Nitric oxide (NO) inhibits proliferation of subventricular zone (SVZ) neural precursor cells in adult mice in vivo under physiological conditions. The mechanisms underlying this NO effect have now been investigated using SVZ-derived neural stem cells, which generate neurospheres in vitro when stimulated by epidermal growth factor (EGF). In these cultures, NO donors decreased the number of newly formed neurospheres as well as their size, which indicates that NO was acting on the neurosphere-forming neural stem cells and the daughter neural progenitors. The effect of NO was cytostatic, not proapoptotic, and did not involve cGMP synthesis. Neurosphere cells expressed the neuronal and endothelial isoforms of NO synthase (NOS) and produced NO in culture. Inhibition of NOS activity by Nω-nitro-l-arginine methylester (l-NAME) promoted neurosphere formation and growth, thus revealing an autocrine/paracrine action of NO on the neural precursor cells. Both exogenous and endogenous NO impaired the EGF-induced activation of the EGF receptor (EGFR) tyrosine kinase and prevented the EGF-induced Akt phosphorylation in neurosphere cells. Inhibition of the phosphoinositide-3-kinase (PI3-K)/Akt pathway by LY294002 significantly reduced the number of newly formed neurospheres, which indicates that this is an essential pathway for neural stem cell self-renewal. Chronic administration of l-NAME to adult mice enhanced phospho-Akt staining in the SVZ and reduced nuclear p27Kip1 in the SVZ and olfactory bulb. The inhibition of EGFR and PI3-K pathway by NO explains, at least in part, its antimitotic effect on neurosphere cells and may be a mechanism involved in the physiological role of NO as a negative regulator of SVZ neurogenesis in adult mice.
The capacity to generate new neurons from stem cells is preserved in the subventricular zone (SVZ) of the adult rodent brain. Quiescent astrocytes (B cells), identified as the SVZ stem cells , become activated and generate rapidly dividing transit-amplifying C cells, which are the precursors of neuroblasts [1, –3]. These neuroblasts migrate rostrally and differentiate as inhibitory interneurons in the olfactory bulb [4, 5]. In vitro, the transit-amplifying C cells retain stem cell properties, in terms of multipotentiality and self-renewal , and generate neurospheres when cultured in the presence of the epidermal growth factor (EGF) . Intracerebroventricular administration of EGF produces an expansion of SVZ dividing cells [7, 8], and mice null for transforming growth factor-α, another endogenous ligand of the EGF receptor (EGFR), exhibit reduced neurogenesis . These observations indicate that EGFR activation is a crucial step in SVZ precursor proliferation and that EGFR may therefore be a molecular target for physiological regulators of neurogenesis in the SVZ.
Nitric oxide (NO) is an intercellular messenger in the nervous system . NO exerts antiproliferative effects on several tumoral cell lines as well as on embryonic cells of neural origin in vitro and facilitates cell differentiation [11, 12]. It was recently reported that both intracerebral and systemic administration of NO synthase (NOS) inhibitors significantly enhance neurogenesis in the SVZ of adult mice and rats [12, –14]. However, it is not clear whether these in vivo results are due to a direct cytostatic action of NO on the SVZ neural precursors or whether they are an indirect consequence of changes in cerebral blood flow or synaptic activity produced by NOS inhibition. A previous report using primary cultures of SVZ explants  pointed to a direct effect of NO on SVZ cells, although the specific cellular target was not identified. The mechanism involved in the NO antineurogenic action is also unknown at present. Based on our previous finding that NO inhibits the EGFR tyrosine kinase in neuroblastoma cells, we hypothesized that NO may participate in the control of neural precursor proliferation by modulating their EGFR activity .
We demonstrate here that both exogenous and endogenous NO directly inhibit the proliferation of EGF-responsive multipotent precursor cells, with a concomitant inhibition of their EGFR tyrosine kinase and subsequent reduction of Akt phosphorylation. In agreement with these in vitro data, chronic NOS inhibition increased phospho-Akt and reduced p27Kip1 in the adult mouse SVZ. Together, the results point to the EGFR-PI3K/Akt-p27Kip1 pathway as the mechanism of action of NO as a physiological regulator of SVZ neurogenesis.
Materials and Methods
Postnatal (P7) CD1 mice were used throughout this study. Care and handling of animals were performed in accordance with the Guidelines of the European Union Council (86/609/EU) and Spanish regulations (BOE 67/8509-12) for the use of laboratory animals.
The following products were used: diethylamine/nitric oxide (DEA/NO) and diethylenetriamine/nitric oxide (DETA/NO) adducts, S-nitroso-N-acetylpenicillamine (SNAP), 8Br-cGMP, Nω-nitro-l-arginine methyl ester (l-NAME), and 5-bromo-2′-deoxyuridine (BrdU) from Sigma-Aldrich (St. Louis, http://www.sigmaaldrich.com); EGF (mouse recombinant) from Upstate Biotechnology (Lake Placid, NY, http://www.upstatebiotech.com); 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4–1 (LY294002) and 1,4-diamino-2,3-dicyano-1,4-bis(2-aminophynyltio)butadiene (U0126) from Cell Signaling Technology, Inc. (Danvers, MA, http://www.cellsignal.com); 1H-[1,2,4]oxadiazol[4,3-a]quinoxalin-1-one) (ODQ) from Alexis Laboratories (San Diego, http://www.axxora.com); 4,5-diaminofluorescein diacetate ester (DAF-2 DA) from Calbiochem (San Diego, http://www.emdbiosciences.com); culture media, saline solutions, glutamine, fetal bovine serum, trypsin/EDTA, gentamicin, and bovine serum albumin from Invitrogen (Carlsbad, CA, http://www.invitrogen.com); trichloroacetic acid (TCA), glycerol, inorganic salts, and concentrated acids, bases, and alcohols from Merck (Darmstadt, Germany, http://www.merck.com); and SDS, acrylamide/bis-acrylamide solutions, and BioRad protein assay kits from Bio-Rad (Hercules, CA, http://www.bio-rad.com).
SVZ Cell Isolation and Culture
Neurospheres were obtained from P7 mice SVZ and maintained in culture essentially as reported [1, 3]. The lateral walls of the lateral ventricles were removed and enzymatically dissociated in aCSF-Ca2+low (5 mM KCl, 124 mM NaCl, 3.2 mM MgCl2, 100 μM CaCl2, 26 mM NaHCO3, and 10 mM glucose) containing 1 mg/ml trypsin and 0.2 mg/ml kinurenic acid at 37°C for 15 minutes. The tissue was centrifuged at 150g for 5 minutes, rinsed in aCSF (5 mM KCl, 124 mM NaCl, 1.3 mM MgCl2, 2 mM CaCl2, 26 mM NaHCO3, and 10 mM glucose), and centrifuged again in the same conditions. Then, the cells were resuspended in Dulbecco's modified Eagle's medium (DMEM)/F-12 (1:1) containing 0.7 mg/ml ovomucoid and mechanically disaggregated with a fire-polished Pasteur pipette. The dissociated cells were centrifuged, resuspended in defined medium (DM) composed of DMEM/F-12 (1:1), 100 μg/ml transferrine, 30 nM sodium selenite, 60 μM putrescine, 20 nM progesterone, 25 μg/ml insulin, 2 mM glutamine, 33 mM glucose, 5 μg/ml gentamicine, 5 mM HEPES, 20 ng/ml EGF, and 0.125 μg/ml fungizone, and maintained in an atmosphere of 5% CO2, at 37°C. After 1–2 days, cell aggregates known as neurospheres were formed. Subcultures were performed every 4–5 days by centrifugation of the neurospheres and mechanical dissociation of the cells in 1 ml of DM; then, the single-cell suspension was replated in new culture flasks in fresh medium to obtain new neurospheres. Experiments were performed between passages 4 and 20.
For those experiments that required quantitative image analysis, neurosphere cells were grown as a monolayer. For that, neurospheres were mechanically dissociated as above, and cells were seeded on a poly(ornithine) substrate in DF12 with 1% fetal calf serum for 4 hours to facilitate adhesion. Then, cells were washed and maintained in DM for 48 hours, unless otherwise indicated. These cells, which will be referred to as “adhered cells” in the text, maintained properties of undifferentiated cells, as assessed by nestin staining, and responded to NO drugs in a manner similar to the floating neurospheres. We also used the human neuroblastoma cell line NB69 for comparison purposes in some cases. The culture conditions and the functional properties of these cells have been described in detail elsewhere .
Cell Proliferation Assays
Neurospheres were centrifuged, and the cells were resuspended in DM and seeded at a density of 105 cells per milliliter in 96-well plates. This relatively high density was selected to allow cell communication by autocrine factors, which may be essential in a physiological process such as proliferation control. Drugs under study were added at the time of seeding. The number of newly formed neurospheres per well was counted under phase microscopy 48 hours later. To measure the neurosphere size, images of five fields per well were obtained, and the volume of the newly formed neurospheres (at least 30 neurospheres per condition) was estimated using the Microimage analysis system from Olympus (Tokyo, http://www.olympus-global.com). To measure proliferation in adhered cells, drugs were added immediately after the 4-hour adhesion period, when cells were exposed to EGF; at the end of the treatment period, cells were trypsinized and counted using a Neubauer chamber. Alternatively, cells were exposed to BrdU for the last 8 hours, fixed and immunostained using a specific anti-BrdU antibody.
Programmed Cell Death
Adhered cells grown on coverslips were exposed to the different treatments. Eight and 24 hours later, cells were fixed with 4% paraformaldehyde in 0.1 M phosphate buffer for 10 minutes. The detection of apoptotic cells was performed using the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) staining method and a commercial kit (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) in accordance with the standard protocol provided by the supplier. Apoptotic cells (fluorescein-12-dUTP DNA-labeled cells) were counted directly under a fluorescence microscope (BX-60; Olympus) and were expressed as percentage of the total number of cells, which were counterstained with DAPI (4,6-diamidino-2-phenylindole). A total of 25 fields were counted per coverslip.
The production of NO in living cell cultures was estimated using the NO-sensitive fluorescent dye DAF-2 [17, 18], which reacts with NO in the presence of oxygen to form the highly fluorescent triazolofluorescein (DAF-2T). We used a membrane-permeable form of the dye, DAF-2 DA, which can be taken up by the cells and hydrolyzed by cellular esterases to again form the membrane-impermeable compound DAF-2. Cells grown on coverslips for 48 hours were treated or not treated with the NOS inhibitor l-NAME (300 μM, overnight) and loaded with 10 μM DAF-2 DA (Calbiochem) for 30 minutes in the dark at room temperature. Some untreated coverslips received 1 mM DEA/NO for the last 2 minutes. Measurements of relative NO concentrations were performed using an inverted epifluorescence microscope (Olympus) with a ×40 objective lens. Fluorescence images of the cells were captured using a DP50 digital camera (Olympus); in all cases, the same camera setting conditions, with an exposure time of 200 ms, were used. To avoid photobleaching, coverslips were maintained in the dark, and each field was exposed to the excitation wavelength only during the capture time. Quantification of the mean signal intensity in individual cells was performed using the software Microimage from Olympus. A total of at least 25 individual cells from five to seven video frames were analyzed for each condition.
Cells from either floating neurospheres or adhered cultures were lysed with ice-cold lysis buffer (50 mM Tris/HCl, pH 7.4, 1% [vol/vol] Triton X-100, 0.5% [wt/vol] sodium deoxycholate, 1 mM phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 10 μg/ml pepstatin A, and 10 μg/ml aprotinin). Supernatants were collected by centrifugation (16,000g), and their protein concentration was measured by the Bradford-based Bio-Rad microassay method. Equal amounts of total protein from each cellular extract were separated by a 7% SDS-PAGE and transferred onto a polyvinylidene difluoride (PVDF) membrane. Immunodetection was carried out using a 1:5,000 dilution of rabbit polyclonal antibodies against neuronal NOS (nNOS), kindly provided by Dr. J. Rodrigo (CSIC, Madrid, Spain) , and endothelial NOS (eNOS) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, http://www.scbt.com). Thereafter, the membranes were sequentially incubated for 90 minutes with a biotin-conjugated anti-rabbit antibody (1:10,000; Sigma-Aldrich) for 30 minutes with the ABC kit and for 2 minutes with the ECL solution. Bands were made visible by exposing the membranes to light-sensitive films.
Detection of Tyrosine Phosphorylation and Phosphorylated Proteins
Neurospheres were collected by centrifugation, and cells were mechanically dissociated in DM without EGF and distributed homogeneously on a six-well culture plate (approximately 5 × 105 cells per well). After 1 hour in the absence of growth factors, cells were treated with or without an NO donor (DEA/NO or SNAP, 0.1–2 mM) for 5–15 minutes and then were exposed to 20 ng/ml EGF for 5–30 minutes. Controls in the absence of EGF were also included. NOS inhibition was accomplished by addition of l-NAME (300 μM) at the seeding time. The reaction was stopped with ice-cold 10% (wt/vol) TCA. Adhered cells were preincubated for 1 hour in growth factor-free DM, treated as above, and scraped to detach them from the culture dish surface.
Western blots were performed as described elsewhere . Briefly, cellular precipitates were collected by centrifugation (16,000g, 15 minutes) and each pellet was dissolved in 50 μl of loading buffer (0.02% [wt/vol] bromophenol blue, 5% [vol/vol] 2-mercaptoethanol, 20% [wt/vol] glycerol, 60 mM Tris/HCl, pH 6.8, and 10% [wt/vol] SDS). After boiling for 5 minutes, proteins were separated by SDS-PAGE (5%–20% linear-gradient polyacrylamide gels for EGFR tyrosine phosphorylation and 10% gels otherwise), transferred to a PVDF membrane, and stained with the dye Fast Green FCF to register the total protein content in each track. Thereafter, immunodetection was carried out using the following primary antibodies: peroxidase-conjugated anti-phosphotyrosine monoclonal antibody (1:2,000; Transduction Laboratories, Lexington, KY, http://www.bdbiosciences.com), anti-EGFR and extracellular signal-regulated kinase (ERK) 1/2 (1:1,000; Santa Cruz Biotechnology, Inc.), pp-ERK (p44/p42, Thr202/Tyr204), pp-Akt (Ser437), and Akt (1:1,000; Cell Signaling Technology). After washing with T-TBS (Tris-buffered salt solution with Tween), the membrane was incubated, when necessary, with the secondary antibody and processed as described above for NOS isoforms. Quantification was done by photodensitometry with a computer-assisted scanner and the Microimage software. The optical densities measured were normalized considering the total protein loaded, as assessed by Fast Green staining, although only minor differences in protein content were detected.
RNA Isolation and Reverse Transcription-Polymerase Chain Reaction
Total RNA was isolated from floating neurospheres and adhered cells using a high pure RNA isolation kit according to manufacturer instructions (Roche). For reverse transcription-polymerase chain reaction (RT-PCR), 1 ìg of total RNA was reverse-transcribed using first-strand cDNA synthesis kit (Roche) with random hexamers. One-thirtieth of the cDNA obtained was amplified using Faststart Taq DNA polymerase (Roche). Primer sets used for amplifying specific cDNA were designed in different exons as follows: for nNOS forward (5′-GGA ACC CGA CAG GCC AAA GAA ATA-3′) and reverse (5′-TCC TCG TGG TAC CGG TTG TCA TCC-3′), yielding a 237-base pair product size; for eNOS forward (5′-ATA TGT TTG TCT GCG GCG ATG TCA-3′) and reverse (5′-CTC TGG GTG CGT ATG CGG CTT GTC-3′), yielding a product size of 202 bp, for inducible NOS (iNOS) forward (5′-CCG CAC CAC CCT CCT CGT TC-3′) and reverse (5′-GGG GGC AGC CTC TTG TCT TTG A-3′), yielding a product size of 216 bp, and for β-actin forward (5′-ATC GTG CGT GAC ATC AAA GAG AAG-3′) and reverse (5′-CAG CAC TGT GTT GGC ATA GAG GTC-3′), yielding a product size of 276 bp. After a 10-minute activation period at 95°C, 45 cycles of 95°C for 10 seconds, 60°C for 5 seconds and 72°C for 10 seconds were performed using the Lightcycler instrument (Roche). The size and amount of PCR products were verified by electrophoresis in 1.5% agarose gels containing ethidium bromide and photographed under UV light illumination.
For immunocytochemical studies, neurospheres were adhered onto coverslips and fixed with 4% (wt/vol) paraformaldehyde in 0.1 M phosphate buffer. Then, they were incubated for 30 minutes in 2.5% (wt/vol) bovine serum albumin (BSA) in phosphate-buffered saline (PBS) and with primary (16 hours at 4°C) and secondary (1 hour at room temperature) antibodies. After washing, the coverslips were mounted on slides with Vectashield and fluorescent signals were detected using a BX60 Olympus epifluorescence microscope and a Leica Spectra confocal microscope. For BrdU detection, cells were first subjected to DNA denaturation by treatment with 0.07 N NaOH at room temperature for 2 minutes and then processed for immunocytochemistry. The primary antibodies used were glial fibrillary acidic protein (GFAP) (rabbit polyclonal, 1:500), BrdU (mouse monoclonal, 1:100) from Dako Denmark A/S (Glostrup, Denmark, dako.dkhttp://www.dako.dk), doublecortin (goat polyclonal, 1:100; Santa Cruz Biotechnology, Inc.), β-III-tubulin (mouse monoclonal, 1:1,000; Promega Corporation, Madison, WI, http://www.promega.com), EGFR (sheep polyclonal, 1:200; Upstate Biotechnology), nNOS (rabbit polyclonal, 1:5,000; generous gift from Dr. J. Rodrigo), eNOS (rabbit polyclonal, 1:100; Santa Cruz Biotechnology, Inc.), and nestin (rabbit polyclonal, 1:5,000; generous gift from Dr. M. Vallejo (CSIC), and goat polyclonal, 1:100; Santa Cruz Biotechnology, Inc.). To detect oligodendrocyte precursors, living cells were incubated for 15 minutes with anti-A2B5 (mouse immunoglobulin M [IgM], 1:20; generous gift from Dr. C. Guaza), washed, and further incubated for 15 minutes with cyanine 3 (Cy3)-labeled anti-mouse IgM. Cells were then fixed and processed for triple labeling with the corresponding antibodies. The secondary antibodies used were anti-rabbit IgG labeled with fluorescein isothiocyanate (FITC) or Cy5, anti-goat IgG labeled with Cy5, and anti-sheep IgG labeled with Cy3, all of which were purchased from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA, http://www.jacksonimmuno.com), and anti-mouse IgG labeled with Cy3 (GE Healthcare, Little Chalfont, Buckinghamshire, U.K., http://www.gehealthcare.com) or FITC (Sigma-Aldrich). All of the secondary antibodies were adsorbed against several species to prevent undesired cross-reactions. Omission of primary antibodies resulted in no detectable staining in all cases.
In Vivo Experiments
Adult CD1 male mice (2–4 months old) were injected with l-NAME (90 mg/kg per day, i.p.) for 15 days, a condition that results in reduction of cerebral NOS activity and increase of neural precursor proliferation in the SVZ . At the end of the treatment period, mice injected with l-NAME or vehicle (PBS) were deeply anesthetized with chloral hydrate (0.5 g/kg, i.p.) and transcardially perfused with 4% (wt/vol) paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. The brains were removed, postfixed for 2 additional hours, and cryoprotected by immersion in 30% sucrose solution overnight. Coronal sections (30 μm thick) of the SVZ and olfactory bulb were obtained using a cryostat and stored at −20°C in a cryoprotectant solution (glycerol and PBS, pH 7.4, 1:1 in volume). For phospho-Akt and EGFR double-immunolabeling, free-floating sections were incubated for 30 minutes with a solution containing 2.5% (wt/vol) BSA, 0.25% (wt/vol) sodium azide, and 0.1% (vol/vol) Triton X-100 in PBS (PAAT) overnight with the primary antibodies (rabbit polyclonal anti-phospho Akt, 1:100 (Cell Signaling Technology) and sheep polyclonal anti-EGFR, 1:200), and for 2 hours with the corresponding fluorescence-labeled secondary antibodies. The SVZ from control and treated mice were visualized by confocal microscopy, and their relative phospho-Akt staining was estimated by comparison of the relative mean fluorescence intensity using the Leica confocal analysis software LCS Lite (Leica, Heerbrugg, Switzerland, http://www.leica.com).
For p27Kip1 detection, brain sections through the SVZ and olfactory bulb were first treated with 2% H2O2 and 60% (vol/vol) methanol in PBS for 30 minutes to block endogenous peroxidase activity. After overnight incubation with rabbit anti-p27Kip1 (1:100; Santa Cruz Biotechnology, Inc.) and 2-hour incubation with a biotinylated anti-rabbit IgG secondary antibody (1:250; Sigma), the tissue was exposed to the avidine-biotin-peroxidase complex (Pierce Chemical, Rockford, IL, http://www.piercenet.com). The peroxidase reaction was made visible with diaminobenzidine (DAB) (0.25 mg/ml) and hydrogen peroxide (0.003%, vol/vol). Sections were then mounted on slides, dehydrated, coverslipped with DePeX, and analyzed under light microscopy.
Data are presented as the mean ± SEM of values obtained from three or more experiments. Comparisons between values obtained in control and treated samples were analyzed using the Student's t test. Differences were considered significant when p < .05.
Characterization of Cell Cultures
Postnatal mouse SVZ progenitors isolated and exposed to EGF divided and formed neurospheres, as previously described . After multiple (4–20) subcultures, part of the neurosphere cells (1 in 22.7 ± 1.2 seeded cells; n = 15 independent cultures) still generated new neurospheres with similar characteristics, which indicates the existence of cells with self-renewal properties. Neurospheres were constituted by nestin-positive precursor cells (Fig. 1A), and most of them also contained a small number of daughter cells that expressed either glial (GFAP or A2B5) or early neuronal (β-III-tubulin or doublecortin) markers (Fig. 1B; Table 1), indicating that the neurosphere-forming cells were multipotent. EGFR was expressed by neurosphere cells (Fig. 1C); confocal analysis revealed the presence of this receptor in all nestin-positive (Fig. 1D–1F) and in some GFAP-positive (Fig. 1G–1I) neurosphere cells but not in cells expressing β-III-tubulin (Fig. 1J–1L). Adhered cells (Fig. 1M) presented a precursor-like phenotype, with most of them expressing nestin (Fig. 1N) and EGFR (Fig. 1O).
Table Table 1.. Neurosphere phenotypic composition in untreated cultures and in cultures treated with an nitric-oxide synthase (NOS) inhibitor or an NO donor
SVZ-Derived Cells Expressed the Neuronal and Endothelial Isoforms of NOS and Produced NO in Culture
Analysis of mRNA by RT-PCR indicated that the neuronal and endothelial isoforms of NOS were expressed in both floating neurospheres and adhered cells (Fig. 2A). The presence of the two proteins was detected by immunoblotting, using specific antibodies for each isoform, in cell lysates from both types of culture (Fig. 2B). Immunocytochemical detection of both NOS isoforms did indeed occur in neural precursors, which also expressed nestin (Fig. 2C, 2D).
Spontaneous endogenous production of NO was demonstrated by measuring intracellular DAF-2 fluorescence in untreated cultures. The fluorescence intensity was significantly reduced when cultures were pretreated with l-NAME and was enhanced in the presence of the short half-life NO donor DEA/NO (Fig. 2E–2G). The mean fluorescence intensity detected in individual cells, which is indicative of intracellular NO concentration, was quantified and expressed in arbitrary units for comparison (Fig. 2H, 2I).
Exogenous and Endogenous NO Inhibited Proliferation of SVZ Cells in Culture
Addition of the long half-life NO donor DETA/NO to neurosphere cell suspensions produced a concentration-dependent reduction in the number (Fig. 3A, 3I, 3J) and size (Fig. 3B, 3I, 3J) of the new neurospheres formed 48 hours later. In these cultures, the number of newly formed neurospheres is indicative of the number of stem-like cells that enter the cell cycle and start dividing, whereas the neurosphere size gives information about the mitotic rate of the daughter cells that constitute the neurosphere (stem-like cells and committed progenitors). DETA/NO also decreased proliferation in adhered cells, as estimated by cell counting (Fig. 3C) and analysis of BrdU incorporation (Fig. 3D).
Addition of the NOS inhibitor l-NAME to floating neurospheres produced the opposite effect: that is, a significant increase in both neurosphere number (Fig. 3E, 3J, 3K) and size (Fig. 3F, 3J, 3K). The proliferative action of l-NAME was also demonstrated in adhered cells, as shown by the increase in cell number and in the percentage of cells incorporating BrdU (Fig. 3G, 3H). When cultures were exposed to either the NO donor or the NOS inhibitor, a large majority of neurospheres still contained mixed phenotypes, as occurred in control conditions (Table 1), thus suggesting that NO was acting preferentially on the multipotent and self-renewing precursor cells and not on committed progenitors.
NO Did Not Affect Apoptosis of SVZ Cells in Culture
Analysis of apoptosis by the TUNEL technique indicated that, 8 hours after seeding the cells, 11.7% ± 6% (n = 3) of the cell nuclei presented signs of DNA fragmentation. The proportion of apoptotic nuclei was not modified when 300 μM l-NAME or 60 μM DETA/NO was added to the cultures at the seeding time (9.0% ± 1%; n = 2 and 10.1% ± 3%; n = 3, respectively). Similar results were obtained 24 hours after seeding.
NO Antiproliferative Action Was Not Mediated by cGMP
Because many physiological actions of NO are mediated by activation of guanylyl cyclase and subsequent increases in intracellular cGMP concentration, this mechanism of action was evaluated in neural precursor cultures. Addition of the membrane-permeant, cGMP analog 8-Br-cGMP to floating neurospheres did not modify the number or the average size of the neurospheres formed 48 hours later. Furthermore, inhibition of soluble guanylyl cyclase by ODQ, at concentrations that prevented vascular NO actions , had no effect on spontaneous cell growth and did not prevent the inhibitory effect of DETA/NO (Fig. 4).
Exogenously and Endogenously Synthesized NO Inhibited EGF-Induced Tyrosine Transphosphorylation of the EGFR
Because neurosphere formation was strictly dependent on EGF, we evaluated whether NO may affect EGFR signaling in these cells. In preliminary experiments, the activation kinetics of the EGFR was established by stimulation of the cells with EGF for increasing periods of time and evaluation of the receptor transphosphorylation, which is an indicative of its tyrosine kinase activity. Maximal transphosphorylation was detected 5 minutes after EGF addition (not shown), and therefore the following experiments were performed under this condition. Figure 5A shows that a protein of 170 kDa, which is the molecular mass of the EGFR, was tyrosine-phosphorylated in cells stimulated with EGF and that the tyrosine phosphorylation band was reduced in the presence of the NO donor SNAP. The band at 170 kDa did indeed correspond to EGFR, as demonstrated in Figure 5B, in which tyrosine phosphorylation is shown in immunoprecipitated EGFR. The inhibitory effect of NO donors on EGFR transphosphorylation was concentration-dependent, as can be seen in Figure 5C and 5D. Densitometric reading of the 170-kDa band corresponding to the transphosphorylated EGFR in different experiments allowed the calculation of an apparent inhibition constant for DEA/NO (K′i [DEA/NO]) of 20 μM. Partial inhibition of NO synthesis by 1-hour incubation with l-NAME produced an enhancement in EGF-dependent EGFR transphosphorylation of approximately 50% (Fig. 5E, 5F), thus indicating that the receptor can also be regulated by endogenously produced NO.
NO Inhibited the EGF-Activated Phosphoinositide-3-Kinase Signaling Pathway in SVZ Neurosphere Cells
We next investigated whether inhibition of the EGFR tyrosin kinase by NO had functional consequences on its downstream effectors. The phosphorylation state of Akt and ERK 1/2 were analyzed to evaluate the activity of the phosphoinositide-3-kinase (PI3-K) and mitogen-activated protein kinase (MAPK) signaling pathways, respectively. Cells were incubated in the absence of any growth factor for 1 hour and then exposed to EGF for different periods of time and to DEA/NO alone or in combination with EGF. Phosphorylation of Akt was EGF-dependent, with similar activation found between 5 and 30 minutes after addition of the growth factor, and was prevented when cells were preincubated with the NO donor; total Akt was not modified by the treatments (Fig. 6A, 6B). On the contrary, EGF was not required for MAPK pathway activation, and neither EGF nor DEA/NO produced significant changes in ERK 1/2 phosphorylation (Fig. 6C, 6D). As a control, we also show the result obtained in NB69 cells, in which ERK 1/2 phosphorylation occurred only after EGF stimulation (Fig. 6C).
The PI3-K and MAPK Pathways Differentially Affected Neurosphere Formation and Growth
To investigate the role of PI3-K/Akt and MAPK signaling pathways in neurospheres, the PI3-K inhibitor LY294002 or the MAPK inhibitor U0126 was added to dissociated cells, and the number and size of the newly formed neurospheres was analyzed 48 hours later. Addition of LY294002 reduced the neurosphere number in a concentration-dependent manner (Fig. 6E, 6I) and had a mild effect on neurosphere size at the highest concentration used (Fig. 6F, 6I). On the contrary, U0126 did not affect the number of newly formed neurospheres (Fig. 6G, 6I), although it also reduced the neurosphere size. (Fig. 6H, 6I). These results indicated that PI3-K, but not MAPK, activation was essential for stem-like cell replication (neurosphere initiation), whereas both pathways contributed to neurosphere growth.
In Vivo NOS Inhibition Enhanced Phospho-Akt Staining in the SVZ and Decreased p27Kip in the SVZ and Olfactory Bulb of Adult Mice
To assess whether the mechanism of action of NO on isolated neural progenitors might be operative also in vivo, l-NAME was administered to adult mice for 15 days, a treatment that significantly inhibits cerebral NOS activity and increases cell proliferation in the SVZ and olfactory bulb , and the presence of phospho-Akt was analyzed in brain sections. As shown in Figure 7A–7C, chronic NOS inhibition significantly enhanced phospho-Akt staining in the SVZ. Many of the cells that contained phospho-Akt expressed EGFR, and the frequency of EGFR+ cells that were also positive for phospho-Akt was higher in animals treated with l-NAME, as can be seen in the lateral wall of the lateral ventricles shown in Figure 7D–7G.
Proliferation induced by the PI3-K pathway is due to Akt-mediated phosphorylation of the cyclin-dependent kinase (CDK) inhibitor p27Kip, which is retained in the cytosol. In the absence of mitogens, p27Kip re-enters the nucleus, where it prevents cell cycle progression. Immunohistochemical detection of p27Kip showed that, in control animals, this protein was present in the nuclei of a few cells in the SVZ (Fig. 7 H) and a large number of cells around the olfactory bulb glomeruli (Fig. 7J, 7L), the final differentiation site of the SVZ precursors. Interestingly, p27Kip staining almost disappeared in the SVZ (Fig. 7I) and was clearly reduced in the olfactory bulb (Fig. 7K, 7M) of animals chronically treated with l-NAME.
In this work, we show that NO, a physiological inhibitor of SVZ neurogenesis in adult rodents, exerted a direct, cGMP-independent, antiproliferative effect on SVZ neural precursors in vitro without affecting cell survival. We also demonstrate that NO prevented the EGF-induced transphosphorylation of the EGFR and the subsequent phosphorylation of Akt, which are required for multipotent progenitor self-renewal and neurosphere formation. In vivo, NOS inhibition enhanced SVZ phospho-Akt and reduced nuclear p27Kip in the SVZ and olfactory bulb of adult mice. We propose that inhibition of the EGFR tyrosine kinase and the PI3-K/Akt pathway, and the subsequent p27Kip translocation into the nucleus, are determinant steps in the control of SVZ neurogenesis by NO.
The neurospheres used in this study were generated from SVZ stem cells (B cells) or transit-amplifying C cells that in vitro maintain stem cell properties and, as a result of asymmetrical division, were constituted by daughter B and/or C cells, as well as a reduced number of progenitors of neuronal or glial lineages ( and present results). Exposure of SVZ cell cultures to NO impaired neurosphere formation, reduced neurosphere size, and decreased cell division, indicating that NO has a direct antiproliferative effect on the neural precursors. Particularly interesting is the fact that NO impaired neurosphere formation, because it allows the identification of the neurosphere-forming cells (that is, the B- and/or C-cell subpopulations) as the cellular targets for NO. This finding is in agreement with our previous results obtained in vivo in adult mice in which the population of SVZ cells most affected by NO contained nestin, but not yet neuronal or glial specific antigens , and may correspond to the same cells that generate neurospheres in vitro.
Neurosphere cells expressed nNOS and eNOS and tonically synthesized NO, which by an autocrine/paracrine mechanism controlled cell proliferation. Therefore, NO effects on neural precursors were not just pharmacological responses to NO donors but were also elicited by endogenous NO production. Expression of nNOS and/or eNOS, but not iNOS, has been reported in cells derived from embryonic brains cultured as neurospheres  or in differentiating conditions . Induction of nNOS by EGF or basic fibroblast growth factor , as well as by the neurotrophic factors nerve growth factor [11, 22] and brain-derived neurotrophic factor , has also been reported in both embryonic and tumoral neural cells. Although in the adult brain SVZ precursors do not express nNOS, they are exposed to NO produced by well-differentiated nitrergic neurons lying in the proximity of the neurogenic area [13, 23]. This anatomical condition explains the proliferative effect of NOS inhibitors also observed in vivo in the adult SVZ.
Because neurosphere cells, as well as the transit-amplifying C cells in vivo, express the EGFR and proliferate when stimulated by EGF, we hypothesized that EGFR might be the molecular target for NO. This hypothesis was based on our previous finding that NO directly inhibits the tyrosine kinase activity of the EGFR in tumoral cells [16, 24]. To directly demonstrate this possibility, the receptor tyrosine phosphorylation was measured in neurosphere cells exposed or not exposed to EGF and NO donors. It is well known that, upon ligand binding, there is a great increase of the EGFR tyrosine kinase activity, which can be determined by measuring the tyrosine phosphorylation of the C-terminal domain of the EGFR itself . EGF-induced transphosphorylation of the EGFR in neural precursors was inhibited by NO donors and was potentiated by NOS inhibition, thus indicating that both exogenous and endogenous NO regulate the first and crucial step in EGF signaling in these cells.
Akt is a serine/threonine protein kinase and a downstream effector of PI3-K; both proteins are part of a signaling pathway that can be initiated by EGFR activation . We observed that, in our neurosphere cultures, Akt phosphorylation was EGF-dependent and was inhibited in the presence of an NO donor. This finding is most likely a consequence of the EGFR tyrosin kinase inhibition produced by NO, although additional effects of NO on some of the intermediate molecules in this pathway cannot be ruled out. The integrity of the PI3-K/Akt pathway was necessary for neurosphere formation, which indicates that Akt is a major mediator in the proliferation and/or survival of neurosphere-forming cells (that is, the multipotent neural precursors [B and/or C cells]). This pathway is also active in vivo in the adult SVZ, where an increase in Akt phosphorylation has been reported after stroke, a condition in which BrdU incorporation is also enhanced . Our finding that systemic NOS inhibition, which significantly increases SVZ precursor cell proliferation , also results in enhanced Akt phosphorylation, strongly suggests that a similar mechanism of action mediates the NO effect on SVZ precursors both in vivo and in vitro.
The role of the MAPK pathway in neural precursor proliferation is uncertain. MAPK inhibition decreased neurosphere size but not neurosphere number, thus suggesting that this pathway participates in the proliferation of daughter cells other than the transit-amplifying C cells (i.e., glial or neuronal precursors). Interestingly, this pathway was activated in the absence of EGF, which suggests that it may be functional in cells devoid of the EGFR, such as the neuronal precursors. Finally, the fact that NO did not significantly inhibit ERK 1/2 phosphorylation suggests that its selective action on Akt is a consequence of EGFR inactivation.
It was recently demonstrated that Akt phosphorylates the CDK inhibitor p27Kip1 and prevents its translocation to the nucleus [27, 28], thus allowing cell cycle progression. Given that p27Kip1 has been identified as a key regulator of the cell cycle specifically in transit-amplifying C cells , this is probably the mechanism by which NO-induced inhibition of Akt results in decreased multipotent precursor proliferation and neurosphere formation. The physiological relevance of this pathway is suggested by the finding that p27Kip1 is present in the SVZ olfactory bulb neurogenic system of adult mice, as previously reported in rats [26, 30]. It is interesting to note the dissimilar distribution of p27Kip1, which was scarce in the highly proliferative SVZ and abundant in the olfactory bulb, where precursors that migrated from the SVZ arrest proliferation and differentiate. Two different conditions have revealed a correlation between p27Kip1 and neurogenesis. After induction of ischemia in rats, cell proliferation is enhanced and p27Kip1 disappears in the ipsilateral SVZ . Also, p27Kip1-null mice present a selective increase in the number of the transit-amplifying progenitors concomitantly with a reduction in the number of neuroblasts in the SVZ . Our previous and present results obtained by analyzing the effects of chronic NOS inhibition on adult SVZ neurogenesis provide several pieces of evidence supporting the view that the EGFR-Akt-p27Kip1 pathway is involved in the physiological role of NO as a negative regulator of SVZ neurogenesis. First, NOS inhibition increased proliferation selectively in transit-amplifying cells and reduced the mitotic rate of neuroblasts , the same effect that was observed in p27Kip1-null mice . Second, only cells provided with the EGFR increased their proliferation rate upon NOS inhibition . Third, chronic NOS inhibition significantly enhanced the presence of phospho-Akt and reduced that of p27Kip1 in the neurogenic regions (present results).
The present results demonstrate that physiological concentrations of NO exert a direct antimitotic action on postnatal SVZ multipotent precursors in vitro. This effect is mediated, at least in part, by inhibition of the EGFR tyrosine kinase and its downstream signaling pathway, PI3K-Akt, which is necessary for neurosphere formation. Chronic depletion of NO in adult mice resulted in enhanced Akt activation in the SVZ, with a parallel disappearance of the Akt-regulated cell cycle inhibitor p27Kip1, both in the SVZ and OB. These in vivo observations suggest that the mechanism of action proposed for NO in SVZ neurosphere cultures is also relevant in the adult animal under physiological conditions.
The authors indicate no potential conflicts of interest.
We thank Drs. C. Guaza, J. Rodrigo, and M. Vallejo for kindly providing A2B5, nNOS, and nestin antibodies, respectively. This work was supported by grants from Fondo de Investigaciones Sanitarias (00/1080), Ministerio de Ciencia y Tecnología (SAF2002-02131), and Junta de Andalucía (CTS-2005/883). E.R.M. is currently affiliated with Departamento de Fisiología y Zoología, Facultad de Biología, Universidad de Sevilla, Sevilla, Spain.