Interferon γ Induces Neurite Outgrowth by Up-regulation of p35 Neuron-specific Cyclin-dependent Kinase 5 Activator via Activation of ERK1/2 Pathway*

Interferon gamma (IFN-γ) is a cytokine predominantly involved in antiproliferative and antiviral responses, immune surveillance, and tumor suppression. However, it has been shown that IFN-γ is also involved in central nervous system development. Here we studied the underlying mechanism for IFN-γ-induced neuronal differentiation using the human neuroblastoma Paju cell line. Our results indicate that IFN-γ treatment led to neurite outgrowth followed by growth arrest in the G1 phase of the cell cycle. IFN-γ induced ERK1/2 phosphorylation and subsequently the transcription factor early gene response 1, which in turn up-regulated p35 expression and increased cyclin-dependent kinase 5 (Cdk5) activity. IFN-γ-induced neurite outgrowth was abolished by the treatment of MEK1/2 kinase inhibitors, such as U0126 and PD98059, which inhibit the ERK1/2 activation and subsequently prevent the up-regulation of p35 expression and Cdk5 activity. In agreement with the role of p35-Cdk5 in neuronal differentiation, small interfering RNA targeting Cdk5 abrogate the IFN-γ-induced neurite outgrowth. Together, these results demonstrate for the first time that IFN-γ-triggered neuronal differentiation mediated through the up-regulation of p35-associated Cdk5 depends on the activation of the ERK1/2 pathway. Therefore, the present study suggests that IFN-γ is not only involved in tumorigenicity but also involved in neurogenesis by regulating cell proliferation and differentiation.

Interferon gamma (IFN-␥) 1 is a cytokine with pleiotropic actions that is profoundly involved in antiproliferative and antiviral responses, immune surveillance, and tumor suppression (1). However, it has also been reported that IFN-␥ plays a role in embryonic neural differentiation and neuronal survival. For example, IFN-␥ enhances nerve growth factor (NGF)-induced neurite outgrowth (2), increases neuronal numbers in embryonic cortical and hippocampal cultures (3), and promotes cholinergic differentiation of septal nucleus and basal forebrain neurons (4). IFN-␥ also appears to affect the differentiation of neural stem cells (5). While in combination with tumor necrosis factor (TNF) superfamily death ligands, such as TNF␣, Fas ligand/CD95 ligand, and TNFrelated apoptosis-inducing ligand (TRAIL), IFN-␥ can induce apoptosis in both human brain cells and neuroblastoma cells (6 -8). However, the potentiation mechanisms of IFN-␥ on death receptor-mediated apoptosis are not fully understood, and no study has yet illustrated the molecular mechanisms of IFN-␥-induced neuronal differentiation.
To elucidate the effects of IFN-␥ on cell survival and death and to further discern the signaling pathways that are responsible for IFN-␥-induced neuronal differentiation, we conducted a series of experiments using Paju cells, a neuroblastoma cell line that was established from human neural crest-derived tumor (9). As observed in some other neuroblastomas (10,11), Paju cells also can be differentiated into neuronal phenotype by extracellular stimuli (12). Cyclin-dependent kinase 5 (Cdk5) is a small serine/threonine kinase that has been implicated in neuronal migration, differentiation, and survival during central nervous system development. Cdk5 kinase activity has been shown to be dependent upon p35, a neuron-specific Cdk5 activator (13). Expression of p35 and Cdk5 kinase activity have been shown to increase during in vitro neuronal differentiation, and mutant mice lacking either Cdk5 or p35 exhibit defective neuronal migration and cortical lamination in the developing mouse brain (14,15). Increased p35 expression and Cdk5 activity are essential for NGF-induced neurite outgrowth in rat pheochromocytoma PC12 cells and are also required for Fasmediated neuronal differentiation in vivo (16,17).
We found that treatment with IFN-␥ also resulted in neuritic outgrowth of Paju cells and generated nearly pure populations of human neuron-like cells, which provided a model to explore the molecular mechanisms of IFN-␥-induced neuronal differentiation. Treatment of Paju cells with IFN-␥ significantly increased the expression of p35 and Cdk5 activity through the activation of the extracellular-signal regulated kinase1/2 (ERK1/2) and the early gene response 1 (Egr1) pathway. In addition, co-treatment with Fas monoclonal antibody (mAb), but not TRAIL, enhanced IFN-␥-induced neuronal differentiation. In this study, we demonstrated that the signaling pathway initiated by IFN-␥ occurred through the ERK-Egr1-p35-Cdk5 pathway.
Death Receptor Analysis-Expression of Fas, DR4, and DR5 receptors was measured by flow cytometry. Briefly, 0.1 g of phycoerythrinconjugated anti-human Fas, DR4, and DR5 (mouse IgG1) or mouse IgG1 (negative control) was added to 1 ϫ 10 6 cells in 200 l of PBS containing 2% fetal bovine serum and 0.02% sodium azide. Cells were incubated for 1 h in the dark at 4°C, washed with PBS, and suspended in 500 l of PBS. Samples were analyzed using a BD Biosciences FACScan (Mountain View, CA), and the data were processed using Cell Quest software (BD Biosciences).
Cell Viability, Apoptosis, and Cell Cycle Analysis-To examine cell viability, apoptotic death, and cell cycle progression, cells (15 ϫ 10 3 / cm 2 ) were plated in 6-well plates or 100-mm dishes, and treatments were started at 24 h after plating. Cells were harvested at predetermined time points after treatments, and viable cells were determined using the trypan blue exclusion assay (18). The number of viable cells in culture was also measured with acid phosphatase assay in 96-well plates (19). Apoptotic cells were analyzed by flow cytometry for sub-G 1 content (20). Cell cycle analysis was carried out by staining the DNA with propidium iodide. Briefly, cells grown in a subconfluent culture were collected by trypsinization, washed with PBS, and fixed by incubation in 1% paraformaldehyde and then in 70% ethanol solution at 4°C. The washed pellet was stained using a propidium iodide-RNase solution (PBS containing 20 g/ml propium iodide, 200 g/ml Dnasefree RNase A, and 0.1% Triton X-100) for 30 min at 20°C in the dark. The cell cycle status was analyzed with a flow cytometer using ModFit LT software (Verity Software House Inc.).
Cell Staining and Neurite Outgrowth-Cells were fixed in 4% paraformaldehyde in PBS at room temperature for 20 min, washed three times with PBS containing 0.1% Triton X-100, and blocked for 40 min with PBS containing 1% bovine serum albumin and 0.1% Triton X-100. The cells were incubated overnight with monoclonal rabbit antihuman NF-H (1:100; 200K; Sigma), diluted in PBS containing 1% bovine serum albumin and 0.1% Triton X-100, washed three times in PBS with 0.1% Triton X-100, and incubated at room temperature for 40 min with fluorescein isothiocyanate-conjugated anti-rabbit monoclonal antibody (1:200 in PBS with 0.1% Triton X-100; Jackson Immuno-Research). Finally, the cells were washed three times with PBS and mounted with Cytoseal (Richard-Allan Scientific, Kalamazoo, MI). The cells were observed with a Leica fluorescence microscope. Paju cell differentiation was determined by scoring for neurite outgrowth. Cells possessing one or more neurites with a length Ͼ1.5-fold of the diameter of the cell body were scored as positive.
Western Blot Analysis-Cells grown in a subconfluent culture were lyzed in lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, and complete protease inhibitor mixture). After centrifugation at 18,000 ϫ g for 15 min at 4°C, supernatants were collected. Equal amounts of protein from each sample were separated by 15% SDS-PAGE gel and transferred to nitrocellulose membranes. The membranes were blocked with blocking buffer (5% nonfat dry milk in TBST (20 mM Tris-HCl, pH 7.5, 500 mM NaCl, and 0.5% Tween 20)) for 1 h at room temperature and then incubated overnight at 4°C with the appropriate primary antibody diluted in the blocking buffer. The membranes were washed in TBST and incubated for 1 h at room temperature with the following secondary antibodies diluted in TBST: anti-mouse IgG2b-HRP (1:5,000) and antimouse IgG1-HRP (1:10,000) or anti-rabbit IgG-HRP (1:5,000) according to the primary antibody used. The blots were then washed and developed by chemiluminescence.
Transfection of cDNA and siRNA-Paju cells were grown in 6-well plates at a density of 2.5 ϫ 10 5 cells/well. The cells were transiently transfected with 5 g of wild-type or dominant-negative Cdk5 cDNA using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Transfection of siRNA targeting Cdk5 (SMARTpool TM containing four pooled SMARTselected siRNA duplexes) was performed using TransMessenger transfection reagent from Qiagen according to the manufacturer's protocol. Cells were allowed to grow for 36 h following transfection and then were experimentally treated as outlined under "Results." Cdk5 Kinase Assay-Cells were lyzed in 10 mM HEPES, pH 7.5 (or 7.4), containing 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and complete protease inhibitor mixture. The lysates were centrifuged at 18,000 ϫ g for 15 min. For Cdk5 immunoprecipitation, the supernatant was incubated with glutathione-Sepharose 4B beads (Amersham Biosciences) and anti-Cdk5 (c-8) for 4 h at 4°C. The Cdk5 immunoprecipitates were placed in 20 mM MOPS (pH 7.4) with 1 mM EGTA, 1 mM dithiothreitol, 20 mM MgCl 2 , 100 M ATP, and then 10 g of histone-H1 and 5 Ci of [␥-32 P]ATP were added. After incubation for 30 min at 30°C, the reaction was terminated by spotting 30 l of the reaction mixture on P81 phosphocellulose pads (Upstate) that were washed five times with 0.75% phosphoric acid followed by rinsing with acetone. Radioactivity was measured with a liquid scintillation counter. In some experiments, in vitro Cdk5 kinase activity was also analyzed by Western blotting.

Neuroblastoma Paju Cells Are Resistant to Death Receptormediated
Apoptosis-A recent study demonstrated that IFN-␥ sensitizes some neuroblastoma cells to death receptor-mediated apoptosis (7). To investigate whether IFN-␥ enhances apoptosis induced by Fas and TRAIL, we first analyzed the expression of death receptors in Paju cells. Flow cytometry analysis showed that Paju cells expressed cell surface Fas and DR5 but failed to reveal DR4 expression (Fig. 1a). Exposure of the cells to IFN-␥ for 24 h significantly increased Fas expression, whereas no increase in DR4 and DR5 expression was evident (Fig. 1a).
Next, we examined whether Fas and TRAIL can induce apoptosis in Paju cells. As we previously reported (18,21), Jurkat T leukemia cells are susceptible to apoptosis induced by agonistic Fas mAb and TRAIL (Fig. 1b). However, parallel cultures of Paju cells did not show any evidence of apoptosis following treatment with Fas mAb or TRAIL (Fig. 1b). In general, the lack or absence of caspase-8 expression is often correlated with the resistance of neuroblastoma cells to death receptor-mediated apoptosis. It has been proposed, however, that IFN-␥ sensitizes cells to undergo apoptosis by increasing the endogenous expression of caspase-8. In light of this, we decided to examine whether IFN-␥ can sensitize Paju cells to undergo apoptosis. Western blot analysis showed that lack of caspase-8 expression in Paju cells and exposure of the cells to IFN-␥ resulted in an increase of caspase-8 at 24 h; this expression level was sustained for at least 72 h after the treatment (Fig.  1c). However, no cleavage forms of caspase-8 were detected. The expression of downstream products of caspase-8, caspase-3, and DNA fragmentation factor 45 was not changed following the treatment with IFN-␥. Additionally, no activated forms of the caspase-3 and DNA fragmentation factor 45 were detected following the treatment.
To further investigate whether IFN-␥ sensitizes Paju cells to death receptor-induced apoptosis, cell death was measured after treatment with IFN-␥ in combination with either Fas mAb or TRAIL. Cells were pretreated with IFN-␥ for 24 h before the application of Fas mAb or TRAIL, and cell viability was assessed by the acid phosphatase assay 24 h later. Fas mAb or TRAIL either alone or in combination with IFN-␥ failed to induce apoptosis in Paju cells (Fig. 1d). The activation of caspase-8 and casapse-3 was also analyzed by Western blotting. Treatment with IFN-␥ alone or in combination with these ligands increased caspase-8 expression (Fig. 1d). However, these treatments failed to show cleavage forms of caspase-8 and caspase-3. These results indicate that IFN-␥ is capable of restoring caspase-8 expression but not sufficient enough to induce apoptosis, implying that a number of factors may be involved in the resistance of neuroblastoma cells to apoptosis.
IFN-␥ Induces Neuronal Differentiation in Paju Cells-Surprisingly, treatment with IFN-␥ induced neuronal differentiation in Paju cells as characterized by neurite outgrowth. Following treatment with IFN-␥, neurite outgrowths began to appear on day 1 (Fig. 2a). The increases in length and number of the neurites were rapid. On day 5 after IFN-␥ treatment, extensive networks were formed by the neuritic processes and the cells had differentiated into nearly pure populations of human neuron-like cells. Neurofilament staining confirmed that these neuron-like cells expressed the marker typical of neurons (Fig. 2b). In contrast to IFN-␥, neither Fas mAb nor TRAIL treatment enhanced the neurite outgrowth in Paju cells.
To investigate whether IFN-␥ treatment causes the transition from proliferation to neuronal differentiation, Paju cells were exposed to IFN-␥ for different periods of time, and cell proliferation was examined by analyzing changes in cell numbers. IFN-␥ treatment significantly inhibited cell proliferation, and this inhibition was seen on day 1 (Fig. 2c). Treatment with Fas mAb also inhibited cell proliferation, but its inhibitory action was weaker than IFN-␥. Further, Fas mAb-mediated growth inhibition began 2 days after treatment. In contrast, treatment with TRAIL had no inhibitory effects on cell proliferation. The inhibitory actions of IFN-␥ and Fas mAb on cell proliferation were also confirmed with cell viability assay (Fig. 2d).
To confirm IFN-␥ mediated transition of neuronal differentiation from proliferation, the effect of IFN-␥ on cell cycle progression of Paju cells was examined. Results from flow cytometry with propidium iodide staining showed that cell cycle arrest occurred at G 1 stage (Fig. 2e). Hence, these results indicate that IFN-␥-induced neuronal differentiation is initiated in the G 1 phase of the cell cycle, and then differentiation proceeds as shown by neurite outgrowth (Fig. 2, a and b).
IFN-␥-induced Neuronal Differentiation Requires p35-associated Cdk5 Activity-p35/Cdk5 has been shown to be essential for NGF-induced neurite outgrowth in PC12 cells and FasLinduced neuronal differentiation in animals (16,17). We therefore investigated whether Cdk5-p35 plays a role in IFN-␥induced neuronal differentiation of Paju cells. We first Flow cytometry analysis showed cell surface expression of Fas is up-regulated by treatment with IFN-␥ (100 ng/ml for 16 h), but not DR4 and DR5. Isotypematched control IgG mAb was used as a negative control (N.C.). b, sensitivity of Paju cells to apoptosis. Cells were treated with Fas mAb or TRAIL for 16 h in a dose-dependent manner. Apoptosis was evaluated by flow cytometry for sub-G 1 content (mean Ϯ S.D.; n ϭ 3). Jurkat T cells were used as a positive control. c, Western blot analyses for caspase-8, caspase-3, and DNA fragmentation factor 45 after IFN-␥ treatment for the time indicated. d, effect of IFN-␥ alone or in combination with Fas mAb or TRAIL on apoptosis. Paju cells were pretreated with IFN-␥ for 24 h to restore caspase-8 expression, and then Fas mAb (3 g/ml) or TRAIL (100 ng/ml) was added. The cells were collected at 6 h after the addition of Fas mAb or TRAIL. These treatments failed to show cleaved caspases in Paju cells, whereas cleaved forms of caspase-8 (p43/41, p18) and caspase-3 (p24, p20, p17) were found in Jurkat T cells. Treatment with IFN-␥ alone or in combination with Fas mAb or TRAIL (for 24 h) failed to trigger apoptosis in Paju cells. examined the expression of p35 and Cdk5 in Paju cells. Following treatment with IFN-␥, the level of p35 increased at 1 h and peaked at 5 h of IFN-␥ treatment (Fig. 3a). The level of p35 then gradually decreased and returned to basal level on day 1 after IFN-␥ treatment. However, the level of Cdk5 did not change significantly following the IFN-␥ treatment. Subsequently, we examined the changes of Cdk5 activity following IFN-␥ treatment. Cdk5-p35 complex was immunoprecipitated with anti-Cdk5 antibody, and the activity of Cdk5 was assessed using histone H1 as a substrate (Fig. 3a). The activity of Cdk5 increased markedly after IFN-␥ treatment, and this change paralleled the changes of p35. This observation is in agreement with previous findings that Cdk5 activity is absolutely dependent on its activator, p35 (15,16,22).
We then studied whether Cdk5 is involved in cell differentiation. Two series of experiments were carried out. First, we examined whether the transfection of dominant-negative Cdk5 gene (dn-Cdk5) into Paju cells blocks IFN-␥-induced neurite outgrowth. Treatment of the Paju cells with IFN-␥ at 24 h after Cdk5 gene transfection resulted in an increase in neurite outgrowth, whereas IFN-␥ failed to induce neurite outgrowth in dn-Cdk5-transfected cells (Fig. 3, b and c). In the next set of experiments, we studied whether inhibition of Cdk5 expression can block cell differentiation. Cdk5 expression was suppressed using siRNA duplex targeting Cdk5 (siRNA-Cdk5). Transient transfection of siRNA-Cdk5 significantly reduced Cdk5 protein expression and inhibited its kinase activity (Fig. 3d). Similarly, IFN-␥-induced neurite outgrowth was also inhibited when Cdk5 expression was inhibited using siRNA-Cdk5 (Fig. 3e). In addition, we examined whether treatment of IFN-␥ in combination with either Fas mAb or TRAIL can enhance cell differentiation. Results from this study showed that co-treatment with Fas mAb, but not TRAIL, enhanced IFN-␥-induced cell differentiation as evidenced by the occurrence of neurite outgrowth (Fig. 3f). Western blotting analysis also showed p35 expression was higher following the combined treatment (IFN-␥ and Fas mAb) than IFN-␥ alone (data not shown). Treatment with Fas mAb or TRAIL alone did not cause an increase in p35 expression; these treatments also did not enhance neurite outgrowth. Therefore, we conclude that increased p35-Cdk5 kinase activity is required for IFN-␥-induced neuronal differentiation in Paju cells.

ERK1/2-Egr1 Signaling Pathway Is Involved in IFN-␥-induced
Neuronal Differentiation-In a previous study, sustained activation of ERK1/2 was shown to be necessary and sufficient for NGF-induced neurite outgrowth; up-regulation of p35 by sustained ERK1/2 activation is an essential component of this pathway (16). Furthermore, it has also been shown that ERK1/2-induced up-regulation of p35 is mediated by transcription factor Egr1. To identify whether ERK1/2 activation also mediates IFN-␥-induced cell differentiation in Paju cells, we examined whether IFN-␥ treatment in the present system resulted in the activation of ERK1/2 and Egr1 pathway. ERK1/2 kinases are activated through dual phosphorylation by MAP kinase/ERK kinase (MEK) (23). Western blot analyses revealed that following IFN-␥ treatment, robust phosphorylation of ERK1/2 p44/42 proteins starting at 15 min and peaking at 1 h was evident in Paju cells (Fig. 4a). The phosphorylation of ERK1/2 then appeared to decrease at 5 h, although it was still higher than the control samples. Egr1 could not be initially detected in the control samples. However, a significant increase was observed 1 h after IFN-␥ treatment; a lower level of Egr1 was also seen at 3 and 5 h after treatment (Fig. 4a).
We then studied whether ERK1/2 activation mediated IFN-␥-induced neuronal differentiation in Paju cells using several specific inhibitors: PD98059 (30 M), a specific inhibitor of MAP-kinase kinase 1 (MEK1); U0126 (10 M), a specific inhibitor of MEK1 and MEK2; SB203580 (10 M), a specific inhibitor of p38 mitogen-activated protein kinase. The Paju cells were treated with IFN-␥ for 5 h in the presence or absence of each inhibitor, and then the cells were collected for Western blot analysis (Fig. 4b). Treatment with U0126 or PD98059 completely suppressed ERK1/2 phosphorylation and p35 expression (Fig. 4b). Alternatively, SB203580 treatment did not have a significant effect on the suppression of ERK1/2 phosphorylation or p35 expression (Fig. 4b). The decrease in the p35 level upon inhibition of ERK1/2 activity correlates with the changes in Cdk5 activity in Paju cells (Fig. 4c). Moreover, treatment with ERK1/2 inhibitors significantly blocked the neurite outgrowth in Paju cells (Fig. 4d). We also examined Cdk5 downstream Akt pathway (24). Akt activation occurs through PI3K phosphorylation (25). In Paju cells, Akt activation was constantly high either with or without IFN-␥ treatment as seen in Akt phosphorylation (Fig. 4a). Treatment with LY294002, a specific PI3K inhibitor, revealed cytotoxicity in a dose-dependent manner, and pretreatment with a nontoxic concentration of 10 g/ml LY294002 failed to prevent IFN-␥-induced neurite outgrowth and p35 up-regulation (Fig. 4b). Taken together, these results suggest that the ERK1/2-Egr1-p35-Cdk5 signaling pathway is required for neuronal differentiation induced by IFN-␥ in Paju cells. DISCUSSION Cytokines including TNF␣ superfamily death ligands (TNF␣, FasL, and TRAIL) and IFN-␥ have been shown to be involved in apoptotic death of the brain cells during traumatic and ischemic central nervous system injury (26,27). Alternatively, it has been shown that these cytokines may also play a significant role in neurogenesis (5,28,29). Fas and TRAIL have an ability to induce apoptosis; however, these cytokines also promote cell proliferation and differentiation (30,31), which implies that the role of Fas and TRAIL in the central nervous system is the subject of controversy. Here we have demonstrated the effects of these cytokines on apoptosis in Paju cells. Treatment of Paju cells with Fas mAb or TRAIL failed to induce apoptosis. As shown here as well as in previous studies (32,33), the inability of Fas and TRAIL to induce apoptosis may be attributed to the inefficient expression of caspase-8 in Western blot analysis shows that treatment with IFN-␥ significantly increased p-ERK and Egr1, indicating activation of the ERK-Egr1 pathway. Treatment with IFN-␥ also caused p35 up-regulation and the subsequent onset of Cdk5 activity. Akt was spontaneously expressed in Paju cells, and a constant level of p-Akt was also observed either with or without IFN-␥ treatment. b, ERK inhibition down-regulated p35 expression. Cells were pretreated with U0126, PD98059, SB203580, or LY294002 for 16 h before application of IFN-␥ and collected for Western blot analysis at 5 h after application of IFN-␥. Treatment with U0126 or PD98059 attenuated IFN-␥-induced increase of p-ERK and also inhibited p35 expression. c, changes in in vitro Cdk5 kinase activity. Pretreatment with U0126 and PD98059, but not SB203580 and LY294002, inhibited IFN-␥-mediated activation of Cdk5 (mean Ϯ S.D.; n ϭ 3). d, effect of ERK inhibitors, U0126 and PD98059, on IFN-␥-induced neuronal differentiation. Paju cells were pretreated either with U0126 or PD98059 before application of IFN-␥. Morphological changes of Paju cells were analyzed under light microscopy at 48 h after the treatments. Treatment with these ERK inhibitors prevented IFN-␥-induced neurite outgrowth.
Paju cells (Fig. 1). IFN-␥ has been shown to up-regulate death receptor expression and increase caspase-8 expression via the Stat1 pathway, thus increasing sensitivity to death receptormediated apoptosis (7). Although IFN-␥ increased Fas receptor (but not DR4 and DR5) expression and restored caspase-8 expression in Paju cells (Fig. 1, a and d), the resistance of Paju cells to apoptosis was not reversed by co-treatment of IFN-␥ in combination with either Fas mAb or TRAIL (Fig. 1, e and f). Thus, IFN-␥-modulated apoptosis seems to be cell type-specific in the nervous system, and additional mechanisms might be involved in the resistance of Paju cells to apoptotic death.
In search of the resistant mechanism of Paju cells, we unexpectedly found that treatment with IFN-␥ induced the cells to differentiate into a neuronal phenotype. It has been shown that the differentiating cells are much more resistant to apoptosis than the undifferentiated parent cells (34,35), perhaps because of the activation of antiapoptotic systems including PI3K/Akt, ERK, and Bcl-2 family molecules. Similarly, Paju cells showed constant activation of PI3K/Akt during IFN-␥ treatment, and treatment with the PI3K inhibitor LY294002 caused cytotoxicity in a dose-dependent manner (Fig. 4a). Therefore, Paju cells were insensitive to apoptosis, which may be the result of constitutive activation of PI3K/Akt pathway.
Next we examined IFN-␥-mediated cell cycle changes because cell cycle arrest is often required for neuronal differentiation, which is regulated by Cdk inhibitors such as p15, p16, p21, and p27 that inhibit phosphorylation of the retinoblastoma protein (36). IFN-␥ treatment led to growth arrest in the G 1 phase of the cell cycle, accompanied by a decrease in S-phase (Fig. 2). This proliferation inhibition may be correlated with occurrence of the neurite outgrowth (Fig. 2). These results are in agreement with previous observations that in many cell types the decision of cells to differentiate is often made in the G 1 phase of the cell cycle and that the transition from proliferation to differentiation is an inversely correlated process at cell cycle exit (37).
We subsequently elucidated the signal pathway responsible for the IFN-␥-induced neuronal differentiation in Paju cells. Results from the present study indicate that treatment with IFN-␥ up-regulated p35 expression, the neuron-specific activator for Cdk5, and thus activated Cdk5. p35-associated Cdk5 activation is required and sufficient for the IFN-␥-induced neuronal differentiation in Paju cells. As shown in Fig. 3, b-e, inhibition of Cdk5 activity by either transfection of siRNA-Cdk5 or dominant-negative Cdk5-expressing vector inhibited IFN-␥-induced neurite outgrowth. We further studied whether ERK-Egr1 pathway is necessary for IFN-␥-induced p35/Cdk5 activation. Results from Fig. 4 demonstrated that treatment of U0126 or PD98059 not only reduced p35 expression and inhibited Cdk5 activity but also blocked IFN-␥-induced neurite outgrowth. Additionally, ERK1/2 activation resulted in the induction of Egr1, a member of a zinc finger transcription factor family that binds to the promoter region of p35 and induces p35 expression (16).
Collectively, the present results thus suggest that IFN-␥mediated p35-Cdk5 activity is implicated in neurogenesis but independent from the pro-apoptotic role of IFN-␥. In previous studies, the activation of ERK and up-regulation of p35 have been observed in NGF-induced neuronal differentiation in PC12 cells and also in Fas-induced neuronal differentiation in vivo (16,17,38). These cytokines are also up-regulated during neuronal development and neurological diseases. Therefore, these observations, together with the present results, support the notion that IFN-␥-induced ERK1/2-dependent p35-Cdk5 activity may be implicated in neural regeneration in the nervous system.