As shown recently by our group (
Koegler et al.) transplantation of USSC in a non-injury model, the preimmune fetal sheep, resulted in more than 20% albumin-producing human parenchymal hepatic cells with absence of cell fusion. Here we examined whether we are able to trigger USSC in vitro into the an endodermal differentiation pathway applying protocols described for both embryonal as well as adult stem cell. The following matrix/growth factors/organic substances were applied: Neurobasal medium/supplement B27/bFGF with and without nicotinamid, on fibronectin, laminin or matrigel for differentiation into the pancreatic development as well as HGF, NGF, bFGF and HGF, FGF4 with or without Oncostatin M to induce liver development. 15 different USSC lines were differentiated for 1, 2, 3, 4 and 6 weeks. Primers were designed with the strategy to define stages of endodermal development on the basis of the embryonic cell development from mouse and human. Therefore the following markers were established: GATA 4, HNF-1, HNF3ß and HNF 4a to define the embryonic and visceral (extra-embryonic) endoderm, a common precursor phenotype for both liver/exocrine pancreas development. To further assess differentiation into liver cells, a-1 antitrypsin, a-fetoprotein, albumin, HGF, Cyp2B6, Cyp3A4, Gys2 and PDX-1, PAX4, ISL-1, NKx6.1, NeuroD, insulin to determine differentiation into the pancreatic development as well as epithelial markers as CK8, CK18, CK19 were analyzed by RT-PCR and subsequent hybridization using gene-specific oligonucleotide probes. USSC were tested negativ for the majority of the markers, only HGF and cytokeratin markers CK8/18 and CK19 were tested positiv. In vitro differentiation shows that USSC never expressed Neuro D, HNF1, HNF3b, PDX-1, PAX4, insulin and a-fetoprotein, but do express depending on the culture conditions common endodermal precursor markers HNF4a, GATA4 (but not HNF1 or HNF3ß) as well as albumin, Cyp2B6, Cyp3A4, Gys2 (liver development) and Nkx6.1 and ISL-1 (pancreatic development. It is very interesting to note that ISL-1, required for the formation of the dorsal mesenchym and essential for the dorsal exocrine pancreas development is strong expressed, although all other markers as PDX-1 and PAX4 were always tested negative. We have never observed the expression of a-fetoprotein, which might be explained by the kinetic of this factor (Young-Yang, Nat.Cell Biology 2004). In summary, the results show that under the vitro conditions applied, only an endodermal precursor phenotype could be established. Therefore at present co-cultivation experiments with injured hepatocytes from sheep/rat/human are performed to mimic the biological niche in vitro as shown by Young-Yang et al. Western blots with an human albumin specific antibody (moAb, clone HAS-11, Sigma) revealed that this approach resulted in an strong expression of albumin in USSC. Moreover, this co-cultivation model is useful for the identification and characterization of factors who are responsible for the biological niche in vivo.
J. Exp. Med.
2005, The American Society of Hematology