Upload
tranthuy
View
222
Download
0
Embed Size (px)
Citation preview
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
Nonstochastic Reprogrammingfrom a Privileged Somatic Cell StateShangqin Guo,1,2,* Xiaoyuan Zi,3,10 Vincent P. Schulz,4 Jijun Cheng,2,5 Mei Zhong,1,2 Sebastian H.J. Koochaki,1,2
Cynthia M. Megyola,1,2 Xinghua Pan,2,5 Kartoosh Heydari,6 Sherman M. Weissman,2,5,7 Patrick G. Gallagher,4,5,8
Diane S. Krause,1,2,9 Rong Fan,2,3,7 and Jun Lu2,5,71Department of Cell Biology2Yale Stem Cell Center3Department of Biomedical Engineering4Department of Pediatrics5Department of Genetics6Department of Immunobiology, Yale Flow Cytometry Core Facility7Yale Comprehensive Cancer Center8Department of Pathology9Department of Laboratory MedicineYale University, New Haven, CT 06520, USA10Department of Cell Biology, Second Military Medical University, Shanghai 200433, China
*Correspondence: [email protected]
http://dx.doi.org/10.1016/j.cell.2014.01.020
SUMMARY
Reprogramming somatic cells to induced pluripo-tency by Yamanaka factors is usually slow and ineffi-cient and is thought to be a stochastic process. Weidentified a privileged somatic cell state, from whichacquisition of pluripotency could occur in a nonsto-chastic manner. Subsets of murine hematopoieticprogenitors are privileged whose progeny cells pre-dominantly adopt the pluripotent fate with activationof endogenous Oct4 locus after four to five divisionsin reprogramming conditions. Privileged cells displayan ultrafast cell cycle of �8 hr. In fibroblasts, a sub-population cycling at a similar ultrafast speed isobserved after 6 days of factor expression and isincreased by p53 knockdown. This ultrafast cyclingpopulation accounts for >99% of the bulk reprog-ramming activity in wild-type or p53 knockdownfibroblasts. Our data demonstrate that the stochasticnature of reprogramming can be overcome in a priv-ileged somatic cell state and suggest that cell-cycleacceleration toward a critical threshold is an impor-tant bottleneck for reprogramming.
INTRODUCTION
Somatic cells can be reprogrammed into pluripotency by ex-
pression of defined transcription factors (Lowry et al., 2008;
Park et al., 2008; Takahashi et al., 2007; Takahashi and Yama-
naka, 2006; Wernig et al., 2007). Although most cell types can
be reprogrammed, this dramatic cell fate conversion occurs
only at low frequency following long latency, even when all cells
are engineered to express the reprogramming factors (Carey
et al., 2010; Stadtfeld and Hochedlinger, 2010; Wernig et al.,
2008). The prevailing theory for this low efficiency and long
latency is a stochastic model, which calls upon stochastic
changes to help subvert the various barriers limiting the fate tran-
sitions (reviewed in Hanna et al., 2010; Stadtfeld and Hochedlin-
ger, 2010; Yamanaka, 2009). Mathematic modeling suggests the
existence of a single major bottleneck event, although additional
non-rate-limiting events may also exist (Hanna et al., 2009, 2010;
Smith et al., 2010; Stadtfeld andHochedlinger, 2010; Yamanaka,
2009). However, the nature of such a bottleneck event has not
been clearly defined.
Although the reprogramming behavior of many cell types fol-
lows a stochastic model, it is possible that rare and/or transient
somatic cells may exist in a post-bottleneck state and can
progress toward reprogramming in a nonstochastic manner.
We term such putative post-bottleneck somatic cells the
privileged cells for reprogramming. Owing to the absence of
the rate-limiting stochastic events, these somatic cells should
display certain unique reprogramming behaviors (Figure 1): a
privileged somatic cell should produce progeny that mainly
progress toward pluripotency rather than adopt alternative
cell fates; their progeny should transition into pluripotency
rapidly in a largely synchronous fashion. Figure 1 depicts the
key differences between privileged and stochastic reprogram-
ming. Identification of a post-bottleneck cell state would help
to define the nature of the stochastic events restricting Yama-
naka reprogramming.
In this study, we provide evidence for the existence of privi-
leged somatic cells and describe a key feature of the privileged
cell state as an unusually fast cell cycle. The fast cycling cells
could exist naturally or could be induced from fibroblasts by
Yamanaka factors and are responsible for essentially all reprog-
ramming activities. Our data suggest a modified view for the role
of cell-cycle regulation in reprogramming and refine the conven-
tional stochastic versus elite models of reprogramming.
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 1
= Privileged somatic founder cell
= Pluripotent progeny= Progeny failed to reprogram
= Somatic founder cell
Latency(number of cell divisions)
% p
roge
ny
in p
lurip
oten
cy
0
50
100B Privileged
Stochastic
Stochasticreprogramming
Privilegedreprogramming
Few Most or all
Long Short
Asynchronous Largely synchronous
Reprogrammingbehavior
Progeny resulted in pluripotency
Latency
Kinetics
Hypothetic lineagescheme
A
Common RarePrevalence
Figure 1. Comparison between Stochastic and Privileged Reprogramming
(A) Hypothetic cell lineages with respect to the somatic founder cells and pluripotent progeny. The number of cell generations depicted is for illustration purpose
and does not represent the actual situations.
(B) Contrasting stochastic and privileged reprogramming with regard to their efficiency and latency.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
RESULTS
Nonstochastic Reprogramming from a Subpopulation ofBone Marrow GMP CellsTo identify the existence of privileged somatic cells, we first took
a live-cell imaging approach, with which the behaviors of single
cells can be faithfully tracked with high resolution (Megyola
et al., 2013). We focused on the well-defined granulocyte mono-
cyte progenitors (GMP) because they support rapid and efficient
reprogramming (Eminli et al., 2009; Megyola et al., 2013) and are
more likely to contain privileged cells. Specifically, GMPs from
mice that carry both Rosa26:rtTA and Oct4:GFP alleles were
used as source cells for reprogramming (FACS-sorting scheme
in Figure S1B available online) so that activation of endogenous
Oct4 locus can be detected as green fluorescence in live cells.
The Yamanaka factors were introduced by a doxycycline
(Dox)-inducible polycistronic lentivirus (Carey et al., 2009) so
that factor expression could be initiated by adding Dox, with
image acquisition starting about 1 hr later (the time required to
calibrate the imaging system). The reprogramming cultures
were imaged at 5–15min intervals for�5 days, when Oct4:GFP+
cells display typical features of mouse pluripotent cells. These
Oct4:GFP+ cells, though still Dox dependent, progress with re-
programming highly efficiently (Table S1) without any obvious
bottleneck restrictions, reaching a pluripotent state that can
support chimeric mice formation and germline transmission
(Megyola et al., 2013).
Using this imaging approach, we mapped the entire fate tran-
sition process from single founder GMPs to Oct4:GFP+ progeny
2 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
(Movie S1 and Figures S1A and 2A) and constructed 14 success-
fully reprogrammed cell lineages from 5 independent experi-
ments. Strikingly, these reprogrammed GMP lineages displayed
a behavior that is consistent with features of the privileged state
(Figure 1). All live progeny from theseGMPs turned onOct4:GFP,
yielding multiple sister colonies with homogeneous Oct4:GFP
fluorescence (Movie S1). No progeny retained hematopoietic
appearance or became ‘‘stuck’’ in intermediate steps, which is
prevalent in the stochastic reprogramming systems (Chen
et al., 2013; Mikkelsen et al., 2008). Because we added Dox
within 1 hr of imaging and because it would take time before fac-
tor expression amounted to significant levels, we could ascertain
with good confidence that the initial somatic cell state was
captured in the imaging process. Taken together, live-cell imag-
ing at single-cell resolution supports that some GMPs exist in a
privileged state (Figures 2A and S1A and Movie S1).
To further validate that a subset of GMPs is privileged, factor-
transduced GMPs were FACS sorted as single cells into 96-well
plates to achieve clonal reprogramming (experimental scheme in
Figure S5A). When 420 total wells were examined after 5 days of
Dox induction, 71 wells (17%) gave rise to Oct4:GFP+ colonies.
The majority of wells (97%) positive for Oct4:GFP+ colonies did
not contain noticeable numbers of cells bearing hematopoietic
morphology (Figure 2B), confirming our observation that most
progeny became reprogrammed using the imaging approach
(Movie S1). Many of these wells contained multiple sister
colonies, with small numbers of round-shaped cells (also
Oct4:GFP+) in close proximity to the larger colonies (Figure 2B),
which we have described previously to result from the dynamic
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
migrative behavior associated with pluripotency (Megyola et al.,
2013). To ensure that all GMP-derived cells were scored, we
repeated single-cell reprogramming with GMPs from a trans-
genic mouse line that expresses an H2B-GFP fusion protein
(Hadjantonakis and Papaioannou, 2004) so that all progeny
could be identified based on their expression of H2B-GFP (Fig-
ure S1C). In wells that contained GMP-derived colonies, all
H2B-GFP+ cells were part of or close to mESC-like colonies
and were positive for alkaline phosphatase activity (Figure 2C,
top). In contrast, single lineage-negative-Kit+Sca+ HSPCs
(referred to as LKS cells, sorting strategy in Figure S1B) gave
rise to many H2B-GFP+ cells that scattered away from the
colonies and were negative for alkaline phosphatase activity
(Figure 2C, bottom). These data from clonal reprogramming con-
ditions indicate that a subset of GMPs produce essentially all-
reprogrammed progeny.
To further ensure that all GMP-derived progeny were
analyzed, including some that might have adopted alternative
cell fate, we trypsinized the entire reprogramming culture initi-
ated from GMPs and assessed the complete cellular constitu-
ents by FACS on day 6 of factor expression. We used a pan
hematopoietic marker, CD45, to mark the somatic cell state
and Oct4:GFP for the pluripotent state. Consistent with the
observation that the progeny of 17% of GMPs all turned on
Oct4:GFP+, the entire culture was largely binary, consisting of
two major populations: the CD45+ hematopoietic cells and the
Oct4:GFP+ reprogrammed cells (Figures 2D and S1D). These
two populations accounted for �97% of all cells in the culture,
with the remaining 3% likely being the feeder cells, as indicated
by their noncharacteristic scattering profile (Figure S1D). This
analysis demonstrates that alternative cell fates are not prevalent
in GMP-initiated reprogramming.
Another anticipated feature of nonstochastic reprogramming
is a synchronous short latency, which could be reflected
as either the number of cell divisions or the amount of time pre-
ceding endogenous pluripotency activation (Figure 1B). We
assessed these parameters using the imaging data and found
that most of these progeny (85%, total n = 38) activated endog-
enous Oct4 after four to five cell divisions (Figure 2E). The re-
maining 15% of events took six or seven cell divisions, but not
more. Expressed in absolute amount of time, all progeny
became Oct4:GFP+ (detectible by fluorescence microscopy)
within 46.0 ± 6.8 hr (n = 38), indicating a largely uniform fast
kinetics. Importantly, this latency is highly consistent among
the 14 GMP lineages obtained across five experiments. These
results support the privileged nature of the founder GMPs and
demonstrate that Yamanaka reprogramming could occur much
more rapidly and synchronously than previously appreciated
(Hanna et al., 2009).
Taken together, the data above demonstrate that a subset of
GMPs exist as privileged cells for Yamanaka reprogramming.
Privileged GMPs Display an Ultrafast Cell CycleWe next asked whether the privileged cells display any unusual
characteristics. We examined the live-cell imaging data, quanti-
fied the timing of events in the GMP-initiated reprogramming lin-
eages, and noticed that the first cell cycle (defined as the time
between the first and second mitotic divisions from the initial
cells; diagram in Figure S1A and Figure 2A) on average took
only�8 hr (Figure 3A). In contrast, a typical mammalian cell cycle
lasts 16–24 hr or longer (Lodish et al., 2000). mESCs and iPSCs
display one of the fastest cell cycles, which lasts�12 hr, owning
to a truncated G1 phase (Hanna et al., 2010; Wang and Blelloch,
2009). Thus, it was striking that the first cell cycle during privi-
leged reprogramming was even substantially faster than that of
mESCs. The three subsequent cell cycles were also fast but
became increasingly longer and stabilized at the speed of
�12 hr per cycle (Figure 3A). Using a cell-cycle phase reporter
(Sakaue-Sawano et al., 2008), we observed a short G1 duration
(�2 hr) for the first cell cycle (Figures 3B and S2 and Movie S2).
These data suggest that G1 shortening underlies the fast cell
cycle. Importantly, because this ultrafast cycle occurred within
the first 10 hr of Dox addition and the subsequent cycles actually
lengthened when factors were continued to be induced (Fig-
ure 3A), the fast cycling property should be intrinsic to the subset
of GMPs and independent from Yamanaka factors. Indeed,
�20% of all GMPs examined (n = 54) display such an ultrafast
cell cycle. Thus, privilegedGMPs are associated with an ultrafast
initial cell cycle.
If privileged GMPs are distinguished by an ultrafast cell cycle,
one would predict that prospectively isolated faster cycling
GMPs reprogram with higher efficiency than do their slower
cycling counterparts. To test this, we fractionated GMPs based
on their cell-cycle speed, using a dye dilution approach (Koche
et al., 2011; Takizawa et al., 2011). Specifically, GMPs were
labeled with a stable dye CFSE, and the number of divisions
that each cell experienced was indicated by the remaining
amount of dye after proliferation-based dye dilution (Figure S3).
Thus, the cells that retained the least amount of dyes repre-
sented the most rapidly cycling cells (Figure 3C). With this
strategy, we sorted factor-transduced GMPs into CFSE-low
and CFSE-high single cells (representing relatively fast and
slow cycling GMPs, respectively) (Figure 3C) and compared their
reprogramming efficiency (Figure 3D). The fast cycling GMPs re-
programmed with higher efficiency than the slower cycling ones
(Figure 3D and Tables S2A and S2B). To rule out the possibility
that this difference was due to the cycling status affecting viral
transduction (even though lentivirus infect both cycling and
quiescent cells), we confirmed this finding using GMPs from
a mouse strain harboring a knockin cassette of the same
polycistron (referred to as the iPS mouse) (Carey et al., 2010)
(Table S2C). Taken together, the data above support that
privileged GMPs are associated with an ultrafast cell cycle,
and the subpopulation of faster cycling GMPs reprogram more
efficiently.
Accelerating Hematopoietic Progenitor Cell CyclingIncreases Reprogramming Efficiency and Induces theEmergence of Privileged CellsThe unusual cell-cycle speed that associated with privileged
GMPs raises the possibility that the ultrafast cell cycle repre-
sents a path to overcome the bottleneck restricting reprogram-
ming. We reasoned that, for a cell that cycles slowly and
reprograms inefficiently, increasing its cycling speed may
enhance its reprogramming and privileged cells could be poten-
tially induced. LKS cells cycle slowly and contain much fewer
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 3
Oct4:GFP ON
Time
A
From a single GMP on Day6
B Phase Oct4:GFP
Div.1 Div.2 Div.3 Div.4
2:10
9:05
9:45
21:30
19:00
16:50
17:05
31:25
31:20
28:50
28:50
27:20
26:20
27:30
1st cycle 2nd cycle 3rd cycle 4th cycle
G1 G1 G1
200 μm 200 μm
0
20
40
60
4 5 6 7
% o
f tot
al e
vent
s
# of divisions before Oct4:GFP ON
E
H2B-GFP AP
GMP
LKS
CC
D45D
Oct4:GFP
200 μm 200 μm
200 μm 200 μm
Figure 2. Nonstochastic Reprogramming from a Subset of GMPs
(A) A representative lineagemap from a single GMP toOct4:GFP+ progeny. The color of the circles corresponds to the color of arrows inMovie S1 and Figure S1A.
Lines denote lineage relationship. Filled green circles denote Oct4:GFP fluorescence as detectible by time-lapse imaging. The numbers (hr:min) at each
branching point indicate the time when mitosis occurred and were used to derive cell-cycle lengths. Red blocks on the horizontal block arrow indicate reporter
signals of a G1 phase reporter (see Figure S2 and Movie S2).
(B) GMPs were transduced with Dox-inducible Yamanaka factors and single cell sorted into 96-well plates in reprogramming conditions (scheme in Figure S5A).
Representative images of the reprogramming product of a single GMP are shown. Note the presence of multiple sister colonies and that many of the round-
shaped cells in the vicinity of Oct4:GFP+ colonies are also GFP+ (zoom in, red arrows).
(C) GMP and LKS cells from H2B-GFPmice were reprogrammed as single cells. Representative images of the reprogramming culture from a single GMP (top) or
LKS cell (bottom) after 6 days of Dox induction are shown. Note the presence of H2B-GFP+, alkaline phosphatase (AP)-negative cells in LKS-initiated culture, but
not in the GMP-initiated culture. A slight increase in colony sizes was noted after fixation/staining.
(legend continued on next page)
4 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
p = 4.4 x 10-8
RandomGMPsCell cycles of successfully
reprogrammed GMPs
Cel
lcyc
ledu
ratio
n(h
ours
)
0
5
10
15
20
25
30
1st 2nd 3rd 4th
A
0
5
10
15
1st 2nd 3rd
G1
dura
tion
(hou
rs)
B
**
* Unlabeled controlCFSE
Fast
C
Slow
CFSE
Post label
Post 24h dilution
0
2
4
6
Fast Slow
D
Rep
rogr
amm
ing
effic
ienc
y(%
) GMP single cell reprogramming
Cell cycles of successfullyreprogrammed GMPs
%of
Max
%of
Max
23/360
7/360
Figure 3. Privileged GMPs Display an Ultrafast Cell Cycle
(A) Cell cycle lengths of successfully reprogrammedGMP lineages. Each dot represents amitotic event that gave rise to Oct4:GFP+ cells. More dots were scored
for cell cycles 2 and 3 due to cell number increase following previous divisions. Within the same imaging experiments, the first cell cycle of randomGMPs (n = 54)
was also measured. Pooled data of 14 lineages from five independent imaging experiments are shown. *p < 0.001.
(B) A G1 phase reporter was used to measure G1 duration by time-lapse imaging. Each dot represents a mitotic event that gave rise to Oct4:GFP+ cells.
(C) Representative FACS plots of GMPs immediately after CFSE label (top) and following 24 hr of dye dilution (bottom). The gating strategy for sorting fast and
slow cycling cells is shown.
(D) The fast and slow cycling GMPs were single cell sorted into 96-well plates in reprogramming conditions. The number of wells that contained Oct4:GFP+
colonies and the total number of plated wells are indicated, separated by a slash.
See also Figures S2, S3, Table S2, and Movie S2.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
ultrafast cycling cells as compared to GMPs (Figures 4A, black
bars, and S4 and Table S3). Accordingly, LKS cells reprog-
rammed with much lower (7.6-fold) efficiency than did GMPs in
single-cell reprogramming assays (Figure 4B, black bars and
Table S4), consistent with a previous report (Eminli et al.,
2009). Importantly, LKS cells mostly reprogrammed in a sto-
chastic manner, i.e., single LKS cells gave rise to both reprog-
(D) GMP-initiated reprogramming culture were trypsinized after 6 days of Dox in
shows that CD45+ hematopoietic cells and Oct4:GFP+ cells make up �97% of
(E) The number of mitotic divisions before Oct4:GFP became detectible by imag
See also Figures S1, S2, S5A, Table S1 and Movies S1, S2.
rammed cells and substantial numbers of nonreprogrammed
cells (Figure 2C, bottom).
While the fast cycling of GMPs is required to sustain rapid tis-
sue turnover, stem cells remain largely quiescent to sustain life-
long tissue homeostasis and injury repair and only begin cycling
in response to proper cues (Greco and Guo, 2010). Thus, we
tested whether culturing LKS cells in growth factors/cytokines
duction and were stained with a CD45 antibody. A representative FACS plot
the culture.
ing (n = 38).
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 5
Rep
rogr
amm
ing
effic
ienc
y (%
)
Fresh cells CulturedB
LKS
GMP
4-8 8-11 9-15 >=12 Dead
A
% o
f pop
ulat
ion
% o
f pop
ulat
ion
Phas
e O
ct4:
GFP
C
Progeny of a single LKS cellDay 5 on Dox0
5
10
15
20
25
LKS GMP
0
20
40
60
0
20
40
60
HSPC cycle speed (hours/cycle)
77/420
2/42010/420
102/420
300 μm
300 μm
Figure 4. Accelerating HSPC Cycling Increases Reprogramming Efficiency and Induces the Emergence of Privileged Cells
(A) The cell-cycle speeds of LKS andGMP cells weremeasured, either fresh (black bars) or after 5 days of culture (red bars). The percentage of cells with indicated
cell-cycle speed is plotted (details in Extended Experimental Procedures and Figure S4). Note that fresh GMPs and cultured LKS cells contain ultrafast cycling
cells (red boxes), which were low/absent in fresh LKS cells and cultured GMPs.
(B) The reprogramming efficiency of fresh or cultured LKS andGMP cells were compared in single cell reprogramming assays. The number of wells that contained
Oct4:GFP+ colonies and the total number of plated wells are indicated, separated by a slash.
(C) Singly sorted LKS cells were cultured for 5 days and their somatic progeny transferred into reprogramming conditions for another 5 days (scheme shown in
Figure S5B). Shown is one representative well dominated by Oct4:GFP+ colonies and lacking hematopoietic-like cells.�15% of the wells containing Oct4:GFP+
colonies (or 3.6% of total wells) displayed this privileged reprogramming behavior.
Images were captured with 103 magnification. See also Figures S4 and S5 and Tables S3 and S4.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
increases the ultrafast cycling cells (Figure S4). Indeed, after
5 days of culture, an ultrafast cycling population (>16%)
emerged from the LKS culture to a level even higher than freshly
isolated GMPs (Figure 4A, red bars). In contrast, GMPs cultured
under the same condition resulted in a drastic loss of the ultrafast
6 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
cells (Figure 4B, red bars). We then tested the reprogramming
efficiency of these cultured cells in the stringent clonal reprog-
ramming assay (scheme in Figure S5B). As a comparison, the
transduced LKS or GMP cells from the same experiment were
single cell sorted and reprogrammed directly without preculture
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
(scheme in Figure S5A). Consistent with the increase in the
ultrafast cycling population, cultured LKS cells became highly
efficient (24.3%) in reprogramming, more so than the freshly
isolated GMPs (18.3%) (Figure 4B, red bars, and Table S4). In
addition, the drastic decrease of the ultrafast cycling population
in cultured GMPs was accompanied by a steep drop of reprog-
ramming efficiency to 0.5% (Figure 4B). These data indicate that
changes in cycling speed parallel the changes in reprogramming
efficiency.
The cultivation turned a single LKS cell into tens to hundreds of
cells (scheme in Figure S5B). Sampling a pool of the mixed prog-
eny from many LKS cells revealed ultrafast cycling behavior in a
fraction of them (Figure 4A). It is conceivable that, if the ultrafast
cycling behavior is shared by the daughters of a common LKS
cell, privileged reprogramming might occur simultaneously
from the many daughter cells. To test this, we transferred the
somatic progeny from single LKS cells into reprogramming con-
ditions (Figure S5B). After 5 days of Dox induction, we observed
the astounding appearance of privileged reprogramming (Fig-
ure 4C and Table S4). In �15% of the wells that contained
Oct4:GFP+ colonies (or 3.6% of total wells), hematopoietic-like
cells were absent. Instead, numerousOct4:GFP+ colonies domi-
nated these wells. Again, we confirmed that all progeny were
accounted for by using clonal LKS cells from the H2B-GFP-
transgenic mice (Figures S5C and S5D). As such, reprogram-
ming from these somatic cells was no longer a slow or rare event
but, instead, fast and prevalent. These wells provide further vali-
dation for the existence and unique behavior of privileged cells.
Thus, mitogenic activation of the slowly cycling LKS cells
increased the ultrafast cycling population as well as their reprog-
ramming efficiency. This activation also induced the emergence
of privileged cells for reprogramming.
Ultrafast Cycling Cells Emerge from Fibroblasts afterProlonged Factor Induction and Harbor the Majority ofReprogramming ActivityTo extend our finding of the ultrafast cycling cells in the hemato-
poietic system, we examined reprogramming from mouse em-
bryonic fibroblasts (MEFs). Because it is well established that
MEFs reprogram with low efficiency (�0.1%) and long latency
(>8 days), we did not anticipate that a substantial fraction of
MEFs naturally cycle at the ultrafast speed. Instead, we asked
whether an ultrafast cycling population can be induced by
Yamanaka factors. Because MEFs do not proliferate well at
clonal density, we estimated their cycling speed following a
similar dye dilution approach (Figure S3). Specifically, we trans-
duced MEFs with the Dox-inducible factors and added Dox to
induce their expression. On day 4 of Dox treatment, the cultures
were trypsinized, labeled with CFSE, and plated back in the
presence of Dox to allow the dye to be diluted for 2 additional
days (Figure S6A). This time frame was chosen based on the
observation that >7 days of Dox induction led to the appearance
of a small percentage of Oct4:GFP+ cells (although Oct4:GFP+
colonies were not visible at this time), whereas 6 days of Dox
treatment yielded no such cells (Figure S6B). Consistent with
MEFs being a more heterogeneous population, the CFSE peak
immediately following dye labeling was wider than that of
hematopoietic progenitors (compare Figures 5A and 3C). Strik-
ingly, a small population of cells (1%–8%) that retained much
less CFSE emerged after a total of 6 days of Dox treatment (Fig-
ure 5A, red box). In contrast, such a population was not observed
in the absence of Dox from the same transduced MEFs (Fig-
ure S6C), indicating that the CFSE-low population was induced
by factor expression. The mean CFSE fluorescence intensity
for this minor population was at least 20-fold less than that of
the bulk cells, suggesting that the fast cells had undergone, on
average, four more divisions than the bulk MEFs. Given that
the bulk MEFs divide every 25–30 hr (Sage et al., 2000; Wang
et al., 2001; White and Dalton, 2005) and would divide, on
average, once to twice during 48 hr, the CFSE-low cells must
have undergone approximately five to six divisions during the
48 hr of dye dilution period. This estimation suggests that
some of the CFSE-low cells had divided at a speed close to
8 hr per cycle. Thus, an ultrafast cycling population emerges
from MEFs in the presence of prolonged factor expression.
We then asked whether the ultrafast cycling cells are more
efficient in reprogramming. Cells of three different CFSE levels
(reflecting different cycling speed)were FACSsorted and their re-
programming efficiencies compared. Due to their poor ability to
endure clonal culture and overall low reprogramming efficiency,
we were unable to perform clonal reprogramming assays with
MEFs. Instead, we used the generally adopted definition for re-
programming efficiency, which is the number of Oct4:GFP+ col-
onies divided by the number of cells plated. The fast cycling cells
displayed �1,150 -fold increase in reprogramming efficiency as
compared to the cells cycling at a medium speed characteristic
of the bulk culture (Figures 5B and 5C). The slow populations
completely lacked reprogramming activity (Figures 5B and 5C).
Overall, we estimated that the minor but fast population con-
tained�99.7%of the reprogrammingactivity of theentire culture.
Taken together, an ultrafast cycling population was induced in
MEF-initiated reprogramming after 6 days of factor expression.
The reprogramming activity was exclusively confined to the
rapidly cycling cells.
p53 Knockdown Expands the Ultrafast CyclingPopulation, which Harbors the Majority ofReprogramming Activity among p53 Knockdown MEFsAlthough genetic perturbations that result in hastened prolifera-
tion have been well documented to increase reprogramming
efficiency (Banito et al., 2009; Hong et al., 2009; Ruiz et al.,
2011; Utikal et al., 2009), it remains unclear whether the role of
cell-cycle acceleration is to increase the number of daughter
cells, each of which possesses an independent probability to
reprogram, or to confer a higher probability selectively to a sub-
set of cells to accomplish the fate transition.
To distinguish these two possibilities, we examined reprog-
ramming in the context of p53 knockdown, which enhances re-
programming via a cell-division-dependent mechanism (Hanna
et al., 2009). Indeed, higher reprogramming efficiency occurred
with p53 knockdown (Figures 6A and 6B). We then performed
CFSE dilution assay to measure the ultrafast cycling fractions
(Figure 6C). Consistent with a role of p53 in regulating cell cycle,
p53 knockdown led to overall decreased CFSE level (i.e.,
increased cycling) among the bulk population, as evidenced by
a shift of the main CFSE peak toward left (Figure 6C). As in
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 7
Nega ve Post label Post dilu on
>4 more div.
Fast Med SlowA
CFSE
B
Rep
rogr
amm
ing
effic
ienc
y (%
)
0
0.2
0.4
0.6
0.8
Fast Med Slow
1150x
% o
f Max
C
Phase
Oct4:GFP
AP
Fast Med Slow
300 μm
300 μm
300 μm
300 μm
300 μm
300 μm
(legend on next page)
8 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
A
β-actin
p53
shCtrl shp53B
CFSE
Post
labe
lPo
st48
hdi
lutio
n
shControl
shp53
C
0123456
shControl shp53
Fast
cells
(%)
D
0
0.5
1
1.5
2
ControlFast
ControlSlow
shp53Fast
shp53SlowR
epro
gram
min
gef
ficie
ncy
(%)
E
0
0.01
0.02
0.03
0.04
shControl shp53
Rep
rogr
amm
ing
effic
ienc
y(%
)
shControl shp53
Unlabeled MEFs
Fast
Fast
Slow
Slow
Figure 6. The Ultrafast Cycling Population from MEFs Is Increased by p53 Knockdown and Accounts for Most Reprogramming Activity
MEFs were transduced with either control (shCtrl) or shRNAs targeting p53 (shp53), along with the Dox-inducible reprogramming factors.
(A) p53-knockdown increased reprogramming in unfractionated MEFs.
(B) Western blot analysis confirmed p53 protein downregulation by p53-targeting shRNAs.
(C and D) Factor-transduced MEFs on 4 days of Dox treatment were labeled with CFSE (top) and were allowed to dilute the dye for another 48 hr. The cultures
were then trypsinized andCFSE intensity analyzed by FACS. (C) Representative FACSplots are shown for CSFE levels right after labeling or following 48 hr of dye-
dilution. Gating for fast and slow cells is shown. (D) Quantification of fast cycling cells (n = 3 per condition).
(E) Reprogramming efficiency of the sorted fast and slow cycling MEFs as shown in (C) (n = 3 per condition).
Error bars indicate SD.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
control knockdown, a distinct fast cycling population retaining
much less CFSE was clearly identifiable in p53 knockdown cells.
Importantly, this fast cycling population increased with p53
knockdown as compared to the control (Figures 6C and 6D).
These data indicate that p53 knockdown increases the number
of cells cycling past a certain threshold speed.
We then asked whether reprogramming activity is confined
within the fast cycling population or nondiscriminatively in p53
knockdown cultures. With p53 knockdown, the great majority
of the reprogramming activity was again confined to the fast
cells, similar to the situation in controls (Figure 6E). These data
Figure 5. Ultrafast Cycling Cells Emerge from Fibroblasts after Yaman
Activity
Factor-transduced MEFs were treated with Dox for 4 days, labeled with CFSE, a
(A) The levels of CSFE in nonlabeled MEFs (negative) right after labeling (postla
medium (med), and slow cycling cells are shown. Note that the fluorescence inten
division (div.) differences.
(B) Reprogramming efficiency of cells sorted on CFSE. Oct4:GFP+ colonies we
indicate SD).
(C) Representative images of the reprogramming cultures from cells of different cy
and alkaline phosphatase (AP)-stained dishes were from whole 60 mm plates.
See also Figure S6.
demonstrate that p53 impacts reprogramming by controlling
the emergence of the ultrafast cycling cells. Outside of the fast
cycling population, however, other cells do not become more
likely to reprogram. Taken together, our data suggest that cell-
cycle acceleration toward a critical threshold could be an impor-
tant bottleneck for reprogramming.
Molecular Characterization Confirms Enhanced CellCycle as the Predominant Feature of a Unique Cell StateTo gain insights into the molecular nature of the unique cell state
associated with the extraordinary efficiency to transition into
aka Factor Expression and Harbor the Majority of Reprogramming
nd allowed for dye-dilution for 48 hr in the presence of Dox.
bel) or following 48 hr of dilution (postdilution) are shown. The gates for fast,
sity difference between fast and medium populations indicates more than four
re scored on day 20 from initial Dox induction (n = 3 per condition; error bars
cling speed. Phase and Oct4:GFP images were captured at 103magnification,
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 9
0
0.0001
0.0002
0.0003
0.0004
SlowMEF1
FastMEF1
SlowMEF2
FastMEF2
iPS1 iPS2
Endogenous Sox2
Nor
mal
ized
exp
ress
ion
(2Δ
CT )
BA First 2 PCA components
GO Category Name Related Cellular ProcessDNA_REPLICATION Cell CycleCHROMOSOME Cell CycleDNA_DEPENDENT_DNA_REPLICATION Cell CycleM_PHASE Cell CycleCHROMOSOMAL_PART Cell CycleMITOSIS Cell CycleM_PHASE_OF_MITOTIC_CELL_CYCLE Cell CycleREPLICATION_FORK Cell CycleDNA_PACKAGING Cell CycleCELL_CYCLE_PROCESS Cell CycleCELL_CYCLE_PHASE Cell CycleDNA_METABOLIC_PROCESS Cell CycleCONDENSED_CHROMOSOME Cell CycleSPINDLE Cell CycleSPLICEOSOME RNA ProcessingDNA_REPAIR Cell CycleCHROMOSOMEPERICENTRIC_REGION Cell CycleRNA_PROCESSING RNA ProcessingRIBONUCLEOPROTEIN_COMPLEX RNA ProcessingSPINDLE_POLE Cell CycleSMALL_NUCLEAR_RIBONUCLEOPROTEIN_COMPLEX RNA ProcessingMITOTIC_CELL_CYCLE Cell CycleSTRUCTURAL_CONSTITUENT_OF_RIBOSOME Protein TranslationCHROMOSOME_SEGREGATION Cell CycleNUCLEAR_PART Cell CycleCHROMATIN_BINDING RNA Transcription
C
D F
β-actin
p57
shCtrl shp57
00.20.40.60.8
11.2
E
Rep
rogr
amm
ing
effic
ienc
y (%
)
0
0.1
0.2
0.3
0.4
shCtrl shp570
0.5
1
1.5
shCtrl shp57
*LKS GMP
p57 RNA
Nor
mal
ized
exp
ress
ion
(legend on next page)
10 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
pluripotency, we performed transcriptome analysis by RNA
sequencing (RNA-seq) (Pan et al., 2013). The transcriptomes of
fast GMPs and the fast cells arising from MEFs after 6 days of
Dox induction (referred to as ‘‘fast MEFs’’ from now on) were
compared to slow GMPs, slow MEFs, bulk MEFs, and estab-
lished iPSCs. Principle component analysis (PCA) showed that
biological replicates of the same cell states cluster in close prox-
imity (Figure 7A). Furthermore and as expected, fast MEFs are
enriched for genes associated with MEFs permissive to reprog-
ramming, whereas the slow MEFs display signatures associated
with MEFs refractory to reprogramming (Polo et al., 2012) (Fig-
ure S7A). These data indicate that our RNA-seq data are of
good quality.
Activation of endogenous Sox2was known to mark the begin-
ning of the nonstochastic phase of reprogramming (Buganim
et al., 2012). However, neither GMPs nor fast MEFs expressed
endogenous Sox2, contrasting iPSCs (Figures 7B and S7B).
Thus, the fast cycling cells exist in a cell state earlier than that
marked by endogenous Sox2 activation, indicating that the non-
stochastic trajectory of reprogramming can be attained earlier
than that demonstrated previously (Buganim et al., 2012).
The high reprogramming efficiency of both fast GMPs and fast
MEFs implies that they may share certain molecular similarities
that contribute to their unique reprogramming behavior even
though they are well separated on the PCA plot (Figure 7A). To
identify such similarities, we performed gene set enrichment
analysis (GSEA) comparing the fast populations with the bulk
MEFs and the slowMEFs, which were similar in gene expression
(Figure 7A). Strikingly, when all gene ontology (GO) categories
were examined, those associated with enhanced cell prolifera-
tion appear to be the predominant feature (Figures 7C and
S7C). Other enriched GO categories reflect enhanced RNA tran-
scription/processing and protein translation, which are consis-
tent with molecular needs in association with fast cell cycle. Of
note, enhanced cell cycle and protein translation were also en-
riched in fast GMPs as compared to slow GMPs, albeit to a
lesser extent (data not shown), due to their overall similarity in
gene expression (Figure 7A). In addition to unbiased GSEA anal-
ysis, we also examined enrichment in mesenchymal-epithelial
transition (MET), apoptosis, and metabolism. Results show that
fast GMPs and fast MEFs are depleted for mesenchymal signa-
tures (Table S5). There is no consistent difference in apoptosis
gene sets, whereas metabolism gene sets enriched in fast cells
are predominantly related to nucleoside metabolism, again sup-
porting a difference primarily in cell cycle (Table S5). These
molecular analyses suggest that fast cell cycle is indeed the
major attribute shared between the two cell states capable of
Figure 7. Molecular Characterization of the Cellular States
(A) Principle component analysis (PCA) was performed on RNA sequencing da
independent reprogrammed cells (iPS) and bulk MEFs (MEF) were included as c
(B) The endogenous Sox2 level was measured by qRT-PCR using primers speci
(C) Gene set enrichment analysis (GSEA) was performed comparing the fast cells
gene ontology (GO) categories enriched in both fast cell populations are shown.
(D) qRT-PCR analysis of p57 mRNA level in HSPCs.
(E) Western blot analysis confirms p57 protein downregulation by p57 targeting
(F) Reprogramming efficiency of LKS and GMP cells following treatment with co
Error bars indicate SD. See also Figure S7.
high-efficiency reprogramming. Loss of mesenchymal gene
expression appears as an additional contributor, which is impor-
tant in early reprogramming (Li et al., 2010; Samavarchi-Tehrani
et al., 2010).
To further understand the molecular regulator(s) of the privi-
leged cell state, we thus focused on cell-cycle regulators and
asked which cell-cycle genes are differentially expressed be-
tween LKS and GMP cells, which display drastically different
cell-cycle and reprogramming behaviors (Figure 4). We exam-
ined all expressed cyclins, cyclin-dependent kinases (CDKs),
and CDK inhibitors using published data (Krivtsov et al., 2006).
The only gene that showed a major difference was Cdkn1c, or
p57 (Figures S7D–S7F). p57 is a G1/S-blocking CDK inhibitor
that was expressed >10-fold higher in the slow cycling LKS
cells as compared to GMPs, a result confirmed by qRT-PCR
(Figure 7D). We next tested whether reduced p57 promotes
LKS cell reprogramming. We inhibited p57 expression by
sequence-specific shRNAs and confirmed downregulation of
the protein (Figure 7E). p57 shRNAs in LKS cells led to signifi-
cantly higher reprogramming (Figure 7F, LKS), consistent with
previous reports that loss of p57 leads to enhanced HSC cycling
(Matsumoto et al., 2011; Zou et al., 2011). In contrast, similar
treatment in GMPs did not yield further increase in reprogram-
ming (Figure 7F, GMP), suggesting that the endogenous p57
level in GMPs is already sufficiently low. These data support
p57 as one of the molecular roadblocks limiting HSPCs to enter
a privileged cell state.
DISCUSSION
We demonstrated that Yamanaka reprogramming could occur
nonstochastically from a privileged somatic cell state. Our data
support a model in which privileged and stochastic reprogram-
ming coexist as part of a continuum (Figure S7G). Because the
privileged cell state is highly efficient in reprogramming, it might
suggest an ‘‘elite’’ status. However, the privileged state differs
from the conventional ‘‘elite’’ cells in several key aspects. First,
the conventional ‘‘elite’’ model assigns different probability of re-
programming to different cells and assumes it as a fixed property
(Hanna et al., 2009, 2010; Yamanaka, 2009). Ourmodel suggests
that the privileged state is rather dynamic, which could exist
naturally but could also be gained via alternative means. Yama-
naka factors may facilitate the emergence of a similar state that
could be viewed as an ‘‘acquired privilege.’’ Our data also sug-
gest that the exact molecular means to reach a privileged state
might be different in different cell types. Specifically, p57 pre-
vents LKS cells to become privileged, but it is conceivable that
ta. Isolation of fast and slow cells was performed as described above. Dox-
ontrols. Three to five replicates were used for each cell type.
fic for its 30UTR region. Two biological replicates for each cell type are shown.
(FastGMP and FastMEF) versus the slow cells (MEF and SlowMEF). The top 25
shRNAs.
ntrol (shCtrl) or p57-targeting shRNAs (shp57).
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 11
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
other cell types may utilize alternative means to reach the
extreme speed of cycling. Although inhibition of p57 in HSCs
enhances their cycling (Matsumoto et al., 2011; Zou et al.,
2011), MEFs deficient in p57 do not become overtly proliferative
(Takahashi et al., 2000). Consistently, shRNA inhibition of p57 in
MEFs resulted in no detectible increase in ultrafast cycling cells
or iPS colony numbers (data not shown). Second, the conven-
tional ‘‘elite’’ cells are considered for their ability to produce
any pluripotent progeny without a specific latency. In contrast,
our privileged cells produce essentially all pluripotent progeny
with a synchronous short latency. Depending onwhen andwhich
cells reach a privileged state (somatic or acquired) within a cell
lineage, reprogramming could occur with varying efficiency
and latency (Figure S7G). Privileged somatic cells represent an
extreme case when the bottleneck is overcome prior to their
exposure to Yamanaka factors.
One of the key features of the privileged cell state is an unusu-
ally fast cell cycle, a mechanism that is distinct from previously
suggested ones, as GMPs are nonepithelial (Aasen et al.,
2008; Li et al., 2010; Samavarchi-Tehrani et al., 2010), do not
endogenously express any of the Yamanaka factors (Kim et al.,
2008; Krivtsov et al., 2006; Wakao et al., 2011), are more differ-
entiated than stem cells (Eminli et al., 2009; Stadtfeld and
Hochedlinger, 2010), and do not have reduced MBD3 levels
(Luo et al., 2013; Rais et al., 2013) (Figure S7H–S7J). The cellular
states of naturally fast cycling somatic cells, such as GMPs, and
those that arise fromMEFs in response to Yamanaka factors are
drastically different (Figure 7A). However, the fact that both types
of fast cycling cells display extraordinarily high reprogramming
efficiency suggests that a fast cell-cycle kinetic has a dominant
role in determining how the epigenome responds to remodeling.
A connection between cell-cycle and fate outcome has been
noted in other cellular systems (Kueh et al., 2013; Pauklin and
Vallier, 2013; Pop et al., 2010; Tsunekawa et al., 2012). Detailed
studies are required to reveal the connections linking cell-cycle
dynamics to epigenetic remodeling.
Although ultrafast cell cycle is a key feature of the privileged
state and ultrafast cycling cells reprogrammuchmore efficiently,
we also notice that not all fast cycling cells become reprog-
rammed. A number of reasons, both biological and technical,
could limit the observation of reprogramming even though the
major bottleneck may have been overcome. Biologically, this
may relate to proper establishment of E-cadherin-mediated
cell-cell adhesion (Chen et al., 2010; Li et al., 2010; Megyola
et al., 2013; Redmer et al., 2011; Samavarchi-Tehrani et al.,
2010) and optimal level or stoichiometry of continued factor
expression (Carey et al., 2011; Papapetrou et al., 2009). Such
events could impact a large number of cells that collectively
reduce the overall reprogramming efficiency. Technically,
although the CSFE dilution method greatly enriches for ultrafast
cycling cells based on divisional history, it is not an absolute or
instant measure of cell-cycle speed. The resolving power of
this method is heavily influenced by the inherent populational
heterogeneity and the duration of dye dilution. This explains
why much larger difference was seen between populations
isolated from MEFs as compared to that between different
GMPs (GMPs are more homogeneous and were allowed a
shorter period for dye dilution). Defining the additional require-
12 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
ments for the ultrafast cycling cells to secure the transition into
pluripotency will help to further elucidating the mechanisms of
reprogramming.
EXPERIMENTAL PROCEDURES
Mice and Cells for Reprogramming
All mouse work was approved by the Institutional Animal Care and Use
Committee of Yale University. The Oct4:GFP 3 Rosa26:rtTA mice were
derived by crossing Oct4:GFP mice with Rosa26:rtTA mice (Eminli et al.,
2009; Megyola et al., 2013) and were used to isolate/derive hematopoietic
cells, MEFs, and ESCs. The iPS mouse (R26rtTA;Col1a14F2A) (Carey et al.,
2010) (stock# 011004) and H2B-GFP transgenic line (stock# 006069) were
purchased from the Jackson Laboratory.
Cell Sorting, Culture, and Reprogramming
Hematopoietic populations were sorted as previously (Guo et al., 2010).
Hematopoietic growth factors included 100 ng/ml mSCF, 50 ng/ml mIL3,
50 ng/ml Flt3L, and 50 ng/ml mTPO (PeproTech) and were added to a
serum-free base medium X VIVO15 (BioWhittaker). Reprogramming condi-
tions included inactivated MEF feeders, complete mESC medium, and
2 mg/ml of Dox (Sigma). CFSE was purchased from Life Technologies. A work-
ing concentration of 7.5–8 mM was used following manufacturer’s instruction.
CFSE-labeled GMPs after 24 hr of dye dilution were sorted for the brightest
and dullest 15%–20% of cells as slow and fast GMPs, respectively. CFSE-
labeled MEFs were cultured in mESC medium with Dox for 48 hr before sort-
ing. MEFs cultured in the absence of Dox were used as controls.
Transcriptome Analysis by RNA Sequencing
1,000 of each cell types were directly sorted into lysis buffer in 96-well
plates for RNA isolation. The mRNA selection and reverse transcription were
performed as described previously with some modifications (Pan et al.,
2013). Additional details are described in the Extended Experimental
Procedures.
ACCESSION NUMBERS
The RNA-seq data described in Figures 7 and S7 have been deposited at the
GEO database with accession number GSE53074.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Extended Experimental Procedures, seven
figures, five tables, and two movies and can be found with this article online at
http://dx.doi.org/10.1016/j.cell.2014.01.020.
ACKNOWLEDGMENTS
This work was supported by NIH grants K01 DK082982 (to S.G.),
R01CA149109 (to J.L.), P30 DK0724429 and R01 DK086267 (to D.S.K.),
R01HL106184 (to P.G.G.), U54 CA143798 (to R.F.), P01GM099130 (to
S.M.W.), the State of Connecticut (to J.L., R.F., and S.M.W.), and the Packard
Fellowship for Science and Engineering (to R.F.).
Received: June 7, 2013
Revised: November 4, 2013
Accepted: January 10, 2014
Published: January 30, 2014
REFERENCES
Aasen, T., Raya, A., Barrero, M.J., Garreta, E., Consiglio, A., Gonzalez, F., Vas-
sena, R., Bili�c, J., Pekarik, V., Tiscornia, G., et al. (2008). Efficient and rapid
generation of induced pluripotent stem cells from human keratinocytes. Nat.
Biotechnol. 26, 1276–1284.
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
Banito, A., Rashid, S.T., Acosta, J.C., Li, S., Pereira, C.F., Geti, I., Pinho, S.,
Silva, J.C., Azuara, V., Walsh, M., et al. (2009). Senescence impairs successful
reprogramming to pluripotent stem cells. Genes Dev. 23, 2134–2139.
Buganim, Y., Faddah, D.A., Cheng, A.W., Itskovich, E., Markoulaki, S., Ganz,
K., Klemm, S.L., van Oudenaarden, A., and Jaenisch, R. (2012). Single-cell
expression analyses during cellular reprogramming reveal an early stochastic
and a late hierarchic phase. Cell 150, 1209–1222.
Carey, B.W., Markoulaki, S., Hanna, J., Saha, K., Gao, Q., Mitalipova, M., and
Jaenisch, R. (2009). Reprogramming of murine and human somatic cells using
a single polycistronic vector. Proc. Natl. Acad. Sci. USA 106, 157–162.
Carey, B.W., Markoulaki, S., Beard, C., Hanna, J., and Jaenisch, R. (2010).
Single-gene transgenic mouse strains for reprogramming adult somatic cells.
Nat. Methods 7, 56–59.
Carey, B.W., Markoulaki, S., Hanna, J.H., Faddah, D.A., Buganim, Y., Kim, J.,
Ganz, K., Steine, E.J., Cassady, J.P., Creyghton, M.P., et al. (2011). Reprog-
ramming factor stoichiometry influences the epigenetic state and biological
properties of induced pluripotent stem cells. Cell Stem Cell 9, 588–598.
Chen, T., Yuan, D., Wei, B., Jiang, J., Kang, J., Ling, K., Gu, Y., Li, J., Xiao, L.,
and Pei, G. (2010). E-cadherin-mediated cell-cell contact is critical for induced
pluripotent stem cell generation. Stem Cells 28, 1315–1325.
Chen, J., Liu, H., Liu, J., Qi, J., Wei, B., Yang, J., Liang, H., Chen, Y., Chen, J.,
Wu, Y., et al. (2013). H3K9 methylation is a barrier during somatic cell reprog-
ramming into iPSCs. Nat. Genet. 45, 34–42.
Eminli, S., Foudi, A., Stadtfeld, M., Maherali, N., Ahfeldt, T., Mostoslavsky, G.,
Hock, H., and Hochedlinger, K. (2009). Differentiation stage determines poten-
tial of hematopoietic cells for reprogramming into induced pluripotent stem
cells. Nat. Genet. 41, 968–976.
Greco, V., and Guo, S. (2010). Compartmentalized organization: a common
and required feature of stem cell niches? Development 137, 1586–1594.
Guo, S., Lu, J., Schlanger, R., Zhang, H., Wang, J.Y., Fox, M.C., Purton, L.E.,
Fleming, H.H., Cobb, B., Merkenschlager, M., et al. (2010). MicroRNA miR-
125a controls hematopoietic stem cell number. Proc. Natl. Acad. Sci. USA
107, 14229–14234.
Hadjantonakis, A.K., and Papaioannou, V.E. (2004). Dynamic in vivo imaging
and cell tracking using a histone fluorescent protein fusion in mice. BMC
Biotechnol. 4, 33.
Hanna, J., Saha, K., Pando, B., van Zon, J., Lengner, C.J., Creyghton, M.P.,
van Oudenaarden, A., and Jaenisch, R. (2009). Direct cell reprogramming is
a stochastic process amenable to acceleration. Nature 462, 595–601.
Hanna, J.H., Saha, K., and Jaenisch, R. (2010). Pluripotency and cellular re-
programming: facts, hypotheses, unresolved issues. Cell 143, 508–525.
Hong, H., Takahashi, K., Ichisaka, T., Aoi, T., Kanagawa, O., Nakagawa, M.,
Okita, K., and Yamanaka, S. (2009). Suppression of induced pluripotent
stem cell generation by the p53-p21 pathway. Nature 460, 1132–1135.
Kim, J.B., Zaehres, H., Wu, G., Gentile, L., Ko, K., Sebastiano, V., Arauzo-
Bravo, M.J., Ruau, D., Han, D.W., Zenke, M., and Scholer, H.R. (2008). Plurip-
otent stem cells induced from adult neural stem cells by reprogramming with
two factors. Nature 454, 646–650.
Koche, R.P., Smith, Z.D., Adli, M., Gu, H., Ku, M., Gnirke, A., Bernstein, B.E.,
and Meissner, A. (2011). Reprogramming factor expression initiates wide-
spread targeted chromatin remodeling. Cell Stem Cell 8, 96–105.
Krivtsov, A.V., Twomey, D., Feng, Z., Stubbs, M.C., Wang, Y., Faber, J., Lev-
ine, J.E., Wang, J., Hahn, W.C., Gilliland, D.G., et al. (2006). Transformation
from committed progenitor to leukaemia stem cell initiated by MLL-AF9.
Nature 442, 818–822.
Kueh, H.Y., Champhekar, A., Nutt, S.L., Elowitz, M.B., and Rothenberg, E.V.
(2013). Positive feedback between PU.1 and the cell cycle controls myeloid
differentiation. Science 341, 670–673.
Li, R., Liang, J., Ni, S., Zhou, T., Qing, X., Li, H., He,W., Chen, J., Li, F., Zhuang,
Q., et al. (2010). Amesenchymal-to-epithelial transition initiates and is required
for the nuclear reprogramming of mouse fibroblasts. Cell Stem Cell 7, 51–63.
Lodish, H., Berk, A., and Zipursky, S.L. (2000). Molecular Cell Biology. Cell-
Cycle Control in Mammalian Cells (New York: W. H. Freeman).
Lowry, W.E., Richter, L., Yachechko, R., Pyle, A.D., Tchieu, J., Sridharan, R.,
Clark, A.T., and Plath, K. (2008). Generation of human induced pluripotent
stem cells from dermal fibroblasts. Proc. Natl. Acad. Sci. USA 105, 2883–2888.
Luo, M., Ling, T., Xie, W., Sun, H., Zhou, Y., Zhu, Q., Shen, M., Zong, L., Lyu,
G., Zhao, Y., et al. (2013). NuRD blocks reprogramming of mouse somatic cells
into pluripotent stem cells. Stem Cells 31, 1278–1286.
Matsumoto, A., Takeishi, S., Kanie, T., Susaki, E., Onoyama, I., Tateishi, Y.,
Nakayama, K., and Nakayama, K.I. (2011). p57 is required for quiescence
andmaintenance of adult hematopoietic stem cells. Cell StemCell 9, 262–271.
Megyola, C.M., Gao, Y., Teixeira, A.M., Cheng, J., Heydari, K., Cheng, E.C.,
Nottoli, T., Krause, D.S., Lu, J., and Guo, S. (2013). Dynamic migration and
cell-cell interactions of early reprogramming revealed by high-resolution
time-lapse imaging. Stem Cells 31, 895–905.
Mikkelsen, T.S., Hanna, J., Zhang, X., Ku, M., Wernig, M., Schorderet, P.,
Bernstein, B.E., Jaenisch, R., Lander, E.S., andMeissner, A. (2008). Dissecting
direct reprogramming through integrative genomic analysis. Nature 454,
49–55.
Pan, X., Durrett, R.E., Zhu, H., Tanaka, Y., Li, Y., Zi, X., Marjani, S.L.,
Euskirchen, G., Ma, C., Lamotte, R.H., et al. (2013). Two methods for full-
length RNA sequencing for low quantities of cells and single cells. Proc.
Natl. Acad. Sci. USA 110, 594–599.
Papapetrou, E.P., Tomishima, M.J., Chambers, S.M., Mica, Y., Reed, E.,
Menon, J., Tabar, V., Mo, Q., Studer, L., and Sadelain, M. (2009). Stoichio-
metric and temporal requirements of Oct4, Sox2, Klf4, and c-Myc expression
for efficient human iPSC induction and differentiation. Proc. Natl. Acad. Sci.
USA 106, 12759–12764.
Park, I.H., Zhao, R., West, J.A., Yabuuchi, A., Huo, H., Ince, T.A., Lerou, P.H.,
Lensch, M.W., and Daley, G.Q. (2008). Reprogramming of human somatic
cells to pluripotency with defined factors. Nature 451, 141–146.
Pauklin, S., and Vallier, L. (2013). The cell-cycle state of stem cells determines
cell fate propensity. Cell 155, 135–147.
Polo, J.M., Anderssen, E., Walsh, R.M., Schwarz, B.A., Nefzger, C.M., Lim,
S.M., Borkent, M., Apostolou, E., Alaei, S., Cloutier, J., et al. (2012). A molec-
ular roadmap of reprogramming somatic cells into iPS cells. Cell 151, 1617–
1632.
Pop, R., Shearstone, J.R., Shen, Q., Liu, Y., Hallstrom, K., Koulnis, M., Grib-
nau, J., and Socolovsky, M. (2010). A key commitment step in erythropoiesis
is synchronized with the cell cycle clock through mutual inhibition between
PU.1 and S-phase progression. PLoS Biol. 8, e1000484.
Rais, Y., Zviran, A., Geula, S., Gafni, O., Chomsky, E., Viukov, S., Mansour,
A.A., Caspi, I., Krupalnik, V., Zerbib, M., et al. (2013). Deterministic direct re-
programming of somatic cells to pluripotency. Nature 502, 65–70.
Redmer, T., Diecke, S., Grigoryan, T., Quiroga-Negreira, A., Birchmeier, W.,
and Besser, D. (2011). E-cadherin is crucial for embryonic stem cell pluripo-
tency and can replace OCT4 during somatic cell reprogramming. EMBO
Rep. 12, 720–726.
Ruiz, S., Panopoulos, A.D., Herrerıas, A., Bissig, K.D., Lutz, M., Berggren,
W.T., Verma, I.M., and Izpisua Belmonte, J.C. (2011). A high proliferation
rate is required for cell reprogramming and maintenance of human embryonic
stem cell identity. Curr. Biol. 21, 45–52.
Sage, J., Mulligan, G.J., Attardi, L.D., Miller, A., Chen, S., Williams, B., Theo-
dorou, E., and Jacks, T. (2000). Targeted disruption of the three Rb-related
genes leads to loss of G(1) control and immortalization. Genes Dev. 14,
3037–3050.
Sakaue-Sawano, A., Kurokawa, H., Morimura, T., Hanyu, A., Hama, H.,
Osawa, H., Kashiwagi, S., Fukami, K., Miyata, T., Miyoshi, H., et al. (2008).
Visualizing spatiotemporal dynamics of multicellular cell-cycle progression.
Cell 132, 487–498.
Samavarchi-Tehrani, P., Golipour, A., David, L., Sung, H.K., Beyer, T.A., Datti,
A., Woltjen, K., Nagy, A., and Wrana, J.L. (2010). Functional genomics reveals
a BMP-driven mesenchymal-to-epithelial transition in the initiation of somatic
cell reprogramming. Cell Stem Cell 7, 64–77.
Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc. 13
Please cite this article in press as: Guo et al., Nonstochastic Reprogramming from a Privileged Somatic Cell State, Cell (2014), http://dx.doi.org/10.1016/j.cell.2014.01.020
Smith, Z.D., Nachman, I., Regev, A., and Meissner, A. (2010). Dynamic single-
cell imaging of direct reprogramming reveals an early specifying event. Nat.
Biotechnol. 28, 521–526.
Stadtfeld, M., and Hochedlinger, K. (2010). Induced pluripotency: history,
mechanisms, and applications. Genes Dev. 24, 2239–2263.
Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells
from mouse embryonic and adult fibroblast cultures by defined factors. Cell
126, 663–676.
Takahashi, K., Nakayama, K., and Nakayama, K. (2000). Mice lacking a CDK
inhibitor, p57Kip2, exhibit skeletal abnormalities and growth retardation.
J. Biochem. 127, 73–83.
Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K.,
and Yamanaka, S. (2007). Induction of pluripotent stem cells from adult human
fibroblasts by defined factors. Cell 131, 861–872.
Takizawa, H., Regoes, R.R., Boddupalli, C.S., Bonhoeffer, S., and Manz, M.G.
(2011). Dynamic variation in cycling of hematopoietic stem cells in steady state
and inflammation. J. Exp. Med. 208, 273–284.
Tsunekawa, Y., Britto, J.M., Takahashi, M., Polleux, F., Tan, S.S., and Osumi,
N. (2012). Cyclin D2 in the basal process of neural progenitors is linked to non-
equivalent cell fates. EMBO J. 31, 1879–1892.
Utikal, J., Polo, J.M., Stadtfeld, M., Maherali, N., Kulalert, W., Walsh, R.M.,
Khalil, A., Rheinwald, J.G., and Hochedlinger, K. (2009). Immortalization elim-
inates a roadblock during cellular reprogramming into iPS cells. Nature 460,
1145–1148.
Wakao, S., Kitada, M., Kuroda, Y., Shigemoto, T., Matsuse, D., Akashi, H.,
Tanimura, Y., Tsuchiyama, K., Kikuchi, T., Goda, M., et al. (2011). Multiline-
14 Cell 156, 1–14, February 13, 2014 ª2014 Elsevier Inc.
age-differentiating stress-enduring (Muse) cells are a primary source of
induced pluripotent stem cells in human fibroblasts. Proc. Natl. Acad. Sci.
USA 108, 9875–9880.
Wang, Y., and Blelloch, R. (2009). Cell cycle regulation by MicroRNAs in
embryonic stem cells. Cancer Res. 69, 4093–4096.
Wang, Z., Yu, R., and Melmed, S. (2001). Mice lacking pituitary tumor trans-
forming gene show testicular and splenic hypoplasia, thymic hyperplasia,
thrombocytopenia, aberrant cell cycle progression, and premature centro-
mere division. Mol. Endocrinol. 15, 1870–1879.
Wernig, M., Meissner, A., Foreman, R., Brambrink, T., Ku, M., Hochedlinger,
K., Bernstein, B.E., and Jaenisch, R. (2007). In vitro reprogramming of fibro-
blasts into a pluripotent ES-cell-like state. Nature 448, 318–324.
Wernig, M., Lengner, C.J., Hanna, J., Lodato, M.A., Steine, E., Foreman, R.,
Staerk, J., Markoulaki, S., and Jaenisch, R. (2008). A drug-inducible transgenic
system for direct reprogramming of multiple somatic cell types. Nat.
Biotechnol. 26, 916–924.
White, J., and Dalton, S. (2005). Cell cycle control of embryonic stem cells.
Stem Cell Rev. 1, 131–138.
Yamanaka, S. (2009). Elite and stochastic models for induced pluripotent stem
cell generation. Nature 460, 49–52.
Zou, P., Yoshihara, H., Hosokawa, K., Tai, I., Shinmyozu, K., Tsukahara, F.,
Maru, Y., Nakayama, K., Nakayama, K.I., and Suda, T. (2011). p57(Kip2) and
p27(Kip1) cooperate to maintain hematopoietic stem cell quiescence through
interactions with Hsc70. Cell Stem Cell 9, 247–261.