Amnion-Derived Pluripotent/Multipotent Stem Cells
Toshio Miki and Stephen C.Strom*
Department of Pathology,University of Pittsburgh,200 Lothrop,St.Pittsburgh,PA 15213
Amniotic epithelium is derived from the epiblast by approx 8 d after fertilization. Other
parts of the placenta are derived from extraembryonic tissue. In addition to this devel-
opmental difference, amniotic epithelial (AE) cells are known to have unique charac-
teristics, such as low level expression of major histocompatibility complex antigens, and
a less restricted differentiation potential. The differentiation of the AE cells to the neu-
ral lineage is well documented. Recently, we reported that AE cells from term placenta
express several stem cell surface markers that are commonly found on pluripotent stem
cells such as embryonic stem cells, and that in culture, AE cells differentiate into cell
types from all three germ layers. In this review, we describe the unique characteristics
of the AE stem cells and summarize previous work concerning the stem cell nature of
cells from amnion.
Index Entries: Amnion; pluripotent; stem cell; Oct-4; placenta.
Stem Cell Reviews
Copyright © 2006 Humana Press Inc.
All rights of any nature whatsoever are reserved.
ISSN 1550–8943/06/2:133–142/$30.00 (Online) 1558–6804
embryonic stem (ES) cells such as SSEA-3,
SSEA-4, TRA 1-60, and TRA 1-81. In addi-
tion, AE cells express molecular markers of
pluripotency, Oct-4 and Nanog, and differ-
entiation into all here germ layers was
demonstrated in vitro. These data suggest
that, like ES cells, AE cells may be a useful
stem cell for cell transplantation and regen-
Placenta is discarded after a live birth.
Therefore, the use of placenta as a stem cell
source would obviate most, if not all, ethi-
cal, religious, or political concerns. If it were
definitively demonstrated that each of us
arrived at birth with “spare parts” that could
be derived from placenta, these stem cells
would be a unique recycling biotechnology.
In theory, placental stem cells could be
isolated from every newborn baby. In the
UnitedStates, more than 4 million live births
occur per year (e.g., 4,115,590 births in 2004).
Once technology is established to develop
pluripotent stem cell lines from each pla-
centa, it would be easy to complete banks
with all major histocompatibility complex
(MHC) immunotypes for further clinical
There is an increasing evidence that
human placenta contains pluripotent and/or
multipotent stem cells. Specific types of stem
cells, such as trophoblastic, hematopoietic,
and mesenchymal stem cells, will be dis-
cussed in other articles in this issue. Here,
we will review stem cells derived from
amnion of human placenta, specifically
amniotic epithelial (AE) cells. Unlike other
parts of placenta, amniotic epithelium is
derived from pluripotent epiblast at day 8,
far before gastrulation (days 15–1 7), which
is a “tipping point” at which cell fate is spec-
ified. The other components of placenta such
as chorion differentiate from extraembry-
onic trophectorderm. Because the amnion
differentiates from the epiblast at a time
when it retains pluripotency, it is reasonable
to speculate that AE cells may have escaped
the specification that accompanies gastru-
lation and that AE cells may preserve some
or all of the characteristics of the epiblast
such as pluripotency. Previously, we
reported that some AE cells express cell sur-
face antigens characteristic of pluripotent
*Correspondence and reprint
Stephen C. Strom
Department of Pathology,
University of Pittsburgh,
200 Lothrop, St. Pittsburgh,
134 ______________________________________________________________________________________________ Miki and Strom
Stem Cell Reviews ♦ Volume 2, 2006
applications. These potential advantages lead us to investigate
the possible stem cell nature of AE cells.
Development of the Amnion
About the 8 d after fertilization, the human blastocyst is
partially embedded in the endometrial stroma. The outer cells
(trophoblast) of the blastocyst differentiate into two layers and
migrate into the stroma. The inner cell mass or embryoblast
also differentiates into two layers, the hypoblast and epiblast.
Epiblast is the source of all three germ layers and eventually
forms the developing embryo. At the same time, a small cavity
(amniotic cavity) appears within the epiblast. Epiblast cells
adjacent to the cytotrophoblast are called amnioblasts (Fig. 1).
Amnioblasts bordering the amniotic cavity differentiate to
amniotic epithelia and the cavity enlarges.
Concomitantly, a layer of extraembryonic mesenchyme
around the AE cell layer develops in to the amniotic connec-
tive tissue. At 31–35 d of gestation, the amnion is made up of
three layers. The innermost layer toward the amniotic cavity
is the amniotic epithelium. Outside of this is a layer of con-
nective tissue cells. The connective tissue consists of very fine
argyrophil fibers at early stages, which are replaced with sev-
eral types of collagens produced by fibroblasts. The outermost
layer separating the amnion from embryonic coelom consists
of a solid, single layer of strongly flattened cells with marked
cell borders. This mesothelium-like coating disappears when
the amnion and chorion start to unite (at the end of 3 mo). At
full term, the placenta is a discoid with a diameter of 15–20 cm,
is approx 3 cm thick, and weighs about 500–600 g. The fetal
surface of the placenta is covered entirely by the amnion, which
has a surface area is approx 700–1200 cm2.
It is a common misconception that the amnion is derived
from the extraembryonic endoderm or extraembryonic tro-
phectoderm such as the yolk sac or chorion. However, the
amnion is derived from the epiblasts, the pluripotent cells,
which eventually give rise to all of the cell types of the embryo.
Isolation and Characterization
of Amniotic Epithelial Cells
Amnion is easily stripped from the underlying chorion and
deciduas at the coarse intermediate layer of connective tissue
(Fig. 2; arrow). The thin, nearly transparent amnion membrane,
which contains AE cells and mesenchymal fibroblasts (AMF),
is obtained after several washing steps to remove blood. For
isolation of AE cells, amniotic membrane is trypsinized, which
releases the AE cells from the supporting connective tissue and
the AMF. In our laboratory, membrane is incubated with 0.05%
trypsin with 0.53 mM EDTA for 10 min to dissociate debris
from the surface of the membrane, followed by two sequen-
tial 40-min trypsinizations. The “after” photo (Fig. 2) shows
that most of the epithelial cells are removed from the amniotic
membrane by trypsinization and that the connective tissue
(AMF) layer remains intact. With this simple protocol, 200–300
million AE cells can be isolated from one placenta.
The number of AE cells increases in the course of develop-
ment (1). The rapid surface growth of the amniotic epithelium
is caused by mitotic division. The mitosis rate peaks at around
day 20 and decreases thereafter. By 180 d of development, mito-
sis is rarely observed (2). However, isolated human AE cells
showed robust proliferation in culture with numerous mitotic
events (3). Despite a lack of robust growth in vivo, these data
indicate that, some or all AE cells preserve their proliferation
potential. The proliferation of AE cells in vivo may be repressed
during late gestation in response to interleukin (IL)-6 super-
family ligands (3).
In the presence of epidermal growth factor (EGF) or trans-
forming growth factor-α, AE cells proliferate robustly and form
a confluent monolayer of cobblestone-shaped epithelial cells.
Fig.1.Illustration of embryo at day 8.At 8 d after fertilization,inner cell mass differentiates into epiblast and hypoblast.An amniotic cavity appears
in the middle of the epiblast.The epiblasts adjacent to the cavity,termed amnioblasts,eventually form the amniotic epithelial layer.
Amnion-Derived Pluripotent/Multipotent Stem Cells _______________________________________________________________135
Stem Cell Reviews ♦ Volume 2, 2006
Proliferation ceases in the absence of EGF and giant palm-
shaped multinucleated cells form, which are reminiscent of
the description of trophoblast differentiation of ES cells (4). It
is known that the amount of EGF in amniotic fluid increases
during preganancy. It has been suggested that a balance of EGF
and IL-6 superfamily ligand signaling regulate AE cell prolif-
eration in vivo.
Characterization of Amniotic Epithelial Cells
Cytokeratins and vimentin are used as markers to identify
cells of different lineages. Vimentin is normally expressed by
mesenchymal cells such as fibroblasts or myocytes as well as
vascular endothelial cells. Recent reports suggest that vimentin
may be a marker of stem/progenitor cells, especially neural
stem cells (5,6).
In monolayer cultures of AE cells, virtually 100% of the cells
react with antibodies to low-molecular weight cytokeratins, con-
firming their epithelial nature. Interestingly, although initially
vimentin-negative, AE cells become vimentin-positive during
cell culture. Vimentin-positive AE cells remain positive for cytok-
eratins. These data indicate that in vitro culture or environmental
influences may induce dedifferentiation of AE cells.
Cell Surface Antigens
In culture, AE cells proliferate to form a population of epithe-
lial cells with uniform morphology. However, even morpho-
logically homogenous monolayer cultures may contain cell
types with different growth or differentiation potential.
Following isolation, AE cells express cell surface antigens
known to be expressed on other stem cells (Table 1). In addi-
tion to the typical stem cell markers such as SSEA-3, SSEA-4,
TRA 1-60, and TRA 1-81, AE cells share many cell surface
antigens with human ES cells. It has to be noted that intitially
isolated, naive AE cells are not homogenously positive for all
of these antibodies. Some surface markers such as CCR4- and
CD117-positive cells are very rare, while others such as CD9,
integrin α6, and integrin β1are expressed on virtually 100% of
the cells. These data suggest that naive AE cells are a het-
erogenous cell population with respect to cell surface profil-
ing. The profile of cultured AE cells is similar to that of ES cells
at a time where they start differentiating within an embryonic
body. Undifferentiated ES cells are not a completely homoge-
nous population (7). Afluorescence-activated cell sorting analy-
sis of a typical ES culture will show that up to 20% of the cells
are positive for SSEA1 and 50–80% of the cells will react with
antibodies to SEEA-3, SSEA-4, TRA1-60, or TRA1-81. As with
the AE cultures, even clonally expanded ES cells may contain
cells at different stages of differentiation and with a different
profile of surface markers. Clearly, more research will be needed
to determine if the culture of AE cells can be optimized to main-
tain their stem cell characteristics and a homogenous profile
of surface markers.
ABCG2 Expression and Side Population Cells
One of the cell surface antigens identified on AE cells is
ABCG2, a member of the ATP-binding cassette superfamily.
Expression of this protein by AE cells suggests that they might
have similar properties to the side population cells identified
in bone marrow. The ABCG2 gene product is a multidrug resist-
ance transport protein which effluxes, among other things, the
Hoechst dye, 33342, from cells (8). In 1996, Goodell et al. (9)
reported a method for isolating hematopoietic stem cells in
a single step based on the retention of this dye. A small
subpopulation of cells termed the side population (SP) cells
Fig.2.Histology of amnion.Normal human placenta sample as sectioned and stained with hematoxylin and eosin.In the left panel,the three major
components, amnion, chorion, and decidua, were shown.The arrow indicates the coarse intermediate layer of connective tissue.The right panel
shows photos of membrane “before” and “after” trypsinization. Only amniotic epithelial cells were dissociated by trypsinization.Although not
shown,subsequent collagenase digestion would release the amniotic mesenchymal cells from the connective tissue.
136 ______________________________________________________________________________________________ Miki and Strom
Stem Cell Reviews ♦ Volume 2, 2006
showed rapid efflux of the dye which was dependent on the
expression of ABCG2. In subsequent experiments, it was
determined that hematopoietic stem cells were contained
within the SP. Later, the SPphenotype was observed in a wide
variety of stem cells including neural stem cells (10) and
muscle-derived stem cells (11). Recently, SP cells were identi-
fied in both the AE and AMF fractions from amnion (12). The
significance of these observations is not known.
Molecular Markers of Stem Cells
Along with the stem cell-specific surface markers, AE cells
express molecular markers of stem cells. There is consensus
agreement that pluripotent human ES cell lines express Oct-4,
SOX-2, Lefty-A, fibroblast growth factor (FGF)-4, rex-1, and
TDGF-1 (cripto, 57; 13), and nanog (14). When examined by
reverse-transcription (RT)-PCR, all stem cell marker genes were
found to be expressed in the freshly isolated AE cells and/or
in cultured cells.
Among those molecular stem cell markers, Oct-4 is known
as one of the transcription factors that play a critical role in
maintaining pluripotency and self-renewal. Oct-4 belongs to
the Pit-Oct-Unc (POU) family of transcriptional regulators
(15–17) and regulates the pluripotency of human and mouse
ES cells (18). In the early mouse embryo, Oct-4 is only expressed
in pluripotent cell types such as cleavage stage blastomeres,
the inner cell mass of the blastocyst, and the epiblast of the egg
cylinder. At gastrulation, Oct-4 expression is downregulated
in somatic cells and was thought to be maintained mainly in
the primordial germ cells. During the postnatal period, Oct-4
expression is detected in growing oocytes and spermatogonia
(18). Pluripotent stem cells in culture, ES cells, embryonal car-
cinoma cells, and embryonic germ (EG) cells all express Oct-4.
This pattern of expression reflects the key role played by
Oct-4 in the maintenance of pluripotency (19). The expression
of Oct-4 is controlled epigenetically by hypermethylation of
the enhancer/promoter region (20).
Epiblasts, which redirect developmental origin of AE cells,
express Oct-4 as long as they remain undifferentiated (17). The
expression of Oct-4 in AE cells suggests that they maintain the
pluripotency of the undifferentiated epiblast. Oct-4 expression
is observed in most AE cells. Although some display nuclear-
localized Oct-4, in the majority of AE cells, the expression is
Summary of Cell Surface Marker Expression
Am SCell surface antigens Em S Neu S Mes S Hem S
Angiotensin converting enzyme
CCR4 (CC chemokine receptor)
CXCR4 (CXC chemokine receptor)
Frizzled-9 (Wnt ligand binding receptor)
VEGF R2 (FLK-1)
Am S, amnion-derived stem cells; Em S, embryonic stem cells; Neu S, neural stem cells; Mes S, mesenchymal stem cells; Hem S, hematopoi-
etic stem cells.
Stem cell and progenitor cell surface marker expression were summarized and compared with the expression of other stem cells. Marker
expression on AE cells was analyzed in our laboratory and data from other stem cells were obtained from published literature. Blank cells
Amnion-Derived Pluripotent/Multipotent Stem Cells _______________________________________________________________137
Stem Cell Reviews ♦ Volume 2, 2006
cytoplasmic. EG cells also show cytosol-positive Oct-4 stain-
ing, indicating that AE cells may be at a developmental stage
similar to that of EG cells.
Further investigation will be required in order to determine
whether the stem cell marker-positive cells are remnants of the
pluripotent cells from the fetus or if amniotic cells maintain
their stem cell nature for a separate, specific function which
has yet to be determined. If placental stem cells are maintained
throughout the pregnancy, the mechanism and the functional
implications of this will be the basis of future exploration.
Multipotent adult progenitor cells (MAPCs) were identi-
fied from adult bone marrow mesenchymal stem cells of
human, rat, and mouse. This rare cell population has pluripo-
tent differentiation ability in vitro and in vivo and the cells
express Oct-4. However, the expression of Oct-4 in MAPCs is
1000-fold lower than that of ES cells (21). In contrast, monkey
(Macaca fascicularis)AE cells expressed Oct-4 mRNAin almost
the same manner as ES cells (unpublished observations). These
data suggest that AE cells may be more like ES cells than adult
Pluripotency of Amniotic Epithelium-Derived
Awide variety of investigations performed since 1980 have
provided evidence of the existence of pluripotent stem cells in
amniotic fluid. The ultimate approach to determine the pluripo-
tency of amniotic epithelium-derived stem cells is to generate
chimeric animals by injecting the single stem cell into a blas-
tocyst. If the stem cell contributes all germ layer cells in the
chimeric animal, pluripotency will be confirmed. Of course, it
would be unethical to perform this experiment with human
blastocyst. Tamagawa et al. (22)established cell lines from cells
isolated from whole amnion, which contain AE cells and AMFs.
The cell line HAM-1 was mixed with mouse early embryonic
stem cells (EES-7) to form an aggregation chimera. The xeno-
geneic chimera embryo was maintained until all three pri-
mordial germ layers were formed. The contribution of
amnion-derived stem cells to all three germ layers of the xeno-
geneic aggregation chimera was demonstrated. Tamagawa’s
report is a remarkable demonstration of pluripotency in vitro.
When AE cells were cultured as an adherent monolayer for
several weeks, small clusters or “spheroids” or cell clusters
were noticed over the cobblestone pavement of epithelial cells.
These spheroids are similar in structure to embryoid bodies
(EB) described cultures of ES cells. Stem cell markers were
examined by immunohistochemistry in the spheroid struc-
tures from long-term (24 d) cultures. Immunohistofluorescent
staining revealed that the rim of cells on the outer edge of each
EB-like spheroid structure expressed stem cell-specific cell sur-
face antigens SSEA-3, SSEA-4, TRA 1-60, and TRA 1-81 (23).
Data presented in Fig. 3 shows surface staining with antibod-
ies to SSEA4 and cytoplasmic activity for alkaline phosphatase.
Expression of the cell surface stem cell markers was mainly
restricted to the spheroid-like structures, whereas more dif-
ferentiated cells in the epithelial monolayer surrounding the
EB-like structures did not react with these antibodies. The stem
cell molecular marker genes, Oct-4 and Nanog, are also pre-
dominantly expressed in the cells of the spheroids. These data
indicate that AE cells form EB-like structures over the mono-
layer of attached AE cells which better support and maintain
their stem cell nature in culture. The spheroid structures grow-
ing over the monolayer of other cells is reminiscent of ES cells
growing over a feeder layer of stromal cells. Miyamoto et al.
(24) reported a feeder layer function of AE cells. Primate ES
cells (Cynomolgus monkey CMk6 cell line) were cultured on
human AE cells, and the ES cells were maintained in an undif-
ferentiated state as was shown by Oct-4 expression and for-
mation of teratomas after injection into immunodeficient mice.
These data are completely consistent with our findings. Unique
cell-to-cell interaction between stem cells and a feeder-like
monolayer of AE cells may be useful to maintain the “stem-
ness” of AE stem cells.
Localization of Stem Cells in Human Placenta
Recently, we succeeded in the localization of the stem cell
marker-positive cells in a whole mount of the amnion mem-
brane (25). The stem cell marker-positive cells were uniformly
scattered over the membrane. Interestingly, the stem cells were
not found in clusters, but rather were surrounded by stem cell
marker negative cells. This observation stimulates our imag-
ination of the existence of a “stem cell niche.” Most tissue stem
cells have their own “niche,” i.e., a specific microenvironment
which helps to maintain the stem cells in an undifferentiated
state. Much like the feeder layer experiments described pre-
viously, the stem cell marker-negative AE cells may be play-
ing the role of a “niche” for the stem cell marker-positive cells.
The specific cell-to-cell contact and signaling pathways
involved in this process are currently under investigation.
Other explanations are also possible. During development,
when the amnioblasts first differentiate from the pluripotent
epiblast, the cells stray from the organized development of the
embryo. Without restrictions or other external regulatory sig-
nals, some of the cells of the amnion may randomly differen-
tiate, whereas others preserve their stem cell characteristics.
of Amniotic Epithelial Cells
We have previously shown the potential of AE cells to dif-
ferentiate into ectoderm, endoderm, and mesoderm lineage
cells in vitro (23). Although our report showed for the first time
that AE cells from one placenta can differentiate into all three
germ layer cells, it has also been demonstrated by other groups
that AE cells have the potential to differentiate into various
types of cells. Sakuragawa and coworkers, (26) the leading
investigators in this field, demonstrated that cultured AE cells
express markers of glial and neuronal progenitor cells. Further
experiments succeeded to induce the differentiation AE cells
into the cells that synthesize and release acetylcholine (27) and
catecholamines (28). Furthermore, dopamine-producing cells
were produced from AE cells (29) and transplanted into the
brain of animals with a model of Parkinson’s disease (30).
Tyrosine hydroxylase-positive cells were immunohistochem-
ically confirmed at the transplanted site of recipients and a
therapeutic effect was observed for about 1 mo.
Because the amnioblast is sometimes also referred to as
the amniotic ectoderm perhaps there is a propensity for the
AE cells to differentiate toward the neural lineage (ectoder-
mal). It is an interesting corollary that ES cells have the same
138 ______________________________________________________________________________________________ Miki and Strom
Stem Cell Reviews ♦ Volume 2, 2006
propensity to differentiate along a neural lineage. Our RT-PCR
data indicate that some of the neuronal progenitor marker
genes are expressed in the naive AE cells (23).
As with ectoderm differentiation, the potential for AE cells
to differentiate to endoderm was also reported. In 2000,
Sakuragawa et al. (31) showed albumin secretion from cul-
tured human AE cells and when LacZ-labeled human AE cells
were transplanted into the liver of an immunodeficient mouse,
the transplanted are cells were found to integrate into the
hepatic plate. Our group has also performed human AE cell
transplants into immunodeficient mice, and observed cells
with the morphology of hepatocytes, which expressed human
albumin or α-1 antitrypsin. Furthermore, we detected human
α-1 antitrypsin circulating in the serum of recipient mice, which
confirmed the engrafted human AE cells function as hepato-
cytes in mouse liver. Although these are encouraging prelim-
inary data, a more complete study of the fate of AE cells
following transplantation into the liver is needed and is cur-
rently underway. Nakajima et al. (32) established a method to
isolate rat amniotic cells. The transplanted rat AE cells sur-
vived in the liver following allogeneic transplantation for at
least 30 d. There are reports of the induction of early markers
of hepatic differentiation of AE cells following the addition of
specific growth factors to culture media. Takashima et al. (33)
used hepatocyte growth factor, FGF-2, heparin sodium salt,
and oncostain M to induce albumin-producing hepatocyte-
like cells. The same group also reported that they could induce
AE cells to produce insulin and could normalize blood glu-
cose in diabetic model animals following transplantation of
the AE cells in to the spleen (34).
There is no report of mesoderm lineage differentiation
except by our group (23). We applied culture conditions that
were used to induce cardiac differentiation of ES cells. Although
functional assays were not performed, the amntiotic epithelium-
derived cardiomyocytes expressed cardiac-specific genes and
immunostaining of α-actinin protein was demonstrated. The
α-actinin expression pattern was identical to that reported for
ES-derived cardiomyocytes (35).
The experiments summarized so far demonstrate that dif-
ferentiation of AE cells can be induced and somewhat directed
by exposure to exogenous growth factors or chemicals. The
plasticity of AE cells is further shown by growth factor-induced
differences in gene and protein expression and also by changes
in cell morphology. The effect of a number of growth factors
on AE cells is summarized in (Fig. 4).
Amniotic Mesenchymal Fibroblasts
At the hypoblast stage, a primary yolk sac is formed.
Extraembryonic mesoderm cells derived from the hypoblast
form a new cell layer beneath the amnioblast and parietal endo-
derm. The extraembryonic mesoderm cells eventually form
the amniotic mesenchymal layer (36). Therefore, the amnion
consists of epiblast-derived AE cells and hypoblast-derived
AMF. Mesenchymal stem cells can differentiate into multiple
mesenchymal lineages, such as adipocytes, osteoblasts, chon-
drocytes, and myocytes. Human MSCs are mainly isolated
from bone marrow, but many other sources are proposed. AMF
also express MSC cell surface markers and differentiation capa-
bility. Avariety of protocols are used to isolate AMFs from pla-
centa. Most reports use collagenase digestion. However, as
shown in Fig. 2, it is difficult to obtain a pure AMF population
without contamination of AE cells. Although trypsinization
releases only AE cells from the connective tissue, subsequent
collagenase digestion will release all remaining cells. Aside
from the reports on the mesenchymal stem cell characteristics
of AMFs, some reports suggest that AMF may be pluripotent,
with the capability to differentiate into cells beyond the mes-
enchymal lineage such as pancreatic cells, cardiomyocytes, or
neural cells. Wei et al. (34) induced pancreatic differentiation
of AE cells and AMF cells with similar culture conditions. The
results showed AMF-derived insulin-producing cells achieved
one-third the efficacy of AE cell-derived insulin-producing
cells upon transplantation. Because the starting material was
a mix of AE cells and AMF, these data may be explained by
the presence of pluripotent AE cell contamination in the AMF.
Senescent Nature of Amniotic Epithelial Cells
Continuous growth of human AE cells in culture leads to
replicative senescence. Under the culture conditions used to
date, after 6–10 passages, AE cells fall into a terminally non-
dividing state. If AE cells are cultured at very low densities,
senescence occurs even earlier. This discrepancy indicates that
the senescence mechanism is not simply controlled by telo-
mere length. Integrin-dependent epidermal growth factor
receptor (EGFR) activation is one of the signaling mechanisms
involved in cell proliferation (37). AE cells express a number
Fig. 3. Characterization of spheroid.Amniotic epithelial (AE) cells form spheroid-like structure on the monolayer AE cells.The cells of the spheroid
express stem cell marker cell surface antigens,such as SSEA-4.Monolayer AE cells were visualized with F-actin counterstaining (rhodamine–phalloidin).
Right, a phase-contrast image shows spheroid-specific alkaline phosphatase expression.Alkaline phosphatase activity was detected only in the
Amnion-Derived Pluripotent/Multipotent Stem Cells _______________________________________________________________139
Stem Cell Reviews ♦ Volume 2, 2006
of integrins. Although it remains to be proven, AE cell prolif-
eration may depend on the integrin and EGFR-associated
molecular complex and cell-to-cell interactions, which are facil-
itated by higher density culture (38). In our hands, human
telomerase reverse transcriptase mRNA expression was not
detected by RT-PCR in naive AE cells. This might suggest that
AE cells are sufficiently quiescent to preserve the characteris-
tics of differentiated somatic cells. To address this question,
the length of AE telomeres was investigated. Mosquera et al.
(39) performed precise evaluation of telomeres length and
telomerase activity on amniotic fluid cells. They described that
telomerase activity was 34–48% of that of a Hela cell line at 22 d
after the initiation of culture, and decreased thereafter.
According to Southern blotting of terminal restriction frag-
ment length, the telomere length initially observed progres-
sively decreased after each round of DNAreplication. Amniotic
fluid cells are a heterogenous cell population; however,
many are cytokeratin-positive cells like the AE cells that we
describe (40). If the data on amniotic fluid cells are represen-
tative of AE cells, it may be useful to release the telomerase
constraints with specific factors. In fact, exposure of AE cells
to trichostatin A, a histone deacetylase inhibitor, temporarily
and reversibly reactivated human telomerase reverse tran-
scriptase expression on AE cells (unpublished observations).
This suggests that optimized culture conditions might be cre-
ated which may allow establishing continuously growing lines
of pluripotent AE cells.
Cytogenetic Stability and Tumorigenicity
Telomerase activity is detectable in human ES cells,
MAPCs, humangerm cells, and 80–90% of human tumor sam-
ples. Telomerase-active stem cells, including ES cells and
MAPCs, sometimes become aneuploid. Aneuploidy is seen
more frequently once mouse MAPCs have been expanded for
greater than 60–70 population doublings and following
repeated freeze–thaw episodes. Furthermore, telomerase-active
ES cells form teratomas when the stem cells are injected into
the muscles of severe combined immunodeficient-beige mice.
By contrast, when from half a million to 2 million human AE
cells were injected into more than 50 individual mice and
observed for an average of 60 d (max. 516 d), no AE cell trans-
plants led to the development of tumors by any route of admin-
istration in severe combined immunodeficient-beige mice or
Rag-2 knockout mice. No cytogenetic abnormalities were
observed in cultured AE cells by simple analysis of the kary-
otype. The safety of AE cells upon transplantation has been
shown in a clinical setting. Some physicians focused on the
storage of glycogen by AE cells and transplanted minced
amnion or rolled tissue into glycogen-deficient patients in an
attempt to correct the metabolic defect. It is well known that
amnion does not express HLA class II antigens and only
expresses class I antigens at low levels. These observations
lead researchers to speculate that AE cells would not be rec-
ognized by the immune system and that allogeneic transplants
of AE cells would survive indefinitely. Researchers trans-
planted AE cells into the forearm of volunteers to determine
whether an immunological reaction would occur. No immuno-
logical reaction was noted, and importantly, in neither case
was there a report of tumorigenicity of the AE cells (41–43).
Drug Discovery and Toxicology
Stem cells are expected to become an important new tool
for drug development (44). Amnion-derived stem cells may
be an ideal cell source for these in vitro assays. The successful
differentiation of AE to cells with mature cardiac or hepatic
functions would be critical. Current preclinical models of car-
diotoxicity for drug candidates are difficult because of prob-
lems with appropriate cell models and the availability of the
most useful cell types. Primary human cardiomyocytes are
excellent tools with which to screen new chemical entities for
efficacy and safety. However, the availability of healthy human
cardiomyocytes is limited. AE cells have the potential to
Fig. 4. Summary of gene induced by exogenous growth factors. Naive AE cells were cultured with several growth factors for 1–2 wk. The
expression of key genes was evaluated by real-time quantitative reverse-transcription PCR and compared with the gene expression of start-
ing (naive) AE cells.
140 ______________________________________________________________________________________________ Miki and Strom
Stem Cell Reviews ♦ Volume 2, 2006
differentiate into cardiomyocytes. Another essential cell type
for drug development is the hepatocyte because of its high
level expression of drug-metabolizing enzymes. AE cells have
been shown to differentiate into hepatocyte-like cells that
demonstrate hepatic gene expression as well as some drug-
metabolizing functions. In addition to the expression of basal
levels of all of the hepatic genes examined to date, amniotic
epithelium-derived hepatocyte-like cells express regulated,
inducible cytochrome P450 genes. Importantly, the specific
drug-metabolizing enzymes are regulated and induced by pro-
totypical inducing agents in a manner similar, if not identical,
to authentic human hepatocytes. Once culture conditions are
optimized to produce mature cardimyocytes or hepatocytes
from AE cells, this may be a nearly unlimited source of cells
for drug metabolism and toxicology purposes.
The use of de-epithelialized human amniotic membrane
has been well documented in ophthalmic surgery for the treat-
ment of Stevens-Johnson syndrome, ocular cicatricial pem-
phigoid, acute thermal and alkali burns, pterygium surgery,
and limbal stem cell transplantation (45,46). For this applica-
tion, only the basement membrane of amnion is needed to
serve as a biological scaffold for epithelial cell migration as
well as for anti-inflammatory properties. An important point
is that human amnion has been approved as a medical mate-
rial by the Food and Drug Administration.
The low antigenicity of amnion may be an advantage for
amniotic epithelium-derived stem cell transplantation or cell
replacement therapy. Human MHC antigens are expressed
very weakly (class I) or not at all (class II) on AE cells (47,48).
This potential advantage was utilized during AE cell trans-
plantation to correct lysosomal storage disease (42). More than
50 cases of AE cell/tissue transplantation had been performed
in various institutes. Although the therapeutic effect for the
correction of lysosomal storage disease was varied and was
transient at best, there were no clear evidences of immuno-
logical reaction to, or rejection of, the transplanted cells. Finally
from a safety standpoint, as we have described in the section
on tumorigenicity, AE cells are clearly nontumorigenic when
transplanted into immunodeficient animals and no tumor
formation has been reported when human AE cells or tissues
were transplanted into patients.
The basis of most of the current ethical concerns with ES
cells is dependent on the need to derive the cell type from
developing blastocysts. Because the blastocyst could poten-
tially develop into an embryo if it were transferred to a suit-
ably prepared recipient, the argument can be made that the
derivation of ES cells interrupts normal development, which
might have produced a human life. Derivation of ES-like stem
cells from placenta would not be expected to raise the same
objections. Currently normal full-term placenta is evaluated
after the birth of the baby and is discarded at the hospital as
medical waste. Because the tissue is only made available fol-
lowing a live birth and is not used for any other purpose, the
isolation and use of stem cells from a discarded placenta would
not be expected to elicit any ethical concerns. No potential
human life is interrupted to derive the stem cells; in fact, the
cells only become available following a normal live birth. Once
technology is established to isolate and/or propagate pluripo-
tent stem cells from the placenta, this medical waste may turn
into valuable property. The precedent and current use of umbil-
ical cord blood stem cells has helped establish protocols for
the possible future use of amnion-derived stem cells. It has
been established that the placenta belongs to the baby and
that parents execute control over the use of the placental
tissues/cells on behalf of the baby. Therefore, the use of human
placenta-derived cells for research purposes and clinical appli-
cations would not be expected to elicit current or future ethi-
cal or legal conflicts.
Here, we review the stem cell characteristics of amnion cells,
especially AE cells. During development, amnion differenti-
ates early form the pluriptoent epiblast prior to cell type spec-
ification, which occurs at gastrulation. Additionally, the amnion
remains separate from the developmental signals, which reg-
ulate differentiation of the epiblast to form all of the cell types
and organs of the developing embryo. At least some of the AE
cells seem to have preserved the pluripotent characteristics of
the epiblast. Cell surface marker expression suggests that
approx 10% of the AE cells express markers which are com-
monly found on ES cells.
Eight years form the first report of the establishment of
human ES cells, the classification of stem cells is now gener-
ally held only for cells which fulfill the dual requirements of
self-renewal and pluripotency. However, even this definition
does not enjoy complete agreement (49). Plasticity or differ-
entiation potential is an essential characteristic of stem cells,
which could be used for further clinical applications. On the
other hand, “unlimited self-renewal” may not be an essential
requirement. Cells in adult tissues do not have unlimited
growth potential and most organs function normally without
clearly defined stem cell compartments. Extended growth
potential and self-renewal are only required when the initial
stem cell source is limited and not able to generate sufficient
numbers of cells for therapeutic use without expansion in cul-
ture. If abundant pluripotent cells could be obtained from the
starting material for differentiation protocols or for cell trans-
plantation and regenerative medicine applications, the require-
ment for extended self-renewal capacity becomes less
important. Furthermore, extended culture carries its own risks.
Long-term culture of cells is frequently accompanied by genetic
changes in the karyotype and tumorigenicity.
Placenta, the source of AE cells, is abundantly available as
a discarded tissue and is free of the ethical concerns of other
stem cells. It could easily be imagined that banks of AE cells
could be established which contained a precise MHC match
for every possible transplant recipient without the need for
extensive technical manipulations such as somatic nuclear
transfer. So far, no stem cells are able to differentiate into ther-
apeutically useful cell types in vitro or their differentiation is
not well controlled. As with the other stem cells, further inves-
tigations will be required to induce AE cells to differentiate to
therapeutically useful cells. However, because they are abun-
dantly available without ethical concerns and because of the
Amnion-Derived Pluripotent/Multipotent Stem Cells _______________________________________________________________141 Download full-text
Stem Cell Reviews ♦ Volume 2, 2006
advantages of pluripotency, low immunogenicity, and lack of
tumorigenicity, amniotic epithelium-derived stem cells may
be an extremely useful cell source for transplantation and organ
and tissue regeneration.
1. Hoyes AD. J Obstet Gynaecol Br Commonw 1968;75(20):164–171.
2. Klinger HP, Schwarzacher HG. Nature 1958;181(4616):
3. Terada S, Matsuura K, Enosawa S, et al. Cell Transplant
4. Xu RH, Chen X, Li DS, et al. Nat Biotechnol 2002;20(12):
5. Klassen H, Ziaeian B, Kirov II, Young MJ, Schwartz PH.
J Neurosci Res 2004;77(3):334–343.
6. Schwartz PH, Bryant PJ, Fuja TJ, Su H, O’Dowd DK, Klassen H.
J Neurosci Res 2003;74(6):838–851.
7. Henderson JK, Draper JS, Baillie HS, et al. Stem Cells
8. Zhou S, Schuetz JD, Bunting KD, et al. Nat Med 2001;
9. Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC.
J Exp Med 1996;183(4):1797–1806.
10. Terskikh AV, Easterday MC, Li L, et al. Proc Natl Acad Sci USA
11. Asakura A, Rudnicki MA. Exp Hematol 202;30(11):1339–1345.
12. Sakuragawa N. US Pregnant Abst 20050089513 2004;919310(10).
13. Brivanlou AH, Gage FH, Jaenisch R, Jessell T, Melton D,
Rossant J. Stem cells Sci 2003;300(5621):913–916.
14. Chambers I, Colby D, Robertson M, et al. Cell 2003;113(5):
15. Scholer HR, Ruppert S, Suzuki N, Chodhury K, Gruss P. Nature
16. Rosner MH, Vigano MA, Ozato K, et al. Nature 190;345(6277):
17. Pesce M, Wang X, Wolgemuth DJ, Scholer H. Mech Dev
18. Pesce M, Scholer HR. Stem Cells 2001;19(4):271–278.
19. Niwa H. Cell Struct Funct 2001;26(3):137–138.
20. Hattori N, Nishino K, Ko YG, et al. J Biol Chem 2004;279(17):
21. Jiang Y, Jahagirdar BN, Reinhardt RL, et al. Nature 2002;
22. Tamagawa T, Ishiwata I, Saito S. Hum Cell 2004;17(3):125–130.
23. Miki T, Lehmann T, Cai H, Stolz DB, Strom SC. Stem Cells
24. Miyamoto K, Hayashi K, Suzuki T, et al. Stem cells 2004;
25. Miki T, Mitamura K, Ross M, Stolz DB, Strom SC. Localization
of stem cell marker positive cells in human term placentae. 2006,
26. Sakuragawa N, Thangavel R, Mizuguchi M, Hirasawa M,
Kamo I. Neurosci Lett 1996;209(1):9–12.
27. Sakuragawa N, Misawa H, Ohsugi K, et al. Neurosci Lett
28. Elwan MA, Sakuragawa N. Neuroreport 1997;232(1):53–56.
29. Kakishita K, Elwan MA, Nakao N, Itakura T, Sakuragawa N.
Exp Neurol 2000;165(1):27–34.
30. Kakishita K, Nakao N, Sakuragawa N, Itakura T. Brain Res
31. Sakuragawa N, Enosawa S, Ishii T, et al. J Hum Genet 2000;45(3):
32. Nakajima T, Enosawa S, Mitani T, et al. Cell Transplant 2003;12(5):
33. Takashima S, Ise H, Zhao P, Akaike T, Nikaido T. Cell Struct
34. Wei JP, Zhang TS, Kawa S, et al. Cell Transplant 2003;
35. Takahashi T, Lord B, Schulze PC, et al. Circulation 2003;107(14):
36. Carlson BM. Human embryol & development biology. Mosby,
37. Moro L, Venturino M, Bozzo C, et al. EMBO J 1998;17(22):6622–6632.
38. Schlaepfer DD, Hunter T. Mol Cell Biol 1996;16(10):6623–5633.
39. Mosquera A, Fernandez JL, Campos A, Goyanes VJ, Ramiro-
Diaz J, Gosalvez J. J Med Genet 1999;36(6):494–496.
40. Chen W. J Med Genet 1982;19(6):433–436.
41. Scaggiante B, Pineschi A, Sustersich M, Andolina M, Agosti E,
42. Sakuragawa N, Yoshikawa H, Sasaki M. Brain Dev 1992; 4(1):7–11.
43. Akle CA, Adinolfi M, Welsh KI, Leibowitz S, McColl I. Lancet
44. Davila JC, Cezar GG, Thiede M, Strom S, Miki T, Trosko J. Toxicol
45. Dua HS, Gomes JA, King AJ, Maharajan VS. Surv Ophthalmol
46. Gomes JA, Romano A, Santos MS, Dua HS. Curr Opin
47. Adinolfi M, Akle CA, McCol I, et al. Nature 1982;295(5847):
48. Sakuragawa N, Tohyama J, Yamamoto H. Cell Transplant
49. Zipor D. Stem cells 2005;23(6):719–726.