HEP (2007) 180:243–262
© Springer-Verlag Berlin Heidelberg 2007
N. Leviˇ car1(u) · I. Dimarakis1· C. Flores2· J. Tracey1· M. Y. Gordon2·
N. A. Habib1
1Department of Surgical Oncology and Technology, Faculty of Medicine, Imperial College
London, Hammersmith Hospital, Du Cane Road , London W12 0NN, UK
2Department of Haematology, Faculty of Medicine, Imperial College London,
Hammersmith Hospital, Du Cane Road, London W12 0NN, UK
Cell Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Haematopoietic Stem Cells . . . . . . . . . . . . . . . . . . . . . . . . . .
Injured Liver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1Generation of Hepatocytes by Haematopoietic Stem Cells . . . . .
3.2 Haematopoietic Stem Cell Therapy for Liver Diseases . . . . . . . .
3.2.1 Animal Studies . . . . . . . . . . . . . . . . . . . . . . .
3.2.2 Human Studies . . . . . . . . . . . . . . . . . . . . . . .
Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1 Haematopoietic Stem Cells forβ-Cell Generation . . . . . . . . . .
4.2 Haematopoietic Stem Cells for Diabetes . . . . . . . . . . . . . . .
4.2.1 Animal Studies . . . . . . . . . . . . . . . . . . . . . . .
4.2.2 Human Studies . . . . . . . . . . . . . . . . . . . . . . .
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract Advances in stem cell biology and the discovery of pluripotent stem cells have
made the prospect of cell therapy and tissue regeneration a clinical reality. Cell therapies
better than any pharmacological or mechanical device. There is an accumulating body of
origin, to liver and pancreatic islet cell regeneration. In this review, we will focus on the
cell therapy for the diseased liver and pancreas by adult haematopoietic stem cells, as well
as their possible contribution and application to tissue regeneration. Furthermore, recent
progress in the generation, culture and targeted differentiation of human haematopoietic
stem cells to hepatic and pancreatic lineages will be discussed. We will also explore the
possibility that stem cell technology may lead to the development of clinical modalities for
human liver disease and diabetes.
Keywords Stem cell therapy · Haematopoietic stem cells · Liver disease · Diabetes
244 N. Leviˇ car et al.
or enhance the biological function of damaged tissues or organs by utilising
cells and bioactive molecules to trigger, enhance, support and complement
the residual capacity for repair. This can be achieved by the transplantation of
cells, which are typically manipulated ex vivo, into a target organ in sufficient
numbers for them to survive long enough to restore the normal function of
organs and tissues. Possible candidate cells to be used include autologous
primary cells, cell lines, various stem cells including bone marrow stem cells,
cord blood stem cells and embryonic stem (ES) cells (Fodor 2003).
In recent years, advances in stem cell biology, including embryonic and
somatic stem cells, have made the prospect of tissue regeneration a potential
clinical reality, and several studies have shown the great promise that stem
cells hold for therapy (Assmus et al. 2002; Wollert et al. 2004). Despite the un-
questioned totipotency of ES cells, there are numerous unanswered biological
questions as to the regulation of their growth and differentiation. The safety
profile of unselected ES cells for transplantation early on demonstrated dys-
regulated cell growth resulting in teratoma formation (Reubinoff et al. 2000).
Moreover, the ethical and legal issues associated with ES cells have shifted
the focus to adult stem cells, and their regenerative potential has been under
The main role of adult stem cells, which are present in approximately
1%–2% of the total cell population within a specific tissue, is to replenish
of the tissue’s functional cells in appropriate proportions and numbers in re-
sponse to ‘wear and tear’ loss or direct organ damage (Fang et al. 2004). They
are vital in the maintenance of tissue homeostasis by continuously contribut-
ing to tissue regeneration and replacing cell lost during apoptosis or direct
injury (Li and Xie 2005). Adult stem and progenitor cells possess the capacity
of self-renewal and differentiation into one or more mature cell types. They
are able to maintain their populations within the human body through asym-
undifferentiated progeny (Preston et al. 2003). These properties make them
ideal candidates for stem cell-based therapies and tissue engineering.
Haematopoietic stem cells (HSC) are multipotent bone marrow cells that sus-
tain the formation of the blood and immune system throughout life. First
identifiedin 1961, HSC have beenby far the best characterisedand most stud-
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 245
ied example of adult stem cells (Till and McCulloch 1961). The bone marrow
stromal cells [called mesenchymal stem cells (MSC)] that have the ability to
2001). It was previously thought that adult stem cells were lineage restricted,
but recent studies demonstrated that bone marrow-derived progenitors in
addition to haematopoiesis also participate in regeneration of ischaemic my-
neurogenesis (Mezey et al. 2000).
into osteoblasts, adipocytes and chondrocytes (Pittenger et al. 1999).
Liver diseases impose a heavy burden on society and affect approximately
17% of the population. Cirrhosis, the end result of long-term liver damage,
has long been an important cause of death in UK and showed large rises
in death rates over the past 20 years (Ministry of Health 2001). The main
causes of cirrhosis globally are hepatitis B and C and alcohol abuse. Changing
patterns of alcohol consumption and the increasing incidence of obesity and
non-alcohol steatohepatitis (NASH) will continue to increase (Fallowfield and
Cirrhosis is a progressive liver disease and is marked by the gradual de-
struction of liver tissue over time. Persistent injuries lead to hepatic scarring
End-stage liver fibrosis is cirrhosis, whereby normal liver architecture is dis-
ruptedbyfibroticbands,parenchymal nodulesand vasculardistortion. Portal
hypertension and hepatocyte dysfunction are the end results and give rise to
plantation (OLT). However, the increasing shortage of donor organs restricts
liver transplantation. With the widening donor-recipient gap, the increasing
incidence of liver disease, life-long dependence on immunosuppression and
the poor outcome in patients not supported by liver transplantation, there is
obviously a demand for new strategies to supplement OLT.
of cells that can respond to liver injury and loss of hepatocytes (Sell 2001).
First, mature hepatocytes, which are numerous, respond to mild liver injury
by 1 to 2 cell cycles. Second, intra-organ ductal progenitor cells, which are
less numerous, respond by longer and limited proliferation. Third, stem cells
246 N. Leviˇ car et al.
entering from the circulation participate in liver regeneration. These cells, in
injury (Petersen et al. 1999). In this latter mode, responding to severe injury,
they enter first as an intermediate cell population, which then mature into
New strategies for generating a viable source of healthy hepatocytes for
treating hepatic disorders are under investigation. Potential sources include
the expansion of existing hepatocytes, ES cells, progenitor/stem cells in the
liver, and bone marrow stem cells. Direct transplantation of hepatocytes has
already been successfully tried in a small number of patients as a bridge to
OLT or as a therapeuticalternative but there remain several limitations,which
mainly include limited hepatocyte amplification, the fact that replacement of
success fuelled the interest in haematopoietic stem cells for hepatic disorders.
In vitro studies have demonstrated the potential of bone marrow stem cells
to differentiate towards the hepatic lineage. Oh et al. (2000) and Miyazaki
et al. (2002) have shown that rat bone marrow contains a subpopulation (3%)
of cells co-expressing haematopoietic stem cell markers (CD34, c-kit, Thy-1),
α-fetoprotein (AFP) and c-met. They have also demonstrated the expression
of albumin, a marker of terminally differentiated hepatocytes, after culturing
crude bone marrow in the presence of hepatocyte growth factor (HGF) and
epidermal growth factor (EGF). Similar observations were made by Okumoto
et al. (2003), where rat bone marrow cells enriched for Sca-1 began expressing
that a subpopulation of murine mononuclear bone marrow cells isolated by
chemotaxis in response to theα-chemokine stromal-derived factor-1 (SDF-1)
expressed messenger RNA (mRNA) for AFP and a population enriched for
Sca-1 expressed mRNA for AFP, c-met and CK19. Purified murine HSC were
HNF1α) and mature hepatocyte markers (CK18, albumin, transferrin) when
are responsible for conversion (Jang et al. 2004).
into liver-like cells. When cultured on collagen matrix and in the presence of
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 247
Furthermore, when selected by SDF-1 chemotaxis they appear to be multi-
potent and express AFP (Kucia et al. 2004). Numerous cytokines and growth
commonly used in the majority of studies and have been shown to promote
hepatic differentiation in vitro (Block et al. 1996; Michalopoulos et al. 2003).
genitor cells (MAPC) has the potential to differentiate into hepatocyte-like
cells in the presence of HGF and fibroblast growth factor-4 (FGF-4) (Schwartz
et al. 2002). However, the MAPC culture is fastidious, with a substantial delay
between the isolation and the appearance of hepatocyte-like cells, which calls
into question their use in the clinic. In vitro hepatic differentiation of MSC in
the presence of HGF and oncostatin M was confirmed by another group (Lee
et al. 2004). Again, the differentiation process was lengthy but they demon-
strated the expression of liver-specific genes in differentiated cells and other
characteristics of liver cells, including albumin production, glycogen storage,
cytochrome P450 activity.
Liver has long been known to exhibit considerable regenerative potential,
but it has been only recently that we began to understand the implication of
stem/progenitor cells in this process. The extrahepatic stem cells such as HSC
are of particular interest since they are easily accessible.
Petersen et al. (1999) were first to show that bone marrow stem cells could
be a potential source of hepatic oval cells. The liver injury was induced with
hepatic cells were shown to be of bone marrow origin, demonstrated by using
markers for the Y chromosome, dipeptidyl peptidase IV enzyme, and L21-6
antigen to identify donor-derived cells.
The other demonstration of hepatocyte regeneration from bone marrow
et al. 1995), a lethal hereditary liver disease (Lagasse et al. 2000). Intravenous
and restored the biochemical function of its liver. The liver repopulation by
bone marrow cells was slow, although significant. The first hepatocytes of
248 N. Leviˇ car et al.
bone marrow origin appeared 7 weeks after transplantation. However, later
on, at 22 weeks, one-third of the liver comprised bone marrow-derived cells.
This suggested that bone marrow stem cells contribute to hepatocyte genera-
tion in the presence of injury, where the regenerative potential of hepatocytes
is impaired. Furthermore, recent data have demonstrated that bone marrow
stem cells injected during liver injury can reduce the resulting liver fibrosis
(Sakaida et al. 2004). The exact mechanism of this therapeutic effect is not
fully understood yet, but it may be facilitated by the matrix metalloproteases,
which enable degradation of hepatic scars and are expressed by bone marrow
stem cells. More studies have shown that HSC engraft, repopulate and have
survival advantage when transplanted into injured liver. Mallet et al. (2002)
used JO2antibody, the murine anti-Fas agonist, to induce hepatic apopto-
sis. Unfractionated bone marrow cells expressing Bcl-2 under the control of
a liver-specific promoter were transplanted into normal mice. Some mice re-
ceived repeated weekly injections of JO2antibody to induce liver injury. Bone
mice, which received JO2antibody injections. Moreover, in mice with induced
liver cirrhosis, 25% of the recipient liver was repopulated in 4 weeks by bone
murine HSC converted into viable hepatocytes with increasing liver injury.
They noticed that liver function was restored 2–7 days after transplantation of
that HSC contribute to the regeneration of injured liver by differentiating into
functional hepatocytes. Misawa et al. (2006) used desialylated bone marrow
cells in order to increase their accumulation in rat livers in a rodent model
of human hepatic Wilson’s disease. They demonstrated that direct accumula-
tion of desialylated bone marrow cells into liver increased the proportion of
shown that bone marrow stem cells can repopulate liver even in the absence
of liver injury. In one study, Theise et al. (2000a) transplanted unfractionated
male bone marrow or CD34+lin−cells into irradiated female mice and looked
for bone marrow-derived hepatocytes. They identified up to 2.2% Y-positive
hepatocytes at 7 days and 2 months or longer post-transplantation. More-
over, positive fluorescent in situ hybridization (FISH) for the Y-chromosome
and albumin mRNA confirmed male-derived cells were mature hepatocytes,
suggesting that hepatocytes can be derived from bone marrow cells in the
absence of severe acute injury. Krause et al. (2001) injected single male HSC
into irradiated mice and obtained engraftment in several organs, including
of the findings, Wang et al. (2003) found albumin-expressing hepatocyte-like
cells in the livers of mice transplanted with highly purified HSC.
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 249
leukocytes of irradiated mice, but did not contribute to non-haematopoietic
tissues, including liver, brain, kidney, gut, and muscle. Kanazawa and Verma
[TgN(Alb1HBV)]—they concluded there was little or no contribution of bone
concludedthat bonemarrow cell infusion wasnot abletoenhance liver regen-
is only a small contribution of bone marrow stem cell to liver regeneration
after chronic liver injury. Vig et al. (2006)have foundonly 4.7%bonemarrow-
derived hepatocytes at 3 months after transplantation and even less (1.6%) at
6 months after bone marrow transplantation.
Although views concerning the contribution of HSC to hepatocyte lineages
in vivo still remain divided, the differences between the studies may in part
reflect the types of cells used, different injury models used and the method
used to detect engrafted stem cells.
Several studieshave also shownthe presence of cellsof bone marrow origin in
derived from the stem cells originated in bone marrow. They examined livers
a male donor and have found Y-chromosome and CK8-positive hepatocytes,
suggesting that extrahepatic stem cells can colonise the liver. Similarly, Theise
et al. (2000b) investigated archival autopsy and biopsy liver specimens from
recipients of sex-mismatched therapeutic bone marrow transplantation and
orthotopic liver transplantations. They identified hepatocytes and cholangio-
and CK19 and FISH analysis for the Y-chromosome. Using double staining
analysis, they found a large number of engrafted hepatocytes (4%–43%) and
cholangiocytes (4%–38%), showing that they can be derived and differenti-
ated from bone marrow and replenish hepatic parenchymal cells. Although
Korbling et al. (2002) confirmed bone marrow-derived hepatocytes in liver
showed that most of the recipient-derived cells in the liver allografts were
macrophages/Kupffer cells and only a small proportion of hepatocytes (1.6%)
was recipient derived. Two other studies of liver transplant patients did not
The difference and inconsistent results of the published studies could be
due to use of different techniques to identify recipient-derived hepatocytes in
250 N. Leviˇ car et al.
transplanted patients. Also, various markers can be used for hepatocyte iden-
tification, and the accuracy of the methods used for identification is variable.
hepatocytes, the use of bone marrow stem cells as therapeutic agents is still in
use of bone marrow stem cells for liver insufficiency. The first clinical study
CD133+cells subsequent to portal vein embolisation of right liver segments.
Computerised tomography scan volumetry showed a 2.5-fold increase in the
growth rate of the left lateral segments compared to the control group of 3 pa-
stem cell administration. Despite the small number of patients and the lack of
an adequately sized randomised control group, these data suggested that cell
therapyenhancesandacceleratesliver regeneration. Thesecondclinicalstudy
out by our group (Gordon et al. 2006). We have performed a phase I clinical
trial in which five patients with liver insufficiency were given granulocyte
the portal vein or hepatic artery. Of the 5 patients, 3 showed improvement in
serum bilirubin and 4 in serum albumin. Clinically, the procedure was well
tolerated with no observed procedure-related complications. The data clearly
suggested the contribution of stem cells to the regeneration of liver damage
and are encouraging for the future development of stem cell therapy for liver
diseases. A more recent study confirmed the observations from the previous
twostudies.Terai etal.(2006)treatedninelivercirrhosispatients withautolo-
gous bonemarrow. Theyinfused5.20+/−0.63×109mononuclear cells(CD34+,
CD45+, c-kit+) via the peripheral vein and followed the patients for 24 weeks.
Child-Pugh scores were observed at 24 weeks after therapy, suggesting that
bone marrow cell therapy should be considered as a novel treatment for liver
Type 1 diabetes accounts for only 5%–10% of all diabetes cases worldwide
but its incidence is increasing with current estimates of those affected at
approximately 105in the UK and 106in the USA (Burns et al. 2004; Dane-
man 2006). Type 1 (insulin-dependent) diabetes is a chronic disease affecting
genetically predisposed individuals in which insulin-secreting β-cells within
pancreatic islets of Langerhans are selectively and irreversibly destroyed by
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 251
all affected—to a state of absolute insulin deficiency (Devendra et al. 2004).
Type 2 diabetes was once considered a disease of wealthy nations but now it is
a global affliction. The incidence of type 2 (adult onset or insulin resistant) di-
abetes is increasing globally and has reached pandemic proportions (Petersen
and Shulman 2006). Estimates by the International Diabetes Federation (IDF)
anticipate that the worldwide incidence of diabetes among those 20–79 years
old will increase by around 70% in the next 20 years (International Diabetes
Federation 2006). Although the aetiology of type 2 diabetes remains obscure,
obesity and a sedentary lifestyle are the most common epidemiologic factors
associated with development of the disease.
The common denominator in both types 1 and 2 diabetes is a decrease in
β-cell mass resulting in an absolute or relative state of insulin insufficiency,
respectively. Long-term normalisation of glucose metabolism is a prerequisite
for prevention of secondary complications and to date has only been achieved
with transplantation of the whole organ or with a reasonable number of islets.
Recent success with the Edmonton protocol in 2000 (hepatic portal vein infu-
However, the chief limitation to transplantation, either whole gland or islets,
is the paucity of donors. Current protocols for islet transplantation require
up to 1×106primary human islets per recipient the approximate equivalent
of 2–4×106β-cells (Burns et al. 2004). Thus, such treatment can be offered
to only an estimated 0.5% of needy recipients (Lechner and Habener 2003).
Additionally, differentiatedβ-cells cannot be expanded efficiently in vitro and
senesce rapidly (Halvorsen et al. 2000).
ing cells in vitro either by genetic engineering ofβ-cells or by utilising various
and to differentiate into insulin producing cells and ultimately into β-cells
(Hess et al. 2003). ES cell usage has been the subject of both ethical and scien-
tific debate (Frankel 2000). The demonstration of plasticity of haematopoietic
stem cells has been published and has led to their study in the treatment of
diabetes (Hess et al. 2003). Insulin secreting cells (β-cells or otherwise) could
be transplanted into patients to help maintain blood glucose homeostasis, re-
duce the burden of diabetes-related complications, overcome the limitation of
donor organs and provide benefit to millions of diabetics.
an ideal cure. Studies of the growth, development and differentiation of pan-
252 N. Leviˇ car et al.
lular and molecular mechanisms ofβ-cell differentiation, but recent work has
begun to pose speculative hypotheses on pancreatic cell development.
Several investigators have isolated multipotent bone marrow-derived cell
expressing a pancreaticβ-cell phenotype under differentiation culture condi-
tions. Such populations include marrow-isolated adult multilineage-inducible
cells (Lee and Stoffel 2003; Tang et al. 2004), as well as a subpopulation of pe-
ripheral blood cells of monocytic origin which have been induced to express
characteristics of pancreatic cells (Ruhnke et al. 2005). In addition, a sub-
population of multipotent human mesenchymal stem cells (hMSC) has been
identified which constitutively expresses pancreatic islet transcription factors
and phenotypically resemble pancreatic isletβ-cells.
MIAMI cells, derived from whole bone marrow after 14 days in culture, are
described as a homogeneous population expressing a unique set of markers.
Interestingly, these cells are negative for the haematopoietic stem cell marker
markers for cells of mesodermal, ectodermal and endodermal lineages. With
of undergoing neural, osteogenic, chondrogenic, adipogenic and endodermal
differentiation. Topromotetheexpressionofpancreatic isletgenes,D’Ippolito
et al. (2004) treated MIAMI cells under conditions previously known to pro-
lowed by exposure to butylated hydroxyanisole (BHA) and exendin-4. Nicoti-
namide, exendin-4 and activin-A were then added as factors known to induce
β-cell transcription factors Beta2/NeuroD, Nkx6.1 and Isl1 in differentiation-
cells may have the potential for pancreatic β-cell differentiation and potential
use in vivo.
As with MIAMI cells, manipulation of BMDS cell culture medium with
high glucose, followed by treatments with nicotinamide and exendin, has
been shown to induce cells into expressing features of pancreatic islet cells
(Tang et al. 2004). Genetic analysis showed the upregulation of pancreatic
β-cell genes such as insulin I and II, Glut-2, nestin, Pdx-1, glucose kinase
and Pax-6 in differentiation-treated cells. Immunocytochemical analysis also
showedinsulinand c-peptidesynthesisin upto20%of differentiation-treated
cells. Induced cells also responded to glucose challenge and secreted insulin
according to glucose concentration. These data imply that BMDS cells are
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 253
phenotypic expression, as well as insulin-secreting function.
Monocytes derived from peripheral blood, known as programmable cells
of monocytic origin, have also been shown to express characteristics of pan-
EGF, HGF and nicotinamide (Ruhnke et al. 2005). After exposure to islet cell-
conditioned medium, these cells expressed transcription factors involved in
monocytic cells which were differentiated into pancreatic neo-islet cells were
responsive to glucose challenge, as determined by insulin release pre- and
post-differentiation induction. Programmable cells of monocytic origin were
found to be genetically and functionally similar to pancreatic β-cells in vitro.
However, invivoexperimentshaveyettovalidatecellengraftment andinsulin
production in a diabetic model.
Moriscot et al. (2005) have identified plastic adherent hMSCs which have
an Oct-4+pluripotent phenotype and which constitutively express Nkx6.1, an
conditionedmedium or co-culturedwith human islets. Variousculturecondi-
co-culture with human islets led to insulin secretion as well as upregulation
of NeuroD and insulin I. Although the described combination of adenoviral
infection and co-culture led to insulin secretion, not all transcription factors
hMSCs have not terminally differentiated into pancreatic cells, or that culture
conditions need to be further optimised for properβ-cell differentiation. hM-
SCs have the capacity of at least partial differentiation into β-cells and may
be a potential target for therapeutic purposes, as they are easily isolated from
In vitro work proves that adult BMDS cells are capable of differentiation
opment. Several investigators have been able to induce bone marrow-derived
cells into cells which phenotypically express pancreatic β-cell characteristics
and respond to glucose challenge. However, cells manufactured in vitro have
not yet demonstrated insulin secretion to the same degree as endogenous β-
cells. Insulin-secreting cells generated in vitro have been found to secrete only
about 1% of the level insulin produced by endogenous β-cells (Bonner-Weir
and Weir 2005). This information supports the notion that in vitro cell cul-
ture may not adequately mimic the physiological microenvironment where
endogenous cells develop and regenerate in the body. Importantly, assessing
of the follow-up to ex vivo studies.
254 N. Leviˇ car et al.
(2003). Irradiated female wild-type mice were injected with BMDS cells ex-
pressing enhanced green fluorescent protein (EGFP) via a Cre-LoxP system
controlled by the active transcription of the murine insulin gene. All donors
were male mice sharing the same C57BL/6 background with recipient ani-
mals. At 4–6 weeks post-transplantation, pancreatic islet tissue was harvested
and EGFP-positive cells were isolated by means of fluorescence-activated cell-
analysis. Up to 3% of total cells per islet were found to express EGFP in trans-
planted animals with no evident expression observed in peripheral blood and
bone marrow samples. RT-PCR analysis of sorted cells revealed besides in-
sulin I and insulin II, the expression of otherβ-cell markers including GLUT2,
IPF-1, HNF1α, HNF1β, HNF3β and Pax-6. At the same time, cells lacked the
common haematopoietic/leukocyte marker CD45. Finally, demonstration of
glucose-dependent and incretin-enhanced insulin secretion was reported as
proof of functionality.
Using the same experimental strategy as Ianus et al. (2003), another group
was unable to confirm their findings (Taneera et al. 2006). Whole BMDS cells
under the control of the murine insulin promoter failed to demonstrate any
signs of GFP expression within the pancreatic parenchyma of transplanted
series of experiments was conducted using BMDS cells expressing GFP under
thecontrolβ-actin promoter. Despitethelarge degreeof engraftment, none of
the GFP-positive cells co-expressed insulin or the β-cell transcription factors
Pdx-1 or Nkx6.1, while more than 99.9% expressed the pan-haematopoietic
BMDS cells demonstrated efficient pancreatic engraftment, a haematopoietic
cell fate was almost exclusively adopted.
tus, Hess et al. (2003) utilised a murine model of streptozotocin-induced pan-
intravenously with BMDS cells from GFP transgenic donor mice. Irradiation
status and c-kit expression served as intergroup comparison criteria. One of
the most interesting observations made was that pancreatic injury was a pre-
requisite for BMDS cells taking on an insulin-producing phenotype within
the pancreatic parenchyma, an observation also made by other researchers
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 255
(Mathews et al. 2004). No improvement was witnessed in the c-kit−trans-
planted subgroup in comparison to the c-kit+group or the subgroup using
whole bone marrow, which points to the c-kit+stem cell population as the
potent subpopulation. The lack of any clinical improvement in the subgroup
transplanted with irradiated whole bone marrow revealed the unlikelihood
of paracrine factors being responsible for any observed improvement in the
non-irradiated groups. Even though 2.5% of insulin-positive cells from islets
of streptozotocin-treated animals were positive for GFP, there was no expres-
sion of Pdx-1 in these cells. In addition to their low frequency, these cells were
noticed to be absent during the onset of hyperglycaemic reduction. Finally,
a large proportion of donor cells documented in ductal or islet regions were
of endothelial lineage, thus associating the regenerative process with various
endothelial interactions. Lechner et al. (2004) transplanted BMDS cells from
in one experiment and animals that had undergone partial pancreatectomy in
another. Both GFP immunostaining and Y chromosome FISH failed to detect
significant evidence of donor BMDS cell trans-differentiation intoβ-cells.
c-peptide and insulin (Tang et al. 2004). In vivo transplantation of these cells
into the left renal capsule and the distal tip of the spleen of streptozotocin-
mediated diabetic mice not only reversed the existing hyperglycaemia, but
also restored the animals’ ability to respond to in vivo glucose challenges.
of BMDS cells in order to restore normoglycaemia (Banerjee et al. 2005).
Banerjee et al. (2005) demonstrated improved glycaemic control in addition
intravenous injections of BMDS cells in streptozotocin-treated mice. The fact
that BMDS cells were harvested from experimental-diabetic mice is evidence
of the retained regenerative potential of the bone marrow.
It is quite striking that a number of groups have not been able to reproduce
the results of Ianus et al. (2003). Most authors noted a number of technical
considerations in an attempt to explain this. Many groups have reversed the
process of identifying pancreatic-specific markers in engrafted cells by try-
ing to identify cellular populations expressing such markers prior to any in
reported to express a vast number of cell markers and transcription factors
necessary for β-cell differentiation (Pessina et al. 2004). When transplanted
into non-obese diabetic mice with autoimmune type 1 diabetes, human um-
bilical cord blood mononuclear cells were able to reduce blood glucose levels
and improve survival compared to untreated animals (Ende et al. 2004). The
argument regarding the true nature and fate of candidate ‘insulin-producing’
cells remains unresolved (Sipione et al. 2004), as they may not be true β-cell
precursors and furthermore may be unsuitable for clinical transplantation.
256 N. Leviˇ car et al.
Pre-transplantation in vitro manipulation (Oh et al. 2004; Tang et al. 2004)
may also provide a useful tool for delivering large numbers of predefined
cells, thus avoiding complications such as suboptimal delivery rates and cell
differentiation down non-pancreatic pathways.
murine β-cells (Dor et al. 2004), thus casting a shadow of uncertainty over
scientists’ expectation of multipotent stem cells. The existence of unknown
molecular and cellular pathways is highly likely. Endothelial signals have been
linked with induction of pancreatic differentiation (Lammert et al. 2001), an
association that may underlie observations in some of the above studies(Choi
et al. 2003; Hess et al. 2003; Mathews et al. 2004). Many studies have failed
to demonstrate expression of insulin or Pdx-1 in donor bone marrow-derived
cells seen in peri-ductal or peri-islet locations (Choi et al. 2003; Hess et al.
2003; Mathews et al. 2004; Taneera et al. 2006), whilst many of these cells
express endothelial markers (Choi et al. 2003; Hess et al. 2003; Mathews et al.
2004). Further to identifying actual potent subpopulations (Hess et al. 2003),
a few technical considerations may come in useful. A dose-dependent effect
has been noted (Ende et al. 2004), suggesting possible population expansion
to be required prior to clinical transplantation. Finally, the demonstration of
multi-step cell delivery being more efficient long-term (Banerjee et al. 2005)
also underlines the possibility of introducing this principle in experimental
and possibly clinical work.
The first reported data on cellular therapy for type 1 and 2 diabetes were
presented by Fernandez Vina et al. (2006a, b). They have treated 23 patients
with type 1 diabetes with CD34+CD38−cells isolated from bone marrow and
sugar decreased by 9.7% and c-peptide significantly increased by 55%. It was
by 17% suggesting that autologous bone marrow stem cells could improve
pancreatic function. Similar results were obtained for type 2 diabetes patients
where autologous CD34+CD38−cells were transplanted via spleen artery into
16 patients. The blood sugar significantly decreased by 27% 90-days post-
transplantation, while c-peptide and insulin increased by 26% and 19%,re-
spectively. Even more impressive is the fact that 90 days post-transplantation,
84% of treated patients did not need any more anti-diabetic drugs or insulin.
Stem Cells as a Treatment for Chronic Liver Disease and Diabetes 257
Over the past 5 years attention has been focussed sharply on stem cells and
the extraordinary potential that they offer in treating a number of currently
intractable human diseases. However, for liver diseases and diabetes, the stem
can be applied to its fullest potential in the clinic. In order to achieve the goal
of cell therapy, the source and types of cells, generation of cells in sufficient
numbers, maintenance of the differentiated phenotype and cell engraftment
cells should expand extensively in vitro, have minimal immunogenicity and
be able to reconstitute tissue when transplanted in damaged tissue. Defining
which patient groups are suitable for this therapy and which stem cell types
are the most effective given the underlying pathology is also important. The
optimum timing and method of delivery need to be determined as they may
have a significant influence on the outcome of cell transplantation. Long-term
side-effects of treatment are unknown as most of the clinical studies are very
recent. In addition, there is growing evidence that transplanted cells, being
multipotent, do not simply replace missing tissue but also trigger local mech-
anisms to initiate a repair response. Paracrine effects and immune regulation
of the transplanted cells may also play a role in functional restoration of the
tissue (Pluchino et al. 2003).
As more scientific knowledge is gained in this field, hopefully some of the
technical concerns will be answered and we will soon see stem cell therapy in
more clinical applications.
Alison MR, Poulsom R, Jeffery R, Dhillon AP, Quaglia A, Jacob J, Novelli M, Prentice G,
Williamson J, Wright NA (2000) Hepatocytes from non-hepatic adult stem cells. Nature
am Esch JS 2nd, Knoefel WT, Klein M, Ghodsizad A, Fuerst G, Poll LW, Piechaczek C, Bur-
chardt ER, Feifel N, Stoldt V, Stockschlader M, Stoecklein N, Tustas RY, Eisenberger CF,
marrow cells to the liver: a novel concept to support hepatic regeneration. Stem Cells
Urbich C, Martin H, Hoelzer D, Dimmeler S, Zeiher AM (2002) Transplantation of pro-
genitorcells andregenerationenhancement inacute myocardial infarction(TOPCARE-
AMI). Circulation 106:3009–3017
Banerjee M, Kumar A, Bhonde RR (2005) Reversal of experimental diabetes by multiple
bone marrow transplantation. Biochem Biophys Res Commun 328:318–325