Granulocyte colony-stimulating factor stimulates neurogenesis via vascular endothelial growth factor with STAT activation

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Granulocyte colony-stimulating factor stimulates neurogenesis via vascular endothelial growth factor with STAT activation Keun-Hwa Jung a,b,1 , Kon Chu a,c,1 , Soon-Tae Lee a , Se-Jeong Kim a , Dong-In Sinn a , Seung U. Kim d,e , Manho Kim a , Jae-Kyu Roh a,⁎ a Stroke and Neural Stem Cell Laboratory in Clinical Research Institute, Department of Neurology, Seoul National University Hospital, Seoul, South Korea b Department of Disease Investigation and Surveillance, Korea Center for Disease Control and Prevention, Seoul, South Korea c Center for Alcohol and Drug Addiction Research, Seoul National Hospital, Seoul, South Korea d Brain Disease Research Center, Ajou University, Suwon, South Korea e Division of Neurology, Department of Medicine, UBC Hospital, University of British Columbia, Vancouver, Canada

A R T I C LE I N FO

AB S T R A C T

Article history:

The adult brain harbors multipotent stem cells, which reside in specialized niches that

Accepted 6 December 2005

support self-renewal. Granulocyte colony-stimulating factor (G-CSF) induces bone marrow

Available online 19 January 2006

stem cells proliferation and mobilization from their niche, and activates endothelial cell proliferation, which might help to establish a vascular niche for neural stem cells (NSCs).

Keywords:

Here, we show that G-CSF induced receptor-mediated proliferation and differentiation of

Granulocyte colony-stimulating

neural precursors in human NSCs cultures and in adult rat brain in vivo. In human NSCs

factor (G-CSF)

cultures, G-CSF activated STAT3 and 5, and increased VEGF and its receptor, VEGFR2 (Flk-1)

Neurogenesis

expression, and VEGFR2 tyrosine kinase inhibitor blocked the neurogenesis stimulated by G-

G-CSFR

CSF. G-CSF also activated endothelial cell proliferation in adult rat brain in vivo. Our results

STAT

indicate that G-CSF stimulates neurogenesis through reciprocal interaction with VEGF and

VEGF

STAT activation. © 2005 Elsevier B.V. All rights reserved.

Abbreviations: NSC, neural stem cell DG, dentate gyrus GCL, granule cell layer SGL, subgranular layer SVZ, subventricular zone RMS, rostral migratory stream OB, olfactory bulb

⁎ Corresponding author. Fax: +82 2 3672 4949. E-mail address: [email protected] (J.-K. Roh). 1 The first two authors contributed equally to this study. 0006-8993/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.12.037

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

1.

Introduction

A variety of adult stem cells were localized in restricted areas of several organs known as “niche” where a subset of tissue cells and extracellular substrates can indefinitely house one or more stem cells and control their selfrenewal and progeny production. The adult brain harbors multipotent stem cells, which reside in specialized niches that support self-renewal (Luskin, 1993; Magavi et al., 2000; Shen et al., 2004). Astrocytes, ependymal cells and endothelial cells collaborate to define niches for cell genesis, and reciprocal paracrine interactions between these cells can both permit and regulate neurogenesis or gliogenesis from resident precursor cells. Agents as diverse as epidermal growth factor (EGF) (Reynolds and Weiss, 1992), fibroblast growth factor-2 (FGF-2) (Ray et al., 1993), brainderived neurotrophic factor (BDNF) (Pencea et al., 2001), erythropoietin (Shingo et al., 2001), insulin-like growth factor-1 (IGF-1) (Aberg et al., 2000), stem cell factor (SCF) (Jin et al., 2002a), and vascular endothelial growth factor (VEGF) (Cao et al., 2004; Jin et al., 2002b; Sun et al., 2003) have been implicated as permissive factors for neurogenesis in the adult brain. Granulocyte colony-stimulating factor (G-CSF), a member of the cytokine family of growth factors, induces BM stem cells (BMSCs) proliferation and mobilization from their niche (Petit et al., 2002). G-CSF exerts its activity via a receptor (GCSFR) of the hematopoietin receptor superfamily (Cosman, 1993). Binding to G-CSFR recruits cytoplasmic tyrosine kinases of both the Jak and Src kinase families and activates STAT proteins (Shimoda et al., 1997; Tian et al., 1994, 1996). Activated STAT translocates to the nucleus and regulates specific target gene expression, which allows cells to proliferate, differentiate and mobilize or to obtain enough trophic support to survive (Shuai et al., 1993; Tian et al., 1996). By this complex cascade, G-CSF might also regulate the proliferation, migration, or differentiation of the adult neural stem cells (NSCs). On the other hand, G-CSF directly activates endothelial cell proliferation (Bussolino et al., 1991; Natori et al., 2002). The release of BDNF from endothelial cells can establish a vascular niche that favors the proliferation of neuronal precursors (Louissaint et al., 2002). Besides, STAT3 has been reported to directly regulate VEGF expression and hence angiogenesis in the human pancreatic cancer (Wei et al., 2003). Recently, VEGF has been implicated as a factor that promotes neurogenesis in the adult brain (Jin et al., 2002b; Louissaint et al., 2002), and the kinase receptors, VEGFR1 (Flt-1) and VEGFR2 (Flk-1; KDR), have been known to be also expressed by nonvascular cells such as neurons or NSCs (Maurer et al., 2003; Sondell et al., 1999). GCSF-induced STAT activation might operate an autocrine VEGF loop in NSCs, and subsequently enhance the neurogenesis. Little is known regarding the existence of G-CSFR on human NSCs or possible role of G-CSF in the adult brain. In the present study, we show that it stimulates the proliferation of neuronal precursors in human NSCs cultures and in adult rat brain in vivo. We then elucidate the plausible mechanisms of G-CSF action in the adult brain.

191

2.

Results

2.1.

G-CSF stimulates neurogenesis in human NSCs in vitro

HB1.F3 NSCs were bipolar, tripolar or multipolar in morphology and 8–10 μm in size (Fig. 1a). The cells expressed phenotypes specific for NSCs, nestin (100%, Fig. 1b). Most cells expressed markers associated with immature neurons, β-III tubulin (Tuj-1; 68%, Fig. 1c) and doublecortin (Dcx; 65%). The cells were also immunopositive to proliferating cell markers; 60% showed positivity to BrdU and 58% to Ki-67 (Fig. 1d). Within the BrdU+ population, little change in the relative proportion of neurons and astrocytes was observed (Tuj-1: 69%, Dcx: 67%). When these cultures were treated with G-CSF for 72 h, they showed increased incorporation of BrdU into cells (Figs. 1e and f; up to +22% vs. control) that expressed the immature neuronal marker, Tuj1 or Dcx (Fig. 1g; up to +30% vs. control). This effect was concentration-dependent and was associated with an increase in cell number (DAPIstained nuclei, Fig. 1h; up to +28% vs. control, P = 0.003, n = 4). The concentration-response relationships for stimulation of BrdU labeling and for elevation of cell number by G-CSF were similar, with maximal effects at 50 ng/ml. At higher G-CSF concentrations (N50 ng/ml), the effect tapered off. To document further the mechanisms by which G-CSF induces the cell growth, the cell cycle distribution was determined by measuring DNA content with flow cytometry in cells with or without G-CSF treatment. As shown in Fig. 1i, G-CSF at 50 ng/ ml induced a significant reduction in the percentage of cells in G0/G1 (53.7%) and a parallel increase in the percentages of cells in S (31. 2%) and G2/M (15.1%) in comparison with untreated control (67.4, 23.6 and 9.0%, respectively) and G-CSFtreated cells at 10 ng/ml (62.2, 26.6 and 11.1%, respectively).

2.2. G-CSF-induced neurogenesis is mediated via G-CSF receptor with subsequent activation of STAT pathway To identify the receptor mechanism involved in the proliferation of NSCs in vitro by G-CSF, NSCs cultures were immunostained with antibodies against G-CSFR, and total cell lysates from the NSCs treated with or without G-CSF were harvested for the western blot analysis. G-CSFR was expressed in almost all cells (Fig. 2a), and the levels were not altered by G-CSF treatment (Fig. 2b). To explore if STAT activation mediates the effects of G-CSF on cell proliferation, HB1.F3 cell extracts were further analyzed to detect the activated tyrosine-phosphorylated forms of STAT3 (pSTAT3) and STAT5 (pSTAT5). STAT3 and -5 were phosphorylated to some extent even at the steady state. G-CSF treatment increased STAT3/5 phosphorylation (Figs. 2c and d), while total STAT3/5 protein levels were not sensitive to treatment. The optical density analysis (Fig. 2e) showed a 4.2 and 2.7-fold higher expression for pSTAT3 and pSTAT5 respectively, in the G-CSF-treated cells compared with controls.

2.3. G-CSF-induced neurogenesis is potentiated by VEGF secreted from NSCs with upregulation of VEGFR2/Flk-1/KDR We next investigated whether G-CSF influences VEGF production and VEGF receptor expression in human NSCs. ELISA

192

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

Fig. 1 – G-CSF-induced neurogenesis in human NSCs, HB1.F3 cell cultures. (a) Phase-contrast microscopy and immunostainings (green) of HB1.F3 cells for (b) the NSC marker, nestin, (c) the immature neuronal marker, Tuj-1, and (d) the cell proliferation marker, Ki-67 with DAPI staining (blue). (e) G-CSF stimulated incorporation of BrdU into cells. Quantitative analysis showed (f) increased BrdU incorporation and (g) production of cells of neuronal lineage in the G-CSF-treated cells, maximally at 50 ng/ml. (h) G-CSF (50 ng/ml) significantly induced an increase in the number of HB1.F3 cells measured by counting DAPI. (i) When the percentage of cells in each phase of the cell cycle was determined, G-CSF induced a significant reduction in the percentage of cells in G0/G1 and a parallel increase in the percentages of cells in S and G2/M, with maximal effect at 50 ng/ml. Data shown are representative fields (a–e) or mean ± SD (f, g, h, n = 3). *P b 0.05 compared with untreated control. Bar = 10 μm.

analysis revealed that VEGF levels rose from 137.4 ± 13.2 pg/ml at baseline to 163.4 ± 9.6 pg/ml at G-CSF-treated NSCs cultures (50 ng/ml), which was statistically significant (Fig. 3a; P = 0.02). To detect both constitutively expressed and induced receptors, either of which could be involved, western blotting was performed in our HB1.F3 cell cultures. The VEGFR2/Flk-1/KDR receptors were expressed in the untreated control, while VEGFR1/Flt-1 receptors were scarcely expressed. VEGFR1/Flt-1 receptor was highly expressed in the human endothelial cell line ECV.304 which was used as a positive control of our study. The expression of VEGFR2/Flk-1/KDR receptors was upregulated in the G-CSF treated cultures (Fig. 3b). The optical density analysis (Fig. 3c) showed a 2.5-fold higher expression for Flk-1/ KDR in the G-CSF-treated cultures compared with controls. To determine if locally secreted VEGF contributes to the neuroproliferative effect of G-CSF, we next examined the effects of VEGF or VEGFR2/Flk-1 receptor tyrosine kinase inhibitor on the neurogenesis stimulated by G-CSF (Fig. 4). The combinatorial treatment of G-CSF (50 ng/ml) and VEGF (50 ng/ml) significantly increased the total number of HB1.F3 cells (Fig. 4a; +19% vs. GCSF alone, P = 0.002; +16% vs. VEGF alone, P = 0.01; +43% vs. control, P b 0.001). The combinatorial treatment of G-CSF and VEGF also induced a significant reduction in the percentage of cells in G0/G1 (51.8%) and a parallel increase in the percentages

of cells in S (33. 2%) and G2/M (15.0%) in comparison with G-CSF or VEGF alone (Fig. 4b). In addition, the VEGFR2/Flk-1 receptor tyrosine kinase inhibitor SU1498 blocked the G-CSF ability to increase the stem cell proliferation (Fig. 4a; −14% vs. G-CSF alone, P b 0.001, n = 4) or to shift the cell cycle to S and G2/M phases from G0/G1 (S: 23.6%, G2/M: 8.4%, G0/G1: 68.0%).

2.4. G-CSF stimulates neurogenesis in SVZ, SGL, and OB in vivo To evaluate if G-CSF acts as a neurogenic factor in adult brain in vivo, we first examined whether G-CSFR is expressed in proliferative zones in the rat brain. G-CSFR was detectable in the SVZ of adult rats where G-CSF or vehicle is infused in (Figs. 5a and b). Numerous scattered neurons and glial cells adjacent to these areas were also stained with the G-CSFR antibody. No difference in the pattern and intensity of G-CSFR immunoreactivity was evident between the two experimental groups. To ascertain whether G-CSF induces in vivo neurogenesis, as demonstrated above in vitro experiments, brain sections from G-CSF- and vehicle-treated rats were processed for doublelabeling immunohistochemistry with antibodies against BrdU and cell-type-specific markers (n = 6). Although the distribution of the newly generated cells was similar in both sets of animals,

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

193

Fig. 2 – Association of G-CSF-stimulated neurogenesis with G-CSF receptor and STAT pathway in HB1.F3 cell cultures. (a) Immunostaining of G-CSFR (red) with DAPI staining (blue) showed a variable degree of cytoplasmic staining of most HB1.F3 cells. (b–d) The Western blot was sequentially probed for human G-CSFR (b), STAT3 (c, lower pannel), STAT5 (d, lower pannel), phospho-STAT3 (c, upper panel), and phospho-STAT5 (d, upper panel). (e) From the quantitative analysis by relative optical density (n = 3), G-CSF led to the phosphorylation of STAT3 and STAT5 proteins, while G-CSFR, STAT3 and STAT5 levels were not altered by G-CSF treatment. *P b 0.05 compared with untreated control. Scale bar = 10 μm.

the BrdU labeling in G-CSF-infused brains was upregulated with subsequent expansion of the SVZ, SGL and rostral migratory stream (RMS) (Fig. 5). From the quantitative analysis, G-CSF increased BrdU-positive cells by 1.5- to 2-fold in the SGL and SVZ, and by 3-fold increase in the olfactory bulb (OB) (Fig. 5; P b 0.05, n = 6). Cells that incorporated BrdU by 14 days after GCSF infusion included Tuj-1-expressing immature neurons in the SVZ, SGL and RMS and CB-positive mature neurons in OB (Fig. 6). Quantitative analysis from the OB revealed that G-CSF treatment produced greater amount of cells of neuronal lineage (Fig. 6g; Tuj-1: 1076 ± 92 cells; CB: 1425 ± 123 cells; GFAP: 475 ± 41 cells, P b 0.05, n = 6) compared to vehicle (Tuj-1: 321 ± 34 cells; CB: 425 ± 45 cells; GFAP: 141 ± 15 cells).

2.5. G-CSF increases vascular density and induces mitotic angiogenesis in the brain To address the neurogenic mechanisms of G-CSF in vivo, we next asked whether the intraventricular administration of GCSF also affects the cerebral microvascular profiles and endothelial proliferation. To confirm the identity of endothelial cell division, sections were double-immunolabeled for both BrdU and laminin, a marker of endothelial cells. Figs. 7a–c show that BrdU+ nuclei in the SVZ and OB of G-CSF-infused rats colocalize with immunoreactivity for laminin. Double-labeled cells were common after G-CSF infusion, but extremely rare in vehicle-infused animals, consistent with the normal absence of significant angiogenesis in the SVZ and OB. When counted in enlarged and thin-walled vessels, containing BrdU+ endothelial cells (Fig. 7d), the fraction of BrdU+ endothelial cells in the SVZ

and OB was 8.4 ± 2.6% and 3.5 ± 0.7% in the G-CSF treated rats and 2.6 ± 0.7% and 1.7 ± 0.5% in the controls (P b 0.05, n = 6). We also noticed a dramatic change in blood vessels after G-CSF treatment compared to vehicle (Figs. 7e, f). The mean diameter of the cerebral microvessels was significantly larger in the SVZ and OB of the G-CSF treated rats, as compared with the corresponding value for the control rats (SVZ: 7.6 ± 2.1 μm vs. 4.6 ± 1.9 μm, OB: 6.3 ± 1.7 μm vs. 4.2 ± 0.9 μm; P b 0.05, n = 6).

3.

Discussion

The major finding of this study is that G-CSF, identified originally as a regulator of hematopoietic stem cells, and more recently as an angiogenic factor, can also stimulate neurogenesis, identified by increased BrdU labeling of neuronal precursor cells in vitro and in vivo. G-CSF-induced neurogenesis was associated with phosphorylation of STAT3 and -5, which was probably mediated by interaction with G-CSFR. In addition, G-CSF induced VEGF and VEGFR2/flk-1/KDR expression in NSCs cultures, and the G-CSF effect was blocked by VEGFR2 tyrosine kinase inhibitor. To identify G-CSF-induced neurogenesis, we employed BrdU labeling in human NSCs cultures. G-CSF increased the NSCs number with dose response and stimulated the incorporation of BrdU into the cells that co-expressed phenotypic markers of neurons. G-CSF might rescue the newly formed cells that would otherwise undergo cell death. However, in our study, simply the prevention of cell death seemed insufficient to account for the immense number of BrdU-positive cells and

194

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

Fig. 3 – ELISA for VEGF and western blot analysis of VEGFR. (a) VEGF concentrations (mean ± SD) were determined in supernatants using ELISA and expressed as pg/ml. VEGF was secreted by HB1.F3 cells to some extent at the steady state, and this was significantly increased by G-CSF (P = 0.02, n = 3). (b) Protein from control cultures or from cultures exposed to G-CSF for 72 h was probed for the principal VEGF receptors with anti-VEGFR2/Flk-1 or anti-VEGFR1/Flt-1 antibodies. (c) VEGFR2/Flk-1 receptors were upregulated by G-CSF treatment, while VEGFR1/Flt-1 receptors are scarcely expressed in the untreated or treated groups. The experiment was repeated three times with similar results. *P b 0.05 compared with untreated control.

the cell cycle shift to S and G2/M phases. The additional finding that G-CSF-induced cell proliferation tapered off at higher concentration than 50 ng/ml, suggested that the

neurogenesis by G-CSF might be critically dependent on proper concentration. G-CSF (10–100 ng/ml) significantly induced phosphorylation and activation of Jak2 and STAT3, and to a lesser extent, STAT 1 and 5 on bone marrow precursors (Hofer et al., 2005; Li et al., 2005; Liu et al., 2001). The dose of G-CSF (50 ng/ml) that is effective in our in vitro study was similar to that which works on bone marrow precursors. We also tested the in vivo neurogenic potential by the intraventricular administration of G-CSF. The intraventricular dose of G-CSF was adjusted to the results of the in vitro experiment along with other growth factors (Jin et al., 2002b). The intraventricular route of G-CSF administration could partly exclude a possibility that BrdU-labeled cells in the brain could originate from BM. Neurogenesis persists in restricted adult mammalian forebrain structures, including the SGL of the hippocampal dentate gyrus (Keyt et al., 1996), the SVZ of the lateral ventricle (Luskin, 1993), and the parenchyma of the adult forebrain (Magavi et al., 2000). The neural progenitors, which originate from stem cells located in the SVZ, follow an intricate migration path and reach their final position in the OB (Luskin, 1993). They move tangentially, in clustered chains, along the entire extent of the RMS. Once in the bulb, they turn to move radially out of the RMS into outer layers, when they differentiate into CB-positive GABAergic interneurons (Luskin, 1993). Because the OB is one of the few structures in the adult forebrain in which there is a continuous supply of newborn neurons, we examined the effect of G-CSF on the distribution and phenotype of newly generated cells mainly in SVZ-RMS-OB stream. G-CSF infusion resulted in numerous BrdU-positive cells, not only in the SVZ and SGL, but moreover, in OB. In each region, most of the newly generated cells expressed the neuronal marker, Dcx, Tuj-1 or CB. To elucidate the mechanism underlying these findings, we examined the G-CSFR and STAT activation. G-CSFR is a single-chain member of the cytokine receptor superfamily, which lacks tyrosine kinase activity (Ihle, 1995). Binding of

Fig. 4 – The effects of VEGF or VEGFR2/Flk-1 receptor tyrosine kinase inhibitor on the neurogenesis by G-CSF. (a) The total cell number was measured by counting DAPI. (b) Cell cycle was analyzed by flow cytometry using propidium iodide. Maximally effective concentrations (50 ng/ml) of G-CSF and VEGF significantly increased the total cell number (P b 0.01, n = 3), and the cell percentage in the S and G2/M phases compared with the treatment of G-CSF or VEGF alone. G-CSF-stimulated cell proliferation in HB1.F3 cell cultures was inhibited by the VEGFR2/Flk-1 receptor tyrosine kinase inhibitor SU1498 (P b 0.01, n = 3). P b 0.05 compared with G-CSF alone (*) or VEGF alone (#) group, respectively.

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

195

Fig. 5 – G-CSFR expression and BrdU incorporation into neuroproliferative zones of rat brain in vivo. (a, b) Brain sections through SVZ were immunostained with an Ab against G-CSFR, which was visualized with DAB. G-CSFR was expressed in SVZ of both G-CSF- and vehicle-treated animals. Brain sections through SVZ, RMS, OB, and SGL were immunostained with anti-BrdU antibody 14 days after intraventricular infusion of vehicle (c–f) or G-CSF (g–j). G-CSF increased the number of BrdU+ cells in SVZ (c, g), RMS (d, h), OB (e, i) and SGL (f, j). Quantitative analysis (k–m) showed that the number of newly generated cells in the SVZ, SGL and OB was significantly higher in G-CSF-infused rats than in vehicle-infused ones (P = 0.002, P = 0.01, P b 0.001, n = 6, respectively). (a, c–j) ×200. (b) ×400 (HP); *P b 0.05 compared with G-CSF alone; SVZ: subventricular zone, SGL: subgranular layer, RMS: rostral migratory stream, OB: olfactory bulb.

G-CSF to its receptor, induces the activation of Jak-1 and Jak-2. In addition to the activation of these Jaks, this interaction also leads to the phosphorylation of STAT (Ihle, 1995; Rane and Reddy, 2000). On phosphorylation, STAT translocates to the nucleus and initiates transcription of its target genes, which in turn exhibit pleiotropic effects on many common processes regulating cell fate (Ihle, 1996). The cell cycle is regulated tightly at G1-S and G2-M checkpoints by several protein kinases composed of a cdk subunit and corresponding regulatory cyclin subunit, and cdk inhibitors (Grana and Reddy, 1995). The G1-S checkpoint is controlled by the CIP/KIP and INK4 families of cdk inhibitors, including p21waf1/cip1, p27kip1, and p18ink4c, which bind to G1 cyclin/cdk complexes and inhibit their catalytic activity, thus preventing the transition of cells from G1 to S phase (Sherr and Roberts, 1999). STATs have been shown to control cellular proliferation by regulating p27kip-1 expression (Kaplan et al., 1998). G-CSFR was expressed in our cultures and the neuroproliferative zones of the adult rat brain as reported previously (Schneider et al., 2005). While total STAT3 and -5 protein levels were not changed by G-CSF treatment, the phosphorylated STAT3 and -5 protein levels increased, supporting that STAT phosphorylation depends on G-CSFR signaling. Cell cycle analysis revealed a significant stimulation of cell cycle

progression by G-CSF at the G1-S checkpoint. This result suggests that the G-CSF-induced NSCs proliferation is mediated, at least in part, through cell cycle stimulation at the G1-S checkpoint. Stimulus-induced G-CSF expression has been reported in human astrocytes (Aloisi et al., 1992; Wesselingh et al., 1990), and Jak-STAT signaling pathways have been implicated in the promotion of astrocytic differentiation and induction of the astrocytic marker, GFAP (Bonni et al., 1997). Thus, G-CSF might be involved in gliogenesis. However, a recent report demonstrated that G-CSF and GCSFR was expressed predominantly in neurons or neural stem cells in normal or ischemic rat brain rather than astrocytes (Schneider et al., 2005). Moreover, exogenous GCSF could stimulate neuronal differentiation of adult neural stem cells in vitro rat hippocampal cultures and in vivo rat brain (Schneider et al., 2005). These actions of endogenous or exogenous G-CSF support an involvement of neurons or neuronal precursors in neurogenesis. To clarify the involvement of glial cells in neurogenesis, the receptor–ligand interaction in various cell types and the precise mechanism of cooperation between signaling pathways must be determined. Many leukemia and cancer cells exhibit constitutive activation of STAT5, which was suggested to contribute to

196

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

Fig. 6 – Phenotype of newly generated cells in the adult rat brain. (a–f) Brain sections were immunostained for BrdU (green) and marker of immature (Tuj-1) or mature (Calbindin) neurons (red). Merged images (yellow) showed that a large amount of newly generated cells in the SVZ, RMS and SGL and within the OB expressed a neuronal phenotype. (g) At 14 days after G-CSF infusion, G-CSF treatment produced numerous cells of neuronal lineage in the OB. *P b 0.05 compared with vehicle-treated rats. Scale bars, 50 μm; SVZ: subventricular zone, SGL: subgranular layer, RMS: rostral migratory stream, OB: olfactory bulb.

Fig. 7 – Capillarization and endothelial proliferation in the SVZ and OB. (a–c) Representative photographs show that some BrdU-positive cells (arrows) are laminin immunopositive in the SVZ and OB. (d) Treatment with G-CSF significantly increased laminin+BrdU+ cells in the SVZ and OB. (e, f) The vascular density was more increased in G-CSF-infused rats than vehicle-infused ones. (g) The mean diameter was significantly greater in the G-CSF-infused rats than in controls. *P b 0.05 compared with vehicle-treated rats. n = 6; Scale bars: (a–c), 50 μm (e, f) 100 μm; SVZ: subventricular zone, OB: olfactory bulb.

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

autonomous growth and tumorigenicity (Moucadel and Constantinescu, 2005). A few studies have identified a neuroprotective effect for erythropoietin, which utilizes STAT5 for transcriptional regulation of anti-apoptotic genes (Renzi et al., 2002; Wen et al., 2002). However, the biological effects of Jak/STAT pathway in human NSCs have not been specifically determined. Because we found preferential STAT5 activation on NSCs by G-CSF treatment, we suggest the potential role of STAT5 in neurogenesis. However, posttranslational modifications, cofactor interactions and downregulation might result in cell type specific responsiveness. The fidelity of STAT5-mediated gene regulation needs to be further defined. In the present study, the upregulation of VEGF and Flk1/KDR was linked with G-CSF treatment. The recent study has indicated that human pancreatic cancer cells exhibited constitutively activated STAT3, with its activation correlated with the VEGF expression level (Yahata et al., 2003). Constitutively activated STAT3 directly activated the VEGF promoter, and blockade of activated STAT3 via ectopic expression of dominant-negative STAT3 suppressed VEGF expression (Yahata et al., 2003). Similarly, the upregulation of VEGF on NSCs by G-CSF might depend on G-CSFR-JakSTAT3 pathway. VEGF is not only an angiogenic factor but a directly-acting neurogenic factor, and it regulates the proliferation and differentiation on neural precursors independent of its effects on endothelial cells (Bonni et al., 1997; Yahata et al., 2003). The recent evidence suggested an internal autocrine loop mechanism, by which VEGF could regulate the proliferation and differentiation of NSCs when co-expressed with its receptors (Maurer et al., 2003). VEGF is regulated by the Flt-1 and Flk-1/KDR receptors, which are part of the tyrosine kinase family. Flk-1/KDR is involved in cell proliferation, while Flt-1 is thought to mainly regulate the formation of cellular architecture (Keyt et al., 1996). In our study, G-CSF induced VEGF secretion from NSCs. NSCs expressed Flk-1/KDR receptors, but not Flt-1, and Flk-1/KDR could be upregulated by G-CSF. VEGFR2 receptor tyrosine kinase inhibitor blocked the effects of G-CSF on cell proliferation. These observations suggest that G-CSF-induced neurogenesis needs VEGF signaling. The combination of G-CSF (50 ng/ml) and VEGF (50 ng/ml) significantly increased the total number of NSCs (+43% vs. control) compared to G-CSF alone (+24% vs. control) or VEGF alone (+27% vs. control). The effect (+43%) of combinational treatment of G-CSF and VEGF was lower than that (+51%) of G-CSF alone plus VEGF alone. This non-additive finding indicated that they might act through a shared signal transduction pathway (Bartoli et al., 2000; Yahata et al., 2003). It is very important to test whether G-CSF effect is blocked and VEGF level is changed by inhibiting Jak/STAT pathway. Further experiments are needed to dissect out the signal transduction mechanisms implicated, e.g. inhibitor peptides, genetic controls including dominant negative STAT or KDR, and ideally genetic knockdown of the growth factor and receptor. We have found that significant angiogenesis takes place within the SVZ and OB after G-CSF treatment. G-CSF stimulates the proliferation of cerebral endothelial cells directly (Bussolino et al., 1991). On the other hand, our findings that

197

VEGF and Flk-1/KDR were upregulated on NSCs themselves by G-CSF suggest an additional mechanism that NSCs may contribute to neovascularization by direct angiogenic effects. In normal brain, the neurogenesis in the adult brain occurs in intimate association with angiogenesis in neurogenic zones of the hippocampus (Palmer et al., 2000), and songbird HVC (Louissaint et al., 2002). Furthermore, factors released by endothelial cells in vitro trigger NSCs proliferation (Shen et al., 2004). G-CSF might act upon the local microvascular bed to provide a permissive environment for neuronal recruitment into the brain. Our results indicate that G-CSF stimulates neurogenesis through reciprocal interaction with VEGF and STAT activation. The coordinate induction of angiogenesis and neurogenesis by a common factor can have critical value, by efficiently assisting the cell replacement therapy. Taken together with the recent evidences that G-CSF can rescue dying neurons (Schäbitz et al., 2003; Schneider et al., 2005), G-CSF might potentially serve to promote brain recovery and repair.

4.

Experimental procedure

4.1.

Human neural stem cell cultures

All experimental procedures were approved by the Care of Experimental Animals Committee of Seoul National University Hospital and by institutional review board for the human cell use. Primary dissociated cell cultures were prepared from embryonic human brains of 15 weeks gestation as described previously in detail (Cho et al., 2002; Chu et al., 2003, 2004; Flax et al., 1998; Jeong et al., 2003; Ourednik et al., 2002). Maintaining NSCs in a proliferative state in culture did not subvert their ability to respond to normal developmental cues in vivo following transplantation (Chu et al., 2003, 2004; Jeong et al., 2003). The cells were initially grown as a polyclonal population first in serum-supplemented and then in serum-free medium containing basic FGF and/or EGF. The cultures were retrovirally transduced with the vmyc oncogene and subsequently cloned. One of the isolated clones, HB1.F3, was expanded for the present study. The cells were grown on poly-L-lysine-coated culture dishes or flasks as single cells or large clusters, which could be subcultured and passaged weekly over a period of 6 months. The cultures were maintained in a serum-free medium consisting of Dulbecco's modified Eagle's medium (DMEM) containing insulin (10 μg/ml), transferrin (10 μg/ ml), sodium selenite (30 nM), hydrocortisone (50 nM), and triiodothyronine (0.3 nM). The human recombinant G-CSF ((Kirin pharmaceuticals, Tokyo, Japan) was added to cultures for 72 h at various concentrations. 4.2.

Quantification of viable cells

Cell viability was assayed by MTT absorbance and by counting cells on photomicrographs of 4′,6-diamidino-2- phenylindole (DAPI, Vector Laboratories). For MTT assays, cultures were incubated with a stock solution of MTT (5 mg/ml in PBS, pH 7.4, Sigma) at 37 °C for 4 h at a final concentration of 1 mg/ml, and absorbance at 570 nm was measured in solubilized cells on the ELISA reader. For cell counting, the number of intact DAPI-stained nuclei in five 200×

198

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

microscope fields per well (at the 3-, 6-, 9-, and 12-o'clock positions

4.6.

Western blot analysis

and in the center) was recorded. In both cases, results were expressed as a percentage of values obtained in control cultures

We examined the potential signaling molecules involved in

not treated with G-CSF.

neurogenesis only at maximally effective concentration. Western blotting was used to determine the expression of G-CSFR on HB1.F3

4.3.

BrdU labeling in vitro

cells and subsequent activation of STAT proteins following G-CSF treatment. We also tested what changes are made by G-CSF

Cells were plated on 12 mm round Aclar plastic coverslips

treatment in the expression of VEGFR1/Flt-1 or VEGFR2/Flk-1 on

previously coated with 10 μg/ml polylysine and housed in 35

HB1.F3 cells. The cells cultured on the 100 mm plate were washed

mm dishes. BrdU (10 μg/ml; Sigma) was added for 72 h to the

with 4 °C PBS and collected. They were homogenized in a lysis

cultures exposed to G-CSF or not. BrdU-positive cells in culture

buffer (100 mmol/l NaCl, 10 mmol/l Tris (pH 7.5), 1 mmol/l EDTA) to

were counted in five fields per well (center and at 3, 6, 9, and 12

which the protease inhibitors (1 mM phenylmethylsulfonyl

o'clock). Cells containing densely red-stained nuclei were consid-

fluoride, 1 mmol/ml aprotinin) were freshly added. The protein

ered BrdU positive, and profiles with smaller or dot-like labeling

concentrations were determined using the Bradford method (Bio-

were considered necrotic cells and were excluded in the analysis.

Rad, Richmond, CA). The protein extracts (30 μg) were separated by

Results were expressed as a percentage of the number of intact

7–15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis

DAPI-stained nuclei obtained in the same fields.

and transferred to a nitrocellulose membrane. They were blocked in 5% non-fat dry milk in TBS (0.15 M NaCl, 25 mM Tris–HCl, 25 mM

4.4.

Immunohistochemistry

Tris) for 2 h and then incubated overnight at 4 °C with anti-G-CSFR (1:500, BD Biosciences), anti- VEGFR-1/Flt-1 (1:500, Santa Cruz

Cell cultures or brain sections were processed for immunocyto-

Biotechnology), anti-VEGFR-2/Flk-1 (1:500, Santa Cruz Biotechnol-

chemistry as described previously (Cho et al., 2002; Chu et al., 2003,

ogy), anti-STAT3 (1:1000, Cell Signaling Technology), anti-pSTAT3

2004; Jeong et al., 2003; Jung et al., 2004). The specimens were fixed

(1:1000, Cell Signaling Technology) or anti-pSTAT5 (1:1000, Cell

in 4% paraformaldehyde in PBS (pH 7.5) for 40 min, washed with

Signaling Technology) antibodies. After washing three times in

PBS, incubated for 2 h with blocking buffer (2% horse serum/1%

TBST (TBS + 0.5% Tween-20), the membrane was incubated with

BSA/0.1% Triton X-100 in PBS, pH 7.5), incubated overnight at 4 °C

the secondary antibody conjugated to horseradish peroxidase at a

with primary antibodies, and then incubated for 1 h at room

1:5000 dilution for 1 h at room temperature. The blots were

temperature with secondary antibodies. Primary antibodies were

developed with enhanced chemiluminescence (Pierce), digitally

as follows: monoclonal anti-BrdU (1:300, Pharmingen); sheep

scanned (GS-700, Bio-Rad) and analyzed (Molecular analyst®, Bio-

polyclonal anti-BrdU antibodies (1:300, BioDesign); anti-Ki67

Rad). The relative optical densities were obtained by comparing the

antibody (1:200, Boeringer-Ingelheim; ABC kit); anti-G-CSFR

measured values with the mean value of the normal control

(1:500, BD Biosciences); anti-Tuj1 (1:200, Chemicon); anti-Dcx

cultures.

(1:100, Santa Cruz Biotechnology); anti-NeuN (1:200, Chemicon); anti-GFAP (1:1000, Sigma); anti-Calbindin (1:200, Sigma); anti-

4.7.

VEGF enzyme-linked immunosorbent assay (ELISA)

laminin (1:100, Sigma). Biotinylated goat anti-mouse IgG (ABC, sigma) and FITC-conjugated anti-sheep IgG (1:100, Biodesign) or

Concentrations of VEGF in the culture supernatants treated with

Cy3-conjugated anti-mouse IgG antibodies (1:300, Jackson Immu-

or without G-CSF were measured using the Quantikine VEGF ELISA

noresearch) were used for the secondary antibodies. DAPI (Vector

kit (R&D Systems, Minneapolis, MN), which is a quantitative

Laboratories) was used to counterstain nuclei in some experi-

immunometric sandwich enzyme immunoassay. A curve of the

ments. The colocalization was analyzed using a laser scanning

absorbance of VEGF versus its concentration in the standard wells

confocal microscopy with a Bio-Rad MRC 1024 (argon and krypton).

was plotted. By comparing the absorbance of the samples with the

The fluorescence signals were detected at excitation/emission

standard curve, we determined the VEGF concentration in the

wavelengths of 550/570 nm (Cy3, red), 488/522 (FITC, green), and

unknown samples.

360/400 (DAPI, blue). Biotin signals were detected with 3,3′diaminobenzidine (DAB; brown) and Vector VIP (purple).

4.8.

Co-treatment with G-CSF and VEGF or VEGFR tyrosine kinase

inhibitor 4.5.

Cell cycle analysis To investigate the potential role of VEGF on G-CSF action, the

The cell cycle of HB1.F3 cells treated with or without G-CSF was

cultures treated with G-CSF were also placed with the recombi-

analyzed by FACScalibur flow cytometer (Becton Dickinson,

nant VEGF165 (50 ng/ml, Calbiochem) or VEGFR tyrosine kinase

Mountain View, CA). At 72 h after treatment, the cells were

inhibitor, SU1498 (10 μM, Calbiochem) throughout the time course

harvested by centrifugation, counted, and prepared for cell cycle

of the experiments (Jin et al., 2002b). Viable cell number was

analysis. Cells (5 × 104) were suspended in 200 μl of hypotonic

quantified by MTT assay and by counting cells on DAPI staining

propidium iodide (PI) solution (0.1% [w/v] sodium citrate, 0.1% [v/v]

and cell cycle was analyzed in each experiment.

Triton X-100, 0.05 mg/ml PI) and were incubated in the dark at 4 °C for 24 h before analysis. The cell histogram FL-2 was divided into 3

4.9.

regions according to cell cycle phase: G0/G1, S, or G2/M. Doublets

containing G-CSF or saline alone

Implantation and use of osmotic minipumps

were eliminated by gating on a peak/area plot of PI fluorescence. The data were analyzed with WinCycle software (Phoenix Flow

We investigated whether the intraventricular infusion of G-CSF

Systems, San Diego).

increases the proliferation of new neurons in the adult rodents.

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

199

Twelve Sprague–Dawley male rats (n = 6 for G-CSF or vehicle

bulb were determined. Endothelial cell division was detected by

alone-treated groups) weighing 220 to 240 g (Genomics) were used.

double staining for BrdU and laminin. BrdU-immunostained

The osmotic minipumps (Alzet, model 2002, Palo Alto, CA)

sections were digitized using a ×40 objective (Olympus BX51) via

contained 0.9% saline or recombinant G-CSF at the concentrations

the image analysis system (Image-Pro Plus™). The numbers of

of 10 μg/ml in saline. The pumps were designed to deliver the

total endothelial cells and the numbers of BrdU+ endothelial cells

contents at 0.5 μl/h for 7 days. The G-CSF concentration would,

in 20 enlarged and thin-walled vessels in the SVZ and OB were

therefore, deliver a total of 1 μg G-CSF over the week. It had been

counted from each rat (n = 6 per group), from which the

tested for safety in a pilot project, and no significant side effects

proportions of BrdU-positive endothelial cells were calculated.

were observed. The filled pumps were incubated overnight in sterile saline at 37 °C before use. Surgery was performed under the

4.12.

Statistical analysis

anesthesia with an intraperitoneal injection of 1% ketamine (30 mg/kg) and xylazine hydrochloride (4 mg/kg) as reported previ-

All data in this study are presented as mean ± standard devia-

ously (Jung et al., 2004). The cannula was placed in the left lateral

tions. Data were analyzed by repeated measures of analysis of

ventricle 4.0 mm deep to the pial surface and +0.0 mm

variance, and unpaired Student's t test, if they were normally

anteroposterior relative to bregma and 1.8 mm lateral to the

distributed (Kolmogorov–Smirov test, P N 0.05). Otherwise, we

midline. All cannula (30 gauge, 4 mm length; Plastics One,

used Mann–Whitney U test and specified the test used. P values

Roanoke, VA) were cemented (Cranioplastic cement; Plastics

of less than .05 were considered to indicate a statistically

One) to the skull, anchored with machine screws (Plastics One),

significant difference.

and connected to the osmotic minipump by medical grade vinyl tubing (Biolab Inc., Decatur, GA). All pumps were implanted subcutaneously in the nape of the neck, and the cannula was

Acknowledgments

lowered after the pumps began pumping. 4.10.

Quantification of BrdU-positive cells in vivo

For in vivo studies, BrdU (100 mg/kg) was administered intraper-

This study was supported by the 21st Century Frontier Research Fund of the Ministry of Science and Technology (SC3060), Republic of Korea. G-CSF was kindly provided by Kirin Pharmaceuticals. Conflict of interest: None to disclose.

itonealy once daily for 7 consecutive days from the day of pump insertion, as reported previously. The rats were sacrificed 7 days after G-CSF or vehicle withdrawal (day 14) to permit newly

REFERENCES

generated cells to integrate in the host brain (n = 6 for each group). BrdU-positive cells were characterized by double-labeling with neuronal or glial cell markers. BrdU-positive cells were counted in five to seven 30 μm coronal sections per animal, spaced 200 μm apart, by two investigators blinded for group allocation. We selected for comparisons corresponding coronal sections, determined using the rat atlas (Paxinos et al., 1997) in both the G-CSFinfused and saline-infused brains. The counts were made, using a 40× objective, by placing an optical grid (field size, 250 μm × 250 μm). The density of BrdU-labeled cells, expressed as cells per cubic millimeter, was determined in the SVZ, SGL, RMS and OB. The number of double-labeled BrdU+Tuj1+, BrdU+CB1+, and BrdU+GFAP+ cells in the SVZ, SGL, RMS, and OB was counted with a conventional or confocal microscope. 4.11.

Capillary parameters and endothelial proliferation

The vasculature was analyzed using a rabbit anti-laminin IgG (1:100, Sigma). Sections were permeabilized by 0.1% saponin for 15 min, blocked with 10% goat serum for 1 hr, and exposed to antilaminin IgG overnight at 4 °C. After washing, the primary antilaminin antibody was detected with Alexa Fluor 566-conjugated goat anti-rabbit IgG (1:400, Molecular Probes). In each sampled olfactory bulb, every laminin-positive profile was traced and recorded manually into image analysis system (Image-Pro Plus™). A profile was defined as the net surface area of laminin immunolabeling. Capillary counts were derived from these profiles, and no correction was made for tangential cuts. This was based upon an assumption of randomly distributed capillary orientations in coronal sections. Each traced outline was then analyzed for its diameter, and the average values per olfactory

Aberg, M.A., Aberg, N.D., Hedbacker, H., Oscarsson, J., Eriksson, P. S., 2000. Peripheral infusion of IGF-I selectively induces neurogenesis in the adult rat hippocampus. J. Neurosci. 20, 2896–2903. Aloisi, F., Care, A., Borsellino, G., Gallo, P., Rosa, S., Bassani, A., Cabibbo, A., Testa, U., Levi, G., Peschle, C., 1992. Production of hemolymphopoietic cytokines (IL-6, IL-8, colony-stimulating factors) by normal human astrocytes in response to IL-1 beta and tumor necrosis factor-alpha. J. Immunol. 149, 2358–2366. Bartoli, M., Gu, X., Tsai, N.T., Venema, R.C., Brooks, S.E., Marrero, M.B., Caldwell, R.B., 2000. Vascular endothelial growth factor activates STAT proteins in aortic endothelial cells. J. Biol. Chem. 275, 33189–33192. Bonni, A., Sun, Y., Nadal-Vicens, M., Bhatt, A., Frank, D.A., Rozovsky, I., Stahl, N., Yancopoulos, G.D., Greenberg, M.E., 1997. Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway. Science 278, 477–483. Bussolino, F., Ziche, M., Wang, J.M., Alessi, D., Morbidelli, L., Cremona, O., Bosia, A., Marchisio, P.C., Mantovani, A., 1991. In vitro and in vivo activation of endothelial cells by colonystimulating factors. J. Clin. Invest. 87, 986–995. Cao, L., Jiao, X., Zuzga, D.S., Liu, Y., Fong, D.M., Young, D., During, M.J., 2004. VEGF links hippocampal activity with neurogenesis, learning and memory. Nat. Genet. 36, 827–835. Cho, T., Bae, J.H., Choi, H.B., Kim, S.S., McLarnon, J.G., Kim, H.S., Kim, S.U., Min, C.K., 2002. Human neural stem cells: electrophysiological properties of voltage-gated ion channels. NeuroReport 13, 1447–1452. Chu, K., Kim, M., Jeong, S.W., Kim, S.U., Yoon, B.W., 2003. Human neural stem cells can migrate, differentiate, and integrate after intravenous transplantation in adult rats with transient forebrain ischemia. Neurosci. Lett. 343, 129–133. Chu, K., Kim, M., Park, K.I., Jeong, S.W., Park, H.K., Jung, K.H., Lee, S. T., Kang, L., Lee, K., Park, D.K., Kim, S.U., Roh, J.K., 2004. Human

200

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

neural stem cells improve sensorimotor deficits in the adult rat brain with experimental focal ischemia. Brain Res. 1016, 145–153. Cosman, D., 1993. The hematopoietin receptor superfamily. Cytokine 5, 95–106. Flax, J.D., Aurora, S., Yang, C., Simonin, C., Wills, A.M., Billinghurst, L.L., Jendoubi, M., Sidman, R.L., Wolfe, J.H., Kim, S.U., Snyder, E.Y., 1998. Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign genes. Nat. Biotechnol. 16, 1033–1039. Grana, X., Reddy, E.P., 1995. Cell cycle control in mammalian cells: role of cyclins, cyclin dependent kinases (CDKs), growth suppressor genes and cyclin-dependent kinase inhibitors (CKIs). Oncogene 11, 211–219. Hofer, M., Vacek, A., Weiterova, L., 2005. Action of granulopoiesis-stimulating cytokines rhG-CSF, rhGM-CSF, and rmGM-CSF on murine haematopoietic progenitor cells for granulocytes and macrophages (GM-CFC). Physiol. Res. 54, 207–213. Ihle, J.N., 1995. Cytokine receptor signaling. Nature 377, 591–594. Ihle, J.N., 1996. STATs: signal transducers and activators of transcription. Cell 84, 331–334. Jeong, S.W., Chu, K., Jung, K.H., Kim, S.U., Kim, M., Roh, J.K., 2003. Human neural stem cell transplantation promotes functional recovery in rats with experimental intracerebral hemorrhage. Stroke 34, 2258–2263. Jin, K.L., Mao, X.O., Sun, Y., Xie, L., Greenberg, D.A., 2002a. Stem cell factor stimulates neurogenesis in vitro and in vivo. J. Clin. Invest. 110, 311–319. Jin, K.L., Zhu, Y., Sun, Y., Mao, X.O., Xie, L., Greenberg, D.A., 2002b. Vascular endothelial growth factor (VEGF) stimulates neurogenesis in vitro and in vivo. Proc. Natl. Acad. Sci. U. S. A. 99, 11946–11950. Jung, K.H., Chu, K., Kim, M., Jeong, S.W., Song, Y.M., Lee, S.T., Kim, J.Y., Lee, S.K., Roh, J.K.K., 2004. Continuous cytosine-bD-arabinofuranoside infusion reduces ectopic granule cells in adult rat hippocampus with attenuation of spontaneous recurrent seizures following pilocarpine-induced status epilepticus. Eur. J. Neurosci. 19, 3219–3226. Kaplan, M.H., Daniel, C., Schindler, U., Grusby, M.J., 1998. Stat proteins control lymphocyte proliferation by regulating p27Kip1 expression. Mol. Cell. Biol. 18, 1996–2003. Keyt, B.A., Nguyen, H.V., Berleau, L.T., Duarte, C.M., Park, J., Chen, H., Ferrara, N., 1996. Identification of vascular endothelial growth factor determinants for binding KDR and FLT-1 receptors. Generation of receptor-selective VEGF variants by site-directed mutagenesis. J. Biol. Chem. 271, 5638–5646. Li, K., Li, C.K., Chuen, C.K., Tsang, K.S., Fok, T.F., James, A.E., Lee, S.M., Shing, M.M., Chik, K.W., Yuen, P.M., 2005. Preclinical ex vivo expansion of G-CSF-mobilized peripheral blood stem cells: effects of serum-free media, cytokine combinations and chemotherapy. Eur. J. Haematol. 74, 128–135. Liu, C., Chu, I., Hwang, S., 2001. Factorial designs combined with the steepest ascent method to optimize serum-free media for CHO cells. Enzyme Microb. Technol. 28, 314–321. Louissaint, A., Rao, S., Leventhal, C., Goldman, S.A., 2002. Coordinated interaction of neurogenesis and angiogenesis in the adult songbird brain. Neuron 34, 945–960. Luskin, M.B., 1993. Restricted proliferation and migration of postnatally generated neurons derived from the forebrain subventricular zone. Neuron 11, 173–189. Magavi, S.S., Leavitt, B.R., Macklis, J.D., 2000. Induction of neurogenesis in the neocortex of adult mice. Nature 405, 951–955. Maurer, M.H., Tripps, W.K., Feldmann, R.E., Kuschinsky, W., 2003. Expression of vascular endothelial growth factor and its receptors in rat neural stem cells. Neurosci. Lett. 344, 165–168.

Moucadel, V., Constantinescu, S.N., 2005. Differential STAT5 signaling by ligand-dependent and constitutively active cytokine receptors. J. Biol. Chem. 280, 13364–13373. Natori, T., Sata, M., Washida, M., Hirata, Y., Nagai, R., Makuuchi, M., 2002. G-CSF stimulates angiogenesis and promotes tumor growth: potential contribution of bone marrow-derived endothelial progenitor cells. Biochem. Biophys. Res. Commun. 297, 1058–1061. Ourednik, V., Ourednik, J., Flax, J.D., Zawada, W.M., Hutt, C., Yang, C., Park, K.I., Kim, S.U., Sidman, R.L., Freed, C.R., Snyder, E.Y., 2002. Segregation of human neural stem cells in the developing primate forebrain. Science 293, 1820–1824. Palmer, T.D., Willhoite, A.R., Gage, F.H., 2000. Vascular niche for adult hippocampal neurogenesis. J. Comp. Neurol. 425, 479–494. Paxinos, G., Watson, C., 1997. The Rat Brain in Stereotaxic Coordinates. Academic Press, Inc., San Diego. Pencea, V., Bingaman, K.D., Wiegand, S.J., Luskin, M.B., 2001. Infusion of brain-derived neurotrophic factor into the lateral ventricle of the adult rat leads to new neurons in the parenchyma of the striatum, septum, thalamus, and hypothalamus. J. Neurosci. 21, 6706–6717. Petit, I., Szyper-Kravitz, M., Nagler, A., Lahav, M., Peled, A., Habler, L., Ponomaryov, T., Taichman, R.S., Arenzana-Seisdedos, F., Fujii, N., Sandbank, J., Zipori, D., Lapidot, T., 2002. G-CSF induces stem cell mobilization by decreasing bone marrow SDF-1 and up-regulating CXCR4. Nat. Immunol. 3, 687–694. Rane, S.G., Reddy, E.P., 2000. Janus kinases: components of multiple signaling pathways. Oncogene 19, 5662–5679. Ray, J., Peterson, D.A., Schinstine, M., Gage, F.H., 1993. Proliferation, differentiation, and long-term culture of primary hippocampal neurons. Proc. Natl. Acad. Sci. U. S. A. 90, 3602–3606. Renzi, M.J., Farrell, F.X., Bittner, A., Galindo, J.E., Morton, M., Trinh, H., Jolliffe, L.K., 2002. Erythropoietin induces changes in gene expression in PC-12 cells. Brain Res. Mol. Brain Res. 104, 86–95. Reynolds, B.A., Weiss, S., 1992. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255, 1707–1710. Schäbitz, W.R., Kollmar, R., Schwaninger, M., Juettler, E., Bardutzky, J., Scholzke, M.N., Sommer, C., Schwab, S., 2003. Neuroprotective effect of granulocyte colony-stimulating factor after focal cerebral ischemia. Stroke 34, 745–751. Schneider, A., Kruger, C., Steigleder, T., Weber, D., Pitzer, C., Laage, R., Aronowski, J., Maurer, M.H., Gassler, N., Mier, W., Hasselblatt, M., Kollmar, R., Schwab, S., Sommer, C., Bach, A., Kuhn, H.G., Schabitz, W.R., 2005. The hematopoietic factor G-CSF is a neuronal ligand that counteracts programmed cell death and drives neurogenesis. J. Clin. Invest. 115, 2083–2098. Shen, Q., Goderie, S.K., Jin, L., Karanth, N., Sun, Y., Abramova, N., Vincent, P., Pumiglial, K., Temple, S., 2004. Endothelial cells stimulate self-renewal and expand neurogenesis of neural stem cells. Science 304, 1338–1340. Sherr, C.J., Roberts, J.M., 1999. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 13, 1501–1512. Shimoda, K., Feng, J., Murakami, H., Nagata, S., Watling, D., Rogers, N.C., Stark, G.R., Kerr, I.M., Ihle, J.N., 1997. Jak1 plays an essential role for receptor phosphorylation and Stat activation in response to granulocyte colony-stimulating factor. Blood 90, 597–604. Shingo, T., Sorokan, S.T., Shimazaki, T., Weiss, S., 2001. Erythropoietin regulates the in vitro and in vivo production of neuronal progenitors by mammalian forebrain neural stem cells. J. Neurosci. 21, 9733–9743. Shuai, K., Ziemiecki, A., Wilks, A.F., Harpur, A.G., Sadowski, H.B., Gilman, M.Z., Darnell, J.E., 1993. Polypeptide signalling to the nucleus through tyrosine phosphorylation of Jak and Stat proteins. Nature 366, 580–583.

BR A I N R ES E A RC H 1 0 7 3–1 0 7 4 ( 2 00 6 ) 1 9 0 –2 01

Sondell, M., Lundborg, G., Kanje, M., 1999. Vascular endothelial growth factor has neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and Schwann cell proliferation in the peripheral nervous system. J. Neurosci. 19, 5731–5740. Sun, Y., Jin, K.L., Xie, L., Childs, J., Mao, X.O., Logvinova, A., Greenberg, D.A., 2003. VEGF-induced neuroprotection, neurogenesis, and angiogenesis after focal cerebral ischemia. J. Clin. Invest. 111, 1843–1851. Tian, S.S., Lamb, P., Seidel, H.M., Stein, R.B., Rosen, J., 1994. Rapid activation of the STAT3 transcription factor by granulocyte colony-stimulating factor. Blood 84, 1760–1764. Tian, S.S., Tapley, P., Sincich, C., Stein, R.B., Rosen, J., Lamb, P., 1996. Multiple signaling pathways induced by granulocyte colony-stimulating factor involving activation of JAKs, STAT5, and/or STAT3 are required for regulation of three distinct classes of immediate early genes. Blood 88, 4435–4444. Wei, D., Le, X., Zheng, L., Wang, L., Frey, J.A., Gao, A.C., Peng, Z.,

201

Huang, S., Xiong, H.Q., Abbruzzese, J.L., Xie, K., 2003. Stat3 activation regulates the expression of vascular endothelial growth factor and human pancreatic cancer angiogenesis and metastasis. Oncogene 22, 319–329. Wen, T.C., Sadamoto, Y., Tanaka, J., Zhu, P.X., Nakata, K., Ma, Y.J., Hata, R., Sakanaka, M., 2002. Erythropoietin protects neurons against chemical hypoxia and cerebral ischemic injury by up-regulating Bcl-xL expression. J. Neurosci. Res. 67, 795–803. Wesselingh, S.L., Gough, N.M., Finlay-Jones, J.J., McDonald, P.J., 1990. Detection of cytokine mRNA in astrocyte cultures using the polymerase chain reaction. Lymphokine Res. 9, 177–185. Yahata, Y., Shirakata, Y., Tokumaru, S., Yamasaki, K., Sayama, K., Hanakawa, Y., Detmar, M., Hashimoto, K., 2003. Nuclear translocation of phosphorylated STAT3 is essential for vascular endothelial growth factor-induced human dermal microvascular endothelial cell migration and tube formation. J. Biol. Chem. 278, 40026–40031.