Download - Spiral - Imperial College London

Copy and paste this link to your website, so they can see this document directly without any plugins.



Keywords

cells, were, al.,, cell, with, that, clones, oxygen, MSCs, from, bone, marrow, stem, which, BALB/c, differentiation, BALB/b, also, have, This, tension, into, their, cultured, genes, Figure, growth, population, this, studies

Transcript

Isolation and Characterisation of
Mesenchymal Stem Cells
Rasilaben Jethalal Vaghjiani
A thesis submitted for the Doctor of Philosophy to Imperial
College of Science, Technology, and Medicine
September 2008
The Kennedy Institute of Rheumatology
Imperial College London
1
Abstract
MSCs have great therapeutic potential and are currently used in various clinical trials.
However, the extremely low frequency of MSCs and the absence of a known cell-specific
marker have made their purification and identification a highly challenging goal. Our
hypothesis is that identification of a MSC specific cell surface marker would facilitate
isolation of a pure population of MSCs, which in clinical studies may enhance the
subsequent regenerative effect in comparison to a heterogeneous MSC population. Our
aim therefore was to attempt to identify such MSC-specific markers using a
transcriptomics approach. Individual clones were isolated after seeding bone marrow
mononuclear cells from BALB/b and BALB/c mice in 5% oxygen tension. All clones had
the ability to differentiate into adipocytes, chondrocytes, and osteoblasts, and to selfrenew, and were therefore functionally characterised as MSCs. All murine MSC clones
consistently expressed very high levels of Sca-1. Cytogenetic analysis of clones revealed
an abnormal karyotype of 69 chromosomes and transplantation of cells into immunocompromised SCID mice revealed no evidence of tumour formation after 7 months
indicating that cells were not malignant. A rigorous genome-wide supervised microarray
analysis revealed six genes were differentially expressed on the MSC clones in
comparison to various controls - STEAP1, STEAP2, ly6f, versican, vitamin D receptor,
and H2-M9. Two-dimensional hierarchical clustering of 28 different arrays revealed both
STEAP genes and vitamin D receptor also had a similar expression pattern to Sca-1. Thus
STEAP1 and STEAP2 were the only two cell membrane protein encoding genes
identified by both analysis methods. Importantly, flow cytometry analysis revealed
STEAP2 was differentially expressed in normal diploid multipotent human bone marrow
stromal cells (hBMSCs) compared to fibroblasts and freshly isolated bone marrow cells.
Furthermore, western blot analysis revealed STEAP1 was significantly expressed in
hBMSCs, but not by human fibroblasts or human chondrocytes. STEAP1 depletion by
RNA interference resulted in decreased cell adherence to tissue culture plastic. Results
suggest STEAP1 and STEAP2 may be novel MSC markers in murine and in human cells.
Further work is needed to elucidate their role in MSCs and to establish their usefulness as
potential cell-specific markers.
2
Table of Contents
Title Page….……………………..………………….…...……………….………….…1
Abstract ............................................................................................................................... 2
Table of Contents................................................................................................................ 3
List of Tables ...................................................................................................................... 7
List of Figures ..................................................................................................................... 8
Declaration........................................................................................................................ 10
Acknowledgements........................................................................................................... 11
Dedication ......................................................................................................................... 12
Abbreviations.................................................................................................................... 13
CHAPTER 1 Introduction…………………………………………………………….17
1.1 Adult Stem Cells ......................................................................................................... 18
1.1.1 Epidermal stem cells ........................................................................................ 18
1.1.2 Intestinal stem cells.......................................................................................... 18
1.1.3 Haematopoietic stem cells ............................................................................... 19
1.1.4 Mesenchymal stem cells .................................................................................. 20
1.1.5 Adult stem cells vs. Embryonic stem cells ...................................................... 20
1.1.6 Inducible pluripotent stem cells ....................................................................... 21
1.2 History of MSCs ........................................................................................................ 22
1.3 Concept of the stem cell niche ................................................................................... 23
1.3.1 The MSC niche in the bone marrow............................................................... 23
1.3.2 Tissue specific MSC niche.............................................................................. 24
1.3.3 The perivascular MSC niche............................................................................ 25
1.4 In vitro stem cell characteristics of MSCs ................................................................. 25
1.4.1 Self- renewal of MSCs.................................................................................... 25
1.4.2 Differentiation capacity of MSCs ................................................................... 26
1.4.2.1 Osteogenesis ................................................................................................. 26
1.4.2.2 Chondrogenesis............................................................................................. 27
1.4.2.3 Adipogenesis................................................................................................. 27
1.4.2.4 Alternative differentiation............................................................................. 27
1.5 In vivo differentiation of MSCs ................................................................................. 28
1.6 Homing ability of MSCs............................................................................................. 29
1.7 Immunosuppressive activity of MSCs ....................................................................... 30
1.8 Clinical studies........................................................................................................... 30
1.8.1 Orthopaedic studies.......................................................................................... 30
1.8.2 Graft-Versus-Host-Disease ............................................................................. 31
1.8.3 Genetic disorders ............................................................................................. 32
1.8.4 Gene therapy .................................................................................................... 33
1.8.5 Conclusion and future outlook for clinical studies .......................................... 34
1.9 Isolation of mesenchymal stem cells ......................................................................... 34
1.10 Surface markers ........................................................................................................ 35
1.10.1 STRO-1.......................................................................................................... 35
3
1.10.2 Other MSC candidate markers....................................................................... 36
1.10.3 MSC negative ‘markers’ ................................................................................ 37
1.11 MSCs, BMSCs, or MAPCs?.................................................................................... 37
1.12 Identification of MSCs............................................................................................. 39
1.12.1. Gene profiling ............................................................................................... 39
1.12.1. Microarray vs. Proteomics ............................................................................ 40
1.13 Cell culture and oxygen tension................................................................................ 41
1.13.1 Effect of oxygen tension on stem cells ......................................................... 42
1.14 Aims of present study ............................................................................................... 43
CHAPTER 2 Materials and Methods ………………………………………………..45
2.1 Reagents..................................................................................................................... 46
2.1.1 . Tissue and cell culture reagents ..................................................................... 46
2.1.2. SDS PAGE and western blotting reagents...................................................... 47
2.1.3 Molecular biology reagents.............................................................................. 47
2.2 Solutions ..................................................................................................................... 47
2.2.1 General buffers................................................................................................. 47
2.2.2. Protein extraction buffers................................................................................ 48
2.2.3 Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE)
buffers ....................................................................................................................... 48
2.2.4. Western blot buffers........................................................................................ 49
2.2.5. Flow cytometry buffers................................................................................... 49
2.3 Tissue culture ............................................................................................................. 49
2.3.1. Isolation and expansion of murine clones....................................................... 49
2.3.1.1 Trypsinisation of clones................................................................................ 50
2.3.2. Isolation of murine BMMNCs and bone marrow adherent cells (BMACs)... 50
2.3.3. Isolation of mouse embryonic fibroblasts (MEFs) ......................................... 51
2.3.4 Culture of human bone marrow stromal cells (hBMSCs) ............................... 51
2.3.5 Culture of neonatal human dermal fibroblasts (NHDFs)............................... 52
2.4 Adipogenic differentiation assay ............................................................................... 52
2.5 Osteogenic differentiation assay ................................................................................ 53
2.5.1 Alizarin Red S Staining ................................................................................... 53
2.6 Chondrogenic differentiation assay ........................................................................... 53
2.6.1 Type II collagen immunostaining .................................................................... 54
2.7 Colony forming unit-fibroblast (CFU-f) assay .......................................................... 55
2.8 Determination of population doublings of clonally-derived cultures ........................ 55
2.9 Biochemical Techniques............................................................................................ 55
2.9.1. Preparation of whole cell protein extracts for immunoblotting ...................... 55
2.9.2. Bradford assay for protein quantitation .......................................................... 56
2.9.3. SDS- PAGE .................................................................................................... 56
2.9.2 Western blotting............................................................................................... 57
2.10 Fluorescence-activated cell sorting analysis............................................................. 58
2.11 Telomere length assay.............................................................................................. 61
2.12 Karyotype analysis................................................................................................... 61
2.13 Microarray analysis.................................................................................................. 62
2.14 RNA extraction, reverse transcription and Real Time PCR .................................... 63
4
2.15 Transplantation of clonally-derived cells................................................................. 63
2.16 Histological analysis of murine tissues.................................................................... 64
2.16.1 Masson Trichrome staining........................................................................... 64
2.16.2 Haematoxylin and Eosin staining ................................................................. 64
2.17 RNA interference experiments ................................................................................ 65
CHAPTER 3 Isolation and Verification of Stem Cell Status of Clones…...………..66
3.1 Introduction................................................................................................................ 67
3.2 Isolation of clones ...................................................................................................... 67
3.3 Cell doubling time of clonal cells .............................................................................. 69
3.4 Determining the multipotency of clonal cells............................................................. 69
3.4.1 Adipogenic differentiation ............................................................................... 69
3.4.2 Chondrogenic differentiation .......................................................................... 78
3.4.3 Osteogenic differentiation............................................................................... 78
3.5 Colony forming ability of clones. .............................................................................. 82
3.6 Clonal cells have the ability to self-renew................................................................. 82
3.7 Discussion .................................................................................................................. 89
CHAPTER 4 Characterisation of MSC Clones……………………………….……102
4.1 Introduction............................................................................................................... 103
4.2 Surface phenotype analysis of BALB/b clones ....................................................... 103
4.3 Surface phenotype analysis of BALB/c clones....................................................... 107
4.4 Growth rate of clones............................................................................................... 107
4.5..Effect of oxygen tension on the growth rate of the clones ...................................... 111
4.6 MSC clones express telomerase and maintain telomere length............................... 111
4.7 Clones are cytogenetically abnormal ....................................................................... 114
4.6 Clonal cells are not malignant.................................................................................. 120
4.7 Despite an abnormal karyotype, clones display regulated differentiation............... 120
4.10 Discussion ............................................................................................................... 126
CHAPTER 5 Molecular Signature of MSC Clones ………………………………..138
5.1 Introduction.............................................................................................................. 139
5.2 Microarray design .................................................................................................... 139
5.3 Cell culture and microarray experiments................................................................. 141
5.4 Principal component analysis .................................................................................. 141
5.5 Unsupervised two-dimensional cluster analysis of microarray data........................ 141
5.6 Supervised analysis of microarray data .................................................................... 147
5.6.1 Supervised analysis of microarray data for BALB/b clones......................... 150
5.6.2 Supervised analysis of microarray data for BALB/c clones......................... 153
5.7 Supervised analysis yields six MSC-specific candidate genes ................................ 157
5.8 STEAP1 and -2 cluster with Sca-1 .......................................................................... 157
5.9 Validation of microarray results ............................................................................... 165
5
5.10 Discussion .............................................................................................................. 165
CHAPTER 6 Human Bone Marrow Stromal Cells express STEAP1 and -2…......164
6.1 Introduction.............................................................................................................. 179
6.2 Isolation of human bone marrow stromal cells........................................................ 179
6.3 Growth rate of hBMSC3.......................................................................................... 180
6.4 Human bone marrow stromal cells are multipotent................................................. 180
6.5 Surface antigen profile of human BMSCs ............................................................... 184
6.6 Cytogenetic analysis of human BMSCs .................................................................. 184
6.7 Human BMSCs express STEAP1............................................................................ 191
6.8 Human BMSCs express STEAP2............................................................................ 195
6.9 STEAP1 depletion in human BMSCs...................................................................... 195
6.10 Discussion .............................................................................................................. 198
CHAPTER 7 Conclusions and Future Work……………………….………………209
7.1 Conclusions and future work ................................................................................... 210
7.1.1 Conclusions................................................................................................... 210
7.1.2 Future work................................................................................................... 212
CHAPTER 8 Bibliography……………………….…………………….……………209
8.1 Bibliography ............................................................................................................ 210
6
List of Tables
CHAPTER 2 Materials and Methods
Table 2.1 Murine antibodies for FACS analysis 59
Table 2.2 Human antibodies for FACS analysis 60
CHAPTER 3 Isolation and Verification of Stem Cell Status of Clones
Table 3.1 BALB/b1 clones 70
Table 3.2 BALB/b2 clones 71
Table 3.3 BALB/b3 clones 72
Table 3.4 BALB/c clones 73
Table 3.5 Clones are multipotent 85
CHAPTER 4 Characterisation of MSC Clones
Table 4.1 Surface antigens examined by flow cytometry for BALB/b
and BALB/c clones 104
Table 4.2 Flow cytometry characterisation of BALB/b clonal cells 106
Table 4.3 Flow cytometry characterisation of BALB/c clonal cells 109
Table 4.4 Clones have an abnormal karyotype 118
Table 4.5 Clones display regulated differentiation 123
Table 4.6 Sca-1 expression decreases upon adipogenic differentiation of
MSC clones 124
Table 4.7 Adipocyte clones are distinct to the undifferentiated clones 125
CHAPTER 5 Molecular Signature of MSC Clones
Table 5.1 BALB/b cell membrane (associated) genes 151
Table 5.2 BALB/b secreted proteins 151
Table 5.3 BALB/b cell cycle associated genes 152
Table 5.4 BALB/b nuclear proteins 152
Table 5.5 BALB/b intracellular genes 154
Table 5.6 BALB/c cell membrane (associated) genes 154
Table 5.7 BALB/c secreted proteins 155
Table 5.8 BALB/c nuclear proteins 155
Table 5.9 BALB/c intracellular proteins 156
Table 5.10 Cell membrane (associated) genes clustering with Sca-1 161
Table 5.11 Genes encoding for secreted proteins clustering with Sca-1 162
Table 5.12 Cell cycle associated genes clustering with Sca-1 163
Table 5.13 Validation of Microarray Data by Real Time PCR for BALB/b and
BALB/c data sets 168
CHAPTER 6 Human Bone Marrow Stromal Cells express STEAP1 and -2
Table 6.1 Human BMSCs are multipotent 183
Table 6.2 Surface antigens examined by flow cytometry for human BMSCs 185
Table 6.3 Surface phenotype of human BMSCs 188
7
List of Figures
CHAPTER 3 Isolation and Verification of Stem Cell Status of Clones
Figure 3.1 Diagram of clonal isolation procedure 68
Figure 3.2 Morphology of BALB/b clones 74
Figure 3.3 Morphology of BALB/c clones 75
Figure 3.4 Adipogenic differentiation of BALB/b clones 76
Figure 3.5 Adipogenic differentiation of BALB/c clones 77
Figure 3.6 Chondrogenic differentiation of BALB/b clones 79
Figure 3.7 Chondrogenic differentiation of BALB/c clones 80
Figure 3.8 Alkaline phosphatase staining of BALB/b1 clone 37 81
Figure 3.9 Osteogenic differentiation of BALB/b clones 83
Figure 3.10 Osteogenic differentiation of BALB/c clones 84
Figure 3.11 Clones have the ability to form colonies 86
Figure 3.12 Clones maintain differentiation capacity with passage 87
Figure 3.13 Clones maintain morphology with passage 88
Figure 3.14 MSC clones exposed to low oxygen tension express HIF-1α 96
CHAPTER 4 Characterisation of MSC Clones
Figure 4.1 Surface antigen profile of BALB/b1 clone 37 105
Figure 4.2 Surface antigen profile of BALB/c clone 35 108
Figure 4.3 BALB/b and BALB/c clones showed extensive proliferation 110
capacity.
Figure 4.4 Effect of oxygen tension on the growth rate of BALB/b3 clone 33 112
Figure 4.5 Effect of oxygen tension on the growth rate of BALB/c clone 35 113
Figure 4.6 Clones express high levels of telomerase 115
Figure 4.7 Passaged clones maintain telomere length 116
Figure 4.8 Clones have an abnormal karyotype 117
Figure 4.9 Bone marrow adherent cells display chromosomal aneuploidy 119
Figure 4.10 Clonal cells are not malignant 121
CHAPTER 5 Molecular signature of MSC Clones
Figure 5.1 Microarray experimental design used to characterise the molecular
signature of MSC clones derived for BALB/b and BALB/c mice 140
Figure 5.2 Principal component analysis of microarray data 142
Figure 5.3 Dendrogram showing Thy1/CD90 clustering 143
Figure 5.4 Dendrogram showing PPARγ clustering 144
Figure 5.5 Dendrogram showing Csf1r clustering 145
Figure 5.6 Dendrogram showing BMMNC highly expressing genes 146
Figure 5.7 Schematic of microarray analysis to find genes specific to the MSC
undifferentiated clone 148
Figure 5.8 Molecular signature of MSC clones 158
Figure 5.9 Dendrogram showing genes clustering with Sca-1 159
8
Figure 5.10 Genes highly expressed on MSC clones 164
CHAPTER 6 Human Bone Marrow Stromal Cells express STEAP1 and -2
Figure 6.1 Growth rate of human BMSCs 181
Figure 6.2 Human BMSCs have multilineage differentiation potential 182
Figure 6.3 Surface antigen profile of human bone marrow stromal cells 186
Figure 6.4 Human BMSCs have a normal karyotype 189
Figure 6.5 Cytogenetic analysis of human BMSC2 190
Figure 6.6 Human bone marrow stromal cells (hBMSC1) express STEAP1 192
Figure 6.7 Human BMSC1 and 2 express STEAP1 193
Figure 6.8 Human BMSC1-3 express STEAP1 194
Figure 6.9 Human bone marrow stromal cells express STEAP2 195
Figure 6.10 STEAP2 expression is retained in adipocyte differentiated human
BMSCs 197
Figure 6.11 Validation of STEAP depletion by real-time PCR 199
Figure 6.12 STEAP1 depletion in human BMSCs 200
9
Declaration
This dissertation is the result of my own work. All collaborations have been
acknowledged in the appropriate place within the text.
10
Acknowledgements
I am very grateful to my supervisor Dr Chris Murphy for his continued guidance and
support during my work, and also for always encouraging me to become a better scientist.
I would also like to thank Professor Jeremy Saklatvala for his very useful advice
throughout my PhD. Special thanks also to members of my group Christine Hopfgarten,
Sonia Talma, and Dr. Jérome Lafont for their continued support and friendship.
I would also like to thank the many people at the Kennedy Institute, especially in the Cell
Signal lab who have helped me during my study and made my time here so memorable. I
am also vey thankful to members of the Matrix Biology lab, particularly for the many
inspired, philosophical, and sometimes bizarre conversations in tissue culture. They made
tissue culture even more enjoyable.
Importantly, I would like to thank my family, especially my brother Girish, for their
continued support in all my endeavours.
11
Dedication
In loving memory of Lalji Vishram…
…mara bapuji
12
Abbreviations
2D Two dimensional
A 2-P Ascorbic acid 2-phosphate
ABS Adult bovine serum
AKT V-akt murine thymoma viral oncogene homolog 1
AMC Adipose-derived mesenchymal stem cell
B/b BALB/b
B/c BALB/c
BMAC Bone marrow adherent cell
BMMNC Bone marrow mononuclear cell
BMP2 Bone morphogenetic protein-2
BMPR1a bone morphogenetic protein receptor, type IA
BMSC Bone marrow stromal cell
BMT Bone marrow transplantation
Ccl7 Chemokine (C-C motif) ligand 7
CD Cluster of Differentiation
Cdk4 Cyclin dependent kinase 4
Cdkn2a Cyclin-dependent kinase inhibitor 2a
CEBP CCAAT/enhancer binding protein
CF Cystic fibrosis
CFTR Cystic fibrosis transmembrane receptor
CFU-f Colony forming unit- fibroblast
cMET c met proto-oncogene
CNS Central nervous system
Crabp2 Cellular retinoic acid binding protein II
Csfr1 Colony stimulating factor receptor 1
CSPG-2 Chondroitin sulphate proteoglycan core protein-2
Cxcl1 Chemokine (C-X-C motif) ligand 1
Cxcl5 Chemokine (C-X-C motif) ligand 5
DAVID Database for Annotation, Visualization and Integrated Discovery
DC Differentiated clone
DCAMKL1 Doublecortin-like kinase 1
DFO Desferrioxamine
DMEM Dulbecco’s modified Eagle’s medium
DNA Deoxyribonucleic acid
DU145 Human prostate cancer cell line
ECM Extracellular matrix
EDTA Ethylenediaminetetraacetic acid
EGF Epidermal growth factor
13
ESCs Embryonic stem cells
FACS Flow activated cell sorting
FCS Fetal calf serum
FGF2 Fibroblast growth factor
FGF7 Fibroblast growth factor-7
FITC Fluorescein Isothiocyanate
FNO F420H2:NADP+ oxidoreductase
Foxa1 Forkhead box A1
Gfzf3 Glis family zinc finger 3
β-GP β -glycero phosphate
GLUT4 Glucose transporter 4
GPI-AP Glycosyl phosphatidylinositol-anchored protein
GVHD Graft versus host disease
H2-M9 Histocompatability 2, m region locus 9
hBMMNC Human bone marrow mononuclear cell
hBMSC Human bone marrow stromal cell
HBSS Hank’s Balance Salt Solution
HGF Hepatocyte growth factor
HIF1a Hypoxia inducible factor-1 alpha
HIF2a Hypoxia inducible factor-2 alpha
hMSC-TERT Human mesenchymal stem cell- Telomerase reverse transcriptase
HS68 human newborn foreskin fibroblasts
HSC Hematopoietic stem cell
HSCT Hematopoietic stem cell transplantation
IBMX 3- isobutyl-1-methyl-xanthine
IGFBP6 Insulin-like growth factor binding protein 6
IGFBP6 insulin growth factor binding protein-6
IL1Ri Interleukin 1 receptor i
IL-2Rb Interleukin 2 receptor b
ISCT International Society for Cellular Therapy
ITGA5 Integrin alpha 5
ITS+1 Insulin transferrin selenium
JAK1 Janus activated kinase 1
LIF Leukaemia inhibitory factor
LNCaP Androgen sensitive human prostate adenocarcinoma cell line
LNGFR Low affinity growth factor receptor
M113 Melanoma cell line
Maged2 Melanoma antigen family D, 2
MALDI-TOF Matrix-assisted laser desorption/ionization- time-of-flight
MAPC Multipotent adult progenitor cells
MCP-3 Monocyte chemotactic protein-3
14
MEF Mouse embryonic fibroblasts
α-MEM alpha-minimal essential media
MLD Metachromatic Leukodystrophy
MSC Mesenchymal stem cell
Nanog Nanog homeobox
NCAM-1 Neural cell adhesion molecule -1
NK-Cell Natural killer- cells
NOD Non-obese diabetic
OCT4 POU domain, class 5, transcription factor 1
OI Osteogenesis Imperfecta
OM Osteogenesis media
p Passage
PE Phycoerythrin
P/S Penicillin/Streptomycin
PBS Phosphate buffered saline
PCA Principal component analysis
PCR Polymerase chain reaction
PDGF-BB Platelet derived growth factor PPAR-γ Peroxisome proliferator-activated receptor- γ
PTEN Phosphatase and tensin homolog
RIPA Radio-immunoprecipitation assay
RNA Ribonucleic acid
ROS Reactive oxygen species
RPLPO Human acidic ribosomal phosphoprotein PO
RT-PCR Reverse transcription-polymerase chain reaction
Runx2 runt-related transcription factor 2
Sca-1 Stem cell antigen -1
SCID Severe combined immunodeficiency
SD Standard deviation
SDS-PAGE Sodium dodecyl sulphate-polyacrylamide gel electrophoresis
SOX4 SRY (sex determining region Y)-box 4
SOX9 SRY (sex determining region Y)-box 9
spl split homolog of drosphila E
STAT5a/b Signal transducer and activator of transcription 5a/b
STEAP1 Six-transmembrane epithelial antigen of the prostate1
STEAP2 Six-transmembrane epithelial antigen of the prostate2
TBS Tris buffered saline
TERC Template RNA component
Tf Transferrin
TfR1 Transferrin receptor 1
TGF-b Transforming growth factor- beta
15
TM4SF1 Transmembrane 4 superfamily member 1
UC Undifferentiated clone
ULB Urea Lysis Buffer
USSC Unrestricted somatic stem cell
Vdr Vitamin D receptor
VEGF Vascular endothelial growth factor
α-MEM alpha-minimal essential media
β-GP β -glycero phosphate
16
Chapter 1
Introduction
17
1.1 Adult Stem Cells
Stem cells are defined by two basic properties: the ability to self-renew and the capacity
to differentiate into one or more cell types. Adult stem cells are multipotent in their
differentiation capacity and are found in many tissues of the body (Wagers et al., 2002).
They are thought to serve as reservoirs of reparative cells that maintain cell turnover, and
during injury or disease migrate to the site of tissue damage and participate in the repair
process.
1.1.1 Epidermal stem cells
The outer layer of the epidermis is continually being shed, thus the lifetime of a skin cell
is about one month depending on the region (Watt et al., 2006). Skin consists of two main
parts: the outer epidermal layer and the connective tissue layer beneath. This contains the
collagen-rich dermis and has an underlying fatty subcutaneous layer. The constant
epidermal cell turnover can only be maintained if there is a population of self-renewing
cells that can remain undifferentiated, but continuously divide via amplifying cells, which
go on to terminally differentiate into distinct epidermal lineages. These lineages include
the intermolecular epidermis, hair follicles, and sebaceous glands (Watt and Hogan,
2000). Epidermal stem cells are typically localised in specialised areas such as the hair
follicle, at the base and or at the bulge (Panteleyev et al., 2001). They are well
characterised and have been demonstrated to express high levels of β1 actin and α6
integrin, and low levels of CD71, the transferrin receptor (Jones and Watt, 1993) and
more recently found to express CD200 (Watt et al., 2006). This has enabled a good
understanding of the lineage pathway for each differentiated cell type.
1.1.2 Intestinal stem cells
The epithelial lining of the small intestine renews itself faster than any other tissue, with a
turnover time of a week or less. The lining of the small intestine is covered with many
different cell types each with a specialised function. This includes (1) absorptive cells,
which are densely packed with microvilli for absorption of nutrients (2) goblet cells,
18
secrete mucus (3) paneth cells, which play a role in the innate immune system (4)
enteroendocrine cells, which secrete serotonin and various peptide hormones (Wright,
2000). There are over 15 subtypes of paneth cells (Ganz, 2000). This constant turnover
process is maintained by intestinal stem cells located in the crypts of the epithelium that
descend into the underlying connective tissue. The stem cell divides to produce more
intermediate cells which go through several rounds of division, with each division the cell
travels towards the basal epithelium and becomes more committed to a specific lineage.
Once the cell has reached the epithelium it has terminally differentiated into one of the
above cell types. To date, there are no known definitive markers to identify intestinal
stem cells. They are known to express Musashi (Kayahara et al., 2003), and other
potential markers are BMPR1a (a phospho-PTEN), DCAMKL1, Eph receptors and
integrins (Montgomery and Breault, 2008).
1.1.3 Haematopoietic stem cells
The haematopoietic stem cell (HSC) is the best characterised adult stem cell. HSCs have
the potential to give rise to all the blood cells, each with very different functions from
transporting oxygen around the body to the production of antibodies. In the adult, the
majority of the haematopoietic process takes place in the highly dynamic environment of
the bone marrow where HSCs are localised. The differentiated cells then exit into the
bloodstream and are transported to where they are needed. HSCs can differentiate into
myeloid lineages (macrophages, monocytes, neutrophils, basophils, eosinophils, dendritic
cells, erythrocytes, megakaryocytes) and lymphoid lineages (T-cells, B-cells, NK-cells).
Despite the large number of differentiated cell types, each lineage pathway has been well
documented for HSCs (Orkin, 2000) (Weissman et al., 2001).
This is because HSC markers are quite well defined for each population of HSC. Human
HSCs are CD34+, CD59+, CD90+, CD38low/-, C-kit+, lin-. Whilst mouse HSCs are
CD34low/-, Sca-1+, CD90low/+, CD38+, C-Kit+, lin- (Wognum et al., 2003). However, it
should be noted HSC populations consist of subsets with varying differentiation
potentials, and this list does not cover the surface phenotype of them all.
19
1.1.4 Mesenchymal stem cells
The bone marrow also contains a non-haematopoietic stem cell called a mesenchymal
stem cell (MSC) that can differentiate into classical mesenchymal lineages adipocytes,
chondrocytes and osteoblasts, at least in vitro (Pittenger et al., 1999). Before the term
‘MSC’ was coined this cell type was thought to function predominantly as a stromal
supportive cell which played an important role in the haematopoietic process (Dexter et
al., 1977), (Calvi et al., 2003). Friedenstein and colleagues provided the earliest evidence
for the existence of its stem cell properties by demonstrating that bone marrow derived
adherent cells were capable of osteogenesis (Friedenstein, 1961). These studies have been
supported by in vivo work (Kuznetsov et al., 1997; Lennon et al., 2001). Although its
absolute in vivo function has not been clarified, it is logical to think that there is such a
connective tissue progenitor cell involved in tissue maintenance of the skeletal system.
However, the lack of a known MSC specific marker has hindered better understanding of
this stem cell.
In summary, adult stem cells are vital for normal tissue maintenance and for repair after
injury for the duration of an organism’s life. It is not surprising that the better understood
stem cells are those whose markers have been identified. This enables isolation and
subsequent studies on a pure population of stem cells and yields relevant results,
compared to stem cells for which no specific marker has been identified.
1.1.5 Adult stem cells vs. Embryonic stem cells
Advantages of using adult stem cells over embryonic stem cells (ESCs) for repair
therapies include relatively stable differentiated phenotypes with no teratoma formation
and the possibility of using autologous cells directly, thus avoiding immune response
concerns. While ESCs may represent future promise for treating certain intractable
diseases; for tissues where there are known adult stem cells available, the latter would
seem a more promising and feasible approach. In contrast to the ESC, one type of adult
stem cell currently used in clinical studies for a increasing number of diseases is the
mesenchymal stem cell (MSC) (Tae et al., 2006). The multipotent nature of MSCs and
their ability to proliferate significantly in culture has rendered them a potentially useful
20
candidate for tissue repair and gene therapy (Bianco et al., 2001; Caplan and Bruder,
2001). In addition, their potential immunosuppressive properties may increase their
therapeutic application (Sotiropoulou and Papamichail, 2007).
1.1.6 Inducible pluripotent stem cells
Recent studies have reported a type of stem cell that has the pluripotency of the
embryonic stem cell, but without the ethical dilemmas (Okita et al., 2007) (Nakagawa et
al., 2007). So-called induced pluripotent stem cells (iPSCs), these cells are derived from
non-pluripotent cells such as adult skin fibroblasts, by inducing them to express stem
cell-associated genes. The Okita group cultured human skin fibroblasts and transduced
them with a retroviral vector containing 4 stem cell genes- OCT3/4, SOX2, KLF4, and cMyc (Okita et al., 2007). The subsequent cells behaved similarly to ES cells. They had a
similar cell doubling time, expressed telomerase and expressed many cell surface markers
and genes characteristic of ES cells. In addition, the iPSCs could differentiate into cells of
the three primary germ layers (ectoderm, endoderm, and mesoderm) in vivo and in vitro.
In a separate study, Yu and colleagues also reported very similar results (Yu et al., 2007).
They also transduced fibroblasts with 4 stem cell genes, except this time with NANOG
and LIN28 instead of KLF4 and c-myc. The resultant iPSCs exhibited similar ES cell
properties to those mentioned above. In a later study Nakagawa reported that iPSCs could
be formed from transduction of both mouse and human fibroblasts using only 3 genes
(Nakagawa et al., 2007). This was an important study as it showed that the gene that was
omitted was the oncogene c-myc which had made iPS cells prone to tumour formation
(Okita et al., 2007).
This is an exciting research area, however much work needs to be done before the cells
can used in a clinical setting as there are concerns over the safety of using retrovirally
transduced cells.
21
1.2 History of MSCs
The bone marrow has long been known to be the source of haematopoietic stem cells
(HSCs). In 1878 the German pathologist Clonhein was the first to suggest that nonhaematopoietic cells were also present in the bone marrow (Chamberlain et al., 2007)
However, it took a further 90 years before Friedenstein and colleagues demonstrated that
the bone marrow contains cells which can differentiate into cells of the mesenchyme
(Friedenstein et al., 1968). These investigators seeded murine whole bone marrow
aspirate on plastic culture dishes. The non-adherent cells (which contain most of the
haematopoietic cells) were removed after 4 hours. The remaining cells were
heterogeneous in terms of morphology. However, they report the presence of spindle
shaped cells that formed little clusters of 2-4 cells, which after a lag period of 2-4 days
began to proliferate rapidly. The cells became more homogeneous in morphology after
they had been passaged several times, i.e. the spindle shaped cells presumably out-grew
the other cell types. They also found that these cells could differentiate into bone and
cartilage. Furthermore, this differentiation characteristic was demonstrated both in vitro
and in vivo (Friedenstein, 1961; Friedenstein et al., 1966). In the next decade these results
were confirmed by others (Ashton et al., 1980; Bab et al., 1986; Castro-Malaspina et al.,
1980). In addition, the latter group reported that these cells were rare, constituting 1 in
100 000 nucleated cells from the bone marrow. Subsequent studies found these cells had
the ability to differentiate into the adipocytes and myoblasts (Dennis et al., 1999;
Wakitani et al., 1995). Thus, these plastic adherent bone marrow cells that had the
potential to differentiate into cells of the mesenchyme were called mesenchymal stem
cells (MSCs). This term was used by Caplan in the 1990s (Caplan, 1991), and early
studies by his group report that differentiation potential of MSCs diminished with
extensive passage (Nakahara et al., 1990). However, more recent studies report that
MSCs have a much broader differentiation potential (Jiang et al., 2002). Since then there
has been a dramatic increase in the number of studies involving MSCs. For instance,
between 1997 and 2007 there has been an over 40-fold increase in the number of pubmed
entries containing ‘mesenchymal stem cell’ (Wagner and Ho, 2007).
22
1.3 Concept of the stem cell niche
The stem cell ‘niche’ is a specialised microenvironment where the stem cells reside. It is
more than just the location of the stem cell as it has a multi-faceted role. It is a highly
dynamic environment containing different types of cells which interact with the stem cell
in many ways (Watt and Hogan, 2000), for example, the niche is important for
maintaining the stem cell in its quiescent state. In response to injury or during normal
tissue homeostasis the microenvironment can signal the stem cell to mobilise, self-renew
and differentiate into specific cell type(s). The niche also prevents depletion of the stem
cell and un-controlled proliferation; therefore, if the niche becomes dysregulated it may
also cause aberrant stem cell function. The stem cell receives many different types of
signals including cell-cell interactions via adhesion molecules and extracellular matrix, as
well as signals via cytokine and growth factors. The physiochemical nature of the niche
including the pH, calcium concentration, and oxygen tension is also important (MartinezAgosto et al., 2007).
Throughout adult life, the in vivo niche maintains the adult stem cell in an
undifferentiated state until required for normal tissue maintenance or repair in response to
injury. It is important for scientists to understand the many properties of the niche and be
able to replicate at least to some extent, the in vivo conditions in vitro. In addition, after a
number of population doublings during in vitro culture, the stem cell undergoes an
‘aging’ process. The proliferation rate and the differentiation potential of the stem cell
decreases and there is a morphology change. This is not ideal for regenerative medicine,
where a sufficient number and quality of stem cell needs to be grown before
transplantation to the patient. Current culturing conditions need to be improved so that
they better mimic the in vivo microenvironment thus enabling stem cell properties to be
better maintained.
1.3.1 The MSC niche in the bone marrow
Traditionally the main source from which MSCs are obtained is the bone marrow. The
bone marrow is primarily composed of haematopoietic cells, HSCs, adipocytes,
23
endothelial cells, fibroblasts, osteoprogenitors, and stromal cells (Castro-Malaspina and
Jhanwar, 1984; Wang and Wolf, 1990). Little is known about the in vivo MSC niche as
there is no known marker of MSCs that can be used to identify the cell in vivo. However,
there are a few theories. The HSC niche is slightly better understood, and since MSCs are
required for the regulation of haematopoietic process (Mitsiadis et al., 2007), it is likely
MSCs share a somewhat similar niche to HSCs. There are thought to be three HSC
niches within the bone marrow (Dazzi et al., 2006). The first is the ‘bony niche’ or
‘endosteal niche’ located at the trabeculae, and lined with osteoblasts (Wilson et al.,
2007). The second is the ‘stromal niche’ composed of reticular cells, fibroblasts,
endothelial cells, and macrophages. The third is the ‘adipocyte niche’ which occupies the
space within the bone marrow. MSCs are thought to be present largely in the stromal
niche as they produce osteoblasts, fibroblasts, and adipocytes which interact with the
HSC.
1.3.2 Tissue specific MSC niche
Another possibility to the one above is that MSCs or MSC-like cells are present in most
tissue types. This is based on the fact that cells with the in vitro characteristics of MSCs
have been isolated from many different tissues including: bone marrow, adipose
tissue(De Ugarte et al., 2003), peripheral blood (Kuznetsov et al., 2001), synovial
membrane (De Bari et al., 2001), human umbilical cord perivascular cells (Sarugaser et
al., 2005), skeletal muscle (Williams et al., 1999), deciduous teeth (Tuan et al., 2003),
trabecular bone (Noth et al., 2002) and amniotic fluid (De Coppi et al., 2007).
This could be due to the premise that there is only one source of MSCs, the bone marrow,
and when mobilised MSCs exit the niche and circulate in the blood to replenish cell
populations during tissue maintenance or injury. This could be achieved in two ways, (1)
the MSC may differentiate into the organ-specific cells, or (2) it may secrete cytokines to
facilitate tissue regeneration. In addition, the immunosuppressive properties of MSCs
may also be favourable to tissue repair.
Transdifferentiation is when a cell differentiates into a lineage outside its developmental
origin, for example when a MSC differentiates into a neural cell. There are some
24
investigators that say this could be due to cell fusion (Terada et al., 2002; Ying et al.,
2002). However, recent cell tracking studies have disputed this suggestion for different
cell types (Forbes et al., 2002). Since, undifferentiated MSC-like cells have been isolated
from various tissues the latter possibility may be more likely.
1.3.3 The perivascular MSC niche
In more recent studies, investigators have postulated that MSCs are the bone marrow
stromal supportive cells, also called marrow adventitial reticular cells or pericytes.
Furthermore, they suggest the perivascular niche which is the cell lining the outer surface
of blood vessels, is where the in vivo MSC is located. (da Silva Meirelles et al., 2008;
Jones and McGonagle, 2008; Short et al., 2003). Supporting this suggestion is a study
that showed MSCs sorted by STRO-1+CD106+ expressed the pericyte specific marker αsmooth muscle actin (Shi and Gronthos, 2003). However, cells with MSC characteristics
have been isolated from articular cartilage which is avascular (Barbero et al., 2003;
Dowthwaite et al., 2004) which contradicts the above possibility. Therefore, more work
needs to be performed in this area and significant progress can only be accomplished
once a MSC specific marker is identified.
1.4 In vitro stem cell characteristics of MSCs
1.4.1 Self- renewal of MSCs
The self-renewal potential of a stem cell is defined as the ability to generate identical
copies of themselves with the same multilineage potential as the mother cell. This can
occur during the entire lifetime of an organism. The in vitro self-renewal ability of MSCs
has been demonstrated. The cells have the ability to form CFU-f and to differentiate into
multiple lineages. However, studies have reported MSCs lose their differentiation
potential with passage in culture ((Gronthos et al., 2003; Muraglia et al., 2000; Pittenger
et al., 1999).
25
1.4.2 Differentiation capacity of MSCs
Differentiation is a series of events involved in the progression of a stem/progenitor cell
to a more specialised cell type, in order to perform a specific function. This process of
differentiation involves co-ordinated regulation. Genes specific to the mode of
differentiation are switched on and/or upregulated, whilst expression of stem cell genes
and genes for other lineages is repressed. Differentiation can result in a change of cell
size, morphology, metabolic activity, polarity, biomechanical properties, extracellular
matrix, and surface phenotype, as well as the way a cell responds to signals (Tsonis,
2004). Therefore, it is a tightly regulated process that occurs in a stepwise manner. This
way a limited number of stem cells can give rise to a highly expanded number of
differentiated cells. This could explain why the fraction of undifferentiated mesenchymal
stem cells is so small in the total bone marrow population. As mentioned above, many
studies have reported that MSCs have the capacity to differentiate into adipocytes,
chondrocytes, and osteoblasts (Tropel et al., 2004). This is the main functional criterion
by which MSCs are defined in vitro.
1.4.2.1 Osteogenesis
The classical method of osteogenic differentiation in vitro involves culturing MSCs in a
monolayer in the presence of ascorbic acid-2- phosphate, a source of phosphate such as
β- glycerophosphate and dexamethasone. Bone morphogenetic proteins (BMPs) can also
be added to induce osteogenesis. Although there are species differences, BMPs work
better in murine MSCs (Edgar et al., 2007) than in human MSCs (Diefenderfer et al.,
2003; Hanada et al., 1997). During culture the MSCs acquire an osteoblastic morphology
as they form aggregates or nodules. There is an increase in their expression of alkaline
phosphatase and the cells deposit a calcium-rich mineralised matrix which can be
visualised histochemically with either Alizarin Red or a Von Kossa stain. The gene
expression of MSCs differentiated into osteoblasts also changes; there is upregulation of
osteoblast genes such as bone sialoprotein, osteocalcin, Runx2, and osteopontin
(D'Ippolito et al., 2004).
26
1.4.2.2 Chondrogenesis
Chondrogenesis can be induced when MSCs are pelleted to form a micromass which is
cultured in a serum-free medium in the presence of TGF-β, dexamethasone, sodium
pyruvate, insulin-transferrin-selenium, and ascorbic acid-2-phosphate (Mackay et al.,
1998). Under these conditions the cells lose their spindle shape, begin to aggregate and
increase expression of cartilage-specific extracellular matrix components including
glycosaminoglycan, aggrecan, link protein, chondroitin sulphate and type II collagen.
Histochemically the presence of glycosaminoglycans and chondroitin sulphate can be
demonstrated with toluidine blue and alcian blue staining respectively. A more specific
method for demonstrating chondrogenesis is showing type II collagen staining via
immunohistochemisty. MSCs differentiated into chondrocytes show increased gene
expression levels of type II collagen, SOX9, and aggrecan (Kafienah et al., 2007; Tropel
et al., 2004).
1.4.2.3 Adipogenesis
To promote adipogenesis MSCs are cultured in a monolayer in a cocktail of
dexamethasone, insulin, 3- isobutyl-1-methyl-xanthine, and indomethacin. After one
week there is evidence of lipid filled droplets in the cells which stain red with Oil Red O.
Eventually the lipid droplets will merge and fill the cell. The cells express nuclear
receptor transcription factor peroxisome-activated receptor γ-2, fatty acid binding protein
aP2, and lipase (Tropel et al., 2004).
1.4.2.4 Alternative differentiation
MSCs also have the ability to differentiate into muscle cells. Rat MSCs exposed to 5azacytidine and amphotericin B have been reported to differentiate into myoblasts that
fuse into rhythmically beating myotubes (Wakitani et al., 1995). This has also been
demonstrated in mouse MSCs (Phinney et al., 1999). An increasing number of
27
investigators have reported neural differentiation of MSCs (Azizi et al., 1998; SanchezRamos et al., 2000). In addition, studies have also reported MSCs can be induced to
differentiate into tendon (Young et al., 1998), ligament, endothelial cells (Davani et al.,
2003) and insulin- expressing cells (D'Ippolito et al., 2004) when exposed to the
appropriate environmental cue(s).
1.5 In vivo differentiation of MSCs
It is difficult to have a true understanding of the absolute in vivo function of the MSC
since it is not possible to identify it in vivo. However, given its ability to differentiate into
the mesenchymal lineages, one would conjecture that that is at least one function. It is
also possible to speculate the prospective function of the MSC from in vivo studies.
Studies have shown that when MSCs are injected into animals, they home to various
tissues and engraft and differentiate there. For instance, Gojo and colleagues
demonstrated that murine MSCs differentiated into cardiomyocytes, endothelial cells,
smooth muscle cells and pericytes when injected directly into a healthy adult heart (Gojo
et al., 2003). They also contributed to the vasculature in the skeletal muscle and lung. In
another study, human MSCs were injected into fetal lamb early in gestation via in utero
transplantation (Liechty et al., 2000). The investigators report the MSCs differentiated
into adipocytes, bone marrow stromal cells, chondrocytes, cardiomyocytes, myocytes,
and thymic stroma. Jiang and colleagues have demonstrated a sub-population of murine
MSC, called multipotent adult progenitor cells (MAPCs) contribute to most, if not all,
somatic cell types when injected into an early blastocyst, (Jiang et al., 2002). MAPCs can
engraft into various tissues and differentiate not only into mesenchymal cells but cells of
mesodermal, neuroectodermal, and endodermal derived cells. On transplantation into a
non-irradiated host, MAPCs engraft and differentiate to haematopoietic lineages, in
addition to the epithelium of liver, lung and gut.
Rat MSCs have the ability to differentiate into bone and cartilage in vivo (Lennon et al.,
2001). Indeed the ability of MSCs to form bone has been extensively reported
(Friedenstein et al., 1966), (Kuznetsov et al., 1997). This is typically performed with
ceramic implants loaded with MSCs and implanted into an animal. Allogeneic canine
28
MSCs have been reported to form bone when implanted into canines with a segmental
defect in the femur (Bruder et al., 1998). Furthermore, the regenerated bone could be
assessed by histology and radiology.
1.6 Homing ability of MSCs
MSCs have the ability to home to sites of injury. Many studies have reported the
migration, engraftment and differentiation of MSCs into specific cell types at sites of
injury. In a mouse model of pulmonary fibrosis where mice were exposed to bleomycin,
murine MSCs migrated to the lung and adopted an epithelium-like phenotype (Ortiz et
al., 2003). In addition the cells reduced inflammation and collagen deposition in the lung
tissue, thus ameliorating its fibrotic effects. In another example, in rats with cerebral
ischemia, culture expanded human MSCs were injected intravenously and were found to
migrate to sites of brain injury (Li et al., 2002). The investigators report that there was
significant recovery of function in rats treated with MSCs in comparison to controls. This
was partly due to an increase in growth factors in the ischemic tissue, a decrease in
apoptosis in the lesion area, and proliferation of endogenous cells in the sub-ventricular
zone. Orlic and colleagues demonstrated the efficacy of MSCs to treat coronary heart
disease (Orlic et al., 2001). They reported that murine bone marrow cells injected into the
heart can generate de novo myocardium suggesting one in vivo function of the MSC may
be in wound repair.
Despite these promising results, there are a few issues that need to be resolved. One is
whether there is a difference in the level of MSCs in these injury/diseased sites compared
to non-affected tissues. Another is whether the level of biological improvement is directly
due to the MSC as it may be that the number of MSCs present post-transplantation is too
few to account for the significant regeneration. In fact, Caplan and colleagues have
postulated that the regenerative effect maybe due to the trophic properties of MSCs
(Caplan and Dennis, 2006), which include secreting cytokines to assist the repair and
regenerative process and their immunosuppressive properties as well.
29
1.7 Immunosuppressive activity of MSCs
It is not just the ability of MSCs to home, engraft and differentiate at sites of tissue
damage that could contribute to tissue repair. The immunomodulatory effect of MSCs is
now well documented (Aggarwal and Pittenger, 2005). MSCs secrete a wide range of
bioactive molecules that are immunoregulatory and also structure regenerative
microenvironments in areas of tissue injury (Caplan, 2007). Initial studies reported
human MSCs suppress T-cell activation and proliferation when exposed to alloantigens
and non-specific mitogens (Di Nicola et al., 2002; Krampera et al., 2003). A subsequent
study reported this suppressive effect is independent of the major histocompatability
complex (Le Blanc et al., 2003). Further studies have found that the immunodulatory
effect is extended to impairing the maturation and function of dendritic cells (Jiang et al.,
2005), as well as the proliferation and differentiation of B-cells (Corcione et al., 2006).
Thus the immunosuppressive effect of MSCs appears to be wide-ranging. This is an
advantageous property because it could be exploited therapeutically. For example, Osiris
Therapeutics have clinical trials in place that are utilizing the immunoregulatory capacity
of MSCs to treat graft-versus-host-disease (GVHD) and Crohn’s disease (Taupin, 2006).
1.8 Clinical studies
The multilineage differentiation potential of MSCs, their immunosuppressive activity,
and the fact that they can be easily isolated from the patient and expanded in culture
makes them very attractive candidates for regenerative medicine and tissue repair. MSCs
are currently being tested in clinical studies for their potential use in cell and gene
therapy for a number of diseases which have no known cure.
1.8.1 Orthopaedic studies
The ability of MSCs to differentiate into bone and cartilage makes them very attractive
candidates for orthopaedic medicine. In clinical studies MSCs were used to treat children
with Osteogenesis Imperfecta (OI), a genetic bone disorder involving defective
production of type I collagen. There is no cure for OI and the only drugs that have shown
30
any therapeutic potential are bisphosphonates, however the effect is only partial (Bembi
et al., 1997). In a clinical trial, two infusions of cultured allogeneic MSCs were given to
children who had already undergone standard bone marrow transplantation for severe OI
(Horwitz et al., 1999), (Horwitz et al., 2001), (Horwitz et al., 2002). The preliminary
results have been promising, as there was skeletal engraftment of the transplanted MSCs
as well as osteogenic differentiation. There was also an increase in total bone mineral
density, and an accelerated growth during the first 6 months post-infusion. The results
suggest MSCs offer a feasible cell source for the therapy of bone disorders.
In a separate study the effectiveness of autologous MSC transplantation in the repair of
articular cartilage defects in the patellae of two adults was examined (Wakitani et al.,
2004). Bone marrow cultured MSCs were embedded in a collagen gel and transplanted
into the articular cartilage defect. Six months post-transplantation the clinical symptoms
of the disease, pain and walking ability, had improved significantly. Furthermore, this
improvement was maintained five years and nine months in one case, and 4 years in the
other. Results suggest MSC transplantation is a promising approach for promoting
articular cartilage repair. However, the quality of the repair is still in question. For
example, the investigators found via arthroscopy that the defect was repaired with
fibrocartilage, which does not have the same biomechanical qualities of articular
cartilage.
1.8.2 Graft-Versus-Host-Disease
The immunosuppressive property of MSCs renders them particularly suitable for clinical
application in haematopoietic pathologies. Haematopoietic stem cell transplantation
(HSCT) is an important medical procedure used to treat many conditions such as
leukaemia, multiple myeloma, severe combined immunodeficiency, congenital
neutropenia, aplastic anemia, and myelodysplastic syndrome. A severe complication of
HSCT is graft-versus-host-disease (GVHD). This is unique to allogeneic transplantation;
it is a form of rejection where conversely it’s the donor cells that attack the recipient
tissue. Immunosuppressive treatment can be given to the patient, but this can lead to
deadly infections. GVHD also causes inflammation, and may cause development of
31
fibrosis, scar tissue and functional disability. It is mediated by donor T cells which react
to the foreign peptides presented on the MHC proteins of the host cell. As mentioned
above, MSCs can inhibit T-cell activity, and in addition MSCs secrete cytokines that
support haematopoiesis and enhance bone marrow recovery after chemotherapy (Koc et
al., 2000).
One recent example of MSC transplantation is in a trial of 46 patients with haematologic
malignancies (Lazarus et al., 2005). HSCs and culture expanded MSCs obtained from
HLA-matched sibling donors were infused intravenously. Haematopoietic recovery
following transplantation occurred quickly for most of the patients. Furthermore, 23 out
of 46 patients did not develop GVHD. The authors concluded infusion of cultureexpanded MSCs with concomitant infusion of HSCs is safe and can potentially reduce
severe side-effects of HSCT.
1.8.3 Genetic disorders
MSCs are also currently being used for the lysosomal disorders metachromatic
leukodystrophy (MLD) and Hurler’s syndrome (Koc et al., 2002a). Both are autosomal
recessive diseases. MLD is caused by a deficiency of the enzyme arylsulfatase A. In the
absence of this enzyme there is a build up of sulphatides in many tissues of the body.
This eventually leads to the destruction of myelin (the covering around nerve fibers that
acts as insulators) and thus the nervous system. As a consequence, patients show mental
retardation, tetraplegia, and spasticity. Hurler’s syndrome is caused by a deficiency of the
enzyme alpha-L iduronidase, which leads to a buildup of mucopolysaccharides in
lysosomes. This causes a multitude of symptoms: muscle diseases, cardiac failure,
hydrocephalous and mental retardation which can lead to death during infancy (Peters et
al., 1998).
Previous therapies to treat patients with the above diseases involved haematopoietic bone
marrow transplantation (BMT) to halt liver and heart abnormalities (Field et al., 1994;
Krivit et al., 1999). However, muscle alteration and degeneration still progressed and
BMT was not always successful. There was a high incidence of graft failure leading to
morbidity. Koc and colleagues hypothesised that MSC implantation may improve the
32
efficacy of BMT (Koc et al., 2002b). In addition the multilineage differentiation capacity
of MSCs may be utilised for tissue repair. The premise being MSCs would migrate to,
and differentiate into, different tissue types: cartilage, bone, muscle, and peripheral and
central nervous system. Five patients with Hurler syndrome and six patients with MLD
were infused intravenously with allogeneic cultured MSCs. The investigators reported no
infusion-related toxicity. There were significant improvements in nerve conduction
velocities in the patients with MLD after MSC infusion. There was also maintained or an
improved bone mineral density in all patients. However, there was no clinical
improvement in mental and physical change in patients. The investigators believe the
results are promising and further evaluations need to be performed.
More recently, there has been a very promising case report of a female with adult form of
MLD suffering from progressive neurological deterioration (Meuleman et al., 2008). A
combination of non- myeloablative haematopoietic stem cell transplantation (HSCT) with
(MSCs) infusion was performed with no significant toxicity and/or side-effects. Forty
months post-infusion the patient displayed stabilisation in all neurological aspects of her
disease
1.8.4 Gene therapy
MSCs are potentially useful for gene therapy. This research area is largely in the
developmental phase with exciting possibilities. One disease for which genetically altered
MSCs are being used is Cystic Fibrosis (CF) - this is the most prevalent fatal genetic
disorder in the Caucasian population. CF is caused by mutations in the CF
transmembrane conductance regulator (CFTR) gene. There is no cure for CF and most
patients die young, in their 20s and 30s. Wang and colleagues have reported that MSCs
have the ability to differentiate into airway epithelial cells (Wang et al., 2005a).
Furthermore, MSCs isolated from patients with CF are amenable to CFTR gene
correction without loss of differentiation capacity. The allogeneic CFTR-corrected MSCs
can contribute to apical chloride secretion in response to cAMP agonist stimulation.
Thus, MSCs may provide a potential cell-based approach for treatment of CF.
33
1.8.5 Conclusion and future outlook for clinical studies
Taken together, these observations suggest that MSCs have great potential for use in cell
and gene therapy. MSCs offer an alternative cell source for a wide range of therapies
such as OI, Hurler’s syndrome, CF, and MLD. Although preliminary results are
promising and fuelling a growing enthusiasm for MSCs in regenerative medicine and
tissue repair, caution still needs to be exercised. There are still a number of issues
involving the safety and efficacy of stem cell therapy. Patients may contract severe
bacterial and viral infections from donor cultured cells. There is also an issue with long
term engraftment and maintenance of the differentiated phenotype. There has been
extensive research on MSCs in the past decade. On this basis there is increasing pressure
to start clinical trials. However, knowledge about the basic biology of MSCs is still in its
infancy. In all the clinical trials undertaken, a heterogeneous population of adherent bone
marrow cells termed ‘MSCs’ have been used. These issues need to be addressed for
MSCs to be a truly effective cell source for future trials.
1.9 Isolation of mesenchymal stem cells
A main source of MSCs is the bone marrow. However, MSCs are present at a very low
frequency in the bone marrow, estimated at 1 in 100 0000 nucleated cells (CastroMalaspina et al., 1980). The combination of a lack of a known specific MSC marker and
its rarity in the bone marrow has caused great difficulty in MSC research and has
prevented its reproducible isolation and purification.
In general, human MSCs are isolated from a bone marrow aspirate that is taken from the
superior iliac crest (Pittenger et al., 1999). Mouse and rat MSCs are usually isolated from
the bone marrow harvested from the femoral and tibial marrow (Lennon et al., 2001;
Peister et al., 2004) In the case of human bone marrow, the bone marrow mononuclear
cell fraction is typically isolated via density gradient centrifugation (Pittenger et al.,
1999). The subsequent cells are most often cultured in basal media (usually Dulbecco’s
with 10% serum). The non-adherent cells consisting largely of haematopoietic cells are
34
removed by changing the media usually after 24 - 48 hours. Hence, MSC-enriched
cultures are then obtained via the Friedenstein method of simple plastic adherence. A
problem with this method is contamination by a wide variety of adherent cells present in
the bone marrow such as endothelial cells, lymphocytes, adipocytes, osteoblasts and
smooth muscle cells. Using this conventional method, the basis for enrichment of MSCs
is by expansion and passaging of adherent stromal cells. In terms of cultivation of the
subsequent cells there is also huge variation. From cell density, to the length of time the
adherent cells are left in culture (D'Ippolito et al., 2004), the tissue culture surface, to the
addition of various serum types and concentrations (Peister et al., 2004) as well as
various growth factors (Jiang et al., 2002; Muraglia et al., 2000). Some groups also
culture cells in 3D scaffolds (Grayson et al., 2006). This all contributes to a myriad of
cell populations with varying characteristics, all termed MSC, which is obviously suboptimal for research purposes. Thus, a more stringent purification method is required to
identify and distinguish MSCs from other cell types present in the bone marrow.
1.10 Surface markers
Many groups have attempted to identify prospective MSC markers by raising monoclonal
antibodies against adherent bone marrow stromal cells. Due to variations in isolation
techniques and cell culture conditions, the results of these studies are inconsistent. The
antigens identified to date are expressed on a variety of cells and do not provide the
specificity required for MSC purification. For example, the SH2 antibody developed by
the Haynesworth group binds to the TGF-β receptor endoglin (CD105) found on many
cells (Haynesworth et al., 1992). This group also identified the SH3 and SH4 antibodies
which bind to distinct epitopes on CD73, an ecto-5'-nucleotidase, which is also not
restricted to MSCs.
1.10.1 STRO-1
The best known antibody to date used to enrich for MSCs is STRO-1 which was first
identified 17 years ago (Simmons and Torok-Storb, 1991). Surprisingly, its antigen has
35
still not been identified. Although STRO-1 does enrich for CFU-fs approximately 10-fold
from harvested bone marrow cells (Gronthos et al., 2003) it is unlikely it is a specific
marker of MSCs as it reacts with other cell types. Especially as human myofibroblasts
express STRO-1 (Strakova et al., 2008). STRO-1 positive cells have the ability to
differentiate in vitro into adipocytes, chondrocytes, osteoblasts, and HSC supporting
fibroblasts, and smooth muscle cells (Dennis et al., 2002). However, it is possible the
STRO-1 antibody is binding to progenitor or committed cells rather than the primitive
MSCs, which would explain the multilineage potential of the cells. More over, it is
possible the STRO-1 antigen is not expressed by all MSCs and a sub-set of MSC is
missing.
Furthermore, STRO-1 on its own is not sufficient to enrich for MSCs, it is often used in
combination with another antibody, for example, CD106/VCAM-1 (vascular cell
adhesion molecule-1). The investigators work on the premise that MSCs are pericytes
and localised in a perivascular niche in vivo (see in vivo niche above) (Gronthos et al.,
2003). This is because many studies have reported MSCs express CD106, which is also
expressed by pericytes (Bagley et al., 2006). Using a combination of the two selects for
high STRO-1 expressing cells, which form CFU-f, have a high proliferation capacity, and
are multipotential in vitro. Although potentially useful, a drawback with this strategy is
that an assumption is made about the surface antigen expression of MSC, which is not
known. Another limitation of the STRO-1 antibody is due to the initial problem with the
approach, which is STRO-1 was raised against a heterogeneous population of bone
marrow adherent cells (Simmons and Torok-Storb, 1991) rather than a pure population of
clonal, tripotent MSCs.
1.10.2 Other MSC candidate markers
Other studies choose a marker by screening a long list of antibodies against the mixed
population of cells they have derived. If one marker is highly expressed by their cell
population, they utilize this marker to enrich for MSCs. One group selected CD271 (lowaffinity nerve growth factor receptor (LNGFR)) as their marker of choice due to its high
expression on an MSC containing adherent population in comparison to other marrow
36
cell populations (Jones et al., 2002). Another group prefer CD133 (prominin-1) which is
a marker of primitive haematopoietic and neural stem cells (Martin-Rendon et al., 2007).
Other investigators attempt to enrich MSCs from bone marrow with SSEA-1 (AnjosAfonso and Bonnet, 2007) or SSEA-4 (Gang et al., 2007). Another marker used for MSC
enrichment is CD200. These cells were reported to have the potential to differentiate into
adipo-, chondro-, and osteogenic lineages (Wright et al., 2001). However, this marker is
also expressed by endothelial cells, thymocytes, B and T lymphocytes.
According to the literature MSC containing populations are also variably positive for
CD90/Thy-1, CD44, CD29, CD13, Flk-1/CD309, and CD10 (Haynesworth et al., 1992;
Pittenger et al., 1999) and mouse MSCs for stem cell antigen-1 (Sca-1) (Peister et al.,
2004). This marker is commonly used to isolate mouse MSCs and HSCs in combination
with other markers. Unfortunately, none of these markers are exclusive to MSCs and
cannot be used to identify the stem cell in vivo.
1.10.3 MSC negative ‘markers’
MSCs are reported to be negative for the haematopoietic markers CD11b (a leucocyte
antigen), glycophorin-A (a red blood cell marker), CD45 (a marker of all haematopoietic
cells), and CD34 (a primitive HSC marker), although the latter two are sometimes
expressed on murine MSCs (Peister et al., 2004). Both human and mouse MSC
populations are almost always negative for CD31 ((PECAM), expressed on both
endothelial and haematopoietic cells) and CD117 ((c-Kit) a HSC marker).
No single isolation method described above is therefore optimal, and there appears to be
no clear biological difference or advantage between the cells isolated from one method
and another.
1.11 MSCs, BMSCs, or MAPCs?
It is difficult to discern the value from one MSC study to another. This is because over
the years there are many different methods utilised to isolate a heterogeneous population
of adherent cells from different sources and showing varying capacities to differentiate
37
into the mesenchymal lineages - bone, fat, and cartilage. A myriad of names have been
given to these cells. These include mesenchymal stem cell/marrow stromal cell (MSC),
bone marrow stem cell (BMSC), adipose-derived mesenchymal cell (AMC), unrestricted
somatic stem cell (USSC) (Kogler et al., 2004), marrow-isolated adult multilineage
inducible cell (MIAMI) (D'Ippolito et al., 2004), and multipotent adult progenitor cell
(MAPC) (Jiang et al., 2002). The latter two are claimed to be more primitive subsets of
MSCs, with a greater differentiation potential. This is reported to be partially due to the
unique isolation procedure the groups adopted. MIAMI cells were derived by culturing
bone marrow cells on a fibronectin-coated surface in the presence of 3 % oxygen tension.
Two weeks later the non-adherent cells were removed and the adherent cells re-seeded
and expanded (D'Ippolito et al., 2004). MAPCs were derived by culturing CD45 and
Glycophorin-A depleted bone marrow cells in the presence of EGF, PDGF-BB, and low
serum concentration, on laminin coated plates. Leukaemia inhibitory factor (LIF) was
also required for the growth of murine and rat MAPCs (Jiang et al., 2002).
It is possible that the differences in isolation methods and addition of various growth
factors select for a distinct population of cells with varied differentiation potential.
Alternatively it is possible that there is no fundamental difference between the cells
derived using the various protocols, as observed differences could be due to specific
responses to various growth factor treatments. It should also be noted that the results have
been difficult to replicate. In a preliminary study, CD45-/Glycophorin A- bone marrow
cells, CD105 positive bone marrow cells and bone marrow adherent cells were compared
(Lodie et al., 2002). The investigators reported that the cells derived from each method
were virtually indistinguishable in terms of differentiation potential, cell surface antigen
expression, and proliferative expansion.
Recently the International Society for Cellular Therapy (ISCT) met to propose a
guideline for the identification of MSCs in order to standardize the definition of a MSC
(Dominici et al., 2006). There were 3 recommended criteria: (1) ability to adhere to tissue
culture substrate; (2) capacity to differentiate in vitro into the adipogenic, chondrogenic,
and osteogenic lineage; (3) a surface antigen expression of more than 95% positive for
CD73, CD90, and CD105, and less than 2% for CD14, CD19, CD34, CD45, and HLA38
DR. This identification is designed to facilitate MSC research. One drawback with the
standardisation is there are no criteria to demonstrate colony forming unit- fibroblast
assay or self-renewal. Another is the guidelines are based on a summary of studies on
mixed population of MSCs, not a clonal, i.e., single cell derived, and thus pure
population. It is possible a pure population of MSCs may have a distinct phenotype to
that stated. Although these guidelines may be useful in practice, they are not sufficient.
Since, differentiated cells such as human fibroblasts HS68 have a similar surface
phenotype to that described above (Wagner et al., 2005). A better method to set criteria is
to isolate a homogeneous population of MSCs, identify their characteristics, and then set
appropriate guidelines based on those cells. Also there are no guidelines on cell isolation
or cultivation. Therefore cells can be isolated using a wide range of procedures as
described above (1.10-1.12) which may exclude or select for a distinct population of
MSCs.
1.12 Identification of MSCs
1.12.1. Gene profiling
A problem with the studies described in (1.10) is that they were performed on
heterogeneous populations of cells, resulting in phenotyping of such cells rather than pure
MSCs. Even if a pure population of MSCs is used the monoclonal antibody may bind to
an antigen that is highly expressed by the cells but not necessarily specific to MSCs. Thus
these approaches are not optimal for the purpose of identifying a cell-specific marker.
We believe a more useful approach is to apply transcriptomics to characterise the surface
phenotype of a pure MSC population in an endeavour to discover MSC candidate
markers. A transcriptomic approach is suitable as it is hypothesis independent and has a
high level of sensitivity which may enable detection of novel MSC surface markers as
well as important cell regulators.
Previous studies have attempted to catalogue a MSC-specific transcriptome, but the
results were limited as they were again performed on a heterogeneous population of cells
(Jia et al., 2002; Song et al., 2006), (Wagner et al., 2006), (Wieczorek et al., 2003).
Tremain and colleagues published the gene expression profile of a single human clonal
39
MSC population where 2,353 genes were identified (Tremain et al., 2001). The results
were limited as the differentiation potential of the clone was not investigated and
therefore the stem cell status of the clone not determined. Also, only a single clone from
one donor was examined. The gene expression profile of a control population, for
example fibroblasts was not analysed in parallel, thus preventing the identification of
potential MSC-specific genes. However, the results demonstrate that gene expression
profiling is a very powerful technique and when used with a more biologically relevant
starting material and with relevant controls, may shed light on the identity and basic
biology of MSCs.
1.12.1. Microarray vs. Proteomics
One of the major drawbacks of proteomics is the limited sensitivity of the technique, at
least at present. It favours the identification of the most abundant proteins over the less
abundant (Park et al., 2007). Therefore, important proteins that are present at a low
frequency are often missed. Also hydrophobic protein lysates from cellular compartments
are a problem for 2-dimensional electrophoresis (Wagner et al., 2006).
For example, Roubelakis and colleagues compared the proteome of amniotic fluid MSCs,
and bone marrow MSCs, and identified 261 and 175 proteins respectively (Roubelakis et
al., 2007) using MALDI-TOF technology. From this number 116 were present on both
samples. Although this seems a significant number this was for the total number of
proteins of the cell. Since there are approximately 2000 polypeptides in a mammalian cell
the above number is only 13% of the total (Duncan and McConkey, 1982).
One method to characterise the surface antigen profile of MSCs is to isolate and analyse
crude cell membrane fractions. Using a more refined methodology consisting of a
combination of liquid chromatography and tandem mass spectrometry Foster and
colleagues resolved to identify changes in membrane protein markers before and after
osteoblast differentiation in MSCs (Foster et al., 2005). They identified over 400 unique
proteins. Although a significant number of surface proteins were identified, the
significance of the result was limited because the proteomic study was performed on an
hMSC-TERT cell line.
40
One great advantage of microarrays is the much greater sensitivity compared to massspectrometry-based proteomics. With trancriptomics, the expression of all the genes in
the cell can be identified. Furthermore, it is much easier, quicker and reproducible than
proteomic analysis.
The absence of a MSC-specific marker is a fundamental problem in MSC biology for
many reasons. It prevents their identification in vivo. It also prevents knowledge of their
anatomical location and distribution. Due to the lack of a single definitive marker it is
difficult to identify and hence purify MSCs from other cell types. This is extremely
important for research, as the use of a mixed population of cells makes it difficult to
attribute the reported results solely to the MSC. As discussed earlier (1.9) at best mixed
populations containing MSCs are currently being used in clinical studies. It is possible
that a pure population may enhance the results. Therefore, it is paramount that a MSC
specific marker be identified if the full therapeutic potential of this stem cell is to be
realised.
1.13 Cell culture and oxygen tension
In 1907, Ross Granville Harrison was one of the first investigators to grow cells in
culture, which consisted of tissue fragments of frogs (Csete, 2005). The science of tissue
culture for in vitro studies has undergone a revolution since then. Whilst many efforts
have been made to mimic various aspects of the in vivo microenvironment of cells such
as tissue culture surfaces, a myriad of media and serum types, growth factors, 3Dscaffolds, and co-culture systems, little attention has been paid to the atmosphere in
which the cells are cultured. The air is largely composed of nitrogen (78%), oxygen
(21%), argon (1%), and trace amounts of carbon dioxide (0.03%), and other gases. 5%
carbon dioxide is used in in vitro tissue culture to maintain the appropriate pH and it also
reflects levels in vivo. However, the oxygen tension found in mammalian tissue is
considerably less than 20%. The highest oxygen concentration present in vivo is 10-13%
in arteries, lung and liver (Grant and Smith, 1963; Kofoed et al., 1985). Whilst in
avascular tissue such as cartilage it can be as low as 1% oxygen (Brighton and
41
Heppenstall, 1971). Therefore the oxygen tension (20%) cells are exposed to in vitro is
not normoxic, even though in conventional tissue culture this is the name given to it.
1.13.1 Effect of oxygen tension on stem cells
Until recently (D'Ippolito et al., 2004; Grayson et al., 2006; Lennon et al., 2001), the
effect of cultivating MSCs at a more physiologically relevant oxygen tension has been
largely ignored. The majority of in vitro studies on MSCs have been performed at 20%
oxygen tension. Observations from previous studies on MSCs and other stem cells
suggest that cultivation of MSCs in a reduced oxygen environment could enhance their
proliferation, self-renewal and differentiation capacity.
In the bone marrow MSCs exist in a relatively low oxygen environment (4-7%) (Kofoed
et al., 1985). Zwartouw and Westwood were among the first to document that exposure to
low oxygen tension enhances the proliferation of some cells (Zwartouw and Westwood,
1958). Experimental evidence from more recent studies suggest that cultivation of stem
cells at a reduced oxygen tension (5%) enhances proliferation in comparison to normoxic
controls (20%) for rat CNS-derived multipotent stem cells (Studer et al., 2000) and rat
marrow-derived mesenchymal cells (Lennon et al., 2001). Studies by Parinello and
colleagues suggest that reduced oxygen tension may delay or halt proliferative
senescence because there is less DNA damage under these conditions (Parrinello et al.,
2003). This is because there are less reactive oxygen species (ROS).
Cultivation of stem cells in a low oxygen environment also reduces apoptosis for many
cell types including fetal rat derived CNS stem cells (Studer et al., 2000), murine muscle
satellite cells (Csete, 2005) and CD34+ marrow progenitor cells (Mostafa et al., 2000).
Mostafa suggests that this anti-apoptotic effect could be due to a reduction in reactive
oxygen species under hypoxic conditions. Lowering the oxygen tension in culture also
affects the differentiation patterns of stem cells. The Csete group examined the effect of
oxygen tension on myogenic and adipogenic differentiation ability of skeletal muscle
satellite cells (Csete, 2005). They reported that hypoxia induced myogenesis, whereas
normoxia led to adipogenesis. This suggests that oxygen tension is an important regulator
of differentiation. It is also known that oxygen tension is a critical signal between
‘stemness’ and differentiation, as haematopoietic-committed precursors are located near
42
blood vessels where the oxygen concentration is higher, and more primitive precursors
inhabit areas of lower oxygen tension (Mason et al., 1989)
Lennon and colleagues also found that maintaining rat MSCs in reduced oxygen
enhanced their in vivo and in vitro osteogenic and chondrogenic potential as well as their
ability to form colonies relative to control cultures at 20% oxygen (Lennon et al., 2001).
Chondrocytes also react to changes in oxygen tension. De-differentiated cells redifferentiate in hypoxic conditions (Murphy and Polak, 2004). The reason for the
enhanced effect of low oxygen on stem cells in vitro is, perhaps at least two-fold. The
first is that it better approximates the in vivo microenvironment of the cell, the second is
that there are less reactive oxygen species present. High concentrations of oxygen lead to
the generation of reactive oxygen species, which are very small molecules that are highly
reactive due to the presence of unpaired electrons.
In summary, reports from the above studies suggest that low oxygen tension plays an
important role in many aspects of stem cell biology including proliferation, apoptosis,
maintenance of stemness, and differentiation. Therefore, we hypothesise that cultivation
of MSCs in a low oxygen environment which better approximates their in vivo niche may
enhance their stem cell properties and this in turn may enable derivation of more
physiologically relevant stem cell populations.
1.14 Aims of present study
The major aim of this thesis was to identify candidate MSC-specific cell surface
marker(s). We believe identification of a MSC specific marker would have many
important advantages. Firstly it would allow identification of the in vivo localisation of
the MSC, therefore its niche. It would also allow subsequent studies on a pure population
of MSCs and thus enable a better understanding of its true function. The molecular
signature of the MSC could be characterised and enable identification of its regulatory
and self-renewal genes. Identification of a MSC specific marker would also allow
isolation of a pure MSC population which could subsequently be used in clinical studies
43
and may enhance the regenerative effect compared to the standard heterogeneous
population of MSCs currently used.
A microarray approach was adopted to attempt this aim. The rationale was to identify
MSC specific genes by comparing the gene expression profile of murine bone marrow
derived MSCs clonal cells to that of relevant control cell populations. The subsequent
MSC specific genes would then be validated and investigated further in normal human
bone marrow derived MSCs.
In addition, the effect of oxygen tension on MSC biology was to be investigated. This is
because previous studies have reported low oxygen tension enhances the CFU-f ability,
proliferation, and differentiation capacity of MSCs (D'Ippolito et al., 2004; Grayson et
al., 2006; Lennon et al., 2001).
44
Chapter 2
Materials and Methods
45
2.1 Reagents
2.1.1 . Tissue and cell culture reagents
All general laboratory reagents and chemicals were purchased from Sigma-Aldrich
(Dorset, UK) unless otherwise stated. Materials were purchased from the following
sources: Dulbecco’s modified Eagle’s medium (DMEM) from PAA (USA), fetal calf
serum (FCS) from Biosera (USA). DMEM + Glutamax, α- minimum essential media (αMEM) + Glutamax, adult bovine serum (ABS) from Invitrogen (USA). FCS and ABS
were aliquoted and stored at -20ºC until required. Hank’s Balance Salt Solution (HBSS),
trypsin/EDTA, penicillin/streptomycin from Biowhittaker (Wokingham, UK). 70µm cell
strainers and 27 gauge needles from BD Biosciences (Europe), disposable scalpels from
Swann-Morton (Sheffield, England). 2X chamber slides from Nalge Nunc (Thermo
Fisher Scientific, Europe).
Reagents used in differentiation assays were purchased from: TGF-β3 from (Insight
Biotechnology, UK), BMP-2 from Peprotech (England). All tissue culture plastics were
purchased from Becton-Dickinson (Oxford, UK). Oil Red O solution was from Raymond
A Lamb Ltd., (East Sussex, UK), the type II collagen antibody (Ab2031) from Chemicon
(USA), and Alizarin Red S solution from (BDH, Poole, UK).
BALB/b and BALB/c mice strains were purchased from Harlan UK Ltd. (Oxon, UK).
Neonatal human dermal fibroblasts (NHDFs), and human bone marrow mononuclear
cells (hBMMNCs) from donor 1 were from Cambrex Bioscience (Walkersville, USA)
and donor 4 from Lonza Wokingham, Ltd (Berkshire, UK). A sample of bone aspirate
was obtained (with informed patient consent and local ethics committee approval) from
Hammersmith Hospital (London) and another from Stanmore Hospital (Greater London).
Human articular chondrocytes were a kind donation from a member of the laboratory (Dr
Jérome Lafont), derived as previously described in (Lafont et al., 2007). The
chondrocytes were derived from the articular cartilage in the femoral condyle of a 50 year
old female patient. The tissue was obtained with informed patient consent and ethical
approval.
Cells were cultured in a Galaxy triple gas incubator (CRS Biotech, Ayrshire, Scotland,
UK). All cells were cultured in 5% oxygen unless stated otherwise.
46
2.1.2. SDS PAGE and western blotting reagents
Materials were purchased from the following sources: NuPage 4-12% gradient gels from
Invitrogen (Paisley,UK). XCell SureLock Mini-Cell electrophoresis apparatus from
Novex (San Diego, USA). Electrophoresis apparatus from Cambridge Electrophoresis
(Cambridge, UK). Trans-blot chamber from Biorad (Hercules, USA). Polyscreen®
PVDF Transfer Membrane from Millipore (Bedford, MA, USA), ECL western blotting
detection reagents from Amersham (Little, Chalfont, UK). FUJI Medical X-Ray film,
Super RX form Fujifilm (Bedford, UK). Rainbow Molecular weight markers form GE
Healthcare (Europe). HIF-1α antibody was from Novus Biological ((NB100-134) USA),
HIF-1β antibody from BD Biosciences (USA), Telomerase from Abcam ((ab23699)
Cambridge, UK), STEAP1 from Santa-Cruz ((sc-10262), USA). Anti-horseradish
peroxidase secondary antibody was from DakoCytomation (Denmark).
2.1.3 Molecular biology reagents
QIAamp® RNA Blood Minikit from Qiagen Ltd. (Crawley, West Sussex, UK) 1 x Optimem reduced serum media, Lipofectamine 2000 (Invitrogen, Paisley, UK). RNAse-Free
DNAse Set from Qiagen (Crawley, England). Promega kit Reverse Transcription from
Promega (Southampton, UK). Pre-developed primer/probe sets for the following genes
were purchased from Applied Biosystems (Foster City, CA, USA) for mouse: Ly6a/Sca1, STEAP1, STEAP2, TM4SF1, Versican, PPARγ, and for human: RPLPO, STEAP1,
STEAP2. TaqMan 2X PCR master mix from Applied Biosystems.
2.2 Solutions
2.2.1 General buffers
A 10X sterile phosphate buffered saline (PBS) solution (43mM Na2HPO4, 14mM
KH2CO3, 14mM KCL and 1.37M NaCl at pH 7.3) was obtained from VWR
(Lutterworth, UK). A 1X PBS solution was generated by diluting the 10X stock with
ultrapure ddH20.
47
Tris buffered saline (TBS) 20mM Tris-HCL
pH 7.6 137 mM NaCl
2.2.2. Protein extraction buffers
Radio-immunoprecipitation assay (RIPA) 100mM Trizma base pH7.4
300mM NaCl
20mM NaF
2mM EDTA
0.2mM β-glycerophosphate
2% (w/v) Na deoxycholate,
0.5% (w/v) SDS
1% (v/v) NP-40
Prior to use:
10μl/ml Protease inhibitor cocktail was added to the RIPA and ULB buffer
Urea Lysis Buffer (ULB) 8M Urea in ddH2O
10% (v/v) glycerol
1% (w/v) SDS
5mM DTT
10mM Tris HCl pH6.8
2.2.3 Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDSPAGE) buffers
4 x SDS-PAGE sample buffer 0.4M Tris/Hcl pH 6.8
40% (v/v) glycerol,
0.4% (w/v) SDS
2% (v/v) β-mercaptoethanol
0.02% bromophenol blue
Bradford reagent 0.1% (w/v) Coomassie Brilliant Blue G
0.05% (v/v) Ethanol
8.5% (v/v) H3PO4
Running buffer 25 mM Tris
(National Diagnostics, Atlanta, USA) 192 mM glycine
0.1% (w/v) SDS
48
2.2.4. Western blot buffers
Western blot transfer buffer 25 mM Tris
192 mM glycine
20% (v/v) methanol
2-mercaptoethanol stripping buffer 62.5 mM Tris-HCl pH 6.7
2% (w/v) SDS
100 mM 2-mercaptoethanol
2.2.5. Flow cytometry buffers
FACS wash buffer 10% (10X PBS)
1% (w/v) FBS
0.1% (w/v) NaN3
FACS fixing buffer 1X PBS
0.5% (v/v) paraformaldehyde
2.3 Tissue culture
2.3.1. Isolation and expansion of murine clones
Murine clones were derived from the bone marrow of 5-week old male mice (BALB/b
and BALB/c strains). Mice were sacrificed by CO2 asphyxiation for 5 minutes. The tibiae
and femora were aseptically removed and the ends cut to expose the interior of the
marrow shaft. The bone marrow cells were collected by flushing the shaft with Hank’s
Balance Salt Solution (HBSS) using a syringe with a 27 gauge needle attached. The cell
suspension was passed through a 70µm cell strainer and then centrifuged at 1500 rpm for
5mins. After discarding the supernatant, the red blood cells (rbc) were lysed by mixing
the pellet in 25ml rbc lysis buffer for 7-10 min. An equal volume of HBSS was added and
the cell suspension was passed through a cell strainer into a 50ml tube and centrifuged as
above. The supernatant was discarded and the pellet was re-suspended in 50 ml complete
media (DMEM + Glutamax, 10% adult bovine serum albumin, and 1%
penicillin/streptomycin) (hereby called complete media) and centrifuged as before. The
49
bone marrow-derived mononuclear cells (BMMNCs) were seeded at a ‘low’ density (10
x 106 cells in a 15cm dish) and cultured at both 20% oxygen, and in a 5% oxygen, 5%
CO2, triple gas incubator. The non-adherent cells were re-plated 24 hours later and fresh
media added to the adherent cultures.
Cells were cultured for approximately one month, and 50% of the medium was replaced
each week with fresh, pre-equilibrated medium. Cells failed to adhere and survive in the
low density cultures exposed to 20% oxygen. However, in the 5% oxygen cultures, after
approximately 4 weeks there was the appearance of individual colonies. Great care was
taken to isolate (with cloning rings) only distinct individual colonies. Initially each
individual colony was passaged into a 3.5cm dish, and then a 15cm dish. At this stage
cells from each clone were counted and subsequently cultured at a seeding density of 1 x
103 cells per cm2. Cultures were passaged at approximately 70% confluency.
2.3.1.1 Trypsinisation of clones
Cells were trypsinised by washing twice with 1X PBS and then incubating with prewarmed trypsin/EDTA for 1 min at 37oC. The dishes were gently shaken to detach the
cells, which lifted off the tissue culture substrate easily. The trypsin was neutralised by
adding an equal volume of complete medium. Cell suspension was then centrifuged at
1500 rpm and the supernatant discarded. Cells were then re-suspended in an appropriate
volume of complete media (typically 1-2ml for a 70% confluent 15 cm dish) and counted
using a haemocytometer.
2.3.2. Isolation of murine BMMNCs and bone marrow adherent cells
(BMACs)
Primary BMACs were cultured by seeding freshly obtained BMMNCs from 9 BALB/b
and BALB/c mice (each strain separately) at a high density (30 x 106) in two 10cm
dishes. Non-adherent cells were re-plated the next day. Cells were grown to ~80%
confluence (2 weeks).
50
2.3.3. Isolation of mouse embryonic fibroblasts (MEFs)
MEFs were generated from embryos obtained from 2 BALB/b and 2 BALB/c mice
(Abbondanzo et al., 1993). Briefly, pregnant mice at day 13 gestation were sacrificed by
CO2 asphyxiation for 5 mins. The peritoneal cavity was opened and the embryos removed
from the uteri (approximately 4-6 embryos per mouse). The embryo was separated from
its placenta and its surrounding membranes and the head and the liver removed using
forceps and scalpel. The resulting embryos were finally minced by passing through a 2ml
syringe with media (DMEM /10% FBS and 1% P/S) into a 20 ml universal. In the tissue
culture hood a more homogeneous suspension was prepared by passing the minced
embryos through a 70 µm cell strainer. The resulting cell suspension was seeded into 4 x
15cm dishes and incubated at 20% oxygen tension. After 3-4 days when the cells were
confluent they were trypsinised and re-plated to expand their numbers.
2.3.4 Culture of human bone marrow stromal cells (hBMSCs)
A commercially obtained frozen vial of human bone marrow mononuclear cells
(hBMMNCs) was thawed in a 37ºC water bath and transferred into a 15 ml tube
containing 12 ml pre-warmed complete media. This was centrifuged at 1500 rpm, the
supernatant was removed and cells re-suspended in complete media and counted using a
haemocytometer. 10 x106 to 30 x106 cells seeded into a 15 cm dish (‘low’ density) in
order to attempt to derive clones. The non-adherent cells were removed after 24 hours
and fresh medium added. In most cases there were very few adherent cells in the low
density cultures and these did not survive. This method was repeated several times, with
some modifications such as leaving the non-adherent cells for 2-3 days, with the
hypothesis that they may be required to aid the growth of the adherent cells. Once there
was the appearance of adherent cell clusters, these were isolated using cloning rings and
transferred to 3.5 cm dishes. There was a morphology change in these cells from a
spindle shape to a more, flattened spread-out shape. These cells did not survive.
Therefore we proposed to derive human bone marrow stromal cells (hBMSCs) which
would contain a population of MSCs, but were not clonally derived, and hence not
necessarily homogeneous. Commercially obtained hBMMNCs were thawed and 10 x106
51
- 30 x106 cells were seeded in a 3.5 cm dish. The non-adherent cells were removed the
next day. Medium was changed twice a week. After approximately 3 days there was the
presence of spindle shape cells. Cells were passaged when approximately 70% confluent
and expanded further in culture until there was a sufficient number for freezing. This
method was repeated with a fresh human bone marrow aspirate that had been ficollseparated that we received from a local hospital. Human BMSCs isolated from the first
donor were called hBMSC1, and this naming method was used for subsequent donors.
hBMSC1 was derived from ficoll-separated hBMMNCs purchased from Cambrex,
(obtained from a healthy 25 year old male). hBMSC2 was cultured from ficoll-separated
hBMMNCs obtained from a 60 year old female (through Hammersmith hospital).
hBMSC3 was cultured from a fresh bone marrow aspirate (through Stanmore hospital),
from an 11 year old male. hBMSC4 was derived from ficoll separated BMMNCs
obtained from Lonza (Wokingham Ltd) (donor, 25 year old male).
2.3.5 Culture of neonatal human dermal fibroblasts (NHDFs)
NHDFs were thawed as described above and seeded at 2000 cells /cm2 and grown to
~80% confluence in 15 cm dishes containing 20 ml DMEM supplemented with 10%
(v/v) FCS and 1% (v/v) Pen/Strep. Culture medium was replaced twice a week.
2.4 Adipogenic differentiation assay
For adipogenic differentiation, cells were seeded at 6000/cm2 in 6 well plates in the
presence of DMEM, 10% FBS and 1% P/S, with 10μg/ml insulin, 2 nM dexamethasone,
5 µM 3- isobutyl-1-methyl-xanthine (IBMX), 30 μM indomethacin at the three different
oxygen tensions. The medium was changed twice every week for 2 weeks. The cells were
fixed in 10% formaldehyde for 4 min and then covered with 1X PBS, and stored at 4oC
until stained with Oil Red O. Wells were rinsed in 60% isopropyl alcohol, and incubated
in Oil Red O solution for 10 mins before rinsing again in 60% isopropyl alcohol,
followed by water.
52
2.5 Osteogenic differentiation assay
For osteogenic differentiation, cells were seeded at 1 x 103 cells/cm2 and primed in a
minimal media of α-MEM with 5% FBS and 5µg/ml A2-P for 3-4 days. Cells were then
re-seeded into 6 well plates at a cell density of 6 x 103 cells/cm2 in the presence or
absence of osteogenic media (α-MEM + Glutamax/10% FBS, 1% P/S, 50μg/ml A2-P and
10mM β-GP) and or 10ng/ml BMP2, and cultured at 20%, 5% and 1% oxygen tensions.
In addition, human cells were cultured with 10 nM dexamethasone. Osteogenic medium
was replaced with fresh medium every 3-4 days. Immature osteoblast differentiation was
visualised after one week by alkaline phosphatase staining (Sigma), following the
manufacturer’s instructions. Mineralization was assessed after 4-5 weeks by alizarin red
staining of calcified deposits.
2.5.1 Alizarin Red S Staining
A 40 mM solution of Alizarin Red S was made in ddH20. Fixed cells in 6 well plates
were rinsed twice with ddH20. One ml of Alizarin Red S solution was added to each well
and incubated for 15 min with gentle stirring. Alizarin Red S solution was drained and
wells briefly rinsed in ddH20 and overturned on tissue paper to dry at room temperature.
2.6 Chondrogenic differentiation assay
To investigate the chondrogenic capacity of the clones a monolayer system was adopted.
Cells were seeded at a cell density of 6 x 103 cells/cm2 in a 2 well chamberslide and
stimulated for 4-5 days with DMEM-HG, 1% FBS, 1% insulin/transferin/selenium +
linoleic acid (ITS+1), 1mM sodium pyruvate, with or without 0.1 mM ascorbic acid 2phosphate, and 5ng/ml TGF-β3 at 20%, 5%, and 1% oxygen tension. Cells were
incubated until the formation of chondrogenic nodules (4-5 days at 1% and 5% oxygen,
and 1 week at 20% oxygen). Cells were then washed twice in PBS and air-dried.
Chondrogenic differentiation was assessed by type II collagen immunostaining.
53
2.6.1 Type II collagen immunostaining
Type II collagen staining was performed by the Histology department, in Charing Cross
hospital. Chamber slides were incubated for 1 hour at 60ºC to melt the glue that adhered
the plastic to the slides and then fixed for 10 min in acetone. To prevent endogenous
peroxidase activity, they were blocked in a solution of 30% hydrogen peroxide in 300 ml
of methanol for 10 min and then rinsed in running tap water for 2 min. Slides were then
treated with 200 μl of swine serum (Serotec, Oxford, UK) blocking solution containing
0.1% sodium azide for 10 min to minimise non-specific background staining. 200 μl of
the primary type II collagen antibody (used at a dilution of 1:200) was applied to each
slide and incubated for 1 hour and then washed twice in TBS. Slides were then incubated
with 200 μl of the secondary antibody DAKO biotinylated swine-anti-rabbit IgG
antibody (DAKO- Denmark) for 35 min and washed twice in TBS again. During the
incubation a 1:1000 dilution of Vectastain Elite avidin-biotin complex (ABC) was
prepared (Vectar Labs, Petersborough, UK). 200 μl of ABC was added to each section
and the slides incubated for 35 min. The ABC binds to the biotinylated secondary
antibody. In order to visualise the antibody a chromagen was applied to the slides. Vector
laboratories peroxidase substrate DAB SK-4100 was used as per manufacturer’s
instructions. Into every 5 ml of distilled water 2 drops of pH 7.5 buffer, 4 drops of
diaminobenzidine (DAB) and 2 drops of hydrogen peroxide were added. 200 μl of this
solution was applied to each section for 5 min, and the reaction was stopped by
immersing the slides under running tap water for 2 min. The slides were placed into prefiltered Surgipath Harris’s hemotoxylin (Surgipath-Cambridgeshire, UK) for 1 min,
removed and washed in running tap water. The slides were immersed in 300 ml0.3% acid
alcohol (200ml of 70% IMS and 6 ml of concentrated hydrochloric acid) for
approximately 10 seconds each, differentiating the haematoxylin counterstain. Having
been placed back into running tap water the slides were left to ‘blue’ for 10 min. The
slides were placed into 300 ml 70% alcohol for 2 min, followed by 2 min in 300 ml of
90% alcohol. To completely dehydrate the cells 3 changes of 99% IMS for 2 min each
were performed followed by 5x 2 min changes of xylene in order to clean the slides and
remove all traces of alcohol. The sections were mounted in PERTEX mountant
54
(CellPath-Powys, UK), covered with coverslips, and left to dry overnight and then
examined under the light microscope.
2.7 Colony forming unit-fibroblast (CFU-f) assay
Murine clones were seeded (in triplicate) at densities ranging from 100-1000 cells in a
10cm dish and incubated at three different oxygen tensions, 20%, 5%, and 1%. Media
was changed once a week. The appearance of colonies was typically observed by 10-14
days. Cells were rinsed with PBS, air-dried and stained with 10% giemsa stain solution
(BDH, Poole, UK) for 20 min with gentle agitation. The stain was rinsed with water and
then fixed with 70% ethanol. Most colonies had more than 50 cells, and only these were
counted. The number of stained colonies divided by the number of cells seeded allowed
calculation of the percentage of CFU-f.
2.8 Determination of population doublings of clonally-derived cultures
BALB/b1 clone 37, BALB/b3 clone 33, and BALB/c clone 35 were seeded at a cell
density of 500-1000/cm2 and passaged every 3-5 days when the cells were approximately
70% confluent. Cells were counted at each passage using a haemocytometer. In order to
examine the effect of oxygen tension on the population doubling rate of each clone the
same method was used, except cells were incubated at three different oxygen tensions,
20%, 5%, and 1%. The population doubling was calculated after each passage by using
the formula: log N / log 2, where N is the number of cells after 3-5 days in culture
(counted after trypsinisation) divided by the number of initial cells seeded. This was
calculated to determine the total population doublings over time.
2.9 Biochemical Techniques
2.9.1. Preparation of whole cell protein extracts for immunoblotting
Cells were trypsinised as discussed above and centrifuged at 1500rpm at 4ºC. The
supernatant was removed and cell pellet washed once with 10 ml ice-cold PBS and
55
centrifuged as stated above. The supernatant was then discarded and cells lysed in 100 µl
1X RIPA buffer containing protease inhibitors and stored at -20 ºC until further use.
Alternatively for HIF-1α analysis, 0.2 x 106 cells were seeded in a 3.5 cm tissue culture
dish and cultured in 2.5 ml complete media at each oxygen tension (20%, 5%, and 1%
oxygen). One day later, 250 µl desferrioxamine (DFO) (from a 10 X 10mM solution) was
added to one 20% oxygen dish and returned to the 20% oxygen incubator. Cells were
cultured for another day and then rapidly washed in 1X in ice-cold PBS and immediately
lysed in 150 µl urea lysis buffer containing protease inhibitors and stored at -20ºC until
further use.
2.9.2. Bradford assay for protein quantitation
Protein concentration was estimated using the Bradford assay. Frozen cell lysates were
thawed on ice. 10 µl of lysate was mixed with 990 µl of Bradford reagent in a cuvette.
The blank cuvette contained 10 µl of the appropriate buffer (RIPA or urea lysis buffer).
Absorbance at 595 nm (A595) was measured in a spectrometer. Protein concentrations
were estimated by comparison of the A595 obtained with a standard curve generated using
known concentrations of bovine serum albumin (BSA).
2.9.3. SDS- PAGE
Proteins were separated by a discontinuous stacking/separation gel electrophoresis
system. Gels were cast in 18.5 cm x 20 cm glass plates with 1.5 mm spacers. The
resolving gel comprised 380 mM Tris-HCl pH 8.8, 0.1% (w/v) SDS, 10% (w/v)
acrylamide for a 10% gel and 12% (w/v) acrylamide for a 12% gel. A 10% gel was made
for Telomerase and HIF1-α analysis and a 12% gel for STEAP1. Whilst the gel was
polymerising it was coated with a thin layer of H2O-saturated isobutanol to ensure no
bubbles formed across the top of the gel. Once the resolving gel had set the H2Osaturated isobutanol was poured off and the top of the gel rinsed with dH2O prior to
pouring the stacking gel. This comprised 125mM Tris-HCl pH 6.8, 0.1% (w/v) SDS, 5%
(w/v) acrylamide. Once polymerised the gels were place in the electrophoresis apparatus
(Cambridge Electrophoresis, Cambridge, UK) in running buffer.
56
50 µg of the protein lysate per sample were denatured prior to loading by addition of 4X
gel sample buffer and incubating for 5 min at 95ºC. Samples were then briefly pulsed in
the centrifuge to bring any condensed sample to the bottom of the eppendorf. Protein
samples were loaded into the well of the stacking gel. 10 µl of molecular weight markers
were run in parallel to the samples to allow size determination of proteins. Gels were run
at 100V (constant voltage) for approximately 2-3 hours depending on the separation
required. Alternatively, when better separation of proteins was required a 4-12% gradient
gel was used using a NuPAGE mini-gel system. The gel was composed of a 4%
acrylamide stacking gel and a 4-12% acrylamide gradient separating gel. Samples were
loaded in the wells of the gradient gel and run as above using 1X NuPAGE MOPS-SDSrunning buffer, in the XCell SureLock Mini-Cell electrophoresis apparatus which is
designed for mini-gels.
2.9.2 Western blotting
Once electrophoresis was complete, the gel was removed from the electrophoresis
apparatus. The stacking part and dye front were cut off, and that remaining briefly soaked
in transfer buffer. The PVDF membrane was pre-hydrated for 1 min in 100% methanol
and then washed in transfer buffer. The gel and PVDF membrane were sandwiched
between two transfer buffer soaked Whatman filter papers and placed in a Trans-blot
chamber filled with cold (4ºC) transfer buffer. Transfer of proteins was carried out at
100V (constant voltage) for 90 min. The transfer buffer in the Tran-blot chamber was
kept at 4ºC by a circulating cooling pump. Membranes were blocked in 4% (w/v) Marvel
(low fat dry milk) and 1% (w/v) bovine serum albumin in 1 X PBS + 0.1% Tween-20
(PBS-T) for 1 hour on a gentle shaker. Next, membranes were incubated in the same
buffer overnight with the appropriate antibody HIF-1α (dilution 1:250), Telomerase
(dilution 1:1000), and STEAP1 (dilution 1:500) on a rotator. The next day the
membranes were washed 3 times for 10 min in 1 X PBS and then incubated for 1 hour in
the appropriate secondary horseradish peroxidase-conjugated antibody (dilution 1:2000).
This was in 5% (w/v) Marvel milk in 1X PBS solution, performed with gentle shaking.
Next, membranes were washed three times in 1X PBS and then incubated 1 min in ECL
57
solution. Protein bands were visualised by autoradiography using Fuji X-Ray film.
STEAP 1 antibody from Santa-Cruz (sc-10262) detected human STEAP 1 but not
murine. Two other STEAP1 antibodies (Santa-cruz (sc-25514) and Abcam (ab3679) were
tested on human and murine cells, but the antibodies were not specific. The SDS-PAGE
and western blotting for the HIF-1α protein analysis was performed by Miss Sonia
Talma, a research assistant in the laboratory.
2.10 Fluorescence activated cell sorting analysis
Cells cultured in complete medium were trypsinised and counted when ~70% confluent
(2.3.1.1). After washing once with FACS buffer, 1 x 105 cells were transferred into 5 ml
FACs tubes and pelleted by centrifugation for 5 min at 1500 rpm. The cells were resuspended in 200μl FACS buffer and 3.5 µl of FITC or PE-labelled antibody was added.
The cells were incubated at 4ºC for 30 min before pelleting as above. The cells were then
fixed with FACS fixing buffer and stored in the fridge until analysis. Cells labelled with
primary antibodies that were not conjugated with a fluorochrome were labelled for a
second time as above with the appropriate secondary fluorochrome-conjugated antibody
(mouse: FITC-conjugated anti-rat (553896) BD Pharmingen, human: PE-conjugated antimouse (M31504) (Caltag Laboratories, San Francisco, USA)). Mouse cells were analysed
with FACScan cytometer using CellQuest software. This was then replaced with BD
FACSCantoII, which was used to analyse the human cells with BD FACSDiva™
software. At least 10,000 events were collected and CellQuest and FLOWJO software
used to create the histograms for the mouse cells and human cells respectively. Table 2.1
lists the mouse antibodies used, and table 2.2 the human antibodies used. STEAP2
antibody from Novus Biological detected human STEAP2, but not murine. ProsSci
(Poway, CA, USA (4307)) STEAP2 antibody was tested on human and murine cells, but
it was not specific.
58
Table 2.1 Murine antibodies used with FACS analysis
Antibody Label Catalogue Number Source
Sca-1 PE 553108 BD Pharmingen, San Diego, CA, USA
SSEA-1 PE FAB2155P R&D Systems
CXCR-4 FITC 551967 BD Pharmingen, San Diego, CA, USA
MHC-I PE 557000 BD Pharmingen, San Diego, CA, USA
MHC-II PE CBL1324P Chemicon, UK
CD3 PE 555275 BD Pharmingen, San Diego, CA, USA
CD11b PE 557397 BD Pharmingen, San Diego, CA, USA
CD29 - 558741 BD Pharmingen, San Diego, CA, USA
CD31 FITC CBL1337F Chemicon, UK
CD34 PE 551387 BD Pharmingen, San Diego, CA, USA
CD44 PE 553134 BD Pharmingen, San Diego, CA, USA
CD45 PE 553081 BD Pharmingen, San Diego, CA, USA
CD49e PE 557447 BD Pharmingen, San Diego, CA, USA
CD73 PE 550741 BD Pharmingen, San Diego, CA, USA
CD80 PE 553769 BD Pharmingen, San Diego, CA, USA
CD90 PE 553014 BD Pharmingen, San Diego, CA, USA
CD105 FITC FAB1320F R&D Systems
CD106 FITC 553332 BD Pharmingen, San Diego, CA, USA
CD117 PE CBL1359P Chemicon, UK
CD166 PE MAB1172 R&D Systems
59
Table 2.2 Human antibodies used with FACS analysis
Antibody Label Catalogue No. Source
STEAP2 - BC100-2517 Novus Biological
STRO-1 - Developmental Studies Hybridoma Bank, University of Iowa, IW
SSEA-1 PE FAB2155P BD Pharmingen, San Diego, CA, USA
CXCR4 PE 555974 BD Pharmingen, San Diego, CA, USA
HLA-ABC FITC MAB1275F Chemicon, UK
HLA-DR FITC 555811 BD Pharmingen, San Diego, CA, USA
CD11b PE 555388 BD Pharmingen, San Diego, CA, USA
CD13 PE 555394 BD Pharmingen, San Diego, CA, USA
CD29 PE 556049 BD Pharmingen, San Diego, CA, USA
CD34 FITC 555821 BD Pharmingen, San Diego, CA, USA
CD45 FITC 555482 BD Pharmingen, San Diego, CA, USA
CD49e PE 555617 BD Pharmingen, San Diego, CA, USA
CD73 PE 550257 BD Pharmingen, San Diego, CA, USA
CD90 PE 555596 BD Pharmingen, San Diego, CA, USA
CD105 FITC CBL418F BD Pharmingen, San Diego, CA, USA
CD106 PE 555647 BD Pharmingen, San Diego, CA, USA
CD122 PE 554525 BD Pharmingen, San Diego, CA, USA
CD133 - AB5558 Abcam, Cambridge, UK
CD166 PE 559263 BD Pharmingen, San Diego, CA, USA
60
2.11 Telomere length assay
Cells were cultured until ~70% confluence then trypsinised and centrifuged at 1500rpm
for 5 min. Cells were then washed in PBS, pelleted and frozen at -20ºC. Genomic DNA
was isolated using DNA isolation Kit (Qiagen) and 1µg of DNA was measured using
spectrometer. Telomere Length was measured by Telo TAGGG Telomere Length Assay
kit (Roche Applied Science) according to manufacturer’s instructions.
2.12 Karyotype analysis
Cells in metaphase were karyotyped by TDL Cytogenetics (London, UK). Cells in culture
were incubated for 20 min in 10 µg/ml colcemid, and then removed by trypsinisation and
placed in a tube containing hypotonic solution (1:12 FBS/ultrapure water) and mixed by
inversion. The tube was incubated at 37°C for 25 min. The cell suspension was pre-fixed
with 5 drops of fresh cool fixative (3:1 methanol/glacial acetic acid), centrifuged for 10
min at 1500rpm, and the supernatant discarded and the cell pellet loosened by flicking
and re-fixed. This step was repeated once more. The slide was then made. The
supernatant was tipped off and the cells re-suspended in a few drops of fixative. 40 μl of
suspension was applied to the top of the wet washed slide and the excess water drained
off. The cells were then allowed to dry onto the slide with the aid of gentle warming from
an anglepoise lamp with gentle blowing. This allowed the metaphase chromosomes to
burst out of their nuclear envelope and spread sufficiently to allow analysis. Slides were
then assessed with low power phase-contrast microscopy and cell suspension
concentration adjusted to allow best spreading. G-banding was then performed. Slides
were aged by drying in an oven (92°C for 40 min or 63°C overnight) and then rinsed in
HBSS for 5 min. Slides were then treated with Trypsin (1 ml 10x Trypsin in 49 ml
HBSS). This is a critical step, and the timing is variable and determined for each batch
experimentally. The slides were rinsed in HBSS containing serum to stop trypsin activity
and stained in 3:1 Leishman/Giemsa (Merck) (2.5 ml in 47.5 ml pH 6.8 buffer). Slides
were then rinsed in purified water, gently blotted dry, and mounted with coverslips. For
analysis, slides were analysed by conventional bright field microscopy. They were
scanned for metaphases at low power (~100x total magnification) and examined with an
61
oil immersion high power objective (~1000x total magnification).The number of murine
chromosomes in each cell was then counted, typically 20 or more cells per sample were
examined and any obviously abnormal chromosomes, such as double-minute ones noted.
A typical metaphase was digitised and stored using a Cytovision image analysing system.
2.13 Microarray analysis
Clonally-derived cells and MEFs were seeded at 1000cells/cm2 and cultured at 5%
oxygen tension (for 3-7 days). BMACs were cultured as described in section 2.3.2 and
lysed when ~80% confluence. Cells were lysed when 70% confluent as described in
section 2.3.1.1. Care was taken to perform the trypsinisation quickly and in cold
conditions, the cells were centrifuged at 4ºC at 1500 rpm, and cell pellet washed in 10 ml
ice cold 1 X PBS and centrifuged again. The supernatant was removed by aspiration and
the cell pellet lysed in 350µl RLT buffer (with β-Mercaptoethanol) (Qiagen RNeasy Kit),
and frozen at -80°C. Four clones and two MEF, BMAC, and BMMNC cultures per
mouse strain were lysed. Total RNA was extracted from cells using Qiagen RNeasy Kit.
RNA was measured (A260/A280) using a Nanodrop ND-1000 spectrophotometer, before
assessing RNA integrity on the Agilent Bioanalyser RNA 6000 Nano system (Agilent
Technologies, Berkshire, UK).
Ten µg of total RNA from each sample was used to synthesise double-stranded cDNA
using GeneChip one-cycle target labelling and control reagents (Affymetrix, Santa Clara,
CA) following manufacturer’s instructions. In vitro transcription was then performed to
produce biotin-labelled cRNA. Ten μg of the labelled cRNA was fragmented and
hybridised to the GeneChip Mouse Genome 430 2.0 Array (Affymetrix). The chip
contains over 45,000 probe sets representing over 34,000 well substantiated mouse genes.
The cRNA was fragmented, hybridisation cocktails prepared and denatured at 99ºC. The
hybridisation cocktails were put on the array and hybridised for 16 hrs. Hybridisation and
scanning were carried out using Affymetrix 450 fluidics and scanning using the
Affymetrix 3000 7G scanner. Data analysis was performed using the Rosetta Resolver®
gene expression data analysis system (version 7).
62
2.14 RNA extraction, reverse transcription and Real Time PCR
Frozen samples were thawed on ice. RNA was then extracted and prepared using the
RNeasy Mini Kit. Complementary DNA (cDNA) was generated from 1 µg of total RNA
using a reverse transcription kit (Promega, USA), and following the manufacturer’s
instructions. 2% of the reversed transcribed solution was used for each real-time PCR
reaction (done in triplicates) using the following gene-specific Taq-Man primer/probes
for mouse: Ly6a/Sca-1, STEAP1, STEAP2, TM4SF1, Versican, PPARγ and the
normalising reference 18S, and for human: RPLPO, STEAP1, and STEAP2. After an
initial denaturation at 95 ºC for 10 min, a 2-step cycle procedure was used (denaturation
at 95 ºC for 2 sec, annealing and extension at 60 ºC for 30 sec) for 40 cycles on a
CORBETT Research Real-Time thermocycler. Relative gene expression levels were
calculated using the delta-delta threshold cycle (ΔΔCt) method using Rotor-Gene 6000
series software (version 1.7).
2.15 Transplantation of clonally-derived cells
To determine the ability of the clonal cells to induce tumour growth in vivo, 1 x 106 cells
in sterile 1X PBS were injected into 6-8 week old, male BALB/c SCID mice. Mice
analysis of murine tissues were either injected intra-muscularly in the right quadriceps
and sub-cutaneously in the back, or intravenously. (This procedure was kindly performed
by Dr. Gabriel Criado). As a control, BALB/b MEFs (which have a normal karyotype)
were likewise injected in separate animals. Mice injected with BALB/b clones and
BALB/c clones were sacrificed after 4.5 months and 7 months respectively. There was no
macroscopic sign of tumour formation. Lungs, heart, liver, kidneys, and the right and left
quadricep were removed and fixed in neutral phosphate buffered formalin for histological
analysis.
63
2.16 Histological analysis of murine tissues
All sections were stained with haematoxylin and eosin, and masson trichrome and the
lungs and the muscle for Alizarin Red S and Alcian Blue to detect whether the clonal
cells had differentiated. This was performed by the Histology department, in Charing
Cross hospital.
2.16.1 Masson Trichrome staining
For masson trichrome staining formalin embedded sections were cut at 4μM using
Microm HM325 microtome. Sections were baked for approximately 1 hour before dewaxing in 2 changes of xylene, 2 minutes each. Next sections were hydrated in 3 changes
of absolute alcohol, 2 minutes each and then washed in distilled water. Sections were
then immersed in 5% Chromic Acid for 1 hour at room temperature followed by washing
in distilled water. After immersion in Celestine Blue for 5 minutes, sections were washed
in distilled water followed by immersion in haematoxylin for 5 minutes. Sections were
then washed in running tap water and immersed in 0.5% Acid Alcohol (70% Alcohol,
30% distilled water, plus 0.5% concentrated hydrochloric acid) for 20 seconds followed
by a washing step in running tap water for 5 minutes. Sections immersed in 1%
Chromazone Red in 1% Acetic Acid for 5 minutes and washed in distilled water. After
immersion in 1% Phosphomolybdic Acid for 5-10 minutes, controlled microscopically,
they were washed in distilled water followed by immersion in 1% Light Green in 1%
Acetic Acid for 3 minutes. Next they were washed in distilled water, blot dried, then
dehydrated in 3 changes of absolute alcohol, 2 minutes each.
Sections cleared in xylene, 2 changes, 2 minutes each and mounted using Pertex
mountant and an appropriate cover slip.
2.16.2 Haematoxylin and Eosin staining
Formalin fixed embedded sections were cut at 4μM and baked and de-waxed as
described above (2.16.1). Sections were then re-hydrated in 3 changes of absolute
alcohol, 2 minutes each and washed in running tap water. Sections were then immersed in
Haematoxylin solution (Surgipath) for 8 minutes, followed by washing in running tap
64
water. After immersion in 0.5% Acid alcohol (70% Alcohol, 30% distilled water, plus
0.5% concentrated hydrochloric acid) for 20 seconds, they were washed in running tap
water. Sections were then immersed in ‘bluing solution’ ammoniated water ~0.1%
concentrated ammonia for 2 minutes followed by washing in running tap water. Next,
sections were blot dried, then dehydrated in 3 changes of absolute alcohol, 2 minutes
each. Sections were then cleared in xylene and mounted as described above (2.16.1).
2.17 RNA interference experiments
Human bone marrow stromal cells (hBMSCs) were seeded at a cell density of 1,300
cells/cm2 (approximately 50% confluence) in 6 well plates and cultured in complete
medium at 5% oxygen. On day two cells were transfected with small interfering RNA
(siRNA) oligos directed against luciferase (Dharmacon, Lafayette, CO, USA), a
scrambled sequence, and two different STEAP1 (called STEAP1.1 and STEAP1.2) and
STEAP2 (called STEAP2.1 and STEAP2.2) (MWG Biotech, Ebersberg, Germany)
transcripts (Table 2.3) or with lipofectamine alone.
Prior to transfection the appropriate amounts of OptiMEM were incubated at 37ºC.
Medium was removed from the cells and 3 ml of OptiMEM added to each well and cells
returned to the incubator. Next the siRNAs were prepared. An appropriate volume of
lipofectamine 2000 was incubated at room temperature with OptiMEM I for 5 min to
give a final 10 nM concentration. A calculated volume of siRNA was also added to
OptiMEM I to give a 10nM final concentration, and this mixture was incubated for 5 min
at room temperature. After incubation these two mixtures were combined and mixed by
pipetting 2-3 times and incubated for a further 20 min at room temperature to allow for
liposome formation. The mixture of lipid and siRNA was carefully added dropwise to
each well to give a final volume of 3.5 ml per well, and plates swirled very gently to mix.
After 4 hours the medium was removed and 2.5 ml of complete medium was added to
each well. Cells were incubated at 5% oxygen for a further 2 days. Cells were then
washed once with 1X PBS and lysed in 350μl RLT buffer (for real-time PCR assays) and
or lysed in 100μl RIPA (for western blot analysis) and stored at -80ºC.
65
Chapter 3
Isolation and Verification of Stem Cell
Status of Clones
66
3.1 Introduction
As discussed in chapter 1, most groups work with a heterogeneous population of cells
they call mesenchymal stem cells (MSCs). It is then difficult to know whether the
experimental results are directly due to different populations of cells or specifically to a
single population of MSCs. In the present study we developed a protocol to isolate clones
to ensure a pure population of cells were being studied.
MSCs cultured under standard conditions tend to decrease their proliferative rate, and
lose their differentiation capacity (Muraglia et al., 2000). A large number of parameters
need to be considered when culturing MSCs. MSCs are typically cultured at 20% oxygen,
however, in the bone marrow the average oxygen tension is approximately 4-7% as
measured by the Kofoed group (Kofoed et al., 1985). A study by Lennon and colleagues
reported that rat MSCs cultivated in a reduced oxygen atmosphere exhibited increased
proliferation rates, higher colony formation, as well as enhanced in vivo osteo/chondrodifferentiation potential relative to culture at 20% oxygen (Lennon et al., 2001). This
suggests that low oxygen tension plays an important role in the maintenance of the stem
cell phenotype. Therefore, to examine whether 5% oxygen was preferential to the
traditional 20% we performed experiments at both oxygen environments. We
endeavoured to produce a detailed characterisation of the clonally-derived MSC
populations in terms of cell morphology, growth rate, differentiation capacity, and selfrenewal.
3.2 Isolation of clones
Bone marrow was isolated from the long bones of BALB/b mice and depleted of red
blood cells. The subsequent bone marrow mononuclear cells (BMMNCs) were cultured
at a low cell density of 56,000 cells/cm2 at 20% oxygen tension as well as 5% (figure
3.1). There were very few adherent cells in cultures that had been exposed to 20%
oxygen tension and they did not survive. In contrast, after one month there was the
appearance of individual colonies in the cultures that had been exposed to 5% oxygen
tension (figure 3.1). These colonies were subsequently isolated and expanded in culture at
67
Figure 3.1 Diagram of clonal isolation procedure. Bone marrow mononuclear cells isolated
from the long bones of 3 BALB/b mice were seeded at a low density and cultured at 20% and
5% oxygen tension. After one month in culture there were no viable cells in the 20% cultures.
In the 5% oxygen cultures individual colonies appeared that were isolated using cloning rings
and re-seeded into 3.5cm dishes and then 15cm dishes. Cells were counted when 70%
confluent and frozen. This method was repeated to isolate clones from two further BALB/b
isolations and a BALB/c (each isolation from 3 mice). In total 119 BALB/b clones were
isolated from 3 separate isolations and 35 BALB/c clones.
Culture ~1 month
Bone Marrow
(3.5cm dish)
Expand
colonies in
5% oxygen
BALB/b clones = 119
BALB/c clones = 35
(15 cm dish)
Cloning
ring
No viable cells
Oxygen
Tension
20%
Oxygen
Tension
5%
68
5% oxygen. During expansion cells were passaged upon reaching approximately 70-80 %
confluence to prevent loss of multipotency which can occur at confluency. Clones were
derived from the first isolation (3 mice combined per isolation) were called BALB/b1.
Using this method we isolated clones from two further BALB/b isolations (referred to as
BALB/b2 and BALB/b3) and 35 from BALB/c mice. A total of 119 BALB/b clones
(tables 3.1-3.3) and 35 BALB/c clones were obtained (table 3.4).
3.3 Cell doubling time of clonal cells
The doubling time calculated for the clones averaged 1.9 ± 0.2 days for BALB/b1, 1.4 ±
0.3 days for BALB/b2, 1.7 ± 0.6 days for BALB/b3, and 2.9 ± 1.7 days for BALB/c
(mean ± SD) (tables 3.1-3.4). The doubling time included an initial lag phase before
colonies are formed as reported by previous investigators (Colter et al., 2001). All
clonally derived cells had a characteristic spindle shape and were relatively small (figures
3.2-3.3).
3.4 Determining the multipotency of clonal cells
MSCs are typically characterised by their ability to differentiate into three specific
lineages - adipocytes, chondrocytes, and osteoblasts. Thus, this differentiation potential
of the clones was examined to further confirm their stem cell status. Strain-matched
mouse embryonic fibroblasts (MEF) were used as an additional control in differentiation
experiments.
3.4.1 Adipogenic differentiation
For adipocyte differentiation cells were cultured for 1-2 weeks in the presence or absence
of adipogenic media. This media contained the classic hormonal cocktail of
dexamethasone, 3- isobutyl-1-methyl-xanthine, indomethacin, and insulin. Media was
changed twice a week. By day 4, cells containing lipid-rich vacuoles which showed
intense Oil Red O staining were detectable in the BALB/b and BALB/c clonal cultures
that had been exposed to the differentiation media at 5% oxygen tension. Another week
was usually necessary for all the clonal cells to commit to this lineage (figures 3.4-3.5).
In the 20% oxygen cultures it usually took 3 weeks for all the cells to differentiate. There
69
Table 3.1 BALB/b1 clones. Bone marrow mononuclear cells from 3 BALB/b mice were
seeded at low density (56,000 cells/cm2) in 5% oxygen. After one month individual colonies
appeared which were isolated and expanded further in culture. Clonal cells were counted and
the population doublings and cell doubling time calculated and then clones were frozen. This
was the first BALB/b isolation hence clones referred to as BALB/b1 clones.
BALB/b1
(Clone)
Cell Count
(1x106)
Population
Doublings
Doubling Time
(Days)
C1 3.65 22 1.6
C2 2.07 21 1.7
C3 2.02 21 1.7
C4 1.06 20 1.7
C5 3.8 22 1.6
C6 2.47 21 1.7
C7 0.66 19 1.9
C8 0.545 19 2.1
C9 1.45 20 1.8
C10 1.54 21 2.2
C11 2.22 21 1.8
C12 4.23 22 1.5
C13 4.79 22 1.5
C14 2.84 21 1.6
C15 3.33 22 1.6
C16 2.3 21 1.8
C17 2.47 21 1.6
C18 1.81 21 1.7
C19 2.47 21 2.3
C20 2.1 21 2.0
C21 2.21 21 1.8
C22 1.7 21 2.1
C23 2.6 21 2.0
C24 1.6 21 2.1
C25 1.9 21 2.1
C26 3.17 21 2.1
C27 2.85 21 2.1
C28 2.17 21 2.2
C29 2.41 21 2.1
C30 0.56 19 1.9
C31 0.28 18 1.9
C32 0.94 20 1.8
C33 0.365 18 1.9
C34 1.23 20 1.8
C35 1.7 21 1.8
C36 0.3 18 2.5
C37 0.9 20 2.17
Average 2.0 20.7 1.9
70
Table 3.2 BALB/b2 clones. Bone marrow mononuclear cells from 3 BALB/b mice were
seeded at low density (56,000 cells/cm2) in 5% oxygen. After one month individual colonies
appeared which were isolated and expanded further in culture. Clonal cells were counted and
the population doublings and cell doubling time calculated and then clones were frozen. This
was the second BALB/b isolation hence clones referred to as BALB/b2 clones.
BALB/b2
(Clone)
Cell Count
(1x106)
Population
Doublings
Doubling Time
(Days)
C1 1.9 21 1.1
C2 1.9 21 1.1
C3 2 21 1.1
C4 2.2 21 1.1
C5 1.8 21 1.3
C6 3.76 22 1.2
C7 3.41 22 1.2
C8 1.18 20 1.3
C9 3.3 22 1.2
C10 2.53 21 1.2
C11 3.65 22 1.2
C12 3.43 22 1.2
C13 3.69 22 1.2
C14 4.07 22 1.2
C15 2.09 21 1.2
C16 3.3 22 1.2
C17 1.81 21 1.3
C18 2.58 21 1.2
C19 1.84 21 1.2
C20 2.58 21 1.3
C21 1.78 21 1.3
C22 2.52 21 1.3
C23 1.5 21 1.4
C24 2.33 21 1.4
C25 1.76 21 1.4
C26 1.45 20 1.4
C27 1.55 21 1.4
C28 2.13 21 1.4
C29 2.63 21 1.5
C30 1.29 20 1.6
C31 1.31 20 1.6
C32 1.94 21 1.5
C33 1.54 21 1.6
C34 1.92 21 1.7
C35 1.7 21 1.7
C36 0.83 20 1.8
C37 2.08 21 1.7
C38 1.56 21 1.7
C39 3.17 22 1.6
C40 2.74 21 1.7
C41 3.08 22 1.7
C42 2.92 21 1.8
C43 1.19 20 1.9
C44 1.55 21 1.9
C45 1.62 21 1.9
Average 2.2 21 1.4
71
Table 3.3 BALB/b3 clones. Bone marrow mononuclear cells from 3 BALB/b mice were
seeded at low density (56,000 cells/cm2) in 5% oxygen. After one month individual colonies
appeared which were isolated and expanded further in culture. Clonal cells were counted and
the population doublings and cell doubling time calculated and then clones were frozen. This
was the third BALB/b isolation hence clones referred to as BALB/b3 clones.
BALB/b3 Cell Count
(1x106)
Population
Doublings
Doubling
Time (Clone)
(Days)
C1 2.09 21 1.0
C2 2.6 21 1.0
C3 3.3 22 1.0
C4 1.09 20 1.1
C5 1.04 20 1.1
C6 0.72 19 1.1
C7 1.28 20 1.1
C8 0.82 20 1.1
C9 3.26 22 1.0
C10 2.4 21 1.0
C11 0.49 19 1.2
C12 0.61 19 1.1
C13 0.22 18 1.2
C14 0.29 18 1.2
C15 0.75 20 1.1
C16 0.33 18 1.2
C17 0.18 17 1.3
C18 0.458 19 1.6
C19 0.44 19 1.7
C20 0.52 19 1.6
C21 3 22 2.0
C22 0.82 20 2.2
C23 1.4 20 2.1
C24 2 21 2.1
C25 1.47 20 2.1
C26 0.64 19 2.3
C27 1.67 21 2.1
C28 0.69 19 2.3
C29 0.77 20 2.3
C30 0.9 20 2.1
C31 1.21 20 2.1
C32 0.27 18 2.3
C33 0.225 18 2.3
C34 0.54 19 2.2
C35 0.85 20 2.6
C36 0.46 19 2.8
C37 5.44 22 2.3
Average 1.2 19.7 1.7
72
Table 3.4 BALB/c clones. Bone marrow mononuclear cells from 3 BALB/c mice were seeded
at low density (56,000 cells/cm2) in 5% oxygen. After one month individual colonies appeared
which were isolated and expanded further in culture. Clonal cells were counted and the
population doublings and cell doubling time calculated and then clones were frozen.
BALB/c Cell Count
(1x106)
Population
Doublings
Doubling Time
(Clone) (Days)
C1 0.6 19 2.8
C2 0.72 19 2.8
C3 0.92 20 2.7
C4 1.2 20 2.7
C5 0.8 20 2.8
C6 0.9 20 2.7
C7 0.71 19 2.8
C8 2.23 21 2.6
C9 1 20 2.7
C10 0.74 19 2.8
C11 0.35 18 3.0
C12 0.47 19 2.9
C13 0.42 19 2.9
C14 0.91 20 2.9
C15 1.07 20 2.8
C16 0.43 19 3.0
C17 0.43 19 3.0
C18 0.41 19 3.1
C19 1.54 21 2.8
C20 1.28 20 2.9
C21 1.21 20 2.9
C22 1.51 21 2.8
C23 0.94 20 2.9
C24 0.81 20 3.0
C25 0.25 18 3.2
C26 1.32 20 2.9
C27 0.95 20 2.9
C28 0.55 19 3.0
C29 0.54 19 3.0
C30 0.8 20 3.0
C31 0.46 19 3.2
C32 0.3 18 3.4
C33 0.86 20 3.1
C34 1.24 20 3.0
C35 1.67 21 3.0
Average 0.87 19.6 2.9
73
BALB/b
Clone 5
Clone 15
Clone 33
Clone 37
Figure 3.2 Morphology of BALB/b Clones. BALB/b clones were cultured in
complete medium at 5% oxygen tension and passaged every 3-5 days when ~70%
confluent. Photos of clones at passage 5-7.
74
BALB/c
Clone 3
Clone 11
Clone 19
Clone 35
Figure 3.3 Morphology of BALB/c Clones. BALB/c clones were cultured in
complete medium at 5% oxygen tension and passaged every 3-5 days when ~70%
confluent. Photos of clones at passage 5-7.
75
Fi gu re 3
.4
A
di po ge ni c di ff er en tia
tio
n of B
A
L
B
/b c lo ne s. C
lo na l c
el ls a nd m ou se e m br yo ni c fib
ro bl as ts (
M
EF
s) w er e se ed ed at 6
00
0c el ls /c m 2
in t he pr es en ce o r ab se nc e of a di po ge ni c m ed ia (
de xa m et ha so ne , 3
-
is ob ut yl -1
-m et hy l-x
an th in e , i
nd om et ha ci n, a nd i ns ul in )
an d cu ltu
re d at 2
0%
, 5
%
, a
nd 1
%
ox yg en te ns io n. C
lo na l c
el ls c on ta in in g lip
id d ro pl et s w er e se en 4
da ys a fte
r in du ct io n in th e 5%
o xy ge n cu ltu
re s an d by 2
w ee ks lip
id -r ic h dr op le ts w er e pr es en t i
n th e m aj or ity
o f c
el ls a nd s ta in ed w ith
O
il R
ed O
. T
he sa m e pr oc es s to ok 3
w ee ks in th e 20
%
o xy ge n cu ltu
re s. C
el ls in th e 1%
c ul tu re s be ga n to sl ou gh o ff a fte
r 10
d ay s. C
on tro
l ce lls
:
cl on al c el ls c ul tu re d in t he a bs en ce o f ad ip og en ic m ed ia a nd M
EF
s cu ltu
re d in i ts p re se nc e, w er e st ai ne d w ith
H
ae m at ox yl in a nd E
os in in o rd er to to vi su al is e th e un di ff er en tia
te d c el ls . P
ho to s a
re ta ke n at 2
w ee k tim
e po in t f
or 5
%
a nd 1
%
o xy ge n cu ltu
re s a
nd 3
w ee ks fo r t
he 2
0%
.
1% 20
%5%
+
A
di po ge ni c M
ed ia -A
di po ge ni c M
ed ia B
A
LB
/b 1
C
37
B
A
LB
/b 3
C
31
B
A
LB
/b 1
C
37
+
A
di po ge ni c M
ed ia B
A
LB
/b M
EF
50
μM
76
Fi gu re 3
.5
A
di po ge ni c di ff er en tia
tio
n of B
A
L
B
/c c lo ne s. C
lo na l c
el ls a nd m ou se e m br yo ni c fib
ro bl as ts (M
EF
s) w er e se ed ed at 6
00
0c el ls /c m 2
in th e pr es en ce o r a
bs en ce o f a
di po ge ni c m ed ia (d ex am et ha so ne , 3
-i so bu ty l-1
-m et hy l-x
an th in e , i
nd om et ha ci n, a nd in su lin
) a
nd c ul tu re d at 2
0%
, 5
%
, a
nd 1
%
ox yg en te ns io n. C
lo na l c
el ls c on ta in in g lip
id d ro pl et s w er e se en 4
da ys a fte
r in du ct io n in th e 5%
o xy ge n cu ltu
re s an d by 2
w ee ks li pi dric
h dr op le ts w er e pr es en t i
n th e m aj or ity
o f c
el ls a nd s ta in ed w ith
O
il R
ed O
. T
he s am e pr oc es s to ok 3
w ee ks in th e 20
%
o xy ge n cu ltu
re s. C
el ls in th e 1%
c ul tu re s be ga n to s lo ug h of f a
fte
r 1
0
da ys . T
he s am e pr oc es s to ok 3
w ee ks in th e 20
%
o xy ge n cu ltu
re s. C
el ls in th e 1%
c ul tu re s be ga n to sl ou gh o ff a fte
r 10
da ys . C
on tro
l c
el ls : c
lo na l c
el ls c ul tu re d in th e ab se nc e of a di po ge ni c m ed ia a nd M
EF
s cu ltu
re d in it s pr es en ce , w
er e st ai ne d w ith
H
ae m at ox yl in a nd Eo si n in o rd er to to vi su al is e th e un di ff er en tia
te d c el ls . P
ho to s a
re ta ke n at 2
w ee k tim
e po in t f
or 5
%
a nd 1
%
o xy ge n cu ltu
re s a
nd 3
w ee ks fo r t
he 2
0%
.
1% 20
%5%
+
A
di po ge ni c M
ed ia -A
di po ge ni c M
ed ia +
A
di po ge ni c M
ed ia B
A
LB
/c C
3
B
A
LB
/c C
35
B
A
LB
/c C
3
B
A
LB
/c M
EF
50
μM
77
was differentiation evident in cultures exposed to 1% oxygen tension; however after 10
days cells began to slough off the tissue culture dish. Thus adipogenic differentiation was
optimal at 5% oxygen tension. Control cells cultured in the absence of adipogenic media
or fibroblasts cultured in its presence displayed no signs of adipogenic differentiation, but
were viable.
3.4.2 Chondrogenic differentiation
To determine whether the clonal cells had the ability to differentiate into chondrocytes
we seeded the cells at a density of 5000 cells/cm2 and treated them with media containing
TGF-β3 at the three different oxygen tensions, 20%, 5%, and 1%. There was a greater
number of chondrogenic nodules formed in the cultures exposed to both 5% and 1%
oxygen tension and these appeared after 4-5 days. In comparison it took one week for the
nodules to form at 20% oxygen tension. Staining for type II collagen which is a
chondrocyte marker was also more intense in the low oxygen cultures for both mouse
strains (figures 3.6-3.7). There was no type II collagen staining in clonal cultures which
had been cultured in the absence of chondrogenic media or in the fibroblasts which had
been cultured in its presence. In fact, there was no sign of cell aggregation in these
cultures- the cells remained as a simple monolayer.
3.4.3 Osteogenic differentiation
To promote osteogenic differentiation cells were first primed in a minimal media (αMEM media 5% FBS and 5 µg/ml ascorbic acid 2-phopshate (A2-P)) for 3-4 days. Cells
were then re-seeded into 6 well plates and cultured in the presence of osteogenic media
(OM) which consisted of α- MEM media 10% FBS, and 50 µg/ml ascorbic acid 2phopshate and 10 mM β- glycerol phosphate (β-GP). Ascorbate is required for
osteogenesis and β-GP is the source of phosphate required for mineralization. Cells were
also cultured in OM supplemented with 1 or 10 ng/ml of bone morphogenetic protein-2
(BMP-2), a potent stimulator of osteogenesis in murine cells (Rawadi et al., 2003). Media
was changed twice a week. After 1 week, cells cultured in the presence of 10 ng/ml
BMP-2 (at all oxygen tensions) stained positive for alkaline phosphatase. This is an
enzyme that has a role in the mineralization of bone and is an indication of immature
78
79
Fi gu re 3
.6
C
ho nd ro ge ni c di ff er en tia
tio
n of B
A
L
B
/b c lo ne s. C
lo na l c
el ls a nd m ou se e m br yo ni c fib
ro bl as ts (
M
EF
s) w er e se ed ed at 50
00
ce lls
/c m 2
an d cu ltu
re d in th e pr es en ce o r a
bs en ce o f c
ho nd ro ge ni c m ed ia (T
G
F-
β3
) a
t 2
0%
, 5
%
, a
nd 1
%
o xy ge n te ns io n. It to ok 4
-
5
da ys fo r c
ho nd ro ge ni c no du le s t
o ap pe ar in th e 5%
a nd 1
%
o xy ge n cu ltu
re s a
nd 1
w ee k in 2
0%
o xy ge n cu ltu
re . T
he p re se nc e of ty pe II
co lla
ge n in th e ch on dr og en ic n od ul es w er e vi su al is ed b y im m un os ta in in g us in g a m on oc lo na l a
nt ib od y. 50
μM
+
C
ho nd ro ge ni c M
ed ia -C
ho nd ro ge ni c M
ed ia +
C
ho nd ro ge ni c M
ed ia B
A
LB
/b 1
C
37
B
A
LB
/b 3
C
31
B
A
LB
/b 1
C
37
B
A
LB
/b M
EF
1% 20
%5%
Fi gu re 3
.7
C
ho nd ro ge ni c di ff er en tia
tio
n of B
A
L
B
/c c lo ne s. C
lo na l c
el ls a nd m ou se e m br yo ni c fib
ro bl as ts (M
EF
s) w er e se ed ed at 5
00
0c el ls /c m 2
an d cu ltu
re d in th e pr es en ce o r ab se nc e of c ho nd ro ge ni c m ed ia (
TG
F-
β3
)
at 2
0%
, 5
%
, a
nd 1
%
o xy ge n te ns io n. I
t t
oo k 4-
5
da ys f or c ho nd ro ge ni c no du le s to a pp ea r in th e 5%
a nd 1
%
o xy ge n cu ltu
re s an d 1
w ee k in 20
%
o xy ge n cu ltu
re . T
he p re se nc e of t yp e II
c ol la ge n in t he c ho nd ro ge ni c no du le s w
er e vi su al is ed b y im m un os ta in in g us in g a m on oc lo na l a
nt ib od y. B
A
LB
/c M
EF
-C
ho nd ro ge ni c M
ed ia B
A
LB
/c C
3
+
C
ho nd ro ge ni c M
ed ia 1%
B
A
LB
/c C
3
B
A
LB
/c C
35
20
%5%
+
C
ho nd ro ge ni c M
ed ia 50
μM
80
20% 5% 1%
Control
OM +
OM
OM +
1ng/ml BMP-2
10ng/ml BMP-2
Figure 3.8 Alkaline Phosphatase Staining of BALB/b1 Clone 37.
Clonal cells were first primed in a minimal media (α-MEM media 5%
FBS and ascorbic acid 2-phosphate (A2-P)) for 3-4 days. Cells were reseeded and cultured for 1 week in osteogenic media (OM) containing A2P and β-glycerol phosphate. In addition 1ng/ml or 10ng/ml of BMP-2 was
added as indicated. Cells were cultured at 20%, 5%, and 1% oxygen
tension. Differentiation of the clonal cells into immature osteoblasts was
visualised by alkaline phosphatase staining. Photos are indicative of
alkaline phosphatase staining of BALB/b and BALB/c clones.
81
osteoblast formation (figure 3.8). Therefore, in subsequent experiments cells were
cultured in OM supplemented with 10 ng/ml BMP-2 to induce osteogenesis. After 4
weeks of osteogenic induction, clonal cells exposed to 5% and 1% oxygen tension
formed aggregates or nodules which stained positive for alizarin red (figures 3.9-3.10).
This is a dye that stains the calcium phosphate in the mineralised matrix. Cultures
exposed to 20% oxygen tension were viable but did not stain positive. Clonal cells
cultured in the absence of osteogenic media and strain-matched fibroblasts cultured in its
presence did not show any positive Alizarin Red staining. It is important to emphasise
that all BALB/b clones (10) and BALB/c clones (7) examined had the ability to
differentiate into all three lineages (table 3.5).
3.5 Colony forming ability of clones.
Expanded clones were tested for their colony forming ability via a colony forming unit –
fibroblast (CFU-f) assay. The percentage of cells capable of forming CFU-f from early
passage (p6-7) and late passage (p52-3) clones was assessed. 1000 cells were seeded in a
10 cm dish and cultured at 20%, 5%, and 1% oxygen tension (figure 3.11(A)). Individual
colonies with more than 100 cells were counted. The majority of colonies at this time
point had more than 100 cells as shown in figure 3.11 (A). Figure 3.11(B) shows that the
clones had a CFU-f of 65 – 75%. One reason the percentage CFU-f is less than 100% is
because not all the cells adhere to the tissue culture substrate. There was no heterogeneity
in the morphology of the clones at 5% oxygen tension; they were all spindle shaped cells.
At the low seeding densities used for these CFU-f assays, cells cultured at 20% and 1%
oxygen tension did not form colonies, and there was a morphology change in these cells
from a spindle shape to a broader, flatter cell.
3.6 Clonal cells have the ability to self-renew
All clones retained their ability to differentiate into the three lineages after extensive
passaging (p42) (figure 3.12) and to form colonies with a CFU-f of 65 -75% at a high
passage (p52-53) (figure 3.11). Thus, the clones demonstrated self-renewal by their
ability to go through numerous cycles of cell division whilst maintaining their stem cell
82
Fi gu re 3
.9
O
st eo ge ni c di ff er en tia
tio
n of B
A
L
B
/b c lo ne s. C
lo na l c
el ls a nd m ou se e m br yo ni c fib
ro bl as ts (M
EF
s) w er e pr im ed in m in im al m ed ia
-
M
EM
m ed ia 5
%
F
B
S
an d as co rb ic a ci d 2-
ph op ha te (A
2-
P)
) f
or 3
-4
d ay s. C
el ls w er e re -s ee de d at 5
00
0c el ls /c m 2
an d cu ltu
re d in th e pr es en ce o r a
bs en ce of o st eo ge ni c m ed ia (
A
2-
P,
β
-g ly ce ro l ph os ph at e, a nd 1
0
ng /m l B
M
P-
2)
at 2
0%
,
5%
,
an d 1%
o xy ge n te ns io n fo r 4
w ee ks .
M
in er al is ed m at rix
pr od uc tio
n w as v is ua lis
ed b y al iz ar in re d st ai ni ng .
B
A
LB
/b 1
C
37
B
A
LB
/b 3
C
31
1% 20
%
5%
B
A
LB
/c M
EF
+
O
st eo ge ni c M
ed ia -O
st eo ge ni c M
ed ia B
A
LB
/c C
3
+
O
st eo ge ni c M
ed ia -O
st eo ge ni c M
ed ia 10

M
83
Fi gu re 3
.1
0
O
st eo ge ni c di ff er en tia
tio
n of B
A
L
B
/c c lo ne s. C
lo na l c
el ls a nd m ou se e m br yo ni c fib
ro bl as ts (M
EF
s) w er e pr im ed in m in im al m ed ia
-
M
EM
m ed ia 5
%
F
B
S
an d as co rb ic a ci d 2-
ph op ha te (A
2-
P)
) f
or 3
-4
d ay s. C
el ls w er e re -s ee de d at 5
00
0c el ls /c m 2
an d cu ltu
re d in th e pr es en ce o r a
bs en ce of o st eo ge ni c m ed ia (A
2-
P,
β
-g ly ce ro l p
ho sp ha te , a
nd 1
0
ng /m l B
M
P-
2)
at 2
0%
, 5
%
, a
nd 1
%
o xy ge n te ns io n fo r 4
w ee ks . M
in er al is ed m at rix
p ro du ct io n w as v is ua lis
ed b y al iz ar in re d st ai ni ng .
B
A
LB
/c C
3
B
A
LB
/c C
35
B
A
LB
/c M
EF
+
O
st eo ge ni c M
ed ia -O
st eo ge ni c M
ed ia B
A
LB
/c C
3
+
O
st eo ge ni c M
ed ia +
O
st eo ge ni c M
ed ia 1% 5% 20
%
10

M
84
Table 3.5 Clones are multipotent. All BALB/b and BALB/c clones analysed had the
ability to differentiate into adipocytes, chondrocytes and osteoblasts when cultured
under the appropriate conditions.
Clone Adipogenesis Chondrogenesis Osteogenesis
BALB/b1 C5
BALB/b1 C15
BALB/b1 C20
BALB/b1 C32
BALB/b1 C36
BALB/b1 C37
BALB/b2 C7
BALB/b3 C3
BALB/b3 C29
BALB/b3 C33
BALB/c C3
BALB/c C4
BALB/c C11
BALB/c C19
BALB/c C25
BALB/c C29
BALB/c C35
85
Figure 3.11 Clones have the Ability to Form Colonies. Cells were seeded at a low density
(1000cell/ 10cm dish) and after two weeks colonies were stained with the giemsa stain (A). The
number of colonies was counted and the CFU-f calculated (B). CFU-f: colony forming unitfibroblast.
Clone Passage CFU-f (5% O2)
BALB/b1 C37
BALB/b1 C37
BALB/b3 C33
BALB/b3 C33
BALB/c C35
BALB/c C35
p6 72
p52 67
p7 70
p53 75
p6 65
p52 68
Colony Forming Unit –
fibroblast (CFU-f)
Cells
10 cm dish Colony
A
B
86
Figure 3.12 Clones Maintain Differentiation Capacity with Passage. Clonal
cells were cultured in complete medium at 5% oxygen tension and passaged
every 3-5 days when ~70% confluent, until passage 42. The differentiation
ability of the clonal cells was assessed as indicated in figures 3.4-10.
Chondrogenic
Differentiation
Osteogenic
Differentiation
Adipogenic
Differentiation
BALB/c Clone 35BALB/b3 Clone 33
100μM
50μM
50μM
87
Passage 5 Passage 57
BALB/b1
Clone 37
BALB/c
Clone 35
Figure 3.13 Clones Maintain Morphology with Passage. BALB/b and BALB/c
clones were cultured in complete medium at 5% oxygen tension and passaged
every 3-4 days when ~70% confluent.
88
status. In addition clonal cells cultured at 5% oxygen showed no changes in their
morphology with time. (figure 3.13).
3.7 Discussion
Mesenchymal stem cells (MSCs) are currently being used in clinical studies to treat a
myriad of diseases such as Osteogenesis Imperfecta (Horwitz et al., 1999), coronary heart
disease (Orlic et al., 2001), Crohn’s disease (Garcia-Olmo et al., 2003; Garcia-Olmo et
al., 2005), and metachromatic leukodystrophy (Koc et al., 2002a) among others (Barry,
2003). Results have been promising, however, given the increasing therapeutic use of
MSCs, surprisingly little is known about their basic biology. In fact, a specific surface
marker that can be used to identify and purify them from other cells types has yet to be
identified. The lack of a specific marker means a mixed population of cells, termed
‘MSC,’ are transplanted in the patient. Obviously it is not ideal to transplant such an
undetermined population of cells, and it may be that a purer population may enhance the
regenerative effect.
Protocols for isolating MSCs vary considerably due to the absence of a known MSC
marker. The rarity of MSCs in the bone marrow further complicates matters (estimated
0.001-0.01% MSCs in total BMMNC population) (Pittenger et al., 1999). Murine MSCs
are sometimes isolated by simply culturing the whole bone marrow or by centrifuging cut
long bones to obtain all marrow cells (Peister et al., 2004), which can result in
contamination by other adherent cells. In both cases the resulting ‘MSCs’ are obtained by
culturing the adherent cells. Variations in cell culture protocols such as tissue culture
surface (D'Ippolito et al., 2004), seeding densities (Colter et al., 2001), media (Peister et
al., 2004), serum, and the addition of various growth factors (Jiang et al., 2002; Muraglia
et al., 2000) further complicates matters. These procedures result in cultures with varying
characteristics, but all termed MSCs. However, it is crucial to obtain a pure population of
cells if we are to further our understanding of fundamental MSC biology. Thus, a
consistent unbiased purification method is required for the routine isolation of these cells.
89
A major difficulty in culturing murine MSCs is haematopoietic contamination. One group
tried a slightly extreme approach by culturing murine MSCs with a cytotoxin potassium
thiocyanate to poison macrophages and other haematopoietic cell types- no assessment
was made regarding the multipotentiality of the resulting cells (Modderman et al., 1994).
Another group attempted to isolate MCSs from the bone marrow by negative depletion
based on known haematopoietic markers (Baddoo et al., 2003). The problem with this
approach is that it is biased, and since it is possible that MSCs also express these markers,
the true MSC population may be missed. Wang and Wolf demonstrated that when murine
bone marrow cells are seeded at a low density colonies arise that appear to be free from
haematopoietic contamination. However, the method they used was impractical as it
generated only 27 spindle shape colonies of 5 or more cells from a total of 200 culture
dishes (Wang and Wolf, 1990). They cultured cells at the standard 20% oxygen tension.
We hypothesized that culturing marrow-derived cells at the more physiologic 5% oxygen
may prove to be more efficient. Indeed, observations from previous studies on MSCs
have suggested that cultivation of MSCs in a reduced oxygen environment could enhance
its colony forming ability, proliferation, self-renewal, and differentiation capacity
(Grayson et al., 2006; Lennon et al., 2001)
Two important aspects of our study are (i) the need to obtain pure populations of stem
cells and (ii) their cultivation under physiologically relevant conditions, specifically
reduced oxygen tension. Murine bone marrow was isolated by flushing the marrow from
the long bones which is a more sterile removal of the bone marrow. To better
approximate the in vivo stem cell environment the cultures were exposed to a
physiologically relevant oxygen tension of 5% as well as the standard 20%. This is an
important aspect of our work. We found cell survival at 20% oxygen in low density
cultures was very poor. In fact, in low density cultures, clones could only be derived in
reduced oxygen conditions. This suggests that adherence and/or initial growth of clonal
cells is enhanced at 5% oxygen tension.
Using this simple but effective protocol clones were obtained from mouse strains,
BALB/b and BALB/c. Despite the initial lag phase required for clonal growth (Peister et
90
al., 2004), the cell doubling time of the clones is quite short ranging from 1.4 ± 0.3 days
to 2.9 ± 1.7 days (mean ± SD). Murine MSCs isolated by the Baddoo group had a cell
doubling time of 4 to 7 days (Baddoo et al., 2003). FGF2 was required to enhance cell
proliferation; however, exposure to FGF2 inhibited the differentiation ability of the cells.
No growth factor was added to the clonal cells in our study. They were simply exposed to
an oxygen environment which better approximates their in vivo environment.
Clonal cells were first isolated from BALB/b mice due to initial studies by colleagues in
our institute on MSCs from this strain and our preliminary studies on these cells
highlighted their ease of use i.e. their colony forming ability and capacity to differentiate.
Clonal cells were also isolated from the BALB/c strain due to their use in previous MSC
studies (Peister et al., 2004) (Zhou et al., 2006). We did attempt to derive clonal cells
from other common mouse strains, FVB/N and C57Bl/6. Unfortunately, clonal cells from
FVB/N mice became infected and we were not successful deriving clonal cells from
C57Bl/6 mice. Further attempts were not successful for the FVB/N strain. It is possible
that the culture conditions used were not optimal for colony formation from these strains.
Indeed, a previous study has reported that MSCs from different mouse strains vary in the
requirements for optimal media conditions, which include α-MEM, DMEM, RPMI-1640,
and IMDM (Peister et al., 2004).
Mets and Verdonk were the first to report the presence of two cell types in human bone
marrow stromal cell cultures (Mets and Verdonk, 1981). Type I cells were spindle-shaped
and rapidly dividing. Type II cells were flatter and broader and divided more slowly.
They found the majority of cells in primary culture are type I. However, with passage the
number of type II cells increases. Their interpretation was that type I cells are progenitor
cells which give rise to type II terminally differentiated non-dividing cells, by
asymmetric cell division. Further studies from various groups have found that the
‘quality’ of the MSC can be correlated to the cell morphology (Bruder et al., 1998;
Sekiya et al., 2002). In addition the spindle shape cells are thought to expand rapidly, as
they do not synthesize the large amounts of extracellular matrix that the broader, flatter
cells do (Phinney et al., 1999).
91
The clonal cells that were isolated in this study were very similar in phenotype, they were
relatively small and spindle shaped. In contrast, most groups report a variation in the size
and morphology of clones (Digirolamo et al., 1999; Muraglia et al., 2000; Pittenger et al.,
1999). It should be noted that these studies isolated colonies from a mixed population of
bone marrow derived adherent cells cultured at 20% oxygen.
The differences between our clonal cells and the clonal cells in the above studies are
twofold. Firstly, our cells were derived in a 5% oxygen environment which is more
physiologically relevant (Kofoed et al., 1985). Cultures exposed to supra-physiological
(20%) oxygen tensions may be incurring cell damage, e.g. due to reactive oxygen species
(Csete, 2005). The higher oxygen concentration may also be inducing differentiation of
the stem/progenitor cells. These stem cells are receiving signals they would not normally
in the in vivo niche. Secondly, we obtained clonal cells directly from freshly isolated
BMMNCs that were seeded at a low density which selects for cells with a high CFU-f
ability, typically stem cells - this is an important aspect of our approach. Other studies
seed BMMNCs at a high initial density, which also tends to encourage growth of other
adherent cell types. Such adherent heterogeneous population therefore contain stem cells,
progenitor cells, and committed cells. Typically, after subsequent passaging, colonies are
then derived from these cultures and it is not surprising that they are heterogeneous in
terms of morphology and differentiation. Finally, culturing cells at 20% is not conducive
to maintenance of the stem cell phenotype. Many studies have now reported that
cultivation of cells at low oxygen maintains cell ‘stemness’ (D'Ippolito et al., 2006). In
our method we are not only selecting for stem cells, we are culturing them in a low
oxygen environment which maintains their stem cell phenotype.
The Post group have recently reported the presence of 2 independent pre-osteoblastic and
pre-adipocytic clonal populations in murine MSCs (Post et al., 2008). Both cell types’
exhibit different morphologies, the pre-adipocytic had a round oval shape and the preosteoblastic a fusiform, fibroblast-like morphology. In addition to the spindle shaped
cells, clonal cultures also contained a number of very small cells. These have previously
92
been reported by Colter and colleagues in their human MSC cultures (Colter et al., 2001).
They found these cells exhibited a high self-renewal ability and greater
multidifferentiation potential.
The results of the present study suggest that low oxygen has maintained morphology of
clonal cells. At 5% oxygen the cell morphology is the same at high passage (after over
200 population doublings) as it was at low passage. The finding was further highlighted
in the CFU-f assay where cells were seeded at a low density. At 5% oxygen cells formed
colonies and were spindle shaped. However, at 20% and 1% oxygen tension the cells
became broader, flatter and failed to form colonies. Thus, it seems the morphology of the
MSC is an important indicator of the quality of the MSC. Many studies have reported that
low oxygen increases CFU-f ability of stem cells (Fehrer et al., 2007; Grayson et al.,
2006; Lennon et al., 2001). Thus, oxygen tension is an extremely important signal and
should be given a greater degree of consideration when cultivating stem cells.
Each colony in a CFU-f assay is assumed to arise from a single cell since the cells are
seeded at a low cell density and the subsequent colonies arise in distinct locations in the
dish. Thus it follows that the total number of colonies obtained reflects the number of
MSCs in the BMMNCs. In the BALB/b1 isolation thirty-seven individual colonies were
obtained from 20 x 106 BMMNCs. This gives a MSC percentage of 0.0002 % in the
BMMNCs. This figure is most likely an underestimation for two reasons: the first is there
were still some red blood cells present after the cells had been processed with a red blood
cell lysis buffer, and may have been included in the cell count of BMMNCs. The second
is that not all the colonies were isolated since those partially merged were omitted, as
were smaller sized colonies. Therefore the MSC percentage is underestimated. However,
clones in this study were isolated from murine BMMNCs whilst the percentage
expression that the Pittenger group found (0.001-0.01%) was obtained from human
BMMNCs (Pittenger et al., 1999). It is possible that there is a species difference in the
quantity of MSCs present in the bone marrow. Also in this study the MSCs are derived
from immature mice, whereas the MSCs were derived from adults by the Pittenger group.
93
A defining property of mesenchymal stem cells is their ability to differentiate into the
adipogenic, chondrogenic and osteogenic lineages. All the clones examined displayed a
capacity to differentiate into these lineages when given the appropriate environmental
cues (figures 3.4-3.10 and table 3.5). Adipogenesis was demonstrated by the formation of
abundant lipid-rich vacuoles in cells. The formation of chondrogenic nodules, that
contained type II collagen confirmed chondrogenesis. Mature osteoblast formation was
visualised by the presence of a mineralised matrix which stains with Alizarin red. It is
important to note that a low concentration of chemicals was used for adipogenesis in
comparison to previous studies. Also chondrogenesis was induced by the addition of just
5 ng/ml TGF- β3 and without dexamethasone which is commonly added. Osteogenesis
was also induced in the absence of dexamethasone which is widely used. Instead the
cytokine BMP-2 was added.
In previous studies the control used for differentiation experiments was ‘MSCs’ cultured
in the absence of the differentiation media. It is then difficult to tell whether the
differentiation has physiological meaning or has occurred due to the addition of
particularly strong pharmacological stimuli, which could induce differentiation in many
different cell types. Thus in the present study strain-matched mouse embryonic
fibroblasts (MEFs) were also used. The MEFs displayed no signs of differentiation when
cultured in induction media, indicating that the differentiation capacity was specific to the
clones.
Oxygen tension had a significant effect upon chondrogenesis. Low oxygen levels
enhanced the number of chondrogenic nodules formed as well as the intensity of type II
collagen staining in comparison to 20% cultures (figures 3.6-7). A further study in which
1 ng/ml TGF-β3 was added resulted in no or few nodule formation in the 20% cultures.
However, it only took slightly longer (5-7 days) for the cultures at 5% and 1% to form
nodules with 1ng/ml TGF-β3 compared to 10ng/ml. This corresponds with the recent
results of Kanichai and colleagues who found that hypoxia (2% oxygen) promotes
chondrogenesis in rat mesenchymal stem cells possibly through an AKT and HIF-1α
mechanism (Kanichai et al., 2008).
94
We also found that the clonal cells expressed HIF-1α, at both 5% and 1% oxygen tension
in comparison to 20% where there was negligible expression (figure 3.14), which is
consistent with the above findings. Recent studies in our laboratory have demonstrated
that HIF-2α is important for hypoxic induction of cartilage matrix genes in human
articular chondrocytes (Lafont et al., 2007). In future studies it will be interesting to see
which isoform, HIF-1α or HIF-2α or both are important for chondrogenic differentiation
from MSCs.
Low oxygen tension (5%) also enhanced adipogenesis. A greater percentage of cells
contained lipid-rich droplets compared to 20% cultures. However, in the lower 1%
oxygen tension, although differentiation occurred, a confounding factor was that many
cells sloughed off the tissue culture substrate. Hence 5% oxygen tension was optimal for
adipogenesis from MSC clones.
Osteogenesis was enhanced by low oxygen tension. All clones displayed the ability to
differentiate into immature osteoblasts, as all the clones that were tested stained positive
for alkaline phosphatase (figure 3.8) (Jaiswal et al., 1997). This was evident at all oxygen
tensions. The deposition of a mineralised matrix is a sign of a mature osteoblast. All
clones formed a mineralised matrix in the presence of osteogenic media containing 10
ng/ml of BMP-2 at 5% and 1% oxygen tension. Strikingly, there was no staining evident
in cells cultured at 20% oxygen tension although the cells were viable. We used alizarin
red staining to demonstrate mature osteoblast formation as it stains the calcium matrix
deposited by mature osteoblasts. Some groups show alkaline phosphatase staining, which
is a non-specific stain and, at best, an indicator of immature osteoblast staining.
An interesting point to note is that osteogenic differentiation only occurred when the cells
were in 3 ml of differentiation media and not 2 ml. This could be due to the fact that bone
nodules are only thought to form at a very narrow pH range (7.2- 7.4). A decrease in pH
was found to abolish mineralised bone nodule formation in rat osteoblasts (BrandaoBurch et al., 2005). We found the increased culture media volume better buffered the pH
and promoted mineralised matrix formation by the clonal cells. When 2 mls of
differentiation media was used the culture media became yellow indicating increased
95
DFO
20% 5% 1% 20% 20% 1% 20% 1% 1%
BALB/c C35
BALB/c
MEF
FVB/N
Chondro
Human
Chondro
HIF-1α
HIF-1β
DFO DFO
Figure 3.14 MSC clones exposed to Low Oxygen Tension express
HIF-1α. Cells were seeded at a density of 0.2 x 106 cells in a 3.5 cm
tissue culture dish and cultured at the oxygen tension (20%, 5%, and
1% oxygen) indicated. One day later, 250 µl desferrioxamine (DFO)
was added to one 20% oxygen dish and returned to the 20% incubator.
Cells were cultured for another 24 hours and then lysed. HIF1-α protein
expression was examined by western blot analysis. FVB/N chondro
(chondrocytes) were derived from the hip bones of 6 week old FVB/N
mice. Human chondrocytes from the articular cartilage in the femoral
condyle Chodrocytes, (Chondro). All cells at passage 2-4.
96
acidity. In summary the enhanced osteogenesis at low oxygen tension is consistent with
the findings by Lennon and colleagues that found osteogenic differentiation of rat MSCs
is enhanced at 5% oxygen tension compared to 20% (Lennon et al., 2001). However, 3%
oxygen tension has been reported to inhibit osteoblast differentiation of human marrowisolated adult multilineage inducible (MIAMI) cells (D'Ippolito et al., 2006). This could
be due to a difference in species, however, it is perhaps more likely to be due to different
in vitro culture conditions. The rat MSCs from the Lennon group and the clonal cells in
this study are cultured in basal medium with 10% serum. In contrast, the MIAMI cells are
cultured at 3% oxygen in the presence of EGF, PDDGF-BB, 2% FBS, and on fibronectin
coated plates. Apart from hypoxia, the presence, let alone concentration of these growth
factors in the in vivo MSC niche are unknown. Interestingly, in a previous paper the
group reported that MIAMI cells did differentiate into the osteogenic lineage when
cultured at 3% oxygen (D'Ippolito et al., 2004).
Our studies have found low oxygen tension enhances the differentiation potential of
clonally-derived MSCs. There are some studies which report opposing results however.
Adipose-derived murine MSCs show enhanced proliferation, but reduced chondrogenesis
and osteogenesis when cultured in 2% oxygen tension (Malladi et al., 2006a). The same
group do report in a later study that priming adipose-derived murine MSCs in 2% oxygen
tension enhanced early chondrogenesis, but diminished early osteogenesis (Xu et al.,
2007). The inhibiting effect of low oxygen on the differentiation potential of the adiposederived MSCs may be explained by a few reasons. As mentioned above adipose-derived
cells are exposed to a greater oxygen tension in vivo compared to bone marrow MSCs, as
it is a vascularised tissue. This could be why the cells did not respond to the reduced
oxygen atmosphere. In addition, there may be functional differences between MSC
populations isolated from different tissues. The Wagner group found that MSCs isolated
from the bone marrow were superior to adipose derived MSCs in terms of adhesion of
haematopoietic precursor cells and their maintenance in an undifferentiated state (Wagner
et al., 2007). Also, they found bone marrow MSCs are superior in their haematopoietic
supportive function. Studies with cells isolated from the bone marrow exhibit enhanced
colony-forming capabilities, proliferation, and differentiation in response to low oxygen
97
tension (Fehrer et al., 2007; Kanichai et al., 2008; Lennon et al., 2001). Indeed, recent
studies by the Grayson group on human MSCs grown in 3D scaffolds at 2% oxygen
tension found that cells subsequently exhibited enhanced colony-forming ability, better
maintained stem cell property, and enhanced adipogenesis and osteogenesis (Grayson et
al., 2007; Grayson et al., 2006). Our results are more consistent with those from the
Lennon group which found that 5% oxygen enhanced most properties of the rat MSC
including colony forming ability, in vivo and in vitro osteogenesis, and in vivo
chondrogenesis. In addition we found low oxygen enhances initial colony formation, in
vitro adipogenesis and chondrogenesis.
Another functional criterion by which a stem cell is defined is its ability to self renew.
The stem cell property of the HSCs can be demonstrated by their ability to re-constitute
the haematopoietic system after bone marrow transplantation. This is not possible for
MSCs. The best way this can be demonstrated in vitro is by their ability to maintain their
differentiation potential and to form colonies from a single cell after extensive
proliferation. The clones derived in this study had the ability to self-renew, as witnessed
by the fact that they retained their ability to form colonies at a very high passage without
loss of multi-lineage differentiation capability. The spindle shaped morphology of the
clone was maintained which is consistent with the literature in being high in self-renewal,
i.e. colony forming ability and differentiation (Sekiya et al., 2002). The CFU-f assay is
reported to be a good predictor of the life span of the cells (Digirolamo et al., 1999), with
high CFU-f efficiency indicating the greatest replicative potential. This seems to be true
for the clones as they were cultured to passage 58 with no sign of replicative senescence.
Thus the clonal cells exhibited both of the functional characteristics used to define the
MSC; they were tri-potent and had the continued capacity to self-renew.
It is important to note that of the seventeen clones tested all were tri-potent. In addition,
the differentiation was very comprehensive. As evidenced by the data in figures 3.4-3.10,
almost every cell differentiated into the specific lineage, as they all stained with the
appropriate stain. Also there were no spindle shaped cells characteristic of MSCs present;
there is a morphology change in all cultures to the differentiated phenotype. There was
98
also no heterogeneity in the differentiation potential of the clones and this was not lost
over time, in contrast to previous studies (Colter et al., 2001; Digirolamo et al., 1999;
Kuznetsov et al., 1997; Muraglia et al., 2000; Pittenger et al., 1999). Pittenger and
colleagues found that only one third of their clonal cells were tri-potent. This could be
because the clones were isolated from previously passaged cultures, and were
continuously exposed to 20% oxygen. As discussed above such cultures could contain
committed cells, or MSCs that had lost their multilineage differentiation potential with
culture at such supra-physiological levels of oxygen. Extensive studies by the Muraglia
group also reported similar results (Muraglia et al., 2000). The investigators report that
17% of clones were tri-potent whilst the remaining were bi-potent or uni-potent.
However, 34% of the clones were tri-potent when they had been cultured in the presence
of FGF2, prior to the differentiation assays. This growth factor is known to preferentially
select for a certain subset of cells, one with a higher proliferation and osteogenic potential
(Martin et al., 1997). Indeed the investigators report all their clones had this
differentiation potential, if not the adipogenic and the chondrogenic potential. One
disadvantage of adding growth factors is that it is biased to selective enrichment of
subsets of cells, which may not necessarily contain the stem cell of interest.
It has almost become an accepted fact that bone marrow derived clones are
heterogeneous in terms of morphology and differentiation potential. The Tuan group
discusses the heterogeneity of clones at length, hypothesising that ‘MSC’ population
contain not just the putative tripotent MSC but also subpopulations that are bi- and unilineage (Baksh et al., 2004). Another way to look at this is that the cell population termed
MSC is usually heterogeneous because it contains other committed cell types such as
immature osteoblasts, adipocytes etc, which can only differentiate into one lineage. A
recent report has documented the presence of two clonal MSC lines (mMSC1 and
mMSC2) derived from murine bone marrow (Post et al., 2008). Both cell lines exhibit
different morphologies as mentioned previously. Interestingly, mMSC1 could only
differentiate into adipocytes and mMSC2 into osteoblasts. This supports the hypothesis
that MSC populations can contain sub-population of cells that are committed to one
lineage. It is important to note that the committed progenitor cells did not have the
99
characteristic spindle shape of our multipotent clones. Yet another possibility is that there
is something missing from the cell culture conditions that maintain the ‘stemness’ of the
stem cell. In its absence the cells lose their ability to maintain their multi-lineage
differentiation potential and become more committed progenitor cells. We propose that
reduced oxygen tension is such a parameter.
One likely reason our clones are so convincingly tri-potent may be due to the fact that
they are single-cell derived from primary bone marrow cultures – isolated and cultured at
a more physiological oxygen tension which helps maintain the ‘stemness’ of the cells.
Indeed we found differentiation and morphology of the clones was very similar, and both
were retained with extensive passaging.
Oxygen tension has an effect on many aspects of stem cell biology, such as proliferation,
differentiation, self-renewal, cell death, and migration. The oxygen concentration in
arterial blood, lungs and liver is 10-13% (Grant and Smith, 1963; Kofoed et al., 1985). It
can be as low as 1% in avascular tissue such as cartilage. The oxygen tension in the bone
marrow is 4-7% (Kofoed et al., 1985). In conventional cell culture 20% oxygen tension is
used, which is supra-physiological tension for the cells. Cells have evolved ways to
defend themselves from the toxic effects of free radicals derived from high oxygen
concentrations. These include the enzymes glutathione peroxidase, catalases, and
superoxide dismutase, as well as antioxidants ascorbic acid (vitamin C), uric acid, and
vitamin E (Frank and Massaro, 1980; Frank et al., 1980). However, these mechanisms
may not be effective when the oxygen tension is as high as 20%, especially under chronic
exposure. Thus, it is possible that such prolonged exposure to an unusually high oxygen
tension may hinder many cell processes and this is why we see greater proliferation,
differentiation, and colony formation when the cells are exposed to a more physiological
oxygen tension (5%).
There are differences between this study and previous ones which make direct
comparisons difficult. This is the first study that has isolated a large number of individual
murine clones and functionally defined them as MSCs. Another particular aspect of the
present study is that we took great care to isolate clones from primary (not passaged)
100
bone marrow cultures, and in 5% oxygen tension. The effect of oxygen tension on tripotent differentiation of the clones was examined under 3 different oxygen conditions hypoxic conditions (1%); low, but more physiological levels (5%); and under standard in
vitro conditions (20%). From these studies we found that cultivation of these cells at 5%
oxygen is optimal. The clones were very similar in morphology, size, growth rate, and
differentiation capacity, and importantly these qualities were maintained with extensive
subculture. We believe that these results originate from the fact that we isolated a pure
population of cells from the bone marrow using an unbiased purification method and
cells were cultured at a physiological oxygen tension (5%) which helped maintain their
stem cell properties. The subject of the next chapter will be the characterization of the
clonal cells and investigation of the effect of oxygen tension on their growth rate.
101
Chapter 4
Characterisation of MSC Clones
102
4.1 Introduction
The purpose of the work detailed in this chapter was to investigate the in vitro
characteristics of clonal cells described in chapter 3. Their mesenchymal stem cell status
has been functionally determined by their ability to differentiate into the three classical
mesenchymal lineages (adipocytic, chondrocytic and osteoblastic) and by their capacity
to self-renew. The surface antigen profile of the clones was extensively examined using
flow cytometry. In the previous chapter we reported that low oxygen tension (5%)
promoted colony formation and differentiation. We hypothesised it may also enhance
growth rate. Therefore the growth rate of cells was examined at three different oxygen
tensions. Stem cells are known to have the capacity to maintain their telomere lengths
with extensive passaging and to express telomerase. Both of these were examined, as was
the genetic stability of the clones, and their ability to form tumours was assessed
following local and systemic administration.
4.2 Surface phenotype analysis of BALB/b clones
The surface antigen profile of 6 BALB/b clones was examined by flow cytometry using
the markers listed in Table 4.1 and compared to strain matched mouse embryonic
fibroblast (MEF) and bone marrow mononuclear cells (BMMNCs). All the clones and
the MEFs were positive for CD49e and to a lesser extent to CD29, CD44, and CD80 and
MHC-Class I (figure 4.1, table 4.2). An important point to note here (as demonstrated by
the FACS histograms) is that the clones were homogeneous as they stained uniformly for
the surface epitopes, 100% of the population is present in one tight peak in the histogram,
and they are not distributed throughout the histogram in sub-populations. The BALB/b
clones are also positive for CD34, the haematopoietic marker, and indeed murine MSCs
have previously been reported to express CD34 (Peister et al., 2004). In agreement with
the literature the clones were negative for CD117 (c-Kit) and MHC Class II (Peister et
al., 2004). Previous studies have reported that murine MSCs do not express the
haematopoietic marker CD45. However, two of the clones expressed very low levels of
this marker. Clones were also negative for CXCR-4 (Wynn et al., 2004). There were two
103
Table 4.1 Surface antigens examined by flow cytometry for BALB/b
and BALB/c clones.
Antibody Antigen
Sca-1 Stem cell antigen-1 / Ly6a
SSEA-1 Stage specific embryonic antigen 1
CXCR-4 Chemokine (C-X-C motif) receptor 4 (Fusin)
MHC-I Major histocompatability class -I
MHC-II Major histocompatability class -II
CD3 Thymocyte subset
CD11b Leucocyte antigen
CD29 Integrin β1
CD31 Platelet endothelial cell adhesion molecule (PECAM)
CD34 Cluster of differentiation 34
CD44 Hyaluronan receptor
CD45 Leukocyte common antigen, Ly5
CD49e Integrin α5
CD73 Ecto-5′-nucleotidase
CD80 Tetraspanin
CD90 Thy-1
CD105 Endoglin/ TGF-βR
CD106 Vascular cell adhesion molecule -1 (VCAM-1)
CD117 C-kit receptor (KIT)
CD166 Activated leukocyte cell adhesion molecule (ALCAM)
104
Figure 4.1 Surface antigen profile of BALB/b1 clone 37. Flow cytometric analysis of the
surface proteins expressed by BALB/b1 clone 37 was performed by labelling cells with
fluorochrome-conjugated antibody specific for each marker. Labelled cells were analysed by
FACS using Cell Quest software. Results are representative of the surface antigen profile of
other BALB/b clones. Red lines and blue lines in each histogram represent cells labelled with
isotype matched control and specific monoclonal antibody, respectively.
C
ou nt s Relative Intensity
CD3 CD45
CD29
CD31
CD34
CD44
CD73
CD80
CD90
CD106
CD117
MHC II
Sca-1
105
Table 4.2 Flow cytometry characterisation of BALB/b clonal cells. BALB/b MSC clonal cells were
harvested at 70% confluence and labelled with PE- or FITC- antibodies or isotype matched controls and
analysed by flow cytometry at passage 3-6. The surface antigen profile of strain matched mouse
embryonic fibroblasts (MEFs) and freshly isolated bone marrow mononuclear cells (BMMNCs) was
also examined. Symbols refer to intensity of expression. Samples were scored as ‘-’ if the population
was in the first decade (same as the isotype control), ‘+’ if it was between the first and second decade,
‘+’ if it was in the second decade, ‘++’ if it was in the third decade and ‘+++’ if it was in the fourth
decade, relative to the isotype control. * denotes clone at passage 53.
Antigen UC5 UC15 UC20 UC32 UC36 UC37 UC37* MEF BMMNC
Sca-1 + + + + + + + + + + + + + + + + + + + + -
SSEA-1 - - - - - - - - -
CXCR-4 - - - - - - - - -
MHC-I + - ± ± + + + + -
MHC-II - - - - - - - - -
CD3 ± - - + ± + - - -
CD11b - - + - - - - - + +
CD29 + + + - ± + + ± -
CD31 - - - - ± - - - -
CD34 - ± ± ± - ± + ± -
CD44 + + + + + + + + + + +
CD45 - - - - - ± ± - + +
CD49e + + + + + + + + + + + + + + + + ±
CD73 - + - - - + + ± ±
CD80 + + + + + + + + -
CD90 - - - - - - - + + + -
CD105 - - - - - - - ± -
CD106 - - ± - - - - ± -
CD117 - - - - - - - - -
CD166 - - - - - - - - -
106
significant differences between the clones and the MEFs. All the clones were consistently
very positive for the stem cell antigen -1 (Sca-1), which the fibroblasts expressed at a low
level. Another difference was the clones were negative for the fibroblast marker CD90,
which the MEFs expressed at a high intensity. The results demonstrate that the clones had
a distinct phenotype to that of the MEFs. The surface antigen profile of clone 37 was also
examined after extensive passaging (passage 53). The clone retained its surface antigen
profile. The BMMNCs were positive for CD49e and CD73, to a lesser extent CD11b,
CD44, CD45.
4.3 Surface phenotype analysis of BALB/c clones
The surface antigen profile of six BALB/c clones was also analysed by flow cytometry
(figure 4.2, table 4.3). The clones displayed a remarkably similar surface antigen profile
to the BALB/b clones (table 4.2). The clones were highly positive for CD49e and to a
lesser extent for CD29, CD34, CD44, and CD80, which corresponded with the surface
phenotype of the BALB/b clones. The 6 clones displayed variable expression of CD3,
CD45, CD73, and CD106. Only two of the clones expressed fibroblast marker CD90 at a
very low level, whereas the MEFs expressed it at a very high intensity, which is
consistent with the result from the BALB/b MEFs. Like the BALB/b clones, the BALB/c
clones were also negative for CD31, a marker for endothelial cells and they also lacked
expression of MHC Class II. Again, a significant difference between the clones and the
MEFs was that the clones expressed very high levels of Sca-1. BMMNCs were positive
for CD45, CD49e, CD73, and slightly positive for the haematopoietic markers CD11b
and CD117. The surface antigen profile of clone 35 was also examined after extensive
passaging (passage 53). The clone retained its surface antigen profile during this period.
4.4 Growth rate of clones
To determine the growth capacity of the clones their population doubling rate was
followed over time, for over 200 days. Cells were seeded at a density of 500-1000 cells/
cm2 and passaged every 3-5 days when they were ~70% confluent thus ensuring cells
were not contact-inhibited. Figure 4.3 shows the population doubling rate for BALB/b3
107
Figure 4.2 Surface antigen profile of BALB/c clone 35. Flow cytometric analysis of the
surface proteins expressed by BALB/c clone 35 was performed by labelling cells with
fluorochrome-conjugated antibody specific for each marker. Labelled cells were analysed by
FACS using Cell Quest software. Results are representative of the surface antigen profile of
other BALB/b clones. Red lines and blue lines in each histogram represent cells labelled with
isotype matched control and specific monoclonal antibody, respectively.
C
ou nt s CD3 CD45
CD29
CD31
CD34
CD44
CD73
CD80
CD90
CD106
CD117
MHC II
Sca-1
Relative Intensity
108
Table 4.3 Flow cytometry characterisation of BALB/c clonal cells. BALB/c MSC clonal cells were
harvested at 70% confluence and labelled with PE- or FITC- antibodies or isotype matched controls and
analysed by flow cytometry at passage 3-6. The surface antigen profile of strain matched mouse embryonic
fibroblasts (MEFs) and freshly isolated bone marrow mononuclear cells (BMMNCs) was also examined.
Symbols refer to intensity of expression. Samples were scored as ‘-’ if the population was in the first
decade (same as the isotype control), ‘+’ if it was between the first and second decade, ‘+’ if it was in the
second decade, ‘++’ if it was in the third decade and ‘+++’ if it was in the fourth decade, relative to the
isotype control. * denotes clone at passage 53.
Antigen UC3 UC4 UC11 UC19 UC25 UC35 UC35* MEF BMMNC
Sca-1 + + + + ++ + + + + + + + + + + + + + + + +
SSEA-1 - - - - - - - - -
CXCR-4 - - - - - - - - -
MHC-I ± + + ± + + + ± ±
MHC-II - - - - - - - - ±
CD3 - + ± - ± + - - -
CD11b - - - - - - - - + +
CD29 ± + ± ± + ± + ± -
CD31 - - - - - - - - -
CD34 + + ± + + ± + - -
CD44 + + ± ± + + + ± + +
CD45 ± + ± - - ± ± - +
CD49e + + + + + + + + + + + + + + + +
CD73 ± - ± - - + - ± +
CD80 + ++ + + + + + ± -
CD90 - + ± - - - - + + + -
CD105 - - - - - - - ± -
CD106 ± ± - - - - - ± -
CD117 - + + - - - - - +
CD166 - - - - - - - - -
109
0
20
40
60
80
100
120
140
160
180
200
p0 -p2 p4 p6 p8 p1
0
p1 2
p1 4
p1 6 p1
8
p2 0
p2 2
p2 4
p2 6
p2 8
p3 0
p3 2
p3 4
p3 6
p3 8
p4 0
p4 2
p4 4
p4 6
p4 8
p5 0
p5 2
p5 4
p5 6
0
20
40
60
80
100
120
140
160
180
200
220
p0 -p4 p6 p
8
p1 0
p1 2
p1 4
p1 6
p1 8
p2 0
p2 2
p2 4
p2 6
p2 8
p3 0
p3 2
p3 4
p3 6
p3 8
p4 0
p4 2
p4 4
p4 6 p4
8
p5 0
p5 2
p5 4
p5 6
0
20
40
60
80
100
120
140
160
180
200
p0 -p 3
p5 p7 p9 p1
1
p1 3
p1 5
p1 7
p1 9
p2 1
p2 3
p2 5
p2 7
p2 9
p3 1
p3 3
p3 5
p3 7
p3 9
p4 1
p4 3
p4 5
p4 7
p4 9
p5 1
p5 3
p5 5
p5 7
Figure 4.3 BALB/b and BALB/c clones showed extensive proliferation
capacity. The growth rate of BALB/b1 clone 37, BALB/b3 Clone 33, and
BALB/c Clone 35 was followed for over 200 days in 5% oxygen tension. Cells
were seeded at a density of 500-1000 cells/cm2 and passaged every 3-5 days
when they were 70% confluent. Cells were counted at each passage to calculate
population doubling time (standard deviation, s.d.).
BALB/b3
Clone 33
BALB/b1
Clone 37
BALB/c
Clone 35
Cell Doubling Time
1.10 days ± 0.28
Cell Doubling Time
1.10 days 0.34
Cell Doubling Time
1.25 days ± 0.41
(mean ± s.d.)
(mean ± s.d.)
(mean ± s.d.)
Passage
Po pu la tio
n D
ou bl in gs 110
clone 33, BALB/b1 clone 37, and BALB/c clone 35 cultured at 5% oxygen tension.
Beginning at passage 0 to passage 58, the growth rate is constant of BALB/b3 clone 33.
The same trend applies for the BALB/b1 Clone 37 and the BALB/c clone C35. Cell
doubling time averaged 1.10 ± 0.28 days for BALB/b3 clone 33, 1.10 ± 0.34 days for
BALB/b1 clone 37, and 1.25 ± 0.41 days BALB/c clone 35 (mean ± SD). The clones
have been expanded far beyond the 50 cell doublings - the Hayflick limit (Hayflick and
Moorhead, 1961). These results demonstrate the enhanced growth rate of the clones
compared to differentiated cells, and since there was no evidence of cell death, the data is
indicative of cell proliferation.
4.5 Effect of oxygen tension on the growth rate of the clones
To explore further the effect of oxygen tension on the clones, their growth rate was
examined at 3 different oxygen tensions 20%, 5%, and 1%. The population doubling time
for BALB/b3 clone 33 cultured at 5% oxygen tension was consistently shorter than that
cultured at 20% oxygen tension, with 1% being intermediate (figure 4.4). Cell doubling
time in 5% oxygen averaged 1.2 days, 1.7 days in 20% oxygen, and 1.6 days in 1%
oxygen. The trend was similar for the BALB/c clone C35 in that the population doubling
time was shortest when cultured at 5% oxygen tension compared to 1% oxygen with 20%
being intermediate for this clone (figure 4.5). The cell doubling time in 5% oxygen
averaged 1.3 days, 1.5 days in 20%, and 1.6 days in 1% oxygen. Taken together the
results suggest that low oxygen tension (specifically 5%) enhances the proliferative
capacity of the MSC clones.
4.6 MSC clones express telomerase and maintain telomere length
Given the rapid expansion demonstrated by the clonal cells without evidence of
senescence, their telomere length and telomerase expression were next examined. Clones
111
Passage
Po pu la tio
n D
ou bl in gs 0
10
20
30
40
50
60
70
80
90
100
110
p6 p8 p1
0
p1 2
p1 4
p1 6
p1 8
p2 0
p2 2
p2 4
p2 6
p2 8
p3 0
p3 2
p3 4
p3 6
p3 8
p4 0
20%
5%
1%
Figure 4.4 Effect of oxygen tension on the growth rate of BALB/b3 clone 33.
Cells were seeded at a cell density of 500-1000 cells/cm2 and cultured at 20%,
5%, and 1% oxygen tension. Cells were passaged every 3-4 days when they were
70% confluent. Cells were counted at each passage with a hemocytometer.
112
0
10
20
30
40
50
60
70
80
90
p9 p1
1
p1 3
p1 5
p1 7
p1 9
p2 1
p2 3
p2 5
p2 7
p2 9
p3 1
p3 3
p3 5
p3 7
p3 9
p4 1
Passage
Po pu la tio
n D
ou bl in gs 20%
5%
1%
Figure 4.5 Effect of oxygen tension on the growth rate of BALB/c clone 35. Cells
were seeded at a cell density of 500-1000 cells/cm2 and cultured at 20%, 5%, and
1% oxygen tension. Cells were passaged every 3-4 days when they were 70%
confluent. Cells were counted at each passage with a hemocytometer.
113
from both strains at low passage (p9-11) and high passage (p38-46) expressed significant
levels of telomerase compared to the strain-matched fibroblasts (figure 4.6).
Telomerase is known to prolong the replicative potential and thus life span of cells by
adding telomeric repeats to telomeres. Figure 4.7 shows that the telomere length of the
clones was constant at >21 kbp from first passage clones to low passage clones (p9-11) to
extensively passaged clones (p58-59) In comparison passaged neonatal human dermal
fibroblasts had distinctly shorter telomeres.
4.7 Clones are cytogenetically abnormal
Recent studies have linked high expansion kinetics of MSCs to an abnormal karyotype
(Miura et al., 2006). Karyotypic analysis of clones in this study did indeed reveal
chromosomal aneuploidy, with all clones displaying an average of 69 chromosomes with
the addition of 1 or 2 mini-chromosomes (figure 4.8).
A number of BALB/b clones from 3 separate isolations (BALB/b1, BALB/b2, BALB/b3)
were karyotyped, and all revealed the same chromosomal aneuploidy (table 4.4). It is
particularly interesting that this karyotype manifested as early as passage 1 and remained
the same after extensive passaging to passage 53. Thus after being cultured for over half a
year and reaching over 180 population doublings the karyotype remained constant.
Cytogenetic analysis of two of the BALB/c clones also revealed the same chromosomal
abnormality suggesting the transformation was not a strain specific event. The BALB/b
MEFs had a normal karyotype of 40 chromosomes suggesting that chromosomal
aneuploidy of the clones could be due to the selection procedure adopted for the
derivation of clones. It is also possible, although less likely that the cells had the
abnormal karyotype in vivo.
BMMNCs from FVB/N mice were seeded at a high cell density to investigate if this
prevented transformation of the cells. Cells were cultured for a week until they were
~80% confluent. Karyotype analysis of these cells revealed they were tetraploid, with 1
mini chromosome (figure 4.9). This suggests cell transformation occurred very early in
culture, as with the clonal cells, and was irrespective of seeding density. In order to
determine whether low oxygen tension affected transformation, BALB/c BMMNCs were
114
Figure 4.6 Clones Express High Levels of Telomerase. Western blot analysis of
two BALB/b and a BALB/c clone showed high telomerase levels were maintained
with continuous passaging (passage 9-11 to 38- 46) compared to mouse embryonic
fibroblasts (MEFs). BALB/b (B/b), BALB/c (B/c).
Telomerase
α- Tubulin
B/
b C
33
p1 1
B/
c C
35
p1 1
B/
b C
37
p9 B/
b C
33
p4 6
B/
c C
35
p4 6
B/
b C
37
p3 8
B/
c M
EF
p5 B/
b M
EF
p4 MSC Clones
115
Figure 4.7 Passaged Clones Maintain Telomere Length.
Telomere length of the clones did not change significantly from
low passage clones (p1-11) to high passage (p58-59). Fibroblast,
normal neonatal human dermal fibroblasts. BALB/b (B/b),
BALB/c (B/c), Kilobase pairs (Kbp).
KbpB/b
C
1 p
1
B/
b C
33
p1 1
B/
b C
37
p9 B/
c C
35
p1 1
B/
b C
37
p5 8
B/
b C
37
p5 8
B/
c C
35
p5 9
Fib
rob
las
ts p8 21.2
4.9-5.1
1.9-2.0
Low Passage High Passage
116
MSC CLONE
BALB/c Clone 3
69 Chromosomes
MEFs
40 Chromosomes
Figure 4.8 Clones have an abnormal karyotyope. Representative
metaphase spread from MSC clonal cells showing 67 chromosones
plus two mini-chromosomes (marked by arrows) (A). BALB/b
MEFs had a normal karyotype of 40 chromosomes (B) (20
metaphases were analysed per sample). Mouse embryonic
fibroblasts (MEFs).
117
Table 4.4 Clones have an abnormal karyotype. The karyotype of a number of BALB/b and BALB/c
clones was analysed and their chromosome number summarised. All clones from p1 to p51 displayed
similar numerical chromosomal aberrations of approximately 69 with the addition of 1-2 minichromosomes. Twenty metaphases of passage 1 freshly isolated FVB/N bone marrow adherent cells
(BMACs) was also analysed and found to be tetraploid. MEFs had a normal karyotype of 40 chromosomes.
Abbreviations: B/b, BALB/b (1-3 denotes 3 separate isolations); MEFs, Mouse embryonic fibroblasts; pd,
population doublings.
Strain Cells Passage Population Doublings Chromosome Number
MiniChromosome
FVB/N BMAC 1 6 82 2
BALB/b MEF 1 8 40 0
BALB/b1 Clone 2 1 18 69-71 2
BALB/b1 Clone 1 3 20 68-71 1
BALB/c Clone 3 4 23 68-69 2
BALB/b 2 Clone 3 4 24 ~68 2
BALB/b 2 Clone 2 4 25 ~69 2
BALB/b 2 Clone 1 5 27 ~69 2
BALB/b 2 Clone 4 6 28 69 1
BALB/c Clone 4 7 31 67 2
BALB/b 3 Clone 31 24 89 ~69 2
BALB/b 3 Clone 31 51 180 ~69 1
118
5% Oxygen, BALB/c BMAC
BALB/c BMAC
79 chromosomes
20% Oxygen
FVBN BMAC
83 chromosomes
5% Oxygen
Figure 4.9 Bone Marrow Adherent Cells display Chromosomal
Aneuploidy. Bone marrow mononuclear cells from BALB/c or FVB/N mice
were seeded at a high density and cultured at 20% or 5% oxygen tension as
indicated. After one week cells were re- seeded into a flask and
cytogenetically analysed (passage 1). Representative metaphase spread from
each sample is shown, FVB/N bone marrow adherent cells (BMACs) were
tetraploid, BALB/c BMACs cultured at 20% were tetraploid, BALB/c
BMACs cultured at 5% oxygen tension were normal except one cell contained
59 chromosomes plus one mini-chromosome (marked by arrow) (20
metaphases were analysed per sample).
1 cell with 59 chromosomes40 chromosomes
119
seeded at a high density and cultured at both 20% and 5% oxygen tension. Cells cultured
at 20% oxygen tension were found to be tetraploid whereas cells exposed to 5% oxygen
were normal, with the exception of one cell that contained 59 chromosomes with one
mini-chromosome, suggesting that the transformation event was beginning, but was at
least delayed in reduced oxygen levels.
4.6 Clonal cells are not malignant
Since the clonal cells have an abnormal karyotype, they were transplanted into
immunodeficient mice to determine whether they were actually malignant. Four different
BALB/c clones were injected sub-cutaneously and intra-muscularly (i.m.) (n=2) or intravenously (i.v) (n=2). MEFs, which have a normal karyotype were used as a control, and
injected in a similar way. Previous groups have found tumour formation by 4 weeks after
i.m. injection, usually at the site of the injection, and in the lungs by 8 weeks after i.v
injection of cytogenetically abnormal MSCs into immuno-compromised mice (Miura et
al., 2006; Tolar et al., 2007; Wang et al., 2005b; Zhou et al., 2006). In contrast to this, we
found no sign of tumour growth macroscopically or microscopically after 4½ and 7
months after the mice were injected with BALB/c and BALB/b clones respectively.
Histological evaluation showed all tissues were healthy. Figure 4.10 shows that the heart,
lung, kidney, and prostate of SCID mouse injected with BALB/b1 clone C37 that was
sacrificed at 7 months is normal with no signs of sarcoma formation.
Normal BALB/c mice were injected with 4 BALB/c clones as well as normal MEFs and
were sacrificed at 4½ months. Results were the same as above there was no sign of
tumour formation and or differentiation.
4.7 Despite an abnormal karyotype, clones display regulated
differentiation
Transformed cells have been reported to exhibit dysfunctional patterns of differentiation
120
Figure 4.10 Clonal cells are not malignant. BALB/b clonal cells C37 (1
x 106) were injected sub-cutaneously and intra-muscularly, or intravenously through the tail vein, into severe combined immuno-deficiency
mice (SCID). Mice were sacrificed 7 month post injection. There was no
sign of tumour formation macroscopically or microscopically. Photos are
representative of all SCID mice injected with all BALB/b and BALB/c
clones. Sections were stained with haemotoxylin and eosin, and masson
trichrome. Magnification 25X.
Heart
Lung
Kidney
Liver
Prostate
Haemotoxylin and Eosin Masson Trichrome
121
(Bjerkvig et al., 2005). However, in contrast to this, the differentiation of clonal cells in
our study was a highly regulated process. Table 4.5 shows the average gene expression
intensities (arbitrary units) of Sca-1 and the adipocyte-specific transcription factor PPARγ during adipogenic differentiation of the clones. The undifferentiated clones from both
strains express relatively high levels of Sca-1 in comparison to control cells,
differentiated clonal cells, MEFs, bone marrow adherent cells (BMACs), BMMNCs.
Adipocyte differentiated clones however, display high levels of PPAR-γ in comparison to
the same undifferentiated clones and the controls. The results are particularly significant
as they are the average of 4 MSC clones and 2 control samples.
The same trend is also found at the protein level, via flow cytometry (table 4.6). The
surface antigen profile of the BALB/b1 clone 37 was examined in the presence of
adipogenic media at day 1, 4, and 2 weeks when all the cells had differentiated.
Interestingly, after 1 day in differentiation media Sca-1 expression decreased and after
complete differentiation its expression was very low. This supports its role as a stem cell
antigen as its expression was hugely down-regulated upon differentiation.
Microarray data comparing undifferentiated to adipocyte differentiated clones also
revealed adipocyte genes to be highly upregulated in the differentiated clones. For
example, fat specific gene 27 was upregulated over 200 fold in the differentiated clones
(table 4.7A). This occurred with a concomitant decrease in stem cell and cell cycle genes,
Sca-1 and Cdk4, which were downregulated 9 and 3 fold, respectively (table 4.7B). There
is also a down-regulation in the cytokines HGF (5-fold) and FGF-7 (4-fold), which have
a role in proliferation (Forte et al., 2006) (Cho et al., 2008).
122
Table 4.5 Clones display regulated differentiation. The relative gene expression
levels for Sca-1 and PPARγ for 4 BALB/b and 4 BALB/c clones, and the equivalent
clone differentiated along the adipogenic pathway was examined via microarray
analysis. Strain matched MEFs, BMACs, and BMMNCs were also examined. In
comparison to undifferentiated clones (UCs), differentiated clones (DCs) expressed low
levels of Sca-1 and alternatively high levels of the adipocyte marker PPARγ. For each
strain n = 2 for mouse embryonic fibroblast (MEF), bone marrow mononuclear cells
(BMMNCs), and bone marrow adherent cells (BMACs), and n= 4 for UC and DC.
Cell Type BALB/b BALB/c
Sca-1 PPARγ Sca-1 PPARγ
UC 1 1 1 1
DC 0.01 4.97 0.22 2.75
MEF 0.64 0.20 0.41 0.21
BMAC 0.06 0.48 0.55 1.64
BMMNC 0.03 0.02 0.02 0.06
123
Table 4.6 Sca-1 expression decreases upon adipogenic differentiation of MSC
clones. The surface antigen profile of BALB/b1 clone 37 was examined by flow
cytometry at day 0, and in adipogenic differentiation media at day 1, day 4, and then
after 2 weeks when all the cells had differentiated into adipocytes. Samples were
scored as ‘-’ if the expression level was in the first decade (same as the isotype
control), ‘+’ if it was between the first and second decade, ‘+’ if it was in the second
decade, ‘++’ if it was in the third decade and ‘+++’ if it was in the fourth decade,
relative to the isotype control.
Adipogenic Media
Day 0 Day 1 Day 4 2 Week
Sca-1 +++ ++ ++ ±
CXCR4 - - - -
SSEA-1 - - - -
MHC-I ± + + +
MHC-II - - - -
CD3 ± - ± -
CD11b - - - -
CD29 - + - ±
CD31 - - - ±
CD34 ± - ± -
CD44 + + + ±
CD45 - - - ±
CD49e ++ ++ + ++
CD73 - ± ± ++
CD80 + + ± ±
CD90 - - - -
CD105 - - - -
CD106 - - - -
CD117 - - - ±
CD166 - - - -
124
Table 4.7 Adipocyte clones are distinct to the undifferentiated clones. The gene expression profile
of 4 BALB/b and 4 BALB/c clones, and the equivalent clone differentiated along the adipogenic
pathway was examined via microarray analysis. In comparison to undifferentiated clones (UCs),
differentiated clones (DCs) expressed high levels of adipocyte genes (A). In comparison, DCs
expressed low levels of Cdk4, stem cell marker Sca-1 and cytokines involved in cell proliferation (B).
For each strain n = 4 for mouse embryonic fibroblast (MEF), bone marrow mononuclear cells
(BMMNCs), and bone marrow adherent cells (BMACs), and n= 8 for UC and DC.
A
Sequence Name Gene Accession Number UC DC
Fold Change
(UC vs DC)
fat specific gene 27 Cidec BB221402 19 4008 212
hydroxysteroid 11-beta
dehydrogenase 1 Hsd11b1
NM_008288
23 3469 151
adipocyte complement related
protein Adipoq
NM_009605
63 9428 149
fatty acid binding protein 4 Fabp4 BC018558 8 836 108
lipase, hormone sensitive Lipe NM_010719 17 1349 77
fatty acid Coenzyme A ligase,
long chain 2 Acsl1
BI413218
56 4079 73
B
Sequence Name Gene Accession Number UC DC
Fold Change
(DC vs UC)
chemokine (C-C motif) ligand 7 Ccl7
AF128193
1352 113 12
Stem cell antigen- 1/Ly6a Sca-1
BC002070
9177 1072 9
ELK3, member of ETS
oncogene family Elk3
BC005686
200 32 6
hepatocyte growth factor Hgf
AF042856
356 76 5
fibroblast growth factor 7 Fgf7
NM_008008
2092 571 4
cyclin dependent kinase 4 Cdk4
NM_009870
3837 1330 3
histone deacetylase 7A Hdac7a
BB277517
605 223 3
125
4.10 Discussion
The characteristics of mesenchymal stem cells have not been well clarified. Varying
reports about their proliferative ability and surface antigen profiles are perhaps due to
variation in cell isolation procedures and cell culture methods used. In addition, a
thorough characterisation of MSCs is desirable prior to therapeutic application. In the
present study the surface antigen profile of the clones was examined to further
characterise these clonally-derived (hence pure), multipotent cell populations. The
method used to denote the level of expression for FACS analysis is described in table 4.2.
This differs from that used in most previous studies. For example, if the population is to
the far right in the fourth decade, the population is denoted +++ meaning very highly
expressed. The Peister group denoted a population +++ if it was 100% outside the isotype
control and in the second decade (Peister et al., 2004). In our results such a population
would be denoted as +.
All clones analysed displayed a remarkably similar surface antigen phenotype, and clones
from both mouse strains (BALB/b and BALB/c) were also very similar. All clones
consistently expressed high levels of Sca-1 in comparison to the MEFs and BMMNCs.
Sca-1 is a glycosyl phosphatidylinositol-anchored cell surface protein (GPI-AP). Sca-1,
as its name states (stem cell antigen), is known to be expressed by haematopoietic stem
cells and mesenchymal stem cells. In addition, it is also reported to be expressed by stem
and progenitor cells in the mammary gland (Welm et al., 2002), prostate (Xin et al.,
2005), heart (Oh et al., 2003), liver (Petersen et al., 2003), and skeletal muscle (Lee et al.,
2000). The function of Sca-1 has not been fully elucidated, however there are reports that
it is involved in cell-cell adhesion and signalling (Holmes and Stanford, 2007). A study
of Sca-1 knockout mice found that Sca-1 is required for self-renewal of MSC progenitors
and has an indirect role in osteoclast differentiation (Bonyadi et al., 2003).
The results in this study are consistent with previous results in MSCs (Li et al., 2008),
(Lamoury et al., 2006), (Peister et al., 2004). Although most studies did not find their
MSC- containing populations express Sca-1 to the high intensity that the clones did. One
126
reason for this may be that the clones are a pure population of MSCs (as discussed in
chapter 3), whereas the other studies typically involved use of mixed cell populations.
Another group reported their BALB/c MSCs were negative for Sca-1 (Peister et al.,
2004). This could be for the reasons just stated. In contrast to Sca-1, clones expressed low
levels of the fibroblast marker CD90 which the MEFs expressed at significant levels. The
clones and MEFs were positive for the haematopoietic antigen CD34 (Peister et al.,
2004), and to a lesser extent, CD45. Although human MSCs are reported to be negative
for CD34 and CD45, CD34 expression by murine MSCs has been reported (Peister et al.,
2004). Expression of CD45 is more controversial as most groups report murine MSCs as
CD45 negative (Jiang et al., 2002). However, the Phinney group reported positive
expression of CD45 (Phinney et al., 1999). This group report much higher CD45
expression levels, which they postulate is due to lympho-haematopoietic contamination,
whereas the clones in our study expressed relatively low levels. As the cells are clonallyderived and homogeneous, as evidenced by their morphology and the FACS data, we
believe contamination is unlikely to be the reason for the low level of CD45 expression.
It could simply be due to the fact that pure populations of MSCs cultured at more
physiological oxygen tension (5%) do, in fact, express this marker, albeit at low levels.
Many groups have reported gain or loss of markers with culture (Eslaminejad et al.,
2007). We found that our clones retained their surface antigen profile when examined up
to passage 53. Again this could be due to the reasons mentioned above - our MSCs are
clonally-derived and hence pure populations, and furthermore were derived and cultured
in more physiologic oxygen conditions.
The ability of multipotent MSCs to expand significantly in culture is one of the reasons
that they are such an attractive candidate for tissue engineering (Abdallah and Kassem,
2008). However, there is a paucity of literature on the growth kinetics of clonal adult
stem cells as most studies are performed on heterogeneous populations. The growth rate
of our clones was followed from passage 0 until passage 58 for the BALB/b and BALB/c
clones in 5% oxygen tension (figure 4.3). This was a period of over 6 months by which
point the clones reached over 180 population doublings. The results demonstrated that
there is a linear increase in the numbers of population doublings of the clones with time
in culture. Therefore the growth rate was constant, i.e., there was no sign of a senescence127
associated decrease in growth with passage. Previous studies on the growth rate of MSCs
have reported that the initial growth rate was fast, but slowed down with passage and
cells began to senescence after 50 population doublings (Muraglia et al., 2000). In
contrast, the cell doubling time of our clones at both high and low passage is the same
with no sign of senescence.
In a recent study a murine ‘MSC’ culture called multipotent adult progenitor cell (MAPC)
was expanded over 120 population doublings (Jiang et al., 2002). This population was
grown in the presence of high concentrations of exogenous growth factors (EGF, LIF,
and, PDGF-BB), whereas it is important to point out that no such factors were added to
our cultures. A further difference was that we culture the cells in 5% oxygen, whereas the
above studies were performed in 20% oxygen. Indeed, cultivation of clones at low
oxygen tension had an impact on the growth rate. Clones that were cultured at 5% oxygen
tension consistently were more proliferative in comparison to 20% and 1% cultures
(figures 4.4- 4.5). Taken together the results demonstrate that low oxygen tension (5%)
enhances the growth rate of these clonally-derived, multipotent murine MSCs.
There are two types of cellular senescence, one is ‘replicative senescence’, i.e., reaching
the ‘Hayflick limit’. The Hayflick limit is the number of times a cell will divide before
senescing due to critical shortening of the telomeres. (52 population doublings) (Hayflick
and Moorhead, 1961). The other is ‘stress-induced’ senescence caused by extrinsic
factors. This is ‘premature’ senescence as it may occur before the Hayflick limit is
reached. Culturing cells in 20% oxygen is likely to be an oxidative stress to cells, as
MSCs are exposed to much lower oxygen tension in vivo (Kofoed et al., 1985). This
hypothesis was tested by Moussavi-Harami and colleagues for MSCs and chondrocytes
(which are exposed to an even lower oxygen tension in vivo, <2% oxygen) (MoussaviHarami et al., 2004). The group found that cultivation of three different human
chondrocyte populations at 5% oxygen was associated with a two-fold increase in growth
rate compared to 20% oxygen cultures. A similar increase was found with human MSCs.
The investigators also found cells cultured at high oxygen levels were associated with
significantly greater oxidant production than low oxygen cultures. Their results support
the hypothesis that high oxygen levels (20%) used in standard cell culture conditions are
128
harmful to cells that are normally exposed to much lower oxygen levels in vivo. These
conditions induce premature senescence as chondrocytes cultured at 5% oxygen reached
an average of 77 population doublings, while cells cultured at 20% oxygen senesced at 38
population doublings. Initial studies by other groups also report enhanced growth of
MSCs cultured at low oxygen compared to the standard 20% oxygen tension (Fehrer et
al., 2007; Grayson et al., 2006; Lennon et al., 2001). However, we are the first to study
the effect of three different oxygen tensions on MSCs so extensively.
As mentioned in the introduction, standard cell culture conditions of 20% oxygen tension
were developed for the cultivation of fibroblasts and then cells lines. While high oxygen
conditions may be suitable for these cell types, a more specialised environment is
required for cells such as MSCs and chondrocytes in order to prevent oxidant-induced
premature senescence. In addition, cultivation at low oxygen tension has been reported to
cause re-differentiation of chondrocytes (Murphy and Polak, 2004) and maintain the
‘stemness’ of MSCs (D'Ippolito et al., 2006). As discussed in chapter 3 low oxygen
enhanced and maintained the differentiation capacity of the clonal cells in the present
study. Thus, oxygen level is an elementary signal to cells, low oxygen tension does not
just enhance the growth rate of cells, it also maintains their intrinsic biology. This could
be why the growth of clonal cells is enhanced at low oxygen tension (5%). We are
minimising culture induced stress, and thus ‘premature senescence’. However, it should
be noted that growth of clonal cells at 1% oxygen is lower than 5% oxygen. As MSCs are
not thought to be exposed to such a low oxygen tension in vivo, perhaps the optimal
results at the more physiologically relevant 5% oxygen are not surprising.
Stem cells appear to have the ability to prevent or delay replicative senescence, allowing
them to continue dividing throughout the lifespan of the organism. One mechanism by
which this is achieved is the telomere length is maintained by the enzyme telomerase.
Telomeres are guanine-rich DNA repeat sequences on the ends of every eukaryotic
chromosome that provide chromosomal stability (Hiyama and Hiyama, 2007). Each
round of cellular replication causes their loss due to incomplete replication of linear
chromosomes, called the ‘end replication problem’. Telomerase is a remarkable enzyme
129
as it is composed of a reverse transcriptase protein and a template RNA TERC (template
RNA component). Telomerase adds telomeric repeats to the ends of the telomere to help
maintain the length. The telomere length of the clonal cells was maintained from 18 to
200 population doublings. Furthermore, the clonal cells expressed high levels of
telomerase enzyme (compared to MEFs) throughout culture. This could, therefore be one
mechanism by which clonal cells avoided the Hayflick limit.
Recent studies have reported that murine MSCs acquire chromosomal abnormalities
whilst in culture and subsequently form tumours when injected into immunocompromised mice (Miura et al., 2006) (Tolar et al., 2007) (Zhou et al., 2006). Hence the
karyotype of the clonal cells in this study was examined. Cytogenetic analysis revealed
all clonal cells did, in fact have an abnormal karyotype of approximately 69
chromosomes usually with the addition of 1 or 2 mini-chromosomes. The presence of
these mini-chromosomes was also reported by Miura and colleagues who called them
double-minute chromosomes and speculated that they were associated with elevated
levels of c-myc (Miura et al., 2006). Little is known about mini-chromosomes, one
hypothesis being that they are amplifications of oncogenes.
Notably, the chromosome number of 69 per cell present in the clones was constant from
18 to180 population doublings. Therefore assuming the initial single bone marrowderived stem cell had a normal karyotype, the cytogenetic alterations occurred during the
first month of culture. Since the MEFs had a normal karyotype and were seeded at a high
initial density it is possible that the high selection pressure of seeding cells at a low
density to form clones may have caused the chromosomal alteration. However, an
abnormal karyotype was also observed in high density, non-clonally derived first passage
FVB/N bone marrow adherent cultures. Thus, clonal selection pressure cannot be the
only reason for the cytogenetic alterations, and, it seems the transformation was not a
strain-specific event. It is possible a combination of low oxygen tension and low density
seeding was responsible for the observed chromosomal instability. However, BALB/c
BMMNCs that were cultured at a high density in 20% oxygen were tetraploid, whereas
the equivalent cells cultured at 5% oxygen were normal, although beginning to show
signs of transformation with the presence of a cell with 59 chromosomes and 1 minichromosome.
130
It is difficult to understand the mechanism by which clonal cells gained the abnormal
karyotype of 69 chromosomes. Zhou and colleagues also found their MSCs gained a
similar number of chromosomes, although theirs ranged from 60-84 (Zhou et al., 2006).
The investigators postulated that transformation might have occurred via cell fusion or
chromosome replication without cytokinesis. However, this would result in tetraploid
cells, therefore there must have been an additional event where approximately 11
chromosomes were lost. The wide range of chromosome number present in the cells of
Zhou’s group suggests continued chromosomal instability. Whereas, in the present study,
once the clonal cells had gained 69 chromosomes this number remained stable through
extensive subculture.
Parinello and colleagues suggest that reduced oxygen tension may halt or delay
proliferative senescence due to less oxidative stress-induced DNA damage (Parrinello et
al., 2003). They also claim that the greater ability of human cells to prevent or repair
oxidative DNA damage contributes to the major differences in the incidence of cancer
and the rate of ageing between mice and humans. Previous studies have also reported that
mouse cells spontaneously immortalize in culture (Todaro and Green, 1963), (Prowse and
Greider, 1995). This has been the case as far back as the early studies of Earle and
colleagues in 1943 (Earle, 1962). Prowse and colleagues suggest mouse cells are more
prone to immortalization and transformation in culture than human cells, due to the
presence of high telomerase activity in most adult mouse tissue in comparison to human
(Prowse and Greider, 1995). Telomeres in human cells consistently shorten with age in
vivo and in vitro, whereas in the adult mouse the telomeres in different tissues vary in
length. The investigators found mouse skin fibroblasts transformed spontaneously in
culture and these cells had longer telomeres and higher telomerase activity compared to
untransformed cells.
Another reason murine cells acquire transformations quite rapidly is because they have
telocentric chromosomes where the centromere is at the end of the chromosome. The
biological advantage for this is unknown and it makes identification of mouse
chromosomes much more difficult than with human chromosomes. Telocentric
chromosomes are more likely to undergo chromosomal translocations, such as
131
Robertsonian translocation (Kalitsis et al., 2006). In humans only the Y chromosome is
telocentric.
In agreement with the above studies we did not detect any chromosomal instability in
human MSC cultures (discussed in chapter 6), and indeed few studies report
transformation of human cells. In mouse cultures it is a common event, but it was
reported in only 1 out of 40 samples in human cultures (Wang et al., 2005b). The
investigators reported spontaneous immortalization of a clonally distinct, highly
proliferative population at passage 3. Wang and co-workers found higher telomerase
activity compared to normal human MSC cultures. However, Rubio and colleagues
reported it took place after long-term in vitro culture (4-5 months), where cells went
through cell crisis phase and then expressed increased levels of c-myc (Rubio et al.,
2005). Also there was no telomerase activity detected in normal human MSCs, while it
was found in transformed MSCs. In addition, studies reporting chromosomal abnormality
in human MSCs usually note the addition and/or translocation of 1 or 2 chromosomes,
never a tetraploidy or a 75% increase in chromosome number.
In most studies MSCs are not karyotyped. Interestingly, so-called MIAMI (marrowisolated adult multilineage inducible) cells (D'Ippolito et al., 2004), which are of human
origin show many similarities to the clonal MSC of the present study. Both are
established and maintained in low oxygen tension, and display a high proliferative
capacity over a prolonged period without evidence of senescence. In addition, neither cell
type produced tumours when injected in immuno-compromised mice. The karyotype of
MIAMI cells has not been investigated, but it would be interesting to discover if they
display any cytogenetic alterations. Multipotent adult progenitor cells (MAPCs) only
grow under very specific conditions in the presence of LIF, EGF, PDGF and have a high
proliferation capacity. Murine MAPCs (mMAPCs) have been cultured to over 120
population doublings as mentioned above. In addition, the telomere length was
maintained at 27 kilo base pairs (kbp) in cells at 40 population doublings to 102
population doublings (Jiang et al., 2002). Murine MSCs have been cultured to 100
population doublings (Meirelles Lda and Nardi, 2003), while our clonal cells were
cultured to 180 population doublings, but were found to be karyotypically abnormal. One
mMAPC population was hyperdiploid after 45 population doublings. No tumour
132
formation was found when mMAPCs were transplanted into immuno-deficient mice.
Human MSCs tend to proliferate to 40-50 population doublings (Baksh et al., 2004),
(Barry, 2003). Human MAPCs can be expanded for more than 80 population doublings
(Reyes et al., 2001). Cytogenetic analysis of two MAPC populations revealed a normal
karyotype, although this was at 40 population doublings. It will be interesting to see if
this was still the case at 80 population doublings since the majority of human MSCs tend
to only proliferate to 40-50 population doublings. The telomere length of one donor was
also examined from 15 to 35 to 45 population doublings and reported to remain constant.
Most studies investigating telomere length in human MSCs report a gradual loss with
culture, typically 1kbp with 10 population doublings (Baxter et al., 2004; Parsch et al.,
2004). The lack of proliferative senescence and maintenance of telomere length are
suggestive of an immortalized cell, and it is possible this may be the mechanism for the
high proliferative capacity of the clones.
To retain stem cell properties it would, perhaps, be ideal to avoid tissue culture
altogether. However, due to the low frequency of MSCs in vivo it is necessary to expand
them for research and for tissue repair purposes. One other explanation for the recent
evidence of chromosomal abnormality observed in murine and human MSCs could be
that in vivo the bone marrow MSCs exist in a dynamic environment surrounded by stem
and differentiated cells. There is a constant interplay between the cells and the
environment. In in vitro tissue culture although we are attempting to mimic aspects of the
in vivo environment by culturing cells at a low oxygen tension we cannot truly mimic the
in vivo niche. Cultivation at low oxygen seems appropriate for colony formation,
proliferation, differentiation, and self-renewal of MSCs as observed in chapter 3.
However, we are not adding any growth factors in order not to bias the cell type and
presumably MSCs are exposed to a variety in their in vivo niche. In addition, we are
selecting for highly proliferative MSCs, whilst in vivo MSCs are mainly in a quiescent
state.
In vitro, the stem cell is out of its original environment, and in consequence is quite
possibly under a significant amount of stress and will not behave the same way it would
in its natural environment. In vitro conditions may also introduce experimental artifacts.
Thus, it is not entirely surprising that the cells began to acquire genetic instabilities when
133
cultivated in such a foreign environment. Results highlight the need for quality control
measures for MSCs before clinical use.
As mentioned above, studies which found chromosomal abnormalities in their MSCs,
then transplanted them into immuno-compromised mice to investigate whether the cells
were malignant (Miura et al., 2006; Tolar et al., 2007; Wang et al., 2005b). The
investigators found that sarcomas developed. In the present study, four clonally-derived
populations from two mouse strains were injected locally or intra-venously into severe
combined immunodeficiency (SCID) mice. No tumour formation was observed at 4½
months (BALB/c) and 7 months (BALB/b) post-transplantation. Although, it cannot be
ruled out that tumours may occur at later time-points, others have reported tumour
formation within 3-4 weeks following local injection (Miura et al., 2006; Tolar et al.,
2007; Wang et al., 2005b). Sarcomas have been reported to have complex karyotypes and
are chromosomally unstable in vivo and in culture (Tolar et al., 2007). In contrast, the
clonal cells in this study had a stable karyotype after the initial transformation. A major
difference between our clonal cells and the transformed and malignant cells in the above
studies is that our clonal cells were derived and maintained at low oxygen tension. Also
the clonal cells did not express significant levels of c-myc. It is possible that culturing
cells at low oxygen prevented DNA damage and mutations caused by reactive oxygen
species (ROS). Loss of DNA repair mechanisms has been reported to make cells more
susceptible to malignant transformation directly or by the occurrence of cancer-prone
stem cells (Kenyon and Gerson, 2007).
C-myc is a proto-oncogene, dysregulated expression of c-myc is found in many human
cancers (Pelengaris et al., 2002). It is notable that in the Miura study, that there was an
increase in c-myc levels and telomerase in the murine MSCs compared to the normal
MSCs where negligible levels were present (Miura et al., 2006). Transformed and
malignant human MSCs also expressed increased levels of c-myc, as well as VEGF
which is found in many tumours, in comparison to normal MSCs (Rubio et al., 2005).
Furthermore, c-myc is reported to activate telomerase activity (Wang et al., 1998).
However, microarray analysis of clonal cells in this study did not reveal increased c-myc
expression compared to controls, i.e., strain-matched BMACs, MEFs, and freshly isolated
134
BMMNCs. The reason for this could be that our cells, although immortalised, are not
malignant. It is possible with extended passaging that c-myc levels may increase and the
cells become malignant, however, it should be noted that cells cultured to passage 43 still
retained the spindle shaped morphology, ability to self-renew, and capacity to
differentiate. The surface antigen profile was also maintained. In addition, the cell
doubling time was constant as was the chromosome number per cell.
The malignant murine MSCs reported in other studies typically display a decrease in
growth rate when the cell is going through ‘crisis phase’ and then a morphology change
to a flattened cuboidal, cellular morphology at passage 3-5 (Miura et al., 2006). Crisis
phase is when the telomere length of cells becomes critically short causing chromosomal
instability which leads to apoptosis (Chin et al., 1999). The majority of the cells apoptose,
but there are few that do not and they gradually display an increased growth rate as they
overcome the crisis period and retain the large cell morphology. Zhou and colleagues
report there was an increase in the growth rate of their murine MSCs at passage 3
followed by contact inhibition and growth in 3-dimensional clusters (Zhou et al., 2006).
Human MSCs went through crisis phase 4-5 months after isolation (Rubio et al., 2005)
which is later than the murine MSCs but in agreement with previous studies (Lloyd,
2002). The cells overcame crisis phase with the increased growth rate described above.
The two significant differences between the clonal cells in this study and the abnormal
MSCs in the ones above are; first ours were clonal and derived and cultured at 5%
oxygen tension. The second is they did not express elevated levels of oncogene c-myc
and of VEGF that some of the cells in the above studies did. Although our clonal cells did
express telomerase and the transformed MSCs did also, this is a characteristic of stem
cells as well as transformed cells. Based on the above results we can conclude our clonal
cells are immortalised but not malignant, as they did not form tumours in vivo. The term
immortalisation was first used to describe cancer cells that express telomerase and are
therefore able to avoid apoptosis and divide continuously. It is more commonly applied to
cell lines Hela and Jurkat which are immortalized cancer cell lines, as well as stem and
germ cells. It is possible our clonal cells are a type of cell line.
135
Despite their abnormal karyotype clonal cells did not show characteristics of transformed
cells. This is evident by their differentiation which is a highly regulated process. Many
studies have reported the dysfunctional patterns of differentiation of transformed cells
(Bjerkvig et al., 2005). Our clonal cells expressed significant levels of Sca-1 compared
to control cells. Gene expression analysis revealed this virtually disappears in the
adipocyte differentiated clones. In contrast, adipocyte differentiated clones expressed
significant levels of PPAR-γ compared to undifferentiated clones. This is an adipocyte
marker, and is a member of the nuclear hormone receptor superfamily functioning as a
transcription factor regulating the expression of adipocyte genes (Tontonoz et al., 1994).
In combination with CEBP, PPAR-γ induces expression of adipocyte genes - adipsin,
fatty acid synthase, aP2, and GLUT4, which establishes terminal adipogenesis of cells
(Wu et al., 1999). These gene expression analysis results were significant as they were
the average of four clones and two control cell populations from each mouse strain; both
strains exhibiting very similar results.
As mentioned earlier Sca-1 is not only expressed by MSCs but also by HSCs and other
tissue-specific stem cells (Holmes and Stanford, 2007). Sca-1 also has a functional role in
stem cells. Upon differentiation of the clonal cells there was a dramatic decrease in Sca-1
expression. This was further confirmed at the protein level by flow cytometry, where
there was quite a dramatic loss of Sca-1 after just one day in adipogenic differentiation
media. This demonstrates how quickly the cells responded to extrinsic signals. After 2
weeks in the differentiation media there were negligible levels of Sca-1 present. In
contrast, to Sca-1 there was an increase in CD73 expression with adipogenesis. This is a
traditional MSC candidate marker, which was expressed at low levels by the clonal cells.
This increased slightly after one day in differentiation media and was expressed at
significant levels after complete differentiation. CD73 is, in fact, also expressed on a
variety of other cell types. It is also a marker which the International Society for Cellular
Therapy (ISCT) (who met to propose a guideline for the identification of MSCs) set as a
marker for MSCs (Dominici et al., 2006). Results stress the need for better markers as
this one is expressed on adipocytes but not MSCs in this study. Further microarray
analysis revealed differentiated cells expressed significantly greater levels (over 70 fold)
136
of adipocyte genes: fat specific gene 27, adipocyte complement related protein, fatty acid
binding protein 4, lipase, and fatty acid coenzyme A ligase. Concomitantly the stem cell
and cell cycle (associated) genes: Sca-1, Ccl7, Hgf, Fgf7 and Cdk4 were down regulated
in the differentiated clones compared to the undifferentiated clones.
Also, as shown in chapter 3, clonally-derived MSCs exhibited highly regulated patterns
of differentiation, and the extent of differentiation was very comprehensive. This is
shown histologically not just for adipogenesis, but for chondrogenesis and osteogenesis.
All clonal cells examined had this tri-lineage ability and this was maintained with
extensive passaging up to passage 43. Also clonal cells demonstrated contact inhibition,
as they formed a monolayer. There was no evidence of multiple layers, i.e. lack of
contact inhibition which is a characteristic of cancer (stem) cells. In contrast, when Zhou
and colleagues injected abnormal MSCs via intra-venous injection in immunocompromised mice lung tumours formed which contained areas of bone formation and
melanocytes (Zhou et al., 2006). The transformed cells differentiated into two different
lineages, which in turn were different from the tissue they were in, thus displaying
unregulated differentiation. Furthermore, Miura and colleagues found that their
transformed MSCs lost their osteogenic potential after they passed through the crisis
phase (Miura et al., 2006).
In conclusion, we believe the clonally-derived cells are a very useful model to investigate
MSC-specific markers since we observed murine MSC differentiation to be a highly coordinated and regulated process, despite chromosomal alterations in the cells. Any
candidate markers that are identified would be subsequently assessed in normal
multipotent, human MSCs.
137
Chapter 5
Molecular Signature of MSC Clones
138
5.1 Introduction
As there is no known MSC marker that can be used to distinguish the rare MSC
population from the many other cells in the bone marrow most groups work with a
heterogeneous population of bone marrow adherent cells which they call ‘MSCs’. Certain
groups have attempted to advance the understanding of MSC biology by characterising
MSCs using a microarray approach. However, the significance of the results has been
limited due to the paucity of the cell model and lack of suitable controls.
The major aim of this study was to identify MSC specific markers and regulators. To this
end we have derived multiple clonal populations from mouse bone marrow and
subsequently proven their stem cell status through functional analysis, i.e., by
demonstrating their ability to self-renew and to differentiate into multiple lineages. FACS
analysis demonstrated the cells have a distinct surface phenotype (in comparison to
fibroblast controls (chapter 4)).
In an effort to define the molecular signature of these cells an extensive microarray
approach was adopted. Potentially MSC specific genes were identified by comparing the
gene expression profile of the multipotent, self-renewing, clonal cells to various relevant
controls. The implications of our findings are discussed.
5.2 Microarray design
A transcriptomic approach was adopted in an attempt to identify markers unique to MSCs
(figure 5.1 shows the design implemented). In total, 28 whole genome-wide arrays were
used. The gene expression profile of eight (four per mouse strain) undifferentiated, multipotent, self-renewing clones were examined. We also utilised four distinct controls to
ensure the final gene list would be relevant, i.e., would enable us to better distinguish
novel, and potentially MSC-specific genes. The first control group consisted of the same
eight clones, but after differentiation into the adipocytes. Therefore these populations had
lost their stem cell properties. The second and third controls were strain-matched (i.e.,
BALB/b and BALB/c) mouse embryonic fibroblasts (MEFs) and bone marrow adherent
cells (BMACs). The former do not have the differentiation potential of the clones, while
the latter are a heterogeneous population of cells, containing, but not exclusively the
139
Control 4: BMMNC (n=2 per mouse strain)
Control 3: BMAC, mixed cell population (n=2 per mouse strain)
Control 2: MEF, does not have differentiation potential (n=2 per mouse strain)
Figure 5.1 Microarray experimental design used to characterise the molecular
signature of MSC clones derived for BALB/b and BALB/c mice. The gene expression
profile of undifferentiated, multipotent clones (for BALB/b and BALB/c) was compared to
4 distinct controls: the equivalent adipocyte differentiated clones; primary bone marrow
adherent cells (BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and
mouse embryonic fibroblasts (MEF). Two samples of BMAC, MEF, and BMMNC were
used for each mouse strain and four undifferentiated clones and differentiated clones per
strain. In total, 28 samples were analysed.
Undifferentiated
Clone
Un-differentiated, multipotent,
self-renewing clone
Control 1: Differentiated clone,
no stem cell properties
Differentiated Clone
(n=4 per mouse strain)
140
multipotent cells called ‘MSCs’. There were two biological replicates used for these
controls for each mouse strain.
5.3 Cell culture and microarray experiments
To avoid oxygen-induced gene expression changes, the cell culture for these experiments
was performed in the more physiologically relevant 5% oxygen tension. MEFs were
initially cultured at 20% oxygen but prior to microarray analysis cells were cultured at
5% oxygen tension for 4 days. BMMNCs were not cultured - they were lysed
immediately after removal of red blood cells. Clonal cells were always cultured in 5%
oxygen, as were the adipocyte-differentiated cultures. BMACs were cultured in 5%
oxygen for two weeks. Total RNA isolated from these cells was labelled and hybridized
to the GeneChip Mouse Genome 430 2.0 Array. The chip contains over 45,000 probe sets
representing over 34,000 well substantiated mouse genes as described in Materials and
Methods.
5.4 Principal component analysis
Principal component analysis (PCA) is a multivariate analysis technique that is used to
reduce multidimensional data sets to lower dimension for analysis. This enables detection
of major patterns in the dataset. The PCA of the 28 datasets (figure 5.2) very clearly
shows that the 8 undifferentiated clones are very similar and thus cluster very tightly,
distinct from each of the control populations including the differentiated clones. Each
control sample also clustered tightly together despite differences in mouse strain.
5.5 Unsupervised two-dimensional cluster analysis of microarray data
A two-dimensional (2D) hierarchical unsupervised cluster analysis was also performed
on the complete data set of 28 arrays. Results are shown in dendrogram form, and are
similar to that of the PCA mentioned above, i.e., each population of cells clustered tightly
together (figures 5.3-5.6). The red colour is indicative of relatively high gene expression
141
Figure 5.2 Principal component analysis of microarray data. Principal component
analysis of the whole-genome expression data of all 28 samples shows that the
undifferentiated clone (UC) (ringed) represent a unique cell type, distinct from each of the
four controls. Differentiated clones (adipocytes) (DC); primary bone marrow adherent
cells (BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse
embryonic fibroblasts (MEF).
UC (8/8)
DC (8/8)
MEF (4/4)
BMAC (4/4)
BMMNC (4/4)
Undifferentiated
Clones
142
Figure 5.3 Dendrogram showing Thy1/CD90 clustering. An unbiased, genome wide, two-dimensional
hierarchical cluster analysis was performed on the complete data set of 28 arrays as described in figure 5.1.
Shown are genes clustering closest to fibroblast marker Thy1/CD90. The color scale ranges from saturated red
log ratios + 1.8 decreasing to saturated green for log ratios -1.8. Each gene is represented by a single row of
colored boxes; each individual control is represented by a single column. Undifferentiated clone (UC);
differentiated clones (adipocytes) (DC); primary bone marrow adherent cells (BMAC); freshly isolated bone
marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF); BALB/b (B/b); BALB/c (B/c).
U
C
37 (B
/b)
U
C
36 (B
/b)
U
C
15 (B
/b)
U
C
5 (B/b)
U
C
3 (B/c)
U
C
35 (B
/c)
U
C
4 (B/c)
U
C
11(B
/c)
D
C
36 (B
/b)
D
C
37 (B
/b)
D
C
15 (B
/b)
D
C
5 (B/b)
M
EF1 (B/b)
M
EF1 (B/b)
M
E
F1 (B
/c)
M
EF2 (B/c)
D
C
3 (B/c)
D
C
35 (B
/c)
D
C
4 (B/c)
D
C
11 (B
/c)
B
M
AC
1 (B
/b)
B
M
AC
2 (B
/b)
B
M
AC
1 (B
/c)
B
M
A
C
2 (B
/c)
BM
M
N
C
1 (B/b)
BM
M
N
C
2 (B/b)
BM
M
N
C
1 (B/c)
BM
M
N
C
2 (B/c)
Thumbnail view
/CD90
143
U
C
5 (B/b)
U
C
37 (B
/b)
U
C
36 (B
/b)
U
C
15 (B
/b)
U
C
3 (B/c)
U
C
35 (B
/c)
U
C
4 (B/c)
U
C
11(B
/c)
D
C
36 (B
/b)
D
C
37 (B
/b)
D
C
15 (B
/b)
D
C
5 (B/b)
M
E
F1 (B
/b)
M
EF1 (B/b)
M
E
F1 (B
/c)
M
E
F2 (B
/c)
D
C
3 (B/c)
D
C
35 (B
/c)
D
C
4 (B/c)
D
C
11 (B
/c)
B
M
AC
1 (B
/b)
B
M
AC
2 (B
/b)
B
M
AC
1 (B
/c)
B
M
A
C
2 (B
/c)
BM
M
N
C
1 (B/b)
BM
M
N
C
2 (B/b)
BM
M
N
C
1 (B/c)
BM
M
N
C
2 (B/c)
Thumbnail view
Figure 5.4 Dendrogram showing PPARγ clustering. An unbiased, genome wide, two-dimensional hierarchical
cluster analysis was performed on the complete data set of 28 arrays as described in figure 5.1. Shown are genes
clustering closest to adipocyte markers PPARγ and PPARγc1b. PPARγ: peroxisome proliferator-activated receptorgamma. The color scale ranges from saturated red log ratios + 1.8 decreasing to saturated green for log ratios -1.8.
Each gene is represented by a single row of colored boxes; each individual control is represented by a single
column. Undifferentiated clone (UC); differentiated clones (adipocytes) (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF);
BALB/b (B/b); BALB/c (B/c).
144
Thumbnail view
U
C
37 (B
/b)
U
C
36 (B
/b)
U
C
15 (B
/b)
U
C
5 (B/b)
U
C
3 (B/c)
U
C
35 (B
/c)
U
C
4 (B/c)
U
C
11(B
/c)
D
C
36 (B
/b)
D
C
37 (B
/b)
D
C
15 (B
/b)
D
C
5 (B/b)
M
E
F1 (B
/b)
M
EF1 (B/b)
M
E
F1 (B
/c)
M
EF2 (B/c)
D
C
3 (B/c)
D
C
35 (B
/c)
D
C
4 (B/c)
D
C
11 (B
/c)
B
M
AC
1 (B
/b)
B
M
AC
2 (B
/b)
B
M
AC
1 (B
/c)
B
M
A
C
2 (B
/c)
BM
M
N
C
1 (B/b)
BM
M
N
C
2 (B/b)
BM
M
N
C
1 (B/c)
BM
M
N
C
2 (B/c)
Figure 5.5 Dendrogram showing Csf1r clustering. An unbiased, genome wide, two-dimensional hierarchical
cluster analysis was performed on the complete data set of 28 arrays as described in figure 5.1. Shown are genes
clustering closest to stromal cell marker Csf1r: colony stimulating factor 1- receptor. The color scale ranges from
saturated red log ratios + 1.8 decreasing to saturated green for log ratios -1.8. Each gene is represented by a single
row of colored boxes; each individual control is represented by a single column. Undifferentiated clone (UC);
differentiated clones (adipocytes) (DC); primary bone marrow adherent cells (BMAC); freshly isolated bone
marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF); BALB/b (B/b); BALB/c (B/c).
145
U
C
37 (B
/b)
U
C
36 (B
/b)
U
C
15 (B
/b)
U
C
5 (B/b)
U
C
3 (B/c)
U
C
35 (B
/c)
U
C
4 (B/c)
U
C
11(B
/c)
D
C
36 (B
/b)
D
C
37 (B
/b)
D
C
15 (B
/b)
D
C
5 (B/b)
M
E
F1 (B
/b)
M
EF1 (B/b)
M
E
F1 (B
/c)
M
EF2 (B/c)
D
C
3 (B/c)
D
C
35 (B
/c)
D
C
4 (B/c)
D
C
11 (B
/c)
B
M
AC
1 (B
/b)
B
M
AC
2 (B
/b)
B
M
AC
1 (B
/c)
B
M
A
C
2 (B
/c)
BM
M
N
C
1 (B/b)
BM
M
N
C
2 (B/b)
BM
M
N
C
1 (B/c)
BM
M
N
C
2 (B/c)
Thumbnail view
Figure 5.6 Dendrogram showing BMMNC highly expressing genes. An unbiased, genome wide, twodimensional hierarchical cluster analysis was performed on the complete data set of 28 arrays. The color scale
ranges from saturated red log ratios + 1.8 decreasing to saturated green for log ratios -1.8. Each gene is
represented by a single row of colored boxes; each individual control is represented by a single column.
Undifferentiated clones (UC); differentiated clones (adipocytes) (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF);
BALB/b (B/b); BALB/c (B/c).
146
and green of low level expression.
For example, as shown in figure 5.3 where the fibroblast marker CD90 is
highlighted, all MEF samples showed high (red) expression, compared to all other cell
types. Thus, highlighting, visually the similarity in gene expression profile of the MEF
samples, irrespective of mouse strain differences. This striking similarity within control
groups is repeated for BMACs taking Csfr1 as an example; and the adipocytedifferentiated clones (PPARγ), and for BMMNCs (figures 5.4-6).
With regard to adipocyte differentiation, BALB/c differentiated clones are less
positive compared to the BALB/b for three of the clones (i.e., DC3, DC35 and DC4 – see
figure 5.4). This can also be clearly seen in the PCA analysis where three of the DC
samples are closer to the UCs than the remaining five DCs (see figure 5.2). Furthermore
these findings did indeed match our observation of a more slow differentiation in these
specific three cultures (as assessed by Oil Red O staining). Nevertheless, the impressive
extent of adipocytic differentiation is highlighted by the fact that MEFs cluster closer to
the undifferentiated clones than same clones after undergoing adipogenesis (e.g., see
figure 5.4).
5.6 Supervised analysis of microarray data
We performed a stringent analysis selecting for genes expressed on all clones and absent
or at least 3 fold lower on the same differentiated clones (figure 5.7A). For the other
controls we selected for genes present on all the clones and absent or at least 3 fold lower
on each control, e.g. MEF1 and MEF2. This criterion was repeated for the BMAC and
the BMMNC samples. For the latter a 10 fold cut-off was chosen since MSCs are thought
to be present at a very low level in this population (Castro-Malaspina et al., 1980). This
gave a total of 28 comparisons of the clones to the various controls where these
conditions had to be met in each case, for cultures from each strain. A final list of 78 and
63 transcripts was obtained for the BALB/b and BALB/c data sets, respectively. An
intersection was taken of the resulting genes from each data set. This rigorous analysis
147
Fi gu re 5
.7
Sc he m at ic o f m
ic ro ar ra y an al ys is to fi nd g en es sp ec ifi
c to th e M
SC
u nd iff
er en tia
te d cl on e. (F
ig ur e le ge nd n ex t p
ag e) 78
g en es B
A
L
B
/b >
3
fo ld h ig he r on U
C
>
3
fo ld h ig he r on U
C
>
3
fo ld h ig he r on U
C
>
10
fo ld hi gh er o n U
C
U
C
5
D
C
5
U
C
15
D
C
15
U
C
36
D
C
36
U
C
37
D
C
37
U
C
5
M
EF
1
M
EF
2
U
C
15
M
EF
1
M
EF
2
U
C
36
M
EF
1
M
EF
2
U
C
37
M
EF
1
M
EF
2
U
C
5
B
M
A
C
1
B
M
A
C
2
U
C
15
B
M
A
C
1
B
M
A
C
2
U
C
36
B
M
A
C
1
B
M
A
C
2
U
C
37
B
M
A
C
1
B
M
A
C
2
U
C
5
U
C
15
U
C
36
U
C
37
B
M
M
N
C
1
B
M
M
N
C
2
B
M
M
N
C
1
B
M
M
N
C
2
B
M
M
N
C
1
B
M
M
N
C
2
B
M
M
N
C
1
B
M
M
N
C
2
A
U
C
3
D
C
3
U
C
4
D
C
4
U
C
11
D
C
11
U
C
35
D
C
35
U
C
3
M
EF
1
M
EF
2
U
C
4
M
EF
1
M
EF
2
U
C
11
M
EF
1
M
EF
2
U
C
35
M
EF
1
M
EF
2
B
M
A
C
1
B
M
A
C
2
B
M
A
C
1
B
M
A
C
2
B
M
A
C
1
B
M
A
C
2
B
M
A
C
1
B
M
A
C
2
B
M
M
N
C
1
B
M
M
N
C
2
B
M
M
N
C
1
B
M
M
N
C
2
B
M
M
N
C
1
B
M
M
N
C
2
B
M
M
N
C
1
B
M
M
N
C
2
63
g en es B
A
L
B
/c U
C
3
U
C
4
U
C
11
U
C
35
U
C
3
U
C
4
U
C
11
U
C
35
>
3
fo ld h ig he r on U
C
>
3
fo ld h ig he r on U
C
>
3
fo ld h ig he r on U
C
>
10
fo ld hi gh er o n U
C
148
72
57
6
BALB/b
B
BALB/c
Total
genes
Cell membrane
(associated)
genes
3
1
5
Figure 5.7 (continued) Schematic of Microarray Analysis to Find Genes
Specific to the MSC Undifferentiated Clone . Genes were selected which were
present on all 4 BALB/b UC and expressed at least 3- fold higher than DC, MEF, and
BMAC and 10-fold higher than BMMNC (previous page) (A). This gave a final list
of 78 genes. This was then repeated for the BALB/c strain and yielded a gene list of
63 genes. (B) Venn diagram showing the total number of genes that are shared
between the BALB/b and BALB/c strain, yielding a final unique list of 6 total genes
specific to the MSC undifferentiated clone. Another venn diagram was derived
showing the number of cell membrane (associated) genes with an intensity greater
than 100 (arbitary units) shared between the BALB/b and BALB/c strain, yielding a
final unique list of 5 genes specific to the MSC undifferentiated clone. Differentiated
clones (adipocytes) (DC); primary bone marrow adherent cells (BMAC); freshly
isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic
fibroblasts (MEF). , and the cell membrane (associated) genes, with intensity greater
than 100 (arbitary units)
149
gave a final list of 6 genes (figure 5.7B). Another venn diagram was derived showing the
number of cell membrane (associated) genes with an intensity greater than 100 (arbitrary
units) shared between the BALB/b and BALB/c strain. Interestingly, this showed that 5
of the 6 genes are cell membrane (associated) genes. These results are discussed in
section 5.7 below.
Classification of the subcellular localisations of the genes based on annotations using the
DAVID (Database for Annotation, Visualization and Integrated Discovery) gene
ontology software was performed. Several classes of enriched proteins were revealed:
surface markers, secreted proteins, cell cycle-associated genes, nuclear proteins, and
intracellular proteins. DAVID provides a comprehensive set of functional annotation
tools which allows identification of genes not identified by Rosetta Resolver®, as well as
information about gene function, and cellular location. Genes with signal intensities
greater than 100 (arbitrary units) in the undifferentiated clones are shown (tables 5.1-5).
The average signal intensity for each sample type is shown, thus the 3 fold or 10 fold cutoff is not obvious in all cases.
5.6.1 Supervised analysis of microarray data for BALB/b clones
Twenty-four percent of the BALB/b genes were cell membrane or membrane-associated
proteins, and these included known MSC genes such as versican and unknown genes
such as six-transmembrane epithelial antigen of the prostate 2 (STEAP2), STEAP1 and
IL1Ri, and interestingly Ly6f, a member of the Ly6 family which includes Ly6a/Sca-1
(table 5.1).
Secreted proteins vouched for ten percent of the gene list containing, Ccl7, proliferin, and
Cxcl5, which have not been reported in MSCs, although Cxcl5 has been found in germline stem cells (Mizukami et al., 2008) (table 5.2). Hepatocyte growth factor (HGF) was
also expressed by the UCs. Four percent of the genes were cell cycle-associated genes
including FGF7 and CXCL1 which have been identified in human MSCs (Djouad et al.,
2007; Imabayashi et al., 2003) (table 5.3). A further 11% were nuclear proteins (table
5.4), one of which is vitamin D receptor (Vdr), an osteogenic transcription factor.
Another transcription factor, T-box, 18 about which little is known was also identified. A
150
Table 5.1 BALB/b cell membrane (associated) genes
Cell surface associated genes were selected which were present on all 4 BALB/b UC and absent or expressed at
least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than BMMNC as described
in figure 5.7. The sub-cellular localisation of the genes was classified using the DAVID gene ontology software.
Genes are ranked by their expression intensity in UC (arbitrary units). Undifferentiated clone (UC); adipocyte
differentiated clones (DC); primary bone marrow adherent cells (BMAC); freshly isolated bone marrow
mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
Chondroitin sulphate
proteoglycan 2 (versican) Cspg2 NM_019389 2685 198 264 135 19
Six transmembrane epithelial
antigen of the prostrate 2 STEAP2 AK015015 2085 184 522 242 0
Six transmembrane epithelial
antigen of the prostrate 1 STEAP1 AF297098 2082 273 307 261 1
Interleukin 1 receptor, type i Il1ri NM_008362 395 46 72 0 7
Lymphocyte antigen 6 complex,
locus F Ly6f NM_008530 267 2 25 6 13
Synaptotagmin xvii Syt17 AW048713 188 49 26 25 10
Transmembrane protein 100 Tmem100 NM_026433 182 0 8 0 0
Histocompatability 2, m region
locus 9 H2-M9 NM_008205 107 8 0 14 0
Table 5.2 BALB/b secreted proteins
Genes encoding for secreted proteins were selected which were present on all 4 BALB/b UC and absent or
expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
Chemokine (C-C motif) ligand 7 Ccl7 AF128193 1825 20 180 175 14
Proliferin Plf X75557 1443 37 0 82 15
Chemokine (C-X-C motif)
ligand 5 Cxcl5 NM_009141 869 0 141 21 0
Fibulin 1 Fbln1 BC007140 628 74 138 41 5
Hepatocyte growth factor Hgf AF042856 529 32 88 85 0
151
Table 5.3 BALB/b cell cycle associated genes
Genes associated with the cell cycle were selected which were present on all 4 BALB/b UC and absent or
expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
Fibroblast growth factor 7 Fgf7 BB791906 2858 189 823 760 0
Chemokine (C-X-C motif)
ligand 1 Cxcl1 NM_008176 953 37 109 101 0
Table 5.4 BALB/b nuclear proteins
Genes encoding for nuclear proteins were selected which were present on all 4 BALB/b UC and absent
or expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
vitamin D receptor Vdr AK018206 992 1 93 138 0
transducin-like enhancer of
split 2, homolog of Drosophila
E(spl)
Tle2 AK008793 227 10 58 57 2
t-box 18 Tbx18 BQ175889 225 21 84 41 21
152
Another transcription factor, T-box 18 about which little is known was also identified. A
large proportion of the data set (35%) consisted of genes encoding other intracellular
proteins (table 5.5). However, the majority had low signal intensity. Ellis van creveld
gene homolog (human) had the highest intensity (218) and has not previously been
reported in MSCs. The remaining 17% were unknown genes, or hypothetical proteins.
5.6.2 Supervised analysis of microarray data for BALB/c clones
The BALB/c data set share some of the genes present in the BALB/b dataset and are
discussed below in 5.7. A large proportion of the data set (27%) were surface markers,
including STEAP 1 and -2, versican, and Ly6f (table 5.6). As well as mas-related gpr,
membrane f protein, and histocompatability 2, m region locus 9 (H2-M9) which have not
previously been identified in MSCs, and also little is known about them.
Nine percent were secreted proteins (table 5.7), and these included insulin-like growth
factor binding protein 6 (IGFBP6), which had a high signal intensity in the
undifferentiated clones. Angiopoietin 4 and HGF were also identified. Genes encoding
nuclear proteins comprised 16% of the final list (table 5.8) - Vdr, cyclin- dependent
kinase inhibitor 2a (Cdkn2a), glis family zinc finger 3 (Gfzf3), t-box 18, and transducinlike enhancer of split 2, split homolog of drosphila E (spl) (Tle2) (table 5.8). The last
three have not been previously reported in MSCs. Twenty percent consisted of other
intracellular proteins, and these included cellular retinoic acid binding protein II
(Crabp2), ankyrin 3 epithelial, cystatin E/M (cysteine protease inhibitors) (not previously
identified in MSCs) (table 5.9). Twenty-seven percent of the data set represented
unknown or hypothetical genes. Interestingly Sca-1 is not present on the surface marker
list for either BALB/b or BALB/c data sets. This was due to the stringent criteria set for
potential MSC-specific gene detection. FACS analysis from chapter 4 (table 4.2-4.3)
showed that there was detectable Sca-1 expression in MEFs, and the microarray data for
Sca-1 expression did not make the 3 fold cut-off imposed, thus excluding it from the final
list.
153
Table 5.5 BALB/b intracellular genes
Genes encoding for intracellular proteins were selected which were present on all 4 BALB/b UC and absent
or expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
ellis van creveld gene
homolog (human) Evc BC021483 218 44 98 23 10
echinoderm microtubule
associated protein like 1 Eml1 BC021917 165 7 37 30 1
centrosomal protein 55 cp55 NM_021292 119 0 32 33 37
carboxylesterase 1 Ces1 BC007140 109 1 10 7 19
coiled-coil containing 21
(tsf) Ccdc21 BB147678 102 40 103 114 0
Table 5.6 BALB/c cell membrane (associated) genes.
Cell surface associated genes were selected which were present on all 4 BALB/c UC and absent or
expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
Six transmembrane epithelial
antigen of the prostrate 2 STEAP2 AK015015 1343 226 337 252 0
Six transmembrane epithelial
antigen of the prostrate 1 STEAP1 AF297098 1253 413 156 270 4
Chondroitin sulphate proteoglycan 2
(versican) Cspg2 NM_019389 395 251 85 28 1
Mas-related gpr, membrane f Mrgprf BC019711 344 103 62 54 3
Histocompatability 2, m region
locus 9 H2-M9 NM_008205 120 17 0 8 11
Lymphocyte antigen 6 complex,
locus F Ly6f NM_008530 103 61 17 33 2
154
Table 5.7 BALB/c secreted proteins
Genes encoding for secreted proteins were selected which were present on all 4 BALB/c UC and absent or
expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
Insulin-like growth factor binding
protein 6 Igfbp6 NM_008344 3492 453 520 487 1
Angiopoietin 4 Angpt4 NM_009641 348 66 32 31 1
Hepatocyte growth factor Hgf D10212 162 106 38 56 1
Table 5.8 BALB/c nuclear proteins
Genes encoding for nuclear proteins were selected which were present on all 4 BALB/c UC and absent or
expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
vitamin D receptor Vdr AV328983 609 154 73 149 0
cyclin-dependent kinase inhibitor 2a Cdkn2a BB709140 530 207 87 86 15
glis family zinc finger 3 gfzf3 BB036542 300 172 99 128 3
t-box 18 Tbx18 AK019867 147 158 41 101 11
transducin-like enhancer of split 2,
split homolog of drosphila E (spl) Tle2 NM_007680 133 24 46 5 7
155
Table 5.9 BALB/c intracellular proteins
Genes encoding for intracellular proteins were selected which were present on all 4 BALB/c UC and absent
or expressed at least 3-fold higher than the equivalent DC, MEFs, and BMACs and 10- fold higher than
BMMNC as described in figure 5.7. The sub-cellular localisation of the genes was classified by DAVID
gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary units).
Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts
(MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
cellular retinoic acid binding
protein II Crabp2 AV280494 1812 410 400 101 0
ankyrin 3, epithelial Ank3 BB709140 465 206 82 138 0
cystatin E/M (cysteine protease
inhibitors) Cst6 BB093349 416 117 41 43 1
156
5.7 Supervised analysis yields six MSC-specific candidate genes
Comparing the datasets from the two mouse strains, six genes were expressed at a
differentially high level in all the UCs compared to the various controls – six
transmembrane epithelial antigen of the prostate (STEAP1 and STEAP2), versican,
vitamin D receptor, ly6f, and H2-M9. As the graphs in figure 5.8 show, these genes
exhibit a strikingly similar gene expression pattern. They are all preferentially expressed
in undifferentiated clones and at relatively low levels in all controls suggesting they are
potential MSC-specific genes. Amazingly, of the six genes identified, five are cell
membrane (or cell membrane associated protein in the case of versican). Indeed, the sixth
gene, being vitamin D receptor is also a membrane protein (nuclear). T-box 18 and HGF
which are present in both datasets were not present in the final list as the average
intensity of both did not fulfil the set criteria.
Of the six genes identified, STEAP1 and STEAP2 are expressed at the highest intensity.
These genes were first identified in prostate cancer cells (Hubert et al., 1999; Porkka et
al., 2002), and this is the first mention of STEAP 1 and/or 2 in MSCs. Ly6f is a member
related to Ly6a/Sca-1, but was expressed at a low level (185 arbitrary units). H2-M9 is a
major histocompatability class Ib gene, and also has relatively low signal intensity (114).
Vitamin D receptor (Vdr) has a greater signal intensity of 559 in the UCs. It is expressed
on many cell types including MSCs, osteoblasts, mammary cells, and cells of the immune
system. Versican is relatively highly expressed (1027), but is also known to be expressed
on many other cell types including chondrocytes and osteoblasts.
5.8 STEAP1 and -2 cluster with Sca-1
An unbiased, genome-wide, 2D hierarchical cluster analysis was performed on the
complete data set of 28 arrays. Genes clustering with Sca-1 are shown in figure 5.9.
STEAP1 and -2 and Vdr cluster with Ly6a (Sca-1) and as stated above these genes
were also found to be differentially expressed in the UCs following the supervised
analysis. Thus two different methods of analysis, (one supervised, and one unsupervised)
have identified STEAP1 and -2 and Vdr as differentially expressed on MSC clones.
Several transcripts of cell cycle protein Cdk4 also clustered in this region. Interestingly
the hypoxia inducible factor-1 alpha (HIF-1α) was also present in this section.
157
G
en e Ex pr es si on L
ev el (a rb itr
ar y un its
)
STEAP2
0
200
400
600
800
1000
1200
1400
1600
1800
2000
UC DC MEF BMAC BMMNC
STEAP1
0
200
400
600
800
1000
1200
1400
1600
1800
2000
UC DC MEF BMAC BMMNC
Versican
0
200
400
600
800
1000
1200
1400
UC DC MEF BMAC BMMNC
H2-M9
0
20
40
60
80
100
120
140
UC DC MEF BMAC BMMNC
Vitamin D Receptor
0
100
200
300
400
500
600
700
UC DC MEF BMAC BMMNC
Ly6f
0
50
100
150
200
250
UC DC MEF BMAC BMMNC
Figure 5.8 Molecular signature of MSC clones. Six genes met the stringent conditions for
differentially high expression on all 8 UC compared to all 20 control samples as described in figure
5.7. The average (± SEM) gene expression intensities (arbitary units) from the microarray analysis
for these 6 genes are shown (n=8 samples for UC and for DC; n=4 for MEF, BMAC, and BMMNC).
Undifferentiated clones (UC); differentiated clones (adipocytes) (DC); primary bone marrow adherent
cells (BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic
fibroblasts (MEF).
158
Thumbnail view
U
C
37 (B
/b)
U
C
36 (B
/b)
U
C
15 (B
/b)
U
C
5 (B
/b)
U
C
3 (B
/c)
U
C
35 (B
/c)
U
C
4 (B
/c)
U
C
11(B
/c)
D
C
36 (B
/b)
D
C
37 (B
/b)
D
C
15 (B
/b)
D
C
5 (B
/b)
M
E
F1 (B/b)
M
E
F1 (B/b)
M
E
F1 (B/c)
M
E
F2 (B/c)
D
C
3 (B
/c)
D
C
35 (B
/c)
D
C
4 (B
/c)
D
C
11 (B
/c)
B
M
A
C
1 (B
/b)
B
M
A
C
2 (B
/b)
B
M
A
C
1 (B
/c)
B
M
A
C
2 (B/c)
B
M
M
N
C
1 (B
/b)
B
M
M
N
C
2 (B
/b)
B
M
M
N
C
1 (B
/c)
B
M
M
N
C
2 (B
/c)
STEAP2
STEAP2
STEAP1
/Sca-1
STEAP1
Figure 5.9 Dendrogram showing genes clustering with Sca-1. An unbiased, genome wide, two-dimensional
hierarchical cluster analysis was performed on the complete data set of 28 arrays as described in figure 5.1. Shown
are genes clustering closest to Sca-1 which include STEAP1 and -2. The color scale ranges from saturated red, log
ratios + 1.8 decreasing to saturated green for log ratios -1.8. Each gene is represented by a single row of colored
boxes; each individual control is represented by a single column. Undifferentiated clones (UC); differentiated
clones (adipocytes) (DC); primary bone marrow adherent cells (BMAC); freshly isolated bone marrow
mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF); BALB/b (B/b); BALB/c (B/c).
159
With regard to the cluster analysis, genes which displayed an expression profile similar to
Sca-1 were selected by going up and down one proximal note of Sca-1. This gave a list of
310 genes. Transcripts were categorised into 3 groups based on annotations in the
DAVID Gene Ontology software as described above (tables 5.10-12). Table 5.10 shows
that Sca-1 is the surface protein with the highest signal intensity 9177 (arbitrary units).
This is shown graphically in figure 5.10. Membrane proteins TM4SF1, erythrocyte
protein band 4.1-like 2, Maged2, and oncostatin M receptor were also highly expressed in
the clones compared to controls (table 5.10). Integrins such as, integrin alpha 5
(fibronectin receptor alpha), ephrin B1, microfibrillar-associated protein 3-like, intergrin
alpha 3, were identified in the Ly6a/Sca-1 cluster. Neural cell adhesion molecule 1
(NCAM-1) was also present.
Epidermal growth factor-containing fibulin-like extracellular matrix protein-1 and 2 were
also highly expressed in the UC, although also quite high on MEFs. They have not been
previously found in MSCs. The undifferentiated clones significantly express the secreted
proteins IGFBP6, HGF, FGF7, and Ccl7 in comparison to the controls (table 5.11, figure
5.10). The last three were also identified in the supervised analysis described above.
Table 5.12 shows cell cycle gene Cdk4 was expressed at a high intensity (3837). Growth
factor FGF7 was also highly expressed on the undifferentiated MSC clones. Mdm2 was
also quite highly expressed, but it was also expressed at a similar level in the MEFs and
BMACs (table 5.12). This highlights the danger of sole reliance on unsupervised
analysis, and also the advantage of the supervised method which filters out such high, but
not MSC-specific gene expression. Thus, STEAP1 and -2 were the only two cell
membrane proteins identified by both methods of analysis. From reports in the literature
they are not widely expressed in other cell types and they are present at very low levels in
all our control cell populations. This makes them very attractive candidates for further
analysis.
160
Table 5.10 Cell membrane (associated) genes clustering with Sca-1.
Genes encoding for cell surface markers were selected which showed an expression profile similar to Sca-1
in an unbiased 2-dimensional hierarchical cluster analysis shown in figure 5.9. The sub-cellular localisation
of the genes was classified by DAVID gene ontology software. Genes are ranked by their expression
intensity in UC (arbitrary units). Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary
bone marrow adherent cells (BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and
mouse embryonic fibroblasts (MEF).
Sequence Name Gene Accession Number UC DC MEF BMAC BMMNC
Lymphocyte antigen 6 complex,
locus A/ Sca-1 Ly6a/Sca-1 BC002070 9177 1072 4803 2815 202
Transmembrane 4 superfamily
member 1 TM4SF1 BQ177170 3049 287 1794 397 17
Erythrocyte protein band 4.1-like 2 Epb4.1I2 BE951907 2959 878 1717 1948 317
Melanoma antigen, family D, 2 Maged2 AF319976 2587 435 1180 227 6
Oncostatin M receptor Osmr AB015978 1927 408 1334 565 5
Six transmembrane epithelial
antigen of the prostrate 2 STEAP2 AK015015 1714 298 429 256 1
Six transmembrane epithelial
antigen of the prostrate 1 STEAP1 AF297098 1668 223 232 194 13
Angiomotin Amot AF297098 1210 299 714 147 12
Transmembrane protein 19 Tmem19 BE951907 1072 425 415 712 199
Histocompatibility 60 H60 BG067039 1069 189 183 421 44
Receptor-like tyrosine kinase Ryk BB822862 1030 334 821 320 91
Hyaluronan synthase 2 Has2 455 67 155 262 21
Erythrocyte protein band 4.1-like 3 Epb4.1I3 M98547 454 133 409 171 0
Integrin alpha 5 (fibronectin
receptor alpha) Itga5 BE951907 339 91 173 204 16
Neural cell adhesion molecule 1 Ncam1 BB050303 330 63 175 116 41
Ephrin B1 Efnb1 BB817972 306 119 241 148 28
Solute carrier organic anion
transporter family, member 2a1 Slco2a1 NM_008216 298 77 150 309 14
Membrane associated guanylate
kinase-related (MAGI-3) Magi3 AF177146 295 99 243 136 69
Microfibrillar-associated protein 3like Mfap3I NM_008216 258 53 157 157 13
Fyn-related kinase Frk BB831986 254 80 138 69 2
Interleukin 1 receptor 1 Ilr1 AB015978 241 68 55 0 0
Intergrin alpha 3 Itga3 BB493533 240 77 202 101 30
Calcium channel, voltagedependent, beta 3 subunit Cacnb3 NM_010875 229 64 179 76 1
Angiopoietin 4 Angpt4 BC006797 225 25 0 52 0
Hepatocellular carcinomaassociated antigen 127 Hcca127 NM_033314 180 100 160 119 32
G protein-coupled receptor 85 Gpr85 BF180812 178 66 68 122 5
Endothelial differentiation,
lysophosphatidic acid G-proteincoupled receptor, 2
Edg2 BB529332 160 69 78 65 21
Transmembrane 100 Tmem100 AK017269 135 18 6 4 0
161
Table 5.11 Genes encoding for secreted proteins clustering with Sca-1.
Genes encoding for secreted proteins were selected which showed an expression profile similar to Sca-1 in an
unbiased 2-dimensional hierarchical cluster analysis shown in figure 5.9. The sub-cellular localisation of the genes
was classified by DAVID gene ontology software. Genes are ranked by their expression intensity in UC (arbitrary
units). Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow adherent cells
(BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic fibroblasts (MEF).
Sequence name Gene Accession Number UC DC MEF BMAC BMMNC
Epidermal growth factor-containing
fibulin-like extracellular matrix
protein 1
Efemp1 BC023060 6128 1180 4635 3486 3
Insulin-like growth factor binding
protein 6 Igfbp6 NM_008344 3098 894 633 76 20
Epidermal growth factor-containing
fibulin-like extracellular matrix
protein 2
Efemp2 NM_021474 2659 887 1573 542 0
Fibroblast growth factor 7 Fgf7 BB791906 2092 571 674 613 2
Chemokine (C-C motif) ligand 7 Ccl7 AF128193 1352 113 149 342 13
Cytokine receptor-like factor 1 Crlf1 NM_018827 878 371 726 363 0
Colony stimulating factor 1 Csf1 M21149 541 177 368 449 65
Hepatocyte growth factor Hgf AF042856 356 76 75 77 19
Procollagen, type 1, alpha 2 Col1a2 BB150460 183 36 70 52 3
162
Table 5.12 Cell cycle associated genes clustering with Sca-1.
Genes associated with the cell cycle were selected which showed an expression profile similar to Sca-1 in an
unbiased 2-dimensional hierarchical cluster analysis shown in figure 5.9. The sub-cellular localisation of the
genes was classified by DAVID gene ontology website. Genes are ranked by their expression intensity in UC
(arbitrary units). Undifferentiated clone (UC); adipocyte differentiated clones (DC); primary bone marrow
adherent cells (BMAC); freshly isolated bone marrow mononuclear cells (BMMNC); and mouse embryonic
fibroblasts (MEF).
Sequence name Gene Accession Number UC DC MEF BMAC BMMNC
Cyclin dependent kinase 4 Cdk4 NM_009870 3837 1330 1944 963 708
Fibroblast growth factor 7 Fgf7 NM_008008 2092 571 674 613 2
Transformed mouse 3T3 cell
double minute 2 Mdm2 AK004719 1578 903 1400 1573 513
Cullin 4B Cul4b AW536452 693 383 578 474 362
Histone deacetylase 7A Hdac7a BB277517 605 223 362 161 217
Colony stimulating factor 1 Csf1 M21149 541 177 368 449 65
Cyclin-dependent kinase
inhibitor 2a Cdkn2a NM_009877 530 183 206 108 14
MAD homolog 3 Smad3 B1646741 424 189 313 277 170
Hepatocyte growth factor Hgf AF042856 356 76 75 77 19
Eprin B1 Efnb1 BC006797 306 119 241 148 28
Fyn-related kinase Frk BB787292 254 80 138 69 2
ELK3, member of ETS
oncogene family Elk3 BC005686 200 32 99 88 22
163
G
en e E
xp re ss io n Le ve l (
ar bi tra
ry u ni ts )
Sca-1
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
UC DC MEF BMAC BMMNC
***
***
***
***
TM4SF1
0
500
1000
1500
2000
2500
3000
3500
4000
UC DC MEF BMAC BMMNC
***
*
***
***
MAGED2
0
500
1000
1500
2000
2500
3000
UC DC MEF BMAC BMMNC
***
***
***
***
CCL7
0
200
400
600
800
1000
1200
1400
1600
1800
UC DC MEF BMAC BMMNC
*** ***
***
***
CDK4
0
500
1000
1500
2000
2500
3000
3500
4000
4500
UC DC MEF BMAC BMMNC
***
***
*** ***
0
500
1000
1500
2000
2500
3000
3500
UC DC MEF BMAC BMMNC
IGFBP6
***
***
*** ***
FGF7
0
250
500
750
1000
1250
1500
1750
2000
2250
2500
2750
UC DC MEF BMAC BMMNC
*** *** ***
***
HGF
0
50
100
150
200
250
300
350
400
450
UC DC MEF BMAC BMMNC
*** *** ***
***
FOXA1
0
10
20
30
40
50
60
70
80
90
100
UC DC MEF BMAC BMMNC
***
**
*** ***
Figure 5.10 Genes highly expressed on MSC clones. Genes which clustered with Sca-1
were selected by going up and down one proximal note of the Ly6a/Sca-1 cluster shown in
figure 5.9. This gave a list of 310 genes. Average gene expression intensities in arbitary units
(±S.E.M) for 9 of these genes which show differentially high expression on UC are shown.
*P<0.05, **P<0.01, ***P<0.001 versus UC expression level. Data were compared using oneway analysis of variance (ANOVA) with Donnet’s multiple comparison test using Prism 5
software (Graphpad). Abbreviations: UC - undifferentiated MSC clones; DC – differentiated
clones (adipocytes); MEF - mouse embryonic fibroblasts; BMAC - bone marrow adherent
cells (primary culture); BMMNC – bone marrow mononuclear cells (freshly isolated), (n=8
samples for UC and for DC; n= 4 for MEF, BMAC, and BMMNC).
164
5.9 Validation of microarray results
Microarray results were validated by quantitative RT-PCR for STEAP1 and -2,
Ly6a/Sca-1, versican, TM4SF1, and adipocyte marker PPARγ. The results confirmed the
microarray data (table 5.13).
5.10 Discussion
Many groups have tried to identify MSC specific markers. For example, hybridoma
technology has been utilised to obtain monoclonal antibodies raised against a
heterogeneous population of adherent bone marrow cells (Haynesworth et al., 1992).
Alternatively, investigators have sorted bone marrow cells based on a specific (but
arbitrarily chosen) antigen (Jones et al., 2002). Both approaches suffer from major
shortcomings, as the antibodies derived in the former may not be MSC specific and the
latter method is biased and provides no new information of an MSC marker.
The transcriptome of embryonic stem cells has been well characterised and this has led to
greater understanding of the self-renewal and differentiation mechanism of embryonic
stem cells (Zhan, 2008). The transcription factors OCT4, Nanog, and SOX4 are known to
be essential to the pluripotent and self-renewing phenotype of embryonic stem cells
(Chambers et al., 2003; Nichols et al., 1998). Very little however, is known about the
mechanisms controlling the ‘stemness’ of MSCs. Application of microarray technology
to further the understanding of MSC biology has been hampered by the lack of
physiologically relevant cell models that can be used for characterisation. This is because
most groups work with a mixed population of cells they term ‘MSCs’ (Jia et al., 2002;
Song et al., 2006; Wagner et al., 2006; Wieczorek et al., 2003). Furthermore, this
heterogeneous cell population is ill-defined. Their ability to both self-renew and
differentiate into the mesenchymal lineages of adipocytes, chondrocytes, and osteoblasts
has not been well examined. Therefore any subsequent results gained about this
heterogeneous population of cells is always going to be limited, and it is difficult to
directly compare one such study with another.
165
Table 5.13 Validation of Microarray Data by Real Time PCR for (A) BALB/b, and (B) BALB/c data
sets. Intensities expressed relative to undifferentiated clones (UC) levels. Adipocyte differentiated clones
(DC); primary bone marrow adherent cells (BMAC); freshly isolated bone marrow mononuclear cells
(BMMNC); and mouse embryonic fibroblasts (MEF); Peroxisome proliferator-activated receptor- γ
(PPARγ).
A
Ly6a/Sca-1 STEAP1 STEAP2 TM4SF1 Versican PPARγ
PCR Array PCR Array PCR Array PCR Array PCR Array PCR Array
UC 1 1 1 1 1 1 1 1 1 1 1 1
DC 0.01 0.01 0.08 0.13 0.06 0.09 0 0 0.06 0.07 6.25 4.97
MEF 0.18 0.64 0.05 0.15 0.07 0.25 0.1 0.56 0.05 0.1 0.06 0.20
BMAC 0.01 0.06 0.04 0.13 0.06 0.12 0.08 0.13 0.03 0.05 0.25 0.48
BMMNC 0 0.03 0 0.01 0 0 0 0.01 0.01 0.01 0.06 0.02
B
Ly6a/Sca-1 STEAP1 STEAP2 TM4SF1 Versican PPARγ
PCR Array PCR Array PCR Array PCR Array PCR Array PCR Array
UC 1 1 1 1 1 1 1 1 1 1 1 1
DC 0.08 0.22 0.05 0.14 0.3 0.3 0.19 0.26 0.15 0.21 3.85 2.75
MEF 0.17 0.41 0.07 0.12 0.15 0.25 0.45 0.63 0.16 0.22 0.23 0.21
BMAC 0.16 0.55 0.03 0.1 0.06 0.2 0.06 0.14 0.06 0.22 1.50 1.64
BMMNC 0.02 0.02 0 0 0 0 0 0 0.01 0.01 0.38 0.06
166
One group has examined the gene expression profile of a clonal population of human
MSCs and catalogued 2,353 genes (Tremain et al., 2001). However, the significance of
the results were limited as only one clone was studied, and no other biological replicates
were used. The lack of appropriate control populations is another drawback as it is not
possible to distinguish potentially MSC-specific genes from the typically long gene lists
that are obtained. Thus these studies have failed to delineate the ‘genetic signature’
common to MSCs.
The Wagner group used a slightly better approach as they compared the gene expression
profile of several cell populations. These included human MSCs isolated from bone
marrow, adipose tissue, umbilical cord blood under two different growth conditions;
comparing to terminally differentiated human newborn foreskin fibroblasts (HS68)
(Wagner et al., 2005). A limitation of the study was that all the MSC populations were
heterogeneous, thus making it difficult to distinguish the MSC specific genes from that of
the other cell types. Also, only the adipogenic and osteogenic differentiation capacity of
the cells was examined. In contrast, in the present study, all the clonal cells studied had
tri-lineage differentiation potential as shown in chapter 3. In addition, adipogenesis of
umbilical cord blood cells was not convincing in the Wagner study. Another downside of
that study is that the different cell populations were cultured in different culture
conditions with the addition of various growth factors and chemicals such as
dexamethasone which may have selected for distinct cell types. However, the
investigators did compare the gene expression profile of the MSC populations to
fibroblasts that did not have the differentiation ability of the MSC populations. The group
identified 25 genes that were differentially highly expressed on all MSC populations
compared to the fibroblasts. However, as already stated, the problem with this is that all
the MSC populations were mixed, and each tissue specific MSC was from a different
donor of varying ages. The fact that the umbilical cord blood MSCs were not as
multipotent as the bone marrow and the adipose MSCs may have affected the results and
prevented entry of MSC specific genes into the final gene list.
The Song group examined the gene expression profile of a heterogeneous population of
human MSCs, comparing it to that of differentiated MSCs (Song et al., 2006). The major
167
aim of their study was to identify molecular MSC markers to facilitate the purification of
homogeneous MSC populations. The gene expression profile of undifferentiated MSCs
was compared to MSCs differentiated into adipo-, chondro-, and osteogenic lineages.
Although this was a sound approach there were some limitations; one major one being
different donors were used for the different differentiation assays, thus introducing
unnecessary biological variability into the dataset, and lessening the power of the results.
Also the donors were quite old ranging 59-70 years old. The starting population was
again heterogeneous and although they demonstrated the tri-lineage differentiation ability
of the MSCs, osteogenic differentiation was only demonstrated through alkaline
phosphatase staining which is not, on its own, a robust assay of osteogenesis. Ideally von
kossa or alizarin red should be used to directly detect the mineralised matrix secreted by
the cells. The investigators compared the gene expression profile of undifferentiated
MSCs, differentiated MSCs, and the equivalent de-differentiated MSCs. The analysis
resulted in a large gene list of 460 genes which they classed as ‘stemness’ genes, 91 of
which encoded cell surface antigens. It is difficult to examine the function of all these
genes. Therefore, the investigators selected 5 based on their previous known functions.
Although this is a realistic approach, it is not ideal as potential unknown MSC specific
genes may be missed.
Due to these various issues, we therefore very carefully planned our microarray
experimental design in order to avoid these different shortcomings, and obtain the most
meaningful data possible. As described in chapter one, by combining a low oxygen
environment with low density cultures we isolated clonal cells that fulfil the criteria for
mesenchymal stem cells. The clones have the capacity to differentiate into the three
lineages - adipogenic, osteogenic, and chondrogenic that characterise a MSC. They also
have the ability to self-renew. After further characterisation (detailed in chapter 4) we
found that the cells were karyotypically abnormal but, we also found they were not
malignant, and showed very well-regulated differentiation in vitro, and thus could be
utilised as a good model to study MSC biology. Our final aim (detailed in the following
chapter) was then to use the results from the microarray study to test for the presence of
168
candidate MSC-specific markers thus identified on (karyotypically) normal human
MSCs.
The Song group compared the gene expression profile of undifferentiated MSCs to
differentiated MSCs, and the Wagner group compared the gene expression profile of
MSCs to fibroblasts (Song et al., 2006; Wagner et al., 2005). We performed an even more
extensive analysis. The gene expression profile of the MSC clones in this study was
compared to the equivalent adipocyte differentiated clones (which lacked stem cell
properties), embryonic fibroblasts (that lacked the differentiation potential of the clones),
a mixed population of BMACs, and freshly isolated BMMNCs.
The gene expression profile of the clonal cells was compared to these relevant controls to
characterise the molecular signature of the MSCs. It is important to note here that all cells
were cultured at 5% oxygen for at least 4 days to ensure no change in transcripts occurred
simply due to alterations in the oxygen environment. A previous study has reported a
change in gene transcripts in MSCs in response to variation of oxygen levels at different
time courses (Ohnishi et al., 2007). Also, all cells were cultured in the same cell culture
conditions without the addition of growth factors that may bias the results. This is an
important point as studies have reported that differences in cell culture conditions can
change the gene expression profile of two parallel populations of MSCs (Wagner et al.,
2006). A large percentage of genes were found to be differentially regulated when bone
marrow MSCs were cultured in two different media compositions M1 and M2. M1
consisted of DMEM and MCDB201, 2% FCS, supplemented with L-glutamine, 1%
insulin transferrin selenium, 1% linoleic acid bovine serum albumin, dexamethasone, Lascorbic acid 2 phosphate, PDGF-BB, and EGF previously used by Reyes and colleagues
(Reyes et al., 2001). M2 was a commercially purchased culture medium from Poietics for
human mesenchymal stem cell. MSCs cultured in M1 expressed a greater proportion of
genes in the functional categories of metabolism and mitochondria. Whereas genes
involved in muscle development, neurogenesis, cell differentiation and morphogenesis,
and skeletal developmental were significantly overexpressed in MSCs cultured in M2
media. The results emphasise the huge impact different culture conditions can have on
the intrinsic biology of the cells and the need for standardisation if comparisons are going
to be made between MSC studies.
169
In this study, a total of eight undifferentiated clones (from two different mouse strains)
were used thus providing the wide range of biological replicates needed for generation of
meaningful microarray data. In addition, four different controls were also used, each with
their own biological replicates. This thus represented a major undertaking, and in total,
the gene expression profile of 28 different samples was examined. Principal component
analysis of these 28 genome-wide arrays showed that the undifferentiated clones
clustered very tightly together, thus highlighting that they were very similar, irrespective
of mouse strain differences. This substantiates the fact that they have a similar
morphology, differentiation potential, and surface antigen profile. This was also true for
the other cell types examined. Similar findings were also uncovered using an
unsupervised 2D-hierarchical cluster analysis of the data. Within the control groups the
cell types exhibited very similar expression profiles e.g., the adipocyte differentiated
clones preferentially expressed PPARγ which is an adipocyte marker. The MEFs highly
expressed the fibroblast marker CD90/Thy-1, which corresponds to the flow cytometry
results (table 4.2-4.3) and the BMACs expressed relatively high levels of colony
stimulating factor 1- receptor (Csf1r).
We performed two types of analysis, supervised and unsupervised, allowing comparison
of the two approaches and strengthening the resultant data. The two analytic methods
have different strengths and pitfalls in transcriptional profiling analysis. In the stringent
supervised analysis, the minimum cut-off we used when comparing gene expression
levels of MSCs to the various controls was 3-fold (and 10-fold for the BMMNCs). The
higher value for the latter was applied since MSCs are present at such a low frequency in
the bone marrow. Other groups have looked for 2 fold differences (Tuan et al., 2003), but
we believe when screening for potentially cell-specific genes, more stringent conditions
are necessary. Thus, for cultures from each mouse strain, these conditions had to be met
for all of the 28 individual comparisons made (see figure 5.7A). This is a very strict
regime, but since we are looking for cell-specific genes, we reasoned that these
conditions should be met on every occasion, rather than merely using statistical averages.
We thus found 63, and 78 genes differentially highly expressed on all the undifferentiated
170
clones from the BALB/b and BALB/c mice respectively. A final comparison of the genes
present in both mouse strains gave only 6 genes. Most strikingly, all six are membrane, or
membrane-associated genes, with only one not being associated with the cell membrane
itself, i.e., vitamin D receptor (Vdr) being nuclear. The five cell membrane (associated)
genes are STEAP1, STEAP2, versican, ly6f, histocompatability 2, m region locus 9 (H2M9).
Indeed, a very interesting point is that differentially expressed cell membrane (associated)
genes were highly conserved between the two mouse strains compared to other genes.
For example, 8 genes for cell membrane (associated) markers were differentially
expressed from BALB/c MSC clones, and 6 for BALB/c clones, therefore almost all
(five) were common to both strains (see figure 5.7B). Indeed, it is noteworthy that only
membrane or membrane-associated genes were identified as differentially expressed in
the final analysis.
Interestingly, Ly6f is a member of Ly6a/Sca-1 family, which is highly expressed on the
undifferentiated clones. Unfortunately little is known about Ly6f protein and no reagents
are available. Despite having a low signal intensity, it would be interesting to determine
whether it has a functional role in MSC biology, as Sca-1 has been reported to be
required for self-renewal of MSC progenitors in a study of Sca-1 knockout mice
(Bonyadi et al., 2003). H2-M9 is a major histocompatability class Ib gene, first cloned
and sequenced from BALB/c mice (Arepalli et al., 1998) and in our study (like Ly6f) is
expressed at a relatively low level. Major histocompatability class 1 genes are expressed
on all nucleated cells. These proteins play a major role in the immune system, as they
present fragmented antigen parts on the host cell surface. These may be ‘self’ or ‘nonself’. Although there is little known about this protein, based on its function it is likely to
be expressed on a variety of cell types. This is the first report identifying the two genes in
MSCs. Vitamin D receptor (Vdr) was expressed at a relatively higher level in UCs (than
Ly6f and H2-M9). Vdr is also known as the calcitriol receptor, and is a member of the
steroid hormone receptor family of nuclear receptor proteins. Vitamin D is required for
calcium homeostasis so it is not surprising that Vdr is a transcription factor that plays an
important role in regulating osteogenesis in MSCs (Pochampally et al., 2007), and indeed
some groups add vitamin D to stimulate osteogenesis of MSCs (Malladi et al., 2006b).
171
Vdr has also been reported to have a role in stimulating chondrogenesis of chick limb bud
mesenchymal cells (Tsonis et al., 1996). Vdr is expressed on many cell types including
MSCs, osteoblasts, mammary cells, and cells of the immune system such as T cells and
dendritic cells.
Versican also known as chondroitin sulphate proteoglycan core protein-2 (CSPG-2) was
found to be highly expressed on the UCs. It is composed of modular protein domains
with several glycosaminoglycan side chains. Versican is a component of the extracellular
matrix (ECM) of cells and thus plays a role in many cellular processes such as cell
adhesion, proliferation, migration, angiogenesis as well as tissue morphogenesis and
maintenance (Rahmani et al., 2006). Due to its role in a wide range of cellular processes
it is expressed on many cell types and thus may lack the specificity required to be a
MSC-specific marker. Studies have found versican expression in MSCs increases upon
differentiation into chondrocytes and osteoblasts (Diaz-Romero et al., 2008; Foster et al.,
2005). This could be due to an increase in the ECM of both cell types. Therefore, results
suggest it does not have a functional role in maintaining the stemness of MSC. As both
Vdr and versican are expressed in many cell types in the bone marrow, they are unlikely
to be useful markers to distinguish MSCs from other bone marrow-derived cells. H2-M9
and Ly6f are expressed at relatively low signal intensity in the UCs and very little is
known about these two membrane proteins, not to mention a lack of reagents available.
Six transmembrane epithelial antigen of the prostate 1 (STEAP1) and STEAP2 were both
highly expressed on the undifferentiated clones. STEAP1 and STEAP2 were first
discovered relatively recently, in 1999 and 2002 respectively (Hubert et al., 1999; Porkka
et al., 2002). They were found in high levels in human prostate cancer. Recent studies
have also reported STEAP1 expression in murine prostate cancer (Yang et al., 2001).
Based on their secondary structure (that contains six transmembrane domains) they are
predicted to function as potential ion channels (Dolly and Parcej, 1996), water channels
(Kruse et al., 2006) or transporter proteins (Ohgami et al., 2006). They belong to a family
of STEAP proteins that consists of STEAPs 1-4. All STEAPs contain intra-membrane
haem-binding sites. STEAPs 2-4 contain a N-terminal domain that is similar to the
archael and bacterial F420H2:NADP+ oxidoreductase (FNO) binding proteins (Ohgami et
172
al., 2006). They are reported to function as ferriductases and cupric reductases that
stimulate cellular uptake of both iron and copper in vitro and possibly in vivo. STEAP1
lacks this FNO-like reductase domain, however STEAP proteins form heteromultimers
with STEAP1. Thus, this may be one mechanism by which STEAP1 participates in metal
homeostasis. In agreement with these results, STEAP2 was found to co-localise
completely with transferrin (Tf) and Tf receptor 1 (TfR1) within a Tf endosome where
the reduction of iron takes place. Tf is a major iron chelator and complexes with TfR1 to
be endocytosed. In contrast to STEAP2, STEAP1 was only partially co-localised with Tf
and TfR1, and was also present in other vesicles that did not contain Tf (Ohgami et al.,
2006).
The Ohgami group performed an extensive analysis of STEAP1 and 2 mRNA expression
in a variety of human tissue including brain, colon, duodenum, heart, kidney, liver, lung,
pancreas, placenta, thymus, ileum, stomach, skeletal muscle, bone marrow, and prostate.
They found both STEAP1 and 2 were predominantly expressed in prostate, with low
levels in the bone marrow (10 fold less) (Ohgami et al., 2006). In contrast, STEAP4 was
predominantly expressed in the bone marrow, with very low levels in the prostate. Initial
functional studies by the Challita-Eid group reported STEAP1 function as a transporter of
small molecules (<1 kDa), suggesting it may be involved in intercellular communication
(Challita-Eid et al., 2007). This hypothesis is further strengthened due to its localisation
at cell-cell boundaries (Hubert et al., 1999).
The current study is the first report of STEAP1 and 2 expression in MSCs. We looked for
murine STEAP1 and 2 via western blot analysis and FACS, but unfortunately the
antibodies were not suitable for detection of murine STEAP1 and STEAP2 proteins.
Both STEAP members showed a strikingly similar expression pattern to stem cell
antigen-1 (Sca-1) (figure 5.9). Vdr is also present in the Sca-1 cluster. Hence all four
genes showed a very similar expression pattern, as detected by an unsupervised, genomewide two-dimensional clustering analysis (figure 5.9). Notably, two different analysis
methods have identified both STEAP1 and -2 as differentially highly expressed in the
UCs, and most importantly these two genes are the only cell membrane genes identified
173
by both analysis methods. This is particularly remarkable when one is reminded that the
analysis is based on the entire genome.
A study by Xin and colleagues reported that Sca-1 enriches for prostate-regenerating cells
which possessed stem cells properties (Xin et al., 2005). However, genetic perturbation of
PTEN/AKT signalling in these cells led to tumourigenesis with concomitant increase in
Sca-1. Results suggest there is an interesting relationship between Sca-1 and the STEAP
proteins. In addition, a prostate stem cell antigen (PSCA) which is also found on prostate
cancer is located with Sca-1 in a multigene cluster of approximately 18 well conserved
genes on chromosome 15 (Gumley et al., 1995). Sca-1 message was very highly
expressed as detected in the microarray analysis. Furthermore, the microarray results
correspond to the flow cytometry results (in chapter 4), which show Sca-1 is highly
expressed in the clones in comparison to MEFs and BMMNCs. Also Sca-1 is highly
expressed in UCs compared to BMACs.
The cell cycle gene Cdk4 also clustered with Sca-1/STEAP1/STEAP2. Cdk4 is important
for cell cycle G1 phase progression thus indicating the high proliferative capacity of the
undifferentiated clones compared to the others cells studied. Interestingly, HIF-1α was
also present in this cluster. Since all cell types were cultured in 5% oxygen (except
BMMNCs which were not cultured), this indicates that the endogenous steady state HIF1α mRNA levels are greater in the MSC clones.
Another potential gene of interest present in the Sca-1 cluster is Maged2, which has been
previously reported to be expressed by mesodermal cells (Bertrand et al., 2004). Cell
surface marker TM4SF1 is also expressed at a high intensity, and is a member of the
tetraspanin superfamily of membrane proteins (Virtaneva et al., 1994). Unfortunately
there is no reagent available to further study this. In the literature, both Maged2 and
TM4SF1 have been implicated in tumour development in the lung (Kidd et al., 2006) and
gut (Kao et al., 2003) respectively. Thus, one may speculate that the high levels of
Maged2 and TM4SF1 may characterise the transformed nature of the MSCs and
potentially distinguish them from their normal counterpart. However, TM4SF1 has also
174
been reported to be expressed at significant levels in human MSCs (Song et al., 2006;
Wagner et al., 2005). Indeed, Song and colleagues report the expression of a number of
genes involved in carcinogenesis and metastasis in their normal MSC populations, and
hence these genes may not necessarily be representative of abnormal or cancer stem cells.
Cytokine and growth factor signalling is an important determinant of the functional state
of cells. Undifferentiated clones significantly express the cytokines, HGF, FGF7, Ccl7,
and insulin growth factor binding protein-6 (IGFBP6) in comparison to the controls
studied. Both FGF7 and HGF have been previously reported in MSCs (Djouad et al.,
2007). FGF7 is also known as keratinocyte growth factor, and high levels of FGF7
expression have been previously reported in human MSCs during their growth phase
(Imabayashi et al., 2003). FGF7 is involved in a broad range of activities including cell
growth and morphogenesis, tissue repair, and embryonic development. It is typically
thought to be an epithelial cell growth factor (Finch and Rubin, 2004). MSCs have been
reported to express HGF and its receptor cMET (Rosova et al., 2008). These investigators
report HGF is activated at sites of ischemic injury. This is one mechanism proposed by
which MSCs migrate to damaged tissue, via a HGF gradient. This is the first mention of
Ccl7 (chemokine (C-C motif) ligand 7) in MSCs. Ccl7 expression was also found in the
BALB/b clonal cells via the supervised analysis. Ccl7 was previously called monocyte
chemotactic protein-3 (MCP-3), and is one of the most pluripotent chemokine, attracting
monocytes and regulating macrophage function (Menten et al., 2001). Ccl7 was recently
identified as a differentiation factor for neuron development (Edman et al., 2008). Clonal
cells also expressed significant levels of IGFBP6. A previous study has reported that
secretion of IGFBP6 by stromal cells may be a key modulator of the hematopoietic
response to IGFs (Grellier et al., 1995). The transcription factor Foxa1 was also
significantly expressed by the clonal cells. Foxa1 is a hepatic transcription factor found in
epi-meso progenitor cells (Inada et al., 2008). It is reported to be important in normal
prostate development (Mirosevich et al., 2005). Also, Foxa1 and Foxa2 are also involved
in regulating midbrain dopaminergic neuron development (Ferri et al., 2007).
175
Clonal cells also expressed neural cell adhesion molecule -1 (NCAM-1) at over 2 fold
greater levels than control samples. It is interesting to note that clonal cells expressed low
levels of transcripts that are characteristic of epithelial, neural, and endothelial cell types
as discussed above. Thus, indicating the broad differentiation capabilities of MSCs.
Although we have only demonstrated the tri-lineage differentiation ability of the clones
into the adipogenic, chondrogenic, and osteogenic lineages, it is possible that the cells
have a greater differentiation potential. Previous studies that have analysed the gene
expression profile of human and murine MSCs have also reported the presence of other
lineage specific transcripts (Phinney et al., 2006; Tremain et al., 2001; Wagner et al.,
2005) (Silva et al., 2003).
STEAP is an epithelial antigen and may seem an unlikely MSC marker, however it must
be noted that MSCs, or more specifically multilineage adult progenitor cells (MAPC), at
the single cell level, have been shown to give rise to epithelial lineages of liver, lung and
gut (Jiang et al., 2002). We have not investigated the ability of the MSC clones to
differentiate into other lineages, but the presence of multilineage transcripts would
suggest this is possible given the right environmental cues. As discussed above the
murine clones expressed a variety of multilineage transcripts, including hepatic
differentiation factors Foxa1, HGF. Although the STEAPs are known as prostate cancer
markers, there are other instances where markers that are used to isolate stem cells for
one tissue are also used as cancer stem cell markers for another. For example, CD133 is a
neural stem cell marker used to isolate neural stem cells (Peh et al., 2008). The watt
group also use CD133 to isolate a MSC-enriched population. (Martin-Rendon et al.,
2007). CD133 is also used to isolate a small subset of human brain cancer cells which
have the potential to form tumours in mice identical to the original tumours (Singh et al.,
2004). Also, interestingly a subpopulation of CD133+ human prostate epithelial cells
have a high growth rate in vitro and can reconstitute prostatic-like acini in vivo
(Richardson et al., 2004). CD34 is the major haematopoietic stem cell (HSC) marker,
commonly used in conjunction with other markers to isolate subpopulations of HSCs.
However, the first definitive report about cancer stem cells was about a subpopulation of
CD34+/CD38- leukemic cells that formed tumours in NOD/SCID mice (Bonnet and Dick,
176
1997). These results suggest there is a close relationship between stem cell and cancer
stem cell markers. This is perhaps because some cancers may be caused by stem cells
losing their ability to regulate proliferation. If a stem cell marker has an obvious
functional role in the stem cell such as proliferation or self-renewal, and is dysregulated,
this may be one way it becomes a marker of both stem and cancer stem cell in the same,
or different tissues. Thus, one may speculate that STEAP1 and 2 are markers of cancer
cells in the prostate, but also potential markers of normal mesenchymal stem cells in the
bone marrow.
The purpose of the next chapter will be to examine their expression in multipotent,
karyotypically normal human bone marrow stromal cells.
177
Chapter 6
Human Bone Marrow Stromal Cells
express STEAP1 and -2
178
6.1 Introduction
Gene expression analysis of murine clonal MSCs revealed that they differentially
expressed the cell surface markers STEAP1 and STEAP2, in comparison to relevant
controls. The purpose of the work in this chapter was to derive normal, human,
multipotent bone marrow stromal cells (BMSCs) and to examine whether they also
differentially expressed STEAP1 and 2.
6.2 Isolation of human bone marrow stromal cells
Initially we attempted to isolate clonal human MSCs using the method we had previously
developed (described in chapter 3) with murine MSCs. We cultured the cells in the same
culture conditions at 5% oxygen tension. Commercially obtained human bone marrow
mononuclear cells (hBMMNCs) were seeded at a low density and cells cultured for over
a month like the murine cells had previously been. However, there were typically very
few adherent cells, and these were representative of senescent cells by their flat, spreadout morphology rather than the highly proliferative murine cells which had a more
spindle shaped morphology. Even after several attempts, clonal cells were not apparent,
i.e., cells were not present in tight clusters like the murine cells but slightly dispersed.
These were isolated using cloning rings and re-seeded into a 3.5 cm tissue culture dish.
Most of these cells did not adhere to the tissue culture surface and there was a
morphology change in the ones that did from a spindle shape to the flat, spread-out shape.
Therefore we decided to culture adherent bone marrow stromal cells (BMSCs) which
would contain a population of MSCs and presumably become more enriched for MSCs
with subsequent passage as previously reported (Barry, 2003). Human BMMNCs were
therefore seeded at a high density and cultured at 5% oxygen. The subsequent cells were
spindle shaped although after passage 2 there was an increase in the flatter shaped cells.
Cells were cultured until 70% confluence to ensure there was not a morphology change
and then re-seeded for expansion. These cells were referred to as hBMSC1 as they were
from the first isolation. This method was repeated with a fresh bone marrow aspirate
which had been ficoll-separated that we received from a hospital. This bone marrow was
different from hBMSC1; it contained whitish-yellow parts floating in the media. This
179
could be fatty deposits as the sample came from a 60 year old patient and the quantity of
adipocytes does increase in the bone marrow with age, thus the bone marrow goes from
red to yellow. Human BMSC2 cells from this patient were initially more spindle shaped
but soon turned very flat in culture. The cells subsequently had a very slow growth rate
compared to hBMSC1, and could only be expanded up to passage three. The third culture
(hBMSC3) was derived from a bone marrow aspirate from an 11 year old osteosarcoma
patient. Microscopically all these cells were homogeneously spindle shaped and had a
faster growth rate. Cells were expanded to passage six and retained their spindle-shaped
morphology. The fourth BMSC sample was derived from commercially obtained
BMMNCs as described for hBMSC1. The donor was male and 25 years old. These cells
were all spindle shaped.
6.3 Growth rate of hBMSC3
To determine the growth rate of hBMSC3, the population doubling time was followed at
5% oxygen. Cells were seeded at a density of 1000-2500 cells/cm2 and passaged when
approximately 70% confluent. Figure 6.1 shows the population doubling time for
hBMSC3. Beginning at passage 2 to passage 5 the growth rate is constant. The cell
doubling time averaged 3.42 ± 0.78 days (mean ± SD), and no cell death was observed
during this period.
6.4 Human bone marrow stromal cells are multipotent
The tri-lineage differentiation potential of the hBMSCs was examined to confirm the
presence of MSCs in the population. Neonatal human dermal fibroblasts were used as an
additional control in differentiation experiments. Cells were cultured in the differentiation
media as described in sections 3.4.1-3.4.3 (figure 6.2, table 6.1). All hBMSCs had the
ability to differentiate into adipocytes when cultured in the appropriate media, and it took
4 weeks for all the cells to differentiate in the hBMSC3 and hBMSC4 cultures. Only ~
50% of the cells in hBMSC1 and hBMSC2 cultures differentiated. All hBMSCs
180
Passage
Po pu la tio
n do ub lin
gs 0
1
2
3
4
5
6
7
8
9
2 3 4 5
Cell doubling Time
3.42 ± 0.78 days
mean ± s.d.
Figure 6.1 Growth rate of human BMSCs. The growth rate of human bone
marrow stromal cells from donor 3 (hBMSC3) was followed in 5% oxygen tension.
Cells were seeded at a density of 1000-2500 cells/cm2 and passaged every 5-8 days
when they were 70% confluent. Cells were counted at each passage to calculate
population doubling time (standard deviation, s.d.).
181
Oil Red O
Collagen
Type II
Alizarin
Red
+ Differentiation Media - Differentiation Media
Figure 6.2 Human BMSCs have multilineage differentiation potential. Human bone
marrow stromal cells (hBMSCs) and fibroblasts were cultured in the presence or absence of
differentiation media as described in figures 3.4- 3.10 to assess their multilineage differentiation
potential. Pictures of cultures of hBMSCs from donor 3 (hBMSC3) are shown. NHDF; neonatal
human dermal fibroblasts.
hBMSC3
NHDF
+ Differentiation Media
182
Table 6.1 Human BMSCs are multipotent. All human bone marrow stromal cells had the
ability to differentiate into adipocytes, chondrocytes, and osteoblasts when cultured under
the appropriate conditions, except hBMSC2 where there was not enough cells to perform
the osteogenesis experiment. Importantly, under the same conditions neonatal human
dermal fibroblasts (NHDFs) failed to differentiate into three lineages
Clone Adipogenesis Chondrogenesis Osteogenesis
hBMSC1
hBMSC2 ?
hBMSC3
hBMSC4
NHDF X X X
183
differentiated into chondrocytes. The osteogenic differentiation of hBMSC2 was not
examined as there was not sufficient quantity of cells to perform the differentiation assay.
Dexamethasone had to be added to the differentiation media for osteogenic
differentiation. The remaining hBMSCs differentiated into osteoblasts when cultured in
osteogenic media. Fibroblasts did not differentiate when cultured in the differentiation
media, but were fully viable (figure 6.2)
6.5 Surface antigen profile of human BMSCs
The surface antigen profile of the hBMSCs was examined by flow cytometry using the
markers listed in table 6.2 and compared to the fibroblasts. Figure 6.3 shows the surface
antigen profile of hBMSC3 and the results suggest that it consisted of a homogeneous
population of cells, as a single peak was observed in all histograms. The surface antigen
profile of hBMSC2 was not examined due to lack of sufficient cell numbers. As shown in
table 6.3 all hBMSCs examined were positive for HLA-ABC and negative for HLA-DR
in agreement with the literature (Majumdar et al., 2003). Surface antigen profile analysis
of hBMSCs revealed hBMSC1 and hBMSC4 expressed STRO-1, the typical ‘MSC’
candidate marker, while hBMSC3 and fibroblasts did not. All cells expressed the
fibroblast marker CD90, and CD13 (a myeloid cell marker), which hBMSC3 and the
fibroblasts expressed at greater levels than hBMSC1 and hBMSC 4. hBMSC1 and
hBMSC3 were both positive for candidate MSC markers CD29 and CD166 which
hBMSC 4 was negative for, while the fibroblasts also expressed CD166. All cells were
negative for the haematopoietic markers CD34 and CD45 as well as SSEA1, CD11b,
CD133, and CXCR4. CD122, Interleukin 2 receptor beta was expressed by all cells.
6.6 Cytogenetic analysis of human BMSCs
All hBMSCs were karyotypically normal except for two cells in hBMSC2 (figure 6.4 and
6.5). hBMSC2 showed a normal female chromosome complement and banding pattern.
184
Table 6.2 Surface antigens examined by flow cytometry for human BMSCs.
Antibody Antigen
STRO-1 STRO-1
SSEA-1 Stage specific embryonic antigen 1
CXCR4 Chemokine (C-X-C motif) receptor 4 (Fusin)
HLA-ABC Human leukocyte antigen- ABC
HLA-DR Human leukocyte antigen- DR
CD11b Leucocyte antigen
CD13 Alanine aminopeptidase
CD29 Integrin β1
CD34 Cluster of differentiation 34
CD45 Leukocyte common antigen, Ly5
CD49e Integrin α5 (ITGA5) (fibronectin receptor, alpha polypeptide)
CD73 Ecto-5′-nucleotidase
CD90 Thy-1
CD105 Endoglin/ TGF-βR
CD106 Vascular cell adhesion molecule -1 (VCAM-1)
CD122 Interleukin 2 receptor beta
CD133 Prominin 1
CD166 Activated leukocyte cell adhesion molecule (ALCAM)
185
Figure 6.3 Surface antigen profile of human bone marrow stromal cells. Flow cytometric analysis of
the surface proteins expressed by human bone marrow stromal cells (hBMSCs) was performed by
labelling cells with fluorochrome-conjugated antibody specific for the marker indicated in each histogram.
The surface antigen profile of BMSC3 is shown. Labelled cells were analysed by FACS using BD
FACSDiva™ software, subsequent results analysed by FLOWJO software. Grey lines and magenta lines
in each histogram represent cells labelled with isotype matched control and specific monoclonal antibody,
respectively. (Results continued on next page).
.
CD11b
CXCR4SSEA1
HLA -ABC HLA -DR
CD13 CD29 CD34
STRO-1
Relative Intensity
C
ou nt s 186
CD45 CD49e CD73
CD90 CD105 CD106
CD122 CD133 CD166
Figure 6.3 (continued) Surface antigen profile of human bone marrow stromal cells.
Relative Intensity
C
ou nt s 187
Table 6.3 Surface phenotype of human BMSCs. Human bone marrow
stromal cells (hBMSCs) were harvested at 70% confluence and labelled with
PE- or FITC- antibodies or isotype matched controls and analysed by flow
cytometry at passage 2-4. The surface antigen profile of neonatal human dermal
fibroblasts (NHDFs) was also examined. Symbols refer to intensity of
expression. Samples were scored as ‘-’ if the population was in the first decade
(same as the isotype control), ‘+’ if it was between the first and second decade,
‘+’ if it was in the second decade, ‘++’ if it was in the third decade, relative to
the isotype control.
Antibody hBMSC1 hBMSC3 hBMSC4 NHDF
STRO-1 + - ± -
SSEA-1 - - - -
CXCR4 - - - -
HLA-ABC ± + ± ±
HLA-DR - - - -
CD11b - - - -
CD13 ± ++ ± ++
CD29 ± ± - -
CD34 - - - -
CD45 - - - -
CD49e ± + ± +
CD73 - + - +
CD90 + ++ ± ++
CD105 - ± - +
CD106 - + - -
CD122 ± ± ± -
CD133 - - - -
CD166 ± + - +
188
Figure 6.4 Human BMSCs have a normal karyotype. Human bone marrow
stromal cells (hBMSCs) from donor 1 displayed a normal diploid karyotype as
assessed by G-banding analysis. (A) G-banding pattern is representative of
hBMSC3 and 4. (B) Neonatal human dermal fibroblasts. All cultures were
examined at passage 3.
1 2 3 4 5
6 7 8 9 10 1211
13 14 15 16 1817
19 20 2221 X Y
46,XY
A
B
189
Figure 6.5 Cytogenetic analysis of human BMSC2. The majority (18/20) of
human BMSC2 displayed a normal diploid karyotype as assessed by G-banding
analysis at passage 3 (A). One cell was hyperdiploid with 52 chromosomes and 4
copies of chromosome 5, 3 copies of chromosome 7, 3 copies of chromosome
10, 3 copies of chromosome 15, and 3 copies of chromosome 20 (B). There was
also a single cell with a balanced reciprocal translocation between 1
chromosome 1 at breakpoint p34 and 1 chromosome 5 at breakpoint q31 (C).
A
B
C
190
However, there was a hyperdiploid cell with a complement of 52 chromosomes and four
copies of chromosome 5, three copies of chromosome 7, three copies of chromosome 10,
three copies of chromosome 15, and three copies of chromosome 20. Also a single cell
with an apparently balanced reciprocal translocation between one chromosome 1 at
breakpoint p34 and one chromosome 5 at breakpoint q31 was observed. The neonatal
human dermal fibroblasts (NHDFs) were karyotypically normal.
6.7 Human BMSCs express STEAP1
STEAP1 protein expression was examined in the hBMSCs by western blot analysis.
Initially we only had hBMSC1 and therefore examined STEAP1 in this population. A
band at the correct molecular weight (36kDa) was detected with the Santa-Cruz (sc10262) antibody and there was negligible level in the fibroblasts (figure 6.6). In addition,
to the 36 kDa STEAP1 band there was also a very small band above it at 38 kDa. We
next derived hBMSCs from the second donor (hBMSC2), so STEAP1 expression in this
cell population was analysed. As shown in figure 6.7 all hBMSCs express STEAP1 at
significant levels compared to human fibroblasts and human chondrocytes. The presence
of the STEAP bands were not particularly clearly visible in this western blot, and may be
due to poor separation as a 12% gel was used. We therefore used a 4-12% gradient gel in
the next experiment. At this point we had derived another hBMSC culture, and therefore
STEAP1 expression was examined in all three hBMSC cultures. As shown in figure 6.8
all hBMSCs expressed STEAP1 in comparison to human fibroblasts. Interestingly donor
3 expressed greater levels compared to donor 1 and 2. There was also the presence of the
38 kDa band, and this band and the STEAP1 36kD band were differentially expressed by
the hBMSCs compared to the fibroblasts.
Examination of STEAP1 protein expression in the hBMSC cultures was also attempted
via flow cytometry, however the antibodies were not suitable for this application.
191
52 kDa
38 kDa
31 kDa
NH
DF
hBM
SC1
Figure 6.6 Human bone marrow stromal cells (hBMSC1s)
express STEAP1. Human bone marrow stromal cells from
donor 1 show STEAP1 expression by western blot analysis. In
comparison, normal human dermal fibroblasts (NHDFs)
express negligible levels. All cells at passage 1-3.
STEAP1
α- Tubulin
192
52 kDa
38 kDa
31 kDa
NH
DF
hBM
SC1
hBM
SC2
Cho
ndro
cyte
s hBMSCs
76 kDa
Figure 6.7 Human BMSC1 and 2 express STEAP1. Human bone marrow stromal
cells (hBMSCs) from 2 different donors show STEAP1 expression by western blot
analysis. In comparison, neonatal human dermal fibroblasts (NHDFs), and human
chondrocytes expressed negligible levels. Human chondrocytes were derived from
articular cartilage in the femoral condyle. All cells at passage 1-3.
STEAP1
α- Tubulin
193
Figure 6.8 Human BMSC1-3 express STEAP1. Human bone marrow
stromal cells (hBMSCs) from 3 different donors show STEAP1 expression by
western blot analysis. In comparison, neonatal human dermal fibroblasts
(NHDFs) express negligible levels. All cells at passage 1-3.
52 kDa
38 kDa
31 kDa
α- Tubulin
STEAP1
NH
DF
hBM
SC1
hBM
SC2
hBM
SC3
hBMSCs
194
6.8 Human BMSCs express STEAP2
STEAP2 protein expression in hBMSCs was initially examined by western blot analysis.
Two antibodies were tested as mentioned in chapter 2: Pro-Sci and Novus –Biological
antibodies. Neither antibody was appropriate for western blotting. However, the novusbiological antibody was suitable for FACS analysis. Therefore, flow cytometry was used
to examine STEAP2 expression in hBMSCs. Results show both hBMSC3 and hBMSC4
express STEAP2 (figure 6.9). Whilst neonatal human dermal fibroblasts and hBMMNCs
were negative, (i.e. did not express it).
The murine clones expressed the stem cell antigen-1 (Sca-1) at a high intensity, which
subsequently decreased dramatically when the clones were differentiated into adipocytes.
In order to examine whether this expression pattern was the same for STEAP2 in the
hBMSCs, hBMSC3 and hBMSC4 were cultured in the presence of adipogenic media for
4 weeks to differentiate them into adipocytes. FACS analysis was performed on the
adipocyte differentiated hBMSCs and equivalent control cells cultured in the absence of
adipogenic media for the same length of time. As shown in figure 6.10 STEAP2
expression was similar in both adipocyte differentiated hBMSCs and control hBMSCs.
STEAP2 expression in hBMSC3s was 50% in normal hBMSCs examined at 70%
confluence and remains approximately the same after 4 weeks culture with or without
adipogenic media.
STEAP2 expression in hBMSC4 was detected on 35% of the cells when examined when
the cultures were 70% confluent (figure 6.9). This increased to 81% for overconfluent
cultures, i.e., the controls in the adipogenic differentiation experiments (see figure 6.10).
In adipocyte differentiated hBMSC4 cultures, STEAP2 was present in 47% of the cells.
6.9 STEAP1 depletion in human BMSCs
In order to begin to elucidate the function of STEAP1 we attempted to deplete this
protein using RNA interference. Bone marrow stromal cells from donor 3 were seeded at
approximately 50% confluence and cultured for two days. On day two cells were
transfected with siRNA oligos directed against luciferase, a scrambled, and two different
195
Figure 6.9 Human bone marrow stromal cells express STEAP2. Flow
cytometric analysis of STEAP2 expression on human BMSCs was performed
by labelling cells with fluorochrome-conjugated antibody specific for STEAP2
protein. Labelled cells were analysed by FACS using BD FACSDiva™
software and subsequent results analysed by FLOWJO software. Grey lines and
coloured (magenta and green) lines in each histogram represent cells labelled
with isotype matched control and STEAP2 antibody, respectively. Numbers
over bars represent STEAP2 percentage expression. STEAP2 was not detected
in NHDFs or hBMMNCs. Neonatal human dermal fibroblasts (NHDFs),
human bone marrow stromal cells (hBMMNCs).
hBMSC4hBMSC3
NHDF
0
hBMMNCs
2.11
Relative Intensity
C
ou nt s 196
197
Figure 6.10 STEAP2 expression is retained in adipocyte differentiated
human BMSCs. Human bone marrow stromal cells from donors 3 and 4
(hBMSC3 & 4) were cultured for four weeks in adipocyte differentiation media
or its absence. Flow cytometric analysis of human BMSCs was performed by
labelling cells with fluorochrome-conjugated antibody specific for STEAP2.
Labelled cells were analysed by FACS using BD FACSDiva™ software and
subsequent results analysed by FLOWJO software. Grey lines and magenta lines
in each histogram represent cells labelled with isotype matched control and
specific STEAP2 antibody, respectively. Upper bars represent STEAP2
percentage expression and lower bars isotype matched control.
(-) Adipogenic media (+) Adipogenic media
hBMSC3
hBMSC4
C
ou nt s Relative Intensity
STEAP1 (called STEAP1.1 and STEAP1.2) and STEAP2 (called STEAP2.1 and
STEAP2.2) transcripts. Cells were then cultured for two days and then lysed. Knockdown
efficiency was then examined by analysing STEAP1 and STEAP2 gene expression via
real time RT-PCR. As shown in figure 6.11, STEAP1 expression is depleted by 62% and
80% by STEAP1.1 and STEAP1.2 siRNAs, respectively (relative to control). STEAP2
gene expression is depleted 82% and 83% by STEAP2.1 and STEAP2.2 siRNAs,
respectively relative to control. STEAP1 depletion was further examined at the protein
level via western blot. Results confirm the real-time PCR data in that there is greater
depletion of STEAP1 protein by STEAP1.2 siRNA than by STEAP1.1 siRNA (figure
6.12). In contrast, STEAP1 protein was present in cells transfected with luciferase and
scrambled siRNA oligos and using only lipofectamine. The 38kDa band observed in
figures 6.6 & 6.8 was faintly visible in the lipofectamine alone and the luciferase lanes. It
was absent in the cells transfected with scrambled oligos. However, this could be due to
the fact that the lane was underloaded, as revealed by the loading control α-tubulin. This
band was also absent in the cells transfected with siRNA oligos against STEAP1.
6.10 Discussion
It is difficult to clonally isolate human MSCs, and there are not many reports of clonallyderived MSCs in the literature. Pittenger and colleagues isolated human clonal cells but
these were derived from previously passaged adherent cells. Marrow isolated adult
multilineage inducible cells (MIAMI) are clonally isolated in specific culture conditions,
with the addition of various growth factors (D'Ippolito et al., 2004). Clonal cells were
extensively isolated by the Muraglia group (Muraglia et al., 2000), and one clone was
isolated by the Jia group (Jia et al., 2002). Given the plethora of MSC studies this is a
very small fraction, therefore highlighting the intrinsic difficulty of the task. Initially we
attempted to derive human MSC clones. However, clones formed very rarely and when
they did, they became senescent once they were re-seeded for expansion. Unfortunately,
time was a limiting factor in this study, and due to the length of time taken to isolate and
expand murine clonal MSCs and the subsequent characterisation required we decided to
culture adherent human bone marrow stromal cells, which were not clonally isolated.
198
Figure 6.11 Validation of STEAP depletion by real-time PCR. Human bone
marrow stromal cells from donor 3 were seeded at 1300cells/cm2, ~50%
confluence. After two days cells were transfected with two specific siRNAs
against STEAP1 (S1.1 and S1.2), STEAP2 (STEAP2.1 and STEAP2.2) and
controls; luciferase and scrambled oligo, or lipofectamine alone and cultured for
two days. Cells were then lysed and STEAP1 (A) and STEAP2 (B) gene
expression levels analysed by real-time PCR, normalising expression levels to
endogenous control RPLPO. STEAP1 and 2 gene expression levels relative to
control luciferase alone sample are shown.
0
0.2
0.4
0.6
0.8
1
1.2
Luciferase STEAP1.1 STEAP1.2
R
el at iv e m R
N
A
le ve l A
0
0.2
0.4
0.6
0.8
1
1.2
Luciferase STEAP2.1 STEAP2.2
R
el at iv e m R
N
A
le ve l B
199
Lip
ofe
ctam
ine
alon
e Luc
ifer
ase
Scr
am ble
d S1.
1 1
0nM
S1.
2 1
0nM
38 kDa
31 kDa
STEAP1
α- Tubulin
Figure 6.12 STEAP1 depletion in human BMSCs. Human bone marrow
stromal cells (hBMSCs) from donor 3 were seeded at a denisty of
1300cells/cm2, ~50% confluence and transfected with two specific siRNAs
against STEAP1 (S1.1 and S1.2) and controls; luciferase, scrambled oligo, or
lipofectamine alone and cultured for two days. Cells were then lysed and
STEAP1 protein expression examined by western blot analysis.
200
Studies have reported enrichment of MSCs with passage of such cells (Baksh et al.,
2004). We further hypothesised that cultivation in a low oxygen environment may help
maintain stem cell properties as this was the case for the murine clones (chapter 3).
Human BMSCs were isolated by seeding human bone marrow mononuclear cells
(hBMMNCs) at a high density and removing non-adherent cells the following day. Cells
were cultured at 5% oxygen and were subsequently passaged when they were 70%
confluent and further expanded. Despite the fact that the hBMSCs are a heterogeneous
population to begin with, the morphology of the cells in the hBMSC3 and hBMSC4
cultures was predominantly spindle shaped, while the hBMSC1 and hBMSC2 contained
some flattened shaped cells characteristic of senescent cells, which appeared and
increased in number (especially in hBMSC2) with time in culture.
The ability of the hBMSCs to differentiate into the adipogenic, chondrogenic, and
osteogenic lineages was examined to determine presence of MSCs. The same
differentiation media that was used for the murine clones was used in these experiments.
All hBMSCs had the potential to differentiate into chondrocytes in a manner similar to
the murine clones. However, the osteogenic differentiation media containing A2P, β-GP
and BMP2 was not sufficient to stimulate osteogenesis of hBMSCs. Unlike the murine
clones, dexamethasone (which is typically added for osteogenesis of both human and
murine cells, and for cells from other species) (Peister et al., 2004; Pittenger et al., 1999)
had to be added to the media of hBMSC cultures to induce osteogenesis. Dexamethasone
is a synthetic glucocorticoid belonging to the steroid hormone family. It is very potent approximately 20-30 times more so than hydrocortisone. Glucocorticoids have been
shown to be potent inducers of osteogenic differentiation in vitro (Kirton et al., 2006).
Species differences have been previously reported, e.g., it has been shown that large
doses of BMP-2 are required to have the same osteogenic effect in human cells than
occurs in mouse cells at much lower doses (Rawadi et al., 2003). In vivo osteogenic
differentiation of MSCs has been widely reported and seems to be the default pathway as
it occurs when MSCs are transplanted in ceramic implants and implanted beneath the
skin (Lennon et al., 2001). It seems a factor(s) is missing in the differentiation medium if
201
a potent chemical such as dexamethasone has to be added for in vitro osteogenic
differentiation of MSCs, when in vivo osteogenic differentiation can occur so readily.
The same adipogenic differentiation media used for the murine clones was also used for
the human BMSCs. It should be noted that a low level of chemicals was used for this in
comparison to previously reported studies. Adipogenesis took longer than the two weeks
observed for the murine clones in the present study. It also did not take place to the same
extent in hBMSC2 cultures. This could be due to the cells, which were from an older
donor (60 years old), and contained the flattened cells characteristic of senescent cells,
and were not very proliferative. The default differentiation pathway in murine MSCs is
reported to be adipogenesis, while in human MSCs, it is osteogenesis (D'Ippolito et al.,
2004). This could be one reason why it took the human BMSCs longer to differentiate.
Also, we had optimised the adipogenic media for the murine cells, and thus added a small
quantity of the various chemicals. It is likely that if we used a greater quantity as used in
most other studies, the cells may have differentiated more quickly.
Previous studies have reported that the age of the donor affects the differentiation
potential of BMSCs. There is a decrease in osteoprogenitor cells in the bone marrow
from older animals (Quarto et al., 1995). Passage zero clones were isolated by the
Muraglia group, in the presence and absence of FGF-2 (Muraglia et al., 2000). An
increase in the frequency of bi-potential clones with osteo-chondro potential was found in
the bone marrow from younger donors compared to older donors. The donor ages were 5
months to 30 years. This is half the age of donor two in this study, and therefore it is
possible this is one reason for the decreased differentiation potential of BMSCs from this
donor. The Muraglia group examined the ability of 185 human clones to differentiate,
and all clones had the ability to differentiate into the osteogenic lineage.
The growth rate of hBMSC3 was followed at 5% oxygen and was constant between the
observed passages (2-5) with an averaged cell doubling time of 3.42 days ± 0.78 days
(mean ± standard deviation). The cell doubling time of other human clonal MSCs has
been reported to be 1.3 days in the log phase in the initial 20 days in culture and then 4.4
days when they had entered the plateau phase and this remained constant for 100 days
202
(Muraglia et al., 2000). Typically the growth rate of murine MSCs is quicker than that of
human MSCs. This could be due to species differences, and, of course, the lifespan of a
mouse is much shorter than a human.
The surface antigen profile of the hBMSCs and fibroblasts was examined using flow
cytometry. Human BMSC3 and fibroblasts were positive for CD13 (myeloid cell marker)
CD49e (integrin) and CD90 (fibroblast marker). CD105, a putative MSC marker was
only expressed by hBMSC3 and by the fibroblasts, and CD166 (another putative MSC
marker) was expressed by hBMSC1 and hBMSC3 and the fibroblasts. There did not seem
to be much consistency between the cell surface antigen phenotype expression pattern of
the hBMSCs. In fact hBMSC3 and the fibroblasts had a more similar surface antigen
phenotype. Both populations expressing the markers mentioned above as well as CD73,
CD105, and CD166. In particular, both cell populations expressed CD13 and CD90 at
greater intensity compared with hBMSC1 and hBMSC4. However, there are significant
biological differences between the two cell populations. Human BMSC3 cultures have a
tri-lineage differentiation potential which the fibroblasts lack. In addition, hBMSC3 also
expressed significant levels of STEAP1 and 2 compared to the fibroblast cultures. Thus,
these results highlight the need for more specific MSC markers, as many cell types are
positive for the previously suggested markers.
Only the surface marker CD122 was present on all hBMSCs and absent on the
fibroblasts. CD122 is interleukin-2 receptor β (IL-2Rβ) which is expressed on immune
cells, particularly T-cells. Previously mentioned MIAMI cells which were derived in low
oxygen conditions were also positive for this marker (D'Ippolito et al., 2004). CD122 is
involved in signalling pathways via the JAK1 protein. JAK1 is reported to activate
STAT5a/b transcription factors which stimulate gene expression of D cyclins (Moriggl et
al., 1999). Expression of CD122 by the murine clones was not examined, however,
microarray analysis did show that the cell cycle gene Cdk4 is differentially expressed by
the clones (figure 5.10) and the IL-1 receptor type I by the BALB/b clones (table 5.1). It
would be interesting to see if the murine clones express CD122.
Both murine clones and human BMSCs express CD49e which is an integrin α5 also
known as ITGA5 or fibronectin receptor. CD49e associates with CD29 (β1 integrin) to
203
form VLA-5, a well-established fibronectin receptor that is expressed on many cell types
including fibroblasts, epithelial cells, endothelial cells, platelets, and peripheral T cells
(Huveneers et al., 2007). CD29 was expressed by hBMSC1 and hBMSC3 but not
hBMSC4. Murine clones consistently expressed very high levels of Sca-1 and this has
been previously reported (Peister et al., 2004). No such stem cell marker has been
identified for human MSCs and unfortunately there is no known human homologue of
Sca-1. Although, interestingly, over-expression of mouse Sca-1 affects human
haematopoietic cells in a similar way to mouse cells, suggesting there may be a protein
with a similar role in humans (Bradfute et al., 2005).
All hBMSCs were karyotypically normal except two cells in hBMSC2 cultures (figures
6.5). There was the presence of one cell with a hyperdiploid chromosome number of 52,
and another with a normal chromosome number, but with a balanced translocation
between two chromosomes. Karyotype analysis was performed on cells at passage three.
The transformation is unlikely to be due to tissue culture as the other hBMSCs were
normal. It is possible the abnormal cells were present in patient, as the donor was 60
years old and cells may have acquired the genetic abnormalities with time. The ages of
the other donors (hBMSC1, 3, and 4) were 25, 11, and 25 years respectively. Or it could
be due to fact that older cells are less genetically resilient, when placed out of their in
vivo environment and exposed to foreign conditions (cell culture), genetic abnormalities
can accumulate. However, the majority of hBMSCs were normal, and therefore results
are consistent with previous studies reporting the relative resistance to transformation of
human MSCs compared to murine cultures (Aguilar et al., 2007).
Now that we had isolated normal, multipotent, human bone marrow stromal cells we
examined whether these cells expressed STEAP1 by western blot analysis. As discussed
in chapter 5, STEAP1 has been discovered relatively recently in 1999 in human prostate
cancer cells (Hubert et al., 1999) and we are the first to report them in MSCs. Western
blotting analysis revealed the presence of two differentially expressed bands in the
hBMSCs, one large band at 36 kDa (STEAP1), the other smaller one at 38kDa in two out
of three western blots. The reason it was not observed in the second western blot could be
204
due to poor separation. A gradient gel was used in the next experiment and the two bands
were clearly observed again. Interestingly, hBMSC3 expressed significantly greater
levels of STEAP1 than hBMSC1 and hBMSC2. An observable difference between the
cells from this donor and the others is that they were more proliferative, and
differentiated into the adipogenic lineage more completely compared to hBMSC1 and
hBMSC2. Also all these cells were homogeneously spindle shaped. The specificity of the
antibody for STEAP1 was confirmed when cells were depleted of STEAP1, i.e., the 36
kDa STEAP1 band was absent (compared to the controls). The 38 kDa band was also
absent, however it was only weakly visible in the controls (lipofectamine alone and
luciferase) therefore one cannot conclude whether it was STEAP1 or a non-specific band.
The datasheet for the STEAP1 antibody used for the western blot analysis shows a photo
of STEAP1 expression in the androgen-sensitive human prostate adenocarcinoma cell
line, LNCaP and there does appear to be two very tight bands, with the upper one being
larger than the lower one and they both run lower than 36 kDa. However, this may be due
to the difference in cell type. The same antibody was also used by the Alves group (Alves
et al., 2006) to examine STEAP1 expression in tumour cell lines. There may be two
bands in the DU145 (human prostate cancer cell line) and M113 (melanoma cell line)
lanes, although greater protein separation is required to say with any certainty.
Unfortunately, the LNCaP lane is overexposed so it looks like one very large band. The
presence of two bands is not remarked on by the investigators, nor in the datasheet by
Santa-Cruz. It will be interesting to elucidate whether the two bands are STEAP1 and
what post-translational modifications may be responsible for the shift. Alternatively, it
could be a non-specific protein, however this seems unlikely since it is differentially
expressed in the hBMSCs and not the fibroblast cultures. Also, the doublet is seen in
different cell types, from prostate cancer cells (LNCaP), to hBMSCs. Unfortunately the
STEAP1 antibody was not suitable for FACS analysis. Another two were also tested, but
they also proved unsuitable.
More importantly, the western blot experiments revealed that STEAP1 is differentially
expressed in hBMSCs from three different donors. These cells had also been derived
from quite different sources. The first and fourth were from commercially obtained
205
hBMMNCs. The second hBMSCs from a bone marrow aspirate from an elderly donor
(60 years), and the third hBMSCs derived from hBMMNCs from a young (11 years old)
donor. The results also revealed the specificity of STEAP1 protein for human BMSCs, in
that it was not expressed at detectable levels by both human fibroblasts and human
articular chondrocytes. This is particularly interesting, because both cell types are of
mesenchymal origin and have been reported to express the most commonly used human
MSC marker STRO-1. Although NHDFs in this study did not express it, previous studies
have reported STRO-1 expression in human fibroblasts HS68 (Wagner et al., 2005),
periodontal ligament fibroblasts (Lallier and Spencer, 2007), and myofibroblasts
(Strakova et al., 2008). Flow cytometry analysis of human chondrocytes used in this
study also revealed significant levels of STRO-1 expression.
The expression pattern of STEAP1 in the human BMSCs is similar to that of STEAP1 in
the murine clonal MSCs, where we first discovered STEAPs. Gene expression analysis
showed that STEAP1 was differentially expressed in the murine MSCs compared to the
equivalent adipocyte differentiated clones, embryonic fibroblasts, as well as bone marrow
adherent cells, and freshly isolated bone marrow mononuclear cells. Although murine
MSCs were karyotypically abnormal, they were not malignant as there was no tumour
formation when they were injected into immunodeficient mice. Furthermore, we did
attempt to derive cytogenetically normal murine MSCs, by culturing BMMNCs in
different conditions such as, at a high density, at 20% oxygen, and also a different strain
of mice, FVB/N (as described in chapter 4). Unfortunately, all cells developed karyotypic
abnormalities.
However, STEAP1 expression in normal human multipotent BMSCs proves it is not a
marker of karyotypically abnormal cells. This is the first mention of STEAP1 in MSCs,
and its function is unknown. Due to its secondary structure investigators have speculated
that it may function as an aquaporin (Kruse et al., 2006), ion channel (Dolly and Parcej,
1996), or more recently in cell-cell communication (Challita-Eid et al., 2007). The
Challita-Eid group report STEAP1 function in mediating small molecules between
adjacent cells in culture (Challita-Eid et al., 2007). In vivo these could be nutrients,
206
metabolites, and electrolytes. All of which may play an important role in stem cell
function. This is further collaborated by immunohistochemical staining of STEAP1
which showed that it is concentrated at cell-cell boundaries in the prostate gland, with
much less staining in the luminal side of the epithelia (Hubert et al., 1999).
STEAP2 expression was examined in the hBMSCs by flow cytometry. FACS analysis
revealed that STEAP2 is differentially expressed in the two hBMSC cultures studied
(hBMSC3 and hBMSC4), compared to both human fibroblasts and hBMMNCs.
Results from chapter 4 showed that Sca-1 is expressed at high intensity in the clones and
decreases in adipocyte differentiated clones. We found this expression pattern was not the
same for STEAP2 as the protein level was the same in adipocyte differentiated cultures
and the control cells cultured for the same length of time. Thus, the results suggest
STEAP2 may not have a functional role in maintaining the undifferentiated status of the
stem cells. STEAP2 expression by western blot analysis was examined in hBMSCs, but
the antibody was not compatible with immuno-blotting. Another antibody was also tested
with the same negative results.
It would be interesting to see if STEAP1 expression is altered in the adipocyte
differentiated hBMSCs, as it is not present in chondrocytes which is one lineage into
which MSCs differentiate. It is possible that the two STEAPs have distinct roles in the
MSCs as there are distinct differences in their secondary structure. STEAP2 and other
members of the STEAP family (STEAP3 & 4) contain an oxidoreductase domain that
STEAP1 does not. These domains are thought to function as ferriductases and cupric
reductases that stimulate cellular uptake of both iron and copper in vitro and possibly in
vivo (Ohgami et al., 2006).
In initial studies STEAP1 function was probed using RNA interference. STEAP1 and 2
were successfully depleted in hBMSC3, and although no differences in proliferation were
detected, STEAP1 and 2 depleted cells showed reduced adherence to the tissue culture
substrate.
In summary, results show that STEAP1 and -2, initially identified via gene expression
analysis as candidate MSC markers in murine clonal MSCs are also differentially
207
expressed by human, normal, multipotent BMSCs. Further work is now needed to more
clearly establish their role in MSCs and ascertain their usefulness as cell-specific
markers.
208
Chapter 7
Conclusions and Future Work
209
7.1 Conclusions and future work
7.1.1 Conclusions
The rationale of this study was to identify potential MSC specific surface markers, as this
is needed to enable isolation and purification of MSCs. The intention was to search in
murine clonal MSCs and then to verify subsequent findings in human MSCs. Murine
cells were chosen due to ready availability of large amounts of freshly isolated murine
bone marrow for manipulation and derivation of clonal MSCs. Another advantage was
the derivation of different populations of strain-matched control cell populations. This
would not have been feasible with human cells.
Murine MSC clones were derived to ensure that results from any experiment would be
due directly to the MSCs rather than other cells present, as can be the case if a
heterogeneous population is used (as is commonly the case in MSC studies). Another
unique and essential aspect of the approach was that the selection and expansion culture
protocols were performed under low oxygen conditions (5% oxygen), with the rationale
of providing the conditions that more closely resemble the in vivo niche of the most
primitive marrow-residing mesenchymal stem cells. Subsequent assays proved the
clonally-derived cells had the ability to differentiate into adipocytes, chondrocytes, and
osteoblasts as well as to self-renew, which demonstrates the (at least in vitro) functional
criteria for mesenchymal stem cell status. In contrast, mouse embryonic fibroblasts did
not have this capacity. Flow cytometry analysis revealed clonal MSCs consistently
expressed stem cell antigen-1 (Sca-1). Recent studies have reported that murine cells
readily become transformed in cell culture (Todaro and Green, 1963; Tolar et al., 2007).
In the present study the MSC clones were indeed found to have an abnormal karyotype.
However, they were not malignant, as observed by the fact that there was no tumour
formation when these cells were injected into immuno-compromised mice. In addition,
the MSC clones did not have other properties described of malignant cells, such as
increased c-myc expression (Miura et al., 2006). It appears the murine MSC clonal cells
in this study had become immortalised but not malignant. In addition, the differentiation
210
of the murine MSCs was a highly co-ordinated and regulated process. In particular, in
adipocyte differentiated clones there was a significant down-regulation in proliferation
related genes such as Cdk4 and stem cell-associated gene Sca-1, with concomitant upregulation of many adipocyte markers including PPARγ. This strongly suggests the
murine clones serve as a very useful model to study MSC biology.
Therefore, the next step was to characterise the molecular signature of the clonal MSCs
using a rigorous microarray analysis. This would enable identification of potentially
MSC-specific markers by comparing the gene expression profile of MSC clones to
relevant controls. Two different analysis methods were used - a supervised analysis
examining genes clustering with Sca-1; and an unsupervised analysis where very
stringent criteria were set for differential gene expression on the clones (compared to all
controls). Thus a genome-wide analysis resulted in the identification (from both methods)
of just two cell membrane proteins (STEAP1 and STEAP2) as differentially expressed on
the murine MSCs. It was then most important to assess if human MSCs differentially
expressed these cell surface proteins.
Human bone marrow stromal cells (hBMSCs) were derived and cultured in 5% oxygen.
The human cells were multipotent and had the capacity to differentiate into adipocytes,
chondrocytes, and osteoblasts. Furthermore, the cells were karyotypically normal. Both
STEAP1 and -2 were differentially expressed in the hBMSCs, therefore strongly
suggesting that these antigens are not simply a characteristic of karyotypically abnormal
cells. FACS analysis demonstrated that hBMSC cultures expressed significant levels of
STEAP2 compared to human fibroblasts and human BMMNCs (where there was no
detectable expression). STEAP2 was retained in hBMSCs which had been differentiated
into adipocytes, suggesting that STEAP2 did not have a role in maintaining the
undifferentiated stem cell phenotype. Human BMSCs also differentially expressed
STEAP1 compared to both human fibroblasts and chondrocytes, as shown by
immunoblotting. Both of the latter cell types are of mesenchymal origin and one
(chondrocyte) is a lineage that MSCs differentiate into. Results tentatively suggest that
STEAP1 may have a potential functional role in MSCs as its depletion in hBMSCs by
RNA interference resulted in a decrease in cell adhesion.
211
STEAPs were discovered relatively recently in prostate cancer cells (Hubert et al., 1999;
Porkka et al., 2002). Due to their expression on epithelial cancer cells they may seem
unlikely markers of mesenchymal stem cells. However, MSCs are reported to
differentiate into a wide spectrum of lineages including the epithelial cells (D'Ippolito et
al., 2004; Jiang et al., 2002). Future studies are now required to establish the usefulness
of STEAP1 and -2 as MSC-specific markers.
7.1.2 Future work
The immediate future work would be to FACS sort adherent human bone marrow stromal
cells based on STEAP2 expression. The STEAP2 antibody used in this study was found
to be suitable for FACS analysis, and we report that approximately 50% of human BMSC
cultures expressed STEAP2. Once these cells are sorted, the next step would be to
examine the CFU-f ability and multilineage differentiation potential of both STEAP2
positive and STEAP2 negative populations. Evidence for the usefulness of STEAP2 as a
MSC marker would be strengthened if the STEAP2 positive population greatly exceeded
the negative population in CFU-f and differentiation assays.
Another immediate future work would be to examine STEAP1 expression in
differentiated hBMSCs compared to levels in undifferentiated hBMSCs. In addition, it
would be worth examining STEAP2 expression in chondro- and osteo-differentiated
hBMSCs. This would allow one to more clearly determine whether STEAPs are involved
in maintenance of the undifferentiated stem cell phenotype.
A more long term goal would be to sort human cells freshly isolated from human bone
marrow for STEAP2 and then perform the experiments mentioned above to assess the
CFU-f ability and tri-lineage differentiation capacity of the sorted cell populations. This
would avoid any possible changes in cell surface marker expression due to tissue culture
effects and provide clearer evidence about the eligibility of STEAP2 as a MSC surface
marker. However, this approach is technically very difficult. STEAP2 expression in the
human BMMNCs was negligible (see chapter 6), i.e., there was no detectable STEAP2
212
positive population. Furthermore, reports of the frequency of MSCs in the BMMNCs
vary from 0.001-0.01% (Pittenger et al., 1999) to 0.0001% cells (Castro-Malaspina et al.,
1980). All estimates point to the fact that MSCs are an extremely rare population in the
bone marrow. Thus if STEAP2 is a potential MSC marker, or even a marker that can be
used to enrich for MSCs, these observations suggest that it will still be a rare population
that will be sorted, and large amounts of starting material will be required of both human
BMMNCs and the STEAP2 antibody in order to isolate a sufficient number of STEAP2
positive cells for subsequent analysis.
A FACS-compatible STEAP1 antibody is not commercially available, although one has
been reported in the literature (Challita-Eid et al., 2007). If one was available in the
future, it would be important to perform the above mentioned experiments using STEAP1
antibody. In addition, it would be interesting to sort using both antibodies and determine
whether this further enriches for MSCs compared to individual STEAP sorted cell
populations.
If studies did show that STEAPs were MSC specific cell surface markers there would be
many potential applications. An MSC specific marker would enable identification and
characterisation of the MSC niche. This would be possible both in animals and humans.
Studies on STEAP knock-out mice could be attempted in order to investigate the
importance of the marker, i.e. whether the mouse survived if the marker was knocked out.
If the knock-out mouse did survive one may then investigate the subsequent
physiological effect, for example if there was normal cartilage and bone formation. These
studies may lead to a better understanding of the in vivo function of the marker and/or
MSC. Furthermore, identification of the marker would also enable the cells to be labelled
and subsequent cell tracking studies in in vivo injury models would also be insightful.
Isolation of a pure population of MSCs would also enable rapid, routine and reproducible
in vitro expansion of MSCs and the generation of the large cell numbers required for
clinical studies.
Identification of a MSC cell surface marker will allow isolation of a pure MSC
population. In the clinical setting this would enable comparison to heterogeneous MSC
213
cell populations in the treatment of specific diseases. It is possible that a pure population
of MSCs may enhance the regenerative effect, as all the cells would have a multilineage
differentiation potential, and therefore the ability to differentiate into the required cell
type. In comparison, a mixed population containing various progenitor cells and
committed cells would not. MSCs have the ability to home to sites of tissue damage and
have trophic properties which facilitate tissue repair. It is logical to think these are stem
cell properties lost upon differentiation. Therefore it is possible a pure population may
home to and engraft at sites of tissue damage better than a mixed population and
therefore result in greater clinical improvement. Thus, there are many potential
applications once a MSC-specific marker is confirmed.
214
Chapter 8
Bibliography
215
8.1 Bibliography
Abbondanzo, S.J., Gadi, I., and Stewart, C.L. (1993). Derivation of embryonic stem cell
lines. Methods Enzymol 225, 803-823.
Abdallah, B.M., and Kassem, M. (2008). Human mesenchymal stem cells: from basic
biology to clinical applications. Gene Ther 15, 109-116.
Aggarwal, S., and Pittenger, M.F. (2005). Human mesenchymal stem cells modulate
allogeneic immune cell responses. Blood 105, 1815-1822.
Aguilar, S., Nye, E., Chan, J., Loebinger, M., Spencer-Dene, B., Fisk, N., Stamp, G.,
Bonnet, D., and Janes, S.M. (2007). Murine but not human mesenchymal stem cells
generate osteosarcoma-like lesions in the lung. Stem Cells 25, 1586-1594.
Alves, P.M., Faure, O., Graff-Dubois, S., Cornet, S., Bolonakis, I., Gross, D.A.,
Miconnet, I., Chouaib, S., Fizazi, K., Soria, J.C., et al. (2006). STEAP, a prostate tumor
antigen, is a target of human CD8+ T cells. Cancer Immunol Immunother 55, 1515-1523.
Anjos-Afonso, F., and Bonnet, D. (2007). Nonhematopoietic/endothelial SSEA-1+ cells
define the most primitive progenitors in the adult murine bone marrow mesenchymal
compartment. Blood 109, 1298-1306.
Arepalli, S.R., Jones, E.P., Howcroft, T.K., Carlo, I., Wang, C., Lindahl, K.F., Singer,
D.S., and Rudikoff, S. (1998). Characterization of two class I genes from the H2-M
region: evidence for a new subfamily. Immunogenetics 47, 264-271.
Ashton, B.A., Allen, T.D., Howlett, C.R., Eaglesom, C.C., Hattori, A., and Owen, M.
(1980). Formation of bone and cartilage by marrow stromal cells in diffusion chambers in
vivo. Clin Orthop Relat Res, 294-307.
Azizi, S.A., Stokes, D., Augelli, B.J., DiGirolamo, C., and Prockop, D.J. (1998).
Engraftment and migration of human bone marrow stromal cells implanted in the brains
of albino rats--similarities to astrocyte grafts. Proc Natl Acad Sci U S A 95, 3908-3913.
Bab, I., Ashton, B.A., Gazit, D., Marx, G., Williamson, M.C., and Owen, M.E. (1986).
Kinetics and differentiation of marrow stromal cells in diffusion chambers in vivo. J Cell
Sci 84, 139-151.
Baddoo, M., Hill, K., Wilkinson, R., Gaupp, D., Hughes, C., Kopen, G.C., and Phinney,
D.G. (2003). Characterization of mesenchymal stem cells isolated from murine bone
marrow by negative selection. J Cell Biochem 89, 1235-1249.
Bagley, R.G., Rouleau, C., Morgenbesser, S.D., Weber, W., Cook, B.P., Shankara, S.,
Madden, S.L., and Teicher, B.A. (2006). Pericytes from human non-small cell lung
carcinomas: an attractive target for anti-angiogenic therapy. Microvasc Res 71, 163-174.
216
Baksh, D., Song, L., and Tuan, R.S. (2004). Adult mesenchymal stem cells:
characterization, differentiation, and application in cell and gene therapy. J Cell Mol Med
8, 301-316.
Barbero, A., Ploegert, S., Heberer, M., and Martin, I. (2003). Plasticity of clonal
populations of dedifferentiated adult human articular chondrocytes. Arthritis Rheum 48,
1315-1325.
Barry, F.P. (2003). Biology and clinical applications of mesenchymal stem cells. Birth
Defects Res C Embryo Today 69, 250-256.
Baxter, M.A., Wynn, R.F., Jowitt, S.N., Wraith, J.E., Fairbairn, L.J., and Bellantuono, I.
(2004). Study of telomere length reveals rapid aging of human marrow stromal cells
following in vitro expansion. Stem Cells 22, 675-682.
Bembi, B., Parma, A., Bottega, M., Ceschel, S., Zanatta, M., Martini, C., and Ciana, G.
(1997). Intravenous pamidronate treatment in osteogenesis imperfecta. J Pediatr 131,
622-625.
Bertrand, M., Huijbers, I., Chomez, P., and De Backer, O. (2004). Comparative
expression analysis of the MAGED genes during embryogenesis and brain development.
Dev Dyn 230, 325-334.
Bianco, P., Riminucci, M., Gronthos, S., and Robey, P.G. (2001). Bone marrow stromal
stem cells: nature, biology, and potential applications. Stem Cells 19, 180-192.
Bjerkvig, R., Tysnes, B.B., Aboody, K.S., Najbauer, J., and Terzis, A.J. (2005). Opinion:
the origin of the cancer stem cell: current controversies and new insights. Nat Rev Cancer
5, 899-904.
Bonnet, D., and Dick, J.E. (1997). Human acute myeloid leukemia is organized as a
hierarchy that originates from a primitive hematopoietic cell. Nat Med 3, 730-737.
Bonyadi, M., Waldman, S.D., Liu, D., Aubin, J.E., Grynpas, M.D., and Stanford, W.L.
(2003). Mesenchymal progenitor self-renewal deficiency leads to age-dependent
osteoporosis in Sca-1/Ly-6A null mice. Proc Natl Acad Sci U S A 100, 5840-5845.
Bradfute, S.B., Graubert, T.A., and Goodell, M.A. (2005). Roles of Sca-1 in
hematopoietic stem/progenitor cell function. Exp Hematol 33, 836-843.
Brandao-Burch, A., Utting, J.C., Orriss, I.R., and Arnett, T.R. (2005). Acidosis inhibits
bone formation by osteoblasts in vitro by preventing mineralization. Calcif Tissue Int 77,
167-174.
217
Brighton, C.T., and Heppenstall, R.B. (1971). Oxygen tension in zones of the epiphyseal
plate, the metaphysis and diaphysis. An in vitro and in vivo study in rats and rabbits. J
Bone Joint Surg Am 53, 719-728.
Bruder, S.P., Kraus, K.H., Goldberg, V.M., and Kadiyala, S. (1998). The effect of
implants loaded with autologous mesenchymal stem cells on the healing of canine
segmental bone defects. J Bone Joint Surg Am 80, 985-996.
Calvi, L.M., Adams, G.B., Weibrecht, K.W., Weber, J.M., Olson, D.P., Knight, M.C.,
Martin, R.P., Schipani, E., Divieti, P., Bringhurst, F.R., et al. (2003). Osteoblastic cells
regulate the haematopoietic stem cell niche. Nature 425, 841-846.
Caplan, A.I. (1991). Mesenchymal stem cells. J Orthop Res 9, 641-650.
Caplan, A.I. (2007). Adult mesenchymal stem cells for tissue engineering versus
regenerative medicine. J Cell Physiol 213, 341-347.
Caplan, A.I., and Bruder, S.P. (2001). Mesenchymal stem cells: building blocks for
molecular medicine in the 21st century. Trends Mol Med 7, 259-264.
Caplan, A.I., and Dennis, J.E. (2006). Mesenchymal stem cells as trophic mediators. J
Cell Biochem 98, 1076-1084.
Castro-Malaspina, H., Gay, R.E., Resnick, G., Kapoor, N., Meyers, P., Chiarieri, D.,
McKenzie, S., Broxmeyer, H.E., and Moore, M.A. (1980). Characterization of human
bone marrow fibroblast colony-forming cells (CFU-F) and their progeny. Blood 56, 289301.
Castro-Malaspina, H., and Jhanwar, S.C. (1984). Properties of myelofibrosis-derived
fibroblasts. Prog Clin Biol Res 154, 307-322.
Challita-Eid, P.M., Morrison, K., Etessami, S., An, Z., Morrison, K.J., Perez-Villar, J.J.,
Raitano, A.B., Jia, X.C., Gudas, J.M., Kanner, S.B., et al. (2007). Monoclonal antibodies
to six-transmembrane epithelial antigen of the prostate-1 inhibit intercellular
communication in vitro and growth of human tumor xenografts in vivo. Cancer Res 67,
5798-5805.
Chamberlain, G., Fox, J., Ashton, B., and Middleton, J. (2007). Concise review:
mesenchymal stem cells: their phenotype, differentiation capacity, immunological
features, and potential for homing. Stem Cells 25, 2739-2749.
Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., and Smith, A.
(2003). Functional expression cloning of Nanog, a pluripotency sustaining factor in
embryonic stem cells. Cell 113, 643-655.
218
Chin, L., Artandi, S.E., Shen, Q., Tam, A., Lee, S.L., Gottlieb, G.J., Greider, C.W., and
DePinho, R.A. (1999). p53 deficiency rescues the adverse effects of telomere loss and
cooperates with telomere dysfunction to accelerate carcinogenesis. Cell 97, 527-538.
Cho, J.H., Itoh, T., Sendai, Y., and Hoshi, H. (2008). Fibroblast growth factor 7
stimulates in vitro growth of oocytes originating from bovine early antral follicles. Mol
Reprod Dev.
Colter, D.C., Sekiya, I., and Prockop, D.J. (2001). Identification of a subpopulation of
rapidly self-renewing and multipotential adult stem cells in colonies of human marrow
stromal cells. Proc Natl Acad Sci U S A 98, 7841-7845.
Corcione, A., Benvenuto, F., Ferretti, E., Giunti, D., Cappiello, V., Cazzanti, F., Risso,
M., Gualandi, F., Mancardi, G.L., Pistoia, V., et al. (2006). Human mesenchymal stem
cells modulate B-cell functions. Blood 107, 367-372.
Csete, M. (2005). Oxygen in the cultivation of stem cells. Ann N Y Acad Sci 1049, 1-8.
D'Ippolito, G., Diabira, S., Howard, G.A., Menei, P., Roos, B.A., and Schiller, P.C.
(2004). Marrow-isolated adult multilineage inducible (MIAMI) cells, a unique population
of postnatal young and old human cells with extensive expansion and differentiation
potential. J Cell Sci 117, 2971-2981.
D'Ippolito, G., Diabira, S., Howard, G.A., Roos, B.A., and Schiller, P.C. (2006). Low
oxygen tension inhibits osteogenic differentiation and enhances stemness of human
MIAMI cells. Bone 39, 513-522.
da Silva Meirelles, L., Caplan, A.I., and Nardi, N.B. (2008). In search of the in vivo
identity of mesenchymal stem cells. Stem Cells 26, 2287-2299.
Davani, S., Marandin, A., Mersin, N., Royer, B., Kantelip, B., Herve, P., Etievent, J.P.,
and Kantelip, J.P. (2003). Mesenchymal progenitor cells differentiate into an endothelial
phenotype, enhance vascular density, and improve heart function in a rat cellular
cardiomyoplasty model. Circulation 108 Suppl 1, II253-258.
Dazzi, F., Ramasamy, R., Glennie, S., Jones, S.P., and Roberts, I. (2006). The role of
mesenchymal stem cells in haemopoiesis. Blood Rev 20, 161-171.
De Bari, C., Dell'Accio, F., Tylzanowski, P., and Luyten, F.P. (2001). Multipotent
mesenchymal stem cells from adult human synovial membrane. Arthritis Rheum 44,
1928-1942.
De Coppi, P., Bartsch, G., Jr., Siddiqui, M.M., Xu, T., Santos, C.C., Perin, L.,
Mostoslavsky, G., Serre, A.C., Snyder, E.Y., Yoo, J.J., et al. (2007). Isolation of amniotic
stem cell lines with potential for therapy. Nat Biotechnol 25, 100-106.
219
De Ugarte, D.A., Morizono, K., Elbarbary, A., Alfonso, Z., Zuk, P.A., Zhu, M., Dragoo,
J.L., Ashjian, P., Thomas, B., Benhaim, P., et al. (2003). Comparison of multi-lineage
cells from human adipose tissue and bone marrow. Cells Tissues Organs 174, 101-109.
Dennis, J.E., Carbillet, J.P., Caplan, A.I., and Charbord, P. (2002). The STRO-1+ marrow
cell population is multipotential. Cells Tissues Organs 170, 73-82.
Dennis, J.E., Merriam, A., Awadallah, A., Yoo, J.U., Johnstone, B., and Caplan, A.I.
(1999). A quadripotential mesenchymal progenitor cell isolated from the marrow of an
adult mouse. J Bone Miner Res 14, 700-709.
Dexter, T.M., Wright, E.G., Krizsa, F., and Lajtha, L.G. (1977). Regulation of
haemopoietic stem cell proliferation in long term bone marrow cultures. Biomedicine 27,
344-349.
Di Nicola, M., Carlo-Stella, C., Magni, M., Milanesi, M., Longoni, P.D., Matteucci, P.,
Grisanti, S., and Gianni, A.M. (2002). Human bone marrow stromal cells suppress Tlymphocyte proliferation induced by cellular or nonspecific mitogenic stimuli. Blood 99,
3838-3843.
Diaz-Romero, J., Nesic, D., Grogan, S.P., Heini, P., and Mainil-Varlet, P. (2008).
Immunophenotypic changes of human articular chondrocytes during monolayer culture
reflect bona fide dedifferentiation rather than amplification of progenitor cells. J Cell
Physiol 214, 75-83.
Diefenderfer, D.L., Osyczka, A.M., Garino, J.P., and Leboy, P.S. (2003). Regulation of
BMP-induced transcription in cultured human bone marrow stromal cells. J Bone Joint
Surg Am 85-A Suppl 3, 19-28.
Digirolamo, C.M., Stokes, D., Colter, D., Phinney, D.G., Class, R., and Prockop, D.J.
(1999). Propagation and senescence of human marrow stromal cells in culture: a simple
colony-forming assay identifies samples with the greatest potential to propagate and
differentiate. Br J Haematol 107, 275-281.
Djouad, F., Delorme, B., Maurice, M., Bony, C., Apparailly, F., Louis-Plence, P.,
Canovas, F., Charbord, P., Noel, D., and Jorgensen, C. (2007). Microenvironmental
changes during differentiation of mesenchymal stem cells towards chondrocytes. Arthritis
Res Ther 9, R33.
Dolly, J.O., and Parcej, D.N. (1996). Molecular properties of voltage-gated K+ channels.
J Bioenerg Biomembr 28, 231-253.
Dominici, M., Le Blanc, K., Mueller, I., Slaper-Cortenbach, I., Marini, F., Krause, D.,
Deans, R., Keating, A., Prockop, D., and Horwitz, E. (2006). Minimal criteria for
defining multipotent mesenchymal stromal cells. The International Society for Cellular
Therapy position statement. Cytotherapy 8, 315-317.
220
Dowthwaite, G.P., Bishop, J.C., Redman, S.N., Khan, I.M., Rooney, P., Evans, D.J.,
Haughton, L., Bayram, Z., Boyer, S., Thomson, B., et al. (2004). The surface of articular
cartilage contains a progenitor cell population. J Cell Sci 117, 889-897.
Duncan, R., and McConkey, E.H. (1982). How many proteins are there in a typical
mammalian cell? Clin Chem 28, 749-755.
Earle, W.R. (1962). Some morphologic variations of certain cells under controlled
experimental conditions. Natl Cancer Inst Monogr 7, 213-236.
Edgar, C.M., Chakravarthy, V., Barnes, G., Kakar, S., Gerstenfeld, L.C., and Einhorn,
T.A. (2007). Autogenous regulation of a network of bone morphogenetic proteins
(BMPs) mediates the osteogenic differentiation in murine marrow stromal cells. Bone 40,
1389-1398.
Edman, L.C., Mira, H., and Arenas, E. (2008). The beta-chemokines CCL2 and CCL7 are
two novel differentiation factors for midbrain dopaminergic precursors and neurons. Exp
Cell Res 314, 2123-2130.
Eslaminejad, M.B., Nadri, S., and Hosseini, R.H. (2007). Expression of Thy 1.2 surface
antigen increases significantly during the murine mesenchymal stem cells cultivation
period. Dev Growth Differ 49, 351-364.
Fehrer, C., Brunauer, R., Laschober, G., Unterluggauer, H., Reitinger, S., Kloss, F.,
Gully, C., Gassner, R., and Lepperdinger, G. (2007). Reduced oxygen tension attenuates
differentiation capacity of human mesenchymal stem cells and prolongs their lifespan.
Aging Cell 6, 745-757.
Ferri, A.L., Lin, W., Mavromatakis, Y.E., Wang, J.C., Sasaki, H., Whitsett, J.A., and
Ang, S.L. (2007). Foxa1 and Foxa2 regulate multiple phases of midbrain dopaminergic
neuron development in a dosage-dependent manner. Development 134, 2761-2769.
Field, R.E., Buchanan, J.A., Copplemans, M.G., and Aichroth, P.M. (1994). Bonemarrow transplantation in Hurler's syndrome. Effect on skeletal development. J Bone
Joint Surg Br 76, 975-981.
Finch, P.W., and Rubin, J.S. (2004). Keratinocyte growth factor/fibroblast growth factor
7, a homeostatic factor with therapeutic potential for epithelial protection and repair. Adv
Cancer Res 91, 69-136.
Forbes, S.J., Vig, P., Poulsom, R., Wright, N.A., and Alison, M.R. (2002). Adult stem
cell plasticity: new pathways of tissue regeneration become visible. Clin Sci (Lond) 103,
355-369.
Forte, G., Minieri, M., Cossa, P., Antenucci, D., Sala, M., Gnocchi, V., Fiaccavento, R.,
Carotenuto, F., De Vito, P., Baldini, P.M., et al. (2006). Hepatocyte growth factor effects
221
on mesenchymal stem cells: proliferation, migration, and differentiation. Stem Cells 24,
23-33.
Foster, L.J., Zeemann, P.A., Li, C., Mann, M., Jensen, O.N., and Kassem, M. (2005).
Differential expression profiling of membrane proteins by quantitative proteomics in a
human mesenchymal stem cell line undergoing osteoblast differentiation. Stem Cells 23,
1367-1377.
Frank, L., and Massaro, D. (1980). Oxygen toxicity. Am J Med 69, 117-126.
Frank, L., Summerville, J., and Massaro, D. (1980). Potection from oxygen toxicity with
endotoxin. Role of the endogenous antioxidant enzymes of the lung. J Clin Invest 65,
1104-1110.
Friedenstein, A.J. (1961). Osteogenetic activity of transplanted transitional epithelium.
Acta Anat (Basel) 45, 31-59.
Friedenstein, A.J., Petrakova, K.V., Kurolesova, A.I., and Frolova, G.P. (1968).
Heterotopic of bone marrow.Analysis of precursor cells for osteogenic and hematopoietic
tissues. Transplantation 6, 230-247.
Friedenstein, A.J., Piatetzky, S., II, and Petrakova, K.V. (1966). Osteogenesis in
transplants of bone marrow cells. J Embryol Exp Morphol 16, 381-390.
Gang, E.J., Bosnakovski, D., Figueiredo, C.A., Visser, J.W., and Perlingeiro, R.C.
(2007). SSEA-4 identifies mesenchymal stem cells from bone marrow. Blood 109, 17431751.
Ganz, T. (2000). Paneth cells--guardians of the gut cell hatchery. Nat Immunol 1, 99-100.
Garcia-Olmo, D., Garcia-Arranz, M., Garcia, L.G., Cuellar, E.S., Blanco, I.F., Prianes,
L.A., Montes, J.A., Pinto, F.L., Marcos, D.H., and Garcia-Sancho, L. (2003). Autologous
stem cell transplantation for treatment of rectovaginal fistula in perianal Crohn's disease:
a new cell-based therapy. Int J Colorectal Dis 18, 451-454.
Garcia-Olmo, D., Garcia-Arranz, M., Herreros, D., Pascual, I., Peiro, C., and RodriguezMontes, J.A. (2005). A phase I clinical trial of the treatment of Crohn's fistula by adipose
mesenchymal stem cell transplantation. Dis Colon Rectum 48, 1416-1423.
Gojo, S., Gojo, N., Takeda, Y., Mori, T., Abe, H., Kyo, S., Hata, J., and Umezawa, A.
(2003). In vivo cardiovasculogenesis by direct injection of isolated adult mesenchymal
stem cells. Exp Cell Res 288, 51-59.
Grant, J.L., and Smith, B. (1963). Bone marrow gas tensions, bone marrow blood flow,
and erythropoiesis in man. Ann Intern Med 58, 801-809.
222
Grayson, W.L., Zhao, F., Bunnell, B., and Ma, T. (2007). Hypoxia enhances proliferation
and tissue formation of human mesenchymal stem cells. Biochem Biophys Res Commun
358, 948-953.
Grayson, W.L., Zhao, F., Izadpanah, R., Bunnell, B., and Ma, T. (2006). Effects of
hypoxia on human mesenchymal stem cell expansion and plasticity in 3D constructs. J
Cell Physiol 207, 331-339.
Grellier, P., Yee, D., Gonzalez, M., and Abboud, S.L. (1995). Characterization of insulinlike growth factor binding proteins (IGFBP) and regulation of IGFBP-4 in bone marrow
stromal cells. Br J Haematol 90, 249-257.
Gronthos, S., Zannettino, A.C., Hay, S.J., Shi, S., Graves, S.E., Kortesidis, A., and
Simmons, P.J. (2003). Molecular and cellular characterisation of highly purified stromal
stem cells derived from human bone marrow. J Cell Sci 116, 1827-1835.
Gumley, T.P., McKenzie, I.F., and Sandrin, M.S. (1995). Tissue expression, structure and
function of the murine Ly-6 family of molecules. Immunol Cell Biol 73, 277-296.
Hanada, K., Dennis, J.E., and Caplan, A.I. (1997). Stimulatory effects of basic fibroblast
growth factor and bone morphogenetic protein-2 on osteogenic differentiation of rat bone
marrow-derived mesenchymal stem cells. J Bone Miner Res 12, 1606-1614.
Hayflick, L., and Moorhead, P.S. (1961). The serial cultivation of human diploid cell
strains. Exp Cell Res 25, 585-621.
Haynesworth, S.E., Baber, M.A., and Caplan, A.I. (1992). Cell surface antigens on
human marrow-derived mesenchymal cells are detected by monoclonal antibodies. Bone
13, 69-80.
Hiyama, E., and Hiyama, K. (2007). Telomere and telomerase in stem cells. Br J Cancer
96, 1020-1024.
Holmes, C., and Stanford, W.L. (2007). Concise review: stem cell antigen-1: expression,
function, and enigma. Stem Cells 25, 1339-1347.
Horwitz, E.M., Gordon, P.L., Koo, W.K., Marx, J.C., Neel, M.D., McNall, R.Y., Muul,
L., and Hofmann, T. (2002). Isolated allogeneic bone marrow-derived mesenchymal cells
engraft and stimulate growth in children with osteogenesis imperfecta: Implications for
cell therapy of bone. Proc Natl Acad Sci U S A 99, 8932-8937.
Horwitz, E.M., Prockop, D.J., Fitzpatrick, L.A., Koo, W.W., Gordon, P.L., Neel, M.,
Sussman, M., Orchard, P., Marx, J.C., Pyeritz, R.E., et al. (1999). Transplantability and
therapeutic effects of bone marrow-derived mesenchymal cells in children with
osteogenesis imperfecta. Nat Med 5, 309-313.
223
Horwitz, E.M., Prockop, D.J., Gordon, P.L., Koo, W.W., Fitzpatrick, L.A., Neel, M.D.,
McCarville, M.E., Orchard, P.J., Pyeritz, R.E., and Brenner, M.K. (2001). Clinical
responses to bone marrow transplantation in children with severe osteogenesis
imperfecta. Blood 97, 1227-1231.
Hubert, R.S., Vivanco, I., Chen, E., Rastegar, S., Leong, K., Mitchell, S.C., Madraswala,
R., Zhou, Y., Kuo, J., Raitano, A.B., et al. (1999). STEAP: a prostate-specific cellsurface antigen highly expressed in human prostate tumors. Proc Natl Acad Sci U S A 96,
14523-14528.
Huveneers, S., Truong, H., and Danen, H.J. (2007). Integrins: signaling, disease, and
therapy. Int J Radiat Biol 83, 743-751.
Imabayashi, H., Mori, T., Gojo, S., Kiyono, T., Sugiyama, T., Irie, R., Isogai, T., Hata, J.,
Toyama, Y., and Umezawa, A. (2003). Redifferentiation of dedifferentiated chondrocytes
and chondrogenesis of human bone marrow stromal cells via chondrosphere formation
with expression profiling by large-scale cDNA analysis. Exp Cell Res 288, 35-50.
Inada, M., Follenzi, A., Cheng, K., Surana, M., Joseph, B., Benten, D., Bandi, S., Qian,
H., and Gupta, S. (2008). Phenotype reversion in fetal human liver epithelial cells
identifies the role of an intermediate meso-endodermal stage before hepatic maturation. J
Cell Sci 121, 1002-1013.
Jaiswal, N., Haynesworth, S.E., Caplan, A.I., and Bruder, S.P. (1997). Osteogenic
differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J
Cell Biochem 64, 295-312.
Jia, L., Young, M.F., Powell, J., Yang, L., Ho, N.C., Hotchkiss, R., Robey, P.G., and
Francomano, C.A. (2002). Gene expression profile of human bone marrow stromal cells:
high-throughput expressed sequence tag sequencing analysis. Genomics 79, 7-17.
Jiang, X.X., Zhang, Y., Liu, B., Zhang, S.X., Wu, Y., Yu, X.D., and Mao, N. (2005).
Human mesenchymal stem cells inhibit differentiation and function of monocyte-derived
dendritic cells. Blood 105, 4120-4126.
Jiang, Y., Jahagirdar, B.N., Reinhardt, R.L., Schwartz, R.E., Keene, C.D., OrtizGonzalez, X.R., Reyes, M., Lenvik, T., Lund, T., Blackstad, M., et al. (2002).
Pluripotency of mesenchymal stem cells derived from adult marrow. Nature 418, 41-49.
Jones, E., and McGonagle, D. (2008). Human bone marrow mesenchymal stem cells in
vivo. Rheumatology (Oxford) 47, 126-131.
Jones, E.A., Kinsey, S.E., English, A., Jones, R.A., Straszynski, L., Meredith, D.M.,
Markham, A.F., Jack, A., Emery, P., and McGonagle, D. (2002). Isolation and
characterization of bone marrow multipotential mesenchymal progenitor cells. Arthritis
Rheum 46, 3349-3360.
224
Jones, P.H., and Watt, F.M. (1993). Separation of human epidermal stem cells from
transit amplifying cells on the basis of differences in integrin function and expression.
Cell 73, 713-724.
Kafienah, W., Mistry, S., Dickinson, S.C., Sims, T.J., Learmonth, I., and Hollander, A.P.
(2007). Three-dimensional cartilage tissue engineering using adult stem cells from
osteoarthritis patients. Arthritis Rheum 56, 177-187.
Kalitsis, P., Griffiths, B., and Choo, K.H. (2006). Mouse telocentric sequences reveal a
high rate of homogenization and possible role in Robertsonian translocation. Proc Natl
Acad Sci U S A 103, 8786-8791.
Kanichai, M., Ferguson, D., Prendergast, P.J., and Campbell, V.A. (2008). Hypoxia
promotes chondrogenesis in rat mesenchymal stem cells: a role for AKT and hypoxiainducible factor (HIF)-1alpha. J Cell Physiol 216, 708-715.
Kao, Y.R., Shih, J.Y., Wen, W.C., Ko, Y.P., Chen, B.M., Chan, Y.L., Chu, Y.W., Yang,
P.C., Wu, C.W., and Roffler, S.R. (2003). Tumor-associated antigen L6 and the invasion
of human lung cancer cells. Clin Cancer Res 9, 2807-2816.
Kayahara, T., Sawada, M., Takaishi, S., Fukui, H., Seno, H., Fukuzawa, H., Suzuki, K.,
Hiai, H., Kageyama, R., Okano, H., et al. (2003). Candidate markers for stem and early
progenitor cells, Musashi-1 and Hes1, are expressed in crypt base columnar cells of
mouse small intestine. FEBS Lett 535, 131-135.
Kenyon, J., and Gerson, S.L. (2007). The role of DNA damage repair in aging of adult
stem cells. Nucleic Acids Res 35, 7557-7565.
Kidd, M., Modlin, I.M., Mane, S.M., Camp, R.L., Eick, G., and Latich, I. (2006). The
role of genetic markers--NAP1L1, MAGE-D2, and MTA1--in defining small-intestinal
carcinoid neoplasia. Ann Surg Oncol 13, 253-262.
Kirton, J.P., Wilkinson, F.L., Canfield, A.E., and Alexander, M.Y. (2006).
Dexamethasone downregulates calcification-inhibitor molecules and accelerates
osteogenic differentiation of vascular pericytes: implications for vascular calcification.
Circ Res 98, 1264-1272.
Koc, O.N., Day, J., Nieder, M., Gerson, S.L., Lazarus, H.M., and Krivit, W. (2002a).
Allogeneic mesenchymal stem cell infusion for treatment of metachromatic
leukodystrophy (MLD) and Hurler syndrome (MPS-IH). Bone Marrow Transplant 30,
215-222.
Koc, O.N., Gerson, S.L., Cooper, B.W., Dyhouse, S.M., Haynesworth, S.E., Caplan, A.I.,
and Lazarus, H.M. (2000). Rapid hematopoietic recovery after coinfusion of autologousblood stem cells and culture-expanded marrow mesenchymal stem cells in advanced
breast cancer patients receiving high-dose chemotherapy. J Clin Oncol 18, 307-316.
225
Koc, S., Leisenring, W., Flowers, M.E., Anasetti, C., Deeg, H.J., Nash, R.A., Sanders,
J.E., Witherspoon, R.P., Storb, R., Appelbaum, F.R., et al. (2002b). Therapy for chronic
graft-versus-host disease: a randomized trial comparing cyclosporine plus prednisone
versus prednisone alone. Blood 100, 48-51.
Kofoed, H., Sjontoft, E., Siemssen, S.O., and Olesen, H.P. (1985). Bone marrow
circulation after osteotomy. Blood flow, pO2, pCO2, and pressure studied in dogs. Acta
Orthop Scand 56, 400-403.
Kogler, G., Sensken, S., Airey, J.A., Trapp, T., Muschen, M., Feldhahn, N., Liedtke, S.,
Sorg, R.V., Fischer, J., Rosenbaum, C., et al. (2004). A new human somatic stem cell
from placental cord blood with intrinsic pluripotent differentiation potential. J Exp Med
200, 123-135.
Krampera, M., Glennie, S., Dyson, J., Scott, D., Laylor, R., Simpson, E., and Dazzi, F.
(2003). Bone marrow mesenchymal stem cells inhibit the response of naive and memory
antigen-specific T cells to their cognate peptide. Blood 101, 3722-3729.
Krivit, W., Aubourg, P., Shapiro, E., and Peters, C. (1999). Bone marrow transplantation
for globoid cell leukodystrophy, adrenoleukodystrophy, metachromatic leukodystrophy,
and Hurler syndrome. Curr Opin Hematol 6, 377-382.
Kruse, E., Uehlein, N., and Kaldenhoff, R. (2006). The aquaporins. Genome Biol 7, 206.
Kuznetsov, S.A., Krebsbach, P.H., Satomura, K., Kerr, J., Riminucci, M., Benayahu, D.,
and Robey, P.G. (1997). Single-colony derived strains of human marrow stromal
fibroblasts form bone after transplantation in vivo. J Bone Miner Res 12, 1335-1347.
Kuznetsov, S.A., Mankani, M.H., Gronthos, S., Satomura, K., Bianco, P., and Robey,
P.G. (2001). Circulating skeletal stem cells. J Cell Biol 153, 1133-1140.
Lafont, J.E., Talma, S., and Murphy, C.L. (2007). Hypoxia-inducible factor 2alpha is
essential for hypoxic induction of the human articular chondrocyte phenotype. Arthritis
Rheum 56, 3297-3306.
Lallier, T.E., and Spencer, A. (2007). Use of microarrays to find novel regulators of
periodontal ligament fibroblast differentiation. Cell Tissue Res 327, 93-109.
Lamoury, F.M., Croitoru-Lamoury, J., and Brew, B.J. (2006). Undifferentiated mouse
mesenchymal stem cells spontaneously express neural and stem cell markers Oct-4 and
Rex-1. Cytotherapy 8, 228-242.
Lazarus, H.M., Koc, O.N., Devine, S.M., Curtin, P., Maziarz, R.T., Holland, H.K.,
Shpall, E.J., McCarthy, P., Atkinson, K., Cooper, B.W., et al. (2005). Cotransplantation
of HLA-identical sibling culture-expanded mesenchymal stem cells and hematopoietic
226
stem cells in hematologic malignancy patients. Biol Blood Marrow Transplant 11, 389398.
Le Blanc, K., Tammik, L., Sundberg, B., Haynesworth, S.E., and Ringden, O. (2003).
Mesenchymal stem cells inhibit and stimulate mixed lymphocyte cultures and mitogenic
responses independently of the major histocompatibility complex. Scand J Immunol 57,
11-20.
Lee, J.Y., Qu-Petersen, Z., Cao, B., Kimura, S., Jankowski, R., Cummins, J., Usas, A.,
Gates, C., Robbins, P., Wernig, A., et al. (2000). Clonal isolation of muscle-derived cells
capable of enhancing muscle regeneration and bone healing. J Cell Biol 150, 1085-1100.
Lennon, D.P., Edmison, J.M., and Caplan, A.I. (2001). Cultivation of rat marrow-derived
mesenchymal stem cells in reduced oxygen tension: effects on in vitro and in vivo
osteochondrogenesis. J Cell Physiol 187, 345-355.
Li, Y., Chen, J., Chen, X.G., Wang, L., Gautam, S.C., Xu, Y.X., Katakowski, M., Zhang,
L.J., Lu, M., Janakiraman, N., et al. (2002). Human marrow stromal cell therapy for
stroke in rat: neurotrophins and functional recovery. Neurology 59, 514-523.
Li, Y., Zhang, C., Xiong, F., Yu, M.J., Peng, F.L., Shang, Y.C., Zhao, C.P., Xu, Y.F.,
Liu, Z.S., Zhou, C., et al. (2008). Comparative study of mesenchymal stem cells from
C57BL/10 and mdx mice. BMC Cell Biol 9, 24.
Liechty, K.W., MacKenzie, T.C., Shaaban, A.F., Radu, A., Moseley, A.M., Deans, R.,
Marshak, D.R., and Flake, A.W. (2000). Human mesenchymal stem cells engraft and
demonstrate site-specific differentiation after in utero transplantation in sheep. Nat Med
6, 1282-1286.
Lloyd, A.C. (2002). Limits to lifespan. Nat Cell Biol 4, E25-27.
Lodie, T.A., Blickarz, C.E., Devarakonda, T.J., He, C., Dash, A.B., Clarke, J., Gleneck,
K., Shihabuddin, L., and Tubo, R. (2002). Systematic analysis of reportedly distinct
populations of multipotent bone marrow-derived stem cells reveals a lack of distinction.
Tissue Eng 8, 739-751.
Mackay, A.M., Beck, S.C., Murphy, J.M., Barry, F.P., Chichester, C.O., and Pittenger,
M.F. (1998). Chondrogenic differentiation of cultured human mesenchymal stem cells
from marrow. Tissue Eng 4, 415-428.
Majumdar, M.K., Keane-Moore, M., Buyaner, D., Hardy, W.B., Moorman, M.A.,
McIntosh, K.R., and Mosca, J.D. (2003). Characterization and functionality of cell
surface molecules on human mesenchymal stem cells. J Biomed Sci 10, 228-241.
227
Malladi, P., Xu, Y., Chiou, M., Giaccia, A.J., and Longaker, M.T. (2006a). Effect of
reduced oxygen tension on chondrogenesis and osteogenesis in adipose-derived
mesenchymal cells. Am J Physiol Cell Physiol 290, C1139-1146.
Malladi, P., Xu, Y., Yang, G.P., and Longaker, M.T. (2006b). Functions of vitamin D,
retinoic acid, and dexamethasone in mouse adipose-derived mesenchymal cells. Tissue
Eng 12, 2031-2040.
Martin-Rendon, E., Hale, S.J., Ryan, D., Baban, D., Forde, S.P., Roubelakis, M.,
Sweeney, D., Moukayed, M., Harris, A.L., Davies, K., et al. (2007). Transcriptional
profiling of human cord blood CD133+ and cultured bone marrow mesenchymal stem
cells in response to hypoxia. Stem Cells 25, 1003-1012.
Martin, I., Muraglia, A., Campanile, G., Cancedda, R., and Quarto, R. (1997). Fibroblast
growth factor-2 supports ex vivo expansion and maintenance of osteogenic precursors
from human bone marrow. Endocrinology 138, 4456-4462.
Martinez-Agosto, J.A., Mikkola, H.K., Hartenstein, V., and Banerjee, U. (2007). The
hematopoietic stem cell and its niche: a comparative view. Genes Dev 21, 3044-3060.
Mason, T.M., Lord, B.I., and Hendry, J.H. (1989). The development of spatial
distributions of CFU-S and in-vitro CFC in femora of mice of different ages. Br J
Haematol 73, 455-461.
Meirelles Lda, S., and Nardi, N.B. (2003). Murine marrow-derived mesenchymal stem
cell: isolation, in vitro expansion, and characterization. Br J Haematol 123, 702-711.
Menten, P., Wuyts, A., and Van Damme, J. (2001). Monocyte chemotactic protein-3. Eur
Cytokine Netw 12, 554-560.
Mets, T., and Verdonk, G. (1981). In vitro aging of human bone marrow derived stromal
cells. Mech Ageing Dev 16, 81-89.
Meuleman, N., Vanhaelen, G., Tondreau, T., Lewalle, P., Kwan, J., Bennani, J., Martiat,
P., Lagneaux, L., and Bron, D. (2008). Reduced intensity conditioning haematopoietic
stem cell transplantation with mesenchymal stromal cells infusion for the treatment of
metachromatic leukodystrophy: a case report. Haematologica 93, e11-13.
Mirosevich, J., Gao, N., and Matusik, R.J. (2005). Expression of Foxa transcription
factors in the developing and adult murine prostate. Prostate 62, 339-352.
Mitsiadis, T.A., Barrandon, O., Rochat, A., Barrandon, Y., and De Bari, C. (2007). Stem
cell niches in mammals. Exp Cell Res 313, 3377-3385.
Miura, M., Miura, Y., Padilla-Nash, H.M., Molinolo, A.A., Fu, B., Patel, V., Seo, B.M.,
Sonoyama, W., Zheng, J.J., Baker, C.C., et al. (2006). Accumulated chromosomal
228
instability in murine bone marrow mesenchymal stem cells leads to malignant
transformation. Stem Cells 24, 1095-1103.
Mizukami, T., Kuramitsu, M., Takizawa, K., Momose, H., Masumi, A., Naito, S., Iwama,
A., Ogawa, T., Noce, T., Hamaguchi, I., et al. (2008). Identification of transcripts
commonly expressed in both hematopoietic and germ-line stem cells. Stem Cells Dev 17,
67-80.
Modderman, W.E., Vrijheid-Lammers, T., Lowik, C.W., and Nijweide, P.J. (1994).
Removal of hematopoietic cells and macrophages from mouse bone marrow cultures:
isolation of fibroblastlike stromal cells. Exp Hematol 22, 194-201.
Montgomery, R.K., and Breault, D.T. (2008). Small intestinal stem cell markers. J Anat
213, 52-58.
Moriggl, R., Topham, D.J., Teglund, S., Sexl, V., McKay, C., Wang, D., Hoffmeyer, A.,
van Deursen, J., Sangster, M.Y., Bunting, K.D., et al. (1999). Stat5 is required for IL-2induced cell cycle progression of peripheral T cells. Immunity 10, 249-259.
Mostafa, S.S., Miller, W.M., and Papoutsakis, E.T. (2000). Oxygen tension influences the
differentiation, maturation and apoptosis of human megakaryocytes. Br J Haematol 111,
879-889.
Moussavi-Harami, F., Duwayri, Y., Martin, J.A., and Buckwalter, J.A. (2004). Oxygen
effects on senescence in chondrocytes and mesenchymal stem cells: consequences for
tissue engineering. Iowa Orthop J 24, 15-20.
Muraglia, A., Cancedda, R., and Quarto, R. (2000). Clonal mesenchymal progenitors
from human bone marrow differentiate in vitro according to a hierarchical model. J Cell
Sci 113 ( Pt 7), 1161-1166.
Murphy, C.L., and Polak, J.M. (2004). Control of human articular chondrocyte
differentiation by reduced oxygen tension. J Cell Physiol 199, 451-459.
Nakahara, H., Bruder, S.P., Haynesworth, S.E., Holecek, J.J., Baber, M.A., Goldberg,
V.M., and Caplan, A.I. (1990). Bone and cartilage formation in diffusion chambers by
subcultured cells derived from the periosteum. Bone 11, 181-188.
Nakagawa, M., Koyanagi, M., Tanabe, K., Takahashi, K., Ichisaka, T., Aoi, T., Okita, K.,
Mochiduki, Y., Takizawa, N., Yamanaka, S. (2008). Gneration of induced pluripotent
stem cells without Myc from mouse and human fibroblasts. Nat Biotechnol 1 101-6.
Nichols, J., Zevnik, B., Anastassiadis, K., Niwa, H., Klewe-Nebenius, D., Chambers, I.,
Scholer, H., and Smith, A. (1998). Formation of pluripotent stem cells in the mammalian
embryo depends on the POU transcription factor Oct4. Cell 95, 379-391.
229
Noth, U., Osyczka, A.M., Tuli, R., Hickok, N.J., Danielson, K.G., and Tuan, R.S. (2002).
Multilineage mesenchymal differentiation potential of human trabecular bone-derived
cells. J Orthop Res 20, 1060-1069.
Oh, H., Bradfute, S.B., Gallardo, T.D., Nakamura, T., Gaussin, V., Mishina, Y., Pocius,
J., Michael, L.H., Behringer, R.R., Garry, D.J., et al. (2003). Cardiac progenitor cells
from adult myocardium: homing, differentiation, and fusion after infarction. Proc Natl
Acad Sci U S A 100, 12313-12318.
Ohgami, R.S., Campagna, D.R., McDonald, A., and Fleming, M.D. (2006). The Steap
proteins are metalloreductases. Blood 108, 1388-1394.
Ohnishi, S., Yasuda, T., Kitamura, S., and Nagaya, N. (2007). Effect of hypoxia on gene
expression of bone marrow-derived mesenchymal stem cells and mononuclear cells.
Stem Cells 25, 1166-1177.
Okita, K., Ichisaka, T., Yamanaka, S. (2007). Generation of germline-competent induced
pluripotent stem cells. Nature 448, 313-18.
Orkin, S.H. (2000). Diversification of haematopoietic stem cells to specific lineages. Nat
Rev Genet 1, 57-64.
Orlic, D., Kajstura, J., Chimenti, S., Jakoniuk, I., Anderson, S.M., Li, B., Pickel, J.,
McKay, R., Nadal-Ginard, B., Bodine, D.M., et al. (2001). Bone marrow cells regenerate
infarcted myocardium. Nature 410, 701-705.
Ortiz, L.A., Gambelli, F., McBride, C., Gaupp, D., Baddoo, M., Kaminski, N., and
Phinney, D.G. (2003). Mesenchymal stem cell engraftment in lung is enhanced in
response to bleomycin exposure and ameliorates its fibrotic effects. Proc Natl Acad Sci U
S A 100, 8407-8411.
Panteleyev, A.A., Jahoda, C.A., and Christiano, A.M. (2001). Hair follicle
predetermination. J Cell Sci 114, 3419-3431.
Park, H.W., Shin, J.S., and Kim, C.W. (2007). Proteome of mesenchymal stem cells.
Proteomics 7, 2881-2894.
Parrinello, S., Samper, E., Krtolica, A., Goldstein, J., Melov, S., and Campisi, J. (2003).
Oxygen sensitivity severely limits the replicative lifespan of murine fibroblasts. Nat Cell
Biol 5, 741-747.
Parsch, D., Fellenberg, J., Brummendorf, T.H., Eschlbeck, A.M., and Richter, W. (2004).
Telomere length and telomerase activity during expansion and differentiation of human
mesenchymal stem cells and chondrocytes. J Mol Med 82, 49-55.
230
Peh, G.S., Lang, R., Pera, M., and Hawes, S. (2008). CD133 expression by neural
progenitors derived from human embryonic stem cells and its use for their prospective
isolation. Stem Cells Dev.
Peister, A., Mellad, J.A., Larson, B.L., Hall, B.M., Gibson, L.F., and Prockop, D.J.
(2004). Adult stem cells from bone marrow (MSCs) isolated from different strains of
inbred mice vary in surface epitopes, rates of proliferation, and differentiation potential.
Blood 103, 1662-1668.
Pelengaris, S., Khan, M., and Evan, G. (2002). c-MYC: more than just a matter of life
and death. Nat Rev Cancer 2, 764-776.
Peters, C., Shapiro, E.G., and Krivit, W. (1998). Hurler syndrome: past, present, and
future. J Pediatr 133, 7-9.
Petersen, B.E., Grossbard, B., Hatch, H., Pi, L., Deng, J., and Scott, E.W. (2003). Mouse
A6-positive hepatic oval cells also express several hematopoietic stem cell markers.
Hepatology 37, 632-640.
Phinney, D.G., Hill, K., Michelson, C., DuTreil, M., Hughes, C., Humphries, S.,
Wilkinson, R., Baddoo, M., and Bayly, E. (2006). Biological activities encoded by the
murine mesenchymal stem cell transcriptome provide a basis for their developmental
potential and broad therapeutic efficacy. Stem Cells 24, 186-198.
Phinney, D.G., Kopen, G., Isaacson, R.L., and Prockop, D.J. (1999). Plastic adherent
stromal cells from the bone marrow of commonly used strains of inbred mice: variations
in yield, growth, and differentiation. J Cell Biochem 72, 570-585.
Pittenger, M.F., Mackay, A.M., Beck, S.C., Jaiswal, R.K., Douglas, R., Mosca, J.D.,
Moorman, M.A., Simonetti, D.W., Craig, S., and Marshak, D.R. (1999). Multilineage
potential of adult human mesenchymal stem cells. Science 284, 143-147.
Pochampally, R.R., Ylostalo, J., Penfornis, P., Matz, R.R., Smith, J.R., and Prockop, D.J.
(2007). Histamine receptor H1 and dermatopontin: new downstream targets of the
vitamin D receptor. J Bone Miner Res 22, 1338-1349.
Porkka, K.P., Helenius, M.A., and Visakorpi, T. (2002). Cloning and characterization of
a novel six-transmembrane protein STEAP2, expressed in normal and malignant prostate.
Lab Invest 82, 1573-1582.
Post, S., Abdallah, B.M., Bentzon, J.F., and Kassem, M. (2008). Demonstration of the
presence of independent pre-osteoblastic and pre-adipocytic cell populations in bone
marrow-derived mesenchymal stem cells. Bone 43, 32-39.
Prowse, K.R., and Greider, C.W. (1995). Developmental and tissue-specific regulation of
mouse telomerase and telomere length. Proc Natl Acad Sci U S A 92, 4818-4822.
231
Quarto, R., Thomas, D., and Liang, C.T. (1995). Bone progenitor cell deficits and the
age-associated decline in bone repair capacity. Calcif Tissue Int 56, 123-129.
Rahmani, M., Wong, B.W., Ang, L., Cheung, C.C., Carthy, J.M., Walinski, H., and
McManus, B.M. (2006). Versican: signaling to transcriptional control pathways. Can J
Physiol Pharmacol 84, 77-92.
Rawadi, G., Vayssiere, B., Dunn, F., Baron, R., and Roman-Roman, S. (2003). BMP-2
controls alkaline phosphatase expression and osteoblast mineralization by a Wnt
autocrine loop. J Bone Miner Res 18, 1842-1853.
Reyes, M., Lund, T., Lenvik, T., Aguiar, D., Koodie, L., and Verfaillie, C.M. (2001).
Purification and ex vivo expansion of postnatal human marrow mesodermal progenitor
cells. Blood 98, 2615-2625.
Richardson, G.D., Robson, C.N., Lang, S.H., Neal, D.E., Maitland, N.J., and Collins,
A.T. (2004). CD133, a novel marker for human prostatic epithelial stem cells. J Cell Sci
117, 3539-3545.
Rosova, I., Dao, M., Capoccia, B., Link, D., and Nolta, J.A. (2008). Hypoxic
preconditioning results in increased motility and improved therapeutic potential of human
mesenchymal stem cells. Stem Cells 26, 2173-2182.
Roubelakis, M.G., Pappa, K.I., Bitsika, V., Zagoura, D., Vlahou, A., Papadaki, H.A.,
Antsaklis, A., and Anagnou, N.P. (2007). Molecular and proteomic characterization of
human mesenchymal stem cells derived from amniotic fluid: comparison to bone marrow
mesenchymal stem cells. Stem Cells Dev 16, 931-952.
Rubio, D., Garcia-Castro, J., Martin, M.C., de la Fuente, R., Cigudosa, J.C., Lloyd, A.C.,
and Bernad, A. (2005). Spontaneous human adult stem cell transformation. Cancer Res
65, 3035-3039.
Sanchez-Ramos, J., Song, S., Cardozo-Pelaez, F., Hazzi, C., Stedeford, T., Willing, A.,
Freeman, T.B., Saporta, S., Janssen, W., Patel, N., et al. (2000). Adult bone marrow
stromal cells differentiate into neural cells in vitro. Exp Neurol 164, 247-256.
Sarugaser, R., Lickorish, D., Baksh, D., Hosseini, M.M., and Davies, J.E. (2005). Human
umbilical cord perivascular (HUCPV) cells: a source of mesenchymal progenitors. Stem
Cells 23, 220-229.
Sekiya, I., Larson, B.L., Smith, J.R., Pochampally, R., Cui, J.G., and Prockop, D.J.
(2002). 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, 530541.
232
Shi, S., and Gronthos, S. (2003). Perivascular niche of postnatal mesenchymal stem cells
in human bone marrow and dental pulp. J Bone Miner Res 18, 696-704.
Short, B., Brouard, N., Occhiodoro-Scott, T., Ramakrishnan, A., and Simmons, P.J.
(2003). Mesenchymal stem cells. Arch Med Res 34, 565-571.
Silva, W.A., Jr., Covas, D.T., Panepucci, R.A., Proto-Siqueira, R., Siufi, J.L., Zanette,
D.L., Santos, A.R., and Zago, M.A. (2003). The profile of gene expression of human
marrow mesenchymal stem cells. Stem Cells 21, 661-669.
Simmons, P.J., and Torok-Storb, B. (1991). Identification of stromal cell precursors in
human bone marrow by a novel monoclonal antibody, STRO-1. Blood 78, 55-62.
Singh, S.K., Hawkins, C., Clarke, I.D., Squire, J.A., Bayani, J., Hide, T., Henkelman,
R.M., Cusimano, M.D., and Dirks, P.B. (2004). Identification of human brain tumour
initiating cells. Nature 432, 396-401.
Song, L., Webb, N.E., Song, Y., and Tuan, R.S. (2006). Identification and functional
analysis of candidate genes regulating mesenchymal stem cell self-renewal and
multipotency. Stem Cells 24, 1707-1718.
Sotiropoulou, P.A., and Papamichail, M. (2007). Immune properties of mesenchymal
stem cells. Methods Mol Biol 407, 225-243.
Strakova, Z., Livak, M., Krezalek, M., and Ihnatovych, I. (2008). Multipotent properties
of myofibroblast cells derived from human placenta. Cell Tissue Res 332, 479-488.
Studer, L., Csete, M., Lee, S.H., Kabbani, N., Walikonis, J., Wold, B., and McKay, R.
(2000). Enhanced proliferation, survival, and dopaminergic differentiation of CNS
precursors in lowered oxygen. J Neurosci 20, 7377-7383.
Tae, S.K., Lee, S.H., Park, J.S., and Im, G.I. (2006). Mesenchymal stem cells for tissue
engineering and regenerative medicine. Biomed Mater 1, 63-71.
Taupin, P. (2006). OTI-010 Osiris Therapeutics/JCR Pharmaceuticals. Curr Opin Investig
Drugs 7, 473-481.
Terada, N., Hamazaki, T., Oka, M., Hoki, M., Mastalerz, D.M., Nakano, Y., Meyer,
E.M., Morel, L., Petersen, B.E., and Scott, E.W. (2002). Bone marrow cells adopt the
phenotype of other cells by spontaneous cell fusion. Nature 416, 542-545.
Todaro, G.J., and Green, H. (1963). Quantitative studies of the growth of mouse embryo
cells in culture and their development into established lines. J Cell Biol 17, 299-313.
233
Tolar, J., Nauta, A.J., Osborn, M.J., Panoskaltsis Mortari, A., McElmurry, R.T., Bell, S.,
Xia, L., Zhou, N., Riddle, M., Schroeder, T.M., et al. (2007). Sarcoma derived from
cultured mesenchymal stem cells. Stem Cells 25, 371-379.
Tontonoz, P., Hu, E., Graves, R.A., Budavari, A.I., and Spiegelman, B.M. (1994).
mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer. Genes Dev 8,
1224-1234.
Tremain, N., Korkko, J., Ibberson, D., Kopen, G.C., DiGirolamo, C., and Phinney, D.G.
(2001). MicroSAGE analysis of 2,353 expressed genes in a single cell-derived colony of
undifferentiated human mesenchymal stem cells reveals mRNAs of multiple cell
lineages. Stem Cells 19, 408-418.
Tropel, P., Noel, D., Platet, N., Legrand, P., Benabid, A.L., and Berger, F. (2004).
Isolation and characterisation of mesenchymal stem cells from adult mouse bone marrow.
Exp Cell Res 295, 395-406.
Tsonis, P.A. (2004). Stem cells from differentiated cells. Mol Interv 4, 81-83.
Tsonis, P.A., Sargent, M.T., Del Rio-Tsonis, K., and Jung, J.C. (1996). 9-cis retinoic acid
antagonizes the stimulatory effect of 1,25 dihydroxyvitamin D3 on chondrogenesis of
chick limb bud mesenchymal cells: interactions of their receptors. Int J Dev Biol 40,
1053-1059.
Tuan, R.S., Boland, G., and Tuli, R. (2003). Adult mesenchymal stem cells and cellbased tissue engineering. Arthritis Res Ther 5, 32-45.
Virtaneva, K.I., Emi, N., Marken, J.S., Aruffo, A., Jones, C., Spurr, N.K., and Schroder,
J.P. (1994). Chromosomal localization of three human genes coding for A15, L6, and
S5.7 (TAPA1): all members of the transmembrane 4 superfamily of proteins.
Immunogenetics 39, 329-334.
Wagers, A.J., Christensen, J.L., and Weissman, I.L. (2002). Cell fate determination from
stem cells. Gene Ther 9, 606-612.
Wagner, W., Feldmann, R.E., Jr., Seckinger, A., Maurer, M.H., Wein, F., Blake, J.,
Krause, U., Kalenka, A., Burgers, H.F., Saffrich, R., et al. (2006). The heterogeneity of
human mesenchymal stem cell preparations--evidence from simultaneous analysis of
proteomes and transcriptomes. Exp Hematol 34, 536-548.
Wagner, W., and Ho, A.D. (2007). Mesenchymal stem cell preparations--comparing
apples and oranges. Stem Cell Rev 3, 239-248.
Wagner, W., Roderburg, C., Wein, F., Diehlmann, A., Frankhauser, M., Schubert, R.,
Eckstein, V., and Ho, A.D. (2007). Molecular and secretory profiles of human
234
mesenchymal stromal cells and their abilities to maintain primitive hematopoietic
progenitors. Stem Cells 25, 2638-2647.
Wagner, W., Wein, F., Seckinger, A., Frankhauser, M., Wirkner, U., Krause, U., Blake,
J., Schwager, C., Eckstein, V., Ansorge, W., et al. (2005). Comparative characteristics of
mesenchymal stem cells from human bone marrow, adipose tissue, and umbilical cord
blood. Exp Hematol 33, 1402-1416.
Wakitani, S., Mitsuoka, T., Nakamura, N., Toritsuka, Y., Nakamura, Y., and Horibe, S.
(2004). Autologous bone marrow stromal cell transplantation for repair of full-thickness
articular cartilage defects in human patellae: two case reports. Cell Transplant 13, 595600.
Wakitani, S., Saito, T., and Caplan, A.I. (1995). Myogenic cells derived from rat bone
marrow mesenchymal stem cells exposed to 5-azacytidine. Muscle Nerve 18, 1417-1426.
Wang, G., Bunnell, B.A., Painter, R.G., Quiniones, B.C., Tom, S., Lanson, N.A., Jr.,
Spees, J.L., Bertucci, D., Peister, A., Weiss, D.J., et al. (2005a). Adult stem cells from
bone marrow stroma differentiate into airway epithelial cells: potential therapy for cystic
fibrosis. Proc Natl Acad Sci U S A 102, 186-191.
Wang, J., Xie, L.Y., Allan, S., Beach, D., and Hannon, G.J. (1998). Myc activates
telomerase. Genes Dev 12, 1769-1774.
Wang, Q.R., and Wolf, N.S. (1990). Dissecting the hematopoietic microenvironment.
VIII. Clonal isolation and identification of cell types in murine CFU-F colonies by
limiting dilution. Exp Hematol 18, 355-359.
Wang, Y., Huso, D.L., Harrington, J., Kellner, J., Jeong, D.K., Turney, J., and McNiece,
I.K. (2005b). Outgrowth of a transformed cell population derived from normal human
BM mesenchymal stem cell culture. Cytotherapy 7, 509-519.
Watt, F.M., and Hogan, B.L. (2000). Out of Eden: stem cells and their niches. Science
287, 1427-1430.
Watt, F.M., Lo Celso, C., and Silva-Vargas, V. (2006). Epidermal stem cells: an update.
Curr Opin Genet Dev 16, 518-524.
Weissman, I.L., Anderson, D.J., and Gage, F. (2001). Stem and progenitor cells: origins,
phenotypes, lineage commitments, and transdifferentiations. Annu Rev Cell Dev Biol 17,
387-403.
Welm, B.E., Tepera, S.B., Venezia, T., Graubert, T.A., Rosen, J.M., and Goodell, M.A.
(2002). Sca-1(pos) cells in the mouse mammary gland represent an enriched progenitor
cell population. Dev Biol 245, 42-56.
235
Wieczorek, G., Steinhoff, C., Schulz, R., Scheller, M., Vingron, M., Ropers, H.H., and
Nuber, U.A. (2003). Gene expression profile of mouse bone marrow stromal cells
determined by cDNA microarray analysis. Cell Tissue Res 311, 227-237.
Williams, J.T., Southerland, S.S., Souza, J., Calcutt, A.F., and Cartledge, R.G. (1999).
Cells isolated from adult human skeletal muscle capable of differentiating into multiple
mesodermal phenotypes. Am Surg 65, 22-26.
Wilson, A., Oser, G.M., Jaworski, M., Blanco-Bose, W.E., Laurenti, E., Adolphe, C.,
Essers, M.A., Macdonald, H.R., and Trumpp, A. (2007). Dormant and self-renewing
hematopoietic stem cells and their niches. Ann N Y Acad Sci 1106, 64-75.
Wognum, A.W., Eaves, A.C., and Thomas, T.E. (2003). Identification and isolation of
hematopoietic stem cells. Arch Med Res 34, 461-475.
Wright, G.J., Jones, M., Puklavec, M.J., Brown, M.H., and Barclay, A.N. (2001). The
unusual distribution of the neuronal/lymphoid cell surface CD200 (OX2) glycoprotein is
conserved in humans. Immunology 102, 173-179.
Wright, N.A. (2000). Epithelial stem cell repertoire in the gut: clues to the origin of cell
lineages, proliferative units and cancer. Int J Exp Pathol 81, 117-143.
Wu, Z., Rosen, E.D., Brun, R., Hauser, S., Adelmant, G., Troy, A.E., McKeon, C.,
Darlington, G.J., and Spiegelman, B.M. (1999). Cross-regulation of C/EBP alpha and
PPAR gamma controls the transcriptional pathway of adipogenesis and insulin
sensitivity. Mol Cell 3, 151-158.
Wynn, R.F., Hart, C.A., Corradi-Perini, C., O'Neill, L., Evans, C.A., Wraith, J.E.,
Fairbairn, L.J., and Bellantuono, I. (2004). A small proportion of mesenchymal stem cells
strongly expresses functionally active CXCR4 receptor capable of promoting migration
to bone marrow. Blood 104, 2643-2645.
Xin, L., Lawson, D.A., and Witte, O.N. (2005). The Sca-1 cell surface marker enriches
for a prostate-regenerating cell subpopulation that can initiate prostate tumorigenesis.
Proc Natl Acad Sci U S A 102, 6942-6947.
Xu, Y., Malladi, P., Chiou, M., Bekerman, E., Giaccia, A.J., and Longaker, M.T. (2007).
In vitro expansion of adipose-derived adult stromal cells in hypoxia enhances early
chondrogenesis. Tissue Eng 13, 2981-2993.
Yang, D., Holt, G.E., Velders, M.P., Kwon, E.D., and Kast, W.M. (2001). Murine sixtransmembrane epithelial antigen of the prostate, prostate stem cell antigen, and prostatespecific membrane antigen: prostate-specific cell-surface antigens highly expressed in
prostate cancer of transgenic adenocarcinoma mouse prostate mice. Cancer Res 61, 58575860.
236
Ying, Q.L., Nichols, J., Evans, E.P., and Smith, A.G. (2002). Changing potency by
spontaneous fusion. Nature 416, 545-548.
Young, R.G., Butler, D.L., Weber, W., Caplan, A.I., Gordon, S.L., and Fink, D.J. (1998).
Use of mesenchymal stem cells in a collagen matrix for Achilles tendon repair. J Orthop
Res 16, 406-413.
Yu, J., Vodyanik, MA., Smuga-Otto, K., Antosiewicz-Bourget, J., Frane, JL., Tian, S.,
Nie, J., Jonsdottir, GA., Ruotti, V., Stewart, R., Slukvin, II., Thomson, JA. (2007).
Induced Pluripotent stem cell lines derived from human somatic stem cells. Science 318,
1917-20.
Zhan, M. (2008). Genomic studies to explore self-renewal and differentiation properties
of embryonic stem cells. Front Biosci 13, 276-283.
Zhou, Y.F., Bosch-Marce, M., Okuyama, H., Krishnamachary, B., Kimura, H., Zhang, L.,
Huso, D.L., and Semenza, G.L. (2006). Spontaneous transformation of cultured mouse
bone marrow-derived stromal cells. Cancer Res 66, 10849-10854.
Zwartouw, H.T., and Westwood, J.C. (1958). Factors affecting growth and glycolysis in
tissue culture. Br J Exp Pathol 39, 529-539.
237

PDF Document reader online

This website is focused on providing document in readable format, online without need to install any type of software on your computer. If you are using thin client, or are not allowed to install document reader of particular type, this application may come in hand for you. Simply upload your document, and Docureader.top will transform it into readable format in a few seconds. Why choose Docureader.top?

  1. Unlimited sharing - you can upload document of any size. If we are able to convert it into readable format, you have it here - saved for later or immediate reading
  2. Cross-platform - no compromised when reading your document. We support most of modern browers without the need of installing any of external plugins. If your device can oper a browser - then you can read any document on it
  3. Simple uploading - no need to register. Just enter your email, title of document and select the file, we do the rest. Once the document is ready for you, you will receive automatic email from us.

Previous 10

Next 10