| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
OPEN ACCESS ARTICLE
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TISSUE-SPECIFIC STEM CELLS |
aCentro di Ricerca E. Menni, Fondazione Poliambulanza, Istituto Ospedaliero, Brescia, Italy;
bDepartment of Histology, Embryology and Applied Biology, University of Bologna, Bologna, Italy;
cDepartment of Obstetrics, University Hospital Zurich, Zurich, Switzerland;
dDepartment of Internal Medicine II, University Clinic of Tübingen, Tübingen, Germany;
eRed Cross Blood Transfusion Service of Upper Austria/Ludwig Boltzmann Institute for Experimental and Clinical Traumatology, Linz-Vienna, Austria;
fDepartment of Cell Biology, Institute of Basic Medical Sciences, Beijing, China;
gDepartment of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania, USA;
hInstitute of Medical Science, University of Tokyo, Tokyo, Japan;
iDepartment of Regenerative Medicine, University of Toyama Faculty of Medicine, Toyama, Japan;
jDepartment of Obstetrics and Gynaecology, University of Berne, Berne, Switzerland;
kDepartment of Anatomy, University of Madras, Chennai, India;
lDepartment of Regenerative Medicine, Kitasato University, Kanagawa, Japan
Key Words. Human placenta • Fetal membranes • Amnion • Chorion • Mesenchymal stromal cells • Fetal tolerance
Correspondence: Ornella Parolini, Ph.D., Centro di Ricerca E. Menni, Fondazione Poliambulanza, Istituto Ospedaliero, Via Bissolati 57, 25124 Brescia, Italy. Telephone: 39-030-2455-754; Fax: 39-030-2455-704; e-mail: ornella.parolini{at}tin.it
Received July 24, 2007;
accepted for publication October 18, 2007.
First published online in STEM CELLS EXPRESS November 1, 2007.
 |
ABSTRACT |
|---|
Disclosure of potential conflicts of interest is found at the end of this article.
|
INTRODUCTION |
|---|
From the data presented by the participants, the following points were evident.
Minimal criteria for defining hAMSC and hCMSC are as follows:
|
|
THE PLACENTA |
|---|
|
Fetal membranes continue from the edge of the placenta and enclose the amniotic fluid and the fetus. The amnion is a thin, avascular membrane composed of an epithelial layer and an outer layer of connective tissue, and is contiguous, over the umbilical cord, with the fetal skin. The amniotic epithelium (AE) is an uninterrupted, single layer of flat, cuboidal and columnar epithelial cells in contact with amniotic fluid. It is attached to a distinct basal lamina that is, in turn, connected to the amniotic mesoderm (AM) (Fig. 2). In the amniotic mesoderm closest to the epithelium, an acellular compact layer is distinguishable, composed of collagens I and III and fibronectin. Deeper in the AM, a network of dispersed fibroblast-like mesenchymal cells and rare macrophages are observed. Very recently, it has been reported that the mesenchymal layer of amnion indeed contains two subfractions, one having a mesenchymal phenotype, which is referred to throughout this review as amniotic mesenchymal stromal cells, and the second containing monocyte-like cells [4].
|
Embryological Development
In humans, by days 6–7 after fertilization (during the implantation window), the blastocyst implants and placenta development begins. At this stage, the blastocyst is flattened and composed of an outer wall (trophoblast) that surrounds the blastocystic cavity. A small group of larger cells, the inner cell mass, is apposed to the inner surface of the trophoblastic vesicle. The embryo, umbilical cord, and amniotic epithelium are derived from the inner cell mass. As the blastocyst adheres to the endometrium, invading trophoblasts erode the decidua, facilitating implantation of the blastocyst. By 8–9 days after fertilization, trophoblastic cells at the implanting pole of the blastocyst proliferate robustly, forming a bilayered trophoblast. The outer of the two layers becomes the syncytiotrophoblast by fusion of neighboring trophoblast cells, whereas the inner cells (cytotrophoblast) remain temporally unfused.
The proliferating cytotrophoblasts and the syncytiotrophoblasts give rise to a system of trabeculae intermingled with hematic lacunae. From these trabeculae are generated the primordial villi that are distributed over the entire periphery of the chorionic membrane. Villi in contact with the decidua basalis proliferate to form the leafy chorion or chorion frondosum, whereas villi in contact with the decidua capsularis degenerate into the chorion leave.
At day 8–9 after fertilization, the inner cell mass differentiates into two layers: the epiblast and the hypoblast. Subsequently, from the epiblast, small cells that later constitute the amniotic epithelium appear between the trophoblast and the embryonic disc and enclose a space that will become the amniotic cavity. On the opposite side, between the hypoblast and cytotrophoblast, the exocoelomic membrane and its cavity modify to form the yolk sac. The extraembryonic mesoderm arranges into a connective tissue that surrounds the yolk sac and amniotic cavity, giving rise to amniotic and chorionic mesoderm. Gastrulation, the process through which the bilaminar disc differentiates into the three germ layers (ectoderm, mesoderm, and endoderm) and develops a defined form, with a midline and cranio-caudal, right-left, and dorsal-ventral body axes, occurs during the 3rd week after fertilization [6, 7].
|
FETAL PLACENTA TISSUE CELL POPULATIONS |
|---|
6 and β1, c-met (hepatocyte growth factor receptor), stage-specific embryonic antigens (SSEAs) 3 and 4, and tumor rejection antigens 1-60 and 1-81 [11, 14]. Surface markers thought to be absent on hAEC include SSEA-1, CD34, and CD133, whereas other markers, such as CD117 (c-kit) and CCR4 (CC chemokine receptor), are either negative or may be expressed on some cells at very low levels. Although initial cell isolates express very low levels of CD90 (Thy-1), the expression of this antigen increases rapidly in culture [11, 14]. Additional surface markers are presented in Table 2.
|
-fetoprotein (AFP) and that Alb- and AFP-positive hepatocyte-like cells could be identified integrated into hepatic parenchyma following transplantation of hAEC into the livers of severe combined immunodeficiency (SCID) mice [21]. The hepatic potential of hAEC was confirmed and extended [11, 22, 23], whereby in addition to Alb and AFP production, other hepatic functions, such as glycogen storage and expression of liver-enriched transcription factors, such as hepatocyte nuclear factor (HNF) 3
and HNF4, CCAAT/enhancer-binding protein (CEBP and β), and several of the drug metabolizing genes (cytochrome P450) could be demonstrated. The wide range of hepatic genes and functions identified in hAEC suggests that these cells may be useful for liver-directed cell therapy. Differentiation of hAEC to another endodermal tissue, pancreas, was also reported. Wei et al. [24] cultured hAEC for 2–4 weeks in the presence of nicotinamide to induce pancreatic differentiation. Subsequent transplantation of the insulin-expressing hAEC corrected the hyperglycemia of streptozotocin-induced diabetic mice. In the same setting, hAMSC were ineffective, suggesting that hAEC, but not hAMSC, were capable of acquiring a β-cell fate.
The studies reviewed above indicate that hAEC are unique cells with many stem cell characteristics. Cell types from all three germ layers were produced in vitro. There is currently strong in vitro and in vivo evidence of neural, pancreatic, and hepatic differentiation of hAEC.
Mesenchymal Stromal Cells from Amnion and Chorion: hAMSC and hCMSC
hAMSC and hCSMC are thought to be derived from extraembryonic mesoderm [7]. Extensive phenotypical and functional characterization is available on hAMSC [25–29], whereas there are few reports of investigations on hCMSC [28, 30, 31]. The available details are summarized in Tables 3 and 4.
|
|
hCMSC are isolated from both first- and third-trimester chorion after mechanical and enzymatic removal of the trophoblastic layer with dispase. Chorionic mesodermal tissue is then digested with collagenase [28] or collagenase plus DNase [31]. Mesenchymal cells have also been isolated from chorionic fetal villi through explant culture, although maternal contamination is more likely [30, 31, 34].
Both hAMSC and hCMSC adhere and proliferate on tissue culture plastic and can be kept until passages 5–10. Reports suggest that hAMSC proliferation slows beyond passage 2, although both first-trimester hAMSC and hCMSC proliferate better than third-trimester cells [28]. Theoretically, term amnion may yield up to 5 x 108 hAMSC [35]; however, in practice, yields are typically 4 million hAMSC per 100 cm2 of starting material, with a fourfold expansion after 1 month (two passages) (G. Bilic, S. Zeisberger, and A.H. Zisch, unpublished data).
Phase micrographs (Fig. 3) show the distinct morphology of cultured hAMSC, hCMSC, and hAEC. Expression of CD49d (4 integrin) on hAMSC distinguishes them from hAEC [29]. In culture, neither vimentin nor cytokeratin 18 expression is specific for hAMSC or hAEC, respectively. Transmission electron microscopy of hAMSC shows mesenchymal and epithelial characteristics. This hybrid phenotype is interpreted as a sign of multipotentiality and is not found in hCMSC, which are more primitive and metabolically quiescent. Specifically, compared with hAMSC, transmission electron microscopy of hCMSC show a simpler cytoplasmic organization. The most relevant features include the presence of stacks of rough endoplasmic reticulum cisternae, dispersed mitochondria and glycogen lakes, whereas features of higher specialization, such as presence of assembled contractile filaments, prominence of endocytotic traffic, and junctional communications, are lacking. Overall, the ultrastructural characteristics of hCMSC resemble those found in the hematopoietic progenitors and in the blue small round cell tumors (e.g., Ewing sarcomas), suggesting their position at the higher levels of the stem cell hierarchy [36]. The surface marker profile of cultured hAMSC, hCMSC, and mesenchymal stromal cells (MSC) from adult bone marrow are similar. All express typical mesenchymal markers (Table 1) but are negative for hematopoietic (CD34 and CD45) and monocytic markers (CD14) [26, 28, 29, 31]. Surface expression of SSEA-3 and SSEA-4 and RNA for OCT-4 has been reported [24, 29, 35, 37] (G. Bilic, S. Zeisberger, and A.H. Zisch, unpublished data). However, immunofluorescence staining of mesenchyme tissue did not detect SSEA-3 or SSEA-4 [39]. Both first- and third-trimester hAMSC and hCMSC express low levels of HLA-A,B,C but not HLA-DR [28, 29], indicating an immunoprivileged status.
|
Human amniotic and chorionic cells successfully and persistently engraft in multiple organs and tissues in vivo. Human chimerism detection in brain, lung, bone marrow, thymus, spleen, kidney, and liver after either intraperitoneal or intravenous transplantation of human amnion and chorion cells into neonatal swine and rats was indeed indicative of an active migration consistent with the expression of adhesion and migration molecules (L-selectin, VLA-5, CD29, and P-selectin ligand 1), as well as cellular matrix proteinases (MMP-2 and MMP-9) [41].
|
IMMUNOLOGY OF THE PLACENTA |
|---|
In particular, different explanations have been proposed regarding the regulation of maternal T-cell proliferation at the fetal maternal interface. It has been shown that some cells of the syncytiotrophoblast express the tryptophan catabolizing enzyme indoleamine 2,3-dioxygenase (IDO), resulting in the depletion of tryptophan and inhibition of T-cell proliferation, which may provide protection of the fetus from maternal T-cells [43, 44]. However, in IDO knockout mice, fetus rejection was not observed [45]. Therefore, even though IDO activity has been hypothesized as a key mechanism for protecting the allogenic fetus, it is not the sole determinant since other mechanisms, perhaps redundant in normal mice, can compensate for the loss of IDO activity during gestation.
Furthermore, soluble HLA-G molecules produced by placenta induce apoptosis of activated CD8+ T-cells [46] and inhibit CD4+ T-cell proliferation [46]. Trophoblast cells expressing HLA-G may also inhibit natural killer cells that could induce fetal rejection [48]. In addition, trophoblast cells that express Fas ligand (CD95L) have been demonstrated to induce apoptosis of maternal Fas (CD95)-expressing lymphocytes [49], therefore representing another possible mechanism for contributing to the maintenance of fetomaternal tolerance. The mechanisms discussed above are not definitive, and it is very likely that they all play a role in the complex phenomenon of fetal maternal tolerance.
|
ROLE OF FETAL MEMBRANES IN TOLERANCE |
|---|
|
PLACENTA AS A POTENT NICHE SOURCE FOR HEMATOPOIETIC STEM/PROGENITOR CELLS: INSIGHTS FROM THE MOUSE PLACENTA |
|---|
Mouse placenta serves as a source and a functional niche for HSC [74]. Prefusion allantois and chorion are potent sources for hematopoietic progenitors. Allantois and chorion, isolated prior to the union between the three major circulatory systems (umbilical cord, yolk sac, and cardiovascular) in the conceptus, express Runx1, a key transcriptional factor critical for HSC ontogeny, and CD41, a hallmark for initiation of definitive hematopoiesis [75], and also generate myeloid and definitive erythroid lineages following explant culture [76]. Furthermore, hemogenic potential and Runx1 expression are independent of the union of the tissues, reflecting their intrinsic hematopoietic property. The identification of critical niche components for HSCs from placenta may be useful for ex vivo HSC expansion.
|
PRECLINICAL STUDIES IN ANIMAL MODELS |
|---|
, and C/EBP, HNF1, HNF4, pregnane x receptor, and constitutive androstane receptor, and differentiated liver genes, including albumin, 1-antitrypsin (A1AT), glucose-6-phosphatase, carbamoyl phosphate synthase I (CPS-I), glutamine synthase, phosphoenolpyruvate carboxykinase, tyrosine aminotransferase, transthyretin, and the drug metabolizing genes CYP1A1, 1A2, 2A6, 2B6, 2C8, 2C9, 2D6 2E1, 3A4, 7, and 7A1 [11, 14, 22, 23]. In addition to quantification of RNA, drug metabolism dependent on CYP1A and CYP3A enzymatic activity has been demonstrated [11, 78]. In unpublished work, the list of hepatic genes identified in hAE include transport proteins, including P-glycoprotein, multidrug resistance protein 1, ABCG2, the bile salt export pump, uridine diphosphate glucuronosyl transferase 1A1 (UGT1A1), and ornithine transcarbamylase (OTC) (Miki T, Marongiu F, Strom S, unpublished data). Following transplantation into the liver, successful engraftment and survival of human [21, 78] or rat amniotic epithelial cells (AEC) [79] has been demonstrated. Recent unpublished work extends these observations to more than 6 months. Human A1AT could be detected by Western blot in mouse serum [14], clearly indicating that hAEC perform this important hepatic function in vivo. Human albumin was detected in the sera and peritoneal fluid of SCID mice who received peritoneal implants of human amniotic membrane [22]. At this time, there are no reports of the characterization of hepatic gene expression in long-term recipients of AE transplants or that the transplantation of AEC can support animals with acute or long-term defects in liver function. These important preclinical studies will be needed before the therapeutic potential of hAE can be fully assessed. At least 30 genes known to be expressed in mature human liver are expressed in hAEC directed toward hepatic differentiation in culture, suggesting that transplantation of hAE-derived hepatocytes could be an effective therapy for liver disease. It is significant that among the hepatic genes expressed are A1AT, OTC, CPS-I, and UGT1A1. Mutations in these genes cause metabolic liver diseases that are currently corrected by liver transplantation. Perhaps hAEC transplants could provide a novel cell therapy for patients suffering liver or lung disease from A1AT deficiency, or those at risk for neurological effects due to the inability to metabolize and excrete ammonia (OTC and CPS-I) or bilirubin (UGT1A1).
Cardiac Repair
Myocardial infarction, ischemia, and stroke are important consequences of end-stage occlusive vascular disease. Present-day therapies are inadequate and palliative, so stem cell therapy has been investigated. Coculture experiments with neonatal rat heart explants have confirmed that hAMSC integrate into cardiac tissues and differentiate into cardiomyocyte-like cells. After transplantation into myocardial infarcts in rat hearts, hAMSC survived for 2 months and differentiated into cardiomyocyte-like cells [37].
A mixed ester of hyaluronan and butyric and retinoic acids (HBR) was reported to promote cardiogenic/vasculogenic differentiation of human amniochorionic (AC)-derived cells. HBR enhanced the expression of cardiomyogenic genes (GATA4, NKX 2.5) and proteins (sarcomeric myosin heavy chain and -sarcomeric actinin) but not skeletal muscle myosin D (MyoD) or neurogenic features (Neurogenin). Cells treated with HBR express both cardiac and endothelial markers, such as von Willebrand factor (vWF), and enhance cardiac repair in infarcted rat heart. AC transplants increase capillary density at the infarct border, normalize left ventricular function, and decrease scar formation. Transplantation of HBR-preconditioned AC-derived cells further enhanced capillary density and the yield of human vWF-expressing cells, also decreasing the infarct size. Some engrafted, HBR-pretreated AC-derived cells were also positive for connexin 43 and cardiac troponin I. These improvements may be related to the local or paracrine secretion of angiogenic, antiapoptotic, and mitogenic factors following human MSC transplantation, in addition to the differentiation of AC to vascular cells. These "trophic" effects may provide a major contribution to the therapeutic potential of MSCs for myocardial infarction and dilated cardiomyopathy. These findings provide an innovative approach to cell therapy of heart failure, whereby AC secrete trophic factors and, under the influence of HBR, may show enhanced cardiac and vascular differentiation [35, 80].
Placenta Derived Stem Cells for Treating Neurological Disorders
Human AEC have shown particular potential for treating central nervous system disorders. Since the discovery that hAEC have stem cell properties [11], express neural and glial markers and neural-specific proteins, and also have the capacity to produce and secrete neurotransmitters [9, 17], cell therapy with these cells has been considered [9, 16]. Successful transplants of hAEC into caudate nucleus [19, 20, 81], hippocampus [82], and spinal cord [66] have been reported. Transplantation of hAEC in a rat model of Parkinson's disease reversed the condition and prevented neuronal death [19, 20]. When hAEC were transplanted into ischemic hippocampus, they differentiated into "neuron-like" cells [82]. Following transplantation into the transected spinal cord of monkeys, hAEC aided a robust regeneration of host axons and prevented death of axotomized neurons of the spinal cord [66]. Transplantation of hAEC into the lesioned areas of a contusion model of spinal cord injury in rats was performed without immunosuppression. Cells survived up to 120 days with no evidence of inflammation or rejection. Animals showed gradual functional improvement using the Basso, Beattie, and Brensnahan (BBB) locomotor rating scale and ultimately reached a score of 19, just 2 points below normal animals. Improvement was also observed in lesion control animals; however, improvement was faster during acute and subacute phases of recovery in transplant recipients. Wu et al. reported similar findings [83]. Early improvement in the BBB scale is thought to indicate that hAEC provide neuroprotection [84]. In narrow-beam and inclined-plane tests, the performance of hAEC-transplanted animals was significantly improved compared with lesion control animals. These tests measure corticospinal tract [85] and rubrospinal tract [86, 87] function, respectively. Improvement following hAEC transplants indicates improvement of pyramidal (corticospinal) and extrapyramidal (brain stem) systems of motor tracts controlling locomotion. hAEC secrete neurotrophic factors [88], whereas medium conditioned by hAEC has been shown to be neurotrophic for E18 rat cortical cells. Novel EGF-like neurotrophic factors were thought to mediate this effect [89]. hAEC-conditioned medium also supported survival of E10 chicken neural retinal cells, which were otherwise dependent on fibroblast growth factor-2 (FGF-2) [90, 91]. Although FGF-2 and EGF were not detected in media by immunoblotting, FGF-2 and EGF gene and protein expression was reported in cryopreserved hAEC [89, 92]. Recently, MSCs were found to produce "neuron-like" cells, but their function is yet to be proven [93].
In conclusion, hAEC transplants produce beneficial results in animal models of spinal cord injury. They were found to exhibit neuroprotection in acute phases of injury and facilitate regeneration of long tracts in long-term phases of recovery, as measured by behavioral assessment. The beneficial effects may be mediated through the secretion of novel neurotrophic factors.
Cell Tracking
In preclinical studies, tracking of transplanted cells is essential. Using cell labeling together with imaging, cells can be traced noninvasively [94–102]. Stem cells from different sources have been labeled using radionuclides, magnetic nanoparticles, or reporter genes, in both preclinical and clinical studies [95, 103, 104]. In contrast to other imaging techniques, luminescence imaging detects live cells, since the reporter gene, luciferase, generates photon emission only in the presence of ATP, luciferin, and oxygen [105]. Reporter gene transfection protocols established for adipose-derived stem cells have been successfully applied in hAMSC [102], allowing luminescence imaging of their survival, migration, and distribution in preclinical in vivo models.
|
CELL AND TISSUE BANKING |
|---|
Cell Banking
So far, most experience in preservation of placental tissue-derived cells has been gained with cord blood, which contains both hematopoietic and mesenchymal stem cells. When cord blood transplantation proved effective, many cord blood banks were established, offering collection and banking for public (allogeneic) or private (autologous or allogeneic) use. Cord blood is procured from natural births or caesarean sections [110]. Different methods for the reduction of red blood cells, plasma volume, and cryopreservation exist [111]. Cord blood products containing cryoprotectants (e.g., dimethyl sulfoxide [DMSO]) are frozen at a controlled-rate and can be stored in liquid nitrogen for at least 15 years without the loss of their engraftment potential in vivo [110, 112].
In contrast to cord blood [113, 114], fetal membrane stem cells are presently preserved mainly for research. However, as these cells gain interest for their regenerative and immunomodulatory properties (described above, in Immunology of the Placenta) [29, 115], future medical needs may require concomitant application of cord blood and placental cells from the same donor. After cryopreservation, hAMSC and hCMSC can be differentiated at least along the osteogenic lineage (S. Hennerbichler, unpublished data). However, no conclusive studies are so far available on the influence of cryopreservation on plasticity.
Tissue Banking
The use of amniotic membrane has history spanning almost 100 years. The first reported clinical use of amniotic membrane was in 1910, when it was applied in skin transplantation [116]. Shortly after, application was expanded to treat burned and ulcerated skin [117, 118] and conjunctival defects [119]. Since its rediscovery in 1995 [120], it has been widely applied in ophthalmology, surgery, and wound healing [121]. Besides its nearly unlimited availability, easy procurement, and low processing costs for therapeutic application, many beneficial properties of this tissue, including bacteriostatic, anti-inflammatory, analgesic, wound healing, reepithelialization, reduced scarring, and anatomical and vapor barrier properties, have been reported [56, 122–124].
Currently, freeze-dried, -sterilized, decellularized, glycerol-preserved, and cryopreserved amniotic membranes are used in ophthalmology and wound care [123, 125–128]. Questions remain as to how these different processing and preservation methods influence sterility, viability, and growth factor release [129, 130] (S. Hennerbichler, unpublished data).
When placentae from caesarean sections were collected, 0 of 10 tested were contaminated with aerobic or anaerobic bacteria, whereas bacteria were detected in 4 of 10 naturally born placenta tested (S. Hennerbichler, unpublished data). It may therefore be preferential to collect placentae from caesarean sections.
Membranes were further investigated for viability and growth factor release after glycerol (50% glycerol) or cryopreservation (10% DMSO). Cryopreserved membrane retained cell viability and released several angiogenic growth factors and cytokines [129, 130], whereas storage in glycerol at 4°C resulted in immediate cell death [129]. Others have confirmed that different processing methods such as irradiation influence the growth factor content of amniotic membrane [131].
|
CONCLUSION |
|---|
|
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
Top
Abstract
Introduction
