Adult mesenchymal stem cells: potential for muscle and tendon regeneration and use in gene therapy.

Abstract

The expansion potential and plasticity of stem cells, adult or embryonic, offer great promise for their use in medical therapies. Recent provocative data suggest that the differentiation potential of adult stem cells may extend to lineages beyond those usually associated with the germ layer of origin. In this review, we describe recent developments related to adult stem cell research and in particular, in the arena of mesenchymal stem cell (MSC) research. Research demonstrates that transduced MSCs injected into skeletal muscle can persist and express secreted gene products. The ability of the MSC to differentiate into cardiomyocytes has been reported and their ability to engraft and modify the pathology in infarcted animal models is of great interest. Research using MSCs in tendon repair provides information on the effects of physical forces on phenotype and gene expression. In turn, MSCs produce changes in their matrix environment in response to those biomechanical forces. Recent data support the potential of MSCs to repair tendon, ligament, meniscus and other connective tissues. Therapeutic applications of adult stem cells are approaching clinical use in several fields, furthering the possibility to regenerate damaged and diseased tissue.

Full text

Page 1

309
J Musculoskel Neuron Interact 2002; 2(4):309-320
Review Article
Adult mesenchymal stem cells: Potential for muscle and
tendon regeneration and use in gene therapy
M. Pittenger, P. Vanguri, D. Simonetti, R. Young
Osiris Therapeutics, Inc., Baltimore, Maryland, USA
Abstract
The expansion potential and plasticity of stem cells, adult or embryonic, offer great promise for their use in medical
therapies. Recent provocative data suggest that the differentiation potential of adult stem cells may extend to lineages beyond
those usually associated with the germ layer of origin. In this review, we describe recent developments related to adult stem
cell research and in particular, in the arena of mesenchymal stem cell (MSC) research. Research demonstrates that transduced
MSCs injected into skeletal muscle can persist and express secreted gene products. The ability of the MSC to differentiate into
cardiomyocytes has been reported and their ability to engraft and modify the pathology in infarcted animal models is of great
interest. Research using MSCs in tendon repair provides information on the effects of physical forces on phenotype and gene
expression. In turn, MSCs produce changes in their matrix environment in response to those biomechanical forces. Recent
data support the potential of MSCs to repair tendon, ligament, meniscus and other connective tissues. Therapeutic
applications of adult stem cells are approaching clinical use in several fields, furthering the possibility to regenerate damaged
and diseased tissue.
Keywords: Stem Cells, Mesenchymal Stem Cells, Adult Stem Cells, Differentiation, Tissue Regeneration, Myogenic, Cardiac,
Gene Therapy, Tendon
Introduction
Clearly, one of the most exciting recent developments in
biology and medicine is the many reports that stem cells exist
in the adult that can regenerate damaged and diseased tissue.
Many adult tissues are known to contain stem cells that
provide for the new cells for the normal tissue turnover that is
known to occur. As examples, this includes liver parenchymal
cells that can regenerate large segments of that organ, the
regenerative satellite cells resident in skeletal muscle,
several epithelial cell types such as the intestinal crypt cells
that provide for the continued turnover of the intestinal
epithelia. Stem cells also have been isolated from the embryo
that can likely generate many tissues1,2and the isolation of
the human embryonic stem (ES)3and embryonic germ
(EG)4cells brings this prospect closer to reality. While there
has been some effort to direct ES cells to different lineages5,
the indication that differentiation of ES cells can be
controlled in a predictable manner is not yet evident and
various cell types are seen in the same culture dish. Almost
concurrent with the human ES and EG cell reports were
several publications describing stem cells from adult
tissues6-8. Our own work characterized human bone marrow-
derived cells that could be greatly expanded to a
homogeneous population (or grown clonally), would
differentiate to several discrete lineages and fulfilled the
criteria to be called mesenchymal stem cells6(Figure 1).
Another report offered evidence that while hepatocytes have
great ability to regenerate functioning liver tissue, there are
also stem cells in blood with this capability7. A provocative
study utilized neural stem cells to repopulate the
hematopoietic lineages and rescue a lethally irradiated mouse,
perhaps one of the first reports that stem cells may have
great plasticity beyond the lineages for which they were first
characterized8. Particularly important is the predictability
and uniformity of the isolation process from one preparation
of stem cells to the next. In this regard, the adult stem cells
isolated from bone marrow continue to offer promising
Corresponding author: Mark Pittenger, Osiris Therapeutics, Inc., 2001
Aliceanna Street, Baltimore, MD 21231, USA
E-mail: mpittenger@osiristx.com
Accepted 5 March 2002
Hylonome

Page 2

310
results for tissue regeneration. The challenge is to characterize
these cells, understand their regenerative potential, and to
develop useful therapies. Whether the adult bone marrow-
derived MSCs can be used to regenerate tissues outside the
mesodermal lineage is also under investigation in several
laboratories. Below, we briefly review the developments in
this field that support the prospects for therapeutic use of
MSCs and highlight work on muscle and tendon regeneration.
Adult bone marrow has proven to be a reliable source of
multipotential cells, first the hematopoietic stem cell (HSC)
and now the mesenchymal stem cell (MSC).
Bone marrow stroma is a tissue composed of many cell types
that fills the intramedullary space of bone. It contains
osteoblasts, adipocytes, reticular cells, endothelial cells,
macrophages, monocytes etc., as well as progenitors for these
cell types. Bone marrow contains niches that provide support
for the maintenance of hematopoietic stem cells, and the
mesenchymal stem cells. Therefore, an important function of
marrow stroma is to provide an environment that prevents the
differentiation of stem cells. It is important to keep in mind
that while bone marrow contains mesenchymal stem cells,
clearly not all bone marrow stromal cells are stem cells.
Bone marrow can be drawn easily under local anesthetic
from accessible sites such as the posterior iliac crest. While
other tissues may contain MSCs, their collection is more
difficult and/or traumatic for the donor. It is likely that such
stem cells can be isolated from many tissues, but which tissues
provide the most accessible cells that retain the greatest
potential with the least trauma to the donor? Keeping in mind
that we are only beginning to understand the potential of stem
cells that persist in the adult, much evidence to date would
recommend bone marrow MSCs. We routinely utilize a modified
version of the isolation procedure first developed in the Caplan
laboratory and this has been presented in detail elsewhere6,9,10.
The process of tissue healing has always fascinated the
observer. Literature on new tissue formation before 1900
comes largely from embryologists and from the accounts of
surgeons, often in military service. For the embryologist,
there were few reagents available such as antibodies,
purified proteins, or gene probes that we rely on heavily
today for following cellular and molecular events during
tissue remodeling. Moreover the first antibiotics,
sulfonamides, were not available until the 1930s, so many
in vitro experiments were hampered by the necessity to
maintain absolute sterility. Therefore, many of these early
investigations were performed in vivo where the immune
system provided bacterial surveillance. The surgeons’
descriptions of events in the ongoing healing process were
largely observational at limited time points, and the process
could not easily be followed longitudinally. Moreover,
publications, including textbooks, made limited use of
pictures and figures.
Studies on effects of exposure to radioactive substances in
the 1940s led to the identification of spleen and bone
marrow as the source of progenitor cells for the
hematopoietic system11-13. These early studies led to the
development of marrow ablation and therapeutic bone
marrow transplantation and the realization that there must
be an isolatable hematopoietic stem cell (HSC). The search
for greater understanding of the characteristics of this adult
stem cell is 30-40 years old and continues to be a very active
field at both the research and clinical levels14.
Early tissue transplantation studies revealed that cells
responded to being placed in a new environment by altering
their phenotype15. The transplantation of whole bone
marrow to an ectopic site demonstrated the formation of
bone and cartilage16,17. In the 1960s, Alexander Friedenstein
cultured cells from guinea pig bone marrow and studied the
cells that became attached to the culture dish18,19. These cells
were mitotically very active and formed bone when
implanted in syngeneic hosts. They were described as
fibroblastic colony forming cells (FCFC) or osteogenic
progenitor cells (OPCs), later more commonly referred to as
colony forming units-fibroblastic (CFU-F). Experiments
M. Pittenger et al.: Adult stem cells for connective tissues
Figure 1. Human mesenchymal stem cells isolated from bone marrow taken from the iliac crest can be induced to differentiate to several
lineages. A) phase micrograph of log phase hMSCs B) Confluent hMSCS can be readily differentiated into adipocytes as shown here with
the lipid vacuoles stained with oil red O C) A micromass of hMSCs can be induced to become chondrocytes as shown here, stained with anti-
type II collagen D) Monolayers of hMSCS differentiate to osteoblasts and elevate expression of alkaline phosphatase (red stain) and
accumulate calcium deposits (dark areas) E) hMSCs can serve a stromal function to support HSCs and their progeny which form
"cobblestones" on top of the hMSCS (hMSCS are underneath and not visible).
ABCDE

Page 3

311
placing these marrow-derived cells in diffusion chambers to
separate them from the host tissue proved that the
implanted cultured cells were the source of the new bone.
These studies of cells cultured in vitro from bone marrow
were the first to test their ability to differentiate in vivo.
In vivo cell labeling with 3H-thymidine prior to isolation from
marrow failed to detect labeled cells in subsequent colonies,
showing that the cells giving rise to in vitro colonies were not
cycling in vivo. Further work by Friedenstein and colleagues
characterized similar cells from mice and rabbits, and, more
recently, from humans20. Dr. Friedenstein produced scientific
findings in several areas during his career. With the
continuing interest in marrow-derived stem cells, his legacy
will certainly live on.
During the late 1970s and 80s, there was a period of
characterization of the stromal support cells present in
marrow responsible for maintaining cultures of HSCs. The
culture of human bone marrow cells that could propagate as
attached, fibroblastic cells was first described by Moore and
co-workers21. Experimentally, treatment with high doses of
3H-thymidine at initial isolation of the cells failed to reduce
their initial growth, suggesting the cells were non-mitotic or
quiescent at first isolation. These authors characterized the
cells’ morphology and several of their expressed proteins.
They were distinguished from endothelial cells on the basis
of lacking Weibel-Palade bodies, expression of factor VIII or
basement membrane collagens, or support of granulocytes
and macrophages. However, there was not an attempt to
characterize the ability of these cells to differentiate to other
lineages.
The work of Maureen Owen and colleagues extensively
explored the osteogenic potential of cells cultured from bone
marrow22,23. Bone marrow was known to contain the HSCs
and a variety of cell types, but isolated cultured bone forming
cells were not well studied. It was clear that osteoblasts were
present in marrow but what other cell types could be
identified? Several reports described the osteogenic potential
of isolated, cultured cells from bone marrow that were not
pre-committed osteoblasts. These cells were also referred to
as stromal stem cells. Whether other organs contained such
cells was unclear and perhaps stromal stem cells from
different organs differed in their properties. There was also
the question whether each stromal stem cell was capable of
producing a different stromal cell type, or more than one cell
type. A working diagram initially proposed by Owen24and by
Owen and Friedenstein25and further developed by Arnold
Caplan26and included the stem cells that could give rise to all
mesenchymal lineages –the mesenchymal stem cells or
MSCs. Working with Caplan, Steve Haynesworth was able to
isolate and characterize human cells from adult bone marrow
that formed bone and cartilage in an in vivo setting27. Some
criticism over the term mesenchymal stem cell resulted in
publications using the name mesenchymal progenitor cell
(MPC) to refer to the same cells28.
Over the years, the fibroblastic cells from bone marrow
isolated from different species and described by different
laboratories have been given a variety of names. The
isolation techniques, culture conditions, media additives,
subculturing techniques, and assays vary among the
laboratories studying these cells. It is likely that, similar to
HSCs, the expressed surface molecules will depend on
careful attention to many factors (composition of the
medium, homogeneity of the cells, time in culture, cell
density, time since last feeding, etc.). While the names likely
describe similar cells, it is clear that the cultured cells studied
in various laboratories are not identical. With this in mind we
have extensively characterized the expression of surface
proteins on the human MSCs with which we work6. It is likely,
however, that a short list of surface molecules would be helpful
to characterize the multipotential cells isolated in many
laboratories. We suggest that a useful group of antibodies to
characterize human MSCs by flow cytometry would be
CD90, 29, 105, 73, 166, HLA-A,B,C (positive); CD 45, 54,
106 (positive but variable); CD14, 34, 117 (negative) where
the analysis is done with 3 positives and 1 negative staining
antibodies per tube and the blank consists of the control IgG
run at the time of analysis. It should be recognized that a set
of cell markers will not necessarily identify a stem cell or the
outcome of the differentiation that is to follow (as this is
dependent on many additional parameters), but it speaks to
the identity and uniformity of the cells under study.
Recently, Catherine Verfaillie and colleagues have
characterized cells from adult human bone marrow that
grow in culture through extensive passaging, and
differentiate to mesenchymal lineages but they have
evidence for non-mesenchymal lineages. Under the right
conditions, these cells also express genes normally found in
neural or endothelial lineages. The authors have termed
these cells mesenchymal progenitor cells (MPC) or
multipotential adult progenitor cells (MAPC), both
referring to the same cell population.
Myogenic potential of bone marrow-derived cells
That bone marrow-derived mesenchymal cells should
differentiate to muscle was investigated by Caplan and
colleagues who studied the in vitro differentiation of rat
MSCs to myoblasts. They were able to demonstrate that 24
hours of 5-azacytidine treatment of rat MSCs in culture
resulted in elongated myotubes that stained positively for
myosin29. Further, they were able to demonstrate in other
work that mouse bone marrow-derived MSCs from an
unaffected sibling could provide dystrophin to the muscle
fibers of affected mice30. More recently, Cossu, Mavillo and
co-workers demonstrated active muscle regeneration in vivo
with bone marrow-derived cells from the C57/MlacZ
M. Pittenger et al.: Adult stem cells for connective tissues

Page 4

312
transgenic mice that express the ‚-Gal gene downstream of
a myosin light chain promoter31. This ensures continued
expression of the transgene in muscle tissues. The healing
muscle clearly showed the incorporation of the ‚gal+donor
cells. The results were similar if the bone marrow was first
fractionated into adherent and non-adherent cell types and
then injected. Further, if animals were irradiated and
repopulated with the transgenic bone marrow cells first,
regenerating muscle showed the presence of the ‚-Gal+
cells, further suggesting that recruitment from the bone
marrow compartment through the circulation is a common
mechanism. The bone marrow cells appeared to
differentiate more slowly than similarly implanted muscle
satellite cells, suggesting that the satellite cells are the first
cells recruited to heal muscle damage, while bone marrow-
derived cells may take part in prolonged regeneration. The
potential for gene-modified bone marrow-derived cells or
the more fully characterized MSCs is being evaluated in
several laboratories (see below).
Recent work has characterized a subpopulation of muscle
satellite cells in the mouse that actively exclude nuclear dyes
and can be distinguished by flow cytometry as a "side
population" or SP cells32. These SP cells have a tremendous
ability to proliferate and provide myoblasts for muscle
regeneration. They also appear to be able to differentiate to
additional lineages and therefore represent another example
of a multipotential stem cell that persists in adult tissue33.
Reasoning that the constant recruiting of myogenic
progenitor cells necessary in muscular dystrophies would
exhaust the satellite cell population in a muscle and require
another source of progenitor cells, Gussoni and colleagues
were able to restore dystrophin expression in the mdx mouse
model by a bone marrow transplant, or by using the SP
population from donor marrow34. Male donor Y chromosomes
could be detected in the muscle nuclei of the recipient
females.
Myogenic differentiation of stem cells in heart
The great mortality associated with heart disease has
prompted research into the replenishment of the
myocardium with additional cells, a procedure referred to as
cardiomyoplasty. The majority of the work in this area has
sought to utilize myoblasts but many cell types have been
utilized. Recently, several articles have demonstrated that
cells from bone marrow can engraft in the heart. Bittner et
al. performed bone marrow transplants from male donors
into affected mdx female mice and searched for the donor
cells in muscle and heart tissues. They were able to detect
the Y chromosome by fluorescence in situ hybridization
(FISH) in the mdx animals, in skeletal and heart muscle, and
these cells were often found to express myogenin, myf5 and
dystrophin. Chiu and colleagues have reported the
engraftment of cultured syngeneic bone marrow stromal
cells into the hearts of rats36. The cells were labeled with the
nuclear dye DAPI and then injected directly into the
ventricle wall of recipients. The animals were sacrificed at
times out to 12 weeks. At 4 weeks and beyond, tissue sections
of the injected hearts revealed the presence of engrafted
donor cells. The donor cells were seen to express sarcomeric
myosin heavy chain and the gap junction protein connexin
43, suggesting functional coupling with neighboring host
cardiomyocytes.
In another recent study, Orlic, Anversa and colleagues
tested the ability of mouse bone marrow-derived cells to
engraft in the damaged area of the ischemic mouse heart37.
They utilized bone marrow from transgenic male mice
expressing the enhanced green fluorescent protein (GFP).
The cells to be injected were not cultured in vitro, but were
selected for the presence of c-kitposand the lack of several
hematopoietic lineage markers (Linneg). Infarcts were created
in female mice by coronary artery ligation and cells delivered
to the region by direct needle injection. These selected cells
engrafted in the infarct region of mice and developed into a
robust myogenic tissue with highly developed vasculature.
The implanted cells were identified by the presence of both
the Y chromosome and expression of GFP. These authors
further characterized the expression of cardiac transcription
factors MEF2, GATA-4, CSX/NKX2.5 and gap junction
protein connexin 43 in the reconstituted region, identifying
cardiomyocytic differentiation. Injection of c-kitnegLinnegcells
failed to produce this myogenic tissue.
Over the past several years, we have been studying the
potential of human and rat MSCs to engraft in the heart and
to repopulate regions of the heart damaged by ischemia. We
have utilized human MSCs placed into the SCID/beige
mouse to test the ability of these cells to become
cardiomyocytes38. The approach was to inject the LacZ-
tagged hMSCs into the ventricle chamber, where only a
portion (~5%) are delivered to the myocardium by the
coronary circulation. The hMSCs that enter the vessels do
find their way out of the blood vessels and can be found
dispersed in the ventricle wall, surrounded by
cardiomyocytes. Due to the high metabolic demands of heart
tissue and the rhythmic beating, the engrafted hMSCs are
subjected to an environment that would be difficult to
establish in vitro. The engrafted hMSCs underwent
morphological differentiation to resemble the surrounding
cardiomyocytes38. The cells began to express muscle specific
proteins including desmin, ·-actinin, phospholamban,
cardiac troponin T and myosin heavy chain. Significantly, the
hMSCs expressed these proteins to levels that are
indistinguishable from surrounding cardiomyocytes. In other
experiments, we have created ischemic regions in the hearts
of athymic rats and implanted the human MSCs directly in
the infarct regions. We are able to detect the implanted cells
M. Pittenger et al.: Adult stem cells for connective tissues

Page 5

313
within the infarct region and see low levels of immunostaining
for sarcomeric proteins (Martin, Senechal and Pittenger,
unpublished). With colleagues at Johns Hopkins Hospital
Department of Cardiac Surgery, we moved to a porcine large
animal model of ischemic heart failure (Shake et al., 2002).
The size of the pig heart is similar to man and allows for in situ
physiological measurements through the use of sonocrystals
or echocardiogram. In animals analyzed as late as six months
after cell implantation, porcine MSCs implanted in the
infarct region appeared to prevent the pathological thinning
evident in animals which did not receive MSCs. These results
will be confirmed in additional animals and new animals
included that utilize allogeneic porcine MSCs.
Genetically modified MSCs for gene delivery in
muscle
Ex vivo gene modified cells are attractive for gene therapy
as they avoid the risks associated with the delivery of viral or
other potentially pathogenic vectors. To this end,
hematopoietic and other stem cells, cell lines, fibroblasts,
keratinocytes, and myoblasts have been studied extensively.
A major requirement for such cell-based therapy is the
sustained expression of the gene without adverse effect on
the recipient. The parameters to consider include availability
and expansion capability of the cells, efficiency of gene
modification, length of therapeutic gene expression in vivo,
immunogenicity of the cells, and safety of the cells, which
includes the absence of formation of tumors and ectopic
tissues. The route of delivery of the cells plays a significant
part and is based on whether systemic or local expression of
the gene product is required. For the systemic expression of
proteins, cells can be infused into the general or organ-
specific circulation; or cells can be encapsulated or attached
to various matrices and implanted subcutaneously. The
more simple method of delivery is to directly inject the cells
under the skin, intra-dermally or intra-muscularly.
Intra-muscular delivery is the optimal choice for myoblast-
mediated cell therapy because of the obvious advantage of
the myoblasts fusing with the host muscle cells to
differentiate into muscle fibers. In addition, the inherent
vascularity of the muscle makes it a useful depot to deliver
M. Pittenger et al.: Adult stem cells for connective tissues
Figure 2. Rat MSCs isolated from bone marrow of Fischer rats were transduced with moloney leukemia retrovirus expressing b-galactosidase-
IRES-NeoR. Transduced or non-transduced cells were labeled with CM-DiI or DAPI and injected into the thigh or calf muscle of Fischer
rats. After 3, 14 or 28 days, the animals were sacrificed and frozen sections of the muscle were prepared. The expression of ‚-Gal was
detected with histochemistry using X-Gal as a substrate. Frozen sections were also observed by fluorescence microscopy for red (CM-DiI)
or blue (DAPI) fluorescence. As seen in Fig. 2A and B numerous ‚-Gal- expressing cells are present in the muscle at day 3 (A) and day 14
(B). Fig. 2C shows Di I stained hMSCs at day 14 between muscle fibers away from the injection site. Fig. 2D shows DAPI-positive non-
transduced MSCs in the muscle at day 28. Magnification 20 X.