Sensors for in vitro bone tissue engineering applications

Sensors for in vitro Bone Tissue Engineering Applications Johan Gustavsson Doctoral Thesis Biomedical Engineering Doctoral Programme Supervised by Dr. Elisabeth Engel Department of Materials Science and Metallurgy Universitat Polit`ecnica de Catalunya Barcelona (Spain) April 2011 This page is intentionally left blank Till min morfar och mormor, till min mor och min bror, och till Laura. This page is intentionally left blank It has been said that I am interested in mountaineering. That’s true. The qualities it requires are just those, which I feel we all need today. Perseverance and patience. A firm grip on realities. Careful, but imaginative planning. A clear awareness of the dangers, but also of the fact that faith is what we make it. And that the safest climber is he who never questions his ability to overcome all difficulties. Dag Hammarskj¨old The wine is awful, the people without temperament, and even the sun radiates no warmth. Jean Bernadotte, ex-monarch of Sweden This page is intentionally left blank Abstract The field of tissue engineering (TE) aims to understand structure-function relationships in mammalian tissues and from there on develop biological substitutes that can restore, maintain, or improve tissue functions. To progress such efforts, research into fundamental biology should also be complemented with advances in engineering and manufacturing issues. New or improved enabling tools of predictive, producing, performing, and preserving character need to be invented, developed, and implemented. If correctly applied, enabling tools of performing character, such as different sensing and imaging modalities, could contribute to increased temporal and spatial control of the TE environment by providing indirect information on cellular metabolic activity, scaffold performance, and cell-biomaterial interactions. Along that line, this thesis has explored how ion sensors, when applied to standard in vitro bone tissue engineering environments, can provide information for both improved understanding and control of cellular and material activities. First, commercial ion sensors were used to determine the influence of three different osteoblast-like cell models on their ionic extracellular environment (IEE). Rat-derived mesenchymal stem cells (rMSCs) and SAOS-2 cells expressed high or very high alkaline phosphatase (ALP) activity, and as a consequence they influenced the concentration of total inorganic phosphorus (Pi) in culture medium containing β-glycerophosphate. On the contrary, MG63 cells showed extremely low ALP activity and did not indfluence [Pi]. Moreover, cell-induced calcium deposition in the extracellular matrix was observed both in mature SAOS-2 and rMSC layers but not in MG63 layers, and was correlated with decreased concentration of calcium in the culture medium. Fluctuations in the bone tissue engineering IEE, mainly with respect to Pi, Ca2+, and pH may therefore be indicative of specific osteoblastic activity. Second, the nature of ion reactivity of calcium-deficient hydroxyapatite (CDHA), a scaffold candidate material for bone tissue engineering, was systematically investigated by exposing it during different time periods to standard culture media of different chemical composition. Traditional sorption models were used to describe experimental data, revealing for example significant sorption of calcium onto CDHA as well as acidification of the culture media. Slight variations in the chemical composition of culture medium were observed to provoke different ion reactivity with respect to total phosphorus (Pi). It was further shown that ion reactivity of CDHA was a function of material maturity as certain ion replacement processes were allowed to occur 7 8 during prolonged contact times. Taken altogether, evaluation of ionic concentrations of culture medium exposed to CDHA contributed to improved understanding of the underlying mechanisms of ion reactivity of CDHA. Third, the effects of the dynamic IEE induced by CDHA on cellular behaviour were evaluated by growing osteoblast-like SAOS-2 cells in semipermeable culture inserts placed in close proximity to CDHA. It was revealed that cells proliferated well and that their alkaline phosphatase activity was modified mainly in time rather than in absolute levels. However, while ALP activity of SAOS-2 cells grown in absence of CDHA easily created conditions for cell-induced calcium deposition in the extracellular matrix, presence of CDHA resulted in competition between cells and material for the calcium and phosphorus. The competition delayed calcium deposition in the cell layers, but as sorption of phosphorus onto CDHA gradually decreased with time, conditions for mineralisation were slowly created for SAOS-2 cells also when grown in presence of CDHA. Obtained results indicate that sorption of phosphorus rather than sorption of calcium was the main limiter for bone mineralisation around CDHA, and should be considered in development of osteogenic biomaterials. Following above observations that the IEE can be significantly altered by the presence of standard bone tissue engineering components such as cells or scaffold materials, a generic setup for online potentiometric measurements of ionic concentrations and pH was fabricated. Traditional calcium-selective electrodes for measurements in small volumes (easily down to 0.1 mL) were fabricated in house, and were characterised in standard in vitro TE conditions. They exhibited a Nernstian response to calcium, and were little influenced by other major extracellular ions. Moreover, the electrodes resisted sterilisation through UV irradiation, did not induce any cytotoxic effects in contact with osteoblasts, and the electrode potential was subject only to minor drift during longer measurements. Ca2+-electrodes were used to successfully monitor sorption of calcium onto calcium-deficient hydroxyapatite immersed in cell culture medium, as well as osteoblast-induced calcium deposition in the extracellular matrix during a time frame of 24 hours. Such proof-of-concept measurements demonstrate the potential of these instruments as enabling tools in TE processes. Finally, all-solid-state potentiometric microelectrodes based on anodically grown films of iridium oxide were prepared for real-time monitoring of pH in traditionally inaccessible bone tissue engineering environments. In particularly, the great usefulness of these sensor was demonstrated by measuring 9 pH in the interior of hydrolysing α-tri calcium phosphate (an injectable bone cement), as well as at its immediate interface with extracellular fluid. In both these cases the pH sensors clearly indicated how the material initially provoked a very alkaline environment, which with time became more and more acidic. The absolute pH variation caused by the material was detected to be of such magnitude that it may influence its adjacent biological environment, and should be considered upon drug loading and/or implantation. Taken altogether, the developed sensor setups (i.e. pH microelectrodes and ion-selective electrodes) are applicable to standard in vitro TE environments. Their small size allow to approach local environments previously neglected, and in that way they can contribute to improved temporal and spatial control of the TE environment, and so enhance reproducibility and efficiency of many TE processes. Contents 16 5.5.1 The Ionic Extracellular Environment . . . . . . . . . . 219 5.5.2 Cellular Response . . . . . . . . . . . . . . . . . . . . . 221 5.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 6 Fabrication, Characterisation, and Application of Ion Sensors in Bone Tissue Engineering 229 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 6.1.1 Potentiometric Ion Sensors . . . . . . . . . . . . . . . . 229 6.1.1.1 Ions in solution . . . . . . . . . . . . . . . . . 231 6.1.1.2 Electrodes of first and second kind . . . . . . 233 6.1.2 Ion Selective Electrodes . . . . . . . . . . . . . . . . . 235 6.1.2.1 Electrode bodies . . . . . . . . . . . . . . . . 235 6.1.2.2 Ion selective membranes . . . . . . . . . . . . 236 6.1.2.3 Response mechanism . . . . . . . . . . . . . . 238 6.1.2.4 Sensor characterisation . . . . . . . . . . . . . 240 6.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 6.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . 243 6.3.1 Potentiometric Instrumentation . . . . . . . . . . . . . 243 6.3.1.1 Hardware . . . . . . . . . . . . . . . . . . . . 243 6.3.1.2 Software . . . . . . . . . . . . . . . . . . . . . 244 6.3.2 Ag/AgCl Reference Electrode . . . . . . . . . . . . . . 246 6.3.3 Calcium Selective Electrodes . . . . . . . . . . . . . . . 249 6.3.3.1 ISE characterisation . . . . . . . . . . . . . . 251 6.3.4 Sensor Applications . . . . . . . . . . . . . . . . . . . . 252 6.3.4.1 Cl–-activity during α-TCP-hydrolysis . . . . 252 6.3.4.2 Ca2+-activity of CDHA in cell culture medium253 6.3.4.3 Ca2+-activity during osteoblast mineralisation 253 6.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 6.4.1 Potentiometric Instrumentation . . . . . . . . . . . . . 255 6.4.2 Ag/AgCl Reference Electrode . . . . . . . . . . . . . . 255 6.4.3 Calcium Selective Electrode Characterisation . . . . . . 260 6.4.3.1 Standard characterisation . . . . . . . . . . . 260 6.4.3.2 ISE characterisation in complex solutions . . 263 6.4.4 Sensor Applications . . . . . . . . . . . . . . . . . . . . 266 6.4.4.1 Cl–-activity during α-TCP-hydrolysis . . . . 266 6.4.4.2 Ca2+-activity of CDHA in cell culture medium267 6.4.4.3 Ca2+-activity during osteoblast mineralisation 268 6.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 6.5.1 On the Design of the Sensor Platform . . . . . . . . . . 273 6.5.2 Calcium selective electrodes . . . . . . . . . . . . . . . 275 Contents 17 6.5.3 Sensors Applications . . . . . . . . . . . . . . . . . . . 276 6.5.3.1 Cl−-activity during setting of α-TCP . . . . . 276 6.5.3.2 Ca2+-sensor...................277 6.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 7 pH Microelectrodes Applied in Bone Tissue Engineering 283 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 7.1.1 pH Detection . . . . . . . . . . . . . . . . . . . . . . . 283 7.1.2 Metal-Metal Oxide Electrodes . . . . . . . . . . . . . . 284 7.1.2.1 Iridium oxide pH electrodes . . . . . . . . . . 285 7.1.2.2 Microelectrodes . . . . . . . . . . . . . . . . . 286 7.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286 7.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . 287 7.3.1 Solid-state Microelectrode Body . . . . . . . . . . . . . 287 7.3.2 Iridium Oxide Deposition . . . . . . . . . . . . . . . . 288 7.3.3 pH Electrode Characterisation . . . . . . . . . . . . . . 289 7.3.4 Sensor Applications . . . . . . . . . . . . . . . . . . . . 289 7.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 7.4.1 Microelectrode Characterisation . . . . . . . . . . . . . 291 7.4.2 Iridium Oxide Deposition . . . . . . . . . . . . . . . . 292 7.4.3 pH Electrode Characterisation . . . . . . . . . . . . . . 293 7.4.4 Sensor Applications . . . . . . . . . . . . . . . . . . . . 294 7.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 7.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 8 Conclusions and Future Work 303 8.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 8.1.1 The Cells (Chapter 3) . . . . . . . . . . . . . . . . . . 303 8.1.2 The Scaffold Material (Chapter 4) . . . . . . . . . . . . 305 8.1.3 Cells and Scaffold Combined (Chapter 5) . . . . . . . . 306 8.1.4 Ion Sensors for Online Monitoring (Chapter 6) . . . . . 308 8.1.5 Sensors for Local Measurements (Chapter 7) . . . . . . 309 8.2 Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310 8.3 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314 List of Figures 1.1 Outline of the present thesis . . . . . . . . . . . . . . . . . . . 26 1.2 Concepts of tissue engineering . . . . . . . . . . . . . . . . . . 29 1.3 An illustrative proof of concept . . . . . . . . . . . . . . . . . 30 1.4 Bioreactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 1.5 Prometheus, the symbol of TE . . . . . . . . . . . . . . . . . . 39 2.1 Schematic representation of the metabolic pathways . . . . . . 54 2.2 Intracellular potassium ion release . . . . . . . . . . . . . . . . 65 2.3 Sensing locations . . . . . . . . . . . . . . . . . . . . . . . . . 69 2.4 The local and global TE microenvironments . . . . . . . . . . 70 2.5 Concept of chemical sensors . . . . . . . . . . . . . . . . . . . 72 2.6 Oxygen sensors for TE applications . . . . . . . . . . . . . . . 74 2.7 pH sensors for TE applications . . . . . . . . . . . . . . . . . . 76 3.1 Bonetissue ............................ 93 3.2 Osteoblastic cell functions with time . . . . . . . . . . . . . . 94 3.3 Fundamental requirements for mineralisation . . . . . . . . . . 96 3.4 Lineage potential of mesenchymal stem cells . . . . . . . . . . 99 3.5 Calcium distribution in serum . . . . . . . . . . . . . . . . . . 103 3.6 Experimental osteoblast cell culture model system . . . . . . . 109 3.7 Colourimetric method for determination of total phosphorus . 114 3.8 Calibration curve for absorbance of inorganic phosphorus . . . 114 3.9 Ion concentration of osteoblastic cell culture media . . . . . . 115 3.10 Ionic stability of cell culture medium . . . . . . . . . . . . . . 118 3.11 Osteoblast morpholgy . . . . . . . . . . . . . . . . . . . . . . . 120 3.12 The ionic extracellular environment of osteoblast cultures . . . 123 3.13 Quantitative calcium deposition in extracellular matrix . . . . 128 3.14 Early calcium deposition in extracellular matrix . . . . . . . . 129 3.15 Late calcium deposition in extracellular matrix . . . . . . . . . 130 3.16 Calcium depostion using non-corresponding medium . . . . . . 131 3.17 Estimation of the osteoblastic influence on the ionic extracellular environment . . . . . . . . . . . . . . . . . . . . . . . . . 131 18 List of Figures 19 3.18 Alkaline phosphatase activity . . . . . . . . . . . . . . . . . . 132 4.1 Bone tissue engineering . . . . . . . . . . . . . . . . . . . . . . 150 4.2 Calcium phosphate cement . . . . . . . . . . . . . . . . . . . . 153 4.3 Theoretical scheme on α-TCP hydrolysis . . . . . . . . . . . . 153 4.4 Fabrication steps of α-TCP and CDHA . . . . . . . . . . . . . 157 4.5 Medium compositions . . . . . . . . . . . . . . . . . . . . . . . 158 4.6 Scheme: experimental layor . . . . . . . . . . . . . . . . . . . 159 4.7 Calculation of regression coefficient . . . . . . . . . . . . . . . 161 4.8 Short-term CDHA/Culture Medium Interactions . . . . . . . . 165 4.9 Alternative view on Ca2+-sorption . . . . . . . . . . . . . . . . 168 4.10 Kinetics of hydrolysis of β-GP..................169 4.11 Ca/P sorption ratio . . . . . . . . . . . . . . . . . . . . . . . . 171 4.12 Long-term CDHA/culture medium interactions . . . . . . . . 172 4.13 Release kinetics of phosphorus from CDHA into DMEM . . . 175 4.14 SEM micrographs of CDHA exposed to cell culture medium . 176 4.15 FTIR of CDHA exposed to cell culture medium . . . . . . . . 177 4.16 Summary of CDHA ionic interactions with cell culture medium179 5.1 Experimental setup to study the chemical influence of CDHA on cellular response . . . . . . . . . . . . . . . . . . . . . . . . 194 5.2 Medium compositions . . . . . . . . . . . . . . . . . . . . . . . 195 5.3 Experiment layour . . . . . . . . . . . . . . . . . . . . . . . . 198 5.4 Ion concentration of osteoblastic cell culture media, 21 days . 203 5.5 Ion concentration of osteoblastic cell culture media, 48 hours . 206 5.6 Cell proliferation . . . . . . . . . . . . . . . . . . . . . . . . . 209 5.7 Total protein . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 5.8 Alkaline phosphatase activity . . . . . . . . . . . . . . . . . . 212 5.9 Calcium depostion during long-term cell culture . . . . . . . . 214 5.10 Quantitative evaluation of calcium depostion . . . . . . . . . . 218 6.1 The potentiometric principle . . . . . . . . . . . . . . . . . . . 230 6.2 Walther Hermann Nernst . . . . . . . . . . . . . . . . . . . . . 231 6.3 Traditional ISE designs . . . . . . . . . . . . . . . . . . . . . . 236 6.4 The phase boundary potential . . . . . . . . . . . . . . . . . . 239 6.5 Interpretation of experimental ISE response . . . . . . . . . . 240 6.6 Experimental potentiometric setup . . . . . . . . . . . . . . . 243 6.7 LabView flowcharts . . . . . . . . . . . . . . . . . . . . . . . . 245 6.8 Ag/AgCl reference electrode designs . . . . . . . . . . . . . . . 247 6.9 Setup for deposition of AgCl . . . . . . . . . . . . . . . . . . . 248 6.10 Calcium selective membrane components . . . . . . . . . . . . 249 List of Figures 20 6.11 Calcium selective electrode design . . . . . . . . . . . . . . . . 250 6.12 Setup for measuring Cl−activity during setting of α-TCP . . 252 6.13 Setups for measuring Ca2+ in bone TE applications . . . . . . 254 6.14 User interface of sensor software . . . . . . . . . . . . . . . . . 256 6.15 Micrographs of Ag/AgCl reference electrode . . . . . . . . . . 257 6.16 Potential response of Ag/AgCl electrode . . . . . . . . . . . . 258 6.17 Ag/AgCl electrode calibration curve . . . . . . . . . . . . . . 258 6.18 Stability of Ag/AgCl electrode . . . . . . . . . . . . . . . . . . 259 6.19 Calcium selective electrode: potential response . . . . . . . . . 261 6.20 Calcium selective electrode: response time . . . . . . . . . . . 261 6.21 Calcium selective electrode: selectivity . . . . . . . . . . . . . 262 6.22 Calcium selective electrode: stability . . . . . . . . . . . . . . 262 6.23 Calcium selective electrode: UV sterilisation . . . . . . . . . . 264 6.24 Calcium selective electrode: drift . . . . . . . . . . . . . . . . 264 6.25 Calcium selective electrode: biofouling . . . . . . . . . . . . . 265 6.26 Calcium selective electrode: cytotoxicity . . . . . . . . . . . . 265 6.27 Calcium selective electrode: cell proliferation . . . . . . . . . . 266 6.28 Cl−-activity during hydrolysis of α-TCP ............267 6.29 Real-time Ca2+-activity of culture medium exposed to CDHA 268 6.30 Calcium uptake by cells grown in DMEM with 3 mM β-GP . . 269 6.31 Calcium deposition in SAOS-2 cultures . . . . . . . . . . . . . 270 6.32 Real-time monitoring of cellular influence on Ca2+-activity . . 271 6.33 Aspect and viability of cells exposed to Ca2+-sensors . . . . . 272 7.1 Schematic view of pH microelectrode body . . . . . . . . . . . 287 7.2 Setups for pH measurement of curing α-TCP and set CDHA . 290 7.3 Voltammograms of Au microelectrodes . . . . . . . . . . . . . 291 7.4 IrO2-deposition on Au microelectrodes . . . . . . . . . . . . . 292 7.5 SEM of IrO2microelectrodes . . . . . . . . . . . . . . . . . . . 293 7.6 Open-circuit potential response of IrO2electrodes. . . . . . . . 294 7.7 Interior pH measurement during curing of α-TCP . . . . . . . 295 7.8 Exterior pH measurement during curing of α-TCP . . . . . . . 296 7.9 Exterior pH measurement of CDHA in cell culture medium . . 296 7.10 Calibration of IrO2electrodes in cell culture medium . . . . . 297 8.1 Cell cultures directly onto CDHA . . . . . . . . . . . . . . . . 314 8.2 Setup for measurements in local environments . . . . . . . . . 314 8.3 Local microenvironments around CDHA . . . . . . . . . . . . 315 List of Tables 1.1 The development of tissue engineering. . . . . . . . . . . . . . 31 1.2 Scaffold materials commonly used in tissue engineering . . . . 36 1.3 Strategic priorities in TE. . . . . . . . . . . . . . . . . . . . . 42 2.1 Glutaminolysis . . . . . . . . . . . . . . . . . . . . . . . . . . 59 2.2 Cell metabolites for TE monitoring . . . . . . . . . . . . . . . 61 2.3 Metabolic rates in mammalian cell cultures . . . . . . . . . . . 62 2.4 Intra- and extracellular ionic levels . . . . . . . . . . . . . . . 63 2.5 Performing enabling tools . . . . . . . . . . . . . . . . . . . . 71 3.1 The main components of bone tissue . . . . . . . . . . . . . . 98 3.2 Standard osteoblast culture models . . . . . . . . . . . . . . . 100 3.3 Standard composition of osteogenic medium . . . . . . . . . . 101 3.4 Inorganic salts content of cell culture media . . . . . . . . . . 105 3.5 Cell culture medium formulations. . . . . . . . . . . . . . . . . 106 3.6 Characteristics of ion selective electrodes . . . . . . . . . . . . 107 3.7 Ionic levels of complete media used for cell cultures . . . . . . 119 4.1 Bone-graft substitutes . . . . . . . . . . . . . . . . . . . . . . 149 4.2 Heating protocol to obtain α-TCP. ...............156 4.3 Ion levels of fresh cell culture medium . . . . . . . . . . . . . . 162 4.4 Sorbate model data . . . . . . . . . . . . . . . . . . . . . . . . 170 5.1 Extinction coefficients of alamarBlue . . . . . . . . . . . . . . 197 5.2 Ionic levels of cell culture media . . . . . . . . . . . . . . . . . 199 5.3 Total Protein, statistics . . . . . . . . . . . . . . . . . . . . . . 211 5.4 Alkaline phosphatase activity, statistics . . . . . . . . . . . . . 212 6.1 Ion electrodes of first and second kind categories . . . . . . . . 235 6.2 Conductive polymers applied to solid-state ISEs . . . . . . . . 236 6.3 Traditional ion-selective membrane components . . . . . . . . 238 6.4 Calcium selective membrane composition . . . . . . . . . . . . 249 6.5 Sensitivity and stability of Ag/AgCl electrode . . . . . . . . . 259 21 List of Abbreviations 22 List of Abbreviations Abbreviation Details AA Ascorbic acid aDMEM advanced Dulbecco’s Modified Eagle Medium AIROF Anodic iridium oxididation film ARS Alizarin red S β-GP β-glycerophosphate CDHA Calcium-deficient hydroxyapatite CPC Calcium phosphate cement DAQ Data aquisition Dex Dexamethasone DMEM Dulbecco’s Modified Eagle Medium ECM Extracellular matrix EGTA Ethylene glycol-bis(2-aminoethylether)-N-N-N’-N’-tetraacetic acid EMF Electromotive force FBS Foetal bovine serum FTIR Fourier transform infrared IIonic strength IEE Ionic extracellular environment ISE Ion selective electrode ISFET Ion selective field effect transistor McCoy McCoy’s 5A medium modified MMO Metal-metal oxide og Osteogenic PiTotal inorganic phosphorus PBS Phosphate buffered saline PEG Poly(ethylene glycol) rMSC Rat-derived mesenchymal stem cell SBF Simulated body fluid SEM Scanning electron microscope TCP Tricalcium phosphate TE Tissue engineering This page is intentionally left blank This page is intentionally left blank Chapter 1 Introduction ‘If engineering is the making of things useful to the human race, then tissue engineering is the making of tissues useful in some way to us’. (Curtis & Riehle, 2001) It has been estimated that one of five people reaching the age of 65 will benefit from some kind of tissue replacement therapy during their remaining life span (Vunjak-Novakovic & Kaplan, 2006). While these patients suffer or die every year waiting for a suitable transplant, medical doctors are facing two main problems: (1) limited availability of donors, and (2) counterproductive immunological reactions induced by allogenic transplants. To reduce the shortage of available organs, and to be able to custom-make transplants for patients, a new field of research called tissue engineering (TE) has emerged during the last decades. The concept of replacing or regenerating lost tissue with personalised tissue grown in vitro has until recently only been possible in science fictional stories. However, the growing clinical need for “spare parts” to damaged or diseased tissue, in combination with the scientific challenges contained within it have recruited scientists from various disciplines such as medicine, biology, chemistry, materials science, and engineering, to produce revolutionary products for the full spectrum of biotechnology; from early diagnostic testing to advanced stages of therapy. The human value of these products has turned TE into a high impact life science discipline, and the efforts made so far have made it possible to launch on the health care market commercial products such as artificial skin, bladders, and cartilage. However, in the shadaw of its progress, the field of tissue engineering has also suffered many failures during its development, and should still today be considered a field in its infancy in need for improvements in all its aspects. 25 Chapter 1 - Introduction 32 1.2.3 The Components of Tissue Engineering Repair and regeneration of tissue is a multi-level process of physical, chemical, biological, and engineering character. The minimal set of material required are the appropriate types of cell, but it is also common to use non-living substrata that can provide structure for tissue development. Also, drugs to induce cellular activity as well as bioreactors that control the cellular microenvironments, facilitate mass transport to the cells, and that provide the necessary biochemical and biophysical signals are commonly applied in tissue engineering (Radisic et al., 2008). These four components, (1) cells and extracellular matrix, (2) biomaterials, (3) drugs, and (4) bioreactors, are interdependent as each component and its intrinsic parameters have to be choosen based on the other components. Therefore it is important to understand the role of each one of them. 1.2.3.1 Cells and the extracellular matrix Being the architect of tissue, cells are the functional and fundamental elements of tissue science. They can be recruited from the patient itself (autologous) or be supplied from an external source (allogenic, or xenogenic if comes from a different species). In a few cases, controling cells is all that may be needed for a successful TE application, but the complete understanding of the process by which cells self-assemble into tissue and organs still remains as the main challenge for the TE field. Although specific functional properties of cells vary from tissue to tissue, tissue engineers heavily rely on cells capability to proliferate,differentiate,migrate, and communicate. The proliferative capacity of cells, i.e. when one cell divides into two daughter cells, is controlled by the extracellular environment and the state of cell differentiation. With other words, there are no single genes or molecular switches that control cellular proliferation, but this process is rather controlled by interactions with several extracellular regulatory molecules. These extracellular factors include stimulatory and inhibitory cytokines, the extracellular matrix (ECM) or substratum which cells make contact to, as well as the type and degree of proximity of neighboring cells. Very often, these same factors are also important regulators of differentiation (Oshima & Campisi, 1991). The understanding of cell functions means understanding of cell-cell and cell-ECM communication mechanisms (Rosso et al., 2004). While cell-cell interactions largely happens through transmembrane glycoproteins (cadherins), Chapter 1 - Introduction 33 cell-ECM interactions happen either via receptor-mediated signaling or via mobilization of growth or differentiation factors from the ECM. The ECM, which is composed of a great variety of molecules (e.g., proteins, glycosamineoglycans, proteoglycans, and adhesive glycoproteins) provides structure, localization, and guidance for cells. Many of the ECM molecules also play critical roles during development and postnatal tissue repair as they bind to receptors in the cell membrane to induce cytoskeletal deformations that activate second-messenger systems that can alter gene expression, adhesion, migration, proliferation, growth and cell death. Thus, much of the secret to control tissue formation under controlled laboratory conditions relies on providing conditions for creation of an ECM that can send appropriate signals to cells. The field of tissue engineering reclaims much of its potential to stem cells (Bianco & Robey, 2001), which are cells that can (1) self-renew, and (2) given the appropriate physical and chemical signals be induced to differentiate into one or more committed descendants, including fully functional mature cells (L.Weissman et al., 2001). Since long time it has been known that skin, other epithelia, and the bone marrow host stem cell populations that persist throughout life, but except these locations it was believed that most stem cells die out during tissue development. However, during the last ten years it has been demonstrated that many tissues of the adult body actually continue to host stem cells (Verfaillie, 2002), and that it therefore exist sources of cells with considerable potential for differentiation and multiplication, which makes stem cells an ideal candidate for use in tissue and regenerative medicine. Based on its origin one distinguishes between embryonic and adult stem cells, which have different capacity for self-renewal and differentiation. Embryonic cells are obtained from the inner cell mass of early embryons or the blastocyst, and they are able to proliferate in long-term cultures while maintaining capacity to form almost any cell type, and as such they are referred to as totipotent or pluripotent stem cells. On the other hand, stem cells obtained from adult tissues are more accesible than embryonic stem cells, but they can also be more difficult to expand in culture, and they are usually also more limited in their differentiation capacity only giving rise to a few specific cell types. The most common source of adult stem cells is the bone marrow which contains two types of stem cells: hematopoietic and mesenchymal cells. While hematopoietic cells are committed to differentiate only into mature blood cells, stromal mesenchymal stem cells have the ability to differentiate into a variety of adult mesenchymal tissues such as bone, cartilage, adipose, and muscle cells both in vivo Chapter 1 - Introduction 34 and in vitro. This plasticity renders mesenchymal stem cells potential for use in several clinical treatments. Despite the advantage of using intrinsic stem cells for tissue repair, there still exist many fundamental biological and engineering challenges, as well as ethical concerns, that must be overcome before clinical use (Weissman, 2000). It has to be aquicerd knowledge on how to control the microenvironment and signalling pathways leading to efficient differentiation and self-renewal of stem cells. It also has to be resolved how to perform in vivo delivery and integration to the host milieu. And finally there is the critical question whether one will be able to grow large enough populations from a patient rapidly enough when needed. 1.2.3.2 Biomaterials as scaffolds When damaged tissue is to be repaired (either in situ or ex vivo) it is of greatest concern that the structure and biological function of native extracellular matrix and environment is mimicked as much as possible both in terms of chemical composition and physical properties. Since initially the volume of cells and intercellular material is likely to be considerably less than that of the mature tissue, a supporting substrate that provides temporary mechanical properties will in general be necessary. Such non-living structural device, that on one hand is supposed to define the geometry of the replacement tissue, and on the other hand provide environmental cues to promote tissue regeneration, is referred to as the scaffold. At a minimum, the ideal scaffold for engineered tissue should (see for example Hutmacher (2001); Chen & Boccaccini (2008)): •posses adequate mechanical properties, •posses interconnecting pores of appropriate scale to favour supply of nutrition, cellular infiltration, tissue integration, vascularization, as well as removal of cellular waste products, •be made from material with controlled biodegradability or bioresorbability, so that the scaffold is replaced in such a way and at such a rate that the mechanical strength of the tissue remains constant, •have appropriate surface chemistry for cellular attachment, differentiation and proliferation, •be easily handled and fabricated into a variety of shapes and sizes. Chapter 1 - Introduction 35 In addition one obviously anticipates proper biocompatibility of the scaffold material itself, i.e. that the material does not exert a toxic effect on its environment, nor provoke any immunological or foreign body reaction (Williams, 2008). To ensure this issue, the scaffold is made from biomaterials that provide stable interactions with biological systems. The TE concept has gradually shifted focus of biomaterials research to extend from rigid prostheses to nowadays include materials that fulfill more than only one action. Nevertheless, the research activity and knowledge obtained from many decades of biomaterials development (e.g., hip implants, knee replacements, intraocular lenses, cardiovascular implants, etc) is of greatest importance also in tissue engineering, and it contributes to deeper understanding of how materials interact with the body on a cellular and molecular level. Therefore, the need for standard biomaterials can be well envisioned throughout at least this century (Ratner & Bryant, 2004). There is no such biomaterial that is capable of meeting with all criteria necessary for successful application towards engineering many tissues. Rather, the choice of biomaterial(s) to fabricate the scaffold depends on the particular type of tissue it will help to regenerate. In order to meet with the spatial and temporal constraints imposed by growing tissue, scaffold materials with tuned physico-chemical properties and biomimicking architectures have been developed to provide cells with the appropriate physical, chemical, and biological cues. The main materials studied and used can be divided into two main categories: synthetic or natural materials. Further classifications can be based on whether the material is organic or inorganic, inert or bioactive, durable or subject to remodeling, etc. Engineering solutions have been offered to obtain composite materials that can provide a wide array of properties in which the cellular niche can be tailored to guide tissue development. Table 1.2 provides a brief summary of some common materials used to produce scaffolds for different kinds of tissue (also see Hubbell (1995)). Chapter 1 - Introduction 36 POLYMERS poly(lactic acid), poly(glycolic acid), poly(lactic acid-co-glycolic acid), poly(caprolactone) poly(urethane) NATURAL MATERIALS Collagen, elastin, silk, chitosan, alginate HYDROGELS poly(HEMA), poly(ethylene glycol), gelatin, agar, fibrin CERAMICS Calcium phosphate, hydroxyapatite, bioglass, corals COMPOSITES Polymer/bioglass FABRICS Gore-Tex Table 1.2: Main classes and some examples of scaffold biomaterials used in tissue engineering. 1.2.3.3 Chemical and physical stimuli The environment where cells and scaffold are brought together to produce functional tissue is subject to many external parameters with significant importance for proper tissue development. Since tissue organisation is a process known to depend on adequate biomechanical and biochemical signalling in both space and time (Wang & Thampatty, 2006), it is necessary to provide cells with an appropriate physicochemical environment. This becomes especially evident when translating the biological models of two dimensional cell cultures into three dimensions. While traditional 2D-cell cultures rely on simple diffusion of gases and molecules for nutrient supply and metabolic activity, bigger and more complex tissue structures will require more sophisticated and efficient routes for supply and removal of metabolites. Considering that the distance between cells and blood capillaries ranges from 20 to 200 µmin vivo, distances which are largely overriden in vitro, together with the poor diffusion capacity and solubility of oxygen in aqueous solutions, hypoxia becomes one of the main limiting factors to obtain 3D cultures in vitro (Volkmer et al., 2008). One of the technical means to partly overcome such obstacles and to provide suitable physicochemical environments is the use of bioreactors, which are devices where biochemical processes can develop under closely monitored Chapter 1 - Introduction 37 and tightly controlled conditions. Bioreactors are nowadays considered a fundamental component in the ex vivo tissue growth process, as it provides dynamics to the cell culture. In short, bioreactors have had three main applications in TE (Butler et al., 2000): •Improving seeding efficiency and cell distribution of 3D scaffolds by convection or perfusion of the cell suspension. •Enhancing tissue quality throughout the full scaffold geometry by reducing negative concentration gradients within the construct through improved supply of nutrient and oxygen. •Applying defined regimes of physical forces to yield construct with regulated and improved strength and/or functionality. To meet with these purposes different bioreactor designs have been developed (Bilodeau & Mantovani, 2006), among which can be distinguished perfusion systems, rotating systems, and systems for mechanical loading (Figure 1.4). It is also possible to combine two or more systems. (1) (2) (3) Figure 1.4: Different types of bioreactors; (1) perfusion systems, (2) rotating systems, (3) systems for mechanical loading. Adapted from Martin et al. (2004). Perfusion systems allow cell culture medium to flow directly through a porous scaffold, and is used to both seed and maintain cell cultures in a 3D environment. The perfusion of medium is mainly supposed to avoid diffusional limitations and to help maintain cell viability and homogeneity Chapter 1 - Introduction 38 throughout the whole construct, and not only in the periphery which often happens during static cultures. Obviously, the effects of direct perfusion is dependent on medium flow-rate and the maturation stage of the constructs. Therefore, optimizing a perfusion bioreactor for the engineering of 3D tissues must address a careful balance between the mass transfer of nutrients and waste products to and from cells, as well as the retention of newly synthesised extracellular matrix components within the constructs (Martin et al., 2004). Also it has to be taken into account the fluid-induced shear stresses within the scaffold pores, and that the fluid flows along paths of least resistance and therefore do not solve the problem of heterogenity in tissue outcoumes (Lovett et al., 2009). Typically, and compared to static cultures, direct perfusion enhance growth, differentiation, and matrix production (e.g. Fassina et al. (2005)). As bioreactor designs continue to improve, and as one learns more on how mechanical forces can translate into biological response, bioreactors will not only provide improved physicochemical conditions for tissue development, but will also significantly contribute to make the TE process more reliable and to produce reproducable outcomes with decreased contamination risks (Martin et al., 2009). 1.2.4 Clinical Tissue Engineering - Can It Work? The capacity to auto-regenerate severely damaged tissue has been described for a few species, for example newt and zebra fish, but is absent or suppressed in almost all human organs. One well-known exception is the human liver which indeed regenerates even after significant loss or injury (Ankoma-Sey, 1999), and which was described already in the classical drama Prometheus (Wolfson, 2006). Prometheus, a titan in Greek mythology, was punished for bringing the powers of the Gods to the people, and for this he had his liver torn out by an eagle (Figure 1.5). When regenerated, the liver was repeatedly torn out again by the eagle. Mythological stories apart, and until it is revealed if the body possesses unleashed innate powers to regenerate other tissues than the liver, clinicians will basically have to rely on TE approaches. It is not until recently that tissue engineered products actually has turned into an alternative in clinical situations. Place et al. (2009) recently summarised the products commercially available, and this revealed a dominance in areas of skin, bone, and cartilage regeneration but the clinical basis extends beyond these common tissues. Below two particular cases illustrate how the top-down TE approach has been successfully implemented in clinical situations. Chapter 1 - Introduction 39 Figure 1.5: The sculpture Prometeo Engadenat by Josep Campeny i Santamaria is exposed in Igualada, Catalonia. It represents the story of how Prometheus has his liver eaten by an eagle. Example 1: A tissue-engineered bladder Using a top-down TE approach, Atala et al. (2006) successfully treated young patients suffering from end stage bladder disease. The aim was to decrease intravesical pressure and improve bladder compliance and continence, and for this purpose surgeons obtained autologous urothelial and muscle cells from bladder biopsies (1-2 cm2) from each patient. The cells were cultivated and expanded during several weeks to obtain large quantities that could be seeded on a biodegradable scaffold made from polyglycolic acid and collagen. The scaffolds had been designed specifically for each patient using templates based on CT scans. The initial size of the scaffold was 70-150 cm2, with a total thickness of about 2 mm, and it was seeded with about 7×108cells of each type. First the smooth muscle cells were seeded on the exterior of the scaffold, then a couple of days later the urothelial cells were seeded on the interior part. The seeded construct was then incubated for about four days before wrapped with omentum, which is commonly used in reconstructive surgery to enhance vascularization thanks to its rich blood supply, before implanted in the patients. The patients were followed up during a period of several years, and it was generally observed that the TE construct improved bladder function. The engineered bladder did not only show an adequate structural architecture and phenotype, it also reduced the leak point pressure. Neither was observed any negative metabolic consequences nor urinary calculi formation. The bowel function returned rapidly after implantation. Chapter 1 - Introduction 40 Example 2: A tissue-engineered allograft airway A young woman suffering from severe difficulties to breathe due to an obstructed airway caused by prolonged tuberculosis was treated by Macchiarini et al. (2008) using a tissue-engineered human trachea. In detail, a 7 cm tracheal segment was obtained from a 51-year-old female who had died of cerebal haemorrhage. The trachea was subsequently decellularised during a period of six weeks, before the graft was inserted in a bioreactor. A biopsy from the patient was done to isolate chondrocytes and epithelial cells respectively. The outside of the scaffold was seeded with chondrocytes while the interior was lined with epithelial cells as the construct rotated around its own longitudinal axis in the bioreactor. The total period of bioreactor culture was 96 hours, and the adherence to the matrix was estimated to be near 100%. The bioreactor phase was then followed by implantation. The patient recovered rapidly after operation. Lung-function tests done two months after implantation were all within normal range, and the patient has been able to return to normal life. Follow-up measurements interestingly showed a healthy, adjacent microvascular bed suggesting proper vascularization. 1.2.5 Need For Critical Assessment Expectation on the field of TE has been huge from both the public and private sector of healthcare. The business sector entered the field in an early stage, attracted by the potential market for TE products. For example, it has been estimated that the demand for tissue engineering as a consequence of loss and/or failure of tissues will account for as much as half of all medicalrelated problems in the United States, and taking into account that by 2040 as much as 25% of the US GDP is expected to be related to healthcare, translating this technology into clinical use is associated with enormous financial reward (Boehrs et al., 2008). But so far TE therapies have far from achieved the commercial goals, and the majority of launched TE products is related to acellular products where the regenerative response is due to either the biomaterial itself or the delivery of bioactive agents from biomaterials. The question to ask therefore is, why have the expectations not been lived up to yet? One has to understand that the tissue engineering treatment is an entity that spans widely separate phases such as cell recovery and manipulation to grafting. Although a great deal of new knowledge has been accumulated, and even some new therapies and treatments have been brought into practice, it is obvious that the field is still in its infancy with major limitations. Undoubtedly, a main problem is the limited mass transfer into the newly cre- Chapter 1 - Introduction 41 ated tissue. While most cells in our body are in close vicinity of capillaries that supplies oxygen and nutrients, all engineered tissue of today lack instant blood supply, and therefore engineered structures remain very primitive in relation to what is required. There exist a variety of strategies to circumvent the problem of vascularization (Lovett et al., 2009), but even if successful, still will remain the sub-sequent problem of integration to the host vascular system in vivo. To achieve clinical success, continued interdisciplinary efforts are required, and alliences between industry and academia need to continue to promote innovative, cost-effective, and high quality research (Pangarkar & Hutmacher, 2003). Organisations with the purpose to unite tissue engineers play important roles by providing community platforms and as well as strategic directions. One such organisation, MATES-IWG2which includes representatives from several American agencies (FDA, NIH, NASA, NSF, etc), published in 2007 a strategic plan3for investments in tissue science and engineering that identified four overreaching goals that needs to be developed for clinical success to be reached: 1. Understanding and controlling the cellular response. How do cells receive and respond to information from their local environment in establishing and maintaining tissues? 2. Formulating biomaterial scaffolds and the tissue matrix environment. To better understand the biology underlying the relationship between scaffold properties and cell fate. 3. Developing enabling tools. Complex, multiparametric inputs are required to assess the state of a tissue and the cells within it. This information will be supplied by improvements in high-throughput assays and instrumentation, imaging modalities, fabrication technologies, computational modeling, and bioinformatics. Additionally, tissue preservation technologies and bioreactors will facilitate the generation of tissues of demand. 4. Promoting scale-up, translation, and commercialization. Establishing protocols and instruments that contribute to increase reproducibility, robustness, and user-friendliness that can enable broad distribution of products. 2Multi-Agency Tissue Engineering Society Interagency Working Group 3Available at http://www.tissueengineering.gov/welcome-s.htm Chapter 1 - Introduction 48 Neuhof, H. (1917). Surgery, Gynecology and Obstetrics, 25, 383. Nichol, J. W. & Khademhosseini, A. (2009). Soft Matter, 5, 1312–1319. Oshima, J. & Campisi, J. (1991). Journal of Dairy Science, 74, 2778–2787. Pancrazio, J. J., Wang, F., & Kelley, C. A. (2007). Biosensors and Bioelectronics, 22, 2803–2811. Pangarkar, N. & Hutmacher, D. W. (2003). Tissue Engineering, 9(6), 1313–1322. Place, E. S., Evans, N. D., & Stevens, M. M. (2009). Nature Materials, 8, 457–470. Radisic, M., Marsano, A., Maidhof, R., Wang, Y., & Vunjak-Novakovic, G. (2008). Nature Protocols, 3(4), 719–736. Ratner, B. D. & Bryant, S. J. (2004). Annual Review of Biomedical Engineering, 6, 41–75. Rosso, F., Giordano, A., Barbarisi, M., & Barbarisi, A. (2004). Journal of Cellular Physiology, 199, 174–180. Semple, J. L., Woolridge, N., & Lumsden, C. J. (2005). Tissue Engineering, 11, 341–356. Sia, S. K., Gilette, B. M., & Yang, G. J. (2007). Birth Defects Research (Part C), 81, 354–361. Skalak, R., Fox, C., & Fung, B. (1988). Tissue Engineering chapter Preface, p. 1. Alan R. Liss, Inc. Stupp, S., LeBonheur, V., Walker, K., Li, L., Huggins, K., Keser, M., & Amstutz, A. (1997). Science, ,276, 384–389. Tysseling-Mattiace, V. M., Sahni, V., Niece, K. L., Birch, D., Czeisler, C., Fehlings, M. G., Stupp, S. I., & Kessler, J. A. (2008). The Journal of Neuroscience, ,28 (14), 3814–3823. van Blitterswijk, C., Stamatialis, D., Unandhar, H., Papenburg, B., Rouwkema, J., Truckenmuller, R., van Apeldoorn, A., Wessling, M., & de Boer, J. (2008). Tissue Engineering Part A, ,14 (5), 796. Verfaillie, C. M. (2002). Trends in Cell Biology, ,12 (11), 502–508. Chapter 1 - Introduction 49 Volkmer, E., Drosse, I., Otto, S., Stangelmayer, A., Stengele, M., Kallukalam, B. C., Mutschler, W., & Schieker, M. (2008). Tissue Engineering, ,14 (8), 1331–1340. Vunjak-Novakovic, G. & Kaplan, D. L. (2006). Tissue Engineering, ,12 (12), 3261–3263. Wang, J.-C. & Thampatty, B. (2006). Biomechanics and Modelling in Mechanobiology, ,5, 1–16. Weissman, I. L. (2000). Science, ,287, 1442–1446. Williams, D. F. (2008). Biomaterials, ,29, 2941–2953. Wolfson, W. (2006). Chemistry & Biology, ,13, 233–234. Wolter, J. & Meyer, R. (1984). Transactions of the American Ophtalmological Society, ,82, 187. This page is intentionally left blank Chapter 2 The Role and Use of Sensors for in vitro Tissue Engineering Applications ‘Much of the information is not missing, just not used’ (Vunjak-Novakovic & Kaplan, 2006) One of the keys for future clinical success in tissue engineering (TE) will be the production of consistent and reproducible cell-based products of high quality and high yield. To accomplish such products, the production process needs to be extremely precise and preferably adjusted continuously to the changing demands imposed by the developing tissue. As of today, the general TE process is still essentially manual, which often leads to sub-optimal productivity. Therefore, development and application of performing enabling tools, for example different sensing modalities, that provide quantitative information on product fate, is highly acknowledged. Performing enabling tools can be beneficial to the TE process in two ways. First, since tissue growth/regeneration evolves with time, any TE application would benefit from real-time monitoring. In that way, important parameters such as growth and different bioreactions can be controlled, just as well as early indications of catastrophic events, such as contamination or cessation of growth, can be obtained. Second, performing enabling tools could provide information from TE environments crucial for proper tissue development but which are normally inaccessible for standard measuring technologies. With other words, properly designed and applied performing enabling tools could contribute with both spatial and temporal information on many TE processes that until today have been overlooked, or even impossible to obtain. 51 Chapter 2 - The Role and Use of Sensors in Tissue Engineering 52 2.1 What to Measure? The tissue engineering environment is complex by nature, and complexity increases since each type of tissue requires its specific environment for adequate development and function. The standard in vitro TE environment is composed of cells and scaffold materials usually contained in some kind of bioreactor. Each of these standard components (i.e., cells, materials, bioreactor) can by themselves influence the environment they reside in. For example, cellular activity influences total biomass, consumption of nutrients, and production of bioproducts (matrix); scaffold materials are often designed to be either biodegradable or bioactive, resulting in changing material composition with time; bioreactors are commonly used to establish a stable environment (e.g., temperature and gas levels) but they can also induce changes in flow rates, mechanical stresses, or chemical gradients. It is therefore of greatest importance to know the activity of each of these standard components to be able to predict possible synergistic or agonistic effects that may be induced when different TE components are combined in one and the same environment. In the following sections it is discussed how activity of cells (Section 2.1.1- 2.1.2) and materials (Section 2.1.3) may influence the final in vitro TE environment. Basically this environment is an aqueous solution containing all the essential molecules for proper cell function. It is often designed to mimick the in vivo biological environment with respect to parameters such as osmolarity, pH, gas levels, vitamin, amino acid and protein content. Between these parameters and any standard TE component may arise reactions that can be directly and/or indirectly indicative of specific TE events, and that could be advantageously measured by correctly applied performing enabling tools. However, monitoring a specific parameter should be considered worthwhile only if that parameter (1) has importance for the TE construct development, (2) is produced/consumed at such amount that the initial concentration is significantly changed, and (3) that the rate of activity is neither too fast nor too slow to be detected. 2.1.1 Relevant Cell Metabolites for TE Monitoring Cellular metabolic activity drives the conversion of energy stored in carboncontaining fuels to adenosine triphosphate (ATP), which is a versatile source of chemical energy that can be used for production of complex molecules needed for construction, restoration, and maintenance of tissues and organs within the body. In presence of oxygen, cells derive their ATP mainly through cellular respiration (2.1): Chapter 2 - The Role and Use of Sensors in Tissue Engineering 53 C6H12O6(aq) + 6O2(g) →6CO2(g) + 6H2O(l) + 36ATP (2.1) The cellular respiration oxidises glucose into carbon dioxide, and releases free energy stored as ATP. But what appears as a straight-forward and efficient energy-transfer (2880 kJ per mole of glucose is obtained) involves an intricate series of chemical transformations before the full amount of ATP is harnessed: 1. Glycolysis, an anaerobic process which mainly transfers the chemical energy of glucose into pyruvate, but also creates a small amount of ATP (see Section 2.1.1.1). 2. Pyruvate oxidation, which converts glycolytic pyruvate into acetyl-CoA (see Section 2.1.1.3). 3. Citric acid cycle, which converts acetyl-CoA into CO2, with the creation of additional ATP (see Section 2.1.1.3). 4. Respiratory chain, where finally the energy stored in different transfer molecules is harvested and large amount of ATP is created, and which the cells can use for anabolic reactions (see Section 2.1.1.4). Each of above sub-reactions influences, and is influenced by, the instantaneous composition of both the intra- and extracellular environment, and therefore also the TE environment. In addition to above reactions, there also exist alternative or complementary metabolic routes for cells to produce energy. One of these pathways is fermentation which is anaerobic production of ATP following the glycolytic pathway (see Section 2.1.1.2). Another is the glutaminolytic pathway which can contribute with moderate levels of ATP derived from glutamine (see Section 2.1.1.5-2.1.1.6). All these main metabolic pathways are discussed below, and they are also schematically illustrated in Figure 2.1. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 54 Glucose Glucose Pyruvate Lactic acid Lactate H + Glutamine Acetyl-CoA O2 NH + Lactate H+ NH + CO2 CO2 H + - 3 HCO - 3 HCO Figure 2.1: Schematic representation of the main metabolic pathways in mammalian cells. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 55 2.1.1.1 Glucose Glucose serves the metabolic demands for survival, growth, proliferation and function in most mammalian cells. After entering the cell through specific transport membrane proteins, the molecule is subject to several sequent enzyme reactions. First, the glucose molecules gain a phosphate group derived from ATP, and is cleaved into two identical sub-compounds. Each of these sub-compounds binds phosphate while NAD+is reduced to NADH + H+,1. The bound phosphate is later transferred to ADP to form ATP. Finally, the remaining structure is dehydrated andconverted into two molecules of pyruvate (which is the anion of pyruvic acid, C3H4O3). Thus, the net reaction of glycolysis can be written as in (2.2): C6H12O6+ 2NAD++ 2Pi→2pyruvate + 2NADH + 2H++ 2ATP + 2H2O (2.2) Although the energetic yield from glucolysis only contributes with about 5% of total ATP-yield from the complete glucose catabolism, glycolysis can by itself account for a major part of energy production. For example, both neutrophils and myoblasts derive up 95% of their energy from glycolytic activity (Wiley & Beeson, 2002; Gaitain et al., 1997). Therefore, glucose consumption is a highly relevant indicator of cellular capacity to produce fuel for cellular work. However, metabolic consumption rate of glucose (qglucose) varies with cell phenotype and also between species (see Table 2.3), as was clearly demonstrated by Schop et al. (2009) who studied mesenchymal stem cells obtained from human, rat, and goat, respectively. Even more important is that the glucose consumption rate varies with level of cell differentiation. Komarova et al. (2000) measured qglucose of osteoblasts throughout a period of 14 days, and they observed fluctuations between about 0.2-0.5 µmol h−1 / 106cells, with the higher glycolytic activity detected in highly mature osteoblast cultures producing nodule mineralization. In a similar way, Collins et al. (1998) observed that glycolytic activity increased its role in energy production as haemopoietic cells matured. On the contrary, Tsao et al. (2005) observed that consumption rate of glucose was higher in the beginning of ovary cell cultures in suspension when cells were still actively growing. Taken together, glucose consumption is indicative of cellular metabolic activity, but its rate cannot be expected to be constant over time. Therefore no direct relationship between glucose consumption and cell growth can be clearly established. 1NAD+, nicotinamide adenine dinucleotide, is a coenzyme that works as a major redox carrier and universal energy intermediate in cells. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 56 2.1.1.2 Lactate Glycolysis is a completely anaerobic process, which to continue requires constant access to NAD+. In an anaerobic cellular ambient (e.g., muscle tissue), NAD+can be regenerated through incomplete oxidation of glucose. In that case, pyruvate is reduced by lactate dehydrogenase to lactic acid (2.3). At physiological pH, lactic acid is dissociated into lactate and H+(Gladden, 2004), with the former being extruded to the extracellular environment via monocarboxylate carriers (Poole & Halestrap, 1993) and anion exchange proteins. pyruvate + NADH + H+→lactate + NAD+(2.3) In strictly anaerobic conditions all consumed glucose will be converted to lactate, and the molar ratio of produced lactate to utilised gluose (YLac/Glu) reaches 2. However, under aerobic conditions lactate production is normally a secondary process for cells to obtain energy and to regenerate NAD+, and YLac/Glu naturally decreases to levels below 2. The yield coefficent YLac/Glu therefore can be used as a preliminary indicator of which metabolic routes cells use to produce energy (Obradovic et al., 1999; Komarova et al., 2000; Schop et al., 2009). The cellular lactate production rate (qlactate) can vary with many parameters, among them oxygen pressure, cell type and differentiation level, and due to environmental influence on qlactate, lactate production does not directly relate to cell growth. However, in cases where lactate can be converted back to pyruvate (which is the case for mammalian cells that undergo the metabolic Cori cycle), Tsao et al. (2005) demonstrated experimentally the direct relationship at any given timepoint (t) between the integral of viable cells and the cumulated amount of glucose and lactate (QGL) calculated according to Equation 2.4; QGL = (G0+L0)−(Gt+Lt) (2.4) where G0and L0represent initial glucose and lactate concentrations, respectively. In most mammalian cells, lactate will however not regenerate pyruvate, but rather accumulate in the extracellular environment where it can reach significant levels (>30 mM). Accumulation of lactate has been mentioned to be responisble for toxic/inhibitory effects on cell growth (Chen et al., 2009), but since lactate production is associated with decreased pH, it has been Chapter 2 - The Role and Use of Sensors in Tissue Engineering 57 debated which of the two effects is the main responsible for the provoked effects. Among others, Patel et al. (2000) demonstrated that lactate accumulation without lowering the pH has less influence on cell growth than the combined effect of lactate accumulation and pH decrease. 2.1.1.3 Carbon dioxide Reaching the mitochondrion, pyruvate can be enzymatically converted to acetyl-coenzyme A (acetyl-CoA). The reaction also produces CO2and it reduces NAD+to NADH and H+(2.5); pyruvate + NAD++ CoA →AcetylCoA + NADH + H++ CO2(2.5) The produced acetyl-CoA is then the main input for a series of cyclic reactions known as the citric acid cycle (or the Krebs cycle, or the tricarboxylic acid cycle) that completes the conversion of glucose (as well as other carbon-containing fuels such as fatty acids and certain amino acids) to CO2 and electron carrier molecules such as NADH and FADH2, and with the additional creation of a small amount of ATP (2.6); AcetylCoA + 3NAD++ FAD + ADP + Pi+ 2H2O →CoASH + 3NADH + 3H++ FADH2+ ATP + 2CO2 (2.6) In its gaseous state CO2is a higly labile molecule that rapidly reacts with water to form carbonic acid (H2CO3), which further dissociates into hydrogen and bicarbonate ions (HCO− 3) (see equilibrium reaction 2.7). CO2(g) ⇋CO2(aq) + H2O⇋H2CO3⇋HCO− 3+ H+⇋CO2− 3+ H+ K1K2 (2.7) At physiological conditions (temperature 37◦C, pH 6.8, ionic strength 0.1M), Goudar et al. (2007) estimated the equilibrium constants to be K1= 10−6.07, and K2= 10−10.04, which would mean that carbon dioxide produced by cells exists as combination of HCO− 3(about 84%) and CO2and H2CO3 (together about 16%). While the presence of CO2in the atmosphere is extremely low (ppm), in vitro cultures are normally done at 5-7% CO2to reflect levels in vivo (about 5.3%), and to act as a buffer (Csete, 2005). As gaseous CO2diffuses freely across cell membranes, and since HCO− 3 can enter or leave the cell through specific transport membrane proteins, Chapter 2 - The Role and Use of Sensors in Tissue Engineering 64 phosphate (Beck, 2003), and magnesium (Walker, 1994; Romani, 2007) are known to influence different cellular events, calcium is undoubtedly the ion with highest documented influence on cellular life. Calcium acts both as a secondary messenger to influence intracellular Ca2+ concentration (through the actions of calcium channels, exchangers and pumps in the cellular membrane), as well as a first messenger through membrane-located Ca2+-receptors (Hofer, 2005). Via further intracellular processing cellular functions such as spreading (ONeill & Galasko, 2000), proliferation and differentiation (Eklou- Kalonji et al., 1998), apoptosis (Lin et al., 1998; Lorget et al., 2000), matrix production (Nakade et al., 2001), hormone regulation (Ahlstrom et al., 2008), bone tissue development (Dvorak et al., 2004), and angiogenic events (Aguirre et al., 2010) can be regulated. Although ionic strength of the cell culture environment is normally maintained constant, single ionic activity can still be subject to local significant changes during tissue development. For example, in Chapter 3 and 6 of this thesis it is investigated how extracellular concentrations of calcium and phosphorus are subjected to osteoblast activity. Disturbance of homeostasis may also be used as an indicator of catastrophic cellular events, as is discussed in the following section. 2.1.2.2 Cell death As a consequence of cell death, metabolic activity is ceased and the cellular membrane looses its permselective properties. Therefore, upon cell death, both the intra- and extracellular ionic environment will be disturbed. The major change in the extracellular environment would be an increase of both potassium and phosphorus. This increse, ∆[X]o, where [X] is either K+or Pi, would depend on the extent of damage (i.e., number of damaged cells, N) as well as the total volume of the cell culture environment (V, dm−3), and could be roughly estimated with Equation 2.9: ∆[Xo] = [Xnew]−[Xbkg] = NVcell[X]i+V[X]bkg NVcell +V−[Xbkg] (2.9) where Vcell is the volume of one single cell, [X]iis the intracellular ion concentration of X, and [X]bkg is the standard background extracellular concentration. In Figure 2.2(a) it is visualised how the extracellular potassium concentration is affected by release of intracellular potassium for different number of damaged cells (0 < N < 107) in different cell culture volumes (0 < V < 10 cm−3). The background potassium concentration was set to 5.5 mM and Chapter 2 - The Role and Use of Sensors in Tissue Engineering 65 [K+]ito 140 mM, and Vcell was set to 1.0×10−12 dm−3which would correspond to a radius of 6.2 µm for a perfectly spherical cell. The validity of Equation 2.9 was tested experimentally with a fixed sample volume (V= 0.30×10−3dm−3), and by lysing different numbers of SAOS- 2 osteoblast cells grown on tissue culture polystyrene with 0.1% Triton-X diluted in phosphate buffered saline solution. The concentration of potassium was then measured using a commercial potassium-selective electrode 2, and [K+]bkg of the lysis solution was determined to 2.49 ±0.04 mM. Experimental data confirmed that potassium was released, and that the release could be well correlated to number of lysed cells (Figure 2.2(b)). 0 2 4 6 8 10 0 0.002 0.004 0.006 0.008 0.01 0 0.5 1.0 1.5 Sample volume (l) Number of damaged cells (x 10 ) ∆[K ] (mM) + 6 (a) Potassium release due to membrane damage as a function of cell number and culture volume. 0.0 0.0 0.2 0.4 0.6 0.8 1.0 Experimental data Model Number of cells (x 10 ) 5 5.0 10.0 15.0 20.0 ∆[K ] (mM) + (b) Experimental data compared to model data for Vsample = 0.3 ml. Data is mean ± standard deviation, n= 4. Figure 2.2: Damage to the cellular membrane causes release of intracellular K+. 2See Section 3.3.2 for details on ion measurements. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 66 2.1.3 Biomaterials Activity The general concept behind the use of biomaterials in tissue engineering is to support cells with an appropriate environment that favours cellular growth and function. To some extent it is possible to control such cellular response through biomaterials properties. For example, it is well established that material properties such as elasticity,wettability, and topography play very important roles and are known to influence events such as cell attachment, contractility, motility, spreading, protein adsoprtion, and cell differentiation (Discher et al., 2005; Bao & Suresh, 2003; Vogler, 1998; Dalby et al., 2007). In addition to such physical materials properties, many biomaterials also interact (bio)chemically with the TE environment, and doing so they can provide (or deplete) the environment with soluble factors that cells metabolise or use as messenger molecules. One of the main material classes in TE is polymeric materials, which are often designed to degrade through one or another mechanism as new tissue is formed (Woodruff & Hutmacher, 2010). Biodegradability is not necessarily restricted only to polymers, but can also be prescribed certain metals (Yun et al., 2009). In brief, these materials are degraded through hydrolysis or corrosion reactions to produce monomeric components that can be removed by natural pathways (but not always). Antoher major category of chemical interaction that can be induced by certain biomaterials is through ionic substitution/exchange which causes ions to be either released into the extracellular environment or absorbed by the material. Among materials that induce such interaction are the ceramics, bioglasses, and apatitic materials, as well as composites containing any of these materials as one of its components. Materials that release or absorb ions to and from the TE environment has been reported for a variety of ions of different biological relevance; e.g. calcium and phosphorus (Engel et al., 2008), magnesium (Yang & Zhang, 2009), strontium (Qiu et al., 2006), zinc (Storrie & Stupp, 2005), copper (Barralet et al., 2009), and fluoride (Itota et al., 2004). A third approach to tune the biochemical properties of a biomaterial is to load it with drugs or other promotor molecules that are then released upon implantation or during in vitro cell culture prior to implantation (Cartmell, 2009; Lee & Shin, 2007). Local release improves efficiency by delivering higher concentrations and can also decrease systemic toxicity and possible side effects of the molecule released. Materials popularly used as delivery vehicles are polymers (Luo & Prestwich, 2001), proteins (Friess, 1998), or ceramics (Ginebra et al., 2006) since these materials possess inherent properties (such as biodegradability or that they favour ionic exchange) that can facilitate controlled release of molecules previously bound/trapped/encapsulated Chapter 2 - The Role and Use of Sensors in Tissue Engineering 67 to the scaffold. This approach has been explored both to deliver antibiotics to the implant site to impede bacterial infections (Jiang et al., 2010), as well as bioactive molecules to accelerate extracellular matrix production and tissue integration (Porter et al., 2009). An alternative approach to suppress bacterial infections in the vicinity of implanted biomaterials is through material-induced release of nitric oxide (NO) from polymeric and sol-gel coatings (Gupta & Kumar, 2008). Monitoring and characterisation of biomaterials activity (let it be either through degradation, ionic exchange, or release mechanisms) distinguishes as an important sub-field in biomaterials research. Independent of the mechanism of interaction, any biomaterial designed to interact with its environment should do so at favourable rate and magnitude. While the rate of reaction is principally determined by the actual reaction mechanism, the magnitude of reaction is commonly an equilibrium between amount of unreacted material and amount of reactor material at any time, implying that the chemical activity of a biomaterial often can vary with its mass and exposed surface area. Furthermore, external physical parameters such as pH, temperature, and ionic strength, may significantly influence the reactivity. Moreover, the biological milieu with which the biomaterial is supposed to interact is usually of compex nature containing proteins, inorganic salts, vitamins, amino acids, and other cells. That imposes difficulties to precisely predict a priori the biochemical response of the biomaterial upon implantation, as only slight variations in the chemical composition of the aqueous environment may provoke completely different interaction responses (see further Chapter 4). As a consequence, characterisation of biochemical activity of biomaterials should ideally be done in conditions that simulate the in vivo environment as much as possible, and it is emphasised once for all that the biochemical activity of any material always should be put in perspective of the environment in which it has been characterised. Finally, it is desirable that the chemical characterisation also provides information on spatial differences of the material reactivity. The latter can be provoked either by the material itself (e.g. changing porosity) or for example through cell-material interactions (e.g. biofilm-coating). Until today, limited information about local microenvironments created around the cell-material interace have and can be obtained. Yet, this is the particular zone where the chemical activity of the biomaterial must be optimised in order to provide adequate cellular response. This issue will be further discussed in Section 2.2.1, and is also illustrated in Figure 2.4. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 68 2.2 How to Measure? When knowing which parameters that are of interest to measure, the next decision that has to be taken is how to detect them. This decision is however dependent on where in the TE environment the detection is to be done. 2.2.1 Where to Measure? Traditionally, four different detection approaches, or sensing locations, have been described in literature (Figure 2.3): (1) invasive sensing, (2) noninvasive sensing, (3) sampling or shunt sensing, and (4) differential measurements (Rolfe, 2006). While methods (1)-(3) are direct measurements of the TE environment, differential measurements are of indirect character as they measure the parameter of interest as a difference at the inlet and the outlet of the TE environment during medium flow. From such measurements the amount of biomass, Cx(t), is determined according to Equation 2.10: F V([X]in −[X]out) = Cx(t)q0(2.10) where Fis the flow rate of the medium, Vis the volume of the environment, [X]in and [X]out are the concentrations or partial pressures of the parameter of interest at the inlet and the outlet respectiviely, and q0is the consumption or production rate of the parameter of interest (Janssen et al., 2006b). Difference between [X]in and [X]out indicates activity from any of the TE components included in the volume V. Normally all four sensing locations mentioned above provide representative information on the average, or the global, extracellular tissue engineering environment, and they are appropriate when the environment is homogenous. However, increased complexity in cellular cultures (e.g., co-cultures) and materials designs (e.g., porosity, topography, degradation, etc) provokes that the global environment becomes less and less representative of the infinite numbers of microenvironments created immediately adjacent to cells growing in proximity to another TE component (Figure 2.4). Examples of local TE microenvironments are the interiors of porous scaffolds and interfaces between cells and the surface of scaffold materials. These local environments are exceptionally small (approaching size and volume of individual cells) and may therefore be subject to higher influence from changes caused by any TE component. For example, the distance between cells attached to a solid material is of the order of 5-30 nm (Curtis, 2001), thus creating a volume of nl-µl Chapter 2 - The Role and Use of Sensors in Tissue Engineering 69 between cells and the surface of the material where signals crucial for cell survival and functioning are transmitted and received. Measurement of local environments is important because these are the environments that cells experience and that decide cellular fate, and it can provide information on the spatial variations in the TE environment. Obviously, getting access to these microenvironments require not only sophisticated techniques, but also an increased number of measuring points to obtain representative data. Therefore, in addition to the decision on the actual sensing location, it is also important to determine the appropriate number and exact position of the sensors to effectively monitor the culture parameters and to obtain information-rich data (Lim et al., 2007, 2008). (a) Invasive plastic optically active sensor layer TE area Cells (b) Non-invasive 3 0 0 2 5 0 1 5 0 1 0 0 5 0 0 (c) Sampling Sensor 2 (outlet) Bioreactor Scaffold with cells Cell culture medium flow Sensor 1 (inlet) (d) Differential Figure 2.3: Different sensing locations. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 70 Biomaterial surface Global environment Local environment Figure 2.4: Two different TE environment can be defined; the global and the local environments. 2.2.2 Sensors for Tissue Engineering Applications A multitude of techniques, among them different imaging and sensing modalities, can be used to assess the TE environment at different levels (see Table 2.5, and reviews by Vojinovic et al. (2006), Starly & Choubey (2007), Mason & Hoare (2007), Lim et al. (2007), and Mather et al. (2007)). Ideally, such techniques or tools should (1) possess adequate sensitivity, specficity, and stability towards the analyte of interest, (2) not interfere with any of the TE components, (3) allow for real-time monitoring of data, and (4) be possible to combine in order to obtain multi-parametric data on the TE process. Most sensors rely on a common principle (Figure 2.5); a chemical reaction between analyte and recognition element (receptor) that produces a signal that can be transduced and detected with for example electrochemical or fibre optics methods (Marazuela & Moreno-Bondi, 2002). The heart of the sensor is obviously the receptor which should (1) react only with the analyte of interest, bind the analyte (2) irreversibly, and (3) at a concentration that corresponds to the range of interest. Commonly used receptors are based on biological components such as proteins, enzymes, antibodies, lipid bilayers, liposomes, DNA, whole cells, or even intact organs or tissue. In such cases the sensor is referred to as a biosensor (Kissinger, 2005). Non-biological receptor elements include synthetically derived ionophores, different metal oxides, fluorophores, and phosphorescent molecules. In Sections 2.2.2.1 - 2.2.2.3 it is discussed the major sensor classes, and their possible application in the field of tissue engineering. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 71 Parameter Enabling tool Detection technique Reference Glucose Enzyme-based sensors Ampereometry Wang (2008) Nano-particles Fluorescence Billingsley et al. (2010) Lactate Enzyme-bases sensors Ampereometry Romero et al. (2010) CO2Severinghaus electrode Potentiometry Zhao & Cai (1997) Optical film sensors Fluorescence Ge et al. (2003) O2Clark electrode Ampereometry Andreescu & Sadik (2005) Microparicles Fluorescence Acosta et al. (2009) Thin film sensors Optical Thomas et al. (2009) pH ISFET Potentiometry Poghossian et al. (2009) MMO electrodes Potentiometry Glab et al. (1989) ISE Potentiometry Ions ISE Potentiometry Radomska et al. (2008) Optodes Fluorescence Hisamoto & Suzuki (1999) Amino acids and Porphyrin-tailored electrode Optical absorbance Awawdeh et al. (2003) proteins Cu-tailored electrodes Ampereometry Luque et al. (2007) Table 2.5: Parameters of interest for TE monitoring, and corresponding performing enabling tools. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 72 Receptor Analyte Transducer Figure 2.5: The concept of chemical sensors. 2.2.2.1 Gas sensors On-line gas analysis is an attractive method to monitor cultivation of growing cells, and it can also be used as the basis for changing medium or as indicator of contamination. The most common means for measurement of dissolved oxygen and carbon dioxide are the Clark and the Severinghaus electrodes, respectively, which both have been described in detail by, for example, Hahn (1980). These electrodes contain a gas-permeable membrane that separates the site of detection from the sample subject to investigation. When gas passes the permeable membrane of the Severinghaus electrode it encounters a buffer of bicarbonate solution which pH obviously is under influence of CO2. As the pH of the buffer solution is potentiometrically measured by an internal pH electrode, an indirect measurement of the concentration of CO2is obtained. In the Clark electrode, the detection mechanism is based on reduction of oxygen at a cathode (normally made from platinum) which is negatively polarised against a reference electrode. As oxygen that passes the permeable membrane is reduced (Equation 2.11), a current flow in the circuit and a linear relationship between current and PO2can be established. O2+ 2H2O + 4e−→4OH−(2.11) A consequence of the Clark electrode detection mechanism is that oxygen is consumed by the electrode during operation, and this can potenially alter the oxygen concentration in culture. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 73 Although the underlying detection principle of both the Clark and the Severinghaus electrodes limit their use to be of either invasive or differential character, still they can be miniaturised down to a few micrometers which could allow for measurements in extremely small volumes (Lee et al., 2007; Zhao & Cai, 1997; Suzuki et al., 1999), as well as high-troughput applications (Andreescu & Sadik, 2005). Both sensors have been applied in bioprocess monitoring, and to some extent also in tissue engineering applications. For example, Janssen et al. (2006a,b) demonstrated the usefulness of the Clark oxygen electrodes to monitor cell growth inside a closed perfusion bioreactor designed to produce bone tissue for clinical applications. For this purpose they applied low-consuming electrodes in a differential-mode, and from Equation 2.10 they could obtain accurate data on cell growth within the bioreactor during the exponential growth phase. An alternative mean to measure gases is based on quenching of fluorescence or phosphorescence. In that case, gas sensitive luminophores (commonly variants of phosphorescent metal-porphyrin dyes or fluorescent ruthenium(II) complexes for oxygen, and 8-hydroxypyrene-1,3,6-trisulfonic acid, HPTS, for carbon dioxide) are incorporated either within (1) an optically transparent and gas-permeable matrix which is directly deposited onto an optical sensor, or (2) microparticles that are suspended in the TE environment. In both cases, the emission of the luminescent dye is quenched by the gas at a rate dependent upon gas concentration. Commonly optical gas sensors contain an additional reference dye which is independent of gas concentration, and so the intensity (I) of the luminescence is calculated as the ratio of luminescence obtained from the indicator dye (Iind) and the reference dye (Iref ), and which further obey the Stern-Volmer equation (Equation 2.12), I=Iind Iref =I0 1 + Ksv[Q] (2.12) where I0is the intensity of fluorescence in absence of gas at partial pressue [Q] (atm), and Ksv is the Stern-Volmer quenching constant (atm−1). Often, when the luminophore is embedded in a polymeric support its response deviates from the Stern-Volmer equation, and corrections for this has to be taken into account. Optical gas electrodes have been proven successful candidates for longterm continuous monitoring in cell culture conditions. For example, Gao Bibliography Acosta, M. A., Ymele-Leki, P., Kostov, Y. V., & Leach, J. B. (2009). Biomaterials, 30, 3068–3074. Aguirre, A., Gonzalez, A., Planell, J., & Engel, E. (2010). Biochemical and Biophysical Research Communications, 393, 156–161. Ahlstrom, M., Pekkinen, M., Riehle, U., & Lamberg-Allardt, C. (2008). Bone, 42, 483–490. Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., & Walter, P. (2002). Molecular Biology of the Cell. Garland Science (Taylar & Francis Group), 4th edition. Aledo, J. C. (2004). BioEssays, 26, 778–785. Andreescu, S. & Sadik, O. A. (2005). Methods, 37, 84–93. Antonisse, M. M. & Reinhoudt, D. N. (1999). Electroanalysis, 11 (14), 1035–1048. Atkuri, K. R., Herzenberg, L. A., Niemi, A.-K., Cowan, T., & Herzenberg, L. A. (2007). PNAS, 104 (11), 4547–4552. Awawdeh, M. A., Legako, J. A., & Harmon, H. J. (2003). Sensors and Actuators B, 91, 227–230. Bao, G. & Suresh, S. (2003). Nature Materials, 2, 715–725. Barralet, J., Gbureck, U., Habibovic, P., Vorndran, E., Gerard, C., & Doillon, C. J. (2009). Tissue Engineering: Part A, 15 (7), 1601–1609. Baxamusa, S. H., Montero, L., Dubach, J. M., Clark, H. A., Borros, S., & Gleason, K. K. (2008). Biomacromolecules, 9, 2857–2862. Beck, G. R. (2003). Journal of Cellular Biochemistry, 90, 234–243. 80 Chapter 2 - The Role and Use of Sensors in Tissue Engineering 81 Berger, C., Horrocks, B., & Datta, H. (1999). Electrochimica Acta, 44, 2677–2683. Bergveld, P. (2003). Sensors and Actuators B, 88, 1–20. Beyenbach, K. W. & Wieczorek, H. (2006). The Journal of Experimental Biology, 209, 577–589. Billingsley, K., Balaconis, M. K., Dubach, J. M., Zhang, N., Lim, E., Francis, K. P., & Clark, H. A. (2010). Analytical Chemistry, 82, 3707–3713. Blum, J. S., Temenoff, J. S., Park, H., Jansen, J. A., Mikos, A. G., & Barry, M. A. (2004). Biomaterials, 25, 5809–5819. Bobacka, J., Ivaska, A., & Lewenstam, A. (2008). Chemical Reviews, 108, 329–351. Brandao-Burch, A., Utting, J. C., Orriss, I. R., & Arnett, T. R. (2005). Calcified Tissue International, 77, 167–174. Brown, D. A., MacLellan, W. R., Wu, B. M., & Beygui, R. E. (2007). Annals of Biomedical Engineering, 35, 1885–1897. Brown, G. C. (1992). Biochemical Journal, 284, 1–13. Cartmell, S. (2009). Journal of Pharmaceutical Sciences, 98, 430–441. Chen, T., Zhou, Y., & Tan, W.-S. (2009). Cell Biology and Toxicology, 25, 573–586. Collins, P. C., Nielsen, L. K., Papoutsakis, S. D. P. A. T., & Miller, W. M. (1998). Biotechnology Progress, 14, 466–472. Cordat, E. & Casey, J. R. (2009). Biochemical Journal, 417, 423–439. Csete, M. (2005). Annals of the New York Academy of Sciences, 1049, 1–8. Curtis, A. (2001). eCells & Materials Journal, 1, 59–65. Dalby, M. J., Gadegaard, N., Tare, R., Andar, A., Riehle, M. O., Herzyk, P., Wilkinson, C. D., & Oreffo, R. O. (2007). Nature Materials, 6, 997–1003. Despopoulos A, S. S., ed (2003). Color atlas of physiology. New York: Thieme. deZengotita, V. M., Kimura, R., & Miller, W. M. (1998). Cytotechnology, 28, 213–227. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 82 Discher, D. E., Janmey, P., & li Wang, Y. (2005). Science, 310, 1139–1143. Dubach, J. M., Harjes, D. I., & Clark, H. A. (2007). Journal of the American Chemical Society, 129, 8418–8419. Duran, C., Thompson, C. H., Xiao, Q., & Hartzell, H. C. (2010). Annual Review of Physiology, 72, 95–121. Dvorak, M. M., Siddiqua, A., Ward, D. T., Carter, D. H., Dallas, S. L., Nemeth, E. F., & Riccardi, D. (2004). PNAS, 101 (14), 5140–5145. Eklou-Kalonji, E., Denis, I., Lieberherr, M., & Pointillart, A. (1998). Cell and Tissue Research, 292, 163–171. Engel, E., del Valle, S., Aparicio, C., Altankov, G., Asin, L., Planell, J. A., & Ginebra, M. P. (2008). Tissue Engineering, 14, 1341–1351. Frick, K. & Bushinsky, D. (1998). American Journal of Physiology, 275 (5), F840–F847. Frick, K., Jiang, L., & Bushinsky, D. (1997). American Journal of Physiology, 272 (5), C1450–1456. Friess, W. (1998). European Journal of Pharamceutics and Biopharmaceutics, 45 (2), 113–136. Gaitain, S., Escribano, S., Sancho, P., Cuenllas, E., & Tejero, C. (1997). British Journal of Haematology, 96, 559–565. Gao, F. G., Jeevarajan, A. S., & Anderson, M. M. (2004). Biotechnology and Bioengineering, 86 (4), 425–433. Ge, X., Kostov, Y., & Rao, G. (2003). Biosensors and Bioelectronics, 18, 857–865. Ge, X., Kostov, Y., & Rao, G. (2005). Biotechnology and Bioengineering, 89 (3), 329–334. Generelli, S., Jacquemart, R., de Rooij, N. F., Jolicoeur, M., Koudelka-Hep, M., & Guenat, O. T. (2008). Lab on a Chip, 8(7), 1210–1215. Genzel, Y., Ritter, J. B., Konig, S., Alt, R., & Reichl, U. (2005). Biotechnology Progress, 21, 58–69. Ges, I. A., Ivanov, B. L., Werdich, A. A., & Baudenbacher, F. J. (2007). Biosensors and Bioelectronics, 22, 1303–1310. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 83 Ginebra, M., Traykova, T., & Planell, J. (2006). Journal of Controlled Release, 113, 102–110. Glab, S., Hulanicki, A., Edwall, G., & Ingman, F. (1989). Critical Review in Analytical Chemistry, 21 (1), 29–47. Gladden, L. (2004). The Journal of Physiology, 558 (1), 5–30. Goudar, C. T., Matanguihan, R., Long, E., Cruz, C., Zhang, C., Piret, J. M., & Konstantinov, K. B. (2007). Biotechnology and Bioengineering, 96 (6), 1107–1117. Guenat, O. T., Generelli, S., de Rooij, N. F., 2, M. K.-H., Berthiaume, F., & Yarmush, M. L. (2006). Analytical Chemistry, 78 (21), 7453–7460. Gupta, R. & Kumar, A. (2008). Biomedical Materials, 3(3), 1–15. Gustavsson, J., Altankov, G., Errachid, A., Samitier, J., Planell, J. A., & Engel, E. (2008). Journal of Materials Science - Materials in Medicine, 19 (4), 1839–1850. Hafner, F. (2000). Biosensors & Bioelectronics, 15, 149–158. Hahn, C. (1980). Journal of Physics E: Scientific Instruments, 13, 470–482. Hendrikse, J., Olthuis, W., & Bergveld, P. (1998). Sensors and Actuators B, 53, 97–103. Hisamoto, H. & Suzuki, K. (1999). Trends in analytical chemistry, 18 (8), 513–524. Hofer, A. M. (2005). Journal of Cell Science, 118 (5), 855–862. Hopewell, B. & Urban, J. P. (2003). Biorheology, 40, 73–77. Hornig, S., Biskup, C., Grafe, A., Wotschadlo, J., Liebert, T., Mohr, G. J., & Heinze, T. (2008). Soft Matter, 4, 1169–1172. Itota, T., Carrick, T. E., Rusby, S., Al-Naimi, O. T., Yoshiyama, M., & McCabe, J. F. (2004). Journal of Dentistry, 32, 117–122. Janssen, F., Hofland, I., van Oorschot, A., Oostra, J., Peters, H., & van Blitterswijk, C. (2006a). Journal of Biomedical Materials Research Part A, 79A (2), 338–348. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 84 Janssen, F. W., Oostra, J., van Oorschot, A., & van Blitterswijk, C. A. (2006b). Biomaterials, 27, 315–323. Jiang, P.-J., Patel, S., Gbureck, U., Caley, R., & Grover, L. M. (2010). Journal of Biomedical Materials Research Part B: Applied Biomaterials, 93 (1), 51–58. Johnson, R. D. & Bachas, L. G. (2003). Analytical and bioanalytical chemistry, 376, 328–341. Karp, F. B., Bernotski, N. A., Valdes, T. I., Bhringer, K. F., & Ratner, B. D. (2008). IEEE Sensors Journal, 8, 104–112. Kazzaz, J. A., Xu, J., Palaia, T. A., Mantell, L., Fein, A. M., & Horowitz, S. (1996). The Journal of Biological Chemistry, 271 (25), 15182–15186. Kellner, K., Liebsch, G., Klimant, I., Wolfbeis, O. S., Blunk, T., B.Schulz, M., & Gpferic, A. (2002). Biotechnology and Bioengineering, 80, 73–83. Kim, D.-H., Viventi, J., Amsden, J. J., Xiao, J., Vigeland, L., Kim, Y.- S., Blanco, J. A., Panilaitis, B., Frechette, E. S., Contreras, D., Kaplan, D. L., Omenetto, F. G., Huang, Y., Hwang, K.-C., Zakin, M. R., Litt, B., & Rogers, J. A. (2010). Nature Materials, 9, 511–517. Kissinger, P. T. (2005). Biosensors and Bioelectronics, 20, 2512–2516. Koivusalo, M., Kapus, A., & Grinstein, S. (2009). The Journal of Biological Chemistry, 284 (11), 6595–6599. Komarova, S. V., Ataullakhanov, F. I., & Globus, R. K. (2000). American Journal of Physiology - Cell Physiology, 279, C1220–C1229. Koschwanez, H. E. & Reichert, W. M. (2007). Biomaterials, 28, 3687–3703. Kreft, O., Javier, A. M., Sukhorukov, G. B., & Parak, W. J. (2007). Journal of Materials Chemistry, 17, 4471–4476. Kurzweil, P. (2009). Sensors, 9, 4955–4985. Lee, J.-H., Lim, T.-S., Seo, Y., Bishop, P. L., & Papautsky, I. (2007). Sensors and Actuators B, 128, 179–185. Lee, S.-H. & Shin, H. (2007). Advanced Drug Delivery Reviews, 59, 339–359. Lee, Y.-E. K., Smith, R., & Kopelman, R. (2009). Annual Reviews Analytical Chemistry, 2, 57–76. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 85 Lehmann, M. & Baumann, W. (2005). Experimental Cell Research, 305, 374–382. Li, C. M., Dong, H., Cao, X., Luong, J. H., & Zhang, X. (2007). Current Medicinal Chemistry, 14, 937–951. Lim, M., Ye, H., Drakakis, E. M., Yue, X., Cass, A. E., Panoskaltsis, N., & Mantalaris, A. (2008). Biochemical Engineering Journal, 40, 1–7. Lim, M., Ye, H., Panoskaltsis, N., Drakakis, E. M., Yue, X., Cass, A. E., Radomska, A., & Mantalaris, A. (2007). Biotechnology Advances, 25, 353–368. Lin, K.-I., Chattopadhyay, N., Bai, M., Alvarez, R., Dang, C. V., Baraban, J. M., Brown, E. M., & Ratan, R. R. (1998). Biochemical and Biophysical Research Communications, 249, 325–331. Lin, Z., Cherng-Wen, T., Roy, P., & Trau, D. (2009). Lab on a Chip, 9, 257–262. Liu, J., Barradas, A., Fernandes, H., Janssen, F., Papenburg, B., Stamatialis, D., Martens, A., van Blitterswijk, C., & de Boer, J. (2010). Tissue Engineering: Part C, 16 (3), 479–485. Logeart-Avramoglou, D., Oudina, K., Bourguignon, M., Delpierre, L., Nicola, M.-A., Bensidhoum, M., Arnaud, E., & Petite, H. (2010). Tissue Engineering: Part C, 16 (3), 447–458. Lorget, F., Kamel, S., Mentaverri, R., Wattel, A., Naassila, M., Maamer, M., & Brazier, M. (2000). Biochemical and Biophysical Research Communications, 268, 899–903. Luo, Y. & Prestwich, G. D. (2001). Expert Opinion on Therapeutic Patents, 11 (9), 1395–1410. Luque, G. L., Ferreyra, N. F., & Rivas, G. A. (2007). Talanta, 71, 1282– 1287. Maharbiz, M. M., Holtz, W. J., Howe, R. T., & Keasling, J. D. (2004). Biotechnology and Bioengineering, 85 (4), 376–381. Malda, J., Klein, T. J., & Upton, Z. (2007). Tissue Engineering, 13 (9), 2153–2162. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 86 Marazuela, M. D. & Moreno-Bondi, M. C. (2002). Analytical and bioanalytical chemistry, 372, 664–682. Maria, C. A. D., Bogoyevitch, M. A., McKitrick, D. J., Arnolda, L. F., Hool, L. C., & Arthur, P. G. (2009). Journal of Molecular and Cellular Cardiology, 47, 49–56. Marxer, S. M. & Schoenfisch, M. H. (2005). Analytical Chemistry, 77, 848–853. Mason, C. & Hoare, M. (2007). Tissue Engineering, 13 (2), 301–311. Mather, M. M., Morgan, S. P., & Crowe, J. A. (2007). Regenerative Medicine, 2(2), 145–160. Medlock, K., Harmer, H., Worsley, G., Horgan, A., & Pritchard, J. (2007). Analytical and bioanalytical chemistry, 389, 1533–1539. Mishra, A. & Starly, B. (2009). Microfluidics and Nanofluidics, 6, 373–381. Nakade, O., Takahashi, K., Takuma, T., Aoki, T., & Kaku, T. (2001). Journal of Bone and Mineral Metabolism, 19, 13–19. Obradovic, B., Carrier, R. L., Vunjak-Novakovic, G., & Freed, L. E. (1999). Biotechnology and Bioengineering, 63 (2), 197–205. O’Hare, D., Parker, K. H., & Winlove, C. P. (2006). Medical Engineering & Physics, 28, 982–988. ONeill, C. A. & Galasko, C. S. B. (2000). Calcified Tissue International, 67, 53–59. Orlowski, J. & Grinstein, S. (1997). The Journal of Biological Chemistry, 272, 22373pp. Patel, S. D., Papoutsakis, E. T., Winter, J. N., & Miller, W. M. (2000). Biotechnology Progress, 16, 885–892. Pedersen, P. L. (1994). Current Biology, 4(12), 1138–1141. Peng, C. & Palsson, B. (1996). Annals of Biomedical Engineering, 24 (3), 373–381. Poghossian, A., Ingebrandt, S., Offenhusser, A., & Schning, M. (2009). Seminars in Cell & Developmental Biology, 20, 41–48. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 87 Poole, R. & Halestrap, A. (1993). American Journal of Physiology, 264 (4 Pt 1), C761–C782. Porter, J. R., Ruckh, T. T., & Popat, K. C. (2009). Biotechnology Progress, 25 (6), 1539–1560. Qiu, K., Zhao, X. J., Wan, C. X., Zhao, C. S., & Chen, Y. W. (2006). Biomaterials, 27, 1277–1286. Radomska, A., Singhal, S., Ye, H., Lim, M., Mantalaris, A., Yue, X., Drakakis, E. M., Toumazou, C., & Cass, A. E. (2008). Biosensors and Bioelectronics, 24, 435–441. Rolfe, P. (2006). Measurement Science and Technology, 17, 578–583. Romani, A. (2007). Archives of Biochemistry and Biophysics, 458, 90–102. Romero, M. R., Ahumada, F., Garay, F., & Baruzzi, A. M. (2010). Analytical Chemistry, 82, 5568–5572. Ruffieux, P.-A., von Stockar, U., & Marison, I. W. (1998). Journal of Biotechnology, 63, 85–95. Schneider, M., Marison, I. W., & von Stockar, U. (1996). Journal of Biotechnology, 46, 161–185. Schop, D., Janssen, F. W., van Rijn, L. D., Fernandes, H., Bloem, R. M., de Bruijn, J. D., & van Dijkhuizen-Radersma, R. (2009). Tissue Engineering Part A, 15, 1877–1885. Starly, B. & Choubey, A. (2007). Annals of Biomedical Engineering, 36 (1), 30–40. Stasio, E. D. (2004). Biophysical Chemistry, 112, 245–252. Sterling, D. & Casey, J. R. (2002). Biochemistry and Cell Biology, 80, 483–497. Storrie, H. & Stupp, S. I. (2005). Biomaterials, 26, 5492–5499. Sukhorukov, G. B., Rogach, A. L., Garstka, M., Springer, S., Parak, W. J., Munoz-Javier, A., Kreft, O., Skirtach, A. G., Susha, A. S., Ramaye, Y., Palankar, R., & Winterhalter, M. (2007). Small, 3(6), 944–955. Suzuki, H., Arakawa, H., Sasaki, S., & Karube, I. (1999). Analytical Chemistry, 71, 1737–1743. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 88 Thomas, P. C., Halter, M., Tona, A., Raghavan, S. R., Plant, A. L., & Forry, S. P. (2009). Analytical Chemistry, 81, 9239–9246. Tsao, Y.-S., Cardoso, A., Condon, R., Voloch, M., Lio, P., Lagos, J., Kearns, B., & Liu, Z. (2005). Journal of Biotechnology, 118, 316–327. Utting, J., Robins, S., Brandao-Burch, A., Orrissa, I., Behar, J., & Arnett, T. (2006). Experimental Cell Research, 312, 1693–1702. Vogler, E. A. (1998). Advances in Colloid and Interface Science, 74, 69–117. Vojinovic, V., Cabral, J., & Fonseca, L. (2006). Sensors and Actuators B, 114, 1083–1091. Vunjak-Novakovic, G. & Kaplan, D. L. (2006). Tissue Engineering, 12 (12), 3261–3263. Walker, G. M. (1994). Critical Reviews in Biotechnology, 14 (4), 311–354. Wang, J. (2001). Electroanalysis, 13 (12), 983–988. Wang, J. (2008). Chemical Reviews, 108, 814–825. Wang, M., Yao, S., & Madou, M. (2002). Sensors and Actuators B, 81, 313–315. Wiley, C. & Beeson, C. (2002). Analytical Biochemistry, 304, 139–146. Wilson, G. S. & Gifford, R. (2005). Biosensors and Bioelectronics, 20, 2388–2403. Wisniewski, N. & Reichert, M. (2000). Colloids and Surfaces B: Biointerfaces, 18, 197–219. Woodruff, M. A. & Hutmacher, D. W. (2010). Progress in Polymer Science, Article in press, Corrected Proof. Wu, M.-H., Lin, J.-L., Wang, J., Cui, Z., & Cui, Z. (2009). Biomed Microdevices, 11, 265–273. Yang, L. & Zhang, E. (2009). Materials Science and Engineering C, 29, 1691–1696. Yang, Y. & Balcarcel, R. R. (2004). BioTechniques, 36, 286–295. Yu, B., Wang, C., Ju, Y. M., West, L., Harmon, J., Moussy, Y., & Moussy, F. (2008). Biosensors and Bioelectronics, 23, 1278–1284. Chapter 2 - The Role and Use of Sensors in Tissue Engineering 89 Yun, Y., Dong, Z., Lee, N., Liu, Y., Xue, D., Guo, X., Kuhlmann, J., Doepke, A., Halsall, H. B., Heineman, W., Sundaramurthy, S., Schulz, M. J., Yin, Z., Shanov, V., Hurd, D., Nagy, P., Li, W., & Fox, C. (2009). Materials Today, 12 (10), 22–32. Zhang, F., Ali, Z., Amin, F., Feltz, A., Oheim, M., & Parak, W. J. (2010). ChemPhysChem, 11, 730–735. Zhao, P. & Cai, W.-J. (1997). Analytical Chemistry, 69, 5052–5058. Chapter 3. The Osteoblastic Ionic Extracellular Environment 96 Figure 3.3: The fundamental factors for mineralisation to occur. Image adapted from Allori et al. (2008a). been proposed that cellular mechanisms gradually enrich the phosphate-rich vesicles with calcium (Mahamid et al., 2011). After being released from the cells, and once the crystals have reached a certain size, the vesicle membrane is ruptured and the crystals grow out into the extracellular matrix. It has been shown that these MVs origin from microvilli from the apical plasma membrane in hypertrophic chondrocytes (Hale & Wuthier, 1987), as well as in osteoblasts (Thouverey et al., 2009). The initial calcium salts to be deposited in the ECM are normally not hydroxyapatite crystals but rather amorphous compounds (noncrystalline), typically a mixture of CaHPO4x 2 H2O, Ca3(PO4)2x 3 H2O, and others (Guyton & Hall, 2005, Ch.79). Then by a process of substitution and addition of atoms, or reabsorbtion and reprecipitation, these salts are converted into hydroxyapatite crystals over a period of weeks or months. However, a few per cent remain permanently in the amorphous form which can be rapidly absorbed when there is need for extra calcium in the extracellular fluid. After the onset of mineralisation, osteoblasts continue to express bonespecific proteins. One such protein is osteocalcin (OCN) which is characterised by its γ-carboxyglutamic acid residues, and also known as bone Gla protein. OCN is both released into circulation, and incorporated into the bone matrix. In presence of calcium, the γ-carboxyglutamic acid residues allow specific conformational changes in the protein, which in turn promotes osteocalcin to bind to bone minerals with consequent accumulation in the matrix (Chenu et al., 1994; Fassina et al., 2005). Its exact function is un- Chapter 3. The Osteoblastic Ionic Extracellular Environment 97 known at present, and although the OCN-gene only has been identified in osteoblasts, the protein is not an absolute requirement for mineralisation, but should rather be considered as a late marker of mature osteoblasts. In vitro experiments however show that osteocalcin has chemotactic activity for a number of cells, including monocytes which are known to be related to osteoclast precursor cells, and it has been proposed that OCN has a potential role in the induction of bone resporption (Mundy & Poser, 1983; Chenu et al., 1994). Another bone-related protein synthesised by mature cells of the osteoblastic lineage is osteonectin (ON), or SPARC. This protein is 10000 times more abundant in bone tissue than in other connective tissue. Osteonectin was given its name (osteonectin = bone connector) due to its strong affinity for both collagen, hydroxyapatite, and calcium (Kelm et al., 1994). Its specific function is still unknown, and it has been shown to act both as a promotor and inhibitor of mineralisation (Romberg et al., 1986; Boskey, 1996). In Table 3.1 are summarised the main components of mammalian bone tissue. Chapter 3. The Osteoblastic Ionic Extracellular Environment 98 CELLS Osteoblasts Synthesis and regulation of bone ECM deposition and mineralisation. Osteocytes Calcification of the osteoid matrix. Blood-calcium homeostasis. Osteoclasts Bone resorption. ORGANIC MATERIAL Fibrinous Collagen type I Provide tensile strength proteins Framework for skeletal structure ≈90% Non-fibrinous Bone sialoprotein (BSP) Adhesion; proteins Constituent of cement line ≈10% Osteopontin (OPN) Adhesion Constituent of cement line Involved in bone remodelling Osteocalcin (OCN) Inhibitor of nucleation Involved in bone remodeling Osteonectin (ON) Enhance mineral deposition Binds Ca2+ and collagen Fibronectin (FN) Adhesion, mechanochemical signaling Matrix Gla Protein (MGP) Inhibitor of nucleation Proteoglycans Chondroitin sulfate Compressive strength Hyaluronic acid Dermatan sulfate Byglican Enzymes Alkaline phosphatase (ALP) ↑Pi Regulate inhibitor Matrix metalloproteinases (MMPs) Growth factors Tumour Growth Factor-β(TGF-β) Bone morphogenic protein (BMP) Cytokines INORGANIC MATERIAL Mineral Hydroxyapatite (HA) Mechanical resistance HA-precursors Table 3.1: The main components of bone tissue. Partly adapted from Salgado et al. (2004). Chapter 3. The Osteoblastic Ionic Extracellular Environment 99 3.1.3 In Vitro Osteoblast-like Model Systems Osteoblast-like culture model systems fall into two main categories: primary cell cultures, or cloned cell lines (Table 3.2). Primary osteoblast models can be derived from different sites (typically trabecular bone or calvariae) and from different species, among which the most common ones include rat and human (Table 3.2). A major part of in vitro cell culutres are however performed with cloned cell lines. On one hand, this may be considered as a step away from the in vivo state, but on the ohter hand, commercially available and established cell lines derived from normal or malignant cell cultures often exhibit a similar pattern and time frame of gene expression as compared to the fully mature osteoblastic phenotype. In addition, the heterogenous character and maturity of primary cells is avoided, which improves the phenotypic stability and reproducibility of the results. However, when working with cell lines one should be aware that marked differences between cell lines of the same phenotype may exist. Osteoblasts can also be obtained by inducing differention of mesenchymal stem cells (MSCs). MSCs reside as adult stem cells at various sites in the body, for example in adipose tissue, blood, dermis, trabecular bone, and periosteum. The most common host and source of MSCs is however the bone marrow (Tuan et al., 2002). As adult stem cells, MSCs have the potential to differentiate into a great variety of phenotypes, including chondrocytes, adipocytes, fibroblasts, and osteoblasts, all being of mesodermal lineage (Figure 3.4). Interestingly, transdifferentiation into phenotypes of endoderm and enctoderm lineages has also been achieved. (Jiang et al., 2002). Epithelial cell Neuron Bone marrowBone MSC Self-renewal Ectoderm Fat cell Bone cell Muscle cell Gut epithelial cell Mesoderm Cartilage cell Connective stromal cell Lung cell Endoderm Osteoblast Figure 3.4: Lineage potential of mesenchymal stem cells. Image is adapted from Uccelli et al. (2008). Chapter 3. The Osteoblastic Ionic Extracellular Environment 100 Primary culture models Human Yamanouchi et al. (1997); ONeill & Galasko (2000) Beloti & Rosa (2005); Jones et al. (2007) Pig Eklou-Kalonji et al. (1998) Rat Chang et al. (2000); Dvorak et al. (2004) Hempel et al. (2004); Valerioa et al. (2004) Cell lines Human HOS TE-85 Clover & Gowen (1994); Habel & Glaser (1998) MG-63 Jukkola et al. (1993); Takamizawa et al. (2004) Gregory et al. (2004); Takagishi et al. (2006) Kim et al. (2007) OHS-4 Fournier & Price (1991) SAOS-2 Hale et al. (2000); Fassina et al. (2005) Ayers et al. (2006); Orimo & Shimada (2008) Thouverey et al. (2009) U2-OS Abed & Moreau (2007) Mouse MC3T3-E1 Fratzl-Zelman et al. (1998); Balint et al. (2001) Beck (2003); Nakano et al. (2007) Xiao et al. (2007); Addison et al. (2007) Rat ROS 17/2.8 Mikami et al. (2007) UMR 106.06 Stanford et al. (1995); Hale et al. (2000) Table 3.2: Examples of osteoblast-like cells (of different origins) that are commonly used to study cell differention (see indicated references). It is important to be aware of that slight differences in cellular behaviour may exist between these models. 3.1.3.1 Osteogenic medium Both primary cells and established cell lines usually require specific osteogenic growth medium to be able to execute the full osteoblast differentiation program. This is especially true for MSCs which are severely dependent on the addition of certain osteogenic factors in order to develop towards the osteoblastic lineage (Tuan et al., 2002). Among the most common osteogenic supplements are organic phosphate, ascorbic acid (AA), and dexamethasone. Other factors with demonstrated effect on osteoblast function include bone morphogenetic protein (BMP) and vitamin D (Arnett & Henderson, 1998). Ascorbic acid is a water-soluble vitamin that has been observed to promote collagen maturation, influence cell growth and stimulate differentiation of osteoblasts and MSCs (Takamizawa et al., 2004; Choi et al., 2008). Being Chapter 3. The Osteoblastic Ionic Extracellular Environment 101 a cofactor for the hydroxylation of proline and lysine residues of collagen, AA excerts a direct effect on the deposited matrix rather than affecting osteoblastic gene expression. The use of dexamethasone, a synthetic glucocorticoid, as a supplement to induce osteoblast differentiation is somewhat contradictory as this agent is known to cause severe bone loss (osteoporosis) in vivo (Reid, 1997). However in vitro, dexamethasone has been demonstrated to possess both positive and negative effect on osteoblast differentiation and mineralisation, depending on the level of cell maturity or cell density. For example, dexamethasone promotes phenotypic markers of osteoblast differentiation, such as alkaline phosphatase (ALP), osteopontin (OPN), osteocalcin (OC), and bone sialoprotein (BSP) (Yamanouchi et al., 1997). Also, Mikami et al. (2007) demonstrated that Runx2 transcriptional activity was increased in osteoblasts upon treatment with dexamethasone, and they speculated that this may be followed by activation of specific osteoblast genes. While common rich growth media (such as DMEM, α-MEM, RPMI, Mc- Coy’s, etc) containing serum provide conditions for osteoblasts in long-term cultures to proliferate and progressively develop collagen matrix and express bone-specific proteins, in vitro mineralisation does not occur spontaneously. The failing component in this case is lack of phosphate (recall Figure 3.3). Therefore, osteogenic medium is elaborated to contain an additional source of phosphate, typically 5-10 mM of β-glycerophosphate (β-GP), which is hydrolysed by alkaline phosphatase into inorganic phosphate (Pi) and glycerol. β−glycerophosphate + H2O→Pi+ glycerol (3.1) Supplement Abbreviation Working concentration β-glycerophosphate β-GP 5-10 mM Ascorbic acid AA 50 µg/mL Dexamethasone Dex 1-100 nM Table 3.3: Typical composition of osteogenic medium (Arnett & Henderson, 1998) Chapter 3. The Osteoblastic Ionic Extracellular Environment 102 3.1.3.2 Assesment of in vitro mineralisaton Mineralisation, the final stage of osteoblastic differentiation, can be assessed by a number of means. The fastest and most easy way is to evaluate calcium deposition in the extracellular matrix. An alternative would be to assay inorganic phosphate deposition but this approach is less adopted. Evaluation of calcium distribution in the ECM is routinely done by methods such as von Kossa staining (Fratzl-Zelman et al., 1998), Alizarin red S (ARS) incorporation (Stanford et al., 1995), o-cresolphthalein complexone method (Fujita et al., 2001; Takagishi et al., 2006), or through fluorescent calcein binding (Hale et al., 2000). ARS staining is considered particularly versatile in the sense that the dye can be extracted from the stained monolayer and then quantified (Gregory et al., 2004). Also fluorescence analyses of calcein bound to calcium phosphate can allow for direct quantitation of extracellular matrix mineral content in monolayer cultures of bone-forming cells. Precise stoichiometric evaluation of calcium salts is however only achieved by more sophisticated techniques, e.g. Fourier-transform infrared (FTIR) spectroscopy, atomic adsorption spectroscopy (Chang et al., 2000), energydispersive X-ray (EDX) spectroscopy, and micro-Raman spectroscopy (Gentleman et al., 2009). Beside providing detailed information on chemical composition, the latter technique is also non-destructive and therefore can be applied to living cells and tissue. 3.1.4 The Ionic Extracellular Environment From Section 3.1.2 it follows that the composition of the ionic extracellular environment (IEE) is important for osteoblasts to fully execute its differentiation program. Among the many ions that constitute the IEE, calcium and phosphate undoubtedly accentuate as the most fundamental ones in the bone formation process. In blood serum there is about 2.5 mM of total calcium. This amount is distributed in three forms: ionised,protein bound, and complexed when united to citrate, phosphate, bicarbonate, or other anions. The ionised calcium is diffusible, the citrate calcium is not ionised but is also diffusible, while the calcium chelated to proteins (mainly albumin and globulin) is neither ionised nor diffusible (Figure 3.5). In the end, the free ionised calcium of blood serum is maintained at a very constant level (≈1.4 mM). The calcium-protein interaction is complex, depending on pH, temperature, protein structure, as well as the concentration of phosphate, carbonate, citrate, etc. Regarding pH, the calcium binding of proteins (albumin) in- Chapter 3. The Osteoblastic Ionic Extracellular Environment 103 creases almost linearly with incresing pH up to about pH 8 as the carboxyl groups of proteins are being ionised and thus the electrostatic interaction is increased (Neuman & Neuman, 1958). Total Calcium Nondiffusible Diffusible 2,5 mM 0,82 mM 1,63 mM Globulin Albumin 0,17 mM 0,65 mM Ionized calcium Complexes 1,33 mM HCO 0,16 mM PO4 - 3 - 0,06 mM 0,30 mM Citrate 0,07 mM Others ? Figure 3.5: Calcium distribution in serum. Adapted from (Neuman & Neuman, 1958). The distribution of inorganic phosphate in serum is less complicated than that of calcium. Protein-binding is negligible and except for a small fraction complexed by divalent ions (≈12%) all the inorganic phosphate is ionised (Behari, 2009, p.26). Phosphorus exists in aqueous solutions in different forms of orthophosphate ( 3.2), and their distribution is pH dependant. H3PO4↔H2PO− 4↔HPO2− 4↔PO3− 4(3.2) While H2PO– 4prevails in acidic conditions and HPO42−is more predominant under basic conditions, at a neutral pH both forms are present. In human serum, the relative distribution of orthophosphate is estimated to be 20% H2PO– 4and 80% HPO42−, but unlike serum calcium, it is difficult to precisely specify a normal value for the total inorganic phosphate of human serum as the phosphate concentration varies a great deal with age and metabolic state particularly. Normal variation is from 1-2 mmol/l. Chapter 3. The Osteoblastic Ionic Extracellular Environment 104 3.2 Objective and Strategy The objective of the work presented in this chapter was to understand how the in vitro ionic extracellular environment (IEE) might be influenced by osteoblast-like cell activity. For that purpose, it was used culture system comprised of three different osteoblast-like cell models: human osteosarcoma cell lines SAOS-2 and MG63 respectively, and mesenchymal stem cells derived from rat bone marrow. Each cell model was maintained in its specific cell culture medium, and the extracellular concentrations of Na+, K+, Ca2+, and total inorganic phosphorus (Pi), as well as pH, were evaluated every second day during a period of three weeks. The steps taken to characterise osteoblastic influence on the IEE were the following: 1. characterisation of the initial ionic composition of the three cell culture media used in the study, 2. evaluation of how different storage and preparation conditions influenced the ionic stability of the cell culture medium with time, 3. characterisation of the IEE of osteoblast cell cultures throughout a time range that covered cellular proliferation and differentiation events. Chapter 3. The Osteoblastic Ionic Extracellular Environment 105 3.3 Materials and Methods 3.3.1 Preparation of Osteoblastic Culture Media Three different basal cell culture media commonly used for osteoblast cultures have been used in this study: (1) Dulbecco’s Modified Eagle’s Medium (DMEM; Invitrogen 11960), (2) advanced Dulbecco’s Modified Eagle’s medium (aDMEM; Invitrogen 12491), and (3) McCoy’s 5A Modified Medium (McCoy; Sigma, M8403). These media present differences in their chemical composition with respect to amino acid, vitamin, and glucose content. And as shown in Table 3.4, also the inorganic salt composition of McCoy and the both DMEMs is different. McCoy’s 5A DMEM & aDMEM [g/L] [g/L] CaCl2·2 H2O 0.132 CaCl2(anhydrous) 0.2 MgSO4(anhydrous) 0.098 MgSO4(anhydrous) 0.098 KCl 0.4 KCl 0.4 NaHCO32.2 NaHCO33.7 NaCl 6.5 NaCl 6.4 NaH2PO4(anhydrous) 0.504 NaH2PO4·H2O 0.125 Table 3.4: Inorganic salts content of the three different basal media, McCoy’s and DMEMs respectively. For the record, the growth media also contain organic material, e.g. amino acids, vitamins, and glucose. For optimal cell growth each medium was further supplemented with additional components according to Table 3.5; l-glutamine (Invitrogen, 25030- 024), penicillin and streptomycin (Invitrogen, 15140-122), sodium pyruvate (Invitrogen, 11360-039), and foetal bovine serum (FBS) (Invitrogen, 10270- 106). This preparation was based on previous experience and protocols derived in our own laboratory. For induction of bone differentiation, osteogenic medium was prepared by supplementing abovementioned cell culture media with 50µg/mL ascorbic acid (AA; Sigma, A4034), 10 mM β-glycerophosphate (β-GP; Sigma, G9891), and 10−8M dexamethasone (Dex; Sigma, D8893). AA was dissolved in ultrapure water (5.0 mg/mL), Dex in 2% ethanol (5.2 ×10−5M), and β-GP in Phosphate Buffered Saline (PBS) (1 M). The solutions were filtered (0.22 µm) before being diluted into the cell culture medium to yield the desired concentrations (see Table 3.5). Chapter 3. The Osteoblastic Ionic Extracellular Environment 112 medium. At day eleven cell layers were fixed and stained with Alizarin Red according to Section 3.3.8. 3.3.11 Cellular Influence on the IEE at Different Total Volumes To estimate the relative influence of osteoblast activity on the cell culture environment, this was measured varying the total volume of the environment. In detail, SAOS-2 cells were grown on porous polycarbonate membranes attached to specific inserts (Nunc, 137052). The inserts (10 mm diameter, pore size 0.4 µm) were initially located in separate wells of a standard 24- well culture plate, and were cared for as was described earlier in Section 3.3.7. However, at day 20 inserts were moved from the culture plate and placed in plastic tubes (15 mL). Each tube was filled with a different volume of osteogenic media (3-13 ml). The tap of the tube was loosely attached to allow for gas exchange, and cells were incubated for another 20 hours before 0.50 mL of supernatant was harvested and assayed according to Sections 3.3.2 and 3.3.3. Chapter 3. The Osteoblastic Ionic Extracellular Environment 113 3.4 Results 3.4.1 Ionic Composition of Osteoblastic Culture Media While concentrations of Na+, K+, Ca2+, and pH of cell culture medium were measured using ion-selective electrodes, concentrations of inorganic total phosphorus was evaluated using the colourimetric molybden-ascorbic acid method (Figure 3.7). The calibration curve used to translate absorbance units to concentration of Piis shown in Figure 3.8, and it reveals a linear relationship between the two units in the range 0.01 mM to 1.0 mM. This applied for both PO3− 4and PO2− 4ions. For example, with PO3− 4ions, y= 2.6064[Pi] + 0.0048 with R2= 0.999 and n= 12. The ionic composition of the different cell culture media is presented in Figure 3.9(a-c). Considering only the base media, i.e. without any supplements, it was observed that the ionic composition of DMEM and aDMEM was similar with respect to all investigated ions. The only marked differences were observed for calcium and phosphorus levels which were slightly higher in DMEM than in aDMEM. Much greater variation was however observed between the both DMEMs and McCoy medium. First, concentration of sodium was significantly lower in McCoy medium than in DMEMs (134.3 ±1.1 mM in McCoy while 152.8 ±0.55 mM in DMEM and 152.1 ±0.16 mM in aDMEM). Second, concentration of calcium was more than halfed in McCoy medium compared to DMEM (0.64 ±0.07 mM in McCoy, 1.46 ±0.07 mM mM in DMEM, 1.35 ±0.05 mM in aDMEM). And third, concentration of Piwas more than four times higher in McCoy medium than in DMEMs (4.49 ±0.21 mM in McCoy, while 0.93 ±0.07 mM in DMEM, and 0.82 ± 0.04 mM in aDMEM). Among the different supplements added to cell culture medium it was FBS that induced the major changes in the ionic composition. In summary, upon adding supplements to any of the three base media it was observed an increase in potassium concentration (+19.7% in complete McCoy, +6.5% in complete DMEM, and +11.1% in complete aDMEM), and a strong decrease in total phopshorus (-57.2% for complete McCoy, -40.3% for complete DMEM, and -54.2% for complete aDMEM). For McCoy medium it was also observed how both sodium and calcium concentrations increased when adding FBS (+4.1% and +38.5%, respectively). The same effect was not observed in any of the both DMEM media. Rather the sodium concentration decreased upon addition of FBS, while the concentration of calcium maintained the same. Chapter 3. The Osteoblastic Ionic Extracellular Environment 114 0 mM 10 mM -5 10 mM -3 10 mM -4 I n c r e a s i n g t o t a l P Figure 3.7: Colourimetric method for determination of total phosphorus. [P ] (mM) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.001 0.01 0.1 10 1 i [P ] (mM) i 0.0 0.2 0.4 0.8 0.6 1.0 1.2 Spectroscopic units Figure 3.8: Calibration curve for quantification of total inorganic phosphorus using NaH2PO4(◦) and Na2HPO4(). The data for NaH2PO4is visualised twice: on a linear scale (empty ◦, bottom x-axis) and on a logarithmic scale (filled ◦, upper x-axis). Chapter 3. The Osteoblastic Ionic Extracellular Environment 115 [Na ] (mM) + 170 160 150 140 130 120 [K ] (mM) + 7,0 6,0 5,0 4,0 3,0 2,0 1,0 0,0 aDMEM DMEM McCoy Base medium l-glutamine pen/strep sodium pyruvate FBS Compete medium + + - - - + + - - - + + - - - + + - - - + - - - - Figure 3.9a Figure 3.9: (a) Na+and K+concentration of osteoblastic cell culture media. Chapter 3. The Osteoblastic Ionic Extracellular Environment 116 [Ca ] (mM) 2+ aDMEM DMEM McCoy Base medium l-glutamine pen/strep sodium pyruvate FBS Compete medium + + - - - + + - - - + + - - - + + - - - + - - - - 2,5 2,02,0 1,5 1,0 0,5 0,0 [P ] (mM) i 0,0 1,0 2,0 3,0 4,0 5,0 6,0 Figure 3.9b Figure 3.9: (b) Ca2+ and Piconcentration of osteoblastic cell culture media. Chapter 3. The Osteoblastic Ionic Extracellular Environment 117 pH aDMEM DMEM McCoy Base medium l-glutamine pen/strep sodium pyruvate FBS Compete medium + + - - - + + - - - + + - - - + + - - - + - - - - 6,0 6,5 7,0 7,5 8,0 Figure 3.9c Figure 3.9: (c) pH levels of osteoblastic cell culture media. 3.4.2 Stability of Cell Culture Medium With Time Concentrations of Na+, K+, Ca2+, pH, and inorganic phosphorus of supplemented McCoy’s 5A modified media with or without FBS, and stored either at 4◦C (fridge) or at 37◦C with 5% CO2(cell culture incubator), were repeatedly evaluated during a period of 15 days. The results are presented in Figure 3.10. While medium stored at 37◦C presented slightly higher concentrations of both Na+and K+than medium stored at 4◦C, it was generally observed that both [Na+] and [K+] maintained relatively stable during 15 days, no matter if the medium contained FBS or not. In contrast, concentration of Ca2+ in FBS-containing medium stored at 37◦C decreased gradually after seven days until the end of the experiment. Also when stored at 4◦C FBS influenced [Ca2+] as the standard variation was observed to increase with time beyond day seven. However, in absence of FBS the concentration of calcium was preserved over the full length of experiment in both storage conditions. Chapter 3. The Osteoblastic Ionic Extracellular Environment 118 Moreover, it was observed that storage of medium at physiological temperature and at controlled gas levels stabilised the pH of the medium while medium stored in the fridge was subject to drastic increases in pH. Finally, preparation and storage condition had no important influence on the concentration of total phosphorus. 120 125 130 135 4,0 4,5 5,0 5,5 0,0 0,2 0,4 0,6 0,8 2,0 3,0 4,0 5,0 Time (days) 7,0 7,2 7,4 7,6 7,8 8,0 8,2 4 C 37 C 4 C FBS 37 C FBS o o o o Time (days) 0 2 4 6 8 10 12 14 160 2 4 6 8 10 12 14 16 Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi Pi pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH pH Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Na Figure 3.10: Ionic stability of supplemented McCoy cell culture medium without and with FBS (left and right column, respectively) at 4◦C (black symbols) and 37◦C (white symbols). Concentrations are expressed in units of mM. Chapter 3. The Osteoblastic Ionic Extracellular Environment 119 3.4.3 Ionic Levels of Cell Culture Media During Osteoblast Cultures Cell cultures of MG63, SAOS-2, and rat-derived mesenchymal stem cells (rMSCs) lasted for 21 days and were performed in polystyrene 24-well plates. Cells were maintained in one of the following media: complete DMEM (MG63), advanced DMEM (rMSC), or McCoys 5A media (SAOS-2). The different cell types all attained a similar aspect marked by an extended, triangular shape when adhered to the substrate. MG63 cells appeared slightly larger than both SAOS-2 and rMSC, with the latter cell type being recognised as slightly thinner and more elongated (Figure 3.11). After one day in culture, and from then on every second day, the cell culture medium was exchanged for fresh media. After five days in culture, one set of each cell type was provided osteogenic media instead of normal media. The ionic composition of fresh media was monitored before addition to cells throughout the full length of the experiment (Table 3.7). Upon medium renewal, harvested supernatants were assayed for ionic and total phosphorus concentrations. The data is presented in Figure 3.12(a-e), where the ion concentrations of fresh medium given in Table 3.7 are drawn as dashed lines in red (normal media) or dotted lines in green (osteogenic media). Each one of Figure 3.12(a-e) is organised in three rows and two columns. Rows present data obtained from a different cell phenotype: SAOS-2 (top), MG63 (middle), and rMSC (bottom), respectively. Left columns show data obtained from cells maintained in normal medium throughout the full length of experiments, while right columns contain data obtained from cells maintained in osteogenic medium. Each graph also contains data on ionic concentration of cell culture medium both in presence and in absence of cells. Na+K+Ca2+ pH Pi DMEM 149.0 ±1.18 5.30 ±0.03 1.41 ±0.03 7.42 ±0.03 0.36 ±0.01 DMEM osteogenic 156.9 ±1.02 5.10 ±0.02 1.17 ±0.03 7.49 ±0.03 0.45 ±0.01 aDMEM 151.2 ±2.02 5.73 ±0.10 1.44 ±0.03 7.41 ±0.02 0.46 ±0.01 aDMEM osteogenic 159.8 ±2.54 5.49 ±0.11 1.17 ±0.02 7.49 ±0.03 0.44 ±0.01 McCoy 132.9 ±1.70 5.51 ±0.06 0.77 ±0.04 7.37 ±0.03 2.77 ±0.05 McCoy osteogenic 140.7 ±1.35 5.29 ±0.04 0.64 ±0.03 7.43 ±0.02 2.91 ±0.04 Table 3.7: The ionic levels of each medium type (supplemented and with FBS) was measured before it was provided to cells. The values (expressed in units of mM, except for pH) are given as mean ±standard deviation. For osteogenic media n = 8, otherwise n= 11. Chapter 3. The Osteoblastic Ionic Extracellular Environment 120 MG63 SAOS-2 rMSC Figure 3.11: Representative images of the different osteoblast models used in this study were obtained after one day in culture when cells still had not reached confluency. Scale bar represents 200 µm, and applies to all three images. Chapter 3. The Osteoblastic Ionic Extracellular Environment 121 As is shown in Figure 3.12(a & b), levels of Na+and K+followed similar trends for all three cell types in both normal and osteogenic medium. Incubated media presented increased concentrations of both Na+and K+ compared to initial concentrations of these ions. While the levels of Na+increased about 2-8 mM between measurements, K+levels normally increased 0.1-0.2 mM. These changes represent increases of about 1-2%, and were due to evaporation of medium during the incubation period2. Beside the evaporation effect, the most important observation was that no difference in neither [Na+] nor [K+] was observed between cell culture medium that had been in contact with cells and cell culture medium not in contact with cells, except for MG63 cells maintained in osteogenic DMEM. In contrast to the concentrations of sodium and potsasium, extracellular concentration of calcium in cell culture medium was influenced by osteoblast cultures (Figure 3.12c). After five days in culture and until the end of experiment [Ca2+] decreased in both MG63 and rMSC cultures when maintained in normal medium. The decreases corresponded to about 0.10-0.13 mM (7- 9%) for MG63 cells, and about 0.05-0.10 mM (3-7%) for rMSC cells. This effect was however not observed in SAOS-2 cells maitained in normal McCoy medium. When exposing cells to osteogenic medium, a massive decrease in [Ca2+] of cell culture medium was observed in both SAOS-2 and rMSC cultures, but not in MG63 cultures. At the same time osteogenic medium prepared from any of the three basal media maintained its calcium concentration level constant in absence of cells. The decrease in calcium in presence of SAOS-2 cells was about 0.17 mM (≈24%) between day 5 and day 7, and then increased to about 0.4 mM (≈60%) between day 7-9, 9-11, 11-13, and 13-15. After that, the decrease in [Ca2+] continued, but at a gradually slower rate. More or less the same tendency was observed for rMSC. Initially (i.e. day 5-7), the decrease in [Ca2+] was small, but then it gradually increased and beyond day 9 it reached about 0.75 mM (≈65%). Like [Na+] and [K+], concentrations of total phosphorus remained unchanged in cell cultures maintained in normal medium (Figure 3.12d). However, unlike normal medium without cells, osteogenic medium without cells repeatedly increased its [Pi]. In osteogenic McCoy medium [Pi] raised on average 1.7 mM (≈58%), while in osteogenic DMEM and aDMEM, [Pi] in- 2It was observed that evaporation was higher in wells located at the border of the multi-well plate (data not shown). To compensate for the evaporation effect, samples and controls were always placed symmetrically in the culture plate. Chapter 3. The Osteoblastic Ionic Extracellular Environment 128 3.4.4 Calcium Deposition in Extracellular Matrix Calcium deposition in the extracellular matrix was evaluated by Alizarin Red staining of fixed cell layers. At the end of experiment, i.e. after 21 days in culture, cell layers were stained with Alizarin Red. The amount of incorporated dye was quantitatively evaluated after extraction using CPC. The fold increase of calcium in the extracellular matrix was calculated as the ratio of calcium in cell layers grown in osteogenic medium compared to cells maintained in normal medium. The data is presented in Figure 3.13, and which shows that no calcium deposition occurred in MG63 layers while calcium deposition in SAOS-2 and rMSC cell layers had increased six to seven times. MG63 SAOS-2 rMSC Ca-deposition in ECM using osteogenic medium (Fold increase) 0 1 2 3 4 5 6 7 8 9 10 Figure 3.13: Quantitative calcium deposition in extracellular matrix. Fold increase of calcium in the extracellular matrix using osteogenic medium after 21 days in culture (n=6) Representative images of stained cell layers were taken after nine days in culture are presented in Figure 3.14(a-i). Cells maintained in normal medium did not show any incorporation of Alizarin Red (a-c). However, when using osteogenic medium both SAOS-2 (e,h) and rMSC (f,i) cultures demonstrated large areas of calcium deposition distributed throughout the samples. In contrast, no calcium deposition was detected in MG63 cultures at this time point using osteogenic medium. Chapter 3. The Osteoblastic Ionic Extracellular Environment 129 MG63SAOS-2rMSC a d g b e h c f i Figure 3.14: Alizarin Red staining of calcium in MG63 (top row), SAOS-2 (middle row), and rMSc (bottom row) cell layers after nine days in culture. Left column (a-c) show ARS stained cells maintained in normal medium. Middle (d-f) and right (g-i) columns show cells cultured in osteogenic medium. Scale bars represent 500 µm (white) and 200 µm (black), respectively. Chapter 3. The Osteoblastic Ionic Extracellular Environment 130 a c b d Figure 3.15: (a,c) Phase contrast images of Alizarin Red stained SAOS-2 cell layers cultured for 16 days. (b,d) Epifluorescent images (excitation 510-560 nm, emission >590 nm) of the same locations as in (a) and (c), respectively. Scale bars represent 500 µm (white) or 200 µm (black). To elucidate the role of medium composition on calcium deposition, SAOS- 2 cells and MG63 cells were cultured in either McCoy or DMEM media. As is shown in Figure 3.16, MG63 cells did not induce calcium deposition in any of the two medium compositions. 3.4.5 Ionic signals as function of cell number and media volume Any change of the ionic extracellular environment induced by cells will depend on the experimental conditions, and ultimately on the ratio between amount of cells and the total cell culture volume. In Figure 3.17 it is demonstrated how the induced change on the extracellular ionic environment from SAOS-2 cell activity during 20 hours varied with total cell culture volume. Chapter 3. The Osteoblastic Ionic Extracellular Environment 131 Osteogenic DMEM Osteogenic McCoy SAOS-2MG63 Figure 3.16: Alizarin Red staining of SAOS-2 and MG63 cells grown in both corresponding and non-corresponding medium. Images taken after eleven days in culture. µ ( g protein / mL) 0 100 200 300 400 500 % change 20 60 100 140 ∆pH 0.10 0.20 0.30 Calcium Phosphorus pH Figure 3.17: Double Y-axis. Estimation of SAOS-2 influence on the ionic extracellular environment in osteogenic medium. Chapter 3. The Osteoblastic Ionic Extracellular Environment 132 3.4.6 Alkaline Phosphatase Activity ALP activity of cell cultures was measured after nine days in culture, and at that time point great differences between the three different culture models were observed (Figure 3.18). Highest ALP activity was detected in SAOS-2 cells, while the lowest activity was observed in MG63 cells. The differences in ALP activity between SAOS-2 cells and rMSCs was in the order of about two magnitudes. More or less the same difference was then observed between rMSCs and MG63 cells. In all cell models, osteogenic medium had a positive influence on ALP activity. The effect was more pronounced in MG63 cells (267.6 ±45.3%), than in rMSCs (97.8 ±38.1%) and in SAOS-2 cells (24.8 ±9.7%). MG63 SAOS MSC 0.001 0.01 0.1 1 10 100 Normal Osteogenic * * * Figure 3.18: Alkaline phosphatase activity of MG63, SAOS-2 and rMSC cells grown in normal medium and osteogenic medium. ALP activity was measured after nine days in culture, and is expressed in units of µmol p-nitrophenol min−1 (mg protein)−1 Chapter 3. The Osteoblastic Ionic Extracellular Environment 133 3.5 Discussion For proper cell growth and tissue development, cells depend on receiving the appropriate cues and information from their extracellular environment. Among the many signals actuating in that environment, ions are known to play an important role. They not only contribute to establish the cellular membrane potential, but also function as essential messenger and transporter molecules. Some ions, like calcium and phosphates, may also be fundamental building blocks for tissue development, especially in bone tissue. In this chapter it has been studied how the ionic extracellular environment is influenced by osteoblast-like cellular activity. The reason behind this study was two-folded. First, the desire was to investigate if fluctuations in any of the extracellular inorganic ions could be considered as an indirect parameter of tissue development or cellular activity. Second, since in more complex bone tissue engineering situations one may also anticipate intervention of bioactive scaffold materials that by themselves interact with the ionic extracellular environment (see Chapter 4 and 5), our ambition was first to clarify to what extent osteoblast-like cells alone influence that specific environment. 3.5.1 Cell Culture Media Characterisation Studies of the ionic extracellular environment initially required that appropriate measurement methods were established. Throughout this work ionselective electrodes have been used as a reference method for cation analysis, while the colourimetric molybden-ascorbic acid method was used for phosphate analyses. The usefulness of these two techniques was clearly demonstrated through the cell culture media analysis in Section 3.4.1 where the compositional differences between the three cell culture media used in this study were reproduced; low levels of calcium and high levels of phosphate in McCoy medium, and high levels of calcium but low levels of phosphate in both DMEM and aDMEM. Furthermore, both methods produced consistent and reproducible results, and they could detect fluctuations in physiological ranges. Beside the actual compositional differences between the different cell culture media, the most important findings regarding medium composition was the effect of FBS on the stability of ionic levels, and how to best store the medium throughout long-term experiments. In Section 3.4.2 it was demonstrated that ionic stability, mainly that of calcium, was better preserved in serum-free medium. Furthermore, if stored at 37◦C and with 5% CO2, the pH of the medium was stabilised over a much longer time period. The most likely explanation of the observed calcium instability of FBS-containing medium Chapter 3. The Osteoblastic Ionic Extracellular Environment 134 would be spontaneous precipitation of calcium phosphate induced by alkaline phosphatase present in FBS. This explination is further supported by the fact that the decrease is faster at 37◦C than at 4◦C, where enzymatic activity is lower. One possible way to confirm this would be to centrifuge the medium samples, and analyse or stain the precipitated material for calcium and phosphorus. The results presented in Sections 3.4.1-3.4.2 demonstrate the importance of appropriate preparation and storage of cell culture medium when studying the ionic extracellular environment, and this applies to studies both in absence and presence of cells. 3.5.2 Calcium Deposition Provoked by Osteoblast Activity Using osteogenic factors it was demonstrated that both SAOS-2 cells and rMSCs provoke significant calcium deposition in the extracellular matrix (Section 3.4.4). As opposed to SAOS-2 and rMSCs, but under identical conditions, MG63 cell layers did not show any sign of calcium deposition. The documented observations agree well with previous studies on similar systems and under similar conditions. SAOS-2 cells are relatively stable cells (Hausser & Brenner, 2005) known to undergo the entire osteoblastic differentiation program, including production of a collagenous extracellular matrix (McQuillan et al., 1995), an extremely high ALP enzymatic activity (Rodan et al., 1987; Stinson et al., 1993), as well as spontaneous release of mineralisation-competent matrix vesicles (Thouverey et al., 2009). Furthermore, an extractable bone-inducing agent (BIA) is present in lysates of SAOS-2 (Anderson et al., 1998). As a result, calicum deposition in SAOS- 2 layers has been commonly reported when using osteogenic medium (e.g. Fassina et al. (2005); Thouverey et al. (2009)). In a similar way, many studies have demonstrated the in vitro mineralisation potential of adult, marrowderived mesenchymal stem cells (Ohgushi et al., 1996; Nauman et al., 2003; Shimko et al., 2004; Maeda et al., 2007). Recalling Figure 3.3, the absent calcium deposition in MG63 cell layers could be due to inappropriate culture medium ionic composition, inadequate alkaline phosphatase activity, defect collagen production, or a combination of each of these factors. In this study, any influence of medium composition was discarded as it was observed that MG63 cells failed to deposit calcium also when grown in osteogenic McCoy medium (Figure 3.13). In addition, the ionic composition of DMEM was close to identical to that of aDMEM, Chapter 3. The Osteoblastic Ionic Extracellular Environment 135 which in turn was successfully used to induce calcium deposition by rMSCs. Instead, a more important factor was that MG63 cells expressed extremely low levels of ALP activity (Figure 3.18), a fact previously described by other authors (Clover & Gowen, 1994). One plausible reason for the absent calcium deposition by MG63 cells could be its abnormal extracellular matrix which contains significant amounts of collagen type III, a protein normally absent in osteoblast cultures but rather present in fibroblast-like cultures (Jukkola et al., 1993). Consequently, the likely reason for absent calcium deposition in MG63 cultures may be a combination of low ALP activity, and a defect osteoblast-matrix. Although literature clearly emphasises that behaviour of MG63 cells deviates from standard osteoblast-like cells, still one can find a few studies where these cells are used as a model of osteoblast mineralisation. Gregory et al. (2004) reported mineral deposition in MG63 cell layers after 21 and 28 days in culutre. Calcium deposition by MG63 cells (referred to as mineralisation) was also reported by Takagishi et al. (2006) and Kim et al. (2007). However, for calcium deposition to occur at all, Takagishi et al. (2006) had to force it by adding exogenous calcium at non-physiological levels (8 mM). In the study by Kim et al. (2007), glucosamine sulfate was provided to cells in combination with β-GP, and induced an almost three-fold increase of mineralisation after already seven days in culture compared to their control. In another study, Sun et al. (2009) reported faint Alizarin Red staining of MG63 cell layers maintained in non-osteogenic medium for up to 12 days. Taken altogether, MG63 cells seem to be a less appropriate cell model to use for studies of typical osteoblastic differentiation processes such as alkaline phosphatase activity and mineralisation events. 3.5.3 Osteoblastic Activity Observed Through the Ionic Extracellular Environment By careful preparation and storage of cell culture medium it was possible to study the effects of cellular activity on the ionic extracellular environment (Section 3.4.3). For all three cell models, as well as in both normal and osteogenic medium, cellular metabolic activity caused acidification of the cell culture medium already at early stages of the culture period. McCoy medium used for SAOS- 2 cultures maintained its pH stable in absence of cells, and therefore a clear effect of cellular respiration could be observed in both normal and osteogenic McCoy medium from around day 13. Interestingly, pH of SAOS-2 cultures continued to decrease until the end of the experiment, indicating that cells Chapter 3. The Osteoblastic Ionic Extracellular Environment 136 either continued to proliferate throughout the full experiment, or that the metabolic rate increased with stage of differentiation. More or less the same effect was observed in rMSC cultures maintained in osteogenic aDMEM medium. However, in normal aDMEM, the pH- decrease was less pronounced. Rather than an effect of drastic changes in metabolic activity, this observation could be ascribed to an expermintal artefact. Upon reaching confluency, two-dimensional rMSC cultures in normal medium tended to detach from the culture substrate and collapse into a three-dimensional, ball-like structure. It is not known whether cells continued to be alive within this structure, but with time the empty culture surface was repopulised by proliferating cells not engaged to the collapsed structure. Using osteogenic medium, the cell layer maintained much better anchored to the culture substrate. It is possible that rMSC cultures exposed to osteogenic medium reinforce their matrix both through specific protein production and through calcium deposition. In both DMEM media the pH spontaneously became more basic during incubation in absence of cells. Taken this effect into account, pH of DMEM in presence of cells clearly indicated cellular respiration. MG63 cells maintained in normal DMEM acidified the culture medium to a higher degree than cells maintained in osteogenic medium, indicating that osteogenic medium slowed down either the proliferation rate or the metabolic activity. It was further demonstrated that cell-induced calcium deposition in the extracellular matrix was related to significant alterations in concentrations of both calcium and inorganic phosphorus in osteogenic medium. SAOS-2 cells and rMSCs both responded to osteogenic medium by hydrolysing β- GP which in turn increased the concentration of inorganic phosphorus. The hydrolysation was highly efficient in SAOS-2 cultures, reaching a turnover efficiency of about 90% already after a few days with osteogenic medium, and this was maintained until the end of the experiment. This data is in agreement with the study of Chung et al. (1992) who reported that 80% of 10 mM β-GP was hydrolysed by osteoblastic activity within 24 hours. In rMSC cultures, βGP-hydrolysation reached a maximum at early stages of cultures (around day 9) and then stabilised or slowed down, indirectly indicating decreased total ALP activity with time. In parallel to the increased levels of inorganic phosphate, calcium concentration decreased in both SAOS-2 and rMSC cultures. In both cases, the decrease was relatively small during the first days, coinciding with the lower βGP-hydrolysation rate observed at this time point. Thereafter, the magnitude of calcium decrease stabilised in rMSC cultures throughout the full experiment, while in SAOS-2 cultures the calcium uptake was observed to Chapter 3. The Osteoblastic Ionic Extracellular Environment 137 decrease during the last days of culture. Although no calcium deposition was detected in the extracellular matrix of MG63 or rMSC cultures maintained in normal medium, still it was observed a slight decrease in extracellular calcium. This suggests that some of the ionised calcium was bound by soluble proteins produced by these cells, or accumulated intracellularly. Exactly which those proteins could be requires further investigation. 3.5.4 Biological Significance of Cell-induced Calcium Deposition βGP-mediated induction of calcium and phosphate deposition in the extracellular matrix is widely used as an indicator of bone tissue differentiation in vitro. However, it should be clarified once for all that its use and relevance as a biological model system has been debated. Basically, the question is if mineralisation occurring only in presence of high concentrations of exogenous organic phosphate can be considered physiological, and not mainly being a consequence of spontaneous mineral precipitation. The latter observation has been documented by several authors (Khouja et al., 1990; Chung et al., 1992; Beresford et al., 1993; Hamlin & Price, 2004), but is in many studies not even considered. In this study it was observed that hydrolysis of 10 mM βGP occurs in both presence and absence of cells, as indicated by increased concentrations of Pi. In medium without cells, concentration of Piincreased presumably due to endogenous alkaline phosphatase activity of FBS, as has been described by Khouja et al. (1990). However, FBS-induced hydrolysis of βGP was not sufficient to form mineral precipitation in any of the three medium compositions used, as evident by maintained calcium concentrations. On the contrary, in presence of ALP-active cells mineral precipitation could be observed already after nine days in culture (of which four was done with osteogenic medium). Considering that bone differentiation is a process that takes several weeks, this has to be considered as very early. Still, extremely rapid mineralisation response in vitro has previously been described (Stanford et al., 1995). In this study, the deposited calcium within mineralising cell cultures was initiated as separated areas being uniformly distributed over the sample. With time, and with continued exposure to β-GP, Alizarin Red staining observed through light microscopy revealed a rather homogenous distribution of calcium within the extracellular matrix. However, when observing the same samples through a narrower wavelength region, positive ARS staining was only observed as clearly separated nodules. Gentleman et al. (2009) applied Chapter 3. The Osteoblastic Ionic Extracellular Environment 144 Murshed, M., Harmey, D., Millan, J. L., McKee, M. D., & Karsenty, G. (2005). Genes & Development, 19, 1093–1104. Nakano, Y., Addison, W. N., & Kaartinen, M. T. (2007). Bone, 41, 549–561. Nauman, E., Sakata, T., Keaveny, T., Halloran, B., & Bikle, D. (2003). Calcified Tissue International, 73, 147–152. Neuman, W. F. & Neuman, M. W. (1958). The Chemical Dynamics of Bone Mineral. : The University Of Chicago Press. Ohgushi, H., Dohi, Y., Katuda, T., Tamai, S., Tabata, S., & Suwa, Y. (1996). Journal of Biomedical Materials Research, 32, 333–340. ONeill, C. A. & Galasko, C. S. B. (2000). Calcified Tissue International, 67, 53–59. Orimo, H. & Shimada, T. (2008). Molecular and Cellular Biochemistry, 315, 51–60. Porter, A. E., Patel, N., & Best, S. (2008). In: Encyclopedia of Biomaterials and Biomedical Engineering, (Wnek, G. E. & Bowlin, G. L., eds) pp. 1451– 1463. Informa Healthcare. Reid, I. R. (1997). European Journal of Endocrinology, 137, 209–217. Rodan, S. B., Imai, Y., Thiede, M. A., Wesolowski, G., Thompson, D., Bar- Shavit, Z., Shull, S., Mann, K., & Rodan, G. A. (1987). Cancer Research, 47, 4961–4966. Romberg, R. W., Werness, P. G., Riggs, B. L., & Mann, K. G. (1986). Biochemistry, 25, 1176–1180. Salgado, A. J., Coutinho, O. P., & Reis, R. L. (2004). Macromolecular Bioscience, 4, 743–765. Shimko, D. A., Burks, C. A., Dee, K. C., & Neauman, E. A. (2004). Tissue Engineering, 10 (9/10), 1386–1398. Stanford, C. M., Jacobson, P. A., Eanes, E. D., Lembke, L. A., & Midura, R. J. (1995). The Journal of Biological Chemistry, 270 (16), 9420–9428. Stinson, R., Thacker, J., & Lin, C. (1993). Clinica Chimica Acta, 221, 105–114. Chapter 3. The Osteoblastic Ionic Extracellular Environment 145 Sun, J., Wei, L., Liu, X., Li, J., Li, B., Wang, G., & Meng, F. (2009). Acta Biomaterialia, 5, 1284–1293. Takagishi, Y., Kawakami, T., Hara, Y., Shinkai, M., Takezawa, T., & Nakamune, T. (2006). Tissue Engineering, 12 (4), 927–937. Takamizawa, S., Maehata, Y., Imai, K., Senoo, K., Sato, S., & Hata, R.-I. (2004). Cell Biology International, 28, 255–265. Terkeltaub, R. A. (2001). American Journal of Physiology and Cellular Physiology, 281, C1–C11. Thouverey, C., Strzelecka-Kiliszek, A., Balcerzak, M., Buchet, R., & Pikula, S. (2009). Journal of Cellular Biochemistry, 106, 127–138. Triffitt, J. (2002). Journal of Biomedical Materials Research, 4, 384–389. Tuan, R. S., Boland, G., & Tuli, R. (2002). Arthritis Research and Therapy, 5(1), 32–45. Uccelli, A., Moretta, L., & Pistoia, V. (2008). Nature Reviews Immunology, 8, 726–736. Valerioa, P., Pereirab, M. M., Goesc, A. M., & Leite, M. F. (2004). Biomaterials, 25, 2941–2948. Wu, L. N., Genge, B. R., Dunkelberger, D. G., LeGeros, R. Z., Concannon, B., & Wuthier, R. E. (1997). The Journal of Biological Chemistry, 272 (7), 4404–4411. Xiao, Z., Camalier, C. E., Nagashima, K., Chan, K. C., Lucas, D. A., de la Cruz, M. J., Gignac, M., Lockett, S., Issaq, H. J., Veenstra, T. D., Conrads, T. P., & Beck, G. R. (2007). Journal of Cellular Physiology, 210, 325–335. Yamanouchi, K., Gotoh, Y., & Nagayama, M. (1997). Journal of Bone and Mineral Metabolism, 15, 23–29. This page is intentionally left blank Chapter 4 Ion Reactivity of Calciumdeficient Hydroxyapatite 4.1 Introduction Sensors for characterisation of biochemical activity of biomaterials may be useful to determine how, and to what extent, a given material interacts with its surrounding environment. Historically, biomaterials have been of inert character, but with the birth of tissue engineering, scaffolds are nowadays rather designed to be multi-functional with respect to both its physical and chemical properties. For example, at some occasions biomaterials interact directly with the environment by sending favourable signals to cells residing within, or in close proximity to it, e.g. through drug-release. At other occasions, the interaction is an indirect consequence, e.g. through biodegradation. No matter the reason for the interaction, it is of greatest importance to understand how the presence of a biomaterial influences the instantaenous composition of the tissue engineering environment. Basically, such knowledge contributes to improved designs of biomaterials able to induce or suppress specific cellular activity. Following this reasoning, in this chapter it is demonstrated how a biomaterial developed for bone tissue engineering, namely calcium-deficient hydroxyapatite (CDHA), interacts ionically on different time-scales, and with two different cell culture media commonly used for in vitro osteoblast cell cultures. 4.1.1 In Search for the Ideal Bone Graft Material Although bone has regenerative capacity, its repair process is impaired in many clinical and pathological situations, something which is emphasised 147 Chapter 4. Ion Reactivity of Calcium-deficient Hydroxyapatite 148 by the fact that bone is the second most common tissue to be transplanted (blood being by far the most common one). According to Giannoudis et al. (2005) more than 2.2 million bone graft procedures in orthopaedics, neurosurgery and dentistry are done annualy all over the world, making it into a highly important socialeconomical event. More important is that behind this number hide the pain and decreased functionality experienced by patients suffering bone loss casued by for example trauma or tumours. Among the bone grafting methods, three different categories can be distinguished: 1. Autografts, i.e. the transplation of tissue from one part of the body to another, is the most successful grafting method. Harvesting bone from the same patient (e.g. from the iliac crest) generally provides excellent conditions for a successful bone repair process, but obviously complicates the operation for both the surgeon and the patient. 2. Allografts, i.e. the transplantion of tissue from a donor different than the patient, is the most common grafting method. Due to differences between donors this method yields variable results. Also one has to consider the immunological issue. To minimize the latter effect, the graft is commonly decellularised, something which severely decreases its mechanical properties. 3. Bone-graft substitutes, i.e. materials of biological or synthetic origin that can provide the necessary mechanical properties of bone, and the fundamental conditions for bone regeneration. Given the clinical and economical complications related to the two former methods, development of appropriate bone-graft substitutes is since the early 1980s a highly active field of research. The common strategy on which such research relies, is to enhance any or several of the material’s osteoinductive, osteoconductive, and osseointegrative properties. This means that primitive, undifferentiated, and pluripotent cells should be stimulated to develop into a bone-forming cell lineage (osteoinduction), and further allowed to grow on its surface and/or towards its interior through pores and channels (osteoconduction), and that finally the construct becomes well anchoraged to the existing bone tissue (osseointegration) (Albrektsson & Johansson, 2001). In addition to these specific requirements, as well as more general issues, such as biocompatibility, minimal fibrotic reaction, sterility and long-term storage, the ideal bone-graft materials should also provide similar mechanical strength and elasticity as healthy bone. Chapter 4. Ion Reactivity of Calcium-deficient Hydroxyapatite 149 The inherent regenerative capacity of bone has definitely contributed to make bone-graft substitutes one of the areas of biomaterials and tissue engineering research with the highest commerial product activity. As a matter of fact, the amount of bone substitutes for cavity filling is large, and there exist an extensive number of products with a wide variation in both origin, composition and mechanism of action (Hing, 2004; Place et al., 2009). The most simple classification of these materials would be based on their origin; naturally derived materials and ceramic composite materials (Table 4.1). Although all these materials demonstrate certain desirable properties for bone regeneration, all of them fall well behind when compared to autologous grafts, and at most they match the results that can be obtained with allogenic bone. Therefore, continued research and development of bone-grafting substitutes is well motivated. Naturally derived materials Demineralised bone matrix Polymers Collagen, fibrin, chitosan, starch Coralline hydroxyapatite Ceramic composites Calcium phosphate ceramics Hydroxyapatite HA Ca10(PO4)6(OH)2 β-tricalcium phosphate β-TCP Ca3(PO4)2 Biphasic calcium phosphate BCP HA + TCP Calcium-deficient hydroxyapatite CDHA Ca10−x(HPO4)x(PO4)6−x(OH)2−x Amorphous calcium phosphate ACP Cax(PO4)y Carbonated apatite CA Calcium phosphate cements Apatite cements End product: HA, CA, or CDHA Brushite cements Bioactive glasses Bioglass SiO2-CaO-Na2-P2O5 Bioactive ceramics Glass ceramics Table 4.1: Common materials used for bone grafts or bone tissue engineering applications. Chapter 4. Ion Reactivity of Calcium-deficient Hydroxyapatite 150 Among the bone-grafting strategies is also included bone tissue engineering (Figure 4.1), where bone-forming cells, e.g. mesenchymal or embryonic stem cells are combined with a scaffold to produce a viable tissue (Jukes et al., 2008; Di-Silvio et al., 2008). The scaffold is typically made from porous ceramics or polymers (Salgado et al., 2004) which should degrade in vivo and be replaced by newly formed bone. Although some progress has been made in this field, better understanding of the spatial and temporal distribution of cells and growth factors necessary for osteogenesis remains to be developed. One approach to improve clinical results include the incorporation of bone morphogenic proteins into the scaffold (Haidar et al., 2009). Ex vivo expansion Attachment to hydroxyapatite/ tricalcium phosphate particles In vivo transplantation into segmental defect Skeletal stem cell Bone marrow Bone matrix Figure 4.1: A typical bone tissue engineering concept. Image adapted from Bianco & Robey (2001). Chapter 4. Ion Reactivity of Calcium-deficient Hydroxyapatite 151 4.1.2 Calcium Phosphate Based Bone-graft Substitutes Calcium minerals based around orthophosphates constitute a particularly interesting candidate group as bone-graft and scaffold material for bone tissue engineering. According to LeGeros (2002), attempts to use calcium phosphates in clincal application dates back to as early as 1920, and the primary interest in these materials obviously stem from its chemical and morphological similarity to the mineral phase of biological apatites of bone, dentin, and enamel. Calcium phosphate compounds differ in chemical composition (it is convenient to classify them according to their Ca/P molar ration, which can vary between 0.5 and 2.0), as well as in their physiological form, being possible to obtain both as ceramics and cements. 4.1.2.1 Calcium phosphate ceramics Among the ceramics, hydroxyapatite (Ca10(PO4)6(OH)2, HA, with Ca/P = 1.67) was early distinguished as a promising bone-graft substitute due to its very high similarity with the inorganic component of bone and tooth mineral. Indeed, HA has been demonstrated to possess very good osteoconductive properties (Chang et al., 2000a), but still has to be considered a too simple model for human bone mineral since biological apatites usually present impurities in their crystal composition, which in turn lower the Ca/P molar ratio to less than 1.67, and therefore is preferably referred to as calcium-deficient hydroxyapatite (Ben-Nissan & LeGeros, 2008). The impurities mainly derive from carbonate substitution with either hydroxyl or phosphate groups of HA. A more accurate representation of biological bone mineral would be as follows (4.1), with 0 <x<1: (Porter et al., 2008) Ca10−x(PO4)6−x(CO3)x(OH)2−x(CO3)x(4.1) Impurities in the mineral crystal may induce changes in the material properties, that influences mechanical strength, thermal stability, or solubility of the material. Pure HA does not only have somewhat week mechanical properties (Chen & Boccaccini, 2008; Liu et al., 2008), like poor tensile strength and toughness, but it also has very low resorbability, which means that it will remain in the body for a long time (years) after implantation. The combination of slow degradation and week mechanical properties makes the implant site a serious focus of mechanical stress. Lowering the Ca/P molar ratio usually enhances the degradation rate, as is the case with tri-calcium phosphate (TCP). TCP exists in two main forms; Chapter 4. Ion Reactivity of Calcium-deficient Hydroxyapatite 152 α-TCP and β-TCP. While β-TCP can be sintered into a solid ceramic body when heated in the range of 700-1125◦C, α-TCP can hardly be sintered, but is used as a cement powder (Ito & Ohgushi, 2008). Both α- and β- TCP have been demonstrated to support osteoblast adhesion, growth, and differentiation, and can therefore be considered osteoconductive (Yuan et al., 2001; Ogose et al., 2005; Seebach et al., 2010). They can be used as bonegraft materials either as individual materials (Kamitakahara et al., 2008), or in combination with other materials, then referred to as biphasic calcium phosphates (BCP) (Daculsi & LeGeros, 2008). The relative dissolution rates of these materials can be ordered like below (Daculsi et al., 2005): α-TCP >> β-TCP >> BCP >> HA (4.2) An increased degradability obviously allows for successive substitution of the bone-graft material with newly formed bone. In addition, it also makes it possible to incorporate into the material certain bone growth factors that are released in a controllable manner via the degradation of the CaP compound. 4.1.2.2 Calcium phosphate cements While HA and β-TCP have demonstrated reasonable biological properties, their fabrication method, usually sintering at high temperature, poses restriction on the shape and size of the device, which in turn may cause problems of adaptation and fixation of it to the bone cavity. To circumvent this problem, calcium phosphate cements (CPC), being a phase mixture of aqueous liquid and finely pulverised calcium phosphate powders that forms hydroxyapatite at room or body temperature, have been developed. The formability of the material introduces major advantages from a surgical point of view as the material is easily shaped and can be injected with low operative invasion. Besides, as it hardens in the body, the good osteoconductive properties of CaP ceramics can be maintained (Habraken et al., 2007). When mixing the cement powder with an aqueous solution, it converts into calcium phosphate ceramic by hydrolysis reaction. Ginebra et al. (1999) demonstrated that α-TCP hydrolysis into CDHA is a two-step process. First the cement powder particles are dissolved by the liquid water, a process which depends on the surface area of the powder particles. The dissolution creates a local saturation of ions that are reprecipitated on the surface of the cement powder particles. Thus, in the second phase the reaction becomes diffusion controlled, and the CDHA become tangled and join together, which leads to hardening of the cement (Figure 4.3). The result is a low crystalline CDHA Chapter 4. Ion Reactivity of Calcium-deficient Hydroxyapatite 153 Figure 4.2: Calcium phosphate cements are injectable, and therefore allow for efficient bone cavity filling. Image adapted from SmartCaP project proposal, p.5. structure, with extremely small particle size, and thus a large surface area, high reactivity, and high adsorption property. In addition, the material is mouldable in situ which allows for perfect adaptation to bone cavities, and it further has self-setting ability at in vivo conditions. α-TCP HO 2 +α-TCP OH- H+ Ca PO 2+ 4 3- α-TCP OH- H+ Ca PO 2+ 4 3- Ca2+ PO4 3- +HPO4 2- α-TCP Ca2+ PO4 3- Ca 2+ OHHPO4 2- α-TCP Ca2+ PO4 3- Ca 2+ OHHPO4 2- α-TCP Ca2+ PO4 3- Ca 2+ OHHPO4 2- CDHA Figure 4.3: The formation of CDHA from α-TCP is a hydrolysis reaction that involves progressive dissolution of the α-TCP particles and precipitation of CDHA. Image adapted from PhD dissertation by Edgar Montufar. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 256 Figure 6.14: The sensor software has three main screens. In the first one (not shown here) the experiment is initiated by activating the sensor channels. In the second screen (top) the sensor signals are visualised in real-time both graphically and numerically. In the third screen (bottom) the measured calibration data is presented for rapid assessment of the sensor response. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 257 (a) Before depositon. (b) After deposition. Figure 6.15: Micrographs of Ag-wire before and after electrodeposition of AgCl. The potential between the Ag/AgCl wire and a double-junction Ag/AgCl reference electrode was monitored upon exposure to different concentrations of potassium chloride, observing that beyond ≈10−5M KCl the potential decreased with increasing concentration of KCl (Figure 6.16). From the same data, when plotting the EMF against the logarithm of the activity of chloride ions, a typical calibration curve for an ion selective electrode was obtained (Figure 6.17). The potential response was further seen to depend mainly on the presence of Cl−ions, as the electrode responded equally well to calcium chloride (CaCl2) as to KCl, while its response to potassium nitrate (KNO3) was close to absent. Furthermore, the durability, or life-time, of the Ag/AgCl electrode was tested by repeated calibrations on a weekly basis. As follows from Table 6.5 both the wire design and the disk design maintained their characteristics for a period of at least three weeks when stored in a solution of ionic strength similar to blood. Also, if exposed to cell culture medium containing proteins, the sensitivity of the Ag/AgCl electrode was maintained. The influence of different sterilisation methods on the electrode response was then tested, and it was observed that both UV irradiation and 70% ethanol can be applied without affecting the electrode sensitivity. However, steam autoclaving caused damage to the electrode body (the epoxy resin), which resulted in unstable signal. Finally it was examined how the electrode signal maintained over a longer time period (four days) in a solution of constant concentration of KCl. As evident from Figure 6.18, no significant drift was observed. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 258 Time (s) 0 3000 6000 9000 12000 EMF (mV) 0 50 100 150 200 250 300 -log a(Cl ) 4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 - Figure 6.16: Potential response of Ag/AgCl electrode when exposed to incrementing concentrations of KCl. The initial period (0 s < t < 1500 s) represents potential response to 10−7M and 10−6M KCl, and is out of range of the right Y-axis. log a -7 -6 -5 -4 -3 -2 -1 EMF (mV) -250 -200 -150 -100 -50 0 50 CaCl2 KNO3 KCl i Figure 6.17: Potential response of Ag/AgCl wire when exposed to varying concentrations of KCl, CaCl2, and KNO3. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 259 Wire electrode Disk electrode (mV decade−1) (mV decade−1) Ionic solution Day 1 -58.24 ±0.43 -57.75 ±0.57 Day 7 -58.86 ±0.24 - Day 14 -60.04 ±0.17 -57.80 ±0.56 Day 21 -59.19 ±0.20 -58.46 ±0.22 Cell culture media 24h exposure -58.21 ±1.37 - Sterilisation methods UV (15 min) -60.19 ±0.89 - Ethanol (70%) -58.70 ±0.21 - Autoclave (steam) - not working Table 6.5: Sensitivity and stability of Ag/AgCl electrode at different time points and in different conditions (n= 3). Time (h) EMF (mV) 120 130 140 150 160 120 24 36 48 60 72 84 Figure 6.18: Stability of Ag/AgCl wire potential exposed to 10−3M KCl during four days. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 260 6.4.3 Calcium Selective Electrode Characterisation Calcium selective electrodes were prepared by attaching a Ca2+-selective membrane on the tip of a plastic micropipette. The membrane separated the test solution from an internal reference electrolyte and an internal reference electrode (Ag/AgCl wire electrode). Characterisation of the electrodes was done both with simple ionic solutions (Section 6.4.3.1), and with cell culture medium containing proteins (Section 6.4.3.2). 6.4.3.1 Standard characterisation When exposed to increasing concentrations of CaCl2(10−7M to 10−1M), the potential of the electrode versus an external reference electrode increased gradually (Figure 6.19). Decreasing the concentration, resulted in a decrease in potential, indicating reversible complexation of Ca2+ ions to the selective membrane. The time required to achieve 90% of steady state potential upon changes in concentration of CaCl2was 31.4 ±4.9 s in the range -5.01 ≤log a(Ca2+)≤-3.47 (Figure 6.20). Plotting the potentials obtained from Figure 6.19 versus the logarithm of the activity (Figure 6.21), it was revealed a typical ISE response with a linear range in the range -5.01 ≤log a(Ca2+)≤-1.722 (which corresponds to a Ca2+-concentration range of 10−5M to 10−1M). The detection limit was around log a(Ca2+) = -6.13 (i.e. [Ca2+]≈10−6M), while the slope of the curve in the linear range was determined to 30.3 ±2.4 mV decade−1(R2 = 0.992, n= 45), which is considered a Nernstian behaviour for positively charged divalent ions. Regarding the selectivity of the Ca2+-selective electrode, it was observed to respond equally well to Ca(NO3)2as to CaCl2(Figure 6.21), and neither did changing the pH of the test solution influence the electrode potential at constant concentration of Ca2+. The selectivity of the electrode was further evaluated by measuring the potential of the electrode as aliquots of different ionic solutions (NaCl, KCl, and MgCl2, respectively and separately) were added to a background reference solution of 10−3M CaCl2. Interference to the electrode potential was detected mainly from potassium with kMP M =1.56, meaning that the electrode was more than 36 times more selective to calcium than to any other of the tested cations. Finally, the life-time of the electrode was evaluated by repeated calibration on a daily basis, and as indicated in Figure 6.22, the electrode maintained its sensitivity for at least up to one month when stored in 0.1 M CaCl2 between measurements. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 261 Time (s) 0 10000 20000 EMF (mV) -150 -100 -50 0 50 6 4 2 -log a(Ca ) 2+ Figure 6.19: Potential response of the calcium selective electrode when exposed to changing concentrations of CaCl2. Time (s) 0 400 800 1200 1600 EMF (mV) -100 -80 -60 -40 3.47 4.04 4.42 5.01 -log a(Ca ) 2+ EMF (experimental) Steday state EMF (-10%) Figure 6.20: Typical potential response of the calcium selective electrode upon changes in concentration under agitation. The potential typically reaches 90% of the steady-state value after about 30 seconds (grey shaded vertical zones). Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 262 0 20 40 60 80 100 120 140 EMF (mV) pH 1 2 3 4 5 6 7 8 9 10 11 log ai -8 -7 -6 -5 -4 -3 -2 -1 CaCl2 KCl NaCl MgCl Fixed [Ca ], varying pH Ca(NO ) 3 2 2+ 2 Figure 6.21: Ca2+-electrode potential response to different ions. Time (days) Sensitivity (mV / decade) 10 20 30 40 50 10 20 30 0 Figure 6.22: Life-time of one specific Ca2+-electrode, evaluated through its sensitivity. Dashed line indicates the mean sensitivity of a larger set of sensors (n= 45). Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 263 6.4.3.2 ISE characterisation in complex solutions For in vitro applications of the Ca2+-selective electrode related to tissue engineering, it was also required to evaluate the influence of issues such as UV irradiation (for sterilisation), protein biofouling, drift, and biocompatibility / cytotoxicity of the sensor when exposed to cell cultures. As shown in Figure 6.23, Ca2+-sensors exposed to UV irradiation during 10 minutes did maintain its sensitivity in the activity range -5.01 ≤log a(Ca2+)≤-1.722. Although it is indicated that the detection limit might have been influenced, more data is required to conclude if this observation is true. After sterilisation through UV irradiation, the sensors were exposed to cell culture medium containing serum proteins. As shown in Figure 6.24, the electrode responded by a rapid decrease in potential. The potential was stabilised within 6 to 24 hours, and from then on all sensors demonstrated a linear drift towards more negative potentials. This intrinsic drift was calculated to 0.095 ±0.02 mV h−1 when sensors were maintained in DMEM with 10% FBS (n= 6). Following UV irradiation and exposure to protein-containing cell culture medium, the sensors were recalibrated according to Section 6.3.3.1, demonstrating not only maintained functionality, but also preserved sensitivity: 29.9 ±2.4 mV decade−1with R2= 0.988, and for n= 14 (Figure 6.25). Finally it was observed that osteoblast-like SAOS-2 cells grown with presence of Ca2+-sensors in their cell culture medium during 72 hours, maintained a healthy aspect (Figure 6.26). It was quantitatively confirmed that cellular proliferation was not influenced by the presence of sensors in the cell culture medium (Figure 6.27). Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 264 log a(Ca ) -7 -6 -5 -4 -3 -2 -1 EMF (mV) 0 20 40 60 80 100 120 2+ Before exposure (n=1) After exposure (n=1) Before exposure (n=3) After exposure (n=3) Figure 6.23: The influence of UV irradiation on Ca2+-selective electrode performance. Calibration was done from 10−7M CaCl2(n= 1), or from 10−5M (n= 3). Time (h) 12 18 24 30 36 42 48 EMF (mV) -35 -30 -25 -20 -15 -10 -5 0 60 Sensor 1 Sensor 2 Sensor 3 Intrinsic drift Conditioning period Figure 6.24: Typical potential response upon initial contact with cell culture medium but in absence of cells (here shown for three different sensors). The conditioning period indicates the time required to assure steady state potential. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 265 log a(Ca ) -5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 EMF (mV) 0 20 40 60 80 100 2+ Best-fit, control Best-fit, biofouling Exp. data (Control) Exp. data (Biofouling) Figure 6.25: Calibration curve, before and after exposure to protein containing cell culture medium during 24-96 hours. Data is expressed as mean ±standard deviation, with n= 14. Figure 6.26: Calcein (green) / Propidium Iodine (Red) viability staining of SAOS- 2 cells grown in presence of a Ca2+-sensor in its cell culture medium. Scalebar = 500 µm. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 272 Figure 6.33: Representaqtive images of the aspect and viability of SAOS-2 cells that had been exposed to real-time Ca2+-measurements were observed through phase contrast (top), as well as with epifluorescence (bottom) after staining with fluorescein diacetate (FDA). Both images represent the same field of view. Scalebar = 500 µm. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 273 6.5 Discussion In previous chapters it has been clearly demonstrated the utility of ion selective electrodes in certain bone tissue engineering applications. In accordance with the conclusions of the second chapter, sensors can be beneficial to the TE process in two ways. They can provide real-time data, for example on vital processes related to growth and state of the developing tissue, as well as information from environments normally inaccessible to standard evaluation technologies. Therefore, the ambition of this chapter has been to develop a general platform for measurements of ions in the TE environment. The platform has in this chapter been evaluated for calcium selective electrodes, and applied in three separate processes related to bone tissue engineering. 6.5.1 On the Design of the Sensor Platform The sensor platform included instrumentation for real-time potentiometric measurements, as well as chemical sensors selective to chloride and calcium, respectively. Regarding the former platform component, it has been demonstrated in this chapter that it provides acceptable means for real-time monitoring of potential differences between indicator and reference electrode, at time scales relevant to most TE applications. If required, the accuracy of the signal could be further improved by band-pass filters following the output signal of the operational amplifier, by decreasing the maximum voltage range of the DAQ device, and / or by implementation of software filters. Further shielding of the electronic device and cables may be beneficial to reduce externally induced noise. Additional channels are easily incorporated according to the same principle as existing channels. On a more sophisticated level, wireless communication of the sensor signals can be considered as an attractive feature for further miniaturisation of the platform (Johannessen et al., 2004). Regarding the chemical sensors, the strategy was to develop a generic sensor design which allowed for minimally invasive measurements in different physical environments, e.g. cell culture microplates or different bioreactors, and which in the future could be modified for detection of other ions than the one described in this chapter. In the end, the adopted design fell on the conventional ISE with an internal reference electrode and electrolyte, and which had been previously described by Radomska et al. (2008). Besides being the most well-established design, the great benefit was that it allowed for fabrication of a high number of sensors in short time and at low cost, which was considered necessary in order to be able to characterise well their behaviour in complex TE environments, and also for developing new formulations of Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 274 the polymeric membranes for future applications. The polymeric membrane was easily attached on the pipette tip and kept the aqueous internal electrolyte separated from the test solution. The membrane could be exchanged for any similar ion selective membrane, thus making the electrode design highly versatile. If desired, the internal electrolyte can be partly solidified by adding agar powder, typically 2-4 weight % (Hassel et al., 1999). Obviously, the use of an internal reference electrolyte puts limits on the final size of the electrode, but it also eliminates the problem of electrode packaging and wire insulation which are two additional requirements for solid-state electrodes. Also, since the main part of the electrode was made from material routinely applied in cell culture handlings, the electrode body was anticipated to not introduce any toxic effects. This was confirmed by the presented results. In the configuration presented in this chapter, the ion selective electrodes had a chemically active area of <1 mm2(this can be decreased using a smaller electrode body), and which made it perfectly possible to apply it to volumes <100 µl. This is an acceptable working volume with respect to cell culture applications and bioreactor designs, but still such measurements have to be considered as global measurements of the TE environment, and further minaturisation of the sensor would be required for use in local microenvironments. The first initiatives of such miniaturisation have been taken through the fabrication of prototypic solid-state electrodes, as was the case with the Ag/AgCl disk electrode which could be integrated in the α-TCP material interface. However, to be able to use such electrode design as an ion selective electrode, it is likely that the electrode surface will require chemical modification for an acceptable attachment of the ion selective membrane. Such treatments and tests have been out of scope in this thesis due to time limitations. In the current design of ion-selective electrodes, inexpensive internal Ag / AgCl reference electrodes were reproducibly prepared from electrolysis of Ag wires in HCl. Other methods to form AgCl on Ag include chemical oxidation, thermal or plasma treatment in chlorine containing atmospheres, or immersion in sodium dichromate dihydrate solution (see references within Polk et al. (2006)). The reference electrode could be made from other silver salts, e.g. Ag3PO4, and then produced e.g from electrolysis using K3PO4 (Ciobanu et al., 2002, 2004). The Ag/Ag3PO4reference electrode is a perfect option when the ion selective polymeric membrane is aimed at phosphate and when the internal electrolyte contains hydrogen phosphate (Kivlehan et al., 2007). Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 275 6.5.2 Calcium selective electrodes Methods for development of liquid membrane ion selective electrodes are well established (Ammann, 1986), and all necessary components are commerically available. Therefore, in this chapter it was used a well-established protocol to produce calcium selective polymeric membranes. The ionophore, a synthetic molecule specifically designed to bind calcium reversibly had been reported to possess sufficient sensitivity and lipophilicity for determination of Ca2+ in blood and human serum (Suzuki et al., 1995). Specifically, it was demonstrated that the ionophore, when housed in a polymeric membrane prepared according to Table 6.4, provided a Nernstian ISE response in the range 10−5 M to 10−1M Ca2+, both in absence and presence of interfering ions typically present in the extracellular environment (Suzuki et al., 1995). That particular data, in combination with the results on to what extent osteoblasts and CDHA may influence the extracellular calcium in cell culture medium (Chapter 3 and 4 of this thesis), motivated the use of that specific ISE membrane configuration. The standard characterisation of the calcium selective electrode carried out in this chapter not only confirmed that the selected ionophore was appropriate for detection of calcium (i.e. Nernstian behaviour, low interference with major extracellular cations), but also that the complete sensing platform (including internal reference electrode, electrolyte, and electrode body) was properly prepared, and that it responded relatively quickly to changes in calcium concentration. Although not reported by Suzuki et al. (1995), it was further observed that the electrodes maintained their calcium selective property during a period well above four weeks, and that they were completely insensitive to changes in pH. All this data indicated that the calcium selective membrane was a promising candidate for real-time monitoring of calcium in conditions related to in vitro TE applications. However, to be really sure about its appropriateness for such applications, it was required a characterisation of the electrode behaviour upon UV irradiation as well as when in contact with proteins present in the test sample. It was also required to determine the intrinsic drift during longer measurements, and to test if the sensor provoked any cytotoxic effects when combined with cellular cultures. As demonstrated in Section 6.4.3.2, neither of above potentially damaging conditions provoked any major deteriorating effects on the sensor functioning. Neither was the intrinsic drift observed to be excessively large, nor did the sensor provoke any cellular damage. Taken altogether, characterisation of the calcium selective electrode indicated that the electrode was possible to use for measurements of calcium in complex ionic solutions containing proteins and cells also during longer time periods. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 276 6.5.3 Sensors Applications Applications of the developed sensors have been aimed at in vitro bone tissue engineering since results obtained in previous chapters have indicated that both materials and cells typically used in bone tissue engineering may interact with its ionic environment. Specifically, the developed sensors were applied to measure (1) Cl−-activity during setting of α-TCP, (2) calcium activity in cell culture medium exposed to CDHA, and (3) osteoblast-induced calcium deposition in the extracellular matrix. 6.5.3.1 Cl−-activity during setting of α-TCP Hydrolysis of α-TCP leads to the formation of calcium-deficient hydroxyapatite (Ginebra et al., 1997). The transformation process of apatitic calcium phosphate cement typically involves active dissolution and precipitation of calcium and phosphate, and usually extends over a period of several days (Ginebra et al., 1999). Also, the pH of the aqueous phase is typically influenced during this period (see further Chapter 7). As most calcium phosphate cements are being developed with the purpose of being injectable, its chemical transformation is obviously supposed to take place in vivo. To somehow mimic biological conditions, setting of α-TCP in vitro has traditionally been done using Ringer’s solutions (Engel et al., 2008). So far, the work presented in this thesis has only concerned ionic interactions provoked by the end product obtained after seven days of hydrolysis of α-TCP maintained in Ringer’s solution. In addition, that interaction has only been characterised with respect to typically biologically relevant extracellular cations and phosphate. Therefore, the measurement of Cl−-activity during hydrolysis of α-TCP is important in several aspects. First, it serves as a reminder that the ionic activity of the studied material may involve other ions than the one mainly focused on in Chapters 4 and 5. Second, it highlights that the material possesses certain biochemical activity during its transformation phase into CDHA. Third, due to the all-solid-state nature of the Ag/AgCl disk electrode, a technological mean to measure the ionic activity in the direct interface between the biomaterial and its aqueous environment is provided. That is, the measurement was performed in the same region where cells would develop their affinity to the material if implanted in vivo. Specifically, it was shown that the forming material provokes a non-linear, relatively rapid, and highly significant, uptake of chloride from the solution. The induced change of the ionic strength of the material is of biological relevance in bone tissue engineering, as chloride ions on one hand can regulate Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 277 protein function (Stasio, 2004), and on the other hand can subsitute for hydroxyl ions in hydroxyapatite mineral (Dykes & Elliott, 1971). The successful modeling of the chloride uptake does not only contribute to elucidate the nature of ionic interaction provoked by the material, but it also opens up a possibility to apply the Ag/AgCl electrode as a potentiometric pseudo-reference electrode (Matsumoto et al., 2002) in applications related to α-TCP and CDHA. Then, and if combined with other solid-state ion electrodes incorporated in the material, such configuration would make it possible to obtain a large variety of ion measurements in the critical interface between material and extracellular environment. 6.5.3.2 Ca2+-sensor Initial evaluation of the developed miniaturised calcium-sensors exposed to standard in vitro conditions was done by provoking depletion of calcium in protein-containing cell culture medium by presence of CDHA. The nature of the calcium-uptake provoked by this material is well-known (see Chapter 4), and by measuring the relation between intial and final concentrations of calcium using commercial calcium electrodes, it was confirmed that the generated EMF signal represented well the calcium activity in the sample with respect to both kinetics and magnitude. The miniaturised Ca2+-sensor therefore can be considered an enabling tool for characterisation of biomaterial activity related to calcium uptake or release in standard in vitro conditions. For such applications, the major benefit of the proposed sensor design is its relatively small size, which easily allows for measurements in volumes down to 0.1 mL. To work with minimal volumes is preferable because it reduces costs, but even more important is that it may improve the representation of the environment that is created upon implantation. CDHA is just one of many biomaterials that interacts with calcium, and has served as an example. The interest to monitor calcium however goes beyond the pure material induced interactions since calcium is known to be a vital molecule in many cellular function (see Section 2.1.2.1). Therefore, to further test the developed calcium-sensor in a cellular application, osteoblasts were stimulated to develop a mineralised matrix, and the sensors were then used to monitor calcium deposition in the extracellular matrix during a period of 24 hours. Since such cell-induced calcium deposition strictly requires the combination of a properly developed collagenous matrix, biological regulation of possible inhibitors of mineralisation, as well as presence of phosphate (typically through alkaline phosphatase activity), a decreasing EMF signal due to calcium deposition could be considered as an indirect indication of the maturation stage of the osteoblasts (Allori et al., 2008a,b). Moreover, Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 278 as the kinetics of calcium deposition was observed to be different in young and more mature cell layers (Figure 6.32a), the continuous EMF signal to some extent may mirror the maturation of the cell layer. The dynamic data provided by the calcium sensors corresponds well with the kinetics of calcium deposition provoked by maturing SAOS-2 cells when obtained using commercial calcium electrodes at discrete time points, and which was earlier presented in Chapter 5 (Figure 5.5). As cells were little affected by the presence of the miniaturised calcium sensors, these sensors may therefore serve perfectly well as enabling tools of performing character in mineralising osteoblast cultures, and possibly also in other cell cultures where fluctuations in extracellular calcium can be related to cellular activity. 6.6 Conclusions In this chapter it has been described the development of a generic setup for online potentiometric ion measurements. The indicator electrode, which was based on an internal Ag/AgCl reference electrode and a polymeric ion selective membrane, was designed to physically allow for ion measurements in relatively small volumes (easily down to 0.1 mL). The applicability of the setup in standard in vitro tissue engineering conditions, i.e. in proteincontaining solutions of physiological ionic strength, was evaluated using a calcium selective polymeric membrane attached to the indicator electrode. The obtained calcium electrode exhibited a Nernstian response to calcium, and was little influenced by other major extracellular ions. Moreover, the electrode resisted sterilisation through UV irradiation, did not induce any cytotoxic effects in contact with osteoblasts, and the signal was subject only to minor drift during longer measurements. The developed Ag/AgCl and Ca2+-sensors were successfully applied to measure ionic activity of different bone tissue engineering components. The Ag/AgCl electrodes, originally aimed as internal reference electrodes, were used to measure Cl−-activity in the immediate interface between chemically transforming α-TCP and Ringer’s solution, revealing a material-induced nonlinear uptake of chloride. Ca2+-electrodes were used to successfully monitor sorption of calcium onto calcium-deficient hydroxyapatite immersed in cell culture medium, as well as osteoblast-induced calcium deposition in the extracellular matrix during a time frame of 24 hours. Bibliography Allori, A. C., Sailon, A. M., & Warren, S. M. (2008a). Tissue Engineering Part B, 14 (3), 259–273. Allori, A. C., Sailon, A. M., & Warren, S. M. (2008b). Tissue Engineering Part B, 14 (3), 275–283. Ammann, D. (1986). Ion-Selective Microelectrodes Principles Design and Application. London: Springer Verlag, first edition. Antonisse, M. M. & Reinhoudt, D. N. (1999). Electroanalysis, 11 (14), 1035–1048. Arnold, M. A. & Solsky, R. L. (1986). Analytical Chemistry, 58, 84R–101R. Bakker, E. (1997). Electroanalysis, 9(1), 7–12. Bakker, E., Bhlmann, P., & Pretsch, E. (1999). Electroanalysis, 11 (13), 915–933. Bobacka, J. (1999). Analytical Chemistry, 71, 4932–4937. Bobacka, J., Ivaska, A., & Lewenstam, A. (2008). Chemical Reviews, 108, 329–351. Ciobanu, M., Wilburn, J. P., Buss, N. I., Ditavong, P., & Lowy, D. A. (2002). Electroanalysis, 14 (14), 989–997. Ciobanu, M., Wilburn, J. P., & Lowy, D. A. (2004). Electroanalysis, 16 (16), 1351–1358. Dykes, E. & Elliott, J. (1971). Calcified Tissue International, 7, 241–248. Engel, E., del Valle, S., Aparicio, C., Altankov, G., Asin, L., Planell, J. A., & Ginebra, M. P. (2008). Tissue Engineering, 14, 1341–1351. 279 Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 280 Evans, A. (1987). Potentiometry and Ion Selective Electrodes. London: ACOL, Thames Polytechnic. Ginebra, M., Fernandez, E., Maeyerl, E. D., Verbeeckl, R., Boltong, M., Ginebra, J., Driessens, F., & Planell, J. (1997). Journal of Dental Research, 76 (4), 905–912. Ginebra, M.-P., Fernandez, E., Driessens, F. C. M., & Planell, J. A. (1999). Journal of the American Ceramic Society, 82 (10), 2808–2812. Hassel, A. W., Fushimi, K., & Seo, M. (1999). Electrochemistry Communications, 1, 180–183. Horvai, G. (1997). Sensors and Actuators B, 43, 94–98. Ives, D. J. G. & Janz, G. J. (1961). Reference electrodes : theory and practice. New York: Academic Press. Johannessen, E. A., Wang, L., Cui, L., Tang, T. B., Ahmadian, M., Astaras, A., Reid, S. W. J., Yam, P. S., Murray, A. F., Flynn, B. W., Beaumont, S. P., Cumming, D. R. S., & Cooper, J. M. (2004). IEEE Transactions on Biomedical Engineering, 51 (3), 525. Johnson, R. D. & Bachas, L. G. (2003). Analytical and bioanalytical chemistry, 376, 328–341. Kikas, T. & Ivaska, A. (2007). Talanta, 71, 160–164. Kivlehan, F., Mace, W. J., Moynihan, H. A., & Arrigan, D. W. (2007). Analytica Chimica Acta, 585, 154–160. Li, H., Luo, X., Liu, C., Jiang, L., Cui, D., Cai, X., & Yang, Q. (2004). Proceedings of 2004 International Conference on Information Acquisition, , 224–227. Maminska, R. & Wroblewski, W. (2006). Electroanalysis, 18 (13-14), 1347– 1353. Matsumoto, T., Ohashi, A., & Ito, N. (2002). Analytica Chimica Acta, 462, 253–259. Polk, B. J., Stelzenmuller, A., Mijares, G., MacCrehan, W., & Gaitan, M. (2006). Sensors and Actuators B, 114, 239–247. Chapter 6. Ion Sensors Fabrication, Characterisation & Applications 281 Radomska, A., Singhal, S., Ye, H., Lim, M., Mantalaris, A., Yue, X., Drakakis, E. M., Toumazou, C., & Cass, A. E. (2008). Biosensors and Bioelectronics, 24, 435–441. Stasio, E. D. (2004). Biophysical Chemistry, 112, 245–252. Stevic, Z., Andjelkovic, Z., & Antic, D. (2008). Sensors, 8, 1819–1831. Sutter, J., Peper, A. R. A. S., Bakker, E., & Pretsch, E. (2004). Analytica Chimica Acta, 523, 53–59. Suzuki, H. (2000). Electroanalysis, 12 (9), 703–715. Suzuki, K., Watanabe, K., Matsumoto, Y., Kobayashi, M., Sato, S., Siswanta, D., & Hisamoto, H. (1995). Analytical Chemistry, 67 (2), 324–334. Toczylowska, R., Pokrop, R., Dybko, A., & Wroblewski, W. (2005). Analytica Chimica Acta, 540, 167–172. Umezawa, Y., Buhlmann, P., Umezawa, K., Tohda, K., & Amemiya, S. (2000). Pure and Applied Chemistry, 72 (10), 1851–2082. Yajima, S., Shiraya, M., & Kimura, K. (2006). Chem. Anal. (Warsaw), 51, 939. Zine, N., Bausells, J., Vocanson, F., Lamartine, R., Asfar, Z., Teixidor, F., Crespo, E., de Oliveira, I. M., & ad A. Errachid, J. S. (2006). Electrochimica Acta, 51, 5075–5079. Chapter 7. pH Microelectrodes 288 and hydrogen evolution regimes (0-1.4 V, scan rate 50 mV second−1) using a platina counter electrode (BASi, MW-1032) and a Ag/AgCl double junction electrode as a reference electrode (Orion 900200). The Au electrode served as the working electrode in this conventional three-electrode electrochemical cell. The Au electrodes were considered clean when the voltammograms were reproducible. The quality of the active surface of the microelectrodes was assessed by potential cycling in an identical setup as above, but now between 0.5 and -0.2 V, and with 10 mM potassium ferricyanide(III) (Sigma, 244023) in supporting electrolyte (0.5 M KCl). The scan rate was 5 mV second−1. 7.3.2 Iridium Oxide Deposition Iridium oxide films were formed on the Au microelectrodes through anodic electrodeposition using an alkaline iridium tetrachloride solution. The solution was prepared following the steps given by Yamanaka (1989) and Bitziou et al. (2008): 1. 0.15 g of iridium(IV) chloride hydrate (IrCl4·H2O, Sigma 516996) was dissolved in 100 ml of milliQ water under magnetic stirring during 30 minutes. 2. 1 ml of aqueous hydrogen peroxide (H2O233% w/v, Panreac 211077 1214) was added to above solution to facilitate the use of low current density upon electrodeposition. The obtained solution was stirred for another 30 minutes. 3. 0.5 g of oxalic acid ((COOH)2·2H2O, Sigma 247537) was then added to the solution to avoid spontaneous precipitation of IrO2. Once again, the solution was left under stirring during 30 minutes. 4. Finally, about 5 g of anhydrous potassium carbonate (Sigma, 590681) was required to gradually adjust the pH of the solution to 10.5. The resulting solution, having a pale green-yellowish colour, was left standing for two days at room temperature until a colour change towards blue was achieved. From that on, the solution was stored cold (4◦C), and could be used during several months to successfully produce AIROFs through anodic electrodeposition. For that process, the cleaned Au-electrodes were inserted into an electrochemical cell containing the degassed iridium oxalate solution, together with a double junction Ag/AgCl reference electrode (Thermo, Orion 900200), and a Pt counter electrode (BASi MW-1032). A potentiostat (CH Chapter 7. pH Microelectrodes 289 Instruments CHI1232) was used to apply a constant potential (0.65 V) during three minutes while reading the current running in the circuit during deposition. After deposition, the coated microelectrodes were washed and left in water for at least two days prior to use as pH sensors. 7.3.3 pH Electrode Characterisation Calibration of the pH electrodes was performed in 0.1 M TRIS (Sigma, 15456- 3)1, to which different volumes of 0.1M or 1M hydrochloric acid (HCl) or sodium hydroxide (NaOH) were added. The voltage between the pH electrode and a reference electrode (Ag/AgCl) was recorded using either the potentiometric instrumentation developed and described in Chapter 6, or a commercial potentiostat (CH Instruments, CHI1232). At the same time, the pH of the analyte solution was measured with a commercial pH meter (Crison GLP 21). All chemical response experiments were performed at room temperature (25◦C) using chemicals of analytical reagent grade. With above procedure the relationship between pH and electrode potential could be characterised. To determine the long-term stability of the sensors, the measurements were repeated during several weeks. Analogously, it was also determined the time required for the sensor to stabilise after a change in pH. Between measurements the sensors were stored in distilled H2O, in dark and at room temperature. 7.3.4 Sensor Applications The developed pH sensors were used to measure pH in solutions containing α-tricalcium phosphate cement of different maturity (see Chapter 4). First, the evolution of pH during setting of α-TCP was measured with the pH sensors mounted directly inside the cement paste which was obtained upon mixing α-TCP with dH2O (liquid to powder ratio = 0.65 ml/g). For that measurement (Figure 7.2a), a miniaturised Ag/AgCl electrode housed within a Teflon body was used as reference electrode (Flex-Ref, Dri-RefTM World Precision Instruments). Second, the evolution of pH was then measured in 0.5 mL simulated body fluid (SBF), which was in direct contact with hydrolysing α-TCP (liquid to powder ration = 0.65 ml/g). For this measurement the sensors were mounted in the immediate interface between cement paste and SBF (Figure 7.2b-c), and potentiometric reading against a Ag/AgCl reference electrode was done during a period of four days, and at room temperature. The complete setup 1tris(hydroxymethyl)aminomethane buffer solution Chapter 7. pH Microelectrodes 290 was placed in a surrounding water bath to provide a humid atmosphere and avoid evaporation of the sample volume. The setup was sealed from below using a RTV silicone adhesive of implant grade (Applied Silicon, PN40064). Finally, the pH of protein-containing (15%) McCoy cell culture medium (Sigma, M4892) in contact with mature/set α-TCP (i.e. CDHA) was measured at standard cell culture conditions (i.e. 37◦C, 5% CO2) during 24 hours, using the same setup as described above (Figure 7.2b-c). To obtain CDHA samples with integrated pH sensors, α-TCP was hydrolysed (L/P = 0.35 ml/g) and then set at 37◦C during seven days in 0.9% NaCl with an electrode dummy placed inside the curing cement. When finally set, the dummy electrode was removed, and the obtained CDHA samples were rinsed and dried before fresh pH microelectrodes were introduced in the sample. The electrode body was first sterilised by immersion in 70% ethanol, and then after mounting and rinsing in dH2O, also through UV irradiation. Before and after treatment, electrodes were subject to calibration in 0.1M TRIS as described in Section 7.3.3. CDHA pH sensors Reference electrode pH sensors Hyrdolysing α-TCP (a) (b) (c) Figure 7.2: Setups for pH measurement of curing α-TCP (a), and of curing as well as cured α-TCP (=CDHA) at the material/liquid interface (b,c). Chapter 7. pH Microelectrodes 291 7.4 Results 7.4.1 Microelectrode Characterisation Contaminants such as hydrocarbons may arise from the environment and adsorb non-specifically to Au surfaces (Amrein & Muller, 1999), making cleaning a necessary step prior to further chemical modifications. For that purpose, electrodes were electrochemically cleaned through repetitive potential cycling in H2SO4, so that contaminants were oxidised at potentials where the metal oxide layer forms or where water discharge commences (Zoski, 2007). The voltammograms did stabilise with increased number of repetitions (Figure 7.3a), indicating a clean Au electrode. The observed peaks correspond to (incipient) formation of oxide at 1.0-1.4 V vs. Ag/AgCl during the positive exploration, and reduction of oxygen (around 0.94 V vs. Ag/AgCl) during the negative exploration (Tian et al., 2003; Vertova et al., 2008). The conductive properties of Au, and the ability of Fe2+/Fe3+ ions to exchange electrons with the bare Au substrate, result in a characteristic current as the potential is varied, and which directly depends on the electrode area (Forster, 1994). As shown in Figure 7.3b, the CV response at slow scan rates for the oxidation of 10 mM potassium ferricyanide(III) in supporting electrolyte (0.5 M KCl) at the Au electrode reproduced a sigmoidal-shaped curve which is typical for microelectrodes (Forster, 1994). Voltage vs. Ag/AgCl (V) 0.00.20.40.60.81.01.21.4 Current (nA) -30 -20 -10 0 10 20 n = 1 n = 3 n = 5 n = 7 n = 9 n = 10 (a) 0.5 M H2SO4, scan rate 50 mV sec−1. Current (nA) 0 50 100 150 Potential vs Ag/AgCl (V) -0.20.00.20.4 (b) 10 mM K3Fe(CN)6, scan rate 5 mV sec−1. Figure 7.3: Characterisation of Au microelectrodes by cyclic voltammetry. Given the observed steady-state current for the disk electrode (iss = 139.3 ±25.6 nA), it could be calculated the effictive electrode radius (r) from equation 7.7, iss = 4nFDCr (7.7) Chapter 7. pH Microelectrodes 292 where nis the number of electrons transferred (= 1), Fis the Faraday’s Constant (96487 C mol−1), Cis concentration (10 mM), and Dthe diffusion coefficient of K3Fe(CN)6. As the typical value of Dfor an aqueous solution is in the range 105cm2s−1, the average effective radii of the electrodes was calculated to r= 36.1 µm, indicating a well prepared microelectrode. 7.4.2 Iridium Oxide Deposition The deposition of iridium dioxide films onto the clean Au microelectrodes was achieved amperometrically using a constant-potential method. During this process (Figure 7.4a), the oxalate ligand becomes oxidised and forms CO2leading to the deposition of a hydrated form of iridium oxide according to reaction (7.8): [Ir(COO)2(OH)4]2−(aq) + 2OH−→[IrO2(OH)2·2H2O](s) + 2CO2+ 2e−(7.8) The cyclic voltammogram of IrO2-modified microelectrode was recorded immediately after deposition (Figure 7.4b), and did show the characteristic reversible redox behaviour Ir(III)/Ir(IV) which have been previously reported in literature (e.g. by Marzouk et al. (1998); Juodkazyte et al. (2005); O’Hare et al. (2006); Bitziou et al. (2008)), indicating a proper formation of an iridium oxide layer. The deposition was also clearly apparent upon investigation with scanning electron microscope (Fig 7.5d). Time (s) 120 160 200 Current (nA) -500 -400 -300 -200 -100 0 0 40 80 (a) Electrodeposition of IrO2. Voltage vs. Ag/AgCl (V) 1.2 Current (µA) -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 1.00.80.60.40.20.0 (b) CV after IrO2-deposition. Figure 7.4: IrO2deposition. (a) Anodic IrO2-deposition on Au microelectrode at constant potential (0.65 V). (b) CV of the IrO2-modified microelectrode in 0.5 H2SO4vs Ag/AgCl reference electrode. Scan rate 50 mV second−1. Chapter 7. pH Microelectrodes 293 a b c d Figure 7.5: (a,c) SEM images of bare Au microelectrode; (b) Incorporation of the microelectrode in CDHA for pH measurements in the material interface; (d) SEM image of IrO2modified microelectrode. Scalebars: white = 3.6 mm, black = 50 µm. 7.4.3 pH Electrode Characterisation After the electrodeposition procedure, the electrodes were left for at least two days to allow for hydration of the IrO2film, before they were subject to characterisation as pH electrodes. As shown in Figure 7.6a, while the pH of TRIS buffer was decreased by addition of HCl, the electrode potential against the reference electrode increased stepwisely. The time required for the sensor to acquire a stable potential upon changes in pH was approximately 10 seconds (Figure 7.6c). From data as the one presented in Figure 7.6a, the calibration curve of pH microelectrodes could be obtained (Figure 7.6b), revealing a perfectly linear relationship between pH and electrode potential. Calibration from pH 3 to 10.5 resulted in a super-Nernstian response (-64.9 ±2.4) mV per decade (n= 51, r2= 0.997) at T= 25◦C. The sensitivity decreased gradually with time, but still after 24 days their behaviour was super-Nernstian (Figure 7.6d). Chapter 7. pH Microelectrodes 294 Time (s) 0 1000 2000 3000 Potential (V) -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 pH 2 4 6 8 10 (a) Potential response to varying pH. pH 3 Potential (V) -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 4 5 6 7 8 9 10 (b) Calibration curve. Time (s) 0 10 20 30 40 50 EMF (mV) 150 155 160 165 170 175 180 pH 6.5 6.6 6.7 6.8 (c) Time reponse. Time (days) 0 Sensitivity (mV / decade) -80 -75 -70 -65 -60 -55 -50 3 6 9 12 15 18 21 24 (d) Stability. Figure 7.6: Calibration of IrO2pH microelectrodes. (a) Typical open-circuit potential response of IrO2electrodes to varying pH. (b) Calibration curve. (c) Time response upon changing pH. (d) pH sensitivity over time. 7.4.4 Sensor Applications The IrO2microelectrodes were applied to measure pH evolution of hydrolysing α-TCP (both inside and outside the cement paste), as well as of cured cement (CDHA) in protein-containing cell culture medium. The mixing of α-TCP powder with water provoked the potential of the IrO2microelectrodes to rapidly decrease until reaching a minima around pH 10 after about 40 minutes (Figure 7.7). Thereafter, the potential increased linearly during the following 2 hours, and from then on increased non-linearly until the end of the experiment. During 24 hours, the material caused pH to change almost 2.5 units. The same trend was observed when measuring the pH of simulated body fluid in contact with hydrolysing α-TCP (Figure 7.8). In that case, the pH sensors were incorporated in the interface between material and the aqueous phase. First, the pH of SBF increased, and reaching a maximum after about 13 hours, it began to decrease in a non-linear way, but did not reach its Chapter 7. pH Microelectrodes 295 steady-state value during the 92 hour-long measurement. Finally it was investigated how set α-TCP (CDHA) influenced the pH of cell culture medium (Figure 7.9). Initially it was observed that in absence of any other component, culture medium by itself became more alkaline when incubated in standard in vitro conditions. However, in presence of CDHA, the pH of cell culture medium was decreased. After normalising (i.e., subtracting the alkaline medium effect), the acidification induced by CDHA was revealed to be a non-linear process which could be successfully modelled (R2 = 0.995) using equation 7.9, EMF =at 1 + bt (7.9) where a= 3.38 mV h−1and b= 2.36×10−2h−1. Equation 7.9 is analog to the second-order sorption model applied in Chapter 4. After measurements, the sensors were calibrated once again to confirm their proper functioning (Figure 7.10). Time (h) 10 15 20 250 EMF (mV) -120 -80 -40 0 40 80 pH 7.5 8.0 8.5 9.0 9.5 10.0 Sensor I Sensor II 5 Figure 7.7: pH measurement performed with two independent IrO2microelec- trodes placed inside curing α-TCP cement prepared with a liquid to powder ratio of 0.65 ml/g. Chapter 7. pH Microelectrodes 296 Time (hours) 0 100 EMF (mV) -60 -40 -20 0 20 40 60 80 pH 6.5 7.0 7.5 8.0 20 40 60 80 Figure 7.8: pH measurement performed with two independent IrO2microelec- trodes placed at the interface between curing α-TCP (L/P = 0.65 ml/g) and simulated body fluid. Time (h) 0 12 18 24 EMF (mV) -40 -20 0 20 40 60 pH 6.6 6.8 7.0 7.2 7.4 7.6 7.8 Medium only + CDHA Normalised Model 6 Figure 7.9: pH measurement performed with two independent IrO2microelec- trodes placed at the interface between CDHA and McCoy cell culture medium with 15% FBS (solid thin lines). Measurements in absence of CDHA are indicated with dashed lines. Chapter 7. pH Microelectrodes 297 pH 10 EMF (mV) -200 -100 0 100 200 300 400 500 600 Before After 6 82 4 Figure 7.10: Calibration curve of IrO2microelectrodes before and after exposure to cell culture medium. 7.5 Discussion Application of sensors in tissue engineering environments can be useful to determine the instantaneous activity of components held in the environment. Just like cellular metabolic activity tends to acidify the extracellular environment, certain scaffold materials may interact chemically with its environment, and induce ionic interchanges between the solid and aqueous phases. One such group of materials is the apatitic calcium phosphate cements, including α-TCP which undergoes hydrolysis reaction in contact with water. The hydrolysis reaction is accompanied by changes in the pH as ions are released and precipitated back onto the material (see Figure 4.3). To appropriately evaluate the effect of the chemical activity of such materials on tissue development, it is necessary to characterise its behaviour in biologically and volumetrically relevant environments. Therefore, sensors applied to such environments should be made small, preferably be applied in the solid/liquid interface, and obviously perform correctly in solutions of ionic strength and protein content similar to that human blood. In this chapter effort has been made to fulfill above requirements when evaluating the chemical influence of α-TCP on the pH of its aqueous environment. Among the different approaches to produce pH and ion sensors, solidstate electrodes are especially interesting as they can be easily miniaturised Chapter 8. Conclusions and Future Work 304 centrations. Cell-induced changes on the extracellular ionic concentrations were detected for calcium and phosphorus. Also, pH of culture medium was influenced by cellular activity. Presence of cells had no detectable influence on extracellular [Na+] or [K+]. a) SAOS-2 cells maintained in normal McCoy medium had no effect on the ionic extracellular environment with respect to investigated ions. However, using osteogenic medium then extracellular [Ca2+] decreased significantly during a time period of ≈48 hours in presence of mature SAOS-2 cells (around 60%). At the same time, [Pi] in osteogenic McCoy medium was significantly increased in presence of SAOS-2 cells. b) Beyond five days in culture, rMSCs maintained in aDMEM were observed to slightly decrease [Ca2+] during a time period of ≈48 hours (3-7%). The decrease in [Ca2+] was greatly enhanced using osteogenic aDMEM (up to about 60%). As with SAOS-2 cells, rMSCs significantly increased [Pi] of osteogenic medium. c) Mature MG63 cells maintained in DMEM medium were observed to decreased [Ca2+] slightly during the time period of ≈48 hours (7-9%). In contrast to SAOS-2 and rMSCs, no or little influence on [Ca2+] and [Pi] was observed when MG63 cells were maintained with osteogenic medium. 4. Cell-induced influence on the extracellular concentrations of calcium and total inorganic phosphorus was associated with deposition of calcium in the extracellular matrix. 5. SAOS-2 cells induced deposition of calcium in their extracellular matrix using either osteogenic McCoy or osteogenic DMEM. In contrast, MG63 cells did not induce calcium deposition in its extracellular matrix using any of the two media. 6. SAOS-2 and rMSCs expressed significantly higher alkaline phosphatase activity than MG63 cells. Thus, the increase of extracellular Piin osteogenic medium can potentially be used as an indicator of cellular alkaline phosphatase activity. 7. The decrease of extracellular ionised calcium could serve as an indirect indication of the combined presence of a collagenous extracellular matrix and alkaline phosphatase active cells. In that way, decrease of extracellular calcium can potentially serve as an indicator of a first step towards mineralisation of the osteoid. Chapter 8. Conclusions and Future Work 305 8.1.2 The Scaffold Material (Chapter 4) 1. Calcium-deficient hydroxyapatite (CDHA) induced changes in the chemical composition of both DMEM and McCoy culture media. The size and nature of the changes were dependent on initial chemical composition of the aqueous phase as well accumulated time of exposure to medium. 2. During the initial 48 hours of contact, CDHA induced sorption of sodium, potassium, and calcium from all investigated media (i.e. normal and osteogenic DMEM and McCoy, respectively). CDHA also caused acidification of all media. Moreover, sorption of total inorganic phosphorus was observed using normal and osteogenic McCoy media and osteogenic DMEM medium, but not using normal DMEM medium. 3. Most sorption processes were successfully described using pseudo-first or pseudo-second-order sorption models. Exceptions were sorption of sodium (all media), and sorption of phosphorus (osteogenic media). In the latter case, serum-induced hydrolysis of β-glycerophosphate increased the concentration of Pilinearly during the contact period. As a consequence, sorption of Pionto CDHA from osteogenic media was better described with a linear model. 4. Upon repeated exposure to fresh culture medium during 21 days, materialinduced acidification of culture medium gradually decreased with time in all media formulation. 5. Sorption of sodium and potassium gradually decreased with time during long-time experiments and was completely undetectable towards the end of the experiment. That observation proposes that CDHA has limited space for incorporation of those ions into the crystal structure. 6. Sorption of calcium onto CDHA from all culture media persisted throughout the full length of the experiment. It follows that the material must have possessed a high number of vacant sites for calcium. 7. While sorption of Pidecreased with time using McCoy medium, CDHA was observed to release Piwhen repeatedly exposed to fresh normal and serum-free DMEM media. The release mechanism involved both sorption and desportion of Pi, but with a net desorption effect during 48 hours. FTIR analyses indicated that phosphate/carbonate replacement processes had occured in all samples. Therefore, such ion-exchange must be taken into account when explaining CDHA reactivity with respect to phosphorus in both DMEM and McCoy media. Chapter 8. Conclusions and Future Work 306 8.1.3 Cells and Scaffold Combined (Chapter 5) 1. It was observed that CDHA decreased the extracellular concentration of calcium in osteogenic McCoy medium in a reproducible manner during the three-week long experiments. Moreover, the decrease could be perfectly described by a second-order sorption model. In contrast, the influence of CDHA on extracellular total phosphorus (i.e. sorption) and pH of culture medium (i.e. acidification) was not constant over time, but was observed to gradually decrease for each time the culture medium was exchanged. 2. SAOS-2 cells maintained in osteogenic medium influenced the extracellular concentration of total inorganic phosphorus by hydrolysis of β-glycerophosphate through its alkaline phosphatase (ALP) activity. In absence of CDHA, cell-induced hydrolysis of β-GP was highly efficient throughout the full length of experiment as [Pi] always increased rapidly in presence of cells. 3. Moreover, SAOS-2 cells grown in absence of CDHA caused calcium in culture medium to be gradually deposited into the extracellular matrix already at an early stage (nine days in culture). 4. Using semi-permeable culture inserts, SAOS-2 cells could be directly exposed to the dynamic changes of the ionic extracellular environment induced by CDHA. Therefore, when using osteogenic medium it was created a situation where CDHA and cells would compete for the same calcium and where CDHA would sorb Piwhich had been liberated by cellular activity. 5. In normal McCoy culture medium that contained cells separated from CDHA, the concentrations of calcium and total inorganic phosphorus were observed to be mainly determined by CDHA. In contrast, when using osteogenic McCoy culture medium the concentration of total inorganic phosphorus was clearly subject to both ion reactivity of CDHA and cellular ALP activity. It was however not possible to determine from the ion measurements if the observed decrease in [Ca2+] of osteogenic medium was due mainly to material-induced sorption or cell-induced calcium-deposition. 6. The dynamic ionic environment induced by CDHA did not influence greatly cellular proliferation using normal McCoy medium. Yet, by the end of the experiment (i.e. after three weeks), higher cell proliferation was observed for cells grown in presence than in absence of CDHA. It Chapter 8. Conclusions and Future Work 307 was also observed that osteogenic medium slowed down cellular proliferation earlier in absence than in presence of CDHA. Both observations could be due to that cells grown in presence of CDHA were delayed in their cellular differentiation compared to cells grown in absence of CDHA. 7. In absence of CDHA, cell-induced deposition of calcium in the extracellular matrix was observed already after nine days in culture using osteogenic medium. In presence of CDHA, such calcium deposition was severely delayed and only to be observed at the very end of the culture period (i.e. day 21). The reason(s) for the suppressed and delayed cell-induced calcium deposition in the extracellular matrix in presence of CDHA was tested: (a) The rate of cell-induced deposition of calcium in the extracellular matrix in absence of CDHA was comparable to the sorption rate of calcium induced by CDHA when cells were highly mature. Thus, when combined in the same culture environment, mature cells and CDHA could in prinicple compete for the same calcium under fair conditions. (b) When growing SAOS-2 cells during five days in presence of CDHA, and thereafter during another four days in absence of CDHA, cellinduced calcium deposition was perfectly comparable to calcium deposition induced by cells grown in absence of CDHA throughout the full period of nine days. That observation indirectly indicated that cells initially grown in presence of CDHA resided in a welldeveloped extracellular matrix. (c) Cellular ALP activity was reduced only at day 3 and day 9 compared to cells grown in absence of CDHA if using normal medium. (d) Osteogenic medium in general increased cellular ALP activity. Maximum ALP activity was very similar both in presence and absence of CDHA, but it occurred at different time points. In presence of CDHA, maximum ALP activity was delayed until day 15. However, even at that time point still insignificant calcium deposition was observed in presence of CDHA. Therefore, absent calcium deposition in the extracellular matrix ought not to be entirely ascribed to delayed cellular ALP activity. (e) The ionic extracellular environment determined by CDHA must have played an important role in the delayed cell-induced calcium Chapter 8. Conclusions and Future Work 308 deposition. Since material-induced sorption of calcium was constant throughout the full length of the experiment, it is tempting to speculate that the main suppresser for cell-induced calcium deposition was the extracellular concentration of Piwhich in contrast to calcium gradually became less affected by CDHA over time. 8.1.4 Ion Sensors for Online Monitoring (Chapter 6) 1. An instrumentation for multi-channel potentiometric measurements was fabricated and successfully tested and used for various types of ion electrodes. 2. Miniaturised Ag/AgCl electrodes were successfully produced in house from electro-oxidation of Ag wires. The Ag/AgCl electrodes (in shape of both wire and disk electrodes) responded to chloride activity according to the Nernst equation and they maintained their sensitivity during several weeks. The electrode potential demonstrated insignificant drift at constant concentration and temperature. The fabricated Ag/AgCl electrodes held promise to be useful both as internal reference electrodes for potentiometric ion-selective electrodes as well as for direct measurements of chloride activity. 3. Ag/AgCl disk electrodes were used to measure chloride activity in realtime at the direct solid/aqueous interface between hydrolysing α-TCP and Ringer’s solution. The electrode signal revealed a fast, non-linear sorption of chloride ions onto the material. The sorption could be perfectly described by second-order sorption models. 4. Calcium-selective electrodes were elaborated from a traditional calciumselective polymeric membrane and a Ag/AgCl wire electrode which was housed in an internal electrolyte of constant concentration. That configuration allowed for fast and cheap fabrication of numerous Ca2+- electrodes of small dimensions, suitable for incorporation in tissue engineering bioreactors or similar cell culture environments. The same design can be used to elaborate electrodes for other ions than calcium by only changing the polymeric membrane and the internal electrolyte solution. 5. The calcium-selective electrodes responded to changes in calcium concentration within reasonable time frames (31.4 ±4.9 seconds) and according to the Nernst equation (30.3 ±2.4 mV decade−1). The sensitivity was preserved for at least four weeks when used with simple Chapter 8. Conclusions and Future Work 309 salt solutions. The electrodes further demonstrated superior selectivity towards calcium ions compared to other major extracellular cations. 6. Characterisation of calcium-selective electrodes in more complex solutions revealed that the fabricated electrodes withstood well both UV irradiation (for sterilisation purposes) and prolonged exposure of protein-containing culture medium (up to 96 hours). The intrinsic drift of potential in protein-containing culture medium was small (0.095 mV h−1), and could be well described by a linear function. 7. Introduction of calcium-selective electrodes into osteoblast-like cell cultures did not influence the cellular proliferation. 8. The calcium-selective electrodes were used to measure sorption of calcium onto CDHA in real-time during 24 hours and in protein-containing culture medium at standard in vitro conditions. The obtained data fitted the predicted sorption behaviour, and initial and final potential readings were in agreement with reference measurements. 9. Regarding cell-induced calcium deposition in osteoblast-like cultures it was first observed that calcium deposition occured in SAOS-2 cultures provided either low or high concentration of β-glycerophosphate (3 or 10 mM, respectively). In both cases, the calcium deposition was monitored in real-time, indicating a clear cellular influence on extracellular calcium during the measurement. 8.1.5 Sensors for Local Measurements (Chapter 7) 1. Solid-state Au microelectrodes were fabricated in house, and were made sensitive to pH using well-known protocols for electrodeposition of iridium oxide onto the metal electrode surfaces. The IrO2-modified electrodes were confirmed to have a super-Nernstian response (-64.9 ±2.4) mV decade−1at room temperature, and it changed little over several weeks if stored in salt buffer solution. 2. IrO2pH microelectrodes were located inside curing α-TCP cement that was prepared with a clinically relevant liquid-to-powder ratio (0.65 mL/g). Continuous potentiometric readings during 24 hours revealed a strong alkalinisation of the cement paste during the first couple of hours from mixture, reaching a minimum pH around 10 units. Thereafter the mix acidified until the end of the experiment but without reaching steady-state. A similar pH evolution was observed when pH microelectrodes were located at the immediate cement/liquid interface. Chapter 8. Conclusions and Future Work 310 3. Finally, IrO2pH microelectrodes were located at the immediate solid/liquid interface of cured α-TCP (i.e. CDHA) and protein-containing cell culture medium. Online measurements during 24 hours revealed nonlinear acidification of culture medium due to presence of CDHA. 8.2 Perspectives As demonstrated, the ionic concentrations of mainly calcium and phosphorus in osteogenic culture medium were significantly influenced by presence of cells that expressed high alkaline phosphatase activity. Real-time (or strategic discrete) measurements of the ionic extracellular environment could therefore to some extent be used to estimate bone tissue development without affecting the sample. Such measurements would be highly beneficial during largescale production of mineralised bone, or during elaboration of new protocols related to bone mineralisation events. The different behaviour of SAOS-2, rMSC, and MG63 cells with respect to both ALP activity and calcium deposition highlights the general importance of choosing an appropriate cell model when studying cellular response mechanisms to a certain biomaterial in vitro. And it becomes even more important if the biomaterial is biochemically active with respect to calcium and phosphorus, as was the case with for example CDHA. The situation is further complicated by the fact that ion reactivity of CDHA was dependent on both the chemical composition of the aqueous phase and accumulated time of exposure to fresh culture medium. Therefore, meaningful extrapolation of cellular response mechanisms to calcium phosphate compounds obtained from studies performed in vitro will minimally require a perfect understanding of all components of the TE envrionment: i.e. cells, scaffold material, and culture medium. The application of ion sensors as enabling tools in TE have in Chapters 3 and 4 been clearly demonstrated to possess a key role in trying to achieve such knowledge. Although difficult to interpret and extrapolate, attempts to learn something concrete about cellular response to CDHA from in vitro studies are far from useless. One such attempt was presented in Chapter 5 of this thesis. It was revealed that the capacity of cells to reinforce the extracellular matrix with calcium phosphate was severly decreased in presence of CDHA. Yet, as the material’s sorption of phosphorus decreased with time, conditions for calcium deposition were slowly created, and by the end of the three-week long experiment significant amounts of calcium was detected in the extracellular matrix of cells cultured in presence of CDHA. That observation proposes that it is phosphorus rather than calcium that initially prevent osteoblastic Chapter 8. Conclusions and Future Work 311 mineralisation to occur in presence of CDHA. As the conditions of the performed experiments were created as a ‘worst-case’ scenario by using the culture medium subject to highest CDHA reactivity with respect to phosphorus, different cellular response may be obtained using a different culture medium (e.g DMEM which releases Pito its environment and therefore should promote calcium deposition rather than prevent). Still, the determining role of phosphorus over calcium should likely be independent of the medium used. Given that ionic interactions provoked by cells and/or biomaterials (like CDHA, but also most other calcium phosphate compounds) at instances can be indirectly indicative of the tissue development process or the conditions for it, it was desired to develop instruments that allowed for real-time monitoring of ions in standard in vitro conditions. For that purpose it was fabricated easy-to-make, cheap, miniaturised ion-selective electrodes, and as a proof-of- concept these electrodes were successfully applied to measure calcium activity in global environments of small-volume containing mineralising osteoblasts and CDHA, respectively. Further miniaturisation of sensors (here as allsolid-state pH microelectrodes) makes it possible to approach traditionally inaccessible TE environments, such as the cell-biomaterial interface. The benefits of the developed system include being able to work with biologically and volumetrically relevant environments, as well as simultaneously obtain data from different ions. Such measurements will prove extremely useful for improved characterisation of ion reactive biomaterials, which in the end also will favour correct extrapolation of its in vivo performance. 8.3 Future Work Looking back at the presented work, there are naturally several aspects of it that could have been done or approached differently, as well as there are many ways to continue developing the discussed topics. One major issue in bone tissue engineering is to produce high quality mineralised bone, but as discussed in Section 3.5.4, the protocols used to induce such bone formation in vitro are highly debated. Although care was taken to include different osteoblast-like cell models and work with to low concentration of β-glycerophosphate (down to 3 mM), an additional cell model which induce characteristic nodule formation would have been highly beneficial in order to compare such mineralisation both quantitatively and qualitatively with the massive calcium deposition provoked by SAOS-2 and rMSC cells. One such cell model is the MC3T3-E1 line derived from mouse (Balint et al., 2001). Chapter 8. Conclusions and Future Work 312 The characterisation of CDHA ion reactivity presented in Chapter 4 was initiated with the objective to determine the biochemical capacity of CDHA to influence cellular behaviour. Therefore, all ionic interactions were monitored in complex protein-containing solutions. Obviously, the ionic interactions would be better characterised if working with simpler ionic solutions, as well as with more precisely specified surface area. Therefore, sorption studies with ion reactive biomaterials are suggested to be complemented with single ionic solutions of varying concentrations in order to establish sorption isotherms of each ion, and from which can be determined the maximal sorption capacity of the material. It is also suggested that the characterisation includes sorption of carbonate. Moreover, as long-term experiments revealed that sorption is a function of material maturity, it would be informative to perform short-term sorption studies with material of different maturity to more carefully determine how ageing of the material influences the ionic interactions with the aqueous surrounding. The material’s capacity to respond differently to small changes in the ionic composition of its aqueous surrounding further suggests that future characterisation of CDHA (or similar materials) should be made in biological relevant solutions (e.g. blood or serum) and be sampled from a big pool. With respect to the developed ion sensors, there are many things to improve and also to further characterise in order to classify them as true enabling tools for certain tissue engineering processes. The intention of this work was only to establish a general platform for potentiometric measurements and to produce a few proof-of-concept measurements. From there on, future modifications of the system are numerous. To begin with, sensors selective to other ions than calcium could be easily achieved only by changing the composition of the polymeric membrane according to well-established protocols in literature. One tempting modification is the development of phosphate-selective sensors for simultaneous detection of Ca2+ and Piin osteoblast cultures and other TE environments that hold chemically active biomaterials such as CDHA. Although not as common as cation-membranes (no commercial alternatives exist), phosphate-selective membranes are indeed possible to develop if using an appropriate ionophore (Ganjali et al., 2006; Kivlehan et al., 2007). Any ion-selective membrane developed for application in standard in vitro environments should withstand the harsh conditions (mainly protein fouling and ionic interference) that it inevitably will be exposed to. In the scope of this work, the calcium-sensor withstood well such conditions during periods at least up to 48 hours. Since longer incubation times possibly will have deteriorating effects on the sensor properties, it could be necessary to evaluate Chapter 8. Conclusions and Future Work 313 different anti-fouling strategies (Gavalas et al., 2006). For example, physically protective coatings or incorporation of poly(ethylene glycol), PEG, into the polymeric membrane have both shown very promising results and are worth exploring (Trouillon et al., 2009; Radomska et al., 2008). In addition to the electrochemical properties of the sensor, also its shape can be greatly modified. In this work was used the conventional design of ion-selective electrodes with an internal electrolyte and reference electrode. Although the size of such sensor design can be further reduced, an all-solid- state design similar to the Ag/AgCl disk electrode or the pH microelectrode would allow for much greater miniaturisation. However, extra care then has to be taken that the polymeric membrane attaches well to the electrode surface. This can be achieved through coating of the electrode with a thin film of conducting polypyrrole that promotes adhesion of the polymeric ionselective membrane and establishes a well defined electrode interface (Zine et al., 2006). Reducing the size of the indicator electrode is however useless if not also the reference electrode can be miniaturised. Different approaches based on Ag/AgX reference electrodes embedded in hydrogels to reduce its sensitivity to chloride ions and pH have been reported (Ciobanu et al., 2002, 2004), and may be interesting for application in small environments. Miniaturised sensors that can perform measurements in small tissue engineering environments can provide a different perspective on how cells and materials integrate with each other, and how local microenvironments may arise at the cell-material interface. One such example is the growth of osteoblast cells directly onto CDHA. As seen in Figure 8.3a, presence of MG63 cells onto CDHA decreases the material-induced uptake of calcium with time. This indicates that those cells and their extracellular matrix act as a partly isolating biofilm to produce a local microenvironment at the material interface (recall Figure 2.4). The same isolating effect could not be observed with SAOS-2 cells which created a more heterogenous biofilm onto the surface of CDHA than MG63 cells (Figure 8.1). Application of sensors in the material interface can therefore be useful to learn more about the local chemical environments created around these biofilms in which cells reside. Chapter 7 includes the initial efforts to develop a system for pH sensing in similar microenvironments, and future work with that system could include the separation of the biofilm from the material through the use of cell culture inserts (Figure 8.2a). The insert can then be perfectly incorporated with the culture chamber used in Chapter 7, creating a minimal space between material and biofilm, in which pH measurements (or other ions if the microelectrode is further modified) can be realised (Figure 8.2b).