A progressive decrease in the pancreatic beta-cell mass is the main cause of diabetes and can be counterbalanced by providing new insulin-producing cells (IPCs). Promising islet cell transplantation methods as a means of reversing type 1 diabetes mellitus (DM) have been hampered by islet availability as well as by allograft rejection1,2. Liver-to-endocrine pancreas transdifferentiation is an attractive strategy for generating beta-cell surrogates since the liver and pancreas share a common bipotential precursor cell within the embryonic endoderm3,4. Most previously reported induction strategies required genetic manipulation, and induced cells were incomplete and non-selective for beta-cell phenotypes and function5-7.
Recent progress has been made demonstrating that some small molecules can promote definitive cellular differentiation to induce pluripotent stem cells (PSCs)6,8 and modulate various stages of beta-cell differentiation from induced PSCs9,10. For example, 5-aza-2'-deoxycytidine (5-AZA), an inhibitor of DNA methylase, was used to successfully induce Ngn3 expression and endocrine differentiation in the PANC-1 human ductal cell line11. Trichostatin A (TSA), a regulator of chromatin remodeling, triggers the process of dedifferentiation and further endocrine pancreatic differentiation12,13. Retinoic acid (RA) facilitates the development of Pdx1 pancreatic endocrine progenitor cells and their further differentiation into beta-cells14,15. Nicotinamide promotes transdifferentiation and maturation of liver stem cells into IPCs16.
There have already been several reports on the generation of IPCs using small molecules from various sources of stem cells, such as embryonic stem cells and mesenchymal stem cells. However, to date, there has been no report demonstrating that pancreatic beta-cells can be induced from liver stem cells in this way. Therefore, by mimicking embryonic pancreatic development, we explored a stepwise protocol for the formation of IPCs from WB cells to control differentiation in vitro. In this study, efforts were focused on the key step of generating Pdx1-expressing cells as pancreatic progenitors. Several methods for inducing differentiation from stem cells into Pdx1-positive cells have been recently reported2,14,17,18, but the combination and sequence of small molecules for induction and differentiation remain to be optimized.
We present here a three-step approach for generating functional IPCs from WB cells within only 17 days. A sequential treatment with 5-AZA, TSA, RA, and insulin, transferrin and selenite (ITS) induced efficient differentiation of WB cells into pancreatic precursor cells（WB-A cells）. More than 5% of the cells became Pdx1-expressing pancreatic progenitors in only 10 days, eventually becoming IPCs in a step requiring nicotinamide for 7 days in vitro or transplantation into diabetic nude rats for 60 days in vivo. These IPCs expressed characteristic pancreatic beta-cell marker genes, such as insulin I, insulin II, GK with no evidence of non-beta-gene expression. Moreover, insulin was released in response to glucose concentration and the induced IPCs in vivo could ameliorate hyperglycemia in diabetic rats. The identification of an attractive chemical strategy that induces hepatocyte-derived IPCs is an important approach for diabetes cell therapy in the near future.