BIOLOGY OF REPRODUCTION 77, 577–588 (2007)
Published online before print 9 May 2007.
DOI 10.1095/biolreprod.106.055244
Stem Cells Derived from Human Fetal Membranes Display Multilineage
Differentiation Potential
Sivakami Ilancheran,2,3 Anna Michalska,3,4 Gary Peh,3,4 Euan M. Wallace,2,3 Martin Pera,3,4 and
Ursula Manuelpillai1,2,3
Center for Women’s Health Research, Department of Obstetrics and Gynecology,2 Monash University, Monash Medical
Center, Clayton, Victoria 3168, Australia
Monash Institute of Medical Research,3 Monash University, Clayton, Victoria 3168, Australia
Australian Stem Cell Center,4 Clayton, Victoria 3168, Australia
fetus lies. The timing of the derivation of amnion from the
epiblast is important, because it has been suggested that tissues
that form prior to gastrulation may retain stem cell or stem celllike capabilities [1].
Indeed, mixed populations of fetal- and amnion-derived
cells obtained by amniocentesis in the second trimester of
pregnancy express the transcription factors POU domain, class
5, transcription factor 1 (POU5F1) and Nanog homeobox
(NANOG) [2], which are necessary for maintaining the
undifferentiated state of pluripotent stem cells. Further, the
cells obtained at amniocentesis display self-renewal properties
forming clonal colonies [2], have decreased telomerase activity
[3], resemble mesenchymal stem cells [4] and neural progenitor
cells [5], and can differentiate into osteocyte-, adipocyte-, and
neural-like cells [2, 6]. While these studies have clearly
demonstrated the stem cell-like properties of amniotic fluid
cells obtained in the second trimester, amniocentesis is usually
undertaken for prenatal genetic diagnosis, and so the majority
of retrieved cells are currently used for this purpose.
Amniocentesis is also associated with a small but significant
excess risk of miscarriage [7]. Thus, if the amnion was to
become a useful source of stem cells, a better strategy may be
to determine whether cells harvested from fetal membranes that
are delivered after birth possess stem cell-like characteristics.
This would provide a vast supply of cells from a tissue that is
easily accessible and currently discarded after birth. However,
to date very few studies have examined whether term amnion
cells, as opposed to midtrimester cells, possess stem cell-like
properties.
Mixed populations of fetal membrane-derived cells express
numerous cell surface and intracellular antigens that are
characteristic of mesenchymal stem cells [8], and in a recent
study it was reported that passaged term human amniotic
epithelial cells (hAECs) can differentiate into cells resembling
those derived from neuroectodermal, mesodermal, and endodermal lineages [9].
It has also been shown that native hAECs express the
nonpolymorphic, nonclassical human leukocyte antigen G
(HLA-G) [10] but lack the polymorphic antigens HLA-A,
HLA-B, and HLA-C (class IA) and HLA-DR (class II) on their
surfaces [11]. These findings suggest that hAECs may be
immunologically inert and would have a reduced risk of
rejection upon transplantation. Term amnion is a highly
abundant and easily accessible tissue that may potentially be
an important source of transplantable stem cells. We
investigated whether term hAECs express markers characteristic of stem cells, self-renewal forming clonal colonies, form
teratomas, differentiate into cells derived from each of the three
primary germ layers, viability of differentiated cells in vitro and
ABSTRACT
The amnion is the inner of two membranes surrounding the
fetus. That it arises from embryonic epiblast cells prior to
gastrulation suggests that it may retain a reservoir of stem cells
throughout pregnancy. We found that human amniotic epithelial
cells (hAECs) harvested from term-delivered fetal membranes
express mRNA and proteins present in human embryonic stem
cells (hESCs), including POU domain, class 5, transcription
factor 1; Nanog homeobox; SRY-box 2; and stage-specific
embryonic antigen-4. In keeping with possible stem cell-like
activity, hAECs were also clonogenic, and primary hAEC cultures
could be induced to differentiate into cardiomyocytic, myocytic,
osteocytic, adipocytic (mesodermal), pancreatic, hepatic (endodermal), neural, and astrocytic (neuroectodermal) cells in vitro,
as defined by phenotypic, mRNA expression, immunocytochemical, and/or ultrastructural characteristics. However, unlike
hESCs, hAECs did not form teratomas upon transplantation into
severe combined immunodeficiency mice testes. Last, using flow
cytometry we have shown that only a very small proportion of
primary hAECs contain class IA and class II human leukocyte
antigens (HLAs), consistent with a low risk of tissue rejection.
However, following differentiation into hepatic and pancreatic
lineages, significant proportions of cells contained class IA, but
not class II, HLAs. These observations suggest that the term
amnion, an abundant and easily accessible tissue, may be a
useful source of multipotent stem cells that possess a degree of
immune privilege.
amnion, developmental biology, fetal membranes, placenta, stem
cells
INTRODUCTION
In pregnancy the human fetus is surrounded by two fetal
membranes: the amnion and the chorion. The chorion is the
outer of the two membranes and, like the placenta, is derived
from the trophectoderm of the implanting blastocyst. The
amnion arises prior to gastrulation from embryonic epiblast
cells that are also destined to develop into the three primary
fetal germ layers: the endoderm, the mesoderm, and the
ectoderm. The amnion forms a fluid-filled sac in which the
1
Correspondence: Ursula Manuelpillai, Department of Obstetrics and
Gynecology, Monash University, Monash Medical Center, 246 Clayton
Rd., Clayton, Victoria 3168, Australia. FAX: 61 3 9594 6389;
e-mail: [email protected]
Received: 2 July 2006.
First decision: 21 August 2006.
Accepted: 30 April 2007.
Ó 2007 by the Society for the Study of Reproduction, Inc.
ISSN: 0006-3363. http://www.biolreprod.org
577
578
ILANCHERAN ET AL.
TABLE 1. Primer sequences and polymerase chain reaction parameters.
Primer sequence (5’! 3’)a
Gene
POU domain, class 5,
transcription factor 1 (POU5F1)
Cripto (CFC1)
Paired box gene 6 (PAX6)
Forkhead box D3 (FOXD3)
Prominin 1 (PROM1)
SRY (sex determining region Y)-box 2 (SOX2)
Developmental pluripotency associated 3 (DPPA3)
Nanog homeobox (NANOG)
Growth differentiation factor 3 (GDF3)
a
F:
R:
F:
R:
F:
R:
F:
R:
F:
R:
F:
R:
F:
R:
F:
R:
F:
R:
CGTTCTCTTTGGAAAGGTGTTC
ACACTCGGACCACGTCTTTC
CAGAACCTGCTGCCTGAATG
GTAGAAATGCCTGAGGAAACG
AACAGACACAGCCCTCACAAACA
CGGGAACTTGAACTGGAACTGAC
GCAGAAGAAGCTGACCCTGA
CTGTAAGCGCCGAAGCTCT
CAGAAGGCATATGAATCC
CACCACATTTGTTACAGC
GGCAGCTACAGCATGATGCAGGAGC
CTGGTCATGGAGTTGTACGCAGG
GTTACTGGGCGGAGTTCGTA
TGAAGTGGCTTGG TGTCTTG
CAGCTGTGTGTACTCAATGATAGATTT
CAACTGGCCGAAGAATAGCAATGGTGT
AGACTTATGCTACGTAAAGGAGCT
CTTTGATGGCAG ACAGGTTAAAGTA
Product
size (bp)
Annealing
temp./time
Extension
temp./time
350
558C, 1 min
728C, 1 min
185
558C, 1 min
728C, 1 min
275
558C, 1 min
728C, 1 min
350
558C, 1 min
728C, 1 min
180 (1)
153 (2)
130
578C, 1 min
728C, 1 min
678C, 1 min
728C, 1 min
174
608C, 30 sec
728C, 30 sec
142
608C, 30 sec
728C, 30 sec
150
608C, 30 sec
728C, 30 sec
F, forward; R, reverse.
whether HLA class IA and class II antigens remain suppressed
following differentiation.
MATERIALS AND METHODS
Tissue Collection and Processing
The study was approved by the Southern Health Human and Animal
Research Ethics Committees of Monash Medical Center. Informed, written
consent was obtained from each patient prior to collection. Fetal membranes (n
¼ 21) were retrieved from healthy women delivered at term by elective cesarean
delivery. Severe combined immunodeficient (SCID) mice (n ¼ 3) were obtained
from Southern Health Animal Facilities, Victoria, Australia. Guidelines relating
to the care and use of animals were approved by Southern Health.
Isolation of hAECs
The amnion was peeled from the chorion and rinsed in phosphate-buffered
saline (PBS). Tissue was digested twice in 0.25% trypsin containing 0.5 mM
EDTA in Hanks’ Balanced Salt Solution for 20 min at 378C with agitation.
Trypsin was inactivated with fetal calf serum (FCS), and the solution was
filtered and centrifuged at 2300 3 g for 10 min. After washing cells in M199
medium, contaminating erythrocytes were lysed in hypotonic solution (8%
ammonium chloride, 0.84% sodium bicarbonate, and 0.37% EDTA) for 10 min
at 378C with gentle shaking. Cells were washed and resuspended in Dulbecco
modified Eagle medium (DMEM)/F12 containing 100 U/ml penicillin/
streptomycin solution and 10% FCS. Media and supplements were purchased
from Invitrogen (Mount Waverly, Australia).
The purity of the isolates was determined by flow analyses for the epithelial
marker cytokeratin 7 (KRT7; Dako, Carpentaria, CA). Briefly, 5 3 105 cells
were incubated with permeabilizing reagent (Dako), 10% goat serum, and Iry
antibody diluted 1:100 for 30 min at room temperature. After washing, cells
were incubated with phycoerythrin-conjugated anti-mouse IgG (1:10; Dako) for
30 min. Isotype-matched mouse IgG at concentrations similar to those of the Iry
antibody was used as a negative control. Isolates that were more than 99%
positive for KRT7 were used for experiments described below.
RT-PCR Analysis
Total RNA was isolated from amnion tissue (n ¼ 5) using Trizol
(Invitrogen). RNA was also isolated from freshly harvested hAECs (n ¼ 5)
and hESC line hES2 using QIAshredder and RNeasy columns (Qiagen,
Doncaster, Australia). Following DNase treatment, 1 lg total RNA was reverse
transcribed into cDNA using random primers and Superscript III (Invitrogen).
The cDNA diluted 1:20 was subjected to 36 cycles of PCR using primers and
parameters shown in Table 1. Human ESCs served as a positive control,
whereas the negative control consisted of reactions omitting reverse
transcriptase enzyme. The resulting products were separated on 1%–2%
agarose gels containing ethidium bromide.
Immunocytochemistry for Stem Cell Markers
Immunocytochemistry was performed on hAECs cultured for 24 h to detect
POU5F1, SOX2, stage-specific embryonic antigen-4 (SSEA-4), and germ cell
tumor marker 2 (GCTM2; n ¼ 5). SSEA-4 is a glycolipid found on hESCs and
cells of the inner cell mass [12, 13]. GCTM2 is a keratin sulphate proteoglycan
present in pluripotent embryonal carcinoma and undifferentiated hESCs [14].
Human AECs were fixed in ice-cold ethanol and blocked in 10% human serum
for 45 min at room temperature. Primary antibodies against human POU5F1
(1:50; Santa Cruz Biotechnology, Santa Cruz, CA), SOX2 (1:200; Chemicon,
Temecula, CA), SSEA-4 (1:200; Developmental Hybridoma Bank, Iowa City,
IA), and GCTM2 (1:2; from Martin Pera, Australian Stem Cell Center,
Melbourne, Australia), were applied and incubated overnight at 48C. Controls
were incubated with corresponding concentrations of isotype-matched IgG in
lieu of Iry antibodies. These were IgG2b (POU5F1), IgG3 (SSEA-4), IgM
(GCTM2), and rabbit IgG (SOX2). Alexa Fluoro dye-conjugated goat antimouse or goat anti-rabbit IIry antibodies (1:100; Invitrogen) were applied to
hAECs for 1 h at room temperature. Cells were mounted in Vectorshield
containing 4 0 ,6-diamidino-2-phenylindole nuclear stain (Vector Laboratories,
Burlingame, CA). Human AECs in 8–10 randomly selected fields were viewed
at 2003 magnification, and the percentage (mean 6 SEM) of positively
staining cells was calculated.
Clonal Culture and Cloning Efficiency
To test for self-renewal properties through clonogenicity, hAECs from five
fetal membranes were seeded at a clonal density of 50 cells/cm2 in 60-mm
diameter Petri dishes (Becton Dickinson Biosciences, Bedford, MA). Control
cultures were maintained in DMEM/F12 with 10% FCS. Cells were also grown
in serum-free DMEM/F12 supplemented with 10 ng/ml epidermal growth
factor (EGF) or basic fibroblast growth factor (bFGF), a combination of EGF þ
FGF (10 ng/ml each), or activin A (10 ng/ml). EGF and bFGF were purchased
from Sigma-Aldrich (St. Louis, MO), and activin A was purchased from R&D
Systems (Minneapolis, MN). Cultures were maintained for 21 days, with media
replaced thrice weekly. Cell colonies were fixed in 10% formalin for 10 min,
immunostained for KRT7 (1:100; Dako), and visualized with 3,3 0 -diaminobenzidine (DAB) chromogen. A cluster was considered to be a clonal colony
when the cell number exceeded 50 [15]. Cell number and cloning efficiency
percentage were calculated as described previously [15]. Briefly, numbers of
cells within each colony were estimated using a 10 3 10 grid mounted on an
eyepiece graticule. The cloning efficiency percentage was calculated using the
formula cloning efficiency (%) ¼ (number of colonies/number of cells seeded)
3 100.
Transplantation of hAECs into Murine Testes
Freshly isolated hAECs (5 3 105 viable cells) were inoculated into one
testis each of three adult SCID mice, whereas saline was injected into the other
testis. Ten weeks after inoculation, animals were killed, and testes were excised
and fixed in 4% paraformaldehyde. To identify hAECs, immunohistochemistry
579
STEM CELLS FROM FETAL MEMBRANES
TABLE 2. Supplements, characterization, and transformation efficiency of human amniotic epithelial cells.
Lineage
Mesoderm
Myocytic
Osteocytic
Adipocytic
Cardiomyocytic
Endoderm
Pancreatic
Hepatic
Ectoderm
Neural
a
b
Supplements
Characterized by
Percentage of cells showing phenotypic
changes and markers/stains used for
characterization after 3–4 wk in culturea
Human serum (5 %)
Hydrocortisone (50 lM)
1,25-Dihydroxyvitamin D3 (0.01 lM)
Ascorbic acid (50 lM)
b-Glycerophosphate (10 mM)
Indomethacin (200 lm)
Isobutyl-methylxanthine (0.5 mM)
Dexamethasone (1 lM)
Insulin (10 lM)
Ascorbic acid (1 mM)
Smooth muscle alpha-actin (ACTA2)
73.75 6 2.27
Von Kossa stain
54.25 6 4.27
Oil Red O stain
29 6 2.91
Troponin T (TNNT)
91.25 6 4.15b
Nicotinamide (10 mM)
Insulin (0.1 lM)
Dexamethasone (1 3 107 M)
Glucagon (GCG)
Albumin (ALB)
Hepatocyte growth factor (HGF)
69.25 6 2.16b
93.75 6 2.18b
61.75 6 2.86b
All-trans retinoic acid (5 3 105 M)
Fibroblast growth factor 4 N-2
and B-27 (10 ng/ml)
Nestin (NES)
Microtubule-associated protein 2 (MAP2)
4.93 6 3.61
11.24 6 5.86
Glial fibrillary acidic protein (GFAP)
84.51 6 4.11
Values are mean 6 SEM, n ¼ 5.
Data obtained from flow cytometric analyses.
for human KRT7 (1:100) was performed on 5-lm sections. Endothelial cells
lining murine blood and lymph vessels were identified using rabbit anti-mouse
PECAM1 (1:100; Dako). Immunostaining was visualized using DAB
chromogen.
Differentiation of hAECs into Multiple Lineages and Their
Characterization
Freshly isolated hAECs (5 3 104 viable cells/well) were plated onto
collagen IV-coated coverslips seated in 24-well plates in standard medium
consisting of DMEM/F12 supplemented with 10% FCS. Plated cells were
divided into three groups. In group 1, cultures were maintained in standard
medium for 24 h only, whereas cells in group 2 were cultured for 4 wk. In
group 3, media was removed after 24 h of plating, and fresh DMEM/F12 with
10% FCS and the supplements shown in Table 2 was added to induce
mesodermal (adipocytic, osteocytic, myocytic, and cardiomyocytic) and
endodermal (pancreatic and hepatic) differentiation.
Neuroectodermal differentiation was induced by plating cells on poly-Dlysine/laminin-coated coverslips and grown in neural basal A medium
(Invitrogen) containing supplements (Table 2). Cultures treated with supplements were also maintained for 4 wk. These experiments were performed in
quadruplicate wells from hAECs isolated from five fetal membranes. The
percentage of viable cells was determined by the trypan blue dye exclusion test.
For evidence of adipogenic differentiation, cells were tested for lipid
granules using Oil Red O stain. Briefly, cells were fixed in 10% neutralbuffered formalin vapor at room temperature for 15 min and rinsed in 60%
isopropanol. The Oil Red O solution was applied for 20 min at room
temperature. Cells were rinsed in 60% isopropanol, then in distilled water, and
were mounted in aqueous mounting medium.
Osteoblastic differentiation was assessed by identifying calcium accumulation using Von Kossa stain. Cells fixed as described above were washed in
distilled water and incubated in 5% silver nitrate solution for 1 h under
ultraviolet light. Cells were washed in distilled water, dehydrated, and mounted
using DPX mounting medium.
Immunocytochemistry was also carried out to identify antigens associated
with differentiated cell phenotypes. These included glucagon (GCG; pancreatic
cells); albumin (ALB) and hepatocyte growth factor (HGF; hepatocytes);
troponin T (TNNT; cardiomyocytes) [16]; smooth muscle alpha-actin (ACTA2;
myocytes); neural progenitor marker nestin (NES); neuronal-specific marker;
microtubule-associated protein 2 (MAP2); and astroglial marker glial fibrillary
acidic protein (GFAP).
Cells fixed in ethanol were blocked for 45 min at room temperature with
10% human serum in PBS (NES, GFAP, and MAP2) or 10% donkey serum
and 1% bovine serum albumin in PBS (GCG, HGF, ALB, ACTA2, and
TNNT). Primary antibodies were applied at the following dilutions: ACTA2
(1:100; Dako); TNNT, GCG, HGF, and ALB (1:10; R&D Systems); NES
(1:200; Chemicon); GFAP (1:400; Chemicon); and MAP2 (1:100; Lab Vision
Corp., Fremont, CA). Corresponding concentrations of mouse IgG1 (ACTA2,
TNNT, NES, MAP2, and GFAP), mouse IgG2a (GCG and ALB), or goat IgG
(HGF) in 1% BSA in PBS were applied to controls. Cells were incubated
overnight at 48C. After several rinses, horseradish peroxidase-labeled biotinstrepavidin (LSAB kit; Dako) was applied at room temperature for 30 min,
followed by DAB to detect ACTA2, GCG, HGF, ALB, and TNNT. To detect
NES, MAP2, and GFAP, 1:1000 diluted Alexa Fluor goat anti-mouse IIry
antibody (Invitrogen) was applied for 1 h, and cells were washed and mounted
in Vectorshield. Cells in 8–10 randomly selected fields viewed at 2003
magnification showing positive staining/phenotypic changes characteristic of
differentiated cells were expressed as a percentage (mean 6 SEM).
Insulin concentrations were measured in media conditioned by cells grown
in pancreatic differentiation media using a commercial assay (Access/DXI Ultra
Sensitive Insulin Assay; Beckman Coulter, Gladesville, Australia).
Human AECs differentiating into hepatic, pancreatic, and cardiomyocytic
cells were further characterized through RT-PCR and/or flow cytometric
analyses and transmission electron microscopy (TEM).
Total RNA isolated from control cultures and differentiated cells was
analyzed (n ¼ 4). For cardiomyocytic cells, mRNA expression of GATAbinding protein 4 (GATA4); atrial natriurertic peptide (ANP); calcium channel,
voltage-dependant, L type, alpha 1C subunit (CACNA1C); potassium voltagegated channel, Shal-related subfamily, member 3 (KCND3); and myosin light
chain 7 (MYL7) was analyzed, and for pancreatic cells, amylase, alpha 2B
(AMY2B) was analyzed. RNA isolation and cDNA synthesis were carried out as
described earlier. The cDNA was amplified for 36 cycles using PCR primers
and conditions described previously [16, 17]. GATA4 is a marker of precardiac
cells [16], ANP is a vasorelaxant hormone secreted by cardiomyocytes [18],
and MYL7 is important in the development of cardiac myofibrillogenesis and
myocardial contraction [19]. CACNA1C and KCND3 are cardiac-specific Ltype calcium channel and transient outward potassium channel proteins,
respectively [20, 21]. AMY2B is an enzyme produced by pancreatic exocrine
cells [22].
For flow analyses, cells were incubated with antibodies against ALB, HGF,
GCG, and TNNT for 45 min at room temperature at concentrations used for the
immunocytochemistry described above. After several washes, 1:100 diluted
allophycocyanin (APC)-conjugated (Becton Dickinson Biosciences) goat antimouse (for ALB, GCG, and TNNT) and donkey anti-goat (for HGF) IIry
antibodies (Becton Dickinson Biosciences) were applied for 30 min. Cells were
washed and analyzed by flow cytometry.
For TEM, cells were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate
buffer for 2 h at room temperature and then left at 48C overnight. Cells were
postfixed in 1% osmium tetroxide and dehydrated in graded acetone, infiltrated,
and embedded in Spurrs resin at 608C for 24 h. Ultrathin sections (80 nm)
stained with 3% uranyl acetate and Renoylds stain were examined.
580
ILANCHERAN ET AL.
Flow Cytometric Analyses for HLA Class IA and II Antigens
Transplantation of hAECs into Mouse Testes
Freshly isolated hAECs, control cultures grown in standard media, and cells
maintained in media inducing hepatic, pancreatic, and cardiomyocytic
differentiation for 4 wk were analyzed (n ¼ 4). Cells (5 3 105 cell/ml) were
incubated with 10 ll APC-conjugated mouse anti-human HLA-A, HLA-B, or
HLA-C (Becton Dickinson Biosciences) or 10 lg/ml mouse anti-human HLADP, HLA-DQ, or HLA-DR (Becton Dickinson Biosciences) for 45 min in PBS
containing 1% BSA. Corresponding concentrations of APC-conjugated IgG1j
and IgG2aj (Becton Dickinson Biosciences) in 1% BSA was applied to the
negative controls. For cells incubated with class II antibody, after several
washes, 1:100 diluted APC-conjugated goat anti-mouse IIry antibody (Becton
Dickinson Biosciences) was applied for 30 min. Cells were washed and
analyzed by flow cytometry.
Following transplantation of freshly isolated hAECs into the
testes of SCID mice, no palpable tumors were detected after 10
wk after injection. However, examination of testes tissue
sections showed monolayers of cells that were positive for
human KRT7 surrounding some blood/lymph vessels distant to
the sites of injection. Murine testicular cells were negative for
KRT7. Murine blood and lymph vessels were identified by
positive immunostaining for murine PECAM1, an antigen
found on endothelial cells lining the blood/lymph vessels (Fig.
3).
Statistical Analysis
Quantified data are presented as the mean 6 SEM. The data were analyzed
by the Kruskal-Wallis one-way analysis of variance followed by Dunn’s
posthoc test using GraphPad Prism software (version 4.03). Significance was
accorded at P , 0.05.
RESULTS
Messenger RNA Expression of Stem Cell Markers
Term amnion and freshly isolated hAECs expressed
POU5F1, SOX2, CFC1, NANOG, DPPA3, PROM1, and
PAX6. The mRNA of the pluripotent markers FOXD3 and
GDF3 was not detected. Undifferentiated hESCs expressed all
pluripotent and multipotent markers tested. Amplified cDNA
was absent in negative control reactions that lacked the reverse
transcriptase enzyme (Fig. 1A).
Localization of Stem Cell Markers
The cell surface antigens SSEA-4 and GCTM2 localized to
the majority of isolated cells (mean 6 SEM percentages: 96 6
1.2 and 98 6 1.5, respectively; n ¼ 5). The transcription factors
POU5F1 and SOX2 were also detected in hAECs cultured
overnight, with nuclear staining restricted to a minority of cells.
The percentages of cells showing nuclear staining for POU5F1
and SOX2 were 4.8 6 1.2 and 6.3 6 1.8, respectively. Control
cultures incubated with isotype-matched IgG lacked immunostaining (Fig. 1B).
Clonogenicity of hAECs
Following overnight incubation, cells seeded at a clonal
density of 50/cm2 had adhered onto the culture dishes and were
sparsely distributed. Colony formation was not observed in
control cultures grown in DMEM/F12 with 10% FCS for up to
21 days. Human AECs were also grown in serum-free DMEM/
F12 media supplemented with EGF, bFGF, and activin A.
Clonal colony formation was not supported by activin A (data
not shown). EGF and bFGF alone and in combination
promoted colony growth, with small clusters of cells visible
by Day 7 and large, flattened, undifferentiated colonies
containing numerous cells by Day 21. Colonies were
successfully subcloned and expanded in the presence of EGF
and bFGF. Cells within the colonies stained positively for
KRT7, a cytoskeletal protein present in hAECs but not
mesenchymal cells and fibroblasts. The cloning efficiency
percentage was calculated on Day 14, when discrete colonies
containing more than 50 cells were clearly visible. Cloning
efficiency was significantly higher in the presence of EGF and
bFGF compared with EGF or bFGF alone (P , 0.01). The
effect of EGF and bFGF on clonal colony formation and
cloning efficiency is shown in Figure 2.
Differentiation of hAECs
Although hAECs did not form teratomas in vivo, we tested
their ability to differentiate into cells derived from primary
germ layers using supplements given in Table 2.
Neuroectodermal Lineage
Freshly harvested hAECs showed strong immunostaining
for the early neural lineage marker NES, the neuronal marker
MAP2, and astrocytic protein GFAP (mean 6 SEM percentages 93.7 6 3.2, 96.4 6 2.9, and 94.8 6 3.3, respectively).
Upon culturing hAECs in neural differentiating medium, cell
numbers declined sharply. After 4 wk, a small percentage of
NES-positive cells and MAP2-positive neurons with large
central bodies and thin, elongated processes were observed.
The majority of cells remaining after 4 wk produced GFAP
(Table 2). These cells had large cell bodies and thin processes
characteristic of astrocytes. The neural lineage markers present
in hAECs soon after isolation and following culture with
supplements is shown in Figure 4.
Endodermal Lineages
Freshly isolated hAECs and cells grown in standard media
for 4 wk lacked GCG, ALB, and HGF. Cells grown with
hepatocyte-inducing supplements were large, with the majority
demonstrating dense, granule-rich cytoplasm and binucleated
cells characteristic of human hepatocytes [23]. The majority of
differentiated cells produced ALB and HGF (Fig. 5 and Table
2). Ultrastructural analyses revealed marked differences
between control and differentiated cells. The double-layered
nuclear membrane, extensive rough endoplasmic reticulum
surrounding the nucleus, and vesicles resembling bile canaliculi characteristic of human hepatocytes were present in
differentiated cells but absent in hAECs grown in standard
media (Fig. 5, E–H).
Cells grown in media promoting pancreatic differentiation
formed masses of cells with a cystlike appearance known to be
characteristic of pancreatic cells [24]. These cells expressed the
pancreatic exocrine cell marker AMY2B and produced the
hormone GCG (Fig. 6 and Table 2). However, insulin was not
detected in the conditioned media. Differentiated cells showed
ultrastructural features characteristic of acinar beta exocrine
pancreatic cells, including bodies resembling beta and
zymogen granules, condensing vacuoles, and distended Golgi
saccules (Fig. 6).
Mesodermal Lineages
Freshly isolated hAECs and cells cultured in standard
DMEM/F12 for 4 wk lacked immunostaining for the myocytic
marker ACTA2 or for mineralized bone or lipid deposits
characteristic of osteocytes and adipocytes, respectively.
581
STEM CELLS FROM FETAL MEMBRANES
FIG. 1. Stem cell markers. A) The mRNAs
of several markers are expressed in amnion
tissue and hAECs. Human ESCs served as a
positive control. B) Immunolocalization of
stem cell markers in hAECs cultured overnight (left panels). Cell nuclei stained with
DAPI are shown in the right panels. Human
AECs showing POU5F1 and SOX2 nuclear
staining only are indicated by arrowheads.
Human ESCs that served as a positive
control show nuclear POU5F1 staining
only. A representative IgG negative control
is also shown. Bars ¼ 50 lm.
However, hAECs cultured with myocytic-promoting supplements formed elongated, striated cells that were positive for
ACTA2.
Compared with controls, under conditions inducing osteocytic differentiation, cells enlarged two to three times, and
many were binucleated and contained mineralized bone
deposits. Cells grown in adipocytic media also enlarged three
to four times, with numerous multinucleated cells containing
lipid droplets scattered throughout the cytoplasm. Control
cultures and those grown with supplements inducing differentiation into mesodermal lineages are shown in Figure 7.
Freshly isolated, control, and cells cultured in cardiomyocytic-promoting medium expressed the cardiomyocytic precursor marker GATA4 and differentiated cell markers ANP, MYL7,
CACNA1C, and KCND3. TTNT was absent in freshly isolated
cells and control cultures. However, TTNT was abundant in
differentiated cells (Fig. 8 and Table 2). Ultrastructural
analyses showed that compared with control cultures, differentiated cells contained T tubules, numerous myofilaments and
myofibrils, and H bands characteristic of relatively mature
cardiomyocytes (Figs. 5E and 8, D–F).
Viability and Transformation Efficiency
Cells showing the differentiated phenotypes and markers
described above were evident within 2 wk of culture. Trypan
blue dye exclusion tests showed that more than 96% of cells
were viable after a further 2 wk in culture in media containing
supplements. However, the efficiency of transformation as
assessed by the percentage of cells displaying the markers and
phenotypic appearances of differentiated cells varied considerably among the mesodermal- and endodermal-derived
lineages. The transformation efficiency into cells resembling
myoctyes, cardiomyocytes, hepatocytes, and pancreatic cells
was high compared with adipocytes (Table 2). Further, more
than one distinct phenotype was observed in cells grown with
supplements inducing hepatic, pancreatic, and cardiomyocytic
differentiation.
HLA Class I and II Antigens
Cells grown in pancreatic, cardiomyocytic, and hepatic
differentiation media were selected for HLA studies on the
basis of high transformation efficiency into these lineages and
potential therapeutic use in clinically important diseases, such
as diabetes, cardiomyopathies, and hepatic disorders. The mean
6 SEM percentage of freshly isolated cells with HLA-A, HLAB, and HLA-C (class IA) was 2.73 6 0.05, and HLA-DR,
HLA-DP, and HLA-DQ (class II) was 0.83 6 0.16 (n ¼ 4). The
percentage of cells with class IA and II antigens after 4 wk in
standard medium was similar to that of freshly isolated cells.
Upon differentiation into pancreatic and hepatic cells there was
a significant increase in the percentage of cells containing HLA
class IA antigens (P , 0.0001 compared to controls).
However, there were no significant changes in the percentage
of cells expressing HLA class II antigens following differentiation into hepatic and cardiomyocytic cells compared with
controls. Representative flow analysis diagrams and the
582
ILANCHERAN ET AL.
FIG. 2. Clonogenicity of hAECs. A) Clonal colonies formed by hAECs seeded at a density of 50 cells/cm2. B) The colonies stain positively for KRT7. C)
The number of cells per colony after 21 days in culture and cloning efficiency after 14 days in culture are shown graphically (mean 6 SEM, n ¼ 5). Data
were analyzed by the Kruskal-Wallis test and Dunn posthoc comparisons. Bars ¼ 100 lm.
STEM CELLS FROM FETAL MEMBRANES
583
FIG. 3. Transplantation of human amniotic
epithelial cells into SCID mouse testes. A)
Absence of immunostaining for KRT7 in a
tissue section taken from a SCID mouse
testis injected with saline. B) A cell monolayer composed of KRT7-positive cells
surrounding a murine blood/lymph vessel
indicated by arrowheads. C) Serial section
with the blood/lymph vessel shown in B
identified by positive immunostaining for
murine endothelial cell antigen PECAM1
(arrowheads). Bars ¼ 100 lm.
percentage of cells expressing HLA antigens are shown in
Figure 9.
DISCUSSION
In this study we have demonstrated that hAECs obtained
from normal term pregnancies 1) express markers usually
ascribed to stem cells, 2) are clonogenic, and 3) can be
differentiated in vitro into endodermal-, mesodermal-, and
ectodermal-derived lineages. We have also demonstrated that
hESCs do not form teratomas in vivo. Taken together, these
observations provide important and novel evidence for the
multipotentiality of hAECs. We have also shown that
undifferentiated cells are likely to possess a degree of immune
privilege but that this may alter following differentiation into
some lineages.
Although hAECs express mRNAs and/or proteins present in
hESCs and multipotent stem cells, the epithelial cells lining the
amnion do not transdifferentiate in vivo. It is possible that there
are endogenous substances and factors in the amniotic fluid
that suppress differentiation, and it may be worthwhile
identifying such factors. This could also assist in understanding
the pathways governing differentiation of adult and embryonic
stem cells.
Our finding that hAECs are clonogenic, forming large,
flattened, undifferentiated colonies containing several hundred
cells within a few weeks of culture is both novel and important.
Cells within the colonies stained positively for KRT7, an
epithelial cytoskeletal protein found in hAECs, indicating that
colonies did not arise from contaminating fibroblast or
mesenchymal cells [25]. Clonogenicity, the ability of a single
cell to form a clonal colony, is a key defining function that
demonstrates the self-renewal properties of stem cells [26].
That hAECs are clonogenic supports our stem cell-like marker
findings. Indeed, the cloning efficiency of the hAECs was
comparable to some hESC lines [27] and higher than some
adult epithelial stem cells, such as those derived from the
human endometrium [15]. However, while we have shown that
hAECs are clonogenic, we have yet to establish their long-term
self-renewal capabilities. This will be important to do in future
studies.
Teratoma formation is an important feature ascribed to
pluripotent hESCs. While the marker expression and clonogenecity suggest some similarity to hESCs, hAECs did not form
teratomas upon transplantation into the testes of SCID mice.
This confirms previous reports [9, 28] and studies using either
amniotic membranes to repair damaged ocular surfaces [29–31]
or amnion cells to improve congenital lysosomal storage
diseases [32] that also found that these cells did not form
tumors. Taken together, the current and previous studies would
suggest that amnion cells are not pluripotent-like hESCs. On
first inspection this may seem to be a limitation of hAECs but,
if proven, the absence of tumor formation following transplantation could actually be an advantage in future clinical
applications. Indeed, an understanding of the mechanisms
whereby hAECs do not form tumors, despite possessing similar
molecular machinery to hESCs, may be useful in directing
lineage-specific and tumor-suppressing differentiation of
hESCs. It is possible that the lack of telomerase activity in
hAECs [3] may contribute to tumor suppression in vivo.
However, equally importantly, the absence of teratoma
formation may indicate that hAECs are unable to differentiate
at all in vivo. It is therefore important to determine their
capacity to differentiate in vivo using appropriate animal
models of disease—the focus of our ongoing studies.
FIG. 4. Differentiation of hAECs into
lineages derived from the neuroectoderm.
The neural lineage markers NES, MAP2, and
GFAP were present in hAECs grown overnight. A few NES- and MAP2-positive cells
resembling neurons and numerous GFAPpositive astrocytes are seen after 4 wk of
culture with supplements inducing neuroectodermal differentiation. Bars ¼ 100 lm.
584
ILANCHERAN ET AL.
FIG. 5. Characterization of hAECs grown
with supplements inducing hepatocytic
differentiation. Cells grown with supplements produce ALB and HGF, as shown by
immunocytochemical staining and flow
analyses (A–D). Transmission electron micrograph of a cell grown in standard
medium for 4 wk (E), and cells grown with
supplements (F–H). F) Features of hepatocytes, including the double-layered nuclear
membrane (arrowheads), extensive rough
endoplasmic reticulum (RER) adjacent to
the nucleus (N), and a prominent electrondense nucleolus (n). The RER, smooth
endoplasmic reticulum (SER), and bile
canaliculi (BC) are shown at higher magnification in G and H. Scale bars ¼ 100 lm (A
and B) and 1 lm (E–H).
Nonetheless, in this study we investigated whether hAECs
could respond to environmental cues and differentiate into cells
derived from the three germ layers in vitro. We thought that
this was likely, as it has been reported that neural progenitor,
neuronal, and glial cell markers are present in the amniotic
epithelium [33, 34]. Consistent with these reports, we found
NES, GFAP, and the neuronal-specific marker MAP2 to be
abundant in freshly isolated hAECs. However, under conditions known to stimulate neural differentiation of hESCs, the
survival and differentiation of hAECs into cells resembling
neurons was poor. Interestingly, in vivo studies in which
hAECs have been injected into injured rodent brain or bonnet
monkeys with spinal cord injury did not show conclusive
FIG. 6. Characterization of hAECs grown
with supplements inducing pancreatic cell
differentiation. AMY2B, a pancreatic exocrine alpha cell marker, is expressed in
hAECs (A). Positive immunostaining for the
hormone glucagon (GCG) is seen in differentiated cells (B). Flow analysis confirms
that the majority of differentiated cells
produce GCG (C). Transmission electron
micrographs of hAECs cultured with supplements show morphologic characteristics
of exocrine acinar beta cells. Beta-granules
(B), zymogen granules (Z), and rough
endoplasmic reticulum (RER) are seen in D,
whereas condensing vacuoles (CV) and
distended Golgi saccule (S) are seen in E.
Bar ¼ 100 lm (B) and 1 lm (D and E).
evidence of differentiation, but that immunosuppressive and/or
neurotrophic factors produced by the cells seem to aid repair
[35–38]. The roles played by neurotrophic factors and
neurotransmitters catecholamines acetylcholine, norepinephrine, and dopamine [39] produced by hAECs during gestation
in fetoplacental growth and function are not widely understood.
However, it is possible that they may hinder differentiation
along neural lineages, at least in vitro.
On the other hand, this is the first report that native hAECs
can differentiate into cells with a phenotype and markers
characteristic of mesodermal-derived myocytes, osteocytes,
and adipocytes. Human AECs also formed elongated TNNTpositive cells resembling cardiomyocytes, consistent with a
STEM CELLS FROM FETAL MEMBRANES
585
FIG. 7. Differentiation of hAECs into
lineages derived from the mesoderm. Human AECs grown overnight (top panel) or 4
wk in control medium (middle panel)
lacked markers characteristic of differentiated cells, whereas hAECs grown with
supplements (bottom panel) displayed morphologic changes and markers. Myocytic
cells contained smooth muscle alpha-actin
(ACTA2). Arrowheads in the lower right
panel show lipid granules. Bars ¼ 100 lm.
previous report describing a-actin staining cardiomyocyticlike
cells derived from passaged term hAECs [9]. Interestingly,
freshly isolated cells and those grown under standard
conditions also expressed the mRNAs of genes that are
important for cardiomyoctyic lineage specification, such as
GATA4 and function, including ANP, MYL7, CACNA1C, and
KCND3. However, ultrastructrual analysis revealed that
features consistent with relatively mature cardiomyocytes,
such as myofilaments, myofibrils, H bands and T tubules
[40], were present only in differentiated cells. These findings
demonstrate that hAECs can attain subcellular specialization in
response to environmental cues. However, differentiated cells
did not contract spontaneously under the culture conditions
used. This may be due to the lack of myofibrillar bundle
organization into densely packed sarcomeres, as is seen in
terminally differentiated cardiomyocytes [40]. It would be
worthwhile investigating whether hAECs can attain terminal
differentiation and their electrophysiologic properties.
We were also able to differentiate hAECs into endodermalderived hepatocytes and pancreatic cells, confirming previous
reports [9, 41]. We have extended those reports by showing
that the differentiated cells have high long-term survival rates
in culture. We have also further characterized the cells,
showing that the hepatic cells produce HGF, a key growth
factor, and that at the ultrastructural level they had the key
morphologic features of normal adult hepatocytes. Similarly,
while we could demonstrate that the hAEC-derived pancreatic
cells produced GCG, we could not coax them to secrete insulin.
Interestingly, at the ultrastructural level the pancreatic cells
looked like exocrine beta-acinar cells consistent with an
FIG. 8. Characterization of hAECs grown
with supplements inducing cardiomyocytic
differentiation. Genes expressed in cardiomyocytes are found in hAECs (A). TNNT, a
marker of differentiated cardiomyocytes, is
abundant in differentiated cells (B and C).
Transmission electron micrographs of
hAECs cultured with supplements show T
tubules (T), myofilaments (MY), myofibrils
(MF), and H bands (H) characteristic of
relatively mature cardiomyocytes (D–F). Bar
¼ 100 lm (B) and 1 lm (D–F).
586
ILANCHERAN ET AL.
FIG. 9. HLAs present in human amniotic
epithelial cells. Representative flow analysis
diagrams of freshly isolated cells and cells
grown with or without supplements inducing differentiation for 4 wk (A). Graphic
representation of the percentage of cells
containing HLA before and after differentiation (mean 6 SEM, n ¼ 4). Data were
analyzed by the Kruskal-Wallis test followed by Dunn posthoc test (B).
inability to secrete insulin. It is also important to note that while
a high proportion of cells displayed markers characteristic of
specialized cells, we observed secondary phenotypes among
some of these lineages. It would be important to characterize
these secondary phenotypes. However, given these promising
preliminary findings, in vivo animal studies investigating
migration, engraftment, and tissue-specific differentiation of
transplanted hAECs would seem warranted. Indeed, mixed
populations of membrane-derived cells have been shown to
synthesize migration and adhesion molecules and proteases
that could facilitate their transmigration and homing to injured
sites in vivo [8].
Native hAECs are thought to have low levels of
polymorphic HLA-A, HLA-B, and HLA-C and negligible
levels of HLA-DR mRNA expression [42]. This is important in
terms of potential clinical applications, as native hAECs would
be expected to have a reduced risk of rejection upon
allotransplantation. Indeed, mixed populations of fetal membrane-derived cells have been found to suppress mixed
allogeneic lymphocyte reactions [8]. Our findings show that
STEM CELLS FROM FETAL MEMBRANES
only a very small fraction of native cells or those maintained
long term under control conditions contain these antigens.
Importantly, however, HLA-A, HLA-B, and HLA-C antigens
are clearly present in significant proportions of hAEC-derived
hepatocytes and pancreatic cells but not cardiomyoctyes.
Understanding the factors stimulating/suppressing HLA antigenicity would be important if these cells were to be useful
clinically. Human AECs also secrete immunosuppressive
factors [37, 43] that could facilitate successful engraftment
and secrete growth factors and cytokines that may stimulate the
differentiation of endogenous stem cell populations and/or
assist in wound healing and repair in vivo [43–48]. Given the
abundance of human term amnion tissue, minimal ethical and
legal issues associated with its usage and our findings that term
hAECs possess stem cell-like characteristics and have
considerable multilineage differentiation potential warrants
further investigation into their differentiation and functional
potential in vivo.
15.
16.
17.
18.
19.
20.
21.
ACKNOWLEDGMENTS
22.
We thank the patients and staff of Monash Medical Center, Nicki Sam
for tissue collection, Budi Marjono and Claire Walker for their assistance
in maintaining cultures, Paul Hutchinson and James Ngui for their
assistance with the flow analyses, and Anna Friedhuber for the TEM and
ultrastructural analyses.
23.
REFERENCES
25.
1. Pera MF, Reubinoff B, Trounson A. Human embryonic stem cells. J Cell
Sci 2000; 113:5–10.
2. Tsai MS, Hwang SM, Tsai YL, Cheng FC, Lee JL, Chang YJ. Clonal
amniotic fluid-derived stem cells express characteristics of both mesenchymal and neural stem cells. Biol Reprod 2006; 74:545–551.
3. Mosquera A, Fernandez JL, Campos A, Goyanes VJ, Ramiro-Diaz J,
Gosalvez J. Simultaneous decrease of telomere length and telomerase
activity with ageing of human amniotic fluid cells. J Med Genet 1999; 36:
494–496.
4. In ’t Anker PS, Scherjon SA, Kleijburg-van der Keur C, Noort WA, Claas
FH, Willemze R, Fibbe WE, Kanhai HH. Amniotic fluid as a novel source
of mesenchymal stem cells for therapeutic transplantation. Blood 2003;
102:1548–1549.
5. Prusa AR, Marton E, Rosner M, Bettelheim D, Lubec G, Pollack A,
Bernaschek G, Hengstschlager M. Neurogenic cells in human amniotic
fluid. Am J Obstet Gynecol 2004; 191:309–314.
6. In ’t Anker PS, Scherjon SA, Kleijburg-van der Keur C, de Groot-Swings
GM, Claas FH, Fibbe WE, Kanhai HH. Isolation of mesenchymal stem
cells of fetal or maternal origin from human placenta. Stem Cells 2004; 22:
1338–1345.
7. Blackwell SC, Abundis MG, Nehra PC. Five-year experience with
midtrimester amniocentesis performed by a single group of obstetriciansgynecologists at a community hospital. Am J Obstet Gynecol 2002; 186:
1130–1132.
8. Bailo M, Soncini M, Vertua E, Signoroni PB, Sanzone S, Lombardi G,
Arienti D, Calamani F, Zatti D, Paul P, Albertini A, Zorzi F, et al.
Engraftment potential of human amnion and chorion cells derived from
term placenta. Transplantation 2004; 78:1439–1448.
9. Miki T, Lehmann T, Cai H, Stolz DB, Strom SC. Stem cell characteristics
of amniotic epithelial cells. Stem Cells 2005; 23:1549–1559.
10. Lefebvre S, Adrian F, Moreau P, Gourand L, Dausset J, Berrih-Aknin S,
Carosella ED, Paul P. Modulation of HLA-G expression in human thymic
and amniotic epithelial cells. Hum Immunol 2000; 61:1095–1101.
11. Terada S, Matsuura K, Enosawa S, Miki M, Hoshika A, Suzuki S,
Sakuragawa N. Inducing proliferation of human amniotic epithelial (HAE)
cells for cell therapy. Cell Transplant 2000; 9:701–704.
12. Lee JB, Song JM, Lee JE, Park JH, Kim SJ, Kang SM, Kwon JN, Kim
MK, Roh SI, Yoon HS. Available human feeder cells for the maintenance
of human embryonic stem cells. Reproduction 2004; 128:727–735.
13. Andrews PW, Casper J, Damjanov I, Duggan-Keen M, Giwercman A,
Hata J, von Keitz A, Looijenga LH, Millán JL, Oosterhuis JW, Pera M,
Sawada M, et al. Comparative analysis of cell surface antigens expressed
by cell lines derived from human germ cell tumours. Int J Cancer 1996;
66:806–816.
14. Herszfeld D, Wolvetang E, Langton-Bunker E, Chung TL, Filipczyk AA,
24.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
587
Houssami S, Jamshidi P, Koh K, Laslett AL, Michalska A, Nguyen L,
Reubinoff BE, et al. CD30 is a survival factor and a biomarker for
transformed human pluripotent stem cells. Nat Biotech 2006; 24:351–357.
Chan RW, Schwab KE, Gargett CE. Clonogenicity of human endometrial
epithelial and stromal cells. Biol Reprod 2004; 70:1738–1750.
Zhao P, Ise H, Hongo M, Ota M, Konishi I, Nikaido T. Human amniotic
mesenchymal cells have some characteristics of cardiomyocytes. Transplantation 2005; 79:528–535.
Lavon N, Yanuka O, Benvenisty N. The effect of overexpression of Pdx1
and Foxa2 on the differentiation of human embryonic stem cells into
pancreatic cells. Stem Cells 2006; 24:1923–1930.
Rubattu S, Volpe M. The atrial natriuretic peptide: a changing view. J
Hypertens 2001; 19:1923–1931.
Moss RL, Fitzsimons DP. Myosin light chain 2 into the mainstream of
cardiac development and contractility. Circ Res 2006; 99:225–227.
Seisenberger C, Specht V, Welling A, Platzer J, Pfeifer A, Kuhbandner S,
Striessnig J, Klugbauer N, Feil R, Hofmann F. Functional embryonic
cardiomyocytes after disruption of the L-type alpha1C (Cav1.2) calcium
channel gene in the mouse. J Biol Chem 2000; 275:39193–39199.
Zicha S, Xiao L, Stafford S, Cha TJ, Han W, Varro A, Nattel S.
Transmural expression of transient outward potassium current subunits in
normal and failing canine and human hearts. J Physiol 2004; 561:735–748.
Ulrich AB, Schmied BM, Standop J, Schneider MB, Pour PM. Pancreatic
cell lines: a review. Pancreas 2002; 24:111–120.
Hussain SZ, Strom SC, Kirby MR, Burns S, Langemeijer S, Ueda T, Hsieh
M, Tisdale JF. Side population cells derived from adult human liver
generate hepatocyte-like cells in vitro. Dig Dis Sci 2005; 50:1755–1763.
Yuan S, Rosenberg L, Paraskevas S, Agapitos D, Duguid WP.
Transdifferentiation of human islets to pancreatic ductal cells in collagen
matrix culture. Differentiation 1996; 61:67–75.
Blaschitz A, Weiss U, Dohr G, Desoye G. Antibody reaction patterns in
first trimester placenta: implications for trophoblast isolation and purity
screening. Placenta 2000; 21:733–741.
Robey PG. Stem cells near the century mark. J Clin Invest 2000; 105:
1489–1491.
Ware CB, Nelson AM, Blau CA. A comparison of NIH-approved human
ES cell lines. Stem Cells 2006; 24:2677–2684.
Miyamoto K, Hayashi K, Suzuki T, Ichihara S, Yamada T, Kano Y,
Yamabe T, Ito Y. Human placenta feeder layers support undifferentiated
growth of primate embryonic stem cells. Stem Cells 2004; 22:433–440.
Tseng SC, Prabhasawat P, Barton K, Gray T, Meller D. Amniotic
membrane transplantation with or without limbal allografts for corneal
surface reconstruction in patients with limbal stem cell deficiency. Arch
Ophthalmol 1998; 116:431–441.
Ucakhan OO, Koklu G, Firat E. Nonpreserved human amniotic membrane
transplantation in acute and chronic chemical eye injuries. Cornea 2002;
21:169–172.
Solomon A, Meller D, Prabhasawat P, John T, Espana EM, Steuhl KP,
Tseng SC. Amniotic membrane grafts for nontraumatic corneal perforations, descemetoceles, and deep ulcers. Ophthalmology 2002; 109:694–
703.
Kosuga M, Takahashi S, Sasaki K, Enosawa S, Li XK, Okuyama S, Fujino
M, Suzuki S, Yamada M, Matsuo N, Sakuragawa N, Okuyama T.
Phenotype correction in murine mucopolysaccharidosis type VII by
transplantation of human amniotic epithelial cells after adenovirusmediated gene transfer. Cell Transplant 2000; 9:687–692.
Sakuragawa N, Thangavel R, Mizuguchi M, Hirasawa M, Kamo I.
Expression of markers for both neuronal and glial cells in human amniotic
epithelial cells. Neurosci Lett 1996; 209:9–12.
Ishii T, Ohsugi K, Nakamura S, Sato K, Hashimoto M, Mikoshiba K,
Sakuragawa N. Gene expression of oligodendrocyte markers in human
amniotic epithelial cells using neural cell-type-specific expression system.
Neurosci Lett 1999; 268:131–134.
Okawa H, Okuda O, Arai H, Sakuragawa N, Sato K. Amniotic epithelial
cells transform into neuron-like cells in the ischemic brain. Neuroreport
2001; 12:4003–4007.
Sankar V, Muthusamy R. Role of human amniotic epithelial cell
transplantation in spinal cord injury repair research. Neuroscience 2003;
118:11–17.
Li H, Niederkorn JY, Neelam S, Mayhew E, Word RA, McCulley JP,
Alizadeh H. Immunosuppressive factors secreted by human amniotic
epithelial cells. Invest Ophthalmol Vis Sci 2005; 46:900–907.
Sakuragawa N, Elwan MA, Uchida S, Fujii T, Kawashima K. Nonneuronal neurotransmitters and neurotrophic factors in amniotic epithelial
cells: expression and function in humans and monkey. Jpn J Pharmacol
2001; 85:20–23.
Elwan MA, Sakuragawa N. Evidence for synthesis and release of
588
40.
41.
42.
43.
ILANCHERAN ET AL.
catecholamines by human amniotic epithelial cells. Neuroreport 1997; 8:
3435–3438.
Baharvand H, Piryaei A, Rohani R, Taei A, Heidari MH, Hosseini A.
Ultrastructural comparison of developing mouse embryonic stem cell and
in vivo-derived cardiomyocytes. Cell Biol Int 2006; 30:800–807.
Takashima S, Ise H, Zhao P, Akaike T, Nikaido T. Human amniotic
epithelial cells possess hepatocyte-like characteristics and functions. Cell
Struct Funct 2004; 29:73–84.
Kubo M, Sonoda Y, Muramatsu R, Usui M. Immunogenicity of human
amniotic membrane in experimental xenotransplantation. Invest Ophthalmol Vis Sci 2001; 42:1539–1546.
Kamiya K, Wang M, Uchida S, Amano S, Oshika T, Sakuragawa N, Hori
J. Topical application of culture supernatant from human amniotic
epithelial cells suppresses inflammatory reactions in cornea. Exp Eye
Res 2005; 80:671–679.
44. Uchida S, Inanaga Y, Kobayashi M, Hurukawa S, Araie M, Sakuragawa
N. Neurotrophic function of conditioned medium from human amniotic
epithelial cells. J Neurosci Res 2000; 62:585–590.
45. Sawhney CP. Amniotic membrane as a biological dressing in the
management of burns. Burns 1989; 15:339–342.
46. Sippel KC, Ma JJ, Foster CS. Amniotic membrane surgery. Curr Opin
Ophthalmol 2001; 12:269–281.
47. Keelan JA, Sato T, Mitchell MD. Interleukin (IL)-6 and IL-8 production
by human amnion: regulation by cytokines, growth factors, glucocorticoids, phorbol esters, and bacterial lipopolysaccharide. Biol Reprod 1997;
57:1438–1444.
48. Hao Y, Ma DH, Hwang DG, Kim WS, Zhang F. Identification of
antiangiogenic and antiinflammatory proteins in human amniotic membrane. Cornea 2000; 19:348–352.