227(hPSC) toward various cell types. However, the induction of endoderm differentia-
tion from mouse embryonic stem cells remains inefficient in comparison with that
of ectoderm and mesoderm cells, requiring the use of endoderm cells that are
induced from embryoid bodies of mouse embryonic stem cells. Embryoid bodies
produce various cell types simultaneously, and the induction of endoderm cells
remains inefficient.
Recent studies of human pluripotent cells have succeeded in efficiently inducing
endoderm differentiation on two-dimensional plastic plates by recapitulating the
developmental cues using stepwise additions of cytokines in cell culture media
(Si-Tayeb et al. 2010b; Kajiwara et al. 2012 ; Sudo 2014 ). Among media supple-
ments, a Rho-associated protein kinase (ROCK) inhibitor prevented cell death dur-
ing seeding of iPS in single cell suspensions, leading to significant improvements of
efficacy, robustness, and purity of differentiation. In addition, activin A was used
with and without Wnt3a to induce definitive endoderm cells that express endoderm
markers such as CXCR4, FoxA2, and Sox17 and do not express Sox7, which is a
marker for extraembryonic endodermal cells. The efficiency of endoderm cell dif-
ferentiation using this method was over 95% according to flow cytometry analyses
of CXCR4 expression. Moreover, during this stage, cells can be differentiated into
hepatic cells, pancreatic cells, or intestinal cells, as discussed in the following sec-
tions (Tremblay and T 2005 ; Franklin et al. 2008 ; Sekine et al. 2012 ; Loh et al.
2014 ; Ikonomou and Kotton 2015 ). From definitive endoderm, hepatic differentia-
tion proceeds more smoothly than pancreatic or intestinal differentiation, reflecting
moderate successes of two-dimensional hepatocyte differentiation.
Hepatic specification is generally monitored according to the expression of
HNF4A and the transcription factors GATA and FoxA. Other key transcription fac-
tors that have been observed during this process include HHEX, PROX1, and TBX3.
Moreover, further differentiation leads to the expression of more specific liver sig-
nature genes, including alpha-fetoprotein, albumin, several glucose metabolic
enzymes such as G6 Pase and PEPCK, and lipid and ammonia metabolic genes.
These cells subsequently express various Cyp genes that are specific to mature
hepatocytes. Although expression levels of these genes remain lower than those in
adult hepatocytes from deceased donors, the quality of these cells varies widely
between product lots, likely reflecting differences between donors and cell prepara-
tion methods. Nonetheless, these endpoint limitations frustrate the use of iPSC-
derived hepatocytes for drug screening and evaluations of hepatic toxicity. Thus,
several issues need to be resolved prior to application of the current strategy to
posthepatic specification. Among obstacles to clinical application, poor differentia-
tion efficiency, immature hepatic function, and poor engraftment rates of mature
hepatocytes are critical issues. Furthermore, freshly isolated adult-donor hepatocytes
may not rescue liver failure in the long term. These limitations are also challenges
to the application of iPSC to other multicellular organs. Consequently, this research
field requires alternative strategies to produce multicellular liver buds that repro-
duce the three-dimensional cellular interactions observed during development
(Ishikawa et al. 2011 ; Sasai 2013 ; Ito et al. 2014 ; Gordillo et al. 2015 ).
12 Liver Regeneration Using Cultured Liver Bud