ELECTRICAL BIOREACTOR DESIGN FOR TISSUE ENGINEERING
291
ELECTRICAL BIOREACTOR DESIGN FOR TISSUE ENGINEERING A thesis submitted to The University of Manchester for the degree of Doctor of Philosophy in the Faculty of Engineering and Physical Sciences 2014 RICHARD BALINT School of Materials
Transcript of ELECTRICAL BIOREACTOR DESIGN FOR TISSUE ENGINEERING
ELECTRICAL BIOREACTOR DESIGN
FOR TISSUE ENGINEERING
A thesis submitted to The University of Manchester for the degree of
Doctor of Philosophy in the Faculty of Engineering and Physical Sciences
2014
RICHARD BALINT
School of Materials
2
CONTENTS
Table of Contents 2
List of Figures 8
List of Tables 16
List of Abbreviations 18
Abstract 19
Declaration 20
Copyright Statement 20
Acknowledgements 21
The Author 22
1. Chapter I: Introduction 24
1.1. The Clinical Background 25
1.2. The Hypothesis 28
1.3. An Introduction to the Disciplines of Regenerative Medicine and
Tissue Engineering 30
1.3.1. Cells 32
1.3.2. Biomaterials 32
1.3.3. Stimulation 32
1.4. Bone 34
1.4.1. Bone as a Material 34
1.4.2. The Cells of Bone 35
1.4.2.1. Mesenchymal Stem Cells 35
1.4.2.2. Osteoblasts 36
1.4.2.3. Osteocytes 37
1.4.2.4. Osteoclasts 37
1.4.3. The Markers of Bone Differentiation 38
1.4.3.1. Bone Morphogenic Proteins 38
1.4.3.2. Cbfa1/Runx2 38
1.4.3.3. Osterix 39
1.4.3.4. Bone Sialoprotein 39
1.4.3.5. Osteonectin, Osteocalcin and Ostepontin 39
1.4.3.6. Alkaline Phosphatase 40
1.4.4. The Expression Profile of the Osteogenic Markers 40
1.5. Electrical Stimulation 44
1.5.1. In Vivo Electricity 44
1.5.2. The Methods of In Vitro Electrical Stimulation 47
1.5.2.1. Types of Electrical Stimulation 47
1.5.2.2. Methods of Delivering the Stimulus 47
1.5.2.2.1. Direct Coupling 47
1.5.2.2.2. Indirect Coupling 47
1.5.2.2.2.1. Capacitive Coupling 48
3
1.5.2.2.2.2. Inductive Coupling 49
1.5.2.3. The Parameters of Electrical Stimulation 50
1.5.3. The Cellular Effects of Electrical Stimulation 52
1.5.3.1. Intracellular Calcium 52
1.5.3.2. The Response of Cells to Weak Electric Fields 53
1.5.3.3. Growth Factors and Receptors under Electrical Stimulation 54
1.5.3.4. Similarities between Electrical and Mechanical Stimulation 54
1.5.3.5. The Mechanisms behind Galvanotaxis 55
1.5.3.6. Intracellular Signalling Pathways 55
1.5.3.7. Sensing in Excitable Cells 56
1.5.3.8. The Structural Effects of High Power Electric Fields 56
1.5.3.8.1. Electro-permeabilization 57
1.5.3.8.2. Electro-fusion 58
1.5.4. The Effects of Electrical Stimulation at the Tissue Level 59
1.5.4.1. Galvanotaxis 59
1.5.4.2. Enhanced Wound Healing 59
1.5.4.3. Improved Nerve Regeneration and Neural Tissue Engineering 60
1.5.4.4. Benefits for Bone 60
1.5.4.5. Effects on the Cardiovascular System 63
1.5.4.6. Skeletal Muscle Tissue Engineering 63
1.5.5. Stimulation through Conductive Scaffolds 65
1.5.5.1. Conductive Polymers 65
1.5.5.1.1. Polypyrrole 66
1.5.5.1.2. Polyaniline 67
1.5.5.1.3. Polythiophene Derivatives 68
1.5.5.2. Electrical Stimulation through the Scaffold 69
1.5.5.3. Further Approaches to Delivering an Electrical Stimulus
through a Biomaterial 70
1.5.5.3.1. Electrets 70
1.5.5.3.2. Piezoelectric Polymeric Materials 70
1.5.5.3.3. Photovoltaic Polymers 71
1.5.6. Future Possibilities in Electrical Stimulation 72
1.6. Bioreactors 74
1.6.1. Electrical Stimulation Bioreactors 75
1.6.1.1. Agarose Bridges 75
1.6.1.2. Bioreactor for Skeletal Muscle Tissue Engineering 76
1.6.1.3. Cardiac Muscle Bioreactor 76
1.6.1.4. Biphasic Current Stimulator 78
1.6.1.5. C-Pace Stimulators 78
1.7. Conclusions, Aims And Objectives 80
2. Chapter 2: Bioreactor Design 83
2.1. Introduction 84
2.2. A Bioreactor for Direct Electrical Stimulation 85
2.2.1. Materials and Methods Used in Building the Bioreactor 86
2.2.2. The Lessons Learned 86
2.3. The First Generation of Capacitive Bioreactors 88
2.3.1. The Bioreactor 88
2.3.2. Evaluation of the First Generation Bioreactor 91
4
2.4. The Second Generation 93
2.4.1. The Bioreactor 93
2.4.2. Evaluation of the Second Generation Bioreactor 93
2.5. The Third Generation Capacitive Bioreactor 95
2.5.1. The Bioreactor 97
2.5.2. Evaluation of the Third Generation Bioreactor 99
2.6. A New Approach to Isolating the Electrodes 100
2.6.1. Finite Element Method Simulations 101
2.6.1.1. Materials and Methods 101
2.6.1.1.1. Geometry 101
2.6.1.1.2. Material Properties 102
2.6.1.1.3. The Mesh 102
2.6.1.1.4. Simulation Parameters 102
2.6.1.2. Results and Discussion 103
2.7. The Fourth Generation Bioreactor 104
2.7.1. Materials and Methods 104
2.7.1.1. The Bioreactor 104
2.7.1.2. Biocompatibility Tests 106
2.7.1.2.1. Cell Culture 106
2.7.1.2.2. Cell Numbers 106
2.7.1.2.3. Metabolic Activity 106
2.7.1.2.4. pH Measurements 107
2.7.1.2.5. Statistical Analysis 107
2.7.2. The Iterative Steps of Improving the Design of the Upper Electrode 108
2.7.3. Improvements to the Auxiliary Components of the Bioreactor System 114
2.7.3.1. Signal Source 114
2.7.3.2. Cables 114
2.7.3.3. Stimulation Stage 115
2.7.3.4. Signal Recording 115
2.7.4. The Final Bioreactor System 116
2.7.5. Evaluation of the Fourth Bioreactor Design 119
2.8. Discussion 120
2.8.1. The Cause behind the Lower Cell Viability 120
2.8.2. Comparison of the Bioreactors 122
2.8.3. A Fifth Generation Bioreactor 123
2.8.4. Perfusion Concepts 125
2.8.4.1. Laminar Flow – Electrical Bioreactor 126
2.8.4.2. 3D Flow – Electrical Bioreactor 128
2.9. Conclusions 130
3. Chapter 3: Computer Simulations 131
3.1. Introduction 132
3.1.1. The Finite Element Method 134
3.1.2. An Introduction to the Computerised Electric Field Simulations –
Why Was COMSOL Multiphysics Chosen? 135
3.1.2.1. Computer Simulations in the MATLAB Environment 135
3.1.2.2. COMSOL Multiphysics 137
3.1.3. The Physics Background of the Electric Field Simulations 138
3.1.3.1. The Maxwell Equations 138
5
3.1.3.2. Electrostatics 139
3.1.3.3. The Absolute Permittivity 140
3.1.3.4. The Electric Field inside the Culture Medium 141
3.1.3.5. The Interpretation of the Results of COMSOL
Simulations – The Importance of the Boundary Conditions 143
3.2. Materials and Methods 145
3.2.1. Computer Simulations of the Electric Field inside the Bioreactor 145
3.2.1.1. The Comparison of the Different Bioreactor Designs 145
3.2.1.2. The Electric Field Strength in the Final Bioreactor
Design (Monolayer Cultures) 147
3.2.1.2.1. Simulation Parameters 147
3.2.1.3. The Electric Field in the Case of a 3D Scaffold 148
3.2.1.3.1. The Relative Permittivity of the Scaffold 148
3.2.1.3.1.1. Micro Computed Tomography of the
Spongostan scaffold 148
3.2.1.3.1.2. Microbalance Weight Measurements 149
3.2.1.3.2. The Electric Field Strength inside the Scaffold 149
3.2.1.3.3. The Electric Field at the Cellular Level 149
3.2.2. The Frequency Response of the Bioreactor 151
3.2.3. Signal Measurements 152
3.3. Results and Discussion 153
3.3.1. The Electric Field in the Four Different Bioreactor Generations 153
3.3.2. The Equation Describing the Electric Field Strength inside the
Fourth Generation Bioreactor 155
3.3.3. The Electric Field in the Case of a 3D Scaffold 158
3.3.4. The Frequency Response of the Bioreactor 163
3.3.5. Signal Measurements 164
3.3.6. Comparison with the Literature 166
3.3.7. Future Possibilities: Biological Cell – Electric Field Interaction 167
3.4. Conclusions 172
4. Chapter 4: In Vitro Experiments 175
4.1. Introduction 176
4.2. Materials and Methods 179
4.2.1. Cell Culture 179
4.2.1.1. Cell Revival 182
4.2.1.2. Sub-culturing 182
4.2.1.3. Cell Freezing 182
4.2.1.4. Differentiation Media 182
4.2.1.4.1. Osteogenic Medium 182
4.2.1.4.2. Adipogenic Medium 183
4.2.2. Electrical Stimulation 184
4.2.3. Experiments 185
4.2.3.1. Low Voltage Experiments 185
4.2.3.2. Expansion Optimisation 185
4.2.3.2.1. Proliferation Rate 186
4.2.3.2.1.1. Proliferation Rate during Expansion 186
4.2.3.2.1.2. Proliferation during Four Days of Culture 186
4.2.3.2.1.3. Cell Numbers after Fourteen Days of Culture 186
6
4.2.3.2.2. Differentiation Potential 187
4.2.3.3. Seeding Optimisation 187
4.2.3.4. High Voltage Experiments 189
4.2.3.4.1. Monolayer Cultures 189
4.2.3.4.2. Spongostan 3D Scaffold Cultures 189
4.2.4. Assays 191
4.2.4.1. PicoGreen DNA Assay 191
4.2.4.1.1. Storage of Monolayer Samples 191
4.2.4.1.2. Storage of Scaffold Samples 191
4.2.4.1.3. The Assay 191
4.2.4.2. Alamar Blue Metabolic Assay 192
4.2.4.3. Alkaline Phosphatase Assay 192
4.2.4.4. Gene Expression 192
4.2.4.4.1. Storage of Samples 192
4.2.4.4.2. RNA Isolation 193
4.2.4.4.3. cDNA Synthesis 193
4.2.4.4.4. Polymerase Chain Reaction 193
4.2.4.5. BMP-2 and BMP-7 Production 194
4.2.4.6. Oil Red O Staining 195
4.2.4.7. Optical Microscopy 196
4.2.4.8. Statistical Analysis 196
4.3. Results and Discussion 197
4.3.1. Low Voltage Experiments 197
4.3.1.1. Three Donor Repeat Experiments 199
4.3.1.1.1. Cell Numbers 199
4.3.1.1.2. Metabolic Activity 199
4.3.1.1.3. Alkaline Phosphatase Activity 201
4.3.1.1.4. Comparison of the Results 201
4.3.2. Investigating the Variation in hMSCs Behaviour 203
4.3.2.1. Optimised hMSC Culture Conditions 203
4.3.2.1.1. Proliferation Rate 204
4.3.2.1.2. Osteogenic Differentiation Potential 206
4.3.2.1.3. Adipogenic Differentiation Potential 208
4.3.2.2. Optimised Cell Seeding 209
4.3.3. High Voltage Experiments 213
4.3.3.1. The Effects of Electrical Stimulation on hMSC Proliferation 213
4.3.3.2. Osteogenic Differentiation 214
4.3.4. Discussion 215
4.3.4.1. Comparison with the Literature 219
4.3.4.1.1. Electric Field Strength 219
4.3.4.1.2. Frequency 219
4.3.4.1.3. Pulse Width 220
4.3.4.1.4. Summary 220
4.3.4.2. Future Possibilities 220
4.4. Conclusions 223
5. Overall Conclusions 225
5.1. The Conclusions of this Thesis 226
5.2. Future Work 228
7
References 230
A. Appendix 256
A.1. Alternative Bioreactor Designs 257
A.1.1. The Parallel Field Bioreactor 257
A.1.1.1. Introduction 257
A.1.1.2. The Design 259
A.1.2. The Direct Bioreactor 260
A.2. The Electrical Capacitance and Resistance of the Three Different
Bioreactors 262
A.2.1. The “Perpendicular” Bioreactor 264
A.2.2. The “Parallel” Bioreactor 264
A.2.3. The “Direct” Bioreactor 265
A.2.4. The Electrical Impedance of the Bioreactors 266
A.3. Simulations of the Electric Field inside the Parallel and Direct Bioreactors 268
A.3.1. Geometries 269
A.3.2. Results 270
A.3.2.1. The Relationship between Electrode Potential and
Current Density 270
A.4. Choosing the Stimulation Parameters for the Direct Bioreactor 273
A.5. Comparison of the Three Bioreactor Designs 275
Engineering Drawings 278
Publications 290
Final word count: 43,063 words
8
LIST OF FIGURES
Figure 1.1 – Incidence of fractures in England and Wales between 1988 and 1998. (Figure
reproduced from van Staa et al, 2001 [7] with the permission of the publisher.) 26
Figure 1.2 – The relationship between bone fractures, MSCs, tissue engineering and electrical
stimulation from the perspective of this study 28
Figure 1.3 – A summary of the tissue engineering process Primary or adult stem cells are
acquired from a patient or a donor. Alternatively embryonic stem cells (ESC) can be used. To
give primary cells ESC-like capabilities, pluripotency can be induced to create induced
pluripotent stem cells (IPSCs). Cells are expanded in culture and then placed into a bioreactor,
generally on a biomaterial scaffold. Chemical and physical stimuli are applied to promote
tissue generation. After weeks of tissue culture the generated tissue construct is implanted into
a patient. 33
Figure 1.4 - Scanning electric microscope images of woven (A) [69] and lamellar bone (B)
[70]. Note the much more organised structure of lamellar bone. Scale bars correspond to 5 μm
(left) and 1 μm (right). Images reproduced with permission of the publisher. 35
Figure 1.5 – The steps of osteogenic differentiation [71] 37
Figure 1.6 – The expression profile of ALP during the osteogenic differentiation
process 42
Figure 1.7 – The four different methods of electrical stimulation.
(Figure reproduced from [183].) 48
Figure 1.8 – The calcium mediated intracellular pathway for sensing electrical signals.
(Figure reproduced from [183].) 53
Figure 1.9 – Macrophages before (A) and after (B) electrofusion [219] (Image reproduced
with the permission of the publisher.) 58
Figure 1.10 – The structure of PPy [250, 251, 290-292]. Little information is known of the
structure of most conductive polymers. This is a result of the difficulty to find a solvent that
produces single crystals of the polymer and the degradation of the polymer in x-ray diffraction
studies [291, 292]. Figure reproduced from Balint et al, 2014 [254]. 67
Figure 1.11 – The structure of PANI [252, 302-305]. Figure reproduced from Balint et al,
2014 [254]. 68
Figure 1.12 – The structure of PEDOT [310].
Figure reproduced from Balint et al, 2014 [254]. 68
Figure 1.13 – The electrical stimulation bioreactors described in the literature: A – The
agarose bridge configuration [337], B – The Donelly bioreactor [338], C – Cardiac muscle
bioreactor [339], D – Biphasic current stimulator [25], E – The C-Pace system (Images were
reproduced with the permission of the publishers. Image of the C-Pace system is the property
of IonOptix LLC.) 77
9
Figure 2.1 – The initial concepts: Sketches of the first ideas for the direct stimulation
bioreactor 85
Figure 2.2 - The initial concept and a top view of the complete
direct electrical bioreactor 86
Figure 2.3 – Computer simulation of the electric field strength between two electrodes
showing an example of the edge effect. The electric field strength is higher (indicated by the
red colour) at the edge of the electrode. 89
Figure 2.4 – Exploded view of the first generation capacitive bioreactor with the stimulation
stage. 1 – Copper wires connecting the upper electrodes to one of “poles” of the stimulation
stage, 2 – Polystyrene plate lid, 3 – Polypropylene sections, 4 – Copper disc upper electrodes,
5 – Polystyrene 6-well plate, 6 – Copper counter electrode connected to the other “pole” of the
stimulation stage, 7 – Stainless steel pole on the stimulation stage,8 – PTFE stimulation
stage 90
Figure 2.5 – The stimulation stage 91
Figure 2.6 – Sketch of the medium “sticking” to the electrodes 92
Figure 2.7 – A completed first generation bioreactor from above (left) and the copper
electrodes on the plate lid from below (right). 92
Figure 2.8 – The second generation bioreactor with the two large rectangular electrodes. 1 – 6-
well plate between the electrodes, 2 – Rectangular copper electrode. 93
Figure 2.9 – Exploded view of the second generation bioreactor. 1 – Polystyrene 6-well plate,
2 – Rectangular copper electrode. 94
Figure 2.10 – An uneven culture medium surface can be avoided by submerging the
electrodes 95
Figure 2.11 – Engineering drawing (left), 3D model (middle) and photography (right) of a
third generation bioreactor upper electrode. 1 – Stainless steel machine screw, 2 – Epoxy
embedding material, 3 – Stainless steel wire: Allows the checking of the electrical
connectivity to the electrode disc even after embedding. 96
Figure 2.12 – Photograph of a third generation bioreactor lid upside down. 1 – PTFE lid, 2 –
Third generation electrode assembly. 97
Figure 2.13 – Exploded view of the third generation bioreactor lid with a 6-well plate bottom.
1 – M4 stainless steel nut 2 – Stainless steel spring washer 3 – PTFE bioreactor lid 4 – Third
generation electrode assembly 5 – Polystyrene 6-well plate bottom. 98
Figure 2.14 – Concept drawings of PTFE cup (A), the PTFE washer (B), the stainless steel
electrode (C) and the electrode assembly (D). 3D models were created in the commercial
software SolidWorks 2008. 100
Figure 2.15 - The geometry used to model the PTFE washer (A) and cup (B) 101
10
Figure 2.16 – 3D models with meshes in COMSOL Multiphysics. A –PTFE washer, B – PTFE
cup 102
Figure 2.17 – Results of the FEM simulation showing the change of the diameters as a
function of temperature. The blue patterned area indicates the overlap between the external
diameter of the PTFE washer and the internal diameter of the PTFE cup. 103
Figure 2.18 – Engineering drawing of the “electrode bridge” 104
Figure 2.19 – Exploded view of the fourth generation bioreactor lid with a 6-well plate bottom.
1 – M4 stainless steel nut 2 – Stainless steel spring washer 3 – Electrode bridge 4 – PTFE
bioreactor lid 5 – Upper electrode 6 – 6-well plate bottom 105
Figure 2.20 – Engineering drawings (left), 3D models (middle) and photographs (right) of the
various iterations of the electrode assembly. A – The original PTFE concept, B – PTFE cups
with raised walls, C - PTFE cups with raised walls and four nebs, D – Bare stainless steel
electrodes 109
Figure 2.21 – Cell numbers (n=3) in a normal 6-well plate (Controls), a bioreactor with bare
stainless steel electrodes (Steel) and with the PTFE cups (PTFE) at
Day 8 (* = p<0.05). 111
Figure 2.22 – Cell numbers (left) and metabolic activity (right) (n=6) in bioreactors with bare
stainless steel electrodes (steel) compared to normal 6-well plates (Controls)
(* = p<0.05). 111
Figure 2.23 – Raising the electrodes to be 6 mm rather than 1 mm from the culture medium
surface (assuming 2 ml of medium) helps avoid any contact between the electrodes and the
culture medium. 112
Figure 2.24 – Cell numbers (left) and metabolic activity (right) (n=6) measured in bioreactors
with raised stainless steel electrodes (Steel) compared to normal 6-well plates (Control) (* =
p<0.05). 112
Figure 2.25 – A comparison of the pH of culture media (n=6) from normal 6-well plate and
bioreactor cultures. 113
Figure 2.26 – The final design of the capacitive electrical bioreactor. Image shows the
bioreactor lid upside down. 1 – PTFE lid, 2 – Stainless steel electrodes. 113
Figure 2.27- The new bottom electrode plates and the cables inside the incubator 115
Figure 2.28 – Photograph (left) and schematic (right) of the bioreactor system. Visible on the
photograph are the low voltage amplifier (1), the function generator (2), the high voltage
amplifier (3), fourth generation bioreactor lids (4) and the incubator (5). 116
Figure 2.29 – 3D view of the assembled fifth generation bioreactor in SolidWorks
2008. 124
Figure 2.30 – Exploded view of the fifth generation capacitive bioreactor concept 125
11
Figure 2.31 – A 3D rendered model of the Laminar flow – Electrical bioreactor
concept 126
Figure 2.32 – Exploded view of the laminar flow – electrical bioreactor. (1 - Modified 6-well
plate, 2 – Bioreactor lid, 3 – Upper electrodes, 4 – Bottom electrode, 5 – Perfusion
tubing) 127
Figure 2.33 – Top view of the bioreactor lid (left) and of the modified 6-well plate (right). A –
Perfusion inlets and outlets, B – Place for the electrodes, C – Flow and culture chamber, D –
Raised areas to optimally direct the flow towards the cells. 127
Figure 2.34 – Rendered model of the 3D flow – electrical stimulation bioreactor 128
Figure 2.35 – Top view of the modified 6- well plate. The arrows indicate the perfusion inlets
and outlets. 129
Figure 2.36 – Exploded view of the 3D flow –electrical bioreactor. 1 – Modified 6-well plate,
2 – Upper electrodes with seals, 3 – Bioreactor lid, 4 – Metal bridge connecting all the
electrodes, 5 – Bottom electrode, 6 – Perfusion tubing 129
Figure 3.1 – The same electrode potential difference (15V in the above example) will result in
a different electric field strength depending on the distance and the material between the
electrodes. The colour legend indicates electric field strength from 0 (blue) to approx. 35 V/m
(red). 132
Figure 3.2 – A sphere discretised into a mesh of a finite number of nodes and elements 134
Figure 3.3 – The charge density on two round electrodes is broken up into finite charges in
MATLAB. Dimensions are in meters, while the scale indicates charge in coulombs. 135
Figure 3.4 – The effect of polarisation on the electric field strength inside the culture medium.
A – If there is no electric field present the water molecules orient themselves randomly. B –
Once an electric field is applied the water molecules will start to orient themselves – the
oxygen facing the positive, while the hydrogen atoms facing the negative electrode. C – This
electric field induced orientation is called the polarisation of the material. Its overall effect can
be viewed as the creation of a positive charge density on the negative electrode facing side of
the material and the creation of a negative charge density on the positive electrode facing side
of the material . D – The two charge denisties generate an antagonist electric field that acts to
weaken the electric field created by the electrodes. A cell placed into the culture medium will
experience the sum of these two fields, which will always be weaker then the one generated by
the electrodes alone. (Ecell – electric field strength exprienced by a cell, E0 – electric field
strength generated by the electrodes, Ep – electric field strength generated by
the polarisation.) 141
Figure 3.5 – The geometries used for the models of the first (A), second (B), third (C) and
fourth (D) generation bioreactors 146
Figure 3.6 – The parameters of the simulations 147
Figure 3.7 – The XRadia Versa XRM-500 system 148
12
Figure 3.8 – The geometry used to simulate the electric field with a Spongostan scaffold in the
bioreactor 149
Figure 3.9 – The model of the bioreactor system in MULTISIM taking into account the
electrical resistance and capacitance of the bioreactor itself and the coaxial cable. 151
Figure 3.10 – The oscilloscope used for the measurements 152
Figure 3.11 – The electric field strength in the four different generations of the bioreactor as a
function of electrode potential difference 153
Figure 3.12 - The electric field strength in the first (A), second (B), third (C) and fourth (D)
generation bioreactor. The colour legend indicates the electric field strength from low (blue) to
relativel high (red). 154
Figure 3.13 – The electric field strength experienced by the cells (Ecells) at 1 V electrode
potential difference as a function of the distance of the electrodes (H) and percentage of this
distance that was filled up by culture medium. 155
Figure 3.14 – The graphical user interface of the electric field strength calculator 156
Figure 3.15 – The reconstructed volume from the 4x magnification scan (left) showing the
gelatine in blue and the empty pore space in red (scale bar corresponds to 500 μm). The image
on the right shows a cross section of the volume from the 20x magnification scan displaying
the structure of the gelatine walls in dark grey (scale bar corresponds to 50 μm). 158
Figure 3.16– A histogram of the different pore sizes in the scaffold 158
Figure 3.17 – The electric field strength experienced by cells within the Spongostan scaffold
as a function of electrode potential difference 159
Figure 3.18 – Image showing the electric field strength in a part of the modelled region. Blue
colour corresponds to low, while red indicates high electric field strength. Note the lighter blue
areas “left and right” and the dark blue areas “above and below” the regions of
gelatine. 160
Figure 3.19 – A graphical summary of the phenomenon observed around gelatine regions
within the scaffold. 161
Figure 3.20 – The effect of the shape and relative permittivity of an object on the surrounding
electric field. The electric field strength around objects of different shape (A). The electric
field strength and field lines around disks with high (B and D) and low (C and E) relative
permittivity. Colour legend corresponds to electric field strength with blue indicating low and
red indicating high values. 162
Figure 3.21 – The Bode-diagram of the electrical bioreactor showing the magnitude (A) and
phase angle (B) of the signal at the bioreactor compared to the output of the signal source as a
function of frequency. 163
Figure 3.22 – The distortions observed with the high-voltage amplifier 164
13
Figure 3.23 – Examples of the minor distortion in the case of some of the low voltage pulses
(A), the distorted shape of 1 μs pulses (B), the overshooting of some of the signals (C) and the
noise present in 1999 μs pulse width signals (D). 165
Figure 3.24 - A graphical explanation of why it is easier to acquire sufficient slices if the cells
are rounded (3D scaffold) (A) compared to when they are more spread out (monolayer cells)
(B). 168
Figure 3.25 – The different angles at which the cells were scanned 169
Figure 3.26 – The process of creating a simulation based on graphical information from an
image stack 170
Figure 3.27 – The path from user input to “cell experience” 173
Figure 4.1 – The four regimes used in this study 177
Figure 4.2 – A 1 cm3 Spongostan scaffold 189
Figure 4.3 – Cell numbers (n=6) in hMSCs cultures after 7 days of 1h/day
electrical stimulation 198
Figure 4.4 – The effect of electrical stimulation on the metabolic activity of hMSCs (n=6) after
7 days of 1h/day stimulation (“*” = p<0.05 compared to Control samples) 198
Figure 4.5 – The effect of electrical stimulation on the alkaline phosphatase activity of hMSCs
(n=6) after 7 days of 1h/day stimulation (“*” = p<0.05 compared to Control samples) 198
Figure 4.6 – The effect of 1 and 10 μs stimulation on the cell numbers, metabolic activity and
alkaline phosphatase activity of hMSCs from three different donors (n=6). (“*” indicates
p<0.05) 200
Figure 4.7 – Sample with hMSCs displaying the spread out, rhomboidal SR morphology (A)
compared to spindle-like, small RS cells (B). The contrast of the images have been modified in
order to enhance visibility. 204
Figure 4.8 - A comparison of the fold increase (A) and fold increase per day (B) in Protocol A
and Protocol B cultures. (“*” indicates p<0.05) 204
Figure 4.9 – Cell numbers in Protocol A and B cultures during four days of expansion
(n=6). 205
Figure 4.10 – Cell numbers in Protocol A and Protocol B samples after 14 days in various
differentiation media (n=4, “*” indicates p<0.05) 206
Figure 4.11 – The fold expression of ALPL, osterix, collagen type I, osteocalcin and
osteopontin mRNA in Protocol A and B cells cultured in osteogenic and growth medium for
14 days (n=4, “*” indicates p<0.05) 207
Figure 4.12 – Alkaline phosphatase activity of Protocol A and Protocol B hMSCs after 14
days in osteogenic and growth medium (n=4, “*” indicates p<0.05). 208
14
Figure 4.13 – Adiponectin (left) and leptin (right) mRNA expression in Protocol A and
Protocol B samples cultured in growth and adipogenic medium for 14 days (n=4, “*” indicates
p<0.05). 208
Figure 4.14 – The amount of lipid per cell in Protocol A and Protocol B samples after days in
adipogenic medium (n=4, “*” indicates p<0.05). 209
Figure 4.15 – The comparison of cell numbers in two 6-well plates after 6 days in culture
(n=6, “*” indicates p<0.05) 210
Figure 4.16 – Cell numbers in three 6-well plates seeded with three different cell
concentrations at day 0 (the day of plating) and at day 6 (after 6 days in culture). (n=6, “*”
indicates p<0.05) 211
Figure 4.17 – Cell numbers (n=6) in monolayer cultures after 7 days of stimulation. (“*”
indicates p<0.05) 213
Figure 4.18 – Cell numbers (n=6) in Spongostan scaffolds after 14 days of stimulation 214
Figure 4.19 – Alkaline phosphatase (left) and BMP-2 (right) mRNA levels after 7 sessions of
stimulation in monolayer samples (n=6). 214
Figure 4.20 – BMP-2 (left) and BMP-7 (right) production per day after 14 sessions of
stimulation in Spongostan scaffold samples (n=6). 215
Figure 4.21 – A possible explanation of the differences observed between the effect of
electrical stimulation in monolayer and scaffold cultures based on the mechanotransduction
pathway. 217
Figure A.1 – Sketch showing the difference between field lines perpendicular and parallel to
the cell monolayer 257
Figure A.2 – The minimum distance between the two electrodes is the smallest, if the
electrodes are placed below and above the monolayer (left) rather than on its two sides (right).
At the same electrode potential this will generate a much stronger electric field
strength. 257
Figure A.3 – 3D rendered image of the parallel plate bioreactor 259
Figure A.4 – Exploded view of the parallel field bioreactor. 1 – Polystyrene plate top, 2 –
Rectangular glass coverslips, 3 – PTFE bioreactor bottom, 4 – Electrodes 259
Figure A.5 – Direct parallel stimulation 260
Figure A.6 – 3D rendered image of the parallel plate bioreactor 260
Figure A.7 – Exploded view of the contact bioreactor 1 – Polystyrene plate top, 2 – Electrodes,
3 – Rectangular glass coverslips, 4 – PTFE bioreactor bottom 261
Figure A.8 – Schematic representation of the circuit with the signal generator supplying one
bioreactor with 6 wells. Each well has an associated resistance (Rn) and a capacitance (Cn)
15
value. Re1 and Re2 represent the resistance of the electrodes, while Rcable is the resistance of
the cables leading to and away from the bioreactor. 262
Figure A.9 – Schematic representations of one well of the “perpendicular” (left), “parallel”
(centre) and “direct” (right) bioreactor. 263
Figure A.10 – A drawing of the geometry used for the parallel bioreactor simulations
(dimensions are in mm) 269
Figure A.11 – A drawing of the geometry used for the direct bioreactor simulations
(dimensions are in mm) 269
Figure A.12 – The electrical field strength in the parallel and direct bioreactors as a function of
electrode potential difference 270
Figure A.13 – The electrical current density inside the direct bioreactor as a function of
electrode potential difference 271
Figure A.14 – A slice taken in the centre of the bioreactor displaying current density at and
electrode potential difference of 450 V. Blue indicates low, while red indicates high current
density. The section corresponding to the culture medium displays an even turquoise colour,
showing that the current density is distributed across the culture medium
in a homogenous manner. 272
Figure A.15 – A graphical summary of the various parameter ranges for the direct
bioreactor 274
Figure A.16 – A graphical comparison of the three bioreactors. The light green triangle
indicate the range where the perpendicular bioreactor can be set depending on electrode
position (“D” = the distance between the top of the culture medium and the upper electrode
assuming 2ml medium.) The blue triangle (indicated by the blue arrow) is the range where the
direct bioreactor can be used with the current signal source. 275
16
LIST OF TABLES
Table 1.1 – The number of tissue engineering related clinical trials found in various registries
around the world (Data last retrieved on 17/02/2014) 31
Table 1.2 – The effect of various electric field strengths 50
Table 1.3 – The effect of various current densities upon osteoblasts and mesenchymal stem
cells 51
Table 1.4 – A summary of bone related in vitro studies carried out with the three different
methods of stimulation 62
Table 1.5 – A list of conductive polymers and their abbreviations [250-253]. Table reproduced
from Balint et al, 2014 [254]. 65
Table 2.1 – The criteria of an ideal ES bioreactor – First generation bioreactor 91
Table 2.2 – The criteria of an ideal bioreactor – Second generation bioreactor 93
Table 2.3 – The criteria of an ideal ES bioreactor – Third generation bioreactor 99
Table 2.4 – The material properties used to model PTFE [340] 102
Table 2.5 – The criteria of an ideal bioreactor – Fourth generation bioreactor 119
Table 3.1 - The relative permittivity of the various materials used in the COMSOL models
[340] 145
Table 3.2 – The results of the microbalance weight measurements 159
Table 4.1 – A brief summary of the electrical stimulatory regimes used in this study 177
Table 4.2 – List of donors 179
Table 4.3 – The electrical stimulatory regimes applied in this study. * - Electrical potential
difference as set on the signal source. ** - Electric field strength only drops to 4 V/m between
pulses. *** - Electric field strength only drops to 44.1 V/m between pulses 184
Table 4.4 – The steps of the monolayer electrical stimulation experiments 185
Table 4.5 – The three examined seeding densities 187
Table 4.6 – The steps of the 3D scaffold experiments 190
Table 4.7 – A list of the genes assayed in this study 194
17
Table 4.8 – A list of the reagents used in the BMP-2 and BMP-7 ELISA assays. All reagents
were purchased from R&D Systems Inc. with the exception of the Tween 20 and
the PBS. 195
Table 4.9 – The metabolic activity of stimulated cells from the three donors compared to
controls 199
Table 4.10 – A graphic summary of the results gained in the two experiments 201
Table A.1 – The parameters used in the calculations [340, 417] 263
Table A.2 – Table A.2 – The electrical conductivity values used in the simulation
[340, 417] 268
Table A.3 – Examples of current densities applied for electrical stimulation 273
18
LIST OF ABBREVIATIONS
2D – Two dimensional
3D – Three dimensional
AC – Alternating current
ALP – Alkaline phosphatase
ALPL – Alkaline Phosphatase (gene)
ANSI - American National Standards
Institute
BMP – Bone morphogenic protein
BMP-2 – Bone morphogenic protein -2
BMP-7 – Bone morphogenic protein -7
BSP – Bone sialoprotein
Cbfa1 - Core binding factor alpha 1
CC – Capacitive coupling
cDNA – Complementary deoxyribonucleic
acid
DC – Direct current
DMSO - Dimethyl Sulfoxide
DNA - Deoxyribonucleic acid
ECM – Extracellular matrix
EGF – Endothelial growth factor
ELISA – Enzyme-linked immunosorbent
assay
ERK – Extracellular signal regulated kinases
ES – Electrical stimulation
FBS - Foetal bovine serum
FEM – Finite element method
FGF – Fibroblast growth factor
HA – Hydroxyapatite
hMSCs – Human mesenchymal stem cells
IC – Inductive coupling
MAP – Mitogen activated protein
MicroCT – Micro computed tomography
mRNA – Messenger ribonucleic acid
MSC – Mesenchymal stem cell
OC - Osteocalcin
ON - Osteonectin
OP – Osteopontin
PANI – Polyaniline
PBS - Phosphate buffered saline
PEDOT - Poly(3,4-ethylenedioxythiophene)
PEMF - Pulsed electromagnetic field
stimulation
PPi - Pyrophosphate
PPy – Polypyrrole
PTFE – Polytetrafluoroethylene
PTh – Polythiophene
qRT-PCR – Quantitative reverse
transcription - polymerase chain reaction
RM – Regenerative medicine
RS – Rapidly self-renewing
Runx2 - Runt-related transcription factor 2
SR – Slow replicating
TE - Tris-EDTA
TENS - Transcutaneous nerve stimulation
TEP - Trans-epithelial potential
TGF-β – Transforming growth factor - β
TNAP - Tissue-nonspecific alkaline
phosphatase
VEGF – Vascular endothelial growth factor
19
ABSTRACT
Richard Balint
Electrical Bioreactor Design for Tissue Engineering
Doctor of Philosophy Dissertation in Biomedical Materials
University of Manchester, 2014
Bone fractures are a major health issue, causing severe pain and disability to millions of
patients. Novel treatments based on tissue engineering, the creation of tissue implants
through the combination of cells and biomaterials, are currently explored, promising faster
and better healing. Electrical stimulation is known to be beneficial for the healing of bone
and is applied regularly in the clinical setting to treat fractures. It is possible that this
electrical modality could also be used to augment the tissue engineering process,
improving the quality of the produced implants. In order to investigate this, the effect of
electrical stimulation on human mesenchymal stem cells, an important bone tissue
engineering cell type, was examined. An autoclavable, reusable, reliable and robust
electrical bioreactor system was designed and built, that allows the delivery of
homogenous capacitive stimulation in both the monolayer and 3D settings. The physical
aspects of the interaction of cells and the electric field generated in the bioreactor was
examined through computer simulations and signal measurements. In vitro experiments
have been carried out demonstrating the ability of electrical stimuli to influence
mesenchymal stem cell behaviour. Important experience has been gained on the principles
governing the effects of electrical stimulation, emphasising the significance of electric field
strength, culture condition, cell type, treatment duration, and signal waveform in defining
the outcome of the stimulation. The knowledge gained in this study will help develop
electrical stimulation into a truly useful tool for bone tissue engineering.
Keywords: bone, tissue engineering, bioreactor, electrical stimulation, computer
simulations
Lay Abstract
In this study it was investigated whether electrical signals could be used to help create
better quality artificially grown bone tissue. Such artificial bone tissue could be used as
implants to treat, for example, bone fractures that do not heal without specialist medical
intervention. A device, a bioreactor, was created that allows artificial tissue to be grown,
while being exposed to an electric field. What sort of electric field the cells in the tissue
experience was explored using computer models. Experiments using the bioreactor show
changes in the behaviour of cells with the electric field, however further studies are
necessary in order to enable electrical signals to be used in a truly effective way for the
creation of artificial bone tissue.
20
DECLARATION
Section 1.5 of Chapter I, titled “Electrical Stimulation”, is based on a literature review written by
the author as part of a Master of Science dissertation at Keele University (“Influence of the
‘PhyBack System’ on Primary Human Mesenchymal Stem Cell Activity”), however the text has
been extensively re-written, expanded and improved upon. No other portion of the work referred to
in the thesis has been submitted in support of an application for another degree or qualification of
this or any other university or other institute of learning.
COPYRIGHT STATEMENT
I. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain
copyright or related rights in it (the “Copyright”) and s/he has given The University of Manchester
certain rights to use such Copyright, including for administrative purposes.
II. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be
made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and
regulations issued under it or, where appropriate, in accordance with licensing agreements which
the University has from time to time. This page must form part of any such copies made.
III. The ownership of certain Copyright, patents, designs, trade marks and other intellectual
property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for
example graphs and tables (“Reproductions”), which may be described in this thesis, may not be
owned by the author and may be owned by third parties. Such Intellectual Property and
Reproductions cannot and must not be made available for use without the prior written permission
of the owner(s) of the relevant Intellectual Property and/or Reproductions.
IV. Further information on the conditions under which disclosure, publication and
commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions
described in it may take place is available in the University IP Policy (see
http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=487), in any relevant Thesis restriction
declarations deposited in the University Library, The University Library’s regulations (see
http://www.manchester.ac.uk/library/aboutus/regulations) and in The University’s policy on
Presentation of Theses.
21
ACKNOWLEDGEMENTS
I would like to thank my supervisors, Professor Sarah H. Cartmell and Dr Nigel J. Cassidy for their
guidance and support in the last four and a half years. Without them I wouldn’t be where I am
today.
I would like to express my gratitude to Dr Araida Hidalgo, Dr James E. A. Dugan, Ms Naa-Dei
Nikoi, Dr Deepak Kumar and Dr Ian Wimpenny for their friendship and advice during my PhD and
during the writing of this thesis. Special thanks goes out to Mr Samuel Jackson for being a true
friend and for all the emergency meetings in the local drinking establishement.
Finally, I would like to thank my family. They are the foundation upon which this thesis was built.
...and thanks for all the fish!
...
22
THE AUTHOR
Richard Balint
A brief summary of the author’s qualifications and main academic achievements during this Doctor
of Philosophy degree.
Qualifications
2011 – Current
(2014)
PhD in Biomedical Materials (final year)
School of Materials, University of Manchester, United Kingdom
2010 MSc in Biomedical Engineering - Distinction
Keele University, United Kingdom
2005 – 2009
BSc in Mechatronics Engineering (Final mark: grade IV - 70-86%)
Budapest University of Technology and Economics (BME),
Faculty of Mechanical Engineering (GPK), Hungary
Employment
2011 – Current
(2014)
Technical Assistant
Henry Moseley X-ray Imaging Facility, University of Manchester
Professional and Leadership Activities
2012 – Current
(2014)
Committee member (Postgraduate Representative) of the Tissue and Cell
Engineering Society UK (TCES)
2012 – 2014
Director of the Biomaterials Discussion Group Meetings, University of
Manchester
Professional Membership
- Tissue and Cell Engineering Society UK, 2011 - Current
- Tissue Engineering and Regenerative Medicine International Society, 2011 - Current
- European Calcified Tissue Society, 2011 – 2012
Awards
- School of Materials – Postgraduate Student of the Year 2014, nominated for Faculty level
award
- Tissue and Cell Engineering Society Travel Award, 2011 and 2012
- Public Engagement Best Presentation Winner, 2011
23
List of Peer-reviewed Publications
1. Balint R and Cartmell SH. A technical note on the culture of human mesenchymal stem
cells. (Manuscript prepared.) To be submitted to the Journal of Regenerative Medicine and
Tissue engineering, 2014
2. Balint R, Cassidy NJ, Cartmell SH. Conductive polymers: Towards a smart biomaterial for
tissue engineering. Acta Biomaterialia, 10, 2341, 2014
3. Shearer T, Rawson S, Castro SJ, Balint R, Bradley RS, Lowe T, Vila-Comamala J, Lee
PD, Cartmell SH. X-ray computed tomography of the anterior cruciate ligament and
patellar tendon. Muscles, Ligaments and Tendons, 2014 (In Press)
4. Balint R, Cassidy NJ, Hidalgo-Bastida LA, Cartmell S. Electrical stimulation enhanced
mesenchymal stem cell gene expression for orthopaedic tissue repair. J. Biomater. Tissue
Eng., 3, 212, 2013
5. Balint R, Cassidy NJ, Cartmell SH. Electrical stimulation: A novel tool for tissue
engineering. Tissue Engineering: Part B, 19, 48, 2013
6. Rupani A, Balint R, Cartmell SH. Osteoblasts and their applications in bone tissue
engineering. Cell Health and Cytoskeleton, 2012, 49, 2012
Conferences
- Oral presentations at international (TERMIS 2011, TERMIS 2012, TERMIS 2013) and
national (TCES 2012) conferences.
- Poster presentations at the TCES 2011, TCES 2013 and TCES 2014 national conferences.
- Co-chair at the TERMIS international conference in 2011 and 2012.
- Co-chair at TCES national conference in 2011, 2012 and 2014.
24
Chapter I
Introduction
“Facts are the air of scientists. Without them you can never fly.”
Linus Pauling...
25
1.1 THE CLINICAL BACKGROUND
Bone fractures are a major health issue. Around the world every year 2.4 people out of a hundred
suffer a fracture [1]. For the European Union in 2010 this meant 3.5 million new fractures [2]. In
the same year the associated economic cost was estimated to be €37 billion [2], out of which the
UK’s share was £3.5 billion [3]. Therefore, fractures are not only the cause of severe pain and
disability to millions of patients, but also have a huge socioeconomic impact [1-6].
Many of the fractures are associated with patients suffering from osteoporosis [2, 5]. This disease
causes patients’ bone to be more fragile, hence predisposing them to fractures [2, 5]. Osteoporosis
also impairs the bone healing process after the trauma, making the treatment of broken bones a
more difficult, lengthy and expensive task [2, 5]. 27.5 million people in Europe are believed to
have osteoporotic bones [2]. Osteoporosis is disease mainly (but not exclusively) affecting the
elderly and in particular women beyond their 50s [2, 5, 7]. The population level effect of this
disease (Figure 1.1) is that the lifetime risk of suffering a fracture for women in their 50s is 53.2%
(for men this is only 20.7%) [7]. This means that a 50 year old woman has more than 50% chance
of suffering a fracture during the remainder of her life. In other terms, more than half of the women
above 50 will suffer a fracture. As the proportion of elderly in the world’s population is predicted
to double by 2050, dealing with freactures will become a more and more serious socioeconomical
problem [8].
Treatment of a fracture is especially difficult in the cases where it pairs with a non-union or a large
size defect. Non-unions are broken bones that fail to regenerate themselves, and are present in
approximately 10% of all fractures [1, 9, 10]. For some tibial fractures this is as high as 50% [10].
Large or critical size defects are defined as extensive bone loss suffered as a result of trauma or
disease that prevents spontaneous healing simply due to its large size [9, 11, 12]. Severe pain, loss
of function, reduced work capability and an overall reduced life quality together with a high socio-
economical cost are associated with both of these situations [1, 4, 9, 10].
26
The current “gold standard” clinical treatment is bone graft transplantation [4, 13]. However, the
use of autologous grafts is hampered by donor site morbidity, the limited availability of tissue and
lengthened operation time, while allografts are difficult to process and carry an additional risk of
infection [4, 6, 14]. Although there are novel treatments currently employed, such as the use of
BMP-7 to promote bone regeneration [10], there still is not a truly effective treatment in existence
[9, 13].
Figure 1.1 – Incidence of fractures in England and Wales between 1988 and 1998. (Figure reproduced
from van Staa et al, 2001 [7] with the permission of the publisher.)
In summary, bone fractures are a widespread medical problem that have a significant impact on the
quality of life of patients and have a huge socioeconomic cost. Fractures especially affect the
elderly who have already impaired bone healing capabilities. Further complications, such as non-
unions and large size defects, make the healing of the fractures an even more difficult task.
One of the emerging therapies promising to overcome these problems is tissue engineering, the
creation of tissue replacements by combining living cells with biomaterials [15]. However, despite
the promising animal and clinical trials, the desired quality of the tissue engineered constructs has
27
not been achieved yet. Problems associated with controlling the behaviour of the cells, nutrient
delivery/waste removal, the achievable construct size, the time necessary to generate these and
their in vivo integration, together with the issues with the structure and functionality of the end-
product, have limited the use of tissue engineered products and are still to be overcome [4, 6, 12,
16-18].
Electrical stimulation, as a novel tool in musculoskeletal tissue engineering, has the potential to be
a significant step towards augmenting the tissue engineering process and bringing these novel bone
implants closer to widespread clinical application.
Over the past decades various electrical/electromagnetic stimuli have been applied successfully to
promote bone growth in both cell culture and animal model experiments and are now utilized in
several orthopedic, dental and maxillofacial applications; such as lumbar spine fusions, treating
osteoporosis, osteoarthrosis, normal and non-union fractures, and promoting the integration of
implanted biomaterials [19-21]. In a study involving 34 patients with tibial non-unions, 89% of the
patients treated with electricity achieved union compared to only 50% of the control group,
showing that electrical stimulation has a significant, beneficial effect on bone healing [22]. In a
recent review of 49 randomised trials on the effects of ES on long bone fracture healing, it was
concluded that there is consistent evidence that this modality has positive effect upon the repair
process [23]. Findings of in vitro studies on osteoblasts and mesenchymal stem cells (MSCs) show
this modality’s ability to enhance proliferation; BMP-2, BMP-4, TGF-β1 and VEGF expression;
ALPL activity and ECM deposition [20, 24-28].
Although the innovative approach of electrical stimulation has proven itself in both the laboratory
and clinical setting as described above, this technique has not been considered yet for bone tissue
engineering despite its great potential. In the literature review conducted for this study only seven
articles dealt with electrical stimulations’s effect on MSCs, demonstrating the lack of knowledge of
this type of application on this important tissue engineering cell type.
28
1.2 THE HYPOTHESIS
Bone fractures are a major health issue. TE is one of the approaches promising to alleviate
this problem.
Electrical stimulation is also used to treat fractures and is especially useful in the cases of
non-unions and large size defects. This implies that ES has an effect beneficial for bone
formation. This was confirmed in laboratory and clinical studies.
Although there have been only seven studies conducted thus far in connection with MSCs
and ES, the findings suggest that ES can influence the behaviour of these adult stem cells
in a way that can be useful for TE.
Therefore there is strong evidence that that ES could become a very useful modality for bone TE
(Figure 1.2).
The aim of this study is:
To develop electrical stimulation into an effective tool for the engineering of tissue, with enhanced
bone formation from bone-marrow derived Mesenchymal Stem Cells being in the main focus.
Figure 1.2 – The
relationship between bone
fractures, MSCs, tissue
engineering and electrical
stimulation from the
perspective of this study
29
In the following sections of this chapter the discipline of tissue engineering and bone as a tissue
will be discussed. This will be followed by an examination of the literature on electrical stimulation
and bioreactors, devices for the delivery of stimulation to cells and tissue, in order to understand
how to best approach the goal of this study. At the end of this chapter the objectives of this thesis
will be set out.
30
1.3 AN INTRODUCTION TO THE DISCIPLINES OF
REGENERATIVE MEDICINE AND TISSUE ENGINEERING
Modern molecular medicine has a great limitation: Once tissue or organ function is lost, it cannot
be returned [29]. To challenge this limitation a new approach, that of regenerative medicine, was
brought to existence during the 1980s [30]. Regenerative medicine (RM) can best be defined as
“process for replacing or regenerating cells, tissues or organs, to restore or establish normal
function” [31]. The emphasis is on regeneration, “the process by which lost specialised tissue is
replaced by the proliferation of undamaged specialised cells” [29]. Only a few tissues of the
human body, for example liver, are capable of regeneration under normal circumstances [32]. Most
tissue types can only “repair”. Repair is undertaken through granulation and, later on, scar tissue
formation. In contrast to regeneration, repair does not return tissue function once lost [32]. RM is a
broad field that has come to encompass (amongst others) cell based and genetic therapies and tissue
engineering.
The first truly tissue engineered construct has been attributed to Professor Eugene Bell from the
Massachusetts Institute of Technology. In collaboration with a pharmaceutical company, Meadox
Medicals [33], Professor Bell created an artificial skin replacement by seeding epidermal cells onto
fibroblasts cast within collagen lattices [34]. This artificial graft was then surgically placed onto
wounds of patients. The artificial graft successfully integrated, became vascularised, and inhibited
wound contraction [34].
Following Professor Bell’s groundbreaking work, through the efforts of the great pioneers of tissue
engineering, Robert Langer, Anthony Atala, Joseph and Charles Vacanti, tissue engineering has
become a widely researched discipline with thousands of publications and many clinical trials
(Table 1.1). A search on the website pubmed.gov using the term “tissue engineering” produced
6150 hits [35]. In 2011 in Europe 1789 patients have undergone treatment using some form of cell
or tissue engineered therapy (excluding the already commonly used procedures such as bone
marrow transplantation) [36].
31
Table 1.1 – The number of tissue engineering related clinical trials found in various registries around
the world (Data last retrieved on 17/02/2014)
Tissue engineering was defined by Langer and Vacanti in 1993 as “an interdisciplinary field that
applies the principles of engineering and the life sciences toward the development of biological
substitutes that restore, maintain, or improve tissue function” [37].
This is quite a broad definition, similar to that of RM. A more practical definition was offered by
the journal of Regenerative Medicine [15], one that has since been endorsed by the British
Standards Institute [31]: Tissue engineering is the “use of a combination of cells, engineering,
materials and methods to manufacture ex vivo living tissues and organs that can be implanted to
improve or replace biological functions”.
As the definition suggests the three cornerstones of tissue engineering are cells, biomaterials and
stimulation (Figure 1.3).
Search term
Database URL Tissue
Engineering
Regenerative
Medicine
Mesenchymal
Stem Cell
ClinicalTrials.gov
(USA)
clinicaltrials.gov
56
119
380
European Clinical
Trials register (EU)
www.clinicaltrialsregister.eu 5 5 28
UK Clinical
Research Network
(UK)
public.ukcrn.org.uk 3 2 6
WHO International
Clinical Trials
Registry Platform
(World)
apps.who.int/trialsearch/Def
ault.aspx
65 21 126
32
1.3.1 Cells
Amongst many others, osteoblasts [38]; chondrocytes [39]; macrophages [40]; adipose [41, 42],
dental follicle, bone marrow and skin derived MSCs [43, 44]; embryonic [45] and induced
pluripotent stem cells [46]; tendon-derived stem cells [47]; endothelial cells [48]; cardiac [49],
muscle [50], and neural progenitors [51], C2C12 myoblasts [52] and NIH3T3 fibroblasts [53] have
all been used as the basis for a tissue engineered construct.
1.3.2 Biomaterials
Cells have been combined with natural (e.g. collagen, gelatine, chitosan, silk fibroin) [41, 48, 54-
56], and synthetic (e.g. PLA, PLLA, PLGA, PLCL, PCL) polymers [44, 53, 55, 57, 58], non-
polymeric (e.g. bioactive glass, hydroxyapatite, calcium phosphate) materials [38, 45, 57] and their
composites [53, 59].
1.3.3 Stimulation
The cell-biomaterial constructs have been stimulated through chemical (e.g. growth factors [48, 60,
61]) and physical means (e.g. mechanical [47, 50], magneto-force [52], perfusion [58] and
topographical [62] cues).
The combination of these three elements (cell-biomaterial-stimulation) yielded many successful
studies into the engineering of bone [38, 40], cartilage [39, 60], tendon [47], cardiac [49], smooth
[42, 44] and skeletal muscle [52], bladder [44], adipose [55], muscle-tendon junction [53], ureteral
[42], vascular [58], intestinal [56], skin [59] and neural [46, 51, 54] tissue.
33
Figure 1.3 – A summary of the tissue engineering process
Primary or adult stem cells are acquired from a patient or a donor. Alternatively embryonic stem cells
(ESC) can be used. To give primary cells ESC-like capabilities, pluripotency can be induced to create
induced pluripotent stem cells (IPSCs). Cells are expanded in culture and then placed into a
bioreactor, generally on a biomaterial scaffold. Chemical and physical stimuli are applied to promote
tissue generation. After weeks of tissue culture the generated tissue construct is implanted into a
patient.
34
1.4 BONE
The main supporting tissue of the human body is bone [63]. As a part of the musculoskeletal
system, bones serve as the attachment site for muscles, allowing the locomotion of the human
body, while also protecting the vital organs from harm [64, 65]. Additional to its mechanical
support role, bone tissue has very important endocrine functions as well [63, 64], such as the
storage of calcium and the maintenance of the ion homeostasis [64].
Bones are not passive elements of the human body. Bone tissue undergoes constant remodelling, a
continuous cycle of bone resorption and formation, following Wolff’s law: Bone is deposited
where needed and resorped where it is not [63, 64, 66]. This allows this calcified tissue to adapt to
many different mechanical loads, supporting the human body much more efficiently.
1.4.1 Bone as a Material
Macroscopically, human bone tissue can be placed into two groups. Cortical bone tissue is dense,
solid and forms the external surface of bones [63]. Trabecular bone on the other hand is a porous,
light weight material with a honeycomb structure [63]. These two bone tissue types are made up of
the same composition, but with different porosities [63].
A characteristic feature of cortical bone is the Haversian system, an interconnected network of
canals surrounded by concentric rings of bone matrix [63]. The Haversian system plays an
important role in the delivery of nutrition to cortical bone, as this bone type is too dense to allow
nutrient/waste exchange through diffusion [63].
Microscopically there are two types of bone (Figure 1.4): Lamellar bone is highly organised,
strong, but relatively slow growing [63]. Woven bone is immature, disorganised and thus weaker
than lamellar bone. Woven bone is only present during embryonic development and in fracture
sites, where fast growth is required [63]. With time woven bone is replaced by lamellar bone,
therefore all adult bones are built up from lamellar bone [63].
35
Both lamellar and woven bone is a composite of an organic protein matrix and an inorganic mineral
phase [63]. The organic matrix is made up in 90% of collagen, a long protein chain, (97% type I,
3% type VII) [63, 67, 68]. The remaining 10% are the non-collagenic proteins, for example
osteopontin, osteocalcin and the bone morphogenic proteins [63, 67, 68]. The organic protein phase
grants bone its elasticity [63]. The mineral phase consists of hydroxyapatite (HA), a form of
crystalline calcium phosphate [66]. HA is what gives bone its rigidity and compressive strength
[66].
Figure 1.4 - Scanning electric microscope images of woven (A) [69] and lamellar bone (B) [70]. Note the
much more organised structure of lamellar bone. Scale bars correspond to 5 μm (left) and 1 μm (right).
Images reproduced with permission of the publisher.
1.4.2 The Cells of Bone
Bone is not just a passive material, but a living, metabolically active organ with its own set of cell
types: Mesenchymal stem cells, osteoblasts, osteocytes and osteoclasts [65, 66, 71]. (Lining cells
(protecting the bone surface) are sometimes also stated as a fifth bone cell type [72].)
1.4.2.1 Mesenchymal Stem Cells
Mesenchymal stem cells (also designated as “mesenchymal stromal cells”, “multipotent stromal
cells” and “colony forming unit – fibroblasts” in various publications) are non-hematopoetic adult
stem cells that have found widespread application in regenerative medicine [73-77]. This is thanks
36
to MSC’s multi-lineage potential, self-renewing capability, immunosuppressive properties and
relative availability [74-79]. Today there are myriads of scientific studies and many clinical trials
(Table 1.1) testing the usefulness of these stem cells in applications ranging from fracture healing
[80], treating stroke [81] to drug delivery to cancer cells [82]. Indeed MSCs are the predominantly
(56%) used cell type in cell and tissue engineering therapies [36].
MSC’s role in vivo is to provide a reservoir of cells that can mobilise, proliferate and differentiate
into multiple lineages if required for tissue repair or maintenance [76]. Originally isolated from
bone marrow [83], MSCs have also been found in a diverse range of locations around the human
body, for example, adipose [84, 85], spleen [84], thymus [84], placenta [86], umbilical cord blood
[78, 79], peripheral blood [87] and even breast milk [88]. This lead to the hypothesis that MSCs are
in fact pericytes, lining the blood vessels all around the human body [89]. MSCs are currently
defined through the antibodies expressed on their cell membranes (positive for: CD73, CD90 and
CD105; negative for: CD11b, CD14, CD19 or CD79a, CD34, CD45 and HLA-DR) [90-92];
adherence in vitro; and their ability to differentiate into chondrocytes, adipocytes and osteoblasts
[76, 90-92]. In addition to these three lineages, MSCs are believed to possess a much wider
differentiation potential [76] and have been reported to differentiate for example into tenocytes
[93], myocytes [94, 95], fibroblasts [96] and neurons [97].
1.4.2.2 Osteoblasts
Osteoblasts arise from MSCs (Figure 1.5) and are the cells responsible for bone formation during
development, remodelling and repair [63, 65, 66, 98]. Osteoblasts line the surface of bone,
synthesising and secreting bone matrix in osteoids and microvesicles [63, 65, 66, 98]. Osteoids are
nodules of uncalcified bone matrix, rich in collagen and other proteins [63]. The microvesicles
secreted by osteoblasts serve as the site of the initial HA crystal formation [66]. The HA crystals
are then deposited in the osteoids, continue growing, eventually leading to the osteoid’s
calcification [63].
37
1.4.2.3 Osteocytes
Around one tenth of the osteoblasts remain “behind” to be surrounded by the advancing, newly
deposited osteoid material [99]. These cells become osteocytes, non-proliferative, terminally
differentiated bone cells [65, 99]. Osteocytes, engulfed by bone, extend long cytoplasmic processes
through the canaliculi in bone [63, 99]. Through these processes ostecytes form a three dimensional
network of cells, communicating with each other and the osteoblasts on the surface [65, 99].
Osteocytes are believed to be responsible for sensing mechanical stress in bone through
mechanotransduction and for guiding the remodelling process [65, 66].
Figure 1.5 – The steps of osteogenic differentiation [71]
1.4.2.4 Osteoclasts
Unlike osteoblasts, osteoclasts are the progeny of heamopoietic stem cells [66, 100]. Osteoclasts
are highly specialised, multinucleated cells that are responsible for the digestion of the bone extra
cellular matrix [100]. During the digestion process, known as resorption, osteoclasts attach to the
surface of bone, where they create an external hemivacuole [67]. Post-attachment osteoclast release
tartrate-resistant acid phosphatase, cathepsin and matrix metalloproteases through the part of their
cell membrane juxtaposed over bone [64, 67]. These enzymes and proteases, together with the low
pH environment created by the hydrogen pumps in the osteocyte’s cell membrane, disassemble and
solubilise the bone matrix [64, 67].
38
1.4.3 The Markers of Bone Differentiation
The osteogenic differentiation of mesenchymal stem cells is marked by the expression of factors
such as Cbfa1/RUNX2, osterix, osteocalcin, ostepontin, osteonectin and the bone morphogenic
proteins (BMPs) [66, 98]. These factors are both regulators and indicators of differentiation
process, as their expression profile (i.e. whether they are expressed and in what quantity) depends
on the stage of the differentiation and the mineralisation [71, 101].
1.4.3.1 Bone Morphogenic Proteins
BMPs are a subfamily of proteins that play an important role in the development of many tissue
types, for example bone, cartilage, neural and epithelial tissues [98, 102].
In the context of bone, BMPs act as osteoinductive agents [71], promoting the osteogenic
commitment, proliferation, migration and maturation of MSCs and osteoblasts [71, 98, 103].
Through this osteoinductive effect BMPs play an important role in the formation, repair and
regeneration of bone tissue. For example, BMP-2, BMP-4 and BMP-7 have been found to be
strongly present in mesenchymal progenitors and proliferating osteoblasts in fracture sites [103-
105]. Their expression returns to baseline only when the formation of lamellar bone is complete
[104]. Osteoblasts laying down woven bone were also stained highly positive for BMPs [105].
BMPs act upstream from the other osteogenic markers. For example, BMP-2 is known to promote
the expression of Cbfa1, osteocalcin, collagen type I and alkaline phopshatase at the mRNA level
[98, 106].
1.4.3.2 Cbfa1/Runx2
Core binding factor alpha 1 (Cbfa1), also designated as runt-related transcription factor 2 (Runx2)
is regarded as the master-switch for osteoblast differentiation [98, 106, 107]. Its presence is the
earliest known sign of an MSC’s commitment to the bone lineage [66, 98, 106]. The presence of
Cbfa1 is essential for bone development as demonstrated by Komori et al [108]: Cbfa1-negative
mice show a complete lack of bone formation with a skeleton comprising solely of cartilage [108].
39
Cbfa1 promotes the expression of a large range of extracellular matrix (ECM) related genes; for
example collagen type I, bone sialoprotein, osterix and osteocalcin [66, 98]; and through them, the
secretion of bone matrix [109].
1.4.3.3 Osterix
Osterix is a zinc finger protein that acts downstream from Cbfa1, committing the cell to the
osteoblast rather than the chondrocyte path [66, 106]. Its presence is essential for bone formation:
This was demonstrated in osterix-negative mice, which failed to produce mineralised bone tissue
during embryonic development [110].
1.4.3.4 Bone Sialoprotein
Bone sialopotein (BSP) constitutes approx. 12% on the non-collageouns proteins of bone [111].
This glycoprotein is believed to serve as the nucleation site, i.e. initiator, of HA crystal formation
[111, 112]. This is supported by the observations that HA crystals form in solutions with below-
precipitation concentrations in the presence of BSP [112].
1.4.3.5 Osteonectin, Osteocalcin and Ostepontin
Osteonectin, osteocalcin and ostepontin are expressed by both osteoblasts and chondrocytes and are
involved in the mineralisation of the ECM [101]. Osteonectin is the most abundant protein
constituent of bone ECM apart from collagen [71]. The purpose of osteonectin (ON) is to link the
mineral and organic phases of bone by specifically binding to collagen, HA and calcium [71, 113].
ON-negative mice show decreased bone formation paralleled by low osteoblast and osteoclast
numbers and activity [114]. ON is also suggested to play a role in HA nucleation [113]. This is
supported by the observation that ON mRNA is expressed in pre-osseus and osseus tissue in vivo
prior to mineralisation [115]. Similarly to ON, osteocalcin is an ECM protein with calcium binding
properties [71] and is considered to be one of the main markers of osteogenic differentiation [116].
Osteocalcin’s (OC) function is to govern the growth of the HA crystals [71] as a negative regulator
[117]. This was demonstrated in OC-negative mice, which produced bones with higher HA content
40
then wild-types [118], thus demonstrating that OC suppresses mineralisation [118]. As a marker it
is important to note that OC is only up-regulated in post-proliferative osteoblasts [106].
Osteopontin (OP) is a highly phosphorylated extracellular glycoprotein that also plays a role in the
mineralisation process [119] and as a mediator of cell-ECM attachment [71, 115]. OP is one of the
main non-collagenous components of the bone matrix and is a potent phosphate-binding molecule,
thus similarly to OC supresses the mineralisation process [119].
1.4.3.6 Alkaline Phosphatase
Alkaline phosphatase (phosphate-monoester phosphohydrolase (alkaline optimum)) (ALP), also
referred to as tissue-nonspecific alkaline phosphatase (TNAP) is a metalloenzyme important for the
mineralisation of bone and is a widely used marker of osteogenesis [120, 121]. ALP is secreted by
osteoblasts through the microvesicles and is essential for the nucleation and formation of HA
crystals [66, 120]. Alkaline phosphatase’s essential nature comes from its ability to hydrolyse the
pyrophosphate (PPi) excreted by cells into inorganic phosphate [119-121]. This does not only
produce a necessary component (inorganic phosphate) for HA formation, but lowers the local
concentration of PPi [119, 120]. This is important as PPi is a potent mineral-binding molecule that
would otherwise suppress mineralisation by compromising inorganic phosphate’s ability to form
HA crystals with calcium [119, 120]. Similarly with PPi, ALP also hydrolyses OP, thus promoting
HA formation by countering both PPi’s and OP’s suppressing effect [119].
1.4.4 The Expression Profile of the Osteogenic Markers
Measuring the expression of osteogenic marker genes and proteins is a valuable tool for assessing
the differentiation and mineralisation state of cells. However, to allow truly accurate conclusions to
be drawn the expression profile of these markers as a function of time must be understood. (I.e. it
must be understood whether the cells are “meant to” express a specific marker at the assayed time
point and what the relationship is between the expression of the gene and the
41
mineralisation/differentiation process. Higher levels of expression does not necessary mean greater
differentiation or more extensive mineralisation.)
Therefore information was gathered from the literature on the expression of osteogenic markers
during the osteogenic differentiation process as a function of time. Information was only included
from articles where an osteoblast precursor was differentiated in culture medium containing only
dexamethasone, β-glycerol phosphate and ascorbic acid as osteogenic supplements. Studies where
growth factors or the forced expression of genes was used were excluded.
The following summary can be made:
Cbfa1/RUNX2, BMP-2, TGF-β, ON, ALP and collagen type I are generally regarded as
early markers of the osteogenic process, while OC and OP are expressed in the late stages
[122].
ALP activity peaks between day 7 and day 20. After the peak the expression of ALP either
diminishes with time or plateaus depending on the cell type and culture conditions (Figure
1.6) [116, 123-128]. This profile has been noted to correlate well with the progression of
mineralization [108].
Cbfa1/RUNX2 shows gradual increase up to a peak between day 3 and day 10. After the
peak the expression of Cbfa1/RUNX2 falls but is still present [87, 116, 123, 124, 126,
129].
The expression of BSP shows gradual increase with time (observations were made up to
day 20) [116, 123, 126].
Collagen type I peaks between day 7 and day 14, after which it continuously falls up to day
35 [87, 124, 126].
OC is expressed in an increasing manner from day 10-12 up to a peak between day 20 and
day 28. Following the peak OC expression somewhat diminishes but is still strongly
present up to day 35 [116, 124, 126-128]. It is interesting to note that adipose derived
MSCs were found to express OC neither at day 14 nor day 28 [131].
42
The expression of BMP2 was shown to increase up to day 30 [116, 126].
ON was found not to increase or to increase only slightly up to day 20 [87, 126].
OP expression increases up to a peak between day 20 and day 28 [87, 126-128, 130]. An
initial smaller peak has also been observed around day 7 [127]. This early peak was
detected by Nakamura et al in mouse MSCs as well, where OP mRNA expression peaked
at Day 8, followed by a decrease until day 12 and a plateau to day 16 [116].
Osterix mRNA expression was observed to increase continuously up to day 16 [116]
The expression of osteogenic markers heavily depends on the cell type and the species,
anatomical location, age, etc. of the source [106, 131].
Figure 1.6 – The expression profile of ALP during the osteogenic differentiation process
Based on the above information the following simplified model of bone generation can be
proposed: Induced by external factors, such as BMPs, MSCs begin their commitment to the
osteoblast lineage. This is initially marked by the expression of Cbfa1/RUNX2 and later on by
osterix, which directs differentiation away from the chondrogenic path and towards the bone
lineage. The cells proliferate and secrete collagen and ALP. ALP provides the inorganic phosphate
43
necessary for HA formation. HA crystals nucleate around BSP in microvesicles and continue their
growth once outside of the vesicles. The growth of the HA crystals is aligned by the collagen fibrils
[117]. Once sufficient cells have been generated and sufficient amount of matrix has been
deposited the secretion of ALP and Collagen dies down. Cells cease proliferating and the
expression of OP, OC and ON up-regulates. ON links together the organic and mineral phases of
bone, while OP and OC bind calcium and phosphate, thus keeping HA crystals from growing
beyond their intended/necessary size. Bone density is maintained through the equilibrium of bone
marker proteins and the activity of osteoblast and osteoclast.
44
1.5 ELECTRICAL STIMULATION
Chemical, mechanical, material-based (i.e., topography, scaffolds) and magnetic cues are now well
established tools in the in vitro creation of tissues and organs. Researchers are now beginning to
develop alternative cell stimuli/activation processes, and electrical stimulation (ES) has become an
active area of research in the engineering of nerves and cardiac and skeletal muscle. Endogenous
electric fields play an essential role in the functioning of all living organisms, not just in the well
known action potentials of nerves and muscles [132, 133], but also in controlling cellular functions,
such as morphology, elongation, gene expression, proliferation and migration [19, 20, 134-139].
Bioelectrical circuits and their ‘wiring’, act as a long range intercellular signalling and controlling
mechanisms in the development, maintenance, repair or regeneration of tissue, and tumour growth
[136, 137, 140-142]. As such, the utilisation of external electrical stimuli as a modality for
improving the quality of the engineered constructs is an important and challenging area of research
for the groups of, for example, Radisic, Vacanti and Langer.
In the following section the various aspects of ES will be discussed. Examples of endogenous
electricity, the different types of stimuli, the various novel methods of delivering these and their
effects on both the cellular and tissue level will be provided, in order to demonstrate the potential
that this technique has for the discipline of tissue engineering.
1.5.1 In Vivo Electricity
The main sources of in vivo electricity are the cells. Through the constant pumping of ion channels
they establish a voltage gradient across their membrane, the membrane potential [143, 144]. By
convention the potential outside the cell membrane is considered to be zero, thus membrane
potentials are specified accordingly. The membrane potential of most cell types is between -60 and
-100 mV [145, 146]. When cells couple together into a continuous layer, they create a resistive
barrier, paralleling the cellular membrane on a bigger scale [140, 144, 147]. Polarised Na+, K
+ and
Cl- ion transport on the two sides of this layer establishes a tissue-level electric gradient across the
45
cellular interface of typically 15-60 mV, an example of which is the trans-epithelial potential (TEP)
found in skin, lens and cornea [140, 144, 147-150]. The TEP mechanism in vertebrate lens is
particularly interesting, as it creates a complex pattern of current loops with magnitudes around 20-
40 mA/cm2. These currents flow inward at the anterior and posterior poles of the lens, and outward
at the equator of the lens, controlling the migration and differentiation of epithelial cells [135, 148].
Another important function of TEP can be observed in injuries of the epithelium. A wound short
circuits the potential difference between the two sides of the epithelium, and, therefore, gives rise
to an electric field (0.04-0.2 V/mm) [136, 137, 140, 142, 144, 150-152] that, in turn, controls the
orientation and frequency of cell division [150] and induces directional migration towards the
injury [136, 142, 147]. This has been shown to override any other influence on the cell’s
functioning such as chemical gradients and population pressure [140, 147]. In regeneration and
embryonic development intra- and extra-cellular electric fields play a pivotal role in regulating
cellular behaviour [140] and the development of spatial patterns such as left-right organ asymmetry
[139, 140, 153, 154]. Disruption of these fields in embryos has been reported to result in serious
defects, for example the absence of the cranium, a malformed head and the loss of eyes [140].
The significance of electricity for regeneration was also demonstrated in studies conducted on
partially amputated Xenopus tails. Electrical current was observed to occur as a result of the injury.
Disrupting this electrical current reduced regeneration, while reversing it completely blocked the
healing process [155].
The significance of bioelectricity is emphasised by its role in one of the rare examples of human
regeneration. Illingworth and Barker [156] reported that the amputated fingertips of children could
be fully regenerated as long as the stump is kept clean and hydrated and observed that electrical
currents of 30 A/cm2
were present in wounds. In their paper, they theorised that moisture in the
wound ensures a continuous electric conductance path, enabling the wound’s electric fields to exert
their regenerating effect [156]. Streaming potentials, streaming currents and the piezoelectricity of
collagen molecules all contribute to the generation of bioelectricity in bone [19, 24, 25, 27, 157-
161]. These electrical phenomena are suggested to be transduced by osteocytes and to play a
46
similar role in remodelling and healing as TEP [20, 160-162]. Bioelectricity is also suggested to be
the ideal mechanism to deliver signals to the chondrocytes in cartilage, as they usually exist in
relative isolation [163]. Other examples of in vivo electricity include the rapid action potentials of
nerves and muscles; long lasting DC voltages around damaged nerves, known as injury potentials
[140] and the trans-endothelial extracellular potential gradients in arteries and veins [151, 164].
The use of ES is well-establish in today’s medicine. A wide range electrical therapeutic system
have been developed to treat specific diseases and/or tissue types: These devices include
implantable and external bone growth stimulators [165, 166], chronic cutanous wound healing
systems [167], functional electrical stimulation devices for the restoration of muscles in paralysed
patients [168], and stimulators for pain relief [169, 170].
Transcutaneous nerve stimulation (TENS) units are primarily used to deliver electro-analgesia and
have been proven to be greatly effective in the treatment of a wide range of pains [169, 170]. In a
clinical study examining the effectiveness of these devices it was shown that TENS treatment
reduced pain by more than a half in 47% of the patients [171]. TENS devices have been shown to
be useful in the treatment of chronic ulcers as well [167]: In a meta-analysis where the effect of
multiple systems (including TENS units) was assayed upon chronic wound healing it was found
that ES produced an enhanced healing rate of 22% per week compared to the 9% in patients not
treated with ES. Upon the comparison of the effect of the various devices, no significant difference
was observed [167]. In another investigation, examining the effectiveness of electrical therapy in
the treatment of pressure ulcers, the wound surface decreased approx. 70% with electrical
treatment, while only 36% without it [172]. ES systems have also been demonstrated to be highly
effective in promoting bone growth: In an extensive study involving 175 patients with non-union
fractures, ES was able to induce solid bone union in 83.7% of the cases [165].
Due to the wide range of bioelectrical presence and its significant influence on in vivo tissue,
researchers are now beginning to explore the potential of using this stimulus in vitro.
47
1.5.2 The Methods of In Vitro Electrical Stimulation
1.5.2.1 Types of Electrical Stimulation
The most basic method of ES is applying a DC voltage, simply generated by batteries [173]. More
complicated stimuli can be in the form of monophasic (DC) or biphasic (AC) [20, 25] sinusoidal
[19, 162, 174, 175], saw tooth [28, 159, 176] or square wave [154, 177] signals, injected in pulses
[28, 159, 177], pulse bursts [19, 178] or continuously [20, 133, 138]. These signals can be
generated by stimulator chips [25], signal generators [26, 154, 175], or dedicated therapeutic
systems [141, 179], such as the Phyback [180], TENS [169, 170] or the EBI Bone healing systems
[19, 26, 178].
1.5.2.2 Methods of Delivering the Stimulus
1.5.2.2.1 Direct Coupling
With direct coupling, the electrode is in direct contact with (e.g. inserted into) the cell culture or
implanted into the patient or laboratory animal [24, 25, 160]. Although this is the simplest way of
delivering ES, there are quite a few disadvantages to the approach. The problem of toxicity arises,
due to the insufficient biocompatibility of the electrodes [181], changes in pH [24], reduced levels
of molecular oxygen [181, 182] and the generation of dangerous Faradic by-products, for example
reactive oxygen species, in the culture medium [20, 24, 181, 182]. The formation of a capacitive bi-
layer around the electrodes [24, 25, 182] and that the effects of the stimulation was observed to
depend heavily on whether it was measured near the anode or the cathode electrode [182], further
hinder the use of this method.
1.5.2.2.2 Indirect Coupling
To avoid the difficulties inherent in direct coupling, an indirect, non-invasive approach is used in
many therapeutic devices and in vitro experimental setups. The three main types of indirect
coupling are the capacitive, inductive and combined coupling methods (Figure 1.7), where the
latter is a combination of a static magnetic and a transient electromagnetic field [19, 20, 25].
48
Figure 1.7 – The four different methods of electrical stimulation. (Figure reproduced from [183].)
1.5.2.2.2.1 Capacitive Coupling
In capacitive coupling (CC) a homogenous electromagnetic field is created between two parallel
layers of metal (e.g. stainless steel, gold) or carbon electrodes, that are placed above and below the
culture environment usually with only a small (0.5-2 mm) gap between them [24, 143, 159, 160,
163, 184-188]. As the electrodes are isolated from oneanother, there is no electrical current,
therefore CC can be utilised without the side-effects of direct coupling. Additionally, as the electric
field is homogenous, an equal amount of stimulation is received by each cell regardless of their
position in the culture vessel. However, in many instances a high voltage (in the range of 100 V) is
needed to be generated between the electrodes for an effective stimulus [189]. Due to this
requirement, more complicated equipment can be necessary when using this technique compared to
the other methods.
A modified version of this method is the semi-capacitive coupling, where the upper capacitor plate
is placed directly on top or slightly immersed in the culture medium [158]. This allows the
49
generation of a more powerful electric field upon the cell layer with the same electrode potentials.
On the other hand, in vitro studies have shown the semi-capacitive coupling solution to be less
effective: In a study where the effect of the capacitive and semi-capacitive methods on osteoblast
ECM formation was compared, the crystal nodules in semi-capacitively stimulated samples were
found to be uncharacteristic and were suggested to be the result of precipitation of electrolytes from
the culture medium, rather than being deposited by the cells [158].
1.5.2.2.2.2 Inductive Coupling
Inductive coupling (IC) utilises dynamic electromagnetic fields that induce small magnitude
electrical currents and potentials in the proximity of the targeted cells. The electromagnetic fields
are generated by coils placed around the cell culture, avoiding the need of delivering the
stimulation through invasive electrodes [21, 24, 141, 151, 160, 174, 190]. In many instances, the
coils are used in pairs, placed in the Helmholtz configuration [26, 28, 162], where the distance
between the coils is equal to their radius [162]. The Helmholtz configuration allows a near
homogenous magnetic field to be generated with uniform electromagnetic field properties across
the cell culture [162].
A subtype of IC is known as pulsed electromagnetic field stimulation (PEMF), where the stimulus
is delivered in pulses (rather than being static or continuously harmonic) in order to mimic, for
example, the natural strain-generated potentials observed in bone [162, 191].
50
1.5.2.3 The Parameters of Electrical Stimulation
The electrical potential on the electrodes and the strength of the generated electric field are,
arguably, the most important parameters of ES. The outcome, especially with regard to cellular
migration, orientation [142, 153, 164, 192] and gene expression [176, 193], has been shown to
depend upon the amplitude of the voltage, with strong indications that a relatively higher setting is
favourable in nearly all circumstances [137, 142, 153, 164, 176, 192-195] – see Table 1.2.
In the case of electrical current this is different: Positive results, such as enhanced proliferation or
increased ECM deposition, were witnessed with electrical currents (or current density) at both
relatively high and low settings (Table 1.3). As such, the strength of the current does not seem to be
important when considering the effectiveness of the stimulus. However, it does act as a limiting
constrain, since any current above a certain threshold will result in cell death [20, 24, 25, 138, 165,
195].
Table 1.2 – The effect of various electric field strengths
0.00000048 V/mmReduced osteoclastogenesis in bone marrow cultures compared to
controls and higher amplitude settings [176]
0.0000006 V/mm Reduced proliferation, but increased ALP activity in osteoblasts [174]
0.000002 V/mm Inhibited cellular growth of osteoblasts [27]
0.002 V/mm Enhanced proliferation and TGF-B1 expression in osteoblasts [186]
0.01 V/mm Migration of fibroblasts above this threshold [153]
0.15 V/mm Orientation of fibroblasts above this threshold [153]
0.4 V/mm Elongation of fibroblasts above this threshold [153]
6 V/mm
Enhanced mineral formation in osteoblasts compared to controls [158]
Enhanced proliferation, ALP activity and mineral formation in
osteoblasts compared to controls [159]
51
A wide range of frequencies (7.5-60000 Hz) have been applied in experiments, though the most of
these were chosen to be below 100 Hz. This is a logical choice as frequencies in this low range, are
the ones that are assumed to occur naturally in vivo and are the most likely to have a beneficial
effect [19, 20, 24, 25, 27, 138, 158, 159, 174, 186, 196]. Frequencies between 10 and 30 Hz [174]
and those less than 100 Hz [162] have been stated to stimulate new bone formation the most, while
7.5 Hz has been observed as best at promoting the proliferation of primary rat osteoblasts [196]
and osteoclast differentiation of bone marrow cells of mice [176].
The maximum pulse width is directly limited by the frequency and indirectly by the current
density. Higher the electrical current, the duration of the ES needs to be shorter to avoid necrosis
(i.e., there is a limit on the electrical energy that can be absorbed by the cell safely) [20, 195].
Nonetheless none of the settings, from 0.025 to 380 ms, that have already been applied in in vitro
experiments, proved ineffective or damaging [20, 26, 138, 158, 159, 162, 177, 185, 190, 197, 198].
Table 1.3 – The effect of various current densities upon osteoblasts and mesenchymal stem cells
0.015 A/m2
Enhanced proliferation, increased VEGF expression, and no upregulation of bone
markers in osteoblasts compared to controls [25]
Enhanced proliferation, increased VEGF expression and upregulation of bone
markers a week after end of stimulation in MSCs [138]
Increased VEGF expression, upregulation of calcium deposition and AP activity a
week after end of stimulation in MSCs compared to controls. Greater proliferation
than with 0.15 A/m2 stimulation. [20]
0.15 A/m2 Worse proliferation of MSCs compared to 0.015 A/m2 samples. [20]
3 A/m2Proliferation and TGF-B1 expression of osteoblasts [186]
Enhanced DNA content increase in osteoblasts [19]
4.2 A/m2 Enhanced proliferation and gene expression of osteoblasts [24]
5 A/m2 Necrosis in osteoblasts above this threshold (DC) [182]
52
1.5.3 The Cellular Effects of Electrical Stimulation
Electric fields have been shown to affect cell metabolism, morphology, protein and DNA synthesis,
alter ionic currents in the cellular membrane, influence mRNA transcription and to promote
proliferation and adhesion [146, 199, 200, 201]. How powerful an effect ES can have on cells has
been well demonstrated in HeLa cells: When an electrical potential of 0.4 V was applied to HeLa
cells cultured on the surfaces of electrodes, the cells completely stopped proliferating and only
returned to multiplication when the electrode potential was decreased to 0.1 V [199, 202, 203].
Several theories exist on how electric fields are recognised at the cellular level. What can be known
is that electric fields seem to be interpreted by the same pathways observed to be involved in
mechano-transduction and chemotaxis [137, 147, 152, 153, 204]. A key component of this process,
that has been observed to be greatly affected by electrical stimuli, is the intracellular calcium level.
1.5.3.1 Intracellular Calcium
Electric signals are believed to be transduced at least partially, through the calcium/calmodulin
pathway (Figure 1.8) [25]. This happens somewhat differently with the various methods of
stimulation: Direct and capacitive coupled stimuli exert their effect mainly on the cellular
membrane, as these cannot overcome its high electrical resistance [144, 151], raising the
intracellular Ca2+
concentration and prostaglandin E2 levels by activating the voltage-gated calcium
channels in the cell membrane [144, 151]. On the other hand, inductively coupled and combined
electromagnetic fields are theorised to generate potentials and currents in the cytoplasm, releasing
intracellular calcium from reservoirs such as the endoplasmic reticulum [151, 196]. The elevated
calcium level in both cases activates the cytoskeletal calmodulin, resulting in, for example,
enhanced proliferation, increased VEGF and TGF-β1 expression [19, 20, 25, 133, 151, 163, 186,
196]. This hypothesis is supported by the fact that blocking the calcium channels by verapamil and
nifedipine [20, 186], the intracellular stores by TMB-8 [196] or calmodulin by W-7 [186, 196],
actively impaired or completely blocked the ES’s effect.
53
Figure 1.8 – The calcium mediated intracellular pathway for sensing electrical signals. (Figure
reproduced from [183].)
Cytosolic calcium has been found to play an additional role in regulating stem cell differentiation
in the form of Ca2+
spikes [205]. Calcium oscillations (i.e. spikes) have been detected both in
osteoblasts and hMSCs, but with different “profiles”. These spikes can serve as an indicator of
differentiation, as during the process, the level of calcium oscillation decreases from what was
observed in hMSCs to that of osteoblasts. Applying an external electric field to hMSCs altered the
calcium oscillation and facilitated osteo-differentiation, presenting an alternate path of transducing
ES in cells and suggesting a new way of controlling stem cell fate [205, 206].
1.5.3.2 The Response of Cells to Weak Electric Fields
Electric fields too weak to open the voltage gated ion channels, or affect other parts of the cells
directly, have been observed to alter cellular behaviour [151, 205, 206]. The actual mechanisms
behind these phenomena are in debate [144, 153]. A simple solution is that, instead of triggering
54
the ion channels or internal Ca2+
release, these small potentials may exert their influence by
keeping the channels open longer or by helping force ions in (or out) of the cell [144, 153].
Alternatively, other ion channels (that may require less stimulation) could be involved in
transducing the electric stimuli. For example, H+ ion exchangers have been associated with
migratory responses in weak electric fields when the activity of Ca2+
was not detected [204]. In
another study, blocking of Na+ channels by tetrodoxin was found to reduce, the otherwise strong,
galvanotactic response of rat prostate cancer cells [149]. This hypothesis is further supported by the
fact that Na+, K
+, H
+ and Cl
- ion pumps are already known to play a role in generating endogenous
electricity such as action potentials, TEP and injury currents [147, 153, 207].
1.5.3.3 Growth Factors and Receptors under Electrical Stimulation
The significance of growth factor receptors in ES transduction has also been highlighted in many
instances, particularly those of EGF, FGF, TGF-β1 and VEGF [144, 149, 164, 193]. This is
supported by the observation that removing these growth factors from the culture medium
disrupted the electric field induced migration of cells [193]. In another study it was found that, as a
result of DC electrical stimulation, EGF receptors on the cell membrane moved to the cathode
electrode facing side of the cell [147].
1.5.3.4 Similarities between Electrical and Mechanical Stimulation
A degree of similarity seems to exist between the transduction of mechanical stimuli (e.g.
ultrasound signals) and that of electric fields: both involve interactions with the cell membrane,
changes in cytosolic calcium and calmodulin activation. One explanation may be that the fields act
upon charged lipids and protein molecules in the cell membrane via electrophoresis or electro-
osmosis [144] resulting in a mechanical like effect [196, 208]. Alternatively electric fields could be
mimicking mechanical stimuli by interfering with focal adhesion proteins, reorganising
microfilament or redistributing and activating integrins [151, 205]. Another explanation is that ES
influences the biomechanical properties of the cells by affecting the cell membrane’s permeability,
55
fluidity and the actin cytoskeleton structure which can be interpreted as a mechanical stimulus
[132, 151, 206].
1.5.3.5 The Mechanisms behind Galvanotaxis
How ES can induce migration (i.e. galvanotaxis) and orientation is a difficult question. Asymmetric
assembly and disassembly of the actin filaments and polarised redistribution of membrane
receptors and integrins have been suggested to play a crucial role in this process [134, 137, 142,
147, 151, 204, 206]. On the other hand, intracellular calcium’s importance in this instance is still
debated: Although intermittent discreet Ca2+
events [137] and waves, originating from the anodal
and heading towards the cathodal side of the cells [204], have been witnessed during galvanotaxis,
researchers were unable to link these to changes in cytoskeletal tension (known to occur in a
pattern corresponding to the direction of the applied fields) before movement [208], or any other
part of the migration process. Furthermore, inhibiting the calcium channels by nifedipine,
gadolinium, nickel or strontium did not alter the migratory rates, although the latter seemed to
influence the directionality of the movement. Similar, orientation disrupting effects were witnessed
with T–type calcium inhibitors. These findings together with reports that blocking PI3K, a member
of an important motility pathway, inhibits electric field driven migration to a greater extent than
orientation, suggests that these two are regulated by a different mechanism [137, 149].
Cathode-wise polarisation of the Golgi apparatus has been proposed as the key factor in defining
the direction of the motion, as the presence of this organelle at the anterior side of the nucleus is a
known prerequisite of forward movement of cells. This polarisation seems to be an overriding cue
during electric field driven migration and disrupting it through chemical means was shown to
significantly reduce directional motility of cells [147].
1.5.3.6 Intracellular Signalling Pathways
There are several intracellular pathways, including the calcium/cadmium pathway mentioned
previously that are, believed responsible for coupling the “direct” effects of electric fields into
cellular responses. The polarised activation of PI3K/Pten and it‘s target Akt and rho, (already
56
known to be important in chemotactic reactions) [137, 147, 152, 207], protein kinase C [144, 149,
205] and the mitogen activated protein kinases p38 and Erk [20, 138, 152, 204-206] are believed to
be involved in this process and responsible for the observed changes in the expression of
proliferative and differentiative genes. This is supported by the strong evidence, presented by Kim
et al, that blocking p38, Erk, PI3K or calcium signalling pathways inhibited biphasic electric
current induced proliferation and VEGF and BMP-2 expression in MSCs [20, 138].
1.5.3.7 Sensing in Excitable Cells
Excitable cells utilise electrical signals during their normal functioning, as such, they are likely to
possess an inherent pathway for sensing ES. Short pulses were found to generate contraction in
muscle cells by mimicking the neuromuscular transmission, while long pulses are theorised to be
transduced through voltage sensor proteins of the excitation-contraction coupling, triggering in the
end calcium release from internal reservoirs [209, 210]. In nerve cells the growth of neurites, the
redistribution of materials in the cytoplasm, increased fibronectin adsorption, accumulation of
surface molecules beneficial for neurite growth on the membrane, and ionic currents around the
tips of growing fibres have all been suggested to play a role in sensing electric fields [146].
Interestingly, same as in non-excitable cells Protein Kinase C have been found to be essential for
the transduction of external electric fields in neurons and astroglial cells; and Erk and p38 Mitogen
Activated Protein Kinases in muscle [146, 210, 211].
1.5.3.8 The Structural Effects of High Power Electric Fields
High power electric fields induce powerful changes in the physical structure of cells. Although
structural changes in cells are of lesser interest to tissue engineering, the knowledge gained in high
power ES applications should not be ignored. The experience gained from these experiments can
be very useful in understanding the mechanisms behind the effects of electrical signals and can
help in the effective use of this modality.
Physical changes induced by ES include amongst others electro-poration/electro-permeablization,
electro-transfection, electro-insertion, electro-adsorption and electro-fusion [212]. Out of these
57
arguably electro-permeabilization and electro-fusion enjoy the most widespread use [199, 212,
213]. Nonetheless, all of these modalities are applied daily in many areas, ranging from gene
therapy, drug delivery, cell hybridization and antibody creation in medical and biological research,
to sterilization in the food industry [212-215].
1.5.3.8.1 Electro-permeabilization
If an electromagnetic pulse above a critical field strength threshold is applied (for eukaryotic cells
one that generates 0.5-1.5 V across the cell membrane) both biological and artificial membranes
start to break down [215, 216]. The electromagnetic pulse generates an additional trans-membrane
potential in the cell [215]. When the sum of this additional potential and the cell’s own membrane
potential reaches the threshold, this usually happens in a matter of micro or nanoseconds, pores
start to form on the sides of the cell closest to the electrodes. [215] This phenomenon is called
electro-poration, and causes the cell membrane to be permeable, fusogenic and allows the
introduction of molecules (e.g. proteins, DNA) into the cell’s interior [216]. The size of injectable
molecules correlates with the size of the pores, as such, can be controlled with the strength of the
electromagnetic field [213-217]. Full breakdown does not occur as the growth of the pores is held
back by the cytoskeleton [213]. Irreversible damage is avoided as the cell’s conductivity (and as
such the integrity of their membranes) return to normal in a few seconds after the end of the
stimulation [213-217].
To introduce the additional trans-membrane potential low frequency strong electric fields are
necessary [218]. These can only be delivered in pulses to avoid any thermal side-effects [218].
High frequency (above 1 MHz) pulses will pass through the membrane without causing an affect,
and will not generate a change in the cell’s membrane potential [218]. However, as these can
penetrate the membrane, high frequency pulses may be able to exert an influence on the inside of
the cell [218].
The most novel application of this modality is electro-chemotherapy, where tumour cells in a
patient are permeabilized to anti-cancer drugs by electrodes placed around the tumour [216]. This
58
allows the well-controlled, specific delivery of these dangerous substances. Another application of
this modality is in electricity-based non-viral gene transfection, which has already been
successfully used in transfecting melanoma, liver, skin and skeletal muscle cells [213]. In electro-
transfection the targeted cells are kept in a suspension with the DNA plasmids. The applied electric
field brings the plasmids in contact with the cell membrane and promotes their diffusion into the
cell [216].
1.5.3.8.2 Electro-fusion
Cells suspended in a solution can be fused together by applying DC electric fields [203] (Figure
1.9). This can be highly useful in, for example, the creation of hybrodomas and new hybrid plant
species [213]. Electro-fuion can be carried out using three main techniques: the agglutinating
substance based macro-technique; the dielectrophoresis based micro-technique; and through the
compression of two cells with micromanipulators and needle electrodes [219]. This modality has
the advantage of avoiding the negative side effects of the biological and chemical methods of
forced cell fusion [213].
Figure 1.9 – Macrophages before (A) and after (B) electrofusion [219] (Image reproduced with the
permission of the publisher.)
59
1.5.4 The Effects of Electrical Stimulation at the Tissue Level
1.5.4.1 Galvanotaxis
Galvanotaxis or electrotaxis is the phenomenon of electric gradient guided cell migration [147,
149, 220]. Cells when undergoing galvanotaxis, depending on the particular type, move either
towards the cathode or the anode of the stimulating electrodes [147, 149, 220]. Control over the
direction and rate of cell movement, orientation and division through this mechanism has the
potential of stimulating or inhibiting tissue construction/reconstruction, immune responses, blood
vessel and neuron growth and could be a be a powerful new tool for tissue engineering [134, 142,
152, 184, 193].
In vitro studies have already shown the ability of electric fields to influence the behaviour of a
variety of cell types in this manner. Corneal and lens epithelial cells, umbilical vein and aortic
endothelial cells, neural crest cells, fibroblasts, osteoblasts, keratynocytes, lymphocytes, mASCs
and MSCs from various species have all been reported to exhibit cathodal migration and
perpendicular orientation in a dose dependant manner [134-136, 142, 148, 152, 153, 164, 192, 208,
221, 222]. Exceptions to this were the responses of rat primary osteoclasts [222] and human retinal
pigment cells [194] which seemed to favour the anode migration.
A recent investigation, conducted by Au et al [184], compared the efficiency of topography and
electricity in influencing the alignment and elongation of cardiac myocytes and fibroblasts. Their
results show that, though cells respond with greater alignment to topography than to ES, the best
results can be achieved with the synergistic application of the two techniques [184].
1.5.4.2 Enhanced Wound Healing
The usefulness of ES in skin tissue engineering and wound treatment is demonstrated by its ability
to induce the re-epithelialisation of cutaneous and corneal wounds through promoting migration
and proliferation of fibroblasts, keratynocytes and epithelial cells, enhancing angiogenesis,
improving blood circulation and blocking edama formation [141, 148, 150, 172, 180, 223, 224].
60
Furthermore, stimulation of cutaneous fibroblasts through conductive polypyrrole and poly-L-
lactide scaffolds yielded a tenfold increase in the expression of the genes IL-6 and IL-8, two
cytokines known to play an important role in wound repair and promoting the growth of new blood
vessels [132].
1.5.4.3 Improved Nerve Regeneration and Neural Tissue Engineering
There are also multiple indications that ES can promote peripherial and central nervous system
regeneration [173, 198]. Already used in numerous clinical applications such as cohlear implants or
the treatment of spinal cord injury and disuse atrophy, novel studies report the ability of ES to
increase nerve fibre and blood vessel density in sciatic nerve models [173], double the amount of
new cells after spinal cord injury [198], and to significantly increase neurite length when applied
through conductive polypyrrole films [225].
The important role of electric fields in the development, functioning and repair of neural tissue also
provides a strong basis for the application of ES in nerve tissue engineering [132, 133, 146, 211].
In laboratory experiments ES was found to promote the proliferation, differentiation and Nerve
Growth Factor (NGF) expression of nerve cells [146, 210, 226], and to be able to guide the
orientation and enhance the rate of neurite sprouting [146, 210, 226]. Interestingly, it was found
that the presence of NGF in the culture medium is not necessary for, but greatly boosts, the
effectiveness of ES [227]. In astroglial cells, in a similar way to what was observed with neurons,
electrical stimuli were observed to significantly increase the production of NGF [211].
1.5.4.4 Benefits for Bone
The positive effects of electricity in healing bone fractures has been noted as early as 1812 [19,
228], but interest in such treatments increased considerably when Fukada and Yasuda demonstrated
the piezoelectric properties of dry bone during the 1950s and 60s, providing a basis for the use of
ES as an osteoinductive tool [19, 24, 25, 157-159]. Such methods have been successful in treating
osteoporosis, osteoarthrosis, normal and non-union fractures and promoting the integration of
implanted biomaterials [19-21, 182, 229].
61
In vitro studies on osteoblasts and MSCs show a certain disparity between the effects of the various
coupling methods (Table 1.4). Capacitive stimulation seems to enhance proliferation, ALP activity
and ECM deposition, while delaying differentiation. Expression of osteopontin, BMP-4, IGF-2 and
other genes appear unchanged, while mRNA levels of BMP-2 and TGF-β1 have risen in a few
instances. VEGF seems to be an exception, where up-regulation has been observed, but only in
connection to biphasic stimulation [20, 24, 25, 138, 158, 159, 186]. The inductive PEMF method
shows a tendency for inhibiting cell growth while enhancing mineralisation and expression of
genes such as BMP-2, BMP-4, TGF-β1, ALP and prostaglandin E [26-28, 176]. A comparison of
capacitive, inductive and combined coupling, by Brighton et al [19], showed that only the
capacitive delivery of the stimulus results in a continuous increase of DNA levels [19]. On the
other hand, both methods improved healing during treatment of non-unions and osteoarthritis in
animal models as well as clinical trials [21, 162, 166, 187, 230-232]. In addition, both methods
seem to promote chondrogenesis, through enhancing proteoglycan and collagen secretion and the
mRNA expression of TGF-β1 and aggrecan in chondrocytes [178, 233].
62
Table 1.4 – A summary of bone related in vitro studies carried out with the three different methods of
stimulation
Cell type(s) Stimulation Results
0.1 V with 10 µA for 2 h using Stimulation on the PPy film produced significantly larger neurite
a Polypyrrole (PPy) film as anode length than in unstimulated and tissue culture plastic control groups.
Spontaneous contractions, expression and sarcomnetric organisation
of troponin T suggest cardiac differentiation.
0.05 V/mm DC for 24-48 h Cells adhered, spread and proliferated on the conductors.
using a PPy/PLLA conductor Cytokine production enhanced 10-fold by stimulation.
Rat bone marrow MSCs 0.2, 0.4 and 0.7 V/mm DC for Changes in morphology observed with both MSCs and fibroblasts,
Rat fibroblasts HT1080 0-60 min on Coll. Type I gel but only the latter aligned in response to the stimulus.
Neonatal rat ventricular 0.5 V/mm, 2 ms rectangular Stimulation resulted in alignment and coupling, increased
myocytes pulses at 1 Hz amplitude contractions and enhanced ultrastructural organisation.
Neonatal rat cardiomyocytes 0,23 and 0.43 V/mm, 1ms Stimulation enhanced elongation of both cell types and fibroblast
and NIH3T3 fibroblasts rectangular pulses at 1 Hz alignment on abraded surfaces. Topographical cues were stronger.
6 V/mm, 62.5 ms sawtooth Significant increase in proliferation in non-confluent and
pulses at 16 Hz for 5-18 days enhanced ECM related protein secretion in confluent cultures.
MC3T3-E1 clonal 0.002 V/mm, 300 μA/cm2 sine Stimulation enhanced proliferation and increased the levels
osteoblastic cells signal at 60kHz for 30min-24 h of TGF-β1.
Neonatal rat calvarial 1 G, 0.225 ms pulses in 4.5 ms Exposure significantly increased the number and size of deposited
osteoblasts bursts for 6 h bone like nodules. Enhanced BMP-2 and BMP-4 expression.
MLO-Y4 osteocyte-like cells 16 G sawtooth pulses in 4.5 ms Increased ALP activity,TGF-β1 and prostaglandin E2 expression,
ROS 17/2.8 cell line bursts at 15 Hz for 8h/day while osteocalcin or cell numbers went unchanged.
Osteosarcoma cell line Exposure limited the normal increase in cell numbers, while
ROS 17/2.8 enhanced ALP activity. Effects were cell-density dependent.
Osteosarcoma cell line Stimulation inhibited cell growth. ALP activity was dependent
ROS 17/2.8 on gap junctional coupling. Results suggest differential effects.
MC3T3-E1 clonal Conductive: 0.002 V/mm All three stimulation types increased DNA content in samples,
osteoblastic cells sinewave at 60Hz for 0.5-24h but only the capacitive one did it significantly and in an ever
Inductive: 22.5 ± 2.5 G pulses increasing manner. Signal transduction was thorugh Ca2+
in 15 Hz bursts for 0.5-24h influx with capacitive coupling, and by the intracellular
Combined: 340 ± 140 mG static release of the same ion with inductive and combined coupling.
and 370 ± 47 mG, 76.6 Hz
alternating field for 0.5-24h
[19]
0.0018 T at 30 Hz for 120 h
0.0018 T at 30 Hz for 12 or 72 h
[26]
[28]
[174]
[27]
Comparison of the modalities
Direct coupling
Inductive coupling
Rat PC-12 cells
Human ESC line H13 1 V/mm at 1 Hz
Human cutenous fibroblasts
Bovine primary osteoblasts
[225]
[154]
[132]
[134]
[197]
[184]
[159]
[186]
Capacitive coupling
63
1.5.4.5 Effects on the Cardiovascular System
One of the greatest limiting factors that hinder the use of engineered cardiac tissue, despite the
successes reported in animal models, is the risk of a potentially fatal arrhythmia due to the
insufficient quality of the implant [234]. Therefore it is essential for tissue engineers to find the
best possible culture conditions [234, 235]. Novel methods of controlling the structure of the
constructs, together with new techniques for assessing the quality of the engineered tissue before
implantation are required [234, 235]. ES is a promising modality that could potentially help
overcome these issues.
In cardiomyocyte monolayer cultures, ES accelerated the growth of cardiocytes, enhanced RNA
accumulation and raised levels of α-actin and myosin heavy chain, atrial natriuretic factor, myosin
light chain 2, connexin-43 and angiotensin II [235-238]. Furthermore, ES was found to promote the
faster maturation of cardiac cells and the greater structural organization of myofibrils [235, 237,
238]. In 3D constructs ES was demonstrated to enhance cellular alignment, mechanical and
electrical coupling and to help synronise the contraction of individual cells [235, 239]. Combining
perfusion with ES is a novel method in 3D cardiac tissue engineering that promises to yield further
benefits [235]. Connexin-43 levels, cell elongation and striation were shown to be further improved
by using the two modalities together [235].
Electricity also enables the quality assurance of engineered cardiac tissue-implants. Examining the
propagation of electrical waves in the engineered construct, together with the measurement of the
generated contraction force after exposure to various amplitude shocks, provide good quality
control information [234].
1.5.4.6 Skeletal Muscle Tissue Engineering
Electrical signals, for instance the impulses from the motor neurons [240], are known to be
essential for the development and maintenance of skeletal muscle [240, 241]. For example, it was
shown that, if skeletal muscle is denervated during in utero development, myotubes do not develop
fully [240].
64
ES can be used to mimick these electrical impulses. This has the added benefit of also forcing the
cultured muscle cells to contract, thus generating their own mechanical stimulation. This makes
this modality a very powerful tool for skeletal muscle engineering. Indeed ES was shown to
enhance both the physiology and the functionality of engineered skeletal muscle constructs [240].
Accelerated sarcomere assembly; enhanced gene expression of VEGF, CYCS, MCAD and PGC-1;
fast-to-slow phenotypic changes; increased contractile properties, fatigue resistance and raised
levels of extracellular adenosine have all been accredited to the beneficial effects of ES [240-246].
Furthermore, ES was observed to improve excitability and force production, thus hastening
maturation in 3D [240]. Similarly to cardiac muscle constructs, the quality of engineered skeletal
muscle can be well assessed with electricity: through measuring the contractility and excitability
during ES [247].
65
1.5.5 Stimulation through Conductive Scaffolds
In recent years, electrically conductive polymer scaffolds have been developed as a new means to
deliver an electrical stimulus [132, 133, 225]. In the future, these multi-functional scaffolds could
act as bioactive substrates for cell attachment, whilst providing a way to better regulate cellular
activities through evenly distributed well-controlled electrical signals both on the surface and
within the scaffold [132].
1.5.5.1 Conductive Polymers
First produced several decades ago [248], today there are over 25 conductive polymer systems
[249]. (For a list of conductive polymers see Table 1.5) They merge the positive properties of
metals and conventional polymers: the ability to conduct charge, great electrical and optical
properties with flexibility in processing and ease of synthesis [248, 249].
Table 1.5 – A list of conductive polymers and their abbreviations [250-253]. Table reproduced from
Balint et al, 2014 [254].
66
The early work on conductive polymers was triggered by the observation that the conductivity of
polyacetylene, a polymer that is normally only semiconducting at best, increases to 10 million fold
greater when polyacetylene is oxidised using iodine vapour [248, 255]. The underlying
phenomenon was named doping and is essential for the conductivity of polymers, as only through
this process do they gain their high conductivity [249]. As polyacetylene was difficult to
synthesise and is unstable in air, the search for a better conductive polymer began [255].
Polyheterocycles since then have emerged as a family of conductive polymers with both good
stability and high conductance [248]. This family contains the currently generally researched
conductive polymers: Polypyrrole (PPy), polyaniline (PANI), and polythiophenes (PTh) [257-262].
1.5.5.1.1 Polypyrrole
Arguably the most studied conductive polymer, reflected by the amount of publication surrounding
its properties and applications, is the conjugated polymer polypyrrole (Figure 1.10) [263-269].
Polypyrrole possesses many excellent qualities and stimulus-responsive properties that make it into
a very promising conductive biomaterial [263, 264]. Most importantly, it has good in vitro and in
vivo biocompatibility [269-272], good chemical stability in, for example, air and water [268, 273]
and has reasonably high conductivity under physiological conditions [269, 272, 274-276]. PPy can
be easily and flexibly synthesised in large quantities at room temperature in a wide range of
solvents including water [265, 269, 272, 277, 278]. It can be fabricated with a high surface area,
different porosities and can easily be modified to make it more suitable for biomedical applications
through the incorporation of bioactive molecules [265, 272, 274, 279, 280]. Additionally, PPy is
stimulus responsive allowing the dynamic control of its properties with the application of an
electrical potential [264, 275]. Unfortunately, polypyrrole is very difficult to further process once
synthesised [146, 268, 278, 281], as its molecular structure makes it non-thermoplastic [268, 271]
mechanically rigid, brittle [146, 271, 282] and insoluble after synthesis [282]. Beyond tissue
engineering [283], PPy today is used in fuel cells [277, 284], corrosion protection [285], computer
displays [284], microsurgical tools [284], biosensors [146, 286], drug delivery systems [146, 275],
neural probes [287], nerve guidance channels [287, 288] and blood conduits [289].
67
Figure 1.10 – The structure of PPy [250, 251, 290-292]. Little information is known of the structure of
most conductive polymers. This is a result of the difficulty to find a solvent that produces single
crystals of the polymer and the degradation of the polymer in x-ray diffraction studies [291, 292].
Figure reproduced from Balint et al, 2014 [254].
1.5.5.1.2 Polyaniline
The second most investigated conductive polymer after polypyrrole is polyaniline (PANI)
also known as aniline black [146, 255, 293]. It exists in various forms based on its oxidation level:
fully oxidized pernigraniline base, half-oxidized emeraldine base and fully reduced
leucoemeraldine base [146, 255] (Figure 1.11). Out of these, PANI emaraldine is the most stable
and conductive [146, 255]. Polyaniline has many advantages such as ease of synthesis, low cost,
good environmental stability and that it can be electrically switched between its conductive and
resistive states [294-298]. Unfortunately, its use in biological applications has been limited by its
low processibility, lack of flexibility, non-biodegradability and that it has been noted to cause
chronic inflammation once implanted [295, 299, 300]. PANI is currently investigated for biosensor,
68
neural probe, controlled drug delivery and tissue engineering applications [261, 301].
Figure 1.11 – The structure of PANI [252, 302-305]. Figure reproduced from Balint et al, 2014 [254].
1.5.5.1.3 Polythiophene Derivatives
A third very interesting conjugated polymer is Poly(3,4-ethylenedioxythiophene) (PEDOT), a
polythiophene derivative [146, 306]. PEDOT is formed by the polymerisation of the bicyclic
monomer 3,4-ethylenedioxythiophene (EDOT) [255]. Compared to polythiophene, PEDOT has a
dioxyalkylene bridging group across the 3- and 4-positions of its heterocyclic ring (Figure 1.12),
greatly improving its properties: lowering its bandgap, reduction and oxidation potential [255,
307]. This is also what grants PEDOT it’s good electrical, chemical and environmental stability
[306] and a better conductivity and thermal stability then that of polypyrrole’s [255, 306].
Figure 1.12 – The structure of PEDOT [310]. Figure reproduced from Balint et al, 2014 [254].
69
Today PEDOT is used in bio-sensing and bioengineering applications [146], for example in neural
electrodes [255, 306, 308], nerve grafts and heart muscle patches [306]. In one interesting example,
a neural electrode was interfaced with the surrounding brain tissue through the in situ
polymerisation of PEDOT [309]. This formed PEDOT filaments extending away from the
electrode, far enough to breach the glial scar around the electrode, forming sensitive contacts with
the plasma membrane of the neurons [309]. PEDOT has also been polymerised within acellular
muscle tissue where it formed a network of elongated tubular structures throughout the tissue [306]
in essence converting it into an extensive conductive three dimensional substrate.
1.5.5.2 Electrical Stimulation through the Scaffold
The most commonly reported result of ES delivered with this technique is greater neurite
outgrowth in nerve cells [261, 266, 281, 311, 312]. For example, PC12 cells have been reported to
have 50% more neurites on NGF doped PPy films when ES was applied [279]. Similarly, PC12
stimulated with 10 mV/cm on PPy–PLGA scaffolds formed an increased number of neurites and
the overall length of these neurites was also greater than without stimulation [257]. Nerve stem
cells stimulated with 100 mV/mm for 60 min on PLLA/PANi scaffolds also showed enhanced
neurite formation [297]. The reason behind this increase of neurite outgrowth has been postulated
to be the enhanced fibronectin adsorption onto the conductive scaffold [313]. ES has been
demonstrated to have other desirable effects on nerve cells: 250 Hz biphasic current delivered via
PPy/PMAS composite films was observed to increase neural differentiation in the presence of NGF
[261]. Similarly, enhanced proliferation and neurite outgrowth was noted when neural cells were
stimulated on nanofibrous PANI-PG scaffolds [314]. Schwann cells cultured on Chitosan-PPy
composites showed greater viability and increased their NGF and BDNF mRNA expression when a
stimulus of 100 mV/mm was applied [315].
Other cell and tissue types could also benefit from electrical stimuli delivered through a conductive
scaffold. The growth of NIH-3T3 fibroblasts was increased by ES delivered through a PANI based
electrospun scaffold [316]. This effect on fibroblast proliferation was also observed with DC
70
current stimulation delivered through PPy-PDLLA scaffolds [317]. In another study, human
cutaneous fibroblasts stimulated on PPy-PLLA films showed greatly increased viability,
mitochondrial activity and IL-6 and IL-8 secretion [318, 319]. Aortic endothelial cells cultured on
fibronectin coated PPy spread out when an oxidising potential was applied and were rounded and
synthesised DNA to a lesser extent when the polymer was reduced to its neutral state [248].
These beneficial effects have been theorised to be the result of negative ion release and positive ion
(e.g. Na+) uptake into the polymer [248, 313], electrophoretic redistribution of cell surface
receptors [313] and the increased adsorption of ECM molecules like fibronectin onto the polymer
[269, 320].
1.5.5.3 Further Approaches to Delivering an Electrical Stimulus through a Biomaterial
1.5.5.3.1 Electrets
Electrets are very interesting electrically active biomaterials that require no power source to deliver
an electrical stimulus [321, 322]. Electrets are dielectric materials that, as a result of containing
trapped monopolar charge carriers, maintain a quasi-permanent electrical charge [321, 322]. They
can be created, for example, from PTFE through corona poling with various charge densities and
polarity [321, 322]. Electrets were demonstrated to be able to affect the cellular orientation, protein
synthesis and absorption of chick embryonic fibroblasts [321]. In a peripheral nervous system
(PNS) regeneration study, PTFE electret nerve guidance channels were observed to enhance the
number of myelinated axons and to enhance the cross-sectional area of the nerve compared to
uncharged PTFE tubes [322].
1.5.5.3.2 Piezoelectric Polymeric Materials
Similarly to electrets, piezoelectric polymers require no external power source to generate an
electrical stimulus [146, 323, 324]. These materials when exposed to mechanical strain generate a
transient surface charge on their surface [146, 322]. One example is polyvinylidene fluoride
(PVDF), a synthetic, semi-crystalline polymer that can be rendered piezoelectric by mechanically
71
stretching it under an intense electric field followed by corona poling [324]. PVDF has been stated
to possess excellent biocompatibility [323], but requires extensive processing and grants no ability
to control the stimulus [146]. Another example is vinylidenefluoide-trifluoroethylene, which unlike
PVDF, does not require mechanical stretching, and thus allows the creation of more complex
piezoelectric biomaterial structures [323]. These piezoelectric polymers are currently used in PNS
regeneration and vascular grafts [146, 324]. PVDF tubes in transacted sciatic nerve models in mice
and rats contained greater number of myelinated axons compared to their non-piezoelectric controls
[323, 324], while neuroblastoma cells cultured on piezoelectric PVDF surfaces displayed enhanced
neuronal differentiation [323].
1.5.5.3.3 Photovoltaic Polymers
Photocurrent stimulation has been proposed as a novel method for delivering an electrical stimulus
[325]: Light shone on a photovoltaic material, for example poly (3-hexylthiophene) (P3HT),
poly(p-phenylene or chlorophyll, will release electrons from the material and can be used to
generate an electrical stimulus [325]. Electrospun photosensitive P3HT–PCL nanofiber composites
were demonstrated to support human bone marrow-derived MSCs and have great potential for
future regenerative medicine application [325].
72
1.5.6 Future Possibilities in Electrical Stimulation
ES has the potential of alleviating some of the problems that currently prevail in tissue
engineering, cancer treatment, regenerative medicine and other bio-medical scientific fields. This
physical stimulatory method could take advantage of the galvanotactic property of cells, directing,
concentrating and isolating them. For example, cells from explants could be isolated through their
various migratory speeds [137] or polarity preferences. Similarly, the localisation of cells could be
controlled in tissue engineering constructs: Different cells in co-cultures could be selectively
controlled and distributed [134]. Furthermore, in 3D constructs, cells could be guided away from
the nutrient rich surface, where they would normally try to position themselves [326], into the
deeper regions of the scaffold resulting in better quality implants. ES could help in growing aligned
and orientated tissue, opening up new possibilities in 3D tissue and organ engineering and allowing
greater complexity in design of the tissue engineered constructs [137]. Problems such as the
inappropriate colonisation of lens epithelial cells after lens or cornea transplants, or when
engineering these organs, could be avoided by replacing the natural electric fields with artificial
ones [135, 136].
ES has the potential to provide better control over the in vitro and in vivo proliferation and
differentiation of cells and the properties of the resultant tissue, both spatially and temporally [132].
Stem cell fate could be defined more precisely, directing them to specific lineages, yielding a more
homogenous cell culture and making their therapeutic application more plausible [139, 151, 206].
Cell proliferation, differentiation [134] and ECM deposition could be enhanced or withheld in
greater geometrical complexity than with any other stimuli (i.e. chemical, mechanical) giving tissue
engineered products that are more similar to their natural counterparts in architecture and function.
The ES technique also has the potential to give investigators the ability to regulate their construct
not just in culture, but also after implantation, therefore, allowing faster and better integration.
.
73
Although there is great potential in controlling cellular behaviour through the direct delivery of
growth factors, problems associated with sustaining the proper level of dosing for longer periods of
time and in a spatially complex way limit the technique’s usefulness. High initial doses raise the
question of toxicity, while successive delivery of smaller amounts is difficult to carry out in vitro
and, even more so, in vivo. Incorporating growth factors into scaffolds, offers a possible solution to
these problems, although good control over the release of these bounded chemicals has yet to be
been achieved [178, 326]. Increasing growth factors levels with genetic manipulation is difficult
and its usefulness is restricted by the numerous technical, ethical and legal problems surrounding it
[326].
As an alternative, ES offers a cheaper, simpler and more flexible way to deliver growth factors by
inducing the cells themselves to produce these materials through natural pathways. Another
advantage of ES is that, unlike treatments that involve incorporated growth factors and gene
therapy, its effects disappear with the discontinuation of the treatment. This has obvious benefits
for clinical usage allowing the clinician better control of the treatment and acting as a safety net to
mitigate post-treatment complications. ES can be used synergistically with other techniques,
reducing the required levels of expensive growth factors [132, 206] and/or other stimuli and,
therefore, the cost of whole process. Furthermore, tissue engineered products could be created with
greater speed, bringing this scientific discipline one step closer to the ultimate goal of widespread
therapeutic application [147, 206, 326]. It is also possible to create programmable, multiple
electrode bioreactor systems that generate complex electric fields customised to the needs of the
particular treatment case. Stimulating in vitro and/or in vivo 2D/3D constructs, with or without
conductive scaffolds, could provide a cheap, flexible and relatively simple cellular enhancement
technique with possible usefulness in mass production. As such, with increased research and
further understanding, electrical stimuli have significant potential for the field of practical tissue
engineering.
74
1.6 BIOREACTORS
In general bioreactors are devices that enable biological or biochemical processes to develop under
tightly monitored conditions [327]. Through the use of these apparatuses temperature, pH, gas
content and nutrient supply can be precisely controlled during cell maintenance [327, 328],
allowing ex vivo culture to be carried out with greater efficiency, reproducibility, scalability and
reduced financial cost compared to traditional in vitro techniques [327, 328]. The enhanced culture
conditions provided by bioreactors are especially useful for 3D tissue cultures, where the traditional
methods have been shown to be inadaquate [329, 330].
Some bioreactors are employed to deliver dynamic culture conditions [329-331]. Examples of these
are the spinner flasks and the rotating wall bioreactors [329-331]. Dynamic, compared to static,
culture was demonstrated to provide a much more homogenous distribution of cells within
scaffolds, together with improved oxygen supply and better mass transport [327, 330, 331]. Mass
transport can be further enhanced by using hollow fibre bioreactors, if necessary, for example for
the culture of high metabolic activity liver cells [327]. Perfusion bioreactors that enable the culture
medium to perfuse homogenously through scaffolds are also widely employed [330].
In tissue engineering bioreactors are not only applied to improve culture conditions, but also to
better mimic the in vivo environment through the application of some form of physical stimulus
[327, 328, 332]. In the literature many examples of these can be found, for instance mechanical
compression bioreactors for bone tissue engineering [333], tension systems for skeletal muscle
constructs [334], and perfusion-based shear stress devices for bone [335] and vascular engineering
[336].
Bioreactors that enable ES to be delivered to cells have also been built. In the next subsection the
desireable properties of ES bioreactors will be discussed. This will be followed by the examination
of the electrical bioreactor desings that have been described in detail in the literature in the light of
these criteria.
75
1.6.1 Electrical Stimulation Bioreactors
The desirable properties of an ES bioreactor are:
Capable of long term culture: In order to suit the needs of tissue engineering applications the
bioreactor should allow cells and tissue constructs to be cultured for weeks or even months.
Low risk of infection: The design should allow the bioreactor to be efficiently sterilised, for
example through the use of autoclaving. The design should also ensure that sterility is
maintained throughout the experiments carried out using the bioreactor.
Reliable: The bioreactor should deliver the same performance throughout its useful lifetime.
Robust: The useful lifetime of the bioreactor should be reasonably long, i.e. in the range of
years. Therefore bioreactors should not be susceptible to faults and should be able to withstand
accidental damage and normal wear and tear as a result of its operation.
Reusable: Connected to the above two criteria, the bioreactor’s design should enable it to
undergo repeated cycles of use (including for example the sterilisation procedures).
Easy to use/handle: A complex or difficult to use design makes experiments more
complicated and lengthy, while increasing the likelyhood of human error. Ease of use is
therefore essential.
Scalable: It is sometimes necessary to deliver stimulation to multiple bioreactors at the same
time. The design should enable this.
Homogenous stimulation: The same stimulus should be delivered to all cells cultured in the
bioreactors.
Not biasing: Cell should not show altered behaviour as a result of being cultured in the
bioreactor compared to traditional in vitro techniques.
1.6.1.1 Agarose Bridges
Agarose bridges are widely used to deliver direct ES, mostly in galvanotaxis studies (Figure1.13A).
Electrodes are placed into separate beakers containing an ionic liquid. The ES from the beakers is
conveyed to the culture environment through agarose bridges. Electrochemical changes are isolated
76
to the beakers, allowing cells to be stimulated using the direct method while avoiding its drawbacks
[337].
Agarose bridge bioreactors are relatively simple systems, using simple components, that effectively
deliver direct stimulation. However ensuring long term reliable and infection free culture and
stimulation can be difficult. Scalability, ease of handling and robustness are also questionable.
1.6.1.2 Bioreactor for Skeletal Muscle Tissue Engineering
Donnelly et al [338] described a bioreactor for stimulating skeletal muscle constructs
(Figure1.13B). Direct stimulation was delivered through U-shaped stainless steel electrodes
embedded into polystyrene plate lids to cultures in a 6-well configuration.
The Donnelly bioreactor is likely to be more suitable for long term cultures and possesses better
scalability compared to the agarose bridge based system. However, this bioreactor is not
autoclavable. As such the plate lids cannot be reused without the risk of an infection. Additionally,
the bioreactor in its current format does not delivery a homogenous stimulus or lends itself to the
stimulation of monolayer cultures.
1.6.1.3 Cardiac Muscle Bioreactor
Radisic’s group designed a bioreactor for a similar application [339]. In their system carbon rods
were placed to the two sides of samples in Petri dishes (Figure1.13C). The wires connecting the
carbon electrodes to the signal source were isolated using copious amounts of silicon glue.
Based on the available information, sample size, scalability and easy of handling are likely to be
limited. Similarly to before, the bioreactor is not autoclavable. This bioreactor is also inefficient in
the sense that only a small proportion of the available culture environment inside the Petri dishes is
exposed to stimulation.
77
Fig
ure
1.1
3 –
Th
e el
ectr
ica
l st
imu
lati
on
bio
rea
cto
rs d
escr
ibed
in
th
e li
tera
ture
: A
– T
he
ag
aro
se b
rid
ge
con
fig
ura
tio
n [
33
7],
B –
Th
e D
on
elly
bio
rea
cto
r [3
38
], C
– C
ard
iac
mu
scle
bio
rea
cto
r [3
39
], D
– B
iph
asi
c cu
rren
t st
imu
lato
r [2
5],
E –
Th
e C
-Pa
ce s
yst
em
(Im
ag
es w
ere
rep
rod
uce
d w
ith
th
e p
erm
issi
on
of
the
pu
bli
sher
s. I
ma
ge
of
the
C-P
ace
sy
stem
is
the
pro
per
ty o
f Io
nO
pti
x L
LC
.)
78
1.6.1.4 Biphasic Current Stimulator
In Kim et al’s solution culture chambers were milled into PTFE blocks in a 6-well configuration
[25] (Figure1.13D). Cells were cultured on gold plated silicon electrodes placed into these wells.
Gold plated reference electrodes were submerged into the culture medium above the cells. This
device delivers direct stimulation homogenously, perpendicular to the cell monolayer.
This bioreactor is comprised of highly biocompatible materials and appears to be at least partly
autoclavable. It enables long term stimulation efficiently with low risks of infection to a relatively
large number of samples. Depending on the robustness of the electrode discs and the wires
connecting them, the bioreactor might be difficult to handle and reuse.
1.6.1.5 C-Pace Stimulators
The C-Pace bioreactors, designed by Ionoptix LLC, is the only electrical stimulation system
commercially available to the author’s knowledge (Figure1.13E). This system delivers direct
electrical stimulation through carbon electrodes to samples cultured in traditional polystyrene well
plates. The rectangular carbon electrodes are soldered onto circuit boards embedded in
polycarbonate plate lids.
This system is robust, easy to use and allows the culture of cells with the same standards as in any
normal well plate. However it also has significant drawbacks. The bioreactors can only be used for
24-48h, after which the electrodes have to be removed, cleaned and placed into deionised water to
equilibrate. The bioreactor cannot be autoclaved without the serious risk of damaging it; hence
ensuring sterility with repeated applications is difficult. These two factors seriously limit the
usability of this system.
In summary, there are many interesting solutions described in the literature for the delivery of ES
in vitro. However, none of these truly fulfil the desirable criteria of an ES bioreactor. It must also
be noted that all of the above bioreactors are for the delivery of direct electrical stimuli, rather than
for the capacitive or the inductive methods. Apart from the agarose-bridge configuration, none of
79
the discussed bioreactors prevented the electrochemical side-effects of direct stimulation. The use
of an indirect coupling method could have helped avoid any bias from these unwanted side-effects.
80
1.7 CONCLUSIONS, AIMS AND OBJECTIVES
In this introductory chapter it has been established that ES can influence the behaviour of many cell
and tissue types in a manner that could be very useful for tissue engineering: Enhanced
proliferation, differentiation and the controlled migration of cells have all been achieved using ES.
The aim of this study is to develop electrical stimulation into an effective tool for the
engineering of tissue, with enhanced bone formation from bone-marrow derived
mesenchymal stem cells being the main focus.
The efficient delivery of the ES will require a bioreactor to be used. The criterias of an “ideal” ES
bioreactor (i.e. low risk of infection, ease of handling, scalability etc.) can probably be best fulfilled
through a simple, but efficient design, based on the 6-well plate configuration.
What type of bioreactor should be used? Out of the three most widely applied delivery methods of
stimulation (i.e. direct, capactive and inductive), the capacitive technique offers the most
advantages. Although direct coupling has been primarily applied in bioreactors discussed in the
literature, the use of this technique has the inherent risk of biasing the stimulation through
electrochemical side-effects. Considering that at the cellular level direct coupling and CC act in a
similar manner, CC offers an alternative with the same benefits as direct coupling while avoiding
its drawbacks. The use of CC is further promoted by the fact that controlling (and calculating the
amount of) the induced small magnitude electrical potentials/currents generated through the
inductive method could potentially be difficult. Additionally, the use of IC might be cumbersome
as this will require large coils to be placed around the sample. Furthermore the electromagnetic
field from the coils will not be isolated to the sample, and will potentially affect the neighbouring
equipment.
81
Therefore the first objective of this study should be to design and build a capacitive electrical
stimulation bioreactor based on the 6-well plate configuration that allows the delivery of an
electrical signal homogenously to monolayer and 2D/3D scaffold cultures. The design should
isolate any negative side-effects from the bioreactor and the stimulation as much as possible.
Furthermore, it should enable the easy, scalable use of the bioreactor with low-risk of infection.
The same ES parameters (e.g. electrode potential, electrical current) will result in a different
stimulus at the cellular level based on the design of a bioreactor. In order to calculate this stimulus,
the electric field generated by the bioreactor has to be evaluated. This is essential not only for the
comparison of different bioreactor designs, but also to enable the experimental results of this study
to be compared to those presented in the literature. Therefore, computerised simulations should be
carried out to determine the exact stimulus received by the cells.
A large range of parameter combinations are possible in ES. Which ones should be tested? From
the literature it can be inferred that high electric field strength is generally better; high electrical
current acts as a limiting factor; no specific pulse width has been shown to be superior; while
frequency in the range of 100 Hz is postulated to be the most beneficial for bone.
ES has been demonstrated to promote the proliferation of bone cells. As such it may be possible to
identify a parameter combination - a regime - that can promote hMSC proliferation. One such
regime could help address cell number scale up issues due to cell biopsy limitations that exist in
tissue engineering.
Electrical signals have been found to promote bone differentiation through increased ALP activity,
mineralisation and the expression of bone related genes such as BMP-2. Therefore, it might be
possible to identify an electrical stimulating regime that has the capability to promote hMSC
differentiation and ECM production. Tissue constructs with increased mechanical integrity and
shortened in vitro culture time could be produced with one such regime.
82
Therefore, the objectives of this thesis are:
Objective #1
Design and build a capacitive electrical stimulation bioreactor based on the 6-well
plate configuration that allows the delivery of an electrical signal homogenously to
monolayer and 2D/3D scaffold cultures.
Objective #2
Carry out computerised simulations to determine the exact stimulus received by the
cells.
Objective #3
Identify a parameter combination - a regime - that can promote hMSC proliferation.
Objective #4
Identify an electrical stimulating regime that has the capability to promote hMSC
differentiation and ECM production.
The next chapter of this thesis, Chapter 2, will discuss the work undertaken to accomplish
Objective #1. Chapter 3 is devoted to the second objective. Objective #3 and #4 are addressed in
the fourth chapter.
83
Chapter II
Bioreactor Design
“Make everything as simple as possible, but not simpler.”
Albert Einstein
84
2.1 INTRODUCTION
At the end of the previous chapter the objectives of this thesis were set out. The first amongst these
is to “design and build a capacitive electrical stimulation bioreactor based on the 6-well plate
configuration that allows the delivery of an electrical signal homogenously to monolayer and
2D/3D scaffold cultures”. This chapter is concerned with the work undertaken in order to
accomplish this.
An ideal electrical stimulation bioreactor should allow the long term culture of cells and tissues
with low risk of contamination, ease of use/handling and good scalability [328]. Robustness and
reusability are also essential. Furthermore, it is also desirable that the stimulation be delivered
equivalently to all cells in the culture and done so with minimal bias from electrochemical side-
effects.
In order to achieve a bioreactor design that fulfils all of these criteria four successive generations of
capacitive bioreactors were created, each building on the experience gained from the previous
iteration. In the following chapter the various design generations will be examined from the
perspective of the above criteria. Possibilities for future improvement will also be presented.
The first section of this chapter is however concerned with the direct electrical stimulation
bioreactor created during the author’s MSc project work, as this bioreactor and the experience
gained during the project had a significant influence of the capacitive bioreactor designs that
followed.
85
2.2 A BIOREACTOR FOR DIRECT ELECTRICAL
STIMULATION
Design developed during the MSc project at the Keele University titled “Influence of the ‘PhyBack
System’ on Primary Human Mesenchymal Stem Cell Activity”
The aim of the work was to assess the effect of electrical stimuli, simple and complex, generated by
a medical device, the Phyback system, upon the proliferation and gene expression of human
mesenchymal stem cells. To this end an inexpensive bioreactor was developed that can sustain cell
cultures for long periods of time and allows electrical stimulation to be efficiently delivered. The
stimulation was decided to be delivered “directly” (i.e. the electrodes are in contact with the culture
medium and there is a conductive connection between them), as this was the simplest method to
use. The electrodes were chosen to be needle-like to minimise their presence (and any issues with
biocompatibility) inside the culture environment (Figure 2.1).
Figure 2.1 – The initial concepts: Sketches of the first ideas for the direct stimulation bioreactor
It was decided that the best solution is to deliver the stimulation through specialty wires (with a
thickness in the 100 µm range) to cells cultured in standard 6-well plates (to allow a large number
of samples to be processed at the same time).
86
2.2.1 Materials and Methods Used in Building the Bioreactor
Figure 2.2 - The initial concept and a top view of the complete direct electrical bioreactor
Small diameter holes were melted into polystyrene 6-well plate lids above the culture area 28 mm
apart, using a heated 500 µm diameter syringe needle. Electrodes were cut from 125 µm stainless
steel wire (SS31605, World Precision Instruments Ltd.) in the necessary length, bent to the suitable
(U-like) shape and placed into the plate lid. Thin copper wires were wrapped around a part of the
stainless steel electrodes on the outer side of the plate lid. These connecting areas were forced
together in terminal blocks in order to maintain a good electrical connection under all
circumstances (e.g. high humidity, thermal expansion) (Figure 2.2). Wires were taped down for
safety and to enable better handling. Finally, crocodile clips were soldered to the end of the copper
wires to allow easy connection to the stimulation source.
2.2.2 The Lessons Learned
The stimulation delivered through the direct electrical stimulation bioreactor was able to alter the
behaviour (e.g. proliferation, gene expression) of hMSCs. However, these findings were potentially
biased by the chemical changes in the culture medium caused by the electrical stimulation. The
electric field generated in the bioreactor was inhomogeneous, thus cells at different locations in the
culture environment received a different stimulus. Additionally, although the bioreactor was
inexpensive to build, it proved to be one-use, as re-sterilising it was difficult and unreliable.
87
Therefore, although electrical stimulation was demonstrated to be promising and the 6-well plate
based design was relatively easy to use, a new method of delivering the stimulation had to be
found.
N.B. Subsequent iterations of the bioreactor design took place as a part of this PhD thesis.
88
2.3 THE FIRST GENERATION OF CAPACITIVE
BIOREACTORS
As the first attempt to create a capacitive bioreactor a simple design was conceived. One that is
built up of inexpensive parts, manufactured from material already found in a laboratory. This
approach allowed experience to be gained quickly on the basic design needs of a capacitive
bioreactor, while requiring minimal financial and temporal investment.
As discussed in the previous chapter, when using capacitive coupling, electrodes are placed usually
above and below the cell culture. The electrodes have to be electrically isolated so that there would
be no electrical current flowing between them. In order to maximise the efficiency of the
stimulation (i.e. maximise the electric field strength generated per volt of electrode potential) the
electrodes have to be placed as close to each other as possible. As such it is generally necessary to
place at least one of the electrodes inside the culture environment.
2.3.1 The Bioreactor
To serve as upper electrodes, 31 mm discs were punched out of 0.4 mm thick sheets of 99.9% pure
copper (RS stock number: 680-965; RS Components Ltd.). Copper was chosen for its high
conductivity, low price, easy machinability, and as it can be easily soldered. Two 1 mm diameter
holes were drilled into the centre of each copper disc 4 mm apart. Through these, 0.71 mm thick
enamelled copper wire (YN83E, Maplin Ltd.) was hooked trough. In order to ensure good
connection under all circumstances the wires were soldered to the electrodes.
To serve as the counter electrode a single, large, rectangular 135 x 90 mm copper sheet was cut out.
Having a single large counter electrode simplifies the design, makes the handling of the bioreactor
easier, and helps to avoid the “edge effects” (inhomogeneity in the electric field due to the
proximity of the edge of the electrode) (Figure 2.3).
89
Figure 2.3 – Computer simulation of the
electric field strength between two
electrodes showing an example of the edge
effect. The electric field strength is higher
(indicated by the red colour) at the edge of
the electrode.
Both electrode types were coated with
clear lacquer (RS 569-307, RS
Components Ltd.), in order to act as an
electrical insulation layer and to protect
the copper from corrosion in the humid environment of the cell culture incubator. This layer was
also necessary to protect the culture medium, and thus the cells, from leached-out copper ions and
corrosion products.
The electrodes were positioned 6.3 mm apart from each other, the disc electrodes inside the culture
environment above the cells, the counter electrode below the 6-well plate (Figure 2.4). This
distance of the electrodes allows cells to be kept in 3.5 ml culture medium while still leaving a 1
mm air gap between the upper electrode and the surface of the culture medium. When choosing this
distance a compromise had to be made as, if the electrodes are positioned closer to each other the
stimulation will be more "efficient" (i.e. a stronger electric field is generated with the same
electrode potential), but less culture medium can be placed into the well. With less culture medium
there is a smaller buffer protecting the cells from any negative side-effects (e.g. copper leached out
from the electrodes).
The disc electrodes were held in position using 12 mm high polypropylene sections cut out from 50
ml centrifuge tubes. The lacquer coated disc electrodes, the polypropylene rings and the plate lid
were glued together using high quality silicon glue (WP Instruments, Co.). The six wires coming
from each well were soldered together and to crocodile clips. The crocodile clips were connected
to the “poles” of the “stimulation stage” (Figure 2.4).
90
Figure 2.4 – Exploded view of the first generation capacitive bioreactor with the stimulation stage.
1 – Copper wires connecting the upper electrodes to one of “poles” of the stimulation stage
2 – Polystyrene plate lid
3 – Polypropylene sections
4 – Copper disc upper electrodes
5 – Polystyrene 6-well plate
6 – Copper counter electrode connected to the other “pole” of the stimulation stage
7 – Stainless steel pole on the stimulation stage
8 – PTFE stimulation stage
91
Figure 2.5 – The stimulation stage
The stimulation stage (Figure 2.5), two 15 cm long
stainless steel threaded rods (the “poles”) on a thick sheet
of plastic, was built to allow the fast exchange of the
bioreactors and their insulation from the incubator’s
metallic parts. The poles of the stimulation stage were connected to the signal source. Electrical
signals were generated using a TG5011 arbitrary function generator (Thurlby Thandar Instruments
Ltd.) and a WA301 amplifier (Thurlby Thandar Instruments Ltd.).
2.3.2 Evaluation of the First Generation Bioreactor
The bioreactor was able to maintain a cell culture
up to a 7 day time point. However, upon placing
the bioreactor lid on the 6-well plate, the culture
medium immediately stuck to the disc electrodes
due to the hydrophilic nature of the lacquer
coating (Figure 2.6). This resulted in an uneven
culture medium surface, and therefore in
inhomogenous stimulation. This also possibly
allowed cytotoxic elements to be leached-out from
the electrodes into the culture medium.
The bioreactor lids were successfully sterilised initially, through a combination of immersion into
70% ethanol and UV irradiation. However, this was not sufficient to remove contaminants in
between uses, resulting in an infection when the bioreactor lids were reused. This was probably due
to a small amount of medium remaining on the surface of the electrodes or in hard to access areas
Table 2.1 – The criteria of an ideal ES
bioreactor – First generation bioreactor
Capable of long term culture?
Low risk of infection?
Reliable?
Robust?
Reusable?
Easy to use/handle?
Scalable?
Homogenous stimulation?
Not biasing?
92
even after cleaning, allowing infectious agents to survive and proliferate there between uses. Some
of these contaminants may have survived the sterilisation procedure and infected the cell cultures
during experiments. This in effect rendered the bioreactors one-use.
Figure 2.6 – Sketch of the medium “sticking” to the electrodes
Additionally, wires and soldering were prone to break as a result of repeated use, further
compromising the reliability and reusability of this design.
In summary, even so this inexpensive bioreactor design (Figure 2.7) was able to maintain a cell
culture long term in a scalable and easy-to-use manner, it failed to fulfil the other criteria. A better
design had to be sought.
Figure 2.7 – A completed first generation bioreactor from above (left) and the copper electrodes on the
plate lid from below (right).
93
2.4 THE SECOND GENERATION
The next generation of the capacitive bioreactor sought to answer the question “Is it necessary to
place electrodes into the culture environment or is it sufficient to place electrodes simply above and
below the well plate?”
2.4.1 The Bioreactor
To answer the question experiments were
carried out, where a single, lacquer coated
135x90 mm copper electrode was placed
above and below standard 6-well plates
(Figure 2.8). The electrodes were
connected to the stimulation stage through
soldered enamel wires and crocodile clips.
The same stimulation stage and signal
source was used as with the previous
bioreactor.
2.4.2 Evaluation of the Second Generation Bioreactor
Although, robustness and reliability were an issue, as
the soldering and the wires were prone to break, this
approach fulfilled nearly all the criteria. However,
this design has one serious limitation: the
inefficiency of the electrical stimulation. In order to
achieve the same electric field strength as with the
first generation bioreactor, an approx. 15 times
higher electrode potential is required (see Chapter 3
Figure 2.8 – The second generation bioreactor with the
two large rectangular electrodes. 1 – 6-well plate
between the electrodes, 2 – Rectangular copper
electrode.
Table 2.2 – The criteria of an ideal
bioreactor – Second generation bioreactor
Capable of long term culture?
Low risk of infection?
Reliable?
Robust?
Reusable?
Easy to use/handle?
Scalable?
Homogenous stimulation?
Not biasing?
94
– Computer simulations). I.e. where the first generation bioreactor required 15 V, this design
requires 225 V. As the signal source has a limited voltage range (max. 15 V), this greatly limits the
range of electric field strengths that can be tested. Considering that electric field strength is a very
important parameter, one that defines whether an effect is present and if present how strong it is,
this is indeed a serious issue that compromises the usefulness of the bioreactor.
In summary, although capacitive electrical stimulation can be delivered through electrodes placed
above and below a well plate, this is a very inefficient way of doing so. As such placing the
electrodes as close to each other as possible, and therefore into the culture environment, is
necessary.
Figure 2.9 – Exploded view of the second generation bioreactor.
1 – Polystyrene 6-well plate, 2 – Rectangular copper electrode.
95
2.5 THE THIRD GENERATION CAPACITIVE
BIOREACTOR
From the previous two generations the following lessons were learned:
Placing the electrodes as close to each other and therefore into the culture environment is
essential, especially considering the limited output capabilities of the current signal
generator.
The upper electrodes will have to be slightly submerged into the culture medium. This is to
avoid any inhomogeneities in the stimulation caused by the medium “sticking” to the
electrode surface (Figure 2.10).
Figure 2.10 – An uneven culture medium surface can be avoided by submerging the electrodes
If the upper electrodes are in contact with the culture medium:
Chemical and electrical isolation of the electrodes from the culture medium has to be
solved.
Reliable sterilisation of the bioreactor is even more important than before. The bioreactor
has to be autoclavable as alternative sterilisation methods, such as submerging in 70%
ethanol, freezing and dry heat sterilisation, were found to be ineffective or impractical.
In order to chemically and electrically isolate the upper disc electrodes from the culture
environment the following approaches were tried:
1. Coating:
a. Lacquer based coatings (Clear lacquer, RS 569-307, RS Components Ltd.)
96
b. Silicon based coatings (Silicon mold release spray, 250-6904, RS Components
Ltd.)
c. Polytetrafluoroethylene (PTFE) based coatings (Dry PTFE Lube FG, product code:
6150009520, Ambersil Ltd.)
2. Embedding:
a. Polyurethane based
b. Araldite
c. High temperature epoxy – The solution that was chosen – Efficient moulding
technique had to be developed (High temperature epoxy encapsulant,
CW1302/HY1300, Araldite, Huntsman Advanced Materials GmbH)
None of the coatings was able to fully chemically isolate the copper electrode or withstand the
hydrolysis effects of the autoclave. Polyurethane based encapsulation was difficult to make, while
the araldite based embedding was unable to withstand autoclave temperatures. High temperature
epoxy on the other hand was relatively easy to work with and was able to endure autoclaving.
Therefore this embedding material was chosen. In order to further improve the biocompatibility of
the upper electrodes, the epoxy on the cell culture facing side of the electrodes was replaced with a
PTFE disc (Figure 2.11).
Figure 2.11 – Engineering drawing (left), 3D model (middle) and photography (right) of a third
generation bioreactor upper electrode. 1 – Stainless steel machine screw, 2 – Epoxy embedding
material, 3 – Stainless steel wire: Allows the checking of the electrical connectivity to the electrode disc
even after embedding.
97
2.5.1 The Bioreactor
29 mm diameter discs were water-jet cut from 0.4 mm thick sheets of 99.9% pure copper (RS stock
number: 680-965; RS Components Ltd.). M4x40 A4 stainless steel machine screws (Buckfast
Tools Ltd.) were point-welded into the centre of the copper discs. A 10 mm piece of stainless steel
wire was also point-welded to the copper discs. This was done to enable the verification of
electrical conductance to the electrodes post-embedding.
The electrode assembly was placed onto the centre of a ø29x1.5mm PTFE disc (lathed to diameter
and cut from PTFE plastic rod, RS Stock No. 752-587) and embedded in a high temperature epoxy
encapsulant (CW1302/HY1300, Araldite, Huntsman Advanced Materials GmbH) (Figure 2.12).
Bioreactor lids were milled out of 20 mm thick PTFE sheets (RS Stock No. 197-0102, RS
Components Ltd.). The electrode assemblies were placed into the bioreactor lids, held in position
by M4 stainless steel nuts and spring washers (Buckfast Tools Ltd.) (Figure 2.13). The electrodes
were connected to the stimulation stage through crocodile clips (1698998, Farnell Ltd.) soldered
onto copper wires. Concept and engineering drawings were created in the commercial software
SolidWorks 2008. Engineering drawings for the various parts can be found in the Appendix.
Figure 2.12 – Photograph of a third generation bioreactor lid upside down.
1 – PTFE lid,
2 – Third generation electrode assembly.
98
Figure 2.13 – Exploded view of the third generation bioreactor lid with a 6-well plate bottom.
1 – M4 stainless steel nut
2 – Stainless steel spring washer
3 – PTFE bioreactor lid
4 – Third generation electrode assembly
5 – Polystyrene 6-well plate bottom.
99
2.5.2 Evaluation of the Third Generation Bioreactor
Third generation bioreactors were found to be able
to maintain a cell culture up to a 7 day time point
with very little risk of infection. The design also
enabled multiple plates to be easily stimulated at
the same time. Slightly submerging the electrode
assemblies into the culture medium solved the
problem of delivering a homogenous stimulation.
Table 2.3 – The criteria of an ideal ES bioreactor – Third generation bioreactor
However, this generation too had a crucial flaw. Although all of the components and materials used
in bioreactor are able to withstand the humidified 121 °C atmosphere of the autoclave, cracks
formed in the embedding material with repeated cycles of sterilisation. This was due to the
difference in the thermal expansion coefficient between the various materials used in the electrode
assembly, mainly the epoxy and the PTFE disc. Through further cycles of autoclaving the cracks
become large enough to expose the copper electrodes to the culture medium and eventually for the
PTFE discs to fall out of the electrode assemblies. This severely limits the reusability and reliability
of the bioreactor. Furthermore, the cracks forming in the embedding material allowed copper ions
to be leached into the culture medium, affecting the cell culture and compromising the experiments.
As such a new method of chemically isolating the electrodes from the culture medium had to be
found.
Capable of long term culture?
Low risk of infection?
Reliable?
Robust?
Reusable?
Easy to use/handle?
Scalable?
Homogenous stimulation?
Not biasing?
100
2.6 A NEW APPROACH TO ISOLATING THE
ELECTRODES
Polytetrafluoroethylene (PTFE) is chemically inert, electrically non-conducting, and
biocompatible; and has a melting point of 323 °C. Therefore, PTFE is the ideal material to isolate
the upper disc electrodes both chemically and electrically. It was envisaged that a PTFE cup and
washer (Figure 2.14) would be placed around the disc electrodes, isolating them from the culture
medium.
Figure 2.14 – Concept drawings of PTFE cup (A), the PTFE washer (B), the stainless steel electrode
(C) and the electrode assembly (D). 3D models were created in the commercial software SolidWorks
2008.
This however requires PTFE to be bonded either to the metal of the electrodes or to the other PTFE
component. Bonding PTFE in a biocompatible way that can withstand the 121 °C temperature and
hydrolysis effect of the autoclave, and is flexible enough to compensate for the different thermal
expansion coefficients is a very difficult engineering problem. After consulting with expert
companies (e.g. Loctite) no appropriate bonding agent was found. As no chemical bonding agent
was available, it was decided that the PTFE parts of the electrode assembly would be held together
by mechanical pressure. The mechanical pressure would be provided by the overlap between the
outer diameter of the PTFE washer and the inner diameter of the PTFE cup. How big this overlap
needs to be was determined using Finite Element Method simulations.
101
2.6.1 Finite Element Method Simulations
Finite Element Method simulations of the expansion of the PTFE washer and cup under various
temperatures were carried out in the commercial software COMSOL Multiphysics 4.0a.
2.6.1.1 Materials and Methods
Two separate models of the two components were created and were used to determine the change
of the inside diameter of the PTFE cup and the outside diameter of the PTFE washer with
temperature. In order to allow easy assembly, the PTFE washer was decided to be shrunk by
freezing at -80 °C before being fitted into the PTFE cup. Therefore the inner diameter of PTFE cup
was chosen to be at room temperature the same as the outer diameter of the PTFE washer at -80 °C.
2.6.1.1.1 Geometry
Figure 2.15 - The geometry used to model the PTFE washer (A) and cup (B)
102
2.6.1.1.2 Material Properties
Young's modulus: 0.5 [GPa]
Coefficient of thermal expansion: 135 x 10 -6
[1/K]
Thermal conductivity: 0.25 [W/(m*K)]
Heat capacity at constant pressure: 1.4 x 10 -3
[J/(kg*K)]
Density: 2200 [kg/m3]
Poisson’s ratio: 0.46
Table 2.4 – The material properties used to model PTFE [340]
2.6.1.1.3 The Mesh
A “Swept” mesh was applied, with pre-defined “Extra Fine” element size to the PTFE washer
model. “Free tetrahedral” mesh, also with the “Extra Fine” element size, was applied to the PTFE
cup (Figure 2.16).
Figure 2.16 – 3D models with meshes in COMSOL Multiphysics. A –PTFE washer, B – PTFE cup
2.6.1.1.4 Simulation Parameters
The Thermal Stress module of COMSOL Multiphysics was used to carry out isotropic, liner elastic
thermal simulations of both models. Temperatures ranging from -100 °C to 140 °C with 10 °C
steps were tested for. The outside radius of the PTFE washer and the inside radius of the PTFE cup
was measured using an averaging Edge Probe registering the furthermost “Y” coordinate.
103
2.6.1.2 Results and Discussion
Figure 2.17 – Results of the FEM simulation showing the change of the diameters as a function of
temperature. The blue patterned area indicates the overlap between the external diameter of the PTFE
washer and the internal diameter of the PTFE cup.
The diameter of the PTFE washer as a function of temperature was determined (Figure 2.17). At -
80 °C the external diameter of the PTFE washer is 28.72 mm. As such the inside diameter of the
PTFE cup was chosen to be 28.8 mm at room temperature. The simulation results also show that
once the two components reach the same temperature there will be an overlap of 220 μm on
average between the two diameters at all of the operating temperatures (room, incubator and
autoclave). The mechanical pressure arising from this overlap was postulated to provide
sufficiently strong forces to hold together the PTFE washer and the PTFE cup.
104
2.7 THE FOURTH GENERATION BIOREACTOR
As the computer simulations showed that mechanically bonding the PTFE cup and washer together
is likely to work, the fourth generation of the capacitive bioreactor was built. However, in trial
experiments problems arose with the electrode assembly. As such the design required further
improvement. In the following subsections the major iterative steps of the journey to the final
electrode configuration are explained.
2.7.1 Materials and Methods
2.7.1.1 The Bioreactor
28 mm diameter discs were water-jet cut out of 316L (A4) stainless steel sheets (RS stock code:
264-7241, RS Components Ltd.). These were to become biocompatible alternatives to the cytotoxic
copper electrode discs. M4 x 45 mm 316L (A4) stainless steel studs were Tungsten Inert Gas
welded into the centre of the discs. PTFE cups were milled out of sections of a 33 mm diameter
PTFE rod, with an internal diameter of 28.8 mm. The electrodes were placed into the PTFE cups
and held in position by placing ø29x1 mm PTFE washers, shrunk at -80 °C, into the PTFE cups. As
the PTFE washers warmed up and expanded, the resultant stress between the cup and the washer
held the two components together, effectively embedding the electrode in PTFE. The electrode
assemblies were placed into PTFE bioreactor lids and held in position using stainless steel nuts and
spring washers (Figure 2.19). The electrodes were electrically connected using water-jet cut
stainless steel “bridges” (Figure 2.18).
Figure 2.18 – Engineering drawing of
the “electrode bridge”
105
Concept and engineering drawings were created in the commercial software SolidWorks 2008.
Engineering drawings for the various parts can be found in the Appendix.
Figure 2.19 – Exploded view of the fourth generation bioreactor lid with a 6-well plate bottom.
1 – M4 stainless steel nut
2 – Stainless steel spring washer
3 – Electrode bridge
4 – PTFE bioreactor lid
5 – Upper electrode
6 – 6-well plate bottom
106
2.7.1.2 Biocompatibility Tests
2.7.1.2.1 Cell Culture
Commercial primary human Mesenchymal Stem Cells (PT-2501, Lonza UK Ltd.) were cultured in
T75 flasks in growth medium (Mesenchymal Stem Cell Growth Medium (MSCGM) Bulletkit, PT-
3001, Lonza UK Ltd.) at 37 °C and 5% CO2 content in humidified atmosphere.
Cells at passage 5 were seeded at the density of 50.000 cells per well in 2 ml growth medium into
6-well plates. Medium was changed at day 4, while samples were harvested on day 8. Three
samples were used per treatment (n=3) in the experiment comparing the biocompatibility of the
bioreactor with and without the PTFE cups. In successive experiments testing the biocompatibility
of the bioreactor lids with bare stainless steel electrodes 6 samples were used (n=6).
2.7.1.2.2 Cell Numbers
Cell numbers at day 8 were determined using the PicoGreen dsDNA Assay kit (Quant-iT™
PicoGreen® dsDNA Assay Kit, P1146, Life Technologies Ltd.). After being washed three times
with 1 ml warm PBS, 1 ml lysis buffer (1% Triton-X (T8787, Sigma-Aldrich Ltd.), 5% TE buffer,
94% ddH2O) was added to each well. After 5 min cells were scraped and the cell lysate was
transferred to 1.5 ml eppendorfs and freeze-thawed once. 100 µl of the lysate (DNA suspension)
was added to wells of a black 96-well plate in dark, together with 100 µl of PicoGreen stain (0.5%
PicoGreen stock, 5% TE buffer, 94.5% ddH2O)) and read in a fluorescent plate reader at 480 nm
excitation and 520 nm emission.
2.7.1.2.3 Metabolic Activity
Samples (n=6) were washed once with 1 ml warm PBS after removing the medium. 1 ml Alamar
Blue solution prepared in dark form 10 % Alamar Blue stain stock (Alamar Blue, DAL1100, Life
Technologies Ltd.) and 90% culture medium was added to each well in dark. Cells were incubated
for 60 min at 37 °C in 5% CO2 content. After incubation 100 µl Alamar Blue solution was
107
transferred from the samples into black 96-well plates in fours and read at 560 nm excitation and
590 nm emission.
2.7.1.2.4 pH Measurements
The pH of the culture medium from normal 6-well plate and bioreactor hMSC cultures (n=6) were
measured after 4 days using a HANNA pH 20 bench top pH meter (HANNA Instruments Inc.).
2.7.1.2.5 Statistical Analysis
Statistical analysis was performed using Two-tailed Student’s T-test and One-way ANOVA
combined with Tukey’s Multiple Comparison Test in the commercial software GraphPad Prism
v5.0. Statistical differences with p-values smaller than p=0.05 were considered significant. Data is
represented as mean ± standard deviation.
108
2.7.2 The Iterative Steps of Improving the Design of the Upper Electrode
Problem #1 – The walls of the PTFE cups deform during autoclaving
After assembly the PTFE cups were firmly held in place by the PTFE washers (Figure 2.20-A).
However after a cycle of autoclaving the cups came loose and were prone to fall off. This was
because of the previously unforeseen softening of the PTFE at the autoclaving temperature. As they
softened the PTFE cups bent under the mechanical pressure from the PTFE washers. When the
PTFE cups cooled down they set in the new bent shape. Therefore, after autoclaving the overlap in
radiuses that held the PTFE cups and washers together diminished and the two components were
prone to come apart. In each trial experiment some of the PTFE cups fell into the plate wells
compromising the cell culture.
The solution
The PTFE cups were autoclaved separate and were added on to the bioreactor lids in a biological
safety cabinet just prior to use. As the PTFE cups were autoclaved without the mechanical pressure
of the washers, the bending of the materials was to be avoided.
Problem #2 – Assembly by hand compromises sterility
The bioreactors had to be assembled after autoclaving by hand, touching the PTFE cups. This
introduced a serious risk of contamination. Furthermore, it was difficult to push the washers into
the cups, and the cups were still prone to fall off during use. Air bubbles trapped under the PTFE
cups were an additional problem that could potentially deprave areas of cells of nutrition,
eventually resulting in cell death.
109
Figure 2.20 – Engineering drawings (left), 3D models (middle) and photographs (right) of the various
iterations of the electrode assembly. A – The original PTFE concept, B – PTFE cups with raised walls,
C - PTFE cups with raised walls and four nebs, D – Bare stainless steel electrodes
110
The solution
The wall of the PTFE cups was raised to 6 mm from the previous 3 mm (Figure 2.20-B).
Previously, the wall of the PTFE cup was just high enough to accommodate the electrode and a
PTFE washer (thus not to restrict the range of heights the electrode can be positioned in). The
raised walls were to help limit the bending of the PTFE cups and were expected to make the PTFE
cups less prone to release the washers even with autoclaving.
Problem #3 – The risk of the PTFE cups falling off is still present
Assembly of the bioreactor was easier thanks to the higher PTFE cup. As expected PTFE cups
were also less likely to fall off the electrodes. However, the problem of the PTFE cups deforming
under the mechanical pressure, if autoclaved assembled with the PTFE washer, was still present.
Autoclaving separate and assembly post-sterilisation was again not a feasible alternative due to the
high risk of infection.
The solution
Four nebs were cut into the heightened wall of the PTFE cups and bent inward after placing the
electrode and the washer (Figure 2.20-C). This “geometrical” bond ensured that the components
stayed together even after several cycles of autoclaving. Pulling the components apart required
considerable force showing the strength of this bond.
Problem #4 – Pockets of air are trapped under the electrodes
A reliable method of keeping the PTFE cups on the electrode assembly has been found. However,
large pockets of air trapped between the electrode assemblies and the bottoms of the culture wells
still posed a major problem compromising cell cultures.
The solution
Removing the PTFE cups from the electrodes was postulated to solve the problem of trapped air
(Figure 2.20-D). However, it was possible that leaving the cell cultures exposed to the stainless
111
steel electrodes would have an adverse effect. In previous bioreactor generations embedding the
cytotoxic copper electrodes was essential. With the electrode material being changed in this
generation to biocompatible, high quality stainless steel this might not be the case anymore. In
order to investigate this, the biocompatibility of a bioreactor with and without PTFE cups was
compared to normal 6-well plates.
Figure 2.21 – Cell numbers (n=3) in a normal 6-well plate (Controls), a bioreactor with bare stainless
steel electrodes (Steel) and with the PTFE cups (PTFE) at Day 8 (* = p<0.05).
Results showed that although cell numbers were lower in bioreactor cultures there was no
significant difference between bioreactors with and without the PTFE cups (Figure 2.21). As such
bioreactors were decided to be used with bare steel electrodes (Figure 2.20-D), solving the problem
of trapped air pockets. The biocompatibility of this configuration was further characterised:
Figure 2.22 – Cell numbers (left) and metabolic activity (right) (n=6) in bioreactors with bare stainless
steel electrodes (steel) compared to normal 6-well plates (Controls) (* = p<0.05).
112
A decrease in both cell numbers and metabolic activity was found in bioreactors compared to
normal 6-well plates (Figure 2.22). This was postulated to be a result of ions leaching out of the
stainless steel and restricted gas exchange due to the electrode disc taking up a large portion of the
culture medium surface. In order to mitigate this, the electrodes were raised away from the culture
medium to be 10 mm from the lid of the bioreactor. This leaves an approx. 6mm air gap between
the culture medium and the electrode compared to the previous 1 mm (Figure 2.23).
Figure 2.23 – Raising the electrodes to be 6 mm rather than 1 mm from the culture medium surface
(assuming 2 ml of medium) helps avoid any contact between the electrodes and the culture medium.
With this extra distance the culture medium would only touch the electrode when the bioreactor
was moved or tilted. In order to determine whether this improves the biocompatibility of the
bioreactor lids, proliferation and metabolic activity was assayed with the new configuration after
eight days of culture.
Figure 2.24 – Cell numbers (left) and metabolic activity (right) (n=6) measured in bioreactors with
raised stainless steel electrodes (Steel) compared to normal 6-well plates (Control) (* = p<0.05).
113
Results showed that raising the electrodes improved biocompatibility from the perspective of
metabolic activity with no significant difference being present between bioreactor and normal 6-
well plate cultures (Figure 2.24). However, cell numbers were still slightly but significantly
lowered in the bioreactors to approx. 95% of
those in the well plate cultures. This was not
due to differences in culture medium pH, as
measurements showed no difference between
6-well plate and bioreactor samples in this
regard (Figure 2.25).
Figure 2.25 – A comparison of the pH of culture media (n=6) from normal 6-well plate and bioreactor
cultures.
As with the greater air gap, electrodes are unlikely to obstruct the gas exchange, perhaps limited
airflow at the bioreactor lid is behind these results. Nonetheless, the biocompatibility of the
bioreactor was deemed sufficient and no further improvements to the design of the bioreactor lids
were sought. Therefore, in in vitro cell culture experiments testing the effects of electrical
simulation this configuration - the fourth generation bioreactor lids with bare stainless steel
electrodes positioned 10 mm from the lids - were used (Figure 2.26).
Figure 2.26 – The final design of the capacitive electrical bioreactor. Image shows the bioreactor lid
upside down. 1 – PTFE lid, 2 – Stainless steel electrodes.
114
2.7.3 Improvements to the Auxiliary Components of the Bioreactor System
2.7.3.1 Signal Source
The range of electric field strengths that could practically be generated with 15 V maximum
amplitude output of the signal source (e.g. 0.2 V/mm – see Chapter 3) was relatively low. A
stronger electric field is more likely to produce an effect on the cells. As such it was desirable to
modify the bioreactor system to allow the delivery of higher electrode potentials and thus a
stronger electric field.
For this a specialist high voltage amplifier is necessary. One that can amplify the stimulating
regimes produced by the signal source to amplitudes in the range of hundred volts, has a very high
slew rate (i.e. can amplify short pulses in the microsecond range) and can be loaded with the high
electrical resistance of the bioreactors. Finding one such device was made difficult by the fact that
many amplifiers available commercially that possess a high voltage output and a high slew rate are
designed to drive piezo-optic devices and Micro-Electro-Mechanical Systems that are considered to
be high capacitance loads. Using an amplifier configured for high capacitance for high resistance
applications can result in a distorted signal and the overloading of the device.
Nonetheless, an amplifier was found that fulfilled all the criteria. The Trek Model 2205 high
voltage amplifier (Trek Inc.) can deliver an output in the range of ±500 V, possesses a slew rate
greater than 150 V/µs and has been calibrated for no load (infinite resistance) and, therefore, should
be able to drive the bioreactors without distorting the stimulating regimes.
2.7.3.2 Cables
The custom-made cables that were used with previous bioreactor generations have proven to be
unreliable. This was due to tendency of the soldering and mechanical force (e.g. crimp)
connections to break or release with repeated use. Furthermore, the copper wiring of the cables was
protected very little from the corrosive, humid environment of the incubators. More importantly the
115
custom-made cables were unable to safely accommodate the high voltage output of the Model 2205
amplifier.
Therefore a new cable system was assembled from commercial components allowing the reliable
and safe delivery of stimulatory signals with amplitudes in the hundred volt range.
2.7.3.3 Stimulation Stage
The “poles” of the previous stimulation stage were
difficult to use, making the exchange of bioreactors a
very cumbersome task. With the new cable system
there was no need for the poles, as such they were
discarded. Previously, lacquer coated copper sheets
were employed as bottom electrodes. These were
connected to the stimulation stage through soldered
wire connections and crocodile clips. These electrical
connections were prone to breaking. Therefore the copper sheets were replaced with a stainless
steel electrode plates connected to the signal source cables through a reliable connection based on
M4 bolts, multiple nuts and spring washers (Figure 2.27).
2.7.3.4 Signal Recording
It is important to verify that the intended stimulation regimes reach the bioreactors undistorted. To
this end a Data Acquisition Card (USB-6009, National Instruments Inc.) was added to the system
that allows the real-time monitoring of the injected signals using a laptop.
Figure 2.27- The new bottom electrode
plates and the cables inside the incubator
116
2.7.4 The Final Bioreactor System
Figure 2.28 – Photograph (left) and schematic (right) of the bioreactor system. Visible on the
photograph are the low voltage amplifier (1), the function generator (2), the high voltage amplifier (3),
fourth generation bioreactor lids (4) and the incubator (5).
The final bioreactor system (Figure 2.28) contains the following components:
1. Signal source: Aim and Thurby Thandar Instruments (TTi) – TG5011 Arbitrary function
generator
Specifications for pulse train signals:
Amplitude – Voltage range: 10 mVp-p to 20 Vp-p
Frequency range: 500 μHz to 12.5 MHz
Pulse width range: 20 ns to 2000 s
Can generate sine-wave, square-wave, pulse-train, noise and arbitrary signals
2. Amplifiers:
a. TTI WA301 Amplifier – Extends the range of the function generator to the
following:
Amplitude – Voltage range: 0 to 30Vp-p
Amplitude – Current range: 0 to ±300 mA peak AC
117
Rise/fall time: <0.5µs
Gain: Vernier adjustment between x1 and x10
b. TREK MODEL 2205 Power Amplifier – Extends the range of the function
generator to the following:
Amplitude - Voltage Range: 0 to ±500 V – Maximum 1000 Vp-p
Amplitude - Current Range: 0 to ±40 mA DC
0 to ±80 mA peak AC
Gain: 50 V/V
Slew Rate (10% to 90% ): Greater than 150 V/μs.
3. Cables
a. Co-axial cables:
RG-58 Co-axial cable with BNC plugs on both ends
(Voltage rating: 500 Veff-max) - Used with WA301 amplifier
Co-axial cable with a High-Voltage Output (SHV) and BNC connector on
the ends - Used with Model 2205 amplifier
b. Pomona - Model 3073 - BNC (Female) To Multi-Stacking Banana Plugs cable
(Hands free testing in controlled voltage environments: 500 WVDC Max.)
c. Hirschmann - MLS WS 50/2,5 - 4 mm safety system, 50 cm long cable, 4 mm
safety Multi-Stacking Banana Plugs on both ends
(Voltage rating: AC/DC 1000 V)
d. Multi Contact - High voltage safety crocodile clip on one end of cable “c”
(Working Voltage: 300 V)
4. Incubator
RS Biotech – Galaxy B CO2 Incubator, Model number: 150-400
118
5. Data acquisition card
National Instruments Inc. USB-6009 DAQ card
6. Stimulation stage
Stainless steel bottom electrode plates held in place and connected to the signal source
cables through stainless steel M4 machine screws, nuts and spring washers.
7. The bioreactor lids
Fourth generation bioreactor lids. Stainless steel disc electrodes positioned 10 mm from the
PTFE lid of the bioreactor using stainless steel nuts and spring washers. Electrodes
connected to signal source cables through water-jet cut stainless steel “bridges”.
8. Normal 6-well plates
119
2.7.5 Evaluation of the Fourth Bioreactor Design
The fourth generation bioreactor allows the long
term culture of monolayer and scaffold cultures
(cells have been maintained up to 16 days) and does
so with very little risk of infection thanks to the
autoclavibility of the bioreactor lids (Table 2.5).
The system is reliable and robust and can be
expected to deliver the same quality stimulation
again and again with repeated uses. Furthermore, it
is completely reusable as autoclaving and the
fatigue of repeated uses has not been witnessed to affect the performance of the system in any way.
The system is also very easy to use: The bioreactor lids just have to be autoclaved, added to the 6-
well plates as any other lid and connected to the simulation source using a crocodile clip. Very
simple and very quick. This also allows any number of bioreactors to be easily connected to the
signal source. Additionally, thanks to the large stainless steel electrodes, the cells in the well plates
receive an equivalent stimulus. Except to a slight decrease in cell numbers, possibly due to
restricted air exchange, the bioreactor system does not bias the behaviour of cells. The slight
reduction in cell numbers is negligible and does not interfere with experiments carried out using the
bioreactor system.
In conclusion, the fourth generation bioreactor system fulfils all the desirable criteria of an ES
bioreactor and is now ready to be used for in vitro experiments testing the effects of electrical
stimulation.
Table 2.5 – The criteria of an ideal
bioreactor – Fourth generation bioreactor
Capable of long term culture?
Low risk of infection?
Reliable?
Robust?
Reusable?
Easy to use/handle?
Scalable?
Homogenous stimulation?
Not biasing?
120
2.8 DISCUSSION
2.8.1 The Cause behind the Lower Cell Viability
The hMSCs samples cultured in the electrical bioreactor showed a 5% drop in cell numbers after
eight days in culture. The cause behind this is not known. There is no data available in this regard
in the scientific literature connected to the other electrical bioreactors discussed in Chapter 1 [25,
327-329, 341, 342].
One plausible explanation is that the materials constituting the bioreactor are not fully cyto-
compatible and have some sort of detrimental effect on cell viability.
However, the type of stainless steel used in the electrodes, ANSI 316L, is well known for its
biocompatibility and is widely used in orthopaedic implantology [343], for example in bone plates
and intravascular stents [344]. In many studies 316L stainless steel serves as the clinical reference
material to which the biocompatibility of other biomaterials is compared [345]. Furthermore, out of
various metal implant materials stainless steel’s biocompatibility was found to be second only to
gold [346]. Osteoblasts directly cultured on the surface of this material show constant proliferation
[343]. Therefore, it can be inferred that it is very unlikely that presence of the stainless steel
electrodes, not even being in prolonged contact with the culture medium, is the cause of the
measured drop of cell viability.
Similar conclusions can be drawn for the PTFE utilised in the bioreactor lid. PTFE too is
extensively used as a biomedical material, for example in vascular grafts [347], for its chemical
stability and biological inertness [347]. Furthermore, PTFE is known to support the adhesion of
MSCs and to be able to encourage their differentiation into various lineages after surface treatment
[348]. A plate lid manufactured out of a highly biocompatible material that is capable of supporting
MSC cultures is very improbable to have a detrimental effect on cell viability.
Can limited gas exchange be the cause instead? If the supply of oxygen is restricted perhaps that
may impair the ability of the hMSCs to self-renew. However, it was found that this stem cell type
121
responds well to hypoxic conditions. Baumgartner et al found no difference in hMCS proliferation
between cultures at 3% and 21% oxygen content through a 42 day culture period [349]. The
proliferation of MIAMI cells, a subtype of MSCs, was even found to be increased by low levels
(3%) of oxygen to as high as three fold [350]. Fehrer et al reported similar findings with 10 extra
population doublings and blocked osteogenic differentiation in hMSCs at 3% oxygen [351].
Hypoxia can restrict multipotent stromal cell proliferation, but this requires oxygen levels as low as
1%, as demonstrated by Holwarth et al [352].
It is highly unlikely that the bioreactor lid could restrict gas exchange to a level that would create
hypoxic conditions comparable to those used in these investigations. Even if it is able to do so, the
cells would most probably respond with enhanced proliferation and not a drop in viability.
Therefore, it can be concluded that a limited oxygen supply is not the cause behind the lower cell
numbers.
However, MSCs are known to be highly sensitive to changes in pH. For example, this cell type was
shown to reduce its proliferation to 40% if the culture medium pH is changed from 7.4 to a more
acidic 6.8 [353]. Similar results were found in a connected study at pH levels of 7.1, 6.8 and 6.5
[354].
In a cell culture incubator, the pH of the culture medium is maintained at the slightly acidic 7.4
with the help of 5% CO2 content and through the buffering effect of the culture medium itself.
However, if CO2 builds up within bioreactor as a by-product of cell metabolism, not being able to
escape fast enough into the incubator, the culture medium would become more acidic. This could
result in a lower proliferation rate providing a good explanation for what was witnessed in vitro.
However, measurements (Figure 2.16) showed no statistical difference between the pH of culture
medium from bioreactor and normal 6-well plate samples. On the other hand, it must be noted that
these samples were exposed to laboratory air during the measurements which could have
potentially removed any difference in pH.
In order to overcome this limitation, in situ measurements could be performed with the bioreactor
122
inside an incubator. Small probes that allow the in situ measurement of pH, such as a Dental
Beetrode Micro pH Electrode (World Precision Instruments Inc.), could be used in these
experiments to minimise any disturbance to the cell culture.
In summary, the 5% drop in cell numbers witnessed in the bioreactor compared to normal 6-well
plate cultures is unlikely to be caused by insufficient biocompatibility or low oxygen levels.
Changes in pH could provide a much better explanation, however no difference was witnessed
between bioreactor and 6-well plate samples. In situ pH measurements could yield further
information in this regard.
It also has to be noted that a 5% decrease in cell numbers is dwarfed by the potential benefit of
utilising the electrical bioreactor, as electrical stimulation has been observed to increase the
proliferation of osteoblasts by 25% [159] and 31% [25]. A similar effect was reported with hMSCs,
enhancing proliferation up to 60% [20].
2.8.2 Comparison of the Bioreactors
The fourth generation capacitive bioreactor developed in this thesis offers many benefits compared
to its counterparts discussed in the literature. Agarose bridges [337] are highly useful in
galvanotactic studies, but are difficult to implement for any other application due to their very
limited culture time capability, one-use nature and susceptibility to infection.
The Donnelly bioreactor [338], and the cardiac muscle bioreactor built by Radisic’s group [339]
both fail in reliability and robustness. Even more importantly, these devices lack the autoclavibility
of the capacitive bioreactor, greatly limiting their reusability.
The commercial bioreactor C-Pace [341, 342] has many major shortcomings as well. Arguably the
most substantial of these is that electrical stimulation can only be delivered for a maximum of two
days when using this system, as after that time the electrodes have to be removed, cleaned and
equilibrated overnight. This poses serious restrictions on the type and duration of experiments that
123
can be carried out using the C-Pace device. No such restrictions are present when using the
capacitive bioreactor. Furthermore, the lids of the C-Pace system cannot be autoclaved rendering
them practically one use. In comparison, the lids of the capacitive bioreactor are fully autoclavable
allowing them to be reused unlimited times.
The device most similar in capabilities to the capacitive bioreactor of this thesis was presented by
Kim et al [25]. However, Kim’s system utilises direct stimulation, potentially biasing the outcome
of experiments through unwanted chemical changes. The ease with which this system can be used
is also questionable.
The bioreactor designed in this study, on the other hand, meets all the criteria of an ideal ES
bioreactor. It is highly biocompatible, autoclavable, reusable, robust and reliable. It allows the long
term delivery of capacitive electrical stimulation to any tissue type and can be adapted to suit the
needs of a large variety of regenerative medicine applications. It is a truly versatile system, one that
is easy and inexpensive to implement, and could one day see widespread use in the discipline of
tissue engineering,
2.8.3 A Fifth Generation Bioreactor
There is of course always place for improvement: The concept of a fifth generation bioreactor has
been developed (Figure 2.29). This design joins up the bioreactor lid and the stimulation stage into
one compact, stand-alone bioreactor. This does not only eliminate the need for large, cumbersome
stimulation stage, but allows the must more efficient synchronous use of multiple bioreactors.
Furthermore, as an added benefit, this design eliminates the possibility of exposing the culture
environment to unsterile conditions by accidentally lifting up the bioreactor lid.
124
Figure 2.29 – 3D view of the assembled fifth generation bioreactor in SolidWorks 2008.
The fifth generation, additional to the components of the fourth generation bioreactor lid, would
include a bottom electrode plate together with the necessary PTFE components to keep it in
position and isolated from the conductive parts of the incubator. Four bolts at the four corners of
the bioreactor would hold the lid and the bottom plate together securely, but allow the removal of
the lid (for example in the case of a culture medium change) easily by hand without the need of
specialist equipment (Figure 2.30).
One such design also has commercial potential and could be developed into a product. A sturdy,
reusable, but inexpensive electrical stimulation bioreactor could become a very useful tool for
cardiac, bone, neural and skeletal muscle tissue engineering. Such a bioreactor also has high
potential for disciplines where the effect of electricity is widely studied, for example, embryonic
development, cardiac and regeneration research. Electro-permeabilization, gene transfection and
even food preservation investigations could also benefit from this technology.
125
Figure 2.30 – Exploded view of the fifth generation capacitive bioreactor concept
2.8.4 Perfusion Concepts
The bioreactors could be further modified to allow the delivery of a second type of stimulation, for
example perfusion [355, 356]. This would also have the added benefit of continuously exchanging
the culture medium, thus improving the biocompatibility of the system. Two combined electrical-
perfusion concepts were created: One for monolayer cultures, a laminar flow-electrical bioreactor,
and one for 3D scaffold based cultures, a 3D flow-electrical bioreactor. Further combining these
bioreactors with a flexible membrane bioreactor (e.g. the Flexercell Tension Plus system (Flexcell
126
International corp.)) or some other system could introduce mechanical stimulation as a third type of
physical stimulus.
2.8.4.1 Laminar Flow – Electrical Bioreactor
Figure 2.31 – A 3D rendered model of the laminar flow – electrical bioreactor concept
This bioreactor would allow the culture of cells in a monolayer while being stimulated with both an
electric field and perfusion (Figure 2.31).
The design consists of five components (Figure 2.32):
1. 6-well plates modified to direct the flow to the cells in an optimal way (Figure 2.33)
2. Bioreactor lids with a place for the electrodes and the perfusion inlets and outlets – adding
the lid onto the modified 6-wellplates creates the flow chamber
3. Upper electrodes
4. Bottom electrodes
5. Perfusion tubing to deliver and remove the medium
127
Figure 2.32 – Exploded view of the laminar flow – electrical bioreactor. (1 - Modified 6-well plate, 2 –
Bioreactor lid, 3 – Upper electrodes, 4 – Bottom electrode, 5 – Perfusion tubing)
Figure 2.33 – Top view of the bioreactor lid (left) and of the modified 6-well plate (right). A – Perfusion
inlets and outlets, B – Place for the electrodes, C – Flow and culture chamber, D – Raised areas to
optimally direct the flow towards the cells.
128
2.8.4.2 3D Flow – Electrical Bioreactor
Figure 2.34 – Rendered model of the 3D flow – electrical stimulation bioreactor
This bioreactor would allow the electrical and perfusion stimulation of 3D tissue constructs (Figure
2.34). Positioning the electrodes at different heights will set the size of the culture chamber. Each
well/culture chamber can be perfused horizontally and stimulated with electricity vertically. This
design consists of the following components (Figure 2.36):
1. Modified 6-well plate with outlets and inlets for perfusion
2. Electrode assemblies with seals
3. PTFE bioreactor lid
4. Metal bridge connecting all the electrodes
5. Bottom electrode plate
6. Perfusion tubing
129
Figure 2.35 – Top view of the modified 6- well plate. The arrows indicate the perfusion inlets and
outlets.
Figure 2.36 – Exploded view of the 3D flow –electrical bioreactor. 1 – Modified 6-well plate, 2 – Upper
electrodes with seals, 3 – Bioreactor lid, 4 – Metal bridge connecting all the electrodes, 5 – Bottom
electrode, 6 – Perfusion tubing
130
2.9 CONCLUSIONS
In conclusion, the aim of this chapter, that to design a capacitive electrical bioreactor based on the
6-well plate configuration, able to deliver a stimulus to both monolayer and scaffold cultures, has
been achieved. Through four successive generations a device was created that fulfils all the
desirable criteria of an electrical stimulation bioreactor. Its autoclavable nature allows cell cultures
to be maintained long term in the bioreactor with very little risk of infection. It is reusable, reliable
and robust, therefore it can be depended upon to deliver the same quality performance when testing
the effects of electrical stimulation. Furthermore, thanks to the bioreactor’s ease-of-use and
scalability, experiments can be carried out without any added difficulty.
The effect of the bioreactor alone on the behaviour of cells was also tested. Results show that cell
numbers are somewhat reduced (by 5%) in cultures maintained in the bioreactor compared to
normal 6-well plates. Although this difference is statistically significant, it will have no impact on
experiments carried out using the bioreactor.
With the design finalised, it is important to characterise the bioreactors electrical behaviour. The
next chapter concerns with the work undertaken in order to achieve this goal.
131
Chapter III
Computer Simulations
“Think? Why think! We have computers to do that for us.”
Jean Rostand
132
3.1 INTRODUCTION
“Objective #2 - Carry out computerised simulations to determine the exact stimulus received by the
cells.”
To be able to ascertain the exact electrical stimulus a cell receives the electrical behaviour of the
bioreactor has to be characterised from two perspectives. Firstly, the relationship between the
electrode potential and the electric field strength that the cells are exposed to must be understood.
As it can be seen in Figure 3.1, the same electrode potential can generate very different electric
field strengths depending on the distance between the electrodes and the materials located in
between them. This is also necessary to allow the optimisation and comparison of various
bioreactor designs and to enable the comparison of the experiments and results of this study to
those discussed in the literature.
Figure 3.1 – The same electrode potential difference (15 V in the above example) will result in a
different electric field strength depending on the distance and the material between the electrodes. The
colour legend indicates electric field strength from 0 (blue) to approx. 35 V/m (red).
The second perspective from which the electrical behaviour of the bioreactor has to be examined is
that of electrical signals. It must be ascertained whether the electrical signals generated by the
function generator reach the cells undistorted. Some properties of the system (e.g. resistance,
capacitance, inductance) can cause the electrical signals to be dampened and their wave shape
133
distorted. Therefore, it needs to be investigated whether this is the case and, if yes, which signals
are affected.
In the following chapter the electric field in all four generations of the electrical bioreactor will be
modelled and simulated in order to allow the comparison of the various designs. This is followed
by a more extensive examination of the relationship between electrode potential, electrode distance,
medium volume and electric field strength in the fourth generation bioreactor. Simulations of the
electric field in case of a 3D scaffold are also presented.
The second half of this chapter is concerned with the electrical behaviour of the bioreactor as a
function of signal frequency (Bode-plot), modelled in the commercial software MULTISIM, and
the signal validation measurements made using a digital NI Data Acquisition Card and an analogue
oscilloscope.
This chapter will finish with a discussion of future possibilities and a summary of the findings.
134
3.1.1 The Finite Element Method
Whether using the direct, the capacitive or the inductive methods for stimulating cells or tissue, the
electromagnetic field generated by the bioreactor has to be evaluated. This can be a complex
problem involving the solution of many complicated ordinary and partial differential equations
[357]. Such a problem can be best solved using numerical analysis methods. The most widely used
of these is the Finite Element Method [358].
The Finite Element Method (FEM) solves complex differential equation based problems by
dividing up the “object” of the problem into a finite number of nodes connected by elements of
finite size through a process known as discretisation [358, 359]. The web of nodes and elements is
called a “mesh” [358] (Figure 3.2). FEM then applies integral functions to the elements of the mesh
to yield a set of linear equations [357, 360]. Solution of the linear equation system provides an
approximated solution to the original differential equation [357, 360].
The family of numerical analysis methods also include the Finite Difference Method for the
solution of ordinary differential equations with boundary-value problems [357], the Rayleigh-Ritz,
the Finite Volume, the Spectral, the Singularities, the Mesh-free and the Boundary Element
Methods [357, 358, 360, 361]. Each of these has its own advantages and disadvantages and are best
suited for different types of problems [358, 360, 361].
Figure 3.2 – A sphere discretised into a mesh of a finite number of nodes and elements
135
3.1.2 An Introduction to the Computerised Electric Field Simulations – Why Was
COMSOL Multiphysics Chosen?
3.1.2.1 Computer Simulations in the MATLAB Environment
The first attempts in this study towards modelling the electric field inside the bioreactor were made
in the commercial software MATLAB. MATLAB (The Mathworks Inc.) is a well-known, high-
level language, interactive software environment that researchers and engineers all around the
world use to carry out complicated numerical calculations, data analysis and visualisation.
The MATLAB models were based on a technique known as the Method of Moments, described in
detail in Bai and Lonngren, 2004 [362]. In brief, the Method of Moments breaks up the charge
density on the electrode into a set number of finite charges, thus allowing the electric field strength
at any point is space then to be calculated as the sum of the electric fields generated by each of
these finite charges (Figure 3.3). This enables the capacitance and potential gradient to be
calculated with relative ease even between electrodes of very complicated shape.
Figure 3.3 – The charge density on two round electrodes is broken up into finite charges in MATLAB.
Dimensions are in meters, while the scale indicates charge in coulombs.
136
The models were able to calculate electric field strength at a set number of points (the density of
these points can be specified by the user) from the electrode potential. Apart from the electrode
potential, and the point density the models require the size and distance of the electrodes to be
specified. The models can accommodate both rectangular and round electrodes.
Although the models are able to calculate the electric field strength, in their current state they
cannot solve problems where more than one material is present between the electrodes.
Furthermore computational resources are used highly inefficiently: At a point density of 8000 (a
20x20x20 matrix) the software can require several ten minutes to carry out the calculations on a
high-end computer. If a point density bigger then this is applied, the memory requirements of the
program can exceed even the capacity of high-end desktop computers.
As commercial physical simulation software that can calculate electric field strength with much
greater efficiency are available, this solution was not pursued any further.
137
3.1.2.2 COMSOL Multiphysics
COMSOL Multiphysics (version 4.0a) (COMSOL Inc.) is a versatile software platform that allows
the simulation of a wide range of physics-based problems. The software uses finite element based
numerical methods to simulate, amongst others, electrostatics, acoustics, heat transfer and static
structural problems. A great advantage of this platform is not only that it accommodates a greater
range of physical domains than the alternative software, but that it also allows the coupling of these
different physical fields. This grants the user the ability to observe the interaction of a wide range
of physical processes.
Compared to the MATLAB based solution, COMSOL Multiphysics contains all the necessary tools
to allow the easily modelling of any geometry, any bioreactor configuration, and can accommodate
a large number of different materials (e.g. PTFE, polystyrene, water) and input parameters (e.g.
electrode potential, charge density). Furthermore, it is much more efficient at using the resources of
a computer and can finish a simulation in a matter of minutes even on an average computer. The
electric field modelling and simulation work discussed in this chapter therefore was performed
using this software.
138
3.1.3 The Physics Background of the Electric Field Simulations
Any computer simulation environment is just a tool in the hand of the scientist or engineer. Its
purpose is to make the solving of physical problems a faster and easier task. It can also help in
dealing with problems too complicated for traditional methods or in better representing data.
However, without an understanding of the underlying physical principles it is very easy to ask the
wrong questions or misinterpret the results. Therefore, prior to the computer simulations, in the
following sections the fundamental equations that describe electric fields will be discussed,
together with how the results of COMSOL simulations should be interpreted.
3.1.3.1 The Maxwell Equations
The following partial differentiation equations are known as the Maxwell equations (derived for
static electric fields) and form the basis of classical electromagnetic theory [363]:
(1)
(2)
(3)
(4)
Where “E” is the electric field strength [V/m],
“B” is the magnetic flux density [Wb/m2]
“H” is the magnetic field [A/m]
“J” is the electrical current density [A/m2]
“D” is the electric flux density [C/m2]
“ρv” is the electrical charge density [C/m3]
”is the curl, a vector operator
” is the divergence, a vector operator
139
3.1.3.2 Electrostatics
Although the stimuli delivered by the bioreactor system are dynamic (i.e. they vary with time), in
order to ascertain the electric field strength (the crucial parameter for the stimulation) it is sufficient
to analyse the bioreactors from the perspective of electrostatics. Electrostatics can be defined as the
specialisation of the Maxwell equations to steady state charges, to the behaviour of electric fields
where there is no time variation [364].
In electrostatics the Maxwell equations can be simplified as [364]:
A,
(5)
Equation 5 displays the electric field strength as a gradient of a potential function and can be
interpreted as a measure of the force an electrostatic field exerts on a charge [364].
B,
(6)
Equation 6 describes electrical charge (charge density, ρv) as the source of electrical flux (D). The
conservation of electrical charge can be derived from this equation [364] and forms the boundary
condition of the electrical simulations performed in COMSOL.
With the constitutive relation between the two being:
C,
(7)
This equation states the connection between flux density and electric field strength, where “ε” is
the absolute permittivity [364, 365]. It must be noted that in COMSOL Multiphysics this equation
forms the basis of electrostatics simulations.
140
3.1.3.3 The Absolute Permittivity
Absolute permittivity can be calculated as [366]:
(8)
Where “ε0” is the permittivity of free space, a constant with a value of 8.85∙10-12
[F/m]
“εr” is the relative permittivity (or dielectric constant), the material property that has to be
specified for the electrostatics simulations in COMSOL.
141
3.1.3.4 The Electric Field inside the Culture Medium
The culture medium is not only electrically conductive (i.e. contains freely moving charges), but is
also a dielectric material. Dielectric materials contain bound charges, that cannot move freely and
thus aid in conduction, but can shift if placed under and electric field [363, 364, 366]. The positive
and negative charges shift in the opposite direction creating polarisation within the material.
Figure 3.4 – The effect of polarisation on the electric field strength inside the culture medium. A – If
there is no electric field present the water molecules orient themselves randomly. B – Once an electric
field is applied the water molecules will start to orient themselves – the oxygen facing the positive,
while the hydrogen atoms facing the negative electrode. C – This electric field induced orientation is
called the polarisation of the material. Its overall effect can be viewed as the creation of a positive
charge density on the negative electrode facing side of the material and the creation of a negative
charge density on the positive electrode facing side of the material . D – The two charge denisties
generate an antagonist electric field that acts to weaken the electric field created by the electrodes. A
cell placed into the culture medium will experience the sum of these two fields, which will always be
weaker then the one generated by the electrodes alone. (Ecell – electric field strength exprienced by a
cell, E0 – electric field strength generated by the electrodes, Ep – electric field strength generated by the
polarisation.)
142
The dielectric nature of the culture medium comes from its main constituent: water. Water
molecules act as molecular dipoles that orient themselves when exposed to an electric field. The
hydrogen atoms will face the negative electrodes, while oxygen atoms will turn towards the
positive electrodes (Figure 3.4).
The polarisation generates a secondary electric field within the material with a magnitude of [363,
364, 366]:
(9)
Where “Ep” is the electric field strength generated by the polarisation [V/m],
“P” is the polarisation [C/m2]
“ε0” is the permittivity of free space
The total electric field strength within the culture medium therefore will be given by the sum of the
electric fields generated by the potential difference between the electrodes (E0) and the polarisation
(Ep):
(10)
As E0 and Ep are of the opposite direction, thus the magnitude of E will be:
(11)
Polarisation can be calculated as:
(12)
Thus from the combination of the above two equations we can see that:
(13)
(14)
(15)
143
Due to the dielectric nature of the culture medium the electric field strength affecting the cells will
be “εr” – fold weaker than what is generated by the electrodes [363, 364, 366]. The relative
permittivity of culture medium is 73, producing a 73-fold reduction of the electric field strength.
3.1.3.5 The Interpretation of the Results of COMSOL Simulations – The Importance of the
Boundary Conditions
Depending on the choice of boundary condition, whether the electrical potential or the charge
density on the electrodes is specified, COMSOL will give a different answer.
For example: A capacitor with an area of 1 m2 and an electrode distance of 1 m is modelled. The
space between the electrodes is filled with a dielectric with a relative permittivity of 100.
The capacitance can easily be calculated as [365]:
(16)
Where “ε0” is the permittivity of space and is equal to 8.854 ∙ 10-12
[F/m],
“εr” is the relative permittivity of the material within the capacitor
“A” is the active area of the electrodes of the capacitor
“d” is the distance between the electrodes
If a potential difference of 100 V is applied to the capacitor the charge density on the electrodes
will be [365]:
(17)
Where “Uel” is the potential difference between the electrodes
In the simulation, if the potential difference between the electrodes is specified (100 V), the
software gives 100 V/m as the electric field strength. Of course this makes sense as, if the potential
difference is 100 V and the electrodes are 1 m apart the potential must change 100 V in 1 m, hence
the field strength is 100 V/m.
144
However, if the charge density on the electrodes is specified, the result will be 1 V/m for the
electric field strength. Which value is correct? No parameter of the capacitor (C, εr, etc.) has
changed and the values are still quite different.
The answer is both are correct. If the electrode potential difference is specified, the results show the
“external” field “E0”, as the potential must change accordingly within the given distance. If the
charge density is specified, the results will indicate the “internal” field “E” (E = E0/εr).
When stimulation is delivered to the bioreactor the parameter that can be set and defines the
strength of the stimulation is the potential difference between the electrodes (Uel). Therefore it is
logical to keep this parameter as the input for the simulations, while it has to be kept in mind that
the electric field strength the cells will be exposed to will be 73-fold weaker (εr) than the value
given by the COMSOL simulation.
(18)
The computer simulations and their results that are presented in the following sections were created
in light of this information.
145
3.2 MATERIALS AND METHODS
3.2.1 Computer Simulations of the Electric Field inside the Bioreactor
The Electrostatics module of COMSOL Multiphysics was used to carry out stationary electrostatics
simulations based on the relative permittivity of the constituent materials. The relative permittivity
values that were used can be seen in Table 3.1. “Free tetrahedral” meshes with the “Extra Fine”
element size were applied to all models.
Material Relative permittivity
Air 1
Copper 999999 (infinity)
Culture medium 73
Epoxy 3.7
Gelatine 3.5
Polystyrene 2.5
PTFE 2.1
Steel 999999 (infinity)
Table 3.1 - The relative permittivity of the various materials used in the COMSOL models [340]
3.2.1.1 The Comparison of the Different Bioreactor Designs
In order to allow the comparison of the different generations, computer simulations of the electric
field inside one well of each bioreactor were carried out. Models were built according to the
geometries in Figure 3.5. The models for the first two generations have their electrodes in their
fixed position, while the electrodes in models of the latter two generations are positioned to be 1
mm away from the culture medium surface. All four models were created assuming 2 ml culture
medium.
146
Figure 3.5 – The geometries used for the models of the first (A), second (B), third (C) and fourth (D)
generation bioreactors
1 – Stainless steel machine screw
2 – Epoxy
3 – Copper electrode
4 – PTFE disk
5 – Culture medium (2 ml)
6 – Polystyrene well (6-well plate)
7 – Copper counter electrode
1 – Stainless steel electrode
2 – Polystyrene well (6-well plate)
3 – Culture medium (2 ml)
4 – Stainless steel counter electrode
1 – Copper electrode
2 – Polymer tube section
3 – Polystyrene well (6-well plate)
4 – Culture medium (2 ml)
5 – Copper counter electrode
1 – Copper electrode
2 – Polystyrene lid
3 – Polystyrene well (6-well plate)
4 – Culture medium (2 ml)
147
3.2.1.2 The Electric Field Strength in the Final Bioreactor Design (Monolayer Cultures)
The electric field strength inside the bioreactor is governed by three distinct parameters: The
electrode potential difference, the distance between the electrodes and the thickness of the culture
medium layer. It is important to find the equation that describes the relationship between these four
variables, as otherwise a new computer simulation has to be created every time the bioreactor is
used with a different configuration.
For practical reasons the inputs for this equation were chosen to be the electrode potential
difference in V (Uel), the distance of the upper electrode from the bioreactor lid in mm (H) and the
volume of culture medium per well in ml (M). The output of the equation will be the electric field
strength the cells experience inside the culture medium in V/m (Ecell).
3.2.1.2.1 Simulation Parameters
The same geometry was used as for the previous simulations of the fourth generation bioreactor
(Figure 3.5-D) with the “H” and “Dm” parameters (Figure 3.6) being varied.
Figure 3.6 – The parameters of the simulations
The potential difference between the electrodes was set to be 1 V, while the “H” parameter was
varied between 1 and 15 mm and the “Dm” parameter was changed between 0 and “Dm-max” (Dm-max
= 18-H mm).
148
3.2.1.3 The Electric Field in the Case of a 3D Scaffold
The electrical stimulation bioreactor can accommodate two and three dimensional scaffolds. In the
cell culture studies (discussed in Chapter 4) experiments will be carried out using a 1cm3
Spongostan scaffold. Therefore, it is important to explore how using a 3D substrate may alter the
strength and behaviour of the electric field. For this the relative permittivity of the Spongostan
scaffold must be ascertained first.
3.2.1.3.1 The Relative Permittivity of the Scaffold
The Spongostan scaffold is a porous material consisting of porcine gelatine (with a relative
permittivity between 3 and 4) and culture medium filling up the pores (with a relative permittivity
of 73) [367]. The equivalent relative permittivity of the two materials combined can be determined
once the ratio of the two volumes is known [367]. The volume ration of the two constituents of the
scaffolds was determined using two methods: Through Micro Computed Tomography (MicroCT)
scanning and through microbalance weight measurements.
3.2.1.3.1.1 Micro Computed Tomography of the Spongostan scaffold
MicroCT scans of a Spongostan scaffold were performed using the XRadia Versa XRM-500
system (Figure 3.7) with a Tungsten target at 4x and 20 x magnification. Pixel size was 1.8368 μm
and 0.3585 μm, while exposure was 4 s and 10 s respectively. Source voltage was 50 kV and
source power was 4 W. 1601 projections were taken
with a binning of 2 in both scans. Volumes were
reconstructed using XRadia XMReconstructor and
analysed using Aviso 8.0. The volume ratio, pore size
distribution and scaffold wall thickness were
determined.
Figure 3.7 – The XRadia Versa XRM-500 system
149
3.2.1.3.1.2 Microbalance Weight Measurements
Results of the tomography scans were verified by measuring the dry and fully wetted weight of
three 1cm3 Spongostan scaffolds. From the weight, knowing the density of the materials (0.99823
g/cm3 for water [340] and 1.3 g/cm
3 for gelatine [368]) the volume and the volume ratios can then
be calculated.
3.2.1.3.2 The Electric Field Strength inside the Scaffold
With the equivalent relative permittivity of the Spongostan scaffold determined, simulations were
carried out in COMSOL of the electric field strength inside the scaffold. The simulations were
based on the previous geometry in Figure 3.6-D. Electrode potential differences from 0 to 150 V
were tested for. The resultant electric field strength values were registered using an averaging
domain probe specified to the Spongostan scaffold (Figure 3.8).
Figure 3.8 – The geometry used to simulate the electric field with a Spongostan scaffold in the
bioreactor
3.2.1.3.3 The Electric Field at the Cellular Level
Can stimulating in 3D make a difference in some way, additional to an altered electric field
strength? Perhaps the shape of the pores or the scaffold material itself can have an effect on the
electric field at the cellular level. In order to determine this, representative cross-sections of the
150
reconstructed volume from the 4x magnification scan were segmented into a binary image. The
images were transferred to ImageJ, where, using an edge detection algorithm, the boundaries
between the material and the pores were detected. The images containing the edges were converted
into vector format using Print2CAD and imported into COMSOL Multiphysics in order to serve as
the basis of 2D simulations of the electric field at the cellular level. The electrical potentials on the
upper and lower boundary of the modelled approx.1800x1800 μm region were set to correspond to
the electric field strength generated by a 150 V electrode potential difference. Gelatine and culture
medium were modelled with a relative permittivity of 3.5 and 73 respectively.
151
3.2.2 The Frequency Response of the Bioreactor
Frequency response analysis is a useful tool for characterising a system from the perspective of
electrical signals, describing the relationship between the magnitude and phase of the output and
the input as a function of frequency. With the help of the frequency response it can be determined
whether a signal (i.e. an electrical stimulation regime) will be altered (e.g. amplified, dampened) as
a result of passing through the bioreactor system. For this the electrical resistance and capacitance
of the bioreactor must be known first. These were calculated to be 5.42∙1013
Ω and 0.076 pF
respectively. Knowing the resistance and the capacitance of the bioreactor its impedance can also
be calculated (For the details of the calculations please see section A.2 of the Appendix):
The impedance of the bioreactors therefore will be:
(19)
Where “s” is the complex frequency which is a complex number
Based on these values the frequency response
of the bioreactor between 1 Hz and 1 GHz
was modelled using circuit simulations in the
commercial software National Instruments
MULTISIM 13.0 (Figure 3.9).
Figure 3.9 – The model of the bioreactor system
in MULTISIM taking into account the
electrical resistance and capacitance of the
bioreactor itself and the coaxial cable.
152
3.2.3 Signal Measurements
It is important to know whether the intended stimulatory regimes reach the cells undistorted.
Although the frequency response analysis yields information in this regard, it does not take into
account the amplifiers. As such in situ measurements were made of various signals on the
bioreactor.
Figure 3.10 – The oscilloscope used for the measurements
Signal measurements were carried out using a NI USB –
6009 Data Acquisition card (National Instruments Inc.),
LabView 2012 Signal Express for DAQ software
(version 6.0.0., National Instruments Inc.), a Hitachi V-
555 oscilloscope (Figure 3.10), a fourth generation bioreactor filled with 2 ml culture medium per
well (no cells) and a Samsung laptop. 1, 10, 100, 250 and 1999 μs pulse train and sine waveforms
were tested at 1, 100, 500 and 9000 Hz frequencies. The signals were delivered at low voltage (5,
10 and 15 V) using the TTi WA301 amplifier and at high voltage (150V) using the Model 2205
amplifier.
153
3.3 RESULTS AND DISCUSSION
3.3.1 The Electric Field in the Four Different Bioreactor Generations
Figure 3.11 – The electric field strength in the four different generations of the bioreactor as a function
of electrode potential difference
The equations describing the relationship between the electrode potential difference (Uel) in V and
the electric field strength (E) inside the culture medium in V/m in the four generations of the
bioreactor are:
(19)
(20)
(21)
(22)
The results show that moving the electrodes outside the culture environment in the second
generation of the bioreactor caused a more than three times drop in electric field strength compared
to the first generation (Figure 3.11). These findings demonstrate the inefficiency of this solution
and emphasise the need to place the electrodes as close to each other as possible.
154
Figure 3.12 - The electric field strength in the first (A), second (B), third (C) and fourth (D) generation
bioreactor. The colour legend indicates the electric field strength from low (blue) to relativel high
(red).
The strongest electric field strength (at the same electrode potential) was generated with the third
generation design, showing a twelve times greater efficiency than the second generation. The
efficiency of the fourth generation bioreactor is lower, as it generates only approx. 60% the output
of the third generation (equations 21 and 22). It is interesting to note that this difference is simply
caused by the fact that the PTFE disk in the third generation was replaced with an air gap in the
fourth generation. This highlights that even the smallest alteration in the design can have a
significant effect on the electric field strength. All four bioreactor generations generated a mostly
homogenous electric field inside the culture medium, however some differences due to edge effects
have been observed (Figure 3.12).
155
3.3.2 The Equation Describing the Electric Field Strength inside the Fourth
Generation Bioreactor
Figure 3.13 – The electric field strength experienced by the cells (Ecells) at 1 V electrode potential
difference as a function of the distance of the electrodes (H) and percentage of this distance that was
filled up by culture medium.
Based on the results of the computer simulation (Figure 3.13) the electric field was found to be
described by the following equations:
If M = Mmax
(23)
If M < Mmax
(24)
)
(25)
156
Where “Ecells“ is the electric field strength the cells experience [V/m],
“H” is the distance of the upper electrode from the bioreactor lid [mm],
“M” is the amount of culture medium used per well [ml],
“Uel“ is the electrode potential difference [V].
It has to be noted that the equations contain an inaccuracy: The limit of equation 24 as “M”
approaches “Mmax” is infinity. Therefore, equation 24 will produce values much higher than
“reality”, if a culture medium volume very close to the maximum (approx. in a radius of 0.1 ml
from “Mmax”) is used. Nonetheless, based on the equations a stand-alone application (Figure 3.14)
was created in the National Instruments Labview software. The application allows the electric field
strength to be easily calculated with a click of a button, while it also warns the user if the input
variables are outside their permitted range.
Figure 3.14 – The graphical user interface of the electric field strength calculator
From these equations it can be inferred that the strength of the electrical stimulation delivered in
the bioreactor can be increased in three ways:
A, by decreasing the electrode distance.
B, by increasing the electrode potential difference.
And C, by filling up the space between the electrodes with culture medium.
157
Maximum electrode potential difference is limited by the capabilities of the signal source, while the
other two options are restricted by the geometry of the bioreactor and carry a risk of lowering the
biocompatibility of the system.
158
3.3.3 The Electric Field in the Case of a 3D Scaffold
MicroCT scans show that only 0.8% of the Spongostan scaffold volume consists of gelatine (Figure
3.15-A). The structure is built up from large pores with an average diameter of 465±242 μm
(Figure 3.16), separated by gelatine walls of approx. 4-10 μm thickness (Figure 3.15-B).
Figure 3.15 – The reconstructed volume from the 4x magnification scan (left) showing the gelatine in
blue and the empty pore space in red (scale bar corresponds to 500 μm). The image on the right shows
a cross section of the volume from the 20x magnification scan displaying the structure of the gelatine
walls in dark grey (scale bar corresponds to 50 μm).
Figure 3.16– A histogram of the different pore sizes in the scaffold
50 μm
159
Gelatine Water Volume ratio
Weight
[mg]
Volume
[cm3]
Weight
[mg]
Volume
[cm3] Gelatine Water
Sample 1 13.37 0.010282 565.93 0.566937 1.8% 98.2%
Sample 2 14.27 0.010974 721.10 0.722379 1.5% 98.5%
Sample 3 14.80 0.011385 671.00 0.67219 1.7% 98.3%
Table 3.2 – The results of the microbalance weight measurements
MicroCT showed that the scaffolds contain approx. 0.8% gelatine and 99.2% water, however the
results of the microbalance measurements (Table 3.2) indicate that these values are different: 1.6%
and 98.4%. The difference between the results of the two techniques can be explained by the
inaccuracy of the methods and by the smaller sample size (field of view) of the MicroCT scans. For
the calculations it will be assumed that the true value is between the two results – 1.2% gelatine.
The equivalent relative permittivity of the Spongostan scaffold therefore will be:
(26)
Based on this value the electric field
strength within the scaffold as a function of
the electrode potential difference was
determined (Figure 3.17) and is described
by the following equation:
(27)
Figure 3.17 – The electric field strength experienced by cells within the Spongostan scaffold as a
function of electrode potential difference
Using equation 27 it can be calculated that an electrode potential difference of 150 V generates an
electric field strength of 56.325 V/m within the scaffold. Monolayer cultures with the same
160
stimulation parameters are exposed to 57.02 V/m. Therefore, cells cultured in monolayer and on
the Spongostan scaffold experience comparable electric field strength.
Stimulating cells on a 3D substrate did not make a difference regarding the overall electric field
strength, can it perhaps have an effect in an alternative way?
Results from cellular level simulations show areas of lower electric field strength (approx. 27 V/m)
on the electrode facing sides and areas of higher electric field strength (approx. 82 V/m) on the
sides parallel to the electric field lines of gelatine regions compared to the overall electric field
strength in the culture medium (approx. 56 V/m) (Figure 3.18). The areas of altered field strength
are of comparable size to the gelatine regions generating them.
Figure 3.18 – Image showing the electric field strength in a part of the modelled region. Blue colour
corresponds to low, while red indicates high electric field strength. Note the lighter blue areas “left and
right” and the dark blue areas “above and below” the regions of gelatine.
These results indicate that cells cultured on a 3D scaffold do indeed experience a different stimulus
compared to monolayer cultures if placed under an electric field. Depending on the size of the
gelatine region and where the cell has attached, a cell can experience one and a half times higher or
EL
EC
TR
IC F
IEL
D
200 μm
161
half as strong stimulation (Figure 3.19). Cells can also experience areas of rapidly changing electric
field strength. What effect this may have on a cell is still not yet known.
Figure 3.19 – A graphical summary of the phenomenon observed around gelatine regions within the
scaffold.
This phenomenon is independent of the shape of the gelatine region (scaffold cross sections) as
shown in Figure 3.20-A. The areas of higher and lower electric field strength are the result of the
markedly different (in this case lower) relative permittivity of the gelatine compared to the culture
medium (Figure 3.20). This effect would also be present, if the permittivity of the gelatine would
be higher than the culture medium’s, only with the high and low field strength regions swapped
over (Figure 3.20-B).
It is also interesting to note that due to the lower electrical permittivity of gelatine, electric field
lines are “pushed away” from the scaffold material and run quasi-parallel to the surface (Figure
3.20-E). This means that cells attached to the gelatine would experience forces parallel rather than
perpendicular to the substrate as in monolayer cultures.
162
Figure 3.20 – The effect of the shape and relative permittivity of an object on the surrounding electric
field. The electric field strength around objects of different shape (A). The electric field strength and
field lines around disks with high (B and D) and low (C and E) relative permittivity. Colour legend
corresponds to electric field strength with blue indicating low and red indicating high values.
EL
EC
TR
IC
FIE
LD
163
3.3.4 The Frequency Response of the Bioreactor
Figure 3.21 – The Bode-diagram of the electrical bioreactor showing the magnitude (A) and phase
angle (B) of the signal at the bioreactor compared to the output of the signal source as a function of
frequency.
The magnitude of the signal (Figure 3.21-A) was unaffected at any of the tested frequencies. Phase
angle (Figure 3.21-B) was slightly altered below 100 Hz. This change in phase angle was below 1°
and as such should not affect the quality of the signal.
From the frequency response analysis it can be seen that the bioreactor will not alter the electrical
signals generated by the signal source and can be freely used at any frequency between 1 Hz and 1
GHz. It is also important to note that the impedance of the bioreactor is very high and as such can
be treated as a circuit break. Therefore, the load parameter on the signal source has to be set to its
maximum setting (in this case 10 kΩ). This also means that there should be no electrical current
A
B
164
flowing between the electrodes and that the signal generator should have no difficulty
generating/maintaining the potential difference.
However, the frequency response analysis did not consider the effect the amplifier may have on the
signals. In order to test this, and to verify the findings of the frequency response analysis, the
electrical signals leaving the amplifiers had to be measured in situ.
3.3.5 Signal Measurements
Figure 3.22 – The distortions observed with the high-voltage amplifier
Signals generated using the low voltage amplifier (TTi WA301) reached the bioreactor with their
intended amplitude, pulse width and waveform shape. Small distortions (Figure 3.23-A) were
observed with the 1 μs and 10 μs pulse trains; however they do not have a significant effect on the
quality of the signal. These small changes in pulse shape are probably the result of the charging
characteristics of the amplifier’s capacitors and, as such, are inherent to the device. They are only
apparent in the 1 and 10 μs pulses as only in these cases are the pulse widths so small that they are
comparable to the charging time of the capacitors.
In contrast to their low voltage counterparts, the high voltage signals (Model 2205 amplifier) were
amplified insufficiently and only reached the bioreactor generally at approx. 93% of their intended
amplitude (Figure 3.22). This was more significant in the case of the 1 μs (Figure 3.23-B) and 10
165
μs pulses which only reached 15.6% and 80% of their intended amplitude respectively. The shape
of the signal was also heavily distorted in the case of these two pulse widths (Figure 3.24-B). Some
signals transiently overshot their maximum voltage, for example by 1% in the case of 100 μs and
by 15% in the case of the 250 μs pulse trains (Figure 3.23-C). The 1999 μs pulse signals failed to
drop to 0 V in between pulses (lowest value was 84%) and were subject to noise (Figure 3.23-D).
Figure 3.23 – Examples of the minor distortion in the case of some of the low voltage pulses (A), the
distorted shape of 1 μs pulses (B), the overshooting of some of the signals (C) and the noise present in
1999 μs pulse width signals (D).
The causes of these distortions are disparate: The inadequate amplification noticed with all the high
voltage regimes is likely to be a result of a calibration error within the Model 2205 device. The
even greater reduction in amplitude and the distorted pulse shape in the case of the 1 μs and 10 μs
pulses are, however, signs of an insufficient slew rate. According to its specification the instrument
possesses a slew rate of min. 150 V/μs. As such, it should be able to raise its output from 0 V to
166
150 V under 1 μs and thus should be able to amplify all of the tested pulse widths. The results
indicate that this is not the case. The overshoots, transient distortions where the signal temporarily
exceeds its final value [369], that were observed with the 100 μs and 250 μs pulse trains are, on the
other hand, results of the signal rising too fast for the amplifier to be truly able to cope with it. The
noise present in the 1999 μs pulse regimes is caused by resonance (i.e. the signal being reflected of
the high resistance of the bioreactor) or perhaps by parasitic oscillations generated by feedback
between the different stages of the amplifier (e.g. input and output of a transistor) [370, 371].
Designing a high voltage amplifier with a high gain and slew rate is a difficult task with many
tradeoffs. The issues discussed above are not uncommon and would be likely present with any
other high voltage amplifier.
In summary, the regimes delivered using the low voltage amplifier will reach the bioreactor as
intended. This is not the case with the high voltage signals as these will suffer some form of
distortion when passing through the amplifier. This is, however, due to the limitations of the
technology and as such can only be taken into account when using the bioreactor system.
3.3.6 Comparison with the Literature
The electrical stimulus that a cell is exposed to in the capacitive bioreactor is now well understood.
Computer simulations have been performed in connection with electric fields and cells before.
Barash et al performed simulations of the current density between two carbon electrodes using
COMSOL, the same software that was applied in this study [372]. COMSOL was also used to
explore the electro-osmotic flow inside cells [373]. Computer simulations have been created to
explore the electric fields and currents generated as a result of in vivo stimulation in order to
improve the efficiency of electro-permeabilisation for electro-chemotherapy and DNA electro-
transfer [374]. Numerous studies have investigated the response of small parts of cell membranes
to voltage pulses using molecular dynamics simulations [375, 376, 377]. However, to the author’s
167
knowledge, such an in-depth investigation of the relationship between electrode potential and cell
experience as in this chapter has not yet been performed.
No data has been published in this regard connected to the other electrical bioreactors discussed in
the literature. The author’s of those scientific articles do state the electric field strength used in their
experiments, but it is not explained how these values were calculated or measured. Donnelly et al
states to have used an electric field strength of 350 and 620 V/m [338]. With the agarose bridges
field strength between 50 and 600 V/m have been applied [337]. These values are quite high. To
generate such an electric field strength using the capacitive bioreactor developed in this thesis an
electrode potential difference in the range of 1000-10000 V would be needed. Taking this into
consideration and the fact that no specialist equipment has been reported to have been used in
connection with the other the five electrical bioreactors discussed in the literature, it can be inferred
that the authors of those publication have not performed computer simulations taking into account
the culture medium or the effect polarisation, but have rather calculated electric field strength by
simply dividing electrode potential difference with the distance of the electrodes. This does not
give a true indication of the electric field strength the cells experience.
This is also the first study to explore the interaction of electric fields and biomaterial scaffolds.
Surface charge and surface electrical potential have been demonstrated to have an effect on cellular
adhesion and biomaterial biocompatibility [378, 379, 380], but the ability of scaffolds to “modify”
the electric fields around themselves has not been considered yet in tissue engineering. This
knowledge further emphasises the difference between monolayer and 3D culture conditions and
may shed important light on cell-extracellular matrix, cell-biomaterial interaction.
3.3.7 Future Possibilities: Biological Cell – Electric Field Interaction
The computer simulations in this chapter have not taken into account the cell itself. Could a cell’s
electrical properties have an effect of the field surrounding it just as the gelatine scaffold has? Or
could perhaps the shape of the cell determine how it experiences the electric field? A flat, spread
168
out cell from a monolayer culture is likely to be subjected to different forces than its more rounded
3D culture counterpart.
In order to investigate this three dimensional models of cells under different culture conditions are
required. But how could such a model be created, one that accurately depicts the complicated shape
of a cell?
The MicroCT method that was applied in the case of the gelatine scaffold cannot be used in this
scenario as cells, due to their small size and very low X-ray attenuation coefficient, are invisible in
scans. Even if visible, their shape would be obscured by the “shadow” of the substrate they are
cultured upon. Staining with heavy metals, such as Osmium Tetroxide, has been shown to render
cells visible in CT scans of 3D culture environments, however individual cells remained
undistinguishable from one another.
Z-stack scanning in confocal microscopy lends itself as a good alternative to MicroCT in this
regard. In this microscopy technique images (slices) are taken of the sample at different positions
(focus distances) along the axis of the objectives (the Z axis) (Figure 3.24). Having multiple images
along this third axis, Z-stack scanning provides three dimensional information, compared to
“traditional” 2D microscopy. The resolution of confocal systems is high enough to be able to
accurately depict cells, while the visibility of cells can be easily enhanced through fluorescent
staining methods. This method works well in the case of 3D cultures. The cells are more rounded,
therefore it easier to gather enough slices during Z-stack scans to be able to accurately reconstruct
the cell in a computer modelling environment.
Figure 3.24 - A graphical explanation of why it is easier to acquire sufficient slices if the cells are
rounded (3D scaffold) (A) compared to when they are more spread out (monolayer cells) (B).
169
Monolayer cultures however pose a greater challenge. The spread out, flat geometry of the cells
make it very difficult to acquire sufficient geometrical information to model the cells (Figure 3.24).
Placing the cells at an angle may provide a solution. The author has performed some preliminary
work demonstrating the feasibility of this method: hMSCs were cultured on glass coverslips and
stained with Cell Mask Deep Red (C10046, Invitrogen, Life technologies Ltd.) dye for the cell
membrane and DAPI (P36931, Invitrogen, Life technologies Ltd.) for the cell nucleus. Coverslips
were placed at 0°, 45°, 65° and 90° (Figure 3.25) and scanned using the Z-stack option on a Leica
confocal microscope.
Out of the four different specimen angles, 65° provided the
greatest amount of geometrical information and the best
contrast. Therefore an image stack taken this way was used as
the basis of the volume reconstruction. The images were
transferred to Aviso 8.0, cells were segmented and the resultant
geometry was transferred to COMSOL Multiphysics for electric
field simulations. Initial simulations were successfully run
proving the practical applicability of this method (Figure 3.26).
Figure 3.25 – The different
angles at which the cells were
scanned
170
Figure 3.26 – The process of creating a simulation based on graphical information from an image stack
It has been demonstrated that confocal microscopy can be used to create three dimensional
computerised models of cells both on 3D scaffolds and in monolayer cultures. Such models could
then be placed into electric field simulations of their respective local environments to gain
information on whether and how the cells interact with the field. The outcome of the simulations
could be verified through in vitro experiments using of voltage sensitive fluorescent dyes [381].
This modelling process (as any other) has its drawbacks. The reconstructed geometries can be too
complicated and therefore difficult to transfer between software or might require too much
computational power. The modelling process is also subject to “artefacts” from the scanning,
171
segmentation and reconstruction phases. If difficulties like these arise, as an alternative, the
reconstructed geometries could be used as a basis for the creation of simplified models that only
represent the crucial geometrical characteristics of the cells.
Independent of how the cellular models are created such simulation could yield important data on
the physical interaction of cells and electric fields, information that, to the author’s knowledge, is
not available yet.
172
3.4 CONCLUSIONS
In this chapter, through computer simulations and electrical measurements, the path from the user’s
input on the signal source to the stimulus the cells experience has been explored (Figure 3.27).
The function generator creates a signal based on the user’s settings. If the signal is passed through
the low voltage amplifier it reaches the bioreactor undistorted. On the other hand, if it is delivered
using the high voltage amplifier it will be altered significantly: Signals are not amplified fully, with
the majority of the signals reaching only 93% of their intended amplitude. In such a case a setting
of 150 V on the function generator will produce an electrode potential difference of 139.5 V.
Signals with pulse widths lower than 10 μs are distorted even more significantly, with their
maximum amplitude dropping proportionally to the shortness of the pulse, for example, to 80%
with 10 μs and to 15.6% with 1 μs. Regimes with short gaps in between the pulses, for example a
1999 μs - 500 Hz pulse train, are subject to noise, probably due to the signal reflecting off the high
resistance of the bioreactor. The frequency response analysis has demonstrated that the bioreactor
itself does not pose any alterations on the signal.
The characteristics of the electric field generated by the electrode potential have also been
examined. The computer stimulations showed that the third and the fourth generations of the
bioreactor produced the strongest electric field strength at the same electrode potential difference.
The electric field strength in the fourth generation bioreactor as a function of the various
experimental parameters (electrode distance, amount of culture medium and electrical potential)
has been determined. Based on this information a standalone application was created that allows
the electric field strength that the cells experience to be easily calculated at any of the bioreactor
configurations without the need to run further computer simulations.
173
Figure 3.27 – The path from user
input to “cell experience”
174
Simulation results also show that cells stimulated on a Spongostan scaffold experience an overall
electric field strength equivalent to those cultured in monolayer with the same bioreactor
configuration. However, their local environment will be much different with cells experiencing
disparate electric field strengths depending on where they have attached.
In summary, it has been determined how the user’s input relates to the electrical stimulus that the
cells experience both in the case of monolayer and 3D cultures and both with low and high voltage
regimes. Therefore, it can be concluded that the aim of this chapter has been fulfilled.
In the next and final chapter of this thesis, it will be explored what sort of biological effect this
“cell experience” might have and whether it can be useful for bone tissue engineering purposes.
175
Chapter IV
In Vitro Experiments
“With an anxiety that almost amounted to agony, I collected the instruments of life around me,
that I might infuse a spark of being into the lifeless thing that lay at my feet.”
Victor Frankenstein
176
4.1 INTRODUCTION
The two remaining objectives of this study were:
Objective #3 - Identify a parameter combination - a regime - that can promote hMSC proliferation.
Objective #4 – Identify a stimulating regime that has the capability to promote hMSC
differentiation and ECM production.
There are an infinite number of possible parameter combinations. Where to start? The results of the
author’s MSc project titled “Influence of the ‘PhyBack System’ on Primary Human Mesenchymal
Stem Cell Activity” served as a good starting point. In that study it was found that:
Direct stimulation with 1 μs shorts pulses delivered with 500 Hz frequency reduces cell
numbers and enhances osteogenic gene expression.
10 μs pulses delivered using the same frequency however increase cell numbers.
From these observations it was inferred that pulse width is the essential parameter that defines the
outcome of the stimulation. Furthermore, the results also suggested that brief “jolts” of electrical
stimulation are potentially what stimulates hMSCs best.
The conclusions of other researcher’s investigations can also aid in making a decision. From the
literature it can be known that higher electric field strength is generally the better choice (Table
1.2) and that frequency within the range of 100 Hz is applied in musculoskeletal tissue engineering
[19, 20, 138, 162, 174, 186]. Sundelacruz et al. [139] presented very interesting findings regarding
the importance of hMSC membrane polarisation. Chemical depolarisation of the cell membrane
was shown to suppress, while hyperpolarisation was demonstrated to enhance osteogenic
differentiation [139]. Mimicking this effect through the application of electrical stimulation could
have great benefits for tissue engineering.
177
Considering the above information it was decided that the following electrical stimulatory regimes
(Table 4.1) would be employed in this study:
Regime 1 1 μs
pulse width at 500 Hz with low (15 V) and
high (150V) voltage.
Regime 2 10 μs
Regime 3 250 μs
Regime 4 1999 μs
Table 4.1 – A brief summary of the electrical stimulatory regimes used in this study
With the help of these regimes (Figure 4.1) it can be ascertained whether the proliferation
enhancing effect of the 10 μs pulse, and the differentiation promoting nature of 1 μs pulse direct
stimulation is still present without an electrical current in the case of capacitive stimulation. The
rapid changes of electrical potential of the 1 and 1999 μs regimes can also yield important
information regarding the “jolt” hypothesis. Additionally, if capacitive electrical stimulation is able
to hyper- or depolarise cellular membranes, some signs of this may be present with the 10 and 250
μs regimes.
Figure 4.1 – The four regimes used in this study
178
In this fourth chapter, the outcome of the in vitro experiments conducted using these four regimes
will be discussed. After a summary of the materials and methods that were employed during these
experiments, the effects of delivering the four regimes at low voltage (15 V) will be presented. This
will be followed by an examination of why the findings of the low voltage experiments showed
such large variation, together with the steps made in order to address it. This chapter will finish
with the discussion of the results of the high voltage experiments and with the conclusions drawn
from the findings.
179
4.2 MATERIALS AND METHODS
4.2.1 Cell Culture
Commercial primary human bone marrow-derived Mesenchymal Stem Cells (PT-2501, Lonza UK
Ltd.) from four different donors (Table 4.2) were purchased at passage 2 and expanded to passage 4
in tissue-culture treated 75cm2 (T75) flasks containing 12 ml growth medium (MSCGM culture
medium, PT-3001, Lonza UK Ltd.) at 37 °C and 5% CO2 content in humidified atmosphere. Cells
were used for experiments at passage 5. The hMSCs from all four donors were expanded according
to the manufacturer’s instructions (Protocol A). Cells from Donor 4 were also expanded using a
low density culture protocol (Protocol B).
Lot no. TAN no. Age Race Sex
Donor 1 7F3915 16057-1 21 Black Male
Donor 2 7F3674 15839 22 Black Female
Donor 3 1F3865 22254 36 Black Male
Donor 4 318006 24935 27 Black Male
Table 4.2 – List of donors
180
PROTOCOL A – The “traditional” protocol that follows the supplier’s recommendations
Initiating the cell culture:
• Cells are plated at the recommended density of 5,000-6,000 cells per cm2. For a T75 flask
this is 375.000-450.000 cells per flask.
Maintenance of the culture:
• The cell culture is inspected 24h after initiation in order to assess cell viability. The culture
medium is changed post-assessment to remove any dead cells.
• Cells are inspected and the culture medium is changed every 3 or 4 days.
Sub-culturing the cells:
• Manufacturers highlight the importance of contact inhibition and recommend sub-culturing
at 90% confluence. However, in this study a cell density –70% confluence – smaller than
the recommended confluence was chosen as the limit.
• Judged by optical microscopy, if the cells are deemed approx. 70% confluent cells are
harvested for further expansion, freezing down or experiments.
• If the cells are not deemed to be at 70% confluence they are kept in culture until they
achieve this cell density.
• Cells are counted at the initiation and at the harvesting of each passage in order to enable
the accurate tracking of the performance of the culture.
Generally speaking commercial hMSCs are delivered at passage 2. In this study, cells were
expanded up to passage 4 and were used at passage 5 to ensure that their multi-potency has not
been compromised due to prolonged monolayer culturing.
181
PROTOCOL B – Low density expansion protocol
Initiating the cell culture:
• hMSCs are plated at a low density of 1000-1350 cells per cm2. For a T75 flask this is
75.000-100.000 cells per flask.
Maintenance of the culture:
• The cell culture is inspected 24h after initiation in order to assess cell viability. The culture
medium is changed post-assessment to remove any dead cells.
• Cell cultures are inspected daily to assess the confluence of the culture.
Sub-culturing:
• After 4 days hMSCs cultures reach approx. 50% confluence and are sub-cultured, frozen
down or used for experiments.
• Very importantly, cells are not allowed to expand above 50% confluence, as permitting the
cells to do so may result in a decreased proliferation rate, altered morphology and a loss in
differentiation potential.
• Culture time is kept consistent throughout all expansions and passages:
I.e. cell are always kept in culture for 4 days and then sub-cultured or used for experiments.
(It is also desirable to keep the culture time consistent between different donors, though
this might not be possible due to variations in proliferation rate.)
• Cells are counted at the initiation and at the harvesting of each passage in order to enable
the accurate tracking of the performance of the culture.
Generally speaking commercial hMSCs are delivered at passage 2. In this study, cells were
expanded up to passage 4 and were used at passage 5 to ensure that their multi-potency has not
been compromised.
182
4.2.1.1 Cell Revival
Frozen cryovials were thawed by holding in a 37 °C warm water bath for 1 min. After this the
contents of the vials were quickly transferred into 50 ml centrifuge tubes containing warm growth
medium. The tubes were centrifuged at 1200 rpm for 5 min. The supernatant was removed and the
cells were re-suspended in fresh Growth medium. The cells were counted using a C-Chip
disposable haemocytometer (Cronus Technologies Ltd.) and seeded into T75 flasks.
4.2.1.2 Sub-culturing
The hMSCs were washed three times with 7 ml warm PBS. After the washes 2 ml of Trypsin-
EDTA was added per flasks, followed by incubation at 37 °C for 5 minutes. After the incubation 7
ml growth medium was added to each flask. The cell suspensions were transferred to 50 ml
centrifuge tubes and centrifuged at 1200 rpm for 5 min. The supernatant was removed and the cell
pellets were re-suspended in fresh growth medium. The cells were counted using a C-Chip
disposable haemocytometer and plated into new T75 flasks.
4.2.1.3 Cell Freezing
Cells were trypsinised following the same steps as during sub-culturing, however, after
centrifugation the cell pellets were re-suspended in a mixture of 10% sterile dimethyl sulfoxide
(DMSO) and 90% foetal bovine serum. Approx. 750.000 cells were added to one cryovial.
Cryovials were frozen at -80 °C in Mr Frosty Freezing Containers (5100-0001, Fisher Scientific
UK Ltd.), followed by storage in liquid nitrogen.
4.2.1.4 Differentiation Media
4.2.1.4.1 Osteogenic Medium
Osteogenic medium was prepared by supplementing low-glucose Dulbecco’s Modified Eagle
Medium (DMEM) (E15-806, PAA Laboratories GmbH) containing L-glutamine, 10% heat
inactivated foetal bovine serum (FBS) and 1% antibiotics and antimycotics, with 10-5
mM
183
dexamethasone (D4902, Sigma-Aldrich Inc.), 10 mM β-glycerolphosphate (G9422-100G, Sigma-
Aldrich Inc.) and 50 µg/ml ascorbic acid (A8960-5G, Sigma-Aldrich Inc.).
4.2.1.4.2 Adipogenic Medium
Adipogenic differentiation medium was prepared by supplementing base medium (DMEM Low-
glucose, 10% FBS, 1% AB) with 0.5 μM dexamethasone (D4902, Sigma-Aldrich Inc.), 0.5 mM
isobutylmethylxanthine (IBMX) (I5879, Sigma-Aldrich Inc.) and 50 μM indomethacin (I7378,
Sigma-Aldrich Inc.).
184
4.2.2 Electrical Stimulation
Electrical stimulation was delivered in fourth generation electrical bioreactors with bare stainless
steel electrodes. Electrodes were positioned 10 mm from the bioreactor lid (electrode distance of
9.2 mm) for monolayer experiments. For 3D scaffold experiments this was 8.2 mm (electrode
distance of 11 mm). Positive electrical potentials were applied to the counter electrode, while the
upper electrodes were set to ground potential. Control samples were cultured in bioreactors with
the same configuration as treated cells, but without any electrical stimulation.
The following electrical stimuli (Table 4.3) were delivered in this study:
Table 4.3 – The electrical stimulatory regimes applied in this study. * - Electrical potential difference
as set on the signal source. ** - Electric field strength only drops to 4 V/m between pulses. *** -
Electric field strength only drops to 44.1 V/m between pulses
Treatment Culture
type
Electrode
potential
difference*
Electrical
field
strength
Pulse width Frequency Duration
1 μs regime
Monolayer 15 V 0.48 V/m
1 μs
500 Hz 1h/day
Monolayer 150 V 0.75 V/m
3D 150 V 8.19 V/m
10 μs regime
Monolayer 15 V 0.48 V/m
10 μs Monolayer 150 V 3.84 V/m
3D 150 V 42 V/m
250 μs regime
Monolayer 15 V 0.48 V/m
250 μs Monolayer 150 V 4.46 V/m
3D 150 V 48.8 V/m
1999 μs regime
Monolayer 15 V 0.48 V/m
1999 μs Monolayer 150 V 4.46 V/m **
3D 150 V 48.8 V/m***
Control N/A
185
4.2.3 Experiments
4.2.3.1 Low Voltage Experiments
50.000 passage 5 hMSCs were seeded in 1 ml growth medium into each well of 6-well plates
already containing 1 ml growth medium per well. Cells were allowed to attach by culturing
overnight at 37 °C and 5% CO2. Next day, the culture medium was changed to 2 ml osteogenic
medium per well, the sterile bioreactor lids were added and the electrical stimulation was started.
Electrical stimulatory regimes were delivered with 15 V electrode potential difference, which
corresponds to an electric field strength of 0.48 V/m. The culture medium was changed on day 4.
On day 7 (after 7 sessions of stimulation) the bioreactors lids were removed (Table 4.4). Cells were
cultured overnight and harvested on the next day for the Alamar Blue (n=6), PicoGreen (n=6) and
alkaline phosphatase (n=6) assays. In the initial experiment Donor 1 cells were treated. Based on
the results of this experiment the 1 μs and 10 μs was selected for further investigation and were
delivered to cells from Donor 1, Donor 2 and Donor 3.
Start Day 0 Seeding of samples
Day 1
Culture medium is changed to osteogenic medium
Bioreactor lids are added
Stimulation is started
Day 4 Culture medium change
Day 7
Last day of stimulation
Bioreactor lids are removed
End Day 8 Sample harvest
Table 4.4 – The steps of the monolayer electrical stimulation experiments
4.2.3.2 Expansion Optimisation
The findings of the low voltage experiments showed great variation and were contradictory. It was
hypothesised that this is potentially due to the compromised stem cell status of the hMSCs.
Inadequate culture conditions during expansion can impair the self-renewing capabilities and
186
differentiation potential of stem cells [382]. Different hMSC cultures impaired to different extents,
could very well show a great disparity in behaviour under the same experimental conditions.
There is strong indication in the literature that the key to maintaining the proliferation rate and the
differentiation potential of hMSCs is low cell density during expansion [382-384]. Therefore,
expansion optimisation experiments were performed comparing the effect of the traditional method
(from 40% to 70% confluence – Protocol A) to a low cell density technique (from 10% to 50%
confluence – Protocol B) on proliferation, and osteogenic and adipogenic differentiation potential.
Donor 4 hMSCs were expanded from passage 2 using either Protocol A or Protocol B. These
hMSCs will be referred to as Protocol A and Protocol B cells.
4.2.3.2.1 Proliferation Rate
The proliferation rate of Protocol A and Protocol B cells was examined from three perspectives.
4.2.3.2.1.1 Proliferation Rate during Expansion
Cell numbers were measured during expansions using C-Chip haemocytometers upon plating and
sub-culturing each passage. Data has been collected from 8 expansions carried out using Protocol
A and 7 using Protocol B.
4.2.3.2.1.2 Proliferation during Four Days of Culture
Protocol A and Protocol B cells were seeded at 1575 and 3150 cell/cm2 culture density into 24-well
plates (3000 and 6000 cells per well). Cells were cultured up to four days in 1 ml growth medium
per well. Cell numbers (n=6) were determined each day using the PicoGreen assay.
4.2.3.2.1.3 Cell Numbers after Fourteen Days of Culture
Protocol A and Protocol B cells were plated into 6-well plates containing 2 ml growth medium per
well at 3150 cells/cm2 culture density. Cells were cultured for 48 h after which the culture medium
was changed to growth, osteogenic and adipogenic medium in the appropriate plates. Cells were
187
cultured for 14 days with culture medium changes at day 4, 7 and 11. Samples (n=4) were
harvested on day 14 for PicoGreen assay.
4.2.3.2.2 Differentiation Potential
In order to assess the differentiation potential, Protocol A and Protocol B cells were seeded at 3150
cell/cm2 into 6-well plates and cultured for 14 days in growth, osteogenic and adipogenic medium
same as during the “proliferation” experiments. Samples at day 14 were assayed for the gene
expression (n=4) of osteogenic (alkaline phosphatase, collagen type I, osterix, osteopontin,
osteocalcin) and adipogenic markers (adiponectin, leptin) (qRT-PCR); alkaline phosphatase
activity (n=4); and lipid formation (Oil Red O staining) (n=4).
4.2.3.3 Seeding Optimisation
Another explanation for the above mentioned variation of the results during the “low voltage”
experiments was believed to be the inaccuracy of the currently employed cell seeding protocol.
Seeding density was believed to be the key parameter in this regard. As such experiments were
carried out comparing three different seeding densities. An optimal seeding density was expected to
produce equivalent cell numbers at the day of the seeding (Day 0) and after six days of culture
(Day 6).
Donor 4 commercial primary human Mesenchymal Stem Cells (PT-2501, Lonza UK Ltd.) were
cultured in T75 flasks in 12 ml growth medium at 37°C and 5% CO2 content in humidified
atmosphere according to Protocol B. At passage 5 cells were re-suspended at the three seeding
densities in Table 4.5.
Table 4.5 – The three examined seeding densities
Seeding volume Density Cell numbers Seeding density 1 100 μl 500000 cells/ml 50,000
Seeding density 2 200 μl 250000 cells/ml 50,000
Seeding density 3 1000 μl 50000 cells/ml 50,000
188
Day 0 - 50.000 cells (hMSC, p5) were seeded into 3 groups of 6 eppendorfs (mimicking three 6-
well plates) using each of the three seeding densities in Table 4.5. The eppendorfs containing the
samples were centrifuged at 7.2g for 5 min. The supernatant was removed and 1 ml of Lysis buffer
was added to each sample. Samples were vortexed and stored at -80 °C until further use. The
number of cells in the samples was determined using the PicoGreen assay (n=6).
Day 6 - 50.000 cells (hMSC, p5) were seeded into three 6-well plates containing 1 ml of growth
medium using each seeding density in Table 4.5. Each well was topped up to contain a final
volume of 2 ml of culture medium. Cells were cultured for 6 days. The medium was changed on
day 3. On day 6 the culture medium was removed and the cells were harvested for the PicoGreen
assay (n=6).
189
4.2.3.4 High Voltage Experiments
In these experiments the four stimulatory regimes were delivered with 150 V electrode potential
difference. This corresponds to an electric field strength of 4.46 V/m in the case of a monolayer
culture and 48.8 V/m for 3D scaffold cultures. Protocol B - Donor 4 cells were used throughout
these experiments.
4.2.3.4.1 Monolayer Cultures
50.000 passage 5 Donor 4 hMSCs were seeded in 100 μl growth medium into each well of 6-well
plates already containing 2 ml growth medium per well. Cells were allowed to attach by culturing
overnight at 37 °C and 5% CO2. Next day, the culture medium was changed to 2 ml osteogenic
medium per well, the sterile bioreactor lids were added and the electrical stimulation was started.
The culture medium was changed on day 4. On day 7 (after 7 sessions of stimulation) the
bioreactors lids were removed. Cells were cultured overnight and harvested on the next day. Cell
numbers were determined using the PicoGreen assay (n=6), while the fold change in mRNA
expression of ALPL and BMP-2 was measured using the qRT-PCR (n=6) technique.
4.2.3.4.2 Spongostan 3D Scaffold Cultures
One day before the experiment Spongostan porcine gelatine sponges were cut into approx. 1cm3
cubes using sterile razor blades. The Spongostan scaffolds (Figure 4.2) were glued into the centre
of 6-well plates using cell culture grade silicon glue. In order to minimise the risk of infection the
scaffolds were washed using 5 ml of 70% Ethanol twice for 15 min. The ethanol was removed and
the scaffolds were washed with 5 ml PBS thrice. The PBS was squeezed out of the scaffold with a
sterile pipette tip between and after the washes. In order to
equilibrate, 9.5 ml of growth medium was added to each
scaffold, and samples were stored overnight at 37 °C and
5% CO2.
Figure 4.2 – A 1 cm3 Spongostan scaffold
190
Start Day -1 Scaffold preparation
Day 0 Seeding of samples
Day 1
Culture medium is changed to osteogenic medium
Bioreactor lids are added
Stimulation is started
Day 4
Cell culture supernatant stored for ELISA
Culture medium change
Day 7
Cell culture supernatant stored for ELISA
Culture medium change
Day 11
Cell culture supernatant stored for ELISA
Culture medium change
Day 14
Last day of stimulation
Bioreactor lids are removed
Cell culture supernatant stored for ELISA
End Scaffold samples stored for the PicoGreen assay
Table 4.6 – The steps of the 3D scaffold experiments
On the next day the culture medium was removed from each well, squeezed out of the scaffolds
and aspirated. 200,000 passage 5 Donor 4 hMSCs was added to each scaffold in four injections of
100 μl cell suspension using a Finnpipette stepper pipette (612-6265, Fisher Scientific UK Ltd).
The samples were incubated for 30 min, after which 9.5 ml growth medium was added to each
well. The hMSCs were incubated overnight at 37 °C in 5% CO2. On the next day of the
experiment, the culture medium was removed and 9.5 ml osteogenic medium was added. The
bioreactor lids were placed onto the plates and the stimulation was started. 1 ml culture medium
was stored per sample for ELISA (n=6) on day 4, 7, 11 and 14 (Table 4.6). The culture medium
was changed on these days post-sample storage with the exception of day 14. The scaffolds were
harvested for PicoGreen assay (n=6) on day 14.
191
4.2.4 Assays
4.2.4.1 PicoGreen DNA Assay
The concentration of DNA, and therefore, correlating to this, the number of cells, was measured
using the PicoGreen assay.
4.2.4.1.1 Storage of Monolayer Samples
After being washed three times with 1 ml warm PBS, 1 ml Lysis buffer (1% Triton-X100 (T8787,
Sigma-Aldrich), 5% Tris-EDTA (TE) buffer, 94% ddH2O) was added to each well. Samples were
incubated for 5 min at room temperature, after which cells were scraped and the cell suspension
was transferred to 1.5 ml eppendorfs. Samples were stored at – 80 °C.
4.2.4.1.2 Storage of Scaffold Samples
The culture medium was removed from each well, after which the scaffolds were washed using 5
ml warm PBS three times. Following each washing step the scaffolds were squeezed using a sterile
pipette tip in order to remove the maximum amount of liquid possible. After the washes scaffold
samples were cut into four pieces using a sterile razor blade and transferred to 1.5 ml eppendorfs.
1ml Lysis buffer was added to each eppendorf. Samples were vortexed, sonicated for 15 min,
vortexed again and finally stored at -80 °C.
4.2.4.1.3 The Assay
All samples were freeze-thawed once in order to ensure that all cells have been lysed. After
repeated cycles of vortexing, 100 µl of the lysate (DNA suspension) was added in fours to wells of
a black 96-well plates in dark, together with 100 µl of PicoGreen stain (0.5% PicoGreen stock, 5%
TE buffer, 94.5% ddH2O) (Quant-iT™ PicoGreen® dsDNA Assay Kit, P1146, Life Technologies
Ltd.). Plates were read in a FLUOstar Optima fluorescent plate reader (BMG Labtech GmbH) at
480 nm excitation and 520 nm emission.
192
4.2.4.2 Alamar Blue Metabolic Assay
The metabolic activity of the cells was measured using the Alamar Blue fluorescent stain method.
After removing the culture medium, samples were washed once with 1 ml warm PBS. 1 ml Alamar
Blue solution, prepared in dark from 10 % Alamar Blue stain stock (AlamarBlue, DAL1100, Life
Technologies Ltd.) and 90 % culture medium, was added to each well in dark. Cells were incubated
for 60 min in 37 °C and 5% CO2 content. Post-incubation 100 µl was transferred from each sample
into a transparent 96-well plate in fours, and read at 560 nm excitation and 590 nm emission in a
FLUOstar Optima fluorescent plate reader (BMG Labtech GmbH).
4.2.4.3 Alkaline Phosphatase Assay
The amount of alkaline phosphatase produced by the cells was measured using the alkaline
phosphatase assay. Samples were washed with 1 ml warm PBS three times. 1 ml ddH2O was added
to each well and cells were scraped. The cell suspension was transferred into 1.5 ml eppendorfs
and lysed by three cycles of freeze/thawing. 20 µl of cell lysate from each sample was added in
fours to wells of transparent 96-well plates together with 200 µl of pNPP solution (1 mg/ml pNPP,
0.2 M Tris buffer in 5ml ddH2O) (SIGMAFAST™ p-Nitrophenyl phosphate Tablets, N1891-
50SET, Sigma-Aldrich Co.) and were read in a Multiskan Ascent colorimetric plate-reader
(Thermo Fisher Scientific Inc.) at 405 nm absorbance every 30 seconds for 30 min.
4.2.4.4 Gene Expression
The fold change in mRNA expression of genes of interest as a result of electrical stimulation was
detected using quantitative Reverse Transcription - Polymerase Chain Reaction (qRT-PCR).
4.2.4.4.1 Storage of Samples
Samples were washed three times with 1 ml PBS. After the wash, 1 ml PBS was added to each
well, cells were scraped and transferred to 1.5 ml eppendorfs. The cell suspensions were
centrifuged at 10g for 5 min, following which the supernatant was removed. The cell pellet was
snap frozen and stored at -80 °C.
193
4.2.4.4.2 RNA Isolation
RNA isolation was performed using the µMACS mRNA isolation kit (Miltenyi Biotec GmbH) and
µMACS mRNA isolation columns (Miltenyi Biotech GmbH) according to the manufacturer’s
specifications. In brief, 1 ml of Lysis/Binding buffer was added to each cell pellet. Samples were
vortexed, transferred to spin columns and centrifuged at 13g for 5 min. 50 μl Oligo Microbeads
was added to each sample post-centrifugation. Each sample was added onto a separate µMACS
mRNA isolation column and was washed two times with 200 μl Lysis/Binding buffer and four
times using 100 μl of Wash buffer.
4.2.4.4.3 cDNA Synthesis
cDNA synthesis was performed using the µMACS cDNA synthesis module (Miltenyi Biotec
GmbH) following the manufacturer’s specifications. In brief, the samples still on the mRNA
isolation columns were washed two times with 200 μl Equilibration/Wash buffer. After the washes
20 μl of lyophilized enzyme was added to each column. The columns were sealed using 1 μl
Sealing solution and incubated at 42 °C for 1h. After the incubation period the columns were rinsed
twice using 100 μl of Equlibration/Wash buffer. 20 μl cDNA release solution was added to each
column and the samples were incubated at 42 °C for 10 min. After this second round of incubation
50 μl of cDNA elution buffer was added to each column. The eluted cDNA samples were collected
in sterile 1.5 ml eppendorfs and stored at -80 °C until further use.
4.2.4.4.4 Polymerase Chain Reaction
1 μl of the cDNA samples was added to wells of MicroAmp Optical 96-Well Reaction plates (Cat.
no.: 4306737, Life Technologies Ltd.) in threes for each gene of interest (Table 4.7). Following
this, a mixture of 1.25 µl of the appropriate Taqman probe (Taqman Gene Expression Assays, Cat.
No. 4331182, Life Technologies Ltd.), 12.5 µl TaqMan Gene Expression Master Mix (4369016,
Life Technologies Ltd.) and 10.25 µl of RT-PCR Grade water (AM9935, Life Technologies Ltd.)
was added to each well. Plates were covered using MicroAmp optical adhesive films (Cat. no.:
194
4311971, Life Technologies Ltd.) and read in a StepOne Plus Sequence detection system (Life
Technologies Ltd.). Data was analysed using the Comparative Ct (ΔΔCT) method.
Table 4.7 – A list of the genes assayed in this study
4.2.4.5 BMP-2 and BMP-7 Production
The amount of BMP-2 and BMP-7 secreted by cells was measured using the Enzyme-linked
Immunosorbent Assay (ELISA) method. 1 ml cell culture supernatant was stored for each sample
by transferring into 1.5 ml eppendorfs and freezing at -80 °C. Bone morphogenetic protein 2 and 7
concentrations in the cell culture supernatant was measured using the DuoSet ELISA kits (human
BMP-2 kit, DY355 and human BMP-7 kit, DY354; R&D Systems Inc.) following to the supplier’s
instructions.
Nunc MaxiSorp 96-well plates (Cat. no.: 12-565-135, Thermo Scientific Ltd.) were coated by
incubating with 100 μl Capture Antibody (1 μg/ml) per well overnight. The capture antibody was
removed and the wells were washed four times using 100 μl of Wash Buffer (Table 4.8). The four
consecutive washes with 100 μl of Wash Buffer will be referred to as the Washing Step. Plates
were blocked by incubating with 300 μl of Reagent Diluent per well for 1 hour at room
temperature. After the incubation, plates underwent a Washing Step, following which 100 μl of the
samples were added to the plates in threes. After this the plates were incubated for 2 h at room
temperature. After a Washing Step, 100 μl of the Detection Antibody was added to each well.
Plates were incubated for 2 h at room temperature. Following another Washing Step, 100 μl of
Gene Gene Symbol Assay ID
Adiponectin ADIPOQ Hs00605917_m1
Alkaline phosphatase ALPL Hs01029144_m1
Bone morphogenetic protein 2 BMP-2 Hs00154192_m1
Collagen type I COL1A Hs00164004_m1
Glyceraldehyde-3-phosphate dehydrogenase GAPDH Hs02758991_g1
Leptin LEP Hs00174877_m1
Osteocalcin BGLAP Hs01587814_g1
Osteopontin SPP1 Hs00959010_m1
Osterix SP7 Hs01866874_s1
195
Streptadivin-HRP solution was added to each well, after which the plates were incubated for 20
min. After the incubation the Straptadivin-HRP solution was removed and plates underwent the last
Washing Step. Post-washing, plates were incubated for 20 min with 100 μl of Substrate Solution
per well. When the incubation was finished 50 μl of Stop Solution was added to each well and
plates were gently tapped to ensure thorough mixing. Plates were immediately read for absorbance
at 450 nm with wavelength correction at 540 nm in a Multiskan Ascent colorimetric plate-reader
(Thermo Fisher Scientific Inc.).
Reagent name Description
Capture Antibody Appropriate capture antibody at 1 μl/mg concentration for BMP-2
and 2 μg/ml concentration for BMP-7 in PBS
Detection Antibody Appropriate detection antibody at 1 μl/mg concentration in Reagent
Diluent for BMP-2 and 0.5 μg/ml concentration for BMP-7
in Reagent Diluent with 0.2% heat inactivated goat serum
Streptadivin-HRP Streptadivin conjugated to horseradish-peroxidase diluted to
1:200 working concentration in Reagent Diluent
Wash Buffer 0.05% Tween 20 (P9416-50ML, Sigma-Aldrich Inc.) in PBS
Reagent Diluent 1 % Bovine Serum Albumin in PBS - Cat. No.: DY995
Substrate Solution 1:1 mixture of H2O2 and Tetremethylbenzidine - Cat. No.: DY999
Stop Solution 2 N H2SO4 - Cat. No.: DY994
Table 4.8 – A list of the reagents used in the BMP-2 and BMP-7 ELISA assays. All reagents were
purchased from R&D Systems Inc. with the exception of the Tween 20 and the PBS.
4.2.4.6 Oil Red O Staining
The amount of lipids accumulated by the adipose differentiated hMSCs was determined using
quantitative Oil Red O staining. After three washes with 1 ml warm PBS, samples were fixed by
incubating in 1 ml neutral buffered formalin (10%) for 30 min at room temperature. After the
incubation the formalin was removed and samples were washed with 1 ml of distilled water. 1 ml
of 60% isopropanol was added to each well and samples were incubated for 5 min at room
196
temperature. The isopropanol was removed and 1 ml of Oil Red O staining solution (60% of 0.3%
Oil Red O stain stock (O0625-100G, Sigma-Aldrich Co.) in distilled water) was added to each
well. Samples were incubated for 15 min at room temperate after which the staining solution was
removed and samples were washed with 1 ml distilled water three times. 1 ml of isopropanol was
added to each well, and samples were incubated at room temperature for 15 min with gentle
agitation. 100 μl of each sample was transferred into transparent 96-well plates in fours and read in
a Multiskan Ascent colorimetric plate-reader (Thermo Fisher Scientific Inc.) at 490 nm absorbance.
4.2.4.7 Optical Microscopy
Optical light microscopy of the hMSCs was performed using a Leica inverted phase-contrast
microscope (Leica Microsystems Wetzler GmbH) and Diagnostic Model 3.2.0 camera (Diagnostic
Instruments Inc.).
4.2.4.8 Statistical Analysis
Statistical analysis was performed in the commercial software GraphPad Prism v5.0 (GraphPad
Software, Inc.). Multiple treatments were compared using One-Way ANOVA followed by Tukey’s
post-test. Student’s T-test (unpaired, two-tailed) was used where only two treatments were
compared. If there were two factors (e.g. donor and medium type), statistical analysis was
performed using Two-Way ANOVA followed by Bonferroni’s post-test. Statistical differences
with a p value smaller than or equal to 0.05 were considered significant. All figures display mean ±
standard deviation values.
197
4.3 RESULTS AND DISCUSSION
4.3.1 Low Voltage Experiments
The effect of 1, 10, 250 and 1999 μs pulse stimulation with 0.48 V/m electric field strength on the
proliferation, differentiation and metabolic activity of hMSCs (Donor 1) was compared. Initial
experiments showed no significant difference between cell numbers after seven sessions of
stimulation with either of the electrical regimes (Figure 4.3). Metabolic activity however appeared
to be significantly enhanced (p<0.001) by capacitive electrical stimuli in a dose dependant manner,
with highest metabolic levels detected with the longest (1999 μs) pulses (Figure 4.4). 1 and 10 μs
pulses performed in a similar manner as there was no significant difference found between these
two pulse widths. There was a significant difference between the 10 and 250 μs pulses and 250 and
1999 μs pulses. Differentiation, as measured by the alkaline phosphatise activity, was significantly
suppressed by the 1 and 1999 μs regimes, but enhanced by the 10 μs regimes (Figure 4.5). These
findings are contradictory to the observations made during the author’s master project where 1 μs
pulses promoted osteogenic differentiation, while 10 μs pulses did not. This difference can be
potentially explained by the facts that different stimulation techniques and different assay types
were used in the two studies.
For their effect on alkaline phosphatise activity the 1 and 10 μs pulses were selected for further
investigation.
198
Figure 4.3 – Cell numbers (n=6) in hMSCs cultures after 7 days of 1h/day electrical stimulation
Figure 4.4 – The effect of electrical stimulation on the metabolic activity of hMSCs (n=6) after 7 days
of 1h/day stimulation (“*” = p<0.05 compared to Control samples)
Figure 4.5 – The effect of electrical stimulation on the alkaline phosphatase activity of hMSCs (n=6)
after 7 days of 1h/day stimulation (“*” = p<0.05 compared to Control samples)
199
4.3.1.1 Three Donor Repeat Experiments
The effect of the 1 and 10 μs pulse regimes on the cell numbers, metabolic activity and alkaline
phosphatise activity of hMSCs from three different donors was examined (Figure 4.6).
4.3.1.1.1 Cell Numbers
Donor 1 and Donor 3 cell numbers were significantly decreased by 1 μs pulse stimulation, with
mean cell numbers lowered to 97% (p<0.05) and 88% (p<0.05) respectively. Cell numbers in
Donor 2 samples was not affected by this regime. However, Donor 2 cell numbers did drop
significantly to 90% with the 10 μs regime. 10 μs had no effect on Donor 3 cells, but increased cell
numbers to 104% in Donor 1 samples (p<0.05).
4.3.1.1.2 Metabolic Activity
Both regimes in the case of all three donors produced a significant change in metabolic activity.
The results are summarised in Table 4.9.
Metabolic activity compared to
controls
1 μs 10 μs
Donor 1 Decrease 80% Decrease 69%
Donor 2 Increase 120% Increase 124%
Donor 3 Increase 110% Increase 108%
Table 4.9 – The metabolic activity of stimulated cells from the three donors compared to controls
Electrical stimulation resulted in a seemingly dose dependent decrease in the case of Donor 1. On
the other hand, treatment with the regimes produces an enhanced metabolic activity in the case of
Donor 2 and 3.
200
Figure 4.6 – The effect of 1 and 10 μs stimulation on the cell numbers, metabolic activity and
alkaline phosphatase activity of hMSCs from three different donors (n=6). (“*” indicates p<0.05)
201
4.3.1.1.3 Alkaline Phosphatase Activity
There was no significant change in the alkaline phosphatase activity with either of the two regimes
with the exception of a significant up-regulation in the case of 1 μs stimulated Donor 3 samples.
4.3.1.1.4 Comparison of the Results
There is great variation between the outcome of the stimuli as summarised in Table 4.10 with no
underlying trend being apparent. More importantly, results of the experiments carried out using
Donor 1 cells are contradictory to one another, with for example 1 μs pulses both increasing and
suppressing the metabolism of the cells.
Table 4.10 – A graphic summary of the results gained in the two experiments
The differences in mean cell number can potentially be explained by “normal” statistical variation:
The differences were small, within the range of a few percentages, and might appear statistically
significant due to the small standard deviation of the various treatment groups. For example, 1 μs
pulses produced a 3% decrease in mean third donor cell number compared to controls, while the
7% increase again with the third donor, but now with 10 μs pulses was not significant. However,
202
this is not the case with metabolic and alkaline phosphatise activity where the observed
(contradictory) changes are more substantial.
Another surprising and perhaps troubling issue is the great difference between the cell numbers,
metabolic activity and alkaline phosphatise activity of the different donors. Donor 1 and Donor 2
samples produced cell numbers in the range of 100,000 after eight days in culture, while this
number was 150,000 for Donor 3 cells. Or, for example, Donor 3 cells showed five times greater
alkaline phosphatise activity than Donor 2 cells. Metabolic activity was similarly different between
the donors.
Based on these findings the following conclusions were drawn:
1. No detectable effect of electrical stimulation was observed on the cell numbers, metabolic
and alkaline phosphatise activity of hMSCs after seven sessions of stimulation at 0.48 V/m
electric field strength.
2. Any effect, if present, was not powerful enough to overcome the “noise” presented by the
great variation in the behaviour of the hMSCs.
3. Furthermore, the variation between the experiments cannot be explained by the “natural”
difference between the behaviour of cells from different donors, as repeat experiments
using cells from the same patient (Donor 1) were also contradictory. The explanation lies
elsewhere.
203
4.3.2 Investigating the Variation in hMSCs Behaviour
It was postulated that the disparate performance of the cells during the low voltage experiments
was a result of two factors: Inadequate culture conditions and inaccurate seeding.
4.3.2.1 Optimised hMSC Culture Conditions
There is no commonly accepted, standardised protocol in existence for the culture and expansion of
mesenchymal stem cells [383-386]. The protocol that has been employed for the expansion of the
hMSCs (Protocol A) is based on the supplier’s instructions and on generally accepted cell culture
practice. However, evidence from the low voltage experiments suggests that this protocol might not
be able to sufficiently maintain the quality of the hMSC cultures.
There is strong evidence in the literature that culturing MSCs at a too high density will impair their
self-renewing capability and differentiation potential. This coincides with a shift in sub-phenotype.
MSCs with high proliferation rate and lineage plasticity are denoted as Rapidly Self-renewing (RS)
MSCs. This sub-phenotype can be found in early passages and low density cultures, displaying a
small, thin, spindle-like, agranular morphology [383, 384, 387, 388]. However, inadequate culture
conditions, for example high cell density, will result in the MSCs taking up a second phenotype
designated in the literature as Slow Replicating (SR) MSCs [382-384, 387-389]. These cells are
large, spread-out, rhomboidal and granular in morphology and have been observed to contain a
high number of vacuoles [384, 387, 390]. SR cells propagate very slowly and have diminished
differentiation capabilities with usually only being able to commit to the osteoblastic lineage [384,
389, 390]. MSC cultures not proliferating well will contain a high proportion of SR cells and a low
proportion of RS cells [383].
If Protocol A was unable to maintain the hMSCs in the RS sub-phenotype, Donor 1, 2 and 3
cultures could be compromised, possibly to different extents. (I.e. Donor 1, 2 and 3 cultures would
contain different proportions of RS and SR cells). This could very well explain the large
differences in cell numbers and alkaline phosphatise activity witnessed in the low voltage
204
experiments. Light microscopy seemingly verified this hypothesis showing large, spread-out SR
type cells in the hMSC cultures (Figure 4.7).
Figure 4.7 – Sample with hMSCs displaying the spread out, rhomboidal SR morphology (A) compared
to spindle-like, small RS cells (B). The contrast of the images have been modified in order to enhance
visibility.
In order to further test this hypothesis and to optimise the protocol with which the hMSCs are
cultured, cells from a fourth donor (Donor 4) were expanded using both the traditional (Protocol A)
method and a low density culture technique (Protocol B). The proliferation rate and osteogenic and
adipogenic potential of the cells expanded with the two protocols was compared.
4.3.2.1.1 Proliferation Rate
Data was collected from 8 expansions carried out using Protocol A and 7 using Protocol B.
Figure 4.8 - A comparison of the fold increase (A) and fold increase per day (B) in Protocol A and
Protocol B cultures. (“*” indicates p<0.05)
205
Figure 4.8-A displays the fold increase in cell numbers in expansions following Protocol A and
Protocol B (i.e. the ratio of the cell numbers counted at the harvest and the initiation of a passage).
A higher column indicates an increased proliferation rate, while a smaller error bar suggests a more
reliable, predictable expansion.
The mean fold increase was higher in the case of Protocol B, although non-significantly different,
compared to Protocol A. The standard deviation on the other hand was much larger, approx. ±40%,
in the case of Protocol A, while only approx ±20% in the case of Protocol B. An even more
marked difference can be seen between the two protocols in Figure 4.8-B. This figure displays the
fold increase per day, which was significantly higher in the case of Protocol B (p=0.004).
Furthermore, based on a qualitative evaluation, in cultures expanded following Protocol B no or
very few SR type cells were present up to passage 4, while Protocol A expansions contained a
varying size but substantial proportion.
These findings show that, although Protocol A is able to produce just as many cells as Protocol B,
it requires substantially more time and does so less predictably. Additionally, microscopy
observations suggest that it is unable to maintain the RS sub-phenotype. This is supported by the
results of further proliferation experiments.
Figure 4.9 – Cell numbers in Protocol A and B cultures during four days of expansion (n=6).
When cultured for four days in growth medium Protocol B cells showed significantly higher cell
numbers from 48 h onward (p<0.001) and a 10% higher overall proliferation rate (Figure 4.9)
compared to Protocol A cells. Similarly, in various differentiation media, after 14 days of culture,
Protocol B samples showed significantly higher cell numbers (p<0.001), with 23% more cells
being present in growth, 12% more in osteogenic and 77% more in adipogenic medium samples
206
compared to Protocol A (Figure 4.10). The fact that Protocol B cell show a higher proliferation rate
compared to cells cultured using the traditional method both in growth and differentiation medium
suggests that this protocol is better able to maintain the self-renewing capability of hMSCs,
possibly through favouring the RS sub-phenotype.
Figure 4.10 – Cell numbers in Protocol A and Protocol B samples after 14 days in various
differentiation media (n=4, “*” indicates p<0.05)
4.3.2.1.2 Osteogenic Differentiation Potential
The ability of Protocol A and Protocol B cells to commit to the bone lineage was measured through
the mRNA expression of osteogenic markers and alkaline phosphatise activity.
qRT-PCR results show no significant difference between the mRNA expression of alkaline
phosphatase (ALPL) and osterix between the two protocols after 14 days of culture (Figure 4.11).
Protocol B cells expressed significantly higher levels of osteopontin in osteogenic medium and
osteocalcin in growth medium. Osteogenic chemical stimulation was able to enhance collagen type
I expression in Protocol B, but not in Protocol A, cultures. However, all of these fold increases
were in the range of, or below, 2-fold. As a rule of thumb, such small fold changes, even if
statistically significant, are not considered decisive. As such, it can be inferred that there was no
overall difference between the osteogenic gene expression of cells from the two protocols.
However, at the secretional level Protocol B cells did show substantially higher alkaline
207
Figure 4.11 – The fold expression of ALPL, osterix, collagen type I, osteocalcin and
osteopontin mRNA in Protocol A and B cells cultured in osteogenic and growth
medium for 14 days (n=4, “*” indicates p<0.05).
208
phosphatase enzymatic activity (Figure 4.12), suggesting that there is a difference between the
differential potential of the two cell cohorts.
Figure 4.12 – Alkaline phosphatase activity of Protocol A and Protocol B hMSCs after 14 days in
osteogenic and growth medium (n=4, “*” indicates p<0.05).
4.3.2.1.3 Adipogenic Differentiation Potential
The above observation is supported by the findings of adipogenic differentiation experiments.
Figure 4.13 – Adiponectin (left) and leptin (right) mRNA expression in Protocol A and Protocol B
samples cultured in growth and adipogenic medium for 14 days (n=4, “*” indicates p<0.05).
Adiponectin mRNA expression was significantly higher (approx. 3-fold) with Protocol B cells in
both growth and adipogenic medium (Figure 4.13). Protocol B samples displayed lower Leptin
levels in growth medium, while there was no statistically significant difference between the two
protocols in adipogenic medium. Lipid formation per cell, measured by Oil Red O staining (Figure
4.14), was significantly higher in Protocol A samples.
209
These findings suggest that, even so there was no major difference between the osteogenic and
adipogenic potential of cells cultured with the two protocols in gene expression, at the secretional
level Protocol B favours bone, while Protocol A promotes adipogenic differentiation.
Figure 4.14 – The amount of lipid per cell in Protocol A and Protocol B samples after days in
adipogenic medium (n=4, “*” indicates p<0.05).
These results demonstrate that cell density during expansion plays an important role in defining the
behaviour of hMSCs. High (above 50%) culture density impairs the self-renewing capability and
osteogenic potential of this cell type. Cultures impaired to different extents would show disparate
cell numbers and osteogenic marker levels in experiments. This is exactly what was witnessed in
the low voltage experiments. Furthermore, with the adaptation of Protocol B as the hMSCs
expansion protocol, the variation between repeats was minimised and experiments were repeatable
in the latter phases of this study. This provides further evidence that indeed it was the inadequate
and inconsistent expansion of hMSCs that resulted in the large variation of cellular performance in
the low voltage studies. It is also worth noting that these findings indirectly support, although do
not prove, the “RS-SR sub-phenotype” hypothesis. Further investigations are necessary in this
regard.
4.3.2.2 Optimised Cell Seeding
Inaccurate cell seeding could have also contributed to the large variation between the performances
of the different cell groups. This was tested by comparing the cell numbers in two equally seeded
210
and cultured 6-well plates. It was found that after six days of culture there was significantly
different cell numbers in the two culture vessels (Figure 4.15). The difference in mean cell number
was approx. 5%, comparable to some of the variation seen in the low voltage experiments. This
suggests that inaccurate seeding could indeed explain some of the inconsistencies.
Figure 4.15 – The comparison of cell numbers in two 6-well plates after 6 days in culture (n=6, “*”
indicates p<0.05)
It was theorised that seeding cell concentration is the important factor in defining the accuracy of a
seeding protocol. In dilute cell suspensions, with substantially more space being available,
gradients in cell density form due to gravity much faster and to greater extent. Using a more
concentrated cell suspension therefore could increase the accuracy of the seeding. In order to test
this hypothesis hMCSs were seeded at three different concentrations, and their cell numbers was
assayed at seeding and after six days in culture (Figure 4.16).
The “original” seeding concentration of 50k cells/ml produced un-equivalent cell numbers not just
after six days (as found before) but right at the seeding as well. Similarly, there were significant
differences (p<0.001) found with a five times more dense cell concentration, 25k cells/ml, at day 0.
However this variation in cell numbers disappeared after six days of culture.
211
Figure 4.16 – Cell numbers in three 6-well plates seeded with three different cell concentrations at day
0 (the day of plating) and at day 6 (after 6 days in culture). (n=6, “*” indicates p<0.05)
A likely explanation of this is that the six days was enough for the cells in each plate to fill up the
available culture area equally, compensating for any initial inconsistencies. The best results were
obtained with the 500k cells/ml concentration as this method provided equivalent cell numbers at
seeding and after six days of culture.
Seeding density (i.e. cell numbers per cm2 or cm
3 of the substrate) has been optimised for many
different monolayer and 3D culture applications [391-393]. However, very little information is
available regarding the importance of seeding concentration (i.e. the concentration (cell/ml) at
which the cells are delivered to the substrate). Wiedman-Al-Ahmad et al published some results
212
regarding the effect of cell concentration during the seeding of bone tissue constructs, however no
data was presented in connection with seeding accuracy [394]. Therefore, there is no true basis
available in the literature (to the author’s knowledge) to which the findings of this experiment can
be compared.
Nonetheless, it can be concluded that seeding concentration does have a significant effect on the
accuracy of the seeding process and that inaccurate seeding could very well have contributed to the
variations in cell numbers observed in the low voltage experiments. With an optimal seeding
concentration found, these issues can now be avoided.
213
4.3.3 High Voltage Experiments
Along with adopting an optimised cell expansion (Protocol B) and cell seeding protocol,
experiments were modified in a third way. It was decided that electrical stimulation will be carried
out using much higher electrode potential (150 V). (This was not possible before as it required the
purchase of a specialist high voltage amplifier.) The higher electrode potential will generate
substantially stronger electric field strength, making any effect the stimulation might have on the
cells much stronger and/or more likely to be present.
4.3.3.1 The Effects of Electrical Stimulation on hMSC Proliferation
1, 10, 250 and 1999 μs pulse width regimes were delivered to hMSCs cultured in monolayer and on
3D Spongostan scaffolds.
Figure 4.17 – Cell numbers (n=6) in monolayer cultures after 7 days of stimulation. (“*” indicates
p<0.05)
In monolayer cultures 1 and 1999 μs pulse regimes significantly lowered cell numbers after seven
session of stimulation to approx. 80% (p<0.001) and 93% (p<0.005) respectively (Figure 4.17). 10
and 250 μs treatment made no significant difference compared to controls. In contrast to these
findings, in scaffold cultures none of the stimulatory regimes had a significant effect on cell
numbers after fourteen sessions (Figure 4.18).
214
Figure 4.18 – Cell numbers (n=6) in Spongostan scaffolds after 14 days of stimulation
4.3.3.2 Osteogenic Differentiation
None of the tested regimes had an effect on ALPL and BMP-2 expression at the mRNA level
(Figure 4.19). Neither was any significant difference found in the levels of BMP-2 and BMP-7
secreted by the cells during fourteen days of differentiation, as measured by the ELISA assays
(Figure 4.20).
Figure 4.19 – Alkaline phosphatase (left) and BMP-2 (right) mRNA levels after 7 sessions of
stimulation in monolayer samples (n=6).
215
Figure 4.20 – BMP-2 (left) and BMP-7 (right) production per day after 14 sessions of stimulation in
Spongostan scaffold samples (n=6).
4.3.4 Discussion
The findings of the high voltage experiments show that capacitive electrical stimulation is able to
affect cell numbers. The brief pulses of the 1 and 1999 μs regimes, albeit having a weaker electric
field strength, were able to lower cell numbers in monolayer culture, while the longer (and
stronger) pulses of the 10 and 250 μs stimuli were not. This observation supports the “jolt”
hypothesis, suggesting that cells are able to sense, or are more susceptible to, rapid changes of
electric field strength.
These results are in agreement with the author’s findings during his Master’s project. In that study
1 μs pulses, although delivered using direct stimulation, caused a similar reduction in hMSC cell
numbers after seven sessions of stimulation. In the Master’s project, however, this reduction in
hMSC numbers paired with an up-regulation of ALPL gene expression, while 10 μs pulses were
able to increase cellular proliferation. The fact that these effects were not present with capacitive
stimulation suggests that they were mediated by the electrical current of the direct method. This is a
very interesting observation, showing that electrical current, albeit carrying a risk of damaging to
the cells, is potentially important for the delivery of some of the beneficial effects of electrical
stimulation. Similar findings were reported by Griffin et al in a recent study comparing the
behaviour of hMSCs after direct and capacitive electrical stimulation [395]. It was found that the
gene expression of IGF-1 and TGF-β1 was significantly higher with direct stimulation compared to
216
capacitive coupling [395]. The capacitive method, on the other hand, significantly enhanced the
expression of MMP2 and MT1-MPP, showing a disparity between the effects of the two techniques
[395]. It can be concluded that the observations of this thesis support the hypothesis that some of
the effects of electrical stimulation are indeed mediated through electrical currents.
It is also interesting to note that, in contrast to the monolayer findings, the 1 and 1999 μs pulse
regimes did not reduce cell numbers in the case of 3D scaffold cultures. The reason behind this is
not known. However, these findings are not completely surprising, considering that cells have been
demonstrated to behave disparately in monolayer and 3D cultures, adopting different cell shapes,
proliferation rates and gene expression profiles [396-399]. For example, embryonic stem cells
express significantly higher levels of proliferation, differentiation and extracellular matrix related
genes in 3D cultures [400]. Studies conducted on MSCs report different cell size, surface antigen
expression, self-renewal capability, osteogenic and adipogenic potential, and widespread changes
in cellular architecture between the two culture methods [401-403]. Microarray experiments
performed by Wang et al showed extensive differences in the gene expression profile of MSCs
cultivated in 3D compared to those maintained in monolayer [403]. These findings suggests that
there is a fundamental difference between 2D and 3D cultured MSCs, providing an explanation to
why cell numbers were lowered in monolayer but not in Spongostan samples. Unfortunately, there
are no publications available, to the author’s knowledge, comparing the effect of electrical
stimulation delivered to monolayer and 3D cultures.
Understanding which “component” of the 3D setting is blocking the effects of electrical stimulation
could shed important light on the underlying mechanisms behind this modality.
The cause of disparate 2D-3D MSC behaviour is believed to be due to differences in substrate
stiffness, cell shape, surface chemistry and chemical gradients between the two culture methods
[398, 404-406]. Substrate stiffness and cell shape are both powerful determinants of stem cell
behaviour. Engler and colleagues have observed that MSCs show a tendency towards the bone
lineage on rigid surfaces, and greater neural and muscle differentiation on more elastic substrates
217
[407-409]. Similar findings were reported regarding cell shape by Kilian et al, demonstrating that,
if MSCs are forced into a rounded morphology they differentiate into adipocytes, while those kept
in a rectangular form commit to the bone lineage [410].
Both substrate stiffness and cell shape are converted into cellular response through the
mechanotransduction pathway [408-410]. Focal adhesions, integrins, microtubules, and the
intracellular factors ERK, Wnt and Rho are known to be key players in this pathway [409-414].
All of these cellular components have also been implicated to be responsible for the sensing of
electric fields [137, 147, 151, 152, 205, 207]. Culturing cells under different conditions (2D-3D)
could have very well caused changes in the mechanotransduction pathway, inhibiting electrical
stimulation’s ability to instigate a cellular response (Figure 4.21). Therefore, it can be inferred that
the findings of this study, that the regimes affected cells differently in monolayer and scaffold
culture, suggest that electrical stimulation is sensed at least partially through the
mechanotransduction pathway.
Figure 4.21 – A possible explanation of the differences observed between the effect of electrical
stimulation in monolayer and scaffold cultures based on the mechanotransduction pathway.
However, alternative explanations are also plausible. Perhaps these brief pulses force cells to
assume a more spread out morphology, restricting the space available to cells. Or perhaps the
signals prevent cells from “piling” upon each other, similarly limiting the maximum cell number
per well. As in scaffold cultures the area available to cells is much greater, and cells can grow in
218
three dimensions, this effect would only be present after substantially longer cultures (i.e. only
when the culture is fully confluent). Or the explanation could be physical one: The thin, spread-out
cell morphology of monolayer cultures is necessary for the regimes to physically take effect on the
hMSCs. A detailed exploration of the mechanism of these stimuli is necessary to answer this
question.
It was hypothesised at the beginning of this chapter that capacitive electrical stimulation might be
able to act by hyper- or depolarising the cell membranes. If true, capacitive electrical signals could
be used to mimic the differentiation altering effect of chemical depolarisation observed by
Sundelacruz et al [139]. As the 10 and 250 μs regimes had no effect on the hMSCs, the findings of
this study do not provide evidence to support the hypothesis. However, they do not disprove it
either. The highest electric field strength applied in this study was approx. 50 V/m. Between the
two sides of a cell this field strength generates an electrical potential difference of 0.05 mV in case
of monolayer, and 0.5 mV in case of 3D culture. Compared to the cellular resting potential of 70
mV, this electrical potential is minute. Approaching the issue from a different perspective, such an
electric field strength, based on a rough estimate, causes an ionic concentration difference in the
range of 10-4
nM. This value is dwarfed by the ionic concentrations (10-80 mM) used by
Sundelacruz et al [139] to induce changes in hMSC differentiation. In summary, the electric field
strength applied in this study was too weak to be able to significantly influence the membrane
potential of cells. Therefore no conclusions can be drawn regarding the “hyper-, depolarisation”
hypothesis. A hundred times more powerful stimulus is required to be able to explore this question.
In Chapter 1, from the literature it was inferred that a stronger stimulus generates a stronger effect
[27, 153, 158, 159, 174]. The findings of this study support this statement. 1 μs pulses with an
electric field strength of 0.75 V/m reduced cell numbers by 20%. The similarly brief jolts of the
1999 μs regime with 0.46 V/m lowered cell numbers by 7%. This suggests that there is indeed a
correlation between electric field strength and the extent of the effect.
219
4.3.4.1 Comparison with the Literature
Many positive effects have been attributed to electrical stimulation in the literature. Why were none
of them observed in this investigation?
4.3.4.1.1 Electric Field Strength
In this study electric field strengths ranging from 0.5 to 50 V/m were used. Such electric field
strength have been stated in the literature to induce galvanotactic effects in MSCs [137], fibroblasts
[153] and epithelial cells [194]; EGF secretion in epithelial cells [193]; and the proliferation of
fibroblasts [318]. However, all of these studies used direct stimulation and under completely
different experimental conditions, making comparison a very difficult task. One inference can be
made nevertheless: The electric field strength applied in this study is strong enough to cause
galvanotaxis, and therefore is powerful enough to exert physical changes on cells.
Zhuang et al’s work provides a better basis for comparison [186]. Zhuang’s group applied
capacitive stimulation with 2 V/m electric field strength to MC3T3-E1 osteoblastic cells, and found
that the cells increased proliferation and TGF-β1 expression in response [186]. Brighton et al [19]
reported very similar findings, showing increased MC3T3-E1 cell numbers in connection with
capacitively delivered 2 V/m stimulation. Such positive effects were not witnessed in this study.
However, this is very easily explained by the difference in cell type, waveform and stimulation
length between the two investigations. This also suggests that these three factors play an important
role in defining the effect of the stimulatory regimes.
4.3.4.1.2 Frequency
Frequencies both higher and lower than 500 Hz have been shown to have a positive effect on bone
cells. 100 Hz capacitive stimulation has been observed to promote the proliferation and VEGF
expression of osteoblasts [20, 138]. Similar results were found with 3000 Hz [25]. These findings
suggest that the choice of frequency in this study (500 Hz) was not incorrect, and that the reason
why these positive effects were not observed here lies elsewhere.
220
4.3.4.1.3 Pulse Width
The shortest pulse width reported in the literature is 25 μs [20]. As such, there is no data available
(to the author’s knowledge) to which the results gained with the 1 and 10 μs pulses can be suitably
compared. Nonetheless, since the delivery of 25 μs pulses is noted to enhance proliferation,
calcium deposition, alkaline phosphatase activity and VEGF expression [20] and 250 μs signals
have been stated to have similar effects [20, 138], it can be assumed that the range of pulse widths
utilised in this study is beneficial to bone formation. Why no positive effects were witnesses in the
high voltage experiments is, therefore, not due to an incorrect choice of pulse width.
4.3.4.1.4 Summary
In summary, many positive, proliferation and differentiation enhancing, effects have been reported
in the literature with the use of stimulatory regimes similar to those used in this study. Why these
beneficial effects were not present in this study is unlikely to be due to the choice of electric field
strength, frequency or pulse width. Perhaps it is the duration of the stimulation that was inadequate,
as many of the studies reporting improved proliferation and differentiation chose to deliver their
stimulus continuously 24 h/day [138]. Or possibly MSCs are unable to sense electric fields in the
same way as osteoblasts, lacking the necessary cellular mechanisms, making them insusceptible to
the same stimulatory regimes. The answer may equally lie in any of the other experimental
parameters or perhaps it is the exact combination of these is what is important. Further
investigations are necessary before these questions can be answered satisfactorily.
4.3.4.2 Future Possibilities
There are many ways in which the effect of capacitive electrical stimulation could be further
explored.
Only a very limited number of possible parameters were examined in this study. Regimes with
higher and lower frequency, different waveforms, pulse width and duration could be tested.
221
Interesting experiments could be carried out comparing the same regime delivered with direct and
capacitive stimulation, or capacitive stimulation parallel and perpendicular to a cellular monolayer.
The high voltage experiments have only been carried out using hMCSs from one donor. These
experiments could be repeated, and any further investigation be carried out, with the use of cells
from more than one donor, in order to ensure that the electrical treatments are to the benefit of all
patients.
Experiments could be carried out in order to identify the cellular mechanisms that take part in
sensing the electrical stimulus. For example, the response of intracellular calcium levels [144, 151]
to the various regimes could be investigated through the use of calcium indicating fluorescent dyes
(e.g. fluo-3 dye). The source of the calcium ions could be determined by selectively blocking the
endo/sarcoplasmic reticulum and the calcium channels [20, 138]. Similar experiments could be
carried out with sodium and potassium ions.
The members of the mechanotransduction pathway, for example ERK, p38 and the MAP kinases
could also be selectively inhibited [20, 137, 138, 147, 413] to investigate their role in sensing
electrical stimulation. Alternatively, superarray technology could be employed to determine how
this intracellular pathway is affected by the regimes.
None of the regimes was observed to have an effect on the production of BMP-2, BMP-7 and
ALPL. This, however, does not mean that other osteogenic markers were not affected. It could still
very well be that the short pulses of the 1 and 1999 μs regimes promote the expression of some
alternative markers of bone differentiation, similarly to what was found with direct stimulation.
Whether these regimes promote osteogenesis could be further tested by measuring their effect on
the expression of RUNX2, osteocalcin, osteopontin, osteonectin and collagen type I.
A more extensive exploration of how electrical stimulation might influence cellular behaviour
could be carried out using microarray assays. With the use of DNA microarrays the effects of the
regimes could be tested on thousands of genes simultaneously. This would not only yield important
information of the underlying mechanisms of electrical stimulation, but would allow changes in
222
gene expression of interest to tissue engineering to be pin-pointed. Consecutive experiments could
then be tailored to exploit these beneficial changes to maximum effect.
223
4.4 CONCLUSIONS
1, 10, 250 and 1999 μs pulse width regimes have been delivered at low (15 V) and high (150 V)
voltage to monolayer and scaffold cultures of hMSCs. The effect of the regimes has been assayed
on cell numbers, metabolic and alkaline phosphatise activity, gene expression and BMP-2 and
BMP-7 secretion. However, no regime has been identified that can promote the proliferation nor
enhance the osteogenic differentiation of hMSCs. As such, the two remaining objectives of this
study remain unfulfilled.
Nonetheless, many important observations have been made regarding the interaction hMSCs and
electrical stimulation:
Capacitive electrical stimulation is able to influence hMSCs behaviour.
1 and 1999 μs pulses lower cell numbers after seven sessions of stimulation.
These findings support the “jolt” hypothesis, suggesting that hMSCs are sensitive to rapid
changes in electric field strength.
The same stimulation had disparate effects in monolayer and scaffold cultures, suggesting
that the mechanotransduction pathway might play an important role in the sensing process.
No evidence was gained in support of the “hyper, de-polarisation” theorem.
The extent of the effect the stimulation might have on a cell culture appears to correlate
with electric field strength.
Electrical current might be necessary to mediate some of the beneficial effects of electrical
stimulation.
Cell type, stimulation duration and waveform appear to be important in defining the
outcome of the stimulation.
Other observations, relevant to tissue engineering, have also been made:
Culture density during expansion effects hMSCs behaviour.
224
Low cell density expansion maintains the self-renewing capability and osteogenic potential
of hMSCs.
This is potentially due to low density culture being able to better maintaining the RS sub-
phenotype. The data gained in this study is, however, insufficient to allow conclusions to
be drawn in this regard.
Traditional culture cell density appears to be beneficial for the adipogenic differentiation of
hMSCs.
Cell concentration plays an important role in defining the accuracy of cell seeding. Higher
concentrations (e.g. 500,000 cells/ml) appear to be more accurate.
225
Overall
Conclusions
226
5.1 THE CONCLUSIONS OF THIS THESIS
In this final, brief chapter, the accomplishments of this study are examined, taking into
consideration the impact that this work has on the discipline of tissue engineering.
The main aim of this study was: “To develop electrical stimulation into an effective tool for the
engineering of tissue, with enhanced bone formation from bone-marrow derived Mesenchymal
Stem Cells being in the main focus.”
An autoclavable, versatile, reliable and reusable bioreactor system for the stimulation of both
monolayer and 3D scaffold cultures has been designed and built. It is the first of its kind, to the
author’s knowledge, to incorporate so many advantages, into one singular device. With this
bioreactor a platform is now available to efficiently and reliably test the effects of capacitive
electrical stimulation. As the design is simple and easy-to-use other research groups will be able to
adapt it for their own scientific investigations with ease. Furthermore, with the many computerised
simulations carried out during this project, the physical aspects of the cell-electric field interaction
is now understood with a clarity not available before. The bioreactor and the computerised
simulations combined provide a truly powerful tool, enabling tissue engineers to investigate the
effects of electrical stimulation on any cell and tissue type with great efficiency.
No bone formation promoting regime has yet been identified. However this study was just the first
step in exploring the effects of capacitive electrical stimulation. Further investigations, examining
additional parameter combinations in greater depth are necessary. Nonetheless, a great wealth of
experience has been gained on delivering electrical stimulation. The importance of cell shape, the
various parameters and experimental conditions is now better understood. This knowledge will
help researchers carry out further investigations, leading to the identification of electrical
stimulatory regimes that are of great benefit to tissue engineering.
227
In conclusion, the main achievements of this study are:
1. A bioreactor system that allows the efficient delivery of capacitive electrical stimulation is
now available.
2. The physical characteristic of how electric field stimulation interacts with cells is better
understood.
3. Further proof is available showing that capacitive electrical stimulation can influence
human mesenchymal stem cell behaviour.
4. Important experience has been gained on the principles governing the effects of electrical
stimulation.
228
5.2 FUTURE WORK
The work detailed in this thesis would be continued in the following way:
1. hMSCs would be expanded from a fifth donor using the low density method. Subsequent
experiments would be carried out using cells from Donor 4 and 5 in order to allow the
assessment of the donor dependence of the effects of electrical stimulation.
2. The most important next step would be to identify parameter combinations that promote
hMSC proliferation and/or osteogenic differentiation. In order to achieve this monolayer
experiments would be carried out delivering a set of new regimes at 150 V electrode
potential difference. The new parameter combinations would be based on those already
proven to be effective in the literature. For example, Hartig et al has described a capacitive
regime based on 16 Hz sawtooth signals that is able to enhance human osteoblast cell
numbers and extracellular matrix synthesis [159]. The same regime was also successfully
used to promote hMSC proliferation and gene expression [395]. The effect of these
regimes would be assayed on cell numbers using the PicoGreen technique, and gene
expression of osteogenic markers using qRT-PCR. A much wider range of bone related
genes (RUNX2, osteocalcin, alkaline phosphatase, collagen type I, osteopontin,
osteonectin) would be assayed compared to this thesis in order to allow the osteogenic
effect of the stimuli to be assessed with much greater accuracy. A smaller sample number
(n=3) would be used in initial experiments to allow the cost- and time-effective testing of
as many parameters combinations as possible. The regimes deemed useful for tissue
engineering purposes would be re-assayed at n=6.
3. This would be followed by the testing of the optimal parameter combinations in
Spongostan scaffold cultures. Samples (n=6) would be assayed after seven days of
stimulation for cell numbers and osteogenic marker expression similarly as in the
monolayer experiments. If the regimes are found to exert their beneficial effect in the 3D
setting as well, they would be delivered in longer experiments. Samples would be
229
stimulated in 3D cultures using the optimal regimes up to 14 or 21 days. Cell numbers
would be assayed using the PicoGreen technique, while the secretion of osteogenic marker
proteins would be measured using the DuoSet ELISA kit technology (R&D Systems Inc.).
4. The involvement of intracellular calcium and the mechanotransduction pathway in the
sensing of the electrical signals would also be investigated. Experiments would be carried
out while blocking calcium channels using verapamil and nifedipine [20, 186], the
intracellular stores by TMB-8 [196], calmodulin by W-7 [186, 196], Erk with PD98059,
and the MAP kinase p38 with SB203580 [20, 138]. Cell numbers and gene expression
would be measured (n=6).
5. The underlying mechanisms of electrical stimulation could be further assayed using
microarray technology either through a commercial provider (Almac Diagnostics Ltd.) or
the in-house service offered by the University of Manchester. Stimulated samples would be
compared to sham treated ones. As this assay is highly expensive, only a very limited
number of samples can be tested. These should be chosen carefully.
6. The various regimes could be further tested using additional cell types. Experiments could
be carried out using both osteoblasts and hMSCs in order to elucidate whether the
differences between the observations of this study and those published in the literature are
due to the disparity in cell types.
MSCs allow electrical stimulation’s effects to be explored on bone, adipose and cartilage
formation. In order examine further lineages, especially those originating from the endo- or
ectoderm, an alternative cell type is required. Mouse embryonic stem cells could provide a
relatively easy-to-use and inexpensive cellular model with pluripotent capabilities for such
investigations.
230
References
231
REFERENCES
1. National Ambulatory Medical Care Survey & American Academy of Orthopaedics.
Fractures/Broken Bones Data Sheet., Last retrieved on: 09.12.2013 -
http://www.schwebel.com/userfiles/files/Fractures%281024%29.pdf
2. Hernlund E, Svedbom A, Ivergård M, Compston J, Cooper C, Stenmark J, McCloskey EV,
Jönsson B, Kanis JA. Osteoporosis in the European Union: medical management, epidemiology
and economic burden: A report prepared in collaboration with the International Osteoporosis
Foundation (IOF) and the European Federation of Pharmaceutical Industry Associations (EFPIA).
Arch. Osteoporos. 8, 1, 2013
3. Ivergård M, Svedbom A, Hernlund E, Compston J, Cooper C, Stenmark J, McCloskey EV,
Jönsson B, Kanis JA. Epidemiology and economic burden of osteoporosis in UK. Arch.
Osteoporos. , 8, 211, 2013
4. Cancedda R, Dozin B, Giannoni P, Quarto R. Tissue engineering and cell therapy of cartilage
and bone. Matrix Biology 22, 81, 2003
5. Court-Brown CM and Caesar B. Epidemiology of adult fractures: A review. Injury, 37, 691,
2006
6. Amini AR, Laurencin CT, Nukavarapu SP. Bone tissue engineering: Recent advances and
challenges. Crit. Rev. Biomed. Eng., 40, 363, 2012
7. Van Staa TP, Dennison EM, Leufkens HGM, Cooper C. Epidemiology of fractures in England
and Wales. Bone, 29, 517, 2001
8. World Health Organisation Website on Aging - http://www.who.int/ageing/about/facts/en/ - Last
retrieved on: 08/07/2014
9. Drosse I, Volkmer E, Capanna R, De Biase P, Mutschler W, Schieker M. Tissue engineering for
bone defect healing: An update on a multi-component approach. Injury, Int. J. Care. Injured. 39S2,
S9, 2008
10. Zimmermann G, Wagner C, Schmeckenbecher K, Wentzensen A, Moghaddam A. Treatment of
tibial shaft non-unions: Bone morphogenetic proteins versus autologous bone graft. Injury, Int. J.
Care. Injured. 40, S50, 2009
11. Schmidmaier G, Capanna R, Wildemann B, Beque T, Lowenberg D. Bone morphogenetic
proteins in critical-size bone defects: what are the options? Suppl 3, S39, 2009
12. Schroeder JE and Mosheiff R. Tissue engineering approaches for bone repair: Concepts and
evidence. Injury, 42, 609, 2011
13. Soucacos PN, Johnson EO, Babis G. An update on recent advances in bone regeneration.
Injury, 39, S1, 2008
14. Ikada Y. Challenges in tissue engineering. J. R. Soc. Interface, 3, 589, 2006
15. Regenerative Medicine Glossary. Regen. Med., 4, S1, 2009
16. Rouwkema J, Rivron NC, van Blitterswijk CA. Vascularization in tissue engineering, Trends in
Biotechnology. 26, 434, 2008
17. Lovett M, Lee K, Edwards A, Kaplan DL. Vascularization strategies for tissue engineering.
Tissue Engineering: Part B, 15, 353, 2009
232
18. Auger FA, Laure Gibot L, Lacroix D. The pivotal role of vascularization in tissue engineering.
Annu. Rev. Biomed. Eng. 15, 177, 2013
19. Brighton CT, Wang W, Seldes R, Zhang G, Pollack SR. Signal transduction in electrically
stimulated bone cells. The Journal Of Bone & Joint Surgery, 83, 1514, 2001
20. Kim IS, Song JK, Song YM, Cho TH, Lee TH, Lim SS, Kim SJ, Hwang SJ. Novel effect of
biphasic electric current on in vitro osteogenesis and cytokine production in human mesenchymal
stromal cells. Tissue Engineering: Part A, 15, 1, 2009
21. Thamsborg G, Florescu A, Oturai P, Fallentin E, Tritsaris K, Dissing S. Treatment of knee
osteoarthritis with pulsed electromagnetic fields: a randomized, double-blind, placebo-controlled
study. Osteoarthritis Cartilage. 13, 575, 2005
22. Simonis RB, Parnell EJ, Raya PS, Peacock JL. Electrical treatment of tibial non-union: A
prospective, randomised, double-blind trial. Injury, 34, 357, 2003
23. Griffin XL, Warner F, Costa M. The role of electromagnetic stimulation in the management of
established non-union of long bone fractures: what is the evidence? Injury, Int. J. Care Injured, 39,
419, 2009
24. Ercan B and Webster TJ. The effect of biphasic electrical stimulation on osteoblast function at
anodized nanotubular titanium surfaces. Biomaterials, 31, 3684, 2010
25. Kim IS, Song JK., Zhang YL, Lee TH, Cho TH, Song YM, Kim DK, Kim SJ, Hwang SJ.
Biphasic electric current stimulates proliferation and induces VEGF production in osteoblasts.
Biochimica Et Biophysica. Acta. 1763, 907, 2006
26. Bodamyali R, Bhatt B, Hughes FJ, Winrow VR, Kanczler JM, Simon B, Abbott J, Blake DR,
Stevens CR. Pulsed electromagnetic fields simultaneously induce osteogenesis and upregulate
transcription of Bone Morphogenetic Proteins 2 And 4 in rat osteoblasts in vitro. Biochemical and
Biophysical Research Communications, 250, 458, 1998
27. Vander Molen MA, Donahue HJ, Rubin CT, Mcleod KJ. Osteoblastic networks with deficient
coupling: differential effects of magnetic and electric field exposure. Bone, 27, 227, 2000
28. Lohmann CH, Schwartz Z, Liu Y, Li Z, Simon BJ, Sylvia VL, Dean DD, Bonewald LF,
Donahue HJ, Boyan BD. Pulsed electromagnetic fields affect phenotype and connexin 43 protein
expression in MLO-Y4 osteocyte-like cells and ROS 17/2.8 osteoblast-like cells, Journal Of
Orthopaedic Research, 21, 326, 2003
29. Mason C and Dunhill P. A brief definition of regenerative medicine. Regen. Med., 3, 1, 2008
30. Viola J, Lal B, Grad O. The emergence of tissue engineering as a research field. Arlington, VA:
National Science Foundation, 2003
31. British Standards Institution: Cell therapy and regenerative medicine glossary. BSI Standards
Limited, 2012
32. Yannas IV. Tissue and organ regeneration in adults. Springer Publishings. 2001
33. Lysaght MJ and Crager J. Origins. Tissue Engineering: Part A, 15, 1449, 2009
34. Bell E, Ehrlich HP, Buttle DJ, Nakatsuji T. Living tissue formed in vitro and accepted as skin-
equivalent tissue of full thickness. Science. 6, 1052, 1981
35. US National Library of Medicine PubMed website - http://www.ncbi.nlm.nih.gov/pubmed/ -
Last retrieved on: 12/02/2014
233
36. Martin I, Baldomero H, Bocelli-Tyndall C, Emmert MY, Hoerstrup SP, Ireland H, Passweg J,
Tyndall A. The survey on cellular and engineered tissue therapies in Europe in 2011. Tissue
Engineering: Part A, 20, 1, 2013
37. Langer R and Vacanti JP. Tissue Engineering. Science, 260, 920, 1993
38. Fan JP, Kalia P, Di Silvio L, Huang J. In vitro response of human osteoblasts to multi-step sol–
gel derived bioactive glass nanoparticles for bone tissue engineering. Materials Science and
Engineering: C, 36, 206, 2014
39. Kreuz PC, Gentili C, Samans B, Martinelli D, Krüger JP, Mittelmeier W, Endres M, Cancedda
R, Kaps C. Scaffold-assisted cartilage tissue engineering using infant chondrocytes from human hip
cartilage. Osteoarthritis and Cartilage, 21, 1997, 2013
40. Dong L, and Wang C. Harnessing the power of macrophages/monocytes for enhanced bone
tissue engineering. Trends in Biotechnology, 31, 342, 2013
41. Correia C, Bhumiratana S, Yan LP, Oliveira AL, Gimble JM, Rockwood D, Kaplan DL, Sousa
RA, Reis RL, Vunjak-Novakovic G. Development of silk-based scaffolds for tissue engineering of
bone from human adipose-derived stem cells. Acta Biomaterialia, 8, 2483, 2012
42. Zhao Z, Yu H, Xiao F, Wang X, Yang S, Li S. Differentiation of adipose-derived stem cells
promotes regeneration of smooth muscle for ureteral tissue engineering. Journal of Surgical
Research, 178, 55, 2012
43. Park BW, Kang EJ, Byun JH, Son MG, Kim HJ, Hah YS, Kim TH, Kumar BM, Ock SA, Rho
GJ. In vitro and in vivo osteogenesis of human mesenchymal stem cells derived from skin, bone
marrow and dental follicle tissues. Differentiation, 83, 249, 2012
44. Tian H, Bharadwaj S, Liu Y, Ma H, Ma PX, Atala A, Zhang Y. Myogenic differentiation of
human bone marrow mesenchymal stem cells on a 3D nano fibrous scaffold for bladder tissue
engineering. Biomaterials, 31, 870, 2010
45. Tang M, Chen W, Weir MD, Thein-Han W, Xu HHK. Human embryonic stem cell
encapsulation in alginate microbeads in macroporous calcium phosphate cement for bone tissue
engineering. Acta Biomaterialia, 8, 3436, 2012
46. Wang A, Tang Z, Park IH, Zhu Y, Patel S, Daley GQ, Li S. Induced pluripotent stem cells for
neural tissue engineering. Biomaterials, 32, 5023, 2011
47. Xu Y, Dong S, Zhou Q, Mo X, Song L, Hou T, Wu J, Li S, Li Y, Li P, Gan Y, Xu J. The effect
of mechanical stimulation on the maturation of TDSCs-poly(L-lactide-co-e-caprolactone)/collagen
scaffold constructs for tendon tissue engineering. Biomaterials, 35, 2760, 2014
48. Odedra D, Chiu LLY, Shoichet M, Radisic M. Endothelial cells guided by immobilized
gradients of vascular endothelial growth factor on porous collagen scaffolds. Acta Biomaterialia, 7,
3027, 2011
49. Gaetani R, Doevendans PA, Metz CHG, Alblas J, Messina E, Giacomello A, Sluijter JPG.
Cardiac tissue engineering using tissue printing technology and human cardiac progenitor cells.
Biomaterials, 33, 1782, 2012
50. Kim BS, Nikolovski J, Bonadio J, Mooney DJ. Cyclic mechanical strain regulates the
development of engineered smooth muscle tissue. Nat Biotechnol., 17, 979, 1999
51. Li YC, Tsai LK, Wang JH, Young TH. A neural stem/precursor cell monolayer for neural
tissue engineering. Biomaterials, 35, 1192, 2014
234
52. Yamamoto Y, Ito A, Kato M, Kawabe Y, Shimizu K, Fujita H, Nagamori E, Kamihira M.
Preparation of artificial skeletal muscle tissues by a magnetic force-based tissue engineering
technique. Journal of Bioscience and Bioengineering, 108, 538, 2009
53. Ladd MR, Lee SJ, Stitzel JD, Atala A, Yoo JJ. Co-electrospun dual scaffolding system with
potential for muscle–tendon junction tissue engineering. Biomaterials, 32, 1549, 2011
54. Baiguera S, Del Gaudio C, Lucatelli E, Kuevda E, Boieri M, Mazzanti B, Bianco A,
Macchiarini P. Electrospun gelatin scaffolds incorporating rat decellularized brain extracellular
matrix for neural tissue engineering. Biomaterials, 35, 1205, 2014
55. Mauney JR, Nguyen T, Gillen K, Kirker-Head C, Gimble JM, Kaplan DL. Engineering
adipose-like tissue in vitro and in vivo utilizing human bone marrow and adipose-derived
mesenchymal stem cells with silk fibroin 3D scaffolds. Biomaterials, 28, 5280, 2007
56. Zakhem E, Raghavan S, Gilmont RR, Bitar KN. Chitosan-based scaffolds for the support of
smooth muscle constructs in intestinal tissue engineering. Biomaterials, 33, 4810, 2012
57. Arafat MT, Lam CXF, Ekaputra AK, Wong SY, Li X, Gibson I. Biomimetic composite coating
on rapid prototyped scaffolds for bone tissue engineering. Acta Biomaterialia, 7, 809, 2011
58. Jeong SI, Kwon JH, Lim JI, Cho SW, Jung Y, Sung WJ, Kim SH, Kim YH, Lee YM, Kim BS,
Choi CY, Kim SJ. Mechano-active tissue engineering of vascular smooth muscle using pulsatile
perfusion bioreactors and elastic PLCL scaffolds. Biomaterials, 26, 1405, 2005
59. Chen H, Huang J, Yu J, Liu S, Gu P. Electrospun chitosan-graft-poly (ɛ-caprolactone)/poly (ɛ-
caprolactone) cationic nanofibrous mats as potential scaffolds for skin tissue engineering.
International Journal of Biological Macromolecules, 48, 13, 2011
60. Fisher MB, Henning EA, Söegaard NB, Dodge GR, Steinberg DR, Mauck RL. Maximizing
cartilage formation and integration via a trajectory-based tissue engineering approach.
Biomaterials, 35, 2140, 2014
61. Shasti M, Jacquet R, McClellan P, Yang J, Matsushima S, Isogai N, Murthy A, Landis WJ.
Effects of FGF-2 and OP-1 in vitro on donor source cartilage for auricular reconstruction tissue
engineering. International Journal of Pediatric Otorhinolaryngology, 78, 416, 2014
62. Nikkhah M, Edalat F, Manoucheri S, Khademhosseini A. Engineering microscale topographies
to control the cell-substrate interface. Biomaterials, 33, 5230, 2012
63. White TD, Black MT, Folkens PA. Human Osteology. Burlington: Elsevier Science, 3rd ed.,
2011
64. Bronner F, Farach-Carson MC, Rubin J (eds.). Bone resorption. Dordrecht: Springer, 2006 -
Bain SD and Gross TS. Chapter 4.: Structural aspects of bone resorption, 58, 2006
65. Rupani A, Balint R, Cartmell SH. Osteoblasts and their applications in bone tissue engineering.
Cell Health and Cytoskeleton, 2012, 49, 2012
66. Sela J and Bab I (eds.). Principles of bone regeneration. Germany: Springer Verlag, 2012
67. Bronner F, Farach-Carson MC, Rubin J (eds.). Bone resorption. Dordrecht: Springer, 2006 -
Oursler MJ. Chapter 2.: Osteolytic enzymes of osteoclasts, 24, 2006
68. Anselme K. Osteoblast adhesion on biomaterials. Biomaterials, 21, 667, 2000
69. Su X, Sun K, Cui FZ, Landis WJ. Organization of apatite crystals in human woven bone. Bone,
32, 150, 2003
235
70. Reznikov N, Shahar R, Weiner S. Three-dimensional structure of human lamellar bone: The
presence of two different materials and new insights into the hierarchical organization. Bone, 59,
93, 2014
71. Lieberman J and Friedlaender GE (eds.). Bone regeneration and repair biology and clinical
applications. Dordrecht: Springer, 2007
72. Seibel MJ, Robins SP, Bilezikian JP (eds.). Dynamics of bone and cartilage metabolism. 2nd
ed., Elsevier Inc., 2006 - Lian JB and Stein GS. Chapter 14: The cells of bone. 221, 2006
73. Sethe S, Scutt A, Stolzing A. Aging of mesenchymal stem cells. Ageing Research Reviews, 5,
91, 2006
74. Augello A and De Bari C. The regulation of differentiation in mesenchymal stem cells. Human
Gene Therapy, 21, 1226, 2010
75. Augello A, Kurth TB, De Bari C. Mesenchymal stem cells: A perspective from in vitro cultures
to in vivo migration and niches. European Cells and Materials, 20, 121, 2010
76. Pountos I and Giannoudis PV. Biology of mesenchymal stem cells. Injury, Int. J. Care Injured,
36S, S8, 2005
77. Wagers AJ and Weissman IL. Plasticity of adult stem cells. Cell, 116, 639, 2004
78. Schneider RK, Puellen A, Kramann R, Raupach K, Bornemann J, Knuechel R, Perez-Bouza A,
Neuss S. The osteogenic differentiation of adult bone marrow and perinatal umbilical mesenchymal
stem cells and matrix remodelling in three-dimensional collagen scaffolds. Biomaterials, 31, 467,
2010
79. Hong SH, Gang EJ, Jeong JA, Ahn C, Hwang SH, Yang IH, Park HK, Han H, Kim H. In vitro
differentiation of human umbilical cord blood-derived mesenchymal stem cells into hepatocyte-like
cells. Biochemical and Biophysical Research Communications, 330, 1153, 2005
80. Udehiya RK, Amarpal, Aithal HP, Kinjavdekar P, Pawde AM, Singh R, Sharma G.
Comparison of autogenic and allogenic bone marrow derived mesenchymal stem cells for repair of
segmental bone defects in rabbits. Research in Veterinary Science, 94, 743, 2013
81. Heo JS, Choi SM, Kim HO, Kim EH, You J, Park T, Kim E, Kim HS. Neural
transdifferentiation of human bone marrow mesenchymal stem cells on hydrophobic polymer-
modified surface and therapeutic effects in an animal model of ischemic stroke. Neuroscience, 238,
305, 2013
82. Hu YL, Fu YH, Tabata Y, Gao JQ. Mesenchymal stem cells: A promising targeted-delivery
vehicle in cancer gene therapy. Journal of Controlled Release, 147, 154, 2010
83. Friedenstein AJ, Deriglasova UF, Kulagina NN, Panasuk AF, Rudakowa SF, Luriá EA,
Ruadkow IA. Precursors for fibroblasts in different populations of hematopoietic cells as detected
by the in vitro colony assay method. Experimental hematology, 2, 83, 1974
84. Krampera M, Marconi S, Pasini A, Galiè M, Rigotti G, Mosna F, Tinelli M, Lovato L,
Anghileri E, Andreini A, Pizzolo G, Sbarbati A, Bonetti B. Induction of neural-like differentiation
in human mesenchymal stem cells derived from bone marrow, fat, spleen and thymus. Bone, 40,
382, 2007
85. Lin TM, Chang HW, Wang KH, Kao AP, Chang CC, Wen CH, Lai CS, Lin SD. Isolation and
identification of mesenchymal stem cells from human lipoma tissue. Biochemical and Biophysical
Research Communications, 361, 883, 2007
236
86. Lee JM, Jung J, Lee HJ, Jeong SJ, Cho KJ, Hwang SG, Kim GJ. Comparison of
immunomodulatory effects of placenta mesenchymal stem cells with bone marrow and adipose
mesenchymal stem cells. International Immunopharmacology, 13, 219, 2012
87. Valenti MT, Carbonare LD, Donatelli L, Bertoldo F, Zanatta M, Cascio VL. Gene expression
analysis in osteoblastic differentiation from peripheral blood mesenchymal stem cells. Bone, 43,
1084, 2008
88. Patki S, Kadam S, Chandra V, Bhonde R. Human breast milk is a rich source of multipotent
mesenchymal stem cells. Human Cell, 23, 35, 2010
89. da Silva Meirelles L, Fontes AM, Covas DT, Caplan AI. Mechanisms involved in the
therapeutic properties of mesenchymal stem cells. Cytokine & Growth Factor Reviews, 20, 419,
2009
90. Croft AP and Przyborski SA. Mesenchymal stem cells from the bone marrow stroma: basic
biology and potential for cell therapy. Current Anaesthesia & Critical Care, 15, 410, 2004
91. Odabas S, Sayar F, Guven G, Yanıkkaya-Demirel G, Piskin E. Separation of mesenchymal
stem cells with magnetic nanosorbents carrying CD105 and CD73 antibodies in flow-through and
batch systems. Journal of Chromatography B, 861, 74, 2008
92. Dominici M, Le Blanc K, Mueller I, Slaper-Cortenbach I, Marini FC, Krause DS, Deans RJ,
Keating A, Prockop DJ, Horwitz EM. Minimal criteria for defining multipotent mesenchymal
stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy, 8,
315, 2006
93. Pittenger M, Vanguri P, Simonetti D, Young R. Adult mesenchymal stem cells: potential for
muscle and tendon regeneration and use in gene therapy. J. Musculoskelet. Neuronal Interact., 2,
309, 2002
94. Orlic D, Kajstura J, Chimenti S, Jakoniuk I, Anderson SM, Li B, Pickel J, McKay R, Nadal-
Ginard B, Bodine DM, Leri A, Anversa P. Bone marrow cells regenerate infarcted myocardium.
Nature, 410, 701, 2001
95. De Bari C, Dell’Accio F, Tylzanowski P, Luyten FP. Multipotent mesenchymal stem cells from
adult human synovial membrane. Arthritis Rheum., 44, 1928, 2001
96. Dicker A, Le Blanc K, Astrom G, van Harmelen V, Götherström C, Blomqvist L, Arner P,
Rydén M. Functional studies of mesenchymal stem cells derived from adult human adipose tissue.
Exp. Cell. Res., 308, 283, 2005
97. Long X, Olszewski M, Huang W, Kletzel M. Neural cell differentiation in vitro from adult
human bone marrow mesenchymal stem cells. Stem Cells Dev., 14, 65, 2005
98. Bronner F and Farach-Carson MC (eds.). Bone formation. Springer-Verlag London Limited,
2004
99. Noble BS. The osteocyte lineage. Archives of Biochemistry and Biophysics, 473, 106, 2008
100. Bronner F, Farach-Carson MC, Rubin J (eds.). Bone resorption. Dordrecht: Springer, 2006 -
Rubin J and Greenfield EM. Chapter 1.: Osteoclast: Origin and differentiation. 1, 2006
101. Nakase T, Sugimoto M, Sato M, Kaneko M, Tomita T, Sugamoto K, Nomura S, Kitamura Y,
Yoshikawa H, Yasui N, Yonenobu K, Ochi T. Switch of osteonectin and osteopontin mRNA
expression in the process of cartilage-to-bone transition during fracture repair. Acta Histochemica,
100, 287, 1998
237
102. Sakou T. Bone Morphogenetic Proteins: From basic studies to clinical approaches. Bone, 22,
591, 1998
103. Onishi T, Ishidou Y, Nagamine T, Yone K, Imamura T, Kato M, Sampath TK, Ten Dijke P,
Sakou T. Distinct and overlapping patterns of localization of Bone Morphogenetic Protein (BMP)
family members and a BMP Type II receptor during fracture healing in rats. Bone. 22, 605, 1998
104. Spector JA, Luchs JS, Mehrara BJ, Greenwald JA, Smith LP, Longaker MT. Expression of
bone morphogenetic proteins during membranous bone healing. Plast Reconstr Surg., 107, 124,
2001
105. Bostrom MPG, Lane JM, Berberian WS, Missri AAE, Tomin E, Weiland A, Doty SB, Glaser
D, Rosen VM. Immunolocalization and expression of Bone Morphogenetic Proteins 2 and 4 in
fracture healing. Journal of Orthopaedic Research, 13, 357, 1995
106. Bilezikian JP, Raisz LG, Martin TJ (eds.). Principles of bone biology. 3rd
ed., Elsevier Inc.,
2008 - Aubin JE. Chapter 4: Mesenchymal stem cells and osteoblast differentiation. 85, 2008
107. Komori T and Kishimoto T. Cbfa1 in bone development. Current Opinion in Genetics &
Development, 8, 494, 1998
108. Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K, Shimizu Y, Bronson RT,
Gao YH, Inada M, Sato M, Okamoto R, Kitamura Y, Yoshiki S, Kishimoto T. Targeted disruption
of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts.
Cell, 89, 755, 1997
109. Karsenty G. Role of Cbfa1 in osteoblast differentiation and function. Cell & Developmental
Biology, 11, 343, 2000
110. Nakashima K, Zhou X, Kunkel G, Zhang Z, Deng JM, Behringer RR, de Crombrugghe B. The
novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation
and bone formation. Cell, 108, 17, 2002
111. Fisher LW, McBride OW, Termine JD, Young MF. Human bone sialoprotein. Journal of
biological chemistry, 265, 2347, 1990
112. Hunter GK and Goldberg HA. Modulation of crystal formation by bone phosphoproteins: role
of glutamic acid-rich sequences in the nucleation of hydroxyapatite by bone sialoprotein. Biochem.
J., 302, 175, 1994
113. Termine JD, Kleinman HK, Whitson SW, Conn KM, McGarvey ML, Martin GR.
Osteonectin, a bone-specific protein linking mineral to collagen. Cell, 26, 99, 1981
114. Delany AM, Amling M, Priemel M, Howe C, Baron R, Canalis E. Osteopenia and decreased
bone formation in osteonectin-deficient mice. J. Clin. Invest., 105, 915, 2000
115. Nakase T, Takaoka K, Hirakawa K, Hirota S, Takemura T, Onoue H, Takebayashi K,
Kitamura Y, Nomura S. Alterations in the expression of osteonectin, osteopontin and osteocalcin
mRNAs during the development of skeletal tissues in vivo. Bone and Mineral, 26, 109, 1994
116. Nakamura A, Dohi Y, Akahane M, Ohgushi H, Nakajima H, Funaoka H, Takakura Y.
Osteocalcin secretion as an early marker of in vitro osteogenic differentiation of rat mesenchymal
stem cells. Tissue Eng Part C Methods, 15, 169, 2009
117. Ogata Y. Bone sialoprotein and its transcriptional regulatory mechanism. J. Periodont. Res.,
43, 127, 2008
238
118. Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, Smith E, Bonadio J, Goldstein S,
Gundberg C, Bradley A, Karsenty G. Increased bone formation in osteocalcin-deficient mice.
Nature. 382, 448, 1996
119. Sapir-Koren R and Livshits G. Bone mineralization and regulation of phosphate homeostasis.
IBMS BoneKEy, 8, 286, 2011
120. Hessle L, Johnson KA, Anderson HC, Narisawa S, Sali A, Goding JW, Terkeltaub F, Millan
JL. Tissue-nonspecific alkaline phosphatase and plasma cell membrane glycoprotein-1 are central
antagonistic regulators of bone mineralization. PNAS, 99, 9445, 2002
121. Golub EE and Boesze-Battaglia K. The role of alkaline phosphatase in mineralization. Curr.
Opin. Orthop.. 18, 444, 2007
122. Vater C, Kasten P, Stiehler M. Culture media for the differentiation of mesenchymal stromal
cells. Acta Biomaterialia, 7, 463, 2011
123. Perinpanayagam H, Martin T, Mithal V, Dahman M, Marzec N, Lampasso J, Dziak R.
Alveolar bone osteoblast differentiation and Runx2/Cbfa1 expression. Archives of Oral Biology,
51, 406, 2006
124. Salim A, Nacamuli RP, Morgan EF, Giaccia AJ, Longaker MT. Transient changes in oxygen
tension inhibit osteogenic differentiation and runx2 expression in osteoblasts. J. Biol. Chem., 279,
40007, 2004
125. Jaiswal N, Haynesworth SE, Caplan AI, Bruder SP. Osteogenic differentiation of purified,
culture-expanded human mesenchymal stem cells in vitro. Journal of Cellular Biochemistry, 64,
295, 1997
126. Frank O, Heim M, Jakob M, Barbero A, Schafer D, Bendik I, Dick W, Heberer M, Ivan
Martin I. Real-time quantitative rt-pcr analysis of human bone marrow stromal cells during
osteogenic differentiation in vitro. Journal of Cellular Biochemistry, 85, 737, 2002
127. Owen TA, Aronow M, Shalhoub V, Barone LM, Wilming L, Tassinari MS, Kennedy MB,
Pockwinse S, Lian JB, Stein GS. Progressive development of the rat osteoblast phenotype in vitro:
reciprocal relationships in expression of genes associated with osteoblast proliferation and
differentiation during formation of the bone extracellular matrix. Journal Of Cellular Physiology,
143, 420, 1990
128. Lian JB and Stein GS. Development of the osteoblast phenotype: molecular mechanisms
mediating osteoblast growth and differentiation. Iowa Orthop J., 15, 118, 1995
129. Qi H, Aguiar DJ, Williams SM, La Pean A, Pan W, Verfaillie CM. Identification of genes
responsible for osteoblast differentiation from human mesodermal progenitor cells. PNAS, 100,
3305, 2003
130. Ogawa R, Mizuno H, Watanabe A, Migita M, Shimada T, Hyakusoku H. Osteogenic and
chondrogenic differentiation by adipose-derived stem cells harvested from GFP transgenic mice.
Biochemical and Biophysical Research Communications, 313, 871, 2004
131. Boeloni JN, Ocarino NM, Goes AM, Serakides R. Comparative study of osteogenic
differentiation potential of mesenchymal stem cells derived from bone marrow and adipose tissue
of osteoporotic female rats. Connect Tissue Res, Early Online: 1, 2014, In Press
132. Shi G, Zhang Z, Rouabhia M. The regulation of cell functions electrically using biodegradable
polypyrrole–polylactide conductors. Biomaterials, 29, 3792, 2008
239
133. Rowlands AS, Cooper-White JJ. Directing phenotype of vascular smooth muscle cells using
electrically stimulated conducting polymer. Biomaterials, 29, 4510, 2008
134. Sun S, Titushkin I, Cho M. Regulation of mesenchymal stem cell adhesion and orientation in
3D collagen scaffold by electrical stimulus. Bioelectrochemistry, 69, 133, 2006
135. Wang E, Zhao M, Forrester JV, McCaig CD. Re-orientation and faster, directed migration of
lens epithelial cells in a physiological electric field. Exp. Eye Res., 71, 91, 2000
136. Wang E, Zhao M, Forrester JV, McCaig CD. Bi-directional migration of lens epithelial cells
in a physiological electrical field. Exp. Eye Res., 76, 29, 2003
137. Hammerick KE, Longaker MT, Prinz FB. In vitro effects of direct current electric fields on
adipose-derived stromal cells. Biochem. Biophys. Res. Commun., 397, 12, 2010
138. Kim IS, Song JK, Song YM, Lee TH, Cho TH, Lim SS, Pan H, Kim SJ, Hwang SJ. Novel
action of biphasic electric current in vitro osteogenesis of human bone marrow mesenchymal
stromal cells coupled with VEGF production. Bone, 43, S43, 2008
139. Sundelacruz S, Levin M, Kaplan DL. Membrane potential controls adipogenic and osteogenic
differentiation of mesenchymal stem cells. PLoS One, 3, 1, 2008
140. McCaig CD, Rajnicek AM, Song B, Zhao M. Controlling cell behaviour electrically: current
views and future potential. Physiol. Rev., 85, 943, 2005
141. Huo R, Ma Q, Wu JJ, Chin-Nuke K, Jing Y, Chen J, Miyar ME, Davis SC, Li J. Noninvasive
electromagnetic fields on keratinocyte growth and migration. J. Surg. Res., 162, 299, 2009
142. Li X and Kolega J. Effects of direct current electric fields on cell migration and actin filament
distribution in bovine vascular endothelial cells. J. Vasc. Res., 39, 391, 2002
143. Sauer H, Bekhite MM, Hescheler J, Wartenberg M. Redox control of angiogenic factors and
CD31-positive vessel-like structures in mouse embryonic stem cells after direct current electrical
field stimulation. Exp. Cell Res., 304, 380, 2005
144. Nuccitelli R. A role for endogenous electric fields in wound healing. Curr. Top Dev. Biol., 58,
1, 2003
145. Gelband CH and Hume JR. Ionic currents in single smooth muscle cells of the canine renal
artery. Circulation Research, 71, 745, 1992
146. Ghasemi-Mobarakeh L, Prabhakaran MP, Morshed M, Nasr-Esfahani MH, Baharvand H,
Kiani S, Al-Deyab SS, Ramakrishna S. Application of conductive polymers, scaffolds and
electrical stimulation for nerve tissue engineering. J. Tissue Eng. Regen. Med., 5, e17, 2011
147. Zhao M. Electrical fields in wound healing: An overriding signal that directs cell migration.
Semin. Cell Dev. Biol., 20, 674, 2009
148. Farboud B, Nuccitelli R, Schwab IR, Isseroff RR. DC electric fields induce rapid directional
migration in cultured human corneal epithelial cells. Exp. Eye Res., 70, 667, 2000
149. Mycielska ME and Djamgoz MBA. Cellular mechanisms of direct-current electric field
effects: galvanotaxis and metastatic disease. J. Cell Sci., 117, 1631, 2004
150. Song B, Zhao M, Forrester JV, McCaig CD. Electrical cues regulate the orientation and
frequency of cell division and the rate of wound healing in vivo. Proc. Natl. Acad. Sci. USA., 99,
13577, 2002
240
151. Titushkin I and Cho M. Regulation of cell cytoskeleton and membrane mechanics by electric
field: role of linker proteins. Biophys. J., 96, 717, 2009
152. Lin F, Baldessari F, Gyenge CC, Sato T, Chambers RD, Santiago JG, Butcher EC.
Lymphocyte electrotaxis in vitro and in vivo. J. Immunol., 181, 2465, 2008
153. Erickson CA and Nuccitelli R. Embryonic fibroblast motility and orientation can be
influenced by physiological electric fields. J. Cell Biol., 98, 296, 1984
154. Serena E, Figallo E, Tandon N, Cannizzaro C, Gerecht S, Elvassore N, Vunjak-Novakovic G.
Electrical stimulation of human embryonic stem cells: Cardiac differentiation and the generation of
reactive oxygen species. Exp. Cell Res., 315, 3611, 2009
155. Reid B, Song B, Zhao M. Electric currents in Xenopus tadpole tail regeneration. Dev. Biol.,
335, 198, 2009
156. Illingworth CM and Barker AT. Measurement of electrical currents emerging during the
regeneration of amputated finger tips in children. Clin. Phys. Physiol. Meas., 1, 87, 1980
157. Petersson CJ, Holmer NG, Johnell O. Electrical stimulation of osteogenesis studies of the
cathode effect on rabbit femur. Acta orthop. scand., 53, 727, 1982
158. Wiesmann HP, Hartig M, Stratmann U, Meyer U, Joos U. Electrical stimulation influences
mineral formation of osteoblast-like cells in vitro. Biochim. Biophys. Acta., 1538, 28, 2001
159. Hartig M, Joos U, Wiesmann HP. Capacitively coupled electric fields accelerate proliferation
of osteoblast-like primary cells and increase bone extracellular matrix formation in vitro. Eur.
Biophys. J., 29, 499, 2000
160. Huang CP, Chen XM, Chen ZQ. Osteocyte: The impresario in the electrical stimulation for
bone fracture healing. Med. Hypotheses, 70, 287, 2008
161. MacGinitie LA, Wu DD, Cochran GVB. Streaming potentials in healing, remodeling, and
intact cortical bone. J. Bone Miner. Res., 8, 1323, 1993
162. McLeod KJ and Rubin CT. The effect of low-frequency electrical fields on osteogenesis. J.
Bone Joint Surg. Am., 74, 920, 1992
163. Xu J, Wang W, Clark CC, Brighton CT. Signal transduction in electrically stimulated articular
chondrocytes involves translocation of extracellular calcium through voltage-gated channels.
Osteoarthritis Cartilage., 17, 397, 2009
164. Zhao M, Bai H, Wang E, Forrester JV, McCaig CD. Electrical stimulation directly induces
pre-angiogenic responses in vascular endothelial cells by signaling through VEGF receptors. J. Cell
Sci., 117, 397, 2004
165. Brighton CT, Black J, Friedenberg ZB, Esterhai JL, Day LJ, Connolly JF. A multicenter study
of the treatment of non-union with constant direct current. J. Bone Joint Surg. Am., 63, 2, 1981
166. Evans RON, Goldberg JA, Bruce WJM, Walsh W. Reoperated clavicular nonunion treated
with osteogenic protein 1 and electrical stimulation. J. Shoulder Elbow Surg., 13, 573, 2004
167. Gardner SE, Franz RA, Schmidt FL. Effect of electrical stimulation on chronic wound
healing: a meta-analysis. Wound Rep. Reg., 7, 495, 1999
168. Shimada Y, Sato K, Kagaya H, Konishi N, Miyamoto S, Matsunaga Y. Clinical use of
percutaneous intramuscular electrodes for functional electrical stimulation. Arch. Phys. Med.
Rehabil., 77, 1014, 1996
241
169. Janardan B. Electrical stimulation in medical sciences from basic to applied. Journal of
Applied Basic Medical Sciences, 3, 49, 2001
170. Deyo RE, Walsh NE, Martin DC, Shoenfeld LS, Ramamurthy S. A controlled trial of
transcutaneous electrical nerve stimulation (TENS) and exercise for chronic back pain. The New
England Journal of Medicine, 322, 1627, 1990
171. Johnson MI, Ashton CH, Thompson JW. An in-depth study of long-term users of
transcutaneous electrical nerve stimulation (TENS). Implications for clinical use of TENS. Pain,
44, 221, 1991
172. Houghton PE, Campbell KE, Fraser CH, Harris C, Keast DH, Potter PJ, Hayes KC, Woodbury
MG. Electrical stimulation therapy increases rate of healing of pressure ulcers in community-
dwelling people with spinal cord injury. Arch. Phys. Med. Rehabil., 91, 669, 2010
173. Mendonça AC, Barbieri CH, Mazzer N. Directly applied low intensity direct electric current
enhances peripheral nerve regeneration in rats. J. Neurosci. Methods, 129, 183, 2003
174. McLeod KJ, Donahue HJ, Levin PE, Fontaine MA, Rubin CT. Electric fields modulate bone
cell function in a density-dependent manner. J. Bone Miner. Res., 1, 977, 1993
175. Bouaziz A, Richert A, Caprani A. Morphological aspects of endothelial cells cultured in the
absence of fetal calf serum under controlled electrical charges applied to the support. Biomaterials,
17, 2281, 1996
176. Chang K, Chang WHS, Huang S, Huang S, Shih C. Pulsed electromagnetic fields stimulation
affects osteoclast formation by modulation of osteoprotegerin, RANK ligand and macrophage
colony-stimulating factor. J. Orthop. Res., 23, 1308, 2005
177. Lam H and Qin YX. The effects of frequency-dependent dynamic muscle stimulation on
inhibition of trabecular bone loss in a disuse model. Bone, 43, 1093, 2008
178. Aaron RK, Wang S, Ciombor D McK. Upregulation of basal TGFβ1 levels by EMF
coincident with chondrogenesis – implications for skeletal repair and tissue engineering. J. Orthop.
Res., 20, 233, 2002
179. Feedar JA, Kloth LC, Gentzkow GD. Chronic dermal ulcer healing enhanced with
monophasic pulsed electrical stimulation. Phys. Ther.,71, 639, 1991
180. Ferroni P, Roselli M, Guadagni F, Martini F, Mariotti S, Marchitelli E, Cipriani C. Biological
effects of a software-controlled voltage pulse generator (PhyBack PBK-2C) on the release of
vascular endothelial growth factor (VEGF). In Vivo, 19, 949, 2005
181. Merrill DR, Bikson M, Jefferys JGR. Electrical stimulation of excitable tissue: design of
efficacious and safe protocols. J. Neurosci. Methods, 141, 171, 2005
182. Spadaro JA and Becker RO. Function of implanted cathodes in electrode-induced bone
growth. Med. & Biol. Eng. & Comput., 17, 769, 1979
183. Balint R, Cassidy NJ, Cartmell SH. Electrical stimulation: a novel tool for tissue engineering.
Tissue Eng. Part B Rev., 19, 48, 2013
184. Au HTH., Cheng I, Chowdhury MF, Radisic M. Interactive effects of surface topography and
pulsatile electrical field stimulation on orientation and elongation of fibroblasts and
cardiomyocytes. Biomaterials, 28, 4277, 2007
185. Antov Y, Barbul A, Mantsur H, Korenstein R. Electroendocytosis: Exposure of cells to pulsed
low electric fields enhances adsorption and uptake of macromolecules. Biophys. J., 88, 2206, 2005
242
186. Zhuang H, Wang W, Seldes RM, Tahernia AD, Fan H, Brighton CT. Electrical stimulation
induces the level of TGF-β1 mRNA in osteoblastic cells by a mechanism involving
calcium/calmodulin pathway. Biochem. Biophys. Res. Commun., 237, 225, 1997
187. Abeed RI, Naseer M, Abel EW. Capacitively coupled electrical stimulation treatment: Results
from patients with failed long bone fracture unions. J. Orthop. Trauma., 12, 510, 1998
188. Zamora-Navas P, Borras Verdera A, Antelo Lorenzo R, Saras Ayuso JR, Peña Reina MC.
Electrical stimulation of bone nonunion with the presence of a gap. Acta. Orthop. Belg., 61, 169,
1995
189. Gargiulo GD and McEwan A (eds.). Applied biomedical engineering. InTech, 2011 -
Meng S, Rouabhia M, Zhang Z. Chapter 3: Electrical stimulation in tissue regeneration. 37, 2011
190. Yonemori, K, Matsunaga S, Ishidou Y, Maeda S, Yoshida H. Early effects of electrical
stimulation on osteogenesis. Bone, 19, 173, 1996
191. Fu Y, Lin C, Chang J, Wang G, Ho M. A novel single pulsed electromagnetic field stimulates
the osteogenesis of human bone marrow derived mesenchymal stem cells. Bone, 55, S322, 2009
192. Cooper MS and Keller RE. Perpendicular orientation and directional migration of amphibian
neural crest cells in DC electrical fields. Proc. Nail. Acad. Sci. USA, 81, 160, 1984
193. Zhao M, Dick A, Forrester JV, McCaig CD. Electric field– directed cell motility involves up-
regulated expression and asymmetric redistribution of the epidermal growth factor receptors and is
enhanced by fibronectin and laminin. Mol. Biol. Cell., 10, 1259, 1999
194. Gamboa OL, Pu J, Townend J, Forrester JV, Zhao M, McCaig C, Lois N. Electrical
stimulation of retinal pigment epithelial cells. Exp. Eye. Res., 91, 195, 2010
195. Franek A, Polak A, Kucharzewski M. Modern application of high voltage stimulation for
enhanced healing of venous crural ulceration. Med. Eng. Phys., 22, 647, 2000
196. Li JKJ, Lin JCA, Liu HC, Sun JS, Ruaan RC, Shih C, Chang WHS. Comparison of ultrasound
and electromagnetic field effects on osteoblast growth. Ultrasound Med. Biol., 32, 769, 2006
197. Radisic M, Park H, Shing H, Consi T, Schoen FJ, Langer R, Freed LE, Vunjak-Novakovic G.
Functional assembly of engineered myocardium by electrical stimulation of cardiac myocytes
cultured on scaffolds. Proc. Natl. Acad. Sci. U S A., 101, 18129, 2004
198. Becker D. Gary DS, Rosenzweig ES, Grill WM, McDonald JW. Functional electrical
stimulation helps replenish progenitor cells in the injured spinal cord of adult rats. Exp. Neurol.,
222, 211, 2010
199. Kojima J, Shinohara H, Ikariyama Y, Aizawa M. Electrically promoted protein production by
mammalian cells cultured on the electrode surface. Biotechnology and Bioengineering, 39, 27,
1992
200. Alikacem N, Marois Y, Zhang Z, Jakubiec B, Roy R, King MW, Guidoin R. Tissue reactions
to polypyrrole-coated polyesters: A magnetic resonance relaxometry study. Artificial Organs, 23,
910, 1999
201. Akkouch A, Shi G, Zhang Z, Rouabhia M. Bioactivating electrically conducting polypyrrole
with fibronectin and bovine serum albumin. J. Biomed. Mater. Res., 92A, 221, 2010
202. Kojima J, Shinohara H, Ikariyama Y, Aizawa M, Nagaike K, Morioka S. Electrically
controlled proliferation of human carcinoma cells cultured on the surface of an electrode. Journal
of Biotechnology, 18, 129, 1991
243
203. Kumagai E, Tominaga M, Harada S. Sensitivity of chronically HIV-1 infected HeLa cells to
electrical stimulation. Appl. Microbiol. Biotechnol., 63, 754, 2004
204. Funk RHW, Monsees T, Özkucur N. Electromagnetic effects – From cell biology to medicine.
Prog. Histochem. Cytochem., 43, 177, 2009
205. Sun S, Liu Y, Lipsky S, Cho M. Physical manipulation of calcium oscillations facilitates
osteodifferentiation of human mesenchymal stem cells. FASEB J., 21, 1472, 2007
206. Titushkin I, Sun S, Shin J, Cho M. Physicochemical control of adult stem cell differentiation:
shedding light on potential molecular mechanisms. J. Biomed. Biotechnol., 2010, 743476, 2010
207. Levin M. Large-scale biophysics: ion flows and regeneration. Trends Cell Biol., 17, 261, 2007
208. Curtze S, Dembo M, Miron M, Jones DB. Dynamic changes in traction forces with DC
electric field in osteoblast-like cells. J. Cell Sci., 117, 2721, 2004
209. Cairns SP, Chin ER, Renaud JM. Stimulation pulse characteristics and electrode configuration
determine site of excitation in isolated mammalian skeletal muscle: implications for fatigue. J.
Appl. Physiol., 103, 359, 2007
210. Kimura K, Yanagida Y, Haruyama T, Kobatake E, Aizawa M. Gene expression in the
electrically stimulated differentiation of PC12 cells. Journal of Biotechnology, 63, 55, 1998
211. Koyama S, Haruyama T, Kobatake E, Aizawa M. Electrically induced NGF production by
astroglial cells. Nature Biotechnology, 15, 164, 1997
212. Astumian D and Berg H. Direct electric field effects and sequential processes in Biosystems.
Bioelectrochemistry and Bioenergetics, 25, 455, 1991
213. Chang DC. Cell poration and cell fusion using an oscillating electric field. Biophys. J.
Biophysical Society, 56, 641, 1989
214. Angersbach A, Heinz V, Knorr D. Effects of pulsed electric fields on cell membranes in real
food systems. Innovative Food Science and Emerging Technology, 1, 135, 2000
215. Aronsson K, Borch E, Stenlof B, Ronner U. Growth of pulsed electric field exposed
Escherichia coli in relation to inactivation and environmental factors. International Journal of Food
Microbiology, 93, 1, 2004
216. Golzio M, Rols MP, Teissie J. In vitro and in vivo electric field-mediated permeabilization,
gene transfer, and expression. Methods, 33, 126, 2004
217. Barati H, Amrollahi R, Ghafurifard H. Conductivity spectra of cellular systems in pulsed
electric fields: Electro-compression disintegration and thermal instability effects. Food and
Bioproducts Processing, 89, 400, 2011
218. Glaser R. Current concepts of the interaction of weak electromagnetic fields with cells.
Bioelectrochemistry and Bioenergetics, 27, 255, 1992
219. Berg H, Augsten K, Bauer E, Forster W, Jacob HE, Muhlig P, Weber H. Possibilities of cell
fusion and transformation by electrostimulation. Bioelectrochemistry and Bioenergetics, 12, 119,
1984
220. Levin, M. Bioelectric mechanisms in regeneration: Unique aspects and future perspectives.
Semin. Cell Dev. Biol., 20, 543, 2009
244
221. Zhao M, Forrester JV, McCaig CD. A small, physiological electric field orients cell division.
Proc. Natl. Acad. Sci. USA, 96, 4942, 1999
222. Ferrier J, Ross SM, Kanehisa J, Aubin JE. Osteoclasts and osteoblasts migrate in opposite
directions in response to a constant electrical field. J. Cell Physiol., 129, 283, 1986
223. Bayat M, Asgari-Moghadam Z, Maroufi M, Rezaie FS, Bayat M, Rakhshan M. Experimental
wound healing using microamperage electrical stimulation in rabbits. J. Rehabil. Res. Dev., 43,
219, 2006
224. Jerčinović A, Hinsenkamp M, Scarceriaux B, Willaert F, de Graef Ch, Heenen M, Goldshmidt
D. Effects of direct constant current (DC) on keratinocytes in vitro. Bioelectrochem. Bioenerg., 39,
209, 1996
225. Schmidt CE, Shastri VR, Vacanti JP, Langer R. Stimulation of neurite outgrowth using an
electrically conducting polymer. Proc. Natl. Acad. Sci. USA, 94, 8948, 1997
226. Aebischer P, Valentini RF, Dario P, Domenici C, Galletti PM. Piezoelectric guidance
channels enhance regeneration in the mouse sciatic nerve after axotomy. Brain Research, 436, 165,
1987
227. Howe CL. Depolarization of PC12 cells induces neurite outgrowth and enhances nerve growth
factor-induced neurite outgrowth in rats. Neuroscience Letters, 351, 41, 2003
228. Peltier LF. A brief historical note on the use of electricity in the treatment of fractures. Clin.
Orthop. Relat. Res., 161, 4, 1981
229. Toth JM, Seim HB, Schwardt JD, Humphrey WB, Wallskog JA, Turner AS. Direct current
electrical stimulation increases the fusion rate of spinal fusion cages. Spine, 25, 2580, 2000
230. Fini M, Cadossi R, Canè V, Cavani F, Giavaresi G, Krajewski A, Martini L, Aldini NN,
Ravaglioli A, Rimondini L, Torricelli P, Giardino R. The effect of pulsed electromagnetic fields on
the osteointegration of hydroxyapatite implants in cancellous bone: a morphologic and
microstructural in vivo study. J. Orthop. Res., 20, 756, 2002
231. Garland D, Holt P, Harrington JT, Caldwell J, Zizic T, Cholewczynski J. A 3-month,
randomized, double-blind, placebo-controlled study to evaluate the safety and efficacy of a highly
optimized, capacitively coupled, pulsed electrical stimulator in patients with osteoarthritis of the
knee. Osteoarthritis Cartilage, 15, 630, 2007
232. Satter Syed A, Islam MS, Rabbani KS, Talukder MS. Pulsed electromagnetic fields for the
treatment of bone fractures. Bangladesh Med. Res. Counc. Bull., 25, 6, 1999
233. Brighton CT, Wang W, Clark CC. Up-regulation of matrix in bovine articular cartilage
explants by electric fields. Biochem. Biophys. Res. Commun., 342, 556, 2006
234. Bursac N, Loo Y, Leong K, Tung L. Novel anisotropic engineered cardiac tissues: Studies of
electrical propagation. Biochemical and Biophysical Research Communications, 361, 847, 2007
235. Barash Y, Dvir T, Tandeitnik P, Ruvinov E, Guterman H, Cohen S. Electric field stimulation
integrated into perfusion bioreactor for cardiac tissue engineering. Tissue Engineering: Part C, 16,
1417, 2010
236. Fishler MG, Sobie EA, Thakor NV, Tung L. Mechanisms of cardiac cell excitation with
premature monophasic and biphasic field stimuli: a model study. Biophysical Journal, 70, 1347,
1996
245
237. Inoue N, Ohkusa T, Nao T, Lee JK, Matsumoto T, Hisamatsu Y, Satoh T, Yano M, Yasui K,
Kodama I, Matsuzaki M. Rapid electrical stimulation of contraction modulates gap junction protein
in neonatal rat cultured cardiomyocytes: Involvement of mitogen-activated protein kinases and
effects of angiotensin II-receptor antagonist. J. Am. Coll. Cardiol., 44, 914, 2004
238. Dusterhoft S and Pette D. Effects of electrically induced contractile activity on cultured
embryonic chick breast muscle cells. Differentiation, 44, 178, 1990
239. Johnson TB, Kent RL, Bubolz BA, McDermott PJ. Electrical stimulation of contractile
activity accelerates growth of cultured neonatal cardiocytes. Circ. Res., 74, 448, 1994
240. Donnelly K, Khodabukus A, Philp A, Deldicque L, Dennis RG, Baar K. A novel bioreactor
for stimulating skeletal muscle in vitro. Tissue Engineering: Part C, 16, 711, 2010
241. Burch N, Arnold AS, Item F, Summermatter S, Brochmann G, Santos S, Christe M, Boutellier
U, Toigo M, Handschin C. Electric pulse stimulation of cultured murine muscle cells reproduces
gene expression changes of trained mouse muscle. PLoS ONE, 5, e10970, 2010
242. Badylak SF, Hinds M, Geddes LA. Comparison of three methods of electrical stimulation for
converting skeletal muscle to a fatigue resistant power source suitable for cardiac assistance.
Annals of Biomedical Engineering, 18, 239, 1990
243. Bayol S, Brownson C, Loughna PT. Electrical stimulation modulates IGF binding protein
transcript levels in C2C12 myotubes. Cell Biochem. Funct., 23, 361, 2005
244. Fujita H, Nedachi T, Kanzaki M. Accelerated de novo sarcomere assembly by electric pulse
stimulation in C2C12 myotubes. Experimental Cell Research, 313, 1853, 2007
245. Kanno S, Oda N, Abe M, Saito S, Hori K, Handa Y, Tabayashi K, Sato Y. Establishment of a
simple and practical procedure applicable to therapeutic angiogenesis. Circulation, 99, 2682, 1999
246. Hellsten Y and Frandsen U. Adenosine formation in contracting primary rat skeletal muscle
cells and endothelial cells in culture. J. Physiol., 504, 695, 1997
247. Dennis RG, Kosnik PE, Gilbert ME, Faulkner JA. Excitability and contractility of skeletal
muscle engineered from primary cultures and cell lines. Am. J. Physiol. Cell Physiol., 280, C288,
2001
248. Guimard NK, Gomez N, Schmidt CE. Conducting polymers in biomedical engineering. Prog.
Polym. Sci., 32, 876, 2007
249. Ateh DD, Navsaria HA, Vadgama P. Polypyrrole-based conducting polymers and interactions
with biological tissues. J. R. Soc. Interface, 3, 741, 2006
250. Kumar D and Sharma RC. Advances in conductive polymers. Eur. Polym. J., 34, 1053, 1998
251. Wan M. Conducting polymers with micro or nanometer structure. Springer Berlin Heidelberg,
1, 2008.
252. Rupprecht L (ed). Conductive polymers and plastics in industrial applications. William
Andrew Publishing - Plastics Design Library, 1999. - Epstein AJ. Electrical conductivity in
conjugated polymers. 1, 1999
253. Kirchmeyer S, Reuter K. Scientific importance, properties and growing applications of
poly(3,4-ethylenedioxythiophene). J. Mater. Chem., 15, 2077, 2005
254. Balint R, Cassidy NJ, Cartmell SH. Conductive polymers: Towards a smart biomaterial for
tissue engineering. Acta Biomaterialia, 10, 2341, 2014
246
255. Zhou DD and Greenbaum E (eds.). Implantable Neural Prostheses 2. Springer
Science+Business Media LLC, 2010. - Zhou DD, Cui XT, Hines A, Greenberg RJ. Conducting
polymers in neural stimulation applications. 217, 2010
256. Cortés MT, Moreno JC. Artificial muscles based on conducting polymers. e-Polymers, 4, 1,
2003
257. Lee JY, Bashur CA, Goldstein AS, Schmidt CE. Polypyrrole-coated electrospun PLGA
nanofibers for neural tissue applications. Biomaterials, 30, 4325, 2009
258. Kim DH, Abidian M, Martin DC. Conducting polymers grown in hydrogel scaffolds coated on
neural prosthetic devices. J. Biomed. Mater. Res., 71A, 577, 2004
259. Guimard NKE, Sessler JL, Schmidt CE. Toward a biocompatible and biodegradable
copolymer incorporating electroactive oligothiophene units. Macromolecules, 42, 502, 2009
260. Li M, Guo Y, Wei Y, MacDiarmid AG, Lelkes PI. Electrospinning polyaniline-contained
gelatin nanofibers for tissue engineering applications. Biomaterials, 27, 2705, 2006
261. Liu X, Gilmore KJ, Moulton SE, Wallace GG. Electrical stimulation promotes nerve cell
differentiation on polypyrrole/poly(2-methoxy-5 aniline sulfonic acid) composites. J. Neural Eng.,
6, 1, 2009
262. Rivers TJ, Hudson TW, Schmidt CE. Synthesis of a novel, Biodegradable electrically
conducting polymer for biomedical applications. Adv. Funct. Mater., 12, 33, 2002
263. Garner B, Georgevich A, Hodgson AJ, Liu L, Wallace GG. Polypyrrole–heparin composites
as stimulus-responsive substrates for endothelial cell growth. Inc. J. Biomed. Mater. Res., 44, 121,
1999
264. Garner B, Hodgson AJ, Wallace GG, Underwood PA. Human endothelial cell attachment to
and growth on polypyrrole-heparin is vitronectin dependent. J. Mater. Sci. Mater. Med., 10, 19,
1999
265. Cui X, Hetke JE, Wiler JA, Anderson DJ, Martin DC. Electrochemical deposition and
characterization of conducting polymer polypyrole/PSS on multichannel neural probes. Sensor
Actuat. A-Phys., 93, 8, 2001
266. Lee JY, Lee JW, Schmidt CE. Neuroactive conducting scaffolds: nerve growth factor
conjugation on active ester-functionalized polypyrrole. J. R. Soc. Interface, 6, 801, 2009
267. Ateh DD, Vadgama P, Navsaria HA. Culture of human keratinocytes on polypyrrole-based
conducting polymers. Tissue Eng., 12, 645, 2006
268. Bousalem S, Mangeney C, Chehimi MM, Basinska T, Miksa B, Slomkowski S. Synthesis,
characterization and potential biomedical applications of N –succinimidyl ester functionalized,
polypyrrole-coated polystyrene latex particles. Colloid Polym. Sci., 282, 1301, 2004
269. Li Y, Neoh KG, Kang ET. Plasma protein adsorption and thrombus formation on surface
functionalized polypyrrole with and without electrical stimulation. J. Colloid Interf. Sci., 275, 488
2004
270. George PM, LaVan DA, Burdick JA, Chen CY, Liang E, Langer R. Electrically controlled
drug delivery from biotin-doped conductive polypyrrole. Adv. Mater., 18, 577, 2006
271. Meng S, Rouabhia M, Shi G, Zhang Z. Heparin dopant increases the electrical stability, cell
adhesion, and growth of conducting polypyrrole/poly(L,L-lactide) composites. J. Biomed. Mater.
Res., 87A, 332, 2008
247
272. Kim DH, Richardson-Burns SM, Hendricks JL, Sequera C, Martin DC. Effect of immobilized
nerve growth factor on conductive polymers: Electrical properties and cellular response. Adv.
Funct. Mater., 17, 79, 2007
273. Cetiner S, Kalaoglu F, Karakas H, A. Sarac AS. Electrospun nanofibers of polypyrrole-
poly(acrylonitrile-co-vinyl acetate). Text. Res. J., 80, 1784, 2010
274. Akkouch A, Shi G, Zhang Z, Rouabhia M. Bioactivating electrically conducting polypyrrole
with fibronectin and bovine serum albumin. J. Biomed. Mater. Res., 92A, 221, 2010
275. Ferraz N, Strømme M, Fellstrom B, Pradhan S, Nyholm L, Mihranyan A. In vitro and in vivo
toxicity of rinsed and aged nanocellulose–polypyrrole composites. J. Biomed. Mater. Res. A, 100,
2128, 2012
276. Zhang X and Manohar SK. Bulk synthesis of polypyrrole nanofibers by a seeding approach. J.
Am. Chem. Soc., 126, 12714, 2004
277. Chronakis IS, Grapenson S, Jakob A. Conductive polypyrrole nanofibers via electrospinning:
Electrical and morphological properties. Polymer, 47, 1597, 2006
278. Bousalem S, Yassar A, Basinska T, Miksa B, Slomkowski S, Azioune A, Chehimi MM.
Synthesis, characterization and biomedical applications of functionalized polypyrrole-coated
polystyrene latex particles. Polym. Adv. Technol., 14, 820, 2003
279. Gomez N and Schmidt CE. Nerve growth factor-immobilized polypyrrole: Bioactive
electrically conducting polymer for enhanced neurite extension. J. Biomed. Mater. Res. A., 81, 135,
2007
280. Song HK, Toste B, Ahmann K, Hoffman-Kim D, Palmore GTR. Micropatterns of positive
guidance cues anchored to polypyrrole doped with polyglutamic acid: A new platform for
characterizie neurite extension in complex environments. Biomaterials, 27, 473, 2006
281. Zhang Z, Rouabhia M, Wang Z, Roberge C, Shi G, Roche P, Li J, Dao LH . Electrically
conductive biodegradable polymer composite for nerve regeneration: Electricity-stimulated neurite
outgrowth and axon regeneration. Artif. Organs, 31, 13, 2007
282. Brahim S and Guiseppi-Elie A. Electroconductive hydrogels: Electrical and electrochemical
properties of polypyrrole-poly(HEMA) composites. Electroanalysis, 17, 556, 2005
283. Bettinger CJ, Bruggeman JP, Misra A, Borenstein JT, Langer R. Biocompatibility of
biodegradable semiconducting melanin films for nerve tissue engineering. Biomaterials, 30, 3050,
2009
284. Lee JW, Serna F, Nickels J, Schmidt CE. Carboxylic acid-functionalized conductive
polypyrrole as a bioactive platform for cell adhesion. Biomacromolecules, 7, 1692, 2006
285. Castano H, O’Rear EA, McFetridge PS, Sikavitsas VI. Polypyrrole thin films formed by
admicellar polymerization support the osteogenic differentiation of mesenchymal stem cells.
Macromol. Biosci., 4, 785, 2004
286. Aoki T, Tanino M, Sanui K, Ogata N, Kumakura K. Secretory function of adrenal chromaffin
cells cultured on polypyrrole films. Biomaterials, 17, 1971, 1996
287. Gomez N, Lee JY, Nickels JD, Schmidt CE. Micropatterned polypyrrole: A combination of
electrical and topographical characteristics for the stimulation of cells. Adv. Funct. Mater., 17,
1645, 2007
248
288. George PM, Lyckman AW, LaVan DA, Hegde A, Leung Y, Avasare R, Testa C, Alexander
PM, Langer R, Sur M. Fabrication and biocompatibility of polypyrrole implants suitable for neural
prosthetics. Biomaterials, 26, 3511, 2005
289. Alikacem N, Marois Y, Zhang Z, Jakubiec B, Roy R, King MW, Guidoin R. Tissue reactions
to polypyrrole-coated polyesters: A magnetic resonance relaxometry study. Artif. Organs, 23, 910,
1999
290. Batz P, Schmeisser D, Gopel W. Electronic structure of polypyrrole films. Phys. Rev. B, 43,
9178, 1991
291. Ribo JM, Acero C, Anglada MC, Dicko A, Tura JM, Ferrer-Anglada N, Albareda A, On the
structure and transport properties of polypyrroles. Butll. Soc. Cat. Cien., 13, 335, 1992
292. Street GB, Lindsey SE, Nazzal AI, Wynne KJ. The structure and mechanical properties of
polypyrrole. Mol. Cryst. Liq. Cryst., 118, 137, 1985
293. Blinova NV, Stejskal J, Trchova M, Prokes J. Control of polyaniline conductivity and contact
angles by partial protonation. Polym. Int., 57, 66, 2008
294. Cullen DK, Patel AR, Doorish JF, Smith DH, Pfister BJ. Developing a tissue-engineered
neural-electrical relay using encapsulated neuronal constructs on conducting polymer fibers. J.
Neural Eng., 5, 374, 2008
295. Borriello A, Guarino V, Schiavo L, Alvarez-Perez MA, Ambrosio L. Optimizing PANi doped
electroactive substrates as patches for the regeneration of cardiac muscle. J. Mater. Sci.: Mater.
Med., 22, 1053, 2011
296. Guo Y, Li M, Mylonakis A, Han J, MacDiarmid AG, Chen X, Lelkes PI, Wei Y. Electroactive
oligoaniline-containing self-assembled monolayers for tissue engineering applications.
Biomacromolecules, 8, 3025, 2007
297. Prabhakaran MP, Ghasemi-Mobarakeh L, Jin G, Ramakrishna S. Electrospun conducting
polymer nanofibers and electrical stimulation of nerve stem cells. J. Biosci. Bioeng., 112, 501,
2011
298. Yu QZ, Shi MM, Deng M, Wang M, Chen HZ. Morphology and conductivity of polyaniline
sub-micron fibers prepared by electrospinning. Mat. Sci. Eng. B-Solid., 150, 70, 2008
299. Huang L, Zhuang X, Hu J, Lang L, Zhang P, Wang Y, Chen X, Wei Y, Jing X. Synthesis of
biodegradable and electroactive multiblock polylactide and aniline pentamer copolymer for tissue
engineering applications. Biomacromolecules, 9, 850, 2008
300. Zhang QS, Yan YH, Li SP, Feng T. Synthesis of a novel biodegradable and electroactive
polyphosphazene for biomedical application. Biomed. Mater., 4, 1, 2009
301. Humpolicek P, Kasparkova V, Saha P, Stejskal J. Biocompatibility of polyaniline. Synthetic
Met., 162, 722, 2012
302. MacDiarmid AG, Epstein AJ. Secondary doping in polyaniline. Synthetic met., 69, 85, 1995
303. MacDiarmid AG, Chiang JC, Richter AF. Polyaniline: A new concept in conducting
polymers. Synthetic met., 18, 285, 1987
304. Pouget JP, Jdzefowiczt ME, Epstein AJ, Tang X, MacDiarmid AG. X-ray structure of
polyaniline. Macromolecules, 24, 779, 1991
249
305. Tourillon G, Garnier F. Morphology and crystallographic, structure of polythiophene and
derivatives. Mol. Cryst. Liq. Cryst., 118, 221, 1985
306. Peramo A, Urbanchek MG, Spanninga SA, Povlich LK, Cederna P, Martin DC. In situ
polymerization of a conductive polymer in acellular muscle tissue constructs. Tissue Eng. Part A,
14, 423, 2008
307. Thomas CA, Zong K, Schottland P, Reynolds JR. Poly(3,4-alkylenedioxypyrrole)s as highly
stable aqueous-compatible conducting polymers with biomedical implications. Adv. Mater., 12,
222, 2000
308. Asplund M, Thaning E, Lundberg J, Sandberg-Nordqvist AC, Kostyszyn B, Inganäs O, von
Holst H. Toxicity evaluation of PEDOT/biomolecular composites intended for neural
communication electrodes. Biomed. Mater., 4, 1, 2009
309. Richardson-Burns SM, Hendricks JL, Martin DC. Electrochemical polymerization of
conducting polymers in living neural tissue. J. Neural Eng., 4, L6, 2007
310. Aasmundtveit KE, Samuelsent EJ, Pettersson LAA, Inganas O, Johansson T, Feidenhans R.
Structure of thin films of poly(3,4-ethylenedioxythiophene. Synthetic Met., 101, 561, 1999
311. Richardson RT, Thompson B, Moulton S, Newbold C, Lum MG, Cameron A, Wallace G,
Kapsa R, Clark G, O'Leary S. The effect of polypyrrole with incorporated neurotrophin-3 on the
promotion of neurite outgrowth from auditory neurons. Biomaterials, 28, 513, 2007
312. Thompson BC, Richardson RT, Moulton SE, Evans AJ, O'Leary S, Clark GM, Wallace GG.
Conducting polymers, dual neurotrophins and pulsed electrical stimulation — Dramatic effects on
neurite outgrowth. J. Control Release, 141, 161, 2010
313. Kotwal A, Schmidt CE. Electrical stimulation alters protein adsorption and nerve cell
interactions with electrically conducting biomaterials. Biomaterials, 22, 1055, 2001
314. Ghasemi-Mobarakeh L, Prabhakaran MP, Morshed M, Nasr-Esfahani MH, Ramakrishna S.
Electrical stimulation of nerve cells using conductive nanofibrous scaffolds for nerve tissue
engineering. Tissue Eng. Part A, 15, 3605, 2009
315. Huang J, Hu X, Lu L, Ye Z, Zhang Q, Luo Z. Electrical regulation of Schwann cells using
conductive polypyrrole/chitosan polymers. J. Biomed. Mater. Res., 93A, 164, 2010
316. Jeong SI, Jun ID, Choi MJ, Nho YC, Lee YM, Shin H. Development of electroactive and
elastic nanofibers that contain polyaniline and poly(L-lactide-co-e-caprolactone) for the control of
cell adhesion. Macromol. Biosci., 8, 627, 2008
317. Shi G, Rouabhia M, Wang Z, Dao LH, Zhang Z. A novel electrically conductive and
biodegradable composite made of polypyrrole nanoparticles and polylactide. Biomaterials, 25,
2477, 2004
318. Shi G, Rouabhia M, Meng S, Zhang Z. Electrical stimulation enhances viability of human
cutaneous fibroblasts on conductive biodegradable substrates. J. Biomed. Mater. Res., 84A, 1026,
2008
319. Shi G, Zhang Z, Rouabhia M. The regulation of cell functions electrically using biodegradable
polypyrrole–polylactide conductors. Biomaterials, 29, 3792, 2008
320. Li Y, Neoh KG, Kang ET. Controlled release of heparin from polypyrrole-poly(vinyl alcohol)
assembly by electrical stimulation. J. Biomed. Mater. Res., 73A, 171, 2005
250
321. Kapur R, Tilien J, Black J. Field-dependent fibroblast orientation on charged surfaces is
independent of polarity and adsorbed serum proteins. Biomaterials, 14, 854, 1993
322. Valentini RF, Sabatini AM, Dario P, Aebischer P. Polymer electret guidance channels
enhance peripheral nerve regeneration in mice. Brain Res., 480, 300, 1989
323. Fine EG, Valentini RF, Bellamkonda R, Aebische P. Improved nerve regeneration through
piezoelectric vinylidenefluoride-trifluoroethylene copolymer guidance channels. Biomaterials, 12,
775, 1991
324. Aebischer P, Valentini RF, Dario P, Domenici C, Galletti PM. Piezoelectric guidance
channels enhance regeneration in the mouse sciatic nerve after axotomy. Brain Res., 436, 165,
1989
325. Jin G, Prabhakaran MP, Liao S, Ramakrishna S. Photosensitive materials and potential of
photocurrent mediated tissue regeneration. J. Photochem. Photobiol. B, 102, 93, 2011
326. Wiesmann HP, Joos U, Meyer U. Biological and biophysical principles in extracorporal bone
tissue engineering Part II. Int. J. Oral Maxillofac. Surg., 33, 523, 2004
327. Martin I, Wendt D, Heberer M. The role of bioreactors in tissue engineering. TRENDS in
Biotechnology, 22, 80, 2004
328. Pörtner R, Nagel-Heyer S, Goepfert C, Adamietz P, Meenen NM. Bioreactor design for tissue
engineering. Journal of Bioscience and Bioengineering, 100, 235, 2005
329. Martin Y and Vermette P. Bioreactors for tissue mass culture: Design, characterization, and
recent advances. Biomaterials, 26, 7481, 2005
330. Yeatts AB and Fisher JP. Bone tissue engineering bioreactors: Dynamic culture and the
influence of shear stress. Bone, 48, 171, 2011
331. Bjerre L, Bunger CE, Kassem M , Mygind T. Flow perfusion culture of human mesenchymal
stem cells on silicate-substituted tricalcium phosphate scaffolds. Biomaterials, 29, 2616, 2008
332. Plunkett N and O’Brien FJ. Bioreactors in tissue engineering. Technology and Health Care,
19, 55, 2011
333. Lee SJ, Park JH, Seol YJ, Lee IH, Kang SS, Cho DW. Development of a bioreactor for the
mechanical stimulation of agarose hydrogels. Microelectronic Engineering, 86, 1411, 2009
334. Boonen KJM, Langelaan MLP, Polak RB, van der Schaft DWJ, Baaijens FPT, Post MJ.
Effects of a combined mechanical stimulation protocol: Value for skeletal muscle tissue
engineering. Journal of Biomechanics, 43, 1514, 2010
335. Cartmell SH, Porter BD, García AJ, Guldberg RE. Effects of medium perfusion rate on cell-
seeded three-dimensional bone constructs in vitro. Tissue Engineering, 9, 1197, 2003
336. Jeong SI, Kwon JH, Lim JI, Cho SW, Jung Y, Sung WJ, Kim SH, Kim YH, Lee YM, Kim
BS, Choi CY, Kim SJ. Mechano-active tissue engineering of vascular smooth muscle using
pulsatile perfusion bioreactors and elastic PLCL scaffolds. Biomaterials, 26, 1405, 2005
337. Gamboa OL, Pu J, Townend J, Forrester JV, Zhao M, McCaig C, Lois N. Electrical
estimulation of retinal pigment epithelial cells. Experimental Eye Research, 91, 195, 2010
338. Donnelly K, Khodabukus A, Philp A, Deldicque L, Dennis RG, Baar K. A novel bioreactor
for stimulating skeletal muscle in vitro. Tissue Engineering: Part C, 16, 711, 2010
251
339. Tandon N, Cannizzaro C, Chao PHG, Maidhof R, Marsano A, Au HTH, Radisic M, Vunjak-
Novakovic G. Electrical stimulation systems for cardiac tissue engineering. Nat Protoc., 4, 155,
2009
340. MATWEB Materials Information Resource – www.matweb.com – Last retrieved 05/04/2014
341. Nedachi T, Fujita H, Kanzaki M. Contractile C2C12 myotube model for studying exercise-
inducible responses in skeletal muscle. Am. J. Physiol. Endocrinol. Metab., 295, E1191, 2008
342. Okutsu S, Hatakeyama H, Kanzaki M, Tsubokawa H, Nagatomi R. Electric pulse stimulation
induces NMDA glutamate receptor mRNA in NIH3T3 mouse fibroblasts. The Tohoku Journal of
Experimental Medicine, 215, 18, 2008
343. Bordji G, Jouzeau JY, Mainard D, Payan E, Delagoutte JP, Netter P. Evaluation of the effect
of three surface treatments on the biocompatibility of 316L stainless steel using human
differentiated cells. Biomaterials, 17, 491, 1996
344. Muller R, Abke J, Schnell E, Macionczyk F, Gbureck U, Mehrl R, Ruszczak Z, Kujat R,
Englert C, Nerlich M, Angele P. Surface engineering of stainless steel materials by covalent
collagen immobilization to improve implant biocompatibility. Biomaterials, 26, 6962, 2005
345. Assad M, Lemieux N, Rivard CH, Yahia L’H. Comparative in vitro biocompatibility of
nickel-titanium, pure nickel, pure titanium, and stainless steel: genotoxicity and atomic absorption
evaluation. Bio-Medical Materials and Engineering, 9, 1, 1999
346. Zitter H and Plenk Jr. H. The electrochemical behavior of metallic implant materials as an
indicator of their biocompatibility. Journal of Biomedical Materials Research, 21, 881, 1987
347. Chen M , Zamora PO, Som P, Peña LA, Osaki S. Cell attachment and biocompatibility of
polytetrafluoroethylene (PTFE) treated with glow-discharge plasma of mixed ammonia and
oxygen. Journal of Biomaterials Science, Polymer Edition, 14, 917, 2003
348. Wang H, Kwok DTK, Xu M, Shi H, Wu Z, Zhang W, Chu PK. Tailoring of mesenchymal
stem cells behavior on plasma-modified polytetrafluoroethylene. Adv. Mater., 24, 3315, 2012
349. Baumgartner L, Arnhold S, Brixius K, Addicks K, Bloch W. Human mesenchymal stem cells:
Influence of oxygen pressure on proliferation and chondrogenic differentiation in fibrin glue in
vitro. Inc. J. Biomed. Mater. Res., 93A, 930, 2010
350. D'Ippolito G, Diabira S, Howard GA, Roos BA, Schiller PC. Low oxygen tension inhibits
osteogenic differentiation and enhances stemness of human MIAMI cells. Bone, 39, 513, 2006
351. Fehrer C, Brunauer R, Laschober G, Unterluggauer H, Reitinger S, Kloss F, Gülly C, Gaßner
R, Lepperdinger G. Reduced oxygen tension attenuates differentiation capacity of human
mesenchymal stem cells and prolong their lifespan. Aging Cell, 6, 745, 2007
352. Holzwarth C, Vaegler M, Gieseke F, Pfister SM, Handgretinger R, Kerst G, Müller I. Low
physiologic oxygen tensions reduce proliferation and differentiation of human multipotent
mesenchymal stromal cells. Cell Biology, 11, 1, 2010
353. Wuertz K, Godburn K, Neidlinger-Wilke C, Urban J, Iatridis JC. Behavior of mesenchymal
stem cells in the chemical microenvironment of the intervertebral disc. SPINE, 33, 1843, 2008
354. Wuertz K, Godburn K, Iatridis JC. MSC response to pH levels found in degenerating
intervertebral discs. Biochemical and Biophysical Research Communications, 379, 824, 2009
252
355. Ratner BD, Hoffman AD, Schoen FJ, Lemons JE (eds.). Biomaterials science. (3rd
ed.),
Academic Press, 2013 - Godier-Furnemont A and Vunjak-Novakovic G. Chapter II Cardiac muscle
tissue engineering.1262, 2013
356. Hecker L, Khait L, Radnoti D, Birla R. Novel bench-top perfusion system improves functional
performance of bioengineered heart muscle. Journal of Bioscience and Bioengineering, 107, 183,
2009
357. Kaw AK and Kalu EE. Numerical methods with applications. 2008
358. Li S and Liu WK. Meshfree and particle methods and their applications. Appl. Mech. Rev.,
55, 1, 2002
359. Zienkiewicz OC, Taylor RL, Zhu JZ. The finite element method: Its basis and fundamentals.
(6th ed.), Elsevier Butterworth-Heinemann, Oxford, 2005
360. Chaskalovic J. Finite element methods for engineering sciences. Springer-Verlag, Berlin
Heidelberg, 2008
361. Ang WT. A beginner’s course in boundary element methods. Universal Publishers, Boca
Raton USA, 2007
362. Bai EW and Lonngren KE. Capacitors and the method of moments. Computers and Electrical
Engineering, 30, 223, 2004
363. Bakshi UA and Bakshi AV. Electromagnetic theory. (1st ed.) - Chapter 5: Conductors,
dielectrics and capacitance. Technical Publications, 5-1, 2009
364. Chen WK (ed.). Electrical engineering handbook. Academic Press, Burlington, 2005 - Diaz
RE. Chapter V-2: Electrostatics. 499, 2005
365. Bird J. Electrical and electronic principles and technology. (3rd
ed.) – Chapter 6: Capacitors
and capacitance. Elsevier Ltd., Oxford, 55, 2007
366. Semat H and Katz R. Physics. - Chapter 25: Capacitance and dielectrics. Rinehart and
Company, Inc., New York, 461, 1958
367. Budenstein PP. (ed.) Digest of literature on dielectrics. National Academies, Washington,
1974
368. The Engineering ToolBox website - http://www.engineeringtoolbox.com/density-solids-
d_1265.html - Last retrieved: 05/05/2014
369. Amos SW and Amos RS. Newnes dictionary of electronics. Elsevier Ltd., 1999
370. Bishop O. Understand amplifiers. - Chapter 9: Imperfections. Elsevier Ltd., 1998
371. Douglas Self. Audio power amplifier design handbook. (5th ed.) - Chapter 8: Compensation,
slew rate, and stability. Elsevier Ltd., 215, 2009
372. Barash Y, Dvir T, Tandeitnik P, Ruvinov E, Guterman H, Cohen S. Electric field stimulation
integrated into perfusion bioreactor for cardiac tissue engineering. Tissue Engineering: Part C, 16,
1417, 2010
373. Andreev VP. Cytoplasmic electric fields and electroosmosis: Possible solution for the
paradoxes of the intracellular transport of biomolecules. PLoS ONE, 8, 1, 2013
253
374. Miklavcic D, Semrov D, Mekid H, Mir LM. A validated model of in vivo electric field
distribution in tissues for electrochemotherapy and for DNA electrotransfer for gene therapy.
Biochimica et Biophysica Acta, 1523, 73, 2000
375. Hu Q, Viswanadham S, Joshi RP, Schoenbach KH, Beebe SJ, Blackmore PF. Simulations of
transient membrane behavior in cells subjected to a high-intensity ultrashort electric pulse. Phys.
Rev. E, 71, 031914, 2005
376. Gumbart J, Khalili-Araghi F, Sotomayor M, Roux B. Constant electric field simulations of the
membrane potential illustrated with simple systems. Biochim. Biophys. Acta., 1812, 294, 2012
377. Pliquett U, Joshi RP, Sridhara V, Schoenbach KH. High electrical field effects on cell
membranes. Bioelectrochemistry, 70, 275, 2007
378. Hallab NJ, Bundy KJ, O'Connor K, Clark R, Moses RL. Cell adhesion to biomaterials:
correlations between surface charge, surface roughness, adsorbed protein, and cell morphology.
Journal of Long-term Effects of Medical Implants, 5, 209, 1995
379. Hunt JA, Flanagan BF, McLaughlin PJ, Strickland I, Williams DF. Effect of biomaterial
surface charge on the inflammatory response: Evaluation of cellular infiltration and TNFα
production. J. Biomed. Mater. Res., 31, 139, 1996
380. Gleixner R and Fromherz P. The extracellular electrical resistivity in cell adhesion. Biophys.
J., 90, 2600, 2006
381. Gross D, Loew LM, Webb WW. Optical imaging of cell membrane potential changes induced
by applied electric fields. Biophys. J., 50, 339, 1986
382. Freshney RI, Stacey GN, Auerbach JM (eds.). Culture of human stem cells. John Wiley &
Sons, Inc., Hoboken, New Jersey, 2007 - Gregory CA and Prockop DJ. Chapter 9: Fundamentals of
culture and characterization of mesenchymal stem/progenitor cells (MSCs) from bone marrow
stroma. 207, 2007
383. Smith JR, Pochampally R, Perry A, Hsu SC, Prockop DJ. Isolation of a highly clonogenic and
multipotential subfraction of adult stem cells from bone marrow stroma. Stem Cells, 22, 823, 2004
384. Colter DC, Class R, DiGirolamo CM, Prockop DJ. Rapid expansion of recycling stem cells in
cultures of plastic-adherent cells from human bone marrow. PNAS, 97, 3213, 2000
385. Bernardo ME, Cometa AM, Pagliara D, Vinti L, Rossi F, Cristantielli R, Palumbo G, Locatelli
F. Ex vivo expansion of mesenchymal stromal cells. Best Practice & Research Clinical
Haematology, 24, 73, 2011
386. Chang J, Lei H, Liu Q, Qin S, Ma K, Luo S, Zhang X, Huang W, Zuo Z, Fu H, Xia Y.
Optimization of culture of mesenchymal stem cells: a comparison of conventional plate and
microcarrier cultures. Cell Prolif., 45, 430, 2012
387. Digirolamo CM, Stokes D, Colter D, Phinney DG, Class R, Prockop DJ. Propagation and
senescence of human marrow stromal cells in culture: A simple colony-forming assay identifies
samples with the greatest potential to propagate and differentiate. British Journal of Haematology,
107, 275, 1999
388. Sekiya I, Colter DC, Prockop DJ. BMP-6 enhances chondrogenesis in a subpopulation of
human marrow stromal cells. Biochemical and Biophysical Research Communications, 284, 411,
2001
254
389. Sekiya I, Larson BL, Smith JR, Pochampally R, Cui JG, Prockop DJ. Expansion of human
adult stem cells from bone marrow stroma: Conditions that maximize the yields of early
progenitors and evaluate their quality. Stem Cells, 20, 530, 2002
390. Colter DC, Sekiya I, Prockop DJ. Identification of a subpopulation of rapidly selfrenewing
and multipotential adult stem cells in colonies of human marrow stromal cells. PNAS, 98, 7841,
2001
391. Holy CE, Shoichet MS, Davies JE. Engineering three-dimensional bone tissue in vitro using
biodegradable scaffolds: Investigating initial cell-seeding density and culture period. J. Biomed.
Mater. Res., 51, 376, 2000
392. Concaro S, Nicklasson E, Ellowsson L, Lindahl A, Brittberg M, Gatenholm P. Effect of cell
seeding concentration on the quality of tissue engineered constructs loaded with adult human
articular chondrocytes. J. Tissue Eng. Regen. Med., 2, 14, 2008
393. Kim K, Dean D, Lu A, Mikos AG, Fisher JP. Early osteogenic signal expression of rat bone
marrow stromal cells is influenced by both hydroxyapatite nanoparticle content and initial cell
seeding density in biodegradable nanocomposite scaffolds. Acta Biomater., 7, 1249, 2011
394. Wiedmann-Al-Ahmad M, Gutwald R, Lauer G, Hübner U, Schmelzeisen R. How to optimize
seeding and culturing of human osteoblast-like cells on various biomaterials. Biomaterials, 23,
3319, 2002
395. Griffin M, Iqbal SA, Sebastian A, Colthurst J, Bayat A. Degenerate wave and capacitive
coupling increase human MSC invasion and proliferation while reducing cytotoxicity in an in vitro
wound healing model. PLoS ONE, 6, e23404, 2011
396. Mace KA and Braun KM (eds.). Progenitor cells: Methods and protocols. Springer New York
, 2012 - Saleh FA, Frith JE, Lee JA, Genever PG. Chapter 4: Three-dimensional in vitro culture
techniques for mesenchymal stem cells. 31, 2012
397. Birgersdotter A, Sandberg R, Ernberg I. Gene expression perturbation in vitro - A growing
case for three-dimensional (3D) culture systems. Seminars in Cancer Biology, 15, 405, 2005
398. Baker BM and Chen CS. Deconstructing the third dimension – how 3D culture
microenvironments alter cellular cues. Journal of Cell Science, 125, 1, 2012
399. Gruber HE and Hanley Jr EN. Human disc cells in monolayer vs 3D culture: cell shape,
division and matrix formation. BMC Musculoskeletal Disorders, 1, 1, 2000
400. Liu H, Lin J, Roy K. Effect of 3D scaffold and dynamic culture condition on the global gene
expression profile of mouse embryonic stem cells. Biomaterials, 27, 5978, 2006
401. Frith JE, Thomson B, Genever PG. Dynamic three-dimensional culture methods enhance
mesenchymal stem cell properties and increase therapeutic potential. Tissue Engineering Part C:
Methods, 16, 735, 2010
402. Kabiri M, Kul B, Lott WB, Futrega K, Ghanavi P, Upton Z, Doran MR. 3D mesenchymal
stem/stromal cell osteogenesis and autocrine signalling. Biochemical and Biophysical Research
Communications, 419, 142, 2012
403. Wang W, Itaka K, Ohba S, Nishiyama N, Chung U, Yamasaki Y, Kataoka K. 3D spheroid
culture system on micropatterned substrates for improved differentiation efficiency of multipotent
mesenchymal stem cells. Biomaterials, 30, 2705, 2009
404. Reilly GC, Engler AJ. Intrinsic extracellular matrix properties regulate stem cell
differentiation. Journal of Biomechanics, 43, 55, 2010
255
405. Pek YS, Wan ACA, Ying JY. The effect of matrix stiffness on mesenchymal stem cell
differentiation in a 3D thixotropic gel. Biomaterials, 31, 385, 2010
406. Tse JR and Engler AJ. Stiffness gradients mimicking in vivo tissue variation regulate
mesenchymal stem cell fate. PLoS ONE, 6, e15978, 2011
407. Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix elasticity directs stem cell lineage
specification. 126, 677, 2006
408. Huebsch N, Arany PR, Mao AS, Shvartsman D, Ali OA, Bencherif SA, Rivera-Feliciano J,
Mooney DJ. Harnessing traction-mediated manipulation of the cell/matrix interface to control
stem-cell fate. Nature Materials, 9, 518, 2010
409. Ingber DE. Mechanosensation through integrins: Cells act locally but think globally. PNAS,
100, 1472, 2003
410. Kilian KA, Bugarija B, Lahn BT, Mrksich M. Geometric cues for directing the differentiation
of mesenchymal stem cells. PNAS, 107, 4872, 2010
411. Mathieu PS and Loboa EG. Cytoskeletal and focal adhesion influences on mesenchymal stem
cell shape, mechanical properties, and differentiation down osteogenic, adipogenic, and
chondrogenic pathways. Tissue Engineering: Part B, 18, 436, 2012
412. McBeath R, Pirone DM, Nelson CM, Bhadriraju K. Cell shape, cytoskeletal tension, and
RHOa regulate stem cell lineage commitment. Developmental Cell, 6, 483, 2004
413. Iqbal J and Zaidi M. Molecular regulation of mechanotransduction. Biochemical and
Biophysical Research Communications, 328, 751, 2005
414. Krendel M, Zenke FT, Bokoch GM. Nucleotide exchange factor GEF-H1 mediates cross-talk
between microtubules and the actin cytoskeleton. 4, 294, 2002
415. Nunes SS, Miklas JW, Liu J, Aschar-Sobbi R, Xiao Y, Zhang B, Jiang J, Massé S, Gagliardi
M, Hsieh A, Thavandiran N, Laflamme MA, Nanthakumar K, Gross GJ, Backx PH, Keller G,
Radisic M. Biowire: a platform for maturation of human pluripotent stem cell–derived
cardiomyocytes. Nature Methods, 10, 781, 2013
416. Sinclair IR and Dunton J. Practical electronics handbook. Newnes, 2007
417. Mazzoleni AP, Sisken BF, Kahler RL. Conductivity values of tissue culture medium from 20
°C to 40 °C. Bioelectromagnetics, 7, 95, 1986
418. Alonzo RJ. Electrical codes, standards, recommended practices and regulations - An
examination of relevant safety considerations. Elsevier, 2010
256
Appendix
257
A.1 ALTERNATIVE BIOREACTOR DESIGNS
A.1.1 The Parallel Field Bioreactor
A.1.1.1 Introduction
Figure A.1 – Sketch showing the difference between field lines perpendicular and parallel to the cell
monolayer
The capacitive electrical stimulation bioreactor used in this study (will be referred to as the
“Perpendicular” bioreactor) was designed to expose samples to an electric field with field lines
perpendicular to the cellular monolayer (Figure A.1). This approach was chosen for the following
reasons:
The initial bioreactor system could only deliver a maximum electrode potential of 15 V. To
generate the strongest possible electric field the electrodes had to be placed as close to each
other as possible. This could only be done if the electrodes are placed above and below the
cell layer rather than on its two sides (Figure A.2).
Figure A.2 – The minimum distance between the
two electrodes is the smallest, if the electrodes are
placed below and above the monolayer (left) rather
than on its two sides (right). At the same electrode
potential this will generate a much stronger electric
field strength.
258
According to one initial hypothesis, electric fields could act by hyper or depolarising the
cellular membrane. If this hypothesis is true, then every cell in the culture would only
receive the same stimulus, if the electrodes are placed above and below. If the electrodes
are placed on the two sides, one part of the culture would be hyperpolarised while the other
would be depolarised.
Galvanotaxis (the electric field induced migration and alignment of cells) can be a useful
tool, but when investigating the effect of electrical stimulation on hMSCs
proliferation/differentiation can also be detrimental. For example, if the cells are forced
into one side of the well, the lack of space can limit their proliferation. This would cause
the culture to contain fewer cells not as a result of the electric field affecting their
proliferative pathways, but by interfering with their spatial distribution. It is possible to
instigate galvanotaxis with parallel field stimulation, but not with perpendicular fields.
In summary, perpendicular field stimulation is the better choice if the effects of electrical signals on
proliferation/differentiation are explored, or when the maximisalisation of the available electric
field strength is essential. However, parallel field stimulation has its own advantages: Its ability to
align cells can help engineer tissues where the direction of alignment is important. Thus it lends
itself for the creation of muscle or peripheral nerve tissue. Stimulation is such way has already been
noted as beneficial for cardiac tissue engineering, promoting the maturation of tissue constructs
[339].
Therefore, an additional bioreactor was designed (Figure A.3), in order to enable the exploration of
the effect of parallel field stimulation.
259
A.1.1.2 The Design
Figure A.3 – 3D rendered image of the parallel plate bioreactor
Bioreactor plates were created by milling eight rectangular wells into blocks of 20mm thick PTFE
(RS Stock No. 197-0102, RS Components Ltd.). Stainless steel electrodes (cut from RS stock code:
264-7241, RS Components Ltd.) were placed into recesses milled into the underside of the plate on
the two sides of the wells (Figure A.4). 24x24mm rectangular glass coverslips (MIC3136,
Scientific Laboratory Supplies Ltd.) were placed into these wells to serve as a substrate for cellular
attachement. The plate is covered with a “normal” polystyrene plate lid.
Figure A.4 – Exploded view of the parallel field
bioreactor.
1 – Polystyrene plate top,
2 – Rectangular glass coverslips,
3 – PTFE bioreactor bottom,
4 – Electrodes
260
A.1.2 The Direct Bioreactor
The direct bioreactor is a modified version of the
parallel field bioreactor. Instead of the capacitive
stimulation that the “perpendicular” and “parallel”
bioreactors deliver, the direct bioreactor delivers direct
stimulation. This is due to the electrodes being in
contact with the culture medium, allowing not just an electric field to be present, but the flow of an
electrical current. Although an electrical current will result in chemical changes in the culture
medium, and has been known to be able to cause cellular death, it might be necessary for the
successful engineering of tissue constructs. For example, Radisic’s group have recently published
in the journal Nature Methods their technique for engineering cardiac muscle employing direct
electrical stimulation [415].
Figure A.6 – 3D rendered image of the parallel plate bioreactor
This design is very similar to the parallel field bioreactor (Figure A.6): 24x24mm coverslips were
placed into eight wells milled into blocks of PTFE (RS Stock No. 197-0102, RS Components Ltd.)
and covered by a polystyrene plate cover. In the case however the electrodes are not isolated from
the culture environment, but are in contact with the culture medium (Figure A.7). Electrodes have
been water-jet cut out of sheets of 316L (A4) stainless steel (RS stock code: 264-7241, RS
Figure A.5 – Direct parallel stimulation
261
Components Ltd.) and bent into the required shape. Electrodes are held in position using self-
tapping stainless steel screws (RS Stock No. 521-383, RS Components Ltd.).
Figure A.7 – Exploded view of the contact
bioreactor
1 – Polystyrene plate top,
2 – Electrodes,
3 – Rectangular glass coverslips,
4 – PTFE bioreactor bottom
262
A.2 THE ELECTRICAL CAPACITANCE AND
RESISTANCE OF THE THREE DIFFERENT
BIOREACTORS
In order to be able to accurately deliver a desired signal (i.e. correct amplitude, pulse shape, no
reflection of the signal, no standing waves) it must be known what sort of load the bioreactor will
pose upon the signal source. Will it be highly resistive or capacitive?
In order to ascertain this, the electrical resistance and capacitance of the three different electrical
bioreactor types - the perpendicular, the parallel and the direct – was calculated (Figure A.8).
Equations and mathematical formulae were taken from [416] and [365].
Figure A.8 – Schematic representation of the circuit
with the signal generator supplying one bioreactor
with 6 wells. Each well has an associated resistance
(Rn) and a capacitance (Cn) value. Re1 and Re2
represent the resistance of the electrodes, while Rcable is
the resistance of the cables leading to and away from
the bioreactor.
Electrical resistance can be calculated from the following equation:
Where “σ” is the conductivity of a material [S/m],
“l” is the length of the material [m]
“A” is the cross-sectional area of the material [m2]
263
Electrical capacitance of a parallel plate capacitor can be calculated as:
ε ε
Where “ε0” is the permittivity of space and is equal to 8.854 ∙ 10-12
[F/m],
“εr” is the relative permittivity of the material within the capacitor
“A” is the active area of the plates of the capacitor [m2]
“d” is the distance between the plates [m]
For the equations the following geometrical parameters and material properties were used:
Figure A.9 – Schematic representations of one well of the “perpendicular” (left), “parallel” (centre)
and “direct” (right) bioreactor.
Table A.1 – The parameters used in the calculations [340, 417]
264
A.2.1 The “Perpendicular” Bioreactor
The electrical resistance of one well of the bioreactor will be:
Ω
Then the resistance of the bioreactor can be calculated as:
The electrical capacitance of one is:
ε ε
ε ε
ε ε
And for the whole bioreactor this is:
A.2.2 The “Parallel” Bioreactor
The electrical resistance of one well can be calculated as follows:
Ω
For the whole bioreactor the resistance then will be:
265
The capacitance of one well of the parallel bioreactor is:
ε ε
ε ε
ε ε
For the whole bioreactor the capacitance will be:
A.2.3 The “Direct” Bioreactor
The electrical resistance of one well of the bioreactor:
Ω
Thus the electrical resistance of the whole bioreactor:
Electrical capacitance of one well of the bioreactor is:
ε ε
From this we can calculate the capacitance of the whole bioreactor:
266
A.2.4 The Electrical Impedance of the Bioreactors
In simple terms electrical impedance is the extension of the concept of resistance to the AC domain
– it represents the opposition a circuit shows to the flow of an electrical current, if an alternating
voltage is applied. It is a complex number consisting of a real and imaginary part. Its real part is
given by the electrical resistance, while is imaginary part consists of electrical capacitance and
inductance [416]:
Ω
Where “R” is the electrical resistance of the circuit [Ω]
“C” is the capacitance [F]
“L” is the inductance [H]
“s” is the complex frequency which is a complex number
The impedance of the bioreactors therefore will be:
]
]
The impedance of both the perpendicular and the parallel field bioreactor is very high and, as such,
can be treated as a circuit break. Therefore the load parameter on the signal source has to be set to
its maximum setting which is 10 kΩ. This also means that there should be virtually no current
flowing between the electrodes and the signal generator should have no difficulty
generating/maintaining the potential difference between the electrodes.
On the other hand, the resistance of the direct bioreactor is quite low (24.5 Ω), which will decrease
further (half to 12.25 Ω) if a further bioreactor is connected to the circuit. With a quick calculation
267
based on Ohm’s law (ignoring the resistance of the cables and electrodes) it can be shown that the
maximum potential difference that can be generated across the electrodes of the bioreactor using
the maximum current output of the signal generator system is:
Ω
It must be noted that this would push the system to its limits and requires the generation of an
electrical current that is lethal to humans. This obviously raises health and safety concerns: Above
10 mA an electrical shock can cause serious muscle contraction making it impossible to let go the
source of the shock [418]. Above 60 mA there is a risk of heart stoppage [418]. It is also important
to note that to the human life alternating current is much more dangerous than direct current. 30
mA of AC current has been reported to be as dangerous as 500 mA DC [418]. Therefore, if an
amplitude higher than 0.245 V is applied (generates a current of 10 mA in the case of one direct
bioreactor), serious care should be employed!
268
A.3 SIMULATIONS OF THE ELECTRIC FIELD INSIDE
THE PARALLEL AND DIRECT BIOREACTORS
Same as in the case of the perpendicular bioreactor, it is important to ascertain the
relationship between the electrical potential upon the electrodes and the electrical field
strength the cells experience in the other two bioreactor types. To this end, simulations of
the electrical field in the parallel and direct bioreactors were carried out in the commercial
software COMSOL Multiphysics. Electrode potentials were varied between 1 and 450 V.
The same relative permittivity values were used for the various materials as for the
perpendicular bioreactor.
Additional current density simulations were carried out for the direct bioreactor. The
electrical conductivity values that were used for these simulations can be seen in Table
A.2.
A “Free tetrahedral” mesh with the “Extra fine” setting was applied to all models.
Electrical field strength and current density (in the case of the direct bioreactor) results
were recorded using an averaging domain probe specified to the culture medium.
Material Electrical conductivity [S/m]
Culture medium 1.7
Steel 1.45 ∙ 106
Table A.2 – The electrical conductivity values used in the simulation [340, 417]
269
A.3.1 Geometries
A model of one well of the parallel bioreactor was
created using the following components
(Figure A.10):
1 – A 26x26x3 mm layer of culture medium
2 – The PTFE wall of the well
3 – Stainless steel electrodes
A model of direct bioreactor was created from these
components (Figure A.11):
1 – A 26x26x3 mm layer of medium
2 – A stainless steel electrodes
Figure A.11 – A drawing of the geometry used
for the direct bioreactor simulations (dimensions are in mm)
Figure A.10 – A drawing of the geometry used for the
parallel bioreactor simulations (dimensions are in mm)
270
A.3.2 Results
Figure A.12 – The electrical field strength in the parallel and direct bioreactors as a function of
electrode potential difference
The electrical field strength that the cells experience (Ecells) inside the two bioreactor types (Figure
A.12) is described by the following equations:
A.3.2.1 The Relationship between Electrode Potential and Current Density
It was previously shown that the electrical resistance of a contact bioreactor is 24.5 Ω. The
maximum current output of the signal source is 300 mA (0.3A).
From these values the maximum potential difference that can be generated between the electrodes
can be calculated using Ohm’s law:
Where “N” is the number of bioreactors connected to the signal source
271
The maximum electrical current flowing through one well of the bioreactor is:
The current density in one well of the contact bioreactor as the function of electrode potential
therefore can be calculated as:
This calculation agrees with the results of the COMSOL simulation:
Figure A.13 – The electrical current density inside the direct bioreactor as a function of electrode
potential difference
The relationship between the electrode potential and the current density inside the culture medium
in the direct bioreactor is linear (Figure A.13) and is described by the following equation:
Where “Jwell” is the electrical fiend strength inside the medium in A/m2,
“Uel” is the electrical potential difference between the electrodes in V.
“N” is the number of bioreactors being stimulated
272
The simulation also shows us that this current density is distributed in a homogenous manner
across the culture medium (Figure A.14), as such all cells in the well will be exposed to the same
current density.
Figure A.14 – A slice taken in the centre of the bioreactor displaying current density at and electrode
potential difference of 450 V. Blue indicates low, while red indicates high current density. The section
corresponding to the culture medium displays an even turquoise colour, showing that the current
density is distributed across the culture medium in a homogenous manner.
273
A.4 CHOOSING THE STIMULATION PARAMETERS FOR
THE DIRECT BIOREACTOR
There are certain factors that limit the range of electrical field strength and current density that can
be applied in the direct bioreactor. As it was mentioned previously, the maximum current output of
the signal source is 300 mA. Applied to a direct bioreactor the maximum potential difference that
this level of electrical current can generate is only 7.35 V. Connecting additional bioreactors to the
signal source will further reduce this value.
This electrical potential difference corresponds to a current density of 480 A/m2
and an electrical
field strength (Ecells) of approx. 3.9 V/m per well. This field strength is around the same magnitude
as the one generated by the “perpendicular” bioreactor at an electrode potential of 150 V.
Furthermore, this is the electrical field strength that Radisic’s group successfully used to engineer
their tissue constructs [415].
But how high is a current density of 480 A/m2? What sort of effect does it have on a cell culture? In
the literature the highest current density that was reported to have been used is 25 A/m2
(Table
A.3), a value significantly lower than 480 A/m2. Square wave stimulation with this current density
aided in the endothelial differentiation of embryonic stem cells [143], therefore can be concluded to
be beneficial to the cell culture. On the other hand, 5 A/m2
DC stimulation has been stated to cause
necrosis in osteoblast cultures [182].
Table A.3 – Examples of current densities applied for electrical stimulation
Article Stimulation type Effect
Kim et al, 2008 0.015 Biphasic Enhanced proliferation and VEGF of osteoblasts
Kim et al, 2009 0.15 BiphasicUpregulation of calcium deposition and alkaline phosphatase
activity of osteoblasts
Zhuang et al, 1997 3 Sinewave Enhanced proliferation and TGF-β1 expression in osteoblasts
Spadaro and Becker, 1979 5 DC Necrosis in osteoblasts above this treshold
Sauer et al, 2004 25 Square wave Endothelial differenctiation of ESCs
A/m2
Current density
274
Based on these two observations it can be inferred that:
Cells can tolerate much higher current density if the stimulation is dynamic (e.g. pulsed,
AC).
480 A/m2
current density is a very high value for cell culture, and, therefore, is likely to
cause necrosis.
The current densities for DC stimulation more then 5-10 A/m2
(which corresponds to an
electrode potential of 0.07-0.15 V) may cause necrosis.
The other limiting factor is the danger that the high electrical currents pose to the user. It has been
established in previous sections that the use of a current over 10 mA are considered to be a risk to
the user. This corresponds (in case of one bioreactor) to an electrical potential difference of 0.245
V, electrical field strength of 0.13 V/m and a current density of approx. 16 A/m2 between the
electrodes.
Therefore, it is recommended that initial stimulations should be carried out with an electrode
potential difference no greater than 0.245 V. In case it is deemed necessary, higher electrode
potentials and current densities can be used up to 7.35 V and 480 A/m2, but special care should be
taken.
Figure A.15 – A graphical summary of the various parameter ranges for the direct bioreactor
275
A.5 COMPARISON OF THE THREE BIOREACTOR
DESIGNS
Figure A.16– A graphical comparison of the three bioreactors. The light green triangle indicate the
range where the perpendicular bioreactor can be set depending on electrode position (“D” = the
distance between the top of the culture medium and the upper electrode assuming 2ml medium.) The
blue triangle (indicated by the blue arrow) is the range where the direct bioreactor can be used with
the current signal source.
It is difficult to compare the three bioreactors as they deliver quite different stimulations,
nonetheless each one has its own advantages and drawbacks:
The direct bioreactor is the most “efficient”: it produces the strongest electrical field at the
same electrode potential. However, it has a very limited useful range (indicated by the blue
triangle on Figure A.16.) with a maximum electrical field strength of 3.9 V/m. Arguably it
is not the electrical field that delivers the stimulus in the case of the direct bioreactor, but
the current density. The direct bioreactor is capable of generating current densities approx.
twenty times higher than the maximum reported in the literature. This device is therefore
276
more than capable of delivering an electrical current stimulus.
Stimulating using an electrical current has its drawbacks: Chemical changes in the culture
medium, reduced culture viability due to the presence of metal electrodes and necrosis are
known side effects of using the direct stimulation method.
Considering the limited range of the direct bioreactor and the present configuration of the
perpendicular bioreactor (H=5.8mm), out of the three the parallel field bioreactor is
capable of delivering the strongest maximum electrical field stimulation. Additionally, it
does so with perfect biocompatibility as its electrodes are completely isolated from the
culture environment.
The direct and parallel bioreactors deliver directional stimulation and can both be used to
align cells using galvanotaxis.
With its current configuration the perpendicular bioreactor delivers the weakest electrical
field strength at the same electrode potential difference. However, its electrodes can be set
to different positions. Lowering the electrodes (decreasing the distance between the culture
medium and the electrode) will significantly increase the electrical field strength generated
in the bioreactr. If the electrodes are set to touch the culture medium (D=0), the
perpendicular bioreactor will be similarly as “efficient” at generating an electrical field as
the direct bioreactor. As it is not limited by the maximum current output of the signal
source, the highest practical electrical field strength can be achieved thus with the
perpendicular bioreactor. Even higher field strengths can be achieved if the electrodes are
submerged into the culture medium.
On the other hand, placing the electrodes closer, in contact or submerged into the culture
medium raises biocompatibility issues. Although no electrical current will flow, the
presence of large metal electrodes in contact with the culture environment may reduce the
viability of the cell culture. Closer the electrodes are to the cell monolayer greater this risk
of this happening is.
277
Although the perpendicular bioreactor cannot be used to align cells (will not generate
galvanotaxis) due to the direction of the electrical field, it may be ideal to study the effects
of hyper- and de-polarisation of cell cultures.
In summary, if one requires an electrical field perpendicular to the cell monolayer, or the strongest
possible stimulation, the perpendicular bioreactor is recommended. If a parallel field is more
advantageous for the specific tissue engineering goal, then the parallel bioreactor should be used. If
an electrical current stimulus is required, it can be delivered using the direct bioreactor. Therefore it
can be concluded that all three bioreactors have their own advantages.
278
Engineering Drawings
279
The following engineering drawings are attached to this thesis:
1. Third generation electrode assembly
2. Bioreactor lid
3. Third generation bioreactor
4. Fourth generation upper electrode
5. Electrode bridge
6. Fourth generation bioreactor
7. Parallel field bioreactor
8. Direct bioreactor
9. Commercial bioreactor concept
10. Commercial bioreactor concept - Exploded view
280
281
282
283
284
285
286
287
288
289
290
Publications
291
The following publications are included with this thesis:
Balint R, Cassidy NJ, Cartmell SH. Electrical stimulation: a novel tool for tissue engineering.
Tissue Eng. Part B Rev., 19, 48, 2013
Balint R, Cassidy NJ, Cartmell SH. Conductive polymers: Towards a smart biomaterial for tissue
engineering. Acta Biomaterialia, 10, 2341, 2014
