Interferon-γ Produced by Microglia and the Neuropeptide PACAP Have Opposite Effects on the Viability of Neural Progenitor Cells

Inflammation is part of many neurological disorders and immune reactions may influence neuronal progenitor cells (NPCs) contributing to the disease process. Our knowledge about the interplay between different cell types in brain inflammation are not fully understood. It is important to know the mechanisms and factors involved in order to enhance regeneration and brain repair. We show here that NPCs express receptors for interferon-γ (IFNγ), and IFNγ activates the signal transducer and activator of transcription (STAT) protein-1. IFNγ reduced cell proliferation in NPCs by upregulation of the cell cycle protein p21 as well as induced cell death of NPCs by activating caspase-3. Studies of putative factors for rescue showed that the neuropeptide, Pituitary adenylate cyclase-activating polypeptide (PACAP) increased cell viability, the levels of p-Bad and reduced caspase-3 activation in the NPCs. Medium from cultured microglia contained IFNγ and decreased the viability of NPCs, whilst blocking with anti-IFNγ antibodies counteracted this effect. The results show that NPCs are negatively influenced by IFNγ whereas PACAP is able to modulate its action. The interplay between IFNγ released from immune cells and PACAP is of importance in brain inflammation and may affect the regeneration and recruitment of NPCs in immune diseases. The observed effects of IFNγ on NPCs deserve to be taken into account in human anti-viral therapies particularly in children with higher rates of brain stem cell proliferation.


Introduction
The nervous system interacts with the immune system during inflammation that is part of many neurodegenerative diseases. Cytokines secreted by immune cells mediate the effects of inflammation in the brain. Increased production of cytokines is observed in different brain disorders in experimental animals and in humans [1]. Our knowledge about the inflammatory process in the brain and the interplay between different cell types in inflammation are not fully understood [1][2][3]. It is important to know the different mechanisms and factors that underlie cell reactions in brain in order to enhance regeneration and brain repair.
NPCs are present in the developing neuroepithelium and in neurogenic areas in the adult brain [4,5]. NPCs are self-renewing cells that give rise to neuroblast and glial cells in the nervous system. Different factors in the local milieu influence cell proliferation and differentiation of NPCs [6][7][8]. NPCs have been shown to react to tissue trauma as a part of the defense mechanism. Chronic inflammation was shown to impair neurogenesis and negatively influence neuronal stem cells in the rodent hippocampus [2,9]. In line with this, reduced brain inflammation using anti-inflammatory drugs restores neurogenesis in rat hippocampus [2] and after brain ischemia [10]. On the other hand, glucocorticoid hormones, which are increased after stress and immune activation, reduce neurogenesis and the proliferation of NPCs [11]. The roles of different cytokines and their interactions in the regulation of NPCs are so far largely unknown.
In this work, we have studied the Interferons (IFN) family of cytokines, which are synthesized and secreted by different cells types during inflammation and in immune reactions [12]. We observed that NPCs express IFNc receptors (IFNcR) in vitro and in vivo, and that stimulation with IFNc activates STAT1 signaling in the NPCs. IFNc caused a decrease in cell viability of NPCs accompanied by reduced cell proliferation and increased cell death. One source of IFNc in the brain is microglial cells that produce increasing amounts of cytokines after cell activation [1,3,12,13]. Cultivation of NPCs with medium from microglia decreased cell viability that was rescued by the addition of the neuropeptide PACAP. These results reveal an important interaction between NPCs and microglial cells that involves the cytokine IFNc and neuropeptide PACAP and which is probably of importance in brain inflammation and disease.

NPCs express receptors for IFNc
In the brain, astrocytes and neurons have been shown to express IFNcR [12,14]. Immunostaining of embryonic NPCs cultured as neurospheres and colabelled with the marker nestin also expressed IFNc receptor-2 (IFNcR2, Fig. 1A). Labeling with BrdU showed that the IFNcR2 positive cells actively divide in the cultures (Fig. 1a). Semiquantitative PCR confirmed the expression of the IFNcR-2 in the NPCs (Fig. 1B). IFNcR2 was also observed with immunoblotting of NPCs and the receptors were present in developing neuroepithelium containing nestin positive NPCs (Fig. 1C).

IFNc decreases viability of NPCs and affects cell proliferation
Treatment of NPCs with 100 ng/ml IFNc induced the rapid phosphorylation of STAT1 ( Fig. 2A), with translocation of the protein into the nucleus (Fig. 2B). IFNc also decreased the viability of NPCs (Fig. 2C). Dose response curve showed that the halfmaximum effect of IFNc was about 3 ng/ml (Fig. 2D). Immuno-staining confirmed that the number of nestin positive NPCs decreased after IFNc treatment (Fig. 2E), and IFNc reduced the number of secondary neurospheres formed in the cultures (Fig. 2F).
A decrease in cell viability may be due to reduced cell proliferation and/or enhanced cell death. Data showed that IFNc significantly reduced the number of BrdU labeled NPCs (Fig. 3A). Similar results were obtained using antibodies against the proliferation antigen Ki67 (Fig. 3B). These results show that IFNc decreases cell proliferation of NPCs that contributes to the reduced number of NPCs after IFNc.
Previous studies have shown that cell cycle regulators, including p53 and cyclinD1 influence cell proliferation in NPCs [11,15,16]. The addition of IFNc did not influence cyclinD1 nor p53 levels, as shown by immunoblotting (Fig. 3C). However, IFNc increased the levels of p21 protein in the NPCs (Fig. 3C), which was also Figure 1. Neural progenitor cells express IFNc receptors. NPCs were prepared from embryonic, E17 old rat brain and cultured as described in Methods. (A) Upper panel. Immunostaining using antibodies against the IFNcR2 receptor (green fluorescence) and against nestin (red fluorescence) as a marker for NPCs. Control without primary antibody showed no staining. Lower panel. BrdU labeling was done as described in Methods. Note expression of IFNcR2 in dividing NPCs. Scale bar, 10 mm. (B) Immunoblot shows the presence of IFNcR2. b-actin was used as control. (C) Sections from E17 rats were double-stained using antibodies against nestin and IFNcR2. Note coexpression in cells in neuroepithelium. Scale bar, 90 mm. doi:10.1371/journal.pone.0011091.g001 observed using RT-PCR (Fig. 3D). To study p21 more closely, we employed silencing RNA (siRNA) against p21. Treatment with p21 siRNA reduced the effect of IFNc on cell proliferation as shown here by analyzing phospho-vimentin (p-vimentin) levels and the number of BrdU labeled cells (Fig. 3E) [17]. In contrast, lowering p53 levels by siRNA did not influence the decrease in cell proliferation caused by IFNc (Fig. 3F). This shows that p53 is indispensable for the IFNc-induced decrease in cell proliferation in the NPCs. In line with this, we observed that the number of p53 deficient mouse embryonic fibroblasts was also reduced by IFNc, showing a p53 independent action for IFNc to regulate cell viability (Fig. 3G).

IFNc induces cell death and activates caspase-3 in NPCs
Cell cycle analyses of NPCs revealed that IFNc decreased the Sphase together with an increase in the sub-G0 phase (data not shown). We therefore studied whether enhanced cell death can contribute to reduced cell viability of NPCs observed with IFNc. Staining of cells using propidium iodide or Terminal deoxynu-cleotidyl transferase-mediated biotinylated UTP nick end labeling (TUNEL) to analyze DNA fragmentation showed that IFNc increased cell death of NPCs ( Fig. 4. A, B). To study the mechanism involved, we analyzed caspase-3, a key caspase involved in cell death. Immunoblotting showed that IFNc caused cleavage of caspase-3 and its downstream substrate, poly-ADP ribose polymerase (PARP) (Fig. 4C). Addition of BAF, a large spectrum caspase inhibitor, reduced cell death induced by IFNc (Fig. 4D). Immunostaining showed that cytochrome-c was present in the cytoplasm of IFNc-treated NPCs but not in control cells (data not shown). The release of cytochrome-c is controlled by the action of Bcl-2 family proteins on mitochondria [18]. As shown below, IFNc increased the levels of the pro-apoptotic protein PUMA in NPCs.

The neuropeptide PACAP increases cell viability of IFNctreated NPCs
To search for putative rescue factors to counteract the negative effects of IFNc we incubated NPCs in the presence of various  proteins and cytokines with receptors on NPCs. The results showed that none of the growth factors examined, including brainderived neurotrophic factor (BDNF) was able to increase viability of NPCs compromised by IFNc (Fig. 5A). In addition, IFNa or IFNb did not counteract the effect of IFNc (Fig. 5A). Cytokines, including IL-10 and TGFb, which have an immune-suppressive function, had no protective effect in this system, nor did antiinflammatory drugs including indomethacin (Fig. 5B,C). Analyses of neuropeptides that are endogenous factors in brain tissue revealed that PACAP [19] increased cell viability (Fig. 5D). PACAP efficiently counteracted the effect of IFNc at nM concentrations that is known to activate the high affinity PACAP receptor, PAC1 (Fig. 5E).

Mechanisms of PACAP action for cell rescue
IFNc signaling is tightly controlled at different levels, including dephosphorylation of p-STAT1 and the induction of the suppressor of cytokine signaling (SOCS) proteins [20]. PACAP did not influence the level of IFNc receptors (data not shown), nor the phoshorylation of STAT1 induced by IFNc (Fig. 6A). The level of SOCS1 increased in the NPCs by IFNc, whereas those of SOCS3 did not (Fig. 6B).
Studies of cell death regulators showed that PACAP did not influence the anti-apoptotic protein Bcl-X in the NPCs, but decreased the level of the pro-apoptotic protein PUMA that was upregulated by IFNc (Fig. 6C). PACAP also increased the phosphorylation of Bad at serine-112 (Fig. 6D), which leads to inactivation of this pro-apoptotic protein [21].
To study the signaling pathways underlying the effect of PACAP, we employed inhibitors against protein kinase A (PKA) and protein kinase C (PKC). The PKC inhibitor, Gö6976, reduced the beneficial effect of PACAP on cell viability, whilst the PKA inhibitor, H89 did not (Fig. 6E). These results show that the effect of PACAP in increasing cell viability in NPCs involves PKC.
PACAP has previously been shown to enhance cell proliferation in adult neural stem cells [22,23]. We observed that PACAP treated cells. Lower panel, BrdU labeling was decreased in control but not in p21-siRNA treated cells. (F) NPCs were treated with siRNA against p53 followed by IFNc as above. IFNc reduces cell proliferation in p53-siRNA treated cells. (G) p53 gene deleted mouse embryonic fibroblasts were treated with 10-100 ng/ml IFNc for 2 days and cell viability determined as described in Methods. doi:10.1371/journal.pone.0011091.g003 increased proliferation of control NPCs, but the effect was not statistically significant in NPCs treated with IFNc (Fig. 6F). Therefore in this situation PACAP acts mainly as a survival promoting peptide reducing the inhibitory effect of IFNc.

Microglia affect NPCs via IFNc
In the brain, microglial cells represent a defense system that takes part in inflammation and in immune reactions. Activation of microglial cells is followed by increased production of various cytokines. We observed that activation of rat microglial cells by lipopolysaccaride (LPS) led to increased levels of IFNc in the culture medium as shown by ELISA (Fig. 7A). Addition of the LPS-conditioned medium to NPCs reduced cell viability by up to 40% compared with media from control cells (Fig. 7B left panels). Supplementation of the conditioned media with 50 nM PACAP counteracted the decrease in viability of NPCs (Fig. 7B right panels). LPS itself had no effect on NPCs viability excluding nonspecific effects of the compound (Fig. 7C). Incubation of the microglia-derived medium with anti-IFNc blocking antibodies largely reduced its negative effect on cell viability (Fig. 7D). These results show that microglia produce IFNc and that PACAP is able to counteract the decrease in cell viability observed with the microglia-conditioned medium.

Discussion
The present results show that IFNc influences NPCs in two ways by decreasing cell proliferation and by increasing cell death. IFNc affected cell proliferation by inducing expression of p21, and cell death by influencing various Bcl-2 family proteins and caspase- 3 activation. Previous studies have shown that cell cycle regulators including p53 and cyclinD1 are important for cell proliferation of NPCs [11,15,16], but we observed no significant change in these proteins in IFNc-treated NPCs. In addition, IFNc reduced cell proliferation also in p53 deficient fibroblasts and in NPCs after downregulation of p53 by siRNA. p21 is a downstream target for p53 but the protein may be regulated also by other factors. We observed that IFNc increased the expression of p21 in NPCs, leading to a decrease in cell proliferation. In line with this, IFNc has been shown to regulate p21 in human breast cancer cells via transcriptional upregulation [24].
IFNs are divided into two major classes: type-1 IFNs, with the structurally related IFNa, IFNb, and type-2 IFN, with IFNc [12]. The IFNs bind to specific IFN receptors on target cells and activate gene transcription through the STAT and Janus tyrosine kinase signaling pathway [25]. IFN receptors are predominantly expressed by the immune cells, but in the brain astrocytes and neurons also show expression [12,14]. Administration of IFNs including IFNc has been shown to modulate neuronal activity and to alter behavior in experimental animals [12]. The exact roles of IFNc in brain physiology and its cellular targets are, however, not fully understood.
In the present study, we observed a high abundance of IFNc receptors in neural precursor cells in neuroepithelium during rat embryonic development and in NPCs cells in culture. IFNc activated STAT-1 signaling and the nuclear translocation of the protein. STAT-1 in turn activates transcription of downstream genes including those affecting inflammation, cell signaling and cell survival [20,25]. IFNc was found to increase levels of the BH-3 only protein PUMA in the NPCs, which can induce the mitochondrial pathway of cell death [18,21,26]. In IFNc treated cells cytochrome-c was released into the cytoplasm with activation of caspase-3, the cleavage of PARP and with increased DNA fragmentation. Incubation with the caspase-inhibitor, BAF protected NPCs against cell death induced by IFNc. These results show that IFNc influences cell death in NPCs by influencing a subset of death-regulating proteins including PUMA with the activation of caspase-3.
The observed decrease in cell viability induced by IFNc may negatively influence the survival and recruitment of NPCs after brain injury or in different diseases. We therefore studied a variety of growth factors and cytokines with described receptors on NPCs for their ability to counteract the effects of IFNc. Of the many factors examined, the neuropeptide PACAP had the largest effect in increasing cell viability. PACAP is a member of the vasoactive intestinal peptide/secretin/glucagon peptide family [19] and is present in embryonic and adult brain tissue [27][28][29][30][31][32]. PACAP binds to its high-affinity PAC1 receptor that is expressed by different target cells [33,34], including NPCs [27]. Previous studies have shown that PACAP increases proliferation/survival of adult neural stem cells [2,23,35]. PACAP also increases neuronal survival in different systems [31,36], however, this action may involve other growth factors, such as insulin growth factor-1 [33] and interleukin-6 (IL-6) [37]. We observed that neither IL-6 nor the other neurotrophic factors studied counteract the IFNcinduced loss of cell viability. In contrast, PACAP at low concentrations exerted a robust protection against cell degeneration of NPCs induced by IFNc, The enhanced cell viability induced by PACAP was accompanied by increases in p-Bad that can inhibit the mitochondrial mediated cell death [21]. In line with this, the activation of caspase-3 by IFNc was reduced in PACAP-treated NPCs. This data suggests that alterations in expression and phosphorylation of Bcl-2 family proteins may underlie the increase in NPC viability induced by PACAP.
Inhibitor studies further showed that increased cell viability by PACAP was mainly mediated by an activation of PKC in NPCs. Molecular cloning has revealed the existence of six different splice variants of the PAC1 receptors, which can activate cell signaling of PACAP either through PKC or PKA [33]. These splice variants are also present in neural stem cells and mediate various cellular responses to PACAP [22,23,35].
Studying the possible crosstalk between PACAP and IFNc, we observed that PACAP did not influence STAT1 signaling nor did PACAP affect the IFNc receptor levels in NPCs (Fig. 6). IFNc receptor signaling is a complex process controlled by several factors including the SOCS proteins [20]. As studied here, SOCS-1 and -3 are expressed by NPCs but these proteins were not altered by PACAP. Although the signaling cascades induced by IFNc and PACAP do not directly overlap, some genes including the Bcl-2 family proteins are differentially regulated by these two factors. Future studies using gene profiling will reveal which other proteins are regulated by PACAP and by IFNc in the NPCs.
Apart from cell survival, PACAP may increase viability of IFNctreated NPCs by enhancing cell division [22]. We observed a slight increase in cell proliferation by PACAP after IFNc treatment but this was not statistically significant as shown using BrdU labeling (Fig. 6G). We therefore conclude that the major effect of PACAP on IFNc-treated NPCs was to promote cell survival by inhibition of caspase-3 activation and increasing p-Bad levels.
Brain tissue is usually though as being protected from immunological reactions in the body by the presence of the blood brain barrier. The immune and nervous systems interact with each other in conditions characterized by increased leakage of the blood brain barrier that occurs after an immune challenge or in brain inflammation. The activation of immune cells occurs in various human brain disorders and neurodegenerative diseases [1]. The levels of IFNc increase in brain in different diseases, such as multiple sclerosis and encephalitis and following immune reactions. IFNc may be produced by invading immune cells or by resident microglial cells that are activated during the inflammatory process. Microglia have recently been shown to interact with neural stem cells and these cells may mediate both positive [38,39] and negative effects on NPCs (this study). NPCs in turn exert potent anti-inflammatory actions in vivo [40,41]. The final outcome of interaction between NPCs and inflammatory cells probably depends on the amount of secreted cytokines at each moment. We observed that LPS-activated microglia cells produce IFNc into the culture medium and this negatively affected the NPCs. Experiments using blocking antibodies against IFNc showed that a part of the activity in the medium is due to IFNc, although other factors cannot be excluded. In line with our data, it was recently reported that the addition of IFNc inhibited neurosphere formation in adult murine NPCs [42]. Previous studies have shown that PACAP may have potent anti-inflammatory function in regulating the production of pro-inflammatory mediators [43]. This has recently been confirmed in PACAP gene deficient mice that showed an increased expression of pro-inflammatory cytokines including IFNc [44]. We show here that PACAP directly acts on NPCs to counteract the effects of IFNc. PACAP is able to pass the blood-brain barrier [45]. PACAP may therefore be useful in treatment of brain inflammation and to enhance recruitment of endogenous NPCs after injury and conditions with high IFNc production. Due to its robust cell survival effects on NPCs PACAP may also be considered as an adjuvant treatment in different NPC transplantation studies.

Animals
Wistar rats were obtained from Harlan (Horst, The Netherlands). All experiments were approved by the local ethical committee and performed in accordance with the European Communities Council Directive (86/609/EEC).
Microglial cells were prepared from newborn rat brain essentially as described [13]. and kept in culture for up to three weeks in Dulbecco's modified Eagle's medium (DMEM)/F12 medium (Gibco) supplemented with 10% fetal bovine serum, and penicillin/streptomycin at 37uC in 5% CO 2 atmosphere. Cells were stimulated for 24 h using bacterial lipopolysaccharides (LPS, Sigma) and conditioned medium (CM) collected and the amount of IFNc analyzed using ELISA and the rat Quantikine assay (R&D Systems, Minneapolis, USA). Blocking anti-IFNc antibodies was from R&D Systems. Unstimulated CM was used as control. CM from microglial cells was added to cultures of NPCs and cell viability determined as below. p53 gene deleted mouse embryonic fibroblasts (kind gift of M Laiho) were cultured in DMEM/10% fetal bovine serum and stimulated with 100 ng/ml mouse IFNc (PeproTech).

NSC viability, cell death and proliferation assays
NSCs were cultured in 96-well cell culture dishes (70,000 cells per well; Costar 3599; Corning) in the presence of 20 ng/ml EGF and different concentrations of IFNc and PACAP. To estimate the viability of cells, we used the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Sigma) assay as described previously [11,46,47]. For TUNEL staining of DNA breaks the In Situ Cell Death Detection Kit (Roche, Basel. Switzerland) was used as described [11,48,49]. BrdU (Sigma) labeling and immunostaining using the Ki67 antibody (1:300; BD Biosciences, Franklin Lakes, NJ, USA) were performed to estimate the number of proliferating cells [11,49]. Cell cycle analyses were done using flow cytometry using a FACS calibur flow cytometer [50]. To estimate the capacity for self-renewal an equal number of NPCs (5000 cells) were incubated for 3 days in the absence and presence of 100 ng/ml IFNc and the number of neurospheres counted [11,47].
For staining of neuroepithelium, 15 mm sections from E17 old rats were cut using a Leitz (Wetzlar, Germany) microtome and placed on SuperFrost Plus glass slides [11]. Anti-IFNcR2 (diluted 1:500) or anti-nestin (1:100) antibodies were added overnight together with followed by washing with PBS. Appropriate secondary antibodies were added for 1 h and sections were washed and mounted with gel mounting medium (Gel Mount; Sigma).
Silencing RNA 100 nM siRNA construct against p53 and against p21 (Dharmacon, Lafayette, CO, USA) was tranfected using the Amaxa Nucleofector system and 5610 6 NPCs [48,49]. Equal number of control and treated cells were incubated for 48 h followed by 100 ng/ml IFNc for additional 24 h. Efficacy of downregulation was analyzed by immunblotting and the levels of p-vimentin that increases during M-phase of the cell cycle [17].

Statistics
Statistical comparisons were performed using Student's t-test when comparing two groups, or one-way ANOVA followed by a Bonferroni post hoc test when comparing three or more groups.