Increased “Vigilance” of Antioxidant Mechanisms in Neural Stem Cells Potentiates Their Capability to Resist Oxidative Stress



Although the potential value of transplanted and endogenous neural stem cells (NSCs) for the treatment of the impaired central nervous system (CNS) has widely been accepted, almost nothing is known about their sensitivity to the hostile microenvironment in comparison to surrounding, more mature cell populations. Since many neuropathological insults are accompanied by oxidative stress, this report compared the alertness of antioxidant defense mechanisms and cell survival in NSCs and postmitotic neural cells (PNCs). Both primary and immortalized cells were analyzed. At steady state, NSCs distinguished themselves in their basal mitochondrial metabolism from PNCs by their lower reactive oxygen species (ROS) levels and higher expression of the key antioxidant enzymes uncoupling protein 2 (UCP2) and glutathione peroxidase (GPx). Following exposure to the mitochondrial toxin 3-nitropropionic acid, PNC cultures were marked by rapidly decreasing mitochondrial activity and increasing ROS content, both entailing complete cell loss. NSCs, in contrast, reacted by fast upregulation of UCP2, GPx, and superoxide dismutase 2 and successfully recovered from an initial deterioration. This recovery could be abolished by specific antioxidant inhibition. Similar differences between NSCs and PNCs regarding redox control efficiency were detected in both primary and immortalized cells. Our first in vivo data from the subventricular stem cell niche of the adult mouse forebrain corroborated the above observations and revealed strong baseline expression of UCP2 and GPx in the resident, proliferating NSCs. Thus, an increased “vigilance” of antioxidant mechanisms might represent an innate characteristic of NSCs, which not only defines their cell fate, but also helps them to encounter oxidative stress in diseased CNS.


The restorative potential of neural stem cells (NSCs) is based on their abilities to provide cell replacement and serve as vehicles for gene therapy, but also, and importantly, on their capacity to stimulate reparative responses in the diseased host while promoting re-establishment of homeostasis [1, [2]–3]. During these events, NSCs and their progeny, together with all the other, more mature cell types, become exposed to the hostile environment of the diseased central nervous system (CNS). Although many investigations are addressing the deteriorating behavior of the mature neurons and glia under such conditions, interestingly, almost nothing is known about the vulnerability and the response of the NSCs which, frequently, appear to better resist environmental stress [4, 5].

Oxygen is necessary for life, but, paradoxically, its metabolism produces reactive oxygen species (ROS) as by-products highly toxic to cells. Because of its elevated metabolic rate, high oxygen consumption, and relatively reduced capacity for cellular regeneration compared with other organs, the brain is believed to be particularly susceptible to the damaging effects of ROS. This becomes evident in diseases such as Parkinson's disease or Huntington's disease, where various indices of ROS damage have been reported within the specific brain regions that undergo selective neurodegeneration. To make matters worse, although mammalian cells have evolved several resistance and repair mechanisms to deal with oxidative stress and the associated damage, the activities of various antioxidant defense molecules that would normally counteract the injurious effects of ROS are reduced in the brain [6]. This is particularly true for the enzymes superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx).

With regard to stem cells, the maintenance of a balance between self-renewal and differentiation is pivotal for their function in development, tissue repair, and homeostasis. In this context, a new role of the cellular redox state control has been recognized, affecting multiple processes related to cell proliferation and differentiation [7]. The redox potential and the concentrations of free radicals in cells are determined by a balance between their rates of production and clearance and are controlled by various antioxidant compounds such as SOD, GPx, and CAT. Moreover, uncoupling of the mitochondrial respiratory chain by proton transporters such as the uncoupling proteins (UCPs) reduces free radical production and consequently regulates the cellular redox state [8]. The control of the latter may also depend on the convergence of different signaling pathways, whereas its changes may, in turn, influence processes controlling self-renewal and differentiation. Thus, as demonstrated recently, the degree of cellular oxidation/reduction in progenitor cells can be altered by growth factor signaling [9]. Depending on the nature of the signaling molecules, it can render the progenitors more reduced or oxidized [9]. In the present study, we therefore hypothesized that NSCs and their postmitotic progeny differ in their free radical household to an extent that also changes their ability to resist oxidative stress. This should be reflected, among others, in distinctive expression patterns of their antioxidant enzymes.

To explore this thought, we examined cultured NSCs and their 7-day-differentiated progeny (postmitotic neural cells [PNCs]) and compared (a) their mitochondrial activities and ROS levels at steady state and after exposure to the mitochondrial toxin 3-nitropropionic acid (3-NP), and (b) the corresponding expression levels of antioxidant molecules and their importance in the response to 3-NP intoxication. Two types of NSCs and PNCs often used in current stem cell research were examined: primary NSCs isolated from the subventricular zone (SVZ) of newborn mice and the immortalized NSC clone C17.2 [10, 11], successfully used in our previous studies of NSC-mediated rescue of neurons [5]. Assessed were their cellular characteristics relating to redox modulation and cell viability, such as mitochondrial activity, concentration of free radicals, ratios of surviving and apoptotic cells, and intensity of cell proliferation. Next, we analyzed the expression of antioxidative enzymes, focusing on the four key modulators UCP2, GPx, manganese-containing mitochondrial superoxide dismutase (Mn-SOD or SOD2), and CAT [8, 12, 13].

This report also includes our pilot immunohistochemical data pertaining to steady-state expression of antioxidant modulators in the subventricular stem cell niche in adult mouse brains. These data will serve as the baseline in future studies addressing the vulnerability of endogenous NSCs to 3-NP-induced oxidative stress.

Materials and Methods

Cell Cultures

Primary Cells.

NSCs isolated from the subventricular zone of newborn C57BL/6 mice were grown under standard conditions in uncoated dishes and serum-free Neurobasal medium supplemented with 2% B27 (NB27), 20 ng/ml epidermal growth factor (EGF), 10 ng/ml basic fibroblast growth factor (bFGF), and 8 μg/ml heparin. After 7 and 14 days, primary and secondary neurosphere cultures were split and plated at 104 cells per ml in NB27 containing EGF, bFGF, and heparin to allow formation of tertiary neurospheres. The latter were used for all the experiments regarding proliferating NSCs or differentiated in culture on poly(l-lysine) for another 7 days without growth factors but in the presence of 1% fetal bovine serum (FBS) for the derivation of PNCs.

Immortalized Cells.

To match the growth conditions of the primary cells, immortalized cells (clone C17.2 [10]) were grown in serum-free Dulbecco's modified Eagle's medium (DMEM)/Ham's F-12 medium supplemented with N2, 50 U/ml penicillin, 50 μg/ml streptomycin, heparin (8 μg/ml), bFGF (20 ng/ml), and EGF (10 ng/ml) as described [14]. Only early-passage (passages 1–3) stocks of the originally immortalized cells were used for the preparation of tertiary neurospheres. To derive PNCs, the latter were grown on poly(l-lysine) without growth factors and in the presence of 1% FBS.

All cells were grown in a standard humidified incubator at 37°C with 20% O2. The state of cell differentiation from tertiary neurospheres was monitored by immunodetection of Ki-67 and nestin, both markers for proliferating and immature cells (Fig. 1A, a–h), class III-β-tubulin (Fig. 1A, i–l, Tuj-1), and glial fibrillary acidic protein (GFAP) (Fig. 1A, m–p, glial marker).

Induction of Oxidative Stress In Vitro

Cultures of NSCs and PNCs at comparable cell densities (2 × 104 cells per well) were treated with 3-NP (0.05 mM) for 24 hours and assayed for the next 5 days in vitro (DIVs) (i.e. 120 hours).

Mitochondrial Activity

The colorimetric 4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay was used to measure mitochondrial functionality in cells [15]. Briefly, cells were incubated with 0.25 mg/ml MTT for 3 hours at 37°C, and mitochondrial enzyme activity was measured in culture supernatants in a spectrophotometer (Axon Instruments/Molecular Devices Corp., Union City, CA, at 570 nm, with a reference wavelength of 630 nm.

ROS Levels

The 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate (CM-H2DCFDA) fluorescence assay [16] was applied to measure the levels of free radicals in cells. CM-H2DCFDA was added at 10 μM for 45 minutes to 3-NP-treated cells (during the last 45 minutes of exposure) and to control cultures. Then, the cells were washed with phosphate-buffered saline (PBS) (0.1 M, pH 7.2), and fluorescence recorded at a wavelength of 485 nm (excitation) and 535 nm (emission).


Cultures exposed to 3-NP for 24 hours were fixed in 4% paraformaldehyde (PFA) and incubated with the Hoechst dye 33342 (10 μg/ml) for 3 minutes in the dark. Using UV illumination, a 4,6-diamidino-2-phenylindole filter, and a ×40 objective, fluorescent (apoptotic) cells were evaluated in 15–20 visual fields with a Zeiss Axioplan-2 microscope and the percentage of apoptotic cells determined (Carl Zeiss, Jena, Germany,

Inhibition of Antioxidant Enzymes

To assess their direct relevance in the cellular antioxidative response, the enzymes UCP2, GPx, and SOD2 that showed significant differences in their expression levels at steady state and after a 3-NP challenge were inhibited with GDP [17], mercaptosuccinic acid (MS) [18], and dethylthiocarbamate (DETC) [19], respectively. The inhibitors were added to cultures 1 hour prior to 3-NP and left for the same duration as the toxin, or to 3-NP-free controls over the same culture period.

Immunocytochemistry In Vitro

Cells grown on poly(l-lysine) (0.1 mg/ml)-coated coverslips were fixed with 4% PFA and rinsed with PBS. Primary antibodies were diluted in blocking solution (5% goat serum supplemented with 0.2% bovine serum albumin, 0.1% Triton X-100 in PBS), and preparations were incubated overnight at 4°C. Specific binding was then revealed with the appropriate secondary antibodies conjugated to Alexa 488, 594, or 647 and diluted 1:500.

The following markers were used: for NSCs, nestin (1:1,000), Musashi (1:500), and Ki-67 (1:500); for PNCs, Tuj-1 (1:500) as neuronal marker and GFAP (1:1,000) as astroglial marker. Antibodies against UCP2, GPx, and SOD2 were diluted 1:200, 1:500, and 1:1,000, respectively. Stains omitting primary or secondary antibodies and recordings through nonspecific filters were used as signal specificity controls.

Histology and Immunohistochemistry of Brain Sections

Adult 16-week-old C57BL/6 mice were perfused with 4% PFA under deep pentobarbital anesthesia. The brains were postfixed in the same fixative for 24 hours and processed for routine cryostat sectioning, and 20-μm-thick serial coronal sections were collected. Rehydrated sections were blocked and immunostained with antibodies against UCP2 (1:100), Gpx (1:250), nestin (1:500), Ki-67 (1:250), and Musashi (1:2,500) using the same conditions as for the cell cultures described above. Stains occurred in coronal sections from levels 25–30 according to the stereotactic atlas of the mouse brain by Sidman et al. [20]. All animals were housed and maintained at Iowa State University, and all animal procedures carried out were approved by the Iowa State University Committee on Animal care and adhered to NIH guidelines [46].

Western Blotting

Cells were harvested, washed with ice-cold Ca2+-free PBS, and resuspended in 2 ml of homogenization buffer (20 mM Tris-HCl, pH 8.0, 10 mM EGTA, 2 mM EDTA, 2 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 25 μg/ml aprotinin, and 10 μg/ml leupeptin). Suspensions were sonicated for 10 seconds and centrifuged at 13,000g for 30 minutes, and the supernatants were collected as whole cell lysate fractions. Samples containing equal amounts of protein were separated by 15% SDS-polyacrylamide gel electrophoresis and blotted onto nitrocellulose membranes. After blocking of nonspecific binding sites with blocking reagent, antioxidant proteins were revealed by incubation of the membranes with antibodies against UCP2 (1:200), GPx (1:1,000), SOD2 (1:2,000), and CAT (1:1,000) overnight at 4°C. Secondary antibodies conjugated to horseradish peroxidase (1:2,000) were applied to visualize bound proteins in Amersham's enhanced chemiluminescence assay (Amersham Biosciences, Piscataway, NJ, Equal protein loading was confirmed by reprobing of the membranes for β-tubulin.


Western blots were scanned and the NIH software ImageJ, version 1.34, used to quantify the densities of immunoreactive bands by calculating the area under the peak curves corresponding to UCP2, GPx, SOD2, and CAT.


3-NP, GDP, DETC, and MS were obtained from Sigma-Aldrich (St. Louis, CM-H2DCFDA, MTT, and Hoechst 33342 stain were obtained from Molecular Probes Inc. (Eugene, OR, DMEM, neurobasal medium, DMEM with Ham's F-12 supplement, FBS, l-glutamine, penicillin/streptomycin, N2 supplement, B27 supplement, EGF, bFGF, and heparin were purchased from Invitrogen (Carlsbad, CA, Primary antibodies against nestin, UCP2, and GPx were received from Chemicon (Temecula, CA,, anti-SOD2 from Upstate (Lake Placid, NY,, anti-CAT from Genetex (San Antonio,, anti-Musashi from CeMines (Evergreen, CO,, anti-Ki-67 from DakoCytomation, Carpinteria, CA,, Tuj-1 from Covance (Berkeley, CA,, and anti-GFAP from Sigma-Aldrich. All secondary antibodies were purchased from Molecular Probes.

Data Analysis and Statistics

All data are expressed as means ± SEM and derived from at least three separate experiments. Statistical significance was determined by Dunnett's post hoc test for multiple comparisons with the control or by Bonferoni's multiple comparison tests performed on data from the different treatment groups. Single comparisons were evaluated using Student's t test or a Welch-corrected unpaired t test, where appropriate.


Steady-State Characteristics of NSCs and PNCs Related to Redox State and Cell Survival In Vitro

To obtain reference points for our study of 3-NP-induced oxidative stress in NSCs and PNCs, we first needed to assess basal levels of the evaluated cellular modalities, namely, mitochondrial activity (MTT assay), ROS production (CM-H2DCFDA oxidation), and apoptosis (Hoechst 33342 staining).

In NSC cultures, approximately 80% of the cells proliferated (Fig. 1A, a, b, 1B, d, white columns) and expressed nestin (Fig. 1A, e, f) but no Tuj-1 and GFAP, characterizing more differentiated cell types (Fig. 1A, i, j, m, n). After differentiation for 7 days, PNCs were characterized by less than 5% of dividing cells (Fig. 1A, c, d, 1B, d, hatched columns), had lost their nestin positivity (Fig. 1A, g, h), and expressed neuronal and glial markers (Fig. 1A, k, l, o, p).

Figure Figure 1..

Steady-state characteristics of NSCs and PNCs related to redox state and cell survival in vitro. (A): Photomicrophotographs comparing the expression of stem cell and postmitotic cell markers in NSCs and PNCs. Most of the cultured NSCs (left panels) expressed the nuclear marker Ki-67 (red), revealing all dividing cells (a, b), and nestin, a prototypic marker for undifferentiated NSCs (e, green; f, red). They were, however, negative for cell type-specific markers, such as the neuronal Tuj-1 (i, j) and the astrocytic GFAP (m, n). PNCs (right panels), on the other hand, had lost their Ki-67 and nestin positivity (c, d, g, h) and began to express Tuj-1 (k, red; l, green) and GFAP (o, green; p, red). The blue signal in a–p indicates nuclear DNA labeled with Hoechst 33342. Scale bar = 40 μm. (B): Quantitative evaluation of cell culture viability. Mitochondrial activity (MTT values; a), levels of free radicals (CM-H2DCFDA reaction; c), number of apoptotic cells (Hoechst 33342 staining; b), and cell proliferation (anti-Ki-67 immunostaining; d) were compared in cultured NSCs (white columns) and PNCs (hatched columns). Both primary and immortalized NSCs displayed stronger mitochondrial activity (a), lower ROS content (c), and less apoptosis (b) than the corresponding PNCs, indicating their better ROS buffering capacity. As expected, approximately 80%–85% of the NSCs was proliferating and stained for the Ki-67 marker (d), whereas PNCs remained almost all negative. ∗, p < .05; ∗∗∗, p < .001. Abbreviations: CM-H2DCFDA, 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate; FU, fluorescent units; GFAP, glial fibrillary acidic protein; MTT, 4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; NSC, neural stem cell; OD, optical density; PNC, postmitotic neural cell.

Striking differences in mitochondrial activity and intracellular ROS levels between NSCs and PNCs were recorded, whether primary or immortalized. NSCs (Fig. 1B, white columns) demonstrated a significantly (p < .05) greater steady-state mitochondrial activity (Fig. 1B, a) than their postmitotic counterparts (hatched columns) and lower ROS production (Fig. 1B, c). Correspondingly, PNCs also counted higher numbers of apoptotic cells (Fig. 1B, b). These data were the first indicators supporting our hypothesis that NSCs and PNCs differ in their basal redox states, with the NSCs being better equipped to control intracellular ROS and resist oxidative stress.

A comparison of the analyzed parameters between the primary and immortalized cells revealed subtle cell type-specific differences (Fig. 1B, a–d, compare left and right plots), although the same trend of their changes between NSCs (white columns) and PNCs (hatched columns) prevailed. Possible reasons for this finding will be discussed.

Response of NSCs and PNCs to 3-NP-Induced Oxidative Stress

After collection of the data pertaining to steady-state ROS metabolism, apoptosis, and cell proliferation in NSCs and PNCs, we next investigated their response to the mitochondrial toxin 3-NP.


No significant changes in NSC number, proliferation, and apoptosis were observed at the time of the toxin removal (t = 0), although the MTT readings had dropped approximately 20%–30% below control values (Fig. 2, dashed lines). Cell behavior remained constant during the next 24 hours, after which, mild deterioration became noticeable. The latter reached a peak at approximately 60–72 hours post-3-NP, when a spontaneous recovery began, resulting in slightly different end values in primary and immortalized cells. Although by the end of the 5th experimental day (120 hours), primary NSCs had returned to values of the untreated controls (Fig. 2A, 2C, 2E, 2G, 2I), immortalized NSCs remained at 75% in their MTT values (Fig. 2B) and also remained affected in their proliferative activity (Fig. 2J). Both translated to a reduction of 20% in their cell numbers (Fig. 2H) and suggested a less efficient control of intracellular ROS than in primary NSCs.

Figure Figure 2..

Response of NSCs and PNCs to 3-nitropropionic acid (3-NP)-induced oxidative stress. After a 24-hour-treatment of the cultures with 0.05 mM 3-NP, NSCs (dashed curve) proved to be less susceptible to the resulting oxidative stress than PNCs (solid curve). Not only were the initial effects (t = 0) less pronounced in the NSCs, but they also were capable, after transient deterioration between 36 and 72 hours, to recover. This was reflected in time-dependent changes of the cells' mitochondrial activity (A, B), increasing ROS levels (C, D), and numbers of apoptotic cells (E, F). Interestingly, although primary NSCs recovered completely within the 5-day experimental period, mitochondrial activity (B), cell numbers (H), and proliferation (J) of immortalized NSCs did not recuperate fully. PNCs (solid line) died within 72 hours (E–H); their death was accompanied by their rapidly degrading mitochondrial activity and rising ROS levels (A–D). (A–H): All values are expressed as percentage of the corresponding controls. (I, J): Control values referred to are those of Ki-67 labeling indices (percentage of labeled cells) in untreated NSC cultures, represented by the horizontal lines. Abbreviations: CM-H2DCFDA, 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate; MTT, 4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; NSC, neural stem cell; PNC, postmitotic neural cell.


The behavior of PNCs in response to 3-NP was radically different from that of the NSC cultures and appeared to perpetuate the initial steady-state difference between both culture types. Immediately after toxin removal (t = 0), PNCs already exhibited obvious detrimental changes in all the measured parameters (Fig. 2, solid lines). In primary and immortalized PNCs, the considerable initial drop in mitochondrial viability quickly became a sigmoid decline, bringing it to almost zero, at 72 hours post-3-NP (Fig. 2A, 2B). This drastic decay of the cultures was mirrored in an exponential increase in ROS content and numbers of apoptotic cells (Fig. 2c–2f). A corresponding decrease in cell numbers left hardly any surviving cells in the wells (Fig. 2G, 2H), which made changes of intracellular ROS and apoptosis after 3 DIVs hardly measurable. At all of the evaluated times, PNCs were almost completely negative for the nuclear proliferation marker Ki-67, irrespective of their treatment with 3-NP (Fig. 2I, 2J).

Differential Expression of Antioxidant Molecules Between NSCs and PNCs During Steady-State and After 3-NP Intoxication

The main enzymatic responses resulting in ROS detoxification are carried within the cells by molecules such as GPx, SOD2, and CAT. UCP2, a mitochondrial proton transporter, can also help lower ROS levels by reducing mitochondrial membrane potential [8, 21]. Thus, part of the contributing reasons explaining the difference in the redox state values between NSCs and PNCs were likely to be found in unequal activities of their antioxidant defense mechanisms. To explore this hypothesis, we analyzed the expression levels of the above enzymes, either in qualitative immunostains of cell cultures (Fig. 3) or quantitatively, from scans of Western blots (Fig. 4).

Figure Figure 3..

Steady-state expression of the antioxidant molecules UCP2 and GPx in NSCs and PNCs in vitro. Cultures of NSCs and PNCs were immunostained for their expression of the antioxidant molecules UCP2 (red) and GPx (red). The blue signal indicates nuclear DNA labeling with Hoechst 33342. Most (∼90%) primary and immortalized NSCs stained intensively for both proteins ([A], a, c, [B], a, c), whereas their PNC counterparts were only weakly labeled ([A], b, d, [B], b, d). Due to the larger size and amount of cytoplasm of the immortalized cells, their red cytoplasmic UCP2 and GPx stains overlap with the nuclear area in the two-dimensional view of their immunostain (in [A] and [B], compare a with c). Steady-state expression of superoxide dismutase 2 and catalase showed no differences between NSCs and PNCs (not shown). Scale bars = 20 μm. Abbreviations: GPx, glutathione peroxidase; NSC, neural stem cell; PNC, postmitotic neural cell; UCP2, uncoupling protein 2.

Figure Figure 4..

Western blot analysis of 3-nitropropionic acid (3-NP)-induced differential expression patterns of antioxidant molecules in NSCs and PNCs. Expression levels of antioxidant molecules (UCP2, GPx, SOD2, and CAT) in NSCs and PNCs were evaluated quantitatively in densitometry scans of Western blots prepared from whole protein extracts 12 and 60 hours after treatment of the cultures with 3-NP. Quantification was achieved by integration of the area under the densitometric peak curve of each protein and the data are plotted in comparison to the untreated controls. The intensity of the 3-NP-induced changes in the levels of antioxidant molecules varied considerably not only between NSCs and PNCs but also between primary and immortalized cells. Although NSCs (white columns) were characterized by strong upregulation of UCP2 and SOD2 (primary) (A, E), as well as GPx (immortalized) (D), no striking response to the toxin was found in PNCs (dashed columns). See main text for more details. ∗, p < .05; ∗∗, p < .01; ∗∗∗, p < .001. Abbreviations: CAT, catalase; GPx, glutathione peroxidase; hrs, hours; NSC, neural stem cell; PNC, postmitotic neural cell; SOD2, superoxide dismutase 2; U, untreated; UCP2, uncoupling protein 2.

Steady-State Expression of Antioxidants

Initial qualitative comparison of antioxidant steady-state levels between NSCs and PNCs revealed a much stronger expression of UCP2 and GPx in the former, for both primary and immortalized cells (Fig. 3). In contrast, both types of cells showed similar levels of SOD2 and CAT activity (not shown). This result was subsequently confirmed and quantified in Western blots, as illustrated in Figure 4 (U values).

3-NP-Induced Expression of Antioxidants

Adaptive changes of antioxidant levels during the first 60 hours post-3-NP appeared less prominent in PNCs (dashed columns) than in NSCs (white columns). In neither of the evaluated enzymes did the PNCs show any significant upregulation compared with their U control values, except in the case of UCP2 after 12 hours (p < .05; Fig. 4A) and some increase in SOD2 at 60 hours (p < .06; Fig. 4E) in primary PNCs.

In contrast to PNCs, NSCs responded to 3-NP with considerable upregulation of the enzymes UCP2 and SOD2 in primary (Fig. 4A, 4E, white columns) and GPx in immortalized cells (Fig. 4D, white columns). Thus, in comparison to the steady-state levels (U values), 3-NP exposure could further increase existing enzymatic differences between NSCs and PNCs (Fig. 4a–4d) and induced molecules such as SOD2, which did not show any differences at steady-state levels (Fig. 4E).

Inhibition of Redox Modulators in NSCs and Its Effects on Their Response to Oxidative Stress

In NSCs, the antioxidative enzymes UCP2 and GPx appeared to play a major role in steady-state ROS metabolism and, together with SOD2, were the ones to be most prominently upregulated in their response to oxidative stress. To test the direct functional relationship between the changing levels of these enzymes and the resistance of NSCs to 3-NP, we inhibited the activity of these molecules with their inhibitors GDP, MS, and DETC, respectively. The resulting cellular behavior (Fig. 5, black lines) was then compared with that of noninhibited NSCs and PNCs (Fig. 5, gray dashed and solid lines). In simultaneously processed 3-NP-free controls, the application of inhibitors alone did not produce any toxic effects (not shown).

Figure Figure 5..

Inhibition of redox modulators in NSCs and its effects on their response to oxidative stress. A pharmacological block of the redox modulators UCP2, GPx, and SOD2 by guanosine 5′-diphosphate, mercaptosuccinic acid, and dethylthiocarbamate, respectively, led to a drastically increased sensitivity of the proliferating NSCs to the detrimental effects of 3-NP. Inhibition of either individual antioxidants (A–C) or all three of them simultaneously (D) occurred. In each of the four panels, the obtained data from the inhibited NSCs (solid black line) pertaining to their mitochondrial viability (a, b), intracellular ROS levels (c, d), and number of apoptotic cells (e, f) were compared with 1) noninhibited NSCs (gray dashed line), and 2) noninhibited PNCs (gray solid line). Although inhibition of UCP2 (A) and GPx (B) resulted in a significant (p < .05) oxidative damage of the NSCs and their decreased viability, the block of SOD2 remained without obvious effects (C). The simultaneous inhibition of all three antioxidants led, on the other hand, to a marked (p < .001) deterioration of NSC behavior and approached it to the values of the noninhibited PNCs exposed to 3-NP. See main text for more details. Abbreviations: CM-H2DCFDA, 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate; GPx, glutathione peroxidase; MTT, 4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; NSC, neural stem cell; PNC, postmitotic neural cell; SOD2, superoxide dismutase 2; UCP2, uncoupling protein 2.

After inhibition of UCP2 and GPx, the viability of primary and immortalized NSCs exposed to 3-NP was strongly reduced, whereas a block in SOD2 activity, interestingly, did not produce any effects. In contrast to noninhibited NSCs (Fig. 5, gray dashed line), the ones deprived of UCP2 or GPx activity (black solid lines) were unable to recover from the neurotoxic effects, and they continued to deteriorate throughout the experimental period of 5 days. Not surprisingly, blocking of all three tested enzymes simultaneously led to the most severe deterioration of NSC viability, approaching that of the noninhibited PNCs (Fig. 5, compare black and gray solid lines).

Expression of Redox Modulators in the SVZ of the Mouse Brain

In the in vitro studies described so far, we found that in their ROS household and tolerance of oxidative stress, NSCs are characterized by more active antioxidant defense mechanisms than more differentiated cell types. Since the largest steady-state differences were found in the expression levels of UCP2 and GPx (U values in Fig. 4A–4D), we examined immunohistochemically the presence of both proteins in the germinative zones of the adult mouse brain [22] at coronal levels 25–30 according to reference 20.

Whereas cells positive for both tested antioxidant molecules were dispersed throughout the brain parenchyma, strongly stained cells were predominantly localized within the anterolateral portion of the SVZ (approximately 5% of the resident cells; Fig. 6J) and in the subgranular zone of the dentate gyrus (not shown). Both regions represent the two main stem cell niches of the forebrain [23].

To characterize these cells, we used the proliferation marker Ki-67 (Fig. 6G–6I) and the stem cell marker Musashi (Fig. 6A–6F, 6D1–6F1) [24]. Cells positive for these markers were consistently characterized by strong expression of UCP2 and GPx.

Figure Figure 6..

Expression of redox modulators in the subventricular zone (SVZ) of the mouse brain. Coronal, 20-μm cryostat sections of adult mouse forebrains were immunostained to reveal expression of UCP2 and GPx in neural cells. NSCs were monitored for the presence of the stem cell marker Musashi (A–F, D1–F1) and the proliferation marker Ki-67 (G–I). Cells maintaining significant levels of UCP2 and GPx were found mainly within the anterolateral regions of the forebrain SVZ and only rarely within the brain parenchyma ([J], white crosses). Their location overlapped with that of NSCs labeled with Mus and Ki-67. See main text for more detail. (A–I): Scale bars = 20 μm; (D1–F1): scale bars = 10 μm. Abbreviations: CC, corpus callosum; CP, caudoputamen; GPx, glutathione peroxidase; LV, lateral ventricle; M, midsagittal line; Mus, Musashi; UCP2, uncoupling protein 2.


Oxidative stress due to excessive presence of ROS is a permanent threat to any cell with aerobic metabolism and accompanies many traumatic CNS pathologies and diseases such as Parkinson's disease and Huntington's disease. Particularly in the brain, where 20% of the oxygen consumed by the body is used, cells have to rely on a variety of potent antioxidant defense mechanisms and a close interaction of glial and neuronal cells for constant removal of ROS [12, 25, [26]–27]. Interestingly, although extensive studies have been performed on the mature cell types populating the CNS, almost no data exist about the effects of oxidative stress on NSCs and their postmitotic progeny. In light of the present hopes to use these cells for therapeutic transplantation and/or take advantage of their endogenous counterparts, the investigation of this issue has become an important necessity, and the present report is a first step in this direction.

To be able to realize their therapeutic purpose, grafted and endogenous NSCs have to resist the hostile pathological microenvironment as much as or even better than the other host cell populations. In the present study, we addressed this possibility by comparing NSCs and their PNC counterparts with respect to their steady-state ROS household, cell viability, expression levels of several key proteins with antioxidative functions, and their changes in response to 3-NP-induced oxidative stress.

By interfering with the mitochondrial electron transfer chain and leading to energy depletion, 3-NP exposure results in increased production of ROS such as superoxide (O2) and hydrogen peroxide (H2O2) [28]. To address both mitochondrial and cytoplasmic antioxidant detoxifier mechanisms controlling intracellular accumulation of ROS, we investigated the expression of the redox modulators UCP2, GPx, SOD2, and CAT.

UCP2 belongs to the family of mitochondrial H+ transporters that regulate oxidative phosphorylation by increasing the uncoupling of electron transport and ATP formation [8]. Recently, UCP2 has been shown to be induced by acute brain injury such as stroke and to participate in neuroprotection and cellular rescue during 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-mediated neuronal damage [13, 29, 30], which made us include this redox modulator in our investigations. The other three antioxidant enzymes, GPx, SOD2, and CAT, are well-characterized scavengers of free radicals; SOD2 is activated by high levels of superoxide and converts it to H2O2, whereas the cytoplasmic GPx and CAT convert the latter to H2O [27].

According to recent literature, it appears that the transcriptome of hematopoietic, neural, and embryonic stem cells includes a subset of genes, the products of which not only help to define “stemness” but can also provide the cells with higher resistance against oxidative stress [31, 32]. Moreover, the intracellular redox state of NSCs also appears to regulate the balance of the cells between their self-renewal, cell differentiation, and apoptosis [9] and may therefore require special regulatory attention. The outcome of the present study is a concrete example of NSC behavior corroborating the first of these suggestions.

Steady-State Characteristics

Cultured NSCs were found to be by default better equipped to control intracellular ROS than PNCs. This was reflected in their significantly higher mitochondrial activity, lower ROS content, reduced apoptosis, and their comparatively higher basal levels of UCP2 and GPx. It has been reported that proliferating cells can actually produce substantial amounts of ROS endogenously [33]. Our results suggest that the maintenance of higher steady-state levels of UCP2 and GPx in stem cells is one of their strategies to offset this effect.

Response to 3-NP Treatment

We observed that NSCs resisted evoked oxidative stress better than the PNCs and were able to recover from its effects within 5 days after the insult. In agreement with these findings, NSCs increased substantially their levels of antioxidant proteins, whereas the latter remained almost unchanged in PNCs. The disparate response of both cell types to oxidative stress suggests, at least in part, an overall higher steady-state “vigilance” of redox control mechanisms in NSCs. Their stem cell characteristics, including proliferation, appear to be associated with a higher “alertness” of proteins such as UCP2 or GPx to changes in intracellular ROS levels. Both proteins can decrease the probability of mitochondrial permeability transition [34], the release of apoptogenic factors [35], and the activation of caspase-3-mediated cell death [36]. GPx has also been shown to protect against 3-NP-induced apoptosis [37]. Hence, a possible interaction of apoptosis-modulating proteins, such as the members of the Bcl family, with antioxidant proteins such as UCP2 or GPx and the ensuing suppression of apoptotic cell death can be one way for cells to counteract the effects of oxidative stress. The fact that we did not see any increase in CAT can be attributed to the role of GPx as the major source of protection against low levels of ROS and its greater affinity for hydrogen peroxide than CAT [27].

We propose that, intriguingly, a molecular network helping NSCs maintain stem cell status [9, 31, 32] may also be responsible for enzymatic “priming” in the control of cellular homeostasis. In the case of UCP2, GPx, and SOD2, this fact appeared to allow the NSCs to react more quickly and more efficiently to external insults related to oxidative stress than PNCs, which were marked by weaker expression and response of all three redox modulators. A similar principle, however, may also apply for other stem cell features, helping NSCs to survive and interact with a pathological environment. Indeed, it has been recently reported that NSCs, by definition lacking differentiation markers and major histocompatibility complex molecules on their surface, show low immunogenicity and thus escape to a great extent the immune response of the brain [38, 39].

In the present study, interestingly, primary and immortalized NSCs displayed some minor behavioral differences: in primary NSCs, the contents in UCP2 and SOD2 increased upon exposure to 3-NP, although both remained unchanged in the immortalized NSCs. The latter, on the other hand, upregulated their GPx activity instead. It has been documented that overexpression of the myc gene can cause elevated ROS production in cells and can downregulate NF-κB-mediated expression of redox regulators such as SOD2 [40, 41]. Such myc-dependent interference with the expression of antioxidants could explain the slightly worse values pertaining to cell viability and ROS household in immortalized NSCs compared with those recorded from primary NSCs, both in their steady-state behavior and after a challenge with 3-NP. It could also play a role in the differential, 3-NP-induced expression patterns of the three tested redox modulators between primary and immortalized NSCs.

Irrespective of the induction patterns of antioxidants, both primary and immortalized NSCs, in comparison with PNCs, showed an equally high resistance to oxidative stress, even if their levels of the tested antioxidants differed. This indicates that other existing defense mechanisms might help NSCs, in general, to stabilize their redox states. Examples of such mechanisms that we are currently investigating are enhanced adaptive controls of nitric oxide (NO) production and thiol pools due to proliferation- and myc-mediated permanently elevated ROS levels [42, [43]–44].

Inhibition of the Antioxidant Response to 3-NP

To test the importance of the 3-NP-induced enzymes UCP2, GPx, and SOD2 for NSC survival, we inhibited their activity before and during the exposure of the cells to the mitochondrial toxin. The blocking of either UCP2 or GPx resulted in a substantial reduction in NSC viability and raised ROS levels and numbers of apoptotic cells. When all three enzymes were inhibited simultaneously, the NSCs died with rapidly increasing ROS levels and decreasing MTT values similarly to PNCs. Thus, although we can assume that additional signaling pathways contribute to the resistance of NSCs to oxidative stress, the cumulative control of membrane potential and efficient ROS clearance appear crucial for their survival and enhanced tolerance.

We found that the inhibition of SOD2 alone did not have any significant effects on the viability of NSCs. Although the steady-state levels of this enzyme were comparable between NSCs and PNCs, it was quickly and strongly induced in primary NSCs responding to 3-NP. No changes were observed in the case of immortalized NSCs, which could have been, as discussed before, the consequence of the inhibitory influence of myc overexpression in these cells. These results suggest that SOD2 is not involved in ROS control during steady-state homeostasis in these cells and that other additional antioxidant mechanisms may have compensated for its inhibition during the stress response. Also, UCP2, which is known to be stimulated by mitochondrial superoxide, may have played an important role in quenching superoxide-induced oxidative stress in the NSCs [45].

Steady-State Expression of Antioxidants in NSCs In Vivo

Stem cells represent an important source of building material in both development and regeneration of the CNS and may very well take advantage of the “stemness-defining,” enhanced control of their redox state to be better prepared than the rest of the brain for the arrival of oxidative stress. This we found confirmed in vitro, where NSCs were characterized by more active antioxidant defense mechanism than their PNC counterparts. Since the largest differences in steady-state levels of antioxidant enzymes between NSCs and PNCs were those of UCP2 and GPx, we performed a pilot study of the expression of both enzymes in the stem cell niche [23] of adult mouse forebrain. Intriguingly, we found that endogenous NSCs localized in the SVZ and the subgranular zone of the dentate gyrus are also characterized by higher steady-state levels of UCP2 and GPx compared with neural cells in other brain areas. This finding, being in harmony with our in vitro results, adds a new and important characteristic to the already special nature of neural stem cell niches: they represent a source of cells that, by the nature of their stem cell transcriptome, are relatively well equipped to resist oxidative stress and, perhaps, other forms of pathological insults as well. A better understanding of such stem cell features is therefore likely to have an important impact on our views of CNS development, activation of endogenous NSCs during injury, and NSC transplantation.


The authors indicate no potential conflicts of interest.


We thank Dr. Petr Ježek for helpful suggestions and comments and Nada Pavlović for her technical help with immunostaining. This work was supported by intramural funding from Iowa State University (J.O., V.O.). V.O. and J.O. contributed equally to this work.