Mild Hypoxia Enhances Proliferation and Multipotency of
Human Neural Stem Cells
Guido Santilli1, Giuseppe Lamorte1, Luigi Carlessi2, Daniela Ferrari1, Laura Rota Nodari1, Elena Binda1,
Domenico Delia2, Angelo L. Vescovi1*, Lidia De Filippis1*
1Department of Biotechnology and Biosciences, University of Milano-Bicocca, Milan, Italy, 2Fondazione IRCCS Istituto Nazionale Tumori, Department of Experimental
Oncology, Milan, Italy
Background: Neural stem cells (NSCs) represent an optimal tool for studies and therapy of neurodegenerative diseases. We
recently established a v-myc immortalized human NSC (IhNSC) line, which retains stem properties comparable to parental
cells. Oxygen concentration is one of the most crucial environmental conditions for cell proliferation and differentiation
both in vitro and in vivo. In the central nervous system, physiological concentrations of oxygen range from 0.55 to 8%
oxygen. In particular, in the in the subventricular zone niche area, it’s estimated to be 2.5 to 3%.
Methodology/Principal Findings: We investigated in vitro the effects of 1, 2.5, 5, and 20% oxygen concentrations on IhNSCs
both during proliferation and differentiation. The highest proliferation rate, evaluated through neurosphere formation
assay, was obtained at 2.5 and 5% oxygen, while 1% oxygen was most noxious for cell survival. The differentiation assays
showed that the percentages of b-tubIII+ or MAP2+ neuronal cells and of GalC+ oligodendrocytes were significantly higher
at 2.5% compared with 1, 5, or 20% oxygen at 17 days in vitro. Mild hypoxia (2.5 to 5% oxygen) promoted differentiation
into neuro-oligodendroglial progenitors as revealed by the higher percentage of MAP2+/Ki67+ and GalC+/Ki67+ residual
proliferating progenitors, and enhanced the yield of GABAergic and slightly of glutamatergic neurons compared to 1% and
20% oxygen where a significant percentage of GFAP+/nestin+ cells were still present at 17 days of differentiation.
Conclusions/Significance: These findings raise the possibility that reduced oxygen levels occurring in neuronal disorders
like cerebral ischemia transiently lead to NSC remaining in a state of quiescence. Conversely, mild hypoxia favors NSC
proliferation and neuronal and oligodendroglial differentiation, thus providing an important advance and a useful tool for
NSC-mediated therapy of ischemic stroke and neurodegenerative diseases like Parkinson’s disease, multiple sclerosis, and
Citation: Santilli G, Lamorte G, Carlessi L, Ferrari D, Rota Nodari L, et al. (2010) Mild Hypoxia Enhances Proliferation and Multipotency of Human Neural Stem
Cells. PLoS ONE 5(1): e8575. doi:10.1371/journal.pone.0008575
Editor: Joseph Najbauer, City of Hope National Medical Center, United States of America
Received September 3, 2009; Accepted December 3, 2009; Published January 5, 2010
Copyright: ? 2010 Santilli et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was financially supported by grants from Cariplo (Cassa di Risparmio delle Provincie Lombarde) Foundation, Neurothon ONLUS
(Organizzazione non lucrativa di utilita ` sociale) Foundation (a nonprofit organization), the Italian Association for Cancer Research, and Stemgen Spa. The funders
had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com (ALV); firstname.lastname@example.org (LDF)
Cultured CNS stem cells are endowed with capacity to self-
renew and differentiate into neurons, astrocytes and oligodendro-
cytes in predictable proportions [1–5]. Thus they have provided a
useful tool to elucidate the pathways leading to generation of
neurons and glia and to study the effects of different extrinsic
factors on the commitment of neural stem cells (NSC) to form such
cell lineages . For these reasons the discovery, isolation and
characterization of multipotent NSC from various locations within
the mammalian brain represents a major recent advancement in
Inevitably, NSCs have become a hot topic of investigation in
translational research for common degenerative diseases. In fact,
an important goal is to accomplish neuroregeneration by
transplantation of exogenous cells that is by cell-mediated therapy.
In clinical settings, gases are appreciated as primary variables in
organ survival, with O2as the critical gas parameter. Indeed,
oxygen plays an essential role in the maintenance of NSC viability
as it is responsible for aerobic metabolism to maintain intracellular
energy balance. Hypoxia, as a state of reduced O2tension below
critical values, triggers intricate and complex mechanisms to
restore O2homeostasis at the cellular, tissue and organism level
and it occurs under physiological as well as pathological
conditions. Markedly severe hypoxia (less than 0.002% O2) is
caused pathologically by stroke, ischemia and increase in solid
tumor size [8,9].
Cerebral ischemia is known to cause acute and delayed
neuronal death through the activation of a complex series of
events leading to severe brain dysfunction both in rodents and
humans [10,11]. Recent studies have shown that both global and
focal ischemia induce increased proliferation and neural differen-
tiation of NSCs residing in the subgranular zone (SGZ) of the
dentate gyrus (DG), the anterior subventricular zone (SVZ) and
PLoS ONE | www.plosone.org1 January 2010 | Volume 5 | Issue 1 | e8575
the posterior periventricular zone adjacent to the hippocampus
[12,13]. A parallel increase of the migration of NSCs along the
neurogenic pathways was also observed , but the mechanisms
involved are still unknown. Hypoxia is among the main factors
causing ischemia-derived injuries. The physiological concentration
of oxygen in the central nervous system (CNS) ranges from as low
as 0.55% in the midbrain to 8% in the pia . In particular, 3–
5% oxygen enhances the proliferation of cultured NSCs and
modulates their differentiation into neurons [16,17].
We have established an immortal human NSC line (IhNSC)
cultured at 5% oxygen that retains normal hNSC features such as
proliferation, self-renewal and multipotency . In particular,
IhNSC can generate fully functional neuronal cells, thus providing
a useful model to study NSC for the therapy of neurodegenerative
diseases or brain injuries like stroke and ischemia, without the
limitations of primary fetal tissue. We recently transplantated
IhNSC-derived progenitors (IhNSC-P) near the hippocampal CA1
layer of adult rats injured by global transient ischemia to evaluate
the integration and maturation of hNSC in a context mimicking
the chronic impairment of neurological function following
hypoxia-induced injuries, and documented their ability to engraft
efficiently, to the point of establishing synaptic contacts with the
host cells (Rota Nodari et al, submitted).
Considering these findings, we have examined the effects of
different oxygen concentrations (1%, 2.5%, 5% and 20%
atmospheric oxygen) on the proliferation, differentiation and
death of IhNSC in order to identify the optimal culture conditions
of non-immortalized human NSCs for NSC-mediated therapy of
CNS injuries characterized by severe hypoxia-associated cell death
that occurs in stroke and ischemia.
Materials and Methods
Generation and Expansion of the IhNSC Line
PK-VM-2 is a replication-defective, infective retroviral vector
described previously [1,19–21] coding for both avian myc (p110
gag-myc or v-myc) driven by the MoLV-long terminal repeat
(LTR) promoter and aminoglycoside transferase (conferring
resistance to neomycin; neor) driven by the SV40 promoter.
The amphotropic retroviral particles were packaged in GP envAM
cells cultured in hNSC medium . The hNSC cultures (parental
cells) used in this study were isolated and propagated from the
diencephalic and telencephalic brain regions of a caucasian
human fetus at 10.5 weeks gestational age and were previously
described by Vescovi et al. . Retroviral transduction with v-myc
was carried as described in Villa et al.  on parental hNSCs that
had undergone 22 passages in vitro. Following G418 selection,
aliquots of these bulk cell lines were cryopreserved in complete
medium containing 10% dimethyl sulfoxide.
Generation of Growth Curves
hNSC lines 1 and 2 were established by fetal human brain as
described ; at passage 14 and 16 respectively, they were shifted
to appropriate oxygen culture conditions. IhNSC were cultured as
hNSC in 5% oxygen and at passage 45 they were split to 1%,
2.5%, 5% and 20% oxygen. In order to allow the adaptation to
the new oxygen concentrations, hNSC and IhNSC were cultured
for 4–5 passages after the shift before use for the experiments.
Thereafter, hNSC and IhNSC were continuously maintained in
the respective oxygen concentrations. hNSC and IhNSC were
cultured in four different humidified CO2 multigas incubators
(Binder), flushed continuously with a N2 gas to maintain
established atmospheric O2concentrations (1%, 2.5%, 5% and
20%) at a constant temperature of 37uC.
The rate of expansion of IhNSC and hNSC was obtained by
plating 16104cells/cm2in growth medium containing FGF2 and
EGF (Peprotech, Rocky Hill, NJ) and grown in the described
oxygen conditions. At each passage (p), NSC-originated neuro-
spheres were dissociated and the logarithmic value of the total
viable-cell number was plotted against the day in vitro (days) since
the beginning of the experiment. For each oxygen condition, the
growth curves were performed in duplicate and generated
Differentiation of IhNSC
To induce IhNSC differentiation, individual spheres were
mechanically dissociated and transferred onto laminin (Roche,
Basel, Switzerland) coated glass coverslips at a density of 16104
cells per cm2in the presence of FGF2 (20 ng/mL). At 3 days,
FGF2 medium was replaced with control medium and IhNSC
differentiated for additional 7 days (10 days) or 14 days (17
Differentiating cells were fixed in freshly buffered 4%
paraformaldehyde at 3, 10 or 17 days along differentiation (see
‘‘Differentiation of IhNSC’’). After blocking with 10% normal goat
serum and treatment with Triton X-100, 0,3% v/v for
intracellular antigen detection, cultures were incubated overnight
at 4uC in the following antibodies: b-tubulin III (b-tubIII, 1:400,
Covance), gamma amino-butyric acid (GABA, 1:500, Sigma), glial
fibrillary acidic protein (GFAP, mouse, 1:500,Chemicon), GFAP
(rabbit, 1:500, Dako), glutamate (1:500, Sigma), microtubular
associated protein type 2 (MAP2, 1:200, Sigma), Ki67 (1:1.000,
Novocastra), galactocerebroside C (GalC, 1:300, Chemicon),
gestin (1:200, Chemicon), vimentin (1:400, Immunological Sci-
After rinsing in phosphate-buffered saline (PBS), cultures were
incubated for 45 minutes at room temperature in the following
secondary antibodies: Cy2 (against mouse or rabbit IgG, 1:200,
Jackson), Cy3 (against mouse or rabbit IgG, 1:800, Jackson), Alexa
546 (against rabbit IgG or mouse IgG1, 1:800, Molecular Probes),
Alexa 488 (against rabbit IgG or mouse IgG1, 1:800, Molecular
Probes), Cy3 (against goat IgG, 1:1000, Jackson). Nuclei were
stained with DAPI. Data are reported as percentages of labeled
cells over the total number of DAPI-labeled nuclei (at least 1500
nuclei per coverslip were scored)+/2 the mean standard error
(SE). Each value represents the average of three independent
experiments unless specified.
Microphotographs were taken using a Zeiss Axiovert 200 direct
epifluorescence microscope (Axioplan 2, Carl Zeiss, Jena,
Germany), or with confocal microscope Leica TCS SP2 DMIRE2.
For mithocondria stain with JC1, IhNSC were plated onto
laminin and after 4 hours cells were stained with JC1 (0.5 mM,
Molecular Probes, Invitrogen) and Hoechst (50 mM, Sigma)
without fixing with PFA. Microphotographs were taken using a
Data are reported as percentages of labeled cells over the total
number of nuclei6SEM. Each value represents the average of
three independent experiments unless indicated into the legend.
Western Blot Analysis
Immunoblots were performed as described  on total cell
extracts prepared in Laemmli buffer (0.125 M Tris-HCl, pH 6.8,
5% SDS) containing 1mM phenylmethylsulfonyl fluoride (PMSF),
10 mg/mL pepstatin, 100 KIU/mL aprotinin, 10 mg/mL leupep-
tin (all from Calbiochem, San Diego, Calif.) and 1 mM Na3VO4.
Extracts (50 mg protein/lane) plus 5% b-mercaptoethanol were
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org2 January 2010 | Volume 5 | Issue 1 | e8575
electrophorezed using SDS-PAGE and electroblotted onto PVDF
membranes (Millipore, Bedford, MA). Membranes were blocked
with 4% non-fat dried milk or with BSA, incubated with
monoclonal antibodies for p53 (Clone DO-7), nestin (Chemicon),
b-actin and GAPDH (Sigma), and rabbit antibodies anti cleaved
PARP, cleaved caspases 3 and cleaved caspases 9 (all from Cell
Signaling, Boston, MA). PVDF membranes were incubated with
antibodies in sealed bags using the X-blot roller hybridization
instrument (Isenet, Milan, Italy). Binding of antibodies was
detected with ECL Super Signal (Pierce, Rockford, IL) and bands
quantified with ImageQuant.
Cell Cycle Studies: BrdU/DNA Analysis and Detection
At 15–20 passages from the shift to the specific oxygen
condition (1%, 2.5%, 5%, 20% O2), IhNSC were dissociated
and kept in growing conditions for 24, 48 or 72 hours. Then
cells were incubated with 20 mM 5-bromo-2-deoxyuridine
(BrdU) for 20 min at 37uC, fixed in 70% ethanol and kept at
4uC before staining . Fixed cells were washed with cold
PBS and the DNA was denaturated with 1mL of 2N HCl for
20 min at RT. DNA denaturation was stopped by adding
3 mL 0.1 M sodium tetraborate pH 8.5. After centrifugation,
the pellet was incubated with 1mL 0.5% (v/v) Tween-20
(Sigma, MO) in PBS containing 1% of BSA (Sigma) for 15 min
Cells were then incubated with anti-BrdU monoclonal
antibody (BD Pharmingen, San Diego, CA) diluted 1:10 in
0.5% (v/v) Tween-20 in PBS containing 1% of BSA for 60 min at
RT in the dark. After centrifugation the pellet was incubated with
1mL of 0.5% (v/v) tween-20 in PBS 1% of BSA for 15 min at RT
and then with Alexa 488 conjugated F(ab9)2 fragment goat anti-
mouse IgG (Invitrogen, Carlsbad, CA), 1:500 dilution in 0.5% (v/
v) Tween-20 in PBS containing 1% of BSA) for 60 min at RT in
the dark. The cells were then resuspended in 1 mL of a solution
containing 2.5 mg/mL of propidium iodide (PI) in PBS and 7 mL
RNAse 3 mg/mL in water, and stained overnight at 4uC in the
dark. Biparametric BrdU/DNA analysis was done on at least
30000 cells for each sample by the FACSCalibur (BD
Biosciences) and the data were analyzed using Summit 4.3
Changes in Mitochondrial Membrane Potential
Mitochondrial membrane potential was evaluated with the
lipophilic cationic probe JC-1 (Invitrogen), which changes
reversibly its color from green to orange as the membrane
potential increases . This property is due to the reversible
formation of JC-1 aggregates upon membrane polarization that
causes shifts in emitted light from 530 nm (i.e., emission of JC-1
monomeric form) to 590 nm (i.e., emission of J-aggregate) when
excited at 490 nm. The apoptotic cells contain somewhat fewer
JC-1 aggregates and more JC-1 monomers and their colour shifts
from orange to green. Cells were stained with 0.5 mM JC-1 and
kept at 37uC for 20 min, washed and resuspended in a total
volume of 500 mL complete medium and analyzed by Cyan ADP
(Coulter, Brea, CA).
Terminal Deoxynucleotidyl Transferase-Mediated dUTP-
FITC Nick-End Labeling Assay
Apoptosis was measured using the TUNEL assay kit (Roche
Diagnostics, Basel, Switzerland) following the manufacturers’
instruction for dual parameter flow cytometry. Analysis was
performed on a FACSCalibur (BD Biosciences – CA USA) and
the data were analyzed using Summit 4.3 software (Coulter).
Statistical analysis has been performed through ANOVA. Data
are reported as means6SEM. Each value represents the average
of three independent experiments, unless otherwise indicated in
the legend. Data are considered statistically significant when
p,0.01 unless otherwise indicated.
Effects of Oxygen Concentrations on the Proliferation of
IhNSCs were initially established and grown as neurospheres in
5% atmospheric oxygen (O2) . For the present experiments,
the cells were split after 45 passages to 1%, 2.5%, 5% and 20% O2
and continuously cultured in the respective oxygen conditions. In
2.5% and 5% of oxygen, IhNSC proliferated similarly and more
rapidly than in 20% oxygen (Fig. 1A). Likewise, the proliferative
rate of the non-immortalized parental hNSCs was maximal in
mild hypoxia (2.5% and 5% oxygen) if compared to 20% oxygen,
with minor variability among different cell lines (Fig. 1B). The
hypoxic condition (1% O2) decreased the growth of IhNSCs and
particularly of hNSCs, which eventually arrested and died. To
determine whether the increased yield of precursors in mild
hypoxia was due to augumented proliferation or reduced cell
death or both, the DNA replication, detected by BrdU
incorporation, and cell viability were simultaneously measured at
different time points after dissociation. BrdU incorporation in cells
during expansion did not appreciably vary at 24, 48 and 72 h from
dissociation (Fig. 1C), and the fractions of cells in G1 and G2
phases were comparable among the different conditions (not
shown), altogether indicating that cell cycle is not altered by O2.
By contrast, the cell viability after dissociation was greater in mild
hypoxia with respect to severe hypoxia (1% O2), becoming
statistically significant after 72h from dissociation (Fig. 1D). Thus,
the increased growth of IhNSCs under mild hypoxia appears to
reflect a reduced apoptotic rate after dissociation, leading to the
enhancement of self-renewal capacity.
Survival of IhNSC at Different Concentrations of Oxygen
To determine to what extent in mild hypoxic conditions the rate
of survival after dissociation accounts for the increased growth rate
over passages, we performed TUNEL analysis at 4, 8, 12 and 24h
from dissociation. The highest percentage of apoptotic cells
(39.3365 at 8 h from dissociation) was found in 1% O2with
respect to the mild hypoxic (10.8760.95 at 2.5% and 18.1863.89
at 5% O2) and normoxic (19.262.85 at 20% O2) conditions
(Fig. 2A), consistent with the previous data from the neurosphere
and the viability assays (Fig. 1D).
Because a decrease of oxygen concentration below the
physiological threshold (1–5%) induces a metabolic shift from
oxidative phosphorylation towards anaerobic glycolysis with the
consequent diminution of mitochondria activity (revised in ),
we assessed the mitochondrial activity of IhNSC cultured at
different oxygen conditions and correlated it with the apoptotic
score. IhNSCs were stained in vivo with JC-1 cationic dye indicator
of mitochondrial membrane potential and analyzed by flow
cytometry at 4 and 24 h from dissociation, but no significant
differences were seen among the various oxygen conditions
(Fig. 2B). These data, which were also confirmed by the
fluorescence microscopy analysis of labeled samples (Fig. 2C,
panel C), indicate that severe hypoxia (1% oxygen) compromises
the survival but not the mitochondrial activity of IhNSCs. Since
hypoxia induces a shift from mitochondrial respiration to
anaerobic glycolysis, insufficient to fully support cell proliferation
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org3 January 2010 | Volume 5 | Issue 1 | e8575
and differentiation, we evaluated the expression of apoptotic
markers in IhNSC during differentiation by Western Blot analysis.
As shown in Fig. 3A, the signals relative to cleaved PARP and
cleaved caspases 3 and 9 were highest when neurospheres were
grown at 1% or 20% O2. Similarly, cleaved PARP, cleaved
Caspase 3 and 9 signals in IhNSCs undergoing differentiation
were much greater in 1% O2than in other oxygen concentrations
(Fig. 3A). Interestingly, p53 accumulation correlated with
increased levels of apoptosis. The frequency of pyknotic nuclei
increased progressively during differentiation in all oxygen
conditions (Fig. 3B, panel C), but the rate of cell death at 1%
O2(84.5611.5%) was dramatically higher than at 2.5% (461.4),
5% (18.263.4) or 20% (39.167.2) O2at 17 div (Fig. 3B). As
previously shown for IhNSC during cell culture, mild hypoxia
(2.5–5% O2) was associated with a modest level of apoptosis even
during differentiation. Since an efficient differentiation is associ-
ated with an increase in aerobic metabolism and number of
mitochondria , we determined the influence of low oxygen on
mitochondrial patterning during differentiation. Differentiated
IhNSC at 17 days were immunolabeled with anti-human
mitochondria antibody and analyzed by confocal microscopy.
Mitochondria aggregates were localized in the cell body and
processes in IhNSC when cultured in mild hypoxia or normoxia
(20% O2) with a ‘‘spaghetti-like’’ distribution (Fig. 3D), whereas
severe hypoxic conditions caused a dramatic reduction of
mitochondria per cell, with a perinuclear speckled pattern.
Interestingly, an evident collapse of mitochondria morphology
and location in the perinuclear area (Fig. 3D), a pattern typical of
cells undergoing apoptosis .
IhNSCs Generate Higher Percentages of Neurons in Mild
Upon removal of mitogens, IhNSCs spontaneously undergo
differentiation and concomitant proliferation arrest . Consid-
ering the effects of oxygen levels on proliferating IhNSCs
(described above), we analyzed the effects of oxygen on
differentiation. The proportion of neural cell lineages derived
from expanded precursors was determined by phenotypic analysis
with antibodies against b-tubIII and MAP2 (specific for early and
late differentiation stage of neuronal cells, respectively), GFAP
(astrocytes), GalC (oligodendrocyte precursors). At 10 days of
differentiation, the percentage of b-tubIII+ neurons derived from
IhNSCs accounted for 19.762.7% at 20%, 19.961.1% at 5% and
21.662.4% at 2.5% O2, but for only 10.160.6% at 1% O2
(Fig. 4A). At 17 days, the b- tubIII+ neurons were higher at 2.5%
(19.561.5%). On the other hand, the proportion of GFAP+
astrocytes was lower in 2.5% (4961.9%), 5% (57.265.3%) and
20% (56.367.4%) compared with 1% O2(72.364.1%) (Fig. 4B).
The b-tubIII and GFAP never co-localized (Fig. 4C), indicating
that the differentiation process of neuronal and astroglial cells was
To evaluate the expression of late neuronal proteins, differen-
tiated IhNSCs were immunostained with the dendritic marker
MAP2. At 10 days in vitro, 2.5% O2induced the highest percentage
of MAP2+ neurons (28.463.2%) relative to 20% (1362.1%), 5%
(17.661.5%) and 1% (21.561%) O2, correlating with the fractions
of early b-tubIII+ neuronal cells. A similar tendency was observed
at 17 days when the relative percentages of MAP2+ cells at 2.5%,
5% oxygen concentration
Figure 1. Mild hypoxia favors proliferation of IhNSC. (A, B) Graphs showing the proliferation rate of IhNSC (A) and hNSC lines 1 and 2 (B) when
cultured at different concentrations of O2. At each passage only 26105cells are plated while the logarithmic value of the total cell number is
calculated on the base of amplification rate at each passage (see material and methods) and plotted against the days in vitro from the beginning of
the experiment (N=2; one representative curve is shown). As shown in this figure, in mild hypoxia, it could be possible to generate more than 161013
(IhNSC) or 161010(hNSC) from 26105cells after 60 days from the first passage. (C) Histogram showing the percentage of IhNSC in S phase (obtained
through BrdU assay, see methods) at 24, 48, 72 hours after dissociation (N=3) at each oxygen culture condition. Values are means6S.E.M. No
significant differences were detected. (D) Histogram showing the number of viable IhNSC after dissociation at 0, 24, 48 and 72 hours after
dissociation. For each oxygen condition 200.000 cells were plated (N=2). Values are means6S.E.M. At 72 hours cells cultured in 2.5% and 5% O2the
percentage of viable IhNSCs was significantly higher than at 1% O2(p,0.001). The difference among all the values at the different oxygen
concentrations was statistically significant (P,0.01) unless indicated (*P,0.05, n.s.=not significant); one-way ANOVA followed by the Student’s t-test.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org4 January 2010 | Volume 5 | Issue 1 | e8575
5% and 20% O2were further increased, except for 1% O2which
determined a marked drop in MAP2+ cells (961.5%) (Fig. 5A and
IhNSCs generated GalC+ oligodendrocytes, with a greater
number of these cells in mild hypoxia than in 1% or 20% O2. In
particular, in 2.5% O2, they reached 13.861.3 already at 10 days
of differentiation and appeared to survive at 17 days (11.561.3),
while in 5% oligodendrocytes increased from 5.260.4 at 10 days
to 14.361.8 at 17 days (Fig. 5B and D). The remaining cells were
nestin positive (see below) or did not react with any of the markers
Mild Hypoxia Supports the Survival of Proliferating
Neuronal and Oligodendroglial Progenitors
IhNSCs undergo a gradual proliferative arrest during differen-
tiation . To assess to what extent residual proliferation of
differentiating IhNSCs is influenced by oxygen levels, we
estimated the fraction of IhNSC expressing the S-phase marker
Ki67. Consistent with the counts of apoptotic nuclei during
differentiation (Fig. 3B), albeit the percentage of Ki67+ cells
decreased with time, IhNSCs grown in 2.5% O2contained the
largest fraction of Ki67+ cells (65.663.5% at 10 days), consistently
higher than at other oxygen concentrations (Fig. 6A).
We further investigated if a selective prolonged survival of
neuronal and oligodendroglial progenitors could account for the
greater percentage of b-tubIII+ (or MAP2+) and GalC+ cells,
generated during differentiation at mild hypoxic conditions
(Figs. 4A, 5A and B). To this aim, differentiating cells we co-
labeled with Ki67 and MAP2 (Fig. 6B) or GalC (Fig. 6C). As
expected, in 2.5% oxygen IhNSC generated a higher percentage
of proliferating neuronal progenitors compared to 1%, 5% and
20% O2. Indeed, at 10 days 11.362% of the cells in 2.5% oxygen
was co-expressing Ki67 and MAP2, relative to 6.660.4 in 5%,
1.861.3 in 1% and 4.961.5 in 20% O2(Fig. 6D, left).
A similar trend was evident for the absolute number of
proliferating oligodendroglial progenitors: at 10 days, 8.562% of
the cells in 2.5% oxygen co-expressed Ki67+ and GalC, compared
to 2.761 in 5%, 3.160.5 in 1% and 1.260.3 in 20% O2(Fig. 6E,
left). Interestingly, of the total number of Ki67+ cells at 17 days in
2.5 and 5% O2, the neuronal MAP2+/Ki67+ (Fig. 6D, right) and
oligodendroglial GalC+/Ki67+ (Fig. 6E, right) accounted for 14%
and 22%, respectively, indicating that mild hypoxia is optimal for
survival and differentiation.
Oxygen Regulates Neurotransmitter Phenotypes of
Recently, we have shown that in 5% O2 IhNSCs generate
GABAergic and glutamatergic neurons, and only few tyrosine
hydroxylase (TH)+ neurons . Since mild hypoxia enhances
IhNSC neurogenesis, we examined the effect of oxygen on
Figure 2. Survival of IhNSC at different O2conditions. (A) Histogram showing the percentage of apoptotic cells (assessed by TUNEL assay, see
methods) at 4, 8, 12 and 24 hours from dissociation (N=3). Values are means6S.E.M. At 1% O2the percentage of apoptotic cells in culture is
significantly higher than at other conditions at each time point (p,0.001 at 4h). The difference among all the values at the different oxygen
concentrations was statistically significant (P,0.01) unless indicated (*P,0.05, n.s.,=not significant); one-way ANOVA followed by the Student’s t-
test. (B) Graph showing the percentages of non apoptotic (na) or apoptotic (ap) cells as assessed by JC1 staining assay for the detection of cells with
depolarized mitochondrial potential, corresponding to apoptotic cells (N=2). Values are means6S.E.M. No significant differences were detected
between the different conditions. (C) Representative images of live dissociated IhNSC in 1%, 2.5%, 5% and 20% oxygen plated onto an adhesive
substrate and stained with the mitochondrial and nuclear dyes JC1 (red) and Hoechst (blue) at 4 hours upon dissociation. The images were obtained
by merging the bright field and fluorescent (for JC1 and Hoechst) pictures of the cells. Scale bars: 10mm.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org5 January 2010 | Volume 5 | Issue 1 | e8575
neuronal differentiation into GABA, glutamatergic, catecholamin-
ergic (TH), cholinergic (ChAT) and serotoninergic (59-HT)
subtypes. At 17 days, the GABA+ cells accounted for 10.360.9
and 9.360.2% in 5% and 2.5% O2respectively, but significantly
decreased to 2.460.9 in 1% O2and 260.4 in 20% O2(Fig. 7A
and C). On the other hand, we observed a tendency of the fraction
of glutamatergic neurons to be maximal in severe and mild
hypoxic conditions (9.661.5% in 1% O2; 10.660.3% in 2.5% O2
Figure 3. Survival and apoptosis during differentiation of IhNSC at different O2conditions. (A) Western blot analysis of the expression of
apoptotic markers at neurosphere stage and at 10 days differentiation. Lysates were prepared from undifferentiated and differentiated IhNSCs grown
at the indicated oxygen concentrations and immunoblotted with antibodies specific for cleaved PARP, cleaved caspases 9 and 3, p53 and b-actin, the
latter to normalize bands for equal loading of proteins per lane. Bands were quantified by densitometry analysis of the ECL-exposed films. (B)
Pyknotic nuclei detected at 10 and 17 days of differentiation (N=3). Values are means6S.E.M. The difference among all the values at the different
oxygen concentrations was statistically significant (P,0.01) unless indicated with an asterisk (*P,0.05, n.s.=not significant); one-way ANOVA
followed by the Student’s t-test. (C) Representative images of pyknotic nuclei (white arrows, Hoechst stain, blue) at 17 days of differentiation. (D)
Representative images of mithocondria (HuMi, red) at 17days of differentiation. Note proliferating cells (Ki67+, green) and the different distribution of
mithocondria at 1% O2relative to other O2concentrations. Scale bars: 20mm (C), 10mm (D). For A and B, the differences at different oxygen
concentrations was statistically significant (P,0.01) unless indicated (*P,0.05, n.s.: not significant); one-way ANOVA followed by the Student’s t-test.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org6January 2010 | Volume 5 | Issue 1 | e8575
and 12.362.4% in 5% O2) and slightly decreased with increasing
oxygen concentrations (7.960.4% in 20% O2) (Fig. 7B and D),
consistent with previous studies . As hypoxia increases
mesencephalic differentiation into TH+ neurons [16,17] we tested
if also telencephalic-diencephalic precursors could acquire a
dopaminergic phenotype by lowered oxygen. Of note, only 1%
of differentiated IhNSCs are TH+ when cultured in 5% oxygen
. We found that no TH+ cells were generated in 20% O2, and
percentages comparable to 5% oxygen were obtained under 2.5%
and 1% O2. Sporadic cholinergic (ChAT+) neurons were observed
in mild hypoxia, whereas 59-HT+ neuronal cells were never
Severe Hypoxia Impairs Differentiation of IhNSCs
Hypoxia promotes an undifferentiated state in several
primitive and precursor cell populations . As lower
percentages of neuronal and glial cells were generated from
IhNSCs cultured at 1% O2, we searched for the presence of
quiescent or immature progenitors being unable to reach
terminal differentiation. To this aim, the intermediate filament
nestin (Fig. 8 A–G) and vimentin (Fig. 8 H–I) were used to
discriminate stem and progenitor cells from the differentiated
progeny, as reported . Nestin expression progressively
decreased with differentiation in mild hypoxia, to the least
extent in 1% O2, as revealed by the greater number of nestin+
cells detected by immunofluorescence (Fig. S1) and Western blot
(Fig. 8C). No co-expression of nestin and b-tubIII was detected
(Fig. 8A and B), indicating that nestin+ cells in 1% O2do not
become immature neuroblasts. Bona fide NSCs in vivo  are
positive for both nestin and GFAP. These markers were co-
expressed in many cells in 1% O2(Fig. 8D), and only in a small
fraction in 20% O2 (Fig. 8G), suggesting an incomplete
differentiation process in stringent oxygen culture conditions.
To exclude that this effect could reflect an aberrant commit-
ment of IhNSC-derived progenitors to the astrocytic lineage, we
assessed the expression of vimentin, an early astroglial marker,
together with the early neuronal marker b-tubIII. While all
cultures contained a consistent number of vimentin+ cells, none
co-expressed b-tubIII (Fig. 8H and I). These experiments thus
indicate that hypoxia influences the differentiation capacity of
IhNSC, forcing the progenitors into a quiescent state.
Oxygen plays a central role in regulating CNS progenitor cell
proliferation, development and homeostasis. In the mammalian
brain, interstitial tissue O2levels range from about 0.55 to 8% 
and there is evidence that control of O2availability is central to the
normal architecture of CNS and underlies the etiology of several
neurological diseases. Here we have shown that mild hypoxia (2.5–
5% O2), which better approximates the physiological setting,
promotes survival of actively replicating IhNSCs as well as the
yield of oligodendroglial and neuronal cells.
Proliferation and Survival in Hypoxic Conditions
Several studies show that low oxygen concentrations (1–5%)
allow cell proliferation through adapted mitochondrial respiration
[30,31]. Under anoxic conditions, mitochondrial respiration is
inhibited and the energy is provided by anaerobic glycolysis,
energetically insufficient to fully support cell proliferation and
differentiation . Here, we have demonstrated that O2
concentration is critical for the growth and survival of IhNSCs.
Mild hypoxia in particular, enhanced the proliferation of IhNSCs,
compared with 1% and 20% O2, in agreement with previous data
on neural crest stem cells, neuronal progenitors or mesencephalic
precursor cells [12,16,17,32]. In accordance to these findings,
Yoshida et al  have recently shown that mild hypoxia is the
optimal condition for both the generation and proliferation of
induced pluripotent stem cells (iPS) from human somatic cells in
comparison with 1% or 21% O2, leading respectively to cytotoxic
effects or to a less efficient reprogramming process. Indeed, in our
study the neurosphere assay showed that in marked contrast to
2.5–5% O2, 1% O2decreased the proliferation of IhNSCs and
raised the rate apoptotic, when compared to 20% O2. In
comparison with immortalized IhNSC , parental hNSCs were
even unable to proliferate in 1% O2, dying after few passages,
presumably because of cell cycle arrest and inhibition of
transcriptional activity [34–36]. Thus, IhNSC provided a unique
Figure 4. Mild hypoxia increases the percentage of IhNSC-derived neurons during in vitro differentiation. IhNSC were differentiated
onto an adhesive substrate in medium without mitogens and fixed for immunocytochemical analysis after 3, 10 and 17 days in vitro. (A)
Quantification of the percentage of neurons (b-tubIII+) over the total nuclei (DAPI+) number. Values are means6S.E.M (N=3). Neurons were
significantly less represented at 1% O2with respect to the other O2concentrations at 10 and 17 days in vitro. The difference among all the values at
the different oxygen concentrations was statistically significant (P,0.01) unless indicated (*P,0.05), n.s.,=not significant; one-way ANOVA followed
by the Student’s t-test.(B) Quantification of the percentage of astrocytes (GFAP+) over the total nuclei (DAPI+) number (N=3). Values are
means6S.E.M. At 10 days in vitro the percentage of astrocytes generated in 1% O2was significantly (*p,0.05) higher with respect to the other
conditions (1% O2vs 2.5% O2, p,0.01). All other values were not significantly different (n.s.,=not significant); one-way ANOVA followed by the
Student’s t-test. (C) Immunocytochemistry of differentiated cells showing the morphology of b-tubIII+ (red) and GFAP+ (green) cells in 2.5% oxygen at
10 days in vitro. Scale bar, 10mm.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org7 January 2010 | Volume 5 | Issue 1 | e8575
tool to study the effects of severe hypoxia on hNSC self-renewal
The proliferation rate of NSCs as determined by the neuro-
sphere assay represents the output of two variables i) the cell cycle
kinetics of the surviving precursors, ii) the rate of survival after
dissociation. To elucidate at what extent oxygen could affect the
cell cycle, we performed BrdU incorporation assay, but no major
differences in the fraction of S-phase cells were detected among
IhNSCs cultured at various O2 concentrations. Furthermore,
viability assay showed a significant fraction of dead cells at 1% and
20% O2, suggesting an enhancement of self-renewal capacity in
mild hypoxia. Taking together, the BrdU incorporation and cell
viability data, and applying Steel’s formula, we found no
appreciable differences in the average cycling times of IhNSCs
cultured at the diverse oxygen conditions, leading to conclude that
oxygen concentration affects IhNSC survival but not the cell cycle
kinetics. To gain further insight into IhNSC survival after
dissociation, we quantified apoptotic cells by TUNEL analysis
and found a dramatic increase of cell death in 1% oxygen, which
evidently accounts for the slow growth curve of neurospheres. In
view of this, we postulated that a less efficient mitochondrial
activity at 1% oxygen could contribute to severe hypoxia-induced
apoptosis . To test this hypothesis, we investigated the
mitochondria activity by JC1 staining assay and unexpectedly
found no impairment at 1% with respect to 2.5, 5 and 20%
oxygen. Furthermore, the amount of mitochondrial JC1 aggre-
gates as detected by fluorescence microscopy in 1% was
comparable to 2.5%, 5% and 20% oxygen.
Cell Survival during Differentiation
We demonstrated that oxygen concentration is also critical for
the neural differentiation of IhNSCs. In marked contrast to 2.5
and 5% O2, 1% O2 decreased the neural differentiation
potential and increased the cell death also when compared with
20% O2.We evaluated the rate of survival of IhNSCs during
differentiation by counting the pyknotic nuclei and found a
significantly higher percentage of apoptotic cells at 1% oxygen,
Figure 5. Mild hypoxia enhances the percentage of IhNSC-derived mature neurons and oligodendrocytes during in vitro
differentiation. IhNSC were differentiated onto an adhesive substrate in medium without mitogens and fixed for immunocytochemical analysis at
3, 10 and 17 days. (A) Quantification of the percentage of mature neurons (MAP2+) over the total nuclei (DAPI+) number. Values are means6S.E.M
(N=3). At 10 and 17 days the percentages of MAP2+ neurons generated in 2.5% O2were significantly higher with respect to the other conditions (B)
Quantification of the percentage of oligodendrocytes (GalC+) over the total number of nuclei (DAPI+). Values are means6S.E.M (N=3). At 10 days in
vitro the percentages of GalC+ oligodendrocytess generated in 2.5% O2were significantly higher with respect to the other conditions, while at 17days
2,5% and 5% O2 conditions generated comparable numbers of GalC+ cells. The differences among all the values at 1%, 2.5%, 5% and 20% oxygen
was statistically significant (P,0.01) unless indicated (*P,0.05, n.s.,=not significant); one-way ANOVA followed by the Student’s t-test for all
experiments. (C–D) Immunocytochemistry of differentiated cells showing the morphology of MAP2+ (red in C) and GalC+ (red in D) cells at 2.5% O2at
10 days. Scale bars in C–D=10mm.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org8 January 2010 | Volume 5 | Issue 1 | e8575
Figure 6. Mild hypoxia supports the survival of IhNSC-derived proliferating neuronal and oligodendroglial progenitors. IhNSCs were
differentiated onto an adhesive substrate in medium without mitogens and fixed for immunocytochemical analysis at 3, 10 and 17 days. (A) Graph
showing the percentage of proliferating (Ki67+) IhNSC cells over total nuclei (DAPI+) at 3, 10 and 17 days during in vitro differentiation. Values are
means6S.E.M (N=3). The 2.5% O2resulted the most permissive condition to the proliferation of IhNSC-derived progenitors at each differentiation
time. (B–C) Immunostaining showing the co-localization (arrows) between the nuclear proliferation marker Ki67 (green) with the neuronal MAP2 (red
in B), and the oligodendroglial GalC (red in C) at 10 days. (D) Quantification of the percentage of proliferating neurons (MAP2+/Ki67+) over the total
nuclei (DAPI+) number (left) and over the total of proliferating cells (Ki67+) number (right) at 3, 10 and 17 days. Values are means6S.E.M (N=3). (E)
Quantification of the percentage of proliferating oligodendrocytes (GalC+/Ki67+) over the total nuclei (DAPI+) number (left) and over the total of
proliferating cells (Ki67+) number (right), at 10 and 17 days. Values are means6S.E.M (N=3). The differences among all the values at 1%, 2.5%, 5% and
20% oxygen was statistically significant (P,0.01) unless indicated (*P,0.05, n.s.,=not significant); one-way ANOVA followed by the Student’s t-test
for all experiments. Scale bars, (B–C) 10mm.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org9January 2010 | Volume 5 | Issue 1 | e8575
consistent with previous studies  showing that severe
hypoxia corresponds to pathopysiological conditions as in
cerebral ischemia . Considering that differentiation is
increased aerobic metabolism [39,40], we investigated the
mitochondria patterning during differentiation. At 1% oxygen
the mitochondrial aggregates showed a perinuclear localization
and pseudo-globular and speckled morphology at 17 days
during differentiation, typical of mitochondrial fission events
correlated with cell death or degeneration . In marked
contrast under mild hypoxia and normoxia the mitochondria
appeared fused and showed ‘Spaghetti-like’ pattern throughout
the cell body and processes.
We further investigated the multipotency of IhNSC, and as
expected, the neuronal differentiation was enhanced under mild
hypoxia, especially at 2.5% O2which yielded significantly higher
proportions of both b-tubIII+ and MAP2+ cells when compared
with 1% and 20% O2. Thus it seems to be a common
phenomenon that proliferation and differentiation of NSCs or
neuronal progenitor cells is higher under physiological conditions
than under non-physiologically high or low oxygen.
We have shown, in accordance with previous studies [16,41,42],
that culturing cells in physiologically normal (2.5–5% O2) than
non-physiologically high (20% O2) or low (1% O2) conditions
drastically enhance neuronal and oligodendroglial differentiation
of hNSC. In particular, at 2.5% O2IhNSC generated the highest
percentages of neuronal cells (up to 28% of b-tubIII+ and 44% of
MAP2+). Consistent with previous studies , we observed an
increase of GABAergic and to less extent of glutamatergic
neuronal phenotypes under mild hypoxia.
Oxygen concentration has been implicated in controlling
oligodendrocyte progenitor proliferation and survival . O2A
progenitors with a more reduced state have a greater likelihood of
self-renewal compared to cells with a more oxidized state. We
addressed the role of oxygen in oligodendroglial differentiation
and showed a marked increase of GalC+ oligodendrocytes in mild
hypoxia (up to 15.5%) compared to other O2concentrations. The
analysis of apoptosis during differentiation suggests that an
enhancement of the survival of IhNSC-derived progenitors could
account for the increase of the neurogenic/oligodendrogenic
potential. It is well established that neuronal and oligodendroglial
cells are characterized by an higher threshold of vulnerability to
Figure 7. Oxygen regulates neurotrasmitter phenotypes of IhNSC-derived neurons. IhNSCs were differentiated onto an adhesive
substrate in medium without mitogens and fixed for immunocytochemical analysis at 3, 10 and 17 days. (A) Quantification of the percentage of
GABAergic neurons (GABA+) over the total nuclei (DAPI+) number, at 17 days. Values are means6S.E.M (N=3). At 17 days, the percentage of GABA+
cells significantly decreased in 1% O2and 20% O2with respect to mild hypoxia conditions. (B) Quantification of the percentage of glutamatergic
neurons (GLUTA+) over the total nuclei (DAPI+) number at 17 days. Values are means6S.E.M (N=3). At 17 days the percentage of GLUTA+ cells was
significantly higher in hypoxic conditions with respect to 20% O2. The differences among all the values at 1%, 2.5%, 5% and 20% oxygen was
statistically significant (P,0.01) unless indicated (*P,0.05, n.s.,=not significant); one-way ANOVA followed by the Student’s t-test for all experiments.
(C–D) Immunostaining showing the morphology of GABA+ cells (C, green) and GLUTA+ cells (D, green) in 2.5% O2at 17 days. Scale bars 10mm.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org10January 2010 | Volume 5 | Issue 1 | e8575
hypoxic-ischemic injury with respect to the astroglial phenotype
. Indeed, we observed an increase of GFAP+ cells generated by
differentiation of IhNSC in 1% O2, with a concomitant decrease
of neuronal and oligodendroglial cells relative to higher oxygen
concentrations. Moreover, the analysis of early markers for
immature progenitors showed that at 1% O2nestin expression
was only slightly down-regulated with differentiation and that a
significant fraction of cells co-expressed nestin and GFAP whereas
faint levels of nestin were detectable at other oxygen concentra-
tions. This observation supports the proposal that self-renewing
divisions of stem cells are conditioned by hypoxic metabolic type
In our study, a major apoptotic tendency of IhNSCs undergoing
differentiation at 1% oxygen is probably paralleled by the rest of
surviving NSC in a state of quiescence  where they remain
GFAP+/nestin+ and with low mitochondrial content, the latter is
typical of severe hypoxic metabolic condition [39,40]. We have
also excluded that any aberrant event could account for the
presence of nestin+/GFAP+ cells during differentiation at 1%
oxygen, by checking the expression of early markers specific for
neuronal and astroglial lineages. The non co-localization of nestin
with b-tubIII and of vimentin with b-tubIII supports the
identification of a major fraction of nestin+/GFAP+ IhNSCs with
In conclusion, our study provides evidence that mild hypoxia,
known to occur in the NSC niches of the adult brain, increases
proliferation, reduces cell death and enhances neuronal and
oligodendroglial differentiation of hNSC. These findings represent
an important advance for studies on brain injuries like stroke and
for the ex vivo generation of specific neurons and oligodendrocytes
for neurodegenerative diseases.
ized human neural stem cells (IhNSCs). IhNSCs were differenti-
ated onto an adhesive substrate in medium without mitogens and
fixed for immunocytochemical analysis at ten days. The graph
shows the percentage of early undifferentiated IhNSCs (nestin+)
over total nuclei (DAPI+ cells) at ten days during in vitro
differentiation. Values are means6S.E.M (N=3). In 1% O2, the
majority of the cells are immature nestin+ progenitors. The
differences among all the values at 1%, 2.5%, 5%, and 20%
oxygen was statistically significant (P,0.01) unless indicated
Severe hypoxia impairs differentiation of immortal-
Figure 8. Hypoxia impairs differentiation of IhNSCs. IhNSCs were differentiated onto an adhesive substrate in medium without mitogens and
fixed for immunocytochemical analysis at 3, 10 and 17 days. (A–B) Immunostaining showing the non co-localization between the neuronal marker b-
tubIII (green) and nestin (red) in IhNSC cells at 10 days in 1% (A) and 2,5% (B) O2. (C) Western blot analysis of the variation of nestin (arrows), during
differentiation. Lysates were prepared from IhNSCs grown and differentiated for 3 and 10 days at the indicated oxygen concentrations. Lysates were
immunoblotted with antibodies specific for nestin and GADPH, the latter to normalize bands for equal loading of proteins per lane. Bands were
quantified by densitometry analysis of the ECL-exposed films. (D–G) Immunostaining showing the co-localization (D, yellow) between GFAP (green)
and nestin in cells at 10 days in 1% O2. The two markers never showed colocalization at 2,5% (E) and 5% (F) O2at the same time point, and only few
cells were co-expressing the two markers at 20% O2(G, yellow). (H–I) Immunostaining showing the non co-localization between the neuronal marker
b-tubIII (red) and vimentin (green) in IhNSC cells at 10 days in 1% (H) and 2,5% (I) O2. Scale bars, (A–B, D–G) 20mm, (H,I) 10 mm.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org 11 January 2010 | Volume 5 | Issue 1 | e8575
(*P,0.05, n.s.=not significant); one-way ANOVA followed by the
Found at: doi:10.1371/journal.pone.0008575.s001 (1.84 MB TIF)
We thank Pietro De Filippis and Patrizia Karoschitz for valuable discussion
and insight regarding this work.
Conceived and designed the experiments: Dd ALV LDF. Performed the
experiments: GS GL LC LRN LDF. Analyzed the data: GS DF EB LDF.
Contributed reagents/materials/analysis tools: Dd ALV. Wrote the paper:
DF Dd LDF.
1. Vescovi AL, Parati EA, Gritti A, Poulin P, Ferrario M, et al. (1999) Isolation and
cloning of multipotential stem cells from the embryonic human CNS and
establishment of transplantable human neural stem cell lines by epigenetic
stimulation. Exp Neurol 156: 71–83.
2. McKay R (1997) Stem cells in the central nervous system. Science 276: 66–71.
3. Gritti A, Parati EA, Cova L, Frolichsthal P, Galli R, et al. (1996) Multipotential
stem cells from the adult mouse brain proliferate and self-renew in response to
basic fibroblast growth factor. J Neurosci 16: 1091–1100.
4. Gritti A, Frolichsthal-Schoeller P, Galli R, Parati EA, Cova L, et al. (1999)
Epidermal and fibroblast growth factors behave as mitogenic regulators for a
single multipotent stem cell-like population from the subventricular region of the
adult mouse forebrain. J Neurosci 19: 3287–3297.
5. Gritti A, Cova L, Parati EA, Galli R, Vescovi AL (1995) Basic fibroblast growth
factor supports the proliferation of epidermal growth factor-generated neuronal
precursor cells of the adult mouse CNS. Neurosci Lett 185: 151–154.
6. Panchision D, Hazel T, McKay R (1998) Plasticity and stem cells in the
vertebrate nervous system. Curr Opin Cell Biol 10: 727–733.
7. Ming GL, Song H (2005) Adult neurogenesis in the mammalian central nervous
system. Annu Rev Neurosci 28: 223–250.
8. Semenza GL (2000) HIF-1: mediator of physiological and pathophysiological
responses to hypoxia. J Appl Physiol 88: 1474–1480.
9. Niizuma K, Endo H, Chan PH (2009) Oxidative stress and mitochondrial
dysfunction as determinants of ischemic neuronal death and survival.
J Neurochem 109 Suppl 1: 133–138.
10. Siesjo BK (1992) Pathophysiology and treatment of focal cerebral ischemia. Part
I: Pathophysiology. J Neurosurg 77: 169–184.
11. Siesjo BK (1992) Pathophysiology and treatment of focal cerebral ischemia. Part
II: Mechanisms of damage and treatment. J Neurosurg 77: 337–354.
12. Burgers HF, Schelshorn DW, Wagner W, Kuschinsky W, Maurer MH (2008)
Acute anoxia stimulates proliferation in adult neural stem cells from the rat
brain. Exp Brain Res 188: 33–43.
13. Park KI, Hack MA, Ourednik J, Yandava B, Flax JD, et al. (2006) Acute injury
directs the migration, proliferation, and differentiation of solid organ stem cells:
evidence from the effect of hypoxia-ischemia in the CNS on clonal ‘‘reporter’’
neural stem cells. Exp Neurol 199: 156–178.
14. Horie N, So K, Moriya T, Kitagawa N, Tsutsumi K, et al. (2008) Effects of
oxygen concentration on the proliferation and differentiation of mouse neural
stem cells in vitro. Cell Mol Neurobiol 28: 833–845.
15. Erecinska M, Silver IA (2001) Tissue oxygen tension and brain sensitivity to
hypoxia. Respir Physiol 128: 263–276.
16. Storch A, Paul G, Csete M, Boehm BO, Carvey PM, et al. (2001) Long-term
proliferation and dopaminergic differentiation of human mesencephalic neural
precursor cells. Exp Neurol 170: 317–325.
17. Studer L, Csete M, Lee SH, Kabbani N, Walikonis J, et al. (2000) Enhanced
proliferation, survival, and dopaminergic differentiation of CNS precursors in
lowered oxygen. J Neurosci 20: 7377–7383.
18. De Filippis L, Lamorte G, Snyder EY, Malgaroli A, Vescovi AL (2007) A novel,
immortal, and multipotent human neural stem cell line generating functional
neurons and oligodendrocytes. Stem Cells 25: 2312–2321.
19. Flax JD, Aurora S, Yang C, Simonin C, Wills AM, et al. (1998) Engraftable
human neural stem cells respond to developmental cues, replace neurons, and
express foreign genes. Nat Biotechnol 16: 1033–1039.
20. Ryder EF, Snyder EY, Cepko CL (1990) Establishment and characterization of
multipotent neural cell lines using retrovirus vector-mediated oncogene transfer.
J Neurobiol 21: 356–375.
21. Villa A, Snyder EY, Vescovi A, Martinez-Serrano A (2000) Establishment and
properties of a growth factor-dependent, perpetual neural stem cell line from the
human CNS. Exp Neurol 161: 67–84.
22. Carlessi L, De Filippis L, Lecis D, Vescovi A, Delia D (2009) DNA-damage
response, survival and differentiation in vitro of a human neural stem cell line in
relation to ATM expression. Cell Death Differ 16: 795–806.
23. Erba E, Bergamaschi D, Bassano L, Damia G, Ronzoni S, et al. (2001)
Ecteinascidin-743 (ET-743), a natural marine compound, with a unique
mechanism of action. Eur J Cancer 37: 97–105.
24. Salvioli S, Ardizzoni A, Franceschi C, Cossarizza A (1997) JC-1, but not
DiOC6(3) or rhodamine 123, is a reliable fluorescent probe to assess delta psi
changes in intact cells: implications for studies on mitochondrial functionality
during apoptosis. FEBS Lett 411: 77–82.
25. Ivanovic Z (2009) Hypoxia or in situ normoxia: The stem cell paradigm. J Cell
Physiol 219: 271–275.
26. Parmar K, Mauch P, Vergilio JA, Sackstein R, Down JD (2007) Distribution of
hematopoietic stem cells in the bone marrow according to regional hypoxia.
Proc Natl Acad Sci U S A 104: 5431–5436.
27. Knott AB, Perkins G, Schwarzenbacher R, Bossy-Wetzel E (2008) Mitochon-
drial fragmentation in neurodegeneration. Nat Rev Neurosci 9: 505–518.
28. Simon MC, Keith B (2008) The role of oxygen availability in embryonic
development and stem cell function. Nat Rev Mol Cell Biol 9: 285–296.
29. Lendahl U, Zimmerman LB, McKay RD (1990) CNS stem cells express a new
class of intermediate filament protein. Cell 60: 585–595.
30. Guzy RD, Schumacker PT (2006) Oxygen sensing by mitochondria at complex
III: the paradox of increased reactive oxygen species during hypoxia. Exp
Physiol 91: 807–819.
31. Jiang BH, Semenza GL, Bauer C, Marti HH (1996) Hypoxia-inducible factor 1
levels vary exponentially over a physiologically relevant range of O2 tension.
Am J Physiol 271: C1172–1180.
32. Shingo T, Sorokan ST, Shimazaki T, Weiss S (2001) Erythropoietin regulates
the in vitro and in vivo production of neuronal progenitors by mammalian
forebrain neural stem cells. J Neurosci 21: 9733–9743.
33. Yoshida Y, Takahashi K, Okita K, Ichisaka T, Yamanaka S (2009) Hypoxia
enhances the generation of induced pluripotent stem cells. Cell Stem Cell 5:
34. Kaidi A, Williams AC, Paraskeva C (2007) Interaction between beta-catenin and
HIF-1 promotes cellular adaptation to hypoxia. Nat Cell Biol 9: 210–217.
35. Koshiji M, Kageyama Y, Pete EA, Horikawa I, Barrett JC, et al. (2004) HIF-
1alpha induces cell cycle arrest by functionally counteracting Myc. Embo J 23:
36. Zhang H, Gao P, Fukuda R, Kumar G, Krishnamachary B, et al. (2007) HIF-1
inhibits mitochondrial biogenesis and cellular respiration in VHL-deficient renal
cell carcinoma by repression of C-MYC activity. Cancer Cell 11: 407–420.
37. Fruehauf JP, Meyskens FL Jr (2007) Reactive oxygen species: a breath of life or
death? Clin Cancer Res 13: 789–794.
38. Semenza GL, Agani F, Feldser D, Iyer N, Kotch L, et al. (2000) Hypoxia, HIF-
1, and the pathophysiology of common human diseases. Adv Exp Med Biol 475:
39. Freyer JP (1998) Decreased mitochondrial function in quiescent cells isolated
from multicellular tumor spheroids. J Cell Physiol 176: 138–149.
40. Radley JM, Ellis S, Palatsides M, Williams B, Bertoncello I (1999) Ultrastructure
of primitive hematopoietic stem cells isolated using probes of functional status.
Exp Hematol 27: 365–369.
41. Morrison SJ, Csete M, Groves AK, Melega W, Wold B, et al. (2000) Culture in
reduced levels of oxygen promotes clonogenic sympathoadrenal differentiation
by isolated neural crest stem cells. J Neurosci 20: 7370–7376.
42. Ohta K, Iwai M, Sato K, Omori N, Nagano I, et al. (2003) Dissociative increase
of oligodendrocyte progenitor cells between young and aged rats after transient
cerebral ischemia. Acta Neurochir Suppl 86: 187–189.
43. Pistollato F, Chen HL, Schwartz PH, Basso G, Panchision DM (2007) Oxygen
tension controls the expansion of human CNS precursors and the generation of
astrocytes and oligodendrocytes. Mol Cell Neurosci 35: 424–435.
44. Cipolleschi MG, Dello Sbarba P, Olivotto M (1993) The role of hypoxia in the
maintenance of hematopoietic stem cells. Blood 82: 2031–2037.
Effects of Oxygen on Human NSC
PLoS ONE | www.plosone.org 12January 2010 | Volume 5 | Issue 1 | e8575