Generation of definitive endoderm from human embryonic stem cells cultured in feeder layer-free conditions.

PURPOSE: The purpose of these studies was twofold: to reduce the level of nonhuman, potentially immunogenic sialic acid N-glycolylneuraminic acid (Neu5Gc) in human embryonic stem cells (hESCs) through culture of the cells in the absence of feeder layers; and to determine whether directed differentiation was preserved under these conditions, that is, using exclusively human-derived products. METHODS: Using a technique developed in our laboratory to culture hESCs in the absence of feeder layers, all nonhuman cell culture reagents were replaced with recombinant or human-derived reagents. The level of the nonhuman sialic acid (Neu5Gc) was measured by high-performance liquid chromatography and monitored over many passages. Subsequently, the cells were subjected to in vitro differentiation into definitive endoderm by lowering the serum concentrations and elevating the amount of activin A. RESULTS: Under standard tissue culture conditions using mouse and other animal products, the basal levels of Neu5Gc were measured between 7 and 10%. After the cell culture reagents were changed to all human products, Neu5Gc levels decreased steadily before leveling below 2%. Upon initiation of the differentiation protocol under these cell culture conditions, we observed robust endoderm formation, as measured by fluorescence-activated cell sorting analysis and the appearance of mRNA for markers of definitive endoderm (Sox17, CXCR4, Goosecoid and FoxA2). CONCLUSION: Consistent with other findings, elimination of nonhuman products in cell culture of hESCs decreases the levels of nonhuman and potentially immunogenic sialic acid levels. Furthermore, our studies demonstrate that in this feeder layer-free system, hESCs undergo directed differentiation into definitive endoderm.

Limited capacity of human adult islets expanded in vitro to redifferentiate into insulin-producing beta-cells.

Limited organ availability is an obstacle to the widespread use of islet transplantation in type 1 diabetic patients. To address this problem, many studies have explored methods for expanding functional human islets in vitro for diabetes cell therapy. We previously showed that islet cells replicate after monolayer formation under the influence of hepatocyte growth factor and selected extracellular matrices. However, under these conditions, senescence and loss of insulin expression occur after >15 doublings. In contrast, other groups have reported that islet cells expanded in monolayers for months progressed through a reversible epithelial-to-mesenchymal transition, and that on removal of serum from the cultures, islet-like structures producing insulin were formed (1). The aim of the current study was to compare the two methods for islet expansion using immunostaining, real-time quantitative PCR, and microarrays at the following time points: on arrival, after monolayer expansion, and after 1 week in serum-free media. At this time, cell aliquots were grafted into nude mice to study in vivo function. The two methods showed similar results in islet cell expansion. Attempts at cell differentiation after expansion by both methods failed to consistently recover a beta-cell phenotype. Redifferentiation of beta-cells after expansion is still a challenge in need of a solution.

Long-term self-renewal and directed differentiation of human embryonic stem cells in chemically defined conditions.

Chemically defined medium (CDM) conditions for controlling human embryonic stem cell (hESC) fate will not only facilitate the practical application of hESCs in research and therapy but also provide an excellent system for studying the molecular mechanisms underlying self-renewal and differentiation, without the multiple unknown and variable factors associated with feeder cells and serum. Here we report a simple CDM that supports efficient self-renewal of hESCs grown on a Matrigel-coated surface over multiple passages. Expanded hESCs under such conditions maintain expression of multiple hESC-specific markers, retain the characteristic hESC morphology, possess a normal karyotype in vitro, as well as develop teratomas in vivo. Additionally, several growth factors were found to selectively induce monolayer differentiation of hESC cultures toward neural, definitive endoderm/pancreatic and early cardiac muscle cells, respectively, in our CDM conditions. Therefore, this CDM condition provides a basic platform for further characterization of hESC self-renewal and directed differentiation, as well as the development of novel therapies.

Activin A maintains pluripotency of human embryonic stem cells in the absence of feeder layers.

To date, all human embryonic stem cells (hESCs) available for research require unidentified soluble factors secreted from feeder layers to maintain the undifferentiated state and pluripotency. Activation of STAT3 by leukemia inhibitory factor is required to maintain "stemness" in mouse embryonic stem cells, but not in hESCs, suggesting the existence of alternate signaling pathways for self-renewal and pluripotency in human cells. Here we show that activin A is secreted by mouse embryonic feeder layers (mEFs) and that culture medium enriched with activin A is capable of maintaining hESCs in the undifferentiated state for >20 passages without the need for feeder layers, conditioned medium from mEFs, or STAT3 activation. hESCs retained both normal karyotype and markers of undifferentiated cells, including Oct-4, nanog, and TRA-1-60 and remained pluripotent, as shown by the in vivo formation of teratomas.

Maintenance of pluripotency in human embryonic stem cells is STAT3 independent.

The preservation of "stemness" in mouse embryonic stem (mES) cells is maintained through a signal transduction pathway that requires the gp130 receptor, the interleukin-6 (IL-6) family of cytokines, and the Janus Kinase-signal transducer and activator (JAK/STAT) pathway. The factors and signaling pathways that regulate "stemness" in human embryonic stem (hES) cells remain to be elucidated. Here we report that STAT3 activation is not sufficient to block hES cell differentiation when the cells are grown on mouse feeder cells or when they are treated with conditioned media from feeder cells. Human ES cells differentiate in the presence of members of the IL-6 family of cytokines including leukemia inhibitory factor (LIF) and IL-6 or in the presence of the designer cytokine hyper-IL-6, which is a complex of soluble interleukin-6 receptor (IL-6R) and IL-6 with greatly enhanced bioactivity. Human ES cells express LIF, IL-6, and gp130 receptors, as well as the downstream signaling molecules. Stimulation of human and mouse ES cells with gp130 cytokines resulted in a robust phosphorylation of downstream ERK1, ERK2, and Akt kinases, as well as the STAT3 transcription factor. Loss of the pluripotency markers Nanog, Oct-4, and TRA-1-60 was observed in hES cells during gp130-dependent signaling, indicating that signaling through this pathway is insufficient to prevent the onset of differentiation. These data underscore a fundamental difference in requirements of murine versus hES cells. Furthermore, the data demonstrate the existence of an as-yet-unidentified factor in the conditioned media of mouse feeder layer cells that acts to maintain hES cell renewal in a STAT3-independent manner.

Characterization and isolation of promoter-defined nestin-positive cells from the human fetal pancreas.

Studies using adult human islets and mouse embryonic stem cells have suggested that the neurepithelial precursor cell marker nestin also identifies and can be used to purify beta-cell precursors. To determine whether nestin can be used to identify beta-cell progenitors in the developing human pancreas, we characterized nestin expression from 12 to 24 gestational weeks, purified nestin+ cells using an enhancer/promoter-driven selection plasmid, and determined whether nestin+ cells can differentiate into beta-cells. Nestin was visualized in the platelet endothelial cell adhesion molecule and alpha smooth muscle actin-positive blood vessels and colocalized with vimentin in the interstitium. Nestin was not observed in pan cytokeratin (pCK)-positive ductal epithelium or insulin cells. Purified nestin+ cells also coexpressed vimentin and lacked pCK immunoreactivity. Purified adult and fetal pancreatic fibroblasts also expressed nestin. The nestin enhancer/promoter used in the selection plasmid was sufficient to drive reporter gene expression, green fluorescent protein, in human fetal pancreatic tissue. Exposure of selected nestin+ cells to nicotinamide, hepatocyte growth factor/scatter factor, betacellulin, activin A, or exendin-4 failed to induce pancreatic and duodenal homeobox gene-1 or insulin message as determined by RT-PCR. Transplantation of nestin+ cells and fetal pancreatic fibroblasts into athymic mice also failed to result in the development of beta-cells, whereas nestin- fetal pancreatic epithelial cells gave rise to functional insulin-secreting beta-cells. We conclude that nestin is not a specific marker of beta-cell precursors in the developing human pancreas.

A novel approach to increase human islet cell mass while preserving beta-cell function.

Human islet expansion in monolayer culture leads to loss of function and senescence. By maintaining the 3-D configuration of islets in fibrin gels, it is feasible to expand beta-cells in response to hepatocyte growth factor (HGF) while preserving physiologic glucose responsiveness both in vitro and in vivo after transplantation into nude mice. Islets were cultured free floating with or without growth factors and nicotinamide and in fibrin gels with the same conditions. Proliferation was observed only in islets cultured in fibrin gels and the cocktail; total insulin increased by threefold, with a concomitant increase in beta-cell mass by morphometry. Insulin release after glucose challenge was also preserved. Islets in fibrin gels gave rise in vivo to large grafts rich in insulin and glucagon, and grafts from free-floating islets were smaller with fewer endocrine cells. Circulating human C-peptide levels were higher than in the mice receiving free-floating islets. In summary, fibrin allows for HGF-mediated cell proliferation while preserving glucose responsiveness in an environment that preserves cell-cell contacts. Limited islet ex vivo expansion under these conditions may improve recipient-donor tissue ratios to equal the functional results of whole-organ transplants.

Exendin 4 up-regulates expression of PDX 1 and hastens differentiation and maturation of human fetal pancreatic cells.

In addition to stimulating insulin secretion, glucagon-like peptide and its long-acting analog exendin 4 have been reported to increase beta-cell mass by both differentiation/neogenesis of precursor cells and enhanced replication of existing beta-cells. Here, we investigated the effect of exendin 4 in the growth and differentiation of beta-cells from undifferentiated precursors in islet-like cell clusters (ICCs) derived from human fetal pancreases. Our results show that the addition of exendin 4 to the culture media stimulates PDX 1 expression in ICCs as shown by immunofluorescence staining. The up-regulation of PDX 1 was not accompanied by changes in insulin expression because we did not find a significant difference in the number of insulin-positive cells in the exendin 4-treated ICCs, compared with controls. We also tested the effects of exendin 4 in the glucose-induced insulin secretion of human ICCs transplanted under the kidney capsule of athymic rats. In the exendin 4-treated rats (given ip during 10 d) 8 wk after the beginning of the treatment, insulin was released in response to glucose as detected by the measurement of circulating human C-peptide. In control (saline-treated) rats, the basal levels of human C-peptide did not change significantly after glucose stimulation. Thus, exendin 4 induces functional maturation of fetal beta-cells in response to glucose. In these rats, serial sections of the kidney-bearing grafts were examined histologically for insulin containing cells. We found a significant increase in beta-cell number, compared with the control rats. Overall, these results show that in vivo exendin 4 causes growth and differentiation of human fetal beta-cells from undifferentiated precursor cells. It also accelerates the functional maturation of fetal beta-cells as evidenced by their glucose-stimulated insulin secretion.

Alternatives to unmodified human islets for transplantation.

Islet transplantation as a procedure to induce insulin independence is still a long way from benefitting the population of more than I million type I diabetic patients in the United States. In addition to the problems involved with immune suppression, the most significant obstacle is a scarcity of human organs for transplantation. In 1999, only 5882 donated pancreases were available, of which only 50% could be expected to produce islet yields suitable for clinical purposes. In this article, we review various sources with the potential to provide tissue for transplantation. These sources include islet and nonislet cells derived from both human and nonhuman sources, with an emphasis on human cells.