DIFFERENTIATION OF HUMAN DERMAL FIBROBLASTS AND...
136
LINKÖPING UNIVERSITY MEDICAL DISSERTATIONS NO. 1202 DIFFERENTIATION OF HUMAN DERMAL FIBROBLASTS AND APPLICATIONS IN TISSUE ENGINEERING Pehr Sommar Division of Surgery Department of Clinical and Experimental Medicine Faculty of Health Sciences SE-581 85 Linköping, Sweden 2010
Transcript of DIFFERENTIATION OF HUMAN DERMAL FIBROBLASTS AND...
LINKÖPING UNIVERSITY MEDICAL DISSERTATIONS
NO. 1202
DIFFERENTIATION OF HUMAN
DERMAL FIBROBLASTS AND APPLICATIONS
IN TISSUE ENGINEERING
Pehr Sommar
Division of Surgery
Department of Clinical and Experimental Medicine
Faculty of Health Sciences
SE-581 85 Linköping, Sweden
2010
Published papers are reprinted with permission from the publisher.
Printed by LiU-Tryck, Linköping, Sweden, 2010
© Pehr Sommar
Cover by Pehr Sommar and Johan PE Junker
Photos by Pehr Sommar, Figure 5; Sofia Pettersson
Illustrations by Pehr Sommar and Kalle Lundgren
ISBN 978-91-7393-326-1
ISSN 0345-0082
To Marie and Ingrid
SUPERVISOR
Gunnar Kratz, Professor
Laboratory for Reconstructive Plastic Surgery
Department of Plastic Surgery, Hand Surgery and Burns
Department of Clinical and Experimental Medicine
Linköping University, Sweden
CO-SUPERVISOR
Thomas Hansson, MD, PhD
Department of Plastic Surgery, Hand Surgery and Burns
University Hospital, Linköping, Sweden
OPPONENT
Anders Lindahl, Professor
Institute of Biomedicine
Department of Clinical Chemistry and Transfusion Medicine
Sahlgrenska University Hospital, Gothenburg, Sweden
COMMITTEE BOARD
Sven Hammarström, Professor
Division of Cell Biology
Department of Clinical and Experimental Medicine
Linköping University, Sweden
Per Wretenberg, MD, PhD
Department of Molecular Medicine and Surgery
Section of Orthopaedics and Sports Medicine
Karolinska Institute
Karolinska University Hospital, Stockholm, Sweden
Avni Abdiu, MD, PhD
Department of Plastic Surgery, Hand Surgery and Burns
University Hospital, Linköping, Sweden
ABSTRACT
Tissue engineering applies principles of biology and engineering to the development of
functional substitutes for damaged or lost tissues. Tools for the neo-generation of tissue in
tissue engineering research include cells, biomaterials and soluble factors.
One main obstacle in tissue engineering is the limited availability of autologous tissue
specific progenitor cells. This has led to interest into using autologous cells with stem cell
plasticity. Bone marrow derived stem cells were the first adult stem cells shown to have
multilineage potential. Since, several reports have been published indicating that cells from
other tissues; fat, muscle, connective tissue e.g., possess potential to differentiate into lineages
distinct from their tissue of origin.
The optimal cell type for use in tissue engineering applications should be easy to obtain,
cultivate and store. The human dermal fibroblast is an easily accessible cell source, which
after routine cell expansion gives a substantial cell yield from a small skin biopsy. Hence, the
dermal fibroblast could be a suitable cell source for tissue engineering applications.
The main aim of this thesis was to investigate the differentiation capacity of human
dermal fibroblasts, and their possible applications in bone and cartilage tissue engineering
applications.
Human dermal fibroblasts were shown to differentiate towards adipogenic, chondrogenic,
and osteogenic phenotypes upon subjection to specific induction media. Differentiation was
seen both in unrefined primary cultures and in clonal populations (paper I). Fibroblasts could
be used to create three-dimensional cartilage- and bone like tissue when grown in vitro on
gelatin microcarriers in combination with platelet rich plasma (paper II). 4 weeks after in vivo
implantation of osteogenic induced fibroblasts into a fracture model in athymic rats, dense
cell clusters and viable human cells were found in the gaps, but no visible healing of defects
as determined by CT-scanning (paper III). After the induction towards adipogenic,
chondrogenic, endotheliogenic and osteogenic lineages, gene expression analysis by
microarray and quantitative real-time-PCR found several master regulatory genes important
for lineage commitment, as well as phenotypically relevant genes regulated as compared to
reference cultures (paper IV).
In conclusion, results obtained in this thesis suggest an inherent ability for controllable
phenotype alteration of human dermal fibroblasts in vitro. We conclude that dermal
fibroblasts could be induced towards adipogenic, chondrogenic, endotheliogenic or osteogenic
novel phenotypes, which suggest a genetic readiness of differentiated fibroblasts for lineage-
specific biological functionality, indicating that human dermal fibroblasts might be a suitable
cell source in tissue engineering applications.
TABLE OF CONTENTS
INTRODUCTION ................................................................................................................................................. 1
TISSUE ENGINEERING................................................................................................................................... 1
CELLS IN TISSUE ENGINEERING ................................................................................................................. 1
Stem cells and cellular differentiation ........................................................................................................... 2
Embryonic stem cells (ESCs) ......................................................................................................................... 3
Adult stem cells (ASCs).................................................................................................................................. 4
Induced pluripotent stem cells (iPSCs) .......................................................................................................... 5
Dermal fibroblasts ......................................................................................................................................... 6
BIOMATERIALS .............................................................................................................................................. 7
SOLUBLE FACTORS ....................................................................................................................................... 8
TISSUE ENGINEERING APPLICATIONS ..................................................................................................... 8
Tissue engineering of cartilage ..................................................................................................................... 8
Tissue engineering of bone ............................................................................................................................ 9
Clinical applications of tissue engineering ................................................................................................. 10
ADIPOSE TISSUE ........................................................................................................................................... 11
Histological and physiological features of adipose tissue ........................................................................... 11
Regulators of adipogenesis .......................................................................................................................... 11
Adipogenic induction factors ....................................................................................................................... 12
CARTILAGE TISSUE ..................................................................................................................................... 13
Histological and molecular features of cartilage ........................................................................................ 13
Regulators of chondrogenesis ..................................................................................................................... 15
Chondrogenic induction factors .................................................................................................................. 16
BONE TISSUE ................................................................................................................................................. 17
Histological and physiological features of bone ......................................................................................... 17
Regulators of osteogenesis .......................................................................................................................... 18
Osteogenic induction factors ....................................................................................................................... 19
ENDOTHELIAL CELLS ................................................................................................................................. 20
Histological and physiological features of the endothelium ........................................................................ 20
Regulators of endotheliogenesis .................................................................................................................. 21
Endothelial induction factors ...................................................................................................................... 22
AIMS .................................................................................................................................................................... 23
MATERIALS AND METHODS ....................................................................................................................... 25
ETHICAL PERMISSION ................................................................................................................................ 25
CELL HARVEST AND INDUCTION ............................................................................................................ 25
Fibroblasts ................................................................................................................................................... 25
Preadipocytes .............................................................................................................................................. 25
Chondrocytes ............................................................................................................................................... 25
Osteoblasts .................................................................................................................................................. 26
Endothelial cells .......................................................................................................................................... 26
Single cell cloning ....................................................................................................................................... 26
Induction of differentiation .......................................................................................................................... 26
CONFIRMATION OF DIFFERENTIATION.................................................................................................. 27
Routine staining ........................................................................................................................................... 27
Immunohistochemistry ................................................................................................................................. 28
Alkaline phosphatase assay ......................................................................................................................... 28
Fluorescence in situ hybridization............................................................................................................... 29
Full expression microarray ......................................................................................................................... 29
Quantative real-time polymerase chain reaction (qRT-PCR) ..................................................................... 30
Flow cytometry ............................................................................................................................................ 30
MICROCARRIERS ......................................................................................................................................... 31
SPINNER FLASK AND THREE-DIMENSIONAL CULTURE ..................................................................... 32
FEMORAL GAP SURGICAL MODEL .......................................................................................................... 33
RESULTS AND DISCUSSION ......................................................................................................................... 35
PAPER I ........................................................................................................................................................... 35
Conclusion ................................................................................................................................................... 37
PAPER II .......................................................................................................................................................... 37
Conclusion ................................................................................................................................................... 40
PAPER III......................................................................................................................................................... 41
Conclusion ................................................................................................................................................... 42
PAPER IV ........................................................................................................................................................ 42
Conclusion ................................................................................................................................................... 45
GENERAL DISCUSSION ............................................................................................................................... 45
FUTURE PERSPECTIVES ............................................................................................................................. 47
CONCLUSIONS ................................................................................................................................................. 49
ACKNOWLEDGEMENTS ................................................................................................................................ 51
REFERENCES .................................................................................................................................................... 53
SELECTED ABBREVIATIONS
α-MEM Alpha minimum essential medium
A2P Ascorbate-2-phosphate
A-FBs Adipogenic induced fibroblasts
ALP Alkaline phosphatase
AM Adipogenic induction medium
ASCs Adult stem cells
BGP β- glycerophosphate
BMP Bone morphogenetic protein
CCs Chondrocytes
CD Cluster of differentiation
C-FBs Chondrogenic induced fibroblasts
CM Chondrogenic induction medium
CPM Chondrocyte proliferation medium
DAPI 4', 6-diamidino-2-phenylindole
DEX Dexomethasone
DMEM Dulbecco´s modified Eagle´s medium
E-FBs Endotheliogenic induced fibroblasts
ECM Extracellular matrix
EM Endotheliogenic induction medium
EPM Endothelial proliferation medium
ESCs Embryonic stem cells
FBs Fibroblasts
FCS Fetal calf serum
FGF Fibroblast growth factor
FITC Fluorescein isothiocyanate
FISH Fluorescence in situ hybridization
FM Fibroblast proliferation medium
HUVECs Human umbilical vein endothelial cells
IBMX Isobutylmethylxantine
IGF Insulin-like growth factor
IHC Immunohistochemistry
iPSCs Induced pluripotent stem cells
MSCs Mesenchymal stem cells
NaCl Sodium chloride
OBs Osteoblasts
O-FBs Osteogenic induced fibroblasts
OM Osteogenic induction medium
OPM Osteoblast proliferation medium
PAs Preadipocytes
PBS Phosphate-buffered saline
PEST Penicillin and streptomycin
PM Preadipocyte proliferation medium
PNPP p-Nitrophenyl phosphate
sccFBs Single-cell cloned fibroblasts
TGF-β1 Transforming growth factor beta, type 1
LIST OF PAPERS
This thesis is based on the following papers, which are referred to in the text by their
Roman numerals.
I.
Johan PE Junker, Pehr Sommar, Mårten Skog, Hans Johnson, Gunnar Kratz
Adipogenic, Chondrogenic and Osteogenic Differentiation of Clonally Derived Human
Dermal Fibroblasts
Cells Tissues Organs. 2010;191(2):105-18
II.
Pehr Sommar, Sofia Pettersson, Charlotte Ness, Hans Johnson, Gunnar Kratz, Johan PE
Junker
Engineering Three-dimensional Cartilage- and Bone-like Tissues using Human Dermal
Fibroblasts and Macroporous Gelatine Microcarriers
Journal of Plastic, Reconstructive and Aesthetic Surgery. 2010 Jun;63(6):1036-46
III.
Pehr Sommar, Johan PE Junker, Eivind Strandenes, Charlotte Ness, Thomas Hansson, Hans
Johnson, Gunnar Kratz
In vivo Implantation of Osteogenic Induced Human Dermal Fibroblasts in a Fracture Model
Manuscript
IV.
Jonathan Rakar, Pehr Sommar, Susanna Lönnqvist, Hans Johnson, Johan PE Junker, Gunnar
Kratz
Evaluating Multi-Lineage Induction of Human Dermal Fibroblasts Using Gene Expression
Analysis
Manuscript
1
INTRODUCTION
TISSUE ENGINEERING
In 1993, tissue engineering was defined by Langer and Vacanti as an “interdisciplinary
field that applies the principles of engineering and the life sciences toward the development of
biological substitutes that restore, maintain or improve tissue function” (Langer and Vacanti
1993). Briefly the term tissue engineering describes the production of tissue replacement
parts. It is closely related to the field of regenerative medicine which is a generic term for
clinical therapies that may involve the use of stem cells, induction of regeneration by
biologically active molecules or in vivo transplantation of engineered tissue. Tissue
engineering contains elements of medicine, material science and engineering. The major
components used for in vitro engineering of tissue include cells, biomaterials and soluble
factors. In many cases all three components are required for the development of clinical
treatments. This thesis will focus on human dermal fibroblasts (FBs) as a possible cell source
for use in tissue engineering applications and in vitro differentiation of these cells using
soluble factors. In paper II-III cells are combined with biomaterials.
CELLS IN TISSUE ENGINEERING
The availability of donor cells is a limiting factor in the treatment of many patients.
Furthermore, patients receiving allogenous cells require treatment with immunosuppressive
drugs. The problem of immunosupression can be avoided by using autologous cells for
transplantation. Cells can, after harvest, be expanded in vitro prior to engraftment to generate
sufficient cell numbers. Examples of early cellular therapies is the use of in vitro grown
keratinocytes for the treatment of large burns (Gallico, O'Connor et al. 1984), use of
melanocytes for the treatment of vitiligo (Olsson and Juhlin 1992), or the use of in vitro
expanded chondrocytes for repair of articular cartilage (Brittberg, Lindahl et al. 1994).
Still current cellular therapies are not yet advanced enough to replace complex tissues or
entire organs. The majority of experimental strategies include only one cell type and a
scaffold, but some include several cell types as seen in vascular engineering; endothelium,
muscle cells, FBs (Iwasaki, Kojima et al. 2008), or keratinocytes, FBs and adipocytes in the
case of skin engineering (Auxenfans, Fradette et al. 2009). The possibility of building whole
organs is limited by the structural complexity and the need for vascularization after in vivo
implantation. Some of the new applications are discussed below in “tissue engineering
applications”.
In some clinical situations there is a lack of donor cells or difficulties in culturing donor
cells in vitro. To overcome these problems the use of stem cells for tissue engineering
applications has been proposed. The possible use of stem cells would in a radical way
facilitate neo-generation of tissues (Bajada, Mazakova et al. 2008).
2
Stem cells and cellular differentiation
In 1961, Hayflick presented results on the limited lifespan of cells, known as Hayflick´s
limit. He showed that normal human FBs in cell culture could divide between 40 and 60 times
before entering a senescence phase (Hayflick and Moorhead 1961). This phenomenon was
later explained by the shortening of telomeres in each division (Sharpless and DePinho 2004).
However, certain cells continue dividing beyond this limit, because of altered transcriptional
activity (Verfaillie, Pera et al. 2002; Boheler 2009).
Stem cells are defined by their capacity of self-renewal, and the capacity to differentiate
into different cell lineages under appropriate conditions. Stem cells divide asymmetrically;
into a more differentiated daughter cell and into a clone to replenish and maintain the stem
cell population. A third criterion is that stem cells functionally repopulate the tissue of origin
when transplanted in a damaged recipient. A final, less well
established criterion is that stem
cells contribute differentiated
progeny in vivo even in the absence of tissue damage
(Verfaillie, Pera et al. 2002). Human stem cells can be categorized into three main types:
embryonic stem cells (ESCs), adult stem cells (ASCs) and induced pluripotent stem cells
(iPSCs). A stem cell is totipotent if it is capable of giving rise to a whole animal, including
germ cells, all cell types of the three germ layers (i.e., ectoderm, mesoderm and endoderm)
and extraembryonic tissues. Zygotes are totipotent. A pluripotent cell can produce all cell
types of the germ layers but not extraembryonic tissues, while a multipotent cell can produce
cells only of the same germ layer. ESCs are generally described as being pluripotent, not
totipotent, reflecting the fact that no animal has been generated by ESCs alone. Adult stem
cells are classically described as multipotent (Verfaillie, Pera et al. 2002).
A single cell, the fertilized egg, gives rise to hundreds of different cell types during
development of the organism. This generation of cellular diversity is termed differentiation.
The traditional view has been that adult cells were terminally differentiated, but during the
last 50 years this notion has been challenged. It has been long known that salamanders and
newts have an extraordinary regenerative capability. These animals can regenerate amputated
limbs, tail, removed retina and lens, heart, spinal cord, brain - virtually any tissue as long as
the animal is kept alive. What is striking is that this regeneration occurs in adult amphibians
by usage of already existing terminally differentiated cells, rather than undifferentiated stem
cells (Tsonis 2000). The inherent ability of adult cells to differentiate was elegantly showed
by Gurdon et al. in 1958 where a cloned frog was created by applying nuclear transfer of frog
FBs into an enucleated oocyte (Gurdon, Elsdale et al. 1958). By performing this cloning they
showed that all information necessary for the differentiation process is withheld in the adult
cell, only demanding the right environmental conditions. In higher mammals, it was thought
that the differentiation process was irreversible until the successful cloning of the sheep Dolly
(Wilmut, Schnieke et al. 1997). These cloning experiments demonstrated that somatic cells
can be reprogrammed back to the totipotent zygotic state by the cellular factors present in
unfertilized eggs and used to generate a mammalian offspring. This could later also be
replicated by cloning somatic cells with ESCs indicating that unfertilized eggs and ESCs
contain factors that can confer totipotency or pluripotency to somatic cells (Tada, Takahama
et al. 2001). A few defined transcription factors have in the last years been detected, and akin
3
to the case of the salamander cells, these factors have made it possible to dedifferentiate
human cells into stem cells (iPSCs) (Takahashi, Tanabe et al. 2007).
Embryonic stem cells (ESCs)
The first ESCs were isolated from the inner cell mass of mouse blastocysts in 1981 by
Evans and Kaufman (Evans and Kaufman 1981) and Martin (Martin 1981) (figure 1). These
cells had the potential to form teratomas when injected subcutaneously in mice. A few years
later, Evans and colleagues showed that they could produce live mice by injecting cultured
ESCs into a developing embryo. ESCs also offered an opportunity to generate live animals
with a desired mutation in every cell, so called knock-out animals, which has had a
revolutionary effect on genetic studies (Thomas and Capecchi 1990).
In 1998 human ESCs (hESCs) were isolated successfully by Thomson et al. creating a
new era of research and hope that stem cell technology may eventually benefit human disease
therapy (Thomson, Itskovitz-Eldor et al. 1998). Initially hESCs were thought to be the best
candidates for use in the tissue engineering field owing to their pluripotency. However, ESCs
are not autologous, hence the risk of host vs. graft rejections. There is also a risk of teratoma
formation, i.e. transplanted cells leading to uncontrollable growth and tumor formation.
Research was also hindered from the start by the ethical debate surrounding their use and the
complex legislation in place to protect their misuse (Frankel 2000).
Figure 1. Early embryonal development. The fertilized egg (zygote) divides and forms the blastocyst. When a
blastocyst implants in the uterus, the ICM eventually develops into a fetus and the surrounding trophoblast
develops into placenta. During gastrulation, the ICM develops into two distinct cell layers, the epiblast and the
hypoblast. The hypoblast forms the yolk sac, while the epiblast differentiates into the germ layers of the embryo;
ectoderm, mesoderm and endoderm.
Pluripotent ESCs for laboratory use are derived from the inner cell mass (ICM) of 5-6 day old blastocysts.
4
Common methods for generating ESCs involves either chemical or enzymatic removal of
the zona pellucida followed by isolation of inner cell mass by immunosurgery of the
blastocyst. In immunosurgery, cells of the trophoblast are destroyed by exposure to antibodies
followed by exposure to complement activity (Solter and Knowles 1975).
ESCs were earlier co-cultured with murine embryonic FBs as feeder-cells. Subsequent
studies have identified multiple factors secreted by the FBs including leukemia inhibitory
factor (LIF), fibroblast growth factors (FGFs), transforming growth factor β (TGF-β), activin,
wingless-type MMTV integration site family members (WNTs), insulin-like growth factor
(IGF), and antagonists of bone morphogenetic protein (BMP) signaling (Williams, Hilton et
al. 1988; Beattie, Lopez et al. 2005; Prowse, McQuade et al. 2007). The removal of feeder-
cells or cytokines leads to spontaneous differentiation and the loss of pluripotency of the
ESCs. The undifferentiated state of ESCs is maintained by the action of transcription factors.
Master regulators of self renewal are octamer binding protein 4 (OCT4), the SRY- related
HMG-box gene 2 (SOX2) and Nanog (Li 2010).
Adult stem cells (ASCs)
As the fertilized egg divides, three germ layers (ectoderm mesoderm and endoderm) are
established, which will ultimately give rise to all types of somatic cells. The ectoderm will
become neural tissues and epidermis. The mesoderm will develop into blood, mesenchyme,
muscle, and notochord. The endoderm will form the respiratory and digestive tracts.
The adult human body maintains an endogenous system of regeneration and repair
through stem cells, which can be found in almost every type of tissue. These non-embryonic
stem cells are called adult stem cells as they are present in the body throughout its lifetime
(Young and Black 2004; Li and Cao 2006). The use of adult stem cells rather than of ESCs in
clinical applications would have several advantages. The autologous source of adult stem cells
prevents host vs. graft rejection, and the former cells are not surrounded with the same ethical
considerations as ESCs. However, ASCs are thought to be multipotent, displaying limited
differentiation potential compared to ESCs (Bajada, Mazakova et al. 2008).
Bone marrow derived stem cells were the first adult stem cells shown to have
multilineage potential. These cells are harvested by bone marrow aspiration, usually from the
posterior iliac crest (Friedenstein, Petrakova et al. 1968; Jaiswal, Haynesworth et al. 1997;
Yoo, Barthel et al. 1998). Bone marrow contains at least two distinct stem cell populations;
hematopoietic stem cells giving rise to all blood cell types, and mesenchymal stem cells
(MSCs). MSCs can differentiate into tissues of the embryos mesoderm including bone,
adipose, cartilage and muscle (Caplan 1991; Prockop 1997; Delorme, Chateauvieux et al.
2006). MSCs have been evaluated in several bone and cartilage tissue engineering
applications (Caplan 2005). Drawbacks with the use of MSCs as a stem cell source is the
invasive and painful harvest and difficulties in culture and expansion (Bruder, Jaiswal et al.
1997). The yield of MSCs from harvested bone marrow is only 0.01 – 0.001 % (Pittenger,
Mackay et al. 1999; Kastrinaki, Sidiropoulos et al. 2008).
5
Figure 2. Differentiation capacity of adult stem cells of mesodermal origin. ASCs of mesodermal origin are
able to differentiate into a range of mesenchymal tissues and cell types, including bone, cartilage, mucle, stroma,
adipose. Evidence has also suggested that ASCs have a greater plasticity, and ability to differentiate into non-
mesencymal tissues including liver and neural tissues.
The optimal stem cell for use in tissue engineering applications should preferably be easy
to obtain, cultivate and store, which does not apply to MSCs. This has led to a search for
alternative cell sources, which has included investigations of the multilineage potential of
normal somatic human cells. It has been suggested that cells with similar plasticity as MSCs
can be isolated from a variety of adult tissues (Young, Mancini et al. 1995), including adipose
tissue (Zuk, Zhu et al. 2001), skeletal muscle (Wada, Inagawa-Ogashiwa et al. 2002) and
various dermal tissue components (Jahoda, Whitehouse et al. 2003; Bartsch, Yoo et al. 2005;
Junker, Sommar et al. 2010). These cell sources have the potential to differentiate into
mesenchymal lineages (figure 2), but during the last years reports of pluripotency of ASCs
have been published; neural cells differentiating into muscle cells (Galli, Borello et al. 2000),
muscle cells to neuronal cells (Romero-Ramos, Vourc'h et al. 2002), dermal FBs to neuronal
cells (Toma, Akhavan et al. 2001) or hepatocytes (Lysy, Smets et al. 2007), and bone marrow
to neuronal cells (Zhao, Duan et al. 2002).
Induced pluripotent stem cells (iPSCs)
The work by Gurdon et al. with nuclear transfer showed that environmental conditions
could alter the cellular phenotype (Gurdon, Elsdale et al. 1958). The phenotype is defined by
the particular pattern of regulated gene expression. Transcriptional programmes associated
with differentiation of specific cell types are controlled primarily by specific transcription
factors. These are proteins with different modes of action that operate at various levels to
facilitate and control the production of RNA (Lander, Linton et al. 2001). Several factors have
6
been identified that regulate cellular differentiation. By analyzing genes highly expressed in
ESCs Takahashi et al. delivered a pool of 24 genes coding for transcription factors which was
used to reprogram adult FBs via retroviral transduction. Eventually, only four transcription
factors; OCT4, SOX2, Kruppel-like factor 4 (KLF4) and cellular myelocytomatosis oncogene
(c-MYC) were deemed sufficient to reprogram FBs into ESC-like cells (Takahashi and
Yamanaka 2006). These ESC-like cells were coined as induced pluripotent stem cells (iPSCs)
to differentiate them from blastocyst-derived pluripotent ESCs. These results were later
repeated in human adult FBs (Takahashi, Tanabe et al. 2007; Lowry, Richter et al. 2008).
iPSCs constitutes a new autologous source of pluripotent stem cells and has already
gained lot of interest for possible clinical applications (Kiskinis and Eggan 2010). Drawbacks
of iPSCs include the use of a retroviral transduction to incorporate genes into the adult FBs.
There is also a very low frequency of iPSC derivation, about 1 in 5000 cells. The safety of
applying transduced cells in the treatment of humans is unknown. Both KLF4 and c-MYC are
well known oncogenes and transfection of such genes raise concerns that induction of tumor
formation could occur in clinical applications. Therefore research is now focusing on
construction of iPSCs without using retroviral vectors and oncogenes, though lower efficiency
has been reported without viral vectors (Sun, Longaker et al. 2010).
Dermal fibroblasts
Dermal FBs were among the first cell types grown in vitro, and have been considered
slightly mundane. The general opinion was that FBs represented a cell population only adding
extracellular matrix to the stroma. Lately however and partly owing to the success of
Takahashi and Yamanaka (Takahashi and Yamanaka 2006) this cell type has regained
interest, and is now considered as a viable alternative for tissue engineering applications.
The FB population is not as homogenous as previously thought. Sorrel and Caplan
describe differences between papillary FBs, which reside in the superficial dermis, and
reticular FBs, which reside in the deeper dermis, highlighting the difference in matrix
molecule production of these two distinct cell populations (Sorrell and Caplan 2004). Two
major populations of FBs have been described in the dermis, termed “mitotically active”, i.e.
replicative progenitor FBs and irreversible postmitotic FBs (Rodemann, Bayreuther et al.
1989). Furthermore, it has been shown that FBs in different parts of the body have different
properties; i.e. oral mucosa-derived FBs proliferate more rapidly and show a higher capacity
for cell doublings compared to dermal FBs. A functional correlation was linked to the fact
that cultured oral FBs secrete more hepatocyte growth factor and keratinocyte growth factor
than skin FBs which may contribute to a “fetal” wound healing phenotype of oral FBs
resulting in less scaring (Gron, Stoltze et al. 2002).
During the last years, there have been reports of adult stem cell populations isolated from
connective tissue from several parts of the body (Young, Mancini et al. 1995). One of the first
reports of pluripotency of dermal FBs was presented by Toma et al. 2001, where stem cell
populations were isolated from the dermis of mice and differentiated into neurons, glia,
smooth muscle cells and adipocytes (Toma, Akhavan et al. 2001). In 2005, this was also
shown in humans (Toma, McKenzie et al. 2005). Several reports followed confirming the
dermal FB as a plausible stem cell source (Bartsch, Yoo et al. 2005; Lorenz, Sicker et al.
7
2008). Other authors have investigated the differentiation of human dermal FBs into cartilage
(Mizuno and Glowacki 1996; Mizuno and Glowacki 1996; French, Rose et al. 2004). The
differentiation potential of FBs is generally mesodermal, but as noticed above there are also
reports of transdifferentiation i.e. neural tissue and hepatic tissue (Toma, Akhavan et al. 2001;
Chen, Zhang et al. 2007; Lysy, Smets et al. 2007). Fernandes et al. suggest that the stem cell
populations isolated from dermis with neural potential are neural crest derived stem cells
(Fernandes, McKenzie et al. 2004). These neural crest-like precursors could also be found in
the hair follicle. Other groups have had special interest in the hair follicle, where stem cell
populations have been found (Lako, Armstrong et al. 2002; Jahoda, Whitehouse et al. 2003).
Irrespective of the origin of stem cells from dermal tissue, the usage of such cells for cell-
based therapies may facilitate the in vitro production of autologous tissues. The human dermal
FB is an easily accessible autologous cell source, where substantial cell yields can be obtained
from a relatively small skin biopsy using minimally invasive procedures and routine cell
expansion techniques. The yield using this cell source is considerable higher as compared to
MSCs or iPSCs. There are also limited ethical considerations surrounding their use as
compared to the case of ESCs. Consequently, human dermal FBs fulfill several criteria for an
optimal cell source for tissue engineering applications.
BIOMATERIALS
Three-dimensional biomaterials serve as space holders preventing encroachment of
surrounding tissues into the graft site. They provide surfaces that facilitate the attachment,
survival, migration, proliferation and differentiation of cells. They also provide a void volume
in which vascularization, new tissue and remodeling can occur. In addition biomaterials
provide a vehicle for delivery of cells to the body part of interest (Muschler, Nakamoto et al.
2004).
There are practically speaking as many different biomaterials as there are tissue
engineering-interested research groups. They can be designed according to the specifications
of the desired tissue to reconstruct, and either used with seeded cells for transplantation or
acellular as a scaffold for the native tissue to grow into. Biomaterials are either organic or
synthetic and can be biodegradable or permanent. Organic biomaterials consist of
extracellular matrix molecules such as collagen, gelatin, fibrin/fibronectin, hyaluronan and
chondroitin sulphate, this group also includes alginate and agarose from algae and chitosan
from arthropods. These are all degradable (Chung and Burdick 2008; Badylak, Freytes et al.
2009). There are also degradable synthetic biomaterials such as polylactic acid (PLA),
polyglycolic acid (PGA), and copolymers like polylactic-co-glycolic acid (PLGA) or
polycaprolactone (PCL) (Gunatillake and Adhikari 2003). Hydroxyapatite and other calcium
phosphates are considered degradable, but have a long degradation time and are usually only
partially degraded and integrated by natural tissues (Coutu, Yousefi et al. 2009). A degradable
scaffold will gradually be replaced by new tissue, generated by the delivered or native cells.
One downside with degradable biomaterials is that they might be too soft to offer necessary
initial support in tissues with mechanical strain.
8
Among permanent biomaterials are metals, glass, and ceramics. These materials will not
be replaced by host tissue over time and therefore have a risk to eventually brake or become
rejected (Weigel, Schinkel et al. 2006; Coutu, Yousefi et al. 2009).
SOLUBLE FACTORS
Soluble factors for tissue engineering applications are chosen either based on the
predicted native environment of the cultured cell type, or based on the effect of the factor on
stimulating a desired differentiation. Soluble factors can be hormones like dexomethasone or
insulin, growth factors like TGF-β, IGF, BMPs, or chemical substances like vitamin D,
indomethacin or ascorbic acid. Soluble factors can be used during cell cultivation by applying
them in induction media. The soluble factors used in this thesis will be discussed more in
detail below.
Soluable factors can also be included in biomaterials and be released to target cells
included in the scaffold or in situ during normal degradation. BMPs are for instance
commonly included in biomaterials in bone tissue engineering applications (Axelrad, Kakar et
al. 2007; Docherty-Skogh, Bergman et al. 2010).
The differentiation potential of cells can also be genetically modified with a variety of
means that either transiently or permanently alters the gene expression of cells, usually using
transfection by viral vectors (Young, Searle et al. 2006). This is a very effective way of
controlling cellular gene expression, but a questionable alternative for clinical applications.
TISSUE ENGINEERING APPLICATIONS
Tissue engineering of cartilage
Cartilage lacks intrinsic capacity to repair itself after trauma or degenerative diseases.
Cartilage defects are commonly treated either by microfracture technique to receive
progenitor cells from blood and bone marrow, osteochondral autograft transfer or by the
implantation of artificial prosthesis (Willers, Partsalis et al. 2007). Microfractures create
fibrocartilage and not hyaline cartilage, and joint prosthesis is not a good choice in the young
patient because of the limited lifespan of prosthesis. Therefore, cartilage would be an optimal
target for tissue engineering applications. Brittberg et al. were the first to report autologous
chondrocyte implantation (ACI) in a clinical model, with autologous chondrocytes expanded
in vitro and subsequent implantation of cells in the joint, secured by a periosteal graft
(Brittberg, Lindahl et al. 1994).
However, the proliferative and functional capacity of chondrocytes is limited, with
dedifferentiation to a fibroblast-like state seen in prolonged in vitro culture (Schnabel,
Marlovits et al. 2002). The chondrocyte phenotype may be rescued by transferring cells to
three-dimensional culture systems (Benya and Shaffer 1982), supplementing culture
conditions with specific growth factors e.g. TGF-β1, BMP-2, IGF-1, or FGF-2 (Bobick, Chen
et al. 2009), and reducing oxygen tension (Domm, Schunke et al. 2002). Extrinsic applied
mechanical stimuli has been proposed for cartilage growth to mimic the natural environment
for articular chondrocytes, either by pressure (Waldman, Spiteri et al. 2004) or fluid shear
forces (Freed, Hollander et al. 1998).
9
To create a natural three-dimensional environment and to prevent dedifferentiation of
chondrocytes, cells are often combined with biomaterials. ACI has lately been used in clinical
trials with chondrocytes grown on collagen scaffolds (Behrens, Bitter et al. 2006). Usually
organic degradable polymers are used in cartilage tissue engineering but also some synthetic
degradable biomaterials like PLA, PGA or PLGA. Biomaterials for cartilage tissue
engineering are used in different forms, the usual scaffold architectures are hydrogels,
sponges or meshes (Chung and Burdick 2008). The large number of arthroscopically
performed procedures has led to interest in injectable biomaterials like hydrogels (alginate,
agarose, fibrin) (Sharma and Elisseeff 2004) or microcarriers (Malda and Frondoza 2006) for
cartilage regeneration.
Because of problems associated with chondrocyte harvest and culture, other cell sources
for cartilage tissue engineering have been proposed. The most commonly used cell source is
MSCs from bone marrow (Johnstone, Hering et al. 1998; Wakitani, Imoto et al. 2002), often
with the use of TGF-β in combination with various biomaterials, but so far no clinical trials
have been conducted (Chung and Burdick 2008). Cartilage like tissue has been created using
adipose derived stem cells in combination with chondrogenic induction medium and fibrin
glue (Dragoo, Carlson et al. 2007). Cartilage like tissue has also been created using dermal
FBs subjected to induction medium (French, Rose et al. 2004) or dermal FBs in combination
with demineralized bone powder and collagen (Mizuno and Glowacki 1996).
Tissue engineering of bone
Bone tissue engineering is a growing therapeutic field. There is a need for bone tissue in
the treatment of fractures with difficulties to heal and for correcting bone defects in
reconstruction after trauma or bone tumor surgery. The classical way of replacing missing
bone has been the use of allogenous bone or autologous iliac crest bone. There has also been
reports of successful percutaneous bone marrow injection in fractures with delayed healing
(Connolly, Guse et al. 1989). Allogenous bone has a low osteoinductive capacity and auto-
grafting from iliac crest has limitations in the amount of tissue available as well as a risk of
donor site morbidity (Enneking and Campanacci 2001; Sasso, LeHuec et al. 2005).
Biomaterials have a central position in bone tissue engineering applications. These
materials can be classified as osteoconductive (biomaterials only) osteoinductive
(biomaterials and soluble factors) or osteogenic (biomaterials and cells) (Hannouche, Petite et
al. 2001). Some of the common osteoconductive biomaterials used are calcium-based
ceramics including hydroxyapatite and tricalcium phosphate, or synthetic biopolymeres as
PLA, PGA or PLGA (Stevens, Yang et al. 2008). These materials are conductive in the sense
that they rely on bone ingrowth that is often confined to the surfaces, and therefore they are
not suitable for larger defects. Osteoinductive biomaterials combine scaffolds and growth
factors. Bone tissue releases several growth factors at the site of fracture, including BMPs,
TGF-β, platelet derived growth factor (PDGF), IGFs and FGFs (Tsiridis, Upadhyay et al.
2007). BMPs are commonly used osteoinductive factors. There are at least 18 BMPs of which
15 are known to be expressed in humans (Axelrad, Kakar et al. 2007). BMP 2 and 7 have
already been used in clinical trials in combination with different scaffolds, e.g. collagen,
PGA/PLA or bone autograft (Hannouche, Petite et al. 2001; Govender, Csimma et al. 2002;
10
Dimitriou, Dahabreh et al. 2005; Axelrad, Kakar et al. 2007). The efficiency of
osteoinductive biomaterials relies on the recruitment of osteocompetent cells from the
surrounding tissues. Their use is therefore limited to cases in which the fracture site can
provide these cells. This precludes their use in necrotic areas or in areas with large defects of
bone. To overcome this problem the use of osteogenic biomaterials; scaffolds loaded with
osteocompetent cells, has been proposed. The most commonly used cell source for seeding
onto osteogenic biomaterials is MSCs from bone marrow (Bruder, Kurth et al. 1998;
Pittenger, Mackay et al. 1999; Granero-Molto, Weis et al. 2009). MSC seeded osteogenic
implants facilitate bone regeneration as compared to unseeded scaffolds (Bruder, Kurth et al.
1998; Dallari, Fini et al. 2006). Some report the construction of bone tissue using other ASCs.
Bone like tissue was created in vivo using adipose derived stem cells in combinations with
PLGA (Cowan, Shi et al. 2004) or tricalcium phosphate and induction medium (Hattori,
Masuoka et al. 2006). Dermal FBs have also been osteogenic induced (Jahoda, Whitehouse et
al. 2003; Lorenz, Sicker et al. 2008). Dermal FBs transfected with BMP-2 (Hirata, Mizuno et
al. 2007; Wang, Zou et al. 2009) or BMP-7 (Rutherford, Moalli et al. 2002) showed
promising effects on fracture healing.
Clinical applications of tissue engineering
The scientific advances in the fields of stem cells, biomaterials, growth- and
differentiation factors have created unique opportunities to fabricate tissues in the laboratory
from combinations of engineered scaffolds, cells, and biologically active molecules. Among
the major challenges now facing tissue engineering is the need for more complex
functionality, as well as both functional and biomechanical stability in laboratory-grown
tissues destined for transplantation. If transplanting tissues, there is a need for blood supply.
Initially vessels are confined to the outer surface of the graft and cells within the graft
compete for nutrients and oxygen. Diffusion is able to support only a limited number of
transplanted cells creating central cell death (Muschler, Nakamoto et al. 2004). For larger
transplants there is a need for vascularization. Strategies for vascularization include the use of
porous biomaterials, application of growth factors like vascular endothelial growth factor
(VEGF) or in vitro prevascularization using endothelial cells (Rouwkema, Rivron et al. 2008).
In some reconstructive case reports, in vivo prevascularization has been achieved by growing
cells and biomaterials in muscles. Warncke et al. have described the microsurgical
reconstruction of mandible with MSCs and BMP-7 in a titanium mesh cultivated in the
latissimus dorsi muscle (Warnke, Springer et al. 2004).
Despite all efforts in tissue engineering research, few applications have so far had any
revolutionary impact on regenerative medicine. The use of stem cells has raised several
questions about how stable this cell source is. A clinical use of ESCs is also implying an
allogenous source with the risk of host vs. graft rejection. The clinical use of iPSCs implies
the use of genetically modified cells, which might have a risk to induce tumors. The same
question is raised using cells with genetic modifications with insertion of genes coding for
growth factors like BMPs.
There are multiple case reports of appealing applications; reconstruction of skull bone
using adipose derived stem cells (Lendeckel, Jodicke et al. 2004), formation of a trachea by
11
an acellular donor trachea colonized by epithelial cells and MSC derived chondrocytes
(Macchiarini, Jungebluth et al. 2008), reconstruction of urinary bladder by combination of
collagen scaffold, urothelial cells and muscle cells (Atala, Bauer et al. 2006), or the
construction of artificial corneas by collagen scaffolds and corneal stromal FBs (Griffith,
Jackson et al. 2009). However, so far no complex tissue engineering application has become
the golden standard of reconstruction.
ADIPOSE TISSUE
Histological and physiological features of adipose tissue
Adipocytes arise from mesodermal FB-like precursors; preadipocytes (PAs) (Rosen and
Spiegelman 2000). Only one third of adipose tissue is made up by mature adipocytes. A
combination of small blood vessels, nerve tissue, FBs, and PAs comprises the remaining two
thirds. Collections of white adipocytes comprise fat lobules, each of which is supplied by an
arteriole and surrounded by connective tissue septa (Avram, Avram et al. 2005). Although
some lipid accumulation can occur in many cell types, adipocytes are morphologically
different from other cells with a rounded shape due to the presence of large lipid droplets
surrounded by the protein perilipin (Blanchette-Mackie, Dwyer et al. 1995). The lipid droplet
occupies the majority of intracellular space, compressing the cytoplasm and nucleus into a
thin visible rim.
Fat storage occurs in white adipose tissue (WAT). Mammals also have brown adipose
tissue (BAT) important in thermogenesis. Heat production is regulated by uncoupling protein-
1 (UCP-1), a mitochondrial membrane protein. BAT is found in newborns and decreases
shortly after birth (Gesta, Tseng et al. 2007). The physiology of WAT can be grouped into
three main categories with potentially overlapping mechanisms: lipid metabolism, glucose
metabolism, and endocrine functions. Free fatty acids (FFA) are stored as triglycerides (TG)
in adipocytes and then released as FFA and glycerol in response to the energy demands of
other tissues. FFAs additionally regulate glucose homeostasis in other tissues. Adipocytes
utilize glucose for TG synthesis via insulin-responsive glucose transporter 4 (GLUT4), a
transmembrane transport protein present in fat cells. Endocrine functions include production
and secretion of peptides including; leptin, adipsin, adiponectin, angiotensin II, plasminogen
inhibitor activator I, prostaglandins and TNFα. Adipocytes also play a role in insulin
sensitivity, immunological responses and vascular diseases (Gregoire 2001; Avram, Avram et
al. 2005).
Regulators of adipogenesis
CCAAT/enhancer binding proteins (C/EBPs) and peroxisome proliferator-activated
receptors (PPARs) are master regulators of adipogenesis (Rangwala and Lazar 2000; Rosen
and Spiegelman 2000; Gregoire 2001). Exposure of PAs to adipogenic induction medium
induces expression of CEBPB and CEBPD, which in turn activate PPARG2 and CEBPA.
C/EBPα binds to and activate promotors of several adipocyte genes, including
fatty acid
binding protein 4 (FABP4), GLUT4, phosphoenolpyruvate carboxykinase (PEPCK), leptin,
and the insulin receptor (Rosen and Spiegelman 2000). Synthetic PPAR agonists like fibrates
12
and thiazolidinediones are used therapeutically to treat dyslipidaemia and diabetes (Ahmed,
Ziouzenkova et al. 2007). Sterol regulatory element binding transcription factor 1
(ADD1/SREBP-1c) is another master regulator known to modulate transcription of
lipoprotein lipase and fatty acid synthetase, as well as promoting adipocyte differentiation
(Osborne 2000).
In addition to C/EBPs, PPARγ2, and ADD1/SREBP-1c, several other transcription
factors, including GATA-binding proteins 2 and 3 (GATA2 and GATA3), and cAMP
response element binding protein (CREB), play a critical role in the molecular control
of the
adipogenic differentiation. GATA2 and GATA3 are specifically expressed in PAs, and
their
mRNAs are down-regulated during adipocyte differentiation. Constitutive expression of
GATA2 and GATA3 suppresses adipocyte differentiation and traps cells at the PA stage (Tong,
Dalgin et al. 2000). The transcription factor CREB is constitutively expressed prior to and
during adipogenesis, and is up-regulated by conventional differentiation-inducing agents such
as insulin and dexomethasone (DEX) (Reusch, Colton et al. 2000).
Signaling molecules such as preadipocyte factor 1 (PREF1) and WNTs also regulate
adipocyte differentiation. PREF1 is an inhibitor of adipocyte differentiation. It is a plasma
membrane protein which activates the mitogen-activated protein kinase/extracellular signal-
regulated kinase (MAPK/ERK) pathways and increases SOX9 which blocks transcription of
CEBPB and D (Sul 2009). PREF1 is highly expressed in PAs, but is not detectable in mature
adipocytes. WNT signaling also appears to be a molecular switch that governs adipogenesis.
PAs that constitutively express WNT1 failed to differentiate when treated with adipogenic
induction medium. Moreover, activation of WNT signaling downstream of the receptor also
inhibits differentiation, indicating that WNT signaling maintains PAs in an undifferentiated
state. This effect seems to be mediated through inhibition of CEBPA and PPARG (Ross,
Hemati et al. 2000).
Adipogenic induction factors
In the experiments presented in this thesis, adipogenic differentiation was achieved by
subjecting cells to treatment with DEX, 3-isobutyl-1metylxanthine (IBMX), indomethacin
and insulin (Pittenger, Mackay et al. 1999).
Dexomethasone: DEX acts by inhibiting PREF1 transcription
and thereby promotes
adipogenesis. Down-regulation of PREF1 is required for adipose conversion. Glucocorticoids
also induce expression of CEBPD. This increase may contribute to the formation of C/EBPδ-
C/EBPβ heterodimers, which in turn may lead to PPARG expression (Wu, Bucher et al.
1996).
Glucocorticoid effects may also be mediated through increased metabolism of
arachidonic acid leading to an increase in production of prostacyclin, which
in turn increases
intracellular cyclic adenosine monophosphate (cAMP) (Gregoire, Smas et al. 1998).
IBMX: Methylxanthines (such as caffeine and theophylline) are an important and widely
used class of drugs believed to mediate many of their physiological effects by increasing
intracellular concentrations of cAMP. These agents are known to inhibit phosphodiesterases
and to block inhibitory A1 adenosine receptors in a competitive manner. It also stimulates
13
adenylyl
cyclase activity by blocking the inhibitory regulatory protein
Gi Thus, the
methylxanthines may increase cAMP accumulation by slowing its degradation or by
enhancing its production (Parsons, Ramkumar et al. 1988). IBMX has also been shown to
increase expression of CEBPB, and this increase is required for subsequent
PPARG expression
and adipocyte differentiation (Gregoire, Smas et al. 1998).
Indomethacin: Indomethacin is a non-steroid anti-inflammatory drug that inhibits the
function of inducible cyclooxygenase (COX-2). The adipogenic activity of indomethacin
cannot simply be ascribed to the inhibition of COX, the concentration of drug required to
induce differentiation is 2-3 orders of magnitude higher than that required to inhibit COX
activity. The probable mode of action in adipogenesis is mediated by indomethacin binding
directly to PPARα and PPARγ and thereby functioning as a ligand for these adipogenic
transcription factors. PPAR-ligands, including the anti-diabetic thiazolidinediones and the
arachidonic acid metabolite 15-deoxy-Δ12,14
-PGJ2, are potent inducers of the differentiation of
several different fibroblastic cell lines to adipocytes (Lehmann, Lenhard et al. 1997).
Insulin: Insulin has considerable homology to IGF-1. Smith et al. reported that IGF-1 and
insulin are essential factors for adipocyte differentiation, using fetal calf
serum depleted of
growth hormone (GH), insulin, and IGF-1 (Smith, Wise et al. 1988). The effect is mediated
by insulin receptors (IR) and IGF-1-receptors (IGF1R) which both belong to the family of
tyrosine kinases. IRs are more important in fuel metabolism whereas IGF1Rs mediate growth.
Insulin at supraphysiological doses can mimic the effects of IGF-1 in PA differentiation,
presumably by activation of IGF1Rs (Bluher, Kratzsch et al. 2005). The adipogenic effect of
IGF and Insulin has been shown to be mediated by phosphatidylinositol 3-kinase activity
(Christoffersen, Tornqvist et al. 1998).
CARTILAGE TISSUE
Histological and molecular features of cartilage
Paraxial mesoderm in vertebrate embryos is condensed into cell clusters called somites.
Cells located in the dorsal domain of the somite become the dermomyotome, a progenitor
tissue that gives rise to skeletal muscle and dermis. Cells located in the ventral domain of the
somite form the sclerotome, a progenitor tissue that gives rise to the cartilage template of the
skeleton (Zeng, Kempf et al. 2002).
Cartilaginous tissues are classified histologically as being hyaline, elastic or
fibrocartilaginous in nature, depending on their molecular composition. Hyaline cartilage is
the predominant form of cartilage and is associated with the skeletal system where it forms
the construct for many bones via endochondral ossification, the growth plate, and the bearing
surface of joints (Roughley 2006). Elastic cartilage is associated with the ear and the larynx,
whereas fibrocartilage is associated with the menisci of the knee and the intervertebral discs.
Fibrocartilage is better suited to resist tensile loads than hyaline cartilage whereas hyaline
cartilage better withstand compressive loads (Almarza and Athanasiou 2004). Adult cartilage
tissue is avascular. It has limited self-repair capacity due to the sparse distribution of highly
14
differentiated, non-dividing chondrocytes, slow matrix turnover and low supply of progenitor
cells (Willers, Partsalis et al. 2007).
Collagen is a family of extracellular matrix (ECM) proteins that make up 2/3 of the dry
weight of cartilaginous tissues. There are 29 described subtypes of collagens (Soderhall,
Marenholz et al. 2007), of which type I and type II are the most common in cartilage.
Collagens are made up of three polypeptide strands forming a triple helix. Collagens give
cartilage strength by forming a heavily cross-linked network. The fibrillar organization
changes with tissue depth (Eyre 2004).
Hyaline cartilage is exclusively composed of chondrocytes, which produce ECM. The
chondrocytes are located in lacunae in the ECM, and the cellular content constitutes
approximately 10 % of the tissue. Articular cartilage is divided into a superficial, middle and a
deep zone. Chondrocytes in the superficial zone are flattened and elongated, while cells in the
middle zone appear rounded, and in the deep zone chondrocytes have an ellipsoid
morphology (Almarza and Athanasiou 2004). In human hyaline cartilage collagens II, IX and
XI form the fibrillar network. Their proportions fall from high levels in the fine fibrillar
networks of growing cartilages (≥10 % IX, ≥10 % XI, ≥80 % II) to the much thicker fibrils of
adult articular cartilage (~1 % IX, ~3 % XI, ≥90 % II). Other collagen types found in small
amounts in hyaline cartilage include types III, VI, X, XII and XIV (Cremer, Rosloniec et al.
1998; Eyre 2004). Hyaline cartilages are characterized by their high content of the
proteoglycan aggrecan. Proteoglycans are glycoproteins that are heavily glycosylated. The
core protein of aggrecan consists of three disulphide bonded globular regions and
glucosaminoglycan (GAG) attachment regions (Roughley 2006). GAGs are polysaccharides
with negatively charged sulphates. Water molecules are bound to these negative charges, and
this hydration increases the compressive resistance of the tissue. The GAG content of hyaline
cartilage is 15-40 % depending on age, gender and location. The most abundant GAGs are
chondroitin-, keratan- and dermatan sulphate (Almarza and Athanasiou 2004). Aggrecan
exists as aggregates in association with hyaluronan and link protein (Roughley 2006). Each
aggregate is composed of a central filament of hyaluronan with up to 100 aggrecan molecules
radiating from it with each interaction stabilized by the presence of a link protein (Morgelin,
Paulsson et al. 1988) (figure 3). Other non-collagenous proteins in cartilage are the
thrombospondins including cartilage oligomeric matrix protein (COMP), fibronectin and
vitronectin. These proteins take part in the interaction between the chondrocytes and the
matrix. In common with all connective tissues, cartilage also contains small leucine-rich
repeat proteoglycans (SLRPs) like decorin, biglycan, fibromodulin, lumican and
proline/arginine-rich end leucine-rich repeat protein (PRELP) (Roughley 2006; Heinegard
2009).
Elastic cartilage shares a lot of the properties of hyaline cartilage. It possesses a similar
ECM composition except for the content of elastin. The elastic fibers of elastin are composed
of the protein tropoelastin and microfibrils (Mithieux and Weiss 2005). Elastic cartilage has a
higher cellular content than hyaline cartilage (Kuhne, John et al. 2010).
15
Figure 3. Aggrecan. Hyaline cartilage has a high content of the proteoglycan aggrecan. The core protein of
aggrecan consists of three disulphide bonded globular regions (G1-G3) and glucosaminoglycan (GAG)
attachment regions. The most abundant GAGs are chondroitin-, keratan- and dermatan sulphate. Aggrecan exists
as aggregates in association with hyaluronan (HA) and link protein (LP). Each aggregate is composed of a
central filament of HA with up to 100 aggrecan molecules radiating from it with each interaction stabilized by
the presence of a LP.
Fibrocartilage is composed of chondrocyte-like and fibroblast-like cells, with a ratio of
about 1:4, but some consider fibrocartilage as composed by one cell type only, named
fibrochondrocytes. The main collagen type is collagen I, accounting for over 80% of total
collagen content. There are also traces of collagen II, III and V. Fibrocartilage has a GAG
content of about 3% on a dry weight basis (Almarza and Athanasiou 2004).
Regulators of chondrogenesis
The earliest marker of chondrogenesis is the chondrogenic master regulator SRY (sex
determining region Y)-box 9 (SOX9) which directly regulates expression of type II collagen
(COL2A1) and aggrecan (ACAN), and is required for expression of genes that encode minor
matrix proteins, including link proteins and type IX and XI collagen. SOX9 function is
enhanced by SOX5 and SOX6, which bind to SOX9 and act as cofactors (Akiyama 2008).
There is a balance between SOX9 and runt-related transcription factor 2 (RUNX2) regulating
the possible differentiation fates in skeletal tissue. High levels of SOX9 will commit cells to
chondrogenesis, whereas high levels of RUNX2 will push cells toward osteogenesis and start
the endochondral ossification. SOX9 has been shown to be dominant to RUNX2 (Eames,
Sharpe et al. 2004). NK3 homeobox 2 (NKX3.2) has a role in enhancing chondrocyte
16
differentiation while, together with SOX9, repressing a terminal hypertrophic chondrocytic
fate by inhibiting RUNX2 expression (Zeng, Kempf et al. 2002). The expression of SOX9 is
regulated by members of the FGF, TGF-β, BMPs, Hedgehog and WNT families (Quintana,
zur Nieden et al. 2009). Sonic hedgehog and BMPs induce expression of NKX3.2 (Zeng,
Kempf et al. 2002).
The pathways of endochondral ossification have been investigated thoroughly, but the
events leading to the formation of joints lined by hyaline cartilage is still a matter of
discussion. It has been shown that the articular cartilage is created by appositional growth by
progenitor cells at the surface of the cartilage (Dowthwaite, Bishop et al. 2004).These cells
have a high NOTCH-1 activity, and NOTCH-related pathways have therefore been discussed
as the regulator of articular cartilage formation (Karlsson and Lindahl 2009). The WNT
pathway has also been discussed. WNT14, acting through the non-canonical pathway has
been shown to be required for the earliest steps of joint specification. It negatively regulates
chondrogenic differentiation creating the interzone of future joints (Hartmann and Tabin
2001). BMP-7 is a potential inhibitory factor for joint formation (Macias, Ganan et al. 1997)
while growth/differentiation factor 5 (GDF-5) is required for joint formation (Storm and
Kingsley 1996).
Chondrogenic induction factors
In the experiments presented in this thesis, chondrogenic differentiation was achieved by
culturing cells in medium containing low amounts of fetal calf serum (FCS) supplemented
with ascorbate-2-phosphate (A2P), insulin and TGF-β1 (Johnstone, Hering et al. 1998). The
low concentration of FCS reflects the limited vascularization of native cartilage tissue with
low perfusion of nutrients to chondrocytes.
A2P: Ascorbate, better known as Vitamin C, has a transcriptional effect on cartilage
formation promoting COL2A1 and Prolyl-4-hydroxylase alpha (P4HA) gene expression
(Clark, Rohrbaugh et al. 2002). At the post-transcriptional level, ascorbate is responsible for
the reduction of iron to its ferrous state. This reaction is important for the hydroxylation of
proline and lysine by P4HA. These are essential amino acids for formation of collagen triple
helices. Additionally collagen secretion from the cells is promoted by ascorbate-dependent
hydroxylation within the endoplasmatic reticulum. The specific form or dose does not seem
very important (Ibold, Lubke et al. 2009).
Insulin: The effect of insulin is mediated by IR and IGF1R. As in the case of adipogenic
induction, chondrogenic induction is believed to be mediated through phosphatidylinositol 3-
kinase activity. Reports indicate that intracellular signaling after chondrogenic induction
includes several MAPK pathways. These effects are mediated by the classical MAP-kinase
pathway by ERK1/2, but also by the p38 (MAPK14) kinase signaling pathway. TGF-β
signaling is thought to co-operate with insulin signaling (McMahon, Prendergast et al. 2008).
17
TGF-β1: In the interdigital region of developing limbs TGF-β induces the expression of
SOX9. This induction is a quick process as the up-regulation of SOX9 has been shown to
occur only 30min after exposure of chicken limb bud cells to TGF-β (Chimal-Monroy,
Rodriguez-Leon et al. 2003). Apart from regulating the expression of SOX9, TGF-β
ultimately initiates the expression of COMP, ACAN, COL2A1, collagen XI (COL11A1) and
fibronectin (Quintana, zur Nieden et al. 2009). TGF-β1 shortens the time course of
chondrogenesis, increases the amount of cartilage formation and negatively regulates
osteogenesis (Iwasaki, Nakata et al. 1993). TGF-β also induces expression of aggrecanases
like ADAM metallopeptidase with thrombospondin type 1 motif, 4 (ADAMTS4) and
promotes the degradation of aggrecan indicating that it may be involved in normal turnover of
proteoglycans in mature cartilage (Moulharat, Lesur et al. 2004). This highlights the complex
role of TGF-β in chondrocyte matrix metabolism.
High density culture: Chondrogenesis is higher in areas with high cell density. It is believed
that the high seeding density mimics the mesenchymal condensations that occur during
embryonic cartilage formation (Bobick, Chen et al. 2009). The culture of chondrocytes in
pellet culture, also termed micro mass, to prevent dedifferentiation of cells was first described
in 1960 (Holtzer, Abbott et al. 1960). It has been shown beneficial to culture cells in a three-
dimensional environment to promote chondrogenic differentiation (Chen, Zhang et al. 2007).
BONE TISSUE
Histological and physiological features of bone
The majority of bones are formed by endochondral ossification. Few bones, such as the
flat bones of the skull and the clavicle are formed by intramembranous ossification (Hartmann
2009). Endochondral bone formation starts by condensation of mesenchymal cells forming
cartilaginous templates, while intramembranous ossification starts without prior formation of
a cartilaginous mold whereby cells within the condensed mesenchymal cell layer differentiate
directly into bone-forming cells. As endochondral bone formation proceeds, proliferating
chondrocytes progressively exit the cell cycle, hypertrophy, and become hypertrophic
chondrocytes. These cells produce a matrix, which stays behind after the cells have undergone
apoptosis and is used as a template for bone formation by osteoblasts. Some authors mean that
it is the cartilaginous cells themselves that differentiate into osteoblasts, others believe that the
outer cells (perichondrium) will give rise to osteoblast precursors and is then called
periosteum (Karsenty 2008; Hartmann 2009).
Bone is a continuously remodeling tissue. This remodeling is started by osteoclasts in
terms of bone resorption. Whereas osteoblasts are of mesenchymal origin, osteoclasts belong
to the monocyte-macrophage cell lineage. At the completion of bone resorption, resorption
cavities contain a variety of mononuclear cells, including monocytes, osteocytes released from
bone matrix, and preosteoblasts recruited to begin new bone formation. Osteoblasts
synthesize
new collagenous organic matrix and regulate mineralization of matrix by releasing small
membrane-bound matrix vesicles that concentrate calcium and phosphate and enzymatically
destroy mineralization inhibitors such as pyrophosphate or proteoglycans. Osteoblasts
18
surrounded by and buried within the matrix become osteocytes with an extensive canalicular
network connecting them to bone surface lining cells, osteoblasts, and other osteocytes
(Clarke 2008). Bone protein is composed of 85 to 90 % collagenous proteins. Bone matrix is mostly
composed of type I collagen, with small amounts of types III and V and FACIT collagens
(Members of this family include collagens IX, XII, XIV, XIX, XX, and XXI) (Clarke 2008).
Noncollagenous proteins compose 10 to 15 % of total bone protein. Approximately 25 % of
noncollagenous protein is
exogenously derived, including serum albumin and α2-HS-
glycoprotein, which bind to hydroxyapatite because of their acidic properties.
Endogenous
glycoproteins present in bone are alkaline phosphatase (ALP), osteonectin, osteopontin, bone
sialoprotein, fibronectin and osteocalcin (Young, Kerr et al. 1992; Clarke 2008). ALP is the
most common and is bound to osteoblast cell surfaces via a phosphoinositol linkage and is
also found free within mineralized matrix. ALP plays a role in mineralization of bone
(Magnusson and Farley 2002). Osteonectin is the most prevalent noncollagenous protein and
accounts for approximately
2% of total protein in developing bone. Osteonectin and
osteocalcin are thought
to affect osteoblast growth and/or proliferation and matrix
mineralization (Young, Kerr et al. 1992). There are also some GAG proteins like aggrecan
and versican present in bone. Bone is composed of 60% mineral. The mineral content of bone
is mostly hydroxyapatite (Ca10(PO4)6(OH)2), with small amounts of carbonate, magnesium,
and acid phosphate (Feng 2009). Matrix maturation is associated with expression of alkaline
phosphatase and several noncollagenous proteins, including osteocalcin and osteonectin. It is
thought that these calcium- and phosphate-binding proteins help regulate ordered
deposition of
mineral by regulating the amount and size of hydroxyapatite crystals formed (Clarke 2008).
Regulators of osteogenesis
As discussed above there is a balance between SOX9 and RUNX2. High levels of SOX9
will commit cells to chondrogenesis, whereas high levels of RUNX2 will push them toward
osteogenesis (Eames, Sharpe et al. 2004). RUNX2 in combination with RUNX3 have an
important function in the terminal differentiation of chondrocytes by inhibiting chondrocytes
from acquiring the phenotype of articular cartilage. RUNX2 and RUNX3 are positive
regulators of chondrocyte hypertrophy which is a prerequisite for endochondral ossification.
RUNX2 regulates the expression of collagen X (COL10A1) in hypertrophic chondrocytes and
the expression of osteopontin (SPP1), integrin-binding sialoprotein (IBSP), and
metalloproteinase 13 (MMP13) in terminal hypertrophic chondrocytes (Komori 2010).
Osteoblastic precursors that produce active RUNX2 are not yet fully committed to become
osteoblasts, for this they require the Kruppel-like zinc-finger transcription factor osterix/SP7.
Osterix is expressed primarily in osteoblasts, but also weakly in prehypertrophic
chondrocytes. Osterix ensures the differentiation into osteoblasts instead of chondrocytes
(Gao, Jheon et al. 2004). In mice, inactivation of osterix results in prenatal lethality owing to
a complete absence of bone formation (Nakashima, Zhou et al. 2002). RUNX2 expression is
regulated by several signaling molecules. Chondrocyte maturation and osteoblast
differentiation in endochondral bone formation are coupled via Indian hedgehog (IHH)
signaling. IHH produced by prehypertrophic chondrocytes induces osteoblast differentiation
19
in adjacent perichondrial cells largely by increasing parathyroid hormone-related protein
(PTH-rP). IHH is required for endochondral but not for intramembranous ossification (St-
Jacques, Hammerschmidt et al. 1999). RUNX2 is negatively regulated by several factors.
RUNX2 is repressed by the basic helix–loop–helix transcription factor TWIST1 in chicken
chondrocytes (Dong, Soung do et al. 2007). Signal transducer and activator of transcription 1
(STAT1) is also a known inhibitor of RUNX2 expression. Accordingly, STAT1-deficient mice
develop a high-bone-mass phenotype (Karsenty 2008).
Reports indicate that WNT signaling, via the canonical β-catenin pathway functions in a
cell autonomous manner to induce osteoblast differentiation and suppress chondrocyte
differentiation in early chondroprogenitors (Hill, Spater et al. 2005). Activating transcription
factor 4 (ATF4 /CREB2) play an important role in assuring that osteoblasts fulfill their
function. ATF4 directly activates the transcription of the osteocalcin gene, and indirect also
regulates proper synthesis of type I collagen. In addition to the role of ATF4 in bone
formation, ATF4 regulates osteoclast differentiation and ultimately bone resorption through
its expression in osteoblasts (Karsenty 2008).
Osteoblastogenesis is regulated by various secreted factors, such as IHH, BMPs, TGF-β,
FGFs, and IGFs (Karsenty and Wagner 2002). The most important seem to be the BMP-
family and TGF-β (Axelrad, Kakar et al. 2007; Deschaseaux, Sensebe et al. 2009)
Osteogenic induction factors
In the experiments presented in this thesis, osteogenic differentiation was achieved by
subjecting cells to treatment with A2P, DEX and β-glycerophosphate (BGP) (Jaiswal,
Haynesworth et al. 1997).
A2P: Ascorbate plays a dual role in bone homeostasis; that is, as an anti-oxidant modulating
osteoclast proliferation and as a cofactor in the activation of transcription factors that promote
osteoblast differentiation (Gabbay, Bohren et al. 2010). In osteoblasts, A2P stimulates cell
growth and is a positive regulator of ECM production, increasing rates of both procollagen
hydroxylation and secretion by the same mechanisms as described in “Chondrogenic
induction factors” (Takamizawa, Maehata et al. 2004).
DEX: The adverse effects of glucocorticoids on bone are well known since the description of
Cushing’s disease. Glucocorticoids regulate several crucial determinants of osteoblast
differentiation, proliferation and function. They exert a permissive effect on early
osteogenesis of bone marrow cells and osteoblasts in vitro. These effects are complex and at
times opposing, and depending upon dose and stage of osteoblast differentiation. A potential
mechanism is modulation of expression and action of BMPs (Cooper, Hewison et al. 1999).
DEX exerts a direct transcriptional effect on ALP and other bone related proteins.
Glucocorticoids also change the sensitivity of osteoblasts to other growth factors and
hormones like IGF-1 and TGF-β (Cooper, Hewison et al. 1999). DEX treatment up-regulates
the parathyroid hormone receptor and increases the basal level of cAMP by activation of
adenylate cyclase (Ogston, Harrison et al. 2002).
20
ΒGP: β-glycerophosphate is an organic phosphate which favors the mineralization of
extracellular matrix by increasing the availability of phosphate ions (Pi). The mechanism by
which BGP induces mineralization is closely linked to the high ALP-activity in bone tissues.
ALP hydrolyzes organic phosphates and elevates the local Pi concentration and thereby
provides the chemical potential for tissue mineralization (Chung, Golub et al. 1992).
Mechanical forces: Fluid induced shear stress from spinner flask culture of microcarriers has
been shown to have a positive impact on osteogenic differentiation (Datta, Pham et al. 2006;
Sommar, Pettersson et al. 2009). These results suggest that shear stress by itself has an impact
on osteogenic differentiation, which is confirmed by several reports (Sikavitsas, Bancroft et
al. 2003; Yu, Botchwey et al. 2004; Holtorf, Jansen et al. 2005; Schwartz, Denison et al.
2007).
ENDOTHELIAL CELLS
Histological and physiological features of the endothelium
The origin of endothelial cells is postulated to be the same as for the hematopoietic cell
lines; the hemangioblast with dual differentiation capacity. Mesodermal cells proliferate and
form solid lumps called blood islands, the peripheral cells of these islands differentiate to
endothelial cells and central cells form hematopoietic cells (Ribatti 2008).
The main functions of endothelial cells are involvement in barrier function, secretion of
active compounds, retaining an antithrombogenic surface, participation in coagulation, a role
in innate immunity and control of the vascular tone (Bachetti and Morbidelli 2000).
The integrity of the barrier depends on junctional zone by which endothelial cells are linked
together to form a monolayer. This zone contains both tight junctions and adherens junctions.
The most important junctional protein in the endothelium is vascular endothelial cadherin
(Ve-cadherin/CDH5) (Vestweber, Winderlich et al. 2009). Multiple inflammatory mediators
also affect the endothelial permeability, and do so by engaging intracellular messengers such
as cAMP, calcium, phosphoinositol lipids and various hydrolase enzymes.
Endothelial cells are active participants in the immune response. Prominent functions are
the role in adhesion and migration of leukocytes but no less important are chemotactic
cytokine release and antigen presentation to lymphocytes. Leucocytes bind to intercellular
adhesion molecules 1 and 2 (ICAM) and vascular cell adhesion molecule (VCAM) (Sumpio,
Riley et al. 2002).
Vascular endothelial growth factor (VEGF) is an endothelial cell mitogen and a potent
regulator of the endothelial phenotype and signaling. VEGF binds to its two receptors,
VEGFR-1 and VEGFR-2/KDR. VEGFR-2 is encoded by the kinase insert domain receptor
gene (KDR) in humans (Olsson, Dimberg et al. 2006). VEGF enhances cellular permeability,
and the induced disruption of the endothelium is a prerequisite for one of the growth factor´s
most important functions; to induce angiogenesis (Mehta and Malik 2006).
Endothelial cells play a crucial role in maintaining an antithrombogenic surface between
blood and tissue. In addition to lining the vessel, endothelial cells actively regulate
thrombosis, thrombolysis and platelet activation by secreting antithrombotic and pro-
21
coagulant factors (Bachetti and Morbidelli 2000). One prominent marker for endothelial cells
is the expression of von Willebrand factor (vWf), an important procoagulative glycoprotein
expressed only by endothelial cells and megakaryocytes. Defects in vWf synthetization or
function is strongly manifested as prolonged bleeding time in patients with von Willebrand’s
disease (Hannah, Williams et al. 2002).
Finally, endothelium is a powerful regulator of blood flow and vascular tone. Endothelial
cells secrete substances that have pronounced effect on smooth muscle cells and act in
proximity of their production site. Here the antithrombogenic and vasoactive functions often
converge, as is the case for prostacyclin and nitric oxide (NO) which both are vasodilators and
inhibit platelet aggregation (Sumpio, Riley et al. 2002). A strong marker for endothelial
phenotype is the presence of endothelial nitric oxide synthase (eNOS). Endothelial cells also
function as guard keepers scanning the blood for non-favorable agents. The clearance of
modified lipoproteins serves as an example of this screening. The endothelial cells express
receptors for normal low density lipoprotein (LDL) and high density lipoprotein (HDL), but
also scavenger receptors for e.g. oxidized LDL.
Regulators of endotheliogenesis
In endothelial cells or progenitors, no single master regulating transcription factor has
been identified. This has led to the suggestion that endothelial-specific gene expression is
controlled combinatorially by multiple transcription factors (De Val and Black 2009). T-cell
acute lymphocytic leukemia 1 protein (TAL1) and GATA2 are expressed during
embryogenesis and are essential for both blood and endothelial cell development. Several
genes in the forkhead family (FOX) (Papanicolaou, Izumiya et al. 2008) and KLF-family
(Atkins and Jain 2007) also play important roles in endothelial differentiation shown by the
disturbed vascularization and mortal faith in animals missing one of these four factors.
Although many factors have important roles, no one seems to have such a central role in
transcriptional programs controlling endothelial cell development as E-twenty six (ETS)
proteins. There are numerous ETS-binding-sites in all known endothelial enhancers and
promoters, but they are not specific for endothelial regulation, why they must function in
combination with other transcription factors. Indeed, many endothelial specific enhancers
contain clustered binding sites for multiple transcription factors suggesting that endothelium-
specific gene expression is combinatorially controlled. FOXC2 and ETV2 seem to have
specific combinatory effect in endothelial enhancers (De Val and Black 2009).
Some reports indicate that HOXA9 is a master regulator of endothelial cell fate
commitment of progenitor cells. HOXA9 is transcribed after activation of histone deacetylase
1 (HDAC1) and it acts as a switch to regulate the expression of genes such as eNOS, KDR,
and CDH5. HOXA5 has been reported to keep endothelial cells in a quiescent state by
inhibiting pro-angiogenic signaling by partially diminishing KDR expression (Rossig, Urbich
et al. 2005).
Several signaling peptides cooperate to promote mesodermal specification for blood
vessel formation. IHH up-regulates BMP-4 and FGF-2. Aberrant vasculogenesis has been
seen without BMP-4, and FGF-2 is thought to promote the expression of the VEGF receptor
KDR that is critical for promoting vasculogenesis (Hofer and Schweighofer 2007; Goldie, Nix
22
et al. 2008). VEGF and FGF-2 exert positive effects on endothelial proliferation, and they are
balanced by retinoic acid (RA) and TGF-β1 which have an anti-proliferative activity. Once
blood vessels have formed, hemodynamic forces/shear stress is an important regulator of
endothelial proliferation and gene expression (Goldie, Nix et al. 2008).
Endothelial induction factors
In the experiments presented in this thesis, endothelial differentiation was achieved by
subjecting cells to treatment with 30 % human serum. Endothelial cells are commonly
cultured in media supplemented with isobutyl-methylxantine (IBMX) and cholera toxin, but
unpublished results from our group (Karlsson et al. submitted) show that human serum in the
dose of 30 % is sufficient alone to drive dermal FBs into endothelial differentiation.
Human serum contains numerous growth factors which can have an effect on endothelial
differentiation and proliferation. In embryonic development BMP-4 and FGF-2 have critical
roles in formation of mesoderm, and mesodermal progenitors are directed into adipogenesis
by VEGF. VEGF and FGF-2 are needed for the induction of endothelial cell proliferation.
Counteractive anti-proliferative signalling by RA and TGF-β1 is an essential requirement for
appropriate blood vessel formation (Goldie, Nix et al. 2008).
23
AIMS
The main aim of this thesis was to investigate the plasticity of human dermal FBs and to
explore their possible role in bone and cartilage tissue engineering applications.
The specific aims were:
To investigate the differentiation potential of human dermal FBs by subjecting them to
adipogenic- chondrogenic- and osteogenic induction media.
To study the possibility of constructing three-dimensional tissues by combining biomaterials
and differentiated human dermal FBs.
To study the in vivo application of osteogenic induced human dermal FBs.
To characterize the differentiation process in human dermal FBs on a gene expression level.
24
25
MATERIALS AND METHODS
All in vitro culture of cells used for experiments described in this thesis was performed
on cell culture grade polystyrene culture flasks unless stated otherwise. Cells were incubated
at 37 °C, 95 % humidity, 21 % oxygen and 5 % carbon dioxide.
ETHICAL PERMISSION
Cells used in this thesis were collected from discarded tissue from routine plastic surgery
operations. No ethical permission was needed since origin of tissue was not traceable. All
patients were asked before surgery for permission to donate tissue. Tissue was delivered to the
lab marked only with sex and age.
The in vivo experiments in paper III were performed in accordance with a protocol
approved by the Norwegian Institutional Animal Care and Use Committee (Project 2008180).
CELL HARVEST AND INDUCTION
Fibroblasts
Human FBs used in paper I-IV were obtained from tissue specimens taken from healthy
male and female patients undergoing routine plastic surgery. The dermal component of the
skin was cut into small pieces and treated over night with collagenase type 1 and dispase.
After centrifugation and resuspension, cells were seeded in culture flasks and cultured in FB
proliferation medium (FM) containing Dulbecco´s minimum essential medium (DMEM), 10
% fetal calf serum (FCS), penicillin and streptomycin (PEST).
Preadipocytes
Human PAs, which have been shown to possess stem cell plasticity (Zuk, Zhu et al.
2001; Zuk, Zhu et al. 2002), were used in paper I as controls regarding the differentiation
potential. PAs were obtained from tissue specimens taken from healthy male and female
patients undergoing routine plastic surgery. The fat was cut into small pieces and treated with
collagenase type 1 for 30 min. After centrifugation and resuspension cells were passed
serially through cell strainer filters to remove debris. Finally, cells were seeded in PA
proliferation medium (PM) containing DMEM/Ham’s F-12, 10 % FCS and PEST.
Chondrocytes
Human articular chondrocytes (CCs) in paper IV were isolated from cartilage tissue
taken from a total knee arthroplasty in a male patient. The cartilage was minced and treated
over night in collagenase type 2 and dispase. After centrifugation and resuspension, cells were
seeded in culture flasks and cultured in chondrocyte proliferation medium (CPM) containing
DMEM, 10 % FCS, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES),
0.1 mM MEM non-essential amino acid solution, 0.4 mM L-proline, 0.2 mM A2P, and PEST.
26
Osteoblasts
Human osteoblasts (OBs) used in paper III and IV were isolated from bone tissue
obtained from a female patient undergoing hip arthroplasty. Blood clots were removed by
agitation in PBS. Bone tissue was minced and serially treated with collagenase type 2 in alpha
minimum essential medium (α-MEM). In paper III, OBs were cultured in osteoblast
proliferation medium (OPM) containing α-MEM, 10 % FCS and PEST. In paper IV, OBs
were cultured in osteogenic induction medium (OM) as shown in table I.
Endothelial cells
Human endothelial cells (HUVECs) used in paper IV were obtained from umbilical
cords taken from newborns. The veins of the umbilical cords were injected with DMEM
containing collagenase type 1. After 20 min, the veins were massaged and perfused with
DMEM. After centrifugation and resuspension, cells were seeded in gelatin coated culture
flasks and cultured in endothelial proliferation medium (EPM) containing DMEM, 30 %
human serum, 3.3 mM IBMX, 0.8 mg/ml cholera toxin and PEST.
Single cell cloning
Single cell cloning was employed to study differentiation on a single cell level. This
method relies on the production of separate cultures which originates from one single cell. By
performing clonal analysis, the possibility of contaminant progenitor cells from other tissues
can be ruled out. The potential presence of subpopulations exhibiting stem cell plasticity can
be studied by comparing results from differentiation experiments performed on primary
culture cells and single-cell clones derived thereof. Studies on this cell population can also
bring insight as to whether fusion of cells is responsible for the phenotype shift associated
with stem cell plasticity.
Single cell human fibroblasts (sccFBs) were prepared in paper I using a TransferMan
micromanipulator and a CellTramAir micropipette (Eppendorff, Germany) fitted to an IX51
inverse light/fluorescence microscope (Olympus, Sweden). FBs were trypsinized and single
cells transferred to separate wells of 24 well cell culture plates. The sccFBs were cultured in
FM until colonies could be detected. The plating efficiency was approximately 15%.
Induction of differentiation
In order to induce the phenotypic shift of FBs, PAs and sccFBs towards adipocyte,
chondrocyte, endothelial and osteoblast-like cell types in paper I-IV specific induction media
were used, which are shown in table I.
27
Table I. Induction media used for differentiation of FBs.
Medium
Media Serum Supplement
Adipogenic (AM) DMEM 10 % Fetal calf serum 1 µM dexamethasone
0.5 mM isobutyl-methylxantine
10 µM insulin
200 µM indomethacin
50 U/ml penicillin
50 g/ml streptomycin
Chondrogenic (CM) DMEM 1 % Fetal calf serum 50 nM ascorbate-2-phosphate
1.125µM insulin
10 ng/ml TGF-1
50 U/ml penicillin
50 g/ml streptomycin
Osteogenic (OM) DMEM 10 % Fetal calf serum 50 M ascorbate-2-phosphate
0.1 M dexamethasone
10 mM β-glycerophosphate
50 U/ml penicillin
50 g/ml streptomycin
Endotheliogenic (EM)
DMEM 30% Human serum
CONFIRMATION OF DIFFERENTIATION
Uninduced FBs were used as control samples in experiments included in paper I-IV. In
paper I uninduced PAs were also included as controls. When staining, cultures and tissue
samples were dehydrated through ethanol-xylene series, embedded in paraffin, sectioned and
mounted on microscopy slides. When analyzing adipogenic induction, endothelial induction,
chondrogenic induction in monolayer and respective reference cultures, cells were fixed in
paraformaldehyde (PFA) and analysis performed in the cell culture flasks.
Routine staining
Routine staining was performed in paper I-IV using previously described protocols; Oil
red O as an indicator of intracellular lipid accumulation (Kutt and Tsaltas 1959; Hausman
1981), Alcian blue stain as an indicator of sulphated GAG rich extracellular matrix (Lev and
Spicer 1964; Mason 1971), and von Kossa staining as an indicator of calcified extracellular
matrix (Bills, Eisenberg et al. 1974). Uptake of flurochrome-labelled LDL (Dil-Ac-LDL) was
used as an indicator of endothelial cell function (Voyta, Via et al. 1984). Oil red O staining in
paper I was quantified by examining randomly selected high power fields and counting
stained and unstained cells. Alcian blue and von Kossa staining in paper I was scored by
independent blinded observers judging the staining intensity on a four-degree scale. In paper
III hematoxylin and eosin was applied for routine stain of sections of bone (Ziegler 1944).
28
Immunohistochemistry
Immunohistochemistry (IHC) was performed in paper I-IV using primary antibodies
towards a number of antigens (table II). Briefly, slides were rehydrated and in the case of
chondrogenic and osteogenic induction antigens were retrieved by treatment with pepsin in
Tris-HCl. Slides were incubated in blocking serum (5% bovine serum albumin) for 45 min
followed by incubation for 1 h with primary antibodies followed by incubation for 1 h with
ALEXA-conjugated secondary antibodies. Sections were mounted using mounting medium
containing 4', 6-diamidino-2-phenylindole (DAPI). Control samples where primary antibodies
were omitted were included in all experiments to rule out unspecific binding of secondary
antibodies.
Table II. Antigens targeted in IHC.
Lineage
Antigen Function
Adipogenic
Perilipin Surface protein of intracellular lipid droplets in
adipocytes (Blanchette-Mackie, Dwyer et al.
1995)
Chondrogenic
Aggrecan Major proteoglycan in cartilage ECM (Roughley
2006)
Collagen II Major collagen in hyaline cartilage ECM (Eyre
2004)
Osteogenic
Osteonectin Proosteoblastic glycoprotein in bone vital in
mineralization (Young, Kerr et al. 1992)
Osteocalcin Proosteoblastic hormone from bone vital in
mineralization (Pagani, Francucci et al. 2005)
Endotheliogenic von Willebrandt factor Glycoprotein involved in hemostasis (Sadler
1998)
Endothelial nitric
oxide synthase
Enzyme in endothelial cells generating nitric
oxide (Dudzinski, Igarashi et al. 2006)
Alkaline phosphatase assay
Alkaline phosphatase (ALP) is an enzyme involved in the mineralization process. It has
the highest enzymatic activity in an alkaline environment. ALP is expressed in several human
tissues, but most prominent in bone liver and kidney. In bone it is expressed in osteoblasts
(Coleman 1992; Golub 1996). ALP activity was measured in paper I by adding paranitro-
phenylphosphate (pNPP) to cell cultures. pNPP is a substrate for alkaline phosphatase and is
cleaved into the yellow colored para-nitrophenol. The reaction can be quantified by using a
spectrophotometer measuring absorbance at a wavelength of 405 nm (Magnusson and Farley
2002).
29
Fluorescence in situ hybridization
To detect human cells present in the femoral defects in paper III, fluorescence in situ
hybridization (FISH) was performed using the Vysis Inc. Spectrum green female total human
genomic DNA probe and kit (Abbott Scandinavia AB, Sweden). Analysis was performed
following the manufacturer’s instructions. FISH is used to detect and localize the presence of
specific DNA sequences on chromosomes. We used FISH to detect human cells by specific
binding to fragments of the X chromosome (Schwartz, Depinet et al. 1997).
Full expression microarray
Microarray analysis was used in paper IV. Microarray has established itself as a valid
method for gene expression analysis. The advantage is the vast number of genes that can be
investigated in the same experimental run (Petersen and Kawasaki 2007). A prerequisite to a
successful microarray assay is a good yield of RNA in the starting material. We used the
NanoDrop 1000 (Thermo Scientific, MA) spectrophotometer to determine purity and yield of
RNA isolated from cell cultures. The instrument measures absorbance at the wavelengths 230,
260, 280 and 320 nm. 260 nm is specific for nucleic acids and 280 nm for proteins.
The basic principle of DNA microarray analysis is hybridization of samples to pre-
manufactured nucleic acid probes attached to a solid substrate. The sample is fragmented at
predicted sites that are taken into consideration when manufacturing the probes, labeled with
biotin conjugated to a fluorescent dye and hybridized to the probes (Mandruzzato 2007). After
hybridization of samples, chips are scanned and raw data is composed of an image of probe
cells with differing signal intensities, corresponding to the amount of sample hybridized to it
(Affymetrix.com). The microarray assay employed in this study was GeneChip Human Gene
1.0ST array (Affymetrix, UK). Gene expression was analysed at the final TP for each group
using 2 chips per group. The experimental procedure was performed according to
manufacturer’s protocol (Affymetrix GeneChip WT ST Labeling Assay Manual rev 5, P/N
701880, with addendum rev 1, P/N 702577) for 300 ng of starting RNA. After subsequent
RNA to cDNA conversion steps each sample was hybridized to a separate array and the arrays
were grouped according to culture medium. GeneChips were washed and stained in a Fluidics
Station 450 (Afymetrix, UK). Chips were scanned using GeneChip Scanner 3000 7G
(Affymetrix, UK). The raw data was imported into GeneSpring GX11 software (Agilent
Technologies Inc., CA) where multiple probe cells were normalized by quantile normalization
to obtain a gene-wise signal (Reimers 2010). Robust multichip average (RMA) was used for
the initial normalization of gene expression values. Quality control and array correlations
were checked. Probe sets with signals below the 20th
percentile of raw signal were filtered out
to minimise background as per manufacturer’s recommendation.
Entity significance was tested using ANOVA with Tukey’s Honestly Significant Difference
test and Benjamini-Hochberg corrections. Gene expression values were compared between
each induced culture group and the FB control group and reported as fold change.
Hierarchical clustering was created using Euclidian metric with single linkage in GeneSpring
GX11.
30
Quantative real-time polymerase chain reaction (qRT-PCR)
qRT-PCR was used in paper IV to validate selected genes in the microarray assay and to
study expression over time. Lineage specific genes were selected and analysed at 3 or 4 time
points. Gene expression assays, reagents and instrumentation were purchased from Applied
Biosystems (Foster City, CA). Two samples were taken from all time points. Samples were
analysed 4 times each for method control resulting in eight samples per time point.
Endogenous control for all assays was hydroxymethyl-bilane synthase (HMBS). The PCR
was performed by using TaqMan Fast 96-well plates in the HT7900 Fast Real-time PCR
machine. The number of cycles of the PCR run was set to 50. For data processing we used the
2-∆∆Ct
method. The formula is based on the generalization that one cycle difference represents
a doubling of product in the PCR (Fleige and Pfaffl 2006). Relative differences between
samples, meaning difference in number of cycles required to reach cycle threshold, are
assessed while relating the difference to firstly, an endogenous control, and then to an internal
calibrator that is the same for all assays (Livak and Schmittgen 2001). All samples are
normalized against the calibrator, in our case the untreated, control FBs. Analysis regarding
significance was performed by comparing medians in populations applying Kruskal-Wallis
with Dunn’s post test.
Flow cytometry
CD stands for cluster of differentiation, and the CD system is used to describe cell
surface proteins, allowing cells to be defined based on what molecules are present on their
surface; receptors, signaling peptides or cell adhesion molecules. The nomenclature was
originally developed in the eighties for describing leukocyte antigens, but since then, its use
has expanded and now includes many other cell types. Today there are over 350 CD-markers
described (www.hcdm.org).
In paper I FBs and sccFBs were analysed with regard to the expression of CD surface
markers. The CD markers chosen have previously been used to study mesenchymal stem cells
(table III) (Pittenger, Mackay et al. 1999; Zuk, Zhu et al. 2002; Liu, Akiyama et al. 2008). A
number of CD markers included are specific for hematopoietic cells and were used in order to
exclude any progenitor cells derived from circulating stem cell populations.
Analyses were performed using a BD LSR flow cytometer (Becton Dickinson, Denmark).
The mean fluorescence intensity was compared between CD-marker and the corresponding
isotype control and the fold increase was calculated. Furthermore, the percentages of
positively labeled FBs and sccFBs were compared. Normal FBs and three sccFB populations
from separate donors were analysed at 8 different occasions.
31
Table III. CD-markers used to analyse FBs and sscFBs
Antigen Protein Function
CD4 Co-receptor for T-cell receptor. Binds MHC II
CD8 Co-receptor for T-cell receptor. Binds MHC I
CD9 Motility related protein-1, tetraspanin 29 Adhesion
CD13 Alanine aminopeptidase Trimming of peptides
CD14 Co-receptor, Pattern recognition receptor Recognition of lipo-
polysaccaride
CD19 LEU-12, Lymphocyte surface antigen Co-receptor together with
CD21 and CD81
CD29 Integrin β1 Migration
CD34 Adhesion factor Attachment and migration of
hematopoietic stem cells
CD44 Hyaluronic acid receptor, Lymphocyte homing
receptor
Homing and migration of
stem cells
CD45 Leukocyte common antigen Involvement in TCR signaling
CD90 THY-1 Cell-cell and cell-matrix
interactions
CD105 SH2, Endoglin Component in TGF-β receptor
system
CD106 VCAM-1 Endothelial adhesion
molecule, binds VLA-4
MICROCARRIERS
Microcarriers were first introduced in the late sixties (van Wezel 1967) and since then a
wide range of microcarriers for tissue engineering applications has been developed.
Microcarriers may be manufactured from collagen, gelatin, dextran, plastic, glass or cellulose,
and may have porous or non porous structure. Porous microcarriers provide a large surface
area for cell growth and are easily handled in suspension and therefore offer a simple way to
fill a defect by injecting cell-seeded microcarriers (Malda and Frondoza 2006).
CultiSpher-GL, which was used in paper II and III is a macroporous porcine gelatine
microcarrier with a diameter of 166-370 µm, and average internal pore size of approximately
30 µm (figure 5) (Percell Biolytica AB, Sweden). This macroporous gelatin microcarrier offer
substantial surface area for cell adhesion and proliferation due to the porous structure, and is
easily degraded by the growing tissues.
32
CultiSpher microcarriers have been used as a biodegradable scaffold for guided tissue
regeneration and cell based therapies in vivo (Ohlson, Branscomb et al. 1994; Malda,
Kreijveld et al. 2003; Liu, Hafner et al. 2004; Stover, Bakay et al. 2005; Melero-Martin,
Dowling et al. 2006; Fernandes, Fernandes et al. 2007; Huss, Junker et al. 2007)
Microcarriers have also previously been used in bone tissue engineering applications (Malda
and Frondoza 2006; Tielens, Declercq et al. 2007).
Figure 5. CultiSpher-GL. Scanning electron
microscope image of CultiSpher-GL, a macroporous
porcine gelatine microcarrier with a diameter of 166-
370 µm, and average internal pore size of
approximately 30 µm.
Image captured by Sofia Pettersson.
SPINNER FLASK AND THREE-DIMENSIONAL CULTURE
There is a growing interest in using bioreactors when culturing tissues of the
musculoskeletal system. It is believed that growing conditions imitatating the natural
exposure to biomechanical forces of these tissues has a positive impact on tissue formation.
There are specifically designed bioreactors that transmit forces like shear, hydrostatic
pressure, compression, and combinations thereof (Abousleiman and Sikavitsas 2006; Schulz
and Bader 2007). In paper II and III cells were cultured on microcarriers in spinner flasks. A
spinner flask is a stirrer, with a magnetic rod inside the flask. Cells grown in this environment
are continuously exposed to fluid induced shear stress. Studies have shown positive effects of
fluid induced shear stress on osteoblast differentiation (Holtorf, Jansen et al. 2005; Datta,
Pham et al. 2006). To transfer FBs to spinner flasks, cells were enzymatically detached,
washed and resuspended together with microcarriers in FM in a Techne MCS 104S spinner
flask (figure 6A). FBs were seeded at approximately 18 x 106 cells/g microcarriers. The first
day cells were allowed to attach to microcarriers under intermittent agitation, and on the
second day agitation was adjusted to continuous stirring (20 RPM).
To create three-dimensional tissues in paper II human dermal FBs seeded on
microcarriers in spinner flasks were combined with platelet rich plasma (PRP) which was
used as a biological glue (Pettersson, Wettero et al. 2009). PRP was prepared from citrated
whole blood from healthy donors. Microcarriers and PRP were mixed 1:1 (vol:vol) and plated
at 400 µl in 12-well cell culture inserts with 3 µm pores (BD Biosciences, MA) (figure 6B).
33
Clotting was activated by adding 15 µl 0.5 M CaCl2 in each insert. Samples were allowed to
clot for 30 min before FM was added to each well and insert.
Figure 6. Cell culture for osteogenic and chondrogenic induction. Spinner flasks (A) and cell culture inserts
in 12-well plate (B) used in paper III.
FEMORAL GAP SURGICAL MODEL
The femoral gap surgical model has been used extensively to study long-bone repair
(Einhorn, Lane et al. 1984; Werntz, Lane et al. 1996; Bruder, Kurth et al. 1998). A segmental
gap is created in the femoral diaphysis and transplanted with biomaterial and/or cells. In
paper III this model was used to evaluate the survival of osteogenic induced human dermal
fibroblasts on microcarriers (O-FBs). 24 female athymic NIH nude rats (Taconic, NY) were
randomly divided into 6 groups, receiving different filling of the femoral defects (table IV).
Table IV. Filling of femoral defects in paper IV.
Group Number
of
animals
Filling Cells/Defect
1 4 400 µl NaCl 0.9 % 0
2 4 400 µl PRP 0
3 4 200 µl microcarriers (MCs)
200 ml PRP
0
4 4 200 µl human dermal fibroblasts on microcarriers (FBs)
200 µl PRP
6.0 x 105
5 4 200 µl human osteoblasts on microcarriers (OBs)
200 µl PRP
1.4 x 104
6 4 200 µl osteogenic induced human dermal fibroblasts on
microcarriers (O-FBs)
200 µl PRP
6.0 x 105
Total 24
34
In isoflurane anesthesia both femurs were exposed by a lateral approach. A custom made
ultra high molecular weight polyethylene fixation plate (20x4x3mm) was attached anteriorly
on the mid-femoral diaphysis and fixed with four 0.9 mm Kirschner wires and two 0.6 mm
cerklage wires. A 5 mm osteoperiostal mid-diaphyseal segment was removed using a high
speed osteotomy burr. Microcarriers and PRP were mixed 1:1 (vol:vol) and 400µl was
delivered to each femoral defect. In the first group 400 µl NaCl was put in each defect and
400 µl PRP in the second group. Defects were sealed by wrapping with cell free dermal tissue
which was closed by sutures (figure 7). Muscle fascia and skin was sutured. Weight bearing
was started immediately post-operative and analgesia was provided. At four weeks rats were
killed by CO2 inhalation. Immediately afterwards, both femurs were collected and fixed in 4
% PFA.
Figure 7. Femoral gap surgical model. (A) Perioperative picture showing the femur of an athymic NIH nude
rat exposed by a lateral approach. The polyethylene fixation plate has been attached anteriorly on the mid-
femoral diaphysis and fixed with four 0.9 mm Kirschner wires and two 0.6 mm cerklage wires. Thereafter the 5
mm long defect has been created by using a high speed osteotomy burr. Cell free dermal tissue is put under the
gap before transplantation. (B) Femurs collected from pilot operation of cadaver animal showing femurs with
fixation plate, and defects sealed with cell free dermal tissue.
35
RESULTS and DISCUSSION
PAPER I
After two weeks of induction in adipogenic medium FBs and PAs had adopted a rounded
shape and displayed intracellular droplets. These droplets accumulated the Oil red O stain
indicating intracellular lipid accumulation, and showed a positive perilipin IHC. Two weeks
after induction of FBs and PAs in chondrogenic medium, nodules and ridges with a high
density of cells and extracellular matrix was observed. This extracellular matrix displayed
positive Alcian blue stain indicating presence of sulphated glucosaminoglycans, and positive
IHC for aggrecan and collagen II. After two weeks of induction in osteogenic medium FBs
and PAs displayed clusters with high cell density and large amounts of extracellular matrix.
Spectrophotometric results showed an increase of alkaline phosphatase activity. After four
weeks of induction, FBs and PAs had formed dense extracellular matrix. This extracellular
matrix displayed positive von Kossa stain indicating calcified matrix, and positive IHC for
osteocalcin and osteonectin. Uninduced control cells did not stain positive in any
examinations showed above (figure 8).
Figure 8. Routine stain of induced FBs. Adipogenic induced FBs (A) and uninduced FBs (D) stained with Oil
red O. Arrows indicate intracellular lipid droplets. Arrowheads indicate cells without lipid accumulation.
Chondrogenic induced FBs in pellet (B) and uninduced FBs (E) stained with Alcian blue. Arrowheads indicate
glucosaminoglucan-rich extracellular matrix. Osteogenic induced FBs (C) and uninduced FBs (F) stained with
von Kossa. Arrows indicate mineralized extracellular matrix.
36
All sccFBs showed signs of chondrogenic and osteogenic differentiation, and five out of
seven showed signs of adipogenic differentiation. SccFBs showed statistically significant
higher fold increase of CD13 (p<0.01), CD29 (p<0.05), CD44 (p<0.005), CD 90 (p<0.01) and
CD105 (p<0.01) when compared to primary culture FBs. The percentage of positive cells as
compared to matched isotype controls was higher when staining for CD13 (p<0.01), CD44
(p<0.005) and CD105 (p<0.05). CD29 and CD90 did not show any statistically significant
difference between FBs and sccFBs when comparing percentages of positively labeled cells.
Several recent publications describe the inherent plasticity of human dermal FBs (Young,
Mancini et al. 1995; Toma, Akhavan et al. 2001; Bartsch, Yoo et al. 2005; Chen, Zhang et al.
2007; Takahashi, Tanabe et al. 2007; Lowry, Richter et al. 2008). The results presented in this
thesis further strengthen this notion. The differentiation capacity of human dermal FBs shown
by us has been consistent in repeated experiments. Human PAs were used as controls
regarding the differentiation potential (Zuk, Zhu et al. 2001). We find a similar differentiation
potential in FBs and PAs; accept for the adipogenic lineage, where PAs have a somewhat
higher efficiency of induction probably due to adipogenic priming. On the other hand,
observers´ scoring of chondrogenic and osteogenic induced FBs was higher than
corresponding PAs.
The phenomenon that causes a cell to change phenotype, and the underlying molecular
mechanisms, has been a focus of intense debate. One theory relies upon the fusion of two
different cell types to generate a new cell type (Eisenberg and Eisenberg 2003; Jin and
Greenberg 2003; Camargo, Chambers et al. 2004). We have used single cell cloning to
answer this question in our work, and these results indicate that fusion between different cell
types does not play a significant role in the phenotypical shift of FBs. Furthermore, the results
obtained using clonal analysis exclude the presence of contaminant tissue-specific progenitor
cells as an explanation of our findings. The flow cytometry analysis showed that cells did not
express hematopoietic stem cell markers, indicating that our cells are not derived from a
circulating stem cell population.
The flow cytometry of FBs and sccFBs showed expression of several MSC-specific
markers, but not CD106. MSCs have been defined as CD29+, CD44
+, CD90
+, CD105
+,
CD106+ and negative for hematopoietic markers CD14
-, CD31
-, CD34
-, CD45
- (Delorme,
Chateauvieux et al. 2006; Liu, Akiyama et al. 2008). Our findings are in line with results
from other groups. Lorentz et al. isolated FBs from juvenile foreskins with adipogenic and
osteogenic potential, which were CD14-, CD29
+, CD31
-, CD34
-, CD44
+, CD71
+, CD73
+,
CD90+, CD105
+, CD133
- and CD166
+ (Lorenz, Sicker et al. 2008). Zuk et al. have shown a
similar profile in stem cells isolated from human adipose tissue. These cells were induced
towards adipogenic, osteogenic, chondrogenic, myogenic and neurogenic lineages (Zuk, Zhu
et al. 2001; Zuk, Zhu et al. 2002).
Interestingly other groups show conflicting results, where FBs have no differentiation
capacity when subjected to induction media (Pittenger, Mackay et al. 1999; Wagner, Wein et
al. 2005). These studies have used FBs as negative controls compared to MSCs. These cells
are immortal cell-line FBs. Probably these cells are terminally differentiated, and have no
remaining plasticity as compared to the fresh isolated cells from dermal specimens used in
this thesis.
37
The fact that five of the seven sccFBs successfully differentiated towards three distinct
lineages implies that at least a subpopulation of the FBs is multipotent. This is in line with
publications exploring the inherent plasticity of dermal derived FBs while employing clonal
analyses (Toma, Akhavan et al. 2001; Bartsch, Yoo et al. 2005; Chen, Zhang et al. 2007).
Why do not all clones have the ability to differentiate into all lineages? There is probably a
sliding scale of ability for cells to differentiate. Some are multipotent and others less potent.
Our results indicate that the multilineage potential of dermal FBs is more uniform than results
presented in previous reports. Chen et al. showed by clonal analysis that 6.4 % of 47 single
cell clones tested were tripotent, 19.1 % were bipotent and 10.6 % were unipotent. One of the
three tripotent clones exhibited neurogenic and hepatogenic differentiation potential (Chen,
Zhang et al. 2007). It is difficult to answer why our clones have a higher differentiation
capacity. Our results are more in line with those from Bartsch et al. where all four
investigated FB clones were tripotent (adipogenic, osteogenic, myogenic). We have a higher
clonal yield because we have used micromanipulation whereas Chen et al. and Bartsch et al.
have used limited dilution and have a plating efficiency of approximately 0.5 %. Chen et al.
evaluate after 13 passages, and we analyze after 3-4 passages. Bartsch et al. preserve the skin
specimens at 4 °C for 6 days after harvest, to select only robust cell types for isolation.
FBs receiving chondrogenic induction medium were grown in both monolayers as well as
in pellets. Pellets were achieved by scraping the cells of the bottom of the culture flask after 3
weeks of monolayer induction, and thereafter cells were grown as a pellet one additional
week. The cartilage-like tissue formation occurred to the highest extent in areas with high cell
density. This implies the need for a three-dimensional structure in the case of chondrogenic
differentiation, which is consistent with reports from other observers (Chen, Zhang et al.
2007).
Conclusion
In conclusion, paper I show that normal human dermal FBs can differentiate to
adipocyte-, chondrocyte- and osteoblast-like cells in vitro when subjected to specific
induction media. Differentiation was of the same extent as in PAs. Furthermore, this was not
due to contamination of progenitor cells or cell fusion, as revealed by experiments performed
using single-cell cloned FBs.
PAPER II
By the employment of FB-seeded microcarriers and PRP, three-dimensional tissue
constructs were created in cell culture inserts. The constructs were induced towards
chondrogenic and osteogenic lineages using induction media. Osteogenic induction was also
performed using FB-seeded microcarriers cultured in spinner flasks. Samples were evaluated
after 4, 8 and 12 weeks.
Chondrogenic induction media stimulated formation of extensive amounts of ECM, most
prominent in the outer borders of the three-dimensional constructs. This matrix stained
positive using Alcian blue indicating presence of sulphated glucosaminoglycans, a stain
38
which intensified with time. Aggrecan IHC was positive, but there was no sign of collagen II
synthesis at either time point
Osteogenic induction medium stimulated extensive proliferation and a morphological
transformation of the FBs into cuboidal shaped cells and the formation of a dense ECM. After
4 weeks of incubation, the osteogenic induced cultures in inserts and spinner flasks displayed
positive von Kossa stain, which confirmed the ECM to be calcified. This matrix was most
prominent in samples from induced spinner flasks, and increased visually in amount after 8
and 12 weeks in both inserts and spinner flask cultures. Osteogenic induced cells showed
positive IHC for osteocalcin and osteonectin. Spectrophotometric results showed a
significantly higher increase of ALP activity in induced cells compared to uninduced cells.
Uninduced FBs on microcarriers showed a low ALP activity and a positive von Kossa
stain after 12 weeks in spinner flasks. All other uninduced controls were negative in all
aspects (figure 9).
Figure 9. Chondrogenic and osteogenic induced FBs in three-dimensional culture. Alcian blue staining of
chondrogenic induced FBs (A) and uninduced FBs (D) after 12 weeks of culture on microcarriers + PRP in
inserts. Von Kossa staining of osteogenic induced FBs (B) and uninduced FBs after 12 weeks of culture on
microcarriers + PRP in inserts, and osteogenic induced FBs (C) and uninduced FBs (F) after 12 weeks of culture
on microcarriers in spinner flasks.
Chondrogenic induced cells showed a rich ECM production, which had high amounts of
GAG, confirmed by Alcian blue stain and aggrecan IHC. There was no detectable production
of collagen II. This is interesting, because collagen II was found after induction of human
dermal FBs in monolayer (Junker, Sommar et al. 2010). The explanation for this could be that
39
cells have produced fibrocartilage and not hyaline cartilage in this three-dimensional model.
The general opinion though is that it is beneficial with three-dimensional culturing in the case
of chondrogenic differentiation (Chen, Zhang et al. 2007) because monolayer culturing of
chondrocytes causes dedifferentiation of the phenotype. Benya et al. showed that the
application of agarose gels caused reexpression of dedifferentiated chondrocytes (Benya and
Shaffer 1982). Microcarriers has also been used to grow chondrocytes and has been shown to
prevent dedifferentiation (Frondoza, Sohrabi et al. 1996) and to stimulate GAG synthesis
compared to monolayer growth (Malda, Kreijveld et al. 2003). In the studies by Frondoza et
al. and Malda et al., chondrocytes were cultured using spinner-flask culture. Our cultures
have been static after implantation into inserts. Maybe a positive effect on production of
collagen II could have been achieved by growing human dermal FBs in continuous stirring,
which has an effect on chondrogenesis (Smith, Donlon et al. 1995; Vunjak-Novakovic,
Obradovic et al. 2002). The culture of three-dimensional constructs created accumulation of
matrix in the superficial part, but less in the central part of the constructs. This is likely due to
insufficient diffusion of induction medium. A possible approach for achieving higher matrix
content in the inner parts of the constructs would be by applying perfusion of induction
medium through the constructs (Pazzano, Mercier et al. 2000; Vunjak-Novakovic, Obradovic
et al. 2002; Wendt, Stroebel et al. 2006). Wendt et al. found that constructs cultured under
continuous perfusion had a homogenous distribution of cells and ECM throughout the
constructs (Wendt, Stroebel et al. 2006). Dynamic compression may facilitate cartilage
formation, and could have been applied to the three-dimensional constructs in order to mimic
the natural environment of articular cartilage (Mauck, Soltz et al. 2000; Waldman, Spiteri et
al. 2004). The chondrogenic differentiation capacity of human dermal FBs shown by us is
consistant with results of other groups. Rabbit dermal FBs have been induced towards
chondrogenic lineage using pretreatment with IGF-1 and plating on aggrecan (French, Rose et
al. 2004). Chondrogenic induction of human dermal FBs has also been orchestrated by the
combination of human dermal FBs, collagen and demineralized bone powder (Mizuno and
Glowacki 1996).
The osteogenic differentiation of human dermal FBs on microcarriers produced high
amounts of calcified ECM. There was a minor positive von Kossa staining, as well as a minor
ALP activity in the uninduced FBs cultured 12 weeks in spinner flasks. Other studies have
shown positive effects of fluid induced shear stress on OB differentiation (Sikavitsas,
Bancroft et al. 2003; Holtorf, Jansen et al. 2005). These results suggest that shear stress by
itself has an impact on osteogenic differentiation, and may explain the findings of
mineralization in the uninduced FB culture in this study. However, unlike the osteogenic
induced FBs, uninduced constructs showed no positive staining for osteocalcin or osteonectin.
This distinction, along with the difference in calcification and ALP activity level between
induced and uninduced FBs, indicates that the stimulation factors in the induction media were
necessary for the ability of human dermal FBs to generate bone-like tissue in vitro.
Furthermore, it seems evident that shearing forces greatly enhances this process (Sikavitsas,
Bancroft et al. 2003; Holtorf, Jansen et al. 2005). The osteogenic differentiation capacity of
human dermal FBs shown by us is consistant with results of other groups. Genetically
modified human FBs; BMP-7 (Rutherford, Moalli et al. 2002) or RUNX2 (Phillips, Guldberg
et al. 2007) in combination with biomaterials have been shown to create bone like tissues.
40
Human PAs have been induced towards osteogenic lineage grown in tricalcium-phosphate
and cultured in osteogenic induction medium (Hattori, Masuoka et al. 2006).
The selection of microcarriers for construction of three-dimensional-tissues was based
upon collaboration with the manufacturer, and our and other group’s previous experience on
good cell adhesion, proliferation and expansion (Del Guerra, Bracci et al. 2001; Liu, Hafner
et al. 2004; Fernandes, Fernandes et al. 2007; Huss, Junker et al. 2007). Two weeks after
seeding on microcarriers our dermal FBs had expanded 4.5 times. The gelatin carrier is easy
to handle and is a degradable biomaterial. Microcarriers can be injected using syringe, which
could facilitate arthroscopical applications. Microcarriers have previously been used in
several cartilage and bone applications (Malda and Frondoza 2006). There is a diversity of
available biomaterials, and it is not possible to determine which biomaterial that would be the
“optimal” scaffold for bone and cartilage applications. Some find microcarriers questionable
as candidates for reconstruction of defects; the carriers might not stay at the desired location
within the defect after injection and may not provide any mechanical stability either. With
appropriate fixation of the fracture, mechanical stability should not be a problem, although
load-bearing should be avoided. Some of the common osteoconductive biomaterials,
hydroxyapatite and tricalcium phosphate add more stability, but are on the other hand more
difficult for cell adhesion, and they are not totally degraded which prevents tissue integration
(Hannouche, Petite et al. 2001).
PRP was used as a blood-derived biological glue to encapsulate cell-seeded microcarriers
(Pettersson, Wettero et al. 2009). The autologous source of the plasma makes it a strong
candidate for clinical purposes. PRP contains multiple growth factors, such as TGF-β, PDGF,
IGF-1 and epidermal growth factor (EGF), and these growth factors induce biological changes
in cell proliferation and matrix metabolism of a variety of connective tissues (Anitua, Andia et
al. 2004). Beneficial effects from platelets used as glue in tissue engineering approaches have
been reported. PRP stimulates cell proliferation and matrix metabolism in CCs (Akeda, An et
al. 2006) and OBs (Slapnicka, Fassmann et al. 2008) when measured after 72h. However this
effect is probably time dependent, Thorwarth et al. showed some effect of PRP on bone
regeneration in vivo after 2 weeks but no long-term effects (4, 12 and 26 weeks) (Thorwarth,
Wehrhan et al. 2006). Our results were analysed at 4, 8 and 12 weeks. Growth factors from
PRP probably have positive effects on initial cellular adhesion and proliferation, but no effect
on long time differentiation results. The fact that our results improve with time further
confirm that differentiation is due to the soluble factors used, and not PRP.
Conclusion
In conclusion, paper II show that three-dimensional constructs of normal human dermal
FBs seeded on microcarriers in combination with PRP create cartilage- and bone-like tissue
upon induction with chondrogenic or osteogenic induction media. Furthermore, fluid induced
shear stress had positive effects on osteogenic differentiation.
41
PAPER III
Fibroblasts (FBs), osteoblasts (OBs) and osteogenic induced fibroblasts (O-FBs) on
microcarriers were respectively transplanted in combination with PRP into femoral defects in
nude rats. Control defects were transplanted with NaCl, PRP or acellullar microcarriers/PRP.
Radiographs taken four weeks post surgery did not reveal detectable callus formation in any
of the experimental groups. In defects transplanted with O-FB seeded microcarriers, dense
cell clusters with large amounts of extracellular matrix partially filled seven out of eight gaps
at 4 weeks (figure 10). These clusters showed a highly positive staining for osteocalcin and
osteonectin. Cell clusters were not present in the gaps of the other groups. In defects
transplanted with OB seeded microcarriers, FB seeded microcarriers or acellular microcarriers
HTX stain showed a randomly dispersed cell/matrix tissue with partially degraded
microcarriers. No immunohistochemical staining was observed in the acellular microcarriers,
while group FB and OB stained moderately positive for osteocalcin and osteonectin. This
stain was very discrete compared to the highly positive IHC-stain observed in the group
seeded with O-FBs. HTX stain revealed fibrous tissue in gaps filled with PRP and NaCl.
These groups displayed negative IHC stain regarding osteocalcin and osteonectin.
FISH analysis displayed viable human cells in defects filled with cell-seeded
microcarriers confirming survival of transplanted cells in the gaps.
Figure 10. HTX stain of osteogenic induced FBs in vivo. Panels displaying HTX-stained sections 4 weeks
after transplantation of femoral gaps with osteogenic induced FBs on microcarriers + PRP (A+B). Note the
dense cell clusters marked with asterisks. Bar = 1 mm (A), 0.4 mm (B).
This study focused on the survival of osteogenic induced human dermal FBs in vivo,
using a femoral gap surgical model in athymic rats. The defects transplanted with O-FB
seeded microcarriers displayed cell clusters containing large amounts of extracellular matrix
and were labeled positively using antibodies towards osteocalcin and osteonectin. These cell
clusters partially filled seven out of eight gaps at four weeks. The number of animals is too
small to statistically conclude the beneficial effects of the O-FBs on microcarriers, but it is a
probable effect of the osteogenic induction as cell clusters were not present in the gaps of the
42
other groups. This suggests unique properties of the osteogenic induced FBs. It should be
noted though that FBs and OBs used only come from one single donor, which makes general
conclusions hazardous. To overcome this, cells from several donors could be used, but this
dramatically increases the amount of animals needed. Furthermore, previous experiments in
vitro using dermal FBs have shown similar differentiation pattern using several donors
(Bartsch, Yoo et al. 2005; Richardson, Arnott et al. 2005; Junker, Sommar et al. 2010).
The cell clusters could arguably be pre-fabricated in the spinner flask, and transplanted as
ready-made clusters into the defects. Clinically this is of minor importance; the important
finding is that viable human cells were detected in the defects 4 weeks post surgery, and that
these cells display osteogenic properties. This suggests a favorable milieu in the defect, which
has maintained the osteogenic induced FBs in an osteogenic differentiated state over time
despite omission of osteogenic induction medium after transplantation into the defects, and
confirms the phenotypical stability after in vitro induction. The phenotypical stability in vivo
was also shown by Bartsch et al. when transplanting osteogenic differentiated FBs
subcutaneously in athymic mice (Bartsch, Yoo et al. 2005).
There was a difference in cell numbers in the gaps between study groups. 6.0 x 105 O-
FBs and FBs respectively, but only 1.4 x 104 OBs were implanted per gap. This is due to
different rates of proliferation. The same amount of cells were seeded on microcarriers in all
three groups, but because of the much higher proliferation capacity of FBs and O-FBs a much
higher cell count is achieved in these groups before seeding of the gaps. This could have been
corrected by applying an initial higher seeding density in OBs. This though reflects the
clinical situation, where OBs proliferate slowly, whereas dermal FBs show excellent in vitro
features.
The CT-investigation at 4 weeks showed non-union in all defects. To be able to show
potential healing of defects longer culture periods must be applied. We had some problems
with rigidity of fixation, why a later time point using this method is questionable. One way to
overcome the effects of mechanic forces is to use a cranial defect model (Cowan, Shi et al.
2004; Hirata, Mizuno et al. 2007).
Conclusion
In conclusion, paper III show that osteogenic induced FBs transplanted to an in vivo
environment survive in this new niche and maintain their induced osteogenic properties
confirming the phenotypical stability 4 weeks after in vitro induction. CT investigation at four
weeks showed non-union of defects.
PAPER IV
Human dermal FBs were induced towards adipogenic-, chondrogenic- endotheliogenic
and osteogenic phenotypes. The gene expression profile of these cells was compared to
normal adipocytes, chondrocytes, endothelial cells and osteoblasts by employing genome
wide expression analysis using microarray technology. Selected gene expression was also
evaluated over time using real-time PCR. Several master regulatory genes important for
43
lineage commitment, as well as phenotypically relevant genes, were found regulated in the
respective induced cultures.
C/EBPs and PPARγ2 are known as master regulatory factors for adipocyte differentiation
involved in terminal differentiation by their subsequent activation of adipocyte-specific genes
(Rangwala and Lazar 2000; Rosen and Spiegelman 2000; Gregoire 2001). We detected up-
regulation of PPARG in adipogenic induced FBs, but only the delta-form of CEBP. Exposure
of PAs to adipogenic induction medium induces expression of CEBPB and CEBPD, which in
turn activate
PPARG2 and CEBPA (Rosen and Spiegelman 2000). Absent CEBPA,
ADD1/SREB-1c, CREB and GLUT4 expression in our dermal FBs may indicate that the
adipogenesis is not completed.
The transcription factor SOX9 is considered a master regulator of cartilage. SOX5 and
SOX6 are co-expressed with SOX9 and required for activation of the COL2A1 gene (Akiyama,
Chaboissier et al. 2002). We find almost no changes in the master regulators of
chondrogenesis accept for NKX3-2 in the microarray. In contrast SOX9 showed 49-fold up-
regulation when analysed with qRT-PCR. COL2A1 and ACAN were also found up-regulated
in the qRT-PCR, but only COL2A1 in the microarray. The CCs showed signs of
dedifferentiation/down-regulation of several lineage specific genes. This has been described
in monolayer chondrocyte cultures (Schnabel, Marlovits et al. 2002).
In endothelial cells or progenitors, endothelial-specific gene expression is probably
controlled combinatorially by multiple transcription factors. TAL1, GATA2, the FOX-family
and ETS transcription factors are among those that have been suggested to control
differentiation (De Val and Black 2009). Some report that HOXA9 is the master regulator of
endothelial differentiation (Rossig, Urbich et al. 2005). We find low changes of GATA2 and
HOX9, but a higher up-regulation of ETV1. The microarray showed that several important
genes are up-regulated, but CDH5, KDR, VWF were only up-regulated in HUVECs. In
contrast all these three genes were up-regulated in E-FBs when analysed with qRT-PCR, but
to a lower extent than in HUVECs.
During osteoblast differentiation, RUNX2 and osterix/SP7 play essential roles in the
commitment of pluripotent mesenchymal cells to the osteoblastic lineage (Komori 2010).
RUNX2 is essential for the commitment of the osteoblast lineage. We find up-regulation of
RUNX2, but not SP7. There is a moderate up-regulation of several genes coding for important
bone related proteins. A similarity between FBs and OBs has previously been suggested. In
cell culture, OBs are nearly indistinguishable from FBs. All the genes expressed in FBs are
also expressed in OBs, and conversely, only two OB-specific transcripts have
been identified:
one encoding RUNX2, and osteocalcin. Thus, genetically, the osteoblast
can be viewed as a
sophisticated fibroblast (Ducy, Schinke et al. 2000). This may explain the moderate changes
of gene expression in qRT-PCR and microarray of O-FBs and OBs compared to FBs.
There is a very complicated interplay between different transcription factors, and usually
there is a combination of inhibitory and enhancing signals. Several of the master regulatory
factors presented above have different effects on different lineages. Transcription factors
controlling adipogenesis have been shown to inhibit osteoblastogenesis, and vice versa.
PPAR-γ is proadipogenic, but also has an antiosteoblastogenic role. For example, endogenous
ligands for PPARγ, such as thiazolidinediones, promote adipogenesis and inhibit
osteoblastogenesis in bone marrow stem cells (Ali, Weinstein et al. 2005). Further on, early
44
stages of osteoblastogenesis are regulated by several WNT ligands. In adipogenesis, WNT
ligands act as inhibitors of adipocyte differentiation (Bowers and Lane 2008). C/EBPβ and γ
may inhibit chondrocyte differentiation by blocking transcription of COL2A1, ACAN, and
other cartilage-specific genes by direct or indirect mechanisms (Okazaki, Li et al. 2002) and
WNT signaling suppresses chondrocyte differentiation in early chondroprogenitors (Goldring,
Tsuchimochi et al. 2006). PREF1 is highly expressed in PAs, and is an inhibitor of adipocyte
differentiation which activates the MEK/ERK pathway and increases SOX9 which blocks
transcription of CEBPB and D (Sul 2009). On the other hand, SOX9 is the master regulator of
chondrogenesis (Akiyama 2008). WNTs do also have effects on endotheliogenesis. WNT1
has been suggested important for the formation of progenitor cells for the hematopoietic and
vascular systems (Woll, Morris et al. 2008). FOXO1 has a regulatory influence on
endotheliogenesis, and is required for vascular development, but is at the same time also a
positive regulator of bone formation (De Val and Black 2009; Rached, Kode et al. 2010).
These described genes are only a small selection from the plethora of regulatory mechanisms
defining phenotypical determination. It is not difficult to imagine that external stimulation
with hormones and other soluble factors may influence this delicate network of regulatory
genes.
The microarray and PCR showed up-regulation of several lineage specific genes in the
induced cultures, but also several genes with minor or no up-regulation when compared to
respective reference cultures. This shows that the induction is turning the cells into a different
phenotypical direction, but that these induced cells do not have the exact same expression
pattern as reference cells. When hierarchical clustering (figure 11) was performed based on
gene expression differences in groups compared to FBs, the adipogenic and osteogenic
lineages showed close relation. This is interesting as discussed above where transcription
factors seem to induce either towards adipogenic or osteogenic lineages (Takada, Kouzmenko
et al. 2009). Even though the chondrogenic and osteogenic lineages are developmentally
much related, for instance the close relationship during endochondral ossification, they are not
closely related when clustered hierarchically based on gene differences compared to FBs. This
might indicate that OBs are not generated by chondroblasts, but that these are two separate
populations after leaving the osteochondro progenitor stage (Karsenty 2008; Hartmann 2009).
Figure 11. Hierarchical clustering. Hierarchical clustering of microarray groupings using a sub-set of 1268
genes found significantly different in all groups compared to control FBs. The clustering displays the
relationships/differences between groups.
45
In general, we found very low fold changes in the microarray assays. But these values are
probably underestimated when compared to the more sensitive qRT-PCR. In all
differentiational groups we could detect higher values in the qRT-PCR when compared to
microarray; FABP4 (adipogenic lineage) showed 3.6 up-regulation in the microarray, but a
162-fold change in the qRT-PCR in adipogenic induced fibroblasts (A-FBs) compared to FBs,
COL2A1 (chondrogenic induced) 1.6 vs. 42. CDH5 (endotheliogenic induced) 1.0 vs. 135,
and osteonectin (osteogenic induced) -1.1 vs. 1.1. Therefore we have chosen not to filter the
microarray results based on fold-change thresholds; instead we have explored specific sets of
genes known to be important in the reference phenotypes. The results from qRT-PCR come
from selected RNA sequences which are amplified 50 times, compared to no amplification of
RNA in the microarray. This results in an inability in microarray to detect differences in low
expressed genes (Wren and Conway 2006). A vast majority, approximately 70% of all
protein-coding human genes are expressed with less than one transcript per cell on average
(Kuznetsov, Knott et al. 2002). Furthermore FBs are well known to have a very unspecific
gene expression, and do express several master regulators. Many of these master regulating
proteins need further phosphorylation, cleavage, modification or degradation to become
activated and achieve their lineage determinant properties. Modified proteins cannot be
detected by microarray assay (Wren and Conway 2006). Therefore it could be interesting to
use immunoprecipitation and western blot to study phosphorylation of expressed master
regulators to further confirm the lineage determinant effects of induction media.
Conclusion
In conclusion, paper IV show that several master regulatory genes important for lineage
commitment, as well as phenotypically relevant genes, are found regulated in FBs induced
towards adipogenic, chondrogenic, endotheliogenic and osteogenic phenotypes.
GENERAL DISCUSSION
All cells have the same DNA, so the different cell types are defined by their particular
pattern of regulated gene expression. Cellular differentiation involves a switch from one
pattern of gene expression to another (Ben-Tabou de-Leon and Davidson 2007). One vital
component in the differentiation process is the expression of different transcription factors.
These are proteins with different modes of action that operate at various levels to facilitate
and control the production of RNA. There may be as many as 3,000 transcription factors in
humans (Lander, Linton et al. 2001). Over the last decades, many factors have been identified
that regulate cellular differentiation. This process has had a number of pioneering events
beginning with Gurdon and nuclear transfer (Gurdon, Elsdale et al. 1958), the successful
cloning of Dolly (Wilmut, Schnieke et al. 1997), the identification of master regulators of
ESCs (Li 2010), and finally the creation of iPSCs (Takahashi and Yamanaka 2006;
Takahashi, Tanabe et al. 2007; Lowry, Richter et al. 2008) confirming that somatic cells
(fibroblasts) can be reprogrammed back to the totipotent zygotic state by defined cellular
factors. We have explored the plasticity of human dermal FBs using soluable factors, and
found that FBs are not “terminally differentiated” as previously believed, but have a certain
46
differentiation potential upon subjection to specific induction media. Several transcription
factors and lineage specific genes are regulated by this induction.
Cellular phenotypes are defined by their particular pattern of regulated gene expression,
and the differentiation process involves a change of gene expression. We have used defined
soluble factors to influence the differentiation process (Jaiswal, Haynesworth et al. 1997;
Johnstone, Hering et al. 1998; Pittenger, Mackay et al. 1999). The induction media used have
been described by other groups, and were initially used for differentiation of MSCs but have
also been used for other adult stem cell populations (Zuk, Zhu et al. 2001). These media
contain potent hormones like insulin and dexamethasone, growth factors like TGF-β, or
chemical substances like indomethacin or ascorbic acid. These are all used in various clinical
applications, and therefore considered safe. The levels we use are several times higher than
therapeutic levels used, but this stimulation ends after the in vitro phase, and should not have
any major influence in vivo. Several groups use genetic modification instead of soluble
factors. The highly potent iPSCs are generated using viral vectors (Takahashi and Yamanaka
2006), but the safety of using these modified cells in clinical applications is a matter of
extensive debate. Retroviruses tend to insert into active genes, and it has been suggested that
their use may increase the risk of cancer. Adenoviruses may elicit a strong immune response
directed against the engineered cells (Young, Searle et al. 2006; Conrad, Gupta et al. 2007).
We find that medium induction via soluble factors serves as a viable alternative to genetic
alteration of transplanted cells.
The cells used have been isolated from tissue specimens from human dermis. The cells
have after subcultivation displayed a typical fibroblastic morphology, with spindle-shaped
cells in parallel clusters. The rapid growth of FBs creates a selection for this cell type. Other
components of skin like adipocytes, endothelial cells or neural cells are negatively selected as
these cells need different optimal growth environments. The use of single cell cloning
excludes phenotypical differentiation into other lineages by contaminant cells. Flow
cytometry data showed no expression of hematopoietic stem cell markers, indicating that the
cell population used is not derived from a circulating stem cell pool. Furthermore, there was
no difference in differentiation or growth pattern between single cell clones or the unrefined
FB populations, indicating that the differentiation potential might be a general inherent ability
in all FBs or subpopulations thereof. This was confirmed by sorting FBs using CD13, CD44,
CD 90 and CD105, four distinct CD markers up-regulated on sscFBs compared to FBs. This
refined population did not display a differentiation potential surpassing that of the unrefined
FBs (data not shown), wich is consistant with results by Bartch et al. (Bartsch, Yoo et al.
2005).
Cells isolated from donors of different age or sex did not seem to have differences in
growth rate or differentiation (data not shown). Clones or unrefined FBs have not been
cultured more than 40 passages, so it is not possible to know if these cells enter a senescence
phase or not. Other studies with human dermal FB populations have shown that isolated cells
could be expanded for over 100 population doublings with retention of their chromosomal
complement and multipotency (Bartsch, Yoo et al. 2005).
We have shown that human dermal FBs can be differentiated towards 4 different
lineages. This means that our cells are multipotent. We have only shown differentiation in the
mesodermal germ layer, and no transdifferentiation. In fact from a therapeutic point of view
47
pluripotency is not a desirable criterion for cell replacement strategies. A multipotent or
unipotent state might be sufficient and much safer for clinical applications compared to ESCs
or iPSCs. The results presented in this thesis and by other research groups indicate that the
morphology, differentiation patterns and the flow cytometic profiles of human dermal FBs,
adipose derived stem cells and bone marrow derived MSCs are quite similar (Zuk, Zhu et al.
2002; Delorme, Chateauvieux et al. 2006; Lorenz, Sicker et al. 2008). Some authors go as far
as describing both mesenchymal stem cells and fibroblasts as members of the same family
(Haniffa, Collin et al. 2009). However, from a clinical point of view, knowledge of the exact
source is secondary to the fact that it is possible to change the phenotype of human dermal
FBs in vitro.
FUTURE PERSPECTIVES
Further studies investigating additional features of adipogenic, chondrogenic,
endotheliogenic and osteogenic differentiated FBs are needed before the employment of these
cells in clinical applications. The stability of phenotypic shift after induction must be further
assessed, to investigate if this shift remains after induction is withheld, or if cells return to a
FB like phenotype, as phenotypic stability is a requisite for succeeding in clinical
applications.
Further differentiation abilities should be investigated. There are reports about
transdifferentiation of dermal FBs, and future studies exploring the differentiation spanning
across germ layers, for instance neurogenic differentiation would be interesting. Assessment
of further lineages of differentiation is necessary to explore the full potency of dermal FBs.
Application of differentiated cells in vivo will be necessary to study the therapeutic
qualities of these cells. The osteogenic lineage was tested in this thesis, but need long-term
follow-up. Other applications could be the employment of chondrogenic induced cells in an
articular cartilage defect model, or the subcutaneous injection of adipose induced cells to fill
voids.
48
49
CONCLUSIONS
In summary, the findings presented in this thesis suggest an inherent differentiation
potential in human dermal FBs. We have shown that human dermal FBs upon stimulation
with specific induction media differentiate towards adipogenic, chondrogenic and osteogenic
phenotypes. Chondrogenic and osteogenic induced cells on biomaterials create three-
dimensional tissues with cartilage and bonelike properties. The bone like tissue survives in an
in vivo environment, and retains its induced osteogenic character. After the induction towards
adipogenic, chondrogenic, endotheliogenic and osteogenic lineages, several master regulatory
genes important for lineage commitment, as well as phenotypically relevant genes were found
regulated as compared to reference cultures. Results presented in this thesis indicate that
human dermal FBs might be a suitable cell source in tissue engineering applications.
50
51
ACKNOWLEDGEMENTS
There are several persons which have helped and supported me during the years of
running this project, to which I would like to express my sincere gratitude;
Gunnar Kratz, for introducing me to the world of science and plastic surgery. You have lots
of energy and enthusiasm. Thanks for all these years!
Thomas Hansson. My supervisor in hand-surgery and co-supervisor for this thesis. Thanks
for advice and fruitful discussions during this process.
Hans Johnson, for good times in Linköping and Bergen. You have taught me to strive
towards a goal. Thank you also for all effort invested in our Bergen stay.
Johan Junker. You started out as my research student, but have already defended your thesis.
Thank you for all assistance on my way to the doctorial grade. You are a really good friend,
always willing to help with computer problems, always willing to carry my furniture when I
have been moving. Always up for a beer. Wish you all the best with your career at Harvard.
Hopefully we will work together again!
Jonathan Rakar. Thank you for intellectual conversations and all work you have put into
paper IV, and in proofreading this book.
Charlotte Ness, for all help with projects in Linköping and Bergen. Thank you also for
coaching in running.
Susanna Lönnqvist, for fun times at the lab running microarrays.
Sofia Pettersson, for fruitful collaboration in paper III.
Mårten Skog, for your assistance with flow cytometry.
Fredrik Huss, for valuable clinical and scientific discussions, and discussions about life in
general. Looking forward to future celebrated articles.
Anita Lönn. You were the first person I met on a lab as a student. Thanks for all help starting
up this project and for teaching me laboratory skills!
Kristina Briheim, for assistance in the lab. Always smiling when I came up with more work!
Eivind Strandenes, for excellent surgical skills.
52
Colleagues at the Department of Plastic and Reconstructive Surgery, Karolinska
University Hospital, Stockholm, for your way of welcoming me to the clinic and supporting
me in my development.
Especially;
Kalle Lundgren. My roommate at Karolinska. Thank you for your never ending energy and
strive, for all help in the scientific field, and for being a dedicated friend. Thank you also for
all help in proofreading and making drawings for this thesis.
Martin Halle, for all late flap nights and dissociated discussions. I really enjoy working with
you!
Jessica Gahm, for fun co-surgery and for all valuable advice about the choices of life.
Ann-Charlott Docherty Skogh. My supervisor in plastic surgery. Thank you for all time and
effort you have invested in my surgical training. Thank you also for sailing and skiing events!
Filip Farnebo, for research coaching in tuff times.
Johan Rinder. My boss. Thanks for your confidence in my scientific and clinical
development.
Marie Wickman, for believing in my abilities and giving me the chance to work with plastic
surgery.
Colleagues at the Department of Hand- and Plastic Surgery, University Hospital
Linköping, for support in the “early years”.
Parents and sisters, for accepting that Christmas and Easter is lab time.
Marie Sommar. My beloved wife! Without you, your never ending love, positivism and
energy I would not have come this far.
Ingrid Sommar, for enlightening my world and for speeding up the process of writing this
thesis.
53
REFERENCES
Abousleiman, RI and Sikavitsas, VI (2006). Bioreactors for tissues of the musculoskeletal system. Adv Exp Med
Biol 585: 243-59.
Affymetrix.com. "Technical note " Retrieved 12.05, 2010, from www.affymetrix.com.
Ahmed, W, Ziouzenkova, O, Brown, J, Devchand, P, Francis, S, Kadakia, M, Kanda, T, Orasanu, G, Sharlach,
M, Zandbergen, F, et al. (2007). PPARs and their metabolic modulation: new mechanisms for
transcriptional regulation? J Intern Med 262(2): 184-98.
Akeda, K, An, HS, Okuma, M, Attawia, M, Miyamoto, K, Thonar, EJ, Lenz, ME, Sah, RL and Masuda, K
(2006). Platelet-rich plasma stimulates porcine articular chondrocyte proliferation and matrix
biosynthesis. Osteoarthritis Cartilage 14(12): 1272-80.
Akiyama, H (2008). Control of chondrogenesis by the transcription factor Sox9. Mod Rheumatol 18(3): 213-9.
Akiyama, H, Chaboissier, MC, Martin, JF, Schedl, A and de Crombrugghe, B (2002). The transcription factor
Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required
for expression of Sox5 and Sox6. Genes Dev 16(21): 2813-28.
Ali, AA, Weinstein, RS, Stewart, SA, Parfitt, AM, Manolagas, SC and Jilka, RL (2005). Rosiglitazone causes
bone loss in mice by suppressing osteoblast differentiation and bone formation. Endocrinology 146(3):
1226-35.
Almarza, AJ and Athanasiou, KA (2004). Design characteristics for the tissue engineering of cartilaginous
tissues. Ann Biomed Eng 32(1): 2-17.
Anitua, E, Andia, I, Ardanza, B, Nurden, P and Nurden, AT (2004). Autologous platelets as a source of proteins
for healing and tissue regeneration. Thromb Haemost 91(1): 4-15.
Atala, A, Bauer, SB, Soker, S, Yoo, JJ and Retik, AB (2006). Tissue-engineered autologous bladders for
patients needing cystoplasty. Lancet 367(9518): 1241-6.
Atkins, GB and Jain, MK (2007). Role of Kruppel-like transcription factors in endothelial biology. Circ Res
100(12): 1686-95.
Auxenfans, C, Fradette, J, Lequeux, C, Germain, L, Kinikoglu, B, Bechetoille, N, Braye, F, Auger, FA and
Damour, O (2009). Evolution of three dimensional skin equivalent models reconstructed in vitro by
tissue engineering. Eur J Dermatol 19(2): 107-13.
Avram, AS, Avram, MM and James, WD (2005). Subcutaneous fat in normal and diseased states: 2. Anatomy
and physiology of white and brown adipose tissue. J Am Acad Dermatol 53(4): 671-83.
Axelrad, TW, Kakar, S and Einhorn, TA (2007). New technologies for the enhancement of skeletal repair. Injury
38 Suppl 1: S49-62.
Bachetti, T and Morbidelli, L (2000). Endothelial cells in culture: a model for studying vascular functions.
Pharmacol Res 42(1): 9-19.
Badylak, SF, Freytes, DO and Gilbert, TW (2009). Extracellular matrix as a biological scaffold material:
Structure and function. Acta Biomater 5(1): 1-13.
Bajada, S, Mazakova, I, Richardson, JB and Ashammakhi, N (2008). Updates on stem cells and their
applications in regenerative medicine. J Tissue Eng Regen Med 2(4): 169-83.
Bartsch, G, Yoo, JJ, De Coppi, P, Siddiqui, MM, Schuch, G, Pohl, HG, Fuhr, J, Perin, L, Soker, S and Atala, A
(2005). Propagation, expansion, and multilineage differentiation of human somatic stem cells from
dermal progenitors. Stem Cells Dev 14(3): 337-48.
Beattie, GM, Lopez, AD, Bucay, N, Hinton, A, Firpo, MT, King, CC and Hayek, A (2005). Activin A maintains
pluripotency of human embryonic stem cells in the absence of feeder layers. Stem Cells 23(4): 489-95.
Behrens, P, Bitter, T, Kurz, B and Russlies, M (2006). Matrix-associated autologous chondrocyte
transplantation/implantation (MACT/MACI)--5-year follow-up. Knee 13(3): 194-202.
Ben-Tabou de-Leon, S and Davidson, EH (2007). Gene regulation: gene control network in development. Annu
Rev Biophys Biomol Struct 36: 191.
Benya, PD and Shaffer, JD (1982). Dedifferentiated chondrocytes reexpress the differentiated collagen
phenotype when cultured in agarose gels. Cell 30(1): 215-24.
Bills, CE, Eisenberg, H and Pallante, SL (1974). Complexes of organic acids with calcium phosphate: the Von
Kossa stain as a clue to the composition of bone mineral. Johns Hopkins Med J 128(4): 194-207.
Blanchette-Mackie, EJ, Dwyer, NK, Barber, T, Coxey, RA, Takeda, T, Rondinone, CM, Theodorakis, JL,
Greenberg, AS and Londos, C (1995). Perilipin is located on the surface layer of intracellular lipid
droplets in adipocytes. J Lipid Res 36(6): 1211-26.
Bluher, S, Kratzsch, J and Kiess, W (2005). Insulin-like growth factor I, growth hormone and insulin in white
adipose tissue. Best Pract Res Clin Endocrinol Metab 19(4): 577-87.
Bobick, BE, Chen, FH, Le, AM and Tuan, RS (2009). Regulation of the chondrogenic phenotype in culture.
Birth Defects Res C Embryo Today 87(4): 351-71.
54
Boheler, KR (2009). Stem cell pluripotency: a cellular trait that depends on transcription factors, chromatin
state and a checkpoint deficient cell cycle. J Cell Physiol 221(1): 10-7.
Bowers, RR and Lane, MD (2008). Wnt signaling and adipocyte lineage commitment. Cell Cycle 7(9): 1191-6.
Brittberg, M, Lindahl, A, Nilsson, A, Ohlsson, C, Isaksson, O and Peterson, L (1994). Treatment of deep
cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 331(14): 889-
95.
Bruder, SP, Jaiswal, N and Haynesworth, SE (1997). Growth kinetics, self-renewal, and the osteogenic potential
of purified human mesenchymal stem cells during extensive subcultivation and following
cryopreservation. J Cell Biochem 64(2): 278-94.
Bruder, SP, Kurth, AA, Shea, M, Hayes, WC, Jaiswal, N and Kadiyala, S (1998). Bone regeneration by
implantation of purified, culture-expanded human mesenchymal stem cells. J Orthop Res 16(2): 155-62.
Camargo, FD, Chambers, SM and Goodell, MA (2004). Stem cell plasticity: from transdifferentiation to
macrophage fusion. Cell Prolif 37(1): 55-65.
Caplan, AI (1991). Mesenchymal stem cells. J Orthop Res 9(5): 641-50.
Caplan, AI (2005). Review: mesenchymal stem cells: cell-based reconstructive therapy in orthopedics. Tissue
Eng 11(7-8): 1198-211.
Chen, FG, Zhang, WJ, Bi, D, Liu, W, Wei, X, Chen, FF, Zhu, L, Cui, L and Cao, Y (2007). Clonal analysis of
nestin(-) vimentin(+) multipotent fibroblasts isolated from human dermis. J Cell Sci 120(Pt 16): 2875-
83.
Chimal-Monroy, J, Rodriguez-Leon, J, Montero, JA, Ganan, Y, Macias, D, Merino, R and Hurle, JM (2003).
Analysis of the molecular cascade responsible for mesodermal limb chondrogenesis: Sox genes and
BMP signaling. Dev Biol 257(2): 292-301.
Christoffersen, CT, Tornqvist, H, Vlahos, CJ, Bucchini, D, Jami, J, De Meyts, P and Joshi, RL (1998). Insulin
and insulin-like growth factor-I receptor mediated differentiation of 3T3-F442A cells into adipocytes:
effect of PI 3-kinase inhibition. Biochem Biophys Res Commun 246(2): 426-30.
Chung, C and Burdick, JA (2008). Engineering cartilage tissue. Adv Drug Deliv Rev 60(2): 243-62.
Chung, CH, Golub, EE, Forbes, E, Tokuoka, T and Shapiro, IM (1992). Mechanism of action of beta-
glycerophosphate on bone cell mineralization. Calcif Tissue Int 51(4): 305-11.
Clark, AG, Rohrbaugh, AL, Otterness, I and Kraus, VB (2002). The effects of ascorbic acid on cartilage
metabolism in guinea pig articular cartilage explants. Matrix Biol 21(2): 175-84.
Clarke, B (2008). Normal bone anatomy and physiology. Clin J Am Soc Nephrol 3 Suppl 3: S131-9.
Coleman, JE (1992). Structure and mechanism of alkaline phosphatase. Annu Rev Biophys Biomol Struct 21:
441-83.
Connolly, JF, Guse, R, Tiedeman, J and Dehne, R (1989). Autologous marrow injection for delayed unions of the
tibia: a preliminary report. J Orthop Trauma 3(4): 276-82.
Conrad, C, Gupta, R, Mohan, H, Niess, H, Bruns, CJ, Kopp, R, von Luettichau, I, Guba, M, Heeschen, C, Jauch,
KW, et al. (2007). Genetically engineered stem cells for therapeutic gene delivery. Curr Gene Ther
7(4): 249-60.
Cooper, MS, Hewison, M and Stewart, PM (1999). Glucocorticoid activity, inactivity and the osteoblast. J
Endocrinol 163(2): 159-64.
Coutu, DL, Yousefi, AM and Galipeau, J (2009). Three-dimensional porous scaffolds at the crossroads of tissue
engineering and cell-based gene therapy. J Cell Biochem 108(3): 537-46.
Cowan, CM, Shi, YY, Aalami, OO, Chou, YF, Mari, C, Thomas, R, Quarto, N, Contag, CH, Wu, B and
Longaker, MT (2004). Adipose-derived adult stromal cells heal critical-size mouse calvarial defects.
Nat Biotechnol 22(5): 560-7.
Cremer, MA, Rosloniec, EF and Kang, AH (1998). The cartilage collagens: a review of their structure,
organization, and role in the pathogenesis of experimental arthritis in animals and in human rheumatic
disease. J Mol Med 76(3-4): 275-88.
Dallari, D, Fini, M, Stagni, C, Torricelli, P, Nicoli Aldini, N, Giavaresi, G, Cenni, E, Baldini, N, Cenacchi, A,
Bassi, A, et al. (2006). In vivo study on the healing of bone defects treated with bone marrow stromal
cells, platelet-rich plasma, and freeze-dried bone allografts, alone and in combination. J Orthop Res
24(5): 877-88.
Datta, N, Pham, QP, Sharma, U, Sikavitsas, VI, Jansen, JA and Mikos, AG (2006). In vitro generated
extracellular matrix and fluid shear stress synergistically enhance 3D osteoblastic differentiation. Proc
Natl Acad Sci U S A 103(8): 2488-93.
De Val, S and Black, BL (2009). Transcriptional control of endothelial cell development. Dev Cell 16(2): 180-
95.
55
Del Guerra, S, Bracci, C, Nilsson, K, Belcourt, A, Kessler, L, Lupi, R, Marselli, L, De Vos, P and Marchetti, P
(2001). Entrapment of dispersed pancreatic islet cells in CultiSpher-S macroporous gelatin
microcarriers: Preparation, in vitro characterization, and microencapsulation. Biotechnol Bioeng
75(6): 741-4.
Delorme, B, Chateauvieux, S and Charbord, P (2006). The concept of mesenchymal stem cells. Regen Med 1(4):
497-509.
Deschaseaux, F, Sensebe, L and Heymann, D (2009). Mechanisms of bone repair and regeneration. Trends Mol
Med 15(9): 417-29.
Dimitriou, R, Dahabreh, Z, Katsoulis, E, Matthews, SJ, Branfoot, T and Giannoudis, PV (2005). Application of
recombinant BMP-7 on persistent upper and lower limb non-unions. Injury 36 Suppl 4: S51-9.
Docherty-Skogh, AC, Bergman, K, Waern, MJ, Ekman, S, Hultenby, K, Ossipov, D, Hilborn, J, Bowden, T and
Engstrand, T (2010). Bone morphogenetic protein-2 delivered by hyaluronan-based hydrogel induces
massive bone formation and healing of cranial defects in minipigs. Plast Reconstr Surg 125(5): 1383-
92.
Domm, C, Schunke, M, Christesen, K and Kurz, B (2002). Redifferentiation of dedifferentiated bovine articular
chondrocytes in alginate culture under low oxygen tension. Osteoarthritis Cartilage 10(1): 13-22.
Dong, YF, Soung do, Y, Chang, Y, Enomoto-Iwamoto, M, Paris, M, O'Keefe, RJ, Schwarz, EM and Drissi, H
(2007). Transforming growth factor-beta and Wnt signals regulate chondrocyte differentiation through
Twist1 in a stage-specific manner. Mol Endocrinol 21(11): 2805-20.
Dowthwaite, GP, Bishop, JC, Redman, SN, Khan, IM, Rooney, P, Evans, DJ, 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(Pt 6): 889-97.
Dragoo, JL, Carlson, G, McCormick, F, Khan-Farooqi, H, Zhu, M, Zuk, PA and Benhaim, P (2007). Healing
full-thickness cartilage defects using adipose-derived stem cells. Tissue Eng 13(7): 1615-21.
Ducy, P, Schinke, T and Karsenty, G (2000). The osteoblast: a sophisticated fibroblast under central
surveillance. Science 289(5484): 1501-4.
Dudzinski, DM, Igarashi, J, Greif, D and Michel, T (2006). The regulation and pharmacology of endothelial
nitric oxide synthase. Annu Rev Pharmacol Toxicol 46: 235-76.
Eames, BF, Sharpe, PT and Helms, JA (2004). Hierarchy revealed in the specification of three skeletal fates by
Sox9 and Runx2. Dev Biol 274(1): 188-200.
Einhorn, TA, Lane, JM, Burstein, AH, Kopman, CR and Vigorita, VJ (1984). The healing of segmental bone
defects induced by demineralized bone matrix. A radiographic and biomechanical study. J Bone Joint
Surg Am 66(2): 274-9.
Eisenberg, LM and Eisenberg, CA (2003). Stem cell plasticity, cell fusion, and transdifferentiation. Birth Defects
Res C Embryo Today 69(3): 209-18.
Enneking, WF and Campanacci, DA (2001). Retrieved human allografts : a clinicopathological study. J Bone
Joint Surg Am 83-A(7): 971-86.
Evans, MJ and Kaufman, MH (1981). Establishment in culture of pluripotential cells from mouse embryos.
Nature 292(5819): 154-6.
Eyre, DR (2004). Collagens and cartilage matrix homeostasis. Clin Orthop Relat Res(427 Suppl): S118-22.
Feng, X (2009). Chemical and Biochemical Basis of Cell-Bone Matrix Interaction in Health and Disease. Curr
Chem Biol 3(2): 189-196.
Fernandes, AM, Fernandes, TG, Diogo, MM, da Silva, CL, Henrique, D and Cabral, JM (2007). Mouse
embryonic stem cell expansion in a microcarrier-based stirred culture system. J Biotechnol 132(2):
227-36.
Fernandes, KJ, McKenzie, IA, Mill, P, Smith, KM, Akhavan, M, Barnabe-Heider, F, Biernaskie, J, Junek, A,
Kobayashi, NR, Toma, JG, et al. (2004). A dermal niche for multipotent adult skin-derived precursor
cells. Nat Cell Biol 6(11): 1082-93.
Fleige, S and Pfaffl, MW (2006). RNA integrity and the effect on the real-time qRT-PCR performance. Mol
Aspects Med 27(2-3): 126-39.
Frankel, MS (2000). In search of stem cell policy. Science 287(5457): 1397.
Freed, LE, Hollander, AP, Martin, I, Barry, JR, Langer, R and Vunjak-Novakovic, G (1998). Chondrogenesis in
a cell-polymer-bioreactor system. Exp Cell Res 240(1): 58-65.
French, MM, Rose, S, Canseco, J and Athanasiou, KA (2004). Chondrogenic differentiation of adult dermal
fibroblasts. Ann Biomed Eng 32(1): 50-6.
Friedenstein, AJ, Petrakova, KV, Kurolesova, AI and Frolova, GP (1968). Heterotopic of bone marrow. Analysis
of precursor cells for osteogenic and hematopoietic tissues. Transplantation 6(2): 230-47.
Frondoza, C, Sohrabi, A and Hungerford, D (1996). Human chondrocytes proliferate and produce matrix
components in microcarrier suspension culture. Biomaterials 17(9): 879-88.
56
Gabbay, KH, Bohren, KM, Morello, R, Bertin, T, Liu, J and Vogel, P (2010). Ascorbate synthesis pathway: dual
role of ascorbate in bone homeostasis. J Biol Chem 285(25): 19510-20.
Galli, R, Borello, U, Gritti, A, Minasi, MG, Bjornson, C, Coletta, M, Mora, M, De Angelis, MG, Fiocco, R,
Cossu, G, et al. (2000). Skeletal myogenic potential of human and mouse neural stem cells. Nat
Neurosci 3(10): 986-91.
Gallico, GG, 3rd, O'Connor, NE, Compton, CC, Kehinde, O and Green, H (1984). Permanent coverage of large
burn wounds with autologous cultured human epithelium. N Engl J Med 311(7): 448-51.
Gao, Y, Jheon, A, Nourkeyhani, H, Kobayashi, H and Ganss, B (2004). Molecular cloning, structure,
expression, and chromosomal localization of the human Osterix (SP7) gene. Gene 341: 101-10.
Gesta, S, Tseng, YH and Kahn, CR (2007). Developmental origin of fat: tracking obesity to its source. Cell
131(2): 242-56.
Goldie, LC, Nix, MK and Hirschi, KK (2008). Embryonic vasculogenesis and hematopoietic specification.
Organogenesis 4(4): 257-63.
Goldring, MB, Tsuchimochi, K and Ijiri, K (2006). The control of chondrogenesis. J Cell Biochem 97(1): 33-44.
Golub, EE (1996). Enzymes in mineralizing systems: state of the art. Connect Tissue Res 35(1-4): 183-8.
Govender, S, Csimma, C, Genant, HK, Valentin-Opran, A, Amit, Y, Arbel, R, Aro, H, Atar, D, Bishay, M,
Borner, MG, et al. (2002). Recombinant human bone morphogenetic protein-2 for treatment of open
tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone
Joint Surg Am 84-A(12): 2123-34.
Granero-Molto, F, Weis, JA, Miga, MI, Landis, B, Myers, TJ, O'Rear, L, Longobardi, L, Jansen, ED, Mortlock,
DP and Spagnoli, A (2009). Regenerative effects of transplanted mesenchymal stem cells in fracture
healing. Stem Cells 27(8): 1887-98.
Gregoire, FM (2001). Adipocyte differentiation: from fibroblast to endocrine cell. Exp Biol Med (Maywood)
226(11): 997-1002.
Gregoire, FM, Smas, CM and Sul, HS (1998). Understanding adipocyte differentiation. Physiol Rev 78(3): 783-
809.
Griffith, M, Jackson, WB, Lagali, N, Merrett, K, Li, F and Fagerholm, P (2009). Artificial corneas: a
regenerative medicine approach. Eye (Lond) 23(10): 1985-9.
Gron, B, Stoltze, K, Andersson, A and Dabelsteen, E (2002). Oral fibroblasts produce more HGF and KGF than
skin fibroblasts in response to co-culture with keratinocytes. APMIS 110(12): 892-8.
Gunatillake, PA and Adhikari, R (2003). Biodegradable synthetic polymers for tissue engineering. Eur Cell
Mater 5: 1-16; discussion 16.
Gurdon, JB, Elsdale, TR and Fischberg, M (1958). Sexually mature individuals of Xenopus laevis from the
transplantation of single somatic nuclei. Nature 182(4627): 64-5.
Haniffa, MA, Collin, MP, Buckley, CD and Dazzi, F (2009). Mesenchymal stem cells: the fibroblasts' new
clothes? Haematologica 94(2): 258-63.
Hannah, MJ, Williams, R, Kaur, J, Hewlett, LJ and Cutler, DF (2002). Biogenesis of Weibel-Palade bodies.
Semin Cell Dev Biol 13(4): 313-24.
Hannouche, D, Petite, H and Sedel, L (2001). Current trends in the enhancement of fracture healing. J Bone
Joint Surg Br 83(2): 157-64.
Hartmann, C (2009). Transcriptional networks controlling skeletal development. Curr Opin Genet Dev 19(5):
437-43.
Hartmann, C and Tabin, CJ (2001). Wnt-14 plays a pivotal role in inducing synovial joint formation in the
developing appendicular skeleton. Cell 104(3): 341-51.
Hattori, H, Masuoka, K, Sato, M, Ishihara, M, Asazuma, T, Takase, B, Kikuchi, M, Nemoto, K and Ishihara, M
(2006). Bone formation using human adipose tissue-derived stromal cells and a biodegradable scaffold.
J Biomed Mater Res B Appl Biomater 76(1): 230-9.
Hausman, GJ (1981). Techniques for studying adipocytes. Stain Technol 56(3): 149-54.
Hayflick, L and Moorhead, PS (1961). The serial cultivation of human diploid cell strains. Exp Cell Res 25:
585-621.
Heinegard, D (2009). Proteoglycans and more--from molecules to biology. Int J Exp Pathol 90(6): 575-86.
Hill, TP, Spater, D, Taketo, MM, Birchmeier, W and Hartmann, C (2005). Canonical Wnt/beta-catenin signaling
prevents osteoblasts from differentiating into chondrocytes. Dev Cell 8(5): 727-38.
Hirata, K, Mizuno, A and Yamaguchi, A (2007). Transplantation of skin fibroblasts expressing BMP-2
contributes to the healing of critical-sized bone defects. J Bone Miner Metab 25(1): 6-11.
Hofer, E and Schweighofer, B (2007). Signal transduction induced in endothelial cells by growth factor
receptors involved in angiogenesis. Thromb Haemost 97(3): 355-63.
Holtorf, HL, Jansen, JA and Mikos, AG (2005). Flow perfusion culture induces the osteoblastic differentiation
of marrow stroma cell-scaffold constructs in the absence of dexamethasone. J Biomed Mater Res A
72(3): 326-34.
57
Holtzer, H, Abbott, J, Lash, J and Holtzer, S (1960). The Loss of Phenotypic Traits by Differentiated Cells in
Vitro, I. Dedifferentiation of Cartilage Cells. Proc Natl Acad Sci U S A 46(12): 1533-42.
Huss, FR, Junker, JP, Johnson, H and Kratz, G (2007). Macroporous gelatine spheres as culture substrate,
transplantation vehicle, and biodegradable scaffold for guided regeneration of soft tissues. In vivo
study in nude mice. J Plast Reconstr Aesthet Surg 60(5): 543-55.
Ibold, Y, Lubke, C, Pelz, S, Augst, H, Kaps, C, Ringe, J and Sittinger, M (2009). Effect of different ascorbate
supplementations on in vitro cartilage formation in porcine high-density pellet cultures. Tissue Cell
41(4): 249-56.
Iwasaki, K, Kojima, K, Kodama, S, Paz, AC, Chambers, M, Umezu, M and Vacanti, CA (2008). Bioengineered
three-layered robust and elastic artery using hemodynamically-equivalent pulsatile bioreactor.
Circulation 118(14 Suppl): S52-7.
Iwasaki, M, Nakata, K, Nakahara, H, Nakase, T, Kimura, T, Kimata, K, Caplan, AI and Ono, K (1993).
Transforming growth factor-beta 1 stimulates chondrogenesis and inhibits osteogenesis in high density
culture of periosteum-derived cells. Endocrinology 132(4): 1603-8.
Jahoda, CA, Whitehouse, J, Reynolds, AJ and Hole, N (2003). Hair follicle dermal cells differentiate into
adipogenic and osteogenic lineages. Exp Dermatol 12(6): 849-59.
Jaiswal, N, Haynesworth, SE, Caplan, AI and Bruder, SP (1997). Osteogenic differentiation of purified, culture-
expanded human mesenchymal stem cells in vitro. J Cell Biochem 64(2): 295-312.
Jin, K and Greenberg, DA (2003). Tales of transdifferentiation. Exp Neurol 183(2): 255-7.
Johnstone, B, Hering, TM, Caplan, AI, Goldberg, VM and Yoo, JU (1998). In vitro chondrogenesis of bone
marrow-derived mesenchymal progenitor cells. Exp Cell Res 238(1): 265-72.
Junker, JP, Sommar, P, Skog, M, Johnson, H and Kratz, G (2010). Adipogenic, chondrogenic and osteogenic
differentiation of clonally derived human dermal fibroblasts. Cells Tissues Organs 191(2): 105-18.
Karlsson, C and Lindahl, A (2009). Notch signaling in chondrogenesis. Int Rev Cell Mol Biol 275: 65-88.
Karsenty, G (2008). Transcriptional control of skeletogenesis. Annu Rev Genomics Hum Genet 9: 183-96.
Karsenty, G and Wagner, EF (2002). Reaching a genetic and molecular understanding of skeletal development.
Dev Cell 2(4): 389-406.
Kastrinaki, MC, Sidiropoulos, P, Roche, S, Ringe, J, Lehmann, S, Kritikos, H, Vlahava, VM, Delorme, B,
Eliopoulos, GD, Jorgensen, C, et al. (2008). Functional, molecular and proteomic characterisation of
bone marrow mesenchymal stem cells in rheumatoid arthritis. Ann Rheum Dis 67(6): 741-9.
Kiskinis, E and Eggan, K (2010). Progress toward the clinical application of patient-specific pluripotent stem
cells. J Clin Invest 120(1): 51-9.
Komori, T (2010). Regulation of bone development and extracellular matrix protein genes by RUNX2. Cell
Tissue Res 339(1): 189-95.
Kuhne, M, John, T, El-Sayed, K, Marzahn, U, Aue, A, Kohl, B, Stoelzel, K, Ertel, W, Blottner, D, Haisch, A, et
al. (2010). Characterization of auricular chondrocytes and auricular/articular chondrocyte co-cultures
in terms of an application in articular cartilage repair. Int J Mol Med 25(5): 701-8.
Kutt, H and Tsaltas, TT (1959). Staining properties of oil red O and a method of partial purification of the
commercial product. Clin Chem 5(2): 149-60.
Kuznetsov, VA, Knott, GD and Bonner, RF (2002). General statistics of stochastic process of gene expression in
eukaryotic cells. Genetics 161(3): 1321-32.
Lako, M, Armstrong, L, Cairns, PM, Harris, S, Hole, N and Jahoda, CA (2002). Hair follicle dermal cells
repopulate the mouse haematopoietic system. J Cell Sci 115(Pt 20): 3967-74.
Lander, ES, Linton, LM, Birren, B, Nusbaum, C, Zody, MC, Baldwin, J, Devon, K, Dewar, K, Doyle, M,
FitzHugh, W, et al. (2001). Initial sequencing and analysis of the human genome. Nature 409(6822):
860-921.
Langer, R and Vacanti, JP (1993). Tissue engineering. Science 260(5110): 920-6.
Lehmann, JM, Lenhard, JM, Oliver, BB, Ringold, GM and Kliewer, SA (1997). Peroxisome proliferator-
activated receptors alpha and gamma are activated by indomethacin and other non-steroidal anti-
inflammatory drugs. J Biol Chem 272(6): 3406-10.
Lendeckel, S, Jodicke, A, Christophis, P, Heidinger, K, Wolff, J, Fraser, JK, Hedrick, MH, Berthold, L and
Howaldt, HP (2004). Autologous stem cells (adipose) and fibrin glue used to treat widespread
traumatic calvarial defects: case report. J Craniomaxillofac Surg 32(6): 370-3.
Lev, R and Spicer, SS (1964). Specific Staining of Sulphate Groups with Alcian Blue at Low Ph. J Histochem
Cytochem 12: 309.
Li, X and Cao, X (2006). BMP signaling and skeletogenesis. Ann N Y Acad Sci 1068: 26-40.
Li, YQ (2010). Master stem cell transcription factors and signaling regulation. Cell Reprogram 12(1): 3-13.
Liu, F, Akiyama, Y, Tai, S, Maruyama, K, Kawaguchi, Y, Muramatsu, K and Yamaguchi, K (2008). Changes in
the expression of CD106, osteogenic genes, and transcription factors involved in the osteogenic
differentiation of human bone marrow mesenchymal stem cells. J Bone Miner Metab 26(4): 312-20.
58
Liu, JY, Hafner, J, Dragieva, G, Seifert, B and Burg, G (2004). Autologous cultured keratinocytes on porcine
gelatin microbeads effectively heal chronic venous leg ulcers. Wound Repair Regen 12(2): 148-56.
Livak, KJ and Schmittgen, TD (2001). Analysis of relative gene expression data using real-time quantitative
PCR and the 2(-Delta Delta C(T)) Method. Methods 25(4): 402-8.
Lorenz, K, Sicker, M, Schmelzer, E, Rupf, T, Salvetter, J, Schulz-Siegmund, M and Bader, A (2008).
Multilineage differentiation potential of human dermal skin-derived fibroblasts. Exp Dermatol 17(11):
925-32.
Lowry, WE, Richter, L, Yachechko, R, Pyle, AD, Tchieu, J, Sridharan, R, Clark, AT and Plath, K (2008).
Generation of human induced pluripotent stem cells from dermal fibroblasts. Proc Natl Acad Sci U S A
105(8): 2883-8.
Lysy, PA, Smets, F, Sibille, C, Najimi, M and Sokal, EM (2007). Human skin fibroblasts: From mesodermal to
hepatocyte-like differentiation. Hepatology 46(5): 1574-85.
Macchiarini, P, Jungebluth, P, Go, T, Asnaghi, MA, Rees, LE, Cogan, TA, Dodson, A, Martorell, J, Bellini, S,
Parnigotto, PP, et al. (2008). Clinical transplantation of a tissue-engineered airway. Lancet 372(9655):
2023-30.
Macias, D, Ganan, Y, Sampath, TK, Piedra, ME, Ros, MA and Hurle, JM (1997). Role of BMP-2 and OP-1
(BMP-7) in programmed cell death and skeletogenesis during chick limb development. Development
124(6): 1109-17.
Magnusson, P and Farley, JR (2002). Differences in sialic acid residues among bone alkaline phosphatase
isoforms: a physical, biochemical, and immunological characterization. Calcif Tissue Int 71(6): 508-
18.
Malda, J and Frondoza, CG (2006). Microcarriers in the engineering of cartilage and bone. Trends Biotechnol
24(7): 299-304.
Malda, J, Kreijveld, E, Temenoff, JS, van Blitterswijk, CA and Riesle, J (2003). Expansion of human nasal
chondrocytes on macroporous microcarriers enhances redifferentiation. Biomaterials 24(28): 5153-61.
Mandruzzato, S (2007). Technological platforms for microarray gene expression profiling. Adv Exp Med Biol
593: 12-8.
Martin, GR (1981). Isolation of a pluripotent cell line from early mouse embryos cultured in medium
conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci U S A 78(12): 7634-8.
Mason, RM (1971). Observations on the glycosaminoglycans of aging bronchial cartilage studied with Alcian
Blue. Histochem J 3(6): 421-34.
Mauck, RL, Soltz, MA, Wang, CC, Wong, DD, Chao, PH, Valhmu, WB, Hung, CT and Ateshian, GA (2000).
Functional tissue engineering of articular cartilage through dynamic loading of chondrocyte-seeded
agarose gels. J Biomech Eng 122(3): 252-60.
McMahon, LA, Prendergast, PJ and Campbell, VA (2008). A comparison of the involvement of p38, ERK1/2 and
PI3K in growth factor-induced chondrogenic differentiation of mesenchymal stem cells. Biochem
Biophys Res Commun 368(4): 990-5.
Mehta, D and Malik, AB (2006). Signaling mechanisms regulating endothelial permeability. Physiol Rev 86(1):
279-367.
Melero-Martin, JM, Dowling, MA, Smith, M and Al-Rubeai, M (2006). Expansion of chondroprogenitor cells
on macroporous microcarriers as an alternative to conventional monolayer systems. Biomaterials
27(15): 2970-9.
Mithieux, SM and Weiss, AS (2005). Elastin. Adv Protein Chem 70: 437-61.
Mizuno, S and Glowacki, J (1996). Chondroinduction of human dermal fibroblasts by demineralized bone in
three-dimensional culture. Exp Cell Res 227(1): 89-97.
Mizuno, S and Glowacki, J (1996). Three-dimensional composite of demineralized bone powder and collagen
for in vitro analysis of chondroinduction of human dermal fibroblasts. Biomaterials 17(18): 1819-25.
Morgelin, M, Paulsson, M, Hardingham, TE, Heinegard, D and Engel, J (1988). Cartilage proteoglycans.
Assembly with hyaluronate and link protein as studied by electron microscopy. Biochem J 253(1): 175-
85.
Moulharat, N, Lesur, C, Thomas, M, Rolland-Valognes, G, Pastoureau, P, Anract, P, De Ceuninck, F and
Sabatini, M (2004). Effects of transforming growth factor-beta on aggrecanase production and
proteoglycan degradation by human chondrocytes in vitro. Osteoarthritis Cartilage 12(4): 296-305.
Muschler, GF, Nakamoto, C and Griffith, LG (2004). Engineering principles of clinical cell-based tissue
engineering. J Bone Joint Surg Am 86-A(7): 1541-58.
Nakashima, K, Zhou, X, Kunkel, G, Zhang, Z, Deng, JM, Behringer, RR and de Crombrugghe, B (2002). The
novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and
bone formation. Cell 108(1): 17-29.
59
Ogston, N, Harrison, AJ, Cheung, HF, Ashton, BA and Hampson, G (2002). Dexamethasone and retinoic acid
differentially regulate growth and differentiation in an immortalised human clonal bone marrow
stromal cell line with osteoblastic characteristics. Steroids 67(11): 895-906.
Ohlson, S, Branscomb, J and Nilsson, K (1994). Bead-to-bead transfer of Chinese hamster ovary cells using
macroporous microcarriers. Cytotechnology 14(1): 67-80.
Okazaki, K, Li, J, Yu, H, Fukui, N and Sandell, LJ (2002). CCAAT/enhancer-binding proteins beta and delta
mediate the repression of gene transcription of cartilage-derived retinoic acid-sensitive protein induced
by interleukin-1 beta. J Biol Chem 277(35): 31526-33.
Olsson, AK, Dimberg, A, Kreuger, J and Claesson-Welsh, L (2006). VEGF receptor signalling - in control of
vascular function. Nat Rev Mol Cell Biol 7(5): 359-71.
Olsson, MJ and Juhlin, L (1992). Melanocyte transplantation in vitiligo. Lancet 340(8825): 981.
Osborne, TF (2000). Sterol regulatory element-binding proteins (SREBPs): key regulators of nutritional
homeostasis and insulin action. J Biol Chem 275(42): 32379-82.
Pagani, F, Francucci, CM and Moro, L (2005). Markers of bone turnover: biochemical and clinical perspectives.
J Endocrinol Invest 28(10 Suppl): 8-13.
Papanicolaou, KN, Izumiya, Y and Walsh, K (2008). Forkhead transcription factors and cardiovascular
biology. Circ Res 102(1): 16-31.
Parsons, WJ, Ramkumar, V and Stiles, GL (1988). Isobutylmethylxanthine stimulates adenylate cyclase by
blocking the inhibitory regulatory protein, Gi. Mol Pharmacol 34(1): 37-41.
Pazzano, D, Mercier, KA, Moran, JM, Fong, SS, DiBiasio, DD, Rulfs, JX, Kohles, SS and Bonassar, LJ (2000).
Comparison of chondrogensis in static and perfused bioreactor culture. Biotechnol Prog 16(5): 893-6.
Petersen, DW and Kawasaki, ES (2007). Manufacturing of microarrays. Adv Exp Med Biol 593: 1-11.
Pettersson, S, Wettero, J, Tengvall, P and Kratz, G (2009). Human articular chondrocytes on macroporous
gelatin microcarriers form structurally stable constructs with blood-derived biological glues in vitro. J
Tissue Eng Regen Med 3(6): 450-60.
Phillips, JE, Guldberg, RE and Garcia, AJ (2007). Dermal fibroblasts genetically modified to express
Runx2/Cbfa1 as a mineralizing cell source for bone tissue engineering. Tissue Eng 13(8): 2029-40.
Pittenger, MF, Mackay, AM, Beck, SC, Jaiswal, RK, Douglas, R, Mosca, JD, Moorman, MA, Simonetti, DW,
Craig, S and Marshak, DR (1999). Multilineage potential of adult human mesenchymal stem cells.
Science 284(5411): 143-7.
Prockop, DJ (1997). Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 276(5309): 71-4.
Prowse, AB, McQuade, LR, Bryant, KJ, Marcal, H and Gray, PP (2007). Identification of potential pluripotency
determinants for human embryonic stem cells following proteomic analysis of human and mouse
fibroblast conditioned media. J Proteome Res 6(9): 3796-807.
Quintana, L, zur Nieden, NI and Semino, CE (2009). Morphogenetic and regulatory mechanisms during
developmental chondrogenesis: new paradigms for cartilage tissue engineering. Tissue Eng Part B Rev
15(1): 29-41.
Rached, MT, Kode, A, Xu, L, Yoshikawa, Y, Paik, JH, Depinho, RA and Kousteni, S (2010). FoxO1 is a
positive regulator of bone formation by favoring protein synthesis and resistance to oxidative stress in
osteoblasts. Cell Metab 11(2): 147-60.
Rangwala, SM and Lazar, MA (2000). Transcriptional control of adipogenesis. Annu Rev Nutr 20: 535-59.
Reimers, M (2010). Making informed choices about microarray data analysis. PLoS Comput Biol 6(5):
e1000786.
Reusch, JE, Colton, LA and Klemm, DJ (2000). CREB activation induces adipogenesis in 3T3-L1 cells. Mol
Cell Biol 20(3): 1008-20.
Ribatti, D (2008). Hemangioblast does exist. Leuk Res 32(6): 850-4.
Richardson, GD, Arnott, EC, Whitehouse, CJ, Lawrence, CM, Hole, N and Jahoda, CA (2005). Cultured cells
from the adult human hair follicle dermis can be directed toward adipogenic and osteogenic
differentiation. J Invest Dermatol 124(5): 1090-1.
Rodemann, HP, Bayreuther, K, Francz, PI, Dittmann, K and Albiez, M (1989). Selective enrichment and
biochemical characterization of seven human skin fibroblasts cell types in vitro. Exp Cell Res 180(1):
84-93.
Romero-Ramos, M, Vourc'h, P, Young, HE, Lucas, PA, Wu, Y, Chivatakarn, O, Zaman, R, Dunkelman, N, el-
Kalay, MA and Chesselet, MF (2002). Neuronal differentiation of stem cells isolated from adult muscle.
J Neurosci Res 69(6): 894-907.
Rosen, ED and Spiegelman, BM (2000). Molecular regulation of adipogenesis. Annu Rev Cell Dev Biol 16:
145-71.
Ross, SE, Hemati, N, Longo, KA, Bennett, CN, Lucas, PC, Erickson, RL and MacDougald, OA (2000).
Inhibition of adipogenesis by Wnt signaling. Science 289(5481): 950-3.
60
Rossig, L, Urbich, C, Bruhl, T, Dernbach, E, Heeschen, C, Chavakis, E, Sasaki, K, Aicher, D, Diehl, F, Seeger,
F, et al. (2005). Histone deacetylase activity is essential for the expression of HoxA9 and for endothelial
commitment of progenitor cells. J Exp Med 201(11): 1825-35.
Roughley, PJ (2006). The structure and function of cartilage proteoglycans. Eur Cell Mater 12: 92-101.
Rouwkema, J, Rivron, NC and van Blitterswijk, CA (2008). Vascularization in tissue engineering. Trends
Biotechnol 26(8): 434-41.
Rutherford, RB, Moalli, M, Franceschi, RT, Wang, D, Gu, K and Krebsbach, PH (2002). Bone morphogenetic
protein-transduced human fibroblasts convert to osteoblasts and form bone in vivo. Tissue Eng 8(3):
441-52.
Sadler, JE (1998). Biochemistry and genetics of von Willebrand factor. Annu Rev Biochem 67: 395-424.
Sasso, RC, LeHuec, JC and Shaffrey, C (2005). Iliac crest bone graft donor site pain after anterior lumbar
interbody fusion: a prospective patient satisfaction outcome assessment. J Spinal Disord Tech 18 Suppl:
S77-81.
Schnabel, M, Marlovits, S, Eckhoff, G, Fichtel, I, Gotzen, L, Vecsei, V and Schlegel, J (2002).
Dedifferentiation-associated changes in morphology and gene expression in primary human articular
chondrocytes in cell culture. Osteoarthritis Cartilage 10(1): 62-70.
Schulz, RM and Bader, A (2007). Cartilage tissue engineering and bioreactor systems for the cultivation and
stimulation of chondrocytes. Eur Biophys J 36(4-5): 539-68.
Schwartz, S, Depinet, TW, Leana-Cox, J, Isada, NB, Karson, EM, Park, VM, Pasztor, LM, Sheppard, LC,
Stallard, R, Wolff, DJ, et al. (1997). Sex chromosome markers: characterization using fluorescence in
situ hybridization and review of the literature. Am J Med Genet 71(1): 1-7.
Schwartz, Z, Denison, TA, Bannister, SR, Cochran, DL, Liu, YH, Lohmann, CH, Wieland, M and Boyan, BD
(2007). Osteoblast response to fluid induced shear depends on substrate microarchitecture and varies
with time. J Biomed Mater Res A 83(1): 20-32.
Sharma, B and Elisseeff, JH (2004). Engineering structurally organized cartilage and bone tissues. Ann Biomed
Eng 32(1): 148-59.
Sharpless, NE and DePinho, RA (2004). Telomeres, stem cells, senescence, and cancer. J Clin Invest 113(2):
160-8.
Sikavitsas, VI, Bancroft, GN, Holtorf, HL, Jansen, JA and Mikos, AG (2003). Mineralized matrix deposition by
marrow stromal osteoblasts in 3D perfusion culture increases with increasing fluid shear forces. Proc
Natl Acad Sci U S A 100(25): 14683-8.
Slapnicka, J, Fassmann, A, Strasak, L, Augustin, P and Vanek, J (2008). Effects of activated and nonactivated
platelet-rich plasma on proliferation of human osteoblasts in vitro. J Oral Maxillofac Surg 66(2): 297-
301.
Smith, PJ, Wise, LS, Berkowitz, R, Wan, C and Rubin, CS (1988). Insulin-like growth factor-I is an essential
regulator of the differentiation of 3T3-L1 adipocytes. J Biol Chem 263(19): 9402-8.
Smith, RL, Donlon, BS, Gupta, MK, Mohtai, M, Das, P, Carter, DR, Cooke, J, Gibbons, G, Hutchinson, N and
Schurman, DJ (1995). Effects of fluid-induced shear on articular chondrocyte morphology and
metabolism in vitro. J Orthop Res 13(6): 824-31.
Soderhall, C, Marenholz, I, Kerscher, T, Ruschendorf, F, Esparza-Gordillo, J, Worm, M, Gruber, C, Mayr, G,
Albrecht, M, Rohde, K, et al. (2007). Variants in a novel epidermal collagen gene (COL29A1) are
associated with atopic dermatitis. PLoS Biol 5(9): e242.
Solter, D and Knowles, BB (1975). Immunosurgery of mouse blastocyst. Proc Natl Acad Sci U S A 72(12):
5099-102.
Sorrell, JM and Caplan, AI (2004). Fibroblast heterogeneity: more than skin deep. J Cell Sci 117(Pt 5): 667-75.
St-Jacques, B, Hammerschmidt, M and McMahon, AP (1999). Indian hedgehog signaling regulates proliferation
and differentiation of chondrocytes and is essential for bone formation. Genes Dev 13(16): 2072-86.
Stevens, B, Yang, Y, Mohandas, A, Stucker, B and Nguyen, KT (2008). A review of materials, fabrication
methods, and strategies used to enhance bone regeneration in engineered bone tissues. J Biomed Mater
Res B Appl Biomater 85(2): 573-82.
Storm, EE and Kingsley, DM (1996). Joint patterning defects caused by single and double mutations in members
of the bone morphogenetic protein (BMP) family. Development 122(12): 3969-79.
Stover, NP, Bakay, RA, Subramanian, T, Raiser, CD, Cornfeldt, ML, Schweikert, AW, Allen, RC and Watts, RL
(2005). Intrastriatal implantation of human retinal pigment epithelial cells attached to microcarriers in
advanced Parkinson disease. Arch Neurol 62(12): 1833-7.
Sul, HS (2009). Minireview: Pref-1: role in adipogenesis and mesenchymal cell fate. Mol Endocrinol 23(11):
1717-25.
Sumpio, BE, Riley, JT and Dardik, A (2002). Cells in focus: endothelial cell. Int J Biochem Cell Biol 34(12):
1508-12.
61
Sun, N, Longaker, MT and Wu, JC (2010). Human iPS cell-based therapy: considerations before clinical
applications. Cell Cycle 9(5): 880-5.
Tada, M, Takahama, Y, Abe, K, Nakatsuji, N and Tada, T (2001). Nuclear reprogramming of somatic cells by in
vitro hybridization with ES cells. Curr Biol 11(19): 1553-8.
Takada, I, Kouzmenko, AP and Kato, S (2009). Molecular switching of osteoblastogenesis versus adipogenesis:
implications for targeted therapies. Expert Opin Ther Targets 13(5): 593-603.
Takahashi, K, Tanabe, K, Ohnuki, M, Narita, M, Ichisaka, T, Tomoda, K and Yamanaka, S (2007). Induction of
pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131(5): 861-72.
Takahashi, K and Yamanaka, S (2006). Induction of pluripotent stem cells from mouse embryonic and adult
fibroblast cultures by defined factors. Cell 126(4): 663-76.
Takamizawa, S, Maehata, Y, Imai, K, Senoo, H, Sato, S and Hata, R (2004). Effects of ascorbic acid and
ascorbic acid 2-phosphate, a long-acting vitamin C derivative, on the proliferation and differentiation
of human osteoblast-like cells. Cell Biol Int 28(4): 255-65.
Thomas, KR and Capecchi, MR (1990). Targeted disruption of the murine int-1 proto-oncogene resulting in
severe abnormalities in midbrain and cerebellar development. Nature 346(6287): 847-50.
Thomson, JA, Itskovitz-Eldor, J, Shapiro, SS, Waknitz, MA, Swiergiel, JJ, Marshall, VS and Jones, JM (1998).
Embryonic stem cell lines derived from human blastocysts. Science 282(5391): 1145-7.
Thorwarth, M, Wehrhan, F, Schultze-Mosgau, S, Wiltfang, J and Schlegel, KA (2006). PRP modulates
expression of bone matrix proteins in vivo without long-term effects on bone formation. Bone 38(1): 30-
40.
Tielens, S, Declercq, H, Gorski, T, Lippens, E, Schacht, E and Cornelissen, M (2007). Gelatin-based
microcarriers as embryonic stem cell delivery system in bone tissue engineering: an in-vitro study.
Biomacromolecules 8(3): 825-32.
Toma, JG, Akhavan, M, Fernandes, KJ, Barnabe-Heider, F, Sadikot, A, Kaplan, DR and Miller, FD (2001).
Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nat Cell Biol 3(9): 778-84.
Toma, JG, McKenzie, IA, Bagli, D and Miller, FD (2005). Isolation and characterization of multipotent skin-
derived precursors from human skin. Stem Cells 23(6): 727-37.
Tong, Q, Dalgin, G, Xu, H, Ting, CN, Leiden, JM and Hotamisligil, GS (2000). Function of GATA transcription
factors in preadipocyte-adipocyte transition. Science 290(5489): 134-8.
Tsiridis, E, Upadhyay, N and Giannoudis, P (2007). Molecular aspects of fracture healing: which are the
important molecules? Injury 38 Suppl 1: S11-25.
Tsonis, PA (2000). Regeneration in vertebrates. Dev Biol 221(2): 273-84.
Wada, MR, Inagawa-Ogashiwa, M, Shimizu, S, Yasumoto, S and Hashimoto, N (2002). Generation of different
fates from multipotent muscle stem cells. Development 129(12): 2987-95.
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(11): 1402-16.
Wakitani, S, Imoto, K, Yamamoto, T, Saito, M, Murata, N and Yoneda, M (2002). Human autologous culture
expanded bone marrow mesenchymal cell transplantation for repair of cartilage defects in
osteoarthritic knees. Osteoarthritis Cartilage 10(3): 199-206.
Waldman, SD, Spiteri, CG, Grynpas, MD, Pilliar, RM and Kandel, RA (2004). Long-term intermittent
compressive stimulation improves the composition and mechanical properties of tissue-engineered
cartilage. Tissue Eng 10(9-10): 1323-31.
van Wezel, AL (1967). Growth of Cell-strains and Primary Cells on Micro-carriers in Homogeneous Culture.
Nature 216: 64-65.
Wang, R, Zou, Y, Yuan, Z, Wang, Y, Chen, Y, Mao, Y, Zhu, ZA, Li, H, Tang, X, Lu, J, et al. (2009). Autografts
and xenografts of skin fibroblasts delivering BMP-2 effectively promote orthotopic and ectopic
osteogenesis. Anat Rec (Hoboken) 292(6): 777-86.
Warnke, PH, Springer, IN, Wiltfang, J, Acil, Y, Eufinger, H, Wehmoller, M, Russo, PA, Bolte, H, Sherry, E,
Behrens, E, et al. (2004). Growth and transplantation of a custom vascularised bone graft in a man.
Lancet 364(9436): 766-70.
Weigel, T, Schinkel, G and Lendlein, A (2006). Design and preparation of polymeric scaffolds for tissue
engineering. Expert Rev Med Devices 3(6): 835-51.
Wendt, D, Stroebel, S, Jakob, M, John, GT and Martin, I (2006). Uniform tissues engineered by seeding and
culturing cells in 3D scaffolds under perfusion at defined oxygen tensions. Biorheology 43(3-4): 481-8.
Verfaillie, CM, Pera, MF and Lansdorp, PM (2002). Stem cells: hype and reality. Hematology Am Soc Hematol
Educ Program: 369-91.
Werntz, JR, Lane, JM, Burstein, AH, Justin, R, Klein, R and Tomin, E (1996). Qualitative and quantitative
analysis of orthotopic bone regeneration by marrow. J Orthop Res 14(1): 85-93.
62
Vestweber, D, Winderlich, M, Cagna, G and Nottebaum, AF (2009). Cell adhesion dynamics at endothelial
junctions: VE-cadherin as a major player. Trends Cell Biol 19(1): 8-15.
Willers, C, Partsalis, T and Zheng, MH (2007). Articular cartilage repair: procedures versus products. Expert
Rev Med Devices 4(3): 373-92.
Williams, RL, Hilton, DJ, Pease, S, Willson, TA, Stewart, CL, Gearing, DP, Wagner, EF, Metcalf, D, Nicola,
NA and Gough, NM (1988). Myeloid leukaemia inhibitory factor maintains the developmental potential
of embryonic stem cells. Nature 336(6200): 684-7.
Wilmut, I, Schnieke, AE, McWhir, J, Kind, AJ and Campbell, KH (1997). Viable offspring derived from fetal
and adult mammalian cells. Nature 385(6619): 810-3.
Woll, PS, Morris, JK, Painschab, MS, Marcus, RK, Kohn, AD, Biechele, TL, Moon, RT and Kaufman, DS
(2008). Wnt signaling promotes hematoendothelial cell development from human embryonic stem cells.
Blood 111(1): 122-31.
Voyta, JC, Via, DP, Butterfield, CE and Zetter, BR (1984). Identification and isolation of endothelial cells based
on their increased uptake of acetylated-low density lipoprotein. J Cell Biol 99(6): 2034-40.
Wren, JD and Conway, T (2006). Meta-analysis of published transcriptional and translational fold changes
reveals a preference for low-fold inductions. OMICS 10(1): 15-27.
Wu, Z, Bucher, NL and Farmer, SR (1996). Induction of peroxisome proliferator-activated receptor gamma
during the conversion of 3T3 fibroblasts into adipocytes is mediated by C/EBPbeta, C/EBPdelta, and
glucocorticoids. Mol Cell Biol 16(8): 4128-36.
Vunjak-Novakovic, G, Obradovic, B, Martin, I and Freed, LE (2002). Bioreactor studies of native and tissue
engineered cartilage. Biorheology 39(1-2): 259-68.
www.hcdm.org. "HCDM; HLDA workshops and CD Molecules."
Yoo, JU, Barthel, TS, Nishimura, K, Solchaga, L, Caplan, AI, Goldberg, VM and Johnstone, B (1998). The
chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells. J Bone Joint
Surg Am 80(12): 1745-57.
Young, HE and Black, AC, Jr. (2004). Adult stem cells. Anat Rec A Discov Mol Cell Evol Biol 276(1): 75-102.
Young, HE, Mancini, ML, Wright, RP, Smith, JC, Black, AC, Jr., Reagan, CR and Lucas, PA (1995).
Mesenchymal stem cells reside within the connective tissues of many organs. Dev Dyn 202(2): 137-44.
Young, LS, Searle, PF, Onion, D and Mautner, V (2006). Viral gene therapy strategies: from basic science to
clinical application. J Pathol 208(2): 299-318.
Young, MF, Kerr, JM, Ibaraki, K, Heegaard, AM and Robey, PG (1992). Structure, expression, and regulation
of the major noncollagenous matrix proteins of bone. Clin Orthop Relat Res(281): 275-94.
Yu, X, Botchwey, EA, Levine, EM, Pollack, SR and Laurencin, CT (2004). Bioreactor-based bone tissue
engineering: the influence of dynamic flow on osteoblast phenotypic expression and matrix
mineralization. Proc Natl Acad Sci U S A 101(31): 11203-8.
Zeng, L, Kempf, H, Murtaugh, LC, Sato, ME and Lassar, AB (2002). Shh establishes an Nkx3.2/Sox9
autoregulatory loop that is maintained by BMP signals to induce somitic chondrogenesis. Genes Dev
16(15): 1990-2005.
Zhao, LR, Duan, WM, Reyes, M, Keene, CD, Verfaillie, CM and Low, WC (2002). Human bone marrow stem
cells exhibit neural phenotypes and ameliorate neurological deficits after grafting into the ischemic
brain of rats. Exp Neurol 174(1): 11-20.
Ziegler, EE (1944). Hematoxylin-Eosin Tissue Stain. Arch Path(37): 68.
Zuk, PA, Zhu, M, Ashjian, P, De Ugarte, DA, Huang, JI, Mizuno, H, Alfonso, ZC, Fraser, JK, Benhaim, P and
Hedrick, MH (2002). Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 13(12):
4279-95.
Zuk, PA, Zhu, M, Mizuno, H, Huang, J, Futrell, JW, Katz, AJ, Benhaim, P, Lorenz, HP and Hedrick, MH
(2001). Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng
7(2): 211-28.
Fax +41 61 306 12 34E-Mail [email protected]
Original Paper
Cells Tissues Organs 2010;191:105–118 DOI: 10.1159/000232157
Adipogenic, Chondrogenic and Osteogenic Differentiation of Clonally Derived Human Dermal Fibroblasts
Johan P.E. Junker a Pehr Sommar a Mårten Skog a Hans Johnson b Gunnar Kratz a
a Laboratory for Reconstructive Plastic Surgery, Department of Plastic and Hand Surgery, and Department of Clinical and Experimental Medicine, University Hospital, Linköping , Sweden; b Department of Surgery, Section of Plastic Surgery and Burn Center, Haukeland University Hospital, Bergen , Norway
ence of ‘stem-cell-specific’ surface antigens was analyzed using flow cytometry. The results reveal that sccFBs have several of the markers associated with cells exhibiting stem cell plasticity. The findings presented here are corroborated by the findings of other groups, and suggest the use of hu-man dermal FBs in cell-based therapies for the reconstruc-tion of fat, cartilage and bone. Copyright © 2009 S. Karger AG, Basel
Key Words
Adipogenesis � Chondrogenesis � Fibroblasts � Osteogenesis � Tissue engineering
Abstract
The apparent need of an autologous cell source for tissue engineering applications has led researchers to explore the presence of cells with stem cell plasticity in several human tissues. Dermal fibroblasts (FBs) are easy to harvest, expand in vitro and store, rendering them plausible candidates for cell-based therapies. The aim of the present study was to observe the effects of adipogenic, chondrogenic and osteo-genic induction media on the phenotype of human FBs. Human preadipocytes obtained from fat tissue have been proposed as an adult stem cell source with suitable charac-teristics, and were used as control cells in regard to their dif-ferentiation potential. Routine staining, immunohistochem-ical analysis and alkaline phosphatase assay were employed, in order to study the phenotypic shift. FBs were shown to possess multilineage potential, giving rise to fat-, cartilage- and bone-like cells. To exclude contaminant progenitor cells or cell fusion giving rise to tissue with adipocyte-, chondro-cyte- and osteoblast-like cells, single-cell cloning was per-formed. Single-cell-cloned FBs (sccFBs) displayed a similar differentiation potential as primary-culture FBs. The pres-
Accepted after revision: March 14, 2009 Published online: July 28, 2009
Dr. Johan P.E. Junker Center for Clinical and Experimental Research Department of Clinical and Experimental Medicine Faculty of Health Sciences, Linköping University, SE–58185 Linköping (Sweden) Tel. +46 13 224 485, Fax +46 13 223 706, E-Mail [email protected]
© 2009 S. Karger AG, Basel1422–6405/10/1912–0105$26.00/0
Accessible online at:www.karger.com/cto
Abbreviations used in this paper
ALP alkaline phosphataseBSA bovine serum albuminCD cluster of differentiationCM chondrogenic mediumdH2O distilled waterDMEM Dulbecco’s modified Eagle’s mediumFB fibroblastIHC immunohistochemistryNBP neutral buffered paraformaldehydeOD optical densityPA preadipocytePBS phosphate-buffered salinesccFB single-cell-cloned fibroblast
J.P.E.J. and P.S. contributed equally to this work.
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 106
Introduction
The loss or failure of organs and tissues is one of the most frequent, devastating and costly problems in hu-man health care. Tissue engineering, which is one of the fastest growing areas in medicine today, applies the prin-ciples of biology and engineering to the development of functional substitutes for damaged tissues. During the last decades, the area has gained a lot of interest and a number of cell-based therapies have been developed [Gal-lico et al., 1984; Olsson and Juhlin, 1992; Peterson et al., 2000; Fossum et al., 2004]. However, one of the main ob-stacles has been the limited availability of autologous tis-sue-specific progenitor cells. This has drawn attention to the possibility of using autologous cells with stem cell plasticity [Pittenger et al., 1999; Jaiswal et al., 2000; Ales-sandri et al., 2004; Guilak et al., 2006; Schultz and Lucas, 2006; Chen et al., 2007; Gao et al., 2007; Passier et al., 2008]. During the last years, the concept of adult cells be-ing restricted in their ability to generate only the differ-entiated cell phenotypes of their own tissue has been challenged. Bone-marrow-derived stem cells were the first adult stem cells shown to have multilineage potential [Jaiswal et al., 1997; Yoo et al., 1998], and initial studies were followed by several reports indicating that cells from other tissues transplanted into a new environment result-ed in detectable contribution in several lineages distinct from their tissue of origin [Zuk et al., 2001, 2002; Guilak et al., 2006; Chen et al., 2007; Dragoo et al., 2007; Bajada et al., 2008; Heydarkhan-Hagvall et al., 2008; Svendsen, 2008]. The discovery of several cluster of differentiation (CD) markers has facilitated the classification and isola-tion of adult stem cells [Jaiswal et al., 1997; Young et al., 1999; Gronthos et al., 2001; Zuk et al., 2002; Young et al., 2004; Bartsch et al., 2005]. When employing antibodies towards these markers in flow-cytometry-based applica-tions, a large number of cells can be analyzed and sorted, thus generating viable populations of adult stem cells with a high degree of purity.
Although promising, adult stem cells have major im-pediments for use in tissue engineering applications. The difficulties associated with their harvest, culture and storage render problems in the development of clinically relevant procedures. The dermal fibroblast (FB) is one of the easiest cell types to process in vitro, and is relatively easy to expand until sufficient numbers are reached. Fur-thermore, the inherent plasticity in FBs has been pro-posed, suggesting that they are able to differentiate to-wards several mesenchymal lineages [Bartsch et al., 2005; Chen et al., 2007]. The usage of FBs in cell-based thera-
pies may facilitate the in vitro production of autologous tissues in a most dramatic way and thus pave the way for new treatment regimes for the reconstruction of dam-aged tissues and organs.
There are intense ongoing discussions about the mech-anisms behind the differentiation of adult stem cells and a variety of different hypotheses on the subject [Eisen-berg and Eisenberg, 2003; Jin and Greenberg, 2003]. For instance, the role of transdifferentiation has been com-pared to the possible fusion of cells. A way to further study the underlying mechanisms is by employing single-cell-cloning technology. This method relies on the pro-duction of separate cultures which originate from one single cell. Studies on this cell population can bring in-sight as to whether fusion of different cell types is respon-sible for the phenotype shift associated with stem cell plasticity. Irrespective of the mechanisms involved, the possibility of changing the phenotype of normal adult human cells may have a significant impact on severalcell-based therapies.
By performing clonal analysis, the possibility of con-taminant progenitor cells from other tissues can be ruled out. The potential presence of subpopulations exhibiting stem cell plasticity can be studied by comparing results from differentiation experiments performed on primary-culture cells and single-cell clones derived thereof. Sev-eral groups have shown the presence of multipotent cell subpopulations in dermal tissue derived from the fore-skin [Young et al., 1995; Bartsch et al., 2005; Chen et al., 2007]. This study uses cells derived from dermal abdom-inal and breast skin.
The aim of this study was to investigate the possibility of changing the phenotype of normal human dermal FBs into adipocyte-, chondrocyte- and osteoblast-like cells using specific induction media. Previously, the induction media have been used for the differentiation of mesen-chymal stem cells [Grigoriadis et al., 1988; Pittenger et al., 1999; Zuk et al., 2002]. Human preadipocytes (PAs), which have been shown to possess stem cell plasticity, were used as controls regarding the differentiation poten-tial [Zuk et al., 2001, 2002]. Changes in cell phenotypes were continuously corroborated by routine histological staining, immunohistochemical (IHC) analysis and, in the case of osteogenic-induced cells, also by measuring alkaline phosphatase (ALP) activity. In order to further study the plasticity of FB populations, and to exclude con-taminant progenitor cells in primary culture, we also performed the experiments using single-cell-cloned FBs (sccFBs) derived from FB culture via micromanipula-tion.
Differentiation of Human Dermal Fibroblasts
Cells Tissues Organs 2010;191:105–118 107
Furthermore, the FBs and sccFBs were analyzed in re-gard to the expression of CD surface markers. The CD markers chosen have previously been employed to study mesenchymal stem cells [Pittenger et al., 1999; Zuk et al., 2002; Liu et al., 2008]. A number of CD markers included are specific for hematopoietic cells and were used in order to exclude any progenitor cells derived from circulating stem cell populations.
Materials and Methods
Isolation and Culture of Cells Human FBs and PAs were obtained from tissue specimens tak-
en from healthy patients undergoing routine plastic surgery. Tis-sue was obtained fresh from the operating theater, transported in sterile NaCl-soaked gauzes and processed within 24 h.
Primary cultures of human dermal FBs were isolated by dis-secting the dermal layer of the skin samples and subsequently incubating the fragments at 37 ° C overnight in Dulbecco’s modi-fied Eagle’s medium (DMEM) containing 2% fetal calf serum, 165 U/ml collagenase (type 1; Worthington, Lakewood, N.J., USA) and 2.5 U/ml dispase (Gibco BRL, Life Technologies, Eggenstein, Germany). Cells were expanded by incubation in FB proliferation medium ( table 1 ) in 25-cm 2 culture flasks at 37 ° C, 5% CO 2 and 95% humidity. Media were changed every other day.
Human PAs were isolated and processed based on the protocol described by Entenmann and Hauner [1996]. Fat was dissected from tissue samples and digested at 37 ° C for 30 min, being con-tinuously stirred in phosphate-buffered saline (PBS) with 260 U/ml collagenase (type 1; Worthington) and 20 mg/ml bovine se-rum albumin (BSA). The suspension was centrifuged for 5 min at 200 g , resuspended in DMEM, centrifuged for 5 min at 200 g , dis-solved in PBS and filtered serially through 100- and 70- � m cell strainers to remove debris. Finally, cells were seeded in PA prolif-eration medium ( table 1 ) in 25-cm 2 culture flasks and incubated at 37 ° C, 5% CO 2 and 95% humidity. Media were changed every other day.
The experiments were performed using cells derived from 8 separate donors.
Generation of Single-Cell Clones To produce single-cell clones, FBs were detached from the cul-
ture flask by trypsinization. A microscope equipped with a Trans-ferMan/CellTramAir micromanipulator (Eppendorf, Stockholm, Sweden) was used to transfer single cells into separate culture wells containing FB proliferation medium. Cells were allowed to attach to the culture dish and to start proliferate. The obtained single-cell clones were expanded through serial passages until sufficient amounts for the start of experiments were reached. A total of seven distinct sccFBs were studied in regard to their mul-tilineage differentiation capabilities.
Table 1. Induction and proliferation media
Medium Media Serum Supplement
Fibroblast DMEM 10% FCS 50 U/ml penicillin 50 �g/ml streptomycin
Preadipocyte DMEM/Ham’s F-12(v/v 1/1)
10% FCS 50 U/ml penicillin50 �g/ml streptomycin
Adipogenic DMEM 10% FCS 1.0 �M dexamethasone0.5 mM isobutyl-methylxanthine200 �M indomethacin10 �M insulin50 U/ml penicillin 50 �g/ml streptomycin
Chondrogenic DMEM 1% FCS 50 nM ascorbate-2-phosphate1.125 �M insulin10 ng/ml TGF-�150 U/ml penicillin 50 �g/ml streptomycin
Osteogenic DMEM 10% FCS 50 �M ascorbate-2-phosphate0.1 �M dexamethasone10 mM glycerophosphate50 U/ml penicillin 50 �g/ml streptomycin
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 108
Induction of Transdifferentiation Human FBs, sccFBs and PAs were stimulated with induction
media to induce adipogenic, chondrogenic and osteogenic differ-entiation ( table 1 ). The induction was started after 2–4 passages of the primary culture. FBs, sccFBs and PAs cultured in FB and PA proliferation medium, respectively, served as uninduced con-trol cells in all experiments.
Differentiation was confirmed by histological staining, indi-rect IHC and spectrophotometric measurement of ALP activity ( table 2 ). All experiments were performed at room temperature. Samples were visualized using Olympus BX41 and IX51 light/flu-orescence microscopes ( ! 20/0.50; ! 40/0.75, and ! 100/1.30 oil) and images captured using an Olympus DP70 CCD camera (Sol-na, Sweden).
Confirmation of Transdifferentiation Adipogenesis Adipogenic differentiation was induced by culturing FBs,
sccFBs and PAs for 2 weeks in adipogenic medium ( table 1 ). Oil red O stain was used as an indicator of intracellular lipid accumu-lation [Kutt and Tsaltas, 1959]. Briefly, cells were rinsed in PBS, fixed for 60 min in 4% neutral buffered paraformaldehyde (NBP) and again rinsed in PBS. Subsequently, cells were stained with 0.3% oil red O solution (Sigma, St. Louis, Mo., USA) in 36% tri-ethylphosphate for 30 min and rinsed in distilled water ( d H 2 O).
Expression of perilipin in intracellular lipid droplets was visu-alized using IHC. Briefly, cells were fixed for 15 min in 4% NBP, rinsed in PBS and incubated with blocking serum (5% BSA) for1 h. Cells were incubated for 1 h with polyclonal antibodies spe-cific for perilipin (dilution 1: 200; Research Diagnostics, Flanders, N.J., USA). After being rinsed in PBS, cells were incubated for 1 h with a fluorescein-isothiocyanate-conjugated secondary anti-body (dilution 1: 400; Research Diagnostics). Finally, cells were rinsed in PBS and mounted using fluorescent mounting medium (DakoCytomation, Stockholm, Sweden). Negative controls in-cluded omission of primary antibodies. Primary and secondary antibodies used in this study have been employed in several other studies and tested with regard to both specificity and cross-reac-tivity [Blanchette-Mackie et al., 1995; Souza et al., 1998; Marti-nez-Botas et al., 2000; Wolins et al., 2001].
Ten high power fields ( ! 40) of adipogenic-induced FBs and PAs and uninduced control cells stained with oil red O were ob-
served. Stained and unstained cells were counted in order to ob-tain percent differentiation. Groups were compared using a non-parametric Kruskal-Wallis test with Dunn’s post test. A p value ! 0.05 was considered significant.
Chondrogenesis Chondrogenic differentiation was induced by culturing FBs
and PAs in monolayer for 4 weeks in chondrogenic induction me-dium (CM; table 1 ) and analyzed using Alcian blue staining as an indicator of sulfated glucose-amino-glycan-rich extracellular matrix [Mason, 1971; Goldstein and Horobin, 1974]. Briefly, cells were rinsed in PBS, fixed for 15 min in 4% NBP and rinsed in d H 2 O. Subsequently, cells were stained for 15 min with 1% Alcian blue (pH 1.0; Sigma), rinsed in d H 2 O and counterstained for5 min with 0.1% nuclear fast red (Sigma).
Additionally, tissue pellets were obtained by culturing FBs,sccFBs and PAs in CM for 3 weeks. Cells were detached from the culturing surface with a cell scraper and centrifuged at 200 g for 5 min to form a pellet, which was cultured 1 week in CM. The pel-let was fixed with NBP for 24 h, dehydrated, embedded in paraf-fin, sectioned, mounted on microscope slides and analyzed using Alcian blue stain.
The Alcian blue stainings on chondrogenic-induced FBs and PAs and uninduced control cells were subsequently studied by eight independent blinded observers using a light microscope. The observers estimated the staining on a four-degree scale (from 0 to 3), where 0 represented no detectable staining and 3 strong detectable staining. Groups were compared using a non-paramet-ric Kruskal-Wallis test with Dunn’s post test. A p value ! 0.05 was considered significant.
The expression of aggrecan and collagen II was detected using IHC. Briefly, cells were fixed for 15 min in 4% NBP, rinsed in PBS and incubated in blocking serum (5% BSA) for 1 h. Cells were in-cubated for 1 h with monoclonal antibodies to aggrecan or col-lagen II (dilution 1: 50 and 1: 100, respectively; Abcam, Cambridge, UK). After being rinsed in PBS, cells were incubated for 1 h with Alexa488-conjugated secondary antibodies (dilution 1: 500; Mo-lecular Probes). Finally, cells were rinsed in PBS and mounted us-ing ProLong � Gold mounting medium (Molecular Probes) con-taining 4 � ,6-diamidino-2-phenylindole.
Table 2. Confirmation of differentiation
Lineage Lineage-specific determinant Method of confirmation
Adipogenic general lipid accumulationperilipin expression
oil red O stainperilipin-specific polyclonal antibodies
Chondrogenic sulfated glucose-amino-glycan-rich matrixaggrecan expressioncollagen II expression
Alcian blue stainaggrecan-specific monoclonal antibodiescollagen-II-specific monoclonal antibodies
Osteogenic calcified matrix productionosteocalcin and osteonectin expressionALP activity
von Kossa stainosteocalcin- and osteonectin-specific polyclonal antibodiesALP assay
Differentiation of Human Dermal Fibroblasts
Cells Tissues Organs 2010;191:105–118 109
Osteogenesis Osteogenic differentiation was induced by culturing FBs,
sccFBs and PAs in monolayer in osteogenic medium ( table 1 ). Af-ter 2 weeks, cells were analyzed for ALP activity using a spectro-photometric assay adapted from Magnusson and Farley [2002]. Briefly, a buffer (pH 9.8) containing 0.4% p-nitrophenyl phos-phate (Sigma), 10.5% diethanolamine (Sigma) and 1 m M MgCl 2 was added to confluent cell populations. Optical density (OD) was measured at 405 (absorbance maximum of p-nitrophenyl phos-phate) and 490 nm (control wavelength) after 5, 10 and 15 min using a DU640 spectrophotometer (Beckmann Instruments, Ful-lerton, Calif., USA) by adding 100 � l of supernatant to a cuvette. OD 490 nm was subtracted from OD 405 nm to compensate for unspe-cific absorbance fluctuations. The obtained results were statisti-cally compared using linear regression analysis in GraphPad Prism 5.0 (GraphPad Software, La Jolla, Calif., USA).
After 4 weeks in osteogenic medium, cells were analyzed using von Kossa staining as an indicator of calcified extracellular ma-trix [Bills et al., 1974]. Cells were rinsed in PBS, fixed for 15 min
in 4% NBP, rinsed in d H 2 O and stained with a 5% silver nitrate solution for 1 h under a 60-watt light. Cells were again rinsed in d H 2 O, treated with 5% sodium thiosulfate (Na 2 O 3 S 2 ) for 5 min and counterstained with 0.1% nuclear fast red for 5 min.
Additionally, tissue pellets were obtained by culturing FBs,sccFBs and PAs in osteogenic medium for 4 weeks. Cells were de-tached from the culture surface with a cell scraper and centri-fuged at 200 g for 5 min. The obtained pellet was fixed with 4% NBP for 24 h, dehydrated, embedded in paraffin, sectioned, mounted on microscope slides and analyzed using von Kossa stain.
The von Kossa stainings of osteogenic-induced FBs and PAs and uninduced control cells were subsequently studied by eight independent blinded observers using a light microscope. The ob-servers estimated the staining on a four-degree scale (0–3), where 0 represented no staining and 3 strong staining. Groups were compared using a non-parametric Kruskal-Wallis test with Dunn’s post test. A p value ! 0.05 was considered significant.
Table 3. CD marker antibodies and isotype controls used in the present study
Antigen Protein Function Cells Supplier Catalog No.
CD4IgG1
coreceptor for TCR binds MHC II Th cells, T-regulatory cells ImmunoTools 21270043S
CD8IgG2a
coreceptor for TCR binds MHC I Tc cells ImmunoTools 21270083S
CD9IgG1
motility relatedprotein-1, tetraspanin 29
adhesion ? ImmunoTools 21270093S
CD13IgG1
alanineaminopeptidase
trimming of peptides early committed granulocyteand monocyte prog.
ImmunoTools 21330133S
CD14IgG1
coreceptor, patternrecognition receptor
recognition oflipopolysaccaride
macrophages ImmunoTools 21270143S
CD19IgG1
Leu-12, lymphocytesurface antigen
coreceptor together withCD21 and CD81
B cells ImmunoTools 21270193S
CD29IgG1
integrin �1 migration most cells ImmunoTools 21270293S
CD34IgG1
adhesion factor attachment and migration ofhematopoietic stem cells
endothelial cells BD Biosciences 555821
CD44IgG1
hyaluronic acidreceptor
homing and migration ofstem cells
early T cells ImmunoTools 21270443S
CD45IgG1
involved in TCR signaling leukocytes ImmunoTools 21270453S
CD90IgG1
Thy-1 cell-cell and cell-matrixinteractions
stem cells Serotec MCA90F
CD105IgG1
SH2 component in TGF-�receptor system
endothelial cells, stem cells ImmunoTools 21271053S
CD106IgG1
VCAM-1 endothelial adhesionmolecule, binds VLA-4
bone marrow mesenchymal stem cells
BD Biosciences 551146
IgG1 isotype ImmunoTools 21335013SIgG2a isotype ImmunoTools 21335023S
MHC = Major histocompatibility complex; TCR = T-cell receptor; TGF = transforming growth factor; prog. = progenitor.
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 110
The expression of osteocalcin and osteonectin was detected us-ing IHC. Briefly, cells were fixed for 15 min in 4% NBP, rinsed in PBS and incubated in blocking serum (5% BSA) for 1 h. Cells were incubated for 1 h with polyclonal antibodies to osteocalcin (dilu-tion 1: 500) or osteonectin (dilution 1: 1,000; Chemicon, Temec-ula, Calif., USA). After being rinsed in PBS, cells were incubated for 1 h with Alexa546-conjugated secondary antibodies (dilution 1: 500; Molecular Probes). Finally, cells were rinsed in PBS and mounted using fluorescent mounting medium (DakoCytoma-tion). Negative controls included omission of primary antibodies.
Flow-Cytometric Analysis Normal FBs and three sccFB populations from separate do-
nors were analyzed using flow cytometry at eight different occa-sions. Each cell population was divided into several samples and each reaction only contained one fluorescently labeled antibody. All experiments were performed in duplicate. Three confluent 75-cm 2 culture flasks seeded with FBs or sccFBs were used in each experiment. Cells were enzymatically detached using trypsin/EDTA and washed subsequently in FB proliferation medium and ice-cold PBS. Cells were filtered through a 100- � m cell strainer to remove cell clusters and debris. The cells were counted in a Bürker chamber to ensure that 0.5 ! 10 6 to 1 ! 10 6 cells were used for each reaction; 10 � l of fluorescein-isothiocyanate-conju-gated antibodies were added to each reaction with a total volume of 0.5 ml. The antibodies used in this study are listed in table 3 . With support of previous experiments where antibodies directed towards CD4, CD9, CD14, CD19, CD34, CD45 and CD106 did not stain the FBs or sccFBs, these antibodies were pooled in one sam-ple. Controls included omission of antibodies and matched iso-type controls. Samples were incubated for 30 min on ice and cen-trifuged at 200 g for 5 min. The obtained cell pellet was washed twice in PBS and resuspended in 0.5 ml of ice-cold PBS before analysis using a BD LSR flow cytometer (Becton Dickinson, Brondby, Denmark). The mean fluorescence intensity was com-pared between CD marker and the corresponding isotype con-trol, and the fold increases were calculated. Furthermore, thepercentages of positively labeled FBs and sccFBs were compared. Statistical analysis was performed using a non-parametric Mann-Whitney test. A p value ! 0.05 was considered significant.
Results
Normal human FBs and PAs obtained from tissue specimens taken from patients undergoing routine plas-tic surgery were used in the present study. Cells were ex-panded in vitro and showed typical morphological ap-pearance for both cell types. Cells from every donor were split into four groups, which were incubated in induction media optimized for adipogenic, chondrogenic and os-teogenic differentiation as well as FB or PA proliferation media ( table 1 ). When transferring single cells to separate culture wells, approximately 15% of the cells proliferated and resulted in sccFB populations. Seven of the single-cell clones obtained were split into four groups and incu-
bated in induction or proliferation media in the same manner as the primary FBs and PAs.
Adipogenesis Adipogenic differentiation was evaluated using oil red
O staining as well as IHC, using antibodies to perilipin ( fig. 1 ). After 2 weeks of induction with adipogenic cells in the FB, sccFB and PA cultures had adopted a more rounded shape and displayed intracellular droplets. These droplets accumulated, the oil red O stain indicat-ing the presence of lipids. The droplets showed specific immunostaining when antibodies to perilipin were used, further strengthening the hypothesis that the cells had differentiated towards adipocytes. Uninduced control cells did not stain positively for oil red O or perilipin. Five out of seven single-cell clones showed signs of adipogen-ic differentiation.
The percentages of oil-red-O-positive cells were 23(n = 10, SD = 2.281) and 45% (n = 10, SD = 6.6) in the ad-ipogenic-induced FBs and PAs, respectively ( fig. 2 a). No oil-red-O-positive cells could be found in the uninduced control cells. The difference in numbers of oil-red-O-pos-itive cells was statistically significant when comparing ad-ipogenic-induced and uninduced control cells (p ! 0.01). The difference in oil-red-O-positive cells in the adipogen-ic-induced FBs and PAs was not statistically significant.
Chondrogenesis Chondrogenic differentiation was evaluated using Al-
cian blue staining. After 2 weeks of induction in CM, FBs, sccFBs and PAs displayed a radical change in the growth pattern, and formation of nodules and ridges with a high density of cells and extracellular matrix was observed. By staining with Alcian blue, the extracellular matrix was shown to be composed of sulfated glucose-amino-gly-cans, indicating an ongoing synthesis of cartilage-like tissue ( fig. 3 ). Chondrogenic-induced FBs and sccFBs dis-played a more intense staining compared to chondrogen-ic-induced PAs. Uninduced control cells kept their phe-notypes throughout the experiments and did not stain positively using Alcian blue in all evaluations. The Alcian blue staining was most prominent in the areas with high cell density. To further evaluate chondrogenic differen-tiation, FBs, sccFBs and PAs were induced using CM and cultured in a pellet. In all cultures, this resulted in a large amount of extracellular matrix stained intensively using Alcian blue stain ( fig. 3 ). Uninduced control cell pellets did not form any Alcian-blue-positive extracellular ma-trix. All seven single-cell clones showed signs of chondro-genic differentiation.
Differentiation of Human Dermal Fibroblasts
Cells Tissues Organs 2010;191:105–118 111
a b c
d e f
g h i
Fig. 1. Adipogenic differentiation was induced in FBs ( a ), PAs ( d ) and sccFB ( g ) stained with oil red O to display intracellular lipid droplets (arrows). Arrowheads indicate cells without lipid accu-mulation. No lipid droplets were detected in uninduced FBs ( b ),
PAs ( e ) or sccFBs ( h ) (arrowheads). Adipogenic-induced FBs ( c ), PAs ( f ) and sccFBs (i) stained with antibodies towards perilipin. Note immunoreactivity around the intracellular lipid droplets. Bar = 100 �m ( a , b , d , e ), 10 �m ( c , f , i ) and 50 � m ( g, h ).
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 112
The observers’ scoring of Alcian blue stainings showed significant differences when comparing chondrogenic-induced and uninduced FBs (p ! 0.001, n = 8) and PAs(p ! 0.05, n = 8; fig. 2 b). The observers’ scoring of chon-drogenic-induced FBs was higher than the scoring of chondrogenic-induced PAs. This difference was not sig-nificant.
Furthermore, chondrogenic differentiation of FBs and PAs cultured in pellet with CM was shown by positive immunostaining for aggrecan ( fig. 4 ). The FBs also dis-played positive immunostaining for collagen II ( fig. 4 ). Uninduced control cells did not stain positively for ag-grecan or collagen II.
Osteogenesis After 2 weeks of culture, FBs, sccFBs and PAs induced
with osteogenic induction medium displayed clusters with high cell density and large amounts of extracellular matrix. Linear regression analysis of spectrophotometric results showed a significantly higher increase in ALPactivity in osteogenic-induced cells compared to unin-duced cells (p ! 0.05, r 2 1 0.89 in all observations; fig. 2 c).
After 4 weeks of induction, FBs, sccFBs and PAs had formed dense extracellular matrix which positive von
Kossa staining confirmed to be calcified ( fig. 5 ). To fur-ther visualize the calcification, cells and extracellular matrix were scraped from the culture dish and centri-fuged to form pellets, which after analysis displayed positive von Kossa staining. The staining was seen in the extracellular matrix and not in cells. Uninduced control cells did not produce any calcified extracellular matrix.
The observers’ scoring of von Kossa stainings showed significant differences when comparing osteogenic-in-duced and uninduced FBs (p ! 0.001, n = 8) and PAs (p ! 0.05, n = 8; fig. 2 d). The observers’ scoring of osteogenic-induced FBs was higher than the scoring of osteogenic-induced PAs. This difference was not significant.
Furthermore, osteogenic differentiation of FBs, scc-FBs and PAs was shown by positive immunostaining for osteocalcin as well as for osteonectin ( fig. 5 ). The immu-nostaining was most evident in close proximity to the cells and was uniformly spread throughout the observed specimens. Uninduced control cells did not stain posi-tively for osteocalcin or osteonectin ( fig. 5 ).
All seven single-cell clones showed signs of osteogenic differentiation.
0
20
40
60
Oil
-re
d-O
-po
siti
ve
ce
lls
(%)
A-FB A-PA
***
FB PA
**
a
0
1
2
3
Ob
serv
er
sco
re
C-FB C-PA
*
FB PA
***
b
0
1
3
4
OD
(40
5–
49
0n
m)
0 105 15
c
0
1
2
3
Ob
serv
er
sco
re
O-FB O-PA
**
FB PA
***
d
2
FBO-FBPAO-PA
Time (min)
Fig. 2. Scoring of routine histological staining and ALP measurements. a The percentages of oil-red-O-stained cells in adipogenic-induced FBs (A-FB), adipo-genic-induced PAs (A-PA), uninduced FBs (FB) and uninduced PAs (PA) are shown. n = 10 in all groups. * * * p ! 0.005, * * p ! 0.01. b Independent observers’ scoring of Alcian blue staining intensity of chondro-genic-induced FBs (C-FB), chondrogenic-induced PAs (C-PA), uninduced FBs (FB) and uninduced PAs (PA). n = 8 in all groups. * * * p ! 0.005, * p ! 0.05. c Linear regression analysis of measurements of spectrophotometric analysis of ALP activ-ity in osteogenic-induced FBs (O-FB) and PAs (O-PA). Uninduced FBs and PAs are included as reference. The y-axis indicates the difference in OD between 405 and 490 nm. d Independent observers’ scoring of von Kossa staining intensity of osteo -genic-induced FBs (O-FB), osteogenic-induced PAs (O-PA), uninduced FBs (FB) and uninduced PAs (PA). n = 8 in all groups. * * * p ! 0.005, * * p ! 0.01.
Differentiation of Human Dermal Fibroblasts
Cells Tissues Organs 2010;191:105–118 113
a b c
d e f
g h
Fig. 3. Chondrogenic differentiation was induced in FBs ( a ) and PAs ( d ) grown in monolayer stained with Alcian blue. Note extra-cellular matrix (arrowheads) staining. No positive Alcian blue staining could be detected in uninduced FBs ( b ) and PAs ( e ).
Chondrogenic-induced FBs ( c ), PAs ( f ) and sccFBs ( h ) cultured for 3 weeks in monolayer and 1 week in pellet stained with Alcian blue. Uninduced sccFBs cultured in monolayer and pellet did not stain positively ( g ). Bar = 100 � m.
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 114
Flow-Cytometric Analysis SccFBs showed a statistically significantly higher in-
crease in CD13 (p ! 0.01), CD29 (p ! 0.05), CD44 (p ! 0.005), CD 90 (p ! 0.01) and CD105 (p ! 0.01) compared to primary-culture FBs ( fig. 6 ). No other tested CD mark-er displayed a significant change in fold increase. The percentage of positive cells compared to matched isotype
controls was higher when staining for CD13 (p ! 0.01), CD44 (p ! 0.005) and CD105 (p ! 0.05) ( fig. 7 ). CD29 and CD90 did not show any statistically significant difference between FBs and sccFBs. No statistically significant dif-ference was detected in FBs analyzed after 2–4 passages or 6 10 passages (data not shown).
a b
c d
e f
Fig. 4. Chondrogenic differentiation, IHC towards aggrecan and collagen II: chon-drogenic-induced FBs ( a ), uninduced FBs ( b ), chondrogenic induced PAs ( c ) and un-induced PAs ( d ) stained with antibodies towards aggrecan. Chondrogenic induced FBs ( e ) and uninduced FBs ( f ) stainedwith antibodies towards collagen II. Note extracellular immunoreactivity (arrows). 4 � ,6-Diamidino-2-phenylindole-stained nuclei are indicated by arrowheads. Bar = 50 � m.
Differentiation of Human Dermal Fibroblasts
Cells Tissues Organs 2010;191:105–118 115
Discussion
In recent years, PAs have been proposed as a novel source of autologous adult stem cells [Zuk et al., 2001, 2002; Heydarkhan-Hagvall et al., 2008]. The results pre-sented here demonstrate that the differentiation potential of FBs seems at least as potent as that of PAs, at least in the case of chondrogenic and osteogenic induction. The results of the observers’ scoring in the case of cartilage- and bone-like tissue formation imply that the induced FBs produce matrix resembling cartilage and bone to a higher extent than PAs. In the case of adipogenic differ-entiation, adipogenic-induced PAs showed a higher num-ber of positively stained cells when using the oil red O stain compared to FBs (non-significant). These findings may be explained by the evident adipogenic ‘priming’ of PAs.
The cartilage-like tissue formation occurred to the highest extent in areas with high cell density. This implies the need for a 3D structure in the case of chondrogenic differentiation. This statement is further strengthened by the higher degree of matrix staining of cells cultured in a pellet and consistent with reports from other observers [Chen et al., 2007].
During the last years, there has been an intense ongo-ing debate about the mechanisms behind the phenome-non of stem cell plasticity and there are a variety of dif-ferent hypotheses on the subject [Eisenberg and Eisen-berg, 2003; Jin and Greenberg, 2003; Camargo et al., 2004]. For instance, the role of transdifferentiation has been compared to the possible fusion of cells. The results obtained from our studies on single-cell-cloned FBs in-dicate that fusion between different cell types does not play a significant role in the phenotypic shift of FBs. The
a b c d e
f g h i j
k l m n o
Fig. 5. Osteogenic differentiation was induced in FBs ( a ), PAs ( c ) and sccFBs ( e ). Uninduced FBs ( b ) and PAs ( d ) are also shown. Von Kossa staining reveals mineralized extracellular matrix (long arrows). Osteogenic-induced FBs ( f ), PAs ( h ) and sccFBs ( j ), and uninduced FBs ( g ) and PAs (i) stained with antibodies towards osteocalcin. Arrowheads indicate the presence of osteocalcin. Os-
teogenic-induced FBs ( k ), PAs ( m ) and sccFBs ( o ), and uninduced FBs ( l ) and PAs ( n ) stained with antibodies towards osteonectin. Note intracellular immunoreactivity (arrowheads) and granule-like structures (short arrows). Asterisk marks the location of the nuclei. Bar = 100 �m ( a–e ) and 20 � m ( f–o ).
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 116
0
0.5
1.0
1.5
Fo
ldin
cre
ase
FB sccFB
CD4, CD9, CD14, CD19,CD34, CD45, CD106
0 0 0
0.5 5 1
1.0 10 2
1.5 15 3
Fo
ldin
cre
ase
Fo
ldin
cre
ase
Fo
ldin
cre
ase
FB FB FBsccFB sccFB sccFB
CD13 CD29CD8
** *
0
10
30
40
Fo
ldin
cre
ase
FB sccFB
CD44
***
0 0
10
2
304
50 6
Fo
ldin
cre
ase
Fo
ldin
cre
ase
FB FBsccFB sccFB
CD90 CD105
** **
20
40
20
Fig. 6. The mean fold increases in CD markers in FBs and sccFBs are shown. * p ! 0.05, * * p ! 0.01, * * * p ! 0.005.
0
20
100
80
Po
siti
ve
ce
lls
(%)
FB sccFB
CD13
**
40
60
0 0
20 20
100 100
80 80
Po
siti
ve
ce
lls
(%)
Po
siti
ve
ce
lls
(%)
FB FBsccFB sccFB
CD29 CD44
***
40 40
60 60
0
20
100
80
Po
siti
ve
ce
lls
(%)
FB sccFB
CD90
40
60
0
20
100
80
Po
siti
ve
ce
lls
(%)
FB sccFB
CD105
*
40
60
Fig. 7. The percentages of positively labeled FBs and sccFBs compared to matched isotype controls are shown. * p ! 0.05, * * p ! 0.01, * * * p ! 0.005.
Differentiation of Human Dermal Fibroblasts
Cells Tissues Organs 2010;191:105–118 117
ulation derived from adult human abdomen and breast skin. Future studies focusing on other lineages of differ-entiation might show how extensive the FB differentia-tion potential is. Studies which focus on the stability of the phenotypic shift after induction, both in vitro and in vivo, are needed. However, the findings presented here, showing that one of the most basic cell types in the hu-man body can differentiate into at least three other dif-ferent cell types, underlines the biological and medical importance of this study. These findings may be of great value when engineering human tissues and may become useful in a plethora of clinical situations to regenerate lost and damaged tissues.
Conclusion
In conclusion, the current study shows that normal human dermal FBs can differentiate into adipocyte-, chondrocyte- and osteoblast-like cells in vitro when sub-jected to specific induction media. Furthermore, thiswas not due to contamination of progenitor cells or cell fusion, as revealed by experiments performed using sccFBs.
Acknowledgments
This work was supported by grants from the Swedish Founda-tion for Strategic Research. We thank Dr. F. Huss for valuable discussions, J. Rakar for assistance with immunohistochemical stainings, J.-I. Jönsson and F. Sjögren for advice regarding flow cytometry, and A. Lönn and K. Briheim for excellent technical assistance.
fact that all of the seven sccFBs successfully differentiated towards several distinct lineages implies that at least a subpopulation of the FBs is multipotent. This is in line with publications exploring the inherent plasticity of der-mally derived FBs while employing clonal analysis [Toma et al., 2001; Bartsch et al., 2005; Chen et al., 2007]. Our results indicate that the multilineage potential of dermal FBs is more uniform than results presented in previous reports. Furthermore, the results obtained using clonal analysis exclude the presence of contaminant tissue-spe-cific progenitor cells as an explanation of our findings.
The flow-cytometric analysis of the FBs and sccFBs revealed a set of CD markers upregulated on the sccFBs. CD13, CD44 and CD105 were present on a higher num-ber of cells in the sccFB population as compared to the FB population. CD29 and CD90 were not, and the higher fold increase is attributed to a higher antigen density on the cell surfaces. These markers have been employed in other studies focusing on adult stem cells from various tissues [Pittenger et al., 1999; Young et al., 1999; Zuk et al., 2002; Bartsch et al., 2005]. The CD marker profile can be used in order to further characterize and purify the sccFB populations. The lack of CD106 expression in the present study suggests that the cells investigated are not mesenchymal stem cells according to Liu et al. [2008].
Clearly, more remains to be learned about the ability of normal human dermal FBs to change their phenotype. The inherent ‘stem cell plasticity’ of human FBs has been proposed by other groups and our results provide further proof for this statement [Young et al., 1995; Toma et al., 2001; Bartsch et al., 2005; Chen et al., 2007; Takahashi et al., 2007; Lowry et al., 2008]. To our knowledge, this is the first study that shows the potential plasticity in a FB pop-
References
Alessandri, G., S. Pagano, A. Bez, A. Benetti, S. Pozzi, G. Iannolo, M. Baronio, G. Invernici, A. Caruso, C. Muneretto, G. Bisleri, E. Para-ti (2004) Isolation and culture of human muscle-derived stem cells able to differenti-ate into myogenic and neurogenic cell lin-eages. Lancet 364: 1872–1883.
Bajada, S., I. Mazakova, J.B. Richardson, N. As-hammakhi (2008) Updates on stem cells and their applications in regenerative medicine. J Tissue Eng Regen Med 2: 169–183.
Bartsch, G., J.J. Yoo, P. De Coppi, M.M. Siddiqui, G. Schuch, H.G. Pohl, J. Fuhr, L. Perin, S. Soker, A. Atala (2005) Propagation, expan-sion, and multilineage differentiation of hu-man somatic stem cells from dermal progen-itors. Stem Cells Dev 14: 337–348.
Bills, C.E., H. Eisenberg, S.L. Pallante (1974) Complexes of organic acids with calcium phosphate: the von Kossa stain as a clue to the composition of bone mineral. Johns Hopkins Med J 128: 194–207.
Blanchette-Mackie, E.J., N.K. Dwyer, T. Barber, R.A. Coxey, T. Takeda, C.M. Rondinone, J.L. Theodorakis, A.S. Greenberg, C. Londos (1995) Perilipin is located on the surface lay-er of intracellular lipid droplets in adipo-cytes. J Lipid Res 36: 1211–1226.
Camargo, F.D., S.M. Chambers, M.A. Goodell (2004) Stem cell plasticity: from transdiffer-entiation to macrophage fusion. Cell Prolif 37: 55–65.
Chen, F.G., W.J. Zhang, D. Bi, W. Liu, X. Wei, F.F. Chen, L. Zhu, L. Cui, Y. Cao (2007) Clonal analysis of nestin(-) vimentin(+) multipotent fibroblasts isolated from human dermis. J Cell Sci 120: 2875–2883.
Dragoo, J.L., G. Carlson, F. McCormick, H. Khan-Farooqi, M. Zhu, P.A. Zuk, P. Ben-haim (2007) Healing full-thickness cartilage defects using adipose-derived stem cells. Tissue Eng 13: 1615–1621.
Eisenberg, L.M., C.A. Eisenberg (2003) Stem cell plasticity, cell fusion, and transdifferentia-tion. Birth Defects Res C Embryo Today 69: 209–218.
Junker /Sommar /Skog /Johnson /Kratz
Cells Tissues Organs 2010;191:105–118 118
Entenmann, G., H. Hauner (1996) Relationship between replication and differentiation in cultured human adipocyte precursor cells. Am J Physiol 270: C1011–C1016.
Fossum, M., A. Nordenskjold, G. Kratz (2004) Engineering of multilayered urinary tissue in vitro. Tissue Eng 10: 175–180.
Gallico, G.G., 3rd, N.E. O’Connor, C.C. Comp-ton, O. Kehinde, H. Green (1984) Permanent coverage of large burn wounds with autolo-gous cultured human epithelium. N Engl J Med 311: 448–451.
Gao, J., J.Q. Yao, A.I. Caplan (2007) Stem cells for tissue engineering of articular cartilage. Proc Inst Mech Eng [H] 221: 441–450.
Goldstein, D.J., R.W. Horobin (1974) Surface staining of cartilage by Alcian blue, with ref-erence to the role of microscopic dye aggre-gates in histological staining. Histochem J 6: 175–184.
Grigoriadis, A.E., J.N. Heersche, J.E. Aubin (1988) Differentiation of muscle, fat, carti-lage, and bone from progenitor cells present in a bone-derived clonal cell population: ef-fect of dexamethasone. J Cell Biol 106: 2139–2151.
Gronthos, S., D.M. Franklin, H.A. Leddy, P.G. Robey, R.W. Storms, J.M. Gimble (2001) Sur-face protein characterization of human adi-pose tissue-derived stromal cells. J Cell Physiol 189: 54–63.
Guilak, F., K.E. Lott, H.A. Awad, Q. Cao, K.C. Hicok, B. Fermor, J.M. Gimble (2006) Clonal analysis of the differentiation potential of human adipose-derived adult stem cells. J Cell Physiol 206: 229–237.
Heydarkhan-Hagvall, S., K. Schenke-Layland, J.Q. Yang, S. Heydarkhan, Y. Xu, P.A. Zuk, W.R. MacLellan, R.E. Beygui (2008) Human adipose stem cells: a potential cell source for cardiovascular tissue engineering. Cells Tis-sues Organs 187: 263–274.
Jaiswal, N., S.E. Haynesworth, A.I. Caplan, S.P. Bruder (1997) Osteogenic differentiation of purified, culture-expanded human mesen-chymal stem cells in vitro. J Cell Biochem 64: 295–312.
Jaiswal, R.K., N. Jaiswal, S.P. Bruder, G. Mbala-viele, D.R. Marshak, M.F. Pittenger (2000) Adult human mesenchymal stem cell differ-entiation to the osteogenic or adipogenic lin-eage is regulated by mitogen-activated pro-tein kinase. J Biol Chem 275: 9645–9652.
Jin, K., D.A. Greenberg (2003) Tales of transdif-ferentiation. Exp Neurol 183: 255–257.
Kutt, H., T.T. Tsaltas (1959) Staining properties of oil red O and a method of partial purifica-tion of the commercial product. Clin Chem 5: 149–160.
Liu, F., Y. Akiyama, S. Tai, K. Maruyama, Y. Kawaguchi, K. Muramatsu, K. Yamaguchi (2008) Changes in the expression of CD106, osteogenic genes, and transcription factors involved in the osteogenic differentiation of human bone marrow mesenchymal stem cells. J Bone Miner Metab 26: 312–320.
Lowry, W.E., L. Richter, R. Yachechko, A.D. Pyle, J. Tchieu, R. Sridharan, A.T. Clark, K. Plath (2008) Generation of human induced pluripotent stem cells from dermal fibro-blasts. Proc Natl Acad Sci USA 105: 2883–2888.
Magnusson, P., J.R. Farley (2002) Differences in sialic acid residues among bone alkaline phosphatase isoforms: a physical, biochemi-cal, and immunological characterization. Calcif Tissue Int 71: 508–518.
Martinez-Botas, J., J.B. Anderson, D. Tessier, A. Lapillonne, B.H. Chang, M.J. Quast, D. Go-renstein, K.H. Chen, L. Chan (2000) Absence of perilipin results in leanness and reverses obesity in Lepr(db/db) mice. Nat Genet 26: 474–479.
Mason, R.M. (1971) Observations on the glycos-aminoglycans of aging bronchial cartilage studied with Alcian blue. Histochem J 3: 421–434.
Olsson, M.J., L. Juhlin (1992) Melanocyte trans-plantation in vitiligo. Lancet 340: 981.
Passier, R., L.W. van Laake, C.L. Mummery (2008) Stem-cell-based therapy and lessons from the heart. Nature 453: 322–329.
Peterson, L., T. Minas, M. Brittberg, A. Nilsson, E. Sjogren-Jansson, A. Lindahl (2000) Two- to 9-year outcome after autologous chondro-cyte transplantation of the knee. Clin Or-thop Relat Res 374: 212–234.
Pittenger, M.F., A.M. Mackay, S.C. Beck, R.K. Jaiswal, R. Douglas, J.D. Mosca, M.A. Moor-man, D.W. Simonetti, S. Craig, D.R. Marshak (1999) Multilineage potential of adult hu-man mesenchymal stem cells. Science 284: 143–147.
Schultz, S.S., P.A. Lucas (2006) Human stem cells isolated from adult skeletal muscle dif-ferentiate into neural phenotypes. J Neurosci Methods 152: 144–155.
Souza, S.C., L.M. de Vargas, M.T. Yamamoto, P. Lien, M.D. Franciosa, L.G. Moss, A.S. Green-berg (1998) Overexpression of perilipin A and B blocks the ability of tumor necrosis factor alpha to increase lipolysis in 3T3-L1 adipocytes. J Biol Chem 273: 24665–24669.
Svendsen, C. (2008) Stem cells and Parkinson’s disease: toward a treatment, not a cure. Cell Stem Cell 2: 412–413.
Takahashi, K., K. Tanabe, M. Ohnuki, M. Narita, T. Ichisaka, K. Tomoda, S. Yamanaka (2007) Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell 131: 861–872.
Toma, J.G., M. Akhavan, K.J. Fernandes, F. Bar-nabe-Heider, A. Sadikot, D.R. Kaplan, F.D. Miller (2001) Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nat Cell Biol 3: 778–784.
Wolins, N.E., B. Rubin, D.L. Brasaemle (2001) TIP47 associates with lipid droplets. J Biol Chem 276: 5101–5108.
Yoo, J.U., T.S. Barthel, K. Nishimura, L. Solcha-ga, A.I. Caplan, V.M. Goldberg, B. Johnstone (1998) The chondrogenic potential of human bone-marrow-derived mesenchymal pro-genitor cells. J Bone Joint Surg Am 80: 1745–1757.
Young, H.E., C. Duplaa, M. Romero-Ramos, M.F. Chesselet, P. Vourc’h, M.J. Yost, K. Er-icson, L. Terracio, T. Asahara, H. Masuda, S. Tamura-Ninomiya, K. Detmer, R.A. Bray, T.A. Steele, D. Hixson, M. el-Kalay, B.W. To-bin, R.D. Russ, M.N. Horst, J.A. Floyd, N.L. Henson, K.C. Hawkins, J. Groom, A. Parikh, L. Blake, L.J. Bland, A.J. Thompson, A. Kirincich, C. Moreau, J. Hudson, F.P. Bow-yer, 3rd, T.J. Lin, A.C. Black, Jr. (2004) Adult reserve stem cells and their potential for tis-sue engineering. Cell Biochem Biophys 40: 1–80.
Young, H.E., M.L. Mancini, R.P. Wright, J.C. Smith, A.C. Black, Jr., C.R. Reagan, P.A. Lu-cas (1995) Mesenchymal stem cells reside within the connective tissues of many or-gans. Dev Dyn 202: 137–144.
Young, H.E., T.A. Steele, R.A. Bray, K. Detmer, L.W. Blake, P.W. Lucas, A.C. Black, Jr. (1999) Human pluripotent and progenitor cells dis-play cell surface cluster differentiation mark-ers CD10, CD13, CD56, and MHC class-I. Proc Soc Exp Biol Med 221: 63–71.
Zuk, P.A., M. Zhu, P. Ashjian, D.A. De Ugarte, J.I. Huang, H. Mizuno, Z.C. Alfonso, J.K. Fraser, P. Benhaim, M.H. Hedrick (2002) Human adipose tissue is a source of multipo-tent stem cells. Mol Biol Cell 13: 4279–4295.
Zuk, P.A., M. Zhu, H. Mizuno, J. Huang, J.W. Futrell, A.J. Katz, P. Benhaim, H.P. Lorenz, M.H. Hedrick (2001) Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 7: 211–228.
IIPAPER
Engineering three-dimensional cartilage- and bone-like tissues using human dermal fibroblasts andmacroporous gelatine microcarriers
Pehr Sommar a, Sofia Pettersson a, Charlotte Ness a,b, Hans Johnson b,Gunnar Kratz a, Johan P.E. Junker a,*
a Laboratory for Reconstructive Plastic Surgery, Division of Plastic and Hand Surgery, Department of Clinical andExperimental Medicine, Linkoping University, Linkoping, Swedenb Department of Surgery, Section of Plastic Surgery and Burn Centre, Haukeland University Hospital, Bergen, Norway
Received 30 June 2008; accepted 7 February 2009
KEYWORDSDermal fibroblast;Chondrogenesis;Osteogenesis;Microcarrier;Tissue engineering;Regenerative medicine
Summary The creation of tissue-engineered cartilage and bone, using cells from an easilyavailable source seeded on a suitable biomaterial, may have a vast impact on regenerativemedi-cine. While various types of adult stem cells have shown promising results, their use is accompa-nied by difficulties associated with harvest and culture. The proposed inherent plasticity ofdermally derived human fibroblasts may render them useful in tissue-engineering applications.In the present study, human dermal fibroblasts cultured on macroporous gelatine microcarriersencapsulated in platelet-rich plasma into three-dimensional constructs were differentiatedtowards chondrogenic and osteogenic phenotypes using specific induction media. The effectof flow-induced shear stress on osteogenic differentiation of fibroblasts was also evaluated.The generated tissue constructs were analysed after 4, 8 and 12 weeks using routine and immu-nohistochemical stainings as well as an enzyme activity assay. The chondrogenic-induced tissueconstructs were composed of glycosaminoglycan-rich extracellular matrix, which stained posi-tive for aggrecan. The osteogenic-induced tissue constructs were composed of mineralisedextracellular matrix containing osteocalcin and osteonectin, with cells showing an increasedalkaline phosphatase activity. Increased osteogenic differentiation was seen when applyingflow-induced shear stress to the culture. Un-induced fibroblast controls did not form cartilage-or bone-like tissues. Our findings suggest that primary human dermal fibroblasts can be usedto form cartilage- and bone-like tissues in vitro when cultured in specific induction media.ª 2009 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published byElsevier Ltd. All rights reserved.
* Corresponding author. Centre for Clinical and Experimental Research, Department of Clinical and Experimental Medicine, Faculty ofHealth Sciences, Linkoping University, SE-58185 Linkoping, Sweden. Tel.: þ46 13 224485; fax: þ46 13 223706.
E-mail address: [email protected] (J.P.E. Junker).
1748-6815/$-seefrontmatterª2009BritishAssociationofPlastic,ReconstructiveandAestheticSurgeons.PublishedbyElsevierLtd.All rightsreserved.doi:10.1016/j.bjps.2009.02.072
Journal of Plastic, Reconstructive & Aesthetic Surgery (2010) 63, 1036e1046
Bone and cartilage transplants are important elements inreconstructive surgery, but shortage of autologous tissueand possible donor sites presents a major barrier. In addi-tion, the harvesting procedures for these tissues are oftenhighly invasive. For tissues that lack an intrinsic regenera-tive ability, such as cartilage, this is a major concern.Cartilage defects require either introduction of progenitorcells or implantation of expanded autologous chondrocytesin order to restore tissue-like function.1e3 Bone tissue canbe regenerated, but treatment of large defects may benefitfrom implantation of cell-seeded biomaterials to facilitatethe natural repair.4,5 The vast number of cells required, incombination with the complex issue of tissue harvesting,remains an obstacle for these cell-based therapies.
The search for alternative cell sources has led to inves-tigations of the multilineage potential of normal adulthuman cells. Mesenchymal stem cells (MSCs) derived frombone marrow are known to contribute to the regenerationof mesenchymal tissues such as bone, cartilage, muscle,ligament, tendon, fat and stroma, and have been studiedextensively earlier.6e10 Nevertheless, the use of MSCs isaccompanied with some disadvantages, most noticeablythe potentially invasive procedures related to their harvestand the difficulty in isolating and expanding them in vitro.
Recent studies have suggested that cells with similarplasticity as MSCs can be isolated from a variety of adulttissues, including adipose tissue, skeletal muscle anddermal tissue components, demonstrating that bonemarrow is not the only source for pluripotent adult cells.11e17
In response to changes in the micro-environment, thesecells are able to differentiate and express tissue markersdistinct from their own origin, making them an interestingalternative for cell-based therapies.12,13,15,16,18
Human dermal fibroblasts represent an easily accessibleand abundant cell source, as a substantial cell yield can beobtained from a relatively small biopsy following minimallyinvasive procedures and routine cell-expansion techniques.We have previously shown that dermal fibroblasts arecapable of altering their phenotypical expression in mono-layer culture under the influence of osteogenic, chondro-genic and adipogenic induction media, suggesting a certainplasticity in this cell population (Junker et al., submitted).For further clinical use, the creation of three-dimensional(3D) tissue construct is needed, leading to the incorpora-tion of suitable biomaterials. The biomaterial shouldpromote attachment, migration and proliferation of seededcells. Optimally, the biomaterial is biodegradable andfacilitates perfusion of nutrients and oxygen beforea functioning vasculature is developed.
Macroporous microcarriers with interconnected poresoffer substantial surface areas for cell adhesion on each ofthem, thereby facilitating a homogeneous distribution ofcells and paracrine signals throughout the entire scaffold.Porcine gelatine microcarriers support in vitro growth ofvarious mammalian cells and have also been used asbiodegradable scaffolds for guided tissue regeneration andcell-based therapies in vivo.19e26 Several additionaladvantages have also been reported for this type ofmicrocarrier, including decreased de-differentiation ofchondrocytes and osteoblasts, and novel cell culturestrategies that rely on microcarrier inoculation rather thanenzymatic detachment of cells.21e23 The gelatine carriers
are easily handled in suspension and this trait enables anefficient way to fill a void with cell-seeded microcarriersupon injection. The addition of a fixative to the cell-seededmicrocarriers after implantation may enhance their void-filling properties, thus minimising the risk of displacementof the construct in vivo.
Platelet-rich plasma (PRP) has previously been proposedas a blood-derived biological glue to encapsulate cell-seededmicrocarriers in a 3D in vitro model (Pettersson et al.submitted). In this work we found that PRP preserved the 3Dstructure until sufficient extracellular matrix productioncould occur to uphold the structural integrity. The autolo-gous source of the plasma, minimising the risk of immuneresponse and the ease with which the blood can be collectedand the plasma separated, makes it a strong candidate forclinical purposes. Beneficial effects of the platelets havealso been reported, as chondrocytes and osteoblast-like cellscultured in the presence of PRP showed enhanced cellproliferation and extracellular matrix production.27,28
The aim of this study is to investigate whether primaryhuman dermal fibroblasts, subjected to osteo- and chon-drogenic-induction media, can form bone- and cartilage-like tissue in a combined microcarrier and PRP model invitro. With the support of previous work, we also aim toinvestigate whether possible synergistic effects betweenthe osteogenic-induction medium and flow-induced shearstress further enhances the osteogenic differentiation.29,30
Materials and methods
Isolation of cells
Fibroblasts were obtained from dermal tissue taken fromhealthy patients undergoing routine plastic surgery. Thetissue was stored in 0.9 % NaCl-soaked gauzes and pro-cessed within 24 h. Primary cultures of human fibroblastswere isolated by dissecting the dermal layer of the skinsamples and incubating the fragments at 37 �C overnight inDulbecco’s modified Eagle‘s medium (DMEM) (Gibco�,Invitrogen Corporation, NY, USA) containing 2% foetal calfserum (FCS) (Gibco�, Invitrogen Corporation, NY, USA),165 Uml�1 collagenase (type 1) (Gibco�, InvitrogenCorporation, NY, USA) and 2.5 Uml�1 dispase� (Gibco�,Invitrogen Corporation, Germany) with repeated agitationto dissolve tissue fragments. The resulting cell suspensionwas centrifuged at 200 g for 10 min, and the cell pellet re-suspended in fibroblast proliferation medium (FM; Table 1)and plated in 75 cm2 culture flasks in 37 �C, 5 % CO2 and 95% humidity. The medium was changed 3 times per week.
Microcarriers
CultiSpher-GL is a macroporous porcine gelatine micro-carrier with a diameter of 166e370 mm, and averageinternal pore size of approximately 30 mm (Percell BiolyticaAB, Sweden). Microcarriers were prepared according to themanufacturer’s recommendations. Briefly, dry micro-carriers were rehydrated in phosphate-buffered solutionPBS, 50 ml g�1 dry weight carriers, for a minimum of 1 h inroom temperature. Without removing the PBS, micro-carriers were sterilised by autoclaving (121 �C, 20 min,
Engineering three-dimensional cartilage- and bone-like tissues 1037
15 psi). Sterilised microcarriers were washed once withPBS, and twice with FM, and stored at 4 �C until use.
Seeding of cells on microcarriers
Fibroblasts at passage 4 were washed with 0.02 % ethyl-endiaminetetraacetic acid (EDTA) and enzymaticallydetached by incubation with EDTA (0.01%)/trypsin (0.125%)solution at 37 �C for approximately 15 min. The suspensionwas centrifuged at 200 g for 5 min, supernatant removedand cells re-suspended in 50 ml FM.
Before seeding, cells were counted using a Guava� PCA�cell counter (Guava Technologies, Inc., CA, USA). Threesamples of 100 ml each were taken from the cell suspension.Samples were diluted 1:10 in Guava ViaCount� Reagent andmeasured after 5 min. The cells were seeded at a finalconcentration of 17.6� 106 cells g�1 microcarriers and 0.2 gmicrocarriers (dry weight) were cultured in each spinnerflask (MCS 104S, 250 ml, Techne, UK).
FM was added to a final volume of 100 ml in each spinnerflask. The cells were allowed to attach to the microcarriersunder intermittent agitation (20 rpm for 5 min every h) in37 �C, 5% CO2 and 95% humidity. On the second day, theculture medium volume was adjusted to 200 ml per spinnerflask, and agitation adjusted to continuous stirring (20 rpm).A 100 ml of FM was changed three times per week. Asa control group, acellularmicrocarriers were prepared in thesame manner and transferred to spinner flasks.
To assess the viability and distribution of fibroblasts in themicrocarriers, MTT staining (3-[4.5-dimethylthiazol-2-yl]-2.5-diphenyltetrazolium bromide) (Sigma Aldrich Co, MO,USA) was performed. A 400 ml of the microcarrier suspensionwas taken from each spinner flask. A 40 ml MTT (5 mg/ml inPBS) was added to each sample and incubated in 37 �C for45 min. Staining was evaluated using an Olympus IX51inverted light microscope (�40/0.75). The MTT stainrevealed optimal cell density after approximately 14 days.
Encapsulation of microcarriers in cell cultureinserts
Cells were counted using a Guava� PCA� cell counter.A total of five 0.5 ml samples were removed from themicrocarrier suspension. A 0.3 ml of the supernatant was
extracted and replaced with EDTA (0.01%)/trypsin (0.125%)for 15e25 min until the microcarriers were dissolved.Samples were diluted 1:10 in Guava ViaCount� Reagent andmeasured twice. The mean count of viable cells was78.8� 106 g�1 microcarriers.
Cells and microcarriers in spinner flasks were allowed tosediment for 30 min following which the medium wasremoved. Microcarriers and cells were centrifuged at 100 gfor 5 min and supernatant removed.
PRPwas prepared from citratedwhole blood fromhealthydonors. Briefly, the whole blood was drawn to Vacutainer�
collection tubes (9NC) (BD Diagnostics, Plymouth, UK) con-taining citrate. The PRP was prepared by centrifugation ofthe citrated whole blood for 20 min at 220 g.
Microcarriers and PRP were mixed 1:1 (vol:vol) andplated at 400 ml in 12-well cell culture inserts with 3-mmpores (BD Biosciences, MA, USA) in order to create 3D tissueconstructs. Clotting was initiated by adding 15 ml 0.5 MCaCl2 in each insert. Samples were allowed to clot for30 min before the FM was added to each well and insert.
Induction of chondrogenic and osteogenicdifferentiation
Induction of 3D constructs started with chondrogenic orosteogenic medium (CM and OM, respectively) 24 h afterplating in inserts (Table 1). In the case of osteogenicinduction, cells were also kept in spinner flask culture in OMthroughout the experiment. Fibroblast-seeded micro-carriers maintained in FM were kept in cell culture insertsand spinner flask culture for the duration of the experiment(un-induced controls). Acellular microcarriers in cellculture insert and spinner flask culture received CM, OM orFM (acellular controls). The cell culture continued for 12weeks, and analyses were performed at 4, 8 and 12 weeks.The media was changed three times per week. All experi-ments were performed in triplicates.
Confirmation of chondrogenic and osteogenicdifferentiation
Chondrogenic and osteogenic differentiation was inducedby culturing Fibroblasts for 4, 8 and 12 weeks in chondro-genic or osteogenic medium, respectively. At each time
Table 1 Induction and proliferation media
Medium Media Serum Supplement
Fibroblast proliferation (FM) DMEM 10% FCS 50 U/ml penicillin50 mg/ml streptomycin
Chondrogenic (CM) DMEM 1% FCS 50 nM Ascorbate-2-phosphate6.25 mg/ml Insulin10 ng/ml TGF-b150 U/ml penicillin50 mg/ml streptomycin
Osteogenic (OM) DMEM 10% FCS 50 mM Ascorbate-2-phosphate1 mM Dexamethasone10 mM Glycerophosphate50 U/ml penicillin50 mg/ml streptomycin
1038 P. Sommar et al.
point, samples were fixed with 4% neutral buffered para-formaldehyde (NBP) (Apokteket, Stockholm, Sweden) for24 h, dehydrated, embedded in paraffin, sectioned andmounted on microscope slides. Chondrogenic differentia-tion was analysed by Alcian blue stain and indirect immu-nohistochemistry (IHC) directed towards aggrecan andcollagen II. Osteogenic differentiation was analysed by vonKossa stain and indirect IHC directed towards osteocalcinand osteonectin at all time points, and spectrophotometricmeasurement of alkaline phosphatase (ALP) activity after12 weeks. All experiments were performed at roomtemperature. Samples were visualised using an OlympusBX41 microscope (�20/0.50, �40/0.75 and �100/1.30 oil)and images captured using an Olympus DP70 CCD camera(Solna, Sweden).
Chondrogenesis
Alcian blue stain was used as an indicator of glycosamino-glycan rich extracellular matrix.31e33 In brief, slides wererehydrated, rinsed in dH2O and stained for 30 min with 1%Alcian blue (pH 1.0) (Sigma Aldrich Co., MO), rinsed in dH2Oand counter-stained for 5 min with 0.1 % nuclear fast red(Sigma Chemical Co., MO, USA).
The expression of aggrecan and collagen II was detectedusing IHC. Briefly, slides were rehydrated, rinsed in PBS andtreated with pepsin (1 mgml�1) (Sigma Aldrich Co., MO,USA) in Tris-HCl (0.2 M) for 15 min. Slides were rinsed in PBSand incubated in blocking serum (5% BSA) for 45 min fol-lowed by incubation for 1 h with monoclonal antibodiesdirected towards aggrecan (Abcam, UK) and collagen II(Chemicon Inc., CA, USA), respectively. Sections wererinsed in PBS and incubated for 1 h with ALEXA488-conju-gated secondary antibodies (Molecular Probes, OR). Finally,sections were rinsed in PBS and mounted using Pro-Long�Gold mounting medium (Molecular Probes, OR)containing DAPI (40, 6-diamidino-2-phenylindole).
Osteogenesis
The von Kossa stain was used as an indicator of calcifiedextracellular matrix.34 In brief, slides were rehydrated,rinsed in dH2O and stained with a 5% silver nitrate solutionfor 1 h under a 60 W light. Sections were rinsed in dH2O,treated with 5% sodium thiosulphate (Na2O3S2) for 5 minand counter-stained with 0.1% nuclear fast red (SigmaChemical Co., MO, USA) for 5 min.
The expression of osteocalcin and osteonectin wasdetected using IHC. In this process, slides were rehydrated,rinsed in PBS and treated with pepsin (1 mgml�1) (Sigma-Aldrich Co., MO, USA) in Tris-HCl (0.2 M) for 15 min. Slideswere rinsed in PBS and incubated in blocking serum(5% BSA) for 45 min followed by incubation for 1 h withpolyclonal antibodies directed towards osteocalcin andosteonectin (Chemicon Inc., CA, USA), respectively.Sections were rinsed in PBS and incubated for 1 h withALEXA488-conjugated secondary antibodies (MolecularProbes, OR). Finally, sections were rinsed in PBS andmounted using ProLong�Gold mounting medium (Molec-ular Probes, OR) containing DAPI.
After 12 weeks of culture spectrophotometric analysis ofALP activity was performed by adding a buffer (pH 9.8)containing 0.4% p-nitrophenyl phosphate (PNPP) (SigmaAldrich Co., MO, USA), 10.5% diethanolamine (DEA) (SigmaAldrich Co., MO, USA) and 1 mM MgCl2.
35 Optical density(OD) was measured at 405 nm (absorbance maximum ofPNPP) and 490 nm (control wavelength) every 5 min for30 min using a DU640 spectrophotometer (Beckman, CA,USA). The OD at 490 nm was subtracted from OD at 405 nmto compensate for unspecific absorbance fluctuations. Theresults obtained were statistically compared using linearregression analysis in GraphPad Prism 5.0 (GraphPad Soft-ware Inc., CA, USA).
Results
Normal human dermal fibroblasts from patients undergoingroutine plastic surgery were expanded in vitro. Observationof cells revealed typical fibroblast-like morphologicalcharacteristics.
By the employment of fibroblast-seeded microcarriersand PRP, 3D tissue constructs were created in cell cultureinserts. The constructs were induced towards chondrogenicand osteogenic differentiation using induction media.Osteogenic induction was also performed using fibroblast-seeded microcarriers cultured in spinner flasks. Sampleswere evaluated after 4, 8 and 12 weeks.
Chondrogenic-induction media stimulated formation ofextensive amounts of extracellular matrix, most prominentin the outer borders of the 3D constructs. By staining withAlcian blue, the extracellular matrix was shown to becomposed of glycosaminoglycans. The staining intensifiedwith time, indicating an ongoing cartilage synthesis(Figure 1). Neither the un-induced fibroblast-seededmicrocarriers nor the acellular microcarriers stained posi-tive using Alcian blue.
To further evaluate the chondrogenic differentiation,we performed immunostaining using antibodies directedtowards aggrecan (Figure 2) and collagen II (not shown).The constructs receiving induction medium were composedof aggrecan-containing extracellular matrix, but there wasno sign of collagen II synthesis in neither time point. Therewas no positive immunostaining of the un-induced fibro-blast-seeded microcarriers or the acellular microcarriers.
Osteogenic-induction medium stimulated extensiveproliferation and a morphological transformation of thefibroblasts into cuboidal-shaped cells and the formation ofa dense extracellular matrix (Figure 3). After 4 weeks ofincubation, the osteogenic-induced cultures in inserts andspinner flasks displayed positive von Kossa stain, whichconfirmed the extracellular matrix to be calcified(Figure 3). This extracellular matrix was most prominent insamples from induced spinner flasks. The extracellularmatrix increased visually in amount after 8 and 12 weeks inboth insert and spinner flask cultures. A positive staining ofthe un-induced fibroblast-seeded microcarriers cultured inspinner flasks could be detected after 12 weeks.Un-induced fibroblasts in inserts and the acellular micro-carriers showed no positive stain.
Furthermore, osteogenic differentiation was shown bypositive immunostaining for osteocalcin (Figure 4) and
Engineering three-dimensional cartilage- and bone-like tissues 1039
osteonectin (Figure 5). There was no positive immunos-taining in either un-induced fibroblast-seeded or acellularmicrocarriers.
Linear regression analysis of spectrophotometric resultsshowed a significantly higher increase of alkaline phos-phatase activity in induced cells compared to un-inducedcells (p< 0.001, r2> 0.84 in all observations). There wasa low ALP activity in un-induced fibroblasts cultured inspinner flasks and no ALP activity in un-induced fibroblastscultured in inserts (Figure 6). There was no ALP activity inthe acellular microcarriers.
Discussion
We have previously shown that normal human dermalfibroblasts, or a subpopulation of these cells, havea phenotypic potential that goes beyond the cell types oftheir resident tissue (Junker et al., submitted). In thepresent study we wanted to investigate if a 3D scaffoldallows cells from human dermis to generate cartilage- andbone-like tissues.
With the phenotypical shift of human dermal fibroblaststowards chondrocyte-like cells, a prominent production ofdense extracellular matrix was observed. This matrixstained positive for glycosaminoglycans and aggrecan. Theamount of matrix was most prominent in the superficial
parts of the 3D constructs and increased with time. Therewas no detectable production of collagen II, indicatingimmature cartilage. The accumulation of matrix in thesuperficial parts of the 3D constructs is likely due toinsufficient diffusion of induction medium. A possibleapproach for achieving higher matrix content throughoutthe constructs is by applying perfusion of induction mediumthrough the constructs.36e38 Applying dynamic compressionto the 3D constructs could also facilitate cartilageformation.39,40
Our findings are supported by results obtained in studiesconducted by groups using cells from skin and subcutaneoustissue for chondrogenic differentiation. Earlier studies haveaimed at growing cartilage using rabbit fibroblasts41 as wellas cells from rabbit adipose tissue42,43 and applied this ina clinical model. Mizuno and Glowacki have experimentallyshown extracellular matrix production resembling chon-drogenesis using a combination of human dermal fibro-blasts, collagen and de-mineralised bone powder.44,45
With the phenotypical shift of human dermal fibroblaststowards bone there was a prominent production of extra-cellular matrix in constructs cultured in inserts and spinnerflasks, which increased with time. The extracellular matrixwas distributed all through the specimens and von Kossastain confirmed the matrix to be calcified. The productionof a bone-like extracellular matrix was confirmed by posi-tive staining for osteocalcin and osteonectin. Furthermore,
Figure 1 Alcian Blue staining of chondrogenic induced fibroblasts and control fibroblasts after 4, 8 and 12 weeks. Chondrogenicinduced fibroblasts after 4(A), 8(B) and 12(C) weeks of culture in cell culture inserts stained using the Alcian Blue stain. Note thepositive staining of the extracellular matrix indicating it to be composed of glucoseaminoglycans (indicated by arrows). Thefibroblast controls cultured in fibroblast proliferation medium for 4(D), 8(E) and 12(F) weeks did not stain positively using AlcianBlue. (Original magnification �40).
1040 P. Sommar et al.
the increased ALP activity in osteogenic-induced culturesprovides more evidence of the differentiation towardsa bone-like tissue.
The extracellular matrix production was most prominentin the osteogenic-induced cells cultured in spinner flasks.For this group, the von Kossa stain and IHC analysis revealedhigh amounts of dense nodules visually resembling spon-gious bone. There was a minor positive von Kossa stainingand ALP activity in the un-induced fibroblasts cultured inspinner flasks after 12 weeks. Other studies have shownpositive effects of fluid-induced shear stress on osteoblastdifferentiation in the absence of dexamethasone.29,30
These results suggest that shear stress by itself has animpact on osteogenic differentiation, and may partiallyexplain the findings of mineralisation in the un-inducedfibroblast culture in this study. However, unlike the osteo-genic-induced fibroblasts, un-induced constructs showed nopositive staining for osteocalcin or osteonectin. Thisdistinction, along with the difference in calcification andALP activity level between induced and un-induced fibro-blasts, indicates that the stimulation factors in the induc-tion media were necessary for the ability of human dermalfibroblasts to generate bone-like tissue in vitro. Further-more, it seems evident that shearing forces greatlyenhance this process.
The results obtained using osteogenic-induced fibro-blasts in this study are supported by the research of othergroups. Earlier studies aiming at creating bone using human
adipocytes46 and genetically modified human fibro-blasts47,48 have been published.
This study is, to our knowledge, the first report of theformation of 3D cartilage- and bone-like tissues in vitrousing chondrogenic- or osteogenic-induced human dermalfibroblasts cultured on macroporous gelatine microcarriers.
The theoretical possibility of contaminant cartilage and/or bone progenitor cells giving rise to cartilage- and bone-like tissues in this study is apparent. However, ours andprevious work by others on clonal populations derived fromdermis suggest either the presence of subpopulationsexhibiting stem cell plasticity in dermis, or an inherentplasticity in normal human fibroblasts (Junker et al.,submitted).14,17
In conclusion, our results indicate the possibility ofgenerating 3D cartilage- and bone-like tissues using dermalcells. In order to study this process on the intracellularlevel, the investigation of the expression of several lineage-specific genes is another target for future studies, whichwill assist in building an understanding of the molecularmechanisms behind ‘transdifferentiation’ However, froma clinical point of view, knowledge of the mechanisms issecondary to the fact that it is possible to change thephenotype of human dermal fibroblasts in vitro. Before anyclinical use of fibroblast-derived tissues is undertaken, thetransplantation of in vitro induced cells into a host isnecessary. We have addressed this by initiating an in vivomodel to study the effects of osteogenic-induced
Figure 2 Immunostaining of chondrogenic induced fibroblasts and control fibroblasts after 4, 8 and 12 weeks. Chondrogenicinduced fibroblasts stained with antibodies towards Aggrecan after 4(A), 8(B) and 12(C) weeks of culture in cell culture inserts.Note the positive staining of the extracellular matrix indicating the presence of Aggrecan (indicated by arrows). The fibroblastcontrols cultured in fibroblast proliferation medium for 4(D), 8(E) and 12(F) weeks did not show any detectable immunostaining.Nuclei are stained blue using DAPI in all slides. (Original magnification �40).
Engineering three-dimensional cartilage- and bone-like tissues 1041
Figure 3 Von Kossa staining of osteogenic induced fibroblasts and control fibroblasts after 4, 8 and 12 weeks. Osteogenic inducedfibroblasts after 4(A), 8(B) and 12(C) weeks of culture in inserts. Note the positive staining of the extracellular matrix showing it tobe mineralised (indicated by arrows). The fibroblast controls cultured in fibroblast proliferation medium for 4(D), 8(E) and 12(F)weeks did not stain positively using von Kossa. The osteogenic induced fibroblasts cultured in stirring vessels for 4(G), 8(H) and 12(I)weeks showed large intensely stained areas throughout the specimens. The fibroblast controls cultured in fibroblast proliferationmedium in stirring vessels for 4(J), 8(K) and 12(L) weeks were gradually increasingly stained using von Kossa. The staining was not asintense as that of the osteogenic induced fibroblasts. (Original magnification �40).
1042 P. Sommar et al.
Figure 4 Immunostaining of osteogenic induced fibroblasts and control fibroblasts after 4, 8 and 12 weeks using antibodiestowards Osteocalcin. Osteogenic induced fibroblasts stained with antibodies towards osteocalcin after 4(A), 8(B) and 12(C) weeksof culture in cell culture inserts. Note the positive staining of the extracellular matrix indicating the presence of Osteocalcin(indicated by arrows). The fibroblast controls cultured in fibroblast proliferation medium for 4(D), 8(E) and 12(F) weeks did notshow any detectable immunostaining. The osteogenic induced fibroblasts cultured in stirring vessels for 4(G), 8(H) and 12(I) weeksshowed large areas displaying positive immunostaining for Osteocalcin. The fibroblast controls cultured in fibroblast proliferationmedium in stirring vessels for 4(J), 8(K) and 12(L) weeks did not stain positively using the same antibodies. Nuclei are stained blueusing DAPI in all slides. (Original magnification �40).
Engineering three-dimensional cartilage- and bone-like tissues 1043
Figure 5 Immunostaining of osteogenic induced fibroblasts and control fibroblasts after 4, 8 and 12 weeks using antibodiestowards osteonectin. Osteogenic induced fibroblasts stained with antibodies towards osteonectin after 4(A), 8(B) and 12(C) weeksof culture in cell culture inserts. Note the positive staining of the extracellular matrix indicating the presence of Osteocalcin(indicated by arrows). The fibroblast controls cultured in fibroblast proliferation medium for 4(D), 8(E) and 12(F) weeks did notshow any detectable immunostaining. The osteogenic induced fibroblasts cultured in stirring vessels for 4(G), 8(H) and 12(I) weeksshowed large areas displaying positive immunostaining for Osteonectin. The fibroblast controls cultured in fibroblast proliferationmedium in stirring vessels for 4(J), 8(K) and 12(L) weeks did not stain positively using the same antibodies. Nuclei are stained blueusing DAPI in all slides. (Original magnification �40).
1044 P. Sommar et al.
fibroblasts on fracture healing. If we can obtain cartilageand bone from a small skin biopsy, this will dramaticallyfacilitate the use of tissue engineering methods in recon-structive surgery even before we can explain the phenom-enon in detail.
Conflict of interest/funding
None.
References
1. Brittberg M, Lindahl A, Nilsson A, et al. Treatment of deepcartilage defects in the knee with autologous chondrocytetransplantation. N Engl J Med 1994;331:889e95.
2. Beiser IH, Kanat IO. Subchondral bone drilling: a treatment forcartilage defects. J Foot Surg 1990;29:595e601.
3. Sledge SL. Microfracture techniques in the treatment ofosteochondral injuries. Clin Sports Med 2001;20:365e77.
4. Cancedda R, Giannoni P, Mastrogiacomo M. A tissue engi-neering approach to bone repair in large animal models and inclinical practice. Biomaterials 2007;28:4240e50.
5. Kose GT, Korkusuz F, Korkusuz P, et al. In vivo tissue engi-neering of bone using poly(3-hydroxybutyric acid-co-3-hydroxyvaleric acid) and collagen scaffolds. Tissue Eng 2004;10:1234e50.
6. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potentialof adult human mesenchymal stem cells. Science 1999;284:143e7.
7. Jaiswal N, Haynesworth SE, Caplan AI, et al. Osteogenicdifferentiation of purified, culture-expanded human mesen-chymal stem cells in vitro. J Cell Biochem 1997;64:295e312.
8. Jaiswal RK, Jaiswal N, Bruder SP, et al. Adult human mesen-chymal stem cell differentiation to the osteogenic or adipo-genic lineage is regulated by mitogen-activated protein kinase.J Biol Chem 2000;275:9645e52.
9. Prockop DJ. Marrow stromal cells as stem cells for non-hematopoietic tissues. Science 1997;276:71e4.
10. Yoo JU, Barthel TS, Nishimura K, et al. The chondrogenicpotential of human bone-marrow-derived mesenchymalprogenitor cells. J Bone Joint Surg Am 1998;80:1745e57.
11. Young HE, Mancini ML, Wright RP, et al. Mesenchymal stemcells reside within the connective tissues of many organs. DevDyn 1995;202:137e44.
12. Zuk PA, Zhu M, Mizuno H, et al. Multilineage cells from humanadipose tissue: implications for cell-based therapies. TissueEng 2001;7:211e28.
13. Zuk PA, Zhu M, Ashjian P, et al. Human adipose tissue isa source of multipotent stem cells. Mol Biol Cell 2002;13:4279e95.
14. Bartsch G, Yoo JJ, De Coppi P, et al. Propagation, expansion,and multilineage differentiation of human somatic stem cellsfrom dermal progenitors. Stem Cells Dev 2005;14:337e48.
15. Jahoda CA, Whitehouse J, Reynolds AJ, et al. Hair follicledermal cells differentiate into adipogenic and osteogeniclineages. Exp Dermatol 2003;12:849e59.
16. Warejcka DJ, Harvey R, Taylor BJ, et al. A population of cellsisolated from rat heart capable of differentiating into severalmesodermal phenotypes. J Surg Res 1996;62:233e42.
17. Young HE, Steele TA, Bray RA, et al. Human reserve pluripotentmesenchymal stem cells are present in the connective tissuesof skeletal muscle and dermis derived from fetal, adult, andgeriatric donors. Anat Rec 2001;264:51e62.
18. Asakura A, Komaki M, Rudnicki M. Muscle satellite cells aremultipotential stem cells that exhibit myogenic, osteogenic,and adipogenic differentiation. Differentiation 2001;68:245e53.
19. Huss FR, Junker JP, Johnson H, et al. Macroporous gelatinespheres as culture substrate, transplantation vehicle, andbiodegradable scaffold for guided regeneration of soft tissues.In vivo study in nude mice. J Plast Reconstr Aesthet Surg 2007;60:543e55.
20. Fernandes AM, Fernandes TG, Diogo MM, et al. Mouse embry-onic stem cell expansion in a microcarrier-based stirred culturesystem. J Biotechnol 2007;132:227e36.
21. Melero-Martin JM, Dowling MA, Smith M, et al. Expansion ofchondroprogenitor cells on macroporous microcarriers as analternative to conventional monolayer systems. Biomaterials2006;27:2970e9.
22. Malda J, Kreijveld E, Temenoff JS, et al. Expansion of humannasal chondrocytes on macroporous microcarriers enhancesredifferentiation. Biomaterials 2003;24:5153e61.
23. Ohlson S, Branscomb J, Nilsson K. Bead-to-bead transfer ofChinese hamster ovary cells using macroporous microcarriers.Cytotechnology 1994;14:67e80.
24. Liu JY, Hafner J, Dragieva G, et al. Autologous cultured kera-tinocytes on porcine gelatin microbeads effectively healchronic venous leg ulcers. Wound Repair Regen 2004;12:148e56.
25. Stover NP, Bakay RA, Subramanian T, et al. Intrastriatalimplantation of human retinal pigment epithelial cellsattached to microcarriers in advanced Parkinson disease. ArchNeurol 2005;62:1833e7.
26. Del Guerra S, Bracci C, Nilsson K, et al. Entrapment ofdispersed pancreatic islet cells in CultiSpher-S macroporousgelatin microcarriers: preparation, in vitro characterization,and microencapsulation. Biotechnol Bioeng 2001;75:741e4.
27. Kanno T, Takahashi T, Tsujisawa T, et al. Platelet-rich plasmaenhances human osteoblast-like cell proliferation and differ-entiation. J Oral Maxillofac Surg 2005;63:362e9.
28. Akeda K, An HS, Okuma M, et al. Platelet-rich plasma stimu-lates porcine articular chondrocyte proliferation and matrixbiosynthesis. Osteoarthr Cartil 2006;14:1272e80.
29. Datta N, Pham QP, Sharma U, et al. In vitro generated extra-cellular matrix and fluid shear stress synergistically enhance 3Dosteoblastic differentiation. Proc Natl Acad Sci U S A 2006;103:2488e93.
30. Holtorf HL, Jansen JA, Mikos AG. Flow perfusion cultureinduces the osteoblastic differentiation of marrow stroma cell-scaffold constructs in the absence of dexamethasone.J Biomed Mater Res A 2005;72:326e34.
31. Lev R, Spicer SS. Specific Staining of Sulphate Groups withAlcian Blue at Low Ph. J Histochem Cytochem 1964;12:309.
32. Goldstein DJ, Horobin RW. Surface staining of cartilage byAlcian blue, with reference to the role of microscopic dye
Figure 6 Graph visualising linear regression analysis ofmeasurements of spectrophotometric analysis of alkalinephosphatase activity in osteogenic induced fibroblasts andcontrol fibroblasts after 12 weeks. X-axis indicates time pointin minutes. Y-Axis indicates the difference in optical densitybetween 405 and 490 nm.
Engineering three-dimensional cartilage- and bone-like tissues 1045
aggregates in histological staining. Histochem J 1974;6:175e84.
33. Mason RM. Observations on the glycosaminoglycans of agingbronchial cartilage studied with Alcian Blue. Histochem J 1971;3:421e34.
34. Bills CE, Eisenberg H, Pallante SL. Complexes of organic acidswith calcium phosphate: the Von Kossa stain as a clue to thecomposition of bone mineral. Johns Hopkins Med J 1974;128:194e207.
35. Magnusson P, Farley JR. Differences in sialic acid residuesamong bone alkaline phosphatase isoforms: a physical,biochemical, and immunological characterization. CalcifTissue Int 2002;71:508e18.
36. Pazzano D, Mercier KA, Moran JM, et al. Comparison of chon-drogensis in static and perfused bioreactor culture. BiotechnolProg 2000;16:893e6.
37. Vunjak-Novakovic G, Obradovic B, Martin I, et al. Bioreactorstudies of native and tissue engineered cartilage. Biorheology2002;39:259e68.
38. Wendt D, Stroebel S, Jakob M, et al. Uniform tissues engineeredby seeding and culturing cells in 3D scaffolds under perfusion atdefined oxygen tensions. Biorheology 2006;43:481e8.
39. Mauck RL, Soltz MA, Wang CC, et al. Functional tissue engi-neering of articular cartilage through dynamic loading of chon-drocyte-seeded agarose gels. J Biomech Eng 2000;122:252e60.
40. Waldman SD, Spiteri CG, Grynpas MD, et al. Long-term inter-mittent compressive stimulation improves the composition andmechanical properties of tissue-engineered cartilage. TissueEng 2004;10:1323e31.
41. French MM, Rose S, Canseco J, et al. Chondrogenic differen-tiation of adult dermal fibroblasts. Ann Biomed Eng 2004;32:50e6.
42. Dragoo JL, Carlson G, McCormick F, et al. Healing full-thicknesscartilage defects using adipose-derived stem cells. Tissue Eng2007;13:1615e21.
43. Masuoka K, Asazuma T, Hattori H, et al. Tissue engineering ofarticular cartilage with autologous cultured adipose tissue-derived stromal cells using atelocollagen honeycomb-shapedscaffold with a membrane sealing in rabbits. J Biomed MaterRes B Appl Biomater 2006;79:25e34.
44. Mizuno S, Glowacki J. Three-dimensional composite of dem-ineralized bone powder and collagen for in vitro analysis ofchondroinduction of human dermal fibroblasts. Biomaterials1996;17:1819e25.
45. Mizuno S, Glowacki J. Chondroinduction of human dermalfibroblasts by demineralized bone in three-dimensionalculture. Exp Cell Res 1996;227:89e97.
46. Hattori H, Masuoka K, Sato M, et al. Bone formation usinghuman adipose tissue-derived stromal cells and a biodegrad-able scaffold. J Biomed Mater Res B Appl Biomater 2006;76:230e9.
47. Phillips JE, Guldberg RE, Garcia AJ. Dermal fibroblasts genet-ically modified to express Runx2/Cbfa1 as a mineralizing cellsource for bone tissue engineering. Tissue Eng 2007;13:2029e40.
48. Rutherford RB, Moalli M, Franceschi RT, et al. Bone morpho-genetic protein-transduced human fibroblasts convert toosteoblasts and form bone in vivo. Tissue Eng 2002;8:441e52.
1046 P. Sommar et al.
IIIPAPER
1
In Vivo Implantation of Osteogenic Induced
Human Dermal Fibroblasts in a Fracture Model
Pehr Sommar, MD1, Johan P E Junker, PhD1, 2, Eivind Strandenes, MD3, Charlotte Ness3,
Thomas Hansson, MD, PhD4, Hans Johnson, MD, PhD3, Gunnar Kratz, MD, PhD1
1Laboratory for Reconstructive Plastic Surgery, Department of Clinical and Experimental Medicine, Linköping University,
Linköping, Sweden 2Laboratory for Tissue Repair and Gene Therapy, Brigham and Women’s Hospital, Harvard Medical School, Boston, US 3Department of Surgery, Section of Plastic Surgery and Burn Centre, Haukeland University Hospital, Bergen, Norway 4Department of Plastic Surgery, Hand Surgery and Burns, University Hospital, Linköping, Sweden
Corresponding author
Pehr Sommar
Department of Reconstructive Plastic Surgery
Karolinska University Hospital Solna
171 76 Stockholm
Email: [email protected]
Phone: +46-739-811696
Fax: +46-8-51772505
Abstract
Fracture healing is a complex event involving cells and growth factors. When healing is impaired it substantially affects quality of life and increases medical costs. To overcome difficulties with impaired bone healing several methods using biomaterials have been tested. Osteogenic biomaterials, which are scaffolds loaded with osteocompetent cells, have been proposed when the defect is large. In this study we wanted to investigate the potential of osteogenic induced human dermal fibroblasts grown on gelatin microcarriers combined with platelet rich plasma (PRP) in a femoral gap surgical model in athymic rats. The gaps were transplanted with one of the following six combinations: 1; NaCl, 2; PRP, 3; microcarriers + PRP, 4; human dermal fibroblasts on microcarriers + PRP, 5; human osteoblasts on microcarriers + PRP, 6; osteogenic induced human dermal fibroblasts on microcarriers + PRP. The gaps were analysed 4 weeks postoperatively with computer tomography, routine histological staining, fluorescence in situ hybridization (FISH) and polyclonal antibodies directed towards osteocalcin and osteonectin. Radiographs taken 4 weeks post surgery did not reveal callus in any of the groups. Gaps transplanted with osteogenic induced human dermal fibroblasts on microcarriers (group 6) contained dense cell clusters with large amounts of extracellular matrix. These cell clusters were not found in the other groups and stained highly positive for osteocalcin and osteonectin. FISH analysis revealed viable human cells in gaps filled with cell-seeded microcarriers confirming survival of transplanted cells. In conclusion osteogenic induced human dermal fibroblasts survive in this new niche and display bonelike structures in the gaps.
2
Introduction Bone tissue has the ability to regenerate
following a fracture. Cells participating in this process include platelets, inflammatory cells, fibroblasts, chondrocytes, endothelial cells, osteoblasts, osteoclasts and the recruitment and transformation of mesenchymal stem cells (Schindeler et al., 2008). Several peptides and growth factors controlling cellular events in bone formation and repair have been identified; the most important being the bone morphogenetic protein (BMP) family and transforming growth factor β (TGF-β) (Axelrad et al., 2007; Deschaseaux et al., 2009). Human fracture repair is typically a robust process and treatment often relies solely on stabilization and pain management. However, with fractures that are difficult to heal, challenges often arise. Disruption of the healing process has major impact on both the associated medical costs and the quality of life for the patient. Approximately 10 % of all fractures are complicated by impaired healing (Einhorn, 1995).
To overcome difficulties with impaired bone healing, several methods have been tested. The most common is plate- or intramedullary fixation and grafting at the fracture site with iliac crest bone. The method of auto-grafting from iliac crest has limitations in the amount of tissue available and also with the risk of donor site morbidity (Sasso et al., 2005). To overcome the donor site morbidity, the use of biomaterials instead of crista bone has been suggested. The biomaterials can be classified as osteoconductive, osteoinductive or osteogenic (Hannouche et al., 2001). Some of the common osteoconductive biomaterials are calcium-based ceramics including hydroxyapatite and tricalcium phosphate. These materials are conductive in the sense that they rely on bone ingrowth that is often confined to the surfaces, and therefore they are not suitable for larger defects. Osteoinductive biomaterials combine scaffolds and growth factors. Bone tissue releases several growth factors at the site of fracture, including BMPs, TGF-β, platelet derived growth factor, insulin-like growth factor 1 and 2 and fibroblast growth factors (Tsiridis et al., 2007). BMPs are the most used osteoinductive factors. There are at least 18 BMPs of which 15 are known to be expressed in humans (Axelrad et al., 2007). BMP 2 and 7 have been used in clinical trials in combination with different scaffolds, e.g. collagen, polylactic acid (PLA), polyglycolic acid (PGA) or bone autograft (Hannouche et al., 2001; Govender et al., 2002; Dimitriou et al., 2005; Axelrad et al., 2007).
The efficiency of osteoinductive biomaterials relies on the recruitment of osteocompetent cells from the surrounding tissues. Their use is therefore limited to cases in which the fracture site can provide these cells. This precludes their use in necrotic areas or in areas with large defects of bone. To overcome this problem the use of osteogenic biomaterials, scaffolds loaded with autologous osteocompetent cells, has been proposed.
An osteogenic biomaterial should promote attachment, migration and proliferation of seeded cells and thereafter facilitate perfusion of nutrients and oxygen. The cells should be able to replace and degrade the biomaterial during bone formation. Degradation of the biomaterial is also necessary for accurate x-ray pictures, which for instance is a problem with calcium based biomaterials.
A promising osteogenic biomaterial is the gelatin macroporous microcarrier. Gelatin microcarriers offer substantial surface areas for cell adhesion and proliferation due to the pourous structure, and are easily degraded by the growing tissues. They support in vitro growth of various mammalian cells (Del Guerra et al., 2001; Malda et al., 2003; Liu et al., 2004; Liu et al., 2006). Microcarriers have previously been used in bone tissue engineering applications (MaldaFrondoza, 2006; Tielens et al., 2007) and the fluid induced shear stress from spinner flask culture of microcarriers has been shown to have a positive impact on osteogenic differentiation (Datta et al., 2006; Sommar et al., 2009). The gelatin microcarriers are easily handled in suspension and therefore offer a simple way to fill a defect by injecting cell-seeded microcarriers.
Platelet rich plasma (PRP) has been proposed as a blood-derived biological glue to encapsulate cell-seeded microcarriers in a three-dimensional in vitro model (Pettersson et al., 2009). PRP preserved the three-dimensional structure of the tissue construct until sufficient extracellular matrix was produced to uphold the structural integrity. The autologous source of plasma makes it a strong candidate for clinical purposes. Beneficial effect from the platelets has been reported, as osteoblast-like cells cultured in the presence of PRP showed enhanced cell proliferation and extracellular matrix production (Slapnicka et al., 2008), whereas some reports show no positive or only little effect of PRP on bone regeneration in vivo (Thorwarth et al., 2006).
The most commonly used cell source for seeding onto osteogenic biomaterials is mesenchymal stem cells (MSCs) from bone marrow
3
(Bruder et al., 1998; Pittenger et al., 1999; Granero-Molto et al., 2009). MSC seeded osteogenic implants facilitate bone regeneration as compared to unseeded scaffolds (Bruder et al., 1998; Dallari et al., 2006). However, the optimal cell type for use in tissue engineering applications should preferably be easy to obtain, cultivate and store. Therefore, bone marrow cells are suboptimal candidates since they are associated with difficulties in harvest, culture and expansion (Bruder et al., 1997). The search for alternative cell sources has included investigations of the multilineage potential of normal adult human cells. Studies have suggested that cells with similar plasticity as MSCs can be isolated from a variety of adult tissues, including adipose tissue, skeletal muscle and various dermal tissue components. These cells are able to differentiate and express tissue markers distinct from their own origin, in response to changes of the microenvironment (Young et al., 1995; Wada et al., 2002; Zuk et al., 2002; Jahoda et al., 2003).
The human dermal fibroblast is an easy accessible cell source. Substantial cell yields can be obtained from a relatively small skin biopsy using minimally invasive procedures and routine cell expansion techniques. We have previously shown that human dermal fibroblasts may be induced to differentiate towards several mesenchymal lineages when subjected to osteogenic, chondrogenic and adipogenic induction media (Junker et al., 2010) as well as endothelial cell induction (Karlsson et al., 2009). This inherent plasticity is supported by results from other groups (Bartsch et al., 2005; Lorenz et al., 2008).
We have furthermore combined human dermal fibroblasts, gelatin macroporous microcarriers and PRP to form a three-dimensional osteogenic biomaterial, which upon subjection to osteogenic induction medium produced mineralized extracellular matrix in vitro (Sommar et al., 2009).
However, before this concept can be used clinically, the cells ability to survive and keep their osteogenic phenotype in an in vivo fracture environment must be elucidated. In the present study the fate of osteogenic induced human dermal fibroblasts on gelatin microcarriers in combination with PRP was investigated in a fracture model in nude rats. The primary aim was to investigate if osteogenic induced fibroblasts transplanted to an in vivo environment could retain their phenotype and growth pattern despite omission of induction medium.
Materials and Methods
Isolation of cells Fibroblasts were obtained from dermal tissue
taken from a healthy female patient undergoing routine plastic surgery. The tissue was stored in 0.9 % NaCl soaked gauzes and processed within 24 hrs. Primary cultures of human dermal fibroblasts were isolated by dissecting the dermal layer of the skin samples and incubating the fragments at 37° C over night in Dulbecco´s Modified Eagle`s Medium (DMEM) (Gibco, Invitrogen corporation, NY) containing 2% Fetal Calf Serum (FCS) (Gibco, Invitrogen corporation, NY), 165 U/ml Collagenase
Table I. Induction and proliferation media
Medium Media Serum Supplement
Fibroblast proliferation (FM) DMEM 10 % FCS 50 U/ml Penicillin
50 g/ml Streptomycin
Osteoblast proliferation (OPM) α-MEM 10 % FCS 50 U/ml Penicillin
50 g/ml Streptomycin
Osteogenic (OM) DMEM 10 % FCS 50 M Ascorbate-2-phosphate
0.1 M Dexamethasone
10 mM Glycerophosphate
50 U/ml Penicillin
50 g/ml Streptomycin
4
(type 1) (Gibco, Invitrogen corporation, NY) and 2.5 U/ml Dispase (Gibco, Invitrogen corporation, Germany) with repeated agitation to dissolve tissue fragments. The resulting cell suspension was centrifuged at 200 g for 10 min, the cell pellet resuspended in fibroblast proliferation medium (FM; table I) and plated in 75 cm² culture flasks at 37° C, 5 % CO2 and 95 % humidity. Medium was changed three times per week.
Osteoblasts were obtained from bone tissue taken from hip arthroplasty of a female patient. Bone tissue was stored in 0.9 % NaCl and processed within 24 hrs. Blood clots were removed by agitation in phosphate buffered saline (PBS). Bone tissue was minced and treated with 1 mg/ml Collagenase (type 2) (Gibco, Invitrogen corporation, NY) in α-MEM (Gibco, Invitrogen corporation, NY) at 37° C for 15 min with intermittent agitation. This procedure was repeated three times. The bone fragments were transferred to a 75 cm2 culture flask and cultured in osteoblast proliferation medium (OPM; table I). Medium was changed three times per week.
Microcarriers CultiSpher-GL microcarriers were obtained from
Percell Biolytica AB (Sweden). Microcarriers were prepared according to the manufacturer´s recommendations. Briefly, dry microcarriers were rehydrated in PBS for a minimum of 1 h at room temperature. Without removing the PBS, microcarriers were sterilized by autoclaving (121° C, 20 min, 15 psi). Sterilized microcarriers were washed once with PBS, twice with FM, and stored at 4°C until use.
Seeding of Cells on Microcarriers Human dermal fibroblasts at passage 4 (four
confluent 75 cm2 culture flasks) were washed with 0.02 % ethylenediaminetetraacetic acid (EDTA) and enzymatically detached by incubation with EDTA (0.01 %)/Trypsin (0.125 %) solution at 37° C for approximately 15 min. The suspension was centrifuged at 200 g for 5 min, the supernatant removed, cells resuspended in 200 ml FM with 0.4 g microcarriers (dry weight) and transferred to two spinner flasks at 100 ml per flask (MCS 104S, 250 ml, Techne, UK). Cells were allowed to attach to microcarriers under intermittent agitation (20 RPM for 5 min every h) at 37° C, 5 % CO2 and 95 % humidity. On the second day, culture medium volume was adjusted to 200 ml per spinner flask,
and agitation adjusted to continuous stirring (20 RPM). 100 ml FM was changed three times per week.
The same procedure was repeated for the seeding of osteoblasts on microcarriers (two confluent 75 cm2 culture flasks for 0.2 g microcarriers). The osteoblasts were kept in spinner flasks in OPM.
As a control group, acellular microcarriers were prepared in the same manner and kept in FM in spinner flasks.
Induction of osteogenic differentiation Before osteogenic induction was started, the
viability and distribution of fibroblasts on microcarriers was analysed using 3-(4.5-dimethylthiazol-2-yl)-2.5-diphenyltetrazolium bromide (MTT) staining (Sigma Aldrich Co, MO). 400 µl of the microcarrier suspension was taken from spinner flasks. 40 µl MTT (5 mg/ml in PBS) was added to each sample and incubated at 37° C for 45 min. Staining was evaluated using an Olympus IX51 inverted light microscope (x40/0.75) (Olympus, Bromma, Sweden).
MTT stain revealed optimal cell density after approximately 14 days, and thereafter induction was started using osteogenic induction medium (OM; table I) in one of the two spinner flasks with fibroblast seeded microcarriers, the other flask continued in FM. Induction of fibroblasts with OM was maintained for 8 weeks.
At the end of the in vitro culture period, cells were counted using a Guava PCA cell counter (Guava Technologies, Inc. CA). A total of five 0.5 ml samples were removed from the microcarrier suspensions. 0.3 ml of the supernatant was extracted and replaced with EDTA (0.01 %)/Trypsin (0.125 %) for 15-25 min until microcarriers were completely dissolved. Samples were diluted 1:10 in Guava ViaCount Reagent and measured twice.
Femoral gap surgical model All procedures involving animals were
performed in accordance with a protocol approved by the Norwegian Institutional Animal Care and Use Committee (Project 2008180). The animals were kept in standard cages (4 animals per cage) under controlled conditions of temperature (20 ± 1° C), humidity (55 ± 5 %) and a standard dark-
5
Table II. Filling of femoral defects
Group Number of
animals
Filling Cells/Defect
1 4 400 µl NaCl 0.9 % 0
2 4 400 µl PRP 0
3 4 200 µl microcarriers (MC)
200 ml PRP
0
4 4 200 µl human dermal fibroblasts on microcarriers (FB)
200 µl PRP
6.0 x 105
5 4 200 µl human osteoblasts on microcarriers (OB)
200 µl PRP
1.4 x 104
6 4 200 µl osteogenic induced human dermal fibroblasts on
microcarriers (O-FB)
200 µl PRP
6.0 x 105
Total 24
light cycle (12/12 hrs). The air in the animal room was recirculated once every 22 hrs. All rats had access to food and water (RM3 pellets sterilized at high temperature; Beekay Universal Ltd., UK) ad libitum.
The femoral gap surgical model has been used extensively in athymic rats to study long-bone repair (Einhorn et al., 1984; Werntz et al., 1996; Bruder et al., 1998). Briefly, 24 female NIH nude rats (Taconic, NY) (mean weight 248.2 g), were randomly divided into 6 groups, receiving different filling of the femoral defects: 1; NaCl, 2; PRP, 3; microcarriers + PRP (MC), 4; human dermal fibroblasts on microcarriers + PRP (FB), 5; human osteoblasts on microcarriers + PRP (OB), 6; osteogenic induced human dermal fibroblasts on microcarriers + PRP (O-FB) (table II). There was no statistically significant difference in mean weight between groups. In isoflurane anesthesia both femurs were exposed by a lateral approach. A custom made ultra high molecular weight polyethylene fixation plate (20x4x3mm) was attached anteriorly on the mid-femoral diaphysis and fixed with four 0.9 mm Kirschner wires (Zimmer, IN) and two 0.6 mm cerklage wires. A 5 mm osteoperiostal mid-diaphyseal segment was removed using a high speed osteotomy burr (Zimmer, IN) and the defect was rinsed with 0.9 % NaCl.
Microcarriers were allowed to sediment for 30 min and medium was removed. Microcarriers were centrifuged at 100 g for 5 min and supernatant was removed.
PRP was prepared from citrated whole blood from a male healthy donor. Briefly, whole blood was collected into Vacutainer collection tubes (9NC) (BD Diagnostics, Plymouth, UK) containing citrate. PRP was prepared by centrifugation of the citrated whole blood for 20 min at 220 g. Microcarriers and PRP was mixed 1:1 (vol:vol) and 400 µl was delivered to each femoral defect. In the first group 400 µl NaCl was put in each defect and 400 µl PRP in the second group.
All defects were sealed by wrapping them in 15x30 mm Karoderm human cell free dermal tissue (Karocell, Sweden), which was closed by suturing with Vicryl 5/0 (Johnson & Johnson, Belgium). Muscle fascia and skin was sutured with Vicryl 5/0. Weight bearing was started immediately post-operative and analgesia was provided.
At 4 weeks rats were killed by CO2 inhalation. Immediately afterwards, both femurs were collected and fixed in 4 % neutral buffered paraformaldehyde (Apoteket, Sweden) for 48 hrs.
Evaluation Bones were examined using dual source
computer tomography (CT) (Somatom Definition, Siemens, Germany) (temporal resolution 75 ms). Thereafter, bones where demineralized by treatment with 10 % EDTA solution in distilled water (dH2O) pH 7.4 for a period of 3 weeks, with renewal of EDTA every week. After decalcification, samples were dehydrated, embedded in paraffin, sectioned 7 μm and mounted onto microscopy slides.
6
Routine histology was assessed by hematoxylin and eosin stain (HTX). In brief, slides were rehydrated, rinsed in dH2O and treated with Mayer´s HTX (Histolab, Sweden) for 6 min, rinsed in water and treated in 0.2 % Eosin (Histolab, Sweden) for 2 min. Slides were rinsed in dH2O, dehydrated and mounted.
The expression of osteocalcin and osteonectin was detected by using immunohistochemisty (IHC). Briefly, slides were rehydrated, rinsed in PBS, and treated with pepsin (1 mg/ml) (Sigma-Aldrich Co., MO) in Tris-HCl (0.2 M) for 15 min. Slides were rinsed in PBS and incubated in blocking serum (5 % bovine serum albumin in PBS) for 45 min followed by incubation for 1 h with polyclonal antibodies directed towards osteocalcin (OC) (dilution 1:500; Chemicon Inc., CA) or osteonectin (ON) (dilution 1:1000; Chemicon Inc., CA) respectively. Sections were rinsed in PBS and incubated for 1 h with ALEXA 488-conjugated secondary antibodies (Molecular Probes, OR). Finally sections were rinsed in PBS and mounted using ProLong Gold mounting medium (Molecular Probes, OR) containing 4', 6-diamidino-2-phenylindole (DAPI). Negative control included omission of primary antibodies.
To detect human cells present in the femoral defects, fluorescence in situ hybridization (FISH) was performed using the Vysis Inc. Spectrum green female total human genomic DNA probe and kit (Abbott Scandinavia AB, Sweden). Briefly, tissue sections on microscopy slides were left over night on a hot plate (58° C), deparaffinized in xylene and rehydrated in ethanol series. The sections were incubated in 0.2 M HCl for 20 min, rinsed in dH2O and in Washing Buffer. Slides were incubated in Pre-treatment solution for 30 min at 82° C, rinsed in dH2O and subsequently in Washing Buffer. Sections were incubated in Protease solution (pH 2.0) for 15 min in 37° C, rinsed in Washing Buffer and dehydrated in ethanol series. A mixture of probe (1.5 µl), Hybridization Buffer (7 µl) and dH2O (1.5 µl) was applied to each section and cover-slips were mounted using DPX mounting agent (KEBO Lab, Sweden). The hybridization process was carried out for 16 hrs in a Vysis HYBrite (Abbott Scandinavia AB, Sweden). The cover-slips were removed and sections transferred to a washing solution containing 2xSSc and 0.3 % NP40 (Abbot Scandinavia AB, Sweden) for two minutes at 82° C. Sections were mounted using ProLong Gold mounting medium (Molecular Probes, OR) containing DAPI.
All experiments were performed at room temperature unless otherwise stated. Samples were visualized using an Olympus BX41 microscope (x20/0.50; x40/0.75; x100/1.30 oil) and images captured using an Olympus DP70 CCD camera (Olympus, Sweden)
Results
In vitro cultivation and differentiation Normal human dermal fibroblasts from routine
plastic surgery were expanded in vitro. Observation of cells revealed typical fibroblast-like morphological characteristics. Osteogenic induction medium stimulated extensive proliferation of the human dermal fibroblasts and production of extracellular matrix in the spinner flasks, which increased over time forming macroscopic aggregates in the culture. No aggregates were seen in control fibroblasts or osteoblast cultures.
In vivo transplantation of differentiated cells One rat in group 5 (OB) died due to
complications related to surgery. 4 weeks after implantation in the femoral defect model, femurs of the remaining 23 animals were collected and analysed. There was compromised fixation of the defect in one out of eight femurs in group 2 (PRP), group 3 (MC), and group 4 (FB) respectively. These were excluded from further analysis.
Analysis of femoral defects Radiographs taken 4 weeks post surgery did not
reveal detectable callus formation in any of the experimental groups. All groups had the same radiological appearance showing unhealed defects. When measuring the remaining distance of the defects, a small reduction was seen in all groups, but the reduction of defects could reflect mechanical compression rather than healing why no further conclusion was drawn from this.
In the microcarrier groups (groups 3-6) there was a clearly visible ingrowth of capillaries in the defects (fig. 1). The microcarriers were partially degraded after 4 weeks.
7
Figure 1. HTX stain. Panels displaying HTX-stained sections 4 weeks after transplantation of gaps with NaCl (A), PRP (B), microcarriers + PRP (C), human dermal fibroblasts on microcarriers + PRP (D), human osteoblasts on microcarriers + PRP (E), osteogenic induced human dermal fibroblasts on microcarriers + PRP (F). Note the dense cell clusters in the osteogenic induced human fibroblasts (F), marked with asterisks. Bar = 1 mm. Insets in C-F show capillaries in the central region of the gaps. Bar = 40 µm.
Figure 2. Osteocalcin. Panels showing sections stained with antibodies towards osteocalcin 4 weeks after transplantation of gaps with NaCl (A), PRP (B), microcarriers + PRP (C), human dermal fibroblasts on microcarriers + PRP (D), human osteoblasts on microcarriers + PRP (E), osteogenic induced human dermal fibroblasts on microcarriers + PRP (F). Negative control with the omission of primary antibodies (G). DAPI-stained nuclei are indicated by arrowheads. Bar = 50 µm. Dense cell clusters were found in gaps transplanted with osteogenic induced human dermal fibroblasts (F). These clusters stained highly positive for osteocalcin (between arrows). Clusters were not present in the gaps of the other groups. There was some immunoreactivity also in the extracellular matrix of human dermal fibroblasts (D) and human osteoblasts (E) (arrows). It should be noted that this stain was discrete compared to the highly positive IHC-stain observed in the group seeded with osteogenic induced human dermal fibroblasts (F).
8
Figure 3. Osteonectin Panels showing sections stained with antibodies towards osteonectin 4 weeks after transplantation of gaps with NaCl (A), PRP (B), microcarriers + PRP (C), human dermal fibroblasts on microcarriers + PRP (D), human osteoblasts on microcarriers + PRP (E), osteogenic induced human dermal fibroblasts on microcarriers + PRP (F). Negative control with the omission of primary antibodies (G). DAPI-stained nuclei are indicated by arrowheads. Bar = 50 µm. Dense cell clusters were found in the gaps transplanted with osteogenic induced human dermal fibroblasts (F). These clusters stained positive for osteonectin interspersed in the extracellular matrix (arrows). Clusters were not present in the gaps of the other groups. There was minor immunoreactivity in the extracellular matrix of human dermal fibroblasts (D) and human osteoblasts (E) (arrows). It should be noted that this stain was very discrete compared to the IHC-stain observed in the group seeded with osteogenic induced human dermal fibroblasts (F).
In defects transplanted with O-FB seeded
microcarriers (group 6), dense cell clusters with large amounts of extracellular matrix partially filled seven out of eight gaps at 4 weeks (fig. 1). These clusters showed a highly positive staining when applying polyclonal antibodies directed towards OC (fig. 2) and ON (fig. 3). Cell clusters were not present in the gaps of the other groups.
In defects transplanted with OB seeded microcarriers (group 5), FB seeded microcarriers (group 4) or microcarriers only (group 3), HTX stain showed a randomly dispersed cell/matrix tissue with partially degraded microcarriers (fig. 1). No immunohistochemical staining was observed in group 3, while group 4 and 5 stained moderately positive for OC (fig. 2) and ON (fig. 3). It should be noted though that this OC and ON stain was very discrete compared to the highly positive IHC-stain observed in the group seeded with O-FB (group 6). HTX stain revealed fibrous tissue in gaps filled with PRP (group 2) and NaCl (group 1) (fig. 1). These groups displayed negative IHC stain regarding OC (fig. 2) and ON (fig. 3).
FISH analysis displayed viable human cells in defects filled with cell-seeded microcarriers (groups
4 – 6) confirming survival of transplanted cells in the gaps (fig. 4).
In all groups, cells at the fracture margin stained positive for OC and ON, which was interpreted as out-growth of cells from the rat bone marrow.
Discussion The present study focuses on the survival of
osteogenic induced human dermal fibroblasts in vivo, using a femoral gap surgical model in athymic rats. We have previously shown that normal human dermal fibroblasts can be induced to differentiate towards an osteoblast-like cell type upon subjection to specific induction medium (Junker et al., 2010). Furthermore, these cells can give rise to three-dimensional tissue resembling bone in vitro, a process which is further enhanced by the addition of fluid induced shear stress (Sommar et al., 2009). The aim of this study was to investigate if osteogenic induced fibroblasts transplanted to an in vivo environment could retain their phenotype and growth pattern despite omission of induction medium.
9
Figure 4. FISH. Panels (A-C) showing fluorescence in situ hybridization (FISH) in sections 4 weeks after transplantation using Spectrum green female total human genomic DNA probe. Panels (D-F) showing corresponding gaps stained with DAPI. (A) and (D) gaps filled with human dermal fibroblasts on microcarriers + PRP. (B) and (E) gaps filled with human osteoblasts on microcarriers + PRP. (C) and (F) gaps filled with osteogenic induced human dermal fibroblasts on microcarriers + PRP. Bar = 50 µm. Arrowheads indicate human cell nuclei.
The defects transplanted with O-FB seeded
microcarriers (group 6) displayed cell clusters present in seven out of eight gaps. These cell clusters contained large amounts of extracellular matrix and were labeled positively using antibodies towards OC and ON. Cell clusters were not found in the other groups, which clearly state the unique properties of the osteogenic induced fibroblasts. The cell clusters could arguably be pre-fabricated in the spinner flask, and transplanted as ready-made clusters into the defects. Clinically this of minor importance; the important finding is that viable human cells were detected in the defects 4 weeks post surgery (FISH), and that these cells display osteogenic properties. This suggests a favorable milieu in the defect, which has maintained the osteogenic induced fibroblasts in an osteogenic differentiated state over time despite omission of osteogenic induction medium after transplantation into the defects.
CT investigation could not detect any mineralized callus in the gaps at the 4-week postsurgery radiographs, even not in defects receiving microcarriers seeded with O-FB (group 6) despite the histological occurrence of cell clusters. A possible explanation for this is the relatively short time allowed for healing in the current experimental setup. A recent study performed by Peters et al employing rat mesenchymal stem cells in a similar femoral gap surgical model showed periostal bridging in only a few gaps after 8 weeks (Peters et al., 2009). However, the aim of the present study was to investigate survival and function of the transplanted cells. To further elucidate their effect on the healing process studies designed with this in focus have to be performed.
Microcarrier-seeded groups demonstrated in-growth of capillaries in the defects, indicating that the gelatin microcarriers are suitable for use as a biodegradable scaffold, allowing neoangiogenesis to take place which is necessary for tissue survival. The
10
cell-seeded microcarriers were partially degraded and not visible in the x-ray investigation, thus displaying excellent features for an osteoinductive biomaterial.
To our knowledge, this study is the first in vivo study applying human dermal fibroblasts in a fracture model. Other studies have described subcutaneous/intramuscular injection of osteogenic induced dermal fibroblasts. Rutherford et al. described the use of human and rat fibroblasts transfected with BMP-7. The BMP-transfected human fibroblasts created ossicles when injected subcutaneously in mice. The rat fibroblasts, but not the human fibroblasts, were transplanted into a femoral gap surgical model with a defect of 2-3 mm with signs of cartilage/bone healing of all defects after 3 weeks (Rutherford et al., 2002). Hirata et al. described mice dermal fibroblasts transfected with BMP-2 in a calvarial defect model. 93 % of the animals had partial healing of the defects at 4 weeks, and all animals at 6 weeks (Hirata et al., 2007). These results were supported by Wang et al. in a study were human dermal fibroblasts transfected with BMP-2 created calcified osseous noduli upon intramuscular injection in rats. This study also used rat fibroblasts in a cranial defect model with about 50 % healing of defects after 10 weeks (Wang et al., 2009).
Neither Rutherford, nor Wang, has used human dermal fibroblasts in fracture models. They have injected BMP-transfected human fibroblasts subcutaneously or intramuscular but only used rat fibroblasts in fracture models. We can only speculate why human fibroblasts have not been employed in fracture models before. One likely explanation is that healing using human cells in fracture models may be a more complex process, yielding less positive results.
The use of genetic modification is a powerful tool in osteogenesis but the safety of using genetically modified cells in clinical applications is a matter of extensive debate. Retroviruses tend to insert into active genes, and it has been suggested that their use may increase the risk of cancer. Adenoviruses may elicit a strong immune response directed against the engineered cells (Young et al., 2006; Conrad et al., 2007). We have therefore chosen not to work with transfection of cells but with osteogenic differentiation by means of medium induction via soluble factors. We find that this serves as a viable alternative to genetic alteration of transplanted cells.
The addition of a cell component in order to enhance fracture healing in large defects of bone is an appealing option. Results presented here indicate that osteogenic induced human dermal fibroblasts seeded on a scaffold are interesting candidates for cell delivery in fractures, even though CT investigation at 4 weeks showed non-union of defects. The demonstrated survival and maintained osteogenic properties of cells 4 weeks after implantation confirms the phenotypical stability after in vitro induction of the cells and indicates their possible value in new regimes for the treatment of fractures.
Acknowledgements We thank A. Lönn and K. Briheim for their
expert technical assistance, and Anders Hansson, Assoc. Professor, MD, PhD, Center for Medical Image Science and Visualization (CMIV), university hospital, Linköping, Sweden for providing CT-scan
investigations.
Competing interests statement The authors declare that they have no
competing financial interests.
References
Axelrad TW, Kakar S, et al. (2007). New technologies for the enhancement of skeletal repair. Injury 38 Suppl 1: S49-62.
Bartsch G, Yoo JJ, et al. (2005). Propagation, expansion, and multilineage differentiation of human somatic stem cells from dermal progenitors. Stem Cells Dev 14(3): 337-48.
Bruder SP, Jaiswal N, et al. (1997). Growth kinetics, self-renewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. J Cell Biochem 64(2): 278-94.
Bruder SP, Kurth AA, et al. (1998). Bone regeneration by implantation of purified, culture-expanded human mesenchymal stem cells. J Orthop Res 16(2): 155-62.
Conrad C, Gupta R, et al. (2007). Genetically engineered stem cells for therapeutic gene delivery. Curr Gene Ther 7(4): 249-60.
Dallari D, Fini M, et al. (2006). In vivo study on the healing of bone defects treated with bone marrow stromal cells, platelet-rich plasma, and freeze-dried bone allografts, alone and in combination. J Orthop Res 24(5): 877-88.
Datta N, Pham QP, et al. (2006). In vitro generated extracellular matrix and fluid shear stress synergistically enhance 3D osteoblastic differentiation. Proc Natl Acad Sci U S A 103(8): 2488-93.
11
Del Guerra S, Bracci C, et al. (2001). Entrapment of dispersed pancreatic islet cells in CultiSpher-S macroporous gelatin microcarriers: Preparation, in vitro characterization, and microencapsulation. Biotechnol Bioeng 75(6): 741-4.
Deschaseaux F, Sensebe L, et al. (2009). Mechanisms of bone repair and regeneration. Trends Mol Med 15(9): 417-29.
Dimitriou R, Dahabreh Z, et al. (2005). Application of recombinant BMP-7 on persistent upper and lower limb non-unions. Injury 36 Suppl 4: S51-9.
Einhorn TA (1995). Enhancement of fracture-healing. J Bone Joint Surg Am 77(6): 940-56.
Einhorn TA, Lane JM, et al. (1984). The healing of segmental bone defects induced by demineralized bone matrix. A radiographic and biomechanical study. J Bone Joint Surg Am 66(2): 274-9.
Govender S, Csimma C, et al. (2002). Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Joint Surg Am 84-A(12): 2123-34.
Granero-Molto F, Weis JA, et al. (2009). Regenerative effects of transplanted mesenchymal stem cells in fracture healing. Stem Cells 27(8): 1887-98.
Hannouche D, Petite H, et al. (2001). Current trends in the enhancement of fracture healing. J Bone Joint Surg Br 83(2): 157-64.
Hirata K, Mizuno A, et al. (2007). Transplantation of skin fibroblasts expressing BMP-2 contributes to the healing of critical-sized bone defects. J Bone Miner Metab 25(1): 6-11.
Jahoda CA, Whitehouse J, et al. (2003). Hair follicle dermal cells differentiate into adipogenic and osteogenic lineages. Exp Dermatol 12(6): 849-59.
Junker JP, Sommar P, et al. (2010). Adipogenic, chondrogenic and osteogenic differentiation of clonally derived human dermal fibroblasts. Cells Tissues Organs 191(2): 105-18.
Karlsson LK, Junker JP, et al. (2009). Human dermal fibroblasts: a potential cell source for endothelialization of vascular grafts. Ann Vasc Surg 23(5): 663-74.
Liu JY, Hafner J, et al. (2006). High yields of autologous living dermal equivalents using porcine gelatin microbeads as microcarriers for autologous fibroblasts. Cell Transplant 15(5): 445-51.
Liu JY, Hafner J, et al. (2004). Autologous cultured keratinocytes on porcine gelatin microbeads effectively heal chronic venous leg ulcers. Wound Repair Regen 12(2): 148-56.
Lorenz K, Sicker M, et al. (2008). Multilineage differentiation potential of human dermal skin-derived fibroblasts. Exp Dermatol 17(11): 925-32.
Malda J and Frondoza CG (2006). Microcarriers in the engineering of cartilage and bone. Trends Biotechnol 24(7): 299-304.
Malda J, Kreijveld E, et al. (2003). Expansion of human nasal chondrocytes on macroporous microcarriers enhances redifferentiation. Biomaterials 24(28): 5153-61.
Peters A, Toben D, et al. (2009). Locally applied osteogenic predifferentiated progenitor cells are more effective than undifferentiated mesenchymal stem cells in the treatment of delayed bone healing. Tissue Eng Part A 15(10): 2947-54.
Pettersson S, Wettero J, et al. (2009). Human articular chondrocytes on macroporous gelatin microcarriers form structurally stable constructs with blood-derived biological glues in vitro. J Tissue Eng Regen Med 3(6): 450-60.
Pittenger MF, Mackay AM, et al. (1999). Multilineage potential of adult human mesenchymal stem cells. Science 284(5411): 143-7.
Rutherford RB, Moalli M, et al. (2002). Bone morphogenetic protein-transduced human fibroblasts convert to osteoblasts and form bone in vivo. Tissue Eng 8(3): 441-52.
Sasso RC, LeHuec JC, et al. (2005). Iliac crest bone graft donor site pain after anterior lumbar interbody fusion: a prospective patient satisfaction outcome assessment. J Spinal Disord Tech 18 Suppl: S77-81.
Schindeler A, McDonald MM, et al. (2008). Bone remodeling during fracture repair: The cellular picture. Semin Cell Dev Biol 19(5): 459-66.
Slapnicka J, Fassmann A, et al. (2008). Effects of activated and nonactivated platelet-rich plasma on proliferation of human osteoblasts in vitro. J Oral Maxillofac Surg 66(2): 297-301.
Sommar P, Pettersson S, et al. (2009). Engineering three-dimensional cartilage- and bone-like tissues using human dermal fibroblasts and macroporous gelatine microcarriers. J Plast Reconstr Aesthet Surg.
Thorwarth M, Wehrhan F, et al. (2006). PRP modulates expression of bone matrix proteins in vivo without long-term effects on bone formation. Bone 38(1): 30-40.
Tielens S, Declercq H, et al. (2007). Gelatin-based microcarriers as embryonic stem cell delivery system in bone tissue engineering: an in-vitro study. Biomacromolecules 8(3): 825-32.
Tsiridis E, Upadhyay N, et al. (2007). Molecular aspects of fracture healing: which are the important molecules? Injury 38 Suppl 1: S11-25.
Wada MR, Inagawa-Ogashiwa M, et al. (2002). Generation of different fates from multipotent muscle stem cells. Development 129(12): 2987-95.
Wang R, Zou Y, et al. (2009). Autografts and xenografts of skin fibroblasts delivering BMP-2 effectively promote orthotopic and ectopic osteogenesis. Anat Rec (Hoboken) 292(6): 777-86.
Werntz JR, Lane JM, et al. (1996). Qualitative and quantitative analysis of orthotopic bone regeneration by marrow. J Orthop Res 14(1): 85-93.
Young HE, Mancini ML, et al. (1995). Mesenchymal stem cells reside within the connective tissues of many organs. Dev Dyn 202(2): 137-44.
Young LS, Searle PF, et al. (2006). Viral gene therapy strategies: from basic science to clinical application. J Pathol 208(2): 299-318.
Zuk PA, Zhu M, et al. (2002). Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 13(12): 4279-95.
IVPAPER
1
Evaluating Multi-Lineage Induction of Human Dermal Fibroblasts
Using Gene Expression Analysis
Jonathan Rakar, MSc1, Pehr Sommar, MD1, Susanna Lönnqvist, MSc1, Hans Johnson, MD, PhD2, Johan P E Junker, PhD1, Gunnar Kratz, MD, PhD1
1Laboratory for Reconstructive Plastic Surgery, Division of Hand and Plastic Surgery, Department of Clinical and Experimental
Medicine, University Hospital, Linköping, Sweden 2Department of Surgery, Section of Plastic Surgery and Burn Centre, Haukeland University Hospital, Bergen, Norway
Corresponding author
Jonathan Rakar
Centre for Clinical and Experimental Research
Department of Clinical and Experimental Medicine
Faculty of Health Sciences, Linköping University
SE-58185 Linköping, Sweden
Tel: +46 10 1034485
Fax: +46 10 1037465
Abstract
During the past decades, several adult stem cell populations from a range of tissues have been characterized. Since principally all human cells contain the same genetic material, the specific gene expression profile determines the cell phenotype. The notion of terminally differentiated somatic cells being necessarily restricted to one phenotype has been challenged, and instead an inherent range of plasticity for any given cell type has been suggested. We have in previous work shown that normal human dermal fibroblasts have an inherent plasticity and can be induced to differentiate towards adipogenic, chondrogenic, endotheliogenic and osteogenic lineages when subjected to defined induction media. The aim of the present study was to further study the differentiation of human dermal fibroblasts on a gene expression level. This was achieved by employing genome wide expression analysis using microarray technology. Selected gene expression was also evaluated over time using real-time PCR. Several master regulatory genes important for lineage commitment, as well as phenotypically relevant genes, were found regulated in the respective induced cultures. The results obtained in this study strengthen previously published results showing an inherent ability for controllable phenotype alteration of human dermal fibroblasts in vitro. We conclude that adipogenic, chondrogenic, endotheliogenic and osteogenic induction results in novel phenotypes that show a genetic readiness for lineage-specific biological functionality.
2
Introduction
A classic concept of biology dictates that adult somatic cells are restricted to produce only cells of their own type. Research over the past decades has shown that certain populations have the ability to change their phenotypes and produce differentiated progeny. Several such populations have been identified and characterized as adult stem cells with the ability to differentiate into several lineages distinct from their tissue of origin [1-7]. The notion of terminally differentiated somatic cells being necessarily restricted to one phenotype has been challenged, and instead an inherent range of plasticity for any given cell type has been suggested. We have in previous work shown that normal human dermal fibroblasts have an inherent plasticity, both in unrefined primary cultures and as clonal populations [8]. We have shown that these human dermal fibroblasts may be induced to differentiate towards adipoytes, chondrocytes, endothelial cells and osteoblasts using defined induction media [8, 9]. This inherent plasticity of fibroblasts is supported by results from other groups [6, 10, 11].
All cells in the human body are progenies of the zygote, thus nearly all cells contain the exact same genome. The different phenotypes that subsequently arise are defined by their particular pattern of regulated gene expression. Differentiation involves switching from one pattern of gene expression to another [12]. Certain transcription factors are required for the development of specific tissues and control phenotype-defining expression. These are termed master regulatory genes [13, 14].
Peroxisome proliferator-activated receptor gamma 2 (PPARγ2) and the CCAAT/enhancer binding proteins (C/EBP) are known as master regulatory factors for adipocyte differentiation involved in terminal differentiation by their subsequent activation of adipocyte-specific genes [15-17]. PPARγ also plays a role in osteogenic contra adipogenic differentiation.
The transcription factor sex determining region Y-box 9 (SOX9) is considered a master regulating factor in mesenchymal condensation, leading to the formation of the cartilaginous template. SOX5 and SOX6 are co-expressed with SOX9 and are required for activation of the collagen II (COL2A1) gene and the production of cartilaginous matrix [18].
No single transcription factor has been found to be a master regulator of endothelial cell differentiation, leading to speculation that control of the phenotype is strictly combinatorial [19]. The transcription factors T-cell acute lymphocytic leukemia protein 1 (TAL1), GATA binding protein 2 (GATA2), forkhead box (FOX) family and E twenty-six (ETS) family have been suggested to be important components [19]. Some reports indicate that homeobox (HOX) A9 is the master regulator of endothelial cell fate commitment of progenitor cells. HOXA9 regulates the expression of genes such as endothelial nitric oxide synthase (eNOS), vascular endothelial growth factor receptor type 2 (KDR), and vascular endothelial (Ve)-cadherin (CDH5).
During osteoblast differentiation runt-related tran-scription factor 2 (RUNX2), osterix (SP7), and canonical wingless-type MMTV integration site family members (WNTs) play essential roles in the commitment of pluripotent mesenchymal cells to the osteoblastic lineage [20].
The aim of the present study was to investigate gene expression changes involved in differentiation of human dermal fibroblasts following exposure to specific induction media. This was achieved by using microarray analysis to investigate the full expression profile of cultures of fibroblasts before and after adipogenic, chondrogenic, endotheliogenic and osteogenic induction. Selected genes were further studied with semi-quantitative real-time PCR (qRT-PCR) analysis.
3
Materials and Methods
Primary human cell cultures Patient material used in this study was handled in accordance with ethical standards at the University of Linköping. All cells used in this study were cultured in polystyrene cell culture flasks at 37° C, 5 % CO2 and 95 % humidity unless otherwise stated.
Human dermal fibroblasts (FBs) were isolated from tissue obtained from abdominal areas of healthy patients undergoing routine surgery. Tissue was obtained fresh from the operating theatre and processed within 2 hrs. Primary cultures of FBs were prepared by dissecting the dermal layer of skin samples and subsequently incubating the fragments at 37° C over night in Dulbecco´s Modified Eagle´s Medium (DMEM) containing 2 % fetal calf serum (FCS), 165 U/ml collagenase (type 1) (Worthington, NJ) and 2.5 U/ml dispase (Gibco BRL, Life Technologies, Germany). Cells were expanded by incubation in FB proliferation medium composed of DMEM, 10 % FCS, 50 U/ml penicillin and 50 µg/ml streptomycin (PEST).
Human pre-adipocytes (PAs) were isolated and processed based on the protocol described by Entenmann et al. [21]. Briefly, sub-dermal adipose tissue was dissected and incubated in 1.15 mg/ml collagenase (type 1) for 45 min under continuous stirring. After removal of enzyme solution cells were expanded in DMEM/HAM’s F-12 (1/1 vol/vol) with 10 % FCS and PEST. PAs were differentiated to adipocytes (ACs) for 14 days using an adipocyte differentiation medium composed of DMEM supplemented with 10 % FCS, PEST, 1 µM dexamethasone (DEX), 0.5 mM isobutyl-methylxantine (IBMX), 10 µM insulin and 200 µM indomethacin.
Human articular chondrocytes (CCs) were isolated from surgical waste material following total knee arthroplasty. Millimeter-sized pieces were excised and enzymatically digested overnight in DMEM, 2 % FCS, 350 U/ml collagenase (type 2) (Gibco, Invitrogen corporation, NY) and 2.5 U/ml dispase. Cells were expanded in DMEM supplemented with 10 % FCS, PEST, 10 mM 4-(2-hydroxyethyl)-1-piperazineethane-sulfonic acid (HEPES), 0.1 mM MEM non-essential amino acid solution, 0.4 mM L-proline, and 0.2 mM ascorbate-2-phosphate (A2P).
Human Umbilical Vein Endothelial Cells (HUVECs) were obtained from umbilical cord veins received from the neonatal unit at Linköping University Hospital. The vein was cannulated and injected with
DMEM containing 165 U/ml collagenase (type 1). After 20 min, the veins were massaged and perfused with DMEM. Enzymatically detached cells were cultured in gelatin (0.2 %) coated culture flasks in DMEM supplemented with 30 % heat-inactivated human serum, PEST, 3.3 mM IBMX and 0.8 mg/mL cholera toxin.
Human osteoblasts (OBs) were obtained from surgical waste material after total hip arthroplasty. Bone was minced and incubated for 45 min in 1 mg/ml collagenase (type 2). Cells were expanded in DMEM supplemented with 10 % FCS, PEST, 1 µM DEX, 50 µM A2P and 10 mM β-glycerophosphate (BGP).
Induction of differentiation Adipogenic induction FBs were cultured in adipogenic induction medium which consisted of DMEM supplemented with 10 % FCS, PEST, 1 µM DEX, 0.5 mM IBMX, 10 µM insulin and 200 µM indomethacin for 4 weeks. Time points (TPs) for analysis were 2, 3 and 4 weeks.
Chondrogenic induction FBs were cultured in chondrogenic induction medium which consisted of DMEM supplemented with 1 % FCS, PEST, 1.125 µM insulin, 50 nM A2P and 10 ng/ml transforming growth factor β1 (TGF-β1) for 4 weeks. TPs for analysis were 1, 2, 3 and 4 weeks.
Endotheliogenic induction FBs were cultured in endotheliogenic induction medium consisting of DMEM with 30 % human serum and PEST for 15 days. TPs for analysis were 5, 10 and 15 days.
Osteogenic induction FBs were cultured in osteogenic induction medium consisting of DMEM supplemented with 10 % FCS, PEST, 0.1 µM DEX, 50 µM A2P and 10 mM BGP for 4 weeks. TPs for analysis were 1, 2, 3 and 4 weeks.
Histology and Immunohistochemistry Initial culture analyses were performed according to previously published methods [8, 9]. Briefly, adipogenic-induced FBs (A-FBs) were stained for intracellular lipid accumulation using Oil red O (ORO) [22] and analysed using antibodies towards perilipin [23] (dilution 1:200; Sigma, Sweden).
Chondrogenic-induced FBs (C-FBs) were stained with Alcian blue stain for matrix content of sulphated glucosaminoglycans [24, 25] and analysed using
4
antibodies towards aggrecan [26] (dilution 1:50) and collagen II [27] (dilution 1:100) (both from Abcam, UK).
Endotheliogenic-induced FBs (E-FBs) were assayed for uptake of Dil-labeled acetylated low density lipoprotein (Dil-Ac-LDL; 10 µg/ml, Biomedical Technologies Inc, MA) [28] and analysed using antibodies towards von Willebrand factor [29] (vWf; dilution 1:200, Dako Cytomation, Denmark) and eNOS (dilution 1:50, Santa Cruz Biotechnologies, CA).
Osteogenic induced FBs (O-FBs) were stained for calcified extracellular matrix using von Kossa stain [30] and analysed using antibodies towards osteocalcin [31] (dilution 1:500) and osteonectin [32] (dilution 1:1000) (both from Chemicon, CA).
All primary antibodies were detected using Alexa Fluor conjugated secondary antibodies (dilution 1:500; Molecular Probes, OR) and cell nuclei visualized using 4',6-diamidino-2-phenylindole (DAPI) through mounting of slides with ProLong Gold DAPI (Invitrogen, Sweden). Images of all specimens were captured using a BX51 light/fluorescence microscope (Olympus, Sweden) fitted with appropriate excitation/emission filters (U-MWIB2). Uninduced control FB cultures were analysed in the same manner as induced FB cultures. Controls used for immunohistochemical analysis included omission of primary antibodies.
RNA Isolation RNA was isolated from discrete replicate cultures, each sample containing approximately 106 cells. RNA of A-FBs, C-FBs, E-FBs and O-FBs was isolated at each TP. RNA was isolated from ACs after 14 days of preadipocyte differentiation, and from CCs, HUVECs and OBs at passage 5. Briefly, cells were detached by trypsin-EDTA treatment and washed in PBS. RNA was isolated using the NucleoSpin RNA II kit (Macherey-Nagel GmBH, Germany) according to the manufacturer’s protocol (rev. 10, 2009). RNA was eluted in 60 µl dH2O and analysed using a NanoDrop spectrophotometer (Thermo Scientific, MA) to determine purity and yield. RNA quality was further assessed by the Bioanalyzer 2100 (Agilent Technologies Inc., CA), where only samples with RNA integrity number above 9 were used. RNA samples were stored at -70° C for up to five months pending analysis.
Microarray Analysis Gene expression was analysed at the final TP for each group using GeneChip Human Gene 1.0ST arrays (Affymetrix, UK). The procedure was carried out according to manufacturer’s protocol (Affymetrix GeneChip WT ST Labeling Assay Manual rev 5, P/N 701880, with addendum rev 1, P/N 702577). A starting amount of 300 ng RNA from each sample was used. After subsequent RNA to cDNA conversion steps each sample was hybridized to a separate array and the arrays were grouped according to culture medium.
The raw data was imported into GeneSpring GX11 software (Agilent Technologies Inc., CA) and analysed using RMA16 for the initial normalization. Quality control and array correlations were checked. Probe sets with signals below the 20th percentile of raw signal were filtered out to minimize background as per manufacturer’s recommendation. Entity significance was tested using ANOVA with Tukey’s Honestly Significant Difference test and Benjamini-Hochberg corrections. Gene expression values were compared between each induced culture group and the FB control group and reported as fold change. Hierarchical clustering was created using Euclidian metric with single linkage in GeneSpring GX11.
cDNA Synthesis and Relative quantification with Real-Time PCR (qRT-PCR) Reverse transcription was carried out using High Capacity RNA-to-cDNA Synthesis Kit (Applied Biosystems, CA), according to the manufacturer’s protocol modified linearly for 50 µl reactions. Hydroxymethylbilane synthase (HMBS) was selected as a reference gene based on candidate reference gene testing (TaqMan Express Human Endogenous Control Plate, Applied Biosystems, CA) supported by microarray results. For ACs and A-FBs the following genes were investigated: fatty-acid binding protein 4 (FABP4), lipoprotein lipase (LPL), and PPARG; for CCs and C-FBs: aggrecan (ACAN), COL2A1, collagen X (COL10A1), proline/arginine-rich end leucine-rich repeat protein (PRELP) and SOX9; for HUVECs and E-FBs: angiotensin II receptor, type 1 (AGTR1), CDH5, KDR, purinergic receptor P2X, ligand-gated ion channel, 4 (P2RX4), prostaglandin E synthase (PTGES) and VWF; for OBs and O-FBs: bone-specific cadherin 11 (CDH11), inhibin beta A (INHBA), osteonectin (SPARC), osteopontin (SPP1), osteoprotegerin (TNFRSF11B) and RUNX2. Primer assays (TaqMan Gene Expression Assays 20X; Applied Biosystems, UK) containing buffered forward and reverse primers with TaqMan MGB reporter-quencher probes were
5
purchased for each gene of interest. For a complete list of primers used see table I. qRT-PCR was performed at all TPs using Fast Universal PCR MasterMix, fast 96-well plates and the ABI 7900HT Fast Real-Time PCR system were used (all purchased from Applied Biosystems).
Fold changes were evaluated using the ∆∆C(T) method [33], where induced FBs, as well as pheno-
type controls, were compared to each calibrator FB value. Results for each plate and culture were analysed statistically in GraphPad Prism 5 (GraphPad Software, CA) using the Kruskal-Wallis method with Dunn’s Multiple Comparison post-test, with p<0.05 considered significant. Graphs displaying fold change values were constructed in GraphPad Prism. Fold change values for all samples were related to control FBs, which were set to 1.
Table I. Assays used for qRT-PCR in present study.
Gene symbol NCBI Gene Reference Assay ID Context sequence
ACAN NM_013227.2 Hs00202971_m1 GCACTGTAACATAGACCAGGAGGTA
AGTR1 NM_000685.4 Hs99999095_m1 AGTCGGCACCAGGTGTATTTGATAT
CDH11 NM_001797.2 Hs00156438_m1 TTGTGGGCAGGCTTCATTCAGATAT
CDH5 NM_001795.3 Hs00174344_m1 CATCAAGCCCATGAAGCCTCTGGAT
COL10A1 NM_000493.3 Hs00166657_m1 CCATAAAGAGTAAAGGTATAGCAGT
COL2A1 NM_033150.2 Hs01060345_m1 GCTCCAATGGCAACCCTGGACCCCC
FABP4 NM_001442.2 Hs01086177_m1 ACAGGAAAGTCAAGAGCACCATAAC
HMBS NM_000190.3 Hs00609297_m1 AATGCGGCTGCAACGGCGGAAGAAA
INHBA NM_002192.2 Hs00170103_m1 ACGTTTGCCGAGTCAGGAACAGCCA
KDR NM_002253.2 Hs00911700_m1 ACACAGCAGGAATCAGTCAGTATCT
LPL NM_000237.2 Hs00173425_m1 CAGATGCCCTACAAAGTCTTCCATT
P2RX4 NM_002560.2 Hs00175706_m1 ACACACGTGCCACAACCTGCTTTTT
PPARG NM_138711.3 Hs01115512_m1 TGAAGGATGCAAGGGTTTCTTCCGG
PRELP NM_002725.3 Hs00160431_m1 CAGGTGTGTGCAGGTGCATCACCTG
PTGES NM_004878.4 Hs00610420_m1 GAAGAAGGCCTTTGCCAACCCCGAG
RUNX2 NM_001015051.2 Hs01047976_m1 GGCGCATTTCAGGTGCTTCAGAACT
SOX9 NM_000346.3 Hs00165814_m1 GGCGAGCACTCGGGGCAATCCCAGG
SPARC NM_003118.2 Hs00277762_m1 CCCATTGACGGGTACCTCTCCCACA
SPP1 NM_001040058.1 Hs00167093_m1 GAAAAGCAGCTTTACAACAAATACC
TNFRSF11B NM_002546.3 Hs00900360_m1 ATCATCCAAGATATTGACCTCTGTG
VWF NM_000552.3 Hs00169795_m1 GTCTCATCGCAGCAAAAGGAGCCTA
6
Results
Adipogenic induction Adipogenic differentiation was investigated after 4 weeks of induction, at which time the morphology of the cultures had already changed to show cells filled with characteristic lipid inclusions (fig 1A). ORO stain confirmed the presence of intracellular lipid accu-mulation (fig 1A). Furthermore A-FBs showed positive
staining for perilipin around the lipid-containing vesicles (fig 1A). Uninduced FBs did not show any such vesicles and were negative in both ORO and perilipin staining.
FABP4 and LPL were significantly up-regulated in A-FBs confirmed by qRT-PCR (fig 1B). Expression of FABP4, LPL and PPARG increased over time.
Figure 1. Adipogenic induction. A. A-FBs showed accumulation of characteristic lipid vesicles, verified by lipid staining using Oil Red O (ORO). FBs show no lipid-body inclusions. A-FBs were highly positive for perilipin, located around the lipid-containing vesicles, whereas FBs were negative. Nuclei are visualized with DAPI (blue) and the image annoted negative displays the absence of cross-reactivity of secondary antibodies. Scale bar is 100 µm for the top two images and 50 µm in the lower five. B. Graphs showing fold-changes of the genes FABP4, LPL and PPARG in A-FBs and ACs compared to FBs. Significance is indicated by * (p<0.05). TP=time point. C. Hierarchical clustering of differentially regulated genes in A-FBs and ACs compared to FBs with importance for the adipocyte phenotype, partially presented also in table II. White indicates low and black high expression.
Microarray analysis showed 1019 up-regulated genes in A-FBs and 3563 up-regulated in ACs compared to FBs, of which 202 were shared by both A-FBs and ACs. The number of genes down-regulated in A-FBs and ACs compared to FBs were 888 and 2405 respectively, of which 591 were shared. In total, expression of 1907 genes was found significantly different in A-FBs compared to FBs with a false discovery rate (FDR) of
95 at p (corrected) <0.05. The highest fold-change in A-FBs compared to FBs was 19.9 (Nik related kinase) and the lowest was -16.9 (matrix metallopeptidase 1). Median fold-changes were 1.3 up and -1.2 down. A selection of genes important for adipocyte differ-entiation and phenotype is shown in table II and a hierarchical clustering using an extended set of selected genes is shown in fig 1C.
7
Table II. Selection of regulated genes in adipogenic induced fibroblasts (A-FBs) and adipocytes (ACs) compared to control fibroblasts (FBs).
Gene symbol Full name Reference
Fold change compared to FBs
A-FBs ACs
Master regulating transcription factor genes
CEBPD CCAAT/enhancer binding protein (C/EBP), delta Cao et al. 1991 [34] 1.2 1.2
PPARG peroxisome proliferator-activated receptor gamma Tontonoz et al. 1995 [35] 1.8 4.7
PPARGC1A PPARG co-activator 1 alpha Sudden et al. 2009 [36] 3.0 2.2
SIRT1 sirtuin 1 Sugden et al. 2010 [36] 1.1 -1.2
Genes coding for important proteins
ADIPOQ adiponectin Hu et al. 1996 [37] 1.5 26.7
ADIPOR1 adiponectin receptor 1 Rasmussen et al. 2006 [38] 1.1 no change
ANGPTL4 angiopoietin-like 4 Gregoire 2001 [16] 2.3 3.15
CFD adipsin Cook et al. 1987 [39] 2.1 19.3
FABP4 fatty acid binding protein 4 Spiegelman et al. 1993 [40] 3.6 25.5
IRS1 insulin receptor substrate 1 Wu et al. 1999 [41] 1.2 1.2
IRS2 insulin receptor substrate 2 Wu et al. 1999 [41] 1.3 -1.0
KLF9 Kruppel-like factor 9 Brey et al. 2009 [42] 1.4 1.4
KLF15 Kruppel-like factor 15 Brey et al. 2009 [42] 1.4 3.6
LEP leptin Friedman et al. 1998 [43] 2.0 1.3
LEPR leptin receptor Friedman et al. 1998 [43] 1.5 -1.2
LPIN3 lipin 3 Péterfy et al. 2001 [44] 1.2 3.3
LPL lipoprotein lipase Jonasson 1984 [45] 5.0 96.3
PLIN perilipin Blanchette-Mackie et al. 1995 [23] 1.2 13.2
Chondrogenic induction Chondrogenic differentiation was investigated after 4 weeks of induction using Alcian blue stain, whereby the extracellular matrix of C-FBs was confirmed to be composed of sulphated glucosaminoglycans (fig 2A). Furthermore C-FBs showed positive immunostaining for aggrecan and collagen II (fig 2A). Uninduced FBs did not stain positively for Alcian blue, aggrecan or collagen II.
ACAN, COL2A1, COL10A1, PRELP and SOX9 were significantly up-regulated in C-FBs confirmed by qRT-PCR (fig 2B). All examined genes with exception of SOX9 seem to have a time dependent pattern of up-regulation. ACAN, COL2A1, COL10A1 and PRELP showed increase over time.
8
Figure 2. Chondrogenic induction. A. C-FBs stained with Alcian blue show matrix content of glucosaminoglycans. FBs show no staining of matrix components. C-FBs were highly positive for aggrecan and collagen II whereas FBs were negative. Nuclei are visualized with DAPI (blue) and the image annoted negative displays the absence of cross-reactivity of secondary antibodies. Scale bar for all images is 50 µm. B. Graphs showing fold-changes of the genes ACAN, COL2A1, COL10A1, PRELP and SOX9 in C-FBs and CCs compared to FBs. Significance is indicated by * (p<0.05). TP=time point. C. Hierarchical clustering of differentially regulated genes in C-FBs and CCs with importance for the chondrocyte phenotype, partially presented also in table III. White indicates low and black high expression.
Microarray analysis showed 506 up-regulated genes in C-FBs and 589 up-regulated in CCs compared to FBs, of which 246 were shared by both C-FBs and CCs. The number of genes down-regulated in C-FBs and CCs compared to FBs were 1027 and 944 respectively, of which 684 were shared. In total, expression of 1533 genes was found significantly different in C-FBs compared to FBs with a FDR of 76 at p (corrected)
<0.05. The highest fold-change in C-FB compared to FB was 12.9 (cathepsin H) and the lowest was -13.8 (cadherin 13). Median fold-changes were 1.2 up and -1.5 down. A selection of genes important for chondrocyte differentiation and phenotype is shown in table III and a hierarchical clustering using an extended set of selected genes is shown in fig 2C.
9
Table III. Selection of regulated genes in chondrogenic induced fibroblasts (C-FBs) and chondrocytes (CCs) compared to control fibroblasts (FBs).
Gene symbol Full name Reference
Fold change compared to FBs
C-FBs CCs
Master regulating transcription factor genes
SOX9 SRY (sex determining region Y)-box 9 Akiyama et al. 2002 [18] 1.1 -1.3
SOX5 SRY (sex determining region Y)-box 5 Akiyama et al. 2002 [18] 1.0 1.4
RUNX2 runt-related transcription factor 2 Eames et al. 2004 [46] 1.1 -1.6
NKX3-2 NK3 homeobox 2 Yamashita et al. 2009 [47] 1.5 1.1
Collagens
COL1A1 collagen, type I, alpha 1 Komori et al. 2009 [20] -1.1 -1.7
COL2A1 collagen, type II, alpha 1 Lefebvre et al. 1997 [48] 1.6 -1.1
COL3A1 collagen, type III, alpha 1 Eyre 2004 [49] 1.2 1.0
COL6A6 collagen type VI alpha 6 Eyre 2004 [49] 4.1 -1.1
COL9A3 collagen, type IX, alpha 3 Genzer et al. 2007 [50] 1.4 1.1
COL10A1 collagen, type X, alpha 1 Komori 2009 [20] 1.2 -2.4
COL11A1 collagen, type XI, alpha 1 Bridgewater et al. 1998 [51] -2.1 -11.4
COL11A2 collagen, type XI, alpha 2 Bridgewater et al. 1998 [51] 1.2 1.1
COL12A1 collagen, type XII, alpha 1 Cremer et al. 1998 [52] -2.0 -6.0
COL15A1 collagen, type XV, alpha 1 Cremer et al. 1998 [52] -2.3 -26.5
Genes coding for important proteins
ACAN Aggrecan Seikya et al. 2000 [53] -4.8 3.3
ADAMTS5 ADAM metallopeptidase with thrombospondin type 1 motif, 5 (aggrecanase-2)
Gendron et al. 2007 [54] 1.3 1.3
BGN Biglycan Heinegard 2009 [55] -1.6 -1.4
CILP2 cartilage intermediate protein 2 (trombospondin2) Neame et al. 1999 [56] 1.3 -1.4
COMP cartilage oligomeric matrix protein Heinegard 2009 [55] -1.5 -5.6
DCN Decorin Heinegard 2009 [55] 1.7 1.4
HAPLN1 hyaluran and proteoglycan link protein 1 Roughly 2006 [57] 1.4 1.6
LUM Lumican Heinegard 2009 [55] 1.6 1.3
MMP13 metalloproteinase 13 Leeman et al. 2002 [58] 1.1 1.5
PRELP proline/arginine-rich end leucine-rich repeat protein Oh et al. 2010 [59] 2.3 -1.0
Endotheliogenic induction Endotheliogenic differentiation was investigated after 15 days of induction. E-FBs stained positive for LDL uptake, vWf and eNOS (fig 3A). Uninduced FBs were not positive for LDL uptake, vWf or eNOS stain.
The expression of AGTR1, CDH5, KDR, P2RX4, PTGES and VWF was significantly up-regulated in E-FBs compared to FBs (fig 3B). AGTR1, CDH5, P2RX4 and
PTGES showed increased expression over time. Expression levels of CDH5, KDR and VWF were higher in HUVECs compared to E-FBs. Expression of P2RX4 was comparable in E-FBs and HUVECs, while HUVECs showed lower expression of AGTR1 and PTGES compared to FBs.
10
Figure 3. Endotheliogenic induction. A. Up-take and internalization of Dil-labeled acetylated low density lipoprotein (dil-ac-LDL) is shown by red fluorescence in E-FBs, but is not seen in FBs. E-FBs were positive for endothelial nitric oxide synthase (eNOS) and von Willebrand factor (vWf). FBs were negative for eNOS and vWf. Nuclei are visualized with DAPI (blue) and the image annoted negative displays the absence of cross-reactivity of secondary antibodies. Scale bar indicates 50 µm. B. Graphs showing fold-changes of the genes AGTR1, CDH5, KDR, P2RX4, PTGES and VWF in E-FBs and HUVECs compared to FBs. Significance is indicated by * (p<0.05). TP=time point. C. Hierarchical clustering of genes found differentially expressed in E-FBs and HUVECs compared to FBs, including genes important for endothelial phenotype partially presented also in table IV. White indicates low and black high expression.
Microarray analysis showed 1471 up-regulated genes in E-FBs and 1431 up-regulated in HUVECs compared to FBs, of which 890 were shared by both E-FBs and HUVECs. The number of genes down-regulated in E-FBs and HUVECs compared to FBs were 1951 and 1991 respectively, of which 1410 were shared. In total, expression of 3422 genes was found significantly different in E-FBs compared to FBs with a FDR of 171 at p (corrected) <0.05. The highest fold-
change in E-FBs compared to FBs was 17.3 (alcohol dehydrogenase 1B) and the lowest was -26.1 (EGF-like repeats and discoidin I-like domains 3). Median fold-changes were 1.2 up and -1.5 down. A selection of genes important for endothelial differentiation and phenotype is shown in table IV and a hierarchical clustering using an extended set of selected genes is shown in fig 3C.
11
Table IV. Selection of regulated genes in endotheliogenic induced fibroblasts (E-FBs) and endothelial cells (HUVECs) compared to control fibroblasts (FBs).
Gene symbol Full name Reference
Fold change compared to FBs
E-FBs HUVECs
Master regulating transcription factor genes
BMP4 bone morphogenetic protein 4 Moser et al. 2005 [60] 1.1 8.6
HOXA9 homeobox A9 Rossig et al. 2005 [61] 1.1 2.6
WNT1 proto-oncogene Wnt1 Woll et al. 2008 [62] 1.5 1.2
GATA2 endothelial transcription factor GATA2 Lee et al. 1991 [63] 1.1 1.5
ETV1 ETS translocation variant 1 De Val et al. [19] 1.8 -1.4
Genes coding for important proteins
AGTR1 angiotensin II receptor, type 1 Chysant 2010 [64] 3.0 -1.1
CD34 hematopoietic progenitor cell antigen CD34 Ishikawa et al. 2004 [65] 1.1 11.6
CDH5 cadherin 5, type 2 (vascular endothelium) Westweber et al. 2009 [66] 1.0 36.3
COLEC12 collectin sub-family member 12 Ohtani et al. 2001 [67] 5.8 2.7
JUP junction plakoglobin Westweber et al. 2009 [66] 1.1 5.2
KDR kinase insert domain receptor (VEGFR2) Ohtani et al. 2001 [67] 1.0 3.5
NOS3 nitric oxide synthase 3 (endothelial cell) Heller et al. 1999 [68] no change no change
P2RX4 P2X purinergic receptor, ligand-gated ion channel 4 Yamamoto et al. 2006 [69] 2.0 2.5
PECAM1 platelet/endothelial cell adhesion molecule (CD31) Ilan et al. 2003 [70] 1.1 65.0
PRCP prolylcarboxypeptidase (angiotensinase C) Mallela et al. 2009 [71] 3.7 3.4
PROCR protein C receptor, endothelial Esmon 2005 [72] 1.2 4.5
PROM1 prominin 1 (CD133) Krenning et al. 2008 [73] 1.0 4.3
PROS1 protein S (alpha) Esmon 2005 [72] 2.7 -2.8
PTGES prostaglandin E synthase Sugimoto et al. 2000 [74] 2.8 -2.5
SELE selectin E Lawrence et al. 1993 [75] 1.1 2.0
TEK TEK tyrosine kinase, endothelial (TIE2) Seegar et al. 2010 [76] 1.1 3.4
THBD Thrombomodulin Esmon 2004 [72] 1.1 2.4
TIE1 tyrosine kinase with immunoglobulin-like and EGF-like domains 1
Seegar et al. 2010 [76] 1.0 14.1
TJP2 tight junction protein 2 (zona occludens 2) Tsukita et al. 2009 [77] 1.3 5.7
VWF von Willebrand factor Heiman 1975 [78] 1.0 52.6
Osteogenic induction Osteogenic differentiation was investigated after 4 weeks of induction using von Kossa stain, confirming the presence of a calcified matrix (fig 4A). Furthermore O-FBs were positive for osteocalcin and osteonectin (fig 4A). Uninduced FBs were not positive for von Kossa stain, osteonectin or osteocalcin staining. Fig 4B shows fold changes derived from qRT-PCR analyses of the six selected genes for the
osteogenic lineage. CDH11, SPARC and TNFRSF11B showed up-regulation at most TPs, but showed no consistent time-dependency. INHBA showed increasing down-regulation over time whereas SPP1 showed strong initial down-regulation at TP1 only to increase over time, reaching up-regulation at the final TP (not significant). RUNX2 varied over time but did not reach significance in the qRT-PCR.
12
Figure 4. Osteogenic induction. A. O-FBs were positive for von Kossa, osteocalcin and osteonectin, whereas FBs were negative for all staining. Nuclei are visualized with DAPI (blue) and the image annoted negative displays the absence of cross-reactivity of secondary antibodies. Scale bar represents 100 µm in the top two images and 50 µm in the lower five. B. Graphs showing fold-changes of the genes CDH11, INHBA, RUNX2, SPARC, SPP1 and TNFRSF11B in O-FBs and OBs compared to FBs. Significance compared to FBs indicated by * (p<0.05). TP=time point. C. Hierarchical clustering of genes found differentially expressed in O-FBs and OBs compared to FBs, including genes important for the osteoblast phenotype partially presented also in table V. White indicates low and black high expression.
Microarray analysis showed 514 up-regulated genes in O-FBs and 354 up-regulated in OBs compared to FBs, of which 221 were shared by both O-FBs and OBs. The number of genes down-regulated in O-FBs and OBs compared to FBs were 534 and 694 respectively, of which 401 were shared. In total, expression of 1048 genes was found significantly different in O-FBs compared to FBs with a FDR of 52 at p (corrected)
<0.05. The highest fold-change in O-FB compared to FB was 31.6 (STEAP family member 4) and the lowest was -19.7 (matrix metallopeptidase 3). Median fold-changes were 1.27 up and -1.30 down. A selection of genes important for osteoblast differentiation and phenotype is shown in table V and a hierarchical clustering using an extended set of selected genes is shown in fig 4C.
13
Table V. Selection of regulated genes in osteogenic induced fibroblasts (O-FBs) and osteoblasts (OBs) compared to control fibroblasts (FBs).
Gene symbol Full name Reference
Fold change compared to FBs
O-FBs OBs
Master regulating transcription factor genes
RUNX2 runt-related transcription factor 2 Komori 2009 [20] 1.5 1.1
SP7 Osterix Lian et al. 2006 [79] no change no change
ATF4 activating transcription factor 4 Rached et al. 2010 [80] 1.1 1.3
FOXO1 forkhead box O1 Rached et al. 2010 [80] 2.0 1.3
Genes coding for important proteins
ALPL alkaline phosphatase, placental Cheng et al. 1994 [81] -1.2 1.7
BGLAP Osteocalcin Wada et al. 2004 [82] -1.1 1.2
CDH11 cadherin 11, type 2, OB-cadherin (osteoblast) Cheng et al. 1998 [83] 1.1 -1.7
COL8A1 collagen, type VIII, alpha 1 Pagani et al. 2005 [84] 1.8 3.1
DCN Decorin Sullivan et al. 2006 [85] 1.8 1.7
DMP1 dentin matrix acidic phosphoprotein 1 Noble 2008 [86] no change no change
IBSP integrin-binding sialoprotein Roca et al. 2005 [87] 1.0 6.1
INHBA inhibin beta A Hopwood et al. 2007 [88] -3.7 -12.2
OGN Osteoglycin Rhen et al. 2008 [89] 1.1 3.4
OMD osteomodulin (osteoadherin) Rhen et al. 2008 [89] 3.2 13.3
POSTN Periostin Oshima et al. 2002 [90] -1.6 -4.1
SMOC1 SPARC related modular calcium binding 1 Vannahme et al. 2002 [91] 1.1 2.6
SMOC2 SPARC related modular calcium binding 2 Zhang et al. 2008 [92] 12.6 1.8
SPARC Osteonectin Termine et al. 1981 [93] -1.1 -1.0
SPARCL1 SPARC-like 1 (hevin) Sullivan et al. 2006 [85] 1.4 1.2
SPP1 Osteopontin Butler 1989 [94] -2.8 -1.3
TNFRSF11B Osteoprotegerin Vidal et al. 1998 [95] 1.1 1.2
TWIST1 twist homolog 1 Yousfi et al. 2002 [96] 1.5 1.5
VIM Vimentin Lian et al. 2009 [97] -1.5 -1.2
Discussion
The results obtained in this study further confirm the previously reported differentiation capacity inherent in FBs [8, 9]. Furthermore, gene expression analysis reveals both the up-regulation of master regulatory transcription factors and important phenotypical markers relevant for the investigated phenotypes.
Microarray analysis at the end-point of differentiation was chosen to evaluate phenotype maturity rather than the sequence of expression events defining each lineage. Important gene regulation was analysed by
the more sensitive qRT-PCR approach where time-dependent expression was taken into consideration. Fold changes in the microarray were substantially underestimated compared to the qRT-PCR, which was further augmented by quantifying qRT-PCR results with a non-efficiency-corrected ∆∆C(T) method. SOX9, for example, showed 1.1-fold up-regulation in the microarray (p<0.05), but as much as 49-fold change in the qRT-PCR in C-FBs compared to FBs. In lieu of this we opted to not filter microarray results based on fold-change thresholds and instead explore specific
14
sets of genes in relation to the control FBs and the phenotype references.
Adipogenic induction Expression of PPARG was significantly higher in both ACs and A-FBs compared to control FBs (table II). PPARG was found to be increasingly up-regulated over the four weeks of induction with adipogenic medium, directionally confirming the findings of the microarray (fig 1B). Forced PPARG expression is sufficient to confer adipocyte characteristics to fibroblasts [98, 99], and has been shown also to be required for adipogenesis to occur [100]. qRT-PCR showed increased expression over time of FABP4 and LPL (fig 1B) which are known targets of PPARγ transcriptional activity [101]. Additionally, the microarray showed up-regulation of PPARGC1A and SIRT1 in the adipogenic-induced FBs indicating a priming of functions related to glucose, lipid, and energy metabolism in the cells [102]. Expectedly, we found no differential regulation relating to the genes uncoupling protein 1 (UCP1) or either cyclooxygenase 2 or 4 (COX2 or COX4) subunits (results not shown), suggesting no altered thermogenic activity of the cells [103].
Interestingly, after four weeks of differentiation only CEBPD and CEBPZ of the CEBP-family of transcription factors showed up-regulation in A-FBs compared to FBs. CEBPα is important for reciprocal induction together with PPARγ, for establishment of insulin sensitive glucose uptake (via glucose transporter type 4 (GLUT4), and activation of IRS), and the CEBPA-deficient adipose phenotype shows less lipid accumulation [41]. Absent CEBPA up-regulation after four weeks of induction may explain why increase in PPARG expression is relatively modest, why GLUT4 expression does not seem to be up-regulated and why the adipogenesis therefore does not seem completed in the induced cultures. Nonetheless, we find that both IRS 1 and 2 are up-regulated in our adipogenic induced FBs (table II).
Further genes found up-regulated in the microarray were KLF9 and KLF15, important for adipogenic differentiation [42], but surprisingly not KLF5. Both LEP and LEPR, as well as ADIPOQ and CFD are seen up-regulated in the microarray (table II) showing an inclination of the induced cells towards adipose-related physiological mechanisms. LPIN3 up-regulation indicates functional maturation of induced cells [44]. The PLIN gene, coding for perilipin, is up-regulated in both A-FBs and ACs as compared to FB controls [23].
Taken together with IHC analysis and morphological appearance, the expression analysis of the induced cultures shows obvious adipocyte characteristics, confirming an adipogenic phenotype of the A-FBs.
Chondrogenic induction SOX9 directly regulates expression of important chondrogenic markers such as COL2A1 [48, 104] and ACAN [53], and is required for expression of NKX3-2 [47], PRELP [59], and type IX and XI collagen [50, 51]. We found significant up-regulation of SOX9 in C-FBs which was validated by qRT-PCR analysis (fig 2B). SOX9 function is enhanced by SOX5 and SOX6, but we do not find significant up-regulation of SOX5 or SOX6 in C-FBs although SOX5 is up-regulated in CCs [18]. It is plausible that the expression changes are too small for detection in the current analysis.
We see a slight increase in RUNX2 expression in C-FBs, but a down-regulation in CCs. There is a balance between SOX9 and RUNX2, higher levels of SOX9 will commit cells to chondrogenesis, whereas higher levels of RUNX2 will lead to hypertrophy and osteogenesis [46]. NKX3-2 has a role in enhancing chondrocyte differentiation while, together with SOX9, repressing a terminal hypertrophic fate by inhibiting RUNX2 expression [105]. NKX3-2 is up-regulated in C-FBs and CCs, indicating together with up-regulated SOX9, that essential factors affecting early chondrogenesis is up-regulated in C-FBs (table III).
Cartilage is characterized by high content of the proteoglycan aggrecan [57]. The aggrecan gene is transcribed in CCs, but suppressed in C-FBs. On the other hand, the more sensitive qRT-PCR detects elevated levels of ACAN in C-FBs compared to FBs, even though CCs show a higher fold-change (fig 2B). Aggrecan was also found in IHC analysis of C-FBs indicating presence of the proteoglycan (fig 2A).
Collagens make up 2/3 of the dry weight of cartilaginous tissues. Collagens II, IX and XI are the most abundant in human hyaline cartilage, together with small amounts of collagens III, VI, XII and XIV [49, 52]. Hypertrophic chondrocytes produce collagen X [20]. Microarray results showed up-regulation of most of these collagens in C-FBs (table III), but many were down- regulated in CCs which together with down-regulation of SOX9 could indicate dedifferentiation of the chondrocyte phenotype. This has been described in monolayer chondrocyte cultures [106].
Many of the genes coding for non-collagenous cartilage proteins were down regulated in both C-FBs and CCs. This could be because of dedifferentiation,
15
but some reports show that non collagenous proteins are more abundant in tissue exposed to compression [107]. Our cultures of CCs and C-FBs are static, and do not in this way mimic the natural environment for chondrocytes.
Finally, the aggrecanase ADAMTS5 [54] and the collagenase MMP13 [58] are up-regulated in both C-FBs and CCs (Table III). These genes code for proteolytic enzymes involved in ECM turnover, indicating an ongoing remodeling process.
In conclusion, results from routine staining and IHC together with the expression analysis of the induced cultures shows obvious chondrocyte characteristics, confirming a chondrogenic phenotype of the C-FBs.
Endotheliogenic Induction Expression of genes important for vasculogenesis; BMP4, HOXA9, WNT1 and GATA2 was up-regulated in E-FBs and HUVECs (table IV). Of the important ETS family of transcription factors, ETV1 was found most up-regulated. Knock-out of BMP4 results in defects in mesodermal patterning and leads to reduced vasculogenesis in mice and embryonic death by day nine [108]. BMP4 has been shown to induce the expression of KDR, crucial for further development of the vasculature [108]. According to Rössig et al. the epigenetic regulation of endothelial determination is coupled to HOXA9 by linking the inhibition of histone deacetylases (HDACs) to repression of HOXA9 [61]. HOXA9 has also been reported to participate in the activation of expression of E-selectin in activated endothelial cells [109]. Over expression of WNT1 has been shown to accelerate the formation of cells with hematopoietic potential and endothelial cells in embryonic stem cell models with stromal cell co-culture, an acceleration not taking place when over-expressing the non-canonical WNT5. This suggests WNT1 signaling to be of importance for formation of progenitor cells for the hematopoietic and vascular systems [62]. A number of enhancers for genes specific for endothelial cells include GATA binding. GATA2 is essential for both blood and endothelial cell development [63].
Requirements set on endothelial cells include the expression of PECAM-1, CD34, SELE, THBD, KDR, CDH5, eNOS, VWF, TIE1 and TEK [65]. E-FBs show up-regulation of genes coding for PECAM-1, CD34, SELE, THBD, CDH5, VWF and TIE1 and thus display gene expression important for mature endothelial cells (table IV). qRT-PCR showed increased expression of CDH5, KDR and VWF in E-FBs (fig 3B). KDR expression differed significantly only after 15 days of induction,
indicating late onset of KDR expression in the induced culture. Besides the traditional endothelial cell markers, E-FBs displayed markers connected to endothelial cell functions. Main functions of endo-thelial cells are involvement in vascular barrier function, regulation of coagulation and antithrombo-genesis, participation in innate immunity and control of the vascular system regarding vascular tone and circulating agents [110]. Genes representing these functions were found up-regulated, including AGTR1, P2RX4 and PTGES (table IV). Responsiveness to angiotensin II of the blood vessels is of substantial importance in mediation of the vasopressive effects of angiotensin II and E-FBs show a supporting expression profile [64].
Additionally, E-FBs express genes designated to endothelial progenitor cells (EPCs). EPCs are considered to originate from bone marrow, and have been described as cells expressing both progenitor and mature endothelial markers [111-113]. This is the case with E-FBs which express endothelial markers and the progenitor marker prominin 1 gene (PROM1, CD133). Control HUVECs also express PROM1, which probably mirrors the fetal origin of the cells. The lack of consensus among the seemingly diverse population of cells that EPCs form aggravates categorization of E-FBs. Still, one identified subgroup of EPCs is the endothelial colony forming cells (ECFCs) also called late outgrowth endothelial progenitor cells due to the late appearance of colonies when culturing human umbilical cord blood cells. As defined by Krenning et al., ECFCs are positive for CD34, PECAM-1, KDR, PROM1 and negative for mature endothelial markers as CDH5, VWF, eNOS and TEK and lacking the monocytic lineage markers CD14 and CD45 [73]. E-FBs display expression of both progenitor and mature endothelial cell markers, as well as absence of monocytic markers.
Taken together, results imply a potential of the induced cell type for differentiation to an endothelial cell phenotype.
Osteogenic induction Expression of RUNX2 in conjunction with SP7 is an important early event for OB determination of mesenchymal precursor cells. Regulation of RUNX2 activity is modulated not so much by transcriptional control as by phosphorylation and recruitment by co-factors, illustrated by mere 1.5- to 2-fold up-regulation in osteo-committed cells compared to their mesenchymal progenitors [114]. RUNX2 is reported to increase early in differentiation and decrease with
16
maturation of the OBs [20]. Our findings are compliant with both of these data. 4ATF4 and FOXO1 play important enhancing roles in osteoblastogenesis [80]. TWIST1 is a modulator of osteogenesis, repressing RUNX2 and downstream RUNX2 targets to delay onset of osteogenesis [115]. We see a similar up-regulation of RUNX2, ATF4, FOXO1 and TWIST1 in both our OB and O-FB cultures whereas the expression of SP7 is indistinguishable in either group from the FB expression levels (table V). RUNX2 enables expression of a number of known genes [20] that are similarly expressed in control FBs as in OBs and O-FBs, making expression changes based on the microarray results difficult to ascertain. A substantial similarity between FBs and OBs has been previously reviewed by Ducy et al. [116]. Osteomodulin and osteoglycin were both up-regulated compared to FBs, further indicating osteogenic differentiation [89]. SPARC was seen up-regulated at early TPs in the qRT-PCR (fig 4B), but no significant up-regulation was detectable at TP4 in neither the qRT-PCR nor the microarray. SPARCL1 can regulate collagen fibril assembly and decorin levels [85], and indeed we find increased expression of SPARCL1 and DCN in both O-FB and OB cultures. Of the collagens, COL8A1 showed the greatest up-regulation in both O-FBs and OBs, whereas COL15A1 and COL5A1 showed the greatest down-regulation, indicating similar expression of collagens in O-FBs and OBs (table V).
In conclusion, results from routine staining and IHC together with the expression analysis of the induced cultures shows obvious osteoblast characteristics, confirming an osteogenic phenotype of the O-FBs.
Figure 5. Hierarchical clustering of microarray groupings using a sub-set of 1268 genes found significantly different in all groups compared to control FBs. This indicates that the differences in regulation of induced FBs are biologically relevant.
Each lineage was initially analysed separately, using uninduced FBs as the starting point, and the reference cultures for each lineage as approximate reference end-points. Oftentimes the similarities between control FBs and induced FBs were predominant in hierarchical clustering. However, when clustering all groups based on the gene expression differences in all groups compared to FBs, the dendrogram becomes very interesting (fig 5). This shows us that the induced differences are relevant for their respective biological context.
In this study we have further confirmed the media-induced adipogenic, chondrogenic, endotheliogenic and osteogenic differentiation of FBs on a gene expression level. This is important for further characterization of the regulatory mechanisms in-volved in the differentiation process, and will lead to a better understanding of how to control in vitro differentiation of FBs.
Acknowledgements
This work was supported by grants from The Swedish Foundation for Strategic Research. The authors wish to thank K. Briheim and A. Lönn for excellent technical assistance and Professor M. Sigvardsson for fruitful discussions.
References
1. Asakura, A., M. Komaki, and M. Rudnicki, Muscle satellite cells are multipotential stem cells that exhibit myogenic, osteogenic, and adipogenic differentiation. Differentiation, 2001. 68(4-5): p. 245-53.
2. Bjornson, C.R., et al., Turning brain into blood: a hematopoietic fate adopted by adult neural stem cells in vivo. Science, 1999. 283(5401): p. 534-7.
3. Jaiswal, N., et al., Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J Cell Biochem, 1997. 64(2): p. 295-312.
4. Johnstone, B., et al., In vitro chondrogenesis of bone marrow-derived mesenchymal progenitor cells. Exp Cell Res, 1998. 238(1): p. 265-72.
5. Pittenger, M.F., et al., M ultilineage potential of adult human mesenchymal stem cells. Science, 1999. 284(5411): p. 143-7.
6. Toma, J.G., et al., Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nat Cell Biol, 2001. 3(9): p. 778-84.
7. Zuk, P.A., et al., Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell, 2002. 13(12): p. 4279-95.
8. Karlsson, L.K., et al., Human dermal fibroblasts: a potential cell source for endothelialization of vascular grafts. Ann Vasc Surg, 2009. 23(5): p. 663-74.
9. Junker, J.P., et al., Adipogenic, chondrogenic and osteogenic differentiation of clonally derived human dermal fibroblasts. Cells Tissues Organs, 2010. 191(2): p. 105-18.
10. Bartsch, G., et al., Propagation, expansion, and multilineage differentiation of human somatic stem cells from dermal progenitors. Stem Cells Dev, 2005. 14(3): p. 337-48.
11. Lorenz, K., et al., Multilineage differentiation potential of human dermal skin-derived fibroblasts. Exp Dermatol, 2008. 17(11): p. 925-32.
12. Ben-Tabou de-Leon, S. and E.H. Davidson, Gene regulation: gene control network in development. Annu Rev Biophys Biomol Struct, 2007. 36: p. 191.
13. Lander, E.S., et al., Initial sequencing and analysis of the human genome. Nature, 2001. 409(6822): p. 860-921.
14. Pei, D., Regulation of pluripotency and reprogramming by transcription factors. J Biol Chem, 2009. 284(6): p. 3365-9.
15. Rosen, E.D. and B.M. Spiegelman, Molecular regulation of adipogenesis. Annu Rev Cell Dev Biol, 2000. 16: p. 145-71.
17
16. Gregoire, F.M., Adipocyte differentiation: from fibroblast to endocrine cell. Exp Biol Med (Maywood), 2001. 226(11): p. 997-1002.
17. Rangwala, S.M. and M.A. Lazar, Transcriptional control of adipogenesis. Annu Rev Nutr, 2000. 20: p. 535-59.
18. Akiyama, H., et al., The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev, 2002. 16(21): p. 2813-28.
19. De Val, S. and B.L. Black, Transcriptional control of endothelial cell development. Dev Cell, 2009. 16(2): p. 180-95.
20. Komori, T., Regulation of bone development and extracellular matrix protein genes by RUNX2. Cell Tissue Res. 339(1): p. 189-95.
21. Entenmann, G. and H. Hauner, Relationship between replication and differentiation in cultured human adipocyte precursor cells. Am J Physiol, 1996. 270(4 Pt 1): p. C1011-6.
22. Hausman, G.J., Techniques for studying adipocytes. Stain Technol, 1981. 56(3): p. 149-54.
23. Blanchette-Mackie, E.J., et al., Perilipin is located on the surface layer of intracellular lipid droplets in adipocytes. J Lipid Res, 1995. 36(6): p. 1211-26.
24. Lev, R. and S.S. Spicer, Specific Staining of Sulphate Groups with Alcian Blue at Low Ph. J Histochem Cytochem, 1964. 12: p. 309.
25. Mason, R.M., Observations on the glycosaminoglycans of aging bronchial cartilage studied with Alcian Blue. Histochem J, 1971. 3(6): p. 421-34.
26. Mackay, A.M., et al., Chondrogenic differentiation of cultured human mesenchymal stem cells from marrow. Tissue Eng, 1998. 4(4): p. 415-28.
27. Grigoriadis, A.E., J.N. Heersche, and J.E. Aubin, Differentiation of muscle, fat, cartilage, and bone from progenitor cells present in a bone-derived clonal cell population: effect of dexamethasone. J Cell Biol, 1988. 106(6): p. 2139-51.
28. Voyta, J.C., et al., Identification and isolation of endothelial cells based on their increased uptake of acetylated-low density lipoprotein. J Cell Biol, 1984. 99(6): p. 2034-40.
29. Denis, C.V., Molecular and cellular biology of von Willebrand factor. Int J Hematol, 2002. 75(1): p. 3-8.
30. Bills, C.E., H. Eisenberg, and S.L. Pallante, Complexes of organic acids with calcium phosphate: the Von Kossa stain as a clue to the composition of bone mineral. Johns Hopkins Med J, 1974. 128(4): p. 194-207.
31. Beresford, J.N., et al., Production of osteocalcin by human bone cells in vitro. Effects of 1,25(OH)2D3, 24,25(OH)2D3, parathyroid hormone, and glucocorticoids. Metab Bone Dis Relat Res, 1984. 5(5): p. 229-34.
32. Malaval, L., et al., Cellular expression of bone-related proteins during in vitro osteogenesis in rat bone marrow stromal cell cultures. J Cell Physiol, 1994. 158(3): p. 555-72.
33. Livak, K.J. and T.D. Schmittgen, Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods, 2001. 25(4): p. 402-8.
34. Cao, Z., R.M. Umek, and S.L. McKnight, Regulated expression of three C/EBP isoforms during adipose conversion of 3T3-L1 cells. Genes Dev, 1991. 5(9): p. 1538-52.
35. Hu, E., P. Tontonoz, and B.M. Spiegelman, Transdifferentiation of myoblasts by the adipogenic transcription factors PPAR gamma and C/EBP alpha. Proc Natl Acad Sci U S A, 1995. 92(21): p. 9856-60.
36. Sugden, M.C., P.W. Caton, and M.J. Holness, PPAR control: it's SIRTainly as easy as PGC. J Endocrinol. 204(2): p. 93-104.
37. Hu, E., P. Liang, and B.M. Spiegelman, AdipoQ is a novel adipose-specific gene dysregulated in obesity. J Biol Chem, 1996. 271(18): p. 10697-703.
38. Rasmussen, M.S., et al., Adiponectin receptors in human adipose tissue: effects of obesity, weight loss, and fat depots. Obesity (Silver Spring), 2006. 14(1): p. 28-35.
39. Cook, K.S., et al., Adipsin: a circulating serine protease homolog secreted by adipose tissue and sciatic nerve. Science, 1987. 237(4813): p. 402-5.
40. Spiegelman, B.M., et al., Regulation of adipocyte gene expression in differentiation and syndromes of obesity/diabetes. J Biol Chem, 1993. 268(10): p. 6823-6.
41. Wu, Z., et al., Cross-regulation of C/EBP alpha and PPAR gamma controls the transcriptional pathway of adipogenesis and insulin sensitivity. Mol Cell, 1999. 3(2): p. 151-8.
42. Brey, C.W., et al., Kruppel-like family of transcription factors: an emerging new frontier in fat biology. Int J Biol Sci, 2009. 5(6): p. 622-36.
43. Friedman, J.M. and J.L. Halaas, Leptin and the regulation of body weight in mammals. Nature, 1998. 395(6704): p. 763-70.
44. Peterfy, M., et al., Lipodystrophy in the fld mouse results from mutation of a new gene encoding a nuclear protein, lipin. Nat Genet, 2001. 27(1): p. 121-4.
45. Jonasson, L., et al., Immunohistochemical localization of lipoprotein lipase in human adipose tissue. Atherosclerosis, 1984. 51(2-3): p. 313-26.
46. Eames, B.F., P.T. Sharpe, and J.A. Helms, Hierarchy revealed in the specification of three skeletal fates by Sox9 and Runx2. Dev Biol, 2004. 274(1): p. 188-200.
47. Yamashita, S., et al., Sox9 directly promotes Bapx1 gene expression to repress Runx2 in chondrocytes. Exp Cell Res, 2009. 315(13): p. 2231-40.
48. Lefebvre, V., et al., SOX9 is a potent activator of the chondrocyte-specific enhancer of the pro alpha1(II) collagen gene. Mol Cell Biol, 1997. 17(4): p. 2336-46.
49. Eyre, D.R., Collagens and cartilage matrix homeostasis. Clin Orthop Relat Res, 2004(427 Suppl): p. S118-22.
50. Genzer, M.A. and L.C. Bridgewater, A Col9a1 enhancer element activated by two interdependent SOX9 dimers. Nucleic Acids Res, 2007. 35(4): p. 1178-86.
51. Bridgewater, L.C., V. Lefebvre, and B. de Crombrugghe, Chondrocyte-specific enhancer elements in the Col11a2 gene resemble the Col2a1 tissue-specific enhancer. J Biol Chem, 1998. 273(24): p. 14998-5006.
52. Cremer, M.A., E.F. Rosloniec, and A.H. Kang, The cartilage collagens: a review of their structure, organization, and role in the pathogenesis of experimental arthritis in animals and in human rheumatic disease. J Mol Med, 1998. 76(3-4): p. 275-88.
53. Sekiya, I., et al., SOX9 enhances aggrecan gene promoter/enhancer activity and is up-regulated by retinoic acid in a cartilage-derived cell line, TC6. J Biol Chem, 2000. 275(15): p. 10738-44.
54. Gendron, C., et al., Proteolytic activities of human ADAMTS-5: comparative studies with ADAMTS-4. J Biol Chem, 2007. 282(25): p. 18294-306.
55. Heinegard, D., Proteoglycans and more--from molecules to biology. Int J Exp Pathol, 2009. 90(6): p. 575-86.
56. Neame, P.J., H. Tapp, and A. Azizan, Noncollagenous, nonproteoglycan macromolecules of cartilage. Cell Mol Life Sci, 1999. 55(10): p. 1327-40.
57. Roughley, P.J., The structure and function of cartilage proteoglycans. Eur Cell Mater, 2006. 12: p. 92-101.
18
58. Leeman, M.F., S. Curran, and G.I. Murray, The structure, regulation, and function of human matrix metalloproteinase-13. Crit Rev Biochem Mol Biol, 2002. 37(3): p. 149-66.
59. Oh, C.D., et al., Identification of SOX9 interaction sites in the genome of chondrocytes. PLoS One. 5(4): p. e10113.
60. Moser, M. and C. Patterson, Bone morphogenetic proteins and vascular differentiation: BMPing up vasculogenesis. Thromb Haemost, 2005. 94(4): p. 713-8.
61. Rossig, L., et al., Histone deacetylase activity is essential for the expression of HoxA9 and for endothelial commitment of progenitor cells. J Exp Med, 2005. 201(11): p. 1825-35.
62. Woll, P.S., et al., Wnt signaling promotes hematoendothelial cell development from human embryonic stem cells. Blood, 2008. 111(1): p. 122-31.
63. Lee, M.E., et al., Cloning of the GATA-binding protein that regulates endothelin-1 gene expression in endothelial cells. J Biol Chem, 1991. 266(24): p. 16188-92.
64. Chrysant, S.G., Current status of dual Renin Angiotensin aldosterone system blockade for the treatment of cardiovascular diseases. Am J Cardiol. 105(6): p. 849-52.
65. Ishikawa, M. and T. Asahara, Endothelial progenitor cell culture for vascular regeneration. Stem Cells Dev, 2004. 13(4): p. 344-9.
66. Vestweber, D., et al., Cell adhesion dynamics at endothelial junctions: VE-cadherin as a major player. Trends Cell Biol, 2009. 19(1): p. 8-15.
67. Ohtani, K., et al., The membrane-type collectin CL-P1 is a scavenger receptor on vascular endothelial cells. J Biol Chem, 2001. 276(47): p. 44222-8.
68. Heller, R., et al., L-Ascorbic acid potentiates nitric oxide synthesis in endothelial cells. J Biol Chem, 1999. 274(12): p. 8254-60.
69. Yamamoto, K., et al., Impaired flow-dependent control of vascular tone and remodeling in P2X4-deficient mice. Nat Med, 2006. 12(1): p. 133-7.
70. Ilan, N. and J.A. Madri, PECAM-1: old friend, new partners. Curr Opin Cell Biol, 2003. 15(5): p. 515-24.
71. Mallela, J., J. Yang, and Z. Shariat-Madar, Prolylcarboxypeptidase: a cardioprotective enzyme. Int J Biochem Cell Biol, 2009. 41(3): p. 477-81.
72. Esmon, C.T., The impact of the inflammatory response on coagulation. Thromb Res, 2004. 114(5-6): p. 321-7.
73. Krenning, G., et al., Generating new blood flow: integrating developmental biology and tissue engineering. Trends Cardiovasc Med, 2008. 18(8): p. 312-23.
74. Sugimoto, Y., S. Narumiya, and A. Ichikawa, Distribution and function of prostanoid receptors: studies from knockout mice. Prog Lipid Res, 2000. 39(4): p. 289-314.
75. Lawrence, M.B. and T.A. Springer, Neutrophils roll on E-selectin. J Immunol, 1993. 151(11): p. 6338-46.
76. Seegar, T.C., et al., Tie1-Tie2 interactions mediate functional differences between angiopoietin ligands. Mol Cell. 37(5): p. 643-55.
77. Tsukita, S., et al., Roles of ZO-1 and ZO-2 in establishment of the belt-like adherens and tight junctions with paracellular permselective barrier function. Ann N Y Acad Sci, 2009. 1165: p. 44-52.
78. Heiman, M.F., Police suicides revisted. Suicide, 1975. 5(1): p. 5-20.
79. Lian, J.B., et al., Networks and hubs for the transcriptional control of osteoblastogenesis. Rev Endocr Metab Disord, 2006. 7(1-2): p. 1-16.
80. Rached, M.T., et al., FoxO1 is a positive regulator of bone formation by favoring protein synthesis and resistance to oxidative stress in osteoblasts. Cell Metab. 11(2): p. 147-60.
81. Cheng, S.L., et al., Differentiation of human bone marrow osteogenic stromal cells in vitro: induction of the osteoblast phenotype by dexamethasone. Endocrinology, 1994. 134(1): p. 277-86.
82. Wada, S., T. Fukawa, and S. Kamiya, [Osteocalcin and bone]. Clin Calcium, 2007. 17(11): p. 1673-7.
83. Cheng, S.L., et al., Human osteoblasts express a repertoire of cadherins, which are critical for BMP-2-induced osteogenic differentiation. J Bone Miner Res, 1998. 13(4): p. 633-44.
84. Pagani, F., C.M. Francucci, and L. Moro, Markers of bone turnover: biochemical and clinical perspectives. J Endocrinol Invest, 2005. 28(10 Suppl): p. 8-13.
85. Sullivan, M.M., et al., Matricellular hevin regulates decorin production and collagen assembly. J Biol Chem, 2006. 281(37): p. 27621-32.
86. Noble, B.S., The osteocyte lineage. Arch Biochem Biophys, 2008. 473(2): p. 106-11.
87. Roca, H., et al., Cooperative interactions between RUNX2 and homeodomain protein-binding sites are critical for the osteoblast-specific expression of the bone sialoprotein gene. J Biol Chem, 2005. 280(35): p. 30845-55.
88. Hopwood, B., et al., Microarray gene expression profiling of osteoarthritic bone suggests altered bone remodelling, WNT and transforming growth factor-beta/bone morphogenic protein signalling. Arthritis Res Ther, 2007. 9(5): p. R100.
89. Rehn, A.P., et al., Osteoadherin is upregulated by mature osteoblasts and enhances their in vitro differentiation and mineralization. Calcif Tissue Int, 2008. 82(6): p. 454-64.
90. Oshima, A., et al., A novel mechanism for the regulation of osteoblast differentiation: transcription of periostin, a member of the fasciclin I family, is regulated by the bHLH transcription factor, twist. J Cell Biochem, 2002. 86(4): p. 792-804.
91. Vannahme, C., et al., Characterization of SMOC-1, a novel modular calcium-binding protein in basement membranes. J Biol Chem, 2002. 277(41): p. 37977-86.
92. Zhang, M., et al., Microarray analysis of perichondral and reserve growth plate zones identifies differential gene expressions and signal pathways. Bone, 2008. 43(3): p. 511-20.
93. Termine, J.D., et al., Osteonectin, a bone-specific protein linking mineral to collagen. Cell, 1981. 26(1 Pt 1): p. 99-105.
94. Butler, W.T., The nature and significance of osteopontin. Connect Tissue Res, 1989. 23(2-3): p. 123-36.
95. Vidal, N.O., et al., Osteoprotegerin mRNA is expressed in primary human osteoblast-like cells: down-regulation by glucocorticoids. J Endocrinol, 1998. 159(1): p. 191-5.
96. Yousfi, M., F. Lasmoles, and P.J. Marie, TWIST inactivation reduces CBFA1/RUNX2 expression and DNA binding to the osteocalcin promoter in osteoblasts. Biochem Biophys Res Commun, 2002. 297(3): p. 641-4.
97. Lian, N., et al., Vimentin inhibits ATF4-mediated osteocalcin transcription and osteoblast differentiation. J Biol Chem, 2009. 284(44): p. 30518-25.
98. Spiegelman, B.M., et al., PPAR gamma and the control of adipogenesis. Biochimie, 1997. 79(2-3): p. 111-2.
99. Tontonoz, P., E. Hu, and B.M. Spiegelman, Regulation of adipocyte gene expression and differentiation by peroxisome proliferator activated receptor gamma. Curr Opin Genet Dev, 1995. 5(5): p. 571-6.
100. Rosen, E.D., et al., PPAR gamma is required for the differentiation of adipose tissue in vivo and in vitro. Mol Cell, 1999. 4(4): p. 611-7.
101. Tontonoz, P., E. Hu, and B.M. Spiegelman, Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell, 1994. 79(7): p. 1147-56.
19
102. Sugden, M.C., P.W. Caton, and M.J. Holness, PPAR control: it's SIRTainly as easy as PGC. J Endocrinol, 2010. 204(2): p. 93-104.
103. Puigserver, P., et al., Involvement of the retinoblastoma protein in brown and white adipocyte cell differentiation: functional and physical association with the adipogenic transcription factor C/EBPalpha. Eur J Cell Biol, 1998. 77(2): p. 117-23.
104. Bell, D.M., et al., SOX9 directly regulates the type-II collagen gene. Nat Genet, 1997. 16(2): p. 174-8.
105. Zeng, L., et al., Shh establishes an Nkx3.2/Sox9 autoregulatory loop that is maintained by BMP signals to induce somitic chondrogenesis. Genes Dev, 2002. 16(15): p. 1990-2005.
106. Schnabel, M., et al., Dedifferentiation-associated changes in morphology and gene expression in primary human articular chondrocytes in cell culture. Osteoarthritis Cartilage, 2002. 10(1): p. 62-70.
107. Wong, M., M. Siegrist, and X. Cao, Cyclic compression of articular cartilage explants is associated with progressive consolidation and altered expression pattern of extracellular matrix proteins. Matrix Biol, 1999. 18(4): p. 391-9.
108. Rhoads, K., et al., A role for Hox A5 in regulating angiogenesis and vascular patterning. Lymphat Res Biol, 2005. 3(4): p. 240-52.
109. Bandyopadhyay, S., et al., HOXA9 participates in the transcriptional activation of E-selectin in endothelial cells. Mol Cell Biol, 2007. 27(12): p. 4207-16.
110. Bachetti, T. and L. Morbidelli, Endothelial cells in culture: a model for studying vascular functions. Pharmacol Res, 2000. 42(1): p. 9-19.
111. Mund, J.A., et al., Endothelial progenitor cells and cardiovascular cell-based therapies. Cytotherapy, 2009. 11(2): p. 103-13.
112. Medina, R.J., et al., Molecular analysis of endothelial progenitor cell (EPC) subtypes reveals two distinct cell populations with different identities. BMC Med Genomics. 3: p. 18.
113. Hristov, M., W. Erl, and P.C. Weber, Endothelial progenitor cells: mobilization, differentiation, and homing. Arterioscler Thromb Vasc Biol, 2003. 23(7): p. 1185-9.
114. Franceschi, R.T., et al., Transcriptional regulation of osteoblasts. Ann N Y Acad Sci, 2007. 1116: p. 196-207.
115. Bialek, P., et al., A twist code determines the onset of osteoblast differentiation. Dev Cell, 2004. 6(3): p. 423-35.
116. Ducy, P., T. Schinke, and G. Karsenty, The osteoblast: a sophisticated fibroblast under central surveillance. Science, 2000. 289(5484): p. 1501-4.
