Phenotypic analysis of human stem cells and their differentiation potential towards pancreatic cells

Abstract

Type 1 diabetes is an autoimmune disorder in which the immune system specifically destroys the insulin-producing β-cells of the pancreas, resulting in a loss of glycemic control. There is no cure for the disease, and patients have the difficult task of managing their blood glucose levels by daily injection of insulin, and monitoring of carbohydrate intake, levels of exercise, and stress. The field of stem cells gives hope to type 1 diabetes patients as stem cells are plastic and can be differentiated into the pancreatic lineage to form insulin-producing β-cell-like cells for transplantation to restore the endocrine function of the pancreas. There are many different stem cell populations described in the published literature that can be easily obtained from the blood, tissues, and organs of adults. They include peripheral blood-insulin producing cells (PB-IPCs) that can be readily isolated due to their unique attachment to hydrophobic surfaces, monocytes that can be dedifferentiated into programmable cells of monocytic origin (PCMOs) and fibroblast-resembling-macrophages (f-MOs), and adipose stem cells (ASCs). There are also other populations of stem cells that can be ethically obtained after childbirth from neonatal tissues that are normally discarded, and include human amniotic stem cells (HASCs) and cord blood hematopoietic stem cells (CB-HSCs). It has been reported in the published literature that each of the above stem cell populations except f-MOs can be induced to form insulin-producing cells. However, in most cases there has been no independent validation of the published data. The aim of this project was to study the phenotype of the above stem cells and determine which stem cell population would be an ideal source from which to derive insulin-producing cells. The published cell differentiation protocols were to be validated, and novel ways of improving upon the protocols were to be examined with the ultimate aim to produce cells that closely resembled pancreatic β-cells. In this study, the above stem cell populations were either produced from the blood of human donors, or purchased directly from commercial companies. They were phenotyped for stem cell, pancreatic cell, and other cell lineage markers. The stem cells were subjected to a variety of published and novel pancreatic differentiation protocols, and examined for the expression of markers that specify different stages of pancreatic differentiation. The ability to transdifferentiate several of the stem cell populations into other cell lineages including the endothelial, neuronal, and lymphocytic lineages was also examined. Phenotyping of the stem cell populations revealed that some undifferentiated stem cells already expressed certain markers of pancreatic progenitor cells. Monocytes dedifferentiated into PCMOs expressed SOX-17 and PDX-1, in addition to the monocytic markers CD14 and CD68. A subset of monocytes dedifferentiated into f-MOs expressed PDX-1, while the remaining cells expressed CD14 indicating the presence of two different cell populations. Therefore it appeared that the monocyte-derived stem cell populations are in an intermediate state of transdifferentiation expressing multiple cell lineage markers. Undifferentiated HASCs expressed PDX-1 and SOX-17 whilst also expressing the stem cell markers Nestin and Oct-4. Several CB-HSCs expressed SOX-17, GPR40, and Nestin. ASCs expressed SOX-17, in addition to the characteristic ASC stem cell markers CD29 and CD90. The expression of pancreatic markers by undifferentiated HASCs, CB-HSCs, and ASCs indicates they could have the potential to differentiate towards the pancreatic lineage. PB-IPCs expressed the pancreatic progenitor cell markers PDX-1, CXCR4, HNF6, and SOX-17 following culture in the absence of additional growth factors other than foetal bovine serum. They highly expressed the dendritic cell marker CD11c and appeared to be in a state of activation as they strongly expressed HLA-DR. The PB-IPCs were phagocytic. Their function and phenotype therefore suggests they are monocyte-like. Attempts to induce their transdifferentiation towards the pancreatic lineage with previously published protocols and novel combinations of differentiation agents were not successful. Attempts to differentiate the monocyte-derived PCMOs and f-MOs towards the pancreatic lineage using a published protocol based on an islet cell conditioning medium (ICM) were not successful. Differentiated PCMOs continued to express CD14 and HLA-DR, and differentiated f-MOs retained their expression of PDX-1 or CD14. Thus, these stem cell populations did not show plasticity towards the pancreatic lineage beyond their expression of PDX-1. f-MOs showed plasticity towards the endothelial lineage after treatment with VEGF165 resulting in vWF-expressing cells, towards the neuronal lineage after treatment with NGFβ resulting in neuron-specific enolase-expressing cells, and towards the lymphocytic lineage after treatment with IL-2 resulting in an increase in CD3- and CD4-expressing cells. Therefore f-MOs may be a source of cells for the generation of these latter cell types. Attempts to differentiate CB-HSCs towards the pancreatic lineage with a commercial medium and matrix from Celprogen Inc. resulted in the generation of pancreatic progenitor cells which expressed GIPR, PDX-1, and GPR40, but there was no evidence of further maturation into insulin-producing endocrine cells. ASCs were subjected to stage-specific differentiation protocols designed to mimic the natural differentiation stages of pancreatic cell formation in vivo. The differentiated ASCs showed evidence of differentiation towards the pancreatic lineage as reflected in the upregulation of the posterior foregut markers HNF6 and PDX-1. The highest amount of PDX-1 was expressed after culture of ASCs with a combination of 100 nM IDE-1, followed by KAAD-cyclopamine and fibroblast growth factor (FGF)-10, and then KAAD-cyclopamine, FGF10 and all trans-retinoic acid. These PDX-1-expressing cells represent a source of pancreatic progenitor cells differentiated up to the posterior foregut stage. HASCs purchased from Celprogen Inc. displayed the greatest potential to be differentiated towards the pancreatic β-cell lineage. They could be transdifferentiated with media and extracellular matrices provided by Celprogen Inc. into cells which expressed PDX-1, SOX-17, Arx, CXCR4, HNF6, glucagon, and PCSK1. However, they did not express insulin. In contrast, ASCs cultured in a published medium on ultra-low attachment plates contained high levels of insulin detectable by a widely used guinea-pig anti-insulin antibody. However, this result is confounded by the fact that insulin expression was not detected by two other anti-insulin antibodies. The differentiated ASCs also did not express proinsulin, amylin, and C-peptide. Therefore the insulin detected by the guinea pig antibody appears not to be genuine, and may result from the previously unreported non-specific binding of the anti-insulin antibody to a related protein. Phenotyping of the differentiated cells revealed some cells expressed tolerogenic or immunosuppressive markers. Differentiated HASCs and PB-IPCs expressed AIRE suggesting they might serve as tolerogenic antigen presenting cells. However, differentiated PB-IPCs, which expressed AIRE and FOXP3, did not inhibit a mixed lymphocyte response, indicating they were not directly immunosuppressive or potentially tolerogenic. Differentiated CB-HSCs and HASCs expressed the immunosuppressive marker TGF-β, in accord with the published report that a population of cord blood stem cells selected by hydrophobic surface attachment from whole cord blood are immunosuppressive and can induce the differentiation of naïve mouse splenic T cells into T regulatory cells (Tregs) that are capable of reversing type 1 diabetes in mice and humans. These studies have independently re-examined protocols for promoting the pancreatic differentiation of a variety of human adult and neonatal stem cells. Only the differentiation of HASCs produced cells that contained detectable levels of pancreatic hormones, in this case glucagon. Most of the published protocols were ineffective at differentiating the stem cell populations to pancreatic hormone-producing cells, contrary to the published data. This raises questions about the reproducibility of stem cell protocols and the need for more robust methods that can be independently replicated.