Direct reprogramming of mouse fibroblasts
to neural progenitors
Janghwan Kima,b, Jem A. Efea, Saiyong Zhua, Maria Talantovac, Xu Yuana, Shufen Wangd,e, Stuart A. Liptonc,
Kang Zhangd,e, and Sheng Dinga,f,1
aDepartment of Chemistry, The Scripps Research Institute, La Jolla, CA 92037;bDevelopment and Differentiation Research Center, Korea Research Institute of
Bioscience and Biotechnology, Yuseong-gu, Daejeon, 305-806, Republic of Korea;cDel E. Webb Center for Neuroscience, Aging, and Stem Cell Research,
Sanford-Burnham Medical Research Institute, La Jolla, CA 92037;dInstitute for Genomic Medicine and Shiley Eye Center, University of California at San Diego,
La Jolla, CA 92093;eMolecular Medicine Research Center and Department of Ophthalmology, West China Hospital, Sichuan University, Chengdu 610065,
China; andfGladstone Institute of Cardiovascular Disease, Department of Pharmaceutical Chemistry, University of California, San Francisco, CA 94158
Edited* by Clifford J. Tabin, Harvard Medical School, Boston, MA, and approved March 29, 2011 (received for review February 24, 2011)
The simple yet powerful technique of induced pluripotency may
eventually supply a wide range of differentiated cells for cell
therapy and drug development. However, making the appropriate
cells via induced pluripotent stem cells (iPSCs) requires reprogram-
ming of somatic cells and subsequent redifferentiation. Given how
arduous and lengthy this process can be, we sought to determine
whether it might be possible to convert somatic cells into lineage-
specific stem/progenitor cells of another germ layer in one step,
bypassing the intermediate pluripotent stage. Here we show that
transient induction of the four reprogramming factors (Oct4, Sox2,
Klf4, and c-Myc) can efficiently transdifferentiate fibroblasts into
functional neural stem/progenitor cells (NPCs) with appropriate
are directly converted from fibroblasts), transdifferentiated NPCs
have the distinct advantage of being expandable in vitro and
retaining the ability to give rise to multiple neuronal subtypes and
glial cells. Our results provide a unique paradigm for iPSC-factor–
ified to serve as a general platform for transdifferentiation.
and in vitro (3, 4) has been reported, until recently these methods
ectoderm, mesoderm, andendoderm. However,the generation of
iN cells (5) using neural-specific transcription factors has estab-
lished that interlineage transdifferentiation is also possible in
vitro. These transdifferentiation schemes entail overexpression of
different sets of lineage-specific transcription factors. A more re-
cent example reported single-factor transdifferentiation of fibro-
blasts into blood precursors using long-term ectopic expression of
OCT4 (6); through extensive binding to the regulatory regions of
regulating hematopoietic programs acting as a lineage-specific
transcription factor in this context. An important aspect of this
study is the ability to generate a mitotically active progenitor
cells—a critical feat that has yet to be accomplished in trans-
differentiation to neural and endoderm lineages.
In an effort to devise a more general transdifferentiation strat-
egy that might give rise to a broad array of unrelated cell types—
including lineage-specific precursors—we attempted to direct
conventional four iPSC-factor–based reprogramming (7, 8) to-
ward alternative outcomes. Specifically, studies indicating that
iPSCs are generated in a sequential and stochastic manner (9–11)
early and epigenetically highly unstable state induced by the
rise to a multitude of cell types (12) with more stable epigenetic
profiles. In this context, induced pluripotency is only one—and
perhaps among the less likely—of many possible outcomes. In-
deed, studies have found partially or incompletely reprogrammed
cells expressing multiple lineage-specific markers (7, 13–17),
lthough successful transdifferentiation from one cell type to
although these cells did not appear to represent physiologically
relevant cell types. Accordingly, we hypothesized that it might be
possible to deliberately bias the early reprogramming process to-
ward a defined cell type by using inductive and/or permissive sig-
naling conditions, after which the desired cells could be selected
and/or expanded. On the basis of this same hypothesis and using
a similar methodology, our group has recently shown that direct
reprogramming into cardiomyocytes can be achieved (18).
In the present study, we have directly reprogrammed fibro-
blasts to functional neural stem/progenitor cells (NPCs) over an
abbreviated period of four-factor induction. This direct reprog-
ramming process is clearly distinct from conventional reprog-
ramming to iPSCs or forward differentiation of pluripotent cells.
Our findings not only represent a unique successful transdiffer-
entiation of somatic cells into proliferating NPCs, but also form
the basis of a methodology for interlineage transdifferentiation
into multi- or oligopotent cells.
To rigorously test our hypothesis, we attempted an interlineage
transdifferentiation from fibroblasts to NPCs using the doxycy-
cline (dox)-inducible secondary mouse embryonic fibroblast
(MEF) system (11, 19, 20). Inducible overexpression allows pre-
cise temporal control over the expression of the conventional
iPSC-reprogramming factors, avoiding potentially detrimental
effects arising from their constitutive overexpression. To ensure
the survival of MEFs during the beginning of the reprogramming
procedure, they were kept in MEF and reprogramming initiation
medium (RepM-Ini; without leukemia inhibitory factor, LIF) for
the first 3–6 d of dox treatment. Thereafter, neural reprogram-
ming medium (RepM-neural) was applied to induce the genera-
tion and/or proliferation of nascent NPCs. RepM-neural contains
FGF2, EGF, and FGF4 to support NPCs (21, 22). We tried dox
reprogramming factor expression (Fig. 1A). We initially found
that a minimum of 3 d of dox treatment was sufficient to obtain
Pax6+colonies after an additional 8–9 d in culture in RepM-
neural (Fig. 1B). These colonies typically contained several hun-
dred cells that nearly homogenously expressed promyelocytic
leukemia zinc finger (PLZF), a rosette NSC marker (23), and
Pax6, an early neural transcription factor (24) (Fig. 1 B–D).
Extending dox treatment to 6 d increased the number of colonies
(0.69 and 0.5% colony generation efficiency, n = 2). Surprisingly,
Author contributions: J.K., K.Z., and S.D. designed research; J.K., J.A.E., S.Z., M.T., X.Y.,
S.W., and S.A.L. performed research; J.K., J.A.E., S.Z., M.T., X.Y., S.A.L., and S.D. analyzed
data; and J.K., K.Z., and S.D. wrote the paper.
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
1To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| May 10, 2011
| vol. 108
| no. 19www.pnas.org/cgi/doi/10.1073/pnas.1103113108
we found that almost 100% of colonies showed neural trans-
differentation regardless of the tested duration of dox treatment
(95.9% ± 4.6 SEM, n = 3, Fig. 1C).
We found typical neural rosette structures, PLZF expression,
and luminal expression of ZO-1 (23) in the transdifferentiated
some of which coexpress the early cortical neuronal marker Dcx,
andeven thedopaminergic neuronal marker tyrosine-hydroxylase
(Fig. 1D). Importantly, these expression profiles are not observed
in iN cells (5). Flow cytometry analysis of the transdifferentiated
cells revealed that a population of cells expressing Prominin-1,
C). This population, in which various neural progenitors (25) ap-
in Fig. S1A). Notably, this population did not contain a significant
probably reflects parallel paths of reprogramming, some of which
result in a more direct transdifferentiation to mature neuronal
fates—e.g., because mature neuronal cells coexist with trans-
differentiated NPCs (Fig. 1D) and PSA-NCAM–expressing cells
were generated as early as day 7 in the newly emerging population
which have typically been found to express Pax6 (23)—were iso-
lated by manual picking and subcultured en bloc in conventional
NPC medium containing FGF2 and EGF for expansion of NPCs
(Fig. 1B). Each small cluster could form a colony within several
days after isolation; however, despite mild enzymatic passaging
and the avoidance of single cell dissociation, the NPCs appeared
to lose their ability to form colonies within 3–5 passages. At this
point, the NPCs were dissociated into single cells then cultured in
N2 medium without any cytokines for further differentiation.
After 1–2 wk of spontaneous differentiation, we observed NeuN
and Map2-expressing mature neurons, GABAergic neurons, and
GFAP-expressing astrocytes (Fig. 1E). The fully differentiated
neurons that had formed by day 20 showed characteristic ex-
pression of synapsin I (Fig. 1F). To more rigorously characterize
these apparently mature neurons, we examined their electro-
physiological properties by patch clamp technique. The neurons
showed evoked action potentials and fast Na+currents (INa+)
(n = 3 out of 4 cells recorded, Fig. 1 G and H). We also detected
established synaptic connections and spontaneous excitatory
postsynaptic currents (EPSCs, Fig. 1I), confirming that they are
functional neurons capable of forming synapses. These results
collectively suggest that the Pax6- and PLZF-expressing NPCs are
functionalandcanbedifferentiated intomatureneurons andglial
cells. Compared with the circuitous process of reprogramming to
iPSCs and subsequent redifferentiation to NPCs, our induced
transdifferentiation method—requiring only a single step that is
complete within 13 d and showing almost 100% newly generated
colonies that are mostly composed of NPCs—is a highly efficient,
direct, and rapid process for generating NPCs.
Importantly, the Nanog-GFP marker harbored by the second-
ary MEF cells was never expressed during transdifferentiation
(Fig. S2). Indeed, its activation was found to require at least 9
d of dox treatment followed by a week of subsequent culture (19).
Thus, we concluded that no fully reprogrammed iPSCs were
generated within the time span of our experiment, and that the
Pax6 and PLZF-expressing cells were directly reprogrammed
from fibroblasts instead of redifferentiation from intermediate
pluripotent cells. To better understand and characterize this
transdifferentiation process, we performed the following addi-
First, we sought to more definitively rule out the possibility
that the generation of NPCs might first require the formation
of transient pluripotent intermediates. To this end, we tested
RepM-neural treatment directly on the iPSCs (NGFP1), which
were used in blastocyst injections to make the secondary MEF
cells, hypothesizing that we would get results similar to MEFs if
the generation of NPCs relied on the generation of a small
number of pluripotent cells early in the transdifferentiation
process. However, we found that using iPSCs as a starting pop-
Scheme for the transdifferentiation of dox-inducible secondary MEF cells into
the indicated number of days. Different media were added sequentially as
described in Materials and Methods. (B) Pax6 immunostaining on day 13 of
colonies arising from the indicated durations of dox treatment; 8 μg/mL dox
was used for the 3-d treatment. (C) Number of PLZF-expressing colonies
generated with different durations of dox treatment, as analyzed on day 13.
Percentages of PLZF+colonies over total colonies are shown. (D) Immunos-
taining of colonies on day 13 with various neural or neuronal markers. (E)
Immunostaining of spontaneously differentiated cells from isolated colonies
on day 13 with various mature neuronal or glial markers. Long-term differ-
entiated neurons showed characteristic synapsin I expression patterns (F) and
generated a full train of action potentials during injection of 20 pA current
(whole-cell configuration in the current-clamp mode) (G) as well as fast Na+
currents and outward K+currents (whole-cell configuration in the voltage-
clamp mode) (H). Mature neurons also showed glutamate-mediated excit-
neurons. EPSC amplitude was partially blocked by the AMPA-type receptor
antagonist NBQX (I, ii), and the remaining component was completely abro-
gated by the NMDA-type receptor antagonist APV (I, iii) (whole-cell configu-
ration in the voltage-clamp mode). (Scale bars, 100 μm.)
Transdifferentiation by transient expression of the conventional four
Kim et al. PNAS
| May 10, 2011
| vol. 108
| no. 19
ulation resulted in a highly complex mixture of neuroectodermal
(Sox1+and Pax6+), endodermal (Sox17+), and mesodermal
(T+) cells (Fig. 2 A and B). On the contrary, most colonies
transdifferentiated from MEFs were almost entirely composed
of cells expressing Sox1, the earliest neuroectodermal marker
(21), and Pax6 (Fig. 2 A and B). RT-PCR analysis corroborated
these results, as Sox1 and Pax6 expression was exclusive in
transdifferentiation (Fig. 2D). As above, cells generated from
iPSCs displayed a much more arbitrary profile of lineage-specific
marker gene expression (Fig. 2 E and F). Although conventional
neural differentiation does not rely on NPC-supporting cyto-
kines, we could not do a comparison in their absence because no
cell growth of transdifferentiated cells was ever observed without
these cytokines. These results show that transdifferentiated NPCs
arise directly from fibroblasts without any dependence on the
generation of pluripotent intermediates.
Second, we analyzed neural marker expression over time to
pinpoint the onset of neural specification during transdifferen-
tiation. As early as day 3, Pax6 expression was increased by 5.6-
fold compared with day 0 (D0) MEFs. Sox1 expression started on
day 5 and increased dramatically thereafter (Fig. 2D). Critically,
during iPSC differentiation, Sox1 expression also begins on day 5
(Fig. 2F). This latter finding implies that if transdifferentiation
relied on the generation of a pluripotent intermediate, this in-
termediate would have to arise at least 5 d before initial Sox1
expression. Because Sox1 expression commences on day 5 during
transdifferentiation, this leaves no time for the reprogramming of
MEFs to pluripotency. Further, even minute amounts of Nanog
expression are not detected until much later in the transdiffer-
entiation process (Fig. 2C). In short, Sox1+NPCs cannot be
arising from putative pluripotent intermediates generated during
We also investigated the possible generation of epiblast stem
cells (EpiSCs) during the transdifferentiation because FGF2
supports self-renewal of EpiSCs, but could not detect any FGF5
expression (26) at any point in the process (Fig. 2D). In addition,
Sox17 and Brachyury (T), which are not only lineage-specific
markers but are also highly expressed in EpiSCs (27), were not
detected either (Fig. 2D). Collectively, our results strongly sup-
port the notion that transdifferentiated NPCs are derived from
nonpluripotent intermediate cells arising during the early phase
of the process.
Third, we analyzed changes in histone modification marks in
the Sox1 and Oct4 promoters by chromatin immunoprecipitation
(Fig. 3). It was found that Sox1 promoter was characterized by
the bivalent chromatin domains enriched in both trimethyl his-
tone H3 Lys-4 (H3-K4me3) and trimethyl histone H3 Lys-27
(H3-K27me3) (28). Our analysis showed that the Sox1 promoter
region underwent a burst of activating H3K4 trimethylation on
day 8, the level of which reached that of control adult NPCs by
day 12. Repressive H3K27me3 marks showed a concomitant
down-regulation from day 4 onward. Slightly higher amounts of
H3K27me3 than adult NPCs on day 12 may be the result of some
unreprogrammed cells from the whole culture harvest. Impor-
tantly, the Oct4 promoter was persistently marked by H3K27me3
and showed a level of H3K4me3 similar to adult NPCs. These
neural progenitor cells. (A) Experimental overview and cartoon representa-
tion of results from panels B, D, and F. Cells after 9 d of differentiation from
iPSCs, or after 13 d of transdifferentiation show the indicated marker ex-
pression profiles. (B) Day 9 immunostaining of cells differentiated from
iPSCs, or transdifferentiated NPCs on day 13, with antibodies against the
indicated markers. Pax6 and Sox1 demarcate early neuroectoderm. Sox17 is
indicative of endoderm. T expression is early mesodermal. (Scale bars, 100
μm.) (C–F) qRT-PCR analysis of the indicated markers’ expression in cells
harvested at multiple time points during the direct reprogramming process
(C and D) and differentiation from iPSCs (E and F). Pluripotency genes (C and
E) and lineage-specific genes (D and F) are shown in separate graphs. All
values are relative to expression in iPSCs.
Direct reprogramming is a highly efficient method of deriving pure
promoter loci were analyzed by chromatin immunoprecipitation with anti-
bodies against the epigenetic marks of H3K4me3 and H3K27me3 during
direct reprogramming. Samples were obtained on day 0 (D0), day 4 (D4), day
8 (D8), and day 12 (D12). Adult NPCs (aNPC) were used as a control. Blue and
red bars indicate the extent of immuoprecipitation from normal IgG and
each of the antimodified histone antibodies, respectively. Data are per-
centage of input chromatin amount, analyzed by qPCR (mean ± SEM, n = 3).
Histone modification during direct reprogramming. Sox1 and Oct4
| www.pnas.org/cgi/doi/10.1073/pnas.1103113108 Kim et al.
results indicate that epigenetic commitment to the formation
of NPCs starts as early as day 8 in our direct reprogramming,
without a parallel commitment to pluripotency.
Fourth, we surmised that the fate choice between NPCs and
iPSCs is determined by exposure to signaling specific to each cell
type. To test this hypothesis, we examined the effects of using
LIF-containing medium (Fig. 4A). Strikingly, even brief exposure
(as little as 1 d) to LIF clearly decreased the number of PLZF-
expressing colonies on day 9 (Fig. 4B). Conversely, flow cyto-
metric analysis showed that LIF exposure increases the number
of SSEA1-expressing cells (Fig. S3 B and C). Gene expression
analysis also showed down-regulation of Sox1 and Pax6 with
concomitant up-regulation of Nanog and Rex1 under the same
conditions (Fig. 4 C and D). The correlated expression between
SSEA1 and Nanog as well as Rex1 implies an overall increase in
pluripotent character by extended LIF treatment. Accordingly,
we observed that small clusters of Nanog-GFP expressing cells
were generated when we used RepM-Pluri (essentially ESC
maintenance medium) throughout, instead of RepM-neural.
Interestingly, Sox17, T, and FGF5 expression on days 9, 11, and
13 was also increased by LIF exposure (Fig. S4B). This increase
in expression of genes specific to other lineages might result from
the spontaneous redifferentiation of Nanog-GFP–expressing
cells generated by LIF exposure or perhaps from an induction of
EpiSC-like cells. These results imply that intermediate cells may
embark on the path toward becoming iPSCs or NPCs by
responding to LIF or NPC-supporting medium, respectively.
This cytokine-dependent cell-type specification during direct
reprogramming has also been shown in other recent studies (6,
29). Thus, it is clear that the choice between conversion to NPCs
and complete reprogramming to iPSCs could be influenced at
a very early point during the transdifferentiation process.
Finally, we did not observe any expression of pluripotency
(Oct4), neural (Sox1 and Pax6), or neuronal (Tuj1) markers in
the starting cells or in the cells cultured in RepM-neural media
without dox induction (Fig. S5), strongly suggesting an absence of
contaminating neural cells in the starting MEF population.
Nonetheless, to rigorously rule out any contribution from con-
taminating neural crest cells and mesenchymal stem cells that
might exist in rather heterogeneous MEF populations (30), we
reprogrammed adult tail-tip fibroblasts (TTFs) using the dox-in-
ducible STEMCCA system (31) following the same method de-
scribed above. We successfully transdifferentiated TTFs into
(±0.006%, n = 2) using just 6 d of induction—shorter than the
8 d required for the generation of pluripotent cells from TTF with
this system (31). These reprogrammed NPCs from adult TTFs
showed the same characteristics as reprogrammed cells obtained
from secondary MEFs: both populations expressed neural rosette
markers (PLZF and ZO-1, Fig. 4G) and were able to differentiate
neurons (Fig. 4 H–L). These results confirm that exogenous genes
delivered lentivirally were able to successfully reprogram adult
TTFs. In parallel with four-factor reprogramming, we also tried
three factors (i.e., omitting c-Myc). However, otherwise identical
conditions failed to produce any neural cells in the latter case.
Because short-term LIF treatment seems to increase the plu-
ripotent character of the reprogrammed cells—although levels of
(Fig. 4D)—a small number of LIF-responsive cells might exist in
our cultures. To strictly eliminate the possibility of these cells, we
inhibited Janus kinase (JAK)/Stat3 signaling, which is directly in-
a small-molecule inhibitor of JAK. Although this inhibition de-
remaining colonies homogeneously expressed Pax6 (Fig. 4E).
These results imply that our transdifferentiation takes a strictly
be blocked by JAK inhibitor 1 treatment. When we analyzed the
properties of neurons differentiated from the NPCs derived using
JAK inhibitor treatment with the secondary MEF system, they
showed very similar results to cells not treated with JAK inhibitor
(Fig. S6). Thus, JAK inhibitor treatment does not adversely affect
direct transdifferentiation into functional NPCs.
In summary, we have successfully transdifferentiated fibro-
blasts into functional and proliferating NPCs. Such four iPSC-
factor–based transdifferentiation can be guided by modulating
transgene expression time and the culture environment (such
as specific cytokines) and will likely prove useful for trans-
differentiation to other lineages as well.
by different environmental cues. (A) Schematic of experiments involving
differential exposure to LIF- containing medium (RepM-Pluri) from day 4
onward for the indicated number of days. Dox treatment was for 5 d, be-
ginning on D0. (B) PLZF+colony number expressed as a percentage of the
total from each indicated sample on day 9. (C and D) Quantitative analysis of
mRNA levels of indicated marker genes in each sample (harvested on day 9).
All values are relative to expression in iPSCs. Pax6 immunostaining (E) and
total colony number (F) of TTFs transdifferentiated in the presence or ab-
sence of JAK inhibitor. Immunostaining of reprogrammed NPCs from TTFs;
neural rosette formation (G) and their long-term differentiation into various
neuronal subtypes (H–L). (Scale bars, 100 μm.)
Fate choice is dictated early during the transdifferentiation process
Kim et al. PNAS
| May 10, 2011
| vol. 108
| no. 19
We have shown a direct cell type switch from fibroblasts to
functional NPCs by transient expression of the four reprogram-
ming factors, whereby the process is clearly distinct from and
does not depend on the generation of iPSCs. This trans-
differentiation process is highly specific and efficient, reaching
completion within 9–13 d. Our method achieves an interlineage
cell type change much like that of iN cells (5), with one critical
advantage: the resulting cells are expandable progenitor cells.
Another advantage of our method is the use of general reprog-
ramming factors instead of lineage-specific transcription factors.
Because the four Yamanaka factors were chosen for pluripotent
cell generation (7), they are generally only considered useful for
the derivation of iPSCs. However, our results suggest a different
paradigm in which various developmentally plastic intermediate
cells may be generated in the process and that iPSCs are perhaps
only one of many possible outcomes of the four-factor reprog-
ramming. The ultimate result may depend largely on extrinsic
signaling inputs (Fig. 5). Interestingly, in the newt, Sox2, Klf4,
and c-Myc are temporarily up-regulated after lens removal or
limb amputation to initiate the regeneration process (33). This
observation lends support to our hypothesis that the four con-
ventional iPSC factors not only induce reprogramming to iPSCs,
but may also be capable of mediating direct fate switching be-
tween differentiated cells. Indeed, we have also found that
fibroblasts can be transdifferentiated into spontaneously con-
tracting cardiac cells by temporary expression of the same four
reprogramming factors under different culture conditions within
11 d (18). These results collectively imply a generally nonspecific/
undirected reprogramming process induced by the four iPSC
factors (i.e., not specifically directed toward the pluripotent state
as it has been regarded) and suggest a unique strategy/paradigm
to significantly expand and exploit iPSC-factor–based reprog-
ramming. Changing the duration of transgene expression and
culture conditions may allow a transient, plastic developmental
state established early to effectively serve as a cellular platform
for transdifferentiation toward various lineages.
Although iN cells are functional neurons (5), they lack the
potential to generate the diverse neuronal subtypes that can be
derived from iPSCs. Furthermore, these postmitotic neurons
may not be very suitable to the study of certain neurological
diseases, due not only to their nonproliferative state (which se-
verely limits their numbers), but also to their inability to re-
capitulate disease phenotypes occurring at the neural progenitor
stage (34). Currently, iPSCs derived from patients with late-onset
neurological diseases (e.g., those with Parkinson disease and
amyotrophic lateral sclerosis) do not readily recapitulate disease
phenotypes when redifferentiated (35, 36), although some
patients’ iPSCs with genetic defects show their symptoms (37,
38). These findings imply that complete reprogramming to
a pluripotent state may reset certain epigenetic hallmarks of the
disease state, thereby necessitating long-term aging under con-
ditions of stress for its repeated manifestation. Considering this
negative correlation between the degree of reprogramming and
the manifestation of a particular disease phenotype, we think
that the transdifferentiated NPCs—which are derived with lim-
ited reprogramming, thus avoiding the establishment of a plu-
ripotent state—could eventually prove to be a better-suited
model system than iPSCs when studying such late-onset diseases.
For secondary MEF cells, the transdifferentiation process may
at first glance seem inefficient compared with iPSC reprogram-
ming using the same cells (19). However, considering the ab-
breviated induction period, the efficiency is reasonable, i.e.,
comparable to initial reprogramming as confirmed by retro-
spective analyses (39).
We observed no significant up-regulation of EpiSC markers
such as Sox17, Brachyury (27), and FGF5 (26) during our direct
reprogramming. Nonetheless, a recent study by the Schöler
group showed direct reprogramming to EpiSCs can be achieved
in 3–5 wk (29), which is substantially slower than our method.
Thus, a temporary emergence of EpiSCs during our direct
reprogramming—which can be achieved within 12 d—is highly
unlikely. Interestingly, 5 d of induction followed by long-term LIF
treatment caused increased expression of Sox17 and FGF5 with
extremely low levels of Brachyury and Nanog (Fig. S4). These cells
may partially resemble EpiSCs; however, the absence of LIF in
our transdifferentiation media makes the generation of such cells
nearly impossible. As expression of exogenous factors is tightly
regulated in the inducible system (ref. 20 and Fig. S4A), to ach-
ieve fully reprogrammed EpiSCs or iPSCs status, a much longer
induction of reprogramming factors may be required.
In a recent development, lentiviral expression of OCT4 alone
was shown to mediate direct reprogramming of human fibroblasts
to blood progenitors (6). Although OCT4 expression is down-
regulated after long-term differentiation, long-term induction is
required and there is always a risk of lentiviral reactivation of
OCT4 expression, which may induce dysplastic lesions (40). Our
study shows that temporary expression of four reprogramming
factors is sufficient to induce lineage-specific transdifferentiation.
and temporary expression methods such as transient transfection
transdifferentiation across a broad lineage spectrum.
Materials and Methods
For direct reprogramming, secondary MEFs were sequentially cultured in MEF
medium, reprogramming initiation medium, and neural reprogramming
medium with 3–6 d of induction by doxycycline treatment beginning on day
0. Lentivirally transduced adult TTFs (using an inducible STEMMCCA system)
were reprogrammed as secondary MEFs. The detailed methods, cells, and
reagents concerning direct reprogramming with secondary MEFs and TTFs,
as well as all analysis such as quantitative PCR, immunocytochemistry, flow
cytometry, chromatin immunoprecipitation, and electrophysiology are de-
scribed in SI Materials and Methods.
ACKNOWLEDGMENTS. We thank Wenlin Li, Per Stehmeier, and Tongxiang
Lin for critical reading of the manuscript. We appreciate the assistance of
Ella Kothari and Jun Zhao at the University of California at San Diego, La
Potential of differentiation
fibf f f f f f f fi i i i i i i i i ib b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b
progenitor cells. By adding neural medium to four factor-induced in-
termediate cells comprising various epigenetic states, a fate switch to neural
stem/progenitor cells (NPCs) can be achieved. Alternatively, iPSCs can be
generated by prolonged expression of the four factors with concomitant
incubation in ESC medium. Cells belonging to other lineages could likely also
be isolated, depending on the type of medium used.
Model for direct reprogramming of fibroblasts (Fib.) to neural stem/
| www.pnas.org/cgi/doi/10.1073/pnas.1103113108 Kim et al.
Jolla, CA Stem Cell Core in blastocyst injections and Chao Zhao in caring for
mice. We thank Debbie Watry for flow cytometry and general assistance.
The pHAGE2-TetOminiCMV-STEMCCA plasmid was a generous gift from
Dr. Gustavo Mostoslavsky. S.D. and K.Z. are supported by the National
Institutes of Health (NIH) Director’s Transformative R01 Program (R01
EY021374). S.D. is supported by funding from the National Institute of Child
Health and Human Development, National Heart, Lung, and Blood Institute;
the National Institute of Mental Health/NIH; the California Institute for Re-
generative Medicine; the Prostate Cancer Foundation, Fate Therapeutics;
the Esther B. O’Keeffe Foundation; and The Scripps Research Institute. K.Z.
is supported by grants from the Chinese National 985 Project to Sichuan
University and West China Hospital, the National Eye Institute/NIH, Veterans
Administration Merit Award, the Macula Vision Research Foundation, Re-
search to Prevent Blindness, a Burroughs Wellcome Fund Clinical Scientist
Award in Translational Research, and the Dick and Carol Hertzberg Fund.
S.A.L. and M.T. were supported in part by NIH Grants P01 HD29587, P01
ES016738, P30 NS057096, and R01 EY05477 and by the California Institute
for Regenerative Medicine.
1. Takeuchi JK, Bruneau BG (2009) Directed transdifferentiation of mouse mesoderm to
heart tissue by defined factors. Nature 459:708–711.
2. Zhou Q, Brown J, Kanarek A, Rajagopal J, Melton DA (2008) In vivo reprogramming of
adult pancreatic exocrine cells to beta-cells. Nature 455:627–632.
3. Graf T, Enver T (2009) Forcing cells to change lineages. Nature 462:587–594.
4. Ieda M, et al. (2010) Direct reprogramming of fibroblasts into functional cardiomyocytes
by defined factors. Cell 142:375–386.
5. Vierbuchen T, et al. (2010) Direct conversion of fibroblasts to functional neurons by
defined factors. Nature 463:1035–1041.
6. Szabo E, et al. (2010) Direct conversion of human fibroblasts to multilineage blood
progenitors. Nature 468:521–526.
7. Takahashi K, Yamanaka S (2006) Induction of pluripotent stem cells from mouse
embryonic and adult fibroblast cultures by defined factors. Cell 126:663–676.
8. Takahashi K, et al. (2007) Induction of pluripotent stem cells from adult human
fibroblasts by defined factors. Cell 131:861–872.
9. Stadtfeld M, Maherali N, Breault DT, Hochedlinger K (2008) Defining molecular
cornerstones during fibroblast to iPS cell reprogramming in mouse. Cell Stem Cell 2:
10. Brambrink T, et al. (2008) Sequential expression of pluripotency markers during direct
reprogramming of mouse somatic cells. Cell Stem Cell 2:151–159.
11. Hanna J, et al. (2009) Direct cell reprogramming is a stochastic process amenable to
acceleration. Nature 462:595–601.
12. Artyomov MN, Meissner A, Chakraborty AK (2010) A model for genetic and
epigenetic regulatory networks identifies rare pathways for transcription factor
induced pluripotency. PLOS Comput Biol 6:e1000785.
13. Mikkelsen TS, et al. (2008) Dissecting direct reprogramming through integrative
genomic analysis. Nature 454:49–55.
14. Meissner A, Wernig M, Jaenisch R (2007) Direct reprogramming of genetically
unmodified fibroblasts into pluripotent stem cells. Nat Biotechnol 25:1177–1181.
15. Silva J, et al. (2008) Promotion of reprogramming to ground state pluripotency by
signal inhibition. PLoS Biol 6:e253.
16. Maherali N, et al. (2007) Directly reprogrammed fibroblasts show global epigenetic
remodeling and widespread tissue contribution. Cell Stem Cell 1:55–70.
17. Sridharan R, et al. (2009) Role of the murine reprogramming factors in the induction
of pluripotency. Cell 136:364–377.
18. Efe JA, et al. (2011) Conversion of mouse fibroblasts into cardiomyocytes using
a direct reprogramming strategy. Nat Cell Biol 13:215–222.
19. Wernig M, et al. (2008) A drug-inducible transgenic system for direct reprogramming
of multiple somatic cell types. Nat Biotechnol 26:916–924.
20. Hanna J, et al. (2008) Direct reprogramming of terminally differentiated mature B
lymphocytes to pluripotency. Cell 133:250–264.
21. Ying QL, Stavridis M, Griffiths D, Li M, Smith A (2003) Conversion of embryonic stem
cells into neuroectodermal precursors in adherent monoculture. Nat Biotechnol 21:
22. Hitoshi S, et al. (2004) Primitive neural stem cells from the mammalian epiblast
differentiate to definitive neural stem cells under the control of Notch signaling.
Genes Dev 18:1806–1811.
23. Elkabetz Y, et al. (2008) Human ES cell-derived neural rosettes reveal a functionally
distinct early neural stem cell stage. Genes Dev 22:152–165.
24. Walther C, Gruss P (1991) Pax-6, a murine paired box gene, is expressed in the
developing CNS. Development 113:1435–1449.
25. Alcock J, Sottile V (2009) Dynamic distribution and stem cell characteristics of Sox1-
expressing cells in the cerebellar cortex. Cell Res 19:1324–1333.
26. Greber B, et al. (2010) Conserved and divergent roles of FGF signaling in mouse
epiblast stem cells and human embryonic stem cells. Cell Stem Cell 6:215–226.
27. Tesar PJ, et al. (2007) New cell lines from mouse epiblast share defining features with
human embryonic stem cells. Nature 448:196–199.
28. Bernstein BE, et al. (2006) A bivalent chromatin structure marks key developmental
genes in embryonic stem cells. Cell 125:315–326.
29. Han DW, et al. (2011) Direct reprogramming of fibroblasts into epiblast stem cells. Nat
Cell Biol 13:66–71.
30. Weston JA, et al. (2004) Neural crest and the origin of ectomesenchyme: Neural fold
heterogeneity suggests an alternative hypothesis. Dev Dyn 229:118–130.
31. Sommer CA, et al. (2009) Induced pluripotent stem cell generation using a single
lentiviral stem cell cassette. Stem Cells 27:543–549.
32. Yang J, et al. (2010) Stat3 activation is limiting for reprogramming to ground state
pluripotency. Cell Stem Cell 7:319–328.
33. Maki N, et al. (2009) Expression of stem cell pluripotency factors during regeneration
in newts. Dev Dyn 238:1613–1616.
34. Marchetto MC, Winner B, Gage FH (2010) Pluripotent stem cells in neurodegenerative
and neurodevelopmental diseases. Hum Mol Genet 19:R71–R76.
35. Soldner F, et al. (2009) Parkinson’s disease patient-derived induced pluripotent stem
cells free of viral reprogramming factors. Cell 136:964–977.
36. Dimos JT, et al. (2008) Induced pluripotent stem cells generated from patients with
ALS can be differentiated into motor neurons. Science 321:1218–1221.
37. Lee G, et al. (2009) Modelling pathogenesis and treatment of familial dysautonomia
using patient-specific iPSCs. Nature 461:402–406.
38. Ebert AD, et al. (2009) Induced pluripotent stem cells from a spinal muscular atrophy
patient. Nature 457:277–280.
39. Smith ZD, Nachman I, Regev A, Meissner A (2010) Dynamic single-cell imaging of
direct reprogramming reveals an early specifying event. Nat Biotechnol 28:521–526.
40. Hochedlinger K, Yamada Y, Beard C, Jaenisch R (2005) Ectopic expression of Oct-4
blocks progenitor-cell differentiation and causes dysplasia in epithelial tissues. Cell
41. Okita K, Nakagawa M, Hyenjong H, Ichisaka T, Yamanaka S (2008) Generation of
mouse induced pluripotent stem cells without viral vectors. Science 322:949–953.
42. Jia F, et al. (2010) A nonviral minicircle vector for deriving human iPS cells. Nat
43. Zhou H, et al. (2009) Generation of induced pluripotent stem cells using recombinant
proteins. Cell Stem Cell 4:381–384.
44. Kim D, et al. (2009) Generation of human induced pluripotent stem cells by direct
delivery of reprogramming proteins. Cell Stem Cell 4:472–476.
45. Warren L, et al. (2010) Highly efficient reprogramming to pluripotency and directed
differentiation of human cells with synthetic modified mRNA. Cell Stem Cell 7:
46. Zhu S, et al. (2010) Reprogramming of human primary somatic cells by OCT4 and
chemical compounds. Cell Stem Cell 7:651–655.
Kim et al.PNAS
| May 10, 2011
| vol. 108
| no. 19