oa Enhancement of differentiation of multipotent pancreatic β cell precursors from human embryonic stem cells

Scalable production of human pluripotent stem cell (hPSC)-derived β cells in vitro would greatly facilitate transplantation therapy and drug discovery for treating diabetes. Employing step-wise differentiation protocols, hPSCs can be differentiated through consecutive stages of endoderm, foregut, pancreatic and endocrine progenitors to ultimately give insulin secreting β cells. Pancreatic progenitors co-expressing the two key transcription factors (TFs), PDX1 and NKX6.1, are recognized as the indispensable precursors of functional, mono-hormonal β cells. Here, we established an optimized protocol for maximizing PDX1+/NKX6.1+ co-positive pancreatic progenitors from hESCs in monolayer culture and increasing their proliferative capacity. Our technique of dissociating densely formed endodermal cells and re-plating them in lower densities on fresh matrigel matrix followed by an augmented duration of retinoid and FGF10 signaling strikingly increased the expression of NKX6.1, which is exclusive only to β cells amongst all endocrine cells. This high induction of NKX6.1 resulted in an increased proportion of PDX1+/ NKX6.1+ population, generating up to >90% PDX1+/ NKX6.1+ co-positive progenitors in monolayer, higher than previously published protocols. In contrast to multiple studies showing negligible induction of NKX6.1 at lower densities, we provide evidence that higher folds of NKX6.1 can be induced in dissociated cells re-plated lower densities compared to aggregations in non-dissociated culture if the duration of retinoid and FGF signaling is prolonged. Our optimized protocol enhanced pancreatic differentiation efficiency by up-regulating pancreatic progenitor TFs such as PDX1, SOX9, HNF6 and FOXA2 and increased the mRNA levels of endocrine TFs such as NEUROG3, NKX2.2 and NEUROD1. Additionally, we show that manipulating cell-cell attachment following endoderm generation in vitro during pancreatic differentiation dramatically inhibited alternate hepatic fate specification by down-regulating hepatic markers like AFP and ALB expression in our optimized protocol in comparison to recently published protocols for generating pancreatic progenitors. Notably, cell cycle and BrdU incorporation assays revealed that our method increased the proliferative capacity of pancreatic progenitors throughout the differentiation stages by increasing the fraction of cells entering S phase of cell cycle and a comparative increase in Ki67 expression, the proliferation marker. As a result, we obtained >70% Ki67+ /SOX9+ pancreatic progenitors in monolayer confirming an increased self-replicating capacity of the generated PDX1+/ NKX6.1+ progenitors. Furthermore, using our optimized protocol for pancreatic differentiation, we were able to enrich a novel and uncharacterized NKX6.1+ /PDX1- population, devoid of Chromogranin A (CHGA) expression, which are therefore proposed to be more mature precursors of β cell. This population re-arranged themselves in embedded, highly compact three-dimensional structures that showed high expression of Ki67. Continuation of our optimized protocol into endocrine differentiation stage validated the ability of our PDX1+/ NKX6.1+ to generate NGN3+/ NKX6.1+ co-positive endocrine progenitors in vitro with a high expression of CHGA and NKX2.2. Therefore, here we show that manipulating the cellular density, cell-cell attachment and cues from extracellular matrix plays a major role in improving pancreatic differentiation efficiency and proliferation thereby providing a cost-effective method for generating pancreatic progenitors in vitro in adherent culture. Indeed, our novel method for maximizing PDX1+/ NKX6.1+ progenitors from hPSCs in monolayer culture could serve as a source of highly proliferative pancreatic progenitors aiding scalable production of functional β cells in vitro.