.4.6.1 ATP after culture with genistein in growth media 151 5.4.6.2 ATP after culture with genistein in beta-glycerophosphate media. 152 5.4.6.3 ATP after culture with genistein in dex media 153 5.4.6.4 ATP after culture with genistein in osteogenic media. 154 5.4.6.5 Summary of ATP availability in all media. 154 5.4.7 ALP activity as an early measurement of differentiation 155 5.4.7.1 ALP activity after genistein treatment in growth media 155 5.4.7.2 ALP activity after genistein treatment in beta-glycerophosphate 156 media 5.4.7.3 ALP activity after genistein treatment in dex media 157 5.4.7.4 ALP activity after genistein treatment in osteogenic media 158 5.4.7.5 Summary of ALP activity after genistein treatment. 159 5.4.8 Collagen production after culture with genistein containing media. 159 5.4.8.1 Collagen production after culture with genistein in growth media. 159 5.4.8.2 Collagen after culture with genistein in beta-glycerophosphate media. 160 5.4.8.3 Collagen production after culture with genistein in dex media. 161 5.4.8.4 Collagen production after culture with genistein in osteogenic media. 162 5.4.8.5 Summary of collagen production after culture with genistein. 162 5.4.9 Quantitative alizarin red staining of mineralisation. 163 5.4.9.1 ARS after culture with genistein in growth media. 163 5.4.9.2 ARS after culture with genistein in beta-glycerophosphate media 164 5.4.9.3 ARS after culture with genistein in dex media 165 5.4.9.4 ARS after culture with genistein in osteogenic media 166 5.4.9.5. ARS summary. 166 5.5 Discussion 167

5.5.1 Proliferation and cell death 167 5.5.2 Differentiation 172 5.6 Conclusions. 174

CHAPTER 6. THE OSTEOGENIC EFFECT OF SOY EXTRACT ON THE CBMSC MODEL. 6.1 Introduction. 175 6.2 Hypothesis 178 6.3 Methods 179 6.4 Results 179 6.4.1 HPLC quantification of soy extract isoflavone content. 179 6.4.2 Direct cell counts. 180 6.4.2.1 Live cell counts of CBMSCs cultured in growth media and soy 180 extract. 6.4.2.2 Live cell counts of CBMSCs cultured in growth media with inhibitor 181 and soy extract. 6.4.2.3 Live cell counts of CBMSCs cultured in osteogenic media and soy 182 extract. 6.4.2.4 Live cell counts of CBMSCs cultured in osteogenic media with 183 inhibitor and soy extract. 6.4.2.5 Summary of soy extract and inhibitor effect on cell number. 183 6.4.3 Inverted photomicroscopy of CBMSC cultures. 184 6.4.3.1 Control CBMSC cultures. 184 6.4.3.2 HPI stained CBMSCs after culture in osteogenic media and soy 184 extract. 6.4.4 Quantification of the activity of ALP. 185 6.4.4.1 ALP activity after culture in growth media and soy extract. 185 6.4.4.2 ALP activity after culture in growth media with inhibitor and soy 186 extract. 6.4.4.3 ALP activity after culture in beta media and soy extract. 187 6.4.4.4 ALP activity after culture in beta media with inhibitor and soy 188 extract. 6.4.4.5 ALP activity after culture in osteogenic media and soy extract. 189 6.4.4.6 ALP activity after culture in osteogenic media with inhibitor and soy 190 extract. 6.4.4.7 Summary of ALP activity after addition of soy extract. 190

6.4.5 Analysis of terminal differentiation by cbfa-1 expression. 191 6.4.5.1 The effect of soy extract on the expression of cbfa-1 in CBMSCs 191 cultured in growth media and growth media with inhibitor. 6.4.5.2 The effect of soy extract on the expression of cbfa-1 in CBMSCs 192 cultured in osteogenic media and osteogenic media with inhibitor 6.4.5.3 Summary of the effect of soy extract on cbfa-1 expression. 192 6.4.6 Alizarin red staining (ARS) of mineralised ECM 193 6.4.6.1 ARS of monolayers after culture with soy extract in growth media. 193 6.4.6.2 ARS of monolayers after culture with soy extract in growth media 194 and inhibitor. 6.4.6.3 ARS of monolayers after culture with soy extract in osteogenic 195 media. 6.4.6.4 ARS of monolayers after culture with soy extract in osteogenic media 196 and inhibitor. 6.4.6.5 Summary of the effect of soy extract on mineralisation. 196 6.4.7 Summary of results. 196 6.5 Discussion 197 6.5.1 Comparison with literature. 197 6.5.2 Comparison with previous chapters 199 6.6 Conclusions. 200

CHAPTER 7. DIFFERENTIATION OF CBMSCS ON CALCIUM PHOSPHATE CEMENTS. 7.1 Introduction. 201 7.2 Hypothesis 205 7.3 Methods 205 7.4 Results 206 7.4.1 Visual comparison of cements. 206 7.4.1.1 Photographs of set cements. 206 7.4.1.2. SEM of set cements. 207 7.4.1 Proliferation of CBMSCs on cements. 208 7.4.2 Expression of cbfa1 after culture of CBMSCs on cements. 209 7.4.3 SEM of cements with CMBSCs. 210 7.4.2 Soy/gelatine cement remodelling. 211 7.4.3 Gelatine cements after culture with CBMSCs 212

xii

7.5 Discussion. 213 7.6 Conclusions 216

CHAPTER 8. CONCLUSIONS AND FUTURE WORK 8.1 Scientific background of the work 217 8.2 The hypothesis of this work 218 8.3 Findings of the work. 219 8.3.1 MG63 cell line observations 219 8.3.2 Primary human BMSCs 219 8.4 Novel contribution 220 8.5 Future work 221

REFERENCES 223

APPENDICES Appendix A 270 Appendix B 274

List of tables

CHAPTER 2 MATERIALS AND METHODS Table 2.1.1 Equipment for extraction of BMSCs 26 Table 2.1.2 Extraction consumables 26 Table 2.1.3 Cell culture plastic 26 Table 2.1.4 Cell culture consumables 27 Table 2.1.5 Wash buffers 27 Table 2.1.6 Isolation and culturing medium 27 Table 2.1.7 Osteogenic medium 27 Table 2.1.8 Adipogenic medium 27 Table 2.1.9 Chondrogenic medium 28 Table 2.1.10.1 Growth medium 28 Table 2.1.10.2 -glycerophosphate medium 29 Table 2.1.10.3 Dex medium 29 Table 2.1.10.4 Osteogenic medium 29

xiii

Table 2.1.11.1 Alizarin red stain, for calcium anions 30 Table 2.1.11.2 Von Kossa stain. Stains the carbonate and phosphate anions in 30 mineralised tissue Table 2.1.11.3 Oil red O stain for the assessment of adipocyte differentiation 30 Table 2.1.11.4 Cartilage stain, for the assessment proteoglycans in chondrocyte 30 differentiation Table 2.1.11.5 Collagen staining 30 Table 2.1.11.6 HPI stain for membrane integrity 31 Table 2.1.12 Cell lysate buffer 31 Table 2.1.13 Solutions for determination of ALP activity 31 Table 2.1.14 Total protein determination by Bradford reagent 31 Table 2.1.15 LDH cytotoxicity testing 31 Table 2.1.16 Immunohistology 32 Table 2.1.17 Immunohistology antibodies 32 Table 2.1.18 Antibody diluent 32 Table 2.1.19 Cryostat sectioning 33 Table 2.2.10.7.3a Production of cDNA from total RNA; reaction mix 48 Table 2.2.10.7.3b Production of cDNA from total RNA; master mix 49 Table 2.2.10.7.5a Experimental design for qPCR; master mix 50 Table 2.2.10.7.5b Experimental design for qPCR; run program 51 Table 2.2.12.3 Components required for production of the gelatine and soy 54 cements. Table 2.2.13.4 Experimental seeding densities for production of a confluent 55 monolayer. Table 2.2.13.5 Quantity of genistein on the soy extract 56

CHAPTER 4. AN IN VITRO EVALUATION OF THE OSTEOGENIC EFFECT OF GENISTEIN ON A CONFLUENT DIFFERENTIATING BMSC MODEL Table 4.1. The differentiation potential of isolates 105 Table 4.2. Analysis of multipotent single cell clones 105

Table of Figures.

CHAPTER 1 GENERAL INTRODUCTION Fig 1.1 Factors involved in the adhesion and condensation of mesenchymal 7 cells.

xiv

Fig 1.2 regulation and function of BMPs in bone growth 8 Fig 1.3 Regulation of mineralisation in the cartilage template. 9 Fig 1.4 Drawing of generic isoflavone and the main modifications 20

CHAPTER 3. THE DEVELOPMENT AND CHARACTERISATION OF A CONFLUENT DIFFERENTIATING IN VITRO MODEL FOR THE ANALYSIS OF OSTEOGENESIS DURING CULTURE WITH GENISTEIN. Fig 3.1 LDH derived cell number after culture in growth media and genistein. 61 Fig 3.2 LDH derived cell number after culture in beta media and genistein. 62 Fig 3.3 LDH derived cell number after culture in dex media and genistein. 63 Fig 3.4 LDH derived cell number after culture in osteo media and genistein. 64 Fig 3.5 The relative LDH release after culture in growth media with genistein 66 Fig 3.6 The relative LDH release after culture in beta media with genistein 67 Fig 3.7 The relative LDH release after culture in dex media with genistein 68 Fig 3.8 The relative LDH release after culture in osteo media with genistein 69 Fig 3.9 ALP activity after culture with genistein in growth media 71 Fig 3.10 ALP activity after culture with genistein in beta media 72 Fig 3.11 ALP activity after culture with genistein in dex media 73 Fig 3.12 ALP activity after culture with genistein in osteo media 74 Fig 3.13 Collagen production after culture with genistein in growth media 75 Fig 3.14 Collagen production after culture with genistein in beta media 76 Fig 3.15 Collagen production after culture with genistein in dex media 77 Fig 3.16 Collagen production after culture with genistein in osteo media 78 Fig 3.17 Alizarin red staining after culture with genistein in growth media 80 Fig 3.18 Alizarin red staining after culture with genistein in beta media 81 Fig 3.19 Alizarin red staining after culture with genistein in dex media 82 Fig 3.20 Alizarin red staining after culture with genistein in osteo media 83

CHAPTER 4. AN IN VITRO EVALUATION OF THE OSTEOGENIC EFFECT OF GENISTEIN ON A CONFLUENT DIFFERENTIATING BMSC MODEL. Fig 4.1 BMSC growth curves in isolation and culturing medium 100 Fig 4.2 Antibody staining for the characterisation of surface epitopes 101 Fig 4.3 Alizarin red stained bulk cultures 102 Fig 4.4 Oil red O stained bulk cultures 103

Fig 4.5 Alcian blue stained bulk cultures 103 Fig 4.6 Temporal characterisation of differentiation markers 106 Fig 4.7 Cell number after culture with genistein in growth media 107 Fig 4.8 Cell number after culture with genistein in beta media 108 Fig 4.9 Cell number after culture with genistein in dex media 109 Fig 4.10 Cell number after culture with genistein in osteo media 110 Fig 4.11 LDH release after culture with genistein in growth media 111 Fig 4.12 LDH release after culture with genistein in beta media 112 Fig 4.13 LDH release after culture with genistein in dex media 113 Fig 4.14 LDH release after culture with genistein in osteo media 114 Fig 4.15 ALP activity after culture with genistein in growth media 115 Fig 4.16 ALP activity after culture with genistein in beta media 116 Fig 4.17 ALP activity after culture with genistein in dex media 117 Fig 4.18 ALP activity after culture with genistein in osteo media 118 Fig 4.19 collagen observed after culture with genistein in growth media 119 Fig 4.20 collagen observed after culture with genistein in beta media 120 Fig 4.21 collagen observed after culture with genistein in dex media 121 Fig 4.22 collagen observed after culture with genistein in osteo media 122 Fig 4.23 Alizarin red staining after culture with genistein in growth media 123 Fig 4.24 Alizarin red staining after culture with genistein in beta media 124 Fig 4.25 Alizarin red staining after culture with genistein in dex media 125 Fig 4.26 Alizarin red staining after culture with genistein in osteo media 126

CHAPTER 5. THE USE OF CULTURED CBMSCS IN A CHARACTERISED CONFLUENT DIFFERENTIATING MODEL OF OSTEOGENESIS TO DETERMINE THE IN VITRO EFFECT OF GENISTEIN IN SEVERAL MEDIA. Fig 5.1 LDH derived cell number after culture in growth media 141 Fig 5.2 LDH derived cell number after culture in beta media 142 Fig 5.3 LDH derived cell number after culture in dex media 143 Fig 5.4 LDH derived cell number after culture in osteo media 144 Fig 5.5 Summary of cell counts and live/dead measurements. 145 Fig 5.6 Summary of Brd-U Incorporation into replicating DNA 146 Fig 5.7 LDH release after culture in growth media 147 Fig 5.8 LDH release after culture in beta media 148

xvi

Fig 5.9 LDH release after culture in dex media 149 Fig 5.10 LDH release after culture in osteo media 150 Fig 5.11 ATP after culture with genistein in growth media 151 Fig 5.12 ATP after culture with genistein in beta media 152 Fig 5.13 ATP after culture with genistein in dex media 153 Fig 5.14 ATP after culture with genistein in osteo media 154 Fig 5.15 ALP activity after genistein treatment in growth media 155 Fig 5.16 ALP activity after genistein treatment in beta media 156 Fig 5.17 ALP activity after genistein treatment in dex media 157 Fig 5.18 ALP activity after genistein treatment in osteo media 158 Fig 5.19 Collagen production after culture with genistein in growth media 159 Fig 5.20 Collagen production after culture with genistein in beta media 160 Fig 5.21 Collagen production after culture with genistein in dex media 161 Fig 5.22 Collagen production after culture with genistein in osteo media 162 Fig 5.23 ARS after culture with genistein in growth media 163 Fig 5.24 ARS after culture with genistein in beta media 164 Fig 5.25 ARS after culture with genistein in dex media 165 Fig 5.26 ARS after culture with genistein in osteo media 166

CHAPTER 6. THE OSTEOGENIC EFFECT OF SOY EXTRACT ON THE CBMSC MODEL Fig 6.1 Representative HPLC trace of the soy extract. 179 Fig 6.2 Live cell counts of CBMSCs cultured in growth media and soy extract 180 Fig 6.3 Live cell counts of CBMSCs cultured in growth media with inhibitor 181 and soy extract Fig 6.4 Live cell counts of CBMSCs cultured in osteo media and soy extract 182 Fig 6.5 Live cell counts of CBMSCs cultured in osteo media with inhibitor and 183 soy extract Fig 6.6 Images of confluent controls 184 Fig 6.7 HPI stained cultures 184 Fig 6.8 ALP activity after culture in growth media and soy extract 185 Fig 6.9 ALP activity after culture in growth media with inhibitor and soy extract 186 Fig 6.10 ALP activity after culture in beta media and soy extract 187 Fig 6.11 ALP activity after culture in beta media with inhibitor and soy extract 188

xvii

Fig 6.12 ALP activity after culture in osteo media and soy extract 189 Fig 6.13 ALP activity after culture in osteo media with inhibitor and soy extract 190 Fig 6.14 cbfa1 expression after culture with soy extract in growth media and 191 growth media with inhibitor. Fig 6.15 cbfa1 expression after culture with soy extract in osteo media and osteo 192 media with inhibitor Fig 6.16 ARS after culture with soy extract in growth media 193 Fig 6.17 ARS after culture with soy extract in growth media and inhibitor 194 Fig 6.18 ARS after culture with soy extract in osteo media 195 Fig 6.19 ARS after culture with soy extract in osteo media and inhibitor 196

CHAPTER 7. DIFFERENTIATION OF CBMSCS ON CALCIUM PHOSPHATE CEMENTS Fig 7.1 Photographs of set cements 206 Fig 7.2 SEM of set cements 207 Fig 7.3 LDH derived cell number after culture on cements 208 Fig 7.4 cbfa1 expression after culture on cements 209 Fig 7.5 SEM of cements with cells attached 210 Fig 7.6 SEM of soy cements before and after culture 211 Fig 7.7 SEM of gelatine cements before and after culture 212

xviii

Acknowledgements

I would like to acknowledge the support of my supervisors and the help and friendship of colleagues at the Blond McIndoe Research Foundation and University of Brighton.

This work was generously funded by The Guinea Pig Club and the Blond McIndoe

Research Foundation, to whom I am indebted.

I am especially grateful to Juliet for her support and encouragement during the last three years. Finally I wish to thank my family for their enthusiasm and support through it all.

xix

Declaration

I declare that the research contained in this thesis, unless otherwise formally indicated within the text, is the original work of the author. The thesis has not been previously submitted to this or any other university for a degree, and does not incorporate any material already submitted for a degree.

Matthew Illsley April 2011

Abbreviations

αMEM Alpha minimal essential medium ISCT International society for cellular therapy ALP Alkaline Phosphatase LDH Lactate dehydrogenase AP1 activator protein 1 MAPC multipotent adult progenitor cell ARS Alizarin Red Stain MAPK mitogen activated protein kinase ATF4 Activating transcription factor 4 MG63, Cell lines MCF-2 ATP Adenosine-5'-triphosphate MMPI matrix metalloproteinase I Beta β-glycerophosphate MSC Variously, mesenchymal stem cell, marrow stromal cell and combinations. BMD Bone mineral density MSX1 muscle segment homeobox 1 BMP Bone morphogenic protein NCAM Neural Cell Adhesion Molecule BMSC Bone Marrow Stromal cell NOD/ nonobese diabetes/ severe SCID combined immunodeficiency BrdU Bromodeoxyuridine (5-bromo-2'- OC osteocalcin deoxyuridine) BSP Bone sialoprotein OP osteopontin cAMP Cyclic adenosine monophosphate ON osteonectin Cbfa1 Core binding factor alpha OPG osteoprotegerin CBMSC Cambrex bone marrow stromal Osteo osteogenic medium cells CDHA Calcium deficient hydroxyapatite Osx osterix cFKH1 Chicken winged helix Pax1/2/ Paired box proteins transcription factor 9 CNS Central nervous system PB- peripheral blood mononuclear BMSC stem cell or other MSC Coll 1α1 Collagen, type I, alpha 1 PDGH platelet derived growth factor DCPA dibasic calcium phosphate PMMA Poly(methyl methacrylate) anhydrate DCPD dibasic calcium phosphate PPARγ Peroxisome proliferator-activated dehydrate receptor gamma Dex Dexamethasone PPi inorganic pyrophosphate DMEM Dulbecco's modified Eagle's Prx1 Paired-related homeo box gene medium EGF Epidermal growth factor Prx2 Paired-related homeo box gene 2 EMEM Eagle's minimal essential medium PS Phosphatidylserine

xxi

ER Oestrogen receptor PSR Phosphatidylserine receptor ERE Oestrogen response element PTH Parathyroid hormone ERK Extracellular signal regulated RANK/ Receptor activator of nuclear kinase RANK factor kappa-B / ligand L ERR Oestrogen related receptor Runx2 Is cbfa1 FABP4 fatty acid binding protein 4 SAOS2 Cell line FACs Fluorescence-activated cell SERM selective oestrogen receptor sorting modulator FCS Foetal calf serum Shh sonic hedgehog FGF Fibroblast growth factor SMAD transcription factor G-292 Cell line Smurf1 E3 ubiquitin-protein ligase GM growth medium SP1 Specificity Protein 1 HA Hydroxyapatite TCF T cell factor Hoxa2 homeobox protein a2 TGF Transforming growth factor HRT Hormone replacement therapy TTCP tetracalcium phosphate HSC hematopoietic stem cells U2OS Cell line IGF insulin like growth factor WNT Wingless-type MMTV integration site family

xxii

Chapter 1

General Introduction

1.1 Motivation

Autograft is the gold standard bone reconstruction material; often bone is removed from the patient’s iliac crest, shaped and used to reconstruct defects (Dusseldorp and Mobbs

2009). However this technique is limited in the availability of donor sites and donor site morbidity (Giannoudis et al 2005). Therefore a current focus of biomaterials research is adjunctive and replacement materials for the reconstruction of defect and donor sites

(Brydone et al. 2010). One avenue of biomaterials research is the use of calcium phosphate cements to fill defects or join gaps between the graft and wound (Ambard and

Mueninghoff 2006). An improvement in the efficacy of calcium phosphate cements has been observed with the addition of osteogenic agents (Heliotis et al. 2006). The naturally occurring isoflavones in Soy are osteogenic and therefore their use in calcium phosphate cements is postulated to accelerate the differentiation of native cells to mineralising osteoblasts (Santin and Ambrosio 2008). Currently there is no in vitro evidence of increasing osteogenic differentiation in primary cells when cultured with soy extracts or calcium phosphate cements containing soy extract. Therefore an understanding of bone biology is required for the isolation of primary osteogenic cells and analysis of soy extract induced osteogenesis.

1.2 Bone formation.

There are two non-pathogenic bone formation mechanisms, intramembranous and endochondral ossification. Intramembranous bones are often flattened or twisting with complex structures; the bones comprising the cranial vault and the scapula are formed in this manner (Gray 1918). In intramembranous ossification a signalling point develops 1

between two membranes that attracts mesenchymal cells, these mesenchymal cells directly differentiate into osteoblasts that subsequently mineralise the area between the two membranes, hence the availability of complex shapes (Alborzi et al. 1996). As well as de novo bone synthesis intramembranous ossification also occurs at the growth plate when injury occurs to endochondral bones during childhood (Xian et al 2004).

Endochondral ossification has been studied extensively and requires the formation of a collagen template; the chondrocytes that produce the template are under strict regulation from a myriad of transcription factors and diffusion gradients. Hypertrophic chondrocytes differentiate into osteoblasts that subsequently mineralise the collagen scaffold (Pazzaglia et al. 2011). During mineralisation several types of bone are formed.

1.3 Bone mineral composition.

The inorganic element of the mineralised bone matrix comprises 58% of the dry weight, most of it is calcium phosphate almost all of which is calcium hydroxyapatite,

Ca10(PO4)6(OH)2. The remainder of the inorganic elements are 7% calcium carbonate,

1% to 2% of calcium fluoride, magnesium phosphate and sodium chloride; the remaining

(sodium, potassium and strontium) trace minerals less than 1%. 85% of the body’s total phosphate is held in bone, with 15% in cells and less than 1% in the blood (Gray 1918,

Grabowski 2009). Bone is the most important organ involved in calcium homeostasis,

0.1% of available calcium is in the blood, about 1% in cells and the remainder in bone.

Between 0.4% and 1% of total bone calcium is readily exchangeable so that the skeleton can support the maintenance of blood pH by deposition or release of as much as 500mg of calcium per day (Gray 1918, Grabowski 2009).

The organic content of bone is 42% of dry weight, about 10% of it is the phosphoproteins, osteocalcin and osteonectin, 2% lipids and glycosaminoglycans, the remaining 88% is type I collagen (Loveridge 1999).

1.4 Bone types

1.4.1 Cortical bone

Cortical, lamina or compact bone comprises the majority of the weight of the organ.

Consisting of organised elements called osteons, cortical bone ensures the excellent transmission of forces away from impact points (Yacoub et al. 2002). The osteon is a process of the antagonistic relationship between the bone dissolving osteoclast and bone creating osteoblast (Javed et al. 2010). The osteons are formed in layers, hence lamina, around the outside of the bone. This arrangement serves to increase bone strength and forms a platform to which tendons and condoyle attach; inside the cortical layer trabecular bone predominates.

1.4.2 Trabecular bone

Trabecular, cancellous or spongy bone occurs on the inside of cortical bone towards the central medullar cavity; forming a three dimensional lattice structure of very high strength to weight ratio (Vanderoost et al. 2011). Because of the relatively large spaces between the bone structures osteons are not formed, instead flat planar lamellae occur.

Due to the high surface area occurring in trabecular bone turnover is much higher than in cortical bone (Holland et al. 2011). Therefore cortical bone provides the majority of the strength while the trabecular bone provides the metabolic activity. Although trabecular bone has a smaller contribution to strength the more elastic trabecular region does provide increased overall mechanical flexibility (Goncalves et al. 2011).

1.4.3 Bone Marrow.

The intramedullary canal is the central bone space and is full of bone marrow (Hak and

Pittman 2010). Marrow is also available in all the channels throughout trabecular and cortical bone, including the osteon Haversian and Volkmann channels. Marrow composition varies dependant on bone type and age (Gurevitch et al. 2007).

In long bones yellow marrow is more prevalent and increases with age; the yellow marrow is comprised of 95% fat and the remaining 5% is cell excaudate, salts and connective tissues. The marrow connective tissue is known as the stroma, mononuclear cells attach to the stroma, in yellow marrow these cells are mostly adipocytes (Gray

1918).

In the articular ends of the long bones, sternum and ribs red marrow prevails; comprised of 75 percent water and the remainder globins, salts, excaudate, and nucleoprotein, the red marrow contains a similar amount of connective tissue to the yellow marrow (Gray

1918). However there are far fewer fat cells in the red marrow and therefore other mononucleate cells are relatively more numerous; red marrow stroma is the haematopoietic stem cell niche and provides all the cells of the blood (Ehninger and

Trumpp, 2011).

1.4.3.1 Bone marrow stroma.

This is the niche of the haematopoietic stem cell (HSC) which is supported by the marrow stroma (Ehninger and Trumpp 2011). There is now a large body of evidence that supports the presence of the multipotent Bone Marrow Stromal Cell (BMSC), which is capable of differentiation to cells of the stroma, osteoblasts, chondrocytes, and adipocytes

(Majumdar et al. 2000, Bianco et al. 2011). Another cell type arises in BMSC culture but has not been isolated directly from an organism, the multipotent adult progenitor cell

(MAPC) is observed in carefully maintained BMSC cultures of between 25 and 30 CPD’s and they are capable of differentiating into both HSC and BMSC cells (Verfaillie 2005).

1.5 Bone maintenance

Constant bone remodelling occurs throughout life, maintained by a continual breakdown and deposition of bone by osteoclasts and osteoblasts respectively (O'Brien 2010). The turnover of bone ensures that trace minerals are made available, fatigue factures do not occur and that density can change in response to mechanical stimuli (Banfi et al. 2010).

Bone turnover is about 10% of the total mass every year; therefore whole turnover occurs at approximately ten year intervals (Banfi et al. 2010). Bones in astronauts lose density because the constant transmission of force is a major factor in bone deposition and maintenance (Mano et al. 2010). There are well defined mechanisms by which this process operates (Matsuo 2009).

1.5.1 Bone turnover.

Bone digestion occurs as the initial step in the formation of an osteon. Multinucleate osteoclasts initially digest a channel through the bone in a roughly cylindrical shape, lining the cut surface above them are osteoblasts which deposit an extracellular matrix which is subsequently mineralised. As the channel becomes thinner, due to the production of successive layers of mineralised matrix, some of the osteoblasts become entombed in lacunae where they become osteocytes, communication and fluid exchange between lacunae occurs via the canaliculi that link lacunae to each other and the central channel. Following the osteoblasts down the central Haversian channel are endothelial cells stimulating angiogenesis and allowing the free exchange of nutrient and waste.

Haversian channels and lamellae are linked together with cross cutting Volkmann canals that often span more than one osteon (Gray 1918, Loveridge 1999).

1.5.2 Bone digestion

Osteoclasts operate at the interface of bone and cell to dissolve bone. They have a tightly sealed skirt at the edge of the cell attached to the strongly alkaline hydroxyapatite surface; into this space lysosomal enzymes and carbonic acid are produced in order to digest both the protein and mineralised extra cellular matrix.

+ The hydroxyapatite dissolving reaction can be described as [Ca3(PO4)2]3Ca(OH)2 +8H

+2 -2  10Ca +6HPO4 +2H2O. The resulting dissolved and degraded material from the osteoclast skirt is transported in vesicles away from the digestion area and released behind the cell (Riches and Ralston 2010).

1.5.2.1 Control of bone resorption.

Bone resorption and deposition is controlled by the interaction of osteoblasts and osteoclasts. Osteoblasts express RANK ligand (RANKL) and osteoprotegrin (OPG). OPG binds to RANKL and blocks it from interacting with RANK. The binding of RANKL to

RANK (facilitated by a decreased amount of OPG) stimulates osteoclast precursors to fuse, forming new osteoclasts which ultimately dissolve bone. Increased OPG or decreased RANKL reduces osteoclast differentiation (Riches and Ralston 2010).

Osteoclasts also influence the differentiation of osteoblasts (Boyce et al. 2009).

These mechanisms can be exploited to treat disease. Inhibition of bone resorption by osteoclast digestion is possible with the addition of bisphosphonates (Matsuo and Irie

2008, Su et al. 2011). Decreasing the differentiation of osteoclasts is possible with the addition of selective oestrogen mimics, TGFb and RANKL antibodies (Turner et al.

1988, Trouvin and Goeb 2010, Cicek et al. 2011). A reduction in the number of osteoclasts can also be achieved by decreasing PTH thereby increasing bone density by reducing osteoblast OPG production (Trivedi et al. 2010).

It is known that the osteocytes in lacunae have a signalling role in bone maintenance and mechanical stimulation prevents their loss (Fonseca et al. 2011). Increased differentiation of osteoblasts has also been shown to occur after mechanical stimulation and independent from osteocyte signalling (Maul et al. 2011).

1.6 Genetic regulation of bone.

Fig 1.1, Factors involved in the adhesion and condensation of mesenchymal cells. Adapted from Hall and Miyake, 2000.

The regulation of bone formation and maintenance is complicated by the multiple functions required from the organ. For such a complicated system there is an array of growth and transcription factors orchestrating gene expression and protein function. The regulation of scaffold condensation and subsequent mineralisation is complex; initial adhesion of mesenchymal cells and initiation of skeletal development is thought to occur by aggregation of cells and enhanced proliferation of the condensate (fig 1.1). The condensation of cells is mediated by adhesion proteins, N-CAM (Neural Cell Adhesion

Molecule), fibronectin, tenascin and syndecan. These are in turn regulated by TGFb and

FGFb. The shape of the condensate is precisely controlled by syndecan and tenascin regulation of fibronectin and exclusion of Hoxa2 from the template. The cells within the

template are induced to proliferate by FGFb, while template size is under the control of

BMPs (Hall and Miyake 2000, Wagner and Karsenty 2001).

Regulation of proliferation within the differentiating template is mediated by Pax1 and

Pax9, transcription factors which are regulated by BMPs (fig 1.2). Noggin binds and inactivates the BMPs regulating differentiation; therefore the Noggin/BMP relationship controls the size of the template and the initiation of differentiation and mineralisation.

Fig 1.2, regulation and function of BMPs in bone growth, Shh (sonic hedgehog) and Noggin regulate BMP activity in condensates. BMPs regulate patterning transcription factors, hox and msx genes. BMP-Cbfa1-SMAD interactions regulate differentiation. Adapted from Wagner and Karsenty, (2001).

BMP regulation of Cbfa1 activity is mediated by SMAD and TGFb (fig 1.3); both Cbfa1 and SMAD are inactivated by Smurf1. BMP interactions with Cbfa1 and SMAD can protect from Smurf1, this interaction is thought to redundantly regulate mineralisation and ensure that it is confined within the template (Stein et al. 2004, Paredes et al. 2004).

Fig 1.3 The mineralisation of the cartilage template. The BMP/SMAD/Cbfa1 interaction mediates differentiation to a mineralised phenotype. Adapted from Stein et al. (2004), Paredes et al. (2004) Nishio et al. (2006) and Jensen et al. (2010).

Although there is upstream regulation of Cbfa1 it is required for osteogenic differentiation and is directly involved in all major osteogenic gene function, including collagen I (Col IaI), matrix metalloproteinase I (MMPI), bone sialoprotein (BSP), TGFb,

RANKL, alkaline phosphatase (ALP), Osterix (Osx), osteopontin (OP), osteocalcin (OC);

Cbfa1 expression is greatest in mature osteoblasts (Stein et al. 2004).

1.6.1 BMP

Current approaches to bone engineering involve the use of transcription factors to tip the healing process in the favour of the desired outcome; many papers have been published regarding the use of BMPs to generate de novo bone. (Heliotis et al. 2006, Brandoff et al.

2008, Chen et al. 2011) BMPs regulate hox genes and are at peak effectiveness during cell proliferation, artificially adding BMPs can tip the balance of transcription factors in

favour of osteogenesis, potently up regulating mineralisation. BMPs are members of the

TGFb superfamily, which are all (over 20 identified), involved in morphogenesis, embryogenesis, cartilage and skeletal development. At the nuclear level increasing extracellular BMP influences Cbfa1 mediation of differentiation increases expression of

ALP, osteopontin, osteocalcin, bone sialoprotein and collagen α-1 (Leboy 2006).

1.6.2 Cbfa1

When binding to promoters in complex with BMPs and vitamin D3 Cbfa1 (Runx2) regulates osteoblast and chondrocyte differentiation; causing cells to switch to the osteoblastic pathway (Leboy 2006). In mice targeted knockdowns of Cbfa1 result in the absence of skeletal mineralisation (Komori et al. 1997).

Cbfa1 is a human homolog of the drosophila runt gene and is similarly essential to development; runx1 is essential for haematopoiesis and runx3 for neurogenesis and regulation of gastric epithelial cell growth. There is thought to be some cross over between the functions of Cbfa1 (runx2) and runx3, because both regulate chondrocyte proliferation and maturation (Cohen 2009, Swiers et al. 2010).

During development expression of Cbfa1 is mediated by TGFb, several BMPs, vitamin D and indirectly by sonic hedgehog (Shh). There are several isotypes of Cbfa1; two main promoters give rise to isotypes that are found predominantly in mature osteoblasts and hypertrophic chondrocytes (Cbfa1 type II). Type I Cbfa1 is found in less differentiated osteoblasts and chondrocytes, alternate splicing and regulation via post translational modifications has also been demonstrated (Enomoto et al. 2000, Bronckers et al. 2005).

Cbfa1 is involved in the differentiation of chondrocytes to hypertrophic chondrocytes, knockout mice exhibit delayed or non-existent completion of the chondrocyte template

essential for endochondral bone formation. Double knockdowns have no hypertrophic chondrocytes at all and no mineralisation (Ito and Miyazono 2003).

Regulation of osteoblast differentiation by Cbfa1 can be shown to occur by the removal of post-translational phosphorylation by protein kinases such as MAPK and ERK.

Activation of the MAPK/ERK pathways increases the availability of Cbfa1 and so increases differentiation. MAPK also mediates interactions between smad and Cbfa1

(Jonason et al. 2009).

Cbfa1 is not essential for the differentiation or function of osteoclasts, however in Cbfa1 knockout mice the regulation of the production of RANKL is abnormal; they produce more OPG and less RANKL than wild type mice. The balance of OPG-RANKL is the main regulator of osteoclastogenesis (Enomoto et al. 2003).

Ultimately as with all developmental transcription factors the regulation of expression and interaction of proteins is extremely complicated and contains many feedback systems that effectively regulate the production of bone. Increased expression of Cbfa1 mRNA has been shown repeatedly by many authors during osteogenic differentiation and is accepted late marker of differentiation (Gabbay et al. 2006, Crippa et al. 2008, Tsai et al.

2009, Curran et al. 2009, Bessa et al. 2009, Huang et al. 2010).

1.6.3 Alkaline phosphatase

Alkaline phosphatase (ALP) is a common marker of osteogenesis and is upregulated in many differentiating cell types from the colon to osteoblasts and osteoclasts (Matsumoto et al. 1990, Heinemann et al. 2000). ALP has been used in the clinic as a marker for osteoblastic function since 1994 and levels increase commensurate with bone turn over and mineralisation (Shankar and Hosking 2006, Tuck et al. 2008).

Approximately 50% of circulating ALP is from the bone, without functional ALP mineralisation fails (Orimo 2010, Saad and Lipton 2010). ALP hydrolyses phosphomonoesters at an alkaline pH, and is the sole method of breaking down inorganic pyrophosphates. Inorganic pyrophosphate (PPi) is the most important inhibitor of calcium phosphate turnover at the bone/cell interface; ALP removes these inhibiting effects and locally increases available phosphate (Whyte 2010).

ALP is used as an early indicator of increased osteoblast differentiation because levels of

ALP spike at a point between 10 and 16 days after initiation of differentiating conditions.

(Hoemann et al. 2009)

1.7 Isolation of marrow stromal cell osteoprogenitors

In the 1960s McCulloch and Till identified the haematopoietic stem cell by the clonal nature of spleen colonies found in nodules after injection with bone marrow derived cells.

Later they demonstrated that each colony was produced by a single bone marrow derived cell (Till and McCulloch 1961, Siminovitch et al. 1963). Friedenstein et al used an extension of this work and with colleagues Chailakhan and Lalykina discovered inducible osteoprogenitor cells residing in the spleen (Friedenstein et al. 1970). The spleen is the only organ in the body to be entirely comprised of mesenchymal elements, inducible osteoprogenitors were thought to retain the potential to become osteoblasts when stimulated correctly and therefore are not committed to a cell fate (Owen and

Friedenstein 1988).

In 1970 Friedenstein et al. isolated from a bone marrow cell suspension an adherent, clonogenic, fibroblast like cell culture which he named the CFU-F, these cells were demonstrated to be multi-potential were able to support myelopoiesis and could also differentiate into adipocytes, chondrocytes and osteoblasts (Friedenstein et al. 1992). In

vivo experiments demonstrated that 60% of clones are able to produce de novo bone

(Friedenstein et al. 1987). After this demonstration by Friedenstein et al. adherent and colony forming fibroblast like cells have been investigated as a potential source of autologous cells for clinically relevant applications; as well as investigations into the plasticity of adult cell populations (Korda et al. 2010).

The first method published by Friedenstein et al. (1970) consists of initial harvest of the bone marrow, sieve homogenisation and culture adherent cells. For the formation of the fibroblast colony forming unit, CFU-F, cells were seeded at a density of 1-5x104 cells per

25cm2, Friedenstein noted the colony forming efficiency at approximately 3-10 cells per

1x104 cells in the bone marrow. Recent human cell data puts this number between 52.2 ±

4.1 (children) and 32.3 ± 3.0 (adults) cells per 1x105 mononuclear cells (Kuznetsov et al.

2009).

Improvements in the method of Friedenstein have been attempted; a trabecular bone wash method was proposed (Grundle and Beresford, 1994). Collagenase digestions of trabecular bone and mechanically harvested cortical bone have been attempted (Tuli et al., 2003). In culture the mRNA phenotype of collagenase released cells decreases for markers of osteoblastic differentiation, Cbfa1 and osterix, and also for the adipogenic

PPARγ and FABP4 indicating a loss of determined cells in vitro (Sakaguchi et al 2004).

The loss of expression over culture time without loss of differentiation phenotype suggests heterogeneity even in clonally derived cultures, or may indicate a form of culture selection is occurring (Wagner and Ho 2007).

It has also been shown that multipotent stromal cells, also MSC, can be isolated from the blood, harvesting from 50 ml of buffy coat using a cell elutriator, total cell numbers of monocytes on average were 2.14 ± 0.22 x107 and from this between one thousand and ten

thousand PB-MSCs survived harvesting, attached to tissue culture plastic and proliferated

(Zvaifler et al. 2000, Kuznetsov et al. 2001).

The simplest method of separation is when a single antibody is used to isolate a cell strain from a heterogonous population. Modern fluorescent antibody cell sorting (FACS) technology allows for the isolation of selected populations of cells by surface epitope characterisation. Unfortunately the whole BMSCs population has defied characterisation by these methods, authors have reported screening hundreds of surface epitopes with no consensus found, either the patterns change with age of culture or they involve subjective judgements of labelling dimness (diGirolamo et al. 1999, Harichandan and Buhring

2011).

1.7.1 Culture of BMSC.

The pluripotency of BMSCs can be maintained in vitro given the correct culture conditions, cells are grown at sub-confluent levels ranging from 1-5 x10³ cells per cm², in

DMEM (αMEM is also common) culture media with various quantities of FCS reported as being necessary, from 7-20% (Meuleman et al. 2006, Lange et al. 2007). EGF, PDGH, and FGF have been shown to ensure non-differentiation of cells over longer culture periods with less emphasis on the concentration of cells per cm² (Maegawa et al. 2007).

1.7.2 in vitro differentiation

BMSC cultures differentiate into four major cell types, adventitial reticular cells (stromal fibroblasts), adipocytes, chondrocytes and osteoblasts (Bianco et al. 2008, Sarugaser et al.

2009). There have also been more contentious reports of several other cell types, including some involving transdifferentiation across a germ line (Bianco. 2007, Bianco et al. 2008).

The first loss of multipotency from cultures is the stromal fibroblast. Depending on the origin of the tissue the potential for differentiation to adipocytes is lost from BMSC cultures; this occurs most often in bone marrow derived cultures and least often in adipose derived cultures (Xiao et al. 2010). Adipogenesis is achieved with stimulation by insulin, indomethacin and dexamethasone, lipid accumulation occurs and can be observed after staining with oil red O (Bianco et al. 1999). The next potential lost from the cell cultures is the osteoblast or chondrocyte lineage, most often chondrocytes.

Chondrocytes are differentiated in pellet culture, increased density in three dimensions is a potent stimulator of differentiation; addition of TGF-b is the most effective chondrogenic supplement (Vater et al. 2011). Osteoblasts are normally the terminal lineage of bone marrow derived MSCs and therefore differentiate from almost all populations of marrow stromal cells (Vater et al. 2011). Since osteoblasts differentiate in situ the major osteogenic stem cell niche is thought to be in close proximity to the majority of osteoblast activity (Bianco 2011).

1.8 BMSC in vivo.

Multipotent skeletal progenitor cells have been found in a large number of tissues, organs and the peripheral blood of humans (Hilfiker et al. 2011), kidneys (Park et al. 2010), umbilical cord blood (Erices et al. 2000), adipose (Zuk et al. 2002), heart and dermis

(Riekstina et al. 2009). All these tissues are highly vascularised, leading some to consider that the MSC niche is the vasculature (Augello et al. 2010). Vasculature pericytes, cells associated with capillaries, isolated from the retina of cows have been shown to express the specific osteogenic markers osteopontin, osteonectin, and bone sialoprotein and when inserted into in vivo diffusion chambers produced de novo bone and cartilage (Canfield et

al. 1996). Experiments on vascular smooth muscle cells have produced similar results

(Yan et al. 2011).

1.8.1 BMSC transplants

Injection of cultured non-sex matched BMSCs have shown that engraftment to the host occurs predominantly to areas of growth and repair, the literature reports engraftment to bone, cartilage, the CNS (astrocytes and neurons), lung, spleen, muscle, and thymus after systemic injection (Tolar et al. 2010). Allogeneic BMSCs have been shown to reduce and even stop the incidence of graft versus host disease after HSC transplant, indicating interaction and modulation of cells within the immune system (Porta et al. 2004).

Transplantation of carefully selected BMSCs has been shown to restore some function to pancreatic islets in NOD/SCID mice (Bell et al. 2011), restore cartilage (Richter 2009) and support angiogenesis (Casteilla et al. 2010). BMSC have been shown to enhance bone regeneration (Ma et al. 2011).

1.9 Bone replacements

The gold standard for bone replacement is autografted bone however there are problems associated with its use, including lack of donor tissue, graft site morbidity and resorption of graft or underlying bone (Stellingsma et al. 2004, Giannoudis et al 2005, Brandoff et al. 2008, Dusseldorp and Mobbs 2009). When looking at alternatives to the autograft the surgeon takes into account the biocompatibility of the implant, availability, ease of shaping and molding during surgery, the long term stability and radio translucence

(Brydone et al. 2010).

There are several materials currently available on the market that meet some of these requirements, polymethyl methacrylate (PMMA), bioactive glass, allograft or demineralised bone tissue and calcium phosphate cements (Brydone et al. 2010).

Additionally some materials are or purport to be integrating, osteoconductive or biomimetic, this generation of materials has been designed purposefully with the intent to replace or fully integrate the graft with the patient’s own tissue (Lu et al. 2010).

Bone cements have been in use for several decades; the most commonly used is PMMA

(Acharya et al. 2011, Lewis 2011). PMMA sets in an exothermic reaction and does not integrate with the surrounding tissues. This is both positive and negative; for temporary repair of bone tissue in the young or for cosmetic surgery the use of a non-integrating polymer can be beneficial (Acharya et al. 2011, Lewis 2011). However for long term repair of tissue a regenerative or integrating polymer is of more use and there are marketed materials available for the surgeon in this case (Ambard and Mueninghoff

2006). PMMA has a reputation as being more difficult to keep clean and as there is no integration with the host tissue and therefore no blood supply it is difficult to remove an infection once it has compromised the cement (Winkler 2009).

For some time hydroxyapatite has been used as a bone filler in the shape of blocks or packed in as granules, HA has some disadvantages, it is not remodelled fully if made too thick for example, however it does have the advantage that it can be soaked in growth factors that can promote the production of new bone (Heliotis et al. 2006, Beekmans et al.

2008).

Calcium phosphate cements are well known and commercially available, discovered in the 1980’s they are extensively researched with several thousand papers published and were a focus of research in the “bone and joint decade” (Woolf 2000, Dreinhofer et al.

2004). Calcium phosphate cements are entering a second generation of development as bioactive, fast setting, drug eluting scaffolds for tissue regeneration and delivery of a desired pharmaceutical (Chow 2009).

Due to extensive research into calcium phosphate cements over the last 25 years it is well known that setting times can be manipulated easily by adjusting the powder sizes, the concentration of accelerant, total volume of liquid phase and the addition of nucleation foci or crystallisation inhibitors (Chow 2009). It is also known that porosity is malleable and that lower porosity equates to stronger materials and increased persistence in the defect, however reducing the porosity inhibits remodelling of the cement (Takagi et al.

1998, Chow 2009).

The major benefits of calcium phosphate cements are that they are injectable, can be drug eluting throughout the matrix of the cement, it has good compressive strength, low setting times at body temperature, it is moulded into the defect and therefore provides a perfect fit, it exhibits porosity which can be used to increase in growth of new tissue. It can be utilised instead of packing with HA because of these qualities (Takagi et al. 1998, Low et al. 2010).

In most available calcium phosphate cements there are two end point setting products; brushite and apatite. Apatite cements in vivo continually increase mechanical strength compared with surrounding tissue after the setting period. Brushite cements decrease their mechanical strength after the setting time (Moseke and Gbureck, 2010).

Cements on the market are made with two components, all use tetra calcium phosphate

(TTCP) as the main dissolving component, the faster setting cement systems with high early strength were developed using the more soluble DCPD (brushite) instead of DCPA

(monetite) as the second nucleating cement component. Some experimental cements use nanocrystaline hydroxyapatite as the nucleating agent, in an attempt to form calcium deficient hydroxyapatite (CDHA) as an end product (Johnsson and Nancollas 1992,

Barralet et al. 2004, Burguera et al. 2006).

Cements have been developed with addition of agents for enhanced osteogenesis, infection control, and reduction in resorption (Brandoff et al. 2008, Panzavolta et al.

2010, Neut et al. 2010). The addition of selective oestrogen receptor modulators has been demonstrated, but the effect on primary cell culture is unknown (Perut et al. 2011).

1.10 Soy phytoestrogens.

Glucose conjugated isoflavones (fig 1.4a), glycosides, and malonates (fig 1.4c) are the prevalent phytoestrogens in the soybean, Glycine max and are the storage form within the plant (Barnes 2010). In Asian food the use of soy, historically and today, is mostly as a fermented product, soy sauce, miso, soy paste, and tempeh are all produced in this manner. Fermentation of soy results in an increased concentration of the aglycones and decreased conjugates (Barnes 1998). Using boiling water to make soy milk hydrolyses the malonates to glycosides and toasting dry soy decarboxylates malonates to produce acetylates (fig 1.4b) (Friedman and Brandon 2001). Therefore isoflavones and their conjugates are different in different foods depending on processing.

After ingestion of isoflavones the aglycones are readily absorbed by the human gut and glycosides are quickly broken down by cells expressing lactase (Day et al. 2000). The remaining conjugates are metabolised and modified by the bacteria in the large intestine, metabolites are dependent on gut microflora (Chien et al. 2006, Nielsen and Williamson

2007, Barnes 2010).

Once in the blood isoflavone aglycones are active in the human body as mimics of oestrogen, and are therefore phytoestrogens (Lindner 1976). High dietary intake of phytoestrogens has been linked with a reduction in breast cancer risk, an increase in bone density and reduction in the incidence of fractures (Usui 2006, Castelo-Branco and

Cancelo Hidalgo 2011). Some studies show an 8-14% increase in bone density, soy

consumption has been linked to the inhibition of bone loss in menopausal and postmenopausal women (Lagari and Levis 2010). Phytoestrogens have been shown to bind to and activate oestrogen receptors α and β, ERα ERβ (Pilsakova et al. 2010).

Isoflavone R1 R2 genistein OH H

daidzein H H

glycitein H OCH3

A B C

D E F

Fig 1.4, Drawing of a generic isoflavone and the three main isoflavonoid aglycones, modifications to carbon 7 occur in the soy plant and after heat treatment of aglycones. Modifications can include A) glycosides, B) acetyl glycosides, C) malonyl glycosides. For comparison the human hormones D) 17β-estradiol, E) estrone, D) progesterone are also represented. Molecules are orientated as found in vivo at ERβ (Pike et al. 1999).

Genistein and daidzein bind ERβ with greater efficiency than ERα, once bound they are reported as activating transcription of an overlapping but different set of responsive genes than oestrogen, dependant on the model used (Katzenellenbogen et al. 2000, Kushner et al. 2000, Akbas et al. 2004, Schreihofer 2005, Tang et al. 2008, Rando et al. 2009,

Hartman et al. 2009, Ni et al. 2010). The selectivity in ER ligand binding has led to the term selective oestrogen receptor modulator, SERM. (Pilsakova et al. 2010)

Genistein has agonist activity for both ERα and ERβ, but has an affinity between 7-30 times greater for ERβ, dependant on the conditions of the study (Morito et al. 2001).

Similarly daidzein has affinities of 420nM and 100nM for α and β respectively, although the isoflavones never reach saturation levels in vivo (Krishnan et al. 2000, Morito et al.

2001). The SERMs ER ligand has been shown to modulate the expression of different genes compared with oestrogen (Akbas et al. 2004).

In developing human bone both ER types are expressed, in adult bone ERα is predominantly expressed in compact bone and ERβ is expressed at higher levels in trabecular bone (Robinson et al. 1997). ERα deficiency in men causes non fusion of the growing bone epiphyses, and therefore continuous longitudinal growth of the long bones and low BMD (Smith et al. 2010). ERβ -/- female mice have a masculine bone phenotype which is not different from the male wild type (Smith et al. 2010).

Oestrogen receptors were first reported as nuclear transcription factors that are ligand activated with specificity dependent on ligand (Nilsson and Gustafsson 2011). ERs have also been found outside the nucleus, bound to membranes and acting as transcription factors without ligand binding (Giguere 2002, Mendelsohn and Karas 2010).

While ERα is most highly conserved ERβ has multiple isoforms, including variants that have; extended N termini, an 18 amino acid insert in the ligand binding domain or carboxyl terminal variants unable to form ligands or modulate transcription (Foryst-

Ludwig and Kintscher 2010). In mice double knockouts survive to adulthood, suggesting that there may be other ERs (McCauley et al. 2002). There is evidence for oestrogen related receptors that may offer redundancy in the system (Giguere 2002).

The activity of the dimerised transcription factor once bound to a ligand varies dependant on the conformational change induced in the receptor. The variation in conformational

change between the two receptor types can be most fully seen when comparing the action of a ligand forming agent such as the SERM tamoxifen. When bound to ERα tamoxifen shows oestrogen agonist activity, however upon binding ERβ tamoxifen is an antagonist, differences in the N terminal region of the receptors are thought to explain the agonist or antagonist effect on the ERE (Chang et al. 2008, Alexi et al. 2009).

The ERs have been shown to bind the promoter of some genes thereby modulating their activation (Leclercq et al. 2006). Activated ERs can also bind other transcription factors influencing the transcription of genes that they cannot directly mediate. Indirect activity of ERα in the absence of a ligand binding factor and therefore transcription factor activity has been demonstrated subsequent to activation of membrane receptors for IGF-I, EGF and TGF (Kushner et al. 2000, Bondar et al. 2009). Most of these authors suggest the mechanism of activation is due to a phosphorylation of the N terminal of the ER, suggesting that variation occurs in receptor isotypes and perhaps between cell types

(McCauley et al. 2003, Ho 2004).

Oestrogens have an indirect effect on osteoclasts by increasing the osteoblast production of OPG and thereby reducing osteoclast activation by the RANK ligand, oestrogen may also directly effect osteoclast differentiation, maturation and activity by modulating cytokine production (Martin-Millan et al. 2010).

Evidence for indirect regulation of genes binding other transcription factors has been substantiated, for instance AP-1, Sp1 and some c-AMP responding elements are all managed in similar activated ER to protein binding processes (Safe and Kim 2008). AP-1 is up-regulated by ERα ligand but down-regulated by the active ERβ with possible methods including sequestering of cofactors that combine with fos and jun (Jakacka et al.

2001, Zhao et al. 2010). AP-1 regulation may explain the reported proliferation effects of

isoflavone doping, and go some way to explain the selective effect of isoflavones on tissue and cell types (Schreihofer 2005).

1.11 Measurements of isoflavone concentrations in vivo:

After studying Japanese men that consume a soy rich diet Adlercreutz et al. report a mean concentration of total, free plus un-conjugated, genistein at 276nM with approximately

4% (9.64 nM) available to activate ERs; Setchell and Cassidy describe the free genistein levels to be between 1 and 1.5% temporally relative to consumption of soybeans (Setchell and Cassidy 1999, Adlercreutz 2002).

Further studies by Adlercreutz observed the free genistein levels in plasma to be between

20 and 40nM for 2 to 8 hours after consumption, no residual genistein is found after 24 hours (Adlercreutz et al. 2004). Comparing these levels of free genistein with those found to activate ERα and β it is apparent that activation of ERβ is easily possible with physiological levels of genistein, however the concentration of genistein for ERα activation is sub-optimal (Rickard et al. 2003). Modulation of ER expression is apparent in some studies at these concentrations (Berner et al. 2010).

In a long term fully randomised study undertaken by Squadrito et al. in 2007 supplementing 55mg of daily dietary genistein compared to HRT and placebo, in all 90 postmenopausal women took part and their responses were investigated at 6 and 12 months (Squadrito et al. 2003, Atteritano et al. 2007). The group that received the isoflavone supplement were measured as having 1.5μM serum genistein at 6 months and

1.7μM at 12 months, using the Aldercreutz report as a guide they are likely to have about

20nM free genistein, sufficient to activate ERβ.

At six months the genistein group demonstrated significantly increased serum levels of

ALP and osteocalcin and at 12 months the OPG/RANKL ratio was reduced to a greater

extent in the isoflavone group than the HRT group. Measurements of BMD showed increases of 3-4% after 12 months in both the HRT and genistein groups relative to minor losses in the placebo group. Key to this study is the length of time the participants were followed, the number of participants, and the concentration of free genistein (Squadrito et al. 2003, Atteritano et al. 2007). In this study free genistein was available in sufficient quantities to activate ERβ. However not all studies provide measurements of the bioavailability of the administered genistein, therefore the level of free genistein is likely the reason for the reported variability in the effects of genistein supplementation.

It is apparent from the study of isoflavones in the human diet that systemic concentrations are almost always lower than full saturation of ERs. Therefore pharmaceutical doses of isoflavones in cements should be considered for maximum effect in order to influence the local environment.

1.12 Hypothesis.

The overall hypothesis of this study is that the addition of soy extracts to a calcium phosphate cement increases the osteogenic differentiation of primary human cells.

This overall hypothesis can also be separated into several sections; that pure isoflavone and soy extracts will increase osteogenesis of a confluent cell model; the increase in osteogenesis will be observable in culture medium, osteogenic medium, and components of osteogenic medium; the addition of a competitive ER inhibitor will reduce the observed effect demonstrating that phytoestrogens are responsible for the osteogenesis.

1.13 The aim of the work.

These hypotheses will be pursued by observing the effect of a wide range of concentrations of genistein on an osteosarcoma cell line, developing the methodology for analysis of proliferation and differentiation. Once a model is established primary human cells will be isolated and characterised prior to addition of soy extract to the primary model. The soy extract will be incorporated into osteogenic cements and the results related to the observations with pure isoflavone and soy extract.

CHAPTER 2

MATERIALS AND METHODS.

2.1. Materials.

2.1.1. Equipment for extraction of BMSCs.

Product Source Surgical Nail clippers Swann-Morton® Ltd, Sheffield, UK Scalpels Needles. All gauges NHS direct. UK syringes

2.1.2 Extraction consumables

Product Source Ficoll Paque plus – 1073g/ml Sigma-Aldrich Company Ltd., Dorset, UK

2.1.3 Cell culture plastic

Product Source Cellstar® 25cm2, 75cm2, 175cm2 Filter Cap Greiner Bio-One Ltd, Stonehouse, Glos, Tissue Culture Flasks UK 30 ml sterile universal plastic tubes BD Falcon® 1ml, 5ml, 10ml, 25 ml Becton Dickinson Labware, NJ, USA. Serological Pipettes Supplied by Marathon Laboratory Supplies BD Falcon® Blue Max™ 50ml Ltd, London, UK Polypropylene Conical Tubes BD Falcon® Blue Max™ Jr. 15ml Polystyrene Conical Tubes BD Falcon® Integrid™ Petri Dishes BD Falcon® Multiwell™ 96-well culture plates

Product Source Nunc™ CryoTube™ 1.8ml Vials Fisher Scientific UK, UK

Non and plugged glass Pasteur pipettes

2.1.4 Cell culture consumables.

Product Source Hanks’ Balanced Salts Solution (HBSS) Gibco™ Invitrogen Life Technologies Ltd, without CaCl2 and MgCl2 Paisley, Scotland, UK

0.05% Trypsin-EDTA

Dimethyl Sulphoxide (DMSO) Hybri- Sigma-Aldrich Company Ltd., Dorset, UK

Max®

2.1.5 Wash buffers

Product Source Penicillin/Streptomycin solution, stock Gibco™ Invitrogen Life Technologies activity of 5000U/ml & 5000µg/ml Ltd, Paisley, Scotland, UK clindamycin 5000U/ml Sigma-Aldrich Company Ltd., Dorset, UK

2.1.6 Isolation and Culturing Medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd,

MEM) + 1000mg/l Glucose + Paisley, Scotland, UK

GlutaMAX™ + Pyruvate

10% Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes)

2.1.7 Osteogenic medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd, MEM) + 1000mg/l Glucose + Paisley, Scotland, UK GlutaMAX™ + Pyruvate 10% Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes) 10nM Dexamethasone Sigma-Aldrich Company Ltd., Dorset, UK 50µg/ml ascorbate-2-phosphate 10mM β-glycerophosphate

2.1.8 Adipogenic medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd, MEM) + 4000mg/l Glucose + Paisley, Scotland, UK GlutaMAX™ + Pyruvate 10% Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes) 1µM dexamethasone Sigma-Aldrich Company Ltd., Dorset, UK 10 µg/ml insulin 60µM indomethacin 500µM isobutyl-methylxanthine

2.1.9 Chondrogenic medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd,

MEM), 4000mg/l Glucose, GlutaMAX™ Paisley, Scotland, UK

Pyruvate

10% Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes)

Product Source 100nM dexamethasone Sigma-Aldrich Company Ltd., Dorset, UK 50µg/ml ascorbate-2-phosphate 40 μg/ml L-proline 100 μg/ml pyruvate 10 ng/ml transforming growth factor (TGF)-β3 50 mg/ml ITS+ Premix Becton Dickinson

2.1.10 Experimental media

2.1.10.1 Growth medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd, MEM) + 1000mg/l Glucose + Paisley, Scotland, UK GlutaMAX™ + Pyruvate Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes) 50µg/ml ascorbate-2-phosphate Sigma-Aldrich Company Ltd., Dorset, UK

2.1.10.2 β-glycerophosphate medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd,

MEM) + 1000mg/l Glucose + Paisley, Scotland, UK

GlutaMAX™ + Pyruvate

10% Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes)

50µg/ml ascorbate-2-phosphate Sigma-Aldrich Company Ltd., Dorset, UK

10mM β-glycerophosphate

2.1.10.3 Dex medium

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd,

MEM) + 1000mg/l Glucose + Paisley, Scotland, UK

GlutaMAX™ + Pyruvate

10% Foetal Calf Serum, New Zealand (heat inactivated at 56°C for 35 minutes)

10nM Dexamethasone Sigma-Aldrich Company Ltd., Dorset, UK

50µg/ml ascorbate-2-phosphate

2.1.10.4 Osteogenic medium

2.1.11. Staining solutions.

2.1.11.1 Alizarin red stain, for calcium anions

Product Source 40mM Alizarin Red S (ARS), CI: 58005 Sigma-Aldrich Company Ltd., Dorset, UK

10% Acetic acid

10% ammonium hydroxide

2.1.11.2 Von Kossa stain.

Product Source 5% sodium thiosulfate Sigma-Aldrich Company Ltd., Dorset, UK 1% Silver Nitrate 0.1% nuclear fast red ammonium sulphate

2.1.11.3 Oil Red O, lipid stain.

Product Source 4% Oil Red O Sigma-Aldrich Company Ltd., Dorset, UK 2-Propanol (isopropanol)

2.1.11.4 Cartilage stain, for the assessment of proteoglycans.

Product Source 0.5%, w/v Alcian Blue (C.I. 74240) pH 2.5 Sigma-Aldrich Company Ltd., Dorset, UK in 3% acetic acid Weigert's Iron Hematoxylin

2.1.11.5 Collagen staining

Product source 0.1% Sirius red F3B (C.I. 35782) Sigma-Aldrich Company Ltd., Dorset, UK Saturated aqueous solution of picric acid

HCl acid

2.1.11.6 HPI stain for membrane integrity

Product Source Dulbecco's Modified Eagle's Media (D- Gibco™ Invitrogen Life Technologies Ltd, MEM) + 1000mg/l Glucose + Paisley, Scotland, UK GlutaMAX™ + Pyruvate Propidium iodide (1mg/ml) Sigma-Aldrich Company Ltd., Dorset, UK Hoechst (5mg/ml)

2.1.12 Cell lysate buffer

Product Source 0.1% Triton X-100 Sigma-Aldrich Company Ltd., Dorset, UK 10 mM Tris-HCl pH 7.8

0.5mM MgCl2

2.1.13 Solutions for determination of ALP activity

Product Source p-Nitrophenol phosphate (p-NPP) Sigma-Aldrich Company Ltd., Dorset, UK 0.1M NaOH shrimp ALP, (EC: 3.1.3.1) p-NP

2.1.14 Total protein determination by Bradford reagent

Product Source Bradford reagent Sigma-Aldrich Company Ltd., Dorset, UK BSA

2.1.15 LDH cytotoxicity testing

Product Source Promega's CytoTox 96 non-radioactive Promega UK Ltd. Southampton, UK cytotoxicity LDH assay PBS Sigma-Aldrich Company Ltd., Dorset, UK

2.1.16 Immunohistology

Product Source 10% neutral-buffered formol saline Genta Medical, York, UK Menzel-Glaser® Microscope Slides Surgipath Europe Ltd., Peterborough, UK Microscope Glass Cover Slips Chance Propper Ltd, Smethwick, Warley, UK ImmEdge Hydrophobic Barrier Pen Vector Laboratories, Ltd., Peterborough, UK Vectabond™ Reagent

Product Source Acetone/methanol 50:50 Sigma-Aldrich Company Ltd., Dorset, UK Haematoxylin (Gill’s, Harris’) BDH, VWR International Ltd., Poole, UK Eosin Y

2.1.17 Immunohistology antibodies.

Product Source 1º Monoclonal Mouse Anti-human STRO-1 R&D Systems, Abingdon Science Park, UK 1º Monoclonal Mouse Anti-human SSEA-4 Abcam plc. Cambridge UK 2° Sheep Anti-Mouse IgG (Whole Sigma-Aldrich Company Ltd., Dorset, UK Molecule) FITC Conjugate Normal sheep serum (for the 2° FITC Anti- Mouse IgG non-specific staining elimination)

2.1.18 Antibody diluent

Component Source 1% BSA (blocking & stabilizer) Sigma-Aldrich Company Ltd., Dorset, UK 0.5% Triton X-100 (penetration enhancer) 0.05% sodium azide (preservative) 0.1% cold fish skin gelatine (blocking) 0.01M PBS, pH 7.2-7.4

2.1.19 Cryostat sectioning

Product Source O.C.T.® Compound, embedding medium Tissue-Tek, Europe for frozen tissue specimens Vectabond™ Reagent Vector Laboratories, Ltd., Peterborough, UK Menzel-Glaser® Microscope Slides Surgipath Europe Ltd., Peterborough, UK Liquid Nitrogen BOC gases, Guildford, UK

2.2 Methods.

For all studies a LEEC incubator operating at 37ºC and 5% CO2 was used to maintain constant culture conditions. In order to maintain sterility of the culture environment all cell manipulations were undertaken using GLP and aseptic technique in a class II microbiological hood.

2.2.1 BMSC Extraction protocol.

After ethical approval and consent was received (Brighton and mid sussex REC

06/Q1907/81) routinely discarded tissue was collected from the operating theatres at the

Queen Victoria Hospital, East Grinstead. After removal of any attached soft tissue, the samples were washed for 15 minutes in Hanks Buffered Salt Solution (HBSS) with penicillin, streptomycin and clindamycin added at 5000Uml-1/5000µgml-1/5000Umg-1 respectively. Subsequently the marrow channel was flushed with DMEM with added penicillin, streptomycin and clindamycin, at the same concentrations as in the Hanks using a 10ml syringe and a 19G needle, the marrow channel flushings were collected aseptically and left to sediment. The bone was opened using surgical nail clippers and the marrow channel was scraped clean of bone fragments. Both the flushings and trabecular bone fragments were pooled and minced with a scalpel. The cell suspension was diluted to a known volume then homogenised by pushing through decreasing graduations of needle diameter, starting with 16G and working down to 20G.

After homogenisation the solution was layered over the top of the same volume of Ficoll-

Paque PLUS (1.073g/ml) in a 50ml falcon tube. The resulting tubes were spun in a 20°C cooled centrifuge at 400g for 30-40 minutes with the brake off. The Ficoll-Paque PLUS gradient produces a buff coloured layer at the interface known as the buffy coat. This layer consists of the mono-nuclear cells including the bone marrow stromal cells

(BMSCs) and stem cells. The buffy coat was carefully removed from the Ficoll-Paque

PLUS and the cells were washed in HBSS and counted in a haemocytometer. BMSCs occur at a frequency of between 1 and 50 cells in every 100 thousand mononuclear cells

(Jaiswal et al. 1997, Kuznetsov et al. 2009). Therefore the mononuclear cells were plated into tissue culture plastic flasks with an area of 75cm2 at a seeding density of 5x103/cm2 with isolation medium and left to attach for 18 hours. Any BMSCs will attach to the tissue culture plastic in this time; the remaining cells are discarded in the first medium change. Due to the rarity of BMSC in the initial culture the cells grew in clonal colonies and appeared to have a fibroblast like phenotype. At the periphery of the colony a few cells had the appearance of osteoblasts, square and with noticeable actin bundles in the cytoplasm. After between several days to weeks, the colonies in the flasks achieved 80% confluence and were passaged. All subsequent flasks were seeded at 3x103 cells per cm2 and harvested at 80% confluent.

2.2.2 Cell maintenance

Medium was completely exchanged every 2-3 days; the old medium was removed and replaced with fresh warm medium until the cells grew to the correct confluence for passaging. Cells were seeded at 3-5x103 cells per cm2 and harvested at 80-90% confluent.

2.2.3 Cell Passaging.

Before the cells reach the point at which confluence retards cell growth (for BMSCs this is approximately 80-90%) the cells require passaging. The medium was removed from the monolayer which was then washed in HBSS to remove remaining serum. After washing, the cell layer was covered in a small amount of trypsin solution (for a 75cm2 flask about 1ml of solution is sufficient) and then placed into the 37°C incubator until the cells ball up and begin to detach from the tissue culture plastic. This takes about 5 minutes. The tissue culture flask was tapped gently to aid removal of the cells from the

surface. The trypsin was then “inactivated” with the addition of a known volume of growth medium; being careful to rinse all the cells from the plastic and into suspension.

A sample was taken from the suspension and counted on a haemocytometer. Cells were then used experimentally or a sample seeded into a new flask at a known cell number.

Alternatively, where cell numbers were higher samples of cells were frozen and stored above liquid N2.

2.2.4 Cell freezing

BMSCs are frozen in much the same way as any other human primary cells. The freezing medium is prepared by adding 10% DMSO to growth medium, in this case DMEM with

10% FCS, stored at 4ºC prior to use. The cells are passaged and counted on a haemocytometer. After counting the cells are centrifuged to a pellet and re-suspended in the cold freezing medium at 3x106 cells per ml and pipetted into a freezing vial. The freezing vial is placed into a controlled rate freezing vessel that regulates the drop in temperature to 1°C per minute down to -80ºC when put into a -80°C freezer overnight.

After overnight freezing the vial is removed the following morning and stored above liquid nitrogen. Prolonged storage in a -80°C freezer decreases viability.

2.2.5 Cell thawing

Cells were removed from liquid nitrogen storage and thawed in the hand, once the medium inside the vial was liquid, but still cold, the cell suspension was removed and placed into a sterile 30ml universal. 4°C DMEM was added to the freezing vial and any remaining cells were washed into the universal. The volume was made up to 10ml and the universal centrifuged at 400g for 5 minutes. The supernatant is removed and the cells re-suspended in 4°C DMEM with 10% FCS then centrifuged and re-suspended in warm

DMEM with 10% FCS. After a viable cell count on a haemocytometer, the cells were seeded into cell culture flasks at the desired concentration per cm2.

2.2.6 Quantification of cell number by haemocytometer.

Cells were removed from the homogeneous suspension using a sterile glass pipette, and placed into the chamber of a prepared haemocytometer. Live cells in the four outside corner squares of the chamber grid were counted and the number divided by four.

Because of the known volume of the liquid over the chamber grid (1x10-4 ml) this cell number (n) represents the number of cells x104 per ml of liquid in the original sample of cell suspension. For example, a count of 300 cells divided by 4, is 75 (n=75), so

75x104/ml in 10 ml. Therefore 7.5x105ml-1 and 7.5x106 cells total.

2.2.7 Growth curves.

The number of population doublings (PDs) achieved by the cells was determined as described by the equation.

Log(x/y) Log2

Where x is the total number of cells harvested and y the total number of cells seeded. PDs are combined at each passage and these cumulative PDs (CPDs) can plotted against the time in culture.

2.2.8 Cambrex BMSC maintenance protocols.

Cambrex primary human mesenchymal stem cells (BMSCs) were purchased from

Poietics (POPT-2501). Cambrex cells are positive for CD105, CD166, CD29, and CD44 and negative for CD14, CD34 and CD45. After maintenance in commercial culture medium (POPT-3001) pluripotency was demonstrated by differentiation into

chondrocytes, adipocytes and osteoblasts in the media described at 2.1.1.7, 2.1.1.8 and

2.1.1.9.

2.2.9 Characterisation of putative BMSCs

2.2.9.1 Immunohistochemistry

1x104 cells were transferred to a glass slide using a cytospin centrifuge and fixed in formal saline or 50/50 acetone and methanol (2.1.1.16). After fixation the cells were rinsed in PBS and placed into a humidified chamber for 30 minutes incubation at 20°C with the normal serum from the species used for secondary antibody (2º) preparation

(Sheep/Goat) at a concentration of 1:30 diluted in PBS (2.1.1.17). The use of this serum prevented non-specific binding of the secondary antibody. A wax pen was used to mark out the area to which the antibody was to be added. To stain the cells the primary antibody (1º) was diluted to the predetermined optimal concentration in antibody diluent and added to the sample. After overnight incubation with the primary antibody at 4°C the slide was washed once in Tween PBS (1/4000) for one minute followed by three washes in PBS for 5 minutes each. The secondary antibody was added at predetermined optimum dilution to minimise cross reactivity and determine best use of stock and incubated at room temperature for 1 hour. After staining was complete the slide was washed once in

Tween PBS (1/4000) for one minute and could be mounted in DPX for long term storage.

Visualisation of the fluorescein tagged antibody was undertaken on a fluorescent microscope. Antibodies were diluted in antibody diluent.

2.2.9.2 Single cell cloning.

A limiting dilution of BMSC suspension was plated at 0.3 cells per well in a 96 well plate, after 4 hours any cells that had not attached or wells containing duplicate cells were discarded and the medium was changed. Cells were cultured in growth medium and when

the well reached approximately 75-80% confluence they were trypsinised, counted, and the total cell suspension was split into 4 and grown in either growth, osteogenic, chondrogenic or adipogenic media with medium changes every three days. SSCs to be differentiated in chondrogenic medium cultures were plated in a small volume in the centre of the tissue culture plastic and allowed to attach in a small area before addition of the medium. These cultures were maintained for 21-30 days or until visual analysis indicated that cells were differentiated.

2.2.9.3 Differentiation

2.2.9.3.1 Attached monolayer culture differentiation.

Homogeneous single cell clones or enriched heterogeneous cell populations were differentiated by passaging into and maintaining the cells in the appropriate differentiation medium. Cells were monitored for changes in their morphology, indicating a changing phenotype. After 21-30 days in differentiation media the different monolayers were stained for either adipocyte, chondrocyte or osteoblast differentiation.

2.2.9.3.2 Pellet culture for the differentiation of chondrocytes.

1ml of a cell suspension of 2x105ml-1 was pipetted into a 15ml polypropylene falcon tube in chondrocyte differentiation media. The tube was centrifuged at 500g for 5 minutes and placed in the incubator where it remained for 21-30 days. Medium changes occurred at 3-

4 day intervals.

2.2.9.3.3 Qualitative assessment of mineralisation by Von Kossa.

5% sodium thiosulfate and 1% Silver Nitrate aqueous solutions were prepared for the

Von Kossa stain. 0.1% nuclear fast red counter stain was also prepared with 0.1g nuclear fast red and 5g ammonium sulphate per 100ml dh20. The mineralised monolayers were immersed in the 1% silver nitrate solution and placed on the window sill in direct

sunlight for half an hour or, if dark, next to a 60-100W bulb for about an hour. When a darkening of the monolayer was observed the flask was washed in several changes of distilled water followed by a 5 minute wash with 5% thiosulfate solution to remove any un-reacted silver. After several washings in distilled water the counter stain was applied for about 5 minutes or until the staining was notable.

Using this stain any calcium deposits show as black or brown-black, cell nuclei as red and cytoplasm as pink.

2.2.9.3.4 Qualitative assessment of adipocytes by Oil red O staining.

Adipocytes were stained using 4% w/v oil red O dissolved in isopropyl alcohol to make a stock stain. The working stain solution was made by mixing the stock stain 40/60 with dH20, mixing for 2 minutes and leaving to settle for 10 minutes before filtering the stain through a Whatman's number 1 filter and using immediately. Cell monolayers were very gently fixed in formal saline for 10 minutes and washed in PBS to remove debris. After fixing the monolayer was overlaid with the working stain solution for 5 to 10 minutes on the bench. Lipid vacuoles in the cell are visible as red globules. Burst cells were avoided as they spread stained lipid over the cells and obscured the field.

2.2.9.3.5 Qualitative assessment of Chondrocyte differentiation.

2.2.9.3.5.1 Preparation of chondrocyte pellets for cryostat sectioning.

Chondrocyte pellets were removed from culture and placed in foil boats with an appropriate volume of OCT for their size. The foil was then placed into the vapour phase immediately above a quantity of liquid nitrogen. After 10-15 minutes the frozen OCT had become hard and opaque. The now frozen samples were stored at -80ºC until ready to cut.

2.2.9.3.5.2 Cryostat sectioning and staining of pellet cultures

After warming the frozen tissue block to -20ºC it was mounted and sectioned at a thickness of 10-15µm onto Vectabond treated glass slides and left to dry for 24 hours at room temperature. After fixation in 10% formal saline the slides were stained.

2.2.9.3.5.3 Assessment of chondrocyte differentiation by cartilage staining.

Formal saline fixed samples were stained in Weigert's Iron Hematoxylin for 10 minutes followed running tap water for 10 minutes to de-stain. An 0.5% w/v alcian blue solution

(pH 2.5) was added and left to stain for 25 minutes followed by a dip destain in 1% acetic acid. The 3% acetic acid solution (pH2.5), Alcian blue stains both sulphated and carboxylated acid mucopolysaccharides and sulphated and carboxylated sialomucins

(glycoproteins). Using this stain nuclei are black, while cartilage (glycosaminoglycans) is blue-green.

2.2.10. Quantitative analysis of experimental cultures.

2.2.10.1 Quantification of mineralization, osteoblast differentiation staining.

Using a modification of the method described by Gregory et al. (2004) an Alizarin Red quantitative stain directly determines the amount of mineralization that has occurred in cells grown in osteogenic medium. Cell monolayers were fixed in formal saline and stored in PBS. Before staining the monolayer was washed twice with excess distilled water (dH20) and blotted dry. 200µl of 40mM Alizarin Red S (ARS) at pH 4.1 was applied per well of a 96 well plate which was then incubated at room temperature for 20 minutes. Unincorporated dye was aspirated, the plate washed four times with dH20 for five minutes per wash, then finally aspirated and blotted dry. At this point the protocol could be continued or if time is short the stained cell monolayers could be frozen at -

20ºC.

The dye quantification was achieved by incubating the stained monolayer with 200µl of

10% acetic acid for 30 minutes at room temperature, or until the monolayer became loosely attached to the plate. The plate was scraped and the suspension was transferred to an epindorff tube. The tubes were then heated to 85ºC for 10 minutes and transferred to ice for five. After heating the slurry was centrifuged at 15,000g for 30 minutes and the supernatant was removed to a new tube. 100µl of 10% ammonium hydroxide was added to the supernatant to neutralise the acid and the pH was checked to ensure it was between pH 4.1 and 4.5. Supernatants were read at 405nm and quantified using a standard curve of ARS at the same pH.

2.2.10.2 Analysis of osteoblast differentiation by alkaline phosphatase (ALP).

The alkaline phosphatase (ALP) assay is based on the procedure used by H. Declercq et al. (2004) and measures the hydrolysis of p-Nitrophenol phosphate (p-NPP) to p-

Nitrophenol (p-NP) catalysed by the ALP enzyme, the quantity of enzyme present is directly measureable by the speed of p-NP production. Shrimp ALP of known activity was used as a positive control.

The differentiated osteoblast monolayer was lysed with 0.1% Triton X-100 in dH2O at

37ºC. After lysis has been observed the lysate was removed in a total volume of 250µl using buffer. The buffer comprised of; 10 mM Tris-HCl pH 7.8, 0.5mM MgCl2 and 0.1%

Triton X-100. The lysate solution was then stored at -80ºC until testing.

To test ALP activity 50µl of ALP substrate containing p-NPP from Sigma (N7653) was added to 50µl of cell lysate solution in an optical 96 well plate and the time noted. The plate was incubated at 37ºC until the colour developed. Absorbance measurements were checked at 405nm to ensure levels of at least 0.2, at an absorbance of 0.2 slightly visible yellowing can be observed with the eye. To stop the reaction 50µl of 0.1M NaOH was added. The time was noted and final measurements were taken at 405nm. As a positive

control shrimp ALP (sigma, P9088) was used at physiological levels of between 7 and 13

U/L, to produce a 45 minute end point absorbance of approximately 0.2. A standard curve was produced from serial dilution of p-NP (sigma, N7660) in 0.1M NaOH. Only linear data from the standard curve was used to determine the levels of ALP in the lysates. The final ALP activity was measured in µM of p-NP per minute per mg protein, enough enzyme to produce the product at 1µM/minute = 1 Unit, after adjustments for total protein levels to minimise fluctuations due to cell number or metabolism, ALP is expressed as Umg-1 protein.

2.2.10.3 Determination of total protein content by Bradford reagent.

Protein levels in the lysate from the ALP assay were quantified using Bradford reagent.

5µl of lysate was added to 195µl of x1 Bradford reagent. The absorbance measurements were made at 595nm within 10 minutes. The standard curve was calculated on a diluted

BSA 2mg/ml stock solution and only linear data was used to determine protein content.

2.2.10.4 Quantification of collagen by Sirius red staining.

Cell culture lysate was centrifuged at 500g for 5 minutes and 50µl of supernatant was plated into a 96 well plate, covered, and allowed to sit in a humidified 37ºC incubator for

24 hours. After 24 hours had elapsed, the plate was uncovered and placed into a non- humidified drying oven at 37ºC for a further 24 hours or until dry. After drying, the plates were washed repeatedly in 100µl of dH2O and tapped dry. 0.1% Sirius red stain in saturated Picric acid (dissolved in dH2O) was applied at 100µl per well and developed on the bench for 1 hour. After 1 hour the residual stain was removed from the wells and discarded. The stained collagen was washed with three further 100µl quantities of 0.1M

HCl. After tapping the plates dry, 100µl of 1M NaOH was added to the wells and left

with mild shaking for fifteen minutes. Collagen production was measured on a plate reader at 540nm.

2.2.10.5 Quantification of cell proliferation and death.

2.2.10.5.1. LDH determination.

2.2.10.5.1.1 LDH Standard curve

To plot a standard curve, the optimum cell number and the maximum sensitivity of

Promega's CytoTox 96 non-radioactive cytotoxicity LDH assay were determined. Cell dilutions of 50 to 1x106 cells per 100µl were prepared in PBS and added to wells of a 96 well plate. The cells were ruptured by two freeze thaw cycles. 50µl of supernatant was removed from each well and transferred to a fresh plate. The samples were incubated in the dark at room temperature for 30 minutes with an equal volume of substrate mixture provided in the CytoTox 96 non-radioactive cytotoxicity assay. After incubation, 50µl of the stop solution was added to each well and the change in colour was measured spectrophotometrically at 490nm. Only linear data from the standard curve was used to determine the cell number in the lysates

2.2.10.5.1.2 Quantification of cell number by LDH assay

Further experimental samples tested for LDH activity were lysed using the lysis procedure in the ALP assay. The monolayer was lysed with 0.1% Triton X-100 in dH2O at 37ºC. After lysis had been observed the cell lysate was removed in a total volume of

250µl of buffer containing 10 mM Tris-HCl pH 7.8, 0.5mM MgCl2 and 0.1% Triton X-

100. 50µl of these lysates were each added to a well of a 96 well plate and incubated in the dark at room temperature for 30 minutes with an equal volume of substrate mixture provided in the CytoTox 96 non-radioactive cytotoxicity assay. After incubation 50µl of the stop solution was added to each well and the change in colour was measured

spectrophotometrically at 490nm. Comparisons were made with the control samples and cell numbers were quantified using the previously defined standard curve.

2.2.10.5.1.3 Relative quantification of cell death by LDH release.

LDH is a highly conserved mitochondrial enzyme involved in the Krebs cycle. LDH can be used as a measurement of cell number but is also released from a cell when it dies.

Therefore measurement of supernatant LDH can quantify cell death relative to monolayer

LDH and produce a value that is relative to both cell death and proliferation. This quantification of the culture is free as the supernatant would otherwise be discarded and allows inference of the proliferation and cell death that occurs between measurements of the cell number in the monolayer.

LDH was used in this study as a relative measurement of cell death occurring between harvesting points. The LDH available in the monolayer combined with the LDH activity of the supernatant, harvested at the same time, was used as the measure of total LDH. The supernatant LDH was expressed as a percentage of total LDH. Therefore cultures in which cells have died heavily and have small numbers of cells remaining in the monolayer relative to large amounts of LDH activity in the supernatant will have a value approaching 100%. The method used was as described here.

Media supernatants were harvested from cultures prior to lysis of monolayers.

Supernatant LDH was measured by addition of 50µl to a well of a 96 well plate and incubation in the dark at room temperature for 30 minutes with an equal volume of substrate mixture provided in the CytoTox 96 non-radioactive cytotoxicity assay. After incubation 50µl of stop solution was added to each well and the change in colour was measured spectrophotometrically at 490nm.

The total LDH output of the culture was combined from measurements of the lysis LDH and the supernatant LDH, the supernatant LDH was expressed relative to total LDH.

2.2.10.5.2 Measurement of proliferation by Brd-U incorporation

Brd-U incorporation was undertaken using a BrdU Cell Proliferation Assay Kit (QIA58-

1000TEST) from calbiochem. The manufacturer’s instructions were followed. 10,000x

BrdU stock was diluted 1:2000 into tissue culture media. 20µl of this stock was added to the monolayer of BMSCs attached to each well of a 96 well plate. The plate was returned to the incubator with the BrdU for 18 hours to label all the actively dividing cells. This was followed by several washes in PBS and fixing in formal saline for 30 minutes at room temperature. The quantification of BrdU incorporation was performed using an

ELISA as provided in the Calbiochem kit. The primary anti-BrdU antibody was diluted

1x1000 in antibody diluent and 100µl was added directly to the washed monolayers for 1 hour. After repeated washings, 100µl of the peroxidase goat anti-mouse IgG HRP conjugate was added to the cells for 30 minutes. The plate was washed in wash buffer x3 and then once in dH2O before 100µl of the substrate solution was added to the plate and incubated for 15 minutes in the dark. After incubation, 100µl of the stop solution was added and the plate dual read in a spectrophotometer at 450-540nm within 30 minutes of addition of stop solution.

2.2.10.5.3 Quantification of proliferation by the uptake of radio labelled thymine.

2x104 cells were pulse labelled with 2mCi of [3H]-thymine for 18 hours in medium, after this period the remaining radio label was removed and the cells harvested on membranes.

Cells were lysed using a lysis buffer supplied by Perkin and Elmer and the cell lysate was passed through a membrane mat. The mat was dried and heat sealed into a plastic package with just enough scintillate to cover the mat. Reading of the incorporated thymine was undertaken in a microbeta trilux from Perkin and Elmer.

2.2.10.5.4 Live and dead staining with Hoechst Propidium Iodide.

Hoechst Propidium Iodide (HPI) stain was prepared by adding 5mg/ml Hoechst and

1mg/ml Propidium Iodide to growth medium. To use the stain the medium was removed from the cultured cells and replaced with the staining solution, staining was visualised using an inverted fluorescent microscope with UV and Texas red blocks for respective visualisation of Hoechst and Propidium Iodide. Live cells stain blue (H), dead cells stained red (PI) and late apoptotic cells stained both blue and red. 6-8 representative micrographs of the stained cells were taken combined with phase contrast photographs of the same.

2.2.10.6 Quantification of levels of intracellular Adenine triphosphate

ATP was quantified as a measurement of cell activity by using the PerkinElmer ATPlite kit, the protocol followed was as per the manufactures recommendations. Cells were grown in PerkinElmer CulturPlate's, opaque tissue culture treated plastic plates. To each

100µl volume of medium 50µl of lysis solution was added and the plate shaken at

700rpm for 5 minutes. After shaking, 50µl of substrate solution was added to the cell lysate and shaken for a further 5 minutes at 700 rpm. The plate was then dark adapted for

10 minutes and the luminescence was measured in a microbeta trilux from Perkin and

Elmer.

2.2.10.7. Molecular biological techniques for the determination of mRNA quantities.

2.2.10.7.1. Total RNA extraction from monolayers.

RNA was extracted using the Rneasy mini kit from Qiagen, following the manufacturers recommended methods. First the monolayer of cells was washed several times in excess hanks, followed by the addition of a lysis buffer (buffer RLT) at 350µl per 5x106 cells.

After lysis was observed, the lysate was homogenised by repeated pipetting through a

200µl eppendorff tip. An equal volume of RNase/DNase free 70% ethanol was added and 47

the solution was homogenised. 700µl of the homogenised lysate was added to each silica spin column provided in the kit and centrifuged (8,000G for 15-30 seconds). Each column was washed with 700µl wash buffer (RW1) and centrifuged. This was followed by two washes in 500µl of buffer 2 (RPE). The column was centrifuged for 5 minutes at full speed to remove all remaining buffers and to dry the column. Elution from the column into a clean tube was achieved with 30-50µl of RNase free dH2O and centrifugation at 8,000G for 2 minutes. For extremely high levels of RNA (greater than

30µg) a repeat elution using the same volume of RNase free dH2O was used.

2.2.10.7.2. Quantification of total RNA.

1µl of the extracted total RNA was quantified using a RNA Quantit kit from Invitrogen and a Qubit fluorimeter following the manufacturer’s protocols. 1µl of isolated RNA was added to 199µl of RNA quantification buffer containing the fluorescent marker Quant-iT

RNA Br at room temperature and analysed against the standards provided in the kit to find the concentration in ng/ml.

2.2.10.7.3. Production of cDNA from total RNA.

1ng of RNA was used in a reverse trascriptase (RT) PCR reaction to create cDNA. The following was added per reaction tube;

Component Quantity 1ng of total RNA x µl 50ng of random hexamers 1µl 10mM dNTP mixture 1µl dH2O to a total volume of 10µl x µl

This mixture was homogenised and incubated at 65ºC for 5 minutes and then placed directly on ice for at least 1 minute. 9µl of master mix was added to the primer template mix and incubated for 2 minutes. Per run the master mix consisted of;

Component Quantity 10x Buffer 2µl

25mM MgCl2 4µl 0.1M DTT 2 µl ribonuclease inhibitor 1µl

To the 19µl volume per reaction a 1µl volume or 50U of reverse transcriptase

SuperScript II was added and incubated for 10 minutes at 25ºC, followed by 42ºC for 50 minutes and a termination step of 70ºC for 15 minutes and placed on ice immediately. cDNA is much more stable than RNA and was either frozen at -80ºC for long term storage or used immediately. Prior to qPCR the RNA strand was removed from the

RNA:DNA hybrid by incubation with 1µl RNase H at 37ºC for 20 minutes before continuing to the qPCR protocol.

2.2.10.7.4. qPCR primer design.

The gene of interest was found on the NCBI gene website. Searching for cbfa-1 results in the display of the Homo sapiens gene runt-related transcription factor 2 (RUNX2). The geneID, 860, was taken and this information was used in an online primer design program (www.RTPrimerDB.org). Such programs are much quicker than designing primers by hand, eliminate human error, provide comparisons with multiple miss-priming databases and contain interon spanning primers which eliminate genomic products. The primers were compared with the available literature for efficacy. RUNX2(CBFA1) primers were F- CACCTCTGACTTCTGCCTC, R- GACTGGCGGGGTGTAAGTAA producing a 178bp product. While GAPDH was F-GAGTCCACTGGCGTCTTCA and

R-GGGGTGCTAAGCAGTTGGT and a product of 190bp

2.2.10.7.5 Experimental design for qPCR qPCR is a test for the relative abundance of mRNA species, it is a significant improvement over standard PCR techniques because it is quantitative in real time. SYBR green detection of double stranded DNA product from qPCR is the most commonly used as well as the cheapest and easiest method of detection. Because SYBR green is a non- specific detection technique it is particularly important that the design of primers does not produce product from gDNA. Although, since cDNA was made, this technique is only semi-quantitative accurate measurements of gene expression are possible for most applications. For low copy number mRNA transcripts the most efficient method of quantification is to compare the number of transcripts of the gene of interest with the number of copies of a housekeeping gene that has a more static level of expression.

GADPH is a commonly used housekeeping gene that was chosen as the relative standard in these experiments. For the qPCR reaction a master mix was made using the volume of

25µl per well and making enough to provide for an extra well. The master mix contained;

Component Volume (µl) x10 buffer 2.5 10mM dNTP’s 1.25 x 4 (5) X1000 SYBR 0.25 Primers 1 x 4 (4) dH2O x Taq 1 cDNA 1ng x Total 25

The mix was divided into the wells of a qPCR plate (25µl/well) and run through the RT-

PCR machine programmed as follows;

Repetitions Temperature (°C) Time (seconds) 1 95 180 40 95 10 60 30

Followed by a melt curve analysis of double stranded products from 50ºC to 95ºC after

40 cycles had elapsed.

2.2.10.7.6. Analysis of qPCR results.

The melt curve for each PCR was examined, the melt curve should be a steady declining line from the coldest, and therefore most double stranded products, to 95ºC, where all the products are melted. Towards the end of the melt, at approximately 80ºC depending on the length of product, the most abundant product in the PCR should melt all at once producing a step in the graph. When this melt graph was plotted in the accepted inverse way there appeared to be a large spike at this temperature. Double peaks or no peak at all indicated a problem with the reaction and the result was discarded. Any peaks in the no cDNA control and the result was discarded. Next the threshold was set; the threshold was the major factor that impacted on the Ct, i.e. the point at which the exponential PCR reaction crossed the threshold. The threshold was set by the controls at the linear part of the data by adjusting the line, using the data plotted on a log scale. This allowed easy evaluation of the time at which the fluorescence measured above background. The same threshold was used for all the samples in the experiment.

2.2.10.7.7. The ΔΔCt analysis method.

This method utilises the difference in levels of reference mRNA and experimental mRNA to determine the relative change in expression of the experimental mRNA. The reference mRNA acts as a loading control to eliminate any variation within each sample. In the experimental design we compared treated cells with untreated cells and without a stable

point of reference in the qPCR any differences could have been due to template loading in the reaction mixture. Therefore for one sample the difference between the reference and the experimental is the ΔCt. The ΔΔCt is the difference between the experimental samples after the adjustment for the reference genes has been taken into account.

2.2.11. Extraction and characterisation of isoflavones from soy.

2.2.11.1 The extraction of isoflavones from soy flour.

500g of Soy flour was de-fatted with 500ml of hexane overnight. Soy isoflavones were extracted from defatted soy flour by shaking at 30°C in an ethanol/water, 20/80 solvent mix overnight. The supernatant was removed, centrifuged and then layered over a column packed with 3 cm glass wool and 1 cm silica gel. After vacuum rotary evaporation of the ethanol at 30°C the remaining extract was freeze dried to produce a stable isoflavone concentrate. The efficacy of the extraction process was determined by HPLC analysis using internal forononetin standards and bookend standard runs as described at 2.2.11.2.

2.2.11.2 The determination of the soy extract composition by HPLC.

The freeze dried crude soy extract from 2.2.11.1 was dissolved in 80/20 DMSO/dH2O and its composition determined by HPLC in the following manner.

The HPLC setup consisted of a 25ºC, 21.2 minute program running at 1ml/minute and consisting of a solvent gradient composed of A and B (A/B) where A is Acetonitrile and

0.1% glacial acetic acid, B is ultra pure water and 0.1% glacial acetic acid. At the start of the run 10µl of sample was injected into solvents in the following ratio (A/B) 10/90 for 0 minutes, followed by 15/85 for 0.1 minutes, 20/80 for 4 minutes, 40/60 for 9 minutes and

60/40 for 4.1 minutes and finally 100/0 for 4 minutes. After the sample has run off the column at 21.2 minutes a 10 minute period was allocated for the column to return to

10/90 solvent ratios in preparation for the next sample run. Samples were run through a

262nm UV spectrophotometer and the area under the peak integrated and used to determine relative composition with the standards. Bookend standards included genistein, daidzein and formononetin, an internal formononetin spike was included with every unknown.

Using this reverse-phase HPLC setup it is expected that the first to elute will be the glucosides of isoflavones in the order, puerarin, daidzein, glycitin, and genistein.

Followed by the malonylglucosides, acetlyglucosides and the aglycones in the same running order. Metylated aglycones, of daidzein (formononetin) and genistein (biochanin

A), are retained for longest.

2.2.12. Production of Biomaterials.

2.2.12.1 Production of calcium phosphate cement containing soy extract.

The components for the cement are described in 2.2.12.3. The 10% gelatine solution was warmed until molten in a 50ºC water bath. 3.636g of the CaP cement was weighed, which corresponds to 2ml of the liquid phase. 1ml of the soy solution and 1ml of the gelatine solution were mixed and frothed at 50ºC for 1 minute. The cement and the liquid phase were then combined and mixed gently. The cement was placed into a 5ml syringe and then injected into a mould, for the purposes of the cell studies this was the well of a 96 well plate. After 2 hours incubation in the dark at 37ºC the cement was immersed in saline solution at 37ºC for 12 days. The saline was changed every 2 days until the cements were fully set at day 12.

2.2.12.2 Production of calcium phosphate cement containing gelatine.

The components for the cement are described in 2.2.12.3. The 20% gelatine solution was warmed until molten in a 50ºC water bath. 3.636g of the CaP cement was weighed, which corresponds to 2ml of the liquid phase. 2ml of the gelatine solution was mixed and

frothed at 50ºC for 1 minute. The cement and the liquid phase were then combined and mixed gently. The cement was placed into a 5ml syringe and then injected into a mould, for the purposes of the cell studies this was the well of a 96 well plate. After 2 hours incubation in the dark at 37ºC the cement was immersed in saline solution at 37ºC for 12 days. The saline was changed every 2 days until the cements were fully set at day 12.

2.2.12.3 Components required for production of the gelatine and soy cements.

Cement components 10% gelatine in dH2O 250mM Na2HPO4 (accelerant)

20% gelatine. in dH2O Calcium phosphate cement; 98% α- 40% soy extract in accelerant TCP(5.4µm) and 2% hydroxyapatite w/w.

2.2.13 Experimental method for the application of isoflavones to cell cultures.

2.2.13.1 Initial seeding densities and media.

After passaging the experimental cells were seeded at confluent density in growth medium as per the table at 2.2.13.4 and left overnight at 37ºC and 5% CO2 to form confluent monolayers in 96 well tissue culture grade plates. After attachment the medium was removed and replaced with experimental medium containing genistein or soy extract and then incubated in a humidified incubator at 37ºC and 5% CO2. The experimental medium consisted of low glucose (1g/l) D-MEM (invitrogen 21885-108) for BMSCs and

α-MEM (invitrogen A10490-01) for MG63s, with the addition of 10% selected heat inactivated FCS and 50µg/ml ascorbate-2-phosphate, as in the tables at 2.1.1.10. To this medium was added either 10mM β-glycerophosphate (β-gly) (sigma G6501), or 1x10-8M dexamethasone (dex) (sigma D4902), or osteogenic medium containing both β-gly and dex (osteo).

2.2.13.2 Experimental setup for analysis of the effect of pure genistein.

For the addition of pure genistein to the cells a stock solution of 300mM Genistein

(Sigma G6776) was prepared in DMSO (Sigma D4540), this stock solution was diluted to the required concentrations in the experimental media. After dilutions the DMSO vehicle was at most 1µl/10ml of media, this concentration was added to media applied to the controls. Medium was removed and replaced every 3 days. After the required incubation time the medium was removed from the monolayer and stored at -80ºC. The monolayers were washed in PBS and the remaining cells assayed as described in the results section using the methods described at 2.2.10.

2.2.13.3 Experimental setup for the culture of soy extract with CBMSCs.

For the application of crude soy isoflavone extract produced as per the instructions at

2.2.11.1 the amount of genistein in the extract was characterised by HPLC as per the instructions at 2.2.11.2 and the quantity of soy extract that contains 300mM of genistein was determined and dissolved in DMSO, this stock solution was prepared freshly each time, diluted as required in experimental medium and used immediately. DMSO vehicle was applied to the controls at a concentration of 1µl/10ml. In a duplicate experiment fulvestrant was added to the media at 100nM in addition to soy extract or vehicle.

Medium was removed and replaced every 3 days. The quantity of genistein in the extract is described by the table at 2.2.13.5. After the required incubation time the medium was removed from the monolayer and stored at -80ºC. The monolayers were washed in PBS and the remaining cells assayed as described in the text using the methods at 2.2.10.

2.2.13.4 Experimental seeding densities for production of a confluent monolayer.

Cell type Cell number (cm-2) MG63 passage 120-123 1x105 Passage three BMSCs, purchased from Cambrex™ 2x104

2.2.13.5 Quantity of Genistein in the soy extract.

Genistein concentration (µM) Soy extract (mg/ml) 0.94 8.44 1.88 16.89 3.75 33.78 7.5 67.57 15 135.14 30 270.27 300 540.54

2.2.14 Experimental method for the application of BMSCs to cement biomaterials.

4.5x104 CBMSCs were seeded in growth medium to the surface of the cements and left overnight at 37ºC and 5% CO2 to attach; methods for cement production are at 2.2.12.

After attachment the medium was exchanged with fresh growth medium and then incubated at 37ºC and 5% CO2. Medium was exchanged every 2 days for 24 days, wells were harvested on days 0, 12 and 24. Analysis of results was achieved by use of the methods described at 2.2.10.

2.2.15. Preparation and visualisation of biomaterials by SEM.

Medium was removed from the cements and they were washed twice in PBS prior to fixing in 1% glutaraldehyde in PBS for 3 hours followed by 5 washes in dH2O. Cements were subsequently freeze dried under vacuum overnight at -80°C prior to sputter coating in a thin layer of palladium and visualisation in a JEOL JSM-6310 scanning electron microscope.

CHAPTER 3.

THE DEVELOPMENT AND CHARACTERISATION OF A

CONFLUENT DIFFERENTIATING IN VITRO MODEL FOR

THE ANALYSIS OF OSTEOGENESIS DURING CULTURE

WITH GENISTEIN.

3.1 Introduction

Immortal cell lines are often used for large scale screening and testing of compounds

(Godl et al 2003). They are an effective compromise between gauging the accurate effects of compounds, achieving adequate cell numbers and reproducing the same test conditions over long periods of time (Rubinstein et al 1990). Cell lines originating in human osteosarcomas have been tested with many oestrogenic compounds (Dubin et al.

2001, Kian et al. 2004, Stossi et al. 2004).

There is discussion in the literature regarding osteosarcoma cell lines and the level of ER expression. U2OS cell lines are often used to study the effect of activating oestrogen receptors because they are not constitutively expressed, allowing levels of expression to be controlled. (Kallio et al 2008, Djiogue et al 2010). SAOS2 cells have been reported to have an abnormal receptor expression profile with regards to sex-steroid response elements, whereas MG63s have been shown to behave in a more native manner and are the cells of choice for this study (Chen et al. 2006, Saraiva et al. 2008).

The MG63 cell line expression of oestrogen receptors is normally weighted in a ratio of

ERβ to ERα for adult trabecular bone. Activation of the ERα ligand in MG63s increases

57 proliferation (Morris et al 2006, Kameda et al 2010). Genistein, a SERM, binds ERα with low efficiency, and signalling by the differentiating ERβ ligand predominates. Because

ERs are expressed in a tissue specific manner and the effect of genistein is dependent on cell line derivation and ER expression binding of genistein to ERs effects the expression of genes containing ERE regions in their promoters. For example, studies undertaken in human MCF-7 breast carcinoma and G-292 osteosarcoma cell lines cells show that genistein and daidzein are greater than 100 fold more effective at stimulating ERβ than

ERα. MCF-7 cells have a 208 fold excess of ERα over β where as G-292 cells have a 1.3 fold excess of β over α. Clearly there is a large variation in immortalised cell lines and their expression and possible responses to SERMs (Chrzan 2007).

MG63s exhibit a relatively normal early osteoblast expression phenotype, they are capable of differentiation and although they proliferate more quickly than primary human osteoblasts, and have a less inducible ALP expression, they are an ideal cell line model to study the effects of SERMs such as genistein (Nakamura et al 1988, Farley et al 1989,

Fournier and Price 1991, Dang 1994, Clover 1994).

MG63s have been used to study the effects of genistein previously. In 2006 Morris et al reported the effects at 48 hours finding an increase in collagen and mineralisation, markers of osteogenesis (Morris et al. 2006). In this chapter this study will be expanded to include a confluent MG63 model and analysis of 4 and 6 days genistein treatment.

Chen et al (2003) used a mouse cell line, MC3T3-E1, to demonstrate the up-regulation of

ALP to a maximum expression at 6 days with and without genistein. Consequently we used the incubation time of 6 days to monitor the effects of genistein on differentiation markers in human MG63s.

In the literature, genistein is often used as an osteogenic supplement to osteogenic medium. To clarify the contribution of genistein to the differentiation that occurs in osteogenic medium and to increase understanding of genistein induced differentiation several different media formulations were tested; growth medium (GM) alone, GM containing β-glycerophosphate (Beta), GM containing dexamethasone (dex) and osteogenic medium containing both dexamethasone and β-glycerophosphate (Osteo).

From this study such information should allow a comparison of the use of genistein as an osteogenic agent alone and in combination.

3.2 Hypothesis

Addition of genistein to a confluent cell line model accelerates osteogenic differentiation when compared with confluent cells in growth medium, osteogenic medium and osteogenic medium components.

3.2.1 Aims.

To determine an optimum concentration range of genistein in both growth medium and osteogenic medium for the efficacious differentiation of osteoblasts in a confluent differentiating immortal cell model, identifying relevant methods for the continued study of in vitro effects in more relevant cell types.

3.3 Methods.

3.3.1 Experimental procedure.

MG63s at passage 120-123 were seeded at 1x105 cells cm-2 in growth medium (alpha-

MEM, 10% heat inactivated FCS and 50µg/ml ascorbate 2-polyphosphate) and incubated overnight to form confluent monolayers in 96 well plates. After attachment the medium was removed and replaced with experimental media formulations comprising growth medium (GM), GM + β-glycerophosphate (beta), GM+ dexamethasone (dex) or

59 osteogenic medium (osteo) as described at 2.1.10. Genistein was added at concentrations

0-300 µM to each of these media types and the culture continued for 2, 4, and 6 days with media changes every three days where appropriate. Full methodology is at 2.2.13

After the required incubation time the used media were removed and stored at -80ºC for later analysis. The monolayers were washed in PBS and the remaining cells assayed as described at 2.2.10.

3.3.2 Statistics

Statistical analysis was performed in Minitab 15. The analysis of data was achieved using a one-way analysis of variance (ANOVA) with multiple comparison tests; Tukey’s family error rate was used to determine the significance between groups without a defined control. i.e. Comparisons of a single concentration of genistein across time points. Dunnett’s family error rate was used to compare groups with a defined control i.e. genistein compared with vehicle. The confidence level was set at 95 and P values of

<0.05 were considered significant.

3.4 Results

3.4.1 The effect of genistein on cell number as derived by analysis of lactate dehydrogenase (LDH) content.

A standard curve was prepared by lysis of known MG63 cell numbers and quantification of the LDH they contained. Experimental monolayers were lysed and the LDH activity was converted into cell numbers, the cell number was plotted against the concentration of genistein for each media. Full methods can be found at 2.2.10.5.1.

Microscopic evaluation of the monolayers was undertaken prior to LDH assessment,

LDH derived cell numbers were commensurate with observations.

3.4.1.1 The LDH derived cell number found after MG63s were cultured in formulations of growth media containing genistein for 2, 4 and 6 days.

1.2E+05 GM Day 0 * GM Day 2 * 1.0E+05 * GM Day 4 * GM Day 6 8.0E+04 *

6.0E+04 number

Cell 4.0E+04

2.0E+04

0.0E+00 0 1.88 3.75 7.5 15 30 300

Genistein concentration (µM)

Fig 3.1 Total MG63 cell number available in the monolayer after culture in growth media and genistein, derived from LDH activity and assessed on Days 2, 4 and 6. 3x104 cells were seeded per well on Day 0. N=2, error bars are SD, * indicates statistically significant compared with control (p<0.05).

The vehicle control MG63s, cultured in growth medium (Fig 3.1), did not proliferate over the 6 day experiment.

The addition of genistein to growth medium had no significant effect until Day 6 when the cell number increased significantly from earlier time points. This significant increase in cell number on Day 6 was observed in genistein concentrations from 1.88µM to 30µM but no increase was seen at the highest genistein concentration, 300µM.

3.4.1.2 The LDH derived cell number after MG63s were cultured in formulations of beta-glycerophosphate media containing genistein for 2, 4 and 6 days.

1.2E+05 Beta Day 0 * Beta Day 2 * 1.0E+05 Beta Day 4 * * * * Beta Day 6 8.0E+04

6.0E+04 Number

Cell 4.0E+04

2.0E+04

0.0E+00 0 1.88 3.75 7.5 15 30 300 Genistein (µM)

Fig 3.2 Total MG63 cell number available in the monolayer after culture in beta media and genistein, derived from LDH activity and assessed on Days 2, 4 and 6. 3x104 cells were seeded per well on Day 0. N=2, error bars are SD, * indicates statistically significant compared with control (p<0.05).

Control MG63s cultured in beta medium (Fig 3.2) with vehicle did not significantly change from seeded cell number at any time over the experiment.

In genistein supplemented media MG63s did not significantly differ from the control or seeded cell number until Day 6. At Day 6 all treatment cultures were significant when compared with the control.

3.4.1.3 The LDH derived cell number after MG63s were cultured in formulations of dexamethasone media containing genistein for 2, 4 and 6 days.

1.2E+05 Dex Day 0 * Dex Day 2 1.0E+05 Dex Day 4 Dex Day 6 8.0E+04

6.0E+04 number

Cell 4.0E+04

2.0E+04

0.0E+00 0 1.88 3.75 7.5 15 30 300

Genistein concentration (µM)

Fig 3.3 Total MG63 cell number available in the monolayer after culture in dex media and genistein, derived from LDH activity and assessed on Days 2, 4 and 6. 3x104 cells were seeded per well on Day 0. N=2, error bars are SD, * indicates statistically significant compared with control (p<0.05).

In dex medium control cultures (Fig 3.3) the cell number remained statistically the same until Day 6 when cell number was significantly larger than at the previous time points.

In genistein treated cultures a trend of increasing cell number was observable at Day 4, an earlier time point than when genistein was supplemented to growth media (fig 3.1) and beta-glycerophosphate media (Fig 3.2). At Day 6 cell number was significantly higher than previous time points for all treatments; 0-300µM. In genistein treatments cell numbers were equivalent to growth and beta media (P>0.05).

3.4.1.4 The LDH derived cell number after MG63s were cultured in formulations of osteogenic media containing genistein for 2, 4 and 6 days.

1.2E+05 Osteo Day 0 Osteo Day 2 1.0E+05 Osteo Day 4 Osteo Day 6 * * * * 8.0E+04

6.0E+04 number

Cell 4.0E+04

2.0E+04

0.0E+00 0 1.88 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.4 Total MG63 cell number available in the monolayer after culture in osteogenic media and genistein, derived from LDH activity and assessed on Days 2, 4 and 6. 3x104 cells were seeded per well on Day 0. N=2, error bars are SD, statistical significance is represented by * (p<0.05).

Control cultures in osteogenic media (fig 3.4) do not significantly increase in cell number until Day 6 of the experiment.

In genistein treated cultures a trend of increasing cell number was observed at Day 4, earlier than GM (fig 3.1) and beta media (fig 3.2) and in keeping with the dex media (fig

3.3). At Day 6 a significant increase in cell number, compared with controls, was apparent (P<0.05).

3.4.1.5 Summary of cell number found after treatment of MG63s with genistein.

After treatment of cells with genistein added to media without dex, i.e. GM and beta media, (fig 3.1 - 3.2) cell number increased compared with controls when measured at

Day 6. In media with added dex, i.e. dex and osteogenic media, (fig 3.3 – 3.4) cell number was observed to increase at Day 4 and Day 6. There was little difference between concentrations of genistein, however in growth media cells treated with 300µM did not behave the same as in the other media types; they proliferated significantly more than the controls.

3.4.2 Relative cell death measured by release of supernatant LDH.

LDH is a highly conserved mitochondrial enzyme involved in the Krebs cycle. LDH activity can be used as a measurement of cell number (3.2.1) but is also released from the cell when it dies. Therefore measurement of supernatant LDH can quantify cell death relative to monolayer LDH and produce a value that is relative to both cell death and proliferation. This quantification of the culture is free, as the supernatant would otherwise be discarded, and allows an inference of the proliferation and cell death that occurs between measurements of the cell number in the monolayer.

LDH was used in this study as a relative measurement of cell death occurring between harvesting points. The LDH available in the monolayer combined with the LDH activity of the supernatant, harvested at the same time, was used as the measure of total LDH. The supernatant LDH was expressed as a percentage of total LDH and presented against genistein concentration. Therefore treatment wells in which cells have died heavily and have small numbers of cells remaining in the monolayer relative to large amounts of

LDH activity in the supernatant will have a value approaching 100%. The methods used can be found in full at 2.2.13

3.4.2.1 Relative LDH release after culture in growth media with genistein.

70 GM Day 2 GM Day 4

LDH 60 * * GM Day 6 * total 50 * * to

* * * * 40 * relative

* * * 30 * LDH * * 20

10 Supernatant

% 0 0 1.88 3.75 7.5 15 30 300 Genistein concentration(µM)

Fig 3.5 Relative LDH released from MG63 cultures after treatment in growth media supplemented with genistein for 2, 4 and 6 days. The data is expressed as the percentage of LDH in the supernatant compared with the total LDH found in both the monolayer and the supernatant. N=3, error bars indicate SD, bars marked with * are significant (P<0.05) compared with the control.

Growth medium vehicle controls (Fig 3.5) increased the LDH found in the supernatant across all time points. Since the cell number remained the same throughout (fig 3.1) this indicates proliferation and cell death occurred equally at all time points but that cell death increased over the experimental time period.

In the genistein treatments at Days 2 and 4 there was a dose response to increasing genistein supplementation with a larger percentage of cell death at low levels of genistein, 1.88 – 15µM are significant at both time points. At Day 6 the percentage of

LDH in the supernatant was reduced, corresponding with a large increase in the cell number (fig 3.1). 300µM genistein has similar LDH release to the controls at Day 2 while at Days 4 and 6 LDH release was lower than controls, and the opposite effect observed at lower concentrations of genistein.

3.4.2.2 Relative LDH release after culture in beta media with genistein.

Beta Day 2 70 Beta Day 4 * Beta Day 6 LDH 60 * *

total * * 50 to

*** 40 ** relative

30 LDH

10 Supernatant

% 0 0 1.88 3.75 7.5 15 30 300

Genistein concentration(µM)

Fig 3.6 Relative LDH released from MG63 cultures after treatment in beta media supplemented with genistein for 2, 4 and 6 days. The data is expressed as the percentage of LDH in the supernatant compared with the total LDH found in both the monolayer and the supernatant. N=3, error bars indicate SD, bars marked with * are significant (P<0.05) compared with the control.

In beta medium controls (Fig 3.6) there was an increase in the amount of cell death over time. At Day 6 the percentage of total LDH found in the supernatant was nearly 60%.

There was no change in the cell number over the experiment (Fig 3.2) indicating there was faster proliferation than in growth medium.

In genistein treated cells from concentrations 1.88 – 30µM at Day 2 relative supernatant

LDH was 50% of the total, higher than the control.

Day 4 LDH was the same as controls in all the concentrations. At Day 6 there was a reduction in relative LDH release with commensurate increase in cell number (fig 3.2) found at all treatments 1.88µM to 30µM are significantly lower than the control.

3.4.2.3 Relative LDH release after culture in dex media with genistein.

70 Dex Day 2 Dex Day 4 LDH 60 * Dex Day 6

total * 50 to * * 40 * relative

* 30 * *

LDH * * 20 *

10 Supernatant 0 % 0 1.88 3.75 7.5 15 30 300 Genistein concentration(µM)

Fig 3.7 Relative LDH released from MG63 cultures after treatment in dex media supplemented with genistein for 2, 4 and 6 days. The data is expressed as the percentage of LDH in the supernatant compared with the total LDH found in both the monolayer and the supernatant. N=3, error bars indicate SD, bars marked with * are significant (P<0.05) compared with the control.

In Dex medium controls (Fig 3.7) LDH release was lower than the growth (fig 3.5) and beta media (fig 3.6), cell number increased at Day 6 (Fig 3.3) reducing the relative LDH.

After treatment with genistein in dex media at Day 2 LDH release was similar from

1.88µM to 15µM and reduced with increasing concentration. At Day 4 the same dose response was repeated, as cell number increased (fig 3.3) relative LDH release decreased.

By Day 6 cell number increased significantly (fig 3.4) and relative LDH release reduced commensurately, the dose response persists at Day 6.

3.4.2.4 Relative LDH release after culture in Osteo media with genistein.

Osteo Day 2 70 Osteo Day 4 LDH 60 Osteo Day 6

total * 50 * to * * * 40 *

relative * * * 30 * * LDH

10 Supernatant

% 0 0 1.88 3.75 7.5 15 30 300 Genistein concentration(µM)

Fig 3.8 Relative LDH released from MG63 cultures after treatment in osteogenic media supplemented with genistein for 2, 4 and 6 days. The data is expressed as the percentage of LDH in the supernatant compared with the total LDH found in both the monolayer and the supernatant. N=3, error bars indicate SD, bars marked with * are significant (P<0.05) compared with the control.

In osteogenic media (Fig 3.8) control cultures were very similar to dex media, cell number increased at Day 6, reducing the relative LDH release.

After 2 Days treatment with genistein in osteogenic media relative LDH release was higher than controls for all concentrations, lowest at 300µM. By Day 4 there was a larger cell number in the monolayers (fig 3.4) and relative LDH release drops commensurately.

At Day 6 cell number is much larger and relative LDH release is low, almost inverse of the Day 2 effect, 1.88 – 30µM is lower than the control and 300µM is not significantly different.

3.4.2.5 Summary of the relative LDH release after culture with genistein.

Cell death occurs in all the supplemented media types and in all the controls. In the media without Dex added, i.e. GM and beta (3.4.2.1 and 3.4.2.2), the relative release of LDH was similar until proliferation increased and consequently relative LDH was suppressed.

In cultures where no increase in cell number occured; i.e. the growth medium control, growth medium with 300µM and the beta medium control, the relative LDH release does increase, but not significantly from Day 4 to 6. In the beta media relative LDH release is similar within concentrations at Days 2 and 4, decreasing at Day 6 due to an increase in cell number. Therefore the amount of LDH released from non-dex cultures is relatively similar across time points.

In dex containing media, i.e. dex and osteogenic, proliferation occurs more quickly.

Observations show an increase at Day 4 (fig 3.3-3.4) so there was a greater decrease in relative LDH release. In 300µM concentrations of genistein LDH release does not change significantly over time, but the cell number does increase, suggesting a greater amount of cell death.

3.4.3 Measurements of ALP activity after treatment with genistein.

Alkaline phosphatase (ALP) is a well known and characterised marker of osteogenic differentiation, monitoring activity of ALP has been shown to be a determinant of cell fate (Zipori 2006). ALP is an enzyme that is predominately found in matrix vesicles during mineralisation of extracellular collagen. Units of ALP activity are measured in the quantity of enzyme that can change 1µmol of PNPP to PNP in 1 minute. Since the activity of ALP is relative to the cell number total, protein levels are used to adjust the activity. Therefore ALP units are Umg-1 protein.

Levels of ALP were measured using the methods described in 2.2.11 and 2.2.12

3.4.3.1 ALP activity after culture with genistein in growth media.

3 GM Day 2 GM Day 4 * 2.5 GM Day 6

2 * * Protein) ‐ 1 1.5 * * * * * * * * * * * (Umg 1 *

ALP * *

0.5

0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.9. ALP activity in MG63 monolayers cultured in growth media supplemented with genistein for 2, 4 and 6 days. ALP is expressed as U mg-1 protein, where 1 unit is the amount of enzyme that converts 1µmol of substrate to product per minute adjusted for total protein. N=3, error bars indicate SD. * indicates significance when compared with controls P<0.05.

In growth medium controls (fig 3.9) there was a reduction in activation of ALP over time, by Day 6 activity was almost non-detectable.

All of the genistein treatments and time frames studied produced significantly greater activation than the controls. Relatively ALP activity was similar at all concentrations of genistein and time points.

3.4.3.2 ALP activity after culture with genistein in beta media.

3 Beta Day 2 Beta Day 4 Beta Day 6 2.5

2 Protein)

‐ 1 1.5 (Umg 1 ALP

0.5

0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.10 ALP activity in MG63 monolayers cultured in beta media supplemented with genistein for 2, 4 and 6 days. ALP is expressed as U mg-1 protein, where 1 unit is the amount of enzyme that converts 1µmol of substrate to product per minute adjusted for total protein. N=3, error bars indicate SD. all genistein treatments are significant when compared with controls P<0.05.

After culturing in beta media ALP activity in controls (fig 3.10) declined from Day 2 to almost undetectable at Day 6.

All genistein containing experiments were significantly greater from controls. However overall there was little significant difference between the genistein treatments over all the time points tested.

3.4.3.3 ALP activity after culture with genistein in dex media

3 Dex Day 2 Dex Day 4 Dex Day 6 2.5

2 Protein)

‐ 1 1.5 (Umg

ALP 1

0.5

0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.11 ALP activity in MG63 monolayers cultured in dex media supplemented with genistein for 2, 4 and 6 days. ALP is expressed as U mg-1 protein, where 1 unit is the amount of enzyme that converts 1µmol of substrate to product per minute adjusted for total protein. N=3, error bars indicate SD. all genistein treatments are significant when compared with controls P<0.05.

When MG63s were cultured in dex medium (fig 3.11) activity reduced with time until nearly undetectable at Day 6.

After culture with all concentrations of genistein, ALP activity at Day 2 was more than double the control. At Day 4 ALP was nearly uniformly 1.5 Umg-1, with only 30µM genistein higher at 2 Umg-1. On Day 6 a dose response was discernable, decreasing ALP with increasing genistein. Activation increases from Day 4 to Day 6

At concentrations of genistein from 1.88 – 7.5µM there was a stepwise activity increase over time, at concentrations from 15µM to 300µM the peak activation was at Day 4.

3.4.3.4 ALP activity after culture with genistein in osteogenic media.

3 Osteo Day 2 Osteo Day 4 2.5 Osteo Day 6

2 Protein

‐ 1 1.5 Umg

ALP 1

0.5

0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.12 ALP activity in MG63 monolayers cultured in osteo media supplemented with genistein for 2, 4 and 6 days. ALP is expressed as U mg-1 protein, where 1 unit is the amount of enzyme that converts 1µmol of substrate to product per minute adjusted for total protein. N=3, error bars indicate SD. all genistein treatments are significant when compared with controls P<0.05.

The activity of ALP in fully osteogenic medium controls (fig 3.12) started low and reduced until Day 6. With the addition of genistein to osteogenic media ALP activity increases over time in a similar manner to the dex media; however peak activities are greater in osteogenic media. 1.88µM and 3.75µM of genistein activate most strongly and with the greatest increase between time points. 300µM does not increase as strongly as the other concentrations and activity reduces from Day 4 to Day 6.

3.4.3.5 Summary of the activation of ALP with the addition of genistein.

In all control media tested ALP activity decreases over the course of the experiment. In media without added dex ALP activation occurs to the same degree at all genistein concentrations and time points. In media with added dex; i.e. dex and osteo media,

74 activation is greater at lower concentrations of genistein and increases at each time point, at high concentrations of genistein ALP activity peaks at Day 4.

3.4.4 Measurements of total collagen production after treatment with genistein.

Collagen is an important precursor for mineralisation, production of collagen is one of the earliest markers activated during differentiation to osteoblastic cells. Collagen was measured using the method described in 2.2.10.4.

3.4.4.1 Collagen production after culture with genistein in growth media.

200 GM Day 2 180 GM Day 4 GM Day 6 160 140 120 * (µg) 100 80 * * * *

Collagen * 60 * 40 20 0 0 1.88 3.75 7.5 15 30 300 Genistein (µM) Fig 3.13 The quantity of collagen observed after MG63 monolayers were cultured in growth media supplemented with genistein for 2, 4 and 6 days. N=8, error bars indicate SD. All treatments are significantly less than controls, * indicates significant increase over previous time point. P<0.05.

Growth media controls (Fig 3.13) deposit more collagen than in the genistein supplemented wells an increase in the controls was observed between Days 2 and 4 and there was no change in deposition after Day 4.

In growth media supplemented with genistein there was a slight upward trend over time in culture, significant at day 4 in most cultures.

3.4.4.2 Collagen production after culture in beta media with genistein.

200 Beta Day 2 180 Beta Day 4 Beta Day 6 160 140 120

(µg) * 100

* * 80 * * * * * Collagen * * * 60 40 20 0 0 1.88 3.75 7.5 15 30 300 Genistein (µM)

Fig 3.14 The quantity of collagen observed after MG63 monolayers were cultured in beta media supplemented with genistein for 2, 4 and 6 days. N=8 error bars indicate SD. All treatments are significantly less than controls, * indicates significant increase over previous time point P<0.05.

In beta media controls (fig 3.14) there was an increase in collagen deposition over time in culture.

After genistein treatment there was an increase in collagen at all the concentrations and at all the days tested, although levels remained below the deposition in the controls.

3.4.4.3 Collagen production after culture with genistein in dexamethasone media.

200 Dex Day 2 Dex Day 4 180 Dex Day 6 160

140 * 120 §

(µg) § § § § § § 100 § § † † † † † † 80 Collagen † † † † † 60 40 20 0 0 1.88 3.75 7.5 15 30 300 Genistein (µM)

Fig 3.15 The quantity of collagen observed after MG63 monolayers were cultured in dex media supplemented with genistein for 2, 4 and 6 days. N=8 error bars indicate SD. † denotes significant decrease when compared to controls, * indicate significant increase compared with control. § indicates significant increase over previous time point. P<0.05.

In dex media controls (fig 3.15) collagen deposition increased from Day 2 to 4 and remained the same from Day 4 to 6.

After treatment with genistein supplemented media at Day 2 there was a similar level of collagen at all the concentrations of genistein except 300µM which was higher than the other concentrations. At day 2 no treatments were higher than the controls.

After treatment with all genistein concentrations collagen increased from Day 2 to Day 4.

At Day 6 a further increase occurred at ≥15µM while at ≤7.5µM there was no increase from Day 4.

3.4.4.4 Collagen production after culture with genistein in osteogenic media.

200 §* Osteo Day 2 180 Osteo Day 4 Osteo Day 6 160 § * *§ *§ *§ * * 140 * 120 * * (µg)

§ § 100 80 Collagen 60 40 20 0 0 1.88 3.75 7.5 15 30 300 Genistein (µM)

Fig 3.16 The quantity of collagen observed after MG63 monolayers were cultured in osteogenic media supplemented with genistein for 2, 4 and 6 days. N=8 error bars indicate SD. † denotes significant decrease when compared to controls, * indicate significant increase compared to control. § indicates significant increase over previous time point. P<0.05.

In osteogenic media controls collagen accumulation increased from Day 4 to 6. In genistein treatments on Day 2 there was an increase in collagen commensurate with an increase in genistein concentration, significant at 30µM and 300µM. At Day 4 total collagen was similar from 1.88µM-15µM, while at 30µM and 300µM there was a large increase in collagen.

After 6 days culture in osteogenic media supplemented with genistein all concentrations induced a significant increase in collagen when compared with controls.

3.4.4.5 Summary of collagen production after culture in media containing genistein.

After genistein treatment in media without dex (GM, beta) collagen deposition in all the experiments was lower than found in the controls. Although lower, similar stepwise collagen accumulation did occur at all concentrations of genistein over the course of the experiment.

In media that contained dex (dex, osteo) some of the genistein concentrations did induce collagen deposition above controls, most often collagen increased with time in culture.

30µM and 300µM genistein induced the greatest increase in collagen.

Therefore it would appear there is a synergistic effect of genistein and dex that effects differentiation to a greater degree than either acting alone.

3.4.5 Alizarin Red staining for quantification of mineralisation.

Alizarin red forms a salt with calcium ions and is therefore more selective than a Von

Kossa stain which stains phosphates. The Alizarin red stain has been described as a quantitative method by Gregory et al. (2004) and an adaptation of their method was used here.

The method for alizarin red staining and quantitative determination of the mineralized

ECM can be found at 2.2.10.1

3.4.5.1 Alizarin red staining after culture with genistein in growth media.

200 GM Day 2 GM Day 4 180 GM Day 6 160 * * * * 140 * (µM)

120 Red 100 * * * * * * 80

Alizarin 60 40 20 0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.17 The quantity of Alizarin red staining observed after MG63 monolayers were cultured in growth media supplemented with genistein for 2, 4 and 6 days. N = 3 error bars indicate SD. * = significance compared with controls P<0.05.

Extra cellular matrix calcium staining in growth medium control cultures increases slightly between Day 2 and 4 and then remained similar over the rest of the experimental period (fig 3.17).

In the genistein treatments at Day 2 mineralisation was less than the control and not different from each other. At Day 4 mineralisation was the same as the control in all the concentrations of genistein. On Day 6 mineralisation was greater than the control between 1.88µM and 7.5µM, while 300µM was lower than control values, a reduction from Day 4.

In cultures containing concentrations of genistein between 1.88µM and 15µM there was an increase in mineralisation over the experimental time period.

3.4.5.2 Alizarin red staining after culture with genistein in beta media.

200 Beta Day 2 Beta Day 4 180 * Beta Day 6 160 * * * * 140 * (µM) 120 * * * *

Red * 100 * * 80

Alizarin 60 40 20 0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.18 The quantity of Alizarin red staining observed after MG63 monolayers were cultured in beta media supplemented with genistein for 2, 4 and 6 days. N = 3 error bars indicate SD. * = significance compared with controls P<0.05.

When mineralisation was observed in MG63s cultured in beta medium (Fig 3.18) there is no change in alizarin red staining of the control over time.

In the genistein treatments Day 2 was lower than controls and not different when compared with each other. At Day 4 mineralisation was higher between 1.88µM and

15µM. At Day 6 mineralisation was higher than controls at 1.88µM to 15µM and lower at 300µM.

3.4.5.3 Alizarin red staining after culture with genistein in dex media.

Dex Day 2 200 Dex Day 4 * * 180 Dex Day 6 * * 160 * * * * 140 (µM)

120 Red

100 * 80

Alizarin 60 40 20 0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.19 The quantity of Alizarin red staining observed after MG63 monolayers were cultured in dexamethasone media supplemented with genistein for 2, 4 and 6 days. N=3, error bars indicate SD. * = significance compared with controls P<0.05.

Dex medium alone (Fig 3.19) induced no change in the mineralisation of MG63s over the time course of the experiment. Similarly in genistein treated MG63s at Day 2 there was no difference between the concentrations of genistein tested or with the control.

At Day 4 in all genistein concentrations mineralisation occurred above controls, significantly at 1.88 µM to 7.5µM, but not different from each other. On Day 6 mineralisation increased again between 1.88 and 30µM, while 300µM was reduced when compared with controls.

3.4.5.4 Alizarin red staining after culture with genistein in osteogenic media.

Osteo Day 2 200 Osteo Day 4 Osteo Day 6 180 * * * * * * * * * 160 * 140

(µM) 120

Red 100

Alizarin 60 40 20 0 0 1.875 3.75 7.5 15 30 300 Genistein concentration (µM)

Fig 3.20 The quantity of Alizarin red staining observed after MG63 monolayers were cultured in osteogenic media supplemented with genistein for 2, 4 and 6 days. N=3, error bars indicate SD. * = significance compared with controls P<0.05.

In osteogenic medium the controls accumulate mineralisation over the course of the experiment. After treatment with genistein at Day 2 mineralisation was higher at low genistein ≤15µM. At Day 4 mineralisation was the same as Day 2 for concentrations of genistein from 1.88 to 15µM, at 30 and 300µM there was an increase in mineralisation from Day 2. On Day 6 there was equal mineralisation for all genistein concentrations.

3.4.5.5 Summary of the effect of genistein on alizarin red staining.

Generally at genistein concentrations between 1.88µM and 15µM mineralisation above control levels accumulated over the course of the experiment. In osteogenic media mineralisation occurred to a detectable degree over the studied time period, however addition of genistein accelerated the observed effects.

3.5 Results summary.

3.5.1 Cell death.

LDH data is complicated by growth and cell death in the intervening period between the start of the experiment and the assay. Because the data is a net figure comparing monolayer and supernatant availability it can be more informative than a cell count, monolayer number may be unchanged but a large supernatant availability of LDH suggests cell death occurred that was masked by proliferation.

Therefore genistein at the tested concentrations has an effect on cell proliferation and death. Cell death in the vehicle controls increased over the course of the experiment in all the media formulations. In the growth and β-glycerophosphate media with genistein treatments relative LDH release increased from Day 2 to Day 4 and decreased from Day

4 to Day 6, suggesting an even amount of cell death masked by proliferation between

Days 4 and 6. After culturing in dex and osteogenic media observations show a shorter delay prior to proliferation.

3.5.2 ALP activity

ALP activity in this chapter has been shown to be relatively low for controls, in keeping with MG63 ALP expression reported in the literature (Masters, 1998). MG63s can be induced to increase ALP by genistein. The data presented here for growth media and beta media shows genistein mediated increase of ALP compared with controls at Day 2, activity remains constant at all the time points and concentrations tested. Dex and osteogenic media containing genistein induces activation increases at each time point.

Genistein at 1.88 and 3.75µM are the most potent activators studied in these media.

3.5.3 Collagen synthesis.

While trends of increasing deposition over time are observable in non-dex containing media (GM and beta) collagen synthesis was significantly lower than controls. In dex containing media (dex and osteo) observations of genistein treatments show a significant collagen increase at high concentrations of genistein (30µM and 300µM), and at all concentrations by Day 6 when compared with the controls.

3.5.4 Mineralisation.

In control MG63 cultures alizarin red staining of mineralisation was similar across the time period studied, in osteogenic media an accumulation of mineralisation occurred.

For genistein treatments mineralisation accumulated over time. Genistein at 1.88µM to

15µM mineralisation increased at each time point, concentrations of genistein 1.88µM to

7.5µM mineralised most strongly. At 300µM genistein mineralisation was below control or reduced over time in culture.

3.6 Discussion

As discussed in the introduction immortalised cell lines are used in much of the literature that seeks to elucidate the effects of genistein on the differentiation of osteoblasts. Morris et al. (2006) have previously used MG63s to test the osteogenic effects of genistein.

To summarise their findings they suggested application of genistein for two days at concentrations from 2.5–30 µmol dm-3 induced MG63s to present phosphatidylserine on the outside of the cell membrane, binding calcium and increasing osteogenesis by nucleation, ALP activity was not changed and collagen synthesis was increased in non- osteogenic media and reduced in osteogenic media.

This mechanism of mineralisation is not regarded as the normal means by which mineralisation occurs. Reports suggest that although phosphatidylserine can bind Ca2+

85 directly the normal mechanism of osteogenesis is calcium dependant phosphatidylserine- annexin binding in mineralising matrix vesicles introduces calcium channels through the membrane allowing control of Ca2+ influx (Tait et al. 1992, Wuthier et al. 1992, Kirsch et al. 1997 & 2000). In fact the lipid component of matrix vesicles has been studied and it is reported by Boyan and Schwartz that phosphatidylserine is 4.8 times more concentrated in the matrix vesicle than on the plasma membrane (Boyan et al. 2009, Golub 2009).

Other matrix vesicle components and collagen have been shown to combine with amorphous calcium phosphate and regulate its formation into a precisely controlled crystalline hydroxyapatite structure (Gajjeraman et al. 2007). It is certain that cell mediated calcification in matrix vesicles is tightly regulated by highly conserved processes, for example, it is known the loss of the matrix vesicle protein ALP in vivo results in loss of mineralisation and neonatal death (Fedde 1999).

The mechanism for phosphatidylserine reversal from cytoplasmic membrane lipid to external presentation is caspase mediated and regarded as one of the first signs of apoptosis (Martin et al 1996). Phosphatidylserine may directly cause phagocytosis or may work indirectly by binding other signalling agents (Ravichandran et al. 2007). A review of the role of phosphatidylserine receptors in apoptosis by Hart et al (2004) summarises the view that serum opsonins such as antibodies and complement components can bind to apoptotic cells, presenting phosphatidylserine, and mediate phagocytosis by “traditional” phagocytic mechanisms without the need for specific phosphatidylserine responsive phagocyte receptors.

Therefore phosphatidylserine is involved in mineralisation but not necessarily in apoptosis, however the precise mechanisms for both mineralisation and apoptosis are contentious, many theories are under investigation. Two strains of phosphatidylserine receptor deficient mice have been developed, neither study reported bone defects. PSR-/-

86 is a lethal neonatal mutation effecting the development of some tissues; phosphatidylserine and its receptors may mediate tissue specific functions (Kunisaki

2004, Li 2003).

Several tissue specific phosphatidylserine receptors have been discovered that can regulate cell engulfment in the brain, spleen, lymph, liver and kidney, so far no bone specific PSR have been found (Park 2007, Miyanishi 2007, Ichimura 2008). So although phosphatidylserine might mediate cell engulfment in some tissues, it is enriched in mineralising matrix vesicles and has a function regulating Ca2+ concentration

Morris et al observe that phosphatidylserine presentation occurs at 2.5µM while nuclear condensation is detectable with TEM at >10µM genistein, but offer no assay to quantify the degree of DNA fragmentation at lower concentrations. Although it seems available they did not include data showing MG63s positive for phosphatidylserine but negative for propidium iodide; using a membrane integrity study over time would provide definitive proof for either necrosis or apoptosis. This is necessary because pathogenic mineralisation of extracellular collagen can occur during PCD and does not require ALP or annexins (Kirsch 2003).

Measurements of ALP in Morris et al were noted to be effected by high protein levels caused by collagen deposition as cell number decreased. They reported that the activity of

ALP remained the same even though cell number was falling with increasing genistein.

An increase in protein caused a relative decrease in ALP IUmg-¹ of protein. Therefore while the paper by Morris et al. 2006 demonstrates phosphatidylserine presentation on the external membrane surface it may be present due to apoptosis or release of mineralising matrix vesicles from cell membranes they do not present any other evidence of genistein mediated osteogenesis.

There are several differences between the model employed by Morris et al and the one presented here. Firstly the model in the paper utilises proliferating cells which have been cultured prior to the experiment in osteogenic media containing; EMEM, +10% FCS,

250µM ascorbic acid and 10mM β-glycerophsophate (Sandrini 2005). The data presented in this chapter uses MG63s that have been maintained in culture in non-osteogenic media,

αMEM +10% FCS. For the experimental media, 50µM ascorbate-2-phosphate was added as well as either or both 10mM β-glycerophosphate and 10nM dexamethasone.

Dexamethasone is a potent osteogenic agent that acts on differentiation pathways separate from the action of genistein (Terkeltaub 2008, Tait 1992). This formulation of osteogenic media is used more widely in the current literature (Decaris and Leach 2010, Brey 2011).

A differentiating cell model, rather than a differentiated model may behave in a completely different manner to inducers of osteogenesis (Slootweg et al. 1996).

In this chapter it was undertaken to elaborate these findings by using a confluent culture system. This system was selected to try to limit the effect of genistein on proliferation and cell death. The Day 2 collagen data presented here is consistent with that presented in the paper. When we compare the paper with the ALP measurements shown in this chapter all results presented here are significantly higher than controls and display a greater increase in activity at Day 2. This more pronounced effect is either; due to the difference in reporting the units of ALP, lower protein production, or because the change to a confluent differentiating system changed the influence of ER-genistein ligands on differentiating cells, rather than a pre-differentiated model.

In 2005 Kim et al demonstrated that 10-9M genistein increased MG63 proliferation in non-confluent conditions by 153.1% and at 10-6M inhibited proliferation vs controls

(61% of control) after 5 days. This proliferative effect was reduced to control values with

88 the addition of a competitive inhibitor, the pure anti-oestrogen ICI 182,780 (trademarked as Fulvestrant and Faslodex) that due to its pure anti-oestrogenic nature had no effect on its own. This data was used to prove that in MG63s genistein binds to the oestrogen receptor and that the concentrations of genistein tested here should inhibit the proliferation of MG63s. Depending on the time in culture the data presented here either agrees or disagrees with Kim et al., at time points earlier than 5 days we observe increased proliferation most often. However after 5 days, relative to controls, cell number is reduced in our culture system.

Whilst there is little other available record of MG63s treated with genistein other cell lines have been exploited to test the effects of genistein. KS483, a differentiating mouse cell line that can form both osteoblasts and adipocytes was cultured with daidzein by

Dang et al. (2004). Their findings show an increase in ALP activity at low doses (1-

20µM), as well as concurrent nodule formation and inhibition of adipocyte formation. At high concentrations this situation was reversed (30-75µM), ALP and mineralisation was inhibited and the differentiation of adipocytes increased.

Schmidt et al (2007), investigated genistein as a topoisomerase inhibitor in several malignant human glioma lines at concentrations between 5-100µM. EC50 was between

25-80 µM associated with cell cycle arrest at G2/M, decreasing proliferation combined with an increase in LDH release, no differentiation was detected. Higher concentrations resulted in apoptosis independent of caspase induction, although this effect was tissue type specific and is affected by expression of ERs. Cells without topoisomerase are resistant to this genistein effect (López-Lazaro 2007).

Kang (2001) report that in the immortalised rat hepatic stellate cell line HSC-T6 proliferation and collagen type I synthesis was inhibited by 38.1% after 48 hour

89 incubation of with 35µM genistein. It was demonstrated by Heugarten et al. (1999) that estradiol can suppress the expression of collagen I by modulating the expression of c-fos and regulation of the transcription factor AP-1 via the MAP kinase cascade. The MAP kinase cascade has been previously shown to be rapidly activated by oestrogen (1nM-

1pM) in the rat osteoblast cell line ROS 17/2.8 (Endoh et al. 1997).

Uchiyama & Yamaguchi (2007) treated sub-confluent mouse MC3T3-E1 osteoblasts with

10µM genistein in β-glycerophosphate medium and found increased proliferation, protein synthesis and ALP but not collagen or mineralisation at 1, 2 and 3 days. The same authors report (2008) that treatment of MC3T3-E1 with oestrogen caused no increase of

ALP at day 3, but 106M genistein did up-regulate ALP. Cbfa-1 was reported to be upregulated by oestrogen but not genistein.

In agreement with Uchiyama’s (2007) mouse cell line but not Kang’s (2001) rat the collagen levels in growth media in this study are similar across all the tested concentrations of genistein and mineralisation levels are static Day 2 and 4 as well, however on Day 6 mineralisation did increase compared to controls.

Much work has investigated the mechanism of action of oestrogens on human osteoblast like cell lines, U2OS cells have often been transiently transfected with oestrogen receptors to further understand the mechanisms involved in the differentiation of osteoblasts. Salvatori et al (2009) described the effects of both oestrogen (10nM) and genistein (1µM) on such a model. Their findings indicate that oestrogens cause the cessation of cell growth at G1/G0 and induce apoptosis; they also observed increased differentiation in arrested cells.

Much of the variation in responses to genistein in the literature can be explained by the use of many different cell lines. Genistein is certainly reported as toxic in the range of

90 concentrations tested in this chapter, however in this range there is also increased differentiation and even selection of differentiated phenotypes. These different effects are at least partly due to the effects phytoestrogens exert as SERMs. Therefore different tissue types and organisms will be affected relative to their expression of ERs and the affinity of the species ER for genistein.

Osteosarcoma and immortalised cell lines have, by definition, an unusual cell death mechanism, since genistein in this concentration range has a toxic effect a primary differentiating cell should used to investigate effects. It is preferable to use a human cell strain since Rat and Mouse cell lines have different responses to genistein (Kang et al.

2001, Uchiyama et al. 2007). A primary human BMSC strain has a distribution of oestrogen receptors that is more relevant to an in vivo situation and is more likely to undergo cell death in a normal manner. Therefore in the next few chapters the use of a differentiating primary human cell type is explored to provide a more realistic model.

3.7 Conclusions:

The hypothesis that genistein accelerates the osteogenic differentiation of a MG63 culture was proved to be correct. The range of genistein concentration utilised in these experiments was found to have relevant effects on the confluent cell model.

The confluent MG63 osteosarcoma cell line described here was useful to describe the action of genistein on markers of differentiation alone, and in combination with components of osteogenic medium after short culture periods.

The use of the MG63 cell line was not thought to offer the most precise tool for the extrapolation of genistein effects from in vitro to in vivo. Therefore a primary cell model was investigated for further study.

CHAPTER 4

AN IN VITRO EVALUATION OF THE OSTEOGENIC EFFECT OF

GENISTEIN ON A CONFLUENT DIFFERENTIATING BMSC MODEL.

4.1 Introduction

Bone marrow stromal cell (BMSC) cultures, as described in the introduction contain osteogenic progenitors. Since genistein has been shown to up regulate markers of osteogenesis in MG63s understanding the in vitro effect on primary isolates of BMSCs would provide further evidence of the likely in vivo effect. Therefore in this chapter the application of genistein to primary human multipotent BMSCs was investigated as a confluent differentiating model. This primary differentiating model is more likely to be relevant to real osteogenesis than an osteosarcoma model.

Since the initial report by Friedenstein describing what became known as the BMSC population the development of techniques to isolate and characterise BMSCs has been well studied with no consensus (Friedenstein et al 1970, Caplan 1991). Authors have refined the Friedenstein method to select proportions of the whole bone marrow population by density gradient, collagen treatment, sieving and phenotypically by surface epitope sorting using magnetic bead or fluorescence-activated techniques (Castro-

Malaspina et al 1980, Sakaguchi et al 2004, Leong et al. 2004, Bianco et al. 2008, Yu et al. 2010). All of these techniques isolate populations of cells that are distinct at isolation but, at least in vitro, share the stem cell property of pluripotency.

There is no consensus for the isolation of BMSCs because no single marker has been discovered that is constitutively expressed; they are phenotypically plastic and share features with many other cell types including endothelial, epithelial and muscle cells,

their isolation has been reported from every vascularised tissue type and their function in vivo is unknown (Tárnok et al. 2010). In the last 40 years of research there has been no significant gain in a definitive technique to select the BMSC population from the in vivo environment, making extrapolation of in vitro effects to an in vivo setting difficult.

Because of the overall lack of consensus non-comparable data can be the result, both between the array of culturing methods and from culture to organism.

Therefore after each isolation the selected population must be fully characterised for its individual capabilities of multi-potentiality and reproduction. Currently in the literature many authors use antibody screening to concurrently select for and characterise their cell population and to this end many antibody panels have been developed that select interesting sub-populations (Pountos, et al. 2007). A promising marker of a subset of bone marrow stromal cells is MCAM (CD146), a marker of classically described adventitial reticular cells from the stroma (Cattoretti et al. 1993). MCAM is a cell adhesion molecule also expressed by pericytes residing in the microvasculature, it has been postulated that pericytes could be in vivo BMSCs although they are spontaneously myogenic in vitro and are not osteogenic in vivo (Shi and Gronthos 2003, Dellavalle et al.

2007, Feng et al. 2010).

The international society for cellular therapy (ISCT) has proposed minimal markers for the selection of BMSCs. According to the ISCT the BMSC should be; plastic adherent, positive for surface epitopes CD73, CD90 and CD105 and negative for CD34, CD45,

HLA-DR, CD11b or CD14, CD19 or CD79α and also capable of differentiation to osteoblasts, chondrocytes and adipocytes (Horwitz et al. 2005, Dominici et al. 2006).

Although a basic level of characterisation in the literature is welcome there are some problems with the ISCT definition; it does not recommend the classic method of initial

seeding at clonal density. Perhaps most importantly it does not uniquely recognise the

BMSC population; Sabatini et al (2005) reported that bronchial fibroblasts are indistinguishable from BMSCs by the above criteria.

The ISCT definition does not require analysis of the culture to determine if any cells within it have the self renewal and differentiation capacity of a stem cell. Single cell cloning to analyse the number of self renewing cells that are tri-potent is a fundamental of the Friedenstein method. Also not recommended is the final pillar in the Friedenstein method, proof of in vivo bone formation ability from a cloned single cell, a demonstration of true stem cell behaviour.

The ISCT only recommends an initial characterisation of the cell population resulting from the selection criteria but that “…each cell preparation does not need to be re- evaluated.” BMSCs lose differentiation capacity over time in culture, in order to provide appropriate controls re-evaluation of the capacity for pluripotency should be an essential step at every passage prior to experimental use. Since initial CFU-F formation ability at isolation varies between authors and with donor age and sex, culture propensity for CFU-

F and single cell cloning determination is a vital characterisation step (Kuznetsov et al.

2009). The CFU-F formation capacity has been shown to directly relate to the number of cells in the culture that have in vivo stem cell capacity (Friedenstein et al. 1974). The lack of even a mention of CFU-F characterisation and differentiation capacity before experimental use makes the ISCT criteria misleading and potentially useless (Kuznetsov et al. 2009).

With regards to the ISCT ruling it’s worth noting that in the literature it has been reported that cultures of some BMSC isolates from adipose tissue are CD34 positive (Yoshimura et al. 2006). Endothelial cells (positive for CD34 and CD31) cultured with BMSC media

and seeding densities become positive for the ISCT criteria (da Silva Meirelles et al.

2009).

There is evidence available that suggests isolation of plastic adherent cells in selective conditions may represent a homogenous culture. The use of some batches of FBS in isolation media can produce cell isolates that after culture selection alone individually display similar expression of 48 known BMSC markers (Lennon et al 1996, Pittenger et al 1999). Therefore heavy reliance on the use of selection by marker panel does not guaranty adequate characterisation or even selection of the correct cell type; depending on the tissue cells were isolated from, the seeding densities and media composition

(Dominici et al. 2006, Meirelles et al. 2006, Kuznetsov et al. 2009).

It has been controversially reported that dedifferentiation and transdifferentiation can be induced to occur in BMSC populations when they are cultured in specialised environments, although the potential of single BMSCs to differentiate into tissues other than those of skeletal origin has not been formally proven in vivo (Krabbe et al. 2005,

Martin et al. 2009, Roobrouck et al. 2011).

There are also different reports of culture limits for BMSC growth without differentiation, most authors point to about 20 population doublings before bulk cultures loose capacity for differentiation (Kozhevnikova et al. 2008). However, the Verfaille group reports that a subpopulation of BMSCs, which they call multipotential adult progenitor cells (MAPCs), can be cultured almost indefinitely, certainly to above 120 population doublings (Serafini and Verfaillie 2006). It is the definition of the term stem cell that it has a large capacity for self renewal with maintenance of potential, therefore it is expected that CPDs should be large or that current culture conditions initiate differentiation.

This culture induced plasticity of human BMSCs has made it difficult to extrapolate the in vitro behaviour to an in vivo niche, in vivo function is also elusive; in other stem cell literature there is consensus for characterisation. The gold standard proofs of stem cell function are based on work on haematopoietic stem cells (HSC). For this proof HSC scientists utilise a single ex vivo HSC to serially reconstitute multi-lineage haematopoiesis in irradiated mice (Kim et al. 2010). This assay establishes several important things; there are no in vitro steps, in vivo assessment is critical, multipotentiality is based on data from the single cell level and serial self renewal indicates that there are identical daughter cells with identical properties propagating over extremely large numbers of population doublings (Kim et al. 2010).

This definition might not hold true for skeletal stem cells because even at the very outset of non haematopoietic stem cell research McCulloch and Till, using single cell clones, found de novo bone speck formation in the spleen (Becker et al. 1963). Indicating that unlike HSCs, BMSCs do not home to their original niche and that to repopulate a skeletal defect BMSCs are required in greater numbers than HSCs. This difference may well be due to the longevity of the tissue type, bone has a low turnover relative to haematopoiesis. Ideally an in vivo experiment replicating the precision of the HSC experiments would be required for definitive stem cell proofs.

Therefore, in conflict with the ISCT and directly because of the limitations of antibody screening any work concerning primary BMSCs should, until the in vivo niche is extrapolated, include characterisation methods that undertake to prove in vitro self renewal with concurrent maintenance of pluripotency, CFU-F determination or single cell cloning is still the gold standard method to determine this (Kuznetsov et al. 2009,

Sarugaser et al 2009). Cloning allows determination of the number of cells within the culture that have self renewal potential and removes the possibility that the culture

contains multiple lineages that provide the illusion of multipotentiality (Friedenstein et al.

1974, Kuznetsov and Robey 1996). Because the number of pluripotent self renewing cells in the culture is low it is likely that the population of cells, although initially homogenous for the selection criteria, is heterogeneous in age and functionality (Sarugaser et al 2009).

Once single cell clones (SCCs) have been shown to be pluripotent and capable of in vitro self renewal the culture has been shown to contain a proportion of in vitro stem cells.

However, the last step to fully characterise the potential of the isolated population, and to overcome the issue of in vitro plasticity, is by an in vivo experiment. In order to achieve this full characterisation of the population the multipotent SCCs can be placed into in vivo chamber models; those cells that mediate the production of de novo bone are true skeletal stem cells. Only about 10% of those human in vitro cells capable of self renewal and pluripotency are true in vivo stem cells (Bianco et al. 2006). It is this small population of skeletal stem cells in the BMSC culture that is commonly referred to as the

“mesenchymal stem cell”, although there is still no proof of its in vivo function or differentiation to all mesenchymal tissues. Although some of the culture may have stem cell functionality the whole culture is heterogeneous for potential and should not be described as a population of stem cells.

Most publications in the literature do not use the final proof of in vivo de novo bone formation due to implications of cost, ethics and time. Perhaps a more important issue is that subsequently there will be changes in the population and further sub culturing would require a repeat in vivo experiment. Changes in the cultured cells can occur no matter how extensively characterised, re-affirming phenotype takes time and uses cells that could be spent experimentally. However, the extensive characterisation of the population is extremely important for the reproducibility of data between cultures and for analysis of data compared to robust controls.

In this chapter the STRO-1 and SSEA-4 antibodies were used for further characterisation of the isolate. The STRO-1 antibody is the classical BMSC antibody that has been used extensively as either a characterisation or isolation marker of BMSC populations; however as with the other BMSC markers it cannot be used alone as it is expressed by a variety of cell types that includes BMSCs (Simmons and Torok-Storb 1991, Salem and

Thiemermann 2010). STRO-1 positive sorting by FACS increases the number of clonogenic cells in the culture; since at isolation cultures were seeded at clonal density it is likely that this selection criteria also concurrently increased STRO-1 positivity in the culture (Sacchetti et al. 2007).

Much of the available isolation and characterisation literature based on the production of an antibody marker panel for BMSCs does not include the final in vivo testing of clones characterisation. However of the recent papers that do undertake in vivo testing the most complete is the paper by Gang et al (2007) concerning the use of SSEA-4 as a marker of

BMSCs. Therefore SSEA-4 was also utilised here as a marker of BMSCs, in combination with STRO-1, initial seeding at clonal density, adherence to tissue culture plastic, single cell cloning and pluripotent differentiation.

After the described characterisation the primary human BMSCs were used experimentally to determine the effect of genistein on osteogenic differentiation markers, as described.

4.2 Hypothesis

Primary bone marrow stromal cells isolated from discarded human tissue differentiate to an osteogenic phenotype in an accelerated manner after the addition of genistein. The osteogenic effect of genistein is observable in growth medium, osteogenic medium, and in medium containing osteogenic components.

4.2.1 Aims

Isolation of BMSC cultures by limiting dilution and adherence at clonal densities to tissue culture plastic. Characterisation of the primary cell strains for cell surface epitopes, in vitro proliferation potential, differentiation capacity and morphology was undertaken to achieve a characterised and confluent in vitro model. These BMSC cultures were observed for production of ALP, collagen, mineralisation and Cbfa-1 before and during osteogenic differentiation to produce a differentiation profile. Concurrent to analysis of differentiation phenotype genistein was applied to BMSC cultures in order to confirm the induction of the osteogenesis observed in MG63’s and to study the effect on proliferation, cell death, ALP, collagen, mineralisation. The effect of genistein on BMSCs was compared to MG63 observations and the literature.

4.3 Methods

Isolation and culture of BMSCs is described in full at 2.2.1 – 2.2.7. Methods for the characterisation of BMSCs are described at 2.2.9. In brief, mononuclear cells were isolated from the buffy coat and cultured at a seeding density of 5x103 cells cm-2, since

BMSCs occur at a frequency of approximately 1-50 every 1x105 cells this seeding density results in colony formation. Colonies were allowed to combine and were characterised for expression of STRO-1, SSEA-4, quantity of single cell clones and differentiation capabilities prior to culturing in media containing genistein.

4.3.1 Experimental procedure

The experimental methods for determination of effects of genistein are described in full at

2.2.13. In brief; the BMSCs were seeded at confluent density. The medium was removed and replaced with media containing genistein, harvesting subsequently occurred at Days

2, 4 and 6. Analysis methods are described in full at 2.2.10.

4.3.2 Statistical analysis.

Statistical analysis was performed in Minitab 15. The analysis of data was achieved using a one-way analysis of variance (ANOVA) with multiple comparison tests; Tukey’s family error rate was used to determine the significance between groups without a defined control. i.e. Comparisons of a single concentration of genistein across time points. Dunnett’s family error rate was used to compare groups with a defined control i.e. concentrations of genistein compared with vehicle. The confidence level was set at 95 and P values of <0.05 were considered significant.

4.4 Results

4.4.1 Characterisation of BMSC cultures.

Plastic adherent mononuclear bone marrow cells were isolated at clonal density and characterised for growth (Fig 4.1), staining for STRO-1 and SSEA-4 antibodies (Fig 4.2), and cultured in media that induced differentiation of osteoblasts (fig 4.3), adipocytes (fig

4.4) and chondrocytes (fig 4.5).

4.4.2 BMSC Growth curves in isolation and culturing medium.

Fig 4.1 The cumulative population doublings of BMSCs in isolation and culturing medium from isolation until the cell number no longer increased.

100

4.4.3 Antibody staining for the characterisation of surface epitope phenotype.

A B

C D

Fig 4.2 Photomicrographs of BMSCs after cytospin preparations were stained. Photographs show staining with antibodies for A) STRO-1 (x20), B) SSEA-4 (x40) and negative control staining for C) STRO-1 and D) SSEA-4.

BMSC cultures were stained for SSEA-4 and STRO-1 using methods described at

2.2.9.1. BMSC isolated by adherence to tissue culture plastic at clonal densities were positive to SSEA-4 and STRO-1 antibodies. Indicating that selection by classic methods also selects cells that stain for recognised antigen production and have been shown to demonstrate pluripotent differentiation.

101

4.4.4 Characterisation of the multipotentiality of isolated populations.

The pluripotency of bulk cultures was demonstrated by sub-culturing the isolates in differentiation medium for 21-30 days or until morphology and culture observations indicated BMSCs had differentiated, as shown in figures 4.3 to 4.5, Alizarin red staining showing accumulation of mineralisation during osteoblast differentiation (fig 4.3), Oil

Red O staining of cells with lipid vacuoles after adipocyte differentiation (fig 4.4) and

Alcian blue staining of glycosaminoglycans in the collagen of pellet cultured cells during chondrocyte differentiation (fig 4.5).

4.4.4.1 Characterisation of the BMSC culture for osteogenic potential

Fig. 4.3 BMSC differentiated monolayers cultured in osteogenic media stained with Alizarin red to determine mineralisation of the extra cellular matrix. A) Single mineralised nodule (x20) B) whole culture mineralisation (x10).

102

4.4.4.2. Characterisation of the BMSC culture for adipogenic potential

Fig 4.4 BMSC differentiated monolayers cultured in adipogenic media stained with Oil red O to show lipid accumulation (x20).

4.4.4.3 Characterisation of the BMSC culture for chondrogenic potential.

C A

Fig 4.5, BMSC differentiated pellets cultured in chondrogenic media stained with Alcian blue for glycosaminoglycans in the collagen matrix A) the stained pellet, B) close up, C) a section of the pellet (x20).

103

4.4.5. The differentiation potential of BMSCs with time in culture and the analysis of the culture potential for multipotent single cells.

Table 1 lists the differentiation potential of isolated cell cultures. The colony forming efficiency of the cultures was, on average 30.33 cells per 1x105 seeded mononuclear cells. After between 1 and 9 passages there was a loss in the differentiation potential of the cultures. Often the loss in potential corresponded with reduced proliferation compared to the normal doubling rate of the culture.

After the BMSCs were characterised for surface epitope production and differentiation potential the method for single cell cloning (2.2.9.2) was utilised to determine the self renewal properties of individuals within the culture, which allowed an inference of culture stem cell capacity. Table 2 outlines the expected and found numbers of self renewing cells within the plated cells. For a culture of cells 1056 wells were seeded with

352 cells at a density of 0.3 cells per well, any wells with more than one attached cell after 4 hours were discarded.

On average, compared to the literature, there was a normal distribution of colony formation from the isolated monocytes, where as a high percentage (5.3%) of single cell clones were capable of multiple lineage differentiation and more than 30 population doublings (Akintoye et al 2006, Kuznetsov et al 2009, Sarugaser et al 2009).

104

Differentiation to multiple cells types at passage Cell Strain CFU-F 1 2 3 4 5 6 7 8 9 10 B381 44 3 3 3 3 3 3 3 3 3 2 B382 20 3 3 3 2 2 2 1 1 - - B383 32 3 3 3 3 3 3 2 2 2 1 B386 36 3 3 3 3 3 3 3 3 2 2 B388 37 3 3 3 3 3 2 2 2 1 1 B389 12 2 2 1 1 ------B390 19 3 3 2 2 1 1 - - - - B392 38 3 3 3 3 3 1 1 - - - B393 52 3 3 3 3 3 3 3 3 3 3 B404 41 3 3 2 2 2 1 1 - - - B409 32 3 3 3 3 3 3 3 2 2 2 B423 51 3 3 3 2 2 2 2 2 1 1 B428 22 3 3 3 3 1 1 - - - - B438 39 3 3 3 3 3 3 2 2 2 2 B442 17 3 3 3 2 2 2 1 1 - - Table 4.1. Showing the capability of isolated cells to form colonies at isolation, /105 cells seeded, and differentiate into multiple cell lineages at each passage, loss of potential follows differentiation of the cell population tested. Only B389 was not capable of differentiation to osteoblasts, chondrocytes and adipocytes. All other isolates were capable of differentiation to all three lineages until at least passage 2.

Clones % of Wells seeded Cell Expected Returned surviving Tripotent original at 0.3 cells per strain colonies colonies at least 10 clones seeding well CPDs density B386ss 1056 352 60 48 21 5.97 B409ss 1056 352 39 31 15 4.26 B438ss 1056 352 55 47 20 5.68 Table 4.2. Table showing the single cell clones capable of in vitro self renewal and potential for multiple differentiations from three primary cell populations, the mean capacity of the populations seeded for stem cells was 5.3% ± 0.91 SDs. The probability that each returned colony originated from a single seeded cell was 95.54%.

105

4.4.6 Temporal characterisation of differentiation markers

Once established, multipotent BMSC cultures were cultured in osteogenic medium, differentiation was monitored by analysis of several osteogenic markers; collagen, ALP,

Cbfa1, mineralisation, cell proliferation and death. A profile of the expression found in differentiating BMSC cultures can be found at Fig 4.6.

Fig 4.6 The activity profile of some of the common markers used to identify the osteoblastic differentiation of BMSC. Primary BMSC were seeded into flasks at normal seeding and harvesting densities and cultured in osteogenic media for 35 days. Samples of the cultures were tested according to the relevant methods and data expressed as a percentage of the maximum value.

4.4.7. Summary of BMSC cultures.

The cultures described and utilised in this chapter were isolated from bone marrow by density gradient and adherence to tissue culture plastic at clonal densities, they are positive for known BMSC antigens. Single cell clones (SCCs) within the culture are capable of in vitro self renewal and differentiation to osteoblast, chondrocytes and adipocytes.

106

4.4.8 Osteogenic effect of genistein after culture with primary human BMSCs in several medium types for 2, 4 and 6 days.

4.4.8.1. Genistein effect on total cell number in the monolayer

LDH from the lysed monolayer was plotted against a graph of the LDH found in known cell numbers. The data presented is expressed as the cell number found in the monolayer.

The method is described at 2.2.10.5.1

4.4.8.1.1 Cell number after culture with genistein in growth media.

* * * * * *

Fig 4.7, BMSC number derived from the content of LDH extracted from the monolayer after 0, 2, 4 and 6 days culture in growth medium and growth media with genistein. N=2, error bars indicate SD. * is significant from control P<0.05.

In growth media controls (fig 4.7) cell number increased slightly over the course of the experiment. Genistein treatment media at 1.88µM induced a trend of increased proliferation compared with cells seeded at T0 over the course of the experiment.

Concentrations greater than 7.5µM induced a reduction in cell number, significant at 30 and 300µM.

107

4.4.8.1.2 Cell number after culture with genistein in beta media.

* * * * * *

Fig 4.8, BMSC number derived from the content of LDH extracted from the monolayer after 0, 2, 4 and 6 days culture in beta medium and beta media with genistein. N=2, error bars indicate SD. * significant from control P<0.05.

Beta medium induced a slight increase in cell number in the controls (fig 4.8) over the course of the experiment. 1.88µM and 3.75µM genistein supplemented beta media induced a trend of increasing cell number at each time point. Cell death was apparent at

≥7.5µM concentrations; cell death occurred significantly and at early time points after addition of genistein concentrations 30 and 300µM.

108

4.4.8.1.3 Cell number after culture with genistein in dex media.

* * * *

* * * * * *

Fig 4.9, BMSC Cell number derived from the content of LDH extracted from the monolayer after 0, 2, 4 and 6 days culture in Dex medium and Dex media with genistein. N=2, error bars indicate SD. * is significant from control P<0.05.

In dex medium controls a significant increase in cell number can be observed at Day 2 of the experiment (fig 4.9). A dose response is apparent when genistein is added to dex media, lower concentrations initiate the greatest increase in cell number at Day 2.

However the increase in cell number is not sustained over the period of the experiment.

The toxic effect of genistein apparent in growth media and beta media is delayed to 30 and 300µM of genistein in dex media.

109

4.4.8.1.1 Cell number after culture with genistein in osteogenic media.

* * *

* * * *

* * * * * *

Fig 4.10, BMSC Cell number derived from the content of LDH extracted from the monolayer after 0, 2, 4 and 6 days culture in osteogenic medium and osteogenic media with genistein. N=2, error bars indicate SD. * significant from control P<0.05.

In osteogenic medium controls (fig 4.10) cell number increased significantly at Day 2, dropping again to T0 values at Day 6. After addition of genistein to osteogenic media proliferation occurred at 1.88 – 15 µM, higher concentrations were toxic. The increase in cell number was not sustained over the experimental time period.

An increased effect on proliferation compared with the other media was observed in osteogenic media formulations.

110

4.4.8.2 The effect of genistein on LDH release compared with total LDH

Because measurements of the supernatant LDH are “free” and comparisons with the monolayer are readily available LDH was used in this study as a measurement of cell death over time in culture. The amount of LDH available in the monolayer at the end of the experiment was combined with the supernatant LDH and this was used as a measure of 100%. The LDH available in the supernatant was directly compared with the total available LDH. The methods used were as described by the manufacturer and can be found in full at 2.2.10.5.1

4.4.8.2.1 LDH release after culture with genistein in growth media.

* * *

* *

Fig 4.11, LDH content of the supernatant as compared to total LDH after BMSCs were cultured for 2,4 and 6 days in growth medium and growth media with added genistein. N=2, error bars indicate SD. * significant from control P<0.05.

Growth medium controls (fig 4.11) did not induce a change in LDH release over the experimental time period. After genistein treatment LDH release occurred at similar levels in all time points and from concentrations of gen from 0 to 15µM. At gen concentrations 30 and 300µM LDH release increased over the control values.

111

4.4.8.2.2 LDH release after culture with genistein in beta media.

* * *

Fig 4.12, LDH content of the supernatant as compared to total LDH after BMSCs were cultured for 2,4 and 6 days in beta medium and beta media with added genistein. N=2, error bars indicate SD. * significant from control P<0.05.

In beta medium controls (fig 4.12) cell death remained similar throughout the experiment.

Genistein treatment induced release of LDH into the supernatant at similar levels in concentrations of gen from 0 - 15µM across all time points. At gen concentrations 30 and

300µM LDH release increased over the control values.

112

4.4.8.2.3 LDH release after culture with genistein in dex media.

* * * * *

* *

Fig 4.13, LDH content of the supernatant as compared to total LDH after BMSCs were cultured for 2,4 and 6 days in dex medium and dex media with genistein. N=2, error bars indicate SD. * significant from control P<0.05.

In dex medium control cultures (fig 4.13) lower cell death was observed at Day 2 than in growth (fig 4.11) or beta media (fig 4.12), at Days 4 and 6 the LDH release was similar to previous observations.

At Day 2 addition of gen at concentrations between 1.88 and 7.5µM, caused a reduction in LDH release and an increase at 30-300µM when compared with controls. At Day 4 and

6 LDH release matched controls until 30 and 300 µM where an increase was observed.

113

4.4.8.2.2 LDH release after culture with genistein in osteogenic media.

* * *

* *

Fig 4.14, LDH content of the supernatant as compared to total LDH after BMSCs were cultured for 2,4 and 6 days in osteogenic medium and osteogenic media with genistein. N=2, error bars indicate SD. * significant from control P<0.05.

Osteogenic medium controls (fig 4.14) induced an increase in cell death over time in culture in a similar manner to observations of dex medium. In gen supplemented osteogenic media the protective effect from 1.88-15µM was sustained until Day 4 and at

Day 6 LDH release matched controls. LDH release at 30µM and 300µM was much higher than controls at all time points and increased stepwise over concentrations and time points.

114

4.4.8.3 ALP activity after addition of genistein to BMSC cultures.

ALP is a well known and characterised marker of osteogenic differentiation, monitoring activity of ALP has been shown to be a determinant of cell fate (Zipori 2006). Levels of

ALP were measured using the method described in 2.2.10.2 and 2.2.10.3

4.4.8.3.1 ALP activity after culture with genistein in growth media.

* * * *

* *

Fig 4.15 ALP Umg-1 protein found in growth media cultured BMSC monolayers after 2, 4 and 6 days treatment with Genistein. N=3, error bars indicate SD, * significant from control P<0.05

In growth medium controls (fig.4.15) there is no significant increase in ALP activity by

Day 6. In the gen treated cells there is little detectable activity at Day 2 rising to above control levels at Day 4. By Day 6 there is a significant increase in ALP activity. ALP at

30 and 300µM genistein was reduced compared to ≤15µM genistein.

115

4.4.8.3.2 ALP activity after culture with genistein in beta media.

* * * * * *

* *

Fig 4.16 ALP Umg-1 protein found in beta media cultured BMSC monolayers after 2, 4 and 6 Days treatment with Genistein. N=3, error bars indicate SD, * significant from control P<0.05

Beta medium controls (fig 4.16) have low observable ALP activity similar with growth medium controls. In the gen treatment cells there is no significant increase in activity at all the concentrations tested until Day 6.

On Day 6 there is a significant increase in activity of ALP above controls. There is no significant difference in activation between concentrations 1.88µm to 7.5µm. 30 and

300µM concentrations are significant compared to controls but lower than 1.88-7.5µM.

116

4.4.8.3.2 ALP activity after culture with genistein in dex media.

* * * * * *

Fig 4.17 ALP Umg-1 protein found in dex media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=3, error bars indicate SD, * significant from control P<0.05

In dex medium (fig 4.17) activation of ALP was similar to the growth and beta control media formulations, no significant differences were detected. After Gen treatment in dex media there is a general trend of rising activity from controls to 300µm, increasing from

Day 2 to Day 4. At Day 6 activation tends towards an increase with increasing concentration. When compared to the growth (fig 4.15) and beta (fig 4.16) media formulations application of gen in dex medium causes earlier ALP upregulation.

117

4.4.8.3.3 ALP activity after culture with genistein in osteogenic media.

* * * * * * * * * *

* *

Fig 4.18 ALP Umg-1 protein found in osteogenic media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=3, error bars indicate SD, * significant from control P<0.05

In osteogenic medium controls (fig 4.18) there is no significance increase in ALP activity over the time in culture.

After adding gen to osteogenic media formulations an increase in ALP activity is observed at Day 2. At Day 4 treatment with gen induces a significant increase in ALP activity at concentrations of gen from 1.88µM to 15µM, inversely proportional to concentration. By Day 6 there is a significant increase in ALP activity above controls, activation is higher gen concentrations from 1.88-7.5µM when compared with concentrations ≥15µM.

118

4.4.8.3.4 Summary of the effect of Genistein on ALP activity.

Comparing medium types it is apparent that those media containing dex and genistein have increased activity of ALP over time (figs 4.15 to 4.18), there is no difference in ALP activity in the vehicle controls. Genistein combined with dex increases ALP activity earlier when compared with media that do not contain dex.

4.4.8.4 Total collagen observed after the addition of genistein.

Collagen was measured using the method described in 2.2.10.4

4.4.8.4.1 The collagen observed after culture with genistein in growth media.

† † † † † *

Fig 4.19 Staining of collagen in growth media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=8, error bars indicate SD, * is significant from control, † denotes significant increase from Day 2 to Day 6. P<0.05

Collagen production in growth media control cultures (fig 4.19) does not significantly change over time. Significant difference from the controls only occurs with 30µM genistein at Day 6, after treatment with ≤30µM deposition increases from Day 2 to Day

119

4.4.8.4.2 Quantity of collagen observed after culture with genistein in ß- glycerophosphate media.

†

† † † †

Fig 4.20 Staining of collagen in beta media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=8, error bars indicate SD, † denotes significant increase from Day 2 to Day 6. P<0.05

In beta medium controls (fig 4.20) there is a significant difference in collagen deposition over time, from Day 2 to Day 6. Collagen in the genistein treatments is not different from the controls but is higher at Day 6 after addition of between 1.88µM and 15µM.

120

4.4.8.4.3 Quantity of collagen observed after culture with genistein in dex media.

† † * * † † † † †

Fig 4.21 Staining of collagen in dex media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=8, error bars indicate SD, * is significant from control, † denotes significant increase from Day 2 to Day 6. P<0.05

In dex containing medium (fig 4.21) the controls do increase collagen production over time in culture, from Day 2 to Day 6. After genistein treatment collagen is the same at

Day 2 and 4. At Day 6 there is an increase in deposition, highest at 30 and 300µM genistein which are also significantly compared with controls.

121

4.4.8.4.4 Quantity of collagen observed after culture with genistein in osteogenic media.

† † † † † † * * * * * * * * * * * * * *

Fig 4.22 Staining of collagen found in osteogenic media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=8, error bars indicate SD, * is significant from control, † denotes significant increase from Day 2 to Day 6. P<0.05

After growth of BMSC in osteogenic medium (fig 4.22) there is no change in the controls. In osteogenic media with genistein supplemented there is increased collagen production with increase in genistein and time.

122

4.4.8.5 Alizarin Red staining for the quantification of mineralisation.

Alizarin red collates with calcium ions and is therefore more selective than Von Kossa stain which stains phosphates. The method used here has been developed for quantitative analysis of staining, a full description can be found at 2.2.10.1.

4.4.8.5.1 Alizarin red staining observed after culture with genistein in growth media

* * *

* * * * * *

Fig 4.23 Alizarin red staining of mineralisation in growth media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=2, error bars indicate SD, * is significant from control P<0.05

In growth medium controls (fig 4.23) alizarin red staining of the mineralised ECM was the same, there was a slight increasing trend in mineralisation with time.

In the gen treatments at Day 2 levels of mineralisation were higher than controls but comparable to each other. Mineralisation increased step wise at the next two time points in 1.88–7.5µM. Day 6 mineralisation was comparable to earlier time points at concentrations higher than 15µm.

123

4.4.8.5.2 Alizarin red staining observed after culture with genistein in ß- glycerophosphate media

* * * * * * * * * * * * * * * * *

Fig 4.24 Alizarin red staining of mineralisation in beta media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=2, error bars indicate SD, * is significant from control P<0.05

In beta medium the control mineralisation (fig 4.24) of the monolayers remained similar throughout the experiment. In the genistein treatments a trend of increasing mineralisation at each time point was observed.

124

4.4.8.5.3 Alizarin red staining observed after culture with genistein in dex media

Fig 4.25 Alizarin red staining of mineralisation found in dex media cultured BMSC monolayers after 2, 4 and 6 Days treatment with Genistein. N=2, error bars indicate SD, all experimental conditions are significant from controls P<0.05

In dex medium controls (fig 4.25) levels of mineralisation were similar throughout. In the genistein treated cells mineralisation accumulated at each time point in all the concentrations studied.

125

4.4.8.5.4 Alizarin red staining observed after culture with genistein in osteogenic media

* * * *

* * * * * * * * * *

Fig 4.26 Alizarin red staining of mineralisation in osteogenic media cultured BMSC monolayers after 2, 4 and 6 days treatment with genistein. N=2, error bars indicate SD, * is significant from control P<0.05

In osteogenic medium controls (fig 4.26) there was a small increase in mineralisation at each time point. In the genistein treated cells there was a significant increase above controls, but no change with concentration of genistein at Day 2 and 4. By Day 6 a trend of increasing mineralisation with increasing genistein was apparent, ≥15µM is significantly higher than the control.

126

4.5. Discussion

4.5.1 Characterisation.

Cells were isolated for use in this chapter from bone marrow by a combination of methods based on the original publications of Friedenstein and Caplan, mononuclear cells were ficoll hypaque density gradient separated from human bone marrow and plated at clonal densities. This technique does not separate the cell population into categories of phenotypic expression, as described by the ISCT, but CFU-F selection is classically recognised as producing a BMSC culture (Friedenstein et al 1970, Colter et al 2001).

Further characterisation of the cell population is important no matter the epitope phenotype because differentiation potential is heavily related to culturing conditions and population age (Kuznetsov et al. 2009). Much research in the literature uses a known antibody panel to define the studied population and even if the antibodies used overlap there are some difficulties comparing work, it may even be found subsequently that populations have distinct properties (Buhring et al. 2007 & 2009, Arufe et al. 2010).

Characterisation of cells for epitope presentation may only become important once the in vivo niche is known and direct selection is required (Deschaseaux et al. 2003, Seeberger et al. 2008).

In this study we have used only two of the most significant antibodies to characterise the cells, they have been previously shown to select for the whole BMSC population (STRO-

1) and extremely selectively for embryonic stem cells and recently BMSCs (SSEA-4) followed by self renewal of single cell clones and differentiation to three commonly used phenotypes (Simmons and Torok-Storb 1991, Gang et al 2007). Because of cost implications and the possibility of change over time in the cell population no in vivo testing was undertaken, this is compatible with most of the recent published literature

127

which do not undertake this step either (da Silva Meirelles et al. 2006, Roobrouck et al.

2011).

Therefore as well as screening the cultures for expression of the surface markers STRO-1 and SSEA-4 two differentiation studies were under taken to improve the characterisation of the culture. The isolate was split at each experimental passage and differentiation to chondrocytes, adipocytes and osteoblasts was undertaken concurrent with further culturing, this characterisation of the BMSC culture is undertaken by almost all literature concerning “MSCs” to prove culture potential (Aicher et al 2010). It was undertaken in this chapter to allow analysis of culture potential at the beginning of any experimental methodology and therefore provides proper controls. Secondly single cell cloning was undertaken as an extension of this analysis of potential; single cell cloning in combination with analysis of differentiation potential eliminates the possibility that some of the many cell types that mimic BMSC in vitro are providing the culture with multipotentiality

(Bianco et al. 2008).

Single cell cloning was achieved using a limiting dilution technique which has long standing in cell biology, having first been used for the cloning of irradiated HeLa cells; but has not been carefully used for analysis of human BMSCs until Sarugaser et al. in

2009; in the Sarugaser paper a comparison is made between techniques that pick single cell colonies (Puck et al. 1955, Sarugaser et al. 2009). For many authors the modern technique of cell sorting using a FACs machine is regarded as the best method to select single cells into individual wells; however single cell cloning by limiting dilution is more easily assessable (Salem and Thiemermann 2010).

Sarugaser reports that single cell cloning at low frequency is more reliable than FACS cloning, using MoFlo FACS they attributed a return of nil colonies from 768 sorted cells.

128

Whereas use of a limiting dilution of 0.5 cell per well returned 17.5% viable and 8.75% tri-potent. However the confidence that a single cell clone is returned is only 75% when

0.5 cells are seeded. These authors continued, seeding 0.2 cells per well resulted in 7.14% viable clones, and 2.92% tri-potent. The confidence that these clones are derived from a single cell is greater than 99.9%.

In the work in this chapter the cells were accurately seeded at a dilution of 0.3 cells per well, the confidence that there will be a single cell seeded into the well at this dilution will be 95.54% according to Poisson distribution (Morley et al. 1983). The return of viable clones (Table 4.2) was 11.9%±2.7 SD and tri-potent single cell clones were found at a frequency of 5.3%±0.91 SD which favourably compares with the Saragaser data.

The single cell clones (SCC) from the cultures studied here were cloned from the bulk culture at approximately 15 population doublings from isolation. In total testing of the clones required an estimated 15 population doublings, from a single cell to 3.5x104. The total cumulative population doublings of the SCCs was analogous to the longevity of the bulk cultures. Unfortunately after cloning, self renewal analysis and testing pluripotency there is little proliferation capacity with concurrent multipotentiality left for experimenting and the original bulk culture must be used.

There are many practical technical issues with the cloning of BMSCs that have much to do with the relatively small in vitro longevity and pluripotency that most authors report,

Sarugaser et al. overcame this in part by using cord blood isolates (Bruder et al 1997,

Muraglia et al 2000, Baxter et al 2004). The findings of this chapter are consistent with the longevity of others cultures, in that differentiation and cell growth seems dependant on the isolate (Table 4.1). These findings agree with other authors; this may well indicate

129

that each culture comprises of a homogeneous population of cells that is at various stages of age and differentiation (DiGirolamo et al. 1999).

Clonal isolation of the BMSC population is key; clonal isolation selects those mononucleocytes whose growth is not contact inhibited, it is likely that any stem cell residing in the bone marrow stroma would be isolated by this technique (Bianco et al

2008). A fraction of cells isolated by this technique are multipotent in vivo and clonogenic cells can be increased by selection with STRO-1 and MCAM (Sacchetti et al

2007). Because only a tiny fraction of cells in these differently isolated populations are capable of in vivo multipotency it currently makes no difference how you select for them; as long as proof is provided that the culture contains single cells capable of differentiation to multiple phenotypes. However classical methods as described by Bianco are more efficient and select a larger CFU-F proportion (Bianco et al 2008).

Because cultures of BMSC have heterogeneous capacity for self renewal and lose their multipotentiality over time in culture it is important to use them experimentally before the loss occurs. Therefore, concurrent with the single cell cloning assay and pluripotency testing the original heterogeneous populations were grown in large numbers, tested for pluripotency and ultimately used in the experiments outlined here before their usefulness expires.

Therefore we know the multipotent cells used to generate the data in this chapter were clonally isolated from human bone marrow. These bulk cultures were themselves capable of pluripotent differentiation for several passages, however contained within them was a cell population that was capable of in vitro self renewal and differentiation to multiple phenotypes at a frequency of around 5% of the heterogeneous population.

130

4.5.2 Genistein mediated differentiation

After testing the BMSC pluripotency and self renewal capability some preliminary genistein testing was undertaken. Conditions were replicated from Chapter 3 in order to observe any differences in the effects of genistein on MG63s compared with primary human BMSCs in this confluent model. Results from this chapter can be compared to the graph (fig 4.1) which shows the change in markers of osteoblast differentiation when

BMSCs are cultured in osteogenic media.

Overall the observations in this chapter show collagen expression does not change; while in media with dex at Day 4 and without dex at Day 6 ALP increases considerably. With genistein mineralisation is above controls and increases at all time points, indicating that treatment with genistein in this model compresses the differentiation process compared with control cultures. This effect can be seen in growth media but more significantly in a combination of osteogenic media and supplementation with ≤15µM genistein, indicating a combination effect.

After addition of 30µM and 300µM of genistein in all media types there is a large amount of cell death, however there is also a high level of ALP activity and mineralisation. As postulated by Morris et al. (2006) this effect may be due to pathogenic mineralisation. As the effect of 30µM and 300µM on expression of ALP and mineralisation is similar to the effect of 15µM of genistein; which does not cause increased cell death when compared with controls, it is unlikely that pathogenic mineralisation accounts for the observations at that concentration.

Heim et al (2004) isolated mononucleocytes from human bone marrow aspirates and cultured them in high glucose DMEM. After demonstrating bulk culture multipotency they experimentally cultured the BMSCs sub-confluently in osteogenic media with

131

genistein at 1µM and 10nM (Frank et al 2002, Heim et al 2004). The Heim data suggests that during osteogenic differentiation 1µM of genistein induces an increase in ALP activity 3.5 fold relative with controls, increases Cbfa-1 expression 3 fold and increases mineralisation at Day 15. During adipocyte differentiation 1µM genistein increases TGFb expression and decreases markers of adipogenesis; these observations were removed by addition of 5µM of Fulvestrant, ICI 182780 (Frank et al 2002, Heim et al 2004).

Heim did not offer any selection criteria other than non-clonal adherence to tissue culture plastic in high glucose medium followed by characterisation of bulk differentiation potential. As expounded upon at length in the chapter introduction it is uncertain what populations of cells may adhere in this manner and culture as if BMSC. Although it cannot be said that the culture in the Heim paper contains similar cells the data is quite strikingly similar to the data presented in this chapter. Genistein at 1µM in the Heim paper and 1.88µM in this chapter is effective at accelerating osteogenic differentiation when applied in osteogenic medium.

Liao et al (2007) report that a culture of non-clonally isolated mouse bone marrow cells in osteogenic medium with genistein added at 1µmol/litre increased expression of ALP and deposition of mineralisation compared with controls. Reported increases are lower than the increases reported here and controls activate in a more rapid manner than human

BMSC. These differences could be due to the difference in initial isolation of non- clonogenic populations or are animal specific.

Pan et al (2005) also report applying genistein at 10nM to 10µM in osteogenic medium to non-clonally derived mouse marrow stromal cells resulting in a dose dependant increase in proliferation, ALP, Cbfa1 and mineralisation. This data compares favourably with the data presented in this chapter, Pan et al. report 1µM genistein application in osteogenic

132

medium at 24 hours increases proliferation by 1.8 fold, the data presented here shows proliferation increased by 3.33 fold at 1.88 µM genistein and 2 days. Pan reports that the differentiation effect can be removed using Fulvestrant at 100nM.

Kim et al. report isolation of MSC from the Wharton's jelly (WJ-MSC) of the umbilical cord; they isolated an un-reported number of cells at an unknown initial seeding density.

These cells, when cultured for 2 passages, express some commonly used markers of

BMSCs; i.e. CD44, CD105, SH2, SH3, CD29 and CD51, whilst they do not express

CD34, CD45, and can be differentiated into chondrocytes, osteocytes, adipocytes and cardiomyocytes (Kim et al. 2010).

Using these cells Shieh et al, applied genistein at 40µM and 200µM retarding proliferation of the WJ-MSCs without reducing cell number at 72 hours (Shieh et al.

2010). Both concentrations induced accumulation of cells at the G0/G1 phase together with a reduction in S phase cells by 48 hours. Although cell number in control cells appeared to increase over the time of the experiment protein levels remained constant.

The authors attribute the effect of genistein on the reduction of intracellular Ca+2 they also provide evidence for inhibition of β-catenin signalling and postulate mediation of actin filament reordering that has been shown to occur in differentiating cells (Andl and

Rustgi 2005, Lilien and Balsamo 2005).

Although the data presented in this chapter should be based on comparable cell types the effect of the reported concentrations is completely different. Here 30µM is sufficient to induce a reduction in cell number at Day 2, whereas Shieh et al. only report cessation of proliferation between Day 2 and 3 with 40µM and no proliferation with 200µM. The reported differences in expression, proliferation and cell death could be because the cells used by Shieh et al are from a post-natal source; however as the isolation procedure is

133

unknown the cells are not comparable with any other literature. The Sarugaser data suggested that post-natal cells have a much larger capacity for in vitro growth; this increased capacity may indicate why WJ-MSCs are not affected by genistein in the same manner as adult BMSCs (Sarugaser et al 2009).

When compared to the MG63 data it is clear that the effect of genistein is different in the two models, MG63’s show a completely different proliferation response. ALP is up regulated at earlier time points, but to a lower degree, while the difference between media types is not as apparent. Collagen production is linked to genistein dose in MG63s, but not BMSCs, finally MG63 mineralisation is not as significant as BMSCs and occurs faster in the controls.

As described, after treatment with genistein BMSCs have been observed to differentiate similarly to the data presented here. The MG63’s, although providing a large cell number to fine tune methods and characterise the confluent model, do not react to genistein in the same way as BMSCs.

The bone marrow isolates characterised here are similar to much of the data available in the literature concerning the use of clonally derived BMSCs. Growth potential, CFU-F capacity, surface marker expression, single cell cloning and pluripotency are reported at levels that are within range of what is widely reported in the literature.

The use of BMSCs to characterise the effect of genistein on the differentiation of a confluent cell model to an osteoblast phenotype is novel. At 1.88µM of genistein in osteogenic medium, proliferation, ALP, collagen and mineralisation increased over the experimental time period. This shortening of the duration of osteogenesis is not in the literature instead authors report that genistein arrests growth when greater than 40µM or reduces adipogenesis at 1µM (Frank et al 2002, Heim et al 2004). 134

It has been shown here that genistein can increase osteogenic differentiation when added to growth media, however when dex is also added to the media there is an additional, complementary effect on osteogenesis.

The use of an antibody panel for isolation of pluripotent cells has been used by many authors that study BMSCs. To provide complete data analysis with large numbers of cells, BMSCs from Cambrex were purchased to provide a reliable cell source that has been clonally derived, cultured and characterised for differentiation and antibody markers that are well studied in the literature. A commercially available population of BMSCs was considered in addition to the clonally derived population because of the lack of surface epitope marker characterisation and the low number of repetitions in the data.

4.6 Conclusions

The application of <15µM genistein proved the hypothesis that genistein accelerates the differentiation of primary human BMSCs true; whilst >15µM was sufficient to reduce cell number. The observed effects of genistein in growth media were cumulative when combined with osteogenic media.

Classical isolation of BMSCs produces cultures which have sub-populations that are pluripotent at the single cell level, the use of BMSC in a confluent differentiating model of osteogenesis is possible.

Although MG63’s and primary human BMSCs both differentiate when exposed to genistein MG63’s react to a smaller degree with regards to cell death and activation of

ALP.

135

CHAPTER 5

THE USE OF CULTURED CBMSCS IN A CHARACTERISED

CONFLUENT DIFFERENTIATING MODEL OF OSTEOGENESIS TO

DETERMINE THE IN VITRO EFFECT OF GENISTEIN IN SEVERAL

MEDIA.

5.1 Introduction

In previous chapters the effect of genistein was studied on the MG63 osteosarcoma cell line and classically isolated primary human multipotent BMSCs. However in the literature many authors use antibodies for the selection and concurrent characterisation of primary human bone marrow cells. In order to provide comparable and reproducible data a commercial BMSC source was purchased so that observations could be made using cells isolated by identical methods and available to all researchers. The effect of genistein was compared between the initial data from classically isolated and characterised primary human BMSC and antibody characterised BMSCs to find any differences between the two isolates and to increase repetition and confidence in the data.

Commercial BMSCs (cBMSC) were purchased from Lonza. Lonza sell cBMSCs under licence from, and produced by Osiris Therapeutics, the production of cBMSCs is covered by several US Patents (US 5,486,359). Each cell isolate is taken from healthy individuals between 18 and 30 years of age and is extensively screened and tested using FDA approved tests. Prior to harvest, donors are screened for transmittable diseases, the medical and social history of each donor is obtained to find any signs, symptoms or behaviours that are consistent with a high risk of disease. The health of the donor is then monitored for up to five years after donation to further ensure his/her health status.

136

Because of this donor analysis the ethical, health and practical concerns that hamper the development of other stem cell treatments are avoided (Hanson 2001). This also provides researchers with a bone marrow derived cell product that is as normal as possible, without a transmittable disease phenotype or harvesting bias.

Currently Osiris Therapeutics has Phase 3 clinical trials underway using BMSCs under the trademark “Prochymal”. The trials are evaluating cBMSCs use for the treatment of several diseases including acute graft versus host disease (GvHD) and Crohn's disease.

Each donation of bone marrow provides 10,000 cryovials each containing 7.5x105 cells, this equates to 7.5x109 cells in total. This large cell number from each donor can only be explained if the cells undergo a selection and culture period prior to sale.

In the past several important BMSC researchers have worked at Osiris Therapeutics, including, Jaiswal, Pittenger and Bruder, co-authors on several seminal papers with

Haynesworth and Caplan, founders of Osiris Therapeutics. Caplan is often quoted as the originator of the term Mesenchymal Stem Cell (Kuznetsov et al. 2009). Osiris

Therapeutics state in patents and FDA licences that the technique that they use to isolate the cBMSC is based on the 1999 Pittenger paper (Pittenger et al. 1999).

The Pittenger paper used density gradient mononuclear cell selection criteria followed by initial seeding at a density of 2x105 cells cm-2 and culturing in DMEM with 10% FCS for several passages. Flow cytometry was used after several passages to analyse the cells, which were also capable of differentiation to chondrocytes, adipocytes and osteoblasts.

Over 50 available antibodies were used in the characterisation of the cells after some culture time, a selection panel was developed and the cultures were quality controlled based on this phenotype (Pittenger et al. 1999).

137

Pittenger reports no spontaneous differentiation, but it is not clear from this statement if they include a spontaneous loss of differentiation potential. Such a loss would only occur by differentiation, although not to a defined terminal lineage. To test for individual cell potential Pittenger et al used a colony forming cloning assay from isolation and colony picking to select single cell clones. Clonal isolation of BMSCs is still the gold standard isolation technique. However the colony picking method has since been disproved as a single cell selection method because the mobility in culture of BMSC is large and multi- cell clones have been demonstrated from colonies picked in the manner described by

Pittenger (Kobayashi et al. 2005).

The colony picking method is still in use in most of the literature. They do report that single cell clones picked in this manner have exactly the same antibody phenotype as the bulk culture, but that not all are capable of tri-lineage differentiation. Therefore cBMSCs have the same in vitro challenges that are apparent of cells in the literature and the isolates from chapter 4.

In Chapter 4 the terminology used in the literature was outlined in depth, the term mesenchymal stem cell utilised by Caplan and Osiris therapeutics is now an outdated descriptor. Pluripotent cells from the bone marrow have not yet been proved to be capable of in vivo differentiation to all tissues of the mesenchyme, therefore use of mesenchymal is not valid. Although there might be a small population of single cells within the bone marrow that demonstrate in vitro/in vivo pluripotency the whole population cannot be described as stem cells.

Therefore the use of the term MSC to describe this whole population is wrong. Since the isolated cells are from the bone marrow stroma the use of the term bone marrow stromal

138

cell (BMSC) is more descriptive and maintains the MSC acronym. So although the cells are sold as mesenchymal stem cells, here they are re-named cBMSC.

For an example of why the use of this nomenclature is important we can say that some of the mononucleocytes from the bone marrow can be isolated clonally (BMSC), from these

BMSC cultures a small population (~5-10%) can be observed to clonally tri-lineage differentiate in vitro (multipotent BMSCs). A small number (~10%) of the clonally isolated BMSCs are also capable of in vivo differentiation to multiple lineages (stem cells). Both the multipotent BMSC sub-population and the stem cell sub-population overlap, however not all multipotent BMSCs are stem cells (Pittenger et al 1999, Bianco et al 2006 & 2007).

The often quoted clonal frequency of 1-50 in 105 bone marrow mononucleocytes which are capable of forming a BMSC culture is not the fraction of the culture which is true stem cells. Very few studies of properly cloned multipotent BMSC cultures have been placed into in vivo models to determine the overlap between single cell clones and in vivo tri-potent clones (Bianco et al 2006, Sarugaser et al. 2009). Although the nomenclature is correct in the outline above, the proportions of the cultures that are in vitro and in vivo multipotent may fluctuate according the methodologies used and the individual isolates.

To confirm observations of the effect of genistein on human primary BMSCs commercial cells were obtained that are available to all researchers, isolated in a standardised manner, characterised in a way that predominates in the literature and is approved by the ISCT, are disease free and have no harvesting bias. These cells were used to confirm the osteogenic effects of genistein in a manner that is achievable by all researchers.

139

5.2. Hypothesis

The previously observed effects of genistein are the same with more replicates and therefore greater statistically certainty in a disease and harvesting bias free BMSC model.

5.2.1 Aim

Further to the previous chapters there was a need to investigate the multiple effects of genistein on a primary human BMSC strain with more statistical certainty. The cBMSC strain was purchased in order to improve the statistical certainty of the data and to compare this characterisation method with the classical method. The osteogenic effect of genistein on the proliferation, differentiation and mortality of cMBSCs will be observed.

5.3 Experimental procedure

The experimental methods for determination of effects of genistein are described in full at

2.2.13. In brief; the BMSCs were seeded at confluent density and left to attach. The medium was replaced with experimental media containing genistein, harvesting occurred after Days 2, 4 and 6 of culturing. Methods are described in full at 2.2.10.

5.4 Results

5.4.1 T0 samples and confirmation of cell attachment.

Light microscopy confirmed that the cells were attached and confluent before the attachment medium was exchanged for experimental media. At this time point (T0) samples were also taken which was essential to quantify some of the results and to ensure that the cells remain in a non-differentiated state at the start of the experiment.

5.4.2 LDH measurement of cell number.

An LDH standard curve was produced from known cell numbers and used to attribute the activity of the cellular LDH to cell number. Unknown samples were lysed and the LDH activity was used as a measure of the total cell number present in the lysis suspension. 140

5.4.2.1 LDH derived cell number after culture in growth media.

* * * * * * * * * * * *

Fig 5.1 The cell number present in the total monolayer, derived from the LDH content after 0, 2, 4 and 6 days culture with genistein in growth medium. n=6, error bars are SD. * indicates the value is significant when compared with the control P<0.05.

cBMSC were cultured at confluent density in growth media (fig 5.1) for 6 days, in the control cultures the cell number increased from Day 0 to Day 2 and fell afterwards. A significant increase in cell number occurred at Day 2 after culture with genistein at 1.88 and 3.75µM when compared with controls. At Day 4 cell number reduced in all treatments but were significantly higher than controls in 1.88 and 7.5µM, at 300µM cell number was significantly lower than controls. All genistein treatments were significant at

Day 6, 1.88-15µM higher than the control and 30-300µM lower.

141

5.4.2.2 LDH derived cell number after culture in beta-glycerophosphate media.

* * * * * * * * * * *

Fig 5.2 The cell number present in the total monolayer, derived from the LDH content after 0, 2, 4 and 6 days culture with genistein in beta medium. n=6, error bars are SD. * indicates the value is significant when compared with the control (P<0.05).

After culture in beta medium (fig 5.2) control cBMSCs increased in number from Day 0 to Day 2, after Day 2 LDH derived cell number reduced. At Day 2 in the genistein treatment media cell number increased from 1.88 to 15µM, significantly at 7.5 and

15µM, while decreasing significantly at 30 and 300µM.

On Day 4 cell number was significantly greater at lower genistein <15µM and lower at

30 and 300µM. By Day 6 only 30 and 300µM were significantly different from controls.

142

5.4.2.3 LDH derived cell number after culture in dex media.

* * * * * * * *

* *

Fig 5.3 The cell number present in the total monolayer, derived from the LDH content after 0, 2, 4 and 6 days culture with genistein in dex medium. n=6, error bars are SD. * indicates the value is significant when compared with the control (P<0.05).

After control cBMSCs were treated with dex medium (fig 5.3) LDH derived cell number did not significantly differ. When genistein was added to dex media at concentrations between 1.88 and 30µM LDH derived cell number was significantly greater at Days 2 and 4. After addition of 300µM cell number was reduced compared with controls over the same period. At Day 6 cell number was not different from the controls.

143

5.4.2.4 LDH derived cell number after culture in osteogenic media.

* * * * * * * *

* *

Fig 5.4 The cell number present in the total monolayer, derived from the LDH content after 0, 2, 4 and 6 days culture with genistein in osteogenic medium. n=6, error bars are SD. * indicates the value is significant when compared with the control (P<0.05).

Control cultures in osteogenic medium (fig 5.4) were not significantly different. When genistein was added to osteogenic media between 1.88 and 30µM cell number was significantly greater at Days 2 and 4. After addition of 300µM a reduction was apparent when compared with controls.

5.4.2.5 LDH derived cell number summary.

General trends across all media types are apparent; in control cultures there is some proliferation of the cultures, increasing on average 3x103 (±8.35x102) from seeded. The increased cell number is not maintained, by Day 6 cell number returns to the seeded density. After genistein treatments from 1.88–15µM, cell number is largely increased in dex and osteogenic media, in growth and beta media cell number increases but the proliferation is not as large. Cell number is reduced after culture with genistein at 30 and

300µM in all media.

144

5.4.3 Summary of cell counts and live/dead measurements.

The method used to determine the live dead proportion was described at 2.1.11.6. In order to measure the total number of cells at each time point as well as differentiate between living and dead a HPI live/dead cell count was used.

Fig 5.5 Average Hoechst stained cell counts of the monolayer after genistein treatment in media at Day 6. N=6. Error bars equate to SD. * indicates significant when compared with relevant control, † indicates significant difference when compared with growth media (P<0.05).

Total cell numbers and live/dead HPI stains enhance our understanding of the balance between toxicity and proliferation as well as providing confirmation of the LDH data.

The graph of direct live cell counts (fig 5.5) confirms the LDH data. When genistein is added to dex and osteogenic media cell number increases significantly at <15µM and decreases significantly at 30 and 300µM. This data is in concurrence with the information from the LDH assay.

145

5.4.4 Summary of Brd-U Incorporation into replicating DNA.

Using the method described in 2.2.10.5.2 the incorporation of BrdU into replicating DNA was measured.

* * * *

* * *

Fig 5.6 Incorporation of BrdU into cBMSCs after treatment with genistein for 6 days. The error bars represent the standard deviation, n=24. * indicates significant when compared with relevant control P<0.05.

The representative data measured by BrdU incorporation (Fig 5.6) into cBMSCs at Day 6 after culture in the four media shows no significant difference between controls. After treatment with genistein from 1.88µM to 15µM there was a significant increase relative to the controls while 300µM was observed to have significantly less proliferation. After treatment with 1.88 and 3.75µM genistein dex and osteogenic media proliferated significantly more than GM and beta, while after 15µM GM was greater than dex and osteo, at 30 and 300µM both GM and beta were significant when compared with dex and osteo (P<0.05).

146

5.4.5 LDH in the supernatant as a measurement of cell death.

The method used were as described by the manufacturer and can be found in full at

2.2.10.5.1. The amount of LDH available in the monolayer combined with the supernatant LDH was used as a measure of total LDH. The LDH available in the supernatant was directly compared with the total LDH. Therefore a relative LDH value approaching 100% indicates a small number of monolayer cells combined with a large release of LDH into the supernatant.

5.4.5.1 Supernatant LDH after culture in growth media.

* * * * * * *

* * * * * * * * *

Fig 5.7 Supernatant LDH as a proportion of total LDH after cBMSCs were cultured with genistein in growth media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control (P<0.05).

In the growth media controls (fig 5.7) the relative LDH release is not significantly different indicating that over the experiment there is low cell death relative to total live cells in the monolayer. In the genistein treatments at Day 2 only 1.88 and 3.75µM are not significantly greater than controls. At Day 4 and 6 all genistein concentrations induced increased LDH release when compared with the control.

147

5.4.5.2 Supernatant LDH after culture in beta-glycerophosphate media.

* * *

* * * * * * * * * * *

Fig 5.8 Supernatant LDH as a proportion of total LDH after cBMSCs were cultured with genistein in beta media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control. (P<0.05).

In beta medium controls (fig 5.8) LDH release is not significantly different from Day 2 to

4 but drops at Day 6. After culture for 2 days in beta media containing genistein from

1.88µM to 15µM cell death was not different from controls, at 30µM and 300µM cell death was significantly higher than controls. At Day 4 and 6 cell death was significantly greater than controls in all the treatments. There was a trend of increasing cell death with time and concentration.

148

5.4.5.3 Supernatant LDH after culture in dex media.

* * * * * * *

* * * * *

Fig 5.9 Supernatant LDH as a proportion of total LDH after cBMSCs were cultured with genistein in dex media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control (P<0.05).

After culture with dex medium (fig 5.9) controls were similar to growth and beta- glycerophosphate media at Day 2, at Day 4 and 6 relative LDH release was significantly lower. At Day 2 culture in dex media supplemented with genistein concentrations 1.88–

30µM were significantly lower than controls. At Day 4 30 and 300µM were significantly higher than controls. By Day 6 all concentrations of gen cause increased relative LDH release. The observable trend was that at early time points relative LDH release was reduced when genistein was added, however LDH release increases with both time in culture and concentration of genistein.

149

5.4.5.4 Supernatant LDH after culture in osteogenic media.

* * * * * * *

* * * * * * * *

Fig 5.10 Supernatant LDH as a proportion of total LDH after cBMSCs were cultured with genistein in osteogenic media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control (P<0.05).

Osteogenic medium controls (fig 5.10) do not release large quantities of LDH when compared to the other medium types, over time in culture the controls do not change significantly. After culture in osteogenic media supplemented with genistein a trend of increasing relative LDH release with time and increasing concentration of genistein was observable. At Day 2 all treatments have significantly more relative release than controls, by Day 4 concentrations greater than 3.75µM are significantly higher than controls. At

Day 6 all treatments are significantly higher than controls.

5.4.5.5 Summary of relative LDH release.

Control LDH release was lower in the dex containing media. Generally across all media the relative release of LDH was lower at ≤15µM. By Day 6 all treatments were significantly greater than the controls and increased with time in culture.

150

5.4.6 Measurements of ATP availability.

The method used was as described by the manufacturer and is explained at 2.2.10.6. The availability of ATP is indicative of the metabolic state of the cell, it is well established that the level of intracellular ATP determines cell fate; ATP depletion is a manifestation of type III programmed cell death (Tsujimoto 1997, Nicotera and Leist 1997, Orrenius et al 1997). It is possible to see very early effects of a compound on cells by analysis of the levels of ATP (Talha and Harel 1985, Jeong and Seol 2008).

5.4.6.1 ATP after culture with genistein in growth media

* * * † † † † * * † * * †

* * †

Fig 5.11. ATP available after cBMSCs were cultured with genistein in growth media for 2 and 72 hours. Control n=24/48 and experimental n=14, error bars represent SD. * indicates significance compared with control. † indicates significantly larger than the alternative time point (P<0.05).

In growth media supplemented with genistein (fig 5.11) up to 7.5µM there was a significant increase after 2 hours in the availability of ATP compared with controls. After culture with 15µM to 300µM genistein there is significantly less ATP available compared with the controls. After 72 hours incubation with ATP the controls are significantly lower than 2 hours although the means are the same, this effect is due to the larger value of n at

151

72 hours (48 vs 24). There are no differences between the control and 1.88µM to 7.5µM; from 15µM to 300µM ATP is significantly lower than the control.

5.4.6.2 ATP after culture with genistein in beta-glycerophosphate media.

* * * † † † † * * † * * †

* * †

Fig 5.12. ATP available after cBMSCs were cultured with genistein in beta media for 2 and 72 hours. Control n=24/48 and experimental n=14, error bars represent SD. * indicates significance compared with control. † indicates significantly larger than the alternative time point (P<0.05).

In beta media (fig 5.12) containing genistein after 2 hours there are clear similarities to growth media, significantly higher ATP is found at concentrations from 1.88µM to

7.5µM, whereas higher concentrations are significantly lower than controls. Although there is lower available ATP in controls after 72 hours in treatments there is a dose response similar to 2 hours; after treatment with 15µM to 300µM ATP is significantly lower. After 72 hours culture with 300µM ATP cannot be differentiated from the blanks and is therefore depleted.

152

5.4.6.3 ATP after culture with genistein in dex media

* * * † † † † * * † * * †

* * †

Fig 5.13. ATP available after cBMSCs were cultured with genistein in dex media for 2 and 72 hours. Control n=24/48 and experimental n=14, error bars represent SD. * indicates significance compared with control. † indicates significantly larger than the alternative time point. (P<0.05).

cBMSCs cultured in dex medium (fig 5.13) showed a decrease in the control levels of

ATP from 2 hours to 72 hours. Compared to growth media there was an increase in the level of ATP in cBMSCs when grown in dex media.

Similar to the two previous treatment regimen (fig 5.11 and 5.12) there was a decrease in the level of ATP at ≥15µm, 300µM was significantly lower than the next lowest. By 72 hours treatments contained less ATP overall but the dose response curve was very similar to observations at 2 hours, no significant differences were observed between 1.88 and

7.5µM and the control. After 72 hours culture with 300µM genistein ATP levels were indistinguishable from background.

153

5.4.6.4 ATP after culture with genistein in osteogenic media.

* * * † † † † † * * †

* * †

Fig 5.14. ATP available after cBMSCs were cultured with genistein in osteogenic media for 2 and 72 hours. Control n=24/48 and experimental n=14, error bars represent SD. * indicates significance compared with control. † indicates significantly larger than the alternative time point (P<0.05).

After 2 hours culturing in osteogenic media (fig 5.14) ATP levels in the control were greater than in growth media alone. Similarly the combination of osteogenic media and genistein has a cumulative effect on ATP levels at 2 hours. After culture with 1.88µM to

7.5µM genistein ATP is significantly greater than the control, 30µM and 300µM are significantly lower. After 72 hours the application of 30 and 300µM significantly reduced

ATP when compared with controls. In osteogenic media application of 300µM genistein results in near depletion of ATP at 72 hours.

5.4.6.5 Summary of ATP availability in all media.

In all tested media 1.88µM-7.5µM of genistein significantly increases ATP at 2 hours and is the same as the controls at 72 hours. Concentrations greater than 15µM significantly reduce ATP in all tests, after culture with 300µM ATP is either depleted or near depletion at 72 hours.

154

5.4.7 ALP activity as an early measurement of differentiation

Alkaline phosphatase (ALP) is an early marker of osteoblast differentiation; measurements were made using the methods described at 2.2.10.2

5.4.7.1 ALP activity after genistein treatment in growth media

† † † † † † † † † Θ † †

Fig 5.15. ALP activity after cBMSCs were cultured with genistein in growth media for 2, 4 and 6 days. n=6, error bars represent SD. Θ indicates not significant compared with control, all other data is significant when compared with the relevant control. † indicates significance compared with the previous time point. (P<0.05).

After culture in growth medium (fig 5.15) controls show increased expression of ALP from Day 2 to 4 and no significant change thereafter. All the genistein supplemented media at Day 2 and 4 induced a significant increase in the activity of ALP. At Day 6 all concentrations except 300µM were significantly greater than the control.

Peak ALP was observed in the mid range (3.75-15µM) of the concentrations tested.

155

5.4.7.2 ALP activity after genistein treatment in beta-glycerophosphate media

† † † † † † † † † † † †

Fig 5.16. ALP activity after cBMSCs were cultured with genistein in beta media for 2, 4 and 6 days. n=6, error bars represent SD. all experimental data is significant. † indicates significance compared with the previous time point. (P<0.05).

In beta medium controls (fig 5.16) ALP activity increased significantly over time from

Day 2 to 4 and was not significantly different from Day 4 to 6. After supplementing beta- glycerophosphate medium with genistein there was a significant increase in activity of

ALP at all concentrations and time points.

At Day 2 and 4 ALP activity was very similar at all genistein concentrations. By Day 6

3.75µM was most highly activated, generally activity reduced with increase in genistein

>3.75µM.

156

5.4.7.3 ALP activity after genistein treatment in dex media

† † † † † † † † †

† † † † Θ †

Fig 5.17. ALP activity after cBMSCs were cultured with genistein in dex media for 2, 4 and 6 days. n=6, error bars represent SD. Θ indicates not significant compared with control, all other data is significant. † indicates significance compared with the previous time point (P<0.05).

After culture in dex medium the controls (fig 5.17) show an increase in ALP activity over time; dex media alone increases the activity of ALP over time when compared to growth media. When dex media is supplemented with genistein only 300µM at Day 6 is not significant when compared with the controls. There is a stepwise increase with time in culture after addition of genistein from 1.88 to 30µM.

157

5.4.7.4 ALP activity after genistein treatment in osteogenic media

† † † † † † † † † †

† Θ † †

Fig 5.18. ALP activity after cBMSCs were cultured with genistein in osteogenic media for 2, 4 and 6 days. n=6, error bars represent SD. Θ indicates not significant compared with control, all other data is significant. † indicates significance compared with the previous time point. (P<0.05).

In osteogenic medium controls (fig 5.18) there is increasing ALP with time in culture, significant increases occur at each time point.

When compared with the controls ALP is significantly increased at all time points after culture with genistein concentrations between 1.88µM and 30µM. The addition of

1.88µM initiates peak activation of ALP, reducing with increasing concentration, while concurrently increasing with time.

300µM of genistein increases ALP significantly at Day 2, however at Day 4 there is no difference from the control, at Day 6 there is a significant reduction in activation when compared to the control.

158

5.4.7.5 Summary of ALP activity after genistein treatment.

Genistein at concentrations 1.88–30µM induce an increase in ALP activity at all time points and in all media. Activation of ALP by genistein in growth and beta media is about halved from the dex containing media.

5.4.8 Collagen production after culture with genistein containing media.

Collagen deposition is a marker of osteogenesis. An increase in production of collagen is directly attributable to an increase in osteogenic phenotype when studied in conjunction with differentiation markers and mineralisation data. The method to determine collagen production is described in full at 2.2.10.4.

5.4.8.1 Collagen production after culture with genistein in growth media.

* * * * * * * * * * * * * * * * *

Fig 5.19. collagen deposition after cBMSCs were cultured with genistein in growth media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control (P<0.05).

In growth media collagen controls (fig 5.19) increase from Day 4 to Day 6. In the genistein treatments at Day 2 1.88-30µM are significantly lower than controls. At Day 4 all are significantly greater than controls. At Day 6 cultures 3.75 to 30µM are significantly lower than the control.

159

5.4.8.2 Collagen after culture with genistein in beta-glycerophosphate media.

* * * * * * * *

Fig 5.20. Collagen deposition after cBMSCs were cultured with genistein in beta media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control (P<0.05).

After culture with beta medium (fig 5.20) there is an increase in collagen production over the experiment, significant from Day 2 to 4 and from 4 to 6. Compared with controls after

2 days culture with the addition of genistein at concentrations of 1.88 and 7.5-30µM there is a significant reduction in collagen. At Day 4 only 30-300µM induced a significant increase in collagen. On Day 6 none of the genistein concentrations were different from the control.

160

5.4.8.3 Collagen production after culture with genistein in dex media.

* * * * * * * * * * * *

Fig 5.21. collagen deposition after cBMSCs were cultured with genistein in dex media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control (P<0.05).

After culture with dex medium (fig 5.21) there is a significant increase in collagen production over the experiment, from Day 2 to Day 6. After 2 days culture with the addition of genistein at 3.75µM there is a significant reduction in collagen, while and

300µM induces a significant increase. At Day 4 all genistein induced a significant increase in collagen. On Day 6 only 3.75 and 7.5µM were not greater than the controls.

161

5.4.8.4 Collagen production after culture with genistein in osteogenic media.

* * * * * * * * * * * *

Fig 5.22. collagen deposition after cBMSCs were cultured with genistein in osteogenic media for 2, 4 and 6 days. n=24, error bars represent SD. * indicates significance compared with control. (P<0.05).

The osteogenic medium control cultures (fig 5.22) significantly increased collagen production at each time point of the experiment. At Day 2 after culture in osteogenic media with 1.88µM and 3.75µM genistein a significant reduction in collagen was observed when compared with the control. At Day 4 all concentrations of genistein induced collagen production that was significantly higher than the control. At Day 6 all treatments induced a further increase in collagen, only 30µM was not significantly higher than the control.

5.4.8.5 Summary of collagen production after culture with genistein.

General observations of all media indicate that genistein induces an increase in collagen over time. In dex containing media supplemented with genistein collagen was deposited in greater quantities than the controls.

162

5.4.9 Quantitative alizarin red staining of mineralisation.

Mineralisation of the ECM was determined using the method described at 2.2.10.1

5.4.9.1 ARS after culture with genistein in growth media.

Fig 5.23. Alizarin red staining of calcium after cBMSCs were cultured with genistein in growth media for 0, 2, 4 and 6 days. n=6, error bars represent SD. * indicates no significance compared with control, all other points are significantly lower than controls. (P<0.05).

In growth medium control cultures (Fig 5.23) increase stained monolayer calcium over time in culture. In genistein supplemented media almost all treatments and time points were significantly less than the controls, only 15µM at Day 6 was not different from the control mineralisation.

163

5.4.9.2 ARS after culture with genistein in beta-glycerophosphate media

* * * * * * *

Fig 5.24. Alizarin red staining of calcium after cBMSCs were cultured with genistein in beta media for 0, 2, 4 and 6 days. n=6, error bars represent SD. * indicates significance compared with control (P<0.05).

In beta medium control cultures (Fig 5.24) mineralisation accumulated over time in culture. When genistein is added to the media there is no difference from controls at Day

2. After culture with genistein at 30 and 300µM at Day 4 mineralisation is significantly lower than controls.

At Day 6 after treatment with 1.88, 7.5 and 15µM of genistein mineralisation is lower than the control. 30 and 300µM induce significantly increased mineralisation at Day 6 when compared with the control.

164

5.4.9.3 ARS after culture with genistein in dex media

* * * * * * * * * *

Fig 5.25. Alizarin red staining of calcium after cBMSCs were cultured with genistein in dex media for 0, 2, 4 and 6 days. n=6, error bars represent SD. * indicates significance compared with control (P<0.05).

In dex medium control cultures (fig 5.25) there is an accumulation of mineralisation over time in culture. When genistein is added to dex media at Day 2 there is significantly more mineralisation from 7.5-30µM. At Day 4 only 15µM is significant. By Day 6 all concentrations of genistein induce increased mineralisation compared with controls.

165

5.4.9.4 ARS after culture with genistein in osteogenic media

* * * *

* *

Fig 5.26. Alizarin red staining of calcium after cBMSCs were cultured with genistein in osteogenic media for 0, 2, 4 and 6 days. n=6, error bars represent SD. * indicates significance compared with control. (P<0.05).

In osteogenic medium control cultures (fig 5.26) there is a relatively small amount of mineralisation, increasing over time in culture. In the genistein treated cultures mineralisation is significantly greater than controls at 30µM at all time points, at Day 2 and 4 all other genistein is not significantly different from controls

At Day 6 genistein from 3.75-300µM induces a significant increase in mineralisation.

5.4.9.5. ARS summary.

After treatment with genistein, at Day 2 and 4, mineralisation was mostly either lower or the same as controls. At Day 6 in all media other than growth medium there was a significant increase in mineralisation, largest at 30 and 300µM of genistein.

166

5.5 Discussion

Previous chapters have shown that genistein is able to stimulate the differentiation of an osteoblast phenotype in both immortalised cell lines and primary cultures of human

BMSC. This chapter more fully investigates the proliferative, toxicity and differentiation effects on primary BMSC when genistein is supplemented into both osteogenic and non- osteogenic media.

For this section of the work cBMSC were purchased from Cambrex for several reasons; firstly the cells we all derived from healthy donors and were therefore free of disease phenotypes. The primary cells used in previous chapters were from donated tissues remaining from surgical procedures, which have an unknown status for disease phenotype and an inherent harvesting bias; as they were discarded from elective surgical procedures.

Secondly, a purchased cell source has no limitations of supply and work was not halted because of lack of cells. Thirdly, the cells were characterised for standardised surface epitopes and therefore the cells used in these studies are directly comparable to the work of other groups who also use cells purchased from this supplier or whose cells are similarly screened for these markers.

Compliance with these points can also be obtained by the use of cell lines such as SAOS2 and MG63's, however as shown in previous chapters these cell lines do not have a primary phenotype or produce a response contemporaneous with primary cells.

5.5.1 Proliferation and cell death

As the effect of genistein on cell toxicity was an aim of this chapter the immediate effect of genistein on the cell was considered, it is known that the level of available ATP in a

167

cell is an indicator of whether a cell initiates cell death and is used extensively as a measurement of cellular activity and vitality (Tsujimoto 1997, Mueller et al 2004). It is possible to directly measure the change in levels of available ATP very quickly after addition of the isoflavone, attributing the addition of genistein with a measurable and persistent outcome independent of metabolites (Talha and Harel 1985, Mueller et al.

2004). An ATP assay differs from many other commonly used cytotoxicity assays as

ATP is used by the cell rapidly and requires functional mitochondria to maintain

(Orrenius et al. 1997, Nicotera and Leist 1997). For example, the MTT assay is based on the production of NADH by the mitochondrial enzyme succinate dehydrogenase and can show viable cells even after mitochondrial activity is poisoned (Loveland et al. 1992).

The LDH assay uses the same chemistry for an indirect measurement of NADH production by lactate dehydrogenase (Mueller et al 2004).

A clear relationship between dose and ATP availability can been seen in the figures at

5.4.6 indicating that there is an immediate effect of application of isoflavone on the availability of ATP, at low doses there is an increase in the level of available ATP dropping below control levels at high doses. At the longest time point and the highest dose there is very little detectable ATP, a correlation between low ATP and cell death is apparent in the literature where authors report that apoptotic and necrotic cell death maybe in part regulated by ATP availability (Stefanelli et al 1997, Tsujimoto 1997, Jeong and Seol 2008).

Genistein induced apoptosis has been often reported in the literature as a cellular response to treatment with the isoflavone (Li 1999). Many of the papers published that examine the apoptotic or enzyme inhibition effects of genistein do so at non-physiological levels which are thought to achieve 10µM at the very most. It is also debatable that these

168

experiments always use the correct cytotoxicity testing to quantify cell death, komissarova (2005) suggests that ending experiments after a short interval (less than 48 hours) can lead to an underestimation of cell death and utilises a long term clonal cell proliferation assay to better quantify the level of toxicity (Komissarova et al 2005).

In order to better observe the cell death and proliferation occurring during our experiments further measurements were made by both imaging the cells in vitro and quantifying the amount of cell death by total cell counts. The results in figs 5.1-5.10 show that low concentrations of genistein stimulate proliferation of BMSCs even when they are confluent, only high concentration of genistein result in a net loss of cells when compared to controls.

When we studied the total cell counts and phase micrographs alongside the HPI data it establishes that there is a net cell increase at low genistein (<15µM) and decrease at high doses (>30µM) of genistein. Viewing the monolayers it is apparent that there are morphological differences between the control cells and the treated cells. Control cells appear contact inhibited, spindle shaped and fibroblastic. In growth media the phenotype in the genistein treated cBMSCs changes to stacking at confluence rather than the contact inhibited growth seen in the controls. In osteogenic media the treated cell morphology becomes square and osteoblast like by Day 4.

The use of the HPI stain provides us with cell numbers from the initial seeding and final time points, live and dead counts of the monolayer together with the remaining cell number. Therefore this membrane integrity test is often used as an indication of cell viability. However cells often have leaky membranes and are still able to produce progeny therefore the data should not be taken on its own to indicate viability (Belloc et al 1994).

169

Therefore an assessment of proliferation was undertaken by analysis of the incorporation of BrdU into copying DNA. Fig 5.6 shows that the rate of proliferation in the lower concentrations of genistein was significantly higher than controls, only at 300µm of genistein was proliferation significantly lower than the control.

In the literature there is discussion about the effect of genistein on the proliferation and cell fate of many cell types, much of this discussion is centred on tumour cell types, due to the SERM nature of phytoestrogens. There is evidence to suggest that in breast tumour cell lines expressing the marker Brac1 the application of 15µM genistein is sufficient to cause significant cell death after 7 days of exposure this was attributed to the accumulation of polyploidy cells (Tominaga et al 2007).

In NIH 3T3 cells a decrease in cell number was observed after 24 hours at a concentration of 7.5µM genistein by HPI staining and MTT assay with no significant differences at lower concentrations (Rucinska et al 2008). In BMSCs previous studies have investigated the effect of genistein on differentiation and found it to increase osteoblast differentiation and decrease adipocyte differentiation, with morphological phenotype apparent days after treatment (Frank et al 2002, Heim et al. 2004).

As a further indication of cell death and proliferation of the course of the experiments cell turnover was measured using the total release of LDH into the media. The LDH method allows enumeration of the number of cells lost during the experiment and also takes into account the amount of proliferation that has occurred, therefore this data should be complementary to the proliferation data that was already collected.

As noted previously assays for the level of cell metabolism and ATP availability are often reported as being utilised in an inefficient manner. LDH, MTT and MTS assays are often used in periods of time up to 48 hours, but very rarely at time points above this (Klein & 170

King 2007). Essentially, long term survivability is best monitored over long time points.

The data presented here shows the changes in LDH availability over longer time points and also accounts for the variation in media types.

The LDH data shows the accumulating enzyme in the culture conditions, the more LDH in the supernatant the greater the amount of release from the cells from which we infer an increase in cell death. In the non-treated control cultures the amount of LDH released into the media remains at a level much lower than 100%, indicating that the control cells released less LDH than is available in the cell monolayer i.e fewer cells died than are left in the monolayer. Because we have cell counts from day 0 to day 6 that are not significantly different it is possible to infer that the control cells were confluent and contact inhibited, and only undertook light proliferation to replace cells.

In the growth media experiment genistein treatment gives us an indication of the effect on the cells that is attributable to just the effect of genistein. In the treatments that are lower than 15µM there is no significant change in supernatant LDH until Day 6 when there is significant release into the supernatant indicating a significant spike in cell death.

However when we look at the number of cells in the monolayer there is no significant change in cell number between these time points, this indicates that there is also significant proliferation during this time to maintain cell numbers.

LDH data from the treatment with greater than 30µM of genistein produces a trend of increasing LDH release into the supernatant when compared to monolayer LDH, this trend provides evidence of increasing cell death and proliferation. At 300µM there is significant increase of LDH accumulation over time.

The combination of these data sets indicates that genistein induces proliferation at concentrations from 1.88µm to 15µM, has no net effect on cells at 30µM and that cell 171

death occurs in this system at the 300µM. These findings are consistent with the findings of others, that physiological concentrations (<7.5µM) of genistein are not toxic (Branca

2003, Frank et al. 2002, Heim et al. 2004).

5.5.2 Differentiation

In our model the genistein acts to significantly increase the abundance of osteogenic markers; the marker ALP increased by Day 4 even at the lowest concentration of

1.88µM. When supplemented in the concentration range 3.75µM to 15µM the up regulation in activity of ALP is not significantly different from observations of osteogenic media controls.

In this model we observed accelerated differentiation with detectable levels of ALP well before the previously reported 7 day limit, the markers in growth medium control cultures did not significantly increase from T0, and osteogenic differentiation medium controls differentiated in a normal manner (Heim et al. 2004).

The literature shows that genistein supplementation of osteogenic medium is significant at lower levels than those tested here; we do show that there is a reduction in ALP activity from the lowest concentration of genistein to the highest in osteogenic media at

Day 4 and 6. When genistein is added to osteogenic media there is a double activation effect indicating separate mechanisms are responsible for the action of the media components, an effect not tested for by anyone else.

The media component responsible for initiating the increase in ALP activity in osteogenic media is the dex. Beta media has similar ALP activity after genistein treatment to growth media. Dex is a steroid and activates ALP by the steroid response element, akin to the oestrogen response element but activating a different pathway (Shur et al. 2005). Dual

172

activation by stimulating both the steroid and oestrogen response elements has produced nearly double the effect of each acting solely. To further test the hypothesis that this dual activation is due to the double effect of the action of the steroid and the isoflavone the addition of a pure anti-oestrogen would act as an inhibitor, blocking the effect of genistein and reducing the activity of ALP in a competitive manner.

Collagen production is another frequently used marker of differentiation, because bone formation requires that type I collagen be used extensively in the extracellular matrices prior to the formation of mineralisation nodules. In all of the experiments used for the evaluation of collagen production it is apparent that there is a general trend of increased collagen production with increasing genistein concentration and that the production of collagen continues over time, increasing the total available collagen in the monolayer that has the potential to be mineralised.

In 2004 Heim et al published one of very few studies investigating the effects of genistein on human BMSCs, their findings were that genistein increases the level of early markers of osteogenesis such as ALP and has no effect on late markers. The findings from this chapter clearly show that genistein concentrations tested here are efficacious at increasing the production of ALP and in a more timely fashion than Heim, this is probably due to the culture conditions utilised by us. Since we cultured our cells at a confluent level we effectively shortened the proliferation step from the first 7 days of differentiation, resulting in our finding significant increases in ALP by Day 4 rather than Day 18 as shown by Heim. We also observed an increase in mineralisation, a late marker of osteogenesis (Frank et al. 2002, Heim et al. 2004).

In the model presented here a relationship between dose, differentiation, cell death and mineralisation is evident. At non-toxic concentrations of genistein, <15µM, there is an

173

increase in proliferation, in ALP activity and in mineralisation over the controls. This effect is lower or neutral towards 30µM of genistein. At 300µM genistein there is lower cell numbers, cell death and relatively lower ALP activity however there is also a small trend towards increased collagen and a large and significant increase in mineralisation.

5.6 Conclusions.

The hypothesis that the osteogenic effect of genistein can be shown with greater statistical certainty in a disease and harvesting bias free BMSC population is true; genistein is demonstrably effective at increasing proliferation, cell number and osteogenesis in the cBMSCs confluent culture model.

The effect of the application of genistein is observable after 2 hours and persists for at least 6 days.

Addition of an oestrogen receptor inhibitor would demonstrate the effect is due to the phytoestrogen interaction with an oestrogen receptor.

174

CHAPTER 6

THE OSTEOGENIC EFFECT OF SOY EXTRACT ON THE CBMSC

MODEL.

6.1 Introduction.

Soy is a quality protein source, therefore soy extracts are used extensively in the food industry (Heiser and Trentelman 1989). The year on year increase in soy crop and its relative cheapness has encouraged the large scale usage of soy in a variety of food applications including as gelation agents and emulsifiers (Genovese et al. 2007).

Industrially soy is sold as hexane defatted white flake; this technology exists and has been in use since the 1930’s (Heiser and Trentleman 1989, Horan 1974).

There is longstanding use of soy as a food stuff, more recently evidence suggests that there are positive effects on peri-menopausal symptoms and the “western diseases” including osteoporosis (Adlercreutz and Mazur 1997, Sayegh and Stubblefield 2002,

Lethaby et al. 2007, Kwack et al 2009, Kim et al 2009, Barnes 2010). This evidence has led to soy being used as a “wellbeing” product, with the intent to reduce the incidence of the western diseases and increase quality and longevity of life (Setchell and Cassidy

1999, Pandey and Rizvi 2009, Blake et al 2011).

Due to these many perceived benefits of soy isoflavones they have been studied intensively as pharmacologically active dietary supplements (Adlercreutz et al 2004,

Hedelin et al 2008). However, in order to reach pharmacologically relevant doses large quantities of soy or isoflavone extracts often have to be consumed (Adlercreutz et al

2004, Joshi et al 2007, Marini et al 2007). The Asian diet contains up to 50mg of isoflavone per day, soy beans contain approximately 1mg isoflavone per gram (Reinli and

175

Block 1996). There is some evidence that the western diseases can occur in this group when the diet is changed to contain a low phytoestrogen content (Barnes 1998).

A method to extract soy isoflavones from raw soy flour was published in 2003 when Choi and Koo used ethanol extraction, in a similar manner to the method used in this chapter.

After addition of the extract to mouse MC3T3-E1 cells for 48 hours they observed an increase in proliferation, ALP and reduced TNF-α production. Although this report is based on the analysis of mouse osteoblastic cells it is the only literature reference describing the treatment of osteoblasts with soy extract prepared in a similar manner to the extract described here. Other reports include McCall et al. (2010) who derived a genistein and polysaccharide combination (GCP) from soy extract which induced several human lymphoma cell lines to inhibit proliferation and increase apoptosis.

Soy extracts and pure soy isoflavones have been observed to have different properties. In

2006 Fuchs et al used pure genistein at 25µM and soy extract, adjusted to a genistein content of 25µM, to alter the expression levels of different and defined groups of genes known to inhibit apoptosis in human endothelial cells. Therefore, according to Fuchs et al, extract containing 25µM of genistein should inhibit apoptosis and increase proliferation. There are currently no published in vitro studies of the effect of an extracted soy product with increased concentrations of isoflavones on human BMSCs. The only available reference which has studied soy extract on osteoblastic cells is the Choi paper the majority of work undertaken in this field is in vivo (Choi and Koo 2003).

An in vivo study analysing the effect of clover isoflavone extract on ovariectomised rats was undertaken by Kawakita et al in 2009. They observed that clover extract reduced bone resorption compared to controls and mild alkalinisation. This study follows on from some of the first isoflavone studies from the 1940’s, Bennetts et al (1946), showed that

176

sheep infertility was directly related to the presence of equol in the blood of sheep grazing on clover. Later research indicated that the equol was primarily metabolised from methoxy-daidzein (Davies and Hill 1989).

In an interesting Rat study by Cho et al (2009) soy saponins were identified as a potential compound that may cause the observed in vivo differences between purified soy isoflavones and soy extracts. The water and ethanol extracted soy used by Cho et al was

94% saponins and 6% isoflavones. Soy saponins have been shown to stimulate antibody production and show antigenotoxic and antitumor activity (Rao and Sung 1995, Berhow et al 2000, Rowlands et al. 2002, Oda et al. 2003). Cho et al identified the lowest observed adverse effect level (loael) in F344 rats as 725mg/kg. Although they did not attempt to separate the effects of the soy isoflavones from soy saponins in the soy extract used, previous studies have identified the loael of orally administered synthetic genistein in Wistar rats to be 500mg/kg (McClain et al. 2005 & 2007).

Further long term rat studies have been undertaken; Hagiwara et al. (2010) found no adverse effect after feeding F344 rats hot water soy extract for 28 days at “3,618 mg/kg bw/day for males and 4,066 mg/kg bw/day for females” indicating that extraction methodology may have an effect on efficacy. Brolling et al. (2009) also observed that solvent processing can effect the presence of phenols and the efficacy of the resulting solution for antioxidant scouring.

These in vivo studies demonstrate that soy extracts can be very different dependent on extraction methods and animal model. The relevance of the outlined animal models to this work can also be questioned for several other reasons. For example, some isoflavones will be metabolised in the gut before entering the blood stream, absorption and pharmacological doses are different between animal models, and as shown in previous

177

chapters the difference of human cell culture systems compared with the rat and mouse calls into question that these models metabolise the active ingredients in soy in the same manner as human cells do (Adegunloye et al. 2003, Soucy et al. 2006, Tamura et al.

2011). In cell culture several human cell types have been shown to metabolise the conversion of flavonoid glycosides to active aglycones including, epithelial, intestinal and hepatic (Murota et al. 2002, Nemeth et al. 2003, Walle et al. 2005, Chen et al. 2011).

As well as the well studied isoflavones genistein and daidzein and their conjugates, the extract also contains other phytoestrogens that are less well studied including saponins.

Discoveries of novel soy phytoestrogens are still reported in the literature, daidzein derivatives glycinol and glyceollin have been reported recently (Bolling et al. 2009).

Since the soy extract is not concentrated or purified in any way it is likely that it contains unknown components including phytoestrogens with unknown oestrogen receptor binding efficiencies and unknown transcription effects of the ER-ligand.

In an attempt to provide evidence that the observed effects are due to the phytoestrogen component of the soy extract the pure oestrogen receptor antagonist Fulvestant was added to the cell culture. Fulvestant competitively forms ligands with oestrogen receptors; however the formed ligand is inactive. The removal of observed effects after application of Fulvestant is an indicator that binding to oestrogen receptors is responsible for the observations (Chen and Anderson 2001).

6.2 Hypothesis

Soy extracts have an osteogenic effect on the primary human cBMSC differentiating cell model characterised in the preceding chapters and the use of an oestrogen receptor inhibitor will competitively inhibit any observations.

178

6.3 Methods

The experimental method was undertaken as described in 2.2.13.3

6.4 Results

6.4.1 HPLC quantification of soy extract isoflavone content.

Soy extract was prepared as per the instructions in the methods described at 2.2.11 and the HPLC was undertaken as described at 2.2.11.2

10 1

15 3

4 5

20 6

Fig.6.1 A representative HPLC trace of extracted soy protein over time (minutes), the isoflavones are labelled: 1, daidzin; 2, genistin; 3, acetlyglucosides and malonylglucosides; 4, daidzein; 5, genistein; 6, formononetin (internal standard). The marked peaks provide quantitative data for the determination of soy isoflavone content by the area beneath the peak relative to the knowns.

Six batches of soy extract were quantified in triplicate using HPLC and observed to have

140 mg (±2.34 mg) genistein per gram of dried extract.

179

6.4.2 Direct cell counts.

Cell number was counted from HPI stained cell cultures, examples of the monolayer cultures and photographs of the HPI stains used to create the following graphs can be found at 6.4.3

6.4.2.1 Live cell counts of cBMSCs cultured in growth media and soy extract.

450 GM Day 2 GM Day 6 400 GM Day 8 GM Day 10 350

300

250 * 200

150 Live cell count cell Live /field ofview 100

0 0 8.45 16.89 33.79 67.57 135.14 Soy extract (mg/ml) Fig 6.2 Direct live cell counts (HPI) of 9 fields of view per experimental conditions after cBMSCs were cultured with growth media and soy extract. N=24, error bars indicate SD. * Indicates not significant when compared with the control (P<0.05).

After culture in growth media control (fig 6.2) cBMSCs increased in number, but not to a significant degree.

After treatment with soy extract at 8.45 mg/ml cell number increased at each time point over the course of the experiment. After culture in media with 16.89 and 33.79 mg/ml there was an initial large increase in cell number followed by a reduction over time. The addition of soy extract at 67.57 and 135.14 mg/ml caused a reduction in cell number.

180

6.4.2.2 Live cell counts of cBMSCs cultured in growth media with inhibitor and soy extract.

450 GM+I Day 2 400 GM+I Day 6 GM+I Day 8 350 GM+I Day 10 300

250

200

150

100 Live cell count cell Live /field ofview 50

0 0.00 8.45 16.89 33.79 67.57 135.14 Soy extract (mg/ml)

Fig 6.3 Direct live cell counts (HPI) of 9 fields of view per experiment after cBMSCs were cultured with growth media with inhibitor and soy extract. N=24, error bars indicate SD. All points are significant when compared with the controls P<0.05.

Addition of the inhibitor Fulvestrant to growth medium reduced the proliferation of the control cultures (fig 6.3) when compared with controls without inhibitor (fig 6.2).

The effect of soy extract on cell number observed in media without inhibitor (fig 6.2) was attenuated by the addition of inhibitor (fig 6.3); both peak proliferation and cell death was lower and higher respectively. Although the difference between cell proliferation and death was reduced with the addition of inhibitor otherwise observations matched those found in growth media without inhibitor.

181

6.4.2.3 Live cell counts of cBMSCs cultured in osteogenic media and soy extract.

450 Osteo Day 2 Osteo Day 6 400 Osteo Day 8 Osteo Day 10 350

300

250 * 200

150

100

Viable cell number cell Viable of /field view 50

0 0 8.45 16.89 33.79 67.57 135.14

Soy extract (mg/ml)

Fig 6.4 Direct live cell counts (HPI) of 9 fields of view per experimental conditions after cBMSCs were cultured with osteogenic media and soy extract. N=24, error bars indicate SD. * Indicates not significant when compared with the control P<0.05.

When control osteogenic medium was applied to cultures cell number increased at each time point (fig 6.4).

Addition of soy extract between 8.45 and 33.79 mg/ml increased cell number between

Day 2 and 6 where it remained until the end of the experiment; 67.57 mg/ml induced initial proliferation followed by cell death. After culture with 135.14 mg/ml cell number was higher than controls at Day 2, the same as controls at Day 6, and cell death occurred at all subsequent time points.

182

6.4.2.4 Live cell counts of cBMSCs cultured in osteogenic media with inhibitor and soy extract.

450 Osteo+I Day 2 Osteo+I Day 6 400 Osteo+I Day 8 Osteo+I Day 10 350 300 250 * 200 150 100

Live cell count cell Live /field ofview 50 0 0.00 8.45 16.89 33.79 67.57 135.14 Soy extract (mg/ml)

Fig 6.5 Direct live cell counts (HPI) of 9 fields of view per experiment after cBMSCs were cultured with osteogenic media with inhibitor and soy extract. N=24, error bars indicate SD. * Indicates not significant when compared with the control P<0.05.

The addition of inhibitor to osteogenic media significantly reduced proliferation in the control cultures, although controls did increase in cell number at each time point. (fig 6.5)

After addition of soy extract cell number increased when compared to controls. Cell number did not increase to the level found in media without inhibitor, cell number increases were more apparent at larger doses of soy extract.

6.4.2.5 Summary of soy extract and inhibitor effect on cell number.

Cell number was increased by the addition of soy extract at between 8.45 and 33.79 mg/ml and reduced at larger quantities. This effect can be either reduced or attenuated by the addition of an ER inhibitor, indicating oestrogens are potentially responsible for the observations.

183

6.4.3 Inverted photomicroscopy of cBMSC cultures. 6.4.3.1 Control cBMSC cultures. A B

Fig 6.6 Images of confluent controls, cultured in A) growth medium and B) osteogenic medium for 10 days (x10).

6.4.3.2 HPI stained cBMSCs after culture in osteogenic media and soy extract.

A B

C D

E F

Fig 6.7, Black field Inverted phase micrographs of cBMSCs after 10 days culture with soy extract in osteogenic media, A) control, B) 8.45mg/ml, C) 16.89 mg/ml, D) 33.79 mg/ml, E) 67.57 mg/ml, F) 135.14 mg/ml (x10).

184

6.4.4 Quantification of the activity of ALP.

ALP was measured as per the method described in 2.2.10.2 and protein by the Bradford method as described in 2.2.10.3

6.4.4.1 ALP activity after culture in growth media and soy extract.

200 GM Day 2 GM Day 6 180 GM Day 8 160 GM Day 10 140

120

Protein) 1 - 100 80 * * * * * * * * * * * * * * * ALP (Umg ALP 60 40 20 0 0 8.45 16.89 33.78 67.57 135.14 Soy extract (mg/ml)

Fig 6.8 The activity of ALP measured after culturing cBMSCs with soy extract in growth media for 2, 6, 8 and 10 days. N=4 and the error bars indicate SD. * indicate significance when compared to the control P<0.05.

After culture in growth medium the control cultures (fig 6.8) increased expression of

ALP over time in culture. After culture with media containing soy extract ALP increased from Day 0 to Day 8 at 8.45 and 16.89 mg/ml. At concentrations 33.78-135.14 mg/ml activation of ALP occurred at Day 6 and reduced at subsequent time points.

185

6.4.4.2 ALP activity after culture in growth media with inhibitor and soy extract.

200 GM +I Day 2 GM +I Day 6 180 GM +I Day 8 160 GM +I Day 10 140

120

Protein) 1 - 100 80

ALP (Umg ALP 60 * * * * * * * * * * * * * * 40 20 0 0 8.45 16.89 33.78 67.57 135.14 Soy extract (mg/ml)

Fig 6.9 The activity of ALP measured after culturing cBMSCs with soy extract in growth media and Fulvestrant (ICI 182780) for 2, 6, 8 and 10 days. N=4 and the error bars indicate SD. * indicate significance when compared to the control P<0.05.

After addition of inhibitor to growth media (fig 6.9) ALP activity did not change.

When growth media was supplemented with inhibitor and soy extract the peak activity was reduced. ALP activation was greater than control at Day 2 and 6 after addition of all soy extracts. After addition of soy extracts between 8.45 and 16.89 mg/ml soy extract induced a significant increase when compared to the control.

186

6.4.4.3 ALP activity after culture in beta media and soy extract.

200 Beta Day 2 180 Beta Day 6 Beta Day 8 160 Beta Day 10 140 120

Protein) * 1 - 100 80 * * * * * * * * * * * * * *

ALP (Umg ALP 60 40 20 0 0 8.45 16.89 33.78 67.57 135.14 Soy extract (mg/ml)

Fig 6.10 The activity of ALP measured after culturing cBMSCs with soy extract in beta media for 2, 6, 8 and 10 days. N=4 and the error bars indicate SD. * indicate significance when compared to the control P<0.05.

In beta control medium (fig 6.10) ALP activity was relatively similar over the period of the experiment.

After addition of soy extract there is a large activation of ALP at early time points at all concentrations, this effect is reduced at later time points, but is still significant compared to controls.

187

6.4.4.4 ALP activity after culture in beta media with inhibitor and soy extract.

200 Beta +I Day 2 180 Beta +I Day 6 Beta +I Day 8 160 Beta +I Day 10 140

120

Protein) 1 - 100 80

ALP (Umg ALP 60 * * * * * * * * * * * * * * * * 40 20 0 0 8.45 16.89 33.78 67.57 135.14 Soy extract (mg/ml)

Fig 6.11 The activity of ALP measured after culturing cBMSCs with soy extract in beta media and Fulvestrant (ICI 182780) for 2, 6, 8 and 10 days. N=4 and the error bars indicate SD. * indicate significance when compared to the control P<0.05.

Addition of inhibitor to the beta medium control cultures (fig 6.11) did not significantly influence ALP expression.

Addition of the inhibitor to beta media with soy extract reduced the peak activation of

ALP; the activation profile was similar at all soy extract concentrations, reducing from

Day 6 to Day 10.

188

6.4.4.5 ALP activity after culture in osteogenic media and soy extract.

200 Osteo Day 2 180 Osteo Day 6 * Osteo Day 8 160 Osteo Day 10 140 * 120

Protein) * * * * 1 - 100 80

ALP (Umg ALP 60 40 * * * 20 0 0 8.45 16.89 33.78 67.57 135.14 Soy extract (mg/ml)

Fig 6.12 The activity of ALP measured after culturing cBMSCs with soy extract in osteogenic media for 2, 6, 8 and 10 days. N=4 and the error bars indicate SD. * indicate significance when compared to the control P<0.05.

Osteogenic media controls (fig 6.12) were observed to increase production of ALP over the period of the experiment.

After addition of soy extract to osteogenic media ALP activity increased at all concentrations by Day 6; significantly at less than 67.57 mg/ml and falling at later time points. At 67.57 and 135.14 mg/ml activity was significantly less than controls at Day 8 and 10.

189

6.4.4.6 ALP activity after culture in osteogenic media with inhibitor and soy extract.

200 Osteo +I Day 2 180 Osteo +I Day 6 Osteo +I Day 8 160 Osteo +I Day 10 140

120

Protein) 1 - 100 80

ALP (Umg ALP 60 * * * * 40 20 0 0 8.45 16.89 33.78 67.57 135.14 Soy extract (mg/ml)

Fig 6.13 The activity of ALP measured after culturing cBMSCs with soy extract in osteogenic media and Fulvestrant (ICI 182780) for 2, 6, 8 and 10 days. N=4 and the error bars indicate SD. * indicate not significant when compared with control P<0.05.

Addition of the inhibitor to osteogenic medium (fig 6.13) results in a reduction ALP when compared to medium without inhibitor (Fig 6.12). After addition of soy extract to the media ALP activity is not upregulated to the same degree as osteogenic media alone.

Peak activation is delayed until Day 8 at 8.45 and 16.89 mg/ml.

6.4.4.7 Summary of ALP activity after addition of soy extract.

In all media soy extract induced an increase in ALP activity, this increase was reduced after addition of inhibitor.

190

6.4.5 Analysis of terminal differentiation by cbfa-1 expression.

Expression of the osteoblast differentiation marker cbfa-1 was semi-quantitatively analysed using RT-PCR according to the method described at 2.2.10.7.

6.4.5.1 The effect of soy extract on the expression of cbfa-1 in cBMSCs cultured in growth media and growth media with inhibitor.

7 GM GM+I 6

3 * * * * * 2 * * 1 *

Relative expression Relative (control =0) 0 8.45 16.89 33.79 67.57 135.14 -1

-2 Soy extract (mg/ml) Fig 6.14 The observed relative fold change in expression of cbfa-1 mRNA after treatment of cBMSCs with soy extract for 10 days in growth media and growth media with inhibitor. The change in expression is plotted relative to the control, which is arbitrarily set at 0 and is therefore a plot of the ΔΔCt. N=3, error bars are the SD of the ΔCt. * indicates significance when the ΔCt is compared with the control P<0.05

Expression of cbfa-1 is increased in soy extract supplemented media when compared to growth media controls (Fig 6.14). The peak increase of 2.3 fold was observed after addition of 16.89 mg/ml of soy extract. The increase in expression fell with increasing concentration of soy extract up to 135 mg/ml where there is a significant decrease in expression of cbfa-1. The addition of the inhibitor reduced the effect.

191

6.4.5.2 The effect of soy extract on the expression of cbfa-1 in cBMSCs cultured in osteogenic media and osteogenic media with inhibitor

7 * Osteo 6 Osteo+I * 5

4 * 3 * 2 *

1 *

0 Relative expression Relative (control =0) 8.45 16.89 33.79 67.57 135.14 -1

-2 Soy Extract (mg/ml)

Fig 6.15 The observed relative fold change in expression of cbfa-1 mRNA after treatment of cBMSCs with soy extract for 10 days in osteogenic media and osteogenic media with inhibitor. The change in expression is plotted relative to the control, which is arbitrarily set at 0 and is therefore a plot of the ΔΔCt. N=3, error bars are the SD of the ΔCt. * indicates significance when the ΔCt is compared with the control P<0.05.

After culturing with soy extract in osteogenic media there was a 5.3 fold increase in cbfa-

1 expression after addition of 16.89 mg/ml. expression was lower than controls after addition of 135.14 mg/ml.

The addition of inhibitor to the experiment led to an observed overall decrease in the expression of the cbfa-1 marker, although the dose response was maintained. The peak expression up-regulation occurred at 16.89 mg/ml and was a 1.4 fold increase.

6.4.5.3 Summary of the effect of soy extract on cbfa-1 expression.

Cbfa-1 expression was increased after addition of 8.48 and 33.79 mg/ml and reduced at higher concentrations; the increase was reduced after addition of inhibitor.

192

6.4.6 Alizarin red staining (ARS) of mineralised ECM

6.4.6.1 ARS of monolayers after culture with soy extract in growth media.

300 GM Day 6 GM Day 10 250

200

150

100 Alizarin red stain Alizarin (µM)

0 0 8.45 16.89 33.79 67.57 135.14 Soy extract (mg/ml)

Fig 6.16 The mineralisation measured by alizarin red staining after culturing cBMSCs with soy extract in growth media for 6 and 10 days. N=10 and the error bars indicate SD. All experiments are significant when compared to the control P<0.05.

Culture in growth medium induced the controls (fig 6.16) to increase mineralisation between Day 6 and 10. After addition of soy extract to the media mineralisation was increased at all time points and for all concentrations. A similar dose effect could be observed at both time points.

193

6.4.6.2 ARS of monolayers after culture with soy extract in growth media and inhibitor.

300 GM+I Day 6 GM+I Day 10 250

200

* 150 * * * * *

100 Alizarin red stain Alizarin (µM)

0 0 8.45 16.89 33.79 67.57 135.14 Soy extract (mg/ml)

Fig 6.17 The mineralisation measured by alizarin red staining after culturing cBMSCs with soy extract in growth media and the inhibitor Fulvestrant (ICI 182780) for 6 and 10 days. N=10 and the error bars indicate SD. * indicates significance when compared to the control P<0.05.

After addition of inhibitor to growth media mineralisation of the control cultures did not increase. Addition of extract and inhibitor to the culture media reduces the increase in mineralisation. 33.79 mg/ml and 135.14 mg/ml are the only concentrations that are significantly higher than controls at both time points.

194

6.4.6.3 ARS of monolayers after culture with soy extract in osteogenic media.

300 Osteo Day 6 * * Osteo Day 10 250 * 200

150

100 Alizarin redstain (µM) 50

0 0 8.45 16.89 33.79 67.57 135.14

Soy extract (mg/ml)

Fig 6.18 The mineralisation measured by alizarin red staining after culturing cBMSCs with soy extract in osteogenic media for 6 and 10 days. N=10 and the error bars indicate SD. * indicates significance when compared to the control P<0.05.

In osteogenic media the control cultures (fig 6.18) slightly increased mineralisation over the course of the experiment. After addition of soy extract to the media there was a similar dose response to the observations of growth media. Addition of 135.14 mg/ml was the only extract combination that caused a significant increase in mineralisation at both time points.

195

6.4.6.4 ARS of monolayers after culture with soy extract in osteogenic media and inhibitor.

300 Osteo+I Day 6 Osteo+I Day 10 250

200

150 * * * *

100 Alizarin redstain (µM)

0 0 8.45 16.89 33.79 67.57 135.14

Soy extract (mg/ml) Fig 6.19 The mineralisation measured by alizarin red staining after culturing cBMSCs with soy extract in osteogenic media and Fulvestrant (ICI 182780) for 6 and 10 days. N=10 and the error bars indicate SD. * indicates significance when compared to the control P<0.05.

Addition of inhibitor to the osteogenic medium controls (fig 6.19) had little effect on the mineralisation of the culture. After addition of inhibitor and soy extract to the media the effects observed in media without inhibitor was removed.

6.4.6.5 Summary of the effect of soy extract on mineralisation.

Soy extract increased mineralisation of cultures, an effect that was removed with the addition of inhibitor.

6.4.7 Summary of results.

Addition of soy extract increases cell number, ALP, cbfa-1 expression and mineralisation. These observations can be removed or reduced with the addition of an inhibitor.

196

6.5 Discussion

6.5.1 Comparison with literature.

The osteogenic effect of soy extract when added to cultured human cBMSCs has not been reported in the literature. However the addition of soy extract to immortalised mouse osteoblasts was reported by Choi and Koo in 2001 & 2003. Choi and Koo (2003) studied the effect of soy extract on proliferation, apoptosis and ALP expression on confluent

MC3T3-E1 cells after culture in growth media with 0.05 g/l (50µg ml-1) of soy extract for

48 hours. In their report they observed that the addition of the ethyl acetate fraction of soy extract increases the proliferation and differentiation of mouse osteoblasts. The observed induction in differentiation occurs to a greater degree than with soy extract alone, therefore they suggest that less polar (hydrophobic) compounds are responsible for the increased osteogenic efficacy of the extract. Although they do not test the extracts for isoflavone content genistein and genistin are considered moderately polar compounds

(Zhang et al. 2010).

The HPLC trace at fig 6.1 is run in a manner that increases acetonitrile in the solvent as time increases. Therefore high polarity (hydrophilic) compounds are removed from the column initially, followed at some length by the most polar (hydrophobic) compounds.

Therefore according to the data presented at fig 6.1 we would postulate that the compounds extracted and concentrated in Choi and Koo’ paper were the soy aglycones.

The concentrating of available isoflavones in this fraction of the isolate would account for the authors observed increase in osteogenic effects. This chapter goes further than Choi and Koo, in that the observed effects of the soy extract can be blocked by the addition of an oestrogen receptor inhibitor, therefore the osteogenic effect of soy extract is directly attributed to the presence of phytoestrogens.

197

When Choi and Koo studied cell death in their model they found that addition of soy extract reduced apoptosis and expression of TNF-α. Therefore the authors suggest that soy extract protects against nitrite induced apoptosis, which can be a product of electron chain blockage in mitochondria. The data presented here in growth media at Day 2 contains soy extract at 1000 fold more concentrated, at this concentration and time point there is increased cell number and ALP activity; data which fits with the Choi and Koo data. Because Choi and Koo do not characterise their extract for phytoestrogen content it is difficult to pick apart the reasons for this discrepancy, it could be a difference between immortalised mouse cells and primary human or simply that they extraction performed by

Choi and Koo did not concentrate the isoflavones to the same degree as reported in this chapter.

The data presented here also shows that the choice of media is important for measuring the effects of soy extract; there is a significant difference between growth media and osteogenic media. The similarity to previous chapters and the significant reduction in efficacy of the soy extract when used in conjunction with the inhibitor suggests that the majority of the osteogenic effect of the soy extract is due to the presence of soy phytoestrogens. Perhaps due to the use of a combination of flavones the effect of soy extract in this chapter is significantly greater than the effect of pure soy isoflavones in the previous chapters.

Other studies have suggested that soy extract is more potent than pure soy isoflavones, for example in Kim et al (2008) published an in vivo study observing that human breast cancer cells injected into athymic nude mice proliferated when treated with pure genistein but were significantly reduced with soy extracts. Similarly Gallo et al (2008) found that a soy extract double selected for isoflavones is effective at delaying the progression of lung

198

cancer in the same mouse model. The observed increased potency of soy extract is potentially due to the combination of phytoestrogens or perhaps processing has reduced conjugated isoflavones into active aglycones (de Lima Toccafondo Vieira et al. 2008).

6.5.2 Comparison with previous chapters

The results presented in this chapter have on the whole no comparable published studies, although studies do exist for pure isoflavones or combinations thereof there are few human in vitro studies using soy extract and no observations of osteogenesis. As can be seen from the table at 2.2.13.5 the soy extract added to the media was adjusted for genistein content, therefore the data presented here can be directly compared with previous chapters. When we compare this chapter with the previous chapters we find ongoing trends; there is a dose response from the soy extract that indicates osteogenic effects at the lowest concentration studied 8.45 mg/ml, because this effect can be removed by the application of inhibitor we can deduce that the effective components of the extract are phytoestrogens.

The cumulative effects observed when we supplement genistein to osteogenic media are also maintained when soy extract is added to the osteogenic media, although this effect is only apparent at concentrations ≤33.78 mg/ml (Fig 6.13). In this chapter we use the combination of ALP and cbfa1 expression to monitor the differentiation state of cBMSCs. The osteoblast marker cbfa-1 has been often used as an indicator of phenotype in differentiating cBMSC systems as its expression is obligatory during osteogenesis and it is known to regulate the production of osteocalcin (Guillot et al 2008, Maromet al

2005). Treatment with soy extract at 16.89 mg/ml in growth media is shown here to be sufficient to increase ALP activation by more than 1 fold and cbfa1 expression by more than 2 fold. When supplemented to osteogenic media 16.89 mg/ml soy extract increases

199

the activity of ALP by 4.5 fold and cbfa1 expression by 5.3 fold. Adding inhibitor to these experiments causes a reduction in the effectiveness of the isoflavones by about

50%, as is expected from a competitive inhibitor (Heim et al 2004). Overall these observations indicate that the most effective method to differentiate cBMSCs to osteoblasts is to supplement osteogenic media with 16.89 mg/ml of soy extract, containing 1.88 µM of genistein.

6.6 Conclusions.

The hypothesis that soy extract is more efficacious at differentiating BMSCs than purified genistein alone and that the osteogenic effect of the soy extract can be removed with the addition of an oestrogen receptor inhibitor is proved.

The removal of effect with the addition of inhibitor indicates that the observations are due to the phytoestrogen content of the soy extracts.

Cell death is less pronounced after culture with soy extract than with pure genistein.

Soy extract is significantly osteogenic in both growth media and osteogenic media; this indicates there is a cumulative effect in addition to the osteogenesis mediated by the dexamethasone and β-glycerophosphate in the osteogenic media.

Cbfa-1 can be used as a marker of osteogenesis in this model as it is upregulated along with ALP.

200

CHAPTER 7

DIFFERENTIATION OF CBMSCS ON CALCIUM PHOSPHATE

CEMENTS.

7.1 Introduction.

Calcium phosphate (CaP) ceramics have been used to fill small bone defects since 1920

(Albee 1920). Since the 1980’s research has concentrated on biphasic calcium phosphate

(BiCaP) which is a combination of two ceramics, hydroxyapatite (HA) and β-tricalcium phosphate (β-TCP) (Jarcho 1981). The ratio of HA to TCP was optimised during the late

1980’s and early 1990’s by Daculsi et al (reviewed Daculsi et al 2003). Both block and granule forms of BiCaP ceramics have been used to fill bone defects (Eggli et al. 1988,

Gauthier et al. 2005).

Block and granular BiCaP are useful because they can be carved or compressed into bone defects. In addition to block or granular BiCaP, a self setting cement was introduced that can be injected and moulded to fit defects perfectly, the setting of the cement used the work by Daculsi to optimise the reaction for the production of calcium deficient hydroxyapatite (CDHA) (Chow and Takagi 2001). However all forms have problems with mechanical stability during normal body stresses (Habibovic et al. 2006,

Schumacher et al. 2010). Therefore these materials are mostly used where adequate mechanical stability and host bone tissue is already present, foremost in craniofacial surgery.

Materials made from BiCaP are bioactive and osteoconductive; they form a direct interconnection with the surrounding tissue and encourage cell penetration and remodelling (Levengood et al. 2010). It is known that the success or failure of these

201

materials is due to the porosity and solubility of the construct, resorption and remodelling of the CaP cement into bone is a critical clinical property (LeGeros et al. 2003,

Kobayashi et al. 2009).

Because the porosity of the cement is critical to the speed of tissue remodelling an injectable CaP foam was investigated by Ginebra et al. (2006) as a mechanism of introducing macroporosity. Previously reports had followed either subtractive logic to introduce porosity; they were partially comprised of sugars or polylactose fibres that wash out of the set cement (Markovic et al 2000, Xu and Quinn 2002). Alternatively porosity can be generated within the cement during setting by breakdown of bicarbonate or similar (del Real et al 2002, Hesaraki et al 2007). Ginebra introduced the use of a biological surfactant to stably foam the cement prior to injection without introducing toxic or incompatible reagents (Ginebra et al 2006).

Further work has shown that the introduction of gelatine as a foaming agent decreases the setting time and increases the strength and pore size of the composite when compared with the cement alone (Montufar et al 2010). In good agreement with the literature

(Fujishiro et al 2001, Bigi et al. 2004). Happily the authors also report that the use of gelatine increases the injectability of the cement paste without compromising the formation of CDHA (Montufar et al. 2010). Other reports in the literature indicate that addition of gelatine to CaP cements was detrimental to the setting time but agreed that compressive strength increased significantly (Shie et al. 2008).

Because CaP cements set at low temperature it is possible to introduce bioactive molecules into the cement prior to use to increase performance and for drug delivery. If the cement is intended for use in a patient with a history of infection then the use of a slow release antibiotic in combination with the cement could allow the control of

202

bacterial load while integration with the host tissue occurs (Jiang et al. 2010). The addition of an osteogenic component like bone morphogenic protein (BMP) to a CaP cement can induce osteogenesis, actively recruiting and committing progenitor cells to an osteogenic lineage (Polak et al. 2011, Wang et al. 2011). These osteoinductive BiCaP cements are the next generation of bone regenerative templates. Since isoflavones are oestrogenic their incorporation in a cement should also increase osteogenesis; and when cultured with SAOS-2 cells soy cements did increase osteogenesis (Perut et al. 2011).

Soy products have a long standing use as engineering materials; famously Henry Ford used soy based bioplastics in the 1940’s to construct car parts (Murray 2007). More recently soy has been incorporated into composite plastics to enhance biodegradation

(Reddy and Yang 2011). Because soy based plastics have enhanced biodegradability they have also been developed to replace aliphatic polyester biomaterials (Vaz et al. 2003).

Soy has also been added to biomaterials in order to enhance the strength and pore size of the structure (Luo et al. 2008). Therefore soy has properties that make it ideal for use in a bone regenerative material; it contains phytoestrogens, allows modulation of porosity and can enhance degradation.

The formulation of a soy based osteogenic CaP cement has allowed the localised delivery of pharmaceutically relevant concentrations of soy extract that have the potential to initiate osteogenesis in cell lines (Santin et al. 2008).

Combining osteoconductive cement with a pharmaceutical concentration of an active and eluted ingredient allows optimisation of the material for regeneration and integration into bone (Heliotis et al. 2006, Luvizuto et al. 2011). The use of soy extract as an initiator of osteogenic differentiation was proven in the previous chapters with the intent that the extract would be combined with a CaP cement that would have an intrinsically greater

203

osteogenic effect than the original formulation alone. The addition of human osteoprogenitors may accelerate the engraftment process (McGonagle et al 2007). The use of human bone marrow stem cells is already recommended in clinic for treatment of host vs graft disease; alternative clinical uses for BMSCs may be easier to licence once efficacy and safety has been proven over the mid to long term (Sato et al. 2010).

Human BMSCs are at the forefront of bone research, therefore there is literature available describing the effect of cements on differentiation as well as BMSC growth on cements.

A comparative study of BMSC loaded grafts using either biphasic CaP or natural bovine bone mineral (bio-oss) was undertaken in dogs by Jafarian et al (2008), finding no significant difference in efficacy. Also in 2008 Mankani et al demonstrated that HA/TCP and human BMSC can be directly injected into the dorsal subcutaneous space of mice resulting in de novo bone formation without the necessity of open surgery. More recently

Li et al. (2011) cultured human adipose-derived stem cells on BiCaP ceramic discs resulting in osteogenic differentiation and increased mineralisation of the scaffolds.

Human clinical trials have described the use of autologous human BMSCs in bespoke manufactured macroporous HA constructs, finding that healing was accelerated by about

6 months in two patients (Quarto et al. 2001). Bone was also successfully regenerated in clinic by Schimming and Schmelzeisen (2004) who were successful in transplanting 18 patients with a construct comprising of autologous periosteum derived osteoblasts on a resorbable Ethisorb matrix, although the scaffold was not CaP based this study showed stem cell derived de novo bone production is possible in multiple patients. Also in 2004,

Kitoh et al. used autologous BMSC and platelet-rich plasma to form an injectable fast setting gel in patients undergoing leg lengthening and observed an acceleration in healing times.

204

BMSC have been analysed after culture on CDHA cements which were used as a substrate to culture rabbit BMSCs (isolated at unknown density) after pre-treatment in osteogenic medium for one week. After addition to CDHA and hydroxyapatite (HA) control substrates and continued culture in osteogenic medium there was an increase in

ALP activity and proliferation. When these scaffold and cell constructs were implanted into rabbit mandibles for 8 weeks no differences were observed between normal bone and the remodelled scaffold (Gou et al 2009).

Human BMSC isolated at clonal density were cultured on CDHA cements by Park et al

(2011), who observed increased proliferation, collagen, ALP, and osteopontin production after 6 weeks culture. The effect of a soy extract calcium phosphate cement on human

BMSC differentiation is unknown.

7.2 Hypothesis

As outlined in the introduction injectable foamed cements for the replacement of bone tissue have been reported in the literature previously. We postulated that the addition of soy extract containing the phytoestrogens described in the previous chapters would be of benefit, that increased differentiation could be achieved in vitro compared with gelatine control cements. Therefore we hypothesise that a soy cement biomaterial will increase the proliferation and osteogenic differentiation of a primary human cBMSC culture.

7.3 Methods

Cements were produced by the method described at 2.2.12, outlined briefly here. Gelatine was warmed to 50°C; soy extract was added for soy cements, for the control it was left out. The warmed gelatine solution was then foamed and combined with the CaP, mixed, and then injected into moulds. The cements were immersed in saline for 12 days prior to cell culture. cBMSCs were added to the cements using the protocol at 2.2.14; 4.5x104

205

cells were seeded to set cements in growth medium and cultured for up to 24 days prior to analysis of the cultures for cell number and differentiation by the methods described at

2.2.10. Cements were imaged by SEM, method at 2.2.15, in brief; fixed and freeze dried materials were sputter coated with Palladium and visualised under high vac in an SEM.

7.4 Results

7.4.1 Visual comparison of cements.

Soy and gelatine mixtures foamed more easily than gelatine alone, the amount of foaming was more consistent and durable. Injection of the mixed cements to moulds was more easily accomplished in the soy containing cements.

7.4.1.1 Photographs of set cements.

In the finished soy/gelatine (SG) cements (fig 7.1B) there was a reduction in the friable nature of the gelatine (G) cement, the surface was more robust. The bubbles in both foams resulted in a visible pore structure but pores were more numerous in SG cements.

SG cements were better able to fill the moulds without introducing voids or requiring subsequent packing.

A B

Fig 7.1, photographs for comparison of A) gelatine control B) soy/gelatine cement.

206

7.4.1.2. SEM of set cements.

SEM at relatively low magnification confirmed that SG cements had increased porosity compared with G cements. At higher magnification (fig 7.2) there is increased pore size and frequency in SG cements, while in the G control hydroxyapatite crystals appear more numerous, smaller and densely packed.

Fig 7.2, SEM of cement biomaterials A) gelatine B) soy/gelatine.

207

7.4.1 Proliferation of cBMSCs on cements.

500 gelatine cement *

450 )

3 soy/gelatine cement 400 350 300 250 200 150 * LDH derived LDH number cell (x10 100 50 0 0 12 24

Time (Days) Figure 7.3, LDH derived cell number after cBMSCs were cultured on materials for 0, 12 and 24 days. Error bars indicate SD. n=6. Cell number increases significantly on both materials at each time point. * indicates a significantly greater cell number when cements are compared at each time point P<0.05.

The proliferation of cBMSCs on cements was quantified using LDH content (fig 7.3).

Cells were seeded at 4.5x104 cells per biomaterial, at Day 12 there is no difference in cell number between the two biomaterial types. By Day 24 there was a statistically greater number of cells on the SG cements when compared with the G cements.

208

7.4.2 Expression of cbfa-1 after culture of cBMSCs on cements.

7 *

1 in CBMSCs in CBMSCs 1 - 4

cement controls. cement 2 *

Relative expression cbfa of expression Relative 0 cultured on SG cements when compared G to compared when cements SG on cultured 0 12 24

Time (Days)

Fig 7.4. Semi-quantified expression of the osteogenic marker cbfa-1 in cBMSCs cultured on biomaterials for 0, 12 and 24 days. The data is expressed as the ΔΔCt, where expression of cbfa-1 is adjusted for loading error to the expression of gapdh (the ΔCt) a comparison is then made of the ΔCt‘s from cBMSCs cultured on G cements relative to SG cements (ΔΔCt). n=5 and error bars indicate SD. * indicate significantly greater ΔCt when cultured on SG cements than G cements.

Analysis of the expression of cbfa-1 in cBMSCs on SG cements compared to G cements

(Fig 7.4) shows that although there is no comparative increase in cell number after 12 days incubation there is a significant 1 fold (100%) increase in cbfa-1 RNA production.

At 24 days there is a 5.3 fold increase in cbfa-1 expression in SG cultured cBMSCs relative to the G cement.

209

7.4.3 SEM of cements with CMBSCs.

After culture of cBMSC with the biomaterials for 24 days the cements were harvested and the surface was observed by SEM. Cell sheets covering the surface of the cements are visible (fig 7.5B), it is apparent that there is deposition on the surfaces of both cements. The surface of the SG cements appears more homogenous and consists of cells and a surface coating. The G cement appears to have less incorporation of the cells into the modification of the surface.

Figure 7.5, SEM of cements after 24 days culture with CMBSCs. A) G, B) SG. ECM covering the surface of the cements is visible in both images; however the cells cultured on the SG cement appear completely covered when compared with the G cement.

210

7.4.2 Soy/gelatine cement remodelling.

Concurrent with cells differentiating on the SG cement there was also a remodelling effect on the cement surface. SEM shows that after cBMSCs have been cultured on the surface of the SG cement the plate shaped crystals are replaced with a smaller scale amorphous surface (fig 7.6).

Figure 7.6, SEM of SG cement before (A) and after (B) the growth of cBMSC on the surface for 24 days.

211

7.4.3 Gelatine cements after culture with cBMSCs

The fine structure of the surface of the G cements under the SEM (fig 7.7) appears coated and is partially concealed. The modifications observed in SG cements are not apparent.

Fig 7.7, Gelatine cements A) before and B) after culture with cBMSC for 24 days.

212

7.5 Discussion.

The gelatine and soy/gelatine cements used in this chapter have previously been described in the literature. Foamed CaP cements were described by Ginebra et al (2006) using albumen as a natural foaming agent to impart additional porosity, this also lowered compressive strength. Addition of albumen did not prevent the formation of calcium deficient hydroxyapatite (CDHA), which is the mechanism by which the cement is set.

The use of albumen foam as a porogen decreased the microporosity by 10% and increased macroporosity by up to 56%, increasing total porosity as much as 77% overall.

The foaming technique was developed in the subsequent report, Montufar et al (2010a), using polysorbate-80 as a foaming reagent the porosity was found to be controlled by the ratio of polysorbate used in the liquid phase of the cement. Use of low concentrations of polysorbate (<5%) had no effect on the setting reaction, but increased both injectability and porosity. The tailored use of this porogen allowed optimisation of the cement for the growth of the human osteosarcoma SAOS-2 cell line, with a concurrent increase in expression of osteoblastic markers without addition of osteogenic medium.

Gelatine foamed cements were also described by Montufar et al. (2010b) with improvements in injectability, setting time, porosity and compressive strength. While addition of gelatine maintained the microporosity of the cement it also increased the macroporosity, total porosity reached as much as 74%. The authors did not culture cells on or with the cements.

The foamed soy extract and gelatine combined calcium phosphate cement was described in 2011 by Perut et al. The 20S5G cements described in the paper by Perut et al are identical to the SG cements presented in this chapter. The liquid to powder (L/P) used in

213

this chapter was the adjusted value of 0.55 ml g-1 as described in the paper, this ratio was used to ensure cohesion and injectability, and to provide comparability between the work.

This SG cement has a wide range of pore sizes, attributed to a combination of the foam and the innate porosity of the CaP cement. The most common pore size was 20µm, this was attributed to the narrowest interconnectivity between pores as at the surface pores were larger than 20µm. X-ray diffraction analysis of the setting reaction products indicated that addition of soy extract increased the setting time in the early period. After 8 hours the SG cements were 15% CDHA, compared with 40% in the G cements. During this setting period all the cement formulations demonstrated an increase in surface area, this increase is related to the formation of plate like HA crystals (fig 7.2) which in soy cements were larger and had better separation than gelatine cements (fig 7.6A).

The SG cements increased surface area to a greater degree than is accountable for by just cement precipitation, the increased porosity and pore interconnectivity compaired with G cements may account for this observation. It is an interesting observation that this increase in surface area and porosity develops over the period of setting time. At 168 hours the SG cement is 85% set and has the same surface area as the already set non- foamed control cement. When the SG cement is completely set the surface area has increased significantly indicating that either there is a very large increase in nucleation of

HA crystals as the last αTCP dissolves and precipitates or there is a washing out of soy extract, freeing more nucleation sites.

The authors use the human osteosarcoma cell line SAOS-2 to culture cells both directly and indirectly in contact with the cements. When cultured indirectly with SG cements the proliferation and collagen deposition of SAOS-2 cells were not different from the control cements, however ALP was observed to increase significantly. When SAOS-2 cells were

214

cultured directly on the SG cements an increase in Alamar blue reduction was observed, this reduction was attributed to proliferation on the cement. This observation was confirmed by direct visualisation of the Hoechst stained cells on the cement surface.

Relative to control cements culture of SAOS-2 cells directly on SG cements did not induce an increase in ALP; however, when compared to indirect culture all cements induced an increase in ALP.

Increased differentiation of cells on many CaP cements has been described in the literature previously and has been attributed to the localised increase of Ca+2 ions in solution, surface topology and the type of HA deposited (Pi et al. 2005, Guo et al. 2009,

Park et al. 2011, Ewald et al. 2011). The effect of the active soy isoflavones present in the cement, which was significant when cultured indirectly, would be expected to increase differentiation markers more than controls in direct contact, even in non-osteogenic medium. As demonstrated in Chapters 4-6 human osteosarcoma cell lines do not respond in the same manner as primary multipotent BMSCs when cultured with isoflavones.

Soy biomaterials have also been used in a pre-clinical rabbit model Giavaresi et al (2010).

The material comprised of 320 mg of soy extract, prepared as described at 2.2.11.1, which was formed into a hydrogel by the addition of CaCl2 solution. This hydrogel was mixed 50/50 with thermoset soy curd granules and was implanted into critical sized defects in the distal femur of New Zealand rabbits. In the control treatments the defects were non-healing at 24 weeks. When the granules in the soy material were not packed densely increased trabecular growth was observed at 8 and 24 weeks compared with commercially available Fisiograft. Compaction of the packing commonly occurs when

HA granules are used in clinic; some reports indicate almost no resorption several years after the surgery (Mendelson et al. 2010).

215

The aim of combining the cement with the soy extract was to create a product that was superior to the cement alone. When primary human cBMSCs were cultured directly on the surface of the SG cement cell numbers increased and osteogenic markers were up regulated. In this chapter ALP, collagen and mineralisation were not considered as markers. This was because ALP extraction would co-extract protein from the soy extract and gelatine present in the cements, skewing the results when they were presented adjusted for protein. Collagen has been measured by absorption of picrosirius red in previous chapters, which also stains gelatine; this signal drowns the very small differences that have been observed in previous chapters. Measuring increased mineralisation on CaP cements is similarly not possible.

Therefore semi quantitative RT-PCR was used to measure the relative expression of cbfa1, which is upregulated during osteogenesis and regulates the production of osteopontin and other mineralisation genes (chapter 6). The data in this chapter indicates that the oestrogenic effect observed when extract was directly applied to cBMSCs was delayed.

Further work by the group suggests that after setting there is a 4 day isoflavone release spike followed by sustained release for a further 10 days, further time points were not taken (appendix 2). This slow release of isoflavones from the SG cements may account for the slower osteogenesis on cements compared with direct application of extract.

7.6 Conclusions.

The hypothesis that the addition of soy extract to CaP cements will increase in vitro proliferation and osteogenic differentiation of primary human multipotent BMSCs was proven to be true. Soy CaP cements are likely to have a beneficial osteogenic effect in vivo.

216

CHAPTER 8

CONCLUSIONS AND FUTURE WORK

8.1 Scientific background of the work

Bone regenerative biomaterials are in demand as adjuncts and replacements to autograft

(Brandoff et al. 2008). Current synthetic materials are often not resorbed or integrated into the patient tissue (Stellingsma et al. 2004). The addition of pharmaceuticals to biomaterials can accelerate, modulate or prevent host responses to the implant and therefore the selection of materials tailored to a desired patient outcome becomes relevant

(Brydone et al. 2010).

Calcium phosphate cements are considered to be low strength osteoconductive materials that can be loaded with growth factors to increase the differentiation of in situ progenitors to osteoblasts (Brydone et al. 2010). Some authors have added osteoprogenitors to cements, thereby creating a true osteogenic material (Park et al. 2011).

The use of growth factors in vivo is problematic, recombinant BMP is expensive and therefore only indicated in the most severe cases (Garrison et al. 2010). However calcium phosphate cements are only indicated for use in non-load bearing craniofacial surgery, with mechanical stabilisation (Ambard and Mueninghoff 2006). It is unlikely that the outcomes with BMP enhanced cements would be so significant that its use would be endorsed by NICE (Garrison et al. 2007).

An additional cement that is osteoinductive without the added cost of growth factor treatment is therefore desirable. Soy has potential as an enhancer of osteoinduction because; isoflavones are oestrogenic and have been shown to improve outcomes in osteoporotic menopausal women, soy protein is a commonly used emulsifier and is a

217

natural amphiphilic surfactant and soy is very cheap and abundant (Barnes 2010, Bolanos and Francia 2010). Therefore soy was incorporated into a CaP cement with the intention of improving outcomes compared to cement without osteoinduction and without the associated cost of recombinant human growth factors (Perut et al. 2011).

The osteogenic capabilities of the soy extract and the soy cement on the primary differentiating cells that are present in bone was unknown.

8.2 The hypothesis of this work

We hypothesised in section 1.12 that the addition of soy extract to calcium phosphate cements would prove to be osteogenic when primary human BMSCs were cultured in direct contact. This hypothesis has proven to be correct.

This study has contributed to our understanding of some of the gaps in the literature.

Previous reports of MG63 cultures with genistein failed to report any increase in markers of osteogenesis (Morris et al 2006). It was reasoned that the culture of MG63s in a confluent model would accelerate differentiation as well as limiting proliferation; therefore any osteogenic effect attributable to genistein would be highlighted using the model described here. The confluent model shows that genistein has an osteogenic effect on MG63s, in agreement with the hypothesis.

The osteogenic effect on BMSCs of supplementary isoflavone in more than one medium has also not been previously reported, most authors use osteogenic media supplemented with isoflavones. It was hypothesised that isoflavones are sufficiently osteogenic alone and that other components of the medium are not required for differentiation, this proved to be the case.

218

The osteogenic effect of complex soy extracts on human BMSCs was previously unknown, as was the combination in a calcium phosphate cement. Therefore it was hypothesised that soy extract containing a complex mixture of phytoestrogens would have a greater osteogenic effect than single isoflavones, both alone and in combination with known osteogenic agents. This was shown to be correct.

It has previously been shown that pure soy isoflavones are osteogenic in primary human cell culture and cell line culture. This osteogenic effect can be modulated by the addition of an oestrogen receptor inhibitor; therefore the observed effects are attributable to isoflavones (Heim et al. 2004, Morris 2006). We utilised this research to hypothesise that the addition of an ER inhibitor would modulate the osteogenic effect of soy extracts demonstrating that the active components of the extract is phytoestrogens. The results presented here clearly show that this hypothesis is correct.

8.3 Findings of the work.

8.3.1 MG63 cell line observations

The confluent MG63 osteosarcoma cell line model described here was useful to observe the action of genistein on markers of differentiation alone and in combination with components of osteogenic medium after short culture periods. A previous report has shown addition of genistein at the concentrations used in this study reduced proliferation and early markers of osteogenesis, while increasing mineralisation (Morris 2006). Our findings in this model are that genistein significantly increases osteogenic markers and shortens differentiation time in osteogenic medium.

8.3.2 Primary human BMSCs

In primary human BMSC pure genistein induces a greater response than in the MG63 cells, again a larger effect in osteogenic media, but genistein is osteogenic alone. The

219

addition of genistein up regulates the relevant osteogenic markers and there is an immediate and persistent metabolic response after only 2 hours incubation. These results agree with the literature, genistein increases osteogenic markers when added to osteogenic medium (Heim et al. 2004). We also observed that genistein increased osteogenesis in culture media without further supplementation, which has not previously been reported.

In the same human BMSCs strain soy extract is more osteogenic than pure genistein. The osteogenic effect of the soy extract can be decreased with the addition of a competitive inhibitor, indicating that the isoflavones within the extract are responsible for the osteogenesis. This osteogenic effect occurs in culture medium, but is greater in osteogenic media.

When cultured on the surface of calcium phosphate cements containing either soy and gelatine or gelatine alone, BMSCs significantly increased markers of osteogenesis on the soy extract cements when compared with the gelatine cements, the surface of the cement was also remodelled more after culture with the soy extract.

8.4 Novel contribution

This thesis has shown that genistein is osteogenic when cultured with MG63s; it is inherently osteogenic but also acts sympathetically with osteogenic media to increase markers of osteogenesis. The model and culture methods outlined in chapter 3 were also shown to be effective when observing the differentiation of primary human BMSCs.

Observations of human BMSCs after addition of soy extract allowed the conclusion that the extract induces the osteogenic differentiation; alone or as a supplement to osteogenic medium. The osteogenesis was proven to be due to the presence of isoflavones in the soy

220

extract because the addition of an estrogen receptor binding competitive inhibitor reduced the observed markers of osteogenesis.

The addition of soy extract to calcium phosphate cements significantly increased the osteogenic differentiation of human BMSCs that were cultured on the surface. This finding encourages the development of osteogenic soy cements that are efficacious and economical in use.

8.5 Future work

The co-injection of autologous BMSCs and cements is the next logical step in the production of a bone filler to replace autograft. In order for this to be a feasible proposition the cells would need to be protected from the cement in some manner; this could conceivably be done by the formation of micro gelatine beads containing cells. It has already been reported in the literature that encapsulation in alginate is effective at protecting human BMSCs from the calcium phosphate setting reaction (Zhao et al. 2011).

An adaptation of this method for the proposed soy/gelatine cement should be achievable.

Further tailoring of the soy cement should be considered; this study has provided data showing the effect of a wide range of concentrations on the osteogenic differentiation of

BMSCs. The data provided could be the basis for further work to perfectly balance the isoflavone content and release from the cement. To achieve this goal experiments should include a combination of in vitro cement culture as described in Chapter 7 and in vivo experimentation of optimised cement and cement with BMSCs should be undertaken.

In addition to the incorporation of autologous cells in CaP cements there is also a great deal of work still to be undertaken in order to further our understanding of the mechanisms of isoflavone action. Proteomic studies of different effects of oestrogen receptor ligands may help design or isolate the most potent isoflavone. There are tissue

221

specific effects of the activated oestrogen receptor isoflavone ligand, therefore the proteomic studies should be carried out in a differentiating human primary cell strain capable of osteogenesis. BMSCs are a prime candidate for such studies.

222

References

Acharya, A., A. J. Timperley, et al. (2011). "Properties of composite specimens of old and new bone cement." Proc Inst Mech Eng H 225(1): 86-93.

Adegunloye, B. J., X. Su, et al. (2003). "Sensitivity of rabbit aorta and mesenteric artery to norepinephrine: role of tyrosine kinases." Eur J Pharmacol 476(3): 201-9.

Adlercreutz, H. (2002). "Phyto-oestrogens and cancer." The Lancet Oncology 3(6): 364-373.

Adlercreutz, H., S. M. Heinonen, et al. (2004). "Phytoestrogens, cancer and coronary heart disease." Biofactors 22(1-4): 229-36.

Adlercreutz, H., P. Kiuru, et al. (2004). "An isotope dilution gas chromatographic-mass spectrometric method for the simultaneous assay of estrogens and phytoestrogens in urine." J Steroid Biochem Mol Biol 92(5): 399-411.

Adlercreutz H and W Mazur (1997). "Phyto-oestrogens and Western diseases." Ann Med 29(2): 95-120.

Aggarwal, S. and M. F. Pittenger (2005). "Human mesenchymal stem cells modulate allogeneic immune cell responses." Blood 105(4): 1815-22.

Aicher, W., HJ Bühring, et al. (2010 Dec). "Regeneration of cartilage and bone by defined subsets of mesenchymal stromal cells--Potential and pitfalls." Advanced Drug Delivery Reviews In Press, Corrected Proof.

Akbas GE, J Song et al. (2004)"A HOXA10 estrogen response element (ERE) is differentially regulated by 17 beta-estradiol and diethylstilbestrol (DES)." J Mol Biol 340(5): 1013-23.

Akhyari P, PWM Fedak et al. (2002). "Mechanical Stretch Regimen Enhances the Formation of Bioengineered Autologous Cardiac Muscle Grafts." Circulation 106(90121): I-137-142.

Akintoye, S. O., T. Lam, et al. (2006). "Skeletal site-specific characterization of orofacial and iliac crest human bone marrow stromal cells in same individuals." Bone 38(6): 758-768.

Albee, F. H. (1920). "STUDIES IN BONE GROWTH: TRIPLE CALCIUM PHOSPHATE AS A STIMULUS TO OSTEOGENESIS." Ann Surg 71(1): 32-9.

223

Alborzi, A., K. Mac, et al. (1996). "Endochondral and intramembranous fetal bone development: osteoblastic cell proliferation, and expression of alkaline phosphatase, m-twist, and histone H4." J Craniofac Genet Dev Biol 16(2): 94-106.

Alexi, X., K. M. Kasiotis, et al. (2009). "Differential estrogen receptor subtype modulators: assessment of estrogen receptor subtype-binding selectivity and transcription-regulating properties of new cycloalkyl pyrazoles." J Steroid Biochem Mol Biol 117(4-5): 159-67.

Altavilla, D., A. Crisafulli, et al. (2004). "Cardiovascular effects of the phytoestrogen genistein." Curr Med Chem Cardiovasc Hematol Agents 2(2): 179-86.

Ambard, A. J. and L. Mueninghoff (2006). "Calcium phosphate cement: review of mechanical and biological properties." J Prosthodont 15(5): 321-8.

Andl, C. and A. Rustgi (2005). "No one-way street: Cross-talk between E-cadherin and receptor tyrosine kinase (RTK) signaling-A mechanism to regulate RTK activity." Cancer Biol Ther 4(1): 28 - 31.

Arca, T., J. Proffitt, et al. (2011). "Generating 3D tissue constructs with mesenchymal stem cells and a cancellous bone graft for orthopaedic applications." Biomed Mater 6(2): 025006.

Arinzeh, T. L., S. J. Peter, et al. (2003). "Allogeneic mesenchymal stem cells regenerate bone in a critical-sized canine segmental defect." J Bone Joint Surg Am 85-A(10): 1927-35.

Arinzeh, T. L., T. Tran, et al. (2005). "A comparative study of biphasic calcium phosphate ceramics for human mesenchymal stem-cell-induced bone formation." Biomaterials 26(17): 3631-8.

Arufe, M. C., A. De la Fuente, et al. (2010). "Chondrogenic potential of subpopulations of cells expressing mesenchymal stem cell markers derived from human synovial membranes." Journal of Cellular Biochemistry 111(4): 834-845.

Atteritano, M., H. Marini, et al. (2007). "Effects of the phytoestrogen genistein on some predictors of cardiovascular risk in osteopenic, postmenopausal women: a two-year randomized, double-blind, placebo-controlled study." J Clin Endocrinol Metab 92(8): 3068-75.

Augello, A., T. B. Kurth, et al. (2010). "Mesenchymal stem cells: a perspective from in vitro cultures to in vivo migration and niches." Eur Cell Mater 20: 121-33.

Banfi, G., G. Lombardi, et al. (2010). "Bone metabolism markers in sports medicine." Sports Med 40(8): 697-714. 224

Barnes, S. (1998). "Evolution of the health benefits of soy isoflavones." Proc Soc Exp Biol Med 217(3): 386-92.

Barnes, S. (2010). "The biochemistry, chemistry and physiology of the isoflavones in soybeans and their food products." Lymphat Res Biol 8(1): 89-98.

Barralet, J. E., K. J. Lilley, et al. (2004). "Cements from nanocrystalline hydroxyapatite." J Mater Sci Mater Med 15(4): 407-11.

Barry, F., R. Boynton, et al. (2001). "The SH-3 and SH-4 antibodies recognize distinct epitopes on CD73 from human mesenchymal stem cells." Biochem Biophys Res Commun 289(2): 519-24.

Barry, F., R. E. Boynton, et al. (2001). "Chondrogenic differentiation of mesenchymal stem cells from bone marrow: differentiation-dependent gene expression of matrix components." Exp Cell Res 268(2): 189-200.

Barry, F. P. (2003). "Biology and clinical applications of mesenchymal stem cells." Birth Defects Res C Embryo Today 69(3): 250-6.

Barry, F. P. (2003). "Mesenchymal stem cell therapy in joint disease." Novartis Found Symp 249: 86-96; discussion 96-102, 170-4, 239-41.

Barry, F. P., R. E. Boynton, et al. (1999). "The monoclonal antibody SH-2, raised against human mesenchymal stem cells, recognizes an epitope on endoglin (CD105)." Biochem Biophys Res Commun 265(1): 134-9.

Barry, F. P. and J. M. Murphy (2004). "Mesenchymal stem cells: clinical applications and biological characterization." Int J Biochem Cell Biol 36(4): 568-84.

Battula, V., P. Bareiss, et al. (2007). "Human placenta and bone marrow derived MSC cultured in serum-free, b-FGF-containing medium express cell surface frizzled-9 and SSEA-4 and give rise to multilineage differentiation." Differentiation. 75(4): 279-91.

Baxter, M. A., R. F. Wynn, et al. (2004). "Study of Telomere Length Reveals Rapid Aging of Human Marrow Stromal Cells following In Vitro Expansion." STEM CELLS 22(5): 675-682.

Becker, A. J., E. A. McCulloch, et al. (1963). "Cytological Demonstration of the Clonal Nature of Spleen Colonies Derived from Transplanted Mouse Marrow Cells." Nature 197: 452 - 454

Beekmans, H. C., G. J. Meijer, et al. (2008). "The hydroxylapatite-bone interface: 10 years after implant installation." Int J Oral Maxillofac Surg 37(8): 768-72.

225

Beggs, K. J., A. Lyubimov, et al. (2006). "Immunologic consequences of multiple, high-dose administration of allogeneic mesenchymal stem cells to baboons." Cell Transplant 15(8-9): 711-21.

Bell, G. I., H. C. Broughton, et al. (2011). "Transplanted human bone marrow progenitor subtypes stimulate endogenous islet regeneration and revascularization." Stem Cells Dev.

Belloc, F., P. Dumain, et al. (1994). "A flow cytometric method using Hoechst 33342 and propidium iodide for simultaneous cell cycle analysis and apoptosis determination in unfixed cells." Cytometry 17(1): 59-65.

Bennetts, H. W., E. J. Underwood, et al. (1946). "A specific breeding problem of sheep on subterranean clover pastures in Western Australia." Br Vet J 102(11): 348-52.

Berhow, M. A., E. D. Wagner, et al. (2000). "Characterization and antimutagenic activity of soybean saponins." Mutat Res 448(1): 11-22.

Berner, C., E. Aumuller, et al. (2010). "Epigenetic control of estrogen receptor expression and tumor suppressor genes is modulated by bioactive food compounds." Ann Nutr Metab 57(3-4): 183-9.

Bessa, P. C., M. T. Cerqueira, et al. (2009). "Expression, purification and osteogenic bioactivity of recombinant human BMP-4, -9, -10, -11 and -14." Protein Expression and Purification 63(2): 89-94.

Bianco, P. (2007). "Life in plastic is fantastic." Blood 110(9): 3090.

Bianco, P. (2011). "Bone and the hematopoietic niche: a tale of two stem cells." Blood.

Bianco, P., S. A. Kuznetsov, et al. (2006). "Postnatal Skeletal Stem Cells." Methods in Enzymology Volume 419: 117-148.

Bianco, P., M. Riminucci, et al. (1999). "Multipotential cells in the bone marrow stroma: regulation in the context of organ physiology." Crit Rev Eukaryot Gene Expr 9(2): 159-73.

Bianco, P., P. G. Robey, et al. (2008). "Mesenchymal Stem Cells: Revisiting History, Concepts, and Assays." Cell Stem Cell 2(4): 313-319.

Bianco, P., B. Sacchetti, et al. (2011). "Osteoprogenitors and the hematopoietic microenvironment." Best Practice & Research Clinical Haematology 24(1): 37-47.

226

Bigi, A., B. Bracci, et al. (2004). "Effect of added gelatin on the properties of calcium phosphate cement." Biomaterials 25(14): 2893-9.

Bitto, A., B. P. Burnett, et al. (2008). "Effects of genistein aglycone in osteoporotic, ovariectomized rats: a comparison with alendronate, raloxifene and oestradiol." Br J Pharmacol 155(6): 896-905.

Blake, C., K. M. Fabick, et al. (2011). "Neuromodulation by soy diets or equol: Anti-depressive & anti-obesity-like influences, age- & hormone-dependent effects." BMC Neurosci 12: 28.

Bolanos, R. and J. Francia (2010). "Isoflavones versus hormone therapy for reduction of vertebral fracture risk: indirect comparison." Menopause 17(6): 1201-5.

Bolling, B. W., J. B. Blumberg, et al. (2009). "Extraction methods determine the antioxidant capacity and induction of quinone reductase by soy products in vitro." Food Chem 116(1): 351-355.

Bondar, G., J. Kuo, et al. (2009). "Estradiol-induced estrogen receptor-alpha trafficking." J Neurosci 29(48): 15323-30.

Boot-Handford, R. P., T. M. Michaelidis, et al. (1998). "The bcl-2 knockout mouse exhibits marked changes in osteoblast phenotype and collagen deposition in bone as well as a mild growth plate phenotype." Int J Exp Pathol 79(5): 329-35.

Boué, S. M., S. L. Tilghman, et al. (2009). "Identification of the Potent Phytoestrogen Glycinol in Elicited Soybean (Glycine max)." Endocrinology 150(5): 2446-53.

Boyan BD, S. Z. (2009). "1,25-Dihydroxy vitamin D3 is an autocrine regulator of extracellular matrix turnover and growth factor release via ERp60-activated matrix vesicle matrix metalloproteinases." Cells Tissues Organs. 189(1-4): 70-4.

Boyce, B. F., Z. Yao, et al. (2009). "Osteoclasts have multiple roles in bone in addition to bone resorption." Crit Rev Eukaryot Gene Expr 19(3): 171-80.

Brandoff, J. F., J. S. Silber, et al. (2008). "Contemporary alternatives to synthetic bone grafts for spine surgery." Am J Orthop (Belle Mead NJ) 37(8): 410-4.

Bravo, L. (1998). "Polyphenols: chemistry, dietary sources, metabolism, and nutritional significance." Nutr Rev 56(11): 317-33.

227

Brey DM, M. N., Diamond SL, Mauck RL, Garino JP, Burdick JA. (2011). "High-throughput screening of a small molecule library for promoters and inhibitors of mesenchymal stem cell osteogenic differentiation." Biotechnol Bioeng. 108(1): 163-74.

Bronckers, A. L., K. Sasaguri, et al. (2005). "Expression of Runx2/Cbfa1/Pebp2alphaA during angiogenesis in postnatal rodent and fetal human orofacial tissues." J Bone Miner Res 20(3): 428-37.

Bruder, S. P., D. J. Fink, et al. (1994). "Mesenchymal stem cells in bone development, bone repair, and skeletal regeneration therapy." J Cell Biochem 56(3): 283-94.

Bruder, S. P., M. C. Horowitz, et al. (1997). "Monoclonal antibodies reactive with human osteogenic cell surface antigens." Bone 21(3): 225-35.

Bruder, S. P., N. Jaiswal, et al. (1997). "Growth kinetics, self-renewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation." Journal of Cellular Biochemistry 64(2): 278-294.

Bruder, S. P., N. Jaiswal, et al. (1998). "Mesenchymal stem cells in osteobiology and applied bone regeneration." Clin Orthop Relat Res(355 Suppl): S247-56.

Bruder, S. P., K. H. Kraus, et al. (1998). "The effect of implants loaded with autologous mesenchymal stem cells on the healing of canine segmental bone defects." J Bone Joint Surg Am 80(7): 985-96.

Bruder, S. P., A. A. Kurth, et al. (1998). "Bone regeneration by implantation of purified, culture-expanded human mesenchymal stem cells." J Orthop Res 16(2): 155-62.

Bruder, S. P., N. S. Ricalton, et al. (1998). "Mesenchymal stem cell surface antigen SB-10 corresponds to activated leukocyte cell adhesion molecule and is involved in osteogenic differentiation." J Bone Miner Res 13(4): 655-63.

Brusnahan, S. K., T. R. McGuire, et al. (2010). "Human blood and marrow side population stem cell and Stro-1 positive bone marrow stromal cell numbers decline with age, with an increase in quality of surviving stem cells: correlation with cytokines." Mech Ageing Dev 131(11-12): 718-22.

Brydone, A. S., D. Meek, et al. (2010). "Bone grafting, orthopaedic biomaterials, and the clinical need for bone engineering." Proc Inst Mech Eng H 224(12): 1329-43.

228

Bu, L., K. D. Setchell, et al. (2005). "Influences of dietary soy isoflavones on metabolism but not nociception and stress hormone responses in ovariectomized female rats." Reproductive biology and endocrinology : RB&E 3: 58.

Bubnic, S. J., A. Nagy, et al. (2005). "Donor hematopoietic cells from transgenic mice that express GFP are immunogenic in immunocompetent recipients." Hematology 10(4): 289-95.

Buhring, H., V. Battula, et al. (2007). "Novel Markers for the Prospective Isolation of Human MSC." Annals of the New York Academy of Sciences 1106(1): 262-271.

Bühring, H.-J., S. Treml, et al. (2009). "Phenotypic Characterization of Distinct Human Bone Marrow–Derived MSC Subsets." Annals of the New York Academy of Sciences 1176(1): 124-134.

Burguera, E. F., H. H. Xu, et al. (2006). "Injectable and rapid-setting calcium phosphate bone cement with dicalcium phosphate dihydrate." J Biomed Mater Res B Appl Biomater 77(1): 126-34.

Canfield, A. E., A. B. Sutton, et al. (1996). "Association of thrombospondin-1 with osteogenic differentiation of retinal pericytes in vitro." J Cell Sci 109 ( Pt 2): 343-53.

Caplan, A. I. and J. E. Dennis (2006). "Mesenchymal stem cells as trophic mediators." Journal of Cellular Biochemistry 98(5): 1076-1084.

Casteilla, L., V. Planat-Benard, et al. (2010). "Vascular and endothelial regeneration." Curr Stem Cell Res Ther 5(2): 141-4.

Castelo-Branco, C. and M. J. Cancelo Hidalgo (2011). "Isoflavones: effects on bone health." Climacteric 14(2): 204-11.

Castro-Malaspina, R. G., G Resnick, N Kapoor, P Meyers, D Chiarieri, S McKenzie, HE Broxmeyer, and MA Moore (1980). "Characterization of human bone marrow fibroblast colony-forming cells (CFU-F) and their progeny." Blood 56: 289-301.

Cattoretti, G., R. Schiro, et al. (1993). "Bone marrow stroma in humans: anti-nerve growth factor receptor antibodies selectively stain reticular cells in vivo and in vitro." Blood 81(7): 1726-1738.

Chang, E. C., T. H. Charn, et al. (2008). "Estrogen Receptors {alpha} and {beta} as Determinants of Gene Expression: Influence of Ligand, Dose, and Chromatin Binding." Mol Endocrinol 22(5): 1032-1043.

229

Chauhan, S., J. Jyoti, et al. (2011). "Non-ionic surfactant interactions in aqueous gelatin solution: A physico-chemical investigation." Journal of Molecular Liquids 159(3): 196-200.

Chen FP, H. T., Hu CH, Wang WD, Wang KC, Teng LF. (2006 ). "Expression of estrogen receptors alfa and beta mRNA and alkaline phosphatase in the differentiation of osteoblasts from elderly postmenopausal women: comparison with osteoblasts from osteosarcoma cell lines." Taiwan J Obstet Gynecol. 45(4): 307-12.

Chen, J., H. Chen, et al. (2011). "Simultaneous regeneration of articular cartilage and subchondral bone in vivo using MSCs induced by a spatially controlled gene delivery system in bilayered integrated scaffolds." Biomaterials.

Chen, L., G. Remondetto, et al. (2008). "Kinetics of the breakdown of cross-linked soy protein films for drug delivery." Biomaterials 29(27): 3750-3756.

Chen, X. and J. J. Anderson (2001). "Isoflavones inhibit proliferation of ovarian cancer cells in vitro via an estrogen receptor-dependent pathway." Nutr Cancer 41(1-2): 165-71.

Chen X, G. S., Quarles LD, Anderson JJ. (2003 ). "Effects of genistein on expression of bone markers during MC3T3-E1 osteoblastic cell differentiation. ." J Nutr Biochem. 14(6): 342-9.

Chen, Y., J. Wang, et al. (2011). "Role of Intestinal Hydrolase in the Absorption of Prenylated Flavonoids Present in Yinyanghuo." Molecules 16(2): 1336-1348.

Cheng, L., P. Qasba, et al. (2000). "Human mesenchymal stem cells support megakaryocyte and pro-platelet formation from CD34(+) hematopoietic progenitor cells." J Cell Physiol 184(1): 58-69.

Chien, H. L., H. Y. Huang, et al. (2006). "Transformation of isoflavone phytoestrogens during the fermentation of soymilk with lactic acid bacteria and bifidobacteria." Food Microbiol 23(8): 772-8.

Chiu, T. M., C. C. Huang, et al. (2009). "In vitro and in vivo anti-photoaging effects of an isoflavone extract from soybean cake." J Ethnopharmacol 126(1): 108-13.

Cho, Y. M., T. Imai, et al. (2009). "A 13-week subchronic toxicity study of dietary administered saponin-rich and isoflavones-containing soybean extract in F344 rats." Food Chem Toxicol 47(8): 2150-6.

Choi, E. M. and S. J. Koo (2002). "Effects of soybean ethanol extract on the prostaglandin E2 and interleukin-6 production in osteoblastic cells." Food Research International 35(9): 893-896. 230

Choi, E. M. and S. J. Koo (2003). "Effects of soybean ethanol extract on the cell survival and oxidative stress in osteoblastic cells." Phytother Res 17(6): 627-32.

Choi, E. M., K. S. Suh, et al. (2001). "Soybean ethanol extract increases the function of osteoblastic MC3T3-E1 cells." Phytochemistry 56(7): 733-739.

Chow, L. C. (2009). "Next generation calcium phosphate-based biomaterials." Dent Mater J 28(1): 1-10.

Chow, L. C. and S. Takagi (2001). "A natural bone cement - A laboratory novelty led to the development of revolutionary new biomaterials." Journal of Research of the National Institute of Standards and Technology 106(6): 1029-1033.

Chrzan BG, B. P. (2007). "Phytoestrogens activate estrogen receptor beta1 and estrogenic responses in human breast and bone cancer cell lines." Mol Nutr Food Res. 51(2): 171-7.

Cicek, M., A. Vrabel, et al. (2011). "TGF-beta Inducible Early Gene 1 Regulates Osteoclast Differentiation and Survival by Mediating the NFATc1, AKT, and MEK/ERK Signaling Pathways." PLoS One 6(3): e17522.

Clover J, G. M. ( 1994 ). "Are MG-63 and HOS TE85 human osteosarcoma cell lines representative models of the osteoblastic phenotype? ." Bone. 15(6): 585-91.

Cohen, M. M., Jr. (2009). "Perspectives on RUNX genes: an update." Am J Med Genet A 149A(12): 2629-46.

Colter, D. C., I. Sekiya, et al. (2001). "Identification of a subpopulation of rapidly self-renewing and multipotential adult stem cells in colonies of human marrow stromal cells." Proceedings of the National Academy of Sciences of the United States of America 98(14): 7841-7845.

Crippa, G. E., M. M. Beloti, et al. (2008). "Effect of growth hormone on in vitro osteogenesis and gene expression of human osteoblastic cells is donor-age-dependent." J Cell Biochem 104(2): 369-76.

Curran, J. M., Z. Tang, et al. (2009). "PLGA doping of PCL affects the plastic potential of human mesenchymal stem cells, both in the presence and absence of biological stimuli." J Biomed Mater Res A 89(1): 1-12. da Silva Meirelles, P C Chagastelles, et al. (2006). "Mesenchymal stem cells reside in virtually all post-natal organs and tissues." J Cell Sci 119(11): 2204-2213.

231

da Silva Meirelles, L., A. M. Fontes, et al. (2009). "Mechanisms involved in the therapeutic properties of mesenchymal stem cells." Cytokine & Growth Factor Reviews 20(5-6): 419-427. da Silva Meirelles, L., T. T. Sand, et al. (2009). "MSC Frequency Correlates with Blood Vessel Density in Equine Adipose Tissue." Tissue Engineering Part A 15(2): 221-229.

Daculsi G, L. O., Malard O, Weiss P (2003). "Current state of the art of biphasic calcium phosphate bioceramics." Journal of Materials Science: Materials in Medicine 14: 195-200.

Dancer, J. Y., S. P. Henry, et al. (2010). "Expression of master regulatory genes controlling skeletal development in benign cartilage and bone forming tumors." Human Pathology 41(12): 1788-1793.

Dang, Z. a. L., C. W. (2004). "The Balance Between Concurrent Activation of ERs and PPARs Determines Daidzein-Induced Osteogenesis and Adipogenesis. ." Journal of Bone and Mineral Research 19: 853-861.

Dang, Z.-C., V. r. Audinot, et al. (2003). "Peroxisome Proliferator-activated Receptor Î³ (PPARÎ³) as a Molecular Target for the Soy Phytoestrogen Genistein." Journal of Biological Chemistry 278(2): 962-967.

Davies, H. L. and J. L. Hill (1989). "THE EFFECT OF DIET ON THE METABOLISM IN SHEEP OF THE TRITIATED ISOFLAVONES FORMONONETIN AND BIOCHANIN-A." Australian Journal of Agricultural Research 40(1): 157-163.

Day, A. J., F. J. Canada, et al. (2000). "Dietary flavonoid and isoflavone glycosides are hydrolysed by the lactase site of lactase phlorizin hydrolase." FEBS Lett 468(2-3): 166-70. de Lima Toccafondo Vieira, M., R. F. Duarte, et al. (2008). "Comparison of the estrogenic potencies of standardized soy extracts by immature rat uterotrophic bioassay." Phytomedicine 15(1-2): 31-7.

Deans, R. J. and A. B. Moseley (2000). "Mesenchymal stem cells: biology and potential clinical uses." Exp Hematol 28(8): 875-84.

Decaris, M. L. and J. K. Leach (2011). "Design of experiments approach to engineer cell-secreted matrices for directing osteogenic differentiation." Ann Biomed Eng 39(4): 1174-85.

Declercq, H., N. Van den Vreken, et al. (2004). "Isolation, proliferation and differentiation of osteoblastic cells to study cell/biomaterial interactions: comparison of different isolation techniques and source." Biomaterials 25(5): 757-768. 232

del Real, R. P., J. G. Wolke, et al. (2002). "A new method to produce macropores in calcium phosphate cements." Biomaterials 23(17): 3673-80.

Dellavalle, A., M. Sampaolesi, et al. (2007). "Pericytes of human skeletal muscle are myogenic precursors distinct from satellite cells." Nat Cell Biol 9(3): 255-267.

Deschaseaux, F., F. Gindraux, et al. (2003). "Direct selection of human bone marrow mesenchymal stem cells using an anti-CD49a antibody reveals their CD45med,low phenotype." British Journal of Haematology 122(3): 506-517.

DiGirolamo, C. M., D. Stokes, et al. (1999). "Propagation and senescence of human marrow stromal cells in culture: a simple colony-forming assay identifies samples with the greatest potential to propagate and differentiate." British Journal of Haematology 107(2): 275-281.

Djiogue S, N. D., Halabalaki M, Kretzschmar G, Beyer A, Mbanya JC, Skaltsounis AL, Vollmer G. (2010). "Estrogenic properties of naturally occurring prenylated isoflavones in U2OS human osteosarcoma cells: Structure-activity relationships." J Steroid Biochem Mol Biol. 120(4-5): 184-91.

Dominici M, Le Blanc K, et al. (2006). "Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement." Cytotherapy 8(4): 315-7.

Dreinhofer, K. E., J. M. Feron, et al. (2004). "Orthopaedic surgeons and fragility fractures. A survey by the Bone and Joint Decade and the International Osteoporosis Foundation." J Bone Joint Surg Br 86(7): 958-61.

Dubin RL, H. C., Pileri CL, Kudlacek PE, Li XY, Yee JA, Johnson ML, Anderson RJ. ( 2001 ). "Thermostable (SULT1A1) and thermolabile (SULT1A3) phenol sulfotransferases in human osteosarcoma and osteoblast cells. ." Bone. 28(6): 617-24.

Ducy, P., T. Schinke, et al. (2000). "The Osteoblast: A Sophisticated Fibroblast under Central Surveillance." Science 289(5484): 1501-1504.

Duncan Bassett, J. H. and G. R. Williams (2003). "The molecular actions of thyroid hormone in bone." Trends in Endocrinology and Metabolism 14(8): 356-364.

Dusseldorp, J. R. and R. J. Mobbs (2009). "Iliac crest reconstruction to reduce donor-site morbidity: technical note." Eur Spine J 18(9): 1386-90.

Eggli, P. S., W. Muller, et al. (1988). "Porous hydroxyapatite and tricalcium phosphate cylinders with two different pore size ranges implanted in the cancellous bone of rabbits. A 233

comparative histomorphometric and histologic study of bony ingrowth and implant substitution." Clin Orthop Relat Res(232): 127-38.

Ehninger, A. and A. Trumpp (2011). "The bone marrow stem cell niche grows up: mesenchymal stem cells and macrophages move in." J Exp Med 208(3): 421-8.

Endoh H, S. H., Maruyama K, Takeyama K, Waga I, Shimizu T, Kato S, Kawashima H. (1997). "Rapid activation of MAP kinase by estrogen in the bone cell line." Biochem Biophys Res Commun. 235(1): 99-102.

Enomoto, H., M. Enomoto-Iwamoto, et al. (2000). "Cbfa1 is a positive regulatory factor in chondrocyte maturation." J Biol Chem 275(12): 8695-702.

Enomoto, H., S. Shiojiri, et al. (2003). "Induction of osteoclast differentiation by Runx2 through receptor activator of nuclear factor-kappa B ligand (RANKL) and osteoprotegerin regulation and partial rescue of osteoclastogenesis in Runx2-/- mice by RANKL transgene." J Biol Chem 278(26): 23971-7.

Erices, A., P. Conget, et al. (2000). "Mesenchymal progenitor cells in human umbilical cord blood." British Journal of Haematology 109(1): 235-242.

Espitalier, F., C. Vinatier, et al. (2009). "A comparison between bone reconstruction following the use of mesenchymal stem cells and total bone marrow in association with calcium phosphate scaffold in irradiated bone." Biomaterials 30(5): 763-769.

Ewald, A., K. Helmschrott, et al. (2011). "Effect of cold-setting calcium- and magnesium phosphate matrices on protein expression in osteoblastic cells." J Biomed Mater Res B Appl Biomater 96(2): 326-32.

Farley JR, K.-N. E., Tarbaux NM, Hall SL, Strong DD. (1989 ). "Alkaline phosphatase activity from human osteosarcoma cell line SaOS-2: an isoenzyme standard for quantifying skeletal alkaline phosphatase activity in serum. ." Clin Chem. 35(2): 223-9.

Fedde KN, B. L., Silverstein J, Coburn SP, Ryan LM, Weinstein RS, Waymire K, Narisawa S, Millán JL, MacGregor GR, Whyte MP. (1999). "Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. ." J Bone Miner Res 14(12): 2015-26.

Federici, E., A. Touche, et al. (2003). "High isoflavone content and estrogenic activity of 25 year-old Glycine max tissue cultures." Phytochemistry 64(3): 717-24.

234

Feng, J., A. Mantesso, et al. (2010). "Perivascular cells as mesenchymal stem cells." Expert Opinion on Biological Therapy 10(10): 1441-1451.

Flint, F. O. and K. Pickering (1984). "Demonstration of collagen in meat products by an improved picro-Sirius Red polarisation method." Analyst 109(11): 1505-1506.

Fonseca, H., D. Moreira-Goncalves, et al. (2011). "Voluntary Exercise has Long-Term In Vivo Protective Effects on Osteocyte Viability and Bone Strength Following Ovariectomy." Calcif Tissue Int.

Foryst-Ludwig, A. and U. Kintscher (2010). "Metabolic impact of estrogen signalling through ERalpha and ERbeta." J Steroid Biochem Mol Biol 122(1-3): 74-81.

Fournier B, P. P. (1991). "Characterization of a new human osteosarcoma cell line OHS-4. ." J Cell Biol. 114(3): 577-83.

Franco-Marques, E., J. A. Mendez, et al. (2009). "Evaluation of the influence of the addition of biodegradable polymer matrices in the formulation of self-curing polymer systems for biomedical purposes." Acta Biomater 5(8): 2953-62.

Frank, O., M. Heim, et al. (2002). "Real-time quantitative RT-PCR analysis of human bone marrow stromal cells during osteogenic differentiation in vitro." Journal of Cellular Biochemistry 85(4): 737-746.

Friedenstein, A., R. Chailakhjan, et al. (1970). "The development of fibroblast colonies in monolayer cultures of guinea-pig bone marrow and spleen cells." Cell Tissue Kinet. 3(4): 393-403.

Friedenstein, A. J., R. K. Chailakhyan, et al. (1987). "Bone marrow osteogenic stem cells: in vitro cultivation and transplantation in diffusion chambers." Cell Tissue Kinet 20(3): 263-72.

Friedenstein, A. J., R. K. Chailakhyan, et al. (1974). "STROMAL CELLS RESPONSIBLE FOR TRANSFERRING THE MICROENVIRONMENT OF THE HEMOPOIETIC TISSUES: Cloning In Vitro and Retransplantation In Vivo." Transplantation 17(4): 331-340.

Friedenstein, A. J., N. V. Latzinik, et al. (1992). "Bone marrow stromal colony formation requires stimulation by haemopoietic cells." Bone Miner 18(3): 199-213.

Friedman, C. D., P. D. Costantino, et al. (1998). "BoneSource hydroxyapatite cement: a novel biomaterial for craniofacial skeletal tissue engineering and reconstruction." J Biomed Mater Res 43(4): 428-32.

235

Friedman, M. and D. L. Brandon (2001). "Nutritional and health benefits of soy proteins." J Agric Food Chem 49(3): 1069-86.

Fuchs, D., B. Dirscherl, et al. (2006). "Soy extract has different effects compared with the isolated isoflavones on the proteome of homocysteine-stressed endothelial cells." Mol Nutr Food Res 50(1): 58-69.

Fujishiro, Y., K. Takahashi, et al. (2001). "Preparation and compressive strength of alpha-tricalcium phosphate/gelatin gel composite cement." J Biomed Mater Res 54(4): 525-30.

Gabbay, J. S., J. B. Heller, et al. (2006). "Osteogenic potentiation of human adipose-derived stem cells in a 3-dimensional matrix." Ann Plast Surg 57(1): 89-93.

Gajjeraman S, N. K., Hao J, Qin C, George A. (2007 ). "Matrix macromolecules in hard tissues control the nucleation and hierarchical assembly of hydroxyapatite. ." J Biol Chem. 282(2): 1193-204.

Gang, E. J., D. Bosnakovski, et al. (2007). "SSEA-4 identifies mesenchymal stem cells from bone marrow." Blood 109(4): 1743-1751.

Garrison, K. R., S. Donell, et al. (2007). "Clinical effectiveness and cost-effectiveness of bone morphogenetic proteins in the non-healing of fractures and spinal fusion: a systematic review." Health Technol Assess 11(30): 1-150, iii-iv.

Garrison, K. R., I. Shemilt, et al. (2010). "Bone morphogenetic protein (BMP) for fracture healing in adults." Cochrane Database Syst Rev(6): CD006950.

Gauthier, O., R. Müller, et al. (2005). "In vivo bone regeneration with injectable calcium phosphate biomaterial: A three-dimensional micro-computed tomographic, biomechanical and SEM study." Biomaterials 26(27): 5444-5453.

Genovese, M. I., A. C. Barbosa, et al. (2007). "Commercial soy protein ingredients as isoflavone sources for functional foods." Plant Foods Hum Nutr 62(2): 53-8.

George J, B. N., Ray SK. (2010 ). "Genistein induces receptor and mitochondrial pathways and increases apoptosis during BCL-2 knockdown in human malignant neuroblastoma SK-N-DZ cells. ." J Neurosci Res. 88(4): 877-86.

Giavaresi, G., M. Fini, et al. (2010). "Bone regeneration potential of a soybean-based filler: experimental study in a rabbit cancellous bone defects." J Mater Sci Mater Med 21(2): 615-26.

236

Giguere, V. (2002). "To ERR in the estrogen pathway." Trends Endocrinol Metab 13(5): 220-5.

Ginebra, M. P., J. A. Delgado, et al. (2007). "Factors affecting the structure and properties of an injectable self-setting calcium phosphate foam." J Biomed Mater Res A 80(2): 351-61.

Ginebra, M. P., M. Espanol, et al. (2010). "New processing approaches in calcium phosphate cements and their applications in regenerative medicine." Acta Biomaterialia 6(8): 2863-2873.

Ginebra, M. P., T. Traykova, et al. (2006). "Calcium phosphate cements as bone drug delivery systems: a review." J Control Release 113(2): 102-10.

Golub, E. (2009 ). "Role of matrix vesicles in biomineralization. ." Biochim Biophys Acta. 1790(12): 1592-8.

Goncalves Coelho, P., P. Rui Fernandes, et al. (2011). "Multiscale modeling of bone tissue with surface and permeability control." J Biomech 44(2): 321-9.

Göpferich, A. (1996). "Mechanisms of polymer degradation and erosion." Biomaterials 17(2): 103-114.

Gordon, S. L., S. R. Oppenheimer, et al. (2001). "Recovery of human mesenchymal stem cells following dehydration and rehydration." Cryobiology 43(2): 182-7.

Gordon, S. L. and J. N. Weinstein (1998). "A review of basic science issues in low back pain." Phys Med Rehabil Clin N Am 9(2): 323-42, vii.

Grabowski, P. (2009). "Physiology of bone." Endocr Dev 16: 32-48.

Grammatopoulos, G. A., E. Bell, et al. (2000). "Homeotic transformation of branchial arch identity after Hoxa2 overexpression." Development 127(24): 5355-5365.

Gray, H. (1918 (2000)). Anatomy of the Human Body. Philadelphia(New York), Lea & Febiger.

Gregory, C. A., W. Grady Gunn, et al. (2004). "An Alizarin red-based assay of mineralization by adherent cells in culture: comparison with cetylpyridinium chloride extraction." Analytical Biochemistry 329(1): 77-84.

Gregory, C. A., J. Ylostalo, et al. (2005). "Adult bone marrow stem/progenitor cells (MSCs) are preconditioned by microenvironmental "niches" in culture: a two-stage hypothesis for regulation of MSC fate." Sci STKE 2005(294): pe37.

237

Gronthos, S., A. C. W. Zannettino, et al. (2003). "Molecular and cellular characterisation of highly purified stromal stem cells derived from human bone marrow." J Cell Sci 116(9): 1827-1835.

Guo, H., J. Su, et al. (2009). "Biocompatibility and osteogenicity of degradable Ca-deficient hydroxyapatite scaffolds from calcium phosphate cement for bone tissue engineering." Acta Biomater 5(1): 268-78.

Gurevitch, O., S. Slavin, et al. (2007). "Conversion of red bone marrow into yellow - Cause and mechanisms." Medical Hypotheses 69(3): 531-536.

Habibovic, P., H. Yuan, et al. (2006). "Relevance of osteoinductive biomaterials in critical-sized orthotopic defect." J Orthop Res 24(5): 867-76.

Habraken, W. J., J. G. Wolke, et al. (2007). "Ceramic composites as matrices and scaffolds for drug delivery in tissue engineering." Adv Drug Deliv Rev 59(4-5): 234-48.

Hagiwara, A., N. Imai, et al. (2010). "A twenty eight-day repeated dose toxicity study of black soybean extract in Sprague-Dawley rats." J Toxicol Sci 35(1): 87-96.

Hak, D. J. and J. L. Pittman (2010). "Biological rationale for the intramedullary canal as a source of autograft material." Orthop Clin North Am 41(1): 57-61; table of contents.

Hall, B. K. and T. Miyake (2000). "All for one and one for all: condensations and the initiation of skeletal development." Bioessays 22(2): 138-47.

Hanson, M. J. (2001). "Cloning for therapeutic purposes: ethical and policy considerations." Univ Toledo Law Rev 32(3): 355-65-.

Harichandan, A. and H. J. Buhring (2011). "Prospective isolation of human MSC." Best Pract Res Clin Haematol 24(1): 25-36.

Hart SP, S. J., Dransfield I. (2004). "Phagocytosis of opsonized apoptotic cells: roles for 'old-fashioned' receptors for antibody and complement. ." Clin Exp Immunol 135(2): 181-5.

Hartman, J., A. Ström, et al. (2009). "Estrogen receptor beta in breast cancer--Diagnostic and therapeutic implications." Steroids 74(8): 635-641.

Hawse, J. R., M. Subramaniam, et al. (2008). "Estrogen-TGFbeta cross-talk in bone and other cell types: role of TIEG, Runx2, and other transcription factors." J Cell Biochem 103(2): 383-92.

238

Hawse, J. R., M. Subramaniam, et al. (2008). "Estrogen-TGFβ cross-talk in bone and other cell types: Role of TIEG, Runx2, and other transcription factors." Journal of Cellular Biochemistry 103(2): 383-392.

Hayami, T., Q. Zhang, et al. (2007). "Dexamethasone's enhancement of osteoblastic markers in human periodontal ligament cells is associated with inhibition of collagenase expression." Bone 40(1): 93-104.

Hedelin, M., M. Lof, et al. (2008). "Dietary phytoestrogens are not associated with risk of overall breast cancer but diets rich in coumestrol are inversely associated with risk of estrogen receptor and progesterone receptor negative breast tumors in Swedish women." J Nutr 138(5): 938-45.

Heim, M., O. Frank, et al. (2004). "The phytoestrogen genistein enhances osteogenesis and represses adipogenic differentiation of human primary bone marrow stromal cells." Endocrinology 145(2): 848-59.

Heinemann, D. E., H. Siggelkow, et al. (2000). "Alkaline phosphatase expression during monocyte differentiation. Overlapping markers as a link between monocytic cells, dendritic cells, osteoclasts and osteoblasts." Immunobiology 202(1): 68-81.

Heiser, J. and T. Trentelman (1989). "Full-Fat Soya Products—Manufacturing and Uses in Foodstuffs." Proceedings of the World Congress: Vegetable Proteins Utilization in Human Foods and Animal Feedstuffs: 52–54.

Heliotis, M., K. M. Lavery, et al. (2006). "Transformation of a prefabricated hydroxyapatite/osteogenic protein-1 implant into a vascularised pedicled bone flap in the human chest." Int J Oral Maxillofac Surg 35(3): 265-9.

Henry-Vitrac, C., A. Desmouliere, et al. (2006). "Transport, deglycosylation, and metabolism of trans-piceid by small intestinal epithelial cells." European Journal of Nutrition 45(7): 376-382.

Hertrampf, T., J. Seibel, et al. (2008). "Analysis of the effects of oestrogen receptor Î± (ERÎ±)- and ERÎ²-selective ligands given in combination to ovariectomized rats." Br J Pharmacol 153(7): 1432-7.

Hesaraki, S., F. Moztarzadeh, et al. (2007). "Formation of interconnected macropores in apatitic calcium phosphate bone cement with the use of an effervescent additive." J Biomed Mater Res A 83(1): 80-7.

239

Hilfiker, A., C. Kasper, et al. (2011). "Mesenchymal stem cells and progenitor cells in connective tissue engineering and regenerative medicine: is there a future for transplantation?" Langenbecks Arch Surg 396(4): 489-97.

Ho, S. M. (2004). "Estrogens and anti-estrogens: key mediators of prostate carcinogenesis and new therapeutic candidates." J Cell Biochem 91(3): 491-503.

Hoemann, C. D., H. El-Gabalawy, et al. (2009). "In vitro osteogenesis assays: influence of the primary cell source on alkaline phosphatase activity and mineralization." Pathol Biol (Paris) 57(4): 318-23.

Holland, J. C., O. Brennan, et al. (2011). "Subchondral trabecular structural changes in the proximal tibia in an ovine model of increased bone turnover." J Anat.

Hong, L., H. Sultana, et al. (2009). "Steroid regulation of proliferation and osteogenic differentiation of bone marrow stromal cells: A gender difference." The Journal of Steroid Biochemistry and Molecular Biology 114(3-5): 180-185.

Hong, S. and P. L. Pedersen (2008). "ATP Synthase and the Actions of Inhibitors Utilized To Study Its Roles in Human Health, Disease, and Other Scientific Areas." Microbiol Mol Biol Rev 72(4): 590-641.

Horan, F. E. (1974). "SOY PROTEIN PRODUCTS AND THEIR PRODUCTION." Journal of the American Oil Chemists Society 51(1): A67-A73.

Horwitz EM, Le Blanc K, et al. (2005). "Clarification of the nomenclature for MSC: The International Society for Cellular Therapy position statement." Cytotherapy 7(5): 393-5.

Hsu, A., T. M. Bray, et al. (2010). "Differential effects of whole soy extract and soy isoflavones on apoptosis in prostate cancer cells." Exp. Biol. Med. 235(1): 90-97.

Huang, C. H., W. Y. Tseng, et al. (2010). "Glucosamine promotes osteogenic differentiation of dental pulp stem cells through modulating the level of the transforming growth factor-beta type I receptor." J Cell Physiol 225(1): 140-51.

Ichimura T, A. E., Humphreys BD, Gunaratnam L, Duffield JS, Bonventre JV. (2008 ). "Kidney injury molecule-1 is a phosphatidylserine receptor that confers a phagocytic phenotype on epithelial cells. ." J Clin Invest. 118(5): 1657-68.

Im, G. I., J. M. Lee, et al. (2011). "Wnt inhibitors enhance chondrogenesis of human mesenchymal stem cells in a long-term pellet culture." Biotechnol Lett.

240

Ito, Y. and K. Miyazono (2003). "RUNX transcription factors as key targets of TGF-beta superfamily signaling." Curr Opin Genet Dev 13(1): 43-7.

Itthichaisri, C., M. Wiedmann-Al-Ahmad, et al. (2007). "Comparative in vitro study of the proliferation and growth of human osteoblast-like cells on various biomaterials." J Biomed Mater Res A 82(4): 777-87.

Jafarian, M., M. B. Eslaminejad, et al. (2008). "Marrow-derived mesenchymal stem cells-directed bone regeneration in the dog mandible: a comparison between biphasic calcium phosphate and natural bone mineral." Oral Surg Oral Med Oral Pathol Oral Radiol Endod 105(5): e14-24.

Jaiswal, N., S. E. Haynesworth, et al. (1997). "Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro." Journal of Cellular Biochemistry 64(2): 295-312.

Jaiswal, R. K., N. Jaiswal, et al. (2000). "Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogen-activated protein kinase." J Biol Chem 275(13): 9645-52.

Jakacka, M., M. Ito, et al. (2001). "Estrogen receptor binding to DNA is not required for its activity through the nonclassical AP1 pathway." J Biol Chem 276(17): 13615-21.

Jarcho, M. (1981). "Calcium phosphate ceramics as hard tissue prosthetics." Clin Orthop Relat Res(157): 259-78.

Javed, A., H. Chen, et al. (2010). "Genetic and transcriptional control of bone formation." Oral Maxillofac Surg Clin North Am 22(3): 283-93, v.

Jeng YJ, MY Kochukov, et al. (2009) "Membrane estrogen receptor-α-mediated nongenomic actions of phytoestrogens in GH3/B6/F10 pituitary tumor cells" J Mol Signal 4: 2.

Jensen, E. D., R. Gopalakrishnan, et al. (2010). "Regulation of gene expression in osteoblasts." Biofactors 36(1): 25-32.

Jeong, S. Y. and D. W. Seol (2008). "The role of mitochondria in apoptosis." BMB Rep 41(1): 11-22.

Jiang, P. J., S. Patel, et al. (2010). "Comparing the efficacy of three bioceramic matrices for the release of vancomycin hydrochloride." J Biomed Mater Res B Appl Biomater 93(1): 51-8.

241

Jiang, Y., B. N. Jahagirdar, et al. (2002). "Pluripotency of mesenchymal stem cells derived from adult marrow." Nature 418(6893): 41-49.

Jo, I., J. M. Lee, et al. (2007). "Bone tissue engineering using marrow stromal cells." Biotechnology and Bioprocess Engineering 12(1): 48-53.

Johnsson, M. S. and G. H. Nancollas (1992). "The role of brushite and octacalcium phosphate in apatite formation." Crit Rev Oral Biol Med 3(1-2): 61-82.

Jonason, J. H., G. Xiao, et al. (2009). "Post-translational Regulation of Runx2 in Bone and Cartilage." J Dent Res 88(8): 693-703.

Jones, E. and D. McGonagle (2008). "Human bone marrow mesenchymal stem cells in vivo." Rheumatology 47(2): 126-131.

Joshi, J. V., R. A. Vaidya, et al. (2007). "Plasma levels of genistein following a single dose of soy extract capsule in Indian women." Indian J Med Res 125(4): 534-41.

Kadiyala, S., H. Lo, et al. (1999). "Formation of highly porous polymeric foams with controlled release capability : a phase-separation technique." Methods Mol Med 18: 57-65.

Kadiyala, S., S. Nagaba, et al. (2001). "Metabolites and analogs of 1alpha,25-dihydroxyvitamin D(3): evaluation of actions in bone." Steroids 66(3-5): 347-55.

Kadiyala, S., R. G. Young, et al. (1997). "Culture expanded canine mesenchymal stem cells possess osteochondrogenic potential in vivo and in vitro." Cell Transplant 6(2): 125-34.

Kallio A, G. T., Lamminen E, Seppänen J, Kangas L, Väänänen HK, Härkönen P. (2008). "Estrogen and the selective estrogen receptor modulator (SERM) protection against cell death in estrogen receptor alpha and beta expressing U2OS cells." Mol Cell Endocrinol. 289(1-2): 38-48.

Kameda C, N. M., Tanaka H, Yamasaki A, Kubo M, Tanaka M, Onishi H, Katano M. (2010). "Oestrogen receptor-alpha contributes to the regulation of the hedgehog signalling pathway in ERalpha-positive gastric cancer." Br J Cancer. 102(4): 738-47.

Kang, J., T. M. Badger, et al. (2010). "Non-isoflavone Phytochemicals in Soy and Their Health Effects." Journal of Agricultural and Food Chemistry 58(14): 8119-8133.

Kang LP, Q. L., Zhang JP, Shi N, Zhang M, Wu TM, Chen J. (2001). "Effect of genistein and quercetin on proliferation, collagen synthesis, and type I procollagen mRNA levels of rat hepatic stellate cells. ." Acta Pharmacol Sin. 22(9): 793-6.

242

Karaoz, E., A. Okcu, et al. (2011). "A comprehensive characterization study of human bone marrow mscs with an emphasis on molecular and ultrastructural properties." J Cell Physiol 226(5): 1367-82.

Karp, J. M. and G. S. Leng Teo (2009). "Mesenchymal Stem Cell Homing: The Devil Is in the Details." Cell Stem Cell 4(3): 206-216.

Katzenellenbogen, B. S., I. Choi, et al. (2000). Molecular mechanisms of estrogen action: selective ligands and receptor pharmacology. J Steroid Biochem Mol Biol. England. 74: 279-85.

Kavalkovich, K. W., R. E. Boynton, et al. (2002). "Chondrogenic differentiation of human mesenchymal stem cells within an alginate layer culture system." In Vitro Cell Dev Biol Anim 38(8): 457-66.

Kawakita, S., F. Marotta, et al. (2009). "Effect of an isoflavones-containing red clover preparation and alkaline supplementation on bone metabolism in ovariectomized rats." Clin Interv Aging 4: 91-100.

Kerrigan, L. A. and J. T. Kadonaga (2001). "Purification of sequence-specific DNA-binding proteins by affinity chromatography." Curr Protoc Protein Sci Chapter 9: Unit 9 6.

Kian Tee M, R. I., Tzagarakis-Foster C, Cvoro A, An J, Christy RJ, Yamamoto KR, Leitman DC. (2004). "Estradiol and selective estrogen receptor modulators differentially regulate target genes with estrogen receptors alpha and beta. ." Mol Biol Cell. 15(3): 1262-72.

Kim, D. W., K. Y. Yoo, et al. (2009). "Soy isoflavones mitigate long-term femoral and lumbar vertebral bone loss in middle-aged ovariectomized mice." J Med Food 12(3): 536-41.

Kim, H., H. Suh, et al. (2005). "In vivo bone formation by human marrow stromal cells in biodegradable scaffolds that release dexamethasone and ascorbate-2-phosphate." Biochem Biophys Res Commun 332(4): 1053-60.

Kim, H. J. (2010). "New understanding of glucocorticoid action in bone cells." BMB Rep 43(8): 524-9.

Kim, J. W., K. H. Choi, et al. (2011). "Bone formation of block and particulated biphasic calcium phosphate lyophilized with Escherichia coli-derived recombinant human bone morphogenetic protein 2 in rat calvarial defects." Oral Surg Oral Med Oral Pathol Oral Radiol Endod.

243

Kim, M. H., J. S. Park, et al. (2010). "Genistein and daidzein repress adipogenic differentiation of human adipose tissue-derived mesenchymal stem cells via Wnt/β-catenin signalling or lipolysis." Cell Proliferation 43(6): 594-605.

Kim, S., N. Kim, et al. (2010). "High-Throughput, Sensitive Quantification of Repopulating Hematopoietic Stem Cell Clones." J. Virol. 84(22): 11771-11780.

Kim, S. H., S. H. Kim, et al. (2009). "Genistein inhibits cell growth by modulating various mitogen-activated protein kinases and AKT in cervical cancer cells." Ann N Y Acad Sci 1171: 495-500.

Kim, S.-H., S.-H. Kim, et al. (2009). "Involvement of Both Extrinsic and Intrinsic Apoptotic Pathways in Apoptosis Induced by Genistein in Human Cervical Cancer Cells." Annals of the New York Academy of Sciences 1171(1): 196-201.

Kim , U. S., Woo J, Kim JY, Kim HA, Jang KH, Kang SA, Lim BO, Kang I, Choue RW, Cho Y. (2005). "Comparative effect of seeds of Rhynchosia volubilis and soybean on MG-63 human osteoblastic cell proliferation and estrogenicity." Life Sci. 78(1): 30-40.

Kirsch T, H. G., Golub EE, Nah HD. (2000). "The roles of annexins and types II and X collagen in matrix vesicle-mediated mineralization of growth plate cartilage. ." J Biol Chem. 275(45): 35577-83.

Kirsch T, N. H., Demuth DR, Harrison G, Golub EE, Adams SL, Pacifici M. (1997 ). "Annexin V-mediated calcium flux across membranes is dependent on the lipid composition: implications for cartilage mineralization." Biochemistry. 36(11): 3359-67.

Kirsch T, W. W., Pfander D. (2003). "Functional differences between growth plate apoptotic bodies and matrix vesicles." J Bone Miner Res. 18(10): 1872-81.

Kitoh, H., T. Kitakoji, et al. (2004). "Transplantation of marrow-derived mesenchymal stem cells and platelet-rich plasma during distraction osteogenesis - a preliminary result of three cases." Bone 35(4): 892-898.

Klaus Godl, J. W., Alexander Kurtenbach, Peter Habenberger, Stephanie Blencke, Heidrun Gutbrod, Kostadinos Salassidis, Matthias Stein-Gerlach, Andrea Missio, Matt Cotten, and Henrik Daub (2003). " An efficient proteomics method to identify the cellular targets of protein kinase inhibitors." PNAS 100 (26): 15434-15439.

Klein CB. and AA King (2007). "Genistein genotoxicity: critical considerations of in vitro exposure dose." Toxicol Appl Pharmacol 224(1): 1-11.

244

Klyushnenkova E, JD Mosca, et al. (2005). "T cell responses to allogeneic human mesenchymal stem cells: immunogenicity, tolerance, and suppression." J Biomed Sci 12(1): 47-57.

Knight, D. M., A. B. Chapman, et al. (1987). "Requirements for Triggering of Adipocyte Differentiation by Glucocorticoids and Indomethacin." Mol Endocrinol 1(1): 36-43.

Kobayashi, H., T. Fujishiro, et al. (2009). "Long-term evaluation of a calcium phosphate bone cement with carboxymethyl cellulose in a vertebral defect model." J Biomed Mater Res A 88(4): 880-8.

Kobayashi, T., H. Watanabe, et al. (2005). "Motility and growth of human bone-marrow mesenchymal stem cells during ex vivo expansion in autologous serum." J Bone Joint Surg Br 87-B(10): 1426-1433.

Komissarova, E. V., S. K. Saha, et al. (2005). "Dead or dying: the importance of time in cytotoxicity assays using arsenite as an example." Toxicol Appl Pharmacol 202(1): 99-107.

Komori, T. (2003). "Requisite roles of Runx2 and Cbfb in skeletal development." J Bone Miner Metab 21(4): 193-7.

Komori, T. (2010). "Regulation of bone development and extracellular matrix protein genes by RUNX2." Cell and Tissue Research 339(1): 189-195.

Komori, T., H. Yagi, et al. (1997). "Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts." Cell 89(5): 755-64.

Korda, M., J. Hua, et al. (2010). "The effect of mesenchymal stromal cells on the osseoinduction of impaction grafts." Tissue Eng Part A 16(2): 675-83.

Kostura, L., D. L. Kraitchman, et al. (2004). "Feridex labeling of mesenchymal stem cells inhibits chondrogenesis but not adipogenesis or osteogenesis." NMR Biomed 17(7): 513-7.

Kozhevnikova, M., A. Mikaelyan, et al. (2008). "Comparative characterization of mesenchymal bone marrow stromal cells at early and late stages of culturing." Biology Bulletin 35(2): 132-138.

Krabbe, C., J. Zimmer, et al. (2005). "Neural transdifferentiation of mesenchymal stem cells – a critical review." APMIS 113(11-12): 831-844.

Krebsbach PH, SA Kuznetsov, et al. (1999). "Bone marrow stromal cells: characterization and clinical application." Crit Rev Oral Biol Med 10(2): 165-81.

245

Krinner A, M Hoffmann, et al. (2010). "Individual fates of mesenchymal stem cells in vitro." Bmc Systems Biology 4: 9.

Krishnan, V., H. Heath, et al. (2000). "Mechanism of action of estrogens and selective estrogen receptor modulators." Vitam Horm 60: 123-47.

Kumar, A. A., S. R. Kumar, et al. (2009). "Autologous bone marrow derived mononuclear cell therapy for spinal cord injury: A phase I/II clinical safety and primary efficacy data." Exp Clin Transplant 7(4): 241-8.

Kunisaki Y, M. S., Noda M, Inayoshi A, Sanui T, Harada M, Sasazuki T, Fukui Y. (2004 ). "Defective fetal liver erythropoiesis and T lymphopoiesis in mice lacking the phosphatidylserine receptor." Blood. 103(9): 3362-4.

Kushner, P. J., D. Agard, et al. (2000). "Oestrogen receptor function at classical and alternative response elements." Novartis Found Symp 230: 20-6; discussion 27-40.

Kuznetsov S. and P Robey (1996). "Species Differences in Growth Requirements for Bone Marrow Stromal Fibroblast Colony Formation In vitro." Calcified Tissue International 59(4): 265-270.

Kuznetsov, S. A., A. J. Friedenstein, et al. (1997). "Factors required for bone marrow stromal fibroblast colony formation in vitro." Br J Haematol 97(3): 561-70.

Kuznetsov SA, PH Krebsbach, et al.(1997). "Single-colony derived strains of human marrow stromal fibroblasts form bone after transplantation in vivo."J Bone Miner Res 12(9):1335-47.

Kuznetsov, S. A., M. H. Mankani, et al. (2009). "Enumeration of the colony-forming units-fibroblast from mouse and human bone marrow in normal and pathological conditions." Stem Cell Research 2(1): 83-94.

Kuznetsov SA, MH Mankani et al.(2001)."Circulating skeletal stem cells."J Cell Biol 153(5): 1133-40.

Kwack, S. J., K. B. Kim, et al. (2009). "Risk assessment of soybean-based phytoestrogens." J Toxicol Environ Health A 72(21-22): 1254-61.

Kwak, H. S., S. Y. Park, et al. (2009). "Marked Individual Variation in Isoflavone Metabolism After a Soy Challenge Can Modulate the Skeletal Effect of Isoflavones in Premenopausal Women." J Korean Med Sci 24(5): 867-73.

246

Lagari, V. S. and S. Levis (2010). "Phytoestrogens and bone health." Curr Opin Endocrinol Diabetes Obes 17(6): 546-53.

Lajeunesse, D. (1994 ). "Effect of 17 beta-estradiol on the human osteosarcoma cell line MG-63. ." Bone Miner. 24(1): 1-16.

Lan Levengood, S. K., S. J. Polak, et al. (2010). "Multiscale osteointegration as a new paradigm for the design of calcium phosphate scaffolds for bone regeneration." Biomaterials 31(13): 3552-63.

Lange, C., F. Cakiroglu, et al. (2007). "Accelerated and safe expansion of human mesenchymal stromal cells in animal serum-free medium for transplantation and regenerative medicine." J Cell Physiol 213(1): 18-26.

Leboy, P. S. (2006). "Regulating bone growth and development with bone morphogenetic proteins." Ann N Y Acad Sci 1068: 14-8.

Leclercq G, M Lacroix et al. (2006)"Estrogen receptor alpha: impact of ligands on intracellular shuttling and turnover rate in breast cancer cells" Curr Cancer Drug Targets 6(1):39-64.

Lee, H. L., T. Yi, et al. (2010). "Msx2 mediates the inhibitory action of TNF-alpha on osteoblast differentiation." Exp Mol Med 42(6): 437-45.

Lee, J.-S., D. M. Thomas, et al. (2006). "HES1 Cooperates With pRb to Activate RUNX2-Dependent Transcription." Journal of Bone and Mineral Research 21(6): 921-933.

Lee, K., M. K. Majumdar, et al. (2001). "Human mesenchymal stem cells maintain transgene expression during expansion and differentiation." Mol Ther 3(6): 857-66.

LeGeros RZ, S Lin et al. (2003). "Biphasic calcium phosphate bioceramics: preparation, properties and applications." Journal of Materials Science: Materials in Medicine 14(3):201-209.

Lennon, D., S. Haynesworth, et al. (1996). "Human and animal mesenchymal progenitor cells from bone marrow: Identification of serum for optimal selection and proliferation." In Vitro Cellular & Developmental Biology - Animal 32(10): 602-611.

Leong, T.-W. D., F.-T. Chew, et al. (2004). "Isolating Bone Marrow Stem Cells Using Sieve Technology." STEM CELLS 22(6): 1123-1125.

Lethaby, A. E., J. Brown, et al. (2007). "Phytoestrogens for vasomotor menopausal symptoms." Cochrane Database Syst Rev(4): CD001395.

247

Lewis, G. (2006). "Injectable bone cements for use in vertebroplasty and kyphoplasty: State-of-the-art review." Journal of Biomedical Materials Research Part B: Applied Biomaterials 76B(2): 456-468.

Lewis, G. (2011). "Viscoelastic properties of injectable bone cements for orthopaedic applications: State-of-the-art review." J Biomed Mater Res B Appl Biomater.

Li, J., S. Mareddy, et al. (2009). "A minimal common osteochondrocytic differentiation medium for the osteogenic and chondrogenic differentiation of bone marrow stromal cells in the construction of osteochondral graft." Tissue Eng Part A 15(9): 2481-90.

Li MO, S. M., Mehal WZ, Rakic P, Flavell RA. (2003). "Phosphatidylserine receptor is required for clearance of apoptotic cells." Science. 302(5650): 1560-3.

Li, X., H. Liu, et al. (2011). "Osteogenic differentiation of human adipose-derived stem cells induced by osteoinductive calcium phosphate ceramics." J Biomed Mater Res B Appl Biomater 97(1): 10-9.

Lian, J. B., G. S. Stein, et al. (2006). "Networks and hubs for the transcriptional control of osteoblastogenesis." Rev Endocr Metab Disord 7(1-2): 1-16.

Liao, Q.-C., Y.-L. Li, et al. (2008). "Inhibition of adipocyte differentiation by phytoestrogen genistein through a potential downregulation of extracellular signal-regulated kinases 1/2 activity." Journal of Cellular Biochemistry 104(5): 1853-1864.

Lilien, J. and J. Balsamo (2005). "The regulation of cadherin-mediated adhesion by tyrosine phosphorylation/dephosphorylation of [beta]-catenin." Current Opinion in Cell Biology 17(5): 459-465.

Lindner, H. R. (1976). "Occurrence of anabolic agents in plants and their importance." Environ Qual Saf Suppl(5): 151-8.

Liu, F., H. Zhong, et al. (2010). "Effect of high glucose levels on the calcification of vascular smooth muscle cells by inducing osteoblastic differentiation and intracellular calcium deposition via BMP-2/Cbfalpha-1 pathway." J Zhejiang Univ Sci B 11(12): 905-11.

Liu, H., H. Yazici, et al. (2008). "An in vitro evaluation of the Ca/P ratio for the cytocompatibility of nano-to-micron particulate calcium phosphates for bone regeneration." Acta Biomaterialia 4(5): 1472-1479.

Liu, X., N. Rapp, et al. (2000). "Molecular cloning and chromosomal mapping of a candidate cytokine gene selectively expressed in human CD34+ cells." Genomics 65(3): 283-92. 248

Livingston TL, S Gordon et al. (2003)."Mesenchymal stem cells combined with biphasic calcium phosphate ceramics promote bone regeneration."J Mater Sci Mater Med14(3):211-8.

López-Lazaro M, W. E., Austin CA. (2007 ). "Cells lacking DNA topoisomerase II beta are resistant to genistein." J Nat Prod 70(5): 763-7.

Loveland BE, TG Johns, et al. (1992). "Validation of the MTT dye assay for enumeration of cells in proliferative and antiproliferative assays." Biochem Int 27(3): 501-10.

Loveridge, N. (1999). "Bone: more than a stick." J Anim Sci 77 Suppl 2: 190-6.

Low, K. L., S. H. Tan, et al. (2010). "Calcium phosphate-based composites as injectable bone substitute materials." J Biomed Mater Res B Appl Biomater 94(1): 273-86.

Lu, H. H., S. D. Subramony, et al. (2010). "Tissue engineering strategies for the regeneration of orthopedic interfaces." Ann Biomed Eng 38(6): 2142-54.

Luo, L. H., X. M. Wang, et al. (2008). "Physical properties and biocompatibility of cellulose/soy protein isolate membranes coagulated from acetic aqueous solution." J Biomater Sci Polym Ed 19(4): 479-96.

Luvizuto ER, S Tangl, et al. (2011). "The effect of BMP-2 on the osteoconductive properties of beta-tricalcium phosphate in rat calvaria defects." Biomaterials 32(15): 3855-61.

Ma, D., H. Yao, et al. (2011). "Enhancing bone formation by transplantation of a scaffold-free tissue-engineered periosteum in a rabbit model." Clin Oral Implants Res.

Mackay, A. M., S. C. Beck, et al. (1998). "Chondrogenic differentiation of cultured human mesenchymal stem cells from marrow." Tissue Eng 4(4): 415-28.

Maegawa N, K Kawamura et al.(2007)."Enhancement of osteoblastic differentiation of mesenchymal stromal cells cultured by selective combination of bone morphogenetic protein-2 (BMP2) and fibroblast growth factor-2"J Tissue Eng Regen Med 1(4):306-13.

Majumdar, M. K., M. Keane-Moore, et al. (2003). "Characterization and functionality of cell surface molecules on human mesenchymal stem cells." J Biomed Sci 10(2): 228-41.

Majumdar MK, MA Thiede et al. (2000). "Human marrow-derived mesenchymal stem cells express hematopoietic cytokines and support long-term hematopoiesis when differentiated toward stromal and osteogenic lineages." J Hematother Stem Cell Res 9(6): 841-8.

249

Majumdar, M. K., M. A. Thiede, et al. (1998). "Phenotypic and functional comparison of cultures of marrow-derived mesenchymal stem cells (MSCs) and stromal cells." J Cell Physiol 176(1): 57-66.

Mankani, M. H., P. H. Krebsbach, et al. (2001). "Pedicled bone flap formation using transplanted bone marrow stromal cells." Arch Surg 136(3): 263-70.

Mankani MH, SA Kuznetsov et al. (2008)."Creation of new bone by the percutaneous injection of human bone marrow stromal cell and HA/TCP suspensions"Tissue Eng Part A 14(12):1949-58.

Mankani MH, SA Kuznetsov, et al. (2006). "In vivo bone formation by human bone marrow stromal cells: reconstruction of the mouse calvarium and mandible." Stem Cells 24(9): 2140-9.

Mano, T., N. Nishimura, et al. (2010). "Sympathetic neural influence on bone metabolism in microgravity (Review)." Acta Physiol Hung 97(4): 354-61.

Marini, H., L. Minutoli, et al. (2007). "Effects of the phytoestrogen genistein on bone metabolism in osteopenic postmenopausal women: a randomized trial." Ann Intern Med 146(12): 839-47.

Markovic, M., S. Takagi, et al. (2000). "Formation of macropores in calcium phosphate cements through the use of mannitol crystals." Bioceramics 192-1: 773-776.

Marom, R., I. Shur, et al. (2005). "Characterization of adhesion and differentiation markers of osteogenic marrow stromal cells." Journal of Cellular Physiology 202(1): 41-48.

Martin SJ, F. D., Amarante-Mendes GP, O'Brien GA, Green DR. (1996 ). "Phosphatidylserine externalization during CD95-induced apoptosis of cells and cytoplasts requires ICE/CED-3 protease activity." J Biol Chem 271(46): 28753-6.

Martin, U., I. Gruh, et al. (2009). "Transdifferentiation of Stem Cells: A Critical View." Engineering of Stem Cells 114: 73-106.

Martin-Millan, M., M. Almeida, et al. (2010). "The estrogen receptor-alpha in osteoclasts mediates the protective effects of estrogens on cancellous but not cortical bone." Mol Endocrinol 24(2): 323-34.

Masters J, B. Ø. P. (1998). Cancer Cell Lines Part 1:, Springer.

250

Matsumoto H, RH Erickson et al. (1990) "Biosynthesis of alkaline phosphatase during differentiation of the human colon cancer cell line Caco-2." Gastroenterology 98(5;1):1199-207.

Matsuo K(2009). "Cross-talk among bone cells." Curr Opin Nephrol Hypertens 18(4): 292-7.

Matsuo, K. and N. Irie (2008). "Osteoclast-osteoblast communication." Arch Biochem Biophys 473(2): 201-9.

Maul, T. M., D. W. Chew, et al. (2011). "Mechanical stimuli differentially control stem cell behavior: morphology, proliferation, and differentiation." Biomech Model Mechanobiol.

Mbalaviele, G., Y. Abu-Amer, et al. (2000). "Activation of peroxisome proliferator-activated receptor-gamma pathway inhibits osteoclast differentiation." J Biol Chem 275(19): 14388-93.

Mbalaviele, G., N. Jaiswal, et al. (1999). "Human mesenchymal stem cells promote human osteoclast differentiation from CD34+ bone marrow hematopoietic progenitors." Endocrinology 140(8): 3736-43.

McCall, J. L., R. A. Burich, et al. (2010). "GCP, a genistein-rich compound, inhibits proliferation and induces apoptosis in lymphoma cell lines." Leuk Res 34(1): 69-76.

McCauley, L. K., T. F. Tozum, et al. (2003). "Transgenic models of metabolic bone disease: impact of estrogen receptor deficiency on skeletal metabolism." Connect Tissue Res 44 Suppl 1: 250-63.

McCauley, L. K., T. F. Tozum, et al. (2002). "Estrogen receptors in skeletal metabolism: lessons from genetically modified models of receptor function." Crit Rev Eukaryot Gene Expr 12(2): 89-100.

McClain, R. M., E. Wolz, et al. (2007). "Reproductive safety studies with genistein in rats." Food Chem Toxicol 45(8): 1319-32.

McClain, R. M., E. Wolz, et al. (2005). "Subchronic and chronic safety studies with genistein in dogs." Food Chem Toxicol 43(10): 1461-82.

McGonagle, D., A. English, et al. (2007). "(iii) The relevance of mesenchymal stem cells in vivo for future orthopaedic strategies aimed at fracture repair." Current Orthopaedics 21(4): 262-267.

Mendelsohn, M. E. and R. H. Karas (2010). "Rapid progress for non-nuclear estrogen receptor signaling." The Journal of Clinical Investigation 120(7): 2277-2279.

251

Mendelson, B., S. Jacobson, et al. (2010). "The Fate of Porous Hydroxyapatite Granules Used in Facial Skeletal Augmentation." Aesthetic Plastic Surgery 34(4): 455-461.

Merolli, A., L. Nicolais, et al. (2010). "A degradable soybean-based biomaterial used effectively as a bone filler in vivo in a rabbit." Biomed Mater 5(1): 15008.

Merolli, A., M. Santin, et al. (2009). "BONE RESPONSE TO SOYBEAN-BASED BIOMATERIALS." J Bone Joint Surg Br 91-B(SUPP_I): 97-d-.

Mersereau, J. E., N. Levy, et al. (2008). "Liquiritigenin is a plant-derived highly selective estrogen receptor Î² agonist." Mol Cell Endocrinol 283(1-2): 49-57.

Meuleman, N., T. Tondreau, et al. (2006). "Human marrow mesenchymal stem cell culture: serum-free medium allows better expansion than classical α-MEM medium." European Journal of Haematology 76(4): 309-316.

Miadoková, E. (2009). "Isoflavonoids - an overview of their biological activities and potential health benefits." Interdiscip Toxicol 2(4): 211-8.

Middleton, E., C. Kandaswami, et al. (2000). "The Effects of Plant Flavonoids on Mammalian Cells:Implications for Inflammation, Heart Disease, and Cancer." Pharmacological Reviews 52(4): 673-751.

Miyanishi M, T. K., Koike M, Uchiyama Y, Kitamura T, Nagata S. (2007). "Identification of Tim4 as a phosphatidylserine receptor. ." Nature. 450(7168): 435-9.

Montufar, E. B., T. Traykova, et al. (2010). "Self-hardening calcium deficient hydroxyapatite/gelatine foams for bone regeneration." J Mater Sci Mater Med 21(3): 863-9.

Moreau, J. L. and H. H. K. Xu (2009). "Mesenchymal stem cell proliferation and differentiation on an injectable calcium phosphate - Chitosan composite scaffold." Biomaterials 30(14): 2675-2682.

Morito, K., T. Hirose, et al. (2001). "Interaction of phytoestrogens with estrogen receptors alpha and beta." Biol Pharm Bull 24(4): 351-6.

Morley AA, Trainor KJ, et al. (1983). "Cloning of human lymphocytes using limiting dilution." Exp Hematol. 11(5): 418-24.

Morris C, T. J., Ambrosio L, Santin M. (2006 ). "The soybean isoflavone genistein induces differentiation of MG63 human osteosarcoma osteoblasts." J Nutr 136 (5): 1166-70.

252

Mosca, J. D., J. K. Hendricks, et al. (2000). "Mesenchymal stem cells as vehicles for gene delivery." Clin Orthop Relat Res(379 Suppl): S71-90.

Moseke, C. and U. Gbureck (2010). "Tetracalcium phosphate: Synthesis, properties and biomedical applications." Acta Biomater 6(10): 3815-23.

Moura, P. J. and S. L. Petersen (2010). "Estradiol acts through nuclear- and membrane-initiated mechanisms to maintain a balance between GABAergic and glutamatergic signaling in the brain: implications for hormone replacement therapy." Rev Neurosci 21(5): 363-80.

Mueller, H., M. U. Kassack, et al. (2004). "Comparison of the Usefulness of the MTT, ATP, and Calcein Assays to Predict the Potency of Cytotoxic Agents in Various Human Cancer Cell Lines." Journal of Biomolecular Screening 9(6): 506-515.

Muraglia, A., R. Cancedda, et al. (2000). "Clonal mesenchymal progenitors from human bone marrow differentiate in vitro according to a hierarchical model." J Cell Sci 113(7): 1161-1166.

Murota, K., S. Shimizu, et al. (2002). "Unique uptake and transport of isoflavone aglycones by human intestinal caco-2 cells: comparison of isoflavonoids and flavonoids." J Nutr 132(7): 1956-61.

Murphy, J. M., K. Dixon, et al. (2002). "Reduced chondrogenic and adipogenic activity of mesenchymal stem cells from patients with advanced osteoarthritis." Arthritis Rheum 46(3): 704-13.

Murphy, J. M., D. J. Fink, et al. (2003). "Stem cell therapy in a caprine model of osteoarthritis." Arthritis Rheum 48(12): 3464-74.

Murphy, J. M., R. Heinegard, et al. (1999). "Distribution of cartilage molecules in the developing mouse joint." Matrix Biol 18(5): 487-97.

Murray, M. P. (2007). "Green Plastics: An Introduction to the New Science of Biodegradable Plastics." Natural Areas Journal 27(3): 283-285.

Nade, S., L. Armstrong, et al. (1983). "Osteogenesis after bone and bone marrow transplantation. The ability of ceramic materials to sustain osteogenesis from transplanted bone marrow cells: preliminary studies." Clin Orthop Relat Res(181): 255-63.

Nakamura T, N. K., Stinson RA. (1988). "Release of alkaline phosphatase from human osteosarcoma cells by phosphatidylinositol phospholipase C: effect of tunicamycin. ." Arch Biochem Biophys. 265(1): 190-6. 253

Nemeth, K., G. W. Plumb, et al. (2003). "Deglycosylation by small intestinal epithelial cell beta-glucosidases is a critical step in the absorption and metabolism of dietary flavonoid glycosides in humans." Eur J Nutr 42(1): 29-42.

Neugarten J, M. I., Lei J, Silbiger SR. (1999). "Estradiol suppresses mesangial cell type I collagen synthesis via activation of the MAP kinase cascade. ." Am J Physiol. 277(6 Pt 2): F875-81.

Neuhuber, B., S. A. Swanger, et al. (2008). "Effects of plating density and culture time on bone marrow stromal cell characteristics." Experimental Hematology 36(9): 1176-1185.

Neut, D., O. S. Kluin, et al. (2010). "Gentamicin release from commercially-available gentamicin-loaded PMMA bone cements in a prosthesis-related interfacial gap model and their antibacterial efficacy." BMC Musculoskelet Disord 11: 258.

Newman, R. E., D. Yoo, et al. (2009). "Treatment of inflammatory diseases with mesenchymal stem cells." Inflamm Allergy Drug Targets 8(2): 110-23.

Ni, Y.-D., W.-J. Hong, et al. (2010). "Dual effects of daidzein on chicken hepatic vitellogenin II expression and estrogen receptor-mediated transactivation in vitro." Steroids 75(3): 245-251.

Nicotera, P. and M. Leist (1997). "Energy supply and the shape of death in neurons and lymphoid cells." Cell Death Differ 4(6): 435-42.

Nielsen, I. L. and G. Williamson (2007). "Review of the factors affecting bioavailability of soy isoflavones in humans." Nutr Cancer 57(1): 1-10.

Nilsson, S. and J. A. Gustafsson (2011). "Estrogen receptors: therapies targeted to receptor subtypes." Clin Pharmacol Ther 89(1): 44-55.

Nilsson, S., S. MÃ¤kelÃ¤, et al. (2001). "Mechanisms of Estrogen Action." Physiological Reviews 81(4): 1535-1565.

Nishio, Y., Y. Dong, et al. (2006). "Runx2-mediated regulation of the zinc finger Osterix/Sp7 gene." Gene 372: 62-70.

Noriko, O., Y. Takafumi, et al. (2006). "Bone formation-promoting effect of genistein on marrow mesenchymal cell culture." Bio-Medical Materials and Engineering 16(1): 23-32.

O'Brien, C. A. (2010). "Control of RANKL gene expression." Bone 46(4): 911-9.

254

Oda, K., H. Matsuda, et al. (2003). "Relationship between adjuvant activity and amphipathic structure of soyasaponins." Vaccine 21(17-18): 2145-51.

Orimo, H. (2010). "The mechanism of mineralization and the role of alkaline phosphatase in health and disease." J Nippon Med Sch 77(1): 4-12.

Orrenius, S., D. H. Burgess, et al. (1997). "Mitochondria as the focus of apoptosis research." Cell Death Differ 4(6): 427-8.

Owen, M. and A. J. Friedenstein (1988). "Stromal stem cells: marrow-derived osteogenic precursors." Ciba Found Symp 136: 42-60.

Pan, W., L. D. Quarles, et al. (2005). "Genistein stimulates the osteoblastic differentiation via NO/cGMP in bone marrow culture." Journal of Cellular Biochemistry 94(2): 307-316.

Pan, Z., P. Jiang, et al. (2007). "Mechanical and biocompatible influences of chitosan fiber and gelatin on calcium phosphate cement." J Biomed Mater Res B Appl Biomater 82(1): 246-52.

Pandey, K. B. and S. I. Rizvi (2009). "Plant polyphenols as dietary antioxidants in human health and disease." Oxid Med Cell Longev 2(5): 270-8.

Pandjaitan, N., N. Hettiarachchy, et al. (2000). "Enrichment of Genistein in Soy Protein Concentrate with b-glucosidase." Journal of Food Science 65(3): 403-407.

Panzavolta, S., P. Torricelli, et al. (2010). "Functionalization of biomimetic calcium phosphate bone cements with alendronate." J Inorg Biochem 104(10): 1099-106.

Paredes, R., G. Arriagada, et al. (2004). "The Runx2 transcription factor plays a key role in the 1alpha,25-dihydroxy Vitamin D3-dependent upregulation of the rat osteocalcin (OC) gene expression in osteoblastic cells." J Steroid Biochem Mol Biol 89-90(1-5): 269-71.

Parikka, V., A. Vaananen, et al. (2005). "Human mesenchymal stem cell derived osteoblasts degrade organic bone matrix in vitro by matrix metalloproteinases." Matrix Biol 24(6): 438-47.

Park D, T.-T. A., Elliott MR, Lu M, Haney LB, Ma Z, Klibanov AL, Mandell JW, Ravichandran KS. (2007). "BAI1 is an engulfment receptor for apoptotic cells upstream of the ELMO/Dock180/Rac module." Nature. 450(7168): 430-4.

Park, H.-C., K. Yasuda, et al. (2010). "Renal capsule as a stem cell niche." American Journal of Physiology - Renal Physiology 298(5): F1254-F1262.

255

Park, S. H., E. S. Gil, et al. (2010). "Relationships between degradability of silk scaffolds and osteogenesis." Biomaterials 31(24): 6162-72.

Park, S. H., A. Tofighi, et al. (2011). "Calcium phosphate combination biomaterials as human mesenchymal stem cell delivery vehicles for bone repair." J Biomed Mater Res B Appl Biomater 97(2): 235-44.

Pazzaglia, U. E., G. Beluffi, et al. (2011). "A Review of the Actual Knowledge of the Processes Governing Growth and Development of Long Bones." Fetal Pediatr Pathol.

Perut, F., E. B. Montufar, et al. (2011). "Novel soybean/gelatine-based bioactive and injectable hydroxyapatite foam: Material properties and cell response." Acta Biomaterialia 7(4): 1780-1787.

Petite, H., V. Viateau, et al. (2000). "Tissue-engineered bone regeneration." Nat Biotech 18(9): 959-963.

Pi, M., P. Faber, et al. (2005). "Identification of a novel extracellular cation-sensing G-protein-coupled receptor." J Biol Chem 280(48): 40201-9.

Pike, A. C. W., A. M. Brzozowski, et al. (1999). "Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist." EMBO J 18(17): 4608-4618.

Pilsakova, L., I. Riecansky, et al. (2010). "The physiological actions of isoflavone phytoestrogens." Physiol Res 59(5): 651-64.

Piotrowska, E., J. Jakóbkiewicz-Banecka, et al. (2010). "Different amounts of isoflavones in various commercially available soy extracts in the light of gene expression-targeted isoflavone therapy." Phytotherapy Research 24(S1): S109-S113.

Pittenger, M., P. Vanguri, et al. (2002). "Adult mesenchymal stem cells: potential for muscle and tendon regeneration and use in gene therapy." J Musculoskelet Neuronal Interact 2(4): 309-20.

Pittenger, M. F., A. M. Mackay, et al. (1999). "Multilineage Potential of Adult Human Mesenchymal Stem Cells." Science 284(5411): 143-147.

Pittenger, M. F. and B. J. Martin (2004). "Mesenchymal stem cells and their potential as cardiac therapeutics." Circ Res 95(1): 9-20.

256

Pittenger, M. F., J. D. Mosca, et al. (2000). "Human mesenchymal stem cells: progenitor cells for cartilage, bone, fat and stroma." Curr Top Microbiol Immunol 251: 3-11.

Polak, S. J., S. K. Levengood, et al. (2011). "Analysis of the roles of microporosity and BMP-2 on multiple measures of bone regeneration and healing in calcium phosphate scaffolds." Acta Biomater 7(4): 1760-71.

Ponticiello, M. S., R. M. Schinagl, et al. (2000). "Gelatin-based resorbable sponge as a carrier matrix for human mesenchymal stem cells in cartilage regeneration therapy." J Biomed Mater Res 52(2): 246-55.

Porta, M. D., G. M. Rigolin, et al. (2004). "Dendritic cell recovery after allogeneic stem-cell transplantation in acute leukemia: correlations with clinical and transplant characteristics." Eur J Haematol 72(1): 18-25.

Pountos, I., D. Corscadden, et al. (2007). "Mesenchymal stem cell tissue engineering: Techniques for isolation, expansion and application." Injury 38: S23-S33.

Quarto, R., M. Mastrogiacomo, et al. (2001). "Repair of large bone defects with the use of autologous bone marrow stromal cells." N Engl J Med 344(5): 385-6.

Rai, B., J. L. Lin, et al. (2010). "Differences between in vitro viability and differentiation and in vivo bone-forming efficacy of human mesenchymal stem cells cultured on PCL-TCP scaffolds." Biomaterials 31(31): 7960-70.

Ralston, S. H. (2010). "Genetics of osteoporosis." Ann N Y Acad Sci 1192: 181-9.

Rando, G., B. Ramachandran, et al. (2009). Differential effect of pure isoflavones and soymilk on estrogen receptor activity in mice. Toxicol Appl Pharmacol. United States. 237: 288-97.

Rao, A. V. and M. K. Sung (1995). "Saponins as anticarcinogens." J Nutr 125(3 Suppl): 717S-724S.

Rao, L. G., M. S. Kung Sutherland, et al. (1996). "Positive interaction between 17 beta-Estradiol and parathyroid hormone in normal human osteoblasts cultured long term in the presence of dexamethasone." Osteoporos Int 6(2): 111-9.

Ratanavaraporn, J., S. Damrongsakkul, et al. (2011). "Osteogenic Differentiation of Bone-Marrow-Derived Stem Cells Cultured with Mixed Gelatin and Chitooligosaccharide Scaffolds." Journal of Biomaterials Science-Polymer Edition 22(8): 1083-1098.

257

Ravichandran KS, L. U. (2007). "Engulfment of apoptotic cells: signals for a good meal." Nat Rev Immunol. 7(12): 964-74.

Ray, R., N. M. Novotny, et al. (2008). "Sex steroids and stem cell function." Mol Med 14(7-8): 493-501.

Reddy, N. and Y. Q. Yang (2011). "Completely biodegradable soyprotein-jute biocomposites developed using water without any chemicals as plasticizer." Industrial Crops and Products 33(1): 35-41.

Reinli, K. and G. Block (1996). "Phytoestrogen content of foods--a compendium of literature values." Nutr Cancer 26(2): 123-48.

Reiter, E., E. Reiter, et al. (2009). "Isoflavones are safe compounds for therapeutical applications - evaluation of in vitro data." Gynecol Endocrinol 25(9): 554-80.

Ren, J., P. Jin, et al. (2011). "Global transcriptome analysis of human bone marrow stromal cells (BMSC) reveals proliferative, mobile and interactive cells that produce abundant extracellular matrix proteins, some of which may affect BMSC potency." Cytotherapy.

Rhodes, N. P. (2007). "Inflammatory signals in the development of tissue-engineered soft tissue." Biomaterials 28(34): 5131-6.

Richardson, S. M., R. V. Walker, et al. (2006). "Intervertebral disc cell-mediated mesenchymal stem cell differentiation." Stem Cells 24(3): 707-16.

Riches, P. L. and S. H. Ralston (2010). "Recent insights into the biology of bone turnover." J R Coll Physicians Edinb 40(1): 66-9.

Richter, W. (2009). "Mesenchymal stem cells and cartilage in situ regeneration." J Intern Med 266(4): 390-405.

Rickard, D. J., D. G. Monroe, et al. (2003). "Phytoestrogen genistein acts as an estrogen agonist on human osteoblastic cells through estrogen receptors α and β." Journal of Cellular Biochemistry 89(3): 633-646.

Riekstina, U., I. Cakstina, et al. (2009). "Embryonic stem cell marker expression pattern in human mesenchymal stem cells derived from bone marrow, adipose tissue, heart and dermis." Stem Cell Rev 5(4): 378-86.

Robey, P. G., S. A. Kuznetsov, et al. (2007). "Skeletal ("mesenchymal") stem cells for tissue engineering." Methods Mol Med 140: 83-99.

258

Robinson, J. A., S. A. Harris, et al. (1997). "Estrogen regulation of human osteoblastic cell proliferation and differentiation." Endocrinology 138(7): 2919-27.

Roobrouck, V. D., K. Vanuytsel, et al. (2011). "Culture Mediated Changes in Fate and/or Potency of Stem Cells." STEM CELLS Epub ahead of print: N/A-N/A.

Ropero, A. B., P. Alonso-Magdalena, et al. (2006). "Rapid endocrine disruption: environmental estrogen actions triggered outside the nucleus." J Steroid Biochem Mol Biol 102(1-5): 163-9.

Rossman, T. G. (2009). "Inappropriate cytotoxicity measurements." Environ Mol Mutagen 50(2): 81.

Rowlands, J. C., M. A. Berhow, et al. (2002). "Estrogenic and antiproliferative properties of soy sapogenols in human breast cancer cells in vitro." Food Chem Toxicol 40(12): 1767-74.

Rubinstein, L., Shoemaker, RH, Paull, KD, R. M. Simon, S. Tosini, P. Skehan, D. A. Scudiero, A. Monks, and M. R. Boyd (1990). "Comparison of In Vitro Anticancer-Drug-Screening Data Generated With a Tetrazolium Assay Versus a Protein Assay Against a Diverse Panel of Human Tumor Cell Lines " JNCI J Natl Cancer Inst 82(13): 1113-1117.

Rucinska, A., M. Roszczyk, et al. (2008). "Cytotoxicity of the isoflavone genistein in NIH 3T3 cells." Cell Biol Int 32(8): 1019-23.

Saad, F. and A. Lipton (2010). "Bone-marker levels in patients with prostate cancer: potential correlations with outcomes." Curr Opin Support Palliat Care 4(3): 127-34.

Sabatini, F., L. Petecchia, et al. (2005). "Human bronchial fibroblasts exhibit a mesenchymal stem cell phenotype and multilineage differentiating potentialities." Lab Invest 85(8): 962-971.

Sacchetti, B., A. Funari, et al. (2007). "Self-Renewing Osteoprogenitors in Bone Marrow Sinusoids Can Organize a Hematopoietic Microenvironment." Cell 131(2): 324-336.

Safe, S. and K. Kim (2008). "Non-classical genomic estrogen receptor (ER)/specificity protein and ER/activating protein-1 signaling pathways." J Mol Endocrinol 41(5): 263-75.

Sakaguchi Y, S. I., Yagishita K, Ichinose S, Shinomiya K, Muneta T. (2004 ). "Suspended cells from trabecular bone by collagenase digestion become virtually identical to mesenchymal stem cells obtained from marrow aspirates." Blood 104(9): 2728-35.

259

Salem, H. K. and C. Thiemermann (2010). "Mesenchymal Stromal Cells: Current Understanding and Clinical Status." STEM CELLS 28(3): 585-596.

Salvatori L, C. F., Coroniti G, Starace G, Frati L, Russo MA, Petrangeli E. (2009). "Down-regulation of epidermal growth factor receptor induced by estrogens and phytoestrogens promotes the differentiation of U2OS human osteosarcoma cells." J Cell Physiol. 220(1): 35-44.

Sandrini E, M. C., Chiesa R, Cigada A, Santin M. (2005 ). "In vitro assessment of the osteointegrative potential of a novel multiphase anodic spark deposition coating for orthopaedic and dental implants. ." J Biomed Mater Res B Appl Biomater. 73(2): 392-9.

Santin, M. and L. Ambrosio (2008). "Soybean-based biomaterials: preparation, properties and tissue regeneration potential." Expert Review of Medical Devices 5: 349-358.

Santin, M., C. Morris, et al. (2007). "A New Class of Bioactive and Biodegradable Soybean-Based Bone Fillers." Biomacromolecules 8(9): 2706-2711.

SANTIN, M., L. NICOLAIS, et al. (2008). Soybean-based thermoplastics as biomaterials. UNIV BRIGHTON [GB]. international.

Santos, T. C., A. P. Marques, et al. (2010). "Chitosan improves the biological performance of soy-based biomaterials." Tissue Eng Part A 16(9): 2883-90.

Saraiva PP, T. S., Conde SJ, Nogueira CR. ( 2008). "The importance of hormone receptor analysis in osteosarcoma cells growth submitted to treatment with estrogen in association with thyroid hormone. ." Cell Biochem Funct. 26(1): 107-10.

Sarugaser, R., L. Hanoun, et al. (2009). "Human mesenchymal stem cells self-renew and differentiate according to a deterministic hierarchy." PLoS One 4(8): e6498.

Sato, K., K. Ozaki, et al. (2010). "Mesenchymal stromal cells for graft-versus-host disease : basic aspects and clinical outcomes." J Clin Exp Hematop 50(2): 79-89.

Sayegh, R. A. and P. G. Stubblefield (2002). "Bone metabolism and the perimenopause overview, risk factors, screening, and osteoporosis preventive measures." Obstet Gynecol Clin North Am 29(3): 495-510.

Schimming, R. and R. Schmelzeisen (2004). "Tissue-engineered bone for maxillary sinus augmentation." Journal of oral and maxillofacial surgery : official journal of the American Association of Oral and Maxillofacial Surgeons 62(6): 724-729.

260

Schmidt F, K. C., Frank B, Wolburg H, Weller M. (2008 ). "The topoisomerase II inhibitor, genistein, induces G2/M arrest and apoptosis in human malignant glioma cell lines." Oncol Rep. 19(4): 1061-6.

Schreihofer, D. A. (2005). "Transcriptional regulation by phytoestrogens in neuronal cell lines." Mol Cell Endocrinol 231(1-2): 13-22.

Schumacher, M., U. Deisinger, et al. (2010). "Indirect rapid prototyping of biphasic calcium phosphate scaffolds as bone substitutes: influence of phase composition, macroporosity and pore geometry on mechanical properties." J Mater Sci Mater Med 21(12): 3119-27.

Seebach, C., J. Schultheiss, et al. (2010). "Comparison of six bone-graft substitutes regarding to cell seeding efficiency, metabolism and growth behaviour of human mesenchymal stem cells (MSC) in vitro." Injury-International Journal of the Care of the Injured 41(7): 731-738.

Seeberger, K. L., A. Eshpeter, et al. (2008). "Epithelial cells within the human pancreas do not coexpress mesenchymal antigens: epithelial-mesenchymal transition is an artifact of cell culture." Lab Invest 89(2): 110-121.

Sellheyer, K. and D. Krahl (2010). "Cutaneous mesenchymal stem cells: status of current knowledge, implications for dermatopathology." J Cutan Pathol 37(6): 624-34.

Serafini, M. and C. M. Verfaillie (2006). "Pluripotency in Adult Stem Cells: State of the Art." Semin Reprod Med 24(05): 379,388.

Setchell, K. D. and A. Cassidy (1999). "Dietary isoflavones: biological effects and relevance to human health." J Nutr 129(3): 758S-767S.

Shankar, S. and D. J. Hosking (2006). "Biochemical assessment of Paget's disease of bone." J Bone Miner Res 21 Suppl 2: P22-7.

Shi, S. and S. Gronthos (2003). "Perivascular Niche of Postnatal Mesenchymal Stem Cells in Human Bone Marrow and Dental Pulp." Journal of Bone and Mineral Research 18(4): 696-704.

Shie, M. Y., D. C. Chen, et al. (2008). "Immersion behavior of gelatin-containing calcium phosphate cement." Acta Biomater 4(3): 646-55.

Shieh, D.-B., R.-Y. Li, et al. (2010). "Effects of genistein on β-catenin signaling and subcellular distribution of actin-binding proteins in human umbilical CD105-positive stromal cells." Journal of Cellular Physiology 223(2): 423-434.

261

Shur, I., R. Socher, et al. (2005). "Dexamethasone regulation of cFos mRNA in osteoprogenitors." J Cell Physiol 202(1): 240-5.

Siggers, K., H. Frei, et al. (2010). "Effect of bone graft substitute on marrow stromal cell proliferation and differentiation." Journal of Biomedical Materials Research Part A 94A(3): 877-885.

Siminovitch, L., E. A. McCulloch, et al. (1963). "The distribution of colony-forming cells among spleen colonies." Journal of Cellular and Comparative Physiology 62(3): 327-336.

Simmons, P. J. and B. Torok-Storb (1991). "Identification of stromal cell precursors in human bone marrow by a novel monoclonal antibody, STRO-1." Blood 78(1): 55-62.

Simon, C. G., Jr., W. F. Guthrie, et al. (2004). "Cell seeding into calcium phosphate cement." J Biomed Mater Res A 68(4): 628-39.

Singh-Gupta, V., H. Zhang, et al. (2010). "Daidzein effect on hormone refractory prostate cancer in vitro and in vivo compared to genistein and soy extract: potentiation of radiotherapy." Pharm Res 27(6): 1115-27.

Sipos, W., P. Pietschmann, et al. (2008). "Strategies for novel therapeutic approaches targeting cytokines and signaling pathways of osteoclasto- and osteoblastogenesis in the fight against immune-mediated bone and joint diseases." Curr Med Chem 15(2): 127-36.

Slootweg, M. C., J. P. Salles, et al. (1996). "Growth hormone binds to a single high affinity receptor site on mouse osteoblasts: modulation by retinoic acid and cell differentiation." J Endocrinol 150(3): 465-72.

Smith, E. P., B. Specker, et al. (2010). "Recent experimental and clinical findings in the skeleton associated with loss of estrogen hormone or estrogen receptor activity." J Steroid Biochem Mol Biol 118(4-5): 264-72.

Smith, P., N. P. Rhodes, et al. (1999). "Influence of the alpha1-adrenergic antagonist, doxazosin, on noradrenaline-induced modulation of cytoskeletal proteins in cultured hyperplastic prostatic stromal cells." Prostate 38(3): 216-27.

Smith, P., N. P. Rhodes, et al. (2000). "Modulating effect of estrogen and testosterone on prostatic stromal cell phenotype differentiation induced by noradrenaline and doxazosin." Prostate 44(2): 111-7.

262

Smith, P., N. P. Rhodes, et al. (2002). "Upregulation of estrogen and androgen receptors modulate expression of FGF-2 and FGF-7 in human, cultured, prostatic stromal cells exposed to high concentrations of estradiol." Prostate Cancer Prostatic Dis 5(2): 105-10.

Smith, P., N. P. Rhodes, et al. (2004). "Relationship between upregulated oestrogen receptors and expression of growth factors in cultured, human, prostatic stromal cells exposed to estradiol or dihydrotestosterone." Prostate Cancer Prostatic Dis 7(1): 57-62.

Sohier, J., P. Corre, et al. (2010). "Hydrogel/calcium phosphate composites require specific properties for three-dimensional culture of human bone mesenchymal cells." Acta Biomater 6(8): 2932-9.

Soucy, N. V., H. D. Parkinson, et al. (2006). "Kinetics of genistein and its conjugated metabolites in pregnant Sprague-Dawley rats following single and repeated genistein administration." Toxicol Sci 90(1): 230-40.

Soung, D., L. Devareddy, et al. (2006). "Soy Affects Trabecular Microarchitecture and Favorably Alters Select Bone-Specific Gene Expressions in a Male Rat Model of Osteoporosis." Calcified Tissue International 78(6): 385-391.

Squadrito, F., D. Altavilla, et al. (2003). "Effect of genistein on endothelial function in postmenopausal women: a randomized, double-blind, controlled study." Am J Med 114(6): 470-6.

Stein, G. S., J. B. Lian, et al. (2004). "Runx2 control of organization, assembly and activity of the regulatory machinery for skeletal gene expression." Oncogene 23(24): 4315-29.

Stellingsma, C., A. Vissink, et al. (2004). "Implantology and the severely resorbed edentulous mandible." Crit Rev Oral Biol Med 15(4): 240-8.

Stossi F, B. D., Frasor J, Komm B, Lyttle CR, Katzenellenbogen BS. ( 2004). "Transcriptional profiling of estrogen-regulated gene expression via estrogen receptor (ER) alpha or ERbeta in human osteosarcoma cells: distinct and common target genes for these receptors." Endocrinology. 145(7): 3473-86.

Su, K., X. Shi, et al. (2011). "Transplantable delivery systems for in situ controlled release of bisphosphonate in orthopedic therapy." Expert Opin Drug Deliv 8(1): 113-26.

Subramaniam M, C. D., Keeting PE, Rasmussen K, Riggs BL, Spelsberg TC. (1992). "Glucocorticoid regulation of alkaline phosphatase, osteocalcin, and proto-oncogenes in normal human osteoblast-like cells." J Cell Biochem. 50(4): 411-24.

263

Swain, S. N., S. M. Biswal, et al. (2004). "Biodegradable Soy-Based Plastics: Opportunities and Challenges." Journal of Polymers and the Environment 12(1): 35-42.

Swiers, G., M. de Bruijn, et al. (2010). "Hematopoietic stem cell emergence in the conceptus and the role of Runx1." Int J Dev Biol 54(6-7): 1151-63.

Tait JF, G. D. (1992 ). "Phospholipid binding of annexin V: effects of calcium and membrane phosphatidylserine content." Arch Biochem Biophys. 298(1): 187-91.

Takada, I., A. P. Kouzmenko, et al. (2009). "Molecular switching of osteoblastogenesis versus adipogenesis: implications for targeted therapies." Expert Opin Ther Targets 13(5): 593-603.

Takagi, S., L. C. Chow, et al. (1998). "Formation of hydroxyapatite in new calcium phosphate cements." Biomaterials 19(17): 1593-9.

Talha, S. and L. Harel (1985). "Early stimulation of ATP turnover induced by growth factors. Synergistic effect of EGF and insulin and correlation with DNA synthesis." Exp Cell Res 158(2): 311-20.

Tamura, M., S. Hori, et al. (2011). "Lactobacillus rhamnosus JCM 2771: Impact on Metabolism of Isoflavonoids in the Fecal Flora from a Male Equol Producer." Curr Microbiol 62(5): 1632-7.

Tang, X., X. Zhu, et al. (2008). "Phytoestrogens induce differential estrogen receptor β-mediated responses in transfected MG-63 cells." Endocrine 34(1): 29-35.

Tárnok, A., H. Ulrich, et al. (2010). "Phenotypes of stem cells from diverse origin." Cytometry Part A 77A(1): 6-10.

Terkeltaub, R. (2008). "Macrophage glucocorticoid receptors join the intercellular dialogue in atherosclerotic lesion calcification. ." Arterioscler Thromb Vasc Biol. 28(12): 2096-8.

Till, J. E. and C. E. Mc (1961). "A direct measurement of the radiation sensitivity of normal mouse bone marrow cells." Radiat Res 14: 213-22.

Tolar, J., K. Le Blanc, et al. (2010). "Concise review: hitting the right spot with mesenchymal stromal cells." Stem Cells 28(8): 1446-55.

Tominaga, Y., A. Wang, et al. (2007). "Genistein inhibits Brca1 mutant tumor growth through activation of DNA damage checkpoints, cell cycle arrest, and mitotic catastrophe." Cell Death Differ 14(3): 472-9.

264

Toquet, J., R. Rohanizadeh, et al. (1999). "Osteogenic potential in vitro of human bone marrow cells cultured on macroporous biphasic calcium phosphate ceramic." Journal of Biomedical Materials Research 44(1): 98-108.

Trainor, K. J. and A. A. Morley (1983). "Cloning of lymphocytes from whole blood by limiting dilution." Journal of Immunological Methods 65(3): 369-372.

Traktuev, D. O., S. Merfeld-Clauss, et al. (2008). "A population of multipotent CD34-positive adipose stromal cells share pericyte and mesenchymal surface markers, reside in a periendothelial location, and stabilize endothelial networks." Circ Res 102(1): 77-85.

Trivedi, R., R. Goswami, et al. (2010). "Investigational anabolic therapies for osteoporosis." Expert Opin Investig Drugs 19(8): 995-1005.

Trouvin, A. P. and V. Goeb (2010). "Receptor activator of nuclear factor-kappaB ligand and osteoprotegerin: maintaining the balance to prevent bone loss." Clin Interv Aging 5: 345-54.

Tsai, M. T., W. J. Li, et al. (2009). "Modulation of osteogenesis in human mesenchymal stem cells by specific pulsed electromagnetic field stimulation." J Orthop Res 27(9): 1169-74.

Tsialogiannis, E., I. Polyzois, et al. (2009). "Targeting bone morphogenetic protein antagonists: in vitro and in vivo evidence of their role in bone metabolism." Expert Opin Ther Targets 13(1): 123-37.

Tsiridis, E., A. Bhalla, et al. (2006). "Enhancing the osteoinductive properties of hydroxyapatite by the addition of human mesenchymal stem cells, and recombinant human osteogenic protein-1 (BMP-7) in vitro." Injury 37(3, Supplement 1): S25-S32.

Tsujimoto, Y. (1997). "Apoptosis and necrosis: intracellular ATP level as a determinant for cell death modes." Cell death and differentiation 4(6): 429-34.

Tuck, S. P., A. C. Scane, et al. (2008). "Sex steroids and bone turnover markers in men with symptomatic vertebral fractures." Bone 43(6): 999-1005.

Turner, R. T., G. K. Wakley, et al. (1988). "Tamoxifen inhibits osteoclast-mediated resorption of trabecular bone in ovarian hormone-deficient rats." Endocrinology 122(3): 1146-50.

Uchida, A., S. Nade, et al. (1987). "Growth of bone marrow cells on porous ceramics in vitro." J Biomed Mater Res 21(1): 1-10.

Uchiyama S, Y. M. (2008). "Anabolic effect of beta-cryptoxanthin in osteoblastic MC3T3-E1 cells is enhanced with 17beta-estradiol, genistein, or zinc sulfate in vitro: the unique effect

265

with zinc on Runx2 and alpha1(I) collagen mRNA expressions." Mol Cell Biochem 307(1-2): 209-19.

Uchiyama S, Y. M. (2007 ). "Genistein and zinc synergistically enhance gene expression and mineralization in osteoblastic MC3T3-E1 cells." Int J Mol Med 19(2): 213-20.

Uskokovic, V. and D. P. Uskokovic (2011). "Nanosized hydroxyapatite and other calcium phosphates: Chemistry of formation and application as drug and gene delivery agents." Journal of Biomedical Materials Research Part B-Applied Biomaterials 96B(1): 152-191.

Usui, T. (2006). "Pharmaceutical prospects of phytoestrogens." Endocr J 53(1): 7-20.

Valls, J., S. Millán, et al. (2009). "Advanced separation methods of food anthocyanins, isoflavones and flavanols." Journal of Chromatography A 1216(43): 7143-7172. van den Bos, C., J. D. Mosca, et al. (1997). "Human mesenchymal stem cells respond to fibroblast growth factors." Hum Cell 10(1): 45-50. van den Bos, C., S. Silverstetter, et al. (1998). "p21(cip1) rescues human mesenchymal stem cells from apoptosis induced by low-density culture." Cell Tissue Res 293(3): 463-70.

Vanderoost, J., S. V. Jaecques, et al. (2011). "Fast and accurate specimen-specific simulation of trabecular bone elastic modulus using novel beam-shell finite element models." J Biomech.

Vater, C., P. Kasten, et al. (2011). "Culture media for the differentiation of mesenchymal stromal cells." Acta Biomater 7(2): 463-77.

Vaz, C. M., L. A. de Graaf, et al. (2003). "In vitro degradation behaviour of biodegradable soy plastics: effects of crosslinking with glyoxal and thermal treatment." Polymer Degradation and Stability 81(1): 65-74.

Verfaillie, C. M. (2005). "Multipotent adult progenitor cells: an update." Novartis Found Symp 265: 55-61; discussion 61-5, 92-7.

Verron, E., I. Khairoun, et al. (2010). "Calcium phosphate biomaterials as bone drug delivery systems: a review." Drug Discovery Today 15(13-14): 547-552.

Vogel, S., T. Trapp, et al. (2010). "Hepatocyte growth factor-mediated attraction of mesenchymal stem cells for apoptotic neuronal and cardiomyocytic cells." Cellular and Molecular Life Sciences 67(2): 295-303.

266

Wagner, E. F. and G. Karsenty (2001). "Genetic control of skeletal development." Current Opinion in Genetics & Development 11(5): 527-532.

Wagner, W. and A. D. Ho (2007). "Mesenchymal stem cell preparations--comparing apples and oranges." Stem Cell Rev 3(4): 239-48.

Walle, T., A. M. Browning, et al. (2005). "Flavonoid glucosides are hydrolyzed and thus activated in the oral cavity in humans." J Nutr 135(1): 48-52.

Wang, L., Y. L. Huang, et al. (2010). "Osteogenic Responses to Different Concentrations/Ratios of BMP-2 and bFGF in Bone Formation." Annals of Biomedical Engineering 38(1): 77-87.

Wang, L., D. Zou, et al. (2011). "Repair of bone defects around dental implants with bone morphogenetic protein/fibroblast growth factor-loaded porous calcium phosphate cement: a pilot study in a canine model." Clin Oral Implants Res 22(2): 173-81.

Watson, C., R. Alyea, et al. (2007). "Nongenomic Actions of Low Concentration Estrogens and Xenoestrogens on Multiple Tissues." Mol Cell Endocrinol 274(1-2): 1-7.

Weir, M. D. and H. H. Xu (2010). "Culture human mesenchymal stem cells with calcium phosphate cement scaffolds for bone repair." J Biomed Mater Res B Appl Biomater 93(1): 93-105.

Weir, M. D. and H. H. Xu (2010). "Human bone marrow stem cell-encapsulating calcium phosphate scaffolds for bone repair." Acta Biomater 6(10): 4118-26.

Wen, H. B., J. Moradian-Oldak, et al. (2000). "Dose-dependent modulation of octacalcium phosphate crystal habit by amelogenins." J Dent Res 79(11): 1902-6.

Whyte, M. P. (2010). "Physiological role of alkaline phosphatase explored in hypophosphatasia." Ann N Y Acad Sci 1192: 190-200.

Winkler, H. (2009). "Rationale for one stage exchange of infected hip replacement using uncemented implants and antibiotic impregnated bone graft." Int J Med Sci 6(5): 247-52.

Woolf, A. D. (2000). "The bone and joint decade 2000-2010." Ann Rheum Dis 59(2): 81-2.

Wuthier RE, W. L., Sauer GR, Genge BR, Yoshimori T, Ishikawa Y. ( 1992 ). "Mechanism of matrix vesicle calcification: characterization of ion channels and the nucleational core of growth plate vesicles." Bone Miner. 17(2): 290-5.

267

Xian, C. J., F. H. Zhou, et al. (2004). "Intramembranous ossification mechanism for bone bridge formation at the growth plate cartilage injury site." J Orthop Res 22(2): 417-26.

Xiao, Y., S. Mareddy, et al. (2010). "Clonal characterization of bone marrow derived stem cells and their application for bone regeneration." Int J Oral Sci 2(3): 127-35.

Xu, H. H. and J. B. Quinn (2002). "Calcium phosphate cement containing resorbable fibers for short-term reinforcement and macroporosity." Biomaterials 23(1): 193-202.

Xu, H. H., L. Zhao, et al. (2010). "Stem cell-calcium phosphate constructs for bone engineering." J Dent Res 89(12): 1482-8.

Yacoub N., Y. H. Ismail, et al. (2002). "Transmission of bone strain in the craniofacial bones of edentulous human skulls upon dental implant loading." J Prosthet Dent 88(2): 192-9.

Yan, J., S. E. Stringer, et al. (2011). "Decorin GAG synthesis and TGF-beta signaling mediate Ox-LDL-induced mineralization of human vascular smooth muscle cells." Arterioscler Thromb Vasc Biol 31(3): 608-15.

Yang, Z., C. Y. Huang, et al. (2011). "Sox-9 facilitates differentiation of adipose tissue-derived stem cells into a chondrocyte-like phenotype in vitro." J Orthop Res.

Yoshimura K, T Shigeura et al(2006). "Characterization of freshly isolated and cultured cells derived from the fatty and fluid portions of liposuction aspirates." J Cell Physiol 208(1): 64-76.

You, M., G. Peng, et al. (2011). "Chondrogenic differentiation of human bone marrow mesenchymal stem cells on polyhydroxyalkanoate (PHA) scaffolds coated with PHA granule binding protein PhaP fused with RGD peptide." Biomaterials 32(9): 2305-13.

Young, R. G., D. L. Butler, et al. (1998). "Use of mesenchymal stem cells in a collagen matrix for Achilles tendon repair." J Orthop Res 16(4): 406-13.

Yu, J., H. He, et al. (2010). "Differentiation potential of STRO-1+ dental pulp stem cells changes during cell passaging." BMC Cell Biology 11(1): 32.

Zaia, J., R. Boynton, et al. (2001). "Posttranslational modifications to human bone sialoprotein determined by mass spectrometry." Biochemistry 40(43): 12983-91.

Zaia, J., R. E. Boynton, et al. (1997). "Post-translational modifications in cartilage oligomeric matrix protein. Characterization of the N-linked oligosaccharides by matrix-assisted laser desorption ionization time-of-flight mass spectrometry." J Biol Chem 272(22): 14120-6.

268

Zaia, J., B. Liu, et al. (2000). "Structural analysis of cartilage proteoglycans and glycoproteins using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry." Anal Biochem 277(1): 94-103.

Zannettino, A. C. W., S. Paton, et al. (2010). "Comparative Assessment of the Osteoconductive Properties of Different Biomaterials In Vivo Seeded with Human or Ovine Mesenchymal Stem/Stromal Cells." Tissue Engineering Part A 16(12): 3579-3587.

Zhang, D.-Y., S. Zhang, et al. (2010). "Negative pressure cavitation extraction and antioxidant activity of genistein and genistin from the roots of pigeon pea [Cajanus cajan (L.) Millsp.]." Separation and Purification Technology 74(2): 261-270.

Zhang Y, WF Chen, et al. (2008). "Soy isoflavones and their bone protective effects." Inflammopharmacology 16(5): 213-5.

Zhao C, H Gao, et al. (2010). "Genome-wide mapping of estrogen receptor-beta-binding regions reveals extensive cross-talk with transcription factor activator protein-1." Cancer Res 70(12): 5174-83.

Zhao L, M Tang, et al. (2011). "Osteogenic Media and rhBMP-2-Induced Differentiation of Umbilical Cord Mesenchymal Stem Cells Encapsulated in Alginate Microbeads and Integrated in an Injectable Calcium Phosphate-Chitosan Fibrous Scaffold." Tissue Eng Part A 17(7-8): 969-79.

Zimmermann MC, SL Tilghman, et al. (2010). "Glyceollin I, a Novel Antiestrogenic Phytoalexin Isolated from Activated Soy." J Pharmacol Exp Ther 332(1): 35-45.

Zipori D (2006). "The stem state: mesenchymal plasticity as a paradigm." Curr Stem Cell Res Ther 1(1): 95-102.

Zuk PA, M Zhu, et al. (2002). "Human Adipose Tissue Is a Source of Multipotent Stem Cells." Mol. Biol. Cell 13(12): 4279-4295.

Zvaifler NJ, L Marinova-Mutafchieva, et al. (2000). "Mesenchymal precursor cells in the blood of normal individuals." Arthritis Res 2(6): 477-88.

269

Appendix A. Statistics.

Chapter 3

3.4.1 Cell number. P values Day/media GM Beta Dex Osteo 0 0.950 0.691 0.963 0.963 2 0.003 0.410 0.998 0.014 4 0.889 0.956 0.271 0.271 6 0.000 0.006 0.129, Dunnett’s 0.044

3.4.2 Cell death. P values Day/media GM Beta Dex Osteo 2 0.000 0.001 0.000 0.000 4 0.000 0.823 0.080 0.834 6 0.000 0.080 0.001 0.002

3.4.3. ALP. P values Day/media GM Beta Dex Osteo 2 0.001 0.000 0.000 0.000 4 0.000 0.001 0.000 0.000 6 0.000 0.000 0.000 0.000

3.4.4 Collagen. P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.000 0.000 0.005 0.000 6 0.000 0.000 0.000 0.000

3.4.5 ARS. P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.000 0.000 0.005 0.000 6 0.000 0.000 0.000 0.000

Chapter 4

4.4.8.1 Cell number. P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.030 0.002 0.000 0.000 6 0.002 0.001 0.001 0.000

270

4.4.8.2 LDH release. P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.000 0.000 0.000 0.000 6 0.000 0.000 0.002 0.000

4.4.8.3 ALP. P values Day/media GM Beta Dex Osteo 2 0.016 0.944 0.124, Dunnett’s 0.037 4 0.707 0.035 0.352 0.000 6 0.000 0.000 0.000 0.000

4.4.8.4 Collagen. P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.039 0.001 0.000 0.000 6 0.033 0.000 0.000 0.000 Comparison of days between concentrations

Media/µM 0 1.88 3.75 7.5 15 30 300 GM 0.804 0.000 0.000 0.000 0.042 0.014 0.116 Beta 0.035 0.004 0.005 0.001 0.002 0.049 0.215 Dex 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Osteo 0.234 0.000 0.000 0.002 0.015 0.006 0.073 Comparison of 1 concentration between days

4.4.8.5 ARS. P values Day/media GM Beta Dex Osteo 2 0.099 0.021 0.000 0.000 4 0.000 0.008 0.001 0.001 6 0.054, Dunnett’s 0.002 0.000 0.028 Comparison of days between concentrations

Media/µM 0 1.88 3.75 7.5 15 30 300 GM 0.261 0.078 0.134 0.117 0.292 0.370 0.828 Beta 0.241 0.365 0.264 0.626 0.922 0.336 0.576 Dex 0.156 0.972 0.143 0.137 0.087 0.076 0.077 Osteo 0.177 0.136 0.093 0.071 0.065 0.054 0.066 Comparison of 1 concentration between days

271

Chapter 5

5.4.2. Cell number P values Day/media GM Beta Dex Osteo 0 0.809 0.709 0.851 0.462 2 0.000 0.000 0.000 0.000 4 0.000 0.000 0.000 0.000 6 0.000 0.000 0.018 0.111

5.4.3 cell number HPI counts P values Media/µM 0 1.88 3.75 7.5 15 30 300 All media 0.000 0.000 0.000 0.000 0.000 0.000 0.001

5.4.4 Brd-U incorporation P values Media/µM 0 1.88 3.75 7.5 15 30 300 All media 0.153 0.000 0.000 0.008 0.003 0.000 0.001

5.4.5. Cell death by LDH P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.000 0.000 0.000 0.000 6 0.000 0.000 0.000 0.000

5.4.6 ATP P values hour/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 72 0.000 0.000 0.000 0.000

5.4.7 ALP P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.000 0.000 0.000 0.000 6 0.000 0.000 0.000 0.000 Comparison of days between concentrations

Media/µM 0 1.88 3.75 7.5 15 30 300 GM 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Beta 0.001 0.000 0.000 0.000 0.000 0.000 0.001 Dex 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Osteo 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Comparison of concentrations between days

5.4.8 Collagen P values Day/media GM Beta Dex Osteo 2 0.000 0.000 0.000 0.000 4 0.000 0.016 0.000 0.000 6 0.000 0.029 0.000 0.000

272

5.4.9 ARS P values Day/media GM Beta Dex Osteo 2 0.000 0.598 0.000 0.000 4 0.000 0.000 0.000 0.000 6 0.000 0.000 0.000 0.000

Chapter 6

6.4.2 Cell counts P values Day/media GM GM+I Osteo Osteo+I 2 0.000 0.000 0.000 0.000 6 0.000 0.000 0.000 0.000 8 0.000 0.000 0.000 0.000 10 0.000 0.000 0.000 0.000

6.4.4 ALP P values Day/media GM GM+I Beta Beta+I Osteo Osteo+I 2 0.000 0.016 0.000 0.000 0.217 0.000 6 0.000 0.000 0.000 0.000 0.000 0.000 8 0.000 0.000 0.000 0.000 0.000 0.000 10 0.000 0.000 0.000 0.001 0.000 0.000

6.4.5 Cbfa-1 P values Day/media GM GM+I Osteo Osteo+I 10 0.000 0.000 0.000 0.002

6.4.6 ARS P values Day/media GM GM+I Beta Beta+I Osteo Osteo+I 6 0.000 0.000 0.000 0.000 0.000 0.000 10 0.000 0.003 0.000 0.000 0.000 0.000

Chapter 7

7.4.1 Proliferation P values Day Cements Cement All days 0 0.004 Gelatine 0.000 12 0.142 Soy 0.000 24 0.000 Comparison between time points Comparison between cements

7.4.2 Cbfa1 expression P values Day Cements 0 0.004 12 0.018 24 0.000

273

Appendix B

Isoflavone

Days in Daidzin Genistin Daidzein Genistein Saline solution

2 Days 134.026 54.336 2.293 0.870

4 Days 94.936 58.886 1.473 0.665

7 Days 15.189 7.919 0.271 0.165

8 Days 13.139 4.656 0.095 0.092

10 Days 15.437 5.713 0.108 0.089

12 Days 19.464 8.557 0.093 0.094 Table showing the release of isoflavones from a soy extract and gelatine foamed cement

274