No abstract

In vitro differentiation of human pluripotent stem cells (hPSCs) into definitive endoderm (DE) represents a key step towards somatic cells of lung, liver and pancreas. For future clinical applications, mass production of differentiated cells at chemically defined conditions and free of xenogeneic substances is envisioned. In this study we adapted our previously published two-dimensional (2D) DE induction protocol to three-dimensional (3D) static suspension culture in the absence of the xenogeneic extracellular matrix Matrigel. Next, fetal calf serum and bovine serum albumin present in the standard medium were replaced by a custom-made and xeno-free B-27. This yielded in a chemically defined and xenogeneic-free 3D culture protocol for differentiation of hPSCs into DE at efficiencies similar to standard 2D conditions. This novel protocol successfully worked with different hPSC lines including hESCs and hiPSCs maintained in two different stem cell media prior to differentiation. DE cells obtained by our novel BSA-free 3D protocol could be further differentiated into PDX1- or NKX6.1-expressing pancreatic progenitor cells. Notably, upon DE differentiation, we also identified a CXCR4+/NCAM+/EpCAMlow cell population with reduced DE marker gene expression. These CXCR4+/NCAM+/EpCAMlow cells emerge as a result of Wnt/beta-catenin hyperactivation via elevated CHIR-99021 concentrations and likely represent misspecified DE.

Growth factors are important regulators during organ development. For many vertebrates (but not humans) it is known how they contribute to the formation and expansion of PDX1-positive cells during pancreas organogenesis. Here, the effects of the fibroblast growth factors FGF2, FGF7, FGF10, and epidermal growth factor (EGF) on pancreas development in humans were assessed by using human pluripotent stem cells (hPSCs). During this, FGF2 was identified as a potent anti-pancreatic factor whereas FGF7, FGF10, and EGF increased the cell mass while retaining PDX1-positivity. FGF2 increased the expression of the anti-pancreatic factor sonic hedgehog (SHH) while suppressing PDX1 in a dose-dependent manner. Differentiating cells secreted SHH to the medium and we interrogated the cells’ secretome during differentiation to globally examine the composition of secreted signaling factors. Members of the TGF-beta-, Wnt-, and FGF-pathways were detected. FGF17 showed a suppressive anti-pancreatic effect comparable to FGF2. By inhibition of specific branches of FGF-receptor signaling, we allocated the SHH-induction by FGF2 to MEK/ERK-signaling and the anti-pancreatic effect of FGF2 to the receptor variant FGFR1c or 3c. Altogether, we report findings on the paracrine activity of differentiating hPSCs during generation of pancreatic progenitors. These observations suggest a different role for FGF2 in humans compared to animal models of pancreas organogenesis.

For the production and bio-banking of differentiated derivatives from human pluripotent stem cells (hPSCs) in large quantities for drug screening and cellular therapies, well-defined and robust procedures for differentiation and cryopreservation are required. Definitive endoderm (DE) gives rise to respiratory and digestive epithelium, as well as thyroid, thymus, liver, and pancreas. Here, we present a scalable, universal process for the generation of DE from human-induced pluripotent stem cells (hiPSCs) and embryonic stem cells (hESCs). Optimal control during the differentiation process was attained in chemically-defined and xeno-free suspension culture, and high flexibility of the workflow was achieved by the introduction of an efficient cryopreservation step at the end of DE differentiation. DE aggregates were capable of differentiating into hepatic-like, pancreatic, intestinal, and lung progenitor cells. Scale-up of the differentiation process using stirred-tank bioreactors enabled production of large quantities of DE aggregates. This process provides a useful advance for versatile applications of DE lineages, in particular for cell therapies and drug screening.

Aims/hypothesis The aim of this study was to examine the effects of proinflammatory cytokines on cells of different developmental stages during the generation of stem cell-derived beta cells (SC-beta cells) from human pluripotent stem cells (hPSCs). We wanted to find out to what extent human SC-beta cells are suitable as an experimental cellular model and, with regard to a possible therapeutic use, whether SC-beta cells have a comparable vulnerability to cytokines as bona fide beta cells. Methods hPSCs were differentiated towards pancreatic organoids (SC-organoids) using a 3D production protocol. SC-beta cells and non-insulin-producing cells were separated by FACS and differential gene expression profiles of purified human SC-beta cells, progenitor stages and the human beta cell line EndoC-βH1, as a reference, were determined after 24 h incubation with the proinflammatory cytokines IL-1β, TNF-α and IFN-γ via a transcriptome microarray. Furthermore, we investigated apoptosis based on caspase cleavage, the generation of reactive oxygen species and activation of mitogen-activated protein-kinase (MAPK) stress-signalling pathways. Results A 24 h exposure of SC-beta cells to proinflammatory cytokines resulted in significant activation of caspase 3/7 and apoptosis via the extrinsic and intrinsic apoptosis signalling pathways. At this time point, SC-beta cells showed a markedly higher sensitivity towards proinflammatory cytokines than non-insulin-producing cells and EndoC-βH1 cells. Furthermore, we were able to demonstrate the generation of reactive oxygen species and rule out the involvement of NO-mediated stress. A transient activation of stress-signalling pathways p38 mitogen-activated protein kinases (p38) and c-Jun N-terminal kinase (JNK) was already observed after 10 min of cytokine exposure. The transcriptome analysis revealed that the cellular response to proinflammatory cytokines increased with the degree of differentiation of the cells. Cytokines induced the expression of multiple inflammatory mediators including IL-32, CXCL9 and CXCL10 in SC-beta cells and in non-insulin-producing cells. Conclusions/interpretation Our results indicate that human SC-beta cells respond to proinflammatory cytokines very similarly to human islets. Due to the fast and fulminant cellular response of SC-beta cells, we conclude that SC-beta cells represent a suitable model for diabetes research. In light of the immaturity of SC-beta cells, they may be an attractive model for developmentally young beta cells as they are, for example, present in patients with early-onset type 1 diabetes. The secretion of chemotactic signals may promote communication between SC-beta cells and immune cells, and non-insulin-producing cells possibly participate in the overall immune response and are thus capable of amplifying the immune response and further stimulating inflammation. We demonstrated that cytokine-treated SC-organoids secrete IL-32, which is considered a promising candidate for type 1 diabetes onset. This underlines the need to ensure the survival of SC-beta cells in an autoimmune environment such as that found in type 1 diabetes. Graphical abstract