REGULAR ARTICLE
Different stages of pluripotency determine distinct
patterns of proliferation, metabolism, and lineage
commitment of embryonic stem cells under hypoxia
Tiago G. Fernandes, Maria Margarida Diogo, Ana Fernandes-Platzgummer,
Cláudia Lobato da Silva, Joaquim M.S. Cabral⁎
Institute for Biotechnology and Bioengineering (IBB), Centre for Biological and Chemical Engineering,
Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
Received 15 October 2009; received in revised form 14 April 2010; accepted 14 April 2010
Abstract Oxygen tension is an important component of the stem cell microenvironment. Herein, we have studied the effect
of low oxygen levels (2% O2), or hypoxia, in the expansion of mouse embryonic stem (ES) cells. In the presence of leukemia
inhibitory factor (LIF), cell proliferation was reduced under hypoxia and a simultaneous reduction in cell viability was also observed.
Morphological changes and different cell cycle patterns were observed, suggesting some early differentiation under hypoxic
conditions. However, when cells were maintained in a ground state of pluripotency, by inhibition of autocrine FGF4/ERK and
GSK3 signaling, hypoxia did not affect cell proliferation, and did not induce early differentiation. As expected, there was an
increase in lactate-specific production rate and a significant increase in the glucose consumption under hypoxic conditions.
Nevertheless, during neural commitment, low oxygen tension exerted a positive effect on early differentiation of ground-state
ES cells, resulting in a faster commitment toward neural progenitors. Overall our results demonstrate the need to specifically
regulate the oxygen content, especially hypoxia, along with other culture conditions, when developing new strategies for
ES cell expansion and/or controlled differentiation.
© 2010 Elsevier B.V. All rights reserved.
Introduction
Stem cells possess the ability for unlimited or prolonged selfrenewal
and to differentiate into highly distinct cell lineages.
In particular, pluripotent embryonic stem (ES) cells can give
rise to cells derived from the three embryonic germ layers
(ectoderm, endoderm, and mesoderm) (Smith, 2001), which
makes them attractive for a wide range of clinical and
pharmacological applications (Klimanskaya et al., 2008;
Pouton and Haynes, 2005). However, a better understanding
of the mechanisms that control the expansion and differentiation
of ES cells is of great importance for the systematic
production of cells for therapeutic applications (Kirouac and
Zandstra, 2008). Interestingly, the oxygen tensions to which
cells are exposed during development are substantially lower
than atmospheric levels (Jauniaux et al., 2003; Lee et al.,
2001), making O2 content a potentially important parameter
when designing new strategies for stem cell expansion and/or
controlled differentiation.
Mouse ES cells, in particular, can be maintained in culture in
an undifferentiated state in the presence of leukemia
inhibitory factor (LIF) and serum, or under serum-free
⁎ Corresponding author. Fax: +351 21 841 90 62.
E-mail address: joaquim.cabral@ist.utl.pt (J.M.S. Cabral).
1873-5061/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.scr.2010.04.003
available at www.sciencedirect.com
www.elsevier.com/locate/scr
Stem Cell Research (2010) 5, 76–89
conditions with LIF and bone morphogenetic protein (BMP-4)
(Ying et al., 2003a). However, recent evidence has highlighted
the existence of metastable pluripotent states, demonstrating
that ES cells can assume distinct states of pluripotency in vitro
(Hanna et al., 2009). Therefore, two stages of pluripotency can
be defined: a naïve or ground state, and a primed or
committed state of pluripotency (Nichols and Smith, 2009).
Culture conditions for the maintenance of ground-state ES
cells were developed and involve the utilization of smallmolecule
inhibitors of fibroblast growth factor (FGF)/extracellular
signal-regulated kinases (ERK) and glycogen synthase
kinase-3 (GSK3) signaling (Ying et al., 2008). Auto-inductive
FGF4/ERK signaling poises ES cells susceptible for lineage
commitment, and cells enter into a primed pluripotent state
(Silva and Smith, 2008). Thus, inhibition of ERK signaling,
augmented by inhibition of GSK3 to restore cellular growth and
viability, allows the propagation of naïve pluripotent ES cells.
The influence of oxygen levels on this dynamic regulatory
mechanism has not yet been studied, and may potentially
provide new insights into ES cell proliferation, metabolism,
and phenotype retention. Furthermore, in the absence of
serum, elimination of inductive signals for alternative fates
renders ES cells to develop efficiently into neural precursors,
as the result of autocrine FGF signaling (Ying et al., 2003b). It
has also been suggested that culture under low oxygen
tensions allows spontaneous ES cell differentiation (Kurosawa
et al., 2006), and recent evidence has revealed that
reduced oxygen tensions are beneficial for the proliferation
and differentiation of fetal and postnatal neural stem cells
(Morrison et al., 2000; Pistollato et al., 2007). Thus, oxygen
tension control during neural commitment of ground-state ES
cells could potentially be used for improved efficiency in the
generation of ES cell-derived neural progenitors.
The influence of low oxygen levels (2%) on the proliferation
and neural differentiation of ground-state embryonic
stem cells was evaluated in this work. A comprehensive
analysis of ground-state ES cell proliferation kinetics, cell
viability, expression of pluripotency markers, cell cycle, and
metabolic profile was performed for low oxygen tensions and
compared with the same results obtained for atmospheric
oxygen levels. In addition, the conversion of ground-state ES
cells into neural precursors under low oxygen tension was
also studied. Consequently, this work is expected to
contribute valuable information for the improvement of
current strategies for the expansion and neural differentiation
of embryonic stem cells.
Results
Determination of dissolved oxygen levels in liquid
culture medium
The time-dependent response of the liquid-phase oxygen
levels in tissue culture plates under simulated culture
conditions is shown in Figure 1. In the case of incubation
under 20% O2, the dissolved oxygen at the bottom of the
culture medium layer was similar to the gas-phase tension.
Under hypoxia, oxygen levels dropped swiftly, and were
close to 4% within 5 h. The dissolved O2 continued to drop at
later times, and eventually reached equilibrium with the gas
phase (2%). It is possible that oxygen consumption by cells at
the bottom of the tissue culture plate under actual culture
conditions might have accelerated the decrease of dissolved
oxygen levels, thus making the actual response time much
smaller. Nevertheless, cells cultured under hypoxia are
exposed to higher O2 levels for at least 5 h once every
2 days due to cell passaging.
Effect of hypoxia on mouse ES cell expansion and
metabolism in the presence of LIF
To evaluate the influence of low oxygen tension (2%) on
mouse ES cell expansion, cells were cultured for five
consecutive passages in KO-DMEM/SR medium supplemented
with LIF. Under these conditions, LIF/STAT3 signaling should
contribute for self-renewal of mouse ES cells (Niwa et al.,
1998). However, as can be seen in Figure 2, hypoxia resulted
in a reduced cell proliferation rate. The overall cumulative
fold increase (FI) in total cell number was substantially lower
in 46C mouse ES cells expanded under 2% O2 (FI=73±11)
(Fig. 1a), when compared with cells maintained at atmospheric
O2 levels (FI=6535±2732) (Pb0.05). This resulted in a
statistically significant decrease in the specific growth rate
(µ) of mouse ES cells expanded at low oxygen tension
(Fig. 2b). In fact, the µ value calculated for 2% O2 was
about half of the typical specific growth rate value for cells
expanded in the presence of LIF (0.43±0.01 day-1
at 2% O2 and
0.84±0.06 day-1
at 20% O2). Interestingly, proliferation was
partially recovered under hypoxia by supplementation of the
culture medium with CHIR99021, a GSK3-specific inhibitor, in
addition to LIF (µ=0.67±0.04 day-1
at 2% O2 with 5 µM of the
molecule, data not shown). This suggests an important role of
this pathway in the regulation of stem cell functions under
hypoxic conditions. Similar results were also observed for
E14tg2a mouse ES cells (Fig. S1, Supplementary Data).
All proliferation results were further confirmed by cell
division analysis using PKH67 fluorescent dye (da Silva et al.,
2009). The determination of the number of cell divisions by
flow cytometry after 2 days in culture revealed that cells
expanded at low O2 levels divided fewer times than cells
cultured at 20% O2, which reached later generations during
Figure 1 Dissolved oxygen in the liquid phase at the bottom of
a tissue culture plate. Measurements were made in 1 mL of
medium and O2 levels were plotted as a function of time. The
liquid was initially equilibrated at atmospheric oxygen levels,
and the plate was then transferred to a humidified incubator
either at 20 or 2% O2 atmosphere.
77Metabolism and Lineage Commitment of Embryonic Stem Cells
the same period (Fig. 2c). Indeed, most cells (over 80%)
cultured at 2% O2 were at generations 3 and 4 after 2 days of
culture (in comparison with 47% for 20% O2), whereas a
significant fraction of cells (over 35%) at 20% O2 had already
reached generations 5 and 6. Furthermore, low oxygen
tension resulted in a decrease in cell viability throughout
time in culture. In fact, after five consecutive passages, cell
viability (assessed by trypan blue exclusion test) at 2% O2 was
77±2%, and at 20% O2 it was 91±2% (Pb0.05). Together with
a decrease in the number of cell divisions, this result helps in
explaining the reduced overall cell fold increase and the
lower specific growth rates obtained at low oxygen levels.
Analysis of cell morphology also revealed striking differences
between cells expanded at 2% O2 and cells maintained at
atmospheric O2 levels (Fig. 2d). Under normoxia (20% O2), in
the presence of LIF, cells displayed typical compact colony
morphology, a characteristic of undifferentiated ES cells. On
the other hand, when cultured under hypoxia (2% O2), cells
became largely flattened and sharpen at the edges of the
colonies, a fibroblast-like morphology typical of early
commitment. Nevertheless, addition of 5 µM of GSK3specific
inhibitor CHIR99021 was able to recover typical
Figure 2 Hypoxia resulted in reduced mouse ES cell proliferation in the presence of LIF. (a) 46 C mouse ES cells were grown in KO-DMEM/SR
medium supplemented with LIF. Results express the cumulative fold increase in total cell number of five consecutive passages performed in
triplicate. (b) Specific growth rates calculated for each oxygen tension tested. Results are presented as the mean of three independent
experiments. Asterisk denotes statistical significance (Pb0.05). (c) Cell division analysis using PKH67 fluorescent dye. Percentage of cells in
each generation (parent to G9) is shown for each oxygen tension tested. Results are presented as the mean of two independent experiments.
(d) Mouse ES cell morphology under 20 and 2% O2 conditions. Scale bar: 50 µm. (e) Real-time PCR analysis of pluripotency markers (Nanog and
Oct3/4) and primed markers (Afp, Eomes, and Fgf5) following 46C mouse ES cell expansion in KO-DMEM/SR medium supplemented with LIF at
20 or 2% O2 for five consecutive passages. Results are expressed as the average value of two independent experiments and are relative to gene
expression at Day 0. The expression levels of the housekeeping gene Gapdh were used as internal control. (f) Cell cycle pattern analysis of
mouse ES cells expanded in the presence of LIF at 2 and 20% O2. Asterisk denotes statistical significance (Pb0.05). All error bars represent the
standard error of the mean (SEM).
78 T.G. Fernandes et al.
compact ES cell morphology in hypoxia (Fig. 2d), which
provides additional evidence of the positive effect of GSK3
inhibition at low oxygen tensions (Mottet et al., 2003).
Quantification of pluripotency markers (Oct-4 and Nanog) by
flow cytometry did not reveal any statistically significant
differences in the levels of these proteins, as measured by
the peak mean intensity values (data not shown), with more
than 95% Oct-4- and Nanog-positive cells under both
conditions tested (Fig. S2, Supplementary Data).
The results so far indicate that expansion of mouse ES
cells in the presence of LIF at low oxygen tensions can result
in a reduction of the proliferation rate, increase in cell
death, and morphology changes, although pluripotency
markers remained expressed in a high percentage of cells.
To further confirm that under these conditions hypoxia leads
to an early priming of mouse ES cells toward commitment,
quantitative real-time polymerase chain reaction (RT-PCR)
was performed after cell expansion (Fig. 2e). It is possible to
observe a tendency for reduction in the relative expression
of pluripotency markers (Oct3/4 and Nanog) under hypoxia.
At the same time, there was a slight increase in the
expression of primed markers (Afp, Eomes, and Fgf5) at 2%
O2, which is more evident for the primitive ectoderm marker
Fgf5. In addition, the cell cycle profile was also assayed by
flow cytometry for cells expanded at 2 and 20% O2 (Fig. 2f).
Mouse ES cells cultured under normoxia (20% O2) showed a
typical cell cycle profile, displaying a large proportion of
cells in the S phase, resulting from the absence of G1
checkpoints (Burdon et al., 2002). However, when maintained
at low oxygen levels, the proportion of cells in G0/G1
increases significantly (Pb0.05), which may denote a
modification in the G1-phase arrest that is characteristic of
early differentiation (Burdon et al., 2002). Therefore, our
results suggest that hypoxia induces an early priming of
mouse ES cells toward differentiation when LIF is used for
maintenance of the pluripotent state.
In addition to affecting mouse ES cell proliferation and
pluripotency, low oxygen levels also influenced cell metabolism
(Fig. 3). The average consumption/production rates (in
mM/day) of glucose, glutamine, lactate, and ammonia were
not substantially affected by the reduction in oxygen level,
remaining fairly constant for cells expanded at 2% O2 and 20%
O2 (Fig. 3a). More importantly, concentrations of potentially
toxic metabolites (lactate and ammonia) did not reach levels
considered to be inhibitory for mammalian cells (Ozturk
et al., 1992). However, since hypoxia resulted in lower cell
numbers, a better comparison of both conditions should be
made using the specific (per cell) consumption/production
rates. This revealed a significantly higher (Pb0.05) production
of lactate per cell under hypoxia (20.3±1.1 pmol/(cell.
day) versus 11.7±2.1 pmol/(cell.day) at 20% O2), which
might be expected under low oxygen tensions as the result of
increased anaerobic metabolism. In addition, a tendency for
slightly higher consumption of glutamine per cell was
observed at low oxygen levels (1.2±0.1 pmol/(cell.day) at
2% O2 and 0.8±0.1 pmol/(cell.day) at 20% O2, P=0.056).
Evidence of these metabolic changes under hypoxic
conditions can also be seen when analyzing the apparent
metabolic yields of the culture (Figs. 3b and c). Even though
the apparent lactate from glucose yield (Y′lactate/glucose) was
Figure 3 Influence of hypoxia on mouse ES cell metabolism during expansion with LIF. (a) Average concentration change in
glucose, lactate, glutamine, and ammonia. (b) Apparent yields of lactate from glucose (Y′lactate/glucose) and ammonia from
glutamine (Y′ammonia/glutamine) calculated for 2 and 20% O2 conditions. (c) Cell yields from glucose (Y′µ/glucose) and glutamine
(Y′µ/glutamine) calculated for low oxygen levels and for normal (20%) oxygen levels. Cells were expanded in KO-DMEM/SR
medium supplemented with LIF and results express the average values of five consecutive passages performed in triplicate.
Asterisks denote statistical significance (Pb0.05).
79Metabolism and Lineage Commitment of Embryonic Stem Cells
above 2 mol of lactate per mol of glucose (the theoretical
limit) for both oxygen tensions tested, Y′lactate/glucose at 2% O2
was substantially higher than at 20% O2 (Fig. 3b) (Pb0.05).
Production of lactate from other sources, like glutamine,
explains why yields greater than 2 can be obtained (Reitzer
et al., 1979). However, according to our results, the relevance
of alternative metabolic pathways, such as glutaminolysis,
seems to be greater under hypoxia. Interestingly, no differences
were seen in the values of the apparent ammonia from
glutamine yield (Y′ammonia/glutamine) (Fig. 3b), but cell yields
from glucose and glutamine calculated for low oxygen levels
were significantly lower than the ones obtained for atmospheric
(20%) oxygen level (Fig. 3c, Pb0.05), indicating that
mouse ES cell metabolism under expansion conditions in the
presence of LIF is less efficient under hypoxia.
Effect of hypoxia on expansion and metabolism of
ground-state pluripotent mouse ES cells
The results presented in the previous section are relative to
ES cells expanded under conditions that activate LIF/STAT3
signaling in order to maintain cell pluripotency. However, as
previously reported, hypoxia suppresses LIF/STAT3 signaling
via hypoxia-inducible transcription factor 1-alpha (HIF-1α)
(Jeong et al., 2007), which may explain the apparent early
differentiation induced by low levels of O2. Nevertheless,
maintenance of ground-state pluripotency can be attained in
culture by inhibition of FGF4/ERK and GSK3 signaling, a
mechanism that is independent and upstream of LIF/STAT3
(Ying et al., 2008). Thus, the influence of low oxygen
tensions on ground-state ES cell proliferation and metabolic
profile may result in different outcomes.
To study this possibility, mouse ES cells were cultured for
five consecutive passages in iSTEM serum-free medium to
sustain ground-state pluripotency, either under atmospheric
oxygen tension (20% O2) or reduced oxygen levels (2% O2).
The proliferation of mouse ES cells under these conditions
was not affected and the overall cumulative fold increase in
total cell number was similar (FI=3911±1429 for 20% O2 and
FI=2646±488 for 2% O2), independently of the oxygen level
used (Fig. 4a). This resulted in identical growth rate values
for both conditions tested (0.82±0.04 day-1
for 20% O2 and
0.79±0.02 day-1
for 2% O2). Similar results were also
observed for E14tg2a mouse ES cells (Fig. S3, Supplementary
Data). Interestingly, cell viability at 2% O2 was always
significantly lower (Pb0.05 for all data points) when
compared with cell viability at 20% O2 throughout time in
culture (Fig. 4b), indicating a detrimental effect of hypoxia
on cell viability that was already seen for mouse ES cells
expanded in the presence of LIF. In addition, over 95% of
cells were positive for Oct-4 and Nanog, either at low or at
atmospheric oxygen tensions, as quantified by flow cytometry
(Fig. S4, Supplementary Data).
Mouse ES cell expansion at clonal densities was also
performed as a rigorous test to further evaluate the
effectiveness of low oxygen levels in sustaining groundstate
ES cells (Fig. 4c). Cells were seeded at low densities and
maintained in iSTEM serum-free medium for 6 days with
medium change at Days 2 and 4. After single-cell deposition,
both oxygen tensions gave rise to ES cell colonies with a
similar efficiency (23±5 and 23±2% for low and atmospheric
oxygen levels, respectively). However, colony size distribution
was more homogeneous at 20% O2 (Fig. 4d), and the
average colony diameter was smaller (72±11 µm) when
compared to 2% O2 (99±21 µm). Thus, this result indicates
that at clonal densities, cells cultured in a reduced oxygen
atmosphere give rise to larger colonies. In addition, cell
division analysis using PKH67 fluorescent dye by flow
cytometry revealed that cells expanded at low O2 levels
divided more times under these conditions, reaching higher
percentages of cells in later generations when compared
with 20% O2 (Fig. 4e). Moreover, mouse ES cells expanded in
a reduced oxygen atmosphere displayed a cell cycle profile
identical to cells cultured under normoxia. Both populations
showed a profile characteristic of ES cells, with a large
proportion of cells in the S phase (52±2 and 53±2% of cells at
reduced O2 levels and normoxia, respectively), resulting
from the absence of G1 checkpoint. No statistically
significant differences were observed between the cell
percentages in G0/G1 for the O2 levels tested (28±1% of
cells at 2% O2 and 26±1% of cells at 20% O2). Real-time PCR
analysis also shows no major differences in the expression
levels of pluripotency markers (Oct3/4 and Nanog) between
cells maintained under different oxygen conditions. In
addition, hypoxia did not result in increased expression of
primed markers, such as Fgf5, Afp, or Eomes (Fig. S5,
Supplementary Data). These results highlight that oxygen
levels did not affect the ground-state pluripotency of
authentic ES cells. Nevertheless, despite possessing a
negative influence on cell viability under ground-state
culture conditions, hypoxia also appeared to accelerate
cell proliferation, which was confirmed by flow cytometry
using the PKH67 fluorescent dye. Therefore, the overall
proliferation rate was not affected and similar cell numbers
were produced in both oxygen tensions tested.
The effect of low oxygen levels on the metabolism of
ground-state mouse ES cells was also evaluated (Fig. 5). The
average consumption/production rates (in mM/day) of
glucose, glutamine, lactate, and ammonia were calculated
for 2% O2 and 20% O2 conditions (Fig. 5a). The reduction in
oxygen tension did not affect glutamine consumption and
ammonia production. Nonetheless, hypoxia seemed to
increase to some extent glucose consumption and lactate
production. Analysis of the specific (per cell) consumption/
production rates also supports this evidence, and under
hypoxia an increase in glucose consumption (10.2±1.7 pmol/
(cell.day) versus 4.1 ± 0.4 pmol/(cell.day) at 20% O2)
(Pb0.05) was also accompanied with an increase in lactate
production (14.8±2.2 pmol/(cell.day) versus 9.3±1.3 pmol/
(cell.day) at 20% O2), which might be expected under low
oxygen tensions as the result of an increased anaerobic
metabolism. Importantly, concentrations of potentially toxic
metabolites, especially lactate, did not reach inhibitory
levels. The apparent metabolic yields were also calculated
for the expansion of ground-state ES cells at low oxygen
levels (Figs. 5b and c). Interestingly, in contrast with the
results obtained for mouse ES cell expansion with LIF, no
significant differences were seen in the values of the
apparent lactate from glucose yield (Y′lactate/glucose) and
the apparent ammonia from glutamine yield (Y′ammonia/
glutamine) (Fig. 5b). Cell yields from glucose and glutamine
calculated for 2 and 20% O2 were comparable and within the
error, indicating that mouse ES cell metabolism during
80 T.G. Fernandes et al.
expansion under ground-state conditions was not negatively
influenced by a hypoxic environment.
Neural commitment of ground-state mouse ES cells
under hypoxic conditions
We next reasoned that oxygen tension control during neural
commitment of ground-state ES cells could potentially be
used for improved efficiency in the generation of ES cellderived
neural progenitors. Therefore, ground-state ES cells
(expanded in iSTEM serum-free medium) were collected and
plated onto gelatin-coated tissue culture plates and incubated
for 8 days in neural differentiation medium. Under these
conditions, elimination of inductive signals for alternative
fates conducts ES cells to develop into neural precursors
through autocrine FGF signaling (Ying et al., 2003b).
The neural commitment protocol was carried out under
either atmospheric oxygen tension (20% O2) or reduced
oxygen levels (2% O2). Plating efficiency was similar for
both oxygen tensions studied, and total cell numbers
obtained throughout time in culture did not differ
considerably until Day 5 (Fig. 6a). However, starting at
Day 6, 20% O2 conditions promoted significantly higher fold
increase values in total cell number, reaching a value of 51
±7 at Day 8, while for 2% O2 conditions the final fold
Figure 4 Influence of low oxygen levels on ground-state mouse ES cell expansion. (a) Influence of oxygen tension on 46C mouse ES
cell cumulative fold increase during expansion in iSTEM serum-free medium. Results express the cumulative fold increase in total cell
number of five consecutive passages performed in triplicate. (b) Percentage of cell viability for each oxygen tension tested. Results
are presented as the mean of three independent experiments. (c) Mouse ES cell colonies at 20 and 2% O2 conditions. Scale bar: 50 µm.
(d) Colony diameter (in µm) distribution for 20 and 2% O2 conditions. A total of 40 colonies were measured from four independent
experiments. (e) Cell division analysis using PKH67 fluorescent dye. Percentage of cells in each generation (parent to G9) is shown for
each oxygen tension tested. Results are presented as the mean of two independent experiments. All error bars represent the standard
error of the mean (SEM).
81Metabolism and Lineage Commitment of Embryonic Stem Cells
increase was about half of this value (22±2). During the
same period (from Day 6 to Day 8) cell viability at 2% O2 was
substantially reduced (around 47% viability at Day 8),
explaining the differences observed in cell fold increase
values (Fig. 6b).
The use of a green fluorescent protein (GFP) knock-in
reporter ES cell line (46C) allowed the examination of the
process by which ES cells acquire neural identity since the
open-reading frame of the Sox1 gene was replaced with the
coding sequence of GFP. Sox1 is an early marker of
neuroectoderm in the mouse embryo and can be used to
study neural differentiation in vitro (Ying et al., 2003b).
Thus, evaluation of neural commitment kinetics is possible
by flow cytometry analysis (Fig. 6c). This revealed that the
maximum expression of Sox1-GFP was reached faster at 2%
O2 than at 20% O2, and by Day 4 cells maintained under low
oxygen levels were already over 80% Sox1 positive (83.8±
3.1%) whereas under 20% O2 the percentage of Sox1positive
cells was only 65.6±7.3%. Nevertheless, the
maximum level of Sox1-GFP cells was identical for both
oxygen tensions tested (85.9±1.5% at 2% O2 and 83.0±1.9%
at 20% O2, both at Day 5). Interestingly, at atmospheric
oxygen tension the percentage of Sox1-positive cells
remained fairly constant after reaching the maximum at
Day 5. However, a decrease in Sox1 levels occurred for low
oxygen tension, and by Day 8 it was significantly lower than
at 20% O2, possibly indicating a further differentiation into
more mature neural cells. In fact, by Day 8, flow cytometry
quantification of neuronal class III β-tubulin (Tuj1) revealed
a slightly higher percentage of neuronal cells at low oxygen
tension (Fig. 6d).
Sox1-positive neural progenitors are nevertheless a
transient population. Hence, to evaluate the influence of
low oxygen tension on further neuronal differentiation, cells
maintained at 20% O2 were collected at Day 6 of neural
commitment and replated onto tissue culture plates precoated
with polyornithine and laminin, as described previously
(Pollard et al., 2006). At both oxygen tensions tested,
cells readily adhered to the surface and showed typical
morphology, as indicated by the establishment of complex
branches and cell-to-cell contacts (Fig. 6e). However, no
significant differences in terms of neuronal maturation
dynamics were seen when comparing 2% O2 and 20% O2
(Fig. 6f), indicating that oxygen tension did not affect this
process at least under these culture conditions. However,
our results appear to indicate a beneficial effect of hypoxia
on early commitment of ground-state ES cells toward
primitive neuroectoderm.
In parallel, the metabolism of neural commitment of
ground-state mouse ES cells was also evaluated (Fig. 7).
During neural commitment of ground-state mouse ES cells
variations in the consumption/production rates of important
metabolites were observed. In general, for both oxygen
tensions tested, the specific (per cell) consumption/production
rates decreased throughout time in culture (Fig. 7a),
consistent with higher initial cell growth followed by
adaptation of the cells to the culture environment. Interestingly,
very high levels of glucose consumption were observed,
especially during the first days in culture (until Day 2). In fact,
the medium used (RHB-A) displays glucose levels (35 mM) that
are substantially higher than what is typically used in other
culture media (e.g., 20 mM), while at the same time
Figure 5 Influence of hypoxia on the metabolism of ground-state mouse ES cell expansion. (a) Average concentration change in glucose,
lactate, glutamine, and ammonia levels. (b) Apparent yields of lactate from glucose (Y′lactate/glucose) and ammonia from
glutamine (Y′ammonia/glutamine) calculated under 2 and 20% O2 conditions. (c) Cell yields from glucose (Y′µ/glucose) and glutamine
(Y′µ/glutamine) calculated for low oxygen levels and for atmospheric (20%) oxygen levels. Cells were expanded in iSTEM serum-free
medium and results express the mean values of five consecutive passages performed in triplicate.
82 T.G. Fernandes et al.
glutamine concentrations are lower (0.07 mM as compared to
2 mM in typical mammalian cell culture media). This also
resulted in high lactate-specific production rates until Day 4
of neural commitment. However, no differences between the
results obtained for both oxygen tensions were seen during
this period. Nevertheless, starting at Day 4, hypoxia led to
substantially higher lactate production rates. Moreover,
glucose-specific consumption rates were also significantly
higher by Day 8 at low oxygen levels. This result might be due
to an increase in anaerobic metabolism at low oxygen
tensions. Importantly, between Day 4 and Day 6, lactate
concentrations were higher than 20 mM, a value known to be
inhibitory for other stem cells (e.g., mesenchymal) under
hypoxia (Schop et al., 2009). This may have affected cell
viability, possibly explaining, at least in part, the limited cell
growth and increased cell death observed under these
conditions when compared to normoxia (Figs. 6a and b).
The apparent yields of lactate from glucose (Y′lactate/glucose)
during neural commitment were also calculated (Fig. 7b).
Interestingly, no differences were seen between the two
oxygen levels tested. Furthermore, Y′lactate/glucose values were
always below 2 mol of lactate per mol of glucose (the
theoretical limit) for both oxygen tensions tested, suggesting
that alternative metabolic pathways, such as glutaminolysis,
were not so relevant under these conditions. In fact, during the
first 2 days of commitment, Y′lactate/glucose values were
significantly lower when compared with later times in culture.
After this period, Y′lactate/glucose values increased and remained
fairly constant until Day 8. The low glutamine concentrations
in the culture medium used may explain these results,
Figure 6 Influence of low oxygen levels on neural commitment of ground-state mouse ES cells. (a) Effect of oxygen tension on
total cell numbers. Cells were grown in RHB-A serum-free medium. Results express the fold increase in total cell number at each
day of neural commitment. (b) Percentage of cell viability for each oxygen tension tested throughout time in culture. (c) Percentage of
Sox1-GFP-positive cells during neural commitment at different O2 levels. (d) Flow cytometry quantification of neuronal class III β-tubulin
(Tuj1)-positive cells at Day 8 of neural commitment. (e) Cell morphology after replating of Sox1-positive neural progenitors and culture for
6 days in RHB-A serum-free medium. Scale bar: 100 µm. (f) Percentage of Tuj1-positive cells during neuronal differentiation of neural
progenitors. All results are presented as the mean of three independent experiments performed in duplicate. Error bars represent the
standard error of the mean (SEM). Asterisks denote statistical significance (Pb0.05).
83Metabolism and Lineage Commitment of Embryonic Stem Cells
underscoring the importance of this amino acid for biosynthetic
pathways, namely protein synthesis, which is more
evident during the first days in culture.
Discussion
In most mammalian tissues, cells are exposed to oxygen
tensions that are substantially lower than atmospheric
levels (Ivanovic, 2009). Hypoxia, or physiological oxygen
tension, is expected to impact not only cell growth and
phenotype but also cellular energetics and metabolism.
Thus, in order to evaluate how oxygen levels influence the
proliferation and neural differentiation of ground-state
embryonic stem cells, we performed a comprehensive
analysis of proliferation kinetics, cell viability, protein and
gene expression, cell cycle, and metabolic profiles under
different oxygen tensions.
When cells were expanded under serum-free conditions
with LIF, growth of mouse ES cells was significantly reduced
at 2% O2 conditions. However, the fraction of cells expressing
Oct-4 and Nanog, as assessed by flow cytometry, was not
affected, despite morphological changes indicative of some
early differentiation. In addition, quantitative real-time PCR
and cell cycle profile analysis further indicated evidence of
early priming of cells toward differentiation. These results
suggest that hypoxic conditions prompt mouse ES cells early
commitment when LIF is used to maintain cell pluripotency.
In fact, LIF/STAT3 signaling is suppressed by HIF-1α, which
results in hypoxia-induced in vitro differentiation (Jeong
et al., 2007). Our findings are consistent with previous
studies that also reported similar effects of hypoxia on
mouse ES cell expansion (Powers et al., 2008). Nevertheless,
Figure 7 Metabolism of ground-state mouse ES cells during neural commitment at different oxygen levels. (a) Specific
consumption/production rates of glucose, lactate, glutamine, and ammonia under 2 and 20% O2 conditions during neural commitment.
(b) Apparent yield of lactate from glucose (Y′lactate/glucose) calculated under 2 and 20% O2 conditions during neural commitment. Cells
were cultured in RHB-A serum-free medium and results express the mean of three independent experiments performed in duplicate.
Error bars represent the standard error of the mean (SEM). Asterisks denote statistical significance (Pb0.05).
84 T.G. Fernandes et al.
despite the effects of hypoxia on stem cell function (Covello
et al., 2006), the blockage of GSK3 signaling by addition of
CHIR99021 to the culture medium had a positive effect in
cells maintained at low oxygen levels with LIF. Inhibition of
GSK3 signaling resulted in partial recovery of specific growth
rate values and typical compact mouse ES cell morphology.
This may indicate a pivotal role of this pathway in the
regulation of cell functions under hypoxic conditions (Mottet
et al., 2003). Moreover, specific pharmacological inhibition
of GSK3 has been shown to maintain undifferentiated ES cells
through activation of the canonical Wnt pathway (Sato et al.,
2004). Consequently, single blockade of GSK3 restrains
neural differentiation, while enhancing growth capacity
(Ying et al., 2008). This mechanism might have contributed
to the partial recovery of ES cell phenotype under hypoxic
conditions with LIF and CHIR99021 supplementation.
Cell metabolism was also altered under hypoxic conditions in
the presence of LIF. Higher lactate production rates, indicative
of an increased anaerobic metabolism, were obtained for low
oxygen tensions. This effect has also been reported for other
stem cell populations (e.g., mesenchymal), simultaneously with
an increase in glucose consumption (Follmar et al., 2006).
Additionally, production of lactate from other sources, especially
glutamine, was also more relevant in hypoxia, as indicated
by the increase in the apparent lactate from glucose yield to
values well above 2. According to our results, it appears that a
less efficient cell metabolism arises during expansion of mouse
ES cells in the presence of LIF under hypoxic conditions
compared to normoxia.
However, the effect of hypoxia in ground-state mouse ES
cells was somewhat different. Self-renewal of authentic,
ground-state ES cells in culture does not depend on LIF/STA3
signaling. It is rather due to inhibition of FGF4/ERK and GSK3
signaling (Ying et al., 2008), a mechanism that is independent
and upstream of LIF/STAT3. Under these conditions,
hypoxia did not affect cell proliferation, which actually
appeared to be increased, especially at clonal densities.
Nevertheless, low oxygen levels had a negative impact on
cell viability. No signs of differentiation were observed, as
assessed by the analysis of pluripotency markers, cell
morphology, and cell cycle profile. Thus, high numbers of
ground-state ES cells could be generated under hypoxic
conditions, despite the negative influence on cell viability.
The contribution of GSK3 inhibition under ground-state
conditions appears to extend beyond limiting differentiation.
It is known that this pathway modulates metabolic
activity, biosynthetic capacity, and overall viability (Doble
and Woodgett, 2003). Nevertheless, our results show that
GSK3 inhibition does not restore cell viability under hypoxia.
Alternatively, reduced GSK3 activity may exert a global
modulation of cell metabolism and biosynthetic capacity
rather than having a direct antiapoptotic action, since
antiapoptotic factors were not found under ground-state
conditions (Ying et al., 2008). In theory, this could be
overcome if the negative influence of hypoxia-related cell
death could be minimized by the use of chemical inhibitors to
generate higher numbers of pluripotent ES cells (Duval et al.,
2004; Krishnamoorthy et al., 2009; Watanabe et al., 2007).
Under ground-state culture conditions, hypoxia seemed to
increase to some extent glucose and lactate metabolism,
similar to that obtained for mouse ES cell expansion in the
presence of LIF, and to that reported for adult stem cells
(Follmar et al., 2006). Nevertheless, no differences were
observed in the values of the metabolic yields, indicating that
mouse ES cell metabolism during expansion under groundstate
conditions is not negatively influenced by a hypoxic
environment. Overall, our findings indicate that ground-state
ES cells can be maintained under hypoxic conditions and
generate comparable numbers of cells without differentiation,
while maintaining an efficient cell metabolism.
We next focused on evaluating how hypoxia influences the
neural commitment of ground-state mouse ES cells. Differentiation
of ES cells toward the neural lineage is characterized
by progression through different stages resembling
neurogenesis (Sutter and Krause, 2008). When signals for
alternative fates are eliminated, mouse ES cells readily
develop into neuroectodermal precursors in adherent monolayer,
due to autocrine FGF signaling (Ying et al., 2003b).
Since the embryo is subjected to a hypoxic environment
(Rodesch et al., 1992), we reasoned that low oxygen tensions
would impact the neural commitment of ground-state mouse
ES cells.
Hypoxia had a negative effect on cell viability after 6 days
of neural commitment. In fact, potentially inhibitory levels of
lactate were reached in the culture medium during this period,
which may explain this outcome. Nevertheless, Sox1-positive
neural progenitors were generated faster at 2% O2, although
similar maximum levels of Sox1-positive cells were obtained in
both oxygen tensions tested. Furthermore, after reaching the
maximum by Day 5, a decrease in Sox1 levels occurred for low
oxygen tension, and by Day 8 it was significantly lower than at
20% O2. This was accompanied by a slight increase in the
percentage Tuj1-positive cells, indicating a further differentiation
in more mature neural cells. Despite this, oxygen
tension did not affect neuronal maturation after replating of
neural progenitors on laminin-coated surfaces. Thus, hypoxia
only appears tohave a positive effect on the early commitment
of ground-state ES cells toward primitive neuroectoderm.
Cell metabolism during neural commitment was also
influenced by hypoxic conditions. High levels of glucose
consumption and lactate production were seen for 2 and 20%
O2 during the first days of commitment. Nevertheless, later
in culture, hypoxia led to substantially higher glucose and
lactate consumption/production rates due to an increase in
anaerobic metabolism at low oxygen tensions. Interestingly,
very low yields of lactate from glucose were obtained during
the first 2 days of commitment. Also, despite increasing
significantly at later days, Y′lactate/glucose values were always
below 2 mol of lactate per mol of glucose (the theoretic
limit) for both oxygen tensions tested, indicating that
alternative metabolic pathways to glycolysis, such as
glutaminolysis, were not favored under these conditions. In
fact, glutamine concentrations in the culture medium used
are within the range of limiting levels described for
mammalian cell culture (0.5 to 2 mM (Vriezen et al.,
1997)), and under these circumstances cells may only use
glutamine for biosynthesis during neural commitment.
Conclusions
This work revealed that hypoxia, or reduced oxygen tension,
can differently impact mouse ES cells depending on the
specific signaling pathway that is being used for maintenance
85Metabolism and Lineage Commitment of Embryonic Stem Cells
of cell pluripotency. When LIF/STAT3 signaling was responsible
for maintaining ES cells self-renewing, hypoxia had a
negative effect on cell proliferation and viability. Furthermore,
it exerted a proactive effect on differentiation, and
cells showed signs of early commitment. On the contrary,
low oxygen levels did not influence the final cell numbers of
mouse ES cells cultured under ground-state conditions.
These cells are maintained in culture by inhibition of
FGF4/ERK and GSK3 signaling and, despite affecting cell
viability, hypoxia did not induce cell differentiation.
Nevertheless, during neural commitment, low oxygen tension
exerted a positive effect on early differentiation of
ground-state ES cells. Thus, it is possible to conclude that,
depending on the specific signaling pathways explored,
oxygen levels may be a potentially important parameter to
consider during the development of strategies for ES cell
expansion and controlled differentiation.
Materials and methods
Cell Lines
In the present studies, the E14tg2a and 46C mouse ES cell
lines were used as model systems. The 46C cell line
(established at the laboratory of Professor Austin Smith,
Wellcome Trust Centre for Stem Cell Research, University of
Cambridge, England, UK) contains a green fluorescent
protein (GFP) knock-in at the Sox1 locus (neuroepithelial
marker gene) and can be induced to undergo neural
commitment under serum-free conditions (Diogo et al.,
2008; Ying et al., 2003b). E14tg2a and 46C mouse ES cells
were kept cryopreserved in liquid nitrogen until further use.
Determination of dissolved oxygen levels in liquid
medium
To examine the response of dissolved oxygen levels in the
liquid medium under simulated culture conditions, a 24channel
SDR SensorDish Reader (PreSens) for on-line monitoring
was used to measure O2 tension at the bottom of a
liquid layer in multiwell culture plates with integrated
optical-chemical oxygen sensors. The liquid was initially
equilibrated at atmospheric oxygen levels, and the plate was
then transferred to a humidified incubator either at 20 or 2%
O2 gas phase. The dissolved O2 levels were measured as a
function of time.
Expansion of mouse ES cells
On thawing, mouse ES cells were expanded on gelatinized
tissue culture plates (0.1% (v/v) gelatin in phosphatebuffered
saline (PBS; Gibco), diluted from a 2% stock
solution; Sigma) using Knockout Dulbecco's modified Eagle's
medium (DMEM; Gibco) supplemented with 15% (v/v)
Knockout serum-replacement (SR; according to the manufacturer's
instructions, Gibco), 2 mM glutamine (Gibco), 1%
(v/v) penicillin (50 U/mL)/streptomycin (50 μg/mL) (Pen/
Strep; Gibco), 1% (v/v) nonessential amino acids (Sigma),
and 1 mM β-mercaptoethanol (β-ME; Sigma); Knockout
serum-free medium (KO-DMEM/SR) was supplemented with
0.1% (v/v) LIF (human LIF produced in house by HEK-293
cells). In selected experiments, the GSK3-specific inhibitor
CHIR99021 (Stemgent) was also added to the culture medium
at a final concentration of 5 µM. Alternatively, the cells were
expanded in iSTEM serum-free medium (Stem Cell Sciences
Inc.) for culture of authentic, ground-state mouse ES cells.
This culture medium is supplemented with three chemical
inhibitors that eliminate differentiation signals, sustaining
mouse ES cell self-renewal and maintaining the pure ES cell
ground state, as previously described (Ying et al., 2008).
The cells were cultured either under atmospheric oxygen
tension (20% O2) or reduced oxygen atmosphere (hypoxic
chamber connected to a Proox Model 21 controller set up at
2% O2; BioSpherix) at 37 °C in a 5% CO2 humidified incubator.
Cells were dissociated every 2 days using Accutase solution
(Sigma) and replated at the same initial cell density (20 000
cells/cm2
). At each passage, viable and dead cell numbers
were determined by counting in a hemocytometer under an
optical microscope using the trypan blue dye exclusion test
(Gibco). Assuming exponential growth, the apparent specific
growth rate, µ (day-1
), was calculated as µ=ln(Xf/Xi)/Δt,
where Xf and Xi are the viable cell numbers at the beginning
and at the end of any given time interval (Δt). The fold
increase in total cell number was defined as the ratio
between final and initial viable cell numbers for each
passage. The cumulative fold increase was calculated as
the product of fold increase values obtained for five
consecutive passages.
Analysis of ground state, pluripotent stem cell
expansion at clonal densities
Cells were seeded at low densities (400 cells/well) on
gelatinized 24-well tissue culture plates and cultured for
6 days in iSTEM serum-free medium either in 20% O2 or 2% O2
atmosphere. Medium was changed at Days 2 and 4. After
6 days, the colonies were counted and the diameter (in µm)
of the colonies was measured. The efficiency in colony
formation was defined as the ratio between the number of
colonies formed and the number of cells seeded on a well
(400 cells).
Proliferation analysis using PKH67 fluorescent dye
For cell division analysis using PKH67, cells were collected
and stained with the dye according to the manufacturer's
instructions (PKH67 green fluorescent cell linker kit for cell
membrane labeling; Sigma). Since they are stably incorporated
in the cell membrane, PKH67 molecules are equally
distributed between daughter cells during cell division, and
therefore each generation of cells is half as fluorescent as
the previous one, allowing the determination of the number
of cell divisions at any given time point in culture (da Silva
et al., 2009). Between 100 000 and 1 000 000 cells were used
for flow cytometry analysis (Day 0), and the remaining cells
were plated with appropriate culture medium as described
before. After 2 days in culture, cells were collected and the
samples used for flow cytometry (FACSCalibur flow cytometer;
Becton Dickinson). Proliferation analysis was performed
using the CellQuest software and/or the Proliferation
Wizard of ModFit software.
86 T.G. Fernandes et al.
Cell cycle analysis using propidium iodine (PI) nucleic
acid staining
Cells were collected and fixed in 2% (w/v) paraformaldehyde
(Sigma) solution in PBS at room temperature for 20 min.
After washing, cells were then incubated with 25 μg/mL
RNase A (Roche) and stained with 5 μg/mL PI (BD) for 30 min
at 37 °C in the dark. The acquisition of the samples was
performed in a FACSCalibur flow cytometer and the data
were analyzed using the ModFit software.
Neural commitment of mouse ES cells
After expansion, mouse ES cells were dissociated with
Accutase, plated on gelatin-coated tissue culture plates,
and incubated up to 8 days in RHB-A medium (Stem Cell
Sciences Inc.), as described elsewhere (Diogo et al., 2008).
Again, cells were maintained under atmospheric oxygen
tension (20% O2) or reduced oxygen atmosphere (2% O2)
during the neural commitment protocol. To quantify the
population of Sox1-GFP+
neural progenitors in each day, cells
were recovered, counted, and analyzed by flow cytometry
(Ying et al., 2003b). Settings were determined at the start of
the experiment using undifferentiated 46C mouse ES cells as
negative control.
Sox1+
neural progenitors are nevertheless a transient
population. Hence, to evaluate the influence of low oxygen
tension on further neuronal differentiation, cells were
collected at Day 6 of neural commitment and replated onto
tissue culture plates precoated with polyornithine and
laminin (30- to 60-min treatment with 0.01% (v/v) polyornithine
solution in water, washing in PBS, and incubation
for 2 h with 10 µg/mL laminin solution (both from Sigma)
(Pollard et al., 2006)). Cells were maintained in culture in
RHB-A medium for a period of time lower than 9 days with
medium change every 3 days. Cells were analyzed for class III
β-tubulin (TujI) expression at specific time points using flow
cytometry.
Intracellular immunofluorescence staining for flow
cytometry analysis
Cells were collected and fixed in 2% (w/v) paraformaldehyde
(Sigma) solution in PBS. Cells were then permeabilized with
1% (w/v) saponin (Sigma) solution in PBS, followed by
incubation in blocking solution (3% (v/v) normal goat serum
(NGS; Sigma) in PBS) for 15 min. The primary antibody was
added to the samples, and cells were incubated for 2 h at
room temperature. Oct-4 and Nanog-expressing cells were
evaluated by flow cytometry after intracellular staining
using mouse monoclonal anti-Oct-3/4 (Chemicon), 1:30 (v/v)
dilution in blocking solution, or rabbit polyclonal anti-Nanog
(Santa Cruz Biotechnology), 1:500 (v/v) dilution in blocking
solution, respectively. Neuronal differentiation efficiency
was also assessed by flow cytometry with Alexa Fluor 488labeled
Neuronal Class III β-tubulin (Tuj1) (1:200 (v/v)
dilution; Covance).
For Oct-4 and Nanog quantification, a secondary antibody
(Alexa Fluor 488-conjugated goat-anti mouse and goat-anti
rabbit IgG, respectively (Molecular Probes), 1:1000 dilution
in blocking solution) was added to the cells after washing,
followed by incubation for 45 min at room temperature.
Cells incubated with appropriate secondary antibodies were
used as negative controls. After washing, the acquisition of
the samples was performed in a FACSCalibur flow cytometer
and analyzed using the CellQuest software. Cell debris and
dead cells were excluded from the analysis based on
electronic gates using forward scatter (cell size) and side
scatter (cell complexity) criteria. A minimum of 10 000
events was collected for each sample.
Real-time polymerase chain reaction
Quantitative real-time PCR was performed using previously
established protocols (Abranches et al., 2006). Briefly, total
RNA was isolated from N106
cells using the High Pure RNA
Isolation Kit (Roche Diagnostics), and DNAse I (Roche)
treatment according to the manufacturer's instructions.
First-strand cDNA was synthesized using the Transcriptor
First-Strand cDNA Synthesis Kit, also from Roche, and
anchored-oligo(dt)18. Real-time PCR was then performed
on a Roche Light Cycler detection system, in a 20 µl final
volume containing 2.0 mM MgCl2 solution, 0.2 µM of each
primer, 4.0 µl of sample, and SYBR Green I mixture.
Threshold cycle (CT) values were calculated using the
LightCycler software 3.4 (Roche) and the Fit Points method.
Gapdh was used as control housekeeping gene in real-time
PCR experiments. Relative quantification of gene expression
was performed using the delta-delta threshold cycle method
(ΔΔCT). The primer sequences used to assess gene expression
can be seen in Table S1 (Supplementary Data).
Metabolite analysis
Metabolite analysis was performed as previously described
(Fernandes et al., 2010). Briefly, the exhausted medium
was recovered after cell culture and samples were
centrifuged for 10 min at 1500 rpm to remove dead cells
and debris. The supernatants were then collected and kept
frozen until analysis. Fresh medium was also analyzed and
the results were used for the determination of consumption
and production rates. Glucose, lactate, glutamine, and
ammonia concentrations were measured using a multiparameter
analyzer (YSI 7100MBS; Yellow Springs Instruments).
The specific metabolic rates (qMet, pmol/(cell.
day)) were calculated for every time interval using the
following equation: qMet =ΔMet/(Δt.ΔXv), where ΔMet is the
variation in metabolite concentration during the time
period Δt and ΔXv the logarithmic mean of viable cell
number during the same period. The apparent lactate from
glucose yield (Y′lactate/glucose) was calculated as the ratio
between qlactate and qglucose, and the apparent ammonia
from glutamine yield (Y′ammonia/glutamine) was also determined
as the ratio between qammonia and qglutamine. Cell
yields from glucose and glutamine were also calculated as
the ratio between µ and qglucose or qglutamine, respectively.
Statistical analysis
Statistical analysis was performed using the SPSS 16.0
software (SPSS Inc., Chicago, IL, http://www.spss.com.uk).
87Metabolism and Lineage Commitment of Embryonic Stem Cells
The nonparametric Wilcoxon–Mann–Whitney test was used
to evaluate statistical significance of two independent
samples (Pb0.05). Unless otherwise stated, error bars
represent the standard error of the mean (SEM) from
triplicates for each condition.
Acknowledgments
T.G.F. and A.M.P acknowledge support from Fundação para a
Ciência e a Tecnologia, Portugal (SFRH/BD/24365/2005 and
SFRH/BD/36070/2007, respectively). This work was financially
supported by Fundação para a Ciência e a Tecnologia,
through the MIT-Portugal Program, Bioengineering Systems
Focus Area. The authors gratefully acknowledge Domingos
Henrique (Institute of Molecular Medicine, Lisboa, Portugal)
for providing the E14tg2a and the 46C mouse embryonic stem
cell lines, and HEK-293 cells for production of human LIF.
Appendix A. Supplementary data
Supplementary data for this article may be found in the
online version at doi:10.1016/j.scr.2010.04.003.
References
Abranches, E., O'Neill, A., Robertson, M.J., Schaffer, D.V., Cabral,
J.M.S., 2006. Development of quantitative PCR methods to
analyse neural progenitor cell culture state. Biotechnol. Appl.
Biochem. 44, 1–8.
Burdon, T., Smith, A., Savatier, P., 2002. Signalling, cell cycle and
pluripotency in embryonic stem cells. Trends Cell Biol. 12, 432–438.
Covello, K.L., Kehler, J., Yu, H., Gordan, J.D., Arsham, A.M., Hu, C.J.,
Labosky, P.A., Simon, M.C., Keith, B., 2006. HIF-2alpha regulates
Oct-4: effects of hypoxia on stem cell function, embryonic
development, and tumor growth. Genes Dev. 20, 557–570.
da Silva, C.L., Gonçalves, R., Porada, C.D., Ascensão, J.L., Zanjani,
E.D., Cabral, J.M.S., Almeida-Porada, G., 2009. Differences amid
bone marrow and cord blood hematopoietic stem/progenitor cell
division kinetics. J. Cell. Physiol. 220, 102–111.
Diogo, M.M., Henrique, D., Cabral, J.M.S., 2008. Optimization and
neural commitment of mouse embryonic stem cells. Biotechnol.
Appl. Biochem. 49, 105–112.
Doble, B.W., Woodgett, J.R., 2003. GSK-3: tricks of the trade for a
multi-tasking kinase. J. Cell Sci. 116, 1175–1186.
Duval, D., Malaisé, M., Reinhardt, B., Kedinger, C., Boeuf, H., 2004.
A p38 inhibitor allows to dissociate differentiation and apoptotic
processes triggered upon LIF withdrawal in mouse embryonic
stem cells. Cell Death Differ. 11, 331–341.
Fernandes, T.G., Fernandes-Platzgummer, A.M., da Silva, C.L.,
Diogo, M.M., Cabral, J.M.S., 2010. Kinetic and metabolic analysis
of mouse embryonic stem cell expansion under serum-free
conditions. Biotechnol. Lett. 32, 171–179.
Follmar, K.E., Decroos, F.C., Prichard, H.L., Wang, H.T.,
Erdmann, D., Olbrich, K.C., 2006. Effects of glutamine,
glucose, and oxygen concentration on the metabolism and
proliferation of rabbit adipose-derived stem cells. Tissue Eng.
12, 3525–3533.
Hanna, J., Markoulaki, S., Mitalipova, M., Cheng, A.W., Cassady, J.
P., Staerk, J., Carey, B.W., Lengner, C.J., Foreman, R., Love, J.,
et al., 2009. Metastable pluripotent states in NOD-mouse-derived
ESCs. Cell Stem Cell 4, 513–524.
Ivanovic, Z., 2009. Hypoxia or in situ normoxia: the stem cell
paradigm. J. Cell. Physiol. 219, 271–275.
Jauniaux, E., Gulbis, B., Burton, G.J., 2003. Physiological implications
of the materno-fetal oxygen gradient in human early pregnancy.
Reprod. Biomed. Online 7, 250–253.
Jeong, C.H., Lee, H.J., Cha, J.H., Kim, J.H., Kim, K.R., Kim, J.H.,
Yoon, D.K., Kim, K.W., 2007. Hypoxia-inducible factor-1 alpha
inhibits self-renewal of mouse embryonic stem cells in vitro via
negative regulation of the leukemia inhibitory factor-STAT3
pathway. J. Biol. Chem. 282, 13672–13679.
Kirouac, D.C., Zandstra, P.W., 2008. The systematic production of
cells for cell therapies. Cell Stem Cell 3, 369–381.
Klimanskaya, I., Rosenthal, N., Lanza, R., 2008. Derive and conquer:
sourcing and differentiating stem cells for therapeutic applications.
Nat. Rev. Drug Discov. 7, 131–142.
Krishnamoorthy, M., Heimburg-Molinaro, J., Bargo, A.M., Nash, R.J.,
Nash, R.J., 2009. Heparin binding epidermal growth factor-like
growth factor and PD169316 prevent apoptosis in mouse
embryonic stem cells. J. Biochem. 145, 177–184.
Kurosawa, H., Kimura, M., Noda, T., Amano, Y., 2006. Effect of
oxygen on in vitro differentiation of mouse embryonic stem cells.
J. Biosci. Bioeng. 101, 26–30.
Lee, Y.M., Jeong, C.H., Koo, S.Y., Son, M.J., Song, H.S., Bae, S.K.,
Raleigh, J.A., Chung, H.Y., Yoo, M.A., Kim, K.W., 2001.
Determination of hypoxic region by hypoxia marker in developing
mouse embryos in vivo: a possible signal for vessel development.
Dev. Dyn. 220, 175–186.
Morrison, S.J., Csete, M., Groves, A.K., Melega, W., Wold, B.,
Anderson, D.J., 2000. Culture in reduced levels of oxygen
promotes clonogenic sympathoadrenal differentiation by isolated
neural crest stem cells. J. Neurosci. 20, 7370–7376.
Mottet, D., Dumont, V., Deccache, Y., Demazy, C., Ninane, N.,
Raes, M., Michiels, C., 2003. Regulation of hypoxia-inducible
factor-1alpha protein level during hypoxic conditions by the
phosphatidylinositol 3-kinase/Akt/glycogen synthase kinase
3beta pathway in HepG2 cells. J. Biol. Chem. 278, 31277–31285.
Nichols, J., Smith, A., 2009. Naive and primed pluripotent states.
Cell Stem Cell 4, 487–492.
Niwa, H., Burdon, T., Chambers, I., Smith, A., 1998. Self-renewal of
pluripotent embryonic stem cells is mediated via activation of
STAT3. Genes Dev. 12, 2048–2060.
Ozturk, S.S., Riley, M.R., Palsson, B.O., 1992. Effects of ammonia
and lactate on hybridoma growth, metabolism, and antibody
production. Biotechnol. Bioeng. 39, 418–431.
Pistollato, F., Chen, H.L., Schwartz, P.H., Basso, G., Panchision, D.M.,
2007. Oxygen tension controls the expansion of human CNS
precursors and the generation of astrocytes and oligodendrocytes.
Mol. Cell. Neurosci. 35, 424–435.
Pollard, S.M., Conti, L., Sun, Y., Goffredo, D., Smith, A., 2006.
Adherent neural stem (NS) cells from fetal and adult forebrain.
Cereb. Cortex 16 (Suppl. 1), i112–i120.
Pouton, C.W., Haynes, J.M., 2005. Pharmaceutical applications of
embryonic stem cells. Adv. Drug Deliv. Rev. 57, 1918–1934.
Powers, D.E., Millman, J.R., Huang, R.B., Colton, C.K., 2008. Effects of
oxygen on mouse embryonic stem cell growth, phenotype retention,
and cellular energetics. Biotechnol. Bioeng. 101, 241–254.
Reitzer, L.J., Wice, B.M., Kennel, D., 1979. Evidence that
glutamine, not sugar, is the major energy source for cultured
HeLa cells. J. Biol. Chem. 254, 2669–2676.
Rodesch, F., Simon, P., Donner, C., Jauniaux, E., 1992. Oxygen
measurements in endometrial and trophoblastic tissues during
early pregnancy. Obstet. Gynecol. 80, 283–285.
Sato, N., Meijer, L., Skaltsounis, L., Greengard, P., Brivanlou, A.H.,
2004. Maintenance of pluripotency in human and mouse embryonic
stem cells through activation of Wnt signaling by a pharmacological
GSK-3-specific inhibitor. Nat. Med. 10, 55–63.
Schop, D., Janssen, F.W., van Rijn, L.D., Fernandes, H., Bloem,
R.M., de Bruijn, J.D., van Dijkhuizen-Radersma, R., 2009.
Growth, metabolism, and growth inhibitors of mesenchymal
stem cells. Tissue Eng. A 15, 1877–1886.
88 T.G. Fernandes et al.
Silva, J., Smith, A., 2008. Capturing pluripotency. Cell 132, 532–536.
Smith, A.G., 2001. Embryo-derived stem cells: of mice and men.
Annu. Rev. Cell Dev. Biol. 17, 435–462.
Sutter, D.M., Krause, K.-H., 2008. Neural commitment of embryonic
stem cells: molecules, pathways and potential for cell therapy. J.
Pathol. 215, 355–368.
Vriezen, N., Romein, B., Luyben, K.C., van Dijken, J.P., 1997. Effects
of glutamine supply on growth and metabolism of mammalian
cells in chemostat culture. Biotechnol. Bioeng. 54, 272–286.
Watanabe, K., Ueno, M., Kamiya, D., Nishiyama, A., Matsumura, M.,
Wataya, T., Takahashi, J.B., Nishikawa, S., Nishikawa, S.,
Muguruma, K., et al., 2007. A ROCK inhibitor permits survival
of dissociated human embryonic stem cells. Nat. Biotechnol. 25,
681–686.
Ying, Q.-L., Nichols, J., Chambers, I., Smith, A., 2003a. BMP
induction of Id proteins suppresses differentiation and
sustains embryonic stem cell self-renewal in collaboration
with STAT3. Cell 115, 281–292.
Ying, Q.-L., Stavridis, M., Griffiths, D., Li, M., Smith, A., 2003b.
Conversion of embryonic stem cells into neuroectodermal precursors
in adherent monoculture. Nat. Biotechnol. 21, 183–186.
Ying, Q.-L., Wray, J., Nichols, J., Batlle-Morera, L., Doble, B.,
Woodgett, J., Cohen, P., Smith, A., 2008. The ground state of
embryonic stem cell self-renewal. Nature 453, 519–523.
89Metabolism and Lineage Commitment of Embryonic Stem Cells