Alzheimer's disease modifies progenitor cell expression of monoamine oxidase B in the subventricular zone

In the adult brain, progenitor cells remaining in the subventricular zone (SVZ) are frequently identified as glial fibrillary acidic protein (GFAP)‐positive cells that retain attributes reminiscent of radial glia. Because the very high expression of monoamine oxidase B (MAO‐B) in the subventricular area has been related to epithelial and astroglial expression, we sought to ascertain whether it was also expressed by progenitor cells of human control and Alzheimer's disease (AD) patients. In the SVZ, epithelial cells and astrocyte‐like cells presented rich MAO‐B activity and immunolabeling. Nestin‐positive cells were found in the same area, showing a radial glia‐like morphology. When coimmunostaining and confocal microscopy were performed, most nestin‐positive cells showed MAO‐B activity and labeling. The increased progenitor activity in SVZ proposed for AD patients was confirmed by the positive correlation between the SVZ nestin/MAO‐B ratio and the progression of the disease. Nestin/GFAP‐positive cells, devoid of MAO‐B, can represent a distinct subpopulation of an earlier phase of maturation. This would indicate that MAO‐B expression takes place in a further step of nestin/GFAP‐positive cell differentiation. In the early AD stages, the discrete MAO‐B reduction, different from the severe GFAP decrease, would reflect the capacity of this population of MAO‐B‐positive progenitor cells to adapt to the neurodegenerative process. © 2010 Wiley‐Liss, Inc.

In the adult brain, progenitor cells remaining in the subventricular zone (SVZ) are frequently identified as glial fibrillary acidic protein (GFAP)-positive cells that retain attributes reminiscent of radial glia. Because the very high expression of monoamine oxidase B (MAO-B) in the subventricular area has been related to epithelial and astroglial expression, we sought to ascertain whether it was also expressed by progenitor cells of human control and Alzheimer's disease (AD) patients. In the SVZ, epithelial cells and astrocyte-like cells presented rich MAO-B activity and immunolabeling. Nestin-positive cells were found in the same area, showing a radial glia-like morphology. When coimmunostaining and confocal microscopy were performed, most nestinpositive cells showed MAO-B activity and labeling. The increased progenitor activity in SVZ proposed for AD patients was confirmed by the positive correlation between the SVZ nestin/MAO-B ratio and the progression of the disease. Nestin/GFAP-positive cells, devoid of MAO-B, can represent a distinct subpopulation of an earlier phase of maturation. This would indicate that MAO-B expression takes place in a further step of nestin/GFAP-positive cell differentiation. In the early AD stages, the discrete MAO-B reduction, different from the severe GFAP decrease, would reflect the capacity of this population of MAO-B-positive progenitor cells to adapt to the neurodegenerative process. V Neural progenitor/stem cells characterized in adult human brain possess the characteristics of self-renewal, proliferation, and differentiation along all major neural lineages (Gross, 2000;Lie et al., 2004;Taupin, 2006). Their development progresses in a permissive microenvironment and proceeds in several stages characterized by their morphology and by gene expression of specific markers such as glial fibrillary acidic protein (GFAP) and the intermediate filament protein nestin (Wei et al., 2002;Imura et al., 2003;Garcia et al., 2004).
In adult mammalian brain, germinal regions are the subventricular zone (SVZ) of the lateral ventricle and the subgranular zone within the dentate gyrus of the hippocampus, which present abundant multipotent neural stem cells showing structural and biological markers of astroglia (Alvarez-Buylla and Garcia-Verdugo, 2002;Christie and Cameron, 2006;Ihrie and Alvarez-Buylla, 2008). In rodent, the SVZ is the source of new specific types of neurons destined to the olfactory bulb (Kornack and Rakic, 2001;Lledo et al., 2008) and of oligondendrocytes during development (Levison and Goldman, 1993). The human SVZ also harbors abundant multipotent progenitor cells exhibiting markers of adult neurogenesis (Bernier et al., 2000(Bernier et al., , 2002 that correspond to astrocytes (Doetsch et al., 1999;Sanai et al., 2004). In contrast to the case in rodent, human SVZ astrocytes are found not adjacent to the ependymal layer but forming a ribbon of cells lining the lateral ventricle, with no evidence of migrating neuroblasts (Sanai et al., 2004;Quinones-Hinojosa et al., 2006).
Basic questions regarding progenitor cell biology and mechanisms of differentiation remain open. Because very few markers allow differentiating of a multipotent radial-glia-like stem cell from a progenitor one, it remains difficult to identify with enough criteria their future neuronal development. Better knowledge of the specific expression of each cell type is then needed to allow a clear discrimination.
With regard to astroglial markers, expression of GFAP has been found in human adult SVZ progenitor cells, whereas the S-100b presence is restricted to mature cells. However, information concerning other well-identified astroglial markers remains elusive. Characterization of astroglial response based on monoamine oxidase B (MAO-B; EC 1.4.3.4) expression has evidenced very high SVZ labeling in adult brain that could not be limited to epithelial and astroglial cells (Saura et al., 1997). With human aging, brain MAO-B activity increases progressively, beginning at about the age of 50-60 years (Saura et al., 1997;Kumar and Andersen, 2004), associated with increased astrogliosis. In Alzheimer's disease (AD) patients, a further increased SVZ MAO-B expression has been observed (Saura et al., 1994;Emilsson et al., 2002;Kennedy et al., 2003), together with an increased SVZ progenitor activity associated with key pathological and neurochemical substrates (Ziabreva et al., 2006). In this study, we investigated whether increased MAO-B expression in the SVZ of AD is related to endogenous proliferation and differentiation of progenitor cells. To help define the process of progenitor cell differentiation and to begin approaching the underlying mechanisms present in neurodegenerative diseases, we investigated the specific expression of MAO-B in the SVZ and evaluated its relationship with specific progenitor cell and astrocyte markers in AD patients compared with age-matched controls.

MATERIALS AND METHODS Human Post-Mortem Brain Tissue
Human post-mortem tissue samples of SVZ of the lateral ventricle walls corresponded to the anterior horn and body of ventricle regions were selected for this study. They included the head and body of the caudate nucleus (Quinones-Hinojosa et al., 2006) and were obtained from our local Neurological Tissue Bank (Serveis Cientifico-Tècnics, Universitat de Barcelona, Barcelona, Spain) according to the European ethical guidelines and approved by the appropriated Research Ethics Committee. Brains were obtained at autopsy from individuals who had suffered a clinical history of AD (n 5 7; stages II, V, VI; aged 77-91 years) and from nonde-mented controls (n 5 3; aged 44-74 years; see Table I for a summary of case histories). Neuropathological confirmation of the clinical diagnosis was undertaken at the Neurological Tissue Bank according to Braak and Newell criteria (Braak and Braak, 1991;Newell et al., 1999). The investigation was carried out on tissues that had been either fresh-frozen and stored at -808C or paraffin embedded; 35 serial sections where obtained from each. Sections thickness was 12 lm for freshfrozen tissue and 8 lm for paraffin-embedded tissue. All sections were mounted in slices (one section each) and used for histological and immunohistochemical procedures.

Western Blot Analysis of MAO-B
Frozen 100-lg samples from dissected SVZ or brain parenchyma were manually homogenized in 5 volumes of icecold Tris-HCl 50 mM, pH 7.7, and centrifuged at 15,000 rpm for 10 min at 48C. The pellet was resuspended in icecooled Tris-HCl and centrifuged twice more. The final pellet was resuspended in Tris-HCl incubation buffer. Protein content was determined by the method of Bradford using bovine serum albumin as standard. Western blot analysis was performed as previously described (Yeomanson and Billett, 1992) using the mouse monoclonal anti-MAO-B antibody 3F12/ G10/2E3. Immobilon-P membranes were used for electroblotting. Blots were probed with anti-MAO-B antibody (1:500, v/v), and binding was detected with a horseradish peroxidase (HRP)-conjugated anti-mouse IgG antibody. Incubation with no primary antibody was used to control the specificity of results. The immunocomplexes were developed with Lumi-Light ECL.

Histology and Immunohistochemistry
Nissl staining was performed according to standard procedure with cresyl violet on four sections of every paraffinembedded tissue. MAO-B histochemistry was performed according to Arai et al. (1986). Briefly, after being washed with 0.01 M phosphate-buffered saline (PBS), five cryostat sections of every fresh-frozen tissue were incubated in a reaction medium for 48 hr at 48C. The medium consisted of 75 mg tyramine hydrochloride, 5 mg 3,3 0 -diaminobenzidine (DAB), 100 mg HRP, 600 mg nickel ammonium sulfate, and 10 26 M clorgyline hydrochloride for monoamine oxidase A inhibition. MAO-B activity appeared in the tissue sections as dark blue precipitates.
For nestin, GFAP, HLA-DR, and amyloid-b (Ab) immunohistochemical analysis, 12 serial paraffin-embedded SVZ sections of every brain sample were processed with the avidin-biotin peroxidase method. For Ab immunohistochemistry, sections had been previously incubated with 98% formic acid for 3 min to enhance antigenicity. A 30-min preincubation in H 2 O 2 -methanol-PBS (0.3/9.7/90) was performed in all slices to inhibit nonspecific staining in blood vessels and neurons. Sections were incubated at room temperature in blocking solution (0.01 M PBS 1 3% normal goat serum, 0.1% Triton X-100) for 2 hr and separately incubated overnight at 48C with the primary antibody at the appropriate dilution in blocking solution. After washing and incubation with the appropriate biotinilated secondary antibody, sections were incubated with ExtrAvidin (1:250) and developed in DAB and H 2 O 2 . Some sections were counterstained with Mayer's hematoxylin.
For MAO-B immunohistochemistry, sections were processed with the avidin-biotin peroxidase method, with some modifications (Rodríguez et al., 2000). Three serial consecutive frozen SVZ sections of every sample were postfixed with acetone for 3 min at room temperature. After blocking endogenous peroxidases, sections were incubated for 30 min at room temperature in blocking solution containing normal pig serum. Overnight incubation was performed at 48C with mouse monoclonal anti-MAO-B antibody (3F12/G10/2E3; Yeomanson and Billett, 1992) diluted 1:50. Then, sections were processed as described above. In all cases, sections stained only with the secondary antibodies were used as negative controls.
According to previous anatomical classification (Quinones-Hinojosa et al., 2006), four layers were observed throughout the SVZ: a monolayer of ependymal cells (layer I), a hypocellular gap (layer II), a ribbon of cells (layer III) composed of astrocytes, and a transitional zone (layer IV). These layers, I-IV, were observed with an optical microscope, and four areas of interest of 1.0 mm 2 each were randomly selected to perform the cell number estimations at 340 objective magnification. This counting procedure was performed in duplicate in three different sections of every case sample. Positive cells were counted in layers I-IV from the SVZ to the parenchyma boundary, and quantification was made in the Image Pro Plus v.5.1 image and analysis system (Media Cybernetics Inc., Bethesda, MD).
Double immunofluorescence was performed on four fresh-frozen and four paraffin-embedded SVZ sections of every sample. Sections were coincubated overnight at 48C with antinestin antibody and anti-MAO-B antibody (fresh-frozen tissue), anti-GFAP antibody, or anti-HLA-DR antibody. After being washed in PBS, sections were incubated in the dark with a cocktail of Alexa488-conjugated anti-rabbit IgG antibody and Alexa546-conjugated anti-mouse IgG antibody. Human brain autofluorescent lipofuscin artifacts were reduced to near-background levels by immersion in 70% ethanol supplemented with 1% of Sudan black B for 5 min (Schnell et al., 1999). Sections were mounted in Immuno-Fluore Mounting medium (ICN Biomedicals, Barcelona, Spain) and examined with a Leica TCS-SL confocal microscope. Counting of double-stained cells was performed as described above for single immunohistochemistry procedures. Because of the antibody characteristics and difficult preservation of human sample integrity, no GFAP/MAO-B or GFAP/MAO-B/nestin colocalization could be performed. As a result, to estimate the relationship between the three cell markers, a correlation analysis was performed, and the ratios GFAP/MAO-B, nestin/GFAP, and nestin/MAO-B were calculated from the cell counts assessed by single immunohistochemistry.

Statistical Analysis
Kurtosis and skewness moments were calculated to test the normal distribution of data. A one-way analysis of variance followed by the Fisher's least significant difference analysis was performed to detect differences. Correlations between cell markers and the progression of AD were estimated by regression analysis between cell counts and a numerical value assigned to each AD stage. All values are presented as mean 6 SD, and differences were considered to be significant at P < 0.05. Data were analyzed with the statistical analysis package StatGraphics 5.0 (STSC Inc., Rockville, MD).

Characterization of Human SVZ
In all control and AD cases, the anterior horn region and the body of the ventricle region of human SVZ presented a similar cellular organization (Fig. 1a,b), as described elsewere (Quinones-Hinojosa et al., 2006). Ependymal cells were arranged as a one-cell-thick epithelium forming layer I. Layer II, a hypocellular region, formed a gap between layer I and a dense ribbon of cell bodies (layer III of the SVZ) of different size and morphology. Layer IV was observed as a transitional zone to the striatal brain parenchyma.
GFAP immunohistochemistry revealed an abundance of GFAP-positive processes within the hypocellular layer, whereas layer III presented many GFAP-positive cell bodies organized as a dense ribbon (Fig. 1c). Their typical astrocyte morphology was associated with processes of irregular caliber with no specific orientation. In this area, control cases had a higher number of GFAP-positive cells than AD cases (189 6 3 cells/mm 2 and 83 6 22 cells/mm 2 , respectively; P < 0.0001; Table  I, Fig. 2f). No correlation was found between GFAPpositive cell density and AD progression (P 5 0.123), but, in all AD samples, a dense population of welldelineated astrocytes appeared deep in parenchyma of caudate nucleus (data not shown).
Nestin immunohistochemistry identified a dense population of polymorphic small cells, forming groups or short chains, localized mainly in layer III and paren-chyma and whose multiple short processes were oriented radially to the hypocellular layer (Fig. 1d). Other nestinpositive cells had larger cell bodies and unipolar or multipolar organization, with few or no visible processes. Small nestin-positive cells were occasionally detected in layer II and the transitional zone (layer IV), especially in AD cases. When control and AD cases were compared, no significant increase in the density of nestin-positive cells was found (159 6 1.7 cells/mm 2 and 218 6 55 cells/mm 2 , respectively; P 5 0.139; Table I, Fig. 2e), but a positive correlation with AD progression was found (r 2 5 0.593, P 5 0.038; Fig. 3), indicating an AD stage VI. Nissl staining showed the anterior horn region and body of ventricle region with a characteristic layer organization. The ependymal cells formed layer I, whereas layer II, a hypocellular region, formed a gap with layer III that was organized as a dense ribbon of cell bodies. Finally, layer IV represented a transitional zone to the striatal brain parenchyma. c: AD stage II. GFAP immunostaining showed layer II as the most GFAP-immunoreactive area, with the GFAPimmunopositive cell bodies localized in layer III. d: AD stage VI. Nestin-immunopositive cells were localized mainly in layer III (arrowheads) and sometimes formed small chains, with short processes ori-ented radially to the hypocellular layer. e: AD stage II. Scarcely activated microglial cells detected by HLA-DR immunohistochemistry were found mainly in layer II (arrowhead). Microglial-positive cells were also detected in the transitional zone (asterisk; layer IV) and brain parenchyma. f: AD stage II. Cerebral Ab protein deposition was practically absent in the human SVZ, except in AD stage VI cases, in which small amyloid deposits were detected in layer III. In AD cases, amyloid-b deposition was abundantly present in the brain parenchyma. LV, lateral ventricle. Scale bars 5 50 lm in a-e; 200 lm in f. [Color figure can be viewed in the online issue, which is available at www. interscience. wiley.com.] increase of progenitors during the course of the disease. Also, a significant increase of the nestin/GFAP ratio between control and AD cases was observed (0.84 6 0.007 and 2.59 6 0.29, respectively; P 5 0.0013). However, no relationship between nestin-and GFAPpositive cell number alteration with AD progression was observed (Fig. 3a,b).
In all samples, a few activated microglial cells were detected in layers III and IV and also in the hypocellular layer (layer II) in controls and AD cases (Fig. 1e). No significant increase of HLA-DR-positive cells was found in the SVZ when control and AD cases were compared (80.3 6 5.1 cells/mm 2 and 78 6 1.0 cells/mm 2 , respectively; P 5 0.958; Fig. 2h), and no correlation was found between microglial density and disease progression (P 5 0.937) nor between nestin-positive cells and microglia in the SVZ (Fig. 3g,h). In contrast, stronger microglial activation with profusion of ramifications was observed in the brain parenchyma of AD cases (raw data not shown).
In all control and AD cases, Ab protein deposition was mostly absent in layers I-IV of the SVZ (Fig. 1f), with the exception of the stage VI AD cases that showed a few small amyloid deposits in layer III. Extracellular amyloid fibrils were also observed in layer III, in medium-sized and small wall arteries and arterioles of all AD samples (data not shown).

MAO-B Localization in the SVZ
Quantitative MAO-B enzyme autoradiography (Saura et al., 1997) clearly showed that SVZ constitutes a human brain area rich in MAO-B (Fig. 2a) but, because of the technique, no information on the cellular types expressing this enzyme was provided. In this study, MAO-B cellular localization was characterized in the SVZ of all samples by using three different methods. When histochemical procedures were performed, a typical MAO-B-positive cell distribution in the cerebral area was found (Fig. 2b). Positive MAO-B cells were localized mainly in layer III of the SVZ, striatal brain parenchyma, and subcortical white matter. In the SVZ, most MAO-B-positive cells presented a stellate morphology, but-positive processes were occasionally observed in layer II, parallel to the lateral wall of lateral ventricle (inset in Fig. 2b). Small proliferation of MAO-B-positive cells was observed in layer II of samples of AD stage VI (data not shown). MAO-B-positive cells were also detected by immunohistochemistry with similar cell morphology and distribution as described above for MAO-B histochemistry. MAO-B-specific immunostaining was found in cells showing the morphology of astrocytes in the SVZ and in the cerebral parenchyma (Fig.  2c). Parallel-positive MAO-B processes were not detected, but instead-positive fine punctuated cellular ramifications with no consistent orientation were observed. When MAO-B-immunopositive cells were quantified, cell density was significantly decreased in the SVZ of all AD cases compared with controls (mean density 118 6 1.7 cells/mm 2 in controls and 96 6 7.1 in AD samples cells/mm 2 , P 5 0.0178; Table I, Fig. 2e). This decrease did not correlate with AD stage progression (P 5 0.131). Western blot analysis of SVZ and caudate nucleus parenchyma showed a similar specific intense band corresponding to the MAO-B molecular weight (Fig. 2d). The difficult preservation of SVZ integrity, caused by the post-mortem time and tissue dissection, render Western blot quantification unreliable.
When the GFAP/MAO-B ratio was studied, a significant 43.7% decrease was found in AD cases compared with controls (1.6 6 0.003 and 0.9 6 0.08 respectively, P 5 0.014; Fig. 3e,f), but no correlation with AD progression was detected (P 5 0.159), nor between GFAP and MAO-B cell counts. When the nestin/ MAO-B ratio was analyzed, a significant 68% increase was found in AD cases compared with controls (1.35 6 0.01 for controls and 2.27 6 0.23 for pathology, P 5 0.013; Fig. 3c,d). The correlation between this nestin/ MAO-B ratio and the progression of the AD stages reached significance (P 5 0.025). Finally, the nestin/ HLA-DR ratio revealed no difference and no correlation with AD progression (P 5 0.369 and P 5 0.066, respectively).

Colocalization of SVZ Markers
Nestin-GFAP double confocal immunohistochemistry revealed the presence of abundant astrocyte-like cells in layer III and also a significant presence in layer IV, with cell processes or somata exhibiting both nestin and GFAP immunoreactivities. Double-immunoreactive cell bodies had oval or fusiform shapes and exhibited prominent, long, slender processes that developed parallel to the wall of the lateral ventricle of the anterior horn. In the body of the ventricle, these processes appeared with no special organization. Nearly all nestinpositive cells expressed GFAP, and major proportions of GFAP-positive cells were also positive for nestin (Fig.  4a-c). The few cells found to be positive for nestin with no astrocytic features were more abundant in the parenchyma, especially in stage VI of AD samples.
In all control samples, double nestin-MAO-B immunohistochemistry revealed the presence of large cells with stellate morphology in layer III and in the hypocellular layer of the SVZ nearby blood vessels (Fig.  4d-f). Some 35% of nestin-positive cells were also positive for MAO-B, and a similar percentage of MAO-Bpositive cells also expressed nestin. Fewer nestin/MAO-B-positive cells with a smaller round shape were observed in AD samples. In all double-immunostained cells, labeling of nestin was located mostly in soma and MAO-B in soma and processes (Fig. 4g-i), except in AD stage VI, in which MAO-B was more localized in processes. Few double-immunostained round cells and unipolar and bipolar cells were also detected in the transitional zone and striatal parenchyma. Finally, nestin immunoreactivity and HLA-DR-immunolabeling showed no colocalization. In the SVZ, HLA-DR-positive cells were observed, mainly in layer III and IV. No spatial relationship between nestin-positive and HLA-DR-positive cells was observed, except in AD samples, in which some HLA-DR-positive cells surrounded nestin-positive cells (data not shown).

DISCUSSION
The present study gives evidence for the first time that MAO-B is expressed in SVZ progenitor cells of the human brain between 44 and 91 years. Morphological examination of the anterior horn and body of human lateral ventricle confirmed that the adult human SVZ is organized into four specific layers (I-IV), with the SVZ astrocytes separated from the ependyma (I) by a hypocellular region (II) devoid of cells bodies. These astrocytes, localized mainly in layer III, formed a cell ribbon before the transitional zone (IV) to parenchyma (Sanai et al., 2004;Quinones-Hinojosa et al., 2006). Our data indicate an increase in progenitors in the SVZ during the course of AD. Neurogenesis in the SVZ is increased in acute neurological disorders such as ischemia and epilepsy (Blumcke et al., 2001;Felling and Levison, 2003;Kokaia and Lindvall, 2003) or in neurodegenerative disorders such as AD, Creutzfeldt-Jakob disease, or brain tumors (Jin et al., 2004;Mizuno et al., 2006;Ziabreva et al., 2006;Quinones-Hinojosa and Chaichana, 2007;Jackson and Alvarez-Buylla, 2008;Waldau and Shetty, 2008). Here we found that expression of markers of radial cell differentiation is independent of neurodegeneration, in spite of the small deposition of Ab protein and microglial reaction present in layers I-IV of the germinal zone.
According to previous immunohistochemical and enzyme histochemical studies, MAO-B is localized in human brain astrocytes as well as in serotoninergic and histaminergic neurons of the raphe nuclei and posterior hypothalamus (Konradi et al., 1988(Konradi et al., , 1989. In this study, post-mortem delay ranged from 3 to 19 hr, and the sample time of storage prior to experiments ranged from 1 to 6 months. Under these conditions, MAO-B remains a stable protein, and, when present, its variation has been related to a cellular process. We and other authors have previously described a widespread increase of MAO-B expression in human brain aging, as a consequence of a general astroglial hypertrophy and/or hyperplasia (Nakamura et al., 1990;Saura et al., 1994Saura et al., , 1997. This astroglial up-regulation of MAO-B activity is closely related to AD senile plaques in cortical layers (Nakamura et al., 1990;Saura et al., 1994Saura et al., , 1997. However, the high expression of MAO-B in the SVZ under similar brain aging conditions, and its high level in ventricle ependyma, shown previously by quantitative MAO-B enzyme autoradiography (Saura et al., 1997), argue for a new cellular process, different from astrogliosis. Colocalization of MAO-B and nestin indicates the expression of a new marker of human adult SVZ, a zone that remains mitotically active in mammals throughout adult life (Alvarez-Buylla and Garcia-Verdugo, 2002;Merkle et al., 2004). The nestin/GFAPpositive cells were abundant in SVZ layer III and organized as a continuous ribbon, whereas nestin/MAO-Bpositive cells localized in layer III were less abundant and with no clear organization. Some nestin/MAO-Bpositive round cells and unipolar and bipolar cells were also detected in the transitional zone or layer IV of SVZ and brain parenchyma. In this portion of the germinal zone, MAO-B cellular localization appeared increased in the filaments in all the pathological cases, and more especially in the AD cases. The positive correlation between nestin/MAO B ratio and progression of the disease, and the increased nestin/GFAP ratio found in AD, Fig. 4. Colocalization of SVZ markers of AD cases. a-c: AD stage V. Double confocal immunohistochemistry of the body of the ventricle, showing GFAP in green and nestin in red, revealed the presence of oval or fusiform astrocyte-like cells mostly in layer III (merged), with positivity large and extensive, with no special organization. d-f: AD stage VI. Double confocal nestin (red)/MAO-B (green) immunohistochemistry revealed the presence of round cells and unipolar and bipolar cells in layer III and the transitional zone and striatal brain parenchyma (merged) especially in AD samples. g-i: In control samples, the presence of nestin (red)-/MAO-B (green)-positive large cells with stellate morphology was detected in layer III and in the hypocellular layer nearby blood vessels (merged). Scale bars 5 40 lm in a-f; 16 lm in g-i.
[Color figure can be viewed in the online issue, which is available at www.interscience. wiley.com.] could reflect the increased progenitor activity previously described for these patients (Ziabreva et al., 2006). Finally, the nestin-positive cells lacking GFAP that we found in SVZ and parenchyma in AD stage VI samples could represent fully committed migrating neuroblasts (Kronenberg et al., 2003).
GFAP and nestin have been the predominant markers used to describe stem and progenitor cells in mammalian CNS (Doetsch et al., 1997(Doetsch et al., , 1999Bernier et al., 2000;Ihrie and Alvarez-Buylla, 2008). In response to brain injury or degeneration, mature reactive GFAPpositive astrocytes can express nestin and revert to the embryonic phenotype of neuroepithelial stem cells (Lin et al., 1995;Bernier et al., 2000). In our study, nearly all nestin-positive cells also express GFAP, but only 35% express MAO-B. The nestin/GFAP-positive cells devoid of MAO-B may represent a distinct subpopulation that proliferates during an earlier phase of maturation. Dopamine-induced proliferation of precursor cells in the SVZ has been recently reported (O'Keeffe et al., 2009). Because dopamine tissue content depends on MAO-B activity (Youdim et al., 2006), the increased MAO-B expression herein evidenced could be related to modulation of that progenitor cell proliferation. If this is true, MAO-B expression would take place in a further step of nestin/GFAP-positive cell maturation, to limit proliferation and facilitate the subsequent differentiation of progenitor cells. In the early AD stages, the increased nestin-positive cells paralleled by a marked reduction of GFAP immunoreactivity evidence proliferation of progenitor cells and differentiation to neuroblasts (Kronenberg et al., 2003) or to newborn cells degenerated in the niche of the germinal zone. Under these conditions, the discrete MAO-B reduction, different from the severe GFAP decrease, would reflect the capacity of progenitor cells to adapt to the neurodegenerative process at the SVZ level by increasing their differentiation rate. However, further investigation is required to decipher the MAO-B participation in progenitor cell maturation and differentiation under control and neuropathological conditions. Regardless, MAO-B labeling provides a new, reliable tool for SVZ human stem cell study under control and pathological conditions. Finally, because of the marked differences between adult human and other vertebrates SVZ, our work highlights the importance of studying these cells in the human brain, especially when related to CNS diseases.