Inhibition of Src-like Kinases Reveals Akt-dependent and -independent Pathways in Insulin-like Growth Factor I-mediated Oligodendrocyte Progenitor Survival*

Previous work has shown that insulin-like growth factor I (IGF-I)¹ is important for the survival of oligodendrocyte progenitors. Early in vitro experiments demonstrated that IGF-I is required for proliferation and survival of oligodendrocyte progenitors (1, 2) as evidenced by increased numbers of oligodendrocyte progenitors in the cerebral cortex and corpus callosum of transgenic mice overexpressing IGF-I (3, 4). IGF-I has also been found to protect white matter in fetal sheep from ischemic injury by decreasing apoptosis and by promoting regeneration of oligodendrocytes (5, 6). Similarly, IGF-I reduces lesion severity and promotes myelin regeneration in experimental autoimmune encephalomyelitis, an animal model of multiple sclerosis (7). In contrast, there is a significant reduction in central nervous system myelination as well as in the number of oligodendrocyte progenitors in the anterior commissure and corpus callosum of IGF-I-null mutant mice (8). However, despite the extensive literature on IGF-I-stimulated oligodendrocyte progenitor survival, the signaling pathways that mediate this effect remain to be fully elucidated.

IGF-I interacts with the IGF-I receptor (IGF-IR) to initiate downstream responses, including proliferation and differentiation. IGF-IR is a heterotetramer with intrinsic tyrosine kinase activity that phosphorylates insulin receptor substrate-1 and -2. Together with insulin receptor substrate-1/2, IGF-IR activates two main downstream signaling pathways, the phosphatidylinositol 3-kinase (PI3K) and the Ras/Raf/MEK/ERK cascades (9). Several studies have shown that PI3K is necessary for IGF-I-mediated neural (10–13) and Schwann (14–16) cell survival. IGF-I is able to prevent glutamate-induced apoptosis in a PI3K-dependent manner in immature oligodendrocytes (17, 18), confirming the established role of PI3K in survival of oligodendrocyte progenitors (19, 20). Activated PI3K increases phosphatidylinositol 3,4-diphosphate and phosphatidylinositol 3,4,5-trisphosphate in the cytoplasmic membrane. One of the crucial downstream targets of PI3K is the serine/threonine kinase Akt, which is recruited to the membrane by direct binding of its pleckstrin homology domain to the PI3K-produced phosphatidylinositol 3,4,5-trisphosphate. Upstream kinases such as 3-phosphoinositide-dependent kinase (PDK)-1 and PDK2 activate Akt by phosphorylation of Thr³⁰⁸ (21) in its activation loop and Ser⁴⁷³ in the C-terminal regulatory domain (22), respectively. In addition, Src family tyrosine kinases can directly regulate the activity of Akt by phosphorylating Tyr³¹⁵ and Tyr³²⁶ in its activation loop (23, 24). Activated Akt can phosphorylate a number of proteins, including glycogen synthase kinase-3β (GSK3β), 6-phosphofructo-2-kinase, the apoptosis-inducing protein Bad, and the forkhead-related transcription factor FKHR-L1 to regulate glycogen synthesis, glycolysis, and cell survival (25).

The MEK/ERK pathway activated by IGF-I was shown to promote survival in renal epithelial cells following oxidative injury (26). Activated ERK phosphorylates ribosomal S6 kinase (RSK), which can, in turn, phosphorylate and inactivate Bad, an effector of apoptosis. RSK also phosphorylates and activates the cAMP response element-binding protein, a transcription factor that mediates cell survival (27–29). Both the MEK/ERK and PI3K pathways are required in IGF-I-mediated monocytederived dendritic cell survival (30), and other studies (63–65, 67) have elucidated even more complex mechanisms in different cell types. We therefore endeavored to determine whether 1) the PI3K pathway is the only mechanism involved in IGF-I-mediated survival of oligodendrocyte progenitors and 2) whether Akt is the downstream effector of PI3K that mediates this effect.

Our results show that IGF-I-promoted oligodendrocyte progenitor survival upon growth factor deprivation is independent of the MEK/ERK pathway, but does require activation of PI3K, as demonstrated using selective inhibitors of MEK (PD 98059) and PI3K (LY 294002 and wortmannin). The specific Src-like tyrosine kinase inhibitor PP2 abolished Akt activation by IGF-I, but had no significant effects on the blockade of caspase-3 activation or progenitor survival. Furthermore, use of both a selective inhibitor and dominant-negative forms of Akt showed that this enzyme is not the only critical component mediating IGF-I-promoted survival in oligodendrocyte progenitors. An unidentified effector of PI3K, in addition to Akt, thus appears to be required to confer full protection of oligodendrocyte progenitors by IGF-I.

EXPERIMENTAL PROCEDURES

Materials—Dulbecco's modified Eagle's medium (DMEM), Ham's F-12 medium, phosphate-buffered saline, Hanks' balanced salt solution, 7.5% bovine serum albumin (fraction V), fetal calf serum, penicillin, and streptomycin were purchased from Invitrogen. Other reagents and supplies were purchased from the following suppliers: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and 2-bromodeoxyuridine from ICN Canada Ltd. (Montreal, Quebec); Immobilon-P membranes from Millipore (Mississauga, Ontario); an ECL Western blot detection kit from PerkinElmer Life Sciences; human recombinant platelet-derived growth factor AA, IGF-I, and basic fibroblast growth factor from PeproTech Inc. (Rocky Hill, NJ); a TUNEL kit from Roche Diagnostics; a protein assay kit from Bio-Rad; Triton X-100, poly-d-lysine, poly-l-ornithine, human transferrin, insulin, and polyclonal rabbit anti-actin antibody from Sigma; polyclonal rabbit anti-active MAPK (mitogen-activated protein kinase; ERK1/2), polyclonal rabbit phospho-Akt (Ser⁴⁷³/Thr³⁰⁸), anti-Akt, anti-phospho-GSK3β (Ser⁹), and anti-cleaved caspase-3 antibodies from New England Biolabs Inc.; anti-Fyn, anti-Lyn, and anti-procaspase-3 antibodies from Santa Cruz Biotechnology Inc. (Santa Cruz, CA); and anti-Src antibody from Oncogene (Cambridge, MA). Secondary antibodies used for immunostaining or immunoblotting were from Southern Biotechnology Associates (Birmingham, AL), Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA), and Cedarlane (Hornby, Ontario). LY 294002, wortmannin, PD 98059, PP2 (4-amino-5-(4-chlorophenyl)-7-(t-butyl) pyrazolo[3,4-d]pyrimidine), and 1L-6-hydroxymethyl-chiro-inositol 2-(R)-2-O-methyl-3-O-octadecylcarbonate were obtained from Calbiochem. All other reagents were obtained from VWR International Ltd. (Mont-Royal, Quebec) or Fisher (Montreal, Quebec).

Primary Cultures—Primary cultures of oligodendrocyte progenitors were prepared from brains of newborn Sprague-Dawley rats as described (31). The meninges and blood vessels were removed from the cerebral hemispheres in Ham's F-12 medium. Briefly, the tissues were gently forced through a 230-μm nylon mesh. The dissociated cells were gravity-filtered through a 100-μm nylon mesh. This second filtrate was centrifuged for 7 min at 1000 rpm, and the pellet was resuspended in DMEM supplemented with 12.5% fetal calf serum, 50 units/ml penicillin, and 50 μg/ml streptomycin. Cells were plated on poly-l-ornithine-precoated 80-cm² flasks and incubated at 37 °C with 5% CO₂ in air. The culture medium was changed after 3 days and every 2 days thereafter. The initial mixed glial cultures (grown for 9–11 days) were placed on a rotary shaker at 225 rpm for 3 h at 37 °C to remove loosely attached macrophages. Oligodendrocyte progenitors were detached following shaking for 18 h at 260 rpm. The cells were filtered through a 30-μm nylon mesh and plated on bacterial grade Petri dishes for 3 h. Under these conditions, astrocytes and microglia attached to the plastic surface, and oligodendrocyte progenitors remained in suspension. The final cell suspension was plated on multiwell dishes precoated with poly-d-lysine at a density of ∼15 × 10³/cm². Cultures were maintained in serum-free medium (SFM) containing 2.5 ng/ml platelet-derived growth factor AA and 2.5 ng/ml basic fibroblast growth factor to stimulate proliferation, and the medium was changed every 2 days. Ninety-five percent of the cells reacted positively to monoclonal antibody A2B5, a marker for oligodendrocyte progenitors, and <5% were galactocerebro-side-positive oligodendrocytes, glial fibrillary acidic protein-positive astrocytes, or C3-positive microglia.

All experiments were conducted in DMEM alone or in SFM in the absence or presence of the indicated pharmacological agents. SFM consisted of a 1:1 DMEM/Ham's F-12 mixture, 10 mm HEPES, 0.1% bovine serum albumin, 25 μg/ml human transferrin, 30 nm triiodothyronine, 20 nm hydrocortisone, 20 nm progesterone, 10 nm biotin, 5 μg/ml insulin, 16 μg/ml putrescine, 30 nm selenium, 50 units/ml penicillin, and 50 μg/ml streptomycin.

Western Blot Analysis—Cells grown in 6-well culture plates were harvested, after treatment, in 50 μl of ice-cold lysis buffer containing 20 mm Tris-HCl (pH 8), 1% Nonidet P-40, 10% glycerol, 137 mm NaCl, 1 mm phenylmethylsulfonyl fluoride, 1 mm aprotinin, 0.1 mm sodium vanadate, and 20 mm NaF. The protein content of cell lysates was determined with the Bio-Rad protein assay kit, and the samples were adjusted with loading buffer containing 2% SDS, 5% glycerol, 5% β-mercaptoethanol, and 0.01% bromphenol blue and boiled for 5 min. Aliquots containing 25 μg of protein were resolved by SDS-PAGE and transferred to Immobilon-P membranes as described previously (32). The membranes were blocked and probed with an appropriate primary antibody. Bands were visualized with horseradish peroxidase-conjugated secondary antibody used in conjunction with an ECL Western blot detection kit. Blots were scanned and quantified using M4 software. To normalize for sample loading and protein transfer, the membranes were stripped and reprobed with an antibody for β-actin, total Akt, or ERK2 as indicated.

MTT Assay of Cell Viability—Oligodendrocyte progenitors were treated with IGF-I in the presence or absence of PD 98059, LY 294002, wortmannin, or PP2 in DMEM for 18 h. Because wortmannin is unstable in aqueous solutions (33), it was added every 6 h prior to harvest. Mitochondrial dehydrogenase activity assayed by cleavage of MTT was used to determine cell viability as described previously (34). The reaction detects only living cells and is based on cleavage of the tetrazolium ring by active mitochondria, producing a visible dark blue formazan product. Oligodendrocyte cultures were incubated with 125 μg/ml MTT at 37 °C for 3 h. The medium was then aspirated, and the precipitated formazan crystals were solubilized in an acid/isopropyl alcohol mixture. Samples were read on a micro-enzyme-linked immunosorbent assay spectrophotometer at 600 nm. Absolute MTT values were normalized by scaling to the mean of progenitor culture grown in DMEM alone (defined as 100%).

Visualization of Apoptotic Nuclei (TUNEL Labeling)—Oligodendrocyte progenitors (growing in 24-well tissue culture plates with poly-d-lysine-coated glass coverslips) were transferred to DMEM with IGF-I in the absence or presence of an inhibitor of PI3K, Src-like kinase, or MEK1 for 18 h at 37 °C. For immunocytochemical detection of apoptotic cells, cultures were washed with phosphate-buffered saline and then fixed with 4% paraformaldehyde for 20 min at room temperature. Fragmented DNA (high molecular mass or internucleosomal) was detected by incorporating fluorescein-12-dUTP at the 3′-OH ends using the TUNEL assay as described (35). Horseradish peroxidase-conjugated anti-fluorescein Fab fragments detected incorporated fluorescein. A 3,3′-diaminobenzidine substrate kit (Vector Laboratories, Burlingame, CA) was used to detect peroxidase activity. Stained cells were visualized by light microscopy.

Transfection of Oligodendrocyte Progenitors with Akt cDNA Constructs—Plasmids CMV6, CMV6-Akt-HA (wild-type Akt), and CMV6-K179M-Akt-HA (kinase-dead Akt (kdAkt)) were purified with the QIAGEN plasmid midi-kit (36). Oligodendrocyte progenitors were transfected with the wild-type Akt and kdAkt plasmids using the Effectene transfection reagent kit (QIAGEN Inc.) following the manufacturer's instructions. Dominant-negative Akt (dnAkt; Akt-T308A/ S473A) with an HA tag at the N terminus (37) and green fluorescent protein (38) in adenoviral vectors were used at a multiplicity of infection of 10. The infection efficiency was 75–95% as determined by counting green fluorescent protein-positive cells or immunofluorescence with anti-HA antibody. The progenitors were infected with an adenoviral vector containing dnAkt or green fluorescent protein 48 h before IGF-I treatment.

Small Interfering RNAs (siRNAs)—Complementary p59^fyn double-strand siRNAs labeled with Alexa Fluor 488 were synthesized by QIAGEN Inc. The siRNA sequences for fyn-1 and fyn-2 covered nucleotides 482–502 and 1307–1327, respectively (NCBI accession number U35365). The siRNAs were transfected into oligodendrocyte progenitors using RNAiFect™ reagent (QIAGEN Inc.) according to the manufacturer's instructions. Knockdown of fyn was assessed 48 h post-transfection by Western blotting with anti-Fyn and anti-Src antibodies or by fluorescence microscopy of individual Alexa 488-labeled siRNA-positive cells, immunostained with anti-Fyn antibody, followed by Texas Red-conjugated goat anti-rabbit secondary antibody. On the basis of this screen, the concentration of siRNA that suppressed fyn expression most efficiently (70%) was selected for further experiments. A nonspecific siRNA labeled with Alexa 488 (AAT TCT CCG AAC GTG TCA CGT; QIAGEN Inc.) was used as a negative control. Cell nuclei were stained for 15 min with 500 ng/ml 4,6-diamidino-2-phenylindole dihydrochloride to determine whether the various treatments caused DNA condensation or fragmentation, characteristic of apoptosis.

Immunoprecipitation—Oligodendrocyte progenitors were pretreated with PP2 or wortmannin 30 min prior to IGF-I for the indicated times. Cells were harvested in ice-cold radioimmune precipitation assay buffer (50 mm Tris (pH 7.4), 150 mm NaCl, and 1% Nonidet P-40) and centrifuged at 1000 × g for 30 min at 4 °C to remove nuclei and insoluble remnants. Immunoprecipitations were carried out in the same buffer using 0.2 mg of protein, 2 μg of antibodies, and 20 μl of protein A/G Plus-agarose at 4 °C for 18 h. The immunocomplexes were washed three times with radioimmune precipitation assay buffer and two times with phosphate-buffered saline and then finally resuspended in 1× SDS sample buffer (62.5 mm Tris-HCl (pH 6.8), 2% SDS, 10% glycerol, and 0.1% bromphenol blue, with 50 mm dithiothreitol added freshly). Western blotting was performed as describe above with anti-phosphotyrosine, anti-Fyn, anti-Src, and anti-Lyn antibodies.

Data Analysis—Unless indicated otherwise, results are presented as the mean ± S.E. of at least three independent experiments performed in duplicate or triplicate. Statistical significance was determined by one-way or two-way analysis of variance, followed by the Tukey test; p values <0.05 were considered significant.

RESULTS

IGF-I Promotes Survival of Oligodendrocyte Progenitors— Oligodendrocyte progenitors were deprived of growth factors to induce apoptosis. Cells were maintained in DMEM with or without IGF-I for 24 h or in chemically defined SFM, which contained 5 μg/ml insulin. The protective effect of IGF-I on progenitors was determined using the MTT assay. In cells grown in DMEM alone, the viability was only 54% of that of the cells grown in SFM, whereas 100 ng/ml IGF-I increased MTT values to the same levels as those found in SFM (Table I). The survival effect of IGF-I was concentration-dependent, with maximal protection occurring at 50 ng/ml IGF-I (Fig. 1). TUNEL staining was subsequently used to assess the number of apoptotic cells. Twenty percent of the progenitors were TUNEL-positive when grown in DMEM alone. The addition of IGF-I at 100 ng/ml caused a significant reduction in the frequency of apoptosis, reducing the number of TUNEL-positive cells to 6.8%, which is similar to the value obtained for cells grown in SFM (Table I).

IGF-I Activates the PI3K/Akt Pathway—Most of the biological effects mediated by IGF-I involve activation of two downstream signaling pathways: the Ras/Raf/MEK/ERK and PI3K/Akt cascades (9). To investigate whether these pathways are involved in IGF-I signaling in oligodendrocyte progenitors, we assessed the activation levels of ERK1/2 and Akt by Western blotting with phospho-specific antibodies. Akt was phosphorylated at both Ser⁴⁷³ and Thr³⁰⁸ (data not shown) when cultures were treated with 100 ng/ml IGF-I. This activation was maintained for 48 h (Fig. 2). In addition, GSK3β, a downstream target of Akt, was phosphorylated at Ser⁹ after IGF-I treatment in a pattern that paralleled Akt activation. In contrast to the effect of IGF-I on Akt, ERK1/2 was transiently phosphorylated at 5 min and decreased to control levels or below at 1 and 4 h, respectively (Fig. 2 and Supplemental Fig. S1).

IGF-I Inhibits Activation of Caspase-3 Induced by Growth Factor Withdrawal—Growth factor deprivation triggers the activation of caspase-3, a potent effector of apoptosis in oligodendrocyte progenitors (39). Caspase-3 is activated by proteolytic cleavage of the 32-kDa procaspase-3 into active 17- and 12-kDa fragments. Progenitors were incubated in DMEM with or without 100 ng/ml IGF-I for 1–48 h, and the activation of caspase-3 was determined by Western blotting with an antibody directed against the 17-kDa fragment. In DMEM alone, caspase-3 activation was time-dependent, with a significant increase detected at 4 h and maximal activation at 48 h. Treatment with IGF-I was able to suppress caspase-3 activation for as long as 48 h (Fig. 2). These results indicate that growth factor withdrawal-induced caspase-3 activation is involved in progenitor apoptosis and that this activation is prevented by IGF-I.

PI3K Inhibitors Reverse the IGF-I-mediated Rescue of Oligodendrocyte Progenitors from Cell Death, in Contrast to MEK and Src Inhibitors—Cells were pretreated with selected kinase inhibitors for 30 min prior to addition of 100 ng/ml IGF-I, and MTT reduction was assayed 18 h later. At the highest concentrations, these drugs can reduce cell survival on their own. The MTT results are therefore expressed as a percentage of the survival effect calculated from the following formula: (T_IGF-I+I – T_I)/(T_IGF-I – T_control). T_IGF-I, T_IGF-I+I, T_I, and T_control represent the MTT reading for IGF-I alone, IGF-I plus inhibitor, inhibitor alone, and control cells in DMEM alone, respectively.

To assess the involvement of PI3K in IGF-I-mediated cell survival, progenitors were treated with the inhibitors LY 294002 and wortmannin. LY 294002, a selective PI3K inhibitor that acts on the ATP-binding site of the enzyme, caused a concentration-dependent decrease in IGF-I-mediated cell survival (Fig. 3). A significant decrease in MTT reduction was obtained with 10 μm LY 294002, whereas 50 μm LY 294002 almost completely abrogated the protective effect of IGF-I (90%). In addition, 30 μm LY 294002 partially reversed the decrease in the number of TUNEL-positive cells produced by IGF-I (Fig. 4C and Table II). Wortmannin, a selective and irreversible inhibitor of PI3K, decreased IGF-I-mediated survival in a concentration-dependent manner, with a maximal decrease in MTT values (70%) occurring at 1 μm (Fig. 3). In addition, 0.5 μm wortmannin fully reversed the protective effect of IGF-I on TUNEL-positive cells elicited by growth factor withdrawal (Table II). These results demonstrate that PI3K plays an important role in IGF-I-mediated survival of oligodendrocyte progenitors.

To assess whether ERK1/2 plays a role in IGF-I-mediated oligodendrocyte progenitor survival, cells were treated with PD 98059, a selective inhibitor of MEK1, the immediate upstream effector of ERK1/2. PD 98059 at 10 μm (a concentration that completely blocks ERK1/2 activation) had no significant effect on IGF-I-stimulated cell survival as assessed by the MTT or TUNEL assay (Figs. 3 and 4D and Table II). Higher concentrations of PD 98059 (25 and 50 μm) exhibited a tendency to slightly decrease MTT values (Fig. 3). These results suggest that the MEK/ERK pathway is not involved in IGF-I-mediated survival.

In addition to PI3K and ERK1/2, the Src family of tyrosine kinases may also participate in IGF-I signaling (40–42). To examine their role in oligodendrocyte survival induced by IGF-I, cultures were pretreated with PP2, a potent and selective inhibitor of the Src kinases. As determined by the MTT assay, cell survival was not affected by low concentrations of PP2 (5 and 10 μm), whereas higher concentrations (20 and 50 μm) were required to significantly decrease cell survival in the presence of 100 ng/ml IGF-I (Fig. 3). Similarly, lower concentrations of PP2 (5, 7.5, and 10 μm) did not reverse the antiapoptotic effect of IGF-I on growth factor-deprived oligodendrocyte progenitors as assessed by the TUNEL assay (Fig. 4E and Table II). However, higher concentrations of PP2 (12.5 and 15 μm) caused a small but significant increase in the number of apoptotic cells (Table II).

Akt Activation by IGF-I Is Blocked by Inhibitors of PI3K and Src-like Kinases—The correlation between the initial increase in Akt activation and the rescue of oligodendrocyte progenitors from apoptosis by IGF-I was investigated further by applying agents that modulate the various pathways. Oligodendrocyte progenitors were pretreated with LY 294002, wortmannin, PD 98059, or PP2 for 30 min prior to treatment with 100 ng/ml IGF-I (16 h), and the activation of Akt, ERK1/2, and caspase-3 and the inactivation of GSK3β were determined by Western blotting. Both PI3K inhibitors (LY 294002 and wortmannin) blocked the IGF-I-promoted phosphorylation of Akt and GSK3β while increasing caspase-3 activation in a concentration-dependent manner (Fig. 5). At all concentrations (0.1, 0.5, and 1 μm), wortmannin increased caspase-3 activation beyond the levels obtained by growth factor withdrawal alone. Interestingly, in the presence of the PI3K inhibitors, the long-term inactivation of ERK1/2 by IGF-I (Fig. 2) was reversed, increasing ERK1/2 phosphorylation by 2–4-fold above control levels. In contrast, the PI3K inhibitors blocked the transient activation of ERK1/2 by IGF-I (Supplemental Fig. S1). These data indicate that PI3K is required for the activation of Akt, inactivation of GSK3β, and IGF-I suppression of oligodendrocyte progenitor cell death by blocking caspase-3 activation. In addition, PI3K is also required for the transient activation of ERK1/2.

PP2 prevented IGF-I-mediated Akt and GSK3β phosphorylation at both early (5 min and 4 h) (Supplemental Fig. S2) and late (16 h) (Fig. 6) time points and decreased ERK1/2 phosphorylation below control levels (Fig. 6). However, even at a concentration of 50 μm, PP2 did not reverse the IGF-I-mediated inhibition of caspase-3 activation following growth factor withdrawal when measured 16 h after treatment (Fig. 6).

As expected, PD 98059 decreased IGF-I-mediated ERK1/2 activation, but had no effect on phosphorylation of Akt and GSK3β. At the highest concentration (50 μm), PD 98059 exhibited a small tendency to reverse the inhibitory effect of IGF-I on caspase-3 activation (Fig. 7). These results suggest that PI3K lies upstream of Akt, GSK3β, and MEK/ERK in the IGF-I signaling cascade, with Src-like kinases acting upstream of Akt and ERK1/2 to regulate their activity.

Role of Src Tyrosine Kinases—To assess whether IGF-I causes tyrosine phosphorylation of Src kinases, oligodendrocyte progenitors were treated with the growth factor in the absence or presence of 10 μm PP2 or 100 nm wortmannin. Cell lysates were immunoprecipitated with anti-Src, anti-Lyn, or anti-Fyn antibody, followed by immunoblotting with anti-phosphotyrosine antibody. Examination of the three Src kinases expressed in oligodendrocyte progenitors showed that IGF-I induced the most significant increase in the tyrosine phosphorylation of Fyn (2–3-fold above basal levels), with a smaller increase in Lyn phosphorylation (1.5-fold). PP2 completely blocked the IGF-I-induced tyrosine phosphorylation of Fyn and Lyn (Fig. 8). In contrast, treatment with the PI3K inhibitor wortmannin did not block the tyrosine phosphorylation of Src family kinases, suggesting that Src tyrosine kinases are not downstream of PI3K. Subsequently, specific siRNAs were used to knockdown fyn to establish a link between Fyn and Akt activation. Western blotting showed that the levels of Fyn were reduced by fyn-1 and fyn-2 siRNAs in a concentration-dependent manner, with fyn-1 being the most efficient (Supplemental Figs. S3 and S4). Although fyn-1 and fyn-2 siRNAs significantly reduced IGF-I-stimulated Akt phosphorylation levels (∼10% of IGF-I alone) (Fig. 9), caspase-3 activation (data not shown) and the number of fragmented cells (Table III and Fig. S4) were not altered.

IGF-I-promoted Progenitor Survival Is Only Partially Mediated by Akt Activation—Because of the ability of PP2 (10 μm) to completely block the phosphorylation and activation of Akt without significantly affecting IGF-I-mediated cell survival or caspase-3 activation, the role of Akt was further examined in oligodendrocyte progenitors. Cells were transfected with a plasmid containing a kdAkt mutant (K179M-Akt-HA). After 48 h, the expression of kdAkt was confirmed by Western blot analysis using anti-HA antibody. The transfection experiments with kdAkt significantly reduced the IGF-I-mediated phosphorylation of GSK3β (p < 0.01) (Fig. 10A), a substrate of Akt. The expression of kdAkt alone caused a significant increase in caspase-3 activation compared with controls and was largely prevented by IGF-I treatment. However, the effect of IGF-I on inhibition of caspase-3 in cells expressing kdAkt was statistically significant compared with non-transfected controls (p < 0.05). Transient expression of kdAkt significantly decreased progenitor survival, but this decrease was partially reversed by IGF-I (Fig. 11). Because of the low efficiency of plasmid transfection in progenitors (∼30%) and the toxicity associated with the transfection reagent, the above results may have been affected to some extent by remaining endogenous Akt. To circumvent this problem, an adenoviral vector containing a dnAkt mutant (Akt-T308A/S473A) was used at a concentration (multiplicity of infection of 10) infecting anywhere from 75 to 95% of the cells. Overexpression of dnAkt prevented the IGF-I-induced phosphorylation of GSK3β (p < 0.001) (Fig. 10B) and partially released caspase-3 from blockade by IGF-I (p < 0.01). In addition, IGF-I blocked caspase-3 activation induced by dnAkt alone (Fig. 10B). Transient expression of the dnAkt mutant (multiplicity of infection of 10, 20, and 30) significantly decreased progenitor survival, but this decrease was significantly reversed by IGF-I (Fig. 11). Furthermore, immunofluorescence microscopy of kdAkt-transfected or dnAkt-infected cells coexpressed with green fluorescent protein adenovirus (Supplemental Fig. S5) showed that kdAkt and dnAkt alone increased the number of cells exhibiting nuclear condensation/fragmentation by 25 and 37%, respectively, compared with 20% in untransfected cells. These values were reduced by 70% when cells were treated with IGF-I, and the morphology appeared normal. Pretreatment of cells with 1L-6-hydroxymethyl-chiro-inositol 2-(R)-2-O-methyl-3-O-octadecylcarbonate (IC₅₀ ≅ 5.0 μm), an Akt inhibitor, significantly decreased Akt and GSK3β phosphorylation, but had no significant effect on IGF-I-induced caspase-3 inhibition (Fig. 12). Similar to the mutant forms of Akt, the Akt inhibitor decreased progenitor survival in a concentration-dependent manner in the absence of growth factors, but this decrease was partially reversed by IGF-I (Fig. 11).

DISCUSSION

Our results show that the activation of PI3K is essential for IGF-I-promoted oligodendrocyte progenitor survival. Furthermore, downstream activation of Akt-dependent and -independent pathways is involved in this survival effect. The existence of an Akt-independent pathway is revealed by the data showing that PP2, a specific Src-like kinase inhibitor, blocked Akt activation and GSK3β inactivation by IGF-I, but had no significant effects on IGF-I-induced caspase-3 inhibition and progenitor survival. This is further supported by the finding that IGF-I increased progenitor survival in cultures expressing dominant-negative mutants of Akt as well as in the presence of a specific Akt inhibitor.

The importance of IGF-I in promoting survival of oligodendrocyte progenitors was indicated by previous in vivo and in vitro studies (1, 2, 3, 4), although the underlying mechanism was not fully understood. In this study, we observed that IGF-I-promoted survival was correlated with the sustained activation of Akt, inactivation of GSK3β, and inhibition of caspase-3 activation, all of which are reversed by inhibition of PI3K. Therefore, similar to the role of PI3K in neuronal cell survival in the central nervous system (43), our results and previous studies (19, 20) demonstrate that this kinase is critical for survival of oligodendrocyte progenitors. Wortmannin was shown to decrease the number of viable progenitor cells (19, 20) and to block IGF-I-mediated survival (19). Inhibition of PI3K was also shown to block noradrenaline-stimulated ERK1/2 activation (32) and the survival effect mediated by neuregulin (44), cannabinoids (45), nerve growth factor (46), and C5b-9 terminal complement complex (39) in oligodendrocyte progenitors. A recent report showed that IGF-I prevents glutamate-induced apoptosis in a PI3K-dependent manner in immature oligodendrocytes (17, 18). Hence, PI3K is required for maintaining basal and growth factor-stimulated oligodendrocyte progenitor survival.

Src-like tyrosine kinases appear to communicate with many different receptor tyrosine kinases, including the insulin receptor and IGF-IR (47), and to affect cell survival, cell cycle progression, and other cellular functions. In this study, we found that Src-like tyrosine kinases act upstream of Akt and GSK3β in IGF-I signaling, as PP2 abolished the IGF-I-induced phosphorylation of Akt and GSK3β. It has been reported that, in addition to phosphorylation of Thr³⁰⁸ and Ser⁴⁷³, tyrosine phosphorylation is also essential for Akt activation (23, 24). Src, Fyn, and Lyn are all expressed in oligodendrocyte progenitors (Fig. 8) (48, 49), and evidence points to the involvement of Src-like kinases in oligodendrocyte differentiation and myelination. Thus, PP2 inhibits morphological differentiation of oligodendrocytes in culture (48) and reduces the number of mature oligodendrocytes (49). Among the tyrosine kinases, Fyn was found to stimulate transcription of the myelin basic protein gene (50), and oligodendrocytes from fyn-null mice do not mature morphologically in response to IGF-I (49). The forebrains of these mutants contain 50% less myelin at 28 days of age, and hypomyelination persists beyond 1 year of age (51). These data suggest that Fyn may participate in IGF-I-mediated oligodendrocyte development and myelination, yet its possible role in survival had not been considered before our study. Of the three Src-like tyrosine kinases, IGF-I had the most significant effect on tyrosine phosphorylation of Fyn, and this effect was abolished by PP2 treatment. The most interesting observation in this study is that, although 10 μm PP2 blocked IGF-I-stimulated phosphorylation of Akt and its downstream target, GSK3β, it did not increase caspase-3 activation (Fig. 6). Similarly, this concentration of PP2 had no significant effect on IGF-I-promoted cell survival (Fig. 3C and Table II), although higher concentrations of PP2 (12.5–50 μm) caused a significant decrease in survival, increasing the numbers of TUNEL-positive apoptotic cells. Furthermore, fyn knockdown by siRNA slightly reduced IGF-I-stimulated Akt phosphorylation (Fig. 9), but had no significant effect on the number of fragmented cells or activation of caspase-3 under basal or IGF-I-treated conditions (Table III). Because PP2 not only blocked Fyn tyrosine phosphorylation by IGF-I, but also reduced Src and Lyn phosphorylation below basal levels and because fyn siRNA decreased Akt activation by only 10%, the results suggest that Src and Lyn may also contribute to the effect on Akt phosphorylation.

The role of Src-like tyrosine kinases in cell survival is cell type-, growth factor-, and receptor-specific. For example, Src inhibition blocks Akt activation and attenuates neuronal survival induced by the glial cell-derived neurotrophic factor family ligand, but not cell survival induced by nerve growth factor (52). A recent report using siRNA for fyn, lyn, or src showed that fyn knockdown specifically inhibits the platelet-derived growth factor- or neuregulin-mediated oligodendrocyte survival only in the presence of the extracellular matrix, suggesting an integrin-mediated effect (53). Our results are in agreement with this recent study since Fyn activation does not appear to regulate IGF-mediated oligodendrocyte progenitor survival. Future studies will address whether Fyn regulation of Akt is required for other trophic functions of IGF-I on oligodendrocytes, including proliferation and differentiation. Another interesting aspect to examine is whether Src-like tyrosine kinases directly phosphorylate Akt (23) or whether they inhibit a phosphatase involved in the pathway such as PTEN. As shown recently, activated Src inhibits PTEN function, which is a lipid phosphatase that dephosphorylates phosphatidylinositol 3,4,5-trisphosphate at the 3′-position, leading to increased Akt phosphorylation (54).

Because Akt has been reported to be an important downstream effector of PI3K in mediating cell survival in many cellular systems, including oligodendrocytes, we sought to corroborate our finding that PP2 can block Akt activation by IGF-I without affecting survival. Oligodendrocyte progenitors were transduced with kdAkt and dnAkt mutants as well as treated with an Akt inhibitor. This compound blocks Akt activation by competing with phosphatidylinositol 3,4,5-trisphosphate for binding to the pleckstrin homology domain of Akt (55). We observed that the defective Akt mutants and the inhibitor reduced IGF-I-stimulated Akt activation and GSK3β inactivation and partially reversed the effect of IGF-I on caspase-3 activation during growth factor withdrawal. However, IGF-I was still able to protect the progenitors from cell death in the presence of the Akt mutants or inhibitor.

Another interesting observation is that the Akt mutants and Akt inhibitor alone increased caspase-3 activation and decreased oligodendrocyte progenitor survival under conditions of growth factor deprivation. As both mature oligodendrocytes and progenitors secrete a number of growth factors, they may provide trophic support that affects the development and maintenance of nearby oligodendrocytes (56). Among these growth factors, nerve growth factor (46) and neuregulin (44) have been reported to promote oligodendrocyte survival through Akt alone because the dnAkt adenovirus fully blocks the action of these growth factors on oligodendrocyte survival, which is not the case with IGF-I. This Akt-independent effect of IGF-I is also independent of the Src family of kinases and downstream of PI3K since LY 294002 and wortmannin fully reversed the ability of IGF-I to inhibit caspase-3 activation and to promote cell survival. The fundamental role of PI3K in mediating survival by multiple growth factors has been shown in many studies. King et al. (57) demonstrated that caspase-3 is activated by inhibition of PI3K and reduced by stimulation of PI3K. This Akt-independent pathway may compensate for the loss of Akt activation when PP2 blocks Src-like tyrosine kinases. Apart from Akt, PDK1, the downstream effector of PI3K, phosphorylates and activates a number of kinases, including protein kinase Cζ, p70^S6K, RSKs and serum- and glucocorticoid-inducible kinases (58), all of which are involved in cell growth, proliferation, and survival. For example, PDK1 induces the phosphorylation of RSK1, RSK2, and RSK3 without the involvement of ERK, leading to their partial activation (59). Activated RSKs can phosphorylate and inactivate Bad, an effector of apoptosis. In addition, RSKs phosphorylate and activate the cAMP response element-binding protein, a transcription factor that mediates cell survival through upregulation of the prosurvival gene encoding Bcl-2 (27–29). Our results show that the MEK1 inhibitor PD 98059 had no significant effects on Akt activation and GSK3β inactivation or IGF-I-induced cell survival and prevention of apoptosis. In addition, long-term treatment with IGF-I (≥4 h) caused a reduction in ERK1/2 phosphorylation (Fig. 2), which was reversed by pretreating cells with LY 294002 or wortmannin. These data indicate that there is cross-talk between the PI3K and MEK/ERK pathways, as suggested by previous observations (59–62). Further evidence indicates that PI3K activation may down-regulate the Ras/Raf/MEK/ERK pathway at the Raf and Akt levels (59, 61). Together, these studies suggest that PDK1 substrates, in addition to Akt, may be involved in a pathway linking activation of PI3K with progenitor survival.

Among the proteins involved in IGF-I signaling, 14-3-3 has been shown to interact with IGF-IR and to activate and translocate Raf-1 into mitochondria (63–65). Targeting Raf-1 to mitochondria causes inactivation of Bad and subsequent inhibition of apoptosis (66). Moreover, phospholipase C is also required for IGF-I-mediated cell survival in cells deprived of adhesion to the extracellular matrix (67). It would therefore be interesting to determine whether these alternative pathways operate in IGF-I-mediated progenitor survival.

In summary, our data indicate that 1) PI3K is essential for cell survival promoted by IGF-I; 2) Src-like tyrosine kinases participate in IGF-I-induced Akt activation; and 3) Akt activation is involved, but may not be the only key downstream intermediate within the PI3K pathway to mediate IGF-I-induced oligodendrocyte progenitor survival and inhibition of caspase-3 activation. It thus appears likely that an unidentified effector of PI3K is required for conferring the complete protective effect of IGF-I on oligodendrocyte progenitors.