(665c) Reprogramming of Liver Cell Lines to Definitive Endoderm By Understanding and Re-Engineering Developmental Master Regulatory Gene Circuits (DRGC)

Saber Meamardoost¹, Natesh Parashurama^1,2,3

Saber Meamardoost: sabermea@buffalo.edu

Natesh Parashurama: nateshp@buffalo.edu

¹ Department of Chemical and Biological Engineering, Furnas Hall, University at Buffalo, State University of New York at Buffalo, Buffalo NY 14260

² Department of Biomedical Engineering, Bonner Hall, University at Buffalo, State University of New York at Buffalo, Buffalo NY 14260

³ Department of Medicine, Clinical Translational Research Center (CTRC), 875 Ellicott St, Buffalo, NY 14214

Human pluripotent stem cell-derived (hPSC)-derived endoderm, which gives rise to liver, pancreas, and lung, is heterogeneous in both mouse and human hPSC studies, and remains the germ layer that is the hardest to induce and the least understood. Endoderm heterogeneity likely affects downstream differentiation and functional analysis and has already impacted clinical studies. Since there are no methods that address this problem, further investigation into the factors that regulate endoderm differentiation are warranted. The overall objective of our research is to elucidate how Foxa2 and Sox17 regulates developmental regulatory gene circuits (DRGC) composed of master regulatory transcription factors (TF) within endoderm and its progeny. Mouse genetic studies of Foxa1/Foxa2 double knockdown (Foxa1/2^-/-) within hepatic endoderm, demonstrate a complete absence of the liver bud and a failure to initiate the hepatic differentiation (Lee et al., Nature 2005). Similarly, Foxa1/2^-/- phenotype in pancreatic endoderm, severely blocks pancreatic growth and differentiation. Consistent with and prior to these studies, Foxa2 was identified as the first known pioneer TF, and binds to regulatory regions of silent endoderm patterning genes, like albumin, to prime them for rapid gene activation. The binding of Foxa2 appears to open normally inaccessible heterochromatin at developmental enhancers and provides access for other key DRGC TF at regulatory elements, like HNF1, HNF4a, and HNF6, to bind. This enables thousands of genes to be activated, within endoderm derivatives, in a relatively short amount of time (1-2 days) during murine development. This developmental role of Foxa1/2 is in contrast to its role in mature liver tissue, in which Foxa1/2 knockdown in adult hepatocytes has no effect on transcription factors or liver-specific genes. On the other hand, Sox17, a TF that plays a role in DRGC and involved in maintenance of endoderm, is also involved in lineage-specific gene repression in pancreas (Spence et al., Cell Stem Cell 2009) and liver (Pfister et al., 2010) within endoderm. Based on these studies and studies in lower organisms, the data suggests that Foxa1/2 can prevent differentiation towards hepatic and pancreatic endoderm, and Sox 17 can maintain endoderm and suppress further differentiation towards liver and pancreas.

To understand how Foxa1/2 and Sox17 contribute to the endodermal DRGC and orchestrate endoderm differentiation and maintenance, we first analyzed the role of Foxa1/2 in human hepatocellular carcinoma (HepG2) cells. HepG2 are a well-studied cell line which express both alphafetoprotein (AFP) and albumin (Alb), and thus represent a de-differentiated liver cell line. We hypothesized that Foxa1/2 regulates hepatic endoderm transcription factors expressed within the HepG2 cell, thereby serving as a model for endoderm/hepatic endoderm. We used an RNA interference approach for the Foxa1/2 (-/-) phenotype followed by qRT-PCR/western blot to measure mRNA/protein level in components of the DRGC and other master transcription factors involved in differentiation. Interestingly, absence of Foxa1/2 transcriptions factors is associated with downregulation of key hepatic markers Alb, Hhex, TTR and AFP; upregulation of some pancreatic markers (Pax6 and Pdx1) and repression of others (Sox 9). These observations are consistent with the role of Foxa1/2 during endoderm differentiation and supports the notion that HepG2 cells can resemble hepatic endoderm characteristics. To further understand endoderm TF and their role in DRGC, we overexpressed Sox17 in both wild type and Foxa1/2^-/- HepG2. Our preliminary data indicates that in wild type cells, exogenous Sox 17 represses Pax 6 expression, but not Pdx expression or Sox 9, while in Foxa1/2^-/- cells, Sox 17 represses Pax 6, Pdx1 and Sox 9 expression. These data suggest that both liver and pancreatic genes can be manipulated by combining the Foxa1/2 konckdown, which represses liver and pancreatic fates, while adding exogenous Sox17. Future work will be further characterization of the DRGC composed of endodermal TF, and development of approaches to reprogram HepG2 and other lineages to definitive endoderm by deciphering the DRGC states that stabilize endoderm formation and removing main factors that trigger further endoderm differentiation to its downstream derivatives. Finally, we are developing a physical/mathematical model of how endoderm DRGC function in human endoderm-like cells.