FGF signaling in specification of hESC-derived definitive endoderm

Sammanfattning: Diabetes affects around 200 million people worldwide. Curing diabetes would require the prevention of autoimmune destruction of beta cells and restoration of the beta cell mass restored either through regeneration or transplantation of the insulin producing cells. Islet transplantation according to the Edmonton protocol have been the most promising therapeutic option for Diabetes type I patients, however, lack of cadaveric donor islets is a major obstacle and new strategies need to be established. Human embryonic stem cells (hESCs) not only offer an excellent source for establishment of strategies for future regenerative cell therapies and drug discovery but also offers en excellent experimental assay for understanding human pancreas development. Pancreas originates from the definitive endoderm (DE), one of the three germ layers. The path from definitive endoderm to pancreatic progenitor cells and finally the insulin producing beta cells involves sequential cell-fate decisions characterized by the expression of multiple transcription factors. Governed by the developmental biological principles that normally control foregut endoderm and pancreas specification, numerous multifactor protocols for directing mature foregut-derived cells from hESCs have been reported. However, to establish less complex and more robust protocols there is a need to further understand the mechanism of action of individual growth and differentiation factors in specification of human pluripotent stem cells towards foregut derived cell lineages. Here, we describe two studies where we have investigated the role of fibroblast growth factor (FGF) signaling, specifically FGF2 and FGF4, and Retinoic acid (RA) in specification of the hESC-derived DE development. Studies in lower vertebrates have demonstrated that FGF2 acts in a restricted manner primarily patterning the ventral foregut endoderm into liver and lung, whereas FGF4 exhibits broad anteriorposterior and left-right patterning activities. Furthermore, whereas FGF2 is not required for ventral pancreas development an inductive role of FGF2 has been shown during dorsal pancreas formation. Whether FGF2 and FGF4 play a similar role during human endoderm development remained however unknown. In contrast, RA has frequently been employed (together with other growth factors) for directed differentiation of hESCs to pancreatic endoderm. In the first study we show that FGF2 in a dosage-dependent manner specifies hESC-derived DE into different foregut lineages such as liver, lung, pancreatic and intestinal cells. Furthermore, by dissecting the FGF receptor intracellular pathway that regulates pancreas specification, we demonstrate for the first time to our knowledge that induction of PDX1+ pancreatic progenitors in part relies on 8 FGF2-mediated activation of the MAPK signaling pathway. In the second study, we show that FGF4 alone is not sufficient for induction of foregut endoderm but that in combination with RA it efficiently induces PDX1+ cells from hESC-derived DE. Specifically, FGF4 promoted cell survival in the differentiating hESCs. Hence, in contrast to studies in lower vertebrates we demonstrate that FGF4 neither patterns hESC-derived DE, nor induces PDX1+ pancreatic progenitors suggesting that FGF4 is not responsible for anterior-posterior patterning of the primitive gut during human development.. Altogether, these observations suggest a broader gut endodermal patterning activity of FGF2 that corresponds to what has previously been advocated for FGF4, implying a functional switch from FGF4 to FGF2 during evolution. Thus, our results provide new knowledge of how cell fate specification of human DE is controlled – facts that will be of great value for future regenerative cell therapies. Finally, we present a method for efficient gene targeting in hESCs, which allows the monitoring of gene expression in living cells.