Chemically defined and xenogeneic-free differentiation of human pluripotent stem cells into definitive endoderm in 3D culture

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.

FGF2 Inhibits Early Pancreatic Lineage Specification during Differentiation of Human Embryonic Stem Cells

“…Intrigued by the finding that developing hESCs actively contribute to the spatio–temporal milieu of growth factors in the differentiation medium, we continued to assess the secretory profile of hESCs prior to pancreatic induction. For that purpose hESCs were cultured in BSA- and serum-free TeSR-E8-medium and subsequently differentiated in chemically defined stage 1 medium [ 13 , 17 ]. Media supernatants from d0 and d4 were collected and quantitatively analyzed by LC-MS/MS.…”

“…Feeder-free culturing of human HES3 cells was performed as described earlier [ 13 , 14 ]. Differentiation was initiated according to our protocol described in [ 15 ].…”

“…Secretome analysis was performed as previously described [ 17 ]. Briefly, media supernatants conditioned for 24 h from HES3 grown in E8 medium (d0 samples) and from HES3-derived endodermal cells (d4 samples) grown in xenogeneic-free endoderm differentiation medium [ 13 ] were collected and centrifuged at 2000× g for 5 min to remove suspended cells and cell debris. A total of 8–9 mL of cell culture supernatants were then applied for precipitation according to the TCA-NLS method [ 18 , 19 ].…”

“…Retinoic acid-signaling and inhibition of BMP/Wnt-signaling was identified as prerequisite for the patterning of hESC-derived DE into PDX1-positive posterior foregut endoderm. These pancreatic-duodenal cells could then be further differentiated into PDX1+/NKX6.1+ multipotent progenitor cells [ 12 , 13 ]. In the present study we analyzed the additional effects of the growth factors FGF2, FGF7, FGF10, and EGF during the anterior–posterior patterning stage.…”

Non‐xenogeneic expansion and definitive endoderm differentiation of human pluripotent stem cells in an automated bioreactor

“…In fact, the dependence of the DE differentiation outcome on the type of medium used for cell propagation was demonstrated (Diekmann et al, 2019) as well as the development of a xeno‐free formulation with a B27 supplement, excluding two of its animal‐derived components. Despite employing static cultivation of HES3 hESCs and observing oversized clusters (some ~500 μm), the estimated yield of 2.8–5.6 × 10 6 DE cells/well (six‐well plate, Table S4; Diekmann et al, 2019), was comparable to that in our study with 6.4 × 10 6 and 7.2 × 10 6 DE cells/well for IMR90 and H9 hPSCs (Table S4), respectively. Generally, a shift away from supplements such as B27 would be preferable since these contain factors with pleiotropic and hard‐to‐predict effects on the specification outcome.…”