DEVELOPMENT AND STEM CELLSRESEARCH ARTICLE
Substantial headway has been made in our understanding of the
molecular events that control development of the liver from studies
using chick, Xenopus, zebrafish and the mouse as model organisms
(Zaret and Grompe, 2008; Lemaigre, 2009; Si-Tayeb et al., 2010a).
Advances using these models have been considerable because
many of the key pathways that control hepatogenesis are
evolutionarily conserved. Although such models offer numerous
advantages, each has its own restrictions. For example, the ability
to conduct genetic experiments in birds and frogs is limited, the
extent to which biochemical experiments can be performed in
zebrafish is minimal, and although biochemical, molecular and
genetic analyses have been successful in mice, it remains a time
consuming, expensive and often tedious undertaking.
Embryonic stem (ES) cells and induced pluripotent stem (iPS)
cells can be cultured indefinitely and generally maintain a normal
karyotype. In contrast to cultured primary cells, these pluripotent
stem cells retain the capacity to differentiate into all cell types and
this has been definitively established by using mouse ES and iPS
cells to generate viable animals through tetraploid embryo
complementation (Nagy et al., 1993; Boland et al., 2009; Kang et
al., 2009; Zhao et al., 2009). The fact that pluripotent stem cells can
be induced to differentiate in culture into a plethora of somatic cell
types has raised the possibility of using pluripotent stem cells as an
alternative to embryos to investigate the fundamental molecular
processes that govern cell differentiation. Early experiments using
mouse embryonic stem cells relied on the use of embryoid bodies
to produce differentiated cells; however, this system is somewhat
chaotic, resulting in the generation of heterogeneous cell types that
commonly require cell sorting to obtain useful cell populations.
Even with such caveats, the study of mouse embryoid bodies has
successfully provided insight into liver cell differentiation,
and recent analyses using Hex–/–mouse ES cells successfully
recapitulated the phenotype associated with Hex–/–mouse embryos
(Keng et al., 2000; Martinez Barbera et al., 2000; Bort et al., 2004;
Bort et al., 2006; Kubo et al., 2010).
As mouse ES cells are capable of reproducing the differentiation
of mouse hepatocytes, it raises the issue of whether human ES
(huES) cells could be used to model human hepatocyte formation.
Several laboratories have recently described protocols using huES
cells that allow the production of cells that display functional and
gene expression characteristics that are normally associated with
hepatocytes (Cai et al., 2007; Agarwal et al., 2008; Chiao et al.,
2008; Shiraki et al., 2008; Basma et al., 2009). Based on such
studies, we developed a protocol that facilitates differentiation of
hepatocyte-like cells from both huES cells and iPS cells with
efficiencies >85% (Si-Tayeb et al., 2010b). This approach avoids the
use of embryoid bodies, feeder cells, fetal calf serum and other
undefined components within the culture medium, which results in
the differentiation being highly reproducible and synchronous. Cells
generated using this approach can synthesize glycogen, secrete
albumin, synthesize urea, metabolize indocyanine green, form cell-
cell junctions with apical characteristics, store lipid and uptake low
density lipoprotein. Importantly, the formation of hepatocyte-like
cells from huES or hiPS cells closely resembles the process through
which hepatocyte differentiation occurs in vivo (Agarwal et al., 2008;
Si-Tayeb et al., 2010b). In response to specific inductive cues that
are added to the medium, the human pluripotent stem cell-derived
cells sequentially acquire characteristics of ventral endoderm
(FOXA2, GATA4, SOX17), specified hepatic progenitor cells
(HNF4A), hepatoblasts (AFP) and hepatocytes (Albumin). Because
the differentiation takes place ex vivo, this system potentially offers
a means to manipulate and address the molecular events controlling
Development 138, 4143-4153 (2011) doi:10.1242/dev.062547
© 2011. Published by The Company of Biologists Ltd
1Department of Cell Biology, Neurobiology and Anatomy, Medical College of
Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA. 2David
Geffen School of Medicine, University of California, Los Angeles, CA 90095, USA.
*Author for correspondence (firstname.lastname@example.org)
Accepted 27 July 2011
The availability of pluripotent stem cells offers the possibility of using such cells to model hepatic disease and development. With
this in mind, we previously established a protocol that facilitates the differentiation of both human embryonic stem cells and
induced pluripotent stem cells into cells that share many characteristics with hepatocytes. The use of highly defined culture
conditions and the avoidance of feeder cells or embryoid bodies allowed synchronous and reproducible differentiation to occur.
The differentiation towards a hepatocyte-like fate appeared to recapitulate many of the developmental stages normally
associated with the formation of hepatocytes in vivo. In the current study, we addressed the feasibility of using human
pluripotent stem cells to probe the molecular mechanisms underlying human hepatocyte differentiation. We demonstrate (1) that
human embryonic stem cells express a number of mRNAs that characterize each stage in the differentiation process, (2) that gene
expression can be efficiently depleted throughout the differentiation time course using shRNAs expressed from lentiviruses and
(3) that the nuclear hormone receptor HNF4A is essential for specification of human hepatic progenitor cells by establishing the
expression of the network of transcription factors that controls the onset of hepatocyte cell fate.
KEY WORDS: Hepatocyte differentiation, Human pluripotent stem cells, HNF4A
HNF4A is essential for specification of hepatic progenitors
from human pluripotent stem cells
Ann DeLaForest1, Masato Nagaoka1, Karim Si-Tayeb1, Fallon K. Noto1, Genevieve Konopka2,
Michele A. Battle1and Stephen A. Duncan1,*
human hepatocyte differentiation. In the current study, we therefore
sought to (1) define a characteristic mRNA fingerprint that could be
used to follow the differentiation process, (2) determine whether
gene function could be manipulated in this system to facilitate
mechanistic studies and (3) ascertain whether the nuclear hormone
receptor HNF4A is required for the onset of human hepatocyte
differentiation from pluripotent stem cells.
MATERIALS AND METHODS
HuES cell (WA09, H9) and hiPS cell (C2A) culture was approved by the
MCW SCRO committee. Cells were cultured as described elsewhere (Si-
Tayeb et al., 2010b) except Matrigel was substituted for a recombinant E-
cadherin-IgG Fc fusion protein matrix (Nagaoka et al., 2010) (StemAdhere
STEMCELL Technologies, Vancouver, BC), which ensured homogeneity
of pluripotency within the starting stem cell population. The general
protocol used for differentiation has been described in detail elsewhere (Si-
Tayeb et al., 2010b). However, we noted that the efficiency of definitive
endoderm formation using the original protocol had declined after
obtaining new batches of commercially available B-27 supplement. The
possibility that B-27 quality can be variable was supported by data showing
that the efficacy of culturing primary neurons is dependent on specific
batches of B-27 supplement (Chen et al., 2008). It has been demonstrated
that inhibition of PI-3 kinase is essential for differentiation of definitive
endoderm from human pluripotent stem cells (McLean et al., 2007).
Insulin, which can activate the PI-3-kinase pathway, appears to be present
in B-27 at relatively high concentrations (Price and Brewer, 2001). We,
therefore, included the PI-3-kinase inhibitor LY294002 during endoderm
formation and found that, as reported elsewhere (McLean et al., 2007), the
efficiency of definitive endoderm formation was restored to over 80%.
Moreover, B-27 supplement that is free of insulin has recently become
available and this also supports efficient differentiation of human
pluripotent stem cells towards a definitive endoderm fate.
Plasmid construction and generation of lentiviruses
Oligonucleotides encoding shRNAs (HNF4i2: TGCAGATGTGTGTGAG -
GGACTCACACACATCTGCA; HNF4i3: TGAAGATTGCCAGCAT -
GCGATGCTGGCAATCTTCA) were annealed and cloned into pLL3.7
puro (Rubinson et al., 2003). Lentiviruses were produced by Fugene 6
(Roche Indianapolis, IN) -mediated transfection of HEK293T cells with
plasmids encoding helper functions, (VSVG, RSV/REV and RRE) as
described previously (Konopka et al., 2007). Transduced stem cells were
selected using puromycin (2-8 g/ml) added 2 days post-infection and
‘lines’ were maintained as polyclonal cultures with continuous selection.
Oligonucleotide array analysis
Total RNA was isolated from three independent experiments at each stage
of hepatocyte differentiation using the RNeasy mini kit (QIAgen, Valencia,
CA). Biotinylated cRNA was generated and hybridized to GeneChip
Human Genome U133 plus 2.0 arrays (Affymetrix, Santa Clara, CA).
Images were acquired using a GeneChip Scanner 3000 (Affymetrix, Santa
Clara, CA) and normalized data analyzed using DNA-Chip analyzer
software (Li and Wong, 2001). CEL files (GSE25417) are available
through Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/).
Cells fixed with 4% paraformaldehyde for 30 minutes were permeabilized
with 0.5% Triton X-100, blocked with 3% BSA/PBS and incubated at 4°C
overnight with the following antibodies: HNF4A, 1:500 (sc-6556; Santa
Cruz Biotechnology, Santa Cruz, CA); SOX17, 1:250 (AF1924; R&D
Systems, Minneapolis, MN); FOXA2, 1:250 (H00003170-M12, Novus
Biologicals, Littleton, CO); GATA4, 1:250 (sc-1237; Santa Cruz
Biotechnology, Santa Cruz, CA); or OCT4, 1:500 (sc-9081; Santa Cruz
Biotechnology, Santa Cruz, CA).
RNA was isolated from stem cell-derived hepatocytes and adult cadaveric
livers using the RNeasy mini kit (QIAgen, Valencia, CA), and human fetal
RNA was purchased from BioChain (BioChain, Hayward, CA). Real-time
quantitative polymerase chain reaction (RT-PCR) was performed on an
Applied Biosystems StepOne Plus real time PCR machine using Taqman
(Applied Biosystems, Foster City, CA) or PrimeTime (Integrated DNA
Technologies, Coralville, IA) assays following the manufacturers’
protocols. Data analysis was performed using RT2 Profiler PCR Array
Data Analysis software (SABiosciences, Frederick, MD). Semi-
quantitative reverse-transcriptase PCR was performed as described
previously (Duncan et al., 1997). Oligonucleotide sequences are provided
in Table S1 in the supplementary material.
Proliferation and apoptosis
Proliferation assays were performed using the Click-iT EdU imaging kit
(Invitrogen, Carlsbad, CA) and proliferating cell numbers calculated by
FACS using a Guava EasyCyte Mini System (Millipore, Billerica, MA).
Apoptotic cells were measured using Guava ViaCount Reagent (Millipore,
Billerica, MA). Three independent experiments were performed for each
assay at each stage of differentiation.
Total cellular protein extracts were separated using Nu-PAGE Bis-Tris 4-
12% gradient gels (Invitrogen, Carlsbad, CA), transferred to PVDF
membrane (Bio-Rad, Hercules, CA) using NuPAGE transfer buffer
(Invitrogen, Carlsbad, CA) with 10% methanol/0.01% SDS. Antibodies
were used to detect HNF4A (sc6556; Santa Cruz Biotechnology, Santa
Cruz, CA; 1/100) and -actin (A5441, Sigma Aldrich, St Louis, MO;
Expression profiles reflecting differentiation of
huES cells into hepatocyte-like cells
We have previously described a protocol (Fig. 1A) in which the
differentiation of human pluripotent stem cells occurs in a
synchronous and stepwise fashion that recapitulates many of the
steps known to occur during hepatogenesis (Si-Tayeb et al., 2010b).
For convenience, we named each differentiation stage based on the
expression of proteins that are characteristic of defined
developmental time points. As shown in Fig. 1A, the process
initiates with ‘pluripotent stem cells’, is followed by the formation
of ‘definitive endoderm’, then ‘hepatic specification’, production
of ‘immature hepatocytes’ and finally the generation of ‘mature
hepatocytes’. We use the term ‘mature hepatocyte’ to describe the
relative maturation of the cells and recognize that hepatocyte-like
cells derived from human pluripotent stem cells are not fully
mature and more closely resemble a neonatal state (Si-Tayeb et al.,
2010b). The generation of hepatocyte-like cells from huES cells is
controlled by the sequential addition of activin A, bone
morphogenetic protein (BMP4)/fibroblast growth factor 2 (FGF2),
hepatocyte growth factor (HGF) and oncostatin M (OSM) at 5-day
intervals (Fig. 1A).
We believed that a description of the complete gene expression
profile that accompanies each stage of differentiation in this model
would allow us to assemble a characteristic mRNA fingerprint that
would facilitate the phenotypic characterization of human
hepatocyte differentiation. We therefore isolated RNA from each
stage of the differentiation process and used it to probe Affymetrix
U133 plus 2.0 arrays in three independent experiments; data were
analyzed using DNA-Chip array analysis (D-Chip) software (Li
and Wong, 2001). We initiated our study by determining whether
each stage of differentiation could be marked by expression of a
unique gene set using cluster analyses. As illustrated by the heat
map shown in Fig. 1B, we identified several genes with expression
Development 138 (19)
profiles that characteristically identified each specific stage of
differentiation. In addition, clusters of genes were revealed with
expression profiles that initiated at a specific differentiation stage
and remained expressed as differentiation towards a hepatic fate
was completed (Fig. 1B). The complete gene list showing the
names and expression profiles of genes shown in the heat map are
provided in Table S2 in the supplementary material. Gene ontology
analyses of the array data sets revealed that, as differentiation
progresses, there is a corresponding increase in the number of
expressed genes that encode proteins with roles in biological
processes that were typically associated with hepatocytes, such as
lipid and carbohydrate metabolism, which is consistent with the
cells adopting a hepatic character. Among the genes whose
expression dynamically changes are several that have established
roles in controlling hepatocyte formation and gene expression,
including FOXA2, TBX3, HHEX, HNF4A, GATA4 and GATA6,
which gave confidence that the approach could not only identify
markers, but could potentially reveal authentic regulators of
We next applied two criteria with the goal of identifying genes
whose expression would reliably define each differentiation stage.
First, we considered only genes with expression levels that were
predicted to reproducibly increase by at least fourfold compared
with successive stages of differentiation (P≤0.05). Second, we
reasoned that a fourfold difference in expression is not necessarily
physiologically relevant if the mRNA level encoded by the gene is
extremely low; moreover, we expected that genes with robust
expression would most probably represent easily detectable
markers of a given differentiation stage. We, therefore, used
Affymetrix signal values as an indicator of expression levels. As a
comparative standard, we chose to relate signal levels to those
measured for TBX3 because TBX3 has been shown to be required
for mouse liver development (Suzuki et al., 2008; Ludtke et al.,
2009). The raw signal values for TBX3 obtained from the
oligonucleotide array data appear to mimic Tbx3 mRNA levels
described during mouse hepatogenesis (Ludtke et al., 2009), with
an average signal value of 943.76±145 at day 10, decreasing to
291.42±29 at day 20. We, therefore, discarded any genes whose
signal value was 200 or less at stages of differentiation in which
the gene was considered to be expressed. When these criteria were
applied, a limited number of genes were identified whose
expression initiated at each stage of differentiation (Fig. 1C,D; see
Table S2 in the supplementary material).
Generation of an mRNA signature that defines
hepatocyte differentiation from huES cells
Although oligonucleotide array analyses are useful for capturing
large amounts of information, we felt that we could simplify
phenotypic analyses of the formation of hepatocyte-like cells from
pluripotent stem cells using a subset of representative markers
Hepatic specification from human ES cells
Fig. 1. Identification of mRNA
profiles that are characteristic
of the differentiation of
hepatocyte-like cells from
human ES cells. (A)The
differentiation procedure. (B)Heat
map summarizing relative changes
in mRNA levels at each stage of
differentiation (red, high; blue,
low). (C,D)The total number of
genes whose expression is
predicted to (C) increase at least
fourfold (P≤0.05, Affymetrix signal
of at least 200) on a specific day of
differentiation or (D) whose
expression initiates on a specific
day (fourfold, P≤0.05, Affymetrix
signal of at least 200, compared
with the previous stage) and is
whose induction could be measured by qRT-PCR. We first
considered genes that displayed expression that was specific to a
given differentiation stage (fourfold, P≤0.05, Affymetrix signal of
at least 200). A subset of those genes was selected on the basis of
robust induction followed by strong suppression during the
differentiation process. Additionally, we favored genes with known
roles in cell differentiation or an established expression profile that
is consistent with a role in hepatogenesis. Quantitative real-time
RT-PCR (qRT-PCR) was then performed on RNA isolated from
each stage of differentiation in two independent experiments. Fig.
2 shows that a specific mRNA signature could be measured for day
0 (pluripotent cells), day 5 (definitive endoderm), day 10 (hepatic
specification) and day 20 (mature hepatocyte). In contrast to these
stages, we were unable to detect a specific expression profile for
day 15 (immature hepatic) because genes with induced expression
at this stage generally remained expressed at day 20 (Fig. 1B). We
addressed this problem by developing a second series of qRT-PCR
assays that detected mRNAs whose expression initiated at a
specific stage of differentiation and remained elevated as the cells
completed their adoption of a hepatocyte cell fate. As shown in Fig.
3, qRT-PCR analyses revealed the presence of mRNA signatures
that were characteristic of days 5-20, days 10-20 and days 15-20.
Finally, to ensure that the developed signatures were not merely
specific to hepatocyte differentiation from H9 ES cells, we
confirmed that each signature was reliably expressed at the
appropriate stages during hepatic differentiation from the human
iPS cell line C2a that we had generated previously (Si-Tayeb et al.,
2010b) (see Figs S1 and S2 in the supplementary material).
We predicted that if the mRNAs we had selected as being
characteristic of hepatocyte differentiation were bona fide hepatic
markers, then we should be able to identify their expression in
vivo. We therefore first examined the marker sets that identified
huES cell-derived hepatocyte-like cells at days 10 to 20, 15 to 20
and day 20 by performing real time qRT-PCR on two human fetal
liver samples derived from a 20-week-old and a 38-week-old
conceptus, respectively, as well as one male and one female adult
liver sample (Si-Tayeb et al., 2010b). As expected, with the
exception of AFP, the expression of which was dramatically higher
in the fetal liver samples, all markers were robustly expressed in
all samples tested (see Fig. S3 in the supplementary material). In
support of these qRT-PCR results, we were also able to confirm the
presence of the mRNA marker sets in both fetal and adult human
livers by examining published array data (Guo et al., 2009). To test
for the presence of the d10 marker mRNAs in humans is
challenging because access to an appropriate tissue source is
limited. As an alternative, we addressed whether the mRNAs that
we defined as characteristic of the nascent hepatic progenitors were
expressed in E10.5 mouse livers. RT-PCR analyses indeed revealed
the presence of the corresponding mouse mRNAs in isolated liver
buds (see Fig. S3 in the supplementary material). Cumulatively
these data demonstrate the existence of a marker set that can be
used in conjunction with immunostaining to phenotype the
differentiated status of hepatic cells derived from human
pluripotent stem cells. It is important to note, however, that many
of the genes identified in the marker set are not specific to
hepatocytes and so to be useful each set should be considered as a
whole rather than as individual genes. Although it is difficult to
compare differentiation protocols, we note that a subset of the
markers we have identified have encouragingly been described as
being expressed in the endoderm and its derivatives during
hepatocyte differentiation from ES cells by others (McLean et al.,
2007; Chiao et al., 2008).
Depletion of HNF4A prevents the formation of
hepatic progenitors from huES cell-derived
We next sought to test the suitability of using human pluripotent
stem cells to analyze the molecular mechanisms underlying human
hepatocyte formation. We chose to focus our studies on the
transcription factor HNF4A because of our extensive understanding
of the role of HNF4A during development of the mouse liver (Li
et al., 2000; Parviz et al., 2003; Battle et al., 2006). HNF4A is a
member of the nuclear hormone receptor transcription factor
family, and loss of HNF4A results in an extensive disruption to
expression of genes encoding all aspects of mouse hepatocyte
function (Battle et al., 2006; Bolotin et al., 2010). Although
HNF4A directly controls expression of many hepatic genes (Battle
et al., 2006; Bolotin et al., 2010), it is also crucial in maintaining
the network of transcription factors that is essential for normal
hepatocyte function (Kyrmizi et al., 2006).
Previous studies using mouse embryos have shown that,
although HNF4A is expressed at the onset of extra-embryonic
endoderm formation and continues in the extra-embryonic visceral
endoderm, it is not expressed in the definitive endoderm before
Development 138 (19)
Fig. 2. Quantitative RT-PCR
analyses of differentiation stage-
specific mRNAs. (A-D)Changes in
mRNA levels with characteristic
expression profiles at (A) day 0, (B)
day 5, (C) day 10 and (D) day 20 of
differentiation. Graphs represent the
relative mean expression value and
s.d. normalized to GAPDH from two
hepatic specification (Duncan et al., 1994; Taraviras et al., 1994;
Watt et al., 2007). Consistent with the expression profile in the
mouse, we had previously shown by immunocytochemistry that
HNF4A protein is absent from definitive endoderm generated by
our differentiation of human pluripotent stem cells at day 5, but is
detected after specification of hepatic progenitors by day 10 (Si-
Tayeb et al., 2010b). To confirm that expression of HNF4A within
this system did not initiate until after specification of the hepatic
progenitors, we measured HNF4A mRNA levels by real-time qRT-
PCR and protein levels by immunoblot analyses in day 0
pluripotent cells, day 5 definitive endoderm cells and day 10
specified hepatic progenitors. Fig. 4 shows that both HNF4A
mRNA and protein were undetectable in undifferentiated huES
cells and after formation of definitive endoderm (D5). However,
after addition of BMP4/FGF2 and removal of activin A, HNF4A
mRNA and protein were readily detected at day 10 of the
differentiation procedure (Fig. 4A,B). The onset of HNF4A
expression is therefore strictly associated with the specification of
the hepatic lineage from human pluripotent stem cells.
To determine any role for HNF4A in regulating the formation of
hepatocytes from human pluripotent stem cells, we generated stable
polyclonal huES cell lines expressing shRNAs designed to target
HNF4A by lentiviral transduction. We chose to work with
polyclonal cells to avoid the concern that a clonal line may have an
inherent deficiency. To ensure continued expression of the shRNA
during the entire differentiation process, the puromycin N-acetyl-
transferase (pac) gene, which confers resistance to puromycin, was
included within the provirus, and cells were maintained in
puromycin throughout the differentiation time course. Because
HNF4A is not expressed in pluripotent stem cells, the presence of
shRNAs against HNF4A did not affect the culture of
undifferentiated huES cells. We next compared HNF4A expression
by immunocytochemistry in differentiated hepatic cells derived
from control huES cells infected with lentiviruses that lack any
shRNA (vector), that express a control HNF4A shRNA that fails to
deplete HNF4A mRNA in HepG2 cells (HNF4i2), or that express
an shRNA that efficiently depletes HNF4A in HepG2 cells
(HNF4i3). Fig. 4C shows that, in contrast to control cells, in which
HNF4A can be detected in at least 85% of nuclei upon
differentiation (day 20), HNF4A was undetectable in cells derived
from HNF4i3 ES cells. We confirmed that HNF4A mRNA was
absent during the entire time-course of HNF4i3 differentiation
using real-time qRT-PCR (Fig. 4B). Based on the observation that
HNF4A was undetectable in the ES cell-derived definitive
endoderm (Fig. 4A,B), we predicted that HNF4i3 cells should be
capable of effectively generating definitive endoderm in response
to 5 days of treatment with activin A. As expected, HNF4i3 cells
efficiently differentiated to cells that expressed SOX17, FOXA2
and GATA4 protein, which were detected by immunocytochemistry
(see Fig. S4A in the supplementary material). In addition, real-time
qRT-PCR analyses revealed that mRNAs encoding FOXA2,
FOXA3, SOX17, HHEX and GATA4 (see Fig. S4B in the
supplementary material) as well as our panel of endoderm markers
MIXL1, CALB1, FGF17, CER1 and CCL2 (Fig. 5A) were
expressed at levels that were comparable with control cells. These
data show that HNF4A is dispensable for differentiation of
definitive endoderm from human pluripotent stem cells.
We next addressed whether the requirement for HNF4A during
differentiation of mouse hepatocytes was conserved in human cells.
We examined the expression of genes with familiar hepatic
functions by semi-quantitative RT-PCR following day 20 of
differentiation of H9 ES cells, control HNF4i2 cells and HNF4A-
depleted HNF4i3 cells. Expression of hepatic mRNAs were
detected at similar levels in both H9 cells and HNF4i2 cells,
demonstrating that the introduction of shRNA per se has little
impact on the differentiation of the ES cells toward a hepatic fate
(Fig. 4D). However, in contrast to control differentiations, hepatic
mRNA levels were severely diminished following differentiation of
HNF4i3 cells, which lacked detectable HNF4A mRNA. We also
performed oligonucleotide array analyses on HNF4i3 cells after
completing the 20-day differentiation protocol (n3 independent
experiments). Hepatic character was then defined in an unbiased
manner by using a set of 40 genes whose mRNAs were previously
shown to be expressed only in human livers and were induced
following the formation of hepatocyte-like cells from huES and iPS
cells (Ge et al., 2005; Si-Tayeb et al., 2010b). H9 huES-derived cells
reproducibly expressed the majority of these characteristic hepatic
mRNAs by completion of the differentiation protocol. However,
when HNF4A was depleted, the ability of HNF4i3 cells to express
these mRNAs by day 20 of differentiation was completely blocked
(Fig. 4E). The loss of hepatic character associated with HNF4i3
cells was confirmed by gene ontology analysis using Ingenuity
Pathway Analyses software (see Fig. S5A in the supplementary
material). Following differentiation of control cells, the expression
of many genes associated with hepatic functions, including lipid
metabolism, small molecule biochemistry, carbohydrate metabolism
Hepatic specification from human ES cells
Fig. 3. Quantitative RT-PCR analyses of mRNAs with maintained
expression. (A-C)Changes in mRNA levels that are induced at a
specific stage of differentiation and are maintained throughout the
differentiation process: (A) day 5 to 20, (B) day 10 to 20 and (C) day 15
to 20. Graphs represent the mean expression value and s.d. normalized
to GAPDH levels from two independent differentiations.
and molecular transport, was robust. In HNF4i3-derived cells,
however, expression of genes important in such functions was lost.
Finally, we attempted to determine whether the HNF4A-depleted
cells had adopted an alternate fate that represented a specific
lineage. Studies by others have categorized sets of genes that are
uniquely expressed in specific organs (Ge et al., 2005). As expected,
these analyses confirmed loss of expression of liver genes from
differentiated HNF4i3 cells; however, there was no clear evidence
to suggest that the HNF4i3-derived cells switch to an alternative
lineage-specific fate, at least based on the expression profile (see
Fig. S5B and Table S3 in the supplementary material). These data
imply that whereas HNF4A is crucial for formation of hepatocytes,
in its absence the cells adopt a stable state that cannot be easily
Finally, HNF4A has been shown to bind elements within an
extensive list of target genes in both the mouse and human
genomes (Odom et al., 2004; Odom et al., 2007; Bolotin et al.,
2010). We, therefore, sought to identify whether any of the genes
whose expression was downregulated in the HNF4i3-derived
hepatic cells had previously been shown to contain elements that
were occupied by HNF4A. By cross referencing our list of genes
whose expression is impacted by loss of HNF4A with previously
identified HNF4A targets, we found that 108 of the 562 genes
whose expression is reduced have been shown to house elements
that are occupied by HNF4A in HepG2 cells and/or human
hepatocytes (see Table S4 in the supplementary material).
HNF4A is essential for BMP4/FGF2-induced
specification of hepatic progenitors from huES
cell-derived definitive endoderm
We have previously demonstrated that HNF4A is essential for
hepatocyte differentiation and liver morphogenesis in the mouse
(Parviz et al., 2003; Battle et al., 2006); however, it has been difficult
to determine the exact developmental stage at which HNF4A first
acts during liver development because HNF4A is essential for
gastrulation (Chen et al., 1994; Duncan et al., 1997). Using tetraploid
complementation, we were able to rescue these gastrulation defects
and produce Hnf4a–/–embryos that survived until around E10.5 (Li
et al., 2000). Analysis of the liver buds of these embryos revealed a
loss in expression of the majority of hepatic mRNAs examined,
although, with the exception of PXR and HNF1A, the expression of
genes encoding liver-enriched transcription factors seemed to be
relatively unaffected. Nevertheless, because HNF4A is expressed at
the onset of hepatogenesis in both the mouse (Duncan et al., 1994;
Taraviras et al., 1994) and during huES cell differentiation, we
considered the possibility that HNF4A could control the earliest
stages of hepatic progenitor cell formation during the differentiation
of human pluripotent stem cells. We therefore examined the impact
of HNF4A depletion at each stage of the differentiation of
hepatocytes from huES cells by comparing the mRNA signature
(Figs 2, 3) that was characteristic of each stage between control and
HNF4i3 cells (Fig. 5).
As expected, and consistent with the array data, loss of HNF4A
resulted in a complete disruption to expression of all genes
encoding mRNAs that are characteristic of the day 20 (mature
hepatocyte) stage of differentiation (Fig. 5B). Expression of genes
that initiated at day 15 was also dramatically reduced by the
absence of HNF4A, although expression of one gene, DCN, was
not changed (Fig. 5C). HNF4A mRNA is first detected at day 10 of
the differentiation process, which coincides with the formation of
hepatic progenitors in response to BMP4/FGF2 signaling (Si-Tayeb
et al., 2010b). When expression of the four genes that characterize
the hepatic progenitor stage (day 10) was examined, in contrast to
control cells, depletion of HNF4A prevented induction of every one
of these mRNAs (Fig. 5D). We also determined the abundance of
mRNAs whose expression initiated at day10 and remained
expressed throughout the differentiation time course and found that,
Development 138 (19)
Fig. 4. HNF4A is essential for differentiation of hepatocyte-like
cells from human ES cells. (A)Immunoblot analyses comparing
HNF4A protein levels during differentiation of H9 ES cells (day 0) to
definitive endoderm (day 5) and hepatic progenitor cells (day 10).
(B)Real-time qRT-PCR comparing HNF4A mRNA levels during
differentiation of HNF4i3 cells (black bars) and control H9 ES cells
(white bars). Data are mean±s.d. (C)Immunocytochemistry identified
the presence of HNF4A (red) in control cells (Vector, HNF4i2), but not in
HNF4i3 cells, following differentiation. Scale bar: 100m. (D)Semi-
quantitative RT-PCR revealed that, in contrast to hepatocyte-like cells
derived from control cells (H9, HNF4i2), expression of characteristic
hepatocyte mRNAs was severely disrupted in HNF4i3 cells. CYCG was
used as a loading control. (E)Heat map summarizing oligonucleotide
array analyses (red, high expression; blue, low expression) that
confirmed the loss of expression of hepatocyte-specific mRNAs in
differentiated HNF4i3 cells in contrast to control cells (H9).
in contrast to control cells, most of these mRNAs were generally
reduced in HNF4i3 cells following differentiation. Interestingly, we
identified some exceptions to this general finding, in that EPAS1,
EFNA1, IGF2 and IGFBP3 mRNA levels remained expressed after
depletion of HNF4A; however, it should be noted that expression
of these genes is not restricted to hepatocytes. Additionally, the
gene expression profile of the differentiated HNF4A-depleted cells
was considerably altered as early as day 10 of differentiation;
however, the rates of proliferation and apoptotic cell death were
unaffected (see Fig. S6 in the supplementary material), making it
unlikely that the observed change in expression profile reflects a
selective loss of hepatocytes within the cultures.
During midgestation stages of development, HNF4A has been
shown to play a central role in controlling the stability of a network
of transcription factors that regulate hepatocyte gene expression
(Kyrmizi et al., 2006). However, given that we had observed what
amounted to a loss of hepatic character as early as day 10 of
differentiation, we determined whether HNF4A was essential for the
onset of hepatocyte transcription factor expression by examining the
expression of transcriptional regulators with known roles during liver
development (Si-Tayeb et al., 2010a) in control and HNF4A-
depleted ES cell-derived hepatic progenitors throughout the
differentiation protocol. As shown in Fig. 6, qRT-PCR revealed that
the levels of mRNAs encoding FOXA1, FOXA2, FOXA3, HNF1B,
Hepatic specification from human ES cells
Fig. 5. HNF4A is required for hepatic specification of human ES cells. Real-time qRT-PCR identified the levels of mRNAs defined as being
characteristic of (A) mature hepatocyte, (B) immature hepatocyte, (C) hepatic specification and (D) definitive endoderm throughout the
differentiation of control (H9) or HNF4i3 cells. Results plotted are the mean. Error bars represent s.d. generated from two independent
HNF1A, GATA4, GATA6 and HHEX, all of which have crucial roles
during early stages of hepatic development, were reduced to close to
undetectable levels. TBX3 and PROX1 were reproducibly, yet more
modestly, depleted and their expression recovered as differentiation
progressed. From these data, we conclude that HNF4A is essential
for specifying the fate of the earliest hepatic progenitor cells from
huES and is necessary for both establishing and maintaining the
network of transcription factors that controls the onset of human
hepatocyte cell fate (Fig. 7).
Understanding organogenesis clearly requires analyses to be
conducted in an intact developing organism because one must
consider the role of cell-cell interactions, cell movements and the
establishment of tissue architecture before a reasonable
understanding of organ or tissue formation can be achieved.
However, the specific study of cell differentiation historically has
been advanced by molecular analyses using a variety of cell culture
systems. For example, in the case of hepatocyte differentiation, rat
hepatoma-human fibroblast hybrid cell lines were used to reveal
the ability of HNF4A and HNF1A to control hepatic gene
expression (Griffo et al., 1993). Indeed, it took close to 10
additional years before this result could be recapitulated using
conditional knockout mouse models (Hayhurst et al., 2001; Parviz
et al., 2003). Using cell culture rather than animals not only reduces
the time through which valuable data can be generated, but also
provides relatively homogeneous cell populations and large
amounts of experimental material, which in turn increases the
accuracy of high resolution molecular and biochemical analyses.
Unfortunately, most cells in culture are derived from tumors or are
immortalized, and although such cells have been valuable, tumor
cells commonly have chromosomal abnormalities that promote
proliferation and cell survival, which often confounds
interpretation. In addition, differentiation is a dynamic process, a
consequence of sequential changes in gene expression profiles and
the competency of cells to respond to environmental cues. In most
cases, however, tumor cells are poorly differentiated, or even de-
differentiated, and this limits their usefulness. In the current study,
we demonstrate that human pluripotent stem cells can be used to
efficiently analyze the molecular basis of human hepatocyte
differentiation. As part of the study, we determined the global gene
expression profiles for each stage of the differentiation process.
Based on these data, we defined a subset of mRNAs whose
detection can be used to determine phenotypic changes in
differentiation. In addition to being useful for phenotypic analyses,
defining the gene expression profile for each differentiation stage
will probably facilitate the identification of molecular pathways
with undefined roles during human hepatocyte differentiation; such
studies are currently under way.
We have shown previously that the differentiation of human
pluripotent stem cells into hepatocyte-like cells results in
expression of a large repertoire of genes associated with hepatocyte
function; these data were confirmed in the current study. Indeed,
92% of mRNAs detected in adult and 93% detected in fetal human
livers by oligonucleotide array analyses were also present in the
day 20 huES cell-derived hepatocytes (see Fig. S7 in the
supplementary material). Although the expression of such an
extensive repertoire of hepatic mRNAs is encouraging, we also
noted that the stem cell-derived hepatocyte-like cells expressed
several mRNAs that were not normally associated with adult or
fetal livers (see Fig. S7 in the supplementary material). Although
it is difficult to realize the impact of these results without further
analyses, the observation that such ectopic expression of mRNAs
exists is an important caveat that must be considered if pluripotent
human stem cells are to be considered a source of cells for
transplant. Nevertheless, the ability to deplete candidate
developmental factors efficiently by shRNA in human pluripotent
stem cells and effectively examine the impact on hepatic cell
differentiation is likely to accelerate the discovery of novel
molecular events that control human hepatocyte cell fate.
Development 138 (19)
Fig. 6. HNF4A is essential for
expression of transcription
factors with roles in
controlling the formation of
hepatic progenitor cells. The
level of mRNAs encoding
transcription factors that have
been associated with
differentiation of mouse
hepatocytes was measured by
real-time qRT-PCR in either
control H9 (black bars) or
HNF4i3 (gray bars) cells
throughout differentiation. Data
In addition to providing proof-of-concept that human
pluripotent stem cells can be used as an efficient tool to probe
human cell fate, we exploited the model to reveal a requirement
for HNF4A in controlling the onset of human hepatocyte
differentiation. Our previous studies in the mouse revealed that
HNF4A is required for the hepatoblasts to differentiate to a mature
state (Parviz et al., 2003) and that Hnf4a–/–E10.5 liver buds failed
to express many genes that are characteristic of hepatocyte
function (Li et al., 2000). Although this requirement for HNF4A
during differentiation is conserved during the formation of human
hepatocytes from pluripotent stem cells, the phenotype associated
with HNF4A-depleted huES-derived hepatocytes seems to have
an earlier onset and is more severe compared with that observed
in the mouse. In the mouse, the E10.5 liver bud, which, based on
marker expression, appears approximately equivalent to day 15 of
the huES differentiation protocol, was observed to express the
majority of liver-enriched transcription factors; in the huES cells
differentiation model, however, expression of most liver
transcription factors is severely disrupted coincident with
specification of the hepatic progenitor cells. Although we are
unable to answer definitively why there are differences between
the mouse and human models, we can consider a number of
possible explanations. For example, it is possible that there exist
inherent differences between human and mouse and that the
mouse has in place mechanisms that are capable of at least
partially compensating for loss of HNF4A during hepatic
specification. Such an explanation may be supported by the
observation that a haploinsufficiency of HNF4A in humans results
in diabetes (Yamagata et al., 1996), whereas Hnf4a+/–mice are
euglycemic (Chen et al., 1994). It is also possible that in the
tetraploid experiments that we previously performed to generate
Hnf4a–/–mouse embryos, there existed a selective pressure for
cells that initiated expression of transcription factors in the
absence of HNF4A. It is also worth considering that the
differentiation of huES cells towards the hepatic fate occurs under
relatively simple conditions compared with the complex in vivo
environment of the mouse embryo. Because of the defined nature
of the culture conditions, it is possible that the loss of HNF4A
results in a more severe phenotype in the huES cell model because
signals that rely on, for example, matrix interactions and three-
dimensional structure are likely to be lacking in the culture
system. We are currently exploring this possibility.
HNF4A has been shown to directly regulate the expression of a
large number of target genes in hepatocytes (Battle et al., 2006;
Bolotin et al., 2010). It therefore seems likely that the global loss
of expression of hepatocyte mRNAs partly reflects a direct
requirement for HNF4A in transcriptional regulation through target
promoters. However, it is also important to note that the expression
of several transcription factors that have been implicated in
defining hepatic cell fate in mouse embryos is severely reduced in
the HNF4A-depleted cells. Somewhat surprisingly, the depletion of
HNF4A also significantly reduced expression of several factors that
help establish a state of hepatic competence within the ventral
endoderm, including the FOXA proteins (Lee et al., 2005) and
GATA factors 4 and 6. All of these factors are expressed in the
ventral endoderm before the onset of HNF4A expression in the
liver progenitor cells and their expression is not affected in human
endoderm derived from HNF4A-depleted huES cells (Fig. 5; see
Fig. S4 in the supplementary material). This implies that the
transcription factor network that governs the transition from an
endodermal cell to that of hepatic progenitor cell is in a relatively
plastic state and must be actively maintained by HNF4A as cell
The loss of HNF1B in HNF4A-depleted cells may be
particularly important because Lokmane et al. have recently shown
that liver specification is blocked in Hnf1b-null mouse embryos
(Lokmane et al., 2008). Interestingly, Hnf1b–/–embryos failed to
express HNF4A, suggesting that the closely intertwined regulation
between these factors is a primary mechanism through which the
hepatic transcription factor network is initially established during
specification of the hepatic progenitors. Analyses of promoter
regions of HNF1B and HNF4A have identified reciprocal binding
sites within the regulatory regions of each factors, implying that the
regulatory relationship between the two factors is direct. The issue
of why the expression of so many genes is affected by the loss of
HNF4A is an important one. One explanation is that HNF4A
directly controls the expression of these genes through interacting
with their transcriptional regulatory elements. However, given that
the expression of many of these genes, such as GATA4, GATA6,
HHEX and FOXA2, precedes that of HNF4A in the definitive
endoderm we feel that this explanation may at best be only partly
true. Instead, we favor an alternative explanation, in which
depletion of HNF4A results in the complete loss of hepatic
character by preventing the specification of hepatic progenitors. As
Hepatic specification from human ES cells
Fig. 7. Schematic of the role of HNF4A during the differentiation of human ES cells into hepatocyte-like cells. Definitive endoderm
generated from human ES cells expresses key hepatic transcription factors, including FOXA2, FOXA3, GATA4, GATA6 and HHEX that, based on
studies in the mouse, probably regulate the ability of the endoderm to adopt a hepatic fate. In response to inductive cues, including FGFs and
BMPs, the receptive endoderm expresses HNF4A within the nascent hepatic progenitor cells where it is responsible for the initiation and
maintenance of expression of several hepatic transcription factors, including HNF1B, that control formation of hepatoblasts and their differentiation
towards functional hepatocytes.
a consequence of their inability to follow a hepatic developmental
program, the endodermal cells, in response to inductive signals in
the medium, adopt an alternate stable state. Although the mRNA
levels of some genes whose expression is associated with
hepatocyte function, including IGF2, IGFBP3, EPAS1 and EFNA1
is maintained, such genes are commonly expressed in other
additional cell types, which may explain why their mRNAs can
still be detected in HNF4A-depleted cells.
In summary, our data demonstrate that HNF4A is essential for
establishing hepatic progenitor cells from human pluripotent stem
cell-derived definitive endoderm and that human pluripotent stem
cells offer a valid model with which to study the molecular
mechanisms underlying human hepatocyte differentiation.
Although we noted that a subset of genes appears to be ectopically
expressed and that some mRNAs, for example those encoding
phase 1 and phase 2 enzymes, are expressed at low levels
compared with adult livers, we find that the majority of hepatic
mRNAs are expressed in the expected fashion, and their onset
appears to closely recapitulate the normal developmental profile,
as has been discussed previously (Agarwal et al., 2008). This
system is also one of the few that offers direct access to the
differentiation of human hepatocytes. As techniques continue to
improve for the generation of human iPS cells (Anokye-Danso et
al., 2011), we believe that similar procedures will facilitate the
study of the mechanisms and possible treatments of inborn errors
in hepatic metabolism.
The authors thank Stephanie Lohman, Jixuan Li and Thomas Wagner for
This work was supported by gifts from the Marcus Family, from the Phoebe R.
and John D. Lewis Foundation, from the Sophia Wolf Quadracci Memorial
Fund, from the Dr James Guhl Memorial Fund and from the Advancing a
Healthier Wisconsin Fund, and by NIH grants DK55743, DK087377,
HG006398 and HL094857. Deposited in PMC for release after 12 months.
Competing interests statement
The authors declare no competing financial interests.
Supplementary material for this article is available at
Agarwal, S., Holton, K. L. and Lanza, R. (2008). Efficient differentiation of
functional hepatocytes from human embryonic stem cells. Stem Cells 26, 1117-
Anokye-Danso, F., Trivedi, C. M., Juhr, D., Gupta, M., Cui, Z., Tian, Y., Zhang,
Y., Yang, W., Gruber, P. J., Epstein, J. A. et al. (2011). Highly efficient miRNA-
mediated reprogramming of mouse and human somatic cells to pluripotency.
Cell Stem Cell 8, 376-388.
Basma, H., Soto-Gutierrez, A., Yannam, G. R., Liu, L., Ito, R., Yamamoto, T.,
Ellis, E., Carson, S. D., Sato, S., Chen, Y. et al. (2009). Differentiation and
transplantation of human embryonic stem cell-derived hepatocytes.
Gastroenterology 136, 990-999.
Battle, M. A., Konopka, G., Parviz, F., Gaggl, A. L., Yang, C., Sladek, F. M.
and Duncan, S. A. (2006). Hepatocyte nuclear factor 4alpha orchestrates
expression of cell adhesion proteins during the epithelial transformation of the
developing liver. Proc. Natl. Acad. Sci. USA 103, 8419-8424.
Boland, M. J., Hazen, J. L., Nazor, K. L., Rodriguez, A. R., Gifford, W., Martin,
G., Kupriyanov, S. and Baldwin, K. K. (2009). Adult mice generated from
induced pluripotent stem cells. Nature 461, 91-94.
Bolotin, E., Liao, H., Ta, T. C., Yang, C., Hwang-Verslues, W., Evans, J. R.,
Jiang, T. and Sladek, F. M. (2010). Integrated approach for the identification of
human hepatocyte nuclear factor 4alpha target genes using protein binding
microarrays. Hepatology 51, 642-653.
Bort, R., Martinez-Barbera, J. P., Beddington, R. S. and Zaret, K. S. (2004).
Hex homeobox gene-dependent tissue positioning is required for organogenesis
of the ventral pancreas. Development 131, 797-806.
Bort, R., Signore, M., Tremblay, K., Martinez Barbera, J. P. and Zaret, K. S.
(2006). Hex homeobox gene controls the transition of the endoderm to a
pseudostratified, cell emergent epithelium for liver bud development. Dev. Biol.
Cai, J., Zhao, Y., Liu, Y., Ye, F., Song, Z., Qin, H., Meng, S., Chen, Y., Zhou, R.,
Song, X. et al. (2007). Directed differentiation of human embryonic stem cells
into functional hepatic cells. Hepatology 45, 1229-1239.
Chen, W. S., Manova, K., Weinstein, D. C., Duncan, S. A., Plump, A. S.,
Prezioso, V. R., Bachvarova, R. F. and Darnell, J. E., Jr (1994). Disruption of
the HNF-4 gene, expressed in visceral endoderm, leads to cell death in
embryonic ectoderm and impaired gastrulation of mouse embryos. Genes Dev.
Chen, Y., Stevens, B., Chang, J., Milbrandt, J., Barres, B. A. and Hell, J. W.
(2008). NS21: re-defined and modified supplement B27 for neuronal cultures. J.
Neurosci. Methods 171, 239-247.
Chiao, E., Elazar, M., Xing, Y., Xiong, A., Kmet, M., Millan, M. T., Glenn, J. S.,
Wong, W. H. and Baker, J. (2008). Isolation and transcriptional profiling of
purified hepatic cells derived from human embryonic stem cells. Stem Cells 26,
Duncan, S. A., Manova, K., Chen, W. S., Hoodless, P., Weinstein, D. C.,
Bachvarova, R. F. and Darnell, J. E., Jr (1994). Expression of transcription
factor HNF-4 in the extraembryonic endoderm, gut, and nephrogenic tissue of
the developing mouse embryo: HNF-4 is a marker for primary endoderm in the
implanting blastocyst. Proc. Natl. Acad. Sci. USA 91, 7598-7602.
Duncan, S. A., Nagy, A. and Chan, W. (1997). Murine gastrulation requires HNF-
4 regulated gene expression in the visceral endoderm: tetraploid rescue of HNF-
4–/– embryos. Development 124, 279-287.
Ge, X., Yamamoto, S., Tsutsumi, S., Midorikawa, Y., Ihara, S., Wang, S. M.
and Aburatani, H. (2005). Interpreting expression profiles of cancers by
genome-wide survey of breadth of expression in normal tissues. Genomics 86,
Griffo, G., Hamon-Benais, C., Angrand, P. O., Fox, M., West, L., Lecoq, O.,
Povey, S., Cassio, D. and Weiss, M. (1993). HNF4 and HNF1 as well as a panel
of hepatic functions are extinguished and reexpressed in parallel in
chromosomally reduced rat hepatoma-human fibroblast hybrids. J. Cell Biol.
Guo, Y., Zhang, X., Huang, J., Zeng, Y., Liu, W., Geng, C., Li, K. W., Yang, D.,
Wu, S., Wei, H. et al. (2009). Relationships between hematopoiesis and
hepatogenesis in the midtrimester fetal liver characterized by dynamic
transcriptomic and proteomic profiles. PLoS One 4, e7641.
Hayhurst, G. P., Lee, Y. H., Lambert, G., Ward, J. M. and Gonzalez, F. J.
(2001). Hepatocyte nuclear factor 4alpha (nuclear receptor 2A1) is essential for
maintenance of hepatic gene expression and lipid homeostasis. Mol. Cell. Biol.
Kang, L., Wang, J., Zhang, Y., Kou, Z. and Gao, S. (2009). iPS cells can support
full-term development of tetraploid blastocyst-complemented embryos. Cell
Stem Cell 5, 135-138.
Keng, V. W., Yagi, H., Ikawa, M., Nagano, T., Myint, Z., Yamada, K., Tanaka,
T., Sato, A., Muramatsu, I., Okabe, M. et al. (2000). Homeobox gene Hex is
essential for onset of mouse embryonic liver development and differentiation of
the monocyte lineage. Biochem. Biophys. Res. Commun. 276, 1155-1161.
Konopka, G., Tekiela, J., Iverson, M., Wells, C. and Duncan, S. A. (2007).
Junctional adhesion molecule-A is critical for the formation of pseudocanaliculi
and modulates E-cadherin expression in hepatic cells. J. Biol. Chem. 282, 28137-
Kubo, A., Kim, Y. H., Irion, S., Kasuda, S., Takeuchi, M., Ohashi, K., Iwano,
M., Dohi, Y., Saito, Y., Snodgrass, R. et al. (2010). The homeobox gene Hex
regulates hepatocyte differentiation from embryonic stem cell-derived
endoderm. Hepatology 51, 633-641.
Kyrmizi, I., Hatzis, P., Katrakili, N., Tronche, F., Gonzalez, F. J. and Talianidis,
I. (2006). Plasticity and expanding complexity of the hepatic transcription factor
network during liver development. Genes Dev. 20, 2293-2305.
Lee, C. S., Friedman, J. R., Fulmer, J. T. and Kaestner, K. H. (2005). The
initiation of liver development is dependent on Foxa transcription factors. Nature
Lemaigre, F. P. (2009). Mechanisms of liver development: concepts for
understanding liver disorders and design of novel therapies. Gastroenterology
Li, C. and Wong, W. H. (2001). Model-based analysis of oligonucleotide arrays:
expression index computation and outlier detection. Proc. Natl. Acad. Sci. USA
Li, J., Ning, G. and Duncan, S. A. (2000). Mammalian hepatocyte differentiation
requires the transcription factor HNF-4alpha. Genes Dev. 14, 464-474.
Lokmane, L., Haumaitre, C., Garcia-Villalba, P., Anselme, I., Schneider-
Maunoury, S. and Cereghini, S. (2008). Crucial role of vHNF1 in vertebrate
hepatic specification. Development 135, 2777-2786.
Ludtke, T. H., Christoffels, V. M., Petry, M. and Kispert, A. (2009). Tbx3
promotes liver bud expansion during mouse development by suppression of
cholangiocyte differentiation. Hepatology 49, 969-978.
Development 138 (19)
Martinez Barbera, J. P., Clements, M., Thomas, P., Rodriguez, T., Meloy, D., Download full-text
Kioussis, D. and Beddington, R. S. (2000). The homeobox gene Hex is
required in definitive endodermal tissues for normal forebrain, liver and thyroid
formation. Development 127, 2433-2445.
McLean, A. B., D’Amour, K. A., Jones, K. L., Krishnamoorthy, M., Kulik, M. J.,
Reynolds, D. M., Sheppard, A. M., Liu, H., Xu, Y., Baetge, E. E. et al.
(2007). Activin a efficiently specifies definitive endoderm from human embryonic
stem cells only when phosphatidylinositol 3-kinase signaling is suppressed. Stem
Cells 25, 29-38.
Nagaoka, M., Si-Tayeb, K., Akaike, T. and Duncan, S. A. (2010). Culture of
human pluripotent stem cells using completely defined conditions on a
recombinant E-cadherin substratum. BMC Dev. Biol. 10, 60.
Nagy, A., Rossant, J., Nagy, R., Abramow-Newerly, W. and Roder, J. C.
(1993). Derivation of completely cell culture-derived mice from early passage
embryonic stem cells. Proc. Natl. Acad. Sci. USA 90, 8424-8428.
Odom, D. T., Zizlsperger, N., Gordon, D. B., Bell, G. W., Rinaldi, N. J., Murray,
H. L., Volkert, T. L., Schreiber, J., Rolfe, P. A., Gifford, D. K. et al. (2004).
Control of pancreas and liver gene expression by HNF transcription factors.
Science 303, 1378-1381.
Odom, D. T., Dowell, R. D., Jacobsen, E. S., Gordon, W., Danford, T. W.,
MacIsaac, K. D., Rolfe, P. A., Conboy, C. M., Gifford, D. K. and Fraenkel, E.
(2007). Tissue-specific transcriptional regulation has diverged significantly
between human and mouse. Nat. Genet. 39, 730-732.
Parviz, F., Matullo, C., Garrison, W. D., Savatski, L., Adamson, J. W., Ning, G.,
Kaestner, K. H., Rossi, J. M., Zaret, K. S. and Duncan, S. A. (2003).
Hepatocyte nuclear factor 4alpha controls the development of a hepatic
epithelium and liver morphogenesis. Nat. Genet. 34, 292-296.
Price, P. J. and Brewer, G. J. (2001). Serum-free media for neural cell cultures. In
Protocols for Neural Cell Culture (ed. S. Fedoroff and A. Richardson), pp. 255-
264. Totowa, NJ: Humana Press.
Rubinson, D. A., Dillon, C. P., Kwiatkowski, A. V., Sievers, C., Yang, L.,
Kopinja, J., Rooney, D. L., Ihrig, M. M., McManus, M. T., Gertler, F. B. et al.
(2003). A lentivirus-based system to functionally silence genes in primary
mammalian cells, stem cells and transgenic mice by RNA interference. Nat.
Genet. 33, 401-406.
Shiraki, N., Umeda, K., Sakashita, N., Takeya, M., Kume, K. and Kume, S.
(2008). Differentiation of mouse and human embryonic stem cells into hepatic
lineages. Genes Cells 13, 731-746.
Si-Tayeb, K., Lemaigre, F. P. and Duncan, S. A. (2010a). Organogenesis and
development of the liver. Dev. Cell 18, 175-189.
Si-Tayeb, K., Noto, F. K., Nagaoka, M., Li, J., Battle, M. A., Duris, C., North, P.
E., Dalton, S. and Duncan, S. A. (2010b). Highly efficient generation of human
hepatocyte-like cells from induced pluripotent stem cells. Hepatology 51, 297-
Suzuki, A., Sekiya, S., Buscher, D., Izpisua Belmonte, J. C. and Taniguchi, H.
(2008). Tbx3 controls the fate of hepatic progenitor cells in liver development by
suppressing p19ARF expression. Development 135, 1589-1595.
Taraviras, S., Monaghan, A. P., Schutz, G. and Kelsey, G. (1994).
Characterization of the mouse HNF-4 gene and its expression during mouse
embryogenesis. Mech. Dev. 48, 67-79.
Watt, A. J., Zhao, R., Li, J. and Duncan, S. A. (2007). Development of the
mammalian liver and ventral pancreas is dependent on GATA4. BMC Dev. Biol.
Yamagata, K., Furuta, H., Oda, N., Kaisaki, P. J., Menzel, S., Cox, N. J.,
Fajans, S. S., Signorini, S., Stoffel, M. and Bell, G. I. (1996). Mutations in the
hepatocyte nuclear factor-4alpha gene in maturity-onset diabetes of the young
(MODY1). Nature 384, 458-460.
Zaret, K. S. and Grompe, M. (2008). Generation and regeneration of cells of the
liver and pancreas. Science 322, 1490-1494.
Zhao, X. Y., Li, W., Lv, Z., Liu, L., Tong, M., Hai, T., Hao, J., Guo, C. L., Ma, Q.
W., Wang, L. et al. (2009). iPS cells produce viable mice through tetraploid
complementation. Nature 461, 86-90.
Hepatic specification from human ES cells